article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
the tama project held observation experiments for detecting gravitational wave ( gw ) signals since 1999 . by the beginning of 2004 ,
3000 hours of data in total was accumulated through the nine observation experiments .
after the last observation experiment in 2004 , tama detector is being upgraded to reduce the low frequency noises . to reduce the seismic noise , new isolation system called tama seismic attenuation system ( sas ) @xcite
is being installed . to isolate horizontal motions ,
an inverted pendulum is employed . for vertical motion ,
double stage monolithic geometric anti - spring ( mgas ) filters are used .
the test mass mirror is suspended by a double pendulum as a final stage of the system .
the sas installation was started in september , 2005 . in the summer of 2006 , a 300-m fabry - perot cavity formed by the mirrors suspended by sass was operated . since the other arm was still formed by the old suspension system , we could directly compare the performance of the vibration isolation systems . in this locked fabry - perot configuration ,
the old suspension arm was controlled by the laser frequency , and the sas arm was controlled by the coil - magnet actuators located at test mass mirror .
figure [ tama - fig1 ] shows the comparison of cavity length and angular fluctuations with sass and the old suspension systems .
the improvement of about 24db was observed above 2 and 3 hz for the length and pitch motion , respectively .
the actuator response of the test mass for sas was reduced by a factor of 3 so as to reduce the actuator associated noises .
therefore , ability of the lock acquisition and drift level of the interferometer with sass should have been confirmed . to ensure the ability of the lock acquisition ,
a digital servo system was employed for the test mass length control .
this system is driven by a dsp processor with sampling frequency of 200 khz , and the unity gain frequency of about 800hz was achieved .
this system utilizes two techniques for the error signal operation .
the first is gating of the error signal ; the error signal is forced to be zero when the transmitted light level of the cavity is below a preset threshold .
this cleans up undesirable signals caused by resonances of modulation sidebands and higher order spatial modes of the beam .
the other technique is normalization of the error signal with regard to the transmitted light .
this widened linear range of the error signal by a factor of 3 .
the locked fabry - perot interferometer was locked with help of the digital servo system , and operated continuously for the 6.5 hours .
since the detected length fluctuation is the addition of that of the sas arm and the old suspension arm , this means that the actuator range was enough to compensate the drift motion of the sass .
this indicates that tama300 would be operable even with both two arms are formed with sass .
actual improvements of length and angular fluctuations at 100 hz region should be confirmed by the locked fabry - perot configuration formed by four sass . now all of the four test mass mirrors are suspended by sass .
we are currently optimizing the performance of the third and fourth sass .
we continue r&d studies of sas and tama300 operations .
next data taking in the summer of 2007 is planned .
cryogenic laser interferometer observatory ( clio ) is located in kamioka mine of japan as an underground site .
purpose of clio project is technical demonstration of key features of lcgt .
lcgt is planed to locate at kamioka underground site for low seismic noise level , to adopt cryogenic sapphire mirrors for low thermal noise level and to have arms of 3 km long . except for the arm length
, clio has same features of lcgt .
therefore , the detector can demonstrate them as a prototype of lcgt .
construction of clio has been started in 2002 .
all of vacuum pipes , cryostats and cryocoolers were installed by the june , 2005 @xcite .
after the cooling tests of the system , the first operation of the cryogenic interferometer has been demonstrated on 18 february , 2006 .
the details of the interferometer , cryogenic systems and seismic environments have been published in ref .
input power to each cavity was 200 m w. the measured finesse and reflectance of the cavity were about 3100 and 0.01 , respectively .
the accumulated power inside the cavity was estimated to be 200 w. during the lock of 50 minutes , the mirrors kept its temperature around 20k .
after the cryogenic operation was established , noise investigations at room temperature are in progress .
figure [ clio - fig2 ] shows the best noise spectrum on 13 december 2006 . at all of frequency regions ,
the differences from the target sensitivity at room temperature are about a factor of 4 .
the observable ranges with snr=10 for inspiral gw signals can be estimated by the noise spectra as shown in fig .
[ clio - fig3 ] . for
neutron star binaries , clio and tama can observe the event within 49k pc and 73k pc , respectively .
we can say that the two detectors have almost same sensitivity . at over 10 solar mass regions , clio keeps good sensitivities due to its low seismic noises .
it is the greatest benefit of underground site .
several observation experiments at room temperature have been done .
once the displacement noise reaches at thermal noise level , its improvement by cooling should be demonstrated .
tama and clio projects are supported in part by a grant - in - aid for scientific research on priority areas ( 415 ) of the ministry of education , culture , sports , science and technology and grant - in - aid for scientific research ( a ) , no.18204021 , 2006 of japan society for the promotion of science .
developments of tama sas were supported by the advanced technology center of the national astronomical observatory of japan and the u.s .
national science foundation under cooperative agreement no . phy-0107417 .
99 s. mrka et al . , class . quantum grav . * 19 * , 2002 , 1605 - 1614 a. takamori et al . , class . quantum grav . * 19 * , 2002 , 1615 - 1621 m. ohashi et al . , class . quantum grav . * 20 * , 2003 , s599-s607 s. miyoki et al . , class .
quantum grav .
* 21 * , 2004 , s1173-s1181 k. yamamoto et al . , j. phys .
32 , 2006 , 418 - 423 t. uchiyama et al .
, j. phys .
32 , 2006 , 259 - 264 s. miyoki et al . ,
quantum grav .
* 23 * , s231-s237 | current status of tama and clio detectors in japan is reported in this article .
these two interferometric gravitational - wave detectors are being developed for the large cryogenic gravitational wave telescope ( lcgt ) which is a future plan for detecting gravitational wave signals at least once per year .
tama300 is being upgraded to improve the sensitivity in low frequency region after the last observation experiment in 2004 . to reduce the seismic noises
, we are installing new seismic isolation system , which is called tama seismic attenuation system , for the four test masses .
we confirmed stable mass locks of a cavity and improvements of length and angular fluctuations by using two sass .
we are currently optimizing the performance of the third and fourth sass .
we continue tama300 operation and r&d studies for lcgt .
next data taking in the summer of 2007 is planned .
clio is a 100-m baseline length prototype detector for lcgt to investigate interferometer performance in cryogenic condition .
the key features of clio are that it locates kamioka underground site for low seismic noise level , and adopts cryogenic sapphire mirrors for low thermal noise level .
the first operation of the cryogenic interferometer was successfully demonstrated in february of 2006 .
current sensitivity at room temperature is close to the target sensitivity within a factor of 4 .
several observation experiments at room temperature have been done .
once the displacement noise reaches at thermal noise level of room temperature , its improvement by cooling test mass mirrors should be demonstrated .
current status of tama and clio detectors in japan is reported in this article .
these two interferometric gravitational wave detectors are being developed for lcgt which is a future plan for detecting gravitational wave signals at least once per year . |
the study of systems at the anderson metal - insulator transition ( mit ) has been a subject of intensive research activity for several decades .
@xcite in particular , much interest has been focused on the scattering properties of critical systems by analysing the probability distribution functions of the resonance widths @xmath5 and wigner delay times @xmath6 , @xcite as well as the transmission or dimensionless conductance @xmath7 .
@xcite the distribution functions of @xmath5 and @xmath6 have been shown to be related to the properties of the corresponding closed systems , i.e. , the fractality of the eigenstates and the critical features of the mit .
on the other hand , at the mit , the distribution of conductances @xmath8 has been found to be _ universal _ ,
i.e. , size independent , but dependent on the adopted model , dimensionality , symmetry , and even boundary conditions of the system .
@xmath8 has been studied for systems in two and more dimensions with a large number of attached single - channel leads .
@xcite in fact , concerning the conductance of one - dimensional ( 1d ) systems and its statistical distribution , the regime of small number of leads has been left almost unexplored .
@xcite in the present work we study numerically several statistical properties of the scattering matrix and the electronic transport across disordered systems in one and three dimensions described by the power - law banded random matrix ( pbrm ) model at criticality and the three - dimensional ( 3d ) anderson model at the mit , respectively .
we stress that we concentrate on the case of a small number of attached leads each of them supporting one open channel .
the organization of this paper is as follows . in the next subsection
we define the pbrm model , the 3d anderson model , as well as the scattering setup .
we also define the scattering quantities under investigation and provide the corresponding analytical predictions from random scattering - matrix theory ( rmt ) for systems with time - reversal symmetry .
these analytical results will be used as a reference along the paper . in section
ii we analyse the average scattering matrix elements , the conductance probability distribution , the variance of the conductance , and the shot noise power for the pbrm model as a function of its effective bandwidth @xmath0 . in section iii
we compare the results of the pbrm model at criticality with the scattering and transport properties of the 3d anderson model .
finally , section iv is left for conclusions . as we have mentioned above , in the present study
we adopt two models , namely , the pbrm model at criticality and the 3d anderson model at the mit .
the pbrm model @xcite describes 1d samples of length @xmath9 with random long - range hopping .
this model is represented by @xmath10 ( @xmath11 ) real symmetric matrices whose elements are statistically independent random variables drawn from a normal distribution with zero mean , @xmath12 , and a variance decaying as a power law @xmath13 , where @xmath0 and @xmath14 are parameters .
there are two prescriptions for the variance of the pbrm model : the so - called non - periodic , @xmath15 where the 1d sample is in a _ line _ geometry ; and the periodic , @xmath16^{2\alpha } } \ , \label{pbrmp}\ ] ] where the sample is in a _ ring _ geometry .
field - theoretical considerations @xcite and detailed numerical investigations @xcite have verified that the pbrm model undergoes a transition at @xmath17 from localized states for @xmath18 to delocalized states for @xmath19 .
this transition shows all the key features of the anderson mit , including multifractality of eigenfunctions and non - trivial spectral statistics at the critical point .
thus the pbrm model possesses a line of critical points @xmath20 .
we set @xmath17 in our study , i.e. , we work with the pbrm model at criticality . the 3d anderson model with diagonal disorder is described by the tight - binding hamiltonian ( tbh ) @xmath21 labels all the @xmath22 sites of a cubic lattice with linear size @xmath9 , while the second sum is taken over all nearest - neighbour pairs @xmath23 on the lattice .
the on - site potentials @xmath24 for @xmath25 are independent random variables .
when @xmath24 are gaussian distributed , with zero mean and variance @xmath26 , the mit at energy @xmath27 occurs for @xmath28 . see refs .
[ ] . then , for @xmath29 ( @xmath30 ) the system is in the metallic ( insulating ) regime .
we set @xmath31 in our study .
we open the isolated samples , defined above by the pbrm model and the 3d anderson model , by attaching @xmath32 semi - infinite single channel leads .
each lead is described by the 1d semi - infinite tbh @xmath33 using standard methods @xcite one can write the scattering matrix ( @xmath34-matrix ) in the form @xcite @xmath35 where @xmath36 is the @xmath37 unit matrix , @xmath38 is the wave vector supported in the leads , and @xmath39 is an effective non - hermitian hamiltonian given by @xmath40 here , @xmath41 is an @xmath42 matrix that specifies the positions of the attached leads to the sample .
its elements are equal to zero or @xmath43 , where @xmath43 is the coupling strength .
moreover , assuming that the wave vector @xmath44 do not change significantly in the centre of the band , we set @xmath45 and neglect the energy dependence of @xmath46 and @xmath34 . notice from eqs .
( [ pbrmnp]-[pbrmp ] ) that in the limit @xmath3 the pbrm model reproduces the gaussian orthogonal ensemble .
therefore , in that limit we expect the statistics of the scattering matrix , eq . ( [ smatrix ] ) , to be determined by the circular orthogonal ensemble ( coe ) which is the appropriate ensemble for systems with time reversal symmetry .
thus , below , we provide the statistical results for the average @xmath34-matrix and the transport quantities to be analysed in the following sections , assuming the orthogonal symmetry .
in all cases , we also assume the absence of direct processes , i.e. , @xmath47 .
we start with the average of the @xmath34-matrix elements .
it is known that @xmath48 where @xmath49 means ensemble average over the coe . within a scattering approach to the electronic transport ,
once the scattering matrix is known one can compute the dimensionless conductance @xmath50 and its distribution @xmath8 . for @xmath51 ,
i.e. , considering two single - channels leads attached to the sample , @xmath8 is given by @xmath52 while for @xmath53 , @xmath54 for arbitrary @xmath55 , the predictions for the average value of @xmath7 and its variance are @xmath56 and @xmath57 respectively .
for the derivation of the expressions above see for example ref .
another transport quantity of interest is the shot noise power @xmath58 , which as a function of @xmath55 reads @xcite @xmath59 in the following sections we focus on @xmath60 , @xmath61 , @xmath62 , and @xmath63 for the pbrm model and the 3d anderson model , both at the mit . in all cases we set the coupling strength @xmath43 such that @xmath64 in order to compare our results , in the proper limits , with the rmt predictions introduced above .
we attach the @xmath32 leads to the first @xmath32 sites of the 1d sample described by the pbrm model .
that is , in the non - periodic version of the pbrm model , eq .
( [ pbrmnp ] ) , we attach the leads at the _ boundary _ of the system .
see fig .
[ setup](a ) . while in the periodic version , eq . ( [ pbrmp ] ) , we attach them to the _
bulk_. see fig . [ setup](b ) . in the latter case ,
finite size effects are considerably reduced .
however , the quantities we analyse below are @xmath9-independent once @xmath9 is much larger than the number of attached leads for both versions of the pbrm model .
we point out that our setup is significantly different from the one used in refs .
[ ] , where the conductance has also been studied using the pbrm model . there , for example , the leads in the @xmath51 case are attached to sites which are separated a distance of @xmath65 and @xmath9 in the periodic and non - periodic versions of the pbrm model , respectively .
in such situation the scattering quantities are strongly @xmath9-dependent .
leads , shown as blue ( gray ) lines , are attached to @xmath32 sites ( black dots ) of a 1d sample described by the ( a ) non - periodic and ( b ) periodic versions of the pbrm model .
the case @xmath53 with @xmath66 is shown as example.,width=188 ] symbols : average @xmath34-matrix elements @xmath67 and @xmath68 for the periodic [ non - periodic ] pbrm model at criticality as a function of ( a ) @xmath0 and ( b ) @xmath69 for @xmath51 .
the blue ( gray ) dot - dashed lines correspond to 2/3 and 1/3 ; the rmt prediction for @xmath67 and @xmath68 , respectively , given by eq .
( [ coes ] ) .
the black [ red ( gray ) ] dashed lines in ( a ) are eqs .
( [ s11b ] ) and ( [ s12b ] ) with @xmath70 [ @xmath71 . the black [ red ( gray ) ] dashed lines in ( b ) are eqs .
( [ s11d ] ) and ( [ s12d ] ) with @xmath70 [ @xmath71 .
error bars are not shown since they are much smaller than symbol size.,width=302 ] first we consider the case @xmath51 , where the @xmath34-matrix is a @xmath72 matrix . in fig .
[ fig1](a ) we plot the ensemble average of the elements @xmath73 and @xmath74 as a function of the bandwidth parameter @xmath0 , @xmath75 and @xmath76 , for the periodic ( black symbols ) and the non - periodic ( red symbols ) pbrm model .
we concentrate on these two matrix elements since the other two elements give no additional information : @xmath77 and @xmath78 .
notice a strong @xmath0-dependence of the average @xmath34-matrix elements driving them from a localized - like regime [ @xmath79 and @xmath80 ; i.e. the average conductance is close to zero ] for @xmath1 , to a delocalized - like or ballistic - like regime [ @xmath81 and @xmath82 ; i.e. , rmt results are already recovered ] for @xmath83 .
moreover , we have found that @xmath67 and @xmath68 , as a function of the bandwidth @xmath0 , are well described by @xmath84 \ , \end{aligned}\ ] ] where @xmath85 is a fitting parameter .
( [ s11b ] ) is a consequence of the unitarity of the scattering matrix , @xmath86 , while the factor 1/3 in eq .
( [ s12b ] ) comes from eq .
( [ coes ] ) with @xmath51 . in fig .
[ fig1](a ) we plot eqs .
( [ s11b ] ) and ( [ s12b ] ) ( dashed lines ) and compare them with the corresponding results from the pbrm model ( symbols ) in the periodic and non - periodic setups . in the same figure eq .
( [ coes ] ) is also plotted ( dot - dashed lines ) . on the other hand
, it is well known that in systems at the disorder driven mit both the energy spectra and the eigenstates exhibit multifractal characteristics .
@xcite the pbrm model is characterized by the effective bandwidth @xmath0 that drives the system from strong ( @xmath87 ) to weak ( @xmath3 ) multifractality .
multifractality can be quantified by the generalized dimensions @xmath88 which describe the fluctuations of the eigenfunctions .
the multifractal dimensions @xmath88 of the @xmath89-th eigenfunction @xmath90 ( given as a linear combination of the basis states in a system with linear size @xmath9 , @xmath91 ) are defined through the so - called inverse participation numbers , @xmath92 , by the scaling @xcite @xmath93 where @xmath94 is the typical value of @xmath95 . however , among all dimensions , the correlation dimension @xmath69 plays a prominent role . @xcite as @xmath0 transits from zero to infinity , @xmath69 takes values from zero to one . using numerical diagonalization of matrices with sizes @xmath96 , 256 , 512 , and 1024
, we extracted @xmath69 by the use of eq .
( [ dq ] ) . for each system size @xmath9
, we used a number of disorder realizations in order to have at least @xmath97 data for its analysis . for each disorder realization we used @xmath98 of the states at the centre of the spectral band ( where the density of states is almost constant ) in order to avoid boundary effects .
we found good agreement between the numerically obtained @xmath69 and the analytical estimation @xcite @xmath99 moreover , in ref .
[ ] the following phenomenological analytical expression for @xmath69 as a function of @xmath0 was proposed : @xmath100 where @xmath101 is a fitting parameter .
see also ref .
( [ d2ofb ] ) describes well our numerical results for @xmath69 with @xmath102 and @xmath103 for the periodic and non - periodic versions of the pbrm model , respectively .
curves : probability distribution @xmath104 for the periodic [ non - periodic ] pbrm model at criticality for several values of @xmath105 ( @xmath106 , 0.02 , 0.04 , 0.1 , 0.2 , and 0.4 from left to right ) in the case @xmath51 .
( b ) black [ red ( gray ) ] symbols : @xmath107 as a function of @xmath0 for the periodic [ non - periodic ] pbrm model .
the blue ( gray ) dashed line is the best fit of the data to the logarithmic function @xmath108 with @xmath109 .
( c ) @xmath104 for @xmath105 scaled to @xmath110 .
the blue ( gray ) dashed line proportional to @xmath111 is plotted to guide the eye.,width=302 ] curves : conductance probability distribution @xmath8 for the periodic [ non - periodic ] pbrm model at criticality for some _ large _ values of @xmath0 in the case @xmath51 .
blue ( gray ) dashed lines are @xmath112 ; the rmt prediction for @xmath8 given by eq .
( [ goen1]).,width=302 ] curves : probability distribution @xmath104 for the periodic [ non - periodic ] pbrm model at criticality for several values of @xmath105 ( @xmath106 , 0.02 , 0.04 , 0.1 , and 0.2 from left to right ) in the case @xmath53 .
( b ) black [ red ( gray ) ] symbols : @xmath107 as a function of @xmath0 for the periodic [ non - periodic ] pbrm model .
the blue ( gray ) dashed line is the best fit of the data to the logarithmic function @xmath108 with @xmath113 .
( c ) @xmath104 for @xmath105 scaled to @xmath110 .
the blue ( gray ) dashed line proportional to @xmath114 is plotted to guide the eye.,width=302 ] curves : conductance probability distribution @xmath8 for the periodic [ non - periodic ] pbrm model at criticality for some _ large _ values of @xmath0 in the case @xmath53 .
blue ( gray ) dashed lines are @xmath112 ; the rmt prediction for @xmath8 given by eq .
( [ goen2]).,width=302 ] as a function of @xmath55 for the non - periodic pbrm model at criticality for several values of @xmath0 .
@xmath61 for the 3d anderson model at criticality ( 3dam ) is also shown .
( b ) @xmath115 as a function of @xmath116 for the periodic pbrm model at criticality for @xmath117 $ ] .
insert : @xmath85 versus @xmath55 .
@xmath85 is obtained from the fitting of eq .
( [ xb ] ) to the @xmath61 vs. @xmath0 data .
thick full lines correspond to @xmath118 .
blue ( gray ) dashed lines are ( a ) the rmt prediction for @xmath61 given by eq .
( [ goeavt ] ) ; and ( b ) one.,width=302 ] as a function of @xmath55 for the non - periodic pbrm model at criticality for several values of @xmath0 .
@xmath62 for the 3d anderson model at criticality ( 3dam ) is also shown .
( b ) @xmath119 as a function of @xmath116 for the periodic pbrm model at criticality for @xmath117 $ ] .
insert : @xmath85 versus @xmath55 .
@xmath85 is obtained from the fitting of eq .
( [ xb ] ) to the @xmath62 vs. @xmath0 data .
thick full lines correspond to @xmath120 .
blue ( gray ) dashed lines are ( a ) the rmt prediction for @xmath62 given by eq .
( [ goevart ] ) ; and ( b ) one.,width=302 ] as a function of @xmath55 for the non - periodic pbrm model at criticality for several values of @xmath0 .
@xmath63 for the 3d anderson model at criticality ( 3dam ) is also shown .
( b ) @xmath121 as a function of @xmath116 for the periodic pbrm model at criticality for @xmath117 $ ] .
insert : @xmath85 versus @xmath55 .
@xmath85 is obtained from the fitting of eq .
( [ xb ] ) to the @xmath63 vs. @xmath0 data .
thick full lines correspond to @xmath122 .
blue ( gray ) dashed lines are ( a ) the rmt prediction for @xmath63 given by eq .
( [ goep ] ) ; and ( b ) one.,width=302 ] then , by substituting eq .
( [ d2ofb ] ) into eqs .
( [ s11b ] ) and ( [ s12b ] ) we obtain the following expressions for the averages of @xmath73 and @xmath74 as a function of @xmath69 : @xmath123 in fig . [ fig1](b ) we compare eqs .
( [ s11d ] ) and ( [ s12d ] ) ( dashed lines ) with the ensemble - average numerical results at different values of @xmath69 .
the agreement is excellent .
in addition , it is interesting to note that @xmath124 which might be relevant for systems at the mit where @xmath69 can be tuned .
finally we want to remark that concerning @xmath60 for the pbrm model , the rmt limit , expected for @xmath3 , is already recovered for @xmath125 .
see fig .
[ fig1](a ) .
.@xmath107 for the periodic pbrm model at criticality for some values of @xmath0 and @xmath9 in the case @xmath51 . [ cols="<,<,<,<",options="header " , ] we also compute @xmath61 , @xmath62 , and @xmath63 for the 3d anderson model and plot them in figs . [ fig7](a ) , [ fig8](a ) , and [ fig9](a ) , respectively ( red curves labeled as 3dam ) .
we observe that for the 3d anderson model @xmath61 , @xmath62 , and @xmath63 behave as the corresponding quantities for the pbrm model with @xmath0 close to 0.2 .
we want to stress that our results for the 3d anderson model at criticality do not seem to depend on the on - site potential distribution .
here we used gaussian distributed potentials .
however , we observe practically the same results by the use of box distributed on - site potentials ( not shown here ) .
we study the scattering and transport properties of the power - law banded random matrix ( pbrm ) model and of the three - dimensional ( 3d ) anderson model , both at criticality .
we observed a smooth crossover from localized- to delocalized - like ( ballistic - like ) behavior in the scattering properties of the pbrm model by varying @xmath0 from small ( @xmath1 ) to large ( @xmath126 ) values .
for this crossover region we proposed heuristic analytical expressions for @xmath127 and @xmath128 . for small @xmath0 , we have shown that @xmath104 is scale invariant with the typical value of @xmath7 , @xmath129 , as scaling factor .
we realized that rmt results , expected in the limit @xmath3 , are already recovered for relatively small values of the bandwidth : @xmath130 .
however , this fact is closely related to the small number of attached leads we used in this work : the larger the number of leads the larger the value of @xmath0 needed to approach rmt behavior .
our conclusions are valid for leads attached to the bulk of the system as well as for leads attached at the boundary . in this work
we assumed that time - reversal symmetry ( @xmath131 ) is present in our disordered systems .
moreover , the case of broken time - reversal symmetry ( @xmath132 ) has been preliminarily studied in ref . [ ] where similar conclusions to this work were made .
we have also shown that the scattering properties of the 3d anderson model are similar to those for the pbrm model with @xmath4 $ ] .
this result is in agreement with previous studies @xcite were it was shown that several critical quantities related to the spectrum and eigenstates of the pbrm model with @xmath133 are practically the same as for the 3d anderson model .
this makes the pbrm model an excellent candidate to explore the properties of the 3d anderson model at criticality at a low computational cost .
we want to stress that for both models at criticality , the pbrm model and the 3d anderson model , we found the universal behavior @xmath134 for @xmath135 , that we tested here for @xmath51 and @xmath53 .
moreover by the use of the pbrm model with @xmath132 ( not shown here ) we have already verified that the more general expression @xmath136 , derived from @xcite @xmath137 with @xmath138 , holds . finally , we recall that we concentrate here on the case of a small number of single - channel attached leads ( up to ten ) .
the study of the scattering and transport properties of the pbrm and the 3d anderson models in the regime of @xmath139 will be the subject of a forthcoming contribution .
some properties of the transmission for the pbrm model were already studied for @xmath139 . ]
we thank one of the referees for valuable comments .
this work was partially supported by the hungarian - mexican intergovernmental s & t cooperation program under grants mx-16/2007 ( nkth ) and i0110/127/08 ( conacyt ) .
we also acknowledge support form conacyt mexico grant cb-2006 - 01 - 60879 ; the european social fund ; the micinn ( spain ) under project no .
fis 2009 - 16450 and through the ramn y cajal program ; the alexander von humboldt foundation ; and the hungarian research fund ( otka ) grants k73361 and k75529 .
p. w. anderson , phys . rev . * 109 * , 1492 ( 1958 ) ; a. mackinnon and b. kramer , rep .
. phys . * 56 * , 1469 ( 1993 ) ; e. abrahams , p. w. anderson , d. c. licciardello , and t. v. ramakrishnan , phys .
* 42 * , 673 ( 1979 ) ; l. p. gorkov , a. i. larkin , and d. e. khmelnitskii , jetp lett . * 30 * 228 ( 1979 ) .
a. cohen , y. roth , and b. shapiro , phys .
b * 38 * , 12125 ( 1988 ) ; c. m. soukoulis , x. s. wang , q. m. li , and m. m. sigalas , phys
lett . * 82 * , 668 ( 1999 ) ; k. slevin and t. ohtsuki , ibid . *
82 * , 382 ( 1999 ) . c. mahaux and h. a weidenmller , _ shell model approach in nuclear reactions _ , ( north - holland , amsterdam,1969 ) ; j. j. m. verbaarschot , h. a. weidenmller , and m. r. zirnbauer , phys . rep . * 129 * , 367 ( 1985 ) ; i. rotter , rep .
54 * , 635 ( 1991 ) . | we study numerically scattering and transport statistical properties of the one - dimensional anderson model at the metal - insulator transition described by the power - law banded random matrix ( pbrm ) model at criticality . within a scattering approach to electronic transport ,
we concentrate on the case of a small number of single - channel attached leads . we observe a smooth crossover from localized to delocalized behavior in the average scattering matrix elements , the conductance probability distribution , the variance of the conductance , and the shot noise power by varying @xmath0 ( the effective bandwidth of the pbrm model ) from small ( @xmath1 ) to large ( @xmath2 ) values .
we contrast our results with analytic random matrix theory predictions which are expected to be recovered in the limit @xmath3 .
we also compare our results for the pbrm model with those for the three - dimensional ( 3d ) anderson model at criticality , finding that the pbrm model with @xmath4 $ ] reproduces well the scattering and transport properties of the 3d anderson model . |
planets are formed in a disk around a star called a protoplanetary disk , which is composed of gas and dust .
terrestrial planets are formed mainly from the dust component .
their formation process can be divided into three stages .
the first stage is the formation of a large number of kilometer sized bodies called planetesimals by accretion among dust particles ( e.g. , * ? ? ?
* ; * ? ? ?
* ) . in the second stage ,
these planetesimals collide to produce a few tens of mars - sized objects called protoplanets ( e.g. , * ? ? ?
* ; * ? ? ?
the final stage is the formation of terrestrial planets from protoplanets ( e.g. , * ? ? ?
* ; * ? ? ?
the collisions among protoplanets are referred to as giant impacts , and thus this final stage is known as the giant impact stage .
giant impacts have a large influence on the various features such as the number of terrestrial planets formed , their mass and spin state ( e.g. , * ? ? ?
* ; * ? ? ?
giant impacts are highly energetic events , and are responsible for the creation of large satellites , like the moon ( e.g. , * ? ? ?
* ) and planets with extremely large cores such as mercury ( e.g. , * ? ? ?
moreover , giant impacts are closely related to the thermal state such as a magma ocean ( e.g. , * ? ? ?
* ) , and the origins of the terrestrial planet atmospheres @xcite . a large number of simulations of giant impacts have been devoted to the specific giant impact events related to the origin of the moon or mercury .
however , since it is recently believed that multiple giant impacts are common during the last stage of terrestrial planet formation , several studies @xcite have investigated the giant impact simulations under various impact parameters .
@xcite were the first to show that collisions of protoplanets during the giant impact stage are not always merging events , that is , two colliding protoplanets sometimes move apart after the collision .
they called such a collision a hit - and - run collision . except for @xcite , all the previous studies on the orbital and accretional evolution of protoplanets during the giant impact stage have been based on the assumption of perfect accretion , where two colliding protoplanets always merge .
however , the hit - and - run collisions demonstrated by @xcite may have an important influence on many of the physical characteristics of terrestrial planets . in order to investigate the effects of such imperfect accretion of protoplanets on terrestrial planet formation
, the merging criteria during protoplanet collisions must be clarified .
@xcite performed 48 simulations of the collisions between same - sized protoplanets with masses of @xmath7 , where @xmath8 is the earth mass .
they found that hit - and - run collisions occurred when @xmath9 for @xmath10 , and @xmath11 for @xmath12 or @xmath13 , where @xmath3 is the impact velocity , @xmath6 is the two - body escape velocity ( equation [ [ eq : esc ] ] ) , and @xmath2 is the impact angle . since they varied @xmath3 in steps of @xmath14 for low - velocity collisions , the transition between the merging and hit - and - run collisions was estimated to be @xmath15 for @xmath10 , and @xmath16 for @xmath12 or @xmath13 .
subsequently , @xcite investigated collisions between different - sized protoplanets with mass ratios of 1:2 and 1:10 .
their results were presented in @xcite .
the transition was able to be estimated as follows . in the case of a 1:2 mass ratio ,
the transition occurs at @xmath15 for @xmath10 and @xmath16 for @xmath12 or @xmath13 , whereas in a case of 1:10 mass ratio , @xmath17 for @xmath10 , @xmath18 for @xmath12 , and @xmath19 for @xmath20 .
@xcite and @xcite investigated collisions of the rocky and icy super - earths ( up to @xmath21 ) , respectively .
they focused on the stripping of the rocky or icy mantle resulting from a high - velocity impact .
although they did not perform the various simulations with low - impact velocities , we were able to find that the transition occurs at roughly @xmath22 from their figures .
as described above , the merging criteria has already been roughly determined for certain discrete values of the impact parameters . however , in order to carry out n - body orbital calculations with the merging criteria for the giant impact stage , a simple formula describing the dependence of the merging criteria on the impact parameters is required .
to achieve this , it is necessary to determine the merging criteria over a wide range of impact parameters . in the present study
, we performed more than 1000 simulations of giant impacts for various impact parameter sets using the smoothed particle hydrodynamic ( sph ) method in order to formulate the merging criteria . in section [ sec : method ] , we present the sph code and initial conditions used in our giant impact simulations . in section [ sec :
outcomes ] , we show the collision outcomes , and investigate the transition between merging and hit - and - run collisions .
we also perform a resolution test on the simulations . in section [ sec : criteria ] , we derive the merging criteria as a function of the impact parameters , and compare the results with those of previous studies . using the derived criteria , we then discuss the merging probability of protoplanets during the giant impact stage in section [ sec : probability ] .
in order to perform impact simulations for protoplanets , we used the sph method ( e.g. , * ? ? ? * ) , which is a flexible lagrangian method of solving hydrodynamic equations , and has been widely used in previous giant impact simulations .
the sph method can easily deal with large deformations and shock waves .
our numerical code is based on @xcite ; here , we briefly describe its essential points .
the equation of the motion for the _ i_-th sph particle is given by @xmath23 where @xmath24 is the velocity of the _ i_-th sph particle , @xmath25 is the time , and @xmath26 and @xmath27 are the pressure gradient and mutual gravity terms between the _ i_-th and _ j_-th particles , respectively .
several forms have been used for the pressure gradient term , none of which appear to be clearly superior to the others . in this paper , we use the following symmetric expression , @xmath28 where @xmath29 , @xmath30 , and @xmath31 are the mass , pressure , and density of the _ j_-th particle , respectively , @xmath32 is the artificial viscosity , @xmath33 is the kernel function , @xmath34 is the distance between the _ i_-th and _ j_-th particles , and @xmath35 is the average smoothing length of the _ i_-th and _ j_-th particles . for the artificial viscosity @xmath32 , we use a von neumann - richtmyer - type viscosity with parameters of @xmath36 and @xmath37 , as described in @xcite . for the kernel function @xmath33
, we use the spherically symmetric spline kernel function proposed by @xcite : @xmath38 this function satisfies @xmath39 , and has a zero value when @xmath40 . in our code
, @xmath41 is variable for each particle and with time , and determined to satisfy the condition that the number of neighboring particles ( @xmath42 ) within @xmath43 is almost constant , @xmath44 .
we used a maximum value of the smoothing length ( @xmath45 ) to save computational cost when searching for neighboring particles and calculating the pressure gradient term .
the value of @xmath45 is determined from @xmath46 where @xmath47 is the minimum density , which is set to @xmath48 in our simulations . the mutual gravity term in equation ( [ eq : sph_eom ] )
can be written as @xmath49 where @xmath50 is the gravitational constant , and @xmath51 is the effective mass of the _ j_-th particle toward the _
i_-th particle defined by @xmath52 this equation gives @xmath53 when @xmath54 . the mutual gravity term between all sph particles
was directly computed using a special - purpose computer for gravitational n - body systems named grape-6a @xcite .
the grape-6a can search for and produce lists of neighboring particles while simultaneously calculating their mutual gravity .
this list of neighbors is used to compute the pressure gradient term in the equation of motion and the time derivative of the internal energy . in a simulation over a period of @xmath55 sec with 20,000 sph particles ,
the typical cpu time was about 4 hours .
thus , the grape-6a allowed us to systematically explore a wide range of impact parameters .
time integration was performed using a pec ( predict , evaluate , and correct ) scheme with variable time steps ( e.g. , * ? ? ?
* ) , which is second - order accurate in time . here
we describe the modeling method for the pre - impact protoplanets .
all the protoplanets are assumed to be differentiated , with a 30% iron core and 70% silicate mantle by mass . in our sph simulations
, we used the tillotson equation of state @xcite , which has been widely applied to giant impact simulations involving shock waves ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the tillotson equation of state contains ten material parameters , and the pressure is expressed as a function of the density and the specific internal energy , which is convenient for treating fluid dynamics .
we used the parameter sets of granite for the silicate mantle and iron for the iron core , which are listed on page 234 of @xcite .
all sph particles in the protoplanets was set to have the same mass , and the total number of particles used for impact simulations was fixed at 20,000 .
for example , in the case of a collision of protoplanets with a mass ratio of 1:9 , the smaller protoplanet consisted of 2,000 particles , and the larger one 18,000 particles . to model the pre - impact protoplanets
, we placed the sph particles in a 3d lattice ( face - centered cubic ) with iron particles on the inside , and rocky particles outside .
the internal energy of the sph particles was set to @xmath56 j / kg .
beginning with this configuration , we calculated vibrations of the protoplanet until the particle velocities become slower than 100 m / s , which is much less than the impact velocity ( order of km / s ) .
after this operation , we used these relaxed objects as the protoplanets for impact simulation .
as a first step , the protoplanets were assumed to have no spin . in order to set impact parameters such as the impact velocity ( @xmath3 ) and impact angle ( @xmath2 ) , we need to determine the radius of the pre - impact protoplanets . since the surface boundary of the protoplanets described by the sph particles is obscure owing to the smoothing length ( @xmath41 ) , we determined the radius of the pre - impact protoplanet ( @xmath57 ) by the following equation .
@xmath58 we prepared more than 1000 sets of initial conditions for the giant impact simulations .
the parameters used were the mass ratio of the protoplanets ( @xmath59 , where @xmath60 and @xmath61 are the mass of the target and impactor , respectively ) , the total mass of the two protoplanets ( @xmath62 ) , the impact angle ( @xmath2 ) , and the impact velocity ( @xmath3 ) .
we systematically varied the mass ratio as @xmath63 1 , 2/3 , 1/2 , 1/3 , 1/4 , 1/6 , and 1/9 . for @xmath63 1 , 1/4 , and 1/9 , we considered three different values of @xmath1 . in total
, we used 13 different combinations for the two colliding protoplanets ( see table [ mass ] ) . for each mass combination
, we varied the impact angle in the range @xmath64 in @xmath65 steps , and the impact velocity in the range @xmath66 in @xmath67 steps , where @xmath6 is the two - body escape velocity defined as @xmath68 where @xmath69 and @xmath70 are the radius of the target ( larger protoplanet ) and the impactor ( smaller protoplanet ) , respectively .
to precisely determine the transition between merging and hit - and - run collisions , we varied @xmath3 with a smaller step size of 0.02 @xmath6 near the transition . in total , we performed more than 1000 runs , consisting of 13 ( mass combinations ) @xmath71 6 ( angles ) @xmath71 16 ( 11 runs with 0.2@xmath6 steps and @xmath72 5 runs with 0.02@xmath6 steps ) .
the impact parameters @xmath3 and @xmath2 are defined when the two protoplanets are in contact with each other ( see figure [ fig : impact_angle ] ) .
we assumed that the two protoplanets are mass points , and calculated backward the positions of two mass points until their distance was apart at a distance of 3 ( @xmath70+@xmath69 ) .
then , we performed the giant impact simulations over a period of @xmath55 sec .
the outcomes of the collisions between the protoplanets are divided into two types : merging and hit - and - run collisions . according to the previous studies @xcite ,
the collision outcomes are subdivided into several regimes ( e.g. , partial accretion , fragmentation and so on ) . however , we classify collision outcomes into only two regimes of merging and hit - and - run collisions , because it is the most essential to the evolution of protoplanets during the giant impact stage that only one big body is left after the giant impact ( i.e. , merging collision ) or two big bodies are left ( i.e. , hit - and - run collision ) . additionally , we need to classify collision outcomes as simply as possible , in order to incorporate those into _ n_-body simulation .
figure [ fig : snapshots ] shows snapshots of two typical collisions . panels ( a ) to ( h ) in figure [ fig : snapshots ] show the time sequence for a relatively low - velocity collision ( @xmath73 ) of same - sized protoplanets ( @xmath74 ) with @xmath10 .
after the first contact , the protoplanets become separated ( see panel ( e ) ) , but remain gravitationally bound .
although some amount of mantle material is ejected , almost all parts of the colliding protoplanets finally merge .
we refer to this type of collision as `` a merging collision '' . on the other hand
, a relatively high - velocity collision leads to a completely different result .
panels ( i ) to ( l ) show the time sequence of a collision with @xmath75 .
the impact angle and protoplanet masses are the same as in panels ( a ) to ( h ) .
after the first contact , the protoplanets escape from each other and are no longer gravitationally bound .
we refer to this type of collision as `` a hit - and - run collision '' . by examining snapshots at @xmath76
, we could easily classify almost all the simulations performed in this study as either merging or hit - and - run collisions .
however , in some cases , the protoplanets were separated but still gravitationally bound at @xmath77 . since those protoplanets are expected to eventually merge , we classified such a case as a merging collision . here , we consider the mass of the protoplanet after the collision . we briefly describe the method to determine the mass of the gravitationally bound objects from the sph particle data after @xmath77 . in the first step ,
we roughly identify clumps of sph particles using a friends - of - friends algorithm ( e.g. , * ? ? ?
we then iteratively check whether or not any sph particles not belonging to clumps are gravitationally bound to clumps . finally , we iteratively identify pairs of clumps that are gravitationally bound .
such clumps should merge after @xmath77 , and we can thus regard them as a single object .
we define the mass of the largest object as @xmath78 and the second largest object as @xmath79 . figure [ fig : result_1vs1_30deg ] shows the mass of the largest object normalized by the total mass , @xmath80 , as a function of the impact velocity normalized by the escape velocity , @xmath81 , for collisions of same - sized protoplanets with @xmath10 .
this figure also shows the numerical results of @xcite for collisions of same - sized protoplanets with @xmath7 , which are in good agreement with our results for total mass @xmath82 of @xmath83 .
we also performed simulations for @xmath84 and @xmath85 .
as seen in figure [ fig : result_1vs1_30deg ] , @xmath80 does not depend on the total mass , when the normalized impact velocity , @xmath81 , is considered .
the insensitivity to the total mass holds true to collisions between protoplanets , which has been predicted by @xcite .
if the material properties such as strength dominates over gravity ( typically collision between smaller bodies with less than 1 km in radius ) , the collision outcomes would depend on the total mass @xcite .
figure [ fig : result_1vs1_30deg ] indicates that @xmath80 changes sharply around @xmath86 .
collisions at impact velocities less than this velocity result in almost perfect accretion ( i.e. , @xmath87 ) , and thus are classified as merging collisions . for the case of a near head - on collision ( @xmath88 )
, @xmath80 gradually decreases with the increase of @xmath3 .
although such a collision should be classified as a partial accretion collision or a fragmentation collision according to @xcite and @xcite , we here classify those as a merging collision because only one large body remains after the collision .
since the probability of a near head - on collision with high velocity is quite low during the giant impact stage , our treatment would not become a serious problem .
on the other hand , figure [ fig : result_1vs1_30deg ] indicates that the collisions at impact velocities higher than @xmath86 result in @xmath89 .
some amount of material become stripped from the protoplanets , and the protoplanets escape from each other .
these collisions are classified as hit - and - run collisions .
we refer to the impact velocity at the transition between merging and hit - and - run collisions as the critical impact velocity , @xmath4 .
the normalized critical impact velocity , @xmath5 , is not strongly dependent on the total mass , @xmath1 .
we obtain @xmath5 = @xmath90 , @xmath91 , and @xmath91 for @xmath92 and @xmath85 , respectively .
the critical impact velocity is expected to depend on the impact angle .
figure [ fig : result_1vs1_60deg ] is similar to figure [ fig : result_1vs1_30deg ] , but for the collisions with @xmath93 .
as is the same in the case of @xmath10 , @xmath80 for @xmath93 does not depend on the total mass .
however , collisions with @xmath93 result in lower @xmath5 values than the @xmath10 case .
the calculated values are @xmath5 = @xmath94 , @xmath95 , and @xmath95 for @xmath92 and @xmath85 , respectively .
this result implies that collisions at higher impact angles are more like to be hit - and - run collisions . the critical impact velocity is also expected to depend on the mass ratio of the protoplanets ( @xmath0 ) .
figure [ fig : result_1vs4_30deg ] shows the results for @xmath96 ( mass ratio of 1:4 ) and @xmath10 .
it can be seen that in this case also , @xmath80 does not depend on the total mass .
this has also been verified for @xmath97 ( mass ratio of 1:9 ) , although the results are not shown here .
as shown in figure [ fig : result_1vs4_30deg ] , @xmath80 changes sharply around @xmath98 .
the normalized critical impact velocities are @xmath5 = @xmath99 , @xmath100 , and @xmath100 for @xmath101 and @xmath102 , respectively , which are larger than for the case of collisions between same - sized protoplanets .
in addition to the impact simulations with 20,000 sph particles ( standard case ) , we performed simulations with 3,000 ( low - resolution case ) , 60,000 and 100,000 particles ( high - resolution cases ) for certain impact parameters in order to check the dependence of convergence on particle number .
although we found that the critical impact velocity for the low - resolution case was slightly different from that for the standard case , the high - resolution cases yielded the same results .
for example , for @xmath103 , @xmath104 , and @xmath105 , we obtained @xmath106 for both the standard and high - resolution cases . therefore , using 20,000 sph particles is enough for determining the critical impact velocity . in subsection
[ sec : pre - impact ] , the initial internal energy of the sph particles was set to @xmath56 j / kg . to investigate the effect of the initial thermal state of the protoplanets
, we prepared pre - impact protoplanets with an internal energies of @xmath107 ( cold - state case ) and @xmath108 j / kg ( hot - state case ) .
we then performed simulations of collisions between same - sized protoplanets with impact parameters near the transition between merging and hit - and - run collisions .
we found that although the results for the very low - impact angle case ( @xmath109 ) showed a very slight dependence on the initial thermal state , for all other cases no such dependence was observed .
in the previous section , we determined the critical impact velocities for several impact parameters . in figure
[ fig : cr_data ] , we summarize @xmath5 for all parameter sets of the mass ratios ( @xmath0 ) and impact angles ( @xmath2 ) . for @xmath103 , @xmath110 , and @xmath111 , we performed simulations for three sets of @xmath1 ( see table [ mass ] ) . in figure
[ fig : cr_data ] , we plot the average of these three results .
the critical impact velocities estimated by @xcite and @xcite are also plotted in figure [ fig : cr_data ] .
@xcite investigated collisions between same - sized protoplanets with masses of @xmath7 , and found that @xmath112 for @xmath10 , and @xmath113 for @xmath12 or @xmath13 .
our results are in good agreement with those of @xcite , since in our simulations , @xmath106 for @xmath10 , @xmath114 for @xmath12 , and @xmath94 for @xmath20 .
@xcite investigated collisions between different - sized protoplanets with mass ratios of 1:2 and 1:10 .
based on figure 17 in @xcite , for a mass ratio of 1:2 , @xmath112 for @xmath10 , and @xmath115 for @xmath116 or @xmath117 . in the present study ,
for a mass ratio of 1:2 , we obtained @xmath118 for @xmath10 , @xmath114 for @xmath12 , and @xmath95 for @xmath20 , and these values are in good agreement with those of @xcite .
although we did not perform simulations for a mass ratio of 1:10 , our data points for 1:9 ( @xmath97 ) seem to fall within the range obtained by @xcite .
as shown in figure [ fig : cr_data ] , @xmath5 increases with decreasing impact angle or mass ratio , which means that collisions with low impact angles or low mass ratios tend to be merging events .
this can be explained in terms of the size of the overlapping volume of the colliding protoplanets . since this is geometrically smaller for higher impact angle , the fraction of kinetic energy converted to thermal energy of protoplanets and kinetic energy of the fragments is small , resulting in a hit - and - run collision .
in addition , in the case of a small impactor ( i.e. , small @xmath0 ) , most of the volume of the impactor tends to overlap with the target .
therefore , the impactor can not easily be ejected , which leads to be a merging collision . in the following ,
we consider a simple physical model in order to express the critical impact velocity as a function of the impact angle and mass ratio . for two spheres with radii @xmath69 and @xmath70 colliding with an impact angle @xmath2 ( see figure [ fig : simple_model ] ) , the mass fractions of the overlapping volumes for the target and impactor ( @xmath119 and @xmath120 , respectively )
are geometrically given by @xmath121 @xmath122 where @xmath123 and @xmath124 are the masses of the overlapping volumes for the target and impactor , respectively .
a constant density is assumed for simplicity . as illustrated in figure [ fig :
simple_model ] , we divide the spheres into overlapping parts and non - overlapping parts , and consider the momentum exchange between the overlapping parts
. using the parameter of the degree of the momentum exchange ( @xmath125 ) , the post impact velocities of the overlapping parts ( @xmath126 and @xmath127 ) are expressed as @xmath128 where @xmath129 is the velocity of the center of mass of the overlapping parts , and is written as @xmath130 for example , when @xmath131 , the velocities of the overlapping parts ( @xmath126 and @xmath127 ) become @xmath129 owing to the complete momentum exchange .
the post impact velocities of the entire target and impactor can be derived based on the conservation of momentum , and written as @xmath132 where @xmath133 and @xmath134 are the velocities for the post - impact target and impactor , respectivery . combining equations .
( [ eq : overlap_post_velocity ] ) - ( [ eq : post_mass_cons ] ) gives @xmath135 if the relative velocity of post - impact objects is higher than the two - body escape velocity , a hit - and - run collision should occur .
therefore , to solve @xmath136 using @xmath137 , the critical impact velocity is obtained as follows : @xmath138^{-1}. \label{eq : form_cr_phys}\ ] ] the calculated results for @xmath139 are drawn as gray curves in figure [ fig : cr_fit ] .
we could roughly reproduce the dependence of @xmath5 on @xmath2 and @xmath0 , but detail features , especially the case for the low - impact angle and low - mass ratio , could not be reproduced .
for such a collision , the role of fragmentation that has not been considered here may become important .
it should be also noted that @xmath139 would not be applied to collisions other than giant impacts between rocky protoplanets .
in addition to the above physical model , we tried to mathematically fit the numerical data for the critical impact velocity .
of the many possibilities available , we found that the following simple formula with five fitting parameters , @xmath140 to @xmath141 , was most effective , @xmath142 where @xmath143 , and @xmath144 .
the fitting parameters are @xmath145 , @xmath146 , @xmath147 , @xmath148 , and @xmath149 .
the fitting curves produced by equation ( [ eq : form_cr ] ) are shown in figure [ fig : cr_fit ] as thick curves , and are in excellent agreement with numerical results .
the value of the fitting parameters derived here is limited to the collision between protoplanets .
almost all previous n - body simulations of terrestrial planet formation during the giant impact stage have been performed based on the assumption of perfect accretion . in order to investigate the statistical properties of fully formed terrestrial planets
, @xcite considered 10 sets of protoplanet initial conditions , and performed 20 runs for each set under the assumption of perfect accretion .
subsequently , to investigate the spin state of the formed planets , @xcite performed additional 30 runs each for 7 sets of protoplanet initial conditions . using the formula for the merging criteria ( equation [ [ eq : form_cr ] ] ) derived in the present study
, we can now determine whether each of giant impacts was a merging or a hit - and - run event .
figure [ fig : impact_events ] shows the normalized impact velocity as a function of impact angle ( left panel ) and mass ratio ( right panel ) for 635 giant impact events during 50 runs under standard initial conditions ( model 1 in @xcite ) .
the symbols denoted by crosses represent hit - and - run events , as determined by equation ( [ eq : form_cr ] ) .
in fact , 40% of all impact events ( 256 out of 635 ) are expected to be hit - and - run collisions .
this result is consistent with a previous study by @xcite , who estimated a hit - and - run probability of roughly half .
during the giant impact stage of terrestrial planet formation in our solar system , a few tens of mars - sized protoplanets collide with each other to form terrestrial planets .
almost all previous studies on n - body calculations of the giant impact stage have been based on the assumption of perfect accretion .
however , recent impact simulations have shown that collisions of protoplanets are not always merging events . as a first step towards studying the effects of such imperfect accretion on terrestrial planet formation , we investigated the merging criteria for a collision of rocky protoplanets . using the sph method
, we performed more than 1000 simulations of giant impacts for various parameter sets , such as the mass ratio of colliding protoplanets ( @xmath0 ) , the total mass of two protoplanets ( @xmath1 ) , the impact angle ( @xmath2 ) , and the impact velocity ( @xmath3 ) .
we investigated the critical impact velocity ( @xmath4 ) at the transition between merging and hit - and - run collisions .
we found that the normalized critical impact velocity , @xmath5 , depends on @xmath0 and @xmath2 , but does not depend on @xmath1 .
we derived a simple formula for @xmath5 as a function of @xmath0 and @xmath2 ( see equation [ [ eq : form_cr ] ] ) , and applied it to the giant impact events considered by @xcite and @xcite .
we found that 40% of these events should not be merging events .
@xcite was the first to performed n - body simulations of the giant impact stage taking into account the merging criteria shown in equation ( [ eq : form_cr ] ) , and investigated the effects of imperfect accretion on terrestrial planet formation .
they found that some basic properties such as the final number , mass , orbital elements , and growth timescale of planets did not change very much , but the spin angular velocity of the fully formed planets was about 30@xmath150 smaller than that for the perfect accretion model .
they also determined that 49% of all impact events were hit - and - run collisions , which is also consistent with our estimate . in this paper
, we focused on the merging criteria for protoplanet collisions . in the future , we plan to investigate additional collisional phenomena such as mantle stripping and ejection of small particles .
using our simulation data for more than 1000 collisional events , we can estimate the change in the core - mantle ratio during the giant impact stage .
this is highly relevant to the formation of mercury , and the formation probability of such planets with very large cores will be investigated .
the ejection of small particles during each collision in the giant impact stage may influence the orbital evolution of terrestrial planets .
the ejected material may have damped the eccentricities of the terrestrial planets to their present low values , although it depends on the total amount of material ejected during the giant impact stage .
acknowledgments _ we thank sarah t. stewart for valuable comments on the manuscript .
this research was partially supported by jsps , the grant - in - aid for young scientists b ( 22740291 ) , and mext , the grant - in - aid for scientific research on priority areas , and the special coordination fund for promoting science and technology . _ c||c|c|c @xmath63 1 & 0.1@xmath151 vs. 0.1@xmath151 & 0.2@xmath151 vs. 0.2@xmath151 & 0.5@xmath151 vs. 0.5@xmath151 + 2/3 & 0.2@xmath151 vs. 0.3@xmath151 & & + 1/2 & 0.1@xmath151 vs. 0.2@xmath151 & & + 1/3 & 0.1@xmath151 vs. 0.3@xmath151 & & + 1/4 & 0.1@xmath151 vs. 0.4@xmath151 & 0.2@xmath151 vs. 0.8@xmath151 & 0.3@xmath151 vs. 1.2@xmath151 + 1/6 & 0.1@xmath151 vs. 0.6@xmath151 & &
+ 1/9 & 0.05@xmath151 vs. 0.45@xmath151 & 0.1@xmath151 vs. 0.9@xmath151 & 0.2@xmath151 vs. 1.8@xmath151 + | at the final stage of terrestrial planet formation , known as the giant impact stage , a few tens of mars - sized protoplanets collide with one another to form terrestrial planets .
almost all previous studies on the orbital and accretional evolution of protoplanets in this stage have been based on the assumption of perfect accretion , where two colliding protoplanets always merge .
however , recent impact simulations have shown that collisions among protoplanets are not always merging events , that is , two colliding protoplanets sometimes move apart after the collision ( hit - and - run collision ) . as a first step towards studying the effects of such imperfect accretion of protoplanets on terrestrial planet formation , we investigated the merging criteria for collisions of rocky protoplanets . using the smoothed particle hydrodynamic ( sph ) method
, we performed more than 1000 simulations of giant impacts with various parameter sets , such as the mass ratio of protoplanets , @xmath0 , the total mass of two protoplanets , @xmath1 , the impact angle , @xmath2 , and the impact velocity , @xmath3 .
we investigated the critical impact velocity , @xmath4 , at the transition between merging and hit - and - run collisions .
we found that the normalized critical impact velocity , @xmath5 , depends on @xmath0 and @xmath2 , but does not depend on @xmath1 , where @xmath6 is the two - body escape velocity .
we derived a simple formula for @xmath5 as a function of @xmath0 and @xmath2 ( eq . ( [ eq : form_cr ] ) ) , and applied it to the giant impact events obtained by _
n_-body calculations in the previous studies .
we found that 40% of these events should not be merging events . |
the effects of temperature changes currently are the most interesting in spin - glass studies . the coexistence of chaos ( or rejuvenation ) and memory @xcite is particularly puzzling .
a first non - trivial problem is whether these effects are universal , i.e. whether they are present in different real spin - glasses @xcite and in numerical simulations of microscopic models @xcite ; the main problem , however , lies in the theoretical explanation of these phenomena .
if one admits that the static ( equilibrium ) free - energy landscape is somehow relevant to the off - equilibrium dynamics , the two effects of chaos and memory seem to contradict each other : the former points in the direction of absence of correlation between the equilibrium states at different temperature , while the latter points in the opposite direction .
a qualitative explanation of the coexistence of the two effects has been given in terms of a hierarchical picture in which states at different temperatures are organized on the same ultrametric tree , whose details are revealed by lowering the temperature @xcite . recent numerical and theoretical work on the subject has confirmed that the generalized random energy model ( grem ) @xcite , a model with such a landscape structure , displays chaos as well as memory @xcite . however , the grem is defined through its phase space , without any reference to any underlying microscopic description ; therefore it would be interesting to find out whether there are spin - glasses that behave like the grem .
the question is whether equilibrium states at different temperatures are correlated or not and which is the nature of the correlations .
the hypothesis of chaos in temperature affirms that they are uncorrelated in finite dimensional systems @xcite ; this has been studied analytically in a number of paper @xcite ; we have questioned the hypothesis validity in the sherrington - kirkpatrick ( sk ) model in a recent paper @xcite , henceforth labeled [ i ] .
the problem has also been investigated numerically @xcite .
it has also been shown that the random energy model ( rem ) , which is non - chaotic by definition , is capable of producing strong rejuvenation signals @xcite .
on the other hand various attempts are underway to substantiate the idea of some kind of hierarchical structure underlying the phenomenology in the context of real - space theories like the droplet model @xcite . in this paper
we consider the mean - field multi-@xmath1-spin spherical spin - glass models for values of the coupling constants such that they display full replica symmetry breaking ( rsb ) @xcite .
it must be remarked that the analysis of the off - equilibrium dynamics @xcite of these models has revealed the coexistence of chaos and memory @xcite .
we study the correlations between the equilibrium states at different temperatures following the lines developed in [ i ] , where the problem has been put in connection with the existence of a particular class of solutions of the saddle point ( sp ) equations obtained in the replica framework . in section 1
we prove that these solutions exist for the class of models considered , implying strong correlations between states at different temperatures , i.e. no chaos in temperature .
these solutions are built with parisi matrices @xcite and in section 2 we show that this structure implies ultrametric relations between the overlaps of states at different temperatures .
these ultrametric relations determine a one - to - one correspondence ( except at the lower levels ) between the trees of states at two different temperatures in such a way that to a cluster of states at temperature @xmath2 , at a level of the @xmath3 tree labeled by some value @xmath4 of the overlap , corresponds a cluster of states at temperature @xmath5 at a level of the @xmath6 tree labeled by the same value @xmath7 .
we call twins " any two clusters in such a correspondence .
specifically , a state at temperature @xmath2 corresponds to a cluster of states at temperature @xmath8 whose minimal overlap is given by @xmath9 .
therefore each state at temperature @xmath2 can be considered as the father of many states at temperature @xmath5 .
a son has only one father , and it has an overlap lower than @xmath9 with all the other states at @xmath2 .
the meaning of the correspondence between the trees at different temperatures is that any relationship between a state or cluster @xmath10 at @xmath5 and any other state or cluster @xmath11 at @xmath2 or @xmath6 is univocally determined by the relationship between @xmath11 and the twin of @xmath10 at @xmath3 . in particular , given a state @xmath11 at @xmath6 and its father @xmath12 at @xmath13 ,
the overlap between @xmath11 and any state @xmath14 at @xmath2 is given by @xmath15 .
each state of a spin - glass system at a given temperature has a certain statistical weight @xmath16 determined by the gibbs measure . according to the parisi solution @xcite the fluctuations of the weights with the disorder
can be described through a stochastic process involving the free energies @xmath17 of the states , which are defined as @xmath18/\sum _
{ k}\exp [ -\beta f_{k } ] $ ] @xcite .
given the tree of states , one considers a given level @xmath19 and assigns independently to each cluster of states at this level a reference free energy chosen randomly such that the average number of clusters with free energy between @xmath20 and @xmath21 is given by @xmath22df \label{legge1}\ ] ] where @xmath23 is the inverse of the parisi function @xmath24 .
after having applied the procedure to every level of the tree , the free energy of a given state is set to be the sum of the reference free energies of the clusters to which it belongs at the various levels .
for instance , if the state @xmath10 belongs to the cluster @xmath25 at the first level of the tree , to the cluster @xmath26 at the second level of the tree and so on , its free energy is given by @xmath27 the distribution obtained has various interesting features ; for instance , in each sample there are only few states with a finite weight , while there is an infinite number of states that carries an infinitesimal weight . since we found a one - to - one correspondence between the trees of states at different temperatures , it is natural to whether if there are also correlation between the weights . in the replica framework the standard procedure to cope with this problem
requires the computation of sums over all the different solutions of the saddle point equations , in order to obtain the cumulants of the distribution function of the weights . when considering a system at a given temperature ,
this can be done noticing that all the solutions are permutations of the standard parisi solution @xcite ; therefore the sum over solutions can be replaced by a sum over replica indices ; unfortunately when considering systems at different temperatures the solutions are not permutations one of the other and we are unable to sum them all . however , in section 3 we conjecture that the weights distribution is identical to that of the grem class of models . in the grem
the tree of states is the same at all temperatures by definition and the reference free energies of the clusters are constant too @xcite .
their distribution obeys the law ( [ legge1 ] ) with @xmath28 replaced by @xmath29 , a function that depends on the model . in analogy to this we guess that in the models we are considering the reference free energy of a given cluster at @xmath3 is equal to the reference free energy of its twin at @xmath6 ; we notice that since for this models we have @xmath30 , ( where @xmath29 depends on the coupling constants but not on the temperature ) , this is consistent with the fact that the same set of free energies obeys law ( [ legge1 ] ) at all temperatures .
this belief is motivated by the fact that there is a one - to - one correspondence between the solutions of sp equations of the multi-@xmath1 spin spherical model and the solutions one finds in the analogous replica treatment of the grem @xcite .
indeed , the solutions have the same _ formal _ structure expressed in terms of the function @xmath31 and the differences between the models shows up only in the actual values of this function . since for every solution of the multi-@xmath0 spherical spin - glass models there is a correspondent solution of the grem , we expect that the sums over solutions give equal results in the two models ; more precisely we expect the results to have the same dependence on the function @xmath29 .
accordingly , the correlations between the weights at different temperatures will be equal in the two classes of models .
the previous conjecture could be proved if we were able to carry out the sum over all the solutions .
this problem is common to many models where one would like to recover within the replica framework a result which is known _ a priori _ ( e.g. in the rem @xcite ) , or which can be obtained through different methods ( e.g. the spherical @xmath0-spin model where the absence of chaos in easily seen within the tap approach @xcite ) . in section 4
we discuss some technical difficulties connected with this problem . at the end
we give our conclusions .
technical details are skipped to the appendices .
we recall the method discussed in [ i ] to study the correlations between states at different temperatures . the replica trick is usually used in order to compute the average over the disorder of the free energy by computing the moments of the partition function @xmath32 . in the thermodynamic limit saddle point
( sp ) equations are obtained for the order parameter which is an @xmath33 matrix @xmath34 .
an outcome of the computation is that the order parameter is connected to the distribution of the overlaps through the relation @xmath35 where the function @xmath36 is the averaged probability density of finding two states with overlap @xmath37 according to their gibbs weight , and @xmath38 and @xmath39 are two different replica indices ( e.g. 1 and 2 ) . in the r.h.s . of ( [ pq1 ] )
there appears a sum over all the different solutions of the saddle point equation with the same ( maximum ) free energy ; indeed , as soon as replica symmetry is broken , we have many solutions : given a certain solution others can be obtained through a permutation of the replica indices .
however , in the isothermal case , all the solutions are given by the parisi solution plus its natural permutations ; therefore the sum over them can be replaced with an average over the replica indices multiplied by the total number of different solutions , which in the @xmath40 limit goes to @xmath41 @xcite . in [ i ]
it has been shown that the correlations between states at different temperatures can be investigated through the computation of the @xmath42 limit of the quantity @xmath43 . as a matter of fact
, we are not interested in its actual value ( which must be equal to the sum of the free energies at temperature @xmath2 and @xmath5 ) , but in the saddle point equation obtained by applying the replica trick to the whole @xmath44 and not to @xmath45 and @xmath46 separately , the inverse of its temperature . therefore a claim like `` the free energy of the two systems must be equal to the sum of the free energies at temperature @xmath3 and @xmath6 '' is shorthand for `` the logarithm of the partition function of the two systems must be equal to the sum of the logarithms of the partition functions at temperature @xmath3 and @xmath6 '' . ] . in this case one
obtains sp equations for a @xmath47 matrix @xmath48 , where @xmath49 @xmath50 and @xmath51 are @xmath33 matrices .
the advantage of this procedure is that the matrix @xmath51 encodes information on the correlations between states at different temperatures ; indeed , through the same steps that led to ( [ pq1 ] ) it can be shown that @xmath52 where the function @xmath53 is the generalization of the @xmath36 to states at different temperatures .
as above , we have to sum over all the solutions of the sp equations : this is quite a delicate point ; indeed we do not expect that all solutions be natural permutations of a single solution as in the isothermal case ( i.e. the standard parisi solution ) . actually , there are infinite solutions corresponding to different parameterizations and we do nt know how to sum them all .
however , it is possible to reconstruct the function @xmath53 in an indirect way by knowing the weight distribution of the equilibrium states .
we have applied the method to the spherical spin - glass model with multi-@xmath1-spin interaction @xcite .
the hamiltonian of the model is defined as@xmath54 the @xmath55 s are independent gaussian random variable with zero mean and variance @xmath56 .
the spins are subjected to the spherical constraint @xmath57 . introducing the function @xmath58
the replicated free energy reads @xmath59-\tr\ln q - n\ ] ] where the variational parameter is an @xmath33 matrix @xmath34 with diagonal @xmath60 .
the sp equations then read@xmath61 expressing @xmath34 as a parisi function we obtain the solution as @xmath62 where @xmath63 is defined as the inverse of the universal function@xmath64 with the value of @xmath65 fixed by@xmath66 in the presence of a small magnetic field there is also a small plateau of temperature - independent height @xmath67 fixed by the condition@xmath68 when we consider simultaneously two systems at different temperature , we get the following expression for @xmath44 in the thermodynamic limit @xmath69 @xmath70 the corresponding sp equations are similar to those for the isothermal case ( [ spsingle ] ) and are reported in appendix 2 .
a solution of the sp equations is certainly the one with @xmath71 and @xmath72 and @xmath50 equal to the corresponding isothermal solutions .
its free energy is given by the sum of the free energies at temperatures @xmath2 and @xmath5 as expected .
the problem is whether other solutions exist with a non - zero @xmath51 and with the same free energy of the @xmath71 solution . in [ i ]
a particular structure was proposed for these @xmath73 solutions ( called non - chaotic , since their existence implies absence of chaos in temperature ) ; we found that solutions having such a structure do exist for this class of models .
actually there is an infinite number of solutions which are parameterized by the value @xmath74 of the diagonal of @xmath51 , whose value ranges from zero to a maximum one , which , for these models , turns out to be the self - overlap of the states at the higher temperature .
the solution with @xmath75 is included in this set and corresponds to the value @xmath76 .
it is very interesting to notice that they exist also when the two systems at different temperatures are subjected to the same magnetic field ( see appendix 2 ) . here for simplicity we will refer to the @xmath77 case . for a given value of @xmath74 the solutions in terms of the three functions @xmath78 , @xmath79 and @xmath80 are@xmath81 where @xmath63 and @xmath31 are the universal function defined above .
these solutions can be built for any couple of temperatures @xmath82 both below the critical temperature and for any @xmath74 between zero and @xmath83 which is the self - overlap of the states of the system at the higher temperature .
the solutions are sketched in figures 1 and 2 .
notice that @xmath84,@xmath85 and @xmath86 are all equal in the small-@xmath87 region .
this is connected with the fact that this class of models verifies the scaling @xmath88 .
we skip to appendix 2 the demonstration that these functions solve the sp equations and have the correct free energy .
it must be remarked that the solutions ( [ solq],[solp ] ) are formally the same which have been found with a similar treatment of the grem @xcite , with a different model - dependent function @xmath29 . at this stage
we can infer that the function @xmath53 has a non - zero support from zero to the self - overlap of the states at the higher temperature .
the solutions we found imply ultrametricity between states at different temperatures of a given system . to see this we must consider the probability @xmath89 of extracting three states at different temperatures with assigned values of their mutual overlap . if we assume , for instance , that states @xmath90 and @xmath91 are at temperature @xmath2 and state @xmath92 is at temperature @xmath5 , this function is related to the solutions of the sp equation through the following relation@xmath93 we are not able to perform the sum over solutions in r.h.s of the previous expression , but we can infer ultrametricity by simply looking at their structure .
the function @xmath89 can be reconstructed later in an indirect way from the distribution of the weights we will describe below ( section 3 ) .
( 260,260)(0,0 ) ( 0,0)(120,120 ) ( 0,140)(120,120 ) ( 140,0)(120,120 ) ( 140,140)(120,120 ) ( 0,60)(60,-60)2(60,60 ) ( 0,200)(60,-60)2(60,60 ) ( 140,60)(60,-60)2(60,60 ) ( 140,200)(60,-60)2(60,60 ) ( 0,90)(30,-30)4(30,30 ) ( 0,230)(30,-30)4(30,30 ) ( 140,90)(30,-30)4(30,30 ) ( 140,230)(30,-30)4(30,30 ) ( 0,245)(15,-15)8(15,15 ) ( 0,260)(1,-1)120 ( 0,245)(15,-15)8(1,1)15 ( 140,110)(10,-10)12(10,10 ) ( 140,120)(1,-1)120 ( 140,110)(10,-10)12(1,1)10 ( 27,27)@xmath94 ( 87,87)@xmath94 ( 12,72)@xmath95 ( 42,102)@xmath95 ( 72,12)@xmath95 ( 102,42)@xmath95 ( 12,102)(30,-30)4@xmath96 ( -10,-3)@xmath97 ( -12,57)@xmath98 ( -12,87)@xmath99 ( 27,167)@xmath94 ( 87,227)@xmath94 ( 12,212)@xmath100 ( 42,242)@xmath100 ( 72,152)@xmath100 ( 102,182)@xmath100 ( -10,137)@xmath97 ( -12,197)@xmath98 ( -12,227)@xmath99 ( 167,27)@xmath94 ( 227,87)@xmath94 ( 152,72)@xmath100 ( 182,102)@xmath100 ( 212,12)@xmath100 ( 242,42)@xmath100 ( 130,-3)@xmath97 ( 128,57)@xmath98 ( 128,87)@xmath99 ( 167,167)@xmath94 ( 227,227)@xmath94 ( 152,212)@xmath95 ( 182,242)@xmath95 ( 212,152)@xmath95 ( 242,182)@xmath95 ( 152,242)(30,-30)4@xmath96 ( 130,137)@xmath97 ( 128,197)@xmath98 ( 128,227)@xmath99 let us consider the solution for a given @xmath74 : a simple analysis of figure ( 2 ) shows that for any three replica indices @xmath38 , @xmath39 , and @xmath101 , the corresponding overlaps @xmath102,@xmath103 and @xmath104 always form an isosceles triangle with the two equal sides smaller or equal to the third
. however , while @xmath102 can take values from zero to @xmath9 , @xmath103 and @xmath104 can take values from zero to @xmath74 .
considering other solutions , we obtain values of @xmath105 between @xmath106 and @xmath107 .
now , since ultrametricity is a property of all the solutions , it will be a property of the sum over solutions as well ; therefore the overlaps between any three states at any temperature satisfy ultrametric relations .
in general , the maximum overlap between a state at temperature @xmath2 and another state at temperature @xmath5 is equal to the self - overlap of the state at the higher temperature .
perfect ultrametricity between states at different temperatures is a special property of this class of models which relies on the identity of the three function @xmath78 , @xmath79 and @xmath80 in the small-@xmath87 region . as we said in the introduction ,
the ultrametric relations between the overlaps of states at different temperatures define a correspondence between the trees of states .
consider a cluster @xmath108 of equilibrium states at temperature @xmath3 whose overlaps are greater or equal to some @xmath109 . given a state @xmath110 in @xmath108
, we have that the overlap between any other state @xmath111 in @xmath108 and any state @xmath112 outside @xmath108 is simply given by @xmath113 @xcite .
now , given a state @xmath114 at @xmath6 whose overlap with any of the states in @xmath108 is greater or equal to @xmath109 , we have that the overlap between @xmath114 and any state @xmath112 at @xmath3 outside @xmath108 is simply given by @xmath115 ; in other words , from a geometrical point of view we may assume that @xmath114 _ is _ in @xmath108 .
furthermore , this is true for all the states at @xmath6 whose overlap with @xmath114 is greater or equal than @xmath109 ; by definition they form a cluster @xmath116 which we call the twin " of @xmath108 . as far as the overlaps are concerned the two clusters @xmath108 and
@xmath116 can be considered the same ; therefore from a topological point of view there is only one tree of states whose details are revealed by lowering the temperature .
finally , we recall that the solutions we described satisfy the separability property @xcite , which means that the overlap contains all the information between two states at equal or different temperatures .
each state of a spin - glass system at a given temperature has a certain statistical weight @xmath16 .
the fluctuation of the weights with the disorder can be described through the following procedure @xcite involving the free energies of the states defined as @xmath18/\sum _
{ k}\exp [ -\beta f_{k } ] $ ] .
given the tree of states , we consider a given level and assign to each cluster of states at this level a reference free energy chosen randomly such that the average number of clusters reference with free energy between @xmath20 and @xmath21 is given by @xmath22df\ ] ] after applying the procedure to any level of the tree we define the free energy of a given state as the sum of the reference free energies of the cluster to which it belongs at the various levels . for instance , if the state @xmath10 belongs to the cluster @xmath25 at the first level of the tree , to the cluster @xmath26 at the second level of the tree and so on , its free energy is given by @xmath117 we have established that the trees of states at different temperatures are equal : to a cluster of states at @xmath2 corresponds a cluster of states at @xmath5 ; therefore we want to know which is the relation between the corresponding reference free energies .
this correlation can be obtained computing the quantities @xmath118 where the index @xmath119 refers to the clusters at a certain level @xmath37 of the tree and @xmath120 and @xmath121 are the weight of the cluster @xmath119 respectively at temperature @xmath2 and @xmath5 . through standard manipulation
@xcite we obtain @xmath122 again , the problem is reduced to the computation of a sum over all the solutions of the sp equations .
we are unable to perform such a sum but we have a line of reasoning to guess which may be the result .
indeed , the previous expression does not depend on the model under consideration ; it is a general outcome of the replica trick .
therefore , if we have two different models with the same set of solutions the corresponding weight distribution functions will be equal . the solutions ( [ solq],[solp ] ) found above are formally the same which have been found with a similar treatment of the grem @xcite , with a scaling @xmath123 in the small-@xmath87 region . in other words , for any solution with @xmath73 of the spherical model with multi - spin interactions
there is a corresponding solution of the grem with the same parameterization .
since there is a one - to - one correspondence between solutions of the grem and solutions of the multi-@xmath0-spin spherical model , we expect all the quantities of interest to have the same formal dependence on the temperatures and on the universal function @xmath124 .
this correspondence prompt us to conjecture that the structure of the correlations between the weights is the same in the two models , i.e. that _ the reference free energy of a given cluster at temperature @xmath2 is equal to the reference free energy of the corresponding cluster at temperature @xmath5 . _ to be precise one should notice that the @xmath125 of the grem has a qualitatively different shape from that of the models we are considering ; indeed the solutions of the grem does not have a plateau at @xmath126 and furthermore @xmath127 for @xmath128 .
this makes no difference at all : since the solutions has the same formal expression in terms of the function @xmath31 , the formal expressions of quantities like the @xmath53 must be the same , and the difference between the two models is only in the actual value of @xmath31 .
according to the previous argument the reference free energies of the various clusters are fixed at all temperatures , in particular the ordering of the clusters according to them does not change ; however , this does not mean that the ordering of the clusters according to their actual weights is also conserved : this is only true on average .
in other words , it is not true that to the heaviest cluster at a given temperature corresponds the heaviest cluster at a lower temperature .
in the previous sections we saw that in order to obtain the quantities of interest we should perform sums over all the different solutions of the sp equations of the functional ( [ fun2 ] ) . here
we intend to discuss one of the main difficulties which prevents us from performing such sums , i.e. the existence of an infinite set of solutions corresponding to different parameterizations of the @xmath129 order parameter @xmath130 . by applying the same technique of section 2 to the rem @xcite we found the same essential features of the problem . here
the solution of the standard isothermal problem is a very simple @xmath24 whose value is @xmath106 for @xmath131 and @xmath41 for @xmath132 ; this is the simplest case of a solution that verifies the so - called parisi - toulouse scaling @xmath133 @xcite , which is also found in models with full rsb like those we are considering .
as we saw in section 2 the problem is connected to the extremization of the quantity @xmath134 with respect to the order parameter which is a @xmath47 matrix @xmath48 , where @xmath49 @xmath50 and @xmath51 are @xmath33 matrices .
one can show that there exist solutions of the type ( [ solq],[solp ] ) for the rem ; in particular , we have that @xmath135,@xmath136 and @xmath105 are 1rsb matrices with values @xmath106 for @xmath137 and @xmath41 for @xmath138 . in the solutions ( [ solq],[solp ] ) the matrices @xmath49 @xmath50 and @xmath51
are parisi matrices , but there exist other solutions corresponding to different parameterizations .
for instance we can divide each of the three matrices @xmath49 @xmath50 and @xmath51 in four blocks of size @xmath139 .
we have then @xmath140 blocks and we can parameterize each of them as a parisi matrix . in the @xmath141
limit the corresponding sp equations are the same obtained by considering four different system , two of them at @xmath3 and the other two at @xmath6 .
the order parameter in the latter case is a @xmath142 matrix composed of 16 matrices of size @xmath143 @xmath144 when we parameterize each of these matrices through parisi functions @xmath145,@xmath146 and take the limit @xmath40 the corresponding equations are identical as they would be if the size of the matrices were @xmath147 rather than @xmath97 ; in other words , the solutions of the four - system problem are also solutions of the two - system problem , and they offer another way of parameterizing the order parameter @xmath48 . this argument can be extended indefinitely considering at the same time a general number of @xmath148 systems , @xmath0 at temperature @xmath3 and @xmath0 at temperature @xmath6 . in this way
an infinite set of possible parameterizations of the @xmath129 order parameter @xmath130 in terms of parisi matrices is obtained . however , not all the parameterizations give new solutions .
going back to parameterization ( [ par1 ] ) we have @xmath140 parisi matrices to consider ; in the rem we have a solution in which they are all equal 1rsb matrices with a breaking point @xmath149 .
it can be seen that this is nothing but a permutation of the solution in which @xmath135,@xmath136 and @xmath105 are 1rsb matrices with @xmath150 ( see also the discussion in the conclusions ) .
therefore the @xmath151-system parameterization does not add anything new in this case .
however , the 4-system parameterization describes also the following solution @xmath152 this solution can not be obtained through a permutation from the solution of the type ( [ solq],[solp ] ) and should be counted separately .
essentially , this parameterization corresponds to two systems ( system @xmath41 and @xmath153 ) at temperature @xmath3 correlated with a system at temperature @xmath6 ( system @xmath154 ) , while system @xmath151 at @xmath6 is completely uncorrelated to the remaining three .
accordingly , in summing over all the different solutions we must consider the infinite set of solutions whose parameterization corresponds to a generic number of @xmath155 systems at temperature @xmath3 correlated to a generic number of @xmath156 systems at temperature @xmath6 , plus @xmath157 ( if @xmath158 ) uncorrelated systems at temperature @xmath3 ( in order to have the same number of system at each temperature ) . the scaling of the correlated components is @xmath159 , i.e. in the rem the correlated components are described by 1rsb function with @xmath160 . according to [ i ] any solution with a given parameterization
can also be considered as solution of a certain constrained system .
for instance , the blocks @xmath161,@xmath162,@xmath163,@xmath164,@xmath165,@xmath166 , in the solution ( [ sol3 ] ) correspond to a system of three real replicas which are constrained to have maximum overlap ( i.e. @xmath41 in the rem ) among themselves .
when considering a constrained system the problem of summing over all solutions greatly simplifies ; indeed in this case we have only one solution plus its natural permutations and we can perform the sum in the standard way by replacing it with a sum over indices . the result is immediate , for the @xmath167 of the constrained system we have @xmath168 .
we define the `` free energy '' @xmath169 of a state of the global system ( composed of two constrained systems ) as @xmath170/\sum_\beta \exp [ -f_\beta]$ ] ( see footnote 1 ) .
then we obtain that these free energies are independent random variables such that the average number of states with free energies between @xmath171 and @xmath172 is @xmath173df$ ] .
this result is precisely what is obtained noticing that the states of the constrained system are triplets of identical states of the single system ; therefore in the rem their free energy is simply given by @xmath174 ( see footnote 1 ) and the distribution of @xmath175 is proportional to @xmath176=\exp[\beta_2 x_2 f]=\exp[f]$ ] .
the same line of reasoning applies to the models we considered in the previous sections as well : it ensures that the results obtained by considering constrained systems ( the only case in which we are able to sum over the solutions explicitly ) are fully consistent with our guess on the distribution of the free energies made in section 3 .
we proved the existence of correlations between the equilibrium states at different temperatures of the spherical spin - glass models with multi-@xmath0 spin interactions for values of the coupling constants such that they display full rsb : there is no chaos in temperature in such models .
furthermore , the overlaps between states at different temperatures satisfy ultrametric relations .
ultrametricity determines a one - to - one correspondence between the trees of states at different temperature : to any cluster of states at temperature @xmath3 at a given level of the tree corresponds a twin " cluster of states at temperature @xmath6 ; the precise meaning of this correspondence has been discussed in details in the introduction and in section 2 . from a purely geometrical point of view
we may say that there is only one tree of states at all temperatures , whose details are revealed lowering the temperature .
prompted by some technical features of the problem , we conjectured that the reference free energies of two twins clusters are equal at all temperatures , exactly as in the grem
. consequently , we expect that quantities like the @xmath53 in these models are the same as the grem .
more precisely we expect them to have the same formal dependence on the temperatures and on the function @xmath31 , whose actual values will be different in the two classes of models . the scenario of a tree of states which bifurcates when lowering the temperature was suggested early in spin - glass studies @xcite and was later advocated in order to explain the phenomenology of rejuvenation and memory in spin - glasses @xcite . according to our findings , the basic premises of that phenomenological picture of off - equilibrium dynamics hold for the class of models considered here ; we believe that this further increases the need for a complete understanding of the relationship between the equilibrium energy landscape and the off - equilibrium dynamics in spin - glasses . within the tap approach @xcite strong correlations between the equilibrium states at different temperatures
are readily obtained in the spherical @xmath1-spin model with 1rsb @xcite . indeed in this model
the angular and the self - overlap contributions to the tap free energy can be factored ; that is if we write @xmath177 , we have that the angular part @xmath178 enters the expression of the tap free energy only in the form @xmath179 . accordingly ,
the angular components of the states , which are the extrema of the free energy , are the same at all temperatures . to our knowledge
this was the only result on the absence of chaos previously obtained for a spin - glass model .
the previous argument can not be used in dealing with spherical spin - glass models with multi-@xmath0 interactions for values of the coupling constants such that they display either 1rsb or full - rsb .
for instance , in the @xmath180 model there is an angular dependence of the form @xmath181 and no factorization is possible .
indeed , when the temperature changes the self - overlap changes with it causing the angular landscapes corresponding to the two interactions to interpenetrate . it would be interesting to understand how this process leads to the properties we found within the replica approach .
furthermore , we recall that the off - equilibrium dynamics @xcite is connected to the tap free - energy landscape @xcite . as discussed in [ i ] , the solutions ( [ solq],[solp ] ) have the same free energy of the @xmath75 solution .
this can be also derived noticing that they form a continuous line in the space of the matrices @xmath130 parameterized by the value of the continuous parameter @xmath96 .
the value @xmath76 corresponds to the standard parisi solutions at the two temperatures . on this continuous line of solutions
we have @xmath182 by definition ; therefore the free energy is constant and equal to that of the @xmath76 solution .
the previous argument has deep consequences .
it provides an easy way to understand the goldstone theorem for disordered systems with full rsb which has been proved and discussed in @xcite .
this theorem connects the presence of goldstone modes in spin - glasses with continuous rsb with the fact that the discrete permutational symmetry within the replica approach becomes a continuous symmetry when the number of rsb steps goes to infinity .
consider the standard replica formulation of a given model : it deals with a @xmath143 order parameter @xmath183 . instead of making the parisi ansatz on it
, we can divide it in four @xmath184 blocks and make the parisi ansatz on each of these blocks separately , i.e. we assume that the two diagonal blocks are equal to an ultrametric @xmath185 matrix @xmath186 and the two off - diagonal blocks are equal to a @xmath184 ultrametric matrix @xmath187
. then the sp equations are written in terms of two functions @xmath24 and @xmath86 defined on the interval @xmath188 $ ] , but in the @xmath141 limit the equations are identical to those obtained considering @xmath189 replicas with an order parameter @xmath130 parameterized by four @xmath143 matrices , the two diagonal ones equal to @xmath183 and the two off - diagonal ones equal to @xmath190 .
this is a special case ( @xmath191 ) of the problem we considered in this paper . in this case
@xcite the solutions corresponding to ( [ solq],[solp ] ) read @xmath192 where @xmath193 is the standard parisi solution and @xmath96 is a continuous parameter with range @xmath194 $ ] ; every value of @xmath96 specifies a solution . as stated above , these solutions , first obtained considering @xmath189 replicas , solve the standard @xmath97-replica problem too . at this point
we can make the following statements : * these solutions exist in every model with parisi rsb , either discrete of continuous .
this statement has already appeared in @xcite where it is claimed that it is implied by ultrametricity .
we justify this result noticing that these solutions are _ permutations _ of the standard parisi solution , as it is readily understood considering fig .
( [ figure0 ] ) .
* when the model has full rsb the parameter @xmath96 is continuous , therefore the solutions form a continuous line of constant free energy . as a consequences on every point of the line
the hessian has zero eigenvalues .
* acknowledgements*. i thank s. franz , g. parisi and l. peliti for interesting discussions .
it s a pleasure to thank my family for constant help and support .
in this appendix we derive the formulas to compute a generic function ( e.g. inverse , logarithm ) of one or more parisi matrices . using the standard eigenvalue technique
one encounters some difficulties .
the main problem is that in the @xmath141 limit the matrix ceases to be determined univocally by the set of its eigenvalues . for a generic parisi matrix parameterized as @xmath195
the eigenvalues are @xcite @xmath196 two important eigenvalues are @xmath197 and @xmath198 , in this context the bar means integration of the parisi function @xmath199 over the interval @xmath200 $ ] . by direct inspection
one can check that the eigenvalues verify the property @xmath201 , where @xmath202 is the product of parisi algebra @xmath203 it can be easily checked that the @xmath204 matrix @xmath205 that projects on the vector of constant coordinates ( i.e. @xmath206 ) has all zero eigenvalues in the limit @xmath207 . as a consequence
the set of eigenvalues ceases to determine the matrix @xmath208 in a unique way . to recover the function @xmath195 we need further information e.g. @xmath209 . using @xmath209 and the relation @xmath210 (
the dot means derivative with respect to @xmath211 ) which follows from the expressions ( [ eige0],[eige1 ] ) we obtain the following inversion relations @xmath212 @xmath213 given a function @xmath214=\sum_k f_k a^k$ ] we want to compute @xmath215 $ ] for a generic ultrametric matrix a. the eigenvalues @xmath216}$ ] of @xmath215 $ ] are readily obtained @xmath217}(x)=f[\l_a(x)]\ ] ] to use the formulas ( [ a0x],[a0d ] ) we need to know @xmath215(0)$ ] .
the expression for a product is @xmath218 as a consequence we have the following expression for the powers of @xmath208 @xmath219 therefore @xmath220(0)=\sum_k f_k ( a^k)(0)=\sum_k f_k k ( a_d-\overline{a})^{k-1}a(0)=a(0)\frac{df}{da}[a_d-\overline{a}].\ ] ] summing up , the expression of a generic function @xmath215 $ ] is @xmath221(x ) & = & a(0)\frac{df}{da}[a_d-\bar{a}]+\int_0^xdy\ \dot{a}\frac{df}{da}[\l _ a(y)]\label{fax } \\
{ 1\over n } \tr f[a ] & = & a(0)\frac{df}{da}[a_d-\bar{a}]+f[a_d-\bar{a}]-\int_0 ^ 1\frac{f[\l _ a(y)]-f[a_d-\bar{a}]}{y^2}dy\label{fad}\end{aligned}\ ] ] the generalization to a function @xmath222 $ ] of @xmath0 parisi matrices the function is straightforward @xmath223(x ) & = & \sum_{i=1}^pa_i(0){\partial g\over \partial a_i}[a_{1d}-\bar{a}_1,\ldots , a_{pd}-\bar{a}_p]+ \\ & + & \int_0^xdy\sum_{i=1}^p\dot{a}_i{\partial g\over \partial a_i}[\l _ { a_1},\ldots,\l _ { a_p } ] \label{gx}\end{aligned}\ ] ] @xmath224 & = & \sum_{i=1}^pa_i(0){\partial g\over \partial a_i}[a_{1d}-\bar{a}_1,\ldots , a_{pd}-\bar{a}_p]+ g[a_{1d}-\bar{a}_1,\ldots , a_{pd}-\bar{a}_p]+ \\ & - & \int_0 ^ 1\frac{g[\l _ { a_1},\ldots,\l _ { a_p}](y ) -g[a_{1d}-\bar{a}_1,\ldots , a_{pd}-\bar{a}_p]}{y^2}dy\end{aligned}\ ] ]
in this appendix we check that the solutions ( [ solq],[solp ] ) verify the sp equation .
we consider the general situation where a non zero magnetic field is present .
the free energy functional read @xmath225 @xmath226 the dependence on @xmath130 can be simplified trough @xmath227 .
the sp equations then read @xmath228 @xmath229 we recall the definition of the temperature - independent function @xmath29 and of its inverse @xmath124 @xmath230 the self - overlap @xmath231 of the equilibrium states at a certain temperature is given by @xmath232 the minimum overlap @xmath233 in presence of a magnetic field @xmath234 is independent of the temperature and is given by @xmath235 in presence of a magnetic field the generalization of the solutions ( [ solq],[solp ] ) is @xmath236 these solutions are valid for any couple of temperatures @xmath237 both below the critical temperature and for any @xmath96 between @xmath233 and @xmath238 which is the self - overlap of the states of the system at the higher temperature . to check
that they verify the sp equation ( [ spq1],[spp ] ) we need to express their r.h.s . in the parisi form , this is readily done applying equation ( [ gx ] ) @xmath239 @xmath240 in the small-@xmath211 region the three functions are equal so the previous expressions simplify in this region to @xmath241 the various quantities entering the previous expressions read @xmath242 when evaluating the quantities entering the integrals the dependence on @xmath96 disappears so as the discontinuity at @xmath243 :
@xmath244 evaluating the integrals in ( [ inte2],[inte1 ] ) through ( [ exin ] ) we obtain @xmath245 therefore the sp equations are verified in the small-@xmath211 region ; now for @xmath246 we have @xmath247 and @xmath248 so the equation for @xmath0 is immediately verified while @xmath84 and @xmath85 decouple @xmath249 @xmath250 we skip the explicit evaluation of the free energy ; it turns out to be the sum of the free energies at temperature @xmath3 and @xmath6 , as it should .
u. krey , j. mag . mag . mat . * 6 * ( 1977 ) 27 ; r. g. palmer , in heidelberg colloquium on spin glasses " , morgenstern and van hemmen eds .
( springer , 1983 ) ; m. mezard , g. parisi , n. sourlas , g. toulouse and m.a .
virasoro , phys .
* 52 * ( 1984 ) 1156 ; m. mezard , g. parisi , n. sourlas , g. toulouse and m.a .
virasoro , j. physique * 45 * ( 1984 ) 843 - 854 reprint in @xcite .
d. j. thouless , p. w. anderson and r. g. palmer , phil . mag . * 35 * 593 ( 1977 ) , reprint in @xcite ; see r. mulet , a. pagnani , g. parisi , phys . rev .
b 63 , ( 2001 ) 184438 for recent work on tap solutions at different temperatures . | we prove the existence of correlations between the equilibrium states at different temperatures of the multi-@xmath0-spin spherical spin - glass models with continuous replica symmetry breaking : there is no chaos in temperature in these models .
furthermore , the overlaps satisfy ultrametric relations . as a consequence
the parisi tree is essentially the same at all temperatures with lower branches developing when lowering the temperature .
we conjecture that the reference free energies of the clusters are also fixed at all temperatures as in the generalized random - energy model . |
hydrogen ( h ) is a ubiquitous impurity in a wide variety of semiconductors , including elemental ( e.g. , si ) and binary compound ( e.g. , gaas ) materials , which comprise the primary basis for current industrial applications . in these covalent systems , h is known to be amphoteric , forming an acceptor or donor level in @xmath1- or @xmath7-type materials , respectively @xcite . meanwhile , a recent theoretical prediction that h could behave as an independent electron donor to induce @xmath1-type conductivity in zno @xcite , along with subsequent support from various experiments involving muon spin rotation ( @xmath8sr ) @xcite , electron paramagnetic resonance ( epr ) , electron - nuclear double resonance ( endor ) measurements @xcite , and additional conventional techniques @xcite , have prompted the development of a generalized hypothesis of h as a shallow donor dopant in wide gap semiconductors @xcite . as in the case of zno , @xmath8sr studies have made a significant contribution to this subject by providing spectroscopic information on muonium ( mu ; an analog of a neutral h atom where the proton is replaced by a positive muon ) observed in a wide variety of semiconductors . while the dynamical aspects ( e.g. , diffusion ) of mu and h may differ considerably ,
because of the relatively low mass of mu ( @xmath9 , with @xmath10 being the proton mass ) , the local electronic structure of mu is virtually equivalent to that of h if a small correction to account for the difference in the reduced mass ( @xmath50.5% ) is made .
the recent discovery of novel mu states having extremely small hyperfine parameters ( @xmath510@xmath11 times smaller than that of mu in vacuum ) and low ionization energy ( @xmath510@xmath12 mev ) in several compound semiconductors , including cds @xcite , znse @xcite , inn @xcite , and gan @xcite , also supports the hypothesis that mu ( and hence h ) can act as a donor in these compounds . the generalized hypothesis also predicts shallow donor h in various metal oxides , most notably in rutile tio@xmath0 , where the origin of the unintentional @xmath1-type conductivity exhibited by the as - grown crystal is the focus of intense research .
while the electrical and optical properties of tio@xmath0 are subject to both intrinsic and extrinsic defects , this material has a rich variety of potential applications in both electronic and opto - electronic devises , and detailed understanding of the h behavior is crucially important to the material functionality in this context . as shown in fig .
[ fig1](a ) , rutile has a tetragonal structure with ti@xmath13 ions occupying the body - center position , where each ti ion is coordinated by six ligand oxygen atoms ( o@xmath14 ) comprising a slightly distorted octahedron .
the tio@xmath15 octahedra are linked by corner - sharing to form o chains parallel to the @xmath4 axis .
regions of low electron density exist between the chains , forming the so - called @xmath16 channel " .
earlier studies have shown that interstitial h resides in the @xmath16 channel , leading to the formation of an oh bond ( with a length of 0.109 nm ) perpendicular to the @xmath4 axis @xcite .
the open channels cause anisotropy in the diffusion process , and may allow fast diffusion of smaller ions parallel to the tetragonal axis .
it has recently been reported that the h diffusion can be strongly enhanced via irradiation of the crystal by infrared light that matches the oh bond stretch mode @xcite .
crystal structure , showing a unit cell with @xmath17 nm and @xmath18 nm .
( b ) atomic position of interstitial hydrogen ( h ) , where h forms oh bonds ( 0.109 nm ) at four inequivalent h sites .
( c)(e ) signs of hyperfine parameters ( @xmath19 ) expected for the magnetic dipolar field ( @xmath20 ) generated by an electron spin localized on ti atom under an external field ( @xmath21 ) .
( see text ) . ] here , we report on the electronic structure of mu in single - crystalline rutile . although the existence of a mu state with an extremely small hyperfine ( hf ) parameter has been reported elsewhere @xcite , little is known about the local electronic structure . in this study , we show that the hf parameters exhibit strong anisotropy within the @xmath2 plane as well as along the @xmath4 axis .
the angular dependence of the hf parameters is perfectly in line with that expected for mu occupying the sites that are common to the interstitial h and forming omu bonds .
more interestingly , the relative signs of these parameters indicate that the hf interaction is predominantly determined by the magnetic dipolar fields generated by the unpaired mu electron located away from the muon site .
this is qualitatively consistent with a recent report on the h - related paramagnetic center , in which the hf interaction is attributed to the ti@xmath22[oh]@xmath23 complex state with the @xmath24 electron being located on the ti ion nearest to the oh base @xcite .
meanwhile , a detailed comparison of the mu and h hf structures reveals a distinct difference between the complex centers , where a contribution from the residual spin polarization at the nearest neighboring o site ( antiparallel to that of the ti ion ) is suggested in the mu case . through consideration of the temperature ( @xmath25 ) dependence of the mu yield ,
we also show that the unpaired electron bound to the mu complex state requires small energy for promotion to the conduction band . based on these observations
, we argue that interstitial mu / h is one of the primary origins of unintentional @xmath1-type conductivity in rutile tio@xmath0 .
this is despite its complex electronic structure , which differs significantly from that for the conventional effective - mass - like donor .
a conventional @xmath8sr experiment was conducted at the former booster meson facility in tsukuba ( at kek ) and at j - parc muse . a 4-mev muon beam that was almost 100% spin - polarized parallel to the beam direction
was implanted into single - crystalline rutile wafers [ 30 mm @xmath26 30 mm , 1.0-mm thick , grown using the verneuil process ] , which were as received " from a local vendor ( furuuchi chemical co. ) . the definitions of the azimuthal and polar angles ( @xmath27 , @xmath28 , respectively ) for the external magnetic field ( @xmath21 ) are shown in fig . [
fig1](a ) . for the @xmath27-dependence measurements ,
a wafer having a normal axis orientation of @xmath29 ( @xmath30 ) was used , where the initial muon spin polarization ( @xmath31 ) was parallel to @xmath4 , and the @xmath32 axis was rotated contrary to the @xmath21 direction within the @xmath2 plane ( i.e. , @xmath33 ) . meanwhile , the @xmath28-dependence was measured by tilting the @xmath21 direction towards the @xmath4 axis ( except for @xmath34 , for which another @xmath35 crystal was used ) , where the angle was tuned by an additional field parallel to the @xmath4 axis . in both cases , the @xmath21-parallel component of the hf parameter was observed in terms of the frequency shift of the satellite signals .
within the @xmath25 range of the present measurements ( 2 k @xmath36 290 k ) , a single diamagnetic state ( @xmath37 mu@xmath38 ) was observed above @xmath510 k as a signal precessing with a frequency @xmath39 ( comprising the central line ) , where @xmath40 ( @xmath41 mhz / t ) is the muon gyromagnetic ratio and @xmath42 .
the depolarization rate for this signal was almost independent of @xmath25 with a gaussian damping ( @xmath50.02 @xmath8s@xmath43 ) ; this was unambiguously attributed to random local fields exerted by the nuclear magnetic moments of the @xmath44ti ( natural abundance : 7.4% ) and @xmath45ti ( 5.4% ) nuclei . meanwhile , the @xmath8sr time spectra exhibited a remarkable change below @xmath510 k. a typical example ( observed at 5 k , @xmath46 mt ) is shown in fig .
[ fig2](a ) , where a beat pattern due to multiple frequency components is clearly observed in the signal amplitude .
sr time spectrum observed in tio@xmath0 at 5 k under an applied magnetic field of 32.5 mt , where the field ( @xmath21 ) is rotated from @xmath35 axis ( @xmath32 ) by @xmath47 around the @xmath4 axis ( @xmath33 ) .
solid curve shows a result of curve fit .
( b ) fast fourier - transform ( fft ) of the data shown in ( a ) .
( c ) fft of the time spectra with @xmath48 at 5 k. ] as is evident in fig .
[ fig2](b ) , one can trace two pairs of satellite lines in the fast fourier - transformed spectra ( @xmath49 , @xmath50 , 2 ) , which are situated almost symmetrically around the central line ( @xmath51 mhz ) .
a preliminary analysis showed that the least - square curve fits of the @xmath8sr time spectra , assuming one pair of satellites , did not reproduce the data for those obtained with @xmath21 oriented away from the @xmath35 direction , yielding a poor @xmath52 .
however , fits with two sets of satellites were found to yield a satisfactory result with significantly improved @xmath52 in the majority of cases .
it was also found that the splitting ( @xmath53 ) of these satellites remained unchanged when the applied field was reduced to 10 mt , indicating that @xmath54 was in the high field - limit range , where the splitting was independent of the @xmath54 value .
these observations clearly indicate that a significant fraction of implanted muons form a paramagnetic mu state characterized by an extremely small hf parameter ( almost 10@xmath11 times smaller than the vacuum value of @xmath55 mhz ) with significant anisotropy in the electronic structure .
the results of curve fits for the time spectra assuming multiple precession frequency components ( @xmath49 ) are summarized in fig .
[ fig3 ] , for both ( a ) @xmath27 and ( b ) @xmath28 dependence .
( we assumed that the signal amplitudes of @xmath56 and @xmath57 were common for each pair in the curve fits . ) as can be clearly seen in fig .
[ fig3](a ) , @xmath49 exhibits a strong anisotropy with a characteristic angle dependence that is symmetric over @xmath58 .
considering the fact that there are four inequivalent h ( mu ) sites with different principal axis orientations for anisotropic hf parameters [ see fig .
[ fig1](b ) ] , four sets of satellite pairs discerned by different @xmath27-dependence can be expected .
more specifically , it is predicted that the @xmath27 dependence of the hf parameters can be expressed in the form @xmath59 where @xmath60 or @xmath61 , and @xmath62 is the offset angle from the @xmath32 axis .
it must be noted that the double sign correspondence between @xmath49 and the right hand side of eq .
( [ mufrq ] ) is not unique , because the sign of @xmath63 depends on that of @xmath19 and , thus , can change over the @xmath27 range . while our measurements are insensitive to the absolute sign of @xmath19 , they can determine the relative sign between @xmath64 and @xmath65 .
the solid curves shown in fig . [ fig3](a ) are the result of a simultaneous fit for those curves using common values of @xmath19 and @xmath62 , where @xmath66 is inferred .
this strongly suggests that the hf interaction is predominantly determined by the magnetic dipolar field generated by the off - site electron spin(s ) .
( this scenario closely resembles the case of mu - substituted free radicals often observed in unsaturated organic compounds . ) moreover , considering that eqs .
( [ mufrq ] ) and ( [ aperp ] ) predict eight lines when @xmath67 , fig .
[ fig3](a ) indicates that @xmath68 is almost equal to @xmath69 within the frequency resolution , which is limited by the observation time range ( @xmath50.1 mhz ) .
the obtained values for @xmath19 and @xmath27 are shown in table [ tab1 ] .
hereafter , we assume that @xmath70 ( see below ) .
the @xmath28 dependence is shown in fig .
[ fig3](b ) , and is characterized by a narrow splitting of @xmath71 at @xmath34 [ see fig .
[ fig2](c ) ] , along with a monotonous increase with increasing @xmath28 .
only one pair of satellites has been identified in this data set , most likely because of the narrow splitting relative to the frequency resolution .
it is also notable that @xmath72 tends to be smaller than @xmath73 for some unknown reason .
we tentatively attribute this to systematic uncertainty in the present measurement , and adopt @xmath74 in order to deduce the hf parameters using a form similar to eq .
( [ aperp ] ) , such that @xmath75 where the sign of @xmath76 is left undetermined .
the value of @xmath76 deduced from the curve fit is shown in table i , where @xmath77 has been determined to be 1.314(15 ) mhz ( which is in perfect agreement with the value deduced from the @xmath27 dependence ) . ) and ( b ) polar angle ( @xmath28 ) .
filled symbols : @xmath49 satellite lines , open symbols : @xmath78 central line .
the solid curves are obtained via curve fits using general forms of the anisotropic hyperfine parameters , while the dashed curves are calculated for a given set of parameters ( see text ) . ] .mu hyperfine parameters deduced from curve fits of data shown in fig .
[ fig3 ] ( where @xmath79 was assumed ) .
the @xmath19 s for the h - related center @xcite are scaled against the gyromagnetic ratio of @xmath80 ( indicated by an asterisk , see text ) .
the @xmath19 signs for mu are assumed to be common to those for h. [ tab1 ] [ cols="^,^,^,^,^",options="header " , ] we currently speculate that the difference in the electronic structure between mu and h is primarily due to the local environment of the specimen . while the present mu complex state was observed in an as received " crystal , the h - related paramagnetic state was observed after a reduction process ( annealing at 600@xmath81 c for 10 min in a n@xmath0 atmosphere @xcite ) .
considering the fact that the reduction process also yields o vacancies , it may be of interest to recall the recent investigation based on density - functional theory that suggested that h is more stable at the o vacancy site , where it adopts a negatively charged state to form a ti@xmath22h@xmath82 complex state @xcite .
thus , it is natural to consider the possibility that the paramagnetic state observed via epr / endor might correspond to such a state , although the consistency with experimental results must be carefully examined .
the fact that the relative magnitudes of the @xmath19 s for the h - related center correspond to the case of @xmath83 in eq .
( [ dipo ] ) ( i.e. , @xmath64:@xmath65:@xmath76 = @xmath84:@xmath85:@xmath84 ) seems to constitute supporting evidence , as it suggests that the relevant h atom has no neighboring o atom . the fact that the hyperfine interaction is dominated by a magnetic dipolar interaction also places a strong constraint on the diffusion of @xmath80 in the @xmath16 channel , because the @xmath19 are strongly dependent on the distance to the electron ( @xmath86 ) .
the diffusive motion of @xmath80 against a stationary @xmath24 electron at the ti site would immediately lead to the strong damping of satellite signals due to the fluctuation of @xmath87 and hence of @xmath19 .
the @xmath25 dependence of the fractional yield ( @xmath88 ) for the mu complex state is shown in fig .
the sum of the signal amplitude between the mu complex state and the diamagnetic state ( not shown ) is almost independent of @xmath25 , suggesting that the mu complex is converted to a diamagnetic state above @xmath510 k. here , it must be noted that @xmath89 ( @xmath90 ) does not necessarily indicate two different muon sites with different charge states . in non - metallic compounds , it is usually expected that the initial yield of the paramagnetic state upon muon implantation is predominantly determined by the density of the short - lived free electrons that are produced via radiolysis near the end of the muon radiation track .
while this density varies between compounds , it is independent of @xmath25 because the radiolysis is an athermal process .
we assume that the muon site in tio@xmath0 is unique ( as shown in fig .
[ fig1]b ) over the entire observed @xmath25 range , where the initial yield of the mu complex is controlled by the athermal electron density , which is independent of @xmath25 . in general , the diamagnetic state can be either positively or negatively charged .
if the electronic energy levels ( @xmath91 , @xmath92 ) associated with the mu complex are situated near the top of the valence band , a mu@xmath23 state can be expected at lower @xmath25 because the fermi level ( @xmath93 ) is likely to be situated far above the mid - gap in the present @xmath1-type specimen ( i.e. , @xmath94 ) . in that case ,
an _ increase _ in the yield of the mu complex would be observed with elevating temperature due to the hole capture process ( @xmath95 ) ; however , this is contrary to the actual observed behavior .
thus , the behavior of the mu complex strongly suggests that the process relevant to the promotion of the diamagnetic state is electron release , @xmath96 ( i.e. , @xmath97 ) , and that the mu complex state can serve as an electron donor .
provided that the ionization of the mu complex state is driven by an arrhenius - type activation process , the disappearance of the mu signal above @xmath510 k suggests that the activation energy ( @xmath98 ) is of the order of @xmath6 mev .
this is in line with certain earlier reports suggesting unidentified shallow level states ( e.g , via optical absorption @xcite or infrared absorption spectroscopy on deuterated rutile@xcite ) . in any case
, it must be remembered that the interpretation of @xmath98 depends on the detailed neutral charge state formation process of the mu complex , and also on the kind of equilibrium state realized for the mu complex formation . at one end of the range
, @xmath98 represents a direct transition from the defect level to the bottom of the conduction band , while at the other extreme , in equilibrium , it indicates a transition from the defect level to the fermi level ( i.e. , @xmath99 ) .
since the origin of the @xmath1-type conductivity in the present specimen is unknown , the present value of @xmath98 should be interpreted as a lower bound for the actual defect level .
we have demonstrated that the electronic structure of the interstitial mu center in rutile is characterized by extremely small and highly anisotropic hyperfine parameters .
these parameters are predominantly determined by magnetic dipolar interaction with the unpaired mu electron , which is primarily situated at the ti site .
the hyperfine parameters are quantitatively explained by a model that considers a small residual spin polarization of the o atom ( which is antiparallel to that of the ti atom ) , suggesting that the electronic structure should be interpreted as being a ti - o - mu complex state .
the extremely small effective moment size of the unpaired electron ( @xmath50.05@xmath100 at the ti site , @xmath50.0039@xmath100 at the o site ) as well as the small activation energy required for its promotion to the conduction band , implies that the mu complex ( and hence the corresponding h state ) can serve as an electron donor .
this strongly suggests that h is one of the primary origins of unintentional @xmath1-type conductivity in rutile tio@xmath0 .
we would like to thank the kek - msl staff for their technical support during the muon experiment .
we would also like to express our gratitude to k. yoshizawa , y. iwazaki , and s. tsuneyuki for helpful discussion and for providing us with the result of their first - principle calculation for mu / h in tio@xmath0 prior to publication .
this work was partially supported by the kek - msl inter - university research program ( 2012a0110 , 2012b0032 ) .
s. m. myers , m. i. baskes , h. k. birnbaum , j. w. corbett , g. g. deleo , s. k. estreicher , e. e. haller , p. jena , n. m. johnson , r. kirchheim , s. j. pearton , and m. j. stavola , rev .
phys . * 64 * , 559 ( 1992 ) .
k. h. chow , b. hitti , and r. f. kiefl , in _
@xmath8sr on muonium in semiconductor and its relation to hydrogen _ , ed . by m.
j. stavola , _ semiconductors and semimetals _ , vol . 51 ( academic , new york , 1998 ) , p. 137 . s. f. j. cox , e. a. davis , s. p. cottrell , p. j. c. king , j. s. lord , j. m. gil , h. v. alberto , r. c. vilo , j. piroto duarte , n. ayres de campos , a. weidinger , r. l. lichti and s. j. c. irvine , phys . rev .
lett . * 86 * , 2601 ( 2001 ) .
j. m. gil , h. v. alberto , r. c. vilo , j. piroto duarte , p. j. mendes , l. p. foreira , n. ayres de campos , a. weidinger , j. krause , ch .
niedermayer , and s. f. j. cox , phys .
lett . * 83 * , 5294 ( 1999 ) .
s. f. cox , j. l. gavartin , j. s. lord , s. p. cottrell , j. m. gil , h. v. alberto , j. piroto duarte , r. c. vilo , n. ayres de campos , d. j. keeble , e. a. davis , m. charlton , and d. p. van der werf , j. phys . : condens
. matter , * 18 * , 1079 ( 2006 ) . | the hyperfine structure of the interstitial muonium ( mu ) in rutile ( tio@xmath0 , weakly @xmath1-type ) has been identified by means of a muon spin rotation technique .
the angle - resolved hyperfine parameters exhibit a tetragonal anisotropy within the @xmath2 plane and axial anisotropy with respect to the @xmath3 ( @xmath4 ) axis .
this strongly suggests that the mu is bound to o ( forming an oh bond ) at an off - center site within a channel along the @xmath4 axis , while the unpaired mu electron is localized around the neighboring ti site .
the hyperfine parameters are quantitatively explained by a model that considers spin polarization of the unpaired electron at both the ti and o sites , providing evidence for the formation of mu as a ti - o - mu complex paramagnetic state .
the disappearance of the mu signal above @xmath510 k suggests that the energy necessary for the promotion of the unpaired electron to the conduction band by thermal activation is of the order of @xmath6 mev .
these observations suggest that , while the electronic structure of mu ( and hence h ) differs considerably from that of the conventional shallow level donor described by the effective mass model , mu supplies a loosely bound electron , and thus , serves as a donor in rutile . |
in the study of knots and links in @xmath4-dimensional spaces ( such as handlebodies , knot complements , c.c.o .
@xmath4-manifolds ) it can prove very useful to take an approach via braids , as the use of braids provides more structure and more control on the braid equivalence moves
. in @xcite geometric braid equivalence has been given for isotopic knots and links in knot complements and in c.c.o .
@xmath4-manifolds with integral surgery description .
this was done by fixing ( pointwise ) a description for the knot complement or the @xmath4-manifold via a closed braid @xmath5 in @xmath1 .
then , knots and links in such a @xmath4-manifold are represented unambiguously by _ mixed links _ in @xmath1 ( figure [ a mixed link ] ) and braids in the manifold are represented unambiguously by geometric _ mixed braids _ in @xmath1 .
a geometric mixed braid is a braid in @xmath1 that contains @xmath6 as a fixed subbraid ( figure [ mixedbraidslmoves ] ) .
isotopy for links in a c.c.o .
3-manifold @xmath0 is then translated into appropriate moves for mixed links in @xmath1 , which comprise isotopy in the complement @xmath7 together with the _ band moves _ coming from the handle sliding moves in @xmath0 , and are related to the surgery description of @xmath0 ( figure [ band moves ] ) . for the braid equivalence , the authors sharpened first the classic markov theorem giving only one type of moves , the _ @xmath2-moves _ ( figure [ mixedbraidslmoves ] ) , which are geometric as well as algebraic .
then , with the use of the @xmath2-moves and the _ braid band moves _ ( figure [ pbbm ] ) they formulated geometric mixed braid equivalence for knots and links in knot complements and in c.c.o .
@xmath4-manifolds obtained from @xmath1 by integral surgery .
further , in @xcite the same authors provided algebraic formulations for the geometric mixed braid equivalences in knot complements and in c.c.o .
@xmath4-manifolds , using the _ mixed braid groups _
@xmath3 ( eq . [ b ] and
figure [ loops and crossings ] ) , introduced and studied in @xcite , and the techniques of parting and combing mixed braids ( figure [ partingandcombing ] ) .
_ parting _ a geometric mixed braid means to separate its strands into two sets : the strands of the fixed subbraid @xmath6 and the moving strands of the braid representing a link in the @xmath4-manifold .
_ combing _ a parted mixed braid means to separate the braiding of the fixed subbraid @xmath6 from the braiding of the moving strands ( figure [ combing ] ) .
the above techniques have been also applied in @xcite for links in a handlebody .
integral surgery is a special case of rational surgery and , by a classic result of topology , every c.c.o .
@xmath4-manifold can be obtained by surgery ( integral or rational ) along a framed link in @xmath1 . moreover , in the case of integral surgery the components of the framed link may be all assumed to be simple closed curves , giving rise to a closed pure braid .
there are @xmath4-manifolds which have simpler description when obtained from @xmath1 by rational surgery .
representative examples are the lens spaces @xmath8 which with rational surgery description @xmath9 are obtained from the trivial knot , while with integral surgery description a non - trivial link is needed , see for example @xcite p.322 . a simpler surgery description of a @xmath4-manifold @xmath0 is expected to induce simpler algebraic expressions for the braid equivalence in @xmath0 .
as an example , compare section 4 in @xcite with section 5.1 in this paper for the case of lens spaces . in this paper
the braid equivalence is in the mixed braid groups @xmath10 and there is only one expression for the braid band moves . in @xcite
there are many , according to the surgery coefficient of each strand of the surgery pure braid and , on top of that , combing is also needed .
the purpose of this paper is to provide mixed braid equivalence , geometric as well as algebraic , for isotopic oriented links in c.c.o .
@xmath4-manifolds obtained by rational surgery along framed links in @xmath1 .
we follow the setting and the techniques of @xcite and we use the results therein .
more precisely , let @xmath0 be a c.c.o .
@xmath4-manifold obtained by rational surgery along a framed link @xmath5 in @xmath1 .
note that the surgery braid @xmath6 may not be assumed to be a pure braid .
let @xmath11 be a surgery component of @xmath6 with surgery description @xmath9 consisting of @xmath12 strands , @xmath13 .
when a _ geometric braid band move _ on @xmath11 occurs , @xmath12 sets of @xmath14 new strands appear , each one running in parallel to a strand of @xmath11 , and also a @xmath15-torus braid wraps around the last strand , @xmath16 , @xmath17 times ( see figure [ homtref ] for an example ) .
these moves together with the @xmath2-moves lead to the _ geometric mixed braid equivalence _ in @xmath0 ( theorem [ geommarkovrat ] ) .
the geometric braid band moves are much more complicated than in the case of integral surgery @xcite .
however , a sharpened version of the reidemeister theorem for links in @xmath0 ( theorem [ reidemrational ] ; see also @xcite ) , whereby only one type of band moves is used in the isotopy equivalence ( figure [ bmoves ] ) , makes the proof of theorem [ geommarkovrat ] lighter . in order to move toward algebraic statements
we use the notion of a @xmath14-strand cable and we apply the techniques and results from @xcite .
a _ @xmath14-strand cable _ encloses a set of @xmath14 new strands arising from the performance of a geometric braid band move .
we show first that parting a @xmath14-strand cable is equivalent to parting each strand of the cable one by one ; that is , parting and cabling commute ( figure [ stndpartcable ] ) .
treating now each one of the @xmath12 @xmath14-strand cables as one thickened strand leads to the _ parted mixed braid equivalence _ ( theorem [ qparted ] ) , assuming the corresponding result in @xcite .
we continue by providing algebraic expressions for parted cables ( lemma [ cableloop ] ) and also for the loopings between the strands of @xmath6 and the remaining strands of the mixed braid after a braid band move is performed ( figures [ cablesa ] , [ cablesa2 ] ) .
then , we part locally the @xmath15-torus braid and the crossing of a parted mixed braid band move ( figures [ cables2 ] , [ cablesc ] ) , obtaining algebraic expressions for these parts . from the above we obtain the algebraic expression of an _ algebraic braid band move_. finally , we do combing through @xmath6 and
we show that combing a @xmath14-strand cable is equivalent to combing each strand of the cable one by one ; that is , combing and cabling commute ( figures [ combcable1 ] and [ combcableproof2 ] ) .
so , assuming the corresponding result in @xcite , we obtain the _ algebraic mixed braid equivalence _ for links in @xmath0 in terms of the mixed braid groups @xmath3 , which is our main result ( theorem [ algmarkov ] ) .
the paper is organized as follows . in section 1
we recall the setting and the essential techniques and results from @xcite . in section
2 we prove the sharpened version of the reidemeister theorem for knots and links in c.c.o .
@xmath4-manifolds with rational surgery description . in section
3 we give the geometric braid equivalence for links in such @xmath4-manifolds ( theorem [ geommarkovrat ] ) and in section 4 we give the corresponding algebraic version ( theorem [ algmarkov ] ) .
finally , in section 5 we give the concrete algebraic expressions for the mixed braid equivalences in lens spaces , in seifert manifolds and in homology spheres obtained from @xmath1 by rational surgery along the trefoil knot .
also , we provide some figures ( figures [ seifcomb1][seifcomb6 ] ) illustrating in a generic example all the techniques developed and used in this paper step - by - step . the algebraic mixed braid equivalence gives good control over the band moves , so our results can be applied for the study of skein modules of c.c.o .
@xmath4-manifolds obtained by rational surgery along a framed link in @xmath1 and for constructing invariants of links in such @xmath4-manifolds , using algebraic means . for a construction of the analogues of the homflypt polynomial for knots and links in the solid torus via this approach , the reader is referred to @xcite . in a sequel paper
we shall use theorem [ algmarkov ] and the results of @xcite in order to compute the homflypt skein module of the lens spaces @xmath8 with this method .
in this section we recall the topological settings and results from [ lr1 ] and [ lr2 ] for expressing link isotopy in a c.c.o .
@xmath4-manifold @xmath0 in terms of geometric and algebraic mixed braid equivalence in @xmath1 . for the rest of this section
we fix a c.c.o .
@xmath4-manifold @xmath0 , which is obtained by integral surgery on a framed link in @xmath1 given in the form of a closed braid @xmath5 .
we shall denote @xmath18 .
let @xmath2 be an oriented link in @xmath0 .
fixing @xmath5 pointwise , @xmath2 can be represented unambiguously by a _ mixed link _ in @xmath1 denoted @xmath19 , that is , a link in @xmath20 consisting of the _ fixed part _
@xmath5 and the _ moving part _ @xmath2 that links with @xmath5 .
mixed link diagram _ is a diagram @xmath21 of @xmath19 on the plane of @xmath5 , where this plane is equipped with the top - to - bottom direction of the braid @xmath6 , see figure [ a mixed link ] for an example . .
, width=124 ] an isotopy of @xmath2 in @xmath0 can be translated into a finite sequence of moves of the mixed link @xmath22 in @xmath1 as follows .
as we know , surgery along @xmath5 is realized by taking first the complement @xmath7 and then attaching to it solid tori according to the surgery description .
thus , isotopy in @xmath0 can be viewed as certain moves in @xmath1 , namely , isotopy in @xmath7 together with the band moves in @xmath1 , which are similar to the second kirby move .
isotopy in @xmath7 is realized by the classical reidemeister moves and planar moves for the moving part together with the _ extended reidemeister moves_. these are the reidemeister ii and iii moves involving the fixed and the moving part of the mixed link ( cf . definition 5.1 @xcite ) .
band move _ is a non - isotopy move in @xmath23 that reflects isotopy in @xmath0 and is the band connected sum of a component , say @xmath11 , of @xmath2 with the specified ( from the framing ) parallel curve @xmath24 of a surgery component , say @xmath25 , of @xmath5 .
note that @xmath24 bounds a disc in @xmath0 .
there are two types of band moves according to the orientations of the component @xmath11 of @xmath2 and of the surgery curve @xmath25 , as illustrated and exemplified in figure [ band moves ] . in the @xmath26-type
the orientation of @xmath11 is opposite to the orientation of @xmath25 ( and of its parallel curve @xmath24 ) , but after the performance of the move their orientations agree . in the @xmath27-type
the orientation of @xmath11 agrees initially with the orientation of @xmath25 , but disagrees after the performance of the move .
note that the two types of band moves are related by a twist of @xmath11 ( reidemeister i move in @xmath7 ) .
-band moves .
, width=528 ] the above are summarized in the following analogue of the reidemeister theorem for oriented links in @xmath0 .
[ reidemintegral ] two oriented links @xmath28 , @xmath29 in @xmath18 are isotopic if and only if any two corresponding mixed link diagrams of theirs , @xmath30 and @xmath31 , differ by isotopy in @xmath23 together with a finite sequence of the two types @xmath26 and @xmath27 of band moves . in order to translate isotopy of links in the @xmath4-manifold @xmath0 to braid equivalence , we need to introduce the notion of a geometric mixed braid .
a _ geometric mixed braid _ related to @xmath18 and to a link @xmath32 in @xmath0 , is an element of the group @xmath33 , where @xmath34 strands form the fixed _ surgery braid _ @xmath6 and @xmath35 strands form the _ moving subbraid _ @xmath27 representing the link @xmath32 in @xmath0 . for an illustration see the middle picture of figure [ mixedbraidslmoves ] .
we further need the notions of the @xmath2-moves and the braid band moves .
[ @xmath2-moves and @xmath36-braid band moves , definitions 2.1 and 5.6 @xcite ] _ ( i ) _ let @xmath37 be a geometric mixed braid in @xmath1 and @xmath38 a point of an arc of the moving subbraid @xmath27 , such that @xmath38 is not vertically aligned with any crossing or endpoint of a braid strand . doing an _ @xmath2-move
_ at @xmath38 means to cut the arc at @xmath38 , to bend the two resulting smaller arcs slightly apart by a small isotopy and to stretch them vertically , the upper downward and the lower upward , and both _ over _ or _ under _ all other arcs of the diagram , so as to introduce two new corresponding moving strands with endpoints on the vertical line of the point @xmath38 .
stretching the new strands over will give rise to an _
@xmath39-move _ and under to an _
@xmath40-move_. for an illustration see figure [ mixedbraidslmoves ] .
two geometric mixed braids shall be called _
@xmath2-equivalent _ if and only if they differ by a sequence of @xmath2-moves and braid isotopy . _
( ii ) _ a _ geometric @xmath36-braid band move _ is a move between geometric mixed braids which is a band move between their closures .
it starts with a little band oriented downward , which , before sliding along a surgery strand , gets one twist _
positive _ or _ negative _ ( see figure [ pbbm ] ( a ) and ( b ) ) .
, width=480 ] -braid band move ( top part of ( d ) ) .
, width=518 ] [ lmove ] _
( i ) _ in @xcite it is shown that classical braid equivalence in @xmath1 is generated only by the @xmath2-moves .
this implies that braid conjugation and in particular change of the order of the endpoints of a braid can be realized by @xmath2-moves .
a demonstration can be found in @xcite figure 14 . + _
( ii ) _ a geometric @xmath36-braid band move may be always assumed , up to @xmath2-equivalence , to take place at the top part of a mixed braid and on the right of the specific surgery strand ( @xcite lemma 5 ) . in [ lr1 ]
the following theorem was proved for isotopic links in @xmath18 .
[ geommarkov ] two oriented links in @xmath18 are isotopic if and only if any two corresponding geometric mixed braids in @xmath1 differ by mixed braid isotopy , by @xmath2-moves that do not touch the fixed subbraid @xmath6 and by the geometric @xmath36-braid band moves .
let @xmath18 .
we will pass from the geometric braid equivalence to an algebraic statement for links in @xmath0 .
algebraic mixed braid _ is a mixed braid on @xmath41 strands such that the first @xmath34 strands are fixed and form the identity braid on @xmath34 strands and the next @xmath35 strands are moving strands and represent a link in the manifold @xmath0 .
the set of all algebraic mixed braids on @xmath41 strands forms a subgroup of @xmath33 , denoted @xmath3 , and called _ mixed braid group_. in [ la2 ] the mixed braid groups @xmath3 have been introduced and studied and it is shown that @xmath3 has the presentation : @xmath42 where the _ loop generators _ @xmath43 and the _ braiding generators _ @xmath44 are as illustrated in figure [ loops and crossings ] . and the braiding generators @xmath44 of @xmath3 .
, width=528 ] in order to give an algebraic statement for braid equivalence in @xmath0 , we first part the mixed braids and we translate the geometric @xmath2-equivalence of theorem [ geommarkov ] to an equivalence of parted mixed braids . _
parting _ a geometric mixed braid @xmath45 on @xmath41 strands means to separate its endpoints into two different sets , the first @xmath34 belonging to the subbraid @xmath6 and the last @xmath35 to @xmath27 , and so that the resulting braids have isotopic closures .
this is realized by pulling each pair of corresponding moving strands to the right and _ over _ or _ under _ each strand of @xmath6 that lies on their right .
we start from the rightmost pair respecting the position of the endpoints .
this process is called _ parting _ of a geometric mixed braid and the result is a _ parted mixed braid_. if the strands are pulled always over the strands of @xmath6 , then this parting is called _
standard parting_. see the middle illustration of figure [ partingandcombing ] for the standard parting of an abstract mixed braid . for more details the reader
is referred to @xcite .
then , in order to restrict theorem [ geommarkov ] to the set of all parted mixed braids related to the manifold @xmath0 , we need the following moves between parted mixed braids .
_ loop conjugation _ of a parted mixed braid @xmath27 is its concatenation by a loop @xmath43 ( or by @xmath46 ) from above and from @xmath46 ( corr .
@xmath43 ) from below , that is @xmath47 .
as it turns out , two partings of a geometric mixed braid differ by loop conjugations ( cf .
lemma 2 @xcite ) .
a _ parted @xmath2-move _ is an @xmath2-move between parted mixed braids .
further , a mixed braid with an @xmath2-move performed can be parted to a parted mixed braid with a parted @xmath2-move performed .
namely we make the parting consistent with the label of the @xmath2-move : an @xmath39 move will be parted by pulling over all other strands , while an @xmath40 move will be parted by pulling under all other strands ( cf .
lemma 3 @xcite ) .
a _ parted @xmath36-braid band move _ is a geometric @xmath36-braid band move between parted mixed braids , such that it takes place at the top part of the braid and the little band _ starts from the last strand _ of the moving subbraid and it _ moves over _ each moving strand and each component of the surgery braid , until it reaches from the right the specific component , and then is followed by parting ( see figure [ pbbm](c ) ) .
moreover , performing a @xmath36-braid band move on a mixed braid and then parting , the result is equivalent , up to @xmath2-moves and loop conjugation , to performing a parted @xmath36-braid band move ( cf .
lemma 5 @xcite ) .
[ partedcco ] two oriented links in @xmath48 are isotopic if and only if any two corresponding parted mixed braids differ by a finite sequence of parted mixed braid isotopies , parted @xmath2-moves , loop conjugations and parted @xmath36-braid band moves .
we now comb the parted mixed braids in order to translate the parted mixed braid equivalence to an equivalence between algebraic mixed braids .
_ combing _ a parted mixed braid means to separate the knotting and linking of the moving part away from the fixed subbraid using mixed braid isotopy .
more precisely , let @xmath49 denote the crossing between the @xmath50 and the @xmath51 strand of the fixed subbraid .
then , for all @xmath52 and @xmath53 we have : @xmath54 .
thus , the only generating elements of the moving part that are affected by the combing are the loops @xmath43 .
this is illustrated in figure [ combing ] . in lemma
6 @xcite formuli are given for the effect of combing on the @xmath43 s ( see lemma [ combinglem ] below ) .
the effect of combing a parted mixed braid is to separate it into two distinct parts : the _ algebraic part _ at the top , which has all fixed strands forming the identity braid , so is an element of some mixed braid group @xmath3 , and which contains all the knotting and linking information of the link @xmath2 in @xmath0 ; the _ coset part _ at the bottom , which contains only the fixed subbraid @xmath6 and an identity braid for the moving part ( see right hand most illustration in figure [ partingandcombing ] ) .
let now @xmath55 denote the set of parted mixed braids on @xmath35 moving strands with fixed subbraid @xmath6 .
concatenating two elements of @xmath55 is not a closed operation since it alters the braid description of the manifold .
however , as a result of the combing , for the fixed subbraid @xmath6 the set @xmath55 is a coset of @xmath3 in @xmath33 .
fore details on the above the reader is referred to @xcite .
translating the parted braid equivalence into an equivalence between algebraic mixed braids , we will obtain an algebraic statement of theorem [ partedcco ] .
since loop conjugation does not take into account the combing of the loop through the fixed subbraid , we need the notion of _ combed loop conjugation_. a combed loop conjugation is a move between algebraic mixed braids and is the result of a loop conjugation on a combed mixed braid followed by combing , so it can be described algebraically as : @xmath56 for @xmath57 , where @xmath58 is the combing of the loop @xmath43 through the fixed subbraid @xmath6 .
we also define _ algebraic @xmath0-conjugation _ of an algebraic mixed braid to be its conjugation by a crossing @xmath44 ( or by @xmath59 ) .
an _ algebraic @xmath0-move _ is defined to be the insertion of a crossing @xmath60 on the right hand side of an algebraic mixed braid .
finally , an _
algebraic @xmath2-move _ is defined to be a @xmath2-move between algebraic mixed braids .
an algebraic @xmath2-move has the following algebraic expression for an @xmath39-move and an @xmath40-move respectively : @xmath61 where @xmath62 , @xmath63 are elements of @xmath3 and @xmath64 , @xmath65 are obtained from @xmath62 , @xmath63 by replacing each @xmath44 by @xmath66 for @xmath67 .
note that algebraic @xmath0-conjugation , the algebraic @xmath0-moves and the algebraic @xmath2-moves commute with combing . note also that remark [ lmove](i ) applies equally to the case of algebraic mixed braids ( cf .
lemma 4 @xcite ) .
we finally need to understand how a parted @xmath36-braid band move is combed through the surgery braid @xmath6 .
[ twistband ] an _ algebraic @xmath36-braid band move _ is defined to be a parted band move between algebraic mixed braids ( see top part of figure [ pbbm](d ) ) .
setting : @xmath68 an algebraic band move has the following algebraic expression : @xmath69 where @xmath70 and @xmath71 are the words @xmath72 respectively with the substitutions : @xmath73}^{\pm 1 } \\ { a^{\pm 1}_i } & \longleftrightarrow & ( { \lambda^{-1}_{n-1,1 } } { \sigma^{2}_n } \lambda_{n-1,1 } ) \ , { a^{\pm 1}_i } \ , ( { \lambda^{-1}_{n-1,1 } } { \sigma^{2}_n } \lambda^{-1}_{n-1,1 } ) , \mbox { \ \ if \ } i < k \\ { a^{\pm 1}_i } & \longleftrightarrow & { a^{\pm 1}_i } , \mbox { \ \ if \ }
\\ \end{array}\ ] ] further , a _ combed algebraic @xmath36-braid band move _ is a move between algebraic mixed braids and is defined to be a parted @xmath36-braid band move that has been combed through @xmath6 .
so it is the composition of an algebraic @xmath36-braid band move with the combing of the parallel strand and it has the following algebraic expression : @xmath74 where @xmath75 is the combing of the parted parallel strand to the @xmath50 surgery strand through the surgery braid .
the group @xmath3 embeds naturally into the group @xmath76 .
we shall denote @xmath77 and similarly @xmath78 .
we are now in position to give the algebraic markov theorem for @xmath48 .
[ algcco ] two oriented links in @xmath48 are isotopic if and only if any two corresponding algebraic mixed braid representatives in @xmath79 differ by a finite sequence of the following moves : \(1 ) algebraic @xmath0-moves : @xmath80 , \(2 ) algebraic @xmath0-conjugation : @xmath81 , for @xmath82 , \(3 ) combed loop conjugation : @xmath83 , for @xmath57 , where @xmath58 is the combing of the loop @xmath43 through @xmath6 , \(4 ) combed algebraic braid band moves : for for every @xmath84 we have : @xmath74 where @xmath70 and @xmath85 are as in definition [ twistband ] and where @xmath75 is the combing of the parted parallel strand to the @xmath12th surgery strand through @xmath6 .
equivalently , by a finite sequence of algebraic mixed braid relations and the following moves : ( 1@xmath86 ) algebraic @xmath2-moves , ( 2@xmath86 ) combed loop conjugations , ( 3@xmath86 ) combed algebraic @xmath36-braid band moves .
let @xmath0 be a c.c.o .
@xmath4-manifold obtained from @xmath1 by rational surgery , that is surgery along a framed link @xmath5 with rational coefficients .
we shall denote @xmath87 .
let also @xmath2 be an oriented link in @xmath0 . by the discussion in section 1.1 , isotopy in @xmath0
is translated into isotopy in @xmath7 together with the two types , @xmath26 and @xmath27 , of band moves for mixed links in @xmath1 .
the band moves in this case are described as follows .
let @xmath25 be a component of @xmath5 with framing @xmath9 .
the specified parallel curve @xmath24 of @xmath25 is a @xmath15-torus knot on the boundary of a tubular neighborhood of @xmath25 which , by construction , bounds a disc in @xmath0 .
then , a _ @xmath88-band move _ along @xmath25 is the connected sum of a component of @xmath2 with the @xmath15-torus knot @xmath24 and there are two types , @xmath26 and @xmath27 , according to the orientations .
the two types of band moves are illustrated in figure [ bmoves ] , where @xmath25 is a trefoil knot with @xmath89 surgery coefficient and where band move " is shortened to b.m . " .
-band moves .
, width=345 ] clearly , theorem [ reidemintegral ] applies also to @xmath87 . namely we have : [ reidemrationalboth ] two oriented links @xmath28 , @xmath29 in @xmath87 are isotopic if and only if any two corresponding mixed link diagrams of theirs , @xmath30 and @xmath31 , differ by isotopy in @xmath23 together with a finite sequence of the two types @xmath26 and @xmath27 of band moves . in this section
we sharpen theorem [ reidemrationalboth ] .
more precisely , we show that only one of the two types of band moves is necessary in order to describe isotopy for links in @xmath0 . the proof is based on a known contrivance , which was used in the proof of theorem 5.10 @xcite ( recall theorem [ geommarkov ] ) for establishing the sufficiency of the geometric braid band moves in the mixed braid equivalence where the case of integral surgery only is considered ( see figure [ unknotted surgery ] ) .
theorem [ reidemrational ] simplifies the proof of theorem [ geommarkovrat ] .
[ reidemeister for @xmath87 with one type of band moves ] [ reidemrational ] two oriented links @xmath28 , @xmath29 in @xmath87 are isotopic if and only if any two corresponding mixed link diagrams of theirs , @xmath90 and @xmath91 , differ by a finite sequence of the band moves of type @xmath26 ( or equivalently of type @xmath27 ) and isotopy in @xmath7 .
let @xmath2 be an oriented link in @xmath0 . by theorem [ reidemrationalboth ]
, it suffices to show that a band move of type @xmath27 can be obtained from a band move of type @xmath26 and isotopy in the knot complement .
we will first demonstrate the proof for an unknotted surgery component @xmath25 with integral coefficient @xmath17 .
( note that integral surgery description can be considered as a special case of rational surgery description . )
we shall follow the steps of the proof in figure [ unknotted surgery ] where @xmath92 .
we start with performing a band move of type @xmath27 using a component @xmath11 of the link @xmath2 . in figure
[ unknotted surgery](b ) we see the two twists of the band move wrapping around the surgery curve @xmath25 in the righthand sense . then , using an arc of the same link component @xmath11 , we perform a second band move of type @xmath26 .
this will take place within a thinner tubular neighborhood than the first band move .
so , the two twists of the second band move , which also wrap around @xmath25 in the righthand sense , commute with the two twists of the first band move ( see figure [ unknotted surgery](c ) ) .
we arrange all @xmath93 twists in pairs as follows .
we pass one twist from the second band move ( the closest ) through all twists of the first band move , see figure [ unknotted surgery](f ) .
since all twists follow the righthand sense , the two innermost twists coming from the second and the first band move , create a little band ( figure [ unknotted surgery](f ) ) which can be eliminated using isotopy in the knot complement of @xmath25 ( figure [ unknotted surgery](g ) and figure [ twist cancelation ] ) .
this is the cancelation of the first pair of the @xmath93 twists . repeating the same procedure
we cancel all @xmath17 pairs and we end up with the component @xmath11 of the link @xmath2 as it was in the initial position before the band moves ( see figures [ unknotted surgery](i ) , ( a ) ) .
band move follows from a type-@xmath26 band move in the case of integral surgery coefficient .
, width=566 ] for the more general case of rational surgery along any knot @xmath25 we follow the same idea .
more precisely , we perform a @xmath88-band move of type @xmath27 along @xmath25 and we obtain an outer @xmath15-torus knot . then , we perform a @xmath88-band move of type @xmath26 along @xmath25 and we obtain an inner @xmath15-torus knot . in figure [ reidemeister theorem1 ]
we illustrate this for the case where @xmath92 , @xmath94 and @xmath25 a trefoil knot . followed by a band move of type @xmath26 . , width=576 ] without loss of generality ( by isotopy in the complement of @xmath25 )
, the second band move is performed on the innermost arc of the @xmath14 arcs parallel to @xmath25 , creating @xmath14 new parallel arcs even closer to @xmath25 .
after the second @xmath88-band move is performed , the outer arc of the @xmath14 new arcs and the inner arc of the @xmath14 arcs coming from the first band move of type @xmath26 form a band ( see shaded area in figure [ reidemeister theorem2 ] ) .
then , using isotopy in the complement of @xmath25 , we eliminate this band by pulling it along @xmath25 .
this will result in the elimination of @xmath95 pairs of parallel arcs to @xmath25 .
in our example , this is done in two steps ( see figures [ reidemeister theorem2 ] and [ reidemeister theorem3 ] ) . as in the case of integral surgery
the twists coming from the two band moves commute ( see figure [ reidemeister theorem4 ] ) .
arranging these @xmath93 twists pairwise ( figure [ reidemeister theorem5 ] ) , they cancel out by the fact that all twists have the same handiness , but opposite orientation . in the end , @xmath11 is left as in its initial position .
so , a @xmath88-band move of type @xmath27 can be performed using a @xmath88-band move of type @xmath26 and isotopy in the complement of the surgery component @xmath25 .
the proof of theorem [ reidemrational ] is now concluded .
in this section we extend theorem [ geommarkov ] to manifolds with rational surgery description , that is @xmath87 , taking into account the sharpened reidemeister theorem for @xmath0 ( theorem [ reidemrational ] ) .
we first need the following .
a _ geometric @xmath88-braid band move _ is a move between geometric mixed braids which is a @xmath88-band move of type @xmath26 between their closures .
it starts with a little band ( an arc of the moving subbraid ) close to a surgery strand with surgery coefficient @xmath9 .
the little band gets first one twist _
positive _ or _ negative _ , which shall be denoted as @xmath96 and then is replaced by @xmath14 strands that run in parallel to all strands of the same surgery component and link only with that surgery strand , wrapping around it @xmath17 times and , thus , forming a @xmath15-torus knot , figure [ lemma ] .
this braided @xmath15-torus knot is denoted as @xmath97 . a geometric @xmath88-braid band move with a positive ( resp .
negative ) twist shall be called a positive geometric @xmath88-braid band move ( resp .
negative geometric @xmath88-braid band move ) .
-braid band move.,width=288 ] by remark 1(ii ) a @xmath88-braid band move may be assumed to take place at the top part of a mixed braid and all strands from a @xmath88-braid band move may be assumed to lie on the righthand side of the surgery strands .
we shall now prove the following .
[ geometric braid equivalence for @xmath87 ] [ geommarkovrat ] two oriented links in @xmath87 are isotopic if and only if any two corresponding geometric mixed braids in @xmath1 differ by mixed braid isotopy , by @xmath2-moves that do not touch the fixed subbraid @xmath6 and by the geometric @xmath88-braid band moves .
the proof is completely analogous to and is based on the proof of theorem 5.10 @xcite ( theorem [ geommarkov ] ) .
let @xmath99 and @xmath100 be two isotopic oriented links in @xmath0 . by theorem [ reidemrational ] ,
the corresponding mixed links @xmath101 and @xmath102 differ by isotopy in the complement of @xmath5 and @xmath88-band moves of type @xmath26 .
note that , by theorem [ reidemrational ] we do not need to consider band moves of type @xmath27 . by theorem
5.10 @xcite , isotopy in the complement of @xmath5 translates into geometric braid isotopy and the @xmath2-moves .
let now @xmath101 and @xmath102 differ by a @xmath88-band move of type @xmath26 ( recall figure [ bmoves ] ) .
let @xmath103 and @xmath104 be two mixed link diagrams of the mixed links @xmath101 and @xmath102 which differ only by the places illustrated in figure [ markovtheorem ] . as in @xcite , by the braiding algorithm given therein , the diagrams @xmath103 and @xmath104 may be assumed braided everywhere except for the places where the @xmath88-band move is performed .
we now braid the up - arc in figure [ markovtheorem](b ) and obtain a geometric mixed braid @xmath105 corresponding to the diagram @xmath103 ( see figure [ markovtheorem](a ) ) .
note that figure [ markovtheorem](c ) is already in braided form and let @xmath106 denote the geometric mixed braid corresponding to the diagram @xmath104 .
band move ( locally ) and its braiding . ,
width=460 ] we would like to show that the two mixed braids @xmath107 and @xmath106 differ by the moves given in the statement of the theorem . we perform a reidemeister i move on @xmath103 with a _ negative _ crossing and obtain the diagram @xmath108 . then
, the corresponding mixed braids , @xmath107 and @xmath109 , differ by mixed braid isotopy and @xmath2-moves ( see figure [ markovproof](a ) and ( b ) ) .
we then perform a _ positive _ @xmath88-braid band move on @xmath109 and obtain the mixed braid @xmath110 . in the closure of @xmath110 we unbraid and re - introduce the two up - arcs illustrated in figure [ markovproof](b ) , obtaining a diagram @xmath111 with the formation of a reidemeister ii move . performing this move on @xmath111 we obtain the diagram @xmath104 , which is already in braided form and its corresponding mixed braid is @xmath106 ( see figure [ markovproof](c ) and ( d ) ) .
so , the mixed braids @xmath110 and @xmath106 differ by mixed braid isotopy and @xmath2-moves . therefore , we showed that the braids @xmath107 and @xmath106 in figure [ markovtheorem](a ) and ( c ) differ by mixed braid isotopy , @xmath2-moves and a braid band move .
this concludes the proof . .
, width=604 ]
in order to translate the geometric mixed braid equivalence to an equivalence of algebraic mixed braids we follow the strategy in @xcite for integral surgery description .
namely , we apply first parting and then combing to the geometric mixed braids .
what makes things more complicated in the case of rational surgery description is that the surgery braid @xmath6 is in general not a pure braid and when we apply a @xmath88-braid band move on a mixed braid , the little band that approaches the surgery strand is replaced by @xmath14 strands that run in parallel to all strands of the same surgery component . in order to adopt and apply results from @xcite
we need the notion of a _
@xmath88-strand cable_. [ cable ] we define a _
@xmath14-strand cable _ to be a set of @xmath14 parallel strands coming from a @xmath88-braid band move and following one strand of the specified surgery component .
let now @xmath112 be a geometric mixed braid and suppose that a @xmath88-braid band move is performed on @xmath112 . treating the new strands coming from the braid band move as cables running in parallel to the strands of a surgery component ,
that is , treating each cable as one thickened strand , we may part the geometric mixed braid following the exact procedure as in @xcite .
more precisely , we have the following . [ partcable ] cabling and standard parting commute .
that is , standard parting of a mixed braid with a @xmath88-braid band move performed and then cabling , is the same as cabling first the set of new strands and then standard parting .
let @xmath112 be a geometric mixed braid on @xmath41 strands and let a @xmath88-braid band move be performed on a surgery component @xmath11 of @xmath6 .
let also @xmath113 be the numbers of the strands of the surgery component @xmath11 and let @xmath114 denote the @xmath14-strand cables corresponding to @xmath115 .
on the one hand , after the @xmath88-braid band move is performed and before any cablings occur , we part the geometric mixed braid following the procedure of the standard parting as described in section 1.3 ( see figure [ stndpartcable](a ) ) . on the other hand we cable first
each set of @xmath14-strands resulting from the @xmath88-braid band move and then we part the geometric mixed braid with the standard parting , treating each cable as one ( thickened ) strand , see figure [ stndpartcable](b ) . since both cabling and parting a geometric mixed braid respect the position of the endpoints of each pair of corresponding moving strands , it follows that cabling and parting commutes .
recall from section 1.3 that a geometric @xmath2-move can be turned to a parted @xmath2-move .
in order to give the analogue of theorem [ partedcco ] in the case of rational surgery we also need to introduce the following adaptation of a parted @xmath36-braid band move .
[ partqbbm ] a _ parted @xmath88-braid band move _ is defined to be a geometric @xmath88-braid band move between parted mixed braids , such that it takes place at the top part of the braid and on the right of the rightmost strand , @xmath16 , of the specific surgery component , @xmath11 , consisting of the strands @xmath115 . moreover ,
the little band starts from the last strand of the moving subbraid and it moves over each moving strand and each component of the surgery braid , until it reaches the last strand of @xmath11 , and then is followed by parting of the resulting mixed braid .
see figures [ partedqbbm ] and [ partecables ] .
-braid band move.,width=576 ] -braid band move using cables.,width=566 ] then theorem [ geommarkovrat ] restricts to the following .
[ qparted ] two oriented links in @xmath116 are isotopic if and only if any two corresponding parted mixed braids in @xmath117 differ by a finite sequence of parted @xmath2-moves , loop conjugations and parted @xmath88-braid band moves . by lemma [ partcable ]
the cables resulting from a geometric @xmath88-braid band move are treated as one strand , so we can apply theorem [ partedcco ] . moreover , by lemma 9 in @xcite a geometric @xmath88-braid band move may be always assumed , up to @xmath2-equivalence , to take place on the right of the rightmost strand of the specific surgery component . in order to translate theorem [ qparted ] into an algebraic equivalence between elements of @xmath79
we need the following lemmas .
[ combinglem ] the crossings @xmath49 , @xmath53 of the fixed subbraid @xmath6 , and the loops @xmath43 , for @xmath118 , satisfy the following ` combing ' relations : @xmath119 * notation : * we set @xmath120 , for @xmath121 and @xmath122 , for @xmath123 .
we note that @xmath124 .
also , by convention we set @xmath125 .
then we have the following : [ cableloop ] a positive looping between a @xmath14-strand cable and the @xmath126 fixed strand of the fixed subbraid @xmath6 has the algebraic expression : @xmath127 while a negative looping has the algebraic expression : @xmath128 we start with figure [ lemma2](a ) where a positive looping between a @xmath14-strand cable and a fixed stand of the mixed braid is shown . in figure
[ lemma2](b ) the cable is replaced by the @xmath14 strands according to definition [ cable ] .
then , using mixed braid isotopy , we end up with figure [ lemma2](c ) , top , whereby we can read directly the algebraic expression @xmath129 .
the second algebraic expression comes from the bottom illustration of figure [ lemma2 ] .
similarly , in figure [ lemma2b ] we illustrate the case where a negative looping between a @xmath14-strand cable and a fixed strand of the mixed braid occurs .
[ cablecomb ] cabling and combing commute . that is , treating a @xmath14-strand cable as a thickened moving strand and combing it through the fixed subbraid @xmath6 , the result is equivalent to combing one by one each strand of the cable .
according to the combing lemma we have to consider all cases between looping and crossings of the subbraid @xmath6 .
we will only examine the four cases illustrated in figure [ combing ] as representative cases .
all others are completely analogous .
the first case is illustrated in figure [ combcable1 ] , where a positive looping between the cable and the @xmath50 fixed strand of @xmath6 is being considered and the crossing of the fixed strands is positive . for a negative looping
the proof is similar .
we now consider the case illustrated in figure [ combcable2 ] , where a positive looping between the cable and the @xmath130 fixed strand of @xmath6 is being considered , and the crossing in @xmath6 is positive .
we shall prove this case by induction on the number of strands that belong to the cable , since , as we can see from figure [ combcable2 ] , the resulting algebraic expressions are not directly comparable .
the case where the cable consists of one strand is trivial .
for a two - strand cable , combing the cable first and then uncabling ( see top part part of figure [ 2sccomb ] ) results in the algebraic expression : @xmath131 while uncabling first and then combing ( bottom part of figure [ 2sccomb ] ) results in the algebraic expression : @xmath132 -strand cable : case 2 . , width=556 ]
we show below that these algebraic expressions are equal , whereby we have underlined expressions which are crucial for the next step . indeed
: @xmath133 we ended up with one of the defining relations of the mixed braid group @xmath3 , recall ( [ b ] ) .
we now consider a @xmath134-strand cable and we let the first @xmath14 strands form a @xmath14-strand subcable .
we first comb the @xmath14-strand cable and then the @xmath135 strand and the result follows by applying the case of a 2-strand cable and the induction hypothesis for the @xmath14-strand cable ( see figure [ combcableproof2 ] ) . let now @xmath112 be a parted mixed braid and let a parted @xmath88-braid band move be performed on the last strand , @xmath16 , of a surgery component consisting of the strands @xmath115 . recall figures [ partedqbbm ] and [ partecables ] . in order to give an algebraic expression for the parted @xmath88-braid band move , we part locally the subbraids @xmath97 and @xmath136 and the loop generators @xmath137 , @xmath138 , and we use mixed braid isotopy in order to transform @xmath97 into @xmath139 and @xmath136 into @xmath140 . see figures [ cables2 ] , [ cablesc ] , [ cablesa ] and [ cablesa2 ] . then , @xmath139 has the algebraic expression : @xmath141^p\ ] ] and @xmath140 has the algebraic expression : @xmath142 -braid band move is an algebraic @xmath88-braid band move . , width=451 ] of the mixed braid to @xmath140 after a @xmath88-braid band move is performed .
, width=355 ] for @xmath143 after a @xmath88-braid band move is performed .
, width=556 ] for @xmath144 after a @xmath88-braid band move is performed .
, width=576 ] we are now in the position to give the definition of an algebraic @xmath88-braid band move .
[ algbm ] ( i ) an _ algebraic @xmath88-braid band move _ is defined to be a parted @xmath88-braid band move between elements of @xmath145 and it has the following algebraic expression : @xmath146 where @xmath147 is the algebraic mixed braid @xmath27 with the substitutions : @xmath148 @xmath149 \(ii ) a _ combed algebraic @xmath88-braid band move _ is a move between algebraic mixed braids and is defined to be a parted @xmath88-braid band move that has been combed through @xmath6 .
moreover , it has the following algebraic expression : @xmath150 where @xmath151 is the combing of the parted @xmath14-strand cables @xmath114 through the surgery braid @xmath6 ( see figure [ cables ] ) .
-braid band move .
, width=585 ] we are , finally , in the position to state the following main result of the paper .
[ algmarkov ] let @xmath115 be the numbers of the strands of a surgery component @xmath11 and let @xmath114 be the corresponding @xmath14-strand cables arising from a @xmath88-braid band move performed on @xmath11 .
then , two oriented links in @xmath87 are isotopic if and only if any two corresponding algebraic mixed braid representatives in @xmath79 differ by a finite sequence of the following moves : + ( i ) algebraic @xmath0-moves : @xmath152 , for @xmath70 , + ( ii ) algebraic @xmath0-conjugation : @xmath153 , for @xmath82 , + ( iii ) combed loop conjugation : @xmath154 , for @xmath155 , where @xmath58 is the combing of the loop @xmath156 through @xmath6 , + ( iv ) combed algebraic braid band moves : @xmath157 , where the algebraic expressions of @xmath139 and @xmath140 are as in eqs .
( 3 ) and ( 4 ) respectively , @xmath147 is @xmath27 with the substitutions of the loop generators as in definition [ algbm ] and @xmath158 is the combing of the resulting @xmath14-strand cables @xmath159 through the fixed subbraid @xmath6 , or equivalently , by a finite sequence of the following moves : + @xmath160 algebraic @xmath2-moves ( see algebraic expressions in eqs 2 ) , + @xmath160 combed loop conjugation , + @xmath160 combed algebraic braid band moves ( definition [ algbm ] ) .
the arguments for passing from parted braid equivalence ( theorem [ qparted ] ) to algebraic braid equivalence are the same as in those in the proof of the transition from theorem [ partedcco ] to theorem [ algcco ] in the case of integral surgery .
the only part we need to analyze in detail is the algebraization of a parted @xmath88-braid band move .
namely , we will show that the following diagram commutes .
@xmath161 in words , we start with a parted mixed braid @xmath162 and we perform on it a parted @xmath88-braid band move ( definition [ partqbbm ] ) obtaining a parted mixed braid @xmath163 , where @xmath12 is the number of strands forming the surgery component .
we then comb both parted mixed braids obtaining @xmath164 and @xmath165 respectively
. we will show that the corresponding algebraic parts , @xmath166 and @xmath167 differ by the algebraic braid equivalence given in the statement of the theorem .
we apply lemma 8 in @xcite , where the @xmath14 strands of a braid band move are placed in the cable and the cable is treated as one strand .
more precisely , we note that the parted @xmath88-braid band move takes place at the top of the braid , so it forms an algebraic @xmath88-braid band move .
we now comb away @xmath27 to the top of @xmath6 and on the other side we comb away @xmath147 .
since the @xmath14-strands cable of the parted @xmath88-braid band move lie very close to the surgery strands , this ensures that the loops @xmath168 around any strand of the @xmath12 strands of the specific surgery components get combed in the same way before and after the @xmath88-braid band move .
so , having combed away @xmath27 we are left at the bottom with the identity moving braid on the one hand , and with the combing of all cables of the braid band move on the other hand , which is precisely what we denote @xmath169 . finally , by lemma [ cablecomb ] , combing and cable commute .
thus , the theorem is proved .
in this section we give the braid equivalences for knots in specific @xmath4-manifolds that play a very important role to @xmath4-dimensional topology such as the lens spaces @xmath8 , homology spheres and seifert manifolds .
it is known that the lens spaces @xmath8 can be obtained by surgery on the unknot with surgery coefficient @xmath9 .
so , the fixed braid @xmath5 that represents @xmath8 is the identity braid of one single strand and thus , no combing is needed .
we have the following : _ two oriented links in @xmath8 are isotopic if and only if any two corresponding algebraic mixed braids in @xmath170 differ by a finite sequence of the following moves ( compare with @xcite , section 4 ) : _ it is known that a dehn surgery on a knot yields a homology sphere exactly when the surgery coefficient is the reciprocal of an integer ( see [ ro ] p.262 ) . for example , surgery on the right - handed trefoil , with surgery coefficient @xmath180 yields the poincare manifold also known as dodecahedral space " ( for the algebraic braid equivalence in this case see @xcite , section 4 ) . in this subsection
we give the algebraic braid equivalence for knots in a homology sphere @xmath0 obtained from @xmath1 by surgery on the trefoil knot with rational surgery coefficient @xmath181 , where @xmath182 .
@xmath189 @xmath147 _ is the word _
@xmath27 _ with the substitutions : _
@xmath190 _ and _ @xmath191 _ is the combing of the @xmath14-strand cables ( @xmath192 and @xmath193 ) through the fixed braid :
_ @xmath194 @xmath160 if a @xmath88-braid band move is performed on the @xmath126 strand of the fixed braid with rational coefficient @xmath9 ( see figure [ seifcomb ] ) then : @xmath199 , where @xmath200 is the combing of the @xmath201 cable through @xmath6 , it is well - known that a manifold @xmath0 obtained by rational surgery from @xmath1 along an @xmath207-torus knot with rational coefficient @xmath9 is either the lens space @xmath208 , or the connected sum of two lens spaces @xmath209 , or a seifert manifold ( for more details the reader is referred to @xcite ) . _
( 4 ) _ _ combed algebraic braid band moves : _ for @xmath155 we have : @xmath212 where @xmath213^p,\\ c_{\pm } & = & \lambda_{n , n+mq-2}\
\sigma_{n+mq-1}^{\pm1}\ \lambda_{n , n+mq-2}^{-1 } , \end{array}\ ] ] @xmath214 _ is the combing through the fixed braid braid of the parted moving cables parallel to the surgery strands and @xmath147 is the word @xmath27 with the substitutions : _
@xmath215 in figures [ homtref ] and [ hompart ] we illustrate an example where the @xmath207-torus knot is the @xmath216-torus knot , @xmath92 and @xmath94 ( see proposition 3.1 in @xcite for details about the manifold obtained ) .
let @xmath0 be the manifold obtained by rational surgery along a framed link @xmath5 in @xmath1 .
let also @xmath37 be a parted mixed braid representing a link in @xmath0 . in figures [ seifcomb1 ] to [ seifcomb6 ]
we illustrate step - by - step the algebraization of a geometric @xmath88-braid band move .
more precisely , in figure [ seifcomb1 ] a geometric @xmath88-braid band move takes place on the last strand of a surgery component @xmath217 of @xmath6 . in figure [ seifcomb2 ]
we part all cables @xmath218 arising from the geometric @xmath88-braid band move , turning the initial geometric @xmath88-braid band move to a parted @xmath88-braid band move . in figure [ seifcomb2 ]
we also part locally the @xmath15-torus subbraid @xmath97 .
this leads to the algebraic expression @xmath139 of @xmath97 , illustrated in figure [ seifcomb3 ] , where the local parting of the crossing subbraid @xmath96 is also initiated . in figure [ seifcomb4 ] the algebraic expression @xmath140 of @xmath96 is illustrated and the local parting of all loop generators is also initiated .
this leads to the algebraic expressions of the loop generators , in figure [ seifcomb5 ] , where also the preparation for combing of the cables @xmath218 through @xmath6 is illustrated .
note that the top part of figure [ seifcomb5 ] ( above the dotted line ) illustrates an algebraic @xmath88-braid band move .
finally , in figure [ seifcomb6 ] the combing of the cables @xmath218 through @xmath6 is performed and the final result is a combed algebraic @xmath88-braid band move . | in this paper we describe braid equivalence for knots and links in a 3-manifold @xmath0 obtained by rational surgery along a framed link in @xmath1 .
we first prove a sharpened version of the reidemeister theorem for links in @xmath0 .
we then give geometric formulations of the braid equivalence via mixed braids in @xmath1 using the @xmath2-moves and the braid band moves .
we finally give algebraic formulations in terms of the mixed braid groups @xmath3 using cabling and the techniques of parting and combing for mixed braids .
we also provide concrete formuli of the braid equivalence in the case where @xmath0 is a lens space , a seifert manifold or a homology sphere obtained from the trefoil .
the algebraic classification of knots and links in a @xmath4-manifold via mixed braids is a useful tool for studying skein modules of @xmath4-manifolds . |
it is known that the special relativity theory is the physical theory of measurement in inertial frame of reference proposed by albert einstein in 1905 , after substantial contributions of hendrik lorentz and henri poincar ( @xcite,@xcite ) .
this theory gave us @xmath0 which led to the atomic bomb and unlocked the secret of the stars .
at that time the expansion nature of the universe was not discovered yet , and this information was not included in the formulation of the special relativity theory . in 1915 albert einstein proposed the general relativity theory , in which the expansion of the universe was not the main objective .
he proposed that gravity , as well as motion , can affect the intervals of time and of space .
the general relativity theory gave us : the space warps , the big bang , and black holes .
the metric @xmath1 of the space - time proposed in general relativity does not allow the incorporation of the expansion in the the lorentz transformation ( since the increasing function of time @xmath2 , that characterizes the expansion nature of the space , omits the linearity in the considered transformations ) , and then it does not permit to perceive if the laws of nature proposed by the special relativity are affected or not together with the universe expansion .
many essays have been achieved in this purpose in the last hundred years , which can be found easily in the literature .
the only cited papers in this article are those who have a clear and direct relationship with this kind of approach . given that the expansion was studied in the general relativity theory , it seems counter nature to discuss universe expansion in the special relativity theory .
however , discussing the effect of the expansion on the laws of nature in the special relativity is different , and might entail some disturbing new concepts which will challenge deeply our deep - rooted understanding of the relativistic laws of nature . in this paper we will explore the incorporation of a special expansion in the special relativity theory that preserves the notion of inertial frame , and we will rewrite and interpret the relativistic laws of nature in the light of this special expansion . in this article
we introduce a practical quantification of the space expansion in our universe using a metric with a numerical product that quantifies the expansion of the space step by step ( we will call it universe expansion ) , and maintains the notion of inertial frame on it . in the second section we rewrite the lorentz transformation equations using the new metric to obtain a family of lorentz transformation equations @xmath3 , where each transformation @xmath4 represents the lorentz transformation equations at the step n of the space expansion .
this family of transformation leads to a decreasing limiting velocity of moving bodies from one step to another , and then the physical laws are affected by the universe expansion , followed by conclusions derived concerning energy , masses , and momentum in an expanding space . in section three ,
the composition of velocity is derived using @xmath4-lorentz transformation equations to discuss the invariance of the new limiting velocities of matter .
the rest mass of photon is then discussed in section forth and new quantum formalism is introduced .
we conclude this section by discussing an inelastic collision in which we find that the conversion of the rest - energy into kinetic energy is affected by the universe expansion as well as the fact that the total liberated energy by fission of atoms is increasing naturally together with the universe expansion .
we conclude by some comments related to the notion of time and proper time . from the rate at which galaxies are receding from each other in our universe
, it can be deduced that all galaxies must have been very close to each other at the same time in the past .
meanwhile galaxies are receding from each other what really happen to the physical laws , are they affected by the process of expansion or not ? in the purpose to investigate the status of the physical laws together with the universe expansion we propose the following quantification of the universe movement : suppose that we can quantify the whole universe movement , from the beginning of the universe expansion where matter was together until present time with current picture of matter distribution .
suppose that we can subdivide the whole process of expansion from the beginning until present time into n steps , in which we approximate the rate of change of the distance between any two separated events from one step to another together with the universe expansion as follow : @xmath5 @xmath6 : _ if the distance between two events at the step n is equal to @xmath7 , and at the step ( n+1 ) is equal to @xmath8 then @xmath9 for all @xmath10 , where @xmath11 is a sequence such that @xmath12 , @xmath13 @xmath14 , and @xmath15 converges . _ for this quantification
the step 0 represents the beginning of the expansion of our universe ( once @xmath16 ) and the space - time is defined as the set of events together with the notion of squared interval @xmath17 defined between any two events .
an event @xmath18 in a reference frame @xmath19 is fixed by four coordinates @xmath20 , @xmath21 , @xmath22 and @xmath23 , where @xmath24 are the spatial coordinates of the point in space when the event in question occurs at the step 0 , and where @xmath23 fixes the instant when this event occurs .
another event @xmath25 will be described within the same reference frame by four different coordinates @xmath26 , @xmath27 , @xmath28 , and @xmath29 .
the points in space where these two events occur are separated by the distance @xmath30 given by @xmath31 .
the moments in time when these two events occur are separated at the step 0 by a time interval given by @xmath32 .
the squared interval @xmath17 between these two events is given as function of these quantities , by definition , through the generalization of pythagorean theorem @xmath33 where @xmath34 is the maximum speed of signal propagation at the beginning of the universe expansion ( the maximum speed for the transmission of information ) .
at the step 1 of the universe expansion , the event @xmath18 in the reference frame @xmath19 is fixed by four new coordinates @xmath35 and @xmath36 , where @xmath35 are the spatial coordinates of the point in space when the event in question occurs at the step 1 , and where @xmath36 fixes the instant when this event occurs , meanwhile the event @xmath25 is described within the same reference frame by four different coordinates @xmath37 , @xmath38 , @xmath39 , and @xmath40 .
the points in space where these two events occur are separated by the distance @xmath41 given by @xmath42 . while the universe expands from the step 0 to the step 1 at a rate such that @xmath43 , @xmath44 and @xmath45 , then we have @xmath46 .
the moments in time when these two events occur are separated at the step 1 by a time interval given by @xmath47 .
the squared interval @xmath48 between these two events is given as function of these quantities , by definition , through the generalization of pythagorean theorem @xmath49 where @xmath34 is the maximum speed of signal propagation at the beginning of the universe expansion , etc .. at the step n of the universe expansion , the event @xmath18 in the reference frame @xmath19 is fixed by four coordinates @xmath50 , @xmath51 , @xmath52 and @xmath53 , where @xmath54 are the spatial coordinates of the point in space when the event in question occurs at the step n , and where @xmath53 fixes the instant when this event occurs , meanwhile the event @xmath25 is described within the same reference frame by four different coordinates @xmath55 , @xmath56 , @xmath57 , and @xmath58 .
the points in space where these two events occur are separated by the distance @xmath7 given by @xmath59 . while the universe expands from the step ( n-1 ) to the step
n at a rate such that @xmath60 , @xmath61 and @xmath62 , then we have @xmath63 the moments in time when these two events occur are separated at the step n by a time interval given by @xmath64 .
the squared interval @xmath65 between these two events is given as function of these quantities , by definition , through the generalization of pythagorean theorem @xmath66 which gives @xmath67 where @xmath34 is the maximum speed of signal propagation at the beginning of the universe expansion .
the above metrics , from the beginning of the universe expansion ( step 0 ) until the present time , quantify the expansion of the space - time step by step .
indeed , a ball @xmath68 defined by @xmath69 will increase if its radius increases , and we have for @xmath70 , @xmath71 , @xmath72 however , for all @xmath73 , the sphere @xmath74 is an expanding sphere given by @xmath75 where @xmath76 is an increasing and bounded product .
the infinitesimal space - time interval at the step n , @xmath71 , between two neighboring events within a locally inertial frame metric takes then the form @xmath77 in which @xmath78 represents the expansion step of the physical universe . in such locally inertial reference frame
the laws of the special relativity apply at each step .
we propose to rewrite the physical laws in each space - time step by step to investigate any possible evolution in the laws of nature .
the metric ( [ m1 ] ) can be written as @xmath79 with @xmath80 , @xmath81 , and @xmath82 .
let us consider an expanding space @xmath83 in which the proper distance between any distant points is increasing following the above quantification @xmath84 such that _ at small scale as well as at the subgalactic scale the fundamental forces of nature maintain the form of matter we have today_. let us consider that the expansion is totally modeled by an adequate numerical quantification represented by the product @xmath85 introduced above ( that is to say if the position of a body at rest or the location of an event is expressed locally via three space coordinates @xmath86 in an inertial frame at the step 0 , then at the step 1 its location at rest is expressed in the same frame by @xmath87 , at the step 2 is expressed in the same frame by @xmath88 , and at the step n its location at rest is expressed by @xmath89 , where n represents the expansion step of the space movement quantification ) .
let us consider two cartesian frames @xmath90 and @xmath91 in @xmath83 such that at each expansion step the equations of newtonian mechanics hold good .
suppose that an event has cartesian coordinates @xmath92 relative to @xmath93 , and @xmath94 relative to @xmath95 at the same step of expansion .
in general for a linear transformation of the cartesian frames we have @xmath96 where the coefficients @xmath97 depend on the movement . the general properties of homogeneity and isotropy of the space with the choice that the frame @xmath95 moves in the x direction of @xmath93 with uniform velocity @xmath98 ( without losing generality ) , such that the corresponding axes of @xmath93 and @xmath95 remain parallel throughout the motion having coincided at @xmath99 will give @xmath100 suppose that a light signal is emitted from the common center of the two frames at @xmath99 . since the space @xmath83 is expanding , then the observer in the frame @xmath93 will observe that the light travels in all direction in an expanding sphere , at the step n , given by @xmath101 and from the frame @xmath95 , the observer perceives the same @xmath102 and we have @xmath103 substituting formula ( [ e1 ] ) in ( [ e2 ] ) gives@xmath104,which yields the following system @xmath105 the resolution of the system ( [ e3 ] ) leads to the lorentz transformation equations @xmath4 with a new limiting velocity @xmath106 the first immediate consequence from the lorentz transformations in an expanding universe is the following : there exists a limiting velocity for moving bodies given by @xmath107 which depends on the fossil light velocity ( velocity of light at the beginning of the universe expansion ) and the expanding parameter of the universe .
indeed , @xmath108 is an increasing and bounded product greater than one , then the limiting velocity @xmath109 is decreasing along with the space expansion of the universe .
it turns out that the limiting velocity was maximal at the beginning of the expansion of the universe ( step 0 ) and was equal to @xmath34 , that is why we call it the fossil velocity of light .
thus the limiting velocity of moving body , including the light , will decrease in an expanding universe .
however , the physical meaning of the equations ( [ e4 ] ) obtained in respect to moving rigid bodies and moving clocks remain the same as for the classical lorentz transformations ( deformability and lose of synchronization , time transformations , etc ) except the fact that the expansion of the universe will be involved in all formulation of the physical laws as we will see later on . as a consequence of these new transformations and taking into account the dynamic of the universe
, we propose to adjust the einstein s equation relative to the mass - energy as follow : the total energy of moving body at the step n of the space expansion is given by the law @xmath110 and the body s rest - mass energy at the step n of the space expansion is given by the law @xmath111 where @xmath112 represents the mass of a body at rest relative to the observer in an expanding universe at the step n. the relative mass at the step n is given by the law @xmath113 indeed , using the newtonian formula @xmath114 , the equation ( [ e6 ] ) , and the fact that the newton s law remains valid if it is written as @xmath115 give the following : @xmath116 on the other hand we have @xmath117 and @xmath118 which yields in ( [ n1 ] ) @xmath119 then @xmath120 with @xmath121 . by integration
we obtain @xmath122 for @xmath123 , and @xmath124 , we obtain @xmath125 and then @xmath126 which gives equation ( [ e7 ] ) . on the other hand the law ( [ e7 ] ) leads to the law ( [ e6 ] ) , indeed , the work of displacing a body by a distance @xmath127 at the step
n is @xmath128 where @xmath98 is the velocity vector .
if this work serves to increase the energy @xmath129 of the body at the step n , then @xmath130 then @xmath131 since @xmath132 is given by ( [ e7 ] ) , we have @xmath133 by substitution we obtain @xmath134 then @xmath135 which gives together with formula ( [ dm ] ) @xmath136 that is to say the law ( [ e6 ] ) .
the universe dynamic does not affect the mass of a given object at rest .
however , object at rest has a rest energy @xmath137 given by ( [ e0 ] ) in which the universe movement is manifested .
when the object starts moving under any force , its total energy , momentum and mass are directly affected by the expansion parameters of the universe .
if the speed @xmath34 is the fossil velocity of light , then the term @xmath138 in laws ( [ e6 ] ) , ( [ e0 ] ) , and ( [ e7 ] ) represents the current experimental velocity of light , of course if we consider that the present time corresponds to the step n of the universe expansion , and it constitutes a limiting velocity for any motion or transfer of interaction at the step n of the universe expansion . for small values of @xmath139 ,
the total energy ( [ e6 ] ) goes over into the classical expression of the kinetic energy as shown by @xmath140 which gives @xmath141 the first term represents @xmath137 the energy at rest , and the second term represents the newtonian kinetic energy .
the relativistic kinetic energy @xmath142 or the motion energy at the expansion step n can be defined by subtracting the rest energy @xmath137 from the total energy @xmath129 @xmath143 the substitution of formula ( [ e6 ] ) and ( [ e0 ] ) in the last equality gives @xmath144 and using relative mass ( [ e7 ] ) we obtain @xmath145 the kinetic energy of a body is not only related to the increase of masses when they are moving as asserted by the special relativity but also related to the universe expansion .
the use of equation ( [ e7 ] ) in the equation of motion ( force is equal to the rate of change of momentum ) remains valid .
however , here the momentum is now at the step n given by : @xmath146 the equation @xmath147 determines the motion of a body acted on by any force , but now correctly takes expanding effects into account .
clearly , the form of the momentum and force equation is very similar in relativity theory and in our approach ; the effective mass @xmath132 depends on the speed of motion of the body relative to the observer and the n - th step expansion parameter @xmath148 according to the formula ( [ e7 ] ) , while in relativity theory it is independent of the universe expansion .
the crucial feature is that the effective mass @xmath132 diverges as @xmath98 tends to the limiting velocity @xmath149 and so the momentum @xmath150 also diverges . as for relativity theory , the energy and momentum are related . indeed , using formula ( [ e6 ] ) and ( [ e6 ] ) we obtain the following relation @xmath151
like for special relativity we will use the classical definition of velocity .
if we call @xmath152 the velocity of an object in the frame @xmath93 , its components with respect to the three axes are @xmath153 .
the same velocity is measured in the frame @xmath95 as @xmath154 with components given by @xmath155 for simplicity and without losing generality , we will consider only that the velocity @xmath152 is parallel to the x - axes .
a simple differentiation of the equations of lorentz transformations @xmath4 gives @xmath156 then we have @xmath157 which gives @xmath158 since @xmath152 is parallel to the x - axes , then the law of composition of velocity is given by @xmath159 and inversely @xmath160 which has the form of the relativist law of composition of velocity with the appearance of the term @xmath161 that characterizes the expansion of the universe at the step n. in the extreme case where an observer in the frame @xmath95 is emitting light signals in the direction of x - axis , since the frame is in an expanding universe at the step n , then the velocity of these signals in the frame @xmath95 is given by @xmath162 which gives @xmath163 which gives @xmath164 and then the limiting velocity of light @xmath109 at the step
n is invariant , that is to say the velocity of light `` measured '' in a moving frame at the step n , appears to be equal to @xmath138 in any direction .
the speed of light in vacuum , at the step n of the universe expansion , will be found the same ( equal to @xmath109 ) by any two observers in uniform relative motion and this is true for the whole process of universe expansion , meanwhile this speed is decreasing together with the universe expansion . for @xmath165
the lorentz transformation @xmath166 gives @xmath34 as an invariant limiting velocity of matter , meanwhile @xmath167 gives @xmath168 as the invariant limiting velocity of matter , and so on , until the lorentz transformation @xmath4 which gives @xmath169 as the invariant limiting velocity of matter .
it turns out that the limiting velocity of moving bodies is decreasing together with the universe expansion and that the velocity of light in empty space is independent of the velocity of its source _ at each step _ of the universe expansion despite of its decreasing nature together with the universe expansion . in the whole process of universe expansion
the velocity of light remains independent of the reference frame of the observer .
the formula ( [ e6 ] ) can be justified exactly as the famous formula @xmath0 ( einstein @xcite ) , by using the new composition of velocity provided by the lorentz transformation @xmath4 and the momentum conservation at the step n.
the quantification of the expansion introduced above leads us to set up different steps - lorentz transformations @xmath3 that permit to quantify the physical laws evolution together with the universe expansion , and derive a relationship between the space - time views obtained by any two observers in an expanding universe .
the result obtained unifies the fundamental special - relativity results of time dilation , length contraction , and the relativity of simultaneity , in a single relation for different steps of quantified expansion .
however , from these different lorentz transformations , we sort out that the limiting velocity @xmath109 of matter is decreasing from one step to another including the velocity of light . the velocity of light at the beginning of the expansion of the universe ( step 0 ) was constant and equal to @xmath34 an unknown value , and the @xmath166-lorentz transformation equations are given by @xmath170 this transformation @xmath171 leads to the relativistic formulation of the physical laws @xmath172 on the other hand , at the step 1 the @xmath167-lorentz transformation equations become @xmath173 from which we sort out that the matter has a new limiting velocity @xmath174 . at this step
all the equations of newtonian mechanics hold good , and the formulations of the physical laws are obtained under some changes relative to the new limiting velocity .
similar formulations of the physical law equations ( [ t0 ] ) are introduced with the new limiting velocity and denoted by @xmath175 at the step n , the @xmath4-lorentz transformation equations become @xmath176 in which the limiting velocity of matter reaches the value of @xmath177 , and where all formulations of the physical laws are possible to obtain under some changes with the hypothesis that the equations of newtonian mechanics are valid at this step of the universe expansion ( conservation of momentum , conservation of energy , etc ) . the new limiting velocity @xmath177 at the step
n induces some change in all formulations ( as regards both mechanics and electro - magnetism in inertial systems ) , including the maxwell equations , to obtain that the physical law remains the same under the @xmath4-lorentz transformations .
the formulation of the physical law equations at the step n with the new limiting velocity @xmath109 are denoted by @xmath178 where the quantities @xmath129 , @xmath179 , @xmath180 , @xmath181 and @xmath132 take the form ( [ e6 ] ) , ( [ e0 ] ) , ( [ ec ] ) , ( [ e6 ] ) , and ( [ e7 ] ) respectively under the @xmath4-lorentz transformation for all @xmath10 .
the lorentz transformations @xmath171 , @xmath182,@xmath183 , and @xmath184 have the same mathematical form , and all the physical quantities are affected by the space expansion from one step to another , except the rest - mass @xmath185 which is invariant under universe expansion . since the light has no rest frame , and since the formalism must be the same for all matter , the quantity @xmath185 in the above formalism represents in reality the mass of matter which is invariant under universe expansion , we will call it the _ invariant mass_. the family of transformation @xmath3 leads to fatal variation of the limiting velocity @xmath109 of moving bodies according to the universe expansion .
this has interesting consequences . in the expansion process
the rest energy at the step n can be evaluated compared to the rest energy at the beginning of the universe expansion .
indeed , the rest energy at the step n is given by the law @xmath186 and using the rest energy notation we obtain @xmath187 which means that the rest energy of matter is decreasing along with the universe expansion .
however , the relative mass of matter is increasing together with the universe expansion , indeed , from the law ( [ e7 ] ) at the step n and the same law at the step n+1 we sort out that @xmath188 and we obtain for all @xmath189 @xmath190 which gives @xmath191 this means that the relative mass of matter in our universe is increasing along with the universe expansion as well as the momentum , meanwhile the rest energy is decreasing .
it is possible to determine the evolution of energy , momentum and relative mass together with the universe expansion by using the invariant mass @xmath185 , indeed , it is not difficult to write all these quantities at the step n function of their values at the step 0 ( beginning of the expansion ) and we have the law of evolution of the relative mass together with the universe expansion given by @xmath192 meanwhile the evolution of momentum together with the universe expansion is given by the law @xmath193 and the total energy evolution together with the universe expansion is given by the law @xmath194 however , the evolution of the physical laws from one step to another are given by the following formulas : @xmath195 the evolution of the momentum from one step to another is given by @xmath196 and the evolution of the total energy from one step to another is given by the law @xmath197 it turns out that the concept of expansion is a real source of energy , and this will be analyzed later on with a concrete example .
the velocity of light was constant at the beginning of the expansion of our universe and equal to @xmath34 ( an unknown fossil value ) .
what we measure today in our experiment and what we call the velocity of light , corresponds in reality to the velocity of light at the expansion step n , where its value is given by @xmath138 .
the light is described in quantum mechanic as a quanta of zero mass where the relation between energy and frequency is given by the planck - einstein s relation @xmath198 and where the momentum is given by the relation @xmath199 where @xmath200 the wave length given by @xmath201 the rules of translation from corpuscular terminology to wave terminology , and vice versa are based on the fact that an electromagnetic wave of length @xmath200 and intensity @xmath202 is considered as a stream of photons of frequency @xmath203 and intensity @xmath204 where @xmath205 is the number of photons passing per unit time through unit areas , and the direction of motion of the wave front is the direction of motion of the photons . in the purpose to integrate the new formalism of expansion in quantum formulation we introduce the following : we denote @xmath200 the fossil wave length of the light at the beginning of the universe expansion where the celerity of photon was @xmath34 , with @xmath206 the fossil frequency of the light at that time .
we denote @xmath207 the current wave length of the light at the step n of the universe expansion , where the celerity of photon is @xmath208 , with @xmath209 the light frequency at the expansion step n. hence the wave length represents a distance , therefore by using the property of the expansion of the space the wave length expands and we have @xmath210 which leads to the following relation between the fossil light frequency and the light frequency at the expansion step n @xmath211 and then @xmath212 from the relativistic relation between energy and momentum @xmath213 , we sort out the momentum of photon with velocity @xmath214 given by @xmath215 thus @xmath216 and well defined for @xmath214 , then @xmath217 which gives that the rest energy of a photon @xmath218 is not equal to zero .
the substitution of ( [ evl ] ) in the formula ( [ pv ] ) gives @xmath219 the precedent equality can be expressed using the rest energy ( [ en ] ) at the step n and @xmath109 as @xmath220 therefore the energy and momentum of photon are related by the law @xmath221 the law ( [ pn ] ) leads to the following approximation @xmath222 where @xmath223 is the limiting velocity of light measured at the step n. the relation ( [ c ] ) was introduced experimentally by a. h. compton in 1923 , @xcite . using the formulation given by lorentz transformation @xmath171 , the relative mass of moving body
is given by @xmath224 which gives @xmath225 then @xmath226 which gives @xmath227 then for all @xmath228 we have @xmath229 nevertheless , if the invariant mass @xmath185 of a photon was zero at the origin of the expansion when its velocity was @xmath34 , it will remain zero at the step
n when the light move at speed @xmath177 because of its natural invariance under the universe expansion . if we suppose that the invariant mass of a photon @xmath230 at the step n , @xmath231 , ( since for @xmath165 the formula ( [ e8 ] ) is not defined for @xmath232 ) from the universe expansion , then the speed of light is given by @xmath214 , and by substitution in the the formula ( [ e8 ] ) we obtain @xmath233 using the value of @xmath234 given by formula ( [ pf ] ) , the equality ( [ eq ] ) becomes @xmath235 on the other hand we have ( [ evl ] ) , and then equality ( [ eq1 ] ) gives @xmath236 which is equivalent to @xmath237 since the rest energy of a photon at the beginning of the universe expansion ( step 0 ) was not null , then formula ( [ con ] ) leads to the contradiction @xmath238 then the invariant mass of photon @xmath239 . along the universe expansion
the invariant mass @xmath185 of matter remains the same , meanwhile the associated rest energy decreases ( @xmath240 ) , and the momentum of matter increases . if the rest energy of photon was zero , it can not decrease . since @xmath241 , @xmath242
, then we can evaluate the invariant mass @xmath185 for a photon of velocity @xmath214 .
indeed , we have @xmath243 the use of the formula ( [ evl ] ) in the precedent equality gives @xmath244 and we have @xmath245 using ( [ en ] ) and the velocity of light @xmath109 at the step n to obtain @xmath246 which means that the ratio between rest energy and velocity of light is invariant under universe expansion . the formula ( [ ev0 ] ) means that the light has non zero invariant mass . however , the invariant mass of a photon depends on the fossil rest energy ( the rest energy at the step n is impossible to determine since the light is never at rest ) .
we know only that the invariant mass and the rest energy of a photon are not null , they might be small , very small , but not null . in the purpose to sort out the real nature of energy gained by matter in an expanding universe we introduce the following example of inelastic collision : consider two identical particles which move toward each other along straight line , with equal speeds , and equal invariant mass @xmath185 , they collide and stick together in an expanding space at the step n. the conservation of momentum at the step n gives : @xmath247 thus @xmath248 which gives @xmath249 , so the final object is at rest with mass @xmath250 .
the conservation of total energy at the step n from the universe expansion gives @xmath251 since the final object is at rest ( that is to say @xmath249 ) , then it follows that @xmath252 then we obtain @xmath253 the mass of final object is larger than the sum of the original masses . to evaluate the nature of energy created with the increase of mass together with the universe expansion from the step 0 to the step
n we have to compare the same collision at the origin of the expansion ( step 0 ) and at the step n. indeed the universe expansion creates the increase of moving masses as @xmath254 for a small value of @xmath255 we can approximate ( [ mi ] ) as @xmath256 which gives @xmath257 the gain of masses represents a rest energy at the step
n given by @xmath258 this means that the rest energy gained by a moving particles together with the universe expansion is equal approximatively to the sum of the classical kinetic energy of the two particles factor the rate of increase of rest energy due to the universe expansion , that is to say @xmath259 .
indeed , from formula ( [ en ] ) it is easy to evaluate the variation of the rest energy between the beginning of the universe expansion ( step 0 ) and the universe expansion at the step
n , we obtain @xmath260 from which it turns out that @xmath259 represents the rate of decrease of the rest energy together with the universe expansion . since the mass of final object is larger than the sum of the original masses , we can approximate this excess of mass for a small value of @xmath255 , and compare it to the same value at the beginning of the universe expansion .
indeed the approximation of the law ( [ mn ] ) , for all @xmath10 , gives @xmath261 which gives @xmath262 and then we obtain @xmath263 it turns out that the excess mass of the composite object at the step n of the universe expansion compared to the mass of the same composite object at the beginning of the expansion ( step 0 ) is increasing together with the universe expansion and then the kinetic energy brought in also increases . the lost kinetic energy has been converted to rest energy ( mass ) .
the classical explanation for the loss of kinetic energy attributes it to conversion into thermal energy ( heat ) : the final object will have a higher temperature or a large internal energy . now since the mass of the final object ( after collision formula ( [ mn ] ) will have an increasing total mass , then this means that the temperature of the final object will increase together with the universe expansion .
since many evidences prove that early earth was molten from bombardment of rocks and mini planets and others , then we can say that there must exist a natural factor that contributes to the global warming of our planet due to the universe expansion ( the global warming of our planet is not only attributed to human activities ! ) .
suppose that we have an atom of uranium with rest mass @xmath264 measured in an expanding universe at the step n , and suppose that something happens so that the atom flies into two equal pieces moving with speed @xmath98 , so that each part has a relative mass @xmath132 at the step n .
suppose that these pieces encounter enough material to slow them up until they stop and then each part will have an invariant mass @xmath124 .
to reach its rest position each part will give an amount of energy left in the material in some form , whatever . the left energy for one part
is then given by @xmath265 the conservation of total energy in an expanding universe at the step n gives @xmath266 so the total liberated energy at the step n is given by @xmath267 then @xmath268 which can be written as @xmath269 where @xmath270,@xmath271 .
the law given by ( [ en2 ] ) for @xmath165 was used to estimate how much energy would be liberated under fission in the atomic bomb since the mass of uranium atom was known as well as the atoms into which it splits ( iodine , xenon , and so on ) .
however , it turns out from the law ( [ en2 ] ) that the released energy when an atom of uranium undergoes fission is correlated to the universe expansion which has interesting consequences .
as the universe expands we have the law ( [ mc ] ) , which gives @xmath272 and from ( [ en2 ] ) we have @xmath273 using the law ( [ en0 ] ) we obtain @xmath274 this means that the ratio between the total liberated energy by an atom of uranium and its rest energy is increasing together with the universe expansion .
the total liberated energy by an atom of uranium ( [ en3 ] ) can be approximated , indeed , we have @xmath275 then we have @xmath276 by using the law ( [ en ] ) that expresses the correlation between the rest energy at the step n and the rest energy at the step 0 , we obtain @xmath277 for a small value of @xmath255 , we obtain the following approximation @xmath278 which gives @xmath279 then @xmath280 it turns out that the total liberated energy by an atom of uranium is increasing together with the universe expansion and we have for one atom of uranium the following evolution of liberated energy from one step to another @xmath281 this has interesting consequences that will change our understanding of gravity , the missing masses , or the estimation of stars energy and others .
the energy in our universe increases together with the universe expansion , and this excess of energy is naturally due to the recession of galaxies from each other .
the increase of energy and mass in our universe compensates the increase of distance between matters and might explain why the gravity exists at long distance .
our estimation of stars energy , or galaxies masses are erroneous if we omit to take into account the movement of matter due to the universe expansion .
we conclude by raising the fatal fact that there is no stationary safe distance when an atom of uranium undergoes fission in an expanding universe , since the liberated energy increases along with the universe expansion , the security of nuclear reactor must be endlessly reconsidered .
it turns out that the tame of this energy is strewed with risk and peril together with the universe expansion .
the family of transformations @xmath171 , @xmath182 , @xmath282 , @xmath183 , @xmath184 is a mathematical formulation of the lorentz transformation that incorporates the given quantification of the space expansion .
this family of transformations @xmath283 is defined for all @xmath284 , where @xmath109 represents the limiting velocity of moving bodies at the expansion step n. from this limiting velocity it turns out that the velocity of light is decreasing as the universe expands , and the limiting velocity @xmath109 was maximal only at the step 0 ( in the beginning of the universe expansion @xmath232 ) , however , at each step the velocity of light is constant in the sense that it is independent of the reference frame of the observer .
all formulations and laws based on the assumption that the velocity of light in empty space is independent of the velocity of its source remain valid for matter and related physical phenomenon during the universe expansion step by step , and the relativistic formulation of the physical laws remains valid _ experimentally _ since the velocity of light at the step n is experimentally given by @xmath285 which represents in this formalism @xmath138 , that is to say @xmath286 where c is the fossil velocity of light ( the fossil velocity of light could be detected in our space ) . at each step of the universe expansion
the measured velocity of light remains invariant from one frame to another in uniform relative motion .
the meaning of this approximation based on the universe quantification @xmath84 is that the velocity of light is constant locally and variable globally with respect to the universe age , and what we mean by constant locally is that the variation of the velocity of light is too small to be detected in short time with respect to the universe age .
all our experimental tests and applications of the special relativity laws at present time are and remain valid since the substitution of the experimental measure of the light speed in the @xmath166-lorentz transformation will represent @xmath4-lorentz transformations .
indeed , the velocity of light is locally constant at the scale of normal time ( using a short interval of time as unit ) and globally variable at the scale of cosmical time ( using a large interval of time as unit ) .
this local and global behavior can be derived straight forward from the quantification @xmath6 .
indeed , the local behavior is reached in the quantification @xmath6 if we use a big number of subdivisions : the bigger the used number @xmath78 of steps is the shorter the time interval of steps we obtain .
thus , as @xmath78 tends to infinity , we have @xmath287 as a consequence of the convergence of the product @xmath148 , then @xmath288 where @xmath18 is a large positive real number . hence the equations ( [ cv ] ) and ( [ c ] ) give @xmath289 therefore the velocity of light is almost constant for short period of time ( locally ) . however , the equation ( [ cst ] ) is not valid anymore for the large period of time .
indeed , the smaller the used number @xmath78 of steps is the bigger the time interval of steps we obtain , and in that case we have @xmath290 for all @xmath291 and @xmath292 hence the equations ( [ c ] ) and ( [ cv0 ] ) lead to the following inequality : @xmath293 that is to say the velocity of light is decreasing globally together with the universe expansion from one step to another .
the formalism obtained within this work using a special expansion notion in the special relativity theory to study the uniform motion of the observer relative to the source is based on the validity of the special relativity theory step by step from the big bang to nowadays . which leads to the conclusion that this new formalism is an extension of the special relativity theory .
we close this article by some comments on the notion of time .
the notion of relative time as for special relativity remains valid with some adjustment , from the metric ( [ m1 ] ) is clear that time is a required coordinate to specify a physical event and that this quantity can not be defined without the notion of space and the notion of expansion , the dynamic of the universe is involved in the definition of time .
the expansion of the universe affects the time and the space .
if we refer to the proper time @xmath294 as the time which is measured by an observer in a reference frame at the step n of the universe expansion in which the events occur at the same spatial point , then @xmath295 and the metric ( [ m1 ] ) gives @xmath296 meanwhile in another inertial reference frame the same events will verify ( [ m1 ] ) , which gives @xmath297 from which we obtain @xmath298 where @xmath299 .
it turns out that the rate of proper time for a system varies with motion through space as for special relativity , and what is new here is that the rate of the proper time varies together with the universe expansion even if @xmath98 is constant .
if we refer to the universe proper time @xmath300 as the time which is measured by an observer in a reference frame at rest at the step n of the universe expansion in which the events occur at the same spatial point , it is not difficult to extract from the metric ( [ m1 ] ) the relationship between proper time and coordinate time @xmath301 the time depends on the observer and the universe expanding step .
what is new here is to associate space , time and expansion in the definition of event .
any change of reference or steps affects all of them .
if the proper time @xmath300 of the universe is the same from one step to another and if the velocity @xmath98 and @xmath34 are constant from one step to another then it turns out from the step @xmath302 to the step @xmath78 that @xmath303 and then @xmath304 which allows us to compare the coordinate time between successive steps as @xmath305 which means that the clocks at the step n will run slower than an identical type of clocks at the previous step , the time will run slower together with the universe expansion .
one more thing , the product @xmath148 which allows the quantification of the universe expansion was extracted from the fractal manifold model @xcite .
more details about the natural construction of the expanding parameter @xmath148 and the universe expansion as well as the universe shape can be found in @xcite .
f. ben adda , _ mathematical model for fractal manifold . _ , international journal of pure and applied mathematics , * 38 * , n 2 , 159 - 190 , ( 2007 ) .
f. ben adda , _ new understanding of the dark energy in the light of a new space time _ , in proceeding of the invisible universe , aip conference proceedings ,
vol 1241 , pp 487 - 496 , ( 2010 ) .
a. compton , _ a quantum theory of the scattering of x - rays by light elements _ , physical review , * 21 * , 483 - 502,(1923 ) .
a. einstein , _ does the inertia of body depend upon its energy content?_,translated from `` ist die tr@xmath306gheit eines k@xmath307rpers von seinem energiegehalt abh@xmath306ngin ? '' , annalen der physik , * 17 * , ( 1905 ) .
a. einstein , _ kosmologishe betrachtungen zur allgemeinen relativit@xmath306tstheorie _ , _ preussische akademie der wissenschaften , sitzungsberichte _ , 142 - 152 , ( 1917 ) .
h. poincar@xmath308 , _ sur la dynamique de l@xmath308lectron _ , comptes rendus de lacad@xmath308mie des sciences de paris * 140 * , 1504 - 1508 , ( 1905 ) .
h. poincar@xmath308 , _ sur la dynamique de l@xmath308lectron _ , rendiconti del circolo mathematico di palermo * 21 * , 127 - 175 , ( 1906 ) . | in this paper , we present a new formulation of lorentz transformations using a metric that quantifies the space expansion . as a consequence , we sort out that the limiting velocity of moving bodies is decreasing together with the space expansion .
a new adjustment of relativistic laws is added to incorporate the non static nature of space - time .
the conservation of the physical laws at each step of the quantified expansion allows the obtaining of new formalisms for the physical laws , in particular when an object starts moving under any force , its total energy , momentum and mass are directly affected by the expansion of the space .
an example of inelastic collision is studied and several conclusions derived , specially the example of fission of atoms leads to clear correlation between liberated energy and universe expansion , it turns out that the liberated energy is increasing together with the universe expansion . |
supersymmetry ( susy ) , is one of the most favored extensions of the sm which is capable of stabilizing the ino - sector of fundamental scalars against the ultraviolet divergences .
the ( soft ) breaking of susy , around the @xmath2 scale , brings about two new ingredients compared to the standard electroweak theory ( sm ) : first , there are novel sources of flavor violation coming through the off diagonal entries of the squark mass matrices .
second , there are novel sources of cp violation coming from the phases of the soft masses . the first effect , which can not be determined theoretically , is strongly constrained by the fcnc data @xcite , and therefore , as a predictive case , it is convenient to restrict all flavor violating transitions to the charged current interactions where they proceed via the known ckm angles .
however , this very restriction of the flavor violation to the sm one does not evade new sources of cp violation .
indeed , the model possesses various flavor
odd phases contained in the complex @xmath3 parameter , @xmath4 parameters , and gauge fermion masses @xmath5 .
these phases form the new sources of cp violation which shows up in the electric dipole moments ( edms ) of leptons and hadrons ( see @xcite and references therein ) . for heavy quark edms see @xcite and for the rate asymmetries of various heavy
light mesons @xcite .
therefore , it is of fundamental importance to determine appropriate collider processes where all or some of the susy cp phases can be inferred or measured .
in fact , the effects of the susy cp phases on the higgs production has already been analyzed in @xcite . in this work
we will discuss the chargino - pair production at lhc energies and ways of isolating the phase of the @xmath3 parameter from the cross section .
due to the large energy and high luminosity of incoming protons , lhc is a very useful machine to detect charginos .
the dominant production mechanism for chargino pairs at a hadron collider is the quark - antiquark annihilation . in this sense ,
the number of chargino pair production events differ from @xmath0 to @xmath6 colliders . in what follows
we will compute the cross section for @xmath7 as a function of @xmath8 $ ] for various values of @xmath9 and the su(2 ) gaugino mass @xmath10 . as the analyses of edms @xcite make it clear the effects of susy cp phases
are expected to grow with lowering soft masses .
our analysis is similar to that used for the linear collider processes@xcite .
the relevant feynman diagrams are depicted in fig .
1 . in what follows
we mainly deal with the first two diagrams since the third one is suppressed by presumably heavy squarks .
then it is obvious that the amplitude for the process depends exclusively on the phases in the chargino sector , @xmath12 , the phase of the @xmath3 parameter .
here we summarize the masses and couplings of the charginos for completeness ( see @xcite for details ) .
the charginos which are the mass eigenstates of charged gauginos and higgsinos are described by a @xmath13 mass matrix @xmath14 where @xmath10 is the su(2 ) gaugino mass taken to be real throughout the work .
the masses of the charginos as well as their mixing matrices follow from the bi - unitary transformation @xmath15 where @xmath16 and @xmath17 are @xmath13 unitary matrices , and @xmath18 , @xmath19 are the masses of the charginos @xmath20 , @xmath21 such that @xmath22 . it is convenient to choose the following explicit parametrization for the chargino mixing matrices @xmath23 where the angle parameters @xmath24 , @xmath25 , and @xmath26 can be determined from the defining eq .
a straightforward calculation yields @xmath27 in terms of which the remaining two angles @xmath28 and @xmath29 read as follows @xmath30}{\mbox{re}[q_{i}]}\ ] ] where @xmath31 and @xmath32 \nonumber \\ & + & m_{2}\cos \theta _ { l}\cos \theta _ { r}+|\mu |\sin \theta_{l}\sin \theta_{r}e^{i(\varphi _ { \mu } + \varphi _ { r}-\varphi _ { l } ) } \nonumber \\ q_{2 } & = & -\sqrt{2}m_{w}[\cos \beta \sin \theta _ { r}\cos \theta_{l}e^{-i\varphi _ { r}}+\sin \beta \cos \theta _ { r}\sin \theta_{l}e^{i\varphi _ { l } } ] \nonumber \\ & + & m_{2}\sin \theta _ { l}\sin \theta_{r}e^{i(\varphi_{l}-\varphi_{r})}+|\mu |\cos \theta_{l}\cos \theta
_ { r}e^{i\varphi _ { \mu } } \,.\end{aligned}\ ] ] the origin of the phases @xmath24 , @xmath25 , and @xmath33 is easy to trace back .
the angles @xmath34 and @xmath35 would be sufficient to diagonalize , respectively , the quadratic mass matrices @xmath36 and @xmath37 if @xmath38 were real . as a result one needs the additional phases @xmath25 which are identical to the phases in the off diagonal entries of the matrices @xmath36 and @xmath37 , respectively .
however , these four phases are still not sufficient for making the chargino masses real positive due to the bi - unitary nature of the transformation , and hence , the phases @xmath28 and @xmath39 can not also be made real positive . finally , inserting the unitary matrices @xmath16 and @xmath17 into the defining eq .
( 1 ) one obtains the following expressions for the masses of the charginos @xmath40^{1/2}\big\}.\end{aligned}\ ] ] the fundamental susy parameters @xmath10 , @xmath9 , @xmath41 and the phase parameter @xmath42 can be extracted from the chargino @xmath43 parameters @xcite @xmath44 the masses @xmath45 and the two mixing angles @xmath46 and @xmath47 of the left and right chiral components of the wave function .
these mixing angles are physical observables and they can be measured just like the chargino masses @xmath45 in the process @xmath48 .
the two angles @xmath49 and @xmath47 and the nontrivial phase angles @xmath50 @xmath51 @xmath52 define the couplings of the chargino - chargino - z vertices @xmath53\ ] ] @xmath54 @xmath55\ ] ] @xmath56\ ] ] @xmath57 @xmath58\ ] ] where @xmath59 is the weak angle .
note that every vertex here is an explicit function of @xmath60 via the various mixing angles . however , the @xmath61 coupling to unlike charginos @xmath62 is manifestly complex , and its phase vanishes in the cp conserving limit , @xmath63 @xmath64 .
obviously , the photon vertex is independent of the susy phases @xmath65 the process @xmath66 is generated by the two mechanisms shown in fig .
1 . the @xmath67-channel @xmath68 and @xmath61 exchanges , and @xmath69-channel @xmath70 exchange , where the latter is consistently neglected below .
the transition amplitude can be parameterized as @xmath71[\bar{u}(\tilde{\chi}_{i}^{-})\gamma ^{\mu } p_{\beta } v(\tilde{\chi}_{j}^{+})]\ ] ] where the charges @xmath72 are defined such that the first index corresponds to the chirality of the @xmath73 current and the second one to chargino current . for various final states , their expressions
are given by \(i ) for @xmath74 @xmath75 @xmath76 @xmath77 \(ii ) for @xmath78 @xmath79 @xmath80 @xmath81 \(iii ) for @xmath82 @xmath83 @xmath84 @xmath85 \(iv ) for @xmath86 @xmath87 @xmath88 @xmath89 \(v ) for @xmath90
@xmath91 @xmath92 @xmath93 \(vi ) for @xmath94 @xmath95 @xmath96 @xmath97 here all the amplitudes are built up by the @xmath98 and @xmath61 exchanges , and @xmath99 stands for the @xmath61 propagator : @xmath100 . in what follows , for convenience we will introduce four combinations of the charges @xmath101 \nonumber \\ q_{2 } & = & \frac{1}{2}re[q_{rr}q_{rl}^{\ast } + q_{ll}q_{lr}^{\ast } ] \nonumber \\ q_{3 } & = & \frac{1}{4}[\left| q_{rr}\right| ^{2}+\left| q_{ll}\right|^{2}-\left| q_{rl}\right| ^{2}-\left| q_{lr}\right| ^{2 } ] \nonumber \\ q_{4 } & = & \frac{1}{2}im[q_{rr}q_{rl}^{\ast } + q_{ll}q_{lr}^{\ast } ] \end{aligned}\ ] ] so that the differential cross section can be expressed simply as @xmath102q_{1}+4\mu _ { i}\mu_{j}q_{2}+2\lambda ^{1/2}q_{3}\cos \theta \}\ ] ] with the usual two body phase space factor @xmath103[1-(\mu_{i}-\mu _ { j})^{2}]\ ] ] defined via the reduced mass @xmath104 . integrating the differential cross section over the center of mass scattering angle @xmath105 one can obtain the total partonic cross section @xmath106 whose dependence on @xmath107 and @xmath9 will be analyzed numerically . the total cross section for the chargino - pair production through @xmath108 fusion at the lhc center of mass energy @xmath109 tev , can be obtained by integrating partonic cross section @xmath110 , over the quark - antiquark luminosities in the distribution function of proton , which is given in parton model by the formula@xcite @xmath111 with a quark luminosity in proton which is defined as @xmath112 where @xmath113 are the parton distribution functions for parton @xmath114 in hadron @xmath4 with momentum fraction @xmath115 evaluated at the factorization scale @xmath116 , and + @xmath117 . here , @xmath67 is the hadron - hadron center of mass energy squared which is related to @xmath118 , the parton - parton center of mass energy squared , via @xmath119 . to calculate the total cross sections for lhc , one has to know parton distributions as functions of the scaling variables @xmath120 and @xmath121 .
although the distributions have not been measured at such values of @xmath121 , one can obtain them using altarelli - parisi equation@xcite and the parton distributions at some scale @xmath122 . for this purpose , we used cteq4 m parton densities , without any qcd corrections for simplicity in our analysis@xcite .
these results are illustrated in figs .
2 - 4 , where the production cross - section is plotted against the phase angle at the lhc with @xmath123 tev .
besides these , it is necessary to analyze the rate asymmetries for having a better information about @xmath124 . analyzing the effects due to normal polarization of the charginos
, the cp - violating phase @xmath124 can be determined . concerning this point
, we investigate the normal polarization vector of the charginos which are inherently cp odd and therefore exist if cp is broken in the fundamental theory .
the normal polarization vector is defined as @xmath125 for @xmath126 the @xmath127th chargino , and it is defined as @xmath128 & = & \pm 4\lambda ^{1/2}\mu_{j , i}(f_{r}^{2}-f_{l}^{2})\sin \theta \sin 2\phi _ { l } \nonumber \\ & & \times \sin 2\phi _ { r}\sin ( \beta _ { l}-\beta _ { r}+\gamma_{1}-\gamma _ { 2})/n\end{aligned}\ ] ] for non - diagonal pairs @xmath62 where @xmath129 . here
@xmath130\ ] ] and @xmath131 a non - vanishing @xmath132 will be sufficient to establish non - vanishing cp violation in the system .
therefore , the value of non - vanishing @xmath132 implies the strength of the cp invariance breaking in susy .
in this section we will discuss the dependence of the chargino production cross section on @xmath107 and @xmath9 at @xmath109 tev .
we apply everywhere the existing collider constraint that @xmath133 . in table 1
, we give the cross section values for @xmath134 and @xmath135 taking m@xmath136 @xmath137 gev , @xmath138 gev , @xmath139 @xmath140 @xmath141 @xmath142 and @xmath143 in the calculations . in figs . 2 and 3 we show the dependence of the cross sections @xmath144 and @xmath145 on @xmath60 , for m@xmath136 @xmath137 gev , @xmath138 gev , and @xmath146 the variation of the cross section makes it clear that , as @xmath124 varies from 0 to @xmath64 the cross section decreases gradually .
the more spectacular enhancement implies the lighter chargino mass .
the decrease of the cross section is tied up to the variation of the chargino masses with the phases .
it is clear that as @xmath147 the mass splitting of the charginos decrease .
this is an important effect which implies that the cross section is larger than what one would expect from the cp conserving theory @xcite .
apart from the cross section itself , one can analyze various spin and charge asymmetries which are expected to have an enhanced dependence on @xmath124 .
the normal polarization in @xmath148 is zero since the @xmath149 and @xmath150 vertices are real even for non - zero phases in the chargino mass matrix . in fig .
6 we show the normal polarization @xmath132 of unlike charginos in @xmath151 which has a different dependence on the phases . here
again m@xmath152 gev , @xmath153 gev , @xmath154 and @xmath155 the value of @xmath156 is chosen since the normal polarization is maximum at @xmath157 .
this result will be discussed in the next section .
the dependence of the normal polarization on the value of @xmath158 and @xmath159 is shown in fig .
7 , where the normal polarization has its maximum at @xmath160 as expected from the equation 24 , and at @xmath161 as stated above .
we believe that for clarifying the essence of measuring @xmath162 the first quantity to be tested is the cross section itself .
we have analyzed the production of chargino pairs at lhc energies at the aim of isolating the phase of the @xmath3 parameter .
our results ( figs
. 2 - 3 ) suggest that , there is a strong dependence on the phase of the @xmath3 parameter .
the cross section is minimum for the value of the phase @xmath163 .
this result is understandable since the mass of the lightest ( heaviest ) chargino makes a maximum ( minimum ) at this point .
it is particularly clear that @xmath164 production rate is depleted in this region .
the size and dependence on @xmath165 of the cross section both get enhanced for small enough soft masses , @xmath44 @xmath10=150 gev and @xmath1=150 gev . as an example , this is seen for the value of @xmath163 , @xmath166 when @xmath10=150 gev and @xmath1=150 gev and for the same value of @xmath60 , @xmath167 when @xmath10=150 gev and @xmath1=300 gev .
the physical chargino mass increases as the values of @xmath10 and @xmath1 increase .
the cross section values that we have obtained for @xmath134 process at the lhc can reach a few 10 @xmath168 , whereas for the @xmath169 pair productions , the cross sections are in the range of a few @xmath168 . having an annual luminosity of 100 @xmath170 , one may accumulate @xmath171 and 100 events per year for @xmath172 and @xmath169 pair productions respectively .
this is a hard task to obtain a clean signature , but the measurement of these processes will be an important step for determining the cp violation sources of low energy supersymmetry .
the cross sections of the subprocess @xmath173 and @xmath174 as a function of @xmath108 c.m.s .
energy @xmath175 are depicted in figs .
there are sharp rising peaks around @xmath175 @xmath176 300 gev and @xmath175 @xmath176 500 gev for @xmath177 and @xmath178 respectively , due to the threshold conditions @xmath175 @xmath176 @xmath179 ( i , j=1,2 ) . since the cross sections for
@xmath180 have their peaks at the energies less then 1 tev , the upgraded tevatron as well , with a c.m.s energy of 1.8 tev and 30 @xmath181 integrated luminosity , is a useful machine to detect cp violation of susy particles .
in fact , it is more likely to see chargino pairs at such @xmath6 colliders than at the lhc .
this follows from the fact that the anti - quarks needed for chargino pair production are always drained from the sea for @xmath0 colliders . in this sense
our analysis puts a lower bound on the likelihood of observing chargino pairs at hadron colliders . in this respect , the proton - antiproton collider tevatron and , to a greater extent ,
the proton - proton collider lhc , are the almost ideal places of probing such a minimal susy scenario of explicit cp violation . in fig .
6 , we show the dependence of the normal polarization of @xmath182 on @xmath158 , for m@xmath136 @xmath137 gev , @xmath138 gev , and @xmath146 as seen from the figure , @xmath132 is not vanishing for masses of lighter charginos , and this is an indication of the cp invariance breaking in susy . for the higher values of m@xmath183 and @xmath1 , @xmath132 is practically vanishing as the mass of charginos increases .
in true experimental environment , the cross sections we have studied are for subprocess obtained by integration over appropriate structure functions .
however , given the energy span of lhc that it will be possible to probe sparticles up to @xmath184 , it is clear that the center of mass energies we discuss are always within the experimental reach .
if the experiment concludes @xmath185 then , given strong bounds from the absence of permanent edms for electron , neutron , atoms and molecules , one would conclude that the first two generations of sfermions will be hierarchically split from the ones in the third generation . in case
the experiment reports a small @xmath165 then presumably all sfermion generations can lie right at the weak scale in agreement with the edm bounds . in this case , where @xmath165 is a small fraction of @xmath186 , one might expect that the minimal model is uv completed above the tev scale such that the @xmath1 parameter is promoted to a dynamical sm
singlet field ( @xmath187 the @xmath188 models ) .
f. gabbiani , e. gabrielli , a. masiero and l. silvestrini , nucl .
b * 477 * , 321 ( 1996 ) [ arxiv : hep - ph/9604387 ] .
v. d. barger , t. falk , t. han , j. jiang , t. li and t. plehn , phys .
d * 64 * , 056007 ( 2001 ) [ arxiv : hep - ph/0101106 ] .
d. a. demir and m. b. voloshin , phys .
d * 63 * , 115011 ( 2001 ) [ arxiv : hep - ph/0012123 ] .
d. a. demir , a. masiero and o. vives , phys .
d * 61 * , 075009 ( 2000 ) [ arxiv : hep - ph/9909325 ] ; phys .
* 82 * , 2447 ( 1999 ) [ arxiv : hep - ph/9812337 ] ; s. w. baek , j. h. jang , p. ko and j. h. park , phys .
d * 62 * , 117701 ( 2000 ) [ arxiv : hep - ph/9907572 ] ; s. w. baek and p. ko , phys .
lett . * 83 * , 488 ( 1999 ) [ arxiv : hep - ph/9812229 ] ; d. a. demir and k. a. olive , phys .
d * 65 * , 034007 ( 2002 ) [ arxiv : hep - ph/0107329 ] ; m. boz and n. k.
pak , phys .
b * 531 * , 119 ( 2002 ) [ arxiv : hep - ph/0201199 ] .
d. a. demir , phys .
d * 60 * , 055006 ( 1999 ) [ arxiv : hep - ph/9901389 ] ; a. pilaftsis and c. e. wagner , nucl .
b * 553 * , 3 ( 1999 ) [ arxiv : hep - ph/9902371 ] ; t. ibrahim and p. nath , phys .
d * 63 * , 035009 ( 2001 ) [ arxiv : hep - ph/0008237 ] .
d. a. demir , phys .
b * 465 * , 177 ( 1999 ) [ arxiv : hep - ph/9809360 ] ; g. l. kane and l. t. wang , phys .
b * 488 * , 383 ( 2000 ) [ arxiv : hep - ph/0003198 ] .
v. d. barger , t. han , t. j. li and t. plehn , phys .
b * 475 * , 342 ( 2000 ) [ arxiv : hep - ph/9907425 ] ; s. y. choi , j. kalinowski , g. moortgat - pick and p. m. zerwas , eur .
j. c * 22 * , 563 ( 2001 ) [ addendum - ibid .
c * 23 * , 769 ( 2002 ) ] [ arxiv : hep - ph/0108117 ] ; s. y. choi , a. djouadi , m. guchait , j. kalinowski , h. s. song and p. m. zerwas , eur .
j. c * 14 * , 535 ( 2000 ) [ arxiv : hep - ph/0002033 ] ; phys .
b * 479 * , 235 ( 2000 ) [ arxiv : hep - ph/0001175 ] ; s. y. choi , a. djouadi , h. s. song and p. m. zerwas , eur .
j. c * 8 * , 669 ( 1999 ) [ arxiv : hep - ph/9812236 ] ; s. y. choi , a. djouadi , h. k. dreiner , j. kalinowski and p. m. zerwas , eur . phys .
j. c * 7 * , 123 ( 1999 ) [ arxiv : hep - ph/9806279 ] .
l. kneur and g. moultaka , phys .
d * 61 * , 095003 ( 2000 ) [ arxiv : hep - ph/9907360 ] ; d. a. demir , phys .
d * 60 * , 095007 ( 1999 ) [ arxiv : hep - ph/9905571 ] ; p.
nath , phys .
lett . * 66 * , 2565 ( 1991 ) .
altarelli , g. , and g. parisi , 1977 , nucl .
b126 , 29 .
s.m.berman , j.d.bjorken and j.b.kogut , phys.rev.d4(1971)3388 ; e.eichten , i.hinchlife , k.lane and c.quigg , rev.mod.phys.56(1984)579 ; e58,1065(1986 ) | we compute the rates for @xmath0 annihilation into chargino - pairs via drell - yan process taking into account the effects of supersymmetric soft phases , at proton - proton collider .
in particular , the phase of the @xmath1 parameter gains direct accessibility via the production of dissimilar charginos .
the phases of the trilinear soft masses do not have a significant effect on the cross sections .
pacs numbers:14.80.ly , 12.15.ji , 12.60.jv |
the picture of resistive ( current - reducing ) phase slips was first discussed by little @xcite .
langer and ambegaokar then used a ginzburg - landau free - energy functional to analytically obtain the lowest free - energy saddle point between two current - carrying metastable states @xcite .
the time scale of the resistive phase slips have been formulated by mccumber and halperin @xcite . the theory developed in refs . @xcite and @xcite
is generally referred to as the langer - ambegaokar - mccumber - halperin ( lamh ) theory .
recently , new technique has been developed for fabricating superconducting nanowires . since resistive transition region broadens with decreasing cross - sectional area of the wire
, nanowires therefore become ideal samples for a more precise test of the lamh theory @xcite . for this purpose , a quantitative evaluation of the thermal - activation rate of phase slip events
is needed .
in particular , this is the case since the thermal rate serves as the background for distinguishing quantum fluctuations at low temperature , a topic of considerable basic scientific interest @xcite recently , the string method @xcite has been presented for the numerical evaluation of thermally activated rare events .
this method first locates the most probable transition pathway connecting two metastable states in configuration space .
this is done by evolving strings , i.e. , smooth curves with intrinsic parametrization , into the minimal energy path .
the transition rates can then be computed using an umbrella sampling technique which simulates the fluctuations around the most probable path . in this paper
we show that the string method can be employed as an efficient numerical tool for the study of thermally activated phase slips in thin superconducting wires below @xmath0 .
the system is modeled by a one - dimensional ( 1d ) time - dependent ginzburg - landau equation ( tdgle ) with a langevin noise term . applying the string method to this particular system
, we obtain the most probable transition pathway between two neighboring current - carrying metastable states .
this pathway continuously connects the langer - ambegaokar saddle - point state @xcite to a state in which the order parameter vanishes somewhere to allow a phase slip of @xmath1 , as first proposed by little @xcite .
we also give a numerically accurate algorithm to evaluate the prefactors for the rate of resistive phase slips .
to outline the string method @xcite , consider a system governed by the overdamped langevin equation @xmath2 where @xmath3 is the frictional coefficient , @xmath4 denotes the generalized coordinates @xmath5 , @xmath6 , @xmath7 , and @xmath8 is a white noise satisfying @xmath9 , with @xmath10 denoting the boltzmann constant and @xmath11 the temperature .
metastable and stable states are located in configuration space as the minima of the potential @xmath12 .
assume @xmath13 and @xmath14 are the two minima of @xmath15 . in terms of the topography of @xmath12 , the most probable fluctuation which can carry the system from @xmath13 to @xmath14 ( or @xmath14 to @xmath13 ) corresponds to the lowest intervening saddle point between these two minima .
the minimal energy path ( mep ) is defined as a smooth curve @xmath16 connecting @xmath13 and @xmath14 with intrinsic parametrization such as arc length @xmath17 , which satisfies @xmath18 where @xmath19 is the component of @xmath20 normal to the path @xmath16 .
this mep is the most probable pathway for thermally activated transitions between @xmath13 and @xmath14 . to numerically locate the mep in configuration space , a string @xmath21 ( a smooth curve with intrinsic parametrization by @xmath17 ) connecting @xmath13 and @xmath14
is evolved according to @xmath22 a re - parametrization is applied once in a while to enforce accurate parametrization by arc length .
the stationary solution of eq .
( [ string - evolution ] ) satisfies eq .
( [ mep - def ] ) which defines the mep .
once the mep is determined , the lowest saddle point is known and the transition rate can be computed by evaluating the fluctuations around the mep @xcite .
following kramers approach and its generalizations @xcite , the transition rate is given by @xmath23^{1/2 } \exp\left\{-\displaystyle\frac{1}{k_bt } [ v({\bf q}_s)-v({\bf q}_a)]\right\},\ ] ] where @xmath24 is the saddle point found at the mep , @xmath25 denotes the hessian of @xmath12 , and @xmath26 is the negative eigenvalue of @xmath27 .
( by definition , @xmath27 has one and only one negative eigenvalue . )
the determinant ratio in eq .
( [ thermal - rate ] ) are numerically obtained by linear interpolation as follows @xcite .
let @xmath28 and @xmath29 be two @xmath30 positive definite matrices and @xmath4 a column vector in @xmath31 space .
a harmonic potential parametrized by @xmath32 ( @xmath33 ) is constructed as @xmath34{\bf q},\ ] ] with the corresponding partition function given by @xmath35=(2\pi\epsilon)^{n/2 } \left(\det[(1-\alpha)f+\alpha g]\right)^{-1/2}.\ ] ] from the expectation value @xmath36 \exp\left[-\displaystyle\frac{1}{\epsilon}u^\alpha({\bf q})\right]\\ & = \left\langle \left[\displaystyle\frac{1}{2\epsilon}{\bf q}^t(f - g){\bf q}\right ] \right\rangle_\alpha : = q(\alpha ) , \end{array}\ ] ] we have @xmath37 it follows from ( [ partition - z ] ) that @xmath38 the expectation value @xmath39 can be numerically evaluated in the canonical ensemble governed by potential @xmath40 . in practice ,
the ensemble is generated by solving @xmath41 where @xmath42 is a white noise satisfying @xmath43 .
to apply the above technique to the present problem , it is noted that the hessian at the saddle point , @xmath27 , has a negative eigenvalue @xmath26 .
given this @xmath26 and the corresponding normalized eigenvector @xmath44 , the indefinite @xmath27 has to be modified to give a positive definite @xmath45 : @xmath46 where @xmath47 is a positive parameter .
it follows that @xmath48 and @xmath49 are related by @xmath50 if we remember that the determinant is the product of the eigenvalues . from the mep
@xmath16 parametrized by the arc length @xmath17 , the eigenvector @xmath44 can be obtained by evaluating @xmath51 at the saddle point , followed by a normalization , and @xmath26 is then computed from @xmath52^th({\bf q}_s){\bf u}_s$ ] .
the ratio @xmath53 can be readily computed according to eq .
( [ determinant - ratio ] ) because @xmath54 and @xmath45 are both positive definite .
the determinant ratio in the rate expression ( [ thermal - rate ] ) is then obtained from @xmath55
for a superconducting wire below @xmath0 , if the transverse dimension @xmath56 the coherence length @xmath57 , then the variations of the order parameter @xmath58 over the cross section of the wire are energetically prohibited .
the wire sample therefore becomes a 1d superconductor , with @xmath58 being a function of a single coordinate @xmath59 along the wire .
the ginzburg - landau free - energy functional is of the form @xmath60=\sigma\int dx\left [ \displaystyle\frac{k}{2}|\nabla\psi(x)|^2- \displaystyle\frac{\alpha_0(t_c - t)}{2}|\psi(x)|^2 + \displaystyle\frac{\beta}{4}|\psi(x)|^4\right],\ ] ] where @xmath61 is the cross - sectional area of the wire , @xmath62 with @xmath63 the effective mass of the cooper pair , and @xmath64 and @xmath65 are both phenomenological parameters .
the time evolution of @xmath58 is governed by the time - dependent ginzburg - landau equation ( tdgle ) @xmath66}{\delta
\psi}+\zeta = k\nabla^2\psi+\alpha_0(t_c - t)\psi -\beta|\psi|^2\psi+\zeta,\ ] ] where @xmath3 is a viscosity coefficient , and @xmath67 is a langevin white noise , with autocorrelation functions @xmath68 this noise generates a random motion of @xmath58 and stabilizes the equilibrium distribution , which is proportional to @xmath69/k_bt\right\}$ ] . for the convenience of presentation and computation ,
we use the dimensionless form @xmath70=\int d\bar{x}\left [ \displaystyle\frac{1}{2}|\bar{\nabla}\bar{\psi}(\bar{x})|^2- \displaystyle\frac{1}{2}|\bar{\psi}(\bar{x})|^2 + \displaystyle\frac{1}{4}|\bar{\psi}(\bar{x})|^4\right],\ ] ] for the free - energy functional . here the over bar denotes the dimensionless quantities , obtained with @xmath28 scaled by @xmath71 , @xmath58 by @xmath72 , and @xmath59 by the correlation length @xmath73 .
correspondingly , the dimensionless tdgle is of the form @xmath74}{\delta\bar{\psi}}+\bar{\zeta } = \bar{\nabla}^2\bar{\psi}+\bar{\psi}-|\bar{\psi}|^2\bar{\psi } + \bar{\zeta},\ ] ] in which the time is scaled by @xmath75 , and the dimensionless noise @xmath76 satisfies the autocorrelation functions @xmath77 throughout the remainder of this paper , all physical quantities are given in terms of the dimensionless quantities , using @xmath78 for the energy scale ( @xmath79 is the condensation energy density , where @xmath80 is the bulk critical field ) , @xmath72 for the @xmath58 scale , @xmath81 for the length scale , and @xmath82 for the time scale .
note that all the temperature effects are absorbed into these scales .
the over bar will be dropped in the remainder of the paper .
consider a closed superconducting ring .
periodic boundary condition for @xmath83 is imposed by @xmath84 where @xmath85 is the circumference of the ring .
metastable current - carrying states @xmath86 are obtained from the stationary ginzburg - landau equation @xmath87 as local minima of @xmath28 : @xmath88 where @xmath89 is the wave vector , @xmath90 is the amplitude , and @xmath91 is an integer . the dimensionless current density in the @xmath86 state is given by @xmath92 .
the metastability of @xmath86 requires @xmath93 . in the presence of thermodynamic fluctuations ,
the lifetime of these metastable states is finite .
when the lifetime is made sufficiently short , the decay of persistent current becomes observable .
the decay of persistent current in a superconducting ring may be explained by dividing the @xmath58-function space into different subspaces , each labeled by an integer @xmath91 , defined through @xmath94 according to the periodic boundary condition for @xmath95 . in each @xmath58-function subspace
there is a current - carrying metastable state @xmath86 , defined as a local minimum of @xmath28 . on the one hand , there are many low - energy configurations that are frequently accessed by the fluctuating system .
nevertheless , such low - energy fluctuations cause no change of the phase difference @xmath96 across the whole ring , and therefore the global phase coherence persists . on the other hand
, there exist thermodynamic fluctuations that lead to transitions between different @xmath58-function subspaces .
these fluctuations involve large amplitude fluctuations of @xmath58 .
heuristically , if @xmath58 vanishes somewhere , then @xmath96 may change ( slip ) by @xmath1 and hence the system moves from one subspace to another ( from @xmath86 to @xmath97 or @xmath98 ) @xcite .
since larger persistent current means higher free energy , transitions among different metastable states tend to reduce the persistent current on average .
large amplitude fluctuations usually cost free energies much higher than @xmath99 , therefore phase - slip events are rare .
only when the cross section of the ring is very small and the temperature is close enough to @xmath0 , the decay of persistent currents due to infrequent phase - slip fluctuations becomes observable .
based on the ginzburg - landau free - energy functional @xmath28 , langer and ambegaokar have derived the lowest saddle point @xmath100 between the two neighboring metastable states @xmath101 and @xmath102 @xcite .
this @xmath103 state corresponds to the most probable fluctuation which can carry the system from @xmath86 to @xmath98 ( or from @xmath98 to @xmath86 ) .
analytical expressions for the saddle - point state @xmath103 and the dimensionless energy barriers @xmath104-f[\psi_n]$ ] and @xmath105-f[\psi_{n-1}]$ ] are given in appendix [ app_la ] . in sec .
[ numerical ] , we will show that using the string method , @xmath103 and @xmath106 can be numerically obtained from the mep connecting @xmath101 and @xmath102 .
the time scale of the thermodynamic phase - slip fluctuations is determined by the langevin equation ( [ tdgle ] ) .
the transition rates @xmath107 for the transitions @xmath108 and @xmath109 can be written as @xmath110,\ ] ] where @xmath111 are the prefactors which fix the fluctuation time scale , and @xmath112 is the energy unit . based upon kramers formulation and its generalizations @xcite , mccumber and halperin
have derived an analytical expression for the prefactors @xmath111 @xcite . to our knowledge , numerical evaluation of @xmath111 has never been reported . here
we outline a numerical scheme for the evaluation of @xmath111 so that a complete solution of the lamh theory may be obtained .
results based on this scheme will be presented in sec .
[ numerical ] .
we adopt a two - component vector representation for the complex @xmath58 : @xmath113^t= \left[{\rm re}\psi(x)\;\;{\rm im}\psi(x)\right]^t.\ ] ] in terms of @xmath114 , the dimensionless form of the free - energy functional in eq .
( [ gl - free - energy1 ] ) becomes @xmath115=\int dx\left [ \displaystyle\frac{1}{2}(\nabla{\mbox{\boldmath$\eta$}})^2- \displaystyle\frac{1}{2}{\mbox{\boldmath$\eta$}}^2 + \displaystyle\frac{1}{4}({\mbox{\boldmath$\eta$}}^2)^2\right],\ ] ] and the dimensionless tdgle becomes @xmath116+{\mbox{\boldmath$\zeta$}}= \nabla^2{\mbox{\boldmath$\eta$}}+{\mbox{\boldmath$\eta$ } } -{\mbox{\boldmath$\eta$}}^2{\mbox{\boldmath$\eta$ } } + { \mbox{\boldmath$\zeta$}},\ ] ] in which the noise @xmath117^t$ ] satisfies the autocorrelation functions @xmath118 the vector forms for the metastable state @xmath86 and the saddle - point state @xmath103 can be easily obtained .
the hessian of @xmath119 $ ] is given by @xmath120=\left[\begin{array}{cc } -\nabla^2 - 1 + 3\eta_1 ^ 2+\eta_2 ^ 2 & 2\eta_1\eta_2\\ 2\eta_1\eta_2 & -\nabla^2 - 1+\eta_1 ^ 2 + 3\eta_2 ^ 2 \end{array}\right].\ ] ] the general expression ( [ thermal - rate ] ) for the thermal - activation rate can be directly applied to the phase - slip fluctuations , with some elaboration for the symmetry properties of the system . according to eq .
( [ thermal - rate ] ) , the dimensionless form of the prefactors @xmath111 in eq .
( [ mh - rate ] ) can be formally written as @xmath121^{1/2},\;\;\ ; \omega_{-}=\displaystyle\frac{|\lambda_s^{(1)}|}{2\pi } \left[\displaystyle\frac{\det h_n}{|\det h_s| } \right]^{1/2},\ ] ] where @xmath122 , @xmath123 , and @xmath124 are the three hessians evaluated at @xmath98 , @xmath86 , and @xmath103 according to @xmath125 $ ] in eq .
( [ hessian ] ) , and @xmath126 is the lowest ( negative ) eigenvalue of @xmath124 .
the free energy @xmath28 is invariant under the gauge transformation @xmath127 and the translational transformation @xmath128 .
( the gauge invariance in terms of @xmath58 is equivalent to the rotational invariance in terms of @xmath114 ) . as a consequence
, @xmath123 must have a zero eigenvalue @xmath129 ( the lowest one , other eigenvalues are all positive ) , with the corresponding eigenvector @xmath130 { \mbox{\boldmath$\eta$}}_n(x),\ ] ] where @xmath131 is the vector form for @xmath86 and @xmath132 is the normalization factor .
similarly , @xmath124 must have a zero eigenvalue @xmath133 , with the corresponding eigenvector @xmath134 { \mbox{\boldmath$\eta$}}_s(x),\ ] ] where @xmath135 is the vector form for @xmath103 and @xmath136 is the normalization factor .
the translational invariance of @xmath28 leads to another zero eigenvalue @xmath137 because the phase - slip center in the saddle - point state @xmath103 can be continuously shifted ( see eq .
( [ barrier - state ] ) ) .
the presence of the zero eigenvalues @xmath129 , @xmath133 , and @xmath137 requires some extra efforts in evaluating the prefactors in eq .
( [ formal - omega ] ) .
while the gauge invariance comes from an unphysical degree of freedom and hence @xmath129 and @xmath133 are simply discarded , the translational invariance , however , points to the fact that phase - slip fluctuations in different spatial regions are equally probable .
the total rate for a transition , say @xmath108 , should be obtained by summing over all the phase - slip fluctuations across the whole length of the system .
this is achieved as follows .
the prefactors @xmath111 in eq . (
[ formal - omega ] ) appear to diverge because of the presence of @xmath138^{-1/2}$ ] in @xmath139 .
this @xmath138^{-1/2}$ ] is contributed by the integral @xmath140 ^ 2\right)\right]= \left[\lambda_s^{(3)}\right]^{-1/2},\ ] ] where @xmath141 ^ 2 $ ] comes from a taylor expansion at @xmath135 for @xmath142 $ ] , representing the second - order term contributed by the component of @xmath143 in the direction of @xmath144 . here
@xmath145 is the normalized eigenvector corresponding to @xmath137 and @xmath146 is the coordinate in the direction of @xmath145 : @xmath147^t\left [ { \mbox{\boldmath$\eta$}}(x)-{\mbox{\boldmath$\eta$}}_s(x)\right].\ ] ] given @xmath148 , equation ( [ zero-1 ] ) may be rewritten as @xmath149^{-1/2}=\displaystyle\frac{1}{2\pi } \sqrt{\displaystyle\frac{\sigma\xi h_c^2(t)}{k_bt } } \int dc_s^{(3)}.\ ] ] it can be shown that @xmath150 is an integral proportional to the system length @xmath85 : @xmath151 , and hence @xmath152^{-1/2}=\displaystyle\frac{1}{2\pi } \sqrt{\displaystyle\frac{\sigma\xi h_c^2(t)}{k_bt}}\lambda l.\ ] ] thus the prefactors @xmath153 in eq .
( [ formal - omega ] ) are proportional to the system length , as required physically by the translational symmetry of the system . in ref .
@xcite an analytical expression has been derived for @xmath154 ( see appendix .
[ app_mh ] ) .
below we outline a method for evaluating @xmath154 numerically .
the eigenspace of @xmath124 corresponding to the zero eigenvalue is two - dimensional .
an orthonormal basis can be constructed from the saddle - point state @xmath155 .
the translational invariance of @xmath28 gives @xmath156 as an eigenvector of @xmath124 : @xmath157 .
the gauge invariance gives @xmath158 in eq .
( [ us2 ] ) as another eigenvector of @xmath124 : @xmath159 . both @xmath156 and @xmath158
are readily computed from @xmath155 numerically .
based on these two nonorthogonal eigenvectors , the eigenvector @xmath145 , normalized and orthogonal to @xmath158 , is obtained : @xmath160,\ ] ] where @xmath161 is the normalization factor and @xmath162 stands for the inner product @xmath163^t\nabla{\mbox{\boldmath$\eta$}}(x)$ ] .
consider an infinitesimal variation of @xmath155 , @xmath164 , where @xmath165 denotes an infinitesimal translation of the phase - slip center .
the corresponding change of the coordinate in the direction of @xmath145 is given by @xmath166 where @xmath154 is related to the normalization factor @xmath161 in eq .
( [ us3 ] ) by @xmath167^{-1}$ ] .
heuristically , eq .
( [ us3 ] ) defines @xmath145 to represent a special direction in the @xmath114-function space . along this direction , the change of @xmath155 is a `` pure '' translation of the phase - slip center , without any rotation of the global phase angle , which is an unphysical degree of freedom .
then in eq .
( [ dcs3 ] ) , the projection of @xmath168 onto @xmath145 measures the infinitesimal translation of the phase - slip center in the @xmath114-function space , along the physically nontrivial direction of @xmath145 . with the help of eq .
( [ dcs3 ] ) and @xmath169 , eq . ( [ zero-3 ] ) is obtained from eq .
( [ zero-2 ] ) . to summarize ,
the dimensionless expressions for the prefactors @xmath111 are obtained as @xmath170^{1/2},\\ \omega_{-}=\displaystyle\frac{|\lambda_s^{(1)}|}{4\pi^2 } \sqrt{\displaystyle\frac{\sigma\xi h_c^2(t)}{k_bt}}\lambda l \left[\displaystyle\frac{\det ' h_n}{|\det '' h_s| } \right]^{1/2 } , \end{array}\ ] ] where the zero eigenvalues @xmath129 and @xmath133 are omitted and eq .
( [ zero-3 ] ) is used for @xmath138^{-1/2}$ ] . here
@xmath171 in @xmath172 and @xmath173 indicates that the only zero eigenvalue is to be omitted when computing the determinant , and @xmath174 in @xmath175 indicates that the two zero eigenvalues are to be omitted when computing the determinant . using those relevant eigenvectors ( corresponding to the negative and zero eigenvalues ) , the matrices @xmath122 , @xmath123 , and @xmath124
can all be modified into positive definite matrices ( as expressed by eq .
( [ positive - definite ] ) ) , for which the determinant ratio can be computed according to eq .
( [ determinant - ratio ] ) .
the string method has been employed to calculate the meps connecting neighboring metastable states @xmath86 and @xmath98 .
all quantities in the numerical calculation are dimensionless .
the length of the system is @xmath176 .
the local minima of @xmath28 are given in eq .
( [ metastable ] ) with @xmath177 .
( the maximum @xmath178 allowed by @xmath179 is @xmath180 . )
we first show in fig [ fig - mep43 ] the mep which connects @xmath181 to @xmath182 .
the string is discretized by @xmath183 points in the @xmath83-function space .
the initial string is taken from a linear interpolation between @xmath184 and @xmath185 . in order to reach the mep ,
the string is evolved toward the steady state according to eq .
( [ string - evolution ] ) , with the potential force given by @xmath186= \nabla^2{\mbox{\boldmath$\eta$}}+{\mbox{\boldmath$\eta$ } } -{\mbox{\boldmath$\eta$}}^2{\mbox{\boldmath$\eta$}}.\ ] ] during this process , the string is re - parametrized by arc length every 10 steps . in the calculation , @xmath83 is represented by a column vector of @xmath187 entries , with the @xmath59 interval @xmath188 $ ] discretized by a uniform mesh of @xmath189 points .
spatial derivatives in the potential force are discretized using central finite difference . to fix the global rotation of the system ,
a spring force is applied to the endpoint order parameter @xmath190 . in the form of @xmath191 with @xmath192 ,
this force restricts @xmath190 to the real axis .
the first column in fig [ fig - mep43 ] displays a sequence of the configurations along the mep from @xmath184 to @xmath185 , and the second column displays the corresponding sequence of @xmath193 . along this particular mep
, there is a phase slip of @xmath1 , nucleated in the middle of the system . through this phase
slip , the winding number changes from @xmath194 to @xmath195 . from fig .
[ fig - mep43 ] , it is seen that @xmath193 first decreases and reaches zero somewhere ( at @xmath196 , see the fourth figure from the top ) , then the phase slip occurs and @xmath193 rebounds to accomplish the transition .
the third figure from the top shows the saddle point @xmath100 between @xmath184 and @xmath185 , which has a locally diminished amplitude and possesses the highest energy along the mep .
little @xcite first pointed out that a persistent current in a closed loop will not be destroyed , `` unless a fluctuation occurs which is of such an amplitude that the order parameter is driven to zero for some section of the loop '' .
however , the configuration of a vanishing order - parameter amplitude somewhere does not necessarily correspond to the lowest saddle point between two current - carrying metastable states . using the stationary ginzburg - landau equation ,
langer and ambegaokar @xcite have obtained the analytical solution for the free - energy saddle point .
they also pointed out the following : `` it is plausible that , from this state of locally diminished amplitude , the system will run downhill in free energy through a configuration in which the amplitude vanishes somewhere , and finally will achieve the configuration in which one less loop in @xmath58 occurs across the length @xmath197 . '' this picture about the transition pathway has been quantitatively confirmed by the mep obtained here . for comparison ,
we have carried out direct simulations for the motion of @xmath58 in the presence of thermal noise , using the stochastic equation ( [ tdgle2 ] ) . for @xmath198
, reasonably clean transition pathways can be obtained from the rare transition events which carry the system from one metastable state to the other .
figure [ fig - path43 ] displays a sequence of @xmath83 and @xmath193 , collected along a transition pathway from @xmath184 to @xmath185 , calculated for @xmath176 and @xmath199 . a comparison of figs .
[ fig - mep43 ] and [ fig - path43 ] shows remarkable similarities .
the advantage of a mep is also seen from this comparison : as a smooth path in configuration space , the mep reveals the transition behavior better than those noisy pathways obtained from stochastic simulations . while some fine features of the transition may be lost due to the noise in stochastic simulations , they can be well preserved in the mep . in particular , in order to obtain a clean pathway from stochastic simulation , the temperature must be kept low enough to reduce local fluctuations , but a low temperature inevitably makes the transition events rare and difficult to catch , thus requiring very long simulation time .
figure [ fig - ener43 ] shows the energy variation along the mep from @xmath181 to @xmath182 .
the dimensionless free - energy barrier for the transition @xmath200 is obtained as @xmath104-f[\psi_4]=0.338 $ ] .
figure [ fig - energy ] shows the energy profile along the mep from @xmath201 to @xmath202 .
this mep consists of @xmath203 segments , each connecting two neighboring metastable states @xmath86 and @xmath98 , with @xmath91 running from @xmath204 to @xmath205 . in calculating the prefactor @xmath206 in eq .
( [ real - omega ] ) for @xmath194 , we use the following procedure : \(1 ) from the mep calculated in sec .
[ mep ] , the minimum @xmath131 and the saddle point @xmath135 are obtained .
\(2 ) the ( unphysical ) degenerate directions @xmath207 at @xmath131 and @xmath208 at @xmath135 in the @xmath114-function space are obtained by simple rotation according to eqs .
( [ un1 ] ) and ( [ us2 ] ) .
\(3 ) the degenerate direction @xmath144 at @xmath135 in the @xmath114-function space is calculated using eq .
( [ us3 ] ) , and the parameter @xmath154 defined in eq .
( [ dcs3 ] ) is then obtained to be @xmath209 .
\(4 ) the unstable direction @xmath210 at @xmath211 is obtained from the normalized difference between two neighboring configurations , evaluated at the saddle point along the mep . the corresponding negative eigenvalue is obtained as @xmath212 ; \(5 ) the hessians @xmath123 and @xmath124 are modified to give two positive definite matrices @xmath213 \left[{\bf u}_n^{(1)}\right]^t\ ] ] and @xmath214\left[{\bf u}_s^{(i)}\right]^t\ ] ] where @xmath215 and @xmath216 s are all positive parameters ; in our calculation the parameters @xmath215 and @xmath216 s are set to be @xmath217 .
\(6 ) the ratio @xmath218 is calculated using eq .
( [ determinant - ratio ] ) .
figure [ fig - q - alpha ] shows the expectation value @xmath39 , defined in eq .
( [ expecta ] ) as a function of @xmath32 in the interval @xmath219 $ ] .
this interval of @xmath32 is discretized using a non - uniform mesh of 352 points .
since @xmath39 varies rapidly near @xmath220 and @xmath217 , more points are distributed near these two ends , with the grid size @xmath221 . in the middle of
@xmath219 $ ] , a larger grid size @xmath222 is used . for each @xmath32 ,
the stochastic equation ( [ eqn - sde ] ) is simulated with @xmath223 , and @xmath39 is obtained from a time average over @xmath224 realizations .
the calculated value of the ratio @xmath225 is @xmath226 . using this ratio and the computed negative eigenvalue at the saddle point
, we obtain the ratio @xmath227 .
other sets of values have also been used for @xmath215 and @xmath216 in eqs .
( [ eqn - modifyhn ] ) and ( [ eqn - modifyhs ] ) , and @xmath228 in eq .
( [ expecta ] ) , but the final result of @xmath229 is not sensitive to those values . using @xmath176 and @xmath230 ( used in stochastic simulation ) , and the numerical values of @xmath104-f[\psi_4]=0.338 $ ] , @xmath231 , @xmath232 , and @xmath227 obtained in secs .
[ mep ] and [ prefactor ] , we are ready to compute the rate ( in the unit of @xmath233 ) @xmath234^{1/2 } \exp\left[-\displaystyle\frac{\sigma\xi h_c^2(t)}{2\pi k_bt } \delta f_-\right],\ ] ] for the transition @xmath200 .
the exponential factor is @xmath235 .
the prefactor @xmath206 is evaluated as @xmath236 it follows that the dimensionless rate @xmath237 is approximately @xmath238 .
this value is in reasonable agreement with what has been estimated through stochastic simulations .
in this paper we have demonstrated that by using the string method , thermal transition rates as formulated in the lamh theory can be numerically evaluated , even at low temperatures . in particular , the pre - exponential factor
may also be determined to some precision .
thus the `` electrical resistance '' of a 1d superconductor may be evaluated quantitatively .
however , it has to be pointed out that quantum tunneling effect , which can be important at low temperatures , is not taken into account in the present formulation .
work is presently underway to show that quantum tunneling can be similarly treated through the string method , thus enabling a complete quantitative account of the current dissipation phenomenon in 1d superconductors .
this work was partially supported by hong kong rgc grant hkust6073/02p .
by definition , the the lowest saddle point @xmath100 between two neighboring metastable states @xmath86 and @xmath98 satisfies the stationary ginzburg - landau equation ( [ stationarygle ] ) .
langer and ambegaokar have obtained @xmath239-i\sqrt{2}k_s\right\}e^{ik_sx},\ ] ] where @xmath240 is a wave vector determined by the condition @xmath241 satisfying @xmath242 . note that the amplitude of @xmath100 is diminished in a small region around @xmath243 , the phase - slip center . here
we note that translating the phase - slip center from @xmath243 to @xmath244 will produce another saddle point @xmath245 , because the free energy of the system is translationally invariant . from the explicit expressions for @xmath86 and @xmath103 ,
the dimensionless energy barriers can be readily obtained : @xmath246,\\ \delta f_+=\displaystyle\frac{1}{4 } \left[\displaystyle\frac{8\sqrt{2}}{3}\sqrt{1 - 3k_s^2 } + 8k_s(1-k_s^2)\left(\pi-\tan^{-1}\displaystyle\frac { \sqrt{1 - 3k_s^2}}{\sqrt{2}k_s}\right)\right ] .
\end{array}\ ] ] here @xmath104-f[\psi_n]$ ] and @xmath105-f[\psi_{n-1}]$ ] .
since @xmath247 ( for @xmath248 ) , the transition from @xmath86 to @xmath98 is more probable than that from @xmath98 to @xmath86 . as a consequence , thermally activated phase slips are current - reducing dissipative process .
by writing the real and imaginary parts as the two components of a vector , the saddle - point state @xmath103 in eq .
( [ barrier - state ] ) can be written as @xmath249 \left[\begin{array}{c } \sqrt{1 - 3k_s^2}\tanh\left[\sqrt{\displaystyle\frac { 1 - 3k_s^2}{2}}x\right ] \\
-\sqrt{2}k_s \end{array } \right]\\ & = \left[\begin{array}{cc } \cos k_sx & -\sin k_sx \\
\sin k_sx & \cos k_sx \end{array}\right ] \tilde{\mbox{\boldmath$\eta$}}_s(x ) , \end{array}\ ] ] from which @xmath250 { \mbox{\boldmath$\eta$}}_s(x)+\left[\begin{array}{cc } \cos k_sx & -\sin k_sx \\ \sin k_sx & \cos k_sx \end{array}\right ] \nabla\tilde{\mbox{\boldmath$\eta$}}_s(x ) \end{array}\ ] ] is obtained .
this equation indicates that while @xmath251 is an eigenvector of @xmath124 corresponding to the zero eigenvalue , i.e. , @xmath252 by definition , it can be decomposed into two physically distinct components .
the first component , @xmath253 { \mbox{\boldmath$\eta$}}_s(x)=[\alpha_s^{(2)}]^{-1}{\bf u}_s^{(2)}(x),\ ] ] is the eigenvector corresponding to the zero eigenvalue @xmath133 , arising from gauge invariance .
the second component , @xmath254 \nabla\tilde{\mbox{\boldmath$\eta$}}_s(x)= \lambda{\bf v}_s^{(3)}(x),\ ] ] is the eigenvector corresponding to the zero eigenvalue @xmath137 , arising from translational symmetry . here a constant @xmath154 is introduced for normalization .
it is easily seen that the global rotation of the phase angle is achieved by changing @xmath135 in the direction of @xmath208 while the translation of the phase - slip center is achieved by changing @xmath135 in the direction of @xmath255 .
the normalization constant @xmath154 is determined by @xmath256 ^ 2 = \displaystyle\frac{2\sqrt{2}}{3}(1 - 3k_s^2)^{3/2}$ ] .
the change of @xmath155 due to a small change in the location of the phase - slip center , @xmath244 , is @xmath257 \nabla\tilde{\mbox{\boldmath$\eta$}}_s(x)dx_c= { \bf v}_s^{(3)}(x)dc_s^{(3)}.\ ] ] substituting eq .
( [ 2nd - component ] ) into eq .
( [ center - translation ] ) yields @xmath258 and thus @xmath259 .
therefore the sample length dependence arises naturally from the translational degeneracy .
we want to point out that in general , @xmath260 is not orthogonal to @xmath158 . from the inner products @xmath261 { \mbox{\boldmath$\eta$}}_s(x)\right]^t \left[\begin{array}{cc}0 & -1 \\ 1 & 0 \end{array}\right ] { \mbox{\boldmath$\eta$}}_s(x ) \\ = & \int dx \left[{\mbox{\boldmath$\eta$}}_s(x)\right]^t { \mbox{\boldmath$\eta$}}_s(x ) = ( 1-k_s^2)l-2\sqrt{2}\sqrt{1 - 3k_s^2 } , \end{array}\ ] ] @xmath262 \nabla\tilde{\mbox{\boldmath$\eta$}}_s(x)\right]^t \left[\begin{array}{cc } \cos k_sx & -\sin k_sx \\ \sin k_sx & \cos k_sx \end{array}\right ] \nabla\tilde{\mbox{\boldmath$\eta$}}_s(x ) \\ = & \int dx \left[\nabla\tilde{\mbox{\boldmath$\eta$}}_s(x)\right]^t \nabla\tilde{\mbox{\boldmath$\eta$}}_s(x ) = \displaystyle\frac{2\sqrt{2}}{3}(1 - 3k_s^2)^{3/2 } , \end{array}\ ] ] @xmath263 { \mbox{\boldmath$\eta$}}_s(x)\right]^t \left[\begin{array}{cc } \cos k_sx & -\sin k_sx \\ \sin k_sx & \cos k_sx \end{array}\right ] \nabla\tilde{\mbox{\boldmath$\eta$}}_s(x ) \\
= & 2\sqrt{2}k_s\sqrt{1 - 3k_s^2 } , \end{array}\ ] ] we obtain @xmath264^t{\bf v}_s^{(3)}(x)= \displaystyle\frac{3^{1/2}8^{1/4}k_s } { \left[(1-k_s^2)l-2\sqrt{2}\sqrt{1 - 3k_s^2}\right]^{1/2 } ( 1 - 3k_s^2)^{1/4}},\ ] ] which approaches zero as @xmath265 and/or @xmath266 .
so the two are orthogonal only in these limits . | current dissipation in thin superconducting wires is numerically evaluated by using the string method , within the framework of time - dependent ginzburg - landau equation with a langevin noise term .
the most probable transition pathway between two neighboring current - carrying metastable states , continuously linking the langer - ambegaokar saddle - point state to a state in which the order parameter vanishes somewhere , is found numerically .
we also give a numerically accurate algorithm to evaluate the prefactors for the rate of current - reducing transitions .
* current dissipation in thin superconducting wires : + accurate numerical evaluation using the string method * |
the study of electromagnetically bound hadronic pairs is an excellent method to probe qcd at very low energy .
opposite charge long - lived hadrons such as pions and kaons can form hydrogen - like atoms bound by the coulomb interaction .
the strong interaction leads to a broadening of the atomic levels and dominates the lifetime of the atom .
pion - pion interaction at low energy , constrained by the approximate 2-flavour su(2 ) ( u , d ) chiral symmetry , is the simplest and well understood hadron - hadron process @xcite .
the observation of @xmath6-atoms ( @xmath7 ) was reported in ref .
@xcite and a measurement of their mean life in ref .
@xcite .
the low energy interaction between the pion and the heavier ( strange ) kaon is a tool to study the more general 3-flavour su(3 ) ( u , d , s ) structure of hadronic interaction , which is not accessible in @xmath8-interactions .
a detailed study of @xmath1-interaction provides insights into a potential flavour ( u , d , s ) dependence of the crucial order parameter or quark condensate in chiral perturbation theory ( chpt ) @xcite .
a measurement of the @xmath1-atom ( @xmath9 ) lifetime was proposed already in 1969 @xcite to determine the difference @xmath10 of the @xmath5-wave @xmath1-scattering lengths , where the indices 1/2 and 3/2 refer to the isospin of the @xmath1-system .
the @xmath0-atom decays predominantly by strong interaction into the neutral meson pair @xmath11 ( @xmath12 ) .
the decay width of the @xmath1-atom in the ground state is given by the relation @xcite : @xmath13 @xmath14 is the lifetime of the atom in the ground state , @xmath15 the fine structure constant , @xmath16 the reduced @xmath0 mass , and @xmath17 = 11.8 mev / c the outgoing @xmath18 or @xmath19 3-momentum in the @xmath1 center - of - mass system .
the term @xmath20 @xmath21 ( 4 @xmath4 2 ) % @xcite accounts for corrections , due to isospin breaking and the quark mass difference @xmath22 .
hence a measurement of @xmath23 provides a value for the scattering length @xmath10 . the mean life of @xmath1-atoms is predicted to be 3.7 @xmath4 0.4 fs @xcite .
the width @xmath24 can also be determined from the @xmath5-wave phase shifts obtained from @xmath1-scattering , i.e. from the interaction of kaons with nucleons .
the @xmath5-wave phase shifts are , however , poorly known due to the absence of data below 600 mev / c and , correspondingly , the uncertainties in @xmath25 and @xmath26 are substantial .
the overall interaction is attractive ( attractive in the isospin 1/2 and repulsive in the 3/2 state ) .
the roy - steiner equations lead , with the available scattering data , to results @xcite that are neither consistent with the most precise measurements @xcite nor with predictions from chpt @xcite .
the method used by dirac is to produce pions and kaons with a high energy proton beam impinging on a thin target .
pairs of oppositely charged mesons may interact and form electromagnetically bound systems .
their subsequent ionization in the production target leads to mesons emerging from the target with low relative momentum ( thereafter called atomic pairs ) . the mean life of the @xmath1-atom
can then be calculated from the number of observed low - momentum pairs .
this method was first proposed in 1985 @xcite and was successfully applied to @xmath27-atoms in serpukhov on the u-70 synchrotron internal proton beam @xcite , and with dirac - i at the cern - ps on beam line t8 .
data from dirac - i lead to a mean life @xmath28 fs for @xmath6-atoms @xcite .
cross sections for @xmath9-production have been calculated in ref .
@xcite . in this paper
we report on the observation of @xmath1-atoms from the first data with dirac - ii .
details on the initial apparatus ( dirac - i ) used to study @xmath6-atoms can be found in ref .
@xcite . a sketch of the modified spectrometer ( dirac - ii ) used to collect the @xmath1 ( and more @xmath29 ) data is shown in fig .
[ fig : a0smp ] .
the 24 gev / c proton beam * ( 1 ) * from the cern - ps impinges on a 26 @xmath16 m @xmath30-target * ( 2)*. the spill duration is 450 ms with an average intensity of @xmath31 protons / spill .
the proton beam then passes through a vacuum pipe and is absorbed by the beam dump .
the secondary particles are collimated through two steel shielding blocks * ( 3 ) * and * ( 7 ) * , upstream of the microdrift chambers * ( 4 ) * and downstream of the ionization hodoscope * ( 6 ) * , respectively .
they pass through a vacuum chamber * ( 8) * and are bent by the 1.65 t field of the dipole magnet * ( 9)*. the two - arm spectrometer is tilted upwards with respect to the proton beam by an angle of 5.7@xmath32 .
positive particles are deflected into the left arm , negative ones into the right arm .
we now describe in more detail the upgrade which was performed to search for and study @xmath1-atoms @xcite .
the upstream detectors * ( 4 6 ) * , see fig .
[ fig : a0smp ] , were either replaced or upgraded . however , they were not yet fully operational during data taking , and were therefore not used in the analysis presented here .
the tracking is performed by drift chambers * ( 10 ) * which have a spatial resolution of 85 @xmath16 m .
the vertical hodoscope * ( 11 ) * consisting of 20 scintillating slabs with a time resolution below 140 ps is used for timing .
the horizontal hodoscope * ( 12 ) * , made of 16 horizontal scintillating slabs , is used for triggering by selecting oppositely charged particles with a vertical displacement smaller than 75 mm .
the @xmath33-erenkov detector * ( 15 ) * was already used previously to reject electrons and positrons .
the refractive index is @xmath34 and the average number of photoelectrons @xmath35 for particles with velocity @xmath36 = 1 .
the inner part of the original container had to be cut to clear space for the two new erenkov detectors needed for kaon identification .
since the momenta of the two mesons originating from the breakup of the @xmath1-atoms are very small in the center - of - mass system , they have similar velocities in the laboratory system , and hence kaons are less deflected than pions .
typical trajectories are shown in fig .
[ fig : a0smp ] .
the heavy gas @xmath37-erenkov detectors in both arms * ( 14 ) * identify pions but do not respond to kaons nor ( anti)protons @xcite .
four spherical and four flat mirrors each focus the light towards the phototubes .
the alignment of the mirrors was checked with a laser beam @xcite . to keep a constant refractive index of @xmath38 , the gas has to be cleaned permanently with a complex recirculating system @xcite .
the average number of photoelectrons is 28 for particles with @xmath36 = 1 .
the aerogel erenkov detector * ( 13 ) * in the left arm identifies kaons and rejects protons @xcite .
such a detector is required only in the left arm since the contamination from antiprotons in the right arm is small due to their low production rate .
the detector consists of three modules .
two modules of 12@xmath39 each ( refractive index @xmath40 ) cover the relevant momentum range between 4 and 8 gev / c .
the aerogel stacks are 42 cm high and are read out by two 5-photonis xp4570 photomutipliers with uv windows .
the typical number of photoelectrons is 10 for @xmath36 = 1 particles .
a third overlapping module with 13@xmath39 aerogel and @xmath41 covers the small angle region to reject protons with momenta above 5.3 gev / c .
the aerogel tiles are stacked pyramidally to increase the radiator thickness halfway between the photomutipliers and to compensate for light absorption . due to the low light yield for the @xmath41 module , and the strong uv - light absorption
, we use a wavelength shifter .
the aerogel tiles are sandwiched between tetratex foils on which tetraphenylbutadiene ( tpb ) has been evaporated @xcite .
the typical number of photoelectrons is 4 8 for @xmath36 = 1 particles .
the preshower detector * ( 16 ) * provides additional electron / hadron separation in the offline analysis .
a lead converter , typically 10 25 mm thick , is placed in front of a 10 mm thick scintillator .
an additional converter / scintillator is installed to compensate for the drop of efficiency of the n@xmath42-erenkov detector at small angles , in the region covered by the aerogel and heavy gas detectors .
the iron absorber * ( 17 ) * and the array of scintillation counters * ( 18 ) * are used to suppress muons .
pairs such as @xmath43 , @xmath6 , @xmath44 or @xmath45 are selected by a two - level trigger . for @xmath46- and @xmath47-candidates the first level trigger requires no signal in the heavy gas detector of the @xmath48-arm and in the n@xmath42-erenkov detectors .
the two tracks have to cross the same ( or one of the two adjacent ) slab(s ) in each of the horizontal hodoscopes .
the two trajectories being asymmetric ( see fig .
[ fig : a0smp ] ) , the kaon is required to cross slabs of the vertical hodoscope located in front of the heavy gas detector , while the associated pion has to fly at large angles in the opposite arm .
the second level trigger @xcite uses raw hits from the drift chambers and , based on a lookup table , rejects pairs with high relative momentum .
the accepted trigger rate is limited by the buffer memory of around 2000 events per spill .
as mentioned already , only detectors downstream of the dipole magnet are used here for event reconstruction .
the trajectories are determined by the drift chambers , the pattern recognition starting from the horizontal @xmath49-coordinate in the last plane and extrapolating back to the target .
a straight line is first fitted through the hits and extended into the magnet yoke . a deflection algorithm
@xcite calculates the slope and @xmath49-coordinate of the track at the magnet entrance , following the magnetic field map .
the trajectory is then extrapolated linearly to the target , with the constraint that the track origin has to coincide with the center of the beam spot .
this determines the momentum of the particle . for the vertical @xmath50-coordinate
the straight line from the drift chamber information is extrapolated back to the center of the beam spot at the production target @xcite .
the variable of interest in the following analysis is the relative momentum @xmath51 of the @xmath52-pairs in their center - of - mass systems . in the transverse plane , the resolution on the relative momentum @xmath53 ( typically 3 mev / c ) is dominated by multiple scattering , while the resolution on the longitudinal component @xmath54 ( @xmath55 1 mev / c ) is not affected @xcite . for further analysis
we use therefore only @xmath54 .
-mass distribution in the @xmath56-region .
the line shows the gaussian fit.,scaledwidth=40.0% ] the momentum calibration was cross - checked with tracks from @xmath57 decays .
figure [ lambda ] shows the invariant @xmath58-mass distribution .
a gaussian fit is applied leading to a mass of 1115.35 @xmath4 0.08 mev / c@xmath59 ( statistical error ) and a width of @xmath60 = 0.58 @xmath4 0.01 mev / c@xmath59 , dominated by momentum resolution .
we have verified that the new hardware and software were able to reproduce the signal from @xmath6-atoms .
details of the procedure for @xmath6 atoms can be found in ref .
figure [ pipiatoms ] shows the momentum distribution after background subtraction .
the enhancement at low @xmath54 corresponds to @xmath61 atomic @xmath6-pairs .
bound @xmath6-pairs are not expected above @xmath62 mev / c and indeed the distribution in fig .
[ pipiatoms ] is flat and compatible with zero , which validates the background subtraction method .
note that the present data can not be compared directly with those of ref .
@xcite because we used here only the downstream detectors , and a different event selection was performed .
-distribution measured with dirac - ii for part of the @xmath6 data after background subtraction .
the accumulation of events at low @xmath54 is due to @xmath6-atoms .
the curve is a gaussian fit to guide the eye.,scaledwidth=45.0% ]
figure [ diagrampairs ] shows the four mechanisms which contribute to the production of @xmath0-pairs .
accidental pairs are due to particles produced on different nucleons ( fig .
[ diagrampairs]a ) , non - coulomb - pairs are associated with the production of long - lived intermediate states ( fig . [ diagrampairs]b ) . on the other hand ,
@xmath0-pairs which interact electromagnetically form correlated coulomb - pairs ( fig .
[ diagrampairs]c ) , or atomic bound states ( fig .
[ diagrampairs]d ) .
the @xmath63 atoms , while traveling through the target , can either decay , be ( de)-excited or break up into @xmath64 @xmath0-pairs which emerge from the target with very low relative momentum .
we now describe the analysis steps @xcite . for prompt pairs
the time difference between the positive and negative spectrometer arm lies between 0.5 and 0.5 ns .
accidental pairs are first removed using the time information from the vertical hodoscopes .
accidental pairs are also needed for subsequent analysis and we select those pairs with a time difference between 12 and 6 ns . the choice of the negative sign avoids the contamination from slow protons . the events have then to satisfy the following criteria : * no electrons nor positrons , * no muons , * one drift chamber track per arm , * @xmath65 mev / c , * @xmath66 mev / c , * the momentum of the kaon lies between 4 gev / c and 8 gev / c , * the momentum of the pion lies between 1.2 gev / c and 2.1 gev / c . with the excellent time resolution of the vertical hodoscope pions , kaons and protons below 2.5
gev / c can be separated by time - of - flight @xcite . for the @xmath44 analysis the aerogel detector
is used in addition to remove protons in the positive arm , while for the @xmath45 analysis , the time difference between the negative and the positive arm measured with the vertical hodoscope has to be negative in order to remove protons faking pions . once the accidentals have been subtracted the prompt pairs ( @xmath67 ) are composed of the following three types : atomic - pairs ( @xmath64 ) , coulomb - pairs ( @xmath68 ) , and non - coulomb - pairs ( @xmath69 ) .
therefore @xmath70 however , the background arising from @xmath6- and @xmath71-pairs with misidentified particles must be considered since the kaon flux is much lower than the pion and proton fluxes .
pions can be selected with the heavy gas detector in coincidence and protons with the aerogel detector in anticoincidence .
we then determine @xmath54 by assigning to the pion ( or the proton ) the mass of a kaon .
for @xmath6 events this incorrect mass assignment shifts the @xmath54-distribution by 150 mev / c @xcite and therefore does not overlap with the @xmath54-distribution of @xmath1 events . the contribution from @xmath71-pairs , non - coulomb @xmath44-pairs and accidentals have a similar linear @xmath54-distribution .
we assume that the background due to coulomb uncorrelated pairs can be described by the @xmath54-distribution of accidentals , following a similar analysis for @xmath6-atoms @xcite .
the non - coulomb background also includes the background from @xmath58-pairs .
coulomb correlated pairs have to be simulated @xcite . to determine the contribution from coulomb- and non - coulomb pairs we select the momentum range @xmath72 20 mev / c , where no atoms are expected .
the @xmath54-distribution is @xmath73 where @xmath74 and @xmath75 are the ( normalized ) differential probabilities for coulomb correlated or uncorrelated pairs , respectively .
the fit variable @xmath36 is the corresponding number of correlated pairs .
we choose a bin size of 0.25 mev / c .
the @xmath76-function to be minimized is @xmath77 ^ 2,\nonumber\\ & & \label{kicarre44}\end{aligned}\ ] ] where @xmath78 are the corresponding statistical errors in the measured number of prompt pairs .
figure [ k+pi- fit]a shows the minuit results for coulomb- and non - coulomb contributions to @xmath44 events . since the shapes of both contributions are known , one can extrapolate into the @xmath79 mev / c signal region . the difference ( residuals ) between the data and the sum of both contributions
is plotted in fig .
[ k+pi- fit]b . above @xmath80 mev
/ c the residuals are consistent with zero , while the enhancement at low relative momentum is the first evidence for @xmath44-atoms .
a gaussian distribution describes adequately the low momentum enhancement observed in the @xmath6 data @xcite . to guide the eye
we also apply a gaussian fit here ( line in fig .
[ k+pi- fit]b ) .
the integral of the gaussian distribution contains 147 @xmath4 61 atomic @xmath44-pairs .
.left : number @xmath36 of coulomb - pairs outside the signal region ( @xmath81 mev / c ) , number @xmath82 of coulomb - pairs extrapolated into the signal region ( @xmath83 mev / c ) , and number @xmath84 of detected atomic - pairs from the residuals of the fit .
right : expected number of atoms @xmath85 , calculated from the number of detected coulomb - pairs , and expected number of atomic pairs @xmath86 using a breakup probability of 53% .
[ cols="^,^ , > , > , > , > , > " , ] a similar fit based on equ .
( [ kicarre44 ] ) is applied to @xmath45 events . however , the number of events is smaller , due to the lower production cross section for negative kaons .
the fit results are summarized in table [ final fit ] .
the errors on the number of coulomb - pairs @xmath36 are the full ( minos ) errors , while the errors on @xmath84 are given by the square roots of the measured bin contents , statistical fluctuations on the much larger monte - carlo sample being negligible .
figure [ totalrres ] shows the sum of the @xmath44 and @xmath45 residuals .
we obtain @xmath87 detected atomic pairs with a statistical significance of 3.2@xmath60 .
the systematic uncertainty is estimated to be around 5% , much smaller than the statistical one . and @xmath45 .
a gaussian fit has been applied ( solid line ) to illustrate the distribution of atomic - pairs.,scaledwidth=48.0% ] the evidence for the observation of @xmath1-atoms is strengthened by the observation of correlated coulomb - pairs which , a fortiori , implies that atoms have also been produced .
this can be seen as follows , without involving simulation : non - coulomb pairs have a similar @xmath54-distribution as accidentals . hence dividing the normalized distribution for prompt pairs by the one for accidentals one obtains the correlation function @xmath88 describing coulomb - pairs . the function @xmath88 , shown in fig .
[ corr ] for @xmath44 as a function of @xmath89 , is clearly increasing with decreasing momentum , proving that coulomb - pairs have been observed .
in the signal region ( @xmath79 mev / c ) one obtains 858 @xmath4 247 coulomb - pairs from the data in fig .
[ corr ] , without resorting to monte - carlo .
the same procedure can be applied to @xmath45 events , leading to 313 @xmath90 148 coulomb - pairs @xcite . as a function of @xmath89 for @xmath44-pairs .
the deviation from the horizontal dotted line proves the existence of coulomb correlated @xmath46-pairs.,scaledwidth=45.0% ] the ratio @xmath91 of the number of produced atoms to the number of coulomb - pairs with small relative momenta has been calculated : @xmath91 = 0.62 @xcite .
however , one needs to take into account the acceptance of the apparatus and the cuts applied in the analysis . by monte - carlo simulation @xcite
we determine the ratio @xmath92 = 0.24 between the number of atoms produced within acceptance and the number of detected coulomb - pairs below @xmath89 = 3 mev / c ( and below @xmath53 = 8
mev / c ) .
this then leads to the expected number @xmath85 of atoms , 204 @xmath4 59 for @xmath45 , and 74 @xmath4 35 for @xmath44 ( table [ final fit ] , right ) .
the uncertainty on @xmath92 is negligible .
the breakup probability @xmath93 relates the number of atoms to the number of atomic pairs .
a calculation of the breakup probability as a function of mean life ( fig . [ lifetime ] ) has been performed using the born approximation @xcite . for the predicted mean life of 3.7 @xmath4 0.4 fs @xcite
the corresponding breakup probability @xmath94 is 53% ( dotted line in fig . [ lifetime ] ) .
( [ breakup ] ) we then find the number @xmath86 of expected pairs given in table [ final fit ] , right .
these numbers are in good agreement with the number @xmath84 of observed atomic - pairs ( fifth column in table [ final fit ] ) . for the @xmath95 m pt - target as a function of mean life of @xmath1-atoms in the 1s - state .
the horizontal solid line is the measured breakup probability and the horizontal dashed line the @xmath96 lower bound corresponding to a lower limit of 0.8 fs for the mean life .
the excluded area ( 90% confidence level ) is shown in turquoise .
the horizontal dotted line gives the theoretical prediction @xcite .
, scaledwidth=30.0% ] conversely , we use the number of observed atomic - pairs @xmath84 from the @xmath76-minimization and the number @xmath82 of coulomb - pairs below @xmath97 3 mev / c ( table [ final fit ] ) to calculate the breakup probability @xmath94 from equ .
( [ breakup ] ) with @xmath98 .
the result for @xmath0 ( @xmath94 = 64 @xmath4 25 % ) is shown by the horizontal solid line in fig .
[ lifetime ] .
this leads to a lower limit for the mean life of @xmath1-atoms of @xmath99 = 0.8 fs at a confidence level of 90% .
this result can be translated into an upper limit @xmath10 @xmath55 0.58 @xmath100 at 90% confidence level , in agreement with predictions @xcite .
we have presented the first evidence for the production of @xmath1-atoms by detecting @xmath101 atomic - pairs .
the evidence is strengthened by the observation of correlated @xmath1 ( continuum ) coulomb - pairs from which the number of bound states ( atoms ) is predicted and found to be in agreement with observation . a lower limit on the mean life of 0.8 fs is established with a confidence level of @xmath102 .
we note that the choice of pt as production target was driven by the high breakup probability facilitating the observation of @xmath1-atoms .
data are now being collected for a more accurate measurement of the lifetime with e.g. a 98 @xmath16 m ni - target , for which the breakup probability is lower ( @xmath10335% according to ref .
@xcite ) but still rapidly rising around the predicted mean life of 3.7 fs .
the ultimate goal of the experiment is to measure the lifetime of @xmath1-atoms with a precision of about 20% @xcite .
we are grateful to the cern - ps crew , whose efforts permitted us to take advantage of a high quality beam . this work was supported by cern , the grant agency of the czech republic , the istituto nazionale di fisica nucleare ( italy ) , the grant - in - aid for scientific research from the japan society for the promotion of science , the ministry of education and research ( romania ) , the ministery of industry , science and technologies of the russian federation and the russian foundation for basic research , the direccin xeral de investigacin , desenvolvemento e innovacin , xunta de galicia ( spain ) , and the swiss national science foundation . 00 s. weinberg , phys .
17 ( 1966 ) 616 .
j. gasser and h. leutwyler , nucl .
b 250 ( 1985 ) 465 , 517 , 539 .
g. colangelo , j. gasser and h. leutwyler , nucl .
b 603 ( 2001 ) 125 . l. afanasyev et al . , phys .
b 338 ( 1994 ) 478 .
b. adeva et al . , phys .
b 619 ( 2005 ) 50 . s. descotes - genon , l. girlanda and j. stern , jhep 0001 ( 2000 ) 041 ; eur .
j. c 27 ( 2003 ) 115 .
bilenkii , nguyen van hieu , l.l .
nemenov and f.g .
tkebuchava , yad .
( 1969 ) 812 ; ( sov
. j. nucl .
( 1969 ) 469 ) .
j. schweizer , phys .
b 587 ( 2004 ) 33 .
p. bttiker , s. descotes - genon , b. moussallam , eur .
j. c 33 ( 2004 ) 409 .
p. estabrooks et al . ,
b 133 ( 1978 ) 490
. v. bernard et al . , nucl .
b 357 ( 1991 ) 129 . l.l .
nemenov , yad .
41 ( 1985 ) 980 ; ( sov
. j. nucl .
41 ( 1985 ) 629 ) .
o.e . gorchakov et al . , yad . fiz . 63 ( 2000 ) 1936 ( phys . at .
63 ( 2000 ) 1847 ) .
b. adeva et al . , nucl .
instr . meth . a 515 ( 2003 ) 467 .
b. adeva et al . , ( addendum to the dirac proposal ) cern - spsc-2004 - 009 , spsc - p-284 add.4 .
a. kuptsov , dirac note 08 - 01 , + http://dirac.web.cern.ch/dirac/i_notes.html .
v. brekhovskikh et al . , dirac note 08 - 02 , + http://dirac.web.cern.ch/dirac/i_notes.html .
s. horikawa et al . , nucl .
instr . meth . a 595 ( 2008 )
. y. allkofer et al . , nucl .
instr . meth . a 582 ( 2007 )
y. allkofer et al . , nucl .
instr . meth . a 595 ( 2008 ) 84 .
l. afanasyev et al . , nucl .
instr . meth . a 491 ( 2002 ) 376 .
d. drijard , m. hansroul , v. yazkov , the dirac offline user guide , + http://dirac.web.cern.ch / dirac/. y. allkofer , phd thesis , universitt zrich , 2008 . y. allkofer , a. benelli , l. tauscher , dirac note 07 - 09 , + http://dirac.web.cern.ch/dirac/i_notes.html . y. allkofer , a. benelli , l. tauscher , dirac note 07 - 08 , + http://dirac.web.cern.ch/dirac/i_notes.html .
b. adeva , et al , nucl .
. meth . a 491 ( 2002 ) 41 .
sakharov , zh .
fiz . 18 ( 1948 ) 631 .
r. lednicky , prep .
arxiv : nucl - th/0501065 . l. afanasyev and o. voskresenskaya , phys
b 453 ( 1999 ) 302 .
o. gorchakov , p. zrelov , v. yazko , geant - dirac simulation , + http://dirac.web.cern.ch / dirac/. | we present evidence for the first observation of electromagnetically bound @xmath0-pairs ( @xmath1-atoms ) with the dirac experiment at the cern - ps .
the @xmath1-atoms are produced by the 24 gev / c proton beam in a thin pt - target and the @xmath2 and @xmath3-mesons from the atom dissociation are analyzed in a two - arm magnetic spectrometer .
the observed enhancement at low relative momentum corresponds to the production of 173 @xmath4 54 @xmath1-atoms .
the mean life of @xmath1-atoms is related to the @xmath5-wave @xmath1-scattering lengths , the measurement of which is the goal of the experiment . from these first data we derive a lower limit for the mean life of 0.8 fs at 90% confidence level . ,
, , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , dirac experiment , exotic atoms , scattering length , @xmath1-scattering , chiral perturbation 36.10.-k , 32.70.cs , 25.80.nv , 29.30.aj |
high - redshift ( @xmath19 ) quasars are a powerful tool to study the early universe . in recent years
, more than 100 quasars at @xmath20 have been discovered . the sloan digital sky survey ( sdss ; * ? ? ?
* ) pioneered the searches of quasars at these redshifts , followed by the canada - france high - redshift quasar survey ( cfhqs ; * ? ? ?
* ) , the ukirt infrared deep sky survey ( ukidss ; * ? ? ?
* ) , and the panoramic survey telescope & rapid response system 1 ( pan - starrs1 ; * ? ? ?
* ) survey . to date over 40 @xmath0 quasars have been discovered based on the sdss imaging data ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the ukidss has discovered several quasars @xcite , including the most distant quasar known at @xmath21 @xcite .
the cfhqs found 20 quasars over @xmath22 deg@xmath3 of sky @xcite .
the pan - starrs1 covers 3@xmath23 steradians of the sky , and is now producing a large number of high - redshift quasars @xcite , including three quasars at @xmath24 @xcite .
most recently , the vista kilo - degree infrared galaxy ( viking ) survey , the dark energy survey ( des ) , the vst atlas survey , and the subaru high-@xmath25 exploration of low - luminosity quasars ( shellq ) project , have started to yield @xmath19 quasars ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the number of high - redshift quasar discoveries is increasing steadily .
meanwhile , the sample of bright high - redshift quasars , especially luminous sdss quasars , have been studied extensively in multiple wavelength bands from x - ray to radio .
these quasars are very luminous with @xmath26 mag .
deep optical spectra have revealed strong or even complete absorption in the ly@xmath10 forests , indicating that the redshift probed ( @xmath0 ) is close to the epoch of cosmic reionization ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
their infrared spectroscopy show that these luminous quasars harbor billion - solar - mass black holes and emit near the eddington limit , suggesting the rapid growth of central black holes at this early epoch ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the broad emission lines of these quasars exhibit solar or supersolar metallicity , implying that vigorous star formation and element enrichment have occurred in their broad - line regions ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in addition , observations in the mid / far - ir , mm / sub - mm , and radio wavebands have provided rich information about the dust emission and star formation in the host galaxies ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
? * ; * ? ?
? * ; * ? ? ?
* ; * ? ? ?
therefore , high - redshift quasars are a powerful probe for understanding black hole accretion , galaxy evolution , and the intergalactic medium ( igm ) state in the first billion years of cosmic time . in this paper
, we present nine quasars newly found in the sdss , including seven quasars in the sdss main survey area , one quasar in the sdss overlap regions , and one quasar in sdss stripe 82 .
the overlap regions are the regions with overlapping imaging in the sdss , which results in multiple observations of individual sources within these regions .
stripe 82 covers @xmath27 deg@xmath3 , and was repeatedly scanned 7090 times by the sdss imaging survey .
we describe these regions in section 2 . with the discovery of these nine quasars ,
we have completed our survey of @xmath0 quasars in the sdss footprint .
we summarize our survey of sdss quasars in the second half of the paper . with a total of 52 quasars
, we derive the quasar luminosity function ( qlf ) at @xmath0 , and in particular , improve the measurement of the qlf at the bright end .
the layout of the paper is as follows . in section 2 , we review our survey of @xmath20 quasars in the sdss . in section 3 ,
we present the nine new quasars . in section 4 ,
we summarize our complete sample of 52 sdss quasars and calculate the qlf at @xmath0 . in section 5 , we discuss the evolution of luminous quasars at high redshift and the quasar contribution to cosmic reionization at @xmath0 .
we summarize the paper in section 6 . throughout the paper , sdss magnitudes
are expressed in the ab system .
near - ir and mid - ir magnitudes are in the vega system .
we use a @xmath28-dominated flat cosmology with @xmath29 km s@xmath30 mpc@xmath30 , @xmath31 , and @xmath32 .
in this section , we briefly review our survey of @xmath20 quasars selected in the sdss . we will need this information for sections 3 and 4 .
the sdss is an imaging and spectroscopic survey of the sky using a dedicated wide - field 2.5 m telescope @xcite at apache point observatory .
imaging was carried out in drift - scan mode using a 142 mega - pixel camera @xcite which gathered data in five broad bands , @xmath33 , spanning the range from 3000 to 10,000 @xcite , on moonless photometric @xcite nights of good seeing .
the effective exposure time was 54.1 seconds .
an sdss run ( strip ) consists of 6 parallel scanlines ( camera columns ) for each of the five @xmath33 bands .
the scanlines are @xmath34 wide with gaps of roughly the same width , so two interleaving strips make a stripe .
sdss scanlines are divided into fields , and a field is the union of five @xmath33 frames covering the same region of sky .
the images were processed using specialized software @xcite , and are photometrically @xcite and astrometrically @xcite calibrated using observations of a set of primary standard stars @xcite on a neighboring 20-inch telescope .
the initial goal of the sdss imaging survey was to scan 8500 deg@xmath3 of the north galactic cap .
the total unique area was expanded to 14,555 deg@xmath3 , by adding @xmath355000 deg@xmath3 in the south galactic cap ( sgc ) @xcite .
@xcite discovered 19 @xmath0 quasars from the sdss photometry , primarily in the north galactic cap .
most of these quasars are bright ( @xmath2 mag ) , and were selected from single - epoch imaging data ( hereafter referred to as the sdss main survey ) .
they represent the most luminous quasars at @xmath19 .
however , there were main survey regions remaining unsearched , particularly in the sgc . in this paper
we report on the discovery of additional quasars found in these regions .
the quasar selection procedure in the main survey has been discussed in detail in the papers mentioned above . here
we briefly review the procedure . because of the rarity of high - redshift quasars and overwhelming number of contaminants , the procedure consists of four basic steps .
the first step is to select @xmath36-band dropout objects mainly in high galactic latitude @xmath37 .
sources with @xmath38 mag and @xmath25-band error @xmath39 mag ( roughly @xmath40 mag ) that were not detected in the @xmath41 bands are selected as @xmath36-dropout objects .
the simple color cut @xmath38 is used to separate quasars ( and cool brown dwarfs ) from the majority of stellar objects ( e.g. * ? ? ?
* ; * ? ? ?
* ) . beyond the limit of @xmath39 mag , the number of contaminants increases dramatically .
the second step is to remove false @xmath36-dropout objects and improve photometry .
all @xmath36-dropout objects are visually inspected , and false detections such as cosmic rays are removed . if necessary , we also take deeper imaging data to improve the @xmath36 and @xmath25-band photometry to reduce the number of contaminants .
the third step is to take near - ir ( usually @xmath42 band ) photometry of @xmath36-dropout objects with another telescope . in the @xmath43 versus @xmath44 color - color diagram ,
high - redshift quasar candidates are separated from brown dwarfs .
specifically , quasar candidates satisfy the criterion @xmath45 ( see also figure [ fig : colorselect ] ) . the final step is to take spectroscopic observations and identify quasar candidates .
in addition to the above ` standard ' survey to a limit of @xmath46 detections in the sdss @xmath25-band images , we also selected two small samples of quasar candidates using a ` non - standard ' method . the first sample consisted of candidates down to @xmath47 in the @xmath25-band images in part of the ukidss footprint .
this is similar to the test done by @xcite .
we used a more stringent color cut @xmath48 mag to reduce the number of contaminants caused by larger photometric uncertainties .
we further required that the candidates should be detected at a significance level of @xmath49 in the ukidss @xmath50 and @xmath42 bands .
two of the quasars in this paper were selected using this method .
the second ` non - standard ' sample consisted of several candidates with @xmath44 colors between 2.1 and 2.2 mag , slightly bluer than that used for the ` standard ' survey .
one quasar in this paper was selected using this method .
in addition to the single - epoch main imaging survey , the sdss also conducted a deep survey by repeatedly imaging a @xmath27 deg@xmath3 area on the celestial equator in the south galactic cap @xcite .
this deep survey stripe , or stripe 82 , roughly spans @xmath51 and @xmath52 , and was scanned 7090 times in total , depending on r.a . along the stripe .
the multi - epoch data have been used to construct co - added images that can reach two magnitudes deeper than the sdss single - epoch images ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
using these co - added data , we have found 12 @xmath20 quasars in stripe 82 @xcite .
these quasars have @xmath53 mag , and are on average two magnitudes fainter than those found in the main survey . in this paper
we present one new quasar found in stripe 82 .
the quasar selection procedure for stripe 82 is very similar to that for the main survey , except that the survey limit is @xmath54 mag ( @xmath55 mag ) instead of @xmath56 mag .
we have also carried out searches of @xmath0 quasars in the sdss overlap regions , the regions that were scanned by two or more sdss imaging runs .
the sdss imaging runs generally overlap each other , due to the survey geometry and strategy .
the imaging survey in drift - scan mode was along great circles , and had two common poles .
the fields overlap more substantially when they approach the survey poles .
in addition , the two interleaving strips that make any stripe overlap slightly , leading to repeat observations in a small area . furthermore ,
if the quality of a run , or part of a run , did not meet the sdss standard seeing and photometric criteria , the relevant region was re - observed , yielding duplicate observations in this region .
the total area of the overlap regions is more than one - fourth of the sdss footprint .
these overlap regions provide a unique dataset that allows us to select high - redshift quasars more than 0.5 mag fainter ( in the @xmath25 band ) than those found with the sdss single - epoch data .
we have discovered eight quasars in the overlap regions @xcite .
the selection procedure of overlap - region quasars is slightly different from those described above .
the image quality is usually different between the repeat runs . in the first step of the selection procedure ,
the magnitude limit for both ` primary ' and ` secondary ' detections is @xmath57 mag or @xmath58 mag ( @xmath59 detection ) .
the candidates are actually fainter than @xmath60 detections , because otherwise they would have been selected in the main survey .
we focused on high galactic latitude ( @xmath37 ) regions . using repeat observations
ensures that most @xmath36-dropout objects we select are physical sources , rather than artifacts or cosmic rays . in the second step ,
we take deeper @xmath36 and @xmath25 band images to improve photometry for @xmath36-dropout objects .
the rest of the selection procedure remains the same .
the details are given in @xcite .
in this section we present the discovery of nine new quasars in the sdss .
the basic information of the quasars , including their coordinates , redshifts , and broad - band ( @xmath61 ) photometry , is given in table 1 .
one of them ( sdss j083525.76 + 321752.6 ; hereafter we use j0835 + 3217 for brevity ) is found in the overlap regions , and another one sdss j211951.89004020.1 ( hereafter j21190040 ) is found in stripe 82 .
the other seven quasars were found based on the sdss single - epoch data . the naming convention for sdss sources is sdss [email protected] , and
the positions are expressed in j2000.0 coordinates . for brevity , we use jhhmm@xmath62ddmm in the following text .
cccccl j081054.32 + 510540.1 & 5.800.03 & 21.520.13 & 19.340.07 & 18.770.06 & see also @xcite + j083525.76 + 321752.6 & 5.890.03 & > 23.0 & 20.730.20 & 20.500.20 & overlap regions + j114338.34 + 380828.7 & 5.810.03 & 21.970.19 & 19.760.09 & 18.980.09 & see also @xcite + j114803.28 + 070208.3 & 6.3390.001 & 23.200.35 & 20.790.10 & 19.360.11 & see also s. warren et al .
( in prep . )
+ j124340.81 + 252923.9 & 5.850.03 & 23.080.29 & 20.220.10 & 19.210.12 & see also s. warren et al .
( in prep . )
+ j160937.27 + 304147.7 & 6.160.03 & > 22.5 & 20.260.13 & 19.390.14 & see also s. warren et al .
( in prep . )
+ j162100.92 + 515548.8 & 5.710.03 & 21.860.13 & 19.700.07 & 19.110.20 & + [email protected] & 5.870.03 & 23.990.27 & 21.680.10 & 20.870.12 & stripe 82 + j231038.88 + 185519.7 & 6.0030.001 & 21.660.25 & 19.210.09 & 17.940.05 & + we first present the observations of quasar candidates in stripe 82 , which were done in 2009 and 2010 .
the @xmath42-band photometry of @xmath36-dropout objects ( quasar selection procedure step 3 ) was made using the sao widefield infrared camera ( swirc ; * ? ? ?
* ) on the mmt .
the observing strategy is the same as that of @xcite .
the observing conditions were typical , with relatively clear skies and @xmath64 seeing .
the images were reduced using standard iraf routines .
we used bright ukidss or 2mass @xcite point sources in the same images for flux calibration .
based on the @xmath42-band photometry , the final sample of quasar candidates was selected .
we then used the mmt red channel spectrograph ( rcs ; * ? ? ?
* ) to identify these candidates .
the exposure time for each target was 2030 minutes , depending on the object brightness and weather conditions .
if a target was identified as a quasar , several further exposures were taken to improve the spectral quality .
the mmt rcs data were reduced using standard iraf routines .
the observations of quasar candidates in the sdss overlap regions were conducted in 2015 and 2016 .
deeper @xmath36 and @xmath25-band photometry of @xmath36-dropout objects ( quasar selection procedure step 2 ) was made using the wide - field optical imager 90prime on the 2.3 m bok telescope .
the 90prime images were reduced in a standard fashion using our own idl routines .
the details of the bok observations and data reduction can be found in @xcite . for the @xmath42-band photometry
, we used the ukidss data for any candidates that have significant ( @xmath49 in @xmath42 ) detections in the ukidss . for the other candidates
, we obtained their @xmath42-band photometry using the mmt swirc .
we used the mmt rcs and the double spectrograph ( dbsp ) on the hale 5.1 m telescope to identify quasar candidates and obtain high - quality optical spectra .
the hale dbsp data were reduced using standard iraf routines as well .
the observations of quasar candidates in the sdss main survey were done between 2010 and 2015 , except for j1621 + 5155 , which was observed in 2006 .
we used the bok/90prime to improve the @xmath36 and @xmath25-band photometry for the sample of @xmath36-dropout objects with @xmath65 mag . for the @xmath36-dropout objects with @xmath39 mag
, we simply used the sdss data .
the @xmath42-band photometry was made using the mmt swirc , or from the ukidss archive for the objects detected at @xmath49 in the @xmath42 band .
we then used the mmt rcs and the hale dbsp to identify quasar candidates and obtain high - quality optical spectra , as we did for the candidates in the overlap regions .
in addition , we took a deep optical spectrum for j2310 + 1855 in long slit mode using the multi - object double spectrograph ( mods ) on the large binocular telescope ( lbt ) .
the mods spectra were reduced using standard iraf routines .
we also obtained deep near - ir spectra for two quasars , j2310 + 1855 and j1148 + 0702 , using gemini gnirs and magellan fire , respectively .
the gnirs observation of j2310 + 1855 is part of our large gemini gnirs campaign of @xmath66 quasars ( gn-2015b - lp-7 ) .
the gnirs campaign is used to measure the rest - frame properties of a large sample of @xmath0 quasars , including uv continuum slopes , broad emission line properties , black hole masses and mass function , etc .
both gnirs and fire ( in echelle mode ) provide a simultaneous wavelength coverage of @xmath67 m in cross - dispersion mode .
the observing strategies for the two observations were the same .
we used the standard abba nodding sequence between exposures .
the exposure time at each nod position was 5 minutes , and the distance between the two positions was @xmath68 . before or after the exposure of each quasar , a nearby a or f
spectroscopic standard star was observed for flux calibration and to remove telluric atmosphere absorption .
the gnirs spectra were reduced using the iraf gemini package , and the details can be found in @xcite .
the fire spectra were reduced using an idl pipeline developed by the fire instrument team , and the details can be found in @xcite . from the above observations
, we took spectra for about 30 candidates , and identified nine quasars , including one quasar ( j0835 + 3217 ) in the sdss overlap regions , one quasar ( j21190040 ) in stripe 82 , and seven quasars in the main survey .
table 1 lists the coordinates , redshifts , and the broad - band photometry of the quasars .
column 1 shows the j2000 coordinates , or the source names .
column 2 shows the redshifts , which span the range @xmath69 .
the redshifts were mostly measured from the ly@xmath10 emission lines , or from the wavelength where the sharp flux decline occurs .
the measurements can be slightly biased towards higher redshifts due to the ly@xmath10 forest .
the redshift error of 0.03 quoted in column 2 is simply the scatter in the relation between ly@xmath10 redshifts and systemic redshifts at low redshift ( e.g. * ? ? ?
the uncertainties from our fitting process and wavelength calibration are negligible in comparison .
the redshift of j1148 + 0702 is measured from its mgii emission line ( see section 3.2.1 ) .
the redshift of j2310 + 1855 is measured from the co ( 65 ) observations by @xcite .
columns 3 through 5 show the @xmath36 , @xmath25 , and @xmath42-band photometry .
the @xmath36 and @xmath25-band photometry was taken from the sdss , or improved by the bok 90prime .
the @xmath42-band photometry was taken from the ukidss , or obtained from the mmt swirc .
these quasars span a brightness range of @xmath70 and a luminosity range of @xmath71 mag . among the seven quasars found in the main survey area
, three quasars were selected using the ` non - standard ' method mentioned in section 2.1 .
j1148 + 0702 and j1609 + 3041 are fainter than a @xmath60 detection in the sdss @xmath25-band images , and j1621 + 5155 has an @xmath72 color slightly bluer than the 2.2 mag limit .
two quasars , j0810 + 5105 and j1143 + 3808 , in table 1 , were independently discovered by the pan - starrs1 @xcite , as indicated in the last column .
in addition , j1148 + 0702 , j1243 + 2529 , and j1609 + 3041 were independently discovered by the ukidss ( s. warren et al .
, in preparation ; see also @xcite ) .
we also recovered j0100 + 2802 at @xmath73 discovered by @xcite , and two quasars , j1545 + 6028 at @xmath74 and j2325 + 2628 at @xmath75 , found by @xcite .
we missed j23560622 at @xmath76 in @xcite , because this quasar has @xmath77 mag in the sdss .
figure [ fig : spec ] shows the optical spectra of the nine quasars .
all spectra except j2310 + 1855 were observed with the mmt rcs . the total integration time per object except j21190040 was from 40 min to 80 min ( 20 min exposures ) , depending on the quasar brightness and observing conditions .
the total integration time for the faintest quasar j21190040 was 150 min , composed of five 30 min exposures .
the spectrum of j2310 + 1855 was obtained from the lbt mods , and the total integration time was 60 min .
each spectrum in figure [ fig : spec ] has been scaled to match the corresponding @xmath25-band magnitude in table 1 , thereby roughly placing it on an absolute flux scale ( although variability introduces uncertainty into this calibration ) .
the quasar rest - frame uv spectrum , from the ly@xmath10 emission line to the feii bump at @xmath78 , contains strong diagnostic emission lines and provides key information on the physical conditions and emission mechanisms of the broad - line region .
the rest - frame uv band is redshifted to the near - ir range for @xmath19 quasars .
as we mentioned earlier , we also obtained near - ir spectra for j1148 + 0702 ( the highest - redshift quasar in our sample ) and j2310 + 1855 ( the most luminous quasar of the nine ) using magellan fire and gemini gnirs , respectively .
figure [ fig : nirspec ] shows the two near - ir spectra .
the total on - source integration time for each object was 60 min , broken into 12 five - minute exposures .
each spectrum has been scaled to match the corresponding @xmath42-band magnitude in table 1 .
_ j0810 + 5105 ( z=5.80 ) , j1143 + 3808 ( z=5.81 ) , and j1243 + 2529 ( z=5.85)_. these quasars were discovered in the sdss main survey .
they are at relatively low redshift @xmath79 .
they all have prominent ly@xmath10 emission . j0810
+ 5105 has @xmath80 mag , making it one of the brightest @xmath0 quasars known .
_ j0835 + 3217 ( z=5.89)_. j0835 + 3217 was discovered in the sdss overlap regions .
it is relatively faint ( @xmath81 mag ) compared to those found in the main survey .
j0835 + 3217 has a narrow ly@xmath10 emission line . @xcite
reported on the discovery of eight quasars in the overlap regions , and also recovered eight previously known quasars in the same area .
j0835 + 3217 is the last one that we found in the overlap regions . _
j1148 + 0702 ( z=6.339 ) and j1609 + 3041 ( z=6.16)_. j1148 + 0702 and j1609 + 3041 were selected in a ` non - standard ' way , as mentioned in section 2.1 .
they are fainter than @xmath60 detections in the sdss @xmath25-band images .
j1148 + 0702 and j1609 + 3041 are the two highest - redshift quasars in this sample .
j1148 + 0702 is the second highest - redshift quasar found in the sdss . using the near - ir spectrum in figure [ fig : nirspec ]
, we estimate its central black hole mass based on the empirical scaling relations @xcite .
the masses from mgii and civ are @xmath82 m@xmath83 and @xmath84 m@xmath83 , respectively .
the redshift estimated from mgii is @xmath85 . _
j21190040 ( z=5.87)_. j21190040 is the faintest quasar in our sample , found in the sdss stripe 82 .
it has strong ly@xmath10 emission .
we previously discovered 12 quasars in stripe 82 @xcite .
j21190040 is the last one that we found in this area .
the quasars in stripe 82 form a statistically complete sample down to @xmath86 mag . _
j1621 + 5155 ( z=5.71)_. j1621 + 5155 was selected in a ` non - standard ' way , with @xmath87 mag .
it thus has the lowest redshift in our sample .
it is a weak line quasar without obvious ly@xmath10 emission in figure [ fig : spec ] .
it was not detected in moderate deep millimeter and radio observations @xcite . _
j2310 + 1855 ( z=6.003)_. j2310 + 1855 is the brightest quasar in our sample .
it is also one of the most luminous quasars in the full sdss @xmath0 quasar sample .
it has very weak ly@xmath10 emission .
weak line quasars seem to be common at @xmath0 ( e.g. * ? ? ?
* ; * ? ? ?
this quasar has been studied extensively in the mm / submm and radio bands ( e.g. * ? ? ?
its strong detections of the [ cii ] 158 @xmath88 m and co ( 65 ) lines provide a redshift measurement of @xmath89 .
the redshift calculated from it mgii emission line is @xmath90 .
we estimate its central black hole mass from its near - ir spectrum shown in figure [ fig : nirspec ] , and the masses from mgii and civ are @xmath91 m@xmath83 and @xmath92 m@xmath83 , respectively .
cccccccccccccc 1 & j000239.40 + 255034.8 & 5.82 & 21.56 & 18.99 & & > 16.5 & & & 16.20 & 15.45 & -27.61 & @xcite & m+o + 2 & [email protected] & 5.850 & 23.09 & 20.50 & & 19.87 & & & & & -25.86 & @xcite & s82 + 3 & [email protected] & 5.929 & 22.85 & 20.35 & & 19.43 & & & 16.72 & 16.03 & -26.04 & @xcite & o + 4 & j002806.57 + 045725.3 & 6.04 & 24.00 & 20.49 & 19.59 & 19.16 & 19.05 & 18.32 & & & -26.38 & @xcite & o + 5 & j010013.02 + 280225.8 & 6.30 & 20.84 & 18.33 & & 17.00 & 15.98 & 15.20 & 14.45 & 13.63 & -29.10 & @xcite & m + 6 & [email protected] & 5.779 & 24.48 & 22.13 & & 21.78 & & & & & -24.39 & @xcite & s82 + 7 & j014837.64 + 060020.0 & 5.923 & 22.25 & 19.31 & 18.91 & 18.37 & 17.72 & 17.13 & 15.90 & 15.09 & -27.08 & @xcite & m+o + 8 & j020332.38 + 001229.4 & 5.72 & 23.76 & 20.75 & 19.85 & 19.05 & 17.75 & 17.32 & 16.35 & 16.06 & -25.74 & @xcite & s82 + 9 & [email protected] & 5.82 & 24.51 & 22.08 & 21.62 & 21.15 & & & & & -24.50 & @xcite & s82 + 10 & [email protected] & 6.078 & 24.17 & 20.97 & 20.60 & 20.44 & 19.78 & 18.95 & & & -25.31 & @xcite & s82 + 11 & j035349.73 + 010404.6 & 6.072 & 23.22 & 20.51 & 20.12 & 19.45 & 18.53 & 18.16 & & & -26.49 & @xcite & s82 + 12 & j081054.32 + 510540.1 & 5.80 & 21.52 & 19.34 & & 18.77 & & & 16.94 & 15.87 & -26.98 & this paper & m + 13 & j081827.39 + 172251.8 & 6.02 & 22.62 & 19.67 & & 18.54 & & & & & -27.37 & @xcite & m + 14 & j083525.76 + 321752.6 & 5.89 & > 23.0 & 20.73 & & 20.50 & & & & & -25.76 & this paper & o + 15 & j083643.86 + 005453.2 & 5.810 & 20.97 & 18.71 & 18.27 & 17.70 & 17.02 & 16.18 & 15.23 & 14.46 & -27.86 & @xcite & m + 16 & j084035.09 + 562419.9 & 5.844 & 22.43 & 19.76 & & 19.00 & & & & & -26.64 & @xcite & m + 17 & j084119.52 + 290504.4 & 5.98 & 22.47 & 19.86 & 19.74 & 19.08 & 18.62 & 17.84 & & & -27.08 & @xcite & o + 18 & j084229.43 + 121850.5 & 6.069 & 23.31 & 19.56 & & 18.84 & & & 15.76 & 15.42 & -26.85 & @xcite & m+o + 19 & j085048.25 + 324647.9 & 5.867 & > 22.5 & 19.95 & & 18.90 & & & 16.39 & 15.18 & -26.74 & @xcite & o + 20 & j092721.82 + 200123.6 & 5.772 & 22.12 & 19.88 & & 19.01 & & & 16.66 & 15.68 & -26.78 & @xcite & m + 21 & j103027.09 + 052455.0 & 6.309 & 22.90 & 19.62 & 19.27 & 18.85 & 18.37 & 17.78 & 16.49 & 15.44 & -27.53 & @xcite & m+o + 22 & [email protected] & 5.778 & 21.68 & 19.07 & 18.87 & 18.31 & 17.92 & 17.03 & 16.36 & 15.56 & -27.61 & @xcite & m+o + 23 & j104845.05 + 463718.4 & 6.198 & 22.43 & 19.85 & & 18.40 & & & 16.26 & 16.24 & -27.51 & @xcite & m + 24 & j113717.72 + 354956.9 & 6.03 & 22.55 & 19.54 & & 18.41 & & & 16.29 & 15.78 & -27.08 & @xcite & m + 25 & j114338.34 + 380828.7 & 5.81 & 21.97 & 19.76 & & 18.98 & & & 16.93 & 16.03 & -26.76 & this paper & m + 26 & j114803.28 + 070208.3 & 6.339 & 23.20 & 20.80 & 19.74 & 19.36 & 18.39 & 17.51 & 16.39 & 15.48 & -26.41 & this paper & m + 27 & j114816.64 + 525150.3 & 6.419 & 23.18 & 19.98 & & 18.25 & & & 15.66 & 15.18 & -27.80 & @xcite & m + 28 & j120737.43 + 063010.1 & 6.040 & > 23.5 & 20.39 & 19.51 & 19.35 & & 17.50 & 16.53 & 14.82 & -26.60 & @xcite & o + 29 & j124340.81 + 252923.9 & 5.85 & 23.08 & 20.22 & 19.81 & 19.21 & 18.29 & 17.54 & 16.70 & 15.52 & -26.22 & this paper & m + 30 & j125051.93 + 313021.9 & 6.15 & 22.15 & 19.48 & 19.54 & 18.92 & 18.37 & 17.44 & & & -27.11 & @xcite & m + 31 & j125757.47 + 634937.2 & 6.02 & 23.50 & 20.60 & 20.39 & 19.78 & & & 16.71 & 16.48 & -26.14 & @xcite & o + 32 & j130608.25 + 035626.3 & 6.016 & 22.35 & 19.29 & 19.24 & 18.86 & 18.69 & 17.34 & 15.99 & 15.52 & -27.32 & @xcite & m + 33 & j131911.29 + 095051.3 & 6.132 & 22.55 & 19.99 & 19.10 & 18.76 & & & 16.73 & 15.71 & -27.12 & @xcite & o + 34 & j133550.80 + 353315.8 & 5.901 & 22.67 & 20.10 & 19.38 & 18.90 & & 17.61 & 16.81 & 16.13 & -26.81 & @xcite & m + 35 & j140319.13 + 090250.9 & 5.86 & 22.73 & 20.48 & 19.70 & 19.17 & 18.59 & 17.93 & 17.09 & 15.95 & -26.27 & @xcite & o + 36 & j141111.27 + 121737.3 & 5.927 & 22.88 & 19.58 & 19.56 & 19.20 & 18.28 & 17.45 & 16.76 & 15.61 & -26.75 & @xcite & m+o + 37 & j143611.73 + 500707.0 & 5.85 & 22.76 & 20.00 & & 19.04 & & & & & -26.51 & @xcite & m + 38 & j154552.08 + 602824.0 & 5.78 & 21.27 & 19.09 & & & & & 16.00 & 15.16 & -27.37 & @xcite & m + 39 & j160253.98 + 422824.9 & 6.09 & 22.88 & 19.81 & & 18.46 & & & 16.14 & 15.03 & -26.85 & @xcite & m + 40 & j160937.27 + 304147.7 & 6.16 & > 22.5 & 20.26 & 20.01 & 19.39 & 18.72 & 18.15 & 17.52 & 17.08 & -26.62 & this paper & m + 41 & j162100.92 + 515548.8 & 5.71 & 21.86 & 19.70 & & 19.11 & & & 15.71 & 14.86 & -26.94 & this paper & m + 42 & j162331.80 + 311200.6 & 6.247 & 24.50 & 19.67 & 19.72 & 19.16 & 18.45 & 17.86 & 16.85 & 15.44 & -27.04 & @xcite & m+o + 43 & j163033.89 + 401209.7 & 6.058 & 23.28 & 20.34 & & 19.38 & & & & & -26.14 & @xcite & o + 44 & j205321.77 + 004706.8 & 5.92 & 24.13 & 21.34 & & 20.46 & & & & & -25.54 & @xcite & s82 + 45 & [email protected] & 6.038 & 23.23 & 20.74 & & 19.18 & & & & & -26.09 & @xcite & s82 + 46 & [email protected] & 5.87 & 23.99 & 21.67 & & 20.87 & & & & & -24.73 & this paper & s82 + 47 & j214755.42 + 010755.5 & 5.81 & 24.21 & 21.61 & 20.92 & 20.79 & & & & & -25.00 & @xcite & s82 + 48 & j230735.36 + 003149.3 & 5.87 & 25.16 & 21.91 & 20.99 & 20.43 & & & & & -24.71 & @xcite & s82 + 49 & j231038.88 + 185519.7 & 6.003 & 21.66 & 19.21 & & 17.94 & & & 15.80 & 15.42 & -27.61 & this paper & m + 50 & [email protected] & 6.117 & 23.80 & 20.85 & & 19.94 & & & & & -25.41 & @xcite & s82 + 51 & j232514.25 + 262847.6 & 5.77 & 21.62 & 19.42 & & & & & 16.19 & 15.41 & -26.98 & @xcite & m + 52 & j235651.58 + 002333.3 & 6.00 & 24.64 & 21.74 & & 21.18 & & & & & -24.84 & @xcite & s82 +
in this section we summarize our survey of @xmath0 quasars in the sdss , and present the final sample of 52 quasars discovered since 2000 .
we then calculate the survey area coverage and the quasar sample completeness .
this information is used to derive the qlf at @xmath0 and the evolution of luminous quasars at high redshift .
table 2 gives the basic data for the 52 sdss quasars at @xmath0 .
they are ordered by r.a .
column 2 lists the quasar coordinates determined by the sdss .
column 3 shows the quasar redshifts , taken from different resources , including the quasar discovery papers or follow - up observation papers .
the redshifts were mostly measured from emission lines , such as ly@xmath10 in the ( observed - frame ) optical , mgii in the near - ir ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , or co in the radio ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
some quasars have very weak emission lines in the optical and near - ir ( i.e. , the rest - frame uv ) , and their redshifts were measured from the onset of the ly@xmath10 absorption . in column 2 , the redshifts measured from ly@xmath10 emission or absorption features are accurate to the second decimal place , and the redshifts measured from mgii or co lines are accurate to the third decimal place .
columns 49 shows the photometry in the @xmath93 bands , and columns 1011 shows the photometry in the first two wide - field infrared survey explorer ( wise ; * ? ? ?
* ) bands w1 and w2 at 3.4 and 4.6 @xmath88 m , respectively .
the optical magnitudes are expressed in the ab system , and the near - ir and mid - ir magnitudes are in the vega system .
column 12 is the absolute ab magnitude of the continuum at rest - frame 1450 ( @xmath94 ) .
we have converted all published @xmath94 values to the cosmology used in this paper .
column 13 shows the references of the quasar discovery papers .
column 14 indicates if a quasar is in the sdss main survey ( ` m ' ) , overlap region ( ` o ' ) , or stripe 82 ( ` s82 ' ) .
note that seven main - survey quasars are also located in overlap regions , and they are marked as ` m+o ' .
the majority of the 52 quasars were found by fan et al .
( 20002006 ) and jiang et al .
( 2008 , 2009 , 2015 , and this paper ) .
three quasars in the sample , j0100 + 2802 ( @xmath73 ) , j1545 + 6028 ( @xmath74 ) , and j2325 + 2628 ( @xmath75 ) , were reported by @xcite and @xcite .
they were discovered using combination of the sdss and wise imaging data ( see also * ? ? ?
* ; * ? ? ?
* ) , and were also selected by our standard selection criteria mentioned in section 2 .
we thus included these three quasars in our final sample .
we also included j0841 + 2905 ( @xmath95 ) found by @xcite and j1319 + 0950 ( @xmath96 ) found by @xcite .
these two quasars do not meet our selection cut @xmath39 mag in the sdss single - epoch images , but they are located in the overlap regions , and were recovered as overlap - region quasars by @xcite .
we did not include three quasars reported by @xcite , @xcite , and @xcite .
@xcite presented three @xmath0 quasars , and two of them meet our selection criteria as mentioned above .
the third one was not selected by us , because its @xmath97 is slightly larger than 0.1 mag .
we did not include the quasar ( at @xmath98 ) of @xcite , which is significantly fainter than our selection limit in the sdss images .
we did not include the radio - loud quasar ( at @xmath99 ) of @xcite , which was found from its radio emission .
it is bright in the optical , but it is strongly blended with a much brighter neighbor in the sdss images , and was not separately detected by the sdss pipeline .
we also did not include the radio - loud quasar ( at @xmath100 ) of @xcite found in stripe 82 , which is fainter than our selection limit in the co - added stripe 82 images .
figure [ fig : redshift ] shows the redshift distribution of the 52 sdss quasars .
the number of quasars above our flux limits decreases rapidly from @xmath101 to @xmath102 .
figure [ fig : allspec ] in the appendix shows their optical spectra .
most of the spectra were taken from the quasar discovery papers listed in column 13 of table 2 .
the spectrum of j1319 + 0950 was presented in @xcite .
the spectrum of j0841 + 2905 was obtained from the mmt .
note that there exist higher s / n optical spectra for some quasars ( e.g. * ? ? ?
* ; * ? ? ?
* ) that are not shown in figure [ fig : allspec ] . among the 52 quasars ,
47 quasars belong to one of three statistically complete samples : the main survey sample , the overlap region sample , and the stripe 82 sample .
there are 24 quasars in the sdss main survey , 17 in the overlap regions ( 7 of them also belong to the main survey ) , and 13 in stripe 82 .
the other 5 quasars are beyond our standard selection criteria and not part of the complete samples .
they are j1335 + 3533 and j1436 + 5007 from @xcite , and j1148 + 0702 , j1609 + 3041 , and j1621 + 5155 in this paper .
we will now derive the qlf based on the 47 quasars .
in this subsection , we calculate the effective area coverage for our quasar samples .
the calculation of effective area is not straightforward for several reasons , including the sdss imaging survey geometry , possible missing data , existence of very bright stars ( resulting in large ` holes ' in object catalogs ) , and so on .
we estimate the effective area using the hierarchical equal area isolatitude pixelization ( healpix ; * ? ? ?
healpix hierarchically tessellates the spherical sky into a mesh of curvilinear quadrilaterals .
the base resolution , or the lowest resolution level ( we call it level 0 ) , consists of 12 pixels over the celestial sphere . the resolution level increases by dividing each pixel into 4 subpixels with identical area .
at level higher than 1 , each pixel has 8 neighboring pixels , except for 24 pixels ( each of them has 7 neighboring pixels ) .
the effective area of our samples is calculated by adding up the area of the pixels that cover our data points . obviously , the accuracy depends on the resolution level of the healpix pixels and the spatial density of data points .
our data points are drawn from the sdss query casjobs online server .
we use all sdss ` primary ' objects with @xmath103 mag and @xmath104 mag for the main survey .
we only remove a tiny fraction of objects with the sdss processing flags ` bright ' , ` edge ' , and ` satur ' . for the overlap regions ,
we use all ` primary ' and secondary objects down to @xmath105 mag and @xmath106 mag .
the average density of the data points is about 5 objects per square arcminute .
the data points for stripe 82 are drawn from our stacked images / catalogs @xcite , and the number density is much higher than that for the other two regions . for a given dataset ,
the starting resolution level is critical for area calculation .
we have tried different starting resolution levels , and found that healpix level 10 ( i.e. , the healpix base resolution is divided 10 times ) is the best for the sdss .
figure [ fig : coverage ] shows an example that compares the results from three different starting resolution levels , levels 9 ( top panel ) , 10 ( middle panel ) , and 11 ( bottom panel ) .
the added ` holes ' from level 9 to level 10 are almost all real , primarily caused by missing data and very bright stars . on the contrary , the majority of the added ` holes ' from level 10 to level 11 are not real , but represent relatively empty regions of the sky in the sdss imaging .
therefore , we use healpix level 10 as our base resolution or starting resolution , which is 11.8 square arcminutes per pixel .
now we classify all pixels into three categories . in the first category
are the pixels that do not cover any data points , and these regions are beyond the effective coverage of our survey .
the pixels in the second category ( hereafter boundary pixels ) are the close neighbors to the pixels in the first category , i.e. , each boundary pixel has at least one neighboring pixel in the first category .
the boundary pixels include outer boundaries and inner boundaries ( the edges of the inner ` holes ' , see figure [ fig : coverage ] ) .
the third category contains all remaining pixels ( hereafter non - boundary pixels ) .
all non - boundary pixels at level 10 constitute the major part of the total effective coverage .
the accuracy of area calculation is now determined by boundary pixels .
we refine boundary pixels by gradually increasing the resolution level , until the resolution roughly matches the average surface density of the data points . for the main survey and overlap regions , the boundary pixels
are calculated to level 13 , at which the resolution is 0.18 square arcminutes per pixel or 5.5 pixels per square arcminute , matching the density of 5 objects per square arcminute . for stripe 82 ,
the boundary pixels are calculated to level 14 , at which the resolution is 0.046 square arcminutes per pixel .
the effective area of the main survey is @xmath107 deg@xmath3 .
the uncertainty quoted here is the contribution of the boundary pixels . as we mentioned earlier , we mainly focused on high galactic latitude @xmath37 deg ( excluding @xmath108 ) , which has an area of 10,371 deg@xmath3 . we also include the lower galactic latitude region between 20 and 30 deg used by @xcite .
these sdss images were taken before 2005 june , and their area is about 869 deg@xmath3 .
the main survey covers 24 luminous quasars at @xmath0 , and the spatial density of these quasars is very low ( about 1 per 468 deg@xmath3 ) . the effective area of the overlap regions is @xmath109 deg@xmath3 .
the uncertainty , or the contribution of the boundary pixels , is relatively large , due to the more complex geometry of the overlap regions . for the overlap regions
, we only considered high galactic latitude @xmath37 deg .
a total of 17 quasars fall in the overlap regions , and 7 of them also belong to the main survey . in other words ,
10 quasars are fainter than a @xmath60 detection in the sdss single - epoch @xmath25-band images .
the effective area of stripe 82 is @xmath110 deg@xmath3 . for stripe 82
, we did not use the region of @xmath111 deg .
this region is at relatively low galactic latitude ( @xmath112 ) .
in addition , the coadded images in this region are significantly shallower than other regions , due to the smaller number of imaging runs covering this region .
we found 13 quasars in stripe 82 .
the spatial density is about 1 per 21 deg@xmath3 , which is much higher than the density in the main survey .
we use simulations to estimate the completeness of the quasar sample .
the incompleteness comes from our quasar selection criteria , i.e. , the color cuts and survey limits that we applied in section 2 .
we describe the completeness as a selection function , the probability that a quasar with a given magnitude ( @xmath94 ) , redshift ( @xmath25 ) , and intrinsic spectral energy distribution ( sed ) meets our selection criteria .
we generate a grid of model quasars using the simulations by @xcite , which is an updated version of the simulations by @xcite .
the model seds are designed to reproduce the colors of @xmath11360,000 quasars at @xmath114 from the sdss boss survey @xcite .
each model sed consists of a broken power - law continuum , a series of emission lines with gaussian profiles , and a scaled fe emission template .
the distributions of spectral features such as the continuum spectral index , line ew , and line fwhm match those from the boss quasars .
the model also incorporates the relations between spectral features and quasar luminosity , such as the baldwin effect and blueshifted lines ( see @xcite for details ) .
the model does not take into account broad - absorption - line ( bal ) quasars and quasars with weak emission lines @xcite .
these quasars have slightly different colors , but the overall impact on our calculation is negligible compared to the statistical uncertainties derived in the next subsections .
we extend the model to higher redshifts under the assumption that the shape of the quasar sed does not evolve with redshift .
the only significant difference is the increasing neutral h absorption in ly@xmath10 forests towards higher redshifts .
finally , photometry is derived from the sed models and photometric errors appropriate for each survey region are added .
we draw a large representative sample of objects from the sdss archive ( or from our stacked images / catalogs for stripe 82 ) . from this sample ,
we obtain the relation between magnitude and error in the @xmath36 and @xmath25 bands in a 2d space , giving us an error distribution at each magnitude .
the errors are added to the model quasars so that the error distributions match those from the real data above .
the @xmath42-band errors are added in the same way so that the error distributions for the model quasars match those from our @xmath42-band data .
as we did in @xcite and @xcite , we compute the average selection probability , @xmath115 , as a function of @xmath94 and @xmath25 after photometric errors are properly incorporated .
figure [ fig : completeness ] shows the selection function for the main survey sample ( top panel ) , the stripe 82 sample ( middle panel ) , and the overlap - region sample ( bottom panel ) .
the filled circles indicate the locations of the quasars in the two samples .
the probability decline at @xmath116 and @xmath117 is caused by the color cuts on the @xmath44 and @xmath43 colors , and the probability decline at the low luminosity end is due to the survey limit in the @xmath25 band . in the top panel , one quasar ( j1243 + 2529 discovered in this paper )
has a probability below 20% .
the reason is that this quasar is relatively faint with @xmath118 mag , below the nominal limit of @xmath119 mag for @xmath120 mag in the sdss .
but its @xmath25-band error ( @xmath120 mag ) still satisfies our selection criteria . in the middle panel ,
the @xmath121 quasar ( j0203 + 0012 ) has the lowest probability .
this quasar was originally found to be at @xmath98 , and it appears to be a @xmath98 quasar in [ fig : allspec ] .
it was later confirmed to be a bal quasar at @xmath121 , and its ly@xmath10 emission has been largely absorbed @xcite . in the bottom panel ,
the contours are different from those for the other two samples .
this is because the 10 quasars in the overlap regions were selected in a small magnitude range , i.e. , @xmath59 to @xmath60 detections in the sdss @xmath25-band images .
the three brightest quasars in this panel are brighter than the faintest quasars in the top panel .
however , they are located in regions that are much shallower than the nominal depth of the sdss single - epoch images , and thus were not selected in our main survey . in figure
[ fig : colorselect ] , we plot the sdss quasars on the @xmath122 versus @xmath123 color - color diagram .
the figure shows that the @xmath122 colors of the sdss quasars are on average bluer than those of the simulated quasars ( black dots ) .
we check @xmath11310 quasars with the bluest @xmath122 colors in the figure , and find that most of them have very strong ly@xmath10 emission that the model does not fully account for .
another likely reason is that @xmath0 quasars tend to have bluer rest - frame uv continuum colors compared to quasars at @xmath114 . near - ir
observations of a large sample of @xmath0 quasars are needed to confirm this hypothesis .
nevertheless , quasars with blue @xmath122 colors are well within our selection criteria , and have little impact on our sample completeness .
we first derive the binned luminosity function from the three subsamples separately : the main survey sample with 24 quasars , the overlap region sample with 10 quasars , and the stripe 82 sample with 13 quasars . the main survey and stripe
82 samples are divided into 4 and 2 discrete luminosity bins , respectively .
redshift evolution is not considered here , but will be taken into account when we parametrize the qlf later .
the volume densities of the quasars are calculated using the traditional @xmath124 method , with the selection function included .
the binned qlf is shown in figure [ fig : lf ] .
the horizontal locations of the filled symbols represent the centers of the luminosity bins , and the horizontal bars indicate the luminosity ranges that the bins cover . in the main survey sample ,
the @xmath73 quasar of @xcite is much brighter than the others , and is thus put in one luminosity bin .
the other three bins have similar numbers of quasars .
we show the median value of the quasar luminosities in each bin as the open symbols in the figure .
the results are consistent with our previous results based on smaller samples , e.g. , figure 6 in @xcite and figure 3 in @xcite .
they also agree with the results from the cfhqs ( gray crosses in the figure ; see * ? ? ?
the binned qlf from our sdss sample in @xcite is well fit by a single power law @xmath125 , or , @xmath126 with a steep slope @xmath8 around 2.9 .
figure [ fig : lf ] shows that the updated binned qlf can also be described as a power law .
the best fit ( the dotted line in figure [ fig : lf ] ) to all sdss data points results in a slope of @xmath127 .
the fit is dependent on the luminosity bin sizes that we chose . in order to remove this dependence ,
we use a maximum likelihood analysis to find the best fit .
the likelihood function ( e.g. * ? ? ?
* ) is written as @xmath128 \\
+ 2\int_{\delta m}\int_{\delta z } \phi(m , z)p(m , z ) \frac{dv}{dz } dzdm , \ ] ] where the sum is over all quasars in the sample , and the integral is over the full luminosity and redshift space of the sample .
the best fit is @xmath129 , consistent with the above result . at low redshift , qlfs
are commonly characterized using a double power law , @xmath130 where @xmath131 is the characteristic magnitude , and @xmath10 is the slope at the faint end .
our sdss data are apparently not deep enough to reach @xmath132 .
however , the recent discovery of much fainter high - redshift quasars ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) have suggested a flatter faint - end slope , and the faintest data point in our sample may be affected by the turnover . in order to better constrain the bright - end slope @xmath8 , we also fit a single power law to all sdss density points but the faintest one in figure [ fig : lf ] , i.e. , the luminosity range @xmath133 mag .
the best fit is @xmath134 .
this is slightly steeper than the slopes found above from the fit to all data points , but is consistent with the value @xmath135 reported by @xcite .
we thus conclude that the bright - end slope of the @xmath0 qlf is around @xmath136 , and we will adopt @xmath9 in the rest of the paper .
we now parametrize the double power law qlf ( equation [ eq : dpl ] ) at @xmath0 . in order to constrain the slope @xmath10
, faint - end data points are required .
however , only a small number of @xmath0 quasars discovered so far are fainter than @xmath137 mag ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in our analysis ,
we include two faint quasars from @xcite and @xcite .
they are represented as the two faintest data points in figure [ fig : lf2 ] .
their effective area coverage and sample completeness were carefully derived in the above papers , and have been incorporated into our calculation .
we do not include an object in @xcite with a narrow ly@xmath10 line .
it is likely a ly@xmath10-emitter , not a type 1 quasar or agn .
the combined sample is still not sufficient to simultaneously constrain all parameters in equation [ eq : dpl ] , thus we choose to fix some of the parameters .
we assume that @xmath131 is constant over our redshift range , i.e. , @xmath138 .
the steep decline of the quasar density at high redshift can be described as ( e.g. * ? ? ?
* ; * ? ? ?
* ) , @xmath139 here we assume @xmath140 , derived from the density evolution of luminous quasars from @xmath15 to 6 ( see details in section 5.1 ) .
furthermore , we fix the value of @xmath8 to be @xmath141 , as measured in section 4.4 .
we then estimate the faint - end slope @xmath10 and the characteristic magnitude @xmath132 by applying a maximum likelihood analysis to equation [ eq : dpl ] . the results are @xmath12 and @xmath13 mag .
the resultant @xmath142 from the best fit is 9.93 gpc@xmath143 mag@xmath30 .
thus the best - fit qlf at @xmath0 can be written as , @xmath144 in units of gpc@xmath143 mag@xmath30 .
note that this is the observed qlf and does not take into account quasar intrinsic properties such as anisotropic emission and dust extinction ( e.g. * ? ? ?
we perform two - dimensional kolmogorov - smirnov ( k - s ) tests to assess the derived qlf .
we generate a large sample ( @xmath145 ) of quasars drawn from the derived qlf ( equation [ eq : dplfit ] ) , covering the full @xmath94-redshift space shown in figure [ fig : completeness ] .
the sample is then convolved with each of the three selection functions to produce three samples of simulated objects .
the resultant samples are compared with the three observed quasar samples using the k - s test ( e.g. * ? ? ?
* ; * ? ? ?
the probability found in each of the three cases is greater than 0.2 , which means that the hypothesis that two data sets are not significantly different is certainly correct .
this indicates that the maximum likelihood analysis that we did above is reasonable .
the two free parameters @xmath10 and @xmath132 are poorly constrained as figure [ fig : alpha_mstar ] shows ; the uncertainties are due to the small number ( 2 ) of quasars at the faint end and the degeneracy between @xmath10 and @xmath132 .
the real uncertainties are likely to be larger ; by fixing @xmath146 and @xmath8 we have not accounted for the uncertainties in those parameters . at low redshift ( @xmath147 ) , the qlf has a very steep bright - end slope @xmath148 and a much flatter faint - end slope @xmath149 ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the bright - end slope at @xmath150 is found to be quite steep ( @xmath148 ) as well ( e.g. * ? ? ?
* ; * ? ? ?
the steep bright - end slope @xmath136 at @xmath0 does not evolve much from those at relatively lower redshifts . on the other hand ,
the faint - end slope of the optical qlf at @xmath151 has not been well constrained ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
previous studies of the qlf at @xmath0 usually assumed a fixed slope @xmath152 as found for low - redshift quasars .
for example , @xcite fixed @xmath152 and found @xmath135 .
their qlf is consistent with ours at the bright end ( figure [ fig : lf2 ] ) .
our results suggest that the faint - end slope at @xmath0 may be marginally steeper than 1.5 based on the faintest data points shown in figure [ fig : lf2 ] , as already pointed out by @xcite and @xcite .
however , the large uncertainties on @xmath10 and @xmath132 will not be reduced before a sizable faint quasar sample is obtained .
in section 4.5 , we described the quasar density evolution using equation [ eq : denev ] , where @xmath140 . here
we explore in detail the density evolution of luminous quasars at @xmath150 .
we work with the integral of the luminosity function from some fiducial lower luminosity @xmath153 to infinity , @xmath154 in figure [ fig : qsoev ] , the blue circles show the results at @xmath0 for @xmath155 mag , together with the results at @xmath156 and 5 from @xcite .
the slopes of the lines give @xmath157 from @xmath158 to 5 , and @xmath16 from @xmath159 to 6 .
@xcite found @xmath160 for quasars brighter than @xmath161 mag at @xmath162 .
this value ( @xmath160 ) has been frequently used in more recent papers ( e.g. * ? ? ?
* ; * ? ? ?
@xcite noticed that the evolution from @xmath15 to @xmath163 is stronger for quasars brighter than @xmath164 mag , with @xmath140 .
we update the density at @xmath165 and find @xmath16 for the same redshift range , which is the same as the value given by @xcite . alternatively , we add the density evolution @xmath166 to equation [ eq : spl ] , and use the maximum likelihood analysis to find the best fit to a single power law .
the result is @xmath167 , consistent with the above result .
the large uncertainty is due to the limited sample size , short redshift baseline , and degeneracy between @xmath146 and @xmath8 .
therefore , we chose to use @xmath16 shown in figure [ fig : qsoev ] .
the underlying assumption is that the density evolution at @xmath0 is similar to that at @xmath168 .
figure [ fig : qsoev ] shows that the spatial density of luminous quasars from @xmath15 to 6 drops faster than that from @xmath156 to 5 ( see also * ? ? ?
this trend seems to continue towards higher redshift .
so far the qlf at @xmath169 has not been well explored .
we estimate the spatial density of luminous quasars at @xmath170 as follows . the ukidss team reported a @xmath21 quasar with @xmath171 mag @xcite , and a @xmath172 quasar with @xmath173 mag @xcite .
the average effective area is roughly 3370 deg@xmath3 for the two quasars ( private communication with d.j .
mortlock and s.j .
warren ) . in order to include the @xmath21 quasar
, we integrate equation [ eq : cqlf ] down to @xmath174 mag ( instead of @xmath155 mag ) , which is roughly the limit of their quasar selection .
since the selection function for the two quasars has not been calculated , we assume that their selection probability is 1 , which would slightly underestimate the density .
given the large statistical uncertainty , the assumption is reasonable for the luminosity regime considered here ( brighter than @xmath174 mag ) .
the results are shown in figure [ fig : qsoev ] .
the slope @xmath146 derived for the density evolution between @xmath0 and 7 is @xmath175 .
in addition , we estimate the density evolution of less luminous quasars from @xmath165 to @xmath176 , using three quasars with @xmath177 from the vista viking survey @xcite .
the cumulative density is integrated down to @xmath178 mag , which is roughly their quasar survey limit .
as shown in figure [ fig : qsoev ] , the slope @xmath146 is @xmath179 .
the above estimate on the density evolution at @xmath180 is rough , based on two very small samples . nevertheless , the trends seen in figure [ fig : qsoev ] suggest a rapid density decline of luminous quasars from @xmath15 towards higher redshifts . we estimate the quasar contribution to the ionizing background at @xmath0 .
we first calculate the number of ionizing photons provided by quasars based on the qlf derived in section 4 .
we assume a broken power - law quasar sed as follows @xcite , @xmath181 the spectral index may vary with quasar luminosity and background ionization rate ( e.g. * ? ? ?
* ) , but we do not consider these complications here .
we integrate the sed over an energy range of 14 ryd ( photons above 4 ryd are consumed by heii ) , and integrate the qlf over a luminosity range from @xmath182 to 18 mag . in figure
[ fig : reion ] , the blue contours show the computed number of ionizing photons in units of mpc@xmath143 s@xmath30 as a function of @xmath10 and @xmath132 . the number of ionizing photons is related to the minimum luminosity that we use in the integral .
we use 18 mag in the above calculation , a value which is often used in the literature ( e.g. * ? ? ?
* ; * ? ? ?
@xcite found a sample of very faint agn candidates with @xmath183 mag at @xmath184 , and estimated that the faint - end slope of the @xmath185 agn luminosity function was roughly 1.81 , consistent with our result of 1.9 at @xmath0 .
their faint agn sample also suggests that we can integrate the qlf down to at least @xmath186 mag .
if we integrate the qlf from @xmath182 to 16 mag , the number of ionizing photons increases by 17% for the best fits shown in equation [ eq : dplfit ] .
agn luminosities can be even lower ( e.g. * ? ? ?
* ; * ? ? ?
in the above calculation , we have assumed that the escape fraction of ionizing photons is 1 . in low - luminosity agn , however , the escape fraction can be much lower ( e.g. * ? ? ? * ) .
we did not take this into account for the luminosity range that we adopted ( @xmath187 mag ) .
the total photon emissivity per unit comoving volume required to ionize the universe is taken from @xcite , i.e. , @xmath188 where @xmath189 is the clumping factor of the igm , and we have assumed that the baryon density @xmath190 .
the clumping factor @xmath189 is critical as it is closely related to the h recombination rate .
simulations have suggested @xmath191 ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in figure
[ fig : reion ] , we show the required photon emissivity ( red dashed curves ) for three representative @xmath189 values , 1.5 , 3 , and 5 . the figure clearly shows the following . 1 .
the significance of the quasar contribution to the ionizing background strongly depends on @xmath10 , @xmath192 , and @xmath189 .
if @xmath17 , the quasar / agn population can not provide enough photons to ionize the @xmath0 igm ( at @xmath18% confidence ) .
we can also rule out at @xmath193% confidence that the quasar / agn population can provide 50% of the required photons .
if @xmath17 , the quasar / agn population can provide enough photons only if the faint - end slope is significantly steeper than 2 and/or the characteristic luminosity is very low .
the conclusion that a faint - end slope at the steep end of current observations is required for quasars to generate sufficient photons to ionize the universe at @xmath0 is generally robust ; however , it is worth restating some of the key assumptions that went into the above calculations .
we assumed a constant @xmath194 , and fixed @xmath146 and @xmath8 when we calculated the qlf .
these choices will underestimate the uncertainties in @xmath10 and @xmath132 .
we chose a magnitude limit of @xmath195 mag , which extends well below current observations .
we further assumed an escape fraction of unity for ionizing photons , independent of quasar luminosity .
finally , we assumed a relatively small range of clumping factors based on current theoretical models .
it is generally thought that low - luminosity star - forming galaxies may provide enough ionizing photons for cosmic reionization , while the contribution from quasars / agn is probably negligible due to their low spatial density .
figure [ fig : reion ] shows that we can not yet fully rule out the probability that uv light from agn is responsible for ionizing the universe ( see also e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) ; however , current observations do not favor this scenario .
future deep surveys of quasars at @xmath19 will provide a definitive answer .
in this paper , we have presented the discovery of nine quasars at @xmath0 identified in the sdss , including seven quasars in the sdss main survey , one quasar in the overlap regions , and one quasar in stripe 82 .
one of the quasars in the main survey , j1148 + 0702 at @xmath196 , is the second highest - redshift quasar found in the sdss .
this completes our survey of @xmath0 quasars in the sdss footprint .
we summarized our final sample of 52 sdss quasars at @xmath0 . in total , we have found ( 1 ) 29 quasars in 11,240 deg@xmath3 of the sdss main survey ; ( 2 ) 17 quasars in 4223 deg@xmath3 of the overlap regions ( 7 of which are in common with the main survey sample ) ; ( 3 ) 13 quasars in 277 deg@xmath3 of stripe 82 .
the main survey quasars are the most luminous quasars , with @xmath2 mag .
the overlap region quasars are roughly 0.5 mag fainter , and the stripe 82 quasars are 2 mag fainter .
the quasars span a wide luminosity range of @xmath6 mag , and comprise a well - defined quasar sample at @xmath0 . based on the combination of our new quasar sample and two much fainter quasars in the literature
, we obtained parameters for a double power - law fit to the qlf at @xmath0 using a maximum likelihood analysis .
the bright end of the qlf is well constrained , and the slope is steep with @xmath197 .
the best - fitting results for the faint - end slope @xmath10 and the characteristic magnitude @xmath11 are @xmath12 and @xmath13 mag , and the two quantities are strongly correlated .
the large uncertainties are due to the small number of quasars at the faint end .
we calculated the cumulative density of luminous quasars ( @xmath198 mag , @xmath199 mag , and @xmath200 mag ) at @xmath0 and 7 , and compared them with those at @xmath156 and 5 .
we found that the cumulative density at @xmath201 declines rapidly towards higher redshift .
we estimated the quasar contribution to the ionizing background at @xmath0 using the derived qlf . assuming an igm clumping factor @xmath17
, the quasar population can not provide enough photons to ionize the @xmath0 igm ( at @xmath18% confidence ) .
we found that quasars can provide enough photons only if the faint - end slope is steeper than 2 and/or if the characteristic luminosity is very low . a large sample of very faint quasars ( @xmath202 mag ) is needed to provide a better constraint on the quasar contribution to cosmic reionization .
many quasars in our sample have been extensively studied in multiple wavelength bands from x - ray to radio .
more observations are being carried out and being planned .
for example , we are carrying out deep near - ir spectroscopy of @xmath66 quasars at @xmath0 using gemini gnirs .
when this program completes , we will have near - ir ( or rest - frame uv ) spectra for all the sdss quasars .
these spectra will allow us to measure various quasar properties , such as spectral index and emission line properties , metallicity in the broad - line region , central black - hole mass , and so on . the well - defined sample will enable us to derive the black - hole mass function at @xmath0 , and further constrain the birth and growth of the earliest massive black holes .
we are also gathering mm / submm observations using alma , iram , and jcmt for our sdss sample ( e.g. * ? ? ? * ; * ? ? ? * ) .
these observations provide rich information on dust emission , star formation , and dynamical properties of quasar host galaxies , in the context of galaxy black hole co - evolution at the early epoch .
therefore , this unique sdss sample will have a legacy value for exploring the distant universe in the future .
meanwhile , ongoing and future large - area surveys are finding high - redshift quasars in large numbers .
for example , pan - starrs1 has found more than 100 quasars @xcite , and will improve the constraint on the bright - end qlf at @xmath0 .
note that the measurement of the quasar density at the brightest end ( @xmath26 mag ) will not be improved by more than a factor of two , since the sdss already covers one fourth of the whole sky . with near - ir imaging ,
the vista viking survey is able to find higher redshift quasars .
it has found 3 quasars at @xmath203 , and is expecting to find nearly 20 quasars at @xmath204 in the near future @xcite .
des , with imaging data slightly deeper than stripe 82 over @xmath205 deg@xmath3 , has found its first @xmath0 quasar , and has claimed that it would find 50100 quasars at @xmath180 , including 310 quasars at @xmath206 @xcite .
the subaru shellq survey is producing a large number of very faint quasars using deep hyper suprime - cam imaging data @xcite .
these quasars will be used to constrain the faint end of the qlf at @xmath0 .
the future lsst survey @xcite will have an unprecedented power for searches of high - redshift quasars .
it will eventually find thousands of quasars ( assuming that there will be enough resources for follow - up identification ) , and fully constrain the @xmath0 qlf .
the above surveys are providing a golden opportunity for studying high - redshift quasars and the distant universe .
we acknowledge support from the ministry of science and technology of china under grant 2016yfa0400703 , and from the national science foundation of china under grant 11533001 .
lj acknowledges support from the thousand youth talents program of china .
idm and xf acknowledge support from nsf grants ast 11 - 07682 and 15 - 15115 .
ys acknowledges support from an alfred p. sloan research fellowship .
we thank n. kashikawa , c. willott , d.j .
mortlock , b.p .
venemans , and s.j .
warren for providing information for figures [ fig : lf2 ] and [ fig : qsoev ] .
this research uses data obtained through the telescope access program ( tap ) , which has been funded by the national astronomical observatories of china ( naoc ) , the chinese academy of sciences ( the strategic priority research program
the emergence of cosmological structures " grant no .
xdb09000000 ) , and the special fund for astronomy from the ministry of finance .
observations obtained with the hale telescope at palomar observatory were obtained as part of an agreement between naoc and the california institute of technology .
observations reported here were obtained in part at the mmt observatory , a joint facility of the university of arizona and the smithsonian institution .
this work is based in part on data obtained as part of the ukirt infrared deep sky survey .
funding for the sdss and sdss - ii has been provided by the alfred p. sloan foundation , the participating institutions , the national science foundation , the u.s .
department of energy , the national aeronautics and space administration , the japanese monbukagakusho , the max planck society , and the higher education funding council for england .
the sdss web site is http://www.sdss.org/. the sdss is managed by the astrophysical research consortium for the participating institutions .
the participating institutions are the american museum of natural history , astrophysical institute potsdam , university of basel , university of cambridge , case western reserve university , university of chicago , drexel university , fermilab , the institute for advanced study , the japan participation group , johns hopkins university , the joint institute for nuclear astrophysics , the kavli institute for particle astrophysics and cosmology , the korean scientist group , the chinese academy of sciences ( lamost ) , los alamos national laboratory , the max - planck - institute for astronomy ( mpia ) , the max - planck - institute for astrophysics ( mpa ) , new mexico state university , ohio state university , university of pittsburgh , university of portsmouth , princeton university , the united states naval observatory , and the university of washington .
figure [ fig : allspec ] shows the optical spectra of the 52 sdss quasars at @xmath0 , ordered by redshift .
the numbers of the quasars correspond to the numbers in column 1 of table 2 .
most of the spectra were taken from the quasar discovery papers listed in column 13 of table 2 .
all spectra have been binned to 10 per pixel .
the figure can be downloaded from http://kiaa.pku.edu.cn/~jiang/sdss52spectra.eps , or http://kiaa.pku.edu.cn/~jiang/sdss52spectra2.eps .
adelman - mccarthy , j. k. , ageros , m. a. , allam , s. s. , et al .
2007 , , 172 , 634 aihara , h. , allende prieto , c. , an , d. , et al .
2011 , , 193 , 29 annis , j. , soares - santos , m. , strauss , m. a. , et al .
2014 , , 794 , 120 baados , e. , venemans , b. p. , morganson , e. , et al . 2014 , , 148 , 14 baados , e. , decarli , r. , walter , f. , et al .
2015 , , 805 , l8 baados , e. , et al .
2016 , , in press ( arxiv:1608.03279 ) barnett , r. , warren , s. j. , banerji , m. , et al .
2015 , , 575 , a31 becker , r. h. , fan , x. , white , r. l. , et al .
2001 , , 122 , 2850 becker , g. d. , sargent , w. l. w. , rauch , m. , & calverley , a. p. 2011
, , 735 , 93 becker , g. d. , bolton , j. s. , madau , p. , et al . 2015 , , 447 , 3402 blain , a. w. , assef , r. , stern , d. , et al .
2013 , , 778 , 113 bolton , j. s. , haehnelt , m. g. , warren , s. j. , et al .
2011 , , 416 , l70 brown , w. r. , mcleod , b. a. , geary , j. c. , & bowsher , e. c. 2008 , , 7014 , 70142p calura , f. , gilli , r. , vignali , c. , et al .
2014 , , 438 , 2765 carilli , c. l. , wang , r. , fan , x. , et al .
2010 , , 714 , 834 carilli , c. l. , & walter , f. 2013 , , 51 , 105 carnall , a. c. , shanks , t. , chehade , b. , et al . 2015 , , 451 , l16 croom , s. m. , richards , g. t. , shanks , t. , et al .
2009 , , 399 , 1755 cool , r. j. , kochanek , c. s. , eisenstein , d. j. , et al .
2006 , , 132 , 823 de rosa , g. , decarli , r. , walter , f. , et al .
2011 , , 739 , 56 de rosa , g. , venemans , b. p. , decarli , r. , et al .
2014 , , 790 , 145 dipompeo , m. a. , myers , a. d. , brotherton , m. s. , runnoe , j. c. , & green , r. f. 2014 , , 787 , 73 fan , x. 1999 , , 117 , 2528 fan , x. , white , r. l. , davis , m. , et al .
2000 , , 120 , 1167 fan , x. , narayanan , v. k. , lupton , r. h. , et al .
2001 , , 122 , 2833 fan , x. , strauss , m. a. , schneider , d. p. , et al .
2001 , , 121 , 54 fan , x. , strauss , m. a. , schneider , d. p. , et al .
2003 , , 125 , 1649 fan , x. , hennawi , j. f. , richards , g. t. , et al .
2004 , , 128 , 515 fan , x. , strauss , m. a. , richards , g. t. , et al .
2006 , , 131 , 1203 fan , x. , strauss , m. a. , becker , r. h. , et al .
2006 , , 132 , 117 fasano , g. , & franceschini , a. 1987 , , 225 , 155 finlator , k. , oh , s. p. , zel , f. , & dav , r. 2012 , , 427 , 2464 fiore , f. , puccetti , s. , grazian , a. , et al .
2012 , , 537 , a16 fliri , j. , & trujillo , i. 2016 , , 456 , 1359 gallerani , s. , maiolino , r. , juarez , y. , et al .
2010 , , 523 , a85 grski , k. m. , hivon , e. , banday , a. j. , et al .
2005 , , 622 , 759 goto , t. 2006 , , 371 , 769 fukugita , m. , ichikawa , t. , gunn , j. e. , et al .
1996 , , 111 , 1748 giallongo , e. , grazian , a. , fiore , f. , et al .
2015 , , 578 , a83 glikman , e. , djorgovski , s. g. , stern , d. , et al .
2011 , , 728 , l26 gunn , j. e. , carr , m. , rockosi , c. , et al .
1998 , , 116 , 3040 gunn , j. e. , siegmund , w. a. , mannery , e. j. , et al . 2006 , , 131 , 2332 hao , l. , strauss , m. a. , fan , x. , et al .
2005 , , 129 , 1795 ho , l. c. , filippenko , a. v. , & sargent , w. l. w. 1997 , , 112 , 315 hogg , d. w. , finkbeiner , d. p. , schlegel , d. j. , & gunn , j. e. 2001 , , 122 , 2129 ikeda , h. , nagao , t. , matsuoka , k. , et al .
2011 , , 728 , l25 ikeda , h. , nagao , t. , matsuoka , k. , et al .
2012 , , 756 , 160 ivezi , . ,
lupton , r. h. , schlegel , d. , et al .
2004 , astronomische nachrichten , 325 , 583 ivezic , z. , axelrod , t. , brandt , w. n. , et al . 2008 , serbian astronomical journal , 176 , 1 jiang , l. , fan , x. , hines , d. c. , et al .
2006 , , 132 , 2127 jiang , l. , fan , x. , vestergaard , m. , et al .
2007 , , 134 , 1150 jiang , l. , fan , x. , annis , j. , et al .
2008 , , 135 , 1057 jiang , l. , fan , x. , bian , f. , et al .
2009 , , 138 , 305 jiang , l. , fan , x. , brandt , w. n. , et al .
2010 , , 464 , 380 jiang , l. , fan , x. , bian , f. , et al .
2014 , , 213 , 12 jiang , l. , mcgreer , i. d. , fan , x. , et al .
2015 , , 149 , 188 juarez , y. , maiolino , r. , mujica , r. , et al .
2009 , , 494 , l25 jun , h. d. , i m , m. , lee , h. m. , et al . 2015 , , 806 , 109 lusso , e. , worseck , g. , hennawi , j. f. , et al .
2015 , , 449 , 4204 kaiser , n. , burgett , w. , chambers , k. , et al .
2010 , , 7733 , 77330e kashikawa , n. , ishizaki , y. , willott , c. j. , et al .
2015 , , 798 , 28 kim , y. , i m , m. , jeon , y. , et al .
2015 , , 813 , l35 kurk , j. d. , walter , f. , fan , x. , et al .
2007 , , 669 , 32 kurk , j. d. , walter , f. , fan , x. , et al .
2009 , , 702 , 833 leipski , c. , meisenheimer , k. , walter , f. , et al .
2014 , , 785 , 154 lupton , r. , gunn , j. e. , ivezi , z. , knapp , g. r. , & kent , s. 2001 , astronomical data analysis software and systems x , 238 , 269 lyu , j. , rieke , g. h. , & alberts , s. 2016 , , 816 , 85 marshall , h. l. , avni , y. , tananbaum , h. , & zamorani , g. 1983 , , 269 , 35 madau , p. , haardt , f. , & rees , m. j. 1999 , , 514 , 648 madau , p. , & haardt , f. 2015 , , 813 , l8 matsuoka , y. , onoue , m. , kashikawa , n. , et al .
2016 , , 828 , 26 mcgreer , i. d. , becker , r. h. , helfand , d. j. , & white , r. l. 2006 , , 652 , 157 mcgreer , i. d. , jiang , l. , fan , x. , et al .
2013 , , 768 , 105 mcgreer , i. d. , mesinger , a. , & dodorico , v. 2015 , , 447 , 499 mcquinn , m. , oh , s. p. , & faucher - gigure , c .- a .
2011 , , 743 , 82 micheva , g. , iwata , i. , & inoue , a. k. 2016 , , in press ( arxiv:1604.00102 ) mitra , s. , choudhury , t. r. , & ferrara , a. 2016 , , summited ( arxiv:1606.02719 ) morganson , e. , de rosa , g. , decarli , r. , et al .
2012 , , 143 , 142 mortlock , d. j. , patel , m. , warren , s. j. , et al .
2009 , , 505 , 97 mortlock , d. j. , warren , s. j. , venemans , b. p. , et al .
2011 , , 474 , 616 mortlock , d. j. 2015 , arxiv:1511.01107 omont , a. , willott , c. j. , beelen , a. , et al .
2013 , , 552 , a43 padmanabhan , n. , schlegel , d. j. , finkbeiner , d. p. , et al .
2008 , , 674 , 1217 peacock , j. a. 1983 , , 202 , 615 pier , j. r. , munn , j. a. , hindsley , r. b. , et al .
2003 , , 125 , 1559 plotkin , r. m. , shemmer , o. , trakhtenbrot , b. , et al .
2015 , , 805 , 123 reed , s. l. , mcmahon , r. g. , banerji , m. , et al .
2015 , , 454 , 3952 richards , g. t. , strauss , m. a. , fan , x. , et al .
2006 , , 131 , 2766 ross , n. p. , myers , a. d. , sheldon , e. s. , et al .
2012 , , 199 , 3 ross , n. p. , mcgreer , i. d. , white , m. , et al .
2013 , , 773 , 14 schmidt , g. d. , weymann , r. j. , & foltz , c. b. 1989 , , 101 , 713 shen , y. , strauss , m. a. , oguri , m. , et al .
2007 , , 133 , 2222 shen , y. , & liu , x. 2012 , , 753 , 125 shull , j. m. , harness , a. , trenti , m. , & smith , b. d. 2012 , , 747 , 100 simcoe , r. a. , cooksey , k. l. , matejek , m. , et al .
2011 , , 743 , 21 skrutskie , m. f. , cutri , r. m. , stiening , r. , et al .
2006 , , 131 , 1163 smith , j. a. , tucker , d. l. , kent , s. , et al .
2002 , , 123 , 2121 strauss , m. a. , fan , x. , gunn , j. e. , et al .
1999 , , 522 , l61 tucker , d. l. , kent , s. , richmond , m. w. , et al .
2006 , astronomische nachrichten , 327 , 821 venemans , b. p. , mcmahon , r. g. , warren , s. j. , et al .
2007 , , 376 , l76 venemans , b. p. , findlay , j. r. , sutherland , w. j. , et al .
2013 , , 779 , 24 venemans , b. p. , baados , e. , decarli , r. , et al .
2015 , , 801 , l11 walter , f. , riechers , d. , cox , p. , et al .
2009 , , 457 , 699 et al .
2013 , , 773 , 44 wang , f. , wu , x .- b . , fan , x. , et al .
2016 , , 819 , 24 wang , r. , carilli , c. l. , wagg , j. , et al .
2008 , , 687 , 848 - 858 wang , r. , wagg , j. , carilli , c. l. , et al .
2011 , , 142 , 101 wang , r. , wagg , j. , carilli , c. l. , et al .
2013 , , 773 , 44 wang , r. , wu , x .- b . ,
neri , r. , et al .
2016 , , in press ( arxiv:1606.09634 ) warren , s. j. , hambly , n. c. , dye , s. , et al .
2007 , , 375 , 213 white , r. l. , becker , r. h. , fan , x. , & strauss , m. a. 2003 , , 126 , 1 willott , c. j. , delorme , p. , omont , a. , et al .
2007 , , 134 , 2435 willott , c. j. , delorme , p. , reyl , c. , et al .
2009 , , 137 , 3541 willott , c. j. , albert , l. , arzoumanian , d. , et al . 2010a , , 140 , 546 willott , c. j. , delorme , p. , reyl , c. , et al .
2010b , , 139 , 906 wright , e. l. , eisenhardt , p. r. m. , mainzer , a. k. , et al .
2010 , , 140 , 1868 - 1881 wu , x .- b . ,
wang , f. , fan , x. , et al .
2015 , , 518 , 512 wyithe , j. s. b. , & bolton , j. s. 2011 , , 412 , 1926 yan , l. , donoso , e. , tsai , c .- w . , et al .
2013 , , 145 , 55 yang , j. , wang , f. , wu , x. , et al .
2016 , , in press ( arxiv:1607.04415 ) york , d. g. , adelman , j. , anderson , j. e. , jr . , et al .
2000 , , 120 , 1579 zeimann , g. r. , white , r. l. , becker , r. h. , et al . 2011 , , 736 , 57 | we present the discovery of nine quasars at @xmath0 identified in the sloan digital sky survey ( sdss ) imaging data .
this completes our survey of @xmath0 quasars in the sdss footprint .
our final sample consists of 52 quasars at @xmath1 , including 29 quasars with @xmath2 mag selected from 11,240 deg@xmath3 of the sdss single - epoch imaging survey ( the main survey ) , 10 quasars with @xmath4 selected from 4223 deg@xmath3 of the sdss overlap regions ( regions with two or more imaging scans ) , and 13 quasars down to @xmath5 mag from the 277 deg@xmath3 in stripe 82 .
they span a wide luminosity range of @xmath6 .
this well - defined sample is used to derive the quasar luminosity function ( qlf ) at @xmath0 . after combining our sdss sample with two faint ( @xmath7 mag ) quasars from the literature
, we obtain the parameters for a double power - law fit to the qlf .
the bright - end slope @xmath8 of the qlf is well constrained to be @xmath9 . due to the small number of low - luminosity quasars , the faint - end slope @xmath10 and the characteristic magnitude @xmath11
are less well constrained , with @xmath12 and @xmath13 mag .
the spatial density of luminous quasars , parametrized as @xmath14 , drops rapidly from @xmath15 to 6 , with @xmath16 .
based on our fitted qlf and assuming an igm clumping factor of @xmath17 , we find that the observed quasar population can not provide enough photons to ionize the @xmath0 igm at @xmath18% confidence .
quasars may still provide a significant fraction of the required photons , although much larger samples of faint quasars are needed for more stringent constraints on the quasar contribution to reionization . |
the graphs in this paper can be directed or undirected . consider a connected graph @xmath0 with a set @xmath11 of terminal vertices , or simply _ terminals _ , and a length ( or cost ) function @xmath12 .
let @xmath13 be a root vertex ( i.e. there is a path from @xmath5 to any vertex in @xmath14 ) if @xmath4 is directed or a special vertex called _ root _ if @xmath4 is undirected . the ( rooted ) steiner tree problem ( ` steiner - tree ` ) is to determine a directed tree @xmath15 in @xmath4 , rooted at @xmath5 , spanning all terminals of @xmath9 and having a minimum total length .
the undirected steiner tree problem , where one searches for a minimum - length tree spanning the terminals in an undirected graph , has been widely studied and the associated decision problem was one of karp s 21 np - complete problems @xcite .
it also has many applications , as shown in @xcite .
this problem is apx - hard @xcite , but it can be solved in polynomial time when the number of terminals is fixed @xcite , and admits constant ratio approximation algorithms otherwise @xcite .
there are less results about the directed version , which is a generalization of the undirected one and of the set cover problem , and only non constant ratio approximation algorithms are known @xcite .
directed problems occur for instance in vlsi design @xcite or in multicast routing @xcite .
we consider in this paper a generalization of the ( rooted ) steiner tree problem .
assume we are given a _
function @xmath16 , where @xmath8 is an upper bound on the number of paths containing @xmath10 and linking @xmath5 to terminals .
equivalently , for every @xmath17 in a tree @xmath15 rooted at @xmath5 , the subtree of @xmath15 rooted at @xmath18 can not contain more than @xmath8 terminals . without loss of generality ,
we assume that @xmath19 for each edge @xmath10 .
the minimum - length capacitated ( rooted ) steiner tree problem is defined as follows : + * minimum - length ( rooted ) capacitated steiner tree problem * ( ` ml - cap - steiner - tree ` ) _ input_. a connected graph @xmath20 ; a set @xmath21 of @xmath22 terminals ; a root vertex @xmath23 ; two functions on @xmath24 : a nonnegative length function @xmath25 and a positive capacity function @xmath26 . _
objective_. determine , if it exists , a minimum - length directed tree @xmath15 rooted at @xmath5 , that spans all the vertices of @xmath9 and does not violate the capacity constraints .
+ if @xmath0 is undirected and @xmath17 is an arc of @xmath15 , then @xmath27 $ ] must be an edge of @xmath24 .
note that ` steiner - tree ` is the special case of ` ml - cap - steiner - tree ` where @xmath28 for all @xmath7 ( in this case , a feasible solution always exists ) . `
ml - cap - steiner - tree ` appears naturally in several contexts , for example when designing a wind farm collection network @xcite , in the design of telecommunication networks @xcite or in power distribution system optimization @xcite .
when @xmath29 for all @xmath7 , ` ml - cap - steiner - tree ` turns into a decision problem , denoted by ` cap - steiner - tree ` , and consisting of determining whether there exists or not a tree rooted at @xmath5 , spanning all the terminals , and not violating the capacity constraints . when @xmath30 , i.e. a feasible solution is a spanning tree , ` ml - cap - steiner - tree ` is solvable in polynomial time if @xmath31 for all @xmath7 , while it is np - hard if @xmath32 for all @xmath7 @xcite
several authors propose models and methods based on mathematical programming to solve this capacitated spanning tree problem for real - life applications such as telecommunication network design problems @xcite .
their methods allow to solve the case where there is a positive integer demand at each vertex ( instead of a unit demand as in ` ml - cap - steiner - tree ` ) . in @xcite ,
the authors provide approximation algorithms for a variant of ` ml - cap - steiner - tree ` where the capacities are uniform and the problem always admits a feasible solution , since it is assumed that a metric completion of the graph is available .
this paper addresses the problem where the demand is equal to 1 for each terminal vertex and @xmath33 .
as will be made clear in the next sections , there are strong links between`ml - cap - steiner - tree ` and the two following famous problems , namely the minimum - length vertex - disjoint paths problem ( ` ml - vdisj - path ` ) and the minimum - length edge - cost flow problem ( ` edge - cost - flow ` ) . * minimum - length vertex - disjoint paths problem * ( ` ml - vdisj - path ` ) _ input_. a graph @xmath0 ; a nonnegative length function @xmath25 on @xmath24 ; @xmath34 disjoint vertex pairs @xmath35 . _
objective_. find @xmath34 mutually vertex - disjoint paths @xmath36 of minimum total length so that @xmath37 links @xmath38 to @xmath39 @xmath40 .
* minimum edge - cost flow problem * ( ` edge - cost - flow ` ) _ input_. a graph @xmath0 ; a positive integer @xmath41 ; two specified vertices @xmath42 and @xmath43 ; a nonnegative length function @xmath25 on @xmath24 ; a positive capacity function @xmath26 on @xmath24 . _
objective_. find a minimum - length feasible flow of @xmath41 units from @xmath42 to @xmath43 , where the length of a flow is the sum of the lengths of the arcs / edges carrying a positive flow .
when @xmath29 for all @xmath7 , ` ml - vdisj - path ` is known as the vertex - disjoint paths problem and will be denoted by ` vdisj - path ` .
it is np - complete in directed and undirected graphs @xcite and remains np - complete for fixed @xmath34 in directed graphs @xcite , but it can be solved in polynomial time if @xmath34 is fixed and the graph is either undirected @xcite or a directed acyclic graph @xcite .
the np - hardness results for ` vdisj - path ` apply to ` ml - vdisj - path ` as well , but for this latter problem the complexity is unknown in the case where @xmath34 is fixed and the graph is undirected .
however , a polynomial - time probabilistic algorithm for @xmath44 has been recently presented in @xcite . for any graph
theoretical terms not defined here , the reader is referred to @xcite .
we use the term _ path _ both for a chain when the graph is undirected , and for a directed path when the graph is directed , i.e. when it is a digraph .
if the graph is directed , recall that , in the definition of ` ml - cap - steiner - tree ` , @xmath5 is assumed to be a root vertex .
this is a trivial necessary condition for the existence of a feasible solution and can be easily checked . since all trees studied in this paper are directed from @xmath5 towards the terminals , we use the term _ tree _ instead of directed tree .
for a vertex @xmath18 in a tree @xmath15 , we denote by @xmath45 the subtree of @xmath15 rooted at @xmath18 . for a subgraph @xmath46 of @xmath4
, we indifferently denote by @xmath47 or @xmath48 the sum of the lengths of the arcs / edges in @xmath49 .
also , for @xmath7 , a rooted tree @xmath15 in @xmath4 , and two vertices @xmath50 and @xmath18 such that @xmath18 is a descendant of @xmath50 in @xmath15 , we say that @xmath50 is @xmath10__-linked _ _ ( resp .
_ ) to @xmath18 in @xmath15 if @xmath10 belongs ( resp
. does not belong ) to the path @xmath52 from @xmath50 to @xmath18 in @xmath15 .
similarly , when we write that @xmath5 is @xmath51-linked to a subset @xmath53 of terminals in @xmath15 , this means that @xmath10 does not belong to the paths in @xmath15 that link @xmath5 to the terminals of @xmath53 .
the capacity constraints therefore impose that , for all @xmath7 , @xmath5 is @xmath10-linked to at most @xmath8 terminals in any feasible solution @xmath15 to an ` ml - cap - steiner - tree ` instance .
equivalently , @xmath45 contains at most @xmath8 terminals for all @xmath17 in @xmath15 .
the next section gives an overview of our results concerning ` ml - cap - steiner - tree ` and explains how the paper is organized .
in this section we show that our results provide a complete characterization of the complexity of ` ml - cap - steiner - tree ` that allows us to distinguish beween easy and hard cases of the problem for digraphs , directed acyclic graphs ( called dags ) and undirected graphs .
notice that any undirected instance of ` ml - cap - steiner - tree ` can be transformed into a directed one by replacing each edge by two opposite arcs having the same length and capacity .
hence , any positive result ( existence of a polynomial - time algorithm or approximation result ) for directed graphs is also true for undirected graphs , while any negative result for undirected graphs ( np - hardness or non - approximability result ) is also true for directed graphs .
, the following parameters are considered : the number @xmath41 of terminal vertices , the minimum and maximum edge capacities , and the edge lengths .
more precisely , @xmath41 can be fixed or not ; the minimum and maximum edge capacities can be non depending on @xmath41 ( equal to 1 or not ) , they can be greater than or equal to @xmath54 ( @xmath55 ) , and they can be equal ( uniform capacity ) or not ; the edge lengths can be all equal to @xmath56 , all equal to a positive value ( i.e. uniform ) , or non uniform .
we settle all cases except one , namely the undirected case with uniform capacity and fixed @xmath57 , but we prove that ` ml - cap - steiner - tree ` is then equivalent to ` ml - vdisj - path ` in undirected graphs with fixed @xmath34 , whose complexity is a long - standing open problem in this case @xcite .
our results are summarized in four tables .
each line of each table corresponds to a specific case of ` ml - cap - steiner - tree ` and refers to the theorem where the case is settled .
the first table contains results that are valid for the three types of graphs , while the next three tables contain results that are specific to digraphs , undirected graphs , and dags , respectively . in these tables , we denote by @xmath58 the best possible approximation ratio for ` steiner - tree ` , and by @xmath59 the best possible approximation ratio for ` ml - vdisj - path ` with a fixed number of source - sink pairs .
the three trees drawn in figure 1 provide another picture of the possible cases for the three types of graphs ( digraphs , dags and undirected graphs ) .
the numbers assigned to the leaves of these trees refer to the corresponding rows in the tables .
the values of the three parameters appear on the branches and each branching node corresponds to a partition of the possible cases : the value on a branch excludes the values on the branches to the left .
for instance , in undirected graphs , the capacities can be either uniform equal to 1 , or at least @xmath60 , or uniform of value at least 2 and at most @xmath61 , or , finally , any capacities not yet considered .
moreover , if a leaf corresponds to a branch where the values of some parameters are unspecified , then this means that the associated result holds even in the most general case ( if it is a positive , i.e. , tractability result ) or in the most specific case ( if it is a negative , i.e. , hardness result ) with regard to the unspecified values .
for instance , the np - hardness result associated with leaf 7 holds even if @xmath41 is fixed and if all lengths are 0 ( since neither the value of @xmath41 nor the lengths appear on this branch ) , and the result associated with leaf 11 holds for any lengths and any capacities ( since only the assumption on @xmath41 being fixed appears on this branch ) .
therefore , for digraphs , the branch any capacity " includes the case of uniform capacities between 2 and @xmath61 for @xmath41 fixed ( or not ) .
concerning the last line of table 3 , if the capacity is uniform and @xmath41 is fixed , then there exists some constant @xmath62 such that all capacities are equal to @xmath54 : hence , in the tree dealing with undirected graphs in figure [ fig : arbres ] , the branch `` any capacity '' , which leads to case 8 of table [ tabundirectedgraphs ] , excludes the case where @xmath41 is fixed . & & & + & & & + & & & + & & & + & & & + & & and @xmath63 & + [ h ] & & & + & & & + & & & + & & & + & & & + & & & + & & & + we describe in section 3 the structure of optimal solutions to an ` ml - cap - steiner - tree ` instance .
section 4 is devoted to relations between ` ml - cap - steiner - tree ` and ` ml - vdisj - path ` .
we prove in section 5 some np - hardness results for the general case , while special cases where the number @xmath41 of terminals is fixed , or where all capacities are almost equal to @xmath41 , are studied in sections 6 and 7 .
we can assume , without loss of generality , that there is a bijection between the set of 1-degree vertices ( leaves ) in @xmath64 and @xmath9 . indeed ,
if @xmath65 is not a leaf , we can add a new terminal vertex @xmath66 and an edge @xmath67 $ ] ( or an arc @xmath68 ) with capacity @xmath69 and length @xmath56 , and replace @xmath43 by @xmath66 in @xmath9 .
moreover , if there is a leaf @xmath70 in @xmath4 , then @xmath18 can be removed from @xmath4 since the removal of @xmath18 from a solution @xmath15 to an ` ml - cap - steiner - tree ` instance gives a solution @xmath71 which is at least as good as @xmath15 .
a solution @xmath15 ( if any ) to an ` ml - cap - steiner - tree ` instance is a tree rooted at @xmath5 , and defines @xmath41 paths from @xmath5 to the @xmath41 terminals .
the vertices with degree at least 3 in @xmath72 are called _
junction vertices_. to each junction vertex @xmath18 , we associate the set @xmath73 of terminals in the subtree @xmath45 rooted at @xmath18 .
moreover , for an arc @xmath17 in @xmath15 , @xmath74 is the number of terminals to which @xmath5 is @xmath10-linked in @xmath15 .
if there is no directed path linking two vertices @xmath18 and @xmath75 in @xmath15 , then @xmath76 , otherwise @xmath15 would contain a cycle . given a tree
@xmath15 spanning a set @xmath9 of terminals , its _ skeleton _ is the tree obtained from @xmath15 by iteratively contracting vertices @xmath77 with exactly one incoming arc @xmath78 and exactly one outgoing arc @xmath79 ( i.e. , the path @xmath80 is replaced by an arc @xmath81 ) .
this means that there is an arc @xmath78 in the skeleton of @xmath15 if and only if there is a path from @xmath50 to @xmath18 in @xmath15 , each internal vertex of this path being of degree 2 in @xmath15 . when all capacities are 1 , the skeleton of a feasible solution is a star , since the root is the only possible vertex with degree @xmath82 in this skeleton .
we now prove some properties which will be useful later .
[ prop : numbvert ] the skeleton of an inclusion - wise minimal tree @xmath15 rooted at @xmath5 and spanning @xmath41 terminals ( all of degree 1 ) contains at most @xmath83 vertices , where @xmath84 is the degree of root @xmath5 in @xmath15 .
let @xmath85 be the number of junction vertices in the skeleton @xmath86 of @xmath15 .
clearly , @xmath86 contains @xmath87 vertices and @xmath88 edges . since the sum of the degrees of all vertices in @xmath86 is @xmath89 ,
we have @xmath90 , which implies @xmath91 and @xmath92 . [ prop : pathlength ] given an inclusion - wise minimal tree @xmath15 rooted at @xmath5 and spanning @xmath41 terminals ( all of degree 1 ) , the path with minimum number of vertices from root @xmath5 to a terminal in the skeleton of @xmath15 contains at most @xmath93 vertices .
let @xmath86 be the skeleton of @xmath15 , @xmath94 its number of vertices , and @xmath95 the minimum number of vertices on a path from @xmath5 to a terminal in @xmath86 .
* if @xmath5 has degree 1 in @xmath15 , then @xmath86 contains one vertex at levels 1 and 2 , and at least @xmath96 vertices at levels @xmath97 .
hence , @xmath98 , which implies @xmath99 .
* if @xmath5 has degree at least 2 in @xmath15 , then @xmath86 contains at least @xmath100 vertices at levels @xmath101 .
hence , @xmath102 , which implies @xmath103 . in both cases , it follows from property [ prop : numbvert ] that @xmath104 .
notice that , if @xmath15 is a complete binary tree , then @xmath105 : therefore , up to a constant factor , the bound in the previous property can not be improved .
given a graph @xmath4 with a root @xmath5 and @xmath41 terminals , a _ potential skeleton _ in @xmath4 is defined as a tree @xmath106 , rooted at @xmath5 , spanning the @xmath41 terminals , and such that the only vertices without outgoing arc are the @xmath41 terminals , while any other vertex , except possibly @xmath5 , has degree at least 3 in @xmath106 ( and hence in @xmath4 ) . while the skeleton of a solution to an ` ml - cap - steiner - tree ` instance is a potential skeleton , the reverse is not necessarily true , as illustrated in figure [ fig : potential ] .
if we select the left arc incident to @xmath5 in the first potential skeleton of the figure , then there are no three vertex - disjoint paths from the left neighbor of @xmath5 to the terminals in @xmath4 . [ h ] . ]
[ enumskelet ] given a graph @xmath4 with @xmath107 vertices , @xmath41 terminals , and a root vertex @xmath5 , it is possible to enumerate in @xmath108 time all potential skeletons of inclusion - wise minimal trees rooted at @xmath5 and spanning the @xmath41 terminals in @xmath4 .
as shown in the proof of property [ prop : numbvert ] , the skeleton of such a tree contains at most @xmath60 junction vertices .
there are @xmath109 ways of choosing at most @xmath60 junction vertices , and , for each such choice , it follows from cayley s formula that there are at most @xmath110 different labelled trees containing only @xmath5 , the @xmath41 terminals , and the chosen junction vertices .
we can orient the edges of every labelled tree from the root @xmath5 towards the other vertices , which takes @xmath111 time per tree , and reject the labelled rooted trees that do not satisfy the definition of a potential skeleton .
the whole procedure therefore takes @xmath108 time .
we detail in this section several links between ` ml - cap - steiner - tree ` and some vertex - disjoint paths problems . we begin with a simple complexity result in the case of unit capacities . in this case ,
an optimal solution to ` ml - cap - steiner - tree ` necessarily consists of @xmath41 vertex - disjoint paths with minimum total length , each one linking @xmath5 to a terminal , and we obtain the following theorem . [ th : cstp - cap1 ] ` ml - cap - steiner - tree ` is polynomial - time solvable if @xmath112 @xmath113 .
assume the input graph @xmath4 is directed , and let us add to @xmath4 a new vertex @xmath42 and an arc @xmath114 of length 0 and capacity 1 for each terminal @xmath115 .
solving ` ml - cap - steiner - tree ` then amounts to finding @xmath41 internally vertex - disjoint paths from @xmath5 to @xmath42 , with minimum total length .
it is well - known that this can be done in polynomial time , but we briefly recall how .
we consider the graph @xmath116 obtained from @xmath4 by replacing each vertex @xmath117 by an arc @xmath118 of length 0 , and each arc @xmath119 ( resp .
@xmath120 ) by an arc @xmath121 ( resp .
@xmath122 ) having the same length as the original one .
all capacities are set equal to 1 .
it is then sufficient to determine a minimum - cost flow of @xmath123 units from @xmath5 to @xmath42 in @xmath116 by using any min - cost flow algorithm @xcite .
recall that , if the graph @xmath4 is undirected , we can transform it into a directed one by replacing each edge by two opposite arcs . in this case , only one of two opposite arcs associated to an edge carries a positive flow in the solution .
the following problem is a generalization of ` ml - vdisj - path ` .
+ * minimum - length labelled vertex - disjoint paths problem * ( ` ml - lab - vdisj - path ` ) _ input_. a graph @xmath0 ; an integer @xmath124 ; a nonnegative length function @xmath25 on @xmath24 ; a label @xmath125 on every @xmath7 ; @xmath34 disjoint vertex pairs ( @xmath38 , @xmath39 ) , each one being associated with a set @xmath126 of labels .
_ objective _ : find @xmath34 mutually vertex - disjoint paths @xmath36 of minimum total length so that @xmath37 links @xmath38 to @xmath39 and all labels on @xmath37 belong to @xmath127 @xmath40 .
+ when @xmath29 for all @xmath7 , ` ml - vdisj - path ` ( resp . `
ml - lab - vdisj - path ` ) turns into a decision problem , denoted by ` vdisj - path ` ( resp . `
lab - vdisj - path ` ) .
notice that ` ml - vdisj - path ` is the special case of ` ml - lab - vdisj - path ` where @xmath128 for @xmath129 .
we now show several links between ` ml - cap - steiner - tree ` and some variants of ` ml - vdisj - path ` and ` ml - lab - vdisj - path ` .
[ th : reductionfrommvdp ] ` ml - vdisj - path ` with @xmath34 source - sink pairs is polynomially reducible to ` ml - cap - steiner - tree ` with @xmath130 terminals .
assume first that the input graph @xmath0 of the ` ml - vdisj - path ` instance is undirected .
let @xmath46 be defined as follows : @xmath131 is obtained by adding to @xmath14 a vertex @xmath5 and @xmath132 terminals @xmath133 , @xmath134 ; @xmath135 is obtained from @xmath24 by adding an edge of capacity @xmath136 and length 0 between @xmath5 and every @xmath38 , @xmath129 , as well as edges of capacity 1 and length 0 between @xmath39 and every @xmath133 , @xmath134 .
the edges of @xmath24 keep their original length , while their capacity is fixed to @xmath34 .
we prove that solving ` ml - vdisj - path ` in @xmath4 is equivalent to solving ` ml - cap - steiner - tree ` in @xmath49 .
the construction of @xmath49 from @xmath4 is illustrated in figure [ fig : mndpcstp ] for @xmath137 , with the pair @xmath138 on every @xmath139 .
[ htbp ] with @xmath137 to ` ml - cap - steiner - tree ` in @xmath46 with @xmath140 . ] given a solution @xmath15 to ` ml - vdisj - path ` in @xmath4 , one can get a solution @xmath71 to`ml - cap - steiner - tree ` in @xmath49 of same total length by orienting all paths from @xmath38 to @xmath39 , @xmath129 , and then adding the @xmath34 arcs @xmath141 , as well as the @xmath130 arcs incident to the terminals
. now , assume there is a solution @xmath71 for ` ml - cap - steiner - tree ` in @xmath49 . since there are @xmath130 terminals while the sum of the capacities of the edges incident to @xmath5 is precisely this amount , we know that @xmath5 is @xmath141-linked to exactly @xmath136 terminals in @xmath71 , @xmath129 .
in particular , @xmath5 is @xmath142-linked to @xmath34 terminals , and these are necessarily @xmath143 , otherwise @xmath71 would contain a cycle .
using the same argument , with @xmath136 decreasing from @xmath34 to @xmath69 , we get that @xmath5 is @xmath141-linked to @xmath144 .
notice that all paths from @xmath5 to @xmath133 , @xmath145 , use the same sub - path from @xmath38 to @xmath39 .
hence , by removing from @xmath71 all arcs incident to @xmath5 and to the terminals , we get a solution @xmath15 to ` ml - vdisj - path ` with same total length .
the proof for digraphs is obtained by replacing edge " by arc " in the construction of @xmath49 .
[ cor:2mvdp - unif ] given two integers @xmath41 and @xmath26 with @xmath146 and @xmath147 , ` ml - vdisj - path ` with @xmath44 source - sink pairs is polynomially reducible to`ml - cap - steiner - tree ` with @xmath41 terminals and uniform capacity @xmath26 .
the proof is similar to the previous one .
the main difference is the definition of @xmath46 . in the undirected case , @xmath131 is obtained by adding to @xmath14 two vertices @xmath5 and @xmath18 and @xmath41 terminals @xmath148 ; @xmath135 is obtained from @xmath24 by adding the edges @xmath149 , [ r , s_2 ] , [ v , s_1 ] , [ v , t_1 ] , [ s'_1,t_2 ] , [ s'_2,t_i]$ ] for @xmath150 , and @xmath151 $ ] for@xmath152 .
the edges of @xmath24 keep their original length while those in @xmath153 have length 0 .
all capacities are set equal to @xmath26 .
we then prove that ` ml - vdisj - path ` on @xmath4 is equivalent to ` ml - cap - steiner - tree ` in @xmath49 in a similar way as in the previous theorem .
the only path that goes from @xmath5 to @xmath154 contains @xmath155 , and hence the remaining capacity on this edge is @xmath156 : this implies that the paths from @xmath5 to the @xmath26 terminals adjacent to @xmath157 must contain @xmath158 , and the rest of the proof is unchanged .
the proof for digraphs is obtained by adding arcs instead of edges to obtain @xmath49 .
[ cor : pmvdp - unif ] ` ml - vdisj - path ` with @xmath159 source - sink pairs is polynomially reducible to ` ml - cap - steiner - tree ` with @xmath160 terminals and uniform capacity @xmath34 .
again , the proof is similar to the one of theorem [ th : reductionfrommvdp ] . in this case , @xmath46 is constructed as follows .
@xmath131 is obtained by adding to @xmath14 a vertex @xmath5 , @xmath161 vertices @xmath162 and @xmath160 terminals @xmath133 with @xmath163 ; @xmath135 is obtained from @xmath24 by adding the edges @xmath164 $ ] , @xmath165 $ ] and @xmath166 $ ] for @xmath167 , as well as edges between @xmath39 and every @xmath133 with @xmath134 and edges between @xmath168 and every @xmath133 with @xmath169 .
the edges of @xmath24 keep their original length while those in @xmath153 have length 0 .
all capacities are set equal to @xmath34 .
we then prove that ` ml - vdisj - path ` on @xmath4 is equivalent to ` ml - cap - steiner - tree ` in @xmath49 in a similar way as in theorem [ th : reductionfrommvdp ] .
notice that , in this case , given any solution @xmath71 to ` ml - cap - steiner - tree ` in @xmath49 , @xmath5 is necessarily @xmath170-linked to terminals @xmath133 ( @xmath171 ) for all @xmath167 , and @xmath5 is @xmath142-linked to terminals @xmath172 ( @xmath171 ) .
_ the results stated in theorems [ th : reductionfrommvdp ] , [ cor:2mvdp - unif ] and [ cor : pmvdp - unif ] are also valid for ` ml - vdisj - path ` and ` ml - cap - steiner - tree ` with strictly positive lengths , since the arcs or edges added to @xmath4 in order to obtain @xmath49 can have arbitrary lengths .
indeed , the total length of a solution @xmath15 to ` ml - vdisj - path ` will then differ from the total length of the corresponding solution @xmath71 to ` ml - cap - steiner - tree ` by a value equal to the total length of the added arcs or edges . _
we next show that , when the number @xmath41 of terminals is fixed , ` ml - cap - steiner - tree ` is polynomially reducible to ` ml - lab - vdisj - path ` .
[ th : reductiontomlvdp ] when @xmath173 is fixed , ` ml - cap - steiner - tree ` can be reduced in polynomial time to ` ml - lab - vdisj - path ` with a fixed number of source - sink pairs .
we first consider the undirected case .
let @xmath174 be an instance of`ml - cap - steiner - tree ` in a graph @xmath4 containing @xmath41 terminals .
it follows from property [ enumskelet ] that the set of potential skeletons of optimal solutions to @xmath174 can be enumerated in @xmath109 time ( since @xmath41 is a constant ) , where @xmath107 is the number of vertices in @xmath4 . to every such potential skeleton @xmath15 ,
we associate a graph @xmath46 constructed as follows , in order to deal with vertex - disjoint paths ( and not internally vertex - disjoint paths ) . for each arc @xmath78 of @xmath15
, we create a copy @xmath175 of @xmath50 and a copy @xmath176 of @xmath18 ; hence , every vertex @xmath18 of @xmath15 is replaced in @xmath49 by @xmath177 copies of @xmath18 , where @xmath177 is the degree of @xmath18 in @xmath15 .
all vertices of @xmath4 that do not appear in @xmath15 are also put in @xmath131 ( with only one copy of each ) . for each edge
@xmath27 $ ] of @xmath4 we put an edge of same length in @xmath49 between each copy of @xmath50 and each copy of @xmath18 .
this construction is illustrated in figure [ fig : mlndp ] .
we then create a source - sink pair @xmath178 in @xmath49 for all arcs @xmath78 in @xmath15 . from property
[ prop : numbvert ] , @xmath15 contains at most @xmath179 vertices , and there are therefore at most @xmath180 such pairs . for each @xmath181 , let @xmath182 denote the number of terminals in the subtree @xmath45 of @xmath15 rooted at @xmath18 , and let @xmath183 be the set of labels associated with the source - sink pair @xmath178 .
we set @xmath184 . in the example of figure
[ fig : mlndp ] , we have @xmath185 , and @xmath186 . the label @xmath187 associated with an edge @xmath10 in @xmath49
is the capacity of the corresponding edge in @xmath4 .
an optimal solution to ` ml - lab - vdisj - path ` in @xmath49 ( if any ) corresponds to a minimum - length solution to ` ml - cap - steiner - tree ` in @xmath4 having @xmath15 as skeleton .
since we enumerate all potential skeletons , the best solution to ` ml - cap - steiner - tree ` obtained during this enumeration is an optimal solution for @xmath174 .
the proof is similar for digraphs , by replacing edge by arc .
[ htbp ] to ` ml - lab - vdisj - path ` in @xmath188 ( the length of @xmath189 $ ] is equal to 3 , and all other lengths are equal to 1 . ) ] [ cor : mcstpuniftomvdp ] when @xmath190 is fixed , ` ml - cap - steiner - tree ` ( resp . `
cap - steiner - tree ` ) with uniform capacity is polynomially reducible to ` ml - vdisj - path ` ( resp . `
vdisj - path ` ) with a fixed number of source - sink pairs .
the proof is similar to the proof of theorem [ th : reductiontomlvdp ] .
however , since the capacities are all equal to a constant @xmath26 , we do not have to use labels .
consider any potential skeleton @xmath15 : if @xmath5 has at least one successor @xmath18 such that the number of terminals in the subtree @xmath45 of @xmath15 rooted at @xmath18 is strictly larger than @xmath26 , then @xmath15 can be rejected since it can not correspond to the skeleton of a tree satisfying the capacity constraints . otherwise , an optimal ( resp .
a feasible ) solution to ` ml - vdisj - path ` is a minimum - length ( resp . a feasible ) solution to ` ml - cap - steiner - tree `
having @xmath15 as skeleton .
[ htbp ] the relations proved in this section are summarized in figure [ fig : pathsteiner ] , where a trivial polynomial reduction corresponds to a generalization of a special case .
we recall that ` vdisj - path ` is np - complete in digraphs , even with @xmath44 souce - sink pairs @xcite , while it is polynomial - time solvable in undirected graph @xcite and dags @xcite when @xmath34 is fixed .
we close this section by mentioning that the reductions given in theorems [ th : reductionfrommvdp ] and [ cor : pmvdp - unif ] are fpt - reductions @xcite with parameters @xmath34 and @xmath191 .
in this section , we prove some np - hardness and np - completeness results .
we first show that ` cap - steiner - tree ` in digraphs is np - complete even if @xmath192 is fixed , the minimum capacity @xmath193 is any value in @xmath194 , while the maximum capacity @xmath195 is at least 2 .
[ th : cstp - directed - unif - k - fixed ] ` cap - steiner - tree ` is np - complete in digraphs , even if @xmath57 is fixed , for any @xmath196 and @xmath63 , with @xmath197 .
this is a direct consequence of theorem [ th : reductionfrommvdp ] ( for @xmath198 , @xmath199 and @xmath200 ) and theorem [ cor:2mvdp - unif ] ( for @xmath146 ) . indeed , ` vdisj - path ` is np - complete in digraphs with @xmath44 source - sink pairs @xcite , and the two theorems show how to polynomially reduce ` vdisj - path ` in this case to ` cap - steiner - tree ` with the right number of terminals .
notice that we can fix the values of @xmath193 and @xmath195 in the constructed ` cap - steiner - tree ` instances ( with @xmath196 , and @xmath63 ) by assigning these two values to two different arcs incident to terminals . ` cap - steiner - tree ` with uniform capacity @xmath201 is np - complete in digraphs , even if @xmath146 is fixed . for undirected graphs
, we have the following result : [ th : cstp - undirected - non - unif ] ` cap - steiner - tree ` is np - complete in undirected graphs , even if @xmath57 is fixed and if the minimum capacity @xmath202 and the maximum capacity @xmath195 are two fixed constants , with @xmath203 and @xmath204 . we give a polynomial - time reduction from ` sat ` .
assume first that @xmath198 and all edge capacities are equal to 1 or 2 .
let @xmath205 be the set of variables and let @xmath206 be the set of clauses in an arbitrary instance @xmath174 of ` sat ` .
for each variable @xmath207 , we denote by @xmath208 ( resp .
@xmath209 ) the number of occurrences of @xmath207 ( resp .
@xmath210 ) in the clauses . we can assume , without loss of generality , that @xmath211 for every @xmath136 ( by exchanging @xmath207 and @xmath212 everywhere in the clauses , if necessary ) .
the following instance @xmath213 of ` cap - steiner - tree ` is associated to @xmath174 . for each variable @xmath207
, we construct a _ variable gadget _ as follows : we add two vertices @xmath214 and @xmath215 , and two vertex - disjoint paths between them .
the first one , @xmath37 , corresponding to literal @xmath207 , is @xmath216 , where , for each @xmath217 , the edge @xmath218 has capacity 2 , and , for each @xmath219 , the edge @xmath220 has capacity 1 .
the second path , @xmath221 , corresponding to literal @xmath210 , is @xmath222 , where , for each @xmath223 , the edge @xmath224 has capacity 2 , and , for each @xmath225 , the edge @xmath226 has capacity 1 .
the edges @xmath227 and @xmath228 also have capacity 1 .
the variable gadgets are linked together as follows : for each @xmath229 , there is an edge @xmath230 of capacity 1 . for each clause @xmath231
, we construct a _ clause gadget _ as follows : we add two vertices @xmath232 , and , for each literal @xmath207 ( or @xmath210 ) contained in @xmath231 , we add edges @xmath233 ( or @xmath234 ) and @xmath235 ( or @xmath236 ) of capacity 2 , if this literal occurs @xmath237 times in clauses @xmath238 .
the clause gadgets are linked together as follows : for each @xmath239 , there is an edge @xmath240 of capacity 2 .
we complete the construction of @xmath213 by adding a root @xmath5 and 3 terminals @xmath241 , as well as 3 edges @xmath242 , @xmath243 and @xmath244 of arbitrary capacity ( 1 is fine , but any value fits ) , an edge @xmath245 of capacity 1 , and an edge @xmath246 of capacity 2 .
the construction of @xmath213 is illustrated in figure [ fig : satcstp ] for @xmath174 with @xmath247 and @xmath248 .
solid lines have capacity 2 while dotted edges have capacity 1 .
[ htbp ] let @xmath15 be a feasible solution to @xmath213 ( if any ) . to avoid cycles , the paths from @xmath5 to @xmath249 and @xmath250 must use the same sub - path @xmath251 from @xmath5 to @xmath252 , and thus all edges of @xmath251 must have capacity 2 .
so , @xmath251 starts with the edge @xmath253 .
then , the only possibility is to use the edge @xmath254 ( or @xmath255 ) for some @xmath136 , and then the edges @xmath256 and @xmath257 ( or @xmath258 and @xmath259 ) .
the next step is to use the edge @xmath260 . using similar arguments with increasing values of @xmath261
, we get that @xmath251 necessarily contains all edges @xmath240 with @xmath262 , and ends at @xmath263 ( which is adjacent to @xmath249 and @xmath250 ) . since @xmath251 starts with the edge @xmath253 , the path @xmath264 from @xmath5 to @xmath154 starts with the edge @xmath265 . moreover , to avoid cycles , @xmath264 and @xmath251 are internally vertex - disjoint .
since @xmath266 and @xmath267 belong to @xmath251 for all @xmath261 , we conclude that , for each @xmath136 , either @xmath264 contains @xmath37 and then @xmath251 may contain only edges of @xmath221 , or @xmath264 contains @xmath221 and then @xmath251 may contain only edges of @xmath37 .
this means that , for each @xmath261 , there is a subpath of three edges of @xmath251 , from @xmath266 to @xmath267 , containing one edge of @xmath37 ( resp .
@xmath221 ) for @xmath136 such that @xmath207 ( resp .
@xmath210 ) is one of the literals contained in clause @xmath231 , and there is no @xmath123 for which the subpath from @xmath268 to @xmath269 contains one edge of @xmath221 ( resp .
@xmath37 ) . we can therefore define a satisfying truth assignment @xmath270 as follows : for each @xmath136 , if @xmath37 is a subpath of @xmath264 , then @xmath271 , else @xmath272 .
conversely , if there is a satisfying truth assignment @xmath273 for @xmath174 , we construct a feasible solution for @xmath213 as follows .
the path @xmath264 from @xmath5 to @xmath154 begins with the edge @xmath265 and , for each @xmath136 , @xmath264 has @xmath37 as a subpath if @xmath271 , and it has @xmath221 as a subpath if @xmath272 . the path @xmath251 from @xmath5
to @xmath263 begins with the edge @xmath253 and can be constructed sequentially by using edges not contained in @xmath264 ( this is always possible , since @xmath273 is a satisfying truth assignment for @xmath174 ) .
the solution to @xmath213 is then obtained by adding edges @xmath274 , @xmath243 and @xmath244 to @xmath275 .
the solution to @xmath213 corresponding to the truth assignment @xmath276 for @xmath174 is represented in figure [ fig : satcstp ] with bold lines . in order to generalize this reduction to any @xmath57 , any @xmath203 , and any @xmath277
, we attach @xmath278 terminals to @xmath252 ( instead of 2 ) and @xmath279 terminals to @xmath5 ( instead of 0 ) : the edges with capacity 1 and 2 become edges with capacity @xmath193 and @xmath278 , respectively .
moreover , since the edges incident to the terminals can have any capacity , we can set the capacity of one of them to @xmath195 .
notice that the previous result is not valid for dags since the graphs constructed in the above proof possibly contain circuits .
we now prove a complexity result for ` lab - vdisj - path ` , i.e. ` ml - lab - vdisj - path ` with lengths 0 .
[ cor : hardness - mlvdp - undirected ] ` lab - vdisj - path ` with @xmath34 source - sink pairs is np - complete in undirected graphs , and hence in digraphs , for any fixed @xmath159 , even if @xmath127 contains all labels for every @xmath280 and @xmath281 contains all labels but one . for @xmath44 ,
this follows from the proof of theorem [ th : cstp - undirected - non - unif ] . indeed , consider the two source - sink pairs @xmath282 and @xmath283 , identify the label of each edge with its capacity , and set @xmath284 and @xmath285 .
we have shown that there is a feasible solution to the instance @xmath174 of ` sat ` if and only if there are two vertex - disjoints paths @xmath286 and @xmath287 linking @xmath288 to @xmath289 and @xmath290 to @xmath157 , and such that @xmath286 uses edges of label 1 or 2 , while @xmath287 uses only edges with label 2 . for larger values of @xmath34
, we simply add dummy source - sink pairs .
note that this is in contrast with ` vdisj - path ` , which is polynomial - time solvable in undirected graphs @xcite and in dags @xcite when @xmath34 is fixed .
a similar problem has been studied in @xcite .
more precisely , ` twocol - vdisj - path ` is defined as follows : given an undirected graph in which every edge has color 1 or 2 , and two source - sink pairs @xmath291 and @xmath292 , determine whether @xmath4 contains two vertex - disjoint paths , the first one from @xmath288 to @xmath289 using only edges of color 1 , and the second one from @xmath290 to @xmath157 using only edges of color 2 . `
twocol - vdisj - path ` is the special case of ` ml - lab - vdisj - path ` where there are @xmath44 source - sink pairs , all lengths are 0 , and @xmath293 for each @xmath294 .
it is proved in @xcite that this problem is np - complete .
the reduction given in the proof of theorem [ th : cstp - undirected - non - unif ] yields an alternative proof of the np - completeness of ` twocol - vdisj - path ` : indeed , as in the previous corollary , we can fix @xmath295 and @xmath283 , and identify the colors of the edges with their capacities . also , for every variable gadget , we add a path consisting of two edges of capacity ( color ) 1 between the endpoints of the edges of capacity 2 so that @xmath286 ( the path that goes through edges of capacity 1 and 2 ) can avoid edges with color 2 .
+ the next result deals with dags and undirected graphs with uniform capacity .
[ th : unif - dag - sat ] ` cap - steiner - tree ` is np - complete in dags and undirected graphs , even in the case of uniform capacity @xmath26 , for any @xmath296 ( not depending on @xmath41 ) . let ` 3-sat`@xmath297 be the satisfiability problem in which every clause contains at most 3 variables , every variable appears in at most 3 clauses , and every litteral ( a variable or its complement ) appears in at most 2 clauses . `
3-sat`@xmath297 is known to be np - complete @xcite .
we show how to polynomially reduce ` 3-sat`@xmath297 to ` cap - steiner - tree ` with uniform capacity @xmath296 .
let @xmath298 be an instance of ` 3-sat`@xmath297 with @xmath205 as set of variables and @xmath206 as set of clauses . to obtain an instance @xmath213 of ` cap - steiner - tree ` with uniform capacity @xmath26
, we construct the following graph @xmath0 : for every variable @xmath299 , we create three vertices @xmath300 and @xmath26 terminals @xmath301 ; for every clause @xmath302 we create a terminal @xmath303 ; finally we add a vertex @xmath5 .
for the directed case , we consider the following arcs : for every @xmath304 , we create the arcs @xmath305 ; for every @xmath306 we create the arcs @xmath307 ( resp .
@xmath308 ) if @xmath207 ( resp .
@xmath210 ) is in @xmath231 .
all the arcs have capacity @xmath26 .
the resulting graph @xmath4 can clearly be obtained in polynomial time and is a dag .
for the undirected case , we consider the same graph , but we replace each arc by an edge .
the construction is illustrated in figure [ fig : sat3 ] for @xmath309 and the ` 3-sat`@xmath297 instance where @xmath310 and @xmath311 , the arcs being oriented from @xmath5 down to the terminals in the directed case . to ` cap - steiner - tree ` . ]
assume there is a truth assignment @xmath270 for @xmath174 .
we construct a feasible solution @xmath15 to @xmath213 as follows .
if @xmath312 ( resp .
@xmath313 ) then we include @xmath314 ( resp .
@xmath315 ) in @xmath15 . for each clause @xmath231
, we choose one of the true litterals in @xmath231 , say @xmath207 ( resp .
@xmath212 ) , and add the arc @xmath307 ( resp .
@xmath308 ) to @xmath15 .
finally , for all @xmath304 , we add the arcs @xmath170 , @xmath316 , @xmath317 to @xmath15 . clearly , @xmath15 is a tree rooted at @xmath5 and spanning all the terminals ; in fact , @xmath15 is a spanning tree .
for every @xmath318 with @xmath312 ( resp .
@xmath313 ) , @xmath5 is @xmath170-linked ( resp .
@xmath316-linked ) to @xmath319 , and is @xmath316-linked ( resp .
@xmath170-linked ) to at most two terminals @xmath303 associated with clauses containing @xmath212 ( resp .
@xmath207 ) , since there are at most two such clauses .
since @xmath320 , all capacity constraints are satisfied .
the solution @xmath15 associated with the truth assignment @xmath321 is represented in figure [ fig : sat3 ] with bold lines .
now , let @xmath15 be a feasible solution to @xmath213 .
consider first the directed case .
for all @xmath306 , there is at least one index @xmath136 such that either @xmath322 and we then set @xmath323 or @xmath324 and we then set @xmath325 .
if a variable did not get any value , we arbitrarily choose one , say @xmath326 .
this gives a truth assignment satisfying each clause @xmath231 .
let us verify that we have not assigned simultaneously values @xmath326 and @xmath327 to some variable .
notice first that , since @xmath15 has no cycle , @xmath5 is either @xmath170-linked or @xmath316-linked to the @xmath26 terminals @xmath301 .
since all capacities equal @xmath26 , this means that either @xmath328 or @xmath329 has no @xmath303 @xmath330 as successor , which means that we do not assign both values @xmath326 and @xmath327 to a variable @xmath207 .
consider now the undirected case .
if there is an index @xmath136 such that @xmath331 , then @xmath332 since , in this case , @xmath5 must be @xmath315-linked to @xmath319 in @xmath15 . hence , @xmath333 ( otherwise , there would be a cycle in @xmath15 ) and we can replace @xmath334 by @xmath170 . similarly ,
if @xmath335 for some index @xmath136 , we replace this arc by @xmath336 .
assume now that @xmath337 and @xmath333 for some @xmath338 and @xmath339 .
then , we have @xmath340 for some index @xmath341 and we replace @xmath342 by @xmath170 . we make a similar exchange if @xmath343 and @xmath344 . after having performed all these replacements , we get a new spanning tree @xmath71 , since each vertex ( except the root ) still has exactly one incoming arc .
moreover , @xmath71 has the same structure as the tree @xmath15 analyzed in the directed case .
we can therefore obtain a satisfying truth assignment using the same rules as above .
remember ( see section 1 ) that ` edge - cost - flow ` consists in determining aminimum - length feasible flow of @xmath41 units from @xmath42 to @xmath43 in a given graph , where the length of a flow is the total length of the arcs / edges carrying a positive flow . `
edge - cost - flow ` is very close to ` ml - cap - steiner - tree ` . indeed , given an instance of ` ml - cap - steiner - tree ` in a graph @xmath4 with @xmath41 terminals , we can construct a graph @xmath49 obtained from @xmath4 by adding a vertex @xmath345 and linking every terminal to this new vertex . solving ` edge - cost - flow ` in @xmath49 with @xmath346 and @xmath347 is then equivalent to solving ` ml - cap - steiner - tree ` in @xmath4 , except that a feasible solution is not required to be a tree .
the proof of theorem [ th : unif - dag - sat ] provides the following corollary : ` edge - cost - flow ` is np - hard in dags and undirected graphs , even if all lengths are 1 and all capacities are 1 or 2 .
consider first the directed acyclic case .
given an instance @xmath298 of ` 3-sat`@xmath297 , let us construct a graph @xmath0 , as in the proof of theorem [ th : unif - dag - sat ] , setting @xmath309 .
we then add a vertex @xmath345 and link @xmath348 @xmath349 and @xmath303 @xmath330 to @xmath345 .
for all arcs @xmath10 incident to @xmath5 , as well as those linking @xmath38 to @xmath350 and @xmath351 , we set @xmath352 , while @xmath353 is set to @xmath354 for all other arcs @xmath10 .
all arcs have capacity 2 , except those incident to @xmath345 which have capacity 1 .
we then solve an ` edge - cost - flow ` instance , looking for a flow of @xmath355 units from @xmath5 to @xmath345 .
let @xmath213 be this instance .
we prove that @xmath174 is satisfiable if and only if the total length of an optimal solution to @xmath213 is at most @xmath356 .
if there is a satisfying truth assignment for @xmath174 , we construct a solution @xmath15 to @xmath213 as in the proof of theorem [ th : unif - dag - sat ] , except that we add an arc from every terminal to @xmath345 .
it is not difficult to check that the total length of such a solution is at most @xmath357 .
consider now a solution @xmath15 to @xmath213 of total length at most @xmath357 .
since the @xmath358 arcs incident to @xmath345 have capacity 1 , they all carry a positive flow .
hence , for every @xmath359 , there is at least one index @xmath136 such that @xmath307 or @xmath308 carries a positive flow in @xmath15 .
moreover , for every @xmath318 , at least one of the arcs @xmath314 and @xmath315 carries a positive flow in @xmath15 .
therefore , the total length of @xmath15 is at least @xmath360 , which means that no other arc @xmath10 with @xmath361 can belong to @xmath15 . in particular , for every @xmath359 , there is _ exactly _ one arc in @xmath15 with a positive flow linking a vertex in @xmath362 to @xmath303 , and , for every @xmath363 , _ exactly _ one of the arcs @xmath314 and @xmath315 carries a positive flow flow in @xmath15 . if @xmath314 ( resp .
@xmath315 ) carries a positive flow in @xmath15 , then there is no flow on the arcs linking @xmath168 ( resp .
@xmath329 ) to a @xmath303 since @xmath364 ) ( resp .
@xmath316 ) has capacity 2 while 2 units of flow are used to reach @xmath350 and @xmath351 .
hence the structure of @xmath15 is the same as in the proof of theorem [ th : unif - dag - sat ] , and we can set @xmath365 if @xmath314 carries a positive flow in @xmath15 , and @xmath366 otherwise , to obtain a satisfying truth assignment for @xmath174 . to obtain a graph with uniform length 1 ,
we replace each arc @xmath10 by a path with @xmath353 arcs of length 1 .
the proof is similar for the undirected case .
we close this section by considering a final uniform case for ` ml - cap - steiner - tree ` .
[ th : unit - cost - steiner ] ` ml - cap - steiner - tree ` is np - hard in any graph , even if all capacities are equal to @xmath54 for any positive constant @xmath62 and if all lengths are 1 .
it is well - known that ` steiner - tree ` is np - hard even in the case of unit lengths @xcite . given any positive number @xmath62 , we show how to polynomially reduce ` steiner - tree ` with unit lengths to ` ml - cap - steiner - tree ` with uniform capacity @xmath54 and unit lengths .
let @xmath213 be an instance of ` steiner - tree ` in a graph @xmath46 with @xmath367 terminals and @xmath352 for all @xmath7 .
we construct a graph @xmath0 from @xmath49 by adding @xmath62 terminals and linking them to @xmath5 with edges / arcs of length 1 .
we then set all capacities to @xmath367 to obtain an instance @xmath174 of ` ml - cap - steiner - tree ` with @xmath368 terminals , @xmath352 and @xmath369 for all @xmath7 .
clearly , solving @xmath213 is equivalent to solving @xmath174 , and @xmath174 can be built from @xmath213 in polynomial time .
in this section , we assume that the number @xmath41 of terminals is fixed . the first theorem deals with undirected graphs having uniform capacity , and
complements theorems [ th : cstp - undirected - non - unif ] and [ th : unif - dag - sat ] .
[ th : undirected - unif - fixedk ] in undirected graphs with a fixed number @xmath41 of terminals and uniform capacity , ` cap - steiner - tree ` is solvable in polynomial time , and ` ml - cap - steiner - tree ` is polynomially equivalent to ` ml - vdisj - path ` with a fixed number of source - sink pairs .
the first part of the theorem is a direct consequence of theorem [ cor : mcstpuniftomvdp ] , since ` vdisj - path ` is polynomial - time solvable and even fpt in undirected graphs
when the number of source - sink pairs is fixed @xcite . the second part comes from theorems [ cor : pmvdp - unif ] and [ cor : mcstpuniftomvdp ] .
on the one hand , recall that , for a fixed number @xmath34 of source - sink pairs , the complexity of ` ml - vdisj - path ` in undirected graphs is open for a long time @xcite ( and so determining the one of ` ml - cap - steiner - tree ` in this case is as hard as settling this open problem ) ; however , there exists a probabilistic polynomial - time algorithm to solve the case with two source - sink pairs @xcite , although no deterministic one is known yet . on the other hand , corollary
[ cor : hardness - mlvdp - undirected ] shows that ` ml - lab - vdisj - path ` with lengths 0 is already np - complete in undirected graphs when @xmath159 is fixed .
the next theorem shows that ` ml - lab - vdisj - path ` is tractable in dags if @xmath34 is fixed , which will come in handy for proving that ` ml - cap - steiner - tree ` is solvable in polynomial time in dags if @xmath41 is fixed .
[ th : cds - dag - k - fixed ] in dags , ` ml - lab - vdisj - path ` is solvable in polynomial time for any fixed number of source - sink pairs .
ml - lab - vdisj - path ` by using a dynamic programming algorithm .
more precisely , consider a dag @xmath0 with a label @xmath187 on each arc @xmath7 , and with @xmath34 vertex - disjoint pairs @xmath370 and their associated label sets @xmath127 .
we first order the vertices of @xmath4 using a topological ordering , and denote by @xmath371 the position of each vertex @xmath18 in such an ordering ( i.e. , @xmath372 for all @xmath373 ) .
let @xmath374 be the set of @xmath34-tuples @xmath375 of vertices of @xmath4 , and let @xmath376 be the function such that @xmath377 is the minimum total length of a set of @xmath34 vertex - disjoint paths in @xmath4 such that the @xmath136th one goes from @xmath38 to @xmath168 and uses only arcs with labels in @xmath127 . if @xmath378 for all @xmath129 , then @xmath379 if @xmath380 for all @xmath136 , and @xmath381 otherwise .
consider now a @xmath34-tuple @xmath375 such that @xmath382 for at least one index @xmath136 , and let @xmath383 be the index such that @xmath384 .
if @xmath385 for some @xmath386 , then @xmath387 . otherwise , let @xmath388 be the set of vertices @xmath18 such that @xmath389 and there exists an arc @xmath390 whose label is in @xmath391 .
if @xmath392 then @xmath387 ; otherwise , we have @xmath393 the dynamic programming algorithm works as follows . for @xmath394 from 2 to @xmath107 , we enumerate all @xmath34-tuples @xmath395 with @xmath396 and , for each of them , we compute the corresponding value of @xmath397 .
the number of enumerated @xmath34-tuples is thus in @xmath398 and the value of each one can be computed in @xmath399 , which yields an @xmath400-time algorithm .
the optimal value to the ` ml - lab - vdisj - path ` instance is then equal to @xmath401 .
[ th : dag - fixedk ] ` ml - cap - steiner - tree ` is solvable in polynomial time in dags if @xmath41 is fixed .
this is a direct consequence of theorems [ th : reductiontomlvdp ] and [ th : cds - dag - k - fixed ] .
notice that ` ml - lab - vdisj - path ` in dags can not be fpt in @xmath34 ( unless fpt = w[1 ] ) , since ` vdisj - path ` ( i.e. , the special case where all arcs have zero length and @xmath402 for each @xmath136 ) is w[1]-hard with respect to @xmath34 in dags @xcite . moreover , ` ml - cap - steiner - tree ` can not be fpt in @xmath41 ( unless fpt = w[1 ] ) , since theorem [ cor : pmvdp - unif ] shows that ` ml - vdisj - path ` can be fpt - reduced to ` ml - cap - steiner - tree ` ( with respective parameters @xmath34 and @xmath403 ) .
in this section , we study the case where all capacities are almost equal to the number of terminals .
we first consider the case where the minimum capacity @xmath193 is at least equal to @xmath54 , where @xmath404 is an arbitrary constant . in
what follows , we denote by @xmath58 the best possible approximation ratio for ` steiner - tree ` ( @xmath405 in undirected graphs @xcite ) , and by @xmath59 the best possible approximation ratio for ` ml - vdisj - path ` with a fixed number of source - sink pairs .
as mentioned in the previous section , @xmath406 in dags , and determining whether @xmath406 or not in undirected graphs is a long - standing open problem . without loss of generality , we assume in this section that @xmath407 for all @xmath408 . if this is not the case , we modify the lengths as follows : for all @xmath408 , we multiply @xmath353 by @xmath409 if @xmath410 , where @xmath411 is the lowest common multiple of the denominators of the lengths @xmath353 , and we set @xmath352 if @xmath10 has length zero .
the first result obtained in this section complements theorem [ th : unit - cost - steiner ] and generalizes the first part of theorem [ th : undirected - unif - fixedk ] . notice that , from theorem [ th : cstp - directed - unif - k - fixed ] , ` cap - steiner - tree ` is np - complete in digraphs with uniform capacity @xmath413 for any constant @xmath414 .
[ th : cstp - large - capa - unif ] in dags and undirected graphs having uniform capacity @xmath413 , ` cap - steiner - tree ` is solvable in polynomial time and ` ml - cap - steiner - tree ` can be approximated within a ratio of @xmath415 , for any constant @xmath416 .
we first state and prove some useful properties .
let @xmath174 be an instance of ` cap - steiner - tree ` .
[ fact7.1 ] there is a feasible solution @xmath15 for @xmath174 if and only if there is a tree @xmath417 ( called _ reduced _ tree ) rooted at @xmath5 , spanning a subset @xmath418 of terminals , and such that , for every edge @xmath10 incident to @xmath5 in @xmath417 , @xmath5 is @xmath51-linked to at least @xmath62 terminals in @xmath417 .
a feasible solution for @xmath174 is clearly such a tree .
so , assume that such a tree @xmath417 exists .
we iteratively complete @xmath417 to obtain a steiner tree @xmath15 spanning also the terminals in @xmath419 as follows .
for every terminal @xmath420 , we consider any path @xmath421 from @xmath5 to @xmath43 in @xmath4 , and we add to @xmath417 the subpath of @xmath421 from @xmath18 to @xmath43 , where @xmath18 is the vertex in @xmath422 the closest to @xmath43 on @xmath421 . from the hypothesis , given any edge @xmath10 of @xmath4 ( including those not in @xmath417 ) , we know that @xmath5 is @xmath51-linked to at least @xmath62 terminals in @xmath417 . hence , @xmath15 does not violate the capacity constraints and is therefore a feasible solution to @xmath174 .
[ fact7.2 ] if @xmath417 is a _ minimal _ , i.e. inclusion - wise minimal , reduced tree ,
then each of its subtrees rooted at a vertex distinct from @xmath5 contains at most @xmath62 terminals .
let @xmath417 be a minimal reduced tree , and assume it contains a vertex @xmath423 such that @xmath424 contains at least @xmath425 terminals .
we can assume without loss of generality that @xmath18 is a child of @xmath5 .
since all terminals are leaves , we can remove one of the terminals of @xmath424 from @xmath417 to obtain a smaller reduced tree , a contradiction .
[ fact7.3 ] a minimal reduced tree contains at most @xmath426 terminals .
assume a minimal reduced tree @xmath417 contains at least @xmath427 terminals , and consider any child @xmath18 of @xmath5 in @xmath417 .
it follows from the previous claim that @xmath5 is @xmath428-linked to at least @xmath425 terminals in @xmath417 .
all terminals being leaves , we can therefore delete any terminal from @xmath417 to obtain a smaller reduced tree , a contradiction .
we can now prove theorem [ th : cstp - large - capa - unif ] .
we first consider ` cap - steiner - tree ` .
according to claim [ fact7.3 ] and property [ prop : numbvert ] , the undirected skeleton of a minimal reduced tree can contain up to @xmath429 vertices .
we therefore enumerate all labelled trees ( including potential skeletons of reduced trees ) on at most @xmath429 vertices .
we then orient each of them from the root to the leaves , and for each such rooted tree we try to replace the arcs by vertex - disjoint paths in a similar way as in theorem [ cor : mcstpuniftomvdp ] .
if such a replacement is possible , we test whether the extended skeleton is a reduced tree : in such a case , we stop the enumeration since we know from claim [ fact7.1 ] that the ` cap - steiner - tree ` instance @xmath174 has a feasible solution .
if no potential skeleton can be extended to a reduced tree , then @xmath174 has no solution : indeed , if such a solution @xmath15 exists , it contains a minimal reduced tree , whose skeleton is necessarily considered in our enumeration and then extended to a reduced tree , which leads to a contradiction .
there are @xmath430 ways of choosing at most @xmath431 terminals among @xmath41 . for each such choice of at most @xmath431 terminals
, it then follows from property [ enumskelet ] that the set of potential skeletons of minimal reduced trees spanning these terminals can be enumerated in @xmath432 time ( since @xmath62 is a constant ) .
finally , at most @xmath433 arcs must be replaced by vertex - disjoint paths in every potential skeleton ( recall that this can be done in polynomial time in dags and undirected graphs , since @xmath62 is a constant ) .
hence , the whole process takes a polynomial time .
consider now a ` ml - cap - steiner - tree ` instance @xmath213 .
we proceed as above , but instead of choosing arbitrary vertex - disjoint paths , we solve the associated ` ml - vdisj - path ` instance with a @xmath59-approximation algorithm , and , instead of stopping the enumeration when a reduced tree is found , we enumerate all of them and store the best one , denoted by @xmath434 .
notice that the total length of @xmath434 is at most @xmath59 times larger than the total length of an optimal solution to @xmath213 since such an optimal solution contains a reduced tree .
we then use a @xmath58-approximation algorithm to determine a minimum - length steiner tree @xmath435 spanning all the terminals not already spanned by @xmath434 . clearly , the total length of @xmath435 is at most @xmath58 times larger than the total length of an optimal solution to @xmath213 .
we finally build a solution @xmath15 to @xmath213 by removing from @xmath436 all arcs of @xmath435 entering a vertex with in - degree 2 in @xmath436 .
this yields a @xmath437-approximation algorithm to ` ml - cap - steiner - tree ` ( with @xmath406 for dags ) .
it follows from theorem [ th : cstp - undirected - non - unif ] that ` cap - steiner - tree ` with non - uniform capacities is np - complete in undirected graphs when the minimum capacity @xmath193 equals @xmath54 , for any constant @xmath438 .
we show however that theorem [ th : cstp - large - capa - unif ] can be extended to dags with non - uniform capacities .
[ th : cstp - large - capa - non - unif - dags ] in dags with @xmath439 , ` cap - steiner - tree ` is solvable in polynomial time and ` ml - cap - steiner - tree ` can be approximated within a ratio of @xmath440 , for any constant @xmath404 . as was the case for the previous theorem ,
we start with some claims . in particular
, we extend the definition of a reduced tree to take into account the non - necessarily uniform capacities .
let @xmath174 be an instance of ` cap - steiner - tree ` in a dag with @xmath412 for some constant @xmath416 .
if @xmath441 , then the result follows from theorem [ th : dag - fixedk ] .
so , assume @xmath442 .
[ fact : reducedtree ] there is a feasible solution @xmath15 to @xmath174 if and only if there is a tree @xmath417 ( called a _ reduced _ tree ) rooted at @xmath5 , spanning at least @xmath62 terminals , and such that , for each arc @xmath443 with capacity @xmath444 ( not only those incident to @xmath5 ) , @xmath5 is @xmath445-linked to at least @xmath446 terminals in @xmath417 . a solution for @xmath174 is clearly such a tree .
now , assume the existence of a reduced tree @xmath417 that spans a subset @xmath53 of at least @xmath62 terminals .
we complete @xmath417 to obtain a steiner tree @xmath15 spanning also the terminals in @xmath419 , as in claim [ fact7.1 ] . from the hypothesis , for each arc @xmath443 of @xmath417 , @xmath5 is @xmath445-linked to at least @xmath446 terminals in @xmath417 , and , for each arc @xmath447 of @xmath4 not in @xmath417 , @xmath5 is @xmath448-linked to at least @xmath449 terminals of @xmath417 .
hence , @xmath15 does not violate the capacity constraints and is thus a feasible solution to @xmath174 .
[ fact : dagnonunif ] in any _ minimal _ , i.e. inclusion - wise minimal , reduced tree @xmath417 , no vertex has out - degree greater than @xmath425 .
assume that a minimal reduced tree @xmath417 contains a vertex @xmath18 with at least @xmath450 outgoing arcs and let @xmath451 be a child of @xmath18 in @xmath417 .
let @xmath452 denote the subtree obtained from @xmath417 by removing all vertices of @xmath453 .
since every subtree @xmath454 of @xmath417 rooted at a child @xmath75 of @xmath18 contains at least one terminal ( otherwise @xmath417 is not minimal ) , we know that @xmath455 ( and thus also @xmath452 ) spans at least @xmath456 terminals . hence , for every arc @xmath443 of @xmath452 not on the path from @xmath5 to @xmath18 in @xmath417 and not in @xmath455 , we know that @xmath5 is @xmath445-linked to at least @xmath457 terminals in @xmath452 . for every child @xmath75 of @xmath18 in @xmath452 , we know that , for any arc @xmath443 in @xmath458 , @xmath5 is @xmath445-linked to the at least @xmath459 terminals in the subtree of @xmath455 obtained by removing all the vertices of @xmath460 . finally , for any arc @xmath443 on the path from @xmath5 to @xmath18 in @xmath452 , we know that @xmath5 is @xmath445-linked to at least @xmath446 terminals in @xmath452 , since this was the case in @xmath417 .
hence , we have proved that @xmath452 satisfies the definition of a reduced tree , and is included in @xmath417 while being smaller , a contradiction .
[ fact : height ] any directed path in the skeleton of a minimal reduced tree @xmath417 contains at most @xmath425 arcs .
assume that the skeleton of a minimal reduced tree @xmath417 contains a directed path with at least @xmath450 arcs .
without loss of generality , we can choose such a path @xmath421 from @xmath5 to a terminal @xmath461 , since each leaf is a terminal ( otherwise @xmath417 is not minimal ) .
we denote by @xmath18 the predecessor of @xmath43 in @xmath421 . since each internal vertex of @xmath421 has degree at least 3 in the skeleton of @xmath417 ( and hence in @xmath417 ) , and since @xmath417 is minimal , we know that @xmath417 spans at least @xmath462 terminals .
we now remove the path from @xmath18 to @xmath43 in @xmath417 and thus obtain a subtree @xmath452 spanning at least @xmath463 terminals . using arguments similar to those in the proof of claim [ fact : dagnonunif ] , it is easy to check that @xmath452 satisfies the definition of a reduced tree , which means that @xmath417 was not minimal , a contradiction
. we can now prove theorem [ th : cstp - large - capa - non - unif - dags ] .
it follows from claims [ fact : dagnonunif ] and [ fact : height ] that the skeleton of a minimal reduced tree has maximum out - degree @xmath425 and maximum height @xmath425 .
hence , the undirected skeleton of a minimal reduced tree contains at most @xmath464 terminals ( which are its leaves ) , and it follows from property [ prop : numbvert ] that such a skeleton has at most @xmath465 vertices .
we therefore enumerate all trees with at most @xmath465 vertices , keeping only those that span at least @xmath62 and at most @xmath466 terminals , in order to ensure that any potential skeleton of a minimum reduced tree is enumerated .
each such enumerated tree is oriented from @xmath5 to the leaves , and we then try to replace its arcs by vertex - disjoint paths but , unlike in theorem [ th : cstp - large - capa - unif ] , when replacing an arc @xmath78 of a tree by a path , we impose that each arc of the path from @xmath50 to @xmath18 has a label ( capacity ) @xmath467 , where @xmath468 is the number of terminals which are not descendant of @xmath18 in the tree . in other words , instead of solving a ` vdisj - path ` instance , we solve a ` lab - vdisj - path ` instance .
if all arcs can be replaced by labelled vertex - disjoint paths , we test whether the extended skeleton is a reduced tree : in such a case , we stop the enumeration , since we know from claim [ fact : reducedtree ] that @xmath174 has a feasible solution . otherwise , we conclude that @xmath174 has no solution .
let us show that the whole process takes a polynomial time .
@xmath466 being a constant , this means that , from cayley s formula and from the number of ways for choosing at most @xmath465 vertices among @xmath107 , there is a polynomial number of labelled trees to enumerate . besides , since the number of arcs that must be replaced by vertex - disjoint paths is at most @xmath469 in each tree , this means , from theorem [ th : cds - dag - k - fixed ] , that the associated ` lab - vdisj - path ` instance can be solved in polynomial time .
consider now a ` ml - cap - steiner - tree ` instance @xmath213 .
we proceed in a similar way as in theorem [ th : cstp - large - capa - unif ] .
more precisely , instead of choosing arbitrary vertex - disjoint paths , we solve an ` ml - lab - vdisj - path ` instance with a fixed number of vertex pairs , which takes a polynomial time in dags according to theorem [ th : cds - dag - k - fixed ] .
however , instead of stopping the enumeration when a reduced tree is found , we enumerate all of them and store the best one , that we denote by @xmath434 .
the total length of @xmath434 is a lower bound on the total length of an optimal solution to @xmath213 , since such an optimal solution contains a reduced tree .
we then use a @xmath58-approximation algorithm to determine a minimum - length steiner tree @xmath435 spanning all the terminals not already spanned by @xmath434 .
the total length of @xmath435 is at most @xmath58 times larger than the total length of an optimal solution to @xmath213 .
we finally build a solution @xmath15 to @xmath213 by removing from @xmath436 all arcs of @xmath435 entering a vertex with in - degree 2 in @xmath436 .
this yields a @xmath470-approximation algorithm for ` ml - cap - steiner - tree ` in dags .
the results given in this section generalize the main results known about the complexity and approximation of ` steiner - tree ` . if @xmath472 , then any steiner tree is a feasible solution to ` cap - steiner - tree ` , and thus ` ml - cap - steiner - tree ` is equivalent to ` steiner - tree ` , which can be solved in polynomial time when @xmath41 is fixed @xcite .
so consider the case where @xmath473 . in
what follows , we denote by @xmath474 the subset of arcs / edges with capacity at least @xmath41 .
let @xmath174 be an ` ml - cap - steiner - tree ` instance and let @xmath15 be an optimal solution to @xmath174 .
let @xmath75 be the closest vertex to @xmath5 in @xmath15 having out - degree at least 2 ( with possibly @xmath475 ) .
all arcs on the path linking @xmath5 to @xmath75 are in @xmath474 , while those in @xmath476 can have any capacity since @xmath5 is @xmath10-linked to at most @xmath60 terminals for all @xmath10 in @xmath476 .
moreover , @xmath476 spans all terminals and , since we can assume that all vertices without outgoing arcs in @xmath15 are terminals , we know that @xmath15 contains at most @xmath60 vertices of degree at least 3 ( see the proof of property [ prop : numbvert ] ) .
assume we can find in @xmath4 a vertex @xmath75 and two terminals @xmath477 and @xmath478 such that there are three internally vertex - disjoint paths : @xmath479 from @xmath5 to @xmath75 ( with possibly @xmath475 ) with all its arcs in @xmath474 , @xmath480 from @xmath75 to @xmath477 , and @xmath481 from @xmath75 to @xmath478 .
we can then build a feasible solution to @xmath174 by extending @xmath482 arbitrarily to obtain a steiner tree spanning all terminals .
indeed , any arc in @xmath482 has a residual capacity @xmath483 , and there are @xmath61 other terminals to span in order to get a steiner tree .
conversely , if there is a feasible solution to @xmath174 , then there is such a triple @xmath484 .
[ claim : shortpath ] let @xmath15 be an optimal solution to an instance @xmath174 of ` ml - cap - steiner - tree ` in a graph @xmath0 with @xmath485 and @xmath410 for all @xmath7 .
let @xmath75 be the vertex with out - degree at least 2 the closest to @xmath5 in @xmath15 , and let @xmath479 be the path from @xmath5 to @xmath75 in @xmath15 .
then all shortest paths from @xmath5 to @xmath486 in @xmath487 intersect @xmath476 only at @xmath75 , and @xmath479 is one of them .
first notice that @xmath488 and all arcs of @xmath479 belong to @xmath474 .
let @xmath489 be any shortest path from @xmath5 to @xmath75 in @xmath49 , and let @xmath490 be the set of vertices that belong to both @xmath489 and @xmath476 . if @xmath491 then @xmath492 is a feasible solution to @xmath174 , which means that @xmath493 , otherwise @xmath15 would not be optimal .
so assume @xmath494 and let @xmath71 be the tree obtained from @xmath15 by replacing @xmath479 by @xmath489 , and by removing all arcs @xmath495 with @xmath496 , to ensure that each vertex ( except @xmath5 ) still has in - degree 1 .
notice that @xmath497 , since @xmath498 and at least one arc ( of length @xmath499 ) is removed from @xmath476 to obtain @xmath71 .
then , we remove all arcs @xmath78 such that @xmath500 contains no terminal ( since they are useless in a solution to @xmath174 ) . this way , we obtain a new tree @xmath501 rooted at @xmath5 , spanning all terminals , and such that @xmath502 and @xmath503 @xmath504 .
this is illustrated on figure [ fig : cor77 ] .
let @xmath505 be a vertex in @xmath506 , and let @xmath43 be any terminal in @xmath507 . in @xmath501
, there is a path from @xmath5 to @xmath43 , while there is no path from @xmath75 to @xmath43 .
so @xmath508 , which is the closest vertex to @xmath75 on @xmath509 verifying @xmath510 , has outdegree at least 2 in @xmath501 .
let @xmath511 be the vertex with outdegree at least 2 which is the closest to @xmath5 in @xmath501 ( @xmath511 belongs to @xmath512 , from the previous paragraph ) , and let @xmath513 be the path from @xmath5 to @xmath511 in @xmath501 .
note that the only vertices @xmath18 in @xmath501 with @xmath514 are those on @xmath513 , and that all arcs on @xmath513 have capacity at least @xmath41 since they also belong to @xmath489 .
moreover , since @xmath503 for every child @xmath18 of @xmath511 in @xmath501 , all arcs in @xmath515 only need to have capacity @xmath60 .
hence , @xmath501 is a feasible solution to @xmath174 with @xmath516 , a contradiction . .
] we now consider ` ml - cap - steiner - tree ` with @xmath471 and show that , when @xmath41 is fixed , it is solvable in polynomial time .
[ th : large - capas - fixedk ] if @xmath41 is fixed and @xmath517 , ` ml - cap - steiner - tree ` can be solved by an algorithm whose running time is polynomial , and whose only non fpt factor with respect to k is @xmath518 . in particular , ` ml - cap - steiner - tree ` is solvable in polynomial time if @xmath519 .
the case @xmath472 has already been settled at the beginning of this section .
so , assume @xmath473 .
given an instance @xmath174 of ` ml - cap - steiner - tree ` , we solve @xmath174 as follows .
we consider all vertices @xmath75 such that @xmath75 is either the root @xmath5 or a vertex of degree at least 3 in @xmath4 .
for each such vertex @xmath75 : * we determine a shortest path @xmath479 from @xmath5 to @xmath75 in @xmath487 , and we denote by @xmath520 the subgraph of @xmath4 obtained by removing all vertices of @xmath479 , except @xmath75 ; * we consider all pairs @xmath521 of distinct terminals and all subsets @xmath490 of at most @xmath522 vertices @xmath523 of degree at least 3 in @xmath520 .
so , let @xmath524 be such a quadruple .
we first determine two internally vertex - disjoint paths @xmath480 and @xmath481 linking @xmath75 to @xmath477 and to @xmath478 in @xmath520 , and such that @xmath525 contains all vertices of @xmath490 and has minimum total length . as in the proof of theorem [ th : cstp - cap1 ] , this can be done in polynomial time : we add a sink @xmath42 and two arcs @xmath526 and @xmath527 , and we determine two internally vertex - disjoint paths of minimum total length from @xmath75 to @xmath42 by using a min - cost flow algorithm ; in addition , we impose a flow equal to 1 on each arc @xmath118 corresponding to a vertex @xmath18 of @xmath490 in the graph @xmath116 obtained from @xmath520 ( as in the proof of theorem [ th : cstp - cap1 ] ) , to ensure that the paths contain @xmath490 .
assume we are able to find the two internally vertex - disjoint paths @xmath480 and @xmath481 in @xmath520 .
we then consider the graph @xmath528 obtained from @xmath520 by assigning a length 0 to all arcs on @xmath525 , and we determine a directed tree @xmath529 of minimum total length in @xmath528 , rooted at @xmath75 , and spanning all terminals in @xmath530 .
let @xmath531 denote the set of arcs @xmath78 in @xmath529 with @xmath18 not belonging to @xmath525 .
we finally build a solution to @xmath174 by taking all arcs of @xmath532 . among all built solutions ,
we keep the best one , which we denote by @xmath533 .
we now prove that @xmath533 is an optimal solution to @xmath174 .
let @xmath534 be an optimal solution to @xmath174 , and let @xmath75 be the vertex in @xmath534 the closest to @xmath5 with out - degree at least 2 .
let @xmath535 and @xmath536 be two children of @xmath75 in the skeleton of @xmath534 , and let @xmath477 ( resp .
@xmath478 ) be a terminal in @xmath537 ( resp .
@xmath538 ) closest to @xmath535 ( resp .
@xmath536 ) in terms of the number of vertices on the path linking them in the skeleton of @xmath537 ( resp .
@xmath538 ) .
we denote by @xmath539 and @xmath540 the paths in @xmath534 linking @xmath5 to @xmath75 , @xmath75 to @xmath477 , and @xmath75 to @xmath478 , respectively .
finally , let @xmath490 be the set of vertices @xmath523 on @xmath541 having degree at least 3 in @xmath534 , and let @xmath542 denote the set of arcs in @xmath534 that do not belong to @xmath543 .
we first prove that the proposed algorithm considers the quadruple @xmath524 .
since @xmath75 has out - degree at least 2 , it is either the root @xmath5 or a vertex of degree at least 3 in @xmath4 .
it follows from claim [ claim : shortpath ] that @xmath520 contains all vertices of @xmath490 .
hence , we only have to prove that @xmath544 .
let @xmath545 ( resp .
@xmath546 ) be the number of vertices in the skeleton of @xmath537 ( resp .
@xmath538 ) . if @xmath547 , the path from @xmath535 to @xmath477 in the skeleton of @xmath534 is reduced to @xmath548 vertex @xmath549 ; otherwise , @xmath535 has out - degree at least 2 in @xmath534 , and we know from the proof of property [ prop : pathlength ] that the path from @xmath535 to @xmath477 in the skeleton of @xmath534 has at most @xmath550 vertices .
similarly , the path from @xmath536 to @xmath478 in the skeleton of @xmath534 has at most @xmath551 vertices .
hence @xmath490 contains at most @xmath552 vertices .
since @xmath75 has out - degree at least 2 , it follows from property [ prop : numbvert ] that the skeleton of @xmath553 contains at most @xmath180 vertices , which implies @xmath554 .
the sum @xmath552 is therefore maximized for @xmath555 , which implies that @xmath490 contains at most @xmath522 vertices .
we now prove that @xmath556 .
let @xmath532 be the solution returned by the proposed algorithm for the quadruple @xmath524 .
it follows from claim [ claim : shortpath ] that @xmath557 , and that @xmath520 contains all vertices of @xmath553 .
since @xmath558 and @xmath540 are two internally vertex - disjoint paths linking @xmath75 to @xmath477 and to @xmath478 in @xmath520 , we have @xmath559 .
consider now the set @xmath560 of arcs @xmath78 in @xmath542 with @xmath18 not belonging to @xmath525 , and let @xmath15 be the tree obtained from @xmath553 by replacing @xmath558 and @xmath540 by @xmath480 and @xmath481 , and by removing the arcs in @xmath561 .
note that @xmath15 is a tree rooted at @xmath75 , spanning all terminals in @xmath562 , and with total length at most equal to @xmath563 in @xmath528 ( since all arcs in @xmath525 have length 0 in @xmath528 ) . hence , @xmath564 in @xmath528 , which implies @xmath565 in @xmath4 . in summary
, @xmath566 the total number of possible quadruples is @xmath567 . for each of them
, we have to compute a shortest path , a minimum - cost flow , and an optimal steiner tree spanning @xmath61 terminals .
the latter problem can be solved in time fpt with respect to the number of terminals @xcite , and the other two problems can be solved in polynomial time @xcite . together with theorem [ th : undirected - unif - fixedk ] , the previous theorem shows , in particular , that theorems [ th : cstp - directed - unif - k - fixed ] and [ th : cstp - undirected - non - unif ] are best possible
. it also implies , together with theorem [ th : cstp - cap1 ] , that ` ml - cap - steiner - tree ` is polynomial - time solvable if @xmath198 and all capacities are equal .
when @xmath41 is part of the input ( i.e. , not fixed ) , we have the following result , which complements theorem [ th : unit - cost - steiner ] . [
th : large - capas - non - fixedk ] if @xmath568 , ` cap - steiner - tree ` is solvable in polynomial time and ` ml - cap - steiner - tree ` can be approximated within a ratio of @xmath440 .
we use the same ideas as those used in the proof of the previous theorem .
more precisely , for solving an instance @xmath174 of ` cap - steiner - tree ` , we enumerate all triples @xmath484 , where @xmath75 is either the root @xmath5 or a vertex of degree at least 3 in @xmath4 , and @xmath569 both belong to @xmath9 . for each such triple , we determine a shortest path @xmath479 from @xmath5 to @xmath75 in @xmath487 , remove all vertices of @xmath479 , except @xmath75 , to create @xmath520 , and determine two internally vertex - disjoint paths @xmath480 and @xmath481 from @xmath75 to @xmath477 and to @xmath478 in @xmath520 .
this can be done in polynomial time using path and flow techniques similar to those used in the previous proof ( without lengths on the arcs ) .
if we succeed in finding the three paths @xmath570 for a triple @xmath484 , then we greedily complete their union into a steiner tree rooted at @xmath5 and spanning all terminals , which gives a solution to @xmath174 .
otherwise , @xmath174 does not have any feasible solution .
all this can be done in polynomial time , since there are @xmath571 triples @xmath484 to enumerate .
for an instance @xmath213 of ` ml - cap - steiner - tree ` , we again enumerate all triples @xmath484 , and determine for each such triple a shortest path @xmath479 from @xmath5 to @xmath75 in @xmath49 , as well as two internally vertex - disjoint paths of shortest total length , @xmath480 and @xmath481 from @xmath75 to @xmath477 and to @xmath478 in @xmath520 ( as in the proof of theorem [ th : large - capas - fixedk ] ) .
if we succeed in finding the three paths @xmath570 for a triple @xmath484 , we then use a @xmath58-approximation algorithm to determine a directed tree @xmath529 of minimum total length , rooted at @xmath5 , and spanning all terminals in @xmath530 .
let @xmath531 be the set of arcs @xmath78 in @xmath529 with @xmath18 not belonging to @xmath572 ; we build a solution to @xmath213 by taking all arcs of @xmath532 . among all built solutions ,
we keep the best one , which we denote by @xmath533 .
now , let @xmath534 be an optimal solution to @xmath213 , and let @xmath75 be the vertex in @xmath534 the closest to @xmath5 with out - degree at least 2 .
let @xmath535 and @xmath536 be two distinct children of @xmath75 in @xmath534 , and let @xmath477 be a terminal in @xmath537 , and @xmath478 a terminal in @xmath538 .
the triple @xmath484 is considered in our enumeration , and we clearly have @xmath573 and @xmath574 .
hence , @xmath575 .
again , the whole process takes a polynomial time .
indeed , there are @xmath571 enumerated triples , and , for each of them , we have to determine a shortest path , a minimum - cost flow , and a @xmath58-approximate solution to an instance of ` steiner - tree ` .
all these problems can be solved in polynomial time .
we have studied the complexity of ` ml - cap - steiner - tree ` in digraphs , dags and undirected graphs , and we have dealt with any possible case with respect to all the parameters that we considered ( minimum and maximum capacities , lengths , and number of terminals ) . moreover , whenever ` ml - cap - steiner - tree ` was intractable while ` cap - steiner - tree ` , the case with lengths @xmath56 , was not , we have provided approximation results for ` ml - cap - steiner - tree ` nearly as good as the best ones for ` steiner - tree ` .
while we have also obtained some results about the parameterized complexity of ml - cap - steiner - tree , several questions remain open in this area : * the results associated with leaves 11 and 13 in figure [ fig : arbres ] are best possible , since the fpt - reduction from vdisj - path parameterized by @xmath34 described in theorem [ cor : pmvdp - unif ] shows in particular that cap - steiner - tree is w[1]-hard with respect to either @xmath41 or @xmath62 in this case , even with uniform capacities ( note that , in this reduction , we have @xmath191 and @xmath576 ) .
* however , the result associated with leaf 9 in figure 1 may not be the best possible one ( i.e. , this case might actually be fpt with respect to @xmath62 ) , since in undirected graphs vdisj - path is fpt with respect to @xmath34 ( so theorem [ cor : pmvdp - unif ] does not provide any useful information in this case ) .
* finally , we think that the main open problem is related to the result provided in theorem [ th : large - capas - fixedk ] ( and associated to leaf 5 in figure [ fig : arbres ] ) . we have proved that ml - cap - steiner - tree is polynomial - time solvable in this case , hence generalizing the same result already known for steiner - tree , but it may actually be fpt with respect to @xmath41 : in particular , notice that this is indeed the case for steiner - tree .
this work was done with the support of the gaspard monge program for optimization and operations research ( pgmo ) http://www.fondation-hadamard.fr/fr/pgmo .
j. cong , a. b. kahng , k .- s .
leung ( 1998 ) , _ efficient algorithms for the minimum shortest path steiner arborescence problem with applications to vlsi physical design_. ieee trans .
on cad of integrated circuits and systems 17 , 2439 .
g. duan , y. yu ( 2003 ) , _ distribution system optimization by an algorithm for capacitated steiner tree problems with complex flows and arbitrary cost functions _ , international journal of electrical power and energy systems 25 , 515523 .
e. uchoa , r. fukasawa , j. lysgaard , a. pessoa , m. poggi de arago , d. andrade ( 2008 ) , _ robust branch - cut - and - price for the capacitated minimum spanning tree problem over a large extended formulation _ , math . program .
, ser . a 112 , 443472 .
d. watel , m .- a .
weisser , c. bentz , d. barth ( 2015 ) _ an fpt algorithm in polynomial space for the directed steiner tree problem with limited number of diffusing nodes _ , information processing letters 115 , 275279 . | given a graph @xmath0 with a root @xmath1 , positive capacities @xmath2 , and non - negative lengths @xmath3 , the minimum - length ( rooted ) edge capacitated steiner tree problem is to find a tree in @xmath4 of minimum total length , rooted at @xmath5 , spanning a given subset @xmath6 of vertices , and such that , for each @xmath7 , there are at most @xmath8 paths , linking @xmath5 to vertices in @xmath9 , that contain @xmath10 .
we study the complexity and approximability of the problem , considering several relevant parameters such as the number of terminals , the edge lengths and the minimum and maximum edge capacities . for all but one combinations of assumptions regarding these parameters
, we settle the question , giving a complete characterization that separates tractable cases from hard ones .
the only remaining open case is proved to be equivalent to a long - standing open problem .
we also prove close relations between our problem and classical steiner tree as well as vertex - disjoint paths problems .
= 0.5 cm |
the ambiguity of an nfa @xmath0 measures the degree of nondeterminism employed by @xmath0 as a function of the input size : let @xmath8 be the number of accepting computations of @xmath0 on input @xmath9 and define @xmath10 to be the ambiguity of @xmath0 .
there are related complexity measures such as the advice and the leaf complexity of @xmath0 . to describe their definition let @xmath11 be the computation tree of @xmath0 on input @xmath9 .
then advice@xmath12 is the maximum , over all paths in @xmath13 from the root to a leaf , of the number of nodes with at least two children and @xmath14 is the advice complexity of @xmath0 .
the leaf complexity of @xmath0 determines the maximal number of computations for inputs of length @xmath1 .
thus , if leaf@xmath15 is the number of leaves of @xmath11 , then @xmath16 for a minimal nfa @xmath0 these measures are related as follows @xcite @xmath17 and , since @xmath18 is at most linear , leaf complexity and ambiguity are polynomially related , provided both are at least linear .
since leaf complexity is either bounded by a constant or at least linear but polynomially bounded , or otherwise exponential in the input length , we obtain that ambiguity is either bounded by a constant or bounded by a polynomial or at least exponential @xcite .
advice and leaf complexity are rather coarse measures , since advice and leaf complexity of an unambiguous nfa may be linear .
ambiguity on the other hand also influences the tractability of algorithmic questions .
for instance , for any fixed @xmath19 it can be determined efficiently whether two nfa s of ambiguity at most @xmath20 are equivalent , resp . whether the ambiguity of a given nfa is at most @xmath20 @xcite .
how large is the decrease in conciseness , i.e. , the increase in the number of states , if ambiguity is restricted ? to study this question , four classes of nfa s , namely una ( unambiguous nondeterministic automata ) , fna ( finitely ambiguous nfa ) , pna ( polynomially ambiguous nfa ) and ena ( exponentially ambiguous nfa ) are introduced in @xcite . the classification into fna s
, pna s or ena s can be performed efficiently @xcite .
remember that the ambiguity of an nfa @xmath0 is either at least exponential or at most polynomial and hence an nfa is either a pna or an ena .
leung @xcite shows that there are ena s @xmath21 with @xmath1 states such that any equivalent pna has at least @xmath22 states .
hence ena s can be far more succinct than pna s .
subsequently a similar result , applicable to a larger class of languages , was shown in @xcite by using methods of communication complexity .
in particular , the conciseness problem for pna s can be reduced to the following communication result for the iterated language of non - disjointness .
let @xmath23 be the alphabet of all subsets of @xmath24 of size @xmath25 and set @xmath26 thus @xmath27 consists of all strings @xmath28 where all pairs @xmath29 correspond to overlapping subsets .
we assume the standard communication model with two players , alice and bob , where alice receives @xmath30 and bob receives @xmath31 .
( observe that @xmath32 has small nfa s with poly@xmath33 states . )
[ fact ] ( @xcite , pages 51 - 53 ) .
let @xmath34 be arbitrary .
if a deterministic protocol @xmath35 accepts only strings from @xmath36 and if at most @xmath37 messages are exchanged , then @xmath35 accepts at most @xmath38 strings from @xmath36 .
( @xmath39 is a sufficiently small constant ) . of particular interest
are fna s , for instance since their equivalence problem is efficiently solvable .
however a separation of fna s and pna s has remained open for almost twenty years @xcite .
we are able to show such a separation and even prove a hierarchy result for polynomial ambiguity . to describe our result
we introduce the languages used in the separation . for a language @xmath40 of strings of identical
length define @xmath41 thus the input is partitioned into blocks of identical length and an input is accepted iff at least @xmath20 blocks belong to the finite set @xmath40 .
now assume that @xmath40 can be recognized by a small nfa @xmath0 .
since @xmath40 is a finite set , we can recognize @xmath42 by an nfa with ambiguity @xmath6 , if we increase the size of @xmath0 by at most the factor @xmath20 .
how should the languages @xmath40 look like ? in a first attempt set @xmath43 as the language of inequality between @xmath25-bit strings .
then @xmath40 is recognizable by an nfa with @xmath44 states and ( bounded ) ambiguity @xmath25 .
but @xmath45 is also recognizable with @xmath44 states and ambiguity @xmath25 : guess a position @xmath46 and accept @xmath47 if @xmath48 for some @xmath49 . what went wrong ?
few advice bits suffice and these advice bits can be remembered . in our second ( and successful ) attempt
we therefore set @xmath50 , where we work with the iterated language of non - disjointness from fact [ fact ] .
this construction has two advantages .
firstly , @xmath40 has a small nfa .
secondly , at least intuitively , the number of guesses required for @xmath40 increases exponentially with @xmath51 and hence a small nfa s for @xmath52 can not remember sequences of @xmath51 guesses . our main result verifies this intuition . [ main ]
let @xmath53 be arbitrary . set @xmath54 and @xmath55 .
any nfa for @xmath42 with ambiguity @xmath7 has at least @xmath56 states .
however , @xmath42 can be recognized by an nfa with ambiguity @xmath6 and size @xmath57 .
observe that we have obtained the claimed separation of fna s and pna s for @xmath58 , but theorem [ main ] also establishes a hierarchy of polynomial ambiguity .
we start by proving theorem [ main ] for @xmath58 .
let @xmath50 and assume that the nfa @xmath0 recognizes @xmath59 with sublinear ambiguity .
observe that all strings in @xmath40 have length @xmath60 and hence strings in @xmath45 have blocks of identical length @xmath60 .
we set @xmath61 , where @xmath23 is the alphabet of @xmath62 .
finally set @xmath63 thus , as in the definition of @xmath59 , the input is partitioned into blocks and an input is accepted iff _ no _ block belongs to the finite set @xmath40 .
the computationally hardest task for the nfa @xmath0 is to separate @xmath64 from @xmath45 .
the critical part of the argument is to exploit the limitation of sublinear ambiguity .
let @xmath65 be the set of states of @xmath0 . in section [ autocom ]
we construct states @xmath66 such that at least @xmath67 strings in @xmath40 have a computation starting in @xmath68 and ending in @xmath69 .
moreover we show in lemma [ densit ] that for any string @xmath70 there is a string @xmath71 such that strings @xmath72 with period @xmath73 can be `` stored '' in a `` launching cycle '' before reaching @xmath68 and in a `` storage cycle '' after leaving @xmath69 .
the launching cycle has the form @xmath74 and allows to reach @xmath68 via a computation @xmath75 ; analogously the storage cycle is built from computations @xmath76 and @xmath77 .
so far the launching cycle is harmless , since it delivers strings in @xmath78 to state @xmath68 , but these strings can not use computations from @xmath68 to @xmath69 which may be reserved for strings in @xmath40 .
however , if a single occurrence of @xmath79 within @xmath72 is replaced by an impostor string @xmath80 and if the launching cycle does not detect the replacement , then @xmath0 is forced into linear ambiguity , provided the impostor @xmath81 can also hide at a matching position within the storage cycle ( see lemma [ 1 ] ) .
thus the nfa @xmath0 has to solve the `` detection problem '' , namely it has to detect whether an impostor @xmath80 has replaced an occurrence of @xmath70 in both cycles .
the detection problem is set up in such a way that * at least @xmath67 strings from @xmath40 are accepted , namely those strings @xmath80 with a computation @xmath82 , and * all strings @xmath81 which for some @xmath70 survive in matching positions within both cycles are rejected . in particular , all strings in @xmath83 are rejected , since a string @xmath84 is its own impostor .
observe that no string @xmath81 is simultaneously accepted as well as rejected , since all impostors have to be detected .
@xmath0 may try to solve the detection problem unconventionally for instance by allowing a potential impostor @xmath81 to survive undetected within the launching and storage cycle , but not allowing @xmath81 to survive in _ matching positions _ within both cycles .
also @xmath0 does not have to solve the detection problem completely , since it can tolerate an impostor @xmath81 without a computation @xmath85 .
we then simulate @xmath0 in section [ dcom ] by a nondeterministic communication protocol which rejects all strings in @xmath83 , accepts at least @xmath67 strings in @xmath40 and does not simultaneously accept and reject a string in @xmath83 ( see lemma [ com ] ) .
thus we have reduced the problem of avoiding linear ambiguity for nfa s recognizing @xmath59 to a communication problem in which a rather small minority of strings in @xmath40 has to be separated from all of @xmath83 .
we show in lemma [ det ] how to transform such a nondeterministic protocol into a deterministic protocol by increasing the number of messages only subexponentially .
we are left with a deterministic protocol which rejects all strings in @xmath83 and accepts at least @xmath67 strings in @xmath40 .
finally the argument concludes with an application of fact [ fact ] .
thus , as in the case of exponential ambiguity , we again have reduced the conciseness problem to an investigation of deterministic protocols which recognize a `` small , but significant chunk '' of a given product language .
the general case of ambiguity @xmath6 is tackled in section [ hierarchy ] . showing the existence of launching and storage cycles
has now become a more complex problem .
previously it was sufficient that the periodic string @xmath86 was `` living '' in the one launching and the one storage cycle .
now we have to work with a vector @xmath87 of states and have to move @xmath86 to @xmath68 and all the way from @xmath88 to @xmath89 for all @xmath90 and finally from @xmath91 to an accepting state .
we begin by utilizing the special structure of the languages @xmath59 .
let @xmath0 be an nfa for @xmath45 with initial state @xmath92 .
let @xmath93 be an arbitrary state of @xmath0 .
* we say that a string @xmath94 * reaches * state @xmath93 iff there is a string @xmath95 and a computation for @xmath96 which starts in @xmath92 and ends in @xmath93 .
moreover state @xmath93 * accepts * @xmath97 iff there is a string @xmath98 and an accepting computation for @xmath99 starting in @xmath93 . *
a pair @xmath100 of states of @xmath0 is * critical * for the pair @xmath101 iff all strings in @xmath102 reach @xmath68 and all strings in @xmath103 are accepted by @xmath69 .
our next goal is to construct a pair @xmath104 such that for all strings @xmath105 acceptance is `` decided '' by critical pairs .
in particular we construct @xmath106 such that there are accepting computations of the form @xmath107 for a final state @xmath108 and a critical pair @xmath100 for @xmath109 .
the crucial advantage of a critical pair is that all strings in @xmath110 reach @xmath68 and all strings in @xmath111 are accepted by @xmath69 ; in particular , there is no transition @xmath82 for a string @xmath112 and acceptance is indeed decided by @xmath100 .
[ onetwo ] let @xmath0 be an nfa for @xmath45 .
then there are strings @xmath113 such that @xmath114 we process the states of @xmath0 in two phases . in the first phase we construct a string @xmath115 such that each state
@xmath93 is either _ alive _ for @xmath116 ( i.e. , all strings in @xmath102 reach @xmath93 ) or _
dead _ for @xmath116 ( i.e. , no string in @xmath102 reaches @xmath93 ) .
the construction process proceeds iteratively by processing all states @xmath93 of @xmath0 in an arbitrary order .
we begin by setting @xmath117 . when processing state @xmath93 we differentiate two cases .
* case 1 : * all strings in @xmath102 reach @xmath93 .
we do not modify @xmath116 .
observe that @xmath93 is alive for @xmath116 and stays alive for any string in @xmath64 with suffix @xmath116 .
* case 2 : * there is a string @xmath118 such that @xmath119 does not reach @xmath93
. the string @xmath119 does not reach @xmath93 and hence no string in @xmath120 has a computation beginning in the starting state @xmath92 and ending in @xmath93 .
we replace @xmath116 by @xmath119 and @xmath93 is dead for @xmath116 , but also dead for any string in @xmath121 with suffix @xmath116 .
also observe that any already processed state @xmath122 stays alive , resp .
remains dead .
in the second phase we proceed completely analogously , but now construct a string @xmath123 such that each state @xmath93 is either _ alive _ for @xmath124 ( i.e. , @xmath93 accepts all strings in @xmath125 ) or _ dead _ for @xmath124 ( i.e. , @xmath93 does not accept any string in @xmath126 ) .
now consider any string @xmath127 in @xmath128 .
observe that @xmath129 is a subset of @xmath45 .
however @xmath116 can not reach a dead state for @xmath116 and @xmath124 can not be accepted by a dead state for @xmath124 .
thus any accepting computation for @xmath130 has to utilize a transition @xmath85 between alive states @xmath68 for @xmath116 and @xmath69 for @xmath124 . but any pair @xmath100 of alive states is a critical pair and we are done . from now on
we fix a pair @xmath104 for which lemma [ onetwo ] holds .
let @xmath100 be an arbitrary critical pair for @xmath106 .
we now utilize that all strings in @xmath131 reach @xmath68 and all strings in @xmath103 are accepted by @xmath69 .
[ densit ] for all strings @xmath84 there are states @xmath132 , integers @xmath133 ( with @xmath134 ) and a string @xmath95 as well as computations @xmath135 we consider all strings of the form @xmath136 the string @xmath137 has suffix @xmath116 and hence @xmath137 reaches @xmath68 . as a consequence there is @xmath118 and a computation @xmath138 for @xmath139 which begins in the initial state @xmath92 and reaches @xmath68 . after reading @xmath140 ,
computation @xmath138 processes @xmath137 and produces a sequence of @xmath141 states , where we list all states before reading a copy of @xmath142 , resp . after reading the last copy
a state @xmath25 of @xmath0 appears twice in this sequence and we obtain a transition of the form @xmath143 for @xmath144 .
finally @xmath138 , starting in @xmath25 , reaches @xmath68 after reading the remaining @xmath145 copies . to establish ( [ one ] )
, we set @xmath146 and obtain transitions @xmath147 and @xmath148 . thus ( [ one ] ) follows .
part ( [ two ] ) is established by a similar argument , but now applied to @xmath149 .
this time we get transitions @xmath150 and @xmath151 .
but then @xmath152 as well as @xmath153 are transitions for any multiples @xmath154 and the claim follows , if we replace both @xmath155 and @xmath156 by @xmath157 .
let @xmath100 be a critical pair for @xmath109 .
we now introduce the detection problem for @xmath100 in which strings in @xmath40 have to be `` weakly '' separates from strings in @xmath83 .
it turns out that any nfa @xmath0 for @xmath45 solves the detection problems for all critical pairs , provided @xmath0 has ambiguity @xmath158 .
since we show later that @xmath0 can be efficiently simulated by a communication protocol with communication resources related to the number of states and that the detection problem is hard for communication complexity , @xmath0 must have many states .
the detection problem of @xmath100 has the following form : * accept a string @xmath159 iff there is a computation @xmath85 of @xmath0 .
remember that for no @xmath160 there is a computation @xmath161 with the initial state @xmath92 , a final state @xmath108 and strings @xmath162 .
hence no string @xmath84 is accepted . *
reject a string @xmath159 iff there are states @xmath163 , integers @xmath164 ( with @xmath165 ) and strings @xmath166 with computations @xmath167 ( the computations ( [ xx1 ] ) and ( [ xx2 ] ) will be used later to define a launching and storage cycle respectively .
it turns out that @xmath81 is placed within matching positions of the @xmath73- and @xmath168-cycle and hence @xmath81 plays the role of an impostor of @xmath79 . )
* @xmath159 is left undecided iff @xmath81 is neither accepted nor rejected .
to explain the purpose of these transitions consider the string @xmath169 \cdot [ zu \cdot ( z'u)^{a-1}].\end{aligned}\ ] ] if we process the first half @xmath170 of @xmath171 starting in state @xmath172 , then there is a computation @xmath173 of the form @xmath174 as well as a computation @xmath175 from @xmath172 to @xmath68 according to ( [ xx1 ] ) .
when reading the second half of @xmath171 , computation @xmath173 splits into a computation @xmath176 which goes full circle reaching state @xmath172 again and a computation @xmath177 which reaches @xmath68 after completely reading @xmath171 .
now assume that there is a transition @xmath82 .
computation @xmath175 has reached @xmath68 after reading the first half of @xmath171 and now reads the second half @xmath178 of @xmath171 .
it travels from @xmath68 to @xmath69 and subsequently reaches state @xmath179 , if _ additionally _ the string @xmath81 is read .
we have been successful * in `` storing '' a mother computation via computation @xmath176 in state @xmath172 , * preparing for a new `` launch '' in state @xmath68 via computation @xmath177 and * storing offspring computations in state @xmath180 via computation @xmath175 .
we utilize properties ( 1)-(3 ) by defining a sequence @xmath181 with many computations , namely we set @xmath182 = s_m \cdot [ z \cdot ( uz')^{a-1 } \cdot u].\end{aligned}\ ] ] assume inductively that there are computations for @xmath183 which have reached the states @xmath172 and @xmath68 respectively and a computation for @xmath184 which has reached @xmath180 . after reading the suffix @xmath170 of @xmath185 , the computation starting in @xmath172
has split into a computation reaching @xmath172 again and a computation reaching @xmath68 , whereas the freshly launched computation reaches @xmath180 from @xmath68 after reading @xmath186 .
observe that all previously launched computations go full circle after reading @xmath187 and again have reached state @xmath180 . as a consequence ,
there are @xmath188 distinct computations for @xmath189 all reaching state @xmath180 at the same time .
we say that @xmath0 has no redundant states , if each state is part of some accepting computation of @xmath0 . which strings are rejected and which strings are accepted ? [ 1 ] let @xmath0 be an nfa recognizing @xmath45 without redundant states .
also assume that @xmath0 has ambiguity @xmath158 .
* consider the detection problem of an arbitrary critical pair @xmath100 .
then all strings in @xmath83 are rejected and no string in @xmath190 is simultaneously accepted and rejected . *
each string in @xmath40 is accepted in the detection problem of some critical pair .
\(a ) we observe first that every string @xmath84 is rejected .
we may choose @xmath191 and the transitions required in ( [ xx1 ] ) and ( [ xx2 ] ) exist as a consequence of lemma [ densit ] : the states @xmath192 and @xmath193 belong to the @xmath25-cycle and the @xmath130-cycle respectively .
now assume that there is a string @xmath159 which is accepted and rejected .
since @xmath81 is accepted , there is a computation @xmath82 . also , since @xmath81 is rejected , there are computations of the form ( [ xx1 ] ) and ( [ xx2 ] )
. thus we may construct the strings @xmath194 for every @xmath188 and obtain @xmath188 distinct computations which , starting from state @xmath172 , reach state @xmath180 at the same time .
but @xmath0 does not have redundant states and each state , and in particular state @xmath172 , is reachable from the initial state .
also each state , and in particular state @xmath180 , can reach an accepting state .
thus there are strings @xmath195 such that @xmath196 has @xmath188 accepting computations . but @xmath194 is a string with length linear in @xmath188 and hence @xmath0 has at least linear ambiguity .
\(b ) follows from part ( a ) , if we apply lemma [ onetwo ] .
we show that the detection problem has an efficient communication protocol , provided a small nfa @xmath0 with ambiguity @xmath158 recognizes @xmath45 .
remember that @xmath197 and @xmath61 .
we work with the conventional two - party communication model consisting of two players alice and bob .
if @xmath198 is the input of @xmath0 , then alice receives @xmath30 and bob receives @xmath31 as their respective inputs .
alice and bob communicate nondeterministically with computations either being accepting , rejecting or undecided .
we say that an input is accepted if at least one computation is accepting , rejecting if at least one computation is rejecting and undecided if all computations are undecided .
( thus undecided computations play the role of rejecting computations for conventional nondeterminism . )
observe that we allow to simultaneously accept and reject an input .
now assume that the nfa @xmath0 recognizes @xmath45 .
let @xmath199 be two states of @xmath0 and let @xmath159 be an input string .
our first goal is to determine whether @xmath0 has a computation for @xmath81 starting in @xmath122 and ending in @xmath200 .
set @xmath201 .
beginning with @xmath202 , alice simulates @xmath0 for input @xmath203 by starting in state @xmath204 and sends state @xmath205 , if @xmath206 is reached .
bob simulates @xmath0 for input @xmath207 by starting in state @xmath206 and sends state @xmath208 , if @xmath209 is reached . in the last round bob
accepts if additionally @xmath210 holds and otherwise outputs `` undecided '' .
obviously the simulating protocol exchanges at most @xmath211 messages .
it has an accepting computation iff @xmath0 has a computation @xmath212 and otherwise leaves the input undecided .
we say that a protocol solves the detection problem of @xmath100 if the protocol labels each input as accepted , rejected or undecided as prescribed by the detection problem .
[ com ] assume that @xmath0 recognizes @xmath45 and that @xmath0 has ambiguity @xmath158 .
let @xmath100 be a critical pair for @xmath109 .
then there is a nondeterministic protocol @xmath213 which solves the detection problem of @xmath100 with @xmath214 messages .
we begin by describing the protocol @xmath213 . in its first attempt
@xmath213 tries to accept its input @xmath159 by simulating the automaton @xmath0 when reading @xmath81 starting in state @xmath68 . @xmath213
accepts @xmath81 iff state @xmath69 is reached and otherwise leaves @xmath81 undecided . in its second attempt
@xmath213 tries to reject @xmath81 .
alice guesses states @xmath163 as well as strings @xmath215 and integers @xmath216 ( with @xmath134 ) .
then alice verifies the following transitions _ without _ communication , namely * @xmath217 as well as @xmath218 and * @xmath219 as well as @xmath220 . in order to check the remaining transition @xmath221 and @xmath222 , alice guesses additional states @xmath223 and verifies the transitions @xmath224 and @xmath225 by herself .
subsequently alice communicates the states @xmath226 as well as @xmath227 and both alice and bob simulate the automaton @xmath0 on input @xmath81 for starting states @xmath172 and @xmath228 .
bob rejects iff the transitions @xmath229 and @xmath230 have been verified and otherwise labels @xmath81 as undecided .
observe that @xmath213 exchanges at most @xmath214 messages , since @xmath213 uses messages only when simulating @xmath0 on the string @xmath159 .
in lemma [ com ] we have solved the detection problem of a critical pair by a nondeterministic protocol @xmath213 with only @xmath214 messages .
however the detection problem separates @xmath40 from its complement @xmath83 only weakly , since the majority of strings from @xmath40 are either rejected or left undecided .
we begin our analysis by transforming the nondeterministic protocol @xmath213 into a deterministic protocol @xmath35 .
we avoid an exponential blowup in the number of messages by observing the structural limitations of @xmath213 .
in particular , @xmath213 accepts a subset @xmath231 of @xmath232 and rejects a superset @xmath233 of @xmath83 , where @xmath234 and @xmath233 are disjoint .
[ det ] there is a deterministic protocol @xmath35 which accepts at least @xmath67 strings from @xmath40 and rejects all strings from @xmath83 .
no string is left undecided and no string is accepted as well as rejected . moreover , at most @xmath235 messages are exchanged .
we begin by fixing a critical pair @xmath100 such that at least @xmath67 strings are accepted in the detection problem of @xmath100 .
observe that such a critical pair exists as a consequence of lemma [ 1 ] ( b ) , since each string in @xmath40 is accepted in the detection problem of at least one critical pair and there are at most @xmath236 critical pairs .
let @xmath231 be the subset of @xmath40 which is accepted in the detection problem of @xmath100 and let @xmath233 be the superset of @xmath83 of rejected strings . according to lemma [ com ] there is a nondeterministic protocol @xmath213 which solves the detection problem of @xmath100 with at most @xmath214 messages .
thus there are conventional nondeterministic protocols @xmath237 for @xmath234 and @xmath238 for @xmath233 which exchange at most @xmath214 messages each . to obtain a deterministic protocol @xmath35 from @xmath237 and @xmath238
we utilize that deterministic protocols with @xmath239 messages can be built from nondeterministic protocols , provided the protocols recognize a language _ and _ its complement by exchanging at most @xmath129 messages @xcite .
our situation however is more complicated , since @xmath234 is only a subset of the complement of @xmath240 .
we employ the construction in @xcite with the following modifications .
define the communication matrix @xmath138 of @xmath241 by setting @xmath242 = \left\ { \begin{array } { ll } 1 & \;\ ; x_1y_1 \cdots x_ty_t \in l_{\rm{yes } } , \\ 0 & \;\ ; x_1y_1 \cdots x_ty_t \in l_{\rm{no } } \\ { \rm{undecided } } & \;\;{\rm{otherwise . } } \end{array } \right.\ ] ] each message @xmath188 corresponds to a submatrix @xmath129 of @xmath138 defined by the collection of rows for which the message is sent and the collection of columns for which it is accepted .
now let @xmath129 be a submatrix of the communication matrix @xmath138 .
we define @xmath243 to be the maximal size of a submatrix @xmath244 of @xmath129 ,
where @xmath244 , after a suitable permutation of rows and columns of @xmath129 , is a lower triangular matrix with ones on the diagonal and zeroes above the diagonal .
( observe that @xmath244 may contain undecided entries , but these entries have to appear below the diagonal . )
since @xmath234 is accepted by the nondeterministic protocol @xmath237 and since no two diagonal entries can be accepted by the same message , we obtain that @xmath245 is bounded by the number of messages of @xmath237 and hence @xmath246 follows .
we first try to reject the given input by deterministically selecting a sequence @xmath247 of messages from the protocol @xmath238 . as for the conventional transformation to deterministic protocols
, the triangular message complexity will be halved in each step and in particular @xmath248 follows .
we proceed as in the conventional transformation and stop the communication prematurely , if the output `` no '' can be excluded and output `` yes '' . otherwise , after at most @xmath249 rounds
, we obtain @xmath250 . as a consequence
, the submatrix @xmath251 has no triangular submatrix of size two or larger .
in particular , the submatrix @xmath129 of @xmath252 spanned by all rows and columns of @xmath253 with a one , contains all ones of @xmath252 , no zeroes and possibly undecided entries .
if the joint input belongs to @xmath129 , then we stop and accept , resp .
stop and reject otherwise . in each round only messages of @xmath238 and hence at most @xmath214 messages are exchanged .
thus overall at most @xmath254^ { \log_2 \delta_{\rm{yes}}(c ) } = |q|^{o(t^2 \cdot \log_2|q|)}$ ] messages are generated .
remember that @xmath255 , where @xmath62 is the language of non - disjointness for @xmath25-element subsets of @xmath256 .
let @xmath35 be a deterministic protocol which accepts only strings in @xmath40 .
also let @xmath39 be a sufficiently small positive constant .
we apply fact [ fact ] and obtain that @xmath35 accepts at most @xmath257 strings from @xmath40 , provided at most @xmath37 messages are exchanged .
now , if an nfa @xmath0 with sublinear ambiguity recognizes @xmath258 , then we apply lemma [ det ] to obtain a deterministic protocol which exchanges at most @xmath259 messages , accepts at least @xmath260 strings and accepts only strings from @xmath40 .
thus , if @xmath261 for a sufficiently small positive constant @xmath39 , then at most @xmath262 inputs from @xmath40 are accepted .
but the nondeterministic protocol accepts at least @xmath67 strings from @xmath40 and hence @xmath263 follows .
we set @xmath264 .
let @xmath265 be a sufficiently small positive constant .
now either @xmath266 and we are done , since then @xmath267 or @xmath268 holds .
in the latter case @xmath269 and the upper bound on the number of messages in fact [ fact ] is met , provided @xmath265 is sufficiently small .
but then @xmath270 follows from ( [ q ] ) and hence @xmath271 holds for some positive constant @xmath272 .
we obtain @xmath273 and hence @xmath274 , since @xmath264 .
we get a contradiction if @xmath265 is chosen sufficiently small and we have shown [ final ] let @xmath0 be an nfa with sublinear ambiguity recognizing @xmath59
. then @xmath0 has at least @xmath275 states .
let @xmath276 be arbitrary and let @xmath0 be an nfa for @xmath277 .
we again follow the strategy for @xmath58 , however the transition from nfa s to communication protocols is now more involved .
for @xmath278 we have to work with vectors @xmath279 of states and besides reachabilty for @xmath68 and acceptance by @xmath280 we also have to guarantee that computation paths exist between @xmath281 and @xmath88 .
this last requirement requires some further work . *
all strings in @xmath131 reach @xmath68 and all strings in @xmath103 are accepted by @xmath280 * and for all strings @xmath284 and for all @xmath285 ( @xmath286 ) there is a string @xmath287 such that a computation for @xmath288 starts in @xmath88 and ends in @xmath89 .
we construct @xmath116 as in lemma [ onetwo ] and hence for any state @xmath93 of the nfa @xmath0 either all strings in @xmath131 reach @xmath93 or no such string reaches @xmath93 . to construct @xmath124
we first run the procedure of lemma [ onetwo ] and property ( 1 ) is satisfied .
then we process all pairs @xmath289 of states of @xmath0 in some arbitrary order .
if for all strings @xmath71 there is a string @xmath290 such that @xmath291 has a computation beginning in @xmath93 and ending in @xmath122 , then we say that the pair @xmath289 is `` alive '' and @xmath124 is left unchanged .
otherwise there is a string @xmath71 such that no computation for a string in @xmath292 has a computation beginning in @xmath93 and ending in @xmath122 .
we replace @xmath124 by @xmath293 .
the pair @xmath289 is now `` dead '' , since no string in @xmath294 has a computation beginning in @xmath93 and ending in @xmath122 .
also observe that processed pairs do not change their status , i.e. , remain dead , resp . stay alive after updating @xmath124 .
we have generalized lemma [ onetwo ] . for @xmath58 lemma [ densit ]
establishes that a string @xmath86 `` lives '' in a launching cycle for @xmath68 and a storage cycle for @xmath69 .
its generalization requires more work .
let @xmath297 be a critical vector and let @xmath84 be an arbitrary string .
we construct a string @xmath284 for @xmath81 so that some string with period @xmath298 can be launched by @xmath68 , stored and launched in between @xmath299 and @xmath89 and finally stored by @xmath280 .
in particular , we say that a string @xmath284 is appropriate for @xmath81 if the following properties are satisfied : * @xmath300 reaches @xmath68 . * for every @xmath285 , @xmath286 , there is a string @xmath301 and computations @xmath302 .
moreover , @xmath301 starts with a suffix of @xmath298 containing @xmath124 as prefix , followed by @xmath303 and completed by a prefix of @xmath304 . *
state @xmath280 accepts any string @xmath305 which consists of a suffix of @xmath298 containing @xmath124 as prefix , followed by @xmath303 . *
the string @xmath306 has periods @xmath307 and @xmath298 respectively .
now assume that @xmath304 is appropriate for @xmath81 .
we show that the string @xmath308 `` lives '' in appropriate cycles for each @xmath309 .
first observe that @xmath86 has period @xmath307 and hence also period @xmath298 .
the proof of lemma [ densit ] shows that a launching cycle @xmath310 is established , once @xmath300 reaches @xmath68 . also , intermediate cycles in between @xmath299 and @xmath89 exist , since @xmath301 has substring @xmath303 , and a final storage cycle following @xmath280 exists , since @xmath280 accepts a string with suffix @xmath303 .
let @xmath311 be some arbitrary ordering of the states of @xmath0 .
each pair @xmath312 influences the construction of @xmath304 .
assume for the moment that strings @xmath313 are already defined .
we set @xmath314 for all @xmath315 ( @xmath316 ) .
observe that @xmath317 , if @xmath318 is alive , and that @xmath124 is a prefix of @xmath319 .
choose the strings @xmath320 so that there is a computation for @xmath319 from @xmath321 to @xmath89 .
such strings @xmath313 exist with property ( 2 ) of a critical vector , since @xmath124 is a prefix of @xmath319 and @xmath322 is alive .
finally set @xmath323 we show that @xmath304 is appropriate for @xmath81 by first verifying property ( 1 ) .
the string @xmath304 has suffix @xmath116 and hence , by property ( 1 ) of a critical vector , @xmath300 reaches @xmath68 , the first component of the critical vector @xmath324 . observe that each @xmath319 has prefix @xmath124 and hence @xmath325 and @xmath304 have @xmath124 as prefix .
we start the verification of properties ( 2 ) and ( 3 ) by defining @xmath326 and constructing a computation @xmath327 . since @xmath124 is a prefix of @xmath304 ,
there is a computation for @xmath303 which leads from @xmath69 to a state @xmath321 such that the pair @xmath328 is alive .
but then , by definition of @xmath329 , there is a computation for @xmath330 which starts in @xmath69 , reaches @xmath321 after reading @xmath303 and ends in @xmath331 after reading @xmath329 .
we set @xmath332 . by construction
, @xmath329 is a prefix of @xmath333 which itself is a prefix of @xmath304 .
thus there is a string @xmath334 with @xmath335 and @xmath334 has prefix @xmath124 .
we now construct a string @xmath336 and a computation @xmath337 as follows .
since @xmath334 has prefix @xmath124 there is a computation for @xmath338 which reaches a state @xmath339 when starting in state @xmath340 . since the pair @xmath341 is alive
, we obtain the computation @xmath342 and set @xmath343 .
the construction of @xmath301 and verifying a computation @xmath302 for arbitrary @xmath344 proceeds in a completely analogous fashion . finally , again by property ( 1 ) of a critical vector , state @xmath280 accepts any string @xmath305 consisting of a suffix of @xmath298 followed by @xmath300 , since the suffix of @xmath298 has prefix @xmath124 .
the remainder of the argument proceeds completely analogous to the case of @xmath58 .
lemma [ 1 ] shows that an nfa with sublinear ambiguity solves the detection problem for @xmath58 .
to introduce its generalization we firstly introduce the detection problem for @xmath345 : @xmath81 has to be rejected iff there is a string @xmath70 such that @xmath81 , acting as an impostor of @xmath79 , can be placed in matching positions within the @xmath346 individual @xmath168-cycles of @xmath0 .
lemma [ 1 ] was a direct consequence of lemma [ densit ] in the case of @xmath58 .
in the same manner we can now show that an nfa with ambiguity @xmath7 solves the detection problem for general @xmath20 as a direct consequence of lemma [ twok ] .
let @xmath0 be an nfa with ambiguity @xmath7 for @xmath277 .
as in lemma [ com ] we simulate @xmath0 to obtain a nondeterministic protocol @xmath213 solving the detection problem with @xmath347 messages ; the exponent grows by the factor @xmath20 , since @xmath346 instead of two computations of @xmath0 on input @xmath81 have to be simulated .
we transform @xmath213 into a deterministic protocol @xmath35 with @xmath348 messages as in lemma [ det ] . to complete the proof of theorem [ main ] ,
we replace @xmath25 by @xmath349 in the proof of lemma [ final ] ( to compensate for the increase in the number of messages of @xmath35 from @xmath350 to @xmath351 ) and obtain | the ambiguity of a nondeterministic finite automaton ( nfa ) @xmath0 for input size @xmath1 is the maximal number of accepting computations of @xmath0 for an input of size @xmath1 . for all @xmath2
we construct languages @xmath3 which can be recognized by nfa s with size @xmath4poly@xmath5 and ambiguity @xmath6 , but @xmath3 has only nfa s with exponential size , if ambiguity @xmath7 is required . in particular , a hierarchy for polynomial ambiguity is obtained , solving a long standing open problem ( ravikumar and ibarra , 1989 , leung , 1998 ) .
juraj hromkovi georg schnitger |
in the previous eso workshop dedicated to the _ production and distribution of c , n , o elements _ held in garching in may 1985 the information on cno elements coming from absorption systems along distant qso were just at the starting point and the first tentative results were reviewed by max pettini .
we had to wait for one decade for the first n measurement in a damped ly@xmath0 system ( dla ) , even longer for o , and the time for c has probably not arrived yet .
dlas are intervening absorption systems with large hydrogen column densities particularly appropriate for the determination of accurate chemical abundances .
typical abundances are @xmath12.5@xmath7 [ fe / h]@xmath7 @xmath11 with a moderate chemical evolution .
the increase in metallicity is of the order of 0.3 dex per unit of decreasing redshift and suggests that dlas do not trace the population of galaxies responsible for the bulk of star formation in the universe .
a floor of abundances at about -2.5 is observed at all redshifts up to z=5.28 ( songaila et al 2002 ) quite suggestive of a prompt enrichment at very early phases . at redshif 5
the age of the universe is about one billion years and there is hope to find young systems where we can better recognize the imprints of the first generation of stars .
the observed abundances are gas phase abundances and in general suffer from possible dust and ionization corrections .
the large neutral hydrogen column densities of the dlas shield the clouds from the igm field effectively .
new evidence comes from the recently measured ar which is particularly sensitive to ionization effects .
the [ ar / si ] is often very close to solar or mildly deficient , which rules out significant ionization ( molaro et al 2001 , vladilo et al 2003 ) .
dust is definitely present in the dlas as revealed by a number of indications such as the systematic deficiency of fe compared with zn , the anti - correlation between refractory elements and the elemental column density , the correlations between [ fe / zn ] and h@xmath8 and the correlation between [ si / fe ] and [ zn / fe ] ( molaro 2001 and references therein ) .
cno elements are all mildly refractory and are virtually free from this complication ( savage & sembach 1996 ) .
the determination of the abundances of the cno elements have proven to be very challenging .
this because the most accessible lines for cno are either saturated , despite the low metallicities of the dlas , or they fall in spectral regions strongly contaminated by the neutral h absorptions of the ly@xmath0 forest .
the oi 1302 line is generally nicely placed redwards the ly @xmath0 forest but is always found strongly saturated . conversely the oi 1355 has never been detected and provides , together with the previous line , high and low bounds to the oxygen abundance .
a similar situation holds for c where the only suitable lines are the cii resonance lines at 1334.53 and cii 1036.33 which are heavily saturated in all dlas .
a more favourable circumstance is found for ni which has several multiplets offering a relevant dynamical range in the line strenghts
. however , the more accessible ni multiplets are @xmath9 1134.16 , 1134.42 , 1134.98 and 1199.55 , 1200.22 , 1200.71 which fall in the ly@xmath0 forest .
first measurements of o were performed at last through observations of oi 925 and oi 950 in the ly-@xmath0 forest , associated with the z@xmath10=3.39 dla towards qso 0000 - 2621 ( molaro et al 2000 ) .
partial contamination by hydrogen clouds is taken into account using a model for the absorber constructed on lines falling redwards the ly @xmath0 forest .
this approach has been followed in the dla at z@xmath10=4.49 towards the j 0307 - 4945 ( dessaugues - zavadsky et al ( 2001 ) , and towards qso 0347 - 381 through the measurement of a dozen oi lines ( levshakov et al 2002 ) .
probably the best available measure due to the simplicity of the absorber and to the quality of keck data has been realized by prochaska et al ( 2001 ) in the dla at z=2.844 towards qso 1946 + 76 were several oi lines have been detected allowing an oi measure of log(oi)=14.819 @xmath2 0.007 .
pettini et al ( 2002 ) targeted few systems with low elemental abundances and low hi column density and succeded in measuring o from an almost unsaturated oi 1302 in a couple of dlas .
oi measurements have been provided for other dlas towards qso 1202 - 0725 ( dodorico et al 2003 ) qso 2059 - 360 ( dessaugues - zavadsky et al 2003 ) , qso 2243 - 6031 , qso 1104 - 1805 ( lopez et al 1999 , lopez et al 2002 ) .
a total of a dozen of systems with some information on the o abundance is presently available .
the [ o / zn ] and [ o / fe ] are shown in fig 1 with filled and empty circles respectively , and the solar o abundance taken from the recent revision by holweger ( 2001 ) .
o and zn do not require any dust correction and two [ o / zn ] reveal a moderate enhancement in @xmath0-elements at the level of 0.2 dex or so , while in one case the ratio is of @xmath4 0.6 dex .
there are also few [ o / fe]@xmath4 0.2 which give stringent values .
most likely these systems are dust - free , as qso 0000 - 2621 where [ o / zn]@xmath4[o / fe ] .
if dust is present the effect is to increase the intrinsic fe abundance and further lower the [ o / fe ] ratio .
overall there is a low overabundance of o relative to zn , which is consistent with previous indications obtained from [ s / zn ] measurements or from [ si / fe ] once expected dust depletion is accounted for ( centurin et al 2000 , vladilo 2001 ) . in the sample
there are two dlas with o and s measured .
the [ o / s ] ratio is @xmath4 0 showing that there is no particular behaviour of o with respect to the other @xmath0 elements .
si is generally available but it is a refractory element and may be partially depleted onto dust grains . in all but two dlas
the [ o / si ] is found @xmath4 0 , which implies that si is only moderately locked up into dust grains in dlas .
if this is the case it is rather interesting to consider the behaviour of si with comparison of the undepleted element zn .
the average value of the 26 dlas with both elements measured is @xmath7[si / zn]@xmath11=-0.07 @xmath20.2 , providing additional evidence of a moderate , if any , enhancement of @xmath0-elements .
overall the low @xmath0 over iron - peak element ratio appears as a prominent characteristic of the dla .
cases of genuinely @xmath0 enhancement exist such as the dla towards j0347 - 3819 but they look as an exception rather than the rule ( levshakov et al 2002 ) .
in the survey by prochaska et al 2001 there are 14 lower limits of c abundances , all obtained from the saturated cii 1335 line .
ci is observed in a few cases but it traces very cool gas and being strongly ionised is not very informative of the total c. levshakov et al ( 2002 ) and lopez et al ( 2002 ) through detailed modelling of all absorption features in the effort to account for the saturation of the cii lines , claimed measurements with a reasonable error . in these two dlas
zn is also measured and for both [ c / zn]@xmath4 0.25 , which suggests a rather _ normal _ behaviour .
highly supersolar [ si / c ] values at the level of 0.5 - 1.5 dex are predicted by popiii sne , hypernovae or pair - instability sne of very massive stars ( umeda and nomoto 2001 ) for which there is no evidence in dlas where [ si / c ] is solar .
better prospects to measure c rely on the observations of very high redshift dlas if abundances finally decrease , or in the class of sub - dlas ( cfr dessauges - zavadski , these proceedings ) . in the z@xmath10=4.383 dla towards br 1202 - 0725 the cii line is found only moderately saturated and [ c / fe]=-0.03 ( @xmath20.15 ) and [ si / c]=0.15 @xmath20.11 are derived ( dodorico et al 2003 ) . in the z@xmath10=5.8 dla towards sdss 1044 - 0125 identified by songaila and cowie ( 2002 ) , in which @xmath12 n(hi)= 20.5 @xmath13 and [ fe / h][email protected] , the cii 1334 transition is unsaturated providing a [ c / fe]=-0.6 , which suggests a possible drop in the c abundance at very high redshift . at redshift about 6
we may be within 100 million years from the epoch of star formation , which puts tight constraints to the c build - up processes ( molaro et al 2003 ) .
among the cno triplet n was the first to be observed . in their first attempt pettini et al ( 1995 ) failed in detecting the element in the dla towards qso 2348 - 146 but established a very low upper limit at [ n / h]@[email protected] , which has been slightly improved with keck observations to [ n / h]@[email protected] ( prochaska and wolfe 1999 ) .
this failure was followed by the detection of n in the dla at z@xmath10=3.39 towards qso 0000 - 2621 and towards qso 1331 + 17 with the n abundances [ n / h]=-2.77 @xmath2 0.17 , and [ n / h]= -2.73 , respectively ( molaro et al 1996 , kulkarni et al 1996 ) .
the three dlas showed a difference by @xmath4 0.6 dex in their [ n / si ] ratios and introduced since the beginning the notion of significant scatter in the [ n / si ] abundances .
it is just amazing that this became evident with the first two systems and that the first one to be observed was a nitrogen poor system .
the early findings were confirmed and extended on larger samples , of which the most important for the number of objects are those of centurin et al ( 1998 ) , lu et al ( 1998 ) , pettini et al ( 2002 ) and prochaska et al ( 2002 ) . in the following we use the complete and updated compilation of centurin et al ( 2002 , table 9 ) who contributed with some new n observations as well .
the sample of n observations or limits now comprises 33 systems , out of which 28 have at least one of o , si or s measured and provide informative ratios . in fig 2 the [ n/@xmath0 ] are plotted versus [ @xmath0/h ] . for the @xmath0 element
either of o , s or si are taken in the order .
as discussed in the previous section they trace each other rather well so that it does not matter very much which @xmath0 element is used .
the data for the bcg are also shown in the figure with smaller symbols .
the data show a peculiar pattern .
most of the sample form a plateau at [ n / si]@xmath4 @xmath10.82 ( @xmath2 0.13 ) , while few points show lower n / o ratios .
the possibility of a gap between low and high values rather than a pure scatter has been pointed out by prochaska et al ( 2002 ) who suggested a bimodal distribution . to the two measurements and to one upper limit with low n / o considered originally by prochaska et al we add the value towards j0307 - 4945 .. dessauges - zavadsky ( 2001 ) , rejected by prochaska et al ( 2002 ) for a possible ionization effect , and the most recent one towards qso 2059 - 360 from dessauges - zavadsky et al ( 2002 ) .
two new upper limits quite separate from the high values and consistent with a bimodal distribution are provided by pettini et al ( 2002 ) .
the average of the four determinations is [ n / si]@xmath4 @xmath11.45(@xmath2 0.05 ) , strongly suggestive of a second plateau even with a narrower dispersion .
the low n / o values are mostly observed at the lowest metallicities , but with an overlap over the higher plateau in the range -2.0 @xmath7[o / h]@xmath7 -1.5 . in fig 3 the plot of [ n / h ] versus [ si / h ]
clearly shows the primary - like behaviour of n in both plateau .
the high n / o values look as an extension of the bcg towards lower metallicities and give definitive evidence of a plateau , which was not so clear from the bcg only .
the sharing of the same n / o between dlas and bcg suggest that the two populations experienced similar chemical evolution .
n synthesis is thought to occur in intermediate mass stars ( ims ) ( 4 to 7 m@xmath6 ) through hot bottom burning in the agb phase .
the nucleosynthesis by ims implies a delay in the n production when compared to that of o as well as a sensitivity of n production to the low mass regime of the imf .
the general interpretation of the observed data is that n is primary at low metallicities and secondary at high metallicity ( henry et al 2000 ) .
the henry et al model is able to reproduce the plateau by reducing the sfr and thus by preventing the rapid o build up by massive stars . for high sfr
the elemental build up is fast and the system enters very soon in the secondary production regime without undergoing a primary behaviour at all .
thus the presence of an extended plateau in the n / o suggests low sfr in the dla , which may also explain the failure of imaging dla galaxies .
the _ lower _ plateau is a new feature of dla with no other counterparts .
izotov et al ( 1999 ) argued that the low n / si were resulting from an inflated si column density by additional si formed in the ionized region .
however , the recent direct o measurements provided five genuine low n / o obtained by means of truly oxygen observations , shown in fig 2 , which rule out such possibility .
low n / o values have been interpreted in a context of a delayed n released model as the result of a relatively young age ( molaro 2001 , pettini 2002 , ) or as being the system observed close to a recent episode of star formation ( lu et al 1998 , centurin et al 1998 ) . according to henry et al the necessary scale time to attain a significant n production by ims is of at least 250 myrs .
it is therefore possible that dlas are catched in the early formation times when the ims had not yet expelled their n. however , as prochaska et al pointed out one should expect to find values everywhere below the plateau and it is difficult to account for the fact that the low n / o show a characteristic value at [ n / o][email protected] . in the low n / o plateau
the scatter is so small that these systems are unlikely in a transient phase towards the upper plateau . a second possibility to explain the low n / o values is a variation in the imf .
a top - heavy imf or an imf truncated below a mass threshold of @xmath4 5 m@xmath6 which exclude low mass stars has been suggested by prochaska et al ( 2002 ) .
however , it remains unclear what produces the change in the imf .
as discussed by henry at this meeting it is rather appealing to think that it is related to the formation of popiii stars a further possibility is that we are dealing with variable n yields as a function of the metallicity since the low n / o values occur preferably at the lowest metallicities .
this may affect the efficiency of the 3rd dredge - up or of the mass - loss rate along the agb phase somewhat altering the characteristic n yields of the ims . however , it is not clear how to account for the fact that both high and low n / o values occur at equal metallicities .
fresh clues for the interpretation of the low n / o plateau can be obtained plotting the [ n / si ] values with respect to [ n / h ] as shown in fig 4 .
the distribution does not show a degeneracy with the metallicity any longer but reveal two different regimes with a transition occurring at [ n / h]@xmath4 @xmath13 .
the low n / o are in the region with [ n / h ] @xmath7 -3 and the high n / o for [ n / h ] @xmath11 -3 . since there is no obvious link between n yields and the amount of n content in the systems we think that what we see is most likely a temporal effect . in fact , since the o abundance is more sensitive to the sfr history while n depends from ims evolutionary time - scales , n could be a better indicator of the age of system .
thus the two regimes reflect probably two epochs .
we propose that the higher plateau is produced by ims after they had time to evolve and release their n , while the lower plateau is produced by more massive , namely m@xmath11 8 m@xmath6 , and relatively younger stars . at later times ( or [
n / h]@xmath4 -3 ) the contribution from the massive stars is overwhelmed by that from the ims .
in fact a hint of connection between the two plateaux can be recognized in the two dlas which lie just below the primary plateau at [ n / si ] @xmath4 -1.1 and [ n / h]@xmath4 -2.8 as shown in fig 4 .
no objects are expected to be found below the low n / o plateau and therefore it would be rather crucial to improve the three upper limits to the levels at which they can probe the existence of the plateau .
we note that for the dla at z@xmath10=2.456 towards qso 1409 + 095 the upper limit is only 0.17 dex above the plateau and in fact there is a hint of the presence of the ni 1200 lines ( pettini et al 2002 , their fig 2 ) .
standard models do not predict n synthesis by massive stars but the situation is far from being settled ( heger and woosley 2002 ) .
chieffi et al ( 2001 ) found that also pop iii stars of intermediate mass do experience thermal pulses and the third dredge - up arguing that stars with primordial composition are sources of c and n. if we take the models of limongi and chieffi ( 2002 ) for zero metallicity , computed with franec code cf85 , with no mix of the h shell and a mass loss parameter of @xmath15=6 , and we integrate with a salpeter imf over the whole mass spectrum 4 - 80 m@xmath6 we get a [ n / o]=-0.85 , namely very close to the higher plateau . on the other hand
the integration in the mass range between 15 and 80 m@xmath6 gives [ n / o]=-1.5 , which is very close to the lower plateau .
the models for low metallicity stars with rotational mixing or overshooting of meynet and maeder ( 2002 ) predict n production for massive stars and can also reproduce the ratio observed in the lower plateau .
the observations of dlas might actually offer crucial guidelines for n production in massive stars .
the fraction of systems on the two plateaux and their relative abundances can be envisaged as two important tests to probe the origin of the two plateaux .
presenly the fraction of the systems forming the low n / o plateau is of 25% of the total .
we can compute the expected ratio assuming that we need roughly 250 myrs for significant n production and computing the age of the systems from the formation epoch to the observed redshift by means of a look- back time @xmath16= 0.538[(1+z)/10]@xmath17 gyrs .
the formation epoch is unknown but to get the observed statistics we need to place it at z@xmath4 4 , which is quite conceivable .
few systems with low n / o have z@xmath10 of about 2 - 3 suggesting a rather continuous formation of the dla .
the case of qso 1202 - 0725 with high n / o and z@xmath10=4.38 is rather remarkable and pushes back the epoch of star formation considerably .
if the synthesis of n observed by ims require about 0.25 - 0.5 gyrs then the time of star formation is shifted to a redshift greater than six . an @xmath0- element enhancement over iron - peak element is expected in the case that massive stars are responsible for the low n / o plateau .
fe is produced considerably by type ia snae and should lag o similarly to n. in fig .
5 the [ si / zn ] and [ si / fe ] are plotted versus [ n / si ] unfortunately there are no [ si / zn ] values for dlas in the low n / o plateau and the [ si / fe ] are between 0.2 and 0.4 which is rather common among the dlas . in absence of dust
these ratios would imply an @xmath0 enhancement , but unfortunately there is no zn information , and the test is not presently conclusive .
it is a pleasure to acknowledge my collaborators , and in particular miriam centurin for allowing me to present here preliminary results of common work .
marco limongi and alessandro chieffi are acknowledged for making the computation of n / o ratios with their model .
lopez s. , reimers d. , rauch m. , et al 1999 , apj 513 , 598 lopez , s. , reimers d. , dodorico s. , prochaska j. x. 2002 , a&a 385 , 778 lu l. , sargent w. l. w. , & barlow t. a. 1998 , aj , 115 , 55 meynet g. & maeder a. 2002 , a&a , 390 , 561 prochaska j.x . , henry r.b.c .
, omeara j.m .
2002 , pasp , 114 , 933 songaila , a. cowie l.l 2002 , aj 123 , 2183 savage b. d. sembach k. r. 1996 , ara&a 34 , 279 umeda h. , nomoto k. , & nakamura t. 2000 , in the first stars , eds .
a. weiss , t. abel , and v. hill , ( springer , heidelberg ) , 150 vladilo g. 2002a , apj , 569 , 295 vladilo g. , centurin m. , molaro m. , peroux c. , & dodorico , v. , 2003 , in prep . | damped ly@xmath0 clouds provide information on the chemistry of distant and metal poor regions of the universe .
in these clouds for observational difficulties n and o are among the last elements to be measured , and c has still to be measured with precision . in combining the extant cno abundances , we avail of a sample which includes 33 n measurements or significant limits , a dozen of o measurements , and only few tentative measurements for c. o is found to track both s and si rather closely without signatures of any peculiar behaviour and the few [ o / zn ] ratios available show a mild @xmath0/iron - peak enhancement which reinforces earlier evidence .
the few tentative c measurements point to roughly solar [ c / zn ] and [ c / si ] ratios .
n abundances are rather complex with n/@xmath0 ratios showing a bimodal distribution .
the majority of the values form a first plateau at [ n / si][email protected] ( @xmath2 0.13 ) and about 25 % form a second plateau at [ n / si]= @xmath11.45 ( @xmath2 0.05 ) .
the @xmath3 values are at the level of the blue compact dwarfs but they extend further towards lower metallicities .
the _ lower _ plateau is a new feature of dlas with no other counterparts .
the two plateaux are found for different values of the n abundance with the transition at [ n / h]@xmath4 -3.0 .
we argue that this results from the delayed release of n between massive and intermediate mass stars with the lower plateau produced by massive stars ( m@xmath5 8 m@xmath6 only , which accounts for the small dispersion .
the systems on the lower plateau are relatively young and imply a continuous formation of the dlas . while the interpretation of the n / o in the dlas needs to be confirmed , these observations could ultimately prove crucial for the interpretation of the early n nucleosynthesis .
# 1_#1 _ # 1_#1 _ = # 1 1.25 in .125 in .25 in |
eselevich m. and v. @xcite revealed that there is a disturbed region extended along the direction of coronal mass ejection ( cme ) propagation in front of the cme , when its velocity @xmath2 relative to the ambient coronal plasma is below a certain critical velocity @xmath3 . given @xmath4 , a discontinuity in the difference brightness distributions
is formed in the frontal part of the disturbed region . since @xmath3 is close to the local fast - mode mhd velocity the formation of such a discontinuity
may be associated with shock wave formation .
if we managed to resolve this discontinuity in space , determining its thickness @xmath5 ( by measuring the shock front profile , in essence ) , we could be able to clarify the dissipation mechanism in the shock front in the corona .
the purpose of this paper is : 1 ) to justify the correctness of @xmath5 measurements in the solar corona using mark 4 and lasco c2 ; 2 ) to discuss a possible dissipation mechanism in the shock front based on measurements of the shock wave thickness @xmath5 .
this analysis involved coronal images obtained by lasco c2 and c3 onboard the soho spacecraft @xcite , presented as difference brightness @xmath6 , where @xmath7 is undisturbed brightness at a moment @xmath8 , before the event considered ; @xmath9 is disturbed brightness at @xmath10 .
we used calibrated lasco images with the total brightness @xmath9 expressed in units of the mean solar brightness ( p@xmath11 ) .
the difference brightness images were employed to study the dynamics of the cme and its disturbed region . for this purpose , we used maps of @xmath12 isolines as well as sections along the sun s radius at fixed position angles @xmath13 and non - radial sections at various times . in the images ,
the position angle @xmath13 was counterclockwise from the sun s northern pole . at 1.2 r@xmath14 r@xmath1 , we employed polarization brightness images from the ground - based coronagraph - polarimeter mark 4 ( mauna loa solar observatory ) . as was the case with lasco data ,
these images were expressed in terms of difference brightness .
to identify a shock front in cme images is best done by tracing the process of its formation .
let us examine such a process on the example of cme 1 , which occurred at the w limb on 20 september 1997 at about 10:00 ut . in that event ,
the coronal ejection had a distinct three - part structure consisting of a frontal structure ( fs ) , cavity , and bright core .
figure 1 ( three top panels ) presents the difference brightness in the form of isolines for three subsequent instants of time corresponding to the cme motion .
the shape of the cme frontal structure is close to a circle ( dotted circle in figure 1 ) .
the process of shock wave formation can be seen in detail on difference brightness distributions @xmath15 plotted along the direction of the cme motion ( dashed line in the top panels of figure 1 ) .
each difference brightness distribution @xmath16 plotted at a given instant was ( see bottom panel in figure 1 ) : 1 .
normalized to a corresponding maximum value of difference brightness measured in the vicinity of the frontal structure ; 2 . displaced along the distance axis
@xmath17 in such a way that the frontal structure position on the axis coincided with its position at the moment of its first registration by lasco c2
. thus the coordinate system was tied with the frontal structure .
the disturbed region is almost absent at the initial instant ( solid circles ) in front of the frontal structure ( slanting hatching in the figure ) . at the next moment ( empty circles ) a compressed plasma region emerges in front of the fs , bounded by the shock in its frontal part ( crosshatching ) . the shock moves faster than the frontal structure and at subsequent moments of time ( diamonds and triangles )
is seen to be far ahead of the frontal structure .
conversely , the anterior boundary of the core lags behind the frontal structure , because of a lower velocity .
any density inhomogeneity in the magnetized coronal plasma can only be stationary thanks to magnetic field inhomogeneity , which is equivalent to the presence of current on the same scale .
the thickness of quiescent current sheet @xmath18 expanding due to diffusion can be estimated , via time @xmath19 , from the following relation @xcite : @xmath20 where @xmath21 is the mean time between electron - proton collisions ( here @xmath22 in degrees and @xmath23 in @xmath24 ) .
let us estimate @xmath18 for our conditions .
assuming , in accordance with @xcite , that , at @xmath25 2.1 r@xmath1 , plasma temperature @xmath26 k , magnetic field @xmath27 0.55 g , and density @xmath28 @xmath24 we obtain @xmath29 r@xmath1 and @xmath30 3 s. hence during cme propagation in the corona ( for several hours ) , the current sheet thickness does not exceed @xmath31 ( @xmath32 r@xmath1 ) which is much less than the spatial resolution of mark 4 and c2 ( @xmath33 0.02 r@xmath1 ) .
this means that the minimal measured current sheet thickness is , at best , close to the spatial resolution of the instruments in use .
however , the motion of the whole current sheet generates in front a disturbed region due to piled - up background plasma particles and to excitation of plasma density and magnetic field variations . in this case
, the image brightness jump corresponding to the current sheet will have an enhanced size due to the effect of the disturbed region .
the brightness jump in the shock front is also related to the density inhomogeneity on the scale of the front .
inherently , however , it differs significantly from the current sheet : deceleration and heating of the supersonic plasma stream occur in the shock front .
a disturbed region is absent ahead of the shock front as it moves at supersonic speed relative to the environment , and hence the shock front profile does not undergo distortions .
we determined the current sheet thickness at the frontal structure boundary as double the size of the brightness jump at half the jump height ( @xmath18 in the bottom panel in figure 1 ) . besides , it is possible to determine the current sheet thickness @xmath34 at the core boundary .
this value is shown in figure 1 only schematically , because the maximum core brightness , which , in reality , is much larger , was intentionally limited in the plot .
similarly , figure 1 illustrates the determination of the thickness of a shock front the brightness jump in which has a typical size @xmath5 . to understand how we can distinguish the current sheet from a shock wave ,
let us consider cme 2 that occurred at the w limb on 2 june 1998 at about 08:00 ut . in that event ( in contrast to the event of 20 september 1997 ) , the cme velocity was lower than the critical one , and no shock wave formation was observed , at least in the c2 field of view ( i.e. up to @xmath35 6 r@xmath1 ) .
the top panel in figure 2 shows the difference brightness at 10:05 ut for the event . a three - part structure ( fs , cavity and core )
may also be discerned in the event , but there was a rather extended disturbed region in front of the cme in the direction of its motion . approximating the shape of the frontal structure as a circle ( dots in the top panel of figure 2 ) enables us to determine how the current sheet size changes at the front boundary of the frontal structure in various directions . for this purpose , we plotted difference brightness sections from the frontal structure center .
these sections were used to estimate the current sheet size @xmath18 .
the position of each of these sections was specified by the angle @xmath36 drawn from the frontal structure center .
this angle was measured from the direction of cme propagation .
the change in this angle is positive counterclockwise . in the bottom panel of figure 2 ,
black circles indicate the dependence of @xmath18 on angle @xmath36 for cme 2 at 10:05 ut . there was a developed disturbed region ahead of the frontal structure ( the maximum distance of the frontal structure was @xmath35 3.5 r@xmath1 ) . as a result
, @xmath18 was nearly five times as large in the direction of the cme motion than in lateral directions ( @xmath37 ) ( figure 2 , solid circles ) . in cme 1 at 10:19 ut (
the maximum distance of the frontal structure from the solar center at this instant was @xmath35 2.2 r@xmath1 ) , the shock front was not observed yet , and @xmath18 could also be determined at various angles ( empty circles in the bottom panel in figure 2 ) .
the @xmath18 thickness is seen to be approximately constant ( 0.15 r@xmath1 ) , only increasing about twofold in the direction of cme motion ( @xmath38 ) .
hence the development of a disturbed region may result in increased apparent thickness @xmath18 .
the disturbed region is only slightly visible and @xmath18 is the smallest at large angles @xmath36 ( in lateral directions ) .
a similar angle @xmath36 dependence can also be plotted for the shock front thickness @xmath5 . in the bottom panel of figure 2
, crosses mark the @xmath39 plot for cme 1 at 11:09 ut , when the maximum distance of the shock front was @xmath35 4.8 r@xmath1 .
obviously , the behavior of @xmath39 differs from that of @xmath40 .
this difference may be due to the lack of a disturbed region ahead of the shock front . in that case
, the shock front thickness may depend on local parameters of the ambient plasma as well as on the velocity component along the normal to the shock front ( which decreases with increasing angle @xmath36 ) the shock wave type may change in that case .
what is the minimum thickness to be recorded by mark 4 and lasco c2 for the current sheet in the corona ?
evidently the effect of widening in the optically thin corona will be smallest for the current sheet whose size is smallest along the line of sight .
an example is the boundary of the erupting filament in the form of a thin loop , whose size is sufficiently small along the line of sight .
the top panel in figure 3 presents the difference polarization brightness from mark 4 data for the cme that commenced at the w limb on 28 june 2000 at about 19:00 ut . in that event one could observe a filament eruption easily discernible both in mark 4 and lasco c2 images
this event allows us to estimate and compare the mark 4 and c2 resolutions . for the purpose , we plotted difference brightness profiles across the filament loop along the dashed line passing through the filament center ( figure 3 , top panel ) .
difference brightness profiles normalized to the maximum brightness and shifted in such a manner that their maxima coincided ( bottom panel in figure 3 ) .
the plot shows that the spatial size of the brightness jump at the filament boundary remains constant , @xmath41 0.045 r@xmath1 , over the entire range @xmath25 1.3 r@xmath1 to @xmath25 6.1 r@xmath1 .
this value is close to the spatial resolution of mark 4 and c2 ( @xmath33 0.02 r@xmath1 ) .
the main contribution to the jump widening appears to be by the disturbed region .
thus it is possible to assume that the observable shock wave thickness ( @xmath42 0.2 r@xmath1 in the bottom panel of figure 2 ) reflects its real size , as there are no disturbances ahead of the shock front , while the measured shock wave thickness is essentially larger than the spatial resolution of the instrument .
notably , calculations for a simple geometric model of quasi - spherical shock @xcite show that the observable brightness profile width @xmath5 was close to width @xmath43 for the density jump in the shock wave .
the measured front size @xmath5 allows us to answer the question of whether the mechanism of energy dissipation in the shock wave is collisionless @xcite or collisional @xcite . if it is collisionless , energy dissipation in a shock wave is conditioned by collective processes in plasma due to developing instabilities . in this case , it is quasi - parallel shocks that have maximum front thickness in the magnetized plasma , that does not exceed @xmath44 where @xmath45 is the proton larmor radius calculated from undisturbed magnetic field directly ahead of the front @xcite . to estimate the larmor radius , we take the coronal magnetic field @xmath46 0.5 g and the proton velocity @xmath47 km / s for the fastest cme case .
we have : @xmath48 r@xmath49 . obviously , the collisionless shock wave front thickness is well below the resolution limit of modern coronagraphs .
measurements show , however , that the front thickness far exceeds this value .
this suggests that the dissipation mechanism in the shock wave is collisional in the corona . in this case , the shock wave energy dissipation is on the scale of order of the proton mean free path @xmath50 and hence it is also the scale that determines the shock wave thickness @xcite .
the proton mean free path expressed in solar radii is @xcite : @xmath51 where @xmath22 and @xmath23 are respectively the proton temperature ( degrees ) and density ( @xmath24 ) in the undisturbed plasma immediately ahead of the shock front .
compare the observable shock wave front thickness @xmath5 with @xmath50 in the corona .
the upper dashed curve in figure 4 is for @xmath50 calculated for the proton temperature and density measured by @xcite .
the lower dashed line is for @xmath50 calculated for the same density and the temperature half as high as the measurements ( @xmath52 k at 2 r@xmath1 ) , but suffering the same decay with distance .
these two dashed curves define roughly the lower and upper boundaries in estimating the free path depending on the chosen temperature .
the experimental dependence @xmath53 was plotted from the shock wave front thickness measured at different distances for eight cmes with velocities @xmath4 ( symbols in figure 4 ) .
the mean curve fitting these data is the thin solid line in figure 4 .
the characteristic front size is comparable with the free path ( @xmath54 ) , at least up to @xmath335 r@xmath1 .
this confirms the assumption that the dissipation mechanism in the shock wave may be collisional at these distances .
the condition @xmath54 is no valid at greater distances and the shock wave must become collisionless .
we did observe a gradual transition to formation of a collisionless shock front with thickness @xmath55 at @xmath56 10 r@xmath1 .
this will be dealt with in more detail in a future paper .
thus , we appear to encounter a rare situation where we can resolve and examine the collisional shock front structure in plasma .
the measured thickness of the shock wave , excited ahead of a cme , far exceeds the spatial resolution of mark 4 and lasco c2 coronagraphs . up to the distance
@xmath0 5 r@xmath1 from the center of the sun the thickness is of order of the proton mean free path .
this means that the energy dissipation mechanism in a shock front is collisional .
the work was supported by program no .
16 part 3 of the presidium of the russian academy of sciences , program of state support for leading scientific schools ns-2258.2008.2 , and the russian foundation for basic research ( project no .
09 - 02 - 00165a ) .
the soho / lasco data used here are produced by a consortium of the naval research laboratory ( usa ) , max - planck - institut fuer aeronomie ( germany ) , laboratoire dastronomie ( france ) , and the university of birmingham ( uk ) .
soho is a project of international cooperation between esa and nasa .
the mark 4 data are courtesy of the high altitude observatory / ncar .
mann , g. , h. aurass , a. klassen , c. estel , and b. j. thompson ( 1999 ) , coronal transient waves and coronal shock waves , 8th soho workshop `` plasma dynamics and diagnostics in the solar transition region and corona '' , paris , france , 22 - 25 june 1999 , esa sp-466 , 477 - 481 .
strachan , l. , r. suleiman , a. v. panasyuk , d. a. biesecker , and j. l. kohl ( 2002 ) , empirical densities , kinetic temperatures , and outflow velocities in the equatorial streamer belt at solar minimum , _ astrophys .
j. _ , _ 571 _ , 10081014 .
measured from the direction of cme propagation at the cme 1 shock front boundary at 11:44 ut ( crosses ) , cme 1 frontal structure boundary at 10:19 ut ( empty circles ) , cme 2 frontal structure boundary at 10:05 ut ( solid circles ) . ]
variation with distance @xmath57 from the solar center , in eight cmes with high velocities : empty squares
20 september 1997 , @xmath13 = 20@xmath58 ; solid circles 11 june 1998 , @xmath13 = 80@xmath58 ; crosses 3 march 2000 , pa = 230@xmath58 ; solid triangles
28 june 2000 , @xmath13 = 270@xmath58 ; solid diamonds
22 november 2001 , @xmath13 = 247 - 254@xmath58 ; empty triangles 21 april 2002 , @xmath13 = 270@xmath58 ; empty diamonds 26 october 2003 , @xmath13 = 265 - 290@xmath58 ; empty circles 4
november 2003 , @xmath13 = 238@xmath58 ( from mark 4 and lasco c2 , c3 data ) .
the dashed curves are the proton mean free path @xmath50 calculated for two proton temperatures . ] | mark 4 and lasco c2 , c3 coronagraph data analysis shows that , up to the distance @xmath0 5 r@xmath1 from the center of the sun , the thickness of a cme - generated shock front may be of order of the proton mean free path .
this means that the energy dissipation mechanism in a shock front at these distances is collisional . |
in this section we review the general theory of hyperbolic @xmath0-manifolds with geodesic boundary , stating the main results we will need in the sequel . [ [ local - structure - cusps - compactifications ] ] local structure , cusps , compactifications + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + in the rest of this paper we will call _ hyperbolic _ an orientable finite - volume complete riemannian @xmath0-manifold with _ non - empty _ boundary , locally isometric to an open subset of a closed half - space of @xmath1 .
we will always denote such a manifold by @xmath2 .
note that @xmath3 is totally geodesic .
doubling @xmath2 along its boundary and using the description of the ends of the double @xcite , one can show that @xmath2 consists of a compact portion together with some `` toric and annular cusps . ''
a toric cusp is here a space of the form @xmath4 , attached to the rest of @xmath2 along @xmath5 , and an annular cusp is defined analogously .
note that toric cusps are disjoint from @xmath3 , while an annular cusp gives two punctures in @xmath3 .
in particular , @xmath3 is compact if and only if @xmath2 has no annular cusps . given @xmath2 as above , we can naturally get a compact manifold @xmath6 by adding a torus @xmath7 for each toric cusp @xmath4 , and an annulus @xmath8 for each annular cusp @xmath9 . however , it turns out that when there are annular cusps another compactification @xmath10 of @xmath2 is more suited to the geometric situation
. we define @xmath10 as a quotient of @xmath6 , where in each annular cusp @xmath11 $ ] with @xmath12 $ ] , for all @xmath13 $ ] we collapse @xmath14 to one point @xmath15 .
note that @xmath2 is obtained from @xmath10 by removing some toric boundary components and drilling some properly embedded arcs .
[ [ rigidity - and - kojima - decomposition ] ] rigidity and kojima decomposition + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + a key result for computational purposes is the following : [ rigidity : teo ] any two homeomorphic hyperbolic manifolds are isometric .
this result is commonly referred to as _ rigidity theorem _ , and a proof for the geodesic boundary case was spelled out in @xcite .
another very important fact is that the hyperbolic structure determines certain combinatorial data which can be employed to efficiently test two manifolds for homeomorphism .
this result is analogous to the epstein - penner decomposition of cusped hyperbolic manifolds without boundary @xcite , and it was proved by kojima @xcite .
its statement involves the notion of _ truncated polyhedron _
, that we now give .
consider the projective model of hyperbolic @xmath0-space , viewed as the open unit ball @xmath16 in the euclidean @xmath0-space @xmath17 .
let us call _ finite _ the points of @xmath16 , _ ideal _ those of @xmath18 , and _ ultra - ideal _ the other points of @xmath17 .
consider a convex polyhedron @xmath19 in @xmath17 with ideal and/or ultra - ideal vertices , and all edges meeting the closure of @xmath16 .
dual to each ultra - ideal vertex of @xmath19 there is an open hyperbolic half - space , and we define @xmath20 to be @xmath21 minus these half - spaces . any @xmath20 arising like this from some @xmath19 will be called a truncated polyhedron . note that @xmath20 has _ internal _ faces , those coming from faces of @xmath19 , and _ truncation _ faces .
moreover internal and truncation faces lie at right angles to each other .
[ kojima : teo ] any hyperbolic manifold admits a canonical decomposition as a gluing of truncated polyhedra along the internal faces . in the sequel we will need to refer to the geometric argument underlying this result ,
so we briefly sketch it here .
regard @xmath17 as the hyperplane at height @xmath22 in minkowsky @xmath23-space @xmath24 .
suppose first that @xmath2 has no toric cusps , and identify the universal cover of @xmath2 to an intersection of half - spaces of @xmath16 .
dual to each such subspace there is a point having norm @xmath22 in @xmath24 , and the decomposition of @xmath2 is obtained by taking the faces of the convex hull of all these points , projecting first to @xmath17 , truncating , and then projecting to @xmath2 .
when there are toric cusps one must also take suitably small margulis neighbourhoods of these cusps , lift them to horoballs in @xmath16 , consider the duals to these horoballs on the light - cone of @xmath24 , include these duals in the convex hull , and suitably subdivide some of the resulting faces .
we address the reader to @xcite for all the details .
we are now in a position to explain why we have introduced the compactification @xmath10 : [ barbar : prop ] let @xmath2 be hyperbolic . for each truncated polyhedron @xmath20 in the kojima decomposition of @xmath2
, consider the corresponding euclidean polyhedron @xmath19 .
glue these @xmath19 s along the same maps as in the kojima decomposition , and remove open stars of the vertices .
the resulting space is then homeomorphic to @xmath10 .
[ [ topological - obstructions - to - hyperbolicity ] ] topological obstructions to hyperbolicity + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + since dealing with open manifolds is impossible by computer , the idea to enumerate the hyperbolic @xmath2 s is to enumerate the corresponding compact @xmath6 s or @xmath10 s .
the next two results show that the presence of annular cusps makes a dramatic difference .
the former can be found in @xcite , the latter follows from the results discussed in section [ cfmp : section ] .
( recall that we are calling ` hyperbolic ' a manifold with _ non - empty _ geodesic boundary ) .
[ obstructions : teo ] let @xmath25 be a compact orientable @xmath0-manifold with boundary , and let @xmath2 be @xmath25 minus the toric components of @xmath26 .
the following conditions are pairwise equivalent : * @xmath2 is hyperbolic ; * @xmath2 is hyperbolic , it has no annular cusps , and @xmath27 ; * @xmath25 is irreducible , boundary - irreducible , acylindrical and atoroidal , and @xmath28 .
[ bad : annuli : prop ] * for any @xmath29 there exists a hyperbolic @xmath2 such that @xmath6 is the handlebody of genus @xmath30 ; * for any compact @xmath0-manifold @xmath25 there exists a hyperbolic @xmath2 such that @xmath31 .
in this section we recall the basics of the theory of simple and special spines , and the related theory of ideal triangulations , describing how it can be employed to ( partially ) enumerate the class of @xmath0-manifolds we are interested in .
proofs of all results on spines and complexity can be found in @xcite .
[ [ simple - spines - and - complexity ] ] simple spines and complexity + + + + + + + + + + + + + + + + + + + + + + + + + + + + throughout the present section we will employ the pl category for @xmath0-manifolds and use the customary notions of pl topology , see @xcite .
simple polyhedron _ is a compact polyhedron @xmath20 such that the link of each point of @xmath20 can be embedded in the space given by a circle with three radii . in particular , @xmath20 has dimension at most @xmath32 .
finite graphs and closed surfaces are examples of simple polyhedra . a point of a simple polyhedron
is called a _ vertex _ if its link is precisely given by a circle with three radii .
a regular neighbourhood of a vertex is shown in fig .
[ almostspecial : fig]-(3 ) .
[ almostspecial : fig ] from the figure one sees that the vertices are isolated , whence finite in number .
graphs and surfaces do not contain vertices . if @xmath25 is a compact @xmath0-manifold with non - empty boundary , we call _ spine _ of @xmath25 a subpolyhedron @xmath20 of @xmath25 such that @xmath33 is an open collar of @xmath26
. we call _ complexity _ of @xmath25 , and denote by @xmath34 , the minimal number of vertices of a simple spine of @xmath25 .
we say that a spine of @xmath25 is _ minimal _ if it has @xmath34 vertices and it does not contain any proper subpolyhedron which is also a spine of @xmath25 .
[ [ special - spines - and - ideal - triangulations ] ] special spines and ideal triangulations + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + we now introduce a more restrictive type of spine which turns out to have a very clear geometric counterpart .
a simple polyhedron @xmath20 is called _ almost - special _ if the link of each point of @xmath20 is given by a circle with either zero , or two , or three radii . the local aspects of @xmath20 are correspondingly shown in fig .
[ almostspecial : fig ] .
the points of type ( 2 ) or ( 3 ) are called _ singular _ , and the set of singular points of @xmath20 is denoted by @xmath35 .
we will say that @xmath20 is _ special _ if it is almost - special , @xmath35 contains no circle component , and @xmath36 consists of open @xmath32-discs .
we now call _ ideal triangulation _ of a compact @xmath0-manifold @xmath25 with non - empty boundary a realization of the interior of @xmath25 as follows .
we take a finite number of tetrahedra , we glue together in pairs the faces of these tetrahedra along simplicial maps , and we remove the vertices .
equivalently , an ideal triangulation of @xmath25 is a realization of @xmath25 as a gluing of _ truncated _ tetrahedra .
the relation between spines and triangulations is given by the following : [ duality : basic : prop ] the set of ideal triangulations of a @xmath0-manifold @xmath25 corresponds bijectively to the set of special spines of @xmath25 .
the polyhedron corresponding to a triangulation is the @xmath32-skeleton of the dual cellularization , as shown in fig .
[ duality : fig ] .
[ duality : fig ] [ [ manifolds - having - special - minimal - spines ] ] manifolds having special minimal spines + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + special spines have two main advantages if compared to merely simple ones .
first of all , a special spine determines the manifold it is a spine of @xcite , which is false for simple spines .
second , no efficient method for listing simple spines is known , whereas enumerating special spines in increasing order of complexity is very easy ( at least theoretically , see @xcite ) . for these reasons ,
the next result is crucial for computational purposes : [ special : prop ] let @xmath25 be a compact @xmath0-manifold with non - empty boundary .
the following conditions are pairwise equivalent : * @xmath25 is irreducible , @xmath37-irreducible , acylindrical , and not the @xmath0-disc ; * @xmath25 has some special minimal spine ; * all minimal spines of @xmath25 are special .
this result is not quite stated in this form in @xcite or in any of matveev s papers , but it easily follows from the proof of ( * ? ? ?
* theorem 2.2.4 ) .
[ [ naf - enumeration - strategy ] ] naf enumeration strategy + + + + + + + + + + + + + + + + + + + + + + + + + cusped hyperbolic manifolds without boundary were studied and enumerated in @xcite , which explains why we have decided to restrict to manifolds with non - empty boundary .
in addition , we forbid here the presence of annular cusps , because , according to propositions [ bad : annuli : prop ] and [ special : prop ] , the theory of spines , triangulations , and complexity does not appear to be well - suited for the investigation of such manifolds .
see section [ cfmp : section ] .
let us then define @xmath38 as the set of all hyperbolic manifolds having complexity @xmath39 and non - empty _ compact _ geodesic boundary .
we also define @xmath40 as the set of all orientable , compact , irreducible , @xmath37-irreducible , and acylindrical manifolds with negative @xmath41 .
theorem [ obstructions : teo ] implies that if @xmath42 then @xmath43 .
conversely , if @xmath44 , then @xmath25 minus the toric components of @xmath26 belongs to @xmath38 if and only if @xmath25 is atoroidal .
we can then view @xmath40 as the set of _ candidate hyperbolic _ manifolds with complexity @xmath39 and without annular cusps .
note now that proposition [ special : prop ] applies to the elements of @xmath40 .
therefore the following theoretical steps lead to the exact determination of @xmath38 , assuming @xmath45 is known inductively for @xmath46 : 1 .
produce the list of all special spines with @xmath39 vertices and negative @xmath41 ; 2 . remove from the list the spines @xmath20 whose associated manifold @xmath47 is not hyperbolic ; 3 .
remove from the list the spines @xmath20 such that @xmath47 belongs to some @xmath48 for @xmath46 ; 4 . for @xmath20
varying in the list , compare the manifolds @xmath47 for equality , discarding duplicates .
step ( 1 ) does not present any theoretical difficulty , but its practical implementation is quite demanding if no computational shortcuts are employed .
we will discuss these shortcuts in the next paragraph .
and we will explain how to carry out the other steps in the next section .
[ [ pseudo - minimal - spines ] ] pseudo - minimal spines + + + + + + + + + + + + + + + + + + + + + taking into account propositions [ duality : basic : prop ] and [ special : prop ] , the reader may wonder why we have employed special spines rather than triangulations to list the elements of @xmath38 .
the first remark is that one can always restrict to _ minimal _ triangulations without losing any potentially interesting manifold .
however , it turns out that dual to a triangulation which is minimal among triangulations of the same manifold , there is often a special spine which is not minimal among simple spines of the same manifold . of course
this can only happen if the corresponding manifold violates one of the topological constraints of proposition [ special : prop ] , but we know from theorem [ obstructions : teo ] that in this case the manifold is not hyperbolic , so we can discard it .
in other words , _ non - minimality is a much more flexible notion for spines than for triangulations _
, so using spines we can substantially reduce the list of manifolds that we will later need to investigate .
of course it is impossible to check in a direct fashion whether a special spine is minimal , but there are many criteria for non - minimality , which can be used as tests to discard spines which will certainly not bring any relevant manifold .
the tests used in @xcite are based on the moves shown in fig .
[ moves : fig ] , which are easily seen to transform a spine [ moves : fig ] of a manifold into another spine of the same manifold . to be precise ,
let us say that a spine is _ pseudo - minimal _ if it can not be transformed into a spine with fewer vertices by a combination of the moves shown in fig .
[ moves : fig ] .
the key point mentioned above on the flexibility of spines is that moves ( 1 ) and ( 2 ) do not lead to special spines ( in general ) , so they do not have counterparts at the level of triangulations . the first step of the enumeration strategy is then replaced by the following : * produce the list of all pseudo - minimal special spines with @xmath39 vertices and negative @xmath41 .
we also mention that another trick very important for computational purposes is to construct the candidate spines portion after portion , following the branches of a tree , and to apply the non - minimality tests arising from the moves of fig .
[ moves : fig ] also to partially constructed spines , thus `` cutting the dead branches '' at their bases .
in this section we discuss how an ideal triangulation can be employed to construct a hyperbolic structure on a given manifold and to recognize the canonical kojima decomposition of that manifold .
this allows to carry out steps ( 2)-(4 ) of the enumeration strategy for @xmath38 explained in the previous section .
we actually include in the discussion also manifolds with annular cusps , because the methods via ideal triangulations to construct and recognize the hyperbolic structure apply to these manifolds too .
it is only step ( 1 ) of the enumeration strategy ( the listing of special spines ) that is not suited to manifolds with annular cusps , and the reason why these manifold are ruled out from @xmath38 .
we first treat the compact case and then sketch the variations needed for the case where there are also some cusps . for all details and proofs ( and for some very natural terminology that we use here without giving actual definitions )
we address the reader to @xcite .
[ [ compact - case - moduli - and - equations ] ] compact case : moduli and equations + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + the basic idea for constructing a hyperbolic structure via an ideal triangulation is to realize the tetrahedra of the triangulation as truncated tetrahedra in @xmath1 and then require that the structures match when the tetrahedra are glued together .
the following facts show that one can use the dihedral angles as moduli to parameterize the realizations of a tetrahedron and to check consistency : * a hyperbolic structure on a combinatorial truncated tetrahedron is determined by the 6-tuple of dihedral angles along the internal edges ; * the only restriction on this 6-tuple of positive reals is that the angles of each of the four truncation triangles should sum up to less than @xmath49 ; * the lengths of the internal edges can be computed as explicit functions of the dihedral angles ; * a choice of hyperbolic structures on the tetrahedra of an ideal triangulation of a manifold @xmath25 gives rise to a hyperbolic structure on @xmath25 if and only if all matching edges have the same length and the total dihedral angle around each edge of @xmath25 is @xmath50 .
given an ideal triangulation consisting of @xmath39 tetrahedra one then has the _ hyperbolicity equations _ : a system of @xmath51 equations with unknown varying in an open set of @xmath52 which , by rigidity ( theorem [ rigidity : teo ] ) , admits one solution at most .
[ [ canonical - decomposition - and - tilts ] ] canonical decomposition and tilts + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + once a hyperbolic structure has been constructed on a manifold @xmath2 using an ideal triangulation @xmath53 , one natural issue is to decide if @xmath53 is the canonical decomposition of @xmath2 and , if not , to promote @xmath53 to become canonical . these matters are faced using the _ tilt formula _
@xcite , that we now describe .
recall first that the kojima decomposition of @xmath2 is constructed by projecting first to @xmath1 and then to @xmath2 the faces of the convex hull of the set @xmath54 of the duals to the boundary components of the universal cover of @xmath2 .
if @xmath55 is a @xmath56-simplex in @xmath53 , each end of a lifting of @xmath55 to @xmath1 determines a point of @xmath57 .
now let two tetrahedra @xmath58 and @xmath59 share a @xmath32-face @xmath60 , and let @xmath61 and @xmath62 be liftings of @xmath63 and @xmath60 to @xmath1 such that @xmath64 .
let @xmath65 be the @xmath32-subspace of @xmath24 that contains the three points of @xmath57 determined by @xmath66 .
for @xmath67 let @xmath68 be the half-@xmath0-subspace bounded by @xmath65 and containing the fourth point of @xmath57 determined by @xmath69 .
then one can show that @xmath53 is the canonical kojima decomposition of @xmath2 if and only if , whatever @xmath70 , the following conditions are met : * the convex hull of @xmath71 and @xmath72 does not contain the origin of @xmath24 ; * @xmath71 and @xmath72 lie on distinct @xmath0-subspaces of @xmath24 .
in addition , if condition ( a ) is met for all triples @xmath70 , the canonical decomposition is obtained by merging together the tetrahedra along which condition ( b ) is not met .
the tilt formula defines a real number @xmath73 describing the `` slope '' of @xmath74 .
more precisely , one can translate conditions ( a ) and ( b ) into the inequalities @xmath75 and @xmath76 respectively .
since we can compute tilts explicitly in terms of dihedral angles , this gives a very efficient criterion to determine whether @xmath53 is canonical or a subdivision of the canonical decomposition .
even more , it suggests where to change @xmath53 in order to make it more likely to be canonical , namely along @xmath32-faces where the total tilt is positive .
this is achieved by 2-to-3 moves along the offending faces , as discussed in @xcite .
[ [ the - non - compact - case ] ] the non - compact case + + + + + + + + + + + + + + + + + + + + when one is willing to accept both compact geodesic boundary and cusps , the same strategy for constructing the structure and finding the canonical decomposition applies , but many subtleties and variations have to be taken into account .
let us quickly mention which . _ moduli_. as suggested by proposition [ barbar : prop ] , to construct a hyperbolic structure on a manifold @xmath2 one must take an ideal triangulation @xmath53 of the compactification @xmath10 , in which each toric cusp is completed with a torus and each annular cusp is completed with a segment .
moreover one must suppose that the segments arising from the annular cusps of @xmath2 are edges of @xmath53 .
to parameterize tetrahedra one should then proceed as follows : * assign dihedral angle @xmath77 to each edge corresponding to an annular cusp of @xmath2 , which geometrically means that the edge , before truncation , is tangent to the boundary of @xmath16 ; * if three edges meet at a vertex asymptotic to a toric cusp of @xmath2 , assign them dihedral angles summing up to @xmath49 , which geometrically means that the vertex is an ideal one .
switching viewpoint , if one starts from an ideal triangulation @xmath78 of a manifold @xmath25 , one must arbitrarily choose a family @xmath79 of edges of @xmath78 and assign dihedral angles @xmath77 to the edges in @xmath79 , and dihedral angles summing up to @xmath49 to triples of edges asymptotic to toric components of @xmath26 .
if the consistency and completeness equations are satisfied ( see below ) , this leads to a hyperbolic structure on @xmath25 minus @xmath79 and the toric components of @xmath26 . _
equations_. if an internal edge with non - zero dihedral angle ends in a cusp then its length is infinity , so some of the length equations must be dismissed when there are cusps .
there are no consistency issues connected with half - infinite edges but , when an edge is infinite at both ends , one must make sure that the gluings around the edge do not induce a sliding along the edge , which translates into the condition that the _ similarity moduli _ @xcite of the euclidean triangles around the edge have product @xmath22 .
this ensures existence of the hyperbolic structure , but one still has to impose completeness of cusps .
just as in the case where there are cusps only , this amounts to requiring that the similarity tori on the boundary be euclidean , which translates into the _ holonomy equations _ involving the similarity moduli .
note that there is no completeness issue connected to annular cusps . _
canonical decomposition_. when there are cusps , the set of points @xmath57 to take the convex hull of consists of the norm-@xmath22 duals of the boundary components of the universal cover and of some points on the light - cone dual to margulis neighbourhoods of the cusps . the precise discussion of how to choose these extra points is too complicated to be reproduced here ( see @xcite ) , but in practice one has that the choice of _ sufficiently small _ margulis neighbourhoods always leads to the right result .
all the hyperbolic manifolds found by computer in @xcite can be decomposed into genuine tetrahedra , _
i.e. _ tetrahedra with positive dihedral angles ( with ultra - ideal vertices , and possibly some ideal ones ) .
we do not know if all hyperbolic manifolds admit such a genuine decomposition , but we know that if we subdivide the kojima decomposition into tetrahedra we find some genuine tetrahedra and possibly some flat ones , which do not contribute to the volume
. therefore the problem of computing the volume of a hyperbolic manifold is completely reduced to the same problem for a single tetrahedron with arbitrarily assigned dihedral angles .
we discuss in this section various formulae for this volume . in doing this we will employ a unified viewpoint , which includes finite , ideal , and ultra - ideal vertices .
[ [ volumes - of - polyhedra - historical - remarks ] ] volumes of polyhedra : historical remarks + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + the calculation of the volume of a polyhedron in 3-dimensional space is a very old and difficult problem .
the first known result in this direction belongs to tartaglia ( 1494 ) who found a formula for the volume of a euclidean tetrahedron , now known as the cayley - menger formula .
it was recently shown in @xcite and @xcite that the volume of any euclidean polyhedron is a root of an algebraic equation whose coefficients are polynomial functions of the lengths of the edges , the polynomials being determined by the combinatorial type of the polyhedron . in the hyperbolic and spherical spaces the situation is much more complicated .
we confine here to the hyperbolic case , and we first recall that the volume formula for a biorthogonal tetrahedron ( _ orthoscheme _ ) has been known since lobachevsky and schlfli ( see @xcite and @xcite respectively ) .
the volumes of the lambert cube and of some other polyhedra were computed by kellerhals @xcite , derevnin and mednykh @xcite , mednykh , parker , and vesnin @xcite , and others .
the volume of a regular tetrahedron in hyperbolic space was investigated by martin @xcite .
the volume formula for a hyperbolic tetrahedron with a few non - ideal vertices was found by vinberg @xcite . despite these partial results ,
a formula for the volume of an arbitrary hyperbolic tetrahedron has been unknown until very recently . the general algorithm for obtaining such a formula
was indicated by hsiang @xcite , and the complete solution of the problem was given within the space of a few years by several authors , namely cho and kim @xcite , murakami and yano @xcite , and ushijima @xcite .
in all these papers the volume of a tetrahedron is expressed as an analytic formula involving 16 dilogarithm or lobachevsky functions whose arguments depend on the dihedral angles of the tetrahedron and on some additional parameter which is found as a root of some complicated quadratic equation with complex coefficients .
a geometric meaning of the murakami - yano formula was recognized by leibon @xcite from the viewpoint of the so - called regge symmetry .
an excellent exposition of these ideas and a complete geometric proof of the murakami - yano formula was given by mohanty @xcite .
it is worth mentioning that the ideas of regge symmetry and scissors congruence were also partially used by cho and kim in @xcite , where the first general formula was actually obtained .
a remarkable phenomenon is that the volume formulae for the tetrahedron become much easier for _ symmetric _ tetrahedra , _
i.e. _ for tetrahedra with identical dihedral angles at opposite edges .
this fact was first noticed by milnor @xcite , who expressed the volume of an ideal tetrahedron ( which is automatically symmetric ) as the sum of the values of the lobachevsky function on the three dihedral angles .
it was later shown by derevnin , mednykh , and pashkevich @xcite that a rather simple formula also exists for an arbitrary symmetric tetrahedron with finite vertices . in the next paragraph
we will present an elementary integral formula for the volume of an arbitrary tetrahedron in hyperbolic space .
the formula involves some parameters depending on the dihedral angles , and it is very helpful to actually evaluate the volume by computer . the murakami - yano result can be obtained as an easy consequence of this formula .
[ [ volume - of - a - tetrahedron - without - truncation ] ] volume of a tetrahedron without truncation + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + let @xmath80 denote the tetrahedron in hyperbolic @xmath0-space with dihedral angles @xmath81 , @xmath82 , @xmath83 , @xmath84 , @xmath85 , @xmath60 as in fig
. [ tetra : fig ] . [ tetra : fig ] recall @xcite that @xmath78 is determined up to isometry by @xmath81 , @xmath82 , @xmath83 , @xmath84 , @xmath85 , @xmath60 .
moreover if the sum of the triple of angles at a certain vertex is @xmath49 then the vertex is ideal .
similarly , if the sum is less than @xmath49 then the vertex is ultra - ideal , _ i.e. _ the tetrahedron is truncated at the vertex , which is then replaced by a triangle .
the next result is due to derevnin and mednykh @xcite : [ volume : teo ] suppose that all the vertices of @xmath86 are ideal or finite .
then @xmath87 is given by @xmath88 with @xmath89 and @xmath90 given by @xmath91 where @xmath92 and @xmath93 .
moreover the @xmath94 s and @xmath95 s are all real numbers , so the integral is just an ordinary integral on an interval of the real line , and the function to be integrated vanishes at the @xmath95 s .
[ vertices : rem ] _ there is a very transparent geometric interpretation of the sums of dihedral angles @xmath96 , @xmath97 , @xmath98 , and @xmath99 appearing in the numerator of the volume formula , and of the sums @xmath100 , @xmath101 , and @xmath102 appearing in the denominator .
namely , the @xmath103 s correspond to the triples of edges incident to the vertices of the tetrahedron , while the @xmath104 s correspond to the hamiltonian cycles .
_ [ equation : rem ] _ the parameters @xmath89 and @xmath90 appearing in the volume formula can be shown to be roots of the equation @xmath105 while @xmath106 where @xmath107 is the gram matrix of @xmath78 .
the numbers @xmath89 and @xmath90 also have a geometric meaning , as explained in @xcite .
namely , they arise as parameters for the decomposition of an ideal octahedron into four ideal tetrahedra with an edge in common .
the octahedron is canonically defined by the tetrahedron @xmath78 , and its dihedral angles are just linear combinations of those of @xmath78 .
_ recall now that the dilogarithm function is defined by the integral @xmath108 where @xmath109 and @xmath110 is the continuous branch of the logarithm function given by @xmath111 with the constraint @xmath112 .
let @xmath113 as an immediate consequence of theorem [ volume : teo ] we have the following murakami - yano - ushijima formula obtained in @xcite and @xcite : [ dilogarithm : cor ] suppose that all the vertices of @xmath86 are ideal or finite .
then @xmath114 where @xmath115 conerning this formula we note that @xmath116 where @xmath117 is the lobachevsky function defined by the integral @xmath118 this shows , as already announced , that the volume of a tetrahedron is an algebraic sum of 16 lobachevsky functions .
[ [ truncated - tetrahedra ] ] truncated tetrahedra + + + + + + + + + + + + + + + + + + + + according to @xcite and @xcite , the determinant of the gram matrix is strictly negative also for truncated tetrahedra , so the numbers @xmath119 appearing in the statement of theorem [ volume : teo ] are still real .
however a non - trivial issue arises concerning the choice of the analytic branch of the @xmath120 function for the definition of the numbers @xmath89 and @xmath90 .
more precisely , one can be forced to change branch , in order to ensure the continuity of @xmath89 and @xmath90 , when @xmath121 approaches @xmath77 , which indeed can happen when there are some ultra - ideal vertices .
for example , for @xmath122 we have @xmath123 , whereas for @xmath124 we obtain @xmath125 . therefore in this case we have to change the analytic branch of @xmath126 . the most convenient way to do so is to replace @xmath127 by @xmath128 , which yields a continuous variation of the values of @xmath89 and @xmath90 as @xmath121 passes through 0 .
we note that @xmath129 and @xmath130 are always positive , so there is no such problem with @xmath131 .
one can now check that , after appropriately choosing the analytic branches of the @xmath120 function , the formula of theorem [ volume : teo ] still gives the correct value of the volume for truncated tetrahedra .
[ [ symmetric - tetrahedra ] ] symmetric tetrahedra + + + + + + + + + + + + + + + + + + + + we will say that @xmath86 is _ symmetric _ if @xmath132 , @xmath133 , and @xmath134 , and in this case we will denote @xmath78 just by @xmath135 for simplicity .
we begin with the easy case of ideal tetrahedra , which are automatically symmetric .
moreover @xmath135 is ideal if and only if @xmath136 . in this case
the volume turns out @xcite to be given simply by @xmath137 where @xmath138 is the lobachevsky function defined above . for symmetric tetrahedra @xmath135 with finite vertices , _
i.e. _ such that @xmath139 , the following was shown in @xcite : [ symvolumehyp ] if @xmath140 has finite vertices then @xmath87 is given by @xmath141 where @xmath142 satisfies @xmath143 with @xmath144 , @xmath145 , and @xmath146 .
[ sinerule : rem ] _ the value of @xmath147 in the previous statement has a simple geometric interpretation in terms of the `` sine rule '' ( see ( * ? ? ? * theorem 7 ) ) : if @xmath148 are the lengths of the edges of @xmath78 with dihedral angles @xmath149 respectively , then _
@xmath150 we conclude this section by noting that no simple formula is currently known for the volume of a symmetric tetrahedron with ultra - ideal vertices , but it is reasonable to expect that such a formula should exists and have connections with an ultra - ideal version of the sine rule .
in this section we recall the definition and main properties of a certain class of manifolds @xmath151 studied in @xcite , and we describe the experimental results of @xcite on @xmath38 for @xmath152 . recall that @xmath38 contains the hyperbolic manifolds of complexity @xmath39 with non - empty _ compact _ geodesic boundary .
the next section will be devoted to manifolds with non - compact boundary , _
i.e. _ with annular cusps .
[ [ a - special - class - of - manifolds ] ] a special class of manifolds + + + + + + + + + + + + + + + + + + + + + + + + + + + + let us denote by @xmath153 the closed orientable surface of genus @xmath30 .
for @xmath29 and @xmath154 we define @xmath155 as the set of all compact orientable manifolds @xmath25 having an ideal triangulation with @xmath156 tetrahedra , and @xmath157 as a motivation for this definition , we mention here that an ideal triangulation of a manifold whose boundary is the union of @xmath153 and @xmath158 tori contains at least @xmath156 tetrahedra .
so @xmath155 is the set of manifolds @xmath25 having the smallest possible complexity , given the topological constraints on @xmath26 .
our first result shows that the class just introduced is very large : * @xmath155 is non - empty precisely for @xmath159 or @xmath160 and @xmath30 even ; * the values of @xmath161 for small @xmath30 and @xmath158 are as shown in table 1 ; * for any fixed @xmath158 there exist constants @xmath162 such that @xmath163 we now give the main statement we have about the elements of @xmath155 and their topological and geometric invariants .
we address the reader to @xcite for the definition of the _ heegaard genus _ of a triple @xmath164 , and to @xcite for the definition of the turaev - viro invariants @xmath165 of a compact @xmath0-manifold @xmath25 , for @xmath166 .
[ main : teo ] let @xmath167 .
the following holds : 1 .
[ main : hyperbolic : item ] @xmath168 is hyperbolic , and its volume depends only on @xmath30 and @xmath158 ; 2 .
[ main : canonical : item ] @xmath25 has a unique ideal triangulation with @xmath156 tetrahedra , which gives the canonical kojima decomposition of @xmath168 ; 3 . [
main : complexity : item ] @xmath25 has complexity @xmath156 ; 4 .
[ main : genus : item ] the heegaard genus of @xmath169 is @xmath170 ; 5 .
[ main : homology : item ] @xmath171 ; 6 .
[ main : turaev : viro : item ] the turaev - viro invariant @xmath165 of @xmath25 depends only on @xmath172 , @xmath30 and @xmath158 .
the first two points of this theorem are established using precisely the philosophy of moduli , equations , and tilts sketched in section [ equations : section ] .
namely , given a minimal triangulation of @xmath167 , one shows that each tetrahedron has either one or no vertex asymptotic to a toric cusp .
if there is one such vertex , one chooses the dihedral angles to be @xmath173 at the edges ending at that vertex , and to be all equal to some @xmath79 at the other edges .
if there is no such vertex , one chooses the tetrahedron to have all dihedral angles equal to some @xmath174 . using a continuity argument
then one sees that there exist values of @xmath79 and @xmath174 satisfying the consistency equations .
completeness and the computation of tilts is then straight - forward .
we also notice that the previous results show the tremendous power of hyperbolic geometry compared to the topological invariants : the classes @xmath155 are extremely large , and only hyperbolic geometry is able to distinguish their elements from each other .
we refrain from recalling the precise statements and details here , but we want to mention that a thorough analysis of the dehn fillings of the elements of @xmath155 was carried out in @xcite , leading in particular to the solution of a problem raised by gordon and wu @xcite on the maximal distance between a boundary - reducible and a non - acylindrical slope on a `` large '' hyperbolic manifold . [
[ experimental - results ] ] experimental results + + + + + + + + + + + + + + + + + + + + we describe here @xmath38 for @xmath152 .
it is easy to see that @xmath175 is empty .
moreover , @xmath176 was shown in @xcite to have @xmath177 elements , all with the same volume @xmath178 .
therefore @xmath176 coincides with the set @xmath179 discussed above , so all its members share the same invariants , except the hyperbolic structure itself . the strategy mentioned in section [ complexity : section ] has been implemented in @xcite to classify @xmath180 and @xmath181 , leading to the following results .
we denote here by @xmath182 the kojima canonical decomposition of a hyperbolic @xmath2 , and we recall that @xmath40 is the set of candidate hyperbolic @xmath0-manifolds of complexity @xmath39 .
we emphasize that the next results indeed have an experimental nature , but their validity was confirmed by a number of computations by hand and cross - checks .
we also mention that all the values of the volume are approximate ones .
more accurate approximations are available on the web @xcite .
[ h3:teo ] * @xmath180 coincides with @xmath183 and it has @xmath184 elements ; * @xmath185 consists of @xmath186 elements of volume @xmath187 ; * @xmath188 consists of a single manifold of volume @xmath189 ; * the @xmath190 elements of @xmath191 all have boundary @xmath192 , and they split as follows : * * @xmath193 compact @xmath2 s with @xmath182 consisting of three tetrahedra ; the volume function attains on them @xmath194 different values ranging between @xmath195 and @xmath196 , with maximal multiplicity @xmath197 ; * * @xmath0 compact @xmath2 s with @xmath182 consisting of four tetrahedra ; they all have the same volume @xmath198 .
* @xmath181 has @xmath199 elements , and @xmath200 and has @xmath201 more ; * @xmath202 has @xmath203 elements of volume @xmath204 ; * @xmath205 has @xmath206 elements of volume @xmath207 ; * @xmath208 has a single element of volume @xmath209 ; * the @xmath210 elements @xmath2 of @xmath181 not belonging to any @xmath211 split as described in table 2 according to the boundary of @xmath10 ( columns ) and type of blocks of the kojima decomposition ( rows ) . some volume information has also been inserted in each box , namely the unique value of volume if there is one , or the minimum and the maximum of volume , the number of values it attains , and the maximal multiplicity of these values . [ cols="<,^,^,^",options="header " , ] we conclude this section with some information on the actual computer implementation of the enumeration process .
we solve the hyperbolicity equations using newton s method with partial pivoting , after explicitly writing the derivatives of the length function .
convergence to the solution is always extremely fast , and it can be checked to be stable under modifications of the numerical parameters involved in the implementation of newton s method . concerning the kojima decomposition , we mention that the evolution of a triangulation toward the canonical decomposition is not quite sure to converge in general , but it always does in practice .
we also point out that our computer program is only able to handle triangulations : whenever some mixed negative and zero tilts appear , the canonical decomposition must be worked out by hand and actually proved not to be a triangulation .
given a hyperbolic manifold @xmath2 with annular cusps , one may decide to use either @xmath212 of @xmath213 as a definition of the complexity of @xmath2 , the former being more natural from a topological viewpoint , and the latter from a geometric viewpoint .
however , as announced in proposition [ bad : annuli : prop ] and explained in this section , neither definition allows to employ the powerful techniques of special spines and ideal triangulations to actually carry out the enumeration of manifolds in order of increasing complexity .
for this reason the understanding of manifolds with annular cusps is still very limited .
this section is only devoted to the description of a special class of such manifolds , but the properties of this class are already sufficient to show that the set of hyperbolic manifolds with annular cusps is very large , and that there is very little control on the topology of the compactifications of such manifolds .
we will also see that the geometric information on this special class allows to prove some very interesting ( and apparently unrelated ) results .
we address the reader to @xcite for all proofs and further details .
we denote by @xmath214 the set of all possible simplicial pairings between the faces of @xmath39 tetrahedra .
we view two pairings to be the same if they coincide combinatorially , and we note that in general a pairing can fail give an ideal triangulation of a manifold , because the link of the midpoint of an edge can be the projective plane . moreover , if a pairing actually gives a manifold , then this manifold need not be orientable .
however , if we fix an orientation on the tetrahedron and require all the pairing maps to reverse the orientation , the result is an ideal triangulation of an orientable manifold .
we denote by @xmath53 the union of all @xmath214 s .
given @xmath215 we define a space @xmath216 by performing the pairings in @xmath78 and then removing first an open regular neighbourhood of the vertices and then a closed regular neighbourhood of the edges . [ ytbar : rem ] _ the manifold @xmath216 is a non - compact one with boundary , and its topological ends have the shape of the product of an annulus and a closed half - line .
let us define @xmath217 and @xmath218 as the natural compactifications of @xmath216 , as we did in section [ structure : section ] for a hyperbolic @xmath2
. then @xmath217 is a ( possibly non - orientable ) handlebody , with genus @xmath219 if @xmath220 .
moreover , if @xmath78 is an ideal triangulation of a manifold @xmath25 , then @xmath218 is homeomorphic to @xmath25 . _ for @xmath215
the manifold @xmath216 can be viewed as @xmath221 , where @xmath222 is a handlebody and @xmath223 is a system of disjoint loops . at the risk of a little ambiguity , which eventually will not create any serious problem , we then identify @xmath216 to the pair @xmath224 , and , following johannson @xcite , we call such a pair a _ relative handlebody_. we denote by @xmath225 the set of all @xmath216 s as @xmath78 varies in @xmath214 , and by @xmath226 the union of all @xmath225 s . to state our first result we call _ complexity _ of a relative handlebody @xmath224 the minimum of @xmath227 where @xmath84 is a system of disjoint properly embedded discs in @xmath222 such that @xmath228 cuts @xmath229 into a union of pairs of pants .
* for all @xmath215 the relative handlebody @xmath216 is hyperbolic . *
the map @xmath230 gives a bijection between @xmath53 and @xmath226 . * among the hyperbolic @xmath224 s of genus @xmath219 , the elements of @xmath225 can be characterized as those having minimal complexity , equal to @xmath231 , and as those having minimal volume , equal to @xmath232 , where @xmath233 is the volume of a hyperbolic regular ideal octahedron .
the proof of this result entirely depends on the techniques described in section [ equations : section ] .
one first notices that a regular ultra - ideal tetrahedron with all dihedral angles equal to @xmath77 is actually a regular ideal octahedron with a checkerboard coloring of the faces , the white ones being internal faces and the black ones being truncation faces
. then one sees that for each given @xmath220 the corresponding @xmath216 can be constructed by gluing according to @xmath78 the white faces of @xmath39 regular ideal octahedra . to conclude
, one proves that @xmath78 itself is the canonical kojima decomposition of @xmath216 .
remark [ ytbar : rem ] and proposition [ tn : prop ] are already sufficient to prove proposition [ bad : annuli : prop ] which shows , as already mentioned , that there are very many manifolds with annular cusps and that their topology is rather arbitrary .
the next result goes in the same direction .
we call _ tangle _ in a compact 3-manifold with boundary a finite union of disjoint properly embedded arcs .
[ tangles : enlarged : intro : cor ] every tangle in every compact @xmath0-manifold is contained in a tangle whose complement lies in @xmath225 for some @xmath39 ( in particular , it is hyperbolic with annular cusps ) .
[ triangulations : determined : intro : prop ] let @xmath234 and @xmath235 be triangulations of the same compact @xmath0-manifold , with finite and/or ideal vertices , possibly with multiple and self - adjacencies .
assume that the @xmath22-skeleta of @xmath234 and @xmath235 coincide .
then @xmath234 and @xmath235 are isotopic relatively to the 1-skeleton . for @xmath215
, we define @xmath236 to be the `` orientable double '' of @xmath216 , namely either the union of two copies of @xmath216 along the boundary , when @xmath216 is orientable , or the quotient of the orientation covering of @xmath216 under the restriction to the boundary of the involution , when @xmath216 is non - orientable .
we define @xmath237 as the set of all @xmath216 for @xmath215 .
the next result shows that the family @xmath237 is completely classified in combinatorial terms and it is universal for @xmath0-manifolds under the operation of dehn filling : * every member of @xmath237 is an orientable cusped hyperbolic manifold without boundary , and it is the complement of a link in a connected sum of some copies of @xmath238 ; * the correspondence @xmath239 defines a bijection between @xmath53 and @xmath237 .
* every closed orientable @xmath0-manifold is a dehn filling of a manifold in @xmath237 .
since a dehn filling of a hyperbolic manifold `` typically '' is hyperbolic , the class @xmath237 provides a powerful method to construct closed hyperbolic manifolds . as an application of this method
one can establish the following refinement of the main result of @xcite : we now recall that , according to thurston s hyperbolic dehn filling theorem , on each cusp of a finite - volume hyperbolic 3-manifold there is only a finite number of slopes filling along which one gets a non - hyperbolic 3-manifold .
these slopes are called _ exceptional _ , and a considerable effort has been devoted to understanding them @xcite . if @xmath78 is a triangulation , the hyperbolic manifold @xmath236 has a preferred horospherical cusp section , and each component of this section corresponds to an edge of @xmath78 .
moreover the valence of the edge gives a lower bound for the length of the second shortest geodesic on the component .
this fact and the agol - lackenby 6-theorem @xcite imply the following : we conclude this paper by showing that in some cases the combinatorics of a triangulation is already sufficient to prove hyperbolicity of the underlying manifolds . to motivate our result ,
suppose first that a 3-manifold @xmath25 has an _ ideal _ triangulation @xmath78 in which each edge has valence at least @xmath201 .
an easy argument shows that @xmath244 , and that @xmath245 precisely when all valences are @xmath201 .
moreover , in the last case , the boundary of @xmath25 is a disjoint union of tori and klein bottles , and thurston s hyperbolicity equations for cusped manifolds have a very simple solution , given by regular ideal tetrahedra .
analogously , if all edges have one and the same valence @xmath246 , the hyperbolicity equations for the geodesic boundary case have a simple solution , given by regular truncated tetrahedra with dihedral angles @xmath247 . an argument based on the agol - lackenby machinery @xcite allows to generalize these facts as follows : according to the last assertion of this result , the edges of @xmath78 can be straightened to geodesics in @xmath25 without entirely disappearing into infinity , which suggests that @xmath78 itself can be straightened .
we address the reader to @xcite for more information on the general problem of existence of a triangulation which can be straightened to an ideal one .
s. kojima , _ polyhedral decomposition of hyperbolic @xmath0-manifolds with totally geodesic boundary _ , in `` aspects of low - dimensional manifolds , kinokuniya , tokyo , '' adv . stud .
pure math . 20
( 1992 ) , pp .
93 - 112 .
a. d. mednykh , j. parker , a. yu . vesnin , _ on hyperbolic polyhedra arising as convex cores of quasi - fuchsian punctured torus groups _ , seoul national university , 2002 , rim - garc preprint series * 02 - 01 * , 33 pp .
a. ushijima , _ a unified viewpoint about geometric objects in hyperbolic space and the generalized tilt formula _ , in `` hyperbolic spaces and related topics , ii , kyoto , 1999 , '' srikaisekikenkysho kkyroku * 1163 * ( 2000 ) , 85 - 98 . | we describe a natural strategy to enumerate compact hyperbolic @xmath0-manifolds with geodesic boundary in increasing order of complexity .
we show that the same strategy can be employed to analyze simultaneously compact manifolds and finite - volume manifolds having toric cusps . in opposition to this
we show that , if one allows annular cusps , the number of manifolds grows very rapidly , and that our strategy can not be employed to obtain a complete list .
we also carefully describe how to compute the volume of our manifolds , discussing formulae for the volume of a tetrahedron with generic dihedral angles in hyperbolic space .
msc ( 2000 ) : 57m50 . according to thurston s geometrization program
, the theory of hyperbolic manifolds plays a central role in @xmath0-dimensional topology .
hyperbolic manifolds with geodesic boundary , the first example of which was given by thurston himself in @xcite ( and later generalized in @xcite ) , are an important portion of this theory . on the other hand ,
the algorithmic and computer approach to @xmath0-manifolds has been acquiring an increasing popularity in recent years . for cusped hyperbolic manifolds this approach , which was worked out in @xcite and several other papers , again depends on ideas of thurston , namely on the use of moduli for hyperbolic ideal tetrahedra and equations to ensure consistency of the structures . for closed manifolds the basics of the computer approach were set by matveev in @xcite ( see also @xcite ) , and several experimental results were later obtained by himself and other authors ( see @xcite ) . in the present paper we describe the general setting of the algorithmic approach to hyperbolic @xmath0-manifolds with geodesic boundary , concentrating in particular on their enumeration in order of increasing complexity , and on the computation of their volume . |
the authors acknowledge pavel simcik and henry ho for their technical assistance .
the laminates used in the experiment were supplied by rogers corporation .
this research was supported by the australian research council _ discovery projects _ funding scheme ( project number dp1095151 ) .
t. driscoll , g. o. andreev , d. n. basov , s. palit , s. y. cho , n. m. jokerst , and d. r. smith , `` tuned permeability in terahertz split - ring resonators for devices and sensors , '' _ appl .
lett . _ * 91 * , p. 062511
, 2007 .
j. f. ohara , r. singh , i. brener , e. smirnova , j. han , a. j. taylor , and w. zhang , `` thin - film sensing with planar terahertz metamaterials : sensitivity and limitations , '' _ opt .
express _ * 16*(3 ) , pp . 17861795 , 2008 .
j. a. gordon , c. l. holloway , j. booth , s. kim , y. wang , j. baker - jarvis , and d. r. novotny , `` fluid interactions with metafilms / metasurfaces for tuning , sensing , and microwave - assisted chemical processes , '' _ phys .
b : condens . matter _ * 83 * , p. 205130
, 2011 .
h. tao , l. chieffo , m. a. brenckle , s. m. siebert , m. liu , a. c. strikwerda , k. fan , d. l. kaplan , x. zhang , r. d. averitt , and f. g. omenetto , `` metamaterials on paper as a sensing platform , '' _ adv .
( online edition ) .
lee , h .- s .
lee , k .- h .
yoo , and j .-
yook , `` dna sensing based on single element planar double split - ring resonator , '' in _ ieee mtt - s international microwave symposium digest _ , pp . 16851688 , 2009 .
r. e. mudgett , d. i. c. wang , and s. a. goldblith , `` prediction of dielectric properties in oil - water and alcohol - water mixtures at 3000 mhz , 25@xmath26c based on pure component properties , ''
_ j. food sci .
_ * 39 * , pp . 632635 , 1974 . | a multichannel thin - film sensor is implemented from a set of microstrip - coupled split - ring resonators ( srr s ) with different dimensions .
each srr exhibits a unique high - q resonance that is sensitive to the presence of a sample in a particular area .
hence , this srr - based sensor can function ( i ) to detect different samples simultaneously to increase the throughput or ( ii ) to characterise nominally identical samples at multiple frequencies to increase the sensor selectivity .
the design principle is validated with simulation and measurement .
owing to the optimized design , sensing a low - permittivity film with a thickness as small as one thousandth of the operating wavelength is achievable .
withawat withayachumnankul@xmath0 , kata jaruwongrungsee@xmath1 , + christophe fumeaux@xmath2 , and derek abbott@xmath2 @xmath3school of electrical & electronic engineering , the university of adelaide , adelaide , sa 5005 , australia @xmath4school of electronic engineering , faculty of engineering , king mongkut s institute of technology ladkrabang , bangkok 10520 , thailand @xmath5nanoelectronics and mems laboratory , national electronics and computer technology center , pathumthani 12120 , thailand split - ring resonators ( srrs ) are among many fundamental building blocks for metamaterials that can collectively provide customizable values of the permittivity and/or permeability .
a srr is typically made of one or two concentric subwavelength metallic rings ( namely single or double srr , respectively ) with a narrow split in each ring @xcite . in response to an electromagnetic excitation ,
it exhibits a strong magnetic resonance whose frequency is determined by its dimensions , geometry , and constituent materials . on resonance
, the ring develops an intense and localised electric field at the narrow split . the sensitivity of the resonance frequency to constituent materials , together with the field localisation , subwavelength ring size , and high - q resonance , makes srrs ideal for thin - film sensing . owing to these properties , srr s
have recently been implemented for biosensors in different configurations .
planar arrays of srr s were employed for thin - film sensing @xcite .
however , those sensors require a relatively large amount of the sample to uniformly cover an array with tens to hundreds of srrs during the measurement . the constraints of the sample size and uniformity can be eliminated by using either microstrip - coupled srr s @xcite or waveguide - loaded srrs @xcite . although both microstrip and waveguide arrangements can be used for thin - film sensing , electromagnetic coupling is much stronger in the former case , resulting in its higher immunity against noise and fluctuations . for the microstrip - coupling configuration , an srr is positioned in a close proximity to a microstrip transmission line , which builds up a magnetic field around itself in a quasi - tem wave propagation .
this oscillating magnetic field induces circulating current in the srr loop . in the quasistatic limit
, an srr can be approximated by an inductor and a capacitor in the form of a series _ lc _ resonant circuit .
specifically , the ring forms the inductor , and the split forms the capacitor .
the resonance takes place in the srr when the electric energy stored in the capacitor is balanced with the magnetic energy stored in the inductor .
loading a sample onto the srr surface alters the total capacitance of the structure and hence results in a detectable change in the resonance frequency .
this article presents a multichannel thin - film sensor based on microstrip - coupled srr s .
several srr s with different sizes are placed along a microstrip line .
thus , the sensor exhibits multiple resonances , each of which represents a single channel governed by an individual srr .
each resonance is sensitive to the presence of a sample deposited at the narrow split of the corresponding srr .
since every resonator possesses a high - q resonance at a unique frequency , mutual interaction among the resonators is minimal .
therefore , several srrs can be placed densely without compromising the performance . beyond the multichannel sensing capability
, this article proposes additionally a strategy to improve significantly the sensitivity of srr sensors by creating a strong and localized field enhancement in a single spot . in earlier designs , several double srr s @xcite or a single srr with two splits
@xcite have been used . in such structures ,
several distributed capacitors collectively determine the resonance frequency , and a considerable amount of sample is required to change all of these capacitors . if the sample partially covers the structure , the unaffected capacitance reduces the sensitivity in the sensing region as the capacitive change is then not as pronounced . in the present design ,
each resonance is caused by a single srr with only one split , as shown in fig .
[ fig : ba_sensor](a ) , and consequently , only one distributed capacitor is present .
a change in this capacitor by sample loading in the small split region has a direct and large impact on the resonance frequency .
fig .
[ fig : ba_sensor](a ) also depicts some other important features of the implemented resonators .
each srr has a rectangular shape to maximize coupling to the microstrip transmission line .
this leads to a stronger resonance that is robust to measurement uncertainties . at each srr split
, sharp tips are adopted to concentrate the electric field to a small spot @xcite . in terms of a series _ lc _ circuit , the tapered shape of the tips decreases the capacitance at the split , and hence increases the resonance q - factor .
fig .
[ fig : ba_sensor](b ) shows the proposed sensor fabricated by using standard photolithography and chemical etching .
this particular four - channel sensor is realized from four different srrs positioned along a 50-@xmath6 microstrip line .
all of the resonators share the same dimensions excepting the width , @xmath7 , that equals 8 , 6 , 4.5 , and 3 mm for r1 , r2 , r3 , and r4 , respectively .
this causes a difference in the loop inductance and hence in the resonance frequency among the srrs .
the other dimensions common to all of the srrs are as follows : @xmath8 mm , @xmath9 mm , @xmath10 mm , @xmath11 mm , @xmath12 mm , @xmath13 mm , and @xmath14 mm .
the metal for the srrs , microstrip , and ground plane is copper with a thickness of 35 @xmath15 m , coated with gold to prevent oxidization .
the substrate is an rt / duroid 6010.2lm high - frequency laminate ( ceramic - ptfe composite ) with a thickness of 1.90 mm , a relative permittivity of 10.2 , and a loss tangent of 0.0023 .
this sensor specification yields four high - q resonances in the microwave s band . during the experiment
, the sensor is connected to a vector network analyzer , agilent technologies n5230a , and the transmission parameter @xmath16 is registered at room temperature in the range between 1.53.5 ghz with a resolution of 3 mhz . the numerical simulation is carried out with a commercial full - wave electromagnetic solver , cst microwave studio , based on the finite - integration technique ( fit ) .
the gap size @xmath17 in each ring resonator is slightly adjusted in the simulation to reflect the fabrication imperfections , as observed under a microscope .
the measured and simulated transmission profiles of the sensor without any sample are shown in fig .
[ fig : ba_s21ref ] with small discrepancies caused by fabrication tolerances and limits in simulation accuracy .
the results show four non - overlapping resonances corresponding to the four distinct srrs .
the lowest resonance @xmath18 is associated with r1 , which has the largest inductance , the second lowest resonance @xmath19 with r2 , and so on .
no change in the resonances is observed during 5 hours of measurement , demonstrating the thermal stability of the sensor .
further insight can be obtained from the simulation , as shown in fig .
[ fig : ba_field ] , where the electric - field distribution on the sensor surface is calculated for every channel . at each resonance frequency
, only the relevant srr establishes a strong and highly confined electric field around the capacitive split , and no field enhancement is observed elsewhere .
the field enhancement in any active ring is as high as 380 kv / m , given the input power of only 1 w. these calculated field distributions confirm the extreme field confinement achieved with the implemented srrs , which is a basis for high - performance sensing . in the next experiment
, ethanol is used as a sample under test by the multichannel sensor . at 3 ghz
, ethanol has a relative permittivity and loss tangent of 6.0 and 0.96 , respectively @xcite . in each measurement , 5 @xmath15l of ethanol is dropped by using a high - precision pipette onto the split of a selected srr channel , and the sensors response is observed .
the transmission results are shown in fig .
[ fig : ba_s21mult ] . the observed resonance shifting and damping indicate the presence of the sample in the selected channel . by comparing the resonance frequencies of the loaded sensor in fig .
[ fig : ba_s21mult ] with the reference values in fig .
[ fig : ba_s21ref ] , it is clear that non - active srr channels show no detectable change in the resonance frequency .
hence , it is clear that the four sensing channels can operate independently .
each srr can be approximated by a series _ lc _ circuit with a resonance frequency @xmath20 , where @xmath21 denotes the ring inductance and @xmath22 denotes the split capacitance .
the sensitivity of the resonance frequency to the split capacitance can be expressed as @xcite @xmath23 it can be deduced from this equation that if the relative change in the capacitance formed by the split , @xmath24 , is constant , the relative change in the resonance frequency , @xmath25 , is fixed , regardless of the ring inductance . nonetheless , the relative change in the resonance frequency for the active channel varies between -4.6% and -6.1% among the four channels , despite a fixed volume of sample .
this small variation is presumably caused by manual sample deposition , which results in a slight difference in the sample location .
this uncertainty can be readily overcome with an automated process .
the sample volume used in the experiment is considerably small .
nevertheless , additional simulations suggest that only one nanolitre of ethanol , i.e. , a 0.1-mm cube , is unambiguously detectable with a frequency shift of 15 mhz . in other words
, the ratio between the operating wavelength and the film thickness can be as large as three orders of magnitude . in conclusion
, this article presents a thin - film sensor with a series of optimized srr s as sensing elements .
the fabricated sensor successfully performs multichannel detection with an enhanced sensitivity . scaling down the structure to operate at higher frequencies further reduces the minimum amount of detectable sample .
more channels can be added to the sensor provided that resulting resonances are sparse enough to avoid mutual coupling .
the resonance of each channel can be positioned at an abritary frequency where unique dielectric features of the sample are expected .
the realization can be used for on - site disposable sensors , which allow sensing with either high selectivity or high throughput . |
the z(2 ) symmetric quantum field theory of a single scalar field with @xmath0 interaction is the number one textbook prototype model for all kinds of field theoretic methods . at the same time , with its possibility of spontaneous symmetry breaking , it may be viewed as the crudest caricature of the higgs sector of the standard model . in this context the strongly conjectured triviality of the model in four space time dimensions is of physical interest as it leads to upper bounds on the higgs mass .
this is so because triviality means that the cutoff can not be removed from the regularized theory without ending in a free gaussian theory .
the model is then interpreted as an effective theory in which an upper limit on tolerable unphysical cutoff effects implies an upper bound on the interaction strength which in turn is responsible for mass generation by the higgs field .
unfortunately in the four dimensional case , we still have to rely on numerical methods to demonstrate triviality beyond the perturbative regime .
such studies in the lattice regularization have been strongly boosted by a series of papers by lscher and weisz ( lw ) , of which the first two have dealt with the one component model in the symmetric @xcite and in the broken @xcite phase . here
control over the lattice theory was gained by combining large orders in the hopping parameter expansion with careful perturbative renormalization group evolution.@xcite these studies were in addition corroborated by some early monte carlo simulations as for example @xcite , @xcite . in these cases
the ising model was considered as the limit of @xmath0 theory at infinite bare coupling . barring a complicated non - monotonic relation between bare and ( natural ) renormalized couplings on the lattice , this limit is the most interesting case for questions concerning triviality . in recent years one of the authors
has taken up the subject again after some progress had been made in monte carlo methods which allow to achieve a new level of precision in this context with only moderate investments in compute power .
the main new ingredients are on the one hand the use of so - called worm algorithms @xcite , @xcite to simulate arbitrary order contributions of a hopping parameter expansion for observables on finite lattices instead of generating field configurations .
the second ingredient is the use of finite volume renormalization schemes as in @xcite . as triviality is an ultraviolet renormalization effect , more compute power
can be devoted in this way to closely approaching the continuum limit as the thermodynamic limit does not have to be taken .
in other words , the manageable ratios @xmath1 between lattice size and spacing is used to achieve a significant range of small @xmath2 and not for large @xmath3 in physical units . in @xcite , @xcite , @xcite
such a strategy has been explored for the symmetric phase of the model . in this publication
we now offer a finite size scaling study on the other side of the critical line . in section 2 .
we define our renormalization scheme , followed by basic definitions of @xmath0 theory . in section 4 . the numerical method and achieved results
are described followed by a brief summary .
this work is based on the master thesis of the first author at humboldt university , berlin 2013 .
at first glance the title of this subsection might look paradoxical as there is no symmetry breaking in a finite volume .
if we define however an order parameter @xmath4 by the large distance behavior of the z(2 ) symmetric fundamental two point correlation ( @xmath5 is the correlation length ) , @xmath6 then this definition has a smooth thermodynamic limit . to define definite renormalization conditions we employ the fourier transform @xmath7 and extract @xmath8 from @xmath9 where we have assumed a torus of extent @xmath3 in each direction and @xmath10 is the varying part of the correlation ( ` connected ' , although we here avoid the one - point function ) .
we now complete our renormalization scheme by singling out two small torus momenta @xmath11 beside zero momentum .
we match @xmath12 to the form @xmath13 which simultaneously fixes the wave function renormalization factor @xmath14 , a renormalized expectation value @xmath15 and the renormalized mass @xmath16 . by solving these conditions
we obtain @xmath17 and @xmath18 where we have introduced the dimensionless finite size scaling quantities @xmath19 and @xmath20 and the usual lattice momentum @xmath21 it is not difficult to see that in the thermodynamic limit @xmath22 our definitions of @xmath16 and @xmath15 approach those of @xmath23 and @xmath24 in @xcite . apart from this limit however
, each fixed value of @xmath19 defines a different renormalization scheme and the perturbative coefficients of the continuum perturbative callan symanzik @xmath25 function , for instance , will depend on @xmath19 beyond the scheme independent one and two loop terms . as usual in the spontaneously broken theory
we define the renormalized coupling constant in terms of @xmath15 by setting @xmath26
the action in the lattice form is given by @xmath27 - 2 \kappa \sum_{x \mu } \varphi \left ( x \right ) \varphi \left ( x + \hat{\mu } \right)\ ] ] with all dimensionless quantities .
this is equivalent to the field theoretic form @xmath28 with mass dimension one field corresponds to @xmath29 in the lw papers . ]
@xmath30 if we match @xmath31 classically , the symmetric phase arises for @xmath32 where @xmath33 fluctuate around zero with a bare mass given by @xmath34 for @xmath35 the field fluctuates around one of two equivalent nonzero values @xmath36 . the quadratic fluctuations around either constant field are now controlled by the bare mass @xmath37 note that in lw @xmath38 does not appear , as in @xcite it is replaced by @xmath39 while in @xcite the action in terms of @xmath30 is not written and only the @xmath39 for the broken phase appears .
consequently the relations between @xmath39 and @xmath40 differ in the two papers as emphasized in a footnote in @xcite .
following lw we explore the plane of bare parameters by approaching the critical line ( continuum limit ) on trajectories at fixed @xmath41 and define the @xmath25-function @xmath42 this definition entails the following tree level lattice artefact contributions @xmath43 and ( @xmath44 ) @xmath45 although these artefacts are small they may be avoided by switching to modified couplings @xmath46 the perturbative continuum @xmath25 function for the coupling @xmath47 of the previous subsection and also @xmath48 formed from it is given by @xmath49 while the first two coefficients are scheme independent , the three loop term @xmath50 is at the moment not known for our present scheme at finite @xmath19
. the infinite volume case is found in @xcite , @xmath51
the renormalization scheme of section [ scheme ] is defined entirely in terms of the two point correlation .
worm simulations are ideally suited for its numerical computation in the ising limit @xmath52 .
the worm ensemble is given by the partition function @xmath53 .
\label{zens}\ ] ] in this formula we sum over link variables @xmath54 and @xmath55 $ ] enforces the constraint that is most easily described in words : each site except @xmath56 must be surrounded by an even number of @xmath57 links while at @xmath56 ( unless @xmath58 ) this number must be odd .
the fugacity is @xmath59 .
the @xmath60 configurations are in one - to - one correspondence with strong coupling graphs with lines drawn on links with @xmath61 . at the same time we have the connection with the spin formulation @xmath62 where for the ising limit the sum is over @xmath63 and @xmath64
is a normalization factor . in @xcite a lot more details about this reformulation and the efficient simulation of ( [ zens ] ) can be found .
it is obvious now that the two point function is given by @xmath65 with the double angles referring to expectation values with respect to ( [ zens ] ) .
the required fourier transforms can be directly accumulated from @xmath66 for the last step we have used the invariance under individual reflections along each direction .
note that with the small momenta of interest we do not expect very rapid oscillations .
as only ratios of @xmath67 , where the wave function renormalization cancels , are of interest , the denominator @xmath68 ( inverse susceptibility ) is never really needed here .
at first we have explored how @xmath19 depends on the hopping parameter for lattice sizes @xmath69 .
the results are shown in fig.[zkappa ] .
we are here just above the infinite volume critical point which is known @xcite to occur close to @xmath70 .
each data point in the plot corresponds to @xmath71 iterations , where an iteration @xcite consists of one worm move per site . from these results
we have decided to adopt in the following the target value @xmath72 for our study .
our results are summarized in table [ tabg ] .
.simulation results to determine the renormalized coupling in the continuum limit ( growing @xmath1 ) for fixed @xmath72.[tabg ] [ cols="<,<,<,<,<",options="header " , ] each line corresponds to a statistics of @xmath73 iterations . by some tuning we found values of @xmath74 that lead to @xmath72 within errors .
the directly measured couplings ( [ gdef ] ) are given in the fourth column while the rightmost column differs by two tiny corrections . by the first order reweighting technique described in @xcite the value
is adjusted to @xmath72 exactly and then the cutoff correction ( [ gtildedef ] ) is applied .
the first correction is clearly only a change within the error bars , but , although to a much lesser degree than in @xcite , it in addition lowers the statistical error slightly .
these data are plotted in fig.[figg ] .
the dotted ( blue ) , dashed ( red ) and solid ( black ) curves derive from integrating the callan symanzik equation @xmath75 with the continuum @xmath25-function at 1,2,3 loop perturbative precision . beside the universal coefficients ( [ betaf ] ) we here use the infinite volume value ( [ beta2 ] ) for the three loop coefficient .
as discussed before this is only indicative with the presently unknown coefficient for @xmath72 certainly being slightly different .
the experience in the symmetric phase has been , however , that at this size the difference may not be very sizeable .
we have defined a finite size renormalization scheme for @xmath0 theory , which in the infinite volume limit goes over into the one that is standard in the broken phase of the model . in the ising limit of infinite bare coupling ,
we have numerically generated values of the renormalized coupling as a function of the lattice cutoff .
using novel simulation techniques we computed precise values which turn out to closely follow the perturbative renormalization group .
the data points are nicely sandwiched between the one and two loop trajectories .
the three loop curve falls in between and is only about two sigma ( 2 % ) away from our data , although the three loop coefficient is taken for @xmath76 rather than @xmath72 studied here .
if we conclude agreement with perturbation theory in the range studied then this should be even better justified for larger @xmath1 and @xmath48 tends to zero in the continuum limit at a logarithmic rate .
this supports the triviality scenario once more by combining numerical and perturbative methods .
we would like to thank tomasz korzec and peter weisz for very helpful discussions and comments on the manuscript .
10 u. m. heller , m. klomfass , h. neuberger , and p. m. vranas , _ regularization dependence of the higgs mass triviality bound _ , _ nucl.phys.proc.suppl . _ * 30 * ( 1993 ) 685 , [ http://xxx.lanl.gov/abs/hep-lat/9210026 [ ] ] . | we study the standard one - component @xmath0-theory in four dimensions .
a renormalized coupling is defined in a finite size renormalization scheme which becomes the standard scheme of the broken phase for large volumes .
numerical simulations are reported using the worm algorithm in the limit of infinite bare coupling .
the cutoff dependence of the renormalized coupling closely follows the perturbative callan symanzik equation and the triviality scenario is hence further supported
. hu - ep-14/08 sfb / ccp-14 - 18 |
there is solid evidence that most matter in the universe is in the form of non - baryonic dark matter ( dm ) @xcite . from the theoretical point of view
there are several particle physics theories which attempt to explain the yet unknown fundamental nature of dm . in the literature a plethora of dm candidates have been proposed ( see , e.g. , ref .
@xcite ) . depending on their masses and interaction
cross sections with themselves or ordinary matter , they all exhibit a present day abundance in agreement with the dm density determined by planck satellite , @xmath10 @xcite . among all particle physics candidates
the most popular ones belong to the class of weakly interacting massive particles ( wimps ) @xcite , since they are assumed to be stable and to have interactions with the standard model ( sm ) particles , giving a correct relic abundance as observed today .
although the sm describes the elementary particles and their interactions with great success , there are other good reasons , besides the need of a dm candidate , for expecting physics beyond the sm .
one motivation is the so - called _
hierarchy problem_. the mass of the higgs boson acquires large quantum quadratic corrections proportional to the scale where the sm is valid . assuming the sm is valid up to very high energy scales , the parameters in the theory have to be fine - tuned in order to keep the higgs mass at an acceptable value of around 126 gev .
since in the sm a symmetry that relates the various couplings does not exist , this situation is considered to be very unnatural ( e.g. , refs .
one of the best motivated scenarios introduced to solve this problem is supersymmetry ( susy ) , with sparticle masses at the tev scale .
although the first run of the large hadron collider ( lhc ) placed important constraints to light sparticles , and a higgs with 126 gev shifts the scale of susy to larger values requiring a certain amount of tuning ( typically at @xmath11 for the mssm , see e.g. @xcite ) , susy continues being a very attractive possibility .
another interesting feature of susy , mostly related to cosmology and the search for dm , is the existence of a conserved quantum number called r - parity , which assigns at each ( super)partner of the sm particles @xmath12 while each ordinary particle is assigned @xmath13 .
this quantum number implies that supersymmetric particles must be created or destroyed in pairs , and that the lightest supersymmetric particle ( lsp ) is absolutely stable , and hence dm candidates . in many supersymmetric extensions of sm , the lightest neutralino ,
a linear combination of the superpartners of the neutral gauge and higgs bosons , is the favoured dm candidate . with the wimp hypothesis
, the abundance of dm originates from thermal decoupling in the early universe .
when the processes of pair - annihilation and pair - creation of wimps go out of chemical equilibrium due to the hubble expansion , the resulting number density freezes out and remains constant per comoving volume until the present time .
this chemical decoupling , however , does not signal the end of wimp interactions with thermal plasma
. there could still be elastic scattering processes with sm particles , which keep wimps in kinetic equilibrium until later time .
when the rate for elastic scattering processes also falls below the hubble expansion rate , wimps enter the epoch called _ kinetic decoupling_. from this point on ,
wimps are decoupled from the thermal bath , and begin to free - stream . after this stage , first gravitationally bound dm structures begin to form , with the size set by the temperature of kinetic decoupling , related to a small - scale cutoff in the primordial power spectrum of density perturbations .
reference @xcite calculated the the primordial power spectrum by including collisional damping and free - streaming of wimps , and showed that the free - streaming led to a cold dm ( cdm ) power spectrum with a cutoff around a scale corresponding to the earth mass , @xmath210@xmath14 ( see also refs .
@xcite ) .
one of the most challenging goals today is to shed light on the nature of the small - scale cutoff in the primordial power spectrum of density perturbations , often dubbed with the name of _ protohalo_. its properties are relevant for indirect dm searches .
indirect dm detection looks for signatures of dm annihilation , such as gamma - ray photons , from dense celestial environments , where the protohalo mass is a relevant quantity to determine the substructure `` boost '' factor .
direct detection experiments of dm look for energy deposition in underground detectors caused by scattering interactions between target nuclei and wimps around us , giving valuable information about the scattering cross section , and through a correlation that we find in this study , they constrain the mass of the dm protohalos . recently , cornell and profumo @xcite studied scattering cross sections that are relevant for direct detection experiments and protohalo sizes in an mssm context for the neutralino dm .
they based their mssm scan on 9 parameters defined at the electroweak scale .
they found a strong correlation between the kinetic decoupling temperature and the spin - dependent ( sd ) cross section of neutralinos off nucleons . on the contrary
, a weaker correlation was found in the case of the spin - independent ( si ) neutralino - nucleon cross section . in the present paper , we do a forecast on the mass of the protohalos within a supersymmetric scenario by taking into account the latest data from all the relevant particle physics experiments as well as the relic density constraints .
we perform our analyses within a bayesian framework , by adopting 10 mssm fundamental parameters defined at the gauge couplings unification scale , among which 9 of them we allow to vary after requiring the correct electroweak symmetry breaking . in the considered mssm scenario , we find that the kinetic decoupling temperature leads to the protohalo mass most probably residing in a range of @xmath0@xmath15 .
this large variation is due to the range of the kinetic decoupling temperature , @xmath16 , since in the neutralino annihilation processes , both gauge bosons and fermions play a role , and these couplings reveal to be independent from one another .
the range corresponds to two most probable posterior regions : higgsino - like and wino - like neutralinos , for which the most probable neutralino masses are around 1 tev and 3 tev , respectively . in these most probable cases ,
we find that protohalo mass correlates with the both sd and si scattering cross sections .
we show that all higgsino - like neutralino regions , where the probability is higher , such a scattering is dominantly spin - dependent . therefore , any experimental measurement of the sd cross section will imply direct consequences on minimal protohalo mass .
we also show how future direct and indirect detection experiments can play an important role in constraining the ( most probable ) minimal protohalo mass down to @xmath210@xmath17 and the expected value of the boost of the luminosity due to the annihilation of dm in those regions .
complementarity , we study how those predictions change in regions that are disfavored by the posterior probability density function ( pdf ) due to the large tuning , necessary to reproduce the experimental observables ( including @xmath18 ) .
this paper is organized as follows .
we describe the supersymmetric model we adopt in sec .
[ mssm - lhc ] .
the role of the dm protohalo is discussed in sec .
[ high - mass ] : a brief explanation of the smallest dm protohalo mass in sec . [ smallest - mass ] .
the discussion of the most probable regions of the mssm and interactions involved in the annihilation of neutralinos are presented in sec .
[ regions ] .
a profile likelihood map is discussed in sec .
[ low - mass ] .
we comment on the impact of the direct detection experiments on the mass of the protohalo in sec .
[ dd ] , and estimate of the boost of the luminosity due to the annihilation rate in a dm halo with substructures in sec .
we finally give our conclusions in sec .
[ final ] .
despite the expectation around a potential discovery of light susy particles at the first run of the lhc , so far no signal of new physics has been found , which could be considered in tension with the ideas of natural susy . however , the relative large mass of the higgs boson points to a heavier mass spectrum , suggesting that the lack of discovery of sparticles in the first run of the lhc is a consequence of the higgs mass value . in the mssm ,
a lightest higgs boson of around 126gev implies a range of @xmath19 between @xmath210@xmath20 gev and @xmath2 @xmath21 gev , . ] where @xmath19 represents the scale at which susy particles decouple from the sm ( for details , see @xcite ) .
hence , within the mssm framework the higgs mass is in tension with naturalness of the electroweak symmetry breaking , requiring a typical tuning of @xmath22% , see , e.g. , @xcite .
this tension is relaxed going beyond the mssm @xcite .
moreover , since stops give the most important contribution to the higgs mass , the allowed range of @xmath19 could be written as a constraint to the stop sector , where typically stop masses should be larger than @xmath23 tev , unless its mixing parameter reaches its maximal value @xcite , leaving basically the rest of the susy spectrum unconstrained .
on the other hand , one of the beautiful aspects of susy is the apparent unification of gauge couplings in the mssm , because it gives a strong hint in favor of grand unified theories suggesting , as well , that we know how the renormalization group equations ( rge ) behave up to the gauge coupling unification scale , @xmath23 .. a popular approximation is to start the rge running from @xmath24 instead of @xmath25 . for some particular scenarios ,
this approximation is not necessarily correct @xcite .
] taking susy parameters at @xmath24 leads to implicit relations between sparticle masses , in particular the average of stop masses at the scale of 1 tev for @xmath26 , @xmath27 , written as a function of the soft parameters at @xmath24 reads , @xcite : @xmath28 where @xmath29 , @xmath30 and @xmath31 are the bino , wino and gluinos soft mass terms , respectively , @xmath32 and @xmath33 are the third generation of squark soft masses , and @xmath34 is the @xmath35 soft mass .
equation ( [ eq : stopaverage ] ) shows that large stop masses imply large gluino mass ( @xmath36 ) , unless the soft mass terms of the third generation squarks are very large , which leads to a scenario like split susy @xcite .
+ regarding naturalness , the largest tuning required to get the correct electroweak symmetry breaking is applied on the @xmath37 parameter . from the minimization of the higgs potential
one obtains @xmath38 where this expression is valid at a scale of 1 tev for @xmath26 @xcite . as in eq .
( [ eq : stopaverage ] ) , @xmath31 is the responsible for the larger contribution .
the current gluino mass bound from atlas and cms @xcite , @xmath39 tev ( assuming 100% decay to @xmath40 and a mass difference between @xmath41 and @xmath42 of at least 200gev ) , is that more stringent for naturalness . from eqs .
( [ eq : stopaverage ] ) and ( [ eq : mu ] ) , we could also see that naturalness and higgs constraints affect mainly the gluino and squarks sector . on the other hand , sleptons , binos and winos
are basically unconstraint . in a more general framework , where the mssm is parameterized at ew symmetry breaking scale , the pmssm ,
the higgs mass measurements constrain mainly the stop sector , leaving the rest of the spectrum effectively unconstrained . in this case , the main constraints for sparticle masses come from lhc limits and b - physics ( see , e.g. , refs .
+ besides the tuning associated to the ew symmetry breaking , there is also a tuning associated to the requirement of having a good dm candidate .
refs @xcite study the fine tuning required to obtain the correct dm relic density .
in particular , @xcite shows that the region of 1 tev , corresponding to the lightest higgsino - like neutralino , requires very smallest tuning .
typically , regions where the correct annihilation cross section is dominated by resonances or sfermion - neutralino co - annihilations require a large tuning . + to study the mssm parameter space we perform a bayesian analysis .
one of the interesting aspects of this approach is that it is possible to take into account naturalness arguments @xcite .
a fine - tuning associated to the electroweak symmetry breaking is included when we takes the mass of the z boson in the same foot as rest of the experimental data .
effectively , after requiring the correct electroweak symmetry breaking , the posterior pdf appears to be proportional to a term that penalizes regions with a large fine - tuning , independently of the choice of the prior probability .
interestingly , this term is inversely proportional to the barbieri - giudice fine - tuning parameter @xcite .
more specifically the ew fine tuning penalization appears as a jacobian factor that arise from the change of variables @xmath43 evaluated at the measured value of @xmath18 , where @xmath44 and @xmath45 are the gauge and yukawa couplings respectively , @xmath46 is the bilinear higgs coupling , and @xmath37 is the higgs mass term in the superpotential defined at the susy breaking scale .
this jacobian factor is completely independent of the choice of parameters and is not based in a specific definition of fine - tuning . in the same way , a fine - tuning penalization associated to all the other experimental observables is included .
motivated by the fact that this definition does not involve prejudices , ref .
@xcite came up with the idea of using this covariant matrix to penalize regions with large fine - tuning in a @xmath47 analysis . in our analysis
we assume gravity mediated susy breaking and parameterize the mssm with 10 fundamental parameters defined at the unification scale of the gauge couplings as well as sm parameters .
we also assume unification and universality conditions for the squark masses , slepton masses and trilinear terms .
the set of 10 parameters is : @xmath48 where we added , as well , the gauge and yukawa couplings , @xmath44 and @xmath45 , respectively .
@xmath29 , @xmath30 , @xmath31 are the gaugino masses , @xmath49 , @xmath50 and @xmath51 are the soft squark , slepton and higgs masses , @xmath52 and @xmath53 are the squarks and slepton trilinear couplings , @xmath46 is the bilinear higgs coupling , and @xmath37 is the higgs mass term in the superpotential . using a more convenient parameterization , the effective set of parameters reads : @xmath54 where _ s _ stands for sm parameters described in table [ tab : nuis_params ] and , without loss of generality , the sign of @xmath37 is fixed to @xmath55 , allowing @xmath56 to have positive and negative values . in such a way we cover regions with relative phases between @xmath37 and @xmath56 .
let us comment about how strong the predictions of the scenario we consider are with respect to the most general mssm . in our approach
we assume that susy was broken at gauge - coupling unification scale .
although this assumption is reasonable in gravity - mediated susy breaking scenarios , it is not the only possibility ; for example , in gauge mediated scenarios it can happen in principle at any scale .
moreover , the consequence of this assumption depends on the freedom we give to the soft parameters . imposing universality condition ( squark and slepton squared - mass matrices proportional to the @xmath57 identity matrix ) and unification condition
( right sfermion masses equal to left sfermion masses and @xmath58 ) implies a specific mass hierarchy for squarks and sleptons which could be ameliorated if the susy breaking scale is smaller .
hence , the regions of parameters we are missing in using this parameterization of the mssm are the ones with any possible hierarchies of sfermions masses . in our case
, @xmath59 is always the lightest stop and @xmath60 the lightest slepton . on the other hand ,
the universality condition is supported by the strong constraints from fcnc process . using a more general parameterization at ew symmetry breaking scale ,
the pmssm , the sparticle masses do not feel the impact of the renormalization group equations , the correlation between the parameters disappear and the choice of the prior will most likely dominates the results not allowing us to make conclusions about the most probable region .
on the other hand , bayesian analysis has been performed in the pmssm from different perspectives , to be able to identify which are the parameters that are directly constrained by the experimental information , that can be checked by looking at the prior dependency in each parameter ( see , e.g. , refs .
@xcite ) .
[ cols="<,<,<,<",options="header " , ] figure [ fig : binon1 ] shows points that reproduce the experimental data within @xmath61 confidence level .
we show the lightest neutralino mass as a function of the kinetic decoupling temperature , @xmath16 ( top ) , and the protohalo mass , @xmath62 ( bottom ) .
let us describe the mass spectrum .
the characteristics of the electroweakino sector are set mainly by the fact that an efficient neutralino annihilation is needed to reproduce the correct relic density . to identify regions where the lightest neutralino co - annihilates with sfermions in the early universe , in both left panels we highlight points that satisfy a criterion based on the mass difference between the lightest neutralino and the lightest stau ( green points ) , and between the lightest neutralino and the lightest stops ( blue points ) .
to select those points we have required a maximal relative mass difference , @xmath63 , of 5% and a maximal absolute mass difference of 5 gev which are imposed for neutralino masses above and below 100 gev , respectively .
the gray band of the top - right panel shows the range of temperatures where the qcd phase transition occurs , from the critical temperature to four times this one , where the value of @xmath16 represents an upper bound .
those points with a @xmath16 around the qcd phase transition are represented with lighter colors in the @xmath64@xmath62 plane in the bottom - left panel , where in this case the value of @xmath62 represents a lower bound .
( top panel ) and @xmath62 for points that reproduce all the experimental observables within @xmath61 confidence level.,title="fig : " ] ( top panel ) and @xmath62 for points that reproduce all the experimental observables within @xmath61 confidence level.,title="fig : " ] + ( top panel ) and @xmath62 for points that reproduce all the experimental observables within @xmath61 confidence level.,title="fig : " ] ( top panel ) and @xmath62 for points that reproduce all the experimental observables within @xmath61 confidence level.,title="fig : " ] most of the points showed in fig .
[ fig : binon1 ] have a neutralino quasi - degenerate with another sparticle .
higgsino - like and wino - like lightest neutralinos are quasi - degenerated with the lightest chargino , guaranteeing both a very efficient annihilation of neutralinos and co - annihilation with charginos , and selecting rather heavy neutralino masses .
neutralinos with a dominant higgsino or wino component cover the mass region of @xmath65 tev . as we commented in the previous section , for pure - higgsino and pure - wino neutralino the relic density constraint fixes the mass to @xmath21 tev and @xmath23 tev , respectively . as a result of our scan
we have identified different mixed states lying between these regions : higgsino - wino neutralinos , higgsino - like and wino - like neutralino that co - annihilate with staus or stops , and wino - bino neutralinos and higgsino - like neutralino with a mass equal to half of the mass of the pseudoscalar .
some of the points with @xmath64 slightly below 1 tev are higgsino - bino neutralino .
this region is strongly constrained by direct detection experiments like xenon100 and lux .
however , there are some blind spots for @xmath37 and @xmath29 with opposite relative sign , as explained in detail in refs . @xcite .
points with @xmath66 tev have a lightest bino - like neutralino .
for @xmath67 gev it is possible to distinguish two groups of points in the bottom - left panel of fig .
[ fig : binon1 ] .
the first group has smaller @xmath68 , ranging from @xmath210 mev to @xmath2100 mev and is basically aligned to the stau co - anhihilation region .
for these points sleptons are light , and the correct neutralino abundance was reached by slepton - neutralino co - annihilation in the early universe .
charginos and heavier neutralinos are typically much heavier . for the second group of points with larger @xmath68 varying from @xmath2100 mev to @xmath23 gev we checked that the lightest ( bino - like ) neutralino is quasi degenerated with both the lightest wino - like chargino and the second lightest neutralino , guaranteeing the neutralino annihilation .
top - right panel of fig .
[ fig : binon1 ] shows that these two regions are not completely disconnected .
for example , for @xmath69 ( meaning @xmath70 ) sleptons also play a role in the annihilation processes .
the region @xmath71 tev has similar characteristics , but in this case the two regions , that one with light sfermions and the other one with light chargino , have a large overlap for @xmath72 mev .
last but not least , we find that there are very few points for the higgs and z resonance regions .
these two regions require a very large tuning , and therefore , they are very difficult to explore when requiring boundary conditions at gut scale . to understand the dominant process of neutralino - sm scattering in the regions we described above , in the top - right panel of fig .
[ fig : binon1 ] we show the relative mass difference between the lightest first and second generation of sleptons and the lightest neutralino , @xmath73 , while in the bottom - right panel we show the gaugino fraction .. the gaugino fraction is defined by @xmath74 ( see @xcite for details ) . ]
these plots show , for all gaugino - like neutralinos ( bino - like or wino - like ) , a clear correlation between the lightest neutralino mass and the kinetic decoupling temperature for a fixed value of @xmath73 .
higgsino - like neutralinos around 1tev do not show a correlation for specific sleptons masses , as we comment in the previous section ; its interaction with sfermions is proportional to the yukawa coupling , and it is therefore negligible for the first and the second generation of sleptons . in the higgsino - like case
the dominant interaction is the one mediated by the z - boson , as in the case of bino - like and wino - like neutralinos when sfermions are decoupled .
+ as we commented in section [ mssm - lhc ] , we assume universality and unification of squarks and slepton masses .
these conditions imply that the @xmath59 and @xmath60 are the lightest squark and slepton , respectively , which is the reason why we only find neutralino - stop and neutralino - stau co - annihilation regions in our analysis . in more general scenarios where sfermions masses do not unify ,
the possibility of having co - annihilation with any sfermion is open , since any of them could be the next - to - lsp .
if the lightest neutralino is bino - like and the first or second generation sfermions are close enough in mass to the lightest neutralino to guarantee a large enough effective annihilation in the early universe then the dominant interaction in the scattering between the lightest neutralino and the sm particles will be the same interaction ( neutralino - fermion - sfermion ) , producing strong correlation between the mass of the lightest neutralino and @xmath62 .
another consequence of universality and unification is that the first and second generation of squarks are in general very heavy ( due to the higgs mass constraint to the stop sector ) , having , in most of the cases , a negligible contribution to the neutralino annihilation and neutralino scattering with the sm particles in the early universe . without this assumption
the most important constraint to squark masses will come from lhc bounds and direct detection experiments , allowing smaller masses . due to the strong lower bounds on first and second generation squarks masses coming from lhc @xcite
, one will expect that sleptons will still give the dominant contribution to the neutralino annihilation and neutralino scattering with the sm particles in the early universe for neutralinos lighter than 300 gev .
however , for neutralino masses larger than 300 gev , contributions from first and second generation squarks could be sizeable .
interestingly , the cases that set the smaller value of @xmath62 , when sleptons are very close in mass to the mass of the lightest neutralino , and larger value of @xmath62 , when sleptons are decoupled and the scattering is mediated by z - boson , are covered in our analysis . on the other hand ,
the consequence on lhc , direct detection and indirect detection could be different , as we will discuss in the next section .
the understanding of the interactions that play a relevant role in the annihilation and scattering of neutralinos with sm particles helps us identify correlations between @xmath62 and the susy spectrum .
these correlations could be very helpful for constraining @xmath16 indirectly from current dm experiments .
in particular , the region of @xmath75 gev could be potentially tested by the lhc , as commented in appendix [ sec : lhc ] .
direct detection experiments of dm look for energy deposition in the underground detector caused by scattering interactions between target nuclei and wimps around us .
the measurement or the bound on this cross section has direct consequences on the value of the @xmath16 , assuming that the processes involved in the last scattering are the same as the ones mediated the scattering of the dm with the detectors .
reference @xcite analyzed correlations between the mass of protohalos , @xmath62 ( as well as the temperature of kinetic decoupling , @xmath16 ) and the sd and si scattering cross sections . such a correlation appears when the mass of squarks is assumed to be large ( @xmath76@xmath77tev ) , and the dominant process for the scattering is mediated by a z boson . and @xmath78 , respectively.[sdsi],title="fig : " ] and @xmath78 , respectively.[sdsi],title="fig : " ] in fig .
[ sdsi ] , we show the most probable region on the plane of the protohalo mass , @xmath62 , and the sd and the si cross sections computed at tree level .
contrary to fig .
[ tm ] , the probability regions do not have disconnected parts , but they include both higgsino - like ( at @xmath21 tev ) and wino - like ( at @xmath23 tev ) neutralinos . in both cases the dominant scattering process is mediated by the z - boson .
we see how the expected improvement on the si sensitivity by , e.g. , xenon1 t @xcite and lux - zeppelin experiment ( lz ) @xcite , will reduce the most probable range of the minimal subhalo mass down to below @xmath210@xmath79 , while the expected sd sensitivity provides weaker constraints .
since in the analysis we have included the xenon100 limits as constraints on the wimp - nucleon scattering cross section , we see in the right panel of fig . [ sdsi ] that the region around @xmath80 @xmath81 is strongly penalized .
the current lux bound is more stringent on the spin - independent sensitivity , giving an upper bound of @xmath82 for a 1 tev neutralino @xcite , although we did not include it in our analysis .
including the lux bound , therefore , would affect the very right part of the right panel of fig .
[ sdsi ] ( and also fig .
[ fig : scatter_sigmasi ] shown below ) .
however , since the regions affected are tiny , it would not affect our conclusions .
confidence level in the sd cross section @xmath83 versus protohalo mass @xmath62 plane .
the neutralino mass is indicated with colors , as shown in the color bar .
the three panels separate the points in three groups : light squarks and sleptons ( left panel ) , light sleptons and decoupled squarks ( central panel ) , and decoupled squarks and sleptons ( right panel ) .
[ fig : scatter_sigmasd ] ] figure [ fig : scatter_sigmasd ] shows points that reproduce the experimental constraint at 2 @xmath84 confidence level for the minimal protohalo mass versus the tree level sd cross section plane .
the right panel shows the case where the lightest first or second generation of sfermions is at least nine times heavier that the lightest neutralino , @xmath85 .
the thin yellow line corresponds to @xmath21 tev higgsino - like neutralino , while the thin red line to @xmath23 tev wino - like neutralino . in these two cases
the z - boson mediates both scattering processes .
the rest of the points correspond to the bino - like neutralino where , instead of a line , we get scattered points with 100 gev @xmath86 tev .
we remind that for the bino - like case the annihilation cross section and , therefore , the relic density can be adjusted varying the neutralino mass and its mass splitting with the lightest ( wino - like ) chargino . on the other hand ,
even if it is ten times heavier than the lightest neutralino , sleptons mediate the dominant scattering processes that set @xmath16 for most of the points .
the size of the contribution of processes , mediated by the z - boson , depends on how large the higgsino component of the neutralino is .
however , the higgsino component of a bino - like neutralino is highly constrained by si cross sections bounds .
nevertheless , as we comment in subsection [ low - mass ] , there are some blind spots for si cross sections . for those points
the z - boson gives an important contribution to the scattering cross section .
regarding @xmath87 for the bino - like region , the dominant process is mediated by the z - boson .
implies that squarks are typically much heavier than ten times the mass of the lightest neutralino . ] besides the dominant scattering processes for @xmath16 and @xmath87 are different , there is an apparent correlation between the two quantities for a fixed neutralino mass .
we have checked the behavior of the correlation for a specific values of @xmath64 , finding that @xmath16 spreads around one order for a given value of @xmath87 .
another characteristic of the bino - like case , assuming sfermions are heavy , is that the minimal protohalo mass is the one allowed by the free - streaming . on the contrary , for the higgsino - like and wino - like cases , the minimal protohalo mass is the one allowed by the damping scale set by acoustic oscillation .
central panel of fig .
[ fig : scatter_sigmasd ] shows the case where the lightest slepton has a mass smaller than @xmath210 times the lightest neutralino mass .
as expected , the wino - like and bino - like regions spread to larger protohalo masses .
vertex to yield the helicity flip , which is yukawa suppressed . for a review , see ref @xcite . ]
left panel of fig .
[ fig : scatter_sigmasd ] shows the case where the lightest sleptons and squarks are smaller than @xmath210 times the lightest neutralino .
here , squarks are light enough to give important contributions to the scattering with the nucleus , spreading the points to larger values of @xmath88 . for the the si cross section , @xmath89 .
[ fig : scatter_sigmasi ] ] figure [ fig : scatter_sigmasi ] shows points in the minimal protohalo mass versus tree level si cross section plane .
the main contribution to the si cross section comes from the higgs exchange , requiring a non - negligible higgino and wino / bino coupling ( since higgs couplings through neutralinos are @xmath90 and @xmath91 ) . on the other hand , the total neutralino - sm scattering , and therefore @xmath16 and @xmath62 , are dominated by sd interactions .
a a consequence , right panel of fig .
[ fig : scatter_sigmasi ] shows the correlation between the @xmath92 and @xmath93 , for the higgsino and wino case .
central and left panels show the effect of sleptons and squarks in the scattering processes .
figures [ fig : scatter_sigmasd ] and [ fig : scatter_sigmasi ] show the expected sensitivity by xenon1 t and lz assuming the neutralino mass is @xmath21 tev . for a neutralino of @xmath2100 gev ,
the expected sensitivity is around one order of magnitude stronger . as we commented in the previous section , assuming universality and unification of the squark masses and slepton masses , we impose a particular mass hierarchy : @xmath59 is the lightest squark and @xmath60 is the lightest slepton . without this assumption , first and second generation of sfermions
can be lighter and change the phenomenology for direct detection experiments and colliders . in the case
that the lightest neutralino is gaugino - like and the first and second generation of squarks are the lighter sfermions , @xmath16 will be completely correlated with the neutralino - nucleon scattering cross section . still , if they are not the lightest ones but they are significantly lighter than in our analysis the neutralino - nucleon scattering cross section could increase , getting values close to the actual limits .
in addition , if they are lighter or close in mass to the first and second generation sleptons , the scattering of the neutralino with sm particles in the early universe could also increase .
those points will most likely populate the top - right corner of the left panel of figs .
[ fig : scatter_sigmasd ] and [ fig : scatter_sigmasi ] . notice that , as we mentioned above , for the computation of si cross sections we have adopted @xmath94 for the contribution of the strange quark to the nucleon form factors @xcite , derived experimentally from measurements of the pion - nucleon sigma term .
however this value is considerably larger than determinations obtained from lattice qcd , @xmath95 @xcite .
the discrepacy between the two values and its impact in the si cross section is studied in more detail in refs .
@xcite .
finally , we remind that the scattering cross - sections considered in this work were computed at tree level . in the cases
where neutralino approaches to a pure state ( bino , wino or higgsino ) , this approximation may not give a reliable result .
in particular , in the case of the wino - neutralino , one loop corrections give the dominant contributions ( see , e.g. ref .
one of the most reliable methods to model the non - linear evolution of dm is numerical simulation , although it is limited by mass resolution .
in fact , the minimum self - bound mass ( @xmath96 ) of dm halos is expected to be many orders of magnitude below the resolution of current simulations . through numerical simulations such as _ acquarius _
@xcite , we can obtain information on the subhalo hierarchy , although its resolution mass limit @xmath210@xmath97 or @xmath210@xmath98 is far from the predicted protohalo mass shown in sec .
[ high - mass ] .
here we investigate the impact of different values of @xmath96 on the gamma - ray luminosity due to dm annihilation , and compute a boost factor of a given halo of mass @xmath99 due to the substructure inside it , by integrating the subhalo annihilation luminosities from the protohalo mass we have found , @xmath62 , up to the mass of sizable fraction of the host halo @xmath100 .
the total luminosity of the dm halo due to annihilation is proportional to : @xmath101 where @xmath102 is the subhalo mass function , i.e. the subhalo number density per unit mass range .
numerical simulations find that the differential subhalo mass function follows a power law @xmath103 m@xmath104 , with @xmath105 or @xmath106 ( see , e.g. @xcite ) .
we adopt a @xmath107 subhalo mass spectrum as our fiducial subhalo model .
we assume that each individual dm subhalo is described by a navarro - frenk - white ( nfw ) density profile @xcite : @xmath108 where @xmath109 and @xmath110 are the characteristic density and radius , respectively .
@xmath111 is defined as the luminosity of each subhalo in the host halo , which depends on the volume integral of the subhalo density squared , and is given by : @xmath112 following the same approach of ref .
@xcite , we parameterize the scaling relation between the gamma - ray luminosity and subhalo mass as : @xmath113 where above the simulation resolution of @xmath210@xmath114 , the luminosity versus subhalos mass scales as @xmath115 , while below the resolution we assume @xmath116 . here
@xmath117 encodes all the particle physics , i.e. , @xmath118 , where @xmath6 is the velocity - averaged annihilation cross section times the relative velocity . , is almost independent of velocity , @xmath119 . ] in order to obtain the scaling behavior of @xmath120 , we adopted scaling relations among several quantities found in the _ aquarius _ numerical simulation .
since each subhalo is described by a nfw density profile , we related the maximum rotation velocity of the subhalo , @xmath121 , and the radius at which the rotation curve reaches this maximum , @xmath122 , with the characteristic density and radius , @xmath109 and @xmath110 , to obtain them as a function of the subhalo mass @xmath99 . these empirical relations between ( @xmath121 , @xmath122 ) and ( @xmath109 , @xmath110 ) , however , lose validity in mass regions below the resolution limit of the simulation .
for this reason we split eq .
( [ cum ] ) in two terms , above and below the resolution ( 10@xmath123 ) , where in the latter we put @xmath124 as a phenomenological parameter describing the scaling behavior .
the luminosity in eq .
( [ lum ] ) can be then written as : @xmath125 \ , .
\label{luminosity}\ ] ] assuming that the first term dominates , the luminosity is , thus , a function of the protohalo mass : @xmath126 for comparison , we define a reference value for such a luminosity , @xmath127 , as : @xmath128 for values of these reference parameters , we adopt @xmath129 , @xmath130 , and @xmath131 . , versus the velocity - averaged annihilation cross section times the relative velocity , @xmath6 ( left panel ) , and for @xmath132 , obtained by using @xmath131 , versus the dark matter particle mass , @xmath64 ( right panel ) .
for both panels , the region with higher probability density corresponds to a higgsino dm candidate ; in the second region the dm candidate is a wino .
left panel shows that , in the higgsino case , the protohalo mass @xmath62 is lower than the reference one , while @xmath6 does not deviate from @xmath133 .
right panel shows that , in the wino case , the protohalo mass @xmath62 is even lighter and @xmath6 is two orders of magnitude larger than @xmath134 ; thus , there is a substantial enhancement of @xmath132.[sb],title="fig : " ] , versus the velocity - averaged annihilation cross section times the relative velocity , @xmath6 ( left panel ) , and for @xmath132 , obtained by using @xmath131 , versus the dark matter particle mass , @xmath64 ( right panel ) . for both panels
, the region with higher probability density corresponds to a higgsino dm candidate ; in the second region the dm candidate is a wino .
left panel shows that , in the higgsino case , the protohalo mass @xmath62 is lower than the reference one , while @xmath6 does not deviate from @xmath133 .
right panel shows that , in the wino case , the protohalo mass @xmath62 is even lighter and @xmath6 is two orders of magnitude larger than @xmath134 ; thus , there is a substantial enhancement of @xmath132.[sb],title="fig : " ] the left panel of fig .
[ sb ] shows the two - dimensional joint posterior pdf for the protohalo mass @xmath62 and @xmath6 with 68% and 95% credible contours .
these most probable regions fall in a mass range between @xmath135 and @xmath136 , and @xmath137@xmath138 @xmath139 s@xmath9 .
the region with higher probability density again corresponds to a higgsino dm candidate with the annihilation cross section close to the canonical value @xmath140@xmath8s@xmath9 , while the second region corresponds to a dm wino candidate with much larger annihilation cross section @xmath210@xmath141@xmath8s@xmath9 . in the right panel
we show the ratio of the luminosity over the reference one @xmath142 , versus the dm mass , @xmath64 .
we also analyzed the change in the boost by varying the @xmath124-parameter in a range between 0.5 and 0.9 , we only show the case @xmath131 , and found that @xmath132 always got largely boosted by decreasing @xmath124 .
this behavior depends on the normalization made on the protohalo mass , @xmath62 , since it has been normalized to the limit of the numerical simulation ( @xmath143 ) . , for points that reproduce all the experimental observables within @xmath61 confidence level .
left panel shows points which refer to a bino - fraction ( @xmath144 ) larger than 0.8 .
right panel shows point with a bino - fraction smaller than 0.8,title="fig : " ] , for points that reproduce all the experimental observables within @xmath61 confidence level .
left panel shows points which refer to a bino - fraction ( @xmath144 ) larger than 0.8 .
right panel shows point with a bino - fraction smaller than 0.8,title="fig : " ] figure [ fig : scatter_boost ] shows the boost factor , @xmath145 , for points that reproduce all the experimental observables within @xmath61 confidence level .
right panel shows points which refer to a higgisino - like and wino - like neutralinos , while the left panel shows points where the neutralino is mostly bino - like .
bino - like neutralinos have very small @xmath6 in the limit of zero velocity
. co - annihilations , which play a very important role in the efficient annihilation in the early universe , are not present anymore ; this is the reason for which we have a very small boost of the luminosity .
finally we comment that although not included in this work , fermi and hess bounds on @xmath64@xmath6 plane strongly constrain the wino - like region , excluding the region around 2.4 tev ; see refs .
in this work we have studied how the kinetic decoupling of dark matter could improve our knowledge of the properties of the dark matter protohalos within a supersymmetric model , i.e. the minimal supersymmetric extension of the standard model .
such a model is the well - motivated extension of the standard model at the electroweak scale . at first
, it was introduced to solve the hierarchy problem of the standard model , but it revealed to have many other interesting characteristics . in particular , it contains a tempting particle dark matter candidate , the lightest neutralino .
if such a neutralino is the lightest supersymmetric particle and the quantum number r - parity is preserved , it is stable , yielding to a thermal abundance as that indicated by the observed dark matter density . in our analysis
we do a forecast on the mass of the protohalos within a supersymmetric framework realized with 9 independent parameters .
we performed any analysis in the light of the latest data coming from particle physics experiments , as well as the relic density constraints . among them ,
the most important observables involved in the analysis , which give a relevant impact on our results , are the mass of the higgs and the relic density . 1 . the kinetic - decoupling temperature and ,
thus , the minimal protohalo mass result to be not well constrained for wimps , since the interactions involved in the annihilation of neutralinos , that are constrained by the relic density , are not necessarily those which participate in the scattering of neutralinos with first and second generation of fermions . in a supersymmetric framework ,
the minimal protohalo mass is typically @xmath146 , assuming a bino - neutralino annihilating through sfermions with a mass of around twice the neutralino mass .
this resulted in a possible option to get a well tempered neutralino .
in addition , this possibility has been well motivated by constrained scenarios like cmssm , affirming that when the neutralino is mostly bino , it efficiently annihilates through sfermions in the early universe , giving the correct relic density . nevertheless , it was in tension with the experimental data within the cmssm , especially after the first run of the lhc , where a considerable part of this region was excluded .
2 . using a bayesian framework
, we showed that the most probable neutralino mass regions satisfying both the higgs mass and the relic density contraints , are those with the lightest supersymmetric neutralino mass around 1 tev and 3 tev , that correspond to higgsino - like and wino - like neutralino , respectively .
we mentioned that , concerning the higgsino - like neutralino , the annihilation cross section is driven by its higgsino component , while for wino - like neutralino , the annihilation cross section is mainly driven by its wino component .
we also discussed that the part of the region closer to @xmath2 2.4 tev gets important contributions from the neutralino - stau co - annihilation , reducing both the effective annihilation rate of neutralinos in the early universe and the value of the neutralino mass , in order to obtain the correct relic density .
we commented that in the case of wino - like or higgsino - like neutralinos the annihilation products are gauge bosons , whose interactions involve different couplings with respect to the ones of the neutralino - fermion scattering . for that reason kinetic decoupling temperature , @xmath16 , exhibits a considerable range of variation , that reflects , in turn , to a protohalo mass range of @xmath0@xmath147 .
we also discussed the bino like neutralino with masses smaller than @xmath21 tev , where a quasi - degenerated sfermion or chargino , or a light sfermion are necessary to get the correct dark matter abundance .
sleptons give the most important contribution for the kinetic decoupling temperature and therefore to the protohalo mass , setting the range @xmath148@xmath149 .
5 . kinetic decoupling of dark matter , involving elastic scattering of a dark matter particle with standard model particles in the early universe , reveals a relevant process for dark matter direct detection searches . in our analysis
, we showed that the regions where the probability is higher the correlation between the protohalo mass and experimental signatures permits to put constraints on the protohalo mass . we depicted how improvements on the spin - independent sensitivity might reduce the most probable range of the protohalo mass between @xmath210@xmath3 and @xmath210@xmath150 , while constraints associated to the expected spin - dependent sensitivity are weaker . to give this conclusion we computed scattering cross sections at tree - level .
however , specially in the wino - like neutralino case , loop corrections should be considered since the tree level coupling vanishes when approaching the pure wino case .
we discussed , as well , how the interplay among both spin - dependent and spin - independent scattering procesess , strongly depends on the neutralino composition .
for both higgsino - like and wino - like cases , the spin - dependent scattering between higgsino and fermions is mediated dominantly by the z boson at tree level , while for the spin - independent scattering , the interaction is mediated by the higgs boson . regarding the higgsino neutralino , we commented that the spin - independent interaction gives a nonzero tree - level contribution as long as gauginos are not decoupled , a non - negligible bino or wino component is necessary to have a non - negligible coupling with the higgs . on the other hand , for the wino - like neutralino
the requirement of a non - negligible component of higgsino is indispensable to have a tree - level contribution to both scattering processes if sfermions are decoupled .
depending on the nature of neutralino , the value of the annihilation cross section , @xmath151 , changes by different orders of magnitude .
we presented that the annihilation cross section , @xmath6 , in the higgsino case does not deviate from the canonical cross section , @xmath152 @xmath8 s@xmath9 . on the other hand , in the wino case non - pertubative effect
is important , @xmath6 increases up to two orders of magnitude . and
it is much smaller in the bino - like case , where co - annihilations with sfermions played a crucial role to fix the correct abundance .
another way to look for dark matter is through indirect detection methods , which consist to detect , _ indirectly _ , the lightest supersymmetric particle through annihilation processes where standard model particles , including gamma - ray photons , are produced .
since the luminosity of each subhalo in the host halo due to the dark matter annihilation processes depends on the volume integral of the subhalo density squared , smaller and denser substructures provide an enhancement of the luminosity . in this work
, we showed for both neutralino higgsino - like and wino - like cases how the boost of the luminosity due to dark matter annihilation increases , depending on the protohalo mass .
we discussed that in the higgsino case , there is no a significant enhancement of the luminosity : the protohalo mass is lower than the standard value often used in the literature of @xmath210@xmath5 , while @xmath6 does not deviate from @xmath7 @xmath8 s@xmath9 . in the wino case ,
a substantial enhancement of the luminosity is seen : the protohalo mass reaches lighter values , and @xmath6 is two orders of magnitude larger .
the work was supported partly by nwo through vidi grant ( s.a . ) , by university of amsterdam ( r.d . and s.a . ) and
by fundao de amparo pesquisa do estado de so paulo ( m.e.c.c . ) .
r.d . kindly acknowledges francesca calore and sergio palomares ruiz for interesting and stimulating discussions .
also thanks miguel nebot and sebastian liem for useful comments .
m.e.c.c thanks andre lessa and boris panes for useful discussions .
as commented in sec . [ low - mass ] , points with @xmath64 smaller than @xmath21 tev are bino - like and require a light enough next - to - lightest sparticle , in order to guarantee an efficient annihilation in the early universe .
based on the characteristics of the next - to - lightest sparticle , we are going to comment the potential lhc signatures . for neutralinos lighter that
500 gev there are two regions , in addition to z / h / a resonances .
the first one has @xmath153 close in mass to @xmath42 . a light wino - like chargino which annihilates and co - annihilates in the early universe
is required , and is represented by points with @xmath154 in the left panel of fig .
[ fig : n2decay ] . in this region @xmath42 is dominantly bino and @xmath153 and @xmath155 are dominantly winos . the bino and wino mass , @xmath29 and @xmath30 , are close to the values where the tree level decay of @xmath155 to @xmath156 is suppressed , and the branching ratio to @xmath157 acquires a large value , as discussed in detail in refs .
right panel of fig .
[ fig : n2decay ] shows that some of the points can have a dominant @xmath158 decay , giving a characteristic signature at collider .
moreover , the decay channel @xmath159 becomes relevant .
although the photon produced in the @xmath155 and @xmath160 decays is very soft , it could give a clear signature at collider in the boosted regime .
keep in mind that a potential measurement of sleptons will directly constraint the prediction for the protohalo mass for bino - like neutralino .
mass and @xmath155 mass plane .
the colors show the mass of the lightest slepton ( @xmath161 ) .
right panel shows the branching ratio of @xmath155 to photons as a function of @xmath155 mass for points with @xmath162 .
red , blue and green points correspond to @xmath1631tev , 500gev@xmath1641 tev and @xmath165500gev respectively.,title="fig : " ] mass and @xmath155 mass plane .
the colors show the mass of the lightest slepton ( @xmath161 ) .
right panel shows the branching ratio of @xmath155 to photons as a function of @xmath155 mass for points with @xmath162 .
red , blue and green points correspond to @xmath1631tev , 500gev@xmath1641 tev and @xmath165500gev respectively.,title="fig : " ] the second region corresponds to stau co - annihilation , where @xmath60 and @xmath42 are very close in mass . in the left panel of fig .
[ fig : n2decay ] the points outside 5 gev @xmath166 40 gev correspond to this region .
notice that , as a consequence of universality conditions of slepton soft masses , the first and second generation of sleptons is relatively close in mass to the lightest stau and , therefore , to the lightest neutralino .
the authors of ref .
@xcite discuss the status of this region after the first run of the lhc in the framework of the constrained mssm ( cmssm ) , and project the likely sensitivity of the lhc searches in run 2 at 14 tev center of mass energy and 300/fb of integrated luminosity , concluding that the entirely cmssm co - annihilation strip will be tested . for @xmath167 500 gev
new regions arise .
stop co - annihilations , and neutralino annihilations are mediated by sfermions . in this neutralino mass range
the production of colored particles is the most promising . in refs .
@xcite , is studied the stop co - annihilation region , not only by direct stop production but also by gluino production , where direct stop productions constraint light stops ( @xmath168 gev ) ; for heavier stops gluino , the production seems to be more promising . on the other hand , the region where neutralino annihilation is mediated by squarks
is directly constrained by limits on squarks masses . | we study how the kinetic decoupling of dark matter within a minimal supersymmetric extension of the standard model , by adopting nine independent parameters ( mssm-9 ) , could improve our knowledge of the properties of the dark matter protohalos .
we show that the most probable neutralino mass regions , which satisfy the relic density and the higgs mass contraints , are those with the lightest supersymmetric neutralino mass around 1tev and 3tev , corresponding to higgsino - like and wino - like neutralino , respectively .
the kinetic decoupling temperature in the mssm-9 scenario leads to a most probable protohalo mass in a range of @xmath0@xmath1 .
the part of the region closer to @xmath22 tev gives also important contributions from the neutralino - stau co - annihilation , reducing the effective annihilation rate in the early universe .
we also study how the size of the smallest dark matter substructures correlates to experimental signatures , such as the spin - dependent and spin - independent scattering cross sections , relevant for direct detection of dark matter .
improvements on the spin - independent sensitivity might reduce the most probable range of the protohalo mass between @xmath210@xmath3 and @xmath210@xmath4 , while the expected spin - dependent sensitivity provides weaker constraints .
we show how the boost of the luminosity due to dark matter annihilation increases , depending on the protohalo mass . in the higgsino case ,
the protohalo mass is lower than the canonical value often used in the literature ( @xmath210@xmath5 ) , while @xmath6 does not deviate from @xmath7@xmath8s@xmath9 ; there is no significant enhancement of the luminosity . on the contrary , in the wino case ,
the protohalo mass is even lighter , and @xmath6 is two orders of magnitude larger ; as its consequence , we see a substantial enhancement of the luminosity . |
in their classic 1935 paper , erds and szekeres @xcite proved that , for every integer @xmath5 , there is a minimal integer @xmath0 , such that any set of @xmath0 points in the plane in general position contains @xmath1 points in convex position , that is , they are the vertices of a convex @xmath1-gon .
erds and szekeres gave two proofs of the existence of @xmath0 .
their first proof used a quantitative version of ramsey s theorem , which gave a very poor upper bound for @xmath0 .
the second proof was more geometric and showed that @xmath6 ( see theorem [ cupscaps ] in the next section ) .
on the other hand , they showed that @xmath3 and conjectured this to be sharp @xcite .
small improvements have been made on the upper bound @xmath7 by various researchers @xcite , but no improvement in the order of magnitude has ever been made .
the most recent upper bound , due to norin and yuditsky @xcite and mojarrad and vlachos @xcite , says that @xmath8 in the present paper , we prove the following .
[ main ] for all @xmath9 , where @xmath10 is a large absolute constant , @xmath11 the study of @xmath0 and its variants has generated a lot of research over the past several decades . for a more thorough history on the subject
, we refer the interested reader to @xcite .
all logarithms are to base 2 .
in this section , we recall several results that will be used in the proof of theorem [ main ] .
we start with the following simple lemma . [
four ] let @xmath12 be a finite point set in the plane in general position such that every four members in @xmath12 are in convex position .
then @xmath12 is in convex position .
the next theorem is a well - known result from @xcite , which is often referred to as the erds - szekeres cups - caps theorem .
let @xmath12 be a @xmath13-element point set in the plane in general position .
we say that @xmath12 forms a _
( _ @xmath13-cap _ ) if @xmath12 is in convex position and its convex hull is bounded above ( below ) by a single edge .
in other words , @xmath12 is a cup ( cap ) if and only if for every point @xmath14 , there is a line @xmath15 passing through it such that all of the other points in @xmath12 lie on or above ( below ) @xmath15 .
see figure [ cupcappic ] .
[ cupscaps ] let @xmath16 be the smallest integer @xmath17 such that any @xmath17-element planar point set in the plane in general position contains a @xmath13-cup or an @xmath18-cap .
then @xmath19 the next theorem is a combinatorial reformulation of theorem [ cupscaps ] observed by hubard et al .
@xcite ( see also @xcite ) .
a transitive 2-coloring of the triples of @xmath20 is a 2-coloring , say with colors red and blue , such that , for @xmath21 , if triples @xmath22 and @xmath23 are red ( blue ) , then @xmath24 and @xmath25 are also red ( blue ) .
[ comcupcap ] let @xmath26 denote the minimum integer @xmath17 such that , for every transitive 2-coloring on the triples of @xmath20 , there exists a red clique of size @xmath13 or a blue clique of size @xmath18 .
then @xmath27 the next theorem is due to pr and valtr @xcite , and is often referred to as the positive - fraction erds - szekeres theorem ( see also @xcite ) . given a @xmath13-cap ( @xmath13-cup ) @xmath28 , where the points appear in order from left to right , we define the _ support of _ @xmath12 to be the collection of open regions @xmath29 , where @xmath30 is the region outside of @xmath31 bounded by the segment @xmath32 and by the lines @xmath33 , @xmath34 ( where @xmath35 , @xmath36 , etc . ) .
see figure [ tbase ] . in the support of @xmath37 , and segment @xmath38 .
] [ partition ] let @xmath39 and let @xmath40 be a finite point set in the plane in general position such that @xmath41 .
then there is a @xmath13-element subset @xmath42 such that @xmath12 is either a @xmath13-cup or a @xmath13-cap , and the regions @xmath43 from the support of @xmath12 satisfy @xmath44 in particular , every @xmath45-tuple obtained by selecting one point from each @xmath46 , @xmath47 , is in convex position .
note that theorem [ partition ] does not say anything about the points inside region @xmath48 .
let us also remark that in the proof of theorem [ partition ] in @xcite , the authors find a @xmath49-element set @xmath42 , such that @xmath13 of the regions in the support of @xmath12 each contain at least @xmath50 points from @xmath40 , and therefore these regions may not be consecutive .
however , by appropriately selecting a @xmath13-element subset @xmath51 , we obtain theorem [ partition ] .
let @xmath40 be an @xmath17-element planar point set in the plane in general position , where @xmath52 and @xmath9 , where @xmath10 is a sufficiently large absolute constant .
set @xmath53 .
we apply theorem [ partition ] to @xmath40 with parameter @xmath54 , and obtain a subset @xmath55 such that @xmath12 is a cup or a cap , and the points in @xmath12 appear in order from left to right . moreover
since @xmath53 is large , regions @xmath56 in the support of @xmath12 satisfy @xmath57 set @xmath58 for @xmath59 .
we will assume that @xmath12 is a cap , since a symmetric argument would apply .
we say that the two regions @xmath30 and @xmath60 are _ adjacent _ if @xmath61 and @xmath62 are consecutive indices .
consider the subset @xmath63 and the region @xmath30 , for some fixed @xmath64 .
let @xmath65 be the segment @xmath66 .
see figure [ tbase ] .
the point set @xmath67 naturally comes with a partial order @xmath68 , where @xmath69 if @xmath70 and @xmath71 .
set @xmath72 . by dilworth
s theorem @xcite , @xmath67 contains either a chain of size at least @xmath73 or an antichain of size at least @xmath74 with respect to @xmath68 .
the proof now falls into two cases . _
case 1_. suppose there are @xmath75 parts @xmath67 in the collection @xmath76 , such that no two of them are in adjacent regions , and each such part contains a subset @xmath77 of size at least @xmath74 such that @xmath77 is an antichain with respect to @xmath68 .
let @xmath78 be the selected subsets . for each @xmath79 , @xmath80 , the line spanned by any two points in @xmath79 does not intersect the segment @xmath81 , and therefore , does not intersect region @xmath82 for @xmath83 ( by the non - adjacency property ) .
since @xmath1 is sufficiently large , we have @xmath84 , and therefore @xmath85 theorem [ cupscaps ] implies that @xmath79 contains either an @xmath1-cup or a @xmath86-cap .
if we are in the former case for any @xmath87 , then we are done .
therefore we can assume @xmath79 contains a subset @xmath88 that is a @xmath86-cap , for all @xmath89 .
we claim that @xmath90 is a cap , and therefore @xmath91 is in convex position .
let @xmath92 .
since @xmath93 , there is a point @xmath94 such that the line @xmath15 supported by the segment @xmath95 has the property that all of the other points in @xmath88 lie below @xmath15 . since @xmath15 does not intersect @xmath81 , all of the points in @xmath96 must lie below @xmath15 .
hence , @xmath91 is a cap and @xmath97 _ case 2_. suppose we are not in case 1 .
then there are @xmath98 consecutive indices @xmath99 such that each such part @xmath100 contains a subset @xmath101 such that @xmath101 is a chain of length at least @xmath102 with respect to @xmath68 . for simplicity
, we can relabel these sets @xmath103 .
consider the subset @xmath77 inside the region @xmath30 , and order the elements in @xmath104 with respect to @xmath68 .
we say that @xmath105 is a _
right - cap _ if @xmath106 is in convex position , and we say that @xmath107 is a _ left - cap _ if @xmath108 is in convex position . notice that left - caps and right - caps correspond to the standard notion of cups and caps after applying an appropriate rotation to the plane so that the segment @xmath32 is vertical . since @xmath77 is a chain with respect to @xmath68 , every triple in @xmath77 is either a left - cap or a right - cap , but not both .
moreover , for @xmath21 , if @xmath109 and @xmath110 are right - caps ( left - caps ) , then @xmath111 and @xmath112 are both right - caps ( left - caps ) . by theorem [ comcupcap ] , if @xmath113 , then @xmath77 contains either a @xmath13-left - cap or an @xmath18-right - cap .
we make the following observation . [ glue ]
consider the ( adjacent ) sets @xmath114 and @xmath115 . if @xmath114 contains a @xmath13-left - cap @xmath116 , and @xmath115 contains an @xmath18-right - cap @xmath117 , then @xmath118 forms @xmath119 points in convex position . by lemma
[ four ] , it suffices to show every four points in @xmath118 are in convex position .
if all four points lie in @xmath120 , then they are in convex position . likewise if they all lie in @xmath121 , they are in convex position .
suppose we take two points @xmath122 and two points @xmath123 .
since @xmath114 and @xmath77 are both chains with respect to @xmath68 , the line spanned by @xmath124 does not intersect the region @xmath125 , and the line spanned by @xmath126 does not intersect the region @xmath127 . hence @xmath128 are in convex position .
now suppose we have @xmath129 and @xmath130 . since the three lines @xmath131 spanned by @xmath132 all intersect
the segment @xmath133 , both @xmath134 and @xmath135 lie in the same region in the arrangement of @xmath136
. therefore @xmath128 are in convex position .
the same argument follows in the case that @xmath137 and @xmath138 .
see figure [ gluepic ] .
[ gluepic ] and a 4-right - cap in @xmath115 , which forms 7 points in convex position . ]
we have for @xmath139 , @xmath140 set @xmath141 . since @xmath1 is sufficiently large , we have @xmath142 which implies that @xmath143 either contains an @xmath1-right - cap , or a @xmath144-left - cap . in the former case
we are done , so we can assume that @xmath143 contains a @xmath144-left - cap .
likewise , @xmath145 , which implies @xmath146 contains either an @xmath147-right - cap , or a @xmath148-left - cap . in the former case we are done since observation [ glue ] implies that the @xmath144-left - cap in @xmath143 and the @xmath147-right - cap in @xmath146 form @xmath1 points in convex position .
therefore we can assume @xmath146 contains a @xmath148-left - cap . in general ,
if we know that @xmath114 contains an @xmath149-left - cap , then we can conclude that @xmath77 contains an @xmath150-left - cap .
indeed , for all @xmath151 we have @xmath152 since @xmath1 is sufficiently large , ( [ p1 ] ) and ( [ p2 ] ) imply that @xmath153 therefore , @xmath77 contains either an @xmath154-right - cap or an @xmath150-left - cap . in the former case
we are done by observation [ glue ] ( recall that we assumed @xmath114 contains an @xmath149-left - cap ) , and therefore we can assume @xmath77 contains an @xmath150-left - cap .
hence for @xmath155 , we can conclude that @xmath156 contains an @xmath1-left - cap .
this completes the proof of theorem [ main ] .
following the initial publication of this work on arxiv , we have learned that gbor tardos has improved the lower order term in the exponent , showing that @xmath157 . | let @xmath0 be the smallest integer such that any set of @xmath0 points in the plane in general position contains @xmath1 points in convex position .
in their seminal 1935 paper , erds and szekeres showed that @xmath2 . in 1960
, they showed that @xmath3 and conjectured this to be optimal . in this paper , we nearly settle the erds - szekeres conjecture by showing that @xmath4 . |
we call @xmath0 a pseudo - boolean function . the present work discusses approaches to obtain heuristics for the program @xmath1 \text{subject to } & $ \v x\in\b^d$ \end{tabular}\ ] ] using sequential monte carlo techniques . in the sequel
, we refer to @xmath2 as the _ objective function_. for an excellent overview of applications of binary programming and equivalent problems we refer to the survey paper by and references therein .
the idea to use particle filters for global optimization is not new [ , section 2.3.1.c ] , but novel sequential monte carlo methodology making use of suitable parametric families on binary spaces @xcite may allow to construct more efficient samplers for the special case of pseudo - boolean optimization .
we particularly discuss how this methodology connects with the cross - entropy method @xcite which is another particle driven optimization algorithm based on parametric families . the sequential monte carlo algorithm as developed by
is rather complex compared to local search algorithms such as simulated annealing @xcite or @xmath3-opt local search @xcite which can be implemented in a few lines .
the aim of this paper is to motivate the use of advanced particle methods and sophisticated parametric families in the context of pseudo - boolean optimization and to provide conclusive numerical evidence that these complicated algorithms can indeed outperform simple heuristics if the objective function has poorly connected strong local maxima .
this is not at all clear , since , in terms of computational time , multiple randomized restarts of fast local search heuristics might very well be more efficient than comparatively complex particle approaches .
the article is structured as follows .
we first introduce some notation and review how to model an optimization problem as a filtering problem on an auxiliary sequence of probability distributions .
section [ sec : smc ] describes a sequential monte carlo sampler @xcite designed for global optimization on binary spaces @xcite .
section [ sec : pfbs ] reviews three parametric families for sampling multivariate binary data which can be incorporated in the proposed class of particle algorithms .
section [ sec : algorithms ] discusses how the cross - entropy method @xcite and simulated annealing @xcite can be interpreted as special cases of the sequential monte carlo sampler . in section [ sec : applications ] we carry out numerical experiments on instances of the unconstrained quadratic binary optimization problem .
first , we investigate the performance of the proposed parametric families in particle - driven optimization algorithms .
secondly , we compare variants of the sequential monte carlo algorithm , the cross - entropy method , simulated annealing and simple multiple - restart local search to analyze their respective efficiency in the presence or absence of strong local maxima .
we briefly introduce some notation that might be non - standard .
we denote scalars in italic type , vectors in italic bold type and matrices in straight bold type .
given a set @xmath4 , we write @xmath5 for the number of its elements and @xmath6 for its indicator function . for @xmath7
we denote by @xmath8 the discrete interval from @xmath9 to @xmath10 . given a vector @xmath11 and an index set @xmath12 , we write @xmath13 for the sub - vector indexed by @xmath14 and @xmath15 for its complement .
we occasionally use the norms @xmath16 and @xmath17 . for particle optimization
, the common approach is defining a family of probability measures @xmath18 associated to the optimization problem @xmath19 in the sense that @xmath20 where @xmath21 denotes the uniform distribution on the set @xmath22 and @xmath23 the set of maximizers .
the idea behind this approach is to first sample from a simple distribution , potentially learn about the characteristics of the associated family and smoothly move towards distributions with more mass concentrated in the maxima .
we review two well - known techniques to explicitly construct such a family @xmath24 .
we call @xmath18 a tempered family , if it has probability mass functions of the form @xmath25 where @xmath26 .
as @xmath27 increases , the modes of @xmath28 become more accentuated until , in the limit , all mass is concentrated on the set of maximizers .
the name reflects the physical interpretation of @xmath29 as the probability of a configuration @xmath11 for an inverse temperature @xmath27 and energy function @xmath30 .
this is the sequence used in simulated annealing @xcite .
we call @xmath18 a level set family , if it has probability mass functions of the form @xmath31 where @xmath32\leq 1}$ ] for @xmath33 .
indeed , @xmath34 is the super - level set of @xmath2 with respect to the level @xmath35 , for @xmath36 , and @xmath37 is the uniform distribution on @xmath34 . as @xmath27 increases , the support of @xmath28 becomes restricted to the points that have an objective value sufficiently close to the maximum of the @xmath2 . in the limit
, the support is reduced to the set of global maximizers .
the particle - driven optimization algorithms are computationally more involved than local search heuristics since we need to construct a sequence of distributions instead of a sequence of states .
we shall see that this effort pays off in strongly multi - modal scenarios , where even sophisticated local search heuristics can get trapped in a subset of the state space .
while the tempered sequence is based on a physical intuition , the level set sequence has an immediate interpretation as a sequence of rare events since , as @xmath38 increases , the super - level set becomes a ` rare event ' with respect to the uniform measure .
rare event simulation and global optimization are therefore closely related concepts and methods for rare event estimation can often be adapted to serve as optimization algorithms .
particle algorithms for rare event simulation include the cross - entropy method @xcite and the sequential monte carlo sampler @xcite .
the former uses the level set sequence , the latter uses a _ logistic potential family _ @xmath39),\end{aligned}\ ] ] where @xmath40)$ ] and @xmath41^{-1}$ ] denotes the logistic function . did not specifically design their algorithm for optimization but their approach to static rare event simulation is closely related to the particle optimization framework .
we discuss a static sequential monte carlo sampler that uses a transition kernel with independent proposals in the move step based on suitable parametric families . this methodology has been demonstrated to reliably estimate the mean of the posterior distribution in bayesian variable selection problems for linear normal models @xcite . in this section
, we provide a self - contained description of this framework . for a more general overview of sequential monte carlo methods we refer to .
for convenience of notation , we index the sequence of distributions @xmath42 directly by @xmath43 rather that by the parameter of the family @xmath44 .
we refer to @xmath45 and @xmath46^n$ ] with @xmath47 as a _ particle system _ with @xmath48 particles .
we say the particle system @xmath49 _ targets _ the probability distribution @xmath50 if the empirical distribution @xmath51 converges to @xmath50 for @xmath52 .
we sample @xmath53 with @xmath54 and set @xmath55 to initialize the system .
suppose we are given a particle approximation @xmath56 of @xmath57 and want to target the subsequent distribution @xmath58 . for all @xmath59 and @xmath60
, we update the weights to @xmath61 and @xmath62 denotes the unnormalized version of @xmath28 .
the normalizing constants @xmath63 and @xmath64 defined in equations and are unknown but the algorithm only requires ratios of unnormalized probability mass functions .
we refer to @xmath65 as the _ step length _ at time @xmath43 . after updating , the particle system targets the distribution @xmath66
as we choose @xmath65 larger , that is @xmath67 further from @xmath68 , the weights become more uneven and the accuracy of the importance approximation deteriorates .
if we repeat the weighting step , we just increase @xmath65 and finally obtain an importance sampling estimate of @xmath69 with instrumental distribution @xmath70 .
this yields a poor estimator since the probability to hit the set @xmath71 with @xmath48 uniform draws is @xmath72 and decreases rapidly as the dimension @xmath73 grows .
the pivotal idea behind sequential monte carlo is to alternate moderate updates of the importance weights and improvements of the particle system via resampling and markov transitions .
the importance weight degeneracy is often measured through the _ effective sample size _
criterion @xcite defined as @xmath74.\ ] ] the effective sample size is @xmath75 if the weights are uniform , that is equal to @xmath76 ; the effective sample size is @xmath76 if all mass is concentrated in a single particle .
given any increasing sequence @xmath42 , we could repeatedly reweight and monitor whether the effective sample size falls below a critical threshold .
for the special case of annealing via sequential monte carlo , however , the effective sample size after weighting @xmath77 is merely a function of @xmath65 . for a particle system @xmath78 at time @xmath43
, we pick a step length @xmath65 such that @xmath79 that is we lower the effective sample with respect to the current particle approximation by some fixed ratio @xmath80 [ , ] .
this ensures a ` smooth ' transition between two auxiliary distributions , in the sense that consecutive distributions are close enough to approximate each other reasonably well using importance weights ; in our numerical experiments , we took @xmath81 .
we obtain the associated sequence @xmath82 by setting @xmath83 where @xmath84 is a unique solution of .
since @xmath77 is continuous and monotonously decreasing in @xmath65 we can use bi - sectional search to solve . in this section ,
we discuss how to condition the weighted system and improve its quality before we proceed with the next weighting step .
we replace the system @xmath85 targeting @xmath58 by a selection of particles @xmath86 drawn from the current particle reservoir @xmath87 such that @xmath88 where @xmath89 denotes the number of particles identical with @xmath90 .
thus , in the resampled system particles with small weights have vanished while particles with large weights have been multiplied . for the implementation of the resampling step ,
there exist several recipes .
we could apply a multinomial resampling @xcite which is straightforward .
there are , however , more efficient ways like residual @xcite , stratified @xcite and systematic resampling @xcite .
we use the latest in our simulations , see procedure [ algo : resample ] .
@xmath91 + * sample * @xmath92}$ ] + @xmath93 if we repeated the weighting and resampling steps several times , we would rapidly reduce the number of different particles to a very few .
the key to fighting the depletion of the particle reservoir is moving the particles according to a markov transition kernel @xmath94 with invariant measure @xmath58 .
the particle @xmath95 is by construction approximately distributed according to @xmath58 , and a draw @xmath96 is therefore again approximately distributed according to @xmath58 . the last sample of the generated markov chain @xmath97 is , for sufficiently many move steps @xmath98 , almost exactly distributed according to the invariant measure @xmath58 and independent of its starting point .
while we could always apply a fixed number of move steps , we rather use an adaptive stopping criterion based on the number of distinct particles .
we define the _ particle diversity _ as @xmath99.\ ] ] ideally , the sample diversity @xmath100 should correspond to the expected diversity @xmath101 where @xmath102 is the smallest value that solves @xmath103 .
this is the particle diversity we would expect if we had an independent sample from @xmath104 .
therefore , if @xmath94 is fast - mixing , we want to move the system until @xmath105 since the quantity on the right hand side is unknown , we stop moving the system as soon as the particle diversity reaches a steady state we can not push it beyond @xcite .
more precisely , we stop if the absolute diversity is above a certain threshold @xmath106 or the last improvement of the diversity is below a certain threshold @xmath107 .
we always stop after a finite number of steps but the thresholds @xmath108 and @xmath109 need to be calibrated to the efficiency of the transition kernel . for slow - mixing kernels ,
we recommend to perform batches of consecutive move steps instead of single move steps .
if the average acceptance rate @xmath110 of the kernel ( see section [ sec : kernels ] ) is smaller than @xmath109 , it is likely that the algorithm stops after the first iteration although further moves would have been necessary .
we could adaptively adjust the threshold @xmath109 to be proportional to an estimate of the average acceptance rate ; for our numerical experiments , however , we kept it fixed to @xmath111 .
@xmath112 + @xmath113 most transition kernels in monte carlo simulations are some variant of the metropolis - hastings kernel ( see e.g. ) , @xmath114 where we sample from the kernel by proposing a new state @xmath115 and accepting the proposal with probability @xmath116 or returning @xmath90 otherwise .
again , we denote by @xmath117 the unnormalized version of @xmath58 since the kernel only requires the ratio of the unnormalized probability mass functions . on binary spaces ,
a common choice for the proposal distribution is @xmath118 with weight vector @xmath119^d$ ] normalized such that @xmath120 . with probability @xmath121 ,
the kernel proposes a uniform draw from the @xmath3-neighborhood of @xmath90 , @xmath122 we refer to this type of kernel as _ symmetric kernel _ since @xmath123 and equation simplifies .
this class of kernels provide a higher mutation rate than the random - scan gibbs kernel ( see for adiscussion ) .
locally operating transition kernels of the symmetric type are known to be slowly mixing .
if we put most weight on small values of @xmath3 , the kernel only changes one or a few entries in each step .
if we put more weight on larger values of @xmath3 , the proposals will hardly ever be accepted if the invariant distribution @xmath50 is multi - modal .
ideally , we want the particles sampled from the transition kernel to be nearly independent after a few move steps which is often hard to achieve using local transition kernels . for the sequential monte carlo algorithm
, we use _ adaptive independent kernels _ which have proposal distributions of the kind @xmath124 which do not depend on the current state @xmath90 but have a parameter @xmath125 which we adapt during the course of the algorithm .
the adaptive independent kernel is rapidly mixing if we can fit the _
parametric family _
@xmath126 such that the proposal distribution @xmath127 is sufficiently close to the target distribution @xmath58 , yielding thus , on average , high acceptance rates @xmath128 . the general idea behind this approach
is to take the information gathered in the current particle approximation into account ( see e.g. ) .
the usefulness of this strategy for sampling on binary spaces has been illustrated by .
we fit a parameter @xmath129 to the particle approximation of @xmath58 according to some suitable criterion .
precisely , @xmath129 is taken to be the maximum likelihood or method of moments estimator applied to the weighted sample @xmath130 .
the choice of the parametric family @xmath126 is crucial to the implementation of a sequential monte carlo sampler with adaptive independent kernel .
we discuss this issue in detail in section [ sec : pfbs ] .
adaptation could , to a certain extent , also be done for local transition kernels . propose an adaptive kernel which replaces the full conditional distribution of the gibbs sampler by an easy to compute linear approximation which is estimated from the sampled particles .
this method accelerates gibbs sampling if the target distribution @xmath50 is hard to evaluate but does not provide fast mixing like the adaptive independent kernel ( see for a comparison ) .
still , the use of local kernels in the context of the proposed sequential monte carlo algorithm might be favorable if , for instance , the structure of the problem allows to rapidly compute the acceptance probabilities of local moves .
further , batches of local moves can be alternated with independent proposals to ensure that the algorithm explores the neighborhood of local modes sufficiently well .
since the sample space @xmath131 is discrete , a given particle is not necessarily unique .
this raises the question whether it is sensible to store multiple copies of the same weighted particle in our system . in the sequel
, we discuss some more details concerning this issue which has only been touched upon briefly by .
let @xmath89 denote the number of copies of the particle @xmath90 in the system @xmath49 .
indeed , for parsimonious reasons , we could just keep a single representative of @xmath90 and aggregate the associated weights to @xmath132 . shifting weights between identical particles
does not affect the nature of the particle approximation but it obviously changes the effective sample size @xmath133 which is undesirable since we introduced the effective sample size as a criterion to measure the goodness of a particle approximation . from an aggregated particle system , we can not distinguish the weight disparity induced by reweighting according to the importance function and the weight disparity induced by multiple sampling of the same states which occurs if the mass of the target distribution is concentrated .
more precisely , we can not tell whether the effective sample size is actually due to the gap between @xmath134 and @xmath58 or the presence of particle copies due to the mass of @xmath134 concentrating on a small proportion of the state space which occurs by construction of the auxiliary distribution in section [ sec : stat model ] .
aggregating the weights means that the number of particles is not fixed at runtime . in this case , the straightforward way to implement the move step presented in section [ sec : move ] is breaking up the particles into multiple copies corresponding to their weights and moving them separately .
but instead of permanently splitting and pooling the weights it seems more efficient to just keep the multiple copies .
we could , however , design a different kind of resample - move algorithm which first augments the number of particles in the move step and then resamples exactly @xmath48 weighted particles from this extended system using a variant of the resampling procedure proposed by . a simple way to augment the number of particles is sampling and reweighting via @xmath135 where @xmath136 denotes the acceptance probability of the metropolis - hastings kernel .
we tested this variant but could not see any advantage over the standard sampler presented in the preceding sections .
for the augment - resample type algorithm the implementation is more involved and the computational burden significantly higher . in particular , the rao - blackwellization effect one might achieve when replacing
the accept - reject steps of the transition kernel by a single resampling step does not seem to justify the extra computational effort . indeed ,
aggregating the weights does not only prevent us from using the effective sample size criterion , but also requires extra computational time of @xmath137 in each iteration of the move step since pooling the weights is as complex as sorting . with our application in mind ,
however , computational time is more critical than memory , and we therefore recommend to refrain from aggregating the weights .
[ sec : pfbs ] we review three parametric families on @xmath131 . in contrast to the similar discussion in @xcite , we also consider a parametric family which can not be used in sequential monte carlo samplers but in the context of the cross - entropy method . for more details on parametric families on binary spaces
we refer to .
we frame some properties making a parametric family suitable as proposal distribution in sequential monte carlo algorithms .
a. for reasons of parsimony , we prefer a family of distributions with at most @xmath138 parameters like the multivariate normal .
b. given a sample @xmath139 from the target distribution @xmath50 , we need to estimate @xmath140 in a reasonable amount of computational time .
c. we need to generate samples @xmath141 from the family @xmath126 .
we need the rows of @xmath142 to be independent .
d. for the sequential monte carlo algorithm , we need to evaluate @xmath143 point - wise .
however , the cross - entropy method still works without this requirement .
e. we want the calibrated family @xmath144 to reproduce e.g. the marginals and covariance structure of @xmath50 to ensure that the parametric family @xmath144 is sufficiently close to @xmath50 . the simplest non - trivial distributions on @xmath131
are certainly those having independent components . for a vector @xmath145 of marginal probabilities
, we define the _ product family _
we check the requirement list from section [ sec : properties ] : ( a ) the product family is parsimonious with @xmath147 .
( b ) the maximum likelihood estimator @xmath148 is the weighted sample mean .
( c ) we can easily sample @xmath149 .
( d ) we can easily evaluate the mass function @xmath150 .
( e ) however , the product family does not reproduce any dependencies we might observe in @xmath49 . the last point is the crucial weakness which makes the product family impractical for particle optimization algorithms on strongly multi - modal problems .
consequently , the rest of this section deals with ideas on how to sample binary vectors with a given dependence structure .
there are , to our knowledge , two major strategies to this end . 1 . we construct a generalized linear model which permits to compute the conditional distributions .
we apply the chain rule and write @xmath126 as @xmath151 which allows to sample the entries of a random vector component - wise .
we sample from an auxiliary distribution @xmath152 and map the samples into @xmath131 .
we call @xmath153 a copula family , although we refrain from working with explicit uniform marginals . we first present a generalized linear model and then review a copula approach . even for rather simple non - linear models
we usually can not derive closed - form expressions for the marginal probabilities required for sampling according to .
therefore , we might directly construct a parametric family from its conditional probabilities .
we define , for a lower triangular matrix @xmath154 , the _
logistic conditionals family _ as @xmath155^{1-\gamma_i}\end{aligned}\ ] ] where @xmath41^{-1}$ ] is the logistic function .
we readily identify the product family @xmath156 as the special case @xmath157 .
the virtue of the logistic conditionals family is that , by construction , we can sample a random vector component - wise while the full probability @xmath158 of the sample @xmath159 is computed as a by - product of procedure [ algo : sampling ] .
we refer to the online supplement for instructions on how to fit the parameter @xmath160 .
@xmath161 + @xmath162 we check the requirement list from section [ sec : properties ] : ( a ) the logistic conditionals family is sufficiently parsimonious with @xmath163 .
( b ) we can fit the parameter @xmath160 via likelihood maximization .
the fitting is computationally intensive but feasible .
( c ) we can sample @xmath164 using the chain rule factorization .
( d ) we can exactly evaluate @xmath165 .
( e ) the family @xmath166 reproduces the dependency structure of the data @xmath167 although we can not explicitly compute the marginal probabilities .
let @xmath152 be a family of multivariate auxiliary distributions on @xmath168 and @xmath169 a mapping into the binary space .
we can sample from the copula family by setting @xmath170 for a draw @xmath171 from the auxiliary distribution .
most multivariate parametric families with at most @xmath138 parameters appear to either have a rather limited dependency range or they do not scale to higher dimensions @xcite .
therefore , the natural and seemingly only viable option for @xmath152 is the multivariate normal distribution @xcite . for a vector @xmath172 and a correlation matrix @xmath173
, we introduce the mapping @xmath174}(v_1),\dots,\ind_{(-\infty , a_d]}(v_d)),\ ] ] and define the _ gaussian copula family _ as @xmath175 for index sets @xmath176 , the cross - moments @xmath177 are equal the cumulative distribution function of the multivariate normal with respect to the entries indexed by @xmath14 ( see for a more detailed discussion ) .
in particular , the first and second moments are @xmath178 where @xmath179 and @xmath180 denote the cumulative distribution functions of the univariate and bivariate normal distributions with zero mean , unit variance and correlation coefficient @xmath181 $ ] .
we refer to the online supplement for instructions on how to fit the parameters @xmath182 and @xmath183 .
we check the requirement list from section [ sec : properties ] : ( a ) the gaussian copula family is sufficiently parsimonious with @xmath163 .
( b ) we can fit the parameters @xmath182 and @xmath184 via method of moments . however , the parameter @xmath184 is not always positive definite .
( c ) we can sample @xmath185 using @xmath186 with @xmath187 .
( d ) we can not easily evaluate @xmath188 since this requires computing high - dimensional integral expressions which is a computationally challenging problem in itself ( see e.g. ) .
the gaussian copula family is therefore less useful for sequential monte carlo samplers but can be incorporated into the cross - entropy method reviewed in section [ sec : cross entropy ] .
( e ) the family @xmath189 reproduces the exact mean and , possibly scaled , correlation structure .
we briefly discuss a toy example to illustrate the usefulness of the parametric families . for the quadratic function @xmath190 the associated probability mass function @xmath191 has a correlation matrix @xmath192 which indicates that this distribution has considerable dependencies and its mass function is therefore strongly multi - modal .
we generate pseudo - random data from @xmath50 , adjust the parametric families to the data and plot the mass functions of the fitted parametric families .
figure [ fig : toy exa ] shows how the three parametric families cope with reproducing the true mass function .
clearly , the product family is not close enough to the true mass function to yield a suitable instrumental distribution while the logistic conditional family almost copies the characteristics of @xmath50 and the gaussian copula family allows for an intermediate goodness of fit .
in this section , we provide a synopsis of all steps involved in the sequential monte carlo algorithm and connect this framework to the cross - entropy method and simulated annealing . in table
[ tab : seq ] , we state the necessary formulas for the tempered and the level set sequence introduced in section [ sec : stat model ] . for convenience , we summarize the complete sequential monte carlo sampler in algorithm [ algo : smc ] . note that , in practice , the sequence @xmath68 is not indexed by @xmath43 but rather by @xmath44 , which means that the counter @xmath43 is only given implicitly .
the algorithm terminates if the particle diversity sharply drops below some threshold @xmath193 which indicates that the mass has concentrated in a single mode .
if we use a kernel with proposals from a parametric family @xmath194 , we might already stop if the family degenerates in the sense that only a few components of @xmath194 , say less than @xmath195 , are random while the others are constant ones or zeros .
in this situation , additional moves using a parametric family are a pointless effort . we either return the maximizer within the particle system or we solve the subproblem of dimension @xmath196 by brute force enumeration .
we might also perform some final local moves in order to further explore the regions of the state space the particles concentrated on .
* sample * @xmath197 * for all * @xmath59 .
+ @xmath198 , @xmath199 + @xmath200 for the level set sequence , the effective sample size is the fraction of the particles which have an objective function value greater than @xmath201 ; see table [ tab : seq ] and equation .
the remaining particles are discarded since their weights equal zero .
consequently , there is no need to explicitly compute @xmath84 as a solution of .
we simply order the particles @xmath202 according to their objective values @xmath203 and only keep the @xmath204 particles with the highest objective values .
rubinstein @xcite , who popularizes the use of level set sequences in the context of the cross - entropy method , refers to @xmath204 as the size of the _ elite sample_. the cross - entropy method has been applied successfully to a variety of combinatorial optimization problems , some of which are equivalent to pseudo - boolean optimization @xcite , and is closely related to the proposed sequential monte carlo framework . however , the central difference between the cross - entropy method and the sequential monte carlo algorithm outlined above is the use of the invariant transition kernel in the latter .
we obtain the cross - entropy method as a special case if we replace the kernel @xmath205 by its proposal distribution @xmath194 .
the sequential monte carlo approach uses a smooth family of distributions @xmath18 and explicitly schedules the evolution @xmath68 which in turn leads to the proposal distributions @xmath194 .
the cross - entropy method , in contrast , defines the subsequent proposal distribution @xmath206 without any reference sequence @xmath134 to balance the speed of the particle evolution . in order to decelerate the advancement of the cross - entropy method
, we introduce a lag parameter @xmath207 and use a convex combination of the previous parameter @xmath208 and the parameter @xmath209 fit to the current particle system , setting @xmath210 however , there are no guidelines on how to adjust the lag parameter during the run of the algorithm .
therefore , the sequential monte carlo algorithm is easier to calibrate since the reference sequence @xmath134 controls the stride and automatically prevents the system from overshooting . on the upside
, the cross - entropy method allows for a broader class of auxiliary distributions @xmath194 since we do not need to evaluate @xmath211 point - wise which is necessary in the computation of the acceptance probability of the hastings kernel ; see section [ sec : gaussian copula ] .
.formulas for optimization sequences [ cols="<,^,^ " , ] a well - studied approach to pseudo - boolean optimization is simulated annealing @xcite . while the name stems from the analogy to the annealing process in metallurgy , there is a pure statistical meaning to this setup .
we can picture simulated annealing as approximating the mode of a tempered sequence using a single particle .
since a single observation does not allow for fitting a parametric family , we have to rely on symmetric transition kernels in the move step .
a crucial choice is the sequence @xmath44 which in this context is often referred to as the _ cooling schedule_. there is a vast literature advising on how to calibrate @xmath44 where a typical guideline is the expected acceptance rate of the hastings kernel .
we calibrate @xmath44 such that the empirical acceptance rate @xmath212 follows approximately @xmath213 for @xmath214 $ ] .
there are variants of simulated annealing which use more complex cooling schedules , tabu lists and multiple restarts , but we stick to this simple version for the sake of simplicity .
algorithm [ algo : sim ann ] describes the version we use in our numerical experiments in section [ sec : perf algo ] .
@xmath215 ( time elapsed ) + @xmath216 we describe a greedy local search algorithm which works on any state space that allows for defining a neighborhood structure .
the typical neighborhood on binary spaces is the @xmath3-neighborhood defined in .
a greedy local search algorithm computes the objective value of all states in the current neighborhood and moves to the best state found until a local optimum is reached .
the local search algorithm is called @xmath3-opt if it searches the neighborhood @xmath217 ( see e.g. for a discussion )
. the algorithm can be randomized by repeatedly restarting the procedure from randomly drawn starting points .
there are more sophisticated versions of local search algorithms exploit the properties of the objective function but even a simple local search procedure can produce good results @xcite . algorithm [ algo : local search ] describes the @xmath75-opt local search procedure we use in our numerical experiments in section [ sec : perf algo ] .
@xmath218 ( time elapsed ) + @xmath216
it is well - known that any pseudo - boolean function @xmath219 can be written as a multi - linear function @xmath220 where @xmath221 are real - valued coefficients .
we say the function @xmath2 is of order @xmath3 if the coefficients
@xmath222 are zero for all @xmath176 with @xmath223 .
while optimizing a first order function is trivial , optimizing a non - convex second order function is already an np - hard problem @xcite . in the sequel , we focus on optimization of second order pseudo - boolean functions to exemplify the stochastic optimization schemes discussed in the preceding sections . if @xmath2 is a second order function , we restate program as @xmath224 \text{subject to } & $ \v x\in\b^d$ , \end{tabular}\ ] ] where @xmath225 is a symmetric matrix .
we call an unconstrained quadratic binary optimization problem ( uqbo ) ; we refer to for a list of applications and equivalent problems . in the literature it is also referred to as unconstrained quadratic boolean or bivalent or zero - one programming @xcite .
meta - heuristics are a class of algorithms that optimize a problem by improving a set of candidate solutions without systematically enumerating the state space ; typically they deliver solutions in polynomial time while an exact solution has exponential worst case running time .
the outcome is neither guaranteed to be optimal nor deterministic since most meta - heuristics are randomized algorithms .
we briefly discuss the connection to particle optimization against the backdrop of the unconstrained quadratic binary optimization problem where we roughly separate them into two classes : local search algorithms and particle - driven meta - heuristics .
local search algorithms iteratively improve the current candidate solution through local search heuristics and judicious exploration of the current neighborhood ; examples are local search @xcite , tabu search @xcite , simulated annealing @xcite .
particle driven meta - heuristics propagate a set of candidate solutions and improve it through recombination and local moves of the particles ; examples are genetic algorithms @xcite , memetic algorithms @xcite , scatter search @xcite . for comparisons of these methods
we refer to or .
the sequential monte carlo algorithm and the cross - entropy method are clearly in the latter class of particle - driven meta - heuristics .
the idea behind sequential monte carlo is closely related to the intuition behind population ( or swarm ) optimization and genetic ( or evolutionary ) algorithms .
however , the mathematical framework used in sequential monte carlo allows for a general formulation of the statistical properties of the particle evolution while genetic algorithms are often problem - specific and empirically motivated . if we can explicitly derive the multi - linear representation of the objective function , there are techniques to turn program into a linear program . for the uqbo it reads @xmath226 \text{subject to } & $ \v
x\in\b^{d(d+1)/2}$ \\ & $ \left .
\begin{array}{l } \hspace{-1ex}x_{ij}\leq x_{ii } \\
\hspace{-1ex}x_{ij}\leq x_{jj } \\
\hspace{-1ex}x_{ij}\geq x_{ii } + x_{jj } -1 \\
\end{array } \right\}\text { for all}\ i , j\in \dset{1,d}$. \end{tabular}\ ] ] note , however , that there are more parsimonious linearization strategies than this straightforward approach [ , ] .
the transformed problem allows to access the tool box of linear integer programming which consist of branch - and - bound algorithms that are combined with rounding heuristics , various relaxations techniques and cutting plane methods [ , ] .
naturally , the question arises whether particle - driven meta - heuristics can be incorporated into exact solvers to improve branch - and - bound algorithms . indeed ,
stochastic meta - heuristics deliver lower bounds for maximization problems , but particle - driven algorithms are computationally somewhat expensive for this purpose unless the objective function is strongly multi - modal and other heuristics fail to provide good results ; see the discussion in section [ sec : extreme ]
. however , the sequential monte carlo approach in combination with the level set sequence might also be useful to determine a global branching strategy , since the algorithm provides an estimator for @xmath227 which is the average of the super - level set @xmath228 .
these estimates given for a sequence of levels @xmath229 might provide branching strategies than are superior to local heuristics or branching rules based on fractional solutions .
a further discussion of this topic is beyond the scope of this paper but it certainly merits consideration .
the meta - heuristics we want to compare do not exploit the quadratic structure of the objective function and might therefore be applied to any binary optimization program . if the objective function can be written in multi - linear form like there are efficient local search algorithms @xcite which exploit special properties of the target function and easily beat particle methods in terms of computational time .
therefore , the use of particle methods is particularly interesting if the objective function is expensive to compute or even a black box .
the posterior distribution in bayesian variable selection for linear normal models is an example of such an objective function ( see and references therein ) .
we stick to the uqbo for our numerical comparison since problem instances of varying difficulty are easy to generate and interpret while the results carry over to general binary optimization . in the vast literature on uqbo
, authors typically compare the performance of meta - heuristics on a suite of randomly generated problems with certain properties .
proposes standardized performance tests on symmetric matrices @xmath230 with entries @xmath231 drawn from the uniform
@xmath232 the test suites generated by ( * ? ?
? * http://people.brunel.ac.uk/~mastjjb/jeb/orlib/bqpinfo.html[or-library ] ) and follow this approach have been widely used as benchmark problems in the uqbo literature ( see for an overview ) . in the sequel
we discuss the impact of diagonal dominance , shifts , the density and extreme values of @xmath233 on the expected difficulty of the corresponding uqbo problem . generally , stronger _ diagonal dominance _ in @xmath233 corresponds to easier uqbo problems @xcite .
consequently , the original problem generator presented by is designed to draw the off - diagonal elements from a uniform on a different support @xmath234 with @xmath235 . in this context
, we point out that the impact of diagonal dominance carries over to the statistical properties of the tempered distributions we defined in the introductory section [ sec : stat model ] . indeed ,
stronger diagonal dominance in @xmath233 corresponds to exponential quadratic distributions @xmath236 having lower dependencies between the components of @xmath237 .
we can analytically derive a parameter @xmath154 for a logistic conditionals family @xmath166 that approximates @xmath238 where the quality of the approximation increases as the diagonal of @xmath233 becomes more dominant @xcite .
we can accelerate the sequential monte carlo algorithm by initializing the system from @xmath166 instead of @xmath239 .
however , we did not exploit this option to keep the present work more concise . for positive definite @xmath240 ,
the optimization problem is convex and can be solved in polynomial time @xcite ; in exact optimization , this fact is exploited to construct upper bounds for maximization problems @xcite .
we observe a corresponding complexity reduction in statistical modeling . for @xmath240 ,
the auxiliary distribution @xmath241 is a feasible mass function , and we can derive analytical expressions concerning all cross - moments and marginal distributions @xcite which allows to largely analyze the properties of @xmath238 without enumerating the state space .
the global optimum of the uqbo problem is more difficult to detect as we shift the entries of the matrix @xmath233 but the relative gap between the optimum and any heuristic value diminishes .
if we sample @xmath242 from a uniform on the _ shifted _ support @xmath243 we obtain an objective function @xmath244 where @xmath245 means equality in distribution .
hence , with growing @xmath246 the optimum depends less on @xmath233 and the relative gap between the optimum and a solution provided by any meta - heuristic vanishes .
define a related criterion @xmath247\ ] ] and report a significant impact of @xmath248 on the solution quality of their local search algorithms which is not surprising .
the difficulty of the optimization problem is related to the number of interactions , that is the number of non - zero elements of @xmath233 .
we call the proportion of non - zeros the _ density _ of @xmath233 . drawing @xmath231 from the mixture @xmath249\ ] ] we adjust the difficulty of the problem to a given expected density @xmath250 .
note that not all algorithms are equally sensitive to the density of @xmath233 . using the basic linearization
, each non - zero off - diagonal element requires the introduction of an auxiliary variable and three constraints .
thus , the expected total number of variables and the expected total number of constraints , which largely determine the complexity of the optimization problem , are proportional to the density @xmath250 . on the other hand , many randomized approaches , including the particle algorithms discussed in section [ sec : smc ] , are less sensitive to the density of the problem in the sense that replacing zero elements by small values has a minor impact on the performance of these algorithms . rather than the zero / non - zero duality
, we suggest that the presence of extreme values determines the difficulty of providing heuristic solutions . the uniform sampling approach advocated by is widely used in the literature for comparing meta - heuristics .
certainly , particle - driven methods are computationally too expensive to outperform local search heuristics on test problems with uniformly drawn entries ; @xcite confirms this intuition with respect to genetic algorithms versus tabu search and simulated annealing .
however , the uniform distribution does not produce _ extreme values _ and it is vital to keep in mind that these have an enormous impact on the performance of local search algorithms .
extreme values in @xmath233 lead to the existence of distinct local maxima @xmath251 of @xmath2 in the sense that there is no better candidate solution than @xmath252 in the neighborhood @xmath253 even for relatively large @xmath3 .
further , extreme local minima might completely prevent a local search heuristic from traversing the state space in certain directions .
consequently , local search algorithms , as discussed in section [ sec : meta ] , depend more heavily on their starting value , and their performance deteriorates with respect to particle - driven algorithms .
we propose to draw the matrix entries @xmath231 from a discretized cauchy distribution @xmath254 that has heavy tails which cause extreme values to be frequently sampled .
figure [ fig : problem distr ] shows the distribution of a cauchy and a uniform to illustrate the difference .
the resulting uqbo problems have quite distinct local maxima ; in that case we also say that the function @xmath255 is _ strongly multi - modal_. and a uniform @xmath256 distribution .
] in this section , we provide numerical comparisons based on instances of the uqbo problem .
we generated two random test suites of dimension @xmath257 , each having @xmath258 instances .
for the first suite , we sampled the matrix entries from a uniform distribution @xmath259 on @xmath260 ; for the second , we sampled from a cauchy distribution @xmath261 as defined in .
for performance evaluation , we run a specified algorithm @xmath262 times on the same problem and denote the outcome by @xmath263 . since the absolute values are not meaningful , we report the relative ratios @xmath264,\ ] ] where the best known solution is the highest objective value ever found for that instance and the worst solution is the lowest objective value among the @xmath262 outcomes .
we summarize the results in a histogram .
the first @xmath48 bins are singletons @xmath265 for the highest values @xmath266 ; the following @xmath48 bins are equidistant intervals @xmath267 .
the graphs show the bins @xmath268 in descending order from left to right on the @xmath269-axis .
the interval bins are marked with a sign `` @xmath270 '' and the lower bound .
the @xmath271-axis represents the counts . for comparison ,
we draw the outcome of several algorithms into the same histogram , where the worst solution found is the lowest overall objective value among the outcomes . for each algorithm ,
the counts are depicted in a different color and , for better readability , with diagonal stripes in a different angle . to put it
plainly , an algorithm performs well if its boxes are on the left of the graph since this implies that the outcomes where often close to the best known solution .
we study how the choice of the binary parametric family affects the quality of the delivered solutions .
the focus is on the cross - entropy method , since we can not easily use the gaussian copula family in the context of sequential monte carlo .
we use @xmath272 particles , set the speed parameter to @xmath273 ( or the elite fraction to @xmath274 ) and the lag parameter to @xmath275 .
the numerical comparisons , given in figures [ fig : uni mc ] and [ fig : cauchy mc ] , clearly suggest that using more advanced binary parametric families allows the cross - entropy method to detect local maxima that are superior to those detected using the product family .
hence , the numerical experiments confirm the intuition of our toy example in figure [ fig : toy exa ] . on the strongly multi - modal instance [ fig : cauchy mc ]
the numerical evidence for this conjecture is stunningly clear - cut ; on the weakly multi - modal problem [ fig : uni mc ] its validity is still unquestionable .
this result seems natural since reproducing the dependencies induced by the objective function is more relevant in the former case than in the latter .
we compare a sequential monte carlo sampler with parametric family , a sequential monte carlo sampler with a single - flip symmetric kernel , the cross - entropy method , simulated annealing and @xmath75-opt local search as described in section [ sec : algorithms ] . for the cross entropy method , we use the same parameters as in the preceding section . for the sequential monte carlo algorithm
, we use @xmath276 particles and set the speed parameter to @xmath277 ; we target a tempered auxiliary sequence . for both algorithms
we use the logistic conditionals family as sampling distribution . with these configurations ,
the algorithms converge in roughly @xmath278 minutes .
we calibrate the sequential monte carlo sampler with local moves to have the same average run time by processing batches of @xmath258 local moves before checking the particle diversity criterion .
the simulated annealing and @xmath75-opt local search algorithms run for exactly @xmath278 minutes .
the results shown in figures [ fig : uni ac ] and [ fig : cauchy ac ] assert the intuition that particle methods perform significantly better on strongly multi - modal problems .
however , on the easy test problems , the particle methods tend to persistently converge to the same sub - optimal local modes .
this effect is probably due to their poor local exploration properties . since particle methods perform significantly less evaluations of the objective function , they are less likely to discover the highest peak in a region of rather flat local modes .
the use of parametric families aggravates this effect , and it seems advisable to alternate global and local moves to make a particle algorithm more robust against this kind of behavior .
further numerical results are shown in figure [ fig : r250c ] and figure [ fig : r250u ] .
the numerical experiments carried out on different parametric families revealed that the use of the advanced families proposed in this paper significantly improves the performance of the particle algorithms , especially on the strongly multi - modal problems .
the experiments demonstrate that local search algorithms , like simulated annealing and randomized @xmath75-opt local search , indeed outperform particle methods on weakly multi - modal problems but deliver inferior results on strongly multi - modal problems .
using tabu lists , adaptive restarts and rounding heuristics , we can certainly design local search algorithms that perform better than simulated annealing and @xmath75-opt local search .
still , the structural problem of strong multi - modality persists for path - based algorithms . on the other hand , cleverly designed local search heuristics will clearly beat sequential monte carlo methods on easy to moderately difficult problems .
the results encourage the use of particle methods if the objective function is known to be potentially multi - modal and hard to analyze analytically .
we have to keep in mind that multiple restarts of rather simple local search heuristics can be very efficient if they make use of the structure of the objective function . for @xmath278 minutes of randomized restarts , the heuristic proposed by , which exploits the fact that the partial derivatives of a multi - linear function are constant
, practically always returns the best known solution on all test problems treated to create figures [ fig : r250c ] and [ fig : r250u ] .
the numerical work was completely done in http://www.python.org/[python 2.6 ] using http://www.scipy.org[scipy ] packages and run on a cluster with @xmath279 ghz processors .
the sources used in this work and the problems processed in this paper can be found at http://code.google.com/p/smcdss .
this work is part of the author s ph.d .
thesis at crest under supervision of nicolas chopin whom i would like to thank for the numerous discussions on particle algorithms .
i thank the editor and two anonymous referees for their detailed comments which helped to significantly improve this paper .
we briefly summarize how the parameters of the logistic conditionals family and the gaussian copula family can be assessed for a given particle system @xmath139 with weights @xmath280 .
we denote by @xmath281 the weighted first and second sample moments .
the log - likelihood function of the weighted logistic regression of @xmath282 on @xmath283 is @xmath284+(1-y^{(i)}_k)\log[1-\logistic(\v z_{k \bullet}^{(i)}\v a)]\right ] \\ & = \sum_{k=1}^n w_k\left[y^{(i)}_k\v z_{k \bullet}^{(i)}\v a-\log[1+\exp(\v z_{k \bullet}^{(i)}\v a)]\right],\end{aligned}\ ] ] where we used that @xmath285=-\log[1+\exp(\v x\t\v a)]=-\v x\t\v a+\log[\logistic(\v x\t\v a)]$ ] . since @xmath286/\partial\v a=\logistic(\v x\t\v a)\v x$ ] , the gradient of the log - likelihood is @xmath287 = ( \m z^{(i)})\t \diag(\v w)[\v y^{(i ) } - \v p^{(i)}_{\v a}],\end{aligned}\ ] ] where @xmath288 . since @xmath289\v x$ ] ,
the hessian matrix of the log - likelihood is @xmath290\right]\v
z_{k \bullet}^{(i)}(\v z_{k \bullet}^{(i)})\t = -(\m z^{(i)})\t \diag(\v w ) \diag(\v q^{(i)}_{\v a } ) \m z^{(i)},\end{aligned}\ ] ] where @xmath291 $ ] .
the data might suffer from complete or quasi - complete separation @xcite which causes the likelihood function @xmath292 to be monotonic . in that case
there is no maximizer .
we can avoid the monotonicity by assigning a suitable prior distribution to the parameter @xmath182 . recommends the jeffreys prior for its bias reduction which can conveniently be implemented via data adjustment @xcite . for the sake of simplicity
, however , we only assign a simple gaussian prior with variance @xmath293 such that , up to a constant , the log - posterior distribution is the log - likelihood function plus a quadratic penalty term and therefore always convex . the score function and its jacobian
matrix become @xmath294 - \varepsilon\v a,\quad s'(\v a)=-(\m z^{(i)})\t \diag(\v w ) \diag(\v q^{(i)}_{\v a } ) \m z^{(i)}-\varepsilon \m i.\end{aligned}\ ] ] the bias - reduced estimators are known to shrink towards @xmath295 which is an undesired property when fitting a parametric family .
therefore , we attempt to keep the shrinkage parameter @xmath296 as small as possible . the first order condition @xmath297 is solved iteratively @xmath298^{-1}s(\v a^{(t ) } ) = \v a^{(t)}+\v x^{(t)}\end{aligned}\ ] ] where @xmath299 is the vector that solves
@xmath300\v x^{(t)}= \left[(\m z^{(i)})\t \diag(\v w ) [ \v y^{(i ) } - \v p^{(i)}_{\v a^{(t)}}]-\varepsilon\v a^{(t)}\right].\end{aligned}\ ] ] if the newton iteration at the @xmath301th component fails to converge , we can either augment the penalty term @xmath296 which leads to stronger shrinkage of the mean @xmath302 towards @xmath295 or we can drop some covariates @xmath303 for @xmath304 from the iteration to improve the numerical condition of the procedure . in practice , we drop the predictors from the regression model which are only weakly correlated with the explained variable .
this step is important to speed up the algorithm and improve its numerical properties . for a proposal distribution
, it is particularly important to take the strong dependencies into account but it is often sufficient to work with very sparse logistic conditionals families . in particularly difficult cases
, we might prefer to set @xmath305 and @xmath306 , where @xmath307 is defined in , which guarantees that at least the mean is correct .
this is an important issue since misspecification of the mean of @xmath308 also affects the distribution of the components @xmath303 for @xmath309 which are sampled conditional on @xmath308 .
we adjust the gaussian copula family @xmath189 by method of moments .
we need to solve the non - linear equations @xmath310 where @xmath311 and @xmath312 are defined in while @xmath313 and @xmath314 denote the cumulative distribution functions of the univariate and bivariate normal distributions with zero mean , unit variance and correlation coefficient @xmath181 $ ] .
while the parameter @xmath315 is easy to assess , the challenging task is to compute the bivariate variances @xmath316 for all @xmath317 .
recall the standard result [ , p.255 ] @xmath318 where @xmath319 denotes the density of the bivariate normal distribution with correlation coefficient @xmath320 .
we obtain the following newton - raphson iteration @xmath321 starting at some initial value @xmath322 ; see procedure [ algo : fit gaussian ] . in the sequential monte carlo context
, good initial values are obtained from the parameters of the previous auxiliary distributions .
we use a fast series approximation @xcite to evaluate @xmath323 .
these approximations are critical when @xmath316 comes very close to either boundary of @xmath324 $ ] .
the newton iteration might repeatedly fail when restarted at the corresponding boundary @xmath325 . in any event
, @xmath326 is strictly monotonic in @xmath320 since its derivative is positive , and we can switch to bi - sectional search if necessary .
the locally fitted correlation matrices @xmath184 might not be positive definite for @xmath333 because the gaussian copula is not flexible enough to model the full range of cross - moments binary distributions might have ( see for an extended discussion ) .
we obtain a feasible parameter replacing @xmath184 by @xmath334 where @xmath335 is smaller than all eigenvalues of the locally fitted matrix @xmath184 .
this approach evenly lowers the local correlations to a feasible level and is easy to implement on standard software . in practice
, we also prefer to work with a sparse version of the gaussian copula family that concentrates on strong dependencies and sets minor correlations to zero . | we discuss a unified approach to stochastic optimization of pseudo - boolean objective functions based on particle methods , including the cross - entropy method and simulated annealing as special cases .
we point out the need for auxiliary sampling distributions , that is parametric families on binary spaces , which are able to reproduce complex dependency structures , and illustrate their usefulness in our numerical experiments .
we provide numerical evidence that particle - driven optimization algorithms based on parametric families yield superior results on strongly multi - modal optimization problems while local search heuristics outperform them on easier problems . |
however violent or disturbed in youth , mature elliptical galaxies are remarkably well - behaved members of the celestial pageant .
their photometric and structural properties obey well - defined relations over a very large range in mass and luminosity .
the correlations of distance - dependent properties with distance - independent ones then allows for the derivation of galaxy distances . the fundamental plane ( fp ) ( dressler et al .
1987 ; djorgovski & davis 1987 ) works as a distance indicator because elliptical galaxies are dynamically hot systems obeying the virial theorem and have fairly similar , mainly old , stellar populations .
the distance - dependent quantity is a combination of the galaxy effective radius and surface brightness ( or just the galaxy luminosity for the faber - jackson [ 1976 ] relation ) , and the distance - independent one is the stellar velocity dispersion @xmath2 .
sometimes a stellar population term , usually the _ 2 index , is also incorporated into the fp ( e.g. , guzmn & lucey 1993 ) .
the surface brightness fluctuations ( sbf ) distance method ( tonry & schneider 1988 ) relies solely on the stellar population properties of old stellar systems ; this enables cepheid calibration via sbf measurements in spiral bulges .
the distance - dependent quantity for sbf is the luminosity - weighted mean luminosity of the stellar population and the distance - independent one is the colour , or possibly some other stellar population indicator .
this paper examines and compares the various photometric and spectroscopic parameters used in the sbf and fp distance methods .
we first discuss the data samples used for this study before exploring correlations between the distance - independent parameters which go into the methods .
special attention is given to potential systematic effects that could bias the distance estimates .
we then describe our procedure for deriving the distance - dependent fp photometric parameters from sbf survey imaging data and compare the results with external data in order to evaluate their accuracy .
we include a data table listing both the fp and sbf parameters ; these are used in a companion paper ( blakeslee et al .
2002 , hereafter paper ii ) which makes detailed comparisons of the distance estimates for individual galaxies from the two methods .
for this study , we combine data from two recent early - type galaxy distance surveys : the sbf survey of galaxy distances and the streaming motions of abell clusters ( smac ) project .
the sbf survey data and calibration are described in detail by tonry et al .
( 1997 , 2000 , hereafter sbf - i and sbf - ii ) , with further analysis of the data given by blakeslee et al.(1999 , sbf - iii ) .
sbf survey data for 300 galaxies within @xmath3 km s@xmath4 are tabulated by tonry et al .
( 2001 , sbf - iv ) .
the smac project is a cluster fp survey which combines new and literature data into a common system .
as well as providing standardised parameters for cluster galaxies , the smac project derived improved values for many nearby non - cluster galaxies .
the new photometric and spectroscopic data are presented by smith et al .
( 2000 , 2001 , hereafter smac - i and smac - ii ) , and the full smac data set is tabulated by hudson et al .
( 2001 , smac - iii ) .
a preliminary analysis of the peculiar velocity data was given by hudson et al .
( 1999 ) . here
, we match data for galaxies appearing in both surveys and discuss correlations between the various distance - independent galaxy properties .
we matched all galaxies with colours from the sbf survey ( sbf - iv ) against the smac spectroscopic catalogue of homogenized _ 2 and @xmath2 measurements .
the spectroscopic measurements have been normalized to a standard aperture of radius @xmath5 by assuming a fixed radial gradient from jorgensen et al .
( 1995 ) for all the galaxies and using the observed velocities to approximate distances ( see smac - iii ) .
the colours , however , are derived from the regions within the galaxies that were used for the sbf analysis .
because the galaxy centres are usually saturated in the sbf observations and the central regions are frequently affected by dust , these colours roughly correspond to the colour near one effective radius @xmath6 ( aperture effects on the relation are discussed in [ ssec : heredity ] ) .
galactic extinction values and ratios from schlegel , finkbeiner , & davis ( 1998 , hereafter sfd ) were used in deriving the reddening - corrected colours ( see sbf - ii ) .
the adopted reddening law gives @xmath7 .
a total of 209 galaxies were found to have both and _ 2 data .
one of these is m32 , which has an extraordinarily large estimated aperture correction , while two others ( ngc404 and ngc5102 ) have @xmath8 .
we omit these three galaxies , leaving a sample of 206 , all but two ( e358 - 059 and ngc4468 ) of which also have smac values for @xmath2 .
the mean observational uncertainties are 0.018mag in ( not including reddening errors ) and 0.007mag in _ 2 ; sbf - i and smac - iii describe how the errors were estimated from comparisons of repeat observations .
figure [ fig : mg2vimain ] shows the correlation of with _ 2for the cross - matched data set .
we find that ellipticals ( filled squares , defined as having morphological type @xmath9 in the rc3 [ devaucouleurs et al . 1991 ] ) , and s0s ( open squares , effectively @xmath10 for this sample ) obey the same relation .
this is expected for a pair of purely stellar population parameters with similar age
metallicity degeneracy properties ( e.g. , worthey 1994 ) .
however , the 8 highest extinction galaxies ( crosses in figure [ fig : mg2vimain ] ) , which have @xmath11 and @xmath12 , preferentially lie below the mean relation .
the least - squares fits for galaxies with @xmath13 and @xmath14 are given by @xmath15 for all morphological types ( 197 galaxies , rms scatter 0.0275 mag ) and for just the ellipticals ( 125 galaxies , rms scatter 0.0257 mag ) , respectively .
these fits are clearly identical within the errors .
the 72 s0s exhibit a marginally higher scatter of 0.030 mag , however they also have larger measurements uncertainties on average , 0.020 mag compared to 0.017 mag for the ellipticals .
therefore , we assign no physical significance to the marginal difference in scatter . if we perform a bivariate fit using the quoted uncertainties added in quadrature with a cosmic scatter term , we find that the cosmic scatter in at a fixed _ 2 must be 0.021mag to obtain a reduced @xmath16 of unity ( or , formally , 0.020 when just the ellipticals are considered and 0.022 for just the s0s ) .
this cosmic scatter actually would include errors resulting from gradient and aperture effects , incorrect reddening estimates , as well as true peculiarities of the stellar populations . in the following section , we use the residuals from the relation to investigate problems with the extinction estimates . figure [ fig : mgisig ] plots _ 2 and against the stellar velocity dispersion @xmath2 for the cross - matched sample .
the relation has been discussed in detail by many authors ( e.g. , terlevich et al . 1981 ;
dressler et al . 1987 ; guzmn et al .
1992 ; bender , burstein , & faber 1993 ; colless et al .
1999 ) , and we mainly include it here as a comparison to the @xmath17@xmath2 relation . excluding the discordant s0 ngc3156 which has @xmath18 , we find @xmath19 with an rms scatter of 0.022 mag for the full morphological sample ( 0.021 mag for just the ellipticals ) , consistent with most previous studies . for the @xmath2 comparison , we again find that galaxies with @xmath11 mag appear to have had their reddening corrections overestimated by sfd . excluding these and ngc3156
, we find @xmath20 with an rms scatter of 0.031 mag for the full morphological sample ( again , the elliptical subsample gives a near identical relation and a scatter 0.030 mag ) .
we know of no other published derivations of a @xmath21 relation . however , bender et al .
( 1993 ) presented a relation between @xmath22 and @xmath21 , mainly using the highest quality subsample of 7s data from burstein et al .
the slope of that relation is also near 0.2 and the scatter is 0.032 mag when the blue galaxies with @xmath23 are omitted .
bower , lucey , & ellis ( 1992 ) examined the @xmath24@xmath21 relation and found scatters of 0.032 and 0.050 mag for samples comprised of 17 virgo and 31 coma early - type galaxies , respectively .
we can use the relation as a test of the galactic extinction map .
figure [ fig : mg2viext ] shows the residuals with respect to eq.([eq : vimg_ell ] ) and indicates an apparent problem at the highest extinctions : 7 of 8 galaxies ( and all 4 ellipticals ) with @xmath25 have negative residuals .
the offset is significant at nearly the 3@xmath2 level . in order of increasing extinction ,
the 8 galaxies are e208 - 021 , ngc4976 , e322 - 059 , e092 - 013 , e221 - 026 , ngc3136 , ngc2434 , and ngc2380 ; of these , only e322 - 059 has a positive residual .
there is no simple trend in the residuals at lower extinctions , however .
sfd produced a similar plot , showing residuals of the @xmath22_2relation ( data from faber et al .
1989 ) against their @xmath26 values .
although they concluded that there was no significant overall trend , all six galaxies with @xmath27 had negative residuals .
we note that arce & goodman ( 1999 ) also found that the sfd map overestimates the reddening in regions of smooth extinction with @xmath28 mag , although it may underestimate the reddening in regions with steep extinction gradients .
hudson ( 1999 ) has tested for large - scale systematic errors in extinction maps by fitting dipoles to galaxy @xmath22_2residuals as a function of position on the sky .
he used a sample of 311 galaxies of high photometric quality from faber et al .
( 1989 ) ( colours within a radius of 33 ) and the homogenized smac _ 2 values and found weak evidence , at the 92% confidence level , for a dipole residual of 13% amplitude in the sfd extinction map .
however , because the residuals with respect to a smaller , independent data set consisting of galactic globular clusters and rr lyra stars did not show this dipole , hudson concluded that the effect actually resulted from a @xmath290.01 mag systematic error in the galaxy colours across the sky .
we have repeated hudson s analysis , including the monte carlo estimations of the significance , using our sample of 205 galaxies with @xmath14 mag . although the sample is smaller , is @xmath2928% more sensitive to reddening errors .
we do not find a very significant dipole residual .
the confidence level for the dipole term in our fit is only 80% , with fractional amplitude @xmath30 .
the best - fitting direction , @xmath31 in galactic coordinates , is only @xmath32 from that found by hudson .
however , if just the four highest extinction galaxies , all of which have negative residuals in figure [ fig : mg2viext ] , are omitted , then the fitted dipole swings by more than 60@xmath33 and has a similar low significance .
these four highest extinction galaxies range in @xmath34 from 241@xmath33 to 316@xmath33 and in @xmath35 from @xmath36215 to @xmath37 .
we conclude that there is no convincing dipole residual in the sfd extinction map from these data , although the extinction values appear to be systematically overestimated in certain directions .
we now explore the possibility of environmental influences on the relation .
hudson ( 1999 ) found that cluster galaxies are @xmath38 mag redder in @xmath39 than group / field galaxies at a given _ 2 . based on stellar population models
, this could imply that cluster ellipticals are older and less enriched at a given _ 2 , as might be expected in hierarchical formation scenarios .
however , such an effect could also imply environmental variations in the abundance ratios , presumably due to differing time - scales for metal enrichment and the relative importance of type i and type ii supernovae , with shorter time - scales yielding a higher relative abundance of alpha elements such as mg ( see worthey 1998 for a discussion ) . in this case
, one would expect cluster ellipticals to have formed more rapidly and thus to have higher _ 2 at a given , or equivalently , to be bluer at a given _ 2 .
figure [ fig : mg2vienv ] shows the residuals of the relation for galaxies in the field , in the three most prominent clusters , and in other , smaller , groups .
the scatter among the virgo galaxies is the highest , but the sample is weighted more towards lower _
2 values than the other samples and may be affected by a couple moderate outliers .
formally , the offset in the relation between field and group / cluster galaxies is @xmath40 mag , or @xmath41 mag if virgo is omitted .
this is at the same @xmath42 significance level as found by hudson ( 1999 ) , but in the opposite sense : the field galaxies are marginally redder at a given _ 2 , which would be consistent with longer enrichment time - scales .
however , at this level , systematic directional errors in the extinction will become important , so we do not consider this result significant .
the mg abundances of giant ellipticals are enhanced with respect to fe abundances , as compared to the solar ratios ( e.g. , worthey , faber , & gonzalez 1992 ) .
variations in the mg to fe ( logarithmic ) ratio [ mg / fe ] at a fixed total metallicity and age should cause scatter in the relation just as it causes significant scatter in the plots of fe indices against mg indices ( e.g. , worthey 1998 ; kuntschner 2000 ) .
the effects of non - solar abundance ratios on broad - band photometric colours due to isochrone temperature and line blanketing effects is not well - modeled , so we decided to look into this in a little more detail .
first it is necessary to make some correction for the systematic difference in aperture between the _ 2 and measurements .
most of our galaxies are at distances of @xmath43 km s@xmath4 ( virgo , fornax , leo , dorado ) ; at this distance the fiducial radius of 0.6 corresponds to @xmath44 .
we attempted to derive colours near the galaxy centres using the surface photometry data files described in sbf - i . in many cases
this was impossible because of saturation , for instance the images from the ctio runs ( see sbf - i ) with their large 0472 pixels and severe charge bleeding could not be used for this .
for the s0s , the central colours showed large scatter when plotted against the sbf survey colours , owing to dust lanes and irregular gradients associated with disk contamination , etc .
however , the relation for the ellipticals was tighter ; figure [ fig : sbfcenter ] shows at @xmath45 plotted against from sbf - iv for the ellipticals ( regardless of whether or not they have _ 2 from smac ) .
the least absolute deviation line ( to reduce the effect of outliers ) shown in the figure is given by @xmath46\ , , \label{eq : vigrad}\ ] ] where ` ctr ' actually refers to a radius of 10 arcsec .
the aperture correction becomes larger for bigger , redder galaxies , but is always less than 0.02 mag .
however , the _ 2 value at @xmath45 will be @xmath47 mag less than the value measured within this radius , according to the mean gradient derived by jorgensen et al .
thus , we apply this offset to the _ 2 indices before making a comparison to the modified colours .
in addition , we follow kuntschner et al .
( 2001 ) in ` correcting ' the _ 2 indices to what they would be for solar abundance ratios at the same total metallicity @xmath48 .
we use the results from their sample of 72 early - type galaxies to derive @xmath49 where @xmath50 is the _ 2 value corrected to solar - abundance ratios .
the scatter in this relation is 0.005 mag .
figure [ fig : vimg2mod ] shows the modified and _ 2 values as compared to some of the latest ( solar abundance ratio ) stellar population models from vazdekis ( 2001 , in preparation ; see blakeslee , vazdekis , & ajhar 2001 ) , based on extensive empirical stellar libraries .
it is clear that the data do not follow the model locus . in particular , the galaxies with corrected @xmath51 have colours which are still too blue with respect to the models .
the apparent offset in the other direction for @xmath52 is less worrisome , as this region is less well - populated by nice ellipticals with well - defined gradients , and only one of the 72 kuntschner et al .
( 2001 ) galaxies which went into deriving eq.([eq : mg2solar ] ) had @xmath53 .
if the @xmath54 mag offset in the colours at high @xmath50 is due to non - solar abundance ratios , then colour for early - type galaxies must increase , just as the fe indices increase ( see trager et al .
2000 and kuntschner et al . 2001 ) , when corrected to solar ratios ( i.e. , lower mg at fixed @xmath48 ) .
there has been much discussion as to whether it is more proper to call mg and the other alpha elements ` enhanced ' or the fe - peak elements ` depressed ' in giant ellipticals ( e.g. , vazdekis et al .
1997 ; trager et al .
2000 , kuntschner et al . 2001 ) , but both are relative terms . in any case , our results indicate that colour is not ` enhanced ' like mg in these galaxies , but rather is ` depressed ' towards the blue , consistent with the results of salaris & weiss ( 1998 ) who found that isochrones with @xmath55}>0 $ ] were hotter , and thus bluer , at all stages in the stellar evolution than scaled - solar abundance isochrones at a given total metallicity .
the difference in temperature became much more pronounced for the higher - metallicity isochrones .
sbf - iv introduced the distance - independent ` fluctuation star count ' , defined as the difference between the sbf magnitude and the total galaxy magnitude : @xmath56 \ , , \label{eq : nbardef}\ ] ] where @xmath57 is the total luminosity and is the fluctuation luminosity ( the luminosity - weighted mean stellar luminosity ) .
therefore , corresponds to 2.5 times the logarithm of the luminosity - weighted number of stars in a galaxy ; as such , it essentially scales with the galaxy mass .
because and @xmath58 are measured in the same bandpass , this parameter is independent of galactic extinction and , if both are from the same image , independent of photometric zero point .
sbf - iv showed that correlated well with the galaxy velocity dispersion for a sample of 64 ellipticals with velocity dispersions from prugniel & simien ( 1996 ) .
figure [ fig : nbar_sig ] shows the corresponding plot for 174 sbf survey galaxies with @xmath58 values from the fits described in the following section and smac velocity dispersions .
the least - squares fit to the ellipticals and s0s with @xmath59 is @xmath60 with an rms scatter of 0.079 dex ; the 117 ellipticals give a closely consistent result and have a scatter of 0.069 dex .
thus , the relation is even tighter than the @xmath21_2relation , which has a scatter of 0.084 dex for the same sample of galaxies ( both the full sample and elliptical subsample ) .
eq.([eq : nbar_sig ] ) is nearly identical to the relation found by sbf - iv for their smaller sample . rearranging terms , we have @xmath61 which resembles the faber - jackson relation , except that the term has a dependence on mass - to - light ratio that results in the small scatter for eq.([eq : nbar_sig ] ) .
blakeslee et al.(2001 ) discuss the stellar - population aspects of in more detail .
we note that the value of the exponent in eq.([eq : fabernbar ] ) depends on the inverse nature of our least - squares fit
. a ` forward ' fit of as a function of @xmath21 would have given an exponent of about 2.8 ; ideally , one would do a full maximum - likelihood analysis . the use of for an alternative calibration of the sbf method is discussed in paper ii .
relatively few of the sbf survey galaxies with smac spectroscopic parameters also have smac surface photometry .
this is largely because the smac project did not use the surface photometry from the 7s survey ( burstein et al .
1987 ) , which is in the @xmath62-band and uncorrected for seeing effects , while it did include the recalibrated and merged 7s spectroscopic data ( davies et al .
therefore , we have derived fp photometric parameters from sbf survey data .
this presented a number of challenges , as the field sizes were not ideal and the galaxy centres often saturated .
however , sbf - i described a procedure for measuring surface photometry in a series of concentric circular annuli of 3 width for all the sbf survey images after masking and interpolating over the bright stars .
the surface photometry data were used in sbf - i to determine accurate photometric offsets for the many individual observing runs in the sbf survey and to provide accurate calibrations for non - photometric survey data .
we fitted sersic ( 1968 ) @xmath63 profiles to the sbf survey surface photometry data in circular annuli .
a generalization of the devaucouleurs ( 1948 ) law , the @xmath63 profile is convenient because it describes a large range of galaxy types from exponential disks to highly diffuse cd halos .
[ see ciotti & lanzoni ( 1997 ) and ciotti & bertin ( 1999 ) for discussions of the general properties of stellar systems following @xmath63 profiles .
] this profile has been widely used for empirically characterizing galaxy light distributions ( e.g. , davies et al . 1988 ; young & currie 1994 ; caon , capaccioli & donofrio 1993 ; courteau et al .
1996 ) , including specifically for measuring fp distances ( graham & colless 1997 ; graham 1998 ) .
kelson et al .
( 2000 ) also explored @xmath63-law fits and concluded that the difference with respect to the @xmath64-law fits was negligible for their subset of sbf survey data ; however , the full data set is much more morphologically diverse , so we have kept this extra degree of freedom . not all the of the sbf survey galaxies converged to reasonable @xmath63 fits ( reduced @xmath65 ) . the failures to produce a good fit were partly because of morphological irregularity ( 9% of the 299 galaxies in the sample presented by sbf - iv were classified as spirals ) and partly because of problems in the data ( central saturation , bright nearby stars , occasionally poor flattening , etc . )
that could be overcome for the sbf analysis , but not for the surface photometry fitting .
we fitted the @xmath66 and @xmath67 sbf surface photometry down to fixed fractions of the sky brightness , 2.0% in @xmath66 and 0.4% in @xmath67 .
acceptable @xmath63 fits were obtained for a total of 280 galaxies , of which 268 were elliptical or lenticular ( @xmath68 ) ; 257 of these were successfully fitted in both the @xmath66 and @xmath67 bands .
the fits were integrated to find the total magnitudes , half light ( or effective ) radii @xmath69 , and mean surface brightnesses interior to the half light radius .
the surface brightnesses are corrected for galactic extinction from sfd and @xmath70 dimming in all cases .
we use the same @xmath71-corrections as in sbf - i : @xmath72 for @xmath66 magnitudes and @xmath73 for @xmath67 magnitudes .
because the smac survey data are standardised to the @xmath74-band , we transform the @xmath66 and @xmath67 surface brightnesses to @xmath74 according to @xmath75 which comes from the empirically - based models discussed in [ ssec : heredity ] . at the median colour of @xmath76 ,
this implies @xmath77 , identical to the fixed colour transformation used by smith et al .
( 1997 ) and only 0.01 mag different from the fixed color assumed by smac - iii . as a further check
, we derived a rough transformation based on data for 11 early - type galaxies with @xmath78 colours measured by tonry , ajhar , & luppino ( 1990 ) , whose colours span a range of 0.20 mag and share a homogeneous system with our data .
the resulting relation has a slope of @xmath79 , consistent with eq.([eq : vrvirel ] ) , and is bluer in the mean by @xmath80 mag .
this @xmath81 zero - point offset is typical of the offsets among sbf survey runs prior to homogenization ( sbf - ii ) and is in the opposite sense to the 0.01 mag offset with respect to the smac - iii colour .
we conclude that eq.([eq : vrvirel ] ) is accurate at the 0.010.02 mag level . finally , we form @xmath82 , which is the photometric combination entering into the fp .
we have neglected seeing corrections on for this sample of nearby galaxies , all of which have @xmath83 and were observed in a typical seeing of 1 .
table 1 presents the results of our fits to the sbf survey photometry . the colums list : ( 1 ) galaxy name and ( 2 ) heliocentric velocity as in sbf - iv ; ( 3 ) morphological @xmath84-type from the rc3 ; ( 4 ) @xmath62-band extinction from sfd ; ( 5 ) colour from the sbf survey ; ( 6 ) fluctuation count parameter formed from the sbf survey and our fitted @xmath67-band @xmath58 ; ( 7 ) quality parameter defined in sbf - ii and tabulated in sbf - iv ; ( 8) @xmath85 from the @xmath67-band fit ; ( 9 ) mean effective @xmath74-band surface brightness from the @xmath67-band fit ; ( 10 ) @xmath85 from the @xmath66-band fit ; ( 11 ) from the @xmath66-band fit ; ( 12 ) @xmath21 from the smac survey ; and ( 13 ) _ 2 from the smac survey .
the surface brightnesses are corrected for galactic extinction and @xmath70 dimming , @xmath71-corrected , and transformed to the @xmath74-band as described above .
the errors on , @xmath21 , and _ 2 are taken from the sbf and smac catalogues .
for the error , we added the error in quadrature with a nominal 0.2 mag uncertainty in @xmath58 .
we have included the parameter for further information on the quality of the measurement , with high values ( @xmath86 ) indicating poor quality ( see sbf - ii and paper ii for discussions ) .
we have not included error estimates on the @xmath85 and values , as they are too strongly correlated , and we recommend using them only in the combination .
most of the remainder of this paper is dedicated to evaluating the uncertainty in the values .
figure [ fig : xfpsbfv_sbfi ] shows the comparison between the @xmath74-band values determined from the @xmath66 and @xmath67 sbf photometry .
the overall scatter is 0.023 dex for 257 early - type galaxies .
the largest outlier , at 0.014 dex , is ngc4111 , which is a nearly edge - on ( axis ratio @xmath87 ) s0 galaxy with a red bulge and bluer disk .
it does not appear in the 7s or smac catalogues and therefore is not used for the distance comparisons in paper ii .
excluding ngc4111 , the scatter in figure [ fig : xfpsbfv_sbfi ] is 0.021 dex .
for the 122 ellipticals ( @xmath9 ) , the scatter is 0.016 dex .
the internal error per measurements is therefore about 0.015 dex overall , or 0.011 dex for ellipticals .
table [ tab : xfpres ] summarizes the results of this comparison and the external comparisons below .
figures [ fig : xfpsmacsbfi ] and [ fig : xfpsmacsbfv ] compare from the @xmath67 and @xmath66 sbf surface photometry fits of the previous section to from the smac survey .
there are 32 galaxies , including 18 ellipticals , with from both data sets .
although our @xmath66 and @xmath67 fits were done independently of each other , these figures look remarkably similar in detail , not just in their offsets and scatter .
this is because the internal scatter is significantly smaller than the external scatter , which has contributions from systematic effects resulting from the different fitting techniques , etc .
the biggest outlier in both figures is the dorado s0 galaxy ngc1553 , which also has the largest in these figures . the data in the smac catalogue for this galaxy derive from the jorgensen et al .
( 1996 ) study .
as shown in table [ tab : xfpres ] , the scatter in for the smac sbf survey comparisons is 0.044 dex for the full overlap sample , 0.0390.040 dex if ngc1553 is omitted , and 0.0270.029 dex for just the ellipticals .
the offset between the two sets of values is about @xmath88 dex , in the sense of the smac values being larger . in comparison ,
smac - iii did not find significant offsets among the different studies which went into the smac catalogue .
the offset found here is not due to the photometry , as we directly compared photometric zero points for 10 of the 32 galaxies in common ( all having new photometry from the smac survey ) , and the systematic offset was @xmath89 mag , a factor of three too small .
therefore , the offset must come from the difference in fitting procedures used and may partially reflect the lack of seeing corrections .
however , the agreement is reasonable after the application of this systematic correction ; there is no evidence for systematic trends in the residuals with , redshift , or galaxy type ( although the scatter is larger for the s0s ) .
we can use the 7s photometric data set as an additional external check of our measurements .
we downloaded the data from faber et al .
( 1989 ) from the astronomical data center and calculated values from the @xmath62-band surface brightnesses , @xmath39 colours , and effective diameters .
the burstein & heiles ( 1984 ) extinction corrections were removed and the sfd extinction corrections were applied .
the surface brightnesses were corrected for @xmath70 dimming ( the 7s catalogue incorrectly states that this was already done ) , and then transformed to the @xmath74-band using @xmath90 since this is an extrapolation of the colours , it is more uncertain than eq.([eq : vrvirel ] ) , but the range of galaxy colours is small enough that the typical error in @xmath91 , and hence the @xmath74-band surface brightness , should only be a few hundredths of a magnitude , or @xmath47 dex in .
we noted above that the 7s photometric data were omitted from the smac survey because they were not corrected for seeing effects .
figure [ fig : dxfpsmac7s ] plots the difference in the values between the smac and 7s catalogues against redshift and compares an example smac seeing correction curve for a galaxy of fixed physical size @xmath92 ( 5 at coma distance ) observed in 2 seeing .
this model appears to overestimate the correction nearby , but gives a good match to the offset at larger distances .
table [ tab : xfpres ] includes the results of the 7s - smac comparison for galaxies within @xmath93 km s@xmath4 , for which the observed offset is not significant .
all of the sbf survey galaxies have @xmath93 km s@xmath4 ( and only a few have @xmath94 km s@xmath4 ) ; this is what has allowed us to neglect seeing effects in our own analysis . before proceeding with the comparisons however , we note that disagreed for ngc1389 by more than 0.45 dex between the sbf and 7s data sets ( it is not in the smac photometric catalogue ) .
the @xmath85 values agreed to within 0.04dex , but the transformed 7s is fainter by 1.44 mag .
this galaxy has ` quality 3 ' ( the lowest ) photometry in the 7s data set , but other quality 3 galaxies do not show similarly large disagreements , so we omit ngc1389 only from the comparisons below
. figures [ fig : xfpsbfi7s ] and [ fig : xfpsbfv7s ] show the comparisons of the transformed @xmath74-band sbf and 7s values .
the results are again summarized in table [ tab : xfpres ] .
there is no trend of the residuals with redshift .
however , the galaxies which lie significantly below the mean difference line for these sbf-7s comparisons are bluer than average .
figure [ fig : dxfp_vi ] plots the differences against colour .
galaxies with @xmath95 exhibit greater scatter and tend to have 7s values greater than expected from the comparison to the sbf survey photometry .
table [ tab : xfpres ] shows that the scatter decreases by @xmath2925% when only @xmath96 galaxies are considered .
this could be due to stellar population effects , as the 7s photometry comes from the @xmath62-band , while the sbf photometry is in @xmath66 and @xmath67 , and both have been transformed to @xmath74 .
however , it is not a simple linear relation between the residuals and colour ( there is no trend for the redder galaxies which constitute the bulk of the sample ) , and photometric transformation errors should not cause offsets in as large as 0.1 dex .
we can not test for a similar problem in the smac comparison because the overlapping galaxies all have @xmath97 ( lower panel of figure [ fig : dxfp_vi ] ) .
we use the values from the sbf survey data and test for colour - dependent errors in the distance analyses of paper ii . for the purpose of comparison to other works ,
the last column of table [ tab : xfpres ] lists the rms scatters in comparisons of defined with a coefficient of 0.30 instead of 0.33 .
for instance , smith et al .
( 1997 ) used a coefficient of 0.30 in doing these sorts of comparisons , although the actual calculation of fp distances used a coefficient of 0.326 ( hudson et al .
as table [ tab : xfpres ] shows , the magnitude of the scatter is reduced by 2035% for comparisons using the smaller coefficient ( the scatter would increase by a similar amount if a coefficient of 0.36 were used ) .
the systematic offset of our values would change from @xmath290.030 to @xmath290.020 dex for a coefficient of 0.30 ( and to @xmath290.040 dex for a coefficient of 0.36 ) .
we have adopted 0.33 for the coefficient in the fp relation because it is within 0.01 of the values derived from very large data samples by hudson et al .
( 1997 ) , jorgensen et al .
( 1996 ) , colless et al .
( 2001 ) and the smac project . from the various comparisons
, we can formally derive the error on for any of the data sets . specifically , we find that the external errors on our measurements of from sbf survey data are 0.030 - 0.035 dex , or 0.0200.025 dex if just the ellipticals are considered , with the 7s errors being about 0.030 dex for both ellipticals and s0s .
the errors translate to an error of 67% in distance , roughly a factor of 3 smaller than the error expected from the intrinsic scatter in the fp .
thus , our measurements are accurate enough for measuring fp distances , and the systematic correction of 0.03 dex brings them onto the homogeneous smac system .
we have merged the sbf survey data with the smac fp survey data and explored the correlations between the distance - independent galaxy properties measured in the two surveys . using 200 galaxies with data in both catalogues , we derived mean relations between sbf survey colour and smac _ 2 and velocity dispersions .
the intrinsic scatter in the relation ( including the effects of colour gradients and any reddening errors ) is 0.020 mag for the ellipticals and 0.022 mag for the s0s .
we used this relation to test for errors in the sfd reddenings .
there is no convincing evidence from these data for a dipole residual in the sfd map ( our best dipole fit has a 9% amplitude but only 80% confidence ) .
however , we do find that the galaxies in our sample with sfd @xmath98 mag have overestimated extinctions .
because these galaxies lie near the galactic plane in a fairly restricted longitude range , we can not say whether or not the problem is systematic at high extinction or simply indicative of large - scale coherent errors .
we also tested for the influences of environment and @xmath1-element enhancement on the relation .
consistent with expectations from the @xmath1-enhanced isochrone calculations of salaris & weiss ( 1998 ) , we found that the enhancement of the _ 2 values is not mirrored by redder colours .
in fact , comparison to solar - abundance ratio stellar population models indicates that is slightly ` depressed ' at a given total metalicity , and therefore appears to follow the fe - peak abundances more than the @xmath1-elements abundances .
we find no evidence for environmental effects larger than @xmath99 mag on the relation .
we have derived the relation between the stellar velocity dispersion @xmath2 from the smac catalogue and the ` fluctuation star count ' , which was introduced in sbf - iv and corresponds to the luminosity - weighted number of stars in a stellar population .
the scatter is only 0.079 dex in @xmath21 ( 0.069 dex for the ellipticals ) , making this even tighter than the _
2@xmath21 relation .
the -@xmath21 relation is equivalent to the faber - jackson relation with a stellar - population correction term .
finally , we have used the sbf survey @xmath66- and @xmath67-band photometry to derive the fp photometric parameters for the survey galaxies .
comparisons to the smac and 7s data samples allowed us to estimate the external errors on our measurements . a systematic correction of about 0.03 dex is required to bring our values onto the smac photometric system .
we provide a full data table of our results , to be used in a forthcoming analysis paper comparing sbf and fp distances to these galaxies .
we thank our sbf and smac survey collaborators ed ajhar , roger davies , alan dressler , david schlegel , and russell smith for their help in amassing and processing these data sets .
we are grateful to harald kuntschner for providing his _ 2 data in electronic format .
this work was supported at the university of durham by a pparc rolling grant in extragalactic astronomy and cosmology and made use of starlink computer facilities .
jpb thanks the acs project at johns hopkins university for support while finishing this paper .
lrrcccccccccc galaxy & @xmath100 & @xmath84 & @xmath101 & @xmath102 & @xmath102 & & @xmath103 & @xmath104 & @xmath105 & @xmath106 & @xmath21 @xmath102 & mg@xmath107 @xmath102 + e092 - 013 & 1770 & @xmath364 & 0.811 & 1.164 0.033 & 19.43 0.23 & 1.52 & 1.306 & 19.621 & 1.448 & 20.150 & 2.230 0.029 & 0.277 0.008 + e183 - 030 & 2668 & @xmath363 & 0.376 & 1.139 0.014 & 20.94 0.57 & 3.14 & 1.375 & 19.483 & 1.408 & 19.491 & 2.241 0.042 & 0.256 0.010 + e208 - 021 & 1003 & @xmath363 & 0.750 & 1.146 0.018 & 19.49 0.27 & 1.43 & 1.507 & 19.513 & 1.457 & 19.384 & 2.243 0.047 & 0.304 0.012 + e322 - 008 & 3029 & @xmath362 & 0.363 & 1.205 0.023 & 20.81 0.54 & 2.53 & 1.395 & 20.086 & 1.513 & 20.556 &
& + e322 - 038 & 3134 & @xmath363 & 0.560 & 1.242 0.051 & 20.32 0.32 & 2.37 & 1.471 & 20.765 & 1.297 & 20.101 & 2.311 0.041 & 0.315 0.014 + e322 - 059 & 3240 & @xmath362 & 0.796 & 1.220 0.027 & 20.57 0.28 & 3.64 & 1.600 & 20.637 & 1.458 & 20.188 & 2.363 0.041 & 0.311 0.014 + e322 - 088 & 2606 & @xmath362 & 0.484 & 1.156 0.034 & 19.85 0.68 & 2.46 & 1.021 & 18.865 & 0.963 & 18.815 & 2.208 0.041 & 0.278 0.014 + e322 - 101 & 2053 & 0 & 0.493 & 1.167 0.023 &
& 1.99 & & & 1.256 & 20.453 & 2.207 0.021 & 0.305 0.008 + e323 - 034 & 4254 & @xmath365 & 0.535 & 1.165 0.019 & 20.86 0.28 & 2.51 & 1.335 & 19.351 & 1.386 & 19.504 & 2.375 0.013 & 0.293 0.006 + e358 - 006 & 1358 & @xmath364 & 0.043 & 1.068 0.020 & 17.59 0.37 & 0.97 & 1.494 & 21.631 & 1.041 & 20.076 &
& + e358 - 059 & 1030 & @xmath363 & 0.039 & 1.081 0.016 & 17.27 0.27 & 1.02 & 1.067 & 20.145 & 1.008 & 19.948 &
& 0.140 0.008 + i0745 & 1135 & @xmath362 & 0.091 & 0.978 0.010 & 16.63 0.36 & 2.33 & 1.117 & 20.222 & 1.205 & 20.443 & & + i1153 & 744 & @xmath362 & 0.083 & 1.197 0.011 & 18.41 0.28 & 0.83 & 1.268 & 20.276 & 1.117 & 19.752 & & + i1459 & 1624 & @xmath365 & 0.069 & 1.194 0.018 & 22.21 0.33 & 1.49 & 1.684 & 19.574 & 1.718 & 19.752 & 2.492 0.042 & 0.332 0.010 + i2006 & 1340 & @xmath365 & 0.048 & 1.183 0.018 & 19.61 0.34 & 1.45 & 1.259 & 19.246 & 1.274 & 19.300 & 2.086 0.013 & 0.273 0.004 + i2311 & 1836 & @xmath365 & 0.620 & 1.136 0.027 & 19.98 0.24 & 2.09 & 1.275 & 18.975 & 1.238 & 18.896 &
& + i3370 & 2971 & @xmath365 & 0.404 & 1.185 0.022 & 20.91 0.24 & 2.69 & 1.425 & 19.332 & 1.483 & 19.560 & 2.318 0.019 & 0.257 0.008 + i4296 & 3660 & @xmath365 & 0.266 & 1.236 0.034 & 22.64 0.27 & 3.15 & 1.683 & 20.154 & 1.700 & 20.203 & 2.532 0.013 & 0.330 0.005 + i4797 & 2597 & @xmath364 & 0.340 & 1.184 0.018 & 20.59 0.27 & 1.96 & 1.331 & 19.282 & 1.382 & 19.266 & 2.366 0.041 & 0.302 0.014 + i4889 & 2486 & @xmath365 & 0.228 & 1.137 0.016 & 20.64 0.24 & 1.77 & 1.356 & 19.207 & 1.347 & 19.173 & 2.219 0.015 & 0.254 0.004 + i4943 & 2930 & @xmath365 & 0.221 & 1.133 0.023 & 19.23 0.25 & 2.80 & 1.153 & 19.727 & 1.000 & 19.209 & 2.230 0.042 & 0.256
0.010 + i5328 & 3136 & @xmath365 & 0.062 & 1.173 0.023 & 21.16 0.22 & 2.67 & 1.411 & 19.695 & 1.446 & 19.818 & 2.303 0.042 &
0.294 0.010 + n0063 & 1168 & 0 & 0.481 & 0.979 0.018 & 18.25 0.37 & 1.14 & 1.380 & 19.974 & 1.919 & 21.554 & & + n0185 & @xmath36227 & @xmath365 & 0.792 & 1.051 0.017 & 14.17 0.24 & 0.01 & 2.129 & 20.828 & & & & + n0221 & @xmath36200 & @xmath366 & 0.346 & 1.133 0.007 & 15.60 0.21 & 0.01 & 1.475 & 17.079 & 1.453 & 16.893 & 1.828 0.009 & 0.119
0.003 + n0274 & 1729 & @xmath363 & 0.242 & 1.135 0.020 & 18.70 0.49 & 1.68 & 1.228 & 19.624 & 1.519 & 20.631 & & + n0404 & @xmath3636 & @xmath363 & 0.253 & 1.054 0.011 & 16.63 0.21 & 0.01 & 1.711 & 19.858 & 1.575 & 19.424 & 1.676 0.039 & 0.016 0.007 + n0448 & 1917 & @xmath363 & 0.261 & 1.132 0.029 & & 1.41 & & & 1.163 & 19.316 & & + n0524 & 2416 & @xmath361 & 0.357 & 1.221 0.010 & 21.56 0.28 & 1.65 & 1.551 & 19.252 & 1.579 & 19.365 & & + n0584 & 1875 & @xmath365 & 0.182 & 1.157 0.009 & 20.62 0.28 & 1.41 & 1.433 & 18.915 & 1.445 & 18.970 & 2.295 0.009 & 0.268 0.003 + n0596 & 1817 & @xmath364 & 0.159 & 1.135 0.008 & 20.27 0.22 & 1.51 & 1.526 & 19.789 & 1.509 & 19.737 & 2.179 0.024 & 0.237 0.006 + n0636 & 1805 & @xmath365 & 0.110 & 1.156 0.008 & 20.64 0.24 & 1.13 & 1.499 & 20.067 & 1.495 & 20.080 & 2.212
0.011 & 0.259 0.006 + n0661 & 3836 & @xmath364 & 0.296 & 1.164 0.011 & 19.60 0.31 & 2.73 & 1.182 & 19.340 & 1.258 & 19.583 & 2.250 0.026 &
0.293 0.008 + n0680 & 2953 & @xmath364 & 0.331 & 1.189 0.011 & 20.48 0.45 & 2.87 & 1.269 & 19.306 & 1.378 & 19.706 & 2.292 0.026 & 0.281 0.008 + n0720 & 1716 & @xmath365 & 0.070 & 1.214 0.009 & 21.66 0.25 & 1.09 & 1.544 & 19.402 & 1.523 & 19.333 & 2.392 0.009 & 0.324 0.004 + n0821 & 1716 & @xmath365 & 0.475 & 1.196 0.022 & 20.88 0.24 & 1.00 & 1.510 & 19.627 & 1.594 & 19.915 & 2.288 0.009 & 0.294 0.004 + n0855 & 600 & @xmath365 & 0.311 & 1.015 0.018 & 16.25 0.24 & 0.23 & 1.365 & 20.660 & 1.246 & 20.344 &
& + n0891 & 529 & 3 & 0.280 & 1.142 0.017 & 19.23 0.23 & 0.36 & 2.081 & 21.577 & 2.110 & 21.713 &
& + n0936 & 1434 & @xmath361 & 0.152 & 1.213 0.010 & 21.78 0.34 & 0.94 & 1.838 & 20.357 & 1.948 & 20.663 & 2.263 0.014 & 0.276 0.004 + n0949 & 610 & 3 & 0.254 & 0.958 0.011 & 16.96 0.26 & 0.44 & 1.374 & 20.026 & 1.611 & 20.755 &
& + n1023 & 648 & @xmath363 & 0.263 & 1.193 0.017 & 20.70 0.24 & 0.25 & 1.706 & 19.161 & 1.667 & 19.011 & 2.259 0.030 & 0.298 0.009 + n1052 & 1475 & @xmath365 & 0.115 & 1.213 0.010 & 20.63 0.33 & 1.48 & 1.431 & 19.090 & 1.390 & 18.974 & 2.302 0.032 & 0.300 0.007 + n1162 & 2331 & @xmath365 & 0.206 & 1.173 0.032 & 20.63 0.34 & 1.91 & 1.245 & 19.585 & 1.294 & 19.758 &
& + n1172 & 1669 & @xmath364 & 0.275 & 1.112 0.032 & 19.40 0.24 & 1.40 & 1.664 & 21.203 & 1.991 & 22.321 & 2.073 0.039 &
0.230 0.007 + n1199 & 2705 & @xmath365 & 0.234 & 1.188 0.012 & 20.93 0.38 & 2.03 & 1.420 & 19.753 & 1.400 & 19.728 & 2.313 0.039 & 0.293 0.007 + n1201 & 1720 & @xmath362 & 0.070 & 1.178 0.009 & 20.54 0.35 & 2.04 & 1.557 & 19.725 & 1.503 & 19.546 &
& + n1209 & 2619 & @xmath365 & 0.163 & 1.198 0.010 & 21.01 0.30 & 1.72 & 1.253 & 19.050 & 1.187 & 18.829 & 2.321 0.055 & 0.301 0.010 + n1297 & 1550 & @xmath362 & 0.119 & 1.187 0.018 & 20.25 0.46 & 1.42 & 1.578 & 20.915 & 1.853 & 21.832 &
& + n1316 & 1780 & @xmath362 & 0.090 & 1.132 0.016 & 22.65 0.25 & 1.79 & 1.964 & 19.541 & 1.986 & 19.663 &
& + n1332 & 1469 & @xmath363 & 0.137 & 1.222 0.010 & 21.37 0.26 & 1.01 & 1.450 & 18.851 & 1.477 & 18.950 & & + n1336 & 1466 & @xmath363 & 0.049 & 1.124 0.032 & 18.46 0.25 & 1.45 & 1.606 & 21.632 & 1.533 & 21.395 & & + n1339 & 1380 & @xmath364 & 0.057 & 1.134 0.012 & 19.07 0.40 & 0.95 & 1.189 & 19.220 & 1.278 & 19.523 & 2.208 0.012 & 0.294
0.003 + n1344 & 1217 & @xmath365 & 0.078 & 1.135 0.011 & 20.52 0.35 & 0.97 & 1.521 & 19.308 & 1.474 & 19.130 & & + n1351 & 1539 & @xmath363 & 0.061 & 1.148 0.016 & 19.76 0.24 & 1.12 & 1.539 & 20.343 & 1.589 & 20.516 & 2.156 0.014 & 0.262 0.004 + n1366 & 1310 & @xmath362 & 0.073 & 1.095 0.018 & 18.93 0.34 & 1.48 & 1.230 & 19.383 & 1.316 & 19.745 &
& + n1373 & 1385 & @xmath364 & 0.061 & 1.085 0.013 & 17.48 0.50 & 2.05 & 1.004 & 19.797 & 1.028 & 19.863 &
& + n1374 & 1397 & @xmath365 & 0.061 & 1.146 0.016 & 19.77 0.22 & 1.00 & 1.375 & 19.384 & 1.404 & 19.533 & 2.265 0.014 & 0.304 0.004 + n1375 & 773 & @xmath362 & 0.062 & 1.070 0.019 & 17.93 0.22 & 0.51 & 1.174 & 19.943 & 1.156 & 19.960 & & + n1379 & 1408 & @xmath365 & 0.052 & 1.143 0.019 & 19.89 0.23 & 1.00 & 1.465 & 19.722 & 1.476 & 19.807 & 2.094 0.015 & 0.248 0.004 + n1381 & 1751 & @xmath362 & 0.059 & 1.189 0.018 & 19.44 0.28 & 1.16 & 1.200 & 18.842 & 1.238 & 18.960 &
& + n1382 & 1763 & @xmath363 & 0.074 & 1.106 0.013 & 17.80 0.36 & 2.58 & 1.182 & 20.471 & 1.196 & 20.202 & & + n1387 & 1302 & @xmath363 & 0.056 & 1.208 0.047 & 20.79 0.24 & 1.16 & 1.551 & 19.618 & 1.747 & 20.372 & & + n1389 & 950 & @xmath363 & 0.045 & 1.145 0.019 & 19.60 0.26 & 0.91 & 1.166 & 18.718 & & & & + n1395 & 1702 & @xmath365 & 0.100 & 1.215 0.010 & 21.83 0.24 & 1.49 & 1.690 & 19.676 & 1.647 & 19.520 & 2.396 0.014 & 0.312 0.003 + lrrcccccccccc galaxy & @xmath100 & @xmath84 & @xmath101 & @xmath102 & @xmath102 & & @xmath103 & @xmath104 & @xmath105 & @xmath106 & @xmath21 @xmath102 & mg@xmath107 @xmath102 + n1399 & 1422 & @xmath365 & 0.056 & 1.227 0.016 & 21.69 0.24 & 1.34 & 1.581 & 18.920 & 1.619 & 19.061 & 2.509 0.013 & 0.329 0.003 + n1400 & 549 & @xmath363 & 0.279 & 1.170 0.009 & 20.93 0.38 & 0.32 & 1.470 & 19.461 & 1.500 & 19.641 & 2.385 0.018 & 0.287 0.004 + n1404 & 1925 & @xmath365 & 0.049 & 1.224 0.016 & 21.54 0.26 & 2.00 & 1.440 & 18.449 & 1.480 & 18.650 & 2.366 0.012 & 0.304 0.003 + n1407 & 1766 & @xmath365 & 0.298 & 1.222 0.010 & 22.44 0.32 & 1.42 & 1.761 & 19.837 & 1.722 & 19.725 & 2.464 0.019 & 0.328 0.005 + n1419 & 1557 & @xmath365 & 0.056 & 1.110 0.018 & 17.81 0.30 & 1.63 & 1.056 & 19.503 & 0.988 & 19.260 & 2.141 0.029 & 0.242 0.006 + n1426 & 1443 & @xmath365 & 0.072 & 1.161 0.009 & 20.10 0.26 & 1.27 & 1.485 & 20.104 & 1.476 & 20.090 & 2.172 0.025 & 0.258
0.006 + n1427 & 1430 & @xmath364 & 0.051 & 1.152 0.018 & 20.50 0.30 & 1.45 & 1.598 & 20.174 & 1.549 & 20.006 & 2.197 0.018 & 0.244 0.005 + n1439 & 1670 & @xmath365 & 0.129 & 1.131 0.009 & 20.38 0.24 & 1.25 & 1.674 & 20.840 & 1.893 & 21.605 & 2.140 0.022 & 0.271 0.005 + n1521 & 4165 & @xmath365 & 0.176 & 1.152 0.029 & 22.12 0.41 & 2.81 & 1.515 & 20.378 & 1.523 & 20.397 & 2.354 0.029 & 0.274 0.006 + n1527 & 1165 & @xmath363 & 0.054 & 1.230 0.016 & 20.50 0.28 & 0.95 & 1.474 & 19.378 & 1.493 & 19.424 & 2.211 0.025 & 0.262 0.007 + n1537 & 1371 & @xmath363 & 0.107 & 1.096 0.018 & 20.45 0.26 & 1.72 & 1.547 & 19.652 & & & 2.198 0.029 & 0.271 0.006 + n1543 & 1088 & @xmath362 & 0.117 & 1.173 0.016 & 20.92 0.24 & 1.11 & 1.711 & 20.086 & 1.685 & 20.029 & 2.181 0.025 & 0.280 0.007 + n1549 & 1153 & @xmath365 & 0.055 & 1.168 0.016 & 21.54 0.25 & 1.13 & 1.752 & 19.601 & 1.815 & 19.814 & 2.320 0.019 &
+ n1553 & 1280 & @xmath362 & 0.053 & 1.159 0.016 & 21.47 0.24 & 1.31 & 1.533 & 18.403 & 1.555 & 18.480 & 2.230 0.036 & 0.268 0.010 + n1574 & 1042 & @xmath363 & 0.070 & 1.162 0.016 & 20.87 0.28 & 1.10 & 1.590 & 19.455 & 1.717 & 19.884 & 2.329 0.036 & 0.290 0.010 + n1587 & 3667 & @xmath365 & 0.308 & 1.215 0.012 & 21.43 0.87 & 3.33 & 1.339 & 19.578 & 1.323 & 19.502 & 2.341 0.039 & 0.315
0.007 + n1596 & 1523 & @xmath362 & 0.041 & 1.171 0.016 & 19.33 0.24 & 1.57 & 1.211 & 18.563 & 1.209 & 18.599 & 2.212 0.036 & 0.275 0.010 + n1653 & 4343 & @xmath364 & 0.178 & 1.167 0.011 & 21.37 0.32 & 3.00 & 1.407 & 20.151 & 1.408 & 20.157 & 2.342 0.055 & 0.284 0.010 + n1700 & 3881 & @xmath365 & 0.185 & 1.163 0.011 & 21.57 0.24 & 2.88 & 1.344 & 19.213 & 1.301 & 19.027 & 2.372 0.009 & 0.279 0.004 + n2271 & 2588 & @xmath363 & 0.527 & 1.234 0.029 & 20.64 0.32 & 1.93 & 1.241 & 19.300 & 1.246 & 19.296 & 2.194 0.047 & 0.284 0.012 + n2293 & 1994 & @xmath361 & 0.518 & 1.237 0.013 & 20.44 0.45 & 2.03 & 1.664 & 20.284 & 1.488 & 19.714 & 2.424 0.020 & 0.309 0.005 + n2305 & 3571 & @xmath365 & 0.327 & 1.225 0.027 & 20.61 0.27 & 3.60 & 1.561 & 20.254 & 1.473 & 19.958 & 2.361 0.047 & 0.299 0.012 + n2325 & 2248 & @xmath365 & 0.508 & 1.164 0.014 & 21.03 0.24 & 2.14 & 1.747 & 20.609 & 1.568 & 20.059 & 2.143 0.036 & 0.287 0.008 + n2380 & 1775 & @xmath362 & 1.318 & 1.110 0.009 & 21.17 0.40 & 1.63 & 1.613 & 19.567 & 1.537 & 19.313 & 2.291 0.029 & 0.271 0.008 + n2434 & 1327 & @xmath365 & 1.073 & 1.098 0.055 & 20.53 0.25 & 1.43 & 1.633 & 19.868 & 1.565 & 19.674 & 2.321 0.022 & 0.247 0.008 + n2549 & 1056 & @xmath362 & 0.282 & 1.163 0.012 & 18.85 0.34 & 1.39 & 1.229 & 18.687 & 1.238 & 18.663 &
& + n2592 & 1989 & @xmath365 & 0.261 & 1.205 0.010 & 19.37 0.47 & 2.26 & 1.028 & 18.893 & 1.122 & 19.147 &
& + n2634 & 2269 & @xmath365 & 0.094 & 1.154 0.022 & 20.50 0.70 & 2.22 & 1.636 & 21.126 & 1.605 & 21.056 & 2.250 0.017 & 0.279 0.004 + n2681 & 715 & 0 & 0.098 & 1.041 0.010 & 20.27 0.39 & 0.74 & 1.949 & 20.926 & 1.892 & 20.718 & & + n2695 & 1825 & @xmath362 & 0.077 & 1.183 0.051 & 20.32 0.38 & 1.64 & 1.279 & 19.599 & 1.312 & 19.692 & 2.278 0.055 & 0.311 0.010 + n2699 & 1825 & @xmath365 & 0.087 & 1.152 0.051 & 19.27 0.24 & 1.68 & 1.030 & 18.855 & 0.962 & 18.705 & 2.114 0.055 & 0.266 0.010 + n2768 & 1363 & @xmath365 & 0.192 & 1.144 0.027 & 21.04 0.28 & 0.96 & 1.604 & 19.515 & 1.589 & 19.561 & 2.252 0.021 & 0.262 0.005 + n2784 & 708 & @xmath362 & 0.927 & 1.188 0.023 & 20.29 0.30 & 1.37 & 1.608 & 18.743 & 1.671 & 18.961 &
& + n2787 & 689 & @xmath361 & 0.567 & 1.194 0.019 & 18.75 0.40 & 1.18 & 1.392 & 18.622 & 1.520 & 19.111 & & + n2865 & 2581 & @xmath365 & 0.356 & 1.105 0.019 & 20.82 0.26 & 2.14 & 1.439 & 19.838 & 1.488 & 19.954 & 2.246 0.008 & 0.188 0.003 + n2904 & 2387 & @xmath363 & 0.544 & 1.201 0.041 & 19.28 0.24 & 2.07 & 1.203 & 19.658 & 1.146 & 19.483 &
& + n2950 & 1327 & @xmath362 & 0.072 & 1.110 0.019 & 19.32 0.32 & 2.23 & 1.360 & 18.968 & 1.378 & 18.988 &
& + n2962 & 2117 & @xmath361 & 0.249 & 1.176 0.017 & 20.32 0.38 & 2.06 & 1.569 & 20.647 & 1.700 & 21.174 & & + n2974 & 1924 & @xmath365 & 0.235 & 1.203 0.015 & 20.96 0.30 & 2.15 & 1.656 & 20.060 & 1.575 & 19.774 & 2.366 0.023 & 0.290 0.004 + n3031 & @xmath3638 & 2 & 0.347 & 1.187 0.011 & 19.07 0.32 & 0.04 & 1.405 & 16.930 & 1.412 & 17.041 &
& + n3032 & 1561 & @xmath362 & 0.073 & 1.073 0.019 & 18.78 0.33 & 1.71 & 1.525 & 20.979 & 1.311 & 20.195 &
& + n3056 & 1047 & @xmath361 & 0.386 & 1.073 0.023 & 17.90 0.30 & 1.98 & 1.310 & 19.510 & 1.613 & 20.391 &
& + n3073 & 1154 & @xmath363 & 0.044 & 1.007 0.019 & 18.29 0.94 & 1.51 & 1.417 & 21.538 & 1.362 & 21.291 & & + n3078 & 2506 & @xmath365 & 0.307 & 1.209 0.017 & 21.66 0.35 & 2.67 & 1.522 & 19.769 & 1.563 & 19.900 & 2.399 0.019 & 0.321 0.005 + n3087 & 2662 & @xmath364 & 0.454 & 1.164 0.019 & 21.32 0.27 & 2.23 & 1.501 & 19.770 & 1.360 & 19.281 & 2.433 0.055 & 0.279 0.010 + n3115 & 698 & @xmath363 & 0.205 & 1.183 0.010 & 20.46 0.21 & 0.95 & 1.590 & 18.418 & 1.595 & 18.379 & 2.394 0.009 & 0.289 0.003 + n3136 & 1647 & @xmath365 & 1.027 & 1.095 0.033 & 21.24 0.25 & 1.39 & 1.722 & 19.848 & 1.700 & 19.751 & 2.376 0.022 & 0.262 0.008 + n3156 & 1296 & @xmath362 & 0.148 & 1.011 0.011 & 18.28 0.24 & 1.34 & 1.274 & 19.950 & 1.193 & 19.732 & 2.035 0.039 & 0.080 0.007 + n3193 & 1378 & @xmath365 & 0.111 & 1.174 0.009 & 21.36 0.26 & 0.85 & 1.472 & 19.603 & 1.453 & 19.506 & 2.302 0.055 & 0.282 0.010 + n3226 & 1275 & @xmath365 & 0.098 & 1.178 0.015 & 20.59 0.30 & 1.33 & 1.792 & 21.199 & & & 2.296 0.039 & 0.294
0.007 + n3245 & 1358 & @xmath362 & 0.108 & 1.139 0.023 & 20.03 0.26 & 1.79 & 1.302 & 18.844 & 1.381 & 19.129 & & + n3250 & 2883 & @xmath365 & 0.445 & 1.226 0.019 & 22.26 0.24 & 2.14 & 1.451 & 19.322 & 1.418 & 19.212 & 2.431 0.018 & 0.308 0.006 + n3257 & 3023 & @xmath363 & 0.334 & 1.192 0.041 & 19.77 0.27 & 2.61 & 1.100 & 19.409 & 1.060 & 19.304 &
& + n3258 & 2808 & @xmath365 & 0.362 & 1.209 0.034 & 21.20 0.30 & 2.32 & 1.603 & 20.439 & 1.565 & 20.309 & 2.457 0.017 &
0.337 0.005 + n3268 & 2761 & @xmath365 & 0.446 & 1.189 0.023 & 21.29 0.30 & 1.85 & 1.638 & 20.585 & 1.608 & 20.517 & 2.383 0.033 & 0.318 0.009 + n3309 & 4068 & @xmath365 & 0.344 & 1.208 0.021 & 21.79 0.70 & 3.40 & 1.335 & 19.613 & 1.375 & 19.754 & 2.395 0.013 & 0.325 0.005 + n3311 & 3713 & @xmath364 & 0.343 & 1.229 0.034 & 22.76 0.53 & 3.12 & 2.098 & 22.138 & 2.003 & 21.861 & 2.272 0.021 & 0.333 0.010 + n3368 & 899 & 2 & 0.109 & 1.145 0.015 & 20.23 0.28 & 1.63 & 1.698 & 19.140 & 1.713 & 19.241 &
& + n3377 & 689 & @xmath365 & 0.148 & 1.114 0.009 & 19.24 0.21 & 0.89 & 1.610 & 19.715 & 1.618 & 19.748 & 2.105 0.006 & 0.238 0.002 + n3379 & 922 & @xmath365 & 0.105 & 1.193 0.015 & 20.48 0.21 & 1.59 & 1.708 & 19.217 & 1.657 & 18.973 & 2.308 0.005 & 0.289 0.002 + n3384 & 728 & @xmath363 & 0.114 & 1.151 0.018 & 20.13 0.22 & 1.24 & 1.755 & 19.802 & 1.645 & 19.467 & 2.149 0.009 & 0.266 0.003 + n3412 & 867 & @xmath362 & 0.121 & 1.111 0.015 & 18.74 0.23 & 1.40 & 1.375 & 19.035 & 1.700 & 20.063 & 1.995 0.011 & 0.207 0.004 + n3414 & 1476 & @xmath362 & 0.104 & 1.149 0.019 & 20.63 0.38 & 1.66 & 1.546 & 19.927 & 1.619 & 20.200 &
& + n3457 & 1156 & 0 & 0.135 & 1.098 0.015 &
& 1.32 & & & 1.036 & 18.861 & & + lrrcccccccccc galaxy & @xmath100 & @xmath84 & @xmath101 & @xmath102 & @xmath102 & & @xmath103 & @xmath104 & @xmath105 & @xmath106 & @xmath21 @xmath102 & mg@xmath107 @xmath102 + n3489 & 693 & @xmath361 & 0.071 & 1.041 0.023 & 18.87 0.22 & 0.96 & 1.287 & 18.252 & 1.296 & 18.222 & 2.007 0.010 & 0.153 0.004 + n3585 & 1491 & @xmath365 & 0.278 & 1.160 0.016 & 21.46 0.26 & 2.31 & 1.755 & 19.685 & 1.741 & 19.640 & 2.317 0.026 & 0.297 0.007 + n3599 & 850 & @xmath362 & 0.089 & 1.112 0.012 & 19.10 0.26 & 1.14 & 1.620 & 21.183 & 1.720 & 21.574 & 1.883 0.039 & 0.159 0.007 + n3605 & 686 & @xmath365 & 0.089 & 1.118 0.024 & 18.12 0.34 & 1.11 & 0.997 & 19.116 & 0.987 & 19.052 & 2.049 0.039 & 0.184 0.007 + n3607 & 951 & @xmath362 & 0.090 & 1.152 0.010 & 21.35 0.25 & 1.40 & 1.620 & 19.378 & 1.607 & 19.400 & 2.379 0.039 & 0.285 0.007 + n3608 & 1197 & @xmath365 & 0.090 & 1.156 0.009 & 20.51 0.23 & 0.92 & 1.546 & 19.875 & 1.604 & 20.094 & 2.269 0.009 &
0.295 0.004 + n3610 & 1765 & @xmath365 & 0.043 & 1.108 0.015 & 19.91 0.28 & 2.21 & 1.329 & 18.997 & 1.302 & 18.893 & 2.192 0.032 & 0.245 0.006 + n3613 & 2054 & @xmath365 & 0.053 & 1.175 0.015 & 20.97 0.44 & 2.22 & 1.502 & 19.788 & 1.363 & 19.279 & 2.319 0.039 & 0.271 0.007 + n3617 & 2224 & @xmath364 & 0.213 & 1.050 0.023 & 19.15 0.32 & 3.07 & 1.465 & 21.106 & 1.464 & 21.107 &
& + n3640 & 1302 & @xmath365 & 0.191 & 1.140 0.009 & 21.28 0.23 & 1.49 & 1.579 & 19.531 & 1.553 & 19.455 & 2.236 0.039 & 0.247 0.007 + n3641 & 1758 & @xmath365 & 0.182 & 1.131 0.012 & 18.62 0.31 & 1.90 & 1.407 & 21.258 & 1.240 & 20.653 & 2.207 0.055 & 0.284 0.010 + n3818 & 1498 & @xmath365 & 0.155 & 1.124 0.015 & 20.32 0.63 & 1.69 & 1.346 & 19.886 & 1.434 & 20.142 & 2.248 0.010 & 0.305 0.005 + n3904 & 1750 & @xmath365 & 0.311 & 1.156 0.055 & 20.96 0.24 & 1.51 & 1.374 & 19.014 & 1.381 & 19.037 & 2.354 0.021 & 0.293 0.006 + n3923 & 1607 & @xmath365 & 0.358 & 1.194 0.055 & 21.91 0.23 & 1.50 & 1.690 & 19.382 & 1.706 & 19.429 & 2.334 0.021 & 0.297 0.006 + n3928 & 974 & 3 & 0.085 & 1.096 0.015 & 17.95 0.66 & 1.47 & 1.107 & 19.285 & 1.211 & 19.631 &
& + n3941 & 944 & @xmath362 & 0.091 & 1.125 0.013 & 19.28 0.26 & 1.63 & 1.322 & 18.464 & 1.366 & 18.606 &
& + n3962 & 1822 & @xmath365 & 0.201 & 1.145 0.023 & 21.96 0.52 & 2.85 & 1.795 & 20.545 & 1.683 & 20.190 & 2.317 0.039 & 0.298 0.007 + n3990 & 705 & @xmath363 & 0.069 & 1.151 0.019 &
& 0.98 & & & 0.967 & 18.855 &
& + n3998 & 1028 & @xmath362 & 0.069 & 1.194 0.011 & 19.84 0.27 & 1.35 & 1.353 & 18.712 & 1.458 & 19.102 & & + n4024 & 1694 & @xmath363 & 0.182 & 1.141 0.016 & 20.00 0.51 & 3.44 & 1.185 & 19.098 & 1.452 & 20.043 & 2.165 0.039 & 0.241 0.007 + n4033 & 1521 & @xmath365 & 0.205 & 1.113 0.023 & 18.98 0.28 & 2.56 & 1.266 & 19.593 & 1.326 & 19.652 & 2.070 0.039 & 0.236 0.007 + n4105 & 1882 & @xmath365 & 0.263 & 1.171 0.017 & 21.25 0.24 & 1.89 & 1.627 & 19.914 & 1.573 & 19.699 & 2.398 0.028 & 0.292 0.006 + n4111 & 806 & @xmath361 & 0.063 & 1.096 0.015 & 19.07 0.30 & 1.20 & 1.267 & 18.709 & 1.155 & 17.923 &
& + n4125 & 1340 & @xmath365 & 0.082 & 1.174 0.011 & 21.80 0.31 & 1.86 & 1.772 & 19.890 & 1.875 & 20.241 & 2.347 0.055 & 0.281 0.010 + n4138 & 835 & @xmath361 & 0.060 & 1.164 0.013 & 18.95 0.32 & 1.01 & 1.272 & 19.000 & 1.294 & 19.114 &
& + n4150 & 244 & @xmath362 & 0.078 & 1.071 0.017 & 18.06 0.30 & 0.81 & 1.189 & 19.020 & 1.290 & 19.400 &
& + n4203 & 1117 & @xmath363 & 0.052 & 1.195 0.019 & 20.73 0.25 & 1.26 & 1.945 & 20.940 & 1.953 & 20.950 &
& + n4233 & 2371 & @xmath362 & 0.103 & 1.191 0.015 & 20.30 0.95 & 2.43 & 1.203 & 19.364 & 1.342 & 19.891 & & + n4261 & 2200 & @xmath365 & 0.078 & 1.258 0.014 & 22.06 0.27 & 1.28 & 1.505 & 19.321 & 1.380 & 18.892 & 2.483 0.010 & 0.326 0.005 + n4278 & 643 & @xmath365 & 0.123 & 1.161 0.012 & 20.39 0.27 & 0.66 & 1.477 & 18.912 & 1.528 & 19.141 & 2.376 0.010 & 0.276 0.005 + n4283 & 1076 & @xmath365 & 0.109 & 1.178 0.010 & 18.26 0.27 & 1.65 & 0.992 & 18.648 & 1.026 & 18.775 & 1.990 0.039 & 0.251 0.007 + n4291 & 1715 & @xmath365 & 0.158 & 1.175 0.017 & 20.35 0.37 & 1.68 & 1.294 & 19.154 & 1.378 & 19.489 & 2.406 0.055 & 0.299 0.010 + n4339 & 1298 & @xmath365 & 0.110 & 1.200 0.015 & 19.35 0.26 & 2.07 & 1.439 & 19.994 & 1.575 & 20.488 & 1.964 0.039 & 0.242 0.007 + n4346 & 762 & @xmath362 & 0.056 & 1.158 0.012 & 19.00 0.25 & 1.00 & 1.254 & 18.918 & 1.411 & 19.433 &
& + n4365 & 1240 & @xmath365 & 0.092 & 1.222 0.017 & 21.66 0.24 & 1.35 & 1.735 & 19.755 & 1.771 & 19.849 & 2.393 0.009 & 0.307 0.003 + n4373 & 3392 & @xmath363 & 0.348 & 1.135 0.023 & 21.96 0.25 & 2.81 & 1.533 & 19.651 & 1.557 & 19.752 & 2.384 0.019 & 0.286 0.008 + n4374 & 1033 & @xmath365 & 0.175 & 1.191 0.008 & 21.81 0.22 & 1.30 & 1.776 & 19.427 & 1.721 & 19.201 & 2.448 0.006 & 0.290 0.002 + n4379 & 1071 & @xmath363 & 0.101 & 1.185 0.017 & 18.70 0.44 & 1.42 & 1.196 & 19.043 & 1.209 & 19.102 &
& + n4382 & 758 & @xmath361 & 0.134 & 1.150 0.022 & 22.09 0.22 & 0.76 & 2.029 & 20.216 & 2.179 & 20.674 & 2.207 0.015 & 0.193 0.006 + n4386 & 1649 & @xmath362 & 0.166 & 1.196 0.019 & 20.30 0.52 & 1.87 & 1.310 & 19.455 & 1.440 & 19.932 &
& + n4387 & 583 & @xmath365 & 0.143 & 1.163 0.011 & 18.92 0.75 & 1.46 & 1.069 & 18.971 & 1.030 & 18.842 & 1.918 0.031 & 0.208
0.005 + n4391 & 1337 & @xmath363 & 0.078 & 1.112 0.015 & 18.27 0.33 & 1.38 & 1.076 & 19.633 & 1.302 & 20.416 & & + n4406 & @xmath36221 & @xmath365 & 0.128 & 1.167 0.008 & 21.82 0.23 & 0.85 & 1.991 & 20.235 & 2.066 & 20.491 & 2.318 0.008 & 0.247 0.003 + n4419 & @xmath36182 & 1 & 0.143 & 1.126 0.026 & 19.03 0.29 & 1.15 & 1.296 & 18.816 & 1.293 & 18.966 &
& + n4434 & 1068 & @xmath365 & 0.096 & 1.125 0.015 & 19.20 0.25 & 0.61 & 1.075 & 19.017 & 1.122 & 19.128 & 2.079 0.016 & 0.252 0.004 + n4441 & 1439 & @xmath361 & 0.086 & 1.005 0.012 & 17.23 0.48 & 1.51 & 1.206 & 20.286 & 1.199 & 20.306 &
& + n4458 & 662 & @xmath365 & 0.103 & 1.140 0.011 & 18.36 0.22 & 0.66 & 1.222 & 19.709 & 1.313 & 19.983 & 1.975 0.016 & 0.203 0.004 + n4460 & 558 & @xmath361 & 0.082 & 1.011 0.015 & 16.80 0.26 & 0.88 & 1.393 & 20.204 & 1.861 & 21.744 & & + n4468 & 895 & @xmath363 & 0.198 & 1.045 0.015 & 17.40 0.23 & 1.35 & 1.495 & 21.378 & 1.288 & 20.674 & & 0.141 0.010 + n4472 & 997 & @xmath365 & 0.096 & 1.218 0.011 & 22.18 0.21 & 1.08 & 1.796 & 19.020 & 1.846 & 19.187 & 2.452 0.006 & 0.299 0.002 + n4473 & 2236 & @xmath365 & 0.123 & 1.158 0.012 & 20.37 0.23 & 1.62 & 1.542 & 19.169 & 1.569 & 19.284 & 2.255 0.009 & 0.299 0.003 + n4476 & 1955 & @xmath363 & 0.123 & 1.048 0.017 & 17.89 0.24 & 1.40 & 1.234 & 19.744 & 1.210 & 19.700 & 1.564 0.039 & 0.148 0.007 + n4478 & 1370 & @xmath365 & 0.107 & 1.164 0.019 & 19.35 0.33 & 1.49 & 1.097 & 18.320 & 1.254 & 18.896 & 2.122 0.007 & 0.245 0.003 + n4486 & 1292 & @xmath364 & 0.096 & 1.244 0.012 & 22.11 0.24 & 0.74 & 1.793 & 19.192 & 1.779 & 19.178 & 2.535 0.009 & 0.282 0.004 + n4489 & 960 & @xmath365 & 0.121 & 1.046 0.015 & 17.93 0.24 & 1.48 & 1.230 & 19.776 & 1.408 & 20.401 & 1.692 0.010 & 0.191 0.004 + n4494 & 1350 & @xmath365 & 0.093 & 1.139 0.010 & 20.69 0.22 & 0.91 & 1.638 & 19.435 & 1.724 & 19.738 & 2.086 0.055 & 0.262 0.010 + n4526 & 602 & @xmath362 & 0.096 & 1.188 0.021 & 21.06 0.26 & 0.92 & 1.587 & 19.033 & 1.511 & 18.762 &
& + n4546 & 1037 & @xmath363 & 0.147 & 1.155 0.013 & 19.67 0.28 & 1.20 & 1.403 & 18.945 & 1.403 & 18.992 &
& + n4550 & 381 & @xmath362 & 0.168 & 1.078 0.011 & 18.41 0.27 & 1.11 & 1.118 & 18.659 & 1.149 & 18.833 & & + n4551 & 1198 & @xmath365 & 0.166 & 1.170 0.009 & 18.81 0.26 & 1.05 & 1.171 & 19.151 & 1.150 & 19.078 & 2.011 0.016 & 0.251 0.004 + n4552 & 322 & @xmath365 & 0.177 & 1.194 0.015 & 20.49 0.23 & 1.35 & 1.437 & 18.682 & 1.461 & 18.674 & 2.382
0.008 & 0.297 0.003 + n4564 & 1165 & @xmath365 & 0.151 & 1.161 0.009 & 19.28 0.26 & 1.12 & 1.264 & 18.796 & 1.337 & 19.030 & 2.201 0.009 & 0.310 0.003 + n4565 & 1227 & 3 & 0.067 & 1.128 0.027 & 21.04 0.23 & 0.99 & 1.834 & 20.050 & 1.873 & 20.457 &
& + n4578 & 2284 & @xmath362 & 0.092 & 1.127 0.015 & 19.66 0.22 & 1.12 & 1.706 & 20.894 & 1.840 & 21.357 & & + lrrcccccccccc galaxy & @xmath100 & @xmath84 & @xmath101 & @xmath102 & @xmath102 & & @xmath103 & @xmath104 & @xmath105 & @xmath106 & @xmath21 @xmath102 & mg@xmath107 @xmath102 + n4594 & 1128 & 1 & 0.221 & 1.175 0.031 & & 1.04 & & & 1.994 & 19.044 & 2.361 0.022 & 0.310 0.009 + n4600 & 787 & @xmath362 & 0.116 & 1.141 0.017 & 16.08 0.28 & 1.66 & 1.210 & 20.075 & 1.205 & 20.032 & & + n4616 & 4585 & @xmath364 & 0.549 & 1.196 0.021 & 20.31 0.29 & 2.36 & 1.408 & 20.701 & 1.184 & 20.111 & 2.243 0.031 & 0.267 0.014 + n4620 & 1178 & @xmath362 & 0.126 & 1.048 0.019 & 17.89 0.35 & 1.51 & 1.381 & 20.951 & 2.091 & 23.017 &
& + n4621 & 444 & @xmath365 & 0.144 & 1.172 0.018 & 21.22 0.27 & 1.62 & 1.697 & 19.520 & 1.672 & 19.464 & 2.336 0.014 & 0.308 0.004 + n4636 & 937 & @xmath365 & 0.120 & 1.233 0.012 & 21.45 0.23 & 1.29 & 1.974 & 20.493 & 2.185 & 21.091 & 2.297 0.009 & 0.300 0.003 + n4638 & 1148 & @xmath363 & 0.110 & 1.149 0.013 & 19.89 0.32 & 1.23 & 1.190 & 18.542 & 1.246 & 18.758 &
& + n4645 & 2593 & @xmath364 & 0.639 & 1.188 0.025 & 20.66 0.24 & 2.13 & 1.364 & 19.531 & 1.371 & 19.557 & 2.283 0.009 & 0.278 0.004 + n4649 & 1095 & @xmath365 & 0.115 & 1.232 0.023 & 21.83 0.22 & 0.86 & 1.619 & 18.634 & 1.630 & 18.702 & 2.531 0.008 & 0.329 0.003 + n4684 & 1589 & @xmath361 & 0.119 & 1.100 0.015 & 18.35 0.26 & 2.68 & 1.259 & 19.152 & 1.203 & 19.030 &
& + n4696 & 2958 & @xmath364 & 0.490 & 1.203 0.014 & 22.40 0.25 & 2.53 & 1.760 & 20.212 & 1.774 & 20.277 & 2.398 0.018 & 0.276 0.007 + n4697 & 1210 & @xmath365 & 0.132 & 1.157 0.010 & 20.71 0.23 & 2.19 & 1.850 & 19.749 & 1.822 & 19.650 & 2.215 0.008 & 0.278 0.004 + n4709 & 4650 & @xmath365 & 0.513 & 1.199 0.011 & 21.77 0.30 & 1.80 & 1.776 & 20.860 & 1.896 & 21.236 & 2.388 0.008 & 0.328 0.003 + n4725 & 1207 & 2 & 0.051 & 1.209 0.023 & 20.71 0.38 & 1.82 & 1.898 & 20.516 & 1.933 & 20.677 &
& + n4729 & 3293 & @xmath364 & 0.543 & 1.219 0.028 & 20.34 0.34 & 3.04 & 1.454 & 20.451 & 1.428 & 20.367 & 2.154 0.009 & 0.276 0.004 + n4742 & 1321 & @xmath365 & 0.176 & 1.045 0.017 & 18.47 0.24 & 1.84 & 1.164 & 18.588 & 1.158 & 18.421 & 1.969 0.024 & 0.158 0.006 + n4754 & 1374 & @xmath363 & 0.143 & 1.178 0.011 & 20.27 0.23 & 1.82 & 1.539 & 19.510 & 1.661 & 19.921 &
& + n4767 & 3005 & @xmath365 & 0.461 & 1.175 0.028 & 20.85 0.29 & 2.80 & 1.329 & 19.289 & 1.361 & 19.377 & 2.319 0.016 & 0.291 0.005 + n4802 & 1013 & @xmath362 & 0.206 & 1.004 0.018 & 17.66 0.24 & 1.63 & 1.347 & 19.477 & 1.344 & 19.415 &
& + n4826 & 414 & 2 & 0.178 & 1.029 0.011 & 19.88 0.27 & 0.77 & 1.816 & 18.803 & & & & + n4946 & 3036 & @xmath364 & 0.475 & 1.190 0.021 & 21.33 0.24 & 2.45 & 1.449 & 20.295 & 1.359 & 19.964 & 2.309 0.025 & 0.303 0.010 + n5011 & 3099 & @xmath365 & 0.427 & 1.189 0.027 & 21.95 0.27 & 2.73 & 1.595 & 20.112 & 1.466 & 19.645 & 2.390 0.017 & 0.272 0.006 + n5044 & 2704 & @xmath365 & 0.301 & 1.210 0.027 & 21.23 0.32 & 2.18 & 1.406 & 19.367 & 1.412 & 19.422 & 2.365 0.039 & 0.320 0.007 + n5061 & 2040 & @xmath365 & 0.296 & 1.099 0.023 & 21.07 0.30 & 3.13 & 1.667 & 19.666 & & & 2.301 0.022 & 0.241 0.006 + n5090 & 3376 & @xmath365 & 0.621 & 1.210 0.032 & 22.64 0.39 & 2.77 & 2.027 & 21.463 & 2.013 & 21.360 & 2.440 0.021 & 0.315 0.007 + n5193 & 3644 & @xmath365 & 0.243 & 1.164 0.034 & 20.54 0.32 & 2.91 & 1.371 & 19.832 & 1.323 & 19.677 & 2.338 0.019 & 0.300 0.008 + n5273 & 1089 & @xmath362 & 0.043 & 1.142 0.017 & 19.14 0.31 & 1.38 & 1.593 & 20.694 & 2.049 & 22.049 &
& + n5322 & 1804 & @xmath365 & 0.061 & 1.183 0.011 & 22.02 0.30 & 1.74 & 1.723 & 20.042 & 1.490 & 19.168 & 2.342 0.039 & 0.265 0.007 + n5338 & 777 & @xmath362 & 0.118 & 1.019 0.023 & 16.57 0.34 & 2.05 & 1.477 & 21.524 & 1.496 & 21.565 & & + n5485 & 1985 & @xmath362 & 0.072 & 1.180 0.017 & 20.35 0.39 & 2.32 & 1.440 & 19.871 & 1.419 & 19.825 & 2.185 0.055 & 0.283 0.010 + n5557 & 3258 & @xmath365 & 0.025 & 1.202 0.021 & 21.30 0.34 & 3.48 & 1.488 & 19.768 & 1.401 & 19.455 & 2.412 0.039 & 0.305 0.007 + n5574 & 1582 & @xmath363 & 0.135 & 1.054 0.011 & 18.34 0.64 & 2.18 & 1.036 & 19.053 & 1.096 & 19.235 &
& + n5576 & 1555 & @xmath365 & 0.136 & 1.098 0.010 & 20.60 0.23 & 1.71 & 1.585 & 19.909 & 1.545 & 19.722 & 2.264 0.055 & 0.240 0.010 + n5582 & 1314 & @xmath365 & 0.059 & 1.145 0.013 & 20.12 0.30 & 1.21 & 1.526 & 20.573 & 1.537 & 20.635 & 2.124 0.023 & 0.274 0.006 + n5631 & 1950 & @xmath362 & 0.087 & 1.122 0.019 & 20.27 0.30 & 2.41 & 1.372 & 19.467 & 1.479 & 19.865 &
& + n5638 & 1648 & @xmath365 & 0.143 & 1.169 0.011 & 20.43 0.30 & 1.79 & 1.407 & 19.605 & 1.381 & 19.506 & 2.192 0.009 & 0.301 0.004 + n5687 & 2200 & @xmath363 & 0.048 & 1.174 0.013 & 20.26 0.54 & 2.68 & 1.614 & 20.965 & 1.704 & 21.296 &
& + n5770 & 1464 & @xmath362 & 0.167 & 1.120 0.017 & 18.48 0.34 & 1.97 & 1.160 & 19.375 & & & & + n5812 & 2066 & @xmath365 & 0.376 & 1.213 0.015 & 21.18 0.34 & 2.35 & 1.608 & 20.138 & 1.596 & 20.010 & 2.301 0.009 & 0.313 0.004 + n5813 & 1963 & @xmath365 & 0.246 & 1.189 0.014 & 22.06 0.26 & 1.20 & 1.847 & 20.709 & 1.980 & 21.177 & 2.356 0.009 & 0.297 0.004 + n5831 & 1683 & @xmath365 & 0.257 & 1.140 0.010 & 20.32 0.25 & 1.14 & 1.478 & 20.002 & 1.505 & 20.099 & 2.211 0.009 & 0.278 0.004 + n5839 & 1211 & @xmath362 & 0.227 & 1.190 0.011 & 19.12 0.36 & 1.24 & 1.215 & 19.749 & 1.432 & 20.532 &
& + n5846 & 1709 & @xmath365 & 0.238 & 1.227 0.007 & 21.90 0.28 & 1.02 & 1.742 & 19.996 & 1.725 & 19.930 & 2.365 0.008 & 0.313 0.003 + n5866 & 672 & @xmath361 & 0.058 & 1.121 0.009 & 20.27 0.22 & 0.87 & 1.566 & 19.156 & 1.563 & 19.230 & & + n5869 & 2110 & @xmath362 & 0.233 & 1.161 0.014 & 19.73 1.24 & 2.24 & 1.290 & 19.556 & 1.321 & 19.698 & & + n5898 & 2090 & @xmath365 & 0.629 & 1.169 0.009 & 21.05 0.32 & 2.12 & 1.466 & 19.511 & 1.506 & 19.659 & 2.353 0.013 & 0.302 0.004 + n5903 & 2518 & @xmath365 & 0.638 & 1.142 0.011 & 21.37 0.30 & 2.28 & 1.629 & 20.182 & 1.629 & 20.174 & 2.306 0.016 & 0.287
0.005 + n5982 & 2936 & @xmath365 & 0.078 & 1.166 0.010 & 21.77 0.68 & 2.83 & 1.537 & 19.986 & 1.558 & 20.067 & 2.385 0.023 & 0.285 0.006 + n6305 & 2770 & @xmath363 & 0.441 & 1.147 0.051 & 18.92 0.28 & 2.33 & 1.206 & 19.308 & 1.055 & 18.764 & 2.242 0.041 & 0.249 0.014 + n6411 & 3690 & @xmath365 & 0.249 & 1.156 0.008 & 20.13 0.40 & 3.35 & 1.467 & 20.342 & 1.429 & 20.238 & 2.236 0.016 & 0.269 0.004 + n6483 & 4880 & @xmath365 & 0.257 & 1.152 0.051 & 20.62 0.47 & 3.33 & 1.272 & 19.850 & 1.122 & 19.285 & 2.353 0.042 & 0.277 0.010 + n6502 & 4880 & @xmath364 & 0.267 & 1.191 0.041 & 20.84 0.30 & 3.81 & 1.398 & 20.367 & 1.259 & 19.851 &
& + n6548 & 2169 & @xmath362 & 0.353 & 1.244 0.009 & 20.68 0.28 & 1.77 & 1.655 & 20.686 & 1.918 & 21.531 & & + n6684 & 865 & @xmath362 & 0.291 & 1.116 0.015 & 20.05 0.30 & 1.07 & 1.719 & 19.912 & 2.002 & 20.814 & 2.006 0.013 & 0.222 0.004 + n6702 & 4725 & @xmath365 & 0.473 & 1.149 0.006 & 21.07 0.30 & 2.71 & 1.401 & 20.193 & 1.527 & 20.634 & 2.250 0.008 & 0.254 0.004 + n6703 & 2370 & @xmath363 & 0.383 & 1.164 0.006 & 20.85 0.35 & 2.02 & 1.590 & 20.097 & 1.598 & 20.123 & 2.266 0.008 & 0.267 0.003 + n6851 & 3043 & @xmath365 & 0.202 & 1.137 0.016 & 20.49 0.28 & 2.04 & 1.253 & 19.286 & 1.235 & 19.223 & 2.267 0.042 & 0.279 0.010 + n6861 & 2819 & @xmath363 & 0.234 & 1.221 0.019 & 21.04 0.40 & 2.47 & 1.359 & 19.149 & 1.312 & 19.005 &
& + n6868 & 2876 & @xmath365 & 0.237 & 1.230 0.009 & 21.43 0.31 & 2.77 & 1.479 & 19.304 & 1.501 & 19.385 & 2.464 0.030 & 0.333 0.007 + n6869 & 2739 & @xmath362 & 0.790 & 1.164 0.007 & 20.62 0.29 & 2.34 & 1.375 & 19.658 & 1.285 & 19.352 &
& + n6876 & 3951 & @xmath365 & 0.195 & 1.178 0.009 & 21.89 0.43 & 3.36 & 1.429 & 19.520 & 1.467 & 19.650 & 2.374 0.030 & 0.300 0.007 + n6909 & 2753 & @xmath364 & 0.164 & 1.064 0.020 & 19.95 0.23 & 2.00 & 1.420 & 20.267 & 1.231 & 19.585 & 2.178 0.042 & 0.218 0.010 + n7029 & 2818 & @xmath365 & 0.159 & 1.138 0.032 & 20.77 0.50 & 2.27 & 1.249 & 19.133 & 1.270 & 19.189 & 2.308 0.030 & 0.266 0.007 + n7041 & 1877 & @xmath363 & 0.173 & 1.139 0.055 & 20.21 0.27 & 1.66 & 1.347 & 19.337 & & & & + lrrcccccccccc galaxy & @xmath100 & @xmath84 & @xmath101 & @xmath102 & @xmath102 & & @xmath103 & @xmath104 & @xmath105 & @xmath106 & @xmath21 @xmath102 & mg@xmath107 @xmath102 + n7049 & 2158 & @xmath362 & 0.237 & 1.174 0.033 & 21.54 0.25 & 1.58 & 1.448 & 19.002 & 1.483 & 19.237 &
& + n7097 & 2404 & @xmath365 & 0.085 & 1.176 0.022 & 20.26 0.27 & 1.55 & 1.182 & 19.133 & 1.191 & 19.192 &
& + n7144 & 1921 & @xmath365 & 0.091 & 1.161 0.009 & 20.63 0.22 & 1.65 & 1.592 & 20.140 & 1.652 & 20.361 & 2.272 0.042 & 0.291 0.010 + n7145 & 1874 & @xmath365 & 0.091 & 1.133 0.016 & 20.12 0.28 & 1.76 & 1.583 & 20.372 & 1.544 & 20.236 & 2.125 0.030 & 0.258 0.007 + n7168 & 2747 & @xmath365 & 0.101 & 1.184 0.022 & 20.50 0.29 & 2.36 & 1.336 & 19.870 & 1.310 & 19.756 & & + n7173 & 2501 & @xmath364 & 0.114 & 1.150 0.017 & 19.91 0.26 & 1.89 & 1.168 & 19.211 & 1.164 & 19.207 & 2.293 0.030 & 0.280 0.007 + n7180 & 1479 & @xmath362 & 0.138 & 1.109 0.009 & 17.86 0.32 & 1.35 & 1.015 & 19.083 & 1.178 & 19.644 & 1.987 0.026 & 0.187 0.006 + n7185 & 1838 & @xmath363 & 0.135 & 1.080 0.012 & 18.44 0.41 & 2.18 & 1.474 & 20.884 & 1.628 & 21.430 &
& + n7192 & 2879 & @xmath364 & 0.146 & 1.174 0.032 & 21.17 0.34 & 2.60 & 1.401 & 19.633 & 1.414 & 19.681 & 2.275 0.042 & 0.263 0.010 + n7196 & 3007 & @xmath365 & 0.095 & 1.211 0.022 & 21.71 0.34 & 1.86 & 1.389 & 19.604 & 1.414 & 19.685 & 2.451 0.042 & 0.302 0.010 + n7200 & 2897 & @xmath364 & 0.083 & 1.183 0.018 & 19.24 0.37 & 2.03 & 1.062 & 19.741 & 1.026 & 19.527 & 2.299 0.042 & 0.296 0.010 + n7280 & 1903 & @xmath361 & 0.240 & 1.105 0.009 & 19.56 0.29 & 1.67 & 1.576 & 20.823 & 1.950 & 21.989 &
& + n7302 & 2586 & @xmath363 & 0.299 & 1.126 0.016 & 19.19 0.28 & 2.33 & 1.201 & 19.368 & 1.578 & 20.692 &
& + n7331 & 819 & 3 & 0.392 & 1.120 0.017 & 20.75 0.24 & 0.36 & 1.702 & 19.035 & 1.805 & 19.521 &
& + n7332 & 1207 & @xmath362 & 0.161 & 1.107 0.008 & 19.98 0.28 & 0.61 & 1.242 & 18.648 & 1.252 & 18.639 &
& + n7454 & 2007 & @xmath365 & 0.343 & 1.123 0.012 & 19.66 0.35 & 1.23 & 1.474 & 20.278 & 1.325 & 19.753 & 2.018 0.009 & 0.199 0.004 + n7457 & 822 & @xmath363 & 0.228 & 1.104 0.009 & 18.99 0.28 & 0.36 & 1.626 & 20.343 & 1.831 & 21.015 & & + n7507 & 1548 & @xmath365 & 0.206 & 1.196 0.009 & 21.54 0.25 & 1.20 & 1.600 & 19.494 & 1.472 & 19.000 & 2.369 0.033 & 0.324 0.007 + n7562 & 3608 & @xmath365 & 0.453 & 1.187 0.018 & 21.95 0.85 & 2.97 & 1.295 & 19.278 & 1.386 & 19.486 & 2.394 0.009 & 0.280 0.004 + n7619 & 3747 & @xmath365 & 0.347 & 1.229 0.009 & 22.74 0.37 & 2.46 & 1.656 & 20.330 & 1.680 & 20.414 & 2.510 0.009 & 0.333 0.003 + n7743 & 1658 & @xmath361 & 0.305 & 1.080 0.009 & 19.51 0.25 & 1.03 & 1.522 & 20.145 & 2.105 & 22.072 &
& + n7796 & 3252 & @xmath364 & 0.044 & 1.230 0.022 & 22.21 0.43 & 2.43 & 1.571 & 20.325 & 1.608 & 20.450 &
& + n7814 & 1050 & 2 & 0.192 & 1.245 0.017 & 20.23 0.22 & 0.51 & 1.611 & 19.734 & 1.571 & 19.740 &
& + u07767 & 1331 & @xmath365 & 0.089 & 1.152 0.019 & & 1.67 & & & 1.029 & 19.611 & & + | we perform a series of comparisons between distance - independent photometric and spectroscopic properties used in the surface brightness fluctuations ( sbf ) and fundamental plane ( fp ) methods of early - type galaxy distance estimation .
the data are taken from two recent surveys : the sbf survey of galaxy distances and the streaming motions of abell clusters ( smac ) fp survey .
we derive a relation between colour and _ 2 index using nearly 200 galaxies and discuss implications for galactic extinction estimates and early - type galaxy stellar populations .
we find that the reddenings from schlegel et al .
( 1998 ) for galaxies with @xmath0 mag appear to be overestimated by 510% , but we do not find significant evidence for large - scale dipole errors in the extinction map . in comparison to stellar population models having solar elemental abundance ratios , the galaxies in our sample are generally too blue at a given _ 2 ; we ascribe this to the well - known enhancement of the @xmath1-elements in luminous early - type galaxies .
we confirm a tight relation between stellar velocity dispersion @xmath2 and the sbf ` fluctuation count ' parameter , which is a luminosity - weighted measure of the total number of stars in a galaxy .
the correlation between and @xmath2 is even tighter than that between _ 2 and @xmath2 .
finally , we derive fp photometric parameters for 280 galaxies from the sbf survey data set .
comparisons with external sources allow us to estimate the errors on these parameters and derive the correction necessary to bring them onto the smac system .
the data are used in a companion paper which compares the distances derived from the fp and sbf methods .
[ firstpage ] galaxies : distances and redshifts galaxies : elliptical and lenticular , cd galaxies : fundamental parameters galaxies : stellar content |
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 has been a triumph of modern astrophysics @xcite . 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 .
@xcite reported on neon abundances relative to oxygen measured in a sample of nearby solar - like stars from their x - ray spectra .
the results were very similar for all stars and substantially larger than the recommended solar value of ne / o=0.15 by number @xcite .
if the ne / o abundance in these stars is adopted for the sun the models are brought back into agreement with helioseismology @xcite .
however , the stellar sample used by @xcite for their abundance analysis contains mostly active stars and binary systems , which are not comparable to the sun in terms of activity and tend to exhibit different outer atmosphere chemical fractionation patterns .
hence it is also necessary to determine the ne / o ratio in relatively inactive stars that are comparable to the sun . in this work
we analyse a sample three times larger than that of @xcite , including about 20 low activity stars ( @xmath1 ) .
illustrated was computed omitting this point . ]
our sample is based on analyses of high resolution _
chandra _ and _ xmm - newton _ spectra culled from the recent literature for about 60 late - type stars , including dwarfs , giants , tidally - interacting binaries and pre - main sequence stars .
it covers a wide range of activity regimes ( @xmath2 ) and spectral types ( mid - f to mid - m ) .
we examined the coronal fe / o and ne / o ratios for these stars using the abundances derived from differential emission measure distribution analyses performed in the various literature studies . in the following , the coronal fe / o and ne / o abundance ratios are expressed relative to the asplund et al.@xcite solar photospheric mixture for the stellar sample .
the mean ne / o of our stellar sample is about twice the value recently recommended for the solar photosphere @xcite and the average of estimates for the solar corona , and about 0.1 dex lower than the mean estimated by @xcite from line ratios for a sample of mostly active stars , but is within the systematic uncertainties of the latter study ( see fig.1 ) .
there is no significant dependence of coronal ne / o on the stellar activity indicators @xmath3 or rossby number , though the lowest activity stars of the sample present some evidence of a decline in ne / o for @xmath4 and solar - like activity , as has also been suggested in a recent study of the same stars @xcite . for higher activities
ne / o , for dwarfs is statistically consistant with a constant value [ ne / o]=0.3 with observational scatter of about 0.1 dex .
in contrast to ne / o , the fe / o ratios vary by more than order of magnitude with a clear trend of decreasing fe / o with increasing activity ( see fig.2 ) . low to intermediate activity stars with @xmath5 tend to have enhanced fe / o compared with photospheric expectations ( quasi - solar for most of the stars in the sample ) indicative of a solar - like fip effect . at higher activity levels ,
fe / o exhibits a strong decline with increasing activity in all the types of stars sampled , including pre - main sequence stars .
the observed abundance patterns are reminiscent of evidence for a dependence of the solar ne / o and fe / o ratios on active region plasma temperature @xcite and indicate a universal fractionation process is at work .
the apparently constant stellar ne / o over a wide range in activity level combined with variability observed in the solar coronal ne / o leads us to suggest that ne is generally depleted in the solar outer atmosphere , whereas photospheric values are reflected in more active stellar coronae .
if this is the case , the solution to the recent `` solar model problem '' - solar models using the @xcite composition containing too little metal opacity to match helioseismic observations - is a combination of the @xcite o abundance downward revision being too large , and the ne abundance being underestimated for the sun by about a factor of 2 . | a survey of ne , o and fe coronal abundances culled from the recent literature for about 60 late - type stars confirms that the ne / o ratio of stellar outer atmospheres is about two times the value recently recommended by asplund et al .
the mean ne / o remains flat from the most active stars down to at least intermediate activity levels ( @xmath0 ) , with some evidence for a decline toward the lowest activity levels sampled .
the abundances surveyed are all based on emission measure distribution analyses and the mean ne / o is about 0.1 dex lower than that found from line ratios in the seminal study of mostly active stars by drake & testa ( 2005 ) , but is within the systematic uncertainties of that study .
we also confirm a pattern of strongly decreasing fe / o with increasing stellar activity .
the observed abundance patterns are reminiscent of the recent finding of a dependence of the solar ne / o and fe / o ratios on active region plasma temperature and indicate a universal fractionation process is at work .
the firm saturation in stellar ne / o at higher activity levels combined with variability in the solar coronal ne / o leads us to suggest that ne is generally depleted in the solar outer atmosphere and photospheric values are reflected in active stellar coronae .
the solution to the recent solar model problem would then appear to lie in a combination of the asplund et al .
( 2005 ) o abundance downward revision being too large , and the ne abundance being underestimated for the sun by about a factor of 2 .
address = instituto de astrofsica de canarias , e-38205 la laguna , tenerife , spain , altaddress = gtc calp , e-38712 san antonio , la palma , spain , [email protected] address = harvard - smithsonian center for astrophysics , 60 garden street , cambridge , ma 02138 , usa address= |
the ccds used in the wide field planetary camera 2 ( `` wfpc2 '' ) of the _ hubble space telescope _ are known to have imperfect charge - transfer efficiency ( `` cte '' ) .
this property of the camera has adverse photometric consequences that have proven difficult to characterize quantitatively .
the source of the problem appears to be chemical impurities in the crystal lattice of the detectors that trap some photoelectrons during exposure and readout , resulting in an underestimate of the original luminous flux impingent on the camera .
these missing electrons eventually diffuse out of the traps and contribute to the background noise of that same exposure or a subsequent one .
since the charge packets that represent stellar images recorded at high row numbers must be clocked across a large fraction of the chip during readout , it is to be expected that they will encounter more traps and will lose more electrons than images recorded at lower row numbers . on the other hand , in frames where the diffuse sky brightness is high
such as in long exposures through broad - band filters charge traps over the entire detector area can be filled by sky - generated photoelectrons . under these circumstances ,
the net effect of the charge - transfer inefficiency would be a small overall lowering of the diffuse sky brightness perceived in the astronomical scene , while the apparent projection of the stellar images _ above _ the diffuse sky surface brightness could be comparatively unaffected .
an early analysis suggested that in images where the sky brightness is low ( @xmath62030@xmath4e@xmath5/pixel ) a simple `` ramp '' correction proportional to the @xmath0-coordinate position of the star image , amounting to 4% per 800 rows , would approximately correct recorded fluxes for the lost electrons ; in images where the sky was bright ( @xmath7250@xmath4e@xmath5/pixel ) , no correction was needed ; and in cases of moderate sky brightness an intermediate correction was recommended ( holtzman 1995a ) .
while holtzman s ramp model corrects for the first - order effects of charge loss during readout , second - order effects may be significant .
for instance , one would expect any given charge trap to absorb only a limited number of electrons before their mutual repulsion effectively blocked the ingestion of any additional charge .
thus , faint star images consisting of comparatively few photoelectrons should lose a larger fraction of their total charge than images of bright stars containing many photoelectrons .
one would therefore expect faint stars to require a larger fractional correction than bright stars near the same row number and in the presence of the same sky brightness . on the other hand ,
the large electron swarm corresponding to the image of a bright star would occupy a somewhat larger volume within the three - dimensional body of the detector , and might therefore actually enounter more lattice imperfections than would be accessible to smaller electron clouds .
consequently , while we might still expect that brighter stars will require smaller magnitude corrections than fainter ones , it may be that the correction will approach zero more slowly than the expected simple exponential dependence as arbitrarily bright stars are considered .
in addition , while it has been demonstrated that a given star s electron swarm can lose some charge while being clocked through some number of rows along a ccd s parallel register , it is also possible that said packet may lose yet more charge while being transferred through some number of columns along the serial register .
again , this second ramp correction , presumably proportional to the star s @xmath1-coordinate position in the digital image , may also depend upon the sky brightness in the scene and on the brightness of the star image itself .
the numerical coefficients of the @xmath0 and @xmath1 ramp corrections may be different , due to the dissimilar morphologies of the parallel and serial registers within the semiconductor material .
whitmore & heyer ( 1997 ; wh97 ) have estimated the above first- and second - order effects of the charge - transfer inefficiency in wfpc2 , based upon a special sequence of observations of the globular cluster tauri (= ) .
observations were obtained at four pointings of the telescope , with deliberate translations imposed so that a particular piece of the star field could be recorded on the four different chips during the course of the sequence . since the natural coordinate systems of the four detectors are rotated by 90 with respect to each other , this allowed several hundred star images to be used for a purely differential determination of the fraction of charge lost as a function of position on the detector , without a need to know the absolute brightness of any given object .
wh97 presented their results in four alternative formulations , of which the most relevant to the present discussion is given by their eqs . 1 , 2d , and 3d : @xmath8 \hbox{\it cts}_{\hbox{\scriptsize observed}}\ ] ] @xmath9 @xmath10 here ,
_ cts _ represents total wfpc2 data numbers belonging to the stellar image within an aperture 05 in radius , as perceived with the gain = 7@xmath4e@xmath5/dn electronics ; @xmath1 and @xmath0 represent the position of the stellar image in the natural coordinate system of the detector ( pixel ( 1,1 ) being the first one to be read through the amplifier and ( 800,800 ) being the last , with the @xmath1 coordinate varying more rapidly ) ; and @xmath11 is the perceived diffuse sky brightness in the scene , exclusive of recognized stellar images .
( i have opportunistically adopted this version from the four variants presented , because following holtzman ( 1995b ) in reducing wfpc2 data i routinely correct measured magnitudes to their half - arcsecond - aperture equivalents , and because the data files produced by my software contain an estimate of the local sky brightness around and underlying each detected object . ) certain properties of these formulae may be noted .
( 1 ) the correction that is proportional to the @xmath0 position of the star depends only on the diffuse sky brightness in the image , and not on the brightness of the star itself .
in contrast , the alternative @xmath0 ramp corrections presented for magnitudes as measured through 02 synthetic apertures depend upon both sky and star brightness .
this change in formulation stems from the finding by wh97 that the difference between magnitudes measured in 02 and 05 apertures itself depends upon the stellar brightness : `` we find the cte loss is slightly reduced wthen the larger aperture is used , by about 1% for the bright stars and 3% for the faint stars . ''
apparently , in the data studied by wh97 this effect tended to erase the magnitude dependence of the ramp corrections in the larger aperture .
however , it should be noted that a simple differencing of the formulae provided by wh97 indicates that perfect cancellation can occur only at a particular value of the sky brightness , and will not be the case in general .
( 2 ) the correction that is proportional to the @xmath1 coordinate depends only on the brightness of the star and not at all on the diffuse brightness of the sky , a result not necessarily to be predicted from the electron - trap model of the charge - transfer inefficiency .
( 3 ) the @xmath0-coordinate ramp is proportional to an exponential function of the logarithm of the sky brightness , so ( a ) the correction tends asymptotically ( but slowly ) to zero as the sky brightness increases without limit , but ( b ) as the sky brightness decreases to zero , the correction @xmath12 this implies a non - physical result for a star of any apparent magnitude detected in the presence of a very low diffuse sky brightness .
( 4 ) the @xmath1-coordinate ramp is a linear function of the logarithm of the star counts , so ( a ) the correction again diverges as the number of star counts approaches zero .
however , in this case the divergence is not a serious drawback , since photometry of any stellar image containing that few photoelectrons is unlikely to be of genuine astronomical interest .
conversely , ( b ) the correction is identically zero when @xmath13 , and changes sign ( , the lattice imperfections in the serial register appear to act as electron sources rather than electron sinks ) for stellar images containing more than 21,000 dn within a half - arcsecond aperture .
this is approximately the amount of flux contained within a star image whose central pixel just equals the digital saturation level of the analog - to - digital converters in wfpc2 when the gain@xmath14e@xmath5/dn electronics are used .
therefore , this extrapolation , too , is probably of minimal practical relevance except possibly for very bright stars observed with the high - gain electronics bay .
these extrapolations do emphasize , however , that the wh97 corrections are purely empirical constructs , and are not tied directly to a self - consistent physical model of the nature of the charge - transfer inefficiency .
( 5 ) the model correction makes no allowance for the possibility of a charge loss that is independent of position , regardless of whatever dependence it may have on either stellar brightness or sky flux .
this feature of the model is a direct consequence of the experimental design , where apparent flux ratios were observed for individual stars recorded at various positions on the chips , without any consideration of ( a ) flux ratios for different stars with externally known magnitude differences , or ( b ) flux ratios for images of individual stars as recorded in frames with different exposure times .
charge losses that are independent of position may be expected within the lattice - flaw model : simply by virtue of being formed in the body of the silicon wafer in the first place , the stellar image may lose some charge to traps located near the place of its formation .
in addition , any inefficiency inherent in the transfer of charge packets from the top of a parallel register to the serial register , or from the end of the serial register into the on - chip amplifier , would be perceived as a charge loss that is independent of the original location of the stellar image on the detector . a hint that position - independent charge losses _ might _ be significant came from preliminary analyses of wfpc2 images of the globular clusters and palomar 4 ( see w. e. harris 1997 ; stetson 1998 , 1999 ) .
a comparison of short - exposure ( @xmath15 s ) images to ground - based photometry of the same objects implied photometric zero - points consistent with those found by holtzman ( 1995b ) from similarly short exposures of the standard field .
in contrast , comparison of longer ( @xmath16 s ) exposures of the same fields to the same ground - based photometry implied zero points that were different by approximately 0.05 mag , in the sense that the effective quantum efficiency of the detectors was higher in the longer exposures .
a subsequent recalibration of the ground - based photometry , including the addition of many more independent ground - based and hst observations of , suggested that the original estimate of 0.05 mag might in fact have been high ; the additional data suggested that an anomaly of order 0.03 mag on average might be more nearly correct . with the publication of the wh97 corrections ,
it is apparent that the perceived short - minus - long zero - point difference might just be a reflection of the dependence of the fractional charge loss on the total number of recorded stellar photoelectrons and on the perceived surface flux of the sky .
however , it remains possible that some residual position - independent charge loss could occur .
in this paper i reinvestigate the photometric consequences of the charge - transfer inefficiency in wfpc2 , based on extensive series of observations of the galactic globular clusters tauri and .
in addition to testing the validity of the corrections proposed by wh97 on a much larger data set , i investigate certain aspects of the problem not considered by them .
the data analyzed in this paper are all in the public domain , and were acquired from the _ hst _ science archive of the canadian astronomy data centre .
the observations include 318 individual wfpc2 exposures lying within 2 arcminutes of the nominal center of the standard calibration field in the nearby galactic globular tauri ( equinox 2000.0 coordinates 1325370 , 473538 according to h. c. harris 1993 ) . to these
are added 91 exposures mostly centered within a few dozen arcseconds of 738068 , + 385155 : about 1 south of , but including , the center of the far - halo globular cluster .
these latter data include all of the observations discussed by w. e. harris ( 1997 ) , plus some additional exposures of the cluster that were obtained by stsci staff specifically to assist in the calibration of wfpc2 .
some differences between the present data set and that of whitmore & heyer ( 1997 ) deserve particular mention .
( 1 ) the wh97 analysis relied on eight observations made in each of the standard broad - band filters ( f336w , f439w , f555w , f675w , and f814w ) , as well as four observations in f606w intended to allow a comparison with a skyflat developed during the medium deep survey project . from each of the 44 exposures ,
data from only one of the four wfpc2 chips were considered in their analysis .
the present study concentrates on observations made in the f555w and f814w filters (: 152 and 159 exposures , respectively ; : 27 and 58 exposures ) , because these are particularly relevant for the _ key project on the extragalactic distance scale _ and for studies of certain local group star clusters and galaxies in which i am especially interested .
data from all four chips of each of the 402 @xmath17 and @xmath18 exposures were considered in the analysis below .
seven observations of in the intermediate - band f547 m filter were also included in the reductions , to allow the eventual estimation of the fundamental photometric zero - points for this filter for a particular project , but these data were not utilized in the analysis of the charge - transfer inefficiency to be presented below .
( 2 ) the wh97 data were all taken on 1996 june 29 , whereas the data employed here were taken from 1994 january 11 through 1997 june 26 .
the data were all obtained during either 1994 may 2122 ( the 28 w. e. harris observations ) , 1996 december 21 ( 6 exposures ) , or 1997 november 1819 ( 57 exposures ) .
since the observation date determines the spacecraft roll angle required to direct the solar panels toward the sun while the telescope is pointed at a particular spot on the celestial sphere , the wh97 observations were all obtained at the same telescope orientation .
the observations discussed here , having been obtained at essentially all times of the year , represent several full cycles of spacecraft roll angle .
the observations were also taken at several different orientations .
( 3 ) the wh97 data were all taken with wfpc2 operating at a camera temperature of 88 c , whereas 116 of the 318 observations considered here were made at the warmer temperature of 76 c used early in wfpc2 s mission ; the remaining 202 observations and all 91 exposures were taken at 88 c. ( 4 ) the wh97 data were obtained with electronics bay 4 , which gives a nominal gain of 7@xmath4e@xmath5/dn , whereas 264 of the observations analysed here were obtained through electronics bay 3 ( gain approximately 14@xmath4e@xmath5/dn ) . the rest of the and all the observations were taken at 7@xmath4e@xmath5/dn .
( 5 ) the wh97 analysis was purely differential on a star - by - star basis , considering only the flux _ ratio _ between two observations of a given star as a function of the positional difference between the two stellar images in the natural coordinate systems of the detectors .
the new analysis will assume that the true standard - system @xmath17 ( johnson ) and @xmath18 ( kron - cousins ) magnitudes are known _ a priori _ from ground - based observations of the two target fields .
the transformation equations relating the natural - system magnitudes to the groundbased standard system will be determined by a robust least - squares technique , assuming that the color transformation coefficients of holtzman ( 1995b ; their table 7 ) are correct .
the statistical solution of these equations will yield the fundamental zero points of the four wfpc2 ccds , and the residuals of individual stars from these solutions will serve as diagnostics of the charge - transfer inefficiency .
( 6 ) wh97 estimated the mean clear - sky surface brightness for a given frame from a star - free region near the center of the field ; this value was then considered appropriate for every star in the image .
the software used in the present analysis estimates a unique local sky - brightness value for every star , from the mode of brightness values in a surrounding annulus if it is aperture photometry that is being derived , or from the median brightness value at and around the position of the star in a residual image from which all known stars have been subtracted in the profile - fitting analysis .
because these individual , local brightness measurements are conveniently available , they will be used in the following analyses rather than a global average sky for the frame as a whole .
for the globular - cluster images used here , there is little net difference between the two methods ( typically @xmath19e@xmath5/pixel ) . both the present study and that of wh97 include a few observations obtained with internal preflashes , whose purpose was to increase the leverage for determining the dependence of the charge loss on diffuse sky brightness : wh97 had three exposures in the f555w filter with some preflash , while the present sample includes those three plus another nine preflashed exposures of . in each case , the preflash produced a diffuse surface brightness averaging approximately 175@xmath4e@xmath5/pixel . in the long exposures of the sky brightness was typically of order 250@xmath4e@xmath5/pixel , and in all other cases the background level was quite close to zero .
all wfpc2 images studied here were recalibrated by the canadian astronomy data centre , using the best bias and flat - field frames available at the time the data were requested .
the intensities in the frames were corrected for geometric distortion , known defective pixels were masked , and the resulting images were multiplied by 4.0 and coverted to short integers as described by stetson ( 1998 ) .
instrumental magnitudes were determined via the profile - fitting technique with the computer program allframe ( stetson 1994 ) , and selected bright , isolated stars were used to compute magnitude corrections to a system defined by half - arcsecond synthetic apertures via a growth - curve analysis ( stetson 1990 ) .
all transformations from the wfpc2 instrumental system to the standard ground - based @xmath20 system were performed with a version of the program ccdstd ( see stetson 1993 ) .
because the edges of the pyramid mirror in wfpc2 are imaged onto the photosensitive areas of the four camera ccds , this analysis will be confined to that area of each chip which is almost certainly unvignetted ; this has been taken to be @xmath21 in the three wfc chips , and @xmath22 in pc . the ground - based johnson @xmath17 and kron - cousins @xmath18 magnitudes which will be used for calibrating wfpc2 are taken from walker ( 1994 ) for , and from my own analysis of ground - based data for .
the photometry for the latter cluster is based upon 68 @xmath17 and 62 @xmath18 frames obtained during the course of 14 nights in six observing runs .
the images were taken with the kitt peak 4 m and 2.1 m telescopes , the canada - france - hawaii telescope ( data obtained through the courtesy of the canadian astronomy data centre ) , and the isaac newton telescope ( data obtained through the courtesy of the isaac newton group archive ) , and have been rigorously transformed to the photometric system of landolt ( 1992 ) as part of an ongoing program that has now homogenized photometric data from over 40 individual observing runs . is a massive , rich globular cluster located about 100 kpc away , and the field is quite crowded under groundbased seeing conditions . therefore the hst imagery was examined to identify stars that appear to be minimally crowded . wherever possible ,
a star identified in the groundbased images was cross - identified with a star detected in the wfpc2 data .
then the groundbased photometry was differentially corrected for those neighbors that were found in the hst images but were _ not _ identified and accounted for in the analysis of the groundbased data .
any star that required a blending correction in excess of 0.1 mag was rejected as a potential standard .
as the analysis proceeded , any star in either or for which the mean of the calibrated hst photometry differed from the mean of the groundbased data by more than 0.3 mag in either filter was likewise discarded , and the solutions were redone until no further rejections were required .
stars that were rejected at this stage were walker numbers 1 , 9 , 24 , 48 , 49 , 57 , 60 , 77 , 99 , 129 , 139 , 152 , 154 , 174 , 196 , 203 , 238 , 240 , 243 , 247 , 248 , 261 , 274 , and 275 ; this left 184 stars for which the groundbased and hst photometry were concordant to within @xmath23 mag .
fifteen of 366 stars in were similarly discarded .
it may be argued that the inclusion of stars with such large standard errors would vitiate any delicate photometric analysis .
however , this should not be a serious problem . in the analyses that follow all observations
are weighted in accordance with their inferred standard errors , which are fairly well known from the photon statistics and repeatability from multiple images .
then a robust version of least squares is used ( see stetson 1989 ) which permits the solution to follow the main trend of the data without being unduly influenced by outliers .
inclusion of faint stars is necessary to provide the needed leverage on the magnitude dependence of the charge loss , and a more strict censoring of the data might introduce bias rather than reduce it .
in addition , most stars were eventually imaged at many different locations on the various chips .
a star whose photometric errors give a semblance of a large charge loss when imaged at a high value of @xmath1 or @xmath0 would produce a cancelling effect on other occasions when it appeared at low column or row numbers . as in the differential wh97 study , then , gradients in the photometric zero point across the face of the chips , and variations of those gradients with sky brightness and stellar magnitude will not be sensitive to the accuracy of the groundbased magnitudes .
prior knowledge of the standard photometric magnitudes will be useful only in testing for the presence of position - independent photometric non - linearities , and those tests alone will be affected by uncertainty in the groundbased results .
the present data set will be analysed in three ways : ( 1 ) assuming that charge - transfer losses are negligible ; ( 2 ) correcting the observed stellar fluxes according to the precepts of wh97 ; and ( 3 ) solving for a new numerical model of the charge loss as a function of position , stellar brightness , and sky brightness simultaneously with the determination of the photometric zero points .
the results of these three analyses will be presented in
3.1 , .2 , and .3 , respectively , for the data obtained at a camera temperature of 88@xmath24c . a briefer discussion of the data obtained at the warmer temperature will be presented in 4 .
the fundamental assumptions of the present analysis are : 1 .
the charge - transfer inefficiency affects the gain = 7 and the gain = 14 observations identically , provided care is taken to evaluate the charge loss in units of electrons .
the inefficiency is the same for f555w and f814w .
the size of the effect has not evolved significantly between 1994 may and 1997 november .
the charge loss may be significantly different at a camera temperature of 76 c as opposed to 88 c. wh97 did not require assumptions ( 1 ) , ( 3 ) , or ( 4 ) because all their observations were obtained on a single date with a single gain setting and camera temperature .
our assumption ( 1 ) follows from the hypothesis that the charge transfer inefficiency is completely the result of physical processes within the body of the ccd , and not from subsequent effects that take place during amplification and digitization .
assumptions ( 2 ) and ( 3 ) are made primarily for simplicity and ease of modifying existing software to perform the present analysis , but the work of wh97 is relevant for assessing the validity of assumption ( 2 ) , at least . while they found that the gross correction for lost charge does appear to depend upon the filter the corrections being the greatest at short and long wavelengths and being least for the f555w and f675w filters they noted , `` this is probably because the chips are more efficient at these wavelengths [ , near 555 and 675@xmath4 nm ] , hence the background is higher .
the higher background appears to reduce the cte loss ''
thus , when proper quantitative account is taken of the apparent dependence of charge loss on background brightness level , wh97 suggest that the photometric correction does not depend heavily on bandpass . in any case
, the validity of assumptions ( 2 ) and ( 3 ) can be checked _
ex post facto _ once the average baseline effect of charge - transfer inefficiency has been calibrated and removed . as a result of assumption ( 4 ) ,
i have performed separate and independent analyses for the colder and warmer camera temperatures .
a least - squares fit of the basic transformation equations to the images of 36,234 stars recorded in the cold - camera observations of and yielded the photometric zero points listed in table 1 .
as mentioned above , the color coefficients derived by holtzman ( 1995b ) were imposed as givens .
specifically , @xmath25 the first part of the table gives the zero point for each chip / filter combination based upon the full set of observations obtained with electronics bay 3 ( exclusively short exposures of the field ) , and the full set with electronics bay 4 ( a combination of short exposures of and both short and long exposures of ) . in the second part of the table , electronics bay
4 zero points are presented for the short exposures , the short exposures , and the long exposures considered separately .
all of these derived zero points are tabulated in detail because they will be used below to test the validity of the various correction schemes .
the calibration residuals @xmath26(magnitude ) resulting from the 16 combined solutions ( 2 gain settings @xmath27 2 filters @xmath27 4 chips , corresponding to the zero points listed in table 1a ) were all merged into a single file ; the sense of the residual is ( observed instrumental magnitude ) _ minus _ ( magnitude predicted from the standard - system indices and the mean transformation equations ) .
figure 1 shows the individual residuals plotted versus ( top ) @xmath1 coordinate on the chip , ( middle ) @xmath0 coordinate on the chip , and ( bottom ) raw instrumental magnitude @xmath28 (= _ constant _
@xmath29[number of stellar photoelectrons within a radius of 05 ] ) ; for clarity , only unsaturated stars where the combined uncertainty of the ground - based standard magnitude and the wfpc2 instrumental magnitude is less than 0.05 mag have been plotted .
it is apparent that there are significant residual trends with both the @xmath0 and @xmath1 coordinates , but no systematic nonlinearity with magnitude is obvious . to provide a more quantitative , albeit still crude , measure of the magnitude of these effects , a robust , weighted least - squares surface of the form
@xmath30 was fitted to all 36,234 fitting residuals .
the natural - system position variables are @xmath1 and @xmath0 , @xmath28 is the instrumental magnitude as defined above , and @xmath31 is the date of observation , measured in years from heliocentric julian date 2,450,000.0 . to place the amplitude of the position and magnitude trends in a comfortable set of units ,
i take the maximum displacement of any given well - measured star from the mean of all stars to be @xmath32 px , @xmath33 px , @xmath34 mag . from the least - squares fit i estimate the maximum photometric consequences of the neglected charge - loss corrections to be @xmath35
the positive signs of the first three coefficients , representing the linear dependence on position and magnitude , indicate that the instrumental magnitude is too large , , the star is measured too faint with respect to the average zero - point for all stars taken together , when the @xmath1 coordinate is large , the @xmath0 coordinate is large , or the star is faint ( large @xmath28 ) all in accordance with the simple predictions of the charge - loss model .
the positive signs of both second derivatives demonstrate that the effect of charge loss is further compounded for faint stars at large @xmath1 and @xmath0 positions . for a star in a typical exposure ,
any one of these neglected corrections can cause a systematic error of no more than about 0.015 mag , which corresponds to a root - mean - square error of about 0.01 mag .
a star in the wrong corner of position - magnitude space could , however , be subject to an error more than four times larger . note in particular that the result for @xmath36 implies a stretching of the instrumental magnitude scale , on average : faint stars are measured too faint compared to bright ones , by about 0.03 mag per 5 magnitudes ( an approximate figure for the effective dynamic range of wfpc2 ) .
the remaining coefficient determined from the least - squares fit to the residuals is @xmath37 the positive sign of the result indicates that for fixed true - system magnitudes , the observed instrumental magnitudes are becoming larger , , fainter , as time passes .
however , with an indicated rate @xmath38 mag yr@xmath39 , this secular change is not very important even over the more than 3.5 years over which the cold - camera data were obtained . in applying the wh97 corrections to the present data sets
, all star and sky fluxes were converted to units of electrons , assuming gain values of identically 7.0 or 14.0@xmath4e@xmath5/dn , to facilitate combining data obtained with either of the two electronics bays .
since i will be calibrating small , differential magnitude corrections directly from stellar observations , minor differences between the postulated and actual gain factors will not seriously damage the analysis .
the numerical coefficients of the wh97 equations were adjusted to represent fluxes expressed in units of electrons rather than electronics bay 4 data numbers . as noted above
, the logarithm of the sky brightness varies rapidly as the sky brightness approaches zero , and is undefined for any non - positive estimate of the sky flux .
in fact , table 1 of wh97 indicates that seven of their 44 digital images had inferred median sky - brightness values , @xmath40 , that were negative ; a further 18 images had positive median brightness values that were less than 0.5 dn px@xmath39 , where a small inaccuracy in @xmath11 begins to lead to a significant change in @xmath41 .
wh97 make no explicit mention of how images like these are to be treated in their methodology .
the electron - trap model of the charge - transfer inefficiency predicts that a realistic stellar image would lose only a finite number of electrons even in the limit of a hypothetical exposure with a zero effective background ( a short exposure of bright stars through a narrow filter , for instance ) .
therefore , it is apparent that some modification must be made to the formulae so that the predicted photometric correction approaches a finite constant value as the estimated background brightness reaches ( and , in the presence of noise , possibly passes through ) zero . in the present analysis
i do this by simply `` softening '' the inferred sky brightness at low values : @xmath42 ^ 2\right\}^{1\over2}.\end{aligned}\ ] ] ( please note that in the present analysis , the units of _ bkg _ are electrons , not dn as in the wh97 study . )
the softening constant , 1@xmath4e@xmath5 , has been adopted arbitrarily , and has as its sole justification the fact that it is better than doing nothing .
fig . 9 of wh97 indicates that the fraction of charge lost from a stellar image continues to increase down to @xmath43 \sim -1,$ ] or @xmath44e@xmath5/pixel , but diffuse sky fluxes below this have not been sampled .
the imposition of this minimum inferred sky brightness means that all frames where the diffuse sky brightness is small compared to 1@xmath4e@xmath5/pixel will be subject to similar , finite , photometric corrections .
henceforth in this paper , the unqualified symbol _ bkg _ will refer to @xmath45 as defined in eq . 1 .
having made this one modification to the wh97 photometric corrections , i then recomputed the cold - camera zero points for the various combinations of gain factor , chip , and filter , arriving at the results shown in table 2 .
i forgo plotting the residuals from these solutions , but the numerical correlations with the various independent variables are : @xmath46 thus , the wh97 corrections operate in the right direction to reduce the size of the maximum systematic error induced by the electron traps , but they do not eliminate the errors entirely , at least for the present data set : some of the charge - loss effects are slightly over - corrected , others are significantly under - corrected .
most notably , in these data the amount of charge lost in the serial register ( the @xmath1 direction ) appears to be larger than estimated by wh97 , as is the dependence of the @xmath0-coordinate ramp on instrumental magnitude ( which has been set to zero in their model ) . as a result ,
a maximum systematic error in excess of 0.02 mag is still possible for stars in an unfortunate location in @xmath47 space in an average exposure . for internal consistency , unlike wh97
i adopt symmetric , exponential - of - a - logarithm , formulations for the charge - transfer losses in both @xmath1 and @xmath0 as functions of both star and sky flux . in doing this
, i claim no superior knowledge of the detailed physics of charge loss , but rather follow wh97 s lead in postulating a simple , phenomenological , numerical model for the required photometric corrections . at the same time , in reconsidering the wh97 equations , i have tried to improve on some of the aforementioned features of their formulation .
in particular , i have added a term to the formula which allows for the possibility of a charge loss which is independent of the image s position on the chip . in doing this ,
i have arbitrarily assumed that any charge loss which _ is _ independent of position will have precisely the same dependence on sky brightness and stellar flux as the @xmath0-dependent ramp correction .
again , this assumption is arbitrary and has been made to reduce the number of independent coefficients that must be determined ; it is physically reasonable provided that the electron traps contributing to any position - independent charge loss are physically similar to those that produce the well - documented @xmath0 photometric ramp in the body of the detector , which may be different from the traps that produce the less well - understood @xmath1-dependent charge losses that probably occur in the serial register . of course
, if it should happen that there are no position - independent effects , the least - squares analysis will allow the coefficient of the constant term to assume a negligibly small value .
finally , rather than expressing the equations in terms of the image positions and the star and sky brightness in absolute terms , i consider expansions about `` typical '' values of these variables .
this makes the various terms in the formulae more nearly independent of each other , so that the least - squares solution is less ill - determined , and also leads to derived standard errors of the coefficients that can be understood in a simple and direct way . with _ bkg
_ representing the softened sky flux in e@xmath5/pixel , _ cts _ representing the stellar flux within a 05 ( radius ) aperture in e@xmath5 , and @xmath1 and @xmath0 the position of the star image in the natural coordinate system of the chip , the adopted numerical model for the photometric corrections is : @xmath48 - 1 ; \\
\hbox{\it star } & = & \log_{10}{[\hbox{\it cts } } ] - 4 ; \\
y\hbox{-cte } & = & c_1 + c_2\times\hbox{\it sky } + c_3\times\hbox{\it star } ; \\
x\hbox{-cte } & = & c_4 + c_5\times\hbox{\it sky } + c_6\times\hbox{\it star } ; \\ c & = & 1 + 0.01\times\left(c_7 + \delta y\right)\times e^{\hbox{\small $ y$-cte } } + 0.01\times \delta x\times e^{\hbox{\small $ x$-cte}}.\end{aligned}\ ] ] if @xmath28 represents the measured stellar magnitude relative to whatever zero point , then @xmath49 this model is not as formidable as it looks .
@xmath50 is equivalent to @xmath51 a point near the center of the unvignetted area on each of the wfpc2 ccds while @xmath52 , @xmath53 refer to points near the edges of the photometrically reliable field .
the peculiar combination of natural and base-10 logarithms and exponents used here is not entirely whimsical ; for me , at least , it facilitates performing order - of - magnitude estimates in the head .
for instance , the constant @xmath54 approximately represents the fraction of charge lost ( expressed in percent ) from a star image containing 10,000 electrons near the effective center of an image that has a sky surface brightness of 10@xmath4e@xmath5/pixel .
the constants @xmath55 and @xmath56 represent the degree to which such a star would lose more or less charge if measured near the edge of the field rather than near the center as a result of , respectively , the @xmath0 and @xmath1 ramps . for
star images containing either 100,000 or 1,000 electrons , these corrections would differ by @xmath57 and @xmath58 percent , respectively .
likewise , for sky values of 100 or @xmath59e@xmath5/pixel ( `` @xmath6 '' because of the softening ) , the corrections for a 10,000-electron star image would differ by @xmath60 and @xmath61 percent . encoding the corrections as proportional to @xmath62 $ ] and @xmath63 $ ] ensures that they will go asymptotically although not necessarily quickly to zero as @xmath64 or @xmath65 , as long as @xmath66 , @xmath67 , @xmath68 , and @xmath69 are all less than or equal to zero . at the other extreme
, the softening of the sky brightness ensures that the correction approaches a constant value for any given star in the limit as the background flux becomes negligible . no corresponding numerical protection in the limit of negligibly small stellar flux
is required , since the apparent magnitude of a statistically insignificant star is usually of little astronomical interest .
a copy of the computer program ccdstd ( stetson 1993 ) was modified to yield numerical values and uncertainties for the seven model constants @xmath70 , as well as the zero points of the transformation from instrumental to standard magnitudes . as before , the 16 separate zero points @xmath71 and @xmath72 were computed for the four chips and the two gain settings ( eight values for @xmath71 and eight for @xmath72 ) , based on the cold ( camera temperature = 88@xmath4c ) data only .
the same gain = 7@xmath4e@xmath5/dn zero points were assumed to apply for the data and both short and long exposures of . at the same time , the cold data for all four chips , both filters , both gain settings , and both clusters were used to determine a single , unique set of values @xmath70 .
the numerical results of this exercise were the zero points listed in table 3a and @xmath73 all of these terms @xmath74 appear to be highly significant , with the sole exception of the sky dependence of the @xmath1-coordinate ramp ( @xmath68 ) : for a ten - fold change in the sky brightness ( , from 10@xmath4e@xmath5/pixel to either 1 or 100@xmath4e@xmath5/pixel ) the required correction changes by only [email protected]% at the right and left edges of the field , which abundantly justifies the neglect of this term by wh97 .
however , a dependence of the @xmath0 ramp on stellar flux ( @xmath67 ) , which wh97 also omitted , appears to be strongly supported by these data . according to this model ,
a star image consisting of 10,000@xmath4e@xmath5 near the center of an image with a sky brightness of 10@xmath4e@xmath5/pixel loses approximately @xmath76% of its electrons ; at @xmath77 and @xmath78 it loses @xmath79% and @xmath80% yielding , in this case , a ramp of about 3.8% per 750 pixels or 4.0% per 800 pixels . in the hypothetical row @xmath81
the same star would lose @xmath82e^{+0.629 } = 0.1$]% , so a position - independent contribution to the total correction , while apparently statistically significant , is clearly not very important .
decreasing the sky brightness by a factor of ten would increase the @xmath0 ramp by a factor @xmath83 , to about 7.0% per 800 pixels . for a star
2.5 magnitudes brighter , with 100,000@xmath4e@xmath5 in its image , the @xmath0 ramp would be smaller by a factor @xmath84 , yielding 2.2% per 800 pixels for 10@xmath4e@xmath5/pixel in the sky , or 3.9% per 800 pixels for a sky brightness of 1@xmath4e@xmath5/pixel .
the numerical correlations between the fitting residuals and position and magnitude are : @xmath85 these are not identically equal to zero because , ( a ) these linear correlations do not have the same functional forms as the terms defined in the least - squares photometric solution , and ( b ) in the two separate robust least - squares fits one for the model parameters plus independent zero points for 16 distinct data sets , and one for the correlations between residuals and positions and magnitudes for all 16 data sets considered together the weighting schemes can not be perfectly reproduced : the degrees of freedom are different . in particular , in fitting the correction model
no allowance was made for a separate trend of photometric error with magnitude independent of position on the chip , so if such a trend is present it can not be perfectly compensated by the model . as a result , the single most significant residual correlation is @xmath86 .
the sign of this correlation has changed from the uncorrected solutions , indicating that trend of increasing fractional charge loss with decreasing stellar flux has been overcorrected , on a global average over the face of the four chips .
the corrected magnitude scale is now apparently compressed by 0.013 mag per 5 magnitudes , as compared to the groundbased results .
i believe that this level of systematic error is small compared to other unavoidable biases endemic to photometric measurements of very faint stars ( see , , stetson & harris 1988 ,
v(b ) ) , and it is unlikely that a much more satisfactory calibration can be derived from the present data .
certainly , in view of the crowding conditions in these fields , i am unable to guarantee that the ground - based photometry is itself linear to this degree of accuracy .
additional well - established , deep , groundbased photometric sequences in other star fields targeted by hst , such as the palomar - type globular clusters and fields in the magellanic clouds , would be most welcome additions to the study of this problem . still , after application of this set of corrections to the wfpc2 data , the maximum systematic errors associated with the purely positional variables @xmath0 and @xmath1 are now well under 1% , at least averaged over the present data set .
the secular trend of the photometric residuals with time that remains after applying these corrections is @xmath87 this is larger than was estimated in 3.1 presumably because in the uncorrected data the mean trend with time was partially masked by scatter resulting from the neglect of the charge - loss effects but over the 3.5 years of available wfpc2 data this still implies a maximum departure from the mean zero points of order 0.002 mag at the two ends of the time interval covered .
there are several additional sanity checks which can be derived from these data .
for instance , if the corrections truly are valid , then the photometric zero point for a given chip / filter combination should be independent of exposure time .
we can perform this test with the short and long exposures of ; i omit the exposures from consideration here because this eliminates confusion produced by any possible differences between the ground - based photometric systems for the two clusters .
the four chips and two filters provide us with eight independent measures of the ( short long ) zero - point difference ( data listed in tables 1b , 2b , 3b ) . taking the unweighted mean of the eight ( short long ) zero - point differences ,
i find net values of @xmath88 ( standard error of the mean ; 0.0245 standard deviation of one difference ) based on the reductions with no photometric corrections ; @xmath89 ( s.e.m . ;
0.0243 s.d . ) with the wh97 corrections ; and @xmath90 ( s.e.m . ; 0.0224 s.d . ) with the new corrections derived in 2.3 . in other words , the wh97 corrections
largely eliminate the perceived difference in zero points between short and long exposures , but the chip - to - chip and filter - to - filter scatter in the differences is not much reduced by the wh97 corrections .
the newly derived corrections represent a slight improvement on those of wh97 , in that they reduce the residual ( short long ) zero - point difference a bit further , and also slightly reduce the chip - to - chip and filter - to - filter scatter . repeating the experiment with weighted means ( after all , the pc1 zero points are based on roughly one - fourth as many stars as those for the wide - field chips ) we obtain a mean ( short long ) difference of @xmath91 for the uncorrected data ; @xmath92 with the wh97 corrections ; and @xmath93 with the new corrections .
the gain ratio between electronics bay 3 and electronics bay 4 may differ from one ccd to the next , but we may expect that the gain will not depend upon the filter with which a given exposure was made . if this is the case , then the difference between the gain = 14 zero point and the gain = 7 zero point should be the same in @xmath17 and @xmath18 .
for instance , with no photometric corrections , for chip 1 the @xmath17 zero point from the gain = 14 observations of is 1.7673 ( table 1a ) , and the gain = 7 zero point is 0.9974 ( table 1b ; again , this test considers the gain = 7 observations for alone to eliminate the effects of any residual differences between the and ground - based photometry ) , for a gain ratio of 0.7699 mag .
the @xmath18-band gain ratio for the same chip is @xmath94 mag , and the difference between the @xmath17 and @xmath18 gain ratios is 0.0437 mag , rather than the expected zero .
each separate calibration produces four values of this gain - ratio difference , from the four chips .
the unweighted root - mean - square value of these four differences is : 0.0245 mag for the uncorrected reductions , 0.0258 mag with the wh97 corrections , and 0.0285 mag with the new corrections . for each filter ,
the zero points of the gain = 14 observations should be the same in the four chips , because the flat - field observations were made with the gain = 14 electronics and the flat fields were normalized over the four chips taken together so that surface brightnesses measured in the different chips would be restored to a common system .
the standard deviations of the individual chips about the unweighted mean zero points for the two filters ( eight observations , six degrees of freedom ) are 0.0114 mag for the uncorrected reductions , 0.0098 mag with the wh97 corrections , and 0.0142 mag with the new corrections .
the tests described in this section are quite indirect , and rely on some untested assumptions for their validity : that gain ratio is independent of wavelength , for instance .
therefore these results can be no more than suggestive .
there is a hint that the wh97 corrections do better than the new ones when applied to the data alone , as in the gain - ratio test and the test assuming the equality of the gain = 14 zero points .
the new corrections seem to do better when the observations are added to the sample , as in the short long test here , and in the global fitting residuals discussed in
3.2 and .3 .
if the self - consistency tests of this section have any meaning at all , it is that the errors in the present data set are not totally random , but rather contain unmodeled systematic components as well .
for instance , the photometric zero points may vary erratically with time , or with the history of previous exposures , or with some other variable not included in the phenomenological model . if this is true , then the standard errors quoted at various places in 3 and in tables 13 which were estimated simply from the root - mean - square scatter divided by the square root of the number of fitting residuals
are merely attractive fictions ; the true errors for any given selected subset of the data may be considerably larger .
it appears that the photometric zero point for any given chip / filter combination in any given data set is really externally reliable only to @xmath95 mag , regardless of the standard errors listed in the tables .
for instance , the dispersion of zero - point _ differences _ in the ( short@xmath4@xmath4long ) test just performed was @xmath96 mag , so the typical uncertainty of the zero - points themselves would by smaller by @xmath97 , or @xmath98 mag .
similarly , the gain - ratio test involved differences of differences of zero points , with a scatter @xmath99 , implying zero points individually reliable to @xmath100 mag , while the third test , involving zero points directly , implied a typical external uncertainty in the range 0.010 0.014 mag .
whitmore ( 1998 ) has conducted an analysis of the time variability of the charge - transfer inefficiency based on seven images of the standard field obtained with wfpc2 operating at the colder temperature and spanning the dates 1994 april to 1997 june .
he concluded that the cte loss has increased over that period of time , but only for the faintest stars : those containing from 20 to 50 dn within a 02 aperture in very short exposures at the 14@xmath4e@xmath5/dn gain setting appeared to lose @xmath101% of their electrons when clocked out through 800 rows in the detector at the end of that period , as compared to @xmath102% at the beginning .
a flux of 280 to 700@xmath4e@xmath5 within 02 corresponds , on my instrumental magnitude scale , to @xmath103 at gain = 7@xmath4e@xmath5/adu and @xmath104 at gain = 14 .
accordingly , i have selected out the photometric residuals , _ after _ the application of the new corrections derived in 3.3 , for stars with predicted instrumental magnitudes fainter than those limits . for these stars considered separately , i find @xmath105 ( systematic variations of the residuals with stellar brightness are not worth investigating in a sample with this limited range of instrumental magnitude . ) for these faint stars , it seems that the new corrections overcompensate a bit for the mean rate of charge loss with @xmath1 and @xmath0 position on the chip the net trends are slightly negative , indicating that the corrected magnitudes are a little too bright at positive values of @xmath106 and @xmath107 but the maximum error is @xmath108% and the r.m.s .
error averaged over the face of the chip is @xmath109mag in both @xmath1 and @xmath0 .
the residuals for these very faint star images do show a systematic positive trend with time , as found by whitmore : the mean residual increases at a rate of 0.0025@xmath4mag yr@xmath39 , for a total net difference of 0.009@xmath4mag over the 3.5 years spanned by the current data set for a faint star located near @xmath110 or , equivalently , for the average of many stars scattered uniformly over the detector field .
both the overcorrection of the @xmath1 and @xmath0 ramps and the apparent secular increase of the charge loss with time are much smaller than the random poisson noise for any star image containing 280700@xmath4e@xmath5 . indeed , the random observational errors for star images this faint are almost always larger than 0.05@xmath4mag when all error contributors are considered .
that is why these stars are essentially absent from fig .
1 where , for reasons of clarity , only the more precise measurements have been plotted .
the unmodeled systematic trends of these faint - star residuals are also much smaller than the systematic statistical effects discussed , , by stetson & harris ( 1988 ) .
still , while these remaining systematic effects are unimportant on a star - by - star basis , the reader should be aware that ensemble average magnitudes for large numbers of very faint stars observed in short exposures may be subject to these additional ( relatively minor ) sources of systematic error .
whitmore further states that for star images containing at least 2800@xmath4e@xmath5 within 02 ( radius ) the rate of increase of charge loss with time is negligible , even in exposures with very low background levels .
it is probably safe to assume that the secular evolution of the charge losses will be doubly negligible in images that contain significant background levels , and that one generic set of charge - loss corrections will be reasonably valid over the 33.5 years spanned by existing studies .
the results presented here certainly seem to bear that conclusion out .
wfpc2 was operated at a camera temperature of 76 c from installation in 1993 december until 1994 april 23 . between 1994 january 11 and april 21 the tauri standard field was observed a number of times in f555w employing both electronics bays , but observations in f814w are only available for the gain = 14 electronics . determining the mean photometric zero points in the usual way , under the assumption of no charge loss , we arrive at the fitting residuals shown in figure 2 .
there are comparatively few observations from the short period of time when the camera was operated at the higher temperature ( 14,521 residuals are shown in fig .
2 ) , and the data do not span the chips nearly as well as in the more extensive cold data .
it is clear that there is a considerable charge - loss effect in the @xmath0 coordinate , but any systematic trends with @xmath1 and magnitude are too small to be seen at this scale .
i have attempted to determine charge - loss coefficients for the warm data as i did for the cold .
i quickly found that the solution for the fitting parameter @xmath111 rapidly diverged to @xmath112 , indicating that any charge loss in the serial register is too small to measure even in a statistical sense . in order to get a self - consistent solution , therefore , it was necessary to simplify the problem .
accordingly , i effectively turned off the ramp correction in @xmath1 by assigning @xmath111 a value of 20 , and i also eliminated the possibility of a position - independent charge loss by setting @xmath113 identically equal to @xmath114 .
the standard robust solution for the remaining three sigificant terms yielded @xmath115 plus the zero points listed in table 4 . in a frame with a mean background of 10@xmath4e@xmath5/pixel , the total amount of charge lost from a 10,000 electron star at @xmath78 ( @xmath116 ) is @xmath117% , as compared to @xmath118% for the cold data .
for the warm data , the variation of the charge loss with background level appears marginally less than for the cold data , but the dependence on the brightness of the star is about twice as strong .
for , a total of 152 wfpc2 exposures in the f555w filter and 159 exposures in f814w spread over 3.5 years of observations have now been calibrated to a common photometric system .
in contrast , walker s ( 1994 ) groundbased study of was based on only 12 @xmath17 and 10 @xmath18 frames obtained over the course of six nights in 1985 , using an early rca ccd . by combining the hst data with walker s it should be possible to get an improved standard sequence in , `` improved '' at least in the sense of being internally more precise even if not externally more accurate .
figure 3(a ) is a ( , @xmath17 ) color - magnitude diagram for walker s stars in , while fig .
3(b ) shows the same stars , but with the present wfpc2 photometry corrected for charge - transfer inefficiency and transformed to the standard @xmath20 system using the precepts of 2.3 and 3 and the zero points of tables 3 and 4 averaged together with walker s .
given the large body of hst data , it is also possible to define new standard - sequence stars in . accordingly , a list of minimally crowded detections in the standard field was derived from the output of the allframe reductions . from among these relatively isolated stars ,
a total of 1,318 were selected which had been measured at least 10 times in each filter , had standard errors of the mean no larger than 0.03@xmath4mag in each of @xmath17 and @xmath18 , and gave no evidence of instrinsic variability in excess of 0.03@xmath4mag , root - mean - square , averaged over the two filters .
these are shown in fig .
( the figure 1,318 includes stars listed by walker ( 1994 ) ; 1,116 of them are new ones without walker , harris , or woolley identification numbers as tabulated by walker . ) figures 4(a ) and ( b ) show , respectively , the ground - based photometry for , and the -based photometry averaged together with the ground - based data for the same stars . since in i had already selected out a large sample of minimimally crowded stars to define the ground - based standard sequence in the first place , there is no need to pick out additional local standards based on the hst data alone .
there is an advantage in re - deriving final sets of zero points based upon these improved , homogenized standard sequences in the two clusters . with most of the stars having been measured on more than one of the four chips during the course of the observations , as well as
having been measured in both long and short exposures ( in the case of ) , with the gain = 7 and gain = 14@xmath4e@xmath5/dn electronics ( in the case of ) , and with the camera operating at both the warmer and colder temperatures ( ) , any residual systematic errors remaining in the individual data subsets will be diluted when all are averaged together .
add to these considerations the fact that the total number of standards has now been increased from 534 to 1,714 while the random errors have been reduced from a median value of 0.017@xmath4mag per star to a median of 0.007@xmath4mag per star , and it follows that the derivation of new zero points based on the improved standard sequence will yield photometry that is more homogeneous across the various data sets , even though the reanalysis can not lead to photometry that is more absolutely correct on the average over all data sets .
the final zero points that result from calibration of the cte - corrected data with the augmented list of standards are given in table 5 .
again , the reader should note that the formal standard errors listed in the table are outrageously optimistic ; true uncertainties of order 1% or slightly more are far more realistic .
only additional observations will improve this situation .
application of the present cte corrections to other wfpc2 data is straightforward .
provided that the reader has stellar positions referred to the natural coordinate system of each chip ( @xmath119 ) and is careful to use the known gain ratio of the observation to convert the instrumental magnitude and some estimate of the diffuse sky brightness in the image to units of electrons , then it is merely necessary to add nine lines of code , corresponding to eqs .
( 1)(9 ) , to whatever software the reader is now using . if , in addition , the reader happens to be following the key project precepts on the standardization of the raw images ( , the images have been multiplied by 4.0 so that they can be stored as short integers without significant loss of precision ; see stetson 1998 ) , then the zero points of table 5 can be used to convert the corrected half - arcsecond aperture magnitudes to the standard system . if the reader does not use the key project precepts , stetson ( 1998 ) list the steps necessary to relate these zero points to those of holtzman ( 1995b ) .
the external reliability of photometry on this system is still probably of order 0.01@xmath4mag , since the present zero points represent a compromise among gain 7 and gain 14 , warm- and cold - camera , and short- and long - exposure data sets , each of which has zero points individually uncertain at the 0.010.02@xmath4mag level .
internally , however , the various chips should yield photometry consistent at a level somewhat better than this .
i am extremely grateful to dr .
jeremy mould and the entire staff of mount stromlo and siding spring observatories for the financial support and hospitality during the time when much of this work was being done .
additional financial support from the jet propulsion laboratory of the california institute of technology is also very much appreciated . rcccc & cold , gain = 14 , @xmath17 & 1.7673 @xmath75 0.0029 & 1.7515 @xmath75 0.0010 & 1.7824 @xmath75 0.0008 & 1.7593 @xmath75 0.0012 cold , gain = 14 , @xmath18 & 2.6667 0.0021 & 2.6534 0.0007 & 2.6695 0.0010 & 2.6513 0.0010 cold , gain = 7 , @xmath17 & 0.9957 0.0023 & 0.9860 0.0016 & 0.9922 0.0015 & 0.9897 0.0016 cold , gain = 7 , @xmath18 & 1.9728 0.0021 & 1.8963 0.0012 & 1.9240 0.0012 & 1.9473 0.0011 & @xmath120 cen , @xmath17 & 0.9974 0.0030 & 1.0041 0.0024 & 1.0056 0.0029 & 0.9965 0.0024 @xmath120 cen , @xmath18 & 1.9405 0.0031 & 1.9029 0.0017 & 1.9035 0.0025 & 1.9134 0.0016 , short , @xmath17 & 0.9979 0.0044 & 0.9709 0.0026 & 1.0050 0.0020 & 0.9899 0.0023 , short , @xmath18 & 1.9981 0.0026 & 1.8929 0.0015 & 1.9367 0.0014 & 1.9594 0.0013 , long , @xmath17 & 0.9938 0.0061 & 0.9874 0.0032 & 0.9591 0.0031 & 0.9743 0.0037 , long , @xmath18 & 1.9468 0.0039 & 1.8826 0.0038 & 1.8900 0.0023 & 1.9224 0.0023 rcccc & cold , gain = 14 , @xmath17 & 1.7264 @xmath75 0.0026 & 1.7254 @xmath75 0.0009 & 1.7517 @xmath75 0.0008 & 1.7326 @xmath75 0.0014 cold , gain = 14 , @xmath18 & 2.6217 0.0021 & 2.6263 0.0007 & 2.6355 0.0009 & 2.6219 0.0011 cold , gain = 7 , @xmath17 & 0.9610 0.0023 & 0.9594 0.0016 & 0.9666 0.0014 & 0.9649 0.0016 cold , gain = 7 , @xmath18 & 1.9345 0.0018 & 1.8640 0.0010 & 1.8942 0.0010 & 1.9195 0.0010 & @xmath120 cen , @xmath17 & 0.9571 0.0032 & 0.9765 0.0023 & 0.9777 0.0028 & 0.9680 0.0026 @xmath120 cen , @xmath18 & 1.8967 0.0033 & 1.8715 0.0014 & 1.8744 0.0026 & 1.8795 0.0019 , short , @xmath17 & 0.9584 0.0038 & 0.9379 0.0023 & 0.9744 0.0018 & 0.9617 0.0023 , short , @xmath18 & 1.9539 0.0021 & 1.8554 0.0012 & 1.9024 0.0012 & 1.9285 0.0011 , long , @xmath17 & 0.9734 0.0058 & 0.9712 0.0032 & 0.9433 0.0031 & 0.9672 0.0037 , long , @xmath18 & 1.9267 0.0036 & 1.8668 0.0038 & 1.8729 0.0024 & 1.9128 0.0023 rcccc & cold , gain = 14 , @xmath17 & 1.7168 @xmath75 0.0025 & 1.7300 @xmath75 0.0009 & 1.7556 @xmath75 0.0008 & 1.7365 @xmath75 0.0015 cold , gain = 14 , @xmath18 & 2.6096 0.0022 & 2.6315 0.0007 & 2.6368 0.0008 & 2.6275 0.0012 cold , gain = 7 , @xmath17 & 0.9601 0.0026 & 0.9678 0.0016 & 0.9750 0.0014 & 0.9732 0.0016 cold , gain = 7 , @xmath18 & 1.9343 0.0019 & 1.8720 0.0010 & 1.8991 0.0009 & 1.9252 0.0009 & @xmath120 cen , @xmath17 & 0.9478 0.0039 & 0.9850 0.0024 & 0.9861 0.0028 & 0.9760 0.0029 @xmath120 cen , @xmath18 & 1.8891 0.0036 & 1.8776 0.0013 & 1.8844 0.0024 & 1.8897 0.0022 , short , @xmath17 & 0.9621 0.0039 & 0.9466 0.0023 & 0.9824 0.0017 & 0.9709 0.0022 , short , @xmath18 & 1.9542 0.0019 & 1.8646 0.0012 & 1.9059 0.0011 & 1.9337 0.0010 , long , @xmath17 & 0.9818 0.0062 & 0.9791 0.0032 & 0.9525 0.0030 & 0.9720 0.0036 , long , @xmath18 & 1.9359 0.0041 & 1.8743 0.0038 & 1.8815 0.0023 & 1.9187 0.0023 rcccc warm , gain = 14 , @xmath17 & 1.7805 @xmath75 0.0019 & 1.7381 @xmath75 0.0011 & 1.7682 @xmath75 0.0008 & 1.7612 @xmath75 0.0009 warm , gain = 14 , @xmath18 & 2.5957 0.0037 & 2.5852 0.0010 & 2.6099 0.0008 & 2.6244 0.0011 warm , gain = 7 , @xmath17 & 1.0656 0.0047 & 0.9833 0.0043 & 1.0298 0.0025 & 1.0311 0.0037 rcccc cold , gain = 14 , @xmath17 & 1.7222 @xmath75 0.0010 & 1.7282 @xmath75 0.0005 & 1.7475 @xmath75 0.0004 & 1.7285 @xmath75 0.0005 cold , gain = 14 , @xmath18 & 2.6123 0.0008 & 2.6313 0.0003 & 2.6355 0.0004 & 2.6353 0.0003 cold , gain = 7 , @xmath17 & 0.9743 0.0012 & 0.9767 0.0005 & 0.9820 0.0004 & 0.9830 0.0005 cold , gain = 7 , @xmath18 & 1.9088 0.0012 & 1.8792 0.0004 & 1.8957 0.0004 & 1.9140 0.0004 warm , gain = 14 , @xmath17 & 1.7836 0.0007 & 1.7441 0.0004 & 1.7683 0.0004 & 1.7602 0.0005 warm , gain = 14 , @xmath18 & 2.5923 0.0013 & 2.5941 0.0004 & 2.6122 0.0004 & 2.6159 0.0005 warm , gain = 7 , @xmath17 & 1.0643 0.0034 & 0.9997 0.0016 & 1.0356 0.0017 & 1.0264 0.0024 | charge - transfer effects in photometry with wide field planetary camera 2 aboard the _ hubble space telescope _ are investigated by a comparison of wfpc2 observations with groundbased photometry for the galactic globular clusters tauri and .
simple numerical formulae describing the fraction of lost signal as functions of position on the detector , stellar brightness , and the diffuse sky brightness recorded in an image are presented , and the resulting corrections are compared to those previously derived by whitmore & heyer ( 1997 , _ instrument science report wfpc2 97 - 08 _ ) .
significant lost - charge effects are seen that are proportional to both the @xmath0 coordinate ( , the number of shifts along the parallel register during readout ) and the @xmath1 coordinate ( number of shifts along the serial register ) .
a `` typical '' star image ( one containing @xmath2 photoelectrons ) near the center of a `` typical '' intermediate - length exposure ( one with a diffuse sky brightness of @xmath310@xmath4e@xmath5/pixel , obtained at a camera temperature of 88@xmath4c ) loses approximately 2% of its electrons to charge traps during readout ; a star in the corner of the image most remote from the readout electronics loses twice that .
the percentage of charge lost decreases as the star brightness or the diffuse sky brightness increases .
charge losses during the brief period when wfpc2 was operated at a temperature of 76@xmath4c were approximately 85% greater , but apart from that no significant change in the charge transfer losses with time during the first 3.5 years of wfpc2 s mission is evident , except possibly a weak effect for the very faintest star images .
these results are quite similar to those of whitmore & heyer , which were based on a much smaller data set , but there are some differences in detail .
even with the present set of corrections , additional sources of calibration uncertainty which i am unable identify or characterize with the available data probably limit the _ external _ accuracy of photometry from wfpc2 to of order 12% .
= 10000 = 10000 # 1m@xmath4#1 # 1ngc@xmath4#1 # 1 and |
it is now well established that the stack of josephson junctions formed by the atomic layers of high - temperature superconductors ( htsc ) represents a nonlinear system with unique dynamic properties . a magnetic field perpendicular to the layers induces abrikosov - type pancake vortices in the planes .
in contrast , a magnetic field applied along the cuo@xmath7 layers ( parallel to the @xmath6-planes ) creates josephson vortices , each carrying one flux quantum , and whose cores are located between the superconducting layers .
when the parallel magnetic field exceeds the crossover field , @xmath8 , where @xmath9 is the flux quanta , @xmath10 is the anisotropy of the london penetration depth , and @xmath11 is the interlayer spacing , the cores of josephson vortices start to overlap and a josephson lattice is formed .
@xcite at the same time , the vortex ground state and the vortex phase diagram of highly - anisotropic ( layered ) htsc in parallel and oblique magnetic fields remains the open question .
the picture of a phase transition from a solid ( phase - ordered ) state to a liquid state of a combined lattice structure of josephson vortices and pancake vortices in high magnetic fields also remains unclear.@xcite due to the proximity to the mott insulator , phase fluctuation are strong and the temperature of the superconducting transition , @xmath12 , in cuprates may be governed by phase fluctuations . in two dimensions , the phase fluctuations can give rise to a kosterlitz - thouless transition which is described by the thermal unbinding of vortex - antivortex pairs .
however , there is no clear picture for the destruction of superconducting order in layered htsc via the unbinding of vortices .
@xcite . emery and kivelson @xcite have calculated the kosterlitz - thouless transition in layered htsc assuming that each layer is fluctuating independently , even for systems with strongly josephson coupled bilayers . on the other hand , corson _
@xcite using microwave conductivity have confirmed the kosterlitz - thouless nature of the phase transition but concluded that in bi@xmath7sr@xmath7cacu@xmath7o@xmath13 ( bi2212 ) system it is the bilayer which should be considered as a unit , i.e. , the superconducting phase is strongly correlated between the two layers of a bilayer . in view of these findings , it is interesting to look for a kosterlitz - thouless transition in the single - layer high-@xmath12 superconductor bi@xmath14sr@xmath14cuo@xmath3 where the interlayer spacing is considerable more .
as previously reported , @xcite this can be achieved by measuring the @xmath0 characteristics of the samples .
bi2212 single crystals , both in zero and in applied magnetic fields , show non - ohmic , power - law @xmath0 characteristics .
the observed power - law behavior , @xmath15 , with a characteristic jump in the exponent @xmath16 near the superconducting transition has been interpreted as evidence for the kosterlitz - thouless transition .
it is apparent that to gain a better understanding of the vortex ground state and vortex phase diagram in htsc , it is essential to understand the dynamics of josephson vortices . in the generally studied geometry , a @xmath17-axis external current drives josephson vortices in the direction perpendicular to both the current and the magnetic field .
the driven motion of the vortices is responsible for the observed flow resistance .
koshelev @xcite calculated the flux - flow resistivity of the josephson vortex lattice in a layered superconductor and showed that the magnetic field dependence of the flux - flow resistivity is characterized by three distinct regions . at low magnetic fields the flux - flow resistivity grows linearly with field . when the josephson vortices start to overlap the flux - flow resistivity crosses over to a regime of quadratic magnetic field dependence
. finally , at very high fields the flux - flow resistivity saturates at a value close to the @xmath17-axis quasiparticle resistivity .
a josephson lattice driven by a current along the @xmath17 axis , together with the resulting flux - flow resistance , has been intensively studied both experimentally and theoretically because the coherent josephson vortices motion gives rise to strong resonant phenomena which may lead to very important applications of high-@xmath12 superconductors .
@xcite in particular , the flux - flow voltage creates an oscillating current through the josephson effect and the current excites josephson plasma waves at terahertz frequency .
part of the energy may be emitted as electromagnetic waves .
@xcite under appropriate conditions , coherent motion of josephson vortices is possible and this can excite cavity resonances of the stack ( the fiske resonances ) , which in turn influence the vortex motion . such geometric resonances manifest themselves in the @xmath17-axis current voltage ( @xmath0 ) characteristic as a series of steps separated by a constant voltage .
@xcite recently , ooi _ et al . _
@xcite reported novel periodic oscillations of the flux - flow resistance with a @xmath17-axis bias current .
the measured periodicity , @xmath18 , is given by the relation @xmath19 where @xmath20 is the sample size in the @xmath6-plane along the direction perpendicular to the applied magnetic field .
such oscillations of the josephson vortices flow resistance can be explained by a matching effect between the lattice spacing of josephson vortices along the layers and the size of the sample .
it should be noted that in josephson - coupled layered superconductors , vortices exist for any orientation of the magnetic field . in the presence of an electrical current ,
the lorentz force on vortices causes motion which generates both in - plane and interplane electric fields and which induces a flux - flow voltage drop across the superconducting sample . the sliding motion of josephson vortices along the @xmath6-plane is easily driven by a current along the @xmath17 axis . for a driving force exerted along the @xmath17 axis ( current in the @xmath6-plane )
, the cuo@xmath7 planes effectively pin the josephson vortices preventing motion along the @xmath17 axis .
@xcite the dynamics of the josephson lattice along the @xmath17 axis is far from being fully understood . in tilted high fields @xmath21 at small angles between the applied field and the @xmath6-plane , a zigzag structure along the @xmath17 axis arising from the attractive interaction of pancake vortices with josephson vortices , see ref . [ ] . in this case
one might expect a series of maxima in the plasma frequency and the critical current at angles for which the pancake lattice is in resonance with the josephson lattice .
investigations of the dynamics of the vortex lattice in htsc are usually limited to the region near @xmath12 because the large critical current density @xmath22 makes low - temperature studies difficult , whereas in clean systems the vortex lattice is expected to melt due to thermal fluctuations .
values of the upper critical field @xmath23 and @xmath22 in the bi@xmath14sr@xmath14cuo@xmath24 system are relatively low , which allows to observe the effect of strong magnetic fields on the dynamics of the josephson vortex lattice at very low temperatures below and above a melting line where the magnetic field behavior depends on the mechanism of dissipation . here , we present an experimental study of the in - plane @xmath0 characteristics and the in - plane josephson vortex flow resistance in parallel and tilted magnetic fields in single - layer la - free bi@xmath14sr@xmath14cuo@xmath24 single crystals at temperatures down to @xmath4 mk . at small angles
@xmath25 between the applied field and @xmath6-plane , we have found prominent current steps in the @xmath0 characteristics and periodic oscillations in the josephson - vortex flow resistance . to gain a better understanding as to the nature of the observed current steps and periodic oscillations , we have compared the behavior of the in - plane flux flow resistance in bi2201 single crystals with the published data for bi2212 with double cuo@xmath7 layers .
the high - quality la - free bi@xmath1sr@xmath2cuo@xmath3 ( bi2201 ) single crystals used for the present study were grown by a kcl - solution - melt free growth method .
@xcite the preparation and characterization of bi2201 single crystals are described in detail elsewhere . @xcite as before , in this work we have used _ as - grown _ single crystals .
the properties of the samples investigated in this work are summarized in table [ table1 ] .
the temperature dependence of the resistance of the present samples in zero magnetic field are essentially the same as those we reported previously .
@xcite we have measured the resistance , @xmath0 characteristics and the differential resistance @xmath26 characteristics for bi2201 single crystals using a standard four - probe technique with symmetrical positions of the low - resistance contacts on both @xmath6 surfaces of the sample.@xcite the current was applied parallel to the @xmath6-plane in all cases . in the experimental arrangement used , the crystal could be rotated @xmath27 relative to the direction of the magnetic field with an angular resolution better than @xmath28 .
the @xmath29 orientation was precisely determined from the lowest value of the resistance at a fixed temperature .
a configuration with @xmath30 was used always .
measurement procedures in the resistive magnet are described in detail in refs .
.doping ( @xmath31 ) , critical temperature ( @xmath32 ) determined from the 50% point of the resistive transition , and dimensions for the bi2201 single crystals investigated .
optimal doping in bi2201 occurs at @xmath33 holes per cu . [ cols="^,^,^,^",options="header " , ]
as reported in ref . [ ] , a parallel external magnetic field plays a crucial role in the dynamics of the kosterlitz - thouless transition , and in particular , reduces the transition temperature . in sufficiently strong magnetic fields parallel to the @xmath6-plane ( up to 8 t )
, the kosterlitz - thouless transition is suppressed because the magnetic field - induced vortices increase the dissipation and reduces the stability of the vortex - antivortex pairs .
however , subsequently , it has been suggested @xcite that in bi2212 at high parallel magnetic fields , there is a multi - critical point in the @xmath34 phase diagram characterized by a magnetic field @xmath35 , above which exists a novel intermediate kosterlitz - thouless type phase .
the existence of this intermediate phase was also confirmed by the non - ohmic power - law @xmath0 characteristics in the bi2212 system .
chen and hu @xcite very recently investigated theoretically the in - plane @xmath0 characteristics of interlayer josephson vortices in cuprate htsc using a computer simulation and showed that for highly anisotropic systems ( @xmath36 ) and at high magnetic field , the power - law in - plane @xmath0 characteristics changes it s exponent from 1 ( i.e. , ohmic ) at high temperatures to @xmath37 at intermediate temperatures , which can be attributed to a kosterlitz - thouless transition . in layered high-@xmath12 superconductors ,
the anisotropy is usually expressed by the ratio @xmath38 , where @xmath39 is the london penetration depth for super currents along the @xmath17 axis , and @xmath40 is the penetration depth for currents in the @xmath6-plane .
the anisotropy ratio @xmath10 can also be expressed by the ratio @xmath41 between the effective masses of the quasiparticles along the @xmath17-axis and in the @xmath6-plane , which can be related to the transport anisotropy with @xmath42 using @xmath43 for the out - of - plane resistivity @xmath44 and in - plane resistivity @xmath45 . @xcite .
owing to the large value of @xmath23 in parallel magnetic fields , in order to study the power - law behavior of the @xmath0 characteristics of bi2201 single crystals , we have used underdoped samples with a low @xmath12 .
previously , we found @xcite that the normal - state anisotropy ratio @xmath46 in high magnetic fields and at low temperatures is practically temperature independent .
for this reason , we can safely neglect any variation of @xmath10 for the temperatures used here . for the underdoped # 7 ( @xmath47 k ) and heavily underdoped # 2 ( @xmath48 k ) samples , the anisotropy is equal to @xmath49 and @xmath50 , respectively . in this case
the values of the multi - critical magnetic field @xmath51 should be @xmath52 t and @xmath53 t. the in - plane current - voltage characteristics for samples # 7 and # 2 near the magnetic - field induced suppression of superconductivity at 0.6 k and 40 mk are plotted using a logarithmic scale in fig .
[ fig1](a ) and fig .
[ fig1](b ) , for selected magnetic fields ( @xmath54 ) .
the temperature dependence of the resistance @xmath55 of the present samples in zero magnetic field ( not shown ) are essentially the same as those reported previously for samples with the same doping .
@xcite ( color on - line ) ( a ) @xmath0 characteristics plotted using a logarithmic scale for samples # 7 ( a ) and # 2 ( b ) near the magnetic - field induced suppression of the superconductivity at 0.6 k and 40 mk for selected magnetic fields oriented parallel to the @xmath6-plane .
the dashed lines are power - law fits to the data .
( c ) and ( d ) magnetic - field dependence of the power - law exponent for samples # 7 and # 2 , respectively . ]
the @xmath0 characteristics can be fitted to a power law for currents above a critical current @xmath56 ( dashed lines in fig.[fig1](a - b ) ) .
the extracted power law exponent , @xmath57 , of the @xmath0 characteristics , is plotted in figures [ fig1](c ) and [ fig1](d ) as a function of magnetic field for samples # 7 and # 2 , respectively , at 0.6 k , 50 mk and 40 mk . at @xmath58
k , @xmath57 decreases rapidly with increasing field and its value tends to unity in high magnetic fields , while at very low temperature , @xmath57 decreases much more slowly .
however , although the @xmath0 characteristics show a power - law behavior with the change in the value of exponent from 1 to above 5 [ fig .
[ fig1](c ) ] and the magnetic fields far exceed the @xmath51 value , both the figures show that @xmath57 decreases smoothly with increasing field , without any features that might be interpreted as evidence for a transition to the kosterlitz - thouless type phase .
it should be noted that the magnetic fields 23 t and 28 t applied here are far in excess of the fields at which the 3d vortex melting occurs in low @xmath12 bi2201 samples . @xcite
it has been suggested in refs .
[ ] , that the power law dependence of the @xmath0 curves at low temperatures is related to the interplay of the in - plane vortex lines and the two - dimension pancake vortices created by a small component of the magnetic field parallel to the @xmath17 axis . that can appear because of a small misalignment to the @xmath17 axis within the crystal .
however , we also observe a power - law behavior in the @xmath6-plane @xmath0 characteristics when the magnetic field is applied perpendicular to @xmath6-plane . in fig .
[ fig1](d ) , we also plot the power - law exponent versus magnetic field applied along the @xmath17-axis for crystal # 2 at @xmath59 mk .
one can see that @xmath57 lies in the same range for both orientations of the field . in view of this , it is difficult to imagine that a possible small @xmath17-axis component of the magnetic field ( in the parallel geometry ) gives the same change in the value of @xmath57 as the large magnetic field oriented along the @xmath17-axis .
in addition , we can formally exclude macroscopic sample inhomogeneity , as the origin of the observed the power - law behavior in @xmath0 characteristics in fig .
[ fig1](a ) and fig .
[ fig1](b ) , the crystals are of a very high quality judging from the small rocking curve width @xmath60
. in high magnetic fields , the in - plane @xmath0 characteristics in low bias region ( small currents ) showed unusual behavior for a superconductor .
this feature in the @xmath0 characteristic in the vicinity of the zero bias can be more clearly seen the measured first derivatives @xmath26 for the sample # 2 with current and voltage contacts on the same surface of the crystal at a temperature of @xmath59 mk with the magnetic field oriented perpendicular and parallel to the @xmath6-plane .
( color on - line ) differential resistance @xmath26 near the current induced suppression of superconductivity at @xmath59 mk for different magnetic fields applied along the @xmath17-axis of the crystal # 2 ( a ) and along @xmath6-plane ( b ) . for clarity , the curves in ( b ) have been shifted vertically ( downwards ) with respect to the upper curve .
the inset in ( a ) shows the magnetoresistance @xmath61 measured at @xmath59 mk for the sample # 2 with the magnetic field applied along the @xmath17-axis .
the inset in ( b ) shows the magnetoresistance @xmath61 measured at @xmath59 mk for the sample # 2 in a parallel magnetic field ( @xmath54 ) . ] in fig .
[ fig2](a ) ( main panel ) , we plot the voltage dependence of the differential resistance @xmath26 near the current induced suppression of the superconductivity in different magnetic fields applied along the @xmath17-axis of the crystal # 2 at @xmath59 mk . in the inset
, we plot the the magnetoresistance @xmath61 measured at @xmath59 mk for the same sample with the magnetic field applied along the @xmath17-axis . as can be seen , the differential resistance @xmath26 saturates at high fields in accordance with the exponent @xmath62 in fig . [ fig1](c ) and fig . [ fig1](d ) .
the @xmath26 curves have a minimum close to zero bias voltage , the amplitude of which decreases with increasing magnetic field .
this minima should disappear at a magnetic field corresponding to the transition of the sample to the normal state .
what is observed is slightly different since this minimum transforms to a maximum for magnetic fields @xmath63 t after the suppression of the superconductivity ( marked by arrows in the inset and main panel ) .
it can be supposed that such a behavior of the differential resistance in the vicinity of low bias at magnetic fields above the resistive upper critical field @xmath5 is connected with moving vortex - like excitations which have been previously observed in superconducting cuprates at @xmath64 in magnetic fields by the detection of a nernst signal @xcite and measurement of an angular dependence of the magnetoresistance on y@xmath65pr@xmath66ba@xmath7cu@xmath67o@xmath68 single crystals .
@xcite with increasing current , the number of vortices decreases and the differential resistance in fig .
[ fig2](a ) decreases rapidly .
figure [ fig2](b ) ( main panel ) shows the analogous @xmath26 curves for the same sample in the magnetic fields applied in the @xmath6-plane of the crystal at @xmath59 mk . for clarity
, the curves have been shifted vertically with respect to the upper curve . in the inset
we show the magnetoresistance @xmath61 measured at @xmath59 mk for the same sample in a parallel magnetic field ( the onset of the superconducting transition is marked by an arrow ) . despite the different types of vortex lattice ( the pancake vortices in perpendicular and josephson vortices in parallel configurations ) , the @xmath26 curves in fig .
[ fig2](b ) have a general form identical to that in fig .
[ fig2](a ) . at our maximum field
@xmath69 t , the resistance of the sample has only reached 70% of its normal - state value and the minimum of the differential resistance in the vicinity of low bias currents is not completely suppressed .
this suggests that the mechanism of the dissipation in the neighborhood of @xmath5 is common to both magnetic field directions .
these results allows us to conclude that the behavior of the flux flow resistance in the @xmath6-plane in bi2201 single crystals at low temperatures is largely similar to that in bi2212 except for the absence of a transition to the kosterlitz - thouless phase .
apparently , this transition does not take place at low temperatures even in high magnetic fields .
most likely the kosterlitz - thouless transition in bi2201 can be observed only in a very narrow temperature region near @xmath12 .
the interaction of the pancake vortices with the josephson vortices can lead to features in the @xmath0 characteristics with the transport current along the @xmath6-plane when the period of the pancake vortex lattice matches the period of the josephson vortex lattice . in order to rule out the presence of spatial inhomogeneity , which can be present in underdoped samples and which may influence the results found , we have used for these measurements an optimally doped bi2201 single crystal ( # 5 ) with @xmath70 k. to investigate the influence of misalignment between the magnetic field and the superconducting layers , the sample was rotated at small angles around its position of parallel to the layers .
temperatures for these measurements were chosen so that the available magnetic field would suffice to observe the flux - flow voltage while remaining below the melting line . with increasing current in the @xmath6-plane ,
we have found prominent current steps in the @xmath0 characteristics .
the amplitude of the current steps depends on the angle between the applied field and the @xmath6-plane . at very small angles
@xmath25 , as in the parallel configuration of field , no current steps are observed in the @xmath0 characteristics .
this is consistent with the sample remaining free of pancake vortices until the normal component @xmath71 exceeds the perpendicular critical field @xmath72 for pancake formation .
@xcite for a detailed investigation , we chose the orientation of the field for which the amplitude of the current steps was a maximum .
( color on - line ) ( a ) @xmath0 characteristics of sample # 5 with @xmath70 k in various tilted ( @xmath73 ) magnetic fields at 0.8 k with the transport current flowing in the @xmath6-plane perpendicular to the field .
current steps are indicated by arrows .
the inset in fig .
[ fig3](a ) shows similar characteristics for two different magnetic fields at @xmath74 .
( b ) voltage position of the current steps versus magnetic field ( solid symbols ) the number of the current step is indicated for each curve .
open circles show the flux - flow voltage for a current of @xmath75 ma .
the dashed lines are fits to the experimental data .
the inset shows the voltage position of the current steps versus the step number for different fields . ]
figure [ fig3](a ) ( main panel ) displays the @xmath0 characteristics for the single crystal # 5 in various tilted magnetic fields at 0.8 k with transport current flowing in the @xmath6-plane perpendicular to the applied magnetic field .
current steps are clearly observed ( for @xmath76 t , they are marked by arrows ) .
the angle @xmath25 between the applied field and the @xmath6-plane was equal to @xmath77 .
the inset in fig .
[ fig3](a ) shows similar characteristics for two different magnetic fields at @xmath74 . in fig .
[ fig3](b ) ( main panel ) , we plot the voltage position of the current steps versus magnetic field ( the number of the current step are indicated for each curve ) .
it can be seen that the voltage position of each step increases nonlinearly with magnetic field ( solid symbols ) as well as the flux - flow voltage , at the current 10 ma showed by open circles .
it is evident that the data can be fitted quite well to curves of the form @xmath78 where @xmath79 and @xmath80 are constants ( dashed lines ) .
these steps can not be attributed to fiske steps caused by an interaction of the ac josephson effect with electromagnetic modes of a cavity formed by the josephson junctions ( geometric resonance ) for a number of reasons .
firstly , the dc voltage is applied along the @xmath6-plane instead of across the junctions and secondly , the voltage position of the current steps depends on the magnetic field which is not expected for fiske resonances .
furthermore , while the current steps are periodic in the voltage at constant fields , the step period increases with increasing magnetic field . this can be seen in the inset of fig .
[ fig3](b ) where we show the voltage position of the current steps versus the step number for different fields . on the other hand , the range of the magnetic fields in fig .
[ fig3](a ) at @xmath81 exceeds the crossover field @xmath82 for sample # 5 and also the field component perpendicular to the layers exceeds the perpendicular lower critical field @xmath72 for pancake formation . under these conditions , according to the model of bulaevskii _ et al . _
@xcite , the tilted magnetic field induces a triangular lattice of josephson vortices and pancake vortices which form a zigzag structure along the @xmath17 axis . in this case
one might expect a series of maxima in the @xmath17-axis plasma frequency and the @xmath17-axis critical current at angles where the pancake lattice is in resonance with the josephson lattice .
such zigzag structure result in oscillations of the out - of - plane resistance and a sharp increase of the @xmath17-axis critical current in bi2212 single crystal at 30 k in field 20 t at small angles near @xmath29 with width @xmath83 .
@xcite based on the model of bulaevskii__et al . _ _
@xcite , we estimate that the periods of the josephson and pancake lattices in our experiment , for example , at @xmath76 t should be matched at @xmath84 .
this is in disagreement with the observed maximum amplitude of the current steps at @xmath85 .
we do not understand the reason for this discrepancy .
however , subsequent studies @xcite after ref .
[ ] have shown that the maximum of such resonance in bi2212 single crystals at @xmath86 k takes place at the angle @xmath87 and the angular dependence of the resonance field is much more drastic than the result of ref .
[ ] . in our case , with the magnetic field component parallel to the layers and the current flowing in the @xmath6-plane , the lorentz force acting on the vortices drives them parallel to the @xmath17 axis with jumps across the superconducting layers .
the motion of the josephson lattice generates both in - plane and out - of - plane alternating electric fields and currents , which are coupled to electromagnetic plasma waves .
there should exist a strong resonance emission when the velocity of the lattice matches the velocity of the plasma wave .
@xcite the matching of the josephson frequency with the frequency of the emission may manifest itself as the current steps not only in the @xmath17-axis current , but possibly in the in - plane current as well [ fig .
[ fig3](a ) ] . since the behavior of the plasma mode in parallel and tilted fields
is poorly understood at present , in contrast to the perpendicular magnetic field geometry , we are unable to discus here this topic more fully . however , since it was assumed that the periodic steps in the in - plane @xmath0 characteristics may be related to the moving josephson vortex lattice along the @xmath17 axis , we have tried to experimentally establish this fact . as noted above , ooi _ et al .
_ @xcite found periodic oscillations of the @xmath17-axis flow resistance of josephson vortices as a function of the in - plane magnetic field in small - strip bi2212 single crystal in a wide range of temperatures and fields .
a resistance of the strip sample in used geometry was almost equal to the @xmath17-axis resistance of the intrinsic junctions and the contribution of in - plane resistivity was neglected .
these oscillations were related to a matching effect between the lattice spacing of josephson vortices along the layers and the width of the sample .
subsequently , koshelev @xcite and machida @xcite have reproduced the oscillation by calculating the flux flow resistance by taking the surface barrier effect into consideration .
an attempt to observe the periodic oscillations of the @xmath6-plane flow resistance of josephson vortices in a parallel magnetic field similar to ref .
[ ] in our crystal with usual sizes @xcite has been unsuccessful .
because of this , we have measured the in - plane josephson vortex flow resistance versus parallel and tilted magnetic fields in a bi2201 strip single crystal # 4 with @xmath88 k. in spite of the low @xmath12 , a careful characterization of the crystal showed the high quality , the high homogeneity and the structural perfection of the sample . as the magnetic field was oriented parallel to the @xmath6-planes exactly , the @xmath6-flow resistance increased smoothly with increasing field because only the number of vortices increased . at the same time , the resistance begins to oscillate as a function of the magnetic field for any small deviation of the field from the exactly parallel geometry as shown by three curves in fig .
the amplitude of oscillations gradually diminishes with increasing magnetic field and oscillations are not observed at fields above 15 t. the oscillations also rapidly vanish when the angle @xmath25 between the applied field and the @xmath6-plane exceeded @xmath89 .
the maximum amplitude of the oscillations was reached at @xmath90.at the same time , the oscillations are most prominent at higher temperatures where thermal fluctuations reduce the role of background pinning and the crystal behaves as if it contains fewer defects than it actually does .
( color on - line ) flow resistance of the josephson vortices as a function of magnetic field for sample # 4 with an applied current of @xmath91 at different temperatures .
the inset shows the magnetic field position of the resistance minima ( marked by arrows in the main panel ) versus the number ( @xmath92 corresponds to the minima closest to zero bias ) at @xmath93 k. ] it can be seen from the inset in fig .
[ fig4 ] that the period of the oscillations @xmath94 t is to a good approximation constant over a wide range of magnetic fields .
the period also does not depend on the temperature ( see curves for three different temperatures in fig . [ fig4 ] ) .
the oscillations can be observed for almost all temperatures below @xmath12 , but have reduced amplitude at very low temperatures and near @xmath12 .
the flow resistance oscillations in fig .
[ fig4 ] are quite similar to the oscillations of the @xmath17-axis josephson - vortex flow resistance in bi2212 single crystals previously observed by ooi _
the oscillations in ref .
[ ] originate from a matching between the triangular lattice and the width @xmath20 of the junction perpendicular to the magnetic field direction .
the period of the oscillations , @xmath18 , corresponded to adding one flux quantum @xmath9 per two layers in the crystalline lattice .
the minima in the resistance corresponded to the condition where the magnetic field was @xmath95 ( @xmath96 is an integer ) . here
@xmath97 is the field needed to add one vortex per layer in a bi2212 , where @xmath11 is the periodicity of the cuo@xmath7 double layers . in ref .
[ ] , a transport current flowing across the layers exerts a lorentz force on the josephson vortices in the plane of layers .
an important parameter is the critical current above which the lattice starts to move , producing a finite voltage .
this current is determined either by bulk pinning or by interaction with the boundaries .
@xcite for the measured flow resistance in fig .
[ fig4 ] , a transport current flowing along the layers increases the effect of the lorentz force on the josephson vortices across the layers . due to the interaction with the boundaries of the sample , the period of oscillations in fig .
[ fig4 ] should depend on the size of the sample along the @xmath17-axis direction .
the flow resistance shows periodic oscillations with period @xmath94 t which is very close ( within the accuracy of measuring the sample thickness ) to the estimated @xmath98 t , corresponding to the magnetic field needed to add one vortex quantum per layer in the crystalline lattice .
here @xmath99 is the size of our strip single crystal along @xmath17 axis ( thickness ) and @xmath100 is the distance between the superconducting cuo@xmath7 layers in bi2201 .
this result suggests that josephson vortices in bi2201 form a square lattice in the ground state . as before ,
the oscillations are most prominent in the high temperature regime where thermal fluctuations reduce the role of background pinning and the crystal behaves as if it contained fewer defects than it actually does .
we note that somewhat similar commensurate oscillations of the @xmath6-plane resistivity for magnetic fields nearly aligned with the @xmath6-plane has been observed on yba@xmath7cu@xmath67o@xmath68 single crystals in earlier work by gordeev _ _ et al.__@xcite .
these oscillations are due to the oscillatory melting line separating the vortex liquid from the vortex smectic state and they are periodic in @xmath101 .
this is very different from the periodic in @xmath102 oscillations of the vortex - flow resistance reported in ref .
[ ] and observed here , which in addition are observed over a wide range of temperatures and magnetic fields below and above a melting line . finally , it is pertinent to recall here a transport phenomenon which has its origins in a lorentz - force - independent dissipation that has been observed in strongly coupled layered bi- and tl - based materials.@xcite for this case the resistivity in the non - ohmic regime did not depend on the angle between the magnetic field and current when they are both parallel to the layers .
the problem of an orientation - independent dissipation and the role of the lorentz force in the motion of the josephson vortex lattice can not be explained by simple flux - flow theory and is far from being totally understood .
turning back to fig .
[ fig3](a ) , we can suppose that the motion of the josephson lattice generates a traveling electromagnetic wave with the @xmath6-plane wave vector selected by the lattice structure .
one might expect resonance phenomena when the velocity of the lattice matches the plasma wave velocity .
@xcite on the other hand , if the motion of the josephson vortex lattice generates a voltage across the layers , one would expect that the current steps in fig .
[ fig3](a ) are due to the matching of the josephson frequency with the frequency of the plasma mode .
both mechanisms may result in the current steps in the @xmath0 characteristics .
in summary , we have studied the in - plane @xmath0 characteristics and josephson vortex flow resistance in la - free bi2201 single crystals in parallel and tilted magnetic fields at temperatures down to @xmath4 mk . for parallel magnetic fields ,
we find a power - law @xmath0 characteristic with a smooth decrease with magnetic - field of the exponent describing the power - law . at small angles between the applied field and @xmath6-plane , prominent current steps in the @xmath0 characteristics and periodic oscillations of josephson - vortex flow resistance
are observed .
while the current steps are periodic in the voltage at constant fields , the voltage position of the steps , together with the flux - flow voltage , increases nonlinearly with increasing magnetic field .
the @xmath6-flow resistance oscillates as a function of magnetic field with a constant period over a wide range of magnetic fields and temperatures .
the current steps in the @xmath0 characteristics and the flow resistance oscillations can be linked to the motion of the josephson vortices across layers . this work has been partially supported by pics grant no .
one of us ( s.i.v . ) was partially supported by the russian foundation for basic research projects no .
06 - 02 - 22001 and no .
07 - 02 - 00349 .
the work at ghmfl was partially supported by the european 6th framework program under contract number rita - ct-3003 - 505474 . | we have investigated the in - plane @xmath0 characteristics and the josephson vortex flow resistance in high - quality la - free bi@xmath1sr@xmath2cuo@xmath3 ( bi2201 ) single crystals in parallel and tilted magnetic fields at temperatures down to @xmath4 mk . for parallel magnetic fields below the resistive upper critical field @xmath5 ,
the @xmath0 characteristic obey a power - law with a smooth change with increasing magnetic - field of the exponent from above 5 down to 1 .
in contrast to the double - layer cuprate bi2212 , the observed smooth change suggests that there is no change in the mechanism of dissipation ( no kosterlitz - thouless transition ) over the range of temperatures investigated . at small angles between the applied field and the @xmath6-plane , prominent current steps in the @xmath0 characteristics and periodic oscillations of josephson - vortex flow resistance
are observed .
while the current steps are periodic in the voltage at constant fields , the voltage position of the steps , together with the flux - flow voltage , increases nonlinearly with magnetic field .
the @xmath6-flow resistance oscillates as a function of field with a constant period over a wide range of magnetic fields and temperatures .
the current steps in the @xmath0 characteristics and the flow resistance oscillations can be linked to the motion of josephson vortices across layers . |
the pseudo - scalar mesons @xmath0 and @xmath1 represent a subject of considerable interest since some time ( cf .
@xcite for surveys ) .
investigations of various aspects of @xmath0 and @xmath1 mesons are tightly related with several theoretical challenges and can augment the experimental information on different phenomenological model parameters .
for instance , the `` anomalously '' large mass of the @xmath1 meson , as member of the @xmath2 nonet @xcite , can be directly connected with the @xmath3 axial anomaly in qcd .
a combined phenomenological analysis of @xmath0 and @xmath1 production in @xmath4 reactions together with the @xmath5 anomaly provides additional information on the gluon - nucleon coupling , which can be used to describe , e.g. , the so - called `` spin crisis '' .
also , the knowledge of the nucleon - nucleon-@xmath1 coupling strength allows a better understanding of the origin of the ozi rule violation in @xmath4 reactions .
a remarkable fact is that near the threshold the invariant mass of the @xmath6 system in such reactions is in the region of heavy nucleon resonances , i.e. resonances with isospin @xmath7 can be investigated via these processes .
furthermore , the so - called `` missing resonances '' can be studied .
another aspect of @xmath0 and @xmath1 production in elementary hadron reactions is that both mesons have significant dalitz decay channels into @xmath8 . as such
, they constitute further sources of di - electrons .
it is , in particular , the @xmath0 which is a significant source of @xmath9 pairs , competing with @xmath10 dalitz decays and bremsstrahlung , as the analysis @xcite of hades data @xcite shows .
one of the primary aims of the hades experiments @xcite is to seek for signals of chiral symmetry restoration in compressed nuclear matter .
for such an endeavor one needs a good control of the background processes , including the @xmath1 dalitz decay , in particular at higher beam energies , as becoming accessible at sis100 within the fair project @xcite .
the @xmath0 and @xmath1 dalitz decays depend on the pseudo - scalar transition form factor , which encodes hadronic information accessible in first - principle qcd calculations or qcd sum rules .
the dalitz decay process of a pseudo - scalar meson @xmath11 can be presented as @xmath12 .
obviously , the probability of emitting a virtual photon is governed by the dynamical electromagnetic structure of the `` dressed '' transition vertex @xmath13 which is encoded in the transition form factors . if the decaying particle were point like , then calculations of mass distributions and decay widths would be straightforwardly given by qed .
deviations of the measured quantities from the qed predictions directly reflect the effects of the form factors and thus the internal hadron structure .
the present paper reports parameterizations of @xmath0 and @xmath1 production cross sections in nucleon - nucleon collisions near the respective thresholds within a one - boson exchange model .
emphasis is put on the accessibility of transition formfactors encoding the strong - interaction @xmath14 structure .
cross sections of interest are @xcite @xmath15 with two - particle invariant phase space @xmath16 for the production of @xmath17 and @xmath18 we employ here a one - boson exchange model , where the @xmath0 production is described by the diagrams exhibited in fig .
the sum of these diagrams generate the invariant amplitude @xmath22 via interaction lagrangians .
-1 mm within the one - boson exchange model .
a ) dalitz decays of @xmath0 mesons from bremsstrahlung like diagrams . the intermediate baryon @xmath23 ( triple line )
can be either a nucleon or a nucleon resonance ( @xmath24 , @xmath25 , @xmath26 ) .
analog diagrams for the emission from fermion line @xmath27 .
b ) dalitz decay of @xmath0 mesons from internal meson conversion .
exchange diagrams are not displayed .
later on we identify @xmath28 and denote the di - electron invariant mass by @xmath29.,title="fig:",scaledwidth=100.0% ] -3 mm the calculation of @xmath0 uses the same diagram topology as in fig .
[ fig2 ] ( with @xmath30 ) supplemented by @xmath31 exchange in a ) .
the included resonances are @xmath32 with odd parity , and @xmath33 and @xmath34 with even parity .
the employed interaction lagrangians can be represented as follows .
+ ( i ) nucleon currents : @xmath35 ( ii ) spin @xmath36 resonances ( @xmath37 and @xmath38 ) : @xmath39\end{aligned}\ ] ] ( iii ) spin @xmath40 resonances ( @xmath41 and @xmath42 ) : @xmath43 with the abbreviations @xmath44 or @xmath0 or @xmath0 , @xmath45 or @xmath46 , @xmath47 or @xmath48 , and @xmath49 .
furthermore needed interactions , such as @xmath50 , @xmath51 , @xmath52 , and @xmath53 , are listed in @xcite .
strong formfactor are needed to dress the nucleon
nucleon ( resonance ) meson vertices .
these are listed in detail in @xcite .
the electromagnetic formfactors encode non - perturbative transition matrix elements @xmath54 in ( [ dgamma ] ) , basically accessible within qcd . here , however , we contrast a few parameterizations : ( i ) so - called qed formfactor meaning a structure - less particle with @xmath55 , ( ii ) a parametrization suggested by the vector meson dominance ( vmd ) model @xmath56 with @xmath57 , @xmath58 and @xmath59 .
the values of @xmath60 are quoted in @xcite . for the case of @xmath0 ,
the kinematically accessible region is restricted and , as a consequence , the @xmath61 contribution is sufficient .
( iii ) for @xmath0 , a monopole fit @xmath62 @xcite may be used , which does not differ too much from the vmd parametrization .
initial state interactions are accounted for by effective reduction factors for @xmath63 , @xmath64 waves : @xmath65 @xcite .
final state interactions are treated by the jost function formalism , see @xcite for details .
-3 mm ( top ) and @xmath66 ( bottom ) production from @xcite ( left , data for @xmath67 from @xcite ( open circles ) , @xcite ( triangles ) and @xcite ( squares ) and for @xmath66 from @xcite ) .
the new data situation confirms these predictions ( right , with data for @xmath67 from @xcite and for @xmath66 from @xcite , both ones depicted as filled circles).,title="fig : " ] ( top ) and @xmath66 ( bottom ) production from @xcite ( left , data for @xmath67 from @xcite ( open circles ) , @xcite ( triangles ) and @xcite ( squares ) and for @xmath66 from @xcite ) .
the new data situation confirms these predictions ( right , with data for @xmath67 from @xcite and for @xmath66 from @xcite , both ones depicted as filled circles).,title="fig : " ] ( top ) and @xmath66 ( bottom ) production from @xcite ( left , data for @xmath67 from @xcite ( open circles ) , @xcite ( triangles ) and @xcite ( squares ) and for @xmath66 from @xcite ) .
the new data situation confirms these predictions ( right , with data for @xmath67 from @xcite and for @xmath66 from @xcite , both ones depicted as filled circles).,title="fig : " ] ( top ) and @xmath66 ( bottom ) production from @xcite ( left , data for @xmath67 from @xcite ( open circles ) , @xcite ( triangles ) and @xcite ( squares ) and for @xmath66 from @xcite ) .
the new data situation confirms these predictions ( right , with data for @xmath67 from @xcite and for @xmath66 from @xcite , both ones depicted as filled circles).,title="fig : " ] -1 mm these seemingly many ingredients ( coupling strengths , formfactors and their cut - offs , see @xcite ) may cause the impression that the one - boson exchange approach to hadronic observables near threshold does not have too much predictive power .
two counterexamples may lend more credibility to the approach . in fig .
[ obe ] the model results of @xcite are exhibited ( left panels ) . later on the data basis
has been improved confirming the model predictions ( right panels ) .
further applications of the present approach to @xmath67 and @xmath66 production involving a final deuteron and including polarization observables , have been presented in @xcite , while @xcite extends the formalism to virtual bremsstrahlung in @xmath68 reactions .
-3 mm ( top ) and @xmath0 ( bottom ) production as a function of the energy excess in @xmath69 ( left ) and @xmath70 reactions ( right ) . for data quotation and further details cf .
@xcite.,title="fig : " ] ( top ) and @xmath0 ( bottom ) production as a function of the energy excess in @xmath69 ( left ) and @xmath70 reactions ( right ) . for data quotation and further details cf .
@xcite.,title="fig : " ] + -6 mm ( top ) and @xmath0 ( bottom ) production as a function of the energy excess in @xmath69 ( left ) and @xmath70 reactions ( right ) . for data quotation and further details cf . @xcite.,title="fig : " ] ( top ) and @xmath0 ( bottom ) production as a function of the energy excess in @xmath69 ( left ) and @xmath70 reactions ( right ) . for data quotation and further details cf .
@xcite.,title="fig : " ] -3 mm
numerical evaluation of the given formalism results in the total cross sections exhibited in fig .
[ fig1_-2 _ ] .
available data are nicely reproduced in the @xmath71 channel ( a concern could be the region of excess energy @xmath72 mev for @xmath0 ) . since now new parameters enter , the channel @xmath70 represents a prediction , in agreement with data in case of @xmath0 ; data are not ( yet ) available for @xmath0. -6 mm ( left , hades data from @xcite , for @xmath73 gev ) and @xmath0 ( right , for @xmath74 gev ) which give access to the formfactors . , title="fig : " ] ( left , hades data from @xcite , for @xmath73 gev ) and @xmath0 ( right , for @xmath74 gev ) which give access to the formfactors .
, title="fig : " ] -3 mm the cross sections @xmath75 , resulting from the integration of ( [ two ] ) over @xmath76 , are exhibited in fig .
[ fig7 - 8 _ ] .
there is a tiny difference when neglecting the internal strong interaction structure of @xmath0 ( qed formfactor ) or when using the vmd formfactor , see left panel .
the situation changes drastically for @xmath0. here , the account of the internal structure becomes important , see right panel .
precision data would even allow for a sensible test of the vmd hypothesis .
it has been shown in @xcite that the formfactors can be deduced from given cross sections @xmath77 .
in summary we report on calculations of the reaction @xmath78 with @xmath79 and subsequent dalitz decay @xmath80 within a one - boson exchange model .
we point out that isolating @xmath0 and @xmath0 contributions , e.g. , in @xmath71 collisions , allows for an experimental determination of the transition formfactors @xmath54 .
in particular , for @xmath0 the vector meson dominance hypothesis would be testable .
on the other hand , the @xmath0 dalitz decay channel is a strong source of @xmath9 pairs in medium - energy heavy - ion collisions which need to be understood before firm conclusions on possible in - medium modifications of hadrons can be made .
we emphasize that , once the model parameters are adjusted in the @xmath69 channel , the @xmath81 channel is accessible without further parameters .
99 b. diekmann , phys .
159 , 99 ( 1988 ) ; th .
feldmann , int .
phys . a 15 , 159 ( 2000 ) ;
christos , phys . rep .
116 , 251 ( 1984 ) ; l.g .
landsberg , phys . rep .
128 , 301 ( 1985 ) .
s.d . bass , phys .
b 463 , 286 ( 1999 ) . h.w .
barz , b. kmpfer , gy .
wolf , m. zetenyi , nucl - th/0605036 ; m. thomere , c. hartnack , gy .
wolf , j. aichelin , phys .
c 75 , 064902 ( 2007 ) ; e.l .
bratkovskaya , w. cassing , nucl .
phys . a 807 , 214 ( 2008 ) .
g. agakichiev et al .
( hades collaboration ) , phys .
98 , 052302 ( 2007 ) ; phys .
b 663 , 43 ( 2008 ) .
fair home page : _ http://www.gsi.de / fair / reports / index.html_. l.p .
kaptari , b. kmpfer , eur .
j. a 37 , 69 ( 2008 ) .
kaptari , b. kmpfer , eur .
j. a 33 , 157 ( 2007 ) .
k. nakayama , j. speth , t .- s.h .
lee , phys .
c 65 , 045210 ( 2002 ) ; + k. nakayama , j. haidenbauer , c. hanhart , j. speth , phys .
c 68 , 045201 ( 2003 ) .
titov , b. kmpfer , b.l .
reznik , eur .
j. a 7 , 543 ( 2000 ) .
kaptari , b. kmpfer , eur .
phys . j. a 31 , 233 ( 2007 ) .
kaptari , b. kmpfer , eur .
j. a 23 , 291 ( 2005 ) .
s.abd el - salam et al .
( cosy - tof collaboration ) , phys .
b 522 , 16 ( 2001 ) .
f. hibou et al .
( disto collaboration ) , phys .
83 , 492 ( 1999 ) .
f. balestra et al .
( disto collaboration ) , phys .
c 63 , 024004 ( 2001 ) ; phys .
b 468 , 7 ( 1999 ) .
f. balestra et al.(disto collaboration ) , phys .
81 , 4572 ( 1998 ) .
s. barsov et al .
( anke collaboration ) , eur .
phys . j. a 31 , 95 ( 2007 ) .
m. hartmann et al .
( anke collaboration ) , phys .
96 , 242301 ( 2006 ) .
kaptari , b. kmpfer , j. phys .
g 30 , 1115 ( 2004 ) ; eur .
j. a 14 , 211 ( 2002 ) .
kaptari , b. kmpfer , nucl .
phys . a 764 , 338 ( 2006 ) .
i. frhlich et al .
( hades collaboration ) , nucl - ex/0610048 .
k. nakayama , y. oh , h. haberzettl , arxiv:0803.3169 [ hep - ph ] . | the production of @xmath0 and @xmath0 mesons in nucleon - nucleon collisions near thresholds is considered within a one - boson exchange model .
we show the feasibility of an experimental access to transition formfactors . |
the process of formation of early - type galaxies is still largely unknown . under study since @xcite , the choice between formation of elliptical galaxies by a single brief collapse or via merging is still disputed in the astronomical literature .
some properties of early - type galaxies are very uniform , suggesting a homogeneous and probably ancient origin .
elliptical and s0 galaxies have smooth light profiles and very low gas and dust content , like overgrown globular clusters . as a class
, they display scaling relations such as a tight fundamental plane ( fp : @xcite ) among the variables luminosity , size , and velocity dispersion , a narrow mg-@xmath0 relation between integrated starlight absorption line strength and velocity dispersion @xcite , and an orderly color - magnitude relation .
many authors note that elliptical galaxies nearby and at intermediate redshift have properties consistent with old , passively evolving systems @xcite .
confusingly , there is also a great deal of evidence that favors a messier , ongoing formation process for early - type galaxies .
some galaxy mergers are caught in the act @xcite with the theoretical expectation that the remnant will soon relax to resemble an elliptical , with an exponential light profile .
the tails , ripples , shells , and other morphological aftereffects of merging are seen in many nearby elliptical galaxies @xcite .
measurements of the integrated stellar absorption features compared with stellar population models indicate light - weighted mean ages for the near - nuclear regions of elliptical galaxies that range from ancient through a median of @xmath17 gyr to a few that appear less than 1 gyr old @xcite .
young stellar populations are much brighter than old populations , so that a relatively modest burst of young stars may be able to skew the mean age of an essentially old galaxy to appear much younger than a mass - weighted mean age , but the presence of any youthful subpopulation contradicts the uniform - and - old hypothesis . with this paper
we attempt to bring clarity to the issue . to simplify , we consider only the relation between the central velocity dispersion @xmath0 and the strength of the integrated stellar mg and mgh features around 5100 .
the advantages of this relation are its distance independence and its small scatter @xcite .
one disadvantage is that age and metallicity both have a similar effect on the mg strength @xcite : an older galaxy has a stronger mg feature , but mg strength can also be increased by mg abundance .
it is clear that the origin of the relation itself is primarily one of abundance : larger galaxies have more heavy elements . intrinsic scatter exits in the @xmath5-@xmath6 relation ( @xmath6 refers to an aperture - corrected central velocity dispersion ) .
this scatter is not correlated with the appearance of the galaxies , with the degree of velocity anisotropy , with their environment @xcite , or with deviation of the objects within ( or perpendiculiar ) to the fp .
@xcite intrepreted this as a combination of age and/or metallicity differencies among ellipticals with similar luminosity , and @xcite find that both mean age and metallicity vary , probably in a correlated manner with younger - appearing galaxies tending to be somewhat more metal - rich .
this correlation is also seen with balmer - metal indicators @xcite the residuals of @xmath5 have a gaussian core , but the galaxies with weaker @xmath5 at given @xmath6 form a tail .
that is , the residual distribution is skewed with an excess of mg - weak galaxies @xcite .
this is suggestive of late star formation : a youthful galaxy will have weak mg strength from the presence of bright , hot stars with weak mg strength and will appear as a low outlier .
we test two opposing hypotheses . in both cases ,
the mg-@xmath0 relation is a mass - abundance correlation . in the `` ancient formation '' hypothesis , elliptical galaxies formed at the epoch of galaxy formation at high redshift ( we assume @xmath7=5 ) .
the mg-@xmath0 relation was put in place at that time .
furthermore , the symmetric portion of the scatter was put in place at formation .
( we could be even more extreme and suggest that the asymmetric part of scatter was also imprinted at formation by an asymmetric abundance distribution , but the implication of this is trivial : except for the effect of passive stellar evolution the mg-@xmath0 relation would remain unchanged back to @xmath7 . ) in the `` ancient formation '' hypothesis , mergers , as inconsequential as possible , cause occasional bursts of star formation that drive a few galaxies to weak mg strength and cause the asymmetry of the observed relation . in the `` maximum merger '' hypothesis , the underlying mg-@xmath0 relation is very tight , and all scatter , symmetric or not , is caused by merging events ( convolved with observational error ) . to compare these two views
, we first collect and analyze published mg-@xmath0 data sets of e and s0 galaxies from field , group , and cluster environments .
we use these data to find robust statistics to help us evaluate the expected asymmetrical scatter . in @xmath83
we describe special - purpose models that predict mg-@xmath0 relations given input parameters such as the number of mergers and gas fraction of each merger .
results are discussed in @xmath84 , and a summary of conclusions is given in @xmath85 .
we analyse the mg-@xmath0 relation in 7 different data sets for e and s0 galaxies .
there are three lick / ids system mg indices : mg@xmath9 , mg@xmath10 , mg@xmath11 , plus the variant @xmath12mg@xmath13 index introduced by @xcite , and defined by : @xmath14 @xmath12mg@xmath13 uses the additional signal available in mg@xmath9 to improve the precision of the mg@xmath10 index
. there does not appear to be any substantial mismatch between @xmath12mg@xmath13 and mg@xmath10 definitions in ultimate index behavior , although differences are detectable @xcite .
we summarize the literature data that we collected with a few words and a table of characteristics in table [ tab-1 ] .
data set 1 is a subset of data extracted from @xcite , it consists of 256 galaxies observed at the lick observatory between 1972 and 1984 .
average errors for mg@xmath10 , mgb , and @xmath12mg@xmath13 indices were 0.008 mag , 0.23 , 0.006 mag , respectively . for velocity dispersion the error is less than 10 @xmath15 .
the advantage of these data is that they define the lick system and are very homogeneous .
the disadvantage is that no aperture corrections for distance have been included .
this appears to have no measureable consequence ( see table [ tab-1 ] ) .
data set 2 is part of the enear database from @xcite that does not overlap any other samples .
it is separated into 3 data sets according to whether the galaxies were assigned to field , group or cluster environments .
measurements of @xmath0 are accurate to 5%-13% , mg@xmath10 to 0.005 - 0.011 mag .
data set 3 : @xcite have presented 136 cluster and group elliptical and s0 galaxies in the direction of the large - scale streaming flow attributed to the great attractor .
the typical errors are 0.05 dex for log@xmath0 and 0.017 mag for mg@xmath10 . data set 4 : the 528 galaxies available with the mg@xmath10 index in @xcite are from 56 galaxy clusters . uncertainties on log@xmath0 are 5% and 0.009 mag for mg@xmath10 .
data set 5 : we culled almost 150 galaxies taken from the efar data @xcite .
the whole sample is much larger but we considered only those galaxies that had errors of 7% or less in @xmath0 . errors for mg@xmath10 were 0.015 mag and for mg@xmath11 0.37 .
data set 6 : @xcite ( the 7 samurai ) computed @xmath0 and lines indices for 600 galaxies from all environments .
the scatter of measurements indicates an uncertainty of 10 % and 0.009 mag for @xmath0 and @xmath12mg@xmath13 , respectively .
we use an updated 7 samurai data set kindly provided by d.burstein ( 1999 , private communication ) .
data set 7 : this data set of 650 ellipticals and so galaxies was extracted from @xcite .
it is a literature compilation with considerable overlap with the 7 samurai data set but with different procedures for homogenization .
the data are in @xmath0 ( same error : 10% ) , and in mg@xmath10 ( error : 0.009 mag ) . with the data in hand we then sought suitable statistics to describe the scatter and asymmetry of the residuals from a linear fit to the data .
the linear fit itself was tried several ways .
a line fit via least - squares method was too sensitive to the presence of outliers .
even when we clipped the outliers in a rejection loop we judged that the resulting line was too unstable .
finally , we binned the data in log@xmath0 , found the median mg index value of each bin , weighted each datum by @xmath16 , and performed a least - squares fit on the array of medians .
this method is outlier - insensitive and very stable .
unless specified otherwise in this paper , the fit was computed only for @xmath17 km / s ; medium - sized and large galaxies . in table
[ tab-1 ] , we report the number of galaxies really used for our fit .
there is a small amount of systematic drift among the data sets .
the table [ tab-1 ] column marked `` mg 300 '' is the value of the mg index fit for @xmath18 km / s .
this quantity is 0.327 @xmath19 0.039 for mg@xmath10 , 5.1 @xmath19 0.085 for mg@xmath11 , and 0.333 @xmath19 0.007 for @xmath12mg@xmath13 , averaged over the data sets . for illustration , we plot three data sets and the computed median line in figure [ fig1 ] . once the line was fit , we sought statistics to characterize the distribution of mg index residuals in terms of both width and asymmetry . for width , the standard deviation has the advantage that everyone is familiar with it , but it is quite sensitive to outliers due to its @xmath20 dependence .
we prefer the average deviation , or adev ( mean absolute deviation ) .
it is an estimator of width less sensitive to outliers than the standard deviation .
it is defined by : @xmath21 but in this case we substitute the median - fit line value @xmath22 for @xmath23 .
for measuring asymmetric tails we need another statistic .
we considered the skewness , a dimensionless number which describes only the shape of the distribution .
we take the usual definition : @xmath24 where @xmath0 is the distribution s standard deviation .
a positive value of skewness signifies a distribution with an asymmetric tail extending out towards more positive @xmath25 .
a negative value signifies a distribution whose tail extends out towards more negative @xmath25 .
we expect a negative skewness if galaxies scatter preferentially toward weaker mg index values .
because we wish to be outlier - insensitive , we also define an `` average deviation ratio '' ( adr ) of the adevs above and below the fitted line , that avoids the @xmath26 dependence of traditional skewness .
since we expect more weak - lined galaxies than strong - lined , we define the adr so that it will be greater than one for asymmetries toward weak mg index values .
@xmath27 the results are summarised in table [ tab-1 ] .
the value of the standard deviation is relatively stable at 0.027 @xmath19 0.003 for mg@xmath10 , 0.474 @xmath19 0.052 for mg@xmath11 and 0.035 @xmath19 0.003 for @xmath12mg@xmath13 among the different data sets .
the skewness is definitely negative , between -0.052 and -1.75 for mg@xmath10 , around -0.7 for mg@xmath11 and -4 . for @xmath12mg@xmath13 , but shows too much scatter to be very useful .
the adev and adr are more stable due to their relative insensitivity to outliers ( adev : 0.0206@xmath19 0.0014 , 0.346 @xmath19 0.032 , 0.02 @xmath19 0.001 and adr : 1.265 @xmath19 0.058 ,
1.105 @xmath19 .12 , 1.7 @xmath19 0.09 for mg@xmath10 , mg@xmath11 , @xmath12mg@xmath13 respectively ) .
so for the rest of our study we prefer focusing on these two statistics .
@xcite noted that mg residuals could be described by a gaussian core with extra galaxies at weak mg strength .
the width of this gaussian core is also of crucial interest for the `` ancient formation '' hypothesis because it represents the amount of scatter imprinted on the mg-@xmath0 relation at formation . in order to find its width from the various data sets
, we compute the number of galaxies within bins of 0.01 mag in both mg@xmath10 and @xmath12mg@xmath13 residual , and within bins of 0.1 in mg@xmath11 .
we fit with a gaussian the principal peak from mg residual distributions , weighting by @xmath28 in each bin ( see fig [ fig2-a ] , fig [ fig2-b ] and fig [ fig2-c ] ) .
these fits yielded widths that were very close for all the data sets except for the efar and @xcite samples .
but these are also the samples with the largest observational errors , and we conclude that this is why the fitted gaussians are more broad . on average , we found @xmath29 equal to 0.0147 @xmath19 0.002 for mg@xmath10 , and 0.013 @xmath19 0.002 for @xmath12mg@xmath13 . because of the difference of units , @xmath29 for the mg@xmath11 index is broader : @xmath30 .
all data sets give remarkably similar characteristics .
we see no statistically significant difference in width or asymmetry among e and s0 galaxies in different environments ( field , group or cluster ) .
also note that our method of line - fitting yields mg@xmath31 field - group - cluster offsets even smaller than those computed by @xcite from the same data . for the remainder of the paper ,
we prefer the adev and adr statistics for providing the first - order discriminant for the synthetic mg-@xmath0 diagrams , whose generation discuss in the following section .
in order to understand the effects of galaxy merging on the scatter in the mg-@xmath0 relation we constructed hybrid models that contain simple galaxy merging trees and rudimentary chemical evolution but a relatively rigorous treatment of the effect of stellar ages on mg feature strength . the aim of the resultant monte carlo simulation is to make a synthetic mg-@xmath0 diagram where the stretch along the @xmath0 axis is heuristic and approximate but where the scatter in the mg strength direction is modeled well .
also , _ relative _ zeropoint changes in mg strength are reliable ; for example , the weakening of mg feature strength with increasing lookback time is well modeled ( if one can assume that ellipticals today were also ellipticals in the past ) .
there are several simplifying assumptions in the models .
chemical enrichment proceeds in a schematic way . a mass-[mg / h ] relation is assumed , and a pre - merger galaxy fragment is assumed to have the [ mg / h ] appropriate for its mass . upon merging , newly incorporated gas is assumed to become stellar with an [ mg / h ] derived from the post - merger mass .
only the gas enriches in this way ; the stellar portion of the merger fragments retain their original abundance . using this scheme ,
the mg abundance is driven toward higher values in larger galaxies , as is observed .
this is one of the more heuristic elements of the models , the chief intent being to produce a spread of galaxy mg strengths .
almost any mechanism that accomplishes this is sufficient for the purposes of this exercise since it is the spread of mg strength from stellar age effects that is under study , not the origin of the relationship itself .
the choice of [ mg / h ] rather than [ fe / h ] is mandated by observation . at high metallicities ,
elliptical galaxies are seen to have approximately constant fe - feature strength as a function of velocity dispersion @xcite .
mg - feature strengths increase with velocity dispersion , and the tight relation between the two variables is the cause of this investigation .
scaled - solar stellar population models predict that both species should increase together with increasing metallicity ( e.g. @xcite ) , so it is mostly mg ( and presumably other light metals ) that drives the increase of abundance in the largest galaxies .
since the deep mg feature at 5100 is the easily measured , [ mg / h ] becomes our abundance parameter .
we choose [ mg / h]@xmath40.6 for the most extreme @xmath0=390 km / s , @xmath32=10@xmath33 m@xmath34 giant ellipticals , and [ mg / h]@xmath40 for @xmath0=150 km / s , @xmath32=5.7@xmath3510@xmath36 m@xmath34 smaller elliptical galaxies .
the relation between velocity dispersion and mass is a projection of the fundamental plane .
specifically , a relation in kappa - space from @xcite was chosen to pass from compact ellipticals through giant ellipticals , namely @xmath37 .
this translates to a rough relation between galaxy mass and velocity dispersion : @xmath38 .
this relation is used in the simulations to translate mass ( the fundamental variable in the simulations ) to velocity dispersion .
the artificiality of this relation does not affect the conclusions of this paper , and its simplicity allows us to explore the stellar population effects caused by complex merger histories .
the part of the models that is not approximate is the treatment of the stellar populations .
quick interpolation tables were summarized from @xcite models using both the original yale / vandenberg evolution and the @xcite isochrones for younger ages .
lookups can be done for @xmath39 luminosity and the three mg indices .
the simulation delivers a galaxy formed at several ( or many ) epochs .
these are collected as a list of ages , masses , and [ mg / h ] abundances .
index strengths are computed as a function of age and [ mg / h ] and weighted by @xmath39 luminosity computed as a function of age , [ mg / h ] , and mass for each tagged star parcel within the final galaxy assemblage .
the main parameters in the models are four . the number of equal - mass merger fragments that will eventually combine to form the final galaxy , the fraction of the final galaxy that formed at the redshift of formation @xmath40 , the gas fraction of the mergers @xmath41 , and a probablility envelope that is either flat with time between @xmath7 and @xmath42 or is proportional to @xmath43 , where @xmath44 is a free parameter .
other parameters control cosmology , set @xmath7 , set the number of galaxies to be generated and their mass range , introduce artificial scatter into the computed mg-@xmath0 relation , and provide for extra ( or not as much ) mg enrichment at each star formation episode . throughout this paper
we adopt @xmath45 , @xmath46 , @xmath47 , @xmath48 . with these parameters ,
the universe is 13.8 gyr old , and @xmath49 corresponds to a lookback time of 12.3 gyr .
we flagged galaxies that experienced star formation within the last 50 myr as `` star forming '' and they are omitted from all statistical calculations .
as the majority of data sets feature the mg@xmath10 index , we use this index for all of the following discussion even though the models predict all index varieties .
as an introduction to the model output , we focus on examples of the effect of our four main parameters . figure [ fig3-a ] shows the effect of varying the number of equal - mass mergers that eventually form the galaxy . with a lot of mergers ( 80 mergers , for example ) ,
the asymmetry becomes low since all the galaxies have almost the same history .
that is , they all suffer a peppering of small amounts of star formation over the age of the universe . but with very few mergers ( 5 mergers , for example ) the dispersion is larger and the asymmetry is much more pronounced .
each galaxy experiences a distinct timing fingerprint for mergers and thus are more individualistic , some having had recent star formation , others having had none .
the shape of the probability envelope has a strong impact as well .
the frequency of mergers is often parameterized in the literature as being proportional to @xmath43 , where @xmath44 is a free parameter .
figure [ fig4 ] translates some of these choices from redshift to time .
one can see that @xmath50 roughly corresponds to a uniform probability with time ( note that when we say `` uniform probability '' throughout this paper , we mean a perfectly flat [ with time ] curve , avoiding altogether the @xmath43 formalism . ) ; @xmath51 corresponds to a present - day merger frequency about half of that at @xmath40 , and @xmath52 corresponds to merger rates strongly skewed to high redshift .
these numbers change somewhat with cosmology .
as fig [ fig3-b ] shows , it is plain that with @xmath53 the obtained distribution is very thin .
since mergers are concentrated in the past , few are lucky enough to have had recent star formation .
the relation thickens with decreasing @xmath44 .
the parameter @xmath54 is the fraction of gas used in each merger event . in figure [ fig3-c ]
we see that more gas leads to more star formation and hence more young galaxies .
less gas leads to weak starbursts and a narrow mg-@xmath0 relation .
the fraction of galaxy formed at @xmath40 , @xmath55 , has less of a dramatic effect ( see fig [ fig3-d ] ) except for the trivial case of @xmath56 .
there is little difference between forming half of the galaxy at @xmath7 and forming the galaxy entirely by mergers , rather to the initial surprise of the authors .
apparently , it is the last few gyr of history that are the most important .
we now search quantitatively for solutions under the maximum merger hypothesis . for fair statistical comparison
, the artificial data were further randomized by an artificial gaussian scatter @xmath57 mag , representing the observational error only .
that is , @xmath58 , the width of the gaussian core of the residual distribution .
this assumption is modified for the `` ancient formation '' hypothesis described below .
because we compare to the mean of all of the observational data , we analysed data from a similar number of artificial galaxies : roughly 2000 monte carlo galaxies .
the scatter from limited sample size of course decreases with increased number of samples : for 250 galaxies the errors were .0009 and .2 for respectively the adev and the adr ( computed from multiple realizations of the same simulation ) . with 2000 galaxies we obtained .0002 for the adev and .04 for the adr .
these errors are not too different from those seen in the observational data sets , so it is all but certain that sampling error causes most of the statistical deviations between data sets . this leads to important guidelines for future work at higher redshift : large sample size is crucial if the scatter is to be adequately characterized , and errors may not creep too much higher than their nearby brethren . figures [ fig5-a ] through [ fig5-c ] show our main diagnostic statistics , adev and adr , for models that cover the grid of input parameters .
data for a uniform probability are shown in figure [ fig5-a ] , for a @xmath59 envelope in figure [ fig5-b ] , and for a @xmath60 envelope in figure [ fig5-c ] .
different @xmath55 and @xmath54 possibilities are plotted as various symbol types as a function of the number of mergers .
the trend with increasing number of mergers is clear : a tighter , more symmetrical relation is found . if this seems counterintuitive because with many mergers one should expect a more volatile and bursty collection of galaxies , consider instead that each galaxy has a more similar history if the number of mergers grows large .
see also figure [ fig3-a ] . we plot a good sampling of the possibilities , representing f@xmath61 or @xmath62 and f@xmath63 with different symbol types . to compare directly with the observational data , we plot the mean observational values found in @xmath64 with @xmath65 and @xmath66 errors .
other trends are also apparent . at a given f@xmath67 ,
the adev increases with the gas fraction since with more gas available more stars will form , making the mean stellar population more volatile .
similarly , at fixed gas fraction , the adev decreases with higher @xmath55 .
indeed , if @xmath55 is set to one , the relation becomes razor thin since the whole galaxy formed at @xmath40 . except for values approaching one ,
the authors were surprised at the ineffectiveness of @xmath55 to modify the mg-@xmath0 relation by much ; it is the weakest of the four parameters . as for the influence of
when in time the mergers are likely to happen , we compare the three figures .
it is rather difficult to find many points that land simultaneously in the 2.5-@xmath0 regions of both adev and adr statistics , but to the extent that some do , most of them are in the `` uniform probability '' figure , around 40 mergers . any probability exponent @xmath44 greater than zero yields mg-@xmath0 relations that are even thinner than the @xmath68 case , and we are forced to conclude that elliptical galaxy formation that was concentrated to the first half of a hubble time is ruled out under the maximum merger hypothesis .
we now relax our assumption that the mg-@xmath0 relation is inherently thin .
we suppose instead that there is a intrinsic , symmetric scatter @xmath29 equal to that which is observed today , but imprinted at formation . the mean width derived in @xmath82 is @xmath69 mag .
this is rather narrow and not too different from the observational scatter , 0.010 mag .
in fact , consideration of these numbers implies that the true galaxy - intrinsic gaussian - random scatter is around 0.01 mag
. it will be interesting to see if future high - precision data upholds this conclusion . under the ancient formation hypothesis ,
only the asymmetric tail of the distribution is produced by late star formation .
do these models fit better ?
we plot the three different probability cases exactly as in the previous section : uniform probability ( fig [ fig6-a ] ) , @xmath59 envelope ( fig [ fig6-b ] ) and @xmath60 envelope ( fig [ fig6-c ] ) .
we derived the statistics using 2000 galaxies , and show the possibilities allowed by @xmath70 and @xmath71 or @xmath62 .
we find that most of the remarks from last section are still valid .
the cases with 5 or 10 mergers do not fit , and the most numerous solutions still occur for uniform probability and @xmath72 .
the main difference is that the additional artificial scatter calms the asymmetry somewhat , and this allows a few more successful models .
we even find a couple of models in the @xmath68 panel that land within the error bars for both statistics , although mostly we find the width of the mg-@xmath0 relation too narrow .
the astounding similarity between maximum - merger and minimum - merger cases is driven by the high quality and consistency of the observational data .
with the low and high bounds of allowed artificial scatter set at 0.010 and 0.015 mag , tight reigns are held on our freedom to interpret the results . to this point in the paper
we have examined only primary statistics adev and adr for the width and asymmetry of the mg-@xmath0 residuals .
we can explore the situation a little more by also considering the absolute mg strength of the mg-@xmath0 relation , parameterized by mg@xmath31 , the median mg strength at @xmath73 300 km / s . including mg@xmath31 along with adr and adev
, we define a goodness parameter to evaluate the relative merit of the various models : @xmath74 we took @xmath75 , @xmath76 and @xmath77 , for the different scatters we had : @xmath78 , @xmath79 and @xmath80 .
this last value is less well - motivated than the others since we expect considerable uncertainty in model mg zeropoint so that can not adopt the observed uncertainty for @xmath81 . on the other hand
, the overall mg level is an important diagnostic and too - weak overall mg is clearly to be discouraged , so our value of 0.04 is an approximate compromise .
we compute this @xmath82 for the usual 3 cases : uniform merger probability envelope , ( 1+z)@xmath83 and ( 1+z)@xmath84 probability , and for both f@xmath85 or @xmath62 . for each case , we make plots from 5 to 100 mergers , and from f@xmath86 to f@xmath87 , for 40 values using a linear distribution for the fraction of gas and a logarithmic distribution for the number of mergers .
we plot isovalues in 1/@xmath82 so that the best models show as maxima rather than minima .
fig [ fig7 ] shows the results for uniform , ( 1+z)@xmath83 and ( 1+z)@xmath84 probability envelopes . in this plot
there are elongated regions of high @xmath82 values .
they tend to slant , indicating significant degeneracy between gas fraction and number of mergers in determining a successful model .
the amplitude of the band decreases with the @xmath44 of the envelope probability .
we see that the valid possibilities shift to slightly higher values of gas fraction when f@xmath88 .
the uniform envelope probability appears to be the best fit .
we now check to see if the possibly successful models indicated by adev , adr , and mg@xmath31 in figure [ fig7 ] are compatible with the abundance distribution and the v - band light - weighted age observed in nearby galaxies .
the abundance distribution is observed to be narrow in all galaxies the size of m32 and larger @xcite , with a fwhm of less than 0.4 dex @xcite in m32 itself , and about the same in the milky way ( e.g. @xcite ) .
we can check the abundance distributions of the simulations as well .
the chemical evolution of these models is primitive , but if the abundance distributions are wider than about 0.5 dex fwhm then we must regard them with suspicion .
also , if the light - weighted mean age ( the closest our present models can come to approximating a spectroscopic balmer - metal feature mean age ) falls below about 7 gyr ( the approximate median of such studies , e.g. @xcite ) we likewise regard this model with much suspicion .
older ages are allowed because the age zeropoint of the balmer - metal feature technique is still quite uncertain .
what we find is emphatic . for @xmath68 or @xmath89 models ,
no model passes the adr / adev test very well , but they also tend to have an abundance scatter greater than 0.5 dex . for uniform probability ,
the mean galaxy age begins to becomes too young for a combination of high gas fraction and large numbers of mergers , but this never became a critical problem because the adr / adev test excludes these models as well .
the best surviving model was with uniform probability , @xmath90 , @xmath91 , and 50 mergers , with an abundance histogram at the 0.5 dex width limit , mean age of 8.3 gyr , and adr and adev within the error bars . at large number of mergers ,
the @xmath90 models tend to have a bimodal abundance distribution caused by the initial metallicity of the galaxy , to which is added additional material from the gaseous portion of the merger , plus a low - metallicity peak from the stellar component of the relatively low - mass merger fragments .
in fact , all of the @xmath90 high points in fig .
[ fig7 ] are at the borderline of having a too - wide abundance distribution , and we should penalize these models a couple of contour levels for this near - transgression .
the @xmath92 models do not suffer from this : they all have suitably narrow abundance distributions .
we claim roughly equal success for both versions of the uniform - probability models , but resounding failure for models in which galaxy mergers happen preferentially in the past .
being left with only one fully successful model after trying so many hundreds left the authors scratching their heads .
we decided to explore other options .
our first option was to abandon the constant merger mass assumption by adopting a cold dark matter ( cdm ) power spectrum for galaxy merger fragments .
we use the formulae from @xcite and @xcite which are based on @xcite formalism .
cdm clusters hierarchically with preferentially small masses in the early universe , building toward larger structures later on .
we built a lookup table with 40 steps in redshift and 200 steps in mass in order to quickly invert the probability function for halos of a given mass and redshift . within each step
a small randomization assigned the final redshift and mass to the merger fragment . to calibrate the zero point between the circular velocity ( v@xmath93 , which the white & frenk formulae give ) and mass
, we normalized to the milky way .
we took v@xmath94 220 km / s , mass equal to 10@xmath95 m@xmath96 , and white & frenk formulae to obtain m@xmath97 10@xmath98 .
the only parameters in the cdm merger models were gas fraction and the cdm `` bias '' parameter @xmath11 .
the adr statistics from the cdm models were very high unless the gas fraction was set below 10% , at which point the adev becomes too narrow .
the @xmath39-weighted age and abundance scatter for the @xmath99 model is acceptable : 11.4 gyr and 0.4 dex , respectively .
increasing the bias parameter helps a little , but no cdm models were found that matched all statistics simultaneously .
the cdm simulations were also interesting because they predict variations in mg-@xmath0 as a function of galaxy size . for small galaxies compared to large
, cdm predicts similar adev , younger mean ages , and much higher asymmetry .
spectroscopic observations @xcite indicate a younger mean age for smaller galaxies .
our collection of mg-@xmath0 observations indicate higher adev but adr roughly the same . a variant on the cdm models was to throw out the cdm power spectrum and adopt a constant power spectrum ( @xmath100 ) and return to a merger probability uniform with time .
these models have suitable adr and adev if the gas fraction is set at 2% - 3% .
the 2% model has a ( @xmath39-weighted ) mean age of 11.7 gyr , with an abundance distribution fwhm scatter of .3 dex .
this is the only model with a mass spectrum ( as opposed to the @xmath55 plus constant @xmath101 models ) that fits all of the observational constraints .
consequently , even though this constant power spectrum model is not physical , the good fit seems to indicate that there will be _ some _ power spectrum that will satisfy the constraints equally well . other variants were tried .
we tried superimposing two different schemes atop one another .
for instance , suppose half of the galaxies were truly ancient and the other half was allowed to have mergers .
these superposition schemes fail dramatically . unless one is exquisitely artful one ends up with mg-@xmath0 diagrams that are clearly double .
finally , we tried schemes of variable gas fraction , where the early universe was assumed to be gas rich , but later mergers are mostly stellar . similarly to the regular 4-parameter models , envelope parameters @xmath102 are ruled out , but uniform - with - time probability gives good matches for @xmath103 . in fact , one such model , with f@xmath104 varying linearly with time from 0.9 at @xmath7 to 0.1 today matches the mg-@xmath0 data best of all models .
this model is also intriguing because it shows strong mg-@xmath0 asymmetry evolution at only modest redshift ( see next section ) .
the recent proliferation of 8 - 12 m telescopes will greatly accelerate the rate at which mg-@xmath0 data will become available for distant galaxies .
the result of studies to date , that the drift of the mg zero point is `` consistent with passive evolution , '' @xcite will eventually be broadened to study the shape of mg residuals as well .
we therefore present some model predictions for high - redshift studies .
there is a major caveat to consider here : model galaxies are assumed to be `` star forming '' for 50 myr after a burst , but `` early type '' at all other times
. if merging is a dominant process , then premerger fragments are often likely to be spiral galaxies . if semianalytic models like @xcite are correct , then , over time , some ellipticals can accrete a gas disk and become spirals .
so the predictions we show in this section are of questionable validity for merger - dominated models ( i.e. all of those that match the @xmath42 data ! ) , but nevertheless provide a useful first cut for the variety of behavior we might expect for mg-@xmath0 relations at significant lookback times .
figure [ fig8 ] shows a collection of near - best models as they would appear if observed at large lookback times .
the merging trees of the model galaxies are computed to redshift zero , but only the partially complete history of each galaxy is considered for nonzero redshift .
so the further back in time we look , the more incomplete the galaxies are . figure [ fig8 ] at a glance shows that mg-@xmath0 relations survive intact even under fairly severe merging scenarios back to when the universe was a quarter of its present age .
the bottom panel shows mg@xmath31 as a function of redshift . within a zeropoint shift ,
all models show mg evolution very similar to the `` ancient , passive evolution '' case shown in large diamond symbols at strong mg strength .
it is easy to shift any given model track by adjusting the assumed metallicity scale , so the overall mg level by itself is not a strong constraint , but the relative time evolution should be fairly well modeled . the most shallow slope , that is , the most constant mg level with time , is given by the @xmath105 , 40-merger , uniform probability , @xmath90 model ( small open diamonds ) .
this is a merger - heavy model .
we think that some researchers will be surprised by the fact that a merger - driven model is redder , relatively speaking , at @xmath106 , than pure passive evolution scenario because one usually thinks of mergers as causing more star formation activity and hence bluer colors and weaker line strengths . in reality
, it can go either way depending on the model .
the steepest mg track is from a very similar model but with the variable @xmath54 option and @xmath92 ( filled bullets ) .
neither steepest nor shallowest models differ by more than 0.05 mag in mg@xmath31 from the passive evolution case if they are normalized to match at one end or the other of the illustrated redshift range .
the tightness and asymmetry of the mg-@xmath0 relation also do not show large changes with lookback time . both the adr and adev can increase or decrease according to the model , but at a very modest level that will be hard to measure without a very large sample of distant galaxies .
in this paper we analyze the mg-@xmath0 relation in elliptical and s0 galaxies by comparing data and models .
we adapted an `` adr '' statistic for the measuring residual asymmetry and we used the adev statistic to measure the width of the scatter . for the combination of all data sets and for 150 @xmath107 320 km
s@xmath108 , adev @xmath109 .0206 @xmath19 .0014 , adr @xmath109 1.265 @xmath19 0.058 and mg@xmath110 0.329 @xmath19 .0039 .
the width of the distribution increases at low @xmath0 . for 50 @xmath111 150 km
s@xmath108 the adev @xmath109 .0282 @xmath19 .0035 .
the errors of the adev and adr statistics are roughly consistent with the random errors one would expect from finite number counting as estimated from running simulations multiple times ( the observed adr scatter is even somewhat lower that what we would expect ) .
we also fit the central peak of the mg residual distribution with a gaussian , thinking that this would be a useful number to know under an ancient formation hypothesis where all of this scatter was imprinted at formation .
we found a width for the central peak of the mg residual @xmath112 .014 @xmath19 .002 mag , for mg@xmath10 while the observational error was @xmath113 .
so if total scatter is a quadrature sum of observational and intrinsic scatter , that is , @xmath114 , we find @xmath115 .
after performing this exercise we realized that there would be little difference for models that began with the entire scatter or just the observational portion of the scatter , and this turned out to be true : the `` ancient formation '' models were not dramatically different from the `` maximum merger '' models .
merger models were constructed to simulate the mg-@xmath0 relation .
the merger fragments that join to form the final galaxy were assumed to be equal in mass .
mergers occur randomly in time ( uniform probability distribution ) or randomly under a probability envelope of the form @xmath43 , where @xmath44 is a free parameter .
the other three main parameters are the number of mergers , the gas fraction of each event , and the fraction of the galaxy that formed in the distant past , at @xmath116 for this paper .
also included was an artificial scatter in mg strength .
this scatter was set equal to the observational error for the `` maximum merger '' hypothesis , and equal to the entire symmetric scatter for the `` ancient formation '' hypothesis , but since the latter was only 40% larger than the former , the two hypotheses generated similar results .
the models predict a simulated mg-@xmath0 diagram , from which we compute mg@xmath31 , adev , and adr in the same way as we did for the observational data .
we also output v - weighted mean age and an abundance distribution since observational constraints also exist on these quantities .
some immediate conclusions are : @xmath117 to the contrary , with many mergers ( 60 - 100 ) a large number of small events leads to small scatter and small asymmetry because merging histories are similar .
most of these models showed _ too _ tight of a relation , and the gas - rich ones also tended toward unacceptably young mean ages .
@xmath117 for the merger probability envelope , a positive @xmath44 in @xmath43 concentrates galaxy formation to the early universe , making the mg-@xmath0 relation thin .
because we find no successful models for @xmath3 , we clash with many published merger rate results such as @xmath118 , @xmath119 , @xmath120 , or @xmath121 @xcite but we agree with a few such as @xmath122 @xcite .
our @xmath123 result is very robust , so we presume that the difference between this result and others stems from differences in treatment ( i.e. redshift range , galaxy types , definition of merger rate , observables ) .
for example , @xcite find a @xmath124 dependence for early type galaxies , but this is for number evolution , a quantity that we do not model because it requires tranformations to other hubble types
. finally , let us briefly discuss the meaning of these results .
our most successful models involved a moderate number of mergers ( 20 - 60 ) with mergers equally probable over cosmic time .
the model with a constant gas fraction that fit the best was with @xmath125 - 0.3 but a model with gas fraction that varies linearly with time from 0.9 at @xmath7 to 0.1 at the present epoch fit even better .
one other model , a variable - mass scenario with a flat power spectrum and merger probability constant with time fit the constraints if the gas fraction was 2% - 3% .
the overall yield of successful models was far below our initial expectations ; we thought many solutions would be possible .
this is good news for understanding galaxy evolution : tight constraints are better than loose ones .
we also find some clarity for the basic question , `` are ellipticals old or young ? '' to arrive at age information , one must look at the most sensitive indicators . a brief perusal of @xcite or other population model paper weighted by the inverse of the observational error orders the list , from most sensitive to least sensitive : balmer - metal diagrams , mg-@xmath0 relation , fundamental plane , and color - magnitude diagram .
thus balmer - metal diagrams show very large age scatter that is harder to detect with other methods .
the present study of the mg-@xmath0 relation finds that the models that fit every observable have an age range of 7 - 10 gyr .
this does not force a revision of the balmer - based age scale , but it keeps alive the long - held suspicion that the balmer ages are all somewhat too young .
the balmer - derived age scatter is fully supported by our study .
\(1 ) balmer - metal and mg-@xmath0 diagrams are compatible , both implying considerable age scatter and a mean age ( 7 - 10 gyr ) within the uncertainty of the balmer method .
the tightness of the mg-@xmath0 relation is compatible with the less - sensitive fp and color - magnitude diagram tightness , and both are compatible with the balmer - metal diagrams . in short ,
we see no barriers to fairly strong merger hypotheses .
\(2 ) in contrast , we find little support for the ancient merger hypothesis : the tamest merger scenarios that we could find are still much wilder than quiescent evolution since @xmath7 .
the least active model ( 2 - 3% gas fraction and flat mass spectrum ) still involves massive mergers ( albeit without much gas ) up to the present time .
on the other hand , this model s mean age was 11 gyr and it was still able to match the width and asymmetry of the mg-@xmath0 relation , so there may be a related model that can satisfy a relaxed version of the ancient merger hypothesis .
we found this avenue of investigation unexpectedly fruitful in the sense that very strong observational constraints allow only very select interpretations to survive .
future theoretical work should involve more choices for mass power spectrum as a first priority , although many aspects of the modeling could be improved .
observational work at intermediate redshift is also needed , although our results regarding high quality and large @xmath126 of such studies make this job more challenging .
bardelli , s. , de grandi , s. , ettori , s. , molendi , s. , zucca , e. & colafrancesco , s. 2002 , , 382 , 17 bender , r. , burstein , d. & faber , s. m. 1992 , , 399 , 462 bender , r. , burstein , d. & faber , s. m. 1993 , , 411 , 153 bender , r. , saglia , r. p. , ziegler , b. , belloni , p. , greggio , l. & hopp , u. 1998 , , 493 , 529 bernardi , m. , renzini , a. , da costa , l.n.,wegner , g. , alonso , v. , pellegrini , p. s. , rit , c. & willmer , c.n.a .
1998 , , 508 , 143 bertelli , g. , bressan , a. , chiosi , c. , fagotto , f. & nasi , e. 1994 , , 106 , 27 burstein , d. , faber , s. m. & dressler , a. 1990 , , 354 , 18 carlberg , r. g. , cohen , j. g. , patton , d. r. , blandford , r. , hogg , d. w. , yee , h. k. c. , morris , s. l. , lin , h. , hall , p. b. , sawicki , m. , wirth , g. d. , cowie , l. l. , hu , e. , songaila , a. 2000 , , 532 , l1 colless , m. , burstein , d. , davies , r. l. , mcmahan , r. k. , saglia , r. p. , & wegner , g. 1999 , , 303 , 813 conselice , c. j. , bershady , m. a. & dickinson , m. 2001 , , 198 , 3417 davies , r. l. , burstein , d. , dressler , a. , faber , s. m. , lynden - bell , d. , terlevich , r. j. & wegner , g. 1987 , , 64 , 581 djorgovsky , s. & davies , m. 1987 , , 313 , 59 dressler , a. , lynden - bell , d. , burstein , d. , davies , r. l. , faber , s.m . ,
terlevitch , r. j. & wegner , g. 1987 , , 313 , 42 dressler , a. , faber , s. m. & burstein d. 1991 , , 368 , 54 ellis , r. s. , smail , i. , dressler , a. , couch , w. c. , oemler , a. , butcher , h. & sharples , r. m. 1997 , , 483 , 582 forbes , d. a. , ponman , t. & brown , r. 1998 , , 508 , l43 forbes , d. a. , beasley , m. a. , brodie , j. p. & kissler - patig , m. 2001 , , 563 , 143 gonzlez , j. j. 1993 , ph .
d. thesis , university of california , santa cruz goudfrooij , p. , mack , j. , kissler - patig , m. , meylan , g. & minniti , d. 2001 , , 322 , 643 governato , f. , gardner , j. p. , stadel , j. , quinn , t. , & lake , g. 1999 , , 117 , 1651 grillmair , c. j. , lauer , t. r. , worthey , g. , faber , s. m. , freedman , w. l. , madore , b. f. , ajhar , e. a. , baum , w. a. , holtzman , j. a. , lynds , c. r. , oneil , e. j. , jr . , stetson , p. b. 1996 , , 112 , 1975 hudson m. j. , lucey j. r. , smith r. j. , schlegel d. j. & davies r. l. 2001 , , 327 , 265 jrgensen , i. , franx , m. & kjgaard , p. 1996
, , 280 , 167 kauffmann , g. , charlot , s. , & white , s. d. m. 1996 , , 283 , 117 kauffmann , g. & white , s. d. m. 1993 , , 261 , 921 kelson , d. d. , van dokkum , p. g. , franx , m. , illingworth , g.d . & fabricant , d. 1997 , , 478 , l13 kodama , t. & arimoto , n. 1997 , , 320 , 41 kodama , t. , arimoto , n. , barger , a. j. & aragn - salamanca , a. 1998 , , 334 , 99 larson , r. b. 1975 , , 173 , 671 markevitch , m. , gonzalez , a. h. , david , l. , vikhlinin , a. , murray , s. , forman , w. , jones , c. & tucker , w. 2002 , , 567 , 27 okamoto , t. , & habe , a. 2000 , , 52 , 670 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 press , w. h. & schechter , p. l. 1974 , , 187 , 425 prugniel , p. & simien , f. 1996 , , 309 , 749 rana , n. c. 1991 , , 29 , 129 schweizer , f. , seitzer , p. , faber , s. m. , burstein , d. , dalle ore , c. m. & gonzlez , j. j. 1990 , , 364 , l33 schweizer , f. & seitzer , p. 1992 , , 104,1039 stanford , s. , eisenhardt , p. & dickinson , m. 1998 , , 492 , 461 terlevich , a. i. & forbes , d. a. 2002 , , 330 , 547 toomre , a. 1977 , in the evolution of galaxies and stellar populations , ed . b. m. tinsley & r. b. larson ( yale univ . observatory : new haven ) , 401 trager , s. c. , worthey , g. , faber , s. m. , burstein , d. & gonzlez , j. j. 1998 , , 116 , 1 van dokkum , p. g. & franx , m. 1996 , , 281 , 985 van dokkum , p. g. , franx , m. , kelson , d. d. & illingworth , g. d. 1998 , , 504 , 17 wegner , g. , colless , m. , saglia , r. p. , mcmahan , r. k. jr , davies , r. l. , burstein , d. & baggley , g. 1999 , , 305 , 259 white , s. d. m. & frenk , c. s. 1991 , , 379 , 52 worthey , g. 1994 , , 95 , 107 worthey , g. , faber , s. m. , gonzlez , j. j. & burstein , d. 1994 , , 94 , 687 worthey , g. , trager , s. c. , & faber , s. m. 1995 , in fresh views of elliptical galaxies , ed .
a. buzzoni , a. renzini , & a. serrano , asp conference series , 86 , 203 worthey , g. , dorman , b. & jones , l. a. 1996 , , 112 , 948 worthey , g. 1997 , in star formation near and far , ed . s. s. holt & l. g. mundy , ( aip press : woodbury , ny ) , 525 ziegler , b. l. & bender , r. 1997 , , 291 , 527 lrrrrrrrr + prugniel , simien 1996 & 644 & 246 & 0.314 & 0.023 & -1.017 & 0.017 & 1.120 & 0.013 + trager et al .
1998 ( lick ) & 256 & 183 & 0.339 & 0.032 & -1.750 & 0.022 & 1.544 & 0.014 + efar 1999&140 & 130 & 0.331 & 0.021 &
-0.324 & 0.017 & 1.210 & 0.013 + dressler et al .
1991&123 & 107 & 0.318 & 0.023 & -0.722 & 0.018 & 1.538 & 0.011 + bernardi et al .
1998 ( all ) & 721&654 & 0.329 & 0.033 & -0.184 & 0.025 & 1.230 & 0.020 + bernardi et al .
1998 ( field ) & 469&366 & 0.331 & 0.033 & -0.186 & 0.025 & 1.151 & 0.020 + bernardi et al .
1998 ( group ) & 104&91 & 0.333 & 0.031 & -0.369 & 0.024 & 1.361 & 0.015 + bernardi et al . 1998
( cluster ) & 104&83 & 0.331 & 0.032 & -0.052 & 0.023 & 1.128&0.013 + hudson 2001 & 528 & 391&0.332 & 0.018 & -0.369 & 0.014 & 1.105 & 0.013 + + trager et al .
1998 ( lick ) & 256 & 183 & 5.091 & 0.526 & -0.971 & 0.378 & 1.220 & 0.2 + efar 1999 & 164 & 144 & 5.260 & 0.422 & -0.573 & 0.314 & 0.990 & 0.3 + + trager
et al . 1998 ( lick ) & 256&183 & 0.338 & 0.032 & -2.472 & 0.021 & 1.788 & 0.011 + 7-samurai & 577&358 & 0.324 & 0.038 & -5.545 & 0.020 & 1.609 & 0.015 + | available data on the magnesium - velocity dispersion ( mg-@xmath0 ) relation for @xmath12000 early - type galaxies is collected and compared .
as noted previously , the mg residuals from a fitted line are roughly gaussian near the mean but have an asymmetric blue tail , probably from subpopulations of relatively young stars .
we define statistics for scatter and asymmetry of scatter in the mg dimension and find impressive uniformity among data sets .
we construct models of galaxy formation built to be as unbiased as possible toward the question of the importance of mergers in the formation of early type galaxies .
the observational constraints ( mg-@xmath0 width , asymmetry , and mean mg strength , plus mean age and width of abundance distribution ) are severe enough to eliminate almost all of models .
among the casualties are : models with merger rates proportional to @xmath2 with @xmath3 , models that assume early formation followed by recent drizzling of new stars , merger - only models where the number of mergers exceeds @xmath480 , merger - only models with less than @xmath420 mergers , and models with a cold dark matter power spectrum even with biasing included .
the most successful models were those with merger probability constant with time , with the number of mergers needed to form the galaxy around 40 .
these models are characterized by mean light - weighted ages of 7 - 10 gyr ( consistent with spectroscopic studies ) , a narrow abundance distribution , and a lookback time behavior nearly indistinguishable from passive evolution of old stellar populations .
galaxies : abundance galaxies : elliptical galaxies : evolution galaxies : formation galaxies : stellar content . |
the skyrme - faddeev model is a @xmath0-dimension modified @xmath4-sigma model @xcite .
this theory has finite - energy string - like solutions and has a potential application as a low energy effective theory of qcd @xcite .
the theory usually has two components , a sigma term which is a quadratic of derivatives and a skyrme term which is quartic in derivatives .
these are the minimum terms required to provide a scale for derricks theorem , @xcite , in @xmath0-dimensional flat space .
the theory is described by the lagrangian @xmath5 where @xmath6 is a three component unit vector , @xmath7 . at fixed time @xmath6
is the map @xmath8 .
a condition for finite energy solutions of is that the field must tend to a constant value at spatial infinity , which we select to be @xmath9 for all time , @xmath10 .
this allows a one - point compactification of the domain , @xmath11 .
so static , finite - energy , solutions are a map @xmath12 this map , , belongs to an equivalence class characterised by the homotopy group @xmath13 .
this shows that there is an integer topological invariant associated with @xmath6 , known as the hopf invariant . in this case
we will refer to the hopf invariant as the topological charge @xmath14 .
+ the topological charge can be found by firstly defining an area form @xmath15 on @xmath16 .
then we can define @xmath17 , which is the pull - back of @xmath15 by @xmath6 . due to the second cohomology group of the three sphere being trivial , @xmath18 ,
all closed forms on @xmath19 are exact forms .
therefore , we can now re - define the exact form @xmath20 by the @xmath21-form @xmath22 as @xmath23 .
therefore we can express the topological charge , @xmath14 , as @xmath24 this non - local definition of the topological charge , , is not particularly useful in this context .
instead the topological charge of can be found as the linking number of loops in the domain .
these loops are formed as preimages of two distinct points on the target space .
for example if we define the two curves @xmath25 and @xmath26 as the preimages of the points @xmath27 and @xmath28 .
if we then choose a smooth surface @xmath29 , with a boundary @xmath25 , the linking is defined as @xmath30 where the @xmath31 refers to the relative orientations of @xmath29 and @xmath26 .
this definition of the topological invariant can be shown to be the same as , @xcite . + there is a well known lower bound on the energy , @xcite , which is based on an involved argument using sobolev - type inequalities , @xmath32 the fractional power of
is proven to be optimal , but the value of @xmath33 might not be .
ward , @xcite , was motivated by his study of the skyrme - faddeev theory on a unit three - sphere to propose that @xmath34 might be a more optimal value .
+ the hopfion location is commonly defined as the curve @xmath35 , which is the antipodal value to the boundary vacuum value .
+ for @xmath2 there have been many extensive and detailed investigations into the static minimum energy solutions of , @xcite . for charges
one to seven it is believed their respective global minimum energy solutions have all been identified .
it is known @xcite for topological charges one to three that the minimum energy solutions have a planar loop location curve .
topological charges five and six have the minimum energy solutions of two linked hopfions @xcite . for topological charge seven
the minimum energy location curve is a trefoil knot @xcite . in @xcite sutcliffe
devised a new knotted rational map ansatz as initial conditions .
these initial conditions were then energetically minimised to give new minimum energy solution candidates for a large class of topological charges .
we shall describe and make use of this technique later .
the actual model of interest here is a modification of the usual skyrme - faddeev model .
it is modified by an additional potential term , so @xmath36 in and @xmath37 this choice of potential , , is not general but is one of the simplest choices
. a potential term , which breaks global @xmath1 symmetry , is also generated on derivation of the theory as an infrared limit of @xmath0-dimensional @xmath38 yang - mills theory @xcite . if we restrict ourself to the plane this model reduces to the old baby skyrmion model .
also this potential , , meets the finite energy criteria for the chosen boundary condition ; where the single vacuum of @xmath39 is also the chosen boundary value .
the potential , , increases the energy density along the location of the hopfion .
this is because the location curve , @xmath40 , corresponds to the maximum of the potential .
therefore , we expect the hopfion location curve to become smaller in length for increasing @xmath41 .
we also expect the hopfion string to become finer with increasing @xmath41 ; as it is analogous with the baby skyrmion model .
this is best understood by an asymptotic analysis where we approximate the field for large @xmath42 as @xmath43 where @xmath44 . for large @xmath42
we know that , due to finite energy criteria , @xmath45 therefore the energy density associated with becomes @xmath46 where @xmath47 is a solution of the partial differential equation @xmath48 separating @xmath47 into radial and angular components as @xmath49 . where @xmath50 is a purely radial function , @xmath51 is a spherical harmonic @xmath52 and @xmath53 are the usual spherical polar coordinates . then solving we gain an asymptotic approximation for @xmath54 , @xmath55 where @xmath56 is a constant of integration .
this shows that hopfions become increasingly exponentially located as @xmath41 increases .
hence the hopfions in this massive theory will have yukawa type asymptotic tails .
it is already well known that charged hopfions can be knotted objects , @xcite .
one of the most effective ways to create non - trivial knotted initial conditions , to be minimised , is to use the rational map ansatz technique described in @xcite . here
the author used a degree one spherically equivariant map to compactifiy @xmath57 by @xmath58 where @xmath59 , @xmath60 and @xmath61 using these complex coordinates an @xmath62-torus knot can be described as the intersection of a complex curve @xmath63 with a unit three - sphere @xcite . hence we can formulate the rational map ansatz @xmath64 the inverse stereographic projection of the curve @xmath65 produces a @xmath66 closed curve in @xmath67 .
the asymptotic value of @xmath68 in the rational map ansatz is used to fix the boundary conditions of @xmath6 .
therefore , we need @xmath69 so the inverse stereographic projection gives @xmath70 at the spacial boundary of @xmath67 .
+ we can now formulate an axially symmetric hopfion initial condition as @xmath71 which has charge @xmath72 .
these types of unlinked hopfion are labelled @xmath73 .
also a hopfion , with a position curve of an @xmath62-torus knot , can be formed by the rational map ansatz @xmath74 where @xmath75 and @xmath76 .
this gives a closed curve that wraps @xmath22 and @xmath77 times about the two circumferences of a torus , @xcite , and has topological charge @xmath78 @xcite .
these tours knot configurations are labelled @xmath79 . a torus knot with a reducible denominator
can produce linked hopfions .
for example a rational map @xmath80 forms two charged @xmath81 hopfions linked once .
we label this type of configuration as @xmath82 . + this
approach not only produces a non - trivial knotted location curve with an analytically known topological charge , it also gives a smooth field with the correct boundary conditions .
+ to find static hopf solitons we set @xmath83 in the skyrme term of .
this gives a non - relativistic theory which has the same static equations of motions as the those derived from .
this greatly simplifies the corresponding equations of motion by removing a numerically cumbersome matrix inversion .
it also still facilitates time evolution by the second order dynamics derived from the sigma term .
the non - relativistic equation of motion can then be numerically evolved on a discrete lattice using a fourth - order derivative approximation .
we also need an additional lagrangian multiplier , @xmath84 , to constrain @xmath6 to take its value on @xmath16 .
if we periodically remove kinetic energy the potential energy will also become minimised ; this will yield minimum energy static solutions .
this minimisation technique produces static solutions and uses much less cpu time when compared to other similar minimisation algorithms . a numerical grid of @xmath85 points , with @xmath86 ,
was found to be large enough to contain the exponentially located hopfion . on this lattice
the hopfion can smoothly attain the vacuum value at the boundary without a noticeable expense of energy .
this lattice is also fine enough not to lose the fundamental topology .
the definition of the position of a hopf soliton is sensible , but not useful for display purposes
. therefore , all the images of hopf solitons in this paper are plots of an isosurface of the preimage of the curve @xmath87 in the domain .
this gives a surface that is a fine tube in the physical space and produces much clearer images . to show the linking number we also need to plot the preimage of a second loop , but there is no unique loop to choose . in all the plots shown we generate general loops on the target space by choosing a point , @xmath88 , on @xmath16 .
we then find the distance on the surface of @xmath16 between @xmath89 and @xmath6 .
this distance is then normalised by the distance between @xmath89 and the south pole of @xmath16 , @xmath90 , as @xmath91 this gives many loops of constant radius on @xmath16 .
the preimages , of these loops , are tubes of varying thickness in the domain . also , an isosurface of unitary value is known to intersect with the position curve of a hopf soliton . throughout this paper
we choose @xmath92 which is an arbitrary choice for aesthetics .
a hopfion initial condition , where the position curve is contained completely on a plane , can be formed by the rational map ansatz .
so if we set @xmath94 in this gives the initial condition of a topological charge one hopf soliton which is located along a planar loop . using this initial condition , and the above minimisation procedure
, we can find the minimum energy configurations for @xmath95 , as shown in figure [ q1-m=0to5 ] .
.minimum energies for @xmath95 topological charge one hopfions .
[ cols="^,^",options="header " , ] table [ tab - energy - inf - m ] shows that in this @xmath96 model the initial and final minimised location curves rarely differ .
a good example of this is the @xmath97 @xmath98 trefoil knot which under minimization remains as a @xmath98 trefoil knot , but in the @xmath99 and @xmath2 models minimises to a @xmath100 .
the @xmath97 , @xmath101 , is also a lower energy solution for this @xmath96 model .
this reluctance to deform from one location curve to another is due to the reduced self interaction of this compact hopfion model .
+ we have also found a topological charge seven trefoil knot , in the @xmath96 model , which is shown in figure [ hopf - q=7-m = inf ] .
for the range of topological charges we have investigated , in this @xmath96 model , we have found that the charge - specific lowest energy solutions are similar to the @xmath2 and @xmath99 models .
but as in the @xmath99 model we have discovered more local minima or saddle point solutions .
again , as shown in figure [ eoverevsq ] , the energy per - unit charge scales with a similar @xmath3 grows as the @xmath2 @xcite and @xmath99 models .
we have explored the skyrme - faddeev model with a potential term included . we found that including a potential in the model makes the hopf solitons exponentially decay to their vacuum value .
increasing the mass makes the hopf solitons string cross - section increasingly exponentially localized .
we found for @xmath99 the minimum energy solutions are described by similar linking curves as the @xmath2 skyrme - faddeev model @xcite .
+ using a spatial rescaling we were able to formulate an infinite mass model .
this infinite mass model is known to yield compact hopfions @xcite . for this infinite mass compact model
we have presented a number of topologically charged solutions .
we showed that the minimum energy compact solutions have similar location curves as in the usual , @xmath2 , skyrme - faddeev model and the @xmath99 massive model . in both the @xmath99 and @xmath96 models we found new local minimum , or stationary point energetic solutions .
the increasing localization of the strings in the two massive models reduces the string self interaction .
therefore , it is not surprising that these models possess solutions stabilised by symmetry .
+ in figure [ eoverevsq ] we showed that the @xmath99 and the @xmath96 models charge - specific energies seem to grow with a similar @xmath3 trend as the @xmath2 case .
therefore , approximating hopfions as kirchhoff rods @xcite could also be successful in the massive model
. + + * acknowledgement . *
i would like to thank professor p.m. sutcliffe and doctor d. harland for their useful discussion .
i would also like to thank the department of mathematical sciences at durham university where a significant amount of the presented research was performed .
a. f. vakulenko and l. v. kapitanski , _ sov .
* 24 * , 433 ( 1979 ) .
ward , _ nonlinearity _ * 12 * , 241 ( 1999 ) .
f. lin and y. yang , _ commun .
phys . _ * 249 * , 273 ( 2004 ) .
a. kundu and y. p. rybakov , _
_ * a15 * , 269 ( 1982 ) . r.a . battye and p.m. sutcliffe ,
letts . _ * 88 * , 22 ( 1998 ) .
ward , _ phys .
_ * b473 * , 291 ( 2000 ) j. gladikowski and m. hellmund , _ phys . rev . _ * d56 * , 5194 ( 1997 ) . r.s . ward , _ j.phys . _ * a39 * , 105 ( 2006 ) .
j. hietarinta and p. salo , _ phys .
rev . _ * d62 * , 081701(r ) ( 2000 ) .
p.m. sutcliffe , _ proc .
soc . _ * a 463 * , 3001 ( 2007 ) . | the skyrme - faddeev model is a @xmath0-dimensional model which has knotted , string - like , soliton solutions . in this paper
we investigate a skyrme - faddeev model with an @xmath1 symmetry breaking potential .
we then rescale this model and take the mass to infinity .
this infinite mass model is found to have compact knotted solutions . in all of the investigated massive models we find similar charged solutions as in the usual , @xmath2 , model .
we also find that their energies follow a similar @xmath3 power growth as the @xmath2 model . |
during the last two decades ground based studies of seyfert galaxies have firmly established that a close relationship exists between radio and line emission . in particular they showed that their narrow line region ( nlr ) is inevitably cospatial with the radio - emission ( e.g. wilson & ulvestad , 1983 ; haniff , wilson & ward , 1988 ) .
furthermore , galaxies harboring a linear radio source have unusually wide line profiles ( whittle 1985 ) and display a correspondence between the location of the high velocity nlr gas with that of the radio ejecta ( e.g. whittle et al .
1988 ; baldwin , wilson & whittle 1987 ; pedlar et al . 1989 ) .
this spatial and kinematical association has been interpreted as due to the compression and acceleration of the interstellar gas induced by the expansion of the radio - ejecta ( wilson & willis 1980 , booler et al .
1982 , pedlar et al .
1989 , taylor et al .
1989 , 1992 ) . hubble space telescope ( hst ) allowed us for the first time to fully resolve the region from which the narrow emission lines originate .
the results of extensive narrow - band imaging of seyfert galaxies clearly showed that the nlr morphology is completely determined by the presence of radio outflows ( e.g. bower et al .
1994 , 1995 , capetti et al .
1995a , 1995b , 1996 , 1997a , 1997b , falcke et al .
1996 , 1998 ) .
in particular , seyfert galaxies with a lobe - like radio morphology have bow shock shaped emission line regions while those with a jet - like radio structure have jet - like emission line structures .
these observations provide further compelling evidence for strong dynamical interactions between the nlr gas and radio - emitting ejecta . in this framework
it is of great interest to study the effects of this interaction on the nlr dynamics at a high spatial resolution .
this prompted a program of long - slit observations using the faint object camera ( foc ) in its spectrographic mode in order to study in detail the velocity field of the nlr of seyfert galaxies .
the first results from observations of ngc 4151 ( winge et al .
1997 , 1998 ) and ngc 1068 ( axon et al .
1998 ) show clear evidence that the highly perturbed and complex velocity field of the nlr of these objects is strongly influenced by the jets propagation . in particular , high velocity gas components are associated univocally with the jet region .
we now present the results of long - slit hst / foc spectroscopy of mrk 3 , a seyfert 2 galaxy which represents one of the clearest examples of close association between radio and line - emission .
radio - images show a linear structure extending over @xmath7 , dominated by two symmetric and highly collimated jets ( kukula et al .
hst emission - line images of mrk 3 revealed that its nlr has a striking `` s '' shaped morphology composed of a series of knots , sheets or filaments and is basically cospatial with the radio - jets ( capetti et al . 1995a , 1996 ) . in this paper
we show that the nlr of mrk 3 is essentially an expanding shell surrounding the radio - jet whose velocity field is determined by the energy input of the outflowing jets .
the detailed spatial and velocity information derived allow us to investigate the energetics of the jet / nlr interaction and to set constraints on the properties and on the evolution of the radio - source associated to mrk 3 .
we defer the discussion on line ratios and the ionization properties of the emitting gas to a forthcoming paper . throughout this paper
we will adopt a recession velocity for mrk3 of 4000 ( z = 0.0133 ) which , for @xmath8 km s@xmath4 mpc@xmath4 , yields a distance of 80 mpc where 1corresponds to 390 pc .
mrk 3 was observed using the foc f/48 long slit spectrograph on december 9@xmath9 , 1996 .
one pixel corresponds to 1.78 and 00287 along the dispersion and slit directions , respectively .
the f305lp filter was used to isolate the first order spectrum which covers the 36505470 region and therefore includes the 3727 , 4861 and 4959 , 5007 emission lines .
an interactive acquisition 1024x512 zoomed image was obtained with the f/48 camera through the f342w filter to accurately locate the brightest line - emission regions .
the slit , 0063x135 in size , was placed at a position angle of -25@xmath10 .
spectra with exposure times of 1000 seconds were taken in the 1024x512 non zoomed mode at 7 locations separated by 04 .
an additional spectrum of 780 s was taken immediately following the interactive acquisition image , before the telescope small angle maneuver ( position 2 in fig .
1 ) . the data reduction follows the procedure described in detail by @xcite . in summary , all frames , including those used for subsequent calibrations , were geometrically corrected by using the equally spaced grid of reseaux marks etched onto the first photocatode in the intensifier tube .
any remaining internal distortion was corrected tracing the spectra of two stars in the core of the globular cluster 47 tuc .
the distortion along the spatial direction was obtained in a similar way , tracing the brightness distribution of the emission lines of the planetary nebula ngc 6543 .
observations of ngc 6543 were also used to obtain the wavelength calibration .
the instrumental broadening is estimated to be @xmath11 20 .
flux calibration was performed using the observations of the spectrophotometric standard star lds749b . to accurately determine the location of the slits we compared the surface brightness profile derived from the hst foc , f/96 , image of capetti et al .
( 1996 ) with that measured from the spectra at the six slit positions which yielded useful data .
the best match is displayed in fig . 1 and is accurate to within half a slit width ( @xmath12 ) . the slit locations , identified as pos1 through pos6 , cut across the emission - line region of mrk 3 in a direction almost perpendicular to that of the radio - axis , p.a .
+ 84@xmath10 ( kukula et al .
velocities were derived by fitting gaussians profiles to the 5007@xmath13 line at each individual pixel using the task longslit in the twodspec figaro package ( wilkins & axon 1992 ) . in several locations the line is composed of more than one component and in these cases multiple gaussian fitting was performed . in regions of low signal - to - noise , up to three pixels along the spatial direction were coadded .
the resulting velocity profiles across the slit at each location are shown in figs . 2 through 8 , where we also plot line widths and intensities .
the origin of the x - axis for each slit location is set at the intersection with the radio - axis .
@xmath14 pos1 : the line profiles are split into two distinct velocity systems in the region of the radio - jet , with a separation of @xmath0 1000 . toward the south , only the redshifted component is visible .
the region of high velocity is extended over @xmath0 05 .
the slit also crosses the bright southeastern blob .
this blob is detached from the main nlr structure and not directly associated with the radio - jet .
the gas in this region has a velocity of @xmath0 3600 and is significantly blueshifted from the velocity on the east side of the nlr which is@xmath0 4300 ( wagner 1987 , metz 1998 ) .
furthermore the lines are quite broad , @xmath0 700 .
@xmath14 pos2 : this slit position was obtained during the interactive acquisition and was located close to pos1 . on the north side of the nlr the velocity is approximately constant around 3900 . along the radio - axis there is a dramatic split of the lines with a separation which increases very rapidly and reaches a maximum of 1700 . unlike in pos1 ,
the velocity field forms nearly a complete velocity ellipsoid .
we are seeing an expanding shell of size @xmath15 03 around the radio - jet .
the line widths of the individual components are always large , 500 , and significantly wider than the instrumental broadening of 320 .
the large increase in line width just north of the split - line region is probably due to spatial confusion of the two velocity systems at the periphery of the shell . on the opposite side of the jet
the slit grazes the bright se blob .
a smooth gradient of 200 can be seen over a distance of 04 and , again , line widths are large @xmath0 700 .
@xmath14 pos3 : the slit is located on the brightest blob of the nlr of mrk 3 , close to its center of symmetry and , likely , to its hidden nucleus .
the velocity field is remarkably flat on the blob but the lines are characteristically broad ( 500 ) .
the emission from this blob is redshifted by about @xmath0 200 with respect to the galaxy systemic velocity . immediately north of this blob , corresponding to where the radio - jet intersects the slit , lines are again split in two components with a velocity separation of @xmath0 1000 .
@xmath14 pos4 : this location is symmetric to pos1 with respect to the nucleus and the velocity field is very similar but with a reversed symmetry .
again it shows a very large line splitting ( 1200 ) along the radio - jet over @xmath0 04 in the form of a broken velocity ellipsoid .
@xmath14 pos5 : the gas kinematics are complex .
three highly perturbed regions correspond to three bright filaments crossed by the slit .
the first , toward the south , is part of the main nlr structure .
the remaining two , which are blueshifted by @xmath0 600 and have large line widths , are associated with the radio - lobe .
these structures are superposed onto a more regular pattern determined by the diffuse emission .
such complex line profiles has been seen previously on the lobe of the powerful radio - seyfert iras 04210 + 0400 ( holloway et al .
1996 ) .
@xmath14 pos6 : slit location 6 grazes the edges of the nlr and of the western radio - lobe in a region of diffuse line emission .
while the velocity field is essentially quiescent , with a dispersion of less than 250 , the line profiles are significantly broad only in the region along the radio - axis , with a fwhm as high as 1000 .
the most dramatic result is that , kinematically , the nlr of mrk 3 takes the form of a broken velocity ellipsoid centered on the radio - jet .
this is the expected behaviour of an expanding cocoon of gas circumscribing the radio - jet and is unambiguous evidence that the nlr is created by the interaction of the radio - jet through the interstellar medium .
note that while there is a misalignment between the nlr and radio axis ( capetti et al .
1995a ) the velocity ellipsoid is centered firmly on the jet .
line widths in the expanding regions are always broad , 500 , suggesting significant turbulence within the cocoon .
the diameter of the emission line shell ( @xmath0 200 pc ) is much larger than the radio - jet which is unresolved in the merlin images ( @xmath1 pc ) .
this is an indication that the radio - plasma does not interact _ directly _ with the line emitting gas .
this is physically expected since for a shock velocity of @xmath16 1000 the post shock temperature will be of @xmath17 k ( taylor et al .
a hot high - pressure cocoon then develops around the advancing radio - jet and mediates the energy exchange between jets and line emitting gas . at these temperatures the main cooling mechanism is free - free radiation rather than line emission . strong support for this interpretation is the direct detection of the ultraviolet counterpart to the mrk 3 jet ( axon , capetti & macchetto 1998 ) .
the excess pressure of the hot gas initially drives the supersonic nlr expansion .
but , as we shall now explain , we believe that the nlr is currently in a momentum driven phase as the bubble has been punctured .
the two key facts leading to this conclusion are the mirror symmetry about the nucleus of the nlr emissivity and fractures in the velocity ellipsoids .
as described above , on the western jet the bubble is open towards the south and the peak of emission is to the north ( pos1 and pos2 ) while the reverse is seen along the east jet ( pos3 through pos5 ) .
this behaviour can be explained if the jets are propagating in the stratified gas of a disk galaxy and the jets are inclined with respect to the gas disk ( see fig .
the dynamical evolution is very similar to that of the wind driven superbubbles associated to ob associations or supernovae in the ism ( tomisaka & ikeuchi 1986 , mac low & mccray 1988 ) . initially the bubble will expand asymmetrically with respect to the jet axis with the side away from the disk expanding more rapidly .
it will still appear as a closed , although asymmetric , velocity ellipsoid .
the bright emission line region corresponds to the side of the bubble which is snow - plowing into the disk . at this stage
the nlr motions are pressure driven .
when the bubble grows to be large enough it is more profoundly influenced by the stratification of the galactic disk . with the current size of 200 pc the cocoon
has probably expanded to several disk scale heights and the hot gas located above the plane of the disk is blowing out into the halo , puncturing the bubble . at this time
the system is effectively momentum driven . aside from the highly perturbed gas on which we have concentrated so far , at all slit locations
there are extended regions in which the gas is relatively quiescent , i.e. with a velocity field characterized by small velocity gradients , small offset from the galaxy systemic velocity and relatively narrow line profiles .
this gas is probably associated with the galaxy disk .
the presence of an extended rotating gas disk in mrk 3 is clearly seen in ground based data ( metz 1998 ) and it is normally observed in seyfert galaxies .
a similar two - components gas structure in the nlr has been recently seen also in ngc 4151 ( winge et al .
radio outflows disrupt only locally the ordered rotation pattern .
the evolution of an overpressured cocoon driven by a supersonic jet has been studied by several authors in the framework of powerful radio - galaxies .
the same approach can be applied to the expanding cocoon in mrk 3 but , in this case , we can directly measure its volume and expansion velocity which turn out to be crucial parameters to constrain the model .
begelman & cioffi ( 1989 ) showed that the evolution of a radio - source is determined by the balance between the ram pressure of the external medium with the thrust of the jet and , where the lateral expansion is concerned , with the cocoon internal pressure , i.e. @xmath18 , where @xmath19 and @xmath20 are the cocoon pressure and the present expansion velocity respectively and @xmath21 is the external density .
the medium in which the jet propagates is likely to be clumpy ; a radio - source evolves within a non - uniform medium in a way similar to that with which it would interact with a smooth medium with the same average density ( de young 1993 ) .
in this situation the relevant value for @xmath21 is the density averaged over the volume of the radio source .
the cocoon internal pressure is maintained by the energy carried by the jet .
for a jet power @xmath22 , averaged over the radio - source lifetime @xmath23 , we have @xmath24 . the cocoon volume , @xmath25 , is given by @xmath26 where @xmath27 and @xmath28 are its diameter and length .
the source age can be estimated noting that in such a cylindrical symmetry the expansion speed decreases as @xmath29 , i.e. @xmath30 . integrating the cocoon expansion from @xmath31 to @xmath32 , i.e. @xmath33 we can now express the jet power as @xmath34 except for the external density @xmath21 , the above parameters can be directly measured from our imaging and spectroscopy data of mrk 3 .
we have @xmath35 pc , @xmath36 pc and @xmath37 from which we get a jet power @xmath38 where @xmath39 is the average density of the unperturbed gas in units of one particle per cubic centimeter , a value which is appropriate to the interstellar medium of our galaxy ( cox & reynolds 1987 ) . as for the radio - source age ,
the value which corresponds to the expansion law we adopted is @xmath40 years . in reality
the cocoon lateral evolution is very complex and depends on the precise distribution of the external medium and mass loading in the shell .
the issue of mass loading of astronomical flows has been discussed extensively in the literature ( e.g. hartquist et al . 1986 ) and it will be important in the evolution of the cocoon around the jet of mrk 3 .
in addition the bubble has now been punctured and it is in a momentum driven phase . a firm upper limit of @xmath41 years on the radio source dynamical time scale can be set by assuming a constant expansion at the present speed . on the other hand ,
a lower bound to the age of the outflow can be set if one assumes that the ionizing photons switch on at the same time as the radio - ejecta .
the size of the extended narrow line region of mrk 3 ( @xmath42 kpc , pogge & de robertis 1993 ) sets it to @xmath43 years .
this would correspond to an expansion speed of less than 0.1 c , similar to what is derived for powerful radio galaxies .
if the nlr dynamics are indeed dominated by the cocoon expansion the jet must have carried enough energy to accelerated the emitting gas .
we can estimate the total kinetic energy of the high velocity gas , operatively defined as the gas within 02 from the radio - jet with a velocity which differ by more than 300 from the systemic velocity .
a given luminosity @xmath44 corresponds to a mass of ionized gas ( osterbrock 1989 ) @xmath45 adopting an average / ratio of 12.6 and an average density of 700 @xmath46 ( metz 1998 ) , an observed flux in the line of @xmath47 @xmath48 corresponds to a mass of @xmath0 130 m. integrating over the region of interest we find a total flux of @xmath49 @xmath48 ( about 20 % of the total nlr emission ) and a kinetic energy of @xmath50 erg .
since only about 1/5 of the jet length is covered by our slits , the total kinetic energy of the gas associated with the radio - jet is @xmath6 erg . on the other hand , during the radio - source expansion , half of the energy pumped by the jet goes into the cocoon s internal energy , while the other half goes into accelerating the external medium .
this strictly applies only if the radiative losses of the cocoon are negligible .
this is indeed the case since the cooling time for a temperature of @xmath51 k is @xmath52 years , where @xmath53 is the gas density , much longer than the radio - source dynamical time .
the total energy deposited by the jet over the radio - source lifetime is @xmath54 erg , remarkably similar to the kinetic energy of the line emitting gas .
this provides _ quantitative _ support to the idea that the gas is accelerated ( indirectly ) by the jet .
radio outflows are known to be commonly associated with seyferts galaxies : in at least 50 % of the cases the radio - emission is extended and often shows a linear structure ( ulvestad & wilson 1989 , kukula et al .
1995 ) with typical sizes of less than @xmath0 1 kpc and they rarely extend beyond 3 kpc . for mrk 3 , we estimated that the radio source size has increased at a minimum speed of @xmath0 3500 ( @xmath55 pc or 8 @xmath56as per year ) .
recently , hst / foc spectroscopy has been obtained also for ngc 1068 and ngc 4151 ( axon et al .
1998 , winge et al .
the lateral expansion of the nlr gas in these two objects occurs at a similar speed to that in mrk 3 .
in particular ngc 1068 shows a well developed jet - cocoon structure whose transverse size is @xmath0 100 pc .
the radio - source sizes of ngc 1068 and ngc 4151 are also comparable to that of mrk 3 ( 900 and 600 pc respectively , wilson & ulvestad 1987 , pedlar et al .
1993 ) and we therefore naturally get similar dynamical timescales to that of mrk 3 .
this suggests that the rate of linear size expansion of the radio - source of mrk 3 is not unusually high , but is representative of seyferts with linear radio - structures .
if this is the case , the lifetime of radio outflows in seyfert galaxies must be typically @xmath57 years , otherwise we should observe seyfert radio sources extending well beyond 1 kpc . as a comparison , the source age for radio - galaxies ranges from @xmath58 and @xmath59 years ( alexander and leahy 1987 , liu et al 1992 , parma et al .
the longer timescales associated with radio - galaxies are also confirmed by the discovery of a line emitting expanding shell surrounding the jet of the 3c 120 which points an age for this source of @xmath60 years ( axon et al .
it appears that the phase of radio - activity in seyferts is relatively short - lived .
two alternative scenarios are viable at this stage .
the high nuclear luminosity and the radio - activity are causally related and thus last for a similar time - scale in which case the whole seyfert phenomenon is short - lived .
the other possibility is that the radio activity in seyferts is recurrent
. we can compare the radio , optical and jet luminosities of mrk3 with those of other classes of active galactic nuclei .
its luminosity is @xmath61 } = 2.1~10^{42 } { \rm erg~ s^{-1}}$ ] ( koski 1978 ) while the radio - luminosity at 4.8 ghz is @xmath62 ( ulvestad & wilson 1984 ; becker , white & edwards , 1991 ) and @xmath63 at 178 mhz ( gower , scott & wills 1967 ) .
the integrated radio spectral index is then 0.66 and from which we derive a radio luminosity for mrk 3 between 10 mhz and 100 ghz of @xmath64 erg s@xmath4 .
the point representing mrk 3 in the radio / optical luminosities plane falls within the region defined by the _ radio - quiet _ ( rq ) quasars sample with @xmath65 studied by miller , rawlings & saunders ( 1991 ) which covers the range @xmath66 in line luminosity .
conversely , the radio luminosity , @xmath67 of mrk 3 is between 2 and 3 orders of magnitude below the average of the _ radio - loud _ ( rl ) agn with similar line luminosities , @xmath68 ( rawlings et al . 1989 ) . the mrk 3 jet power , @xmath22 , ( estimated from the expansion velocity of the emission line region associated with the radio - jet ) can be compared with estimates based on the equipartition parameters ( rawlings & saunders 1991 ) for rl agns .
mrk 3 lies again between 2 and 3 orders of magnitude below the region defined by the objects having similar values of @xmath68 . on the other hand
the ratio between @xmath22 and @xmath67 is similar for mrk 3 and rl agn , indicating that there is a similar fraction of jet power which is dissipated in radio - emission .
more quantitatively , the efficiency at which the jet power is converted in radio emission in mrk 3 is 0.05 @xmath39 .
we speculate that the low radio - luminosity of radio - quiet objects is due to the presence of intrinsically less energetic outflows than in radio - loud objects , rather than to a lower efficiency in producing synchrotron emission .
clearly , observations of a sample of radio - quiet active nuclei with nuclear outflows are required before firm conclusions can be drawn .
our hst / foc long - slit spectroscopic observations clearly show that the kinematics of the gas in the nlr of mrk 3 are determined by the effects of the interaction between the radio - outflow and the ambient gas . in particular , along the radio - jet the line emitting gas shows two velocity systems separated by as much as 1700 which form broken velocity ellipsoids .
this is a clear indication that the gas is expanding away from the radio - jet axis and that the nlr is essentially a cylindrical shell expanding supersonically .
these results are in close agreement with what we found in the nlr of other seyfert galaxies observed with hst .
the physical picture which emerges is that the high velocity shocks induced by the outflowing plasma are heating and compressing the external gas .
close to the radio - jets , the gas is heated to a high temperature , @xmath51 k , and thus forms a hot cocoon surrounding the jets . while its emission ( mostly in the x - ray ) can be relevant for the ionization properties of the nlr , it is negligible for the overall energy budget of the radio - source and thus the cocoon will undergo an essentially adiabatic expansion . during this expansion
, it acts as a piston , accelerating and compressing the surrounding gas which is responsible for the narrow line emission .
the jet kinetic energy is therefore transferred to the line emitting gas after it is dissipated in thermal energy associated with the cocoon . with the current size of 200 pc
the cocoon has expanded to several disk scale heights .
the hot gas located above the plane of the disk is probably escaping into the halo , puncturing the bubble as indicated by the fractures in the velocity field .
we believe that the nlr is currently in a momentum driven phase . aside from the highly perturbed gas associated with the radio - jet ,
a relatively quiescent component is also present , probably associated with the galaxy disk .
the radio - outflow disrupts only locally its ordered rotation pattern . in the framework in which the gas motions are driven by the radio - jets , it is possible to derive several key physical parameters describing the properties of the radio - source associated to mrk 3 . in particular , from the size of the cocoon and its lateral expansion speed we estimated an upper limit to its age ( @xmath69 years ) and a lower limit to the jet kinetic power ( @xmath3 erg s@xmath4 ) .
this provides quantitative support for the proposed nlr model , since the total energy carried by the jet in its lifetime approximately equals the kinetic energy associated with the perturbed high velocity nlr gas , @xmath6 erg .
furthermore , if the advance speed of the mrk 3 radio - source is taken to be representative , it implies that radio - outflows associated with seyferts galaxies are short lived @xmath57 years , since their typical sizes are smaller than a few kpc .
finally , while the jet kinetic luminosity of mrk 3 is between 2 and 3 orders of magnitude smaller than that derived for radio - loud agns with similar emission line luminosity , the fraction of the jet power dissipated in radio - emission is similar to that of rl - agn .
we thus speculate that the main distinction between radio - quiet and radio - loud agn is to be ascribed to a different jet power rather than to a different efficiency in producing synchrotron emission . exploring the effects of the jet propagation within the nlr of active nuclei
is clearly a very promising tool to study the jet physics and to relate the properties of different classes of agn . in particular it will be important to study the dynamical timescales for a statistically significant number of seyfert and radio - galaxies to further test the proposed model .
alexander , p. , leahy , j.p .
, 1987 , mnras , 225 , 1 axon , d.j . ,
capetti , a. , macchetto , f.d . , 1998 , apj submitted axon , d.j . ,
marconi , a. , capetti , a. , macchetto , f.d .
, schreier , e. , a. robinson , a. , 1998 , apjl 496 , 75 baldwin , j. a. , wilson , a. s. & whittle , m. , 1987 , apj 319 84 becker , r.h .
, white , r.l .
, & edwards , a.l . , 1991 apjs 75 , 1 begelman , m .. , cioffi , d.f . , 1989 , apjl 345 , 21 booler , r.v . , pedlar , a. , & davies , r.d .
, 1982 , mnras 199 , 229 bower , g. , wilson , a.s . , mulchaey , j. , miley , g.k . ,
heckman , t.m . , and
krolik , j.h .
, 1994 , aj 107 , 1686 .
bower , g. , wilson , a.s . , morse , j.a . ,
gelderman , r. , whittle , m. , and mulchaey , j. , 1995 , 454 , 106 .
capetti , a. , macchetto , f.d . , sparks , w.b .
, boksenberg , a. , 1995a , apj 448 , 600 capetti , a. , axon , d.j . , kukula , m. , macchetto , f. , pedlar , a. , sparks , w.b . , & boksenberg , a. 1995b , apjl 454 , l85 capetti , a. , axon , d.j .
, macchetto f.d . , sparks , w.b . ,
boksenberg , a. , 1996 , 469 , 554 .
capetti , a. , axon , d.j . , & macchetto , f.d . , 1997a , apj 487 , 560 capetti , a. , macchetto , f.d . ,
& lattanzi , m.g .
, 1997b , apjl 476 , l67 cox , d.p .
, reynolds , r.j . , 1987 , ara&a 25 , 303 de young , d.s . , 1993 ,
apj 402 , 95 dopita , m.a . ,
sutherland , r.s .
, 1995 apj 455 , 468 dopita , m.a . , sutherland , r.s .
, 1996 apjs 102 , 161 falcke , h. , wilson , a.s . , & simpson , c. , 1996 , apjl 470 , l31 falcke , h. , wilson , a.s . , & simpson , c. , 1998 , apj 502 , 199 gower , j.f.r . , scott , p.l . & wills , d. , 1967 , mmras 71 , 49 haniff , c. a. , wilson , a. s. , & ward , m. j. , 1988 , apj 334 , 104 hartquist , t.w . , dyson , j.e . ,
pettini , m. , smith , l.j . , 1986 ,
mnras 221 , 715 holloway , a.j . , steffen , w. , pedlar , a. , axon , d.j . ,
dyson , j.e .
, meaburn , j. , tadhunter , c.n . , 1996 mnras 279 , 171 koski , a.t . , 1978 ,
apj 223 , 56 kukula , m.j . ,
ghosh , t. , pedlar , a. , schilizzi , r.t . , miley , g.k .
, de bruyn , a.g . , saikia , d.j . , 1993 ,
mnras 264 , 893 kukula , m.j .
, pedlar , a. , baum , s.a .
, odea c.p . , 1995 ,
mnras 276 , 1262 liu , r. , pooley , g.g . ,
riley , j.m .
, 1992 , mnras , 257 , 545 macchetto , f. d. , marconi , a. , axon , d. j.,capetti , a. , sparks , w. b. , crane , p. , 1997 , 489 ,
579 mac low , m .-
& mccray , r. , 1988 , apj 324 , 776 miller , p. , rawlings , p. , & saunders , r. , 1993 , mnras 263 , 425 osterbrock , d.e . , 1990 , astrophysics of gaseous nebulae & active galactic nuclei ( mill valley , ca : university science books ) parma , p. , murgia , m. , morganti r. , capetti , a. , de ruiter , h.r . , fanti , r. 1998 , a&a submitted pedlar , a. , unger , s.w . , &
dyson , j.e .
, 1985 , mnras 214 463 pedlar , a. , meaburn , j. , axon , d. j. , unger , s. w. , whittle , d. m. , meurs , e. j. a. , guerrine , n. , & ward , m. j. , 1989 , mnras 238 , 863 pedlar , a. , et al .
1993 mnras 263 , 471 pogge , r.w .
, de robertis , m.m . , 1993 , apj 404 , 563 rawlings , p. , saunders , r. , eales , s.a . ,
& mackay c.d . , 1989 ,
mnras 240 , 701 rawling , s. , sanders , r. , 1991 , nature 349 , 138 steffen , w. , holloway , a.j . ,
pedlar , a. , 1996 , mnras 282 , 130 taylor , d. , dyson , j.e , axon , d.j . , & pedlar , a. , 1989 , mnras 240 , 487 taylor , d. , dyson , j.e , axon , d.j .
, 1992 , mnras 255 , 351 tomisaka , k. , ikeuchi , s. , 1986 , pasj 38 , 697 ulvestad , j.s . &
wilson , a.s .
, 1984 apj 278 544 ulvestad , j.s . &
wilson , a.s .
, 1989 apj 343 659 wagner , s. j. , 1987 , a&a 185 , 77 whittle , m. , 1985 , mnras 216 , 817 whittle , m. , pedlar , a. , meurs , e.j.a . ,
unger , s.w .
, axon , d.j . , & ward , m.j .
1988 , apj , 326 , 125 wilkins , t. w. , axon , d. j. 1992 , in asp conf .
25 , astronomical data analysis software and systems i , ed .
d. m. worrall , c. biemesderfer , & j. barnes ( san francisco : asp ) , 427 wilson , a.s . &
willis , a.g . , 1980 , apj 240 , 429 wilson , a.s . , & ulvestad , j.s . , 1983 , apj 275 , 8 wilson , a.s . , ulvestad , j.s .
, 1987 apj 319 , 105 winge , c. , axon , d.j .
, macchetto , f.d . ,
capetti , a. , 1997 , apjl 487 , 121 winge , c. , axon , d.j . ,
macchetto , f.d . ,
capetti , a. , marconi , a. , 1998 , apj submitted | we have obtained hst foc f/48 long slit optical spectroscopy of the inner 2 of the narrow line region of the seyfert 2 galaxy mrk 3 with a spatial resolution of 006 .
spectra were taken in six locations with the slit approximately perpendicular to the radio - axis . in the region cospatial with the radio - jet , where the brightest emission line knots are located
, the velocity field is highly perturbed and shows two velocity systems separated by as much as 1700 . in several locations the split lines form almost complete velocity ellipsoids implying that we are seeing an expanding shell of gas .
the diameter of this shell ( @xmath0 200 pc ) is much larger than the width of the radio - jet ( @xmath1 pc ) .
we interpret this to be the consequence of the rapid expansion of a cocoon of hot gas , shocked and heated by the radio - emitting outflow , which compresses and accelerates the ambient gas .
the cocoon mediates the energy exchange between jets and line emitting gas .
the gas motions within the nlr of mrk 3 are therefore clearly dominated by the interaction between the jets and the interstellar medium and the nlr itself is essentially a cylindrical shell expanding supersonically . with its current size of 200 pc
the cocoon has expanded to several disk scale heights . due to the external gas density stratification , the hot gas located above the plane of the disk blows out into the halo , puncturing the bubble and fracturing the velocity ellipsoids .
the system is effectively momentum driven . from the size and velocity of the expanding region
, we derive an upper limit to the radio - source age , @xmath2 years , and a lower limit for the jet power , @xmath3 erg s@xmath4 , required to inflate the cocoon and estimate that the jet minimum advance speed is @xmath5 pc per year .
the total kinetic energy of the high velocity gas associated with the radio - jet can be estimated as @xmath6 erg , comparable to the total energy carried by the jet over its lifetime and this quantitatively supports the idea that the nlr gas is accelerated by the jet .
radio - outflows are associated with at least 50 % of seyferts galaxies with typical sizes smaller than a few kpc .
if the advance speed of mrk 3 is representative of the seyfert population then these sources must also be short lived and probably recurrent .
evidence that this is indeed the case is provided by the fact that the expansion time - scale derived for ngc 1068 is comparable to that seen in mrk 3 .
the jet kinetic luminosity of mrk 3 is between 2 and 3 orders of magnitude smaller than that derived for radio - loud agns with similar emission - line luminosity . on the other hand ,
the fraction of jet power dissipated in radio - emission is similar .
we speculate that the main distinction between radio - quiet and radio - loud agn is ascribed to a difference in jet power rather than to a different efficiency in synchrotron emission production . |
galaxies are not spherical objects .
although it is difficult to measure the intrinsic shapes of astronomical bodies from their projected light distributions , inferences can be make from their observed ellipticities .
many studies of the distribution of observed axis ratios of galaxies and galaxy clusters ( ryden , 1992 , 1996 ; tremblay & merritt 1996 ) have shown that they are well represented by a population of intrinsically triaxial objects . similar statistical analysis performed by alam & ryden ( 2002 ) with a much larger sample of galaxies from the sloan digital sky survey early data release
has strengthened the claim that the observed distributions of ellipticities can not be fit by projections of oblate or prolate spheroids .
this type of analysis is only available in the cases where the mass distribution creates light .
the shapes of dark matter halos , therefore , must be inferred from even more indirect means .
proxy measures of the shape of our own halo offer widely varying results .
the tidal stream of the sagittarius dwarf galaxy implies a nearly spherical halo ( ibata 2001 ) , in contrast to the results of star counts , which give a minor to major axis ratio of @xmath5 0.6 ( siegal 2002 ) .
the dark matter halos created in numerical studies of structure formation are non - axisymmetric systems .
@xmath0-body simulations of dissipationless collapse in the cold dark matter ( cdm ) scenario by dubinski & carlberg ( 1991 ) showed that the resulting virialized halos deviated substantially from spherical symmetry and from axisymmetry , with average axis ratios @xmath6 and @xmath7 , where @xmath8 , @xmath9 and @xmath10 are the major , intermediate , and minor axes respectively .
simulations of larger cosmological volumes , which more adequately reproduce the hierarchical nature of structure formation in cdm , have also shown that dark matter halos are triaxial ( barnes & efstathiou 1987 ; warren 1992 ) .
analysis of the currently favored @xmath11cdm cosmology has shown similar results ( jing & suto 2002 ) .
the simulations cited above consist only of a collisionless dark matter component and include no gas dynamics .
the introduction of a gas and stellar disk in the equatorial plane of the halo would certainly have effects on the internal structure of the halo .
fully self - consistent cosmological simulations of dissipational structure formation have problems reproducing more than general galactic attributes ( murali 2002 ) , and the resulting halo shape has not been the focus of these efforts .
however , dubinski ( 1994 ) showed that the adiabatic growth of a disk - like potential in a triaxial halo did not influence @xmath12 , while the intermediate to major axis ratio was only slightly increased to @xmath13 .
models and simulations of disk galaxies , however , usually employ spherical halos , which do not reflect these results from cdm simulations . the purpose of most current investigations of galaxy - scale simulations has been modeling transformations of galaxies through major mergers ( barnes 1992 ; dubinski , mihos , & hernquist 1999 ) , minor encounters ( struck 1997 ; thakar & ryden 1998 ) and gas - dynamical effects ( mihos & hernquist 1996 ; springel 2000 ) .
the triaxiality of the halo does not immediately effect the results of these simulations and the implications of using triaxial halos have not been explored .
early studies of the stability of disk galaxies ( e.g. ostriker & peebles 1973 ; hohl 1976 ; efstathiou 1982 ) demonstrated that a cold , rotationally supported disk is dramatically unstable to bar formation .
this fact was used as justification for the existence of heavy , spherical , dark matter halos enveloping disk galaxies .
this dynamically hot component of the galaxy inhibits bar formation .
it has also been shown that a dense core gives a galaxy an inner lindblad resonance , which does not allow swing amplification of waves through the center of the galaxy and thus acts as a bar suppressant ( sellwood 1989 ) . in all cases
the stabilizing mass component has been assumed to have spherical symmetry .
if these galaxies were enveloped by triaxial halos , they would be more susceptible to bar instabilities .
a systematic exploration of the stability of disks with non - axisymmetric halos has not been conducted .
curir & mazzei ( 1999 ) presented fully self - consistent gas - dynamical simulations of disk galaxies with triaxial halos , but only used two halo models and did not employ bulges .
the resolution in their simulations was severely limited ( @xmath14 ) , far less than the number of particles used by efstathiou ( 1982 ) which sellwood ( 1989 ) found to be insufficient .
the notion that the potential field of a galaxy may contain non - axisymmetric features is not a new one ; it is , in fact , readily apparent from the bars seen in many galactic disks .
the effect of an oval distortion , like a bar , on a gas disk has been proposed as a driving mechanism for spiral density waves ( lin 1970 ) .
this idea has been explored numerically as well ( sanders & huntley 1976 ; sanders 1977 ; huntley 1978 ) .
these early simulations generally were limited to two - dimensional , massless disks driven by an analytic distortion to the symmetric potential , but were successful in creating significant spiral response in the disk . however , bar forcing does not reproduce tightly bound spiral patterns . and in comparison to sanders & huntley ( 1976 ) , the response of the outer disk is weak when a more realistic bar potential , one that drops off faster with radius , is used in the simulation ( sanders & tubbs 1980 )
observationally spiral patterns are as common in barred galaxies as galaxies without bars . also , there is evidence that bars and spiral arms have different pattern speeds , showing that one feature may not drive the other ( sellwood & sparke 1988 ) .
a triaxial halo would not be subject to these concerns since the potential of the dark matter would dominate the outer regions of the disk , especially low mass disks which are stable to bar formation .
the bulge - to - disk ratio of a spiral galaxy , as well as the tightness of the spiral pattern , are two integral criteria of the hubble classification scheme . on the average , there exists a smooth increase in pitch angle with later hubble types ( kennicutt 1981 ) .
such a correlation , however , is little more than a consistency check between classification parameters , and there exists significant scatter in the measured pitch angle for a given hubble type .
even with this scatter , the correlation between arm pattern and galaxy type has prompted several theoretical studies on the origin of the this relation ( see kennicutt & hodge 1982 , and references therein ) .
kennicutt ( 1981 ) found that the correlation between pitch angle and maximum rotational velocity of the disk was as good as the theoretical model predictions , suggesting that the mass distribution and its resulting rotation curve is an important determinant in producing the shape of spiral arms .
a more massive or more concentrated bulge will induce higher rotation velocities and loosely follow the correlation with hubble type as well .
our simulations include a collisionless stellar disk and a stiff , rotating halo .
most simulations include a bulge , which is also stiff .
our use of dissipationless simulations with halos which are not fully self - consistent allows us to cover a significant range of parameter space for all three galaxy components ; the disk , bulge , and halo .
the structure of this paper is as follows : 2 presents our initial conditions as well as the @xmath0-body techniques used in this study .
3 shows the analysis and results of our fiducial simulation .
4 describes our suite of different galaxy models . in 5
we investigate the stability of our models with different halo and bulge masses , as well as different halo rotation rates .
6 presents our results for the spiral morphology of our simulations .
we present results for models which follow the bulge - to - disk ratios of the hubble sequence and models which vary widely in rotation velocity .
we compare our results to observations of spiral galaxies .
the positions and velocities of the disk particles are initialized in the method outlined by hernquist ( 1993 ) .
the density distribution of the disk is described by @xmath15 where @xmath16 and @xmath17 are cylindrical coordinates , @xmath18 and @xmath19 are the scale length and scale height of the disk , and @xmath20 is the total mass of the disk . the circular velocity and velocity dispersions are obtained from @xmath21 @xmath22 @xmath23 @xmath24 where @xmath25 is radial acceleration , @xmath26 is the surface density of the disk , @xmath27 is the epicyclic frequency and @xmath28 is the angular rotation rate .
the radial acceleration in equation ( 2 ) is the total acceleration contributed by the self - gravity of the disk , the dark matter halo , and a spherical bulge .
the normalization of the radial velocity dispersion , @xmath29 , is obtained by requiring @xmath30 at a specified radius @xmath31 to be a multiple of the critical dispersion at that radius : @xmath32 where @xmath33 is toomre s parameter for gravitational stability in a rotationally supported disk . the disk models in this paper
are defined to have @xmath34 at @xmath35 , which in the milky way galaxy would be approximately the solar radius .
the rotational velocity of each particle is then given by @xmath36 the different halo models used in this paper can be represented by dehnen s ( 1993 ) @xmath37-models ; @xmath38 where @xmath37 is the interior logarithmic slope of the density profile and @xmath39 is the scale length of the halo .
a model with @xmath40 corresponds to a model proposed by hernquist ( 1990 ) for bulges and elliptical galaxies .
a spherical hernquist model is used here to represent the bulges of our galaxies .
dubinski & carlberg ( 1991 ) showed that the hernquist model is a good fit for cdm halos , and is close to the general density profile for dark halos found by navarro , frenk , & white ( 1997 ) , which also asymptotes to @xmath41 at small radii .
the @xmath37-models can be generalized to triaxial systems by replacing the radial coordinate @xmath42 with @xmath43 , where @xmath44 where @xmath8 , @xmath9 , and @xmath10 are once again the principal axes of the system . in practice ,
the disk - particle positions are sampled from the density function in equation ( 1 ) , the accelerations on each particle from the different galaxy components are computed , and then the velocity dispersions and the rotational velocity of each particle are calculated .
the peculiar velocities of each particle are sampled from gaussian probability distributions with dispersions given by equations ( 3)-(5 ) . because @xmath27 and @xmath28 in equations ( 5)-(7 ) are assumed to be azimuthally symmetric , the velocities for the disk particles are computed assuming that the halo is spherical ; therefore when the disk models are placed inside a triaxial halo they are initially out of equilibrium . to compute the self - gravity of the disk particles , we have employed ( springel , yoshida , & white 2001 ) , a tree code based on the hierarchical algorithm of barnes & hut ( 1986 )
. employs a spline - softened force calculation that is expanded to quadrupole order for particle - cell interactions .
the timesteps are continuously varying and individual for each particle .
the spatial and time resolutions are set by the smoothing length , @xmath45 , and the timestep accuracy parameter , @xmath46 . for simulations with @xmath47 , which are the majority of the simulations in this paper ,
these parameters were set to @xmath48 and @xmath49 .
the gravitational contribution of the halo is computed by the self - consistent field ( ) method of hernquist & ostriker ( 1992 ) .
this method is based on the expansion of the galaxy potential in an orthonormal set of basis functions .
the accuracy of the potential calculated is driven by the number of terms in the expansion .
the numbers @xmath50 and @xmath51 determine the number of radial and angular basis functions used .
the integration time , however , is a linear function of the number of terms as well .
we found that the best balance of accuracy and speed was achieved with @xmath50=16 and @xmath51=10 , which were the values used in all the simulations .
the zeroth - order basis function was chosen to be the hernquist profile , and the coefficients of the expansion were computed by monte carlo sampling of the halo density distribution , @xmath52 . since the coefficients only have to be calculated once for each halo model ,
accurate calculations can be made by initially sampling a large number of particles from each halo profile , @xmath53 particles .
the method has been used frequently to model triaxial systems ( merritt & quinlan 1998 ; holley - bockelmann 2001 ) . when comparing the force on a set of test particles calculated from the coefficients with the force calculated from direct summation over the particles used for the coefficients , the rms error was @xmath54 .
this error level was maintained in models where @xmath37 was not equal to unity .
the contribution to the potential of the galaxy from the bulge can be computed analytically .
therefore , the total acceleration of each particle is given by @xmath55 the first term is the acceleration returned by from walking the tree of disk particles .
the middle term is the acceleration from the bulge , with the subscript @xmath9 denoting bulge properties .
the third term in equation ( 10 ) is the acceleration computed from the expansion of the halo potential .
the particle timesteps are inversely proportional to the particle acceleration , @xmath56 .
ccc @xmath20 = 1.0 & @xmath57 = 6.0 & @xmath58 = 0.3 + @xmath18 = 1.0 & @xmath31 = 3.63 & @xmath59 = 0.2 + @xmath19 = 0.2 & @xmath37 = 1.0 & @xmath37 = 1.0 + & @xmath60 = 0.85 & + & @xmath61 = 0.70 & + & @xmath62 = 0.06 & +
table 1 lists the physical parameters of the fiducial simulation . scaling our @xmath0-body units to physical values reasonable for the milky way galaxy as in hernquist ( 1993 ) , we get @xmath63 kpc and @xmath64 , making the unit of time 13.1 myr . with these scalings ,
the value of the rotational speed at the solar radius , @xmath65 , is nearly 220 km s@xmath66 .
the fiducial halo rotation rate , = 0.06 , is equivalent to a pattern speed of 4.5 in these unit .
( for comparison , the simulations of bekki & freeman ( 2002 ) , in which a massless gas disk is embedded in a rotating triaxial halo , use a pattern speed of 3.8 .
) figure [ curves.1 ] shows the contributions of each galaxy component to the total circular velocity .
figure [ curves.1 ] also shows the values of the @xmath28 , @xmath67 , and @xmath68 resonances as a function of radius .
although the simulations of dubinski ( 1994 ) suggest that the equatorial plane of the halo can be as elliptical as @xmath69 , we chose a more moderate value of @xmath1 for our fiducial simulation .
the minor axis ratio , @xmath70 , was set to 0.7 .
figure [ potdiff ] plots the fractional difference in the potential along the @xmath71 and @xmath72 axes of halos with different values of @xmath60 .
the dotted line represents the asymmetry of the potential from the entire fiducial model . at five scale lengths ,
the difference in the halo potential is @xmath73 , while the difference in the total system potential is @xmath74 . to test our numerical method , we ran a simulation with a spherical halo but still computed the halo s contribution to the potential through the method . outside of the shape of the halo , which was set to @xmath75 ,
all the parameters listed in table 1 were the same .
the simulation was run to a time of 500 , which corresponds to a physical time of 6.55 gyr .
figure [ sph - stats ] shows the rotation curve , the surface density , and the vertical particle density for the initial and final states of the simulation .
these curves show remarkably little deviation from their initial values , proving that the integration of the into works well for this type of simulation .
the fiducial model was integrated to a time of 6.55 gyr as well .
the influence of the triaxial halo of the density structure and rotation curve can be seen in figure [ fid - stats ] . as in the simulation with the spherically symmetric halo ,
there is little change in the azimuthally averaged structure of the disk .
the maximum of the rotation curve is slightly lower than the initial velocity , caused by a slight heating of the disk as rotational energy is transformed into thermal .
it is interesting to note that the vertical structure of the disk does not change even after this lengthy integration time .
the disk is not vertically heated by the constant perturbation of the halo potential .
structure in the disk is generated by the influence of the halo .
a time - varying , non - axisymmetric potential creates a torque on the disk , and it is expected that symmetric , @xmath43=2 structures should form .
figure [ fid - img ] shows an image of the fiducial model at a time of 655 myr .
the bisymmetric structure is apparent . to quantify the response of the disk to the halo potential
, we performed a fourier decomposition of the surface density of the disk into logarithmic spirals ( sellwood & athanassoula 1986 ; binney & tremaine 1987 ) of the form @xmath76 , where @xmath77 and @xmath78 is the azimuthal angle .
the amplitude @xmath79 of each fourier component is the fourier transform of the surface density @xmath80 ; @xmath81 in practice , since an @xmath0-body simulation is resolved by discrete particles , the integral signs can be replaced by a summation over all particles and equation ( 11 ) then becomes @xmath82 the parameter @xmath43 was set to 2 to look for bisymmetric structure , and coefficients were calculated for values of @xmath83 .
the decomposition of the fiducial model at 655 myr is shown in figure [ fid - four ] .
the decomposition of the model with the spherical halo at the same time interval is shown for comparison .
power at negative values of @xmath46 corresponds to leading spiral waves .
power at positive values of @xmath46 corresponds to trailing spiral waves , which in the fiducial model have a distinct peak at @xmath84 .
the pitch angle @xmath85 for this logarithmic spiral can be calculated from @xmath86 .
smaller values of @xmath46 correspond to more loosely - bound spiral patterns , or arms with a larger pitch angle .
the peak seen in figure [ fid - four ] centered near @xmath87 is a linear bar structure .
the presence of a bulge in the fiducial model prevents a significant bar from forming .
the peak of the power at @xmath88 is significantly higher than the power at @xmath89 .
an identical simulation to the fiducial model was performed , with the only change being the removal of the bulge component of the galaxy .
the fourier decomposition for this bulgeless model at @xmath90 myr is also plotted on figure [ fid - four ] . the amplitude of the trailing arms is slightly reduced and is at smaller @xmath46 , but the bar component of the disk has been greatly enhanced . the fourier decomposition for the simulation with a spherical halo is also plotted for comparison .
the spectrum of this model shows no significant peaks and is at low amplitude . because the disk consists solely of collisionless particles
, it is expected that the spiral modes excited by the halo should damp away simply due to relaxation effects .
a galactic disk made entirely of stars could not sustain spiral patterns indefinitely , although since the potential of such a system is much smoother than our simulations the relaxation time would be much longer .
we quantify `` arm strength '' as the sum of the amplitudes from @xmath91 . a strong bar would systematically bias this definition of arm strength , since @xmath87 represents linear structures .
however , the simulations for which this calculation was done did not contain strong bars .
the amplitude of the fourier spectrum at @xmath87 in figure [ fid - four ] is not the dominant peak . above @xmath92 , which corresponds to a pitch angle of @xmath93 ,
the fourier spectrum is usually noise .
figure [ fid - arms ] plots arm strength against time for the fiducial simulation .
it shows a rapid growth of the spiral modes , peaking around 0.5 gyr , then exponentially decaying to an asymptotic value of around 0.5 .
the rms value of the arm strength for the model with the spherical halo is 0.2 .
it is important to note that these simulations all begin with the disk being out of equilibrium with the triaxial halo .
several simulations were performed to test the response of the disk to a adiabatic conversion of the halo potential from spherical symmetry to the full triaxial form .
initially , the model is begun with the halo potential being spherical .
the triaxial potential is phased in by the function @xmath94 @xmath95 while @xmath96 .
this function has zero derivatives at @xmath97 and @xmath98 , so that the change in potential is smooth at the limits .
simulations were run with @xmath99 20 and 40 in @xmath0-body units , corresponding to physical times of 0.26 and 0.52 gyr .
these simulations , plotted in figure [ fid - arms ] , reach their peak arm strength near 2@xmath100 and then proceed to decay along the same curve as the fiducial model .
the peak power is reduced monotonically with @xmath100 , from 2.5 to 1.7 and 1.2 .
all the curves in figure [ fid - arms ] begin to overlap at a time of 0.8 gyr and decay to the same arm strength thereafter .
the structure is very similar in all three simulations at @xmath101 gyr .
even though the peak arm strength is reduced as the triaxial potential is phased in , the final state of the disk is unchanged . to explore the @xmath0-dependence of our simulations , we increased the resolution of our fiducial model by an order of magnitude , realizing the disk with @xmath102 particles and lowering the smoothing length to @xmath103 . in comparison to the @xmath47 simulation ,
the structures in the disk , shown in figure [ big - img ] , are extremely apparent and well - resolved . inside three scale lengths ,
strong @xmath104 spiral arms are seen .
these arms are encircled by two different structures at different radii .
these are caustic - like structures , where the particle orbits from different radii have crossed each other , leading to a local density enhancement .
the outer caustic is highly elliptical and is not continuous with azimuth .
it has an axis ratio of 0.7 and a semimajor axis of 6.8@xmath18 .
the inner caustic , which nearly connects with the spiral pattern inside it , has an axis ratio of 0.9 and a semimajor axis of 3.9@xmath18 . the inner lindblad resonance for this model is shown on top of the image in figure [ big - img ] . located just outside the inner caustic , at 5@xmath18
, it connects the inner and outer caustics at the minor axis of the outer caustic .
figure [ big - arms ] shows the fourier decomposition of the @xmath105 simulation in comparison to the fiducial model .
the peaks in the power of both simulations are at the same locations , with a slightly higher peak at @xmath84 and lower peak at @xmath87 for the higher resolution simulation .
figure [ big - arms ] also plots the arm strength as a function of time .
the peak arm strength of the high resolution simulations is slightly higher than that of the fiducial simulation , and the two curves decay at nearly the same rate .
these results demonstrate that the fiducial resolution is adequate for the analysis performed .
cold dark matter simulations do not predict a single shape to dark matter halos ; rather a distribution of axis ratios and triaxialities . for the halo figure rotation ,
the fiducial value of @xmath106 is somewhat arbitrarily chosen and limits on this parameter are not defined .
we ran simulations to test the effects of varying these parameters , first varying @xmath60 with held constant , then varying with @xmath60 held at 0.85 .
figure [ q - results ] shows the fourier decompositions of five models , including the fiducial model , with different levels of flattening .
the most extreme model has @xmath107 and @xmath108 .
the most moderate model has axis ratios of 0.9 and 0.95 .
( the value of @xmath61 for our models is always @xmath109 . )
all the models show a peak near @xmath84 , showing that the pitch angle of the spiral arms created remains nearly the same in these different models .
as @xmath60 gets closer to unity , the response of the disk falls off ; at @xmath110 , the amplitude of the fourier peak is a factor of 2.3 lower than the fiducial model , and the disk itself shows little visible evidence for spiral structure . as @xmath60 gets smaller , however , response of the disk around @xmath87 increases dramatically . in the most flattened model ,
the amplitude at @xmath111 is greater than that at @xmath84 .
figure [ om - results ] shows how the integrated arm strength and the value of each change with the rotation rate of the halo , . as the halo rotation increases , the response of the disk is monotonically increasing when @xmath112 . as increases to = 0.12 , the value of decreases from 12 to 7
this can be seen in figure [ 4disk - om ] , which shows images of four of these models .
the mass distribution of the milky way , although studied for several decades , is not known with certainty .
the mass distribution of other galaxies are not known with much confidence at all , although it is known that there exists a range of halo and disk masses .
low surface brightness galaxies are thought to be dark - matter dominated , while many researchers espouse the maximal disk hypothesis presented by van albada & sancisi ( 1986 ) for galaxies with high surface brightness to explore these parameters , we constructed several sets of models in which all three components of a disk galaxy disk , bulge , and halo were varied in systematic ways .
first we explored models where the disk mass was reduced and the dynamics became more heavily influenced by the potential field of the halo .
second , we ran models where the parameters of the halo , namely the interior slope @xmath37 and the scale radius @xmath39 , were altered such that the mass distribution of the galaxy became disk - dominated at small radii .
the total mass of the halo is constant , or nearly so , if one considers the truncation radius of the halo to be at a value of 100@xmath18 , the value chosen to calculate the moment of inertia of the halo .
we characterize each of these models by its _ mass ratio _ , @xmath113 , the ratio of the disk to halo mass within the solar radius , 8.5 kpc .
table 2 lists the general properties of the disk - dominated models , denoted d1d5 , where d5 is the most nearly maximal disk with a mass ratio of 4.53 , and the properties of the halo - dominated models , denoted h1h5 , where h1 has the least massive disk with a mass ratio of 0.22 .
figures [ rot - curvesd ] and [ rot - curvesh ] show the circular velocity profiles for the halo- and disk - dominated models respectively . in all models the bulge mass
is set to 0.3@xmath20 .
ccccccc h1 & 0.22 & 0.3 & d1 & 0.91 & 0.8 & 3.63 + h2 & 0.38 & 0.5 & d2 & 1.13 & 0.6 & 3.73 + h3 & 0.53 & 0.7 & d3 & 1.51 & 0.4 & 3.98 + h4 & 0.68 & 0.9 & d4 & 2.27 & 0.2 & 4.48 + h5@xmath114 & 0.76 & 1.0 & d5 & 4.53 & 0.0 & 5.69 + third , we created a set of models in which the disk and halo mass distributions were held constant and the bulge mass and scale length were varied .
the disk mass was @xmath20 = 0.8 and the halo parameters were the same as the fiducial model .
the bulge masses were @xmath58 = 0 , 0.1 , 0.3 , 0.5 , 0.7 , 0.9 , 1.3 , and 2.0 .
the value of each @xmath59 was calculated such that the average density inside one bulge scale radius was the same as that of the fiducial bulge , namely @xmath115 where @xmath116=0.2 and @xmath117 .
our last set of models centered on models that had lower rotation velocities , values closer to 100 than to the fiducial value of 220 .
the motivation behind these models will be discussed in 6 .
the specific parameters used are as follows : the halo parameters are identical to the halo in model d5 , which is the model with the lowest central concentration .
the disk mass was 0.5 .
the bulge mass was varied , having values of 0 , 0.075 and 0.15 .
the results of these simulations will be discussed in the following two sections , focusing on the formation of bars and morphology of the spiral arms induced in the disk by the halo .
bars are simple to detect in our simulations , since they represent the @xmath87 modes of the fourier decompositions . models with strong spiral arms , however , can also return high amplitudes at @xmath87 compared to models with no structure at all . to prevent false detection of bars
, we followed the procedure described in efstathiou ( 1982 ) to quantify the strength of the bar .
the particles are binned in radial rings of linear separation and the amplitude and phase of the @xmath87 , @xmath104 component is calculated for each ring .
if a bar is present , it is defined as existing within the rings that have phases which are coherent to within 10% .
the bar strength is the maximum fourier component within this coherence and is denoted by the symbol @xmath3 .
the rings are 0.1@xmath18 in width and contain several thousand particles each . to be considered a valid detection , the coherence must be at least 5 rings in length but does not have to begin at the center - most ring some simulations with obvious bars do not begin having coherent phases until @xmath118 . from this method
we can determine the amplitude , length , and pattern speed of the bar .
the binning of particles in cylindrical rings eliminates false detection of bars in simulations with strong spiral patterns because these models will have high values of 2max but will not have coherent phases .
figure [ 4bar ] shows simulations at four different levels of bar strength : 2max = 0.1 , 0.2 , 0.4 , and 0.6 .
values of 2max less than or equal to 0.2 have weak bars that are generally @xmath119 disk scale length in size . at 2max = 0.4
, the bar is very distinct and longer . at 2max
= 0.6 , the highest value seen in our simulations , the bar dominates the mass of the disk and can be nearly four scale lengths in size .
when integrated with spherical halos , all of the models presented in this section developed no detectable bars .
figure [ bar.b3 ] shows the time evolution of bar strength in models with different values of halo rotation rate . for the fiducial value of @xmath120 , the disk - dominated models are generally stable .
only model d3 develops a detectable bar , one which asymptotes to 2max@xmath121 .
when is increased , however , bar instabilities become much larger . at @xmath122 , the models with the lowest mass ratios ( d1 and d2 ) are stable , but the the other models show significant bar formation . the growth in the bar modes
is very gradual at first , but at 3 gyr models d3 ,
d4 , and d5 rapidly grow their bars , which peak at values of 2max between 0.5 and 0.6 and slowly decline in amplitude . in the bottom panel of figure [ bar.b3 ] ,
has been increased to 0.09 .
all the models begin to show bar structures at @xmath5 2.5 gyr and become bar - dominated by 3 gyr .
models d3 and d4 make a rapid transformation to being bar - dominated , while 2max for the other models grow more gradually .
regardless of the time it takes for the bar to grow , the values of 2max for all the models is close to 0.4 once the bar has reached its peak strength .
the values of 2ave , the bar strength averaged over the last 0.8 gyr of the simulation , do not show any correlation with mass ratio . for model d1 , 2ave = 0.39 , while for model d5 2ave = 0.40 .
the halo - dominated models , including the fiducial model h5 , are all stable against bar formation , even with a halo rotation rate of 0.09 .
as stated above , the stability of the disks shown in figure [ bar.b3 ] is in part due to the mass of the bulge .
we decreased the value of the bulge mass to @xmath123 and re - ran our disk - dominated models to investigate bar formation in galaxies with less prominent bulges .
figure [ bar.b15 ] shows the results . at = 0.06 ,
models d3d5 develop distinct bars while the other models remain stable . at = 0.075 ,
all the models show some level of bar strength , with the average value of 2max loosely correlating with increasing mass ratio . at = 0.09 ,
all the models show rapid bar growth and become bar - dominated between 2 and 2.5 gyr .
the values of @xmath4 also correlate roughly with mass ratio . for model
d1 @xmath124 , increasing to @xmath125 for model d5 .
in these simulations , the growth of the bar begins earlier but is slower than the models with @xmath2 .
the values of 2ave are slightly higher in the @xmath2 simulations for those models which develop bars .
the bar pattern speeds do not appear to correlate with either or 2max .
the simulations that develop strong bars all have bar pattern speeds near 0.5 radians per unit time ( in the same units as ) . the pattern speeds do not change on average once the bar has reached its peak amplitude . since our simulations employ a static halo potential our results do not reflect the dynamical friction and angular momentum transfer that would surely occur in physical galaxies , reducing the pattern speed ( debattista & sellwood 2000 ; valenzuela & klypin 2002 ) .
although in some simulations the peak in the fourier spectrum is well defined and narrow , in others a simple search for the maximum value of @xmath126 can vary significantly in time if there is more than one peak in the spectrum .
pitch angles in spiral galaxies are known to vary with radius and are measured as an average value ( kennicutt 1981 ) . therefore , to quantify the morphology of the spiral structure in our simulations we calculate an average value of @xmath46 , summed from @xmath127 and weighted by the fourier amplitude at each value of @xmath46 .
the simulations for which this quantity , , were calculated did not contain bars as defined in 5 , and so using @xmath87 as the starting point of the weighted average does not bias the results to smaller values of . only amplitudes greater than half the rms amplitude in that range are actually counted in the summation . the value of varies in time about a mean value at a level of @xmath5 5 - 20% .
fourier spectra with higher overall amplitude , usually found in models with higher values of , have smaller rms fluctuations .
the weighted average of @xmath46 is again averaged over the time range 0.41.3 gyr .
the lower of these limits is chosen to allow the disk to reach equilibrium with the halo .
as the simulations progress the arm strength decays , and so the integrations were stopped at 1.3 gyr . in our notation
, is the weighted average of the fourier spectrum for a given moment in time , and is the average of this value over time range mentioned .
the quantity can then be converted into an average pitch angle , .
the top panel in figure [ pave - t ] shows the evolution of with time for the halo - dominated models h1h5 with = 0.09 .
although there is some variation , the average value with time for each simulation is well - defined .
the bottom panel in figure [ pave - t ] shows four realizations of the low - velocity model with @xmath2 , the last of which was resolved at @xmath128 particles rather than the fiducial value of @xmath129 .
the simulations were sampled with much higher frequency to investigate how varies on small time scales .
the results show little difference between all four realizations , showing that one realization at the fiducial resolution , although not ideal , is adequate for our results .
the clear trend in the top panel of figure[pave - t ] is that decreases with decreasing mass ratio .
figure [ aave ] shows the value of plotted against mass ratio for all the models listed in table 2 .
the results for = 0.09 have smaller error bars than those for = 0.06 because of the higher amplitude of the fourier spectra for = 0.09 . for both sets of simulations ,
the value of is maximum at the fiducial mass ratio of 0.76 and decreases as the mass ratio gets bigger or smaller . as expected from the results of figure [ om - results ] , the data for = 0.09
are shifted to lower values of as compared to the data for = 0.06 .
figure [ 4disk.09 ] shows images of models h1h4 at 655 myr with = 0.09 .
the increased response of the disk with decreasing mass ratio is clearly evident . the spiral structure for model h2
can be traced for well over 2@xmath130 radians .
the image of model h1 , with the lowest mass disk , resembles a classic kalnajs disk ( binney & tremaine 1987 ) .
the caustic structures seen in the high - resolution run of the fiducial simulation are seen throughout the disk and are highly elliptical .
differential rotation has caused the oval caustics to overlap and create a spiral pattern .
the third set of models described in 4 , those with varying bulge mass , can be thought of as an ersatz hubble sequence , at least along one dimension of the required criteria .
the properties of the disk and halo were chosen as to produce significant amplitudes in the fourier decomposition , reducing noise in the measured value of the pitch angle .
figure [ 4bulge.105 ] shows images of four models , @xmath131 and 2.0 with = 0.105 .
as the bulge mass varies from small to large , the spiral arms vary from loosely to tightly wound .
the arms themselves vary in thickness ; the larger the bulge mass , the narrower the spiral arms . for the model with @xmath132 ,
the spiral arms can be traced to within one scale radius of the center .
for @xmath133 , the arms do not begin until outside two scale radii ( although a high - resolution simulation could reveal structure inside this radius , as it did for the fiducial simulation ) .
figure [ hubble ] plots the average pitch angle against the logarithm of the bulge - to - disk ratio for each of these models .
( except for the model with @xmath134 of course , which had pitch angles of @xmath135 and @xmath136 for the halo rotation rates 0.075 and 0.105 respectively . )
the plot shows a linear increase in with @xmath137 .
as the bulge mass increases , the spiral arms become tighter .
the pitch angles for the simulations with = 0.105 are systematically higher than those with = 0.075 .
plotted as well are data taken from kennicutt ( 1981 ) for the pitch angles of galaxies along the hubble sequence , from sa - type to sc - type .
the bulge - to - disk ratios corresponding to these classifications were taken from kppen & arimoto ( 1990 ) .
the scatter in the observations , mentioned at the top of this section , is easily apparent .
the simulation data , however , follow the average trend of pitch angle with hubble type .
the slopes of the simulation data are flatter than that of the observations .
several explanations are viable for explaining this : the simulated hubble sequence created here is artificial since it only varies one parameter , the bulge mass
. there could be systematic variations of total galaxy mass and length scales along the hubble sequence which could effect the spiral response of the disk .
only two values of were used , and the steeper slope of the observations could be reproduced if it were assumed that higher bulge mass correlated with lower halo rotation rates .
lastly , our definition of , since it is an average over a range of @xmath46 , would tend to return values that are closer to the center of that range , reducing the slope of the correlation .
the pitch angles in kennicutt ( 1981 ) were measured by obtaining the arm coordinates from h@xmath46 or blue continuum observations and plotting these coordinates on the log @xmath42 vs @xmath78 plane .
a straight line was fitted to each arm and the results for the two main arms were averaged . as noted above , a better predictor than hubble classification of spiral pitch angle
is the rotational velocity of the galaxy ( kennicutt 1981 ) .
the relation found in the kennicutt paper is plotted by the solid lines in figure [ vrot ] .
these lines represent the mean and upper and lower bounds on his data .
the slope and intercept of these lines were calculated by inspection , and are not meant to be quantitatively compared to our simulation results .
qualitatively , however , it is obvious that our simulations follow the same relation
. both sets of bulge simulations lie well within the bounds set by the observations .
what is remarkable as well is that the simulations listed in table 2 also show this trend .
the simulations with the lower rotation velocity generally have higher pitch angles .
although the halo - dominated models do not return the same pitch angles as the disk - dominated models with the same rotation velocity , the trend is still preserved regardless of whether the mass dominating the kinematics is from the halo or the disk or the bulge .
the scatter in our results lies within the scatter seen observationally .
the lowest rotation velocity in the models listed in table 2 is d5 , with @xmath138 .
to explore the response of systems with even lower velocities , we used the models described at the end of 3 with both low mass disks and halos .
three different bulge masses were used : @xmath139 0 , @xmath123 and @xmath140 .
the first two of these models developed a strong and a weak bar respectively .
the method of bar detection used in 4 allowed us to eliminate the particles that contribute to the bar from being used in calculating .
since the length of the bar is known , all particles with radii below the bar length were simply not counted in the summation in equation ( 12 ) .
the results are plotted in figure [ vrot ] , and they follow the observational relation
. a tighter correlation can be acheived by including what we know of the mass distribution and halo rotation .
figure [ energy ] plots against the `` normalized '' rotational energy of the halo , a dimensionless quantity given by @xmath141 where @xmath142 is energy , @xmath143 is the maximum rotation velocity of the disk also used in figure [ vrot ] , and @xmath144 is the total mass of the galaxy .
the energy in the right hand side of equation ( 16 ) is the rotational energy of the halo , @xmath145 , where @xmath146 is the moment of inertia of the halo .
the subscript @xmath147 on the energy and the total mass represents the radius at which these two quantities are calculated .
the four windows in figure [ energy ] plot against calculated at four different values of @xmath147 .
the correlation is apparent in all four panels , but is excellent for @xmath148 .
the choice of @xmath147 significantly influences the slope of the correlation , with smaller values of @xmath147 resulting in steeper slopes .
the scatter changes only moderately with varying @xmath147 for the first four sets of models in figure [ energy ] , but effects the low - velocity models most .
the quantity has a power - law dependence on @xmath147 , which is highest with the low - velocity models .
thus the values of for that set will be affected most by changes in @xmath147 .
we have presented simulations of galaxy disks embedded in rigid , rotating , triaxial halos .
we have constructed a suite of models in which the masses of the disk , bulge and halo have all been varied . with these models , the effects of changing the halo rotation rate and flattening have been investigated .
our results have focused on the formation of bars and spiral structure .
models with a mass ratio @xmath149 and @xmath150 develop a strong bar , even if a significant bulge is present .
for all models , the torque created by the halo induces spiral structure in the disk .
the average pitch angle of this structure , , is strongly anti - correlated with the maximum rotation speed of the model .
it is clear that estimations of the stability of disk galaxies can not ignore the shape of the halo .
models with significant amounts of dark matter and a significant bulge ( like d1 and d2 , which have rotational velocities of over 200 in our system of units ) can become bar dominated with halo rotation rates of = 0.09 and greater .
maximal disk models , like d5 , are very susceptible to bar instabilities if torqued by the halo . in all the simulations presented in 5 , the only set of halo and bulge parameters for which models d3d5 were stable against bar formation was @xmath2 and = 0.06 .
even with those parameters , model d3 still developed a low amplitude but persistent bar .
a simple criterion for determining whether a disk galaxy will develop a bar , such as that determined by efstathiou ( 1982 ) , is not readily apparent from the data . with more free parameters , @xmath58 , @xmath60 , and , in our simulations
, it may be that a simple criterion can not quantify the stability of all our models used in our simulations .
the correlation of rotational velocity and pitch angle in our models provides a physical explanation of the same correlation found in observations of spiral galaxies .
the pitch angle of the arms induced by an external torque is related to the total mass of the system , loosely through @xmath151 , but more tightly to the dimensionless rotational energy parameter .
the differences seen in our results for different halo rotation rates and flattenings fit well with the amount of scatter seen in the correlation of @xmath151 with for real galaxies
. this result does not have to be unique to the properties of a galaxy s halo ; it has been shown that encounters with satellite galaxies , which would also create a time - varying potential field in which to disrupt the disk , induce two - armed spiral structure in disk galaxies ( toomre & toomre 1972 ) .
but encounters with smaller galaxies are by their nature transient and it remains that spiral structure is a ubiquitous feature of disk galaxies both with and without satellite perturbers . in kennicutt ( 1981 ) , and references therein
, the properties of the spiral arms are measured through their blue light , usually from hii regions , dust lanes , or the blue continuum , all of which are associated with star formation .
our collisionless simulations better represent the red stellar population which accounts for most of the mass in a galactic disk .
images of spiral galaxies in the near infrared have shown that spiral structure is evident in old stellar populations ( eskridge 2002 ) . the arms observed are smoother and wider than those seen in blue light , but generally follow the same morphology .
the fourier decomposition method used in this paper has been applied to observational data of hii regions in spiral galaxies ( puerari & dottori 1992 ; garca - gomz & athanssoula 1993 ) . the fourier spectra obtained resemble in many ways the spectra of our simulated galaxies , often with several distinct peaks inside one spectrum .
the pitch angle was calculated by these authors from the value of @xmath46 for which the amplitude is highest .
this definition of @xmath85 reproduces the general trend with hubble type as described by kennicutt ( 1981 ) but does not reflect the natural variations in pitch angle within a single galaxy seen both in the fourier spectra of the observations and in kennicutt s calculations .
kennicutt noted that the variations of @xmath85 for a typical galaxy were significantly larger than the measurement error .
an average value of the pitch angle better represents the multiple spiral components .
hydrodynamic simulations by bekki & freeman ( 2002 ) have shown that the extended spiral structure seen in the blue compact dwarf galaxy ngc 2915 could be due to a rotating triaxial halo .
their simulations contained a disk fully composed of gas particles and a halo 100 times more massive than the disk , effectively making the disk massless .
our collisionless simulations show spiral structure in the inner regions of galactic disks
. since gas fractions in disk galaxies are usually no higher than 30% , the gas should be gravitationally coupled to the stars and follow the spiral patterns seen in our simulations .
the dissipation and subsequent star formation in the gas would lead to self - consistent spiral structure which would outlast a collisionless disk and work to preserve its structure as well .
an open question is the correct value or range of values to use for .
the figure rotation of ngc 2915 has been calculated to be 8.0 ( bureau 1999 ) , which is 0.107 in our @xmath0-body units .
these authors also propose that halo figure rotation , a distinct phenomenon from halo angular momentum , is present in a significant fraction of dark matter halos in cdm simulations .
regardless of the the details in determining or , it has been clearly demonstrated that an external torqueing mechanism can reproduce the type of correlation observed between spiral pitch angle and rotational velocity . even in simulations in which the magnitude of the applied torque is the same , the density profile , flattening , and rotation rate of that halo
are all held constant , the pitch angle of the spiral structure varies according to the total mass of the system .
jlt would like to thank richard pogge for his helpful discussions , as well as volker springel and stein sigurdsson for help with the and codes respectively .
this work was performed at the ohio supercomputer center under grant pas0825 . | we investigate the effect of rotating , triaxial halos on disk galaxies through an extensive set of numerical @xmath0-body simulations .
our simulations use a rigid potential field for the halos and bulges and collisionless particles for the disks .
the triaxiality and the rotation rate of the halo are varied , as well as the masses of all three galaxy components .
we analyze both the bar stability and the spiral response of the disks under these conditions .
we characterize most of our models by the mass ratio of the disk to the halo at 2.3 disk scale lengths , . for models with a mass ratio greater than 0.8 , a halo pattern speed = 6.7 , and a intermediate - to - major axis ratio @xmath1
, a strong bar will develop within 3 gyr , even for models with a bulge mass @xmath2 .
models in which the bulge mass is reduced by half develop bars earlier and with lower .
we create an artificial hubble sequence of disk galaxies by varying the bulge - to - disk ratio of our models from 0 to 2.5 . the torque induced by a rotating ,
non - axisymmetric halo creates bisymmetric spiral structure in the disk .
we find that the pitch angle of the spiral arms in these models follows the same general trend found in observations of spiral galaxies , namely that later type galaxies have higher pitch angles .
our simulations follow closely the observational relation of spiral pitch angle with maximum rotational velocity of the disk , where galaxies with faster rotation have more tightly wound spiral arms .
this relation is followed in our simulations regardless of whether the dominating mass component of the galaxy is the disk , the halo , or the bulge .
2max@xmath3 2ave@xmath4 |
more than forty years after the discovery of radio pulsars , the mechanism by which they emit coherent radio beams is still not fully understood .
also , many properties of this radiation remain a mystery , especially the phenomenon of drifting subpulses .
this puzzling phenomenon was widely regarded as a powerful tool for the investigation of the pulsar radiation mechanism .
recently , this phenomenon received renewed attention , mostly owing to the newly developed techniques for the analysis of the pulsar radio emission fluctuations ( edwards & stappers 2002,2003 ; es02,es03 henceforth ) . using these techniques , weltevrede et al .
( 2006 a , b ; w06a , b henceforth ) presented results of the systematic , unbiased search for the drifting subpulses and/or phase stationary intensity modulations in single pulses of a large sample of pulsars .
they found that the fraction of pulsars showing evidence of drifting subpulses is at least 60 % and concluded that the conditions for the drifting mechanism to work can not be very different from the emission mechanism of radio pulsars .
it is therefore likely that the drifting subpulse phenomenon originates from the so - called inner acceleration region right above the polar cap , which powers the pulsar radiation . in the classical model of ruderman & sutherland ( 1975 ; rs75 henceforth ) the subpulse - associated spark filaments of plasma circulate in the pure vacuum gap
( vg hereafter ) around the magnetic axis due to well known drift of plasma with non - corotational charge density ( see appendix a for more details ) .
there are few periodicities characteristic for this model , called also the pulsar carousel model : the primary period @xmath9 which can be measured as a distance between the observed subpulse drift bands , the secondary period ( apparent when drifting is aliased ; gil & sendyk 2003 for detailed description ) , and the tertiary period @xmath7 ( called also the carousel time in rs75 .
although this symbol is still in use , we advocate to replace it by @xmath7 . ] , as it is the time interval after which the gap plasma completes one full circulation around the magnetic pole ) .
the carousel model is widely regarded as a natural and qualitative explanation of the drifting subpulse phenomenon .
however , its original version published by ruderman & sutherland ( 1975 ; rs75 hereafter ) predicts too high a drifting rate of the sparks around the polar cap , as compared with the observations of drifting subpulses ( e.g. deshpande & rankin 1999 , 2001 ; dr99,dr01 henceforth ) , and too high a heating rate of the polar cap ( pc henceforth ) surface due to the spark - associated back - flow bombardment , as compared with x - ray observations ( e.g. zhang et al .
another difficulty of the rs75 model is that recent calculations strongly suggest that the surface binding energy of both ions and electrons are too low to allow the development of a vacuum gap .
indeed , when the surface magnetic field is purely dipolar , then the gap can develop only in magnetars and several highest b - field pulsars ( medin & lai 2007 ) .
another type of inner accelerator model , named space - charge - limited flow ( sclf , arons & scharlemann 1979 ; harding & muslimov 1998 ) , has been discussed in the literature , which assumes that both ions and electrons can be freely striped off the neutron star surface .
although this approximation is valid for most pulsars assuming a pure dipolar field at the polar cap region , a stronger , multipole magnetic field near the polar cap region ( which is needed to make a large number of radio pulsars above the radio emission death line , ruderman & sutherland 1975 ; zhang et al .
2000 ) would introduce a non - negligible binding energy of ions / electrons ( medin & lai 2007 ) , which renders the sclf approximation no longer valid .
another difficulty of the steady - state sclf model widely discussed in the literature is that it does not predict the existence of any `` sparks '' that could give rise to the drifting sub - pulses .
so , in our opinion , it is not an attractive inner accelerator model to interpret pulsar radio emission . motivated by these shortcomings of the otherwise attractive vg model gil , melikidze & geppert ( 2003 ; g03 henceforth ) developed further the idea of the inner acceleration region above the polar cap by including the partial screening caused by the thermionic flow of ions from the pc surface heated by sparks .
we call this kind of the inner acceleration region a `` partially screened gap '' ( psg hereafter ) .
the psg is thermally self - regulated in such a way that the surface temperature is always close to but slightly lower ( less than 1 percent ) than the critical temperature at which the maximum co - rotational ion outflow occurs and the gap is fully screened ( see appendix and/or g03 for more details ) .
moreover , if the surface temperature was even few percent lower than the critical temperature , there would be a pure vacuum gap , with all the problems discussed above .
since the actual potential drop in the psg is much lower than that of the pure vg model ( rs75 ) , the intrinsic drift rate and pc heating rate are compatible with measurements of @xmath7 and @xmath6 , respectively .
the psg model can be tested if two observational quantities are known : ( i ) the circulational period @xmath7 for drifting subpulses observed in radio - emission and ( ii ) the x - ray bolometric luminosity @xmath6 of thermal bb radiation from the hot polar cap ( see appendix a ) .
radio pulsars were targeted since beginning of x - ray astronomy for various scientific reasons .
zhang , sanwal & pavlov ( 2005 ; z05 henceforth ) were the first who made an attempt to resolve the mystery of drifting subpulses in radio pulsars by observing them in x - rays .
they proposed to detect thermal x - ray photons from the pc heated by sparks of plasma likely to be associated with drifting subpulses observed in radio band .
their choice was the best studied drifting subpulse pulsar b0943 + 10 . using x - ray observatory they detected a weak source coincident with the target pulsar . due to very small number of counts detected , no unambiguous spectrum could be obtained .
however , they were able to fit the bb model to the data , although a power law model was acceptable as well . within a bb model
they inferred a bolometric luminosity @xmath10 erg / s emitted from the hot spot ( few mk ) with a surface area much smaller ( about 60 times ) than the conventional polar cap area as defined by the bundle of last closed dipolar field lines .
this radio pulsar was well studied by dr99 , who described the number of sparks and the circulation time @xmath11 needed for them to complete one full revolution around the pole ( where @xmath12 is the basic pulsar period ) .
these properties as well could not be accounted for by the conventional theory , and some radical modification of rs75 model was required .
it appears that psg model not only resolves all the problems of the rs75 model , but also offers a clean prediction that can be used to test theories of the inner pulsar accelerator .
gil , melikidze & zhang ( 2006b ; paper i henceforth ) reanalyzed the b0943 + 10 case within the psg model .
they derived a very useful formula directly connecting the drifting rate of plasma sparks ( measured by the circulation period @xmath7 ) and the polar cap heating rate by the back - flow spark bombardment ( measured by the bolometric thermal luminosity @xmath6 ) . by assuming that both the measured quantities are determined by the same value of electric field in the psg , they obtained a simple formula relating the so - called efficiency of thermal radiation from the hot polar cap with the circulation time @xmath13 where @xmath14 is the pulsar spin - down ( see eq .
[ a3 ] with @xmath15 in appendix a ) .
psr b0943 + 10 with its data specified in table 1 , fitted this observational curve quite well ( fig .
1 ) . when one observable parameter in equation ( 1 ) is known ( @xmath6 or @xmath7 ) , the other one can be predicted without any free parameters . in paper
i we included b1133 + 16 , the twin pulsar to b0943 + 10 ( at least in the sense of the kinematical properties ; see table 1 ) . in this second case
we speculated that the long periodicity of about @xmath16 revealed by a number of authors ( e.g. w06a , b ) , is actually the circulational period @xmath17 .
this claim was recently confirmed by sophisticated data analysis of herfindal & rankin ( 2007 ; hr07 henceforth ) , although these authors admitted that they did not believe our prediction of @xmath7 value before their own analysis .
the x - rays from b1133 + 16 were detected by kargaltsev , pavlov and garmire ( 2006 ) using chandra x - ray observatory , who found that their properties were similar to those of the twin pulsar b0943 + 10 . because of the small number of counts detected , obtaining an unique spectrum was not possible , like in the case of psr b0943@xmath010 ( zsp05 ) .
however , the bb model was acceptable and gave the bolometric luminosity @xmath18 erg / s emitted from the hot ( few mk ) and very small polar cap ( again much smaller ( about 100 times ) than the canonical one ) .
as one can see in figure 1 , with the inferred values of @xmath7 and @xmath6 the pulsar b1133 + 16 nicely clusters with its twin pulsar along the critical curve expressed by equation ( 1 ) .
note that filled circle represents our prediction and asterisk represents the estimate of @xmath7 by hr07 .
encouraged by the observational confirmation of our prediction of @xmath7 in b1133 + 16 , we applied the same method to two other pulsars for which the measurements or estimates of thermal bolometric luminosity were available ( gil , melikidze & zhang 2007 ; paper ii henceforth ) .
one of the famous three musketeers , psr b0656 + 14 , in which thermal x - rays from small hot polar cap were clearly detected by de luca , caraveo , mereghetti , et al .
( 2005 ; dl05 hereafter ) , was an obvious choice .
the bb thermal luminosity @xmath19 ergs / s ( table 1 ) inserted into equation ( 1 ) , returned the predicted value of @xmath20 .
amazingly , weltevrede et al . ( 2006c ; w06c henceforth ) reported the long - period fluctuation spectral feature @xmath21 associated with the quasi - periodic amplitude modulation of erratic and strong radio emission detected from this pulsar .
thus , it was tempting to interpreted this period as the circulation time @xmath7 . with this value of @xmath7 and @xmath6 shown above , the pulsar b0656 + 14 fits the equation ( 1 ) quite well ( figure 1 ) . although the drifting subpulses were not apparent in this case , the erratic radio emission reported by w06c was similar to the so - called q - mode in psr b0943 + 10 ( showing clearly drifting subpulses in the organized b - mode ) . the low frequency feature in the fluctuation spectra , identical to that of the b - mode , was found by rankin & suleymanova ( 2006 ; see their fig .
asgekar & deshpande ( 2001 ; ad01 hereafter ) also detected this feature in the 35-mhz observations of psr b0943 + 10(see their figs.1 and 2 ) .
this simply means that the carousel plasma drift is maintained in both regular drifting and erratic ( with no drifting subpulses ) pulsar emission modes .
this is a property of plasma and magnetic field interaction in the gap rather than the structure of this plasma .
however , drifting subpulses can be clearly observed only if the gap plasma has some lateral structure , localized sparking discharges for instance . for the second of the three musketeers , psr b1055 - 52 ,
we have just found an evidence of a low frequency feature @xmath22 ( biggs 1990 ; b90 henceforth ) , which can be interpreted as the carousel periodicity @xmath23 . using this interpretation , which was very fruitful in several other cases discussed above and below , we examine thermal x - ray radiation from the small hot spot detected in this pulsar and attempt to test our psg model in section 4.4 .
the third musketeer ( geminga ) is radio quiet , so although it shows thermal bb x - ray emission from the small hot spot , it is not useful for our analysis . another pulsar that we could examine using our method of inferring values of @xmath7 from intensity modulation spectra was psr b0628 - 28 .
as indicated in table 1 , it was detected in x - rays by tepedelenliolu & gelman ( 2005 ; t05 henceforth ) , using chandra and observatories .
this was an exceptional pulsar ( called an overluminous one by becker et al .
2005 ) with efficiency much larger than that of typical pulsars ( becker & trmper 1998 ) . for thermal bb component alone
@xmath24 ( table 1 ) .
this value inserted to equation ( 1 ) gives the predicted value of @xmath25 .
interestingly , w06c reported for this pulsar a relatively short periodicity of @xmath26 ( table 1 ) . if this periodicity is interpreted as the circulation time @xmath7 , then this is pulsar is not exceptional at all .
it lies on the theoretical curve ( eq .
[ 1 ] ) in figure 1 at exactly the right place .
psr b0628 - 28 is just another ( fourth ) pulsar satisfying the predictions of equation ( 1 ) , which relates the efficiency of thermal x - ray radiation from a hot polar cap to the circulational periodicity associated with drifting subpulses observed in radio emission . in order to expand the sample of pulsars that have
both @xmath6 and @xmath7 measured / estimated , we recently launched an observational campaign using the observatory .
we targeted at two old pulsars that had @xmath7 measurements but had no x - ray observations before , and we report on the results of these observations in this paper .
the two pulsars , psr b0826 - 34 and psr b0834@xmath006 were observed during the cycles ao-5 and ao-6 , respectively . simultaneous radio monitoring was also performed and we will report on these observations in the separate paper . psr b0826 - 34 was not detected and we have derived an upper limit for its thermal luminosity .
we clearly detected psr b0834@xmath006 , whose spectrum is best modelled by a bb radiation from a small hot spot .
we interpret this as due to pc heating by the back - flow bombardment , and found that the bolometric @xmath6 agrees well with equation ( 1 ) predicted by the psg model . for completeness , in this paper
we include yet another pulsar psr b1929@xmath010 , whose bolometric thermal luminosity was recently determined by misanovic , pavlov & garmire ( 2007 ) .
we show that this pulsar also satisfies equation ( 1 ) by finding a suitable feature in the modulation spectra data base of w06a , b ( see section 3.3 for some details ) .
the number of pulsars satisfying and/or being consistent with equation ( 1 ) increased to seven . to the best of our knowledge no single counter - example exits .
it is worth emphasizing that only pulsars for which both the bolometric luminosity @xmath6 of thermal x - rays from hot polar cap and circulational periodicity @xmath7 of drifting subpulses observed in radio band are known , can be used for this analysis . in our sample of 8 available cases , 4 pulsars ( b0656@xmath014 , b1055@xmath152 , b0834@xmath006 and b1929@xmath010 ; see footnote 8 related to the latter case ) and 3 others either show an evidence of hot spot thermal emission ( b1133@xmath016 and b0628@xmath128 ) or at least such component can not be excluded ( b0943@xmath010 ) .
the last case ( b0826@xmath134 ) is uncertain as we only have an upper limit for x - ray detection ( consistent with psg model ) .
we have observed two radio pulsars b0834@xmath006 and b0826@xmath034 known for their prominent subpulse drift with the observatory ( jansen , lumb , altieri et al .
we marked them in red color in figure 1 , to distinguish them from previously analyzed four pulsars ( marked in black ) in papers i and ii .
yet another pulsar b1929@xmath010 ( marked in blue in figure 1 ) is discussed in section 3.3 , as its values of @xmath6 and @xmath7 have became recently available .
the pulsar psr b0834@xmath006 was observed with on 2007 november 17 and 18 for a total of @xmath2771.7 ks . the epic - mos ( turner , abbey , arnaud et al .
2001 ) and epic - pn ( strder , briel , dennerl et al .
2001 ) cameras were operated in imaging mode ( see table [ tab - obs ] ) .
the observation was scheduled at the end of the satellite revolution and the detector background strongly increased when the satellite entered the radiation belts . to maximize the signal to noise ratio we rejected the period of high background which resulted in net exposure times around 50 ks ( table [ tab - obs ] ) . for the x - ray analysis we used the science analysis system ( sas ) version 7.1.0 together with xspec version 11.3.2p for spectral modelling .
standard sas source detection based on a maximum likelihood technique was simultaneously applied to the x - ray images obtained from the three epic instruments and five different energy bands ( band b1 [email protected] kev , b2 [email protected] kev , b3 [email protected] kev , b4 [email protected] kev and b5 [email protected] kev ) .
a weak source was found at the position of the pulsar at r.a .
= 08 37 05.71 and dec .
= 06 10 15.8 ( j2000.0 ) with a 1@xmath28 statistical error of 1.7 .
nearly 150 x - ray sources were detected in the epic images and a comparison with catalogues from other wavelength bands yields many correlations within @xmath270.5 of the x - ray positions .
this demonstrates that the systematic uncertainty in the astrometry is small compared to the statistical error of the source position .
the x - ray source position is within 1.3 of the radio position of psr b0834@xmath006 , consistent within the errors .
the positional agreement and other properties of the x - ray source ( see below ) make a chance coincidence very unlikely .
the total epic count rate ( summed for the three instruments in the [email protected] kev band ) obtained from the source detection analysis is ( [email protected])@xmath3010@xmath31 cts s@xmath4 , insufficient for a detailed spectral analysis . to obtain constraints on the shape of the x - ray spectrum we therefore use hardness ratios ( x - ray colours ) derived from the count rates in the standard energy bands and
compare them with those expected from various model spectra .
because the epic - pn detector is more sensitive , in particular at low energies where most of the counts are detected , we use only count rates obtained from epic - pn .
hardness ratios are defined as hr1 = ( r2-r1)/(r2+r1 ) , hr2 = ( r3-r2)/(r3+r2 ) , hr3 = ( r4-r3)/(r4+r3 ) and hr4 = ( r5-r4)/(r5+r4 ) with rn denoting the source count rate in band bn . to compare the measured hardness ratios with those inferred from model spectra , we simulated expected epic - pn spectra ( using xspec and the appropriate detector response files ) and derived expected count rates and hardness ratios . the distance to psr b0834@xmath006
estimated as 643 pc was derived from its dispersion measure of dm = 12.86 pc cm@xmath31 ( from the online atnf pulsar catalog ) . assuming a 10% ionization degree of the interstellar matter along the line of sight to psr b0834@xmath006 , this converts to a hydrogen column density of n@xmath33 = 4.0@xmath34 @xmath35 . because of the low statistical quality of the x - ray data we are not able to derive tight constraints on the absorbing column density .
therefore , we limit our investigated model parameter space to n@xmath33 values between 1.0@xmath34 @xmath35 ( a lower limit which is reached within a distance of 200 pc ; posselt , popov , haberl , et al . 2008 ) and 8.0@xmath34 @xmath35 ( allowing an uncertainty of a factor of 2 in the assumed ionization degree for the conversion from dm to n@xmath33 ) . as model spectra we tested power - law ( pl hereafter ) and blackbody ( bb hereafter ) emission and a combination of the two .
in all model spectra absorption was included , assuming elemental abundances from wilms , allen & mccray ( 2000 ) .
for the absorbed power - law model we explored the parameter space for n@xmath33 between 1.0@xmath34 @xmath35 and 8.0@xmath34 @xmath35 with a step size of 1.0@xmath34 @xmath35 and for the photon index @xmath36 between 1 and 5 in steps of 0.2 .
figure 2a shows the hardness ratios hr1 versus hr2 derived at the parameter grid points .
the measured hardness ratios hr1 and hr2 are drawn with 1@xmath28 ( solid lines ) and 2@xmath28 ( dotted lines ) error bars .
the rectangular boxes around the error bars indicate the corresponding confidence areas , although these are in reality limited by error ellipses which fit inside the boxes . as can be seen
, the power - law model spectra can not reproduce the measured hardness ratios within their 1@xmath28 errors .
allowing 2@xmath28 errors would require a relatively steep power - law with a photon index between 2 and 4 and preferentially high absorption .
the results for a bb model with temperatures varying between @xmath37 ev and @xmath38 ev in steps of 20 ev ( n@xmath33 grid as above ) are plotted in figure 2b .
the measured hardness ratios are best reproduced by the model with n@xmath33 = 4@xmath34 @xmath35 and @xmath39 ev . the @xmath40 ( @xmath41 ) confidence range for the temperature is @xmath42 ev .
we determine bolometric luminosity using the model parameters at the grid points ( normalizing the simulated spectra to match the observed count rate in the [email protected] kev band ) which yielded = 8.6@xmath43 @xmath44 erg s@xmath4 .
it is remarkable that the hydrogen column density derived from the `` best - fit '' bb model of 4.0@xmath34 @xmath35 is fully consistent with the dm assuming 10% ionization along the line of sight to psr b0834@xmath006 .
we also investigated a combination of bb and pl ( with a photon index of 2.0 as typically seen in the x - ray spectra of pulsars ( e.g. kargaltsev et al .
2006 ) , both subject to the same absorbing column density . as first case
the normalization of the power - law component was set to have a flux ( for the [email protected] kev band ) in the pl component of 50% of that in the bb component , i.e. a flux ratio of : = 1:0.5 .
the hardness ratios are shown in figure 2c . as expected , hr2 , which is most sensitive to the shape of the intrinsic spectral shape , increases with respect to the case of the pure bb due to the contribution of the harder pl component .
the @xmath40 ( @xmath41 ) confidence ranges are @xmath45 ev and @xmath46 @xmath44 erg s@xmath4 .
it should be noted here , that the luminosity of the bb component increases , although a power - law component is added to the model spectrum .
this is because the power - law rises toward the low energies and a higher n@xmath33 values is required to compensate for that .
a higher n@xmath33 in turn increases the bolometric luminosity of the bb component in order to match the observed spectrum ( hardness ratios and count rates ) again .
these effects are also evident in the second case , where we used a flux ratio of : = 1:1 ( fig .
2d ) : hr2 increases further and the upper limits for also rise somewhat ( @xmath47 ev ; = 9.9@xmath48 @xmath44 erg s@xmath4 ) .
the above results are summarized in figure 3 which presents @xmath6 versus @xmath49 obtained from the modelled hardness ratios in figures 2b-2d , where symbols ( circle and square for @xmath40 and @xmath41 levels , respectively ) , their colors ( red , blue and green for bb , bb(2/3)+pl(1/3 ) and bb+pl model , respectively ) and related numbers , correspond to those used in figures 2b-2d .
we can summarize that thermal radiation from the hot polar cap of psr b0834 + 06 is described by kt = ( 170@xmath50 ev ( or surface temperature of the polar cap @xmath51 k and @xmath6 = @xmath52 erg s@xmath4 , where the @xmath41 errors are determined by both statistical and model uncertainties .
the pulsar psr b0826 - 34 was observed with on 2006 november 13 and 14 with the epic - mos and epic - pn cameras operated in imaging mode ( table [ tab - obs ] ) . also during this observation ,
strong background flaring activity ocurred near the end of the observation .
after background screening a total exposure time of @xmath27 38.8 ks was obtained .
we selected this source because it was one of the few pulsars with known @xmath7 value ( gupta , gil , kijak et al .
2004 ; g04 hereafter ) . when applying for the observing time we realized that psr b0826 - 34 would be at most a very weak source like psr b0943 + 10 ( or even weaker ) detected by z05 .
indeed the spin - down value was quite low and even our equation ( 1 ) predicted the source luminosity twice lower than that of b0943 + 10 .
however , b0826 - 34 is closer to the earth than b0943 + 10 by the factor of 1.5 . despite relatively large dm=52.9 pc cm@xmath31 we optimistically assumed that the hydrogen column density n@xmath33 will be similar to that of psr b0943 + 10 ( with dm=15.4 pc cm@xmath31 ) .
we speculated that the factor of 3.5 in dm values would be compensated to some degree by the factor of 0.67 in a distance .
we did not detect the pulsar , which probably means that the actual value of n@xmath33 is much higher than assumed , due to some dense cloud of hydrogen along the line of sight to b0826 - 34 .
therefore , we determined the upper limit for thermal x - ray radiation from hot pc from this pulsar . because of the higher sensitivity of the epic - pn camera we used images from this instrument only .
we created images in the energy bands @xmath53 kev , @xmath54 kev and @xmath55 kev and determined @xmath41 upper limit count rates for the expected source position for each energy band .
the total @xmath56 kev upper limit was obtained as 2.3 cts s@xmath4 . assuming the bb model with @xmath57 ev and the absorption column density of 3@xmath34 @xmath35 ( 1.4@xmath58 @xmath35 ) , this converts into an upper limit for the bolometric luminosity of = 1.0@xmath59 erg s@xmath4 ( = 1.45@xmath59 erg s@xmath4 ) .
the latter value was conservatively used in figure 1 ( table 1 ) .
table 1 and figure 1 present the observational data of a number of quantities for seven pulsars , in which both @xmath7 and @xmath6 are known or at least constrained .
these data are confronted with the model curve representing equation ( 1 ) , which is marked by the solid line , accompanied by broken lines describing theoretical errors due to uncertainty with determination of the neutron star moment of inertia ( see appendix a ) . two pulsars : b0943 + 10 and b1133 + 16 have already been discussed in paper i , and two others b0656 + 14 and b0629@xmath128 in paper ii .
as argued in papers i and ii these pulsars strongly support our theory ( they are presented as black dots in our figure 1 ) . in the following
we study and discuss the results from the remaining three pulsars : b0834 + 06 , b0826@xmath134 ( red dots ) and b1929 + 10 ( blue dot ) .
as already mentioned , the circulational ( tertiary ) period @xmath7 is known for a handful of pulsars , and b0834 + 06 is one of them .
the first measurement of tertiary periodicity for this pulsar was made by asgekar & deshpande ( 2005 ; ad05 henceforth ) , who argued that @xmath60 and the number of circulating sub - beams ( sparks ) @xmath61 , implying the aliased subpulse drifting with primary period @xmath62 .
they found a strong low frequency feature in the intensity fluctuation spectrum at @xmath63 in one sequence of 64 single pulses , supported by side tones flanking the primary feature of @xmath64 by @xmath65 .
these results seemed quite robust , although a small derived number of sparks ( 8) as compared with other cases was a bit worrying .
we used @xmath66 in the scientific justification for proposal , predicting from equation ( 1 ) a quite luminous hot pc in psr b0834 + 06 , emitting with = 36@xmath44 erg s@xmath4 .
the model simulations indicated the count rate of about 0.018 cts s@xmath4 , implying a very promising case .
slightly before the scheduled xmm observing session a new estimate was obtained by rankin & wright ( 2007 ; rw07 henceforth ) , who argued , using their new arecibo data and new technique involving a distribution of null pulses , that @xmath67 .
they argued that the number of sparks and/or subbeams involved in the non - aliased subpulse drift with the true primary period @xmath68 is 14 and thus @xmath69 ( table 1 ) . according to equation
( 1 ) this would imply the luminosity 4.16 times lower than = 36@xmath44 erg s@xmath4 given in our proposal , that is = 8.85 @xmath44 erg s@xmath4 or @xmath70 .
amazingly , this is almost exactly the central value of our best fit for hot bb component in psr b0834@xmath006 ( see table 1 and figs . 1 and 3 ) .
thus our measurements interpreted within the psg model ( eq . [ 1 ] ) strongly support the value of @xmath71 obtained by rw07 , while @xmath60 obtained by ad05 is highly unlikely .
the carousel rotation time in this pulsar was obtained by means of computer simulations compared with real single pulse data by gupta , gil , kijak et al .
1984 . according to equation ( 1 ) its valule @xmath72 implies the efficiency @xmath73 .
these values are marked by the red horizontal error bar labelled by b0826 - 34 .
the upper limit 22 @xmath74 is marked as the short arrow above . this pulsar would have to be much more efficient in converting the spin - down power into x - rays to be detected in a 50 ksec exposure , or at least a six times longer exposure time would be required .
recently misanovic , pavlov and garmire ( 2007 ; m07 hereafter ) argued that x - rays from psr b1929@xmath010 include both magnetospheric and thermal components .
the bb fit to the latter gives a temperature k@xmath75=0.3 kev and a projected surface area @xmath76 m@xmath77 or radius @xmath78 of about 33 meters ( much smaller than the canonical @xmath79 m@xmath77 or @xmath80 ) .
this corresponds to the bolometric luminosity @xmath81 ergs s@xmath4 emitted from hot ( @xmath82 k ) polar cap with a radius of about 33 meters.19 ( using different way of data binning resulting in better photon statistics per spectral bin ) and argued that the hot bb component is statistically unjustified .
however , if they allowed the bb radius and temperature of the hot spot as the free parameters , then the best fit resulted in very small hot spot area with a radius @xmath83 meters , perhaps even smaller that the one obtained by m07 .
in opinion of hb08 this is unacceptable small as compared with the canonical pc radius .
however , within our model this is a result of relatively low dipolar surface magnetic field @xmath84 gauss
. the actual non - dipolar magnetic field must be much higher ( about 400 times ) to provide enough binding energy ( ml07 ) for creation of the psg in this pulsar , which results in the hot spot radius @xmath85 meters ( see section 5 for more details ) .
] we used the central value of b1929@xmath010 @xmath86 ergs s@xmath4 with 2@xmath28 errors from m07 ( see the top panel in their figure 11 ) . for each new pulsar with a known value of thermal bolometric luminosity @xmath6 we search the available data bases for a possible value of @xmath7 .
in case of psr b1929 + 10 we found in w06a ( their figure a13 ) a clear but weak low frequency spectral feature at about 0.02 @xmath87 .
this translates into a long periodicity @xmath88 , with errors estimated from half - width of the low frequency feature .
going back to figure 1 we see that the data point ( marked in blue color ) for b1929 + 10 ( table 1 ) fits the theoretical curve very well .
this is an important point , as it extends the parameter space to the low efficiency/(long period ) region in our figure 1 .
the range of parameters for our 7 cases under examination increased to factors of 67 and 7 for the efficiency @xmath89 and the tertiary period @xmath90 , respectively .
this is a bright radio pulsar showing complex patterns of single pulse intensity modulations .
the drifting subpulses are not apparent but this can be the result of a central cut of the line - of - sight ( los ) throughout the emission beam . indeed , this pulsar has a strong interpulse ( ip ) separated from the main pulse ( mp ) by about 145 degrees of longitude ( measured between centroids ) and both these components have complex profiles , consistent with central los traverse .
b90 analyzed the fluctuation spectrum and found in part of the profile a small and broad feature at frequency 0.045 cycles/@xmath12 with @xmath91 .
this frequency and low q can be interpreted as the carousel periodicity @xmath92 .
recently , mitra ( 2008 ) confirmed this feature at parts of both mp and ip in his data taken at gmrt ( privat information ) .
the pulsar psr b1055@xmath152 is a luminous source of x - ray emission .
dl05 identified three spectral components in this radiation : power law magnetospheric emission , cool bb emission from the entire surface of the cooling neutron star , and hot bb emission from a small hot spot .
this latter component is of special interest for us and its parameters along with references are listed in table 1 .
if the `` carousel '' hypothesis discussed above is correct , then we expect a correlation between the carousel period @xmath7 and the bolometric luminosity @xmath6 from the hot spot , according to our equation ( 1 ) . as shown in table 1 and fig .
1 , the bolometric luminosity @xmath93 erg / s , where the errors were estimated from fitting the epic - pn spectrum , extracted from the xmm - newton archival data , with the same model as used by dl05 . although the central values result in the data point lying slightly below the theoretical curve , psr b1055@xmath152 is certainly consistent with equation ( 1 ) . indeed , one can see that values slightly higher than the central one , e.g. , @xmath94 and @xmath95 would result in a very good fit to the equation ( 1 ) .
within the partially screened gap ( psg ) model of the inner acceleration region in pulsars developed by g03 , we derived in paper i a simple and clean relationship ( eq . [ 1 ] ) between the thermal x - ray bolometric luminosity @xmath6 from hot pc heated by sparks and the circulation time @xmath7 of the spark - associated drift detected as the subpulse drift in pulsar radio emission .
this relationship expresses the well justified assumption ( appendix a ) that both the drifting rate and the polar cap heating rate are determined by the same value of electric field within the inner acceleration region .
indeed , the drifting rate described by measurable @xmath7 is determined by the tangent ( with respect to surface magnetic field ) component of the electric field , while the heating rate described by measurable @xmath6 is determined by its component parallel to the surface magnetic field in the ( partially screened ) gap . in paper ii
we showed that psrs b0943@xmath010 , b1133 + 16 , b0628@xmath120 and b0654 + 14 , which were the only pulsars with both @xmath6 and @xmath7 known at that time , satisfied equation ( 1 ) quite well ( see also fig . 1 and table 1 ) .
this suggested that the psg model may indeed be a reasonable description of the inner accelerator region near the polar cap . in this paper
we support this view by demonstrating that another two pulsars ( b0834 + 06 , b1929 + 10 and b1055@xmath152 ) also satisfy the equation ( 1 ) .
yet another pulsar b0826 - 34 , in which only the upper limit for @xmath6 was obtained , demonstrated a consistency with equation ( 1 ) as well . only for a handful of pulsars the circulation ( carousel ) time
was measured or constrained so far .
measurement of @xmath7 by means of modulation spectral analysis requires a strong unevenness in the circulating system , maybe a distinguished group of adjacent sparks or even just a single spark ( see also the scenario discussed by gil & sendyk , 2003 ; gs03 hereafter ) .
moreover , this feature should persist considerably longer than the circulation time .
such favorable conditions do not occur frequently in pulsars and therefore direct or indirect measurements of @xmath7 are very rare . in principle , in a clean case one should be able to detect the primary feature @xmath9 , reflecting the phase modulation of regularly drifting subpulses , flanked by two symmetrical features corresponding to slower amplitude modulation associated with carousel circulation as well as direct low frequency feature @xmath96 ( like in the case of psr b0943 + 10 ; dr01 , ad01 and gs03 ) .
however , results of paper ii clearly showed that @xmath7 can be found also in pulsars without regularly drifting subpulses ( and/or in erratic drifting modes ) .
this strongly suggested that no matter the degree of the organization of spark plasma filaments at the polar cap , the slow circumferential plasma drift was always performed at about the same rate in a given pulsar .
the problem was how to reveal this motion .
two new methods were discussed or at least mentioned in paper i. the 2-d phase resolved modulation spectral analysis developed by es02 and es03 and implemented by w06a , b was the first one .
the second method based on examination of the distribution of nulls in the long sequence of single pulses was recently developed by hr07 and rankin and wright ( 2007 ; rw07 henceforth ) . in view of the main results obtained in this paper the latter method
deserves some more detailed discussion here . as discussed in section 3.1
there is a controversy about the actual value of @xmath7 in psr 0834 + 06 .
ad05 reported that the alias - corrected @xmath62 and @xmath60 , implying the number of sparks @xmath61 .
these authors found just one sequence of 64 pulses in which the fluctuation spectrum analysis revealed the low frequency feature at about 1/15=0.067 . on
the other hand rw07 found the non - aliased primary drift periodicity @xmath97 and the number of sparks @xmath98 , implying the tertiary long periodicity @xmath69 .
this longer cycle with @xmath99 was supported by our measurements of @xmath6 and psg model expressed by equation ( 1 ) .
rw07 examined an interaction between nulls and emission in psr b0834 + 06 .
they found that null pulses are not randomly distributed and that the most likely periodicity in their appearance is about @xmath100 .
following the previous discovery of hr07 that null pulses and drifting subpulses in psr b1133 + 16 are associated with the same long periodicity ( about @xmath101 ) rw07 convincingly argued that short pseudo - nulls ( one pulsar period or less ) are just a result of irregular distribution of subpulse subbeams / sparks that persist on time scales of at least hundreds of pulsar periods .
the short - time pseudo - nulls appear when the line - of - sight cuts through the low - level emission region in the radio beam .
our results on both b1133 + 16 and b0834 + 16 strongly support this picture .
the interesting question is then why ad05 obtained such a strong feature at @xmath102 for a sequence of 64 single pulses from b0834 + 06 .
rw07 admitted that they also found in their data some sequences showing @xmath103 periodicity , which seemed to be a sub - harmonic of @xmath16 cycle .
we noticed yet another problem with the result of ad05 . in our opinion , these authors have used incorrectly their equations ( 2 ) and ( 3 ) . in fact , as @xmath104 they used the longitudinal distance between the profile components and in consequence , the azimuthal magnetic angle between the neighboring subbeams was @xmath105 degs .
this ignored the subpulses appearing in the saddle of the profile .
we believe that they should use @xmath106 degs , and as a result , the number of sparks would be @xmath107 instead 8 .
this is consisted with @xmath108 obtained by rw07 and supported by our results presented in this paper . in summary , we strongly believe that the actual value of @xmath7 in psr b0834 + 06 is close to 30 pulsar periods and that @xmath109 corresponds to a first harmonic of the basic cycle . some evidence of low frequency spectral features at both @xmath110 and @xmath111 can be seen in figure a19 of w06 .
moreover , it seems that 14 sparks inferred by rw07 are more likely than 8 sparks inferred by ad05 .
the essence of the psg pulsar model is the presence of a strong , nondipolar surface magnetic field @xmath112 , although it does not appear explicitly in equation ( 1 ) ; see appendix a for details .
the strong value of @xmath112 is necessary for providing enough binding ( cohesive energy ) to prevent the free flow of iron ions from the surface ( medin & lai 2007 ; ml07 hereafter ) , while the small radius of curvature is needed to develop cascading pair production ( e.g. gil & melikidze 2002 ) .
the latter phenomenon is essential for both shorting out the gap potential drop and providing a dense electron - positron plasma in the radio emission region ( eg .
melikidze & gil , 2000 and gil , lyubarski & melikidze , 2004 ) . when the calculations of ml07 are adapted to the psg model
, then one can derive the dependence of the surface magnetic field on the surface temperature @xmath113 ; ( we will give detailed description of this topic in a separate paper , but see appendix a for some details ) .
for the condensed fe surface this relationship is represented by the solid red line in figure 7 of ml07 .
we can apply this apparatus to our case of psr b0834 + 06 , with @xmath114 erg @xmath115 , @xmath116 k , and the associated effective surface area of the hot spot @xmath117 m@xmath77 . on the other hand
, one can read off from figure 7 in ml07 the range of values @xmath118 g corresponding to @xmath119 k. since the dipolar surface magnetic field and polar cap area are @xmath120 g and @xmath121 m@xmath77 , respectively , we can find the effective surface area @xmath122 m@xmath77 .
this is consistent with our estimate , in which @xmath123 is about 50 smaller than @xmath124 .
theoretically , this results naturally from the flux conservation of the open magnetic field lines . as pointed out in paper ii ( see also references therein ) , the small size of the hot spot relative to the canonical polar cap area is a typical property of hot bb thermal radiation detected in a number of pulsars .
an extreme case was published just recently by pavlov , kargaltsev , wong et al .
( 2008 ; p08 hereafter ) , who reported on the chandra detection of a very old ( 170 myr ) and close to the earth ( 0.13 kpc and 0.184@xmath125 kpc , according to atnf ( manchester , hobbs , teoh et al .
2005 ) and ne2001 ( cordes & lazio 2003 ) database , respectively ) radio pulsar psr j0108 - 1431 , with a very weak dipolar surface magnetic field @xmath126 g and a low spindown @xmath127 erg @xmath115 . during 30 ks exposure they detected 53 counts and found that the spectrum can be described by pl model or bb model equally well . for the latter model they obtained the bolometric luminosity @xmath128 erg @xmath115 ,
@xmath129 k and @xmath130 m@xmath77 , which translates into the hot spot radius as small as 4 meters .
this is the smallest hot polar cap ever observed@xmath131 , with the ratio @xmath132 , equal to 1770 or 923 ( highest ever obtained ) for distances 0.13 and 0.18 kpc , respectively .
accordingly , the actual surface magnetic field @xmath133 ( see gil & sendyk 2000 and ml07 ) is equal to 4.5 or 2.3 @xmath134 g for a distance of 0.13 or 0.18 kpc , respectively .
interestingly , the latter value agrees almost exactly with ml07 ( red solid line in their fig .
7 ) , while the former one implies too high a surface temperature exceeding 5 mk .
thus , the extremely small hot polar cap with @xmath1353.2 mk results from the fact that the actual surface magnetic field must be about 1000 times stronger than the dipolar component , in order to provide enough cohesive energy to develop psg in this pulsar .
we can therefore say that the case of psr j0108 - 1331 supports strongly the psg pulsar model , the ml07 cohesive energy calculations for the condensed fe polar cap surface and ne2001 distance to this pulsar ( about 0.184 kpc ) .
if one adopts 0.184 kpc as the proper distance to psr j0108 - 1331 , then the bolometric bb luminosity is @xmath136 erg @xmath115 and the efficiency @xmath137 . with this value
the equation ( 1 ) predicts the tertiary periodicity @xmath138 12 .
however , the confirmation of this by means of single pulse radio observations of psr j0108 - 1431 seems hopeless with present day possibilities , as the pulsar is also extremely weak in radio band ( tauris , nicastro , johnston et al .
1994 ) .
thus , our psg model seems to account for the physical phenomena at and above the actual pulsar polar cap quite well .
other available inner acceleration models do not match the observations well .
the pure vacuum gap model ( ruderman & sutherland 1975 ) has @xmath139 .
although it also satisfies eq.(1 ) , it predicts a very high polar cap heating rate , typically @xmath140 ( zhang et al .
2000 ) , and therefore a very small @xmath7 .
the predicted high @xmath6 has been ruled out by the x - ray observations of many old pulsars ( zsp05 , to05 , k06 and this paper ) , and the predicted low @xmath7 is also inconsistent with the radio observations . on the other hand ,
as discussed in 1 the steady - state sclf model does not predict the existence of the `` sparks '' whose drifts around the polar cap region provide the most natural interpretation of the observed drifting sub - pulse patterns . a modified unsteady sclf model ( which has not been discussed in the literature )
may be able to introduce a sparking - like behavior .
based on the similar logic ( i.e. the potential drop along the magnetic field line in the gap is equal to the horizontal potential drop across the spark , see appendix ) , a similar equation as eq.(1 ) can be derived for the sclf model . however , since this model introduces a very small effective @xmath141 value ( @xmath142 , harding & muslimov 2001 ) , the predicted polar cap heating rate is too low to interpret the observations , typically @xmath143 ( harding & muslimov 2002 ) .
also the corresponding drifting velocity is too small so that the predicted @xmath7 is too long as compared with the radio data .
the psg model predicts an intermediate particle inflow rate , and gives the clean prediction ( eq.[1 ] ) which allows @xmath6 to be a moderate value .
this is strongly supported by the data . in order to solve the binding energy problem in the canonical dipolar magnetic field at the neutron star surface
, it has been conjectured that drifting subpulse pulsars are bare strange stars ( xu et al .
the simplest model does not allow a hot polar cap because of the high thermal conductivity of the bare strange star surface layer , which is ruled out by the data .
yue et al .
( 2006 ) argued that psr b0943@xmath010 may be a low mass quark star ( @xmath144 ) .
however , pulsar drifting seems to be the most common behavior of radio pulsars ( w06a , b ) , some of which have well measured mass around @xmath145 ( thorsett & chakrabarty 1999 ) .
we regard that the quark star scenario is no longer attractive in view of the latest observations .
the cohesive energy calculations of fe ion chains in ultra - strong magnetic field by ml07 seem to be strongly supported by the x - ray observations discussed in this paper .
finally , we would like to address a hypotheses put forward by becker , kramer & jessner et al .
( 2006 ) that in old pulsars ( @xmath146 yrs ) the magnetospheric emission dominates over thermal emission , including both cooling radiation and hot polar cap emission component .
these authors suggested that the latter radiation component decreases along with the former one , and if so , the hot polar caps in cooling neutron stars could be formed by anisotropic heat flow due to the presence of the magnetic field rather than by particle bombardment .
while in young nss with core temperature @xmath147 k the strong crustal magnetic fields may channel the heat toward the polar cap resulting in @xmath148 of a few mk ( perez - azorin , miralles & pons 2006 ; geppert , kker & page 2006 ) , in pulsars older than @xmath149 years this mechanism is much less efficient and the only viable process that can produce such hot and small polar caps is the back - flow particle bombardment .
almost all pulsars presented and examined in this paper are older than 1 myr ( an exception is 110 kyr psr b0656 + 14 ) .
for instance , psr b0834 + 06 is 3 myr old and its x - ray emission is dominated by hot bb component ( an obvious counter - example arguing against becker s claim ) . in psr b1929 + 10 ( 3.1 myr old )
the luminosity of hot bb component is at least comparable with the magnetospheric x - ray radiation ( m07 ) .
the very old ( 170 myr ) rotation powered non - recycled pulsar j0108 - 1431 clearly shows bb radiation from the hot polar cap ( p08 ) , probably accompanied by the magnetospheric emission , but no evidence of cooling radiation from the whole surface , as expected for such an old pulsar . in summary , both the polar cap full cascade ( zhang & harding 2000 ) and the downward outer gap cascade ( cheng , gil & zhang 1998 ) that have been proposed to interpret non - thermal x - ray emission from spindown - powered pulsars are expected to be less significant in pulsars from our sample with respect to the young pulsars .
the predicted values of x - ray luminosity in these models are typically lower than that of the polar cap heating in the psg model ( eq.[1 ] ) . in view that other available models of the pulsar inner accelerator ( pure vacuum gap model and space - charge - limited flow model ) either overpredict or underpredict the polar cap heating level , we conclude that the pulsar inner accelerator is likely partially screened due to a self - regulated sub - goldreich - julian flow . also , the pure vacuum gap model predicts too fast a drifting and the space - charge - limited flow model has no natural explanation for the subpulse drift phenomenon at all .
we thus strongly believe that thermal radiation associated with a polar cap heating due to partially screened inner accelerator ( psg ) is a common component of pulsar x - ray emission regardless of its age , and this component plays especially significant role in the spectra of old pulsars .
our results are partly based on observations with xmm - newton , an esa science mission with instruments and contributions directly funded by esa member states and the usa ( nasa ) .
we acknowledge the support of the nasa nnx07af07 g and nnx08ac67 g grants .
jg was partially supported by the polish state committee for scientific research grant n n203 2738 33 and gm was partially supported by the polish state committee for scientific research grant n n203 1262 33 , as well as by the georgian nsf st06/4 - 096 and intas 06 - 1000017 - 9258 grants .
the project is supported by the bundesministerium fr wirtschaft und technologie / deutsches zentrum fr luft- und raumfahrt ( bmwi / dlr , fkz 50 ox 0001 ) and the max - planck society .
we thank dr .
dipanjan mitra for stimulating discussions , critical reading of the manuscript and helpful comments .
the charge depleted inner acceleration region above the polar cap results from the deviation of a local charge density @xmath150 from the co - rotational charge density ( goldreich & julian 1969 ) @xmath151 . for isolated neutron stars
one might expect the surface to consist mainly of the iron formed at the neutron star s birth ( e.g. lai 2001 ) .
therefore , the charge depletion above the polar cap can result from binding of the positive @xmath152fe ions ( at least partially ) in the neutron star surface .
if this is really possible ( see medin & lai 2006 , 2007 and paper ii for details ) , then the positive charges can not be supplied at the rate that would compensate the inertial outflow through the light cylinder . as a result , a significant part of the unipolar potential drop develops above the polar cap , which can accelerate positrons to relativistic energies and power the pulsar radiation mechanism , while the electrons would bombard the polar cap surface , causing a thermal ejection of ions , which are otherwise more likely bound in the surface in the absence of additional heating .
this thermal ejection would cause partial screening of the acceleration potential drop @xmath153 corresponding to a shielding factor @xmath154 ( see gmg03 for details ) , where @xmath155 is the charge density of the ejected ions , @xmath156 is the potential drop and @xmath157 is the height of the acceleration region . the gap potential drop is completely screened when the total charge density @xmath158 reaches the co - rotational value @xmath159 . in terms of binding of @xmath152fe ions ,
the screening factor @xmath160 , @xmath161 is the cohesive energy of the condensed iron surface , @xmath148 is the actual surface temperature , @xmath162 is the critical temperature above which the iron ions are ejected with the maximum co - rotation limited rate , and @xmath163 ( medin & lai 2007 ) . because of the exponential sensitivity of the accelerating potential drop to the surface temperature , the actual potential drop should be thermostatically regulated .
in fact , when the potential drop is large enough to ignite the cascading pair production , the back - flowing relativistic charges will bombard the polar cap surface and heat it at a predictable rate .
this heating will induce thermionic emission from the surface , which will , in turn , decrease the potential drop that caused the thermionic emission in the first place . as a result of these two oppositely directed tendencies , the quasi - equilibrium state should be established , in which heating due to electron bombardment is balanced by cooling due to thermal radiation .
this should occur at a temperature slightly lower than the critical temperature above which the polar cap surface delivers thermionic flow at the corotational charge density level .
this is an essence of the psg model . for practical reasons
it is assumed that @xmath164 , while in reality @xmath148 is few thousands k lower than @xmath165 , with the latter being strongly dependent on the surface magnetic field @xmath112 .
this is illustrated by figure 7 in medin & lai ( 2007 ) , which was prepared for the pure vg model .
the psg model is realized along the red ( for fe ) line in this figure , which shows that for a few mk surface temperatures , as suggested by x - ray observations of pulsar hot spots ( see paper ii and references therein ) the surface magnetic field must be close to @xmath166 g in all pulsars . for most pulsars
this is a much stronger field than that inferred from pulsar spindown due to the magnetic dipole radiation .
therefore , the surface magnetic field in neutron stars must be dominated by crust anchored non - dipolar magnetic anomalies .
such strong and curved surface magnetic field is also necessary for development of the cascading pair production via curvature radiation ( e.g. rs75 , gil & melikidze 2002 ) .
several models proposed for generating pulsar radio emission based on the concept of vacuum gaps need radius of curvature of surface magnetic field much smaller than the stellar radius ( see for e.g. gil , melikidze , mitra 2002 ) .
a possibility of generating such fields would be from currents in the neutron stars crust ( e.g. urpin , levshakov & iakovlev 1986 , geppert , rheinhardt & gil 2003 ) .
mitra , konar , & bhattacharya ( 1999 ) examined the evolution of multipole components generated by currents in the outer crust .
they found that mostly low - order multipoles contribute to the required small radii of curvature and that the structure of the surface magnetic field is not expected to change significantly during the radio pulsar lifetime .
the spark plasma inside psg must slowly drift with respect to the polar cap surface due to non - corotational charge density .
this drift will manifest itself by the observed subpulse drifting , provided the spark arrangement is quasi - stable over time scales of hundreds of pulses or so .
the deviation of the charge density from the co - rotational value generates an electric field @xmath167 just above the polar cap surface .
the parallel component causes acceleration of charged particles , while the perpendicular component participates in the subpulse drift . the tangent electric field at the polar cap boundary , @xmath168 ( see appendix a in gmg03 for details ) .
due to the @xmath169 drift the discharge plasma performs a slow circumferential motion around the magnetic axis ( see the next paragraph below ) with velocity @xmath170 . the time interval
to make one full revolution around the polar cap boundary is @xmath171 .
one then has = 2 , [ p3p ] where the coefficient @xmath172 should be close to unity .
if the plasma above the polar cap is fragmented into filaments ( sparks ) , which determine the intensity structure of the instantaneous pulsar radio beam , then in principle , the circulational periodicity @xmath7 can be measured / estimated from the pattern of the observed drifting subpulses ( deshpande & rankin 1999 , gil & sendyk 2003 ) . in practice , @xmath7 is measured from the low frequency features in the modulation spectra obtained from good quality single pulse data of pulsars with drifting subpulses .
according to rs75 , @xmath173 , where @xmath174 is the number of sparks contributing to the drifting subpulse pattern observed in a given pulsar and @xmath9 is the primary drift periodicity ( distance between the observed non - aliased subpulse drift bands ) .
the circumferential motion around the magnetic axis like in rs75 holds only when the magnetic and the spin axes are almost parallel ( almost aligned rotator , in which the line - of - sight trajectory is almost the circumferential tracks of sparks moving around the magnetic axis ) .
many pulsars with drifting subpulses have indeed a very broad profile characteristic of the almost aligned rotators : e.g. b0826 - 34 , b0818 - 41
. others , which are not a broad profile pulsars and show regular drifting must have very high impact angle , i.e. grazing the emission beam . in such cases
one can not exclude the almost aligned geometry . in more general ( inclined ) case
, the spark trajectory does not have to be closed on the polar cap , as sparks should rather follow the trajectory of the line - of - sight projected onto the polar cap , being slightly late behind the star s rotation .
however , observations of drifting subpulses in some pulsars do not support such a scenario , being consistent with the circumferential motion of the spark - associated sub - beams of subpulse radiation , even if pulsar is not an aligned rotator .
indeed , an orderly drifting subpulses always demonstrate a systematic intensity modulation , either increasing or decreasing towards the pulse profile midpoint . also , in pulsars with more central cut of the line - of - sight trajectory the subpulse drift is less apparent ( or none ) but a characteristic phase - stationary modulation of subpulse intensity modulation persist .
these properties strongly suggest that sparks move on closed trajectories on the polar cap , although they do not have to be circular , like in axially symmetric rs75 model , to the extent that in some of the detection of circumferential motion with specified value of @xmath7 periodicity is possible .
a good example of such pulsar with central light - of - sight cut is b0834 + 06 discussed in this paper .
there must be then some agency that makes sparks moving across the the line - of - sight projection on closed trajectories around the local magnetic pole instead around the rotational pole , irrespective of the inclination and impact angles .
the quasi - equilibrium condition is @xmath175 , where @xmath176 is the cooling power surface density by thermal radiation from the polar cap surface and @xmath177 is the heating power surface density due to back - flow bombardment , @xmath178 is the lorentz factor , @xmath179 is the number density of the back - flowing particles that deposit their kinetic energy at the polar cap surface , @xmath141 is the shielding factor , @xmath180 is the charge number density of the thermionic ions and @xmath181 is the corotational charge number density and @xmath182 is the time derivative of the period in @xmath183 .
it is straightforward to obtain an expression for the quasi - equilibrium surface temperature in the form @xmath184 ( paper ii ) , where @xmath185 , the parameter @xmath186 ( gil & sendyk 2000 , medin & lai 2007 ) describes the domination of the local actual surface magnetic field over the canonical dipolar component at the polar cap , and @xmath182 is the normalized period derivative . here
@xmath187 and @xmath188 is the canonical ( rs75 ) and actual emitting surface area , with @xmath189 and @xmath190 being the canonical ( rs75 ) and the actual polar cap radius , respectively . since the typical polar cap temperature is @xmath191 k ( paper ii )
, the actual value of @xmath192 must be much larger than unity , as expected for the highly non - dipolar surface magnetic fields .
using equation ( [ p3p ] ) one can derive the formula for thermal x - ray luminosity as l_b=2.510 ^ 31 ^ -2()()^-2 [ lx ] , or in the simpler form representing the radiation efficiency with respect to the spin - down power @xmath193 erg / s , where @xmath194 g cm@xmath77 is the neutron star moment of inertia and @xmath196 ( see papers i and ii for details ) = 0.63 ( ) ( ) ^-2 [ lx ] .
this equation is very useful for a direct comparison with the observations , since it contains only the observed quantities ( although it is subject to small uncertainty factors related to the unknown moment of inertia @xmath197 and the coefficient @xmath198 .
it does not depend on any details of the sparking gap model like non - dipolar surface magnetic field @xmath199 , the height @xmath157 of the acceleration region and the shielding factor @xmath141 , since they cancel in the derivation procedure , as they suppose to do so .
indeed , this equation reflects the fact that both the subpulse drifting rate ( due to @xmath169 plasma drift ) and the polar cap heating rate ( due to back - flow bombardment ) are determined by the same physical quantity , which is the potential drop across the inner acceleration region just above the polar cap .
no other agency should be involved . in practical application of equation ( a3 ) we will set @xmath200 and @xmath201 .
the former is commonly used and the latter means that the values of the accelerating @xmath202 and perpendicular @xmath203 components of electric field in the psg are almost the same .
it is quite a reasonable assumption , all the more that it seems to be supported observationally ( fig .
1 ) . arons j. , sharlemann e.t .
1979 , apj , 231 , 854 asgekar , a. , & deshpande , a.a .
2001 , , 326 , 1249 becker , w. , kramer , w. , jessner , a. , et al .
2006 , apj , 645 , 1421 biggs , j. d. , 1990 , , 246 , 341 ( b90 ) cheng , k. s. , gil , j. , & zhang , l. 1998 , , 493 , l35 cordes , j.m . , & lazio , t.j.w . 2002 , astro - ph/0207156 de luca , a. , caraveo , p. a. , mereghetti , s. , et al .
2005 , , 623 , 1051 ( dl05 ) deshpande , a.a . , & rankin , j.m .
1999 , , 524 , 1008 ( dr99 ) deshpande , a.a .
, & rankin , j.m .
2001 , , 322 , 438 edwards , r.t . , & stappers , b.w .
2002 , , 393 , 733 ( es02 ) edwards , r.t . , & stappers , b.w .
2003 , , 407 , 273 ( es03 ) geppert , u. , rheinhardt , m. , & gil , j. , 2003 , a&a , 412 , l33 geppert , u. , kker , m . , & page , d. 2006 , , 457 , 937 gil , j. , & melikidze , g. i. 2000 , , 544 , 1081 gil , j. , & sendyk , m. , 2000 , , 541 , 351 gil j. , & melikidze , g.i . , 2002 , , 577 , 909 gil , j. a. , melikidze , g. i. and mitra , d. , 2002 , aap , 388 , 235 g gil j. , & sendyk , m. 2003 , , 585 , 453 gil , j. , melikidze , g.i . , & geppert , u. 2003 , , 407 , 315 ( g03 ) gil j. , lyubarsky , y. , & melikidze , g.i . 2004 , , 600 , 872 gil j. , melikidze , g. , & zhang , b. 2006a , , 650 , 1048 gil j. , melikidze g. , & zhang , b. 2006b , , 457 , 5 ( paper i ) gil j. , melikidze g. , & zhang , b. 2007 , , 376 , l67 ( paper ii ) mitra , d. , konar , s. , bhattacharya , d , 1999 , , 307 , 459 goldreich p. , & julian , h. 1969 , , 157 , 869 gupta , y. , gil j. , kijak , j. , & sendyk , m. 2004 , , 426 , 229 harding , a. k. , & muslimov , a. g. 1998 , , 508 , 328 harding , a. k. , & muslimov , a. g. 2001 , , 556 , 987 harding , a. k. , & muslimov , a. g. 2002 , , 568 , 862 herfindal , j.l . , & rankin , j.m .
2007 , , 380 , 430 ( hr07 ) hui , c.y . , & becker , w. 2008 , astro - ph/07070800 ( hb08 ) jansen , f. , lumb , d. , altieri , b. , et al .
2001 , , 365 , l1 kargaltsev o. , pavlov g.g . , &
garmire , g.p .
2006 , , 636 , 406 ( k06 ) lai , d. 2001 , rev .
phys . , 73 , 629 manchester , r.n . , hobbs , g.b . , teoh , a. , & hobbs , m. 2005 , , 129 , 1993 misanovic , z. , pavlov , g.g . , & garmire , g.p .
2007 , astro - ph/0711417 ( m07 ) medin z. , & lai , d. 2007 , , 382 , 1833 ( ml07 ) medin , z. , & lai , d. 2006 , phys .
rev . a , 74 , 062508 perez - azorin , j.f . ,
mirales , j.a . , & pons , j.a .
2006 , , 451 , 1009 posselt , b. , popov , s. b. , haberl , f. , et al .
2008 , ( in press ) rankin , j.m .
, & suleymanova , s.a .
2006 , , 453 , 679 rankin , j.m . , & wright , g.a.e .
2007 , , 379 , 507 ( rw07 ) ruderman , m.a . , & sutherland , p.g . 1975 , ,
196 , 51 ( rs75 ) strder , l. , briel , u. , dennerl , k. , et al .
2001 , , 365 , l18 tauris , t.m .
, nicastro l. , johnston , s. , et al . 1994 , , 428 , l53 tepedelenliolu e. , & gelman , h. 2005 , , 630 , 57 ( t05 ) turner , m. j. l. , abbey , a. , arnaud , m. , et al . 2001 , , 365 , l27 thorsett , s. e. , & chakrabarty , d. 1999 , , 512 , 288 urpin , v. a. ; levshakov , s. a. ; iakovlev , d. g. , 1986 , mnras , 219,703 weltevrede , p. , edwards , r.i . , & stappers , b.w .
2006a , , 445 , 243 ( w06a ) weltevrede , p. , wright , g.a.e . , stappers b.w . , & rankin , j. m. 2006b , , 459 , 597 ( w06b ) wilms , j. , allen , a. , & mccray , r. 2000 , , 542 , 914 xu , r. x. , qiao , g. j. , & zhang , b. 1999 , , 522 , l109 yue , y. l. , cui , x. h. , & xu , r. x. 2006 , , 649 , l95 zhang , b. , harding a. , & muslimov a. 2000 , , 531 , l135 zhang , b. , & harding a.k .
2000 , , 532 , 1150 zhang , b. , sanwal d. , & pavlov , g.g . 2005 , , 624 , l109 ( z05 ) b0943@xmath010 & 1.09 & 3.49 & @xmath204 & @xmath205 & 1 & @xmath206 & 8&@xmath207 & @xmath208 + b1113@xmath016 & 1.19 & 3.73 & @xmath209 & @xmath210 & 2 & & & & + & & & & @xmath211 & 3 & @xmath212 & 9&@xmath213 & @xmath214 + b0834@xmath006 & 1.27 & 6.8 & @xmath215 & @xmath216 & 4 & @xmath217 & 5&@xmath218 & @xmath219 + b1929@xmath010 & 0.23 & 1.16 & @xmath220 & @xmath221 & 5 & @xmath222 & 10&@xmath223 & @xmath224 + b0656@xmath014 & 0.38 & 55.0 & @xmath225 & @xmath226 & 6 & @xmath227 & 11&@xmath228 & @xmath229 + b1055@xmath152 & 0.19 & 5.8 & @xmath230 & @xmath231&13 & @xmath232 & 11&@xmath233 & @xmath234 + b0628@xmath128 & 1.24 & 7.12 & @xmath235 & @xmath236 & 6 & @xmath237 & 12&@xmath238 & @xmath239 + b0826@xmath134 & 1.85 & 0.99 & @xmath240 & @xmath241 & 7&@xmath242 & 5&@xmath243 & + + 08 28 16.6 & -34 17 07 & 1269 & pn & 2006 - 11 - 13 13:44:24 & 2006 - 11 - 14 09:19:30 & 38.83 + & & & m1 & 13:22:03 & 09:19:35 & @xmath1 + & & & m2 & 13:22:03 & 09:19:50 & @xmath1 + + 08 37 05.6 & 06 10 15 & 1454 & pn & 2007 - 11 - 17 13:44:24 & 2007 - 11 - 18 09:19:30 & 48.95
+ & & & m1 & 13:22:03 & 09:19:35 & 53.30 + & & & m2 & 13:22:03 & 09:19:50 & 54.44 + versus circulation period @xmath7 of drifting subpulses in the radio band .
the solid curve represents the prediction of the psg model ( eq . [ 1 ] ) , while the dotted curves correspond to uncertainties in determining the moment of inertia ( see appendix a ) .
the values of @xmath7 and @xmath6 along with their error bars ( 2@xmath28 ) and references for the data are given in table 1 . ] and photon index @xmath36 compared to the measured values from the epic - pn data of psr b0834 + 06 .
the cross and box drawn with full lines indicate 1@xmath28 and dotted lines 2@xmath28 confidence regions .
* b. * as in figure 2a but for an absorbed blackbody model with @xmath49 ranging from 80 to 480 ev , with a step of 20 ev . * c. * as in figure 2b but for a model with blackbody and power - law component .
both components are absorbed by the same n@xmath33 and the relative ( [email protected] kev ) flux ratio is 1:0.5 , respectively . *
d. * as in figure 2c but for a flux ratio of 1:1 . ] and its efficiency @xmath89 versus the polar cap temperature @xmath49 for b0834 + 06 derived by means of xspec spectral modelling for color marked and numbered points in figures 2b-2d .
the large red circle corresponds to the best fit of the bb model and the error bars include the model uncertainties . ] | we report the x - ray observations of two radio pulsars with drifting subpulses : b0834@xmath006 and b0826@xmath134 using .
psr b0834@xmath006 was detected with a total of 70 counts from the three epic instruments over 50 ks exposure time .
its spectrum was best described as that of a blackbody ( bb ) with temperature @xmath2 k and bolometric luminosity of @xmath3 erg s@xmath4 . as it is typical in pulsars with bb thermal components in their x - ray spectra ,
the hot spot surface area is much smaller than that of the canonical polar cap , implying a non - dipolar surface magnetic field much stronger than the dipolar component derived from the pulsar spin - down ( in this case about 50 times smaller and stronger , respectively ) .
the second pulsar psr b0826@xmath134 was not detected over 50 ks exposure time , giving an upper limit for the bolometric luminosity @xmath5 erg s@xmath4 .
we use these data as well as the radio emission data concerned with drifting subpulses to test the partially screened gap ( psg ) model of the inner accelerator in pulsars .
this model predicts a simple and very intuitive relationship between the polar cap thermal x - ray luminosity ( @xmath6 ) and the `` carousel '' period ( @xmath7 ) for drifting subpulses detected in the radio band .
the psg model has been previously successfully confronted with four radio pulsars whose @xmath6 and @xmath7 were both measured : psr b0943@xmath010 , psr b1133@xmath016 , psr b0656@xmath014 , and psr b0628@xmath128 .
the x - ray data of psr b0834@xmath016 reported here are also in agreement with the model prediction , and the upper limit derived from the psr b0826@xmath134 observation does not contradict with such a prediction .
we also include two other pulsars psr b1929@xmath010 and b1055@xmath8 whose @xmath6 and/or @xmath7 data became available just recently .
these pulsars also follow the prediction of the psg model .
the clear prediction of the psg model is now supported by all pulsars whose @xmath6 and @xmath7 are measured and/or estimated .
[ firstpage ] |
psr j1741@xmath12054 , discovered in a blind pulsation search of a _ fermi _ point source @xcite , is a @xmath8ms , characteristic age @xmath9kyr @xmath2-ray pulsar . the pulsar was then detected in archival parkes radio observations and subsequently studied at gbt @xcite , showing it to be very faint and deeply scintillating .
the radio observations also gave a remarkably low dispersion measure @xmath10 , which for a standard galactic electron density model @xcite implies a distance of @xmath11kpc .
this makes this one of the closest energetic pulsars known . at this distance
the low flux density observed at pulsar discovery , @xmath12jy , also makes this the least luminous radio pulsar known . indeed , since the source scintillates to undetectability at many epochs , the time average radio luminosity is @xmath13 .
the low distance was already suspected from the relatively high @xmath2-ray flux ( for the modest @xmath14 spin - down luminosity ) . at the @xmath15-estimated distance ,
the pulsar would emit 28% of its spin - down power in @xmath2-rays , if the radiation were isotropic .
since psr j1741@xmath12054 is an important addition to the set of nearby , energetic young pulsars , further investigation of its energy loss and environs are needed . in particular , we wish to constrain its motion , true age , non - photon spindown deposition and interaction with its environment .
for example , it may be a significant source of @xmath16 injection in the solar neighborhood .
such injection is generally visible as a pulsar wind nebula ( pwn ) .
the shape of this nebula can constrain the source s distance , ism interaction , and proper motion .
if the spin orientation can be measured @xcite , then we can also model the @xmath2-ray beam shape and connect the observed @xmath2-ray flux to its true luminosity , predicting @xmath17 @xcite .
this , in turn , provides the @xmath2-ray efficiency , a critical test of pulsar models .
we accordingly searched for a pwn , discovering an h@xmath0 bow shock and an x - ray pwn and trail .
like the few other pulsar bow shock nebulae , this is a non - radiative , balmer - dominated shock .
the shock velocity is relatively low , and doppler broadened post - shock emission traces details of its flow , presenting an opportunity for the study of collisionless shock structure under unusual conditions , as well as constraining properties of the pulsar wind .
we describe initial observations and conclusions here .
additional discussion of the x - ray pwn and of the spectral properties of the pulsar point source are presented in @xcite .
6.0truecm -1truecm
our initial detection of the psr j1741@xmath12054 pwn was made in a 600s exposure using the wiyn 3.6 m telescope and the minimo camera with the kp w012 ( @xmath18 ) narrow - band h@xmath0 filter on march 24 , 2009 .
an additional 4@xmath19s exposure plus a continuum frame were obtained the next night . unfortunately all data suffered from poor ( [email protected]@xmath20 ) seeing .
nevertheless , these data revealed a clear elliptical ( @xmath21 ) edge - brightened shell .
as this is close to the galactic bulge ( @xmath22 ) , the continuum frame was crowded with field stars .
figure 1 shows the h@xmath0 image , without continuum subtraction .
we were able to use the spectroscopic flux calibration ( see below ) to estimate the total ( continuum - subtracted ) h@xmath0 flux of the nebula as @xmath23 .
we subsequently ( aug 21 , 2009 ) obtained 3@xmath24600s exposure using the efosc2 camera on the ntt 3.6 m through the h@xmath0#692 filter . while the image quality was somewhat better ( @xmath25 ) , the wider ( @xmath26 ) transmission band and lower peak throughput gave the h@xmath0 nebula sharper edge structure , but lower contrast ( figure [ nttim ] ) . on this image
we show the position of the _ cxo _ point source and the locations of the @xmath27 slits used for the spectroscopic study .
5.9truecm -1truecm we obtained a _ cxo_/acis - s observation ( obsid 11251 ) of the pulsar field on may 21 , 2010 ( = mjd 55337.1 ) , with 48.8ks of exposure .
the pulsar was placed near optimum focus on the backside illuminated s3 chip and the ccds were operated in half - frame vfaint mode , reading out every 1.64s to reduce pileup . a bright point source and compact nebula are seen , locating the pulsar to 17h 41 m 57.28s , @xmath120@xmath28 54@xmath29 11.8@xmath20 ( @xmath30 ) .
the crowded optical field made it difficult to identify field x - ray sources .
however , over the s3 half - chip several soft ( coronal emission ) sources were clearly identified with field stars , confirming the x - ray / optical frame tie at the @xmath31 level .
there is diffuse , faint x - ray emission in a pwn trail extending some @xmath32 at position angle pa=@xmath33 ( n through e ) .
this structure contains @xmath34 counts . from 2@xmath35 from the pulsar
the trail contains an additional 92@xmath36 counts .
finally , the compact core contains @xmath37 counts in a @xmath38 radius region about the pulsar , dominated by the point source .
the emission from these structures is consistent with a simple absorbed power law with spectral index @xmath39 and absorbing column @xmath40 .
the point source clearly shows an additional soft thermal component , and the detailed spectral measurements of these structures are discussed in @xcite .
here we concentrate on the pwn morphology .
the region around the point source appears slightly extended . after subtracting
a psf generated with marx , shifted and scaled to best match the point source component , the excess appears dominated by a diagonal band , @xmath41 , with a minor axis at pa=@xmath42 .
this structure appears to contain @xmath43 counts or 11% of the total counts in this region .
as the minor axis pa is within 5@xmath28 of the large scale x - ray trail axis ( figure [ psfsub ] ) , we associate this structure with the equatorial torus of the pwn . the 3:1 aspect ratio suggests a planar structure seen edge on , and implies that the pulsar spin axis is close to the plane of the sky .
5.5truecm -0.5truecm pulsar h@xmath0 bow shock nebulae are relatively rare , since the pulsar wind must impact a partly neutral medium .
typically , this means that the pulsar has escaped its supernova remnant birthsite and is interacting with the general ism @xcite .
the shock spectra , however , are particularly interesting as the shocked gas is advected downstream on a time short compared to the cooling time .
this means that these are non - radiative shocks and are dominated by balmer emission rather than the usual forbidden lines @xcite . when the up - stream neutral atoms , which are unaffected by the fields in the shock , drift into the shocked ism gas they can be excited , producing line radiation at the ( cold ) radial velocity of the ambient medium gas . for typical conditions
the ratio of excitation to ionization rates is @xmath44 so that we obtain an ism - velocity balmer photon for every @xmath45 neutral atoms .
however , in addition , the atoms can charge - exchange with the hot shocked ions .
this gives rise to excited neutrals in the post - shock flow and a broad line component at the systemic radial velocity of the post - shock gas .
thus spatially - resolved , kinematically - resolved observations of the line emission can provide a wealth of information on the local ism and bow - shock kinematics @xcite .
we were able to obtain two @xmath27 long - slit spectra of the psr j1741@xmath12054 bow shock , using the keck i / lris on june ll 2010 ( mjd 55358.5 ) .
a total of 900s exposure was taken with the slit at @xmath46 ( along the pwn symmetry axis ) and 1200s total exposure with the slit at @xmath47 . both exposures covered the pulsar position . unfortunately , since only one amplifier was available in the lris red camera , the long slit was truncated .
however , we managed to cover the pwn emission ; the slit locations were accurately determined from recorded slit images and are shown on figure 2 .
7.2truecm for these observations , a 5600 dichroic separated the red and blue beams .
the blue camera employed a 600l / mm , 4000 blaze grating and covered 3100 - 5600@xmath48 at a resolution of @xmath49/pixel . for the bow shock exposures ,
the red camera employed the 1200l / mm , 7500 grating for a resolution of @xmath50/pixel . with
the 1@xmath20 slit the achieved resolution as measured from the night sky lines near h@xmath0 was 1.48 or 68@xmath51 ( fwhm ) .
the spatial resolution along the slit was 0.24@xmath20/pixel .
for the blue channel , we were able to employ standard calibrations .
unfortunately the red channel configuration was non - standard for the observing run and a flux standard was not obtained in this configuration .
however , we were able to use flux measurements at a lower resolution setting along with measured grating responses to recover the red flux calibration with a conservative estimate of 15% accuracy .
the wavelength calibration was refined by reference to the night sky lines and line velocities have been converted to the local standard of rest ( lsr ) using the @xcite constants .
figure [ bsspec ] shows the full spectrum of the bright limb in front of the pulsar from the @xmath52 observation .
the strong dominance of the balmer emission is clear , although a few of the brightest forbidden lines are also visible .
these lines were detectable only in the brightest portions of the shock limb , so we have not been able trace this component separately . table 1 contains measurements of the line fluxes . lcrl[h ! ! ]
h@xmath0 & 6563 & 70.2@xmath53 & @xmath54 + h@xmath55 & 4861 & 17.6@xmath56 & + h@xmath2 & 4340 & 5.2@xmath57 & + h@xmath58 & 4102 & 1.3@xmath56 & + n[ii ] & 6583 & 4.0@xmath59 & @xmath54 + n[ii ] & 6549 & 1.0@xmath60 & + o[i ] & 6300 & 1.6@xmath61 & + o[i ] & 6363 & 0.3@xmath62 & + o[ii ] & 3727 & 4.8@xmath56 & + s[ii ] & 6731 & 1.0@xmath62 & + s[ii ] & 6716 & 1.2@xmath61 & + our balmer measurements give h@xmath0/h@xmath63 and h@xmath2/h@xmath64 , and we can use these flux ratios to estimate the extinction toward the nebula . the selective extinction is @xmath65 where the models values ( e.g. @xmath66 ) depend on the shock populations and geometry , especially the shock optical depth of ly@xmath55 .
@xcite have estimated the balmer ratios for the broad component of a radiative shock ( their model 2 ) .
for the optically thin ( case a ) situation , they find @xmath67 and @xmath68 .
our corresponding extinction estimates are @xmath69 and @xmath70 ; the @xmath71 overlap range is @xmath720.46 . however if the ly@xmath55 optical depth is @xmath73 , trapping increases the expected h@xmath0 flux and the predicted broad line ratios are @xmath74 and @xmath75 .
the h@xmath0 values then gives @xmath760 ; physical , positive values are allowed only at the @xmath77 level .
the h@xmath2 flux gives @xmath78 .
thus we conclude that the selective extinction is @xmath79 with smaller values preferred for finite thickness shocks , especially for the h@xmath0 measurement . for reference
, the fitting formulae for extinction values in @xcite give @xmath80 for this direction and @xmath81kpc .
our limit on the extinction can be translated to an effective hydrogen column @xmath82 , in agreement with the x - ray spectral fitting of the pwn . for either our conservative upper limit or the lower extinction inferred for modest optical depth shocks
the neutral column is substantially larger than the low @xmath83 inferred from the dm , suggesting low ionization along this line of sight . with the spectral resolution and s / n available we were able to resolve the h@xmath0 lines , making an initial exploration of the velocity structure in the post - shock flow .
in figure [ slitspec ] we show the h@xmath0 spectra along the two slit axes , with velocity shifts measured relative to the lsr as determined from night sky lines .
emission from the approaching ( blue ) and receding ( red ) sides of the bow shock are clearly seen in the pa=@xmath84 panel .
interestingly the front of the bow shock appears to be dominated by blue - shifted emission .
we discuss a likely interpretation of this peculiar velocity structure below .
7.7truecm -0.5truecm
the spatial and kinematic information provided by these observations give an excellent opportunity to probe the geometry of the pwn outflow @xcite . very helpful in this study
are the elegant solutions to the thin momentum - conserving bow shock provided by @xcite .
although these analytic results strictly only describe the shape of a contact discontinuity for a thin , fully mixed cold flow , they do provide a good approximate shape for the non - radiative forward shock in the ism @xcite .
the general velocity structure also follows the numerical simulations , although the cold analytic model has somewhat denser flows and correspondingly higher velocities downstream from the bow shock apex .
the characteristic standoff angle of the contact discontinuity in the thin shock limit is @xmath85 where for a relativistic pulsar wind the momentum flux is @xmath86 and the numerical value is scaled to the nominal spindown flux and distance of psr j1741@xmath12054 traveling at 100@xmath87 through a medium of number density @xmath88 .
thus we expect a small stand - off angle for the bow shock , as seen in figure 1 .
numerical simulations show that the forward ism shock scale is a factor @xmath89 larger @xcite .
however , figure 1 shows that the observed h@xmath0 shock is very flat at the apex ; it does not fit well to the standard @xcite thin shock solution .
while the numerical thick bow shock models do show a somewhat wider angle spread downstream , they do not display the very small standoff and flattened apex of the psr j1741@xmath12054 bow shock .
effects that may produce such flattened structure are underlying wind asymmetry and projection to the earth line - of - sight .
@xcite has provided useful expressions for the structure of anisotropic winds . in the pulsar case
, we expect that the wind momentum flux will be axisymmetric about the pulsar spin axis . in general
this provides momentum deposition misaligned with the bow shock axis . however , there is good evidence @xcite that pulsar spin and proper motion axis have a tendency to be aligned .
we therefore focus on the aligned , axisymmetric wind .
the pulsar wind is also highly relativistic , so that the mass of the bow shock shell is dominated by the swept - up mass of the ism . in the nomenclature of @xcite
this is @xmath90 , while the aligned case is @xmath91 .
we further restrict to an equatorially symmetric wind ( @xmath92 ) .
an isotropic wind has @xmath93 while an equatorial ( @xmath94 ) wind has @xmath95 . for these conditions
the analytic solution is a surface of rotation with pulsar shock distance @xmath96 ) for @xmath97 the angle to the direction of the pulsar - ism velocity with a cylindrical coordinate @xmath98 we can also write the tangential velocity ( in the bow shock frame ) and surface density of swept - up mass in the cold shell in terms of two quantities @xmath99/4\ ] ] and @xmath100/4\ ] ] with @xmath101^{1/2}/{\tilde { \bar \omega}}^2\ ] ] the tangential surface velocity in the shell , @xmath102 the shell surface mass density , and @xmath103 the mean mass per particle of the ambient ism
. this bow shock structure will be axisymmetric in some angle @xmath104 about the pulsar velocity , which will be inclined by angle @xmath105 to the earth line - of - sight . to complete the model of the bow shock appearance and radial velocity structure we project to the plane of the sky .
this places the emission at projected angles @xmath106 with respect to the ( pulsar ) wind origin with @xmath107 and @xmath108 the emission from each point along the bow shock is proportional to the appropriate @xmath109 and the projected velocity ( in the observer frame ) is given by @xmath110 with @xmath111 $ ] .
note that this velocity solution is for the mixed ( cooled , zero pressure ) flow of the swept up gas . in practice ,
numerical simulations show modest departures from this analytic solution for the tangential velocity @xcite and a detailed solution for the full line structure of the nebula should consider such effects .
however , a more basic difference lies in the fact that the non - radiative h@xmath0 arises from neutrals drifting into , and charge exchanging with , ions in the post - shock flow .
thus the h@xmath0 structure depends strongly on the portion of this flow which is reached by such neutral atoms @xcite .
in particular , immediately behind the forward ( ism ) shock we expect that the gas has velocity @xmath112 , where the angle between the pulsar velocity and the shock normal is @xmath113 $ ] .
only later does the post - shock flow converge to the tangent to the contact discontinuity .
this effect may be seen in the simulations of @xcite .
thus immediately behind the ism shock one expects a line - of - sight velocity @xmath114 since @xmath115 is always positive , this velocity always has the sign of @xmath116 .
an initial question is whether the shock created by an equatorial wind can reproduce the flattened shape of this pulsar bow shock .
one way to parametrize this shape is the ratio of the perpendicular half angular size @xmath117 , measured through the pulsar , to the angle from the pulsar to the projected limb of the wind shock in the forward direction @xmath118 .
note that @xmath119 .
this ratio is very large for psr j1741@xmath12054 s bow shock , @xmath120 .
we have computed this ratio for a number of axisymmetric , aligned wind bow shock models .
figure [ shockshape ] shows that the projected shock limb has a ratio nearly independent of @xmath105 for an isotropic ( @xmath93 ) wind , but that the ratio can reach the observed value for a @xmath121 ( @xmath95 ) wind and a pulsar motion nearly in the plane of the sky .
considering winds that are even more equatorially concentrated or including the flaring effect of the finite pressure in a thick bow shock allows somewhat smaller @xmath105 to be accommodated .
7.9truecm figure [ mod_im ] shows the projected shape of an equatorial relativistic wind emanating from the pulsar position ( circle ) for an inclination @xmath122 . for comparison ,
the line shows the limb of the isotropic wind bow shock ; the standoff distance for the equatorial wind is appreciably reduced .
note that , since formally @xmath94 , the bow shock approaches the star in the forward ( @xmath123 ) direction . since the pulsar wind likely has a jet component , a physical bow shock would be smooth at the apex .
similar matching to the h@xmath0 limb was used by @xcite to argue that the wind of the millisecond pulsar psr j2124@xmath13358 is anisotropic . for psr j1741@xmath12054
we have additionally been able to connect the shock shape with anisotropy detected in the x - ray synchrotron nebula ( 2.2 ) .
we next check if our model can explain the features of figure [ slitspec ] .
most remarkable is the dominance of blue - shifted emission at and in front of the pulsar . for the pa=@xmath84
slit , the bright knot at the apex is at @xmath124 to @xmath125 .
the cross - axis slit has h@xmath0 offset to @xmath126 to @xmath127 , with no equivalent red - shifted component .
this is significantly offset from the systemic ( ism ) velocity .
the lack of any red - shifted emission is at first surprising , given that mass flows around the bow shock .
however , if the pulsar is moving out of the plane of the sky ( @xmath128 ) , we see that the prompt post - shock broad - line emission ( from neutral atom charge exchange ) will show negative radial velocities from both sides of the nebula .
this is illustrated in the top two panels of figure [ mod_spec ] , which trace the velocity structure of the prompt emission .
here we plot a model for @xmath129 to better separate the velocity components .
the @xmath130 models ( for a 3d space velocity @xmath131 ) that match the shock shape also provide the best match to the velocity shifts in the slit spectra .
the immediate post - shock layers can dominate if the neutrals penetrate only partly into the shocked ism , before suffering nearly complete ionization .
this is equivalent to case c of @xcite .
in fact psr j1741@xmath12054 has a relatively large spindown luminosity among pulsars showing ism h@xmath0 shocks ( although below that of psr b0740 - 28 ) , so if the density and ionization fraction of the upstream medium are high , ionization may indeed be strong in the shocked ism .
thus , ahead of the pulsar where the shock is nearly normal , neutral h may only exist in the immediate post - shock layer and blue shifted emission would dominate the observed spectrum .
the pulsar space velocity @xmath132 is modest and as one moves behind the pulsar position , the shock obliquity @xmath133 increases sharply and the post - shock heating drops . as ionization drops , neutral h may reach the bulk flow along the contact discontinuity ( bucciantini & bandiera 2001 case b ) . charge exchange will continue and thus we expect components with negative radial velocity ( near side ) and positive radial velocity ( far side ) . indeed in figure
[ slitspec ] , across the body of the shell , the pa=@xmath84 spectrum shows two components with velocity extrema @xmath134 and @xmath135 from the front and back sides of the flow .
figure [ mod_spec ] shows the computed radial velocity for this ` mixed component ' .
the large positive and negative velocities expected at the pulsar position are not seen in the pa=@xmath136 spectrum or in the pa=@xmath84 spectrum in front of the pulsar , suggesting that near the apex neutral h atoms do not penetrate to this ` mixed ' tangential portion of the flow .
the lower right panel shows a heuristic merged model , where the ` mixed ' component grows as @xmath97 increases downstream .
this smoothed model bears reasonable similarity to the observed 2d spectrum , although it fails to reproduce the ` closed off ' shape along the symmetry axis and the red - shifted ` mixed ' emission from the body of the nebula is less prominent in the data .
clearly additional effects are needed to reproduce the full velocity structure of this bow shock and additional imaging and spectroscopy will be needed to fully explain its origin .
for example , the close - out of the symmetry axis spectrum suggests an expanding bubble structure rather than an open - backed bow shock .
bubble structure has been seen for other pwn bow shocks ( e.g psr b2224 + 65 s ` guitar nebula ' , cordes et al .
1993 ) ; @xcite have proposed that such bubble structures are a natural consequence of instabilities in the post - shock flow of the relativistic pulsar wind .
the fainter and slower red wing ` mixed ' emission 5 - 10@xmath20 behind the pulsar might be due to ism inhomogeneities or , plausibly , instabilities in the pwn trail .
one additional difference to the data is the lack of an obvious ism - velocity ` narrow ' line component in the measured spectra . in the model spectra
, the component is assigned 10% of the broad - line flux .
the limited lris spectral resolution makes it difficult to discern any such component , particularly in the brighter knots of the bow shock emission .
however , we estimate from the pa=@xmath84 spectrum of the nebular body a broad - to - narrow intensity ratio @xmath137 . interestingly
, this ratio provides a measure of @xmath138 , probing electron - ion equilibration in the post - shock emitting region @xcite ; relatively large ratios and fast equilibration appear to occur in low velocity shocks @xcite , supporting a modest velocity for psr j1741@xmath12054 .
if the narrow component can be isolated and kinematically resolved , it provides much additional information on the upstream medium s density and on pre - shock heating of the ambient gas .
7.0truecm 3.9truecm
the leading edge of the bow shock appears structured , especially in the ntt image , so it is somewhat difficult to measure the pulsar - apex angle and cross - axis half angle .
our best estimates are @xmath139 and @xmath140 .
we would like to relate these to the characteristic contact discontinuity standoff angle @xmath141 . in @xcite ,
the ism shock / contact discontinuity angle ratio was estimated as @xmath142 at the shock apex . while finite pressure effects can cause this to grow slightly downstream , here we assume a constant ratio .
also , for the @xmath95 , @xmath143 shape , we infer @xmath144 and @xmath145 .
together these estimates let us infer a characteristic ( isotropic wind ) contact discontinuity standoff angle @xmath146 .
this characteristic offset implies @xmath147 . as noted above the velocity structure
suggests that @xmath148 , so we loosely constrain the product @xmath149 . in turn
this suggests a plausible density @xmath150 if we adopt the dm distance . for the allowed extinction range ( @xmath151 ; i.e @xmath152 )
, we infer an h@xmath0 number flux from the full nebula of @xmath153 .
this may be related to the neutral density and velocity of the upstream medium @xcite .
for example the upstream neutrals will produce a narrow component h@xmath0 number flux @xmath154 for an excitation - to - ionization ratio @xmath155 and effective full nebula angular radius @xmath156 .
this predicts @xmath157 of the observed flux of the ( broad ) h@xmath0 line .
one can make a similar estimate for the flux of the broad component , depending on the ratio of charge - transfer to ionization rates , @xmath158 , i.e. @xmath159 ( depending on the extinction ) . with
the low @xmath160 and @xmath161 estimated above , we would infer a large yield of broad @xmath162 photons from charge transfer interactions , @xmath163 .
figures in @xcite suggest that the broad to narrow ratio grows rapidly at low shock velocities , especially for optically thin shocks , although velocities as small as 150 @xmath164 are not computed .
additional modeling and , especially , an accurate measurement of the narrow line component would help settle whether such high efficiency conversion to h@xmath0 occurs . turning to the nebula distance , we see that the relatively low @xmath165 inferred from the x - ray absorption column and the balmer line ratios supports a close distance for this pulsar , with the hi surveys @xcite showing values twice that inferred for the nebula at distances as small as @xmath166kpc ( _ heasarc nh tool _ ) . with improved spectral measurements of the pwn h@xmath0
we should be able to fit for a precise pulsar space velocity and inclination .
comparison with a proper motion , measurable from _ cxo _ x - ray imaging ( @xmath167 7y baseline ) or possibly hst imaging ( few year baseline ) would then yield a direct kinematic distance to the neutron star .
this is particularly valuable for understanding the apparently very high @xmath2-ray efficiency .
our match to the @xmath168 ratio suggests @xmath169 . for our estimated @xmath170
this is also consistent with the spread of velocities in our slit spectra and with the measured apex blue shift @xmath171 expected if the surrounding medium radial velocity is near the lsr and prompt emission dominates near the apex . if the pulsar spin and velocity are aligned , this implies a pulsar viewing angle @xmath172 . in turn
this may be compared with the viewing angles inferred for the observed @xmath2-ray pulse profile @xcite .
the two pole caustic ( tpc ) model has great difficulty producing the @xmath2-ray pulse width @xmath173 and lag from the radio @xmath174 . in the outer gap ( og )
picture , @xmath2-rays for such an old pulsar are only seen if viewed from near the equator , @xmath175 , although the observed pulse width and @xmath2-radio lag are then easily achieved .
the best - fit prefers @xmath176 , consistent with the angles inferred from the bow shock geometry above .
the discovery of an h@xmath0 bow shock associated with the @xmath2-ray pulsar psr j1741@xmath12054 provides an important opportunity to constrain the geometry and momentum deposition of a pulsar wind .
our initial images and spectra suggest that this pulsar wind has a strong equatorial concentration and that the spin axis ( and space velocity ) are close to the plane of the sky .
the spectroscopy implies a low extinction ( close distance ) for the nebula , consistent with pulsar dispersion measure estimates .
spectra were obtained for two slices across the nebula which are best described by a pulsar velocity of @xmath177 directed @xmath178 out of the plane of the sky .
the variation in the spectrum along the nebula s velocity axis suggests that different layers ( prompt and then mixed ) of the post - shock flow are emitting near the nebula apex and tail respectively .
some unmodeled features are seen and numerical simulation including pressure effects in the shock tail may be needed to explain the full velocity structure
. the apparently equatorially concentrated wind and near edge - on view are consistent with inferences of pwn flow from observations of other pulsars and with the @xmath2-ray emission geometry inferred from lat pulsations .
pulsar h@xmath0 bow shocks often display high symmetry suggesting that careful modeling can exploit their spatial and velocity structure to probe the dynamics of collisionless shocks in partly ionized media at a range of inclination angles .
if , as suggested by our initial imaging and spectroscopy , the portion of the post shock flow probed by the h@xmath0 emission varies along the bow shock surface , then the shock of psr j1741@xmath12054 provides an especially important laboratory to study the structure of collisionless non - radiative shocks . improved measurements
could precisely determine the anisotropy of the relativistic pulsar wind , may constrain the degree of wind - velocity alignment , and can help deliver a kinematic measurement of the pulsar distance .
such studies will , however , require higher spatial and spectral resolution data at good sensitivity .
this might be acquired through intermediate resolution integral field studies or fabry - perot h@xmath0 scans @xcite .
we thank tony readhead for collaboration on _ fermi _ counterpart studies and for the lris integrations , and aaron meisner for help with the wiyn observing and minimo processing .
we also thank pablo saz - parkinson , scott ransom and paul ray for valuable discussions on the detection and distance of the lat pulsars .
finally , we thank the referee , marten van kerkwijk , for a prompt , careful reading and useful suggestions that improved the text . this work was supported in part by nasa grants nnx08aw30 g and nnx10ad11 g and chandra grant go0 - 11097x .
abdo , a.a .
2009 , science , 325 , 840 aldcroft , t.l . ,
romani , r.w . & cordes , j.m .
2002 , apj , 400 , 638 bucciantini , n. & bandiera , r. 2001 , aa , 375 , 1032 bucciantini , n. 2002 , aa , 387 , 1066 chen , b. et al .
1998 , aa , 336 , 137 camilo , f. et al .
2009 , apj , 705 , 1 chatterjee , s. & cordes , j.m .
2002 , apj , 575 , 407 chevalier , r.a .
1980 , apj , 235 , 186 cordes , j.m . &
lazio , t.j.w .
2002 , arxiv : astro - ph/0207156 cordes , j.m . , romani , r.w . & lundgren , s.c .
193 , nature , 362 , 133 .
cormeron , f. & kaper , l. 1998 , aa , 338 , 273 dickey , j.m .. & lockman , f.j .
1990 , araa , 28 , 15 gaensler , b. et al .
2002 , apj , 580 , l137 heng , k. 2010 , pasa , 27 , 23 johnston , s. et al .
2005 , mnras , 364 , 1397 kalberla , p.m. et al .
2005 , aa , 440 , 775 ghavamian , p. et al .
2001 , apj , 547 , 995 mann , e.c . ,
romani , r.w . & fruchter , a.s . 1999 ,
aas , 195 , 4101 mihalas , d. & binney , j. galactic astronomy ng , c .- y . , & romani , r.w .
2008 , apj , 673 , 411 raymond , j.c .
2001 , spascirev , 99 , 209 romani , r.w . & watters , k.p .
2010 , apj , 714 , 810 sivakoff , g. et al .
2011 , in prep van adelsberg , m. et al .
2008 , apj , 689 , 1089 van kerkwijk . m. & ingle , a. 2008 , apj , 683 , 159 watters , k.p .
2009 , apj , 695 , 1289 wilkin , f.p .
1996 , apj , 459 , l31 wilkin , f.p .
2000 , apj , 532 , 400 | we have detected an h@xmath0 bow shock nebula around psr j1741@xmath12054 , a pulsar discovered through its gev @xmath2-ray pulsations .
the pulsar is only @xmath3 behind the leading edge of the shock .
optical spectroscopy shows that the nebula is non - radiative , dominated by balmer emission .
the h@xmath0 images and spectra suggest that the pulsar wind momentum is equatorially concentrated and implies a pulsar space velocity @xmath4 , directed @xmath5 out of the plane of the sky .
the complex h@xmath0 profile indicates that different portions of the post - shock flow dominate line emission as gas moves along the nebula and provide an opportunity to study the structure of this unusual slow non - radiative shock under a variety of conditions .
_ cxo _ acis observations reveal an x - ray pwn within this nebula , with a compact @xmath6 equatorial structure and a trail extending several arcmin behind .
together these data support a close ( @xmath7kpc ) distance , a spin geometry viewed edge - on and highly efficient @xmath2-ray production for this unusual , energetic pulsar . |
the electronic properties of strongly correlated transition - metal oxides strongly depend on spin and orbital degrees of freedom.@xcite both spins and orbitals on neighboring sites interact with each other via superexchange interactions,@xcite giving rise to a complex interplay and potentially to novel states induced by orbital quantum fluctuations . in this context
the occurrence of orbital order and more exotic states , like an orbital peierls state@xcite and an orbital - liquid ground state have been discussed.@xcite one possibility to test materials for those orbital states is to investigate the character of the elementary orbital excitations .
if the superexchange interaction dominates , one expects novel collective elementary excitations , namely orbital waves or orbitons with a significant dispersion,@xcite analogous to spin waves in a magnetically ordered state .
however , the orbitals are also strongly coupled to the lattice,@xcite therefore orbital excitations in many compounds are well described in the limit of local crystal - field excitations .
if both , superexchange interactions and the coupling to the lattice are relevant , the spectral signatures become more difficult to interpret , a complex situation that has hardly been studied thus far.@xcite the experimental observation of orbitons at low energies @xmath5250mev has been claimed based on raman data of lamno@xmath0 , @xmath6tio@xmath0 , and @xmath6vo@xmath0 with @xmath6=rare earth.@xcite however , the orbiton interpretation of these data has caused controversy.@xcite for manganites with partially occupied @xmath3 states in a predominantly octahedral crystal field , it is meanwhile well established that the orbital degree of freedom is quenched by the strong crystal - field splitting of the order of 1ev.@xcite recently , it has been shown by resonant inelastic x - ray scattering ( rixs ) on 1d cuprate chains that such high - energy orbital excitations may show a significant dispersion.@xcite the large dispersion reflects the strong superexchange coupling of the cuprates , and the high excitation energy results from the crystal - field splitting . at the same time , the high excitation energy of the lowest orbital excitation is a clear signature that orbital fluctuations are weak in the ground state , and that this orbitally ordered ground state is not affected by the high - energy dispersion .
in contrast , the crystal - field splitting is much weaker for the partially occupied @xmath2 states in titanates and vanadates . for @xmath6vo@xmath0 with @xmath7 electron configuration
, it has been suggested that orbital quantum fluctuations may be comparatively strong because superexchange interactions between @xmath2 electrons are frustrated on a cubic or nearly cubic lattice.@xcite from magnetic neutron scattering , indications for highly unusual orbital correlations were found and the existence of an orbital peierls state in the intermediate phase of yvo@xmath0 has been proposed.@xcite however , a recent study of the optical excitations across the mott - hubbard gap rules out that orbital fluctuations are strong in @xmath6vo@xmath0.@xcite some of us reported the observation of orbital excitations in the optical conductivity of yvo@xmath0.@xcite an optical absorption band at 0.4ev for polarization of the electric field @xmath8@xmath9@xmath10 has been attributed to the exchange of orbitals on adjacent sites , i.e. , to the direct excitation of two orbitons . a two - orbiton interpretation has also been discussed for the rixs data of the orbital excitations of the @xmath11 compounds latio@xmath0 and ytio@xmath0.@xcite in the titanates , the two - orbiton character is difficult to prove since the orbital excitations are observed at the same energy of 0.20 - 0.25ev in rixs , raman spectroscopy , and infrared absorption,@xcite while there is no clear signature of a one - orbiton excitation at about half the energy of the proposed two - orbiton peak . it has been claimed that the rixs intensity of two - orbiton excitations may exceed that of single - orbiton modes in titanates , in particular if quantum fluctuations are large in the orbital sector.@xcite rixs is the analogue of raman spectroscopy with x - ray photons .
rixs is very sensitive to orbital excitations , in particular in the soft x - ray range.@xcite at the v @xmath12 edge , a photon is resonantly absorbed , exciting a @xmath13 core electron to an empty @xmath14 state , and subsequently this intermediate state decays by re - emitting a photon , leaving the system in , e.g. , an orbitally excited state .
one major advantage of rixs over raman scattering with visible light is the short wavelength of x - rays , allowing to study the excitations as a function of the transferred momentum @xmath15 .
furthermore , by tuning the energy of the incoming light to a particular absorption band , special ions are selected and potentially also specific crystallographic sites . since the probing depth for soft x - rays is comparatively large ,
bulk properties are probed . here
, we report on rixs measurements with a resolution of 60mev at the v @xmath16 edge and 70mev at the o @xmath17 edge .
we observe spin - conserving intra-@xmath2 excitations at 0.1 - 0.2ev , orbital excitations from the high - spin @xmath18=1 @xmath7 ground state to low - spin @xmath18=0 @xmath7 states at 1.07ev and 1.28ev ( these are allowed due to the strong spin - orbit coupling of the core hole ) , and further excitations above 1.8ev .
our central results are the observations of a finite dispersion of the low - energy orbital excitations at 0.1 - 0.2ev and of an additional weak peak at 0.4ev for x - ray energies tuned to the o @xmath17 edge .
the combination of rixs and optical data@xcite gives strong evidence for the interpretation of the 0.4ev peak as a two - orbiton excitation .
the paper is organized as follows .
experimental details are given in sec .
[ experiment ] , followed in sec .
[ struc ] by a short description of the crystal structure , the spin / orbital - ordered phases , and the electronic structure .
section [ results ] describes our results . in sec .
[ secxas ] we briefly discuss x - ray absorption spectra .
the rixs features observed at the v @xmath16 edge are assigned to the different orbital excitations in sec .
[ assign ] . in sec .
[ lowenergy ] we concentrate on the low - energy orbital excitations observed at 0.1 - 0.2ev , focussing on the observation of a finite dispersion and on the competition between superexchange interactions and the crystal field .
section [ kedge ] addresses the rixs data measured at the o @xmath17 edge .
conclusions are given in sec .
[ conclusion ] . in the appendix
, we discuss optical data on crystal - field excitations of v@xmath19 ions in dysc@xmath20v@xmath21o@xmath0 .
single crystals of yvo@xmath0 were grown by the traveling - solvent floating zone method.@xcite prior to the experiments the samples have been freshly cleaved in air .
the measurements were performed using the soft x - ray spectrometer saxes at the adress beam line at the swiss light source , paul scherrer institut , villigen , switzerland , which has been designed especially for high - resolution rixs.@xcite the rixs data were measured with a combined energy resolution ( monochromator and spectrometer ) of @xmath22=60mev at the v @xmath16 edge and @xmath22=70mev at the o @xmath17 edge for scattering angles of @xmath23=90@xmath24 and 130@xmath24 ( see sketch in fig . [ fig1 ] ) .
all measurements were performed in specular geometry to keep the transferred momentum @xmath15 parallel to the crystallographic @xmath25 axis . for this purpose
, the sample was oriented with the @xmath10 axis perpendicular to the scattering plane and the incident x - rays were linearly polarized with @xmath8@xmath9@xmath10 ( @xmath26 polarization ) .
compared to @xmath27 polarization parallel to the scattering plane , @xmath26 polarization has the advantage that the orientation of the electric field with respect to the crystallographic axes is independent of the scattering angle .
the polarization of the scattered light has not been determined .
the attempt to measure a second sample with a different orientation to obtain data for @xmath28 failed due to experimental problems with this sample .
the incident energy has been varied across the v _ _ l__@xmath1 and o _ k _ edges . before the rixs measurements , room - temperature x - ray absorption ( xas )
spectra were measured _ in situ _ in total electron yield ( tey ) mode ( see right panel in fig . [ fig1 ] ) . below about 100k ,
the resistivity of the samples is very high , thus charging effects constricted the measurement of the drain current .
therefore , we additionally measured xas spectra in total fluorescence yield ( tfy ) mode ( inset in fig .
[ fig2 ] ) .
note that the absolute energy scale has been calibrated by a comparison with the xas data of polycrystalline yvo@xmath0 and of v@xmath29o@xmath0.@xcite = 300k for nine different incident energies across the v @xmath16 edge ( 512 - 519ev ) and at the o @xmath17 edge ( @xmath30530ev ) .
the x - ray absorption spectrum ( xas ) measured in total electron yield is shown in the right panel , with color - coded points indicating the excitation energies @xmath31 used for the rixs spectra .
a sketch of the scattering geometry is shown in the inset .
all spectra were measured with @xmath8@xmath9@xmath10 .
the rixs spectra were measured at a scattering angle of @xmath23=130@xmath32 .
the inelastic peaks with energy transfers up to 4ev correspond to orbital excitations .
the steep increase of the rixs signal at 4.5ev for @xmath31=530.1ev at the o _ k _ edge is in good agreement with the onset of charge - transfer excitations observed in the optical conductivity.@xcite however , the large intensity , exceeding the intensity of the elastic peak , points towards x - ray fluorescence emission , possibly superimposed by the charge - transfer excitations mentioned above . ]
the pseudo - perovskite compound yvo@xmath0 shows several phase transitions as a function of temperature . in the room - temperature phase an orthorhombic structure ( _ pbnm _ )
is adopted , while for @xmath33 a monoclinic structure ( _ p2@xmath34/b _ ) with long - range orbital order was observed@xcite(@xmath35 type , possibly with an admixture of @xmath36 type , see discussion in ref . ) .
antiferromagnetic order of @xmath36 type was found below @xmath37=116k .
below 77k , the structure becomes orthorhombic ( _ pbnm _ ) again , accompanied by a change to @xmath36-type orbital and @xmath35-type spin order.@xcite as far as the electronic structure is concerned , yvo@xmath0 is a mott insulator with two localized electrons in the @xmath14 shell of each v@xmath19 ion .
a crystal field of predominantly octahedral symmetry gives rise to a splitting of the @xmath14 levels into a lower - lying @xmath2 level and an @xmath3 level .
these are further split by the deviations from cubic symmetry.@xcite in cubic approximation , the lowest - lying two - electron state shows @xmath38 symmetry ( @xmath7 , @xmath18=1 ) , the lowest excited states are of @xmath39 and @xmath40 symmetry ( @xmath7 , @xmath18=0 ) and are expected above 1ev ( see sec .
[ assign ] ) . at still higher energies , states with the following symmetries
are expected : @xmath41 ( @xmath42 , @xmath18=1 ) , @xmath43 ( @xmath7 , @xmath18=0 ) , as well as @xmath44 , @xmath39 , and @xmath45 ( @xmath42 , @xmath18=0 or 1).@xcite the cubic approximation is sufficient for the overall assignment of all rixs peaks observed above 1ev at the v @xmath16 edge . however , for the discussion of orbital excitations below about 0.5ev we have to take into account deviations from cubic symmetry ( see secs . [ assign ] and [ lowenergy ] ) .
xas spectra have been measured in total fluorescence yield at 22k , 100k , and 300k , i.e. , in all three crystallographic phases ( inset of fig .
[ fig2 ] ) .
the absorption bands peaking at about 516ev , 522ev , and 534ev correspond to the v @xmath16 , v @xmath46 , and o @xmath17 edges , respectively , i.e. , transitions from the spin - orbit split v @xmath47 and v @xmath48 core states to empty v @xmath14 states , and from o @xmath49 to o @xmath13 states , respectively .
empty states within the @xmath13 level of oxygen arise due to hybridization . in particular ,
the lowest peak at the o @xmath17 edge at 530.1ev is attributed to states of mixed o @xmath13 - v @xmath14 character.@xcite a detailed analysis of the x - ray absorption spectra of yvo@xmath0 can be found in refs . and . for @xmath8@xmath9@xmath10
at 22k , 100k , and 300k for @xmath31=517.1ev and 2@xmath50=90@xmath24 .
these data sets have been accumulated over 3h each .
inset : temperature - dependent xas spectra of yvo@xmath0 for polarization @xmath8@xmath9@xmath10 at 22k , 100k , and 300k recorded in total fluorescence yield mode . ] = 512.6ev ( red ) and 517.1ev ( green ) .
the thin magenta curve shows the same data for @xmath51=512.6ev on a larger scale ( thick line : smoothed ) .
the strong high - energy rixs features observed for @xmath31=517.1ev can also be resolved for @xmath31=512.6ev but with strongly reduced spectral weight . ]
l c c c c & & & & + & & & & + + & & & & + & & & & + & & & & + + & & & & + & & & & + + & & & & + & & & & + & & & & + + & & & & + & & & & + & & & & + & & & & + figure [ fig1 ] shows rixs spectra at room temperature as a function of the transferred energy for different incident energies @xmath31 which are indicated by color - coded dots in the xas data ( right panel ) .
each spectrum has been accumulated over 30 minutes .
long - time scans with better statistics , accumulated over 3 hours are given in fig .
[ fig2 ] for @xmath31=517.1ev . at the v @xmath16 edge
, we observe a series of inelastic peaks with energy transfers of about 0.10 - 0.20ev , 1.07ev , 1.25ev , and 1.84ev , and a broad band of overlapping features between 2.2ev and 3.5ev at 300k .
the spectral weights of these peaks strongly depend on @xmath31 but the peak energies do not ( see figs . [ fig1 ] and [ fig3 ] ) , clearly demonstrating the raman character of these modes . with increasing @xmath31 , the spectral weight of the higher - lying bands increases , which reflects the change of the intermediate state and the corresponding change of the resonance conditions .
the main contribution to the rixs process at the v @xmath16 edge can be denoted by @xmath52$].@xcite in the first step a v @xmath47 core electron is excited by a photon with @xmath31 to the open @xmath14 shell . in the second step an electron relaxes to the @xmath13 shell , leaving the system either in the initial ground state ( @xmath53 , elastic peak ) , or in a low - energy excited state ( @xmath54 , inelastic peaks ) . for energies smaller than the mott - hubbard gap of about 1.6ev,@xcite
this low - energy excitation can be a phonon , a magnon , or an orbital excitation .
in general , also combined excitations such as an orbital excitation with phonon ( or magnon ) sidebands are allowed . in yvo@xmath0 phonon energies are smaller than 0.09ev,@xcite and magnon energies do not exceed 0.04ev,@xcite thus we attribute the observed rixs features to orbital excitations .
indeed , the excitation energies observed in rixs are in excellent agreement with the values derived for the orbital excitations by optical absorption measurements.@xcite for the sake of completeness we note that these orbital excitations may show phonon ( or magnon ) sidebands , which may be important for the discussion of the line width ( see sec . [ linewidth ] ) .
yvo@xmath0 shows inversion symmetry on the v sites .
thus orbital ( or @xmath55-@xmath55 ) excitations are forbidden in optics by parity and do not directly contribute to the optical absorption spectrum due to the dipole selection rule .
a weak contribution to the optical spectrum arises by the simultaneous excitation of a phonon . in oxides ,
the o phonon modes of 50 - 80mev are most effective .
in contrast , an orbital excitation in rixs corresponds to two subsequent dipole - allowed transitions , thus orbital excitations contribute directly to the rixs signal . when comparing the results of the two spectroscopies for the energies of the spin - conserving excitations from the @xmath7 @xmath18=1 ground state to @xmath18=1 final states ( see tab .
[ tab1 ] ) , the phonon shift occurring in optical absorption data has to be taken into account . at the v @xmath16 edge , the overall rixs spectrum ( see fig . [ fig2 ] ) can be well described in a local crystal - field scenario . for a @xmath7 high - spin @xmath18=1 ground state , we expect the following orbital excitations with increasing energy : spin - conserving intra-@xmath2 excitations at low energies ; spin - flip intra-@xmath2 excitations to a low - spin @xmath18=0 final state at about @xmath56 , where @xmath57ev denotes hund s coupling;@xcite and @xmath58 excitations at higher energies ( see tab .
[ tab1 ] ) . in the optical data,@xcite
the spin - conserving intra-@xmath2 excitations have been identified with a feature at about 0.20 - 0.26ev . subtracting a phonon shift of 50 - 80mev , this yields 0.12 - 0.21ev for the orbital excitations , in excellent agreement with the rixs peak at about 0.1 - 0.2ev .
these values also agree with various theoretical results predicting the intra-@xmath2 transitions in the range of 0.06 - 0.24ev.@xcite based on raman scattering data , the observation of orbitons has been claimed in @xmath6vo@xmath0 at lower excitation energies of 43mev and 62mev ( refs . ) and at 45mev and 84mev.@xcite however , a more recent study@xcite suggests that these peaks have to be interpreted as multiphonons , in agreement with our results .
a more detailed analysis of the low - energy rixs feature is given in sec .
[ lowenergy ] .
intra-@xmath2 excitations from the high - spin @xmath18=1 ground state to a low - spin @xmath18=0 final state are expected at about @xmath59ev.@xcite in optics , these excitations give rise to a rich structure with dominant peaks at about 1.1 - 1.3ev,@xcite again in excellent agreement with the rixs data .
the value of @xmath60 is hardly screened in a solid , thus very similar energies are observed for @xmath61 ions in different compounds , e.g. , v - doped @xmath62-al@xmath29o@xmath0.@xcite in the intermediate state of the rixs process at the v @xmath16 edge , spin is not a good quantum number due to the large spin - orbit splitting between v @xmath47 and v @xmath48 core - hole states .
therefore , the spin - selection rule does not apply , giving rise to a large spectral weight of such high - spin to low - spin excitations.@xcite this is corroborated by the absence of rixs peaks between 1.1 and 1.3ev in the data obtained at the o @xmath17 edge ( see fig .
[ fig1 ] ) . the @xmath49 core hole in the corresponding intermediate state
does not show an orbital moment , thus the spin - orbit mechanism is switched off and the spin - selection rule applies .
also in optics , the spin - flip excitations are suppressed by the spin selection rule , but they become weakly allowed by spin - orbit coupling or by the simultaneous excitation of a magnon.@xcite therefore , it is not necessary to consider a phonon shift for the comparison of optics and rixs in the case of spin - flip excitations .
the rixs peak observed at 1.84ev at 300k ( see fig .
[ fig2 ] ) corresponds to excitations from the @xmath2 level into the @xmath3 level , @xmath58 , i.e. , to excitations into @xmath63 in cubic approximation .
a value of roughly 2ev is typical for a v@xmath19 ion in an oxygen octahedron.@xcite again , rixs and optical results are in excellent agreement with each other ( see appendix ) .
a detailed assignment of the higher - lying excitations is hindered by the fact that the different contributions strongly overlap in the rixs data . at 300k
, we observe a peak at about 3.0ev with shoulders at about 2.3ev and 2.7ev . for the @xmath53 configuration in a ligand field of intermediate strength
, the tanabe - sugano diagram@xcite predicts in cubic approximation that the energy of the spin - flip excitation @xmath64 is about 2.1 - 2.2 times the energy of the @xmath65 band observed at 1.07 - 1.25ev .
thus we expect the @xmath64 peak roughly at 2.2 - 2.8ev , i.e. , in the range of the shoulders at 2.3ev and 2.7ev . above the @xmath64 peak ,
the tanabe - sugano diagram predicts excitations to @xmath66 , @xmath67 , and @xmath68 .
as noted above for the peaks at 1.07 - 1.25ev , the orbital excitation energies observed here are similar to the ones reported for v@xmath19 ions doped into @xmath62-al@xmath29o@xmath0 , in which the @xmath64 and @xmath66 excitations lie at 2.6ev and 3.1ev , respectively.@xcite for the data measured at the v @xmath16 edge we conclude that the rixs spectrum above about 1ev can be well described in terms of local crystal - field excitations .
our main goal is to clarify whether a local crystal - field scenario fully describes the orbital excitation spectrum , or whether superexchange plays a significant role .
a thorough quantitative analysis has to consider both the crystal field and superexchange interactions simultaneously .
however , this complex problem has hardly been addressed thus far.@xcite therefore we focus on the two limiting cases , either a dominant crystal field or dominant superexchange interactions .
both , superexchange and the non - cubic crystal field lift the threefold orbital degeneracy of the cubic @xmath69 ground state , giving rise to low - energy intra-@xmath2 excitations .
we concentrate on these excitations below 0.5ev , where the energy scale of superexchange interactions may become comparable to the excitation energy itself . figure [ fig4 ] depicts the dependence of the low - energy rixs spectra on temperature @xmath70 ( top panel ) and crystal momentum @xmath15 ( middle panel ) for incident energies of @xmath71ev and 517.1ev , as well as the dependence on @xmath31 ( bottom panel ) . at the v @xmath16 edge .
top : temperature dependence of the intra-@xmath2 excitations with @xmath23=90@xmath24 and @xmath31=512.6ev or 517.1ev at 22k , 100k , and 300k . middle : dependence on the scattering angle @xmath23 with @xmath23=90@xmath24 and 130@xmath24 for @xmath70=100k and different values of @xmath31 .
bottom : dependence on @xmath31 .
data for @xmath31=517.1ev are offset for clarity in the top and middle panels .
all spectra were accumulated over 3 hours with @xmath8@xmath9@xmath10 and @xmath15@xmath9@xmath25 . to estimate the contribution from the elastic peak with a full width at half maximum of 60mev ,
the spectra have been mirrored around zero ( dotted lines ) . ]
the overall line width of about 0.2ev of the features at 0.1 - 0.2ev is certainly larger than the experimental resolution of 60mev .
we identify two main reasons for the large line width : ( i ) a multi - peak structure , i.e. the experimental feature is composed of more than one peak , and ( ii ) the coupling to phonons or magnons , giving rise to incoherent parts of the orbital spectra . in a crystal - field scenario ,
the coupling of the orbital degrees of freedom to phonons gives rise to a vibronic character of the elementary excitations .
the spectrum is composed of a series of phonon sidebands . in a solid , these sidebands
are typically not resolved but form a broad excitation continuum.@xcite concerning point ( i ) , a multi - peak structure arises both from the splitting of the orbitally threefold degenerate @xmath69 state in a non - cubic environment and from the existence of two different v sites within the unit cell in the intermediate phase at 100k . in total , this yields four different excitation energies in a crystal - field scenario .
however , results based on lda@xcite ( local density approximation ) and first principles@xcite indicate that the crystal - field splitting is similar for the two v sites and that the excitation spectrum can be grouped into two bands ( at 0.06 - 0.08ev and 0.20 - 0.24ev according to ref .
, see tab . [ tab1 ] , or at 0.10 - 0.13ev and 0.20ev according to ref . ) which reflect the splitting of the cubic @xmath69 state . in a superexchange scenario ,
one also expects a multi - peak structure , namely two different one - orbiton modes . for the intermediate state at @xmath70=100k with @xmath36-type spin order and @xmath35-type orbital order ,
one mode is expected to show a dispersion for @xmath72 ( see below ) whereas the other one remains dispersionless.@xcite a model based only on superexchange neglects the coupling to phonons and thus neglects the vibronic broadening .
however , orbital excitations are also coupled to magnons via the common superexchange processes , giving rise to broad incoherent parts of the orbital spectra.@xcite even broader features are expected for two - orbiton excitations , which will be discussed in sec .
[ kedge ] . in the rixs data ,
it is not possible to directly distinguish different contributions to the broad feature at 0.1 - 0.2ev .
however , there is an interesting dependence on @xmath31 at the v @xmath16 edge ( see fig .
[ fig4 ] ) .
the rixs peak maximum is at 0.13 - 0.14ev for @xmath31=517.1ev , about 30 - 40mev lower than for @xmath31=512.6ev .
this is valid both for @xmath23=90@xmath24 and 130@xmath24 . at the same time
, the inelastic peaks extend to about 0.4ev for all data obtained at the v @xmath16 edge , irrespective of the shift of the peak maximum .
this can be rationalized by assuming that the rixs feature consists of two different peaks with spectral weights that depend on @xmath31 , similar to the overall behavior depicted in fig .
a strong dependence of the relative spectral weight of orbital excitations was observed also in rixs on tiocl.@xcite we emphasize that it is possible to describe our rixs data by two orbital excitations at 0.10 - 0.13ev and 0.20 - 0.22ev with gaussian line shapes under the assumption that the spectral weights of these two excitations depend on @xmath31 , temperature , and momentum ( see fig . [ fig5 ] ) .
a two - peak scenario is supported not only by the lda and first - principles results mentioned above@xcite but also by the optical conductivity , showing these excitations with a higher energy resolution at about 0.20ev and 0.26ev.@xcite subtracting a phonon shift of 50 - 80mev from the optical results yields orbital excitation energies of 0.12 - 0.15ev and 0.18 - 0.21ev , in agreement with our analysis of the rixs data ( see tab .
[ tab1 ] ) .
the orbital excitation energies observed in rixs agree with the expectations of a crystal - field scenario but are hard to reconcile with a _ pure _ superexchange scenario .
the energy scale for superexchange interactions is given by @xmath73=@xmath74 , where @xmath75 denotes the nearest - neighbor v - v hopping parameter and @xmath76 the on - site coulomb repulsion @xmath76@xmath775ev .
the value of @xmath73 can been determined from , e.g. , the spin - wave dispersion observed in inelastic neutron scattering.@xcite note that @xmath73 sets the overall energy scale for a spin - orbital superexchange model but that the effective spin - exchange coupling constants @xmath78 and @xmath79 are smaller than @xmath73 , they depend on the orbital correlations and thus on the ground state ( see , e.g. , fig . 4 in ref . ) . accordingly , different estimates have been given@xcite for the size of @xmath73 ( 0.02 - 0.04ev ) . for @xmath80 ,
the dispersion of orbital excitations in the intermediate phase between 77k and 116k with @xmath36-type spin order and @xmath35-type orbital order is given by@xcite @xmath81 with @[email protected] and a gap of @xmath83 .
the parameter @xmath84 depends on @xmath82 and on the crystal field and is explicitly given in ref . .
neglecting the crystal field , one finds @xmath84@xmath771 and a maximum energy of orbital excitations of about 0.08ev.@xcite for @xmath73=40mev , the different estimates of the crystal - field contribution yield orbital excitations with a gap of about 0.02 - 0.10ev and a bandwidth of about 0.02 - 0.05ev for yvo@xmath0 . in comparison ,
the observed excitation energies of 0.1 - 0.2ev are rather large and clearly indicate a significant contribution of the crystal field .
we like to add that a superexchange scenario predicts a significantly reduced dispersion for the low - temperature phase below 77k with @xmath35-type spin order and @xmath36-type orbital order.@xcite also for the low - temperature phase , a finite dispersion is only expected for @xmath80 . at first sight
, a superexchange model seems to suggest an alternative scenario in which the lower peak at 0.1ev corresponds to single - orbiton excitations and the upper one at 0.2ev to a two - orbiton contribution , as discussed for the titanates.@xcite here , one has to discuss the relationship between a single local orbital flip on a given transition - metal site ( i.e. , the rixs final state at the v @xmath16 edge ) and the number of orbitons that are excited . in a crystal - field scenario , flipping an orbital is equivalent to a single local crystal - field excitation .
two orbital flips on adjacent sites then correspond to two crystal - field excitations .
this simple correspondence breaks down in a superexchange scenario , reflecting many - body physics and the quasiparticle character of the excitations .
a local orbital flip has to be translated into the eigenstates of the bulk , i.e. , into a superposition of one- and multi - orbiton excitations , similar to one- and multi - magnon contributions in case of a local spin flip.@xcite however , a strong two - orbiton contribution at the transition - metal @xmath12 edge is only expected in case of strong orbital fluctuations.@xcite for yvo@xmath0 , strong orbital fluctuations have been ruled out recently by optical data for the mott - hubbard excitations.@xcite another argument against a two - orbiton interpretation of the peak at 0.2ev is the observation of a two - orbiton peak at 0.4ev in rixs at the o @xmath17 edge ( see sec .
[ kedge ] ) and in the optical conductivity.@xcite finally , the temperature dependence of the peak energies is hard to explain in a scenario with dominant superexchange interactions , as discussed in the next paragraph .
therefore , a two - orbiton interpretation of the rixs feature at 0.2ev can be ruled out .
the rixs data for different temperatures resolve the dependence of the intra-@xmath2 excitations on the crystal structure in the different phases of yvo@xmath0 very well ( see top panel of fig .
[ fig4 ] ) . from a phenomenological point of view
, we may neglect the two - peak structure discussed above and consider solely the maximum of the experimental peak or its first moment , i.e. , the center of mass of the inelastic spectrum . both the maximum and the first moment show shifts of the order of 10 - 40mev as a function of temperature .
similar shifts are obtained both from the fits using two gaussians for the two - peak structure ( see fig .
[ fig5 ] and tab .
[ tab1 ] ) and from the optical data.@xcite a shift of the excitation energy can be explained in a crystal - field scenario by a change of the crystal structure and thus of the crystal field , and in a superexchange - based scenario by a change of the orbital order.@xcite however , the orbital order changes dramatically between the different ordered phases and the orbitally disordered phase above 200k , and orbital fluctuations are only weak above 200k.@xcite accordingly , spin - orbital superexchange models predict a significant change of the excitation energies across the phase transitions between the different phases.@xcite in contrast , the observed shifts of 10 - 40mev are much smaller than the excitation energy of 0.1 - 0.2ev .
this clearly indicates that the excitation energy is not very sensitive to the details of the orbital order , pointing towards a dominant contribution of the crystal field .
this is corroborated by the good agreement between the experimental data and the theoretical results@xcite for local crystal - field excitation energies and their temperature - induced relative shifts ( see tab .
[ tab1 ] ) .
similar shifts of up to 40mev as a function of temperature are observed for the spin - flip intra-@xmath2 excitation at 1.25 - 1.28ev and for the @xmath58 excitations at 1.84 - 1.88ev , see fig .
[ fig2 ] and tab .
[ tab1 ] . not only the peak energy but also the spectral weight of the low - energy intra-@xmath2 excitations depends on temperature , being about 20 - 30% smaller at 300k compared to 100k for @xmath31=512.6ev ( see figs . [ fig4 ] and [ fig5 ] ) and @xmath31=517.1ev ( see fig . [ fig2 ] ) .
this is surprising given that the spectral weight rather increases with temperature for all higher - lying excitations , as shown in fig .
this particular behavior of the low - energy orbital excitations may be hard to understand in a pure crystal - field scenario.@xcite the most obvious way to distinguish a local crystal - field excitation from a propagating orbiton is to study the dependence of the orbital excitation energies on the momentum @xmath15 .
theoretical calculations based on an effective spin - orbital superexchange model predict a significant dispersion ( see eq .
( [ dispersion ] ) ) with a bandwidth of up to 0.05ev for orbital waves propagating along the @xmath10 axis of yvo@xmath0 in the intermediate phase with @xmath35-type spin order and @xmath36-type orbital order,@xcite i.e. , for ferromagnetic alignment of spins along @xmath10 .
perpendicular to @xmath10 , the orbiton dispersion is strongly suppressed by the antiferromagnetic correlations which yield a strong renormalization of the orbiton dispersion in dimensions higher than one.@xcite moreover , a superexchange scenario predicts no dispersion for @xmath15 within the @xmath85 plane if the @xmath86 orbital is occupied on every site,@xcite which is typically assumed for yvo@xmath0 .
this feature is due to the directional hopping properties of @xmath2 orbitals , i.e. , hopping between @xmath87 orbitals ( @xmath88 orbitals ) is finite only along the @xmath89 direction ( @xmath90 direction ) within the @xmath85 plane .
this completely suppresses orbital quantum fluctuations in the plane.@xcite the observed dispersion along the @xmath25 direction thus remains puzzling .
experimentally , the @xmath15 dependence can be studied by varying the scattering angle @xmath23 . for small energy losses ,
the transferred momentum @xmath15=@xmath91 is fully determined by the momentum of the incoming photon @xmath92=@xmath93 and by the scattering angle @xmath23 .
we have used @xmath23=90@xmath24 and @xmath94 , which at @xmath31=512.6ev corresponds to transferred momenta of @xmath95=0.3674@xmath96 and @xmath97=0.4709@xmath96 , respectively . in the specular geometry of our measurement ( see sketch in fig . [ fig1 ] ) , the transferred momentum points along the @xmath25 axis . with a lattice constant of @xmath25=5.273,@xcite
we find @xmath98@xmath7762% and @xmath99@xmath7779% . for an incident energy of @xmath100=517.1ev at @xmath70=100k and 300k for @xmath23=90@xmath24 and 130@xmath24 .
black solid lines depict fits of the spin - flip intra-@xmath2 excitations using two oscillators with gaussian line shape . ] from a phenomenological point of view , we may compare the first moment of the rixs feature for @xmath23=90@xmath24 and 130@xmath24 . at @xmath70=100k
, we find a shift of 13mev with an estimated uncertainty of 5mev for @xmath31=512.6ev and 517.1ev ( see middle panel of fig .
[ fig4 ] ) .
remarkably , the same analysis in the orbitally disordered phase at 300k yields a shift of 6mev for @xmath31=512.6ev but -4mev for 517.1ev , i.e. , at 300k the first moment is independent of the scattering angle within the error bars .
fitting two gaussians for @xmath31=512.6ev yields very similar results , i.e. , no shift at 300k ( see fig .
[ fig5 ] c @xmath101 e ) and shifts of about 20mev for both peaks at 100k ( see fig . [ fig5 ] b @xmath101 d ) .
as discussed above , a bandwidth of 20mev appears to be plausible for yvo@xmath0 with realistic parameters for the crystal field .
however , a superexchange scenario predicts no dispersion for the upper peak and a finite dispersion of the lower peak only for @xmath28 , not for @xmath102 .
the rixs feature at 0.1 - 0.2ev is composed of at least two peaks , and the shift of 13mev discussed above is much smaller than the line width .
therefore we have to consider an alternative scenario . a @xmath15 dependence of the transition matrix elements@xcite of the two different peaks may give rise to a @xmath15 dependence of the overall line shape ( and thus of the first moment ) via a transfer of the spectral weight between the two peaks around 0.1ev and 0.2ev .
however , the data contain three arguments against this scenario : ( i ) the first moment is independent of @xmath23 within
the error bars in the orbitally disordered phase at 300k .
( ii ) fits using two gaussian peaks at 100k yield a very similar shift of about 0.02ev for both peaks .
( iii ) the @xmath23-dependence of the spectral weight of the two gaussian peaks is opposite to the expectations of such a scenario , i.e. , the spectral weight of the lower ( upper ) peak _ increases ( decreases ) _ as the overall feature shifts to higher energy ( see fig .
[ fig4 ] ) .
this gives clear evidence that this alternative scenario of a transfer of spectral weight can be ruled out and supports the interpretation that the observed dependence on @xmath23 indeed reflects an intrinsic @xmath15 dependence , i.e. , a finite dispersion of the low - energy orbital excitations for @xmath102 .
for comparison , we also determined the @xmath23-dependence of the higher - lying rixs peaks at 1.1 - 1.3ev by fitting two gaussian peaks to the data for @xmath31=517.1ev , see fig .
note that these peaks have hardly any rixs intensity for @xmath31=512.6ev , see fig .
[ fig1 ] . thus far , a possible dispersion of these high - energy excitations has not been considered for the vanadates .
the peak position of the well - pronounced lower peak at 1.07ev with a width of about 0.13ev does not depend on the scattering angle @xmath23 within the error bars .
for the less intense , broader peak at 1.28ev , we find that the peak position is about 15mev higher for @xmath23=90@xmath24 than for 130@xmath24
. however , this peak is much broader and shows less intensity , and the fits are affected by the noise level .
therefore , the error bars of the fitted peak position are larger , and we conclude that its position is independent of @xmath23 within the error bars . summarizing this section , we find evidence that neither a pure superexchange model nor a pure crystal - field scenario can explain our data .
the peak energy of 0.1 - 0.2ev and its rather small temperature dependence clearly point towards a significant or possibly even dominant contribution of the crystal field .
the large line width indicates a vibronic coupling to phonons and possibly also the coupling to magnons .
the temperature dependence of the intensity and in particular the finite dispersion strongly suggest that also superexchange interactions play an important role .
we emphasize that our data show a finite dispersion for momentum @xmath102 , which is not expected in the superexchange models considered thus far.@xcite in order to quantify the individual contributions of the crystal field and of superexchange interactions both have to be considered on the same footing , which is a challenging task .
the excitation spectrum is then dominated by an orbiton - phonon continuum.@xcite if we consider also the coupling to magnons,@xcite we have to expect an orbiton - magnon - phonon continuum . a finite dispersion of the orbital excitations is expected to reveal itself via a momentum - dependent _ line shape _ of this continuum.@xcite
a detailed study of the @xmath15 dependence of the line shape remains a challenge to both theory and experiment .
edge ( bottom panel ) and of the optical conductivity @xmath103 of yvo@xmath0 as determined by ellipsometry ( top panel , from ref . ) .
the dashed line in the bottom panel shows data for the v @xmath16 edge from fig .
[ fig2 ] . ]
we propose that a more direct view on the relevance of intersite or superexchange interactions at present can be obtained from the rixs data measured at the o @xmath17 edge , which differs in many respects from the @xmath16 edge data , see figs .
[ fig1 ] and [ fig7 ] .
the spin - flip excitations at 1.1 - 1.3ev are absent ( or very weak ) due to the spin - selection rule , as discussed above .
the low - energy rixs feature with a peak at 0.1 - 0.2ev extends to much higher energies , see bottom panels of figs .
[ fig4 ] and [ fig7 ] .
finally , also the peak positions above 1.5ev are different .
the very large intensity above 4.5ev suggests x - ray fluorescence emission .
we speculate that this strong feature above 4.5ev corresponds to both fluorescence and charge - transfer excitations .
this is based on the astounding overall agreement between the optical conductivity@xcite and the rixs data from the o @xmath17 edge ( see fig . [ fig7 ] ) : ( i ) the absorption edges of mott - hubbard excitations and of charge - transfer excitations are observed at about 1.7ev and 4.5ev , respectively , ( ii ) the dominant contribution is a broad feature peaking at about 2.5ev , ( iii ) between 1.7ev and 3.3ev the spectral weight is smaller in the low - temperature phase , ( iv ) the peak at about 3.8ev is present only in the low - temperature phase , and ( v ) both onset energies at about 1.7ev and 4.5ev show a similar temperature dependence .
this gives clear evidence that rixs at the o @xmath17 edge is sensitive to mott - hubbard and charge - transfer excitations , i.e. , to intersite excitations .
additionally , we expect to observe the _ spin - conserving _
orbital excitations discussed above for the v @xmath16 edge . in the optical conductivity , the spectral weight of orbital excitations is orders of magnitude smaller compared to the mott - hubbard and charge - transfer excitations shown in the top panel of fig .
this explains that the pre - peak at about 1.9ev is more pronounced in the rixs data than in the optical conductivity , in particular in the low - temperature phase .
rixs at the o @xmath17 edge shows both the mott - hubbard excitations and the orbital excitation @xmath58 at about 1.8 - 1.9ev ( see sec .
[ assign ] ) .
we emphasize that the rixs intensity between 2.3ev and 2.7ev at the v @xmath16 edge was assigned to an @xmath64 ( @xmath18=0 ) excited state in sec .
[ assign ] .
this excitation involves a spin flip and thus does not contribute to rixs at the o @xmath17 edge , as shown above for the comparable excitations peaking at 1.1 - 1.3ev .
this strongly corroborates that the broad peak at 2.5ev at the o @xmath17 edge corresponds to mott - hubbard excitations and not to orbital excitations .
edge of yvo@xmath0 .
each diamond corresponds to a vo@xmath104 octahedron with three non - degenerate @xmath2 levels indicated by the horizontal bars . for a single octahedron ,
the ground state is a superposition of @xmath53 and @xmath105 , where denotes a ligand o @xmath13 hole .
see main text for a discussion of intermediate and final states . ] at @xmath70=100k and @xmath23=90@xmath24 for @xmath100=530.1ev ( o @xmath17 edge , blue ) and 512.6ev ( v @xmath16 edge , yellow ) as well as the difference between the two spectra ( red ) .
top panel / right axis : optical conductivity @xmath106 of yvo@xmath0 at @xmath70=100k for @xmath8@xmath9@xmath10 ( dashed green ) and @xmath107=@xmath106(100k)-@xmath106(4k ) ( solid green line ) .
the strong increase of @xmath106 at low and high energies reflects absorption by ( multi- ) phonons and the onset of interband excitations , respectively ( for details see ref . ) .
the two - orbiton excitation is forbidden in optics below t=77k .
thus the 4k data can be used as an estimate for the background of phonons and single orbital excitations .
the difference @xmath106(100k)-@xmath106(4k ) shows an estimate of the two - orbiton contribution at 100k with a peak at 0.4ev , similar to the rixs difference spectrum .
middle and bottom panel : rixs spectra for @xmath100=530.1ev and two different @xmath15 values obtained by changing the scattering angle , i.e. , 2@xmath50=90@xmath24 ( middle panel ) and 130@xmath24 ( bottom panel ) . for an estimate of the peak shape and position of the two - orbiton peak at 0.4ev in rixs , two gaussians at 0.12 and 0.22ev have been fitted ( width and area ) after subtracting the mirrored elastic line . ]
rixs at the o @xmath17 edge has not been studied very much thus far .
it is known to be sensitive to both orbital ( or @xmath55-@xmath55 ) excitations and to intersite charge - transfer excitations.@xcite to the best of our knowledge , this is the first report on the observation of mott - hubbard excitations in rixs at the o @xmath17 edge .
the differences between the rixs data from the o @xmath17 edge and from the v @xmath16 edge originate from the different microscopic excitation processes . in both cases ,
the incoming x - ray photon excites a core electron into the upper hubbard band . at an o site
, this is possible due to the hybridization between v @xmath14 and o @xmath13 states . in a local picture for a single vo@xmath104 octahedron
, the ground state can be described by @xmath62|@xmath108 + @xmath109|@xmath105@xmath110 with @xmath111=1 , where denotes a ligand o @xmath13 hole ( see top panel of fig .
[ fig8 ] ) . the incoming x - ray
thus may promote the ligand o @xmath13 hole to an o @xmath49 core hole at one particular o site which connects two octahedra .
this yields an intermediate state state in the upper hubbard band , where @xmath112 and @xmath113 refer to the two v sites ( see middle panel of fig . [ fig8 ] ) .
the system relaxes by filling the core hole with a valence electron , which may yield different low - energy excited states , see bottom panel of fig .
the final state corresponds to a charge - transfer excitation @xmath114@xmath110 if the core hole is filled by an o @xmath13 electron from below the fermi energy , i.e. , the final state shows both an o hole below the fermi energy and a @xmath105 state within the upper hubbard band .
similarly , a mott - hubbard excitation results if the valence electron is taken from the lower hubbard band , i.e. , @xmath115@xmath110 , which again is possible due to the @xmath13-@xmath14 hybridization expressed by the second term .
alternatively , an orbital excitation @xmath116 is created if the core hole is filled by an electron from the upper hubbard band , similar to the dominant mechanism at the v @xmath16 edge .
our data were measured with linearly polarized incident x - rays with @xmath8@xmath9@xmath10 . the dipole selection rule states that the initial excitation from o @xmath13 to o @xmath49 involves a hole with @xmath117 character . for this @xmath117 hole ,
the @xmath13-@xmath14 hybridization is different for v - o - v bonds along @xmath10 or within @xmath85 .
along @xmath10 , o @xmath117 shows a strong hybridization with the v @xmath3 orbital with @xmath118 symmetry . within @xmath85 ,
o @xmath117 hybridizes with v @xmath2 orbitals with @xmath87 and @xmath88 symmetry .
therefore , v - o - v bonds along @xmath10 and within @xmath85 certainly show different contributions to our o @xmath17 rixs data measured with @xmath8@xmath9@xmath10 .
however , a quantitative statement requires a much more thorough theoretical investigation of the microscopic processes .
most interesting to us is the possibility to create _ two _ orbital excitations simultaneously . in the literature , the excitation of two orbitons has been calculated for rixs at the _ transition - metal _ @xmath17 and @xmath12 edges with a focus on manganites and titanates.@xcite two different mechanisms have been described . in the intermediate state ,
the core hole may shake up the valence electrons on the transition metal site , giving rise to , e.g. , a single orbital flip .
this state has to be translated into the eigenstates of the bulk , e.g. , a local crystal - field excitation or a single ( or multi- ) orbiton .
alternatively , the core hole may modulate the superexchange , giving rise to two orbital flips on adjacent sites .
similarly we expect for rixs at the o @xmath17 edge that the o @xmath49 core hole modulates the v - o - v intersite superexchange and shakes up the valence electrons .
therefore , the intermediate state may evolve from @xmath119;@xmath120;@xmath121 to @xmath122 , which may decay to a final state with two orbital excitations , @xmath123 , see fig .
we expect that the sensitivity to such double orbital excitations is enhanced at the o @xmath17 edge because the excitation occurs at an o site connecting two v sites , and due to hybridization the intermediate state extends over a v - o - v bond . at the v @xmath16 edge ,
the intermediate state is much more localized on a single v site due to the strong interaction with the v @xmath13 core hole .
this is supported by our observation of mott - hubbard excitations at the o @xmath17 edge at 2.5ev , which clearly demonstrates the sensitivity to intersite excitations . moreover ,
similar intersite excitations have been reported for rixs at the o @xmath17 edge of insulating late transition - metal oxides . in nio , a double spin flip has been observed on neighboring sites,@xcite and the excitation of two magnons has been discussed for insulating parent compounds of the high-@xmath124 cuprates.@xcite similar to our case , the two - magnon feature in the cuprates can be revealed by comparison of rixs at the cu @xmath16 edge and the o @xmath17 edge.@xcite in the optical data , a peak at 0.4ev for @xmath8@xmath9@xmath10 in the intermediate monoclinic phase ( see top panel in fig . [ fig9 ] , right axis ) has been interpreted as a two - orbiton excitation , i.e. , an _ exchange of orbitals between two neighboring v sites _ along the @xmath10 axis.@xcite this is equivalent to a double orbital flip , i.e. , from the @xmath87 orbital to the @xmath88 orbital on one site and _ vice versa _ on a neighboring site . the rixs spectra for @xmath31=530.1ev in the bottom panels of figs . [ fig4 ] and [ fig7 ]
indeed reveal an additional contribution at the o @xmath17 edge , which extends up to about 0.9ev , roughly twice the cut - off of the single orbital excitations . after subtracting an estimate of the single orbital excitations ,
the o @xmath17 edge data show a peak at about 0.4ev ( see fig .
[ fig9 ] ) , in good agreement with the optical data .
we stress that only orbital excitations can be expected in this energy range in yvo@xmath0 .
as discussed above , phonons and magnons are relevant below 0.1ev only , and the mott - hubbard gap is larger than 1ev.@xcite the rixs data at the v @xmath16 edge clearly demonstrate that the single orbital excitations are located at 0.10 - 0.22ev for all temperatures measured here , and the data from the o @xmath17 edge strongly corroborate the existence of the two - orbiton excitation . together ,
rixs and optics provide strong evidence for the two - orbiton interpretation of the feature at 0.4ev , demonstrating the relevance of orbital exchange processes in yvo@xmath0 .
in conclusion , we have reported high - resolution rixs results at the v @xmath16 edge of yvo@xmath0 which reveal the orbital excitations with unprecedented quality . by comparison with optical data,@xcite peaks at 0.1 - 0.2ev ,
1.07ev , 1.28ev , 1.86ev , and between 2.2ev and 3.5ev unambiguously can be identified as orbital excitations .
orbital excitations to final states with either @xmath18=1 or @xmath18=0 are very well resolved , demonstrating the superior capability of high - resolution rixs for the study of orbital excitations .
the lowest orbital excitations at 0.1 - 0.2ev correspond to spin - conserving intra-@xmath2 excitations .
the excitation energy , its small temperature dependence , and the large line width indicate a large contribution of the crystal field .
this low - energy rixs feature can be decomposed into two different contributions at 0.1ev and 0.2ev , respectively , which is supported by the dependence of line shape and peak energy on the incident energy .
most remarkably these low - energy excitations show a shift of about 13 - 20mev as a function of the scattering angle .
we argue that this reflects a finite dispersion of orbital excitations for @xmath102 . however , the shifts are much smaller than the line width and the superexchange models studied thus far do not predict any dispersion for @xmath102
. therefore , we can not disentangle quantitatively the contributions from superexchange and from the crystal field at this stage and call for further theoretical investigations of the low - energy orbital excitations in @xmath2 systems . in particular , the microscopic mechanism for a dispersion for @xmath125 needs to be understood .
rixs at the o @xmath17 edge is more sensitive to intersite excitations such as charge - transfer excitations or excitations across the mott - hubbard gap . in particular
, we find an additional contribution to the rixs signal which peaks at about 0.4ev .
we interpret this feature as a two - orbiton excitation , i.e. , an exchange of orbitals on adjacent sites , in agreement with optical results.@xcite a quantitative description of the low - energy orbital excitations in yvo@xmath0 will have to consider both superexchange interactions and the coupling to the lattice .
the xas and rixs measurements were performed at the adress beamline of the swiss light source using the saxes instrument jointly built by paul scherrer institut ( switzerland ) , politecnico di milano ( italy ) , and cole polytechnique fdrale de lausanne ( switzerland ) .
the authors thank n. hollmann , l .- h .
tjeng , and l. braicovich for fruitful discussions .
this work was supported by the dfg via sfb 608 and partially by the u.s .
department of energy , office of basic energy sciences , division of materials science and engineering , under contract no .
de - ac02 - 76sf00515 and by the alexander von humboldt foundation .
above 1.6ev , orbital or @xmath55-@xmath55 excitations can not be detected by optical spectroscopy on yvo@xmath0 because they are masked by the much larger spectral weight of excitations across the mott - hubbard gap.@xcite as an alternative , we have chosen to study the orbital excitations of single v@xmath19 ions substituted into dysco@xmath0 , which is isostructural to yvo@xmath0 . due to the similarity of the ionic radii
, the crystal field on the v site is expected to be very similar in dysc@xmath20v@xmath21o@xmath0 and yvo@xmath0 . at the same time , the band gap is much larger in dysco@xmath0 than in yvo@xmath0 , which allows us to study the @xmath58 excitations .
moreover , sc@xmath19 has a @xmath126 electron configuration , thus all orbital excitations of dysc@xmath20v@xmath21o@xmath0 can be attributed to dy @xmath127-@xmath127 or v @xmath55-@xmath55 excitations .
the dy @xmath127-@xmath127 excitations are well studied,@xcite the peaks observed in our dysco@xmath0 data are indeed typical.@xcite in particular , one does not expect any dy @xmath127-@xmath127 excitations between about 1.7 and 2.6ev .
the v @xmath55-@xmath55 excitations can be identified by comparing dysc@xmath20v@xmath21o@xmath0 with dysco@xmath0 .
single crystals have been grown by the traveling - solvent floating - zone method .
the purity and single - phase structure of the crystals was confirmed by powder and single - crystal x - ray diffraction at room temperature .
both compounds adopt the space group @xmath128 and have similar lattice constants and v - o bond lengths as rvo@xmath0 .
typical dimensions of the as - grown rods are a few millimeter along all three crystallographic axes .
the sample of dysco@xmath0 is transparent yellow , whereas the sample of dysc@xmath20v@xmath21o@xmath0 is opaque and dark - green .
for the transmittance measurements , we cut samples of dysco@xmath0 with [ 100 ] orientation and of dysc@xmath20v@xmath21o@xmath0 with [ 001 ] orientation ( both with the @xmath129 axis in - plane ) .
these were polished to a thickness @xmath55 of 730@xmath13010@xmath131 m in case of dysco@xmath0 and 1312@xmath13010@xmath131 m in case of dysc@xmath20v@xmath21o@xmath0 .
transmittance measurements were performed between 0.1 and 3.0ev with linearly polarized light using a fourier spectrometer . and
dysc@xmath20v@xmath21o@xmath0 for @xmath8@xmath9@xmath129 at 300k .
the additional absorption peaks observed in dysc@xmath20v@xmath21o@xmath0 are due to orbital excitations of the substituted v@xmath19 ions . ] in fig .
[ fig10 ] we show the optical absorption @xmath132 , where @xmath70 denotes the measured transmittance . at 300k ,
the @xmath58 absorption band of v@xmath19 in dysc@xmath20v@xmath21o@xmath0 ( excitation to @xmath63 in cubic notation ) is observed at 1.92ev . subtracting a phonon shift of 50 - 80mev yields 1.84 - 1.87ev for the @xmath58 excitations , in excellent agreement with the rixs data peaking at 1.84ev ( see fig .
[ fig2 ] and tab .
[ tab1 ] ) .
observation of the lower - lying v@xmath19 intra-@xmath2 transitions is impossible because of the superposition with the predominant @xmath127-@xmath127 transitions .
finally , the strong increase of absorption of dysc@xmath20v@xmath21o@xmath0 above 2.2ev ( see fig . [ fig10 ] ) is attributed to higher - lying excitations , again in excellent agreement with rixs .
79ifxundefined [ 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/10.1126/science.288.5465.462 [ * * , ( ) ] @noop * * , ( ) link:\doibase/10.1103/physrevlett.91.257202 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.67.100408 [ * * , ( ) ] link:\doibase/10.1103/physrevlett.91.257203 [ * * , ( ) ] link:\doibase/10.1143/ptps.160.155 [ * * , ( ) ] link:\doibase/10.1103/physrevb.62.2338 [ * * , ( ) ] @noop * * , ( ) http://link.aps.org/doi/10.1103/physrevlett.87.217202 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.76.075108 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.85.064406 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevlett.100.167205 [ * * , ( ) ] @noop * * , ( ) link:\doibase/10.1103/physrevlett.94.076405 [ * * , ( ) ] link:\doibase/10.1103/physrevb.73.224436 [ * * , ( ) ] link:\doibase/10.1103/physrevb.73.020409 [ * * , ( ) ] link:\doibase/10.1103/physrevlett.97.157401 [ * * , ( ) ] @noop * * , ( ) http://stacks.iop.org/1367-2630/7/i=1/a=144 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.75.064303 [ * * , ( ) ] http://stacks.iop.org/1367-2630/10/i=5/a=053027 [ * * , ( ) ] @noop * * , ( ) http://link.aps.org/doi/10.1103/physrevlett.93.147204 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.77.035109 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevlett.107.147201 [ * * , ( ) ] http://dx.doi.org/10.1038/nature10974 [ * * , ( ) ] link:\doibase/10.1103/physrevlett.99.126402 [ * * , ( ) ] link:\doibase/10.1103/physrevb.69.075118 [ * * , ( ) ] link:\doibase/10.1103/physrevlett.86.3879 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.86.125128 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevlett.103.107205 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.81.125118 [ * * , ( ) ] link:\doibase/10.1103/physrevb.77.113102 [ * * , ( ) ] http://www.worldscientific.com/doi/abs/10.1142/s0218625x02003822
[ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.69.125103 [ * * , ( ) ] http://www.sciencedirect.com/science/article/pii/s0925838803005759 [ * * , ( ) ] link:\doibase/10.1103/physrevlett.92.117406 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevlett.95.197402 [ * * , ( ) ] link:\doibase/10.1103/physrevb.73.035111 [ * * , ( ) ] link:\doibase/10.1103/physrevlett.96.067402 [ * * , ( ) ] link:\doibase/10.1103/physrevb.76.125105 [ * * , ( ) ] link:\doibase/10.1103/physrevb.77.045132 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevlett.101.106406 [ * * , ( ) ] http://link.aps.org/doi/10.1103/revmodphys.83.705 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevlett.109.117401 [ * * , ( ) ] link:\doibase/10.1103/physrevb.65.174112 [ * * , ( ) ] http://dx.doi.org/10.1107/s0909049510019862 [ * * , ( ) ] http://link.aip.org/link/?rsi/77/113108/1 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.59.7422 [ * * , ( ) ] link:\doibase/10.1103/physrevb.61.11506 [ * * , ( ) ] http://jpsj.ipap.jp/link?jpsj/71/2086/ [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.80.195107 [ * * , ( ) ] link:\doibase/10.1103/physrevb.73.094440 [ * * , ( ) ] link:\doibase/10.1103/physrevb.74.054412 [ * * , ( ) ] http://stacks.iop.org/0953-8984/20/i=29/a=293201 [ * * , ( ) ] http://jpsj.ipap.jp/link?jpsj/75/124707/ [ * * , ( ) ] @noop _ _ ( , ) _ _ , @noop ph.d
. thesis , ( ) _ _ , @noop ph.d .
thesis , ( ) , http://stacks.iop.org/0953-8984/17/i=35/a=007 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.69.075110 [ * * , ( ) ] http://www.sciencedirect.com/science/article/pii/0022459690902029 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.54.5368 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.56.1780 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.57.14584 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevlett.102.027401 [ * * , ( ) ] @noop _ _ ( , ) http://link.aip.org/link/?jcp/116/471/1 [ * * , ( ) ] @noop _ _ ( , ) http://link.aps.org/doi/10.1103/physrevb.75.184434 [ * * , ( ) ] http://stacks.iop.org/1742-6596/391/i=1/a=012168 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevlett.107.107402 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevlett.105.167404 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.61.4186 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.66.165104 [ * * , ( ) ] http://jpsj.ipap.jp/link?jpsj/75/044702/ [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevlett.105.157006 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevb.85.214527 [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevlett.110.087403 [ * * , ( ) ] | we report on the observation of orbital excitations in yvo@xmath0 by means of resonant inelastic x - ray scattering ( rixs ) at energies across the vanadium _ _ l__@xmath1 and oxygen _ k _ absorption edges . due to the excellent experimental resolution we are able to resolve the intra-@xmath2 excitations at 0.1 - 0.2ev , 1.07ev , and 1.28ev , the lowest excitations from the @xmath2 into the @xmath3 levels at 1.86ev , and further excitations above 2.2ev . for the intra-@xmath2 excitations at 0.1 - 0.2ev ,
the rixs peaks show small shifts of the order of 10 - 40mev as a function of temperature and of about 13 - 20mev as a function of the transferred momentum @xmath4 .
we argue that the latter reflects a finite dispersion of the orbital excitations . for incident energies tuned to the oxygen _
k _ edge , rixs is more sensitive to intersite excitations .
we observe excitations across the mott - hubbard gap and find an additional feature at 0.4ev which we attribute to two - orbiton scattering , i.e. , an exchange of orbitals between adjacent sites .
altogether , these results indicate that both superexchange interactions and the coupling to the lattice are important for a quantitative understanding of the orbital excitations in yvo@xmath0 . |
an exoplanet on an edge - on orbit periodically passes behind its host star .
the decrement in thermal flux that occurs during such an eclipse is a measure of the dayside brightness temperature of the planet .
the brightness temperature of a planet varies with wavelength , primarily because of the atmosphere s wavelength - dependent opacity and vertical temperature profile ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
if different wavelengths probe the same atmospheric layer ( e.g. , a cloud deck ) then the planet will appear to have a blackbody spectrum . in the absence of clouds ,
a planet may still have a blackbody spectrum if the atmospheric layers probed are isothermal .
indeed , the emission spectra of some planets are reported to be featureless : e.g. , tres-2 @xcite , tres-3 @xcite , wasp-18b @xcite , and wasp-12b @xcite . in principle
, the detection of molecular bands in the infrared emission spectrum of a planet enables the retrieval of greenhouse gas abundances and the vertical temperature profile of the planet ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
spectral resolution is critical to such retrieval exercises because a high - resolution emission spectrum is more likely to deviate significantly from a blackbody , and renders the retrieval problem well - constrained .
this bodes well for current and future efforts to perform _ bona fide _ emission spectroscopy .
so far , however , the vast majority of exoplanet emission measurements have been broadband eclipse photometry .
a typical retrieval model uses a dozen parameters to describe the atmospheric composition and vertical temperature profile , while a typical hot jupiter has only been observed in 24 thermal broadbands .
even for the few planets with 6 or 7 thermal measurements , the photometric retrieval problem is under - constrained . a widely noted consequence of the parameter
data mismatch is that exact atmospheric properties can not be uniquely determined , making color - color and color - magnitude diagrams more realistic approaches to atmospheric classification @xcite .
the less - discussed aspects of under - constrained retrieval are that ( 1 ) there is no way to reject erroneous measurements , and ( 2 ) the estimated uncertainties on eclipse depths directly affect the uncertainties on atmospheric parameters .
this is in stark contrast to over - constrained problems such as fitting an occultation model to time - series data , for which it is customary to perform outlier rejection ( e.g. , @xmath2-clipping ) , and for which the photometric uncertainties are estimated in the process of fitting a model to the data , rather than trusting the output of aperture photometry routines .
nonetheless , many exoplanet discoveries have been based on broadband emission spectra : a temperature inversion in the atmosphere of hd 209458b was inferred from 4 broadband eclipse depths @xcite , disequilibrium chemistry was invoked to explain the 6-band emission spectrum of gj 436b @xcite , and high atmospheric c / o was discovered based on 7 broadband eclipses of wasp-12b @xcite .
these successes have led to classifying planets based on temperature inversions ( using 2 broadbands per planet ; * ? ? ? * ) and c / o ratio ( using @xmath3 bands per planet ; * ? ? ?
temperature inversions and non - solar chemistry have since been disputed for each of the exemplar planets due to the re - reduction of existing data @xcite acquisition of new data at the same wavelength @xcite or acquisition of new data at different wavelengths @xcite .
such challenges are not unique to eclipse _ photometry _ : the featureless day - side emission spectrum of hd 189733b @xcite exhibited an absorption feature at a later epoch @xcite , and line emission from the dayside of hd 189733b @xcite has been disputed by @xcite . nor are issues of repeatability limited to superior conjunction : the first thermal phase measurements of an exoplanet @xcite were later found to be off by @xmath4 in phase and more than a factor of 2 in amplitude @xcite ; the first half of the thermal phase measurements of @xcite were later found to be corrupted by detector systematics @xcite .
the situation is similar for transit spectroscopy , where initial claims of molecular absorption bands @xcite were disputed on the basis of data reduction , error estimation , and astrophysical variability @xcite .
indeed , @xcite offers a sobering review of the exoplanet atmospheric characterization field , speculating that many of the extraordinary claims of the past decade may be overturned by better data . in this article
we attempt to reconcile burrows pessimistic view with the growing body of papers making statements about planetary atmospheres based on a handful of eclipse measurements . instead of focusing on a single planet , we perform a holistic analysis of all transiting planets with multiple eclipse measurements .
we consider only broadband measurements ( for which it is easy to quantify the number of independent observational constraints ) of eclipse depths ( which are unaffected by star spots ) .
our approach is to compare the goodness - of - fit and evidence for three classes of models : blackbodies , self - consistent radiative transfer , and spectral retrieval .
since the disputes over atmospheric properties have often revolved around the reliability of eclipse depths , we empirically estimate the accuracy of broadband eclipse measurements .
notably , the dominant `` signal '' in space - based eclipse photometry is usually the detector sensitivity , which must be modeled using the very same observations of the science target .
future observations of transiting planets with the james webb space telescope are likely to resolve many of the current scientific disputes about the nature of hot jupiter atmospheres . attempts to push the observatory to smaller and cooler planets , however ,
will still rely on self - calibration ; error estimation and repeatability will therefore remain critical issues .
a search on exoplanet.org @xcite combined with a careful literature review yields 44 exoplanets with published photometric eclipse measurements in at least two thermal wavelengths ( @xmath5 @xmath6 m ) , summarized in table [ thermal_eclipses ] . in most cases ,
only a single occultation has been measured at each wavelength .
bolded numbers signify measurements based on more data : multiple eclipses and/or an eclipse embedded in phase variations .
.planets with at least 2 thermal eclipse measurements [ thermal_eclipses ] [ cols="<,<",options="header " , ] comparing the @xmath7 and @xmath8 columns of table [ reshoots ] suggests that published eclipse uncertainties are too small : the original researchers , subsequent researchers , or both groups under - estimated the uncertainty in their measurement .
since the latter eclipse measurements are based on more data , we assume that they represent an accurate measurement and uncertainty , while the original measurements , based on a simple occultation , had under - estimated error bars .
alternatively , the planets may be exhibiting weather that changes the eclipse depths from one epoch to the next , as predicted by @xcite .
eclipse depth variability at the level of @xmath9 would invalidate spectral retrieval because multi - band broadband emission spectra are constructed over a span of many planetary orbits .
the weather hypothesis is ruled out in a few cases by the repeat observations discussed in [ best_case ] , however .
we quantify the degree to which eclipse uncertainties have been under - estimated by combining @xmath7 and @xmath8 to obtain an empirical estimate of systematic uncertainty , following 2.1 of @xcite . in the first case
, we assume there is an additional source of noise that affects single - eclipse measurements . physically , this might correspond to how well one can model the detector given only a few hour observation of the science target .
we estimate the magnitude of this systematic uncertainty by considering the distribution of @xmath10 . in the one case where the epoch - to - epoch discrepancy , @xmath7 , was smaller than the total published uncertainty , we set this quantity to zero .
the symmetric - distribution is decidedly asymmetrical : researchers analyzing single - eclipse measurements have over - estimated the eclipse depth more often than not .
identifying the cause of this bias is beyond the scope of the current manuscript , so we limit ourself to properly estimating the empirical eclipse uncertainty . ]
distribution @xmath11 \cup [ -\sqrt{\delta^2-\sigma_{\rm tot}}]$ ] has a standard deviation of @xmath12 .
we adopt @xmath13 for the remainder of this paper ( this is somewhat greater than , but broadly consistent with , the repeatability estimate of @xmath14 based on a pair of 3.6 @xmath6 m transits of hd 189733b ; * ? ? ?
the second approach is to consider the distribution of @xmath15 , which amounts to hypothesizing that single - eclipse uncertainties have been under - estimated by a constant factor .
for example , researchers may under - estimate the degree to which the unknown detector model impacts eclipse depth uncertainty ( numerical experiments have shown that most extant methods underestimate occultation error bars in the presence of correlated noise ; * ? ? ?
the standard deviation of the symmetric distribution @xmath16 \cup [ -|\delta|/\sigma_{\rm tot}]$ ] is @xmath17 .
we adopt @xmath18 in the remainder of this paper .
to summarize the previous section , _
spitzer _ has proven capable of photometry better than @xmath0 and many existing eclipse measurements are likely accurate at that level : specifically , those based on multiple eclipses or taken as part of longer phase measurements ( the bolded numbers in table [ thermal_eclipses ] ) .
single - epoch eclipse measurements of the best and brightest targets have _ not _ been repeatable at this level , however .
this is unfortunate because such single - eclipse measurements represent the vast majority of the broadband emission data ( the unbolded numbers in table [ thermal_eclipses ] ) . , but we add an empirical systematic error of @xmath19 in quadrature to each simple - eclipse measurement . in this hypothesis , there is a floor to how precise an eclipse measurement can be without acquiring more data , so modern eclipse measurements are no more accurate than earlier attempts .
eclipse uncertainties based on multiple eclipse measurements , or an eclipse embedded in a phase measurement , are kept unchanged.[realistic_error],scaledwidth=50.0% ] figure [ realistic_error ] shows the distribution of blackbody bic vs. @xmath20 in light of empirical eclipse depth uncertainties .
values based on multiple eclipse measurements , or obtained as part of phase measurements , are taken at face value .
we add a systematic uncertainty of @xmath21 in quadrature to the quoted uncertainties for all single - eclipse measurements . and @xmath22 would be somewhat smaller , but they would have to be applied across the board , leaving our conclusions essentially unchanged . ]
we then re - fit a blackbody and recompute the bic for each planet using these more realistic error bars . , but we inflate the published single - eclipse uncertainties by the empirical factor @xmath18 .
this scenario accounts for the possibility that modern eclipse measurements , which have much smaller quoted uncertainties than the first generation of eclipses , might really be more accurate than their predecessors .
eclipse uncertainties based on multiple eclipse measurements , or an eclipse embedded in a phase measurement , are kept unchanged.[error_factor],scaledwidth=50.0% ] in figure [ error_factor ] we inflate the published uncertainties of single - eclipse measurements by our empirically determined factor of @xmath18 .
we then re - fit a blackbody and recompute the bic for each planet using these more realistic error bars . under the assumption of realistic eclipse uncertainties , hd 189733b
has the most featured emission spectrum and lies in the green region in both figures [ realistic_error ] and [ error_factor ] .
if spectral retrieval could achieve a perfect fit , @xmath23 , then it would be modestly favored as compared to the blackbody , according to the bic .
obtaining such a good fit is not trivial for this planet because even our realistic noise hypothesis takes the published uncertainties at 3.6 , 4.5 , and 8.0 @xmath6 m at face value .
all other planets lie at / below the dashed green line , suggesting that blackbodies are favored , _ even if spectral retrieval provides a perfect fit to the data_. in any case , a researcher who has gone to the trouble of running a markov chain monte carlo to perform spectral retrieval should also estimate the evidence for their model using the posterior distribution ; bic is merely a way of approximating this .
ideally , the evidence for spectral retrieval models with different numbers of parameters could be compared using , for example , a reversible jump markov chain monte carlo @xcite or nested sampling @xcite .
given the small number statistics , we expect a broad range of @xmath24 values with a significant tail ; the gray zone indicates the @xmath25 ( 68.3% ) interval . nonetheless , a few short period planets lie well above the gray region in figures [ realistic_error ] and [ error_factor ] , suggesting they are poorly fit by a blackbody and hence exhibit spectral features .
these features are either the hints of molecular bands , or remaining astrophysical / detector noise . the only planets that make the cut under both the @xmath26 and @xmath22 hypotheses are corot-2b , hd 189733b , and wasp-5b . in order to put the poorly - fitting blackbodies in perspective , we compare them to self - consistent radiative transfer models .
self - consistent atmospheric radiative transfer models typically have between one and three tunable parameters : recirculation efficiency , optical opacity , and relative abundance of co ( e.g. , * ? ? ?
* ; * ? ? ?
* ) and are usually tuned by eye in order to obtain a decent fit . in
what follows we will quote @xmath27 values from the literature ( i.e. , using published eclipse uncertainties ) .
as such , the values should be compared to the blackbody bic values shown in figure [ blackbody_bic ] . as noted by @xcite ,
corot-2b is so poorly fit by spectral models at 4.5 @xmath6 m that a blackbody fit has a smaller @xmath24 .
in fact , @xcite explain the anomalous eclipse depth by invoking emission from a circumstellar accretion disk contaminating the system flux in the mid - infrared at the level of @xmath28 .
@xcite performed spectral retrieval on wasp-5b , but the authors were unable to obtain a good fit that conserved energy , even when they allowed the atmospheric c / o ratio to vary .
it is hard to imagine that a self - consistent radiative transfer model with only two variables would do any better .
the 3.6 @xmath6 m photometry of hd 189733b is @xmath1 discrepant from the best match 1d radiative transfer model obtained by varying two model parameters @xcite .
the mismatch between the predicted and measured flux at 3.6 @xmath6 m contributes @xmath29 to the @xmath27 budget , making this model a far worse fit than a simple blackbody ( @xmath30 , as shown in figure [ blackbody_bic ] ) .
it is likely that bona fide fits using self - consistent radiative transfer models could provide somewhat better @xmath27 , but this is computationally intensive and has only been performed once , to our knowledge @xcite .
a recent wholesale look at all extant eclipse spectra concluded that the only potentially robust area of agreement between self - consistent models and the data was the `` systematic increase in the ratios to shorter wavelengths '' @xcite .
in other words , the planets poorly fit by blackbodies are also poorly fit by self - consistent radiative transfer models .
the radiative transfer models could simply be wrong .
there have been efforts to compare and validate exoplanet radiative transfer codes @xcite and many have been tested against high quality observations of brown dwarfs , but it is possible that they are missing important physics relevant to irradiated planets .
`` missing physics '' includes atmospheric dynamics and clouds , but these are also omitted from most spectral retrieval models .
we therefore hypothesize that the spectral features in extant broadband spectra are due to a combination of astrophysical + detector noise ; spectral retrieval provides better fits because it is under - constrained .
most recent measurements have not yet been repeated , but one could argue that the various advances in reduction and analysis have made modern eclipse measurements more accurate than their predecessors . in hindsight , it is easy to point out poor judgements made by earlier researchers . in all cases ,
however , the authors were making defensible choices about how to treat the data and how to fit it . in no case
has the original paper been retracted or has an erratum been published . with one exception @xcite , researchers have only questioned the original measurements once better observations were available .
researchers still make choices about their reduction scheme , and the intra - pixel sensitivity variations of _
warm spitzer _ are still modeled using the same few hours of data that are used to measure the eclipse depth .
we should aspire to parametrize these choices and marginalize over them to produce accurate , if less precise , measurements .
a promising avenue is to use gaussian processes to model the intrapixel sensitivity variations , which implicitly marginalizes over the functional form of the detector model .
this strategy has been used for transit spectroscopy @xcite and to model the effect of star spots on thermal phase variations @xcite . moreover , none of the studies reporting secondary eclipse measurements account for how the meta - parameters of reduction and analysis pipelines contribute to uncertainty in eclipse depth . at best ,
researchers experiment with a variety of schemes and adopt the one that minimizes the scatter in the photometry @xcite .
this amounts to optimizing the meta - parameters rather than marginalizing over them .
if different choices of meta - parameters , detector parametrization , or astrophysical parametrization lead to significantly different eclipse depths ( see 4.2 ) , then one should be wary of small quoted uncertainties .
the possibility of multimodal posterior distributions should also give us pause , since neither gradient descent ( e.g. , levenberg - marquardt ) nor markov chain monte carlo routines are well suited to finding global solutions under these circumstances . in short , the current generation of single - eclipse measurements are still systematics - dominated and susceptible to many of the same problems as the previous generation . in the @xmath26 hypothesis , there is a noise floor that affects all single - eclipse measurements , so current single - eclipse measurements are little better than the first generation . in the @xmath22 hypothesis , on the other hand ,
the uncertainties are under - estimated by a constant _ factor _ , so single - eclipse measurements published today ( which tend to have small quoted uncertainties ) are taken to be more accurate than their predecessors .
in other words , the @xmath22 hypothesis assumes that eclipse depth estimates are becoming more accurate with time . that @xmath22 can be a function of time , rather than constant ; one could hope that @xmath22 tends to unity , indicating that observers are getting better at estimating the accuracy of their measurements . ]
our results are independent of which hypothesis we choose , as discussed above .
measurement - to - measurement variance in eclipse depths is only sensitive to systematics that change from epoch to epoch : detector behavior , star spots , and exoplanet weather .
there are other systematics , however , that might remain constant from epoch to epoch but that still introduce an error in our estimate of the planetary flux .
wasp-12b is the poster - child for such astrophysical sources of uncertainty , starting with the possibility of contamination from a circumstellar disk @xcite .
a change in astrophysical assumptions namely the strength of ellipsoidal variations affects the 4.5 @xmath6 m eclipse depth of wasp-12b by @xmath31 @xcite .
moreover , published eclipse measurements of wasp-12b have had to be revised after the discovery of a binary companion that diluted the eclipse measurements , leading to eclipse depth increases of @xmath32 to @xmath33 in the near to mid - infrared @xcite . in short ,
even if the photometry for an exoplanet system were precisely known , there is significant room for error in the dayside emission of the planet , which is the quantity we need to know for spectral fitting .
the retrieval of atmospheric structure and composition from disk - integrated broadband photometry hinges on planets not emitting like blackbodies .
we have considered the 44 short - period planets with emission measurements in multiple broadbands .
if published uncertainties are taken at face value , then seven of these planets have broadband spectra that favor spectral retrieval over blackbody fits , according to the bayesian information criterion
corot-2b benefits the most from the additional model parameters . in order to perform under - constrained spectral
retrieval , however , it is critical to know the actual uncertainty on eclipse measurements . _
spitzer _ is capable of exquisite photometry ( @xmath34 ) , but single eclipses acquired , reduced and analyzed in isolation have only been repeatable at the @xmath25 level of @xmath1 ( or single - eclipse uncertainties have been under - estimated by a factor of 3 ) .
if one adopts such empirical uncertainties for single - eclipse measurements , then blackbody fits are preferable over spectral retrieval for all planets , with the possible exception of hd 189733b . we conclude that statements about atmospheric composition based solely on broadband emission measurements are premature .
if one adopts empirical estimate of single - eclipse accuracy , then hd 209458b and gj 436b are well fit by blackbodies , and wasp-12b is not so poorly fit as to favor spectral retrieval .
this resonates with the cautionary review of @xcite .
temperature inversions and odd compositions were inferred for short period planets based on broadband emission spectra @xcite .
our results call these claims into question . undoubtedly , many planets have stratospheric inversions and non - solar chemistry , but there is no robust evidence for this in the current photometry of short - period planets .
cjh was funded by the illinois space grant .
jcs was funded by an nsf gk-12 fellowship .
nbc is indebted to the participants of exopag-9 for discussions of instrument systematics , as well as to m.r .
line and n. madhusudhan for discussions of spectral retrieval .
the authors received useful feedback from s.j .
carey , d. dragomir , h.a .
knutson , k.b .
stevenson , j.n .
winn , and the anonymous mnras referee .
nbc acknowledges the generous hospitality of the institut the plantologie et dastrophysique de grenoble ( ipag ) . this research has made use of the exoplanet orbit database and the exoplanet data explorer at exoplanets.org . | a planet s emission spectrum contains information about atmospheric composition and structure .
we compare the bayesian information criterion ( bic ) of blackbody fits and idealized spectral retrieval fits for the 44 planets with published eclipse measurements in multiple thermal wavebands , mostly obtained with the spitzer space telescope .
the evidence for spectral features depends on eclipse depth uncertainties . _
spitzer _ has proven capable of eclipse precisions better than @xmath0 when multiple eclipses are analyzed simultaneously , but this feat has only been performed four times . it is harder to self - calibrate photometry when a single occultation is reduced and analyzed in isolation ; we find that such measurements have not passed the test of repeatability .
single - eclipse measurements either have an uncertainty floor of @xmath1 , or their uncertainties have been underestimated by a factor of 3 .
if one adopts these empirical uncertainties for single - eclipse measurements , then the evidence for molecular features all but disappears : blackbodies have better bic than spectral retrieval for all planets , save hd 189733b , and the few planets poorly fit by blackbodies are also poorly fit by self - consistent radiative transfer models .
this suggests that the features in extant broadband emission spectra are due to astrophysical and instrumental noise rather than molecular bands .
claims of stratospheric inversions , disequilibrium chemistry , and high c / o ratios based solely on photometry are premature .
we recommend that observers be cautious of error estimates from self - calibration of small data sets , and that modelers compare the evidence for spectral models to that of simpler models such as blackbodies .
[ firstpage ] |
the nambu jona - lasinio model has been extensively studied for several decades [ 1 - 3 ] . in recent years
there has been strong interest in the study of quark matter at high densities , using the njl model and related models . in particular
, one finds color superconductivity under certain conditions and it has been suggested that some compact stars might be made of superconducting quark matter [ 4 - 10 ] .
it is our belief that in such studies one should use a model which reproduces , as well as possible , know features of qcd . in this work
we wish to generalize the su(3)-flavor version of the njl model to be consistent with the type of momentum - dependent quark masses found in lattice simulations of qcd [ 11 ] .
for example , in figs . 1 and 2 we show some of the results obtained in ref . [ 11 ] .
it may be seen from these figures that , in euclidean space , the quark mass goes over to the current mass for euclidean momentum @xmath2 2 gev . on the other hand , the njl model , in the standard analysis [ 1 ] ,
gives rise to a constant value for the constituent quark mass .
as we will see , the form of the nonlocality used here is different from that used in refs .
[ 6 , 12 - 16 ] .
for example , in reference [ 12 ] the @xmath3 vertex is modified by a form factor which depends on the relative momentum of the quark and antiquark , while , in another scheme , a form factor is associated with each quark line appearing in a diagram . in the latter procedure
no further regularization is needed .
however , for the problem considered in this work , the nonlocal models that appear in the literature are of limited applicability , since , in the limit of zero current quark mass , the quark self - energy is proportional to the form factors used to define the nonlocality [ 12 ] . in contrast , in the current work , we calculate the form of the quark self - energy after introducing a regulator and a nonlocal quark interaction .
we stress that the procedure used here differs from any that appears in the literature .
the organization of our work is as follows . in section
ii we review the standard analysis for the condensates and gap equation " of the su(3)-flavor njl model [ 1 ] .
we then go on to review a procedure for including the contribution to the quark self - energy due to the addition of a confining interaction for the case of the su(2)-flavor model . in section iii
we introduce a nonlocal interaction in the su(3)-flavor njl model while maintaining the separable nature of the interaction .
we also describe the approximation used for the t hooft interaction in the nonlocal model . in section iv and v we present some of the results of our numerical calculations . in section
vi we discuss the density dependence of the quark condensate and the momentum dependent quark mass . finally , section vii contains some additional discussion and conclusions . using an extrapolation of the current quark mass to zero .
( the small dip at @xmath4 gev is not statistically significant [ 11 ] . ]
in order to best introduce the nonlocal model , we will first review the calculation of the quark self - energy in the local su(3)-flavor njl model and then proceed to add a confinement interaction , as was done in an earlier study of the quark self - energy [ 17 ] .
we consider the generalization to a model with nonlocal short - range and t hooft interactions in the next section .
the lagrangian of the model is @xmath5\nonumber\\ & & + \frac{g_d}{2}\{\det[\bar q(1+\gamma_5)q]+\det[\bar q(1-\gamma_5)q]\}\,.\end{aligned}\ ] ] here @xmath6 is the matrix of quark current masses , @xmath7 and the @xmath8 are the gell - mann matrices .
further , @xmath9 , with being the unit matrix in the flavor space .
the quark propagator is written as @xmath10 with @xmath11 we may define @xmath12 and @xmath13 with similar definitions for @xmath14 and @xmath15 . in the absence of a confinement model , we have @xmath16 and @xmath17 , where @xmath18 and and @xmath19 are constants .
( here , we have take @xmath20 . ) in this case we have [ 2 ] @xmath21 @xmath22 @xmath23 these equations are depicted in fig .
3a , where the last term represents the t hooft interaction .
the up quark vacuum condensate is given by @xmath24 here , @xmath25 is a function needed to regulate the integral . in this work
we will use the pauli - villars procedure and evaluate the integral in euclidean space , as was done in ref . [
17 ] . in the general case
we may write @xmath26 where @xmath27 and @xmath28 were defined in eqs .
( 2.4 ) and ( 2.5 ) . in the appendix of ref .
[ 17 ] we considered lorentz - vector confinement , with @xmath29 and @xmath30 ^ 2 } -\frac{4\mu^2}{[(k_e - k_e^{\,\prime})^2+\mu^2]^3}\right\}\,,\ ] ] where @xmath31 denotes the euclidean - space momentum transfer . here
, @xmath32 is a small parameter introduced to soften the momentum - space singularities .
note that the form @xmath33 ^ 2 } -\frac{4\mu^2}{[(\vec k-\vec k^{\,\prime})^2+\mu^2]^3}\right\}\,,\ ] ] represents the fourier transform of @xmath34 , so that for small @xmath35 approximates a linear potential over the relevant range of _ r_. in fig .
3b we show the equation for the self - energy when the confining field is included . in ref .
[ 17 ] we obtained the following coupled equations in the case of the su(2)-flavor model @xmath36a(\kpsq)}{k^{\,\prime \,2}[1-b(k^{\,\prime \,2})]^2-a^2(k^{\,\prime \,2})+i\epsilon}\,,\ ] ] @xmath37v^c(k - k^{\,\prime})}{k^{\,\prime \,2}[1-b(k^{\,\prime \,2})]^2-a^2(k^{\,\prime \,2})+i\epsilon}\,.\ ] ] these equations were solved after passing to euclidean space and including a pauli - villars regulator of the form @xmath38[k_e^{\,2}+a^2(k_e^{\,2})+2\lambda^2]}\ ] ] in euclidean space .
note that the form @xmath39[k_e^{\,2}\left(1-b(k_e^{\,2 } ) \right)^2+a^2(k_e^{\,2})+2\lambda^2]}\ ] ] may also be used .
( 2.15 ) and ( 2.16 ) @xmath40 appears with a sign opposite to that given in ref . [
17 ] , since , in that work , we used a negative value of @xmath41 . for the present work we use a positive value of @xmath41 to be consistent with all of our other publications . )
in this section we describe the procedures we use to create a nonlocal version of our generalized njl model .
we consider the second term of eq .
( 2.1 ) and make the replacement @xmath42\\\nonumber & & \longrightarrow\frac{g_s}{2}\sum_{i=0}^8\{[\bar q(x)\lambda^if(x)q(x)\cdot \bar q(y)\lambda^if(y)q(y)\\[0.4cm]\nonumber & & \hspace{0.9cm}+[\bar q(x)i\gamma_5\lambda^if(x)q(x)\cdot\bar q(y)i\gamma_5\lambda^if(y)q(y)]\}\,.\end{aligned}\ ] ] this replacement corresponds to the use of a separable interaction @xmath43 .
a related modification may be made for the t hooft interaction .
it is useful , however , to describe these modifications as they affect momentum - space calculations . with reference to fig .
4 , we replace @xmath44 by @xmath45 in the evaluation of the second term of fig .
2 we need @xmath46 and choose to write @xmath47 g_s\exp[-(k - k^{\,\prime})^{\,2n}/2\beta]\\\nonumber & = & \:g_s\exp[-(k - k^{\,\prime})^{\,2n}/\beta]\,.\end{aligned}\ ] ] in this work we take @xmath48 and @xmath49 gev@xmath50 . in fig .
5 we exhibit the function @xmath51 $ ] .
it is clear that many other functions may be chosen .
we now rewrite eqs .
( 2.15 ) and ( 2.16 ) for the up , down and strange quarks . for example , for the su(3)-flavor case @xmath52a_u(k^{\,\prime\,2 } ) } { k^{\,\prime\,2}[1-b_u(k^{\,\prime\,2})]^2-a_u^2(k^{\,\prime\,2})+i\epsilon}\ , , \\[0.35 cm ] k^2b_u(k^2)&=&i\myint{\kp } \frac{2(k\cdot\kp)[1-b_u(\kpsq)]v^c(k-\kp ) } { k^{\,\prime\,2}[1-b_u(k^{\,\prime\,2})]^2-a_u^2(k^{\,\prime\,2})+i\epsilon}\,,\end{aligned}\ ] ] with similar equations for @xmath53 etc . again
, these equations are solved after passing to euclidean space and introducing regulator functions : @xmath54 . in euclidean space
we write c_u()= , etc .
note that without the t hooft interaction the equations for the up , down and strange quarks are uncoupled .
, by the nonlocal ( separable ) term , @xmath55 .
the distinction between our separable model and that of ref .
[ 12 ] , for example , is that in ref .
[ 12 ] the alternative replacement @xmath56 was used . ]
$ ] is shown for @xmath48 and @xmath49 . ] our treatment of the t hooft interaction is based upon a generalization of the last term in eqs .
( 2.6)-(2.8 ) .
with reference to the third term on the right in fig .
6 , we introduce a correlation between the quark of momentum _
k _ and the quark of momentum @xmath57 .
we also include a correlation between the quark of momentum _ k _ and that of momentum @xmath58 .
( therefore , our procedure does not introduce a correlation between the two quarks in the separate condensates . at this stage of the development of our model that seems to be
a reasonable approximation and avoids having to define a three - quark correlation function , @xmath59 ) for example , we generalize the term @xmath60 to obtain the contribution to @xmath61:a_u^t(k)= -g_d + in minkowski space .
this expression is then evaluated in euclidean space and added to the right - hand side of eq .
( 3.3 ) . in a similar fashion
, we calculate @xmath62 and @xmath63 .
there is a good deal of flexibility in choosing the regulators @xmath64 , @xmath65 , and @xmath66 .
also , various forms could be chosen for the correlation functions , @xmath67 .
previously , in our minkowski - space studies of the @xmath1 mesons we used @xmath68 gev@xmath69 and @xmath70 [ 18 ] . however , in that work we used a gaussian regulator in minkowski space so that a direct comparison with the present study can not be made .
on the other hand , we do not expect to find radically different parameters , if the constituent masses in the two calculations are similar .
for example , in our earlier work , in which @xmath18 and @xmath19 were parameters , we used @xmath71 , which can be compared to the value of @xmath72 calculated here .
we have also used either @xmath73 [ 19 - 23 ] or @xmath74 [ 18 ] , values which may be compared to @xmath75 . to proceed ,
we take @xmath76 , @xmath77 , @xmath78 , @xmath79 , @xmath80 , @xmath81 and @xmath82 .
we then consider values of @xmath83 , @xmath84 , @xmath85 and @xmath86 .
the results of our calculations are given in table i. recall that the function @xmath87 does not appear in our expression for the condensates .
the calculation of the condensates includes the pauli - villars regulators , @xmath64 , @xmath65 and @xmath66 , however .
[ see eqs .
( 2.9)-(2.11 ) . ] in our calculation of the properties of the @xmath1 mesons [ 18 ] we had @xmath88 , since we used @xmath89 values in the range @xmath90@xmath91 in that work . in order to specify a value of @xmath89 for this work
, we note that a calculation based upon chiral perturbation theory yields @xmath92 [ 24 ] .
inspection of table i suggests that the values of @xmath89 , other than @xmath83 , given in table i are acceptable .
for @xmath93 we have @xmath94 and @xmath95 , which are reasonably close to the phenomenological parameters @xmath96 and @xmath97 used in our earlier work [ 19 - 23 ] .
it is worth noting that , in standard application of the su(3)-flavor njl model , one finds @xmath98 [ 1 ] , so that the results shown in table i are encouraging , given that the value for @xmath99 obtained using chiral perturbation theory is about 1.7 [ 24 ] , as noted above .
.calculated values for the condensates and for @xmath100 are given for the up and strange quarks for four values of @xmath89 .
the parameters @xmath79 gev , @xmath80 gev , @xmath782 , @xmath498 , @xmath81 gev , @xmath76 gev , @xmath77 were used .
the values of @xmath41 and @xmath32 were fixed in earlier work [ 18 - 23 ] .
values of @xmath101 have been suggested [ 28 ] , so we see that our calculated values are at , or near , the lower limit for that quantity . [
cols="<,<,<,<,<",options="header " , ]
in fig . 7 we show @xmath102 , where _ k _ is the magnitude of the euclidean momentum .
the dashed line exhibits the result without the confining interaction ( @xmath103 ) .
it is interesting to see that inclusion of confinement improves the shape of the curve when we compare our results to the lattice results shown in figs . 1 and 2 .
we note that @xmath102 goes over to @xmath79 for large _ k_. in fig .
8 we show @xmath104 .
( recall that @xmath105^{-1}$ ] . )
we remark that @xmath106 when @xmath103 . in figs . 9 and 10
we show @xmath107 and @xmath108 , respectively . as expected , we find that @xmath107 goes over to @xmath80 when _ k _ is large .
are shown for the parameters @xmath79 gev , @xmath80 gev , @xmath78 , @xmath81 gev , @xmath76 gev , @xmath49 , @xmath109 , and @xmath93 .
the dashed line shows the result without confinement ( @xmath103 ) . ]
are shown .
( see caption to fig .
] are shown .
( see caption to fig . 7 . ) ] are shown .
( see caption to fig . 7 . )
as stated earlier , the behavior of the njl model for finite values of the baryon density is an extensively explored topic [ 1 , 25 , 26 ] , with particular recent emphasis on color superconductivity [ 4 - 10 ] . in this section
we explore the behavior of our model at finite baryon density .
( it should be noted that the nature of the phase transition describing chiral symmetry restoration at finite density is quite model dependent .
for example , the inclusion of current quark masses can change a strong first - order transition to a smooth second - order transition [ 26 ] . )
a comprehensive study of the thermodynamics of the three - flavor njl model has been reported in ref . [ 27 ] . there
it is found that the up , down and strange quark masses are essentially constant up to the density where a first - order phase transition appears . at that point ,
the up and down quark masses drop from a value of about 380 mev to about 30 mev .
that behavior differs from the behavior expected at low density .
for example , we have the well - known relation between the value of the condensate and the baryon density of nuclear matter = 1- , where @xmath110 is the pion - nucleon sigma term .
this relation is valid to first - order in the density .
it may be derived , in the case of nuclear matter , by writing |qq_= |qq_0+n||qq|n_b and making use of the definition of the pion - nucleon sigma term , @xmath110 , and the gell - mann oakes renner relation . if we put @xmath111 gev , we have @xmath112 , where @xmath113 is in units . for nuclear matter
@xmath114(0.109 gev)@xmath115 , so we see that the condensate is reduced by about 35% , if we evaluate eq .
( 6.1 ) at nuclear matter density .
we can check whether the density dependence given by eq .
( 6.1 ) is reproduced in our model , since it should not matter whether the scalar density of the background matter is generated by quarks in nucleons or by the presence of free quarks . in the former case ,
we may write , for the baryon density , _ b=4^k_f where the factor of 4 is arises from the product of the spin and isospin factors . in the case of quarks
, we have _ b=4n_c ( ) ^k_f where , in this case , the factor of 4 again arises from the spin and isospin factor .
( both up and down quarks are present in equal numbers . ) the color factor , @xmath116 , is cancelled by the baryon number of 1/3 of each quark .
we need to modify the equations for the quark self - energy to take into account the presence of the fermi seas of up and down quarks whose fermi momentum is @xmath117 .
we take one fermi sea to be composed of on - mass - shell up quarks with constituent mass @xmath72 .
the following term is then added to the equation for @xmath61 .
a_u^()(k)=-(2g_s)n_c2^k_f f^2(k- ) where @xmath118^\frac{1}{2}$ ] .
the second factor of 2 in eq .
( 6.5 ) reflects the spin degeneracy .
we note that @xmath72 is density - dependent and could be written as @xmath119 in keeping with the labelling of fig . 11 , where @xmath120 is used .
also , if @xmath121 , @xmath122 would then represent @xmath123 , where @xmath124 is the scalar density associated with the up - quark fermi sea . in fig .
11 we show @xmath125 calculated for four values of @xmath126 and in fig .
12 we show _
m_(0 ) as a function of @xmath127 . since the quarks in the fermi sea are taken to be on - mass - shell , we would , in principle , require @xmath120 in minkowski space . however , since @xmath117 = 0.268 gev for the case of nuclear matter , only a very modest extrapolation of the curves shown in fig .
11 is needed for the densities considered in this work . in fig .
13 we show the value of the up quark condensate as a function of @xmath127 .
( note that @xmath128 represents the density of nuclear matter . )
it is seen that , for small values of the density , the density dependence of the condensate reproduces what is expected from eq . ( 6.1 ) .
( if we extrapolate the curve using a linear approximation , the condensate is reduced by about 30% at nuclear matter density . )
are shown for @xmath129 [ dotted line ] , @xmath130 [ dashed line ] , and @xmath131 [ dash - dot line ] .
] are shown as a function of @xmath127 .
note that @xmath132 .
note that @xmath132
we have remarked earlier in this work that the large values of the condensate ratio @xmath99 seen in table i play a role in obtaining a good fit to the mixing angles of the @xmath1(947 ) and @xmath133(958 ) mesons [ 18 ] . to understand this remark
we note that the effective singlet - octet coupling constants for pseudoscalar states are [ 3 ] g_00^p&=&g_s-(++ ) , + g_88^p&=&g_s-(-2 - 2 ) , and g_08^p=-(2 - - ) , where @xmath134 , @xmath135 and @xmath136 .
we take @xmath137 , so that g_08^p=-(-).if @xmath138 , the result for @xmath139 is six times larger than when @xmath140 .
in addition to the effects of @xmath139 , singlet - octet mixing is induced by the quantity [ 18 ] e_08(k)=[e_u(k)-e_s(k)],where @xmath141^\frac{1}{2}$ ] , etc .
it is found that , since @xmath139 and @xmath142 tend to cancel in our formalism , the significant singlet - octet mixing generated by @xmath142 is reduced by the values of @xmath139 obtained for the larger value of the ratio @xmath143 , with the result that we reproduce the values of the mixing angles found in other studies that make use of experimental data to obtain values for the mixing angles [ 18 ] . in our earlier work , which was carried out in minkowski space ,
the values of @xmath144 and @xmath19 were taken as parameters .
inspection of our figures which exhibit values of @xmath102 and @xmath107 suggests that an extrapolation into minkowski space may be made if @xmath145 is not too large .
the fact that @xmath72 and @xmath75 are close to our phenomenological parameters for @xmath93 is encouraging and suggests that some support for our choice of quark mass parameters may be found in our euclidean - space analysis .
the full consequences of separating the specification of the nonlocality of the quark interaction from the choice of the regulator of the theory should be explored more fully .
although that feature of our model introduces greater flexibility , that comes with the disadvantage of having to introduce other parameters in the model .
we have made only limited variation of the form of the nonlocality and the regulator .
for further applications it may be of interest to explore a more comprehensive parameter variation .
it is also necessary to extend the calculations reported in figs .
1113 to larger values of the density than those considered here .
that step will require more complex methods for solving our nonlinear equations for the self - energy than the simple iteration scheme we have used thus far .
our work may be compared to that of alkofer , watson and weigel [ 29 ] who have solved the schwinger - dyson equation using a gluon propagator whose low - momentum behavior is enhanced by a gaussian function .
( that modification requires the introduction of two phenomenological parameters [ 30 ] . )
the behavior found for @xmath146 and @xmath147 in euclidean space is similar to that obtained in this work .
. 1 of ref .
those authors also solve the bethe - salpeter equation to obtain the properties of various @xmath148 mesons with generally satisfactory results .
it is of interest to note that the minkowski space solution for @xmath146 and @xmath147 is such that the quark can go on - mass - shell .
that feature may be related to our work [ 18 - 23 ] in which we use on - mass - shell quarks with masses @xmath149 gev and @xmath73 gev ( or 0.585 gev [ 18 ] ) when solving the bethe - salpeter equation in our study of @xmath150 mesons .
99 s. p. klevansky , rev .
@xmath152 , 649 ( 1992 ) .
u. vogl and w. weise , prog .
@xmath153 , 195 ( 1991 ) .
t. hatsuda and t. kunihiro , phys .
@xmath154 , 221 ( 1994 ) . for reviews ,
see k. rajagopal and f. wilcek , in b. l. ioffe festscrift , at the frontier of particle physics / handbook of qcd , m.shifman ed .
( world scientific , singapore , 2001 ) ; + m. alford , hep - ph/0102047 .
m. alford , j. berges and k. rajagopal , nucl . phys .
b @xmath155 , 219 ( 1999 ) ; + j. kundu and k. rajagopal , hep - ph/0112206 ( 2002 ) . c. gocke , d. blaschke , a. khalatyan and h. grigoria , hep - ph/0104183-v2 ( 2002 ) . i. a. shovkovy , hep - ph/0110352 ( 2002 ) . m. alford , r. rajagopal and f. wilcek , phys .
b @xmath156 , 247 ( 1998 ) .
d. t. son , phys . rev .
d @xmath157 , 094019 ( 1999 ) .
t. sh@xmath158fer , e. v. shuryak and m. velkovsky , phys .
@xmath159 , 53 ( 1998 ) .
j. skullerud , d. b. leinweber , and a. g. williams , phys . rev .
d @xmath152 , 074508 ( 2001 ) .
h. ito , w. w. buck and f. gross , phys .
c @xmath160 , 2483 ( 1991 ) ; c @xmath161 , 1918 ( 1992 ) .
s. schmidt , d. blaschke , y. l. kalinovsky , phys .
c @xmath162 , 435 ( 1994 ) .
r. s. plant and m. c. birse , nucl .
a @xmath163 , 607 ( 1998 ) .
r. d. bowler and m. c. birse , nucl .
a @xmath164 , 655 ( 1995 ) .
r. s. plant and m. c. birse , hep - ph/0007340 and references therein .
l. s. celenza , xiang - dong li , and c. m. shakin , phys .
c @xmath165 , 1492 ( 1997 ) .
+ in this reference a negative value of @xmath41 was used .
therefore , the sign on the right - hand sides of eqs .
( 2.4 ) and ( 2.5 ) , appearing in this reference , should be corrected to be positive .
c. m. shakin and huangsheng wang , role of the t hooft interaction in the calculation of the mixing angles of the @xmath1(547 ) and @xmath166(958 ) mesons , brooklyn college + report no .
bccnt : 01/082/307 ( 2001 ) . to be published in physical rev .
c [ dh8173 ] . c. m. shakin and huangsheng wang , phys .
d @xmath167 , 014019 ( 2000 ) .
l. s. celenza , huangsheng wang , and c. m. shakin , phys .
c @xmath167 , 025209 ( 2001 ) .
c. m. shakin and huangsheng wang , phys .
d @xmath167 , 074017 ( 2001 ) . c. m. shakin and huangsheng wang , phys .
d @xmath167 , 114007 ( 2001 ) . c. m. shakin and huangsheng wang , phys .
d @xmath152 , 094020 ( 2001 )
. g. amoros , j. bijnens , and p. talavera , nucl .
b @xmath168 , 87 ( 2001 ) . m. asakawa and k. yazaki , nucl .
a @xmath169 , 668 ( 1989 ) .
v. bernard , ulf - g .
meissner , and i. zahed , phys .
d @xmath170 , 819 ( 1987 ) . f. gastineau , r. nebauer , and j. aichelin , hep - ph/0101289 ( 2001 ) .
m. shifman , a. vainstein and v. zakarov , nucl .
b @xmath171 , 385 ( 1979 ) ; 448 ( 1979 ) ; + l. reinders , h. rubinstein and y. yazaki , phys . rep .
@xmath172 , 2 ( 1985 ) .
r. alkofer , p. watson and h. weigel , hep - ph/0202053 ( 2002 ) .
p. maris and p. c. tandy , phys .
c @xmath173 , 055214 ( 1999 ) . | the nambu jona - lasinio ( njl ) model has been extensively studied by many researchers . in previous work
we have generalized the njl model to include a covariant model of confinement . in the present work we consider further modification of the model so as to reproduce the type of euclidean - space momentum - dependent quark mass values obtained in lattice simulations of qcd .
this may be done by introducing a nonlocal interaction , while preserving the chiral symmetry of the lagrangian . in other work on nonlocal models , by other researchers , the momentum dependence of the quark self - energy is directly related to the regularization scheme .
in contrast , in our work , the regularization is independent of the nonlocality we introduce .
it is of interest to note that the value of the condensate ratio , @xmath0 , is about 1.7 when evaluated using chiral perturbation theory and is only about 1.1 in standard applications of the njl model .
we find that our nonlocal model can reproduce the larger value of the condensate ratio when reasonable values are used for the strength of the t hooft interaction .
( in an earlier study of the @xmath1(547 ) and @xmath1(958 ) mesons , we found that use of the larger value of the condensate ratio led to a very good fit to the mixing angles and decay constants of these mesons . )
we also study the density dependence of both the quark condensate and the momentum - dependent quark mass values . without the addition of new parameters , we reproduce the density dependence of the condensate given by a well - known model - independent expression valid for small baryon density .
the generalization of our model to include a model of confinement required the introduction of an additional parameter .
the further generalization to obtain a nonlocal model also requires additional parameters .
however , we believe our results are of sufficient interest so as to compensate for the introduction of the additional parameters in our formalism . |
anomalous transverse momentum and net charge event - by - event fluctuations have been proposed as indicators of the formation of a quark gluon plasma ( qgp ) in the midst of high - energy heavy ion collisions .
entropy conserving hadronization of a plasma of quarks and gluons could produce a final state characterized by a dramatic reduction of the net charge fluctuations relative to that of a hadron gas @xcite .
measurements of net charge fluctuations were reported by both phenix @xcite and star @xcite collaborations on the basis of @xmath0 data acquired during the first rhic run at @xmath8 130 gev . in this work ,
we report a systematic study of the net charge fluctuations as a function of the system size and their energy using @xmath9 fluctuation measure , defined as follows : @xmath10 where @xmath11 are the number of positively and negatively charged particles in the acceptance of interest .
we present measurements of dynamical net charge fluctuations in @xmath0 collisions at @xmath1 19.6 , 62.4 , 130 , and 200 gev , @xmath12 collisions at @xmath1 62.4 , 200 gev and in @xmath3 collisions at @xmath1 200 gev .
we study the beam energy , system size and collision centrality dependencies quantitatively in order to identify possible signature of the formation of a qgp .
the data used in this analysis were measured using the solenoidal tracker at rhic ( star ) detector during the 2001 , 2002 , 2004 and 2005 data rhic runs at brookhaven national laboratory .
the @xmath13 collisions at @xmath8 62.4 , 130 , and 200 gev data and @xmath2 collisions at @xmath8 62.4 and 200 gev were acquired with minimum bias triggers . for 19.6 gev data ,
a combination of minimum bias and central triggers was used .
technical descriptions of the star detector and its components are published in technical reports @xcite .
this analysis used tracks from the tpc with transverse momentum in the range @xmath14 gev / c with pseudorapidity @xmath15 .
the centrality bins were calculated as a fraction of this multiplicity distribution starting at the highest multiplicities .
the ranges used were 0 - 5% ( most central collisions ) , 5 - 10% , 10 - 20% , 20 - 30% , 30 - 40% , 40 - 50% , 50 - 60% , 60 - 70% , and 70 - 80% ( most peripheral ) for @xmath0 collisions .
similarly , collision centrality slices used in @xmath2 collisions are 0 - 10% ( most central ) , 10 - 20% , 20 - 30% , 30 - 40% , 40 - 50% and 50 - 60% ( most peripheral ) .
each centrality bin is associated with an average number of participating nucleons , @xmath16 , using glauber monte carlo calculation @xcite .
we present , in fig .
[ fig1 ] , measurements of the dynamical net charge fluctuations , @xmath9 , as a function of collision centrality in @xmath17 collisions at @xmath8 19.6 , 62.4 , 130 , and 200 gev , @xmath2 collisions at @xmath8 62.4 and 200 gev .
the dynamical net charge fluctuations , in general , exhibit a monotonic dependence on the number of participating nucleons . at a given number of participants the measured fluctuations also exhibit a modest dependence on beam energy , with @xmath9 magnitude being the largest in @xmath0 collisions at @xmath1 19.6 gev .
the @xmath9 values measured for @xmath3 collisions at @xmath4 200 gev amounts to -0.230 @xmath18 0.019(stat ) .
dependence on the number of participating nucleons can be seen in fig .
[ fig2](b ) , discussed in section iii a. fig . [ fig2](a ) shows the dynamical fluctuations scaled by the measured particle multiplicity density in pseudorapidity space ( @xmath19 ) .
data from @xmath0 and @xmath2 collisions are shown with solid and open symbols , respectively .
values of @xmath19 used for the scaling correspond to efficiency corrected charged particle multiplicities measured by star @xcite and phobos @xcite .
the magnitude of @xmath9 scaled by @xmath19 for @xmath0 200 gev data is significantly different from the results of other systems and energies . the observed @xmath20 increases with the increase in collision centrality . the dashed line in the figure corresponds to charge conservation effect and the solid line to the prediction for a resonance gas .
the data indicates that dynamical net charge fluctuations , scaled by @xmath19 for the most central collisions are qualitatively consistent with resonance gas prediction @xcite .
[ fig2](b ) presents the dynamical fluctuation scaled by the number of participants , @xmath21 as a function of the number of participants .
the measured data scaled by the number of participants ( @xmath16 ) are thus consistent with no or a very weak collision centrality dependence . however , a definite system size and energy dependence is observed .
vertical error bars represent statistical uncertainties . , of particles produced with pseudorapidity @xmath22 scaled by ( a ) the multiplicity , @xmath19 , ( b ) the number of participants .
the dashed line corresponds to charge conservation effect and the solid line to the prediction for a resonance gas.,width=283 ] , of particles produced with pseudorapidity @xmath22 scaled by ( a ) the multiplicity , @xmath19 , ( b ) the number of participants . the dashed line corresponds to charge conservation effect and the solid line to the prediction for a resonance gas.,width=275 ] we plot in fig .
[ fig3 ] values of @xmath23 measured for different ranges of pseudorapidity , @xmath24 , normalized by the magnitude of @xmath23 for a pseudorapidity range @xmath25 ( @xmath26 ) to enable comparison of different centralities , beam energies and system size .
the magnitude of the normalized correlation is maximum for the smallest pseudorapidity ranges and decreases monotonically to unity , at all energies and centralities , with increasing pseudorapidity range .
this shows that the collision dynamics in @xmath3 collisions , 0 - 10% @xmath2 and 0 - 5% @xmath0 collisions are significantly different .
indeed , we find the relative magnitude of the correlations measured for @xmath22 increases by nearly 25% for @xmath0 200 gev relative to those in @xmath3 .
note in particular that the slope ( @xmath27 ) in @xmath3 , @xmath2 and @xmath0 systems depends on the correlation length ( in pseudorapidity ) : the shorter the correlation , the larger the slope .
the observed distributions then indicate that the correlation length is shorter for central collisions and for larger systems , in agreement with the observed reduction of the charge balance function @xcite .
, as a function of the integrated azimuthal range @xmath28 for similar number of participating nucleons for @xmath0 200 gev and @xmath2 200 gev .
[ fig4],width=274 ] , as a function of the integrated azimuthal range @xmath28 for similar number of participating nucleons for @xmath0 200 gev and @xmath2 200 gev .
[ fig4],width=274 ] fig .
[ fig4 ] displays measured values of @xmath9 as a function of azimuthal angle ranges from 10 to 360 degrees for @xmath0 and @xmath29 collisions at 200 gev . @xmath0 and
@xmath30 collisions are compared at similar number of participating nucleons .
one observes that the magnitude of @xmath9 with respect to azimuthal angle , @xmath28 , is similar for similar number of participating nucleons in both systems with best agreement for collisions with more than 20 participants .
we have presented measurements of dynamical net charge fluctuations in @xmath0 collisions at @xmath8 19.6 , 62.4 , 130 , 200 gev , @xmath2 collisions at @xmath8 62.4 , 200 gev and @xmath3 collisions at @xmath8 200 gev , using the measure @xmath9 .
the fluctuations are non vanishing at all energies and exhibit a rather modest dependence on beam energy in the range 19.6 @xmath31 @xmath32 @xmath31 200 gev for @xmath0 as well as @xmath2 collisions .
net charge dynamical fluctuations violate the trivial @xmath6 scaling expected for nuclear collisions consisting of independent nucleon - nucleon interactions . scaled dynamical net charge fluctuations @xmath20 grow by up to 40% from peripheral to central collisions .
this centrality dependence may arise , in part due to the large radial collective flow produced in @xmath0 collisions .
we also studied fluctuations as a function of azimuthal angle and pseudorapidity and found that dynamical fluctuations exhibit a strong dependence on rapidity and azimuthal angular ranges which could be attributed in part to radial flow effects .
+ 50 s. jeon and v. koch , phys .
lett . * 85 * , 2076 ( 2000 ) ; h. heiselberg and a. d. jackson , phys . rev . *
c63 * , 064904 ( 2001 ) ; m. asakawa , u. heinz , b. mueller , phys .
* 85 * , 2072 ( 2000 ) .
( phenix collaboration ) , phys .
89 * , 082301 ( 2002 ) .
j. adams _ et al . _
( star collaboration ) , phys . rev . *
c68 * , 044905 ( 2003 ) .
m. anderson _
et al . _ ,
a499 * , 624 ( 2003 ) ; nucl .
a499 * , 659 ( 2003 ) .
j. adams _ et al .
_ , ( star collaboration ) , [ nucl - ex/0311017 ] .
et al . _ , ( star collaboration ) , phys .
lett . * 87 * , 112303 ( 2001 ) . b. b. back _ et al . _ , ( phobos collaboration ) , phys . rev
. lett . * 87 * , 102303 ( 2001 ) ; phys .
88 * , 022302 ( 2002 ) ; phys . rev . *
c65 * , 061901r ( 2002 ) ; phys . rev . *
c70 * , 021902 ( 2004 ) ; phys .
lett . * 91 * , 052303 ( 2003 ) ; phys .
rev . * c74 * , 021901(r ) ( 2006 ) ; [ nucl - ex/0601026 ] .
j. adams _ et al .
_ , ( star collaboration ) , phys . rev
90 * , 172301 ( 2003 ) . | 1.0 cm 1.0 cm we present measurements of net charge fluctuations in @xmath0 collisions at @xmath1 19.6 , 62.4 , 130 , and 200 gev , @xmath2 collisions at @xmath1 62.4 , 200 gev , and @xmath3 collisions at @xmath4 200 gev using the net charge dynamical fluctuations measure @xmath5 .
the dynamical fluctuations are non - zero at all energies and exhibit a rather modest dependence on beam energy .
we find that at a given energy and collision system , net charge dynamical fluctuations violate @xmath6 scaling , but display approximate @xmath7 scaling .
we observe strong dependence of dynamical fluctuations on the azimuthal angular range and pseudorapidity widths . |
helpful discussions with krisztian palotas are gratefully acknowledged .
the author thanks the royal society london for support .
99 g. binnig , h. rohrer , ch .
gerber , e. weibel , phys .
49 , 57 ( 1982 ) .
g. binnig , h. rohrer , rev .
59 , 615 ( 1987 ) .
e. j. heller , m. f. crommie , c. p. lutz , d. m. eigler , nature 394 , 464 ( 1994 ) . h. gawronski , m. mehlhorn , k. morgenstern , k. , science 319 , 930 ( 2008 ) .
n. neel et al . , phys .
98 , 016801 ( 2007 ) .
k. r. harikumar et al .
, nature nanotechnology 3 , 222 ( 2008 ) .
w. a. hofer , a. s. foster , a. l. shluger , rev .
75 , 1287 ( 2003 ) .
z. t. deng , et al .
96 , 156102 ( 2006 ) .
j. tersoff , d. r. hamann , phys .
b 31 , 805 ( 1985 ) . k .- h .
rieder et al . , phil .
lond . a 362 , 1207 ( 2004 ) .
p. hohenberg , w. kohn , phys .
136 , b864 ( 1964 ) .
w. kohn , l. j. sham , phys .
140 , a1133 ( 1965 ) .
m. born , zeitschr .
37 , 863 ( 1926 ) ; 38 , 803 ( 1926 )
. w. heisenberg , zeitschr .
43 , 172 ( 1927 ) .
werner a. hofer , found .
41 , 754 ( 2011 ) .
e. schrdinger , phys .
28 , 1049 ( 1926 ) .
w. heisenberg , zeitschr .
33 , 879 ( 1925 ) .
james l. park and henry margenau , intl
. j. theor .
phys . 3 , 211 ( 1968 ) .
see in particular page 213 :
many gedankenexperiments have been designed to illustrate heisenberg s famous law ; unfortunately , the false impression is often conveyed that his principle , which is actually a theorem about standard deviations in collectives of measurement results , imposes restrictions on _ measurability_. [ italics in the original text ] a. j. heinrich , j. a. gupta , c. p. lutz , d. m. eigler , science 306 , 466 ( 2004 ) .
david bohm , phys .
85 , 166 ( 1952 ) ; 85 , 180 ( 1952 ) . l. de broglie , non - linear wave mechanics : a causal interpretation .
elsevier , amsterdam ( 1960 ) .
david hestenes , found .
40 , 1 ( 2010 ) .
j. c. lian , et al .
phys . rev .
b 81 , 195411 ( 2010 ) .
a. khrennikov , europhys .
90 , 40004 ( 2009 ) .
g. groessing , s. fussy , j. mesa pascasio , h. schwabl , j. phys .
306 , 012040 ( 2011 ) .
elze , j. phys .
171 , 012034 ( 2009 ) .
g. t hooft , j. phys .
67 , u166 ( 2007 ) . | we show by a statistical analysis of high - resolution scanning tunneling microscopy ( stm ) experiments , that the interpretation of the density of electron charge as a statistical quantity leads to a conflict with the heisenberg uncertainty principle . given the precision in these experiments we find that the uncertainty principle would be violated by close to two orders of magnitude , if this interpretation were correct . we are thus forced to conclude that the density of electron charge is a physically real , i.e. in principle precisely measurable quantity .
keywords : scanning tunneling microscope , electron charge , density functional theory , uncertainty relations today , stms have reached a level of precision which is quite astonishing . while it was barely possible to resolve the positions of single atoms in early experiments @xcite ,
it is now routine not only to resolve the atomic positions but also e.g. the standing wave pattern of surface state electrons @xcite on metal surfaces , and even very subtle effects like inelastic excitations , kondo resonances , or surface charging @xcite . what the stm measures , in these experiments , is the current between a surface and a very sharp probe tip .
the current itself is proportional to the density of electron charge at the surface @xcite .
while one may dispute this claim for some special cases , and while it can be shown for specific situations that an explicit simulation of the scattering process does improve the agreement between experiment and theory ( see , for example @xcite ) , in measurements on metal surfaces the bias range is so low and the dominance of single electron - states at the tip so high , that the tersoff - hamann approximation @xcite , which assumes tunneling into a single tip - state with radial symmetry , is a very good approximation .
then the map of tunneling currents at a surface is , but for a constant , equal to the map of electron charge densities at the same surface .
a standard deviation of the density of charge due to the uncertainty of position and momentum can thus be mapped identically onto a standard deviation of the tunneling current , which can immediately be compared to experimental results . from the viewpoint of a physicist of the golden age of quantum mechanics ,
say the time around 1925 , the precision of stm measurements , where the vertical resolution of the best instruments is close to fifty times smaller ( about 0.05 pm @xcite ) than the compton wavelength of an electron ( or about 2.4 pm ) , must seem like magic . in figure 1 we show a quantum corral of 51 silver atoms on an ag(111 ) surface measured by rieder et al .
@xcite . in some images
the instrument was able to resolve not only the standing wave pattern of surface state electrons , but also the modulation due to the positions of single surface atoms .
the measurement is quite spectacular , and it raises some very fundamental problems in quantum mechanics , which we shall explore in the following .
it is irrelevant for this analysis , whether the remaining experimental imprecision is due to microscopic or macroscopic processes : it is the level of precision in these experiments itself , which is problematic . in density
functional theory ( dft ) a many - electron system is comprehensively described by the density of electron charge @xcite .
however , the density itself , within the framework of second quantization , is thought to be a statistical quantity @xcite . in principle
, this statement can be tested by a statistical analysis of high - resolution stm measurements including the uncertainty relations @xcite .
the reason that such an analysis at this point is necessary is the following : in a recent paper @xcite it was found that an extended model of electrons fully recovers the formulation of quantum mechanics in terms of schrdinger s equation @xcite .
this leads to a very fundamental issue in quantum mechanics .
originally , it was found by born , jordan and schrdinger that the formulation of the theory in terms of the schrdinger equation is equivalent to the formulation in terms of operators and matrices , as proposed by heisenberg @xcite , implying the uncertainty relations . however , this equivalence depends on the interpretation of the electron as a point - particle .
it is not a priori valid for an extended electron . for extended electrons
the density of charge will be a physical property of a microsystem .
the question then arises , whether the uncertainty relations are still meaningful in this context .
this question can in principle be answered by an analysis of measurements of electron charge distributions .
it should be noted that the interpretation of the electron wave as a physical wave makes the usual , purely mathematical , derivation of the uncertainty relations via fourier transforms and a gaussian wave packet untenable . if the electron is a real wave , then it does possess a physical momentum and it is essentially a plane wave confined to some region in space . in this case
a composition of single electrons in terms of a fourier series of partial waves with different momenta , even though mathematically possible , is physically not meaningful .
the uncertainty relations are thus a statement about the physical properties of systems at the atomic scale @xcite . if it turns out that even in our most precise measurements the interpretation of the density of electron charge as a statistical quantity is still possible , then the uncertainty relations have a meaning also in this context . however , if it turns out that standard statistical quantities , depending on the uncertainty of location and momentum , will be much smaller in the actual experiments than allowed for by a theoretical analysis , then the uncertainty relations , and by implication operator mechanics , can not be considered an equivalent description . in this case , it should also be ultimately possible to measure any physical property of a microsystem , i.e. an atomic - scale system , with in principle infinite precision .
the statistical measure we shall use for our analysis is the standard deviation @xmath0 .
the standard deviation for a measurement of position @xmath1 , given @xmath2 measured values @xmath3 , is defined as : @xmath4 where @xmath5 denotes the statistical average .
we shall analyze the variation of the density of electron charge at a metal surface , and compare the local contrast in the experiments , which is given by the change of the density from one point of measurement to the next , with the achievable contrast on the basis of allowed energy , momentum , and position uncertainty .
we shall explicitly treat noble metal ( 111 ) surfaces , where free electrons , confined to the surface layer , exist , which can be treated for all practical purposes as free electrons .
they should thus comply with the following fundamental relation in quantum mechanics : @xmath6 the relation between an uncertainty in momentum and an uncertainty in energy is described by : @xmath7 it is difficult to define the exact uncertainty of momentum of a surface state electron in the experiments . from the energy resolution in today
s best measurements of about @xmath8mev @xcite and using eq .
( 2 ) one could infer an uncertainty of momentum of @xmath9 kgms@xmath10 , and a local uncertainty @xmath11 m , or 3.1 nm .
however , to establish that the uncertainty relations fail in this context , it is only necessary to provide an upper limit of the energy and momentum uncertainty .
thus , for an electron at the fermi level of a ag(111 ) surface the maximum available energy is given by the energy difference between the bottom of the parabolic surface state band and the energy at the fermi level .
this energy is about 80 mev .
this corresponds also to the thermal energy at ambient conditions .
as the experiments are performed at cryogenic temperatures of 5k , the thermal energy is much lower than the band energy .
hence the maximum uncertainty in the energy of the electron is also about 80 mev . from this value it is possible , via the energy uncertainty and the heisenberg uncertainty relations ( eqs .
( 2 ) ) , to derive the uncertainty of the x - position : @xmath12 m or about 3.5 .
this value is about three times larger than the wigner - seitz radius ( 1.06 ) , which describes the radius of a sphere with the typical electron density found in metals .
the uncertainty @xmath13 is also larger than the distance between two individual atoms on a silver surface , which is 2.9 .
the unit cell , the wigner - seitz cell , and the uncertainty are marked in figure 1c .
following the standard interpretation of quantum theory the electron is a point particle .
its appearance at a particular location is a statistical event governed by the schrdinger equation and the probability density , which is equal to the square of the electron s wavefunction , or @xmath14 .
the same statement is also correct for alternative interpretations of quantum mechanics , e.g. based either on the pilot - wave theory @xcite , or a theory of the dirac electron based on zitterbewegung @xcite . in this case
the precision of an experiment will depend on the statistical variation of the electron s location and its convolution over time .
this , however , will also influence the statistical variation of e.g. the apparent height .
if the apparent height can be measured with a precision of less than 0.05pm , as found in experiments @xcite , then the statistical probability of an electron to be found at a particular location can only have a certain standard deviation @xmath0 .
given the local uncertainty derived from the uncertainty relations , this deviation can , in principle , be linked either to a minimum number of events within the measurement interval of an stm , which is in the range of milliseconds , or to the error in the vertical distance measurement combined with the lateral resolution in the experiments .
note that the argument differs substantially from previous analysis of stm measurements , which only considered the vertical precision : given the time resolution of the instrument and a tunneling current of 1na about 10@xmath15 electrons cross the vacuum barrier within the measurement resolution of the instrument .
this , alone , will not be sufficient to establish the statistical facts .
only the _ lateral resolution _ in combination with the _ vertical precision _ will be sufficient to analyze the statistics in these experiments comprehensively .
we have thus two possible ways in which the experimental values can be compared to the standard deviation inferred from the uncertainty relations .
we may either find , that a threshold number of measurements n@xmath16 is necessary to obtain the experimentally observed standard deviation . in this case
we can estimate the minimum distance covered by the electron within the time - resolution limit of stm experiments @xmath17 ( i.e. milliseconds ) , and derive a lower limit for the time - resolution of the instrument without a violation of the uncertainty relations .
such a limit is described by : @xmath18 where @xmath19 is the total path during n@xmath16 measurements , @xmath20 is the average path from one measurement to the next , and @xmath21 is the energy uncertainty . or we may find that the calculated standard deviation is actually quite independent of the number of measurements . in this case
the value of the standard deviation for an energy uncertainty @xmath21 can be directly compared to the maximum standard deviation of the tunneling current given the precision of the instrument . here
, the percentage of error in an stm measurement , calculated from ratio of the error in the distance measurement divided by the apparent height of a feature , can be used , in combination with the lateral resolution of the instrument , to determine the maximum standard deviation . for an error in the experiments of 0.3% ,
which is easily obtained in today s most precise experiments , we find then : @xmath22 where @xmath23 is the lateral resolution of an stm and @xmath21 the maximum energy uncertainty .
if three times the standard deviation inferred from the energy uncertainty is substantially larger than the lateral experimental resolution @xmath23 , then one of the original assumptions going into the analysis of the process must be flawed .
we show that this is indeed the case . in the following
we start the discussion by analyzing events at one particular pixel of the grid on the surface and analyze events from the viewpoint of one electron only .
subsequently , we shall determine how the presence of many electrons at the surface could alter the initial findings .
the apparent height of a feature measured by stm is , for a metal surface , in the range of 10 to 200pm . for the sake of the argument , and to avoid being overly restrictive
, we assume the vertical resolution to be 0.1pm , and the lateral resolution to be 20pm . in a wigner - seitz cell
we then have about 88 individual pixels of stm measurements ( see figure 1d ) .
if we assume that the feature height is 30pm , then the experimental precision is 0.3% .
the apparent height at an individual pixel would be affected , however , if at a given moment during the measurement interval the electron were not found at the location of this pixel , but at a neighboring pixel . thus we may say that the standard deviation of the location measurement is such that the required percentage of measurements must be at this particular location . for a variation of less than 0.3% three times the standard deviation of the measured results
must thus still be smaller than half the distance between the pixels ( see figure 1d ) , or about 10 pm .
the standard deviation must thus be smaller than 3.3pm .
given that the uncertainty @xmath24pm , every single measurement event , at which the electron should be detected at the position of one pixel at @xmath25 , will lead to a statistically distributed measurement in the interval @xmath26 , and @xmath27 .
we have simulated such a statistical distribution of events in figure 2 .
it can be seen that the locations of the electron are indeed randomly distributed in the interval , as required by the uncertainty relations and quantum mechanics .
we find a standard deviation of about 280pm for all sets of data .
increasing the number of measurements does not change the precision , since the precision is limited by the uncertainty relations themselves .
the precision can only be increased by an increase of the electron s energy .
since @xmath0 is about one hundred times the required value , we may say that there is a clear contradiction between the observed precision of the experiments , and the calculated uncertainty for the location of the electron .
it is also clear that the underlying assumption , leading to this contradiction , is the statistical nature of the measurement , encoded in the uncertainty relations .
this contradiction , moreover , is not the result of a small variation of the values obtained , but a variation of two orders of magnitude .
it is straightforward to calculate the energy necessary to obtain the precision found in the experiments . if @xmath0 has to be around 3 pm , then the necessary uncertainty will also be about 3 pm .
this means that the momentum uncertainty has to be around @xmath28kgms@xmath10 , and the necessary energy will be more than 1000 ev . for a measurement taken at cryogenic temperatures of 5k ,
as routinely performed today , the probability of such energy , derived from the fermi - dirac distribution , is essentially zero ; the precision found in the experiments is thus impossible to explain within the framework of standard quantum theory .
the result would not change substantially if the precision in the experiments were somewhat lower .
however , so far we have assumed that the uncertainty leads to a loss of electron detection at a particular point .
it is , at least in principle , also possible , that uncertainty , via the influx of adjacent electrons at this particular point , may lead to a gain in electron detection .
we may calculate this gain from the normal distribution for the distance from adjacent pixels and the ratio of the diameter of a pixel , or 20pm and the circumference of a circle at this distance , so that the normalized gain @xmath29 from a point at distance @xmath30 will be ( units in pm ) : @xmath31 summing up all contributions from pixels in a distance d of less than 280 pm , we obtain a value of 0.65 . for a contrast of 100 ,
corresponding to an apparent height difference of 200pm , which is easily obtained in an stm , the gain at a particular pixel should not be larger than 0.01 .
comparing with the number obtained from a normal distribution , based on a standard deviation of 280pm and a pixel width of 20pm , or 0.65 , we see that such a contrast can not be obtained if the density of charge is assumed to be a statistical quantity complying with the uncertainty relations .
again , the value obtained from the statistical analysis differs by about two orders of magnitude from the value inferred from the experiments .
so that from both perspectives , the loss due to the appearance of the electron at adjacent pixels , and the gain due to the appearance of the electron statistically deviating from adjacent pixels , we arrive at the same conclusion : the precision and contrast of today s stm experiments is well beyond the requirements of the uncertainty principle .
it is thus safe to conclude that the density of charge can not be a statistical quantity . to show in detail , how a statistical density of electron charge complying with the uncertainty relations would affect the image of a single adatom at the silver surface , we have simulated the image of an adatom at very low bias voltage ( 20mv ) , in constant height mode at a distance of 500pm above the centre of the adatom .
figure 3 shows two possible images : figure 3a reflects a non - statistical measurement , where the measured current is exactly proportional to the calculated local density of states ( ldos ) , this image is very close to experimental results ( see fig . 1a , and the measurements by lian et al .
@xcite ) .
figure 3b shows a hypothetical measurement , where the ldos value at every single point is affected by the random distribution of the electrons present at a given point . in this image
we have summed up all contributions from the normal distributions of adjacent measurement points , where the amplitude is equal to the exact density of states at a given point , and its contribution to adjacent points is calculated from a normal distribution with a standard deviation of 280pm .
we have rescaled the ensuing ldos values to the maximum value for the density .
figure 3c shows a simulated linescan across a single adatom . here , the numerical difference is more than one order of magnitude .
the simulation allows the conclusion that measurements of adatoms would not be able to show an apparent height difference of more than about 10 pm .
this is , again , a clear contradiction with experimental results , where the apparent heights are typically larger than 100pm . in summary
, we have shown that modern stm experiments violate the heisenberg uncertainty relations by about two orders of magnitude .
this indicates that the density of electron charge is not a statistical quantity , as currently believed . in a wider context
, it should be noted that the question of a sub - quantum theoretical framework , something which einstein or schrdinger would have called a
hidden variable theory , seems to be increasingly on the agenda today .
one reason for it is certainly the sophistication and precision of today s experiments , as analyzed in this paper .
another , however , is dissatisfaction with a framework which does not seem to allow us insight into fundamental processes at the atomic scale . most notable in this context
are attempts to account for the probabilistic framework of quantum mechanics by an analysis of a deeper level of description , see recent papers by khrennikov on prequantum statistical theory @xcite , by grssing et al .
on subquantum thermodynamics @xcite , by elze on emergent quantum mechanics @xcite , and by thooft on deterministic quantum mechanics @xcite .
it might be that these developments indicate that our understanding of reality is about to undergo a _
quantum leap _ into a new direction . |
0.5 truecm the processes of heavy flavour production on nucleon and nucleus targets at high energies are very interesting from both the theoretical and practical points of view .
these processes at high energies are usually considered in the framework of perturbative qcd in the leading and the next - to - leading order @xmath5 expansion . in the case of a nuclear target
it is usually assumed that the cross section of heavy flavour production , @xmath6 , should be proportional to @xmath7 , with @xmath8 , that is , in agreement with the experimental result of [ 1 ] , where a value @xmath9 was obtained for the case of neutral d - meson production at @xmath10 = 39 gev .
however it is experimentally well - known [ 2 - 8 ] that quark distributions in nucleus are slightly different from the same distributions in free nucleons .
so the value of @xmath1 can differ from unity and it seems interesting to estimate it .
we will present in the section 2 the results obtained using the nucleus to nucleon structure function ratios taken from ref . [ 9 ] .
the standard qcd expression for heavy quark production cross section in a hadron 1 - hadron 2 collision has the form @xmath11\left[x_bg_{b/2}(x_b,\mu^{2})\right ] \hat{\sigma}^{ab\rightarrow q\overline{q}}(\hat{s},m_q,\mu^{2 } ) \ ; , \label{eq : totalpf}\ ] ] where @xmath12 and @xmath13 . here @xmath14 and @xmath15 are the structure functions of partons @xmath16 and @xmath17 inside hadrons @xmath18 and @xmath19 respectively , and @xmath20 is the cross section of the subprocess @xmath21 as given by standard qcd .
the latter depends on the parton center - of - mass energy , the mass of the produced heavy quark @xmath22 and the qcd scale @xmath23 .
( [ eq : totalpf ] ) should account for all possible subprocesses @xmath24 .
the parton cross section @xmath20 can be written in the form @xcite @xmath25 with @xmath26 and @xmath27\;.\ ] ] the functions @xmath28 , @xmath29 and @xmath30 can be found in @xcite
. for the numerical calculations we wrote the nuclear structure function @xmath31 in the form @xmath32 similarly to ref .
@xcite and we take the values of @xmath33 for gluons , valence and sea quarks from ref . [ 9 ] for light sea quarks but an enhancement of gluon distributions in the nuclei at small @xmath2 . ] .
they are presented in fig .
the values of @xmath33 in [ 9 ] are presented for @xmath34 that is not small enough at high energies .
so at @xmath35 we used two variants of the @xmath33 behaviour for gluon distributions : the first is the extrapolation as @xmath36 with @xmath37 = 0.096 and 0.040 for the charm and beauty production , respectively ( solid lines in fig .
such behaviour is in qualitative agreement with the results of @xcite .
the second is the constant frozen at @xmath38 ( dash - dotted lines ) .
three different sets of parton distributions were used , namely mrs-1 @xcite , mt s - dis @xcite and grv ho @xcite that can be found in cern pdflib @xcite , but they give practically the same results for the @xmath1 behaviour ( see table 1 ) .
we have used in the case of charm production the values @xmath39 = 1.5 gev and @xmath40 = 4 gev@xmath41 and in the case of beauty production @xmath42 = 5 gev and @xmath40 = @xmath43 .
the obtained results for @xmath1 determined from the ratios of heavy quark production cross section on a gold target and on the proton are presented in the table 1 for three sets of structure functions and frozen gluon distribution ratios at @xmath38 and in fig . 2 for the grv ho set and two variants of gluon distribution ratios in the small @xmath2 region .
here @xmath44 is the c.m .
energy for the interaction of the incident proton with one target nucleon .
one can see that at fixed target energies the values of @xmath1 are slightly higher than unity , which is not in contradiction with the result of ref . [ 1 ] .
however @xmath1 decreases with increasing energy and this effect is larger in the case of charm production than in the case of beauty .
one can see also that the difference between two variants for @xmath33 at @xmath35 becomes important only at the highest energies .
we also calculate the values of @xmath1 for different feynman-@xmath2 ( @xmath45 ) regions using @xmath46 at energies @xmath44 = 39 gev and 1800 gev .
the results are presented in fig .
3 . at negative and moderate @xmath45 ( in the nucleus fragmentation region )
the values of @xmath1 are slightly higher than unity .
however in the case of charm production in the beam fragmentation region ( positive @xmath45 ) the values of @xmath1 become essentially smaller than unity . for beauty production
the last effect is expected only at very high energies .
5 mm we calculate the a - dependence of charm and beauty production using standard qcd formulas and accounting in the difference of parton distributions for free and bound nucleons . if one parametrize the heavy flavour production cross section as @xmath0 , the value of @xmath1 is slightly different from unity at the available energies . at comparatively low energies
the obtained values of @xmath1 are a little larger than unity .
this should be connected with some nucleon - nucleon correlations which change the large-@xmath2 parton distributions . at higher energies the values of @xmath1 decrease and
become smaller than unity . at @xmath44
= 1800 gev we expect the value of @xmath47 0.95 .
the decrease of the ratios @xmath33 that results in a decrease of @xmath1 can be connected with the effects of parton density saturation @xcite which occurs in heavy nuclei at @xmath2 values higher than in the proton . if we consider two small and different values of @xmath48 and @xmath49 in eq .
( 1 ) , it is clear that the contribution to the inclusive cross section from the region @xmath50 should be larger than the mirrow contribution ( @xmath51 ) because the value of the ratio @xmath33 in the first case is larger .
it means that heavy quark pairs will be produced preferably in the nucleus fragmentation hemisphere , i.e. , asymetrically , that is quite usual and confirmed experimentally in the case of light quark hadron production . from fig .
3 it is clear that charm production on nucleus at lhc energies should give information on the nuclear shadowing of the structure functions at small @xmath2 . let us note also that at hera - b energy ( @xmath44 = 39 gev ) with grv(ho ) parton distributions we obtain @xmath52 = 6.8 nb that is in agreement with the experimental value @xmath53 nb of ref . @xcite . however the experimental point should be decreased by 5% because it was obtained as an extrapolation from the gold target data assuming an @xmath54 dependence whereas our calculations give the value @xmath1 = 1.01 .
we are grateful to m.a.braun for useful discussions and both to k.j.eskola and to i.sarcevic for sending us their numerical results .
we thank the direccin general de poltica cientfica and the cicyt of spain for financial support . n.a . and c.a.s .
also thank the xunta de galicia for financial support .
the paper was also supported in part by intas grant 93 - 0079 .
the functions @xmath56 , @xmath57 and @xmath58 which determine the ratios of the distributions for protons in the nucleus versus free protons for gluons ( solid and dash - dotted curves , see text ) , valence quarks ( dashed curves ) and sea quarks ( dotted curves ) respectively for @xmath59 4 gev@xmath41 ( a ) and @xmath60 ( b ) .
the energy dependence of @xmath1 in the cases of charm and beauty production for grv ho structure functions and using extrapolated ( solid curves ) and frozen at @xmath38 ( dashed curves ) ratios of gluon distributions . the feynman-@xmath2 dependence of @xmath1 values in the cases of charm and beauty production at @xmath44 = 39 gev ( a ) and 1800 gev ( b ) for grv ho structure functions and using extrapolated ( solid curves ) and frozen at @xmath38 ( dashed curves ) ratios of gluon distributions .
m.j.leitch et al .
72 ( 1994 ) 2542 .
, j.j.aubert et al .
b123 ( 1983 ) 275 .
a.bodek et al .
50 ( 1983 ) 1431 ; 51 ( 1983 ) 534 .
r.g.arnold et al .
phys.rev.lett . 52 ( 1984 ) 727 .
bcdms coll .
, g.bari et al .
b163 ( 1985 ) 282 .
, j.ashman et al
. phys.lett .
b202 ( 1988 ) 603 ; m.arneodo et al .
b211 ( 1988 ) 493 .
, m.arneodo et al .
b333 ( 1990 ) 1 .
, p.amaudruz et al . z.phys .
c51 ( 1991 ) 387 . k.j.eskola . nucl.phys .
b400 ( 1994 ) 240 .
p.nason , s.dawson and r.k.ellis . nucl.phys .
b303 ( 1988 ) 607 .
i.sarcevic and p.valerio .
b338 ( 1994 ) 426 .
v.barone , m.genovese , n.n.nikolaev , e.predazzi and b.g.zakharov . z.phys .
c58 ( 1993 ) 451 .
m.b.gay ducati , a.l.ayala and e.m.levin .
preprint cbpf - nf-010/95 , hep - ph/9506349 .
a.d.martin , r.g.roberts and w.j.stirling .
phys.rev.d37 ( 1988 ) 1161 .
j.g.morfin and w .- k.tung .
c52 ( 1991 ) 13 .
m.gluck , e.reya and a.vogt .
c53 ( 1992 ) 127 .
pdflib : nucleon , pion and photon parton density functions and @xmath5 calculations , version 6.06 .
h. plothow - besch , cern - ppe ( 1995 ) .
l.v.gribov , e.m.levin and m.g.ryskin .
phys.rep . 100
( 1983 ) 1 .
d.m.jansen et al .
74 ( 1995 ) 3118 . | we calculate the a - dependence of charm and beauty production cross sections on nuclear targets taking into account the difference of quark and gluon distributions in free nucleons and in nucleus . at comparatively low energies , if @xmath0 , the value of @xmath1 is slightly higher than unity . with the growth of the initial energy the value of @xmath1 decreases and becomes smaller than unity .
we also calculate the values of @xmath1 for different feynman-@xmath2 of the produced @xmath3 pair and obtain that they decrease significantly in the beam fragmentation region . * us - ft/22 - 95 * .8 truecm * a - dependence of heavy flavour production in qcd * + 1.5 truecm * n.armesto , c.pajares , c.a.salgado and yu.m.shabelski@xmath4 * + 0.9 truecm _ departamento de fisica de partculas ,
universidade de santiago de compostela , + 15706santiago de compostela , spain _ 2.5 truecm \2 . truecm * ) permanent address : petersburg nuclear physics institute , + gatchina , sanct - petersburg 188350 , russia .
1.5 truecm august 1995 + .2 truecm * us - ft/22 - 95 * |
one of the important tools for studying the properties of quark - gluon plasma ( qgp ) in ultrarelativistic heavy ion collisions is the analysis of a qcd jet production .
the medium - induced energy loss of energetic partons , `` jet quenching '' , should be very different in the cold nuclear matter and qgp , resulting in many observable phenomena @xcite .
recent rhic data on high - p@xmath0 particle production @xcite ( the suppression of hadron spectra and azimuthal back - to - back two - particle correlations , strong elliptic flow ) are in agreement with the jet quenching hypothesis @xcite . at lhc
, a new regime of heavy ion physics will be reached at @xmath3 tev where hard and semi - hard particle production can stand out against the underlying soft events .
the initial gluon densities in pb+pb reactions at lhc are expected to be significantly higher than those at rhic , implying a stronger partonic energy loss , observable in various new channels @xcite . in the most of available
monte - carlo heavy ion event generators the medium - induced partonic rescattering and energy loss are either ignored or implemented insufficiently @xcite .
thus , in order to analyze rhic data on high - p@xmath0 hadron production , test the sensitivity of lhc observables to the qgp formation , and study the corresponding experimental capabilities of detectors , the development of adequate and fast monte - carlo models for jet quenching simulation is necessary .
in this paper we present the model of jet quenching and discuss its validation basing on rhic data for high-@xmath4 hadron spectra . in sect .
2 we give the physics frameworks of the model . section 3
describes the event - by - event simulation procedure . in sect .
4 the generalization of the model to the case of `` full '' heavy ion event ( superposition of soft hydro - type state and hard multi - jets ) is fulfilled . in sect . 5
the efficiency of the model is demonstrated by means of the numerical analysis of hadron spectra at rhic .
the detailed description of physics frameworks of the developed model can be found in a number of our previous papers @xcite .
our approach bases on an accumulating energy loss , the gluon radiation being associated with each parton scattering in the expanding medium and includes the interference effect using the modified radiation spectrum @xmath5 as a function of decreasing temperature @xmath6 . the basic kinetic integral equation for the energy loss @xmath7 as a function of initial energy @xmath8 and path length @xmath9 has the form @xmath10 where @xmath11 is the current transverse coordinate of a parton , @xmath12 is the scattering probability density , @xmath5 is the energy loss per unit length , @xmath13 is in - medium mean free path , @xmath14 is the medium density at the temperature @xmath6 , @xmath15 is the integral cross section for parton interaction in the medium .
the collisional energy loss due to elastic scattering with high - momentum transfer has been originally estimated by bjorken in @xcite , and recalculated later in @xcite taking also into account the low - momentum transfer loss resulting mainly from the interactions with plasma collective modes .
since the latter process does not contribute much to the total collisional loss in comparison with high - momentum scattering ( due to absence of large factor @xmath16 where @xmath17 is the debye screening mass ) and , in numerical estimates it can be effectively `` absorbed '' by means of redefinition of minimum momentum transfer @xmath18 , we used the collisional part associated with high - momentum transfer only @xcite , @xmath19 and the dominant contribution to the differential cross section @xmath20 for scattering of a hard parton with energy @xmath8 and mass @xmath21 off the `` thermal '' parton with energy ( or effective mass ) @xmath22 . here @xmath23 for @xmath24 , @xmath25 and @xmath26 scatterings respectively , @xmath27 is the qcd running coupling constant for @xmath28 active quark flavors , and @xmath29 is the qcd scale parameter which is of the order of the critical temperature , @xmath30 mev . the integrated cross section @xmath15 is regularized by the debye screening mass squared @xmath31 .
the maximum momentum transfer @xmath32 [ s-(m_p - m_0)^2 ] / s$ ] where @xmath33 .
there are several calculations of the inclusive energy distribution of medium - induced gluon radiation using feyman multiple scattering diagrams .
the relation between these approaches and their basic parameters has been discussed in detail in the recent writeup of the working group `` jet physics '' for the cern yellow report @xcite .
we restrict ourselves to using bdms formalism @xcite . in the bdms frameworks ,
the strength of multiple scattering is characterized by the transport coefficient @xmath34 ( @xmath35 is the gluon mean free path ) , which is related to the elastic scattering cross section @xmath15 ( [ sigt ] ) . in our simulations
this strength is in fact regulated mainly by the initial qgp temperature @xmath36 .
then the energy spectrum of coherent medium - induced gluon radiation and the corresponding dominant part of radiative energy loss of massless parton have the form @xcite : @xmath37 \>\ln{\left| \cos{(\omega_1\tau_1 ) } \right| } \ > , \\
\omega_1 = \sqrt{i \left ( 1 - y + \frac{c_r}{3}y^2 \right ) \bar{\kappa}\ln{\frac{16}{\bar{\kappa } } } } \quad \mbox{with}\quad \bar{\kappa } = \frac{\mu_d^2\lambda_g } { \omega(1-y ) } ~ , \end{aligned}\ ] ] where @xmath38 , @xmath39 is the fraction of the hard parton energy carried away by the radiated gluon , and @xmath40 is the quark color factor . a similar expression for the gluon jet can be obtained by setting @xmath41 and proper by changing the factor in the square brackets in ( [ radiat ] ) , see ref .
@xcite . the integration ( [ radiat ] )
is carried out over all energies from @xmath42 , the minimum radiated gluon energy in the coherent lpm regime , up to initial parton energy @xmath8 .
note that we do not consider here possible effects of double parton scattering @xcite and thermal gluon absorption @xcite , which can be included in the model in the future .
the simplest generalization of the formula for a heavy quark of mass @xmath43 can be done by using the `` dead - cone '' approximation @xcite : @xmath44 one should mention the more recent developments on heavy quark energy loss calculations available in the literature @xcite , which can be also considered as further model improvements .
the medium is treated as a boost - invariant longitudinally expanding quark - gluon fluid , and partons as being produced on a hyper - surface of equal proper times @xmath45 @xcite . in order to simplify numerical calculations
we omit here the transverse expansion and viscosity of the fluid using the well - known scaling solution obtained by bjorken @xcite for a temperature and density of qgp at @xmath46 mev : @xmath47 the internal model parameters are the initial conditions for the qgp formation expected for central au+au ( pb+pb ) collisions at rhic ( lhc ) : @xmath48 , @xmath36 and @xmath28 .
for non - central collisions and for other beam atomic numbers we suggest the proportionality of the initial energy density @xmath49 to the ratio of nuclear overlap function and effective transverse area of nuclear overlapping @xcite .
note that using other scenarios of qgp space - time evolution for the monte - carlo implementation of the model is also envisaged .
in fact , the influence of the transverse flow , as well as that of the mixed phase at @xmath50 , on the intensity of jet rescattering ( which is a strongly increasing function of @xmath6 ) has been found to be inessential for high initial temperatures @xmath51 . on the contrary , the presence of qgp viscosity slows down the cooling rate , that implies a jet parton spending more time in the hottest regions of the medium . as a result the rescattering intensity increases , i.e. , in fact an effective temperature of the medium gets higher as compared with the perfect qgp case .
we also do not take into account here the probability of jet rescattering in nuclear matter , because the intensity of this process and corresponding contribution to the total energy loss are not significant due to much smaller energy density in a `` cold '' nucleus .
another important element of the model is the angular spectrum of in - medium gluon radiation .
since the detailed calculation of the angular spectrum of emitted gluons is rather sophisticated and model - dependent @xcite , the simple parameterization of gluon angular distribution over the emission angle @xmath52 was used : @xmath53 where @xmath54 is the typical angle of the coherent gluon radiation as estimated in @xcite .
other parameterizations are also envisaged .
the model has been constructed as the fast monte - carlo event generator pyquen ( pythia quenched ) , and the corresponding fortran routine pyquen is available via internet @xcite .
the routine is implemented as a modification of the standard [email protected] .
* jet event @xcite .
the following event - by - event monte - carlo simulation procedure is applied .
+ @xmath56 generation of the initial parton spectra with pythia ( fragmentation _ off _ ) .
+ @xmath56 generation of the jet production vertex at the impact parameter @xmath57 according to the distribution @xmath58 where @xmath59 are the distances between the nucleus centers and the jet production vertex @xmath60 ; @xmath61 is the maximum possible transverse distance @xmath62 from the nuclear collision axis to @xmath63 ; @xmath64 is the radius of the nucleus @xmath65 ; @xmath66 is the nuclear thickness function with nucleon density distribution @xmath67 ; @xmath68 is the nuclear overlap function ( see ref .
@xcite for detailed nuclear geometry explanations ) .
+ @xmath56 calculation of scattering cross section @xmath69 ( [ sigt ] ) .
+ @xmath56 generation of the displacement between @xmath70-th and @xmath71-th scatterings , @xmath72 : @xmath73 and calculation of the corresponding transverse distance , @xmath74 .
+ @xmath56 reducing the parton energy by collisional and radiative loss per each @xmath70-th scattering : @xmath75 where the collisional part is calculated in the high - momentum transfer approximation ( [ sigt ] ) , @xmath76 and the energy of a radiated gluon , @xmath77 , is generated according to ( [ radiat ] ) and ( [ radmass ] ) : @xmath78 \>\ln{\left| \cos{(\omega_1\tau_1 ) } \right|}~ , ~~\frac{di}{d\omega } | _ { m_q \ne 0 } = \frac{1}{(1+(\beta \omega ) ^{3/2})^2 } \frac{di}{d\omega } | _ { m_q=0}~. \end{aligned}\ ] ] @xmath56 calculation of the parton transverse momentum kick due to elastic scattering @xmath70 : @xmath79 @xmath56 formation of the additional ( in - medium emitted ) gluon with the energy @xmath80 and the emission angle @xmath81 relative to the parent parton determined according to the parameterization ( [ sar ] ) .
+ @xmath56 halting the rescattering if _
( 1 ) _ the parton escapes the dense zone , or _ ( 2 ) _ qgp cools down to @xmath82 mev , or _ ( 3 ) _ the parton loses so much energy that its @xmath83 drops below @xmath84 .
+ @xmath56 at the end of each event , adding new ( in - medium emitted ) gluons to the pythia parton list and rearrangement of partons to update string formation .
+ @xmath56 formation of the final state particles by pythia ( fragmentation _ on _ ) .
the mean total particle multiplicity in central au+au ( pb+pb ) collisions at rhic ( lhc ) is the input parameter of the model ( instead of @xmath104 we put @xmath64 here for simplicity ) , the total multiplicity for other centralities and atomic numbers being assumed to be proportional to the number of nucleons - participants .
we also set the poisson multiplicity distribution and the following particle ratios : @xmath105 the hard part of the event includes pythia / pyquen hadronic jets generated according to the binomial distribution .
the mean number of jets produced in aa events at a given @xmath57 is proportional to the number of binary nucleon - nucleon sub - collisions and determined as @xmath106 where @xmath107 is the cross section of corresponding hard process in @xmath108 collisions ( at the same c.m.s .
energy , @xmath109 , of colliding beams ) with the minimum transverse momentum transfer @xmath110 .
the latter is another input parameter of the model . in the frameworks of our approximation , partons produced in ( semi)hard processes with the momentum transfer less than @xmath110
are considered as being `` thermalized '' , so their hadronization products are included in a soft part of the event `` automatically '' .
note that we can expect some adequate results only for central and semi - central collisions , but not for very peripheral collisions ( @xmath111 ) where the hydro - type description is not applicable .
besides , the very forward rapidity region ( where other dynamical effects can be important ) is beyond our treatment here .
the extended in such a way jet quenching model has been constructed as the fast monte - carlo event generator , and the corresponding fortran code is also available via internet @xcite .
let us remind in the end of this section , that ideologically our approximation is similar to the model of hirano and nara @xcite .
the difference is that we concentrate here on the detailed simulation of the parton multiple scattering in a qcd - medium ( the scattering - by - scattering generation of parton path length and energy loss in an expanding qgp , taking into account the collisional loss , lund string fragmentation model both for hard partons and in - medium emitted gluons , etc . ) , while the treatment of the hydrodynamic part in @xcite is much more detailed than in our simple ( and therefore fast ) simulation procedure .
in order to demonstrate the efficiency of the model , the jet quenching pattern in au+au collisions at rhic was considered . the comparison of calculated and experimentally measured pseudorapidity @xmath113 and transverse momentum @xmath4 spectra of hadrons together with their dependence of event centrality allows the optimization of the model and specification of main model parameters . the phobos data on @xmath113-spectra of charged hadrons @xcite have been analyzed to fix the particle density in the mid - rapidity region and the maximum longitudinal flow rapidity , @xmath114 . for the calculation of ( multi)jet production cross section
, we used the factor @xmath115 taking into account higher order corrections of perturbative qcd .
the rest of the model parameters have been obtained by fitting phenix data on @xmath4-spectra of neutral pions @xcite : the kinetic freeze - out temperature @xmath116 mev , maximum transverse flow rapidity @xmath117 and minimum transverse momentum transfer of `` non - thermalized '' hard process @xmath118 gev/@xmath119 .
it was found that the nuclear modification of the hardest domain of @xmath4-spectrum ( @xmath120 gev/@xmath119 ) is determined in our case only by the intensity of the medium - induced parton rescattering .
this fact allows us to extract from the data initial conditions of the qgp formation independently on other input parameters : the initial temperature @xmath121 mev , the formation time @xmath122 fm/@xmath119 and the number of active quark flavours @xmath123 .
we will see below , that setting model parameters as it was described above , makes it possible to reproduce the main features of jet quenching pattern at rhic : the @xmath4dependence of the nuclear modification factor @xmath124 and two - particle azimuthal correlation function @xmath125 .
figure 1 shows @xmath113-distribution of charged hadrons in au+au collisions for different centrality sets .
the good fit of phobos data @xcite is achieved excepting very forward rapidities .
the @xmath4-distributions of @xmath126-mesons obtained at phenix @xcite is also well reproduced by our calculations , even for relatively peripheral collisions ( figure 2 ) .
figure 3 shows the nuclear modification factor @xmath124 for neutral pions , which is defined as a ratio of particle yields in @xmath127 and @xmath108 collisions normalized on the number of binary nucleon - nucleon sub - collisions : @xmath128 where @xmath129 mb is the inelastic non - diffractive @xmath108 cross section at @xmath130 gev . in the absence of medium - induced effects in the mid - rapidity region
it should be @xmath131 for high enough @xmath132 gev/@xmath119 ) .
such value of @xmath133 has been observed so for d+au and peripheral au+au collisions , but not for central and semi - central au+au events , where @xmath134 up to maximum measured transverse momenta @xmath135 gev/@xmath119 .
one can see from figure 3 that our model calculations reproduce @xmath4 and centrality dependences of @xmath124 @xcite quite well . another important tool to verify jet quenching is two - particle azimuthal correlation function @xmath125 the distribution over an azimuthal angle of high-@xmath4 hadrons in the event with @xmath136 gev/@xmath137 relative to that for the hardest `` trigger '' particle with @xmath138 gev/@xmath119 .
figure 4 presents @xmath125 in @xmath108 and in central au+au collisions ( data from star @xcite ) .
clear peaks in @xmath108 collisions at @xmath139 and @xmath140 indicate a typical dijet event topology .
note that almost the same pattern has been observed in d+au and peripheral au+au collisions .
however , for most central au+au collisions the peak near @xmath141 disappears .
it can be interpreted as the observation of monojet events due to the `` absorption '' of one of the jets in a dense medium .
such event configuration corresponds to the situation when the dijet production vertex is close to the surface of the nuclear overlap region : then one partonic jet can escape the medium almost without re - interactions and then go to detectors , while second jet loses most of its initial energy due to a large number of rescatterings and therefore becomes unobservable @xcite .
figure 4 demonstrates that measured suppression of azimuthal back - to - back correlations is well reproduced by our model ( the same procedure of uncorrelated background subtraction as in @xcite was applied ) .
we leave beyond the scope of this paper the analysis of such important rhic observables as the azimuthal anisotropy and particle ratios .
since these observables are very sensitive to the soft physics , in order to study them a more careful treatment of low-@xmath4 particle production than our simple approach is needed ( the detailed description of space - time structure of freeze - out region , resonance decays , etc . ) for example , our model can reproduce experimentally measured @xcite @xmath4-dependence of the coefficient of azimuthal anisotropy @xmath142 ( the second harmonic of fourier decomposition of particle azimuthal distribution ) qualitatively , giving rapid hydrodynamical growth up to @xmath143 gev/@xmath119 with the subsequent saturation .
however , the model calculations significantly underestimate the data at @xmath144 gev/@xmath119 .
a solution of baryon - to - meson ratio `` puzzle '' @xcite is also beyond our consideration here .
further development of our model with special emphasis on the more detailed description of low-@xmath4 particle production is planed for the future .
the model of jet quenching in ultrarelativistic heavy ion collisions has been developed .
it includes the generation of the hard parton production vertex according to the realistic nuclear geometry , rescattering - by - rescattering simulation of the parton path length in a dense matter , radiative and collisional energy loss per rescattering , final hadronization with the lund string fragmentation model for hard partons and in - medium emitted gluons .
the model is the fast monte - carlo tool implemented to modify a standard pythia jet event .
the model has been generalized to the case of the `` full '' heavy ion event ( the superposition of soft , hydro - type state and hard multi - jets ) using a simple and fast simulation procedure for soft particle production .
the efficiency of the model is demonstrated basing on the numerical analysis of high-@xmath4 hadron production in au+au collisions at rhic .
the good fit of experimental data on @xmath113 and @xmath4 spectra of hadrons for different event centralities is achieved .
the model is capable of reproducing main features of the jet quenching pattern at rhic : the @xmath4 dependence of the nuclear modification factor @xmath124 , and the suppression of azimuthal back - to - back correlations .
the further development of the model focusing on a more detailed description of low-@xmath4 particle production is planed for the future . _
acknowledgments_. discussions with a.i .
demianov , yu.l .
dokshitzer , a. morsch , s.v .
petrushanko , c. roland , l.i .
sarycheva , j. schukraft , c.yu .
teplov , i.n .
vardanyan , i. vitev , b. wyslouch , b.g . zakharov and g.m .
zinovjev and are gratefully acknowledged .
this work is supported by grant n 04 - 02 - 16333 of russian foundation for basic research .
99 r. baier , d. schiff , b.g .
zakharov , annual rev .
nucl . part .
* 50 * , 37 ( 2000 ) proceedings of 16th international conference on ultrarelativistic nucleus - nucleus collisions quark matter2002 ( nantes , france , 18 - 24 june , 2002 ) , nucl .
a * 715 * ( 2003 ) ; proceedings of 17th international conference on ultrarelativistic nucleus - nucleus collisions quark matter2004 ( oakland , california , 11 - 17 january , 2004 ) , j. phys .
g * 30 * ( 2004 ) x .- n .
wang , phys .
b 579 , * 299 * ( 2004 ) a. accardi et.al .
jet physics , in : _ hard probes in heavy ion collisions at the lhc _
, ed . by m.
mangano , h. satz , u. wiedemann , cern report 2004 - 09 , hep - ph/0310274 cern workshop on monte carlo tools for the lhc , 2003 , + http://mlm.home.cern.ch/mlm/mcwshop03/mcwshop.html i.p .
lokhtin , a.m. snigirev , phys .
b * 440 * , 163 ( 1998 ) i.p .
lokhtin , a.m. snigirev , eur .
j. c * 16 * , 527 ( 2000 ) i.p .
lokhtin , a.m. snigirev , eur .
j. c * 21 * , 155 ( 2001 ) i.p .
lokhtin , s.v .
petrushanko , l.i .
sarycheva , a.m. snigirev , phys . at .
nucl . * 65 * , 943 ( 2002 ) i.p .
lokhtin , a.m. snigirev , phys .
b * 567 * , 39 ( 2003 ) j.d .
bjorken , fermilab publication pub-82/29-thy ( 1982 ) s. mrwczyski , phys .
b * 269 * , 383 ( 1991 ) ; m.h .
thoma , phys .
b * 273 * , 128 ( 1991 ) r. baier , yu . l. dokshitzer , a.h .
mueller , d. schiff , phys .
c * 60 * , 064902 ( 1999 ) ; phys .
c * 64 * , 057902 ( 2001 ) x .- n .
wang , x .- f .
guo , nucl . phys .
a * 696 * , 788 ( 2001 ) i. vitev , e - print hep - ph/0503221 e. wang , x .-
wang , phys .
87 , 142301 ( 2001 ) yu.l . dokshitzer , d. kharzeev , phys
. lett .
b * 519 * , 199 ( 2001 ) m. djordjevic , m. gyulassy , nucl .
a * 733 * , 265 ( 2004 ) n. armesto , c.a .
salgado , u.a .
wiedemann , phys .
d * 69 * , 114003 ( 2004 ) b .- w .
zhang , e. wang , x .-
wang , phys .
lett . * 93 * , 072301 ( 2004 ) j.d .
bjorken , phys .
d * 27 * , 140 ( 1983 ) b.g .
zakharov , jetp lett . *
70 * , 176 ( 1999 ) u.a .
wiedemann , m. gyulassy , nucl .
b * 560 * , 345 ( 1999 ) ; u.a .
wiedemann , nucl .
b * 588 * , 303 ( 2000 ) ; nucl .
a * 690 * , 731 ( 2001 ) m. gyulassy , p. levai , i. vitev , nucl .
b * 571 * , 197 ( 2000 ) ; phys .
85 * , 5335 ( 2000 ) ; nucl .
b * 594 * , ( 2001 ) 371 http://cern.ch/lokhtin/pyquen t. sjostrand , comp .
phys . com .
* 135 * , 238 ( 2001 ) n.a .
kruglov , i.p .
lokhtin , l.i .
sarycheva , a a.m. snigirev , z. phys .
* c 76 * , 99 ( 1997 ) i.p .
lokhtin , l.i .
sarycheva , a.m. snigirev , phys .
* b 537 * , 261 ( 2002 ) e. schnedermann , j. sollfrank , u. heinz , phys . rev .
* c 48 * , 2462 ( 1993 ) s. muroya , h. nakamura , m. namiki , progr .
* 120 * , 209 ( 1995 ) p.f .
kolb , j. sollfrank , u. heinz , phys .
* c 62 * , 054909 , ( 2000 ) ; p.f .
kolb et al . , phys .
* b 500 * , 232 ( 2001 ) ; nucl .
phys . * a 696 * , 175 ( 2001 ) hydjet fast event generator , + http://cern.ch/lokhtin/hydro/hydjet.html t. hirano , t. nara , nucl .
a * 743 * , 305 ( 2004 ) ; phys .
c * 69 * , 034908 ( 2004 ) ; phys . rev .
c * 66 * , 041901 ( 2002 ) b.b .
back et al .
( phobos collab . ) , phys . rev
. lett . *
91 * , 052303 ( 2003 ) s.s .
adler et al .
( phenix collab . ) , phys .
* 91 * , 072301 ( 2003 ) c. adler et al .
( star collaboration ) , phys .
lett . *
89 * , 202301 ( 2002 ) m. plumer , m. gyulassy , x .-
wang , nucl .
a * 590 * , 511 ( 1995 ) s.s .
adler et al .
( phenix collab . ) , phys .
lett . * 91 * , 182301 ( 2003 ) ; e - print nucl - ex/0411040 j. adams et al .
( star collaboration ) , phys .
lett . * 93 * , 252301 ( 2004 ) ; e - print nucl - ex/0409033 s.s .
adler et al . ( phenix collab . ) , phys . rev .
c * 69 * , 034909 ( 2004 ) j. adams et al .
( star collaboration ) , phys .
lett . * 92 * , 112301 ( 2004 ) | the method to simulate the rescattering and energy loss of hard partons in ultrarelativistic heavy ion collisions has been developed .
the model is a fast monte - carlo tool introduced to modify a standard pythia jet event .
the full heavy ion event is obtained as a superposition of a soft hydro - type state and hard multi - jets .
the model is applied to analysis of the jet quenching pattern at rhic .
= -20 mm # 1#23.6pt * a model of jet quenching in ultrarelativistic heavy ion collisions and high - p@xmath0 hadron spectra at rhic * + i.p .
lokhtin@xmath1 and a.m. snigirev@xmath2 + m.v .
lomonosov moscow state university , d.v .
skobeltsyn institute of nuclear physics , + 119992 , vorobievy gory , moscow , russia
+ @xmath1 e - mail : [email protected] + @xmath2 e - mail : [email protected] |
weak nonlinearity is known to cause a redistribution of elastodynamic spectral energy density .
energy present at one or two base frequencies will migrate , under the influence of nonlinearity , to higher harmonics and to sum and difference frequencies . if those sum and difference frequencies originally have little or no energy , their presence can be a sensitive measure of nonlinearity .
the method has been shown to be capable of detecting flaws in otherwise linear - response specimens .
it has been applied to crack detection @xcite and to assessing the quality of adhesive bonds @xcite .
much of this work has been qualitative , with emphases on how the nonlinear effects scale with field amplitude .
the geometry is such , and the fields sufficiently uncontrolled , that attempts to quantify the nonlinearity in absolute terms have not been feasible . in special circumstances , with high amplitude and long distance plane wave propagation
, it is possible to monitor harmonic generation , and thereby assess nonlinearity , and do so quantitatively .
it has been suggested that nonlinear rayleigh wave propagation may be a useful nde technique to measure surface properties @xcite .
lamb waves were also suggested for this purpose @xcite . as an alternative to plane wave configurations
diffuse elastodynamic fields might be appropriate for nde measurements under insufficiently controlled conditions .
diffuse fields span a broad range of applications from optics and microwaves to ultrasonics ( see @xcite and references therein ) .
gross properties of a specimen can be evaluated from integral field parameters @xcite .
extension of diffuse field theory to include nonlinear effects is receiving increasing experimental @xcite and theoretical @xcite attention .
its application to evaluation of material properties may be a future research focus .
the purpose of this paper is to describe effects of weak nonlinearity on the energy spectrum of a diffuse field , as one of the integral quantities available for experimental measurement . with a view toward eventual application in the context of nde , but with an interest in the general problem , we take elastodynamics in a nonlinear solid as an underlying physical system for our diffuse field model .
we start with the governing equations for the chosen field type in section ii , and develop them into a system of odes with quadratic and cubic nonlinear terms describing time - evolution of the modal amplitudes . in section iii
we discuss statistical assumptions for the normal frequencies and modes of the solid . in section iv the nonlinear equations are treated by means of regular perturbation theory and averaged to obtain the spectral power transfered into a frequency band .
results independent of the underlying physical nature of the nonlinearity are considered . in section
v we specialize to the case of an isotropic homogeneous elastic body and an initial field generated by two transient narrow - band signals centered at different frequencies .
quantitative estimates for the nonlinear energy transfer into double and combination frequencies are provided and discussed .
conclusions are presented in section vi .
suppose we have a finite hyperelastic body occupying volume @xmath0 in its natural state referenced by coordinate @xmath1 , and having material density @xmath2 .
strain energy density of the body for a displacement field @xmath3 with a corresponding green s tensor @xmath4 is described to leading orders in strain by @xcite : @xmath5 where @xmath6 , @xmath7 and @xmath8 are linear , second and third - order nonlinear elastic tensors respectively .
let the body be subjected to external forces and tractions that stop acting after a certain cutoff time @xmath9 after which the field in the body freely evolves without dissipation under zero - displacement boundary conditions . according to the hamilton s principle
, the weak form of the governing equation for the field evolution can be written as @xmath10 where @xmath11 is the first piola - kirchoff stress tensor .
it is related to the strain energy density ( [ eq : energydensity ] ) via deformation gradient @xmath12 @xcite:@xmath13 to distinguish parts of equation ( [ eq : governing ] ) responsible for linear and leading - order nonlinear behavior of the system we expand the stress tensor @xmath11 in powers of @xmath14 up to the third - order , and label the respective linear and nonlinear operators as @xmath15 and @xmath16 .
equation ( [ eq : governing ] ) is then written as @xmath17d\mathbf{x}^{3}dt=0.\label{eq : weakform}\ ] ] a complete set of eigenvalues @xmath18 and their normalized eigenfunctions @xmath19 is considered to be known for the linear operator as solution of @xmath20 for clarity of notation we introduce composite greek indices with implied summation upon them:@xmath21 and denote the first partial derivatives of the field as separate functions : @xmath22 .
the displacement field allows decomposition upon the eigenfunctions with time - dependent modal amplitudes @xmath23 @xcite:@xmath24 by employing this representation in equation ( [ eq : weakform ] ) , and using the eigenfunctions as a set of test functions @xmath25 , we restate the governing equation as @xmath26 where @xmath27 , and @xmath28 .
all the specifics of particular type of nonlinear behavior are now contained in modal matrices @xmath29 . only symmetrical with respect to the last two indices parts of matrix @xmath30 and symmetrical with respect to the last three indices parts of matrix @xmath31 survive modal summation in ( [ eq : ode ] ) . without loss of generality we put these matrices equal to their corresponding symmetrical parts , and write them in greek index notation as @xmath32 and @xmath33 nonlinear terms of the stress ( [ eq : piolatensor ] ) yield directional tensors in the above formulae:@xmath34 and @xmath35 due to the major and minor symmetries of elastic tensors * @xmath6 * , @xmath7 and @xmath8 @xcite both matrices @xmath36 can be identified as fully symmetric:@xmath37 a quantity which we term linear energy stored in a single mode @xmath38 at time @xmath39 is @xmath40 in the absence of nonlinearity it is equal to the total energy of a mode , and is constant over time .
it can be written in terms of the complex amplitudes @xmath41:@xmath42 the amplitudes @xmath41 arise from the action of external forces and tractions prior to the cutoff time @xmath43 .
they describe evolution of the linear part of the field by means of the modal amplitudes @xmath44 that are found as solutions of the linearized version of equation ( [ eq : ode ] ) , with matrices @xmath36 put to zero @xcite:@xmath45 the energy flow due to nonlinearity , @xmath46 , is obtained from the governing equation ( [ eq : ode ] ) , with modal power input being@xmath47 it must be noted that the energy quantity @xmath48 ( [ eq : e ] ) is not conserved , because it is not a true energy in the nonlinear case
. analysis of the strain energy density ( [ eq : energydensity ] ) reveals that additional terms @xmath49 must be added to @xmath48 to produce a quantity that is conserved .
however , the terms involve summation upon modal amplitudes other than that of the mode at hand , and do not allow simple interpretation in terms of a single mode . thus they are not used .
we restrict ourselves to a class of systems for which the field excited by external forces and tractions has a fully diffuse nature . in experiments such fields
are practically realized , for example , in an elastic solid of a classically chaotic shape , and have statistical properties close to or indistinguishable from those of a field described by a random hamiltonian @xcite .
the normal modes of the system are taken to be centered gaussian vectors with a certain spatial correlation , as was first theoretically conjectured @xcite , and later numerically and experimentally verified @xcite .
we assume that the mean density of states of the normal frequencies is given by the function @xmath50 in the form of weyl - series @xcite , and frequency - frequency correlations can be neglected . since experimental identification of a particular mode
is complicated by the missing level effect or modal overlap we choose to pursue calculation of average spectral density rather than individual modal amplitudes . in order not to distinguish between individual modes at the frequencies of interest , and
thus deal with the average quantities , we limit observation time of the system @xmath39 to be less than the corresponding break ( heisenberg ) time , @xmath51 . on the other hand ,
time @xmath39 is considered larger than transition ( ballistic ) time in the solid ( mean time between two successive scattering events at the boundaries ) , @xmath52 , so that diffuse regime of the field is established .
@xmath53 and @xmath54 stand for characteristic diameter of and wavespeed in the solid .
putting together the two bounds yields @xmath55 , a condition that can be experimentally realized . in the framework of the adopted statistical model
the modes @xmath56 of the linear operator ( [ eq : eigenproblem ] ) are centered gaussian random vectors with variance given by@xmath57 where @xmath58 represents ensemble average , and @xmath59 is a smooth frequency - dependent correlation matrix .
pairwise correlation of the first partial derivatives of the modes@xmath60 is readily obtained from @xmath61 orthonormality of the modes imposes a normalization condition upon the correlation matrix : @xmath62 from the mode statistics the complex amplitudes @xmath41 describing initial field are found to be centered gaussian random numbers with variance@xmath63 the amplitudes @xmath41 relate to the linear energy ( [ eq : e0 ] ) , and give the smooth function @xmath64 an interpretation as a frequency - normalized spectral energy density .
the linear amplitudes @xmath65 ( [ eq : a ] ) form a centered gaussian ensemble as well .
their pairwise time correlation is calculated as@xmath66 both , mean density of states and correlation matrix , can be calculated in terms of the average green s function : @xmath67 , and @xmath68 @xcite . for time scales under consideration , in particular @xmath69 ,
the characteristic wavelength on the frequencies of interest is much smaller than the diameter of the solid : @xmath70 .
the leading order nonlinear contribution thus comes from the bulk rather than near - boundary region of the solid .
this allows us to neglect effects of the latter , and approximate the exact green s function @xmath71 in the solid by the green s function in an unbounded medium @xmath72 .
this approximation implies that @xmath59 has an infinite correlation radius inherited from @xmath72 , and leads to formal integral divergence in the calculations of the following sections . to mend the problem we consider the scattering of the field inside the solid as a diffuse process with a free mean path on the order of the diameter @xmath53 , which now provides a finite correlation radius for the model : @xmath73 @xcite .
the ansatz is justified as the final results turn out not to depend on the specific choice of the value of @xmath53 , as long as it stays much greater than the wavelength .
we assume nonlinear effects to be small and seek solution of equation ( [ eq : ode ] ) by utilizing small perturbation theory , and expanding amplitudes @xmath23 in orders of magnitude @xmath74 , where @xmath65 are given by ( [ eq : a ] ) .
next order amplitudes @xmath75 arise from the presence of nonlinearity , and are determined solely by the linear field:@xmath76.\end{aligned}\ ] ] to obtain successive corrections to the average spectral density of the power flow we expand ( [ eq : pimodal ] ) in a series of magnitudes of @xmath65:@xmath77 as mentioned in section ii , since the power flow arises from the nonlinear mode coupling , the modal energy of the linear field ( [ eq : e0 ] ) is conserved : @xmath78 . for the assumed statistics of the amplitudes @xmath65 ( [ eq : astatistics ] )
the first - order correction to the average power density is zero as well:@xmath79 the power flow expansion ( [ eq : piexpansion ] ) starts with @xmath80 as the leading term .
we express the fourth moments of the amplitudes @xmath65 as double products of their pairwise correlations ( [ eq : astatistics ] ) , and obtain the power flow in terms of the energy density : @xmath81\nonumber \\ & \qquad \qquad \times \left[\frac{\sin \left(\omega _ { k}-\omega _ { l}\pm \omega _ { m}\right)t}{\omega _ { k}-\omega _ { l}\pm \omega _ { m}}+\frac{\sin \left(\omega _ { k}+\omega _ { l}\pm \omega _ { m}\right)t}{\omega _ { k}+\omega _ { l}\pm \omega _ { m}}\right]\biggr \},\label{eq : pi2integrated}\end{aligned}\ ] ] where @xmath82 , is the average normalized energy of a single mode .
the modal sum is evaluated as a frequency integral with integrand weighed by the mean density of states:@xmath83 we note that matrix @xmath31 responsible for the cubic nonlinearity does not enter ( [ eq : pi2integrated ] ) , for its contribution is proportional to @xmath84 .
the centered gaussian statistics of the modes allows averages in the form of @xmath85 to be expressed as triple products of the pairwise correlations ( [ eq : uprimestatistics ] ) . combining terms with the same modal indices we formally write ( with no implied summation on latin indices )
@xmath86 the coupling functions @xmath87 are defined later . expression ( [ eq : pi2integrated ] ) with the factors ( [ eq : nfactors ] ) inserted , though bulky and cumbersome to analyze , gives the leading term in the power flow due to a weak nonlinearity . in experimental practice , however , it is not uncommon to deal with the fields that have the spectral energy density @xmath88 varying smoothly on frequency scales @xmath89 greater than the averaging bandwidth of the limited observation time : @xmath90 . in this case ( [ eq : pi2integrated ] ) simplifies further .
all the terms directly proportional to rapidly oscillating sine factors yield negligible averages:@xmath91 and the modal sums involving such factors are evaluated as follows:@xmath92 & /\left(\omega _ { n}-\omega \right)\\ = & \pi d\left(\omega \right)f\left(\omega \right)\left[\theta \left(\omega \right)+o\left(1/\delta \omega t\right)\right],\end{aligned}\ ] ] where @xmath93 is a unit step function .
another simplification arises if we pay attention only to the frequencies that carried no energy initially : @xmath94 . for such frequencies
a weak change in the energy density due to @xmath80 is not masked by a strong initial linear field , and is convenient for experimental measurement . with the simplifications
mentioned the power flow expression reduces to @xmath95 the only remaining coupling function is given by contraction of the nonlinear and correlation matrices : @xmath96 we note that@xmath97 the power flow in the form of ( [ eq : pi20density ] ) allows simple interpretation : the energy transfered into a given frequency @xmath98 comes from all pairs of frequencies @xmath99 and @xmath100 that have the given frequency as a combination , i.e. equal to their sum or difference : @xmath101 .
the qualitative result is in agreement with and could have been expected from an elementary theory of nonlinear oscillations @xcite .
the symmetry of the nonlinear matrix ( [ eq : nsymmetry ] ) leads to the symmetry of @xmath102 with respect to any interchange of its arguments .
the coupling strength of any triad of frequencies is thus independent of the energy transfer direction .
nevertheless , the power itself exhibits an overall trend of the energy to be transfered up the frequency spectrum , as it is proportional to the density of states at the target frequency @xmath50 .
the expression ( [ eq : pi20density ] ) implies energy growth that is , if @xmath88 s are approximately constant in time , proportional to elapsed time @xmath39 . in discrete spectrum systems such behavior
is found when a triad of frequencies is locked in internal resonance , producing secular terms in the solution obtained by regular perturbation theory .
this is the behavior observed in our case because of the finite time @xmath103 : since individual modes are not resolved , any combination frequency produced by the source frequencies in the frequency range of interest is indistinguishable from at least one of the normal frequencies of the system , which hence lies in effective internal resonance with them .
we return to elastodynamic displacement fields , and specialize to the case of an elastic body composed of known isotropic homogeneous material , the case that holds premium experimental and theoretical interest .
the linear part of the green s function in the unbounded medium yields the correlation matrix @xmath59 ( see appendix a ) : @xmath104\nonumber \\ & \quad -\left(\delta _ { ij}/3-\delta \hat{x}_{i}\delta \hat{x}_{j}\right)\nonumber \\ & \qquad \times \left[\frac{1}{c_{l}^{3}}j_{2}\left(k_{l}\left|\delta \mathbf{x}\right|\right)-\frac{1}{c_{t}^{3}}j_{2}\left(k_{t}\left|\delta \mathbf{x}\right|\right)\right]\biggr \},\label{eq : k}\end{aligned}\ ] ] where @xmath105 is a separation variable , @xmath106 , @xmath107 is total mass of the solid , @xmath108 and @xmath109 are the longitudinal and transverse wavespeeds respectively , and @xmath110 is the spherical bessel function of order @xmath111 . the mean density of states for clamped boundary conditions is calculated in @xcite : @xmath112\\ & -\frac{s_{0}}{8\pi c_{l}^{2}}\omega \left[2+\left(c_{l}/c_{t}\right)^{2}+3\left(c_{l}/c_{t}\right)^{4}\right]\\ & \qquad \bigm /\left[\left(c_{l}/c_{t}\right)^{2}+1\right]+o\left(l / c\right),\end{aligned}\ ] ] with @xmath113 being the surface area of the solid . for the model of nonlinearity at hand appropriate for an isotropic solid , described by the five - constant theory @xcite , and correlation matrix @xmath59 provided by ( [ eq : k ] ) , the coupling function @xmath102 ( [ eq : n0 ] ) assumes the form @xmath114 @xmath115 is a dimensionless function of the frequency ratio and material properties that characterize nonlinear coupling strength between the source and target frequencies .
the function is independent of the linear dimensions of the body and possesses the symmetries according to ( [ eq : nsymmetry ] ) : @xmath116 .
an analytical expression for @xmath115 is available , but too bulky to be presented here .
details of the calculations are found in appendix a. as a sample distribution of the ( linear ) energy density we take two gaussian peaks with half - width @xmath89 and total energies @xmath117 centered at frequencies @xmath118 and @xmath119:@xmath120 the width of the peaks is restricted by condition @xmath90 , as imposed by applicability requirements for the simplified power input expression ( [ eq : pi20density ] ) .
it is additionally assumed that the peaks do not overlap : @xmath121 .
the suggested form of the spectrum closely models the combined spectra of two narrow - band signals that might be used in an experiment . by adjusting the carrying frequencies
the features of nonlinear mode coupling can be investigated , and the form of @xmath115 measured . as a parameter suitable for characterization of the nonlinear energy transfer strength we consider transfer times defined as the formal time required for entire mean energy of the source peaks ( [ eq : epsgauss ] ) to be transfered into combination frequencies , provided small perturbation theory and equation ( [ eq : pi20density ] ) holds : @xmath122 , and @xmath123 .
here @xmath80 represents the total power input into frequency band supporting the resulting peak at the combination frequency . according to the definition @xmath124 , @xmath125 , and @xmath126
characterize energy transfer into double , sum and difference frequency respectively .
calculations yield these times to be@xmath127 as a measure of absolute strength of the linear field we choose squares of the mach numbers : @xmath128 and @xmath129 . as seen from ( [ eg : transferts ] ) stronger nonlinear effects ( shorter times ) require higher mach numbers , i.e. stronger linear fields . for numerical estimates we take an aluminum block and a fused silica block with the mechanical properties listed in table [ table : properties ] , [ cols="^,^,^,^,^,^,^,^,^",options="header " , ] and choose the transverse wavespeed for each material as the characteristic one : @xmath130 . due to the inversion symmetry
we evaluate the coupling function @xmath115 for frequency range @xmath131 only , and plot it in figure [ fig : n ] . dimensionless coupling function in aluminum ( thick line ) and fused silica ( thin line ) . ] immediately a significant feature of the plot comes into view : the sharp peaks and discontinuities of the coupling function at frequency ratios of @xmath132 and @xmath133 for aluminum , and @xmath134 and @xmath135 for fused silica .
the feature may be understood by considering nonlinear combinations of plane waves .
we recall that from a plane wave perspective appropriate in the limit @xmath70 , in addition to internal frequency resonance between the source frequencies @xmath99 and @xmath136 and the target frequency @xmath98 , there is an additional requirement for wavevector resonance .
we note that there are two plane wave types in the system , namely longitudinal and transverse .
if the types of the three waves ( at @xmath98 , @xmath99 and @xmath100 ) are identical , wavevector resonance ( @xmath137 with @xmath138 ) is always possible , and demands that the three wavevectors are parallel or antiparallel .
if the wave type for one of the frequencies is different from the other two , then wavevector resonance is not always possible ; it depends on the frequency ratio .
if it is possible , there will be a nontrivial angle between the wavevectors .
the transition between possibility and impossibility occurs at certain special frequency ratios , @xmath139 . at these ratios wavenumber resonance occurs with parallel or antiparallel wavevectors
the special frequency ratio depends solely on the poisson ratio of the material . for non - dispersive single wavespeed systems , when @xmath140 , the ratios become @xmath141 and @xmath142 . in this case
the effect is not observed , as one of the source frequencies needs to be zero @xcite .
however , for certain dispersive single wavespeed systems the peaks and discontinuities might still be found , if the dispersion equation of the system is such that the wavenumber resonance is possible .
it is also worth noting that the discontinuity position at leading order is independent of the field strength .
thus field calibration is not required in experimental measurements in order to observe this characteristic feature , for only relative values of @xmath115 are needed .
this makes the method convenient to use in the light of its eventual application to nde . in order to see whether the presented theory is applicable for experimental verification and eventual use
, we need to provide numerical estimates for transfer times ( [ eg : transferts ] ) . for this purpose
we choose the solid to be equivalent in volume and surface area to a cube with a side of @xmath143 .
the typical values of the transition ( ballistic ) time in the system are of the order @xmath144 .
the carrying frequencies of the two narrow - band signals are taken to be @xmath145 and @xmath146 , so that the break ( heisenberg ) time is of the same order as the absorption times of @xmath147 common to experiment .
the transfer times are calculated for the aluminum block and plotted in figure [ fig : t ] as functions of the linear field strength , characterized by mach number .
transfer times in aluminum block for fixed @xmath145 and @xmath146 . ]
it is feasible to generate linear diffuse fields with rms elastic strains corresponding to mach numbers of order @xmath148 that yield transfer times of order @xmath149 ( see figure [ fig : t ] ) .
thus the observation times of the order of the absorption time yield energy densities at combination frequencies some @xmath150 times weaker than the source signal energy , i.e. rms strains of @xmath151 .
these are easily detectable .
we also note that the strain ratio ( combination field to initial field ) is much smaller than unity , so the use of regular perturbation theory is valid for the given time scale .
finally , we calculate transfer times in aluminum and fused silica as a function of source frequency ratio for a fixed mach number ( see figure [ fig : tfixed ] ) .
transfer times in aluminum ( thick lines ) and fused silica ( thin lines ) block for varying source frequency ratio , and fixed @xmath152 and @xmath153 . ]
the frequency dependence of the nonlinear coupling strength in terms of the transfer times reveals the frequency dependence of @xmath115 .
the dips in the transfer times correspond to the peaks of the coupling function , and might be used to estimate poisson s ratio of the material , or as a signature of nonlinearity .
we notice that the energy transfer into near - zero frequencies is inefficient , as manifested by high difference frequency transfer times in the vicinity of @xmath154 .
in the present work we have obtained a formula for the evolution of the average linear energy spectrum ( [ eq : pi20density ] ) of a weakly nonlinear system . except for definition of the modal linear energy @xmath48 ( [ eq : e ] ) , and statistical properties of the normal frequencies @xmath155 and modes @xmath56 ( [ eq : eigenproblem ] ) , the result has no reference to the physical nature of the dynamic system governed by nonlinear equations ( [ eq : ode ] ) . with proper generalization of the matrices @xmath29 describing the physical nature of the nonlinearity it might be applicable beyond the scope of elasticity .
we observe that the energy redistribution occurs for triads of frequencies , with one being the sum or difference of the other two .
the average power input into a narrow frequency band is found to be a cumulative effect coming from such interactions in the initial energy spectrum of the field , and to be proportional to the convolution of the energies stored at two frequencies .
relative weight of the interactions is given by the frequency - dependent coupling function @xmath102 ( [ eq : n0 ] ) , and is calculated as a contraction of the nonlinear and correlation matrices @xmath36 and @xmath156 .
the case of chief experimental interest involving an isotropic homogeneous elastic body with nonlinearity given by the five - constant theory and an initial energy spectrum consisting of two narrow - band signals is discussed in detail .
we find that characteristic times for the full energy transfer from the source into combination frequencies depend on the ratio of the source frequencies and exhibit characteristic dips and peaks and discontinuities at special frequency ratios corresponding to wavenumber resonance .
the position of the resonance depends on poisson s ratio .
we also find that the transfer times are reciprocal to the strength of the initial signal given by the square of the corresponding mach number .
the current theory is derived for zero - displacement boundary conditions of a clamped solid , and is impractical for experimental realization . however , for the frequencies of interest the main contribution to the mixing comes from the bulk of the solid . since the near - boundary regions play a lesser role , the authors believe that the energy spectrum evolution for the traction - free elastic solid would exhibit behavior similar to that discussed here , and thus be accessible for experimental verification .
zaitsev , a.m. sutin , i.yu .
belyaeva , and v.e .
nazarov , j. vibration and control * 1*(3 ) , 335 ( 1995 ) ; alexander m. sutin and dimitri m. donskoy , in _ proceedings of the 8th international symposium on nondestructive characterization of materials , boulder , colorado , 1997 _ , edited by robert e. green jr .
( plenum press , new york , 1998 ) a. wegner , a. koka , k. janser , u. netzelmann , s. hirsekorn , and w. arnold , ultrasonics * 38 * , 316 ( 2000 ) ; m. rothenfusser , m. mayr , and j. baumann , _ ibid . _ * 38 * , 322 ( 2000 ) d.c .
hurley , j. acoust .
106*(4 ) , 1782 ( 1999 ) w.j.n . de lima and m.f .
hamilton , j. sound vib . * 265*(4 ) , 819 ( 2003 ) hans - jrgen stckmann , _ quantum chaos : an introduction _ , ( cambridge university press , 1999 ) , chap . 2 , 6 d.j .
pine , d.a .
weitz , p.m. chaikin , and e. herbolzheimer , phys .
60*(12 ) , 1134 ( 1988 ) ; richard l. weaver and wolfgang sachse , j. acoust .
97*(4 ) , 2094 ( 1995 ) johannes f. deboer , ad lagendijk , rudolf sprik , and shechao feng , phys .
lett . * 71*(24 ) , 3947 ( 1993 ) r. bressoux and r. maynard , europhysics letters * 50*(4 ) , 460 ( 2000 ) ; s.e .
skipetrov and r. maynard , phys .
. lett . * 85*(4 ) , 736 ( 2000 ) ; s.e .
skipetrov , phys .
e * 63 * , 056614 ( 2001 ) r.w .
ogden , _ nonlinear elastic deformations _
( ellis horwood ltd . , chichester , 1984 ) , chap .
3.4 , 4.3 , 6.1 under zero - displacement boundary conditions solution of the nonlinear problem belongs to the function space with a basis formed by the eigenfunctions of the linear operator , thus allowing the proposed decomposition .
for zero - traction boundary conditions such decomposition is no longer valid , since the eigenfunctions no longer span the entire solution space of the nonlinear problem . r.l .
weaver , j. sound vib . * 94*(3 ) , 319 ( 1984 ) r.l .
weaver , j. acoust .
85*(3 ) , 1005 ( 1989 ) c. ellegaard , t. guhr , k. lindemann , h.q .
lorensen , j. nygrd , and m. oxborrow , phys .
lett . * 75*(8 ) , 1546 ( 1995 ) ; c. ellegaard , t. guhr , k. lindemann , j. nygrd , and m. oxborrow , _ ibid . _ * 77*(24 ) , 4918 ( 1996 ) ; a. andersen , c. ellegaard , a. d. jackson , and k. schaadt , phys .
e * 63 * , 066204 ( 2001 ) m.v .
berry , j. phys a * 10*(12 ) , 2083 ( 1977 ) ; p. oconnor , j. gehlen , and e.j .
heller , phys .
lett . * 58*(13 ) , 1296 ( 1987 ) steven w. mcdonald and allan n. kaufmann , phys . rev .
a * 37*(8 ) , 3067 ( 1988 ) ; k. schaadt , t. guhr , c. ellegaard , and m. oxborrow , phys .
e * 68 * , 036205 ( 2003 ) the expressions are obtained by calculating ensemble average of the imaginary part of the green s function @xmath157 ; e.n economou , _
green s functions in quantum physics _
( springer - verlag , new york , 1979 ) , chap
prigodin , b.l .
altshuler , k.b .
efetov , and s. iida , phys .
lett . * 72*(4 ) , 546 ( 1994 ) ; v.n .
prigodin , _ ibid .
_ * 74*(9 ) , 1566 ( 1995 ) leonard mierovitch , _ elements of vibration analysis _ ( mcgraw - hill , new york , 1986 ) , chap .
10 marc dupuis , robert mazo , and lars onsager , j. chem . phys . * 33*(5 ) , 1452 ( 1960 ) f.d .
murnaghan , _ finite deformations of an elastic solid _ ( chapman & hall , new york , 1951 ) , chap . 4 r.t smith , r. stern , and r.w.b stephens , j. acoust .
* 40*(5 ) , 1002 ( 1966 ) d.s .
drumheller , _ introduction to wave propagation in nonlinear fluids and solids _
( university press , cambridge , 1998 ) , app .
b r. bechmann and r.f.s .
hearmon , in _ elastic , piezoelectric , piezooptic , electrooptic constants , and nonlinear dielectric susceptibilities of crystals _ ,
_ _ _ _ edited by k.h .
hellwege and a.m. hellwege , landolt - brnstein , new series ( springer - verlag , new york , 1969 ) , vol .
iii/2 for this reason a discussion of weak nonlinearity in such simple systems would offer little insight and few rich behaviours .
weaver , j. mech .
solids * 38*(1 ) , 55 ( 1990 ) ; j.d .
achenbach , _ wave propagation in elastic solids _ ( elsevier , amsterdam , 1999 ) , chap . 3 b. eckhardt , u. drr , u. kuhl , and h .- j .
stckmann , europhys . lett . *
46*(2 ) , 134 ( 1999 )
elastic deformation of an isotropic homogeneous solid is described by the five - constant theory in terms of the lam constants @xmath158 and @xmath159 , and nonlinear coefficients @xmath160 , @xmath161 and @xmath162 @xcite . according to the theory the linear and second - order nonlinear elastic tensors have the form of@xmath163 the elementary isotropic tensors @xmath164
are given as follows:@xmath165 @xmath166 by substituting tensors ( [ eq : elastictensors ] ) into the strain energy definition ( [ eq : energydensity ] ) , particular form of directional tensors @xmath167 ( [ eq : ntensor ] ) is derived .
the green s function in the unbounded medium @xmath72 is calculated from its spatial fourier - transform @xcite : @xmath168 direct integration of the above expression and its subsequent normalization ( [ eq : knorm ] ) yields particular expression for correlation matrix @xmath59 ( [ eq : k ] ) . as two of its particular limit cases we note , first , the known autocorrelation function for scalar helmholtz equation : @xmath169 @xcite obtained by letting @xmath170 . and
second , the autocorrelation function for purely transverse field such as , for example , electromagnetic field @xcite obtained by letting @xmath171 .
the correlation matrix of the first partial derivatives of the modes is derived from @xmath59 by means of ( [ eq : kprimecalc ] ) .
it is expressed in terms of known directional tensors @xmath172 , characteristic wavenumber @xmath173 , and wavespeed ratios @xmath174 : @xmath175 the sum over longitudinal and transverse wave types is denoted as @xmath176 .
directional tensors @xmath172 are defined by the following expressions:@xmath177,\end{aligned}\ ] ] @xmath178 @xmath179 the quadruple , composite quadruple and @xmath180-order directional moments * @xmath181 * moments yield zero values when integrated upon all spatial directions or contracted upon any two pairs of indices : @xmath182 the coupling function ( [ eq : n0 ] ) is given by contraction of the directional tensors of the nonlinear matrix ( [ eq : ntensor ] ) with correlation matrices of the first derivatives ( [ eq : kprime ] ) .
it is used to find the dimensionless function @xmath183 where constants @xmath184 are @xmath185 and the integral @xmath186 is given by @xmath187 for a finite correlation radius of the coupling matrix @xmath53 , the integral additionally depends on the target frequency @xmath98 .
the dependence turns out to be only significant in the small vicinity of @xmath188 , elsewhere the contribution being small : @xmath189 .
we note that for the mentioned ratios @xmath190 one of the source frequencies @xmath99 or @xmath191 must be close to zero , and the phenomenon is of small practical importance from experimental point of view . for these near - zero frequencies the wavelength becomes comparable to or greater than the diameter of the solid .
thus the substitution of the exact green s function @xmath71 by the green s function in the unbounded medium @xmath72 is no longer valid , and conditions for the time scales made in section iii are not met . with these limitations in mind we set @xmath192 for practical calculations , and
obtain analytical expression for ( [ eq : i ] ) .
as the result of this procedure a singularity of the function @xmath115 at @xmath193 is acquired ( not shown in figure [ fig : n ] ) .
discontinuities of the function at the wavenumber resonances , and discontinuities of its slope at @xmath194 and @xmath195 , are found as well .
we expect in practice to observe sharp transitions over finite ranges in @xmath196 of order @xmath197 , at a characteristic frequency ratio @xmath196 possibly shifted by an amount @xmath189 . | we study the evolution of diffuse elastodynamic spectral energy density under the influence of weak nonlinearity . it is shown that the rate of change of this quantity is given by a convolution of the linear energy at two frequencies .
quantitative estimates are given for sample aluminum and fused silica blocks of experimental interest . |
it is not clear when the mantra `` room temperature superconductivity '' first passed the lips of superconductivity researchers , but it has attained some prominence in the literature since little used it in the title ( `` superconductivity at room temperature '' ) of an overview in 1965@xcite of his electronic polarization mechanism proposed for polymeric systems .
ginzburg immediately transitioned this to two - dimensional metal - insulator sandwich materials and generalized the description.@xcite sad to say , these electronic - mediated superconductors are yet to be realized .
uses of the term _ room temperature superconductivity _ have ranged from general speculations@xcite to glimpses of `` glitchite''@xcite to editorial coverage of public announcements@xcite to preplanning for applications.@xcite in the ensuing decades vitaly ginzburg has been the most visible and persistent advocate that there is no theoretical justification for pessimism .
ginzburg was elaborating more generally on `` high temperature superconductivity '' by@xcite 1967 and his optimism has never flagged .
hts ( cuprate high temperature superconductors ) set the standard with t@xmath0 90 k in 1987 and rising to @xmath1130 k by the mid-1990s , but there have been no further increases . since there is no viable theory of the mechanism of hts , there is no approach that will allow the rational design of new materials with higher @xmath2 in that class .
the phonon mechanism however has an accurate microscopic theory that is valid to reasonably strong coupling , and therefore invites the design of materials with higher t@xmath3 .
the recent and very surprising breakthroughs in phonon - mediated superconductivity can be put to use to provide a plausible recipe for the design of a room temperature superconductor .
this is the purpose of this modest proposal , where a central feature is the emphasis on _ control _ of the coupling , versus the standard approach of increasing the _ brute strength _ of the coupling .
nearing the end of the 1970s there was pessimism about raising the superconducting transition temperature ( t@xmath3 ) significantly , and there were claims ( mostly unpublished ) that the maximum t@xmath3 was around 30 k. the only superconductivity known at the time ( excluding the superfluidity of @xmath4he which was unlike anything seen in metals ) was symmetric ( @xmath5-wave ) pairing mediated by vibrations of the atoms ( phonons ) .
the maximum t@xmath3 known at the time was 23 k in the a15 system .
the maximum t@xmath3 had increased by only 6 k in 25 years , and it was not due to increase for several more years ( 1986 , when hts was announced ) .
the discovery@xcite in 2001 of mgb@xmath6 , with t@xmath3=40 k , was stunning not only because of the unimaginably high t@xmath3 for phonon - mediated superconductivity , but also because it appeared in a completely wrong kind of material according to the accumulated knowledge .
this aspect has been dealt with in some detail@xcite now , and the pairing mechanism , character , and strength is well understood . while there has been some effort to apply what has been learned from mgb@xmath6 to find other similar , or even better , superconductors ,
so far mgb@xmath6 remains in a class by itself .
diamond , when it is heavily doped with b , becomes superconducting@xcite with reports as high as 12 k. b - doped diamond has features in common@xcite with mgb@xmath6 except for the two - dimensionality ( 2d ) of mgb@xmath6 s electronic structure .
it is this 2d character we deal with here , one important aspect of which has not been emphasized previously .
we confine our consideration to ambient pressure ; applied pressure could well provide enhancements .
migdal - eliashberg ( me ) theory of superconductivity is firmly grounded on the material - specific level@xcite and is one of the impressive successes of condensed matter theory . stripped to its basics , ( phonon - coupled )
t@xmath3 depends on the characteristic frequency @xmath7 of the phonons and on the strength of coupling @xmath8 ; in the strong coupling regime of interest here the retarded coulomb interaction loses relevance .
the essence of increasing t@xmath3 lies in making one or both of these characteristic constants larger while avoiding structural instability .
the focus here is on the decomposition@xcite of @xmath8 @xmath9 where the mode-@xmath8 for momentum @xmath10 is given for circular fermi surfaces in two dimensions by @xmath11^{-2 } \sum_k \delta(\varepsilon_k)\delta(\varepsilon_{k+q } ) = \frac{1}{\eta_q\sqrt{1-\eta_q^2 } } \label{xiq}\end{aligned}\ ] ] where @xmath12 .
here @xmath13 is the fermi level ( e@xmath14=0 ) density of states , @xmath15 is the squared electron - displaced ion matrix element averaged over the fermi surface , @xmath16 is the atomic mass , and @xmath17 is the characteristic physical phonon frequency .
the el - ph matrix element @xmath18 involves @xmath19 , @xmath16 , and @xmath17 in the standard way . here the band degeneracy @xmath20 factor ( number of fermi surfaces ) is treated as in mgb@xmath6 , where the two @xmath21 surfaces are equally important . in the conventional adiabatic approximation ,
phonon renormalization is given by the real part of the phonon self - energy at zero frequency @xmath22 , \eta_q > 1 .
\label{chiq}\end{aligned}\ ] ] both here and above the final expression has been expressed for mgb@xmath6-like systems with cylindrical fermi surfaces.@xcite the behavior of this renormalization is pictured in fig .
[ renorm ] , note specifically the independence of the degree of softening on @xmath23 .
it is the last result eq .
[ chiq ] that we emphasize here : the phonon renormalization is constant ( _ i.e. _ controlled ) for @xmath24 , and diminishes quickly for larger @xmath10 . for circular fermi surfaces there is no sharp peak in @xmath25 at @xmath26 nor anywhere else . in mgb@xmath6 ,
phonon coupling arises dominantly from @xmath24 , and only the two bond - stretching branches have the very large matrix elements .
the number of phonon scattering processes from / to the fermi surface is quantified by the `` nesting function '' @xmath27 ( given here in normalized form whose zone sum is unity ) . for the 2d circular dispersion relation ,
it is the simple known quantity in eq .
it has simple integrable divergences at @xmath10=0 and @xmath28 which , most importantly , do not result in @xmath10-dependent softening , hence avoiding instability in spite of contributing positively to @xmath8 .
this behavior can be contrasted with the vast uncertainty inherent in a general fermi surface , which has arisen forcefully in recent discoveries . under pressure
elemental li metal becomes superconducting up to near 20 k as shown by three groups,@xcite in spite of being an @xmath29 metal with a simple fermi surface.@xcite in fig .
[ xiq3d ] @xmath30 is displayed in three planes@xcite in the zone , where extremely sharp structure is apparent .
the sharp structure near the k symmetry point leads to a lattice instability ( large contribution to @xmath31 ) .
the sharp structure occurs in spite of a very simple cu - like fermi surface consisting of spheres joined by necks along @xmath32 directions .
this example demonstrates why _ control _ of the @xmath10-dependence of the coupling is essential ; @xmath8 is not so large@xcite for li at this volume ( @xmath33 ) yet the lattice has been pushed to instability .
this example shows that the overall value of @xmath8 is not the indicator of instability of the system , which occurs when some phonon frequency is renormalized ( softened ) to zero .
the crucial features to be learned from mgb@xmath6 and li under pressure follow .
high frequency is important ; this has long been understood . by beginning from a very stiff unrenormalized lattice ,
a crystal can withstand a great deal of renormalization to lower frequencies that must accompany strong coupling .
+ 2 . very high
mode @xmath34 s ( up to 20 - 25 for mgb@xmath6 ) can arise without instability .
mgb@xmath6 appears not to be near any instability , although analysis shows@xcite that only @xmath115 - 20% stronger coupling would result in instability .
the phonon softening in 2d systems due to strong coupling , and the total @xmath8 , is independent of carrier concentration ( or varies smoothly if the effective mass changes with doping ) .
impressive results can be attained from strong coupling to only a fraction of the modes . for mgb@xmath6 , t@xmath3 = 40 k with
only 3% of the modes strongly coupled .
these modes are the bond - stretch modes ( 2 out of 9 branches ) with @xmath24 ( 12% of the zone ) .
general - shaped fermi surfaces , even the simple one of li , can readily lead to lattice instability from a thin surface of soft modes in q - space .
such instability restricts the achievement of high t@xmath3 .
two - dimensional parabolic bands provide _ ideal control _ of the q - dependence of coupling strength : phonon renormalization is _ constant _ for @xmath24 , not sharply peaking in an unexpected region of the zone as can happen for a fermi surface of general shape ( viz .
for li in fig .
[ xiq3d ] ) .
the superconducting t@xmath3 of mgb@xmath6 is remarkable , and arises from the extreme strong coupling of 3% of its phonons .
the electron - phonon deformation potential @xmath35 is very large for the bond - stretch modes , and mgb@xmath6 gains a factor @xmath36=4 from having two fermi surfaces .
figure [ hex1 ] illustrates ( 1 ) the @xmath21-band fermi surfaces , idealized to identical circular surfaces , and ( 2 ) the kohn - anomaly region of renormalized phonons @xmath24 . in mgb@xmath6
this `` kohn surface '' encloses only 12% of the zone .
previous analysis@xcite has shown that pushing these phonons further , by either increasing the bare coupling strength @xmath37 or varying the lattice stiffness ( bare phonon frequency ) separately , could increase t@xmath3 by perhaps 20% but then would drive the material to a lattice instability .
it is fairly obvious that if one increases both proportionately , then one gains by enhancing the characteristic frequency while keeping @xmath8 constant .
this is the _ metallic hydrogen _
scenario that was discussed originally 40 years ago.@xcite there is however the equally obvious approach : introduce more phonons into the coupling . from this viewpoint ,
mgb@xmath6 with its 3% of useful phonons seems pathetic , yet as noted in sec .
iii , increasing the number of strongly coupled phonons be increasing the doping does not enhance @xmath8 because the individual mode @xmath34 s decrease accordingly . and further increasing the matrix elements , which increases @xmath8 , also readily leads to instability .
the key to increasing @xmath8 is not in doing more ( more carriers , hence more phonons ) , but in doing it _
again _ and _ again _ , with other fermi surfaces and other phonons . here the _
, not over the strength of the coupling but over the _ specific momenta _ [ that is , @xmath34 , or more specifically @xmath25 ] becomes essential .
the requirement is to renormalize frequencies in other parts of the zone _ without further softening _ the ones that are already very strongly renormalized , since that will not lead to lattice instability .
two - dimensional circularly - symmetric dispersion relations provide that control , as demonstrated by the equations above .
adding an additional circular fermi surface centered at @xmath39 gives another region of renormalized phonons , and contributions @xmath34 to @xmath8 , of radius @xmath26 centered at @xmath39 .
let us skip directly to the optimal case , illustrated in fig .
[ hex19 ] . by adding to the mgb@xmath6 zone - centered surface another spherical fermi surface half - way along the @xmath40-k line , with radius 1/8 of the length of the @xmath40-k line , one obtains three new spanning vectors @xmath41 and their symmetry partners that produces an array of close - packed kohn surfaces of radius @xmath26 within which coupling is strong .
given the close - packing fraction in 2d , @xmath42=0.907 , this arrangement manages to utilize 90% of the zone for strongly coupled phonons , an increase in fraction of zone used , compared to mgb@xmath6 ( 12% ) , by a factor of 7.5 . this extension from mgb@xmath6
is not yet optimum , because mgb@xmath6 uses only 2/9 of its branches .
optimally , every branch would be drafted into service in strong coupling , giving another factor of 4.5 , or a total enhancement of @xmath130 - 35 .
the strongly coupled modes in mgb@xmath6 have mode-@xmath34 s of mean value 20 - 25 ( calculations so far have not been precise enough to pin this down ) .
let us not be pessimistic , and therefore use @xmath43=25 ; this value is consistent with the value of @xmath8 from the strongly coupled modes divided by 3% , i.e. 0.7/0.03@xmath123 . then with 90% participation of the phonons @xmath8=25@xmath44=22.5 . using the allen - dynes equation@xcite to account properly for the strong - coupling limit that is being approached , and using the mgb@xmath6 frequency of 60 mev
, one obtains t@xmath3=430 k. the strong coupling limit@xcite of the ratio @xmath45 is 13 , so we can estimate the gap of such a superconductor to be @xmath46 ev .
this will be an interesting superconductor indeed . whereas the architectural design of a room temperature superconductor provided here
is straightforward , the structural engineering necessary to implement this vision will require creativity and knowledge of materials . where does one look ?
figure [ hex19 ] gives an idea of a set of fermi surfaces that could be promising .
this pattern is very much like that of na@xmath47coo@xmath6 , although its t@xmath3=4.5 k , and even that not necessarily from phonons , is not a good example of strong coupling enhanced by fermi surface arrangement .
( the zone - centered fermi surface in na@xmath47coo@xmath6 is much larger than the others . )
the encouraging feature to be emphasized here is that this example shows that the desirable arrangement of figure [ hex19 ] is nothing unusual .
not only are the fermi surfaces very near the midpoint of the @xmath40-k line , but they are nearly circular although not enforced by symmetry .
a more provocative example is that of the layered , electron doped system li@xmath47mncl , which has up to t@xmath3=15 k for m = zr , and t@xmath3=26 k for m = hf . for @xmath48 which probably includes the accessible concentrations
, this system has very nearly circular k - centered fermi surfaces whose nesting possibilities have been noted.@xcite there are surfaces at two inequivalent points k , resulting in kohn regions centered at the zone center ( degeneracy two ) and [ because k@xmath49k scattering involves momentum transfer k in the hexagonal zone ] at each of the pair of points k ( also degeneracy two ) .
the figure of kohn surfaces would look like fig .
[ hex19 ] with circles centered at the @xmath39 and @xmath50 points missing .
indeed , heid and bohnen calculate@xcite phonon softening around k and the zone center , but alas they find that the calculated coupling strength comes well short of accounting for t@xmath3=15 k in li@xmath47zrncl ( not to mention the question of t@xmath3=26 k in li@xmath47hfncl ) .
the superconductivity in this system remains unexplained , but the placement of its 2d circular fermi surfaces illustrates the first step in pulling more phonons into the coupling . while the electronic system we seek is 2d in character ,
we also seek ( strong ) coupling to phonon of arbitrary momentum and polarization .
mgb@xmath6 is a bad example here , since the strong bonding lies solely within the layer ; the @xmath51-axis modes are hardly coupled .
a lattice with strong bonds with substantial component perpendicular to the layers have a better likelihood of coupling to bond stretch modes . from this point of view
the li@xmath47zrncl system seems to be a step in the right direction , as it has zr - n bonds both within the layer and perpendicular to it .
heid and bohnen@xcite found moderate coupling both to two stiff modes ( @xmath165 mev , primarily n ) and to two softer modes ( @xmath125 mev ) .
these modes are however polarized in - plane ; the perpendicular zr - n modes do not couple to the 2d electronic system .
creativity will be required to find how to involve a larger fraction of modes in the coupling .
since there may be people who are truly interested in this topic of room temperature superconductivity , the basic premises and concepts should be made as clear and precise as possible ( `` truth in advertising '' ) .
the following issues should be mentioned .
the use of the term _ optimal _ as used above was disingenuous and incorrect , and that discussion as it stands is unduly pessimistic .
close - packing of @xmath52-diameter kohn circular surfaces does indeed use the area of the zone efficiently ( although mathematicians or engineers would fill the holes with smaller circles , and then again and again ad infinitum , thereby using 100% of the area ) .
however , maximizing the area is _ not _ the objective .
the objective is to maximize t@xmath3 or to simplify slightly for now , to maximize @xmath8 .
phase space in 2d leads to the non - intuitive result that integration of @xmath34 over a kohn region of diameter @xmath52 is independent of @xmath23 , as long as the me theory we are applying holds .
the objective then is to maximize the _ number _ of kohn regions .
since there are likely to be different values of @xmath53 and @xmath17 for the different regions , it will be something like the quantity @xmath54_j}{\omega_j } \label{optimum}\end{aligned}\ ] ] that needs maximizing , where @xmath55 is the ( variable ) number of kohn regions .
care must be taken to keep the kohn regions from overlapping if the combined renormalization of phonons will be so strong as to drive instability .
also , the band degeneracy factor @xmath20 seems to be extremely helpful , but must be watched .
[ optimum ] there would be some factor related to @xmath20 of identical terms . ) for example , in fig .
[ hex19 ] each of the fermi surfaces contributes ( via intrasurface scattering ) to the coupling strength and accompanying phonon softening inside the kohn region at @xmath56 , a factor of 12 in this case .
a good strategy would be to have matrix elements be larger for @xmath41 , which have smaller degeneracy factors , than for @xmath57 with its large multiplier , thereby crafting strong coupling while avoiding the @xmath58 instability .
the validity of the theory bears reconsideration .
it is safe to say that if @xmath59 for every phonon @xmath10 , then me ( second order perturbation ) theory is safe . if this inequality is not satisfied for a small fraction of the coupling strength , corrections are probably minor .
however , for the typical strongly - coupled phonon in mgb@xmath6 , @xmath60 the condition of validity is violated badly for _ every important phonon _ , making me theory in mgb@xmath6 unjustified as pointed out earlier.@xcite nevertheless , its predictions seem to be reasonable for mgb@xmath6 , so it is reasonable for us to extrapolate the theory to include more phonons with similar strength of coupling .
it was noted above that for a frequency @xmath61 60 mev as in mgb@xmath6 , @xmath62 20 is required to reach the vicinity of room temperature .
while there have been many papers addressing the very strong regime and resulting polarons and bipolarons , we are not aware of any that address seriously such coupling in a degenerate electron system .
there is a general expectation that the electron system becomes unstable , but unstable to what is unclear ; a degenerate gas of polarons ( of the order of one per atom ) does not seem like a clear concept .
( if the instability occurs only below t@xmath3 , it may not control or limit the pairing at high temperature . )
the description of such a system is still unknown ; while mgb@xmath6 has a finite concentration ( 3% ) of _ extremely _ strongly coupled phonons , the net value of @xmath8 is less than unity . following
the materials design proposed here will lead to study of a new materials regime as well as higher t@xmath3 .
a blueprint for the design of a room temperature superconductor has been provided here .
the essence is this : very strong q - dependent coupling in a _ controlled _ fashion is possible , and one follows the path toward getting as much as one can out of both the electronic and the phononic systems .
two - dimension electronic structures with circular fermi surfaces give an unsurpassed level of control of the q - dependence , which if uncontrolled can and does lead to structural instabilities even at moderate total coupling strength .
stiff lattices are important , as is getting as many of the branches as possible involved in coupling . on the one hand mgb@xmath6 ,
impressive as it is , seems to be doing a poor job in most respects of making use of the available phase space .
mgb@xmath6 uses only 12% of all the possible phonon momenta , and only 2/9 of its phonon branches , netting only 3% of phonons involved in coupling .
mgb@xmath6 excels at producing extremely large electron - displaced ion matrix elements , and its 2d phase space keeps the extremely large coupling ( to those 3% ) firmly under control .
all things considered , it seems reasonable to expect that materials exist , or can be made , that will improve on mgb@xmath6 s current record .
there is no promise here , nor even expectation , that such improvement will be easy . although one can employ tight - binding models to suggest crystal structures and interatomic interactions that will place band extrema in the desired positions in the zone , synthesizing the corresponding material is more uncertain .
we work with discrete nuclear charges , so this is a diophantine problem rather than a continuous one , and bonding properties can change rapidly from atom to atom .
in addition , incremental improvements lead to even more incremental payoffs .
the behavior@xcite of t@xmath3(@xmath8 ) @xmath63 in the strong - coupling regime , while monotonically increasing , provides a law of diminishing returns : doubling t@xmath3 requires quadrupling @xmath8 , a sobering prospect .
finally , it should re - emphasized that no realistic theory of the degenerate electronic system in the presence of phonon coupling in the regime @xmath64 ( say ) exists , but this is another , perhaps better , physical reason to push to stronger coupling systems . and think of it would nt it be really great to carry around a spool of superconducting d@xmath6e@xmath65z wire in your pocket ?
this work was stimulated in part by preparation for , and attendance of , the workshop on _ the possibility of room temperature superconductivity _
, held june 2005 at the university of notre dame .
i have benefited from discussion and collaboration with numerous workers in the field .
our recent work in this area has been supported by national science foundation grant dmr-0421810 .
support from the alexander von humboldt foundation during the latter part of this work is gratefully acknowledged .
10 w. a. little , scientific american * 212 * , 21 ( 1965 ) . v. l. ginzburg , phys .
* 13 * , 101 ( 1964 ) ; sov .
jetp * 19 * , 269 ( 1964 ) . _ on the problem of high temperature superconductivity _ j. ladik and a. bierman , phys .
a * 29 * , 636 ( 1969 ) .
_ on the possibility of room - temperature superconductivity in double stranded dna _
k. antonowicz , nature * 247 * , 358 ( 1974 ) .
_ possible superconductivity at room temperature _ j. langer , solid state commun . * 26 * , 839 ( 1978 ) . _ unusual properties of aniline black does superconductivity exist at room - temperature ? _
k. s. jayaraman , nature * 327 * , 357 ( 1987 ) .
_ superconductivity at room - temperature _ c. k. n. patel and r. c. dynes , proc .
sci . * 85 * , 4945 ( 1988 ) .
_ towards room - temperature superconductivity
_ v. l. ginzburg and d. a. kirzhnitz , doklady akad .
sssr * 176 * , 553 ( 1967 ) .
_ on high - temperature and surface superconductivity
_ j.m .
an and w.e .
pickett , phys .
lett . * 86 * , 4366 ( 2001 ) y. kong , o. v. dolgov , o. jepsen , and o. k. andersen , phys .
b * 64 * , 020501 ( 2001 ) .
j. kortus , i. i. mazin , k. d. belashchenko , v. p. antropov , and l. l. boyer , phys .
lett . * 86 * , 4656 ( 2001 ) . i. i. mazin and v. p. antropov , physica c * 385 * , 49 ( 2003 ) .
e. a. ekimov , v. a. sidorov , e. d. bauer , n. n. melnik , n. j. curro , j. d. thompson , and s. m. stishov , nature * 428 * , 542 ( 2004 ) .
y. takano , m. nagao , k. kobayashi , h. umezawa , i. sakaguchi , m. tachiki , t. hatano , and h. kawarada , appl .
lett . * 85 * , 2581 ( 2004 ) .
l. boeri , j. kortus , and o. k. anderson , phys .
* 93 * , 237002 ( 2004 ) .
k .- w . lee and w. e. pickett , phys .
lett . * 93 * , 237003 ( 2004 ) .
x. blase , ch .
adessi , and d. conntable , phys .
lett . * 93 * , 237004 ( 2004 ) .
h. j. xiang , z. li , j. yang , j. g. hou , and q. zhu , phys .
b * 70 * , 212504 ( 2004 ) . y. ma , j. s. tse , t. cui , d. d. klug , l. zhang , y. xie , y. niu , and g. zou , phys .
b * 72 * , 014306 ( 2005 )
. p. b. allen , phys .
b * 6 * , 2577 ( 1972 ) ; p. b. allen and m. l. cohen , phys . rev
* 29 * , 1593 ( 1972 ) .
a numerical correction is given in eq .
( 4.27 ) of p. b. allen , in _ dynamical properties of solids _ , ch .
2 , edited by g. k. horton and a. a. maradudin ( north - holland , amsterdam , 1980 ) .
w. e. pickett , j. m. an , h. rosner , and s. y. savrasov , physica c * 387 * , 117 ( 2003 ) .
j. m. an , s. y. savrasov , h. rosner , and w. e. pickett , phys . rev .
b * 66 * , 220502 ( 2002 ) .
i. i. mazin , private communication .
k. shimizu , h. kimura , d. takao , and k. amaya , nature * 419 * , 597 ( 2002 ) . v.
v. struzhkin , m. i. eremets , w. gan , h .- k .
mao , and r. j. hemley , science * 298 * , 1213 ( 2002 ) . s. deemyad and j. s. schilling , phys .
lett . * 91 * , 167001 ( 2003 ) .
d. kasinathan , j. kune , a. lazicki , h. rosner , c. s. yoo , r. t. scalettar , and w. e. pickett , phys .
lett . * 96 * , 047004 ( 2006 ) .
g. profeta , c. franchini , n. n. lathiotakis , a. floris , a. sanna , m. a. l. marques , m. lueders , s. massidda , e. k. u. gross , and a. continenza , phys .
* 96 * , 047003 ( 2006 ) .
w. e. pickett , braz .
* 33 * , 695 ( 2003 ) .
l. boeri , g.b .
bachelet , e. cappelluti and l. pietronero , phys .
b * 65 * , 214501 ( 2002 ) .
l. boeri , e. cappelluti and l. pietronero , phys .
b * 71 * , 012501 ( 2005 ) . | the vision of `` room temperature superconductivity '' has appeared intermittently but prominently in the literature since 1964 , when w. a. little and v. l. ginzburg began working on the _ problem of high temperature superconductivity _ around the same time . since that time the prospects for room temperature superconductivity have varied from gloom ( around 1980 ) to glee ( the years immediately after the discovery of hts ) , to wait - and - see ( the current feeling ) .
recent discoveries have clarified old issues , making it possible to construct the blueprint for a viable room temperature superconductor . |
let @xmath0 be a strictly convex domain , whose boundary @xmath1 is @xmath2 ( for some @xmath3 ) and has non - vanishing curvature everywhere .
there is a induced dynamical system on @xmath4 , the _ dynamical billiard _ , in which a particle moves freely along strict segments in @xmath4 , and changes its velocity according to the law of elastic reflection upon collisions with the boundary @xmath1 .
let @xmath5 be the phase space of the billiard system , which consists of all post - collision statuses with the boundary @xmath1 , and @xmath6 be induced _ billiard map _
, which is a @xmath7 diffeomorphism on @xmath5 .
since the motion is free inside @xmath4 , the dynamics of the billiard systems are determined completely by the shapes of the table @xmath4 .
for example , the system on an elliptic table is completely integrable ( see @xcite ) . in this paper
we consider a generic convex billiard table , and prove the existence of homoclinic orbits on all branches of all hyperbolic periodic points .
there are many results on the existence of homoclinic or heteroclinic orbits .
poincar first appreciated the significance of transversal homoclinic point ( developed later by birkhoff and smale ) .
robinson proved in @xcite that for a @xmath2 map on two - sphere , if the unstable manifold of a hyperbolic fixed point accumulates on its stable manifold , then a @xmath2 small perturbation can create a homoclinic point .
pixton @xcite extended robinson s result to periodic orbits , and proved the generic existence of homoclinic orbit on @xmath8 in the area - preserving category .
using some topological argument , oliveira showed in @xcite the generic existence of homoclinic orbit on @xmath9 , and extended in @xcite to any compact surface ( for the area - preserving systems with irreducible action on the first homology group @xmath10 ) . recently in @xcite
, xia gave a proof of the the generic existence of homoclinic orbit among the class of systems homotopic to identity , and among the class of hamiltonian systems ( also on general compact surface ) .
creation of homoclinic orbits in billiard system turns out to be completely different from the above approaches , since we ca nt perturb the billiard map directly , but deform the underlining billiard table .
such types of perturbations are _ not the local ones _ , and have some unavoidable side effects .
using a variational approach , bangert showed in @xcite that for _ all _ monotone twist maps , for each rotation type @xmath11 , the set @xmath12 of global minimizers are totally ordered , and there exist heteroclinic orbits for all _ neighboring pairs _ of periodic points ( and homoclinic orbits if there is only one periodic point ) . since the billiard map on convex table is a monotone twist map , there always exist some heteroclinic orbits .
however , most periodic orbits are not global minimizers in general , and there may not be any heteroclinic orbits relating them .
it is proved in @xcite that for generic billiard table @xmath4 , if @xmath13 , then there will be a transversal intersection .
see also @xcite .
a priori , the two manifolds may have no intersection at all .
what we show here is that , non - intersection is not the generic case if @xmath14 , and every hyperbolic periodic points does admit some homoclinic orbits .
let @xmath15 . for a @xmath2-generic convex billiard system @xmath4
, we have that 1 .
all periodic points are either hyperbolic or elliptic @xcite ; 2 . for every hyperbolic point @xmath16
, there is a transversal homoclinic point @xmath17 .
note that the properties we listed are open for all finite periods , and @xmath18-properties in general .
so it suffices to prove the theorem in the @xmath19-category .
then @xmath2-genericity follows directly .
it is worth pointing out that the existence of heteroclinic orbits is not generic , since two hyperbolic orbits may be separated by some invariant curves .
specifically , let s consider the elliptic table @xmath20 , on which the billiard system is completely integrable , and the periodic two orbit along the minor axis is parabolic .
a @xmath19-small deformation of @xmath4 can turn this orbit into a hyperbolic periodic orbit .
but there is no heteroclinic orbit connecting the periodic orbits on major and minor axes , since some of the invariant curves survive and separate the two orbits .
let s consider a general diffeomorphism @xmath21 on a closed surface @xmath5 , @xmath16 and @xmath22 be two hyperbolic periodic points of @xmath23 . recall that @xmath23 is said to have a _
homoclinic loop _ at @xmath16 , if one unstable branch of @xmath16 coincides with a stable branch of @xmath16 .
similar , we say that @xmath23 has a _ heteroclinic path _ from @xmath16 to @xmath22 , if one unstable branch of @xmath16 coincides with a stable branch of @xmath22 . in general we may say that the common branch forms a _ path of saddle connection _ of @xmath23 .
a main step proving above theorem is let @xmath23 be an area - preserving diffeomorphism on a closed surface @xmath5 , such that each periodic point is either hyperbolic , or elliptic and moser stable .
then for any branch @xmath24 of the invariant manifolds of any hyperbolic periodic point @xmath16 , one of the following alternative holds : 1 .
@xmath24 contains some recurrent point .
@xmath24 forms a path of saddle connections .
the existence of a path of saddle connection is highly non - generic .
but the point of above proposition is not about generic systems .
it is for all systems satisfying above condition . in particular , for a specific dynamical system , if one can pre - exclude any path of saddle connections , then all stable and unstable branches of _ the given system _ must be recurrent ( no further perturbation required )
. then homoclinic intersection follows if the phase space @xmath5 has simple topology .
in this section we describe the topology on the collection of convex billiard tables , list some generic properties of the billiard maps on these tables , and then introduce the main tool for our study , prime - end extensions .
let @xmath0 be a strictly convex domains , whose boundary @xmath1 is a @xmath19 curve with nonvanishing curvature .
the support function @xmath25 of @xmath4 is given by @xmath26 .
the collection of all such support functions , say @xmath27 , form an open subset of @xmath28 ( with the induced topology on @xmath27 ) .
we denote by @xmath29 the corresponding collection by rigid motions on @xmath30 and by homotheties of @xmath4 , both of which do not change the dynamics at all .
the resulting space is also a baire space ( see ( * ? ? ?
3 ) ) , and shares the same generic properties . ] of convex billiard tables . a property ( p )
is said to be true on a @xmath19-generic convex billiard system , if the set of the support functions of the tables satisfying ( p ) contains a @xmath19-residual subset of @xmath27 .
let @xmath31 be the phase space of the billiard map on @xmath4 , where @xmath1 is parametrized by the arc - length parameter @xmath32 , @xmath33 and @xmath34 measures the angle of the outgoing velocity with the inner normal direction at @xmath35 .
the advantage of choosing this coordinate system is that the billiard map always preserves the lebesgue measure @xmath36 .
it is well known that for convex billiard maps , there is no fixed point in the open annulus @xmath37 , and period-@xmath38 orbits can only lie in the line @xmath39 .
let @xmath5 be a compact surface , @xmath21 be an area - preserving diffeomorphism on @xmath5 and @xmath16 be a hyperbolic periodic point of @xmath23 with period @xmath40 , @xmath24 be a branch of the unstable manifold of @xmath16 .
pick a point @xmath41 .
then @xmath42 $ ] forms a fundamental domain of @xmath24 , and its omega - limit set under @xmath43 is denoted by @xmath44 .
clearly it is independent the choice of @xmath41 and remains the same if we use @xmath45 .
so we may denote it by @xmath46 for short .
similarly we define the alpha - limit set @xmath47 of a stable branch @xmath24 of @xmath16 with respect to @xmath48 .
sometime it is more convenient to use a unified notation .
that is , we may use @xmath46 to denote the _ alpha - limit set _ of @xmath24 , if @xmath24 is a stable branch .
let @xmath49 be two branches of some hyperbolic periodic points of @xmath21
. then * either @xmath50 , or @xmath51 .
* either @xmath52 , or @xmath53
. we will provide a slightly more general statement in lemma [ branch ] with a much shorter proof .
let @xmath54 and @xmath55 be the set of convex tables in @xmath29 such that for every @xmath56 , 1 .
@xmath57 is finite , and all of them are either hyperbolic or elliptic ; 2 . given two hyperbolic periodic points @xmath58 , @xmath59 and @xmath60 either do nt intersect , or have some transversal intersections ; 3 .
every elliptic periodic point in @xmath61 is moser stable .
recall that an elliptic fixed point @xmath16 is said to be _
moser stable _ , if there is a fundamental system of neighborhoods which are invariant closed disks @xmath62 surrounding the point , such that @xmath63 is a minimal rotation .
so moser stable periodic points are isolated from the dynamics and are unreachable from the outside invariant rays .
the above property ( b ) is chosen carefully when defining @xmath64 , since there might be some tangent intersections of two invariant manifolds besides the transversal ones ( even in the generic case , as pointed out in @xcite ) .
this is different from the classical _ kupka - smale condition _ of generic diffeomorphisms , which requires that all intersections of two invariant manifolds are transversal .
this might be a technique difficulty , since only special perturbations can be performed : deformations of the underlining table . [
finite ] for every @xmath65 , the set @xmath55 is @xmath66-open and @xmath19-dense in @xmath29 .
this proposition is the combination of propositions 69 in @xcite and theorems 34 in @xcite .
in fact they made several interesting observations : 1 . to perturb a degenerate periodic orbit ,
a finite induction process is employed to find a local deformation , _ the normal perturbations _ of @xmath4 , that changes only the curvature at some base point of the periodic orbit , while keeping the orbit unchanged .
every @xmath67-periodic orbit @xmath68 of @xmath69 gives an inscribed @xmath67-polygon in @xmath4 .
so at least one of the @xmath70 lies in the compact annulus @xmath71 ( recall @xmath33 in our notation ) .
therefore the finiteness follows from non - degeneracy of periodic orbits .
3 . to make an elliptic periodic orbit moser stable
, it suffices to make a @xmath19-small normal deformation maintaining a third order contact , such that the first birkhoff coefficient is nonzero under the birkhoff normal form .
then moser twist mapping theorem ensures that the orbit is moser stable .
see also @xcite for earlier versions of such deformations .
another classical result we may use is .2 cm * lazutkin theorem . * @xcite _ for @xmath72 large enough , every strictly convex domain @xmath4 with @xmath2 boundary and nowhere vanishing curvature admits a family of invariant curves converging to the boundary @xmath73 .
moreover , these invariant curves have irrational rotation numbers and occupy a positive volume in a neighborhood of @xmath74 .
_ originally this theorem asked for @xmath75 to apply kam theory ; later this was reduced to @xmath76 , see @xcite .
clearly the existence of invariant curves prevents the billiard system to be ergodic .
so all convex billiards are not ergodic if the boundary is @xmath77 and with nowhere vanishing curvature . on the other hand
, it is showed that no such invariant curve can survive if one of the restrictions is relaxed , see mather @xcite and hubacher @xcite .
currently , most known ergodic convex billiard tables have long flat components to apply defocusing mechanism ( see @xcite and the references therein ) .
let @xmath78 be the open unit disk , @xmath79 be a bounded , simply connected domain on the plane .
then there exists a conformal map @xmath80 ( see @xcite ) .
every point @xmath81 is specified by a nested sequence of open arcs @xmath82 with @xmath83 such that the endpoints of @xmath84 lie on each side of @xmath85 , @xmath84 are nested and converge to @xmath86 ( see ( * ? ? ?
* theorem 17.11 ) ) .
the prime - end compactification @xmath87 gives one way to compactify @xmath79 , whose topology is uniquely determined by the extended homeomorphism @xmath88 , such that @xmath89 and @xmath90 ( see ( * ? ? ?
* theorem 17.12 ) ) . here
we point out that : 1 .
a point @xmath85 may cover many points in the boundary @xmath91 , which is compact and connected , and called the _ impression _ of @xmath86 ( * ? ? ?
* theorem 17.7 ) .
a point @xmath92 may be lifted to many points in the prime - end @xmath93 , because we may observe it several times when traveling inside @xmath79 ( ( * ? ? ?
* figure 37-(b ) ) ) .
see also @xcite for various examples .
let @xmath94 be a homeomorphism .
then there exists uniquely an extension @xmath95 .
if @xmath23 is orientation - preserving , then the restriction @xmath96 is an orientation - preserving circle homeomorphism .
therefore the rotation number @xmath97 of @xmath96 is well defined , which is called the _ carathodory rotation number _ associated to @xmath98 , see @xcite .
it is well known that @xmath96 has periodic orbits if and only if it has a rational rotation number .
however , we note that 1 . a fixed point of @xmath99 on @xmath93 does nt imply the existence of a fixed point of @xmath23 on @xmath91 .
for example , let @xmath79 be the unstable set of the new formed source of a da - map on @xmath9 , which is an open set bounded by two ( split ) stable manifolds
. then we can construct @xmath23 on @xmath79 which are translations along the boundary .
so no point on @xmath91 is fixed by @xmath23 .
but the two added points in the prime - end are fixed by @xmath99 .
2 . a fixed point of @xmath23 on @xmath91
does nt imply the existence of a fixed point of @xmath99 on @xmath93 .
for example , we can put an aster set @xmath100 on the north pole @xmath101 , and let @xmath23 rotate @xmath8 along the axis by @xmath102 . although @xmath101 is fixed , the prime - end extension on @xmath103 has rotation number @xmath104 on the added boundary .
prime - end extensions can also be defined for connected closed surface @xmath5 ( possibly with boundary ) .
let @xmath105 be an open connected subset , whose boundary consists of a finite number of connected pieces , each of which has more than one point .
then we can add to @xmath79 a finite number of circles , to get its prime - end compactification @xmath106 .
see also @xcite .
[ downfixed ]
let @xmath21 be an area - preserving diffeomorphism , @xmath79 be an open connected subset and @xmath95 be a prime - end extension . 1 .
if there exists a fixed point of @xmath99 on the added boundary of @xmath106 , then there exists a fixed point of @xmath23 on @xmath91 .
2 . moreover assume that each fixed point of @xmath23 is either hyperbolic , or elliptic and moser stable .
then the fixed point is hyperbolic .
let @xmath107 be a collection of open arcs with @xmath108 and both endpoints on @xmath91 , that define the point @xmath109 .
note that @xmath110 also defines @xmath111 .
let @xmath112 be an accumulating point of @xmath113 .
since @xmath23 is area - preserving , we see that the fundamental domains separated by @xmath113 and @xmath114 have the same area and intersect each other .
so @xmath115 .
therefore , @xmath16 must be fixed by @xmath23 ( since @xmath108 ) .
if @xmath16 is elliptic , then it is also moser stable ( by our assumption ) .
so there are a fundamental system of closed disks @xmath62 surrounding @xmath16 such that @xmath116 is minimal . for any such a disk @xmath62 , if @xmath117 , then @xmath118 ( by the invariance of @xmath79 ) .
so either @xmath119 for some @xmath62 , which implies that @xmath113 are stopped outside @xmath62 and ca nt accumulate @xmath16 ( a contradiction ) ; or @xmath118 for all such @xmath62 s .
since @xmath91 is assumed to have only finite number of connected pieces , there exists some disk @xmath62 with @xmath120 , contradicting the choices of @xmath113
. so @xmath16 must be hyperbolic .
in this section we will study the behavior of the branches of hyperbolic periodic points of a surface diffeomorphism .
more precisely , let @xmath5 be a closed connected surface of genus @xmath121 , @xmath122 be an area - preserving diffeomorphism on @xmath5 .
[ branch ] let @xmath123 be a compact , connected , @xmath23-invariant subset , @xmath124 be a branch of the invariant manifolds of a hyperbolic fixed point @xmath16 .
then either @xmath125 , or @xmath126 . this was proved by mather in @xcite using the ideal boundary closure .
a different proof was given in @xcite for the case @xmath127 .
for completeness we give a proof without using the ideal boundary points . without loss of generality
we assume @xmath124 is a stable branch fixed by @xmath23 .
suppose on the contrary that @xmath124 intersects both @xmath128 and @xmath129 .
let @xmath130 be an open arc in @xmath124 with both endpoints on @xmath128 .
clearly @xmath16 ca nt be the endpoint of @xmath130 ( otherwise @xmath131 since @xmath79 is invariant ) .
let @xmath121 be the genus of @xmath5 , and consider the iterates @xmath132 , @xmath133 , all of which have their endpoints on @xmath128 .
together with @xmath128 , they will separate @xmath5 into ( at least two ) different connected components .
pick any one of components , say @xmath134 , that touches @xmath130 from one side .
fix a point @xmath135 and pick a small @xmath136 such that @xmath137 .
let @xmath138 ( close to a half disk if @xmath139 is small enough ) .
let @xmath140 , and @xmath141 be the @xmath142-interior of @xmath134 , which may consist of several connected components .
pick a point @xmath143 , and let @xmath144 be the connected component of @xmath145 containing @xmath146 .
it is easy to see that @xmath147 is closed in @xmath134 , hence coincides with @xmath134 ( since @xmath134 is connected ) .
pick @xmath142 small enough such that @xmath148 .
in particular , it forces @xmath149 .
pick @xmath101 large enough such that @xmath150 for all @xmath151 . by poincar recurrence theorem
, there exists @xmath152 such that @xmath153 .
note that @xmath154 , @xmath155 are the only pieces of the boundary of @xmath156 that can enter @xmath134 . with both endpoints hanging on @xmath128 ,
all of them are too short ( less than @xmath142 ) to invade the ball @xmath157 ( since @xmath137 ) .
so @xmath158 .
similarly , we have @xmath159 .
this implies that @xmath160 , which contradicts the area - preserving assumption .
this completes the proof . by hartman s theorem , for any hyperbolic fixed point @xmath16 of @xmath23 ,
there exist a @xmath161 local chart @xmath162 around @xmath16 and some @xmath163 , such that the @xmath164-axis gives the local unstable manifold , the @xmath146-axis gives the local stable manifold , and @xmath165 .
next lemma can be proved following the ideas of the proof of ( * ? ? ?
* lemma 6.3 ) ( see also @xcite ) .
suppose that each fixed point of @xmath23 is either hyperbolic , or elliptic and moser stable .
let @xmath123 be a compact , connected , @xmath23-invariant subset , @xmath79 be an invariant , connected component of @xmath166 .
[ open ] if there is some fixed point of @xmath99 on the added boundary of @xmath106 , then there exists a hyperbolic fixed point @xmath16 on @xmath91 and a small open neighborhood @xmath162 of @xmath16 , such that @xmath167 consists of one or two branches of the local invariant manifold of @xmath16 .
let @xmath109 be a fixed point of @xmath99 .
then by lemma [ downfixed ] , there exists a hyperbolic fixed point @xmath16 on @xmath91 .
let @xmath162 be a small neighborhood of @xmath16 given by hartman s theorem .
there are two cases and we treat them separately : * case 1 . *
@xmath168 for some local branch @xmath124 of @xmath16 in @xmath162 . in this case
we have @xmath131 ( by lemma [ branch ] ) , since @xmath79 is a connected component of @xmath166 .
moreover , the two open quadrants on both sides of @xmath124 are contained in @xmath79 ( by the invariance of @xmath79 ) .
now we consider the new - reached branches , say @xmath169 . either @xmath169 does nt intersect @xmath79 ( then it serves as a boundary piece of @xmath79 ) , or we can continue the above argument such that @xmath79 contain a new quadrant adjacent to @xmath169 , and push the boundary to the next local branch .
inductively , we see that either @xmath79 stops at some branch , which serves as the boundary piece of @xmath79 ( so our conclusion holds in this case ) , or we keep going such that all four branches intersect @xmath79 . in the later case , @xmath79 contains the whole open neighborhood of @xmath16 , a contradiction .
* case 2 . *
@xmath170 for all local branches @xmath124 .
let @xmath107 be a collection of open arcs with @xmath108 that define the point @xmath109 and locate the point @xmath16 . passing to a subsequence and relabeling the quadrants , we assume @xmath113 lies in the first quadrant for all @xmath40
. then consider the curves @xmath171 , and the rectangles @xmath172 in the first quadrant bounded by @xmath173 .
pick @xmath174 small enough such that @xmath172 is disjoint from @xmath175 .
still @xmath172 contains @xmath113 for some large enough @xmath176 .
all @xmath113 are chosen from a connected open set @xmath79 .
so @xmath177 , and some component of this intersection separates @xmath175 and
@xmath113 in @xmath79 , say @xmath178 , which is an open arc in @xmath79 , with both endpoints on @xmath128 .
let @xmath179 and @xmath180 be the two components of @xmath79 separated by @xmath178 . since @xmath181 for all @xmath65 , we see that @xmath182 for both @xmath183 ( since they contain @xmath175 and @xmath184 , respectively ) .
being the only boundary piece of @xmath185 in @xmath79 , @xmath186 ca nt lie on one side of @xmath178 ( by the area - preserving property of @xmath23 ) .
so @xmath186 must meet @xmath178 . by the choice of hartman
coordinate system around @xmath16 , we see that @xmath187 consists of a single point @xmath188 , which must lies on @xmath189 . then it is easy to see its pre - image @xmath190 also lies in @xmath178 .
according to the definition of @xmath178 , we see the whole piece @xmath191 is contained in @xmath178 , which is also contained in @xmath79 .
moreover , this holds for all smaller @xmath142 .
so @xmath79 contains a cone in the first quadrant , and then the whole first quadrant by the invariance of @xmath79 . in this case , the boundary consists also the local branches .
this finishes the proof .
[ recurrent ] suppose that each periodic point of @xmath23 is either hyperbolic , or elliptic and moser stable .
then for any branch @xmath24 of the invariant manifolds of any hyperbolic periodic point @xmath16 , we have 1 .
either @xmath24 contains some recurrent point .
2 . or @xmath24 forms a saddle connection . without loss of generality , we assume @xmath16 is fixed and all four quadrants are also fixed by @xmath23 , @xmath24 is an unstable branch , say the positive @xmath164-axis , which is also fixed by @xmath23 .
suppose that @xmath24 contains no recurrent point
. then @xmath24 ca nt approach @xmath16 from the first and fourth quadrants , and @xmath192 .
therefore , @xmath166 contains the first and fourth quadrants of a smaller neighborhood @xmath162 , which will grow to two open strips attaching to @xmath24 ( by the invariance of @xmath166 ) .
let @xmath79 be the connected component of @xmath166 containing the first quadrant .
it is easy to see that @xmath79 is invariant ( being a component ) .
let @xmath193 be the prime - end compactification of @xmath79 , @xmath194 be the prime - end extension and @xmath195 be the induced circle homeomorphism . note that each point @xmath41 is a regular point of the boundary @xmath91 , and gives rise to at least one point in @xmath106 , say @xmath196 ( when observing @xmath24 from the first quadrant )
. denote by @xmath109 the one - side limit of @xmath86 in @xmath93 as @xmath164 moves back to @xmath16 along @xmath24 .
clearly @xmath109 is a fixed point of @xmath197 .
so the rotation number of @xmath197 is zero , the omega - set @xmath198 ( @xmath41 ) consists of a single fixed point of @xmath197 , say @xmath199 .
note that if @xmath24 never comes back to @xmath16 , then @xmath79 also contains the fourth quadrant and we may get two points when observing @xmath24 from the the first and fourth quadrants respectively , say @xmath200 and @xmath201 .
let s take @xmath202 in that case for certainty .
then by lemma [ downfixed ] , there exists a hyperbolic fixed point @xmath203 .
* @xmath14 . according to lemma [ open ] ,
there exists a small neighborhood @xmath162 of @xmath16 , such that @xmath167 consists only of the branch(es ) of local invariant manifolds .
each piece is regular on @xmath91 , and hence contained in @xmath24 .
clearly @xmath24 can not intersect the unstable branches .
so @xmath24 must coincide with one branch of the stable manifold of @xmath16 , which forms a homoclinic loop .
* case 2 . * @xmath204 .
applying lemma [ open ] again , we see that there exists a small neighborhood @xmath134 of @xmath22 , such that @xmath205 consists only of the branch(es ) of the local invariant manifolds of @xmath22 .
each piece is regular on @xmath91 , and hence contained in @xmath24 .
once again it is a stable branch of @xmath22 and forms a path of heteroclinic connections from @xmath16 to @xmath22 .
therefore @xmath24 always serve as a saddle connection under the assumption that @xmath24 contains no recurrent point .
this finishes the proof .
[ closure ] suppose that each periodic point of @xmath23 is either hyperbolic , or elliptic and moser stable .
if @xmath23 does nt have any path of saddle connection , then all four branches of any hyperbolic periodic point are recurrent and have the same closure .
let @xmath24 , @xmath206 , @xmath207 and @xmath208 be the four branches of a hyperbolic periodic point @xmath16 ( ordered counterclockwise ) .
it suffices to show that the closure of any given branch contains the branch next to it .
then by induction we see that the closure contains all four branches and serves as the common closure . without loss of generality , we assume @xmath24 is an unstable branch and show @xmath209 . if this is not the case , then @xmath210 ( by lemma [ branch ] ) .
let @xmath79 be a connected component of @xmath211 containing @xmath206 .
clearly @xmath16 is an accessible boundary point from @xmath79 ( along @xmath206 ) .
then by ( the proof of ) lemma [ open ] , there exists an open neighborhood @xmath212 such that @xmath213 consists of one or two local branches
. clearly it ca nt be the single branch @xmath24 ( by the recurrence from lemma [ recurrent ] ) .
it ca nt contain @xmath206 ( since @xmath210 ) , or @xmath207 ( both @xmath24 and @xmath207 are unstable branches ) .
so it has to contain @xmath208 , which leads to a homoclinic loop @xmath214 and contradicts the assumption that there is no path of saddle connection .
this completes the proof .
let @xmath55 be the set of convex billiard tables given in
[ un ] , and @xmath215
. then by lemma [ finite ] , we see that @xmath216 contains a @xmath19-residual subset of @xmath29 , such that for every billiard table @xmath217 , for the billiard map @xmath69 on the phase space @xmath31 , 1 .
each periodic point of @xmath69 is either hyperbolic or elliptic ; 2 . given two hyperbolic periodic points @xmath218 , @xmath59 and @xmath60 either do nt intersect , or have some transversal intersections ; 3 .
every elliptic periodic point is moser stable . in the following ,
let @xmath16 be a hyperbolic periodic point of @xmath69 , @xmath24 be a branch of the unstable manifold of @xmath16 , @xmath219 , which is compact , connected , @xmath23-invariant , where @xmath220 ( or @xmath221 if necessary ) . according to proposition [ finite ]
, we can assume @xmath222 without loss of generality .
according to lazutkin theorem , there always exist invariant circles accumulating to both boundaries @xmath223 of @xmath5 .
so what happens near the boundary does not affect dynamics in the middle .
we can collapse @xmath74 to two points , say @xmath101 and @xmath224 , and the result surface is homeomorphic to @xmath8 .
clearly both @xmath101 and @xmath224 are surrounded by invariant closed disks enclosed by lazutkin invariant circles and have moser - stable behaviors .
the following is a direct application of proposition [ recurrent ] , since the paths of saddle connections are excluded from systems in @xmath216 .
[ billrecu ] let @xmath225 and @xmath16 be a hyperbolic periodic point .
then the two unstable branches are recurrent under @xmath23 , and two stable branches are recurrent under @xmath226 .
[ accumu ] there always exists an adjacent pair @xmath227 of stable and unstable branches such that they accumulate on each other from the quadrant between them .
let s start with the branch @xmath24 of unstable manifold from the positive @xmath164-axis .
let @xmath22 be a recurrent point given by corollary [ billrecu ] .
so some iterates @xmath228 accumulates on @xmath22 itself , from either the first quadrant @xmath229 or from the fourth quadrant @xmath230 ( possibly both ) .
we may assume that it is the former case , the orbit of q accumulates from @xmath229 .
then it is easy to see the branch @xmath128 of the stable manifold from the positive @xmath146-axis is contained in @xmath46 .
there is a backward recurrent point @xmath22 on @xmath128 , whose backward orbit accumulates on itself either from @xmath229 or the second quadrant @xmath231 . in the former case , we have an adjacent pair of stable and unstable branches @xmath232 that accumulate on each other .
now let s assume the backward orbit of @xmath22 accumulates on itself from @xmath231 , which implies @xmath233 .
now let s consider @xmath207 , which can accumulate to itself from either @xmath231 , or from the third quadrant @xmath234 .
if it accumulates from @xmath231 , then we have an adjacent pair @xmath235 .
otherwise let s check @xmath236 .
either we have an adjacent pair @xmath237 , or @xmath236 accumulates on @xmath24 though @xmath230 and the accumulating relations of four branches form a closed chain .
however , the accumulating relations are transitive . since @xmath236 accumulates to itself though @xmath230
, @xmath24 also accumulates on @xmath236 though @xmath230 .
this still gives rise to an adjacent accumulating pair @xmath238 .
let @xmath225 and @xmath16 be a hyperbolic periodic point with an adjacent accumulating pair .
then there exists a homoclinic orbit of @xmath16 .
let @xmath232 be the adjacent accumulating pair given by proposition [ accumu ] .
without loss of generality , we assume @xmath24 is the branch of the unstable manifold of @xmath16 along the positive @xmath164-axis , @xmath128 is the branch of the stable manifold of @xmath16 along the positive @xmath146-axis , and they accumulate each other from the first quadrant .
let @xmath162 be a small @xmath161 chart around @xmath16 given by hartman s theorem , @xmath139 be a small positive number , @xmath239 be a large integer , and @xmath240 be a small region in the first quadrant of @xmath162 .
let @xmath241 be the position on @xmath24 that hits @xmath242 for first time , and @xmath243 be the position on @xmath128 that hits @xmath242 for first time .
adjacent.pdf ( 13,8)@xmath16 ( 33,7)@xmath24 ( 3,35)@xmath128 ( 20,25)@xmath242 ( 13,60)@xmath241 ( 65,21)@xmath243 according our choice of coordinate system , @xmath241 lies on the part of boundary @xmath244 above the round spot , which is given by @xmath245 similarly , @xmath243 lies on the part of boundary @xmath244 below the square spot , which is given by @xmath246 we already have a homoclinic intersection if either @xmath247 or @xmath248 . in the following we assume this is not the case ( see figure [ adjacent ] ) .
let s consider two simple closed curves @xmath249\ast [ q_u , p]$ ] and @xmath250\ast [ q_s , p]$ ] .
note that on @xmath8 , the algebraic sum of the intersection numbers of two closed curves must be zero .
clearly @xmath251 and @xmath252 cross each other once at @xmath16 .
so @xmath251 and @xmath252 have to cross each other elsewhere , say @xmath22 .
note that @xmath241 and @xmath243 are separated by the two shaded spots ( since @xmath239 ) .
so @xmath22 lies outside @xmath242 , which forces @xmath253 .
that is , @xmath22 is a homoclinic point of @xmath16 .
[ 4branch ] let @xmath225 and @xmath16 be a hyperbolic periodic point .
then there exist some transversal homoclinic orbits on all branches of @xmath16 . as in the previous proof ,
let @xmath22 be a homoclinic intersection of some pair @xmath232 of the branches .
without loss of generality , we assume they intersect transversally at @xmath22 ( by the choice of the table @xmath225 ) .
then @xmath254\ast k[q , p]$ ] is a simple closed curve .
let @xmath62 be the domain enclosed by it .
it is easy to see the other branches , @xmath207 and @xmath236 , have to enter @xmath62 ( by corollary [ closure ] ) , since some parts of @xmath128 and @xmath24 are contained in @xmath62 .
so @xmath207 has to cross @xmath128 , and @xmath236 has to cross @xmath24 .
the crossing points provide the homoclinic points on these branches . again
our choice of @xmath216 implies the existence of transversal homoclinic orbits .
this finishes the proof .
this research is supported in part by national science foundation . | in this paper we investigate some generic properties of a billiard system on a convex table .
we show that generically , every hyperbolic periodic point admits some homoclinic orbit . |
@xmath1 emission is the main tracer of the molecular gas in our galaxy as well as in external galaxies ( see _ e.g. _ * ? ? ?
* ; * ? ? ?
yet , this co emission is ambiguous .
indeed , this co emission is usually associated to cold ( 10 - 20@xmath2 ) , dense ( @xmath3 ) , strongly uv - shielded , molecular gas ( carbon atoms are mostly locked in co ) . however , it became more and more obvious over the last decade @xcite that a large fraction of the co emission in our galaxy comes from warm ( 50 - 100@xmath2 ) , low density ( 100 - 500@xmath4 ) , weakly uv - shielded , diffuse gas ( carbon atoms are mostly locked in @xmath5 ) .
moreover , at the beginning of the history of the @xmath0 factor , @xcite noted that diffuse and dense gas have similar factor .
for instance , @xmath6 , @xmath7 toward @xmath8oph , a prototypical diffuse line of sight , and @xmath9 , @xmath10 toward ori a , a prototypical dense gas line of sight .
this triggers the question of the mean value of the @xmath0 factor in diffuse gas ( * ? ? ?
* for details see ) .
optical depth from @xcite and this work versus the total line of sight reddening from @xcite . * middle , bottom : * total hydrogen column density versus integrated @xmath11 optical depth for the sources studied by @xcite . *
right : * co luminosity versus the total line of sight reddening from @xcite . ]
we quantify the @xmath0 conversion factor in diffuse gas from a sample acquired over the last 20 years .
it is made from the study of whole galactic lines of sight measured in absorption against extragalactic continuum background sources .
there are two different groups of lines : either they have a low visual extinction @xmath13 when they are observed at high galactic latitude @xmath14 or when they lie in the galactic plane , their large total visual extinction @xmath15 can be divided in well separated velocity components , each one having @xmath16 ( see the left column of fig . [
fig : pety:1 ] for an example ) .
hence , all the gas studied here is diffuse .
indeed , the co column density per velocity component on any line of sight is low , typically @xmath17 , which implies that less than 7% of the carbon is locked in co. the computation is then made in three steps .
first , the total hydrogen column density is deduced from the @xmath18 reddening maps of @xcite , using the standard relation @xcite @xmath19 second , we estimate the atomic gas fraction via the @xmath11 absorption measurements by writing it as the product of two terms @xmath20 where the left term is directly measured from the data ( middle , top panel of fig . [
fig : pety:1 ] ) while the right term is calibrated from the careful measurements of @xcite ( middle , bottom panel of fig .
[ fig : pety:1 ] ) .
this gives @xmath21 or @xmath22 , which implies a 30% overall uncertainty on the method . ] .
the third step consists in the measures of the co emission luminosity , @xmath23 , along the line of sight ( right panel of fig .
[ fig : pety:1 ] ) .
we note here that the high @xmath23 values @xmath24 arise from the accumulation of several low-@xmath25 components along the same galactic line of sight .
moreover , co is not reliably detected at low @xmath18 @xmath26 .
the corresponding lines of sight are not used in our estimation because the @xmath0 factor is used to estimate the molecular column density only from the co gas which is detected . in summary , we obtain @xmath27 , @xmath28 and @xmath29 , which give @xmath30 , the same _ mean _ co luminosity per @xmath31 in diffuse and dense gas !
to understand this result , we write @xmath32 where the @xmath33 ratio comes from the co chemistry while the @xmath34 comes from the radiative transfer through the cloud structure . in diffuse gas , more than 90% of the carbon is locked in @xmath5 , which implies that @xmath35 @xcite . in addition , the gas is subthermally excited .
large velocity gradient radiative transfer methods @xcite thus show that 1 ) @xmath36 is large because of weak co excitation in warm gas ( 60 - 100@xmath2 ) , and 2 ) @xmath37 until the opacity is so large that the transition approaches thermalization .
we thus obtain @xmath38 ( see also * ? ? ?
* ) . in dense gas ,
all the carbon is locked in co , @xmath39 . as a consequence , @xmath12 = cst and @xmath37 .
a constant @xmath0 factor thus implies that @xmath37 .
we interpret this as a bulk effect in a turbulent medium , the medium is macroscopically optically thin because of the large velocity gradients due to turbulence . in summary ,
the change of chemistry from diffuse to dense gas is compensated by the inverse change of the radiative transfer , giving the same @xmath0 factor .
as @xmath1 alone can not be used to discriminate between diffuse and dense gas , other tracers as molecules with higher dipole moments ( _ e.g. _ @xmath40 , cs , hcn ) are tried .
however , they often are difficult to detect .
we thus propose to use the co isotopologues , the ratio of the @xmath1 over the @xmath41 emission . indeed , @xmath42 in diffuse , warm gas because of the @xmath5 fractionation ( _ e.g. _ * ? ? ?
* ) , while @xmath43 in dense , cold gas ( _ e.g. _ * ? ? ?
we here aim at estimating the co luminosity of the diffuse molecular gas perpendicular to the galactic plane from our absorption data for which the mean luminosity is @xmath44 .
we assume that the gas is simply ordered in plane - parallel , stratified layers .
the mean number of galactic half - width along integration path is then @xmath45 , where @xmath46 is the latitude of each line of sight of our sample .
we thus obtain @xmath47 .
the co emission galactic survey of @xcite gives a mean co brightness per kpc of @xmath48 at @xmath49 .
assuming a single gaussian vertical distribution of dispersion 60@xmath50 , we obtain @xmath51 .
this is a lower limit because we see large amount of co emission outside area mapped by the galactic surveys of the co emission ( mainly limited to the galactic plane ) .
hence , a large fraction of the co luminosity measured in galactic surveys could come from diffuse gas .
the coincidence between the values of the @xmath0 factor in diffuse and dense gas implies that the mass estimates computed with the standard value are correct .
however , the underlying physical interpretation of the detected gas is very different .
if the gas is dense : it will fill a small fraction of the interstellar volume ; it will be confined by ram or turbulent pressure ( if not gravitationally bound ) ; and it is on the verge of forming stars . if the gas is diffuse : it is a warmer , low pressure medium filling a large fraction of the interstellar volume ; it contributes more the mid - ir or pah emission ; and it is probably not gravitionally bound or about to form stars .
17 natexlab#1#1 , f. 2008 , astron .
j. , 136 , 2846 , r. c. 1978 , apj , 224 , 132 , e. b. 2007 , apj , 658 , 446 , w. b. & gordon , m. a. 1978 , a&a , 63 , 7 , r. w. & dickey , j. m. 1989 , apj , 338 , 841 , p. & kwan , j. 1974 , apj , 189 , 441 , p. f. 2008 , apj , 680 , 428 , c. & troland , t. h. 2003 , apj , 586 , 1067 , a. k. 2008 , astron . j. , 136 , 2782 , h. s. 1982 , apj , 262 , 198 . 2007 , a&a , 461 , 205 , h. s. & lucas , r. 1998 , a&a , 339 , 561 , h. s. , pety , j. , & lucas , r. 2010 , aa , 518 , a45 , j. , lucas , r. , & liszt , h. s. 2008 , a&a , 489 , 217 , b. l. 2009 , astrophys . j. , suppl .
ser . , 180 , 125 , d. j. , finkbeiner , d. p. , & davis , m. 1998 , apj , 500 , 525 , j. s. & scoville , n. 1982 , apj , 258 , 467 | we show that the @xmath0 factor , which converts the co luminosity into the column density of molecular hydrogen has similar values for dense , fully molecular gas and for diffuse , partially molecular gas .
we discuss the reasons of this coincidence and the consequences for the understanding of the interstellar medium . |
holography , based on ads / cft ( anti - de - sitter space / conformal field theory ) correspondence @xcite , has been extensively used to analyze strongly coupled gauge systems .
for instance , qcd can be reformulated based on the gauge - gravity duality , either in the bottom - up approach @xcite modeled as a five - dimensional gauge theory defined in an ads background , or in the top - down approach @xcite starting with a stringy setting
. such models of holographic qcd have succeeded well in reproducing several important properties for qcd hadrons within a theoretically expected size of uncertainties . in particular , the bottom - up approach can be made to reproduce the qcd in the all energy region , from the high energy behavior through the operator product expansion ( ope ) down to the low energy resonance physics : the correct power behaviors in ope for the chiral condensate and the gluon condensate are realized by introduction of the bulk ( chiral non - singlet ) scalar field , @xmath1 , corresponding to the @xmath2 operator and ( chiral singlet ) bulk gluonic field , @xmath3 , respectively @xcite , which is contrasted to the top - down approach having high energy behavior completely different than that of qcd .
the bottom - up holography so constructed can provide us with novel insights into the strong dynamics through the highly nontrivial effects of the gluonic dynamics : it reproduces nicely the known value of the gluon condensate which is otherwise zero , and the mass of @xmath4 meson of the right magnitude .
such a bottom - up holographic qcd model having gluon condensation effects ( hereafter , we will use the phrase gluonic effects " to refer to the gluon condensation effects ) @xmath3 was further applied @xcite to the walking technicolor ( wtc ) model @xcite which has an approximate scale symmetry and a large anomalous dimension @xmath5 , and further predicts a light composite higgs dubbed `` technidilaton ( td ) '' as a pseudo nambu - goldstone boson of the approximate scale symmetry .
the holographic wtc of ref .
@xcite was formulated through a simple replacement of @xmath6 for qcd by @xmath5 in the mass parameter of the bulk scalar @xmath1 in the holographic qcd with @xmath3 , which was found @xcite to have more intriguing gluonic effects than in those of the holographic qcd : large gluonic effects due to @xmath3 in wtc actually realize an idealized limit where td @xcite as a flavor - singlet scalar fermionic bound state ( lowest kaluza - klein ( kk ) mode of @xmath1 ) has no direct relevance to the electroweak symmetry breaking , since the technigluons carry no electroweak charge , and hence can not be identified with the td , or composite higgs , whose constituents must carry the electroweak charge . ]
has a vanishingly small mass compared with the typical symmetry breaking scale @xmath7 ( @xmath8 : techni - pion decay constant ) , and hence it can naturally be as light as 125 gev to be identified as the higgs boson discovered at the lhc @xcite .
this is in sharp contrast to the earlier nonperturbative studies based on the ladder approximation @xcite without nonperturbative gluonic dynamics , which yields a substantially smaller scalar mass than in the qcd case , though it does not have such an idealized massless limit and hence no natural framework for the light td and the decay constant @xmath9 of the td @xmath10 only in the form of a product , @xmath11 , ( not separately ) .
this implies that a massless td limit @xmath12 is formally realized only when @xmath13 , or the coupling vanishes ( decoupled dilaton ) @xcite . nevertheless , it so happened that the mass estimated by the ladder pcdc actually can be parametrically tuned to be 125 gev in such a way as to be consistent with the current lhc data , being still far from the decoupling limit @xcite . ] .
therefore , proper inclusion of gluonic effects is important not only for the study of qcd , but also , or more significantly , for the realistic wtc calculations .
holographic computations are generically done through evaluating green functions constructed from ( techni-)quark and gluon currents , so that the main outputs are made from the current correlators , including full information on masses and couplings for the associated mesons and glueball coupled to those currents .
once the vector current correlator is obtained from the holographic calculation , it is possible to translate it into the electromagnetic current correlator , @xmath14 , from which we can estimate the ( techni-)hadronic leading order ( hlo ) contribution to the anomalous magnetic moment of the muon ( the muon @xmath0 ) . actually , in ref .
@xcite a holographic estimation of such a qcd hlo contribution was done based on an earlier bottom - up holographic model @xcite , in which , however , effects from the gluon condensation are not incorporated . in this paper , we study the gluonic effects on the muon @xmath0 in qcd and wtc through the ( techni-)hlo contribution , based on a recently published holographic model in the bottom - up approach @xcite .
the model has 3 holographic parameters ( @xmath15 ) to be explained later ; @xmath16 and @xmath17 , roughly corresponding to the chiral condensate and the gluon condensate , respectively , at the infrared brane located at @xmath18 . by fixing the values @xmath19 mev and @xmath20 mev as inputs , we have @xmath16 and @xmath18 as functions of @xmath17 .
accordingly , we show all other hadronic observables given as functions of a single holographic gluonic parameter @xmath17 : while the mass of the flavor - singlet scalar @xmath21 and the scalar glueball are insensitive to the value of @xmath17 , the mass of @xmath4 meson and the gluon condensate which are fairly sensitive to @xmath17 so that we can determine the value of @xmath22 nicely fitting the reality of all the observables studied @xcite
. we then estimate the hlo contributions including the gluon - condensation effect , based on a formula for the hlo directly evaluated by the current correlator in the ( theoretically more tractable ) _ space - like momentum _ in contrast to the conventional one converting it to the dispersion integral from the time - like contributions where the experimental data are available .
the results show that the proper inclusion of gluon - condensation effect @xmath22 causes about 6% enhancement of the hlo contribution to the muon @xmath0 : @xmath23 , @xmath24 , which are compared to the previous calculation @xcite without gluon - condensation effect @xmath25 : @xmath26 , @xmath27 , respectively .
this enhancement makes the prediction of holographic calculation of the hlo contribution very close to the known value @xcite from the phenomenological estimate using the experimental inputs @xmath28 and @xmath29 , respectively .
we also calculate the hlo contributions to the electron and the tau lepton @xmath0 , and show that the holographic predictions are quite consistent with the known values as well .
considering the fact that the energy scales important for determining the hlo effects are hierarchically different depending on the lepton species , it is remarkable that the holographic calculations , with proper inclusion of gluon - condensation effect , can reproduce hlo contributions to @xmath0 for all the leptons .
this means that with inclusion of the gluonic contributions the holographic calculation of @xmath14 is quite reliable in a _ wide range of energy scale _ , and remarkably in the ( theoretically tractable ) continuous space - like momentum , not just in the range of the ( discrete ) time - like momentum where the resonance parameters are fitted in the conventional holographic studies in the zero - width approximation ( large @xmath30 limit ) .
encouraged by this success , we apply the same holographic calculations of the hlo - type contributions to the @xmath0 from the wtc , with @xmath6 in qcd simply replaced by @xmath5 through the bulk scalar mass .
we first show the gluonic effects on various observables in wtc , including the mass of the flavor - singlet scalar meson @xmath31 identified with td , which can be as light as the 125 gev boson discovered at the lhc for a large gluonic effect with @xmath32 @xcite .
then we study the effect of such a large @xmath17 on the muon @xmath0 through the techni - hlo contribution , which we find as large as 100 times of the value for @xmath25 : imposing a typical constraint on the peskin - takeuchi @xmath33 parameter @xcite from the electroweak ( ew ) precision test , @xmath34 , the enhancement is about 10 times and by relaxing the constraint on @xmath33 parameter as @xmath35 for the reason described in the text , the techni - hlo contribution can be further enhanced by another factor of about 10 . for all such enhancements , however ,
we show that the techni - hlo contributions from wtc dynamics is negligibly small compared to the hlo contributions from qcd .
it is then very unlikely that the contributions from wtc dynamics can explain the inconsistency by about 3.3 @xmath36 between the experimental value of the muon @xmath0 and the standard model ( sm ) prediction of it .
we also mention the techni - hlo contribution to the tau @xmath0 , which may become relevant in future experiments .
the paper is organized as follows : in the next section , we review the model @xcite and formulas which are needed for the calculations of various physical quantities studied in this paper . in section [ sec : qcd ] , we study the gluonic effect of the muon @xmath0 through the holographic qcd calculation of hlo contribution , based on the successful inclusion of the gluonc effects on the various hadonic observables in ref . @xcite . in section
[ sec : wtc ] , we use similar holographic calculations for the study of wtc effects on the @xmath0 .
section [ sec : conclusions ] is devoted to discussions and summary of the paper .
the model @xcite we shall employ is based on deformations of a bottom - up approach for successful holographic dual of qcd @xcite .
the model is described as @xmath37 gauge theory defined on the five - dimensional ads space - time , which is characterized by the metric @xmath38 with @xmath39 $ ] . here , @xmath40 and @xmath41 ( @xmath42 and @xmath43 )
represent five - dimensional ( four - dimensional ) lorentz indices , and @xmath44 denotes the curvature radius of the ads background .
the fifth direction , denoted as @xmath45 , is compactified on an interval extended from the ultraviolet ( uv ) brane located at @xmath46 to the infrared ( ir ) brane at @xmath47 , i.e. , @xmath48 .
the uv cutoff @xmath49 will be taken to be 0 after all calculations are done .
in addition to the bulk left- ( @xmath50 ) and right- ( @xmath51 ) gauge fields , we introduce a bulk scalar @xmath1 which transforms as a bifundamental representation field under the @xmath52 gauge symmetry , and therefore it is considered to be dual to the quark bilinear operator @xmath53 .
the mass - parameter @xmath54 is then related to @xmath55 as @xmath56 we take @xmath57 when we apply this holographic model to the study of the actual qcd , while we take @xmath58 when we consider the model as a dual of wtc .
the action of the model is given as @xcite @xmath59 where @xmath60 denotes the five - dimensional bulk action , @xmath61 \nonumber \\ & & - \frac{1}{4 } { \rm tr}[l_{mn}l^{mn } + r_{mn}r^{mn } ] \bigg ] \ , , \label{s : bulk}\end{aligned}\ ] ] and @xmath62 the boundary actions which are given in ref .
the covariant derivative acting on @xmath1 in eq.([s : bulk ] ) is defined as @xmath63 , where @xmath64 with @xmath65 being the generators of @xmath66 which are normalized as @xmath67=\delta^{ab}$ ] .
@xmath68 is the five - dimensional field strength which is defined as @xmath69 $ ] , and @xmath70 is defined as @xmath71= ( l / z)^{10}$ ] .
a salient feature of the model is the extra bulk scalar @xmath3 introduced in eq.([s : bulk ] ) in order to reproduce the correct asymptotic behavior of the qcd .
this is a bulk field which is dual to the gluon condensate @xmath72 in qcd . here
, @xmath73 is related to the qcd gauge couping @xmath74 by @xmath75 .
since @xmath76 is a singlet under the chiral @xmath52 symmetry , the dual - bulk scalar @xmath3 has to be a real field .
we take @xmath77 , thus the corresponding bulk - mass parameter becomes @xmath78 .
the gauge coupling @xmath79 and a parameter @xmath80 appearing in the action are fixed as @xmath81 so that the model reproduces the ope in qcd ( see later discussions ) @xcite .
we shall begin with the bulk scalar sector in eq.([s : bulk ] ) . the bulk scalar fields @xmath1 and @xmath3 are parametrized as @xmath82 @xmath83 with the vacuum expectation values ( vevs ) , @xmath84 and @xmath85 .
we hereafter disregard ( techni-)pion fields @xmath86 which will not be relevant for the present study .
the boundary conditions for @xmath87 are @xcite : @xmath88 where @xmath40 stands for the current mass of ( techni-)quarks , and the ir value @xmath16 is related to the ( techni-)quark condensate @xmath89 .
the parameter @xmath90 has been introduced which can arise from the ambiguity of the definition for the current mass @xmath40 , and is fixed to be @xmath91 for qcd and @xmath92 for wtc , by matching the uv asymptotic form of the scalar current correlator to the form predicted from the operator product expansion @xcite .
the boundary conditions for @xmath93 are taken for both qcd and wtc cases as @xmath94 where @xmath95 becomes an external source for the ( techni-)gluon condensate operator @xmath96 , and @xmath17 is a parameter which is associated with the ( techni-)gluon condensate @xmath97 .
the solutions of the vevs @xmath87 and @xmath93 in the limit where @xmath98 and @xmath99 are given as @xcite @xmath100 one can solve the equations of motion for the bulk scalars @xmath101 and @xmath102 with the uv boundary conditions similar to those in eqs .
( [ bc : vs ] ) and ( [ bc : vchi ] ) with @xmath40 and @xmath95 replaced by sources @xmath103 for the scalar and gluonic currents @xmath104 .
putting their solutions back into the action @xmath105 in eq.([s5 ] ) , one can then obtain the generating functional @xmath106 $ ] holographically dual to qcd or wtc .
the chiral condensate and gluon condensate ( @xmath89 and @xmath107 ) are thus calculated by performing @xmath108 and @xmath109 , respectively : @xmath110 where @xmath111 is the chiral condensate renormalized at @xmath112 @xcite .
the chiral condensate renormalized at generic scale @xmath42 is given by @xmath113 with @xmath114 .
one can also calculate the scalar meson ( td ) mass @xmath115 and glueball mass @xmath116 by evaluating the lowest poles of the scalar and gluonic current correlators @xcite to find the following eigenvalue equations : @xmath117 where @xmath118 is the bessel function of the first kind .
we shall move on to the gauge sector in eq.([s : bulk ] ) .
we define the five - dimensional vector and axial - vector gauge fields @xmath119 and @xmath120 as @xmath121 it is convenient to work with the gauge - fixing @xmath122 and take the boundary conditions @xmath123 , @xmath124 and @xmath125 , where @xmath126 and @xmath127 correspond to sources for the vector and axial - vector currents , respectively .
we then solve the equations of motion for ( the transversely polarized components of ) @xmath128 and @xmath129 and substitute the solutions back into the action in eq.([s : bulk ] ) , to obtain the generating functional @xmath130 $ ] holographically dual to qcd or wtc .
then the vector and the axial - vector current correlators are obtained in a way similar to the case of the scalar sector .
the correlators are defined as @xmath131 where the currents are defined as @xmath132 @xmath133 and @xmath134 ( where @xmath135 is the euclidean momentum ) are expressed as @xmath136 where the vector and axial - vector profile functions @xmath137 and @xmath138 are defined as @xmath139 and @xmath140 with the fourier transforms of @xmath126 and @xmath127 .
these profile functions satisfy the following equations : @xmath141 v(q^2 , z ) = 0 , \label{eom : piv}\ ] ] @xmath142 ^ 2 \right]a(q^2 , z ) = 0 , \label{eom : pia}\ ] ] @xmath143 with the boundary conditions @xmath144 and @xmath145 .
it is worth mentioning that both the vector and the axial - vector current correlators involve gluonic effects through @xmath17 .
most notably , we have @xmath146 \ , , \label{piv : complete0}\ ] ] in accord with the ope in qcd : @xmath147 \ , , \label{opeva}\end{aligned}\ ] ] which was the basis for the parameter matching in eq.([matching ] ) . were it not for the @xmath3 term , we would not reproduce the correct gluon condensate term @xmath148 in the qcd asymptotics and also in wtc .
therefore the physical quantities which are related to these correlators , including the muon @xmath0 , are influenced by the existence of gluonic dynamics .
the vector and axial - vector current correlators , @xmath149 and @xmath150 , can be expanded in terms of towers of the vector and axial - vector resonances .
we then identify the lowest poles for @xmath151 as the ( techni-)@xmath152 and @xmath4 mesons .
their masses , @xmath153 and @xmath154 , are calculated by solving the eigenvalue equations for the vector and axial - vector profile functions @xcite : @xmath155 v(z ) = 0 , \label{mrho : eq}\ ] ] @xmath156 ^ 2 \right]a(z ) = 0 , \label{ma1:eq}\ ] ] with the same boundary conditions @xmath157 and @xmath158 .
we thus find @xmath153 and @xmath154 as functions of the model parameters @xmath16 , @xmath17 and @xmath55 with the overall scale set by @xmath18 : @xmath159 and @xmath160 .
the ( techni-)pion decay constant is expressed as @xmath161 , which is , in the case of wtc , related to the ew scale @xmath162 gev as @xmath163 . here
, @xmath164 denotes the number of ew doublets , which is fixed to be 1 in the case of qcd .
the present model enables us to express @xmath165 as a function of @xmath16 , @xmath17 , @xmath55 and @xmath18 @xcite : @xmath166 where @xmath167 .
the @xmath33 parameter @xcite is calculated from @xmath149 and @xmath150 as @xmath168 $ ] , where @xmath169 .
thus @xmath33 is expressed as a function of two parameters @xmath16 and @xmath17 @xcite once @xmath55 is fixed : @xmath170 ^ 2 \right ) \ , , \label{spara } \end{aligned}\ ] ] where @xmath171 . once @xmath172 is calculated in the holographic model , the electromagnetic current correlator @xmath14 is obtained as @xmath173 where the summation runs over all the fermion flavors @xmath174 with their electromagnetic charge denoted as @xmath175 . here , we defined electromagnetic current correlator as @xmath176 where the electromagnetic current is defined as @xmath177 . with the electromagnetic current correlator given
, the ( techni-)hadronic leading order ( hlo ) contribution to the muon @xmath0 is calculated to be @xcite ( for a graphical expression , see fig .
[ hlofig ] . ) .
, width=226 ] @xmath178 where @xmath179 with @xmath180 being the electromagnetic coupling constant .
@xmath181 is the renormalized electromagnetic current correlator defined as @xmath182 , and @xmath183 is a weight function which has the following form : @xmath184 the formula can be rewritten in terms of @xmath149 as follows : @xmath185 where @xmath186 is the prefactor in eq .
( [ eq : piem ] ) , which takes the following values in the case of qcd and one - family technicolor model @xcite : @xmath187 now , all the physical quantities we are interested in have been expressed in terms of calculable holographic quantities . in the following two sections , for the case of qcd and wtc
respectively , we evaluate those physical quantities as functions of the model parameters @xmath188 and @xmath55 , especially focusing on their dependence on @xmath17 closely related to the gluon condensate .
in this section , we apply the holographic calculations explained in the previous section for the estimation of various physical quantities in the real - life qcd . for this purpose , we take @xmath6 so that the model correctly reproduces the uv asymptotic behavior of the qcd . after fixing the value of @xmath55 ,
all the physical quantities are expressed as functions of @xmath18 , @xmath16 and @xmath17 .
we use @xmath189 mev and @xmath19 mev as inputs to fix two of these remaining model parameters .
then , from eqs .
( [ mrho : eq ] ) and ( [ fpi ] ) , one can express @xmath18 and @xmath16 as functions of @xmath17 .
now , the remaining physical quantities , namely @xmath154 , @xmath190 , @xmath191 , @xmath115 , @xmath116 and @xmath192 can be expressed as functions of a single parameter @xmath17 through eqs .
( [ ma1:eq ] ) , ( [ gcond ] ) , ( [ condensates ] ) , ( [ ms ] ) , ( [ mg ] ) and ( [ spara ] ) , which are depicted in fig .
[ fpimrhofixed ] by varying @xmath17 from 0 to 0.4 : with @xmath19 mev and @xmath193 mev fixed .
top - left : @xmath154 @xcite ; top - center : @xmath190 @xcite ; top - right : @xmath194 @xcite ; bottom - left : @xmath195 @xcite ; bottom - center : @xmath196 @xcite ; bottom - right : @xmath192 @xcite .
two dashed red lines in each plot correspond to observed upper and lower values of 1-@xmath197 error band quoted in the corresponding references .
[ fpimrhofixed],title="fig:",width=204 ] with @xmath19 mev and @xmath193 mev fixed .
top - left : @xmath154 @xcite ; top - center : @xmath190 @xcite ; top - right : @xmath194 @xcite ; bottom - left : @xmath195 @xcite ; bottom - center : @xmath196 @xcite ; bottom - right : @xmath192 @xcite .
two dashed red lines in each plot correspond to observed upper and lower values of 1-@xmath197 error band quoted in the corresponding references .
[ fpimrhofixed],title="fig:",width=204 ] with @xmath19 mev and @xmath193 mev fixed .
top - left : @xmath154 @xcite ; top - center : @xmath190 @xcite ; top - right : @xmath194 @xcite ; bottom - left : @xmath195 @xcite ; bottom - center : @xmath196 @xcite ; bottom - right : @xmath192 @xcite .
two dashed red lines in each plot correspond to observed upper and lower values of 1-@xmath197 error band quoted in the corresponding references .
[ fpimrhofixed],title="fig:",width=204 ] + with @xmath19 mev and @xmath193 mev fixed .
top - left : @xmath154 @xcite ; top - center : @xmath190 @xcite ; top - right : @xmath194 @xcite ; bottom - left : @xmath195 @xcite ; bottom - center : @xmath196 @xcite ; bottom - right : @xmath192 @xcite .
two dashed red lines in each plot correspond to observed upper and lower values of 1-@xmath197 error band quoted in the corresponding references .
[ fpimrhofixed],title="fig:",width=204 ] with @xmath19 mev and @xmath193 mev fixed .
top - left : @xmath154 @xcite ; top - center : @xmath190 @xcite ; top - right : @xmath194 @xcite ; bottom - left : @xmath195 @xcite ; bottom - center : @xmath196 @xcite ; bottom - right : @xmath192 @xcite .
two dashed red lines in each plot correspond to observed upper and lower values of 1-@xmath197 error band quoted in the corresponding references .
[ fpimrhofixed],title="fig:",width=204 ] with @xmath19 mev and @xmath193 mev fixed .
top - left : @xmath154 @xcite ; top - center : @xmath190 @xcite ; top - right : @xmath194 @xcite ; bottom - left : @xmath195 @xcite ; bottom - center : @xmath196 @xcite ; bottom - right : @xmath192 @xcite .
two dashed red lines in each plot correspond to observed upper and lower values of 1-@xmath197 error band quoted in the corresponding references .
[ fpimrhofixed],title="fig:",width=204 ] observed values of those quantities are also indicated in the plots : two dashed red lines correspond to upper and lower values of 1-@xmath197 error band .
the observed " glueball mass @xmath116 has been taken from an expected mass range in lattice simulations @xcite .
as for the flavor - singlet two - quark bound state @xmath198 meson , we have chosen @xmath21 and estimated the observed " mass neglecting mixing with four - quark bound state @xmath199 .
( for more detailed discussions , see ref .
@xcite . ) the figures tell us that the @xmath4 meson mass @xmath154 and gluon condensate @xmath190 highly depend on the change of @xmath17 , while other quantities are rather insensitive to it , keeping values at around observed values of each quantity . from these results , we see the optimal value of @xmath17 ( and hence @xmath18 and @xmath16 ) : @xcite @xmath200 for the holographic model with @xmath6 to reproduce the qcd observables . in table
[ fpi : mrho : ob : fixed : tab ] we show the results of holographic calculations for @xmath201 ( and @xmath25 for comparison ) along with observed value of each quantity .
[ h ] .the predicted values of various qcd observables obtained from holographic calculations for @xmath201 ( and @xmath25 ) with @xmath202 mev , @xmath193 mev fixed , in comparison with the observed values .
[ cols="^,^,^,^,^,^,^",options="header " , ] as we expect , @xmath203 is about 6 times larger compared to the ratio in the case of muon @xmath0 .
we have shown the results of holographic calculations of the ( techni-)hlo contributions to the anomalous magnetic moment @xmath0 of leptons .
it was shown that , in the case of the qcd contribution , it was enhanced by about 6% due to the proper inclusion of the gluon - condensation effect ( @xmath201 ) compared to the estimate ignoring it ( @xmath25 ) @xcite , leading to a better agreement with the known value determined by the experimental data @xcite : @xmath204 considering that our estimate was at the chiral limit @xmath205 , the agreement is rather impressive . in the case of wtc , where the gluonic effect plays more significant role compared to the case of qcd to reproduce the observed 125 gev scalar boson ,
the gluon - condensation effects are more dramatic for the muon @xmath0 : the value of the techni - hlo contribution at physical value @xmath32 is more than 100 times larger compared to the estimate ignoring the gluonic effect(@xmath25 ) , if we allow rather large value of @xmath33 up to @xmath206 .
it is quite interesting that the contributions to the anomalous magnetic moment from the vacuum polarization of the electromagnetic current is several orders of magnitude larger than the naive scale - up estimate , which is obtained by simply multiplying the ratio of the squares of typical scales of qcd ( @xmath207 92 mev ) to one - family model ( @xmath208 123 gev ) to the value of @xmath209 , yielding a value of @xmath210 .
this value roughly coincides with the value of holographic calculation with @xmath25 , @xmath211 ( see table [ tab : athlo - comp ] ) .
though this kind of significant enhancement is quite interesting theoretically , the phenomenological interest is whether wtc contribution is visible or not .
even if there is an enhancement by several orders of magnitudes compared to the naive estimate , the contribution from the wtc dynamics to the @xmath0 is still negligibly small compared to the qcd contribution . the magnitude of wtc contribution to the muon @xmath0 is , at most , order of @xmath212 for the cases of setup investigated in this paper .
this value is still quite small compared to the current discrepancy between the experimental value of the muon @xmath0 and the standard model prediction of it : @xmath213 @xcite .
therefore , it seems rather unlikely that the contributions from wtc dynamics can explain the current 3.3 @xmath36 deviation between experiment and the sm prediction . in conclusion
we have shown , in a bottom - up holographic model for qcd / wtc , that the amount of gluonic effects to realize the physical point ( @xmath201 for qcd and @xmath214 for wtc with @xmath215 ) makes significant effects enhancing the ( techni- ) hlo contributions to the lepton @xmath0 . since
the introduction of the gluonic effects in this model is just to make the model consistent with the high energy region of the qcd , it is a highly nontrivial test of this holographic model whether or not the effects also improve the agreements with the low energy hadron physics , particularly in the most relevant momentum region @xmath216 for @xmath217 ( see fig.[fig : pi_wtc ] ) , which is far infrared region for qcd .
note that the quantity studied in this paper is in the space - like momentum region @xmath218 which is theoretically more tractable without limitation of the zero - width constraint of large @xmath30 limit where the holography is so far justified , in contrast to the resonance phenomenology in the time - like region .
our work presents an explicit example of a holographic model which successfully reproduces the qcd physics in all energy region from the very ultraviolet region for ope all the way down to such a deep infrared region in the space - like momentum . the vital role of the gluon condensate in wtc in connection with the light scalar composite can be clarified by the future lattice studies such as an extension of the work which observed a light flavor - singlet scalar meson in large @xmath219 qcd @xcite .
this work was supported by the jsps grant - in - aid for scientific research ( s ) # 22224003 and ( c ) # 23540300 ( k.y . ) .
j. m. maldacena , adv .
* 2 * , 231 ( 1998 ) [ int .
j. theor .
phys . * 38 * , 1113 ( 1999 ) ] .
l. da rold and a. pomarol , nucl .
b * 721 * , 79 ( 2005 ) .
j. erlich , e. katz , d. t. son and m. a. stephanov , phys .
lett . * 95 * , 261602 ( 2005 ) .
t. sakai and s. sugimoto , prog .
phys . * 113 * , 843 ( 2005 ) .
k. haba , s. matsuzaki and k. yamawaki , phys .
d * 82 * , 055007 ( 2010 ) .
s. matsuzaki and k. yamawaki , phys .
d * 86 * , 115004 ( 2012 ) . k. yamawaki , m. bando and k. matumoto , phys .
lett . * 56 * , 1335 ( 1986 ) ; m.
bando , t. morozumi , h. so and k. yamawaki , phys .
lett . * 59 * , 389 ( 1987 ) .
bando , k. matumoto and k. yamawaki , phys .
b * 178 * , 308 ( 1986 ) .
s. chatrchyan _ et al . _
[ cms collaboration ] , phys .
b * 716 * , 30 ( 2012 ) ; g. aad _ et al .
_ [ atlas collaboration ] , phys .
b * 716 * , 1 ( 2012 ) .
m. harada , m. kurachi and k. yamawaki , phys .
d * 68 * , 076001 ( 2003 ) ; m. kurachi and r. shrock , jhep * 0612 * , 034 ( 2006 )
. s. shuto , m. tanabashi and k. yamawaki , in _ proc .
1989 workshop on dynamical symmetry breaking _ , dec .
21 - 23 , 1989 , nagoya , eds .
t. muta and k. yamawaki ( nagoya univ . , nagoya , 1990 ) 115 - 123 ; m. s. carena and c. e. m. wagner , phys .
b * 285 * , 277 ( 1992 ) ; m. hashimoto , phys .
b * 441 * , 389 ( 1998 ) .
m. hashimoto and k. yamawaki , phys .
d * 83 * , 015008 ( 2011 ) .
s. matsuzaki and k. yamawaki , phys .
d * 85 * , 095020 ( 2012 ) ; phys .
d * 86 * , 035025 ( 2012 ) ; phys .
b * 719 * , 378 ( 2013 ) .
d. k. hong , d. kim and s. matsuzaki , phys .
d * 81 * , 073005 ( 2010 ) .
k. hagiwara , r. liao , a. d. martin , d. nomura and t. teubner , j. phys .
g * 38 * , 085003 ( 2011 ) .
m. e. peskin and t. takeuchi , phys .
* 65 * ( 1990 ) , 964 ; phys . rev .
d * 46 * ( 1992 ) , 381 ; b. holdom and j. terning , phys .
b * 247 * ( 1990 ) , 88 ; m. golden and l. randall , nucl .
b * 361 * ( 1991 ) , 3 .
k. haba , s. matsuzaki and k. yamawaki , prog .
theor . phys .
* 120 * , 691 ( 2008 ) .
t. blum , phys .
lett . * 91 * , 052001 ( 2003 ) ; c. aubin and t. blum , phys . rev .
d * 75 * , 114502 ( 2007 ) .
e. farhi and l. susskind , phys .
d * 20 * , 3404 ( 1979 ) .
j. beringer _
[ particle data group collaboration ] , phys .
d * 86 * , 010001 ( 2012 ) .
m. a. shifman , a. i. vainshtein and v. i. zakharov , nucl .
b * 147 * , 385 ( 1979 ) ; nucl .
b * 147 * , 448 ( 1979 ) .
j. gasser and h. leutwyler , phys .
rept . * 87 * , 77 ( 1982 ) .
for a recent review , see w. ochs , arxiv:1301.5183 [ hep - ph ] .
m. harada and k. yamawaki , phys .
b * 297 * , 151 ( 1992 ) .
d. nomura and t. teubner , nucl .
b * 867 * , 236 ( 2013 ) .
s. eidelman and m. passera , mod .
a * 22 * , 159 ( 2007 ) .
y. aoki , t. aoyama , m. kurachi , t. maskawa , k .-
nagai , h. ohki , a. shibata , k. yamawaki and t. yamazaki , phys .
d * 78 * , 094511 ( 2013 ) . g. cacciapaglia , c. csaki , c. grojean and j. terning , phys .
d * 71 * , 035015 ( 2005 ) ; r. foadi , s. gopalakrishna and c. schmidt , phys .
b * 606 * , 157 ( 2005 ) ; r. s. chivukula , e. h. simmons , h. j. he , m. kurachi and m. tanabashi , phys .
d * 72 * , 016008 ( 2005 ) .
g. w. bennett _ et al . _ [ muon g-2 collaboration ] , phys . rev .
d * 73 * , 072003 ( 2006 ) .
y. aoki , t. aoyama , m. kurachi , t. maskawa , k .-
nagai , h. ohki , e. rinaldi , a. shibata , k. yamawaki and t. yamazaki , arxiv:1305.6006 [ hep - lat ] . | we study gluonic effects " ( gluon condensation effects ) on the hadronic leading order ( hlo ) contributions to the anomalous magnetic moment ( @xmath0 ) of leptons , based on a holographic model having explicit gluonic mode introduced for consistency with the operator product expansion of qcd .
we find gluonic enhancement of hlo contributions to the muon @xmath0 by about 6% , which nicely fills in the gap between the holographic estimate without gluonic effects and the phenomenological one using the experimental data as inputs .
similar calculations including the gluonic effects for the electron and the tau lepton @xmath0 are also carried out in good agreement with the phenomenological estimates .
we then apply our holographic estimate to the walking technicolor ( wtc ) where large techni - gluonic effects were shown to be vital for the technidilaton , ( pseudo ) nambu - goldstone boson of the ( approximate ) scale symmetry of wtc , to be naturally as light as 125 gev .
it is shown that the value of the techni - hlo contributions to the muon @xmath0 is 10 - 100 times enhanced by inclusion of the same amount of the gluonic effects as that realizing the 125 gev technidilaton , although such an enhanced techni - hlo contribution is still negligibly small compared with the current deviation of the standard model prediction of the muon @xmath0 from the experiments .
the techni - hlo contributions to the tau lepton @xmath0 is also discussed , suggesting a possible phenomenological relevance to be tested by the future experiments . |
the following effective lagrangian for gluon field and its motivation were discussed in @xcite : @xmath0 here @xmath1 is a scale of qcd related to gluon condensate , @xmath2 , and @xmath3 the minima of the effective potential are achieved for covariantly constant abelian ( anti-)self - dual fields @xmath4 where the matrix @xmath5 belongs to the cartan subalgebra of @xmath6 @xmath7 the minima are connected by discrete parity and weyl transformations , which indicates that the system is prone to existence of solitons in real space - time and kink configurations in euclidean space .
group - theoretical analysis of symmetry breaking and domain wall formation was done in @xcite .
let us consider the simplest kink configuration
kink interpolating between self - dual and anti - self - dual abelian vacua .
if the angle @xmath8 between chromoelectric and chromomagnetic fields is allowed to deviate from the constant vacuum value and all other parameters are fixed to the vacuum values , then the lagrangian reads @xmath9 the corresponding equation of motion is sine - gordon equation @xmath10 which have the following kink solution : @xmath11 the angle @xmath8 interpolates between @xmath12 and @xmath13 .
@xmath14 stays for one of the four euclidean coordinates .
the kink describes a planar domain wall between the regions with almost homogeneous abelian self - dual and anti - self - dual gluon fields .
chromomagnetic and chromoelectric fields are orthogonal to each other on the wall . far away from the wall the field is self - dual or anti - self - dual , and absolute value of the topological charge density
is equal to the value of the gluon condensate .
the topological charge density vanishes on the wall .
we may now construct kink superposition with the standard methods .
let us denote the general kink configuration as @xmath15 where @xmath16 is the inverse width of the kink , @xmath17 is a normal vector to the plane of the wall .
the general kink network is then given by @xmath18 now we consider more general kink - like solution : three parameters @xmath19 and @xmath20 deviate from their vacuum values near the center of the kink ( fig . [ 3-kink ] ) .
one may see that the value of scalar condensate @xmath21 decreases in the center of the kink which may increase size of stable chromomagnetic flux tube .
calculation of one - loop quark contribution to the qcd effective potential demonstrated that strong electromagnetic fields can play the role of a trigger for deconfinement @xcite .
the catalyzing role of strong magnetic fields was also observed in lattice qcd simulations @xcite . during heavy ion collisions ,
the electric @xmath22 and magnetic @xmath23 fields are almost orthogonal to each other @xcite : @xmath24 . for this configuration of the external electromagnetic field ,
the one - loop quark contribution to the qcd effective potential for the homogeneous abelian gluon fields is minimal for the chromoelectric and chromomagnetic fields directed along the electric and magnetic fields respectively @xcite .
the orthogonal chromo - fields are not confining : color charged quasiparticles can move along the chromomagnetic field .
deconfined quarks as well as gluons can move preferably along the direction of the magnetic field even after the electromagnetic field turns off .
this means that a defect in domain wall network may be formed during heavy ion collision ( see fig .
[ fig : chromo_bag ] ) .
let us consider eigenvalue problem for vector field @xmath25q_\nu=\lambda^2 q_\mu\ ] ] with the boundary conditions @xmath26 the corresponding eigenvalues are @xmath27 in the finite trap the lowest eigenvalue is @xmath28 few lowest eigenvalues @xmath29 as functions of @xmath30 are shown in fig .
[ fig : eigen_flow ] .
one concludes that if the dimensionless size @xmath30 of the trap is sufficiently small @xmath31 then there are no unstable tachyonic modes in the spectrum of color charged vector fields . to estimate the critical size
one may use the mean phenomenological value of the gluon condensate ( gauge coupling constant @xmath32 is included into the field strength tensor ) @xmath33 equation leads to the critical radius @xmath34 however , as it follows from fig . [ 3-kink ] , scalar condensate become smaller at the domain wall and may increase the value of critical radius .
[ fig : eigen_flow ] in minkowski space - time the eigenvalue problem turns to the wave equation : @xmath35q_\nu=0.\ ] ] solution for this equation can be found as a linear combination of eigenmodes with @xmath36 .
@xmath37 equation can be treated as the dispersion relation between energy @xmath38 and momentum @xmath39 for the quasiparticles . unlike the chromomagnetic field ,
the ( anti-)self - dual fields characteristic for the bulk of domain network configuration lead to purely discrete spectrum of eigenmodes in euclidean space and do not possess any quasiparticle treatment in terms of dispersion relation between energy and momentum for elementary color charged excitations .
even though a single chromomagnetic trap has critical size , overall volume of deconfinement transition in hadronic matter is not restricted by @xmath40 .
numerous traps may be formed , for instance , due to the influence of strong electromagnetic fields in the whole region of relativistic heavy ion collision @xcite .
the traps may merge into a large chromomagnetic lump developing the flux tube structure inside the lump .
equivalently , one may also think about formation of the flux tubes inside the traps with overcritical size emerging under conditions of high energy or / and baryon number density , strong electromagnetic fields . in the absence of any external impact one
expects that the fraction of the deconfinement phase is statistically negligible as it occupies an essentially three - dimensional sub - manifold in the four - dimensional space @xmath41 .
the influence of external conditions like the high energy density , high baryon density and strong electromagnetic fields can trigger the growth of the relative fraction of the deconfined phase . under certain conditions
high energy density accumulated by colorless collective excitations can be converted into the thermodynamics of color charged quasiparticles .
the growth of the deconfinement phase fraction may be described in terms of creation of either numerous stable chromomagnetic traps or / and the trap with overcritical size , populated by the polarized chromomagnetic tubes @xcite . in the confinement regime ( color charge quasiparticles
are completely suppressed ) one may expect @xmath42 while the deconfinement regime ( color charged quasiparticles are fully activated ) occurs if @xmath43 the scalar gluon condensate may stay practically unchanged during the phase transformation .
the regime of dilute plasma can not be reached unless the scalar gluon condensate density vanishes , which would require complete rearrangement of the quantum effective action . however , as soon as the color charged quasiparticles are activated the structure of the effective action itself in principle becomes dependent on their temperature and density , and it can not be excluded that the scalar condensate vanishes at certain critical values of the temperature and density . in summary ,
the domain wall network representation of qcd vacuum is suggestive of a two - stage deconfinement transition . at the first stage
topological charge density vanishes , but the scalar condensate stays almost unchanged , color charged quasiparticles are activated at this stage while the colorless collective excitations can decay into the color charged ones .
the system is far from being a dilute gas . at the second stage
the scalar gluon condensate vanishes , and the system turns into dilute quark - gluon plasma . | domain model of confinement , chiral symmetry breaking and hadronization is based on description of qcd vacuum gluon configurations as an ensemble of almost everywhere homogeneous abelian ( anti-)self - dual fields .
the ensemble can be explicitly constructed as domain wall network .
we shortly overview this approach and within this framework discuss the effects of qcd vacuum polarization by strong electromagnetic fields .
it is stressed that such polarization effects can play the role of trigger for deconfinement . |
understanding the formation and growth of supermassive black holes ( smbh ) in galaxy evolution is a key problem in astrophysics .
some of the largest puzzles are the origin of m@xmath9-@xmath10 relationship ( @xcite , @xcite , @xcite ) , the role of active galactic nuclei ( agn ) feedback and its effects for quenching star formation ( e.g. , @xcite ) , and how to effectively transport gas to the galactic nuclei to fuel black hole growth ( e.g. , @xcite ) .
the majority of quasi - stellar object ( qso ) host galaxy studies have concentrated on nearby systems ( [email protected] ) using optical observations with hubble space telescope ( hst ) and ground - based facilities ( e.g. , @xcite ) . the key ingredient in these observations is to achieve the high spatial resolutions necessary to disentangle bright qso emission from that of the underlying stellar population and hii regions .
a range of host galaxy parameters has been discovered , implying that qsos are hosted by many different galaxy types with a range of simultaneous star formation activity ( e.g , @xcite ) , qsos have been found in massive inactive elliptical galaxies , late - type spirals and irregulars .
even with the large range of selection effects , there is a coherent picture that luminous nearby qsos are generally found in luminous and massive host galaxies with a range of morphologies @xcite .
however , at high - redshift ( z@xmath111 ) the picture of qso host galaxies is less clear with only a small number of host systems observed .
high redshift qsos have been found in star forming galaxies with morphologies ranging from discs ( e.g. , @xcite ) to mergers ( e.g. @xcite ) , while some studies have shown qsos to reside in passive , elliptical galaxies ( e.g. @xcite ) .
one of the most compelling physical explanations of the co - evolution of the host galaxy and smbh , is negative feedback from agn energetics .
there has been mounting observational evidence supporting star formation quenching via qso / agn activity by expelling large reservoirs of cold gas and/or heating of the gas in massive halos ( @xcite , references therein ) .
recent studies have found that the majority of low redshift type-2 qsos ( z@xmath0 0.2 ) contain evidence of galaxy wide outflows on kpc scales with [ oiii]5007 emission lines @xcite ; however their effects on star formation rates are yet to be understood .
similarly , recent integral field spectrograph ( ifs ) observations of [ oiii ] @xmath12 h@xmath2 emission in z@xmath02 qsos have revealed host galaxies with strong evidence of outflows , and lower star formation rates in regions with the strongest outflows ( gemini nifs : @xcite , vlt sinfoni : @xcite ) .
these observations have given tantalizing clues of qso feedback , yet there is still little known about the z@xmath02 host galaxies ( i.e , stellar mass , dynamics , metallicities ) , and whether they obey the present day black hole mass - galaxy scaling relations @xcite .
as qsos outshine their host galaxies by several orders of magnitude , studying their hosts requires a careful removal of the qso emission , for which a good understanding of the point spread function ( psf ) is required . understanding the psf for ground based observations is very difficult since atmospheric variations cause the psf to change over a time span of a few seconds , making it extremely difficult to model .
there have been some successful attempts to remove the bright qso light using nearby stars as reference to detect extended emission from the host galaxy ( e.g. , seeing - limited:@xcite hst / ao : @xcite ) .
the majority of qso host galaxy observations have used space - based observations where the psfs are stable for qso removal . at low redshift ( z@xmath61 )
there have been several studies that used both artificial and stellar psfs to remove qso light to search for extended emission , which have allowed for several successful studies of low and intermediate redshift qso hosts ( @xcite , @xcite , @xcite ) . at high - redshift ( 1 @xmath6 z @xmath64 ) , these searches have been more challenging because the angular scales of host galaxies are comparable to the psf halo ( @xmath01 ) and psf removal is dominated by residuals , which makes it difficult to disentangle the qso and host galaxy .
the bigger difficulty comes from extracting star formation rates ( sfr ) and metallicities of the host galaxies , since these quantities can be easily contaminated by qso narrow - line emission with a range of spatial and kinematic offsets ( @xmath6 1000 kms@xmath5;@xcite , @xcite ) .
broad - band photometry has been used to model the stellar properties of distant host galaxies , however residual noise from psf subtraction makes it difficult to obtain accurate magnitudes , and there are no reliable tests to distinguish stellar rest - frame optical continuum from the synchrotron emission of the central agn .
a combination of adaptive optics ( ao ) and integral field spectroscopy ( ifs ) provides the necessary spatial and spectral resolutions required to separate qso emission from its host .
having spectral information at each spatial location allows us to extract key information about the galaxy that an imaging survey simply can not achieve .
ifs observations provide a powerful technique to remove the bright qso .
this can be achieved by utilising unresolved emission from the qso ( i.e. , broad - line emission , like h@xmath2 ) to construct a pure qso psf image .
this psf image is normalized and then subtracted per wavelength channel in the data cube , thus leaving only narrow - line emission .
if there is spatially offset narrow - line emission , this can be used directly to infer kinematics , dynamical masses ( assuming virialized gas ) , and nebular emission diagnostics of the gas . recently this technique was proven to be effective in resolving the host galaxy of a redshift z=1.3 qso using sinfoni on the vlt ( @xcite ) .
these authors were successful at detecting the host galaxy and were able to construct a spatially resolved narrow emission line map with identified ionization mechanisms and star formation rates ( 100 m@xmath4 yr@xmath5 ) .
they found that the galaxy dynamical mass and black hole mass obeyed the present - day m@xmath13 vs. m@xmath14 relation within the current scatter .
in contrast , there have been no ifs observations of high - z qsos hosts where the central agn has been shown to regulate star formation . while evidence for qso driven winds at low and high - redshifts have been found ,
only a single case has shown direct evidence that suggests these winds regulate star formation ( @xcite ) . a larger sample of high - z qso host galaxy observations
are needed to build - up a coherent picture .
we have conducted an ifs lgs - ao pilot survey of five z@xmath02 type - i qsos using both keck ii and gemini north facilities to demonstrate the feasibility and limits of qso host galaxy detection at high - redshift , and to obtain a range of qso properties . in [ obs ] we describe observations and target selection . in [ data ] we present the data reduction . in [ extract ]
we describe our psf extraction and removal technique , in [ results ] we discuss our two sources which had a narrow h@xmath2 detection , and describe how we obtained our flux limits in sources with null detections , and in [ disc ] we interpret the results for two of our sources ( sdssj1029 + 6510 & sdssj0925 + 0655 ) and derive dust - corrected star formation rate limits .
we compare our results with studies of qsos at similar bolometric luminosities , and in [ conclusion ] we provide our conclusions . throughout the paper
we assume a @xmath15-dominated cosmology with @xmath16=0.308 , @xmath17=0.692 , and h@xmath18=67.8 kms@xmath5 mpc@xmath5 @xcite .
we used the near infrared integral field spectrographs osiris ( @xcite ) on the keck telescope and nifs @xcite on the gemini north telescope ( program identification gn-2012b - q-53 ) coupled with the observatories laser guide star adaptive optics systems .
we present k - band spectra of 5 quasars at an average redshift of z @xmath19 2.2 ( angular size scale , 8.5 kpc per arcsecond ) with an average total on - target integration time of 3600s . on each night
we observed an a type standard star for telluric correction and flux calibration .
table [ tabobs ] summarizes our observational parameters and setup .
we selected these qsos from the fifth edition of the sdss quasar catalog based on the seventh data release ( @xcite ) .
for this pilot survey we selected sources that would have optimal ao performance to aid in the psf subtraction .
criteria for the keck and gemini north observations were : ( 1 ) all objects must be observable with the altair and keck ao systems based on tip / tilt magnitude and separations ( r mag @xmath20 16.5 within 25 for altair system and r mag @xmath20 18.5 within 45 for keck - ao ) , and ( 2 ) objects must have redshift between 2.016 and 2.427 where h@xmath2 falls in the prime k - band wavelength regime ( @xmath20 2.2 @xmath21 m ) . using these constraints at k - band
allowed only @xmath030 observable qsos .
we made our final selection based on available tip / tilt stars that are bright and close in separation : one with on - axis tip - tilt source correction ( r=16.4 mag ) , and four for off - axis tip - tilt correction .
table [ tabobs ] contains all the information on the tip / tilt stars .
all of our selected sources are type 1 radio - quiet qsos with 1.4 ghz flux @xmath20 0.15 mjy ( @xcite ) with no availible near - ir spectroscopy , making our sample less biased towards qso hosts with high star formation rates .
host galaxies with high star formation rates presented in @xcite , @xcite were pre - selected based on long slit spectra of the [ oiii]5007 line or far - ir observations . for multi - wavelength analysis of our objects
we include archival observations on our sources .
table [ tabphoto ] contains optical to near - infrared archival photometric information on our qso sample , encompassing archival data from the sdss for the optical magnitudes and 2mass for near - infrared . as of data release 10
, sdss has incorporated wise and 2mass photometric data into their catalog , made available in web format on the object explorer website that can be accessed through sdss3.org . in table
[ tabphotoir ] we present photometry for the four wise bands at 3.4 , 4.6 , 12 and 22 .
all five sources are detected in the 3.4 - 12 bands however only three sources have reliable photometry , where the other two suffer from confusion of flux from the bright nearby tip / tilt stars .
three sources are detected in the 22 band , one is undetected and one does nt have reliable photometry due to confusion ; please see table [ tabphotoir ] for details on the individual sources .
two of our sources , sdssj1029 + 651 and sdssj2123 - 005 were observed with the herschel space telescope s spire instrument h.netzer , observation id:1342270338 ] in the 250 , 350 & 500 bands .
we downloaded the fully reduced level 2 maps from the herschel data archive ( http://irsa.ipac.caltech.edu/applications/herschel/ ) , we converted the maps from jy / beam to jy / pixel by dividing the maps by the beam size found in the spire handbook , available at herschel.esac.esa.int and applied standard aperture photometry over the beam size ( 17.6 , 23.9 , 35.2 ) of the telescope in each of the bands at the optical location of the qsos from table [ tabobs ] .
the two sources are undetected in all of the bands and we provide the three sigma limits in table [ tabphotoir ] .
the osiris observations were reduced using the publicly available osiris data reduction pipeline .
dark frames were median combined to produce a master dark frame using the osiris pipeline routine `` combine frames '' .
each science and calibration frame then had the master dark subtracted from it and the following pipeline routines were performed : `` adjust channel levels '' , `` remove crosstalk '' , `` clean cosmic rays '' , `` extract spectra '' , assemble data cube , `` correct dispersion '' . for sky subtraction ,
each science frame had the nearest in time sky frame subtracted using the `` scaled sky subtraction '' routine that accounts for the temporal variability of the oh sky lines ( @xcite ) .
the science and telluric frames were stacked together using a 3@xmath10 mean clip algorithm in the `` mosaic frames '' routine to remove large bad pixels that occur from the `` extract spectra '' routine . a 1d telluric spectrum was then extracted from the highest signal - to - noise spaxels in the telluric cube using the extract star routine .
strong hydrogen absorption lines were masked using the `` remove hydrogen lines '' routine , and the blackbody of the star was subtracted using the `` divide blackbody '' routine . the spectra were normalized and used to correct for atmospheric absorption and the instrumental footprint in the mosaiced science frame .
the final science data was flux calibrated using standard star observations that were taken closest in time , at similar air mass and were reduced in the same manner as described above .
the nifs observations were reduced using the gemini nifs iraf reduction pipeline that operates within pyraf .
some modifications were applied to the standard pipeline and additional routines were written to match our science goals . for each night
we reduced the xe , ar lamp observations to establish the wavelength solution for each of the targets using the gemini nifs pyraf baseline calibration routine .
dark frames for the science observations were median combined and subtracted from each of the science and sky frames .
the science , telluric , and sky frames were then reduced using the nifs science reduction routine .
the end result is a data cube which has been flat fielded , bad pixel masked and reformatted into a 3d cube , which was spatially re - sampled from the native spatial sampling of 0.1 @xmath22 0.04 to square pixels with a size of 0.05 .
the science and telluric frames had the nearest sky frame in time subtracted , with oh emission line scaling between the sky and science frames .
the centroids of the qso and telluric stars were obtained through a 2d gaussian fit to a spectrally collapsed image , and the dithered observations were shifted and stacked using a 3@xmath10 mean clipped algorithm . the 1d telluric star spectrum
was extracted by averaging spatially over the highest signal - to - noise spaxels , its blackbody was subtracted , and the strong hydrogen absorption line were masked .
the 1d telluric spectrum was then divided into the science cube to correct for atmospheric and instrumental absorption features .
using the sdss spectra ( figure [ sloan_spec ] , left ) we derive the bolometric luminosity ( l@xmath23 ) from the rest - frame 1450 continuum using methodology presented in @xcite .
we obtain the black hole mass ( m@xmath13 ) using equation ( 7 ) presented in @xcite , utilizing the 1350 continuum value together with 1549 civ fwhm , derived by fitting a gaussian profile using the curve@xmath24fit function that is part of the scipy package , written for python based on non - linear least squares routine .
table [ tabprop ] contains the above information . using our k - band qso spectra ( figure [ sloan_spec ] , right side ) we derive luminosity of the broad h@xmath2 emission line , black hole mass and the equivalent width ( table [ tabpropha ] ) .
we fit the line using a gaussian profile which assisted in deriving the broad h@xmath2 luminosity , redshift and equivalent width . in deriving the line luminosity and equivalent width
we integrate over @xmath251.3@xmath22 the fwhm of the fitted profile .
the black hole mass was then estimated using equation ( 6 ) from @xcite .
the presented near infrared spectra were extracted from our data cubes using a spatial aperture of approximately the seeing halo .
+ the broad h@xmath2 emission originates from gas located in a compact disk within the central few parsecs making this emission essentially point - like in our observations .
we use spectral channels that confine the broad line emission for psf construction .
our algorithm finds the highest signal - to - noise spectral channels that do not coincide with oh emission lines to be combined to create a master psf image .
generally the selected psf regions are 2.5 - 3 nm ( 10 - 15 spectral channels ) in size and tend to sit near the peak of the broad h@xmath2 line .
we hypothesised that the majority of the extended narrow line emission will be within 400kms@xmath5 from the qso s redshift , where the psf has the highest signal to noise and the greatest potential for contamination from the nlr , so we also select spectral regions that are offset from the peak of the broad emission line ( not including oh sky lines ) , that should have minimal contribution from extended narrow - line emission .
we combine all spectral regions using a 3@xmath10 clipping routine , to mitigate contamination from the extended narrow - line emission .
this way spaxels that do contain narrow emission would be weighted less since spectral channels offsets by 2000 - 3000kms@xmath5 are less likely to contain nlr .
the end result is a 2d image of our observed psf that gets normalized to the flux at the peak pixel .
we then go through individual channels in our data cube , scale the image to the maximum value of the psf at the particular channel and subtract the image .
this routine provided the best residuals post psf subtraction .
some studies have additional steps with psf construction , by initially fitting and subtracting the nuclear continuum with a low order polynomial ( @xcite ) .
the purpose of the linear fit is to remove any continuum emission from the host galaxy . in our work ,
extensive studies of the final psf subtracted cube using both methods does not reveal a continuum emission from the host galaxy at the 3@xmath10 level ( average k mag @xmath26 20.9 ) , hence we decided not to include this additional step in our qso psf construction routine since it adds at least 1.2 times more noise in the psf subtracted cubes . to test the quality of our psf subtraction , we constructed radial profiles at different wavelength channels , before and after psf subtraction , to verify whether the final cube had the central core and seeing halo successfully removed .
figure [ psfsubresults ] shows the results of these tests for two of our targets .
the green and blue radial profile curves are constructed from a spectrally - summed image which contains both broad and narrow h@xmath2 emission ( @xmath272.5 nm or @xmath28v=1142kms@xmath5 ) .
the green curve is constructed from the data cube before psf subtraction , the blue is after the psf is removed and the red curve is constructed from just the psf image ( @xmath292.5 nm , spectrally offset 1 - 5 nm ) .
the points are constructed by taking an average in an annulus with @xmath28r=0.1 at a range of separations from the centroid of the qso .
the radial profile in the post psf subtracted data cube ( blue curves ) have little slope and significantly less flux , and do not strongly correlate with the general shape of the green and red curves .
this demonstrates that the psf subtracted data has a significant portion of the qso flux removed , with only the inner 0.2 being strongly dominated by noise from psf subtraction .
averaging over the data cube along the spectral axis , we find that generally within 0.2 the qso still contributes to about 10 - 20@xmath8 of the total data counts , while only 2 - 5@xmath8 outside 0.2 .
as expected , observations with the smallest psf fwhm showed the best post psf subtraction data cubes producing the best contrast .
however it should be noted that leftover qso continuum / blr light does not affect measured values derived from narrow line emission , since they are derived by fitting the line and any underline continuum left over from psf subtraction simultaneously , at which stage the continuum contamination can be calculated .
to find narrow line emission we searched all of the individual @xmath0 3,000 spaxels in each of the cubes using an algorithm that searches for flux above a predefined threshold , in combination with visual inspection of each cube .
when a line feature is identified we calculated the signal - to - noise by obtaining the standard deviation in the surrounding spatial and spectral pixels , and divided it into the fitted peak of the emission line . for cases where
a faint emission feature is found we bin the data using nearby spaxels to increase the signal - to - noise to distinguish between a faint noise spike versus real emission .
we confirm a detection if the peak of the emission line is greater than 3@xmath10 from the neighboring spaxels and the spectral width is greater than the intrinsic instrumental width of 0.35 nm and 0.20 nm for osiris and nifs respectively . for
bright noise spikes we wrote a routine that parses through the cube and removes them if their counts are 5@xmath10 or higher from the surrounding region ( one spaxel in each spatial direction , and 2 spectral channel two the left and right of the spike ) , some of these features have a fwhm greater than instrumental but given their spatial isolation and significantly higher counts than the surrounding region we quantify them as being `` noise spikes '' .
majority of them are associated with locations of oh sky lines , hence we believe these spikes are residuals caused by sky subtraction .
this routine also confirms faint extended structure in the case of sdssj0925@xmath300655 to be real rather than a combination of separated noise spikes . after searching through the five observed data cubes we identify narrow line emission in two of the systems , sdssj1029 + 6510 and sdssj0925 + 06 . for the given qso redshift ,
the identified emission lines are likely narrow h@xmath2 . if [ nii]6584 were assumed instead the flux ratio between it and undetected h@xmath2 would be @xmath1130 in some regions , this is well beyond what has been found in other galaxies ( e.g , @xcite ) .
once an h@xmath2 line is identified we searched for [ nii]6548,6584 and [ sii]6718,6733 at a similar velocity offset from the broad h@xmath2 line .
the detected narrow h@xmath2 emission lines all lie within 600 kms@xmath5 from of their respective qsos broad h@xmath2 redshift , however the full spectral axis in each spaxel was examined for potential narrow emission lines that could be associated with structure surrounding our qsos .
all of the line fits were done with a single gaussian function using the non - linear least squares routine provided through scipy .
the initial guess for the peak is the value at the location of the maximum flux , the initial guess on wavelength offset is the location of the maximum flux , and initial guess on @xmath10 was 80 kms@xmath5 , no further constraints were put on the parameters .
the radial velocity map is derived from the measured line offset in each spaxel relative to the redshift of the broad h@xmath2 line .
the velocity dispersion map of the gas is derived after removing the instrumental width in quadrature from @xmath10 at each spaxel .
velocity dispersion map is used to dictate the region over which the spectra need to be summed to derive total flux .
figure [ osiris_results ] ( panel i ) shows the k - band image of the sdssj1029 + 6510 qso from the collapsed data cube ( 1.99 - 2.4 @xmath31 ) .
figure [ osiris_results ] ( panel ii ) and figure [ osiris_results ] ( panel iii ) show the 2d kinematics of the extended narrow line emission relative to the broad h@xmath2 emission and the spectra of the individual components .
+ the psf subtracted data cube reveals three extended narrow line emission regions , labeled a , b and c in figure [ osiris_results ] ( panel ii ) .
these emission - line regions have a blue - shifted velocity offset of 10 - 500 kms@xmath5 with respect to broad h@xmath2 emission , and a maximum projected separation of @xmath0 0.6(4.2 kpc ) from the qso .
we bin the individual spaxels in regions a and c to detect a hint of h@xmath2 emission at a signal - to - noise of 3.1 and 2.1 , respectively .
individual spaxels in region b reach a signal - to - noise of @xmath112 , with the central 3 pixels reaching a signal to noise @xmath11 7 . in table
[ tabosiris ] we present the extracted emission - line properties of the individual regions . using [ n ii ] @xmath12
h@xmath2 we adopt the the line ratio separation between star formation and agn to be at log([nii]/h@xmath2)=-0.5 in the hii diagnostic or
bpt " @xcite diagram ( figure [ bpt ] ) .
the majority of the objects in the region log([nii]/h@xmath2)@xmath20 - 0.5 are star forming galaxies ( @xcite , @xcite , @xcite ) .
while low metallicity regions ionized by an agn can be a contaminant at these line ratios , all of the qsos in our sample ( particularly sdssj1029 + 6510 ) show strong uv emission lines in civ , siv+oiv and mgii ( figure [ sloan_spec ] ) that are typical of solar to super - solar metallicity qsos ; hence for this particular system we are not concerned about low metallicity contamination in the region log([nii]/h@xmath2)@xmath20 - 0.5 .
our limits allow us to discard shock contributions to the emission for regions a and b , line ratios of emission due to shocks tend to reside in log([nii]/h@xmath2)@xmath26 - 0.4 on the bpt diagram ( @xcite ) from a gas that is moving at the recorded velocities of our extended emission . based on the ratio of log([nii]/h@xmath2 ) for a
, this region can reside in the transition zone between agn / sf , assuming no extinction , the star formation rate limit for h@xmath2 flux in region a is [email protected] m@xmath4 yr@xmath5 using the schmidt - kennicutt law ( sfr@xmath32 , @xcite ) , this is a limit because agn photo ionization contribution will increase the observed flux , hence the star formation rate is lower than what is quoted .
region b is located well in the star formation position on the bpt ( log([nii]/h@xmath2)@xmath20 - 1.5 ) diagram with a star formation rate limit of [email protected]@xmath4 yr@xmath5 .
region c resides well inside the agn component of the diagram , and therefore is likely narrow line emission from the qso , at a projected radial distance of @xmath02.8 kpc .
figure [ nifs_results ] ( panel i ) is a k - band image of the qso constructed by summing the flux across the entire data cube ( 1.99 - 2.4 @xmath31 ) .
figure [ nifs_results ] ( panel ii and iii ) , show the 2d kinematics of the extended narrow line emission relative to the redshift of the broad h@xmath2 emission and the spectra of the individual components , respectively .
the post - psf subtracted data cube reveals resolved narrow h@xmath2 emission originating from three distinct regions ( a , b , and c ) , that are both spatially offset ( 0.5 - 1 ) and redshifted ( 80 - 250 kms@xmath5 ) from the qso , see table [ tabnifs ] for extracted parameters on individual regions .
we bin by [email protected] for each of these regions to increase the signal - to - noise for kinematic analysis . using ( [ n ii ] @xmath12 h@xmath2 ) ratio diagnostic
, we put the separation between star formation and agn at log([nii]/h@xmath2)=-0.5 , with star formation being the dominant photoionization mechanism in log([nii]/h@xmath2)@xmath20 - 0.5 ( see , section [ osirissdss ] for further discussion ) .
limits on the log([nii]/h@xmath2 ) ratio places regions a , b and c inside the star formation region on the bpt diagram .
our limits allow us to discard shock contributions to the emission for regions a , b , and c , line ratios of emission due to shocks tend to reside in log([nii]/h@xmath2)@xmath26 - 0.4 on the bpt diagram ( @xcite ) from a gas that is moving at the recorded velocities of our extended emission . using the schmidt - kennicutt law ( @xcite ) we obtain un - reddened upper limit star formation rates of [email protected] , [email protected] , [email protected] m@xmath4 yr@xmath5 for regions a , b and c , respectively . assuming these three clumps have virialized we obtain dynamical masses of 8.7 , 1.0
, 0.3 @xmath33m@xmath34 ( table [ tabnifs ] using the standard virial mass equation @xmath35 ) the remaining three targets reveal no narrow - line h@xmath2 emission offset spatially or spectrally from the qso .
null detections may be due to two possibilities : ( 1 ) these sources have heavy extinction azimuthally around the qso @xmath111 kpc ; and/or ( 2 ) these sources have sufficiently low star - formation rates that reside below the sensitivity limit of these observations .
we perform a monte - carlo simulation in which we generate star forming regions with narrow - line h@xmath2 emission surrounding the qso at various spatial separations .
the purpose of this simulation is to find the limiting flux ( and unreddened star - formation rate limits ) of our observations and determine how our psf removal techniques affect our sensitivity versus distance from these qsos . for our simulations , individual star - forming regions
occupy [email protected] in the osiris data cube and [email protected] in the nifs data cube , with each spaxel containing a spectrum consisting of an emission line resembling narrow h@xmath2 with a fixed full width at half maximum of 80 kms@xmath5(not convolved with an instrumental profile ) .
we select a fwhm of 80 kms@xmath5 to match the widths of some of our detected extended narrow line emission , to further test their validity . in a given data cube
the star forming regions have a spatially uniform flux , the integrated flux over all the simulated regions vary between cubes .
we insert these regions uniformly surrounding the qso in a cross shape to resemble resolved extended structure , which ranges from 0.1 to 1.5 in separation from the qso in the nifs data cubes and 0.1 - 0.7 in the osiris cubes .
the star forming regions are always centered on the quasar whose position we obtain by fitting a 2d gaussian to an image of a collapsed data cube along the spectral axis .
the spacing between the star forming regions is 0.1 to allow signal - to - noise estimates surrounding each individual region .
we vary the star formation rates from 0.5 m@xmath4 yr@xmath5 to 40 m@xmath4 yr@xmath5 in each of the narrow - line emission regions . for the osiris data , we insert the simulated star forming regions into a data cube that is created by running the extract spectra routine that simply transforms the two dimensional data into a 3d cube . for the gemini data we run the standard iraf reduction pipeline that extracts the 2d spectra and constructs the 3d data cubes , into which we insert the star forming regions .
we then process the data cubes through the rest of the reduction pipeline as described in [ data ] .
finally we run our psf subtraction routine on the reduced data cubes as described in
[ extract ] .
we attempt to recover each of the narrow - line h@xmath2 emission regions that were artificially inserted .
just as for the real data , emission must be detected with a minimum of 3@xmath10 confidence , and emission lines must have a fwmh greater than the instrumental width
. recovered star - forming regions with minimum star - formation rates at various angular separations are presented in figures [ osiris_limit ] and [ nifs_limit ] , and fluxes of h@xmath2 from sdssj0925 + 0655 and sdssj1029 + 6510 regions a , b , and c are over - plotted for comparison . in general we find that our data reduction procedure is not the main factor for missing narrow h@xmath2 flux ; the dominant effect is the sensitivity of the detector and psf removal within 0.2 from the qso . at separations @xmath36 ,
limiting star - formation rates are an average of 1.4 m@xmath4 yr@xmath5 ( 0.7@xmath37erg s@xmath5 cm@xmath7 ) integrated over a star - forming region for the nifs instrument and 1.5m@xmath4 yr@xmath5 for osiris .
this translates to 0.32m@xmath4 yr@xmath5 kpc@xmath7 and 0.53m@xmath4 yr@xmath5 kpc@xmath7 in the nifs and osiris data cubes respectively . for sdssj1029 + 6510 ,
we show the integrated flux of region b as well as its individual components in figure [ osiris_limit ] , and find they are detected without binning . these simulations and
the limiting fluxes for both of these sources indicate low h@xmath2 flux at near and far angular separations from the qso . for sdssj0925 + 0655 ,
fluxes of the observed components sit well above the star formation distribution ( figure [ nifs_limit ] ) , and in principle we are able to detect fainter emission at smaller separations .
the other three qsos do not show any signs of h@xmath2 narrow - line emission .
we use the bolometric luminosities of our sources to make estimates of dust extinction .
the bolometric luminosities of our sample all sit near 1@xmath39erg s@xmath5 cm@xmath7 , the maximum value for a z@xmath02 qso is around 1@xmath40erg s@xmath5 cm@xmath7 as has been found by studies such as @xcite .
this limit only allows us to correct for 2.5 magnitudes of extinction at 1450 , so the limiting star formation rates get as large as 2.03 m@xmath4 yr@xmath5(0.5m@xmath4 yr@xmath5 kpc@xmath7 ) or 2.2 m@xmath4 yr@xmath5(0.8m@xmath4 yr@xmath5 kpc@xmath7 , using a small magellanic cloud extinction curve from @xcite ) for nifs and osiris respectively ( see [ disc1029 ] , [ disc0925 ] ) .
note that for sdssj0925 + 0655 the limits may be higher as the qso is intrinsically redder than the rest of our sample ( see [ disc0925 ] ) .
we acknowledge that the dust in these scenarios is uniformly distributed , hence the same dust properties that we find along the line of sight to the qso are elsewhere in the galaxy .
most studies that quote star formation rates give them integrated over some angular scale , typically the beam size of their instrument if the sources they are referencing are not resolved . at the angular resolution of our observations we are capable of resolving a typical z@xmath02 galaxy with an angular scale of @xmath01 . integrating these limits over a 1 box we obtain for nifs : 22m@xmath4 yr@xmath5(33m@xmath4 yr@xmath5 with maximum dust extinction ) , osiris : 37m@xmath4 yr@xmath5(54m@xmath4 yr@xmath5 with maximum dust extinction ) .
these limits are a sum of the lowest flux that we detected around the qsos in a 1@xmath41 box in our simulations with the addition of possible dust obscuration .
we believe these are hard limits on the upper value of the star formation rate in these host galaxies .
derived sfr limits include contamination from dust in the agn .
there is a possibly that most of the dust is surrounding the nuclear region rather than distributed in the host galaxy .
archival wise photometry of our sources ( table [ tabphotoir ] ) shows that 3 of our sources are detected at 22(rest frame @xmath0 7 ) , all of the sources are detected in the other 3 wise bands that range from 1 - 3.75 at an average redshift of z=2.2 , however only 3 sources have reliable photometry due to confusion of flux from the nearby bright tip / tilt stars . for the sources that were detected at an observed wavelength of 22 we find that the average flux density is 16.8mjy indicating that the dust is agn heated ( @xcite ) .
limits closer to the value with minimum dust ( 22m@xmath4 yr@xmath5 for nifs and 37m@xmath4 yr@xmath5 for osiris ) may be more realistic , as some previous studies of dust in type-1 luminous qsos near z@xmath02 have found a number of sources with very little ( [email protected] ) to no extinction ( see , @xcite ) .
examining qso spectra extracted over the psf halo ( figure [ sloan_spec ] , right side ) we do not detected any unresolved narrow line region emission in any of our sources .
we find that generally the spectra are well fitted with a single gaussian profile and inclusion of narrow emission is only required for the case of sdssj1029 + 65 due to narrow h@xmath2 emission associated with star formation within 0.2 of the qso .
we place a flux limit of 3 - 4@xmath37 erg s@xmath5 cm@xmath7 which converts to 1 - 1.5@xmath43 erg s@xmath5 , assuming the nlr emission line has a fwhm of 80 kms@xmath5 .
there are two explanations for the null narrow - line h@xmath2 emission detections for three of the sources in our sample .
this could be caused simply by the lack of star formation and/or significant extinction in the host galaxy .
we argue that the main reason we do not see a significant amount of narrow h@xmath2 is likely due to the lack of star formation rather than extinction .
multi - wavelength observations can help estimate the amount of obscuration that is present in the galaxy due to dust . using available multi - wavelength data
we find that our sources do not contain sufficient amounts of dust to cause the observed h@xmath2 limits .
the qsos in our sample are all luminous type-1 agn representing some of the most powerful qsos at z@xmath44 . as we will argue in the following sections , even a small dust correction to these systems will increase the bolometric luminosities of our objects above the observed values at this redshift .
this indicates that the majority of qsos in our sample are hosted inside galaxies that are either transitioning from star forming to quenched galaxies or already reside in quiescent galaxies .
the host galaxy of this object shows compact vigorous star formation within 2 kpc from the qso .
the rest of the galaxy seems to show no narrow h@xmath2 which we attribute to low star formation rates .
sdssj1029 + 6510 is the second most powerful qso in our sample with a bolometric luminosity of 1.39@xmath45erg s@xmath5(table-[tabobs ] ) , in addition to the second longest observation time in our sample .
note that some of the emission in individual spaxels of region b are at the 3@xmath10 level , near the limit of our observations .
the ratio of log([nii]/h@xmath2 ) @xmath46 is located in the hii star formation portion of the diagram ( figure [ bpt ] ) for region b making it a strong candidate for star formation with a formation rate of [email protected]@xmath4 yr@xmath5 .
this indicates rapid star formation within 2kpc of the qso . for region
c , a ratio of [email protected] for log([nii]/h@xmath2 ) puts this source partly in the agn ionization region of the diagram ( figure [ bpt ] ) , and detection of [ nii ] emission with higher signal to noise than h@xmath2 suggests this emission is due to the agn .
lastly for region a , the measured ratio of log([nii]/h@xmath2)=@xmath47 places it partially inside the star - formation region on the diagram .
this source has a lack of extended star - forming regions , with 90% of the star formation activity within 2 kpc from the qso .
this is in stark contrast to other resolved host galaxies in @xcite , @xcite and @xcite , which have extended star forming regions over several kiloparsecs with star formation rates of @xmath0100 m@xmath4 yr@xmath5 .
our limiting flux simulations indicate that we should detect star formation rates as low as 1.4 m@xmath4 yr@xmath5 or down to a flux level of 0.6 - 0.8@xmath48erg s@xmath5 cm@xmath7 , at separations @xmath260.2 from the qso . instead , we detect two `` streams '' ( region b at snr@xmath263 ) of narrow h@xmath2 and nothing else significant around it ( regions a and c are @xmath49 ) .
this indicates that the surrounding ( @xmath262 kpc ) regions have narrow h@xmath2 flux that is below the sensitivity of the instrument .
dust can cause extinction of h@xmath2 flux by re - radiating it at longer wavelength .
qsos in early stages of evolution are thought to be heavily obscured .
after the agn inputs energy / momentum during the blow out " phase , gas and dust can get pushed out allowing the agn & galaxy to be detected in the optical , which otherwise would be obscured .
observations at other wavelengths can provide clues about the level of obscuration .
a strong detection in the far - ir can indicate dust heating due to uv radiation from recent birth of massive stars .
this would indicate that some portion of the uv radiation is absorbed ( suppressed ) and re - emitted at longer wavelength .
qsos that show reddening in their rest - frame uv spectra are good candidates for systems with a considerable level of obscuration , including a number of systems with indicators of outflows through blue shifted broad absorption lines in their rest - frame uv spectra , or broad blue - shifted components in the 500.7 nm [ oiii ] emission line , indicating that some of these systems might be in the blow - out " stage ( @xcite ) . for the case of sdssj1029 + 6510 we are able to put some constraints on the level of obscuration from both far - ir photometry and rest - frame uv - spectrum .
this qso was observed as part of a program with the herschel space telescope to target some of the brightest optical qsos with the spire instrument .
examining the archival data we find that at the optical position of the qso nothing is detected above 3@xmath10 level in the 250 , 350 , and 500 bands .
the flux density limits are ( @xmath010mjy , see table [ tabphotoir ] ) , indicating that this qso s host galaxy is not in a star - burst phase ( l@xmath50 ) .
the rest frame uv spectrum obtained from sdss shows ( figure [ sloan_spec ] ) a continuum slope typical of a type 1 un - obscured qso ( steep blue continuum ) , and a bolometric luminosity of 1.39@xmath51erg s@xmath5 ( table-[tabprop ] ) , which is about an order of magnitude above the average qso bolometric luminosity .
any correction for dust will start pushing the bolometric luminosity beyond the typical value for bright qsos at z@xmath44 ( @xmath52erg s@xmath5 ) .
assuming we need to correct an order of magnitude of flux at rest frame wavelength of 1450 due to dust we would only push the limiting star formation rate to 0.7 m@xmath4 yr@xmath5 kpc@xmath7 ( using a small magellanic cloud extinction curve from @xcite ) , not sufficient to explain the lack of h@xmath2 flux .
we therefore favor the low star formation rate model as the main explanation for the observed h@xmath2 flux in the case of sdssj1029 + 6510 at separations
greater than 2 kpc .
the extended h@xmath2 emission surrounding sdssj0925 + 0655 is a strong candidate for active star formation .
the ratio of log([nii]/h@xmath2 ) for region a is within the star formation region on the diagram ( figure [ bpt ] ) while our limits on regions b and c place them near the ambiguous regions between star formation and agn . the total flux from all
these implies an integrated star formation rate of 29@xmath53 m@xmath4 yr@xmath5 .
the detected narrow h@xmath2 emission regions are compact ( @xmath02kpc ) and we only detect narrow h@xmath2 in these three regions . in other regions of the data cube we are able to reach a sensitivity limit of 0.8@xmath37erg s@xmath5 or a star formation rate of 1.4 m@xmath4 yr@xmath5 at separations @xmath11 0.2 from the qso .
all of the detected regions are at separations @xmath11 0.5 ( 4 kpc ) .
this implies the narrow h@xmath2 flux sits below the sensitivity of the detector at separations between 1.4 to 4 kpc .
we propose that the primary reason for lack of h@xmath2 flux is either from star formation halting , or from obscuration due to dust in the host galaxy ( as introduced in the [ disc1029 ] ) .
the bolometric luminosity ( 5.9@xmath54erg s@xmath5 ) of this qso as calculated from the 1450 continuum is about an order of magnitude below the average value of a qso at this redshift , due to the continuum being heavily reddened .
however the broad h@xmath2 emission of this source agrees with the rest of the objects in our sample ( similar equivalent width and luminosity , see table [ tabprop ] ) that do not show any signs of reddening in their rest - frame uv spectra ( see figure [ sloan_spec ] and table [ tabpropha ] ) .
the average bolometric luminosities of our sample is @xmath55erg s@xmath5(see table [ tabprop ] ) .
the agreement between broad line h@xmath2 properties ( velocity dispersion & intensity ) hints that the bolometric luminosity should be consistent with other members of our sample .
as found in @xcite most reddened qsos are red due to dust in their host galaxies rather than the inter - galactic medium or dust inside the milky way . for this source
we estimate the amount of reddening by invoking the condition that the bolometric luminosity should be at the average value for a qso with such a strong broad h@xmath2 emission ( at least @xmath56erg s@xmath5 ) .
this implies that the flux at 1450 needs to be boosted by @xmath57 implying that a@xmath58 . using extinction curve from @xcite assuming small magellanic cloud ( smc ) like extinction ( r@xmath59=2.74 ) we obtain a@xmath60 .
this implies that the flux at h@xmath2 needs to be corrected by at least @xmath61 , yielding a de - reddened star formation rate limit of 0.45 m@xmath4 yr@xmath5 kpc@xmath7 , and the combined de - reddened star formation rate on a , b , and c of 41 m@xmath4 yr@xmath5 .
this implies that dust attenuation only removes 0.1 m@xmath4 yr@xmath5 kpc@xmath7 if we only correct the bolometric luminosity such that it sits at the average .
overall this level of dust obscuration is not enough to be the primary reason for low h@xmath2 flux . even assuming an extreme case where the bolometric luminosity is near the maximum value for a type-1 qso at z@xmath02 ( @xmath52 )
would only imply a limit of 0.9m@xmath4 yr@xmath5 kpc@xmath7 .
this could imply that there is a low star formation rate in the host galaxy , where the star formation has been nearly shut off within [email protected](1.4 - 4 kpc ) from the qso .
these distant regions ( a , b and c ) are still forming stars at rates that are detectable .
our observations indicate that the host could be in a process of transitioning from a star - forming into a quiescent galaxy .
however the less unlikely possibility is that the star formation is active in a diffuse region at separations of 1.4 - 4kpc rather than in the clumpy regions that we see in regions a , b , c and in other star forming galaxies at this redshift .
there have been a number of multi - wavelength surveys of radio quiet type 1 qsos at z@xmath02 that have presented a range of conclusions about host galaxy star formation properties .
high redshift qso studies have either implied high star formation rates in concurrent high - z type - i qsos or have argued for a lack of star formation activity . in this section
we summarize and compare surveys that share similar qso properties to our sample ( i.e. , smbh mass , bolometric luminosity , unobscured type 1 ) .
herschel pacs observations of agn and qsos in the cosmos extragalactic survey indicate a correlation between their bolometric luminosity and rest - frame 60 host galaxy emission ( @xcite ) . using the mean 60 flux ( 3.4@xmath54erg s@xmath5 ) in the @xmath63erg s@xmath5 z=1.5 - 2.2 bin in table 1 from @xcite
indicate that the mean star formation rate should be of order 200 m@xmath4 yr@xmath5 , using the 70 star formation rate law presented in @xcite .
this is nearly an order of magnitude greater than the mean star formation rate in our sample , as indicated by narrow h@xmath2 emission line detection ( 78m@xmath4 yr@xmath5 and 29m@xmath4 yr@xmath5 ) and limits ( 22m@xmath4 yr@xmath5 for nifs and 37m@xmath4 yr@xmath5for osiris , integrated over a 1@xmath41 box .
see section [ null ] for the discussion ) .
the disagreement between our sample and the herschel results could be due to just the limited - number of sources observed ( 14 in @xcite at a similar bolometric luminosity ( @xmath64erg s@xmath5 ) as the 5 qsos in our sample ) .
it is worth noting that the qsos may be responsible for a significant portion of the total 60 luminosity , so derived 60 star formation rates should be considered as upper limits .
hst observations of radio quiet qsos at z@xmath02 in @xcite indicate an average star formation rate of 100m@xmath4 yr@xmath5 derived from rest - frame uv emission originating from the host galaxy . in their study
they use both stellar and artificial psfs to remove the bright qso .
the number of qsos in our sample is similar to @xcite , which are type-1 and radio quiet .
the star formation rate differences between our sample and @xcite could be due to strong qso contamination from residual emission from their psf subtraction , or that star formation in our hosts are quite diffuse .
in contrast , studies such as @xcite and @xcite find quiescent galaxies that host radio quiet high - z qsos .
these observations are from seeing - limited ( 0.4 - 0.5 ) near - infrared imaging and are limited to disentangling the host galaxy at close angular scales ( @xmath6 4 kpc ) . sdssj0925 + 0655 and sdssj0850 + 5843 share similar rest frame uv photometry to their samples
, however the other half of the qsos in our study are 1 to 1.5 magnitudes brighter . including our results with these two other papers only yields a total of 15 high - z qso that are observed to reside in quiescent " z@xmath02 galaxies in current literature . at even higher redshifts , recent alma observations of z@xmath06 qsos @xcite using
the 158 [ cii ] emission line reveals a detection in nearly 90@xmath8 of the sources observed .
the targets in their samples have similar properties to ours ( i.e. , bh mass , bolometric luminosities @xmath12 eddington ratios ) .
in @xcite they reach a star formation limit of 40 m@xmath4 yr@xmath5 assuming the [ cii ] emission emanates solely from star formation . yet
sources in @xcite reach star formation rates as high as 1000 m@xmath4 yr@xmath5 , which implies that sources with detected [ cii ] have extreme star formation rates in comparison to our detections and sensitivity limits at z=2 .
these z@xmath06 sources are all near the peak of their starburst phase , assuming that most of the [ cii ] emission originates from star formation and not the qso .
according to present day m@xmath65-m@xmath9 relation and theoretical work ( e.g. , @xcite ) there is an expectation of simultaneous smbh and galaxy growth , presumably via mergers at these high ( @xmath66erg s@xmath5 cm@xmath7 ) bolometric luminosities ( @xcite ) .
in contrast , our observations show star formation rates that are well below this expected initial burst and below the typical star - forming galaxies at z@xmath02 @xcite . the essential difference and advantage of our study compared to previous studies , is that our detection and limits of star formation rates can be made at differing spatial and velocity locations away from the qso .
in contrast , the majority of all studies we have discussed have integrated star formation rate limits over a large range of psf and beam sizes . based on our detection limits ,
it is clear that we do not detect the clumpy ( 1 kpc@xmath41 ) , strong star formation regions ( up to @xmath010 m@xmath4 yr@xmath5 kpc@xmath7 ) in current ifs observed z@xmath02 star forming galaxies @xcite . if there is underlying star formation undetected in these host systems , then the surface brightness profiles of the star formation has to be diffuse and integrated across a large area of the galaxy .
if our limits are to match previous inferred star formation rates of z@xmath02 qso hosts , then it would need to be diffuse with significant extinction .
the sample selection in our pilot survey is albeit random , since we were selecting based on achieving the best ao performance for psf subtraction , therefore it is interesting that we would happen to select 3/5 type - i qsos that are quiescent .
the majority of our sample is similar to only a small number of observations of high - z qso hosts residing in quiescent galaxies , and are in disagreement with other work that indicate simultaneous high star formation rates and agn activity .
qso duty cycles are still poorly understood , however it does seem to appear that in a number of cases the qso can still be active while star formation in the host has been effectively turned off .
these results agree well with agn feedback models that require that the feedback mechanism only carry a small portion of the total bolometric luminosity of the qso ( 5 - 10@xmath8 ) to effectively turn off star formation @xcite . on the other hand
this also agrees with non - causal evolution of smbh and their host galaxies ( @xcite ) , where the growth of the smbh and star formation are unrelated and agn feedback is not the main constituent in formation of local scaling relations , possibly because agn and star formation activity happen on different time scales .
our study , @xcite and @xcite are consistent with star formation time scales being significantly shorter than that of the qso .
there are likely numerous high angular resolution observations from hst and ground - based observations that have had null detections of high - redshift qso host galaxies , that would benefit being released to the community to improve these global statistics .
interestingly , this means there is likely a social selection bias of high - z qso host galaxies , where authors typically only publish detections ( hence qso hosts with higher star formation properties ) rather than their null detections . in any case
, it is obvious that there are a large number of selection effects that need to be taken account , but clearly a larger sample of high - redshift qsos would greatly benefit from ifs+ao observations and aid in our understanding of the demographics of high - z qso host galaxies .
we have presented lgs - ao assisted integral field spectroscopy observations of five z = 2 qsos targeted at resolving h@xmath2 nebular emission lines from their host galaxies . using the broad emission line region of the qso
we were able to construct a psf to remove the qso continuum and emission to achieve the necessary contrast to detect h@xmath2 and [ nii ] host galaxy emission ( see [ extract ] ) .
* for two out of five sources ( sdssj1029 + 6510 @xmath12 sdssj0925 + 0655 ) we are able to resolve extended narrow line emission surrounding the qso . * in sdssj1029 + 6510 we detect narrow h@xmath2 ( regions a and b ) that likely originates from star formation at close separations ( 2 - 4 kpc ) from the qso .
if we assume the h@xmath2 flux is from star formation the integrated star formation rate from region a and b is [email protected] m@xmath4 yr@xmath5(110 m@xmath4 yr@xmath5 with dust correction ) . * for sdssj0925 + 06 we detect three distinct star forming regions that are separated from the qso by @xmath0 4 kpc .
the upper limit star formation rate for all three regions combined is [email protected] m@xmath4 yr@xmath5(40.7 m@xmath4 yr@xmath5 with dust corrections ) .
* careful examination of the other three sources in our sample do not detect any narrow h@xmath2 emission post psf subtraction , even in the cases of sdssj1005 + 4356 @xmath12 sdssj2123 - 0050 for which we spent the most integration time per source .
* we ran a monte carlo simulation on our data by inserting extended narrow h@xmath2 at various separations from the qso with varying h@xmath2 fluxes ( star formation rates ) .
we find that we can detect star formation rates down to 1.4 m@xmath4 yr@xmath5 ( see [ null ] ) as close as 0.2 from the qso .
incorporating dust obscuration this value can vary from 2.6m@xmath4 yr@xmath5 - 9m@xmath4 yr@xmath5 ( see [ disc1029 ] & [ disc0925 ] ) depending on the value of a@xmath67 . at the 9m@xmath4 yr@xmath5 limit , after correcting the sdss spectra for dust reddening we are pushing the bolometric luminosities for some of our sources past the typical values for type 1 qsos at this redshift .
even with a star formation rate of 9 m@xmath4 yr@xmath5 it would be difficult to explain the missing narrow h@xmath2 to be due to dust obscuration inside the host galaxy .
hence for these sources low star formation rate is the likely reason for lack of narrow h@xmath2 originating from the host galaxy . *
four sources show low star formation rates at close angular separation of the qso , with no dereddened star formation @xmath11 9 m@xmath4 yr@xmath5 within 2 to 4 kpc of the qso .
* we do not detect any strong evidence for nlr emission ( region c of sdssj1029 + 6510 is only 2.1@xmath10 ) in any of our sources .
we place a luminosity limit of 1 - 1.5@xmath43erg s@xmath5 cm@xmath7 on an emission line originating from the qso s nlr . *
compared to other z=2 qso host galaxy surveys our sample is unique by having little - to - no star formation in high redshift type - i qsos .
this is in agreement with a large fraction of nearby ( z @xmath6 0.5 ) qso host galaxies being quiescent . yet at comparable and higher redshifts to our sample the majority of surveys have found simultaneous star formation activity with qso activity .
clearly a larger z=1 - 3 qso ifs+ao sample will be critical in developing a more coherent picture of qso host galaxies during this important epoch . based on observations obtained at the gemini observatory , which is operated by the association of universities for research in astronomy , inc .
, under a cooperative agreement with the nsf on behalf of the gemini partnership : the national science foundation ( united states ) , the national research council ( canada ) , conicyt ( chile ) , the australian research council ( australia ) , ministrio da cincia , tecnologiae inovao ( brazil ) and ministerio de ciencia , tecnologa e innovacin productiva ( argentina ) .
the authors would like to give our thanks to eric steinbring who served as our gemini phase ii liaison for planning this program ( gn-2012b - q-53 ) .
data was also obtained at w.m .
keck observatory , which was made possible by generous financial support from the w.m . keck foundation .
the authors would like to acknowledge the dedicated members of the keck observatory staff , particularly jim lyke and randy campbell , who helped with the success of our observations .
the authors wish to recognize the significant cultural role and reverence that the summit of mauna kea has always had within the indigenous hawaiian community .
we are most fortunate to have the opportunity to conduct observations from this heiau " mountain .
scg thanks the natural science and engineering research council of canada for support .
alexander , d. m. , swinbank , a. m. , smail , i. , mcdermid , r. , & nesvadba , n. p. h. 2010 , , 402 , 2211 alexander , d. m. , & hickox , r. c. 2012 , , 56 , 93 allen , m. g. , groves , b. a. , dopita , m. a. , sutherland , r. s. , & kewley , l. j. 2008 , , 178 , 20 bahcall , j. n. , kirhakos , s. , saxe , d. h. , & schneider , d. p. 1997
, , 479 , 642 baldwin , j. a. , phillips , m. m. , & terlevich , r. 1981 , , 93 , 5 barai , p. , viel , m. , murante , g. , gaspari , m. , & borgani , s. 2014 , , 437 , 1456 becker , r. h. , white , r. l. , & helfand , d. j. 1995 , , 450 , 559 bennert , n. , canalizo , g. , jungwiert , b. , et al . 2008 , , 677 , 846 calzetti , d. , wu , s .- y . , hong , s. , et al . 2010 , , 714 , 1256 cano - daz , m. , maiolino , r. , marconi , a. , et al .
2012 , , 537 , l8 cardelli , j. a. , clayton , g. c. , & mathis , j. s. 1989 , , 345 , 245 carniani , s. , marconi , a. , biggs , a. , et al .
2013 , , 559 , a29 croom , s. m. , richards , g. t. , shanks , t. , et al .
2009 , , 399 , 1755 cutri , r. m. , skrutskie , m. f. , van dyk , s. , et al . 2003 ,
the irsa 2mass all - sky point source catalog , nasa / ipac infrared science archive davies , r. i. 2007 , , 375 , 1099 di matteo , t. , springel , v. , & hernquist , l. 2005 , , 433 , 604 erb , d. k. , steidel , c. c. , shapley , a. e. , et al .
2006 , , 647 , 128 fabian , a. c. 2012 , , 50 , 455 falomo , r. , kotilainen , j. k. , pagani , c. , scarpa , r. , & treves , a. 2004 , , 604 , 495 falomo , r. , kotilainen , j. k. , scarpa , . , & treves , a. 2005 , , 434 , 469 farrah , d. , urrutia , t. , lacy , m. , et al .
2012 , , 745 , 178 ferrarese , l. , & merritt , d. 2000 , , 539 , l9 floyd , d. j. e. , axon , d. , baum , s. , et al .
2010 , , 713 , 66 floyd , d. j. e. , dunlop , j. s. , kukula , m. j. , et al .
2013 , , 429 , 2 frster schreiber , n. m. , genzel , r. , bouch , n. , et al .
2009 , , 706 , 1364 fynbo , j. p. u. , krogager , j .- k . ,
venemans , b. , et al .
2013 , , 204 , 6 gebhardt , k. , bender , r. , bower , g. , et al .
2000 , , 539 , l13 genzel , r. , newman , s. , jones , t. , et al .
2011 , , 733 , 101 gordon , k. d. , clayton , g. c. , misselt , k. a. , landolt , a. u. , & wolff , m. j. 2003 , , 594 , 279 greene , j. e. , & ho , l. c. 2005 , , 630 , 122 groves , b. a. , heckman , t. m. , & kauffmann , g. 2006 , , 371 , 1559 harrison , c. m. , alexander , d. m. , mullaney , j. r. , & swinbank , a. m. 2014 , arxiv:1403.3086 hopkins , p. f. , & quataert , e. 2011 , , 415 , 1027 hopkins , p. f. , & elvis , m. 2010 , , 401 , 7 jahnke , k. , & macci , a. v. 2011 , , 734 , 92 liu , g. , zakamska , n. l. , greene , j. e. , nesvadba , n. p. h. , & liu , x. 2013 , , 436 , 2576 hamilton , t. s. , casertano , s. , & turnshek , d. a. 2002 , , 576 , 61 hutchings , j. b. , frenette , d. , hanisch , r. , et al .
2002 , , 123 , 2936 inskip , k. j. , jahnke , k. , rix , h .- w .
, & van de ven , g. 2011 , , 739 , 90 kornei , k. a.,shapley , a. e. , martin , c. l. , et al .
2012 , , 758 , 135 kotilainen , j. k. , falomo , r. , decarli , r. , et al .
2009 , , 703 , 1663 kauffmann , g. , heckman , t. m. , tremonti , c. , et al .
2003 , , 346 , 1055 kennicutt , r. c. , jr .
1998 , , 498 , 541 kewley , l. j. , dopita , m. a. , sutherland , r. s. , heisler , c. a. , & trevena , j. 2001 , , 556 , 121 kirhakos , s. , bahcall , j. n. , schneider , d. p. , & kristian , j. 1999 , , 520 , 67 kormendy , j. , & ho , l. c. 2013 , , 51 , 511 larkin , j. , barczys , m. , krabbe , a. , et al .
2006 , , 50 , 362 law , d. r. , steidel , c. c. , erb , d. k. , et al .
2009 , , 697 , 2057 law , d. r. , shapley , a. e. , steidel , c. c. , et al .
2012 , , 487 , 338 lehnert , m. d. , van breugel , w. j. m. , heckman , t. m. , & miley , g. k. 1999 , , 124 , 11 magorrian , j. , tremaine , s. , richstone , d. , et al .
1998 , , 115 , 2285 mrquez , i. , & petitjean , p. 2003 , revista mexicana de astronomia y astrofisica conference series , 16 , 135 matsuoka , y. , strauss , m. a. , price , t. n. , iii , & didonato , m. s. 2014 , , 780 , 162 mcconnell , n. j. , & ma , c .- p .
2013 , , 764 , 184 mcgregor , p. j. , hart , j. , conroy , p. g. , et al .
2003 , , 4841 , 1581 odonnell , j. e. 1994 , , 422 , 158 peng , c. y. 2007 , , 671 , 1098 pettini , m. , & pagel , b. e. j. 2004 , , 348 , l59 planck collaboration , ade , p. a. r. , aghanim , n. , et al . 2014 , , 571 , aa16 quinlan , g. d. , & hernquist , l. 1997 , , 2 , 533 somerville , r. s. , hopkins , p. f. , cox , t. j. , robertson , b. e. , & hernquist , l. 2008 , , 391 , 481 ridgway , s. , heckman , t. , calzetti , d. , & lehnert , m. 2002 , , 46 , 175 riechers , d. a. , walter , f. , carilli , c. l. , bertoldi , f. , & momjian , e. 2008 , , 686 , l9 rosario , d. j. , trakhtenbrot , b. , lutz , d. , et al .
2013 , , 560 , a72 rowan - robinson , m. 1995 , , 272 , 737 runnoe , j. c. , brotherton , m. s. , & shang , z. 2012 , , 422 , 478 sesana , a. , haardt , f. , & madau , p. 2008
, , 686 , 432 scannapieco , e. , silk , j. , & bouwens , r. 2005 , , 635 , l13 schaerer , d. , boone , f. , zamojski , m. , et al .
2014 , arxiv:1407.5793 schneider , d. p. , richards , g. t. , hall , p. b. , et al .
2010 , , 139 , 2360 schramm , m. , wisotzki , l. , & jahnke , k. 2008 , , 478 , 311 steidel , c. c. , rudie , g. c. , strom , a. l. , et al .
2014 , arxiv:1405.5473 thompson , t. a. , quataert , e. , & murray , n. 2005 , , 630 , 167 treister , e. , schawinski , k. , urry , c. m. , & simmons , b. d. 2012 , , 758 , l39 urrutia , t. , lacy , m. , spoon , h. , et al .
2012 , , 757 , 125 veilleux , s. , kim , d .-
c . , rupke , d. s. n. , et al .
2009 , , 701 , 587 vestergaard , m. , & peterson , b. m. 2006 , , 641 , 689 villforth , c. , heidt , j. , & nilsson , k. 2008 , , 488 , 133 wang , r. , wagg , j. , carilli , c. l. , et al .
2013 , , 773 , 44 wang , j .-
g . , dong , x .- b . ,
wang , t .-
, et al .
2009 , , 707 , 1334 willott , c. j. , omont , a. , & bergeron , j. 2013 , , 770 , 13 wurster , j. , & thacker , r. j. 2013 , , 431 , 2513 zakamska , n. l. , strauss , m. a. , krolik , j. h. , et al .
2006 , , 132 , 1496 lcccccccc qso & r.a & dec & observation & integration & psf fwhm & tip / tilt & tip / tilt + sdssj & j2000 & j2000 & date & n@xmath68 @xmath22s & & sep ( ) & r [ mag ] + 085022.63 + 584315.0 & 08:50:22.63 & 58:43:15 & 2013jan04 & 4x600 & 0.15 & 14.1&10.4 + 092547.47 + 065538.9 & 09:25:47.47 & 6:55:38.9 & 2013jan08 & 6x600 & 0.13 & 6.2&13.0 + 100517.43 + 434609.3 & 10:05:17.50 & 43:46:10.9 & 2011dec31 & 16x300 & 0.202 & 21.5 & 13.1 + 102907.09 + 651024.6 & 10:29:07.09 & 65:10:24.6 & 2011dec30 & 12x300 & 0.177 & 43.8&13.8 + 212329.46 - 005052.9 & 21:23:29.46 & -00:50:52.9 & 2012aug06 & 6x600 & 0.11 & 0&16.4 + llllllllll qso & u & g & r & i & z&j&h&k + sdssj0850 + 5843 & 19.301 & 19.071 & 18.939 & 18.977 & 18.619&&& & + sdssj0925 + 0655 & 21.583 & 20.745 & 19.888 & 19.493 & 19.197&&& & + sdssj1005 + 4346 & 16.985 & 16.803 & 16.631 & 16.469 & 16.249&15.474&15.041&14.271 & + sdssj1029 + 6510 & 17.150 & 16.938 & 16.833 & 16.757&16.602&15.881&15.413&14.566 & + sdssj2123 - 0050&17.194 & 16.648 & 16.434 & 16.338&16.121&15.180&14.616&13.904 & + lllllllll qso & w1 3.4 & w2 4.6 & w3 12 & w4 22&spire 250&spire 350&spire 500 & + sdssj & mjy&mjy&mjy&mjy&mjy&mjy&mjy & + 0850 + 5843 & &&&@xmath206.8&&& & + 0925 + 0655 & &&&&&& & + 1005 + 4346 & 1.32 & 2.21 & 8.01 & 13.83 & && & + 1029 + 6510 & 0.87 & 1.49 & 6.82 & 12.91&@xmath2021.8&@xmath2013.3&@xmath2016.5 & + 2123 - 0050&1.48 & 2.37 & 9.82 & 23.85&@xmath2013.8&@xmath208.67&@xmath2011.6 & + lcccc qso & z@xmath69 & l@xmath23 & m@xmath13 & eddington + & & erg s@xmath5@xmath39 & m@xmath70@xmath33 & ratio + sdss0850@xmath305843 & 2.211 & [email protected] & [email protected] & 0.098 + sdss0925@xmath300655 & 2.197 & [email protected] & & + sdss1005@xmath304346 & 2.086 & [email protected] & [email protected] & 0.14 + sdss1029@xmath306510 & 2.163 & [email protected] & [email protected] & 0.14 + sdss2123@xmath620050 & 2.261 & [email protected] & [email protected] & 0.24 + column 3 is the bolometric luminosities ( l@xmath23=4.2@xmath71 ) obtained from rest frame 1450 continuum with corrections from @xcite .
column 4 is the black hole mass obtained from rest frame 1549 civ emission line ( @xcite ) .
column 5 is the ratio of the bolometric luminosity to the eddington luminosity obtained from the measured black hole mass .
lcccc qso & z@xmath3 & l@xmath3 & m@xmath13 & equivalent + & & erg s@xmath5 & m@xmath70@xmath33 & width ( ) + sdss0850@xmath305843 & 2.212 & 3.16@xmath72 & [email protected] & 384@xmath256 + sdss0925@xmath300655 & 2.196 & 3.58@xmath72 & [email protected] & 352@xmath251 + sdss1005@xmath304346 & 2.105 & 7.42@xmath72 & [email protected] & 230@xmath251 + sdss1029@xmath306510 & 2.183 & 3.12@xmath72 & [email protected] & 289@xmath252 + sdss2123@xmath620050 & 2.281 & 3.84@xmath54 & [email protected] & 281@xmath251 + cccccccc component & f@xmath3 & f@xmath73}$]6584 & [ nii]/h@xmath2 & sfr & v@xmath74 & v@xmath75 & m@xmath76 + & & & & m@xmath4 yr@xmath5 & kms@xmath5 & kms@xmath5 & + a&4.22 @xmath25 0.75&@xmath200.951&@xmath200.2310 & & -778@xmath2516 & 163@xmath2536 & + b&22.6 @xmath25 1.92&@xmath200.71&@xmath200.0319 & 67@xmath256 & -355@xmath2519 & 34@xmath2512 & [email protected] + c&4.14@xmath25 1.95&[email protected]&-0.24@xmath25 0.32 & & -39@xmath2542 & 36@xmath2540 & + cccccccc component & f@xmath3 & f@xmath73}$]6584 & [ nii]/h@xmath2 & sfr & v@xmath74 & v@xmath75 & m@xmath76 + & & & & m@xmath4 yr@xmath5 & kms@xmath5 & kms@xmath5 & + a&[email protected]&@xmath200.245&@xmath200.0565 & [email protected] & [email protected] & [email protected] & [email protected] + b&[email protected]&@xmath200.58&@xmath200.1410 & [email protected] & [email protected] & [email protected] & [email protected] + c&[email protected]&@xmath200.148&@xmath200.1222 & [email protected] & [email protected] & [email protected] & [email protected] + , civ , ciii ] , mgii .
near - ir spectra are presented on the right side , where the broad h@xmath2 line is present .
these were extracted from the data cubes , integrating over the spaxels within the seeing halo.,title="fig:",scaledwidth=60.0% ] , civ , ciii ] , mgii .
near - ir spectra are presented on the right side , where the broad h@xmath2 line is present .
these were extracted from the data cubes , integrating over the spaxels within the seeing halo.,title="fig:",scaledwidth=42.0% ] 0655(right ) .
green and blue radial profile curves are constructed from spectrally integrated images which contain both broad and narrow h@xmath2 , while the red curve is constructed from the psf image .
the blue curve is constructed in the same spectral regions as the green curve however post psf subtraction , indicating that our psf removal technique is capable of removing both the ao corrected core as well as the seeing halo.,title="fig:",scaledwidth=45.0% ] 0655(right ) .
green and blue radial profile curves are constructed from spectrally integrated images which contain both broad and narrow h@xmath2 , while the red curve is constructed from the psf image .
the blue curve is constructed in the same spectral regions as the green curve however post psf subtraction , indicating that our psf removal technique is capable of removing both the ao corrected core as well as the seeing halo.,title="fig:",scaledwidth=45.0% ] + ) of extended narrow h@xmath2 emission detected post - psf subtraction .
radial velocity measurements are obtained by fitting narrow h@xmath2 emission line in the individual regions with a gaussian function .
the spatial resolution of each observation is reprsented by the ellipse in the lower left corner obtained through 2d gaussian fitting to the psf image .
bottom : averaged per spaxel spectra of each of the labeled components with some relative flux offset . the light blue curve shows the wavelength dependence of the noise and oh sky emission . dashed red lines
represent the expected wavelength of narrow emission lines .
north is up , east is left.,title="fig:",scaledwidth=45.0%]i ) of extended narrow h@xmath2 emission detected post - psf subtraction .
radial velocity measurements are obtained by fitting narrow h@xmath2 emission line in the individual regions with a gaussian function .
the spatial resolution of each observation is reprsented by the ellipse in the lower left corner obtained through 2d gaussian fitting to the psf image .
bottom : averaged per spaxel spectra of each of the labeled components with some relative flux offset . the light blue curve shows the wavelength dependence of the noise and oh sky emission . dashed red lines
represent the expected wavelength of narrow emission lines .
north is up , east is left.,title="fig:",scaledwidth=45.0%]ii + ) of extended narrow h@xmath2 emission detected post - psf subtraction .
radial velocity measurements are obtained by fitting narrow h@xmath2 emission line in the individual regions with a gaussian function .
the spatial resolution of each observation is reprsented by the ellipse in the lower left corner obtained through 2d gaussian fitting to the psf image .
bottom : averaged per spaxel spectra of each of the labeled components with some relative flux offset .
the light blue curve shows the wavelength dependence of the noise and oh sky emission .
dashed red lines represent the expected wavelength of narrow emission lines .
north is up , east is left.,title="fig:",scaledwidth=90.0%]iii 0655 from the collapsed nifs altair ao cube using 0.05 spatial sampling .
upper right : psf subtracted image showing resolved extended h@xmath2 narrow line emission in contours that stretch from 1.8@xmath10 - 5@xmath10 and the velocity map ( kms@xmath5 ) obtained from fitting the h@xmath2 line in the individual regions using a gaussian function .
the spatial resolution of each observation is represented by the aperture in the lower left corner obtained through 2d gaussian fitting to the psf image .
bottom : averaged per spaxel spectra of each of the labeled components with some relative flux offset . the light blue curve shows the wavelength dependence of the noise and oh sky . dashed red lines
represent the expected wavelength of narrow emission lines .
north is up , east is left.,title="fig:",scaledwidth=45.0%]i 0655 from the collapsed nifs altair ao cube using 0.05 spatial sampling .
upper right : psf subtracted image showing resolved extended h@xmath2 narrow line emission in contours that stretch from 1.8@xmath10 - 5@xmath10 and the velocity map ( kms@xmath5 ) obtained from fitting the h@xmath2 line in the individual regions using a gaussian function .
the spatial resolution of each observation is represented by the aperture in the lower left corner obtained through 2d gaussian fitting to the psf image .
bottom : averaged per spaxel spectra of each of the labeled components with some relative flux offset . the light blue curve shows the wavelength dependence of the noise and oh sky . dashed red lines
represent the expected wavelength of narrow emission lines .
north is up , east is left.,title="fig:",scaledwidth=45.0%]ii + 0655 from the collapsed nifs altair ao cube using 0.05 spatial sampling .
upper right : psf subtracted image showing resolved extended h@xmath2 narrow line emission in contours that stretch from 1.8@xmath10 - 5@xmath10 and the velocity map ( kms@xmath5 ) obtained from fitting the h@xmath2 line in the individual regions using a gaussian function . the spatial resolution of each observation is represented by the aperture in the lower left corner obtained through 2d gaussian fitting to the psf image .
bottom : averaged per spaxel spectra of each of the labeled components with some relative flux offset . the light blue curve shows the wavelength dependence of the noise and oh sky . dashed red lines
represent the expected wavelength of narrow emission lines .
north is up , east is left.,title="fig:",scaledwidth=90.0%]iii 0.2region that was recovered at various separations from the qso in our monte carlo simulation from the osiris observations of sdssj1029 + 6510 ( green ) .
fluxes and distribution of features a , b , c ( light red ) from figure [ osiris_results ] are over plotted .
in addition , flux from individual spaxels of region b are plotted in dark red .
the three inner spaxels surpass the [email protected] box flux limit , while integration of the additional 4 outer spaxels builds a spectrum with a signal - to - noise that is significantly above the noise floor .
note : although the spatial size of region b is greater than [email protected] , 90% of the flux sits in the central 3 spaxels , who individually contain a signal to noise ratio @xmath26 3.,title="fig:",scaledwidth=70.0% ] + 0.25region that was recovered at various separations from the qso in our monte carlo simulations from the nifs observations of sdssj0925 + 0655 ( green ) .
integrated fluxes of features a , b , c ( red ) from figure [ nifs_results ] are over - plotted.,title="fig:",scaledwidth=70.0% ] + /h@xmath79 ) and log([nii]/h@xmath2 ) for sdssj1029 + 6510 regions a , b @xmath12 c ( top ) and regions a , b , @xmath12 c for sdssj0925 + 0655 ( bottom ) are over - plotted . dashed curve represents the empirical boundary between star formation and agn from sdss ( [email protected] ) by @xcite .
the solid curve is the theoretical star formation boundary from @xcite .
our limits indicate that the narrow - line emission that we detect is not strongly dominated by agn or shocks , with the possible exception of the low - signal to noise region c in sdssj1029 + 6510.,title="fig:",scaledwidth=60.0% ] /h@xmath79 ) and log([nii]/h@xmath2 ) for sdssj1029 + 6510 regions a , b @xmath12 c ( top ) and regions a , b , @xmath12 c for sdssj0925 + 0655 ( bottom ) are over - plotted .
dashed curve represents the empirical boundary between star formation and agn from sdss ( [email protected] ) by @xcite .
the solid curve is the theoretical star formation boundary from @xcite .
our limits indicate that the narrow - line emission that we detect is not strongly dominated by agn or shocks , with the possible exception of the low - signal to noise region c in sdssj1029 + 6510.,title="fig:",scaledwidth=60.0% ] | we present integral field spectrograph ( ifs ) with laser guide star adaptive optics ( lgs - ao ) observations of z@xmath02 quasi - stellar objects ( qsos ) designed to resolve extended nebular line emission from the host galaxy .
our data was obtained with w. m. keck and gemini - north observatories using osiris and nifs coupled with the lgs - ao systems , respectively .
we have conducted a pilot survey of five qsos , three observed with nifs+ao and two observed with osiris+ao at an average redshift of z=2.2 .
we demonstrate that the combination of ao and ifss provides the necessary spatial and spectral resolutions required to separate qso emission from its host .
we present our technique for generating a psf from the broad - line region of the qso and performing psf subtraction of the qso emission to detect the host galaxy emission at separation of @xmath00.2@xmath1 ( @xmath01.4 kpc ) .
we detect h@xmath2 narrow - line emission for two sources , sdssj1029 + 6510 ( z@xmath3=2.182 ) and sdssj0925 + 0655 ( z@xmath3=2.197 ) , that have evidence for both star formation and extended narrow - line emission .
assuming that the majority of narrow - line h@xmath2 emission is from star formation , we infer a star formation rate for sdssj1029 + 6510 of 78.4 m@xmath4 yr@xmath5 originating from a compact region that is kinematically offset by 290 - 350 kms@xmath5 . for sdssj0925 + 0655
we infer a star formation rate of 29 m@xmath4 yr@xmath5 distributed over three clumps that are spatially offset by @xmath0 7 kpc .
the null detections on three of the qsos are used to infer surface brightness limits and we find that at 1.4 kpc from the qso the un - reddened star formation limit is @xmath6 0.3 m@xmath4 yr@xmath5 kpc@xmath7 . if we assume typical extinction values for z=2 type-1 qsos , the dereddened star formation rate for our null detections would be @xmath6 0.6 m@xmath4 yr@xmath5 kpc@xmath7 .
these ifs observations indicate that while the central black hole is accreting mass at 10 - 40@xmath8 of the eddington rate , if star formation is present in the host ( 1.4 - 20 kpc ) it would have to occur diffusely with significant extinction and not in compact , clumpy regions . |
elements in the interstellar medium ( ism ) generally exist in the form of gas or dust .
the interstellar gas - phase abundances of elements can be measured from their optical and ultraviolet ( uv ) spectroscopic absorption lines .
the elements `` missing '' from the gas phase are bound up in dust grains this phenomenon is often called `` interstellar depletion '' .
the determination of the dust - phase abundances ( `` depletion '' ) is indirect and more complicated .
one usually relies on the interstellar extinction modeling or the solid - state spectral feature analysis .
both methods require an explicit assumption of the nature ( e.g. grain composition , size and geometry ) and the physical characteristics of the grains ( e.g. their optical properties and the strengths of their characteristic vibrational bands usually determined from their laboratory analogs ) .
more commonly , the dust - phase abundance of an element is derived by assuming a reference abundance total abundance of this element ( both in gas and in dust ) for the ism and then from which subtracting off the observed gas - phase abundance .
interstellar depletions allow us to extract important information about the composition and quantity of interstellar dust ( see 2.11 in li 2004a ) : ( 1 ) the fact that si , fe , mg , c , and o are depleted in low density clouds indicates that interstellar dust must contain an appreciable amount of si , fe , mg , c and o. indeed , all contemporary interstellar dust models consist of both silicates and carbonaceous dust ( see li 2004b for a review ) .
( 2 ) from the depletion of the major elements si , fe , mg , c , and o one can estimate the gas - to - dust mass ratio to be @xmath0165 ( see li 2004a ) , if the interstellar abundances are assumed to be those of the solar values of holweger ( 2001 ) . ( 3 ) in addition to the silicate dust component , there must exist another dust population , since silicates alone are not able to account for the observed amount of extinction relative to h although si , mg , and fe are highly depleted in the ism .
even if all si , fe , and mg elements of solar abundances of holweger ( 2001 ) are locked up in submicron - sized silicate grains , they can only account for @xmath060% of the total observed optical extinction ( see li 2004a ) .
apparently , in interstellar depletion and dust composition studies the knowledge of interstellar reference abundances is critical . historically , the solar abundances have been taken to represent the total interstellar abundances .
but we note that the published solar abundances have undergone major changes over the years and are still subject to major systematic uncertainties .
this is demonstrated in table [ tab : sunabd ] in which the widely used solar abundances over the past 3 decades for the dust - forming elements c , o , mg , si and fe are tabulated .
most notably , the most recent estimates of the solar c ( @xmath1_{\odot}}\approx 245{\,{\rm ppm}}$ ] ; allende prieto , lambert , & asplund 2002 ) and o abundances ( @xmath2_{\odot}}\approx 457{\,{\rm ppm}}$ ] ; asplund et al .
2004 ) are significantly reduced from their earlier values . using the gas - phase abundances measured by the _
satellite for the @xmath3 ophiuchi sightline ( morton et al . 1973 ) and the solar abundances of cameron ( 1973
; see table [ tab : sunabd ] ) as the reference abundances , greenberg ( 1974 ) found that the observed depletion of c , n , and o is significantly greater than could be accommodated by the dust under any reasonable models .
oph were subject to systematic errors arising in the gas - phase abundance analysis and the depletions toward many other stars are actually substantially lower .
] twenty years later , sofia , cardelli , & savage ( 1994 ) found that the interstellar depletions are lowered for c , n , and o if b stars are used as the reference standard .
they argued that the solar system may have enhanced abundances of many elements , and therefore the solar abundances ( of anders & grevesse 1989 ) are not representative of the interstellar abundances .
snow & witt ( 1996 ) analyzed the surface abundances of b stars and field f and g stars and found that not only c , n , and o but also si , mg , and fe and many other elements are underabundant in these stars .
this led them to suggest that the interstellar abundances are appreciably subsolar ( @xmath060%70% of the solar values compiled by anders & grevesse 1989 ) .
it is worth noting that the most recent estimates of the solar c ( @xmath1_{\odot}}\approx 245{\,{\rm ppm}}$ ] ; allende prieto et al .
2002 ) and o abundances ( @xmath2_{\odot}}\approx 457{\,{\rm ppm}}$ ] ; asplund et al .
2004)_{\odot}}$ ] is of particular interest : the original @xmath0851@xmath4 ( anders & grevesse 1989 ) was considered too large to be accounted for in the ism if the solar abundance is taken to be the interstellar reference standard ( mathis 1996 ) .
this was one of the major motivations to invoke subsolar interstellar abundances . ]
are also `` subsolar '' , just @xmath050%70% of the anders & grevesse ( 1989 ) values and close to the `` subsolar '' interstellar abundances originally recommended by snow & witt ( 1996 ) .
these reductions in the amounts of heavy elements available for grains have turned the former surplus of raw materials ( greenberg 1974 ) into a shortage and therefore have profound significance for the nature of interstellar dust ( snow & witt 1996 ) .
if the interstellar abundances are indeed `` subsolar '' like b stars or the newly - determined solar values , there is clearly a crisis for interstellar grain modeling ; in particular , this places severe challenges to the interstellar extinction modeling and the interpretation of the interstellar 9.7@xmath5 feature which is generally attributed to amorphous silicate dust feature strength @xmath6@xmath7 in the local diffuse ism is @xmath8 = \frac{\delta \tau_{9.7{\,{\rm
\mu m}}}}{n_{\rm h } } \frac{1}{\kappa_{\rm sil}^{\rm abs}(9.7{\,{\rm \mu m}})\,\mu_{\rm sil } } = \frac{\delta \tau_{9.7{\,{\rm \mu m}}}}{a_v } \frac{a_v}{n_{\rm h } } \frac{1}{\kappa_{\rm sil}^{\rm abs}(9.7{\,{\rm \mu m}})\,\mu_{\rm sil}}\ ] ] where @xmath9 is the silicate mass absorption coefficient at @xmath10=9.7@xmath5 , @xmath11 is the visual extinction , and @xmath12 is the hydrogen column density , @xmath13 is the silicate molecular weight . with @xmath9@xmath142850@xmath15 and @xmath13@xmath14172@xmath16 for amorphous olivine mgfesio@xmath17
, the local diffuse ism ( @xmath18@xmath14@xmath19@xmath20 ) requires @xmath21@xmath1435@xmath4 , significantly exceeding the b star si / h abundance ( @xmath22 ; sofia & meyer 2001 ) . but also see mathis ( 1998 ) who argued that composite fluffy spheroids with axis ratios greater than 2 could explain the observed silicate features even if the interstellar abundances are subsolar . ]
this is often referred to as interstellar abundance `` budget crisis '' ( e.g. see kim & martin 1996 ) . in the context of these reduced reference abundances , mathis ( 1986 )
re - investigated the composite interstellar dust model and found that fluffy grains consisting of small silicates , oxides , amorphous carbon derived from burning benzene ( `` be '' ) , and vacuum with a volume fraction of @xmath045% could account for the observed interstellar extinction and satisfy the much tighter abundance constraints .
other dust models invoking compact grains require more elements than available ( li & greenberg 1997 , weingartner & draine 2001 , li & draine 2001 , zubko , dwek , & arendt 2004 ) . by comparing the observed dust - to - gas ratio of the @xmath3 oph sightline with that expected from the reduced depletions , snow & witt ( 1996 ) concluded that the dust in this sightline must have a fluffy , porous or fractal structure .
however , can fluffy dust really alleviate the subsolar interstellar abundance budget crisis ?
we will explore this issue in terms of the kramers - kronig relations ( purcell 1969 ) , using the most recently determined ( subsolar ) b star abundances ( sofia & meyer 2001 ; see table [ tab : sunabd ] ) as the reference standard .
it is found in this paper that the kramers - kronig relations readily rule out the b star abundances as the interstellar reference abundances since the available elements are insufficient to account for the observed interstellar extinction , independent of specific grain models and grain fluffiness .
it is also found that if the sun ( of holweger 2001 ) or the field f and g stars ( of sofia & meyer 2001 ) are used as the reference standard , the dust depletion is not inconsistent with the interstellar extinction .
however , given that even the elemental abundances of our own sun are not well determined , we are not at a position to conclude that the sun or the field f and g stars reference abundances are to be preferred ; after all , if the solar mg , si , and fe elements are also close to these of b stars as the newly determined solar c and o abundances , the solar abundances would also fall short in accounting for the interstellar extinction .
as shown by purcell ( 1969 ) , the kramers - kronig relations can be applied to the interstellar space sparsely populated by interstellar grains .
let @xmath23 be the extinction optical depth at wavelength @xmath10 and @xmath24 be the extinction per h nucleon .
the kramers - kronig relations can be used to relate @xmath25 to the total volume occupied by dust per h nucleon @xmath26 through @xmath27 where the dimensionless factor @xmath28 is the orientationally - averaged polarizability relative to the polarizability of an equal - volume conducting sphere , depending only upon the grain shape and the static ( zero - frequency ) dielectric constant @xmath29 of the grain material ( purcell 1969 ; draine 2003a ) .
let @xmath30 be the volume fraction of vacuum ( `` fluffiness '' or `` porosity '' ) contained in a fluffy grain .
let @xmath31_{\rm ism}}$ ] be the total interstellar abundance of element x relative to h , @xmath31_{\rm gas}}$ ] be the amount of x in gas phase , @xmath31_{\rm dust}}$ ] be the amount of x contained in dust ( obviously we have @xmath31_{\rm dust}}$]=@xmath31_{\rm ism}}-{\left[{\rm x / h}\right]_{\rm gas}}$ ] ) , and @xmath32 be the atomic weight of x. let there be @xmath33 dust species in a composite fluffy grain ; let @xmath34 be the mass density of dust species @xmath35 ; let @xmath36 be the fraction of element x locked up in dust species @xmath35 .
for a chosen set of interstellar reference abundances , the total volume of interstellar dust per h nucleon can be estimated from the interstellar depletions @xmath37_{\rm ism}}-{\left[{\rm x / h}\right]_{\rm gas}}\right ) \mu_{\rm x}/\left(1-p\right ) \rho_j\ ] ] where the first summation is over all possible dust species and the second one is over all condensable elements . ) actually slightly overestimates the total dust volume since the @xmath38 enlarging factor has also been applied to the ultrasmall grains [ e.g. polycyclic aromatic hydrocarbon molecules ( pahs ) responsible for the 3.3 , 6.2 , 7.7 , 8.6 and 11.3@xmath5 `` unidentified infrared '' ( uir ) emission features ] , which are separated from the bulk , composite fluffy grains which contain most of the mass .
although small pahs are expected to be planar ( and thus `` highly flattened '' ) , their static dielectric constants @xmath29 are not large ( e.g. @xmath39 for benzene ) , therefore the @xmath28 factors for pahs will be in the order of unity and their contribution to the right - hand - side of eq.([eq : kk ] ) will not be large , after all , in the silicate - graphite - pahs interstellar grain model ( weingartner & draine 2001 ; li & draine 2001 ) the pah population takes over only @xmath40 of the total grain volume ( and @xmath41 of the total grain mass ) . ] the interstellar extinction per h nucleon @xmath24 is known for a limited range of wavelengths . for the diffuse ism , the mean extinction per h nucleon @xmath24 is fairly well - determined from the far uv to infrared ( ir ) : @xmath42 ( draine 2003b ) .
for @xmath43 we will adopt the theoretical @xmath24 values calculated from the silicate - graphite - pah model which has been shown to successfully reproduce the observed interstellar extinction from the far - uv to mid - ir and the observed ir emission ( weingartner & draine 2001 , li & draine 2001 ) .
since @xmath23 is a positive number for all wavelengths , the integration of @xmath24 over a finite wavelength range can be used to obtain a lower bound on @xmath44 @xmath45_{\rm ism}}-{\left[{\rm x / h}\right]_{\rm gas}}\right ) \mu_{\rm x}/\rho_j } ~~.\ ] ] at a first glance of eqs.([eq : kk],[eq : vtot ] ) , there appears to be no problem for dust models with subsolar interstellar abundances to account for the interstellar extinction , provided that the dust is sufficiently fluffy ( i.e. , the porosity @xmath30 is sufficiently large ) .
however , one should keep in mind that for a composite fluffy grain , the increase in the total dust volume @xmath46 will be offset by a decrease in @xmath28 since the effective static dielectric constant becomes smaller when the dust becomes more porous which leads to a smaller @xmath28 ( see fig.1 in purcell [ 1969 ] and fig.15 in draine [ 2003a ] ) . for the diffuse ism ,
the integration of the mean extinction over the wavelength range @xmath47 is approximately @xmath48 .
if we adopt the b star abundances as the reference standard and subtract the observed gas - phase abundances ( see table 2 in sofia 2004 ) , we obtain @xmath1_{\rm dust}}= { \left[{\rm c / h}\right]_{\rm ism}}-{\left[{\rm c / h}\right]_{\rm gas}}\approx 60 $ ] parts per million ( ppm ) , @xmath49_{\rm dust}}\approx 21{\,{\rm ppm}}$ ] , @xmath50_{\rm dust}}\approx 16.8{\,{\rm ppm}}$ ] , and @xmath51_{\rm dust}}\approx 27.5{\,{\rm ppm}}$ ] .
it is reasonable to assume that all the dust - phase si atoms are incorporated into amorphous olivine silicates with a stoichiometric composition of mgfesio@xmath17 .
this also consumes 16.8@xmath4 mg and fe .
we take the remaining 4.2@xmath4 mg to be depleted in mgo , and the remaining 10.7@xmath4 fe evenly tied up in feo , fe@xmath52o@xmath53 , and fe@xmath53o@xmath17
. we will assume the 60@xmath4 dust - phase carbon to be bound up either in graphite or in amorphous carbon .
therefore , from eq.([eq : fp ] ) we obtain @xmath54\right)_{\rm min}\approx 2.07 $ ] if the carbon is in the form of graphite , or @xmath54\right)_{\rm min}\approx 1.94 $ ] if the carbon is in the form of `` be '' amorphous carbon .
in doing so , we take the mass density of amorphous olivine mgfesio@xmath17 , mgo , feo , fe@xmath52o@xmath53 , fe@xmath53o@xmath17 , graphite , and amorphous carbon to be 3.5 , 3.58 , 5.7 , 5.25 , 5.18 , 2.24 , and 1.8@xmath55 , respectively . for a composite fluffy grain consisting of small amorphous silicates , oxides ( mgo , feo , fe@xmath52o@xmath53 , and fe@xmath53o@xmath17 ) , and graphite or amorphous carbon particles
, we use the bruggeman effective medium theory ( bohren & huffman 1983 ; ossenkopf 1991 ) to calculate its effective static dielectric constants @xmath56 .
the static dielectric constants of the constituent dust materials of the composite grain adopted in this work are : @xmath57 for amorphous olivine ( draine & lee 1984 ) , @xmath58 for mgo ( roessler & huffman 1998 ) , @xmath59 for fe@xmath52o@xmath53 ( steyer 1974 ) , @xmath60 for amorphous carbon ( rouleau & martin 1991 ) , and @xmath61 for feo , fe@xmath53o@xmath17 and graphite ( in the calculation , we take @xmath62 to represent @xmath61 ; it is found that there is not much difference among the model results obtained using @xmath63 , @xmath64 , or @xmath62 ) .
interstellar grains are nonspherical as indicated by the detection of interstellar polarization . for simplicity , we approximate these nonspherical grains by prolates or oblates .
interstellar polarization modeling suggests that interstellar grains are modestly elongated : lee & draine ( 1985 ) found that the 3.1@xmath5 ice polarization feature of the becklin - neugebauer ( bn ) object is best fit by @xmath65 oblates , where @xmath66 ( @xmath67 ) is the semiaxis along ( perpendicular to ) the symmetry axis ; similarly , hildebrand & dragovan ( 1995 ) also found that @xmath65 oblates provide the best match to the 9.7@xmath5 silicate polarization feature of the bn object ; using ice - coated core - mantle grains , greenberg & li ( 1996 ) found that @xmath68 prolates are preferred in explaining the 9.7 and 18@xmath5 silicate polarization features of this object . for a composite fluffy grain of a given porosity @xmath30 and a given elongation @xmath69
, we first calculate its effective static dielectric constant @xmath56 and then calculate the @xmath28 factor . in figure
[ fig : fp ] we show the model - predicted @xmath54\right)_{\rm mod}$ ] as a function of @xmath30 obtained for spheres ( @xmath70 ) and prolates ( @xmath71 ) using the b star abundances as the interstellar reference standard .
it is seen that even for @xmath72 , the model values @xmath54\right)_{\rm mod}$ ] are always below the lower limit @xmath54\right)_{\rm min}$ ] , no matter what chemical form the carbon takes ( graphite or amorphous carbon ) .
similar results are obtained for oblate grains .
this indicates that _ if the total interstellar abundances are those of the b stars , there are not enough raw materials to make the dust producing the interstellar extinction _ , unless the grains are highly elongated or flattened ( @xmath73 for prolates or @xmath74 for oblates ) .
it is also seen in figure [ fig : fp ] that the model value of @xmath44 first increases with @xmath30 and reaches its maximum at @xmath750.55 , depending on the grain shape , and then decreases .
this explains why in the latest version of the composite dust model to make the most economical use of the heavy elements mathis ( 1996 ) arrived at a porosity of @xmath76 , in contrast to the original value of @xmath77 ( mathis & whiffen 1989 ) .
peaks at @xmath78 for spherical grains consisting silicates , oxides and amorphous carbon ( see fig.[fig : fp ] ) while it is at @xmath76 in the mathis ( 1996 ) model lies in the fact the b star abundances and the interstellar gas - phase abundances we are using are somewhat different from those adopted in mathis ( 1996 ) . ] for @xmath79 , the model - predicted values of @xmath44 are insensitive to grain shape .
this is because for grains with a porosity @xmath79 , their effective static dielectric constants becomes smaller than @xmath01.8 , while the @xmath28 factor is insensitive to grain shape when @xmath80 ( see fig.1 of purcell [ 1969 ] and fig.15 of draine [ 2003a ] ) .
using the kramers - kronig relations , we have shown in [ sec : kk ] that the b star abundances as the interstellar reference standard fall short in accounting for the interstellar extinction , no matter how fluffy the dust is or what precise chemical form the dust takes , unless the dust is highly elongated or flattened . mathis ( 1996 ) suggested that the interstellar extinction could be accounted for if the dust has a fluffy , porous structure , consisting of small silicates , amorphous carbon , and oxides with vacuum comprising @xmath045% of its volume . but dwek ( 1997 ) argued against this suggestion on the basis of the fluffy dust model emitting too much in the far - ir to be consistent with that observed for the diffuse ism and the neglect of pahs in the model ( which are required to explain the `` uir '' emission bands ) .
times the @xmath28 factor , unless the dust is _ both _ extremely elongated _ and _ conducting , @xmath81 for porous grains does not differ much from that for compact grains of the same mass .
the enhanced short - wavelength absorptivity of porous dust revealed in jones ( 1988 ) also explains why porous dust emits more in the far - ir : the dust absorbs more at short wavelengths , and the absorbed energy must be radiated away at long wavelengths ! ]
while the dwek ( 1997 ) argument involves estimating the relatively uncertain uv and optical interstellar radiation field times the otherwise unobserved absorption ( in contrast to extinction ) of the grains ( cf .
mathis 1998 ) , the discussion presented in this work is more robust since essentially no assumptions are made in using the kramers - kronig relations to rule out the b star abundances as the interstellar reference standard .
the only assumption made in our analysis is that all the dust - phase si atoms , with a similar number of fe and mg atoms , are locked up in amorphous olivine silicates ( mgfesio@xmath17 ) and the remaining fe and mg atoms form oxides .
but we have also considered silicates in the form of amorphous pyroxene ( mg@xmath82fe@xmath83sio@xmath53 ) and amorphous olivine [ mg@xmath84fe@xmath85sio@xmath17 ] with various fe fraction @xmath87 and even with some of the si atoms in sio@xmath52 .
it is found that the conclusion of this work remains unchanged , unless an appreciable fraction of the fe atoms are in the form of metallic iron needles.o@xmath17 ) and even haematite ( fe@xmath52o@xmath53 ) on time - scales of the order of @xmath88 ; the reaction of metallic iron particles with sulphur in the ism will lead to their rapid degradation to sulphide and even sulphate . ]
we have also considered the f and g star abundances or the holweger ( 2001 ) solar abundances ( see table [ tab : sunabd ] ) as the interstellar reference standard .
we take all the dust - phase mg , si , and fe atoms to be in amorphous silicates and assign four o atoms for the average of the mg , si , fe abundances ( based on olivine ) . we take all the c atoms remaining from the gas to form graphite .
the kramers - kronig relations place a lower bound of @xmath89_{\rm min}\approx 0.97 $ ] for models using the f and g star abundances and @xmath89_{\rm min}\approx 0.96 $ ] for models using the solar abundances .
as shown in figure [ fig : fgsun ] , the model - predicted values of @xmath44 always exceed the lower limit @xmath89_{\rm min}$ ] except for very porous dust with @xmath90 , indicating that the f and g star abundances or the solar abundances are not inconsistent with the interstellar extinction . if we do not adopt the holweger ( 2001 ) solar abundances but the most recently determined solar values of c ( allende prieto et al .
2002 ) and o ( asplund et al . 2004 ) , and assume that the solar abundances of other elements like mg , si and fe are also close to those of b stars like c and o , the solar abundances would also fall short in accounting for the interstellar extinction .. if we adopt those of asplund et al .
( 2005 ) , the interstellar extinction can be accounted for by modestly elongated ( @xmath9123 ) grains with a vacuum volmue fraction of @xmath04060% . ]
given that even the elemental abundances of our own sun are still subject to major systematic uncertainties , we are not at a position to conclude that the solar or f and g star abundances are a more preferable reference standard for the ism .
it is very possible that the abundances of refractory elements in stars and in the ism are different , namely , there does not exist any set of abundances of refractory elements derived from _ stellar atmospheres _ , minus the measured gas - phase abundances of these elements in the diffuse ism , can yield the number of dust atoms implied by interstellar extinction . various physical processes such as radiation pressure , ambipolar diffusion , and gravitational sedimentation occurring during the early stages of star formation ( draine 2004 ; snow 2000 ) can act to prevent grains from being incorporated into stars and , therefore , lead to lower abundances of heavy elements in stars than in the ism out of which they have formed .
alternatively , the stellar _ photospheric _ abundances of heavy elements may under - represent the composition of the star , due to the possible underestimation of the degree of heavy - element settling in stellar atmospheres .
the latter is also indicated by the inconsistency between the most recent solar abundance determinations and the helioseismological results ( bahcall et al .
2004 ) .
however , there does exist some evidence for the subsolar nature of the interstellar abundances : some noble gases have subsolar abundances in the ism ( e.g. cardelli & meyer [ 1997 ] found that krypton has an interstellar abundance of 60% of solar ; neon was found to be @xmath075% solar in the ism [ takei et al .
2002 ] ) . since
noble gases are not likely to be depleted onto dust , the abundances of noble gases in the ism should reflect the total interstellar abundances .
the problem of what might be the most appropriate set of interstellar reference abundances is still awaiting a solution , but it is unlikely for the ism to have abundances as low as those of b stars since otherwise there will not be enough material to make the dust needed to explain the observed interstellar extinction .
i thank g.j .
bendo , j.i .
lunine , k.a .
misselt , and the anonymous referee for helpful discussions and/or suggestions .
i thank the university of arizona for the `` arizona prize postdoctoral fellowship in theoretical astrophysics '' .
allende prieto , c. , lambert , d.l . , & asplund , m. 2002 , apj , 573 , l137 anders , e. , & grevesse , n. 1989 , geochim .
acta , 53 , 197 asplund , m. , grevesse , n. , & sauval , a.j .
2005 , in cosmic abundances as records of stellar evolution and nucleosynthesis , ed .
f.n . bash , & t.g barnes ( san francisco : asp ) , in press asplund , m. , grevesse , n. , sauval , a.j .
, allende prieto , c. , & kiselman , d. 2004 , a&a , 417 , 751 bahcall , j.n . , basu , s. , pinsonneault , m. , & serenelli , a.m. 2004 , apj , in press bohren , c.f . , & huffman , d.r .
1983 , absorption and scattering of light by small particles ( new york : wiley ) cameron , a.g.w .
1973 , space sci .
, 15 , 121 cameron , a.g.w .
1982 , in essays in nuclear astrophysics , ed .
barnes , d.d .
clayton , & d.n .
schramm ( cambridge : cambridge univ .
press ) , 23 cardelli , j.a . , & meyer , d.m .
1997 , apj , 477 , l57 draine , b.t .
2003a , in the cold universe , saas - fee advanced course vol.32 , ed .
d. pfenniger ( berlin : springer - verlag ) , 213 draine , b.t .
2003b , ara&a , 41 , 241 draine , b.t .
2004 , in asp conf .
309 , astrophysics of dust , ed .
witt , g.c .
clayton , & b.t .
draine ( san francisco : asp ) , 691 draine , b.t . , & lee , h.m .
1984 , apj , 285 , 89 dwek , e. 1997 , apj , 484 , 779 greenberg , j.m . 1974 , apj , 189 , l81 greenberg , j.m . , &
li , a. 1996 , a&a , 309 , 258 grevesse , n. , & sauval , a.j .
1998 , space sci .
, 85 , 161 hildebrand , r.h .
, & dragovan , m. 1995 , apj , 450 , 663 holweger , h. 2001 , in solar and galactic composition , ed . r.f .
wimmer - schweingruber ( berlin : springer ) , 23 jones , a.p .
1988 , mnras , 234 , 209 jones , a.p .
1990 , mnras , 245 , 331 kim , s .- h . , & martin , p.g .
1996 , apj , 462 , 296 lee , h.m .
, & draine , b.t .
1985 , apj , 290 , 211 li , a. 2004a , in penetrating bars through masks of cosmic dust : the hubble tuning fork strikes a new note , ed .
block , i. puerari , k.c .
freeman , r. groess , & e.k .
block ( dordrecht : kluwer ) , 535 li , a. 2004b , in asp conf .
309 , astrophysics of dust , ed .
witt , g.c .
clayton , & b.t .
draine ( san francisco : asp ) , 417 li , a. , & draine , b.t .
2001 , apj , 554 , 778 li , a. , & greenberg , j.m .
1997 , a&a , 323 , 566 mathis , j.s .
1996 , apj , 472 , 643 mathis , j.s .
1998 , apj , 497 , 824 mathis , j.s .
, & whiffen , g. 1989 , apj , 341 , 808 morton , d.c . ,
drake , j.f . ,
jenkins , e.b .
, rogerson , j.b . , spitzer , l. , & york , d.g .
1973 , apj , 181 , l103 ossenkopf , v. 1991 , a&a , 251 , 210 purcell , e.m .
1969 , apj , 158 , 433 roessler , d.m . , &
huffman , d.r .
1998 , in handbook of optical constants of solids ii , ed .
palik ( boston : academic ) , 919 rouleau , f. , & martin , p.g .
1991 , apj , 377 , 526 snow , t.p .
2000 , j. geophys .
res . , 105 , 10239 snow , t.p . , &
witt , a.n .
1995 , science , 270 , 1455 snow , t.p . , &
witt , a.n .
1996 , apj , 468 , l65 sofia , u.j .
2004 , in asp conf .
309 , astrophysics of dust , ed .
witt , g.c .
clayton , & b.t .
draine ( san francisco : asp ) , 393 sofia , u.j . , & meyer , d.m .
2001 , apj , 554 , l221 sofia , u.j . ,
cardelli , j.a . , & savage , b.d .
1994 , apj , 430 , 650 steyer , t.r . 1974 , phd thesis , univ .
arizona takei , y. , fujimoto , r. , mitsuda , k. , & onaka , t. 2002 , apj , 581 , 307 weingartner , j.c .
, & draine , b.t .
2001 , apj , 548 , 296 whittet , d.c.b . 1984 ,
mnras , 210 , 479 zubko , v.g . , dwek , e. , & arendt , r.g .
2004 , apjs , 152 , 211 | what might be the most appropriate set of interstellar reference abundances of chemical elements ( both in gas and in dust ) relative to hydrogen has been a subject of much discussion in the past decade . while historically the sun has been taken as the reference standard , it has recently been suggested that the interstellar abundances might be better represented by those of b stars ( because of their young ages ) which are just @xmath06070% of the widely adopted solar values ( `` subsolar '' ) . on the other hand ,
the most recent estimates of the solar carbon and oxygen abundances are also close to those of b stars .
if the interstellar abundances are indeed `` subsolar '' like b stars or the newly - determined solar c , o values , there might be a lack of raw material to form the dust to account for the interstellar extinction . in literature
it has been argued that this problem could be solved if interstellar grains have a fluffy , porous structure since fluffy grains are more effective in absorbing and scattering optical and ultraviolet starlight than compact grains ( on a per unit mass basis ) .
however , we show in this work that , using the kramers - kronig relation , _ fluffy dust is not able to overcome the abundance shortage problem_. a likely solution is that the abundances of refractory elements in _ stellar photospheres _ may under - represent the composition of the interstellar material from which stars are formed , resulting either from the possible underestimation of the degree of heavy - element settling in stellar atmospheres , or from the incomplete incorporation of heavy elements in stars during the star formation process . |
[ sec : intro ] the infrared spectra of protostars and obscured background stars show prominent absorption features at 3.0 , 4.25 , 4.7 , 6.0 , 6.85 , 9.7 , and 15 @xmath1 along with a suite of weaker features ( e.g. @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite ; see @xcite for a complete list of features ) .
these are attributed to absorption in the vibrational modes of molecules in ices , except for the 9.7 @xmath1 band which is mostly due to silicates . at the low temperatures ( @xmath920 k ) of dense clouds and circum - protostellar environments atoms and molecules
freeze out rapidly on dust grains .
grain surface chemistry efficiently forms new , simple species , such as h@xmath3o and h@xmath3co ( e.g. @xcite ) . complex species ( e.g. polyoxymethylene [ ` pom ' , -(ch@xmath3-o)@xmath10- ] , and hexamethylenetetramine [ hmt , c@xmath11h@xmath12n@xmath13 )
can be formed through the impact of energetic photons or cosmic rays on the ices , as many laboratory studies have shown ( e.g. @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite ) .
thus far , only the simple species h@xmath3o , co , co@xmath3 , ch@xmath14 , ch@xmath4oh , nh@xmath4 and the @xmath15co and @xmath16 isotopes have been positively identified in the ices toward both low and high mass protostars as well as extincted background stars whose lines of sight trace quiescent dense cloud material .
reasonable evidence exists for the presence of hcooh , ocs and the ions nh@xmath8 , ocn@xmath6 , and hcoo@xmath6 as well , although their existence is sometimes debated because of inaccurate fits with laboratory spectra or the lack of multiple bands for independent confirmation .
the identification of the ions in the ices has been particularly controversial .
the first ion claimed was ocn@xmath6 @xcite and was thought to be produced by heavy energetic processing of co : nh@xmath4 ices .
later it was realized that acid - base chemistry in a hnco : nh@xmath4 ice could yield the same products @xcite .
acid - base reactions are very efficient and occur at temperatures as low as 10 k @xcite .
the created ions are less volatile than neutral species and would be able to form an interstellar salt after the other species have sublimated .
the presence of complex species formed by ultraviolet ( uv ) photon and cosmic ray processing of the simple ices is at least as controversial .
many claims were made @xcite , but the observational evidence is not firm .
it is crucial to determine the complexity of the ices in the circumstellar environment of low mass young stellar objects ( ysos ) , as they may be delivered directly to comets and they may ultimately serve as the source of volatiles in planets . at sublimation fronts in the warm inner regions of disks they are the starting point of a complex gas phase chemistry .
the relative abundances of n- , c- , and o - bearing species coming from the ices determine directly the type of chemistry in hot cores ( or ` hot corinos ' for low mass objects ; e.g. @xcite ) . as a step toward better characterizing the molecular content of icy grain mantles
, this work presents spectra of a large sample of low mass protostars and background stars obtained with the infrared spectrometer ( irs ; @xcite ) at the spitzer space telescope @xcite in the 5 - 35 @xmath1 wavelength range .
these data are complemented with spectra at 2 - 4 @xmath1 obtained with ground - based facilities .
such full coverage over the near and mid - infrared wavelength ranges is essential in determining the composition of the ices ; most species have several vibrational modes and more secure identifications can be made when the features are studied simultaneously .
the identification process is aided by studying large samples of sight - lines facilitating correlation of band strengths and shapes with each other and , if known , with physical characteristics along the sight - lines , such as luminosity and evolutionary stage of any heating source .
the great sensitivity of the irs allows for observations of objects with luminosities that are down by orders of magnitude with respect to what could be observed before in the mid - infrared ( typically 1 l@xmath17 versus @xmath18 l@xmath17 ) .
most data presented in this paper were obtained from spectral surveys of nearby molecular clouds and isolated dense cores in the context of the spitzer legacy program ` from molecular cores to planet - forming disks ' ( ` c2d ' ; @xcite ) .
initial results from this program emphasize the importance of thermal processing in the evolution of the ices , as the ices surrounding the low mass yso hh 46 irs are more processed compared to b5 irs1 @xcite .
ices toward the edge - on disk crbr 2422.8 - 3423 were presented in @xcite and also show signs of heating .
finally , mid - infrared spectra of ices toward background stars tracing quiescent cloud material were published in @xcite , showing for the first time that the ( unknown ) carrier of the 6.85 @xmath1 absorption band is abundant even at the coldest conditions away from star formation .
this paper specifically addresses questions on the identification of the 6.0 and 6.85 @xmath1 bands .
these prominent absorption features are still not fully identified , despite being detected nearly 30 years ago with the kuiper airborne observatory @xcite and commonly observed toward massive ysos with the infrared space observatory ( iso ; @xcite , @xcite ) .
the peak position of the 6.85 @xmath1 band varies dramatically , and is thought to be a function of the ice temperature @xcite .
either one carrier with a pronounced temperature dependence of its absorption bands , or two independent carriers , one much more volatile than the other , must be responsible for the 6.85 @xmath1 feature .
if the former is the case , the ammonium ion ( nh@xmath8 ) is considered a promising candidate @xcite .
the 6.0 @xmath1 band was initially thought to be mainly due to the bending mode of h@xmath3o ( @xcite ; however , see @xcite ) , but analysis of iso spectra indicated that this is not the case toward several massive ysos @xcite .
the depth of the 6.0 @xmath1band is in some cases significantly ( factor of 2 ) deeper than that expected from the 3.0 @xmath1 h@xmath3o stretching mode .
part of this ` excess ' might be due to a strong vibration of the hcooh molecule .
recent observations of background stars show a small excess ( @xmath1925% ) , of which half could be due to the enhanced band strength of the h@xmath3o bending mode in co@xmath3-rich ices @xcite .
finally , it was claimed that much of the 6.0 @xmath1 excess and part of the 6.85 @xmath1 band is due to highly processed ices , i.e. a mixture of complex species with c h and o
h bonds produced after irradiation of simple ices @xcite . in this case , one would expect the 6.0 @xmath1 excess to be enhanced in high radiation environments , such as near more evolved protostars .
thus , fundamental questions remain on the identification of the 6.0 and 6.85 @xmath1 bands , and with the large sample presented in this work the possible answers can be further constrained .
subsequent papers will specifically address the co@xmath3 @xcite , ch@xmath14 @xcite , and nh@xmath4 ( s. bottinelli et al .
, in preparation ) species .
further papers in this series will investigate the ices toward background stars behind large clouds ( c. knez et al .
, in preparation ) and isolated cores ( a. c. a. boogert et al . , in preparation ) .
in [ sec : sou ] the source sample is presented , while the observations and data reduction of the ground - based l - band and spitzer 5 - 38 @xmath1 spectra are described in
[ sec : obs ] . in
[ sec : res ] the continuum level for the multitude of absorption features is determined , and subsequently the best value for the h@xmath3o column density derived , followed by an empirical decomposition of the 5 - 7 @xmath1 absorption complex .
the positively identified species , or correlations of the components with other observables are discussed in
[ sec : id ] .
finally , the abundances of the simple species , and constraints on the carriers of unidentified components are discussed in [ sec : dis ] .
lllllrlll l1448 irs1 & 03:25:09.44 & + 30:46:21.7 & perseus & low & 0.34 & & sl , sh , lh & nirspec + iras 03235 + 3004 & 03:26:37.45 & + 30:15:27.9 & perseus & low & 1.44 & & sl , ll & nirspec + iras 03245 + 3002 & 03:27:39.03 & + 30:12:59.3 & perseus & low & 2.70 & & sl , sh & + l1455 smm1 & 03:27:43.25 & + 30:12:28.8 & perseus & low & 2.41 & & sl , sh , ll1 & + rno 15 & 03:27:47.68 & + 30:12:04.3 & perseus & low & @xmath200.21 & & ll1,sl , sh , lh & nirspec + l1455 irs3 & 03:28:00.41 & + 30:08:01.2 & perseus & low & 0.98 & & sl , sh , ll1 & + iras 03254 + 3050 & 03:28:34.51 & + 31:00:51.2 & perseus & low & 0.90 & & ll1,sl , sh , lh & nirspec + iras 03271 + 3013 & 03:30:15.16 & + 30:23:48.8 & perseus & low & 2.06 & & ll1,sl , sh , lh & nirspec + iras 03301 + 3111 & 03:33:12.85 & + 31:21:24.2 & perseus & low & 0.51 & & sl , sh , lh & nirspec + b1-a & 03:33:16.67 & + 31:07:55.1 & perseus & low & 1.87 & & sl , sh , ll1 & nirspec + b1-c & 03:33:17.89 & + 31:09:31.0 & perseus & low & 2.66 & & sl , sh , ll1 & + b1-b & 03:33:20.34 & + 31:07:21.4 & perseus & low & 0.68 & & sl , ll & + b5 irs3 & 03:47:05.45 & + 32:43:08.5 & perseus & low & 0.51 & & ll1,sl , sh , lh & nirspec + b5 irs1 & 03:47:41.61 & + 32:51:43.8 & perseus & low & 0.78 & & sl , sh , lh & nirspec + l1489 irs & 04:04:43.07 & + 26:18:56.4 & taurus & low & 1.10 & & sl , sh , lh & nirspec + iras 04108 + 2803b & 04:13:54.72 & + 28:11:32.9 & taurus & low & 0.90 & & sl , sh , lh & nirspec + hh 300 & 04:26:56.30 & + 24:43:35.3 & taurus & low & 0.79 & & sl , sh , lh & nirspec + dg tau b & 04:27:02.66 & + 26:05:30.5 & taurus & low & 1.16 & & sl , sh , lh & nirspec + hh 46 irs & 08:25:43.78 & @xmath2051:00:35.6 & hh 46 & low & 1.70 & & sl , sh , lh & isaac + iras 12553 - 7651 & 12:59:06.63 & @xmath2077:07:40.0 & cha & low & 0.76 & & ll1,sl , sh , lh & + iras 13546 - 3941 & 13:57:38.94 & @xmath2039:56:00.2 & bhr92 & low & @xmath200.06 & & sl , sh , lh & + iras 15398 - 3359 & 15:43:02.26 & @xmath2034:09:06.7 & b228 & low & 1.22 & & sl , sh , ll1 & + elias 29 & 16:27:09.42 & @xmath2024:37:21.1 & oph & low & 0.53 & 26700814 & sws01 sp3 & + crbr 2422.8 - 3423 & 16:27:24.61 & @xmath2024:41:03.3 & oph & low & 1.36 & & sl , sh , lh & nirspec + rno 91 & 16:34:29.32 & @xmath2015:47:01.4 & l43 & low & 0.03 & & sl , sh , lh & isaac + iras 17081 - 2721 & 17:11:17.28 & @xmath2027:25:08.2 & b59 & low & 0.55 & & sl , sh , ll1 & nirspec + sstc2dj171122.2 - 272602 & 17:11:22.16 & @xmath2027:26:02.3 & b59 & low & 2.26 & & sl , ll & + 2massj17112317 - 2724315 & 17:11:23.13 & @xmath2027:24:32.6 & b59 & low & 2.48 & & sl , ll & nirspec + ec 74 & 18:29:55.72 & + 01:14:31.6 & serpens & low & @xmath200.25 & & sl , sh , lh & nirspec + ec 82 & 18:29:56.89 & + 01:14:46.5 & serpens & low & 0.38 & & sl , sh , lh & + svs 4 - 5 & 18:29:57.59 & + 01:13:00.6 & serpens & low & 1.26 & & sl , sh , lh & isaac + ec 90 & 18:29:57.75 & + 01:14:05.9 & serpens & low & @xmath200.09 & & sl , sh , lh & + ec 92 & 18:29:57.88 & + 01:12:51.6 & serpens & low & 0.91 & & sl , sh , lh & nirspec + ck4 & 18:29:58.21 & + 01:15:21.7 & serpens & low & @xmath200.25 & & sl , sh , lh & + r cra irs 5 & 19:01:48.03 & @xmath2036:57:21.6 & cra & low & 0.98 & & sl , sh , ll1 & isaac + hh 100 irs & 19:01:50.56 & @xmath2036:58:08.9 & cra & low & 0.80 & 52301106 & sws01 sp4 & + cra irs7 a & 19:01:55.32 & @xmath2036:57:22.0 & cra & low & 2.23 & & sl , sh , lh & isaac + cra irs7 b & 19:01:56.41 & @xmath2036:57:28.0 & cra & low & 1.63 & & sl , sh , lh & isaac + cra iras32 & 19:02:58.69 & @xmath2037:07:34.5 & cra & low & 2.15 & & sl , sh , ll1 & + l1014 irs & 21:24:07.51 & + 49:59:09.0 & l1014 & low & 1.28 & & sl , ll & nirspec + iras 23238 + 7401 & 23:25:46.65 & + 74:17:37.2 & cb 244 & low & 0.95 & & sl , sh , lh & nirspec + & & & & & & & & + w3 irs5 & 02:25:40.54 & @xmath2162:05:51.4 & & high & 3.53 & 42701302 & sws01 sp3 & + monr2 irs3 & 06:07:47.8 & @xmath2006:22:55.0 & & high & 1.66 & 71101712 & sws01 sp3 & + gl989 & 06:41:10.06 & @xmath2109:29:35.8 & & high & 0.52 & 71602619 & sws01 sp3 & + w33a & 18:14:39.44 & @xmath2017:52:01.3 & & high & 1.92 & 32900920 & sws01 sp4 & + gl7009s & 18:34:20.91 & @xmath2005:59:42.2 & & high & 2.52 & 15201140 & sws01 sp3 & + gl2136 & 18:22:26.32 & @xmath2013:30:08.2 & & high & 1.48 & 33000222 & sws01 sp3 & + s140 irs1 & 22:19:18.17 & @xmath2163:18:47.6 & & high & 1.57 & 22002135 & sws01 sp4 & + ngc7538 irs9 & 23:14:01.63 & @xmath2161:27:20.2 & & high & 2.31 & 09801532 & sws01 sp2 & + & & & & & & & & + elias 16 & 04:39:38.88 & @xmath2126:11:26.6 & taurus & bg & - & & sl , sh & + ec 118 & 18:30:00.62 & @xmath2101:15:20.1 & serpens & bg & - & & sl , sh & +
[ sec : sou ] the source sample is selected based on the presence of ice absorption features and consists of a combination of known low mass ysos and new ones identified from their spitzer / irac and mips broad - band seds ( table [ t : sample ] ; @xcite ) . for the known objects , their mid - infrared ( 5 - 20 @xmath1 ) spectra are presented here for the first time , with the exception of b5 irs1 , hh 46 irs @xcite , crbr 2422.8 - 3423 @xcite , and l1489 irs , iras 04108 + 2803b , hh 300 , and dg tau b @xcite . the ysos are located in the perseus , taurus , serpens , and corona australis molecular cloud complexes , and a number of nearby isolated dense cores ( table [ t : sample ] ) .
ysos in the ophiuchus cloud are notably absent , because the spitzer / irs spectra were not available at the time this paper was written .
the spectral energy distributions of the sample of 41 low - mass ysos span a wide range of spectral indices @xmath22 to + 2.70 , with @xmath23 defined as @xmath24 including all photometry ( @xmath25 ) available between the @xmath26 2.17 @xmath1 @xmath27-band from the two micron all sky survey ( 2mass ; @xcite ) and the @xmath26 24 @xmath1spitzer / mips band . in the infrared broad - band classification scheme @xcite , most objects ( 35 out of 41 ) fall in the embedded class 0/i category ( @xmath280.3 ) .
the remaining 6 objects are flat - type objects ( @xmath29 ; @xcite ) .
although most objects fall in the class 0/i category , the dispersion within it is large .
some objects are envelope - dominated class 0 sources undetected in the 2mass @xmath27-band ( e.g. l1455 smm1 , b1-c ) , while others are more likely to be flat - type sources with dispersed envelopes and/or face - on disks extincted by foreground dust .
the latter sources were identified by excluding the @xmath27-band flux in the @xmath23 determination , and are marked as such in table [ t : sample ] .
thus , our sample represents a wide range of yso evolutionary stages , and probably source orientation . to contrast the possibly processed ices toward ysos with ices toward quiescent regions , published mid - infrared spectra of 2 background stars toward the serpens and taurus clouds were included in the analysis @xcite .
in addition , 8 massive ysos and 2 low mass ysos observed with the iso satellite were included in the sample ( table [ t : sample ] ) .
these serve as a reference in the ice feature analysis , because their mid - infrared spectra are well - studied at higher spectral resolution than is possible with spitzer / irs and because the massive ysos may trace different physical conditions that may affect the formation and evolution of the ices .
[ sec : obs ] spitzer / irs spectra were obtained as part of the c2d legacy program ( pids 172 and 179 ) as well as a dedicated open time program ( pid 20604 ) .
in addition , several previously published gto spectra were included @xcite .
the coordinates and irs modules that were used for each source are listed in table [ t : sample ] .
for all modules the 2-dimensional basic calibrated data ( bcd ) spectral images produced by the ssc pipeline version s13.2.0 or later were used as the starting point of the reduction .
the low resolution ` short - low ' and ` long - low ' modules ( sl and ll ; @xmath30 ) were reduced in a way that is customary for ground - based spectra .
first , the spectral and spatial dimensions were orthogonalized , and then the 2-dimensional images of the two nodding positions were subtracted in order to remove extended emission . in some cases the background emission is highly structured or multiple sources are present in the slit , and instead of subtracting nodding pairs , the background emission was determined from neighboring columns in the extraction process .
a fixed width extraction was performed and the 1-dimensional spectra were then averaged .
subsequently the spectra were divided by spectra of the standard stars hr 2194 ( a0 v ; pid 1417 ; aor keys and for sl1 and sl2 respectively ) and hr 6606 ( g9 iii ; pid 1421 ; aor key for ll ) reduced in the same way in order to correct for wavelength - dependent slit losses .
spectral features were divided out using the photospheric models of @xcite .
the spectra of the ` short - high ' and ` long - high ' modules ( sh and lh ; @xmath31600 ) were reduced using the c2d pipeline @xcite .
an important difference with the low resolution modules is the need for the removal of spectral fringes , which is done in the c2d pipeline as well .
many of the spitzer spectra were complemented by ground - based keck / nirspec @xcite and vlt / isaac @xcite l - band spectra at resolving powers of @xmath32=2000 and 600 respectively .
these were reduced in a way standard for ground - based long - slit spectra . finally , the spitzer / irs modules and ground - based spectra were multiplied along the flux scale in order to match spitzer / irac 3.6 , 4.5 , 5.8 , and 8.0 @xmath1 and spitzer / mips 24 @xmath1 photometry @xcite , using the appropriate filter profiles . for the iso
/ sws satellite spectra used in this paper , the latest spd pipeline version 10.1 products were taken from the iso archive and the detector scans were cleaned from cosmic ray hits and averaged .
the final spectra do not show any significant differences with respect to the data published in @xcite .
[ sec : res ] all sources in the sample show numerous absorption features on top of rising , flat , or falling 235 @xmath1 spectral energy distributions ( fig .
[ f : sampl ] ) .
most of these features are due to ices ( [ sec : abs ] ) , and they are often accompanied by strong 9.7 and weaker 18 @xmath1 silicate absorption bands ; but toward some sources , the silicate bands are in emission ( e.g. rno 15 , iras 17081 - 2721 ) .
also , prominent narrow emission lines of h@xmath3 ( s[5 ] , s[6 ] , and s[7 ] ) are observed toward iras 03271 + 3013 , cra iras32 , and l1455 smm1 .
these belong to the most deeply embedded objects ( @xmath33 ; table [ t : sample ] ) in the sample and the line emission is most likely related to molecular outflow activity ( e.g. @xcite ; @xcite ) .
cra iras32 shows the shock - related 5.340 @xmath1emission line of fe@xmath34 @xcite , and a strong line at 6.628 @xmath1 , perhaps also due to fe@xmath34 .
these emission lines are not further analyzed here , but their presence compromises the analysis of the underlying absorption features in these particular sources .
h i emission lines ( most notably br @xmath23 at 4.052 @xmath1 and pf @xmath35 at 3.740 @xmath1 ) are present in many of the ground - based l - band spectra , but at the high spectral resolution contamination with the ice absorption features is not a problem . in sources with strong l - band h i emission ( e.g. rno 91 ) the blended pf @xmath23 ( 7.459 @xmath1 ) and hu @xmath36 ( 7.502 @xmath1 ) lines are weakly visible in the irs spectra ( @xmath53% of the continuum ) , affecting the continuum determination for the low contrast ice features in that region ( [ sec : abs ] ; @xcite ) .
[ f : sampl ] [ f : tau ] [ f : opdepth ] [ sec : cont ] determining the baseline for the absorption features in the spectra of embedded protostars is not trivial .
the continuum sed underlying the many blended absorption features originates from warm dust near the central star and is dependent on poorly characterized parameters such as disk and envelope mass , size , and inclination , and dust composition @xcite .
the continuum applied to each source and indicated in fig .
[ f : sampl ] was derived as follows .
short - ward of 5 @xmath1 the continuum is much better defined than at longer wavelengths , and therefore the 2 - 5 and 5 - 35 @xmath1 regions are treated separately . for the short wavelength part , continuum points selected at 2.3 , 3.9 , and 5.0 @xmath1 are interpolated with a smooth spline function .
inclusion of the 2.3 @xmath1 portion of the spectrum , or @xmath27-band photometry , is essential in choosing the best continuum for the 3.0 @xmath1 absorption band .
the continuum long - ward of 5 @xmath1 is much less constrained , because the entire 5 - 32 @xmath1 region is covered by absorption features .
even the ` emission peak ' at 8.0 @xmath1 is covered by the long wavelength wing of the 6.0 @xmath1 h@xmath3o ice absorption band ( e.g. @xcite ) , and the short wavelength wing of the 9.7 @xmath1 silicate band .
the latter often extends toward shorter wavelengths compared to the silicate spectrum observed in the diffuse medium ( e.g. toward gcs 3 ; @xcite ) .
some sources also show an additional depression at 8 @xmath1 due to yet unknown species ; the best known example of this is the massive yso w33a @xcite .
thus , an interpolation over a large wavelength range is required , with few constraints . when possible , a low order polynomial ( on a @xmath37 versus @xmath38 scale ) was used to interpolate between the 5 and 32 @xmath1 continuum points .
this works best for the most deeply embedded sources ( e.g. b1-c , 2massj17112317 ) . for many sources ,
some guidance for the polynomial fit is provided by scaling the silicate spectrum of the diffuse medium source gcs 3 @xcite to the peak optical depth at 9.7 @xmath1 and by scaling a laboratory h@xmath3o ice spectrum to the 3.0 @xmath1 band .
less embedded sources , i.e. the sources with relatively strong near - infrared emission ( e.g. rno 91 , iras 03235 + 3004 ) , show a decreasing continuum between 5 - 10 @xmath1 and an emission bump at longer wavelengths .
here , a best effort spline continuum was defined , guided by modified black - body curves below 10 @xmath1 . for the background stars
the optical depth spectra presented in @xcite were used .
[ sec : abs ] the derived continua were used to put the spectra on an optical depth scale for the analysis of the wealth of absorption features ( figs .
[ f : sampl ] and [ f : tau ] ) .
the well known 6.0 and 6.85 @xmath1 bands ( e.g. @xcite ) are detected in all sources , along with the 9.7 @xmath1 band of silicates ( either in absorption or emission ) and the 3.0 @xmath1 h@xmath3o and 15 @xmath1 co@xmath3 ice bands when available .
a number of weaker ice features ( 7.25 , 7.40 , 7.67 , 9.0 , and 9.7 @xmath1 ) are detected in several sources as well .
examples in fig .
[ f : opdepth ] show that these same absorption features are present in sources of a very different nature ( background stars , massive ysos , class 0 , i , and flat spectrum low mass ysos ) , but that the depths and the profiles vary .
very deep 6.0 and 6.8 @xmath1 bands , with peak optical depths of up to 2 , are observed in the class 0 sources .
these are the deepest ice bands ever detected in low mass ysos , comparable to the most obscured massive ysos ( e.g. w33a , gl 7009s ) .
[ sec : h2o ] solid h@xmath3o libration mode in the source hh 46 irs .
* panel a * shows a smooth , unbiased polynomial continuum ( thick line ) .
the optical depth spectrum is compared to the silicate spectrum of the galactic center source gcs3 in * panel b * ( thick line ; @xcite ) . after subtracting gcs3 a prominent absorption feature in found at 10 - 20 @xmath1 which is likely due to the strong libration mode of h@xmath3o ( * panel c * ) , as is shown by a comparison to a laboratory spectrum of pure amorphous h@xmath3o ice at @xmath3910 k ( thick line ; * panel d * ) .
this band is particularly sensitive to grain shape effects . and a better fit is obtained with small spherical grains ( thick line ; * panel e * ) .
the column densities of h@xmath3o derived from the 3 and 13 @xmath1 bands are in good agreement , further validating the correctness of the applied method .
the vertical dashed line is the 11.3 @xmath1 absorption feature identified in @xcite and @xcite . ]
[ f : lib ] h@xmath3o is the most abundant molecule in interstellar ices ( e.g. @xcite ) , and strong , broad absorption bands are present over much of the observed spectral range : the o - h stretching mode between @xmath52.7 - 3.6 @xmath1 , the o - h bending mode between @xmath55.4 - 9 @xmath1 , and the libration mode between @xmath510 - 30 @xmath1
. many absorption features of other species overlap with these h@xmath3o bands , especially in the 5 - 8 @xmath1 region ( [ sec:60 ] ) , and in order to separate them , accurate ( 10 - 20% ) solid h@xmath3o column densities are required .
the 3.0 @xmath1 band provides , in principle , the most accurate h@xmath3o column density , because the o - h stretching mode is very strong and dominates over other species @xcite . evidently , the ice band must not be saturated , the continuum signal - to - noise values must be sufficient , and photometry must be available below the 2.8 @xmath1 atmospheric gap for an accurate continuum determination ( [ sec : cont ] ) . none of these requirements are met for the most embedded objects in our sample ( e.g. b1-c , 2massj17112317 - 2724315 ) . for sources that do meet the requirements ,
the long wavelength wing that often accompanies the 3.0 @xmath1 band is not taken into account in the h@xmath3o column determination .
the width of a laboratory pure h@xmath3o absorption spectrum at the best fitting temperature is assumed @xcite , using the integrated band strength @xmath40 cm / molecule @xcite .
thus , the calculated column density represents that of absorption by pure h@xmath3o on small grains .
if the wing were taken into account , the h@xmath3o column density would be larger by at most 15% in some sources .
the wing , which has an integrated optical depth of at most 30% with respect to the rest of the 3 @xmath1 band , is probably caused by a combination of scattering on large ice coated grains ( radius @xmath410.5 @xmath1 ) and absorption by ammonia hydrates @xcite . only the former would contribute to the h@xmath3o column density , but less than expected for small grains because the integrated extinction ( absorption+scattering ) strength per molecule is a factor of 2 larger for large grains than it is for small grains ( e.g. fig . 7 in @xcite ) .
h@xmath3o libration mode can be used to reliably measure the solid h@xmath3o column density .
silicate free optical depth spectra were derived in the same way as shown in fig .
[ f : lib ] for the ysos cra irs7 b ( left panels ) and b1-b ( right panels ) .
for each source the top panel shows the observed spectrum and the polynomial continuum ( dotted line ) , and the bottom panel the optical depth spectrum .
note the very prominent libration mode and the peculiar absence of silicates toward b1-b .
again , small spherical grains of pure h@xmath3o at @xmath42=10 k ( thick line ) fit the peak position of the observed band well , although excess absorption is visible on the long wavelength side .
finally , note that the down - turn above 30 @xmath1 is likely real and due to the onset of the h@xmath3o lattice mode , as the good fits with the laboratory data show . ]
[ f : libonly ] an alternative , and sometimes the only , way to determine the h@xmath3o column density is by measuring the peak strength of the libration mode in the @xmath43 @xmath1 region .
this band is difficult to isolate in the spectrum because it is several @xmath1 wide and wedged between the strong 9.7 and 18 @xmath1 silicate features .
however , application of a low order polynomial continuum to hh 46 irs ( fig .
[ f : lib ] ) and subtraction of a standard interstellar silicate spectrum ( gcs 3 ; @xcite ) shows a broad feature centered on @xmath512.3 @xmath1that is attributed to the h@xmath3o libration mode . using a laboratory spectrum of a thin film of pure h@xmath3o ice at @xmath3910 k @xcite the depth is in good agreement with the 3.0 @xmath1 stretching mode , but there is a discrepancy in the peak position by about 1 @xmath1(fig .
[ f : lib ] ) .
the peak position of the libration mode is quite sensitive to the shape of the grains , and for small spherical grains ( using the formulation of @xcite ) a much better calculated fit is obtained ( fig .
[ f : lib ] ) . at such large shift to shorter wavelengths
, the h@xmath3o libration mode overlaps with the steep edge of the 9.7 @xmath1 silicate band and a substructure is introduced reminiscent of the 11.3 @xmath1 absorption feature reported in @xcite and @xcite . finally , this method of determining the h@xmath3o column density works well for deeply embedded sources ( fig .
[ f : libonly ] ) , but it fails when a silicate emission component is present , filling in the h@xmath3o absorption ( e.g. rno 91 ) . [ sec:60 ] prominent absorption bands are present at 6.0 and 6.8 @xmath1 in all sources . in - between these bands ( @xmath56.5 @xmath1 ) , the optical depth is non - zero and varies relative to the 6.0 and 6.8 @xmath1 peak optical depths for different sources ( e.g. fig .
[ f : opdepth ] ) . beyond 7.0 @xmath1 ,
` featureless ' absorption is present up to 8 @xmath1 that approaches zero optical depth at 8 @xmath1 in some sources , but continues to be significant in others and merges with the 9.7 @xmath1silicate band . here
an attempt is made to decompose this complex blend of absorption features .
distinct , much weaker bands that are present in a subset of sources at 7.25 , 7.4 , and 7.6 @xmath1 are discussed separately ( [ sec:7_8 ] ) .
spectra of four low mass ysos on normalized optical depth scales ( histogram line ) .
the normalization factors are indicated for each source .
the smooth solid line represents a laboratory spectrum of pure , amorphous solid h@xmath3o at the indicated temperature at the column density derived from the 3.0 and 13 @xmath1 bands .
clearly , in all sources a significant fraction of the absorption is not explained by solid h@xmath3o , and it varies from source - to - source . for the 6.0 @xmath1 feature this fraction
is indicated by the value of @xmath44 , defined in eq .
[ eq : e6 ] .
note the variations of the profiles of the 6.0 and 6.85 @xmath1 bands , as well as the presence of weak features at 7.25 , 7.4 , and 7.65 @xmath1 . ]
[ f:58 ] solid h@xmath3o is a major contributor to the optical depth in the 58 @xmath1 region by its o - h bending mode at 6.0 @xmath1 and the overtone of the libration mode extending between 6 and 8 @xmath1 ( e.g. @xcite ) . for consistency with previous work , although formally incorrect , these overlapping bands will be referred to as a single mode , the o - h bending mode of h@xmath3o .
its absorption profile depends on the ice structure .
in particular , as an amorphous ice is heated from @xmath45 to 80 k , the distinct peak at 6.0 @xmath1 becomes weaker @xcite . at higher temperatures ,
the shape remains relatively constant . for almost half of the sample
the ice temperature could be determined from the shape of the 3.0 @xmath1 h@xmath3o stretching mode ( table [ t : tau ] ) .
most other ysos are deeply embedded , and @xmath45 k was assumed .
subsequently a laboratory spectrum of pure h@xmath3o @xcite at the determined temperature was scaled to the column density derived from the 3.0 and 13 @xmath1 bands ( [ sec : h2o ] ) and subtracted from each optical depth spectrum . the shapes and depths of the residual features , and any variations thereof between sources ,
are analyzed .
it should be kept in mind , however , that some features may still be the result of interactions of species with h@xmath3o in solid state matrices , such as the narrow peak induced in h@xmath3o : co@xmath3 mixtures ( [ sec : disc2 ] ; @xcite ) .
( defined in eq .
[ eq : e6 ] ) plotted as a function of the solid h@xmath3o column density for the entire source sample : low mass ysos ( filled circles ) , massive ysos ( open circles ) , and background stars ( stars ) . for sources near the horizontal dotted line
the entire 6.0 @xmath1 absorption band can be explained by pure h@xmath3o ice .
all other sources have significantly deeper 6.0 @xmath1 bands than expected from the 3 or 13 @xmath1 bands . ]
[ f : h2oh2o ] toward most low mass ysos , the observed 6.0 @xmath1 absorption band is deeper than expected from the 3.0 and 13 @xmath1 h@xmath3o ice bands ( fig .
[ f:58 ] ) .
such a discrepancy has been reported toward massive ysos as well @xcite and it is quantified in units of the h@xmath3o column density as follows : @xmath46 which is the integrated optical depth of the 6.0 @xmath1 band in wavenumber ( @xmath47 ) space divided by the h@xmath3o column density derived from the 3.0 and/or 13 @xmath1 bands .
the factor of @xmath48 is the integrated band strength of the o - h bending mode in a pure h@xmath3o laboratory ice @xcite in units of cm / molecule .
the latter extends all the way out to 8.5 @xmath1 , while the numerator of eq .
[ eq : e6 ] integrates only to 6.4 @xmath1 .
therefore the laboratory band strength is reduced by a factor of 0.60 , as derived from the laboratory spectra .
@xmath44 varies significantly : from @xmath492 in @xmath530% of the ysos ( e.g. rno 91 , iras 03301 + 3111 , ec 92 ) to 1.0 ( i.e. no excess ) in other ysos ( table [ t : tau ] and fig .
[ f : h2oh2o ] ) .
absorption complex . in the left panel
are shown averaged , smoothed , h@xmath3o - subtracted spectra of ysos with small 6.0 @xmath1 excess ( * a * ) , ysos with strong 6.0 @xmath1 excess ( * b * ) , and the yso iras 03301 + 3111 which shows strongly red - shifted 6.0 and 6.85 @xmath1 bands ( * c * ) . applying linear combinations of spectra ( * a * ) , ( * b * ) , and ( * c * ) in the left panel , 5 independent absorption components are found and shown in the right panel .
components c1 and c3 are found by subtracting spectrum ( * c * ) , multiplied by a factor of 0.63 , from spectrum ( * a * ) [ * top * ] . components c2 and c4 are identical to spectrum ( * c * ) [ * middle * ] .
component c5 is found by subtracting spectrum ( * a * ) , multiplied by a factor of 0.69 , from spectrum ( * b * ) [ * bottom * ] .
these 5 components , in addition to the bending mode of h@xmath3o , are responsible for the 58 @xmath1 absorption complex in ysos and background stars .
example fits are shown in fig .
[ f : compex ] .
note that c3 and c4 are the previously found short- and long - wavelength components of the 6.85 @xmath1 band @xcite . ]
[ f : decomp ] region for a selection of sources , after subtraction of pure h@xmath3o and component c5 .
the vertical dashed lines at 6.0 and 6.85 @xmath1 are plotted to facilitate comparing the positions of the features in the different sources .
note that the plotted sources represent the most extreme cases of our sample ; most sources resemble the spectrum of hh 46 irs .
the spectra are normalized on optical depth scale , and sorted from top to bottom in increasing wavelength of the 6 @xmath1 residual .
note that at the same time the 6.85 @xmath1 feature shifts to longer wavelengths as well . ]
[ f : resid ] the large source - to - source variations of @xmath44 may hold clues to the nature of the carrier(s ) of the 6.0 @xmath1 excess absorption . as is illustrated in fig .
[ f : decomp ] , these variations are used to separate any independent , overlapping absorption components in the 5 - 8 @xmath1 range .
initially , a selection of sources with low values of @xmath44 and sources with high values of @xmath44 is averaged , and the resulting spectra are scaled and subtracted from each other such that the distinct peak at 6.85 @xmath1 is removed .
what remains is a very broad feature , stretching between 5.8 and 8 @xmath1 .
while the profiles of the 6.0 and 6.85 @xmath1 bands themselves look rather similar for many sources ( e.g. like hh 46 irs ) , significant source to source variations are sometimes observed ( fig .
[ f : resid ] ) .
the peak position of the 6.0 @xmath1 component varies between 5.85 and 6.2 @xmath1 , and broadens as the feature shifts toward longer wavelengths . at the same time
the 6.85 @xmath1 band shifts from 6.8 to 6.95 @xmath1 . by subtracting extreme spectra from each other
, it is found that the latter consists of two components centered at 6.755 and 6.943 @xmath1with widths of fwhm=0.195 and 0.292 @xmath1 , respectively , similar to what was previously found for massive protostars @xcite .
the 6.0 @xmath1 band consists of two components as well , with peak positions of 5.84 and 6.18 @xmath1 and fwhm widths of 0.26 and 0.40 @xmath1 , respectively .
absorption complex with the five components derived in fig .
[ f : decomp ] , as well as h@xmath3o . in the left and right panels
two different ysos are shown .
the dots are the observed data points , and the thin line through them is the sum of all components above . ]
[ f : compex ] it is thus hypothesized that , after h@xmath3o subtraction , the 58 @xmath1absorption in all sources is composed of 5 distinct components ( fig . [ f : decomp ] ) : a set of bands peaking at 5.84 and 6.18 @xmath1(c1 and c2 hereafter ) , a set of bands peaking at 6.755 and 6.943 @xmath1(c3 and c4 ) , and a broad asymmetric feature stretching between 5.8 and 8 @xmath1 and peaking near 5.9 @xmath1 ( c5 ) .
example fits with these components are shown in fig .
[ f : compex ] .
the peak optical depths of the components in all sources are listed in table [ t : tau ] .
the indicated error bars are based on statistical uncertainties in the spectral data points , except for c5 .
the accuracy of the c5 optical depth is affected by the accuracy of the baseline , which at present can only partly be quantified .
the baseline uncertainty is a combination of the error in @xmath50(h@xmath3o ) ( table [ t : colden ] ) and the poorly known error in the continuum choice ( [ sec : cont ] ) . for many sources the latter ,
not included in table [ t : tau ] , is likely at most 0.03 on optical depth scale . lccccccccc l1448 irs1 & 10 & 4.35 ( 1.57 ) & 0.36 ( 0.04 ) & 0.10 ( 0.01 ) & 0.00 ( 0.01 ) & 0.00 ( 0.01 ) & 0.02 ( 0.01 ) & 0.03 ( 0.01 ) & 0.08 ( 0.01 ) + iras 03235 + 3004 & 10 & 2.01 ( 0.31 ) & 1.98 ( 0.20 ) & 1.35 ( 0.07 ) & 0.38 ( 0.15 ) & 0.13 ( 0.15 ) & 0.40 ( 0.07 ) & 0.13 ( 0.07 ) & 0.39 ( 0.06 ) + iras 03245 + 3002 & 10 & 1.76 ( 0.25 ) & 8.11 ( 0.81 ) & 2.92 ( 0.17 ) & 0.52 ( 0.40 ) & 0.00 ( 0.40 ) & 1.19 ( 0.16 ) & 0.55 ( 0.16 ) & 0.84 ( 0.13 ) + l1455 smm1 & 10 & 2.39 ( 0.37 ) & 5.18 ( 0.52 ) & 1.92 ( 0.08 ) & 0.55 ( 0.24 ) & 0.17 ( 0.24 ) & 0.86 ( 0.16 ) & 0.25 ( 0.16 ) & 0.83 ( 0.07 ) + rno 15 & 10@xmath51 & 1.90 ( 0.19 ) & emission & 0.08 ( 0.01 ) & 0.01 ( 0.01 ) & 0.04 ( 0.01 ) & 0.03 ( 0.01 ) & 0.03 ( 0.01 ) & 0.02 ( 0.01 ) + l1455 irs3 & 10 & 2.45 ( 1.00 ) & 0.15 ( 0.05 ) & 0.13 ( 0.01 ) & 0.08 ( 0.04 ) & 0.08 ( 0.04 ) & 0.04 ( 0.03 ) & 0.02 ( 0.03 ) & 0.00 ( 0.03 ) + iras 03254 + 3050 & 40@xmath51 & 1.75 ( 0.23 ) & 1.63 ( 0.16 ) & 0.32 ( 0.02 ) & 0.14 ( 0.06 ) & 0.04 ( 0.06 ) & 0.18 ( 0.02 ) & 0.09 ( 0.02 ) & 0.04 ( 0.01 ) + iras 03271 + 3013 & 10 & 2.75 ( 0.63 ) & 1.29 ( 0.13 ) & 0.95 ( 0.06 ) & 0.39 ( 0.11 ) & 0.16 ( 0.11 ) & 0.33 ( 0.13 ) & 0.07 ( 0.13 ) & 0.26 ( 0.04 ) + iras 03301 + 3111 & 40@xmath51 & 2.61 ( 0.30 ) & 0.49 ( 0.05 ) & 0.06 ( 0.01 ) & 0.00 ( 0.00 ) & 0.03 ( 0.00 ) & 0.00 ( 0.01 ) & 0.03 ( 0.01 ) & 0.02 ( 0.01 ) + b1-a & 10 & 1.49 ( 0.32 ) & 2.35 ( 0.23 ) & 0.73 ( 0.04 ) & 0.33 ( 0.13 ) & 0.12 ( 0.13 ) & 0.39 ( 0.04 ) & 0.29 ( 0.04 ) & 0.03 ( 0.05 ) + b1-c & 10 & 1.48 ( 0.28 ) & 7.03 ( 0.70 ) & 2.04 ( 0.08 ) & 0.65 ( 0.37 ) & 0.14 ( 0.37 ) & 1.01 ( 0.06 ) & 0.64 ( 0.06 ) & 0.24 ( 0.13 ) + b1-b & 10 & 1.30 ( 0.24 ) & 2.30 ( 0.23 ) & 1.15 ( 0.04 ) & 0.51 ( 0.21 ) & 0.05 ( 0.21 ) & 0.52 ( 0.04 ) & 0.35 ( 0.04 ) & 0.00 ( 0.07 ) + b5 irs3 & 10@xmath51 & 1.84 ( 0.18 ) & 0.32 ( 0.04 ) & 0.12 ( 0.02 ) & 0.01 ( 0.01 ) & 0.04 ( 0.01 ) & 0.06 ( 0.02 ) & 0.05 ( 0.02 ) & 0.02 ( 0.02 ) + b5 irs1 & 10@xmath51 & 1.64 ( 0.21 ) & 1.23 ( 0.12 ) & 0.19 ( 0.01 ) & 0.07 ( 0.01 ) & 0.03 ( 0.01 ) & 0.07 ( 0.02 ) & 0.03 ( 0.02 ) & 0.01 ( 0.01 ) + l1489 irs & 40@xmath51 & 1.77 ( 0.21 ) & 1.32 ( 0.13 ) & 0.37 ( 0.02 ) & 0.15 ( 0.06 ) & 0.05 ( 0.06 ) & 0.19 ( 0.02 ) & 0.08 ( 0.02 ) & 0.06 ( 0.02 ) + iras 04108 + 2803b & 40@xmath51 & 1.61 ( 0.23 ) & 0.49 ( 0.05 ) & 0.22 ( 0.01 ) & 0.10 ( 0.03 ) & 0.06 ( 0.03 ) & 0.10 ( 0.01 ) & 0.06 ( 0.01 ) & 0.01 ( 0.02 ) + hh 300 & 10@xmath51 & 2.00 ( 0.20 ) & 1.09 ( 0.11 ) & 0.24 ( 0.01 ) & 0.04 ( 0.03 ) & 0.05 ( 0.03 ) & 0.08 ( 0.02 ) & 0.06 ( 0.02 ) & 0.04 ( 0.01 ) + dg tau b & 100@xmath51 & 1.87 ( 0.32 ) & 2.05 ( 0.20 ) & 0.23 ( 0.02 ) & 0.09 ( 0.02 ) & 0.08 ( 0.02 ) & 0.08 ( 0.02 ) & 0.10 ( 0.02 ) & 0.01 ( 0.01 ) + hh 46 irs & 40 & 2.00 ( 0.20 ) & 2.23 ( 0.22 ) & 0.75 ( 0.04 ) & 0.34 ( 0.13 ) & 0.17 ( 0.13 ) & 0.36 ( 0.02 ) & 0.18 ( 0.02 ) & 0.10 ( 0.02 ) + iras 12553 - 7651 & 10 & 0.99 ( 0.19 ) & 1.40 ( 0.14 ) & 0.15 ( 0.01 ) & 0.01 ( 0.03 ) & 0.02 ( 0.03 ) & 0.09 ( 0.03 ) & 0.07 ( 0.03 ) & 0.00 ( 0.02 ) + iras 13546 - 3941 & 10 & 1.19 ( 0.12 ) & 0.23 ( 0.03 ) & 0.15 ( 0.01 ) & 0.05 ( 0.03 ) & 0.03 ( 0.03 ) & 0.05 ( 0.01 ) & 0.02 ( 0.01 ) & 0.00 ( 0.01 ) + iras 15398 - 3359 & 10 & 2.02 ( 0.54 ) & 3.32 ( 0.33 ) & 1.41 ( 0.08 ) & 0.54 ( 0.17 ) & 0.15 ( 0.17 ) & 0.55 ( 0.03 ) & 0.31 ( 0.03 ) & 0.40 ( 0.09 ) + elias 29 & 40@xmath51 & 1.32 ( 0.13 ) & 1.44 ( 0.14 ) & 0.24 ( 0.02 ) & 0.05 ( 0.02 ) & 0.04 ( 0.02 ) & 0.05 ( 0.02 ) & 0.02 ( 0.02 ) & 0.00 ( 0.02 ) + crbr 2422.8 - 342 & 40@xmath51 & 1.40 ( 0.14 ) & 1.68 ( 0.17 ) & 0.32 ( 0.03 ) & 0.06 ( 0.05 ) & 0.09 ( 0.05 ) & 0.13 ( 0.02 ) & 0.11 ( 0.02 ) & 0.00 ( 0.02 ) + rno 91 & 40@xmath51 & 1.99 ( 0.17 ) & 0.80 ( 0.08 ) & 0.39 ( 0.02 ) & 0.13 ( 0.05 ) & 0.07 ( 0.05 ) & 0.13 ( 0.01 ) & 0.06 ( 0.01 ) & 0.10 ( 0.02 ) + iras 17081 - 2721 & 10@xmath51 & 1.61 ( 0.16 ) & emission & 0.10 ( 0.01 ) & 0.04 ( 0.02 ) & 0.01 ( 0.02 ) & 0.07 ( 0.01 ) & 0.03 ( 0.01 ) & 0.01 ( 0.01 ) + sstc2dj171122.2 - 272602 & 10 & 1.58 ( 0.32 ) & 3.09 ( 0.31 ) & 1.00 ( 0.05 ) & 0.24 ( 0.14 ) & 0.13 ( 0.14 ) & 0.48 ( 0.04 ) & 0.30 ( 0.04 ) & 0.19 ( 0.06 ) + 2massj17112317 - 2724315 & 10 & 1.46 ( 0.24 ) & 4.58 ( 0.46 ) & 1.30 ( 0.06 ) & 0.35 ( 0.24 ) & 0.04 ( 0.24 ) & 0.74 ( 0.04 ) & 0.42 ( 0.04 ) & 0.37 ( 0.07 ) + ec 74 & 10@xmath51 & 1.75 ( 0.31 ) & 0.47 ( 0.05 ) & 0.10 ( 0.01 ) & 0.02 ( 0.03 ) & 0.03 ( 0.03 ) & 0.05 ( 0.02 ) & 0.02 ( 0.02 ) & 0.01 ( 0.01 ) + ec 82 & 10 & @xmath520.63 & emission & 0.04 ( 0.01 ) & 0.02 ( 0.02 ) & 0.00 ( 0.02 ) & 0.02 ( 0.01 ) & 0.01 ( 0.01 ) & 0.00 ( 0.01 ) + svs 4 - 5 & 10 & 2.41 ( 0.48 ) & 1.61 ( 0.16 ) & 0.68 ( 0.04 ) & 0.31 ( 0.11 ) & 0.09 ( 0.11 ) & 0.28 ( 0.02 ) & 0.21 ( 0.02 ) & 0.17 ( 0.03 ) + ec 90 & 10 & 1.01 ( 0.10 ) & emission & 0.13 ( 0.01 ) & 0.05 ( 0.02 ) & 0.02 ( 0.02 ) & 0.06 ( 0.01 ) & 0.03 ( 0.01 ) & 0.00 ( 0.01 ) + ec 92 & 40@xmath51 & 2.28 ( 0.20 ) & emission & 0.24 ( 0.02 ) & 0.15 ( 0.04 ) & 0.07 ( 0.04 ) & 0.10 ( 0.02 ) & 0.02 ( 0.02 ) & 0.02 ( 0.01 ) + ck4 & 10 & @xmath521.08 & emission & 0.08 ( 0.00 ) & 0.03 ( 0.03 ) & 0.04 ( 0.03 ) & 0.01 ( 0.02 ) & 0.01 ( 0.02 ) & 0.00 ( 0.02 ) + r cra irs 5 & 40@xmath51 & 1.62 ( 0.12 ) & 1.02 ( 0.10 ) & 0.29 ( 0.02 ) & 0.11 ( 0.06 ) & 0.09 ( 0.06 ) & 0.13 ( 0.01 ) & 0.08 ( 0.01 ) & 0.00 ( 0.01 ) + hh 100 irs & 10@xmath51 & 2.05 ( 0.20 ) & 0.69 ( 0.07 ) & 0.27 ( 0.03 ) & 0.10 ( 0.05 ) & 0.10 ( 0.05 ) & 0.11 ( 0.03 ) & 0.09 ( 0.03 ) & 0.00 ( 0.02 ) + cra irs7 a & 10 & 0.97 ( 0.17 ) & 3.02 ( 0.30 ) & 0.54 ( 0.04 ) & 0.08 ( 0.09 ) & 0.03 ( 0.09 ) & 0.26 ( 0.02 ) & 0.35 ( 0.02 ) & 0.00 ( 0.04 ) + cra irs7 b & 10 & 1.24 ( 0.22 ) & 3.44 ( 0.34 ) & 0.68 ( 0.04 ) & 0.19 ( 0.12 ) & 0.06 ( 0.12 ) & 0.35 ( 0.02 ) & 0.31 ( 0.02 ) & 0.04 ( 0.04 ) + cra iras32 & 10 & 3.47 ( 1.24 ) & 2.43 ( 0.25 ) & 0.86 ( 0.04 ) & 0.59 ( 0.29 ) & 0.41 ( 0.29 ) & 0.34 ( 0.22 ) & 0.31 ( 0.22 ) & 0.00 ( 0.05 ) + l1014 irs & 10 & 1.52 ( 0.19 ) & 2.08 ( 0.21 ) & 0.62 ( 0.04 ) & 0.36 ( 0.15 ) & 0.21 ( 0.15 ) & 0.34 ( 0.04 ) & 0.16 ( 0.04 ) & 0.00 ( 0.02 ) + iras 23238 + 7401 & 10 & 1.17 ( 0.20 ) & 2.30 ( 0.23 ) & 0.72 ( 0.03 ) & 0.29 ( 0.10 ) & 0.07 ( 0.10 ) & 0.38 ( 0.05 ) & 0.16 ( 0.05 ) & 0.00 ( 0.05 ) + w3 irs5 & 10@xmath51 & 1.03 ( 0.10 ) & 6.28 ( 0.63 ) & 0.29 ( 0.01 ) & 0.00 ( 0.04 ) & 0.00 ( 0.04 ) & 0.14 ( 0.01 ) & 0.16 ( 0.01 ) & 0.00 ( 0.02 ) + monr2 irs3 & 100@xmath51 & 2.41 ( 0.24 ) & 2.25 ( 0.23 ) & 0.20 ( 0.01 ) & 0.01 ( 0.01 ) & 0.02 ( 0.01 ) & 0.06 ( 0.02 ) & 0.22 ( 0.02 ) & 0.09 ( 0.01 ) + gl989 & 40@xmath51 & 1.65 ( 0.08 ) & 0.66 ( 0.07 ) & 0.18 ( 0.01 ) & 0.06 ( 0.03 ) & 0.04 ( 0.03 ) & 0.10 ( 0.01 ) & 0.07 ( 0.01 ) & 0.01 ( 0.01 ) + w33a & 40 & 2.57 ( 0.64 ) & 6.02 ( 0.60 ) & 1.87 ( 0.03 ) & 0.46 ( 0.20 ) & 0.28 ( 0.20 ) & 0.74 ( 0.05 ) & 0.51 ( 0.05 ) & 0.67 ( 0.07 ) + gl7009s & 10 & 1.98 ( 0.40 ) & 3.62 ( 0.36 ) & 1.17 ( 0.05 ) & 0.38 ( 0.18 ) & 0.01 ( 0.18 ) & 0.72 ( 0.04 ) & 0.54 ( 0.04 ) & 0.37 ( 0.06 ) + gl2136 & 100@xmath51 & 1.58 ( 0.16 ) & 3.02 ( 0.30 ) & 0.35 ( 0.01 ) & 0.11 ( 0.05 ) & 0.04 ( 0.05 ) & 0.19 ( 0.02 ) & 0.14 ( 0.02 ) & 0.06 ( 0.01 ) + s140 irs1 & 100@xmath51 & 1.74 ( 0.17 ) & 1.93 ( 0.19 ) & 0.16 ( 0.01 ) & 0.05 ( 0.02 ) & 0.03 ( 0.02 ) & 0.05 ( 0.01 ) & 0.10 ( 0.01 ) & 0.00 ( 0.01 ) + ngc7538 irs9 & 10@xmath51 & 1.58 ( 0.16 ) & 2.25 ( 0.23 ) & 0.48 ( 0.02 ) & 0.12 ( 0.08 ) & 0.07 ( 0.08 ) & 0.27 ( 0.01 ) & 0.15 ( 0.01 ) & 0.04 ( 0.02 ) + elias 16 & 10@xmath51 & 1.48 ( 0.11 ) & 0.73 ( 0.07 ) & 0.18 ( 0.01 ) & 0.04 ( 0.02 ) & 0.04 ( 0.02 ) & 0.07 ( 0.01 ) & 0.03 ( 0.01 ) & 0.00 ( 0.02 ) + ec 118 & 10 & 1.17 ( 0.12 ) & 1.42 ( 0.14 ) & 0.23 ( 0.02 ) & 0.06 ( 0.05 ) & 0.04 ( 0.05 ) & 0.20 ( 0.03 ) & 0.10 ( 0.03 ) & 0.00 ( 0.01 ) + spitzer / irs spectra of sources showing the 7.25 , 7.40 , and 7.67 @xmath1 features .
the top source , w33a , is a massive yso of which a higher resolution ( @xmath32=800 ) iso / sws spectrum is shown .
the features are labeled with their ( possible ) identifications . ] [ f:7_8um ] [ sec:7_8 ] in addition to solid h@xmath3o and the 5 components discussed above , a number of weak features can be discerned in the 78 @xmath1 region of several low mass ysos , much like those identified toward massive protostars @xcite .
these features lie on top of a curved ` continuum ' created by the wings of the prominent 6 and 6.85 @xmath1features .
a high order polynomial ( up to 6@xmath54 ) had to be used to fit a local baseline to typically these wavelength regions : 7.007.14 , 7.37.35 , 7.477.50 , and 7.807.95 @xmath1 .
examples of the resulting features are shown in fig .
[ f:7_8um ] , where they are compared to the higher resolution ( @xmath55 ) iso / sws spectrum of the massive yso w33a .
the 7.25 @xmath1 feature is likely due to the hcooh molecule ( [ sec : hcooh ] ) .
the feature at 7.40 @xmath1 is possibly due to hcoo@xmath6 ( [ sec : hcoom ] ; @xcite ) , while the 7.67 @xmath1 feature is identified with ch@xmath14 @xcite .
[ sec : nh3ch3oh ] several sources show substructures at 9.0 and 9.7 @xmath1 within the strong silicate absorption band ( fig .
[ f:8_10um ] ) .
previous work indicated that these can be attributed to the umbrella mode of solid nh@xmath4 and the c - o stretching mode of ch@xmath4oh ( e.g. @xcite ) . to extract these and any other ice absorption features in the 8 - 12 @xmath1 range , the silicate absorption and possible blended emission needs to be accurately modeled .
this is deferred to a future publication ( s. bottinelli , in prep . ) , and here a local 4@xmath54 order polynomial continuum is fitted to the wavelength regions 8.25 - 8.75 , 9.23 - 9.37 , and 9.98 - 10.4 @xmath1 to extract the features ( fig .
[ f:8_10um ] ) .
column densities of ch@xmath4oh and nh@xmath4 derived from these and other features are further discussed in [ sec : ch3oh ] and [ sec : nh3 ] . and
ch@xmath4oh in the spectra of the low mass ysos hh 46 irs ( left panels ) and iras 15398 - 3359 ( right ) . in the top panels for each source
the dashed line indicates a local polynomial continuum , which was used to derive the optical depth plots in the lower panels . ]
[ f:8_10um ]
[ sec : id ] numerous carriers have been proposed to account for the various features in the 58 @xmath1 wavelength region @xcite . here
new observational constraints to possible carriers are discussed , using the present large source sample and wavelength coverage .
first , the contributions of the most securely identified ( ch@xmath4oh , nh@xmath4 ) or otherwise well constrained species ( h@xmath3co , hcooh , hcoo@xmath6 ) to the observed c1-c5 components are discussed ( [ sec : carr ] ) .
subsequently , the origin of the remaining absorption is constrained by correlating it with a number of other observables ( [ sec : corr ] ) .
[ sec : carr ] [ sec : ch3oh ] solid ch@xmath4oh was previously unambiguously detected by its c@xmath20h stretching mode at 3.54 @xmath1 ( e.g. @xcite ) and its c@xmath20o stretching mode at 9.7 @xmath1 @xcite . in our sample , a gaussian was fitted to the 9.7 @xmath1 feature on an optical depth scale and a ch@xmath4oh detection is claimed if the integrated optical depth has a larger than 3@xmath56 significance and if the peak position and fwhm are similar to previous detections , i.e. within 9.68 - 9.76 and 0.25 - 0.36 @xmath1 respectively .
the 3.54 @xmath1 band , when available , was treated similarly .
thus , ch@xmath4oh was positively identified in 12 low mass ysos ( table [ t : colden ] ; fig .
[ f : abun ] ) . of those ,
6 are based on the 9.7 @xmath1 band only .
column densities were calculated using integrated band strengths of 5.6@xmath57 and 1.6@xmath58 cm / molecule for the 3.54 and 9.7 @xmath1 bands , respectively @xcite . in agreement with previous studies @xcite ,
the ch@xmath4oh column varies considerably with respect to solid h@xmath3o : from low upper limits of a few percent ( 3@xmath56 ) in several sources , to detections of up to 30% in a few others . in the 58 @xmath1 region ,
the c@xmath20h deformation mode of ch@xmath4oh overlaps with the c3 component , and , at the derived column densities , contributes typically 5 - 20% to its strength , but up to 40% for svs 4 - 5 and gl7009s .
oh ( * panel a * ) , hcooh ( * b * ) , and nh@xmath8 ( * c * ) in percentage of h@xmath3o as a function of the 9.7 @xmath1 silicate band optical depth .
the solid ch@xmath4oh abundance is clearly more sight - line dependent than the hcooh abundance .
sources with the largest relative ch@xmath4oh column densities are labeled in the left panel .
open symbols with arrows refer to 3@xmath56 upper limits .
error bars are 1@xmath56 .
the nh@xmath8 abundance in the right panel was calculated assuming both the c3 and c4 components are due to nh@xmath8 , but after subtracting the contributions from ch@xmath4oh and h@xmath3co . ]
[ f : abun ] lrrrr l1448 irs1 & 0.47 ( 0.16 ) & @xmath1912.1 & @xmath5214.9 & 13.9 ( 4.6 ) + iras 03235 + 3004 & 14.48 ( 2.26 ) & @xmath192.7 & 4.2 ( 1.2 ) & 2.2 ( 0.9 ) + iras 03245 + 3002 & 39.31 ( 5.65 ) & @xmath191.2 & @xmath529.8 & 5.5 ( 0.7 ) + l1455 smm1 & 18.21 ( 2.82 ) & 3.3 ( 0.8 ) & @xmath5213.5 & 7.1 ( 1.6 ) + rno 15 & 0.69 ( 0.06 ) & @xmath195.8 & @xmath525.0 & 13.5 ( 2.4 ) + l1455 irs3 & 0.92 ( 0.37 ) & @xmath1915.1 & @xmath5212.5 & 8.9 ( 5.5 ) + iras 03254 + 3050 & 3.66 ( 0.47 ) & @xmath193.2 & @xmath524.6 & 8.2 ( 1.1 ) + iras 03271 + 3013 & 7.69 ( 1.76 ) & @xmath192.5 & @xmath525.6 & 5.8 ( 3.3 ) + iras 03301 + 3111 & 0.40 ( 0.04 ) & @xmath194.9 & @xmath5215.1 & 8.1 ( 3.2 ) + b1-a & 10.39 ( 2.26 ) & @xmath522.8 & @xmath521.9 & 7.7 ( 0.7 ) + b1-c & 29.55 ( 5.65 ) & 1.2 ( 0.3 ) & @xmath527.1 & 6.7 ( 0.3 ) + b1-b & 17.67 ( 3.20 ) & 3.1 ( 0.5 ) & 11.2 ( 0.7 ) & 2.9 ( 0.4 ) + b5 irs3 & 1.01 ( 0.09 ) & @xmath194.4 & @xmath528.1 & 15.3 ( 3.6 ) + b5 irs1 & 2.26 ( 0.28 ) & @xmath192.6 & @xmath523.7 & 5.2 ( 1.3 ) + l1489 irs & 4.26 ( 0.51 ) & @xmath192.9 & 4.9 ( 1.5 ) & 6.2 ( 0.8 ) + iras 04108 + 2803b & 2.87 ( 0.40 ) & @xmath193.1 & @xmath523.5 & 6.9 ( 0.9 ) +
hh 300 & 2.59 ( 0.25 ) & @xmath192.5 & @xmath526.7 & 5.5 ( 1.1 ) + dg tau b & 2.29 ( 0.39 ) & @xmath193.4 & @xmath525.7 & 9.3 ( 1.8 ) + hh 46 irs & 7.79 ( 0.77 ) & 2.7 ( 0.7 ) & 5.5 ( 0.3 ) & 6.3 ( 0.4 ) + iras 12553 - 7651 & 2.98 ( 0.56 ) & @xmath191.8 & @xmath523.0 & 8.4 ( 2.1 ) + iras 13546 - 3941 & 2.07 ( 0.20 ) & @xmath528.3 & @xmath523.9 & 3.6 ( 0.7 ) + iras 15398 - 3359 & 14.79 ( 3.95 ) & 1.9 ( 0.3 ) & 10.3 ( 0.8 ) & 3.9 ( 0.3 ) + elias 29 & 3.04 ( 0.30 ) & @xmath1911.3 & @xmath524.9 & 3.2 ( 1.0 ) + crbr 2422.8 - 342 & 4.19 ( 0.41 ) & @xmath191.0 & @xmath529.3 & 6.7 ( 0.7 ) + rno 91 & 4.25 ( 0.36 ) & @xmath193.2 & @xmath525.6 & 5.4 ( 0.6 ) + iras 17081 - 2721 & 1.31 ( 0.13 ) & @xmath192.4 & 3.3 ( 0.8 ) & 9.9 ( 1.8 ) + sstc2dj171122.2 - 272602 & 13.94 ( 2.82 ) & 3.0 ( 0.6 ) & @xmath521.3 & 6.4 ( 0.5 ) + 2massj17112317 - 2724315 & 19.49 ( 3.20 ) & 2.5 ( 0.5 ) & @xmath523.2 & 6.7 ( 0.4 ) + ec 74 & 1.07 ( 0.18 ) & @xmath193.2 & @xmath529.3 & 9.7 ( 2.8 ) + ec 82 & 0.39 ( 0.07 ) & @xmath192.5 & @xmath5214.2 & 10.2 ( 6.3 ) + svs 4 - 5 & 5.65 ( 1.13 ) & 4.6 ( 1.0 ) & 25.2 ( 3.5 ) & 3.6 ( 0.5 ) + ec 90 & 1.69 ( 0.16 ) & @xmath194.1 & 6.8 ( 1.6 ) & 4.4 ( 0.9 ) + ec 92 & 1.69 ( 0.14 ) & @xmath195.3 & 11.7 ( 3.5 ) & 2.9 ( 1.8 ) + ck4 & @xmath521.56 & & & + r cra irs 5 & 3.58 ( 0.26 ) & @xmath194.1 & 6.6 ( 1.6 ) & 5.1 ( 0.4 ) + hh 100 irs & 2.45 ( 0.24 ) & @xmath192.4 & @xmath529.7 & 11.6 ( 2.1 ) + cra irs7 a & 10.89 ( 1.92 ) & @xmath190.6 & @xmath523.8 & 7.0 ( 0.4 ) + cra irs7 b & 11.01 ( 1.97 ) & @xmath191.5 & 6.8 ( 0.3 ) & 5.8 ( 0.2 ) + cra iras32 & 5.26 ( 1.88 ) & @xmath199.8 & @xmath5218.1 & 15.9 ( 7.9 ) + l1014 irs & 7.16 ( 0.91 ) & 5.4 ( 1.3 ) & 3.1 ( 0.8 ) & 7.2 ( 1.0 ) + iras 23238 + 7401 & 12.95 ( 2.26 ) & @xmath191.5 & @xmath523.6 & 4.8 ( 0.7 ) + w3 irs5 & 5.65 ( 0.56 ) & @xmath190.4 & @xmath522.8 & 7.0 ( 0.4 ) + monr2 irs3 & 1.59 ( 0.15 ) & @xmath192.7 & @xmath525.6 & 25.9 ( 2.2 ) + gl989 & 2.24 ( 0.10 ) & @xmath192.2 & 3.1 ( 0.1 ) & 8.4 ( 0.7 ) + w33a & 12.57 ( 3.14 ) & 5.2 ( 0.7 ) & 14.7 ( 3.6 ) & 8.1 ( 0.8 ) + gl7009s & 11.31 ( 2.26 ) & 2.5 ( 0.7 ) & 31.3 ( 7.8 ) & 4.9 ( 0.7 ) + gl2136 & 4.57 ( 0.45 ) & 5.1 ( 1.2 ) & 8.5 ( 2.1 ) & 7.2 ( 0.9 ) + s140 irs1 & 1.95 ( 0.19 ) & @xmath522.6 & @xmath523.0 & 9.6 ( 1.0 ) + ngc7538 irs9 & 6.41 ( 0.64 ) & @xmath522.0 & 7.5 ( 1.8 ) & 5.7 ( 0.4 ) + elias 16 & 2.26 ( 0.16 ) & @xmath191.4 & @xmath522.3 & 5.2 ( 0.5 ) + ec 118 & 3.57 ( 0.35 ) & @xmath191.4 & @xmath524.7 & 11.6 ( 1.4 ) + [ sec : hcooh ] the main vibrational modes of formic acid ( hcooh ) occur in the 3 - 4 @xmath1 region ( o - h and c - h stretch ) , at 5.85 @xmath1 ( c = o stretch ) , at 7.25 @xmath1 ( c - h deformation ) , and at 8.1 @xmath1 ( c - o stretch ) .
previously , the 7.25 @xmath1 band was detected toward massive ysos , while the 5.85 @xmath1 band , in addition to h@xmath3o , could explain most of the 6 @xmath1 absorption @xcite .
the 7.25 @xmath1 band offers the most secure identification , even though its intrinsic strength ( @xmath59 cm / molecule ) is a factor of @xmath525 weaker than the 5.85 @xmath1 band .
it is least blended with other features and it is unambiguously detected ( @xmath60 ) toward 9 low mass ysos ( fig .
[ f:7_8um ] ; table [ t : colden ] ) . with the aid of laboratory analogs @xcite
, the observed strength of the 7.25 @xmath1 feature is used to determine the contribution of hcooh to the 6.0 @xmath1 band .
pure hcooh , and mixtures with h@xmath3o show 7.25 @xmath1features that are at least a factor of 2 wider than the observed band ( fwhm@xmath515 @xmath61 ) , and that peak at too short wavelengths .
@xcite and recently @xcite note that the band narrows and shifts to longer wavelengths in tertiary mixtures , specifically in h@xmath3o : ch@xmath4oh : hcooh=100:40:12 ( fig .
[ f : hcooh ] ) .
such a mixture would be consistent with the detections and upper limits of solid ch@xmath4oh ( typically @xmath50(ch@xmath4oh)@xmath62(hcooh ) toward ysos ; table [ t : colden ] ) .
for the purpose of this analysis the laboratory and interstellar data are matched by scaling their integrated 7.25 @xmath1 optical depths .
for these sources , the hcooh column densities are 1 - 5% with respect to solid h@xmath3o ( fig .
[ f : abun ] ) and 50 - 100% of component c1 can be explained by the c = o stretching mode of hcooh ( fig .
[ f : hcooh ] ) .
this wide range encompasses the large uncertainty in the band strength : the laboratory 5.85 @xmath1 band is stronger by 60% in h@xmath3o mixtures compared to pure hcooh . for many sources in which the 7.25 @xmath1 band is not detected ( @xmath63 ) , a stronger constraint on the hcooh column density is provided by the strength of the c1 feature .
relative hcooh abundances of a few % are derived by using the c = o band strength of the h@xmath3o - rich mixtures .
finally , in most cases neither the series of shallow , broad hcooh bands in the 3 - 4 @xmath1 o - h and c - h stretching mode region , nor a band at 8.1 @xmath1 , blended with the edge of the prominent 9.7 @xmath1 silicate feature , provide stronger constraints to the hcooh column density .
o stretching mode , the right panels the c - h deformation mode .
from top to bottom four ysos are shown with confirmed 7.25 @xmath1 absorption band detections ( indicated with vertical dotted lines in the right panels ) .
the histograms in the left panels represent component c1 of the interstellar spectra , i.e. after subtracting solid h@xmath3o and components c2 and c5 defined in [ sec:60 ] .
the dashed lines represent the spectra before c2 and c5 subtraction . in each panel
the thick smooth line represents the laboratory ice mixture hcooh : h@xmath3o : ch@xmath4oh=12:100:40 at @xmath64 k after removal of the h@xmath3o bending mode @xcite .
the laboratory spectra are scaled such that the integrated optical depth of the 7.25 @xmath1 band matches the observations .
note that the laboratory spectrum of the 7.25 @xmath1 band is wider than shown in @xcite .
this is due to a different baseline choise at the steep edge of the strong c - h deformation mode of ch@xmath4oh in the laboratory data . ]
[ f : hcooh ] [ sec : h2co ] as was noted for massive ysos ( e.g. @xcite ) , the c1 component shows a sharper edge on the short wavelength side than can be explained by hcooh alone .
the c = o stretching mode of h@xmath3co is a viable candidate , and , assuming an integrated band strength of @xmath65 cm / molecule @xcite , substantially better fits to component c1 are obtained by adding typically an abundance of 6% of h@xmath3co with respect to h@xmath3o ( fig .
[ f : labcombo ] ) .
accurate h@xmath3co abundances and uncertainty estimates are hard to establish due to the strong blend with hcooh .
this abundance estimate is consistent with the non - detection of the c - h stretching bands at 3.34 , 3.47 , and 3.54 @xmath1 and leads to a contribution of 10 - 35% from h@xmath3co to component c1 . at the quoted typical abundance ,
the c - h bending mode vibration of h@xmath3co at 6.68 @xmath1 contributes @xmath1915% to component c3 ( assuming @xmath66 cm / molecule from @xcite ) .
[ sec : nh3 ] the abundance of solid nh@xmath4 has long been a matter of debate , because all of its absorption bands overlap with much stronger bands of h@xmath3o and silicates .
a combined in - depth analysis of all these features is therefore required and accurate column densities will be presented elsewhere ( s. bottinelli et al . , in prep . ) . in this work ,
nh@xmath4 column densities accurate to within a factor of a few are sufficient .
the strength of the 9.0 @xmath1 umbrella mode is used to calculate the strength of the 6.16 @xmath1 n - h bending mode and its contribution to component c2 . the inflection seen at 9.0 @xmath1 in several sources
is fitted with a gaussian on optical depth scale and a detection of nh@xmath4 is claimed if the measured peak position and fwhm width are within 8.91 - 9.05 @xmath1 and 0.20 - 0.45 @xmath1 , respectively .
thus , nh@xmath4 is positively identified in 17 low mass ysos .
column densities were calculated using an integrated band strength of @xmath67 cm / molecule for nh@xmath4 mixed in an h@xmath3o ice @xcite , leading to values of typically 3 - 8% with respect to h@xmath3o . at these abundances ,
10 - 50% of the c2 component is due to solid nh@xmath4 .
[ sec : hcoom ] the formate ion ( hcoo@xmath6 ) is an interesting candidate to consider because its c = o stretching mode at 6.33 @xmath1 overlaps with the c2 component , its c - h deformation mode coincides with the 7.40 @xmath1 band detected in several sight - lines ( fig .
[ f:7_8um ] ) , and because it is chemically related to the hcooh species discussed above @xcite .
its identification is tentative , however , since acetaldehyde ( ch@xmath4hco ) has a feature at 7.40 @xmath1 as well , but not near the c2 component .
if hcoo@xmath6 were the carrier of the 7.40 @xmath1band , its column density would be @xmath50.3% with respect to solid h@xmath3o , an order of magnitude less than hcooh , and its contribution to component c2 would be @xmath68% . to illustrate that the independently derived column densities of solid hcooh , hcoo@xmath6 , h@xmath3co , nh@xmath4 , and ch@xmath4oh are indeed consistent with the 5 - 7 @xmath1 observations , laboratory spectra of these ices are scaled and subsequently added ( fig .
[ f : labcombo ] ) . while no attempt is made to match the band profiles in detail ( profiles are highly dependent on interactions in the ice matrix and a detailed analysis awaits further laboratory work ) , the agreement is encouraging .
a considerable fraction of the interstellar 5 - 8 @xmath1absorption consists of a complex blend of features from these simple ices .
in particular , they explain most of the observed strength of component c1 ( 50 - 100% ) , much of c2 ( 30 - 70% ) and some of c3 ( @xmath1930% ) . at the same time ,
[ f : labcombo ] also shows that component c5 and most of the prominent 6.85 @xmath1 absorption band ( c3 and c4 ) are not accounted for .
nh@xmath8 is a candidate for the latter ( [ sec : disc3c4 ] ) , but its column density can not be determined independently and thus this species is not included in this plot .
bands of the low mass yso hh 46 irs , after h@xmath3o subtraction ( @xmath50(h@xmath3o)=82.8@xmath69 @xmath70 ) .
laboratory spectra of ( from top to bottom ) solid ch@xmath4oh , ch@xmath14 , nh@xmath4 , hcoo@xmath6 , h@xmath3co , and hcooh were scaled to column densities derived independently elsewhere in the spectrum : 5.6% ( ch@xmath4oh ) , 2.2% ( ch@xmath14 ) , 12.7% ( nh@xmath4 ) , 0.4% ( hcoo@xmath6 ) , 6.0% ( h@xmath3co ) , and 2.0% ( hcooh ) with respect to h@xmath3o .
these spectra were subsequently added to the underlying , unidentified c5 absorption component ( dashed line ) and the result is shown as a solid thick line . clearly , the prominent 6.85 @xmath1 band is not explained by these particular species . ]
[ f : labcombo ] [ sec : corr ] to obtain further insight into the origin of the c1-c5 components , their strengths were normalized to the h@xmath3o column density for the entire sample and plotted in fig .
[ f : compcorrel ] . as a measure of the degree of correlation the ratio of the mean
values and standard deviations were calculated ( excluding upper limits ) , yielding values of 1.97 , 1.70 , 3.61 , 1.35 , and 0.87 for c1-c5 respectively .
thus , of all components , c3 correlates best with h@xmath3o , followed by c1 .
the c2 and c4 components show more scatter , and , most interestingly , their strengths are largest at the lowest h@xmath3o abundances .
component c5 correlates poorest with h@xmath3o , especially when taking into account the tight 3@xmath56 upper limits at the bottom end of the distribution . while , of all components ,
c5 is most susceptible to errors in the continuum determination , the c5/@xmath50(h@xmath3o ) variation is too large to be explained by that alone ( [ sec:60 ] ) . therefore , the large variations in the 6.0 @xmath1 excess ( @xmath44 ; fig .
[ f : h2oh2o ] ) can be ascribed primarily to variations in component c5 , and much less to c1 .
components ( fig .
[ f : decomp ] ) , normalized to @xmath50(h@xmath3o ) , plotted as a function of @xmath50(h@xmath3o ) .
error bars are 1@xmath56 , but open symbols with arrows attached refer to 3@xmath56 upper limits . at the top right of each panel
the ratio of the mean over the standard deviation of the sample ( excluding upper limits ) is indicated .
component c3 , i.e. the short - wavelength component of the 6.85 @xmath1 feature , correlates significantly better with @xmath71 than the other components do . of the components that produce the 6.0 @xmath1 excess
, c1 correlates better with @xmath72 than does c5 .
note the increased strengths of c2 and c4 at low @xmath73 . ]
[ f : compcorrel ] the different degrees of correlation of the c1-c5 components with h@xmath3o indicate that they have different origins . to further elucidate their nature ,
their strengths were correlated with a number of tracers of the chemical or physical processing history of the ices .
a well - known tracer is the ratio of broad and narrow 4.7 @xmath1 co ice components ( e.g. @xcite ) , often referred to as ` polar ' and ` apolar ' ices due to the mixing of co with high and low dipole moment species , respectively .
the apolar ices have a much lower sublimation temperature than the polar ices ( @xmath74 versus 90 k ) and thus their abundance ratio traces thermal processing .
higher temperatures are traced by the 15 @xmath1 co@xmath3 bending mode .
characteristic double peaks are visible in the crystalline ices produced after heating co@xmath3:h@xmath3o mixtures to @xmath75 k @xcite , and the fraction of crystalline co@xmath3 ice so derived for many sources in our sample is presented in @xcite .
actual dust temperatures can be derived from far - infrared colors , but to trace the coldest dust components ( @xmath7620 k ) photometry at wavelengths @xmath77 @xmath1 is required . for massive ysos
this was provided by iso / lws spectra , and the far - infrared color was found to correlate with indicators of thermal processing @xcite .
unfortunately , such photometry is not available for most of our sample of low mass ysos .
tracers of ultraviolet radiation and cosmic ray fluxes are even harder to obtain . traditionally , the strength of the 4.62 @xmath1 absorption band has been used for this , since the carrier , an xcn-bearing species , was thought to be produced by the effects of uv photons or cosmic rays ( e.g. @xcite ) .
recent work shows , however , that the potential carrier ocn@xmath6 can be produced by heating an hnco - containing ice formed at cold conditions by grain surface chemistry @xcite .
thus , in this case the strength of the 4.62 @xmath1 band is in fact another tracer of thermal evolution .
band ) plotted against possible ice or dust temperature tracers . along the @xmath78axis are plotted : ( * panel a * ) a measure of the solid h@xmath3o abundance , ( * panel b * ) a measure of the solid co abundance , ( * panel c * ) the polar / apolar solid co ratio , ( * panel d * ) the fraction of crystalline co@xmath3 measured from the 15 @xmath1 bending mode , and ( * panel e * ) the relative abundance of the carrier of the 4.62 @xmath1 feature , ocn@xmath6 .
low mass ysos are indicated by filled dots , massive ysos by circles , and background stars by star symbols .
panels ( a ) and ( b ) show that sources with prominent c4 components have low ice ( h@xmath3o and co ) abundances . ]
[ f : corr685 ] the prominent 6.85 @xmath1 absorption band consists of independent short and long - wavelength components ( c3 and c4 ; fig . [ f : decomp ] ) , whose integrated optical depth ratio varies considerably ( c4/c3=0.5 - 4 ) .
this essentially reflects abundance variations of the carrier of the c4 component , because c4 correlates more poorly with the h@xmath3o column density than does the c3 component ( fig .
[ f : compcorrel ] ) .
the c4/c3 ratio is plotted against the observables discussed above , and while all of the observables show a large spread at c4/[email protected] , sources with stronger c4 components all have low solid h@xmath3o and co abundances ( panels ( a ) and ( b ) of fig . [
f : corr685 ] ) .
these include the low mass ysos iras 03301 + 3111 and dg tau b , and the massive ysos mon r2 irs3 and s140 irs1 .
the weakness of the ice bands in these particular sources hampers the correlation study with some of the above - mentioned observables : the ratio of the polar / apolar solid co components and the fraction of crystalline co@xmath3 can not be determined if these bands are not or weakly detected .
similarly , only a handful of ocn@xmath6 abundance determinations are available for these sources .
it is thus concluded that the c4/c3 ratio is enhanced in environments that have low ice abundances .
this is in agreement with earlier work on massive ysos @xcite , which showed an increase of the ratio in lines of sight with warmer dust temperatures .
potential carriers of the c4 and c3 components are discussed in [ sec : disc3c4 ] .
( h@xmath3o ) plotted against the same ice or dust temperature tracers as in fig . [ f : corr685 ] .
possibly , enhanced polar co ( c ) fractions are observed at larger c5 strengths . ]
[ f : corrc5 ] the strength of the enigmatic c5 component varies widely for both low and high mass ysos . toward massive ysos , a trend of increasing strength of the c5 component with the ocn@xmath6 abundance ( 4.62 @xmath1feature )
was claimed @xcite , but in the current sample this trend is rather weak and is limited by the small number of sources with ocn@xmath6 detections ( panel ( e ) in fig . [
f : corrc5 ] ) .
a stronger trend is observed with the ratio of the polar and apolar co components : sources with strong c5 components show larger polar / apolar co ratios ( panel ( c ) in fig .
[ f : corrc5 ] ) .
no correlations are observed with the h@xmath3o , co , and crystalline co@xmath3 abundances .
potential carriers are further discussed in [ sec : disc5 ]
[ sec : dis ] long - standing key questions in astrochemistry concern the composition of the interstellar ices and the degree to which the composition of the ices is altered by heating and by the impact of energetic photons and cosmic rays .
the 3 - 30 @xmath1 spectra presented in this paper extend studies addressing these questions from the previously well analyzed massive ysos to low mass ysos , as well as to background stars tracing quiescent cloud material .
the 58 @xmath1 region exhibits a very rich absorption spectrum , which , apart from the o - h bending mode of h@xmath3o , consists of at least 5 independent components . while most of component c1 , much of c2 and part of component c3 is due to the simple species ch@xmath4oh , hcooh , nh@xmath4 , and h@xmath3co , the origin of a substantial amount of absorption is still unclear , and this is discussed in more detail below .
[ sec : disc3c4 ] the 6.85 @xmath1 band was decomposed into ` blue ' and ` red ' components ( c3 and c4 ) .
while c3 correlates tightly with the h@xmath3o abundance , the relative strength of the c4 component increases at the lowest h@xmath3o abundances ( fig .
[ f : compcorrel ] ) .
this could be explained by the formation of a new species as the ices are heated and sublimate or it could be a pre - existing species that is less volatile than h@xmath3o whose absorption band ( c4 ) prevails at lower h@xmath3o abundances . in the first scenario heating must play a role , and in the second it may .
the detection of the c4 component in all ysos and toward background stars favors the second scenario , in which the carrier is not formed by heating for its carrier exists already in the earliest stages of chemical evolution , i.e. in quiescent clouds .
higher dust temperatures in the vicinity of ysos lead to sublimation and lower h@xmath3o and co abundances ( fig .
[ f : corr685 ] ) and the abundance of the carrier of the c3 component is reduced more rapidly than that of the c4 component . does the carrier of the c4 component survive in the diffuse interstellar medium ?
a band at 6.85 @xmath1 is detected in lines of sight toward the galactic center ( gc ) , tracing a large column of diffuse dust @xcite .
however , its depth relative to the 9.7 @xmath1 silicate band ( @xmath80 ) is a factor of 5 less compared to the c4 component in the ysos ( factor of 10 compared to c3 ) .
the weak 6.85 @xmath1 band observed toward the gc sources is attributed to the c - h asymmetric deformation mode of aliphatic hydrocarbons @xcite .
such aliphatic hydrocarbons have strong c - h stretching modes with distinct peaks at 3.38 , 3.42 , and 3.48 @xmath1 . a band at 3.47 @xmath1 is commonly detected toward ysos @xcite , but it does not show these distinct peaks and its depth relative to the 6.85 @xmath1 band is a factor of 3 weaker compared to the gc sources .
thus , although the carrier of the c4 component observed in our sample is ( probably ) less volatile than ices , it does not survive in the diffuse ism .
conversely , the hydrocarbons observed in the diffuse ism are not the carrier of the c4 component toward ysos . in a scenario proposed by @xcite
both the c3 and c4 components arise from an nh@xmath8-containing ice .
its abundance would be typically 7% with respect to h@xmath3o ( table [ t : colden ] and fig .
[ f : abun ] ) , assuming the integrated band strength of @xmath81 cm / molecule measured in @xcite .
nh@xmath8 ions can be produced either by acid - base reactions at low temperature ( e.g. from nh@xmath4:hnco at 10 k ; @xcite ) or by ultraviolet or cosmic ray irradiation of simple mixtures ( e.g. h@xmath3o : co@xmath3:nh@xmath4:o@xmath3 ) .
the nh@xmath8 band that is produced shifts to longer wavelengths upon heating , i.e. from the peak wavelength of the c3 component to the peak wavelength of the c4 component .
after additional energetic processing and h@xmath3o sublimation a salt is formed .
salts do not sublimate until temperatures of @xmath5200 k ( e.g. @xcite ) , significantly higher than the h@xmath3o sublimation temperature of 90 k in space .
the processed laboratory spectra show absorption at 6.0 @xmath1 due to anions ( no@xmath82 , no@xmath83 , and hco@xmath83 ) , some of which overlaps with the c2 components centered at @xmath56.15 @xmath1 that are observed in the ysos with strong c4 components ( fig . [
f : resid ] ; [ sec : disc2 ] ) . while these properties are consistent with the observations of ysos , the presence of the c4 component toward all ysos and background stars is hard to explain .
in the experiments of @xcite an nh@xmath8 temperature of almost 200 k is required to match the peak position of this component .
such a high temperature can be excluded for many sources in the sample .
perhaps the impact of cosmic rays is responsible for the shift in peak position .
subsequent heating by nearby protostars leads to the observed increase in the c4/c3 ratio and eventually to the formation of a salt .
[ sec : disc5 ] the enigmatic c5 component that stretches from 5.8 to @xmath58.0 @xmath1is not due to nh@xmath4 , h@xmath3co , hcooh , hcoo@xmath6 , or ch@xmath4oh , because these species do contribute to components c1-c4 ( [ sec : carr ] ) . also , no correlation is found between the strengths of the c5 and the 6.85 @xmath1 red ( c4 ) components .
therefore , these have unrelated origins .
heating might be involved in the formation of the carrier(s ) of the c5 component ( fig .
[ f : corrc5 ] ; [ sec : corr ] ) .
previously , three different origins were considered in the literature .
first , it should be noted that the shape of the average c5 component resembles that of the spectrum of solid h@xmath3o ( fig .
[ f : compex ] ) .
it is particularly reminiscent of heated h@xmath3o ( @xmath84 k ) , because of the weakness of the 6.0 @xmath1 peak with respect to the long wavelength wing ( e.g. @xcite ) . while the shape of the 3 @xmath1 band indicates that heated ices are present along several lines of sight ( e.g. , rno 91 ) , its depth is , by the very definition of the c5 component , insufficient .
possibly , due to the effects of light scattering at the shorter wavelengths , a more direct path to the warm inner regions of the yso is observed at 6 @xmath1 compared to 3 @xmath1 .
monte carlo radiative transfer models of a circumstellar disk show that at inclination angles of @xmath85 the depth of the 6.0 @xmath1 h@xmath3o band is enhanced by a factor of 10 with respect to that at 3.0 @xmath1 @xcite . at higher inclinations ,
the enhancement is a factor of 34 , while at lower inclinations the effect is negligible .
thus , the c5 component may be attributed to heated h@xmath3o ice for some sources in our sample ; for example , rno 91 is known to possess an inclined disk @xcite and has one of the most prominent c5 components . at the other extreme ,
the weakest c5 components are observed toward background stars ( table [ t : tau ] ) .
indeed , in these lines of sight the effect of scattering on the relative h@xmath3o band depths is expected to be small due to the uniformity of the foreground clouds . for many ysos with c5 absorption
, however , independent observational evidence indicates that the scattering scenario is unlikely .
for example , the ch@xmath4oh column density derived from the 3.53 @xmath1 c h stretching mode agrees well with that derived from the 9.7 @xmath1 c h bending mode ( [ sec : ch3oh ] ; @xcite ) .
also , for high mass ysos the 4.25 and 15 @xmath1 bands provided solid co@xmath3 column densities that were in excellent agreement with each other @xcite , and in one line of sight also with the 2.7 @xmath1 co@xmath3 combination mode @xcite .
thus , for these ysos other explanations are required .
a second explanation of the c5 component could be the blend of negative ions suggested in @xcite .
ammonium ( nh@xmath8 ) and the anions hco@xmath83 , no@xmath83 , and no@xmath82 are produced after uv photolysis or cosmic ray bombardment of h@xmath3o / co@xmath3/nh@xmath4/o@xmath3 ice mixtures .
nh@xmath8 has a strong band at 6.85 @xmath1 ( i.e. components c3 and c4 in this work ) , but the bands of the anions overlap such that only one very broad feature is observed in the 5 - 8 @xmath1range . while one may expect that in this scenario the c3/c4 and c5 components correlate well with each other , the observations show they do not .
notably , the background stars have strong 6.85 @xmath1components and little c5 absorption . the nh@xmath8 ion can be produced after warming up a simple ice mixture too , however , while the hco@xmath83 , no@xmath83 , no@xmath82 anions may need more energetic input @xcite .
this may explain the large variation in the strength of the c5 component .
third , extensive processing of simple ices by uv light or cosmic rays produces an organic refractory residue @xcite , whose blended o - h and c - h stretching and bending modes produce a feature similar to the c5 component @xcite . both this scenario and the blend of ions discussed above require a correlation of the strength of the c5 component with independent indicators of the presence of strong uv or cosmic ray fluxes , which are currently unavailable .
also , both the organic residue and the salts are less volatile than the ices they sublimate at temperatures @xmath19500 k in the laboratory ( @xcite , @xcite ; @xmath86=180 k for h@xmath3o ice ) .
therefore , one may expect c5 to be the main absorption component in the 58 @xmath1 region toward some ysos where all h@xmath3o ices have evaporated .
although in several sources the c5 component is the dominant contributor , it always is observed together with h@xmath3o ( except possibly toward l1448 irs1 ) .
some h@xmath3o may be trapped in the salt or organic residue at high temperatures , and some may be present in unprocessed foreground material .
it is also important to note that , similar to c4 ( [ sec : disc3c4 ] ) , the c5 component is not observed in the diffuse ism , confirming that their carrier(s ) are not as refractory as silicates ( @xmath871500 k ) .
finally it is noted that the c5 component may have a different origin in different lines of sight .
an in - depth analysis of the physical conditions in specific lines of sight is required to discriminate between the effects of disk orientation and the various ice processing mechanisms .
[ sec : disc2 ] the c2 component spans the 6.0 - 6.4 @xmath1 wavelength range and is likely due to at least three different carriers .
the n - h bending mode of nh@xmath4 at 6.15 @xmath1 contributes 10 - 50% to this band ( [ sec : nh3 ] ) .
another likely contributor is h@xmath3o .
although the spectrum of pure , amorphous h@xmath3o was subtracted from the 5 - 8 @xmath1region before determining the various component strengths and profiles , the spectrum of interstellar h@xmath3o may deviate from that of a pure , amorphous h@xmath3o ice .
the o - h bending mode of h@xmath3o shifts to longer wavelengths , becomes significantly narrower , and its peak strength increases if the hydrogen bonding network in bulk amorphous h@xmath3o ice is broken by the presence of other species @xcite .
co@xmath3 is a realistic species , because it is abundant ( 15 - 40% with respect to h@xmath3o ) and a significant fraction is known to be mixed with h@xmath3o @xcite . at sufficiently high co@xmath3 concentrations , the h@xmath3o is located in small clusters , leading to monomer , dimer , and multimer bondings that produce a strong peak at 6.12 @xmath1 .
the optical depth of the peak has been calculated for the yso b5 irs1 @xcite and the background star elias 16 @xcite assuming the h@xmath3o ice along these lines of sight consists of two components : pure h@xmath3o and the mixture h@xmath3o : co@xmath3=2:1 .
it was found that when 25% of the h@xmath3o is present in the latter , the 3 and 6 @xmath1 bands of h@xmath3o and the 15 @xmath1 band of co@xmath3 are fitted simultaneously .
comparison with the c2 component in these sources shows that 20 - 50% could be due to this effect .
thus the short wavelength side of the c2 component is both due to h@xmath3o : co@xmath3 mixtures and due to nh@xmath4 . for the long - wavelength side ,
the c = o stretching mode of the hcoo@xmath6 anion was previously considered , contributing @xmath1920% to the c2 strength ( [ sec : hcoom ] ) .
other anions , such as no@xmath82 , no@xmath83 , and hco@xmath83 ( [ sec : disc3c4 ] ) , may contribute as well .
these particular anions may be produced in the energetic processing of simple ices @xcite , along with nh@xmath8 .
indeed , one of the most interesting observational aspects of the c2 component is that it becomes more pronounced for sources with stronger c4 components and lower h@xmath3o abundances ( figs .
[ f : resid ] and [ f : compcorrel ] ) .
overall the two components do not correlate very well , however , which is consistent with the result that three carriers of different nature contribute to the c2 component .
[ sec : dischem ] ch@xmath4oh shows larger abundance variations with respect to h@xmath3o than other solid state species discussed in this work ( hcooh , h@xmath3co , nh@xmath4 ) .
tight 3@xmath56 upper limits of a few % are reported for a number of ysos , but detections above the 10% level for others ( table [ t : colden ] and fig .
[ f : abun ] ) .
the solid hcooh abundance is typically a factor of 2 - 3 lower than ch@xmath4oh and does not show such large variations .
if both ch@xmath4oh and hcooh are formed from co by grain surface reactions , the lower hcooh abundances are understood by the low atomic o / h ratio in the accreted gas , and by the competition for o to react to co@xmath3 rather than with the formyl radical ( hco ) to form hcooh @xcite .
indeed , large solid co@xmath3 abundances ( 15 - 40% with respect to h@xmath3o ) are observed for a wide variety of ysos @xcite .
grain surface chemistry does indeed appear a more likely formation mechanism than uv processing of the ices .
first , while processing of simple ices by uv photons leads to ch@xmath4oh formation , this process is not efficient enough to explain the enhanced abundances toward massive ysos @xcite , while for low mass ysos the uv flux is likely too low regardless .
second , hcooh and ch@xmath4oh are detected even in the most obscured sight - lines in the sample .
in fact , enhanced ch@xmath4oh abundances are detected exclusively in some of the most embedded ysos : the low mass ysos svs 4 - 5 , b1-b , and iras 15398 - 3359 and the massive ysos w33a and gl7009s @xcite .
svs 4 - 5 is known to trace the envelope of the foreground class 0 object smm4 @xcite . on the other hand , toward a few background stars tracing quiescent cloud material a low upper limit to the ch@xmath4oh abundance was determined ( table [ t : colden ] ; @xcite ) .
thus it is suggested that , in the grain surface chemistry scenario , the ch@xmath4oh abundance is enhanced at a time during the collapse ( of some clouds ) when the atomic o needed to form h@xmath3o and co@xmath3 is exhausted , while still sufficient co and h are available to form ch@xmath4oh @xcite . while grain surface chemistry likely forms the simple neutral species ,
different mechanisms are required to produce the ions possibly responsible for several of the components in the 5 - 8 @xmath1 range .
if the observed c3 and c4 components are indeed due to nh@xmath8 , its abundance is typically 7% with respect to h@xmath3o ( fig .
[ f : abun ] ) .
the c4 component is observed even in the coldest sight - lines .
indeed , low temperature acid - base chemistry can efficiently produce nh@xmath8 @xcite .
the abundance of negative counter ions needs to be addressed . while the c5 component could hide the necessary negative ions @xcite ,
its strength does not correlate with the c3/c4 components .
further investigations on the presence of ions in the ices are crucial to understand the ice chemistry .
their physical properties are different from neutral species .
ions have higher sublimation temperatures , and thus the salts that are formed after the volatile species ( e.g. h@xmath3o and co@xmath3 ) have sublimated are able to survive at lower extinctions , and then are subjected to harsher conditions , forming more complex species .
for example , the biologically interesting species urea ( h@xmath3nconh@xmath3 ) is formed after the uv irradiation of a nh@xmath8ocn@xmath6 salt @xcite .
[ sec : concl ] the present work extends the study of ices over the full 3 - 20 @xmath1wavelength range from the previously well studied massive ysos ( @xmath88l@xmath17 ) to low mass ysos ( @xmath51l@xmath17 ) .
the following new insights were obtained : * absorption features are detected at 6.0 and 6.85 @xmath1 in all sources , including high- and low - mass ysos and background stars tracing quiescent cloud material .
an empirical decomposition shows that the 5 - 7 @xmath1 absorption complex consists , in addition to the h@xmath3o bending mode , of a combination of at least 5 independent absorption components . * in a subset of sources additional weak features are detected at 7.25 , 7.40 , and at 9.0 and 9.7 @xmath1 on top of the prominent si - o stretching band of silicates .
these features are associated with ch@xmath4oh , hcooh , nh@xmath4 , and possibly hcoo@xmath6 and indicate abundances of 1 - 30% , 1 - 5% , 3 - 8% , and 0.3% with respect to h@xmath3o , respectively .
the large source - to - source solid ch@xmath4oh abundance variations are likely a result of the conditions at the time of grain surface chemistry . *
component c1 ( 5.7 - 6.0 @xmath1 ) is mostly explained by solid hcooh and h@xmath3co at abundances with respect to solid h@xmath3o of typically 1 - 5% and @xmath56% , respectively .
* component c2 ( 6.0 - 6.4 @xmath1 ) also likely arises from a blend of several species .
solid nh@xmath4 can account for 10 - 50% of the absorption .
a similar amount can be attributed to absorption by monomers , dimers , and small multimers of h@xmath3o mixed with a substantial amount of co@xmath3 .
the long - wavelength side of c2 is potentially due to anions produced by acid - base chemistry ( e.g. hcoo@xmath6 ) or energetic processing .
indeed , a weak correlation with c4 suggests the carriers of these bands ( possibly salts ) are related . * components c3 and c4 , the 6.85 @xmath1 band ,
show the same characteristics for low mass ysos and background stars as was previously found for massive ysos by @xcite .
their ratio is empirically found to be dependent on dust temperature .
the carrier of c4 is likely less volatile than that of c3 and than h@xmath3o , but it is not observed in the diffuse ism .
these characteristics are also consistent with both components being due to nh@xmath8 ( ammonium ion ) .
the detection of strong 6.85 @xmath1 bands toward deeply embedded ysos and background stars requires a production at low temperature ( acid - base chemistry or processing by cosmic rays ) , given the lack of heating sources and stellar uv fields .
* the origin of the very broad component c5 ( 5 - 8 @xmath1 ) is least understood .
it is quite strong in several low and high mass ysos , absent in others and so far undetected toward background stars .
it is possibly related to thermal processing , as a weak correlation with the ratio of the polar / apolar co components is observed and its shape resembles that of warm h@xmath3o .
a blend of absorption by ions , as proposed by @xcite , which would require both heating and energetic processing of the ices , can not be excluded .
the latter could also lead to an organic residue , whose shape resembles that of the c5 component as well . *
weak correlations are found between the absorption components in the 5 - 8 @xmath1 range and tracers of physical conditions . a more thorough understanding of the conditions , the source geometry and local radiation fields in specific lines of sight is required to further constrain the nature of the carriers ( in particular for components c2-c5 ) and the importance of thermal and energetic processing .
to what degree does scattering play a role ?
what are the dust temperatures , cosmic ray fluxes , uv radiation fields and time scales as a function of distance along the line of sight
? measurements of source fluxes above wavelengths of 100 @xmath1 with the herschel space observatory will be valuable , as well as ground - based measurements of the 4.62 @xmath1 band of ocn@xmath6 in more lines of sight .
in addition , further laboratory work , in particular on the effect of cosmic rays on the ices , especially for the identification of components c3 and c4 , is required . finally , study of the 5 - 8 @xmath1 features toward a larger sample of background stars is crucial to make further progress in the identification and processing history of interstellar ices .
we thank the anonymous referee for thoughtful comments on the manuscript and alberto noriega - crespo for help in identifying the emission lines .
support for this work , part of the spitzer legacy science program , was provided by nasa through contracts 1224608 , 1230779 , 1230782 , 1256316 , and 1279952 issued by the jet propulsion laboratory , california institute of technology , under nasa contract 1407 .
astrochemistry in leiden is supported by a nwo spinoza grant , a nova grant , and by the european research training network the origin of planetary systems " ( planets , contract number hprn - ct-2002 - 00308 ) .
the work of jkj is supported by nasa origins grant nag5 - 13050 .
kmp is supported by nasa through hubble fellowship grant # 01201.01 awarded by the space telescope science institute , which is operated by the association of universities for research in astronomy , inc .
, for nasa , under contract nas 5 - 26555 .
we thank the lorentz center in leiden for hosting several meetings that contributed to this paper .
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 . | with the goal to study the physical and chemical evolution of ices in solar - mass systems , a spectral survey is conducted of a sample of 41 low luminosity ysos ( @xmath0 ) using 538 @xmath1 spitzer space telescope and 34 @xmath1 ground - based spectra . the sample is complemented with previously published spitzer spectra of background stars and with iso spectra of well studied massive ysos ( @xmath2 ) .
this paper focuses on the origin of the prominent absorption features in the 5 - 8 @xmath1 spectral region .
the long - known 6.0 and 6.85 @xmath1 bands are detected toward all sources , with the class 0-type low mass ysos showing the deepest bands ever observed . in almost all sources
the 6.0 @xmath1 band is deeper , by up to a factor of 3 , than expected from the bending mode of pure solid h@xmath3o , based on the optical depths of the 3.0 @xmath1 stretching and 13 @xmath1 libration modes .
the depth and shape variations of the remaining 57 @xmath1 absorption indicate that it consists of 5 independent components , which , by comparison to laboratory studies , must be from at least 8 different carriers .
together with information from the 3 - 4 @xmath1 spectra and the additionally detected weak 7.25 , 7.40 , 9.0 , and 9.7 @xmath1 features it is argued that overlapping bands of simple species are responsible for much of the absorption in the 5 - 7 @xmath1 region , at abundances of 1 - 30% for ch@xmath4oh , 3 - 8% for nh@xmath4 , 1 - 5% for hcooh , @xmath56% for h@xmath3co , and @xmath50.3% for hcoo@xmath6 with respect to solid h@xmath3o .
the 6.85 @xmath1 band likely consists of one or two carriers , of which one is less volatile than h@xmath3o because its abundance relative to h@xmath3o is enhanced at lower h@xmath3o/@xmath7 ratios .
it does not survive in the diffuse interstellar medium ( ism ) , however .
the similarity of the 6.85 @xmath1 bands for ysos and background stars indicates that its carrier(s ) must be formed early in the molecular cloud evolution .
if an nh@xmath8 salt is the carrier its abundance with respect to solid h@xmath3o is typically 7% , and low temperature acid - base chemistry or cosmic ray induced reactions must have been involved in its formation .
possible origins are discussed for the carrier of an enigmatic , very broad absorption between 5 and 8 @xmath1 .
it shows large depth variations toward both low- and high - mass ysos .
weak evidence is found that it correlates with temperature tracers .
finally , all the phenomena observed for ices toward massive ysos are also observed toward low mass ysos , indicating that processing of the ices by internal ultraviolet radiation fields is a minor factor in the early chemical evolution of the ices . |
the problem of the masses and mixing of neutrinos ( see refs .
@xcite ) is the first priority problem of neutrino physics .
many experiments searching for neutrino oscillations and neutrinoless double - beta decay and investigating the high - energy part of the tritium beta - spectrum are going on or are under preparation . at present ,
indications that neutrinos are massive and mixed have been found in solar neutrino experiments ( homestake @xcite , kamiokande @xcite , gallex @xcite , sage @xcite and super - kamiokande @xcite ) , in atmospheric neutrino experiments ( kamiokande @xcite , imb @xcite , soudan @xcite and super - kamiokande @xcite ) and in the lsnd experiment @xcite . from the analyses of the data of these experiments in terms of neutrino oscillations
it follows that there are three different scales of neutrino mass - squared differences : @xmath2 where @xmath3 is the neutrino mass - squared difference relevant for short - baseline ( sbl ) experiments , whose allowed range is determined by the positive results of the lsnd experiment .
the two possibilities for @xmath4 correspond , respectively , to the msw @xcite and to the vacuum oscillation solutions of the solar neutrino problem .
thus , at least four light neutrinos with definite masses ( with only three active flavor neutrinos , @xmath5 , @xmath6 , @xmath7 ) must exist in nature in order to accommodate the results of all neutrino oscillation experiments .
this means that there exists at least one non - interacting sterile neutrino @xcite . in this case
, a left - handed neutrino field @xmath8 ( @xmath9 ) is a mixture of the left - handed components @xmath10 of the four fields of neutrinos with definite masses @xmath11 ( @xmath12 ) : @xmath13 where @xmath1 is the @xmath14 unitary mixing matrix . in ref
. @xcite we considered the schemes with four massive neutrinos and we have shown that from six possible mass spectra of four massive neutrinos only the following two mass spectra with two pairs of close masses separated by a gap of about 1 ev ( the `` lsnd gap '' ) are compatible with all existing data ( see also ref .
@xcite ) : @xmath15 in scheme a , @xmath16 and @xmath17 are the mass - squared differences relevant for atmospheric and solar neutrino oscillations , respectively , whereas in scheme b we have @xmath18 and @xmath19 . in both schemes
@xmath20 is relevant for neutrino oscillations in short - baseline experiments . in this paper
we study the compatibility of these two schemes with an upper bound the schemes a and b are trivially compatible with bbn . ]
@xmath21 ( see ref .
@xcite ) on the effective number @xmath22 of light neutrinos relevant in big - bang nucleosynthesis ( bbn ) ( see , for example , ref .
we derive the constraints on some elements of the neutrino mixing matrix @xmath1 that follow from @xmath21 ( see also ref .
@xcite ) and we discuss some possibilities to check these constraints in future short - baseline ( sbl ) and long - baseline ( lbl ) neutrino oscillation experiments .
let us notice that there was recently some discussion in the literature about the validity of the upper bound @xmath21 ( see ref .
this discussion was initiated by conflicting results of different measurements of the deuterium abundance in high - redshift hydrogen clouds @xcite .
the situation is not completely clarified , but there seems to exist now a strong case in favor of @xmath21 @xcite .
let us define the quantities @xmath23 okada and yasuda @xcite have shown that the bound @xmath21 implies stringent limits on @xmath24 in scheme a and @xmath25 in scheme b. this means that the solar neutrino deficit is explained by small angle msw @xmath26 transitions transitions ( see refs .
@xcite ) . ] and the atmospheric neutrino anomaly by @xmath27 transitions disappearance were found . ]
( see ref .
@xcite ) . in this paper
we have reanalyzed the constraints on the parameter @xmath24 that follow from bbn .
we have obtained more stringent upper bounds on @xmath24 in scheme a and @xmath25 in scheme b than those presented in ref .
@xcite and we have investigated the effects of these constraints for sbl and lbl neutrino oscillations , some of which can be tested in the near future . the plan of the paper is as follows . in section [ sec2 ]
we summarize the formalism of sbl neutrino oscillations in the schemes a and b ( for more details see refs .
@xcite ) . in section [ sec3 ]
we investigate in detail the dependence of @xmath22 on @xmath24 . from the bbn upper bound on @xmath22 we obtain upper bounds on @xmath24 in scheme a and @xmath25 in scheme b. in sections [ sec4 ] and [ sec5 ] we obtain various general relations between sbl and lbl oscillation amplitudes and @xmath24 .
these relations can be useful to check the bbn constraints and to test the hypothesis of existence of a sterile neutrino . in section [ sec5 ]
we also introduce a new parameterization of the mixing matrix that is suitable for the schemes under consideration .
in the schemes under consideration the probability of @xmath28 transitions ( @xmath29 ) and the survival probability of @xmath30 in sbl experiments are given by @xcite @xmath31 with the oscillation amplitudes @xmath32 where the index @xmath33 runs over the values @xmath34 or @xmath35 . in the following
, we will consider values of @xmath3 in the wide range @xmath36 the direct comparison of the regions in the @xmath37@xmath3 plane allowed at 90% cl by the results of the lsnd experiments with the regions excluded at 90% cl by the results of other sbl experiments indicate a considerably smaller allowed range for @xmath3 : @xmath38 the results of the combined analysis of the data of all sbl experiments presented in ref .
@xcite show that the lower bound @xmath39 is robust ( if the lsnd indication is correct ) , whereas values of @xmath3 larger than @xmath40 maybe are not excluded .
taking into account the results of solar and atmospheric neutrino experiments , from the exclusion curves of sbl disappearance experiments at each possible value of @xmath3 we have @xcite @xmath41 where @xmath42 and @xmath43 is the upper bound for the amplitude @xmath44 that can be obtained from the exclusion plots of sbl reactor and accelerator disappearance experiments . from the 90% cl exclusion curves of the bugey @xcite @xmath45 reactor experiment and of the cdhs @xcite and ccfr @xcite @xmath46
accelerator experiments it follows that @xmath47 is small ( @xmath48 ) for @xmath49 in the range ( [ widerange ] ) and @xmath50 for @xmath51 ( see fig . 1 of ref .
in this section we derive the bbn constraints on the elements of the mixing matrix in the two schemes a and b. we follow the standard arguments @xcite leading to the bbn constraints on the elements of the mixing matrix @xmath1 if a sterile neutrino exists . according to the standard bbn scenario ( see , for example , ref .
@xcite ) a lepton asymmetry can be neglected and the active neutrinos are in chemical equilibrium .
thus , the following considerations are valid for a temperature range @xmath52 mev with the decoupling temperature @xmath53 mev for electron neutrinos and @xmath54 mev for muon and tau neutrinos . in this temperature region ,
the effective potentials of neutrinos due to coherent forward scattering in the primordial plasma are given by @xcite @xmath55 where @xmath56 is the neutrino momentum , which henceforth will be replaced by its temperature average @xmath57 , and @xmath58 .
the propagation of neutrinos is governed by the effective hamiltonian @xcite @xmath59 where @xmath60 are the effective energy eigenvalues of neutrinos in matter and @xmath61 is the corresponding effective mixing matrix .
the amount of sterile neutrinos present at nucleosynthesis can be calculated using the differential equation @xcite @xmath62 where @xmath63 is the number density of the sterile neutrino relative to the number density of an active neutrino in equilibrium and @xmath64 are the collision rates of the active neutrinos , including elastic and inelastic scattering @xcite , given by @xmath65 let us notice some important characteristics of eq.([master ] ) : 1 .
the probabilities for @xmath66 transitions are averaged over the collision time @xmath67 and also @xmath63 has to be considered as such an averaged quantity .
2 . this equation should approximately hold for non - resonant and adiabatic resonant neutrino transitions , as long as the active neutrinos are in chemical equilibrium and @xmath68 is fulfilled .
the characteristic expansion time of the universe @xmath69 is given by @xmath70 where @xmath71 is the hubble parameter , which is related to the temperature @xmath72 by @xmath73 .
the condition @xmath74 means that neutrino oscillations have to be fast relative to the collision time .
the relation @xmath75 shows that for temperatures larger than 2 mev the collision time is always much smaller than the expansion time @xcite .
chemical equilibrium for the active neutrinos is maintained at temperatures larger than the neutrino decoupling temperature even in the presence of sterile neutrinos , provided the relevant mixing angles in the primordial plasma are not too large @xcite .
using the relation @xmath76 , the evolution equation ( [ master ] ) for @xmath63 can be rewritten as @xmath77 since by definition @xmath22 is the effective number of massless neutrino species at @xmath78 , in order to get a constraint on the mixing of sterile neutrinos we need to calculate the value of @xmath63 at @xmath78 produced by neutrino oscillations . with the initial condition @xmath79 ( @xmath80 mev ) , the integration of eq.([master1 ] )
gives @xcite @xmath81 imposing the upper bound @xmath82 one obtains the condition @xmath83 since @xmath84 is small in scheme a and @xmath85 is small in scheme b ( see eqs.([041 ] ) and ( [ 042 ] ) ) , they do not have any effect on neutrino oscillations during bbn .
hence in this section we use the approximation @xmath86 in scheme a and the approximation @xmath87 in scheme b @xcite . in this case , the elements @xmath88 and @xmath89 @xmath90 of the mixing matrix can be chosen real and can be parameterized by @xmath91 with the abbreviations @xmath92 and @xmath93 for @xmath94 and @xmath95 . in both schemes
a and b we have @xmath96 .
then , it can be shown that @xcite @xmath97 where @xmath98 and @xmath56 is replaced by its temperature average . with the definitions @xmath99 eq .
( [ p ] ) holds in both schemes .
the two terms in eq .
( [ p ] ) correspond to oscillations due to @xmath100 and @xmath3 , respectively .
oscillations due to @xmath4 are neglected in eq.([p ] ) because their contribution to @xmath101 ( see eq.([int ] ) ) is at least two orders of magnitude smaller for @xmath102 . using the expression ( [ p ] )
one can find an analytic form of the integral ( [ int ] ) and the condition ( [ cond ] ) gives the bound @xmath103 with the definitions ( [ defa ] ) this bound holds for both schemes , provided the conditions laid down in i.iv .
are fulfilled .
let us now discuss the implications of the bound ( [ bound ] ) . in order to understand
if this bound is compatible with a large @xmath104 , we first consider the possibility that @xmath104 is close to one .
in this case we have @xmath105 in scheme a and @xmath106 in scheme b. this means that , in order to accommodate the atmospheric neutrino anomaly , @xmath107 can not be small .
however , this is in contradiction with the inequality ( [ bound ] ) because of the second term in the left - hand side .
therefore , we conclude that the bound ( [ bound ] ) implies that @xmath104 is small , i.e. @xmath24 is small in scheme a and @xmath25 is small in scheme b. let us now check the validity of this reasoning considering the oscillations due to @xmath100 which generate the second term in eq.([bound ] ) and taking into account that the method considered here only holds for non - resonant neutrino oscillations or adiabatic transitions through a resonance .
looking at eq.([p ] ) we see that there is a resonant behaviour of the oscillations due to @xmath100 at the temperature where the relation @xmath108 is satisfied .
since @xmath109 is negative , this is only possible if @xmath110 is negative .
however , since we do not know the sign of @xmath110 , we have to check if the resonance is crossed adiabatically in the case of a negative sign .
a measure of the adiabaticity of the resonance is given by the parameter @xmath111 where @xmath112 and @xmath113 are the oscillation frequency and the effective mixing angle in matter . if @xmath114 the adiabatic crossing of the resonance is guaranteed .
clearly , for @xmath115 and @xmath116 this condition is fulfilled .
furthermore , oscillations due to @xmath100 occur around the temperature @xmath117 mev , where @xmath118 .
one can show that for @xmath115 the relation @xmath74 holds for resonant and non - resonant oscillations due to @xmath100 and thus the conditions for the validity of eq.([master ] ) and all relations derived from it are fulfilled as well .
the exclusion of the case @xmath115 with the help of the second term in eq.([bound ] ) is correct .
it is interesting to note that no conclusions can be drawn from the second term in eq.([bound ] ) if @xmath119 because then @xmath120 mev and in this temperature region one has @xmath121 and therefore no oscillations due to @xmath100 are possible because of strong quantum damping .
let us now examine the implications of the first term in eq.([bound ] ) .
the presence of this term implies that either @xmath24 or @xmath122 has to be very small . as we have just shown considering the second term in eq.([bound ] ) , in scheme a @xmath24 is small .
this implies that @xmath123 and only non - resonant oscillations are possible ( see the last term in eq.([p ] ) ) .
hence , eq.([bound ] ) is valid in scheme a and the first term gives the bound @xmath124 on the other hand , considering the second term in eq.([bound ] ) , we have shown that in scheme b the quantity @xmath25 is small .
therefore , in scheme b we have @xmath125 and a resonance occurs at the temperature @xmath126 ) are not fulfilled and the first term of eq.([bound ] ) does not apply . in this case the amount of sterile neutrinos produced at the resonance through non - adiabatic transitions can be calculated with the formula @xcite @xmath127 where @xmath128 ( @xmath129 ) is the effective mixing angle before ( after ) the resonance .
the last approximation in eq.([mswres ] ) follows from the fact that @xmath130 because of the high effective neutrino potential before the resonance and @xmath131 for small @xmath25 .
the quantity @xmath132 is the landau zener probability @xcite and @xmath133 is given by @xcite @xmath134 where @xmath135 is the half - width of the resonance defined in ref .
@xcite . demanding again
that @xmath136 one obtains the condition @xmath137 which gives the bound @xmath138 which is large for small @xmath25 .
hence the linear approximation should apply . in the following table
we show the values of the bounds @xmath139 and @xmath140 obtained from eqs.([bound1a ] ) and ( [ bound1b ] ) and the lsnd lower bound @xmath141 ( see eq.([lsndrange ] ) ) for different values of @xmath142 : @xmath143{0mm}{2 mm } & \ : 4.8 & 3.4 & 2.5 & 1.9 & 1.4 & 1.1 & 0.7 & 0.5 \\ \mbox{scheme b:}\quad & \mathcal{b}^{(\mathrm{b})}/10^{-5}\:\ : & \ : 4.9 & 3.4 & 2.6 & 1.9 & 1.5 & 1.1 & 0.8 & 0.5 \end{array}\ ] ] the bounds ( [ bound1a ] ) for the non - resonant oscillations in scheme a and ( [ bound1b ] ) for the resonant case in scheme b are in rough agreement with the corresponding numerical integrations of the evolution equations of an ensemble of oscillating neutrinos which take into account 2-neutrino oscillations @xcite .
let us now summarize our findings . the constraints on the parameter @xmath24 in scheme a ( eq.([bound1a ] ) ) and @xmath25 in scheme b ( eq.([bound1b ] ) ) has been derived in both schemes from the oscillations due to @xmath3 at temperatures of the order @xmath144 mev where @xmath145 .
the second term in eq.([bound ] ) , which is generated by the oscillations due to @xmath100 , only serves to exclude large @xmath104 .
the transitions of active into sterile neutrinos due to @xmath3 are non - resonant in scheme a , whereas they are resonant in scheme b , leading to a stronger constraint on @xmath25 than the constraint on @xmath24 in scheme a ( see table ( [ table ] ) ) . we have obtained stronger bounds than ref .
@xcite because we have taken into account the complete collision rates ( [ gamma ] ) presented in ref .
@xcite , which are nearly an order of magnitude larger than those used in ref . @xcite .
furthermore , our method shows that the dependence of the bound on @xmath24 in scheme a and @xmath25 in scheme b on the allowed number of neutrino species @xmath22 is logarithmic @xcite .
we want to emphasize that our calculation of the amount of sterile neutrinos brought into equilibrium by oscillations before the onset of bbn is based upon equation ( [ master ] ) in scheme a and upon the landau
zener approximation in scheme b. therefore , the numbers given in table ( [ table ] ) should be viewed as order of magnitude estimates of the upper bounds on the parameter @xmath24 in scheme a and @xmath25 in scheme b. concluding this section , we would like to make some comments on the non - standard bbn scenario with a non - zero lepton asymmetry presented in ref . @xcite .
there it has been shown that for a certain range of the mixing parameters and for resonant transitions due to @xmath3 a lepton asymmetry as small as @xmath146 can be amplified up to @xmath147 and in this way transitions at a lower temperature ( like the transitions due to @xmath100 in the schemes under consideration ) can be prevented .
such a scenario would not change our bound for @xmath25 in scheme b , but could have some effect on our considerations for scheme a. it @xmath24 is large in scheme a , a resonance due to @xmath3 occurs and , if a large lepton asymmetry is generated , the oscillations due to @xmath100 are suppressed and the second term on the left - hand side of eq.([bound ] ) is not valid .
hence a large parameter @xmath24 can not be excluded .
therefore , contrary to our discussion in this section , in this case we would have two possibilities for scheme a : either @xmath24 is small and has to obey the bound ( [ bound1a ] ) or @xmath24 is close to one and @xmath25 has to obey the same bound as given in eq.([bound1b ] ) for scheme b. however , judging from the results of ref .
@xcite it has yet to be shown that with @xmath24 close to one a realistic scenario meeting all experimental constraints can actually be achieved in scheme a and we leave this problem to future investigation .
as we have shown in the previous section , bbn suggests that the parameter @xmath24 in scheme a ( @xmath25 in scheme b ) is small . in this section
we present some possibilities to obtain information on the parameter @xmath24 from future results of sbl and lbl neutrino oscillation experiments .
we consider explicitly scheme a , but the same results are valid in scheme b after the replacement of @xmath24 with @xmath25 .
let us start with the consideration of @xmath27 oscillations , that are presently searched for in the chorus @xcite and nomad @xcite experiments and will be searched for in the cosmos @xcite experiment . from eq.([aab ] ) , for the amplitude of @xmath27 oscillations we have the upper bound @xmath148 the unitarity of the mixing matrix implies that @xmath149 and eq.([06 ] ) can be rewritten as @xmath150 \ , .
\label{080}\ ] ] since the quantity @xmath151 satisfies the inequality ( [ 041 ] ) , if @xmath152 we have @xmath153 on the other hand , if the inequality @xmath154 is satisfied , for the amplitude @xmath155 we have @xmath156 as it is seen from eqs.([08 ] ) and ( [ 11 ] ) , if it will be found that the amplitude @xmath155 satisfies the inequality @xmath157 it will mean that the parameter @xmath24 must satisfy the inequality @xmath158 and is in general not small . in fig.[fig1
] we have plotted the values of @xmath159 ( solid curve ) and @xmath160 ( dashed curve ) obtained from the 90% cl exclusion curves of the cdhs @xcite and ccfr @xcite experiments in the range ( [ widerange ] ) of @xmath3 , that covers the lsnd - allowed region .
the dashed curve representing @xmath160 constitutes an upper bound for @xmath155 due to the conservation of probability .
therefore , if the inequality ( [ 111 ] ) is satisfied , @xmath155 must lie between the solid and dashed curves . in fig.[fig1
] we have also plotted the most recent exclusion curve presented by the chorus collaboration @xcite ( dash - dotted curve ) , the expected final sensitivity of the chorus @xcite and nomad @xcite experiments ( dash - dot - dotted curve ) and the expected sensitivity of the cosmos @xcite experiment ( dotted curve ) .
one can see that a substantial part of the region in which the inequality ( [ 111 ] ) is satisfied is already ruled out by the present chorus exclusion curve , a large part will be excluded when the final sensitivities of the chorus and nomad experiments will be reached and almost all the region will be excluded if the cosmos experiment will not find @xmath27 oscillations . in fig.[fig2 ] we have plotted the values of @xmath161 obtained from the 90% cl exclusion curves of the bugey @xcite , cdhs @xcite and ccfr @xcite experiments .
one can see that for the values of @xmath3 in which @xmath162 , if the inequality ( [ 111 ] ) is satisfied @xmath24 must be large and incompatible with the bbn upper bound . on the other hand , for the values of @xmath3 in which @xmath163 , @xmath24 does not need to be large ( the inequality ( [ 09 ] ) is satisfied because @xmath84 can be larger than @xmath164 ) and there is no contradiction with the bbn upper bound on @xmath24 even if the inequality ( [ 111 ] ) is satisfied . if @xmath155 will be measured and found to satisfy the inequality ( [ 111 ] ) , also @xmath3 will be determined with some accuracy . in this case , depending on the allowed range of @xmath3 , it will be possible to decide if a small @xmath24 compatible with the bbn upper bound is excluded or not .
note , however , that for the lsnd range ( [ lsndrange ] ) of @xmath3 the parameter @xmath24 must be large if the inequality ( [ 111 ] ) is satisfied .
furthermore , if some future experiment will measure a value of @xmath155 such that eq.([111 ] ) is satisfied , then the inequality ( [ 08 ] ) will imply in addition the lower bound @xmath165 which is more stringent than the lower bound ( [ 112 ] ) if @xmath166 .
let us now consider the possibility that a future experiment will measure a small value for the amplitude @xmath155 , i.e. such that the inequality ( [ 11 ] ) is satisfied .
this measurement will not give any information on the value of @xmath24 .
in this case , the bbn bound derived in section [ sec3 ] implies that @xmath167 and for the amplitude @xmath155 we have the upper bound @xmath168 the numerical values of this upper bound in the range ( [ widerange ] ) of @xmath3 obtained from the 90% cl exclusion curve of the bugey @xcite experiment is shown in fig.[fig3 ] ( solid curve ) , together with the most recent exclusion curve presented by the chorus collaboration @xcite ( dash - dotted curve ) , the expected final sensitivity of the chorus @xcite and nomad @xcite experiments ( dash - dot - dotted curve ) and the expected sensitivity of the cosmos @xcite experiment ( dotted curve ) .
thus , we have seen that the investigation of @xmath27 oscillations in the region of @xmath3 that includes the lsnd - allowed region could allow to obtain information on the value of the parameter @xmath24 that is important for bbn .
let us consider now the possibility to obtain information on the parameter @xmath24 from the data of inclusive @xmath46 and @xmath45 experiments . from eqs.([baa ] ) and ( [ 041 ] ) it follows that @xmath169 furthermore , from eq.([07 ] ) we have @xmath170 from eqs.([181])([183 ] ) we obtain the inequality @xmath171 the amplitude @xmath172 is small in the whole range ( [ widerange ] ) of @xmath3 .
if the parameter @xmath24 is small as suggested by bbn , also the amplitude @xmath173 must be small .
the existing data do not exclude , however , large values of @xmath173 for @xmath174 .
hence , our analysis shows that the investigation of inclusive @xmath46 transitions for @xmath174 would be interesting for the check of the constraint that follows from bbn .
let us now consider the amplitude @xmath175 of @xmath26 oscillations . from the upper bound @xmath176 which follows from eqs.([c ] )
, ( [ aab ] ) and ( [ 041 ] ) , it is clear that the bbn bound on @xmath24 implies that @xmath175 is extremely small . in this case
the amplitudes measured in sbl @xmath5 disappearance experiments and sbl @xmath177 and @xmath178 appearance experiments are related by @xmath179 the experimental check of this relation allows to test the bbn bound on @xmath24 . indeed , if sbl @xmath5 disappearance experiments will find neutrino oscillations with a lower limit @xmath180 for the amplitude @xmath172 such that @xmath181 , where @xmath182 and @xmath183 are the experimental upper limit for @xmath37 and @xmath184 , respectively , it will mean that @xmath185 . in this case , taking into account that @xmath186 and that for small @xmath172 we have @xmath187 , the inequality @xmath188 gives the lower bound @xmath189 if , for example , it will be found that , which is given by the results of the bugey @xcite experiment , and with the present allowed range of @xmath37 , which is given by the results of the lsnd @xcite and bugey experiments ( @xmath190 and @xmath191 for @xmath3 in the lsnd range ( [ lsndrange ] ) ) .
notice that the allowed range of @xmath37 will be checked in the near future by karmen @xcite and other experiments @xcite .
furthermore , information on @xmath184 could be obtained in the future in experiments using a @xmath5 beam from a muon collider ( see refs .
@xcite ) . ]
@xmath192 , @xmath193 and @xmath194 , eq.([es2 ] ) will imply the lower bound @xmath195 , which is incompatible with the bbn upper bound .
cp violation in long - baseline neutrino oscillation experiments can also be used to check the bbn constraint on @xmath24 , though in practice this possibility is more remote .
it has been shown in ref .
@xcite that the absolute value of the cp - violating parameter @xmath196 $ ] ( in scheme a , whereas @xmath197 $ ] in scheme b ) could be as large as @xmath198 , which is the maximal value allowed by the unitarity of the mixing matrix . with the method discussed in ref .
@xcite one can show that @xmath199 for small @xmath200 . from the upper bound @xmath201 ( see eq.([041 ] ) ) and the bbn upper bound @xmath202 ( see eq.([bound1a ] ) ) we have @xmath203 consequently , taking into account that @xmath204 is small , the bbn upper bound on @xmath24 implies that cp violation in the @xmath27 channel is suppressed . on the other hand , if cp violation will be observed in lbl @xmath27 experiments , using eq.([cp1 ] ) it will be possible to set a lower bound for @xmath24 .
analogous results are valid in scheme b with the replacements @xmath205 , @xmath206 and @xmath207 .
in this section we propose a parameterization of the @xmath14 mixing matrix @xmath1 which is suitable for the two schemes a and b that allow to accommodate all existing neutrino oscillation data . in these schemes the neutrinos @xmath208 , @xmath209 and the neutrinos @xmath210 , @xmath211 give separate contributions to all observables .
thus , it is natural to consider separately the elements @xmath212 with @xmath213 and the elements @xmath214 with @xmath215 . in the following
we will stick to scheme a for the actual presentation .
the formulas in scheme b can be obtained with the exchanges @xmath216 and @xmath217 of the columns of @xmath1 .
we take into account that all observable transition probabilities are invariant under the phase transformation @xmath218 where @xmath219 and @xmath220 are arbitrary parameters .
therefore , @xmath221 with @xmath215 and @xmath222 with @xmath213 can always be considered as real vectors .
let us write them as @xmath223 with @xmath224 the unit vectors @xmath225 and @xmath226 can be written in the form @xmath227 let us introduce also the orthogonal unit vectors @xmath228 then the vectors @xmath212 with @xmath229 and @xmath230 can be expanded over the orthonormal basis @xmath231 , @xmath232 and the vectors @xmath212 with @xmath233 and @xmath213 can be expanded over the orthonormal basis @xmath234 , @xmath235 . as a result , we have the parameterization @xmath236 for the mixing matrix . using the invariance under the transformation ( [ 401 ] )
, we can choose the parameters @xmath237 and @xmath238 as real . from eqs.([c ] ) and ( [ 404 ] ) it is obvious that the parameters @xmath239 and @xmath240 are connected , respectively , with the parameters @xmath84 and @xmath241 by the relations @xmath242 furthermore , from the unitarity of @xmath1 it follows that the parameters @xmath243 and @xmath244 are related by @xmath245 similarly , we can think of @xmath237 and @xmath238 being fixed by the unit length of the lines of @xmath1 and five of the remaining complex parameters by the orthogonality of different lines . counting the number of remaining real parameters in @xmath1 we obtain 10 versus 9 physical real parameters in a 4@xmath2464 unitary mixing matrix .
a careful inspection of the residual phase freedom reveals that one of the complex parameters @xmath244 , @xmath247 , @xmath243 , @xmath248 , @xmath249 , @xmath250 , @xmath251 , @xmath252 , @xmath253 or @xmath254 can be chosen real in addition .
the significance of the parameters in the matrix ( [ uv ] ) shows up by considering the amplitudes of neutrino oscillations in disappearance and appearance sbl experiments , which are directly connected with the parameters @xmath255 and @xmath256 .
in fact , we have @xmath257 and @xmath258 the sbl oscillation amplitudes do not contain the angles @xmath259 and @xmath260 .
this is connected with the fact that sbl amplitudes are determined by products of vectors in the @xmath34 space or in the @xmath35 space ( see eqs.([aab ] ) and ( [ baa ] ) ) .
however , information on the angle @xmath260 can be obtained from the results of atmospheric and lbl neutrino experiments , whereas the results of solar neutrino experiments give information on the angle @xmath259 . in the four - neutrino scheme
a the parameters @xmath244 , @xmath247 are small because @xmath261 is small ( see eq.([041 ] ) ) .
let us now investigate if it is possible to obtain bounds on the parameters @xmath262 and @xmath263 from measurable quantities .
we will use the unitarity of the mixing matrix and we will work in the approximation @xmath86 . in this case
, the parameterization ( [ uv ] ) of the mixing matrix reduces to @xmath264 and the amplitude of @xmath27 oscillations is given by ( see eq.([aab ] ) ) @xmath265 furthermore , taking into account that @xmath266 we have @xmath267 from eqs.([060 ] ) and ( [ 063 ] ) and taking into account that @xmath268 ( see eq.([041 ] ) ) , we obtain @xmath269 this lower bound for @xmath270 , obtained in the approximation @xmath271 , is consistent with the lower bound ( [ 081 ] ) on @xmath24 . from the comparison of these two lower bounds it is clear that if @xmath24 is large , its value approximately coincides with @xmath263 , whereas @xmath262 is small if @xmath272 is small .
indeed , it is possible to obtain an upper bound for the parameter @xmath262 using the unitarity relation @xmath273 taking into account that @xmath274 we find from eq.([065 ] ) @xmath275 since @xmath276 ( see eq.([041 ] ) ) , we obtain the upper bound @xmath277 therefore , if the experimental upper bound for the amplitude of sbl @xmath46 transitions is small , @xmath262 is small as well .
note that the observation of sbl inclusive @xmath46 transitions together with the observation of sbl @xmath27 and @xmath278 ( lsnd @xcite ) transitions would allow to get information on the transitions of muon neutrinos into sterile states .
in fact , because of the unitarity of the mixing matrix , for the amplitude @xmath279 of @xmath280 transitions we have @xmath281 hence , we think that it is important to continue the search for neutrino oscillations in the inclusive @xmath46 channel , especially in the region of small @xmath282 , below @xmath283 , which has not been explored so far .
let us consider the possibility that the right side of eq.([070 ] ) will be measured to be different from zero , i.e. , that @xmath280 transitions occur in sbl experiments . in the scheme under consideration the amplitude of these transitions
is given by @xmath284 taking into account that @xmath285 , eqs.([182 ] ) , ( [ 070 ] ) and ( [ 071 ] ) imply that @xmath286 finally , if it will be found that not only @xmath287 sbl transitions but also @xmath46 sbl transitions are strongly suppressed , it will mean that the only oscillation channel involving sterile neutrinos which may be not suppressed is the channel @xmath288 .
therefore , in this case it will be possible to obtain information on transitions of active neutrinos into sterile states only if @xmath7 beams ( see ref .
@xcite ) will be available . in this case @xmath289 and @xmath290 let us now consider the possibility that not only @xmath244 , @xmath247 but also @xmath252 , @xmath253 are small , as follows from the smallness of @xmath291 implied by bbn if @xmath21 , as shown in section [ sec3 ] . in this case it is possible to expand all the quantities @xmath255 , @xmath292 , @xmath256 and @xmath293 in powers of the small parameters @xmath244 , @xmath247 , @xmath252 , @xmath253 , up to order @xmath84 , @xmath24 , @xmath294 .
this means that in this approximation it is possible to express all elements of @xmath1 in terms of the parameters @xmath244 , @xmath247 , @xmath252 , @xmath253 , @xmath259 and @xmath260 .
note that one of the phases of the small complex parameters is unphysical and can be transformed away reaching thus the number of 9 physically independent real parameters in @xmath1 . performing this program
we obtain @xmath295 in this approximation , eqs.([bb ] ) and ( [ aa ] ) become @xmath296 and @xmath297 therefore we have @xmath298 the absolute value of @xmath252 is given by latexmath:[\[\label{ps } absence of @xmath7 disappearance experiments ( see eq.([073 ] ) ) , the determination of the parameter @xmath300 needs further information from lbl experiments .
this is not an artifact of the expansion in eq.([elements ] ) .
the reason is that @xmath173 and @xmath155 are small if @xmath301 is small and this is true even if @xmath300 is large , i.e. @xmath24 is large .
this explains the limitations in our search for tests of the bbn constraint on @xmath24 derived in section [ sec3 ] . on the other hand , in the present approximation , for the determination of the values of @xmath302 , @xmath303 and @xmath301 in principle the results of sbl experiments alone are sufficient .
the free parameter @xmath304 is fixed by a measurement of the sbl @xmath305 oscillation amplitude . with the result of the lsnd experiment and the negative result of all other sbl neutrino oscillation experiments @xmath304
is approximately bounded by @xmath306 .
if the bbn constraint on @xmath24 is valid , then @xmath27 transitions are significantly suppressed in sbl neutrino oscillation experiments with an oscillation amplitude of order @xmath307 or smaller .
furthermore , eq.([baa ] ) shows that the oscillation amplitude @xmath184 is at most close to 0.1 . in the discussion in section [ sec3 ]
we have seen that the bbn bound @xmath21 implies that @xmath308 . in this case
we have @xmath309 inspecting eq.([elements ] ) we note that in scheme a the quantity @xmath151 is now of the same order of magnitude as @xmath84 . as mentioned in the previous section , this is particularly interesting for @xmath310 ev@xmath311 where the bound on @xmath173 from the @xmath6 disappearance experiments is not very stringent and , consequently , in this @xmath3 region @xmath173 should be much smaller than the present upper experimental upper bound .
in addition , also @xmath312 is of the same order of magnitude , showing that at zeroth - order in the expansion over the small quantities @xmath84 and @xmath24 , the mixing matrix is given by @xmath313 therefore , we have @xmath314 hence , the @xmath315 @xmath316 and @xmath317 @xmath318 sectors are decoupled and the oscillations of solar and atmospheric neutrinos are independent .
furthermore , since only the small mixing angle msw solutions of the solar neutrino problem seems to be allowed by the data in the case of @xmath26 transitions ( see refs .
@xcite ) , at zeroth - order in the expansion over the small quantities @xmath84 , @xmath24 and @xmath259 we have @xmath319 similar conclusions are valid in scheme b with @xmath320 replaced by @xmath321 and @xmath322 .
let us now discuss lbl neutrino oscillations . keeping only the dominant terms we get @xmath323 where @xmath324 in scheme a and @xmath325 in scheme b. this shows that there must be significant lbl @xmath27 oscillations because of the sizeable @xmath6 disappearance known from the atmospheric neutrino anomaly , which implies that @xmath326 is large .
for @xmath278 and @xmath178 transitions matter effects have to be taken into account . the upper bounds on these transition probabilities derived in ref .
@xcite are of order @xmath147 . if @xmath24 is small , as follows from the bbn constraint , the `` matter - stable '' bounds on @xmath327 and @xmath328 derived in ref .
@xcite are improved for small values of @xmath3 because of the inequalities @xmath329 leading to @xmath330 hence , the bbn constraint on @xmath24 imply that @xmath331 obviously , the observation of a violation of these inequalities in lbl neutrino oscillation experiments would imply that the bbn constraint on @xmath24 is not valid .
in this paper we have focused our discussion on the two four - neutrino schemes a and b ( see eq.([ab ] ) ) which are compatible with the results of all neutrino oscillation experiments @xcite . as shown in ref .
@xcite , the bound @xmath21 for the effective number @xmath22 of light neutrinos before the onset of bbn implies a stringent upper bound on @xmath332 in scheme a and on @xmath25 in scheme b. our limits are more stringent than those of ref .
@xcite because we used the complete collision rates presented in ref . @xcite .
thus , in scheme a we have obtained @xmath333 .
we have demonstrated that in scheme b the limit derives from resonant transitions due to @xmath3 , leading to an upper bound for @xmath25 that is more stringent than the one for @xmath24 in scheme a : @xmath334 .
the validity of these limits from bbn depends to a certain extent on the results of the measurements of the primordial deuterium abundance which are controversial at the moment @xcite .
we have therefore emphasized the importance of terrestrial experiments to get information on the value of @xmath24 in order to check the bbn constraints . in section [ sec4 ]
we have worked out lower bounds on @xmath24 in scheme a ( @xmath25 in scheme b ) involving the following observables in sbl experiments : 1 .
the @xmath335 oscillation amplitude @xmath155 ; 2 .
the oscillation amplitude @xmath173 in @xmath6 disappearance experiments for @xmath310 ev@xmath311 ; 3 .
the oscillation amplitudes @xmath172 in reactor disappearance experiments and @xmath37 and @xmath184 in accelerator experiments .
to proceed further , in section [ sec5 ] we have introduced a parameterization of the 4 @xmath246 4 neutrino mixing matrix @xmath1 ( [ uv ] ) which is particularly suited for the schemes a and b. in this parameterization the quantity @xmath24 is given by @xmath336 in scheme a ( @xmath337 in scheme b ) , where @xmath252 and @xmath253 are in general complex parameters in @xmath1 .
the parameterization ( [ uv ] ) has lead us to the following observations : 1 .
the parameter @xmath301 can be determined in sbl @xmath338 oscillation experiments .
2 . for the determination of @xmath300 in sbl oscillation experiments it is necessary to employ a @xmath7 neutrino beam and perform a disappearance experiment .
3 . for @xmath21
the sbl oscillation amplitude @xmath155 is at most of order @xmath307 whereas the lbl oscillation amplitude must be of order one in the same channel . as already pointed out in ref .
@xcite , the bbn constraint @xmath21 leads to a 4-neutrino mixing picture in which , apart from small corrections of the order of the small quantities @xmath84 and @xmath24 , the mixing matrix @xmath1 is composed by the two decoupled sectors @xmath315 @xmath316 and @xmath317 @xmath318 in scheme a ( @xmath315 @xmath318 and @xmath317 @xmath316 in scheme b ) .
consequently , the oscillations of solar and atmospheric neutrinos are independent .
the solar neutrino problem is solved with @xmath26 oscillations governed by the mixing angle @xmath259 and the atmospheric neutrino anomaly is explained with @xmath27 oscillations governed by the mixing angle @xmath260 ( see eq.([uv ] ) for the definition of these angles ) .
concerning the solar neutrino deficit , it was shown that a vacuum oscillation solution involving sterile neutrinos is rather disfavored by the data ( see refs .
therefore , the specific prediction of a small mixing angle msw solution of the solar neutrino problem with only @xmath26 transitions serves as an indirect check of the schemes a and b under consideration and of the bbn bound .
this prediction implies a characteristic strong distortion of the @xmath339b @xmath5 spectrum and a day - night asymmetry in the neutral current event rate at sno @xcite .
experimentally , this prediction can be checked independently from solar model calculations ( see ref .
@xcite ) . if indeed the bbn constraints on the 4-neutrino mixing matrix @xmath1 are confirmed then the biggest gap in our knowledge of @xmath1 , the sterile neutrino mixing , is considerably narrowed . in this paper
we have shown that it is possible to test the ensuing mixing matrix to some extent in terrestrial experiments .
the most striking feature in this context is the suppression of all sbl oscillation amplitudes .
k. inoue , talk presented at the @xmath340 international workshop on _ topics in astroparticle and underground physics _ , gran sasso , italy , september 1997 ( http://www-sk.icrr.u-tokyo.ac.jp/doc/sk/pub/pub_sk.html ) ; r. svoboda , talk presented at the conference on _ solar neutrinos : news about snus _ , 26 december 1997 , santa barbara , california ( http://www.itp.ucsb.edu/online/snu/ ) .
mikheyev and a.yu .
smirnov , yad .
42 * , 1441 ( 1985 ) [ sov .
. phys . * 42 * , 913 ( 1985 ) ] ; il nuovo cimento c * 9 * , 17 ( 1986 ) ; l. wolfenstein , phys .
d * 17 * , 2369 ( 1978 ) ; _ ibid .
_ * 20 * , 2634 ( 1979 ) .
peltoniemi and j.w.f .
valle , nucl .
b * 406 * , 409 ( 1993 ) ; d.o .
caldwell and r.n .
mohapatra , phys .
d * 48 * , 3259 ( 1993 ) ; z. berezhiani and r.n .
mohapatra , phys .
d * 52 * , 6607 ( 1995 ) ; j.r .
et al . _ , phys .
lett . * 74 * , 2160 ( 1995 ) ; e. ma and p. roy , phys .
d * 52 * , r4780 ( 1995 ) ; r. foot and r.r .
volkas , phys .
d * 52 * , 6595 ( 1995 ) ; e.j .
b * 357 * , 608 ( 1995 ) ; j.j .
gomez - cadenas and m.c .
gonzalez - garcia , z. phys .
c * 71 * , 443 ( 1996 ) ; s. goswami , phys . rev .
d * 55 * , 2931 ( 1997 ) ; a.yu .
smirnov and m. tanimoto , phys .
d * 55 * , 1665 ( 1997 ) ; e. ma , mod .
a * 11 * , 1893 ( 1996 ) .
v. barger , t. weiler and k. whisnant , preprint hep - ph/9712495 ; s.c .
gibbons , r.n .
mohapatra , s. nandi and a. raychaudhuri , preprint hep - ph/9803299 .
copi , d.n . schramm and m.s .
turner , phys .
55 , 3389 ( 1997 ) ; n. hata _
et al . _ ,
phys . rev .
d * 55 * , 540 ( 1997 ) ; c.j .
copi , d.n .
schramm and m.s .
turner , phys .
d * 55 * , 3389 ( 1997 ) ; d.n .
schramm and m.s .
turner , preprint fermilab - pub-97/186-a ( astro - ph/9706069 ) ; keith a. olive , talk given at taup97 , gran sasso lab .
, september 1997 , preprint astro - ph/9712160 .
e.w . kolb and m.s .
turner , _ the early universe _ ( addison - wesley 1990 ) .
et al . _ , mon . not .
268 * , l1 ( 1994 ) and * 278 * , 506 ( 1996 ) ; e.j . wampler
et al . _ ,
. astrophys . *
316 * , 33 ( 1996 ) ; a. songaila _
et al . _ ,
nature * 368 * , 599 ( 1994 ) ; m. rugers and c.j .
hogan , astrophys .
* 459 * , 1 ( 1996 ) and astron .
j. * 111 * , 2135 ( 1996 ) ; j.k . webb _
_ , nature * 388 * , 250 ( 1997 ) .
s. burles and d. tytler , preprint astro - ph/9803071 , to appear in the proceedings of the second oak ridge symposium on atomic & nuclear astrophysics , ( oak ridge , tn , december 2 - 6 , 1997 ) , ed .
a. mezzacappa ( institute of physics , bristol ) .
krastev and s.t .
petcov , phys .
72 * , 1960 ( 1994 ) ; phys .
d * 53 * , 1665 ( 1996 ) .
k. kainulainen , phys .
b * 224 * , 191 ( 1990 ) .
r. barbieri and a. dolgov , phys .
b * 237 * , 440 ( 1990 ) .
k. enqvist , k. kainulainen and j. maalampi , nucl .
b * 349 * , 754 ( 1991 ) .
k. enqvist , k. kainulainen and m. thomson , nucl .
b * 373 * , 498 ( 1992 ) .
d. ntzold and g. raffelt , nucl .
b * 307 * , 924 ( 1988 ) .
cline , phys .
* 68 * , 3137 ( 1992 ) .
k. enqvist , k. kainulainen and j. maalampi , phys .
lett b * 249 * , 531 ( 1990 ) .
x. shi , d. n. schramm and b. d. fields , phys .
d * 48 * , 2563 ( 1993 ) .
l. d. landau , phys .
z. ussr * 1 * , 426 ( 1932 ) ; c. zener , proc .
a * 137 * , 696 ( 1932 ) .
r. foot , m.j .
thomson and r.r .
volkas , phys .
d * 53 * , 5349 ( 1996 ) ; x. shi , phys .
d * 54 * , 2753 ( 1996 ) ; r. foot and r.r .
volkas , phys .
d * 55 * , 5147 ( 1997 ) .
[ fig1 ] fig.[fig1 ] .
plot in the @xmath155@xmath3 plane of the curve @xmath159 ( solid line ) . for @xmath155 to the right of this curve ( see eq.([111 ] ) ) the lower bound ( [ 112 ] ) for @xmath24 applies .
the dashed curve representing @xmath160 constitutes an upper bound for @xmath155 due to the conservation of probability .
the values of @xmath341 and @xmath160 have been obtained from the 90% cl exclusion curves of the cdhs @xcite and ccfr @xcite experiments .
the dash - dotted curve , dash - dot - dotted and dotted curves represent , respectively , the most recent exclusion curve presented by the chorus collaboration @xcite , the expected final sensitivity of the chorus @xcite and nomad @xcite experiments and the expected sensitivity of the cosmos @xcite experiment .
[ fig2 ] fig.[fig2 ] .
plot in the @xmath24@xmath3 plane of the lower limit @xmath161 for @xmath24 that follows from the inequality ( [ 111 ] ) .
the values of @xmath204 and @xmath341 have been obtained from the 90% cl exclusion curves of the bugey @xcite , cdhs @xcite and ccfr @xcite experiments .
[ fig3 ] fig.[fig3 ] .
plot in the @xmath155@xmath3 plane of the upper bound @xmath342 ( solid curve ) for @xmath155 ( see eq.([113 ] ) ) in the case of a very small @xmath24 .
the values of @xmath204 have been obtained from the 90% cl exclusion curve of the bugey @xcite experiment .
the dash - dotted curve , dash - dot - dotted and dotted curves are the same as in fig.[fig1 ] . | we investigate the big - bang nucleosynthesis constraints on neutrino mixing in the framework of the two four - neutrino schemes that are favored by the results of neutrino oscillation experiments .
we discuss the implications of these constraints for terrestrial short and long - baseline neutrino oscillation experiments and we present some possibilities of testing them in these experiments . in particular , we show that from the big - bang nucleosynthesis constraints it follows that the @xmath0 transition is severely suppressed in short - baseline experiments , whereas its oscillation amplitude in long - baseline experiments is of order 1 .
we also propose a new parameterization of the four - neutrino mixing matrix @xmath1 which is appropriate for the schemes under consideration . |
let @xmath3 be a prime number .
let @xmath5 be a number field and let @xmath6 be a uniform @xmath3-extension : @xmath6 is a normal extension whose galois group @xmath7 is a uniform pro-@xmath3 group ( see section [ sectionuniform ] ) .
we suppose moreover that _ the set of places of @xmath5 that are ramified in @xmath6 is finite_. if @xmath8 is a finite subextension of @xmath6 , let us denote by @xmath9 the @xmath3-sylow subgroup of the class group of @xmath10 and put @xmath11 where the limit is taken over all number fields @xmath10 in @xmath6 with respect the norm map .
then @xmath12 is a @xmath13$]-module and , thanks to a structure theorem ( see section [ sectionarithmetic ] ) , one attaches a @xmath2-invariant to @xmath12 , generalizing the well - known @xmath2-invariant introduced by iwasawa in the classical case @xmath14 .
iwasawa showed that the size of the @xmath2-invariant is related to the rate of growth of @xmath3-ranks of @xmath3-class groups in the tower .
for the simplest @xmath15-extensions , i.e. the cyclotomic ones , he conjectured that @xmath16 ; this was verified for base fields which are abelian over @xmath17 by ferrero and washington @xcite but remains an outstanding problem for more general base fields .
iwasawa initially suspected that his @xmath2-invariant vanishes for all @xmath15-extensions , but later was the first to construct @xmath15-extensions with non - zero ( indeed arbitrarily large ) @xmath2-invariants .
it is natural to ask what other @xmath3-adic groups enjoy this property .
our present work leads us to the following conjecture : [ conjecture ] let @xmath1 be a uniform pro-@xmath3 group having a non - trivial fixed - point - free automorphism @xmath18 of order @xmath19 co - prime to @xmath3 ( in particular if @xmath20 is prime , @xmath1 is nilpotent ) .
then @xmath1 has arithmetic realizations with arbitrarily large @xmath2-invariant , i.e. for all @xmath21 , there exists a number field @xmath5 and an extension @xmath6 with galois group isomorphic to @xmath1 such that @xmath22 . our approach for realizing @xmath1 as a galois group is to make use of the existence of so - called @xmath3-rational fields .
see below for the definition , but for now let us just say that the critical property of @xmath3-rational fields is that in terms of certain maximal @xmath3-extensions with restricted ramification , they behave especially well , almost as well as the base field of rational numbers .
as we will show , conjecture [ conjecture ] can be reduced to finding a @xmath3-rational field with a fixed - point - free automorphism of order @xmath19 co - prime to @xmath3 .
these considerations lead us to formulate the following conjecture about @xmath3-rational fields .
[ conj2 ] given a prime @xmath3 and an integer @xmath23 co - prime to @xmath3 , there exist a totally imaginary field @xmath24 and a degree @xmath19 cyclic extension @xmath25 such that @xmath5 is @xmath3-rational .
although we will not need it , we believe @xmath24 in the conjecture may be taken to be imaginary quadratic ; see conjecture [ conjecture - cl ] below .
our key result is : conjecture [ conj2 ] for the pair @xmath26 implies conjecture [ conjecture ] for any uniform pro-@xmath3 group @xmath1 having a fixed - point - free automorphism of order @xmath19 .
one knows that if @xmath19 is an odd divisor of @xmath4 , where @xmath3 is a regular prime , then for any @xmath27 , the cyclotomic field @xmath28 provides a positive answer to the previous question .
we therefore have [ maintheorem ] assume @xmath3 is a regular prime and that the uniform group @xmath1 has a fixed - point - free automorphism @xmath18 of odd order @xmath19 dividing @xmath4 .
then conjecture [ conjecture ] is true for @xmath1 .
this circle of ideas is closely related to the recent work of greenberg @xcite in which he constructs analytic extensions of number fields having a galois group isomorphic to an open subgroup of @xmath29 .
the idea of studying pro-@xmath3 towers equipped with a fixed - point - free action of a finite group of order prime to @xmath3 occurs also in boston s papers @xcite and @xcite .
our work raises the following purely group - theoretical question .
[ gtquestion ] let @xmath1 be a nilpotent uniform pro-@xmath3 group .
does there exist a uniform nilpotent pro-@xmath3 group @xmath30 having a fixed - point - free automorphism of prime order @xmath31 such that @xmath32 ?
a positive answer to this question would imply that for all nilpotent uniform pro-@xmath3 groups @xmath1 , there exist arithmetic realizations with arbitrarily large @xmath2-invariant . in his recent work @xcite
, gras gave some conjectures about the @xmath3-adic regulator in a fixed number field @xmath5 when @xmath3 varies .
in our context , one obtains : let @xmath33 be an infinite set of prime numbers and let @xmath19 be an integer co - prime to all @xmath34 .
let @xmath35 be a family of uniform pro-@xmath3 groups of fixed dimension @xmath36 , all having a fixed - point - free automorphism @xmath18 of order @xmath19 .
assuming the conjecture of gras ( see conjecture [ conjectureregulator ] ) , there exists a constant @xmath37 , such that for all @xmath38 , there exist @xmath39-extensions of number fields with arbitrarily large @xmath2-invariants . in another direction , a conjecture in the spirit of the heuristics of cohen - lenstra concerning the @xmath3-rationality of the families @xmath40 of number fields @xmath41 galois over @xmath17 ,
all having galois group isomorphic to a single finite group @xmath42 , seems to be reasonable ( see @xcite ) . when the prime @xmath43 , the philosophy here is that the density of @xmath3-rational number fields in @xmath40 is positive .
this type of heuristic lends further evidence for conjecture [ conjecture ] .
let @xmath44 be a finitely generated pro-@xmath3 group . for two elements @xmath45 of @xmath44 ,
we denote by @xmath46 the conjugate of @xmath47 by @xmath48 and by @xmath49:=x^{-1}y^{-1}xy = x^{-1}x^y$ ] the commutator of @xmath47 and @xmath48 . for closed subgroups @xmath50 of @xmath44 ,
let @xmath51 $ ] be the _ closed _ subgroup generated by all commutators @xmath52 $ ] with @xmath53 .
let @xmath54 $ ] be the maximal abelian quotient of @xmath44 , and let @xmath55 be its @xmath3-rank
. denote by @xmath56 the @xmath3-central descending series of @xmath44 : @xmath57 , \cdots , \ \g_{n+1}=\g_n^p[\g,\g_n]\cdot\ ] ] the sequence @xmath58 forms a base of open neighborhoods of the unit element @xmath59 of @xmath44 .
if @xmath5 is a number field , let @xmath60 be the @xmath3-sylow subgroup of the class group of @xmath5 .
let @xmath61 be the set of primes of @xmath5 of residue characteristic @xmath3 .
if @xmath62 is any finite set of places of @xmath5 , denote by @xmath63 the maximal pro-@xmath3 extension of @xmath5 unramified outside @xmath62 and put @xmath64 as well as @xmath65 .
two standard references concerning @xmath3-adic analytic and in particular , uniform , pro-@xmath3 groups are the long article of lazard @xcite and the book of dixon , du sautoy , mann and segal @xcite .
let @xmath1 be an analytic pro-@xmath3 group : we can think of @xmath1 as a closed subgroup of @xmath66 for a certain integer @xmath19 . the group @xmath1 is said _ powerful _
if @xmath67\subset \gamma^p $ ] ( @xmath67 \subset \gamma^4 $ ] when @xmath68 ) ; a powerful pro-@xmath3 group @xmath1 is said _ uniform _ if it has no torsion .
let us recall two important facts .
every @xmath3-adic analytic pro-@xmath3 group contains an open uniform subgroup .
[ isomorphismxp ] a powerful pro-@xmath3 group @xmath69 is uniform if and only if for @xmath70 , the map @xmath71 induces an isomorphism between @xmath72 and @xmath73 . let us make some remarks .
let @xmath74 be the dimension of @xmath1 as analytic variety .
\1 ) if @xmath1 is uniform then @xmath75 , where @xmath76 is the @xmath3-cohomological dimension of @xmath1 .
\2 ) ( * ? ? ?
* corollary 1.8 ) suppose @xmath1 is a torsion - free @xmath3-adic analytic group , @xmath77 and @xmath78
. then the group @xmath1 is uniform .
\3 ) for @xmath79 , the pro-@xmath3-group @xmath80 is uniform , where @xmath81 is the the set of @xmath82 matrices with coefficients in @xmath15 .
now let us fix a uniform pro-@xmath3 group @xmath1 of dimension @xmath36 .
recall that @xmath83 is the @xmath3-descending central series of @xmath1 . by theorem [ isomorphismxp ]
, one has @xmath84=p^{dn}$ ] , for all @xmath85 .
let @xmath86:=\lim_{\stackrel{\longleftarrow}{\u \lhd_o \gamma } } \z_p[\gamma/\u]}$ ] be the complete iwasawa algebra , where @xmath87 runs along the open normal subgroups of @xmath1 .
put @xmath88 = \z_p[[\gamma]]/p$ ] .
the rings @xmath89 and @xmath13 $ ] are local , noetherien and without zero divisor @xcite : each of them has a fractional skew field .
denote by @xmath90 the fractional skew field of @xmath89 . if @xmath12 is a finitely generated @xmath89-module , the _
@xmath91 of @xmath12 is the @xmath90-dimension of @xmath92 . for more details ,
we refer the reader to howson @xcite and venjakob @xcite .
let @xmath12 be a finitely generated @xmath13$]-module .
put @xmath93 \ , \ , \ , { \rm and \ } \ , \mu(\x)= \sum_{i\geq 0 } \rg_\omega \x[p^{i+1}]/\x[p^{i}].\ ] ] one has @xmath94 and , @xmath95 if and only if @xmath96 .
there is a large and growing literature on the study of @xmath13$]-modules in the context of iwasawa theory .
we recall a result of perbet @xcite , where , by making use of the work of harris @xcite , venjakob @xcite and coates - schneider - sujatha @xcite , among others , he manages to estimate the size of the coinvariants @xmath97 of @xmath12 . recall that @xmath98 is the largest quotient of @xmath12 on which @xmath99 , the @xmath85th element of the @xmath3-central series , acts trivially .
suppose that @xmath12 is a torsion @xmath13$]-module where @xmath1 is a uniform pro-@xmath3 group of dimension @xmath36 .
then for @xmath100 : @xmath101 and @xmath102 we now turn to applying these formulas in the arithmetic context .
let @xmath6 be a galois extension of number fields with galois group @xmath1 , where @xmath1 is a uniform pro-@xmath3 group .
we assume that _ the set of primes of @xmath5 ramified in @xmath6 is finite_. let @xmath103 be the @xmath3-central descending series of @xmath1 and put @xmath104 for the corresponding tower of fixed fields .
now let @xmath12 be the projective limit along @xmath6 of the @xmath3-class group @xmath105 of the fields @xmath106 .
then , @xmath12 is a finitely generated torsion @xmath13$]-module .
for the remainder of this work , @xmath12 will denote this module built up from the @xmath3-class groups of the intermediate number fields in @xmath6 .
in particular , we put @xmath107 and @xmath108 . for this module , by classical descent ,
perbet proved : for @xmath100 , one has : @xmath109 and @xmath110 one then obtains immediately the following corollary : [ critereminorationmu ] @xmath111 along the extension @xmath6 , the @xmath3-rank of @xmath105 grows linearly with respect to the degree @xmath112 $ ] if and only if @xmath113 .
@xmath114 if there exists a constant @xmath115 such that for all @xmath116 , @xmath117 , i.e. @xmath118 $ ] , then @xmath119 . at this point
, we should recall some standard facts from commutative iawasawa theory .
first , for the cyclotomic @xmath15-extension , iwasawa conjectured that the @xmath2-invariant is 0 for all base fields , and this has been shown to be the case when the based field is abelian over @xmath17 . when the base field @xmath5 contains a primitive @xmath3th root of unity , the reflection principle allows one to give some estimates on the @xmath2-invariant @xcite .
a number field @xmath5 is called _ @xmath3-rational _ if the galois group @xmath120 of the maximal pro-@xmath3-extension of @xmath5 unramified outside @xmath3 is pro-@xmath3 free . from the extensive literature on @xmath3-rational fields , we refer in particular to jaulent - nguyen quang do @xcite , movahhedi - nguyen quang do @xcite , movahhedi @xcite , and gras - jaulent@xcite .
a good general reference is the book of gras @xcite .
if @xmath44 is a pro-@xmath3 group , let us denote by @xmath121 the torsion of @xmath122 . when @xmath123 , put @xmath124 ; when @xmath125 , put @xmath126 .
a standard argument in pro-@xmath3 group theory shows the following : a pro-@xmath3 group @xmath44 is free if and only if @xmath121 and @xmath127 are trivial .
indeed , the exact sequence @xmath128 gives the sequence @xmath129\longrightarrow 0,\ ] ] and to conclude , recall that @xmath130 and @xmath131 .
remark that if @xmath44 is pro-@xmath3 free then @xmath132 , where @xmath36 is the @xmath3-rank of @xmath44 .
this observation shows in particular that if the group @xmath44 corresponds to the galois group of a pro-@xmath3 extension of number fields , then necessarily this extension is wildly ramified .
we now explain how the schur multiplier of @xmath133 is related to the leopoldt conjecture .
let @xmath62 be a set of places of @xmath5 , and for @xmath134 , let @xmath135 be the pro-@xmath3 completion of the local units at @xmath136 .
if the @xmath15-rank of the diagonal image of @xmath137 in @xmath138 is maximal , namely @xmath139 , then @xmath140 . moreover if @xmath141 , these conditions are equivalent .
in particular , assuming the leopoldt conjecture for @xmath5 at @xmath3 , @xmath120 is free if and only if @xmath142 is trivial .
for the case of restricted ramification , see @xcite .
for the general case , see ( * ? ? ?
* corollaire 1.5 ) .
some examples of free @xmath133 are given in
[ splitting - section ] below .
let @xmath143 be the @xmath3-sylow subgroup of the ray class group of modulus @xmath144 of @xmath5 , where @xmath145 , and @xmath146 is the absolute index of ramification of @xmath147 ( in @xmath148 ) .
assume that @xmath5 verifies the leopoldt conjecture at @xmath3 .
then @xmath5 is @xmath3-rational if and only if @xmath149 .
first , as one assumes leopoldt conjecture for @xmath5 at @xmath3 , then the @xmath15-rank of @xmath150 is exactly @xmath151 .
hence @xmath120 is free if and only if @xmath152 , which is equivalent to @xmath153 .
now , by hensel s lemma , every unit @xmath154 is a @xmath3-power in @xmath155 and so @xmath156 .
[ exemple - p=37 ] take @xmath157 and @xmath158 . then , @xmath3 splits totally in @xmath148 and @xmath159 .
a simple computation gives @xmath160 , so @xmath5 is @xmath161-rational .
[ exemple - p=2 ] one can check easily that @xmath162 is not @xmath163-rational but @xmath164 is @xmath163-rational . here is a very well - known case of the situation .
[ p - rational ] suppose that the number field @xmath5 contains a primitive @xmath3th root of unity
. then @xmath5 is @xmath3-rational if and only if there exists exactly one prime of @xmath5 above @xmath3 and the @xmath3-class group of @xmath5 ( in the narrow sense if @xmath68 ) is generated by the unique prime dividing @xmath3 .
see for example theorem 3.5 of @xcite .
in particular , when @xmath3 is regular , @xmath28 is @xmath3-rational for all @xmath27 .
now we will give some more precise statements by considering some galois action .
let @xmath25 be a galois extension of degree @xmath19 , with @xmath165 .
put @xmath166 .
let @xmath167 be the number of archimedean places of @xmath24 .
let @xmath62 be a finite set of places of @xmath24 .
the arithmetic objects that will use have a structure of @xmath168$]-modules . then for a such module @xmath169 ,
one notes by @xmath170 its character .
let @xmath171 be the teichmller character , let @xmath172 be the trivial character and let @xmath173 be the regular character of @xmath174 . [ action ] suppose the field @xmath5 is @xmath3-rational and that the real infinite places of @xmath24 stay real in @xmath5 ( this is always the case when @xmath19 is odd ) .
then @xmath175 it is well - known .
the character of the @xmath174-module @xmath176 is equal to @xmath177\chi_{\mathrm{reg}}$ ] and by dirichlet s unit theorem , the character of @xmath178 is equal to @xmath179 .
then , as @xmath5 is @xmath3-rational , @xmath180\chi_{\mathrm{reg } } - \big(\big(r_1(\k_0)+r_2(\k_0)\big ) \chi_{\mathrm{reg } } - \1\big)=r_2(\k_0 ) \chi_{\mathrm{reg } } + \1\cdot\ ] ]
the literature on genus theory is rich .
the book of gras ( * ? ? ? * chapter iv , 4 ) is a good source for its modern aspects .
all we will need in this work is the following simplified version of the main result .
[ theoriegenre ] let @xmath8 be a galois degree @xmath3 extension of number fields @xmath6 .
let @xmath62 be the set of places of @xmath5 ramified in @xmath8 ( including the infinite places ) .
then @xmath181 we will also need the following elementary fact , whose proof we leave to the reader .
[ ramsurq ] let @xmath182 and let @xmath183 be a finite set of prime numbers such that , for @xmath184 , @xmath185 , where @xmath3 is an odd prime .
then there exists a cyclic extension @xmath186 of degree @xmath3 , exactly and totally ramified at the primes in @xmath62 ( but unramified outside @xmath62 ) .
for this section , our main reference is the book of ribes and zalesskii ( * ? ? ? * chapter 2 and chapter 4 ) .
if @xmath1 is a finitely generated pro-@xmath3 group , denote by @xmath187 the group of continuous automorphisms of @xmath1 . recall that the kernel of @xmath188 is a pro-@xmath3 group and that @xmath189 , where @xmath36 is the @xmath3-rank of @xmath1 .
let @xmath1 be a finitely generated pro-@xmath3 group and let @xmath190 .
put @xmath191 .
[ primetop ] ( i ) obviously , @xmath192 is a closed subgroup of @xmath1 .
\(ii ) for integers @xmath85 , @xmath193 with equality when @xmath85 is co - prime to the order of @xmath18 . for @xmath194 , @xmath195
, one says that @xmath196 is fixed - point - free if @xmath197 .
more generally , if @xmath174 is a subgroup of @xmath187 , one says that the action @xmath198 of @xmath174 on @xmath1 is fixed - point - free if and only if @xmath199 in others words , @xmath174 is fixed - point - free if and only if , for all non - trivial @xmath200 , @xmath196 is fixed - point - free . clearly if @xmath201 is fixed - point - free then @xmath196 is fixed - point - free
; it is an equivalence when @xmath18 is of prime order @xmath202 .
we are interested in instances of groups with fixed - point - free action that arise in arithmetic contexts .
let us recall the schur - zassenhaus theorem for a profinite group @xmath44 : [ schur - zassenhaus ] let @xmath1 be a closed normal pro-@xmath3 subgroup of a profinite group @xmath44 .
assume that the quotient @xmath203 is of order co - prime to @xmath3 .
then the profinite group @xmath44 contains a subgroup @xmath204 isomorphic to @xmath174 .
two such groups are conjugated by an element of @xmath1 and @xmath205 .
see theorem 2.3.15 of @xcite or proposition 1.1 of @xcite . as first consequence ,
one has the following : [ fixedpointfreegeneral ] let @xmath1 be a finitely generated pro-@xmath3 group and let @xmath206 of order @xmath19 , @xmath207 . if @xmath198 is fixed - point - free then @xmath208 is fixed - point - free , where we recall that @xmath209 , the 2nd step in the @xmath3-central series of @xmath1 , is the frattini subgroup .
let @xmath210 .
the group @xmath211 acts on @xmath1 , on @xmath209 and on @xmath212 . by the schur - zassenhaus theorem ( applied to @xmath213 ) , the non - abelian cohomology group @xmath214 is trivial and then the nonabelian cohomology of the exact sequence @xmath215 allows us to obtain : @xmath216 which is exactly the assertion of the proposition .
see also @xcite , @xcite , @xcite .
a main observation for our paper is the converse of the previous proposition when @xmath1 is uniform : [ actionuniforme ] let @xmath1 be a uniform pro-@xmath3 group .
let @xmath217 of order @xmath19 co - prime to @xmath3 .
then @xmath218 is fixed - point - free if and only if @xmath219 is fixed - point - free .
one direction is taken care of by proposition [ fixedpointfreegeneral ] . for the other direction , we first note that @xmath220 is a @xmath211-isomorphism for @xmath221 .
we thus obtain a @xmath211-isomorphism from @xmath212 to @xmath222 .
if @xmath223 is fixed - point - free , then @xmath224 is fixed - point - free for all @xmath27 .
suppose @xmath225 satisfies @xmath226 .
as the action of @xmath212 is fixed - point - free , we have @xmath227 is trivial , so @xmath228 .
continuing in this way , we in fact conclude that @xmath229 .
let @xmath1 be a uniform pro-@xmath3 group and let @xmath190 of order @xmath19 co - prime to @xmath3 . denote by @xmath230 the character of the semi - simple action of @xmath211 on @xmath212 .
then @xmath196 is fixed - point - free if and only , @xmath231 . indeed , one has seen ( proposition [ actionuniforme ] ) that @xmath196 is fixed - point - free if and only @xmath232 is fixed - point - free , which is equivalent to @xmath231 .
suppose that @xmath1 is uniform of dimension @xmath36 .
the restriction of @xmath190 to @xmath233 is an element of @xmath234 .
denote by @xmath235 $ ] its characteristic polynomial .
then the action @xmath196 is fixed - point - free if and only if @xmath236 . when @xmath1 is uniform and @xmath237 is of order @xmath19 co - prime to @xmath3 , testing that the action of @xmath174 is fixed - point - free on @xmath1 is equivalent to testing this condition on the quotient @xmath238
let @xmath230 be the character resulting from the action of @xmath174 on @xmath239 .
then @xmath198 is fixed point - free on @xmath239 if and only if , for all @xmath240 , @xmath241 does not contain the trivial character , where here @xmath242 is the restriction to @xmath211 . by frobenius reciprocity , this condition is equivalent to @xmath243 , where @xmath244 is the induction from @xmath211 to @xmath174 .
we need also of the following proposition which will be crucial for our main result .
[ conjugueauto ] let @xmath1 be a finitely generated pro-@xmath3 group .
let @xmath18 and @xmath245 be two elements in @xmath187 of order @xmath19 co - prime to @xmath3 . if @xmath246 in @xmath212 , then @xmath247 is fixed - point - free if and only if , @xmath248 is fixed - point - free .
more precisely , @xmath249 for a certain element @xmath250 .
one needs the following lemma : [ autoexterieur ] let @xmath1 be a finitely generated pro-@xmath3 group .
let @xmath18 and @xmath245 be two elements of @xmath187 of order @xmath19 , @xmath165 , satisfying @xmath251 .
then there exists @xmath252 such that @xmath253 .
this is to be found in lemma 3.1 of @xcite .
since @xmath18 and @xmath245 coincide as elements of @xmath254 , there exists @xmath255 in the pro-@xmath3 group @xmath256 such that @xmath257 .
consider the group @xmath258 .
since @xmath259 , @xmath260 is a semi - direct product of a pro-@xmath3-group and a group of order @xmath19 . as @xmath245 and @xmath18 are both in @xmath260 , the subgroups @xmath261 and @xmath262 are conjugate to each other ( by the schur - zassenahus theorem [ schur - zassenhaus ] ) : there exists @xmath263 such that @xmath264 for a certain integer @xmath265 , @xmath266 , since @xmath18 and @xmath245 have the same order . moreover , in @xmath212 , @xmath267 , and so we can take @xmath268 . by the previous lemma , @xmath253 , for some @xmath269 in @xmath187 .
moreover , @xmath270 implies that @xmath269 induces the trivial automorphism of @xmath212 .
we have that @xmath48 is a fixed point of @xmath245 if and only if @xmath271 is a fixed point of @xmath18 .
given a uniform pro-@xmath3 group @xmath1 equipped with an automorphism @xmath18 of order @xmath19 prime to @xmath3 , the central question of this subsection is to realize @xmath272 as a galois extension over a number field .
[ relevementlibre ] let @xmath10 be a free pro-@xmath3-group on @xmath36 generators , and let @xmath273 be the natural map @xmath274 . consider a subgroup @xmath275)$ ] of order @xmath19 co - prime to @xmath3 .
then there exists a subgroup @xmath276 isomorphic to @xmath174 such that @xmath277 .
moreover , any two such subgroups are conjugated by an element @xmath278 .
first of all the natural map @xmath279 is onto ( proposition 4.5.4 of @xcite ) . put @xmath280 and recall that @xmath281 is a pro-@xmath3 group .
then , @xmath282 which has order co - prime to @xmath3 . by the schur - zassenhaus theorem [ schur - zassenhaus ] ,
there exists a subgroup @xmath283 , such that @xmath284 .
moreover , two such subgroups are conjugate to each other .
in fact , one needs a little bit more .
the following proposition can be found in a recent paper of greenberg @xcite and partially in an unpublished paper of wingberg @xcite .
[ wingberg1 ] let @xmath285 be a profinite group where @xmath10 is free pro-@xmath3 on @xmath286 generators and where @xmath174 is a finite group of order @xmath19 co - prime to @xmath3 .
let @xmath1 be a finitely generated pro-@xmath3 group on @xmath36 generators , with @xmath287 .
suppose that there exists @xmath288 , with @xmath289 , such that the module @xmath290 is isomorphic to a submodule of @xmath291 .
then there exists a normal subgroup @xmath292 of @xmath10 , stable under @xmath174 , such that @xmath293 since this result is essential for our construction , we include a proof .
[ wingberg0 ] let @xmath10 be a free pro-@xmath3-group on @xmath36 generators and let @xmath1 be a pro-@xmath3 groups generated by @xmath36 generators .
let @xmath294 be a morphism on pro-@xmath3 groups @xmath295 .
assume that there exists a finite group @xmath296 of order @xmath19 co - prime to @xmath3 .
then the action of @xmath174 lift to @xmath10 such that @xmath294 becomes a @xmath174-morphism . for a finitely generated pro-@xmath3 group @xmath292 , denote by @xmath297 the natural map @xmath298 .
recall that @xmath299 is a pro-@xmath3 group .
let @xmath300 be a minimal presentation of @xmath69 .
denote by @xmath301 .
the natural morphism @xmath302 is onto ( see ( * ? ? ?
* proposition 4.5.4 ) ) . put @xmath303 .
then @xmath304 .
now the isomorphism between @xmath305 and @xmath212 induces an isomorphism @xmath306 between @xmath307 and @xmath254 .
hence on @xmath308 , one has : @xmath309 .
in particular @xmath310 is a pro-@xmath3 group .
now let @xmath311 .
then : @xmath111 @xmath312 is a pro-@xmath3 group and @xmath114 @xmath313 which has order co - prime to @xmath3 . by the schur - zassenhaus theorem [ schur - zassenhaus ]
, there exists @xmath314 such that @xmath315 , _
i.e. _ @xmath316 and we are done . as @xmath287 , let @xmath294 be a surjective morphism @xmath317 .
put @xmath318 .
as @xmath207 , the action of @xmath174 on @xmath305 is semi - simple .
let us complete the @xmath168$]-module @xmath212 with a submodule @xmath319 such that @xmath320 as @xmath174-module .
let @xmath30 be the pro-@xmath3 group @xmath321 , where @xmath322 is generated by an @xmath323-basis of @xmath319 .
by lemma [ wingberg0 ] , there exists @xmath314 isomorphic to @xmath174 , such that the morphism @xmath324 is a @xmath204-morphism . by proposition [ relevementlibre ] ,
there exists @xmath325 such that @xmath326 .
we note that @xmath327 is equivalent to @xmath328 .
then we take @xmath329 and observe that @xmath330 is @xmath174-isomorphic to @xmath1 .
we now review a group - theoretic notion that we need for our study of the @xmath2-invariant .
let @xmath44 be a profinite group .
one says that @xmath44 is a frobenius group if @xmath331 , where @xmath1 is a finitely generated pro-@xmath3 group , @xmath174 is of order @xmath19 co - prime to @xmath3 , and such that the conjugation action of @xmath174 on @xmath1 is fixed - point - free . the notion of a frobenius group is a very restrictive one , as illustrated in the following theorem : [ rz ] let @xmath332 be a frobenius profinite group .
then the subgroup @xmath1 of @xmath44 is nilpotent .
moreover if @xmath333 , @xmath1 is abelian , if @xmath334 , @xmath1 is nilpotent of class at most @xmath163 , and more generaly of class at most @xmath335 if the prime number @xmath202 divides @xmath336 .
let @xmath332 , where @xmath1 is a uniform pro-@xmath3 group and such that @xmath337 .
then @xmath44 is a frobenius group if and only if the action of @xmath174 is fixed - point - free on @xmath212 .
it is a consequence of proposition [ actionuniforme ] .
let us recall the motivating question of this paper . given a uniform group @xmath1 of dimension @xmath36 , equipped with a fixed - point - free automorphism of finite order co - prime to @xmath3
, can one realize an arithmetic context for @xmath1 as galois group with arbitrarily large associated @xmath2-invariant ? here
we develop our strategy .
given a uniform pro-@xmath3 group @xmath1 , the key task is to produce an extension @xmath338 of a number field @xmath5 with galois group isomorphic to @xmath1 such that 1 .
there are only finitely many primes that are ramified in @xmath6 ; 2 .
there exist infinitely many primes of @xmath5 that split completely in @xmath6 .
to produce such a situation , we realize @xmath1 , along with its fixed - point - free automorphism , inside the maximal pro-@xmath3 extension @xmath339 of a @xmath3-rational field @xmath5 ; in particular the condition @xmath111 will be automatically satisfied in our situation .
[ principle ] suppose @xmath1 is a uniform pro-@xmath3 group and @xmath338 is a galois extension of a number field @xmath5 such that @xmath340 is isomorphic to @xmath1 and the following conditions hold : 1 .
there are only finitely many primes that are ramified in @xmath6 ; 2 . there
exist infinitely many primes of @xmath5 that split completely in @xmath6 . then there
exist @xmath1-extensions of number fields with arbitrarily large associated @xmath2-invariant .
since only finitely many primes of @xmath17 ramify in @xmath5 , a field of finite degree over @xmath17 , the set @xmath341 of primes of @xmath5 which split completely in @xmath338 and are neither divisors of @xmath3 nor ramified over @xmath17 is infinite . fix a finite subset @xmath62 of @xmath341 of any prescribed order .
let @xmath342 be the residue characteristics of the primes in @xmath62 .
let @xmath343 a cyclic degree @xmath3 extension , described in proposition [ ramsurq ] , which is ramified at each of these @xmath344 .
then @xmath345 is galois extension of group isomorphic to @xmath1 .
put @xmath346 and @xmath347 . then by applying theorem [ theoriegenre ] in @xmath348 ,
one obtains @xmath349\left ( |s|-r_2(\f)\right).\ ] ] one conclude with corollary [ critereminorationmu ] , thanks to the fact that @xmath62 can be arbitrarily large .
we now review the strategy of @xcite ( see also @xcite ) for finding arithmetic situations with large @xmath2-invariant .
let @xmath148 be an imaginary quadratic field .
let us denote by @xmath18 the generator of @xmath351 .
suppose @xmath3 is a rational prime which splits in @xmath352 into two distinct primes @xmath353 and @xmath354 .
let us suppose further that @xmath3 does not divide the class number of @xmath5 .
then for @xmath355 , the maximal pro-@xmath3 extension @xmath356 of @xmath5 unramified outside @xmath357 has galois group @xmath358 isomorphic to @xmath15 .
the automorphism @xmath18 permutes the fields @xmath356 . thus inside the compositum of these two @xmath15-extensions ,
if we denote by @xmath359 the subgroup fixing the field @xmath356 , then the subfield @xmath338 fixed by @xmath360 is galois over @xmath17 with galois group isomorphic to @xmath15 , and the action @xmath361 is dihedral .
[ quadratiquecase ] under the preceding conditions , @xmath362 , where the action of @xmath18 satisfies @xmath363 , @xmath47 denoting a generator of @xmath340 .
thus , all primes @xmath202 of @xmath17 which remain inert in @xmath148 subsequently split completely in @xmath6 .
the first part of the corollary follows from the observations preceding it .
for the second part , remark that the non - trivial cyclic subgroups of @xmath364 are of the form @xmath365 , @xmath366 , or of the form @xmath367 .
hence let @xmath202 be a prime which is unramified in @xmath368 .
if @xmath202 is inert in @xmath148 , then the frobenius automorphism @xmath369 of @xmath202 in @xmath368 is of the form @xmath370 .
as @xmath369 is of order @xmath163 ( or equivalently , @xmath371 ) , the prime @xmath202 splits totally in @xmath6 . by applying proposition [ principle ] to the extension @xmath368
, this construction allows us to produce @xmath15-extensions with @xmath2-invariant as large as desired .
we can say a bit more in the example [ quadratiquecase ] .
let us show that a place of @xmath5 above a prime @xmath202 splits totally in @xmath6 if and only if @xmath202 is inert in @xmath148 . for a prime @xmath372 , denote by @xmath373 the frobenius of @xmath374 in the compositum of the @xmath15-extensions of @xmath5 . let @xmath202 a prime that splits in @xmath148 ; let us write @xmath375 . if we write @xmath376 , then by conjugation , @xmath377 .
the key point is that the maximal pro-@xmath3 extension of @xmath5 unramified outside @xmath3 and totally split at @xmath378 and @xmath379 is finite , see @xcite . then @xmath380 .
note that when @xmath202 is inert in @xmath148 then @xmath381 . by reducing the frobenius modulo @xmath382
, we note that in the case when @xmath202 is inert , @xmath383 but when @xmath202 splits in @xmath148 , @xmath384 .
one can generalize the above discussion for @xmath385 , by considering a large cm - extension abelian over @xmath17 .
see cuoco ( * ? ? ?
* theorem 5.2 ) ) we observe that the basic principle in the constructions above is that there is a positive density of primes of a field @xmath24 which are inert in @xmath5 , and that all of these subsequently split ( thanks to the fact that the action @xmath196 is fixed - point - free ) in the @xmath1-extension .
this is the starting point for the general case .
we will consider a uniform pro-@xmath3 group @xmath1 of dimension @xmath36 having a fixed - point - free automorphism @xmath18 of order @xmath19 co - prime to @xmath3 .
we assume that @xmath386 ; indeed for @xmath387 , @xmath388 ( by theorem [ rz ] of ribes and zaleskii ) .
[ existence ] suppose @xmath389 is a totally imaginary number field admitting a cyclic extension @xmath390 of degree @xmath386 co - prime to @xmath3 such that @xmath5 is @xmath3-rational .
let @xmath1 be a uniform group of dimension @xmath36 having an automorphism @xmath245 of order @xmath19 .
let @xmath85 be an integer such that @xmath391 \cdot p^n \geq 2d$ ] , and let @xmath24 , respectively @xmath5 be the @xmath85th layer of the cyclotomic @xmath15-extension of @xmath389 , respectively @xmath392 .
then there exists an intermediate field @xmath393 such that @xmath338 is galois over @xmath24 with galois group isomorphic to @xmath394 . in particular ,
if @xmath245 acts fixed - point - freely on @xmath212 and if @xmath20 is prime , then @xmath395 is a frobenius group .
the extension @xmath25 is cyclic of degree @xmath19 and the number field @xmath5 is @xmath3-rational : put @xmath396 and @xmath397 .
the extension @xmath398 is a galois extension of group @xmath44 isomorphic to @xmath399 . by proposition
[ action ] , the character of the action of @xmath18 on @xmath305 is @xmath400}{2 } \cdot p^n \cdot \chi_{\mathrm{reg}}+\1}$ ] .
now let @xmath401 be the character of the action of @xmath245 on @xmath212 .
then @xmath402 and , for all @xmath230 , @xmath403 . in particular since @xmath391
\cdot p^n \geq 2 d,$ ] then necessarily , the @xmath404-module @xmath212 is isomorphic to a submodule of @xmath305 . by proposition [ wingberg1
] , there exists a normal subgroup @xmath292 of @xmath10 , stable under @xmath18 , such that @xmath405 .
we now state the key arithmetic proposition we need .
[ splitting ] let @xmath406 be a galois extension of galois group @xmath272 , where @xmath1 is a uniform group of dimension @xmath36 and where @xmath18 is of order @xmath19 co - prime to @xmath3 .
suppose that @xmath196 is fixed - point - free .
then every place @xmath147 which is ( totally ) inert in @xmath25 and is not ramified in @xmath6 splits completely in @xmath6 .
let @xmath147 a prime of @xmath5 inert in @xmath25 which is not ramified in @xmath6 .
let us fix a prime @xmath407 of @xmath338 ( see @xmath408 as a system of coherent primes in @xmath6 ) .
denote by @xmath409 be the frobenius of @xmath147 in @xmath25 and let @xmath410 be an element of order @xmath19 of the decomposition group of @xmath408 in @xmath406 lifting @xmath409 .
then @xmath411 in @xmath254 , for an integer @xmath412 , @xmath413 . by proposition [ conjugueauto ] , there exists @xmath414 such that @xmath415 , the last equality coming from remark [ primetop ] .
let @xmath416 .
as @xmath417 , the prime @xmath408 is unramified in @xmath25 and then the decomposition group of @xmath408 in @xmath406 is cyclic : the elements @xmath418 and @xmath419 commute or , equivalently , @xmath420 . hence if @xmath421 , the element @xmath422 has a fixed - point .
contradiction , and then @xmath423 .
[ main ] suppose @xmath389 is a totally imaginary number field admitting a cyclic extension @xmath390 of degree @xmath386 co - prime to @xmath3 such that @xmath392 is @xmath3-rational .
let @xmath1 be a uniform pro-@xmath3 group having an automorphism @xmath245 of order @xmath19 with fixed - point - free action .
there exists a finite @xmath3-extension @xmath424 unramified outside @xmath3 such that for any given integer @xmath425 , there exists a cyclic degree @xmath3 extension @xmath8 which admits a @xmath1-extension @xmath426 satisfying @xmath427 . by proposition [ existence ] , once we let @xmath24 and @xmath5 by the @xmath85th layer of the cyclotomic @xmath15-extensions of @xmath389 and @xmath392 respectively , where @xmath428\geq 2 \dim(\gamma)$ ] , we are guaranteed of the existence of a galois extension @xmath406 with galois group isomorphic to @xmath429 and @xmath430 . by proposition
[ splitting ] , every inert places in @xmath25 splits completely in @xmath6 .
to finish , we apply the construction of proposition [ principle ] .
let us remark that in the construction of theorem [ main ] ( in fact of proposition [ principle ] ) , we start with a @xmath3-rational field @xmath431 but its cyclic degree @xmath3 extension @xmath10 with many ramified primes is not @xmath3-rational .
indeed as @xmath432 and the extension @xmath8 is ramified at every place of @xmath62 , the extension @xmath8 is not primitively ramified when @xmath62 becomes large .
and then the field @xmath433 is not @xmath3-rational ( see ( * ? ? ?
* chapter iv , 3 ) ) . in his original treatment @xcite ,
iwasawa was able to treat the case @xmath68 alongside odd primes @xmath3 .
the elements of finite order of @xmath434 are of order @xmath163 .
then the question [ gtquestion ] is essential for applying our previous `` co - prime to @xmath3 '' strategy for @xmath435 .
indeed , let us consider the uniform pro-@xmath163 group @xmath436 instead of @xmath435 by noting that @xmath437 has a fixed - point - free automorphism @xmath245 of order @xmath438 .
by example [ exemple - p=2 ] , one knows that the field @xmath439 is @xmath163-rational ; the galois group @xmath440 is free on @xmath441 generators .
moreover , @xmath440 has an automorphism of order @xmath438 coming from the unique cyclic sub - extension @xmath25 of degree @xmath438 .
the character of this action contains the character of the action of @xmath245 on @xmath212 .
hence by proposition [ existence ] , there exists a galois extension @xmath6 with @xmath442 in which every odd inert place @xmath147 in @xmath25 splits completely in @xmath6 .
in particular every such place splits in every @xmath435-quotient of @xmath1 and then the proposition [ principle ] apply for @xmath435 .
the nilpotent uniform groups of dimension @xmath443 are all commutative . in dimension
@xmath438 they are parametrized , up to isomorphism , by @xmath444 and represented ( see ( * ? ? ?
* theorem 7.4 ) ) by : @xmath445=[y , z]=1 , [ x , y]=z^{p^s}\rangle.\ ] ] here the center of @xmath446 is the procyclic group @xmath447 and one has the sequence:@xmath448 let @xmath455 be a primitive third root of the unity and let @xmath456 be the truncation at level @xmath85 of the @xmath3-adic expansion of @xmath455 : @xmath457 .
let us consider @xmath18 defined by : @xmath458 then @xmath190 .
indeed one has to show that the relations defining @xmath1 are stable under the action of @xmath18 , which is obvious for the relations @xmath459=1 $ ] and @xmath460=1 $ ] .
let us look at the last relation .
first , as @xmath1 is uniform , let us recall that @xmath461 .
we have : @xmath462)=\sigma(xyx^{-1}y^{-1})\equiv x^{\zeta^{(n)}}y^{\zeta^{(n ) } } x^{-\zeta^{(n)}}y^{-\zeta^{(n ) } } ( \mod \ \gamma_n ) \equiv z^{\left(\zeta^{(n)}\right)^2 p^s } ( \mod \ \gamma_n ) \longrightarrow_{n } z^{\zeta^2 p^s}.\ ] ] to finish , let us show that the automorphism @xmath18 is fixed - point - free .
indeed the eigenvalues of the action of @xmath18 on the @xmath323-vector space @xmath463 are @xmath455 ( with multiplicity @xmath163 ) and @xmath464 , so @xmath465 is fixed - point - free and we conclude with proposition [ actionuniforme ] .
if @xmath466 , there is no element @xmath467 of order @xmath438 with fixed - point - free action . indeed , let @xmath468 , and suppose that @xmath18 is fixed - point - free .
the eigenvalues of @xmath18 in @xmath469 are @xmath455 and/or @xmath464 with multiplicity .
but , as the trace of @xmath18 is in @xmath323 , then necessarily @xmath470 and @xmath471 . for @xmath158 , which is the smallest irregular prime , we may not resort to the construction above with @xmath473 , but we can still realize @xmath446-extensions with arbitrarily large @xmath2-invariant by applying theorem [ main ] for @xmath474 .
let @xmath6 be a galois extension with @xmath3-adic analytic galois group @xmath1 of dimension @xmath36 .
denote by @xmath475 the set of primes of @xmath352 unramified in @xmath6 .
assume @xmath475 finite .
for @xmath417 , let @xmath476 be the conjugacy class of the frobenius of @xmath147 in @xmath6 and put @xmath477 where @xmath478 .
let @xmath3 be a regular prime and let @xmath1 be a uniform pro-@xmath3 group having a non - trivial fixed - point - free automorphism @xmath18 of prime order @xmath482 .
then there exists a constant @xmath483 and a @xmath1-extension @xmath338 over a number field @xmath5 such that @xmath484 one use the construction of section [ cadregeneral ] . put @xmath485 where @xmath85 is the smallest integer such that @xmath486 .
let @xmath25 be the unique cyclic extension of degree @xmath202 ; @xmath396 .
let @xmath6 be a uniform galois extension with group @xmath1 constructed by the method of proposition [ existence ] .
let @xmath487 be the set of conjugacy classes of @xmath351 of the elements @xmath488 with @xmath489 . by proposition [ splitting ] , a prime @xmath374 of @xmath5
for which the frobenius conjugacy class @xmath490 is in @xmath487 , splits totally in @xmath6 .
hence @xmath491 where @xmath492 is the conjugacy class of the frobenius of @xmath493 in @xmath351 . by using chebotarev density theorem ,
one concludes that @xmath494 by making use of fixed - point - free automorphisms , we have produced some uniform extensions with infinitely many totally split primes . using class field theory it is also possible to produce some free - pro-@xmath3 extensions with some splitting phenomena , but _
we do not know if it is possible to construct free pro-@xmath3 extensions in which infinitely many primes of the base field split completely_. in fact this question is related to the work of ihara @xcite and the recent work of the authors @xcite .
let us show how to produce some free pro-@xmath3 extensions with some splitting .
let @xmath62 and @xmath497 be two finite sets of places of number field @xmath5 such that @xmath498 .
recall that @xmath499 .
let @xmath500 be the signature of @xmath5 .
let @xmath501 be the maximal pro-@xmath3 extension of @xmath5 unramified outside @xmath62 and totally decomposed at @xmath497 .
put @xmath502 ; @xmath503 and @xmath504 . of course , one has @xmath505 and for @xmath506 , @xmath507 .
the first inequality of the previous proposition would allow us to produce a free pro-@xmath3 extension with complete splitting at the primes in @xmath497 if @xmath513 .
on the other hand , if we apply the second line , let us read the second of proposition [ koch - shafarevich ] with the role of @xmath62 and @xmath497 reversed , we find that @xmath514 we conclude that , under this strategy , @xmath515 so this method is rather limited in scope .
let us take @xmath516 when @xmath3 is regular and @xmath125 .
then the group @xmath508 will be pro-@xmath3 free when @xmath517 is trivial .
as the @xmath3-class group of @xmath5 is trivial , one has : @xmath518 where @xmath135 is the pro-@xmath3 completion of the local units at @xmath136 and where @xmath519 is the group of @xmath62-units of @xmath5 .
now , there are several types of scenarios where the quotient @xmath520 is trivial .
let us give one .
suppose that @xmath521 where @xmath202 is inert in @xmath148 .
then @xmath522 is isomorphic to the @xmath3-part of @xmath523 .
hence , the global @xmath3th roots of unity will kill this part when @xmath524 i.e. when @xmath525 is exactly divisible by @xmath3 .
when this is the case , the group @xmath508 is free on @xmath526 generators .
we can take these ideas a bit further by studying the galois action . throughout this subsection , we fix the following notation and assumptions ( we still assume @xmath496 ) .
let @xmath24 be a number field , @xmath25 a cyclic extension of integer @xmath19 co - prime to @xmath3 .
put @xmath166 .
let @xmath167 be the number of archimedean places of @xmath24 .
let @xmath62 and @xmath497 be two finite sets of places of @xmath24 such that @xmath498 and @xmath62 contains @xmath527 . by abuse of notation , the set of places of @xmath5 above @xmath62 and @xmath497
are again called @xmath62 and @xmath497 respectively .
denote by @xmath528 ( resp .
@xmath529 ) the set of places of @xmath62 ( resp . of @xmath497 ) splitting in @xmath25 an by @xmath530 ( resp .
@xmath531 ) the set of places of @xmath62 ( resp . of @xmath497 ) not splitting in @xmath25 . as in [ section - p - rational ] , the arithmetic objects of interest have a structure as @xmath168$]-modules .
we recall the following mirror identity from the book of gras ( * ? ? ?
* chapter ii , 5.4.2 ) : suppose that @xmath5 , which is a cyclic degree @xmath3 extension of @xmath24 , contains a primitive @xmath3th root of unity and is @xmath3-rational .
then @xmath534 , as well as every number field @xmath106 in the cyclotomic extension @xmath535 of @xmath5 is also @xmath3-rational .
then for @xmath536 , by theorem [ caracteres ] , one has : @xmath537 the group @xmath538 corresponds by class field theory to the galois group of the maximal _ abelian _ pro-@xmath3 extension of @xmath24 unramified outside @xmath62 .
the galois group @xmath539 is free on @xmath540 generators .
the action of @xmath174 on @xmath541 being semi - simple , the @xmath168$]-module @xmath542 is isomorphic to @xmath543 on which the @xmath174 action is trivial . by the chebotarev density theorem , we can choose a set @xmath544 of places of @xmath24 all inert in @xmath25 , with @xmath545 , such that the frobenius symbols @xmath546 , @xmath547 , generate the @xmath3-group @xmath548 . by the choice of @xmath497 ,
one has @xmath549 .
thanks to theorem [ caracteres ] one then obtains : @xmath550 then the maximal pro-@xmath3-extension of @xmath5 unramified outside
@xmath497 and splitting at @xmath62 , is trivial .
one then uses proposition [ koch - shafarevich ] to obtain : assume the conditions of proposition [ caractere2 ] .
let @xmath406 be a sub - extension of @xmath556 such that @xmath557 is a uniform pro-@xmath3 group .
then every ( totally ) inert ( odd ) prime @xmath558 in @xmath25 splits completely in @xmath6 .
here we exploit some conjectures on @xmath3-rationality to get further heuristic evidence in direction of conjecture [ conjecture ] .
these are of two types .
the first is in the spirit of the leopoldt conjecture in a certain sense this is a transcendantal topic ; the main reference is the recent work of gras @xcite .
the second one is inspired by the heuristics of cohen - lenstra @xcite .
[ enfamille ] let @xmath33 be an infinite set of prime numbers and @xmath19 an integer co - prime to all @xmath560 .
let @xmath35 be a family of uniform pro-@xmath3 groups all of dimension @xmath36 and all having a fixed - point - free automorphism of order @xmath19 . if conjecture [ conjectureregulator ] is true for some number field @xmath5 of degree @xmath561 , then for all but finitely many primes @xmath560 , the groups @xmath39 have arithmetic realizations as galois groups with arbitrarily large @xmath2-invariant .
let us take @xmath24 be a totally imaginary abelian number field of degree @xmath85 over @xmath17 and let @xmath562 be a degree @xmath19 cyclic extension with @xmath563 .
let @xmath564 be the compositum .
put @xmath565 . under conjecture
[ conjectureregulator ] , one knows that for large @xmath3 depending only on @xmath5 , the field @xmath5 is @xmath3-rational .
one then applies theorem [ main ] .
let us fix @xmath444 and let @xmath33 be the set of primes @xmath566 .
let @xmath567 be the pro-@xmath3 group @xmath446 of dimension @xmath438 introduced in [ exemplenoncommutatif ] .
one can apply proposition [ enfamille ] to the family of groups @xmath568 .
let us consider another idea for studying @xmath3-rationality , in the spirit of the cohen - lenstra heuristics @xcite .
for a prime number @xmath569 and an integer @xmath570 not divisible by @xmath3 , let us consider the family @xmath571 of cyclic extensions of degree @xmath19 over a fixed @xmath3-rational number field @xmath24 .
following the idea of pitoun and varescon @xcite , it seems reasonable to make the following conjecture : [ conjecture - cl ] let us fix @xmath3 and @xmath19 as above .
if @xmath24 is a @xmath3-rational quadratic imaginary field , the proportion of number fields @xmath5 in @xmath572 which are @xmath3-rational ( counted according to increasing absolute value of the discriminant ) is positive .
bolker , _ inverse limits of solvable groups _ , proc .
* 14 * ( 1963 ) , 147152 .
n. boston , _ some cases of the fontaine - mazur conjecture_. j. number theory * 42 * ( 1992 ) , no .
3 , 285291 .
n. boston , _ some cases of the fontaine - mazur conjecture .
ii_. j. number theory * 75 * ( 1999 ) , no .
2 , 161169 .
n. boston , _ explicit deformation of galois representations _ , inventiones math . *
103 * ( 1991 ) , 181 - 196 .
j. coates , p. schneider and r. sujatha , _ modules over iwasawa algebras _ , j. inst.math .
jussieu * 2*:1 ( 2003 ) , 73108 .
h. cohen and h.w .
lenstra , jr , _ heuristics on class groups on number fields _ , number theory , noordwijkerhout 1983 , lecture notes in math . ,
1068 , springer 1984 , 33 - 62 .
generalized iwasawa invariants in a family _ , compositio math .
* 51 * ( 1984 ) , 89 - 103 .
dixon , m.p.f du sautoy , a. mann and d. segal , _ analytic pro-@xmath3-groups _ , cambridge studies in advances mathematics 61 , cambridge university press , 1999 . b. ferrero and l.c .
washington , _ the iwasawa invariant @xmath573 vanishes for abelian number fields _ , annals of mathematics .
second series * 109 * ( 1979 ) , no .
2 , 377395 .
d. gildenhuysn w. herfort and l. ribes , _ profinite frobenius group _ , arch .
* 33 * ( 1979/80 ) , no .
6 , 518528 .
g. gras , class field theory , smm , springer , 2003 .
g. gras , _ les @xmath574-rgulateurs locaux dun nombre algbrique conjectures @xmath3-adiques _ , canadian j. math . , to appear .
g. gras , _ thormes de rflexion _
nombres bordeaux * 10 * ( 1998 ) , no .
2 , 399499 .
g. gras and j .- f .
jaulent , _ sur les corps de nombres rguliers _ , math . z.
* 202 * ( 1989 ) , 343 - 365 .
r. greenberg , _ galois representations with open image _ , ann .
qubec , to appear .
f. hajir and c. maire , _ on the invariant factors of class groups in towers of number fields _ , preprint 2015 . m. harris , _
@xmath3-adic representations arising from descent on abelian varieties _ , compositio math . * 39*:2 ( 1979 ) , 177 - 245 .
with correction : compositio math . * 121*:1 ( 2000 ) , 105 - 108 .
herfort and l. ribes,_on automorphisms of free pro - p - groups i _ , proc .
* 108 * ( 1990 ) , no .
2 , 287295 . s. howson , _ euler characteristics as invariants of iwasawa modules _ , proc .
london math .
society * 85 * ( 2002 ) no . 3 , 634 - 658 . y. ihara ,
_ how many primes decompose completely in an infinite unramified galois extension of a global field ? _ , j. math .
japan * 35 * ( 1983 ) , no . 4 , 693 - 709
. k. iwasawa , _ on the @xmath2-invariants of @xmath575-extensions_. number theory , algebraic geometry and commutative algebra , in honor of yasuo akizuki , pp .
kinokuniya , tokyo , 1973 . j .-
f . jaulent and t. nguyen quang do , _ corps @xmath3-rguliers , corps p - rationnels et ramification restreinte _ , j. thor .
nombres bordeaux * 5 * ( 1993 ) , 343 - 363 .
j. gonzlo - snchez and b. klopsch , _ analytic pro-@xmath3 groups of small dimensions _ ,
j. group theory * 12 * ( 2009 ) , no . 5 , 711734 . b. klopsch and i. snopce , _ a characterization of uniform pro-@xmath3 groups _ , q. j. math .
* 65 * ( 2014 ) , no .
4 , 12771291 .
m. lazard , _ groupes analytiques @xmath3-adiques _ , ihes , publ .
* 26 * ( 1965 ) , 389 - 603 . c. maire , _ sur la dimension cohomologique des pro - p - extensions des corps de nombres _ , j. th .
des nombres de bordeaux * 17 * fasc . 2 ( 2005 ) , 575 - 606 .
a. movahhedi and t. nguyen quang do , _ sur larithmtique des corps de nombres @xmath3-rationnels _ , sminaire de thorie des nombres , paris 198788 , 155200 , progr .
, 81 , birkhuser boston , boston , ma , 1990 .
a. movahhedi , _ sur les p - extensions des corps p - rationnels _ , math . nachr .
* 149 * ( 1990 ) , 163176 .
t. nguyen quang do , _ formations de classes et modules diwasawa _ , noordwijkerhout 1983 , lecture notes in math .
1068 , springer 1984 , 167 - 185 .
g. perbet , _ sur les invariants diwasawa dans les extensions de lie p - adiques _ ( french ) [ on iwasawa invariants in p - adic lie extensions ] algebra number theory * 5 * ( 2011 ) , no . 6 , 819848 . f. pitoun and f. varescon , _ computing the torsion of the p - ramified module of a number field_. ( english summary ) math . comp .
* 84 * ( 2015 ) , no .
291 , 371383 .
l. ribes and p. zalesskii , profinite groups , emg 40 , springer , 2010 . j .-
serre , _ quelques applications du thorme de densit de chebotarev _ , publ . math .
ihes , vol .
* 54 * ( 1981 ) , 123 - 201 .
o. venjakob , _ on the structure theory of the iwasawa algebra of a p - adic lie group _ , j. eur .
* : ( 2002 ) , 271311 .
k. wingberg , _
free pro-@xmath3-extensions of number fields _ , preprint 2005 . | for @xmath0 , iwasawa was the first to construct @xmath1-extensions over number fields with arbitrarily large @xmath2-invariants . in this work ,
we investigate other uniform pro-@xmath3 groups which are realizable as galois groups of towers of number fields with arbitrarily large @xmath2-invariant .
for instance , we prove that this is the case if @xmath3 is a regular prime and @xmath1 is a uniform pro-@xmath3 group admitting a fixed - point - free automorphism of odd order dividing @xmath4 . both in iwasawa s work , and in the present one , the size of the @xmath2-invariant appears to be intimately related to the existence of primes that split completely in the tower .
-rational fields . ] |
magnetic fields and their role in star formation have long been studied in the literature ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
polarization observations in the optical , infrared and submillimeter bands often routinely detect extended polarized emission , revealing magnetic field morphologies across entire star - forming regions . while the higher - frequency optical / near - infrared light is polarized as a result of scattering off dust grains ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , the submillimeter wavelengths directly show the emitted polarized light from dust .
the grains are thought to be aligned with their shorter axes parallel to the magnetic field lines in most situations , thus emitting in a plane perpendicular to the field lines @xcite .
many submillimeter facilities have developed their polarization capabilities .
operating at a wavelengths @xmath10 m , hertz on the caltech submillimeter observatory ( cso ) mapped magnetic fields in star - formation regions with a resolution of about 20 .
a summary catalog with 56 sources is given in @xcite .
its successor instrument , sharp @xcite , is currently being operated on the cso , achieving a resolution of about 10 .
dust continuum polarization was also observed with scupol on the james clerk maxwell telescope ( jcmt ) at @xmath11 m with a binned resolution of about 9 , summarized in the legacy catalog with 83 sources by @xcite .
higher - resolution maps are possible with interferometers .
the berkeley - illinois - maryland association ( bima ) array reached about 2 -4at 1.3 and 3.3 mm @xcite .
more recently , both the combined array for research in millimeter - wave astronomy ( carma ) and the submillimeter array ( sma ) presented results from legacy - type surveys toward 38 star - forming regions and cores ( @xcite , @xmath12 , @xmath13 mm ) and 14 high - mass star - forming clumps ( @xcite , @xmath14 , @xmath15 m ) .
all these observations have in common that they often cover large enough areas ( of polarized emission ) such that connected patches of projected magnetic field orientations become visible .
this is advantageous to investigate magnetic field morphologies .
consequently , several studies are making use of this in combination with other tracers such as outflows , larger scale magnetic field orientations or theoretical models . the class 0 low - mass protostar l1157-mm is found to have an hourglass morphology ( from carma observations ) that aligns both with its outflow and also with the larger scale field detected by sharp @xcite . the magnetic field in the dense clump iras 20126 - 4104 ( sharp , @xcite ) is both aligned with its bipolar outflow as well as the larger scale field from optical polarimetry data .
a comparison with numerical simulations indicates a cloud rotation axis inclined by about @xmath16 with respect to the magnetic field .
a positive correlation between mean field orientations and both pseudo - disk symmetry axes and bipolar outflow axes is derived from seven class 0 protostars in @xcite and @xcite based on sharp polarization detections , sampling @xmath17 au scales . on smaller scales , @xcite find that a subset of their sources has consistent magnetic field orientations between the larger scale ( @xmath18 , probed by single - dish submillieter telescopes ) and the smaller scale ( @xmath19 ) magnetic fields from carma .
this is interpreted as the magnetic field playing a role in regulating infall down to smaller scales ( @xmath20 au ) in these sources .
bipolar outflows ( co 2 - 1 transition ) in their sample of low - mass protostars seem to be randomly oriented with respect to the smaller scale fields @xcite , although there is a trend that low - polarization sources have their outflows perpendicular to small - scale fields , indicating that these fields might being wrapped by rotation @xcite .
the sma sample @xcite indicates that magnetic fields at dense core scales are either aligned within @xmath21 of or perpendicular to the larger parsec - scale fields .
similar to the carma result , they also conclude that outflows ( co 3 - 2 transition ) are randomly oriented with the core magnetic fields .
numerical simulations of collapsing cores , investigating scenarios to align / misalign rotation , outflows and magnetic fields are presented in e.g. , @xcite . in ideal situations , magnetic field theoretical models and simulations
can be fairly closely matched to polarization observations ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
a mathematical model to fit an observed hourglass field morphology is derived in @xcite .
their analytic formulae assuming an axisymmetric purely poloidal field threading a flattened star - forming core and connecting to a uniform background field are using the shape of the hourglass pattern to constrain the ratio of local - to - background magnetic field strength .
provided an independent measurement of the background field strength , local field strength values in a region of interest can be calculated from their analytic approach . many observed star - forming sources are , however , irregular and complicated .
it is , thus , important and desirable to develop new techniques that can directly and independently extract information from detected polarization .
a major shortcoming of dust polarization is that only sky - projected field morphologies ( rotated by 90@xmath22 from measured polarization orientations ) are observed but no magnetic field strengths .
recent developments aiming at improving this situation include the polarization - angle dispersion function leading to an estimate of the turbulent - to - mean magnetic field strength @xcite , the study of turbulent velocity dispersion spectra of co - existing ions and molecules to constrain the ambipolar diffusion scale and the plane - of - sky component field strength @xcite and the polarization - intensity gradient method to derive the local total magnetic field strength @xcite . in this work here ,
we focus on a previously introduced observable : the angle @xmath0 between the local magnetic field orientation and its underlying dust emission gradient orientation .
although not limited to dust polarization measurements , we focus on dust submillimeter observations across a large and diverse sample , with the goal to ( 1 ) illustrate characteristic features and differences in @xmath0 based on a to date largest sample of 50 sources , and ( 2 ) to provide an interpretation for @xmath0 via the polarization - intensity gradient method , illustrating clear trends in the importance of the magnetic field across the sample that are accessible via @xmath0 .
this study is part of the program on the sma @xcite to investigate the structure of the magnetic field from large to small scales .
the paper is organized as follows .
section [ section_sample ] briefly describes selection criteria and properties of the sma and the cso sample .
the main observable the angle @xmath0 its interpretation and the polarization - intensity gradient method are summarized in section [ section_analysis ] , based on earlier results in @xcite .
results for the angle @xmath0 follow in section [ section_result ] , starting from overall - sample to group and source - averaged results .
section [ discussion ] discusses interpretation and comparison with previous results .
conclusion and summary are given in section [ section_conclusion ] .
this work combines data from two different samples and telescopes ( table [ table_quantities ] ) .
hertz on the single - dish telescope cso has published a catalog of polarization observations at a wavelength @xmath23 m with a resolution @xmath24 @xcite .
the sma , operating in various array configurations at @xmath25 m , is sampling smaller scales with resolutions of a few arcseconds down to sub - arcsecond scales .
the sma has been pioneering high - resolution submillimeter polarization observations . whereas the earliest observations were focusing on the most promising and brightest targets ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , later receiver and correlator upgrades made it possible to also observe fainter sources .
this has boosted the sample of high - resolution low - and high - mass star - forming regions which currently totals about 30 sources .
a substantial number of these sources ( 14 ) was observed through a recent dedicated sma polarization legacy program @xcite , sampling star - forming clumps with densities larger than @xmath26 @xmath27 at scales around or smaller than 0.1 pc .
detailed follow - up source - by - source analyses are being conducted , with first results published for dr21(oh ) @xcite , g192.16 - 3.84 @xcite , g35.2 - 0.74 n @xcite . for our analysis here , we have applied selection criteria identical to the ones for the cso sample : coherent polarization patches with at least 10 polarization segments , and a stokes @xmath28 continuum detection strong enough to define a gradient .
besides newly detected sources from the sma polarization legacy program , the sample analyzed here also comprises some earlier observed targets ( table [ table_quantities ] ) .
the cso / hertz polarization catalog contains 56 regions , mostly galactic sources and some galaxies @xcite .
we chose a total of 27 sources from the cso catalog .
sources are selected based on their coverage of detected polarized emission , together with the condition that the stokes @xmath28 continuum emission need to be strong enough to define a gradient . in order to have at least some connected patches
, we request a minimum of 10 measured polarization
intensity gradient pairs for our analysis . @xmath2-maps
together with some basic statistical numbers for each of the cso - selected sources are given in @xcite .
the local angle @xmath0 is measured between a magnetic field orientation and a dust emission gradient ( figure [ schematic_simple ] ) . in our study here ,
an observed dust polariztion orientation which is then rotated by 90@xmath22 reveals the magnetic field orientation .
nevertheless , any measurement showing the magnetic field morphology , as e.g. , line polarization , faraday rotation or zeeman splitting , can be used and what follows is generally applicable .
@xmath0 is uanambiguously defined in the range 0 to @xmath29 ( as the smaller of the two angles complementary to @xmath30 ) , or @xmath31 to @xmath29 when giving it a sense of orientation .
@xmath0 is straighforward to derive from an observation .
a simple inspection by eye can often distinguish between zones of small ( prevailing field orientation rather orthogonal to emission contours ) and zones of large @xmath0 ( prevailing field orientation rather parallel to contours ) .
_ what is the meaning of the angle @xmath0 ?
_ with the help of an ideal magnetohydrodynamics ( mhd ) force equation , by decomposing the magnetic field orientation in an orthonormal system with unity vectors parallel and normal to an emission contour ( figure [ schematic_simple ] ) , the angle @xmath2 with @xmath32 $ ] quantifies the fraction of the available magnetic field tension force along a density gradient direction @xcite .
thus , @xmath2 can be interpreted as a magnetic field alignment deviation or efficiency that measures how the magnetic field is responsible for an observed density morphology along the gradient direction .
magnetic field lines closely aligned with density gradients , as e.g. , seen in polar regions in collapsing systems , yield @xmath33 or small , which indicates that the magnetic field tension force is not or only minimally slowing down infall in this region along the density gradient direction .
contrary to that , in regions where the magnetic field is more bent , as e.g. , in the inner part of an accretion plane where the field becomes close to orthogonal to a density gradient , @xmath2 is significantly larger than 0 . hence , the field tension force can maximally inhibit collapse or any motion along the density gradient direction . therefore , @xmath2-maps contain information on the role of the magnetic field @xcite .
changes in @xmath2 across a star - formation region can reveal zones of major or minor magnetic field influence .
additionally , many sources reveal bimodal distributions in @xmath0 when giving it a sense of orientation , i.e. , when differentiating between a field line rotated clockwise or counter - clockwise with respect to an emission gradient .
this can be interpreted as signs of a collapse ( or the beginning of a collapse ) which will start to bend initially straight magnetic field lines along the accretion direction that subsequently leads to @xmath34 above and below the accretion plane @xcite .
we note that @xmath0 and its characteristic features , found across a sample of about 30 sources in @xcite , are based entirely on observations only . in order to further quantitatively assess the magnetic field , some of the observational results in the following sections will be further analyzed with the polarization - intensity gradient method @xcite . at every position of detected polarization
, this method leads to a local magnetic field strength and a ratio between the local field tension force and gravity .
the background of this method is the ideal mhd force equation , making use of the misalignment angle @xmath2 . in a molecular cloud
, the interaction of all the forces ( magnetic field , gravity , pressure gradients ) leads to the observed magnetic field and dust morphologies .
hence , these forces leave a geometrical imprint , like the misalignment angle @xmath2 , that can be identified with the various terms in the mhd force equation . in particular ,
the direction of the field tension force is orthogonal to the detected field orientation ( i.e. , parallel to the detected dust polarization orientation ) , a field curvature can be derived from adjacent field segments , the direction of the local gravitational pull can be calculated by summing up all the surrounding mass . with this , a force triangle can be constructed at every location of detected polarization . this triangle can be closed and solved for the local field strength under the approximation that the orientation of the emission gradient is a measure for the inertial term in the mhd force equation . by taking additionally into account velocity bulk directions ,
this approximation was further studied in @xcite and confirmed to be robust for field strength and force ratio estimates .
the resulting local field strength is given by @xcite : @xmath35 where @xmath36 and @xmath37 are the gravitational pull and pressure gradient , and @xmath38 is the difference in orientation between the local gravitational pull and/or pressure gradient and the emission gradient ( figure [ schematic_simple ] ) .
@xmath39 is the density and @xmath40 is the magnetic field radius .
it is important to remark that this method provides a total magnetic field strength . with the field tension force @xmath41
, the above equation can be rewritten as : @xmath42 where @xmath43 and @xmath44 are gravitational pull and any possible additional pressure gradient .
@xmath5 quantifies how significant the local field tension force is , as compared to the other forces .
interestingly , this ratio is free of any molecular cloud model assumptions and only relies on two measurable angles that reflect the geometrical imprint of magnetic field and other forces . assuming pressure gradients to be typically small where @xmath45 , @xmath46 and @xmath47 are the molecular cloud mass , radius and temperature .
@xmath48 is related to the cloud gas composition and proportional to the hydrogen mass , @xmath49 , with a numerical factor of order unity . for pure thermal support
, this leads to @xmath50pc .
the sources in our samples comprise hundreds up to thousands of solar masses .
this leads to thermally supported length scales of @xmath5110 pc or larger .
our largest observed scales in the cso sample are about 1 pc , with most of the sources being observed on scales of @xmath52 mpc .
thus , we are very likely sampling sufficiently small scales where thermal pressure is small and negligible compared to gravity .
this also still holds if the molecular cloud temperature is higher with @xmath5320 - 30 k. yet an additional pressure gradient can result from large - scale flows .
compressive flows are suggested to contribute to and shape observed filamentary networks with coherent structures on pc- or larger scales ( e.g. , * ? ? ?
* ; * ? ? ?
significant pressure gradients might be present here , in particular for supersonic flows and turbulence .
omitting such gradients can lead to uncertainties in our analysis .
this shortcoming can be overcome if additional large - scale velocity information is available overlapping with polarization data .
the sample analyzed here , however , is again probing mostly smaller scales which likely eliminates a possibly significant bias from large - scale converging flows . finally , pressure from turbulence with velocity @xmath54 scales as @xmath55 .
with typical velocity dispersions of @xmath56 km / s in our sample , gravity is also dominating over turbulent pressure on scales of @xmath57 pc and smaller . ] compared to gravity in our samples , @xmath5 provides a criterion whether local collapse is possible ( @xmath58 ) or whether the field tension is large enough to inhibit local collapse ( @xmath59 ) .
@xmath5-maps of molecular clouds systematically show zones of smaller- or larger - than - one @xmath5-values .
the same transition can often also be observed in @xmath2-maps which makes them an easy tracer for changes in the field significance . @xmath2- and @xmath5-maps for each of the cso - selected sources ( section [ section_cso_sample ] and table [ table_quantities ] )
are given in @xcite .
finally , both @xmath5 and @xmath60 are little or not at all affected by projection effects .
this is possible because the projection correction enters as a ratio of two inclination angles ( for @xmath38 and @xmath2 ) . for identical inclination angles for magnetic field line and emission gradient
, the projection effect cancels out .
the projection effect is negligible or less than 30% as long as one inclination angle is within 45@xmath22 and the other one is not more than 50@xmath22 different from the first one @xcite .
this section starts by presenting sma - cso - combined overall results ( section [ result_sample ] ) .
this is followed by a refined analysis where sources from the combined samples are categorized based on proposed criteria ( section [ result_category ] ) .
finally , section [ result_individual ] looks into averaged source - by - source trends and compares them to the proposed categorization from section [ result_category ] .
results for both all independent @xmath0 , @xmath38 and @xmath5 , and also for their source - averaged quantities @xmath61 are shown .
we start by looking for general overall trends , based on the combined full sma - cso sample . here
, we intentionally do not distinguish between sma and cso sources or any further source properties because we are aiming at characterizing the magnetic field in any star - forming region . to this purpose
, we show for all sources with all their detected polarization orientations , their angles @xmath2 with their corresponding @xmath5-values ( figure [ figure_sample_scatter ] ) . for the total of 50 sources ,
this amounts to close to 4,000 independent ( @xmath62)-pairs .
this scatter plot clearly traces and follows an upper envelope which is determined by the maximum possible @xmath5-values which are @xmath63 for @xmath64 .
the combined sample fills the entire possible range in @xmath2 with a distribution displaying a broad peak around smaller @xmath2-values with a slow roll - off and an extended tail towards larger values ( lower panel in figure [ source_psi_delta ] ) . binning @xmath2 into intervals of 10@xmath22
shows uniform @xmath2-distributions with gradually fewer data points in subsequent bins towards larger @xmath2-values .
the distributions of @xmath5 show broad peaks around @xmath5-values initially smaller than 1 .
these peaks broaden in subsequent @xmath2 bins and shift eventually toward @xmath5-values larger than 1 . at this point here
, no distinction is made between individual sources . as we will discuss in the sections [ result_category ] and [ result_individual ] , distinct source types
preferentially fall into distinct zones in the scatter plot in figure [ figure_sample_scatter ] . filling the entire possible ranges in @xmath2 and @xmath5
possibly indicates that we might have a fairly complete and unbiased sample representing all possible source types .
we stress that it is not automatically obvious from the beginning that the scatter plot ( figure [ figure_sample_scatter ] ) needs to show a complete ( though not uniform ) coverage .
this is not obvious because the force ratio @xmath65 allows for a vast range of @xmath5-values within @xmath66 . in particular , values could be exclusively always smaller than 1 , or they could be always @xmath67 .
nevertheless , values cover a range across the dividing line @xmath68 . while a single angle @xmath2 can be affected by projection effects due to its unknown inclination angle , for a large number of more than 1,000 independent data points projection effects
are likely averaging out .
the binned uniform @xmath2-distributions together with the gradually broadening @xmath5-distributions suggest that , generally on average , larger @xmath2-values point toward larger @xmath5-values . and
indeed , @xmath2-values averaged for all the binned intervals confirm this .
furthermore , the binned values follow approximately the curve @xmath69 ( figure [ figure_sample_scatter ] ) , which is simply the mean between 0 and @xmath70 . in conclusion , for a large sample without further distinguishing any individual sources , we find a sample - wide trend where growing @xmath2-values indicate growing @xmath5-values following @xmath71 . on average ,
gravity is balanced by the magnetic field tension force , i.e. , @xmath72 , for @xmath73 .
individual sources or groups of sources can still reveal different properties ( section [ result_category ] and [ result_individual ] ) .
@xcite proposed a scenario with an evolutionary magnetic field role which leaves its imprints in the angle @xmath2 .
two distinct intermediate phases are identified ( table [ table_source_type ] ) : sources of type iia ( with their magnetic field roughly perpendicular to the cloud major axis ) , and sources of type iib ( with their magnetic field roughly parallel to the cloud major axis ) .
this categorization was motivated by the very clear examples of mon r2 ( iia ) and m@xmath74 ( iib ) from the cso catalog @xcite .
additional sources from the cso sample were then tentatively categorized in table 2 in @xcite . for completeness , an extended version of this table including the new sma sources is reproduced here ( table [ table_quantities ] ) .
since we are also interested in inter - sample differences ( between the sma and the cso ) we show results here separately for the two instruments .
the top row in figure [ hist_cat ] shows @xmath2-histograms for all the sources that are ( clearly ) identified as iia and iib .
this includes 12 iia and 7 iib sources from the cso sample and 9 iia and 4 iib sources from the sma sample .
the remaining sources are discarded because of their less complete and/or less clear polarization patterns still lists an attempted tentative categorization for all the sources in both samples . ] .
individual source - based histograms in appendix b in @xcite for the cso sample were already hinting differences in distributions ( which was an additional motivation for the classification scheme ) .
the group - based histograms now reveal a very clear difference with a falling distribution with @xmath75 ( sma ) and 30@xmath22 ( cso ) , and an opposite trend with a growing distribution with @xmath76 ( sma ) and 51@xmath22 ( cso ) . as an obvious consequence , a kolmogorov - smirnov ( ks ) test rejects the null hypothesis ( i.e. , probability @xmath77 that the two samples iia and iib originate from an identical distribution ) with @xmath78 ( sma ) and @xmath79 ( cso ) , where the maximum deviations ( @xmath80-value ) between their cumulative distributions are @xmath81 ( sma ) and @xmath82 ( cso ) .
an inter - sample ks test , i.e. , iia@xmath83 versus iia@xmath84 and iib@xmath83 versus iib@xmath84 , gives @xmath85 and @xmath86 for iia and @xmath87 and @xmath88 for iib .
we can , thus , conclude that equal - type distributions from two different instruments show a much closer resemblance than distributions from different types from the same instrument , with iib - type sources even showing a high probability to reflect identical distributions .
this is intriguing because this indicates that the sampling of different scales and densities by the sma and the cso is less relevant in the context here .
independent of that , both instruments seem to reveal the same different types of morphologies .
this can also be seen from the cumulative distributions in column 5 in figure [ hist_cat ] .
the last column shows the results for 5 type - iii sources from the sma sample .
no type - iii sources are found in the cso sample .
the @xmath2- distribution shows an even more pronounced peak towards small values with @xmath89 . besides the angle @xmath2 , the stokes @xmath28 integrated dust emission is another prime observable that reflects mass and morphology in the star - forming regions . since every source has a different absolute flux emission , a peak - normalized emission , @xmath90 $ ] ,
is used in order to compare emission distributions across samples .
the third row in figure [ hist_cat ] displays histograms of @xmath91 for iia and iib in the sma and cso samples .
for these distributions , the normalized stokes @xmath28 emission only at the positions of detected polarized emission ( where @xmath2 is identified ) is used .
as it is evident from the two cso panels , the @xmath91-histograms for iia and iib are very different among the cso sample : the iia - group shows an exponential - like fall off toward the stronger emission end , whereas type - iib sources reveal a broader distribution in @xmath91 with a more gentle decrease with brighter emission . a ks test again clearly discriminates between the two distributions ( @xmath92 , @xmath93 ) .
the mean normalized emissions are @xmath94 and @xmath95 .
the observed stokes @xmath28 emission is an integrated and projected quantity .
nevertheless , the clear difference in distributions between iia and iib is likely pointing to some underlying differences in morphology ; e.g. , iib cso sources show more abundant brighter areas than iia sources , which could be a consequence of flatter or steeper density profiles .
all sma types ( iia , iib , iii ) show exponential - like fall offs where all average values are around 0.30 .
all the sma distributions are fairly similar to the cso - iia distribution , leaving only the cso - iib distribution as clearly distinct ( blue dashed line in the cumulative distribution plot in the third row . with the clear segregation in @xmath2
, the magnetic field seems a plausible candidate to explain these .
therefore , we proceed to analyze and interpret the above observational findings with the polarization intensity gradient method @xcite .
histograms for the field significance @xmath5 are shown in the fourth row in figure [ hist_cat ] .
@xmath5-maps were previously published in appendix b in @xcite for the cso sources . for the histograms in figure [ hist_cat ] , a sample - based ( iia , iib ) @xmath96-clip
is applied to the @xmath5-values in order to remove outliers - value of the order of @xmath97 can significantly boost @xmath3 because typical @xmath5-values are of the order @xmath98 .
we note that outliers are always a consequence of the exponentially growing @xmath99 for small angles @xmath100 . ] .
similarly , values below @xmath101 ( i.e. , 1% field significance ) are removed .
both groups of iia sources show more uniform distributions up to about @xmath102 , followed by a rather sharp drop .
type - iib groups reveal a more exponential - like tail with a significant contribution from the range @xmath103 .
a ks test identifies the two types of distributions as clearly different ( cso iia versus iib : @xmath104 , @xmath105 ; sma iia versus iib : @xmath106 , @xmath107 ) .
average values differ with @xmath108 ( sma ) and @xmath109 ( sma ) and @xmath110 ( cso ) and @xmath111 ( cso ) .
the sma iii sources show an even more pronounced peak at smaller @xmath5-values with @xmath112 .
the systematic difference between iia , iib and iii is also clearly seen in the cumulative distributions with an almost complete absence of values @xmath59 for iii .
thus , the observed differences in @xmath2 between the samples iia and iib seem to point to a different role of the magnetic field , where @xmath5 provides a way to quantitatively discriminate between the samples , identifying iia ( @xmath113 ) and iib ( @xmath114 ) as potentially collapsing and non - collapsing systems on average .
the bottom row in figure [ hist_cat ] illustrates this result with @xmath5-versus-@xmath2 plots : iia sources scatter more densely around smaller @xmath2-values , while iib sources scatter around larger @xmath2-values .
type - iii sources are even more concentrated around smaller @xmath2-values ( table [ table_source_type ] ) . from the category - based properties in section [ result_category ] , we now switch our focus to the analysis of individual sources . figure [ figure_schematic ] displays source - averaged @xmath3-values as a function of the source - averaged @xmath1-values .
a @xmath96-clip is applied to each source separately in figure [ source_sigma_b_delta ] in order to calculate @xmath3 . as a result
, iia sources almost exclusively are found to have @xmath115 , with @xmath116 .
the majority of the iib sources shows @xmath114 , with @xmath117 in the range between @xmath118 to @xmath119 .
this finding is not entirely surprising after the clearly different distributions in figure [ hist_cat ] for iia and iib .
nevertheless , it is not necessarily automatically clear that ensemble - based statistical properties of iia ( @xmath120 ) and iib ( @xmath121 ) also apply to each individual source in each category .
it is , thus , rather striking that there seems to be a clear separation between iia and iib sources which identifies almost every source with @xmath6 as iia and @xmath7 as iib . additionally , it is neither obvious from figure [ hist_cat ] that average @xmath1-values fall strictly on one or the other side of @xmath122 .
we have to note that we have limited the above description to the cases that were visually based on the magnetic field versus source major axis clearly categorized as either iia or iib ( red and blue filled symbols in figure [ figure_schematic ] and [ source_sigma_b_delta ] ) .
what about sources that show less obvious patterns with , e.g. , incomplete polarization coverages , signs of fragmentation or distorted and less organized field morphologies ? the figures [ figure_schematic ] and [ source_sigma_b_delta ] show @xmath3 for all the sma and cso sources listed in table [ table_quantities ] , independent of instrument , attempted categorization or physical scale . ignoring any classification scheme ( table [ table_source_type ] ) , a clear correlation of @xmath3 with @xmath1
is seen as well .
we stress that no iia / iib - categorization is needed to arrive at this result .
our originally proposed distinction between ( clear ) iia- and iib - type sources is based on a crude visual inspection only .
nevertheless , it seems that this simple by - eye assessment is able to fairly reliably distinguish between the two regimes , @xmath115 or @xmath114 in figure [ figure_schematic ] and [ source_sigma_b_delta ] . in return
, this finding indicates : ( 1 ) generally , if magnetic field lines tend to be perpendicular to a clump / core major axis , gravity is dominating on average and overwhelming the magnetic field tension force ( @xmath115 ) .
( 2 ) generally , if the magnetic field lines tend to be aligned with a clump / core major axis , the magnetic field tension force is on average dominating over gravity .
this by - eye inspection is more precisely and quantitatively assessed with the angle @xmath2 which measures the alignment between magnetic field lines and clump / core emission gradients . from figure [ source_sigma_b_delta ]
, the transition between possibly collapsible and non - collapsible systems , on average , occurs at @xmath123 .
furthermore , it is interesting to note that sources that can not be unambiguously categorized ( empty symbols in figure [ figure_schematic ] and [ source_sigma_b_delta ] ) also follow the same trend and fit in together with the categorized sources . additionally , we mark sources that we categorize again simply by eye as hourglass - like with a pinched field morphology ( type iii in our evolutionary scenario ) with black filled symbols in figure [ figure_schematic ] and [ source_sigma_b_delta ] .
these sources presumably hint collapsing systems in an already more evolved stage ( as compared to iia and iib ) .
consistent with that expectation , all these sources show a clearly dominating gravity force with @xmath3 around 30% only , together with @xmath124 which is among the smallest values across the entire sample .
fitting a straight line to all the data ( sma , cso ; type iia , iib , iii ) gives @xmath125 ( section [ discussion_fit ] ) .
it is important to stress that @xmath3 describes a source - averaged property . in its origin , @xmath5 is , nevertheless , a local measurement . @xmath5-maps ( figure
[ figure_schematic_sigma_b ] and @xcite ) clearly show distinct zones within the same source where @xmath5 can be larger or smaller than one .
thus , for an average @xmath114 , local collapse with @xmath58 is still possible .
similarly , @xmath115 , does not necessarily mean that the entire cloud / core is collapsing . in conclusion
, there are clear and systematic differences in @xmath1 across the sample of 50 sources .
this is entirely based on observed magnetic field and dust morphologies only .
when interpreting @xmath2 in the context of the polarization - intensity gradient method , a clear correlation is found between @xmath2 and the significance of the magnetic field .
this correlation allows us to discriminate between , on average collapsible and non - collapsible systems .
despite the fact that the sma ( @xmath126 ) and the cso ( @xmath127 ) are typically probing different scales and densities , there seems to be no difference between the two instruments in figure [ figure_schematic ] and [ source_sigma_b_delta ] .
both the sma ( circles ) and the cso ( squares ) equally sample sources along the observed range in @xmath1 and @xmath3 .
it is only the type - iii sources that were exclusively found among the sma sample ( black filled circles ) .
further implications of these findings are discussed in section [ discussion ] .
finally , a @xmath96-clip is applied in figure [ source_sigma_b_delta ] .
this removes only the largest outliers which are typically found in only 1 or 2 locations in a source . as a consequence ,
the spread in @xmath3 is reduced .
a clear trend and distinct regimes are already apparent in figure [ figure_schematic ] without any clipping .
table [ table_quantities ] summarizes these values .
figure [ source_sigma_b_delta ] displays source - averaged force ratios ( field tension force over gravitational force ) , @xmath3 , versus source - averaged @xmath1 ( field alignment deviation ) .
the clear observed trend leads to a log - linear best - fit with a scaling @xmath128 with a mean error @xmath129 and a standard deviation of 0.19 for @xmath130 and @xmath131 .
_ can we extract any further meaning from exponent and numerical factor in this scaling ? _
the magnetic field significance @xmath65 is derived in @xcite .
@xmath132 is the complementary angle to the field alignment @xmath2 , i.e. , @xmath133 ( figure [ schematic_simple ] ) . in the following ,
we analyze our sample - based scaling for ( i ) @xmath1 small and ( ii ) @xmath1 large . for small angles
@xmath2 , we can write @xmath134 .
@xmath38 is expressed in radians and , if also assumed to be small , it simply becomes @xmath135 . on the other hand , expanding the above scaling , @xmath136 .
here , @xmath1 is expressed in degree , and the zeroth and first term dominate for small @xmath1 .
setting equal the two equations for @xmath137 , yields @xmath138 .
the numerical factor @xmath139 , thus , sets the limit for the sample s smallest @xmath140 if @xmath141 . for the best - fit value @xmath130 , @xmath142 .
the exponent @xmath60 ( together with @xmath139 ) defines the slope of how @xmath140 increases across the sample with growing @xmath1 .
as @xmath60 is very small , @xmath140 increases very slowly . for @xmath143 , @xmath144 . finding small angles
@xmath145 confirms the initial assumption of @xmath38 being small in the above development . for large @xmath1 we can write @xmath146 , where we assume @xmath147 , but still variable from source to source .
higher - order terms in @xmath2 are negligible as @xmath2 is expressed in radians . in the expansion of the scaling relation ,
the linear term is negligible , leading to @xmath148 . for the numerical values @xmath139 and @xmath60 , the quadratic term further dominates over the zeroth order . setting equal
again the two expressions gives @xmath149 . for our numerical values and large @xmath150 ,
this yields @xmath151 .
the angle @xmath152 grows , to a first order , with @xmath1 but depends on @xmath153 , which leads again to small values @xmath145 .
we briefly turn our attention to the angle @xmath38 , the difference between emission gradient and ( local ) gravity direction ( figure [ schematic_simple ] ) .
@xmath145 is generally found to be rather small across the entire sample ( figure [ source_psi_delta ] ) , slowly growing from an average @xmath154 to @xmath155 . in a
perfectly symmetric and circular source , emission gradient and gravity direction coincide and @xmath156 .
more realistically , in close - to - circular , elliptical or still compact systems , @xmath38 still remains small with little variations across the source ( as e.g. , in w51 e2 , iras 16293 a , iras 16293 b , ngc 1333 iras 4a , orion bn / kl main core , i18360 , g34 n , g240 ... ) .
our analysis of the scaling relation together with figure [ source_psi_delta ] possibly indicates a universal source property with @xmath145 being generally small .
this is likely a general trend because we are dealing with confined sources ( displayed with closed emission contours ) that show an accumulation of mass .
overall similar geometries lead to @xmath38 covering similar ranges in each source , consequently leading to a similar spread around similar source - averaged @xmath145 values . in summary
, linking the sample - based scaling of @xmath128 to the original derivation of @xmath5 shows that @xmath3 is largely controlled by the source - averaged magnetic field alignment @xmath1 .
the observed range for @xmath1 is @xmath157 to @xmath16 . for both small and large values in @xmath1 , the average emission gradient to gravity misalignment @xmath145
is estimated to be similar with @xmath158 and @xmath159 , respectively . the really observed range grows from @xmath160 to @xmath155 .
thus , @xmath3 is indeed influenced to a much smaller extent by @xmath145 .
this further supports the interpretation in @xcite where , based on the close resemblance between @xmath2- and @xmath5-maps , @xmath2 was proposed to be a tracer for the change in the magnetic field significance across an observed source .
morphological imprints in low- to high - magnetization simulated molecular clouds , analyzing relative orientations between magnetic field and column density gradients , were studied in @xcite .
their 2d - projected analysis reveals differences in distributions of their angle @xmath161 ( between magnetic field and column density gradients ) as a function of magnetization , column density and evolutionary stage . while histograms for low - magnetization cases ( weak field strength ) peak for all densities and time steps around @xmath162 , indicating magnetic fields that are tracing iso - contours in column densities , high - magnetization cases ( large field strength ) display a shift in distribution peaks as column densities increase from magnetic fields tracing iso - contours to magnetic fields perpendicular to contours ( figure 9 in @xcite ) .
such opposite trends are precisely seen in our @xmath0-histograms for iia- vs iib - type sources ( figure [ hist_cat ] ) .
we note that the angle @xmath161 in @xcite is defined from 0 to 180@xmath22 where @xmath163 indicates a magnetic field orientation perpendicular to a column density gradient .
we have defined @xmath164 $ ] where @xmath165 means the field is parallel to an emission gradient and @xmath166 indicates counter - clock or clockwise rotation .
@xmath161 peaking around 90@xmath22 is , thus , equivalent to @xmath0 peaking around @xmath167 ( iib ) , and @xmath161 peaking around 0 and 180@xmath22 corresponds to a peak in our @xmath0-histograms around 0 ( iia ) . in the framework of the simulations by @xcite
, this indicates that iia sources in both the cso and the sma samples are expected to have strongest magnetic fields and largest column densities .
this will even more apply to the sma type - iii sources because they show an even more pronounced peak around @xmath168 . indeed ,
large field strength values from a few mg to @xmath169 mg are measured in more evolved , collapsing type - iii sources for w51 e2 @xcite , g31.41 + 0.31 @xcite and ngc 1333 iras 4a @xcite .
it needs to be mentioned that iia / iii - type sources having the largest field strengths does not contradict with our finding in figure [ source_sigma_b_delta ] where iia / iii - type sources show @xmath170 .
@xmath5 being a ratio , a field strength can significantly increase as material is compressed while , at the same time , a growing gravitational pull from a denser core can still lead to @xmath171 . the detailed analysis in w51 e2 @xcite
is supporting this picture , revealing growing field strength values from a few mg to about 17 mg in the inner core while @xmath5 is indeed decreasing toward the center . a source - by - source comparison in our sample with the results in @xcite
is non - trivial because individual sources often do not have enough independent detected polarization segments in order to unambiguously categorize a @xmath0-histogram according to their classification scheme .
we have , therefore , focused on the clearly identified groups iia , iib and iii in our data sets . by accumulating distributions in this way , we are able to identify differences in @xmath0-distributions that can be matched with numerical results .
@xcite analyzed a sample of 24 molecular clouds , observed in submillimeter dust emission at 350 @xmath172 m with the cso / hertz with a resolution of about 20@xmath173 @xcite .
their study is motivated by the goal to determine the mean orientation of the ordered magnetic field with respect to the cloud principal axes . in their approach
they compare the mean dust polarization orientation ( derived from all the measured polarization position angles in a cloud ) with the cloud aspect ratio ( derived from the first and second moments of the dust emission ) . a statistical treatment , assuming the orientations of the molecular clouds to be random with respect to the line of sight and equally probable , is adopted to overcome projection effects .
@xcite find that oblate cloud shapes are the most likely geometry for their cso sample while the magnetic field shows a preference to be aligned with the shortest cloud ellipsoidal axis with a deviation of @xmath174 towards the middle and long cloud axes .
while their best - fitting distribution is peaking around this value , they note that long tail features are present , and the relatively small number statistics still sets limits to their conclusion .
they rule out a scenario where the magnetic field orientations are drawn randomly from a uniform distribution at the percent level . in a recent study , @xcite investigated the orientations of 13 filamentary molecular clouds in the gould belt with respect to the intercloud medium ( icm ) using near - infrared dust extinction maps and optical stellar polarimetry data .
these filaments have low column densities ( @xmath175 mag ) with length scales of @xmath176 pc .
@xcite define the orientation of a cloud by the long - axis orientation of the auto - correlation function of the column density map . in order to derive the mean magnetic field orientation of the surrounding icm ( outside of the cloud ) ,
an equal - weight stokes mean of all the detected polarization is adopted .
they find that all mean field and cloud orientations fall within @xmath177 from being either perpendicular ( 7 clouds ) or parallel ( 6 clouds ) to each other .
they exclude the possibility that such a sample can be drawn from random orientations with a probability of less than 1% .
@xcite conclude that their finding of a bimodal distribution supports the scenario of a dynamically important magnetic field , disfavoring super - alfvnic turbulence that would lead to random magnetic field orientations with clouds ( e.g. , * ? ? ?
a sample of 14 high - mass star - forming clouds , observed with the sma in dust polarization at 870@xmath172 m with a resolution of a few arcseconds , serves as a basis of a statistical analysis in @xcite .
the sma sample typically probes scales around 0.1 to 0.01 pc and densities larger than @xmath26@xmath27 . @xcite
determine the position angle of the clump major axis by fitting a 2-dimensional elliptical gaussian to the dust continuum emission .
the magnetic field orientation over the clump is not averaged but rather for every independent polarization measurement , the absolute difference in angle between the clump major axis and every measured polarization segment is computed .
the histogram of these angles reveals a bimodal distribution with two peaks : a first population of magnetic field orientations that is aligned with the clump major axis within about 10@xmath22 , and a second population associated with magnetic field segments that are perpendicular to the clump major axis within about @xmath21 . besides the above summarized studies where polarization is detected across or outside of each cloud or clump with at least several or many independent measurements , earlier studies already probed magnetic field orientations but were limited to a single detection of polarized orientation at the source intensity peak .
these studies are summarized in @xcite .
the results vary from apparent random orientations with respect to cloud axes ( @xcite , 7 elongated cores ) to a strong preference of polarized emission aligned with structures ( @xcite , 6 sources ) , and magnetic field orientations in cloud cores to be parallel in 3 out of 10 cases @xcite .
_ how do our results for the angle @xmath2 ( figures [ hist_cat ] and [ source_sigma_b_delta ] ) compare to all the previous studies ?
_ the first point to note is that all the earlier work is dealing with samples of only a few up to about 10 sources , with the largest sample being 24 sources in @xcite .
our analysis is based on 50 sources .
a second major difference is the analyses and steps taken in order to arrive at the conclusion of aligned or perpendicular fields with respect to a cloud or a core .
essentially every cited work is taking a different approach .
this also includes our work here .
further relevant to these different approaches is the concept of _ average _ ( one single global ) quantity versus _ local _ quantities .
@xcite and @xcite compare a single average magnetic field orientation ( based on all measured polarization p.a.s and equal - weight stokes mean of all detected polarization , respectively ) with a single average structure orientation ( derived from moments aspect ratios and column density auto - correlations , respectively ) .
@xcite compare a single average source orientation ( based on 2-dimensional ellipsoidal fitting ) with every independent local magnetic field orientation .
our work here compares every independent local field with every independent local emission gradient orientation . not only differ all these studies in what they are comparing ( average or local quantities ) , but they also define _ average _ in different ways . while different data sets can be suggestive for different approaches and different authors can have different preferences ,
this obviously can introduce biases leading to a non - trivial comparison .
the starting motivations in @xcite to work with _ local _ quantities was that ( 1 ) the role ( and strength ) of the magnetic field will generally vary with position in a source ; and ( 2 ) averaging might not always be an adequate operation to characterize certain sources . to be more explicit , for close - to - point symmetrical or close - to - radial field morphologies , as e.g. , in w51 e2 @xcite ,
defining an average field orientation is not unambiguous .
such systems would typically allow for two possible average orientations with a 90@xmath22 difference between them . in practise
, sources will not be perfectly symmetric .
non - symmetric features , especially if further weighted by , e.g. , stokes @xmath28 , can then tip the balance in favor of one or the other possible average orientation .
it is , however , questionable , whether such a subtlety is physically motivated .
it is unlikely to provide a robust criterion for such sources .
mathematically , the reason for this ambiguity and complication is the large dispersion between field p.a.s .
this is unlike sources that show close - to - uniform field orientations that lead to a small dispersion .
a second group of sources where averaging is not well motivated ( although it can be computed ) , are complicated systems , such as , e.g. , dr 21(oh ) @xcite where fragmentation and multiple outflows are leading to complex ( but still organized ) field morphologies .
obviously , one could remedy this situation by simply excluding any questionable sources from any statistical survey studies .
we are refraining from that because our goal is to provide a framework to assess the role of the magnetic field in any source .
based on this reasoning , our analysis here is adopting the local quantity @xmath2 is that it is unambiguously defined in the range -90 to @xmath29 ( if giving it a sense of orientation ) or 0 to @xmath29 ( if working with absolute values ) .
this is because @xmath0 and @xmath2 are _ differences _ between 2 p.a.s which makes them rotationally invariant quantities , i.e. , they are independent of any reference coordinate .
this is unlike p.a.s that are measured with respect to some orientation and reference frame .
additionally , p.a.s suffer from a 180@xmath22-ambiguity resulting from the fact that we are measuring orientations only and not directions with vector heads . ] . despite the differences and complications in comparing the various studies
, we are now turning our attention to the question whether there is bimodality in @xmath2 or @xmath1 ( the source - averaged quantity displayed in figure [ source_sigma_b_delta ] ) .
histograms for @xmath1 and @xmath2 are shown in figure [ bimodality_sma ] for the entire sma sample and its different categories iia , iib and iii .
as expected , iia and iib populate the lower and upper end in the @xmath1-histogram .
type - iii sources mostly add to the lower end .
interestingly , some iia and iii sources also fall around @xmath178 . the remaining sources neither classified as iia , iib nor iii exclusively fall around mid - values @xmath179 .
we remark again that the categorization of iia , iib and iii is simply motivated by visual inspection and does in no way affect the derivation of @xmath1 .
while there is an indication of a possible bimodality with peaks around @xmath159 and @xmath180 driven mostly by iia and iib sources the histogram reveals a third peak around @xmath179 .
_ what causes this central peak ?
_ @xmath2 very clearly and effectively differentiates between magnetic fields locally aligned or perpendicular to a source structure ( emission gradient ) .
consequently , when calculating an average @xmath1 , larger or smaller source major - to - minor axes ratios can lead to a spread around typical iia and iib @xmath1-values ; e.g. , for a source with an axes ratio 3:1 with a magnetic field parallel to its minor axis ( type iia ) , @xmath1 will be small @xmath159 .
if the source is less filamentary , becoming more compact and shorter ( but still with the field parallel to its minor axis ) , the axes ratio decreases and @xmath1 will grow to @xmath181 .
similarly , for iib - sources where the field is aligned with the source major axis , @xmath1 will decrease as the source becomes less filamentary .
all this is simply a result of averaging over more or fewer small ( iia ) or large ( iib ) @xmath2-values because of more or less elongated sources .
this then explains a spread in @xmath1 , that can populate mid - values .
studies that are comparing mean source orientations versus mean or individual magnetic field orientations @xcite are less subject to such a spread .
the second group falling onto mid-@xmath1 values are sources that are neither categorized as iia , iib nor iii .
although there are few of these sources , they do exist . in summary , lower- and upper - end peaks in the top panel in figure [ bimodality_sma ] can become more pronounced by taking into account spreading in @xmath1 for filamentary sources .
this would point toward bimodality . at the same time , nevertheless , there are clearly sources that fall in between .
the histogram for the cso sample reveals a very similar result ( figure [ bimodality_cso ] , top panel ) , with an upper - end peak that appears more present .
_ is there bimodality in @xmath2 before averaging ? _ the bottom panels in the figures [ bimodality_sma ] and [ bimodality_cso ] show normalized histograms where @xmath2-histograms for each source are normalized to one and then summed up .
this normalization removes a possible bias where ( nearby ) sources with many more independent measurements could dominate the statistics .
iia- and iib - categories show the expected trends with falling and rising distributions with larger @xmath2 .
this is largely consistent with the histograms in figure [ hist_cat ] that are counting every single independent measurement . combining the two categories results in a distribution falling with @xmath2
this is further amplified when adding type - iii sources for the sma and non - categorized sources .
assuming that the sma and cso samples represent a fair cross - section of star - forming regions , _ this result shows that the prevailing source - magnetic field configuration in both the sma and cso sample is one that statistically prefers small magnetic field misalignments @xmath2 . _ with this additional information we might re -
interpret the upper panels in figure [ bimodality_sma ] and [ bimodality_cso ] : the @xmath1-histograms possibly hint a maximum peak around smallest @xmath1-values with a tail toward larger @xmath1-values .
this interpretation is gaining additional support when combining the sma and cso samples into one @xmath1-histogram ( figure [ bimodality_sma_cso_combined ] ) , which shows a broader distribution .
@xmath1 is less than @xmath182 in 75% of all the sources .
one third of the sources shows @xmath183 .
a uniform distribution seems to be ruled out .
we finally note that earlier studies might already have seen hints of such distributions as well .
@xcite suggest a combination of two distributions ( 0@xmath22 to 40@xmath22 and 80@xmath22 to 90@xmath22 ) with a weight 5:3 to explain their histogram measuring angles between source major axis and polarization orientations . both @xcite and @xcite state bimodality based on fields aligned within 40@xmath22 and orientations being perpendicular within 30@xmath22. these are wide ranges , given only a 0@xmath22 to 90@xmath22 range to fully account for between perfectly aligned and perfectly perpendicular .
bimodality , if at all , is , thus , very unlikely to occur with sharp and narrow peaks . with the to date largest sample of 50 sources it seems to become more plausible that we are dealing with broad distributions peaking at small misalignment values with a tail and a possibly secondary peak or bump toward larger misalignments . with a sample of 24 sources , @xcite also noticed the presence of long - tail features in their analysis . if further confirmed with even larger samples , this finding would then favor a scenario where the dominating configuration in sources from @xmath51pc to @xmath51 mpc scale
has the magnetic field roughly aligned with the source minor axis . on much larger scales ( @xmath184 ) , tracing the interstellar medium ( ism ) with polarized dust emission at 353 ghz , recent _ planck _ all - sky observations @xcite , investigating the relative orientations between the magnetic field and dust structures , show that in the diffuse ism structures are preferentially aligned with the magnetic field .
a glimpse into denser molecular cloud complexes further reveals that this degree of alignment decreases with increasing column density .
the magnetic field tends to become more perpendicular to the interstellar dust ridges , i.e. , aligned with their shorter axes .
our results for @xmath2 on smaller scales and higher densities agree with this trend . having derived a field significance @xmath3 for each source in figure [ figure_schematic ] and [ source_sigma_b_delta ] , we can now proceed to estimate further source properties : star - formation efficiency and mass - to - flux ratio . since the magnetic field is providing resistance against gravity
quantified with the magnetic field tension - to - gravity force ratio @xmath5 gravity is not everywhere equally effective to initiate and drive a collapse .
a concept of magnetic - field - diluted gravity was explored in @xcite .
with @xmath5 we can now define a _ gravity efficiency _ @xmath185 ; i.e. , if @xmath186 or very small , the entire non - diluted gravitational force from an enclosed mass distribution is driving a collapse and the gravity efficiency @xmath187 ; if @xmath5 is large , the magnetic field is holding against gravity and the effectively acting gravity appears reduced with @xmath188 @xcite . the gravity efficiency is limited to the range between 0 and 1 .
we set @xmath189 if @xmath190 where collapse is inhibited .
the free - fall star - formation efficiency is defined from the free - fall collapse time and mass accretion solving a ( pressureless ) momentum equation with only a gravitational force term .
absorbing the gravity efficiency factor @xmath191 into an effective gravitational force leads to a magnetic - field - reduced star - formation efficiency ( as compared to a free - fall efficiency ) @xmath192 where the square root appears because collapse time is inverse proportional to the square root of the enclosed mass @xcite .
figure [ efficiency_mass_flux ] displays magnetic - field - reduced star - formation efficiencies , @xmath193 , for the entire sma and cso samples based on @xmath3 .
it is , nevertheless , worth noting that @xmath191 like the prime observable @xmath0 can be locally measured . in @xcite , azimuthally averaged gravitational efficiencies @xmath194 showed clear radial profiles for w51 e2/e8 from several observations , being close to 1 in the core and dropping to @xmath18% or less in the outer regions . ] as a function of @xmath1 .
the sample - wide trend shows that the magnetic field is able to reduce the star - formation efficiency .
moreover , the efficiency tends to drop with larger @xmath1 .
the sma type - iii sources show a close - to - maximum efficiency , largely unaffected by the magnetic field .
generally , type - iib sources reveal a more significantly reduced average efficiency than iia sources , dropping to 0 ( @xmath195 ) and possibly preventing ( further ) collapse for @xmath196 .
it is important to note that _ local _ collapse and _ local _ star formation are still possible or can be hindered according to the @xmath5-patterns in figure 6 .
sample - averaged efficiencies are around 50% .
finally , it needs to be noted that these reduced efficiencies are upper limits because not all of a cloud s / core s mass might take part in a collapse .
this leads to an additional volume efficiency factor , @xmath197 , which can further reduce the total efficiency as @xmath198 @xcite .
the conservative estimate here is using @xmath199 .
the second source property we are investigating is the mass - to - flux ratio . unlike earlier studies ( e.g. , @xcite and references therein ) that are relying on a combination of magnetic field strength measurements ( like zeeman line splitting ) and a mass proxy to estimate this quantity
, we are again making use of the field tension - to - gravity force ratio @xmath5 .
explicitly writing out the forces and expressing them in terms of magnetic flux and mass enables us to link @xmath5 to the mass - to - flux ratio .
the source - averaged mass - to - flux ratio , normalized to the critical mass - to - flux ratio , can be written as @xmath200 @xcite . the inverse square root results from the magnetic flux @xmath201 being proportional to the field strength @xmath60 whereas the tension force is proportional to @xmath202 .
the numerical factor @xmath203 appears because magnetic flux is defined across a cylindrical cross - section @xmath204 .
normalized mass - to - flux ratios and the gravity efficiency @xmath191 , the mass - to - flux ratio based on @xmath5 can also be locally computed . a clear transition from magnetically subcritical at larger
radii to supercritical towards the core center was found for w51 e2 in @xcite . here
, we are again looking for sample - wide trends adopting source - averaged mass - to - flux ratios . ]
are depicted in the bottom panel in figure [ efficiency_mass_flux ] versus @xmath1 .
the overall result is a mass - to - flux ratio decreasing with larger misalignment angles @xmath1 , leaving most of the type - iia and -iii sources in a magnetically supercritical state while most of the type - iib sources fall below the critical mass - to - flux ratio into a subcritical state .
this trend and transition are largely consistent with the scaling of the field significance @xmath5 in figure [ source_sigma_b_delta ] .
a sample of 50 star - forming regions observed in submillimeter dust polarization with the sma and the cso is studied with the goal of assessing the significance of the magnetic field .
our analysis is making use of the fundamental observable @xmath0 , which is the angle between magnetic field orientation and dust emission gradient .
the key results are summarized in the following . 1
. across the entire sma - cso - combined sample , @xmath2 ( from about @xmath205 independent measurements ) is populating the entire range between 0 to @xmath206 with a broad peak around smaller values ; i.e. , the magnetic field is found to range from locally aligned ( @xmath207 ) to locally maximally misaligned ( @xmath206 ) with respect to the dust emission gradient . for this large sample , projection effects likely average out .
when interpreting @xmath2 with the polarization intensity gradient method , growing @xmath2 generally indicates growing magnetic field significance @xmath5 ( the ratio of magnetic field tension force to gravity and/or pressure gradient ) . for local independent measurements , @xmath5 can be estimated on average with @xmath208 .
categorization : _ various groups ( iia : tendency of magnetic field to be parallel to source minor axis ; iib : tendency of magnetic field to be parallel to source major axis ; iii : symmetrized source ) show striking differences in group - based @xmath2-histograms .
type - iia and iib samples depict opposite trends , peaking at small and large @xmath2-values , respectively .
type - iii sources ( only found in sma observations ) peak at even smaller @xmath2-values .
sources that can not be categorized , can fall into or in between any of these groups .
group - based differences in @xmath2 manifest themselves with clear differences in the group - based field significance : @xmath209 and @xmath210 .
source - averaged scaling across entire sample : _ a clear correlation is found between the source - averaged field misalignment @xmath1 and the source - averaged field significance @xmath3 . the combined sma - cso sample yields a best - fit log - linear scaling @xmath125 with a mean error @xmath211 .
the angle @xmath38 between dust emission gradient and local gravity and/or pressure gradient , which is used in the exact derivation of @xmath65 , shows only small variations across the sample .
this explains the tight correlation between @xmath1 and @xmath3 and makes @xmath1 a prime observable to trace and quantify the role of the magnetic field .
these findings also hold separately for both the sma and the cso sources , possibly hinting a scale - free self - similar magnetic field influence .
individual sources that are clearly categorized as iia ( and iii ) and iib almost exclusively show average @xmath212 ( collapse possible , on average ) and @xmath213 ( no collapse , on average ) . in either case ,
local collapse can be possible or hindered as @xmath5-maps show distinct zones with @xmath5 larger or smaller than one .
comparison with numerical study : _ numerical simulations of low- to high - magnetization molecular clouds find distributions in @xmath0 that can be matched with our observed iia- and iib - type segregation . in this framework ,
iia- ( and iii-)type sources represent higher magnetization scenarios with stronger magnetic fields .
large magnetic field strengths are , indeed , measured for , e.g. , w51 e2 , g31.41 + 0.31 and ngc 1333 iras 4a which are categorized as iia and iii . 5 . _ bimodality : _ distributions of the average magnetic field misalignment @xmath1 for the sma and cso samples possibly are bimodal . however , combining them together broadens the distributions with a broad maximum peak around small @xmath1-values , a populated mid - range and a secondary smaller peak at larger @xmath1-values . assuming the sma and cso samples to be representative ,
the most abundant magnetic field - source configuration has , on average , the magnetic field roughly aligned with the emission gradient .
this is favoring magnetic fields that are roughly aligned with the source minor axis .
75% of the sources show an average misalignment @xmath1 of less than @xmath182 .
6 . _ star - formation efficiency and mass - to - flux ratio : _ based on the field tension - to - gravity force ratio @xmath5
, an effectively acting gravitational force can be defined with a gravity efficiency factor @xmath214 .
consequently , in the presence of a magnetic field , the standard free - fall collapse time grows and the star - formation efficiency is reduced .
the sample - wide trend shows a drop in star - formation efficiency with larger @xmath1 with the average star - formation efficiency being significantly reduced or even suppressed for some sources with @xmath196 .
@xmath5 can also be linked to the normalized mass - to - flux ratio .
a transition from magnetically supercritical ( mostly for type - iia and -iii sources with small @xmath1 ) to subcritical ( for sources with larger @xmath1 ) is found across the sample .
the authors thank the referee for valuable comments and suggestions that further improved this manuscript .
spl and tcc are supported by the ministry of science and technology ( most ) of taiwan with grant most 102 - 2119-m-007 - 004-my3 .
pmk acknowledges support through grant most 103 - 2119-m-001 - 009 .
mp acknowledges the financial support of the ocevu labex ( anr-11-labx-0060 ) and the a*midex project ( anr-11-idex-0001 - 02 ) funded by the `` investissements d@xmath215avenir '' french government programme managed by the anr .
andr , ph . , di francesco , j. , ward - thompson , d. , inutsuka , s. , pudritz , r.e .
, & pineda , j. , arxiv:1312.6232 , accepted for publication as a review chapter in protostars and planets vi , university of arizona press ( 2014 ) , eds .
beuther , r. klessen , c. dullemond , th . henning schematic of the relevant angles to calculate the magnetic field significance @xmath65 .
the angle @xmath0 is measured between magnetic field orientation ( red solid line ) and intensity gradient orientation ( blue solid line ) , covering the range @xmath216 or @xmath217 if additionally given a clockwise or counter - clockwise sense of rotation .
@xmath132 is its complementary angle , i.e. , @xmath133 .
the magnetic field tension force is directed normal to the field line along the unity vector @xmath218 , collinear to the originally detected dust polarization orientation ( red dashed line ) which is rotated by @xmath206 with respect to the magnetic field in the case of ( sub-)millimeter dust polarization emission .
the intensity gradient is normal to the emission intensity contour ( black solid line ) along the unity vector @xmath219 , forming an angle @xmath38 with the ( local ) gravity and/or pressure gradient orientation ( black dashed line ) .
the unity vector tangential to the intensity contour , @xmath220 , forms an orthonormal system together with @xmath221 .
figure reproduced from @xcite . ]
@xmath5 vs @xmath2 for all sources from the combined sma and cso samples ( close to 4,000 independent ( @xmath222)-pairs ) .
bin - averaged values are shown with red squares and are connected with straight lines .
values are binned in 10@xmath22 intervals .
the black dashed line shows the upper envelope @xmath223 .
the averaged value in the last bin around @xmath224 falls beyond the upper envelope because of the exponential growth in @xmath5 for @xmath225 .
the dashed green line is @xmath226 . ]
histograms for categorized source groups : the columns one and two are for sma - iia and cso - iia , columns three and four are for sma - iib and cso - iib and column six is for sma - iii .
distributions for observed absolute @xmath2 , relative @xmath0 and @xmath91 are shown in the top , second and third row .
clear differences are apparent .
@xmath5-histograms , derived from the polarization - intensity gradient method , are displayed in the forth row .
solid vertical lines mark mean values .
the bottom row shows scatter plots of @xmath5 vs @xmath2 , illustrating how the different categories fill in different areas in this parameter space .
column five displays cumulative distributions for all the histograms for @xmath2 , @xmath0 , @xmath91 and @xmath5 for iia ( red ) , iib ( blue ) and iii ( black ) where solid lines are used for the sma and dashed lines for the cso distributions . ]
source - averaged @xmath227-values vs source - averaged @xmath117-values for the combined sma and cso samples .
red and blue filled symbols show the clearly identified iia- and iib - type sources from both the sma ( circles ) and cso ( squares ) sample ( table [ table_quantities ] ) .
empty symbols correspond to sources with less clear features .
black filled circles are the clearly identified sma - iii sources .
averages are calculated including outliers .
schematic magnetic field structures are illustrated for iia- ( red segments ) , iib- ( blue segments ) and iii - type sources ( white segments ) .
white contours display the dust continuum emission .
the black - to - white color grading on the top indicates the local @xmath2 .
clear opposite trends are apparent for type - iia and -iib sources .
all 3 schematic types are generated by over - gridding and interpolating really observed sources ( iia , iib : g240 and g35.2 from @xcite and @xcite ; iii : w51 e2 from @xcite ) .
the systematic local changes in @xmath2-maps reflect changes in @xmath5-maps @xcite which explains the observed correlation between @xmath227 and @xmath117 across the two samples .
properties for the various source types are summarized in table [ table_source_type ] .
@xmath5-maps are shown in figure [ figure_schematic_sigma_b ] . ]
source - averaged @xmath227-values vs. source - averaged @xmath117-values for the combined sma and cso samples .
different from figure [ figure_schematic ] , outliers in @xmath5 are removed for each source individually with a @xmath96 clip .
symbols and colors are identical to the ones in figure [ figure_schematic ] .
the horizontal line marks the force equilibrium @xmath228 , below and above which gravity , _ on average _ , is or is not overwhelming the magnetic field .
@xmath229 possibly discriminates between systems with , _ on average _ , no / slow collapse and faster collapse . in either case ,
_ local collapse _ and _ local star formation _ can be possible or hindered as @xmath5-maps show distinct zones with @xmath5 larger or smaller than one ( figure [ figure_schematic_sigma_b ] and @xcite ) .
clearly identified iia - type ( red ) and iib - type ( blue ) sources appear to fall onto the left and right side of the transition around @xmath230 .
the solid gray line is the log - linear best fit @xmath128 with @xmath130 and @xmath131 .
the dashed and dotted gray lines indicate mean error ( @xmath231 ) and @xmath96 ( @xmath232 ) bounds .
the lower panel is identical to the upper panel but in linear - linear scale . ]
@xmath5-maps for the 3 schematic types iia , iib and iii ( figure [ figure_schematic ] ; iia , iib : g240 and g35.2 from @xcite and @xcite ; iii : w51 e2 from @xcite ) , derived from the effectively observed resolutions listed in table [ table_quantities ] .
the black - to - white color grading indicates @xmath5 .
corresponding black and white patterns between @xmath2 ( figure [ figure_schematic ] ) and @xmath5 are visible . ]
top panel : source - averaged @xmath233-values vs source - averaged @xmath117-values for the combined sma and cso samples .
symbols and colors are identical to the ones in figure [ figure_schematic ] .
vertical lines show the dispersion in @xmath38 for each source .
the thick gray line is a linear fit to the averaged values , revealing an only small average increase in @xmath145 with @xmath117 across the entire sample from @xmath160 to @xmath234 .
this is significantly less than the change in @xmath117 which makes @xmath117 the main parameter to drive changes in the magnetic field significance @xmath5 .
small black dots show all independent @xmath235 measurement pairs .
lower panel : histograms for the sma - cso - combined sample for @xmath2 and @xmath38 for each measured independent position ( not averaged ) .
while @xmath2 shows a flatter distribution , @xmath38 is clearly peaked toward small angles .
this suggests again rather constant , small values for @xmath38 with minor changes across the sample , as it is found in the upper panel for averaged angles . ]
histograms for the sma sample : source - averaged @xmath1 ( top panel ) and @xmath2 ( bottom panel ) .
bins are slightly shifted in the upper panel for a better display of the various groups . _
sample _ refers to the entire sample ( 23 sources ) including sources that are not categorized . in the bottom panel ,
@xmath2-histograms for each source are normalized to 1 , and then summed up in each bin . ]
magnetic - field - reduced star formation efficiency ( top panel ) and normalized mass - to - flux ratio ( bottom panel ) as a function of source - averaged @xmath1 .
the solid black line marks the critical mass - to - flux ratio , separating systems that are , _ on average _ , supercritical or subcritical . in either case ,
_ local collapse _ can be hindered or possible according to the schematic @xmath2- and @xmath236patterns in figure [ figure_schematic ] and [ figure_schematic_sigma_b ] .
symbols and colors are identical to the ones in figure [ figure_schematic ] . ]
g240 & 4.7 & 870 & 2.1 & 48 & & 0.81 & 83 & -83 & 24 & 23 & 1.62 ( 3.71 ) & 0.02 ( 0.02 ) & 0.35 ( 0.41 ) & 0.30 ( 0.52 ) & iii + g34.4.0 & 1.57 & 870 & 1.5 & 11 & & 0.93 & 87 & -61 & 28 & 23 & 1.16 ( 4.27 ) & 0.01 ( 0.001 ) & 0.26 ( 0.32 ) & 0.25 ( 0.54 ) & iii + g34.4.1 & 1.57 & 870 & 1.5 & 11 & & 0.42 & 89 & -89 & 34 & 25 & 10.6 ( 54.8 ) & 0.02 ( 0.007 ) & 0.87 ( 1.84 ) & 1.66 ( 6.55 ) & iia + g35.2 & 2.19 & 870 & 1.5 & 16 & & 0.63 & 90 & -87 & 54 & 24 & 19.4 ( 194 ) & 0.02 ( 0.02 ) & 1.86 ( 5.11 ) & 3.22 ( 25.2 ) & iib + ngc 6334a & 1.7 & 870 & 5.2 & 43 & & 0.60 & 84 & -84 & 56 & 21 & 8.96 ( 10.3 ) & 0.19 ( 0.19 ) & 1.91 ( 2.22 ) & 1.99 ( 2.53 ) & iib + ngc 6334i & 1.7 & 870 & 1.5 & 12 & & 0.81 & 82 & -90 & 50 & 23 & 10.1 ( 136 ) & 0.02 ( 0.003 ) & 1.14 ( 2.56 ) & 1.42 ( 14.0 ) & iib + ngc 6334 in & 1.7 & 870 & 2.4 & 20 & & 0.81 & 79 & -87 & 38 & 27 & 2.52 ( 3.93 ) & 0.04 ( 0.04 ) & 0.85 ( 0.91 ) & 0.52 ( 0.65 ) & iia + ngc 6334v & 1.7 & 870 & 2.9 & 24 & & 0.78 & 79 & -89 & 39 & 24 & 12.5 ( 49.5 ) & 0.02 ( 0.01 ) & 0.84 ( 1.57 ) & 1.86 ( 6.35 ) & iia + i 18360 & 4.8 & 870 & 1.6 & 37 & & 0.70 & 89 & -86 & 41 & 28 & 3.95 ( 6.07 ) & 0.04 ( 0.04 ) & 0.58 ( 0.93 ) & 0.98 ( 1.66 ) & iib + dr 21(oh ) & 1.7 & 870 & 1.3 & 11 & & 0.74 & 88 & -87 & 40 & 26 & 7.04 ( 21.8 ) & 0.02 ( 0.007 ) & 0.99 ( 1.41 ) & 1.30 ( 2.73 ) & - + g5.89 @xmath237 & 2 & 870 & 2.4 & 23 & & 0.72 & 76 & -84 & 35 & 26 & 2.88 ( 8.34 ) & 0.01 ( 0.008 ) & 0.55 ( 0.73 ) & 0.66 ( 1.41 ) & - + orion bn / kl @xmath237 & 0.48 & 870 & 2.8 & 6.4 & & 0.70 & 89 & -90 & 44 & 25 & 30.4 ( 136 ) & 0.02 ( 0.002 ) & 1.26 ( 4.72 ) & 3.99 ( 20.1 ) & - + w51 e2 - ext @xmath237 & 7 & 870 & 0.7 & 24 & & 0.95 & 33 & -47 & 18 & 12 & 1.26 ( 1.35 ) & 0.02 ( 0.007 ) & 0.27 ( 0.30 ) & 0.30 ( 0.33 ) & iii + w51 e2 - sub+comp & 7 & 870 & 2.0 & 68 & & 0.92 & 77 & -34 & 18 & 16 & 0.35 ( 0.35 ) & 0.06 ( 0.002 ) & 0.18 ( 0.17 ) & 0.09 ( 0.10 ) & iia / iii + w51 e8 - sub+comp & 7 & 870 & 2.0 & 68 & & 0.76 & 85 & -74 & 22 & 22 & 1.35 ( 1.35 ) & 0.12 ( 0.12 ) & 0.63 ( 0.63 ) & 0.31 ( 0.31 ) & iia / iii + w51 e2+e8 - sub+comp & 7 & 870 & 2.0 & 68 & & 0.84 & 85 & -74 & 21 & 20 & 2.0 ( 2.57 ) & 0.02 ( 0.02 ) & 0.44 ( 0.50 ) & 0.36 ( 0.51 ) & iia / iii + w51 north - ext @xmath237 & 7 & 870 & 0.7 & 24 & & 0.89 & 82 & -81 & 34 & 23 & 2.48 ( 4.37 ) & 0.03 ( 0.005 ) & 0.64 ( 0.71 ) & 0.54 ( 0.79 ) & iia / iii + w51 north - sub@xmath238comp & 7 & 870 & 2.0 & 68 & & 0.90 & 70 & -53 & 27 & 15 & 1.25 ( 1.25 ) & 0.03 ( 0.003 ) & 0.39 ( 0.36 ) & 0.31 ( 0.32 ) & iia + ngc 1333 iras4a @xmath237 & 0.3 & 870 & 1.2 & 1.7 & & 0.82 & 84 & -89 & 27 & 19 & 1.43 ( 4.27 ) & 0.01 ( 0.002 ) & 0.36 ( 0.43 ) & 0.29 ( 0.58 ) & iia / iii + iras 16293 a @xmath237 & 0.12 & 870 & 2.5 & 1.5 & & 0.86 & 90 & -86 & 36 & 23 & 10.1 ( 177 ) & 0.02 ( 0.004 ) & 0.80 ( 1.48 ) & 1.08 ( 9.66 ) & iii + iras 16293 b @xmath237 & 0.12 & 870 & 2.5 & 1.5 & & 0.57 & 89 & -90 & 39 & 25 & 3.92 ( 10.1 ) & 0.01 ( 0.006 ) & 0.64 ( 0.83 ) & 0.60 ( 1.35 ) & iia / iii + g31.41 @xmath237 & 7.9 & 870 & 0.9 & 34 & & 0.92 & 73 & -82 & 21 & 21 & 1.30 ( 3.83 ) & 0.01 ( 0.05 ) & 0.28 ( 0.35 ) & 0.23 ( 0.51 ) & iii + w3 ( oh)@xmath237 & 2.04 & 870 & 1.5 & 15 & & 0.79 & 77 & -87 & 35 & 22 & 2.50 ( 7.74 ) & 0.02 ( 0.001 ) & 0.54 ( 0.84 ) & 0.57 ( 1.64 ) & iia + & & & & & & & & & & & & & & & + & & & & & & & & & & & & & & & + + co@xmath239/ m@xmath240 & 7.9 & 350 & 20 & 770 & & 0.71 & 89 & -88 & 48 & 25 & 13.9 ( 54.1 ) & 0.02 ( 0.02 ) & 1.69 ( 2.47 ) & 2.21 ( 5.97 ) & i/- + mon r2 & 0.95 & 350 & 20 & 92 & & 0.78 & 67 & -55 & 23 & 18 & 1.17 ( 1.17 ) & 0.01 ( 0.005 ) & 0.34 ( 0.32 ) & 0.29 ( 0.29 ) & iia + m@xmath74 & 7.9 & 350 & 20 & 770 & & 0.61 & 88 & -89 & 57 & 24 & 13.0 ( 38.6 ) & 0.01 ( 0.005 ) & 1.83 ( 2.41 ) & 2.88 ( 5.64 ) & iib + ngc 2068 lbs10 & 0.4 & 350 & 20 & 39 & & 0.62 & 77 & -84 & 51 & 24 & 5.33 ( 5.33 ) & 0.02 ( 0.02 ) & 1.38 ( 1.38 ) & 1.67 ( 1.67 ) & iib + ngc 6334a & 1.7 & 350 & 20 & 170 & & 0.84 & 86 & -86 & 28 & 26 & 2.89 ( 7.27 ) & 0.06 ( 0.002 ) & 0.39 ( 0.59 ) & 0.54 ( 1.31 ) & iia + g34.3@xmath241 & 3.7 & 350 & 20 & 360 & & 0.87 & 89 & -85 & 35 & 24 & 0.98 ( 17.8 ) & 0.01 ( 0.007 ) & 0.30 ( 0.82 ) & 0.22 ( 3.06 ) & iib + ngc 6334i & 1.7 & 350 & 20 & 170 & & 0.63 & 88 & -84 & 38 & 25 & 9.20 ( 34.9 ) & 0.01 ( 0.01 ) & 1.08 ( 1.71 ) & 1.46 ( 4.87 ) & iia + m@xmath242 & 8 & 350 & 20 & 780 & & 0.81 & 71 & -83 & 58 & 26 & 5.06 ( 5.06 ) & 0.39 ( 0.39 ) & 1.60 ( 1.60 ) & 1.56 ( 1.56 ) & iib + w33 c & 2.4 & 350 & 20 & 230 & & 0.74 & 89 & -75 & 40 & 27 & 1.09 ( 7.48 ) & 0.02 ( 0.02 ) & 0.31 ( 0.65 ) & 0.30 ( 1.59 ) & - + w33 a & 2.4 & 350 & 20 & 230 & & 0.87 & 80 & -75 & 36 & 22 & 1.17 ( 1.17 ) & 0.01 ( 0.003 ) & 0.28 ( 0.22 ) & 0.35 ( 0.33 ) & iia + m 17 & 1.6 & 350 & 20 & 160 & & 0.44 & 88 & -90 & 34 & 26 & 7.18 ( 149 ) & 0.01 ( 0.01 ) & 0.75 ( 2.81 ) & 1.12 ( 15.9 ) & iia + w3 & 1.95 & 350 & 20 & 190 & & 0.91 & 78 & -81 & 24 & 23 & 1.72 ( 3.01 ) & 0.01 ( 0.002 ) & 0.37 ( 0.48 ) & 0.42 ( 0.67 ) & iia + omc 1 & 0.414 & 350 & 20 & 40 & & 0.59 & 86 & -90 & 29 & 20 & 11.2 ( 74.3 ) & 0.01 ( 0.001 ) & 0.74 ( 1.08 ) & 1.09 ( 4.50 ) & iia + omc 3 & 0.414 & 350 & 20 & 40 & & 0.59 & 59 & -89 & 25 & 20 & 1.57 ( 31.7 ) & 0.02 ( 0.02 ) & 0.68 ( 1.72 ) & 0.42 ( 5.68 ) & iia + sgr a@xmath237 east & 8 & 350 & 20 & 780 & & 0.83 & 90 & -90 & 47 & 30 & 14.1 ( 138 ) & 0.04 ( 0.04 ) & 2.08 ( 9.48 ) & 3.09 ( 30.0 ) & i/- + iras 05327 & 0.414 & 350 & 20 & 40 & & 0.43 & 81 & -88 & 54 & 28 & 2.77 ( 2.77 ) & 0.11 ( 0.11 ) & 0.86 ( 0.86 ) & 0.92 ( 0.92 ) & - + sgr b2 & 8 & 350 & 20 & 780 & & 0.70 & 90 & -87 & 52 & 21 & 11.5 ( 63.0 ) & 0.02 ( 0.001 ) & 1.07 ( 1.47 ) & 1.46 ( 5.22 ) & iib + sgr b1 & 8 & 350 & 20 & 780 & & 0.76 & 90 & -90 & 48 & 28 & 36.3 ( 204 ) & 0.06 ( 0.06 ) & 1.64 ( 8.33 ) & 4.94 ( 35.5 ) & i / iia + dr21 & 3 & 350 & 20 & 290 & & 0.58 & 86 & -87 & 26 & 22 & 1.74 ( 3.61 ) & 0.01 ( 0.001 ) & 0.52 ( 0.50 ) & 0.37 ( 0.52 ) & iia + mon ob1 iras 12 & 0.80 & 350 & 20 & 78 & & 0.65 & 88 & -88 & 42 & 29 & 5.56 ( 17.2 ) & 0.01 ( 0.01 ) & 0.78 ( 1.61 ) & 1.17 ( 3.74 ) & iib + m@xmath243 & 8 & 350 & 20 & 780 & & 0.71 & 90 & -80 & 47 & 29 & 3.32 ( 123 ) & 0.07 ( 0.006 ) & 1.04 ( 6.73 ) & 0.98 ( 26.6 ) & iib + m@xmath244 & 8 & 350 & 20 & 780 & & 0.89 & 38 & -37 & 19 & 11 & 1.02 ( 1.02 ) & 0.01 ( 0.01 ) & 0.37 ( 0.37 ) & 0.38 ( 0.38 ) & iia + i & irregular & irregular & irregular + iia & elongated & field mostly aligned with source minor axis & systematic : overall small , larger towards and along central major axis + iib & elongated & field mostly aligned with source major axis & systematic : overall large , smaller towards and along central major axis + iii & more circular & field increasingly radial and / or hourglass - like & systematic : overall small , additional azimuthal symmetries in @xmath245 + | submillimeter dust polarization measurements of a sample of 50 star - forming regions , observed with the submillimeter array ( sma ) and the caltech submillimeter observatory ( cso ) covering pc - scale clouds to mpc - scale cores , are analyzed in order to quantify the magnetic field importance .
the magnetic field misalignment @xmath0 the local angle between magnetic field and dust emission gradient is found to be a prime observable , revealing distinct distributions for sources where the magnetic field is preferentially aligned with or perpendicular to the source minor axis . source - averaged misalignment angles @xmath1 fall into systematically different ranges , reflecting the different source - magnetic field configurations .
possible bimodal @xmath1-distributions are found for the separate sma and cso samples . combining both samples
broadens the distribution with a wide maximum peak at small @xmath1-values . assuming the 50 sources to be representative , the prevailing source - magnetic field configuration is one that statistically prefers small magnetic field misalignments @xmath2 .
when interpreting @xmath2 together with a magnetohydrodynamics force equation , as developed in the framework of the polarization - intensity gradient method , a sample - based log - linear scaling fits the magnetic field tension - to - gravity force ratio @xmath3 versus @xmath1 with @xmath4 ( mean error ) , providing a way to estimate the relative importance of the magnetic field , only based on measurable field misalignments @xmath2 .
the force ratio @xmath5 discriminates systems that are collapsible on average ( @xmath6 ) from other molecular clouds where the magnetic field still provides enough resistance against gravitational collapse ( @xmath7 ) .
the sample - wide trend shows a transition around @xmath8 .
defining an effective gravitational force @xmath9 , the average magnetic - field - reduced star - formation efficiency is at least a factor of two smaller than the free - fall efficiency . for about one forth of the sources
the average efficiency drops to zero .
the force ratio @xmath5 can further be linked to the normalized mass - to - flux ratio , yielding an estimate for the latter one without the need of field strength measurements . across the sample , a transition from magnetically supercritical to subcritcal is observed with growing misalignment @xmath1 . |
random walks are widely used in physics so as to model the features of transport processes where the migrating ( possibly massless ) particle undergoes a series of random displacements as the effect of repeated collisions with the surrounding environment @xcite .
while much attention has been given to random walks on regular euclidean lattices , and to the corresponding scaling limits , less has been comparatively devoted to the case where the direction of propagation can change continuously at each collision : for an historical review , see , e.g. , @xcite .
such processes , which are intimately connected to the boltzmann equation , have been named _ random flights _ , and play a prominent role in the description of , among others , neutron or photon propagation through matter @xcite , chemical and biological species migration @xcite , or electron motion in semiconductors @xcite . within the simplest formulation of this model , which was originally proposed by pearson ( 1905 ) @xcite and later extended by kluyver ( 1906 ) @xcite and rayleigh ( 1919 ) @xcite , it is assumed that particles perform random displacements ( ` flights ' ) along straight lines , and that at the end of each flight ( a ` collision ' with the surrounding medium ) the direction of propagation changes at random .
when the number of transported particles is much smaller than the number of the particles of the interacting medium , so that inter - particles collisions can be safely neglected , it is reasonable to assume that the probability of interacting with the medium is poissonian . for the case of neutrons in a nuclear reactor ,
e.g. , the ratio between the number of transported particles and the number of interacting nuclei in a typical fuel@xmath2moderator configuration is of the order of @xmath3 , even for high flux reactors @xcite .
it follows that flight lengths between subsequent collisions are exponentially distributed ( hence we will call this process _ exponential flights _ in the following ) .
we assume that collisions can be either of scattering or absorption type . at each scattering collision , the flight direction changes at random , whereas at absorption events the particle disappears and the flight terminates .
each flight can be seen as a random walk in the phase space of position @xmath4 and direction @xmath5 in a @xmath0-dimensional setup .
the particle density @xmath6 represents the probability density of finding a transported particle at position @xmath4 having direction @xmath5 at a time @xmath7 , up to an appropriate normalization factor . in many applications ,
the actual physical observable is the average of the density @xmath6 over the directions @xmath5 , namely @xmath8 where @xmath9 is a normalization factor corresponding to the surface of the unit @xmath0-dimensional sphere and @xmath10 is the gamma function @xcite .
along with the development of monte carlo methods , numerical solutions @xmath11 to complex three - dimensional linear / nonlinear transport problems coming from applied sciences are becoming accessible to an high degree of accuracy : criticality calculations in reactor cores @xcite , scattering and absorption in heated plasmas @xcite , propagation through anisotropic scattering centers in atmosphere or fluids @xcite , and charge transport in semiconductors under external fields @xcite , only to name a few .
nonetheless , even for the simplest systems , many theoretical questions remain without an answer , so that the study of exponential flights has attracted a renovated interest in recent years ; see , e.g. , @xcite . in particular , it has been emphasized that the dimension @xmath0 deeply affects the nature of @xmath11 , and prevents in most cases from obtaining explicit results . the aim of our work is to investigate exponential flights in a generic @xmath0-dimensional setup , under simplifying hypotheses . here
, we will mostly focus on establishing insightful relationships between space , time and the statistics of particle collisions within a given volume .
a number of new results will be derived , concerning unbounded , bounded , scattering as well as absorbing domains .
this paper is structured as follows . in sec .
[ setup ] we recall the mathematical formalism , introduce the physical variables and derive their inter - dependence for any @xmath0 . in sec .
[ space_moments ] we detail the structure of the spatial moments of the particle ensemble . sec .
[ coll_density_sec ] is devoted to the analysis of the collisions statistics in a given domain .
then , in sec . [ dimensional_analysis ] we examine the distinct cases @xmath12 and @xmath13 .
both the spatial and temporal evolution of the particle ensemble are considered , and results for bounded domains are obtained by resorting to the method of images .
we provide a comparison between analytical ( or asymptotic ) findings and monte carlo simulations .
conclusions will be finally drawn in sec .
[ conclusions ] .
within the natural framework of statistical mechanics , the evolution of the particle density @xmath6 for exponential flights is governed by the linear boltzmann equation @xcite .
linearity stems from neglecting inter - particle collisions . in the hypothesis that an average particle energy can be defined ( the so called one - speed transport ) , and that the physical properties of the medium do not depend on position nor time , the boltzmann equation for the density @xmath6 reads @xcite @xmath14 where @xmath15 is total cross section of the traversed medium ( carrying the units of an inverse length ) , @xmath16 is the scattering cross section , @xmath17 is the particle speed , and @xmath18 is the source .
the total cross section @xmath15 is such that @xmath19 represents the average flight length between consecutive collisions ( the so - called mean free path ) , and is related to the scattering cross section @xmath16 and to the absorption cross section @xmath20 by @xmath21 .
the quantity @xmath22 is the scattering kernel , i.e. , the probability density that at each scattering collision the random direction changes from @xmath23 to @xmath5 .
= 9.0 cm denoting by @xmath6 the solution of the boltzmann equation for a medium without absorptions ( @xmath24 ) , the complete solution with absorption @xmath25 can be easily obtained by letting @xmath26 thanks to linearity @xcite .
this allows primarily addressing a purely scattering medium ( @xmath27 ) , without loss of generality . at long times , i.e.
, far from the source , the direction - averaged particle density @xmath11 is known to converge to a gaussian shape , namely , @xmath28 the quantity @xmath29 playing the role of a diffusion coefficient @xcite . however , eq . is approximately valid for @xmath30 , and can not capture the particle evolution at early times , nor the finite - speed propagation effects . indeed , diffusion implicitly assumes a non - vanishing probability of finding the particles at arbitrary distance from the source .
deviations from the limit gaussian behavior are well known , e.g. , for neutron @xcite as well electron transport @xcite . in the following ,
we outline the relation between @xmath11 and the underlying exponential flight process .
consider a @xmath0-dimensional setup .
a particle , originally located at position @xmath31 in a given domain , travels along straight lines at constant speed @xmath17 , until it collides with the medium .
the position of a particle at the @xmath32-th collision can be expressed as a random walk @xmath33 , i.e. , a sum of random variables @xmath34 .
the flight lengths @xmath35 are assumedly identically distributed and obey an exponential probability density @xmath36 with @xmath37 .
the exponential law in eq . stems from assuming a uniform distribution of the scattering centers in the traversed medium .
heterogeneous materials , such as complex fluids , would generally lead to clustered scattering centers , obeying to , e.g. , negative binomial distributions , and in turn to non - exponential flight lengths @xcite .
however , we will focus our attention on homogeneous media . at each collision , the particle randomly changes its direction according to the scattering kernel @xmath22 . for the sake of simplicity , we assume here that the scattering is isotropic , so that @xmath22 has a uniform distribution , independent of the incident direction @xmath23 .
once a flight length has been sampled from @xmath38 , the new direction @xmath5 is therefore uniformly distributed on the @xmath0-sphere of surface @xmath39 .
therefore , by virtue of the apparent spherical symmetry , the transition kernel , i.e. , the probability density of performing a displacement from @xmath40 to @xmath34 , depends only on @xmath35 , and reads @xmath41 we initially neglect absorptions , so that @xmath42 : in one - speed transport , this condition can be either seen as the particles been scattered , or equivalently being absorbed and then re - emitted ( with the same speed ) at each collision .
this latter interpretation would correspond , e.g. , to a _ criticality _
condition in multiplicative systems for neutron transport .
we define then the free propagator @xmath43 as the probability density of finding a particle at position @xmath4 at the @xmath32-th collision , for an infinite medium , i.e. , in absence of boundaries .
we adopt here the convention that the particle position and direction refer to the physical properties before entering the collision ; for instance , the index @xmath44 refers to uncollided particles , i.e. , particles coming from the source and entering their first collision . assuming that all the particles are isotropically emitted at @xmath45 , the particle density @xmath43 must depend only on the variable @xmath46 , because of the spherical symmetry . on the basis of the properties exposed above
, the particle propagation as a function of the number of collisions is a markovian process in the variable @xmath47 , where for each collision @xmath48 the new propagator is given by the convolution integral @xmath49 with initial condition @xmath50 .
in particular , by direct integration we immediately get the uncollided propagator @xmath51 it is convenient to introduce the @xmath0-dimensional fourier transform of spherical - symmetrical functions , as in the subsequent analysis this will allow easier deriving of the properties of the exponential flights . denoting by @xmath52 the transformed variable with respect to @xmath53 , for any spherical - symmetrical function @xmath54 we have the following transform and anti - transform pair @xmath55 and @xmath56 @xcite @xmath57 where @xmath58 is the modified bessel function of the first kind , with index @xmath59 @xcite . it is apparent from eqs .
that the dimension @xmath0 of the system plays a fundamental role , in that it affects both the transition kernel and the fourier transform kernel itself .
= 9.0 cm the convolution integral in eq . in fourier space
gives the algebraic relation @xmath60 @xmath61 , with initial condition @xmath62 . by recursion
, it follows that in the transformed space @xmath63 it turns out that the fourier transform of the transition kernel @xmath64 can be explicitly performed in arbitrary dimension , and gives @xmath65 where @xmath66 is the gauss hypergeometric function @xcite .
hence , we finally obtain @xmath67^n .
\label{p_z_n}\ ] ]
the quantity @xmath64 is positive for @xmath68 ; moreover , @xmath69 , which ensures normalization and positivity of the propagator . in fig .
[ fig1 ] we visually represent the effects of dimension and number of collisions on the shape of @xmath70 .
remark in particular that the spread of @xmath70 increases with @xmath0 , for a given @xmath32 . on the contrary ,
@xmath70 becomes more peaked around the origin with growing @xmath32 , for a given @xmath0 .
formally , performing the inverse fourier transform of eq .
gives the propagator @xmath71 for an arbitrary @xmath0-dimensional setup .
actually , in some cases this task turns out to be non - trivial .
nonetheless , even in absence of an explicit functional form for the propagator , information can be extracted by resorting to the tauberian theorems . in particular , the expansion of @xmath70 for @xmath72
gives the behavior of @xmath73 for @xmath30 , i.e. , far from the source , in the so - called _ diffusion limit _
@xcite ; viceversa , the expansion of @xmath70 for @xmath74 gives the behavior of @xmath73 for @xmath75 , i.e. , close to the source .
we recall that @xmath66 is defined through the series @xcite @xmath76 at the leading order for @xmath77 we therefore have @xmath78 observe that eq .
can be viewed as the expansion of an exponential function .
then , the inverse fourier transform gives the gaussian shape @xmath79 which is valid for @xmath80 , @xmath81 playing the role of a diffusion coefficient .
this stems from the exponential flights having finite - variance increments , @xmath82 , so that their probability density @xmath73 falls in the basin of attraction of the central limit theorem @xcite .
we mention that clustered scattering centers , with non - exponential flight lengths , may lead to non - gaussian limiting statistics @xcite .
remark the close analogy between eqs . and :
in particular , @xmath83 and @xmath84 differ by a factor @xmath85 , which roughly speaking represents the average number of collisions per unit time .
moreover , at the leading order for @xmath86 we have the expansion @xmath87 where the first term vanishes for @xmath88 . by inverse fourier
transforming , we have for @xmath75 @xmath89 when @xmath90 , and @xmath91 when @xmath88 . here
@xmath92 and @xmath93 are constants depending on @xmath32 and @xmath0 .
it can be shown that the divergence at the origin in @xmath73 due to the dirac delta source disappears after @xmath94 collisions for @xmath90 , and @xmath95 for @xmath88 .
= 9.0 cm assume again that @xmath42 , i.e. , that there are no absorptions .
the free propagator @xmath73 gives information on the position of a transported particle at the moment of entering the @xmath32-th scattering collision .
the link between the travelled distance , the flight time and the number of collisions is provided by the speed @xmath17 .
indeed , once a flight of length @xmath96 between any two collisions has been sampled from @xmath38 , the flight time must satisfy @xmath97 .
hence , the transition kernel , i.e. , the probability density of performing a displacement from @xmath40 at @xmath98 to @xmath34 at @xmath99 , will be given by @xmath100 it follows that inter - collision times are exponentially distributed @xmath101 where @xmath102 represents the average time between collisions .
we define the propagator @xmath103 as the probability density of finding a particle at position @xmath104 at time @xmath7 at the @xmath32-th collision . from the markov property of the process , at each collision @xmath48 we have @xmath105 with initial condition @xmath106 .
in particular , by direct integration we immediately get the uncollided propagator @xmath107 = 9.0 cm we denote the laplace transform of a function @xmath108 by its argument , i.e. , @xmath109 then , from the double convolution integral in eq .
we have the algebraic product in the fourier and laplace space @xmath110 @xmath61 , with @xmath111 . by recursion
, it follows that in the transformed space we have @xmath112 it turns out that the fourier and laplace transform of the transition kernel @xmath113 can be explicitly performed in arbitrary dimension , and gives @xmath114 where @xmath115 .
hence , we finally obtain @xmath116 ^n . \label{propagator_f_l}\ ] ] moreover , the following relation follows : @xmath117 , so that @xmath118 i.e. , @xmath73 can be intepreted as the time average of @xmath103 .
finally , the propagator @xmath119 will be given by the sum of the contributions @xmath103 at each collision , namely @xmath120 taking the fourier and laplace transform of eq .
, we then have @xmath121 with @xmath122 . in presence of absorptions ( @xmath123 )
, the propagator @xmath124 can be obtained from eq . by integrating over directions .
this relation holds true at each collision , so that we have @xmath125 where @xmath126 .
hence , from eq . , by replacing @xmath16 with @xmath15 we get for the propagator @xmath127 @xmath128 where @xmath129 represents the average flight time between any two collisions .
now , observe that from eq .
we have @xmath130 where the quantity @xmath131 , @xmath132 , can be interpreted as the the probability of being scattered , i.e. , _ not _ being absorbed , at any given collision .
then , by remarking that @xmath133 , it follows that @xmath134 the propagator in eq . is then given by the product of the free propagator , with the total cross section @xmath15 replacing the scattering cross section @xmath16 , times the probability of having survived up to entering the @xmath32-th collision .
remark that the total cross section and the non - absorption probability are related by @xmath135 . when the absorption length is infinite , @xmath136 , @xmath137 and we recover the free propagator for pure scattering , with @xmath27 .
so far , we have assumed that the medium where particles are transported has an infinite extension , hence the name free propagator .
more realistically , we might consider finite - extension media with volume @xmath138 enclosing the source , so that boundary conditions come into play and affect the nature of the propagator .
several boundary conditions can be conceived according to the specific physical system , among which the most common are reflecting and leakage .
this issue has been extensively examined , e.g. , for radiation shielding in reactor physics @xcite and for electron motion in semiconductors @xcite . here
we will focus on leakage boundary conditions , where particles are considered lost as soon as their trajectory has traversed the outer boundary @xmath139 of the domain .
while the volume @xmath138 is in principle totally arbitrary , in the subsequent calculations for the sake of convenience we will assume that @xmath140 is a sphere of radius @xmath141 centered in @xmath142 .
from the point of view of the propagator , leakages can be taken into account by assuming that the population density @xmath73 at any @xmath32 vanishes at the so - called extrapolation length @xmath143 , i.e. , @xmath144 @xcite .
because trajectory do not terminate at the boundary , but rather at the first collision occurred outside the volume , the extrapolation length is expected to be larger than the physical boundary of @xmath145 and can be determined from solving the so - called milne s problem associated to the volume @xcite . in general , @xmath143 is of the kind @xmath146 $ ] , the dimensionless constant @xmath147 depending on the dimension of the system @xcite .
when the scattering length is much smaller than the typical size of the volume , i.e. , @xmath148 , the extrapolation length coincides with the physical boundary , @xmath149 .
this means that the inter - collision length is so small as compared to the total travelled distance that the first collision outside the domain will actually be very close to the last collision inside the domain , which corresponds to @xmath73 vanishing at @xmath141 @xcite .
the moments of the propagator provide an estimate of the spatial evolution of the particles ensemble , as a function of the number of collisions or time . due to the supposed spherical symmetry
, we expect all the odd moments to vanish .
we define the @xmath150-th moment of @xmath54 over the spherical shell @xmath151 as @xmath152 performing the integral in eq . would require explicitly knowing @xmath54 .
however , as shown in the appendix [ appendix a ] , the @xmath150-th spatial moment @xmath54 of can be expressed as a function of the @xmath150-th derivative of @xmath153 with respect to @xmath52 @xmath154_{z=0 } \label{spatial_moments}\ ] ] when @xmath150 is even , and @xmath155 otherwise . by setting @xmath156 in eq .
we then have @xmath157 as a function of the number of collisions .
analogously , by setting @xmath158 we get @xmath159 , which gives the evolution as a function of time , upon inverse laplace transforming . in particular , for the spread @xmath160 of the propagator with absorptions we get @xmath161 and @xmath162 e^{-t/\tau_a}.\ ] ] in absence of absorption , @xmath163 and @xmath137 , the particle spread @xmath164 is linear with respect to @xmath32 . on the contrary
, @xmath165 has a ballistic behavior ( @xmath166 ) at earlier times ( where transport is dominated by velocity ) , and a diffusive behavior ( @xmath167 ) at later times ( where transport is dominated by scattering ) .
the transition between the two regimes is imposed by the time scale @xmath168 .
a remarkable feature is that in either case the spread does not explicitly depend on the dimension @xmath0 .
intuitively , this can be understood by considering that ( independently of the dimension @xmath0 of the embedding setup ) the vectors @xmath169 and @xmath170 define a plane with random orientation , so that space is explored by plane surfaces at each collision .
this is in analogy , e.g. , with the behavior of @xmath0-dimensional brownian motion @xcite .
in many physical problems , one is interested in assessing the statistics of the time @xmath171 or the collision number @xmath172 spent inside a given domain @xmath138 .
a prominent role in characterizing a physical system is played in particular by the mean residence time @xmath173 and the mean collision number @xmath174 @xcite . in reactor physics , for instance , the average number of neutron collisions within a region would be related to such issues as the nuclear heating or nuclear damage in fissile as well as structural materials @xcite .
we introduce the collision density @xmath175 @xcite , which is defined as @xmath176 from eq .
, it follows that @xmath175 can be equivalently obtained from the propagator @xmath119 as @xmath177 for an infinite ` observation time ' @xmath178 , the mean residence time @xmath173 within @xmath138 can be expressed in terms of the collision density @xmath175 in the same domain by slightly adapting an argument due to kac @xcite @xmath179 where @xmath180 . as for the collision number
, the probability of performing @xmath172 collisions in @xmath138 is given by @xmath181 hence the moments @xmath182 from the definition of @xmath175 , it follows immediately that @xmath183 i.e. , the integral of the collision density over a volume @xmath138 gives the mean number of collisions within that domain , hence the name given to @xmath175 .
remark that both @xmath173 and @xmath174 depend on the boundary conditions imposed on @xmath139 , which in turn affect the functional form of the propagator , and then @xmath175 . using
a free propagator corresponds to defining a fictitious volume @xmath138 , whose boundaries @xmath139 are ` transparent ' , so that particles can indefinitely cross @xmath139 back an forth . on the contrary , the use of leakage boundary conditions leads to the formulation of first - passage problems , i.e. , determining the distribution of the time , or collision number , required to first reach the boundary .
consequently , @xmath173 is preferentially called mean first - passage time @xcite .
it can be shown that higher order moments of @xmath172 can be obtained by recursion , in terms of kac integrals analogous to those in eq . .
a detailed treatment is beyond the scope of the present paper and is discussed in @xcite .
the previous discussion shows that @xmath175 is key in determining residence times and collisions statistics .
after formally carrying out the summation in eq . , we have @xmath184 $ ] , so that @xmath175 for a free propagator in absence of absorption is defined in terms of the following fourier integral @xmath185 whose convergence depends on the dimension @xmath0 of the system .
it turns out that convergence is ensured for @xmath186 , which means that for @xmath187 and @xmath1 finite - size domains with transparent boundaries @xmath174 and @xmath173 diverge as @xmath188 .
this result is in analogy with plya s theorem , which states that random walks on euclidean lattices are recurrent for @xmath189 @xcite .
as shown in the following , we can nonetheless provide an estimate of such divergence as a function of @xmath190 , i.e. , single out a singular term from a functional form .
for finite domains with leakage boundary conditions and/or absorptions ( @xmath123 ) , @xmath175 is defined also for @xmath187 and @xmath1 systems . for @xmath186 ,
tauberian theorems show that the asymptotic behavior of eq . is given by @xmath191 for large @xmath53 , and @xmath192 close to the source .
in the following , we detail the results pertaining to specific values of @xmath0 . we choose @xmath19 as length scale and we work with dimensionless spatial variables @xmath193 . remark that in absence of absorption the length scale is @xmath194 , since @xmath195 .
the case @xmath88 allows illustrating the general structure of the calculations .
one potential application of this framework could be provided by nanowires or carbon nanotubes ( almost @xmath187 systems ) in electron transport @xcite .
the transition kernel is @xmath196 whose fourier transform is @xmath197 starting from @xmath198 , the free propagator @xmath73 can be explicitly obtained by performing the inverse fourier transform of @xmath199 , and reads @xmath200 where @xmath201 is the modified bessel function of the second kind , with index @xmath59 @xcite .
the same formula has been recently derived , e.g. , in @xcite as a particular case of a broader class of random flights . in fig .
[ figpn_d1 ] we provide a comparison between monte carlo simulation results ( symbols ) , the analytical formula eq .
( solid lines ) , and the diffusion limit , eq .
( dashed lines ) , for different values of @xmath32 .
remark in particular that the diffusion limit is not accurate for small @xmath32 , and becomes progressively closer to the exact result for increasing @xmath32 , as expected . at intermediate @xmath32 ,
the tails of the propagator are always fatter than those predicted by the diffusion approximation . in a @xmath187 setup ,
the collision density @xmath175 for the free propagator diverges .
nonetheless , it is possible to single out the divergence as follows @xmath202 for fixed @xmath190 , the inverse transform can be explicitly performed in terms of hypergeometric functions . retaining the non - vanishing terms for large @xmath190 , we have @xmath203 which is composed of a term diverging with @xmath204 ( not depending on @xmath53 ) , and a functional part which is linear in @xmath53 ( not depending on @xmath190 ) . for the propagator with absorptions , from eq .
we have @xmath205 then , performing the inverse fourier transform , we get @xmath206 remark that eq
. has been derived , e.g. , in @xcite by solving the stationary boltzmann equation in @xmath187 .
when @xmath207 , the particles are almost surely absorbed at the first collision , and we have the expansion @xmath208 so that at the first order the collision density has the same functional form as the uncollided propagator . at the opposite , when @xmath137 the particles are almost surely always scattered ( @xmath209 ) , and we have the expansion @xmath210 and @xmath211 diverges as the collision density associated to the free propagator , as expected . the case of leakage boundary conditions can be dealt with by imposing that the propagator @xmath73 must vanish for any @xmath32 at the extrapolated boundary @xmath143 . for @xmath88 ,
the extrapolated length is given by @xmath212 $ ] with @xmath213 @xcite , i.e. , the propagator must vanish at one scattering length outside the physical border @xmath141 of the domain . by resorting to the method of images @xcite , which allows solving for the propagator in presence of boundaries in terms of the propagator in absence of boundaries
, we therefore have for the collision density @xmath214 for @xmath215 , and @xmath216 elsewhere .
the moments of the residence time ( or , equivalently , of the number of collisions ) within a sphere of radius @xmath141 can be explicitly computed based on eq . .
the moments associated to the free propagator clearly diverge . here
we separately analyze the propagator with absorptions ( in an infinite domain ) and the propagator with leakages at the boundary @xmath217 ( without absorption ) . for the case of absorptions ,
the average residence time within a ( fictitious ) sphere of radius @xmath141 reads @xmath218 assuming that particles can cross the boundaries of the sphere an infinite number of times .
when the radius of the sphere is large as compared to the typical particle displacement , we have @xmath219 , which gives @xmath220 when @xmath207 , and diverges for @xmath137 . for the case of leakages
, the mean first - passage time reads @xmath221 when the radius of the sphere is large as compared to the typical particle displacement , @xmath149 , and we have @xmath222 .
the case @xmath223 has a key interest in assessing , e.g. , the dynamics of chemical and biological species on surfaces @xcite .
moreover , it concerns also quantum wells and inversion layers in electron transport @xcite .
the transition kernel reads @xmath224 whose fourier transform is @xmath225 starting from @xmath198 , the free propagator @xmath73 can be explicitly obtained by performing the inverse fourier transform of @xmath226 , and reads @xmath227 this result was previously established by @xcite , and has later appeared in , e.g. , @xcite . in fig . [ figpn_d2 ] we compare the monte carlo simulation results ( symbols ) with the theoretical formula in eq . , for different values of @xmath32 .
the diffusion limit , eq . , is also shown in dashed lines .
remark that the diffusion limit is not accurate for small @xmath32 , and becomes progressively closer to the exact result for increasing @xmath32 . at intermediate @xmath32 ,
the tails of the propagator are always fatter than those predicted by the diffusion approximation . in a @xmath1 setup
, the collision density @xmath175 diverges .
nonetheless , analogously as done for the @xmath187 case , it is possible to single out the divergence as follows @xmath228 for fixed @xmath190 , the inverse transform can be explicitly performed .
details of the rather cumbersome calculations are left to the appendix [ appendix b ] . retaining the non - vanishing terms for large @xmath190
, we have @xmath229 where @xmath230 is the exponential integral function @xcite . to the authors best knowledge , the formula for the collision density @xmath175 has not appeared before , and might then provide a useful tool for describing the migration of species on @xmath1 environments .
similarly as in the @xmath187 case , @xmath175 is composed of a term diverging with @xmath231 ( not depending on @xmath53 ) , and a functional part in @xmath53 ( not depending on @xmath190 ) . in deriving eq .
we have neglected a constant term which is small compared to @xmath232 , namely , @xmath233/\pi$ ] , where @xmath234 is the euler s gamma constant @xcite .
the collision density with leakages at @xmath235 can be obtained again by the method of images , whence @xmath236 where @xmath237 $ ] and the milne s constant @xmath238 @xcite .
the moments associated to the free propagator clearly diverge . for the propagator with leakages at
the boundary @xmath217 the mean first - passage time within a sphere of radius @xmath141 reads @xmath239 when @xmath240 , we have @xmath241 , and we get @xmath242 in the diffusion limit . as for the collision density with absorptions ,
calculations analogous to the @xmath187 case lead to @xmath243 we could not find an explicit expression for the latter integral in terms of elementary functions . however , the limits for small and large scattering probability can be easily obtained , and read @xmath244 when @xmath207 , and @xmath245 when @xmath137 , respectively .
the former expression gives the uncollided propagator at the leading order , whereas the latter diverges logarithmically as @xmath137 . the case @xmath246 plays a prominent role , among others , in reactor physics , in that it concerns the transport of neutrons and photons through matter @xcite , and is key in describing electron transport in bulk semiconductors @xcite .
on the basis of the strikingly similar form of eqs . and , it would be tempting to postulate an analogous expression for the propagator in @xmath247 .
for @xmath88 we have indeed the functional form @xmath248 , and for @xmath223 @xmath249 .
then we could conjecture an exponent @xmath250 , so that @xmath251 by imposing normalization .
unfortunately , this is not the case , and @xmath252 is not the true @xmath247 propagator .
actually , few explicit results can be derived , and much of the analysis is therefore devoted to the asymptotic behavior .
the transition kernel reads @xmath253 whose fourier transform is @xmath254 the propagator @xmath73 with initial condition @xmath198 involves then the following integral @xmath255^n dz,\ ] ] which can not be carried out explicitly for arbitrary @xmath32 . in the diffusion limit @xmath256
we have @xmath257^n \simeq 1- nz^2/3 $ ] , so that @xmath258 as expected from eq . .
[ figpn_d3 ] we compare the monte carlo simulation results ( symbols ) with the diffusion limit , eq .
( dashed lines ) , and with the approximate propagator , eq .
( solid lines ) .
remark that eq .
provides a fairly accurate approximation of the simulation results , except close to the source . after carrying out the sum over @xmath32 , the collision density @xmath175
is given by the following integral @xmath259 which again can not be performed explicitly @xcite .
as before , we consider then the asymptotic behavior . denoting @xmath260
$ ] , we have @xmath261 in the diffusion limit @xmath262 , and @xmath263 close to the source .
similar expansions appear , e.g. , in @xcite , as derived from the analysis of the boltzmann equation .
correspondingly , by performing the respective integrations we have @xmath264 for @xmath265 , i.e. , far from the source , and @xmath266 + ... \label{close_source_3d}\ ] ] for @xmath267 , i.e. , close to the source .
remark that for @xmath246 @xmath175 does not diverge even for infinite domains without absorptions . in fig .
[ figpsi_3d ] we compare the monte carlo simulation results ( symbols ) with the asymptotic limits close to and far from the source , eqs . and , respectively ( dashed lines ) .
the simulation results progressively approach the asymptotic limits as the number @xmath190 of summed collisions increases .
eqs . and can provide asymptotic estimates for the collision density with leakages at the boundary @xmath268 . by the method of images
, we have that the collision density with boundaries is @xmath269 , with @xmath270 $ ] and @xmath271 @xcite .
the moments of the residence time within a sphere of radius @xmath141 can be explicitly computed based on eq .
for the free propagator @xmath73 , i.e. , when particles can freely cross the surface of the sphere .
we have @xmath272 when @xmath273 , and @xmath274 for @xmath275 .
moreover , for leakage boundary conditions at the surface , from @xmath276 we have @xmath277 when @xmath273 , and @xmath278 for @xmath275 . the case @xmath279 is briefly presented here for the sake of completeness .
the transition kernel reads @xmath280 whose fourier transform is @xmath281 we could not find an explicit representation for the inverse fourier transform of @xmath282 . nonetheless , the propagator @xmath103 is known and reads @xmath283^{n-2}\ ] ] for @xmath284 @xcite .
hence , it follows that the propagator @xmath285 can be obtained from solving the integral @xmath286 this integral can be performed , and gives @xmath287 , \label{propagator_4d}\ ] ] where @xmath288 @xmath289 being an hypergeometric function @xcite . as for the collision density @xmath175
, we have @xmath290 which can be computed explicitly and gives @xmath291 finally , the collision density in presence of leakages at the boundary @xmath268 can be obtained by resorting to the method of images , whence @xmath269 , with @xmath292 $ ] .
the constant @xmath293 has been estimated by running a monte carlo simulation and determining the extrapolation length , and reads @xmath294 .
the moments of the residence time within a sphere of radius @xmath141 can be explicitly computed based on eq . for the free propagator @xmath73 , namely , @xmath295 moreover , for leakage boundary conditions at the surface , from @xmath276 we have @xmath296
in this paper , we have examined the dynamics of exponential flights and their relation with the linear boltzmann equation , a subject that arises in many areas of physics or biology . in particular ,
we have focused on @xmath297 the propagator @xmath73 , which describes the ensemble evolution of the transported particles as a function of the number of collisions , and @xmath298 the collision density @xmath175 , which is related to the particle equilibrium distribution . moreover ,
the connection between the number of collisions and time has been examined .
we have provided the framework for a generic @xmath0-dimensional setup , which allows emphasizing the key role of @xmath0 in determining the properties of @xmath73 and @xmath175 . in this context , we have shown that knowledge of @xmath175 formally allows deriving the moments of the residence time ( or equivalently of the collision number ) within a given volume , which is key in assessing many physical properties of the system at study .
the role of boundary conditions has been explored by considering leakages from a given domain , via the method of images .
the behavior of the spatial moments of the particle ensemble has been examined , as well .
we have then provided specific results for one - speed isotropic transport in infinite as well as bounded domains , and for absorbing or purely scattering media . the case
@xmath88 has been considered as a prototype model of exponential flights along a straight line , where only two directions ( forward or backward ) are possible . due to this simplification
, most quantities can be explicitly derived .
the case @xmath223 has been analyzed in detail : despite the calculations being non trivial , in some cases closed - form results can be obtained . in particular , we have provided an expression for the collision density , which , coupled with the method of images , might be useful for a realistic description of migration on bounded surfaces . the case @xmath246 is key in most real - world applications , such as the propagation of neutrons or photons in matter , or electrons in bulk semiconductors .
unfortunately , this case turns out to be hardly amenable to closed - form analytical formulas , and most results concern the asymptotic behavior of the particles , either close to or far from the source .
finally , the case @xmath279 has been considered for the sake of completeness .
moreover , monte carlo simulations have been performed so as to validate the proposed results and support the analysis of the asymptotic behavior .
a good agreement is found between theoretical predictions and numerical simulations . by virtue of the increasing power of monte
carlo methods in solving realistic three - dimensional transport problems , one might argue that finding closed - form results for simple systems has a limited interest .
however , we are persuaded that analytical and asymptotic formulas may turn out to be useful , in that they can help in improving monte carlo algorithms by a clever use of the relation existing between different variables . moreover , as exponential flights are a transversal field , the cross - over between distinct areas of science might hopefully shed some light at achieving long - standing goals in transport theory , such as full analytical solutions for @xmath247 systems .
the authors wish to thank drs .
f. malvagi and a. rosso for careful reading of the manuscript and useful discussions .
we begin by computing the @xmath150-th coefficient of the taylor expansion of @xmath299 with respect to @xmath52 , which reads @xmath300_{z=0}=\frac{i^m \left ( \frac{r}{2}\right)^{\frac{d}{2}+m-1 } } { \gamma\left ( 1+\frac{m}{2}\right ) \gamma\left(\frac{d+m}{2 } \right ) } , \ ] ] for even @xmath150 , and zero otherwise .
apply now the @xmath150-th derivative to a function @xmath153 such that @xmath54 has a spherical symmetry . recalling then the definition of the moment @xmath301 from eq .
, we have @xmath302_{z=0}=\frac{i^m 2^{1-\frac{d}{2}-m } ( 2\pi)^{\frac{d}{2}}}{\gamma\left ( 1+\frac{m}{2}\right ) \gamma\left(\frac{d+m}{2 } \right ) } \frac{\langle r^m \rangle}{\omega_d}.\ ] ] rearranging the coefficients , we can finally express the spatial moments of @xmath54 in terms of the @xmath150-th @xmath52-derivative of @xmath153 , namely , @xmath154_{z=0}.\ ] ]
in order to find the collision density associated to the free propagator , we begin by decomposing the sum over the collisions @xmath32 into even and odd index , i.e. , @xmath303.\ ] ] then , by remarking that for even @xmath32 @xmath304 we get @xmath305 for large @xmath190 .
we have neglected a constant term of the kind @xmath233/(2\pi)$ ] , which is small compared to @xmath232 . for odd @xmath32 , we can use the series representation @xmath306 now , carrying out the double sum over odd @xmath32 and over @xmath307 , from eq .
we get @xmath308 for large @xmath190 . again , to obtain this result we have neglected a constant term of the kind @xmath233/(2\pi)$ ] , which is small compared to @xmath232 . 10 b. d. hughes , _ random walks and random environments _ vol .
i ( clarendon press , oxford , 1995 ) .
g. h. weiss , _ aspects and applications of the random walk _ ( north holland press , amsterdam , 1994 ) .
w. feller , _ an introduction to probability theory and its applications _
, 3rd edition ( wiley , new york , 1970 ) .
r. metzler and j. klafter , phys . rep . *
339 * , 1 ( 2000 ) .
j. dutka , arch .
exact sci . * 32 * , 351 ( 1985 ) .
g. i. bell and s. glasstone , _ nuclear reactor theory _
( van nostrand reinhold , 1970 ) . m. weinberg and e. p. wigner , _ the physical theory of neutron chain reactors _ ( university of chicago press , 1958 ) . c. cercignani , _ the boltzmann equation and its applications _
( springer , 1988 ) .
h. t. hillen and g. othmer , siam j. appl .
math * 61 * , 751 ( 2000 ) . c. jacoboni and p. lugli , _ the monte carlo method for semiconductor device simulation _
( springer , 1989 ) .
k. pearson , nature * 27 * , 294 ( 1905 ) .
j. c. kluyver , proc .
koninklijke akademie van wetenschappen te amsterdam * 8 * , 341 ( 1906 ) .
j. w. strutt ( lord rayleigh ) , philos . magazine
* 6 * ( 37 ) , 321 ( 1919 ) .
m. abramowitz and i. a. stegun ( eds . ) , _ handbook of mathematical functions with formulas , graphs , and mathematical tables _ , 9th edition ( dover , ny , 1972 ) . i. lux and l. koblinger , _ monte carlo particle transport methods : neutron and photon calculations _ ( crc press , boca raton , 1991 ) .
j. a. wesson , _ tokamaks _ , 3@xmath310 ed .
( clarendon press , 2003 ) .
e. jakeman and r. j. a. tough , j. opt .
am . a * 4 * , 1764 ( 1987 ) .
t. kurosawa , proceedings of the 3@xmath310 international conference on hot carriers in semiconductors ( 1964 ) . c. jacoboni and l. reggiani , rev .
* 55 * , 645 ( 1983 ) .
j. price , semicond .
* 14 * , 249 ( 1979 ) .
j. c. j. paasschens , phys .
e * 56 * , 1135 ( 1997 ) .
e. orsingher and a. de gregorio , j. theor . probab . *
20 * , 769 ( 2007 ) .
a. d. kolesnik , j. stat . phys . * 131 * , 1039 ( 2008 ) .
g. le car , j. stat . phys .
* 140 * , 728 ( 2010 ) .
m. v. fischetti and s. e. laux , phys .
b * 38 * , 9721 ( 1988 ) . m. v. fischetti and s. e. laux , phys .
b * 48 * , 2244 ( 1993 ) .
g. placzek and w. seidel , phys . rev . *
72 * , 550 ( 1947 ) .
i. freund , phys .
a * 45 * , 8854 ( 1992 ) .
s. redner , _ a guide to first - passage processes _ ( cambridge university press , cambridge , 2001 )
. k. m. case and p. f. zweifel , _ linear transport theory _
( addison - wesley , reading , 1967 ) .
m. kac , _ probability and related topics in physical sciences _ ( lectures in applied mathematics , wiley , 1957 ) .
a. m. berezhkovskii , v. zaloj , and n. agmon , phys . rev .
e * 57 * , 3937 ( 1998 ) .
a. zoia , e. dumonteil , and a. mazzolo , arxiv:1102.5291v1 .
g. milton wing , _ an introduction to transport theory _
( wiley , ny , 1962 ) .
w. stadje , j. stat . phys . * 46 * , 207 ( 1987 ) .
b. conolly and d. roberts , eur
res . * 28 * , 308 ( 1987 ) . | in this paper we analyze some aspects of _ exponential flights _ , a stochastic process that governs the evolution of many random transport phenomena , such as neutron propagation , chemical / biological species migration , or electron motion .
we introduce a general framework for @xmath0-dimensional setups , and emphasize that exponential flights represent a deceivingly simple system , where in most cases closed - form formulas can hardly be obtained .
we derive a number of novel exact ( where possible ) or asymptotic results , among which the stationary probability density for @xmath1 systems , a long - standing issue in physics , and the mean residence time in a given volume . bounded or unbounded , as well as scattering or absorbing domains are examined , and monte carlo simulations are performed so as to support our findings . |
very recently , we have considered a generalized thermostatistics based on a multifractal phase space , whose dimension @xmath8 is not equal the number @xmath9 of the degrees of freedom fixed by the number of particles @xmath10 @xcite . in such a case ,
the behaviour of the complex system is determined by the dimensionless volume @xmath11 of the supporting phase space where @xmath12 is dirac - planck constant . according to self - similarity condition ,
the specific statistical weight of such a system is given by the power law function @xcite @xmath13 where @xmath14 is multifractal index , @xmath15 is reduced fractal dimension .
this function should be multiplied by the specific number of monofractals with dimension @xmath2 @xmath16 which are contained in multifractal whose spectrum is determined by a function @xmath1 . as a result , the whole statistical weight , being the multifractal measure , takes the form @xmath17 where @xmath18 is a density distribution over dimensions @xmath2 . using the method of steepest descent , we arrive at the power law @xmath19 which generalizes the simplest relation ( [ 1 ] ) due to replacement of the bare fractal dimension @xmath2 related the index @xmath20 by the multifractal function @xmath21 here
, the fractal dimension @xmath22 relates to given parameter @xmath14 to be defined by the conditions of application of the steepest descent method : @xmath23 inserting the statistical weight ( [ 4 ] ) into the deformed entropy @xcite @xmath24 related to the whole statistical weight @xmath25 arrives at thermostatistical scheme governed by the tsallis formalism of the non - extensive statistics @xcite .
within such a scheme , the non - extensivity parameter is determined by the multifractal index ( [ 5 ] ) which monotonically increases , taking value @xmath26 at @xmath20 and @xmath27 at @xmath28 .
it is appeared , the multifractal function @xmath0 is reduced to the specific heat to determine , together with the inverted value @xmath29 , both statistical distributions and thermodynamic functions of a complex system @xcite . in this way , the entropy ( [ 7 ] ) will be positive definite , continuous , symmetric , expandable , decisive , maximal , concave , lesche stable and fulfilling a generalized pesin - like identity if the exponent ( [ 5 ] ) varies within the interval @xmath30 $ ] @xcite . in this letter , we are aimed to model possible analytical forms of both multifractal index @xmath0 and related spectrum @xmath1 . in section 2
we show that the monotonically increasing function @xmath0 is presented by the hyperbolic tangent deformed in accordance with both tsallis and kaniadakis procedures whose using allows one to describe explicitly arbitrary multifractal phase space .
section 3 is devoted to consideration of the multifractal spectrum @xmath1 which determines the number of monofractals within the multifractal with the specific dimension @xmath2 .
section 4 concludes our consideration .
as the simplest case , we take the function @xmath0 in the form @xmath31,\label{8}\ ] ] being determined by parameter @xmath32 and argument @xmath33 . according to fig.[dq(q)_d ] ,
related multifractal dimension function @xcite @xmath34 ) ( curves 1 , 2 , 3 correspond to @xmath35).,width=302 ] monotonically decreases from maximum value @xmath36 to @xmath37 with @xmath14 increase .
however , maximum value of the fractal dimension @xmath38 is fixed by the magnitude @xmath39 , so that one should put @xmath40 in the dependence ( [ 8 ] ) . as a result , it takes quite trivial form .
as the @xmath0 function increases monotonically within narrow interval @xmath30 $ ] , one has a scanty choice of its modeling .
one of such possibilities is opened with using deformation of the hyperbolic tangent ( [ 8 ] ) at @xmath41 . at present ,
two analytical procedures are extensively popularized .
the first of these is based on the tsallis exponential @xcite @xmath42 where deformation parameter @xmath43 takes positive values .
the second procedure has been proposed by kaniadakis @xcite to determine the deformed exponential @xmath44 with using these definitions , the deformed tangent ( [ 8 ] ) takes the form @xmath45 where the multifractal index @xmath14 varies within the domain @xmath46 .
the @xmath14-dependencies of the multifractal index @xmath0 and its inverted value @xmath47 are shown in fig.[tau(q ) ] at different magnitudes -dependencies of the multifractal index @xmath48 and its inverted value @xmath49 ( solid line corresponds to @xmath50 , dashed curves relate to the tsallis deformation with @xmath51 ; dotted lines correspond to the kaniadakis one at @xmath52).,width=302 ] of both tsallis and kaniadakis deformation parameters @xmath43 .
it is principally important , the first of these deformations arrives at more fast variations of both indices @xmath0 and @xmath53 in comparison with non - deformed hyperbolic tangent @xmath54 , whereas the kaniadakis deformation slows down these variations with @xmath43 increase .
characteristic peculiarity of the tsallis deformation consists in breaking dependencies @xmath0 , @xmath53 in the point @xmath55 where the second terms in both numerator and denominator of the definition ( [ 12 ] ) take the zero value . as a result
, the multifractal index ( [ 12 ] ) obtains the asymptotics @xmath56 for @xmath57 the dependence @xmath58 takes the simplest form : @xmath59 at @xmath60 , and @xmath61 at @xmath62 .
it is worthwhile to notice , the tsallis deformation parameter can not take values @xmath63 because these relate to fractal dimensions @xmath64 at @xmath65 . at the kaniadakis deformation ,
the multifractal index @xmath66 varies smoothly to be characterized by the following asymptotics : @xmath67^{-2/\kappa}\quad\quad\quad { \rm at}\quad\quad \kappa(q-1)\gg 1 .
\end{array } \right .
\label{14}\end{aligned}\ ] ] in contrast to the tsallis case , here the deformation parameter can take arbitrary values to give the simplest dependence @xmath68 at @xmath57 . as
a function of the index @xmath14 , the fractal dimension ( [ 9 ] ) falls down monotonically as shown in fig.[d_q]a . according to eqs.([13 ] ) , at the ) .,title="fig:",width=226 ] ) .,title="fig:",width=226 ] tsallis deformation , one has broken dependence @xmath69 , being characterized by the asymptotics @xmath70 in the limit case @xmath57 , the phase space is smooth @xmath71 within the interval @xmath72 . for the kaniadakis deformation , the fractal dimension @xmath73
is given by smoothly falling down curve whose slope increase with the deformation parameter growth .
according to eqs.([14 ] ) , in this case , one has the asymptotics @xmath74 at @xmath57 , the typical dependence reads @xmath75^{-1/2}$ ] .
at given multifractal index @xmath0 , the spectrum function @xmath1 is defined by inverted legendre transformation ( [ 5 ] ) where the specific dimension reads @xmath76 as shows fig.[d_q]b , the dependence @xmath77 has monotonically falling down form to take the value @xmath78 at @xmath79 for the tsallis deformation . in this case , asymptotical behaviour is characterized by eqs.([13 ] ) , according to which one obtains @xmath80 in the limit @xmath81 , the dependence @xmath82 takes the step - like form being @xmath83 within the interval @xmath60 and @xmath78 otherwise . for the kaniadakis deformation , eqs.([14 ] ) arrive at the asymptotics
@xmath84^{-(2+\kappa)/\kappa}\quad\quad { \rm at}\quad\quad \kappa(q-1)\gg 1 .
\end{array } \right .
\label{19}\end{aligned}\ ] ] the typical behaviour is presented by the dependence @xmath85^{-3/2}$ ] related to @xmath57 .
the multifractal spectrum is defined by the equality @xmath86 being conjugated eq.([5 ] ) . here , the specific multifractal index @xmath87 is determined by eq.([17 ] ) which arrives at the limit relations ( [ 18 ] ) , ( [ 19 ] ) . with its using ,
one obtains the asymptotics @xmath88+(1 + 1/\kappa)d\qquad { \rm at}\quad\ d\ll 1 \end{array } \right .
\label{20}\end{aligned}\ ] ] for the tsallis deformation , and the relations @xmath89+d\quad\ { \rm at}\quad d\ll 1 , \end{array } \right .
\label{21}\end{aligned}\ ] ] characterizing the kaniadakis deformation .
as shows fig.[f(d ) ] , for finite deformation parameters @xmath90 , , dashed curves relate to the tsallis deformation with @xmath51 ; dotted lines correspond to the kaniadakis one at @xmath91).,width=302 ] a spectrum function increases monotonically , taking the minimum value @xmath3 at @xmath4 and maximum @xmath5 at @xmath6 . by this , the derivative @xmath92 equals @xmath93 on the left boundary and @xmath94 on the right one .
it is characteristically , the whole set of the spectrum functions is bounded by the limit dependencies @xmath95 and @xmath96 , the first of which relates to limit magnitude of the tsallis deformation parameter @xmath57 whereas the second corresponds to the kaniadakis limit @xmath97 .
typical form of the spectrum function is presented by the dependencies @xmath98 it may seem , at first glance , that negative values of the spectrum function @xmath1 has not a physical meaning . to clear up this question ,
let us take the set of monofractals with the reduced dimension @xmath4 . obviously , such monofractals relate to the whole set of the phase space points , whose number equals the dimensionless volume @xmath99 .
just such result gives the definition ( [ 2 ] ) in the point @xmath4 where @xmath3 . on the other hand , in opposite case @xmath6 where @xmath5
, we obtain the very thing the number of monofractals with volume @xmath99 equals @xmath100 to give one multifractal of the same volume @xmath99 . in this way , a single monofractal
is contained in the multifractal at condition @xmath101 which takes place at the reduced dimension @xmath102 whose dependence on the deformation parameter @xmath43 is shown in fig.[d0(k ) ] . related to the condition @xmath101 as function of the parameter @xmath43 ( curve 1 corresponds to the tsallis deformation , curve 2 to the kaniadakis one ; positive values @xmath1 relate to the domain @xmath103).,width=264 ] the dependence of the number @xmath104 of monofractals containing in the phase space volume @xmath99 related to the multifractal with the specific dimension @xmath2 is shown in fig.[n(k ) ] .
it is seen , the number of monofractals within the phase space volume @xmath99 related to the specific dimension @xmath2 at the deformation parameters @xmath50 ( a ) and @xmath105 ( b ) ( different levels @xmath106 are shown with thin curves , thick lines relate to @xmath107).,title="fig:",width=226 ] of monofractals within the phase space volume @xmath99 related to the specific dimension @xmath2 at the deformation parameters @xmath50 ( a ) and @xmath105 ( b ) ( different levels @xmath106 are shown with thin curves , thick lines relate to @xmath107).,title="fig:",width=226 ] @xmath104 increases with the @xmath99 volume growth at small dimensions @xmath2 , whereas in the limit @xmath7 the dependence @xmath108 becomes falling down to give infinitely increasing numbers @xmath104 at @xmath109 .
the speed of such increase growths monotonically with both decrease of the tsallis deformation parameter @xmath43 and increase of the kaniadakis one .
the whole set of monofractals within a multifractal related to the phase space , which gives the support of a generalized thermostatistics , is modeled by the exponent @xmath0 that determines the statistical weight ( [ 4 ] ) at given volume @xmath99 . to be the entropy ( [ 7 ] ) concave , lesche stable et cetera
, the exponent @xmath0 should be a function , monotonically increasing within the interval @xmath30 $ ] at multifractal exponent variation within the domain @xmath46 .
the simplest case of such a function gives the hyperbolic tangent @xmath110 whose deformation ( [ 12 ] ) defined in accordance with both tsallis and kaniadakis exponentials ( [ 10 ] ) , ( [ 11 ] ) allows one to describe explicitly arbitrary multifractal phase space . in this way
, the tsallis deformation arrives at more fast variations of the statistical weight index @xmath0 in comparison with non - deformed hyperbolic tangent , whereas the kaniadakis one slows down these variations with increasing the deformation parameter @xmath43 .
all possible dependencies @xmath0 are bounded from above by the linear function @xmath111 at @xmath112 $ ] which is transformed into the constant @xmath113 at @xmath62 .
this dependence relates to the smooth phase space within the tsallis interval @xmath112 $ ] .
the dependence ( [ 2 ] ) of the number of monofractals within the phase space volume @xmath99 related to the multifractal with the specific dimension @xmath2 is determined by the spectrum function @xmath1 .
this function increases monotonically , taking the minimum value @xmath3 at @xmath4 and maximum @xmath5 at @xmath6 ; by this , its derivative equals @xmath93 on the left boundary and @xmath94 on the right one .
the whole set of the spectrum functions is bounded by the limit dependencies @xmath95 and @xmath96 , the first of which relates to limit magnitude of the tsallis deformation parameter @xmath57 and the second one corresponds to the kaniadakis limit @xmath97 .
the number of monofractals within the multifractal increases with the @xmath99 volume growth at small dimensions @xmath2 and falls down in the limits @xmath7 to give infinitely increasing at @xmath114 . | we consider the set of monofractals within a multifractal related to the phase space being the support of a generalized thermostatistics .
the statistical weight exponent @xmath0 is shown to can be modeled by the hyperbolic tangent deformed in accordance with both tsallis and kaniadakis exponentials whose using allows one to describe explicitly arbitrary multifractal phase space .
the spectrum function @xmath1 , determining the specific number of monofractals with reduced dimension @xmath2 , is proved to increases monotonically from minimum value @xmath3 at @xmath4 to maximum @xmath5 at @xmath6 .
the number of monofractals is shown to increase with growth of the phase space volume at small dimensions @xmath2 and falls down in the limit @xmath7 .
, multifractal ; exponent of statistical weight ; fractal dimension spectrum ; deformed exponential . 05.45.df . |
deviations of the triple gauge boson couplings ( tgcs ) from their values predicted by the standard model ( sm ) are a possible indication for new physics ( np ) beyond the sm .
if no light higgs boson exists the mechanism responsible for the restoring the unitarity could well be the strong electroweak symmetry breaking ( sewsb ) mechanism @xcite . as a consequence , at energies below np cut - off scale @xmath5 the effects of np
are reflected in the tgc s values leading to their deviations @xmath6 and @xmath7 from the sm predictions .
since these deviations decrease as @xmath8 increases , their observation requires a very precise measurements , more precise than those at lep and tevatron . with a high event statistics at a @xmath3 collider option at the international linear collider ( ilc ) it is possible to reach a high precision of the tgc measurements .
anomalous tgcs affect both the total production cross - section and the shape of the differential cross - section as a function of the w production angle . as a consequence , distributions of w decay products
are changed also .
thus , the information about tgcs can be extracted from the angular distributions of the reconstructed w boson . in @xmath3 collisions
the tgcs contribute through _ t_-channel w - exchange .
in this study the expected sensitivity for a measurement of the couplings @xmath1 and @xmath2 in @xmath9jets at @xmath10 ( @xmath11 ) is investigated .
there are two possible initial @xmath3 helicity states depending on the photon handedness , denoted as @xmath12 ( if two photons have the same helicities ) and @xmath13 ( if two photons have the opposite helicities ) .
total and differential cross - sections distributions as a function of the anomalous tgcs ( @xmath14 ) , simulated with the tree - level monte carlo ( mc ) generator whizard @xcite , for all possible initial and final state helicity combinations are shown in figures [ fig : total_gg ] and [ fig : differ_gg ] .
as a beam simulation circe2 @xcite is used to describe realistic beam spectra for @xmath3-colliders .
the response of a detector has been simulated with simdet v4 @xcite , a parametric monte carlo for the tesla @xmath15-detector .
it includes a tracking and calorimeter simulation and a reconstruction of energy - flow - objects ( efo ) .
only the efos with a polar angle above @xmath16 are taken for the w boson reconstruction , simulating the acceptance of the photon collider detector as the only difference to the @xmath15-detector @xcite .
the signal and background events are studied on a sample of events generated with whizard and overlayed with low energy @xmath17 events ( _ pileup _ ) @xcite .
the corresponding number of added pileup events per bunch crossing is 1.8 @xcite .
the informations about the neutral particles ( _ neutrals _ ) from calorimeter and charged tracks ( _ tracks _ ) from tracking detector are used to reconstruct the signal and background events . the potential background for both initial @xmath18 states
are @xmath19 events that can mimic the signal with four jets when gluons are radiated in the final state .
the qcd corrections to the @xmath20-pair born level production cross - section are different for the two @xmath18 states : in the @xmath13 state the corrected cross - section is @xmath21 with @xmath22 being of @xmath23(1 ) , resulting in a born cross - section correction of 4 - 5@xmath24 . in this study
this correction is not taken into account . in the @xmath12 state
, the suppression factor @xmath25 @xcite leads to a born level cross - section close to zero but the qcd corrections lead to an enhancement by double - logarithmic terms @xmath26 @xcite . to estimate the corrected cross - section for the @xmath12 state the @xmath27 diagrams are taken into account i.e. the diagrams contributing to @xmath28 and @xmath29 .
the @xmath30 cut parameter ( @xmath31 ; @xmath32 ) for a variable centre - of - mass energy @xmath33 is defined by generating only events with the invariant masses of each parton pair above 30 gev resulting in an emission of hard gluons .
the signal events for both @xmath18 states and background events for the @xmath13 state are generated with omega matrix element generator @xcite taking into account only the lowest order feynman diagrams .
the qcd correction for the @xmath20 pair production in the @xmath12 state is estimated generating the background events with madgraph @xcite . in order to minimise the pileup contribution to the high energy signal tracks ,
the information on the track impact parameters is used in the same way as in the case of @xmath34-collisions @xcite allowing the rejection of @xmath35 of pileup tracks and @xmath36 of signal tracks .
the remaining tracks are combined into four jets and the events with a number of efo greater than 40 and number of charged tracks greater than 20 are accepted only .
the two reconstructed w bosons are denoted as forward ( @xmath37 , @xmath38 ) and backward ( @xmath39 , @xmath40 ) where @xmath41 is a w boson production angle in the centre - of - mass system ( cms ) .
the angle between the two jets belonging to the same w boson , boosted to the cms , is used as a next selection criteria - if the angle is within a given range of @xmath42 , the event is accepted .
further , events with a total mass above 125 gev and the individual w boson mass of @xmath43 are accepted .
that results in efficiencies of approximately 53% for signal and less than 2% for background events i.e. in a purity of 81% in both @xmath18 states .
the top pair production is estimated to be negligible .
the final angular distributions for the @xmath13 state state . ] used for the tgcs error estimation are shown in figure [ fig : final ] .
for the extraction of the tgss from the reconstructed kinematical variables ( fig .
[ fig : final ] ) a binned likelihood fit is used .
a sample of @xmath44 sm signal events is generated with whizard and passed trough the detector simulation . each event is described reconstructing five kinematical variables - the w production angle with respect to the @xmath45 beam direction @xmath41 , the w s polar decay angles @xmath46 ( angle of the fermion with respect to the w flight direction measured in the w rest frame ) and the azimuthal decay angles @xmath47 of the fermion with respect to a plane defined by w and the beam axis . in hadronic w
- decays the up- and down - type quarks can not be separated so that only @xmath48 is measured .
the matrix element calculations from whizard are used to obtain weights @xcite to reweight the angular distributions as functions of the anomalous tgcs where @xmath49 and @xmath50 are the free parameters .
six - dimensional ( 6d ) event distributions over @xmath51 , @xmath52 , @xmath47 and centre - of - mass energy are fitted with minuit @xcite , minimising the likelihood function depending on @xmath53 and @xmath54 : @xmath55 + \frac{(n-1)^{2}}{2(\delta l^{2})},\end{aligned}\ ] ] where _ i , j , k , l _ and _ m _ run over the reconstructed angular distributions @xmath56 and @xmath47 , _ p _ runs over the reconstructed centre - of - mass energy , @xmath57 is the `` data '' which corresponds to the sm mc sample , @xmath58@xmath59 , @xmath60 , @xmath22 , @xmath61 , @xmath62 , @xmath63 ( mc sample ) is the event distribution weighted by the function @xmath64,@xmath65 and @xmath66@xmath59 , @xmath60 , @xmath22 , @xmath61 , @xmath62 , @xmath67 . the factor @xmath68 sets the number of signal events to the expected one after one year of running of an @xmath3-collider . in case where the background is included in the fit
@xmath68 defines the sum of signal and background events and @xmath69 $ ] .
the number of background events is normalised to the effective w boson production cross - section in order to obtain the corresponding number of background events after one year of running of an @xmath3-collider for corresponding @xmath18 state .
it is assumed that the total normalisation ( efficiency , luminosity , electron polarisation ) is only known with a relative uncertainty @xmath70 .
thus , @xmath71 is taken as a free parameter in the fit and constrained to unity with the assumed normalisation uncertainty . per construction
the fit is bias - free and thus returns always exactly the sm as central values . in the @xmath13 state @xmath72
is a realistic precision that can be achieved while for the @xmath12 due to the small number of events where the cross - section is @xmath73 suppressed . ]
, the luminosity is expected to be measured with an error of @xmath74 .
table [ tab : t1 ] shows the estimated statistical errors we expect for the different couplings at @xmath10 for a two - parameter 6d fit at detector level including the pileup and background events in both @xmath18 states .
.estimated statistical errors for @xmath75 and @xmath50 from the 6d fit at detector level for both @xmath18 states in @xmath3 collisions at @xmath76 gev , without pileup , with pileup and with background events . [
cols="<,^,^,^,^,^,^,^,^,^ " , ] concerning the systematic errors the influence of the background and the degree of photon polarisation have been investigated , assuming @xmath77 in the @xmath13 state and @xmath78 in the @xmath12 state . in the @xmath12 state ,
the polarisation uncertainty of @xmath79 for @xmath53 is to less than the statistical error while in the @xmath13 state , the polarisation uncertainty of @xmath80 for @xmath53 is less than three times the statistical error . the uncertainty on @xmath54 in both @xmath18 states is found to be negligible .
in the @xmath13 state the background cross - section should be known to better than 0.8% for @xmath53 and to better than 4% for @xmath54 if the corresponding systematic uncertainty should no be larger than the statistical error .
for @xmath12 the requirement is 0.6% for @xmath54 while there are basically no restrictions for @xmath53 .
the estimated sensitivity of the tgcs measurement in both @xmath3 initial states at @xmath76 gev with integrated luminosities of @xmath81 fb@xmath82 is of order @xmath83 for @xmath6 and higher than @xmath84 for @xmath7 in the @xmath13 state assuming @xmath85 .
the state @xmath12 takes into account a larger error on the luminosity measurement of @xmath86 resulting in a sensitivity to @xmath1 higher than @xmath87 and to @xmath2 higher than @xmath88 .
while @xmath53 can be measured somewhat better in @xmath89 , the @xmath3-collider provides a higher accuracy for a @xmath54 measurement compared to the @xmath89- and @xmath34-colliders .
9 m. chanowitz , m. golden and h. georgi , phys .
d36 ( 1987 ) 1490 ; m.j.g .
veltman and f.j .
ynddurain , nucl .
phys . b325 ( 1989 ) 1 .
w.kilian , `` whizard 1.24 a generic monte carlo integration and event generation package for multi - particle processes '' , lc - tool 2001 - 039 ( revised ) ( 2001 ) . | if there is no the standard model higgs boson , the interaction among the gauge bosons becomes strong at high energies ( @xmath0 ) .
the effects of strong electroweak symmetry breaking could manifest themselves indirectly through the vertices as anomalous gauge boson couplings before they give rise to new physical states like resonances . here
a study of the measurement of trilinear gauge couplings @xmath1 and @xmath2 is presented looking at the hadronic decay channel of the ww boson pair at an @xmath3 - collider .
a sensitivity of @xmath4 can be reached depending on the coupling under consideration and on the initial polarisation state . |
meson physics is a very broad field , and this workshop covers a very large part of it , with about 25 plenary and 50 parallel talks .
i can not possible attempt to summarize all that .
so , i will pick and choose from what was presented , a process which is admittedly very subjective , and add some new and interesting results which were not talked about .
i will also update some results which have appeared in print since my talk .
before i go into physics , let me acknowledge the laboratories and experiments which have contributed to this field , and to this workshop , and briefly mention what is , to the best of my knowledge , their current status . in the following , ` ended ' means that data taking has ended , but physics analysis and publications continue .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ bepc ( bes ii ended , bes
iii operational in 2007 ? ) ; bnl ( e852 ended , phoenix and star at rhic have minor spectroscopy programs ) ; celsius ( wasa ended ) ; cern ( crystal barrel and obelix ended ; compass has potential ) ; cosy ( anke , gem , tof operational ) ; desy ( h1 , zeus , hermes operational ) , fermilab ( selex and focus ended , tevatron operational until 2009 ) , frascati ( at da@xmath0ne ; kloe and dear ended , finuda and siddharta operational ) ; gsi ( hades operational , panda in 2014 ? ) ; jlab ( clas operational , 12 gev upgrade ? ) ; jparc ( operational in 2008 ? ) , kek ( ps and belle ( high potential ) operational ; bonn ( elsa operational ) ; mainz ( mami ii operational ) ; slac ( babar ( high potential ) operational until 2010 ) . _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the sum total of the above summary is that while of the many very productive older programs , some have ended , many continue , and new programs ( bepc ii , gsi ( fair ) , and jparc ) are in the offing .
we can therefore have high expectations for continued exciting discoveries in meson physics in the future .
i will begin my summary in the order of the mass hierarchy .
light quark ( @xmath1 ) mesons , heavy quark ( @xmath2 ) mesons , and then light / heavy ( @xmath3 ) mesons . for good measure
, i will throw in a gluon or two , and say a few words about glueballs .
finally , i will talk about the recently discovered exotics , about which everybody is excited , but whose nature nobody really knows .
pions and kaons are ancient particles , and one would think that there could not possible by anything new to learn about them . actually , there is ! one of the most interesting and elementary questions one can ask about a particle is `` what does it look like ? '' to put it in terms of the simplest physical observable , the question becomes , `` what do the particle s electromagnetic form factors look like ? '' to relate the meson s form factors to its constituent quarks one needs to invoke perturbative qcd ( at least at present ) , and that means that one needs experimental measurements of form factors at large enough momentum transfers at which pqcd might be expected to be valid . let us see where we stand both experimentally and theoretically on this issue .
are shown with square symbols.,title="fig:",width=230 ] are shown with square symbols.,title="fig:",width=230 ] let me recall the great debate about `` when is @xmath4 large enough to validate pqcd ? '' which took place in the mid-1980s between brodsky and colleagues@xcite , who believed @xmath5 was enough , and isgur and llwellyn smith@xcite , who claimed that @xmath6 would be required . the experimental data for pion form factors used in the debate consisted of barely three data points for @xmath7 with @xmath8 errors .
we would all agree that perhaps this was `` much ado about nothing '' .
so , what is new ? what is new is the first precision measurements at large timelike momentum transfer ( @xmath9 ) of pion and kaon form factors at cleo@xcite as shown in fig .
1 point marked `` @xmath10 '' is not a direct measurement , but dervied from the measured branching ratio for @xmath11 by milana et al.@xcite . using the same arguments ,
i have derived @xmath12@xcite .
the agreement of these @xmath13 at @xmath14 with the measured values at @xmath15 lends justification to the procedure . ] .
the solid curves in the figure show the arbitrarily normalized variation of @xmath16 with @xmath4 .
also shown for the pion are pqcd predictions and qcd sum rule predictions , neither of which come even close to the measured values , which says that here is now something real for theorists to chew on .
most of the activity in this category has been confined to the pseudoscalar @xmath17 and @xmath18 mesons .
the obvious reason is that accelerators in the @xmath19 gev region can produce them copiously . at this conference
, a lot of activity was reported on production of @xmath17 and @xmath18 .
detailed reports were presented on photo and electro production measurements being made at mainz , jlab , and elsa .
we also heard about pion - induced production at bnl , production in @xmath20 collisions at frascati , in @xmath21 collisions at julich , and in heavy ion collisions at gsi .
we have an interesting development in the measurements of the mass of @xmath17 .
a recent cosy measurement@xcite reported @xmath22 , which differs by @xmath23 from an earlier na38 measurment@xcite .
@xmath24 mev .
now we have the report from kloe@xcite of @xmath25 , which agrees with na38 within @xmath26 .
the resolution of the @xmath27 kev difference in @xmath28 between cosy and na38/kloe would normally be only of academic interest , but in this era of meson meson molecules , it could be important if somebody comes up with a weakly bound molecule of @xmath29 !
the mass of @xmath18 is known with a @xmath30 kev uncertainty@xcite , but so far we have no reports from cosy or kloe on improved measurements .
kloe@xcite has also reported the long sought value of the ratio @xmath31 , and used it to address the question of the possible glue content of @xmath18 .
they find it to be @xmath32 . while the numerical value is not significant at this level ,
it does indicate that the glue content of @xmath18 is quite small .
the spectroscopy of the low mass vectors , the @xmath33 has been dominated by the precision measurements of the novosibirsk snd and cmd2 groups .
measurements of comparable precision are now being reported by kloe .
hopefully , long standing problems , like the @xmath34 mev uncertainty in the @xmath35 mass can be resolved by these new data , although it is my belief that the problem actually lies with the different formalisms for resonance analysis being used .
the p wave singlets , @xmath36 and @xmath37 , have remained inactive for the last fifteen years .
the focus of the activity has been on the p wave triplets , and among them the greatest attention has been drawn by the isosinglet scalars , primarily because @xmath38 are also the quantum numbers of the lowest mass glueball .
it behooves us , however , to take a look at all three , @xmath39 , @xmath40 , @xmath41 . the problem common to all the states with masses larger than @xmath42 gev is excessive crowding . in the 1 gev interval of mass , @xmath43 gev , according to godfrey and isgur@xcite , 75 ( @xmath44 ) mesons are expected , the average spacing being @xmath45 mev .
the ones which have been observed have widths ranging from @xmath46 mev , which means that states overlap , mix , and interfere horribly , and even the most sophisticated partial width analyses of experimental data have a tough time identifying and characterizing ( i.e. @xmath47 , branching ratios ) the states .
the problem is best illustrated by the isospin singlet scalars , the @xmath48 s . with @xmath44
quarks , one expects two isoscalars , @xmath49 and @xmath50 , and an isovector @xmath51 .
the trouble is that there are _ five _
@xmath49 s known : @xmath52 or @xmath53 , @xmath54 , @xmath55 , @xmath56 , and @xmath57 .
it has become very important to examine these very carefully because one expects only one more isosinglet @xmath58 state beyond @xmath49 and @xmath50 , and that third state is to be identified with the interesting , and ever elusive , glueball .
but we have five !
so , any new experimental information on these is most welcome .
the sigma has been there for a long time , albeit in the shadows .
its most recent observation is by bes in @xmath60 , with the result @xmath61 , @xmath62@xcite .
leutwyler and colleagues@xcite interpret the @xmath53 as a pole in the @xmath63 scattering amplitude and obtain @xmath64 , @xmath65 , and consider its physics to be governed by the goldstone pions , i.e. , the sigma is not a simple @xmath59 meson .
this is consistent with the fact that typical potential model calculations ( e.g. godfrey and isgur@xcite ) predict the lowest mass @xmath49 to have a mass of @xmath66 gev .
lattice has so far nothing to say about the @xmath53 ; it is still struggling with 250 mev pions .
the fact is that the nature of the @xmath53 remains as cloudy as ever , althought pennington@xcite claims that its two photon width precludes its being a tetraquark ( @xmath67 ) or a glueball .
parenthetically , let me mention that bes@xcite has confirmed the observation of sigma s strange partner , the * kappa * , previously reported by fermilab e791@xcite in the reaction @xmath68 with @xmath69 and @xmath70 . in the reaction @xmath71 , bes obtains @xmath72 and @xmath73
. it will be interesting to see if the @xmath74 also emerges from the analysis of @xmath75 scattering length , a la leutwyler .
we had five scalars , and we wanted to weed out two in order to have left two @xmath76 and one glueball .
well , we appear to have put @xmath52 or @xmath53 aside as a pole in the @xmath63 scattering amplitude .
we need to remove one more .
the axe usually falls on @xmath54 . because @xmath77 mev , and @xmath54 strongly couples to @xmath78 , @xmath54 , and its cousin @xmath79 ,
are often dismissed as @xmath78 molecules .
this conjecture , or , for example , the recent conjecture@xcite that like the @xmath53 , @xmath54 is a dynamically generated resonance , make @xmath54 non@xmath76 .
however , there is no hard proof , theoretical or experimental , for this assertion . as a matter of fact , as described later , i believe that the systematics of @xmath80 states argues against this prejudice .
despite the fact that @xmath54 is one of the oldest known ( since 1973 ) resonances , and has been observed in numerous experiments , very little quantitative is known about it .
an average of the latest measurements by e791@xcite , belle@xcite , bes@xcite , and kloe@xcite gives @xmath81 mev , and the average of the e791@xcite and belle@xcite results gives @xmath82 .
besides this , we know very little .
pdg06 summarizes the decays of @xmath54 to @xmath83 , @xmath84 , and @xmath85 as _ seen _ , _ seen _ , _ seen _ , which can be seen as a very unsatisfactory situation .
the only quantitative measure we have is that according to bes@xcite , the partial width for @xmath86 is about @xmath87 that for @xmath88 , and when the relative phase space differences are taken into account , the coupling to kaons is about 3 to 4 times larger than that to pions@xcite . with @xmath52 and
@xmath54 banished as non@xmath76 , the scalar field is reduced to the required three , two @xmath76 and one @xmath89 glueball .
most of the glueball games are played between these three . despite early over optimistic searches for ` pure ' glueballs , it is now generally accepted that ` pure ' glueballs do not exist .
quantum mechanics dictates that the scalar glueball must mix with the scalar mesons .
so , the game now is _ how much _ @xmath89 exists in each of the three candidates .
is there a winner ?
perhaps not by a landslide , but even so !
the answer depends on who you ask , and how they order the _ unmixed _ @xmath90 , @xmath91 and @xmath89 states .
not much new has been forthcoming in this debate which used to be the talk of the town just a few years ago .
the @xmath57 is now firmly established as being @xmath92 , and it has been found to be healthily populated , along with @xmath56 in @xmath93 annihilation at 5.2 gev/@xmath94@xcite .
levels , ( left ) @xmath95 charmonium @xmath80 states in experiment and theory , ( right ) low lying light quark @xmath80 states . ]
there is a vexing problem with the spin
orbit splitting of the lowest light quark @xmath80 states , and i believe that it relates to the non@xmath76 assignment of @xmath54 .
let me elaborate .
the qcd interaction does not depend on quark flavor .
orbit splitting has the same heirarchy for @xmath90 , @xmath95 and @xmath97 , and it is in the order : @xmath58 ( lowest ) , @xmath98 , @xmath99 . as shown in fig .
2 , it is the order predicted for , and experimentally observed for @xmath95 charmonium , and @xmath97 bottomonium .
it is what is predicted for @xmath90 , for example by godfrey and isgur@xcite . in assigning @xmath55 as the @xmath100 member of the lowest @xmath80
, the pdg would have us believe in the completely inverted order , @xmath99 ( lowest ) , @xmath98 , @xmath58 .
this , of course , becomes necessary because the @xmath54 has been dismissed as a non@xmath76 .
it is far easier for me to believe that the theoretically predicted @xmath58 at 1090 mev has been shifted down to 980 mev by non@xmath76 admixtures , than to believe that it moves up by @xmath101 mev to 1370 mev . in other words , i want to keep alive the possibility of @xmath54 having a substantial component of @xmath76 . in doing so ,
i am entirely aware of the width arguments advanced by godfrey and isgur against @xmath54 being @xmath76 .
having dwelt so long on the @xmath58 light
quark mesons , let me simply mention that the situation about @xmath99 tensors is no better . below 2.4 gev in mass ,
six @xmath102 are predicted by godfrey and isgur@xcite , twelve are claimed by different experiments , and of these only six are considered as well established@xcite .
charmonium states .
the dashed lines indicate states predicted by godfrey and isgur@xcite . ]
as stated earlier , the light
quark mesons , have the experimental problem of numerous closely spaced overlapping states with very large widths , and the theoretical problems of being highly relativistic and having too large a value of @xmath103 ( @xmath104 ) to allow application of pqcd . in contrast , the heavy quark mesons are far more tractable , both because they have well resolved , narrow width bound states ( e.g. , for @xmath95 typical level spacing and width are 100 mev and @xmath105 mev , respectively ) , and tractable relativistic problems ( @xmath106 , @xmath107 ) , and small enough @xmath108 @xmath109 ) to permit use of pqcd .
it follows that if you want to relate meson spectroscopy to the fundamental theory of qcd , you must turn to heavy
quark mesons , and this series of workshops must evolve more and more in that direction .
let me begin by drawing your attention to the level diagram for @xmath95 charmonium states ( in fig . 3 ) .
there are eight bound states before the system becomes unstable with respect to decay into open charm @xmath110 mesons ( @xmath111 ) at 3.73 gev .
the @xmath97 bottomonium system is similar , but because of the three times heavier mass of the @xmath112quark ( @xmath113 , @xmath114 ) , it has fourteen bound states before it becomes unstable with respect to decay into open
beauty @xmath115 mesons ( @xmath116 ) at 10.5 gev .
let me talk about the experimental situation .
the bad news is that many of the facilities which produced most of the recent high precision hadron spectroscopy data have ended .
this includes : bnl ( e852 ) , cern ( cbarrel ) , fermilab ( e760/e835 ) .
the good news is that new and promising facilities are just coming on , or are on the horizon .
these include : dafne , gsi ( fair ) , jparc , bepc ii , and cebaf upgrade .
the ( unexpected ) good news is that the slac and kek @xmath117 factories , not built for spectroscopy , have accumulated such huge luminosities that they seem to be capable of making excellent contributions to hadron spectroscopy , almost wherever they choose to .
progress is made as much by improved precision as by discoveries .
so , let me give you a few examples of precision measurements in heavy quark meson spectroscopy . at novosibirsk , the mass of @xmath10 , @xmath118 , has been measured , i.e. , with an unprecedented precision of 1 part in 3 million@xcite . at cleo , in a single measurement , @xmath119 , lepton universiality
is confirmed to @xmath120@xcite . in another measurment , isospin conservation ( violation )
is confirmed to @xmath121 ( 0.5% ) by measuring @xmath122 decay to @xmath123 versus @xmath124 ( to @xmath125 versus @xmath126)@xcite . in the bottomonium sector
new measurements at cleo@xcite of the leptonic branching ratios of @xmath127 states with a precision of @xmath128 have resulted into up to 55% changes in long accepted values . the precision result , @xmath129 can be now used to validate the latest result of unquenched lattice calculation , @xmath130 .
it appears that lattice is getting there . today
@xmath97 , tomorrow @xmath95 ?
let me move on from improved precision to real discoveries . in the charmonium sector ,
these have consisted of the discovery of the spin
singlet states @xmath131 and @xmath132 .
these bound states have long defied attempts to identify them , the main reson being that they are nearly impossible to populate via the vector @xmath133 formed in @xmath20 annihilation .
radiative decay to @xmath134 is expected to be a very weak m1 transition , and that to @xmath135 is forbidden by charge conjugation .
yet the importance of these spin singlet states can not be exaggerated . together with the known spin
triplet states ( or their centroid ) , these provide measures of the qcd hyperfine splitting , and therefore the @xmath59 * spin spin interaction*. so , our only hope of learning about the spin
spin interaction rests on the identification of these states of charmonium .
the only hyperfine splitting known in the heavy quarkonia is the mass difference of @xmath137 charmonium states , @xmath10 and @xmath138 : @xmath139 a model
independent prediction for the hyperfine splitting of the @xmath140 states is that @xmath141 in 1982 , the crystal ball collaboration at slac claimed identification of @xmath134 and reported @xmath142 , which would correspond to @xmath143 mev , about 35% * larger * than expected .
fortunately , or unfortunately , the crystal ball observation was never confirmed , and @xmath134 remained unidentified for 30 years , until now .
spectra from two photon fusion measurements by cleo ( left ) and babar ( right ) .
the @xmath144 peak is prominent in both spectra.,title="fig:",width=240 ] spectra from two photon fusion measurements by cleo ( left ) and babar ( right ) .
the @xmath144 peak is prominent in both spectra.,title="fig:",width=230 ] in 2002 , belle claimed identification of @xmath134 in the decay of 45 million @xmath117 mesons , @xmath145 and reported @xmath146 mev , which would correspond to @xmath147 mev , a factor * two smaller * than expected and a factor * four * smaller than @xmath148 .
could this be possible ?
well , perhaps !
there are two important ways @xmath140 states differ from @xmath137 states .
@xmath137 states , with @xmath149 , lie in the coulombic region ( prop . to @xmath150 ) of the @xmath59 potential , @xmath151 , whereas the @xmath140 states , with @xmath152 , lie in the confinement part ( prop . to @xmath153 ) of the potential .
the spin spin potential in the two regions could be different .
the second difference is that the @xmath140 states , particularly @xmath122 , lie close to the @xmath110 breakup threshold at 3730 mev , and can be expected to mix with the continuum as well as higher @xmath154 states .
all in all , it is important to nail down @xmath134 experimentally , and measure its mass accurately .
this was successfully done by cleo@xcite and babar@xcite in 2004 by observing @xmath134 in two photon fusion , @xmath155 .
the two observations are shown in fig .
the average of all observations is @xmath156 mev , which leads to @xmath157 mev , which is almost a factor * 2.5 smaller * than @xmath148 . explaining this large difference is a challenge to the theorists .
the challenge for the experimentalists lies in completing the spectroscopy of @xmath134 , now that its mass is known .
the interest in identifying @xmath135 is also related to the hyperfine or @xmath159 interaction , but in a different way . in the conventional @xmath59 potential , the confinement part is scalar , and therefore there is no long range spin
spin interaction .
the @xmath159 interaction is only due to the coulombic part , and is a contact interaction proportional to the wave function at the origin . in other words , it is finite only for @xmath160states . for @xmath161wave , or higher @xmath162 states ,
the interaction is zero .
thus , for @xmath163 states , the @xmath164 state should be degenerate with the centroid of the @xmath80 states , or @xmath165 there are many ways this prediction could be wrong , viz the simple way of calculating the centroid as @xmath166/9 $ ] could be ( and is ) wrong , or the @xmath59 potential could have a long range @xmath159 part in it .
it is therefore important to experimentally determine @xmath167 by identifying the @xmath158 state of charmonium .
numerous attempts to identify @xmath135 made during the last 35 years had failed .
now , two attempts have succeeded .
the fermilab e835 @xmath93 annihilation experiment@xcite has claimed @xmath135 observation at @xmath168 level@xcite in the reaction @xmath169 and reported @xmath170 mev .
cleo has reported @xmath135 observation at @xmath171 level in the reaction @xmath172 . both inclusive ( @xmath173 and @xmath174 detected ) and
exclusive analyses were made , with the result @xmath175 mev .
the exclusive spectrum is shown in fig .
it appears that the @xmath163 hyperfine splitting is very small , perhaps even zero . with the discovery of @xmath134 and @xmath135 ,
a milestone has been reached in charmonium spectroscopy .
the bound state spectrum is now complete .
cleo is now the only producer of @xmath97 bottomonium spectroscopy results . besides improving the precision of the branching ratio measurements for the radiative@xcite and leptonic @xcite decays of the upsilon states
, cleo has made the first observation of a non@xmath63 transition between bottomonium states@xcite , and the first observation of a @xmath176state in bottomonium@xcite . since no bottomonium physics
was presented at the workshop , i will dwell no further on it , and return to meson spectroscopy below 5 gev .
the renaissance in hadron spectroscopy to which i referred earlier has been triggered by the discovery , or claimed discovery , of a number of unanticipated states .
this is very exciting , and also confusing , because the nature of these states is still not understood . in chronological order ,
the series consists of the pentaquark ( spring8jan .
2003 ) , @xmath177 ( babar march 2003 ) , x(3872 ) ( belle march 2003 ) , x , y , z(3940 ) ( belle 2005 ) , and y(4260 ) ( babar 2005 ) .
there have been many reports of the sightings of the pentaquark @xmath178 and its heavy quark siblings , and even a greater number of reported failures in finding them .
google tells me that there were 123,000 entries for the pentaquark by march 28 , 2006 , 135,000 by april 10 , and 151,000 by the time of this writing in july 25 , 2006 .
this tells you how exciting and intoxicating the idea is , or rather was .
see , for example , refs .
34 and 34 .
the pentaquark has been on `` life support '' for too long .
we should let it go may it rest in peace ! discovered by belle@xcite , and quickly confirmed by cdf@xcite , d@xcite , and babar@xcite , the x(3872 ) state definitely exists .
its average mass and width are @xmath179 , and @xmath180 , and it decays prominently in @xmath123 .
what is debatable , however , is `` what is it ? ''
numerous speculations have been made .
is it a charmonium ( @xmath181 , @xmath182 or @xmath183 ) or ( @xmath184 , @xmath185 or @xmath186)@xcite ?
is it a @xmath184 charmonium hybrid@xcite ?
is it a vector glueball mixed with vector charmonium@xcite , or is it a @xmath187 molecule@xcite ?
a firm determination of the @xmath47 of x(3872 ) would go a long way in sifting through the above models .
unfortunately , it is difficult to say that any of the claims made so far pin the @xmath47 down unambiguously .
a number of reported decays , e.g. , @xmath188 , @xmath189 , @xmath190 , which would pin down some of the quantum numbers , suffer from limited statistics , and angular distribution studies and the dipion mass distribution in @xmath191 appear to be inconclusive . the present toss
up appears to be between @xmath192 and @xmath193 , with @xmath98 being a slight favorite . with the goal of pinning down the binding energy of @xmath194 ( if it is bound )
, cleo has announced@xcite a preliminary precision determination of the @xmath195 mass , @xmath196 , so that the best estimate of the binding energy of the @xmath194 molecule as @xmath197 is now @xmath198 , which constrains the molecular model considerably more than before .
further constraint would require a more precise measurement of the mass of x. with huge luminosities ( @xmath199 fb@xmath200 ) available to it , belle has been announcing the discovery of resonance after resonance .
by now we have x(@xmath201 ) , y(@xmath202 ) , and z(@xmath203 ) ) , with all masses being consistent with 3935 mev .
the latest results for these states are summarixed in table i. as shown in the table , the errors in the parameters are large enough , and the evidence at least for _ non - observations _ is tentative enough so that not all three , x , y , z , need be distinct and separate states .
this is particularly true for x and z whose widths are similar , and both of them could accomodate different levels of @xmath204 and @xmath194 decays .
it is obvious that lots more experimental data will be needed before the dust settles down . in this respect
, it is very disappointing that babar , with comparable data sets available to it , has maintained an ominous silence on x , y , z. with the excitement these states have generated , one would think that babar would jump right in , and prove or disprove the belle claims ! ; ( left ) by babar , ( right ) by cleo.,title="fig:",width=220 ] ; ( left ) by babar , ( right ) by cleo.,title="fig:",width=196 ] while babar has regretably stayed away from belle s x , y , z states , it has exploited its expertise in isr ( initial state radiation ) , or rr ( radiative return ) , studies to present another surprise@xcite .
it has presented evidence for the observation of a broad resonance formed in @xmath20 annihilation following isr , and decaying into @xmath123 ( see fig . 6 ) .
the parameters of the resonance are @xmath205 , and @xmath206 .
the formation of @xmath207 in isr and the s wave dipion mass distribution essentially ensure that @xmath208 .
this comes as a great surprise , because there is a deep minimum in the @xmath209 measurement at 4260 mev , which would militate against the presence of a vector with this mass .
it is therefore extremely important to confirm babar s observation .
this has been done in two new measurements at cleo . in the first measurement , cleo@xcite has confirmed a strong ( factor @xmath210 ) enhancement of the @xmath211 yield at 4260 mev compared to the yield at 4160 mev and 4040 mev .
lest one argue that enhancement at a single energy point does not confirm a resonance , in a second measurement , cleo has confirmed the y(4260 ) in isr , exactly the same measurement as babar ( see fig .
they obtain @xmath212 , and @xmath213 .
the measurement has very small background , and a very clean ( @xmath214 ) signal .
so , the y(4260 ) @xmath154 state is very real .
but , what is it ?
no charmonium vector is expected with this mass , which deepens the mystery .
the usual speculations , such as misplaced charmonium , a @xmath215 hybrid , a molecule , etc . , abound .
clearly , lots more measurements are needed to sift through these speculations . as a postscript to the discovery of all these new and unanticipated states , and the uncertainty about their nature ,
let me recall that the discovery of @xmath10@xcite was followed by a spate of theoretical papers about its nature@xcite , most of which , even by famous nobelists , turned out to be wrong . that is how physics develops .
let me now move to the non quarkonium results presented at the workshop .
here the heavy quark can be either the charm or bottom quark , and the light quark can be @xmath217 , or @xmath218 .
the convention for labeling the quantum numbers of heavy
light mesons is to couple the spin @xmath219 of the heavy quark with the @xmath220 of the light quark to give the total spin @xmath221 .
we have had considerable activity in this sector , with babar , belle and cleo ( particularly in its reincarnation as cleo - c ) making important contributions .
the first exciting news was babar s discovery@xcite of a new narrow width @xmath223 meson with @xmath224 , @xmath225 , unexpectedly below the @xmath226 threshold .
this was followed by the discovery of its @xmath227 partner with @xmath228 , below the @xmath229 threshold , by cleo@xcite .
these mesons were expected to lie above their respective thresholds , and to be therefore quite wide .
theoretical suggestions to explain the observed characteristics of these mesons again range from simple @xmath230 to @xmath231 molecules , to @xmath230 mixed with @xmath226 and @xmath229 continua , and to tetraquark , with no resolution so far .
the corresponding non - strange @xmath232 states would be expected at 2217 mev ( @xmath233 ) and 2363 mev ( @xmath234 ) .
these are expected to be wide , and have not been firmly identified so far . admittedly , the main interest in heavy
light mesons rests in the study of their weak interaction properties , and cleo - c has made this their major goal .
hopefully precision measurements in this sector will go a long way towards validating the lattice qcd calculations .
cleo has reported preliminary form factor results , @xmath235@xcite and @xmath236@xcite .
the corresponding predicitions from unquenched lattice calculations are @xmath237 and @xmath238@xcite .
cleo has also measured a large number of hadronic and weak decays of @xmath176 and @xmath239 mesons .
this is the domain of cdf and d contributions .
the latest of their contributions relate to the identification and mass measurements of the excited @xmath117mesons . in the notation @xmath241 the masses are : + [ cols= " < , < , < " , ] + and the tour - de - force measurement by cdf is @xmath242 , which compares with the lattice prediction of @xmath243 .
since we had a presentation of the latest results on @xmath244 violation , let me summarize what we know now .
the best values of the angles of the unitary triangle are @xmath245 , @xmath246 , and @xmath247 .
one set of ckm matrix elements is : @xmath248 , @xmath249 , @xmath250 , and @xmath251 , and , of course , @xmath252 .
as i mentioned in the beginning , hadron spectroscopy has had a renaissance during the last five years .
we have heard about exciting new discoveries at this workshop , and the future promises many more to come . | a summary of the present experimental status of meson physics is presented .
the presentation includes the new results presented at the meson06 workshop , as well as other recent experimental developments in the field . |
our aim in physics is not only to calculate some observable and get a correct number but mainly to understand a physical picture responsible for the given phenomenon .
it very often happens that a theory formulated in terms of fundamental degrees of freedom can not answer such a question since it becomes overcomplicated at the related scale .
thus a main task in this case is to select those degrees of freedom which are indeed essential .
for instance , the fundamental degrees of freedom in crystals are ions in the lattice , electrons and the electromagnetic field . nevertheless , in order to understand electric conductivity , heat capacity , etc .
we instead work with `` heavy electrons '' with dynamical mass , phonons and their interaction . in this case
a complicated electromagnetic interaction of the electrons with the ions in the lattice is `` hidden '' in the dynamical mass of the electron and the interactions among ions in the lattice are eventually responsible for the collective excitations of the lattice - phonons , which are goldstone bosons of the spontaneously broken translational invariance in the lattice of ions . as a result ,
the theory becomes rather simple - only the electron and phonon degrees of freedom and their interactions are essential for all the properties of crystals mentioned above .
quite a similar situation takes place in qcd .
one hopes that sooner or later one can solve the full nonquenched qcd on the lattice and get the correct nucleon and pion mass in terms of underlying degrees of freedom : current quarks and gluon fields .
however , qcd at the scale of 1 gev becomes too complicated , and hence it is rather difficult to say in this case what kind of physics , inherent in qcd , is relevant to the nucleon mass and its low - energy properties . in this lecture
i will try to answer this question .
i will show that it is the spontaneous breaking of chiral symmetry which is the most important qcd phenomenon in this case , and that beyond the scale of spontaneous breaking of chiral symmetry light and strange baryons can be viewed as systems of three constituent quarks which interact by the exchange of goldstone bosons ( pseudoscaler mesons ) and are subject to confinement .
it is well known that at low temperature and density the approximate @xmath0 chiral symmetry of qcd is realized in the hidden nambu - goldstone mode .
the hidden mode of chiral symmetry is revealed by the existence of the octet of pseudoscalar mesons of low mass which represent the associated approximate goldstone bosons .
the @xmath1 ( the @xmath2-singlet ) decouples from the original nonet because of the @xmath3 anomaly @xcite .
another consequence of the spontaneous breaking of the approximate chiral symmetry of qcd is that the valence quarks acquire their dynamical or constituent mass @xcite through their interactions with the collective excitations of the qcd vacuum - the quark - antiquark excitations and the instantons .
we have recently suggested @xcite that beyond the scale of spontaneous breaking of chiral symmetry a baryon should be considered as a system of three constituent quarks with an effective quark - quark interaction that is formed by a central confining part and a chiral interaction mediated by the octet of pseudoscalar mesons between the constituent quarks .
the simplest representation of the most important component of the interaction of the constituent quarks mediated by the octet of pseudoscalar bosons in the @xmath4 invariant limit is @xmath5 here @xmath6 represents the flavor @xmath2 gell - mann matrices and the sums run over the constituent quarks . because of the flavor dependent factor @xmath7 the chiral boson exchange interaction ( [ 1.1 ] )
will lead to orderings of the positive and negative parity states in the baryon spectra , which agree with the observed ones in all sectors . for the spectrum of the nucleon and @xmath8 the strength of the chiral interaction between the constituent quarks is sufficient to shift the lowest positive parity states in the @xmath9=2 band ( the @xmath10 and @xmath11 ) below the negative parity states in the @xmath9=1 band ( @xmath12 and @xmath13 ) . in the spectrum of the @xmath14 , on the other hand
, it is the negative parity flavor singlet states ( @xmath15 ) that remain the lowest lying resonances , again in agreement with experiment .
the mass splittings between the baryons with different strangeness , and between the @xmath14 and the @xmath16 which have identical flavor , spin and flavor - spin symmetries arise from the explicit breaking of the @xmath4 symmetry that is caused by the mass splitting of the pseudoscalar meson octet , and the different masses of the u , d and the s quarks .
this lecture has the following structure .
section 2 contains the proof why the commonly used perturbative gluon exchange interaction between the constituent quarks leads to incorrect ordering of positive and negative parity states in the spectra . in section 3
we outline the importance of the spontaneous breaking of chiral symmetry for low - energy qcd , and in section 4 we present a short historical sketch of the role of chiral symmetry in quark based models .
section 5 contains a description of the chiral boson exchange interaction ( [ 1.1 ] ) . in section 6
we describe the spectra of the nucleon , the @xmath8 resonance and the @xmath14 hyperon as they are predicted by the @xmath4 symmetric interaction ( [ 1.1 ] ) .
the effect of @xmath4 breaking in the interaction is considered in section 7 .
section 8 is devoted to an `` exact '' three - body description of baryons where the interaction ( [ 1.1 ] ) is taken into account to all orders . in section 9
we discuss the role of the exchange current corrections to the baryon magnetic moments that are associated with the pseudoscalar exchange interaction .
finally , in section 10 some recent lattice - qcd results are discussed .
it was accepted by many people ( but not by all ) that the fine splittings in the baryon spectrum ( in analogy to atomic physics they are often called `` hyperfine splittings '' as they arise from the spin - spin forces ) are due to the gluon - exchange interaction between the constituent quarks @xcite .
now i shall address myself to a formal consideration why the one gluon exchange interaction can not be relevant to the baryon spectrum .
the most important component of the one gluon exchange interaction @xcite is the so called color - magnetic interaction @xmath17 where @xmath18 are color @xmath2 matrices .
it is the permutational color - spin symmetry of the 3q state which is mostly responsible for the contribution of the interaction ( [ 2.1 ] ) .
indeed , the corresponding two - body matrix element is @xmath19_{\rm c}\times [ f_{ij}]_{\rm s}:[f_{ij}]_{\rm cs}|\vec \lambda^{\rm c}_i \cdot \vec \lambda_j^{\rm c } \vec\sigma_i \cdot \vec \sigma_j|[f_{ij}]_{\rm c } \times [ f_{ij}]_{\rm s}:[f_{ij}]_{\rm cs}>\ ] ] @xmath20_{\rm c},[11]_{\rm s}:[2]_{\rm cs } \\
-{8\over 3 } & [ 11]_{\rm c},[2]_{\rm s}:[11]_{\rm cs}\end{array}\right .. \label{2.2}\ ] ] thus the symmetrical color - spin pairs ( i.e. with a @xmath21_{\rm cs}$ ] young pattern ) experience an attractive contribution while the antisymmetrical ones ( @xmath22_{\rm cs}$ ] ) experience a repulsive contribution .
hence the color - magnetic contribution to the @xmath8 state ( @xmath23_{\rm cs}$ ] ) is more repulsive than to the nucleon ( @xmath24_{\rm cs}$ ] ) and the @xmath8 becomes heavier than the nucleon .
the price is that @xmath25 should be larger than unity , which is bad .
in addition there is no empirical indication in the spectrum for a large spin - orbit component of the gluon - exchange interaction @xcite implied by this big value of @xmath26 .
the crucial point is that the interaction ( [ 2.1 ] ) feels only colour and spin of the interacting quarks .
thus the structure of the @xmath27 and @xmath14 spectra has to be the same as these baryons differ only by their flavour structure .
if one looks at the particle data group tables , however , one immediately sees a different ordering of the positive and negative parity states in both spectra . in the @xmath27 spectrum
the lowest states are @xmath28 ; @xmath29 ; @xmath30 @xmath31 @xmath32 , while in the @xmath14 spectrum the ordering is as follows : @xmath33 ; @xmath34 @xmath31 @xmath35 ; @xmath36 . a weak flavor dependence via the quark masses in ( [ 2.1 ] )
can not explain this paradox .
the interaction ( [ 2.1 ] ) can not explain why the two - quantum excitations of positive parity @xmath10 , @xmath11 , @xmath37 and @xmath38 lie below the one - quantum excitations of negative parity @xmath39 @xmath31 @xmath40 , @xmath41 , @xmath42 and @xmath43 , respectively .
for instance , the positive parity state @xmath10 and the negative parity ones @xmath39 @xmath31 @xmath40 have the same mixed ( @xmath24_{\rm cs}$ ] ) color - spin symmetry thus the color - magnetic contribution to these states can not be very different ( a small difference is only due to the different radial structure of the positive and negative parity states ) . but
the @xmath10 state belongs to the @xmath44 shell , while the @xmath45 pair is a member of the @xmath46 band , which means that the @xmath10 should lie approximately @xmath47 above the @xmath45 . in the @xmath8 spectrum
the situation is even more dramatic .
the @xmath11 positive parity state has a completely antisymmetric cs - young pattern ( @xmath23_{\rm cs}$ ] ) , while the negative parity states @xmath48 have a mixed one .
thus the color - magnetic contribution to the @xmath11 is much more repulsive than to the @xmath48 .
in addition the @xmath11 is the @xmath44 state , while the pair @xmath48 belongs to the @xmath46 band . as a consequence
the @xmath11 must lie much higher than the @xmath48 .
all these features are well seen in the explicit 3-body calculations @xcite .
the qcd lagrangian with three light flavors has a global symmetry @xmath49 if one neglects the masses of current u , d , and s quarks , which are small compared to a typical low - energy qcd scale of 1 gev .
the @xmath3 is not a symmetry at the quantum level due to the axial anomaly .
if the @xmath0 chiral symmetry of the qcd lagrangian were intact in the vacuum state we would observe degenerate multiplets in the particle spectrum corresponding to the above chiral group , and all hadrons would have their degenerate partners with opposite parity . since this does not happen the implication is that the chiral symmetry is spontaneously broken down to @xmath50 in the qcd vacuum , i.e. , realized in the hidden nambu - goldstone mode .
a direct evidence for the spontaneously broken chiral symmetry is a nonzero value of the quark condensates for the light flavors @xmath51 which represents the order parameter . that this is indeed so , we know from three independent sources : current algebra @xcite , qcd sum rules @xcite , and lattice gauge calculations @xcite .
there are two important generic consequences of the spontaneous chiral symmetry breaking .
the first one is an appearance of the octet of pseudoscalar mesons of low mass , @xmath52 , which represent the associated approximate goldstone bosons .
the second one is that valence quarks acquire a dynamical or constituent mass .
both these consequences of the spontaneous chiral symmetry breaking are well illustrated by , e.g. the @xmath53-model @xcite or the nambu and jona - lasinio model @xcite .
we can not say at the moment for sure what the microscopical reason for spontaneous chiral symmetry breaking in the qcd vacuum is .
it was suggested that this occurs when quarks propagate through instantons in the qcd vacuum @xcite .
for the low - energy baryon properties it is only essential that beyond the spontaneous chiral symmetry breaking scale new dynamical degrees of freedom appear - constituent quarks and chiral fields .
the low - energy baryon properties are mainly determined by these dynamical degrees of freedom and the confining interaction .
this is quite in contrast to pseudoscalar mesons . in the chiral limit
, @xmath54 , all members of the pseudoscalar octet ( @xmath52 ) would have zero mass , which is most clearly seen in the gell - mann - oakes - renner @xcite relations @xmath55 @xmath56 @xmath57 @xmath58 @xmath59 that relate the pseudoscalar meson masses to the quark condensates and current quark masses @xmath60 .
thus the nonzero masses of mesons are determined by the nonzero values of the current quark masses . in the baryon case , even in the chiral limit , baryons would have approximately their actual masses of the order of 1 gev , as these masses are mostly determined by the dynamical ( constituent ) masses , the goldstone boson exchange interaction among them , as well as a confining interaction .
the dynamical ( constituent ) masses are in turn determined mainly by the quark condensates , which is most clearly seen from the gap equations of the nambu and jona - lasinio model , and only weakly dependent on current quark masses .
the role of the current quark masses in baryons is to break just the @xmath4 symmetry in the baryon spectrum .
the importance of the constraints posed by chiral symmetry for the quark bag @xcite and bag - like @xcite models for the baryons has been recognized early . in bag models with
restored chiral symmetry on the bag surface , or bag - like models , the massless current quarks within the bag were assumed to interact not only by perturbative gluon exchange but also through chiral meson field exchange . in these models the chiral field has the character of a compensating auxiliary field only , rather than a collective low frequency goldstone quark - antiquark excitation .
the possibility of a nonzero quark condensate was also not addressed .
as it was discussed in a previous section , it is the quark condensate which is the most important characteristic determining the baryon properties . according to the bag philosophy ,
however , there is a perturbative qcd phase inside a bag , where all condensates vanish by definition .
a general limitation of all bag and bag - like models is , of course , the lack of translational invariance , which is of crucial importance for a description of the excited states .
common to these models is that the breaking of chiral symmetry arises from the confining interaction .
this point of view contrasts with that of manohar and georgi @xcite , who pointed out that there should be two different scales in qcd with 3 flavors . at the first one of these , @xmath61 1 gev , the spontaneous breaking of the chiral symmetry occurs , and hence at distances beyond @xmath62 0.2 fm the valence current quarks acquire their dynamical ( constituent ) mass ( called `` chiral quarks '' in @xcite ) , and the goldstone bosons ( mesons ) appear .
the other scale , @xmath63 mev , is that which characterizes confinement , and the inverse of this scale roughly coincides with the linear size of a baryon . between these two scales
then the effective lagrangian should be formed out of the gluon fields that provide a confining mechanism , as well as of the constituent quark and pseudoscalar meson fields .
manohar and georgi did not , however , specify whether the baryons should be described as bound qqq states or as chiral solitons .
the chiral symmetry breaking scale above fits well with that which appears in the instanton liquid picture of the qcd vacuum @xcite . in this model
the quark condensates ( i.e. , equilibrium of virtual quark - antiquark pairs in the vacuum state ) , as well as the gluon condensate , are supported by instanton fluctuations of a size @xmath64 fm .
diakonov and petrov @xcite suggested that at low momenta ( i.e. , beyond the chiral symmetry breaking scale ) qcd should be approximated by an effective chiral lagrangian of the sigma - model type that contains valence quarks with dynamical ( constituent ) masses and meson fields .
they considered a nucleon as three constituent quarks moving independently of one another in a self - consistent chiral field of the hedgehog form @xcite . in this picture
the @xmath8 appears as a rotational excitation of the hedgehog , and no explicit confining interaction is included .
a very similar description for the nucleon was suggested in @xcite .
these types of models are now called `` quark - soliton models '' @xcite . the spontaneous breaking of chiral symmetry and its consequences - dynamical quark mass generation , appearance of the quark condensate , and pseudoscalar mesons as goldstone excitations - are well illustrated by the nambu and jona - lasinio model @xcite .
this model lacks a confining interaction , which , as argued below , is essential for a realistic description of the properties of baryon physics .
in an effective chiral description of the baryon structure , based on the constituent quark model , the coupling of the quarks and the pseudoscalar goldstone bosons will ( in the @xmath4 symmetric approximation ) have the form @xmath65 ( or @xmath66 ) , where @xmath67 is the fermion constituent quark field operator , @xmath68 the octet boson field operator , and @xmath69 is a coupling constant . a coupling of this form , in a nonrelativistic reduction for the constituent quark spinors , will to lowest order give rise to a yukawa interaction between the constituent quarks , the spin - spin component of which has the form @xmath70 here @xmath71 and @xmath72 denote the masses of the interacting quarks , and @xmath73 that of the meson .
there will also be an associated tensor component , which is discussed in ref .
@xcite . at short range the simple form ( [ 5.1 ] ) of the chiral boson exchange interaction can not be expected to be realistic and should only be taken to be suggestive . because of the finite spatial extent of both the constituent quarks and the pseudoscalar mesons the delta function in ( [ 5.1 ] ) should be replaced by a finite function , with a range of 0.6 - 0.7 fm , as suggested by the spatial extent of the mesons .
in addition , the radial behaviour of the yukawa potential ( [ 5.1 ] ) is valid only if the boson field satisfies a linear klein - gordon equation .
the implications of the underlying chiral symmetry of qcd for the effective chiral lagrangian ( which in fact is not known ) , which contains constituent quarks as well as boson fields , are that these boson fields can not be described by linear equations near their source .
therefore it is only at large distances , where the amplitude of the boson fields is small , that the quark - quark interaction reduces to the simple yukawa form .
the latter point is rather important and has to be clarified . the radial dependence in ( [ 5.1 ] )
is a direct consequence of a free @xmath74 green function for the boson field , and a pseudoscalar @xmath65 or pseudovector @xmath66 coupling in both quark - meson vertices .
making use of the usual `` static approximation '' and neglecting the recoil corrections at both vertices in a nonrelativistic reduction for the constituent quark spinors , one arrives at ( [ 5.1 ] ) .
the free green function above comes from the well - known lagrangian for a free boson field @xmath75 , which means that the boson field satisfies a linear klein - gordon equation . on the other hand ,
the underlying @xmath76 chiral symmetry of qcd tells that the goldstone boson field can not be described by such a simple klein - gordon lagrangian .
the transformation properties of the goldstone boson field under @xmath77 and @xmath78 chiral rotations , @xmath79 and @xmath80 , are well defined if the goldstone boson fields @xmath68 are combined in a unitary matrix @xcite @xmath81 @xmath82 the `` kinetic term '' in this case is @xmath83 which in its expansion in powers of @xmath84 contains not only a free klein - gordon kinetic term but also other terms of higher powers .
for instance , the terms of fourth order in the meson field give rise to @xmath85 scattering @xcite , etc .
the full chiral lagrangian describing the goldstone boson field is unknown and should contain a quartic term involving derivatives @xmath86 and higher order terms .
it is clear that a dressed green function for the field @xmath84 should be very different compared to the free klein - gordon green function since the selfinteraction of the goldstone boson field is important . only far away from its source ( the fermion current ) , where the amplitude of the boson field @xmath84 is small , the lowest term in powers of @xmath84 becomes dominant , and therefore only at large distances the quark - quark interaction reduces to its simple yukawa form .
_ at this stage the proper procedure should be to avoid further specific assumptions about the short range behavior of @xmath87 in ( [ 1.1 ] ) , to extract instead the required matrix elements of it from the baryon spectrum , and to reconstruct by this an approximate radial form of @xmath87 .
the overall minus sign in the effective chiral boson interaction in ( [ 1.1 ] ) corresponds to that of the short range term in the yukawa interaction .
_ the flavor structure of the pseudoscalar octet exchange interaction in ( [ 1.1 ] ) between two quarks @xmath88 and @xmath89 should be understood as follows : @xmath90 the first term in ( [ 5.5 ] ) represents the pion - exchange interaction , which acts only between light quarks .
the second term represents the kaon exchange interaction , which takes place in u - s and d - s pair states .
the @xmath91-exchange , which is represented by the third term , is allowed in all quark pair states . in the @xmath4
symmetric limit the constituent quark masses would be equal ( @xmath92 ) , the pseudoscalar octet would be degenerate and the meson - constituent quark coupling constant would be flavor independent . in this limit
the form of the pseudoscalar exchange interaction reduces to ( [ 1.1 ] ) , which does not break the @xmath4 invariance of the baryon spectrum . beyond this limit the pion , kaon and @xmath91 exchange interactions will differ ( @xmath93 ) , because of the difference between the strange and u
, d quark constituent masses ( @xmath94 ) , and because of the mass splitting within the pseudoscalar octet ( @xmath95 ) ( and possibly also because of flavor dependence in the meson - quark coupling constant ) .
the source of both the @xmath4 symmetry breaking constituent quark mass differences and the @xmath4 symmetry breaking mass splitting of the pseudoscalar octet is the explicit chiral symmetry breaking in qcd .
the two - quark matrix elements of the interaction ( [ 1.1 ] ) are : @xmath19_{\rm f}\times [ f_{ij}]_{\rm s } : [ f_{ij}]_{\rm fs } { } ~| -v(r_{ij})\vec \lambda^{\rm f}_i \cdot \vec \lambda_j^{\rm f } \vec\sigma_i \cdot \vec \sigma_j { } ~|~[f_{ij}]_{\rm f } \times [ f_{ij}]_{\rm s } : [ f_{ij}]_{\rm fs}>\ ] ] @xmath96_{\rm f},[2]_{\rm s}:[2]_{\rm fs } \\
-8v(r_{ij } ) & [ 11]_{\rm f},[11]_{\rm s}:[2]_{\rm fs } \\ 4v(r_{ij } ) & [ 2]_{\rm f},[11]_{\rm s}:[11]_{\rm fs}\\ { 8\over 3}v(r_{ij } ) & [ 11]_{\rm f},[2]_{\rm s}:[11]_{\rm fs}\end{array}\right .. \label{6.1}\ ] ] from these the following important properties may be inferred : \(i ) at short range , where @xmath97 is positive , the chiral interaction ( [ 1.1 ] ) is attractive in the symmetric fs pairs and repulsive in the antisymmetric ones . at large distances
the potential function @xmath97 becomes negative and the situation is reversed .
\(ii ) at short range , among the fs - symmetrical pairs , the flavor antisymmetric pairs experience a much larger attractive interaction than the flavor - symmetric ones , and among the fs - antisymmetric pairs the strength of the repulsion in flavor - antisymmetric pairs is considerably weaker than in the symmetric ones . given these properties we conclude , that with the given flavor symmetry , the more symmetrical the fs young pattern is for a baryon the more attractive contribution at short range comes from the interaction ( [ 1.1 ] ) . for two identical flavor - spin young patterns @xmath98_{\rm fs}$ ]
the attractive contribution at short range is larger for the more antisymmetrical flavor young pattern @xmath98_{\rm f}$ ] .
consider first , for the purposes of illustration , a schematic model which neglects the radial dependence of the potential function @xmath87 in ( [ 1.1 ] ) , and assume a harmonic confinement among quarks as well as @xmath99 . in this model @xmath100
if the only interaction between the quarks were the flavor- and spin - independent harmonic confining interaction , the baryon spectrum would be organized in multiplets of the symmetry group @xmath101 . in this case
the baryon masses would be determined solely by the orbital structure , and the spectrum would be organized in an _ alternative sequence of positive and negative parity states . _ the hamiltonian ( [ 6.2 ] ) , within a first order perturbation theory ,
reduces the @xmath101 symmetry down to @xmath102 , which automatically implies a splitting between the octet and decuplet baryons . for the octet states
@xmath27 , @xmath14 , @xmath16 , @xmath103 ( @xmath104 shell , @xmath9 is the number of harmonic oscillator excitations in a 3-quark state ) as well as for their first radial excitations of positive parity ( breathing modes ) @xmath10 , @xmath37 , @xmath38 , @xmath105 ( @xmath44 shell ) the flavor and spin symmetries are @xmath106_{\rm fs}[21]_{\rm f}[21]_{\rm s}$ ] , and the contribution of the hamiltonian ( [ 6.2 ] ) is @xmath107 . for the decuplet states @xmath8 , @xmath108 , @xmath109 , @xmath110 ( @xmath104 shell ) the flavor and spin symmetries , as well as the corresponding matrix element , are @xmath106_{\rm fs}[3]_{\rm f}[3]_{\rm s}$ ] and @xmath111 , respectively . the first negative parity excitations ( @xmath46 shell ) in the @xmath27 and @xmath16 spectra @xmath39 - @xmath112 and @xmath113 - @xmath114
are described by the @xmath24_{\rm fs}[21]_{\rm f}[21]_{\rm s}$ ] symmetries , and the contribution of the interaction ( [ 6.2 ] ) in this case is @xmath115 .
the first negative parity excitation in the @xmath14 spectrum ( @xmath46 shell ) @xmath116 - @xmath117 is flavor singlet @xmath24_{\rm fs}[111]_{\rm f}[21]_{\rm s}$ ] , and , in this case , the corresponding matrix element is @xmath118 .
these matrix elements alone suffice to prove that the ordering of the lowest positive and negative parity states in the baryon spectrum will be correctly predicted by the chiral boson exchange interaction ( [ 6.2 ] ) .
the constant @xmath119 may be determined from the n@xmath120 splitting to be 29.3 mev .
the oscillator parameter @xmath121 , which characterizes the effective confining interaction , may be determined as one half of the mass differences between the first excited @xmath122 states and the ground states of the baryons , which have the same flavor - spin , flavor and spin symmetries ( e.g. @xmath10 - @xmath27 , @xmath37 - @xmath14 , @xmath38 - @xmath16 ) , to be @xmath123 mev .
thus the two free parameters of this simple model are fixed and we can make now predictions . in the @xmath27 and @xmath16 sectors
the mass difference between the lowest excited @xmath124 states ( @xmath10 and @xmath38 ) and @xmath125 negative parity pairs ( @xmath39 - @xmath40 and @xmath113 - @xmath114 ) will then be @xmath126 whereas for the @xmath14 system ( @xmath37 , @xmath116 - @xmath117 ) it should be @xmath127 this simple example shows how the chiral interaction ( [ 6.2 ] ) provides different ordering of the lowest positive and negative parity excited states in the spectra of the nucleon and the @xmath14-hyperon .
this is a direct consequence of the symmetry properties of the boson - exchange interaction discussed at the beginning of this section .
namely , the @xmath106_{\rm fs}$ ] state in the @xmath10 , @xmath11 and @xmath38 positive parity resonances from the @xmath44 band feels a much stronger attractive interaction than the mixed symmetry state @xmath24_{\rm fs}$ ] in the @xmath39 - @xmath40 , @xmath128 - @xmath129 and @xmath113 -@xmath114 resonances of negative parity ( @xmath46 shell ) .
consequently the masses of the positive parity states @xmath10 , @xmath11 and @xmath38 are shifted down relative to the other ones , which explains the reversal of the otherwise expected `` normal ordering '' .
the situation is different for @xmath116 - @xmath117 and @xmath37 , as the flavor state of @xmath116 - @xmath117 is totally antisymmetric . because of
this the @xmath116 - @xmath117 gains an attractive energy , which is comparable to that of the @xmath37 , and thus the ordering suggested by the confining oscillator interaction is maintained .
consider now , in addition , the radial dependence of the potential with the @xmath4 invariant version ( [ 1.1 ] ) of the chiral boson exchange interaction ( i.e. , @xmath130 ) . if the confining interaction in each quark pair is taken to have the harmonic oscillator form as above , the exact eigenvalues and eigenstates to the coinfining 3q hamiltonian are @xmath131 @xmath132_{\rm x}[f]_{\rm fs}[f]_{\rm f}[f]_{\rm s}>,\label{6.6}\ ] ] where @xmath9 is the number of quanta in the state , the elliott symbol @xmath133 characterizes the @xmath2 harmonic oscillator symmetry , and @xmath134 is the orbital angular momentum .
the spatial ( x ) , flavor - spin ( fs ) , flavor ( f ) , and spin ( s ) permutational symmetries are indicated by corresponding young patterns ( diagrams ) @xmath98 $ ] .
all these functions are well known ( see , e.g. , @xcite ) .
note that the color state @xmath23_{\rm c}$ ] , which is common to all the states , has been suppressed in ( [ 6.6 ] ) . by the pauli principle @xmath98_{\rm x}=[f]_{\rm fs}$ ] .
the full hamiltonian is the sum of the confining hamiltonian and the chiral field interaction ( [ 1.1 ] ) .
when the boson exchange interaction ( [ 1.1 ] ) is treated in first order perturbation theory , the mass of the baryon states takes the form @xmath135 where the chiral interaction contribution is @xmath136 and @xmath137 the contribution from the chiral interaction to each baryon is a linear combination of the matrix elements of the two - body potential @xmath138 , defined as @xmath139 represents the oscillator wavefunction with @xmath140 excited quanta . as we shall only consider the baryon states in the @xmath141 bands , we shall only need the four radial matrix elements @xmath142 and @xmath143 for the numerical construction of the spectrum .
the contributions to all nucleon , @xmath8 and @xmath14-hyperon states from the boson exchange interaction , in terms of the matrix elements @xmath144 , are listed in tables 1 and 2 . in this approximate @xmath4-invariant version of the chiral boson exchange interaction the @xmath145 and the @xmath146 mass differences
would solely be ascribed to the mass difference between the s and u , d quarks , since all these baryons have identical orbital structure and permutational symmetries .
the states in the @xmath14-spectrum would be degenerate with the corresponding states in the @xmath16-spectrum which have equal symmetries .
the oscillator parameter @xmath121 and the four integrals are extracted from the mass differences between the nucleon and the @xmath147 , the @xmath11 and the @xmath10 , as well as the splittings between the nucleon and the average mass of the two pairs of states @xmath148 and @xmath149 .
this procedure yields the parameter values @xmath121=157.4 mev , @xmath150=29.3 mev , @xmath151=45.2 mev , @xmath152=2.7 mev and @xmath143=34.7 mev .
given these values , all other excitation energies ( i.e. , differences between the masses of given resonances and the corresponding ground states ) of the nucleon , @xmath8- and @xmath14-hyperon spectra are predicted to within @xmath153 15% of the empirical values where known , and are well within the uncertainty limits of those values .
note that these matrix elements provide a quantitatively satisfactory description of the @xmath14-spectrum even though they are extracted from the @xmath154 spectrum .
the relative magnitudes and signs of the numerical parameter values can be readily understood .
if the potential function @xmath155 is assumed to have the form of a yukawa function with a smeared @xmath156-function term that is positive at short range @xmath157 fm , as suggested by the pion size @xmath158 fm , one expects @xmath152 to be considerably smaller than @xmath150 and @xmath151 , as the radial wavefunction of the excited s - state has a node , and as it extends further into the region where the potential is negative .
the negative value for @xmath143 is also natural , since the corresponding wavefunction is suppressed at short range and extends well beyond the expected node in the potential function .
the model described above has relied on an interaction potential function @xmath87 in ( [ 1.1 ] ) that is flavor independent .
a refined version takes into account the explicit flavor dependence of the potential function in ( [ 5.5 ] ) ( @xmath93 ) . in the following
we show how this explicit flavor dependence provides an explanation of the mass splitting between the @xmath14 and the @xmath16 which have the same quark content and the same fs , f and s symmetries , i.e. , they are degenerate within the @xmath4 version ( [ 1.1 ] ) of the chiral boson exchange interaction . beyond the @xmath4 limit
the ground state baryons will be determined by the @xmath159-exchange radial integral @xmath160 , the k - exchange one , @xmath161 , and by the @xmath91-exchange integrals , @xmath162 , @xmath163 and @xmath164 , where the superscripts indicate quark pairs to which the @xmath91-exchange applies . as indicated by the yukawa interaction ( [ 5.1 ] ) these matrix elements should be inversely proportional to the product of the quark masses of the pair state .
thus @xmath165 we also assume that @xmath166 , which is suggested by the fact that the quark masses are equal in the states in which these interactions act , and by the near equality of the kaon and @xmath91 masses , @xmath167 .
thus we have only two independent radial integrals . to determine the integrals @xmath160 , @xmath161 and the quark mass difference @xmath168 we consider the @xmath169 , @xmath170 and @xmath171 splittings : @xmath172 @xmath173 @xmath174 which imply @xmath161 = 19.6 mev , @xmath175 mev , if the conventional value of 340 mev is given to @xmath176 , @xmath177 mev and the quark mass ratio @xmath178 .
these values of the matrix elements lead to the values of 65 mev and 139 mev for the @xmath179 and the @xmath180 mass differences @xmath181 @xmath182 in good agreement with the empirical values of 77 mev and 125 mev , respectively .
a description of the other parts of the @xmath16 , @xmath103 and @xmath110 spectra can be found in @xcite .
in the previous sections we have shown how the goldstone boson exchange ( gbe ) , taken to first order perturbation theory and without explicit parameterizing the radial dependence , can explain the correct level ordering of positive and negative parity states in light and strange baryon spectra , as well as the splittings in those spectra .
a question , however , arises about what will happen beyond first order perturbation theory . in order to check this
we have numericaly solved three - body faddeev equations @xcite . besides the confinement potential , which is now taken in linear form , the gbe interaction between the constituent quarks
is now included to all orders . these results
further support the adequacy of the gbe for baryon spectroscopy .
in addition to the octet - exchange interaction we include here also the flavor - singlet ( @xmath1 ) exchange . in the large @xmath183 limit
the axial anomaly becomes suppressed @xcite , and the @xmath1 becomes the ninth goldstone boson of the spontaneously broken @xmath184 chiral symmetry in addition to the octet of pseudoscalar mesons @xcite .
+ for the gbe the spin - spin component of the interaction between the constituent quarks @xmath88 and @xmath89 reads : @xmath185 where @xmath186 are flavor gell - mann matrices and @xmath187 . in the simplest case ,
when the boson field satisfies the linear klein - gordon equation , one has the following spatial dependence for the meson - exchange potentials in ( [ 8.1 ] ) : @xmath188 @xmath189 with quark and meson masses @xmath71 and @xmath190 , respectively .
( [ 8.2 ] ) contains both the traditional long - range yukawa potential as well as a @xmath156-function term .
it is the latter that is of crucial importance for baryon physics .
we already discussed in section 5 that it is strictly valid only for pointlike particles , and that it must be smeared out since the constituent quarks and pseudoscalar mesons have finite size , and in addition the boson fields in a chiral lagrangian should in fact satisfy a nonlinear equation . furthermore it is quite natural to assume that at distances @xmath191 , where @xmath192 can be related to the constituent quark and pseudoscalar meson sizes , there is no chiral boson - exchange interaction , since this is the region of perturbative qcd with the original qcd degrees of freedom . the interactions at these very short distances are not essential for the low - energy properties of baryons .
consequently we use a two - parameter `` representation '' for the @xmath156-function term in ( [ 8.2 ] ) @xmath193 following the arguments above one should also cut off the yukawa part of the gbe for @xmath194 . the @xmath195 coupling constant can be extracted from the phenomenological pion - nucleon coupling @xcite as @xmath196 . for simplicity ( and to avoid any additional free parameter ) , the same coupling constant is assumed for the coupling between the @xmath91- meson and the constituent quark .
this is exactly in the spirit of unbroken @xmath4 symmetry . for the flavor - singlet @xmath1 ,
however , we must take a different coupling @xmath197 , as the @xmath1 decouples from the pseudoscalar octet due to the @xmath3 anomaly .
this fact is illustrated best by the failure of the gell - mann oakes renner relations @xcite for the flavor singlet @xcite . lacking a phenomenological value , we treat @xmath198 as a free parameter .
the constituent masses of the u and d quarks are taken to be 340 mev , as suggested by the nucleon magnetic moments . in the present calculation
we neglect tensor meson - exchange forces .
we expect their role to be of minor importance for the main features of the baryon spectra @xcite ( mainly due to the absence of the strong @xmath156-function part in this case ) .
our full interquark potential is thus given by @xmath199 while all masses and the octet coupling constant are predetermined , we treated @xmath192 , @xmath200 , @xmath201 and @xmath202 as free parameters and determined their values to be : @xmath203 notice that we do not need any constant @xmath204 , which is usually added to the confining potential .
in fact only _ four _ free parameters suffice to describe all 14 lowest states of the @xmath27 and @xmath8 spectra , including the absolute value of the nucleon ( ground state ) . at the present stage of determining the qq potential due to gbe
we were led by the principle of working with the smallest possible number of free parameters .
therefore we took the octet coupling constant @xmath205 , and likewise the constituent quark mass @xmath206 , as predetermined .
of course , one may expect that subsequent studies within the gbe model will put further constraints on the parametrization of the @xmath207 potential .
the qq potential ( [ 8.4 ] ) constitutes the dynamical input into our 3-body faddeev calculations of the baryon spectra .
we show our results in fig . 1 for the parametrization of the qq interaction as given above .
it is well seen that the whole set of lowest @xmath27 and @xmath8 states is reproduced quite correctly . in the most unfavourable cases deviations from the experimental values do not exceed 3% !
in addition all level orderings are correct .
in particular , the positive - parity state @xmath10 ( roper resonance ) lies _ below _ the pair of negative - parity states @xmath39 - @xmath112 .
the same is true in the @xmath8 spectrum with @xmath11 and the pair @xmath128 - @xmath129 .
we emphasize again that the qq potential ( [ 8.4 ] ) is able to predict also the absolute value of the nucleon mass . in previous models
an arbitrary constant was usually needed to achieve the correct value of 939 mev . at the present stage of our investigation of the baryon spectra with the gbe interaction
we have left out the tensor forces .
therefore the fine - structure splittings in the @xmath208-multiplets are not yet introduced .
however , it is clear from the observed smallness of these splittings and from the arguments given above , that the tensor component of the gbe can play only a minor role .
here we also note that the yukawa part of the interaction in ( [ 8.2 ] ) is only of secondary importance .
in fact , the pattern of fig . 1 could also be described with the `` @xmath156-part '' ( [ 8.3 ] ) alone ( and a slightly modified set of parameters ) .
it is instructive to learn how the gbe affects the energy levels when it is switched on and its strength ( coupling constant ) is gradually increased ( fig .
2 ) . starting out from the case with confinement
only , one observes that the degeneracy of states is removed and an inversion of the ordering of positive- and negative - parity states is achieved , both in the @xmath27 and @xmath8 excitations . from fig . 2 also
the crucial importance of the chiral interaction @xmath209 becomes evident .
notice that the strength of our confinement , @xmath202= 0.474 @xmath210 , is rather small and the confining interaction contributes much less to the splittings than the gbe .
the relative `` weakness '' of our effective confining interaction could be due to a partial cancellation between the much stronger color - electric confinement and the @xmath53-exchange since they are of opposite sign .
due to the same reason one can not expect that the effective confining interaction between the constituent quarks is strictly of linear form .
a flavor dependent interaction of the form ( [ 1.1 ] ) will imply the presence of an irreducible two - body exchange current operator , as seen , e.g. , directly from the continuity equation , by which the commutator of the interaction and the single particle charge operator is equal to the divergence of the exchange current density @xcite .
the general form of the octet vector exchange current operator , that is associated with the complete octet mediated interaction ( [ 5.5 ] ) , is @xcite @xmath211 here @xmath212 and @xmath213 are dimensionless functions that describe @xmath159 and @xmath214 exchange , respectively , and which include both the pionic ( kaonic ) current and the pair current term at long range .
consider a simplified nonrelativistic constituent quark model .
the impulse approximation expressions for the magnetic moments of the ground state octet baryons and their experimental values are listed in table 3 ( columns `` ia '' and `` exp '' , respectively ) .
a natural approach is to determine the mass ratios @xmath215 and @xmath216 to fit the experimental values of the magnetic moments of the @xmath217 and @xmath218 octet and the @xmath110 and @xmath219 ( @xmath220 , @xmath221 ) decuplet baryons , which are unaffected by the exchange current operator ( [ 9.1 ] ) .
while with only two independent variables it is not possible to fit all four experimental magnetic moments exactly , the best overall fit , @xmath222 @xmath223 @xmath224 @xmath225 , happens to be obtained with precisely the ratios @xmath226 and @xmath227 , which were used for the constituent quark masses to fit ground state baryons ( @xmath228 mev and @xmath229 mev ) .
we find ( see table 3 ) that the meson exchange current contributions systematically improve the predictions of the naive constituent quark model ( i.e. , with one - body quark currents only ) for all known magnetic moments .
.magnetic moments of the baryon octet ( in nuclear magnetons ) .
column ia contains the quark model impulse approximation expressions , column `` exp '' the experimental values , column i the impulse approximation predictions , column ii the exchange current contribution with @xmath230 as the constituent quarks are not too heavy , both their electromagnetic and axial current operators have significant relativistic correction terms .
their effect is to reduce the magnitude of the predicted values of both the axial coupling constants and the magnetic moments of the baryons that are given by the static quark model .
this correction reduces the standard overprediction of the axial current coupling constant of the nucleon ( 5/3 vs 1.24 ) and the strange baryons , but it worsens the mostly satisfactory predictions for the magnetic moments of the baryons that are obtained with the static quark model . in ref .
@xcite it is shown that the exchange current corrections associated with the chiral boson exchange interaction between the quarks can compensate for the relativistic correction in the latter case , while leaving it operative in the case of the axial coupling constants .
this then makes it possible to obtain at least qualitatively satisfactory simultaneous description of both the magnetic moments and the axial coupling constants .
instead of a conclusion we discuss some important recent lattice qcd results in this last section .
it was shown already a few years ago that one can obtain a qualitatively correct splitting between @xmath8 and @xmath27 already within a quenched approximation ( for a review and references see @xcite ) . within the quenched approximation to qcd the sea quark closed loop diagrams generated by gluon lines
are neglected .
thus in the quenched approximation for baryons one takes into account only 3 continuous valence quark lines and full gluodynamics .
this quenched approximation contains , however , part of antiquark effects related to the z graphs formed of valence quark lines .
one can even construct diagrams within the quenched approximation which correspond to the exchange of the color - singlet isospin 1 or 0 @xmath231 pairs between valence quark lines @xcite .
it is also important that these diagrams contribute to the baryon mass to leading order ( @xmath232 ) in a @xmath233 expansion @xcite ( their contribution to the @xmath234 splitting appears , however , to subleading orders ) . from the quenched measurements @xcite it is not clear
what were the physical reason for the @xmath234 splitting : gluon exchanges , instantons , or something else . to clarify this question , liu and dong
have recently measured the @xmath234 splitting in the quenched and a further so - called `` valence approximation '' @xcite . in the valence approximation the quarks are limited to propagating only forward in time ( i.e. , z graphs and related quark - antiquark pairs are removed ) .
the gluon exchange and all other possible gluon configurations , including instantons , are exactly the same in both approximations .
the striking result is that the @xmath234 splitting is observed only in the quenched approximation but not in the valence approximation , in which the @xmath27 and the @xmath8 levels are degenerate within error bars .
consequently the @xmath234 splitting must receive a considerable contribution from the diagrams with @xmath231 excitations , which correspond to the meson exchanges , but not from the gluon exchange or instanton - induced interaction between quarks ( to be precise , the instanton - induced interaction could be rather important for the interactions between quarks and antiquarks ) .
if the observation of liu and dong is confirmed it would be important to measure the relative positions of the lowest excited states of positive and negative parity in the @xmath27 , @xmath8 , @xmath14 and @xmath16 spectra within both the quenched and the valence approximation .
one expects that , if the entire @xmath235 splitting ( i.e. , 300 mev ) is due to the antiquark excitations in the quenched approximation , then @xmath236 should be below the negative parity pair @xmath237 , while in the @xmath14 spectrum the situation should be opposite : the negative parity pair @xmath238 should be below the first positive parity excitation @xmath239 . in the valence approximation
the spin - spin force among quarks , which is due to goldstone boson exchange , is absent , and the relative position of @xmath236 and @xmath237 should be just opposite to the quenched approximation : the @xmath236 should be above the negative parity pair
. however , in the @xmath14 spectrum the negative parity pair @xmath238 should still be below the @xmath239 .
s. weinberg , phys .
d11 * ( 1975 ) 3583 .
g.t hooft , phys .
rev . * d14 * ( 1976 ) 3432
. s. weinberg , physica * 96 a * ( 1979 ) 327 .
a. manohar and h. georgi , nucl .
* b234 * ( 1984 ) 189 .
shuryak , phys.rep .
* c115 * ( 1984 ) 152 .
d. i. diakonov and v. yu .
petrov , phys .
* b147 * ( 1984 ) 351 ; nucl . phys .
* b272 * ( 1986 ) 457 .
glozman and d. o. riska , the baryon spectrum and chiral dynamics , preprint hu - tft-94 - 47 , [ lanl hep - ph 9411279 ] ( 1994 ) ; systematics of the light and strange baryons and the symmetries of qcd , preprint hu - tft-94 - 48 , [ lanl hep - ph 9412231 ] ( 1994 ) .
glozman and d. o. riska , physics reports * 268 * ( 1996 ) 263 .
a. derujula , h. georgi , and s. l. glashow , phys . rev .
* d12 * ( 1975 ) 147 .
n. isgur and g. karl , phys . rev . * d18 * ( 1978 ) 4187 ; * d19 * ( 1979 ) 2653
. b. silvestre - brac and c. gignoux , phys . rev .
* d32 * ( 1985 ) 743 . m. gell - mann , r. j. oakes , and b. renner , phys . rev .
* 175 * ( 1968 ) 2195 .
m. shifman , a. vainstein , and v. zahkarov , nucl . phys .
* b147 * ( 1979 ) 385 , 448 .
d. r. daniel et al , phys . rev .
* d46 * ( 1992 ) 3130 ; m. n. fukugita et al , phys .
* d47 * ( 1993 ) 4739 .
m. gell - mann and m. levy , nuovo cim . * 16 * ( 1960 ) 705 .
y. nambu and g. jona - lasinio , phys .
* 122 * ( 1961 ) 345 .
a. chodos and c. b. thorn , phys .
rev.*d12 * ( 1975 ) 2733 ; g. e. brown and m. rho , phys . lett . *
b82 * ( 1979 ) 177 ; a. w. thomas , adv .
* 13 * ( 1984 ) 1 ; g. a. miller , int .
* 1 * ( 1984 ) 189 .
w. weise , international review of nucl .
phys . , * 1 * ( 1984 ) 116 ; d. robson , topical conference on nuclear chromodynamics , j. qiu and d. sivers ( eds . ) , world scientific , singapore , 174 ( 1988 ) .
d. i. diakonov , v. yu .
petrov , and p. v. pobylitsa , nucl .
* b306 * ( 1988 ) 809 .
s. kahana , v. soni , and g. ripka , nucl . phys .
* a415 * ( 1984 ) 351 .
m. birse and m. banerjee , phys . rev .
* d31 * ( 1985 ) 118 . c. v. christov et al . ,
. phys . * 37 * ( 1996 ) 91 .
r. alkofer and h. reinhard , chiral quark dynamics , springer , 1995 .
u. vogl and w. weise , progr .
* 27 * ( 1991 ) 195 .
s. coleman , j. wess , and b. zumino , phys . rev .
* 177 * ( 1969 ) 2239 .
s. weinberg , phys .
lett . * 18 * ( 1967 ) 188 .
glozman and e. i. kuchina , phys . rev .
* c49 * ( 1994 ) 1149 .
glozman , z. papp , and w. plessas , phys . lett .
* b381 * ( 1996 ) 311 .
e. witten , nucl .
b156 * ( 1979 ) 269 . s. coleman and e. witten , phys .
* 45 * ( 1980 ) 100 .
d. o. riska , phys .
reports * 181 * ( 1989 ) 207 .
k. dannbom , l. ya .
glozman , c. helminen , and d. o. riska , in preparation .
d. weingarten , nucl .
b ( proc . suppl . )
* 34 * ( 1994 ) 29 .
t. d. cohen and d. b. leinweber , comments nucl . part .
* 21 * ( 1993 ) 137 .
g. @xmath240 t hooft , nucl .
* b72 * ( 1974 ) 461 . k. f. liu and s .- j .
dong , quark model from lattice qcd , preprint uk 194 - 03 , hep - lat 9411067 . | beyond the spontaneous chiral symmetry breaking scale light and strange baryons should be considered as systems of three constituent quarks with an effective confining interaction and a chiral interaction that is mediated by the octet of goldstone bosons ( pseudoscalar mesons ) between the constituent quarks . |
one of the best motivated extensions of the standard model consists of introducing extra fermions that are singlets under the standard model gauge group , which we will call for definiteness right - handed ( rh ) neutrinos . with this particle content
, the lagrangian contains a yukawa coupling of the rh neutrinos with the left - handed leptonic doublets and the higgs doublet , which leads to a dirac neutrino mass when the electroweak symmetry is spontaneously broken .
besides , it contains a majorana mass term for the rh neutrinos , which is a priori unrelated to the electroweak symmetry breaking scale .
the most popular scenario in this model consists on assuming that the rh neutrino majorana mass scale is much larger than the electroweak symmetry breaking scale , thus naturally leading to tiny majorana neutrino masses through the renowned see - saw mechanism @xcite .
furthermore , any other low energy effect of the rh neutrinos decouples at least with their mass squared , resulting in a tiny rate for the rare leptonic decays @xcite , a tiny leptonic electric dipole moment @xcite and a tiny deviation of the electroweak observables from the standard model predictions @xcite , in excellent agreement with presently existing experimental data .
it should be borne in mind , however , that the mass of the rh neutrinos is a free parameter which can take any value between zero and the planck scale .
an interesting possibility arises when the rh neutrinos have masses in the range @xmath2 gev .
if this is the case , the new particles could be produced and detected at the large hadron collider ( lhc ) , if their couplings to the standard model particles are sizable @xcite .
this situation , at first sight bizarre in view of the tininess of the neutrino masses , can be realized in some well motivated scenarios , namely when lepton number is approximately conserved ( see , e.g. , @xcite ) .
more importantly , the contributions from the new particles to the low energy phenomena are no longer suppressed , opening the possibility of finding additional low energy signatures of the new physics with experiments at the intensity frontier .
furthermore , the existence of rh neutrinos with masses in the range of @xmath3 gev with lepton number violating interactions can dramatically enhance the rate of neutrinoless double beta decay @xcite , inducing rates that could possibly be observable at gerda @xcite , even when the light neutrinos present a normal hierarchical mass spectrum . in this paper
we analyze the constraints from various experiments on the scenario where the rh neutrinos are accessible to collider searches , as well as the interrelation between the different constraints .
more specifically , we will derive the constraints that follow from the present bounds on the rate of the process @xmath4 , with a special emphasis on the relation to the neutrino mixing and oscillation parameters .
we will also discuss the prospects to observe neutrinoless double beta decay in the next round of experiments , in the view of all present experimental constraints . in section 2
we discuss the formalism and define the parameter space of the theory . in sections 3 and 4
we discuss the constraints on the parameter space which arise from radiative charged lepton decays and the implications for collider searches and electroweak precision observables .
in the subsequent section we perform a detailed analysis of the possible enhancement of the neutrinoless double beta decay rate due to the exchange of the heavy ( rh ) majorana neutrinos . in section 6
we combine the constraints on the parameter space that we obtained in sections 3 , 4 and 5 . finally , we report our conclusions .
we denote the light and heavy majorana neutrinos with definite masses as @xmath5 and @xmath6 , respectively .
the charged and neutral current weak interactions involving the light majorana neutrinos have the form : @xmath7 where @xmath8 is the pontecorvo , maki , nakagawa , sakata ( pmns ) neutrino mixing matrix @xcite , @xmath9 is a @xmath10 unitary matrix which diagonalises the majorana mass matrix of the left - handed ( lh ) flavour neutrinos @xmath11 ( generated by the see - saw mechanism ) , and the matrix @xmath12 characterises the deviations from unitarity of the pmns matrix . the elements of @xmath13 are determined in experiments studying the oscillations of the flavour neutrinos and antineutrinos , @xmath14 and @xmath15 , @xmath16 , at relatively low energies . in these experiments
the initial states of the flavour neutrinos , produced in some weak process , are coherent superpositions of the states of the light massive majorana neutrino @xmath5 only .
the states of the heavy majorana neutrino @xmath17 are not present in the superpositions representing the initial flavour neutrino states and this leads to deviations from unitarity of the pmns matrix .
the charged current and the neutral current weak interaction couplings of the heavy majorana neutrinos @xmath17 to the @xmath18 and @xmath19 bosons read : @xmath20 here @xmath21 is the unitary matrix which diagonalises the majorana mass matrix of the heavy rh neutrinos and the matrix @xmath22 is determined by ( see @xcite ) @xmath23 , @xmath24 and @xmath25 being the neutrino dirac and the rh neutrino majorana mass matrices , respectively , @xmath26 . the matrix @xmath12 which parametrises the deviations from unitarity of the neutrino mixing matrix , can be expressed in terms of the matrix @xmath22 : @xmath27 it is possible to constrain the elements of the hermitian matrix @xmath12 by using the existing neutrino oscillation data and data on electroweak ( ew ) processes @xcite ( e.g. , on @xmath28 decays , invisible @xmath29 decays , universality tests of ew interactions ) . for @xmath30 gev
these constraints read @xcite : @xmath31 the elements of the matrix rv and the masses @xmath32 of the heavy majorana neutrinos @xmath6 should satisfy the approximate constraint on the elements of the majorana mass matrix of the lh flavour neutrinos @xcite , @xmath33 ev , @xmath34 . in the case of the type i see - saw mechanism under discussion this implies : @xmath35 this relation can be satisfied in several situations . the most trivial way to satisfy it
is by imposing that @xmath36 for all @xmath37 and @xmath38 , which renders the observation of all rh neutrinos impossible at the lhc or in any low energy phenomena .
however , this relation can also be fulfilled if one element of the matrix @xmath39 is sizable .
this requires the existence of at least another large matrix element , in order to cancel the large contribution of the former one in eq .
( [ vr1 ] ) . whereas the possibility of cancellations between three different terms can not be precluded , in the simplest case only two terms
will cancel .
then , the large contribution to eq .
( [ vr1 ] ) from one of the rh neutrinos , say @xmath40 with mass @xmath41 , is cancelled by a negative contribution from another rh neutrino , say @xmath42 with mass @xmath43 , provided @xmath44 which is naturally fulfilled if the rh neutrinos @xmath40 and @xmath42 form a pseudo - dirac pair @xcite , i.e. , if there exists an approximately conserved lepton charge ( see , e.g. , @xcite ) . in this scenario
, in order not to spoil the cancellation between these two terms , the contribution from the third neutrino to eq .
( [ vr1 ] ) should be negligible .
therefore in what follows we will work for simplicity in the @xmath45 see - saw scenario , in which the indicated cc and nc weak interaction couplings of @xmath46 are set to zero and @xmath46 is decoupled . in this case
the sm is effectively extended by the addition of 2 rh neutrino fields only . in this class of models ( see , e.g. , @xcite ) one of the three light ( majorana ) neutrinos is massless and hence the neutrino mass spectrum is hierarchical .
two possible types of hierarchical spectrum are allowed by the current neutrino data ( see , e.g. , @xcite ) : @xmath47 normal hierarchical ( nh ) , @xmath48 , @xmath49 and @xmath50 , where @xmath51 and @xmath52 ; + @xmath53 inverted hierarchical ( ih ) , @xmath54 , @xmath55 and @xmath56 , where @xmath51 and @xmath57 . in both cases we have : @xmath58 .
the two heavy neutrino fields satisfy the majorana condition : @xmath59 , @xmath60 .
if the majorana mass matrix @xmath25 of the rh neutrinos is not cp invariant , one can always make the eigenvalues of @xmath25 real and positive , @xmath61 . for @xmath25 respecting the cp symmetry
, the two real eigenvalues of @xmath25 can have the same or opposite signs ( see , e.g. , @xcite ) . with the choice @xmath62 we can always make ( without loss of generality ) , @xmath41 can be positive or negative : @xmath63 and @xmath64 .
one can show , however , that we get the same results for the observables of interest to this study in the two cases .
therefore in what follows we shall work with @xmath61 .
we assume that the heavy rh neutrinos @xmath65 have masses in the range @xmath66 gev , which makes possible , in principle , their production , e.g. , at lhc . in order to be produced with observable rates at lhc , the cc and nc couplings of @xmath67 and @xmath42 in eqs .
( [ ncc ] ) and ( [ nnc ] ) have to be sufficiently large . under this condition
the existing experimental upper bounds on the neutrinoless double beta ( ) decay rate put stringent constraints on the mass spectrum of the rh neutrinos . it can be shown @xcite , in particular , that in the type i see - saw scenario of interest the two heavy ( rh ) majorana neutrinos @xmath65 must be almost degenerate in mass : @xmath68 . if we assume that @xmath69 and @xmath70 , @xmath71 , in order to satisfy the experimental limit on the -decay rate , one should have @xmath72 ( @xmath73 ) for @xmath74 ( @xmath75 ) gev @xcite .
the charged current and the neutral current weak interaction couplings of the heavy majorana neutrinos @xmath17 , @xmath76 , are furthermore constrained by the requirement of reproducing the correct low energy neutrino oscillation parameters after the decoupling of the heavy degrees of freedom .
remarkably , under the condition that @xmath77 are sufficiently large to produce observable effects of the rh neutrinos at low energies , these couplings take a very simple form @xcite .
indeed , the dirac mass matrix @xmath78 in the case under study can be written as @xcite @xmath79 where @xmath80 and @xmath81 is a complex orthogonal matrix . in the scheme with two heavy rh majorana neutrinos
the matrix @xmath81 has the form @xcite : @xmath82 where @xmath83 .
the rh neutrino mixing matrix entering into the cc and nc weak interaction lagrangians ( [ ncc ] ) and ( [ nnc ] ) can be expressed as @xcite @xmath84 the @xmath81-matrix in the case of , e.g. , nh spectrum can be decomposed as follows : @xmath85 one can get a similar expression for the ih spectrum .
the dirac neutrino mass matrix can be decomposed accordingly as @xmath86 , in a self - explanatory notation .
taking for definiteness @xmath87 , it follows that @xmath88 ( @xmath89 ) grows ( decreases ) exponentially with @xmath90 .
, @xmath89 ( @xmath88 ) will grow ( decrease ) exponentially with @xmath90 .
it is possible to show that for sufficiently large values of @xmath91 of interest , the results for the different observables considered in our study do not depend on the choice of the sign of @xmath90 . ] therefore , for sufficiently large @xmath90 it is possible to compensate the huge suppression in eq .
( [ mdfromo ] ) from the tiny observed neutrino masses and the relatively light rh neutrino masses . can not be neglected in the calculation of the majorana mass matrix of the lh flavour neutrinos even though it is exponentially suppressed compared to @xmath88 : the naive approximation @xmath92 leads to @xmath93 , due to @xmath94 ( see @xcite for details ) .
therefore , reproducing the correct light neutrino masses and mixing requires a large amount of fine tuning , unless the rh neutrinos form a pseudo - dirac pair .
demanding @xmath95 and random rh neutrino masses of @xmath96 , for instance , requires a tuning of one part in @xmath97 in order to produce a neutrino mass @xmath98 . ]
we are interested in heavy majorana neutrino couplings to charged leptons and gauge bosons , which are large enough to produce observable low energy signatures . in the limit of `` large '' @xmath90 , the matrix @xmath81 in eq .
( [ opm ] ) can be very well approximated by : @xmath99 for nh spectrum and `` large '' @xmath90 , the matrix @xmath39 takes the form : @xmath100 in a different context similar expressions for the corresponding neutrino yukawa couplings were derived in @xcite in a scheme in which a successful leptogenesis can take place at @xmath101 gev , and in @xcite where a tev scale see - saw model with approximately conserved lepton charge was proposed . in the case of ih light neutrino mass spectrum , the matrix @xmath39 is obtained by replacing @xmath102 and @xmath103 ( @xmath104 ) in eq .
( [ rvpseudo ] ) . for both types of neutrino mass spectrum ,
the elements of the two columns of @xmath39 in the limit of `` large '' @xmath90 are related by the condition : @xmath105 thus recovering eq .
( [ rel0 ] ) . if @xmath25 is cp conserving and @xmath64 , @xmath106 , the expressions for @xmath107 in eq .
( [ rvpseudo ] ) have an additional factor @xmath108 and instead of eq .
( [ rel1 ] ) we have : @xmath109 .
this relation and the relation given in eq .
( [ rel1 ] ) lead to the same expressions for the observables discussed further in the present study .
the overall size of the couplings eq .
( [ rvpseudo ] ) depends crucially on the value of the parameter @xmath90 which has no direct physical interpretation .
it proves convenient to express @xmath90 in terms of the largest eigenvalue @xmath110 of the matrix of neutrino yukawa couplings using the relation : @xmath111 with @xmath112 gev . in terms of @xmath110 and for @xmath113 ,
the heavy majorana neutrino couplings become : @xmath114 where we have used the fact that for the ih spectrum one has @xmath115 . from eqs .
( [ mixing - vs - y ] ) or ( [ mixing - vs - yih ] ) , using the unitarity of the matrix @xmath9 , one can express the neutrino yukawa eigenvalue @xmath110 in ( [ maxy ] ) in terms of heavy neutrino to charged leptons coupling constants : @xmath116 an upper limit on the neutrino yukawa coupling @xmath110 can be derived by assuming the validity of perturbative unitarity .
indeed , the requirement of perturbative unitarity can be easily fulfilled if the see - saw parameter space satisfies the condition : @xmath117 where @xmath118 is the total decay rate of the heavy majorana neutrino @xmath119 , which in the limit @xmath120 is given by the following expression : @xmath121 therefore , taking into account eqs .
( [ ymax2 ] ) and ( [ gammani ] ) and the condition ( [ pertuni ] ) , we get the following upper limit on the parameter @xmath110 from perturbative unitarity : @xmath122
the @xmath124 decay branching ratio in the scenario under discussion is given by @xcite : @xmath125 where @xmath126 is the fine structure constant and @xmath127_{\mu j}^ { * } \,\left[\left(1+\eta\right)u \right]_{e j } g\left(\frac{m_{j}^{2}}{m_{w}^{2}}\right)\nonumber\\ & + & \sum\limits_{k=1}^{2 } \left ( rv\right)_{\mu k}^ { * } \left ( rv\right)_{e k } g\left(\frac{m_{k}^{2}}{m_{w}^{2}}\right)\,.\end{aligned}\ ] ] the loop integration function @xmath128 has the form : @xmath129 it is easy to verify that @xmath128 is a monotonic function which takes values in the interval @xmath130 $ ] , with @xmath131 for @xmath132 . from the definition of the matrix @xmath12 , eq .
( [ eta ] ) , one can write the amplitude @xmath133 as follows : @xmath134 \left [ g(x ) - g(0)\right]\ , , \label{t2}\ ] ] where @xmath135 and we have assumed that the difference between @xmath41 and @xmath43 is negligibly small and used @xmath136 . using eqs .
( [ t2 ] ) and ( [ rel1 ] ) we get : @xmath137 finally , using the expressions of @xmath138 and @xmath139 in terms of neutrino parameters , eqs .
( [ mixing - vs - y ] ) and ( [ mixing - vs - yih ] ) , we obtain the @xmath124 decay branching ratio for the nh and ih spectra : @xmath140^{2}\ , , \label{meg - u - nh } \\\nonumber&&\\ & & { \rm\bf ih:}\;\;\;b(\mu\to e+\gamma)\cong\nonumber\\ & & \frac{3\alpha_{\rm em}}{32\pi } \left(\frac{y^{2}v^{2}}{m_{1}^{2}}\frac{1}{2}\right)^{2 } \left|u_{\mu 2}+iu_{\mu 1 } \right|^{2 } \left|u_{e 2}+iu_{e 1 } \right|^{2 } \left[g(x)-g(0 ) \right]^{2}\ , .
\label{meg - u - ih}\end{aligned}\ ] ] employing the standard parametrisation of the neutrino mixing matrix @xcite it is not difficult to obtain expressions for the factors @xmath141 and @xmath142 , @xmath143 , in terms of the neutrino mixing parameters and the solar and atmospheric neutrino mass squared differences : @xmath144\ , , \label{umu32nh}\\ & & \left| u_{e 2 } + i u_{e 1 } \right|^{2 } = c^2_{13}\left [ 1 + 2 c_{12}\,s_{12}\sin\left(\frac{\alpha_{21}}{2}\right)\right ] \ , , \label{ue21ih } \\ & & \left|u_{\mu 2}+i u_{\mu 1 } \right|^{2 } = c_{23}^2+s_{13}^2\,s_{23}^2 - 2\,c_{12}\,s_{12}\,(c_{23}^2-s_{13}^2\,s_{23}^2 ) \sin\frac{\alpha_{21}}{2}\nonumber\\ & & \hspace{1.4cm}+\,2\,c_{23}\,s_{13}\,s_{23 } \left[s_{12}^2\sin\left(\frac{\alpha_{21}}{2}+\delta\right ) - c_{12}^2\sin\left(\frac{\alpha_{21}}{2}-\delta\right ) \right]\ , , \label{umu21ih}\end{aligned}\ ] ] where @xmath145 , @xmath146 , @xmath147 , @xmath148 and @xmath149 are respectively the solar neutrino , atmospheric neutrino and the chooz angles , @xmath150 is the dirac cp violating phase and @xmath151 and @xmath152 are the two majorana cp violating phases @xcite .
we show in fig .
[ fig1 ] the branching ratio of @xmath124 as a function of the rh neutrino mass @xmath153 , for three different values of the neutrino yukawa coupling eigenvalue : @xmath154 $ ] , blue @xmath155 ( green @xmath156 ) [ red @xmath157 . the figure was obtained by performing a scan of the values of the neutrino mixing angles @xmath147 , @xmath148 and @xmath149 and the solar and atmospheric neutrino mass squared differences @xmath158 and @xmath159 within the corresponding 3@xmath160 bounds ( see table [ tab : mixing2010 ] ) .
the majorana phases @xmath151 and @xmath161 are varied in the interval and @xmath152 enter into the expression for the neutrino mixing matrix in the form @xmath162 and @xmath163 , respectively . ]
@xmath164 $ ] @xcite and the dirac phase @xmath150 is varied in the interval @xmath165 $ ] . shown
are also the current experimental upper limit on the @xmath124 decay branching ratio @xcite , @xmath166 , as well as the prospective limit of the meg experiment @xcite , @xmath167 .
it is apparent from fig .
[ fig1 ] that the data on the process @xmath168 set very stringent constraints on the tev scale see - saw mechanism .
a neutrino yukawa coupling @xmath169 generates a rate for the @xmath124 decay , which is ruled out by the mega experiment , unless the rh neutrino mass is @xmath170 gev .
furthermore , if the meg experiment reaches the sensitivity of @xmath171 without finding a signal , for the same yukawa coupling @xmath169 the rh neutrino mass should be larger than 1 tev . more generally , for @xmath172 ( @xmath173 tev ) and @xmath174 we get the following upper limit on the product @xmath175 of the heavy majorana neutrino couplings to the muon ( electron ) and the @xmath18 boson and to the @xmath19 boson from the current upper limit @xcite on @xmath176 : @xmath177 where we have used eqs .
( [ bmutoeg1 ] ) and ( [ t3b ] ) .
this can be recast as an upper bound on the neutrino yukawa coupling @xmath110 .
taking , @xmath178 the best fit values of the solar and atmospheric oscillation parameters , we get : @xmath179 the upper limit on the neutrino yukawa coupling derived for @xmath180 is reached for @xmath181 ( @xmath182 ) in the case of nh ( ih ) light neutrino mass spectrum . for @xmath183 ,
the upper limit on @xmath110 is obtained for @xmath181 ( @xmath184 ) and @xmath185 ( @xmath186 ) , if the neutrino masses have normal ( inverted ) hierarchy . a possible way to circumvent these stringent upper bounds
consists in assuming a very fine cancellation between the different terms in one of the eqs .
( [ ue32nh])@xmath187([umu21ih ] ) for the factors @xmath141 and @xmath142 , @xmath143 , in the expressions ( [ meg - u - nh ] ) and ( [ meg - u - ih ] ) for @xmath176 .
such a cancellation is possible in the case of nh spectrum ( see also @xcite ) .
indeed , we have @xmath188 if @xmath189 in this case @xmath190 and , thus the upper bound on the yukawa coupling @xmath110 is no longer valid . assuming the neutrino parameters have values within the corresponding present @xmath191 ranges reported in table 1 , we find that the values of @xmath149 implied by the condition in eq .
( [ ue32nh0 ] ) satisfy : @xmath192 , or @xmath193 . these values are outside the 3@xmath160 interval of the experimentally allowed values of @xmath194 ( see table 1 ) .
we consider next the case in which the neutrino mass spectrum is with inverted hierarchy .
the first thing to notice is that the factor @xmath195 can be rather small for @xmath196 since @xmath197 , where we have used the best fit value of @xmath198 . in this case
we have @xmath199 for @xmath200 .
second , we show that , as in the case of normal hierarchical spectrum , it is possible to have a strong suppression of the factor @xmath201 if the chooz mixing angle is close to the corresponding @xmath191 experimental upper bound .
to be more quantitative , we take @xmath202 and @xmath203 .
then , expression ( [ umu21ih ] ) takes the form : @xmath204\ , .
\label{umu21ihb}\end{aligned}\ ] ] it is not difficult to show that , for fixed values of the phases @xmath151 and @xmath150 , @xmath201 has a minimum for @xmath205 at the minimum , using eqs .
( [ umu21ihb ] ) and ( [ s13min ] ) , we get : @xmath206 we will find next for which values of the cp violating phases @xmath150 and @xmath151 this lower bound is equal to zero and if the resulting @xmath149 , obtained from eq . ( [ s13min ] ) , is compatible with the existing limits from the neutrino oscillation data . we have @xmath207 if the dirac and majorana phases @xmath150 and
@xmath151 satisfy the following conditions : @xmath208 and @xmath209 .
taking @xmath210 ( @xmath211 ) and using @xmath212 in eq .
( [ s13min ] ) we get : @xmath213 the solution ( [ s13min2 ] ) is compatible with the @xmath191 upper limit of the chooz mixing angle ( see table 1 ) . in general , one can always find a viable pair of cp violating phases @xmath151 and @xmath150 satisfying the relations given above in order to set the r.h.s . of eq .
( [ umu21ihmin ] ) equal to zero , if the mixing angle @xmath149 is sufficiently large , namely , if @xmath214 .
more precisely , one finds , @xmath178 that @xmath215 for @xmath216 , @xmath217 and @xmath218 .
[ cols="^,^ " , ] + on the majorana phase @xmath219 ( left side ) , @xmath151 ( right side ) for @xmath220 gev , @xmath221 , @xmath222 and @xmath223 .
the gray dash - dotted line shows @xmath224 .
lower panels : @xmath225 versus the degeneracy parameter @xmath226 for @xmath220 gev , @xmath227 , @xmath222 and @xmath228 .
the left ( right ) panels correspond to a nh ( ih ) light neutrino mass spectrum .
[ fig5],title="fig:",width=226,height=75 ]
we now combine the information on the see - saw parameter space that we obtained from the lepton flavour ( @xmath124 ) and lepton number ( @xmath229 ) violating processes studied in the previous sections . in the simple extension of the standard model considered so far , @xmath230 with the addition of two heavy rh neutrinos @xmath67 and @xmath231 at the tev scale , which behave as a pseudo - dirac particle and at the same time are responsible for the generation of neutrino masses via the see - saw mechanism ,
a sizable ( dominant ) contribution of @xmath67 and @xmath231 to the @xmath229-decay rate would imply a `` large '' effect in the muon radiative decay rate .
indeed , if @xmath232 , where @xmath233 is given in eq .
( [ mee3 ] ) , using eqs .
( [ bmutoeg1 ] ) and ( [ t3b ] ) it is easy to show that , given the splitting @xmath234 between @xmath41 and @xmath43 , @xmath232 can be directly related to the @xmath235 decay branching ratio .
more explicitly , we have : @xmath236 where @xmath237 for the nh mass spectrum .
as was pointed out earlier , the corresponding expression for the case of ih spectrum is obtained by replacing @xmath102 and @xmath103 ( @xmath238 ) . for the nh ( ih )
neutrino mass spectrum one has : @xmath239 ( @xmath240 ) .
therefore , for @xmath241 gev and nh spectrum , the predicted rate of the @xmath242 decay can even be larger by up to one order of magnitude than the upper bound on @xmath176 if a positive signal in the current experiments searching for @xmath229-decay is detected , implying @xmath243 ev .
a lower bound on @xmath176 in ( [ bmeg ] ) can be derived for both types of light neutrino mass spectrum . in the case of nh spectrum ,
such lower bound is within the sensitivity of the meg experiment , provided a @xmath229-decay corresponding to @xmath244 ev is observed .
vs @xmath225 for @xmath220 gev , @xmath223 and @xmath47 nh neutrino mass spectrum ( blue dots ) , @xmath53 ih neutrino mass spectrum ( red dots).[fig6],width=510,height=359 ] the analytic relation between @xmath176 and @xmath225 in eq .
( [ bmeg ] ) is confirmed by the results of numerical computation , reported in fig .
the plot shows the correlation between the @xmath124 branching ratio and the effective majorana mass in the case of `` large '' couplings between the rh ( pseudo - dirac pair ) neutrinos and charged leptons in the lagrangian ( [ ncc ] ) .
the effective majorana mass @xmath225 was computed for @xmath223 using the general expression ( [ mee1 ] ) .
the @xmath229-decay nucleus was assumed to be @xmath245ge .
the neutrino oscillation parameters are taken , again , within the corresponding @xmath191 experimental intervals reported in table 1 .
the majorana phase @xmath151 ( @xmath246 ) and the phase @xmath247 in the ih ( nh ) case were varied in the intervals @xmath164 $ ] and @xmath165 $ ] , respectively .
the neutrino yukawa coupling takes values @xmath248 . the correlation between @xmath176 and @xmath232 reported in eq .
( [ bmeg ] ) is satisfied for values @xmath249 .
this is in agreement with figs .
[ fig3 ] and [ fig4 ] , where it is shown that a signal compatible with the gerda sensitivity reach is possible , provided @xmath250 , for both types of neutrino mass spectrum .
moreover , in the case of ih light neutrino mass spectrum , such correlation depends strongly on the value of the majorana phase @xmath151 .
indeed , for @xmath251 gev and @xmath252 we expect that the meg experiment @xcite is able to measure the @xmath124 decay rate ( see fig .
[ fig2 ] ) . if lepton flavour violation is discovered by meg , according to eqs .
( [ mee3 ] ) and ( [ ue21ih ] ) , a positive signal detected by gerda ii , @xmath230 @xmath253 ev , implies : @xmath254 . in the case of @xmath220 gev and @xmath223 ,
used to obtain fig .
[ fig6 ] , we would expect , in general , positive signals to be observed in both meg and gerda ii experiments if @xmath255 ; in the case of @xmath256 , the @xmath229 and @xmath235 decays are predicted to proceed with rates below the sensitivity of these two experiments .
we note , however , that it is not possible to get independent constraints on the degeneracy parameter @xmath226 and the majorana phase from the data on @xmath229 and @xmath235 decays . finally , we notice also that the strong correlation exhibited in fig . [ fig6 ] is a consequence of the constraints imposed by the neutrino oscillation data on the type i see - saw parameter space in the case investigated by us .
we have analyzed the low energy implications of a type i see - saw scenario with right - handed ( rh ) neutrino masses at the electroweak scale and sizable charged and neutral current weak interactions .
this class of scenarios have the attractive feature that the rh neutrinos could be directly produced at the large hadron collider , thus allowing to test in collider experiments the mechanism of neutrino mass generation .
furthermore , and in contrast to the high - scale see - saw mechanism , the rates for the rare leptonic decays are unsuppressed in this scenario , which opens up the possibility of detecting signatures of new physics with experiments at the intensity frontier .
present low energy data set very stringent constraints on this scenario .
namely , reproducing the small neutrino masses requires , barring cancellations , that two of the heavy rh neutrinos must form a pseudo - dirac pair even in the case when there is no conserved lepton charge in the limit of zero splitting at tree level between the masses of the pair . besides , reproducing the experimentally determined values of the low energy neutrino oscillation parameters ( mixing angles and neutrino mass squared differences ) fixes the weak charged current ( cc ) and neutral current ( nc ) couplings of the heavy majorana neutrinos to the charged leptons and @xmath28 and light neutrinos and @xmath257 , @xmath76 ( see eqs .
( [ ncc ] ) and ( [ nnc ] ) ) , up to an overall scale which can be related to the largest eigenvalue @xmath110 of the matrix of neutrino yukawa couplings .
this allows to derive explicit expressions for the rates of the lepton flavour violating ( lfv ) charged lepton radiative decays , @xmath123 , @xmath258 , @xmath259 , in terms of the low energy ( in principle measurable ) neutrino mixing parameters ( including the dirac and majorana cp violating phases ) , the neutrino yukawa coupling @xmath110 and the rh neutrino mass scale . using the present constraint on the rate of the process @xmath123
we have obtained an upper bound on the coupling @xmath110 under the assumption that the rh neutrino mass scale is in the range ( @xmath260 ) gev .
our analysis shows that the restrictions on this scenario from the data on the neutrino mixing parameters and the upper bound on the @xmath123 decay rate imply that the cc and nc couplings @xmath77 of the heavy rh majorana neutrinos are too small to allow their production at the lhc with an observable rate .
other lepton flavour violating processes , such as @xmath261 conversion in nuclei , give complementary constraints on the parameter space of this model , which will be discussed elsewhere @xcite .
we have also analyzed the enhancement of the rate of neutrinoless double beta ( @xmath229- ) decay induced by the rh neutrinos .
we have shown that even after imposing the restrictions on the parameter space implied by the data on the neutrino oscillation parameters and on the lfv charged lepton radiative decays , the contribution due to the exchange of the rh neutrinos in the @xmath229-decay amplitude can substantially enhance the @xmath229-decay rate . as a consequence
, the latter can be in the range of sensitivity of the gerda experiment even when the light neutrinos possess a normal hierarchical mass spectrum .
finally , the rate of the @xmath1 decay , generated by the exchange of the heavy rh neutrinos can naturally be within the sensitivity of the meg experiment .
thus , the observation of @xmath229-decay in the next generation of experiments which are under preparation at present , and of the @xmath1 decay in the meg experiment , could be the first indirect evidence for the tev scale type i see - saw mechanism of neutrino mass generation .
this work was supported in part by the infn program on `` astroparticle physics '' , by the italian miur program on `` neutrinos , dark matter and dark energy in the era of lhc '' , by the world premier international research center initiative ( wpi initiative ) , mext , japan ( s.t.p . ) and by the dfg cluster of excellence `` origin and structure of the universe '' ( a.i . ) .
the work of e.m . is supported by fundao para a cincia e a tecnologia ( fct , portugal ) through the projects ptdc / fis/098188/2008 , cern / fp/116328/2010 and cftp - fct unit 777 , which are partially funded through pocti ( feder ) .
p. minkowski , phys .
b * 67 * ( 1977 ) 421 ; m. gell - mann , p. ramond and r. slansky , _ proceedings of the supergravity stony brook workshop _ , new york 1979 , eds .
p. van nieuwenhuizen and d. freedman ; t. yanagida , _ proceedinds of the workshop on unified theories and baryon number in the universe _ ,
tsukuba , japan 1979 , eds .
a. sawada and a. sugamoto ; r. n. mohapatra and g. senjanovic , phys .
* 44 * ( 1980 ) 912 .
s. m. bilenkii , s. t. petcov and b. pontecorvo , phys .
b * 67 * ( 1977 ) 309 ; t. p. cheng and l. li , phys .
* 45 * ( 1980 ) 1908 ; w. j. marciano and a. i. sanda , phys .
b * 67 * ( 1977 ) 303 ; b. w. lee , s. pakvasa , r. e. shrock and h. sugawara , phys .
* 38 * ( 1977 ) 937 [ erratum - ibid .
* 38 * ( 1977 ) 937 ] . j. p. archambault , a. czarnecki , m. pospelov , phys . rev .
* d70 * ( 2004 ) 073006 .
[ hep - ph/0406089 ] .
p. langacker and d. london , phys .
d * 38 * ( 1988 ) 886 .
e. nardi , e. roulet and d. tommasini , phys .
b * 327 * ( 1994 ) 319 [ arxiv : hep - ph/9402224 ] ; d. tommasini , g. barenboim , j. bernabeu and c. jarlskog , nucl .
b * 444 * ( 1995 ) 451 s. bergmann and a. kagan , nucl .
b * 538 * ( 1999 ) 368 .
f. del aguila , j. de blas , m. perez - victoria , phys . rev . * d78 * , 013010 ( 2008 ) .
[ arxiv:0803.4008 [ hep - ph ] ] .
s. antusch , j. p. baumann and e. fernandez - martinez , nucl .
b * 810 * ( 2009 ) 369 .
s. antusch _ et al .
_ , jhep * 0610 * ( 2006 ) 084 .
t. han , b. zhang , phys . rev .
lett . * 97 * ( 2006 ) 171804 . [ hep - ph/0604064 ] .
f. del aguila , j. a. aguilar - saavedra , r. pittau , jhep * 0710 * ( 2007 ) 047 .
[ hep - ph/0703261 ] .
_ , jhep * 0905 * ( 2009 ) 030 ; f. del aguila , j. a. aguilar - saavedra , nucl .
* b813 * ( 2009 ) 22 - 90 .
[ arxiv:0808.2468 [ hep - ph ] ] .
m. b. gavela , t. hambye , d. hernandez and p. hernandez , jhep * 0909 * ( 2009 ) 038 [ arxiv:0906.1461 [ hep - ph ] ] .
a. halprin , s.t .
petcov and s.p .
rosen , phys .
b * 125 * ( 1983 ) 335 .
j. vergados , nucl . phys .
b * 218 * ( 1983 ) 109 ; w.c .
haxton and j. stephenson , prog .
* 12 * ( 1984 ) 409 ; m. blennow _
_ , arxiv:1005.3240 .
a. ibarra , e. molinaro and s. t. petcov , jhep * 1009 * ( 2010 ) 108 [ arxiv:1007.2378 [ hep - ph ] ] .
b. pontecorvo , zh .
( jetp ) * 33 * ( 1957 ) 549 and * 34 * ( 1958 ) 247 .
z. maki , m. nakagawa and s. sakata , prog .
* 28 * ( 1962 ) 870 .
b. pontecorvo , sov .
jetp * 26 * ( 1968 ) 984 .
[ zh . eksp .
fiz . * 53 * , 1717 ( 1967 ) ] . a. merle and w. rodejohann , phys
. rev .
d * 73 * ( 2006 ) 073012 .
l. wolfenstein , nucl .
b * 186 * ( 1981 ) 147 .
petcov , phys .
b * 110 * ( 1983 ) 245 .
a. kleppe , _ `` extending the standard model with two right - handed neutrinos '' , in * lohusalu 1995 , neutrino physics * , 118 - 125 _ ; e. ma , d. p. roy and u. sarkar , phys .
b * 444 * ( 1998 ) 391 ; p. h. frampton ,
s. l. glashow and t. yanagida , phys .
b * 548 * ( 2002 ) 119 .
m. raidal and a. strumia , phys .
b * 553 * , 72 ( 2003 ) .
v. barger , d. a. dicus , h. j. he and t. j. li , phys .
b * 583 * ( 2004 ) 173 t. endoh , s. kaneko , s. k. kang , t. morozumi and m. tanimoto , phys .
* 89 * ( 2002 ) 231601 .
a. ibarra and g. g. ross , phys .
b * 591 * ( 2004 ) 285 [ arxiv : hep - ph/0312138 ] ; phys .
b * 575 * ( 2003 ) 279 [ arxiv : hep - ph/0307051 ] .
s. t. petcov , w. rodejohann , t. shindou and y. takanishi , nucl .
b * 739 * ( 2006 ) 208 [ arxiv : hep - ph/0510404 ] .
s. m. bilenky and s. t. petcov , rev .
* 59 * ( 1987 ) 671 .
j. a. casas and a. ibarra , nucl .
b * 618 * ( 2001 ) 171 [ arxiv : hep - ph/0103065 ] .
m. raidal , a. strumia and k. turzynski , phys .
b * 609 * ( 2005 ) 351 [ erratum - ibid .
b * 632 * ( 2006 ) 752 ] [ arxiv : hep - ph/0408015 ] .
s. t. petcov , sov .
j. nucl .
* 25 * ( 1977 )
340 [ yad . fiz .
* 25 * ( 1977 ) 641 ] ; s.m .
bilenky , s.t .
petcov and b. pontecorvo , phys .
b * 67 * ( 1977 ) 309 .
t. p. cheng and l. f. li , phys .
* 45 * ( 1980 ) 1908 .
bilenky , j. hosek and s.t .
petcov , phys .
b * 94 * ( 1980 ) 495 . e. molinaro and s. t. petcov , eur .
j. c * 61 * ( 2009 ) 93 [ arxiv:0803.4120 [ hep - ph ] ] .
m. l. brooks _ et al . _
[ mega collaboration ] , phys .
* 83 * ( 1999 ) 1521 [ arxiv : hep - ex/9905013 ] .
a. maki , aip conf .
* 981 * ( 2008 ) 363 .
s. pascoli and s.t .
petcov , phys .
b * 544 * ( 2002 ) 239 ; _ ibid .
_ b * 580 * ( 2004 ) 280 ; phys . rev .
d * 77 * ( 2008 ) 113003 ; see also , e.g. , s.t .
petcov , physica scripta * t121 * ( 2005 ) 94 .
a. ibarra , e. molinaro and s. t. petcov , _ to appear_. | we study a type i see - saw scenario where the right - handed ( rh ) neutrinos , responsible for the light neutrino mass generation , lie at the electroweak scale . under certain conditions ,
the strength of the charged ( cc ) and neutral current ( nc ) weak interactions of the standard model particles with the heavy rh neutrinos can be large enough to allow the production of the latter at the lhc , opening also the possibility of observing other low energy signatures of the new physics in the electroweak precision observables as well as in searches for rare leptonic decays or neutrinoless double beta decay . in this scenario the flavour structure of the indicated cc and nc couplings of the heavy rh neutrinos is essentially determined by the low energy neutrino parameters , leading to fairly strong correlations among the new phenomena . in particular , we show that the present bound on the @xmath0 decay rate makes very difficult the observation of the heavy rh neutrinos at the lhc or the observation of deviations from the standard model predictions in the electroweak precision data . we also show that all present experimental constraints on this scenario still allow i ) for an enhancement of the rate of neutrinoless double beta decay , which thus can be in the range of sensitivity of the gerda experiment even when the light majorana neutrinos possess a normal hierarchical mass spectrum , and ii ) for the predicted @xmath1 decay rate to be within the sensitivity range of the meg experiment . |
the theoretical study of the nonrelativistic heavy quark - antiquark system @xcite and its application to bottomonium @xcite and toponium @xcite physics is of special interest because it relies entirely on the first principles of qcd .
the system in principle allows for a perturbative treatment with the nonperturbative effects being well under control and with no crucial model dependence .
this makes the heavy quarkonium to be an ideal place to determine the fundamental qcd parameters such as the heavy quark mass @xmath0 and the strong coupling constant @xmath1 .
an essential progress has been made in the theoretical investigation of the nonrelativistic heavy quark dynamics based on the effective theory approach @xcite .
the analytical results for the main parameters of the nonrelativistic heavy quark - antiquark system are now available up to next - to - next - to - leading order ( nnlo ) ( see @xcite for a brief review ) .
the nnlo corrections have turned out to be so sizeable that it appears to be indispensable also to gain control over the next - to - next - to - next - to - leading order ( n@xmath2lo ) both in regard of phenomenological applications and in order to understand the structure and the peculiarities of the nonrelativistic expansion .
the calculation of the heavy - quarkonium spectrum to @xmath3 @xcite was the first breakthrough in the n@xmath2lo analysis .
recently , the @xmath4 corrections to the heavy - quarkonium production and annihilation rates have been also obtained @xcite .
the subject of this talk is restricted mainly to the analysis of the spectrum . in the next section
we briefly introduce the basic ingredients of the nonrelativistic effective - theory formalism . in sect .
3 the structure of the third order corrections to the heavy quarkonium ground state energy is outlined and the final result is presented . in the last section
this result is applied for the phenomenological analysis of the top - antitop pair production near the threshold .
the nonrelativistic behavior of the heavy - quark - antiquark pair is governed by a complicated multiscale dynamics . in the nonrelativistic regime , where the heavy - quark velocity @xmath5 is of the order of the strong - coupling constant @xmath1 , the coulomb effects are crucial and have to be taken into account to all orders in @xmath1 .
this makes the use of the effective theory mandatory .
the effective - theory approach allows us to separate the scales and to implement the expansion in @xmath5 at the level of the lagrangian .
let us recall that the dynamics of a nonrelativistic quark - antiquark pair is characterized by four different _ regions _ and the corresponding modes @xcite : ( i ) the hard region ( the energy and three - momentum scale like @xmath0 ) ; ( ii ) the soft region ( the energy and three - momentum scale like @xmath6 ) ; ( iii ) the potential region ( the energy scales like @xmath7 , while the three - momentum scales like @xmath6 ) ; and ( iv ) the ultrasoft region ( the energy and three - momentum scale like @xmath7 ) .
nonrelativistic qcd ( nrqcd ) @xcite is obtained by integrating out the hard modes . subsequently integrating out the soft modes and the potential gluons results in the effective theory of potential nrqcd ( pnrqcd ) @xcite , which contains potential heavy quarks and ultrasoft gluons , ghosts , and light quarks as active particles .
the effect of the modes that have been integrated out is two - fold : higher - dimensional operators appear in the effective hamiltonian , corresponding to an expansion in @xmath5 , and the wilson coefficients of the operators in the effective hamiltonian acquire corrections , which are series in @xmath1 .
the theory of pnrqcd is relevant for the description of the heavy - quarkonium system .
let us recall its basic ingredients . in pnrqcd ,
the ( self)interactions between ultrasoft particles are described by the standard qcd lagrangian .
the interactions of the ultrasoft gluons with the heavy - quark - antiquark pair are ordered in @xmath5 by the multipole expansion . for the n@xmath2lo analysis ,
only the leading - order ( lo ) emission and absorption of ultrasoft gluons have to be considered .
they are described by the chromoelectric dipole interaction .
the propagation of the quark - antiquark pair in the colour - singlet and colour - octet states is described by the nonrelativistic green function of the effective schrdinger equation .
the lo approximation for the green function is given by the coulomb solution , which sums up terms singular at threshold and describes the leading binding effects .
the corrections due to higher - order terms in the effective hamiltonian can be found in rayleigh - schrdinger time - independent perturbation theory as in standard quantum mechanics .
let us now turn to the problem of perturbative calculations in the effective theory . both nrqcd and pnrqcd have specific feynman rules , which can be used for a systematic perturbative expansion .
however , this is complicated because the expansion of the lagrangian corresponds to a particular subspace of the total phase space .
thus , in a perturbative calculation within the effective theory , one has to formally impose some restrictions on the allowed values of the virtual momenta .
explicitly separating the phase space introduces additional scales to the problem , such as momentum cutoffs , and makes the approach considerably less transparent .
a much more efficient and elegant method is based on the expansion by regions @xcite , which is a systematic method to expand feynman diagrams in any limit of momenta and masses .
it consists of the following steps : ( i ) consider various regions of a loop four - momentum @xmath8 and expand , in every region , the integrand in taylor series with respect to the parameters that are considered to be small there ; ( ii ) integrate the expanded integrand over the whole integration domain of the loop momenta ; and ( iii ) put to zero any scaleless integral in step ( ii ) , dimensional regularization , with @xmath9 space - time dimensions , is used to handle the divergences . in the case of the threshold expansion in @xmath5 , one has to deal with the four regions and their scaling rules listed above . in principle
, the threshold expansion has to be applied to the feynman diagrams of full qcd . however , after integrating out the hard modes , which corresponds to calculating the hard - region contributions in the threshold expansion , it is possible to apply step ( i ) to the diagrams constructed from the nrqcd and pnrqcd feynman rules .
equivalently , the lagrangian of the effective theory can be employed for a perturbative calculation without explicit restrictions on the virtual momenta if dimensional regularization is used and the formal expressions derived from the feynman rules of the effective theory are understood in the sense of the threshold expansion . in this way , one arrives at a formulation of effective theory with two crucial virtues : the absence of additional regulator scales and the automatic matching of the contributions from different scales @xcite ( see also@xcite ) .
the second property implies that the contributions of different modes , as computed in the effective theory , can be simply added up to get the full result .
the analysis of the heavy quarkonium spectrum at @xmath10 _
i.e. _ the third order corrections to the coulomb approximation involves two basic ingredients : the effective hamiltonian and the retardation effect associated with the emission and absorption of dynamical ultrasoft gluons .
the general form of the hamiltonian valid up to n@xmath2lo is given in @xcite .
a detailed discussion of the effects resulting from the chromoelectric dipole interaction of the heavy quark - antiquark pair to the ultrasoft gluons relevant for perturbative bound - state calculations at n@xmath2lo can be found in @xcite . for vanishing angular momentum we can write the energy level of the principal quantum number @xmath11 as @xmath12 where the leading order coulomb energy is @xmath13 .
@xmath14 stands for corrections of order @xmath15 .
the first and second order corrections can be found in @xcite for arbitrary @xmath11 .
the @xmath16 corrections to the energy levels arise from several sources : ( i ) matrix elements of the n@xmath2lo operators of the effective hamiltonian between coulomb wave functions ; ( ii ) higher iterations of the nlo and nnlo operators of the effective hamiltonian in time - independent perturbation theory ; ( iii ) matrix elements of the n@xmath2lo instantaneous operators generated by the emission and absorption of ultrasoft gluons ; and ( iv ) the retarded ultrasoft contribution .
the contributions to @xmath17 for vanishing @xmath18-function were computed for general @xmath11 in @xcite . the corresponding correction to the ground state energy
gets contributions from all four sources and reads @xmath19{a_1}\right . }
\nonumber\\ \lefteqn{+\left[-{1\over 36}+\frac{\ln2}{6}+\frac{l_{\alpha_s}}{6}\right ] { c_a^3}+\left[-{49\over36}\right . }
\nonumber\\ \lefteqn{\left.+{4\over3}\left(\ln2+l_{\alpha_s}\right ) \right]{c_a^2c_f}+\left[-{5\over 72}+{10\over 3}\ln2 \right . }
\nonumber\\ \lefteqn{\left.+{37\over6}l_{\alpha_s}+ \left({85\over 54 } -{7\over 6}l_{\alpha_s}\right)s(s+1)\right ] { c_ac_f^2 } } \nonumber\\ \lefteqn{+\left[{50\over9}+{8\over3}\ln2 + 3l_{\alpha_s}-{s(s+1)\over3}\right ] { c_f^3 } } \nonumber\\ \lefteqn{+\left[-{32\over15}+2\ln2+(1-\ln2)s(s+1)\right]{c_f^2t_f } } \nonumber\\ \lefteqn{+{49c_ac_ft_fn_l\over36 } + \left[{11\over18}-{10\over27}s(s+1)\right]{c_f^2t_fn_l } } \nonumber\\ \lefteqn{\left.+{2\over3}c_f^3l^e_1\right\}\ , , } \label{zerobeta}\end{aligned}\ ] ] where the color group invariants are @xmath20 , @xmath21 , @xmath22 , @xmath23 is the number of light - quark flavors , @xmath24 is the spin quantum number and @xmath25 .
the numerical value of the qcd bethe logarithm @xmath26 can be found in @xcite .
@xmath27 scheme for the renormalization of @xmath28 is implied .
the logarithmic @xmath29 part of eq .
( [ zerobeta ] ) has been derived first in @xcite .
the coefficient @xmath30 parameterizes the @xmath31-loop correction to the coulomb potential ( @xmath32 @xcite ) . at present , only pad estimates of the three - loop @xmath27 coefficient @xmath33 are available @xcite . for the bottom and top quark case they read @xmath34 the third order corrections to the ground state energy proportional to the @xmath18-function only get contributions from ( i ) and ( ii ) and read @xmath35l^2_\mu + \left[\left({16\pi^2\over 3}+64\zeta(3)\right)\beta_0 ^ 3\right .
} \nonumber \\ \lefteqn { + { 10}a_1\beta_0 ^ 2 + \left({40}\beta_1+{a_1 ^ 2\over 2}+{a_2}+8\pi^2c_ac_f\right . } \nonumber \\ \lefteqn { \left.\left .
+ \left({21\pi^2\over 2}-{16\pi^2\over 3}s(s+1)\right)c_f^2\right)\beta_0 + { 3}a_1\beta_1\right . }
\lefteqn{+{4\beta_2}\bigg]l_\mu + \left(-{8}+{4\pi^2}+{2\pi^4\over45}+64\zeta(3)\right . }
\nonumber \\ \lefteqn { -{8\pi^2}\zeta(3)+{96}\zeta(5)\bigg)\beta_0 ^ 3 + \left({2\pi^2\over 3}+{8\zeta(3)}\right)a_1\beta_0 ^ 2 } \nonumber \\ \lefteqn { + \left ( \left({8}+{7\pi^2\over 3 } + { 16\zeta(3)}\right)\beta_1 -{a_1 ^
2\over 8}+{3\over 4}a_2+\bigg({6\pi^2}\right . }
\nonumber \\ \lefteqn { \left.-{2\pi^4\over 3}\right)c_ac_f + \left(8\pi^2-{4\pi^4\over 3}+\left(-{4\pi^2\over 3 } + { 4\pi^4\over 9}\right)\right . } \nonumber \\ \lefteqn { \left.\left.\left .
\times s(s+1)\right ) c_f^2\right)\beta_0 + { 2a_1\beta_1}+{4\beta_2 } \right\}\ , , } \label{beta}\end{aligned}\ ] ] where @xmath36 , @xmath37 is riemann s @xmath38-function and @xmath39 is the @xmath40-loop coefficient of the qcd @xmath18 function ( @xmath41 ) .
the terms proportional to @xmath42 in eqs .
( [ beta ] ) have been computed first in @xcite .
the total result for the third order correction to the ground state energy is given by the sum of eqs .
( [ zerobeta ] ) and ( [ beta ] ) adopting the choice @xmath43 one obtains for the bottom and top system ( for s=1 ) in numerical form @xmath44\ , , } \label{eq : dele3num}\end{aligned}\ ] ] where we have separated the contributions arising from @xmath33 and @xmath42 .
the only unknown ingredient in our result for @xmath45 is the three - loop @xmath27 coefficient @xmath33 of the corrections to the static potential entering eq .
( [ zerobeta ] ) . up to
now there are only estimates based on pad approximation which we will use in our analysis .
however , our final result only changes marginally even for a rather large deviations of @xmath33 from its pad estimate .
let us focus on the top quark threshold production .
( the bottom quark case is considered in @xcite ) .
the corresponding experimental data are not yet available and our goal is to present a formula which can be directly used for the top quark mass determination from the characteristics of the cross section of @xmath46 production near threshold @xcite .
one has to distinguish the production in @xmath47 annihilation where the final state quark - antiquark pair is produced with @xmath48 and in ( unpolarized ) @xmath49 collisions where the dominant contribution is given by the @xmath24-wave zero spin final state .
the nonperturbative effects in the case of the top quark are negligible , however , the effect of the top quark width has to be taken into account properly @xcite as the relatively large width smears out the coulomb - like resonances below the threshold .
the nnlo analysis of the cross section @xcite shows that only the ground state pole results in the well - pronounced resonance .
the higher poles and continuum , however , affect the position of the resonance peak and move it to higher energy . as a consequence
, the resonance energy can be written in the form @xmath50 to compute the shift of the peak position due to the nonzero top quark width @xmath51 we use the result of @xcite for the @xmath52 and @xmath53 cross sections at nnlo .
numerically for both processes @xmath54 this value is rather stable with respect to the variation of all input parameters of our analysis . to evaluate the perturbative contribution we use the result of the previous section up to n@xmath2lo . in fig .
[ fig1 ] the perturbative ground state energy in the next - to - leading ( nlo ) , nnlo and n@xmath2lo is plotted as a function of the renormalization scale of the strong coupling constant for @xmath48 .
one can see , that the n@xmath2lo result shows a much weaker dependence on @xmath55 than the nnlo one .
moreover at the scale @xmath56 gev which is close to the physically motivated scale @xmath57 , the n@xmath2lo correction vanishes and furthermore becomes independent of @xmath55 , _
i.e. _ the n@xmath2lo curve shows a local minimum .
this suggests the convergence of the series for the ground state energy _ in the pole mass scheme_. collecting all the contributions we obtain a universal relation between the resonance energy in @xmath46 threshold production in @xmath47 annihilation or @xmath49 collisions ( the result shows very weak spin dependence ) and the top quark pole mass @xmath58 where the weak nonlinear dependence of eq . ( [ ttcorr ] ) on @xmath59 is taken into account in the second term on the right - hand side of eq .
( [ ttcorrnum1 ] ) .
the central value is computed for @xmath60 gev .
the analysis of the theoretical uncertainty in eq .
( [ ttcorrnum1 ] ) can be found in @xcite . due to the very nice behavior of the perturbative expansion for the ground state energy
we do not expect large higher order corrections to our result . as a consequence the top quark pole mass can be extracted from the future experimental data on @xmath46 threshold production with the theoretical uncertainty of @xmath61 mev .
thus we find the n@xmath2lo corrections to stabilize the behavior of the perturbative series for the ground state energy and to be mandatory for a high precision theoretical analysis .
a. pineda and j. soto , phys .
b 420 , ( 1998 ) 391 ; phys .
d 59 , ( 1999 ) 016005 ; a. czarnecki , k. melnikov , and a. yelkhovsky , phys .
rev . a 59 , ( 1999 ) 4316 ; m. beneke , a. signer , and v.a .
smirnov , phys .
b 454 ( 1999 ) 137 .
kniehl and a.a .
penin , phys . rev .
85 ( 2000 ) 1210 , _ ibid .
_ 85 ( 2000 ) 3065 ( e ) ; _ ibid . _ 85 ( 2000 ) 5094 . | in this talk i review the calculation of the third order corrections to the heavy quarkonium spectrum in the nonrelativistic effective theory framework and its application to the phenomenology of top quark threshold production . |
in scalar - tensor theories the gravitational interaction is mediated both by a metric tensor and a scalar field .
the interest in this type of theories of gravity is connected with the expectation that the observed late - time accelerated expansion of the universe may be understood without a dark - energy ( de ) component @xcite .
instead , it is the modified ( compared with einstein s theory ) geometrical sector which is supposed to provide the desired dynamics@xcite .
this may be seen as a geometrization of de .
different aspects of scalar - tensor theories in general or subclasses of them have been investigated in @xcite .
scalar - tensor theories are formulated either in the einstein frame or in the jordan frame .
both frames are related by a conformal transformation .
while matter and scalar field energies are separately conserved in the jordan frame , the dynamics of both components is coupled in the einstein frame for any equation of state ( eos ) different from that of radiation . because of the complex structure of scalar - tensor theories , simple solutions are difficult to obtain , even if the symmetries of the cosmological principle are imposed . hence , in practice , the background expansion rate is usually obtained via numerical integration of the equations of motion . in general , the scalar - tensor - theory based cosmological dynamics may substantially differ from standard cosmology .
our focus here is on the simplest possible extension of the standard @xmath1cdm model that scalar - tensor theory can provide . in this minimalist approach
we remain in the vicinity of the standard model at the present epoch and we aim to quantify the differences between scalar - tensor theory and general relativity ( gr ) by establishing a structure in which the scalar field @xmath0 explicitly enters an analytic solution of the dynamics such that for @xmath2 the standard @xmath1cdm limit is recovered . to this purpose
we construct a simple model which is analytically solved in the einstein frame . with the help of a conformal transformation we then demonstrate how the field @xmath0 , which is given as a certain power of the scale factor , enters the ( jordan - frame ) hubble rate . here
we rely on an effective gr description of the jordan - frame dynamics to determine the geometric equivalent of de . in more detail ,
our starting point is a simple , analytically tractable expression for the coupling between nonrelativistic matter and the ( einstein frame ) scalar field which modifies the standard decay of the matter energy density with the third power of the cosmic scale factor .
this interaction - triggered deviation of the standard decay in the einstein frame is modeled by a power - law behavior in terms of the einstein - frame scale factor . under this condition and
if additionally an effective energy density and an effective pressure , linked by a constant bare " eos parameter , are assumed , an explicit solution of the einstein - frame dynamics is obtained with the mentioned power as an additional parameter .
a straightforward conformal transformation then allows us to obtain the hubble rate and the deceleration parameter in terms of this parameter in the jordan frame as well . for the value zero of such new parameter ,
corresponding to @xmath2 , both frames become indistinguishable and reproduce the dynamics of the standard @xmath1cdm model .
otherwise one has a variable @xmath0 and the dynamics in both frames becomes different , deviating from that of the standard model . the analytic expression for the hubble rate which explicitly clarifies the impact of the scalar field on the cosmological dynamics is the main achievement of this paper .
we shall confront the deviations from the standard model with data from supernovae of type ia ( snia ) , the differential age of old galaxies that have evolved passively ( using @xmath3 , where @xmath4 is the hubble rate and @xmath5 is the redshift parameter ) and baryon acoustic oscillations ( bao ) .
the structure of the paper is as follows . in sec .
[ basic dynamics ] we recall basic general relations for scalar - tensor theories and specify them to the homogeneous and isotropic case . in sec .
[ two - component ] we set up an effective two - component description in the einstein frame , introduce our interaction model and find the einstein - frame hubble rate .
the transformation to the jordan frame is performed in sec .
[ jordan ] where we also discuss the implications of a mapping of the scalar - field degrees of freedom on the effective fluid dynamics in a gr context .
section [ observations ] is devoted to a bayesian statistical analysis on the basis of observational data of snia , @xmath3 and bao . finally , in sec .
[ conclusions ] we summarize our results .
scalar - tensor theories are based on the ( jordan - frame ) action ( see , e.g. , @xcite ) @xmath6 + s_{m}\left(g_{\mu\nu}\right)\ ] ] with a minimally coupled matter part @xmath7 where @xmath8 and @xmath9 denotes the matter lagrangian .
the dynamical field equations are @xmath10 where the energy - momentum tensor of the matter is obtained as usual via @xmath11 the dynamics of the field @xmath0 is dictated by @xmath12 which implies a coupling to the trace @xmath13 of the matter energy - momentum tensor . adopting the conformal transformation @xmath14 of the metric tensor @xmath15 as well as a redefinition of the potential term @xmath16 and @xmath17 one obtains the einstein - frame action @xmath18 + \int d^{4}x\sqrt{-\tilde{g}}\tilde{l}_{m}\left(e^{2b(\varphi)}\tilde{g}_{\mu\nu}\right ) , \
\label{n}\ ] ] in which the matter part is non - minimally coupled to the gravitational sector . throughout the paper ,
quantities without a tilde refer to the jordan frame , quantities with a tilde have their meaning in the einstein frame . restricting ourselves to spatially flat homogeneous and isotropic cosmological models with a robertson - walker metric and assuming a perfect - fluid structure for the energy - momentum tensor of the matter with energy density @xmath19 , pressure @xmath20 and four - velocity @xmath21 , @xmath22 the relevant jordan - frame equations are @xmath23,\ ] ] where @xmath24 is the hubble rate of the jordan frame and @xmath25 is the jordan - frame scale factor , @xmath26,\ ] ] @xmath27 as well as the matter conservation @xmath28 the corresponding relations for the einstein frame are @xmath29 , \label{th2}\ ] ] where @xmath30 is the einstein - frame hubble rate and @xmath31 is the einstein - frame scale factor , as well as @xmath32 , \label{dth}\ ] ] @xmath33 and @xmath34 where we have introduced the notation @xmath35 .
different from the jordan frame , the matter component is not separately conserved here .
the time coordinates and the scale factors of the flrw metrics of both frames are related by @xmath36 respectively . the matter pressure and the matter energy density transform as @xmath37 respectively .
this means @xmath38 , i.e. , the eos parameter of the matter remains invariant .
the einstein - frame equations ( [ th2 ] ) ( [ ddvphi ] ) are of an effective two - component structure in which matter interacts with a scalar field .
we associate an effective energy density @xmath39 and an effective pressure @xmath40 to the scalar field by @xmath41 respectively . equations ( [ drmn ] ) and ( [ ddvphi ] ) can then be written as @xmath42 and @xmath43 respectively , where @xmath44 is the matter eos parameter and @xmath45 is the einstein - frame eos parameter for the scalar field .
notice the difference in notation between the eos parameter @xmath46 and the coupling @xmath47 in the action of the scalar - tensor theory .
the interaction vanishes for the special case @xmath48 . the total energy density @xmath49 and the total pressure @xmath50 are @xmath51 for which the conservation equation @xmath52 holds .
in the special case @xmath53 the solution of ( [ drm+b ] ) can be written as @xmath54 where the function @xmath55 encodes the effects of an interaction between matter and scalar field . comparing ( [ drm+b ] ) and ( [ rhof ] )
, it follows that @xmath56 the absence of an interaction means a constant @xmath55 , equivalent to @xmath57 constant .
writing the time derivative of @xmath58 as @xmath59 eq . ( [ drp+b ] ) becomes @xmath60 for a constant eos parameter @xmath61 the solution of this equation is @xmath62 for a given interaction @xmath63 , the hubble rate ( [ th2 ] ) is then determined by the sum of @xmath64 from ( [ rhof ] ) with ( [ f ] ) and @xmath39 from ( [ rx3 ] ) . to construct an explicitly solvable model we shall assume a power - law behavior of the interaction function @xmath58 ,
@xmath65 clearly , for @xmath66 the function @xmath55 becomes constant and the interaction is absent . with ( [ pansatz ] ) the explicit solution of ( [ rx3 ] ) then is @xmath67 via @xmath68 the interaction term @xmath69 becomes @xmath70 in order to obtain analytical expressions for the dynamics we consider the simple case of a linear dependence @xmath72 where the expression for @xmath73 follows from ( [ phivarphi ] )
this implies also the relations @xmath74 as well as @xmath75 the power @xmath76 is a direct measure of the interaction strength .
the interaction - free case @xmath66 corresponds to @xmath77 constant , i.e. , @xmath78 and , consequently , to @xmath79 , equivalent to the @xmath1cdm model . for @xmath80 one
has @xmath81 while for @xmath82 the opposite , @xmath83 , is valid .
the first equation in ( [ phivarphi+ ] ) relates the scalars @xmath0 and @xmath84 without specifying a potential @xmath85 .
for the special case of an exponential potential , @xmath86 one has , from ( [ phivarphi ] ) , @xmath87 and relation ( [ vu ] ) provides us with the power - law potential @xmath88 with ( [ dvarphi3 m ] ) the balance equation ( [ drp+b ] ) for @xmath39 can be written @xmath89 with an effective eos parameter @xmath90 corresponding to an effective pressure @xmath91 . the interaction modifies the bare " eos parameter @xmath61 . the modification is linear in the interaction parameter @xmath76 .
likewise , the matter balance ( [ drm+b ] ) takes the form of a conservation equation @xmath92 since from ( [ phivarphi+ ] ) one has @xmath93 , it follows that the matter energy density can be written in terms of the scalar field as @xmath94 this exponential structure allows us to write @xmath95 which is of interest in defining an effective potential .
then , eq . ( [ drp+b ] ) ( for @xmath96 ) is equivalent to @xmath97 where @xmath98 now includes the interaction , @xmath99 because of the representation in ( [ rhomphib ] ) , the interaction potential is of the exponential type .
one may introduce effective quantities @xmath100 for which , from ( [ ddphieff ] ) , @xmath101 is valid with @xmath102 for @xmath103 we have @xmath104 as well .
together with ( [ rhomphib ] ) the total energy becomes @xmath105 for the actual value we have @xmath106 .
introducing the fractional quantities @xmath107 the square of the hubble rate is @xmath108\tilde{a}^{-3\left(1+\tilde{w}\right)}\right\}.\ ] ] the non - interacting case @xmath66 corresponds to a @xmath46cdm model . as already mentioned , since then @xmath109 constant , the only possibility here is @xmath79 , i.e. , the @xmath1cdm model . in the following
we shall focus on the case @xmath79 but admitting @xmath76 to be non vanishing . under this condition
@xmath110 consists of a constant part like the @xmath1cdm model but modified by the presence of @xmath76 and a matter part in which the parameter @xmath76 modifies the conventional @xmath111 behavior : @xmath112 these small modifications make the einstein - frame dynamics different from that of the @xmath1cdm model .
( alternatively , this solution may be interpreted in a purely gr context with the @xmath84 component belonging to the matter part of the field equations , interacting with nonrelativistic matter . ) for the deceleration parameter @xmath113 we find @xmath114 } { \tilde{\omega}_{m0}\tilde{a}^{-3\left(1- m\right ) } + \left[1 - m - \tilde{\omega}_{m0}\right]}\,.\ ] ] for @xmath115 its present value consistently reduces to @xmath116 . now , a consistency check can be performed as follows . with ( [ varphitila ] )
we have @xmath117 taking into account @xmath118 results in @xmath119 equivalent to @xmath120 in general , our initial assumption of a constant @xmath61 does not seem to be compatible with the dynamics of @xmath84 in ( [ dvarphik ] ) .
the corrections to the constant value @xmath79 are quadratic in the interaction parameter @xmath76 , there does not appear a term linear in @xmath76 in ( [ tw ] ) .
our approach will therefore be consistent , if we restrict ourselves to @xmath79 and to modifications of the effective equation of state ( [ tilweff ] ) which are linear in @xmath76 .
since @xmath76 quantifies deviations from the @xmath1cdm model , one expects @xmath76 to be small .
with @xmath24 the transformations ( [ ttilb ] ) allow us to establish the relation between the hubble rates of both frames : @xmath121 with @xmath122 and ( [ pansatz ] ) we have @xmath123 i.e. , the power @xmath76 quantifies the difference of the scale factors in both frames . for @xmath66 one
has @xmath124 and the dynamics in both frames reduces to that of the @xmath1cdm model . combining ( [ varphitila ] ) and ( [ ata ] ) yields @xmath125 in terms of the jordan - frame scale factor @xmath25 , @xmath126 for the hubble rate it follows that @xmath127 the matter energy densities in both frames are related by @xmath128 where we have used @xmath129 .
as expected , we recover the characteristic @xmath130 behavior for separately conserved non - relativistic matter in the jordan frame . using ( [ ata ] ) and ( [ tilh2 ] ) in ( [ hth ] ) we obtain the jordan - frame hubble rate square @xmath131\right\ } \,.\ ] ]
obviously , @xmath132 .
this implies @xmath133 .
then , for @xmath134 , @xmath135a^{-\frac{6m}{1 + 3m}},\qquad a\equiv \frac{\left(1 + 3 m \right)^{2}}{1 - m } , \ ] ] or , since @xmath136 the square of the hubble rate can be written as @xmath137\phi .\ ] ] this expression is our main result .
it represents the ( jordan - frame ) hubble rate of our scalar - tensor - theory model . from the structure of ( [ h2phi ] ) it is obvious ,
how the scalar @xmath0 modifies the cosmological dynamics compared with the gr based standard model . for @xmath66 ,
equivalent to @xmath2 , we recover the @xmath1cdm model and the jordan - frame dynamics coincides with the dynamics in the einstein frame . for any @xmath138 ,
equivalent to @xmath139 , the expression ( [ h2phi ] ) ( or ( [ h2 ] ) ) represents a testable , alternative model with presumably small deviations from the @xmath1cdm model .
notice that the modifications of the @xmath1cdm model are different from those in the einstein frame .
in particular , there is no constant part even for @xmath134 .
the corresponding total energy density may be written as @xmath140 or , separating the matter part as in ( [ h2j ] ) @xmath141\phi .\ ] ] the appearance of the scalar @xmath0 in these expressions changes the relative contributions of matter and the dark - energy equivalent .
our model encodes the deviations of the scalar - tensor description from the @xmath1cdm model entirely in the constant parameter @xmath76 which is supposed to be small . to be more definite
we shall assume @xmath142 .
under this condition @xmath0 decays with @xmath25 for @xmath80 while it increases for @xmath82 .
it is only exactly at the present epoch @xmath143 that @xmath144 .
for @xmath80 we have @xmath145 in the past ( @xmath146 ) , for @xmath82 , on the other hand , @xmath0 increases from @xmath147 at @xmath146 to the present @xmath144 . for @xmath148
the energy density approaches @xmath149 in the far - future limit @xmath150 we have @xmath151 depending on the sign of @xmath76 it may either decay ( @xmath80 ) or increase ( @xmath82 ) .
the matter fraction becomes @xmath152\phi}.\ ] ] we may introduce an effective gr description by defining a component @xmath153 , @xmath154a^{-\frac{6m}{1 + 3 m } } + \omega_{m0}a^{-3 } \left[a\,a^{\frac{6m}{1 + 3 m } } - 1\right],\ ] ] equivalent to @xmath155\phi\ ] ] with the fractional contribution @xmath156\phi } { a\phi^{-1}\omega_{m0}a^{-3 } + \left[1 - a\omega_{m0}\right]\phi}.\ ] ] then , at high redshift , @xmath157 for @xmath80 the effective energy density @xmath158 becomes negative for small values of @xmath25 . while the combination @xmath159 remains always positive for small values of @xmath76 and @xmath160 of the order of 0.3 , the combination @xmath161 will change its sign with the increase of the scale factor .
this sign change will occur at a value @xmath162 , given by @xmath163^{\frac{1 + 3m}{6m}}.\ ] ] it is straightforward to check that values @xmath164 drastically change the dynamics compared with that of the standard model .
namely , for @xmath165 the combination @xmath161 is negative for small values of @xmath25 . with increasing @xmath25
one obtains a change to positive values at @xmath166 for @xmath167 , corresponding to a redshift @xmath168 and at @xmath169 ( redshift @xmath170 ) for @xmath171 . for @xmath172
the term @xmath161 is positive for @xmath148 but will change the sign as well . for @xmath173
this happens at @xmath174 ( redshift @xmath175 ) , for @xmath176 the corresponding values are @xmath177 ( @xmath178 ) . the behavior of the fractional energy densities is shown in figs . 1 .
the right panels use logarithmic units showing the asymptotic values of @xmath179 and @xmath180 at early times .
it should be emphasized again that the energy density " @xmath158 does not represent a material substratum , it is of geometric origin .
the fact that the quantity @xmath158 may become negative does not jeopardize the model .
what matters here is the hubble rate which is always well behaved . of the total energy and geometric energy fraction @xmath181 for negative ( top panels ) and
positive ( bottom panels ) values of @xmath76 .
logarithmic units are used in the right panels .
this allows for a better visualization for the asymptotic values at early times .
, title="fig:",scaledwidth=40.0% ] of the total energy and geometric energy fraction @xmath181 for negative ( top panels ) and positive ( bottom panels ) values of @xmath76 .
logarithmic units are used in the right panels .
this allows for a better visualization for the asymptotic values at early times .
, title="fig:",scaledwidth=40.0% ] of the total energy and geometric energy fraction @xmath181 for negative ( top panels ) and positive ( bottom panels ) values of @xmath76 .
logarithmic units are used in the right panels .
this allows for a better visualization for the asymptotic values at early times .
, title="fig:",scaledwidth=40.0% ] of the total energy and geometric energy fraction @xmath181 for negative ( top panels ) and positive ( bottom panels ) values of @xmath76 .
logarithmic units are used in the right panels .
this allows for a better visualization for the asymptotic values at early times .
, title="fig:",scaledwidth=40.0% ] [ figomxn ] to complete the analogy , the component @xmath158 is supposed to obey the conservation equation @xmath182 with an effective eos parameter @xmath183 .
with @xmath184 we obtain @xmath185 a direct calculation yields @xmath186\phi + \omega_{m0}a^{-3}\left[\frac{1+m}{1 + 3m}a \phi^{-1 } - 1\right]}{\left[1 - a\omega_{m0}\right]\phi + \omega_{m0}a^{-3 } \left[a \phi^{-1 } - 1\right]}.\ ] ] equation ( [ wtil-1 ] ) establishes a relation between the constant , bare " eos parameter @xmath134 in the einstein frame and the generally time - dependent effective jordan - frame eos parameter @xmath183 .
for @xmath66 we recover @xmath187 . at high redshift @xmath188}{\left[a \phi^{-1 } - 1\right ] } \qquad\qquad ( a\ll 1).\ ] ] the value of @xmath183 may be close to zero at high redshift , i.e. , mimicking dust , but the effective energy density will be negative for @xmath80 . already a rather small value of @xmath189 will substantially change the behavior of this dark - energy equivalent compared with the standard - model dark energy .
the present value of the effective eos parameter is @xmath190 given that @xmath191 is small , this remains in the vicinity of @xmath192 . in the far future @xmath183 approaches @xmath193 the evolution of the effective eos parameter @xmath183 is visualized in fig .
2 for different values of @xmath76 . for small values of @xmath25 one has @xmath194 but there will be a change to @xmath195 well before the present epoch .
the effective energy density changes from the phantom regime to a later phase with @xmath196 . due to the sign change in @xmath158
the transition point @xmath197 is accompanied by a singularity in @xmath183 at this point for @xmath76 positive .
.,title="fig:",scaledwidth=45.0% ] .,title="fig:",scaledwidth=45.0% ] [ figomxp ] the result for the deceleration parameter @xmath198 is @xmath199\phi } { a\omega_{m0}a^{-3}\phi^{-1 } + \left[1 - a\omega_{m0}\right]\phi}.\ ] ] as shown in fig .
3 , it changes from a high - redshift value @xmath200 close to @xmath201 , to the value close to @xmath202 , @xmath203 at late times .
.,title="fig:",scaledwidth=45.0% ] .,title="fig:",scaledwidth=45.0% ] [ fig5 ] the explicit expression for the pressure is @xmath204\phi + \rho_{0}\omega_{m0}a^{-3}\left[\frac{1+m}{1 + 3m}a\phi^{-1 } -1\right].\ ] ] differentiating and using @xmath205 yields @xmath206\phi + \omega_{m0}a^{-3}\left[\left(\frac{1+m}{1 + 3m}\right)^{2}a\phi^{-1 } -1\right ] \right\}.\ ] ] since @xmath207\phi + \omega_{m0}a^{-3}\left[\frac{1+m}{1 + 3m}a\phi^{-1 } -1\right ] \right\},\ ] ] we find @xmath208\phi + \omega_{m0}a^{-3}\left[\left(\frac{1+m}{1 + 3m}\right)^{2}a\phi^{-1 } -1\right ] } { \frac{2m}{1 + 3m}\left[1 - a\omega_{m0}\right]\phi + \omega_{m0}a^{-3}\left[\frac{1+m}{1 + 3m}a\phi^{-1 } -1\right]}.\ ] ] at high redshift this quantity ( which corresponds to the adiabatic sound speed ) is considerably smaller than 1 , it may even be close to zero .
the far - future limit is @xmath209 while this does not seem to differ substantially from the standard model , the intermediate behavior does . as visualized in fig . 4
, this quantity changes its sign at the points with @xmath210 which implies a singularity in @xmath211 .
recall that the energy density @xmath158 is an effective quantity which simulates de but does not belong to the matter part of the field equations .
it is of entirely geometric origin .
.,title="fig:",scaledwidth=45.0% ] .,title="fig:",scaledwidth=45.0% ] [ fig8 ] we may define a total eos parameter @xmath212 which results in @xmath213\phi + 2 m a\phi^{-1}\omega_{m0}a^{-3}}{\left[1 - a\omega_{m0}\right]\phi + a\phi^{-1}\omega_{m0}a^{-3}}.\ ] ] the limits are @xmath214 and @xmath215 i.e. , at high redshift @xmath46 is close to zero , in the far - future limit it is close to @xmath202 .
the total eos parameter is visualized in figs .
it is well behaved for all values of @xmath76 .
.,title="fig:",scaledwidth=45.0% ] .,title="fig:",scaledwidth=45.0% ] [ figomxp ] with @xmath216\phi,\ ] ] @xmath217\phi\right\ } = - 3h \frac{\rho + p}{\rho_{0}}\ ] ] and @xmath218\ ] ] we find @xmath219\phi + \left(1+m\right)a\phi^{-1}\omega_{m0}a^{-3}}.\ ] ] this effective adiabatic sound speed of the cosmic medium as a whole has the high - redshift limit @xmath220 i.e. , a very small value , and @xmath221 in the far future , a value close to @xmath202 .
while for @xmath80 the derivative @xmath222 is always positive and @xmath223 is always negative , the quantity @xmath224 remains negative in the entire range . for @xmath82 , on the other hand , @xmath222 is always negative but @xmath223 may change its sign for sufficiently large values of @xmath25 , which gives rise to a divergency in @xmath224 at the critical value @xmath225^{\frac{1}{3}\frac{1 - 3|m|}{1+|m|}}\qquad ( m<0).\ ] ] for @xmath226 the ratio @xmath224 is positive , for @xmath227 it is negative since @xmath228 starts to grow at @xmath229 .
this is consistent with the far - future limit in ( [ rhofut ] ) . except for the rather large deviation from the standard model with @xmath173
, these singularities occur in the distant future .
6 shows the dependence of @xmath224 in dependence on the scale factor for various values of m. .,title="fig:",scaledwidth=45.0% ] .,title="fig:",scaledwidth=45.0% ] [ fig12 ]
the purpose of this section is to test the viability of the background expansion predicted by ( [ h2 ] ) , equivalent to ( [ h2phi ] ) , based on the available observations . a particular interest here is to obtain constraints on the parameter @xmath76 which is responsable to dictating deviations of our model from @xmath1cdm . in order to compare ( [ h2 ] ) ( or ( [ h2phi ] ) ) with the data we constrain the model parameters with the following observational data sets . * supernovae * : first , we use supernovae data from the jla compilation @xcite .
we will use the binned data set provided by reference @xcite with the corresponding covariance matrix @xmath230 .
this test is based on the observed distance modulus @xmath231 of each binned sn ia data at a certain redshift @xmath5 , @xmath232 where the luminosity distance , in a spatially flat flrw metric , is given by the formula @xmath233 knowledge of the hubble expansion rate allows us to compute the predicted theoretical value @xmath234 for a given redshift @xmath235 .
the binned jla data contains 31 data - points . for the jla supernova sample the @xmath236 function is constructed according to @xmath237 * differential age * : a second observational source comes from the evaluation of differential age data of old galaxies that have evolved passively @xcite .
indeed , the expansion rate is defined as @xmath238 since spectroscopic redshifts of galaxies are known with very high accuracy , one just needs a differential measurement of time @xmath239 at a given redshift interval in order to obtain values for h(z ) .
the data used in this work consist of @xmath240 data points listed in @xcite , but previously compiled in @xcite .
* baryon acoustic oscillations * : the baryon acoustic oscillation ( bao ) scale is calculated via @xmath241^{1/3},\ ] ] where @xmath242 is the angular - diameter distance . the values for @xmath243 have been reported in the literature by several galaxy surveys . in our analysis
we use data at @xmath244 and @xmath245 from the sdss @xcite , data at @xmath246 and @xmath247 from the wigglez @xcite and one data point at @xmath248 from the 6dfgrs @xcite surveys . for our statistical analysis
we construct the chi - square function of each sample @xmath249 where @xmath250 for the h(z ) and bao datasets , respectively .
the number of data points @xmath251 in each set is , respectively , @xmath252 and @xmath253 whereas @xmath254 is the observational error associated to each observation @xmath255 and @xmath256 is the theoretical value predicted by the scalar - tensor model . adding up information from all data samples we can construct the total chi - square function as @xmath257 we consider the expansion rate ( [ h2 ] ) as a model with three free parameters , @xmath258 and @xmath76 .
our main interest is in constraints on the latter .
we will fix two hypersurfaces for the parameter @xmath259 : the planck prior @xmath260 @xcite and the recent determination from the hubble space telescope ( hst ) @xmath261 @xcite .
these priors on @xmath259 allows us to span the @xmath160 x @xmath76 plane .
this can be seen in the left and central panels in fig .
7 where the @xmath262 and @xmath263 confidence regions are shown in red for the h(z ) data , blue for the supernovae data and green for the bao data , respectively , for both @xmath259 priors .
the combined contours obtained from the total chi - square function ( [ chitotal ] ) are given by the solid black lines .
from these plots it is clear that the preferred @xmath76-values depend on the @xmath259 prior . with the planck prior positive values of @xmath76 are preferred , while the hst prior results in a preference for negative @xmath76 .
it is worth noting that in both two - dimensional plots the @xmath1cdm model , expressed by the horizontal line at @xmath66 , lies outside the @xmath264 confidence level region of the total ( @xmath265 ) function . and display 1@xmath266 and 2@xmath266 confidence level contours in the plane @xmath160 x @xmath76 .
the central panel displays the same contours with the recent hst prior @xmath267 .
h(z ) data only is represented by red regions , bao data in green and supernovae data in blue .
the contours obtained from the total @xmath236 are denoted by the black solid lines . in the right panel
we display the one - dimensional probability distribution function ( pdf ) for the parameter @xmath76 after marginalization over @xmath259 and @xmath160.,title="fig:",scaledwidth=30.0% ] and display 1@xmath266 and 2@xmath266 confidence level contours in the plane @xmath160 x @xmath76 .
the central panel displays the same contours with the recent hst prior @xmath267 .
h(z ) data only is represented by red regions , bao data in green and supernovae data in blue .
the contours obtained from the total @xmath236 are denoted by the black solid lines . in the right panel
we display the one - dimensional probability distribution function ( pdf ) for the parameter @xmath76 after marginalization over @xmath259 and @xmath160.,title="fig:",scaledwidth=30.0% ] and display 1@xmath266 and 2@xmath266 confidence level contours in the plane @xmath160 x @xmath76 .
the central panel displays the same contours with the recent hst prior @xmath267 .
h(z ) data only is represented by red regions , bao data in green and supernovae data in blue .
the contours obtained from the total @xmath236 are denoted by the black solid lines . in the right panel
we display the one - dimensional probability distribution function ( pdf ) for the parameter @xmath76 after marginalization over @xmath259 and @xmath160.,title="fig:",scaledwidth=30.0% ] [ fig1 ] in order to obtain specific constraints on @xmath76 we apply bayesian statistical analysis . with this procedure
we obtain a one - dimensional probability distribution function ( pdf ) after marginalizing the likelihood function @xmath268 over the parameters @xmath259 and @xmath160 .
the resulting pdf for the parameter @xmath76 is shown in the right panel of fig .
although positive values for the parameter @xmath76 are slightly preferred , the peak of the distribution has been obtained at @xmath269 which indicates the agreement of the model with the @xmath1cdm model .
we have established a class of scalar - tensor - theory - based cosmological models which are simple extensions of the @xmath1cdm model . we have discussed in detail the background dynamics of this class .
our main result is an analytic expression ( [ h2phi ] ) for the hubble rate which explicitly quantifies the difference to the standard model through a constant parameter @xmath76 which determines the dynamics of the scalar field @xmath0 .
we identified a geometric equivalent of the de component of the standard model .
the corresponding effective energy density may be positive or negative , including a transition between both regimes .
the effective eos parameter of this component diverges at the transition point but the overall dynamics is well behaved .
a similar comment holds for the effective adiabatic sound speed square of the geometrical de .
such behavior is unavoidable in any model in which the energy density of an de equivalent and its time derivative change their signs during the cosmic evolution .
this seems to limit the usefulness of an effective fluid description of parts of the geometrical sector and may cause computational problems .
but since the overall dynamics and the dynamics of the matter component are smooth , the mentioned apparently unwanted features do not really jeopardize the model .
they just demonstrate the fact that the fluid picture for parts of the geometry is a formal description which may well differ from that of a real fluid .
our tests of the model parameter @xmath76 with data from snia , @xmath3 and bao constrain @xmath76 to values very close to the @xmath1cdm value @xmath66 with a slight preference for positive values .
we expect the existence of the analytic background solution ( [ h2phi ] ) to be useful for future investigation of the perturbation dynamics of this scalar - tensor extension of the @xmath1cdm model .
even a very small non - vanishing value of @xmath191 will certainly modify the standard scenario of structure formation .
+ + * acknowledgement : * partial financial support by cnpq ( brazil ) is gratefully acknowledged .
s. nojiri and s.d .
odintsov , _ where new gravitational physics comes from : m - theory ?
b * 576 * , 5 ( 2003 ) , arxiv : hep - th/0307071 ; _ introduction to modified gravity and gravitational alternative for dark energy _ ,
* 4 * , 115 ( 2007 ) , arxiv : hep - th/0601213 .
s. capozziello , s. nojiri , s.d .
odintsov and a. troisi , _ cosmological viability of f(r)-gravity as an ideal fluid and its compatibility with a matter dominated phase _
, phys.lett.b * 639 * , 135 ( 2006 ) , arxiv : astro - ph/0604431 .
r. jimenez , l. verde , t. treu and d. stern , _ constraints on the equation of sate of dark energy and the hubble constant from stellar ages and the cosmic microwave background _
, astrophys .
j. * 593 * ( 2003 ) 622 .
percival et al . , _ baryon acoustic oscillations in the sloan digital sky survey data release 7 galaxy sample _ , mnras , * 401 * ( 2010 ) 2148 ; n. padmanabhan , x. xu , d.j eisenstein , r. scalzo , a.j .
cuesta , k.t .
mehta , e. kazin , _ a 2% distance to z=0.35 by reconstructing baryon acoustic oscillations - i : methods and application to the sloan digital sky survey _ * 427 * ( 2012 ) 2146 . | we construct a cosmological scalar - tensor - theory model in which the brans - dicke type scalar @xmath0 enters the effective ( jordan - frame ) hubble rate as a simple modification of the hubble rate of the @xmath1cdm model .
this allows us to quantify differences between the background dynamics of scalar - tensor theories and general relativity ( gr ) in a transparent and observationally testable manner in terms of one single parameter .
problems of the mapping of the scalar - field degrees of freedom on an effective fluid description in a gr context are discused .
data from supernovae , the differential age of old galaxies and baryon acoustic oscillations are shown to strongly limit potential deviations from the standard model . |
global analyses do not reliably determine the gluon for @xmath10 at low , yet perturbative , @xmath11 as shown in fig .
[ fig : globalgluons ] .
this is due partly to the lack of precise structure function data for @xmath12 and mainly due to the fact that the data included in global analyses actually probe the quark distribution , while the gluon density is constrained by the @xmath13 dependence of the data , that is by the evolution . in the low @xmath2 region
the available @xmath11 interval decreases and the accuracy of the gluon determination becomes worse . at leading ( lo ) , next - to - leading ( nlo ) and next - to - next - to leading ( nnlo ) order , for the two scales @xmath14 ( left ) and @xmath15 gev@xmath16 ( right panel ) .
lo gluons ( dash dot ) compared are cteq6l @xcite and mrst2004f4lo @xcite .
the two ( long ) dashed lines indicate the error estimate of the cteq6.5 @xcite gluon and the shaded band is the error band for the mrst2001 @xcite global gluon .
central values for the nlo global fits are from cteq6.5 m ( short dashed ) and mrst2004nlo @xcite ( dotted ) .
the solid line represents mrst2006nnlo @xcite.,width=604 ] the strong dependence of the global fits for the gluon on the order of the analysis is clearly demonstrated in fig .
[ fig : globalgluons ] .
note that the recent gluon from the mrst nnlo analysis @xcite receives sizeable corrections both in size and shape compared to the nlo fit , signalling a large uncertainty of the gluon in this regime . in this context
it is also interesting to note that the gluon as obtained from global fits can significantly change , both in normalisation and shape , if small @xmath2 resummations are incorporated into the analysis @xcite .
data for the exclusive @xmath1 process offer an attractive opportunity to determine the low @xmath2 gluon density in this @xmath11 domain , since here the gluon couples _ directly _ to the charm quark and the cross section is proportional to the gluon density _ squared _ @xcite .
therefore the data are much more sensitive to the behaviour of the gluon .
the mass of the @xmath17 vector meson introduces a relatively large scale , amenable to the perturbative qcd ( pqcd ) description not only of large @xmath11 diffractive electroproduction , but also photoproduction of @xmath0 . the available @xmath0 data probe the gluon at a scale @xmath18 in the range @xmath19 gev@xmath16 for @xmath2 in the range @xmath20 ; that is just the domain where other data do not constrain the gluon reliably , see fig .
[ fig : globalgluons ]
. it would be good to have comparable data on exclusive @xmath21 production to determine the gluon at larger scales , but here the available data are sparse , see fig . [
fig : upsilon ] below .
to lowest order the @xmath22 amplitude can be factored into the product of the @xmath23 transition , the scattering of the @xmath17 system on the proton via ( colourless ) two - gluon exchange , and finally the formation of the @xmath24 from the outgoing @xmath17 pair .
the crucial observation is that at high @xmath25 centre - of - mass energy , @xmath26 , the scattering on the proton occurs over a much shorter timescale than the @xmath23 fluctuation or the @xmath24 formation times , see fig .
[ fig : feyndiagram ] .
moreover , at leading logarithmic accuracy , this two - gluon exchange amplitude can be shown to be directly proportional to the gluon density @xmath27 with @xmath28 @xmath11 is the virtuality of the photon and @xmath29 is the rest mass of the @xmath24 . to be explicit ,
the lowest - order formula is @xcite @xmath30 ^ 2 \left(1+\frac{q^2}{m^2_{\j}}\right ) , \label{eq : lo}\ ] ] where @xmath31 is the electronic width of the @xmath0 .
production , @xmath1 .
the factorised form follows since , in the proton rest frame , the formation time @xmath32 is much greater than the @xmath17-proton interaction time @xmath33 . in the case of the simple two gluon exchange shown here , @xmath34 , where @xmath35 is the radius of the proton.,height=188 ] in the leading logarithmic approximation , the integral over the transverse momentum @xmath36 of the @xmath37-channel gluons , see fig .
[ fig : feyndiagram ] , gives rise to the integrated gluon density @xmath38 . as usual in collinear factorisation , the @xmath36 dependence of the integral is completely absorbed in the input gluon distribution ( of the global analyses ) , taken at the factorisation scale @xmath39 .
the integral over the charm quark loop is expressed in terms of the electronic width , @xmath31 , of @xmath0 , and the @xmath40 term in the final brackets reflects the contribution of the longitudinally polarised incoming @xmath41 .
equation ( [ eq : lo ] ) gives the differential cross section at zero momentum transfer , @xmath42 . to describe data integrated over @xmath37 ,
the integration is carried out assuming @xmath43 with @xmath44 the experimentally measured slope parameter . throughout this work the value @xmath45
is used , which is in agreement with @xcite . which is only observed for photoproduction , but which is of the order of differences between measurements of the two experiments h1 and zeus .
the possible uncertainty is smaller than , or at most comparable to , other approximations used . ]
thus it becomes possible to extract the gluon density @xmath38 directly from the measured diffractive @xmath0 cross section .
expression ( [ eq : lo ] ) is a simple , first approximation , justified in the lo collinear approximation using the non - relativistic @xmath0 wave function .
the relativistic corrections were intensively discussed in @xcite .
the problem is that , simultaneously with the relativistic description of the @xmath46 quarks , one needs to consider higher order fock component @xmath47 states of @xmath0 .
hoodbhoy @xcite has studied these two effects to order @xmath48 .
he has shown that relativistic corrections to ( [ eq : lo ] ) , written in terms of the experimentally measured @xmath31 , are small , @xmath49 , see @xcite . so we do not account further for relativistic corrections below .
nlo corrections arise , first , from an explicit integration over the gluon @xmath36 , which goes beyond the leading log contribution , arising from @xmath50 , and , second , from more complicated diagrams with one additional loop . to perform the explicit @xmath36 integration we have to use the unintegrated gluon distribution , @xmath51 , which is related to the integrated gluon by @xmath52 of course
, the infrared contribution can not be treated perturbatively , and so we have introduced a lower limit @xmath53 for the @xmath54 integral in the @xmath0 production amplitude , @xmath55 \:\longrightarrow\ : \int_{q_0 ^ 2}^{(w^2-m_{\j}^2)/4 } \frac{{\rm d}k_t^2\,\alpha_s(k_t^2)}{\bar q^2 ( \bar q^2 + k_t^2 ) } \ , \frac{\partial \left [ xg(x , k_t^2)\right]}{\partial k_t^2 } \ : + \ : \frac{\alpha_s(q_0 ^ 2)}{\bar q^4 } xg(x , q_0 ^ 2)\ , .
\label{eq : nlointegral}\ ] ] expression ( [ eq : nlointegral ] ) replaces the factor @xmath56 in the lo result ( [ eq : lo ] ) . to be precise , the unintegrated distribution @xmath57 embodies the sudakov factor @xmath58 such that @xcite @xmath59/\partial\ln k_t^2\ , .
\label{eq : deftfactor}\ ] ] thus @xmath60 in eq .
( [ eq : unintgluon ] ) is given by @xmath61 and correspondingly for ( [ eq : nlointegral ] ) , @xmath62 . in our numerics
we have chosen @xmath63 . however , in the amplitude ( [ eq : nlointegral ] ) , the dominant contribution comes from the region of @xmath64 where @xmath58 is close to unity . the inclusion of the @xmath65 factor may be considered as an @xmath66 correction to the gluon density and suppresses the gluon in our analysis by 1.7% for photoproduction at @xmath67 .
the contribution coming from @xmath68 is written in terms of the integrated gluon @xmath69 , that is the infrared part is absorbed into the input distribution at the ` transition ' scale @xmath70 .
of course , at low @xmath11 the gluon extracted from a global analysis may be affected by the presence of absorptive corrections which are usually neglected . here
, the absorptive corrections are expected to be smaller .
the transverse size , @xmath71 , of the @xmath72 dipole produced by the ` heavy ' photon in dis has a logarithmic distribution @xmath73 starting from @xmath74 up to some hadronic scale . in the case of @xmath24 production the size of the @xmath17 dipole
is limited by the size of the @xmath24 meson .
even in photoproduction it is of the order of @xmath75 .
since the probability of rescattering is proportional to @xmath76 , we anticipate a much smaller absorptive effect .
a more detailed analysis of the nlo corrections was done in @xcite .
part of these corrections generates the running of @xmath77 , while part is similar to gluon reggeization in the bfkl approach .
indeed , for @xmath24 electroproduction it was shown @xcite using the conventional collinear factorisation scheme , that there is a nlo correction of the form ( ) ( ) .
[ eq : dlog ] in the @xmath36 factorisation approach such a correction may be included by replacing the @xmath37-channel gluon by the reggeized gluon .
however this correction vanishes with a natural choice of the factorisation scale , @xmath63 , which was adopted in our prescription @xcite .
one therefore has reason to believe that the @xmath36 factorisation approach accounts for a major part of the nlo effect , and that the resulting ` nlo ' gluon may be compared to that in a set of nlo global partons .
regularization scheme .
our partons should also be considered to be in the @xmath78 scheme , since we use the @xmath78 definition of @xmath77 , and moreover the factorisation scale which provides the cancellation of the @xmath79 correction is also specified in the @xmath78 scheme . ]
therefore we shall refer to the resulting distributions as nlo gluons .
one also needs to account for the fact that the two gluons exchanged carry different fractions @xmath80 of the light - cone proton momentum , see fig .
[ fig : feyndiagram ] .
that is one has to use the generalised ( skewed ) gluon distribution . )
description this should be considered as a nlo correction to the evolution of @xmath57 . since we do not consider the evolution , but just parametrise the scale dependence of the gluons ( see below ) , this correction is outside our analysis . ] in our case @xmath81 , and the skewing effect can be well estimated from @xcite @xmath82 where @xmath83/\partial\ln x$ ] .
that is in the small @xmath2 region of interest we take the gluon to have the form @xmath84 .
recall that the integral ( [ eq : nlointegral ] ) was written for the discontinuity ( i.e. for the imaginary part ) of the amplitude shown in fig .
[ fig : feyndiagram ] . the real part may be determined using a dispersion relation . in the low @xmath2 region , for our positive - signature amplitude @xmath85 , the dispersion relation can be written in the form @xmath86 both corrections lead to an enhancement of the cross section .
in the following , we present fits to the data for exclusive @xmath24 production from hera using the perturbative description discussed above . in the low @xmath2 region it is expected that the @xmath2 dependence of the gluon density @xmath87 is well approximated by the form @xmath88 .
however , the evolution in @xmath11 modifies this behaviour , enlarging the power @xmath89 as @xmath11 increases .
in particular , in the double leading log ( dll ) approximation , we have the asymptotic form x g ~ ( ) .
thus we need a @xmath11 dependent parametrisation . however , in the limited region of @xmath11 covered by the exclusive @xmath0 data , it is sufficient to use a simple parametric form xg(x,^2 ) = n x^- = a + b(^2/0.45 gev^2 ) .
[ eq : gluonansatz ] the free parameters @xmath90 , @xmath91 and @xmath44 are determined by a non - linear @xmath92 fit to the exclusive @xmath0 data from h1 @xcite and zeus @xcite . , signalling a slight incompatibility between the data
. however , they lead to similar results for the gluon . as the combined fit is very satisfactory ( @xmath93 )
, we will not discuss fits of individual data sets in the following .
] this three - parameter form provides enough flexibility to accurately describe the @xmath2 and @xmath11 behaviour of @xmath24 production in the limited domain covered by the @xmath0 data , namely @xmath94 and @xmath95 , so we will use exactly eq .
( [ eq : gluonansatz ] ) for the lo fit .
however , for the nlo approach , where we have the @xmath54 integral ( [ eq : nlointegral ] ) which runs up to the kinematical limit @xmath96 , we face the problem of a badly convergent @xmath54 integral .
indeed the low @xmath2 gluons @xmath84 obtain a large anomalous dimension @xmath97 . since @xmath98 we get @xmath99 . on the other hand , the expression ( [ eq : nlointegral ] ) was calculated with running @xmath77 .
recall that the scale dependence of the power @xmath89 was generated by the evolution of the form @xmath100 .
accounting for the running @xmath77 , it is natural to replace the second term in the parametric form ( [ eq : gluonansatz ] ) by @xmath101 .
therefore , for the nlo fit , we use xg(x,^2 ) = n x^- = a + b(^2/_qcd^2 ) .
[ eq : gluonansatznlo ] data as described in the text . left panel : cross section compared to some of the h1 @xcite and zeus @xcite data , with values for @xmath102 as indicated ; right panel : gluon compared to global fits for scales as indicated , where the solid ( dashed ) lines are the mrst2004f4lo @xcite ( cteq6l @xcite ) results .
the width of the bands displays the uncertainty of the cross section and fitted gluon respectively , whereas the darker shaded areas indicate the region of the available data.,width=642 ] the gluon densities obtained from the lo fit , using ( [ eq : lo ] ) with ( [ eq : gluonansatz ] ) , together with the skewing and real part corrections , to the exclusive @xmath0 data @xcite are shown in fig .
[ fig : lo ] . , [ fig : nlo ] only a subset of 51 data points used in the fits is displayed , and there are data points at up to @xmath103 gev@xmath16 , corresponding to @xmath104 gev@xmath16 . ]
inclusion of these corrections gives a 22% suppression of the gluon for photoproduction at @xmath105 , with the skewing correction giving the dominant suppression of 18% .
we use a 1-loop running @xmath77 with @xmath106 . in the analysis ,
error bands on the gluon and cross section are generated using the full covariance matrix for the fitted parameters , where as input we have added the statistical and systematic experimental errors of the data in quadrature .
compared to the gluons from the global fits , the gluon from our lo fit is similar in shape , but slightly smaller in normalisation and shows less rise towards smaller @xmath2 with growing scales .
we now present the results of our nlo fit , which we obtain by modifying eq .
( [ eq : lo ] ) by help of ( [ eq : nlointegral ] ) with 2-loop running @xmath77 , and replacing the parametrisation ( [ eq : gluonansatz ] ) by ( [ eq : gluonansatznlo ] ) .
of course we also include skewing and real part corrections as before , and the @xmath65 factor as discussed above .
figure [ fig : nlo ] shows our fit using @xmath108 and @xmath109 .
the dotted lines represent the central values of the gluons obtained using @xmath110 .
the nlo gluon fit shows a better matching to the global gluons at @xmath111 than the lo gluon obtained .
the analysis of the exclusive @xmath24 data indicates that , at the larger scales , the small @xmath2 behaviour of the gluon distribution is slightly flatter than that of the global analyses , both in their @xmath2 behaviour and in their scale dependence of @xmath112 . at low scales ,
the gluon obtained from the @xmath24 data still rises with decreasing @xmath2 , especially in contrast to the mrst fit .
for completeness we also present the values of the parameters for the lo and nlo fits ( tab .
[ parameters ] ) . as can be seen from figs .
[ fig : lo ] , [ fig : nlo ] and the @xmath113 values quoted in tab .
[ parameters ] , our simple ansatz for the form of the gluon , @xmath84 , using an @xmath2 independent power @xmath89 , works very well . to quantify the @xmath2 and @xmath11 behaviour , we tabulate the values of the power of the gluon , @xmath89 , from ( [ eq : gluonansatznlo ] ) at four @xmath11 values , compared to values estimated from mrst2004nlo @xcite and cteq6.5 m gluons @xcite ( tab .
[ lambda ] ) . as is evident from fig .
[ fig : nlo ] and tab .
[ lambda ] , our nlo gluon fit seems to be incompatible with the strength of evolution of the mrst global fit ( e.g. at @xmath67 our gluon increases by a factor @xmath114 from @xmath115 to @xmath116 gev@xmath16 compared to @xmath117 for the mrst2004nlo gluon ) ; however , there is fair agreement in the evolution between our gluon and the cteq fit ( cteq evolves by a factor @xmath118 in the same regime , although in absolute normalisation our nlo @xmath0 prediction for the gluon agrees , on average , much better with the mrst nlo prediction ) .
of course , this should be seen in light of the large uncertainties of both the mrst and cteq gluons at small @xmath2 and scales , see figs .
[ fig : globalgluons ] and [ fig : nlo ] .
this uncertainty persists at the largest scales probed in our fit .
as these scales are rather low , the accuracy of the dglap approach may already be seriously affected by small @xmath2 effects and absorptive and power suppressed corrections .
of course , in the @xmath36 factorisation approach there is some uncertainty arising from the infrared cut - off @xmath70 , below which we can not consider the @xmath36 integration literally .
we have to express this low @xmath36 contribution in terms of the gluon integrated over @xmath119 up to @xmath53 .
however , with our prescription for unintegrated partons , these two contributions match smoothly to each other .
the ambiguity due to the choice of @xmath70 is quite small , as illustrated by the closeness of the dotted lines to our nlo gluons in fig .
[ fig : nlo ] .
note that the difference between the lo and nlo gluons is large at the smallest @xmath2 values , both in the global parton analyses and for the gluons obtained from elastic @xmath24 production . for the global analyses this is due to the absence of the photon - gluon coefficient function at lo . at lo , the photon couples only to the quark parton , which is produced from a gluon at some scale @xmath120 , due to the strong ordering in @xmath36 . at low @xmath2 the gluon grows with @xmath121 , and , to provide the measured values of the proton structure function @xmath122 , we need a much larger gluon distribution in the lo formalism . at nlo
the photon - gluon coefficient function is present , which provides a direct photon - gluon parton coupling at the scale @xmath11 , and , more importantly , a @xmath123 divergence appears in the quark - gluon splitting function , which accelerates the quark evolution and in turn requires less gluon . in elastic @xmath24 production
we have an analogous situation . by carrying out the @xmath36 integration
, we include the interaction with the gluon at large scales of the order of @xmath124 .
moreover , part of this integral has @xmath125 , and may be regarded as one step of backward evolution . in summary , in dglap analyses based on collinear factorisation , the large change in the gluon distribution in going from lo to nlo is due to the strong ordering in @xmath36 and the absence of an additional loop integral in the lo coefficient function and parton evolution .
inclusion of higher - order terms beyond nlo is expected to give a much smaller effect .
they will mainly affect the normalisation and not the @xmath2 dependence of the gluon .
the higher - order corrections in our approach are @xmath126 , but may be enhanced by large logarithms .
the corrections which are enhanced by @xmath127 , and which may lead to some @xmath2 dependence , are absorbed in the form of the gluon distribution by choosing the appropriate factorisation scale @xmath63 , see eq .
( [ eq : dlog ] ) .
then , in our nlo approach , the part of the @xmath36 integral which may be logarithmically large , is accounted for explicitly in the @xmath36 factorisation formalism which is used to obtain the @xmath36 integral .
after this , those higher - order corrections , which are not included in the @xmath36 integral , are concentrated in the domain @xmath128 .
thus the scale dependence of these higher - order corrections is driven mainly by the scale dependence of the running of @xmath77 which we take into account .
the fact that the gluons obtained from our analysis turn out to be close to the gluon distributions coming from the global analyses for @xmath129 , where they are fairly stable , indicates that the omitted higher - order corrections are indeed small .
photoproduction , using our lo and nlo gluon , compared to data .
the dotted lines indicate the error band of the lo prediction , whereas the shaded band is our nlo prediction .
the data points are the published zeus @xcite and h1 @xcite results , and the preliminary results from zeus @xcite.,width=453 ] we have extended our framework to predict elastic @xmath21 photoproduction , using our lo and nlo gluon with our cross section formulae including corrections and @xmath65 factor , see fig .
[ fig : upsilon ] .
although the data is sparse , the cross section predictions are reasonable .
we conclude that this new information coming from our nlo analysis of elastic @xmath24 production data , in which the @xmath130 couple _ directly _ to the low @xmath2 gluon parton and where the cross section is proportional to the _ square _ of the gluon , is especially valuable to constrain the small @xmath2 behaviour of the gluon distribution .
the accuracy of the elastic @xmath24 data is now sufficient , as indicated by the error bands ( arising from the experimental uncertainties ) on the extracted gluon distribution shown in fig .
[ fig : nlo ] , to improve our knowledge of the gluon distribution at small @xmath2 , @xmath131 , considerably .
this is in comparison to the small @xmath2 behaviour of the ` global ' gluon distributions , which is not well constrained by the inclusive dis structure function data .
+ * acknowledgements * we would like to thank robert thorne for valuable discussions .
tt thanks the uk science and technology facilities council for an advanced fellowship .
the work of mr was supported in part by the federal program of the russian ministry of industry , science and technology , rsgss-5788.2006.02 .
x cteq collaboration , j. pumplin _
et al . _ ,
jhep * 0207 * ( 2002 ) 012 .
martin , w.j .
stirling and r.s .
thorne , phys .
lett . * b636 * ( 2006 ) 259 .
cteq collaboration , w .- k .
et al . _ , jhep * 0702 * ( 2007 ) 053 .
martin , r.g .
roberts , w.j . stirling and r.s .
thorne , eur .
j. * c28 * ( 2003 ) 455 .
martin , r.g .
roberts , w.j . stirling and r.s .
thorne , phys .
* b604 * ( 2004 ) 61 .
martin , w.j .
stirling , r.s .
thorne and g. watt , phys .
b652 * ( 2007 ) 292 .
white and r.s .
thorne , phys .
rev . * d75 * ( 2007 ) 034005 .
ryskin , z. phys . *
c37 * ( 1993 ) 89 .
h1 collaboration , a. aktas _
et al . _ ,
j. * c46 * ( 2006 ) 585 .
zeus collaboration , s. chekanov _
et al . _ ,
j. * c24 * ( 2002 ) 345 .
zeus collaboration , s. chekanov _
et al . _ ,
* b695 * ( 2004 ) 3 .
ryskin , r.g .
roberts , a.d .
martin and e.m .
levin , z. phys .
* c76 * ( 1997 ) 231 .
l. frankfurt , w. koepf , m. strikman , phys . rev .
* d54 * ( 1996 ) 3194 .
p. hoodbhoy , phys . rev .
* d56 * ( 1997 ) 388 .
martin , m.g .
ryskin and t. teubner , phys . rev .
* d62 * ( 2000 ) 014022 .
ivanov , m.i .
kotsky and a. papa , eur .
j. * c38 * ( 2004 ) 195 .
ivanov , a. schfer , l. szymanowski and g. krasnikov , eur .
j. * c34 * ( 2004 ) 297 .
levin , a.d .
martin , m.g .
ryskin and t. teubner , z. phys . *
c74 * ( 1997 ) 671 .
martin , m.g .
ryskin and t. teubner , phys . rev .
* d55 * ( 1997 ) 4329 , phys . lett .
* b454 * ( 1999 ) 339 .
k. golec - biernat , a.d .
martin , m.g .
ryskin and a.g .
shuvaev , phys . rev .
* d60 * ( 1999 ) 014015 .
martin , m.g .
ryskin and g. watt , eur .
. j. * c37 * ( 2004 ) 285 .
zeus collaboration , j. breitweg _
et al . _ , z. phys .
* c75 * ( 1997 ) 215 ; eur .
j. * c6 * ( 1999 ) 603 .
zeus collaboration , j. breitweg _
et al . _ ,
b437 * ( 1998 ) 432 .
h1 collaboration , c. adloff _
b483 * ( 2000 ) 23 .
results presented by i. rubinsky at the hep2007 conference in manchester , july 2007 . | exclusive @xmath0 production , @xmath1 , offers a unique opportunity to determine the gluon density of the proton in the small @xmath2 domain .
we use the available hera data to determine the gluon distribution in the region @xmath3 and @xmath4 , where the uncertainty on the gluon extracted from the global parton analyses is large . the gluon density is found to be approximately flat at the lower scale ; it is compared with those of recent global analyses .
ippp/07/50 + dcpt/07/100 + lth 755 + 27 september 2007 + * small @xmath2 gluon from exclusive @xmath0 production * @xmath5 , c. nockles@xmath6 , m. ryskin@xmath7 and t. teubner@xmath8 + @xmath5 _ department of physics and institute for particle physics phenomenology , + university of durham , durham dh1 3le , u.k .
_ + @xmath6 _ department of mathematical sciences , + university of liverpool , liverpool l69 3bx , u.k . _
@xmath9 _ petersburg nuclear physics institute , gatchina , st . petersburg , 188300 , russia _ |
direct - production of two hard photons in nucleon - nucleon collisions ( @xmath9 ) at intermediate energies , unlike production of single photons ( @xmath10 bremsstrahlung ) , belongs to those fundamental processes which are still poorly explored both theoretically and experimentally .
however , a study of this process can provide us with new important information on the underlying mechanisms of the @xmath10 interaction that is complementary to that obtained from investigations of other processes .
in particular , the @xmath9 process can be used as a sensitive probe for experimental verification of the possible existence of @xmath10-decoupled nonstrange dibaryon resonances .
these are two - baryon states @xmath11 with zero strangeness and exotic quantum numbers , for which the strong decay @xmath12 is either strictly forbidden by the pauli principle ( for the states with the isospin @xmath13 and with an even sum @xmath14 , where @xmath15 is the orbital momentum and @xmath16 is the spin ) , or is strongly suppressed by the isospin selection rules ( for the states with @xmath17)@xcite .
such dibaryon states can not be simple bound systems of two color - singlet nucleons , and a proof of their existence would have consequences of fundamental significance for the theory of strong interactions . the @xmath10-decoupled dibaryon states are predicted in a series of @xmath18-inspired models@xcite . among the predicted dibaryons ,
there are those whith masses both below and above the pion production ( @xmath19 ) threshold .
the @xmath10-decoupled dibaryons with masses below the @xmath19 threshold are of special interest , since they may decay mainly into the @xmath20 state , and consequently , their widths should be very narrow(@xmath21 ) .
narrow dibaryon states have been searched for in a number of experiments@xcite , but none has provided any undoubted evidence for their existence . most of the past searches , however , were limited to dibaryons coupled to the @xmath10 channel , and , to our knowledge , only a few ones were dedicated to the @xmath10-decoupled dibaryons@xcite . at the same time , since the latter do not couple to the @xmath10-channel , then , depending on their production and decay modes , they may have escaped detection up to now and may naturally appear in dedicated experiments .
+ if the @xmath10-decoupled dibaryons exist in nature , then the @xmath9 process may proceed , at least partly , through the mechanism that directly involves the radiative excitation @xmath22 and decay @xmath23 modes of these states . in @xmath10 collisions at energies below the @xmath19 threshold ,
these production and decay modes of the @xmath10-decoupled dibaryon resonances with masses @xmath24 would be unique or dominant .
the simplest and clear way of revealing them is to measure the photon energy spectrum of the reaction @xmath9 .
the presence of an @xmath10-decoupled dibaryon resonance would reveal itself in this energy spectrum as a narrow line associated with the formation of this resonance and a relatively broad peak originating from its three - particle decay . in the center - of - mass system , the position of the narrow line(@xmath25 )
is determined by the energy of colliding nucleons ( @xmath26 ) and the mass of this dibaryon resonance as @xmath27 .
an essential feature of the two - photon production in @xmath10 collisions at an energy below the @xmath19 threshold is that , apart from the resonant mechanism in question , there should only be one more source of photon pairs .
this is the double @xmath10-bremsstrahlung reaction . but
this reaction is expected to play a minor role .
indeed , it involves two electromagnetic vertices , so that one may expect that the @xmath9-to-@xmath20 cross section ratio should be of the order the fine structure constant @xmath28 .
however , the cross section for @xmath20 is already small ( for example , the total @xmath29-bremsstrahlung cross section at energies of interest is a few @xmath30 ) .
+ the process @xmath31 at an energy below the @xmath19 threshold provides an unique possibility of searching for the @xmath10-decoupled dibaryon resonances in the mass region @xmath32 with quantum numbers @xmath33 ) or those with @xmath17 ( @xmath34 is the total spin , and @xmath35 is the parity of a dibaryon ) .
the preliminary experimental studies of the reaction @xmath36 at an incident proton energy of about 200 mev@xcite showed that the photon energy spectrum of this reaction had a peculiar structure ranging from about 20 mev to about 60 mev .
this structure was interpreted as an indication of the possible existence of an @xmath10-decoupled dibaryon resonance(later called @xmath6 ) that is produced in the process @xmath7 and subsequently decays via the @xmath8 channel .
unfortunately , a relatively coarse energy resolution and low statistics did not allow us to distinguish the narrow @xmath0-peak associated with the @xmath6 production from the broad @xmath0-peak due to the decay of this resonance and , hence , to determine the resonance mass exactly . to get a rough estimate of the @xmath6 mass
, we admitted that the expected narrow @xmath0-peak is positioned at the center of the observed structure and thus obtained @xmath37 mev .
however , if the uncertainty in the position of the narrow @xmath0-peak in the structure is taken into account , the actual @xmath6 mass might be considerably different from such a crude estimate .
the possible existence of this dibaryon resonance was also probed in the proton - proton bremsstrahlung data taken by the wasa / promice collaboration at the celsius accelerator@xcite for proton energies of 200 and 310 mev .
however , the trigger used in those experimental studies to select events was designed and optimized for investigating the usual @xmath38 bremsstrahlung .
the selected events were those corresponding to two simultaneously detected protons emerging in the forward direction at angles between @xmath39 and @xmath40 with respect to the beam axis , each of which had a kinetic energy exceeding the present threshold ( 40 mev ) .
the analysis of those data resulted only in upper limits on the dibaryon production cross section in the mass range from 1900 to 1960 mev .
therefore , to clarify the situation with the dibaryon resonance @xmath6 , we have decided to measure the energy spectrum of the @xmath41 reaction more carefully .
the experiment was performed using the variable energy proton beam from the phasotron at the joint institute for nuclear research(jinr ) . the schematic layout of the experimental setup is shown in fig . 1 .
the pulsed proton beam ( bursts of about 7 ns fwhm separated by 70 ns ) with an energy of about 216 mev and an energy spread of about 1.5% bombarded a liquid hydrogen target ( @xmath42 ) .
the average beam intensity was on the order of @xmath43 protons / s and was monitored by an ionization chamber ( @xmath44 ) .
the liquid hydrogen target was a cylindrical cryogenic vessel with a length of about 5 cm ( @xmath45 ) and a diameter of 4.5 , cm which had thin entrance and exit windows made from 100 @xmath46 thick mylar foil .
both @xmath0-quanta of the reaction @xmath47 were detected by two @xmath0-ray detectors(@xmath48 and @xmath49 ) placed in a horizontal plane , symmetrically on either side of the beam at a laboratory angle of @xmath3 with respect to the beam direction .
the @xmath0-detector @xmath48 was an array of seven individual detector modules ( see fig .
1 ) : the central detector module ( @xmath50 ) and six detector modules ( @xmath51 ) surrounding the central detector module . each detector module was a 15 cm thick @xmath52 crystal having a hexagonal cross section with an outer diameter of 10 cm .
the detector modules @xmath53 ( ring ) were designed to measure the energy of secondary particles and photons leaving the central detector module @xmath50 .
thus , the energy of each photon detected by the detector @xmath48 was determined by summing up the energy deposited in the central detector module and the energies deposited in the detector modules of the ring . besides , the ring was used as an active shielding from cosmic ray muons .
the energy resolution of a single @xmath52 crystal was measured to be about 13% for @xmath54 mev .
the second @xmath0-ray detector @xmath49 was a cylindrical @xmath55 crystal 15 cm in diameter and 10 cm in length . compared to the detector @xmath48
, it had a poorer energy resolution for @xmath0-rays with energies @xmath56 mev .
the detector @xmath49 was mainly designed to select @xmath4 coincidence events . to reject events induced by charged particles , plastic scintillators ( @xmath57 and @xmath58 ) were put in front of each @xmath0-detector .
both @xmath0-ray detectors were placed in lead houses with walls 10 cm thick which were in turn surrounded by a 5 cm borated parafin shield .
the solid angles covered by the detectors @xmath48 and @xmath49 were 43 msr and 76 msr , respectively . to attenuate prompt particles coming from the target , 15 cm polyethylene bars ( pb ) were placed in between the target and each @xmath0-detector .
+ the electronics associated with the @xmath0-detectors and the plastic scintillators provided amplitude and time signals suitable for digitizing by an @xmath59 and a @xmath60 .
the data acquisition system was triggered by the time signal from the detector module @xmath50 which started all the @xmath61 and opened the @xmath59 gates . in parallel ,
the time signals from the detector module @xmath50 and the @xmath0-detector @xmath49 were sent to a coincidence circuit to select those events which had signals from both the @xmath0-detectors .
events satisfying this criterion were recorded in the event - by - event mode on the hard disk of the computer .
a further selection of @xmath4 coincidence candidate events was done during off - line data processing . the off - line selection procedure for events associated with the process @xmath62 included the following main operations : rejection of events induced by charge particles coming from the target , rejection of events induced by cosmic ray muons , selection of events which are caused by proton beam bursts and selection of events associated with @xmath4 coincidences .
the measurements were performed both with the target filled with liquid hydrogen and with the empty one .
data for the full and empty targets were taken in two successive runs for @xmath6331 h and @xmath6321 h , respectively .
the integrated luminosity of about @xmath64 was accumulated for the measurement with the full target . the resulting number of @xmath4 coincidence events associated with the @xmath62 process as a function of the photon energy measured by the detector @xmath48 is presented in fig .
2 . as can be seen , the energy spectrum of the reaction @xmath62 consists of a narrow peak at a photon energy of about 24 mev and a relatively broad peak in the energy range from about 50 to about 70 mev .
it was found that the statistical significance for the sharp peak as well as for the broad one exceeds 5@xmath5 .
the width ( fwhm ) of the narrow peak was found to be about 8 mev .
this width is comparable with that of the energy resolution of the experimental setup .
the observed behavior of the photon energy spectrum agrees with a characteristic signature of the sought dibaryon resonance @xmath6 that is formed and decays in the radiative process @xmath65 . in that case
the narrow peak should be attributed to the formation of this dibaryon , while the broad peak should be assigned to its three - particle decay . using the value for the energy of the narrow peak @xmath66 mev
, we obtained the @xmath6 mass @xmath67 mev .
having assumed that the @xmath68 process with the @xmath6 mass of 1956 mev is the only mechanism of the reaction @xmath69 , we calculated the photon energy spectra of this reaction for a proton energy of 216 mev .
it was also assumed that the radiative decay of the @xmath6 is a dipole @xmath70 transition from the two - baryon resonance state to a @xmath29-state in the continuum .
the calculations were carried out with the help of monte carlo simulations which included the geometry and the energy resolution of the actual experimental setup .
the photon energy spectra were calculated for two different scenarios of the @xmath6 decay .
the difference between them was that one of these scenarios took into account the final state interaction ( @xmath71 ) of two outgoing protons whereas in the other that interaction was switched off .
each of the scenarios imposed some restrictions on possible quantum numbers of the dibaryon state in question .
the scenario including the @xmath71 implies that the final @xmath29-system is in the singlet @xmath72 state and consequently it should take place , in particular , for the isovector @xmath73 dibaryon state ( the simplest exotic quantum numbers ) , namely , @xmath74 . at the same time , the scenario in which the @xmath71 is switched off , is most likely to occur for the isotensor @xmath75 dibaryon state .
the spectra calculated for these two decay scenarios and normalized to the total number of @xmath4 events observed in the present experiment are shown in fig .
2 . comparison of these spectra with the experimental spectrum indicates that both the calculated spectra are in reasonable agreement with the experimental one within experimental uncertainties .
however , one can see that the energy spectrum for the scenario that can be realized for the isotensor dibaryon(@xmath17 ) is more consistent with the experimental spectrum . at the same time , the statistics of the experiment is insufficient to draw any firm conclusions in favor of one of these scenarios , that is , in favor of either @xmath76 or @xmath17 for the isospin of @xmath6 . in this connection
, we would like to emphasize that the experimental data on the process @xmath77 can shed some light on the properties of the dibaryon state we have observed .
the point is that the @xmath6 dibaryon can in principle be excited in this process via the mode @xmath78 .
however , excitation of the dibaryon @xmath79 with isospin @xmath76 is allowed by the isospin selection rules , whereas excitation of the isotensor @xmath80 resonance is expected to be strongly suppressed@xcite .
therefore , the preliminary results of the study of the reaction @xmath77 carried out by the rmc collaboration at triumf@xcite indicating no enhancement of the two - photon yield in the process @xmath81 compared to the usual one - nucleon mechanism @xmath82 on the proton bound in the deuteron may imply that the @xmath79 dibaryon has isospin 2 .
the @xmath0-ray energy spectrum for the @xmath62 reaction at a proton energy below the pion production threshold has been measured for the first time . the spectrum measured at an energy of about 216 mev for coincident photons emitted at an angle of @xmath3 in the laboratory frame clearly evidences the existence of the @xmath10-decoupled dibaryon resonance @xmath6 with a mass of about 1956 mev that is formed and decays in the process @xmath65 .
the data we have obtained , however , are still incomplete , and additional careful studies of the reaction @xmath83 are needed to get proper parameters ( mass , width , spin , etc . ) of the observed dibaryon state
. 99 s.b.gerasimov and a.s.khrykin,mod.phys.lett.*a8*,2457(1993 ) .
s.b.gerasimov,s.n.ershov,a.s.khrykin,phys.at.nucl.*58*,844(1995 ) .
p.j.g.mlders , a.t.m .
aerts , and j.j .
de swart , phys.rev .
* d 21*,2653(1980 ) .
l.a.kondratyuk , b.v.martemyanov , and m.g.schepkin,sov.j.nucl.phys.*45*,776(1987 ) .
v.b.kopeliovich,phys.at.nucl.*58*,1317(1995 ) .
b.tatischeff et al.,phys.rev.c*59*,1878(1999 ) .
w.brodowski et al.,z.phys.*a355*,5(1996 ) ; r.bilger,nucl.phys.*a629*,141c(1998 ) . l.v.filkov
et al.,phys.rev.c * 61*,044004(2000 ) .
a.s.khrykin,_in proceeding of the xiv international conference on particles and nuclei(panic96 ) , willianisburg , virginia , usa,1996 , _ ed . by carl e.carlson and john j.domingo,p.533 .
a. s. khrykin,@xmath84 newsletter * 13*,250(1997 ) .
h.caln et al.,phys.lett.*b427*,248(1998 ) .
zolnierczuk ( the rmc collaboration ) , acta phys.polon.*b31*,2349(2000 ) .
s.b.gerasimov,in : _ proc . of the 13-th int .
seminar on high energy physics problems `` relativistic nuclear physics and quantum chromodynamics '' _ , edited by a.m. baldin and v.v .
burov , dubna,1998,v.ii , p.3 ; nucl - th/9712064,nucl - th/9812077 , nucl - th/9808070 . | the energy spectrum for high energy @xmath0-rays ( @xmath1 mev ) from the process @xmath2 emitted at @xmath3 in the laboratory frame has been measured at an energy below the pion production threshold , namely , at 216 mev .
the resulting photon energy spectrum extracted from @xmath4 coincidence events consists of a narrow peak at a photon energy of about 24 mev and a relatively broad peak in the energy range of ( 50 - 70 ) mev .
the statistical significance for both the narrow and broad peaks exceeds 5@xmath5 .
this behavior of the photon energy spectrum is interpreted as a signature of the exotic dibaryon resonance @xmath6 which is assumed to be formed in the radiative process @xmath7 followed by its electromagnetic decay via the @xmath8 mode .
the experimental spectrum is compared with those obtained by means of monte carlo simulations .
* study of two - photon production process in proton - proton collisions below the pion production threshold * * a.s.khrykin * + _ joint institute for nuclear research , dubna , 141980 russia + email : [email protected]_ * key - words : * nucleon - nucleon interactions , bremsstrahlung , two - photon production , multiquark states . |
the na49 experiment has collected data on pb+pb collisions at beam energies between 20_a _ - 158_a _ gev with the objective to search for the phase transition to a deconfined phase which is expected to occur in the early stage of the reactions if a sufficient energy density is achieved .
+ several experimental signatures for a qgp have been proposed , among which strange particle yield measurements , in particular , are considered to be a powerful tool .
an enhancement increasing with the strangeness content of the particle was predicted as a consequence of a qgp formation @xcite @xmath2 @xcite . + a non - monotonic energy dependence of the @xmath3/@xmath4 ratio with a sharp maximum close to 30_a
_ gev is observed in central pb+pb collisions . within the statistical model of the early stage @xcite , this is interpreted as a signal of the onset of deconfinement , which causes a sharp change in the energy dependence of the strangeness to entropy ratio .
this observation naturally motivates further study of the production of strange hadrons , in particular hyperons as a function of the beam energy .
+ in this report we show new na49 results on strange particle production in central and minimum bias pb+pb collisions at all available sps energies and compare them to previously shown data @xcite
@xmath2 @xcite at other energies , as well as to model predictions .
the na49 detector @xcite is a large acceptance hadron spectrometer at the cern sps , featuring four large volume tpcs as tracking detectors .
two of them are inside a magnetic field . the ionisation energy loss ( de / dx ) measurements in the tpcs are used for particle identification .
central collisions were selected on the basis of the energy of the projectile spectator nucleons measured by a downstream calorimeter .
[ fig : mtxi ] shows new results on transverse mass spectra of @xmath5 ( left ) and @xmath6 ( right ) integrated over the rapidity range -0.5 @xmath7 y @xmath7 + 0.5 in central pb+pb collisions at different beam energies .
the shape of the spectra is approximately exponential in @xmath8-@xmath9 with an inverse slope parameter of about @xmath10 @xmath11 240 mev and @xmath12 @xmath11 280 mev , and exhibits no significant energy dependence . + ( left ) and @xmath6 ( right ) in central pb+pb ( 7 % most central at 20_a _
@xmath2 80_a _ gev , 10 % most central at 158_a _ gev ) collisions at midrapidity .
the errors shown are statistical only . ]
+ in fig .
[ fig : meanmt ] the energy dependence of the mean transverse mass , @xmath13
@xmath2 @xmath9 , is shown for @xmath1 , @xmath14 , @xmath0 and @xmath15 + @xmath16 .
the mean transverse mass was calculated from the measured spectra and using a fitted function extrapolated to full @xmath8 range .
different extrapolation methods have been tested to estimate systematic error .
the resulting extrapolation factors are small .
+ @xmath2 @xmath9 of @xmath1 , @xmath14 , @xmath0 and @xmath15 + @xmath16 in central pb+pb / au+au collisions from ags @xcite , @xcite to rhic @xcite energies .
open symbols indicate the anti - particles . ]
+ if energy density and hence pressure increase with beam energy , an increase of transverse expansion is expected . assuming that the strength of the transverse expansion is reflected in the mean transverse mass , the variable @xmath13 @xmath2 @xmath9 is expected to rise with @xmath17 .
previously we have reviewed the energy dependence of @xmath13 @xmath2 @xmath9 for pions , kaons and protons @xcite - @xcite . at ags energies
these particles show a strong increase of the mean transverse mass with the beam energy .
however , the mean transverse masses show a weak energy dependence at sps energies . the weak dependence has been attributed @xcite to the constant pressure and temperature when a mixed phase is reached during the dynamical evolution of the colliding system in the early stage of the reaction .
[ fig : meanmt ] shows new results on @xmath13 @xmath2 @xmath9 for heavy strange hadrons . also for these particles no significant energy dependence is observed at sps energies .
measurements of multiply strange hadrons at ags energies are needed to fully study their energy dependence .
the na49 experiment features a large acceptance in the forward hemisphere allowing for measurements of rapidity spectra from midrapidity up to almost beam rapidity .
the rapidity spectra for @xmath0 and @xmath15 in central pb+pb collisions are shown in fig .
[ fig : rap ] .
they are parametrized by a single gaussian function .
na49 results on spectra of @xmath18 , @xmath3 , @xmath19 , @xmath14 , @xmath1 , and @xmath20 were presented earlier @xcite . + and @xmath15 produced in central pb+pb collisions ( 7 % most central at 20_a _
@xmath2 80_a _ gev , 10 % ( @xmath0 ) and 23.5 % ( @xmath15 ) most central at 158_a _ gev ) .
the closed symbols indicate measured points , open points are reflected with respect to midrapidity .
the solid lines represent fits with a single gaussian .
the errors shown are statistical only . ]
+ na49 has measured different particle species at midrapidity in 20_a _ , 30_a _ , 40_a _ ,
80_a _ , and 158_a _ gev central pb+pb collisions .
results on yields at midrapidity in central au+au collisions at @xmath17 = 62.4 , 130 and 200 gev have been published by the star collaboration at rhic @xcite . a compilation of these data is shown in fig . [ fig : midyield ] , together with ags data .
+ = 1.5 ( @xmath18 + @xmath4 ) ) , in central pb+pb / au+au collisions .
the data are compared to string hadronic models @xcite ( urqmd 2.0 : dashed , hsd : solid).,title="fig : " ] = 1.5 ( @xmath18 + @xmath4 ) ) , in central pb+pb / au+au collisions . the data are compared to string hadronic models @xcite ( urqmd 2.0 : dashed , hsd : solid).,title="fig : " ] + one observes a relatively weak energy dependence of the yield of hyperons starting from sps to rhic energies , while a strong increase is seen for the anti - hyperons .
the data on @xmath1 and kaon yields are compared to string - hadronic models @xcite .
both models follow the trend of the measured data , but they fail to reproduce the data quantitatively .
predictions for multi - strange hyperons and the @xmath14-meson are still not available .
+ + figs .
[ fig : particle / pi_yield_midrap_ratio ] and [ fig : particle / pi_yield_4pi_ratio ] show the energy dependence of relative strange particle yields .
the ratio of the midrapidity and total yields normalized to the respective pion yields are compared to model calculations .
while @xmath21@xmath19@xmath22/@xmath21@xmath18@xmath22 , @xmath21@xmath20@xmath22/@xmath21@xmath23@xmath22 , @xmath21@xmath6@xmath22/@xmath21@xmath23@xmath22 , @xmath21@xmath24@xmath22/@xmath21@xmath23@xmath22 and @xmath21@xmath14@xmath22/@xmath21@xmath23@xmath22 ratios rise continuously with energy , a distinct maximum is visible in the energy dependence of the ratios @xmath21@xmath3@xmath22/@xmath21@xmath4@xmath22 , @xmath21@xmath1@xmath22/@xmath21@xmath23@xmath22 and @xmath21@xmath5@xmath22/@xmath21@xmath23@xmath22 .
the same structure is seen for the midrapidity ratios .
hadron gas @xcite ( only available for 4@xmath23 multiplicity ratios ) and microscopic models @xcite do not provide a proper description of the structures observed in the data . on the other hand
, this feature can be understood in a scenario where the onset of deconfinement is reached around 30_a _ gev as proposed in the statistical model of the early stage @xcite .
+ @xmath23@xmath22 = 1.5 ( @xmath21@xmath18@xmath22 + @xmath21@xmath4@xmath22 ) , in central pb+pb / au+au collisions .
the data are compared to string hadronic models @xcite ( urqmd 2.0 : dashed , hsd : solid ) and statistical hadron gas model @xcite ( assuming strangeness in full equilibrium : dotted ) . , title="fig : " ] @xmath23@xmath22 = 1.5 ( @xmath21@xmath18@xmath22 + @xmath21@xmath4@xmath22 ) , in central pb+pb / au+au collisions . the data are compared to string hadronic models @xcite ( urqmd 2.0 : dashed , hsd : solid ) and statistical hadron gas model @xcite ( assuming strangeness in full equilibrium : dotted ) . , title="fig : " ] + the na49 antibaryon / baryon ratios are shown in fig .
[ fig : antibaryon / baryon_ratio ] as a function of the beam energy .
the ratios increase with increasing strangeness content of the hyperons at a given energy .
the energy dependence of the ratio for multiply strange hyperons is much weaker than that for protons and @xmath1 . because the net baryon density at midrapidity is a strong function of the beam energy , the antibaryon to baryon ratio strongly depends on the number of valence up and down quarks in the hadron . for @xmath15 with no valence up / down quarks the energy dependence of the ratio is the weakest .
the open symbols represent measurements from the na57 collaboration @xcite .
while the particle ratios are in good agreement with the na49 data , the absolute yields of na57 tend to be systematically higher by about 30 % , in central pb+pb collisions both at 40_a _ and 158_a _ gev @xcite .
/@xmath25 ) at midrapidity from sps to rhic energies .
the errors shown are statistical only . ]
[ fig : strangeness_enhancement ] shows results on system size and centrality dependence of hyperon yields at midrapidity at 158_a _ gev .
having measured particle yields for p+p , c+c , si+si and pb+pb collisions , one can determine a hyperon enhancement as : + @xmath26 where @xmath27 is the number of wounded nucleons calculated within the glauber model @xcite . a significant enhancement for the @xmath5
is observed at 158_a _ gev when going from p+p to pb+pb . for the @xmath1 hyperon
the enhancement is not as strong as for the @xmath5 hyperon and the centrality dependence is slightly weaker .
together with the results for c+c and si+si @xcite a complete coverage of the system size range from @xmath27 @xmath28 14 is possible .
+ + the na57 collaboration @xcite observes the same trend ( not shown ) but instead of p+p a p+be reference is used . since already in p+a reactions a slight enhancement of strange particle production is observed @xcite , the enhancement relative to p+be is less . + + particle yields up to @xmath29 = 4
gev / c have been measured in central pb+pb collisions .
the baryon / meson ratios qualitatively show the same @xmath29 dependence as at rhic .
as well , the nuclear modification factor @xmath30 is extracted and compared to pqcd calculations @xcite .
+ + the elliptic flow of @xmath1 hyperons has been measured in semi - central pb+pb collisions at 158_a _ gev .
the standard method of correlating particles with an event plane has been used .
the elliptic flow of @xmath1 particles increases linearly up to 2 gev / c .
this increase is weaker at sps than at rhic energies , partly due to different centrality selection .
a linear increase of @xmath31 is also observed for other particles ( @xmath23,proton ) as a function of @xmath29 in mid - central collisions at 158_a _
gev @xcite .
acknowledgements : this work was supported by the us department of energy grant de - fg03 - 97er41020/a000 , the bundesministerium fur bildung und forschung ( 06f137 ) , germany , the virtual institute vi-146 of helmholtz gemeinschaft , germany , fellowship of the german academic exchange service ( daad ) for ph.d .- research studies , germany , the polish state committee for scientific research ( 1 p03b 097 29 , 1 po3b 121 29 , 2 p03b 04123 ) , the hungarian scientific research foundation ( t032648 , t032293 , t043514 ) , the hungarian national science foundation , otka , ( f034707 ) , the polish - german foundation , the korea research foundation grant ( krf-2003 - 070-c00015 ) and the bulgarian national science fund ( ph-09/05 ) .
99 j. rafelski and b. mller , phys . rev
* 48 * ( 1982 ) 1066 .
j. rafelski and b. mller , phys .
* 56 * ( 1986 ) 2334 .
p. koch , b. mller and j. rafelski phys .
* 142 * ( 1986 ) 167 .
m. gadzicki and m. i. gorenstein , acta phys .
polon . b*30 * ( 1999 ) 2705 .
m. gadzicki for the na49 collaboration , j. phys .
* g * 30 ( 2004 ) s701 .
s. v. afanasiev et al .
( na49 collaboration ) , phys .
c * 66 * ( 2002 ) 054902 .
t. susa for the na49 collaboration , nucl
a698 * ( 2002 ) 491 . s. v. afanasiev et al .
( na49 collaboration ) , phys .
b * 538 * ( 2002 ) 275 . c. meurer for the na49 collaboration , j. phys .
* g * 30 ( 2004 ) s1325 .
t. anticic et al .
( na49 collaboration ) , phys .
* 93 * ( 2004 ) 022302 . c. alt et al .
( na49 collaboration ) , phys .
* 94 * ( 2005 ) 192301 . c. alt et al .
( na49 collaboration ) , phys .
* 94 * ( 2005 ) 052301 . c. alt et al .
( na49 collaboration ) , phys .
c * 73 * ( 2006 ) 044910 .
s. v. afanasiev et al .
( na49 collaboration ) , nucl .
instrum . meth
. a * 430 * ( 1999 ) 210 . c. blume for the na49 collaboration , j. phys .
g * 31 * ( 2005 ) s685 . j. l. klay et al .
( e895 collaboration ) , phys .
c * 68 * ( 2003 ) 054905 .
l. ahle et al .
( e802 collaboration ) , phys .
c * 57 * ( 1998 ) 466 .
l. ahle et al .
( e866 and e917 collaboration ) , phys . lett .
b * 490 * ( 2000 ) 53 .
s. albergo et al .
( e896 collaboration ) , phys .
* 88 * ( 2002 ) 062301 .
b. b. back et al .
( e917 collaboration ) , phys .
c * 69 * ( 2004 ) 054901
. m. gorenstein et al . , phys . lett .
* b567 * ( 2003 ) 175 .
j. adams et al .
( star collaboration ) , phys .
* 92 * ( 2004 ) 182301 .
j. speltz for the star collaboration , j. phys .
* g * 31(2005 ) s1025-s1028 .
e. l. bratkovskaya et al . , phys .
c * 69 * ( 2004 ) 054907 .
j. cleymans et al . , phys .
c * 60 * ( 1999 ) 054908 .
f. antinori for the na57 collaboration , j. phys .
* g * 32 ( 2006 ) s427-s441 .
d. elia for the na57 collaboration , j. phys .
* g * 30 ( 2004 ) s1329-s1332 .
r. j. glauber and g. matthiae , nucl .
b * 21 * , ( 1970 ) 135 .
t. schuster for the na49 collaboration , these proceedings .
d. kikola for the na49 collaboration , these proceedings . | we present results on strange particle production from experiment na49 in central and minimum bias pb+pb collisions in the beam energy range 20_a _ - 158_a _ gev .
new results on @xmath0 production in central pb+pb collisions and on @xmath1 , @xmath0 production in minimum bias collisions are shown .
transverse mass spectra and rapidity distributions of strange particles at different energies are compared .
the energy dependence of the particle yields and ratios is discussed .
na49 measurements of the @xmath1 and @xmath0 enhancement factors are shown for the first time . |
the recent trend towards parallel computing in the financial industry is not surprising . as the complexity of models used in the industry grows , while the demand for fast ,
sometimes real - time , solutions persists , parallel computing is a resource that is hard to ignore . in 2009 ,
bloomberg and nvidia worked together to run a two - factor model for calculating hard - to - price asset - backed securities on 48 linux servers paired with graphics processing units ( gpus ) , which traditionally required about 1000 servers to accommodate customer demand .
gpu computing offers several advantages over traditional parallel computing on clusters of cpus .
clusters consume non negligible energy and space , and computations over clusters are not always easy to scale . in contrast , gpu is small , fast , and consumes only a tiny fraction of energy consumed by clusters .
consequently , there has been a recent surge in academic papers and industry reports that document benefits of gpu computing in financial problems .
arguably , the numerical method that benefits most from gpus is the monte carlo simulation .
monte carlo methods are inherently parallel , and thus more suitable for implementing on gpu than most alternative methods . in this paper
we concentrate on monte carlo methods and financial simulation , and discuss computational and algorithmic issues when financial simulation algorithms are developed over gpu and traditional clusters .
the computational framework we use is the estimation of an integral @xmath0 over the @xmath1 dimensional unit cube , using sums of the form @xmath2 in monte carlo and quasi - monte carlo , @xmath3 converges to @xmath4 as @xmath5 in the former the convergence is probabilistic and @xmath6 come from a pseudorandom sequence , and in the latter the convergence is deterministic and the @xmath6 come from a low - discrepancy sequence . for a comprehensive survey of monte carlo and quasi - monte carlo methods ,
see @xcite .
often it is desirable to obtain multiple independent estimates for @xmath7 say @xmath8 so that one could use statistics to measure the accuracy of the estimation by the use of sample standard deviation , or confidence intervals .
let us assume that an allocation of computing resources is done and we choose parameters @xmath9 : the first parameter , @xmath10 is the sample size , and gives the number of vectors from the sequence ( pseudorandom or low discrepancy ) to use in estimating @xmath11 @xmath12 and the parameter @xmath13 gives the number of independent replications we obtain for @xmath3 , i.e. , @xmath14 the grand average @xmath15 gives the overall point estimate for @xmath16@xmath17 in monte carlo , to obtain the independent estimates @xmath18 one simply uses blocks of @xmath19 pseudorandom numbers .
in quasi - monte carlo , one has to use methods that enable independent randomizations of the underlying low - discrepancy sequence .
these methods are called randomized quasi - monte carlo ( rqmc ) methods ( see @xcite ) .
traditionally , in parallel implementations of monte carlo algorithms , one often assigns the @xmath20th processor ( of the @xmath13 allocated processors ) the evaluation of the estimate @xmath21 to do this computation , each processor needs to have an assigned number sequence ( pseudorandom or low - discrepancy ) and methods like blocking , leap - frogging , and parameterization are used to make this assignment .
parameterization is particularly useful when independent replications are needed to compute @xmath15 ( see @xcite , and also @xcite , @xcite , @xcite ) .
if only a single estimate is needed , then blocking or leap - frogging can be used ( @xcite , @xcite , @xcite , @xcite , @xcite ) .
figure [ paral](a ) describes this traditional monte carlo implementation where the @xmath20th processor @xmath22 generates its assigned sequence @xmath23 to compute @xmath24 as @xmath25 . in many applications @xmath19
is typically in millions , and @xmath13 is large enough for statistical accuracy , in the range 50 to 100 . in a massively parallel environment , depicted by the second diagram , where the number of processors @xmath26is much larger than @xmath13
, it can be a lot more efficient to completely `` transpose '' our computing strategy .
now the processors @xmath27 run simultaneously ( for a total of @xmath13 times ) to generate the sequence @xmath28 to compute @xmath24 as @xmath25 , where @xmath29 is part of the sequence @xmath30 which is assigned to the @xmath31th processor .
the choice of the two computing paradigms , which we vaguely name as `` parallel '' and `` massively parallel '' , determines how the underlying sequence ( pseudorandom or low - discrepancy ) should be generated . in the parallel paradigm ,
a recursive algorithm for generating the underlying sequence works best since each processor generates the `` entire '' sequence .
this paradigm is appropriate for a computing system with distributed memory , such as a cluster . for the massively parallel paradigm , a direct algorithm that generates
the @xmath32th term of the sequence from @xmath32 is more appropriate .
@xcite use the term `` counter - based '' to describe such direct algorithms .
the massively parallel paradigm is an appropriate model for gpu computing where prohibitive cost of memory access makes recursive computing inefficient . in section [ section pseudorandom ]
we briefly discuss a counter - based pseudorandom number generator , called * philox * , introduced by @xcite , and the pseudorandom number generators , * mersenne twister * , and * xorwow*. in section [ section rstart ] we introduce a randomized quasi - monte carlo sequence , which we name * rasrap * , and give algorithms for recursive and counter - based implementations of this sequence . in this section , we also give a brief description of a well - known quasi - monte carlo sequence , the sobol sequence .
we will compare the computational time for generating these sequences on cpu and gpu , in section [ section timing comp ] .
most pseudorandom number generators are inherently iterative : they are generated by successive application of a transformation @xmath33 to an element of the state space to obtain the next element of the state space , i.e. , @xmath34 . here
we discuss some of the pseudorandom number generators considered in this paper .
one of the most popular and high quality pseudorandom number generators is the mersenne twister introduced by @xcite .
it has a very large period and excellent uniformity properties .
it is available in many platforms , and recently matlab adopted it as its default random number generator .
a parallel implementation of the mersenne twister was also given by @xcite .
their approach uses parameterization , and it falls under our parallel computing paradigm : each processor in the parallel environment generates a mersenne twister , and different mersenne twisters generated across different processors are assumed to be statistically independent .
there are several parameters that need to be precomputed and stored to run the parallel implementation of mersenne twister .
xorwow is a fast pseudorandom number generator introduced by @xcite .
this generator is available in curand : a library for pseudorandom and quasi - random number generators for gpu provided by nvidia .
however , the generator fails certain statistical tests ; see @xcite for a discussion .
the reason we consider this generator is because of its availability in curand , and that its computational speed can be used as a benchmark against which other generators can be compared .
philox is a counter - based pseudorandom number generator introduced by @xcite .
its generation is in the form @xmath35 , and thus falls under our massively parallel computing paradigm .
a comparison of some counter - based and conventional pseudorandom number generators ( including philox and mersenne twister ) is given in @xcite . in section
[ section timing comp ] , we will present timing results comparing the pseudorandom number generators , and in section [ libor ] and [ mbs ] , we will compare these sequences when they are used in some financial problems .
these numerical results will also include rasrap and sobol , two randomized - quasi monte carlo sequences that we discuss next .
the van der corput sequence , and its generalization to higher dimensions , the halton sequence , are among the best well - known low - discrepancy sequences .
the @xmath36th term of the van der corput sequence in base @xmath37 , @xmath38 , is defined as @xmath39 where @xmath40 the halton sequence in the bases @xmath41 is @xmath42 .
this is a low - discrepancy sequence if the bases are relatively prime . in practice ,
@xmath43 is usually chosen as the @xmath31th prime number .
there is a well - known defect of the halton sequence : in higher dimensions , when the base is larger , certain components of the sequence exhibit very poor uniformity .
this is often referred to as _ high correlation between large bases .
_ as a remedy , permuted ( or , scrambled ) halton sequences were introduced .
the _ permuted van der corput sequence _ generalizes ( [ vdcp ] ) as @xmath44 where @xmath45 is a permutation on the digit set @xmath46 . by using different permutations for each base
, one can define the permuted halton sequences in the usual way .
there are many choices for permutations published in the literature ; a recent survey is given by @xcite . in this paper , we will follow the approach used in @xcite and pick these permutations at random .
the halton sequence can be generated recursively , which would be appropriate for an implementation on cpu , or directly ( counter - based ) , which would be appropriate for gpu .
next we discuss some recursive and counter - based algorithms for the halton sequence .
a fast recursive method for generating the van der corput sequence was given by @xcite .
we now explain his algorithm .
let @xmath47 be a positive integer and @xmath48 arbitrary .
define the sequence @xmath49 by @xmath50 and the transformation @xmath51 by @xmath52 where@xmath53 the transformation @xmath51 is called the von neumann - kakutani transformation in base @xmath54 the orbit of zero under @xmath55 i.e. , @xmath56 is the van der corput sequence in base @xmath47 . in fact
, the orbit of any point @xmath57 under @xmath51 is a low - discrepancy sequence .
if @xmath58 is chosen at random from the uniform distribution on @xmath59 then the orbit of @xmath58 under @xmath51 is called a _ random - start van der corput sequence _ in base @xmath54 the following algorithm summarizes the construction by @xcite of the ( random - start ) van der corput sequence in base @xmath54 it can be generalized to halton sequences in the obvious way .
@xcite[struck alg ] . generates a random - start van der corput sequence with starting point @xmath58 and base @xmath54 1 .
generate the sequence @xmath60 according to ( [ bk ] ) ; 2 . choose an arbitrary starting point @xmath48 ; 3 .
calculate @xmath61 according to ( [ k ] ) ; 4 . @xmath62
repeat step 3 - 4 .
algorithm [ struck alg ] is prone to rounding error in floating number operations due to the floor operation in ( [ k ] ) .
for example , a c++ compiler gives a wrong index @xmath61 after 3 steps of iteration when the starting point is @xmath63 if the rounding error introduced in ( [ bk ] ) is not carefully handled .
we now suggest an alternative algorithm that computes a random - start permuted halton sequence .
the advantages of this algorithm over algorithm [ struck alg ] are : ( i ) it avoids rounding errors , ( ii ) it is faster , and ( iii ) it can be used to generate permuted halton sequences .
[ recursive perm vdc ] ( recursive ) generates a random - start permuted van der corput sequence in base @xmath47 . 1 .
initialization step .
generate a random number @xmath64 and find some integer @xmath32 so that @xmath65 is the @xmath66 term in the van corput sequence in base @xmath47 .
initialize and store a random digit permutation @xmath67 .
expand @xmath32 in base @xmath47 as @xmath68 ( @xmath61 depends on @xmath32 ) .
set @xmath69 for @xmath70 .
store @xmath71 .
calculate and store @xmath72 for @xmath73 .
set @xmath74 for @xmath75 .
set the quasi - random number @xmath76 ; 2 .
let @xmath77 .
find @xmath78 ; 3 .
. set @xmath80 . set @xmath81 for @xmath82 .
the quasi - random number corresponding to @xmath83 is @xmath76 ; 4 .
repeat step 2 - 3 .
algorithm 2 is an efficient iterative algorithm appropriate for the parallel computing paradigm .
however , for the massively parallel computing paradigm , such as gpu computing , we need a counter - based algorithm . for the halton sequence
, this would be simply its definition : ( counter - based ) [ definition vdc ] generates a random - start permuted van der corput sequence in base @xmath47 .
1 . initialization step : choose a small positive real number , @xmath84 .
generate a random number @xmath65 from the uniform distribution on @xmath85 , and find @xmath32 such that @xmath86 ; 2 .
the quasi - random number corresponding to @xmath32 is @xmath87 ; 3 .
let @xmath77 and repeat step 2 - 3 .
the name rasrap is an abbreviation for random - start randomly permuted halton sequence : if in algorithms [ recursive perm vdc ] and [ definition vdc ] , the permutations for each base are generated at random , then we obtain rasrap . the sobol sequence is a well - known fast low - discrepancy sequence popular among financial engineers .
the @xmath88th component of the @xmath31th vector in a sobol sequence is calculated by @xmath89 , where @xmath90 is the @xmath61th digit from the right when integer @xmath31 is represented in base @xmath91 and @xmath92 is the bitwise exclusive - or operator .
the so - called direction numbers , @xmath93 , are defined as @xmath94 to generate the sobol sequence , we need to generate a sequence of positive integers @xmath95 . the sequence @xmath95 is defined recursively as follows : @xmath96 , where @xmath97 are coefficients of a primitive polynomial of degree @xmath98 in the field @xmath99 , @xmath100 .
the initial values @xmath101 can be chosen freely given that each @xmath102 , is odd and less than @xmath103 . because of this freedom , different choices for direction numbers can be made based on different search criteria minimizing the discrepancy of the sequence .
we use the primitive polynomials and direction numbers provided by @xcite .
the counter - based implementation of the sobol sequence introduced here is convenient on gpus , but a more efficient implementation proposed by antonov and saleev based on gray code is used in practice on cpus . for details about this approach ,
see @xcite .
the sobol sequence can be randomized using various randomized quasi - monte carlo methods . here
we will use the random digit scrambling method of @xcite .
more on randomized quasi - monte carlo and some parallel implementations can be found in @xcite , and , @xcite .
mersenne twister , philox , xorwow , rasrap , and sobol sequences are run on intel i7 3770k and nvidia geforce gtx 670 .
we compare the throughput of different algorithms on cpu ( table 1 ) and gpu ( table 2 ) .
table 1 shows that the fastest algorithm for the halton sequence on cpu is algorithm 2 .
it is about 3.8 times as fast as the algorithm by struckmeier ( algorithm 1 ) .
not surprisingly algorithm 3 , the counter - based implementation , is considerably slower on cpu .
mersenne twister uses its serial cpu implementation and it is about 3.4 times faster than algorithm 2 for the halton sequence . and sobol sequence based on gray code is faster than mersenne twister . table 2 shows that the throughput of algorithm 3 on gpu improves significantly compared to the cpu value .
counter - based sobol sequence is twice as fast as rasrap , and the pseudorandom number generator philox is almost 200 times faster than rasrap .
[ gencpu ] the computational speed at which various sequences are generated is only one part of the story .
we next examine the accuracy of the estimates obtained when these sequences are used in simulation . in the next section
, we use these sequences in two problems from computational finance , and compare them with respect to the standard deviation of their estimates and computational speed .
an interest rate derivative is a derivative where the underlying asset is the right to pay or receive a notional amount of money at a given interest rate .
the interest rate derivatives market is the largest derivatives market in the world . to price interest rate derivatives , forward
interest rate models are widely used in the industry .
there are two kinds of forward rate models : the continuous rate model and the simple rate model .
the framework developed by @xcite ( hjm ) explicitly describes the dynamics of the term structure of the interest rates through the dynamics of the forward rate curve .
hjm model has two major drawbacks : ( 1 ) the instantaneous forward rates are not directly observable in the market ; ( 2 ) some simple choices of the form of volatility is not admissible . in practice , many fixed income securities quote the interest rate on an annual basis with semi - annual or quarterly compounding , instead of a continuously compounded rate .
the simple forward rate models describe the dynamics of the term structure of interest rates through simple forward rates , which are observable in the market .
this approach is developed by @xcite , @xcite , @xcite and @xcite .
the london inter - bank offered rates ( libor ) is one of the most important benchmark simple interest rates .
let @xmath104 denote the time-@xmath105 value of a zero coupon bond paying 1 at the maturity time @xmath106 .
a forward rate @xmath107 ( @xmath108 ) is an interest rate fixed at time @xmath105 for borrowing or lending at time @xmath109 over the period @xmath110 $ ] .
an arbitrage argument shows that forward rates are determined by bond prices in accordance to @xmath111 a forward libor rate is a special case of ( [ fwdrate ] ) with a fixed period @xmath112 for the accrual period .
typically @xmath113 or @xmath114 .
thus , the @xmath115-year forward libor rate at time @xmath105 with maturity @xmath106 is @xmath116 so if we enter into a contract at time @xmath117 to borrow 1 at time @xmath106 and repay it with interest at time @xmath118 , the interest due will be @xmath119 . fix a finite set of maturities @xmath120 and let @xmath121 , denote the lengths of the intervals between maturities .
normally we fix @xmath115 as a constant regardless of day - count conventions that would introduce slightly different values for the fractions @xmath122 . for each maturity @xmath123 , let @xmath124 denote the time-@xmath105 value of a zero coupon bond maturing at @xmath125 . and write @xmath126 for the forward rate at time @xmath105 over the period @xmath127 $ ] . equation ( [ fwdlibor ] ) can be then rewritten as @xmath128 the subscript @xmath129 emphasizes we are looking at a finite set of bonds .
the dynamics of the forward libor rates can be described as a system of sdes as follows . for a brief informal derivation ,
see @xcite .
@xmath130 where @xmath131 is a @xmath132-dimensional standard brownian motion and the volatility @xmath133 may depend on the current vector of rates @xmath134 as well as the current time @xmath105 .
@xmath135 is the unique integer such that @xmath136 .
pricing interest rate derivative securities with libor market models normally requires simulations . since the libor market model deals with a finite number of maturities , only the time variable needs to be discretized .
we fix a time grid @xmath137 to simulate the libor market model . in practice
, one would often take @xmath138 so the simulation goes directly from one maturity date to the next . for simplicity
, we use a constant volatility @xmath139 in the simulation .
we apply an euler scheme to ( [ dl_n ] ) to discretize the system of sdes of the libor market model , producing @xmath140 + \hat{l}_n(t_i)\sqrt{t_{i+1 } - t_i } \sigma_n(t_i)^{\top}z_{i+1},\ ] ] where @xmath141 and @xmath142 are independent @xmath143 random vectors in @xmath144 . here
hats are used to identify discretized variables .
we assume an initial set of bond prices @xmath145 is given and initialize the simulation by setting @xmath146 in accordance with ( [ fwdliborn ] ) .
next we use the simulated evolution of libor market rates to price a caplet .
an interest rate cap is a portfolio of options that serve to limit the interest paid on a floating rate liability over a set of consecutive periods .
each individual option in the cap applies to a single period and is called a caplet .
it is sufficient to price caplets since the value of a cap is simply the sum of the values of its component caplets .
we follow the derivation in @xcite .
consider a caplet for the time period @xmath147 $ ] .
a party with a floating rate liability over that period would pay interest @xmath148 times the principle at time @xmath149 .
a caplet is designed to limit the interest paid to a fixed level @xmath150 .
the difference @xmath151 would be refunded only if it is positive .
so the payoff function of a caplet is @xmath152 , where the notation @xmath153 indicates that we take the maximum of the expression in parentheses and zero .
this payoff is exercised at time @xmath118 but determined at time @xmath106 .
there is no uncertainty in the payoff over the period @xmath154 $ ] .
then the payoff function at time @xmath118 is equal to @xmath155 at time @xmath106 .
this payoff typically requires the simulation of the dynamics of the term structure .
@xcite derived a formula for the time-@xmath105 price of the caplet under the assumption of @xmath156 following a lognormal distribution , which does not necessarily correspond to a price in the sense of the theory of derivatives valuation . in practice , this formula @xmath157 is used to calculate the `` implied volatility '' @xmath139 from the market price of caps . to test the correctness of the libor market model simulation , we use the daily treasury yield curve rates on 02/24/2012 as shown in table 3 to initialize the libor market rates simulation .
we first apply a cubic spline interpolation to the rates in table 3 to get estimated yield curve rates for every 6 months .
then the estimated yield curve rates are used to calculate the bond prices for every 6 months in order to initialize the libor rates in ( [ bl ] ) .
we assume the following parameters in libor rates simulation @xmath158 the simulations are run on intel i7 3770k and nvidia geforce gtx 670 respectively . for a fixed sample size @xmath19 , we repeat the simulation 100 times using independent realizations of the underlying sequence .
we investigate the sample standard deviation of the 100 estimates and computing time as a function of the sample size @xmath19 .
we also compare the efficiency of different sequences , where efficiency is defined as the product of sample standard deviation and execution time .
the sobol sequence and a scrambled version of it are provided in the curand library from nvidia .
we use both the single precision version ( sobol-lib(single ) ) and double precision version ( sobol-lib(double ) ) in our simulation .
we also implement our own version of the sobol sequence ( sobol(single ) and sobol(double ) ) for comparison .
figure [ sobolcomp ] plots the sample standard deviation of 100 estimates for the caplet price , computing time , and efficiency , of different implementations of the sobol sequence against the sample size @xmath19 .
we also include the numerical results obtained using the fast pseudorandom number sequence xorwow from curand as a reference .
we make the following observations : 1 .
the convergence rate exhibits a strange behavior and levels off for the curand sobol sequence generators , sobol-lib(single ) and sobol-lib(double ) , as @xmath19 gets large .
our implementation of the sobol sequence gives monotonically decreasing sample standard deviation as @xmath19 increases ; 2 .
the execution time for curand generators sobol-lib(single ) and sobol-lib(double ) is significantly longer than our implementation , and not monotonic for a specific range of @xmath19 ; 3 . the efficiency of curand generators sobol-lib(single ) and sobol-lib(double ) is even worse than the efficiency of the pseudorandom number sequence xorwow .
our sobol sequence implementations have better efficiency than xorwow .
due to the poor behavior of the sobol sequence in the curand library , we will use our implementation of the sobol sequence with single precision in the rest of the paper .
we will denote this sequence simply as sobol " in the numerical results . in section [ section timing comp ] , we compared the computing times of several sequences .
here we compare the performance of mersenne twister , rasrap and sobol , when they are used in simulating the libor market model .
the sequences are run on one cpu core .
figure [ singlecpulibor ] shows that the sample standard deviation of the estimates obtained from rasrap and sobol sequences converge at a much faster rate than the mersenne twister .
the convergence rate for mersenne twister is about @xmath159 , and the rate for rasrap and sobol is about @xmath160 and @xmath161 , respectively .
the recursive implementation of rasrap does not introduce much overhead in running time and gives very close timing results to mersenne twister .
the sobol sequence based on gray code is faster than mersenne twister . as a result ,
the two low - discrepancy sequences enjoy better and flatter " efficiency than that of mersenne twister .
we next investigate how well rasrap and sobol sequence results scale over multi - core cpu .
we implement a parallel version of rasrap and sobol with openmp that can run on 8 cpu cores simultaneously .
figure [ liborcpu ] plots the performance of openmp version of rasrap and sobol on cpu .
it exhibits the same pattern of convergence , running time , and efficiency as in figure [ singlecpulibor ] .
the convergence remains the same as in the one core case , but we gain a speedup of four with the parallelism using openmp . in this section
we compare the counter - based implementations of rasrap and sobol with pseudorandom sequences philox and xorwow , on gpu .
figure [ liborgpu ] plots the sample standard deviation , computing time , and effciency .
we make the following observations : 1 .
the convergence rate for philox and xorwow is about @xmath162 and @xmath163 , respectively ; 2 .
the convergence rate for rasrap and sobol is about @xmath164 and @xmath165 respectively ; 3 .
xorwow is the fastest generator , followed by philox and sobol. rasrap is slightly slower than sobol ; 4 .
the efficiency of sobol is the best among all sequences .
we follow the mortgage - backed securities ( mbs ) model given by @xcite . consider a security backed by mortgages of length @xmath166 with fixed interest rate @xmath167 which is the interest rate at the beginning of the mortgage .
the present value of the security is then @xmath168 , where @xmath169 is the expectation over the random variables involved in the interest rate fluctuations .
the parameters in the model are the following : @xmath170 discount factor for month @xmath171 @xmath172 cash flow for month @xmath171 @xmath173 interest rate for month @xmath171 @xmath174 fraction of remaining mortgages prepaying in month @xmath171 @xmath175 fraction of remaining mortgages at month @xmath171 @xmath176 ( remaining annuity at month @xmath171 ) @xmath177 @xmath178 monthly payment @xmath179@xmath180 random variable .
the model defines several of these variables as follows : @xmath181 @xmath182 @xmath183 @xmath184 the interest rate fluctuations and the prepayment rate are given by @xmath185 @xmath186 where @xmath187 are constants of the model .
the constant @xmath188 is chosen to normalize the log - normal distribution so that @xmath189 .
the initial interest rate @xmath190 also needs to be specified .
we choose the following parameters in our numerical results : @xmath191 figure [ mbscpu ] compares openmp implementations of rasrap and sobol sequences on 8 cpu cores .
the sample standard deviation of estimates obtained by rasrap is smaller than that of sobol for every sample size , however , the sobol sequence gives a better rate of convergence .
we gain a speedup of 6 with the parallelism using openmp compared to the single core version .
rasrap has the better efficiency for all sample sizes .
figure [ mbsgpu ] compares the gpu implementations of rasrap , sobol , philox , and xorwow .
we observe : 1 .
the convergence rate for philox and xorwow is about @xmath192 ; 2 .
rasrap gives lower standard deviation than sobol , however , the convergence rate for sobol ( @xmath164 ) is better than rasrap ( @xmath193 ) ; 3 .
the efficiency of rasrap is the best among all sequences .
in figure [ hist ] , we display the gpu speed - up over cpu for both libor and mbs examples . these results only consider the computing time , and the computing time of cpu - twister is taken as the base value in each example .
the largest speed - up is a factor of 95 and it is due to gpu - xorwow for the libor market model simulation . in the mbs example , gpu - rasrap speed - up is a factor of 250 , and the other gpu sequences give a speed - up of factor 290 . finally , to demonstrate the impressive computing power of gpu , we compare gpu with the high performance computing ( hpc ) cluster at florida state university .
we implement a parallel sobol sequence using mpi , and run simulations for the two examples , libor and mbs .
figure [ cluster ] plots the computing time against the number of cores used by the cluster , when the sample size @xmath19 takes various values . the gpu computing time is plotted as a horizontal line since all the cores of gpu are used in computations .
figure [ cluster ] shows that for the libor example , the gpu we used in our computations has equivalent computing power roughly as 128 nodes on the hpc cluster .
this is about when the hpc computing time plot reaches the level of gpu computing time , for each @xmath19 . in the mbs example , 256 nodes on the hpc cluster are equivalent to the gpu .
we also point out that on a heterogeneous computing environment such as a cluster , continually increasing the number of nodes will not necessarily decrease the running time due to higher cost of communication between nodes and higher probability that slow nodes are used .
but for gpus , a more powerful product with more cores would suggest gains in computing time .
chen , g. , thulasiraman , p. and thulasiram , r.k .
, distributed quasi - monte carlo algorithm for option pricing on hnows using mpc . in _ proceedings of the 39th annual simulation symposium _ , huntsville , usa , 26 april 2006 , pp .
9097 , 2006 .
dedoncker , e. , zanny , r. , ciobanu , m. and guan , y. , distributed quasi monte - carlo methods in a heterogeneous environment . in
_ proceedings of the 9th heterogeneous computing workshop _ ,
cancun , mexico , may 2000 , pp .
200206 , 2000 .
hofbauer , h. , uhl , a. and zinterhof , p. , parameterization of zinterhof sequences for grid - based qmc integration . in j. volkert , t. fahringer , d. kranzlmller , and w. schreiner , editors , _ proceedings of the 2nd austrian grid symposium _ ,
volume 221 of [email protected] , innsbruck , austria , 2007 , pp . 91105 , 2007 .
austrian computer society .
matsumoto , m. and nishimura , t. , mersenne twister : a 623-dimensionally equidistributed uniform pseudorandom number generator .
_ acm transactions on modeling and computer simulations , _ 1998 , * 8(1 ) * , 330 .
kten , g. and srinivasan , a. , parallel quasi - monte carlo applications on a heterogeneous cluster . in k. t.
fang , f. j. hickernell and h. niederreiter , editors , _ monte carlo and quasi - monte carlo methods 2000 _ , springer - verlag , berlin , 2002 , pp .
406421 .
kten , g. , shah , m. and goncharov , y. , random and deterministic digit permutations of the halton sequence . in l. plaskota and h.
woniakowski , editors , _ monte carlo and quasi - monte carlo methods 2010 _ , springer , 2012 .
saito , m. and matsumoto , m. , a deviation of curand : standard pseudorandom number generator in cuda for gpgpu . in _
10th international confrence on monte carlo and quasi - monte carlo methods in scientific computing _ ,
sydney , australia , 1317 february 2012 .
salmon , j.k . ,
moraes , m.a .
, dror , r.o . and shaw , d.e .
, parallel random numbers : as easy as 1 , 2 , 3 . in _ sc11 proceedings of 2011 international conference for high performance computing , networking , storage and analysis _ , ny , usa , 2011 .
vandewoestyne , b. and cools , r. , good permutations for deterministic scrambled halton sequences in terms of @xmath196-discrepancy .
_ journal of computational and applied mathematics _ , 2006 , * 189 * , 341361 . | gpu computing has become popular in computational finance and many financial institutions are moving their cpu based applications to the gpu platform . since most monte carlo algorithms are embarrassingly parallel , they benefit greatly from parallel implementations , and consequently monte carlo has become a focal point in gpu computing .
gpu speed - up examples reported in the literature often involve monte carlo algorithms , and there are software tools commercially available that help migrate monte carlo financial pricing models to gpu .
we present a survey of monte carlo and randomized quasi - monte carlo methods , and discuss existing ( quasi ) monte carlo sequences in gpu libraries .
we discuss specific features of gpu architecture relevant for developing efficient ( quasi ) monte carlo methods .
we introduce a recent randomized quasi - monte carlo method , and compare it with some of the existing implementations on gpu , when they are used in pricing caplets in the libor market model and mortgage backed securities .
gpu , monte carlo , randomized quasi - monte carlo , libor , mortgage backed securities [ [ section ] ] |
in the standard model ( sm ) @xcite leptons are pointlike fermions .
the magnetic moments are predicted with high precision by the theory and the electric dipole moments must be zero .
the moments of the electron are measured with high precision @xcite and found in perfect agreement with the predictions . for the muon magnetic moment a discrepancy of 2.7 standard deviations is found between recent measurements and the prediction @xcite , giving rise to think about a possible indication of physics beyond the sm .
a deviation of the magnetic moments from the predictions of the theory would be a signal for a substructure or new interactions .
furthermore , a measurement of non - zero values of the electric ( electromagnetic or weak ) dipole moments reveals cp violation . for the @xmath0-lepton , about 17 times heavier than the muon ,
only rough limits on the moments were derived before the lep era @xcite . at lep experiments , measurements of the electromagnetic moments
are done using the processes @xmath3 and , recently , @xmath4 . also belle published a measurement of the electric dipole moment of the @xmath0-lepton in @xmath2 .
weak moments are measured from the process @xmath2at centre - of - mass energies near the @xmath5 resonance by several experiments .
the aleph experiment published a new result using the full data statistics .
a review of the experimental results is given .
as a function of @xmath6 = @xmath7 ( top ) and as a function of @xmath8 ( bottom).,title="fig:",width=264 ] as a function of @xmath6 = @xmath7 ( top ) and as a function of @xmath8 ( bottom).,title="fig:",width=264 ] in general a @xmath0-lepton couples to a photon through its electric charge , magnetic moment and electric dipole moment .
the lorentz - invariant ansatz is given by the following form @xcite , @xmath9 the @xmath10 dependent form - factors , @xmath11 , have familiar interpretations for @xmath12 : @xmath13 = @xmath14 , where @xmath13 is the charge of the @xmath0-lepton and @xmath15 the unit charge , @xmath16 = @xmath7 is the anomalous magnetic moment , @xmath7 = @xmath17 , and @xmath18 = @xmath19 , where @xmath8 is the electric dipole moment . in the sm @xmath7
is non - zero due to higher order diagrams and is predicted to be @xmath20 = 0.001 177 3(3 ) @xcite . a non - zero value of @xmath8 is forbidden by both p invariance and t invariance @xcite .
assuming cpt invariance , the observation of a non - zero value of @xmath8 would imply cp violation .
@xmath21 and @xmath22 can be probed in processes with @xmath23 vertices .
hereafter i follow the convention @xmath7= @xmath6 and @xmath8 = @xmath24 . in the process @xmath2 at low energies the photon is virtual with @xmath10 = s , where s is the centre - of - mass energy .
non - zero values of @xmath7 and @xmath8 contribute to the cross section @xcite . exploiting
also the @xmath0 lepton decays in @xmath2 , additional information on @xmath8 can be obtained from triple momentum and spin correlation observables @xcite .
the joint spin - density matrix of the @xmath0 leptons can be written as @xmath25 the term @xmath26 results from the sm amplitudes , @xmath27 and @xmath28 are the interference terms between sm and cp violating amplitudes and @xmath29 is the cp even contribution bilinear in @xmath8 .
the term @xmath27 is cp odd and t odd while @xmath28 is cp odd and t even .
the coupling to real photons appears in @xmath3 ( final state bremsstrahlung ) .
non - zero dipole moments enhance the production of high energy photons in @xmath3 @xcite . for small @xmath7
the effect is due to the interference between the sm and the anomalous amplitudes leading to a linear dependence in @xmath7 .
the electric dipole moment contributions depend on the square of @xmath8 , allowing only the determination of its absolute value . in this process @xmath10
is zero but the radiating @xmath0 is not on the mass shell .
almost real photons are expected in the two - photon process , @xmath4 .
the cross section of @xmath4 depends on a quartic polynomial in @xmath7 @xcite . for small values of @xmath7
again only the linear term in @xmath7 contributes . in case of @xmath8
the lowest order dependence is quadratic . for illustration ,
figure [ fig : ggcs ] shows the change of the total cross section of @xmath4 as a function of @xmath7 and @xmath8 .
weak dipole moments are introduced to the @xmath30 vertex in a similar way as for the electromagnetic ones in eqn.(1 ) @xcite .
the quantity @xmath31 = @xmath32 is the anomalous weak magnetic dipole moment and @xmath33 = @xmath34/@xmath15 is the weak electric dipole moment .
both quantities are zero in the sm at born level .
loop diagrams result in @xmath35 = @xmath36 @xcite and @xmath37 cm @xcite , where @xmath15 is the unit charge .
these numbers are well beyond the sensitivity of current experiments . however , some models beyond sm predict values up to @xmath38 for @xmath35 @xcite and @xmath39 cm for @xmath34 @xcite .
the measurement of @xmath35 and @xmath34 in @xmath40is possible from the transverse and normal polarisations of the @xmath0-leptons @xcite .
the full sensitivity is obtained if the @xmath0-lepton direction of flight is reconstructed .
this is possible when both @xmath0-leptons decay semileptonically @xcite . in some measurements
also the ansatz given in eqn.(2 ) is used .
in the l3 detector . two low momentum tracks point to local energy depositions in the bgo calorimeter .
in addition , two photons originating from a @xmath43 are seen .
the event is identified as a @xmath44 , @xmath45 final state .
, width=264 ] decays into an electron and a non - electron are selected .
, title="fig:",width=264,height=188 ] decays into an electron and a non - electron are selected .
, title="fig:",width=264,height=188 ] the first estimate of @xmath41 was done using data from petra @xcite at @xmath46 1000 gev@xmath47 , resulting in @xmath48 .
indirect limits were inferred from the partial width @xmath49
@xcite measured at lep , yielding [email protected] and @xmath51 cm . the first attempt to derive a value at @xmath10 = 0
was done in ref .
@xcite using data from the l3 experiment at lep .
the latter method was improved , on the basis of the new calculations @xcite and applied to the data by the opal @xcite and l3 @xcite experiments . using the photon energy spectrum and the distribution of the angle between the photon and the nearest @xmath0-lepton , a likelihood fit done by l3 results to @xmath52 and @xmath53 cm .
recently delphi @xcite and l3 @xcite analyzed the @xmath4channel to determine these moments .
both experiments used the data of lep in the centre - of - mass energy range between @xmath54 = 180 gev and 208 gev , corresponding to an integrated luminosity of 650 pb@xmath55 per experiment .
only events are used with non - detected electrons and positrons , ensuring that @xmath56 1 gev@xmath47 .
the selection of @xmath4events requires low deposited energy in the detectors and two non - coplanar low momentum tracks . a typical event in the l3 detector is shown in figure [ fig : l3event ] .
l3 accepted events with one @xmath0-lepton decaying into an electron and the other @xmath0-lepton decaying into a @xmath57-meson .
a sample of 351 events is selected .
the selection efficiency is about 0.1% and the background , mainly from other @xmath0-lepton decays , amounts to 30% .
delphi selected events with one @xmath0-lepton decaying into an electron and the other decaying in any one - prong final state but not an electron .
the identification of electrons and other particles is done using the ionisation along the track in the time projection chamber ( tpc ) .
a pull - quantity is defined for each track by the normalised difference of the expected and measured energy loss depositions for several particle hypotheses . as an example
, the pull for the muon hypothesis and the electron hypothesis is shown in figure [ fig : dpull ] . combining the pull information a sample of 2390 @xmath4events is selected from a large two - photon event sample .
the efficiency is of about 0.8% and the background of about 12% results mainly from other two - photon final states .
the measured cross sections , shown in figure [ fig : ggresults ] as a function of the centre - of - mass energy , are compared to the expectation from the sm and a fit is performed with cross section predictions including @xmath41 and @xmath42 as free parameters . since the measurements are in reasonable agreement with the sm predictions , there is no indication for non - sm values of the @xmath0 lepton moments . for the channels
@xmath3 and @xmath58 , respectively , opal and l3 published only the limits , and + , the results from delphi and l3 measurements for the channels @xmath4 and @xmath3 are summarized in table [ table : tabtau ] . .
the values of @xmath41 and @xmath42 measured by delphi in @xmath4 and by l3 in @xmath3 . [ cols="<,^,^",options="header " , ] averaging these values leads to @xmath59= ( -0.65@xmath60 cm and @xmath61 = ( [email protected])@xmath63 cm , or to 95% cl limits of ( @xmath64 cm ) and ( @xmath65 cm ) .
the most recent measurements of the @xmath0-lepton electromagnetic and weak moments are in agreement with the standard model predictions .
no hint for new physics is observed .
the accuracy is not sufficient to test the sm loop corrections . in case that experiments published measured values of similar accuracy , given in table [ table : tabtau ] and table [
table : dtab ] , the results are combined .
otherwise the most precise measurements are taken .
limits are obtained at 95% c.l . for the electromagnetic moments , + @xmath66 and for the weak moments : @xmath67
i thank francisco matorras and daniel haas for the support given to me to prepare this summary and fred jegerlehner and gunnar klaemke for critical reading of the article .
klaus mnig i thank for helpful suggestions .
also i would like to thank h. hayashii for the excellent organisation of this workshop providing a creative and pleasant atmosphere .
9 s. l. glashow , , s. weinberg , , a. salam , `` weak and electromagnetic interactions '' , in _ elementary particle theory _ , edited by n. svartholm , page 367 , stockholm , 1969 , almqvist and wiksell .
peter j. mohr and barry n. taylor , .
d. hertzog , _ measurement of the muon anomaly to high and even higher precision _ , these proceedings .
d.j . silverman and g.l .
s. weinberg , 1995 , _ the quantum theory of fields _ , cambridge university press , cambridge 452 .
samuel , g. li and r. mendel , , erratum _ ibid _ * 69 * ( 995 ) ( 1992 ) ; f. hamzeh and n.f .
nasrallah , .
barr , w.j .
marciano , _ electric dipole moments _ proceedings _ cp violation _ , ed . c. jarlskog ,
world scientific , ( 1990 ) .
w. bernreuther , o. nachtmann , and p. overmann , .
j. biebel and t. riemann , ; s.s .
gau , t. paul j. swain and l. taylor , . f. cornet and j. illana , . j. bernabeu , g.a .
gonzalez - springberg , j. vidal , ; u. stiegler , .
j. bernabeu , .
w. bernreuther , ; m.j .
booth , universityof chicago report no .
efi-93 - 02 ; hep - ph/9301293
. m. c. gonzalez - garcia , s.f .
novaes , ; p. poulose , s.d .
rindani , preprint prl - th-97/24 , 1997 , hep - ph/9708332 ; t. huang , w. lu and z. tao , .
r. escribano , e. masso , .
j.a . grifols and a. mendez , , erratum _ ibid _ * 259 * ( 512 ) ( 1992 ) .
opal coll .
, k. ackerstaff , .
, m. acciarri , .
delphi coll .
, j. abdallah , .
, p. achard , .
argus collab . ,
h. albrecht , .
belle collab .
, k. inami , .
p. acton , .
w. bernreuther , .
n. wermes , nucl .
b * 55 * c ( proc . supplement ) 313 ( 1997 ) . l3
, m. acciarri , .
aleph coll .
, a. heister , . | the electromagnetic and weak dipole moments of the @xmath0-lepton have been measured by experiments at @xmath1colliders .
data samples of @xmath2 , @xmath3 and @xmath4 events collected at centre - of - mass energies between 10 and 200 gev are used . no deviation from the standard model
is found .
limits on the moments are summarised from the most recent results . |
research in compressed sensing @xcite @xcite is expanding rapidly .
the sufficient condition for @xmath0-recovery based on the restricted isometry property ( rip ) @xcite @xcite is one of the celebrated results in this field .
the design of sensing matrices with small rip constants is a theoretically interesting and challenging problem .
currently , random constructions provide the strongest results , and the analysis of random constructions is based on large deviations of maximum and minimum singular values of random matrices @xcite @xcite . in the present paper , a random construction of bipolar sensing matrices based on binary linear codes
is introduced and its rip is analyzed .
the column vectors of the proposed sensing matrix are nonzero codewords of a randomly chosen binary linear code . using a generator matrix , a @xmath1 sensing matrix can be represented by @xmath2-bits .
the existence of sensing matrices with the rip is shown based on an argument on the ensemble average of the weight distribution of binary linear codes .
the symbols @xmath3 and @xmath4 represent the field of real numbers and the finite field with two elements @xmath5 , respectively .
the set of all @xmath1 real matrices is denoted by @xmath6 . in the present paper ,
the notation @xmath7 indicates that @xmath8 is a column vector of length @xmath9 .
the notation @xmath10 denotes @xmath11-norm @xmath12 defined by @xmath13 the @xmath14-norm is defined by @xmath15 where @xmath16 denotes the index set of nonzero components of @xmath8 .
the functions @xmath17 and @xmath18 are the hamming weight and hamming distance functions , respectively .
let @xmath19 be a @xmath1 real matrix , where the @xmath20-norm of the @xmath21-th @xmath22)$ ] column vector @xmath23 is normalized to one , namely , @xmath24 .
the notation @xmath25 $ ] represents the set of consecutive integers from @xmath26 to @xmath27 . the restricted isometry property of @xmath28 introduced by candes and tao @xcite plays a key role in a sufficient condition of @xmath0-recovery .
a vector @xmath29 is called an @xmath30-sparse @xmath31)$ ] vector if @xmath32 .
if there exists a real number @xmath33 satisfying @xmath34 for any @xmath30-sparse vector @xmath29 , then we say that @xmath28 has the rip of order @xmath30
. if @xmath28 has the rip of order @xmath30 , then the smallest constant satisfying ( [ rip ] ) is called the _ rip constant _ of @xmath28 , which is denoted by @xmath35 .
assume that @xmath28 has the rip with small @xmath35 .
in such a case , any sub - matrix composed from @xmath36-columns @xmath37 of @xmath28 is nearly orthonormal .
recently , candes @xcite reported the relation between the rip and the @xmath0-recovery property .
a portion of the main results of @xcite is summarized as follows .
let @xmath38 $ ] , and assume that @xmath28 has the rip with @xmath39 for any @xmath30-sparse vector @xmath40 ( i.e. , @xmath41 ) , the solution of the following @xmath0-minimization problem @xmath42 coincides exactly with @xmath43 , where @xmath44 .
note that @xcite considers stronger reconstruction results ( i.e. , robust reconstruction ) .
the matrix @xmath28 in ( [ l1recovery ] ) is called a _ sensing matrix_. the incoherence of @xmath28 defined below and the rip constant are closely related .
the incoherence of @xmath28 is defined by @xmath45 , i \ne j } |\phi_i^t \phi_j|.\ ] ] the following lemma shows the relation between the incoherence and the rip constant .
similar bounds are well known ( e.g.,@xcite ) .
[ deltaupper ] assume that @xmath46 is given . for any @xmath38 $ ]
, @xmath35 is upper bounded by @xmath47 an elementary proof ( different from that in @xcite ) is presented in appendix .
in this section , we present a construction method for sensing matrices based on binary linear codes . a sensing matrix obtained from this construction has a concise description . a sensor can store a generator matrix of a binary linear code , instead of the entire sensing matrix .
the function @xmath48 is called a _ binary to bipolar conversion function _ defined by @xmath49 where @xmath43 is an all - one column vector of length @xmath9 .
namely , using the binary to bipolar conversion function , a binary sequence is converted to a @xmath50-sequence .
the following lemma demonstrates that the inner product of two bipolar sequences @xmath51 and @xmath52 is determined from the hamming distance between the binary sequences @xmath26 and @xmath27 .
[ hammingandcorr ] for any @xmath53 , the inner product of @xmath51 and @xmath52 is given by @xmath54 ( proof ) let @xmath55 and @xmath56 .
define @xmath57 and @xmath58 by @xmath59 : a_i = b_i \ } , \quad y_{2 } { \stackrel{\triangle}{=}}\{i \in [ 1,p ] : a_i \ne b_i \},\ ] ] where @xmath60 and @xmath61 .
equation ( [ correqu ] ) is derived as follows : @xmath62 it is easy to confirm that @xmath51 is normalized , i.e. , @xmath63 , for any @xmath64 .
let @xmath65 @xmath66 be a binary @xmath67 parity check matrix where @xmath68 holds .
the binary linear coded @xmath69 defined by @xmath70 is given by @xmath71 where @xmath72 is a zero - column vector of length @xmath73 .
the following definition gives the construction of sensing matrices .
assume that all of the nonzero codewords of @xmath69 are denoted by @xmath74 ( based on any predefined order ) , where @xmath75 .
the sensing matrix @xmath76 is defined by @xmath77 where @xmath78 .
if @xmath79 has the rip of order @xmath30 , the rip constant corresponding to @xmath79 is denoted by @xmath80 .
since the order of the columns is unimportant , we do not distinguish between sensing matrices of different column order ( or choice of codewords from @xmath69 ) .
if the weights of all nonzero codewords of @xmath69 are very close to @xmath81 , then the incoherence of @xmath79 becomes small , as described in detail in the following lemma .
[ ripbound ] assume that @xmath82 is given and that @xmath83 holds for any @xmath84 .
in such a case , the incoherence @xmath79 is upper bounded by @xmath85 ( proof ) for any pair of codewords @xmath86 , the hamming weight of @xmath87 is in the range : @xmath88 due to the linearity of @xmath69 .
this means that @xmath89 holds for any @xmath90 . using lemma [ hammingandcorr ]
, we immediately obtain @xmath91,\quad
-\epsilon \le \beta_p(c_i)^t \beta_p(c_j ) \le \epsilon,\ ] ] where @xmath92 the definition of incoherence and the above inequalities lead to an upper bound on the incoherence : @xmath85 we here consider binary linear codes whose weight distribution is tightly concentrated around the hamming weight @xmath81 . before starting the analysis , we introduce the weight distribution @xmath93}$ ] , which is defined by @xmath94 in the present paper , we consider an ensemble of binary parity check matrices , which is referred to herein as the _
random ensemble_. the random ensemble @xmath95 contains all binary @xmath67 matrices .
equal probability @xmath96 is assigned to each matrix in @xmath95 .
let @xmath97 be a real - valued function defined on @xmath95 , which can be considered as a _
random variable _ defined over the ensemble @xmath95 .
the expectation of @xmath97 with respect to the ensemble @xmath95 is defined by @xmath98 { \stackrel{\triangle}{=}}\sum_{h \in r_{r , p } } p(h ) f(h).\ ] ] the expectation of weight distributions with respect to the random ensemble has been reported @xcite to be @xmath99 = { p \choose w } 2^{-r}.\ ] ] in the following , a combination of average weight distribution and markov inequality is used to show that the rip holds for @xmath79 with overwhelmingly high probability .
[ first ] assume that we draw a parity check matrix from @xmath95 .
the probability of selecting @xmath70 that satisfies @xmath100 is lower bounded by @xmath101 ( proof ) let us define @xmath102 as @xmath103 for @xmath104 .
the condition @xmath105 implies that @xmath106 for any @xmath107 .
namely , if @xmath105 holds , then @xmath108 is proven to be smaller than or equal to @xmath109 by lemma [ ripbound ] .
next , we evaluate the ensemble expectation of @xmath102 : @xmath110 & = & \sum_{w=1}^{\lfloor ( \frac{1-\epsilon}{2 } ) p \rfloor } e_{r_{r , p}}[a_w(h ) ] \\ \nonumber & + & \sum_{w=\lceil ( \frac{1+\epsilon}{2 } ) p \rceil}^{p } e_{r_{r , p}}[a_w(h ) ] \\ \nonumber & = & \sum_{w=1}^{\lfloor ( \frac{1-\epsilon}{2 } ) p \rfloor } 2^{-r } { p \choose w } + \sum_{w=\lceil ( \frac{1+\epsilon}{2 } ) p \rceil}^{p } 2^{-r } { p \choose w } \\ & < & 2^{1-r } \sum_{w=0}^{\lfloor ( \frac{1-\epsilon}{2 } ) p \rfloor } { p \choose w } .\end{aligned}\ ] ] the final inequality is due to the following identity on the binomial coefficients : @xmath111,\quad { p \choose w } = { p \choose p - w } .\ ] ] using the markov inequality , we obtain the following upper bound on the probability of the event @xmath112 : @xmath113 & \le & e_{r_{r , p}}[k_\epsilon(h ) ] \\ & < & 2^{1-r } \sum_{w=0}^{\lfloor ( \frac{1-\epsilon}{2 } ) p \rfloor } { p \choose w } .\end{aligned}\ ] ] since @xmath114 takes a non - negative integer - value , we have @xmath115 > 1 - 2^{1-r } \sum_{w=0}^{\lfloor ( \frac{1-\epsilon}{2 } ) p \rfloor } { p \choose w } .\ ] ] this completes the proof .
the following theorem is the main contribution of the present paper .
[ main ] assume that @xmath70 is chosen randomly according to the probability assignment of @xmath95 . if @xmath116 holds , then @xmath117 holds with probability greater than @xmath118 where @xmath78 .
the constant @xmath119 is given by @xmath120 ( proof ) a simpler upper bound on @xmath121 is required .
using the inequality on binomial coefficients @xcite : @xmath122 we have @xmath123 where @xmath124 is the binary entropy function defined by @xmath125 in order to consider the exponent of an upper bound , we take the logarithm of ( [ simpler ] ) and obtain an upper bound of the exponent : @xmath126 \hspace{-4mm}&<&\hspace{-2mm}1 + \log_2 ( m+1 ) - p + \log_2 p \\ \label{exponentp } & + & \hspace{-2mm}p h\left(\frac{1-\epsilon}{2 }
\right ) \\ & < &
\hspace{-2mm}1 + 2\log_2 ( m+1 ) - \frac{1}{2 } p \epsilon^2 .
\end{aligned}\ ] ] in the above derivation , we used the relation @xmath127 and the assumption @xmath68 . a quadratic upper bound on the binary entropy function ( lemma [ entropybound ] in appendix )
was also exploited to bound the entropy term .
letting @xmath128 we have @xmath129 lemma [ deltaupper ] and lemma [ first ] imply that , in this case , @xmath130 holds with probability greater than @xmath131 . due to lemma [ deltaupper ] , the @xmath0-recovery condition ( [ sqr2 ] ) can be written as @xmath132 from this inequality , we have @xmath133 which proves the claim of the theorem . in this subsection ,
the asymptotic properties of the proposed construction are given .
[ second ] assume that we draw a parity check matrix from @xmath95 .
the probability of selecting @xmath70 that satisfies @xmath100 is upper bounded by @xmath134 ( proof ) here , we use a variant of chebyschev s inequality @xcite : @xmath135 \le \frac{var_{r_{r , p}}(k_{\epsilon}(h))}{e_{r_{r , p}}[k_{\epsilon}(h ) ] ^2},\ ] ] where @xmath136 denotes the variance with respect to @xmath95 . the variance @xmath137 is given by @xmath138 where @xmath139 and @xmath140 .
the covariance of weight distributions denoted by @xmath141 is defined as follows : @xmath142 \\ & - & e_{r_{r , p}}[a_{w_1}(h ) ] e_{r_{r , p}}[a_{w_2}(h ) ] \end{aligned}\ ] ] for @xmath143 $ ] .
the covariance for the random ensemble has the following closed formula @xcite : @xmath144 for @xmath143 $ ] .
applying the covariance formula to ( [ covcov ] ) , we have @xmath145 plugging the expectation of @xmath102 @xmath110 & = & 2^{-r}\left ( \sum_{w=1}^{a } { p \choose w } + \sum_{w = b}^{p } { p \choose w } \right ) \\ & = & 2^{-r}\left ( 2\sum_{w=0}^{a } { p \choose w } -1 \right ) \end{aligned}\ ] ] and the upper bound on the variance ( [ covbound ] ) into ( [ chebyschev ] ) proves the lemma
. the asymptotic behavior of @xmath146 $ ] and @xmath147 $ ] is summarized in the following theorem .
[ asymptoticth ] assume that @xmath148 is fixed @xmath149 .
let @xmath150 \\
f_2(\epsilon , \alpha ) & { \stackrel{\triangle}{= } } & \lim_{p \rightarrow \infty } \frac 1 p \log_2 prob[k_\epsilon(h)\ne0].\end{aligned}\ ] ] the following inequalities give upper bounds on @xmath151 and @xmath152 , respectively : @xmath153 ( proof ) we first discuss ( [ f1 ] ) .
let @xmath154 using the inequality on the binomial coefficients @xmath155 @xmath156 can be bounded from below : @xmath157 the inequality ( [ secondmoment ] ) can be simplified as @xmath158 for sufficiently large @xmath156 .
the right - hand side of the above inequality can be bounded from above using ( [ entropylower ] ) : @xmath159 we are now able to derive the inequality given in ( [ f1 ] ) as follows : @xmath160 \\ & = & \alpha - h\left(\frac{1-\epsilon}{2 } \right).\end{aligned}\ ] ] the inequality given in ( [ f2 ] ) is readily obtained from ( [ exponentp ] ) .
theorem [ asymptoticth ] implies a sharp threshold behavior in the asymptotic regime .
let @xmath161 be @xmath162 which is referred to as the _ critical exponent_. if @xmath163 , ( [ f1 ] ) means that the probability to draw a @xmath164 matrix with @xmath100 decreases exponentially as @xmath9 goes to infinity . on the other hand , ( [ f2 ] ) indicates that the probability _ not _ to select a matrix with @xmath100 decreases exponentially if @xmath165 .
in the present paper , a construction of a bipolar sensing matrix is introduced and its rip is analyzed .
the existence of sensing matrices with the rip has been shown based on a probabilistic argument .
an advantage of this type of sensing matrix is its compactness .
a sensor requires @xmath166-bits in order to store a truly random @xmath1 bipolar matrix . on the other hand , we need only @xmath167-bits to store @xmath79 because we can use a generator matrix of @xmath69 as a compact representation of @xmath69 . however , this limited randomness of matrices results in a penalty on the rip constant .
although the present construction is based on a probabilistic construction , the results shown in theorem [ main ] are weaker than the @xmath0-recovery condition @xmath168 for the truly random @xmath1 bipolar matrix ensemble shown in @xcite .
the condition shown in theorem [ main ] can be written as @xmath169 and is more similar to the conditions of deterministic constructions , such as that given in @xcite .
lemma [ ripbound ] may be useful for evaluating the goodness of a randomly generated instance .
the weight distribution of @xmath69 can be evaluated with time complexity @xmath170 , and an upper bound on the rip constant can be obtained using lemma [ ripbound ] .
[ entropybound ] the following inequality approaches to @xmath171 . ]
@xmath172 holds for @xmath173 .
+ ( proof ) let @xmath174 be @xmath175 the domain of which is @xmath176 .
the first and second derivatives of @xmath174 are given by @xmath177 and @xmath178 respectively .
it is easy to verify that @xmath179 for @xmath176 , which indicates that @xmath174 is convex .
thus , we can obtain the global minimum of @xmath174 by solving @xmath180 , and we have @xmath181 and @xmath182 .
let @xmath36 be an index set satisfying @xmath183 .
for any @xmath184 , we have @xmath185 where @xmath186 is a sub - matrix of @xmath187 composed from the columns corresponding to the index set @xmath36 .
for any @xmath188 , @xmath189 holds since @xmath190 we use this inequality to bound @xmath191 in ( [ cicj ] ) and obtain @xmath192 similarly , @xmath193 can be lower bounded by @xmath194 from the definition of @xmath35 , the lemma is proven .
the author would like to thank the anonymous reviewers of ieee information theory workshop 2009 for their constructive comments .
the present study was supported in part by the ministry of education , science , sports and culture of japan through a grant - in - aid for scientific research on priority areas ( deepening and expansion of statistical informatics ) 180790091 .
e.candes , j. romberg and t.tao , `` robust uncertainty principles : exact signal reconstruction from highly incomplete frequency information , '' ieee trans . on information theory ,
vol.52(2 ) , pp . 489 509 , 2006 .
t.wadayama , `` on undetected error probability of binary matrix ensembles '' , in proceedings of ieee international symposium on information theory ( isit2008 ) , tronto ( 2008 ) ( related preprint : arxiv:0705.3995 ) . | a random construction of bipolar sensing matrices based on binary linear codes is introduced and its rip ( restricted isometry property ) is analyzed based on an argument on the ensemble average of the weight distribution of binary linear codes . |
we will show that the radiative efficiency of an accreting black hole ( @xmath0 , @xmath7 is the luminosity , @xmath8 is the mass accretion rate ) depends on the magnetic field strength .
an investigation similar to ours , but with different conclusions , was carried out by quataert ( 1997 ) .
it was suggested that at low accretion rates , @xmath9 , a two - temperature plasma with @xmath10 forms near the schwarzschild radius of an accreting black hole ( shapiro et al 1976 , phinney 1981 , rees et al 1982 , narayan & yi 1995 ) , and the disk is inflated into a torus by the ion pressure .
the plasma in the torus is collisionless , and coulomb energy exchange between electrons and ions is small .
viscous heating power must be @xmath11 .
if nearly all the heat is released into ions , radiative efficiencies as low as @xmath2 are obtained , because ions take all the heat into the hole .
electrons radiate efficiently , and @xmath12 can not be much smaller than @xmath13 due to adiabatic heating of electrons .
observational estimates for @xmath7 and @xmath8 exist for a number of objects ( narayan et al 1997 , fabian & rees 1995 ) , and some black hole candidates seem to have @xmath14 .
we show here that if indeed @xmath14 , magnetic fields in these accretion flows must be very weak , @xmath1 , where @xmath15 is the gas ( ion ) pressure . for stronger magnetic fields , @xmath4 because electrons receive more than 0.1% of the viscous heat , and this part of the viscous heat is efficiently radiated . at equipartition fields ( @xmath16 ) ,
electrons get most of the heat and the radiative efficiency is close to the standard thin disk estimate , @xmath17 .
theoretical models of black hole accretion that assume both @xmath5 and @xmath14 should be abandoned .
the reason why stronger magnetic fields lead to an increased fraction of viscous energy going into electrons ( and correspondingly to larger radiative efficiency ) is roughly as follows . at strong fields , turbulence in the torus
is dominated by alfven waves which are only weakly damped at the length scales @xmath18 ( the ion larmor radius ) , where damping by ions is maximal .
turbulent energy flux gets through the ion damping barrier at scales @xmath18 and cascades to smaller scales . at length scales
@xmath19 , ions drop out of dynamics because of the larmor circle averaging , and the energy flux heats electrons .
the constraint @xmath20 ( for @xmath14 ) is of interest for modeling the spectrum of an accreting black hole .
the radiation spectrum is determined by the synchrotron emission and its comptonization ; the synchrotron peak frequency @xmath21 changes by a factor of 30 when our estimate of @xmath22 is changed from 1 ( which is the standard assumption ) to @xmath23 .
unless the plasma somehow leaves precisely all of its angular momentum behind by the time it reaches the hole , the ion torus will be rotating at about the keplerian speed .
to accrete , the plasma has to transport its angular momentum outwards ( in a frame moving radially with the plasma ) , and this means that hydromagnetic turbulence should develop in the torus .
whatever the nature of turbulence and the resulting turbulent viscosity are , viscous heat will be generated at a rate @xmath24 ( e.g. , shapiro & teukolsky 1983 ) .
the question is into which plasma species is this heat deposited .
the ions will take their share into the hole , while most of the electrons share will be radiated .
narayan et al ( 1997 ) and mahadevan ( 1997 ) suggested that the ratio of electron to ion heating , @xmath25 , is given by the mass ratio , @xmath26 .
we will show that such small ratios of electron to ion heating are possible only at very weak magnetic fields .
there are two particle heating mechanisms in a magnetized collisionless plasma - landau heating by electric fields parallel to the local magnetic field and landau heating by the time varying magnetic field ( by the curl of the electric field parallel to the local magnetic field , e.g. stix ( 1992 ) ) . in the advection dominated regime ( narayan & yi 1995 ) near the hole , thermal velocities of electrons and ions are @xmath27 , yet the @xmath28-factors are not large , so that the plasma is roughly non - relativistic . at comparable thermal velocities
landau heating by parallel electric fields is proportional to @xmath29 , and hence it is dominated by electrons , the ratio of electron to ion heating being @xmath30 . for magnetic heating , the charge @xmath31 should be replaced by the magnetic moment , @xmath32 , where @xmath33 . the heating ratio is @xmath34 , as narayan and coworkers usually assume .
electron to ion heating ratio will be much greater than @xmath23 if heating by the parallel electric fields is important .
ion heating is also suppressed , and correspondingly electron to ion heating ratio is increased , at small length scales when averaging of the lorentz force over the ion larmor circle becomes important . to calculate the ratio of electron to ion heating , we must follow the turbulent energy as it cascades from the largest energy containing scales to smaller scales and estimate the strengths of the two damping mechanisms for all scales .
since random magnetic fields limit the mean free path of plasma species to corresponding larmor radii , hydrodynamic description is applicable at large scales , and kolmogorov energy cascade develops . at smaller scales , turbulence is dominated by alfven waves , because the nonlinear frequency @xmath35 ( @xmath36 is the energy flux in momentum space , units @xmath37 ) becomes smaller than the linear alfven frequency @xmath38 ( @xmath39 is the alfven speed ) . following goldreich & sridhar ( 1995 ) ( appendix a )
, we assume that for @xmath40 , when alfven frequency is formally larger than the nonlinear frequency , turbulence develops anisotropy , with filamentary perturbations aligned with the local magnetic field , @xmath41 .
degree of anisotropy is determined by a requirement that the spectrum be critically balanced , meaning that hydrodynamic and alfven frequencies are comparable , @xmath42 . in the critically balanced alfven turbulence regime
, it takes about @xmath43 for a turbulent eddy to cascade to twice its wavenumber .
let @xmath44 be the landau damping rate at a given scale ( calculated in appendix b ) .
the fraction of energy flux intercepted at wavenumber scales @xmath45 is of order @xmath46 - the rest of energy goes to scales of order @xmath47 .
the ratio @xmath48 ( damping to real frequency for linear waves ) is all we need .
this ratio determines at what scales the energy flux is damped , and knowing this it is straightforward to predict which plasma species is heated .
figure 1 shows @xmath48 as a function of the perpendicular wavenumber . as seen from the figure , significant fraction of the energy
is damped near the ion larmor radius , @xmath49 , when magnetic fields are weak , @xmath50 for @xmath51 the main damping mechanism of alfven waves is magnetic heating of ions .
the damping rate is given by ( appendix b , for @xmath52 ) @xmath53 at @xmath54 , alfven waves gradually modify into whistlers ( appendices b and c ) , the ions still take most of the heat load , and for @xmath52 , @xmath55 according to the last formula , ion damping becomes small at @xmath56 few .
turbulent energy that gets through the ion damping barrier at @xmath49 is only weakly damped until it reaches very small scales to be damped by electrons ( see appendix c for a discussion of whistler turbulence ) .
ions drop out of the picture at @xmath57 few because of larmor circle averaging .
the ion heating maximum at @xmath49 determines the ratio of electron to ion heating . at equipartition magnetic fields ( @xmath58 ) ,
the maximal value of ion damping is @xmath59 at @xmath49 and much smaller at other scales .
most of the viscous heat gets over the ion damping barrier and is ultimately damped in electrons .
figure 2 shows the turbulent energy flow in the wavenumber space ( arrows are drawn by hand ! ) .
the corresponding radiative efficiency is @xmath17 , similar to thin accretion disk models .
the very small radiative efficiency assumed in advection dominated accretion models , @xmath2 , is possible only if a very small fraction , @xmath60 , of the alfven wave energy at @xmath61 is transmitted to whistlers at @xmath56 few .
whistlers can propagate at @xmath62 as small as @xmath63 . for @xmath64 ,
alfven waves are not fully damped at @xmath65 and will excite whistlers by three - wave interactions .
thus , @xmath20 is necessary for @xmath14 .
fortunately , electron parameters ( @xmath66 and @xmath67 ) do not enter eqs .
( 1 ) and ( 2 ) .
electron thermal velocity and @xmath28-factor are determined by radiation and are not reliably predicted by the advection dominated ( or any other ) accretion model .
however , the only parameter that matters for our calculation is @xmath22 - an input parameter .
our predictions are not limited to any particular model of accretion , as long as @xmath10 and @xmath68 .
a minimal model of turbulence was assumed in this work - no secondary instabilities , no shocks , just a simple kolmogorov cascade with landau damping .
this model predicts substantial heating of electrons and large radiative efficiencies if magnetic fields are close to equipartition .
small radiative efficiency is possible if nearly all the viscous heat is absorbed by ions . for this
, magnetic fields must be very weak , @xmath20 .
incidentally , we can argue that low radiative efficiency is possible only at very weak magnetic fields if we assume that blandford & znajek ( 1977 ) mechanism is at work .
poloidal magnetic field of order @xmath69 near a maximally rotating black hole gives a luminosity @xmath70 .
we get a radiative efficiency @xmath71 , where @xmath72 is the shakura - sunayev parameter . arbitrarily assuming @xmath73 , @xmath74 , we get @xmath75 , so that @xmath76 is required for @xmath77 .
i thank john bahcall and martin rees for suggesting the problem , and peter goldreich , pawan kumar and insu yi for valuable discussions .
this work was supported by nsf phy-9513835 .
this appendix is based on goldreich & sridhar ( 1995 ) .
suppose that @xmath36 is the energy flux in the wavenumber space ( energy pumping rate , units @xmath37 ) .
isotropic kolmogorov turbulence is characterized by a nonlinear frequency @xmath78 at wavenumbers @xmath79 .
if there is a large scale magnetic field with alfven speed @xmath39 , another characteristic frequency of the turbulence is the linear frequency @xmath80 .
if the turbulence were isotropic , the hydrodynamic nonlinear frequency would be much smaller than the linear alfven frequency at small enough scales . according to goldreich & sridhar
this does not happen . instead , turbulent energy propagates to smaller scales within a narrow cone in the @xmath45-space given by @xmath81 . in other words ,
turbulence is dominated by elongated structures , that are aligned with the local magnetic field .
this could be roughly explained as follows .
suppose , that a `` turbulent eddy '' leaves the cone .
on the outside of the cone , turbulence is just a collection of weakly nonlinear waves .
waves propagating parallel and antiparallel to the local magnetic field follow the field lines perturbed by each other , and thus cascade in @xmath82
. turbulent energy from outside of the cone will cascade in @xmath82 without cascading in @xmath83 and will thus enter the cone .
turbulence fills the entire cone , because within the cone the presence of the large - scale magnetic field is not `` felt '' , @xmath84 .
turbulent eddies will try to isotropize themselves , filling as much of the @xmath45-space as possible .
we calculated landau damping of alfven waves ( whistlers at @xmath54 ) numerically using a full dielectric permittivity tensor of a magnetized maxwellian plasma ( e.g. mikhailovskii 1967 ) .
results are shown in fig.1 , where the ratio of imaginary to real parts of the eigenfrequency is plotted versus dimensionless perpendicular wavenumber @xmath62 .
the calculation was carried out for maxwellian electrons and ions with thermal velocities @xmath85 and @xmath86 which are close to the plasma parameters of a two - temperature advection dominated accretion model of narayan & yi ( 1995 ) .
we assumed @xmath87 , and checked that @xmath88 does not depend on @xmath89 as long as @xmath90 .
the four curves correspond to four values of @xmath22 .
damping by ions is maximal at @xmath49 , where damping by electrons is negligible . for strong fields , @xmath16 ,
the damping at @xmath49 is small . the damping increases with decreasing @xmath22 .
at @xmath91 , the damping is so strong that alfven waves do not propagate in a narrow region near @xmath92 .
the region of forbidden propagation is narrow for @xmath91 and widens for @xmath93 .
two remarks are in order .
electrons and ions that resonate with alfven waves and whistlers belong to the bulk of the distribution functions .
deviations from maxwellian distribution , which may be expected in our collisionless plasma , will not strongly affect the results . also , linear and nonlinear landau damping rates are similar , because linear frequencies and nonlinear frequency shifts are of the same order for a critically balanced turbulence .
most of the numerical results shown in fig.1 can be understood analytically . as discussed in
2 , analytical expressions ( 1 ) and ( 2 ) are of importance - they imply that our results are model independent . for @xmath41 , @xmath94 , the following approximate expressions for the permittivity tensor ( mikhailovskii 1967 ) give a good accuracy . for a wavenumber in the @xmath95-plane , with @xmath96 along the local magnetic field ,
@xmath97 here @xmath98 is the phase speed , @xmath99 , @xmath100 are modified bessel functions , and @xmath101 permittivities @xmath102 and @xmath103 are due to ions , @xmath104 is due to electrons , and @xmath105 is due to both species . the dispersion relation for linear waves follows from maxwell equations , @xmath106 for @xmath107 , the parallel permittivity @xmath104 is very large , and both damping and real frequency are accurately predicted by a two dimensional approximation @xmath108 for @xmath109 , we have @xmath110 , and eq .
( b9 ) reduces to @xmath111 , which gives the alfven wave dispersion law @xmath112 for @xmath113 , we have @xmath114 , @xmath115 , and eq .
( b9 ) gives @xmath116 , that is whistlers @xmath117 corrections to the above dispersion laws ( within the 2d approximation , eq .
( b9 ) ) give the damping rates for alfven waves and whistlers .
for alfven waves , in the limit @xmath118 when @xmath103 is pure imaginary , we have approximately @xmath119 giving @xmath120 that is @xmath53 for whistlers at @xmath54 , we have approximately @xmath121 giving @xmath55 at even larger @xmath62 electron damping becomes important .
the ratio of electron to ion damping can be estimated ( for @xmath118 ) as @xmath122 here the wave electric fields should be obtained from the system of linear equations whose determinant is given in eq .
two of these equations can be approximately written as @xmath123 @xmath124 now eq .
( b17 ) gives @xmath125 showing that electron heating dominates for @xmath126 . at these small scales ,
ions are not important either for real part of the frequency or its imaginary part ( heating ) . for perturbations with perpendicular length scales
@xmath127 , ions can be treated as a constant positive background for electrons and electromagnetic fields .
we should make sure that turbulent energy will keep cascading to smaller scales after alfven waves are transformed into whistlers at length scales @xmath18 .
like collisionless alfven waves , collisionless whistlers can be described hydrodynamically .
hydrodynamic equations valid at length scales @xmath128 can be obtained as follows .
electrons move freely along magnetic field lines , and @xmath129 . in the perpendicular direction ,
electrons move with the e@xmath130b - drift velocity , @xmath131 .
these two equations give @xmath132 neglecting displacement currents , @xmath133 from eqs .
( c1 ) and ( c2 ) , we obtain the electron hydrodynamics ( ehd ) equation ( kingsep , chukbar , & yankov 1990 ) @xmath134 linear waves in eq .
( c3 ) are whistlers , their dispersion law can be presented as eq .
turbulent cascade in eq .
( c3 ) was discussed by kingsep , chukbar , & yankov ( 1990 ) and by goldreich & reisenegger ( 1992 ) .
blandford , r. d. , & znajek , r. l. 1977 , mnras , 179 , 433 goldreich , p. , & reisenegger , a. 1992 , apj , 395 , 250 goldreich , p. , & sridhar , s. 1995 , apj , 438 , 763 fabian , a. c. , & rees , m. j. 1995 , mnras , 277 , l55 kingsep , a. s. , chukbar , k. v. , & yankov , v. v. 1990 , rev .
plasma phys . ,
16 , 243 mahadevan , r. 1997 , apj , 477 , 585 mikhailovskii , a. b. 1967 , rev .
plasma phys .
, 3 , 159 narayan , r. , & yi , i. 1995 , apj , 452 , 70 narayan , r. , mahadevan , r. , grindlay , j. e. , popham , r. g. , & gammie , c. 1997 , apj , submitted ( astro - ph/9706112 ) phinney , e. s. 1981 , in plasma astrophysics , ed .
t. d. guyenne , & g. levy ( esa sp-161 ) , 337 quataert , e. 1997 , apj , submitted .
rees , m. j. , begelman , m. c. , blandford , r. d. , & phinney , e. s. 1982 , nature , 295 , 17 .
shapiro , s. l. , lightman , a. p. , eardley , d. m. 1976 , apj , 204 , 187 shapiro , s. l. , & teukolsky , s. a. 1983 , black holes , white dwarfs , and neutron stars ( new york : john wiley & sons ) stix , t. h. 1992 , waves in plasmas ( new york : american institute of physics ) | radiative efficiency , @xmath0 , of a slowly accreting black hole is estimated using a two - temperature model of accretion .
the radiative efficiency depends on the magnetic field strength near the schwarzschild radius . for weak magnetic fields , @xmath1 , the low efficiency @xmath2 assumed in some theoretical models
is achieved . for @xmath3 , a significant fraction of viscous heat
is dissipated by electrons and radiated away resulting in @xmath4 . at equipartition magnetic fields , @xmath5
, we estimate @xmath6 . |
let @xmath0 be a complete hyperbolic surface of infinite type whose boundary components , if any , are closed geodesics or punctures .
assume that there exists a pants decomposition @xmath5 of @xmath0 by simple closed geodesics such that their lengths are bounded from the above by a fixed constant @xmath6 .
we say that such a pants decomposition is _ upper bounded_. by @xcite , any other pants decomposition of @xmath0 by simple closed curves can be straightened to a pants decomposition by simple closed geodesics .
the _ quasiconformal _ teichmller space @xmath4 consists of all quasiconformal maps @xmath7 up to post composition by isometries and up to bounded homotopies which setwise fix boundary components of @xmath2 .
note that bounded homotopies do not fix boundary geodesics pointwise since the distance between any two points ( on a boundary geodesic ) is finite .
thus @xmath4 is the reduced quasiconformal teichmller space of the surface @xmath0 .
the _ length spectrum _ teichmller space @xmath1 consists of all homeomorphisms @xmath8 up to post compositions by isometries and up to bounded homotopies that setwise preserve the boundary components of @xmath2 such that @xmath9 where the supremum is over all simple closed curve @xmath10 on @xmath0 , and where @xmath11 are the lengths of the geodesic representatives of @xmath10 on @xmath12 , respectively .
note that @xmath13 because for each simple closed geodesic @xmath10 on @xmath0 and a @xmath14-quasiconformal map @xmath7 we have ( cf .
wolpert @xcite ) @xmath15 the _ length spectrum metric _
@xmath16 is defined by @xmath17 for @xmath18 .
shiga @xcite was the first to study the length spectrum metric on quasiconformal teichmller spaces of surfaces of infinite type and he proved that @xmath16 in general is not complete on @xmath4 .
this implies that @xmath4 could be a proper subset of @xmath1 . for a fixed upper
bounded pants decomposition @xmath5 on @xmath0 , the assignment of fenchel - nielsen coordinates @xmath19 to each @xmath20 or @xmath21 completely determines the ( marked ) surface @xmath2 .
alessandrini , liu , papadopoulus , su and sun @xcite proved that the fenchel - nielsen coordinates for @xmath4 satisfy @xmath22 and @xmath23 , and that the map from @xmath4 to the fenchel - nielsen coordinates is a locally bilipschitz homeomorphism onto @xmath3 .
alessandrini , liu , papadopuolus and su @xcite proved that the fenchel - nielsen coordinates for @xmath21 satisfy @xmath24 and @xmath25 moreover , they proved that if @xmath0 ( equipped with an upper bounded pants decomposition ) contains a sequence of simple closed geodesics whose lengths go to @xmath26 then @xmath27 and that @xmath4 is nowhere dense in @xmath1 ( cf .
@xcite ) .
let @xmath28 , called the _
normalized fenchel - nielsen map _ , be defined by @xmath29 where if @xmath30 is a boundary geodesic of @xmath0 then we take only the length component of the fenchel - nielsen coordinates . by @xcite ,
the fenchel - nielsen coordinates give a locally bilipschitz homeomorphism between the ( reduced ) quasiconformal teichmller space @xmath4 and @xmath3 which implies that @xmath4 is contractible .
fletcher @xcite used complex analytic methods to prove that the unreduced quasiconformal teichmller space is locally bilipschitz to @xmath3 .
our main result is ( cf .
2 , theorem [ thm : main ] and corollary [ cor : contractible ] ) : .2 cm [ [ theorem-1 . ] ] * theorem 1 . *
+ + + + + + + + + + + + _ the normalized fenchel - nielsen map @xmath31 is a locally bilipschitz homeomorphism . _ in particular
, the length spectrum teichmller space @xmath1 is contractible .
.2 cm thus @xcite , @xcite and theorem 1 imply that the unreduced quasiconformal teichmller space , the reduced quasiconformal teichmller space and the length spectrum teichmller space are locally bilipschitz to @xmath3 and thus to each other . a problem of characterizing the closure of @xmath4 inside @xmath1 for the length spectrum metric @xmath16 is raised in @xcite .
we use the fenchel - nielsen coordinates to characterize the closure @xmath32 of @xmath4 .
.2 cm [ [ theorem-2 . ] ] * theorem 2 .
* + + + + + + + + + + + + _ let @xmath0 be an infinite type hyperbolic surface with an upper bounded pants decomposition @xmath5 .
then @xmath33 if and only if @xmath24 and @xmath34 as @xmath35 . _
we prove that the normalized fenchel - nilesen map is a localy bilipschitz homeomorphism onto @xmath3 .
[ thm : main ] let @xmath0 be an infinite type complete hyperbolic surface equipped with an upper bounded geodesic pants decomposition @xmath36 .
the normalized fenchel - nielsen map @xmath37 for @xmath21 , induces a locally bilipschitz surjective homeomorphism @xmath38 let @xmath39 be such that @xmath40 for each @xmath41 .
.1 cm * step i : * we establish that @xmath42 which is already proved in @xcite .
we give another proof in order to facilitate the rest of the argument . by the definition ,
@xmath21 if there is @xmath43 such that @xmath44 for each simple closed curve @xmath45 on @xmath0 .
in particular @xmath46 is a bounded sequence .
it remains to bound the twists .
the choice of the twists of @xmath0 on @xmath30 are determined up to integer multiples of @xmath47 . without loss of generality ,
we normalize them such that , for each @xmath48 , @xmath49 given this normalization , it is enough to prove that @xmath50 is bounded uniformly in @xmath48 .
using @xcite , there exists a surface @xmath51 which is @xmath14-quasiconformal to @xmath0 such that @xmath52 for all @xmath48 , where @xmath53 ( cf .
@xcite ) . the @xmath14-quasiconformal map @xmath54 maps each pair of pants @xmath55 of @xmath0 onto a geodesic pair of pants of @xmath51 such that on each boundary geodesic the map is affine .
divide each geodesic pair of pants into two right angled hexagons by three _ seams _ , namely three geodesic arcs connecting pairs of boundary curves and orthogonal to them .
each hexagon contains half of each boundary geodesic of the pair of pants which are called _ a - sides_. the other three sides of the hexagon which are seams are called _ b - sides_. the two hexagons are glued along their b - sides to obtain the pair of pants and the pairwise union of their a - sides forms three geodesic boundaries of the pair of pants .
the map @xmath54 maps a - sides of each hexagon of each pair of pants of @xmath0 onto a - sides of hexagons of the corresponding pair of pants of @xmath2 and it is affine on the a - sides .
note that a single @xmath41 is on the boundary of two pairs of pants @xmath56 and @xmath57 of @xmath58 which implies that @xmath30 is divided into a - sides with respect to both @xmath56 and @xmath57 .
the two divisions of @xmath30 into a - sides do not match in general and the distance between the endpoints of two a - sides coming from two pairs of pants is the twist parameter @xmath59 of @xmath0 at the closed geodesic @xmath30 for the pants decomposition @xmath58 .
the map @xmath54 is affine on each @xmath30 , it maps the foots of the seams of the pair of pants @xmath60 to the foots of the corresponding seams of @xmath61 for @xmath62 and it does not introduce any full twisting along @xmath30 by its construction ( cf .
@xcite , @xcite ) .
then for @xmath53 we have @xmath63 let @xmath64 .
then @xmath2 is obtained by a(n infinite ) multi twist on @xmath51 along the family @xmath5 by the amount @xmath65 .
it is enough to prove that @xmath66 is bounded in terms of @xmath67 because @xmath68 .
this is proved in @xcite using results of minsky @xcite and choi - rafi @xcite .
we give a more direct proof of this result below .
fix a cuff @xmath30 and let @xmath69 and @xmath70 be two geodesic pairs of pants with common boundary @xmath30 .
either @xmath71 or @xmath72 and we divide the argument into these two cases .
* assume that @xmath71 .
there exists a unique geodesic arc @xmath73 , for @xmath62 , which starts and ends at @xmath30 that is orthogonal to @xmath30 at both of its endpoints .
let @xmath74 be a closed curve on @xmath51 obtained by concatenating @xmath75 followed by an arc of @xmath30 from an endpoint of @xmath75 to an endpoint of @xmath76 in the direction of the left twist along @xmath30 followed by @xmath76 followed by an arc of @xmath30 connecting other two endpoints of @xmath75 and @xmath76 in the direction of the right twist ( cf .
figure 1 ) . .
] we will give an upper bound for @xmath77 in terms of @xmath78 .
fix three consecutive lifts @xmath79 , for @xmath80 , of @xmath30 under the universal covering @xmath81 .
let @xmath82 be the lift of the geodesic representative @xmath83 of @xmath74 that intersects @xmath79 , for @xmath80 .
moreover , let @xmath84 be the lift of @xmath85 that connects @xmath86 and @xmath87 ( cf .
figure 2 ) .
let @xmath88 and @xmath89 , and @xmath90 .
the lengths satisfy @xmath91 because @xmath2 is obtained from @xmath51 by a multi twist along @xmath92 .
we either have @xmath93 or @xmath94 . consider the case that @xmath95 and the other case is analogous .
let @xmath96 and let @xmath97 be the foot of the orthogonal from @xmath98 to @xmath99 . consider the quadrilateral with vertices @xmath98 , @xmath100 , @xmath101 and @xmath102 ( cf . figure 2 ) .
we get @xmath103 by the collar lemma @xcite , there exists @xmath104 such that @xmath105 since @xmath106 and @xmath107 , and by ( [ eq : * * ] ) and ( [ eq : * * * ] ) , we have @xmath108 which implies @xmath109 note that by construction @xmath110 we estimate @xmath111 from the above using right - angled pentagons .
namely each hexagon of @xmath60 contains a half of the arc @xmath85 and @xmath85 intersects the b - side of the hexagons that connects the two boundary geodesics of @xmath60 different from @xmath30 .
then both hexagons of @xmath60 are divided into two right - angled pentagons by @xmath85 .
the sides of the obtained pentagons are as follows in the cyclic order : a portion of the a - side on @xmath30 , followed by the half of @xmath112 , followed by a portion of a b - side , followed by an a - side on a boundary curve different from @xmath30 and followed by a b - side .
we choose one of the two pentagons such that the portion of the a - side has length at least @xmath113 .
since @xmath40 it follows that any a - side of the hexagon has length at most @xmath114 and a hyperbolic formula for the right - angled pentagons gives @xmath115 the inequality implies that there is @xmath116 such that @xmath117 which in turn implies @xmath118 for some constant @xmath119 . by ( [ eq : * ] ) ,
( [ eq:4 ] ) and ( [ eq : l ] ) we have @xmath120\\max\ { 1,|\log l_{\alpha_n}(x_0)|\}\ ] ] which gives @xmath121 and this finishes the proof of @xmath42 in the case @xmath122 .
* case 2 . *
assume that @xmath123 .
we define @xmath124 to be the unique geodesic arc in @xmath125 starting and ending at @xmath30 and orthogonal to @xmath30 at both of its endpoints .
then we define a closed curve @xmath74 to consists of @xmath124 and an arc of @xmath30 of the size @xmath126 .
the above argument applies to this case as well .
.1 cm * step ii : @xmath28 is surjective . * for @xmath127 the surface @xmath128 obtained by gluing the pants with prescribed cuffs and twists is complete . the marking map for @xmath128 can be chosen to be a homeomorphism because each twist is realized in an annulus containing a given cuff .
let @xmath129 be the surface obtained by a @xmath14-quasiconformal map @xmath130 such that @xmath131 for all @xmath48 as before ( cf .
@xcite ) . then @xmath132 for @xmath48 .
the surface @xmath128 is obtained by a multi twist around @xmath5 by the amount @xmath133 .
note that @xmath134 . by the collar lemma @xcite
, we have that @xmath135 for each simple closed curve @xmath10 on @xmath0
. then @xmath136 and @xmath137 which together with ( [ eq : str ] ) implies that @xmath138 for each @xmath139 . in the exactly the same fashion
, we obtain @xmath140 for each @xmath139 .
thus @xmath141 and @xmath28 is onto .
.1 cm * step iii : @xmath28 is localy lipschitz . *
let @xmath142 be fixed and let @xmath143 be two arbitrary points in the ball of radius @xmath144 centered at @xmath142 .
consequently @xmath145 .
note that @xmath146 for each @xmath48 .
it remains to consider @xmath147 .
by @xcite and @xcite , there exists a @xmath148$]-quasiconformal map @xmath149 such that @xmath150 for each @xmath48 with @xmath151 .
let @xmath152 be such that there exists an integer @xmath153 with @xmath154 note that , for @xmath155 , we have @xmath156 ) and ( [ eq:6 ] ) because normalized twists @xmath157 are bounded from the above and @xmath158 the construction of @xmath149 from @xcite implies that @xmath159 we estimate @xmath160 .
let @xmath161 .
then we have @xmath162 which implies @xmath163 note that @xmath164 which implies that @xmath165 let @xmath166 the surface @xmath167 is obtained from @xmath51 by a multi twist along @xmath92 by the amount @xmath65 .
as before , we divide the argument into two cases : @xmath72 and @xmath71 .
* case 1 . * given @xmath30 , assume first that the pairs of pants with boundary @xmath30 are equal , namely @xmath72 .
let @xmath74 be a closed curve obtained by concatenating the unique arc @xmath124 in @xmath168 orthogonal to @xmath30 at both of its endpoints followed by an arc on @xmath30 of the length @xmath169 .
let @xmath83 be the geodesic representative of @xmath74 . denote by @xmath170 the hyperbolic surface obtained by twisting the amount @xmath171 , for @xmath172 and @xmath173 defined by ( [ eq : relative_t_n ] ) , along the cuffs @xmath30 on the surface @xmath51 .
note that @xmath174 and that by the definition of @xmath173 we have that @xmath175 .
recall that ( cf .
@xcite , @xcite ) @xmath176 where @xmath177 is the angle between @xmath178 and @xmath179 .
let us fix @xmath180 .
note that @xmath181 is either increasing or decreasing from @xmath182 to @xmath183 in @xmath184 for @xmath185 ( depending whether @xmath173 is positive or negative ) due to the fact that the geodesic length along a left earthquake with support @xmath30 is a convex function ( cf .
@xcite , @xcite ) .
assume that @xmath186 . if @xmath187 ( which implies @xmath188 for @xmath189 ) then we set @xmath190 . if @xmath191 then we choose @xmath192 such that @xmath193 as follows .
consider universal covering @xmath194 such that one lift @xmath179 of @xmath30 is the @xmath195-axis .
further we arrange that two lifts @xmath196 and @xmath197 of the arc @xmath124 that are adjacent to the @xmath195-axis from the left and the from the right meet the @xmath195-axis between @xmath198 and @xmath199 .
let @xmath200 be an endpoint on @xmath201 of the hyperbolic geodesic containing @xmath196 and let @xmath202 be an endpoint on @xmath201 of the geodesic containing @xmath197 .
for any @xmath153 , a @xmath203 full left twists on @xmath30 on the surface @xmath51 maps the curve @xmath83 to a new curve @xmath192 .
the curve obtained by the concatenating the arc @xmath124 with the arc which winds around @xmath30 @xmath203-times plus the shear amount @xmath169 is homotopic to @xmath192 .
the lift of the above arc has two orthogonal sub arcs to the @xmath195-axis one from the left which is equal to @xmath196 which meets @xmath195 axis at a point @xmath204 between @xmath198 and @xmath199 , and the other orthogonal arc @xmath205 which meets the @xmath195-axis at a point @xmath206 . by the definition of left twists
, it follows that one endpoint of a lift @xmath207 of @xmath192 is between @xmath98 and @xmath26 , and the other endpoint of @xmath207 is between @xmath208 and @xmath209 . among all the geodesics
whose one endpoint is in the interval @xmath210 and the other endpoint is in the interval @xmath211 , the geodesic with endpoints @xmath98 and @xmath208 subtends the largest angle @xmath212 with the @xmath195-axis .
we have @xmath213 define @xmath214}+2\ ] ] where @xmath215 $ ] is the integer part of @xmath216 .
then we have that @xmath217 which implies that @xmath218 for all @xmath219 $ ] . note that @xmath220 by the mean value theorem there exists @xmath221 such that @xmath222 because @xmath223 .
since @xmath224 , the above gives @xmath225 .1 cm assume now that @xmath226 .
then we use a similar method by considering @xmath227 and @xmath203 full right twist around @xmath30 to replace @xmath83 with @xmath192 .
the proof proceeds analogously .
.2 cm * case 2 . *
the second case is when @xmath71 .
define a closed curve @xmath228 to consists of the unique arc @xmath229 in @xmath168 orthogonal at both of its endpoints to @xmath230 followed by the arc in @xmath30 ( in the direction of the left twist ) of the size at most @xmath231 followed by the unique arc @xmath232 orthogonal to @xmath30 at both of its endpoints followed by an arc on @xmath230 of size at most @xmath231 .
for the convenience of the notation , denote by @xmath74 the closed geodesic homotopic to @xmath74 .
the arcs @xmath112 , for @xmath62 , have lengths comparable to @xmath233 up to positive multiplicative constants . let @xmath79 , for @xmath234 , be two consecutive lifts of @xmath30 .
two lifts @xmath235 , for @xmath236 , of @xmath237 which meet @xmath79 can be chosen such that the distance between their foots on @xmath79 is at most @xmath231 .
assume that @xmath186 .
we perform @xmath203 full left twists along @xmath30 to obtain a new closed curve @xmath192 from the closed curve @xmath83 .
when @xmath238 + 2 $ ] , we get ( similar to case 1 ) for both angles @xmath239 that the new closed geodesic @xmath192 subtends with @xmath30 , @xmath240 then @xmath241 which gives @xmath242 when @xmath226 , the proof proceeds as before .
thus we established that the map @xmath28 is locally lipschitz .
.1 cm * step iv : @xmath243 is locally lipschitz .
* we consider the map @xmath243 and prove that it is also locally lipschitz .
let @xmath244 be fixed .
denote by @xmath245 the surface corresponding to @xmath246 , namely @xmath247 .
let @xmath248 such that @xmath249 and @xmath250 which implies @xmath251 .
there exists a @xmath252-quasiconformal map @xmath253 such that @xmath254 for all @xmath255 ( cf .
@xcite , @xcite ) .
recall that @xmath256 where @xmath153 , @xmath257 and @xmath258 , we have @xmath259 it follows that latexmath:[\[\label{eq : diff - in - twists } since @xmath248 , it follows that there exists @xmath261 such that @xmath262 the inequalities ( [ eq : diff - in - twists ] ) , ( [ eq : star ] ) and ( [ eq : twostars ] ) imply @xmath263 where @xmath264 , and since @xmath265 , we get @xmath266 since @xmath267 is a @xmath268-quasiconformal , it follows that @xmath269 .
moreover , if @xmath270 then ( [ eq : irr - twists - b - b ] ) implies that that @xmath271 . finally , @xmath272 .
it remains to estimate the length - spectrum distance between @xmath273 and @xmath128 .
this part of the argumentt is essentially contained in @xcite . note that @xmath128 is obtained from @xmath274 by multi twist along @xmath30 by the amount @xmath275 .
the estimate ( [ eq : irr - twists - b - b ] ) and the triangle inequality @xmath276 gives that @xmath277 for any simple closed geodesic @xmath10 on @xmath274 , we estimate @xmath278 .
we have @xmath279 and @xmath280 by the collar lemma .
since @xmath141 , it follows that there exists @xmath43 such that @xmath281 .
thus there exists @xmath282 such that @xmath283 the above inequalities imply that @xmath284 and by reversing roles played by @xmath128 and @xmath274 we get @xmath285 this proves that @xmath243 is lipschitz .
since @xmath3 is contractible , we get [ cor : contractible ] the length spectrum teichmller space @xmath1 for any hyperbolic surface @xmath0 with an upper bounded pants decomposition is contractible .
a question of characterizing the closure of the image of @xmath4 inside @xmath1 was raised in @xcite .
we use our understanding of the topology on the fenchel - nielsen coordinates that makes the map @xmath28 into a homeomorphism to give a characterization of the closure of @xmath4 .
let @xmath292 denotes a bounded function and let @xmath293 as @xmath294 .
moreover , @xmath295 denotes a function which converges to @xmath26 as @xmath294 and let @xmath296 .
then @xmath297 are the fenchel - nielsen coordinates of @xmath21 if and only if @xmath298 and @xmath299 by @xcite , the image @xmath300 of the quasiconformal teichmller space @xmath4 consists of all @xmath297 such that @xmath301 or equivalently @xmath302 [ thm : closure - qc ] let @xmath0 be a complete hyperbolic surface with an upper bounded pants decomposition @xmath303 .
then @xmath21 is in the closure of @xmath4 for the metric @xmath16 if and only if @xmath304 and @xmath305 .
we first note that if @xmath0 has a ( geodesic ) pants decomposition which is bounded from the above and from the below then ( cf .
@xcite , @xcite ) @xmath306 .
therefore we assume that there is a pants decomposition of @xmath0 which is upper bounded with a sequence of cuffs whose lengths go to @xmath26 .
let @xmath307 such that @xmath308 in the length spectrum metric @xmath16 as @xmath309 .
then @xmath310 , namely @xmath21 .
let @xmath311 be the set of all geodesics in @xmath58 such that @xmath312 .
then by theorem [ thm : main ] @xmath313 as @xmath309 .
thus for any @xmath314 there exists @xmath315 such that for all @xmath316 we have @xmath317 assume on the contrary that ( [ eq : o(length ) ] ) is false
. then there exists @xmath261 and subsequence @xmath318 such that @xmath319 as @xmath320 and @xmath321 choose @xmath322 .
the above two inequalities give for all @xmath316 @xmath323 which contradicts @xmath324 as @xmath309 .
thus @xmath2 satisfies ( [ eq : o(length ) ] ) .
assume that @xmath21 satisfies ( [ eq : o(length ) ] ) .
we need to find a sequence @xmath307 such that @xmath308 as @xmath309 for the length spectrum metric @xmath16 .
for a given @xmath325 , let @xmath326 be defined by the fenchel - nielsen coordinates @xmath327 and @xmath328 \min\ { |t_{\alpha_n}(x)- t_{\alpha_n}(x_0)|,i\}.\ ] ] by @xcite , we have @xmath307 .
let @xmath329 and choose @xmath314 .
since @xmath2 satisfies ( [ eq : o(length ) ] ) , it follows that there exists @xmath330 such that @xmath331 for all @xmath41 with @xmath332 .
moreover , there exists @xmath333 such that @xmath334 for all @xmath41 with @xmath335 .
d. b. a. epstein and a. marden , _ convex hulls in hyperbolic space , a theorem of sullivan , and measured pleated surfaces _ , analytical and geometric aspects of hyperbolic space ( coventry / durham , 1984 ) , 113 - 253 , london math .
lecture note ser .
, 111 , cambridge univ . press ,
cambridge , 1987 . | let @xmath0 be an infinite type hyperbolic surface ( whose boundary components , if any , are closed geodesics or punctures ) which has an upper bounded pants decomposition .
the length spectrum teichmller space @xmath1 consists of all surfaces @xmath2 homeomorphic to @xmath0 such that the ratios of the corresponding simple closed geodesics are uniformly bounded from below and from above .
alessandrini , liu , papadopoulos and su @xcite described the fenchel - nielsen coordinates for @xmath1 and using these coordinates they proved that @xmath1 is path connected .
we use the fenchel - nielsen coordinates for @xmath1 to induce a locally bilipschitz homeomorphism between @xmath3 and @xmath1 ( which extends analogous results by fletcher @xcite and by allessandrini , liu , papadopoulos , su and sun @xcite for the unreduced and the reduced @xmath4 ) .
consequently , @xmath1 is contractible .
we also characterize the closure in the length spectrum metric of the quasiconformal teichmller space @xmath4 in @xmath1 .
[ section ] [ definition]remark [ definition]example [ definition]proposition [ definition]theorem [ definition]corollary [ definition]lemma |
a longstanding problem in general relativity has been to find the general behavior of singularities .
several results , both analytical@xcite and numerical@xcite have been obtained .
though most of the results are for the case where the spacetimes have one or more symmetries , recent work has been done on the general case where there are no symmetries.@xcite there is a longstanding conjecture@xcite due to belinski , lifschitz and khalatnikov ( bkl ) that states that the generic singularity is spacelike and local .
this conjecture has been reformulated and put more precisely by uggla _
et al_.@xcite the type of local dynamics conjectured by bkl depends on the type of matter . for vacuum and for many types of matter , the bkl conjecture is that the local dynamics is oscillatory , corresponding to the dynamics of a bianchi type ix spacetime .
however , for stiff fluid ( _ i.e. _ fluid with pressure equal to energy density ) the bkl conjecture is that the local dynamics is asymptotically velocity term dominated corresponding to the dynamics of a bianchi type i spacetime . the vacuum version of the bkl conjecture has been supported by the numerical simulations of @xcite which show local and oscillatory dynamics in vacuum spacetimes with no symmetry
. the stiff fluid version of the bkl conjecture has been supported by the theorem of andersson and rendall@xcite which shows the local existence in a neighborhood of the singularity of solutions of the einstein equations with stiff fluid matter with the expected asymptotic behavior and with enough degrees of freedom to be the generic solutions . what is not known is whether generic stiff fluid initial data evolves to a solution of the andersson and rendall class . to address this issue
, we perform numerical simulations of the approach to the singularity for stiff fluid matter with no symmetries .
our methods are those of @xcite using the system of @xcite .
section ii presents the equations and numerical methods used .
the results are given in section iii and conclusions in section iv .
the system evolved here is essentially that of reference@xcite but specialized to the stiff fluid case and with a slightly different choice of gauge . here
the spacetime is described in terms of a coordinate system ( @xmath0 ) and a tetrad ( @xmath1 ) where both the spatial coordinate index @xmath2 and the spatial tetrad index @xmath3 go from 1 to 3 .
it is assumed that @xmath4 is hypersurface orthogonal and that the relation between tetrad and coordinates is of the form @xmath5 and @xmath6 here @xmath7 is the lapse and we have chosen the shift to be zero .
we choose the spatial frame @xmath8 to be fermi propagated along the integral curves of @xmath4 .
the commutators of the tetrad components are decomposed as follows : @xmath9 & = & { { \dot u}_\alpha}{{\bf e}_0 } -(h { { \delta _ \alpha}^\beta } + { { \sigma _ \alpha}^\beta } ) { { \bf e}_\beta } \label{com0i } \\ \left [ { { \bf e}_\alpha } , { { \bf e}_\beta } \right ] & = & ( 2 { a_{[\alpha}}{{\delta _ { \beta ] } } ^\gamma } + { \epsilon _ { \alpha \beta \delta } } { n^{\delta \gamma}}){{\bf e}_\gamma } \label{comij}\end{aligned}\ ] ] where @xmath10 is symmetric , and @xmath11 is symmetric and trace free .
square brackets denote the antisymmetric part of a tensor .
define @xmath12 and @xmath13 , that is @xmath14 is the timelike vector of the tetrad and @xmath15 is the spatial metric corresponding to the choice of @xmath14 as the time direction .
then the stress - energy tensor can be decomposed as @xmath16 where @xmath17 and @xmath18 are orthogonal to @xmath14 and where @xmath18 is symmetric and trace - free .
round brackets denote the symmetric part of a tensor .
scale invariant variables are defined as follows : @xmath19 the matter variables are not all independent , because we assume that the stress - energy is that of a stiff fluid @xmath20 here @xmath21 is the rest frame energy density of the fluid and @xmath22 is the fluid four - velocity , which can be decomposed as @xmath23 where @xmath24 is orthogonal to @xmath14 .
comparison of equations ( [ generalstress ] ) and ( [ stiffstress ] ) yields @xmath25 here @xmath26 and @xmath27 and angle brackets denote the symmetric trace - free part of a tensor .
thus , all scale invariant matter variables can be expressed in terms of @xmath28 and @xmath29 .
finally choose the lapse to be @xmath30 .
the relation between scale invariant frame derivatives and coordinate derivatives is @xmath31 and @xmath32 . from the einstein field equations and the conservation of stress - energy one obtains the following evolution equations : @xmath33 { q^\beta } \bigr ) \\ & - & 2 { v^\alpha } \bigl ( { \partial _ t } \omega - 2 [ q+1 ] \omega \bigr ) \biggr ] \label{ev7 } \\ \nonumber { \partial _
t } q & = & \left [ 2 ( q-2 ) + { \textstyle \frac 1 3 } \left ( 2 { a^\alpha } - { r^\alpha}\right ) { \parb _ \alpha } - { \textstyle \frac 1 3 } { \parb ^\alpha}{\parb _ \alpha } \right ] q \\ \nonumber & - & { \textstyle \frac 4 3 } { \parb _ \alpha}{r^\alpha } + { \textstyle \frac 8 3}{a^\alpha}{r_\alpha } + { \textstyle \frac 2 3 } { r_\beta}{\parb _ \alpha}{\sigma ^{\alpha \beta } } - 2 { \sigma ^{\alpha \beta}}{w_{\alpha \beta } } \\ \nonumber & + & { \frac 2 { g_+ } } \biggl [ 2 \omega { \sigma ^{\alpha \beta } } { v_\alpha } { v_\beta } - 2 ( q-2 ) \omega + { \partial _ t } \omega \\ & - & { \frac { 2 \omega } { g_+ } } { v_\alpha } { \partial _ t } { v^\alpha } \biggr ] \label{ev8}\end{aligned}\ ] ] here we are using units where @xmath34 .
furthermore the quantities @xmath35 , @xmath36 , @xmath37 and @xmath38 are given by @xmath39 in addition to the evolution equations , the variables satisfy constraint equations as follows : @xmath40{{e_\beta } ^i } \bigr ) \\ & - & { n^{\gamma \alpha } } { { e_\alpha } ^i } \label{cn1 } \\
\nonumber 0 & = & { { \cal c}_{\rm g } } \equiv 1 + { \textstyle \frac 1 3 } ( 2 { \parb _ \alpha } - 2 { r_\alpha } - 3 { a_\alpha}){a^\alpha } - { \textstyle \frac 1 6}{n_{\alpha \beta}}{n^{\alpha \beta } } \\ & + & { \textstyle \frac 1 { 12 } } { { ( { { n^\alpha}_\alpha})}^2 } - { \textstyle \frac 1 6 } { \sigma _ { \alpha \beta}}{\sigma ^{\alpha \beta } } - \omega \label{cn2 } \\
\nonumber 0 & = & { { ( { { \cal c}_{\rm c}})}^\alpha } \equiv { \parb _ \beta } { \sigma ^{\alpha \beta}}+ 2 { r^\alpha } - { { \sigma ^\alpha}_\beta } { r^\beta } - 3 { a_\beta}{\sigma ^{\alpha \beta } } \\ & -&{\epsilon ^{\alpha \beta \gamma}}{n_{\beta \delta } } { { \sigma _ \gamma}^\delta } + 3 { q^\alpha } \label{cn3 } \\
\nonumber 0 & = & { { \cal c}_q } \equiv q - { \textstyle \frac 1 3 } { \sigma ^{\alpha
\beta } } { \sigma _ { \alpha \beta}}+{\textstyle \frac 1 3 } { \parb _ \alpha } { r^\alpha } - { \textstyle \frac 2 3}{a_\alpha}{r^\alpha } \\ & - & { \textstyle { \frac 1 2 } } ( \omega + 3 p ) \label{cn4 } \\
\nonumber 0 & = & { { ( { { \cal c}_{\rm j}})}^\alpha } \equiv { \parb _ \beta } { n^{\alpha \beta } } - ( { r_\beta}+ 2 { a_\beta}){n^{\alpha \beta } } \\ & + & { \epsilon ^{\alpha \beta \gamma } } ( { \parb _ \beta}{a_\gamma } - { r_\beta}{a_\gamma } ) \label{cn5 } \\ 0 & = & { { ( { { \cal c}_{\rm w}})}^\alpha } = { \epsilon ^{\alpha \beta \gamma } } \bigl ( { \parb _ \beta}{r_\gamma } - { a_\beta}{r_\gamma } \bigr ) - { n^{\alpha \beta}}{r_\beta } \label{cn6 } \end{aligned}\ ] ] we want a class of initial data satisfying these constraints that is general enough for our purposes but simple enough to find numerically .
recall that on an initial data surface , the spatial metric @xmath41 and extrinsic curvature @xmath42 must satisfy the constraint equations @xmath43 here @xmath44 and @xmath45 are respectively the derivative operator and scalar curvature associated with @xmath41 and @xmath46 and @xmath47 are the components of the stress - energy tensor given in equation ( [ generalstress ] ) .
we use the york method@xcite which begins by defining the quantities @xmath48 and @xmath49 by @xmath50 we choose @xmath51 to be constant , @xmath48 to be the flat metric @xmath52 and @xmath53 to vanish . with these choices , equations ( [ cmom ] ) and ( [ cham ] )
become @xmath54 here @xmath55 is the ordinary derivative with respect to cartesian coordinates and indicies are raised and lowered with @xmath52 .
we choose space to have topology @xmath56 with the cartesian coordinates @xmath57 and @xmath58 each going from @xmath59 to @xmath60 .
we choose the following solution of equation ( [ ymom ] ) @xmath61 with the off - diagonal components of @xmath49 vanishing . here
the @xmath62 and @xmath63 are constants .
note that due to the periodicity of the coordinates and the linearity of equation ( [ ymom ] ) the general solution of equation ( [ ymom ] ) is a fourier series .
the solution that we choose is then essentially the simplest solution of equation ( [ ymom ] ) without symmetries .
the quantity @xmath46 can be freely specified and we choose it to be @xmath64 where the @xmath65 are constants . with these choices for @xmath49 and @xmath46 , equation ( [ yham ] ) is solved numerically ( in a manner to be described later ) to yield @xmath66 and therefore @xmath41 and @xmath42 . from this initial data
, we must then produce the initial values of the scale invariant variables . from equation ( [ com0i ] ) it follows that @xmath67 and since @xmath68 is constant it then follows that @xmath69 vanishes . since the initial spatial metric is conformally flat , we choose the initial spatial tetrad vectors by @xmath70 it then follows from equation ( [ comij ] ) that @xmath71 vanishes and that @xmath72 from equation ( [ com0i ] ) it then follows that @xmath73 while @xmath74 is given by equation ( [ si5 ] ) and @xmath75 by the vanishing of equation ( [ cn4 ] ) .
the numerical method used is as follows : each spatial direction corresponds to @xmath76 grid points with spacing @xmath77 .
the variables on grid points @xmath78 to @xmath79 are evolved using the evolution equations , while at points @xmath80 and @xmath76 periodic boundary conditions are imposed .
the initial data is determined once equation ( [ yham ] ) is solved .
this is done iteratively as follows:@xcite define @xmath81 then equation ( [ yham ] ) takes the form @xmath82 .
we make an initial guess @xmath83 for @xmath84 and solve using the conjugate gradient method @xcite the equation @xmath85 iterating until @xmath86 satisfies equation ( [ yham ] ) to within a set tolerance .
the evolution proceeds using equations ( [ ev1]-[ev8 ] ) with the exception that the term @xmath87 is added to the right hand side of equation ( [ ev8 ] ) to prevent the growth of constraint violating modes and the term @xmath88 is added to the right hand side of equation ( [ ev3 ] ) to make the system well posed.@xcite spatial derivatives are evaluated using centered differences , and the evolution is done using a three step iterated crank - nicholson method@xcite ( a type of predictor - corrector method ) . in equation ( [ ev8 ] )
the highest spatial derivative term is @xmath89 which gives this equation the form of a diffusion equation .
note that diffusion equations can only be evolved in one direction in time , in this case the negative direction which corresponds to the approach to the singularity .
stability of numerical evolution of diffusion equations generally requires a time step proportional to the square of the spatial step .
however , the constant of proportionality depends on the coefficient of the second spatial derivative . to ensure stability , we define @xmath90 to be the maximum value of @xmath91 ( over all space and over all @xmath3 and @xmath2 ) and then define @xmath92 and @xmath93 .
the time step @xmath94 is then chosen to be whichever of @xmath95 and @xmath96 has the smaller magnitude . before presenting numerical results
, it is helpful to consider what behavior to expect as the singularity is approached ( that is as @xmath97 ) . first denote the eigenvalues of @xmath98 by ( @xmath99 ) .
then suppose that at sufficiently early times the time averages of @xmath100 are all positive .
then the time averages of the eigenvalues of @xmath101 are all positive .
since we are evolving in the negative time direction , this should lead ( through equation ( [ ev1 ] ) ) to an exponential decrease in @xmath102 .
however , since all spatial derivatives appear in the equations in the form @xmath103 we would expect the spatial derivatives to become negligible .
that is , the approach to the singularity is local . furthermore , this positivity of the time averages of the eigenvalues of @xmath101 should lead ( through equations ( [ ev2]-[ev3 ] ) ) to exponential decrease in @xmath69 and @xmath104 .
a similar argument applied to equations ( [ ev6 ] ) and ( [ dtqvec ] ) and using equations ( [ cn2 ] ) and ( [ cn4 ] ) indicates that as the singularity is approached @xmath105 should become negligible , but @xmath74 should not , and therefore that @xmath106 should become negligible .
thus , as the singularity is approached , the system should be well described by a simplified set of evolution and constraint equations where spatial derivatives as well as @xmath107 and @xmath106 are negligible .
note that the fact that spatial derivatives are becoming negligible does @xmath108 mean that the spacetime is becoming homogeneous .
rather the considerable spatial variation is becoming a negligible part of the equations of motion since all spatial derivatives appear multiplied by @xmath102 which is becoming negligible .
we now write down this simplified system of evolution and constraint equations where all these terms are neglected . in this limit equation ( [ cn3 ] ) implies that the matricies @xmath98 and @xmath109 commute and therefore have the same eigenvalues .
the evolution equations for @xmath110 and @xmath111 can then be written as evolution equations for their eigenvalues . the non - trivial evolution and
constraint equations then become in this limit @xmath112 @xmath113 here @xmath114 and @xmath115 are the eigenvalues of @xmath98 and @xmath109 respectively and @xmath116 is given by @xmath117 and indicies are not summed over unless explicitly indicated .
suppose that the dynamics is in a period ( called a kasner epoch ) when all the @xmath115 are negligibly small .
then it follows from equations ( [ trcn1 ] ) and ( [ trcn2 ] ) that @xmath118 and therefore , from equations ( [ trev1 ] ) and ( [ trev3 ] ) that @xmath74 and the @xmath114 are constant . from equations ( [ trcn1 ] ) and ( [ trev2 ] ) it follows that there are two possibilities for a kasner epoch : ( i ) all the @xmath114 are @xmath119 in which case the @xmath115 remain negligible and the kasner epoch lasts all the way to the singularity .
( ii ) one of the @xmath114 is @xmath120 in which case the corresponding @xmath115 grows until it is large enough to bring the kasner epoch to an end .
we now look in more detail at possibility ( ii ) .
let @xmath121 be the @xmath114 that is @xmath122 .
then @xmath123 is the @xmath115 that is growing .
we are therefore led to examine equations ( [ trev1]-[trcn2 ] ) neglecting @xmath124 and @xmath125 but not @xmath123 .
in this regime equations ( [ trev1 ] ) and ( [ trev2 ] ) become @xmath126 where @xmath127 .
define @xmath128 .
then equation ( [ sig1ke ] ) becomes @xmath129 while from equations ( [ sig2ke ] ) and ( [ sig3ke ] ) it follows that there are constants @xmath130 and @xmath131 with @xmath132 such that @xmath133 and @xmath134 .
similarly , it follows from equation ( [ omegake ] ) that there is a constant @xmath135 such that @xmath136 .
finally , it then follows from equation ( [ trcn1 ] ) that @xmath137 satisfies the equation of motion @xmath138 z\ ] ] where @xmath139 $ ] .
note that the quantity in square brackets vanishes at @xmath140 and @xmath141 where @xmath142 .
therefore the dynamics is a `` bounce '' from a kasner epoch corresponding to @xmath141 to one corresponding to @xmath140 . use a minus subscript to denote a quantity before the bounce and a plus subscript to denote a quantity after the bounce .
we then have @xmath143 .
however , from the definition of @xmath137 it follows that @xmath144 while from the definition of @xmath145 it follows that @xmath146 .
we then find the following `` bounce rule '' relating a quantity after the bounce to quantities before the bounce .
@xmath147 note that from the bounce rule it follows that @xmath74 increases at each bounce .
furthermore , it follows from equation ( [ trcn1 ] ) ( and from the fact that @xmath148 is trace - free ) that the minimum possible value for a @xmath114 during a kasner epoch is @xmath149 and therefore that no further bounces can happen once @xmath150 ( though bounces may cease at lower values of @xmath74 ) .
thus we expect that at each spatial point there is a last bounce followed by a kasner epoch that lasts all the way to the singularity .
in other words , we expect the approach to the singularity to be of the andersson and rendall class .
all runs were done in double precision on a sunblade 2000 with @xmath151 ( except for a convergence test which also used @xmath152 ) .
the equations were evolved from @xmath153 to @xmath154 . for the initial data ,
the trace of the extrinsic curvature was @xmath155 corresponding to an initial value of @xmath156 for @xmath68 .
the constants @xmath157 and @xmath65 characterizing the initial data were @xmath158 and the constant @xmath159 was @xmath160 .
we would like to know whether @xmath161 and @xmath29 become negligible as the singularity is approached . in figure 1
are plotted the maximum values ( over all space , @xmath162 and @xmath2 ) of @xmath163 and @xmath164 .
note that after a certain amount of evolution , all these quantities steadily decrease .
this indicates that after a certain amount of time spatial derivatives become negligible in the eqautions of motion and that the approximation considered at the end of the previous section becomes valid .
maximum values of @xmath165 ( solid line ) , @xmath166 ( dotted line ) @xmath167 ( dot - dashed line ) and @xmath164 ( dashed line ) _ vs _
@xmath168 @xmath169 _ vs _
@xmath170 for the @xmath151 run ( solid line ) and @xmath171 _ vs _
@xmath170 for the @xmath152 run ( dotted line ) ] components of @xmath148 in the asymptotic frame _ vs _
@xmath121 ( solid line ) , @xmath172 ( dotted line ) and @xmath173 ( dot - dashed line ) ] components of @xmath71 in the asymptotic frame _ vs _
@xmath123 ( solid line ) , @xmath124 ( dotted line ) and @xmath125 ( dot - dashed line ) ] @xmath28 _ vs _ @xmath168 it then follows that the interesting part of the dynamics is the development of the variables at a single point as a function of time .
we now present data of that form .
the behavior at the spatial point chosen is typical .
the results of a convergence test are plotted in figure 2 . here
what is plotted is @xmath169 for the @xmath151 run ( solid line ) and @xmath174 for the @xmath152 run ( dotted line ) .
both quantities are plotted _ vs _ @xmath170 .
note that the two curves roughly agree in magnitude but become out of sync in time .
this indicates second order convergence but with the system having sensitive dependence on initial conditions .
similar results were obtained for the other constraints .
figures 3 and 4 show respectively the diagonal components of @xmath148 and @xmath71 in the asymptotic frame , _
i.e. _ the frame of the eigenvectors that @xmath175 has at the end of the evolution . for that part of the evolution where the approximation made at the end of the previous section is valid , these diagonal components are the eigenvalues of @xmath176 and @xmath109 respectively .
note that the dynamics of the eigenvalues of @xmath176 consists of epochs where they are apporoximately constant ( kasner epochs ) punctuated by short bounces .
furthermore the components of @xmath71 are negligible except at the bounces . also note that there is a last bounce and that this coincides with the most negative eigenvalue of @xmath98 becoming greater than @xmath155 . in figure 5
is plotted @xmath74 _ vs _ @xmath170 .
note that the behavior of @xmath74 is a sequence of constant values that are punctuated by short bounces and that the bounces in @xmath74 occur at the same times as the bounces in @xmath148 .
the sequence of the values of @xmath74 is 0.05956 , 0.1607 , 0.4139 , 0.6231 , while the corresponding values of the most negative eigenvalues of @xmath177 are -1.583 , -1.565 , -1.277 , -0.6596 .
these values obey the `` bounce rule '' of equation ( [ rlbounce ] ) .
these simulations support the expected picture for the approach to the generic singularity in a spacetime where the matter is a stiff fluid . as the singularity
is approached the terms in the equations of motion involving spatial derivatives become negligible .
the dynamics at each spatial point consists of a sequence of kasner epochs punctuated by short bounces .
the sequence of values of @xmath74 obeys the expected bounce rule .
there is a last bounce , after which the dynamics is described by a single kasner epoch all the way to the singularity , thus yielding a spacetime in the class of reference @xcite .
what remains to be done is to treat the approach to the singularity for non - stiff fluids . here
the behavior that is expected is quite different .
the bkl conjecture is that the matter will become negligible and the dynamics of the gravitational variables as the singularity is approached will be that of vacuum spacetimes .
the formalism of @xcite includes a class of non - stiff fluids , and the resulting equations are similar to those of the stiff fluid case .
nonetheless , the numerical methods of this paper are not adequate to treat the case of non - stiff fluids .
that is because in non - stiff fluids shock waves form , while the numerical methods of the present paper are appropriate for smooth solutions .
a shock capturing method would be appropriate to treat the non - stiff fluid case .
this work was partially supported by a grant from the national science and engineering research council of canada and by nsf grant phy-0456655 to oakland university . | numerical simulations of the approach to the singularity in spacetimes with stiff fluid matter are presented here .
the spacetimes examined have no symmetries and can be regarded as representing the general behavior of singularities in the presence of such matter .
it is found that the singularity is spacelike and that as it is approached , the spacetime dynamics becomes local and non - oscillatory . |
max - stable processes are a canonical class of statistical models for multivariate extremes .
they appear in a variety of applications ranging from insurance and finance @xcite to spatial extremes such as precipitation @xcite and extreme temperature .
max - stable processes are exactly the class of non - degenerate stochastic processes that arise from limits of independent component - wise maxima .
this fact provides a theoretical justification for their use as models of multivariate extremes .
however , many useful max - stable models suffer from intractable likelihoods , thus prohibiting standard maximum likelihood and bayesian inference .
this has motivated development of maximum _ composite _ likelihood estimators ( mcle ) for max - stable models @xcite as well as certain approximate bayesian approaches @xcite .
in contrast to their likelihoods , the cumulative distribution functions ( cdfs ) for many max - stable models are available in closed form , or they are tractable enough to approximate within arbitrary precision .
this motivates statistical inference based on the minimum distance method @xcite . in this paper
, we propose an m - estimator for parametric max - stable models based on minimizing distances of the type @xmath0 where @xmath1 is a @xmath2-dimensional cdf of a parametric model , @xmath3 is a corresponding empirical cdf and @xmath4 is a tuning measure that emphasizes various regions of the sample space @xmath5 using elementary manipulations it can be shown that minimizing distances of the type is equivalent to minimizing the _ continuous ranked probability score _ ( crps ) .
[ def:(crps - m - estimator)](crps m - estimator ) let @xmath4 be a measure that can be tuned to emphasize regions of a sample space @xmath6 .
define the crps functional @xmath7 then for independent random vectors @xmath8 with common distribution function @xmath9 we define the following crps m - estimator for @xmath10 .
@xmath11 for simplicity , we shall assume throughout that the parameter space @xmath12 is a compact subset of @xmath13 , for some integer @xmath14 .
the remainder of this paper is organized as follows . in section [ sec : extreme - values - and - nax - stability ] we review some essential multivariate extreme value theory and provide definitions and constructions of max - stable models . in section [ sec : crps - m - estimation ] we establish regularity conditions for consistency and asymptotic normality of the crps m - estimator and provide general formulae for calculating its asymptotic covariance matrix . in section [ sec : crps - for - max - stable ] we specialize these calculations to the max - stable setting . in section
[ sec : simulation ] we conduct a simulation study to evaluate the proposed estimator for popular max - stable models .
let @xmath15 be independent and identically distributed measurements of certain environmental or physical phenomena .
for example , the @xmath16s may model wave - height , temperature , precipitation , or pollutant concentration levels at a site @xmath17 in a spatial region @xmath18 .
if one is interested in extremes , it is natural to consider the asymptotic behavior of the point - wise maxima .
suppose that , for some @xmath19 and @xmath20 , we have @xmath21 for some non - trivial limit process @xmath22 , where @xmath23 denotes convergence of the finite - dimensional distributions .
the class of _ extreme value _ processes @xmath24 arising in the limit describe the statistical dependence of ` worst case scenaria ' and are therefore natural models of multivariate extremes .
the limit @xmath22 in is necessarily a _
max - stable process _ in the sense that for all @xmath25 , there exist @xmath26 and @xmath27 , such that @xmath28 where @xmath29 are independent copies of @xmath22 and where @xmath30 means equality of finite - dimensional distributions ( ch.5 of * ? ? ?
due to the classic results of fisher - tippett and gnedenko , the marginals of @xmath22 are necessarily extreme value distributions ( frchet , reversed weibul or gumbel ) .
they can be described in a unified way through the generalized extreme value distribution ( gev ) : @xmath31 where @xmath32 , and where @xmath33 and @xmath34 are known as the location , scale and shape parameters .
the cases @xmath35 and @xmath36 correspond to frchet , reverse weibull , and gumbel , respectively ( see , e.g. ch.3 and 6.3 in @xcite for more details ) .
the dependence structure of the limit extreme value process @xmath22 rather than its marginals is of utmost interest in practice .
arguably , the type of the marginals is unrelated to the dependence structure of @xmath22 and as it is customarily done , we shall assume that the limit process @xmath22 has been transformed to standard @xmath37-frchet marginals .
that is , @xmath38 for some scale @xmath39 ( ch.5 of * ? ? ?
let @xmath40 be a max - stable process with @xmath37-frchet marginals as in .
then , its finite - dimensional distributions are multivariate max - stable random vectors and they have the following representation : @xmath41 where @xmath42 and where @xmath43 is a finite measure on the positive unit sphere @xmath44 known as the _ spectral measure _ of the max - stable random vector @xmath45 ( see e.g. proposition 5.11 in * ? ? ?
the integral in the expression is referred to as the _ tail dependence function _ of the max - stable law .
we shall often use the notation : @xmath46 where @xmath47 , for the tail dependence function of the max - stable random vector @xmath45 .
it readily follows from that for all @xmath48 , the max - linear combination @xmath49 is @xmath37-frchet random variable with scale @xmath50 .
conversely , a random vector @xmath45 with the property that all its non - negative max - linear combinations are @xmath37-frchet is necessarily multivariate max - stable @xcite .
this invariance to max - linear combinations is an important feature that will be used in our estimation methodology ( section [ sec : crps - for - max - stable ] , below ) .
some max - stable models are readily expressed in terms of their spectral measures while others via tail dependence functions .
these representations however are not convenient for computer simulation or in the case of random processes , where one needs a handle on all finite - dimensional distributions .
the most common constructive representation of max - stable process models is based on poisson point processes @xcite .
see also @xcite for an alternative .
indeed , consider a measure space @xmath51 and let @xmath52 be a poisson point process on @xmath53 with intensity measure @xmath54 .
[ p : de - haan - simple ] let @xmath55 be a collection of non - negative integrable functions and let @xmath56 then , the process @xmath40 is max - stable with @xmath37-frchet marginals and finite - dimensional distributions : @xmath57 the proof of this result is sketched in appendix [ sub : derivation - of - max - stablecdf ] .
relation is known as the _ de haan spectral representation _ of @xmath22 and @xmath58 as the spectral functions of the process .
it can be shown that every separable in probability max - stable process has such a representation ( see @xcite and proposition 3.2 in @xcite ) .
the max - functional in has the properties of an _ extremal stochastic integral_. indeed , we have max - linearity : @xmath59 for all @xmath60 the above max - linear combination is therefore @xmath37-frchet and has a scale coefficient : @xmath61 one can also show that @xmath62 and @xmath63 are independent , if and only if @xmath64 , for @xmath65-almost all @xmath66 .
that is , the extremal integrals defining @xmath62 and @xmath63 are over non - overlapping sets .
this shows that for max - stable process models pairwise independence implies independence .
further , @xmath67 converges in probability to @xmath62 if and only if @xmath68 converges in @xmath69 to @xmath70 , as @xmath71 . for more details , see e.g @xcite and @xcite .
the expressions and may be related through a change of variables ( proposition 5.11 @xcite ) . while the spectral measure @xmath72 in is unique , a max - stable process has many different spectral function representations .
nevertheless , relation provides a constructive and intuitive representation of @xmath22 , that can be used to build interpretable models .
a great variety of max - stable models can be defined by specifying the measure space @xmath51 and an accompanying family of spectral functions @xmath70 or equivalently through a consistent family of spectral measures or tail dependence functions .
we review next several popular max - stable models and their basic features .
@xmath73 _ ( multivariate logistic ) _ let @xmath74 have the cdf @xmath75 for @xmath76 and @xmath77.$ ] the parameter @xmath78 controls the degree of dependence , where @xmath79 corresponds to independence ( @xmath80 ) , while @xmath81 to complete dependence ( @xmath82 , interpreted as a limit ) .
this model is rather simple since the dependence is exchangeable but it provides a useful benchmark for the performance of the crps - based estimators since the mle is easy to obtain in this case ( see table [ tab : logistic - model - simulation ] below ) .
the recent works of @xcite and @xcite develop far - reaching generalizations of multivariate logistic laws by exploiting connections to sum - stable distributions .
@xmath73 _ ( max - linear or spectrally discrete models ) _ let @xmath83 be a matrix with non - negative entries and let @xmath84 be independent standard @xmath37-frchet random variables .
define @xmath85 the vector @xmath86 is max - stable . it can be shown that the cdf of @xmath22 has the form were the spectral measure @xmath87 is concentrated on the normalized column - vectors of the matrix @xmath88 , i.e. on @xmath89 , where @xmath90 , and where @xmath91 stands for the dirac measure with unit mass at the point @xmath92 .
conversely , any max - stable random vector with discrete spectral measure @xmath72 has a max - linear representation as in , where the columns of the matrix @xmath88 may be recovered from .
we shall also call such models spectrally discrete .
since any spectral measure @xmath72 can be approximated arbitrarily well with a discrete one , max - linear models are dense in the class of all max - stable models . as argued in @xcite ,
max - linear distributions arise naturally in economics and finance , as models of extreme losses .
the @xmath93s represent independent shock - factors that lead to various extreme losses in a portfolio @xmath22 depending on the factor loadings @xmath94 .
max - linear models are particularly well - suited for crps - based inference , since their tail dependence function has a simple closed form : @xmath95 see section [ sec : simulation ] below for a simple example of crps - based inference for max - linear models and @xcite for an alternative m - estimation methodology .
@xmath73 _ ( moving maxima and mixed moving maxima ) _ let @xmath96 and @xmath97 , for some non - negative integrable function @xmath98 , @xmath99 .
then yields the so - called _ moving maxima _ random field : @xmath100 the choice of the kernel @xmath101 as a multivariate normal density in @xmath102 yields the well - known _
smith storm model _
, where the @xmath103s may be interpreted as storm locations , @xmath101 is the spatial storm attenuation profile and @xmath104 its strength .
more flexible models can be obtained by taking maxima of independent moving maxima , resulting in the so - called _
mixed moving maxima _ : @xmath105 where @xmath106 is a poisson point process on @xmath107 with intensity @xmath108 , and where @xmath98 is such that @xmath109 . here
@xmath110 is the ` mixing ' measure , which may be continuous or discrete , and the @xmath111s may be viewed as different types of storms .
the mixed moving maxima random fields are stationary , ergodic and , in fact , mixing @xcite . by , their tail dependence functions are @xmath112 @xmath73 _ ( spectrally gaussian models ) _ by viewing @xmath113 as a probability space , in the case @xmath114 , the spectral functions @xmath115 in become a stochastic process . by picking @xmath116 to be non - negative transformations of a gaussian process @xmath117 on @xmath113 , one obtains interesting and tractable max - stable models whose dependence structure is governed by the covariance structure of the underlying gaussian process @xmath118 .
the popular smith , schlather , and brown - resnick random field models are of this type @xcite .
@xmath119 _ ( schalther models ) _ let @xmath120 be a stationary gaussian random field with zero mean and let @xmath121 .
then @xmath62 in has the following tail dependence function @xmath122 where @xmath123 denotes integration with respect to the ` probability ' measure @xmath65 .
@xmath119 _ ( brown - resnick ) _ let @xmath124 be a zero mean gaussian random field with stationary increments .
set @xmath125 , where @xmath126 is the ` variance ' of @xmath117 .
the seminal paper of @xcite introduced this model with @xmath127 the standard brownian motion and showed that , surprisingly , the resulting max - stable process @xmath62 in is stationary , even though @xmath127 is not .
the cornerstone work of @xcite showed that @xmath128 is stationary for a centered gaussian process @xmath127 , with stationary increments .
it also obtained important mixed moving maxima representations of @xmath22 under further conditions on @xmath127 .
the tail dependence function of @xmath22 in this case is @xmath129 it can be shown that the smith model @xcite is a special case of a brown - resnick model with a degenerate random field @xmath130 , @xmath131 , where @xmath132 is a normal random vector in @xmath133 .
the above models can be deemed _ spectrally gaussian _ since their tail dependence functions ( and hence spectral measures ) are expectations of functions of gaussian laws .
one can consider other stochastic process models for the underlying spectral functions @xmath70 and thus arrive at _ doubly stochastic _ max - stable processes .
we comment briefly on some general probabilistic properties of these models .
if @xmath134 is a stationary process in @xmath113 , then the max - stable process @xmath135 is also stationary .
it is , however , non - ergodic . in particular ,
the schlather models are non - ergodic .
this is important in applications , since a single observation of the random field @xmath22 at an expanding grid , may not yield consistent parameter estimates .
@xcite have shown that brown - resnick random fields with non - stationary @xmath136 such that @xmath137 almost surely , have mixed moving maxima representations as in .
they are therefore mixing @xcite and consistent statistical inference from a single realization of such max - stable random fields is possible .
the poisson point process construction in involves a maximum over an infinite number of terms . as a result ,
computer simulations of spectrally gaussian max - stable models necessitates truncation to a finite number . in the case of the brown - resnick model ,
the number of terms required to produce a satisfactory representation is prohibitively large .
accurate simulation of brown - resnick processes is an active area of study @xcite .
consequently , simulation studies for inference under brown - resnick models have yet to appear .
for this reason the remaining discussion of spectrally gaussian max - stable models including simulation and application is restricted to the schlather model .
@xmath138 @xmath73 _ ( co - variation ) _ for @xmath62 as in , define @xmath139 : = \int_{s } g_t \wedge g_s d\nu \equiv \int_s g_t d\nu + \int_s g_s d\nu - \int_s g_t\vee g_s d\nu,\ ( t , s\in t).\ ] ] note that @xmath140 and @xmath141 are the scale coefficients of the frchet random variables @xmath62 and @xmath142 .
the co - variation @xmath143\ge 0 $ ] is non - negative and equals zero if and only if @xmath62 and @xmath63 are independent , analagous to covariance for gaussian processes
. @xmath73 _ ( extremal coefficient ) _ a popular summary measure of multivariate dependence in max - stable models is the extremal coefficient .
define @xmath144 for a process @xmath145 with standard 1-frchet marginals @xmath146 and thus @xmath147 , where @xmath148 corresponds to complete dependence while @xmath149 implies that @xmath150 s , @xmath151 are independent .
it is well know that for a process with standard @xmath37-frchet marginals , @xmath152 $ ] . in the case of the schlather model
there is an explicit formula for the bivariate extremal coefficient in terms of the correlation function : @xmath153 .
figure [ fig : max - stable - realizations ] displays realizations from the schlather model for the different correlation functions given in table [ tab : spectrally - gaussian - max - stable - models ] .
note that these examples are all ( spectrally ) isotropic in the sense that the correlation @xmath154 of the underlying gaussian process depends only on the distance @xmath155 between locations @xmath17 and @xmath156 .
this however is not a requirement in general .
figure [ fig : max - stable - realizations ] also provides some visual evidence of how the covariance structure and smoothness of @xmath127 influence the dependence structure of the resultant max - stable random field @xmath22 .
it is possible to parameterize the dependence structure of the max - stable random field using a large variety of covariance functions available for parameterizing gaussian processes .
.[tab : correlation - functions]correlation functions for gaussian random fields . for the matrn covariance function
, @xmath157 is the modified bessel function of the second kind . [
tab : spectrally - gaussian - max - stable - models ] [ cols="^,^",options="header " , ] note that for this model , the two estimators are comparable in terms of bias but the mcle is more efficient than crps .
this may indicate that mcle is especially well suited for estimation with spectrally gaussian max - stable models .
we have developed a general inferential framework for max - stable models based on the continuous ranked probability score ( crps ) .
it is shown that under mild regularity , crps m - estimators are consistent and asymptotically normal .
simulation studies across common spectrally continuous and discrete max - stable models yield essentially unbiased estimators with close to nominal coverage .
our estimators were about half as efficient versus the mle in the case of the simple multivariate logistic model , where a tractable likelihood exists .
overall the method displays flexibility and broad applicability in the max - stable setting . in the case of the schlather max - stable model , crps estimates were less efficient than mcle .
it is possible that efficiency for crps estimates can be improved through better tuning of the measure @xmath4 in the crps .
for instance consider @xmath158 it can be shown that crps m - estimation remains consistent for all @xmath159 with little complication over the case @xmath160 ( equivalent to the specification ) , which was chosen for analytical simplicity .
this begs the question of specifying @xmath161 to maximize the expected hessian of the crps , which should result in more efficient estimators .
this is beyond the scope of the present paper and it will be studied in a future work .
by , for all @xmath162 , @xmath163 where @xmath164 observe that @xmath165 since @xmath106 is a poisson point process on @xmath166 with intensity @xmath167 , @xmath168 which equals and completes the proof .
the above argument shows that the integrability of the functions @xmath70 implies the @xmath62s in are non - trivial random variables .
[ [ subasymptotics - proof - appendix - proofs - for - section - seccrps - m - estimation ] ] [ sub : asymptotics - proof - appendix ] proofs for section [ sec : crps - m - estimation ] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ fix @xmath178 markov s inequality and another application of fubini gives @xmath179 next , using the identity @xmath180 we have that the rhs of equals @xmath181\mu\left(dx\right)\right\ } \label{eq : consistency_proof_1}\end{gathered}\ ] ] note that @xmath182\le\overline{f}_{\theta_{0}}\left(x\right),$ ] and by condition _
( ii ) _ , @xmath183 thus , by the lebesgue dominated convergence theorem @xmath184\mu\left(dx\right)=\int_{\mathbb{r}^{d}}\lim_{n\to\infty}\mathbb{e}\left[\overline{f}_{n}\left(x\right)\wedge\overline{f}_{\theta_{0}}\left(x\right)\right]\mu\left(dx\right).\ ] ] the strong law of large numbers implies that @xmath185 converges almost surely to @xmath186 .
hence , by applying dominated convergence again , we obtain @xmath187=\overline{f}_{\theta_{0}}\left(x\right),\ \ \mbox{for all } x\in\mathbb{r}^{d}.\ ] ] this , by implies that the right - hand side of vanishes as @xmath71 , which in view of yields the desired convergence in probability and the proof is complete . by a standard argument using the lebesgue dct , condition _ ( iii )
_ of this proposition ensures that integration and differentiation can be interchanged in all that follows .
we proceed by establishing _
( i)-(iii ) _ of theorem [ thm:(asynormcrps ) ] . _
( ii ) _ observe that @xmath193 equals @xmath194\left(f_{\theta_{1}}\left(y\right)-f_{\theta_{2}}\left(y\right)\right)\right\ } \mu\left(dy\right)\right|\\ \le2\int_{\mathbb{r}^{d}}\left|f_{\theta_{1}}\left(y\right)-f_{\theta_{2}}\left(y\right)\right|\mu\left(dy\right)\end{gathered}\ ] ] where the last relation follows from the triangle inequality and fact that @xmath195 then , by the mean value theorem and the cauchy - schwartz inequality @xmath196 where @xmath197 by assumption _
( ii ) _ of this proposition , @xmath198 is finite . hence _
( ii ) _ of theorem [ thm:(asynormcrps ) ] holds where @xmath199 is constant ( and therefore trivially @xmath200 ) .
finally , we derive @xmath208 by considering its @xmath209th entry .
let @xmath210 denote @xmath211 .
@xmath212\\ & = & \mathbb{e}\left\ { \int_{\mathbb{r}^{d}}2\left(f_{\theta}\left(y_{1}\right)-\mathbf{1}_{\left\ { x\le y_{1}\right\ } } \right)\partial_{i}f_{\theta}\left(y_{1}\right)\mu\left(dy_{1}\right)\right.\\ & & \qquad\times\left.\left.\int_{\mathbb{r}^{d}}2\left(f_{\theta}\left(y_{2}\right)-\mathbf{1}_{\left\ { x\le y_{2}\right\ } } \right)\partial_{j}f_{\theta}\left(y_{2}\right)\mu\left(dy_{2}\right)\right|_{\theta=\theta_{0}}\right\ } \\ & = & 4\mathbb{e}\left\ { \left.\int_{\mathbb{r}^{d}}\int_{\mathbb{r}^{d}}b_{\theta}\left(x , y_{1},y_{2}\right)\partial_{i}f_{\theta}\left(y_{1}\right)\partial_{j}f_{\theta}\left(y_{2}\right)\mu\left(dy_{1}\right)\mu\left(dy_{2}\right)\right|_{\theta=\theta_{0}}\right\ } \end{aligned}\ ] ] where @xmath213 expanding the integrand and applying fubini gives @xmath214 where @xmath215 which is exactly the @xmath209th element of , as desired . [ [ proofs - for - section - seccrps - for - max - stablesubproofs - for - section-4 ] ] proofs for section [ sec : crps - for - max - stable][sub : proofs - for - section-4 ] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ observe that @xmath216 for @xmath217 and hence the integrand in vanishes , as @xmath218 .
also , by using a taylor series expansion of the exponential function at zero , it is easy to see that @xmath219 as @xmath220 .
therefore , the integral in is finite .
now , by making the changes of variables @xmath226 and @xmath227 in the last two integrals respectively , we obtain @xmath228 this , in view of , yields the expression in terms of the incomplete gamma function in . for note that @xmath235 is equal in distribution to a @xmath236-frchet random variable with scale @xmath229 which has finite expectation @xmath237 for , applying fubini s theorem , and observing that @xmath238 , with @xmath239 @xmath240 this establishes . for , substituting the espression @xmath241 from lemma [ lem : frechetintegral ] we have @xmath242\ ] ] which , after substituting and yeilds @xmath243.\label{eq : efrechetint1}\ ] ] using the fact that @xmath22 is distributed @xmath37-frchet with scale @xmath229 we have @xmath244 now the substitution @xmath245 gives @xmath246 plugging into yields @xmath247\\ & = 2\sqrt{\pi}\left[\frac{2v_{0}}{\sqrt{v_{0}+v}}-\sqrt{v_{0}}+\frac{2v}{\sqrt{v_{0}+v}}-\sqrt{2v}\right]\\ & = \eqref{eq : efrechetintegral}.\end{aligned}\ ] ] now recall that @xmath253 substituting , we obtain @xmath254 where , in view of lemma [ lem : frechetintegral ] , on can show that @xmath255 our next goal is to calculate @xmath256 where the expectation is taken under @xmath10 . using
the fact that @xmath257 is 1-frchet with scale @xmath258 lemma [ lem : efrechetintegral]_(ii ) _ implies @xmath259 thus , in view of , becomes @xmath260 this , in view of implies and completes the proof . | max - stable random fields provide canonical models for the dependence of multivariate extremes .
inference with such models has been challenging due to the lack of tractable likelihoods .
in contrast , the finite dimensional cumulative distribution functions ( cdfs ) are often readily available and natural to work with .
motivated by this fact , in this work we develop an m - estimation framework for max - stable models based on the _ continuous ranked probability score _ ( crps ) of multivariate cdfs .
we start by establishing conditions for the consistency and asymptotic normality of the crps - based estimators in a general context .
we then implement them in the max - stable setting and provide readily computable expressions for their asymptotic covariance matrices . the resulting point and asymptotic confidence interval estimates
are illustrated over popular simulated models .
they enjoy accurate coverages and offer an alternative to composite likelihood based methods .
example.eps gsave newpath 20 20 moveto 20 220 lineto 220 220 lineto 220 20 lineto closepath 2 setlinewidth gsave .4 setgray fill grestore stroke grestore |
consider two identical rubber balloons , the same one as used at children s parties , filled with air and connected by a hose with a valve . for simplicity , we shall approximate the balloons by spheres .
suppose that the larger balloon , the one that has been inflated more , has radius @xmath0 and contains @xmath1 moles of air , while the smaller one has radius @xmath2 and contains @xmath3 moles of air .
when the valve is opened which balloon is going to grow and which will shrink ?
it seems almost obvious that the larger balloon should shrink , while the smaller one should grow .
this , however , is not what is often found .
for some balloon sizes the air flows from the larger balloon to the smaller one , but for other sizes the direction of the air flow is reversed .
what can explain this seemingly counterintuitive behavior ?
when the valve of the hose connecting the two balloons is opened , the direction of air flow is determined by the two laws of thermodynamics @xcite .
suppose that @xmath4 moles of air are transfered from the bigger balloon to the smaller one .
the work needed to accomplish this is determined by the first law of thermodynamics , @xmath5 where @xmath6 , @xmath7 and @xmath8 are the environment entropy , pressure , and temperature ; @xmath9 , @xmath10 , and @xmath11 are the internal energy , pressure and volume of the gas inside the two balloons @xmath12 ; and @xmath13 is the internal energy of the balloon rubber membranes .
the second law of thermodynamics requires that @xmath14 where @xmath15 and @xmath16 are the entropy of gas and of the rubber membrane of balloon @xmath17 . since the transfer of @xmath18 moles of gas between the two balloons with finite pressure difference is an irreversible process , the total entropy of the universe will increase .
it follows that the work necessary for the transfer is bounded from below by @xmath19 from thermodynamics @xmath20 and @xmath21 where @xmath22 is the volume of rubber , @xmath23 is the surface tension , @xmath24 is the surface area , and @xmath25 is the chemical potential of gas inside the balloon @xmath17 . as a leading order approximation we can consider rubber to be incompressible ,
so that @xmath26 .
substituting eqs .
[ 1a ] and [ 1ac ] into eq .
( [ 1ab ] ) , we obtain @xmath27\;.\ ] ] for a spherical balloon , the variation in the volume and in the surface area are related by @xmath28 while the difference between the internal and the external pressures is governed by the law of laplace @xcite , @xmath29 substituting eqs .
( [ 2 ] ) and ( [ 3 ] ) into eq . ( [ 1b ] ) ,
the amount of work necessary to accomplish the transfer of air is @xmath30 for gas at fixed temperature , the gibbs - duhem equation @xcite is @xmath31 so that @xmath32 using the ideal gas equation of state and integrating , the chemical potential inside each balloon is found to be , @xmath33 where @xmath34 is the reference chemical potential at atmospheric pressure . inserting eq .
( [ 9 ] ) into eq .
( [ 4 ] ) , the amount of work necessary to transfer @xmath18 moles of gas from the larger balloon to the smaller one is , @xmath35 transfer of gas will occur spontaneously if it does not require any external work , @xmath36 .
for the specific case of two interconnected balloons this will be the case if the flow of air is from balloon with high internal pressure to the one with low internal pressure .
furthermore , since the thermodynamics requires that @xmath37 see eq .
( [ 8 ] ) , this conclusion holds for non - ideal gases as well .
although the calculation was performed for spherical balloons , it is evident that the conclusion concerning the direction of air flow , eqs .
( [ 4 ] ) and ( [ 10 ] ) , does not depend on the specific balloon shape . on the other hand ,
if the same balloons are filled with a heavy gas and are placed at different heights in the gravitational field , the flow of air will no longer be governed simply by the difference in internal pressure between the two balloons . nevertheless , while eq .
( [ 10 ] ) will loose its validity , eq .
( [ 4 ] ) will remain correct , however the chemical potential of gas will have to include a gravitational contribution as well .
as air enters into balloon , its rubber membrane stretches and becomes tense .
the helmholtz free energy of a rubber sheet is @xmath38 if the sheet is subjected to stress @xmath39 , its internal energy @xmath40 will vary in accordance with eq .
( [ 1ac ] ) , and its helmholtz free energy as , @xmath41 the surface tension of the sheet is then , @xmath42 it is an experimental fact @xcite that up to fairly large extensions the internal energy of a rubber sheet is independent of its area , @xmath43 this equation should be compared to a similar equation for the ideal gas the internal energy of an ideal gas is independent of the volume that it occupies .
clearly , this is not correct for real gases , but serves as a very good approximation for a gas at low density .
similarly , eq . [ 11d ] holds only for `` ideal '' rubber .
the flory theory @xcite allows us to calculate the change in entropy of an elastic object during a deformation .
suppose that the rubber sheet has dimensions , @xmath44 , @xmath45 , and @xmath46 and that after the deformation the new dimensions are @xmath47 , @xmath48 , and @xmath49 , then @xmath50 \;,\ ] ] where @xmath51 is a constant related to the number of chains and the topological structure of the polymer network . if rubber is subjected to a not very high stress it is reasonable to assume that it is incompressible , so that @xmath52 .
this means that for a uniform isotropic stress in the @xmath53 plane , @xmath54 and @xmath55 .
( [ 11e ] ) then simplifies to @xmath56 \;,\ ] ] where @xmath57 is the surface area prior to the deformation and @xmath58 is the final surface area .
it is important to note that a deformation of an elastic body results in a decrease of its entropy .
microscopically , this is a consequence of the reduction of the conformational volume accessible to a stretched polymer . substituting eq .
( [ 11f ] ) into eq .
( [ 11c ] ) the surface tension of a rubber sheet is @xmath59 \;,\ ] ] where @xmath60 .
we expect eq .
( [ 11 g ] ) to work reasonably well up to extensions on the order of @xmath61 .
if the balloon is inflated beyond this `` ideal '' limit , deviations are to be expected . in general then , the stress - strain relation for a rubber balloon of radius @xmath62 can be written as @xmath63 where @xmath64 is the balloon radius prior to inflation , and the scaling function @xmath65 governs the crossover from the ideal rubber regime to the non - ideal one .
the scale @xmath66 specifies the balloon size at which its rubber starts to behave non - ideally .
here we use the simplest form of the crossover function , @xmath67 where @xmath68 and @xmath69 . note that the function @xmath65 is continuous and differentiable everywhere .
transfer of @xmath18 moles of air from the bigger balloon to the smaller one will occur spontaneously if the pressure inside the bigger balloon is larger than the pressure inside the smaller balloon . to simplify
the notation lets measure all the lengths in units of @xmath64 and define @xmath70 with @xmath71 . using the law of laplace @xmath72 if @xmath73
, once the valve is open there will be a spontaneous transfer of gas from the larger balloon to the smaller .
on the other hand , if @xmath74 the transfer will proceed in the opposite direction . from eqs .
( [ 12 ] ) and ( [ 13 ] ) we see that @xmath75 while @xmath76 therefore , unless the function @xmath77 has a zero on the interval @xmath71 the larger balloon will always shrink , passing its air to the smaller balloon . indeed for a given @xmath65
, we find that if @xmath78 then @xmath73 for all @xmath71 , and the larger balloon always inflates the smaller one , see fig .
the maximum value of the crossover scale @xmath79 and the balloon size @xmath79 for which this behavior occurs is determined by the set of equations , @xmath80 and @xmath81 where the prime denotes differentiation with respect to @xmath82 .
we now distinguish two cases : @xmath83 and @xmath84 . for @xmath84 and @xmath85
the equation @xmath86 has two roots @xmath87 and @xmath88 and a new phase in which small balloons inflate the larger ones appears , see fig .
this phase terminates at a critical point , beyond which a large balloon with size @xmath89 will inflate any smaller balloon . for a given crossover function and crossover scale @xmath66 ,
the location of the critical point is determined by the set of equations , @xmath90 and @xmath91 .
the @xmath92 line is @xmath93 .
arrow shows the direction of the air flux .
the big balloon always inflates the smaller one .
, width=226 ] .
there are two distinct regimes , separated by a phase boundary terminating at a critical point , beyond which a large balloon will inflate any smaller balloon.,width=226 ] as the crossover scale grows , so do the values of @xmath88 and @xmath94 . for @xmath95 @xmath96 where @xmath97 is the root of equation @xmath98
if the crossover scale goes to infinity , @xmath99 , so that balloon rubber always behaves ideally , then @xmath100 and @xmath101 . in this limit
the phase diagram assumes the topology presented in fig .
[ fig6 ] . .
, width=302 ] for @xmath83 the phase diagram becomes even more interesting . for @xmath58 and @xmath66 inside the shaded area of fig .
[ fig7 ] , eq.([18 ] ) has four solutions @xmath87 , @xmath88 , @xmath102 , and @xmath103 , and the possible topologies of the phase diagram are shown in figs .
[ fig4 ] and [ fig5 ] . as the boundary of the shaded region is approached from inside , one of the phases shrinks and disappears and the phase diagram assumes the topology of fig .
[ fig3 ] . outside the shaded region ,
the only possible topologies are the ones presented in figs .
[ fig2 ] , [ fig3 ] , and [ fig6 ] . and [ fig5 ] .
outside the shaded region the topology is the one presented in figs .
[ fig2 ] , [ fig3 ] , and [ fig6 ] . as @xmath104 the shaded area shrinks to zero.,width=302
] the topological change from fig .
[ fig4 ] to fig .
[ fig5 ] occurs as @xmath105 .
as @xmath104 , the shaded area in fig .
[ fig7 ] shrinks to zero and the topology of the phase diagram reduces to the one presented in fig .
[ fig3 ] . .
as @xmath105 there is a topological change with the new phase diagram shown in fig .
[ fig5].,width=302 ] .
, width=302 ]
we have explored the thermodynamics of air transfer between two partially inflated rubber balloons . surprisingly , for such an apparently simple system a very rich phase diagram governing the air transfer between the two balloons is obtained .
we find that depending on the elasticity of balloon rubber and the balloon sizes , the air can flow either from the larger balloon to the smaller one , or _
vice versa_. the topology of the phase diagram is controlled by the crossover function which characterizes the deviation of balloon rubber from the ideal flory behavior . | two identical rubber balloons are partially inflated with air ( to different extent ) and connected by a hose with a valve .
it is found that depending on balloon volumes , when the valve is opened the air will flow either from the larger ( fuller ) balloon to the smaller ( emptier ) balloon , or from the smaller balloon to the larger one . the phenomenon is explained in terms of the non - ideal rubber elasticity of balloons .
the full phase diagram for the air flow dynamics is constructed . |
quantum key distribution ( qkd ) systems @xcite establish shared secret keys between two legitimate users ( alice and bob ) .
an eavesdropper ( eve ) who makes optimal physical measurements on the channel , can know , on average , none of the bits of the secret key without inducing statistical noise that reveals her presence .
after alice and bob obtain samples of a correlated random variable by means of quantum measurements , a reconciliation phase assures agreement of alice and bob s data .
finally , a privacy amplification phase uses universal hash functions to eliminate eve s potential partial information at the cost of shrinking the length of alice and bob s shared data .
continuous variable quantum key distribution makes use of coherent quantum measurements such as balanced homodyne or balanced heterodyne detection .
the result is a continuous spectrum of outcomes , as opposed to the discrete spectrum resulting from the photon counting measurements used in discrete variable quantum key distribution . unlike photon counters , homodyne and heterodyne detectors do not require dead times and thus continuous variable qkd ( cvqkd ) systems @xcite are in principle scalable to standard telecom rates , such as 10 ghz
however , coherent quantum measurements result at minimum in vacuum noise manifesting its self as gaussian distributed randomness with noise power that remains constant as signals attenuate with transmission distance .
this limits the achievable secure link length to a smaller distance than is achievable for discrete qkd .
thus cvqkd is best applied to distances less than 60 km . the security of cvqkd has been studied both for gaussian - distributed modulation@xcite , of alice s signals and for discrete modulation@xcite . for a cvqkd protocol
, the net secret information per channel use can be expressed as @xmath0 , where @xmath1 is the efficiency of the error correcting code used in the reconciliation process .
for an ideal code that achieves channel capacity , @xmath2 .
the quantity @xmath3 is the classical mutual information for the channel and the signaling scheme between alice and bob , and @xmath4 is the holevo information between bob and eve , which bounds the upper limit of eve s accessible information regarding bob s measurements .
our recent work proved the security of a cvqkd protocol designed to relax the primary bottleneck in current implementations : the time required to accomplish reconcilation .
this time is strongly influenced by the value @xmath5 , the minimum reconciliation efficiency , which if exceeded , achieves a net positive secret key rate in terms of bits per channel use . in order to lower the efficiency requirement and increase the speed ( in bits per second ) of the error correcting code that achieves reconciliation , our protocol makes use of discrete signal modulation , reverse reconciliation , post - selection , and quantum state tomography of bob s received density matrix to bound eve s obtainable information .
in this paper we describe an experimental implementation of this protocol .
there are currently reported continuous variable qkd experiments implemented in optical fiber@xcite and others that simulate a lossy channel with a beamsplitter@xcite .
compared to the previous work , this is the first experiment to use discrete signaling over optical fiber .
this paper is also the first continuous variable system that uses a continuous - wave local oscillator ( i.e. the lo and signal are not time multiplexed ) , which we believe to be better suited for higher speed systems . to achieve this
we implement a frequency translation scheme that avoids guided acoustic wave brillouin scattering ( gawbs)@xcite , an effect which otherwise contaminates continuous wave cvqkd .
we believe this work is significant as well for potential future pulsed cvqkd systems where gawbs noise could be present at pulse rates as low as tens of mhz . this paper is also the first experiment to use optical amplification in the receiver for amplification of the lo .
the paper is organized as follows : section 2 contains a brief description of the new cvqkd protocol , a discussion of its security can be found in ref .
section 3 describes the experimental setup and system calibration .
the experimental results are in section 4 , and section 5 contains a summary .
we briefly describe the protocol @xcite in this section . it is similar to those proposed in @xcite but is distinguished by its use of tomographic measurement of bob s quantum state which allows calculation of security when post - selection is used .
it produces a relatively tight bound and allows for large secrecy capacity .
* step 1 : * in each time slot , alice sends randomly and equiprobably one of four weak coherent states , as represented in fig .
1(a ) , where the radius of the circle indicates somewhat arbitrarily the effective radius of quantum fluctuations . *
step 2 : * each timeslot is randomly assigned by bob to `` data '' or `` tomography '' subsets . for each data time slot ,
the measurement axis x or y is randomly chosen by setting local oscillator phase to @xmath6 or to @xmath7 radians .
for each tomography time slot , a random local oscillator phase of @xmath6 , @xmath8 , or @xmath7 is chosen .
bob also chooses a post - selection threshold .
data having absolute value less than this threshold are discarded .
positive valued data is assigned `` 1 '' , negative data is assigned `` -1 '' .
* step 3 : * bob now reveals which time slots contain tomography and those that contain post - selected data .
alice then reveals which state was sent during those tomography time slots . using this information
, bob performs conditional quantum tomography for the four conditional density matrices , which permits calculation of an upper bound on the holevo information for collective attacks on the system .
the protocol is aborted if the the holevo information is too large .
* step 4 : * bob then reveals the local oscillator phase associated with each post - selected data sample . for each of these samples , alice projects the transmitted state onto that axis .
* step 5 : * bob sends checkbits to alice over a public channel , i.e. reverse reconcilation .
alice corrects her data to agree with bob .
* step 6 : * alice and bob perform privacy amplification to distill the final secure key . in order to obtain positive secrecy capacity ,
it is desirable that alice and bob nearly achieve the capacity of the channel given the signal - to - noise ratio .
recently , a new result of classical information theory @xcite shows that for a lossy gaussian channel with given signal - to - noise ratio , when bob quantizes the received data , the optimal way for alice to encode data is to also send quantized data .
specifically , under the condition that bob performs binary quantization , alice needs only send binary data and achieve the channel capacity .
this result is significant for reverse - reconciliation cvqkd because it indicates that if bob quantizes the data received , then alice does not need to send gaussian modulated signals but should send binary signals .
the quantized input - quantized output(qiqo ) cvqkd protocol is described below : * step 1 : * alice randomly picks up a random variable @xmath9 and encodes a coherent state @xmath10,where @xmath11 is a positive real number depending on bob s signal - to - noise ratio and @xmath12 denotes the index of time slot , and sends it through a lossy and noisy quantum channel .
alice s encoding scheme can be described in fig .
[ fig : encoding ] .
* step 2 * bob receives a quantum state from the quantum channel . with probability @xmath13 , each measurement is assigned to channel characterization , where bob randomly chooses a local oscillator phase @xmath14 of @xmath6 , @xmath8 or @xmath15 , makes a homodyne measurement and records the real result @xcite . with probability @xmath16 , that measurement is assigned a data collection index @xmath12 , where bob randomly chooses a local oscillator phase @xmath14 of @xmath6 or @xmath7 before performing homodyne detection . if his measurement result is greater than @xmath17 , where @xmath18 is bob s decision threshold , then he quantizes the result to @xmath19 .
if bob s measurement result is less than @xmath20 otherwise , he quantizes the data to @xmath21 . for other cases where his measurement result is between @xmath20 and @xmath17 , bob
quantizes his data to @xmath22 . when @xmath23 , the data from the corresponding time slot will not be selected in the post processing .
when @xmath24 , it reduces to the case without post - selection . to summarize these two steps ,
alice uses random qpsk signaling but bob s collected data are digitized bpsk .
* step 3 : * when all quantum communication has been finished , bob reveals to alice which time slots that were used for characterization phase measurements .
alice reveals to bob the state that she has sent for those time slots .
then bob performs conditional quantum tomography for each one of the four particular coherent states that alice sent .
only three different collection angles are required to achieve a good estimate of the received state @xcite .
we know that without eve , the channel can be modeled as a beamsplitter with two inputs , one of which is alice s output to the quantum channel and the other one is the excess channel noise mode .
@xmath25 where @xmath26 is the output of beamsplitter going to bob s detectors and @xmath27 is quantum efficiency of the quantum channel . for any field quadrature of bob , we have @xmath28 assume @xmath29 , @xmath30 and @xmath31 are the possibility distributions of the three quadratures , we have @xmath32 where * is a convolution . by fourier
transform techniques , we can find @xmath33 once we know @xmath30 and got @xmath29 from tomography .
therefore , we can also reconstruct @xmath34 based on quantum tomography on @xmath35 . for the protocol
, bob performs quantum conditional tomography for all four cases .
then bob can reconstruct @xmath34 for all four cases .
the protocol requires that the reconstructed @xmath34 for all four cases to be the same .
otherwise , alice and bob abort the protocol .
* step 4 : * for each data collection time slot , bob reveals the local oscillator phase that was chosen .
if bob used @xmath36 , then alice records @xmath37 for the case where @xmath38 or @xmath39 and @xmath40 for the case where @xmath41 or @xmath42 . if bob used @xmath43 , then alice records @xmath37 for the case where @xmath38 or @xmath41 and @xmath40 for the case where @xmath39 or @xmath42 .
* step 5 : * bob sends checkbits to alice over a public channel , i.e. _ reverse reconciliation_. the reconciliation is strictly one - way . * step 6 : * alice and bob perform privacy amplification to distill the final secure key .
schematic of experiment , width=415,height=291 ]
alice transmits a continuous - wave local oscillator with orthogonally polarized weak coherent states , shown schematically in the upper part of fig .
2 . phase modulator pm2 sets the phase shift between the lo and quantum signal .
alice s information is encoded in the relative phase of the lo to coherent - state 50 mhz sidebands , which are created with an amplitude modulator ( am ) driven at 50 mhz . by biasing the am near
the extinction point the signal light is placed entirely in the two sidebands .
these 50 mhz sidebands then undergo a 2 ghz phase modulation while passing through pm1 . the drive voltage for pm1
is chosen so that the phase shift amplitude @xmath44 corresponds to the first root of the bessel function @xmath45 .
the result is that the signal light is ideally entirely frequency shifted away from the optical lo freqency creating sidebands spaced 2 ghz apart . in order to simplify fig .
2 , only two of these sidebands are represented .
two separate phase modulators are used in order to safely limit the rf power per modulator .
the lo and signal are combined on a polarization beam splitter ( pbs1 ) then sent down the transmission channel fiber having linear loss of 5.18db .
in the fiber , guided acoustic wave brillouin scattering ( gawbs ) causes a portion of the lo to scatter into frequencies up to 1.5 ghz .
these frequencies correspond to thermally populated acoustical phonon modes in the fiber that modulate the index of the fiber core .
the scattering is composed of co - polarized and de - polarized components . at the receiver , after 24.2 km of fiber , the lo power is about -6 dbm .
bob separates the lo and signal by means of pbs2 , where the lengths of the lo and signal paths before recombining are matched to less than 1 mm . due to imperfections in the pbs , about @xmath46 of the lo leaks into the signal path .
the small amount of leaked lo contains detectable gawbs noise having frequency components up to 1.5 ghz , but the frequency shifted signal remains uncontaminated as it inhabits frequencies beyond the gawbs noise spectrum . at the lo port @xmath47 of the lo
enters the lo path , passing through pm3 which is modulated in the same way as pm1 , ideally shifting all of the lo to sidebands .
subsequently pm4 performs bob s random selection of phase then an optical amplifier boosts the power of the lo to 15.0 dbm , filtered by a 0.8 nm optical bandpass filter and then passed through a polarizer . the signal and lo are mixed on a 50/50 fiber beam splitter ( 49.8/50.2 in practice ) and guided to two photodiodes ( epitaxx etx75 ) having a 1.1 ghz 3-db bandwidth . because the diodes have no frequency response beyond 1.4 ghz , the different optical sidebands do not beat with each other .
the lo that specifies the single optical mode that undergoes quantum homodyne detection thus corresponds to 50 mhz double sidebands around a sequence of several optical frequencies separated by 2 ghz , but having no component at the center frequency , that of the original lo laser .
a filter follows each photodiode , separating frequencies less than 5 mhz from those greater than 5 mhz .
each photocurrent then enters a 180 degree hybrid bridge ( anzac h-9 ) producing an rf difference photocurrent .
the photocurrent passes through a 25 mhz highpass filter , a 50 db - gain electrical amplifier with noise figure 0.9 db , then a mixer that brings the 50 mhz rf frequencies to the base band .
the downconverted photocurrent is finally filtered by a 1.9 mhz filter for a / d sampling .
the system is controlled by a computer which uses a training frame to perform real - time adjustment of the phase drift between the signal and lo by addition of a constant to bob s phase input to pm4 .
the computer also performs data post - selection and classical data correction .
the detection setup realizes between 64 and 65 db of common mode noise suppression during experimental runs .
this balancing , reached by inducing small loss on one optical fiber and electrical path matching , is sufficient to suppress excess noise due to laser rin noise ( 7 db excess noise ) , gawbs noise on the lo ( 15 db excess noise ) , and the edfa ( 10 db excess noise ) .
the overall quantum efficiency is 0.56 , with pbs efficiency 0.795 , fiber beamsplitter efficiency of 0.98 , photodiode efficiency 0.74 , and effective transmission losses of 0.02 due to imperfect fiber beamsplitter ratio . during experiments ,
7 mw of lo power impinges on each detector . at this power level
the electric noise is 0.069 shot noise units . even when no quantum signal is present , there remains residual excess noise present only when bob s signal path is connected .
this noise is believed to be residual gawbs noise and results in 0.002 to 0.005 shot noise units of excess noise remaining at the receiver .
we hypothesize that the frequency translation scheme is imperfect due to uncertainty in polarization , modulation voltage , and possibly due to modulator waveguide imperfections . by comparing this to the excess noise noise present when the 2 ghz rf signal is turned off and 1.0 extra shot noise unit of gawbs noise
is measured , we estimate 27 db of gawbs noise suppression .
no excess noise is present when the length of the channel is 0 km and when the 2 ghz rf signal is either turned off or left on .
for a @xmath48 km channel , a received signal - to - noise ratio of @xmath49 ( ratio of signal power to standard deviation of the shot noise ) and post - selection threshold @xmath50 shot noise units meet the requirements of the error correction code used in reconciliation . because in principle eve could replace the communications channel with a gawbs - free channel , adding a controlled noise - like source , the excess noise is assumed to be under the control of eve .
according to the protocol , conditional tomography for all four states has been performed .
the results show that the average excess noise of the quantum channel at the detector is @xmath51 shot noise units , of which @xmath51 is due to gawbs noise and any remaining imperfections due to the phase estimation , amplitude modulation , and phase modulation are small and difficult to measure . in fig.[fig : tomography ] the raw homodyne tomography histograms are shown for @xmath52 samples per phase , which show excellent agreement with the expected gaussian distribution for coherent light with very small excess noise . for each of the four signal states transmitted by alice , three angles used by bob s tomography . given the channel transmission , detection quantum effiency , excess noise , the post - selection threshold , and the efficiency of the error correction code ( efficiency @xmath53 , error rate @xmath54 ) , we operate in a secure region@xcite , obtaining a final key rate of 3.45 kb / sec .
we note that unlike previous experiments , this experiment is not constrained by the time required for the error correction code , but by the data rate , which is limited by the 2 ms / s data acquisition and control card ( national instruments pci-6115 ) .
we have not implemented automatic polarization control at the input of bob s pbs2 , so the system operates well for 7 minutes before the polarization needs to be readjusted . the same experiment operating at a 20 mhz clock rate
would leave us with a final key rate of approximately 60 kb / sec , which compares to the best current rate of 2 kb / sec for 25 km of fiber .
we have experimentally implemented a qkd system based on based on discretely signaled continuous variables using a continuous wave lo that is polarization multiplexed with the signal .
gawbs noise scattering from lo to signal was avoided by a frequency conversion technique .
a final key rate of @xmath55 kb / sec at @xmath48 km was achieved , which to the best of our knowledge , is currently the best performing cvqkd system over optical fiber .
it is anticipated that future work will include new modulation schemes to achieve higher rates and the study of finite data size effects on the security of the system .
this research was funded in part by the agence nationale de la recherche as part of the project hqnet .
we thank matthieu bloch for discussions regarding ldpc coding . | we report a continuous variable key distribution system that achieves a final secure key rate of 3.45 kb / sec over a distance of 24.2 km of optical fiber .
the protocol uses discrete signaling and post - selection to improve reconciliation speed and quantifies security by means of quantum state tomography .
polarization multiplexing and a frequency translation scheme permit transmission of a continuous wave local oscillator and suppression of noise from guided acoustic wave brillouin scattering by more than 27 db .
99 c. h. bennett and g. grassard , `` quantum cryptography : public key distribution and coin tossing '' in proceedings of ieee international conference on computers , systems , and signal processing ( ieee , newyork , 1984 ) , pp .
175 - 179 .
a. k. ekert , `` quantum cryptography based on bell s theorem''**67 * * , pp .
661 - 663 ( 1991 ) .
m. a. nielsen and i. l. chuang , `` quantum computation and quantum information '' ( cambridge university press , uk , 2000 ) .
f. grosshans , g. van assche , j. wenger , r. brouri , n. j. cerf , and p. grangier , `` quantum key distribution using gaussian - modulated coherent states '' nature * 421 * , pp . 238 - 241 ( 2003 ) .
n. j. cerf , m. lvy , and g. van assche , `` quantum distribution of gaussian keys using squeezed states''**63 * * , 052311 ( 2001 ) .
f. grosshans and p. grangier , `` reverse reconciliation protocols for quantum cryptography with continuous variables''quant - ph/0204127 ( 2002 ) .
s. lorenz , n. korolkova , g. leuchs , `` continuous - variable quantum key distribution using polarization encoding and post selection '' , * 79 * , 273 - 277 ( 2004 ) . s. l. braunstein and p. van loock , `` quantum information with continuous variables''rev .
mod .
phys . * 77 * , pp .
513 - 577 ( 2005 ) .
f. grosshans , `` collective attacks and unconditional security in continuous variable quantum key distribution''**94 * * , 020504 ( 2005 ) .
m. navascus and a. acn , `` security bounds for continuous variable quantum key distribution '' * 94 * , 020505 ( 2005 )
. r. garca - patrn and n. j. cerf , `` unconditional optimality of gaussian attacks against continuous - variable quantum key distribution '' * 97 * , 190503 ( 2006 ) .
m. navascus , f. grosshans , and a. acn , `` optimality of gaussian attacks in continuous - variable quantum cryptography '' * 97 * , 190502 ( 2006 ) .
r. namiki and t. hirano , `` efficient - phase - encoding protocols for continuous - variable quantum key distribution using coherent states and postselection '' * 74 * , 032302 ( 2006 ) .
m. heid and n. ltkenhaus , `` security of coherent - state quantum cryptography in the presence of gaussian noise '' * 76 * , 022313 ( 2007 ) .
a. leverrier and p. grangier , `` unconditional security proof of long - distance continuous - variable quantum key distribution with discrete modulation '' * 102 * , 180504 ( 2009 ) .
y. zhao , m. heid , j. rigas , and n. ltkenhaus , `` asymptotic security of binary modulated continuous - variable quantum key distribution under collective attacks '' * 79 * , 012307 ( 2009 ) .
z. zhang and p. l. voss , security of a discretely signaled continuous variable quantum key distribution protocol for high rate systems " opt .
exp . *
17 * , pp .
12090 - 12108 ( 2009 ) .
j. lodewyck , m. bloch , r. garcia - patron , s. fossier , e. karpov , e. diamanti , t. debuisschert , n.j .
cerf , r. tualle - brouri , s.w .
mclaughlin , p. grangier , `` quantum key distribution over 25 km with an all - fiber continuous - variable system''**76 * * , 042305 ( 2007 ) .
s. fossier , e. diamanti , t. debuisschert , r. tualle - brouri , and p. grangier , `` field test of a continuous - variable quantum key distribution prototype '' new . j. phys . * 11 * , 045023 ( 2009 ) .
b. qi , l.l .
huang , l. qian , and h.k .
loj , `` experimental study on the gaussian - modulated coherent - state quantum key distribution over standard telecommunication fibers '' * 76 * , 052323 ( 2007 ) .
a.m. lance , t. symul , v. sharma , c. weedbrook , t.c .
ralph , p.k .
lam , `` no - switching quantum key distribution using broadband modulated coherent light''**95 * * , 180503 ( 2005 ) .
t. symul , d.j .
alton , s.m .
assad , a.m. lance , c. weedbrook , t.c .
ralph , p.k .
lam , `` experimental demonstration of post - selection - based continuous - variable quantum key distribution in the presence of gaussian noise''**76 * * , 030303(r ) ( 2007 ) .
a.j .
poustie , guided acoustic - wave brillouin scattering with optical pulses " , no .
8 ( 1992 ) .
d. levandovsky , m. vasilyev and p. kumar , `` near - noiseless amplification of light by a phase - sensitive fibre amplifier '' pramana - j . of phys .
* 56 * , 281 - 285 ( 2001 ) . |
finding an efficient separability criterion for multipartite quantum systems and characterizing entanglement have been an important problem in the field of quantum information theory . since bell formulated the statistical irregularities that can be seen in the correlations between measurement outcomes from two distant , but entangled , systems @xcite , various criteria and related notions have been discovered in the context of non - local correlations in quantum systems @xcite .
recently , it has been shown that it is possible to devise a protocol to extract more work via entangled systems from a heat bath thermodynamically than can be done from any separable state @xcite .
this is of interest because a physically useful quantity , locally extractable work , can be employed to test the existence of entanglement in bipartite quantum systems , leading to the idea of thermodynamical separability criteria . in this paper , we discuss if the thermodynamical separability criteria can be generalized to multipartite systems , particularly tripartite ones , as even the simplest transition from bi- to tripartite systems makes our problems much harder .
it has been known that , unlike bipartite systems , there are two non - equivalent classes of entanglement , i.e. the ghz and the w states , when three quantum subsystems have non - local correlations @xcite .
suppose that we are given an ensemble of tripartite systems , which is in either the separable or the ghz or the w states , where the ghz state @xcite can be in general written as @xmath0 and the w state @xcite , @xmath1 where @xmath2 and @xmath3 are two orthogonal states of a two - level system , e.g. eigenstates of a pauli spin operator @xmath4 .
our task here is to distinguish each of these two ensembles from separable states in terms of extractable work . by extractable work ,
we mean work that can be extracted thermodynamically from local heat baths under local operations and classical communication ( locc ) .
before discussing the case of tripartite entanglement , we here sketch the protocol shown in @xcite for the work - extraction from bipartite quantum systems .
if we have an ensemble of two - level systems , either classical or quantum , we can extract work of amount of @xmath5 $ ] from a heat reservoir thermodynamically @xcite , where @xmath6 , and @xmath7 are the boltzmann constant , the temperature of the heat bath , and @xmath7 is the shannon entropy of a binary random variable @xmath8 , respectively .
the variable @xmath8 corresponds to the outcome of measurement on the system and @xmath7 can be written as @xmath9 , where @xmath10 and @xmath11 are the probabilities for the two outcomes .
we will set @xmath12 for simplicity hereafter and call the unit of work
bit `` . we will also write ' ' extracting work from quantum systems `` for short , instead of ' ' extracting work thermodynamically from a heat bath via quantum systems " .
the extractable work from correlated pairs in @xcite is a simple generalization of the above case .
suppose that two distant parties , alice and bob , have an ensemble of identically prepared pairs of quantum bits ( _ qubits _ ) , which is described by a density operator @xmath13 .
first , we define the extractable work @xmath14 when alice and bob chose @xmath15 and @xmath16 as the directions of their ( projective ) measurement . after dividing the shared ensemble into groups of two pairs , alice measures one of the two qubits in a group with the projector she chose and informs bob of the outcome ( see fig .
[ protocol ] ) .
bob performs the same on his qubit of the other pair in the group . as a result of collective manipulations
, they can extract @xmath17 $ ] bits of work per pair at maximum . and
@xmath18 for their measurement and work extraction . for a half of this ensemble ,
alice measures her state with @xmath19 and bob maximizes the extractable work from his side along the direction of @xmath16 by using alice s measurement results . for the other half , they exchange their roles . ]
second , we consider a quantity @xmath20 , which is an average work extractable when we set @xmath21 and vary @xmath15 continuously over a great circle on the bloch sphere : @xmath22 the great circle is the one that maximizes the integral . the integral in eq .
( [ var_basis ] ) can be taken over the whole bloch sphere , nevertheless , it does not change the essential part of our discussion here . in @xcite , it was shown that an inequality @xmath23 is a necessary condition for a two - dimensional bipartite state @xmath13 to be separable , that is of the form , @xmath24 .
the state @xmath25 in the right - hand side ( rhs ) can be any pure product state @xmath26 .
if we integrate @xmath27 over the whole surface of the bloch sphere , then the condition ( [ theineq ] ) becomes @xmath28 where the subscript @xmath29 stands for the bloch sphere .
this condition , eq . ( [ theineqbs ] ) , has been shown to be more effective than the standard bell - clauser - horne - shimony - holt ( bell - chsh ) inequalities @xcite in detecting the inseparability of the werner state @xcite , which is a state in the form of @xmath30 .
let us now consider the case of tripartite quantum systems . by generalizing the protocol in the previous section in a simple manner , we can have a necessary condition for separability , which is of the same form as eq .
( [ theineq ] ) , straightforwardly .
that is , each one of three parties receives the outcomes of measurements from the other two and we take conditional entropies such as @xmath31 instead of @xmath32 in the definition of @xmath33 and @xmath34 .
then , the inequality ( [ theineq ] ) with @xmath35 on the left - hand side ( and a different numerical value ) holds for all separable states of the form of @xmath36 .
the proof is essentially the same as that in @xcite .
thus any violation of the inequality implies the existence of entanglement between at least two subsystems .
obviously , this extension can be applied to correlations between larger number of subsystems .
however , such an inequality tells little about the properties of multipartite entanglement .
thus , we would like to find a different protocol and focus on distinguishing the two inequivalent classes of tripartite entanglement : in this paper , we will devise a way to distinguish the ghz state , the w state , and separable states , in terms of the extractable work . with three parties ( alice , bob , and charlie ) , we consider a protocol in which one extracts work along the direction of the @xmath37-axis after receiving information on the outcomes of the other two s measurements along the @xmath38- and @xmath37- axes , respectively .
namely , for a subensemble where alice and bob measure their qubits and charlie extracts work from his , the ( average ) amount of work obtainable is @xmath39 in this equation , we denote the direction of axis by a unit vector , such as @xmath40 .
we have chosen two directions , @xmath40 and @xmath41 , for two measurements in anticipation that the difference between the ghz and the w states will be seen by varying the direction of @xmath41 , while keeping @xmath40 fixed .
we do not lose generality by choosing the direction of the work - extraction to be the same as one of the measurement axes , that is @xmath41 in eq .
( [ tri_work01 ] ) , as it is the right choice in detecting singlet - type bipartite entanglement .
also , as we focus on the ghz and the w states , which are symmetric in terms of @xmath42 and @xmath43 , we will assume throughout the paper that it is charlie who extracts work after alice and bob make measurements on their subsystems .
otherwise , we need to permute the roles of each party and take an average . and @xmath41 , respectively , where bob varies the direction of @xmath41 , choosing one of the three orthogonal directions @xmath44 at each round .
charlie extracts work along @xmath41 after receiving measurement results from alice and bob .
they exchange their roles cyclically after three rounds , however , we do not consider the exchange of roles as the states of our interest here are symmetric with respect to three parties . ]
we now show that a quantity below , @xmath45 , achieves our task .
we assume that the direction @xmath40 is given and fixed for everyone of three parties throughout our discussion . instead of varying @xmath41 continuously
, we average the extractable work over three orthogonal directions in the bloch sphere , @xmath46 , for @xmath41 .
this is mainly because of our limited resource for numerical computation , but it does not change our main task at all . with only the direction of @xmath40 fixed , those of @xmath47 and @xmath48 can not be determined .
we use @xmath49 to specify the angle between @xmath47 and a certain predetermined direction in the space . then , the averaged work can be written as @xmath50 and we take the maximum value of @xmath51 over @xmath49 to remove the @xmath49-dependence as @xmath52 we will analyse how we can use @xmath45 to distinguish three classes of correlations in tripartite systems . with a product state , @xmath53
, the extractable work depends only on the subsystem from which we extract work , regardless of the measurements on the other two .
thus , @xmath54 will be determined solely by the geometric relation between the state of the system and the choice of axes .
@xmath55 takes its maximum value of @xmath56 when the bloch vector representing the state coincides with one of the axes , i.e. @xmath57 for a given @xmath40 .
we now show that @xmath56 is indeed an upper bound for the extractable work @xmath58 from any separable state .
an inequality , @xmath59 is a necessary condition for a tripartite state @xmath13 to be separable .
this condition is very similar to the thermodynamical separability criterion for bipartite correlations @xcite , however , @xmath45 is an average of work over only three directions as in eq .
( [ tri_work02 ] ) , while it was over a great circle on the bloch sphere ( or the whole surface of the sphere ) in @xcite .
this results in a difference in the efficiency of detecting weakly entangled states : the condition ( [ tri_criterion ] ) is less efficient than the criterion in @xcite , in this sense .
_ the conditional entropy in eq .
( [ tri_work01 ] ) can be written as @xmath60 where @xmath61 means that alice ( bob ) obtained the outcome @xmath62 by the measurement along @xmath63 and @xmath64 is the probability of a joint event of @xmath65 and @xmath66 . omitting the directions of axes for simplicity ,
we can write eq . ( [ tri_proof01 ] ) as @xmath67 if the state @xmath13 is separable , i.e. @xmath68 , then the density operator for charlie after alice and bob obtained @xmath69 and @xmath70 becomes @xmath71 where @xmath72 is a projection operator for outcome @xmath73 along the direction of @xmath63 and @xmath74 .
therefore , @xmath75 because of the concavity of the shannon entropy .
if we were sure that we had a tripartite product state , @xmath76 , the conditional entropy would be @xmath77 where @xmath78 . since @xmath79\le 1/3 $ ] for all separable states @xmath80 , combining eqs .
( [ tri_proof04 ] ) and ( [ tri_proof05 ] ) , we have @xmath81 \nonumber \\ & \le & \max\left[1-\frac{1}{3}\sum_n \sum_{\vec{u } } p_n h(c^n|a^n , b^n)\right ] \nonumber \\ & \le & \frac{1}{3},\end{aligned}\ ] ] thus eq .
( [ tri_criterion ] ) .
as @xmath45 still has a dependence on the direction of @xmath40 , even a highly entangled state may not violate the inequality ( [ tri_criterion ] ) , however , once it is violated it is surely a manifestation of entanglement .
let us discuss how we can distinguish the ghz , and the w states , using the above criterion . to this end
, we make use of the remaining variable , i.e. the direction of @xmath40 . the extractable work from a ghz state can reach 1 bit , which is the maximum possible violation of the inequality ( [ tri_criterion ] ) .
this maximum is attained when we choose @xmath40 to be perpendicular to @xmath2 in the bloch sphere ( such as @xmath82 ) , the state after alice s measurement still has maximum ( bipartite ) entanglement . because of the perfect correlation between @xmath83 and @xmath43 , charlie can extract maximum work , 1 bit per one set of three subsystems .
we have obtained the value of the minimum extractable work from the ghz as 0.1619 bits numerically .
this occurs when @xmath84 and @xmath85 , for example , where @xmath15 and @xmath49 are the azimuthal and the longitudinal angles in the bloch sphere for the direction of alice s measurement basis @xmath40 .
that is , the direction denoted by @xmath86 is that of @xmath2 and @xmath49 is a relative phase between @xmath2 and @xmath3 in their superpositions .
the maximum work extractable from @xmath87 is @xmath88 , which is attained when the @xmath40-direction is parallel to @xmath2 .
if the outcome of alice s measurement is 0 , which occurs with probability @xmath89 , the other two systems , @xmath83 and @xmath43 , will still have a perfect correlation along the three orthogonal directions , being in the state @xmath90 and thus charlie can get 1 bit of work . if alice s outcome is 1 , then charlie can obtain only @xmath56 bits on average as the remaining two systems are in a product state @xmath91 . therefore , the average extractable work will be @xmath92 bits .
the minimum work from the w state was obtained numerically as 0.1696 bits .
this is achieved when @xmath93 and @xmath94 for the direction of @xmath40 .
now we can distinguish the ghz and the w states , looking at the maximum and the minimum values of @xmath58 after varying the direction of @xmath40 .
the maximum and minimum extractable work from the ghz and the w states have the following relationships : @xmath95 although the difference between the two minima is rather small .
therefore , when the state of a given ensemble is one of the three possibilities , the ghz , the w , and separable states , as we have assumed above , it is possible to specify the state by examining the range of extractable work . if @xmath96 always holds regardless of the direction of @xmath40 , then it is in a separable state due to the above proposition .
suppose that entanglement exists only in two subsystems out of three , such as @xmath97 , where @xmath98 is a maximally entangled state .
if we wish to discriminate such bipartite entanglement in tripartite systems , we need to examine the extractable work from each party , instead of simply computing the average @xmath58 . in this example of state @xmath99 , alice and bob
can obtain the maximum work ( 1 bit ) , regardless of the outcome of charlie s measurement , because of the perfect correlation between @xmath100 and @xmath83 .
however , no information from alice and bob can be useful to maximize the work charlie can extract .
thus , he can have only @xmath56 bits of work at maximum after averaging over three directions .
if one of the subsystems is disentangled from the other two , the extractable work from this site can not exceed @xmath56 , which is precisely what has been proved in the preceding subsection . by comparing the extractable work from each site
, we can distinguish tripartite correlation with bipartite entanglement . as we have mentioned above , the thermodynamical separability criterion for bipartite systems is able to detect more inseparability of the werner state of two qubits than the bell - chsh inequalities .
let us consider a werner - type state , in analogy of the bipartite werner state , with tripartite systems @xcite : we define the ghz - werner - type state as @xmath101 and the w - werner - type state as @xmath102 these states are also referred to as isotropic states @xcite .
the inequality ( [ tri_criterion ] ) turns out to be violated when @xmath103 with the ghz - werner - type state , and when @xmath104 with the w - werner - type state .
are these criteria better than other schemes of multipartite - entanglement - detection ? in @xcite , the ghz - werner - type state with @xmath105 has been shown to be distillable , thus entangled , where @xmath106 with @xmath107 in @xcite , while we have no corresponding data for the w - werner - type state .
let us take mermin s inequality as the multipartite version of bell - chsh inequalities @xcite .
mermin s inequality for a tripartite system can be written as @xmath108 where the operator @xmath109 is defined as ( by omitting the @xmath110 sign ) @xmath111 in eq .
( [ mermin_ineq02 ] ) , @xmath112 is a measurement operator at site @xmath113 and @xmath114 is a unit vector representing the direction of the measurement . the ghz state , eq .
( [ ghz ] ) , can violate the inequality ( [ mermin_ineq01 ] ) maximally as @xmath115 .
this means that the ghz - werner - type state , eq .
( [ ghz_werner ] ) , can violate eq .
( [ mermin_ineq01 ] ) when @xmath116 since @xmath117 similarly , the w - werner - type state , eq .
( [ w_werner ] ) , can violate the inequality when @xmath118 as the maximum value of @xmath119 is 3.046 @xcite .
therefore , the condition ( [ tri_criterion ] ) is less effective in detecting the entanglement in both the ghz - werner - type and the w - werner - type states : there are a class of weakly entangled states that violate mermin s inequality , but do not violate eq .
( [ tri_criterion ] ) .
.the comparison of the thermodynamical separability criterion for tripartite systems , eq .
( [ tri_criterion ] ) , and mermin s inequality , eq .
( [ mermin_ineq01 ] ) , in detecting the inseparability of the werner - type states .
this table shows the minimum values of @xmath120 to violate the inequalities , i.e. the smaller value , the more detection of inseparability . [ cols="^,^,^",options="header " , ] the situation does not change much even if we average the extractable work by varying the @xmath41 over the whole bloch sphere . the inequality ( [ tri_criterion ]
) can be violated by the ghz - werner - type state when @xmath121 and by the w - werner - type state when @xmath122 , after choosing @xmath40 optimally .
the reason why eq .
( [ tri_criterion ] ) is not as effective as mermin s inequality is yet unclear , while a similar thermodynamical criterion is more effective in the case of bipartite entanglement as shown in @xcite .
we have generalized the thermodynamical separability criterion to tripartite quantum systems and shown that it can be used to distinguish two different classes of tripartite entanglement , the ghz and the w , in terms of thermodynamically extractable work under locc .
we have also found that the criterion for tripartite systems is less effective in detecting the werner - type entanglement than that for bipartite systems .
although it is not perfectly clear if the werner - type states above are the proper counterpart of the bipartite werner state we should compare with , it appears that the separability criterion we have obtained here is not as efficient as other criteria , such as mermin s inequality , in detecting weak entanglement .
nevertheless , it is capable of distinguishing the ghz , the w , and separable states with physical quantity and this may lead to an interesting physical or information processing process , in which the w state is more useful than the ghz state , unlike most of the known processes .
we gratefully acknowledge discussions with aslav brukner .
vigui is grateful to peter holdsworth for arranging his stay in london .
maruyama is a boursier de lulb , and is supported by the european union through the project resq ist-2001 - 37559 , the action de recherche concerte de la communaut franaise de belgique , and the iuap program of the belgian federal government under grant v-18 .
vedral acknowledges funding from epsrc and the european union .
99 j. s. bell , physics ( long island city , n.y . ) * 1 * , 195 ( 1964 ) .
m. horodecki , p. horodecki , and r. horodecki , phys .
a * 223 * , 1 ( 1996 ) .
m. horodecki and p. horodecki , phys .
a * 59 * , 4206 ( 1999 ) .
m. a. nielsen , phys .
lett . * 86 * , 5184 ( 2001 ) .
b. m. terhal , phys . lett . * 271 * , 319 ( 2000 )
. k. maruyama , f. morikoshi , and v. vedral , phys .
a * 71 * , 012108 ( 2005 ) .
w. dr , g. vidal , and j. i. cirac , phys .
a * 62 * , 062314 ( 2000 ) .
d. m. greenberger , m. a. horne , and a. zeilinger , in _
bell s theorem , quantum theory , and conceptions of the universe _
, edited by m. kafatos ( kluwer , dordrecht , 1989 ) , p.69 . j. von neumann , _ mathematical foundations of quantum mechanics _ , ( english translation by r. t. beyer , princeton university press , princeton , 1955 ) . c. h. bennett , int .
* 21 * , 905 ( 1982 ) j. oppenheim , m. horodecki , p. horodecki , and r. horodecki , phys .
lett . * 89 * , 180402 ( 2002 ) .
j. f. clauser , m. a. horne , a. shimony , and r. a. holt , phys .
* 23 * , 880 ( 1969 ) .
r. f. werner , phys .
a. * 40 * , 4277 ( 1989 ) .
m. murao , m. b. plenio , s. popescu , v. vedral , and p. l. knight , phys .
a * 57 * , r4075 ( 1998 ) .
e. m. rains , phys .
a * 60 * , 173 ( 1999 ) .
n. d. mermin , phys .
lett . * 65 * , 1838 ( 1990 ) .
a. cabello , phys .
a * 65 * , 032108 ( 2002 ) . | the work extractable from correlated bipartite quantum systems can be used to distinguish entanglement from classical correlation .
a natural question is now whether it can be generalized to multipartite systems . in this paper
, we devise a protocol to distinguish the ghz , the w , and separable states in terms of the thermodynamically extractable work under local operations and classical communication , and compare the results with those obtained from mermin s inequalities . |
gaugino mediation @xcite provides an attractive framework of mediating supersymmetry ( susy ) breaking effects to the observable sector , since this framework is free from the susy flavor changing neutral current problem . in gaugino mediation , only gaugino masses and a @xmath1-parameter ( and the higgs @xmath2-term )
are assumed to be non - vanishing at the high energy scale and all susy particle masses at the low - energy scale are determined mostly by the gaugino masses at the high energy scale .
therefore , this framework is very predictive .
provided that particle contents below the grand unified theory ( gut ) scale are those in minimal supersymmetric standard model ( mssm ) , the observed higgs boson mass of 125 gev is explained with the gluino mass larger than 5 - 6 tev @xcite , which is unfortunately beyond the reach of the lhc experiments .
however , the diphoton excess recently reported by atlas @xcite and cms collaborations @xcite indicates the presence of several pairs of the vector - like matter multiplets around or below the tev scale , in addition to mssm matter contents @xcite . to explain the diphoton excess with the cross section of @xmath3 5fb
, rather large number of vector - like pairs is required as long as relevant yukawa couplings are not larger than unity .
it has been shown that the four pairs of leptonic matters with masses of @xmath3 400 gev and colored ones with masses of @xmath3 800 gev can reproduce the observed diphoton signal under the conditions that those vector - like matters form complete @xmath4 multiplets and the perturbativity of the relevant couplings is maintained up to the gut scale @xcite .
in fact , vector - like multiplets around tev scale significantly change the susy mass spectrum at the low - energy , due to remarkable changes of gauge and gaugino beta - functions @xcite : at the low - energy , ratios of sfermion masses to gaugino masses become much larger than those evaluated in mssm .
moreover , the stop trilinear coupling becomes large . as a result ,
the higgs boson mass of @xmath3 125 gev is explained with the gluino mass accessible at the lhc experiment . in this paper
, we revisit our previous study in the light of the diphoton excess , and show that the gluino mass is likely to be lighter than 2 - 3 tev .
we take account of threshold corrections to gauge couplings from the vector - like matter multiplets lighter than 1 tev , and evaluate mssm mass spectra using two - loop renormalization group equations ( rges ) , which are required due to large couplings at the gut scale .
one of the plausible models to explain the observed diphoton excess is a model which contains several pairs of vector - like fermions .
the vector - like fermions couple to a singlet scalar boson @xmath5 of 750 gev . as a susy realization , we consider the following superpotential : @xmath6 with soft susy breaking terms @xmath7 where @xmath8 and @xmath9 are the superpotential and soft susy breaking terms of the mssm , respectively , @xmath10 is a gauge singlet chiral superfield , @xmath11 and @xmath12 have @xmath13 charges of @xmath14 and @xmath15 , and @xmath16 and @xmath17 have @xmath18 and @xmath19 .
the vector - like multiplets belong to complete @xmath4 multiplets as @xmath20 and @xmath21 .
we introduce four pairs of @xmath22 and @xmath23 . here , @xmath24 , @xmath25 and @xmath26
are taken to be real positive without a loss of generality .
one of bosons , @xmath5 , in the chiral superfield @xmath10 is dominantly produced by the gluon fusion process and it subsequently decays to diphoton radiatively .
using narrow width approximation , the cross section of @xmath27 is estimated as @xmath28 where @xmath29 is a @xmath29 factor , @xmath30tev and @xmath31gev .
using mstw2008nnlo @xcite with the factorization scale of @xmath32 , @xmath33 .
then , @xmath34 where we have used @xmath35 . in our setup ,
the radiatively generated @xmath36 is large and positive ( @xmath37 ) , leading to the large mass splitting of two real states in @xmath10 .
therefore , the only lighter state contributes the diphoton cross section .
provided @xmath38 is real positive , the lighter state is cp - odd , and the cross section is @xmath39 where we take @xmath40 gev , @xmath41 gev , @xmath42 . with a similar mass spectrum and couplings can marginally explain the diphoton excess without the divergences of the coupling constants below the gut scale . ]
the typical value of the k factor is around 1.5 @xcite . with these values , @xmath26 and @xmath43
do not unify at the gut scale .
however , this may not be a problem since such a disparity may be easily generated by gut symmetry breaking terms .
, @xmath44 gev and @xmath45 . ]
in this section , we evaluate susy mass spectra with the vector - like matter fields lighter than one tev in gaugino mediation . in gaugino mediation models , we assume the k " ahler potential is the sequestered form @xcite or the no - scale form @xcite , and a susy breaking field only couples to gauge field - strength superfields directly . then ,
at the high energy scale , sfermion masses vanish at the leading order and only gaugino masses and @xmath1-term are not suppressed .
the higgs @xmath2-term , @xmath46 , may or may not be suppressed , depending on a setup .- term vanishes in e.g. scherk - schwarz susy breaking @xcite and the gravitational susy breaking @xcite . in those cases , the @xmath47-term of the chiral compensator field vanishes at the leading order . ]
gaugino mediation models are parametrized with @xmath48 , @xmath49 and @xmath50 or more conveniently , @xmath51 taking into account conditions for the correct electroweak symmetry breaking ( ewsb ) .
we identify the cut - off scale with the gut scale , @xmath52 , and @xmath48 is the universal gaugino mass at @xmath52 .
if the higgs @xmath2-term , @xmath46 , vanishes at the high energy scale , @xmath53 and @xmath54 are not free - parameters but predictions .
this possibility will be revisited later . before discussing the impact of the light vector - like matters ,
let us briefly show the expected gluino mass in gaugino mediation without vector - like matters . in fig .
[ fig : mh_mssm ] , we plot the mass of the sm - like higgs boson as a functions of the physical gluino mass .
susy mass spectra are computed using suspect v2.43 package @xcite .
the blue solid ( green dashed ) line shows the computed higgs boson mass using susyhd v1.02 code @xcite ( feynhiggs v.2.11.3 code @xcite ) .
we take @xmath44 gev , @xmath55 , @xmath45 and @xmath56 .
the blue and green bands show the uncertainties of the theoretical calculations , including the error of the measured top mass , @xmath57 gev @xcite .
although we see the difference between the results of the two different codes , the higgs boson mass of 125 gev requires a rather heavy gluino of 5 - 8 tev without extra vector - like multiplets . however , as shown in ref
@xcite , the existence of the light vector - like matter fields have significant effects on susy particle masses in gaugino mediation : 1 ) squark / gluino mass ratios as well as 2 ) @xmath58-term / squark mass ratio are enhanced . as a result ,
the predicted gluino mass becomes significantly smaller than that without vector - like matter fields since the observed higgs boson mass does not require the heavy gluino . at the one - loop level ,
the change of the beta - functions is simply given by @xmath59 where @xmath60 is the number of the vector - like matter multiplets , @xmath61 , @xmath62 and @xmath63 are the bino , wino and gluino mass , @xmath64 , @xmath65 and @xmath66 are gauge couplings of @xmath67 , @xmath68 and @xmath69 , and @xmath70 and @xmath71 are beta - functions for gauge couplings and gaugino masses , respectively .
since the susy breaking masses of @xmath16 , @xmath17 , @xmath12 and @xmath11 are much larger than @xmath25 and @xmath24 , the radiative corrections to the gauge couplings from the scalar components and fermion components should be taken into account separately .
the threshold corrections from the vector - like matter multiplets can be included as @xmath72 \nonumber \\ & - & \frac{n_5}{8\pi^2 } \frac{2}{5 } \left [ \frac{1}{4 } \ln \frac{m_{\rm susy}}{m_{l'_1 } } + \frac{1}{6 } \ln \frac{m_{\rm susy}}{m_{d'_1 } } + \frac{1}{4 } \ln \frac{m_{\rm susy}}{m_{l'_2 } } + \frac{1}{6 } \ln \frac{m_{\rm susy}}{m_{d'_2 } } \right ] \ , , \nonumber \\
g_2^{-2}(m_{\rm susy } ) & \to & g_2^{-2}(m_{\rm susy } ) - \frac{n_5}{8\pi^2 } \left [ \frac{2}{3 } \ln \frac{m_{\rm susy}}{m_l } + \frac{1}{6 } \ln \frac{m_{\rm susy}}{m_{l'_1 } } + \frac{1}{6 } \ln \frac{m_{\rm susy}}{m_{l'_2 } } \right ] \
, , \nonumber \\ g_3^{-2}(m_{\rm susy } ) & \to & g_3^{-2}(m_{\rm susy } ) - \frac{n_5}{8\pi^2 } \left [ \frac{2}{3 } \ln \frac{m_{\rm susy}}{m_d } + \frac{1}{6 } \ln \frac{m_{\rm susy}}{m_{d'_1 } } + \frac{1}{6 } \ln \frac{m_{\rm susy}}{m_{d'_2 } } \right ] \
, , \label{eq : thc}\end{aligned}\ ] ] where @xmath73 is a mass scale of the susy particles , and @xmath74 and @xmath75 are mass eigenvalues of the scalar components taking into account mass splittings from @xmath76 and @xmath77 .
now , we discuss the enhancements of sfermion mass and @xmath58-terms .
although two - loop rges are important to evaluate susy mass spectra as shown below , the enhancements of @xmath58-term and sfermion masses can be understood qualitatively by solving one - loop rges .
the @xmath58-terms of the first and second generation up - type squarks are @xmath78 where @xmath79 , @xmath80 and @xmath81 . neglecting @xmath67 and @xmath68 contributions and threshold corrections in eq.([eq : thc ] )
, we have an enhancement of @xmath82 for the fixed gluino mass at @xmath73 where @xmath83 tev and @xmath84 .
similarly , the squark mass becomes larger than that of mssm as @xmath85 accordingly , @xmath86 which is crucial for the enhancement of the higgs boson mass . ) and
gluino mass .
the horizontal axis is @xmath87 , where @xmath88 is a renormalization scale .
the red ( black ) lines show the results at the two - loop ( one - loop ) level . here , @xmath89tev , @xmath90 , @xmath40gev , @xmath41gev , @xmath91 , @xmath92 and @xmath44 gev .
, title="fig : " ] ) and gluino mass .
the horizontal axis is @xmath87 , where @xmath88 is a renormalization scale .
the red ( black ) lines show the results at the two - loop ( one - loop ) level . here , @xmath89tev , @xmath90 , @xmath40gev , @xmath41gev , @xmath91 , @xmath92 and @xmath44 gev .
, title="fig : " ] are the same as in fig .
[ fig : gauge ] . ]
although the above discussion is almost correct , two - loop rges are required to be included since the gauge couplings are quite large at the high energy scale .
a set of the two - loop rges is shown in appendix [ ap : rges ] . in fig.[fig : gauge ] , the rge runnings of the gauge couplings and gluino mass with one- and two - loop rges are shown as functions of the renormalization scale .
the model parameters are taken as @xmath40gev , @xmath41gev , @xmath90 , @xmath92 , and @xmath44 gev , and @xmath26 and @xmath43 , are fixed to be 0.4 at the weak scale .
the universal gaugino mass at the gut scale ( @xmath93gev ) is taken as @xmath94tev .
in the plot , squark masses are around 5 tev and gluino mass is around 2 tev .
the deviation of the gluino masses at the one and two - loop runnings is quite large .
the large correction comes from the following terms : @xmath95 in fact , without the above contributions , the rge running at the two - loop level is not much different from that evaluated at the one - loop level .
the difference of the rge evolutions at the one- and two - loop level affects the squark masses at the low - energy scale . in fig .
[ fig : msq ] , we show the squark mass ( the mass of the @xmath68 singlet up - type squark ) as a function of the physical gluino mass using one- and two - loop rges .
the squark mass evaluated at the two - loop level is larger by about 1 tev than the one - loop computation , which obviously affects the higgs boson mass calculation .
the calculated higgs boson mass is shown in fig .
[ fig : mh ] for @xmath96 .
we compute the higgs boson mass using susyhd , with the theoretical uncertainty including the error of the measured top mass .
the predicted gluino mass is in a range of 1.2 - 2.2 tev , including the theoretical uncertainty , which is accessible at the lhc . and other parameters are the same as in fig . [
fig : gauge ] . ] for @xmath97 gev with @xmath98 .
the other parameters are the same as in fig .
[ fig : gauge ] . ]
as mentioned earlier , the higgs @xmath2-term vanishes at the high energy scale in some setups , and @xmath53 is not a free parameter in those cases .
if the gaugino masses are taken to be universal at the gut scale , @xmath53 is predicted to be 20 - 22 , slightly depending on the susy scale .
however , non - universal gaugino masses arise in e.g. models of product group unification @xcite , where the doublet - triplet splitting problem is naturally solved . with non - universal gaugino masses , the prediction of @xmath53 is in a wider range .
the predicted @xmath53 in the case of non - universal gaugino masses are shown in fig .
[ fig : tanb ] .
we have taken @xmath99 at the gut scale but @xmath62 to be free . [
cols="^,^,^,^,^",options="header " , ] finally , we present some mass spectra in our gaugino mediation model in table [ tab : sample ] . the higgs boson mass is computed using susyhd , and the susy mass spectra are computed using suspect package with modifications of rges and an inclusion of eq.([eq : thc ] ) . in the points * i * and * ii * , the gaugino masses at the gut scale are taken to be universal . in the point
* iii * , the gaugino masses at the gut scale are non - universal , and @xmath100 is imposed and @xmath53 shown in the table is determined by the conditions for the correct electroweak symmetry breaking . in the point * iii * the relic abundance of the lightest neutralino is consistent with the observed value , @xmath101 , thanks to bino - wino coannihilation . here , the relic abundance is calculated using micromegas @xcite . slightly @xcite . ]
we have studied impacts of the light vector - like matter multiplets on the sparticle masses especially on the gluino mass in gaugino mediation , in the light of the diphoton excess .
the existence of the light vector - like matter fields enhances trilinear couplings and squark / gluino mass ratios at the infrared scale .
consequently , the observed higgs boson mass of @xmath3 125 gev is explained with a light gluino : the gluino mass is likely to be smaller than 2 - 3 tev unless @xmath53 is small as @xmath102 .
the predicted gluino is fairly light and the lhc may soon discover the gluino in a final state with multi - jets and missing transverse momentum . in cases of the universal gaugino masses ,
the abundance of the bino - like neutralino is too large compared to the observed dark matter abundance .
therefore , the small @xmath103-parity violation may be required to avoid the over - closure of the universe . however , the non - universal gaugino masses at the gut scale are expected to arise in models of the product group unification , where the doublet - triplet splitting problem is naturally solved .
if the wino mass at the high energy scale is smaller than the bino mass ( and the gluino mass ) and the bino and wino masses are mildly degenerate at the weak scale , the relic abundance of the lightest neutralino can be consistent with the observed dark matter abundance , thanks to the bino - wino coannihilation . the correct electroweak symmetry breaking is explained with the vanishing higgs @xmath2-term at the high energy scale in this case . in our paper , we have introduced susy invariant mass parameters for the singlet and vector - like matter fields . however , these mass parameters are not necessarily required and it is possible to explain their origins by the radiative breaking mechanism .
this is because the soft mass squared of the singlet is naturally driven to be negative at the low - energy scale , leading to a non - zero vacuum expectation value . in this case , the required masses for the singlet and vector - like fields are generated without explicit mass terms in the superpotential .
grants - in - aid for scientific research from the ministry of education , culture , sports , science , and technology ( mext ) , japan , no .
26104009 ( t. t. y. ) ; grant - in - aid no .
26287039 ( t. t. y. ) from the japan society for the promotion of science ( jsps ) ; and by the world premier international research center initiative ( wpi ) , mext , japan ( t. t. y. ) .
the research leading to these results has received funding from the european research council under the european unions seventh framework programme ( fp/2007 - 2013 ) / erc grant agreement n. 279972 `` npflavour '' ( n.y . ) .
the two - loop rges of gauge couplings and gauginos with @xmath104 pairs of @xmath105 and @xmath106 are @xmath107 and @xmath108 where @xmath109 and @xmath110 correspond to @xmath111 and @xmath26 , respectively , @xmath112 .
the two - loop coefficients are @xmath113 @xmath114 let us summarize beta - functions of the susy invariant parameters with @xmath84 .
we have used susyno package @xcite for obtaining the following two - loop rges .
the notation is @xmath115 @xmath116 , \nonumber \\
\beta_{\lambda_d } ^{(1 ) } & = & \lambda_d \bigl(14\lambda_d^2 + 8\lambda_l^2 - \frac{4}{15 } g_1 ^ 2 - \frac{16}{3 } g_3 ^ 2 \bigr ) , \nonumber \\
\beta_{\lambda_d } ^{(2 ) } & = & \lambda_d \bigr [ -50 \lambda_d^4 -16 \lambda_d^2 \lambda_l^2 -16 \lambda_l^4 + \frac{16}{5 } \lambda_d^2 ( g_1 ^ 2 + 20 g_3 ^ 2 ) + \frac{24}{5 } \lambda_l^2 ( g_1 ^ 2 + 5 g_2 ^ 2 ) \nonumber \\ & + & \frac{644}{225 } g_1 ^ 4 + \frac{64}{45 } g_1 ^ 2 g_3 ^ 2 + \frac{176}{9 } g_3 ^ 4 \bigr ] , \nonumber \\
\beta_{m_l}^{(1 ) } & = & m_l\bigl ( 2 \lambda_l^2 - \frac{3}{5}g_1 ^ 2 - 3 g_2 ^ 2 \bigr ) , \nonumber \\
\beta_{m_l}^{(2 ) } & = & m_l \bigr [ -18 \lambda_l^4 -24 \lambda_d^2 \lambda_l^2 + \frac{327}{50 } g_1 ^ 4 + \frac{9}{5 } g_1 ^ 2 g_2 ^ 2 + \frac{39}{2 } g_2 ^ 4 \bigr ] , \nonumber \\
\beta_{m_d}^{(1 ) } & = & m_d \bigl ( 2 \lambda_d^2 - \frac{4}{15}g_1 ^ 2 - \frac{16}{3 } g_3 ^ 2 \bigr ) , \nonumber \\
\beta_{m_d}^{(2 ) } & = & m_d \bigr [ -26 \lambda_d^4 -16 \lambda_d^2 \lambda_l^2 + \frac{644}{225 } g_1 ^ 4 + \frac{64}{45 } g_1 ^ 2 g_3 ^ 2 + \frac{176}{9 } g_3 ^ 4 \bigr ] , \nonumber \\
\beta_{m_s}^{(1 ) } & = & m_s \bigl(16 \lambda_l^2 + 24 \lambda_d^2 \bigr ) , \nonumber \\
\beta_{m_s}^{(2 ) } & = & m_s\bigl [ -48 \lambda_d^4 -32 \lambda_l^4 + \frac{32}{5 } \lambda_d^2 ( g_1 ^ 2 + 20 g_3 ^ 2 ) + \frac{48}{5 } \lambda_l^2 ( g_1 ^ 2 + 5 g_2 ^ 2 ) \bigr ] .
\nonumber \\\end{aligned}\ ] ] the beta - functions for the soft mass squared parameters of new particles are @xmath120 \ , , \nonumber \\
\beta_{m_s^2}^{(2 ) } & = & -64 \lambda_l^4 ( 2 a_{\lambda_l}^2 + m_{l'}^2 + m_{\bar l'}^2 + m_s^2 ) - 96 \lambda_d^4 ( 2 a_{\lambda_d}^2 + m_{d'}^2 + m_{\bar d'}^2 + m_s^2 ) \nonumber \\ & + & \frac{48}{5 } \lambda_l^2 \bigl [ ( m_{l'}^2 + m_{\bar l'}^2 + m_s^2 + a_{\lambda_l}^2 ) ( g_1 ^ 2 + 5 g_2 ^ 2 ) \nonumber \\ & + & 2 ( g_1 ^ 2 m_1 ^ 2 + 5 g_2 ^ 2 m_2 ^ 2 ) - 2 a_{\lambda_l } ( g_1 ^ 2 m_1 + 5 g_2 ^ 2 m_2 ) \bigr ] \nonumber \\ & + & \frac{32}{5 } \lambda_d^2 \bigl [ ( m_{d'}^2 + m_{\bar d'}^2 + m_s^2 + a_{\lambda_d}^2 ) ( g_1 ^ 2 + 20 g_3 ^ 2 ) \nonumber \\ & + & 2 ( g_1 ^ 2 m_1 ^ 2 + 20 g_3 ^ 2 m_3 ^ 2 ) - 2 a_{\lambda_d } ( g_1 ^ 2 m_1 + 20 g_3 ^ 2 m_3 ) \bigr]\ , , \nonumber \\
\beta_{m_{\bar d'}^2}^{(1 ) } & = & 2 \lambda_d^2 ( a_{\lambda_d}^2 + m_{d'}^2 + m_{\bar d'}^2 + m_s^2 ) + \frac{2}{5 } g_1 ^ 2 s \ , , \nonumber \\
\beta_{m_{\bar d'}^2}^{(2 ) } & = & -52 \lambda_d^4 ( 2 a_{\lambda_d}^2 + m_{d'}^2 + m_{\bar d'}^2 + m_s^2 ) \nonumber \\ & - & 16 \lambda_d^2 \bigl [ \lambda_l^2 ( ( a_{\lambda_d } + a_{\lambda_l})^2 + m_{d'}^2 + m_{\bar d'}^2 + m_{l'}^2 + m_{\bar l'}^2 + 2 m_s^2 ) \nonumber \\ & + & \frac{g_1 ^ 2}{5 } m_{\bar d'}^2 - \frac{g_1 ^ 2}{5 } m_{d'}^2 \bigr ] + \lambda_l^2 g_1 ^ 2 \left(\frac{16}{5 } m_{l'}^2 -\frac{16}{5 } m_{\bar l'}^2 \right ) \nonumber \\ & -&\frac{6}{5 } g_1 ^ 2 g_2 ^ 2 \bigl[m_{h_d}^2- m_{h_u}^2 + \sum_i ( m_{l_i}^2 - m_{q_i}^2 ) + 4 ( m_{l'}^2 - m_{\bar l'}^2 ) \bigr ] \nonumber \\ & + & \frac{32}{45 } g_1 ^ 2 g_3 ^ 2 \bigl [ \sum_i ( 3 m_{\bar d_i}^2 + 3 m_{q_i}^2 - 6 m_{\bar u_i}^2 ) + 12 m_{\bar d'}^2 - 12 m_{d'}^2 + 4 m_1 ^ 2 + 4 m_1 m_3 + 4 m_3 ^ 2 \bigr ] \nonumber \\ & + & \frac{2}{75 } g_1 ^ 4 \bigl [ \sum_i ( 8 m_{\bar d_i}^2 + 48 m_{\bar e_i}^2 - 3 m_{l_i}^2 + 3 m_{q_i}^2 -16 m_{\bar u_i}^2 ) - 3 m_{h_d}^2 + 15 m_{h_u}^2 \nonumber \\ & + & 32 m_{\bar d'}^2 -12 m_{l'}^2 + 60 m_{\bar l'}^2 + 644 m_1 ^ 2 \bigr ] \nonumber \\ & + & \frac{4}{5 } g_1 ^ 2 s_1 + \frac{16}{3}g_3 ^ 4 s_3 \ , , \nonumber \\ \beta_{m_{d'}^2}^{(1 ) } & = & 2 \lambda_d^2 ( a_{\lambda_d}^2 + m_{d'}^2 + m_{\bar d'}^2 + m_s^2 ) -\frac{2}{5 } g_1 ^ 2 s \ , , \nonumber \\ \beta_{m_{d'}^2}^{(2 ) } & = & -52 \lambda_d^4 ( 2 a_{\lambda_d}^2 + m_{d'}^2 + m_{\bar d'}^2 + m_s^2 ) \nonumber \\ & - & 16 \lambda_d^2 \bigl [ \lambda_l^2 ( ( a_{\lambda_d } + a_{\lambda_l})^2 + m_{d'}^2 + m_{\bar d'}^2 + m_{l'}^2 + m_{\bar l'}^2 + 2 m_s^2 ) \nonumber \\ & -&\frac{g_1 ^ 2}{5 } m_{\bar d'}^2 + \frac{g_1 ^ 2}{5 } m_{d'}^2 \bigr ] - \lambda_l^2 g_1 ^ 2 \left(\frac{16}{5 } m_{l'}^2 -\frac{16}{5 } m_{\bar l'}^2 \right ) \nonumber \\ & + & \frac{6}{5 } g_1 ^ 2 g_2 ^ 2 \bigl[m_{h_d}^2- m_{h_u}^2 + \sum_i ( m_{l_i}^2 - m_{q_i}^2 ) + 4 ( m_{l'}^2 - m_{\bar l'}^2 ) \bigr ] \nonumber \\ & -&\frac{32}{45 } g_1 ^ 2 g_3 ^ 2 \bigl [ \sum_i ( 3 m_{\bar d_i}^2 + 3 m_{q_i}^2 - 6 m_{\bar u_i}^2 ) + 12 m_{\bar d'}^2 - 12 m_{d'}^2 - 4 m_1 ^ 2 - 4 m_1 m_3 - 4 m_3 ^ 2 \bigr ] \nonumber \\ & + & \frac{2}{75 } g_1 ^ 4 \bigl [ \sum_i ( -24 m_{\bar e_i}^2 + 15 m_{l_i}^2 + m_{q_i}^2 + 48 m_{\bar u_i}^2 ) + 15 m_{h_d}^2 -3 m_{h_u}^2 \nonumber \\ & + & 32 m_{d'}^2 + 60 m_{l'}^2 -12 m_{\bar l'}^2 + 644 m_1 ^ 2 \bigr ] \nonumber \\ & - & \frac{4}{5 } g_1 ^ 2 s_1 + \frac{16}{3}g_3 ^ 4 s_3 \ , , \nonumber \\
\beta_{m_{l'}^2}^{(1 ) } & = & 2 \lambda_l^2 ( a_{\lambda_l}^2 + m_{l'}^2 + m_{\bar l'}^2 + m_s^2 ) -\frac{3}{5 } g_1 ^ 2 s \ , , \nonumber \\
\beta_{m_{l'}^2}^{(2 ) } & = & -36 \lambda_l^4 ( 2 a_{\lambda_l}^2 + m_{l'}^2 + m_{\bar l'}^2 + m_s^2 ) \nonumber \\ & + & \lambda_d^2 \bigl [ -24 \lambda_l^2 ( ( a_{\lambda_d } + a_{\lambda_l})^2 + m_{\bar d'}^2 + m_{d'}^2 + m_{l'}^2 + m_{\bar l'}^2 + 2 m_s^2 ) \nonumber \\ & + & \frac{24}{5 } g_1 ^ 2 ( m_{\bar d'}^2 - m_{d'}^2 ) \bigr ] -\frac{24}{5 } \lambda_l^2 g_1 ^ 2 ( m_{l'}^2 - m_{\bar l'}^2 ) \nonumber \\ & + & \frac{9}{5 } g_1 ^ 2 g_2 ^ 2 \bigl [ m_{h_d}^2 - m_{h_u}^2 + \sum_i ( m_{l_i}^2 - m_{q_i}^2 ) + 4 ( m_{l'}^2 - m_{\bar l'}^2 ) + 2 ( m_1 + m_2)^2 \bigr ] \nonumber \\ & -&\frac{16}{5 } g_1 ^ 2 g_3 ^ 2 \bigl [ \sum_i ( m_{\bar d_i}^2 + m_{q_i}^2 - 2 m_{\bar u_i}^2 ) + 4 ( m_{\bar d'}^2 - m_{d'}^2 ) \bigr ] \nonumber \\ & + & \frac{1}{25 } g_1 ^ 4 \bigl [ \sum_i ( 2 m_{\bar d_i}^2 + 2 m_{q_i}^2 + 56 m_{\bar u_i}^2 - 18 m_{\bar e_i}^2 + 18
m_{l_i}^2 ) + 18 m_{h_d}^2 \nonumber \\ & + & 8 m_{\bar d'}^2 + 40 m_{d'}^2 + 72 m_{l'}^2 + 981 m_1 ^ 2 \bigr ] \nonumber \\ & - & \frac{6}{5 } g_1 ^ 2 s_1 + 3 g_2 ^ 4 s_2 \ , , \nonumber \\
\beta_{m_{\bar l'}^2}^{(1 ) } & = & 2 \lambda_l^2 ( a_{\lambda_l}^2 + m_{l'}^2 + m_{\bar l'}^2 + m_s^2 ) + \frac{3}{5 } g_1 ^ 2 s \ , , \nonumber \\
\beta_{m_{\bar l'}^2}^{(2 ) } & = & -36 \lambda_l^4 ( 2 a_{\lambda_l}^2 + m_{l'}^2 + m_{\bar l'}^2 + m_s^2 ) \nonumber \\ & + & \lambda_d^2 \bigl [ -24 \lambda_l^2 ( ( a_{\lambda_d } + a_{\lambda_l})^2 + m_{\bar d'}^2 + m_{d'}^2 + m_{l'}^2 + m_{\bar l'}^2 + 2 m_s^2 ) \nonumber \\ & - & \frac{24}{5 } g_1 ^ 2 ( m_{\bar d'}^2 - m_{d'}^2 ) \bigr ] + \frac{24}{5 } \lambda_l^2 ( m_{l'}^2 - m_{\bar l'}^2)\nonumber \\ & - & \frac{9}{5 } g_1 ^ 2 g_2 ^ 2 \bigl [ m_{h_d}^2 - m_{h_u}^2 + \sum_i ( m_{l_i}^2 - m_{q_i}^2 ) + 4 ( m_{l'}^2 - m_{\bar l'}^2 ) - 2 ( m_1 + m_2)^2 \bigr ] \nonumber \\ & + & \frac{16}{5 } g_1 ^ 2 g_3 ^ 2 \bigl [ \sum_i ( m_{\bar d_i}^2 + m_{q_i}^2 - 2 m_{\bar u_i}^2 ) + 4 ( m_{\bar d'}^2 - m_{d'}^2 ) \bigr ] \nonumber \\ & + & \frac{1}{25 } g_1 ^ 4 \bigl [ \sum_i ( 10 m_{\bar d_i}^2 + 4 m_{q_i}^2 -8 m_{\bar u_i}^2 + 54 m_{\bar e_i}^2 ) + 18 m_{h_u}^2 \nonumber \\ & + & 40 m_{\bar d'}^2 + 8 m_{d'}^2 + 72 m_{\bar l'}^2 + 981 m_1 ^ 2 \bigr ] \nonumber \\ & + & \frac{6}{5 } g_1 ^ 2 s_1 + 3 g_2 ^ 4 s_2 , \nonumber \\\end{aligned}\ ] ] where @xmath121 + \delta s \ , , \end{aligned}\ ] ] with @xmath122 .
the additional contributions to one - loop beta - functions for soft mass squared parameters of mssm fields are @xmath123 the two - loop beta - functions of the soft mass squared parameters are modified as @xmath124 where @xmath125 beta - functions of new @xmath2-terms are @xmath126 , \nonumber \\
\beta_{m_l}^{(1 ) } & = & \bigl ( 4 \lambda_l^2 a_{\lambda_l } + \frac{6}{5}g_1 ^ 2 m_1 + 6 g_2 ^ 2 m_2 \bigr ) , \nonumber \\
\beta_{m_l}^{(2 ) } & = & \bigl [ -72\lambda_l^4 a_{\lambda_l } -48\lambda_l^2\lambda_d^2(a_{\lambda_l } + a_{\lambda_d } ) \nonumber \\ & - & \frac{654}{25 } g_1 ^ 4 m_1 - \frac{18}{5 } g_1 ^ 2 g_2 ^ 2(m_1+m_2 ) - 78 g_2 ^ 4 m_2 \bigr ] , \nonumber \\
\beta_{m_d}^{(1 ) } & = & \bigl ( 4 \lambda_d^2a_{\lambda_d}+ \frac{8}{15}g_1 ^ 2 m_1 + \frac{32}{3 } g_3 ^ 2 m_3 \bigr),\nonumber \\ \beta_{m_d}^{(2 ) } & = & \bigl [ -104 \lambda_d^4 a_{\lambda_d } -32 \lambda_l^2\lambda_d^2(a_{\lambda_l } + a_{\lambda_d } ) \nonumber \\ & - & \frac{2576}{225}g_1 ^ 4 m_1 - \frac{128}{45}g_1 ^ 2 g_3 ^ 2(m_1+m_3 ) - \frac{704}{9 } g_3 ^ 4 m_3 \bigr ] .
\nonumber \\\end{aligned}\ ] ] k. inoue , m. kawasaki , m. yamaguchi and t. yanagida , phys .
d * 45 * , 328 ( 1992 ) .
d. e. kaplan , g. d. kribs and m. schmaltz , phys .
d * 62 * , 035010 ( 2000 ) [ hep - ph/9911293 ] ; z. chacko , m. a. luty , a. e. nelson and e. ponton , jhep * 0001 * , 003 ( 2000 ) [ hep - ph/9911323 ] .
d. buttazzo , a. greljo and d. marzocca , arxiv:1512.04929 [ hep - ph ] ; j. ellis , s. a. r. ellis , j. quevillon , v. sanz and t. you , arxiv:1512.05327 [ hep - ph ] ; k. m. patel and p. sharma , arxiv:1512.07468 [ hep - ph ] ; n. craig , p. draper , c. kilic and s. thomas , arxiv:1512.07733 [ hep - ph ] ; l. j. hall , k. harigaya and y. nomura , arxiv:1512.07904 [ hep - ph ] ; y. l. tang and s. h. zhu , arxiv:1512.08323 [ hep - ph ] ; f. wang , w. wang , l. wu , j. m. yang and m. zhang , arxiv:1512.08434 [ hep - ph ] ; e. palti , arxiv:1601.00285 [ hep - ph ] ; w. chao , arxiv:1601.00633 [ hep - ph ] ; b. dutta , y. gao , t. ghosh , i. gogoladze , t. li , q. shafi and j. w. walker , arxiv:1601.00866 [ hep - ph ] ; a. djouadi , j. ellis , r. godbole and j. quevillon , arxiv:1601.03696 [ hep - ph ]
. s. f. king and r. nevzorov , arxiv:1601.07242 [ hep - ph ] .
r. franceschini _ et al .
_ , arxiv:1512.04933 [ hep - ph ] .
e. cremmer , s. ferrara , c. kounnas and d. v. nanopoulos , phys .
b * 133 * , 61 ( 1983 ) ; j. r. ellis , a. b. lahanas , d. v. nanopoulos and k. tamvakis , phys .
b * 134 * , 429 ( 1984 ) ; j. r. ellis , c. kounnas and d. v. nanopoulos , nucl .
b * 247 * , 373 ( 1984 ) .
j. scherk and j. h. schwarz , phys .
b * 82 * , 60 ( 1979 ) ; j. scherk and j. h. schwarz , nucl .
b * 153 * , 61 ( 1979 ) .
izawa , t. kugo and t. t. yanagida , prog .
phys . *
125 * , 261 ( 2011 ) [ arxiv:1008.4641 [ hep - ph ] ] ; s. iwamoto , t. t. yanagida and n. yokozaki , ptep * 2015 * , 073b01 ( 2015 ) [ arxiv:1407.4226 [ hep - ph ] ] ; k. harigaya , t. t. yanagida and n. yokozaki , phys . rev .
d * 91 * , no . 7 , 075010 ( 2015 ) [ arxiv:1501.07447 [ hep - ph ] ] .
j. p. vega and g. villadoro , jhep * 1507 * , 159 ( 2015 ) [ arxiv:1504.05200 [ hep - ph ] ] .
s. heinemeyer , w. hollik and g. weiglein , comput .
commun . * 124 * , 76 ( 2000 ) [ hep - ph/9812320 ] ; s. heinemeyer , w. hollik and g. weiglein , eur . phys .
j. c * 9 * , 343 ( 1999 ) [ hep - ph/9812472 ] ; g. degrassi , s. heinemeyer , w. hollik , p. slavich and g. weiglein , eur .
j. c * 28 * , 133 ( 2003 ) [ hep - ph/0212020 ] ; m. frank , t. hahn , s. heinemeyer , w. hollik , h. rzehak and g. weiglein , jhep * 0702 * , 047 ( 2007 ) [ hep - ph/0611326 ] ; t. hahn , s. heinemeyer , w. hollik , h. rzehak and g. weiglein , phys .
lett . * 112 * , no .
14 , 141801 ( 2014 ) [ arxiv:1312.4937 [ hep - ph ] ] .
g. belanger , f. boudjema , a. pukhov and a. semenov , comput .
commun . * 149 * , 103 ( 2002 ) [ hep - ph/0112278 ] ; g. belanger , f. boudjema , a. pukhov and a. semenov , comput .
commun . * 174 * , 577 ( 2006 ) [ hep - ph/0405253 ] . | the 750 gev diphoton excess reported by atlas and cms indicates the presence of several pairs of the vector - like matter multiplets around tev scale .
if that is the case , radiative corrections from the @xmath0 gauge interaction significantly change from those of mssm , and the infrared - free nature of the gauge interaction leads to characteristic susy mass spectra : a ratio of a squark mass to the gluino mass , and scalar trilinear couplings are enhanced at the low - energy scale .
consequently , even in gaugino mediation models , the higgs boson mass of 125 gev is explained with the fairly light gluino of 2 - 3 tev , which can be accessible at the lhc . |
the rare earths , across which the 4@xmath5 level fills , provide an enormous range of interesting and potentially exploitable spin / orbit physics .
in elemental form they are metallic , but they are also prominent as the sites of magnetic order in a variety of insulating compounds , where furthermore transitions among internal spin - orbit configurations in their 4@xmath5 shells allow for a rich optical activity .
recently there has been a renewed interest in these elements in their simple nacl structured mononitrides , with theoretical suggestions of ferromagnetic ( fm ) semiconductors and half metals .
@xcite experimental verification of the predictions is lacking for most of the series , though recent studies have identified several as intrinsic fm semiconductors .
the most thoroughly studied is gdn , lying at the centre of the series .
the half - filled 4@xmath5 shell on the gd@xmath2 ions have ( l , s ) of ( 0,7/2 ) ( @xmath6s@xmath7 ) and a magnetic moment of 7 @xmath8 .
gdn is fm below a curie temperature ( t@xmath9 ) most often found to be near 65 - 70 k , and with 32 aligned gd ions per nm@xmath10 it has one of the largest magnetizations known .
@xcite the next lightest rare - earth , eu@xmath2 , has a j = 0 ground state ( @xmath11f@xmath12 ) , which then prevents any magnetic order , though its propensity for mixed valence leads to unusual magnetic behaviour in eun .
@xcite sm@xmath2 , at one lighter again , has a ground state of @xmath0h@xmath1 and a 4@xmath5 magnetic moment of @xmath13 . in the free ion
the moment is 0.8 @xmath8 , but it is commonly smaller in the crystal fields of solids .
the earliest magnetic investigations of smn , performed forty years ago , found a very small moment below a transition near 25 k. it was then assigned as antiferromagnetic based on a vanishingly small net moment in the ordered phase .
@xcite subsequently a neutron scattering study suggested instead a fm phase with near cancellation between the spin and orbital moments .
@xcite a fm state with a 27 k t@xmath9 has recently been confirmed by magnetisation data on a thin film .
@xcite that study showed that in the crystal field of smn the paramagnetic ( pm ) moment is 0.45 @xmath8 , smaller than the free ion moment . in the fm state it drops to the remarkably small value of @xmath14 .
the moment then couples only weakly to an external field , so that the coercive field rises above the 6 t limit that was available in the reported squid study . in the past decade
there have been a number of reports devoted to metallic sm compounds that similarly have near zero net moments in their fm phases .
@xcite in these , as in smn , the 4@xmath5 spins order ferromagnetically but with the resulting spin moment largely cancelled by the 4@xmath5 orbital contribution .
in addition to the intrinsic fundamental interest in such moment - free fm materials , these show promise as fringe - field free sources of spin - polarised electrons for injection into conventional semiconductors , or for super - hard fm layers to pin the moments of softer magnetic layers . among these small - moment fm materials , smal@xmath15 has had special attention .
it has a modest net moment in the fm phase of 0.26 @xmath8 per formula unit , oriented anti - parallel with the 4@xmath5 spin moment and thus parallel with the 4@xmath5 orbital moment .
its net moment is larger by a factor of about 5 in comparison with , for example , smzn or smcd , but in contrast these have net moments that are parallel to the 4@xmath5 spin moment .
@xcite it is that difference that has encouraged the attention paid to smal@xmath15 , for the ease with which the spin moment can be increased , with the introduction of substitutional gd ions , to fully compensate the orbital moment .
@xcite since in any case the net 4@xmath5 moment on sm@xmath2 shows a temperature dependence from admixture with the j=7/2 excited state at 1500 k , exact compensation finally occurs at only one crossover temperature .
it will be shown below that fm smn shares with smal@xmath15 the character that the net moment lies antiparallel to the 4@xmath5 spin moment .
interestingly an lsda+@xmath4 ( local spin density approximation with hubbard - u corrections ) calculation also suggests a moment of 0.03 @xmath8 per sm ion in the fm phase of smn .
@xcite that remarkably small moment results from not only a near cancellation between 4@xmath5 spin and orbital contributions , each of about 5 @xmath8 , but also has contributions of order 0.1 @xmath8 from the sm 5@xmath16 and n 2@xmath17 shells .
the agreement is surprising , in view of the @xmath18 computational accuracy that would be required .
the calculation gave a net magnetisation coincident with the 4@xmath5 spin moment direction , which would then prevent the use of gd to achieve complete cancellation .
the magnetisation measurements do not reveal whether the 4@xmath5 spin or orbital moment dominates , so for that we turn to the present x - ray magnetic circular dichroism ( xmcd ) investigation .
the conducting characteristics of smn , unlike smal@xmath15 , smzn or smcd , points to it being a semiconductor , as appears to be common among the rare - earth nitrides .
@xcite the implication is that the fm state relies on indirect inter - ion 4@xmath5 exchange involving the sm 5@xmath16 and n 2@xmath17 orbitals , leading to a lower t@xmath9 than in metallic compounds .
clearly fm semiconductors will show distinct advantages for integration with conventional semiconductors , such as the possibility of impedance matching for spin injection .
the very small moment of smn makes conventional magnetometer studies especially difficult , which as discussed above led to a long - term uncertainty concerning the magnetic order in smn .
the problem is exacerbated in thin film geometries that are of current interest for devices , for there is then very little material available for the measurement .
in contrast xmcd measures not the total moment but rather probes the spin- and orbital - moments independently , with atomic - species and orbital - shell selectivity .
we thus have investigated smn using xmcd at the sm l@xmath3 edges , which concerns a 2@xmath19 promotion of an electron on the sm@xmath2 ion .
work at the m@xmath20 edges would access the 4@xmath5 moments directly , but the short attenuation length of 1 kev electrons renders the study problematic through the passivating cap that is required to prevent oxidation in the rare - earth nitrides .
however , the strong intra - ion 4@xmath5 - 5@xmath16 exchange interaction facilitates the l - edge investigation of the 4@xmath5 spin alignment through the passivating cap .
the l - edge results then benefit from comparison with calculated xmcd spectra to extract reliable 4@xmath5 alignment information .
the calculation references the xmcd spectrum to the 4@xmath5 spin alignment , while the data are referenced to the applied field direction , and this permits for the determination of the spin alignment relative to the applied field .
the epitaxial smn film used in this study was grown by evaporating sm in the presence of nitrogen gas onto ( 111 ) mgo held at a temperature of @xmath21400 @xmath22c .
the ( 111)-oriented film was grown at a rate of 100 nm hr@xmath23 with a pressure of 10@xmath24 mbar of n@xmath15 ; under these conditions the ratio of n@xmath15 to sm flux on the substrate is about 200 . after completing the growth ,
the films were capped _ in situ _ with aln .
the film growth was monitored by reflection high - energy electron diffraction ( rheed ) to confirm the epitaxial nature of the films , and subsequent _ ex situ _ x - ray diffraction confirmed the orientation and high crystalline quality .
x - ray - absorption spectroscopy ( xas ) and xmcd measurements were performed at the sm l@xmath3 edges on the id12 beam line at the european synchrotron radiation facility in grenoble .
the signal was detected by total fluorescence yield for which the capping layer is not an impediment at l - edge energies .
the film was mounted at close to grazing incidence on the cold finger of a he gas constant flow cryostat , which has been inserted in the bore of a 6 t superconducting split - coil magnet .
temperature was monitored closely adjacent to the sample on the cold finger .
the band structure and xmcd calculations have been performed for the face - centered cubic structure of sodium chloride ( space group @xmath25 , no .
225 ) with lattice constant @xmath26=5.1 using fully relativistic spin - polarized linear - muffin - tin - orbital ( lmto ) method @xcite with the combined correction term taken into account .
we used the perdew - wang @xcite parametrization for the exchange - correlation potential .
brillouin zone integrations were performed using the improved tetrahedron method @xcite and charge self - consistency was obtained with 349 irreducible @xmath27 points . to improve the potential we include additional empty spheres . the basis consisted of sm @xmath28 , @xmath17 , @xmath16 and @xmath5 ; n @xmath28 , @xmath17 , and @xmath16 ; and empty sphere @xmath28 and @xmath17 lmto s .
the intrinsic broadening mechanisms have been accounted for by folding xmcd spectra with a lorentzian . for the finite lifetime of the core hole a constant width @xmath29 , in general form
, @xcite has been used .
the finite resolution of the spectrometer has been accounted for by convolution with a gaussian of width 0.6 ev . in order to simplify the comparison of the theoretical x - ray isotropic absorption spectra of smn to the experimental ones we take into account the background intensity which affects the high energy part of the spectra and
is caused by different kinds of inelastic scattering of the electron promoted to the conduction band above the fermi level due to x - ray absorption ( scattering on potentials of surrounding atoms , defects , phonons , etc . ) . to calculate the background spectra we used the model proposed by richtmyer _
et al . _
@xcite ( for details see ref . ) .
we have adopted the lsda+@xmath4 method @xcite as a different level of approximation to treat the electron - electron correlations .
we used the rotationally invariant fully relativistic lsda+@xmath4 method .
@xcite the effective on - site coulomb repulsion @xmath4 was considered as an adjustable parameter .
we used @xmath4 = 8 ev . for the exchange integral @xmath30 the value of 0.66 ev estimated from constrained lsda calculations
sm l@xmath3 xas and xmcd spectra taken at a temperature of 15 k and in a 6 t magnetic field are compared in fig .
[ fig : xasandxmcd ] with simulated spectra .
the computed xas spectra were adjusted to be in quantitative agreement with the experimental results to achieve an excellent match .
the data and calculation both refer to the fm phase , but xas data taken at higher temperature show no significant difference in the pm phase .
both the computed and experimental xmcd are normalised to the xas results , so one can make absolute amplitude as well as spectral comparisons . here the agreement is imperfect , so in fig .
[ fig : xasandxmcd ] the computed xmcd spectra have been scaled by the factors listed in the figure to fit at least the most prominent features in the data . in the case of l@xmath15 xmcd ,
the agreement is adequate if we scale by only -1 ; i.e. , the spectral shape and amplitude are in reasonable agreement , but the sign is incorrect .
a similar sign change applies to the l@xmath31 edge , though in this case the measured amplitude is a factor of three smaller and there are further features at lower energy ; these will be discussed below .
the calculation sets the spin axis coincident with that about which the helicity is defined , but in contrast the experimental convention is to set the applied field antiparallel to the x - ray propagation direction .
@xcite the two definitions coincide when the spin moments are aligned directly by the field , so the sign difference here immediately identifies the spin direction as lying along the field ; the spin magnetic moment is aligned in opposition to the field .
the argument is not new ; note that the sign of sm xmcd has already been used to investigate the alignment in smal@xmath15 , @xcite and furthermore it is in agreement with data on gdn , where there is no doubt that the spin moment aligns with the field .
@xcite the net magnetic moment in smn lies along the 4@xmath5 orbital magnetic moment . aside from a sign reversal , it is notable that the xmcd signal is larger by nearly a factor of ten at l@xmath15 than at l@xmath31 .
such an unbalanced branching ratio is not uncommon for the light rare earths , and results from a dependence of the 5@xmath16 orbital cloud on the 4@xmath5 - 5@xmath16 exchange ; the majority spin orbital is slightly more localised toward the nucleus than that of the minority spin , and hence it overlaps the @xmath32 orbital more strongly . @xcite that is also reflected in the lsda+@xmath4 simulation of fig .
[ fig : xasandxmcd ] , although the l@xmath15:l@xmath31 ratio is underestimated to be @xmath33 .
furthermore there is a strong electric quadrupole ( 2@xmath34 ) signal extending to some 12 ev lower in energy . in part
the prominence of the quadrupole signal at l@xmath31 follows the weakening of the 2@xmath19 electric dipole signal in the light rare earths .
@xcite the lowest energy 4@xmath5 states within the band structure of smn do not lie so far as 12 ev below the 5@xmath16 conduction band , @xcite but the 4@xmath5 xas / xmcd features are shifted lower in energy because the 4@xmath5 electrons interact more strongly with the core hole .
@xcite it is notable that the quadrupole signal follows the dipole signal closely as the temperature is raised ; the polarisation of the 5@xmath16 shell is as expected determined by intra - atomic 4@xmath5 - 5@xmath16 exchange .
0.49 and ( b ) the l@xmath15 edges at 15 k in a field of 6 t. xmcd data are quoted relative to the amplitude of the first prominent ( white ) peak in xas .
the calculated xmcd spectra have been sign reversed and in the case of the l@xmath31 spectrum scaled by a factor 1/3.,title="fig:",scaledwidth=100.0% ] [ fig : xassml3 ] 0.49 and ( b ) the l@xmath15 edges at 15 k in a field of 6 t. xmcd data are quoted relative to the amplitude of the first prominent ( white ) peak in xas .
the calculated xmcd spectra have been sign reversed and in the case of the l@xmath31 spectrum scaled by a factor 1/3.,title="fig:",scaledwidth=100.0% ] [ fig : xassml2 ] in fig .
[ fig : xmcdvst ] is a comparison of the l@xmath15 and l@xmath31 edge spectra collected in the fm phase at 15 and 25 k , and at 50 k in the pm phase .
the spectral shape and sign is unchanged across that temperature range , implying that the net moment lies along the 4@xmath5 orbital and antiparallel to the 4@xmath5 spin moment in the pm as well as the fm phase .
the data show a reduction by a factor of 2 between 15 and 25 k , in agreement with the 27 k t@xmath35 in smn .
the signal dies away further in the pm phase at 50 k. the hysteresis at 15 k ( fig .
[ fig : mhsm ] ) confirms the magnetisation measurements , @xcite with a coercive field of @xmath362 t signalling the weak coupling of the small net moment to an external field . 0.49 and ( b ) the l@xmath15 edge xmcd spectra in the ferromagnetic ( 15 , 25 k ) and paramagnetic ( 50 k ) phases .
the magnetic field was 6 t.,title="fig:",scaledwidth=100.0% ] [ fig : xmcdsml3vst ] 0.49 and ( b ) the l@xmath15 edge xmcd spectra in the ferromagnetic ( 15 , 25 k ) and paramagnetic ( 50 k ) phases .
the magnetic field was 6 t.,title="fig:",scaledwidth=100.0% ] [ fig : xmcdsml2vst ] edge xmcd amplitude at 15 k showing wide hysteresis loop.,scaledwidth=80.0% ] to confirm that the sign of the xmcd spectra above signal that the sm@xmath2 spin lies antiparallel to the field , we have investigated also very thin smn layers embedded in gdn , an epitaxial superlattice of 12@xmath37(1.5 nm smn/10 nm gdn ) .
the 7 @xmath8 spin on gd@xmath2 ensures that its alignment is along the applied field .
the inter - ion exchange that precipitates fm behaviour can then be expected to align the sm and gd spin moments , thus reversing the sign of the sm xmcd signatures .
indeed that is exactly what is seen in fig .
[ fig : xmcdsl ] , showing the sm l@xmath31 signal with the sign reversed relative to homogeneous smn .
the dipole ( 2@xmath19 ) amplitude is also a factor of @xmath38 larger in the superlattice ; clearly the 5@xmath16 spin is strongly aligned by inter - ion exchange across the smn / gdn interface .
the quadrupole ( 2@xmath34 ) signal appears marginally stronger , but by only 30 % .
a comparison with the xmcd at the sm l@xmath15 edge is prevented in this case by masking from magnetic exafs ( extended x - ray absorption fine structure ) above the gd l@xmath31 edge .
edge xmcd in 1.5 nm smn layers embedded in gdn at 15 k.,scaledwidth=80.0% ]
we have reported xmcd in smn , focusing on an investigation of the sm 5@xmath16 spin alignment at the l@xmath3 edges .
the data are interpreted with the aid of an lsda+@xmath4 calculation of the l@xmath3 edge xmcd . except for the sign the calculated and measured xmcd spectra are in excellent agreement at the l@xmath15 , but they differ by a factor of three at the l@xmath31 edge , related to the unusual branching ratio for these transitions .
there is also a prominent quadrupolar feature at the l@xmath31 edge , with a visibility that is enhanced by the relatively weak dipolar spectrum at that edge .
smn is known to possess a strikingly small fm moment of about 0.03 @xmath8 per sm ion ; the 4@xmath5 spin and orbital magnetic moments cancel to within less than 1 % of their individual magnitudes .
the sign of the measured xmcd spectra are opposite to that calculated , identifying that the residual magnetic moment in smn is directed anti - parallel to the 4f spin magnetic moment .
the material is in that respect similar to smal@xmath15 , and it is likely that the moment could be reduced to zero by substituting about 0.5 % gd in the cation site .
temperature - dependent xmcd is in excellent agreement with the 27 k t@xmath9 that has been determined by magnetisation studies .
the field dependence shows the same large coercive field determined by magnetisation studies , and as is expected under the very weak coupling of an applied field with the vanishingly small residual moment .
we acknowledge financial support from the nz frst ( grant no .
vicx0808 ) and the marsden fund ( grant no . 08-vuw-030 ) .
the macdiarmid institute is supported by the new zealand centres of research excellence fund .
thanks the alexander - von - humboldt foundation for support through a fellowship .
32ifxundefined [ 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 10.1103/physrevb.75.045114 [ * * , ( ) ] link:\doibase 10.1103/physrevb.69.045115 [ * * , ( ) ] link:\doibase 10.1103/physrevb.72.014427 [ * * , ( ) ] link:\doibase 10.1103/physrevb.73.235335 [ * * , ( ) ] link:\doibase 10.1103/physrevb.83.174404 [ * * , ( ) ] @noop _ _ , vol .
( , ) pp . @noop _ _ , vol .
( , ) pp .
link:\doibase 10.1016/0304 - 8853(79)90135 - 5 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.174406 [ * * , ( ) ] link:\doibase 10.1103/physrevb.59.11445 [ * * , ( ) ] link:\doibase 10.1016/j.ssc.2011.05.004 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.87.127202 [ * * , ( ) ] link:\doibase 10.1103/physrevb.70.134418 [ * * , ( ) ] http://stacks.iop.org/0953-8984/20/i=26/a=265001 [ * * , ( ) ] link:\doibase 10.1103/physrevb.82.174421 [ * * , ( ) ] link:\doibase 10.1063/1.3581049 [ * * , ( ) ] @noop * * , ( ) , link:\doibase 10.1103/physrevb.76.245120 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.235120 [ * * , ( ) ] link:\doibase 10.1103/physrevb.12.3060 [ * * , ( ) ] link:\doibase 10.1016/0304 - 8853(95)01083 - 1 [ * * , ( ) ] link:\doibase 10.1103/physrevb.45.13244 [ * * , ( ) ] link:\doibase 10.1103/physrevb.49.16223 [ * * , ( ) ] @noop _ _ , vol .
( , ) link:\doibase 10.1103/physrev.46.843 [ * * , ( ) ] link:\doibase 10.1063/1.2234809 [ * * , ( ) ] link:\doibase 10.1103/physrevb.44.943 [ * * , ( ) ] link:\doibase 10.1103/physrevb.67.155103 [ * * , ( ) ] link:\doibase 10.1016/j.physrep.2006.05.002 [ * * , ( ) ] link:\doibase 10.1103/physrevb.73.214430 [ * * , ( ) ] link:\doibase 10.1143/jpsj.66.337 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.78.1162 [ * * , ( ) ] | smn is ferromagnetic below 27 k , and its net magnetic moment of 0.03 bohr magnetons per formula unit is one of the smallest magnetisations found in any ferromagnetic material .
the near - zero moment is a result of the nearly equal and opposing spin and orbital moments in the @xmath0h@xmath1 ground state of the sm@xmath2 ion , which leads finally to a nearly complete cancellation for an ion in the smn ferromagnetic state . here
we explore the spin alignment in this compound with x - ray magnetic circular dichroism at the sm l@xmath3 edges .
the spectral shapes are in qualitative agreement with computed spectra based on an lsda+@xmath4 ( local spin density approximation with hubbard - u corrections ) band structure , though there remain differences in detail which we associate with the anomalous branching ratio in rare - earth l edges .
the sign of the spectra determine that in a magnetic field the sm 4@xmath5 spin moment aligns antiparallel to the field ; the very small residual moment in ferromagnetic smn aligns with the 4@xmath5 orbital moment and antiparallel to the spin moment .
further measurements on very thin ( 1.5 nm ) smn layers embedded in gdn show the opposite alignment due to a strong gd - sm exchange , suggesting that the smn moment might be further reduced by about 0.5 % gd substitution . |
the author would like to thank t. suzuki and t. maki for some useful comments .
he would also like to thank a. sugamoto for reading this manuscript .
99 e. fischbach et al . , phys .
* 56 * ( 1986 ) 3 . for a review ,
see : e. fischbach et ai .
( n.y . ) * 182 * ( 1988 ) 1 ; e. g. adelberger et al .
, phys . rev .
* d42 * ( 1990 ) 3267 .
c. m. will , phys . rep .
* 113 * ( 1984 ) 345 .
m. b. green , j. h. schwarz and e. witten , _ superstring theory _ , ( cambridge university press , cambridge , 1986 ) .
d. gross and e. witten , nucl .
b277 * ( 1986 ) 1 .
see for example , e. w. kolb and m. s. turner , _ the early universe _ , ( addison - wesley pub .
, new york , 1990 ) .
e. w. kolb , m. j. perry and t. p. walker .
phys . rev .
* d33 * ( 1986 ) 869 ; j. rich , d. lloyd owen and m. spiro , phys .
* 151 * ( 1987 ) 239 , and references there in .
d. la and p. j. steinhardt , phys .
* 62 * ( 1989 ) 376 ; phys . lett .
* b220 * ( 1989 ) 375 .
a. guth , phys .
* d23 * ( 1981 ) 347 . for a review ,
see : a. d. linde , rep .
prog . phys .
* 47 * ( 1984 ) 925 ; r. h. brandenberger , rev . mod . phys .
* 57 * ( 1985 ) 1 ; k. a. olive , phys . rep .
* 190 * ( 1990 ) 307 , and references there in . c. brans and r. h. dicke , phys
* 124 * ( 1961 ) 925 .
p. j. steinhardt and f. s. accetta , phys .
* 64 * ( 1990 ) 2740 .
t. damour , g. w. gibbons and c. gundlach .
* 64 * ( 1990 ) l23 , and references there in .
r. holman , e. w. kolb and y. wang , phys .
* 65 * ( 1990 ) 17 .
m. j. geller and j. p. huchra , science * 246 * ( 1989 ) 897 .
k. maeda , phys . rev .
* d39 * ( 1989 ) 3159 .
a. vilenkin , phys . rep .
* 121 * ( 1985 ) 263 .
e. witten , phys .
* d30 * ( 1984 ) 272 , and references there in .
t. w. b. kibble , j. phys .
* a9 * ( 1976 ) 1387 .
s. coleman , phys . rev .
* d15 * ( 1977 ) 2929 .
b. zeldovich , i. yu . kobzarev and l. b. okun , sov . phys . jetp .
* 40 * ( 1975 ) 1 .
p. sikivie , phys .
* 48 * ( 1982 ) 1156 .
see ref .
[ 5 ] , p. 218 .
m. morikawa , ap .
j. * 362 * ( 1990 ) l37 , c. t. hill , p. j. steinhardt and m. s. turner , phys .
lett * b252 * ( 1990 ) 343 ; p. j. steinhardt , presented at the iupap conf . on primordial nucleosynthesis and evolutlon of the early universe ( tokyo , japan , sept . 4 - 8 , 1990 ) , preprint upr-0440 t . t. j. broadhurst , r. s. ellis , d. c. koo and a. s. szalay , nature * 343 * ( 1990 ) 726 . m. fukugita , f. takahara , k. yamashita and y. yoshii , ap . j. * 361 * ( 1990 ) l1 .
d. saez and v. j. ballester , phys . rev .
* d42 * ( 1990 ) 3321 . | we investigate cosmological consequences of an extended gravity model which belongs to the same class studied by accetta and steinhardt in an extended inflationary scenario .
but we do not worry about inflation in our model ; instead , we focus on a topological object formed during cosmological phase transitions .
although domain walls appear during first - order phase transitions such as qcd transition , they decay at the end of the phase transition .
therefore the `` domain wall problem '' does not exist in the suitable range of pameters and , on the contrary , the `` fragments '' of walls may become seeds of dark matter . a possible connection to `` oscillating universe '' model offered by morikawa et al .
is also discussed . in recent years
, many ideas on theory of gravitation have been produced by lots of authors . from the viewpoint of experimental physics ,
the existence of the `` fifth force '' has been examined in these days @xcite .
the tests for gravity theory have been managed in various styles , by means of newly - developed instruments and methods @xcite . from the theoretical viewpoint , many authors have studied possible modifications and corrections to einstein s gravity theory , in consideration of an attempt to unify gravitation and quantum mechanics . according to superstring theory @xcite which has attracted particle - physicists attention for several years
, einstein gravity takes some corrections at very high energy @xcite . of course
, we can not discuss gravitation theory apart from evolution of the universe @xcite .
various modifications of gravity theory have been tried in order to solve cosmological problems .
observational and theoretical constraints on the amount of time - variation of newton constant have been investigated and the upper limit is given @xcite .
for planning precise tests of gravity , it is meaningful to study possible consequences of models which admit spatial and temporal variation of the effective newton `` constant '' .
we can express results of experimental check for models of gravitation by putting observational constraints on a few parameters .
a recent idea of `` extended inflation '' scenario @xcite includes an interesting modification of einstein gravity . in the `` extended inflation '' scenario ,
difficulties in `` old '' inflation @xcite may be avoided by introducing a variable newton `` constant '' into the theory . in some models of this type
, theories of scalar fields coupled non - minimally to gravity are considered , and they are very similar to brans - dicke s graviy theory @xcite .
many variants are also considered in the literature @xcite . as a variant of extended inflation model , `` hyperextended inflation ''
@xcite model is known . in this model , self - interaction of the scalar field which is coupled non - minimally to gravity ,
is postulated and a crucial difference from brans - dicke theory is that the scalar field has an equilibrium point of motion . owing to this property , one can take a wide range of values or parameters keeping away from conflicts with observations . in the present article
, we consider a model of this type which permits variations of an effective newton constant in space and time .
this model is a general , and probably simplest model , which can describe the variation of the strength of gravity .
we do not consider inflation in this paper , however .
the reason is that there are too many problems left unsolved in inflationary models to examine here ; they are for example , the problem of the origin of the very large structure of galaxies recently discovered @xcite , etc . throughout this paper , we choose so - called natural units , i.e. , we set @xmath0 .
our starting point is to consider the followmg lagrangian @xmath1 in this equation , @xmath2 is a neutral scalar field coupled non - minimally to gravity and @xmath3 is a function of @xmath2 and gives the manner of non - linear coupling to gravity ; @xmath4 is a constant which has the same dimensions and order of magnitude as the newton constant , i.e. , @xmath5 ; @xmath6 stands for an effective cosmological term which comes from the contribution of the other matter fields .
this term arises , for example , durng a phase transition of a matter field . a weyl transformation of the metric @xmath7 converts the lagrangian to lagrangian including conventional einstein - hilbert term @xcite @xmath8 in the following , we investigate the property of the model based on this lagrangian ( [ ( 2 ) ] ) ; whenever we can describe a phenomenon both in terms of the lagrangian ( [ ( 1 ) ] ) and that of ( [ ( 2 ) ] ) .
if the effective cosmological term @xmath6 is present , nonlinear self coupling of the scalar field @xmath2 appears in the lagrangian ( [ ( 2 ) ] ) . on account of the self - interaction , topological
object in which energy of the scalar field concentrate can form .
such objects take the form of `` domain walls '' @xcite .
we study the properties of the object and the role in cosmology .
.,width=226 ] first we have to specify the functional form of @xmath3 .
we do not expect drastic changes in the effective strength of gravity in the course of evolution of the universe .
accordingly , it is suitable to choose a periodic continuous function , as a trigonometric function ( fig . 1 ) .
the difference between the maximum and minimum of @xmath3 is denoted by @xmath9 , for later use .
first of all , we study domain walls ; thus @xmath3 has only to take the following `` piecewise - linear '' shape ( fig .
2 ) : @xmath10\phi_0\right ) \,,\qquad\quad \mbox{if } \left[\frac{\phi}{\phi_0}\right ] \mbox { is an even integer}\,,\nonumber \\
f(\phi)&=&1-a\left(\phi-\left[\frac{\phi}{\phi_0}\right]\phi_0-\phi_0 \right ) \,,\quad \mbox{if } \left[\frac{\phi}{\phi_0}\right ] \mbox { is an odd integer}\,,\end{aligned}\ ] ] where @xmath11 $ ] denotes an integer part of @xmath12 .
this @xmath3 involves two parameters , @xmath13 and @xmath14 . in this approximation
, @xmath15 equals to @xmath16 . in general , @xmath3 is characterized by @xmath17 and @xmath18 .
.,width=226 ] suppose that some matter field undergoes first - order phase transition . during the transition , an effective cosmologlcal term @xmath6 arises .
then the neutral scalar field @xmath2 can form a configuration of `` kink '' in one direction , say , in the @xmath19-axis . in three dimensions
, this kink seems as an infinitely - spread wall of high energy density , parallel to @xmath12-@xmath20 plane @xcite .
the ( domain ) wall of this type appears in general theories with spontaneously broken discrete symetry @xcite . in our model
we assume symmetry with respect to @xmath21 .
we want to know qualitative features of the object .
we also wish to avoid a complicated situation where background curvature can not be ignored . to this end , let us suppose that a dimensionless combination @xmath22 is small in comparison with unity .
then the following static configuration can be obtained as an approximate solution ( fig .
3 ) @xmath23 where @xmath24 , and it should be read as @xmath25 . in the presence of @xmath26.,width=226 ] this kink solution is lying in the @xmath19-direction ; thus this configuration is understood as infinitely stretching domain wall in the @xmath12-@xmath20 plane .
an effective cosmological term entirely vanishes when the phase transition of the matter field is over .
then , the scalar field loses its self - interaction term and only has the non - minimal kinetic term .
the domain wall generated during the phase transition is no longer stable when the phase transition finishes .
now then , how does the wall decay ?
we assume plane - symmetric walls here again .
there being no potential term , we have only to solve substantially a wave equation with respect to an initial condition
. we take the static solution ( [ ( 4 ) ] ) as an initial condition at the time @xmath27 .
furthermore we assume that the time - derivative of @xmath2 is zero everywhere , at @xmath27 .
if the assumption is taken , the effective lagrangian ( [ ( 2 ) ] ) for @xmath2 reduces to a dominant term ( now @xmath28 ) @xmath29 therefore the wave equation we must solve is approximated as @xmath30 using the kink ( [ ( 4 ) ] ) as an initial configuration , we can solve the equation as @xmath31\,,\quad ( t>0,\ , z>0 ) \,,\ ] ] and @xmath32 , where @xmath33 and @xmath34 ( see fig . 4 ) . at @xmath35.,width=226
] we find that the walls separated from each other soon become to move at nearly light speed .
two essential characteristics of the behaviour of the wall after @xmath6 disappears are : ( i ) the energy stored at the wall can not dissipate suddenly ; ( ii ) the bulk of energy can not stay still , according to the equation of motion .
now , we consider the significance or the domain wall in our model in cosmology .
the only first - order phase transition in the universe ever known is the ( confinement ) transition of quantum chromodynamics ( qcd ) @xcite .
although the order of the phase transition is yet debatable , we presume the first - order transition . the temperature of the phase transition is considered as @xmath36 .
consecquently , the energy density of the false vacuum , that is , the effective cosmological term @xmath6 in our context , is the order of @xmath37 .
when the temperature of the universe decreases below @xmath38 , the false vacuum energy dominates over the thermal energy of radiations and particles . at the same time the scalar field @xmath2 takes the form of the wall due to non - linear self - coupling , since the thermal fluctuations of the field become sufficiently small .
the domain walls are initially generated by the kibble mechanism @xcite .
we expect at least one wall in each horizon - size volume .
we must check the consistency of calling the object as `` wall '' .
if the thickness of the `` wall '' is wider than the horizon scale of the universe at that time , we can no longer can it as a `` wall '' .
moreover , the process of formation of the object becomes closely dependent on the growth of the fluctuation of the field .
the ratio of the thickess of the `` wall '' to the hoizon length is @xmath39 we find there is no contribution as long as the following inequalities are satisfied @xmath40 which we obtain by combining the cohd1tiort adopted when we led the approxmate solution of the `` wall '' .
note that the conditions are independent of the value of @xmath6 . at the same time
, we can safely ignore the effect of the expansion of the universe .
the mass density of the wall @xmath41 is given by @xmath42 along with progress of the phase transition , bubbles of the true vacuum are born and expand @xcite . the phase transition is over when all the spatial regions come to belong to the true vacuum as a result of the percolation of the bubbles .
in the true vacuum where the effective cosmological constant is nearly zero , the domain wall made of the configuration of @xmath2 decays . therefore the wall begins to break inside the bubbles .
consequently , the evolution of fragments of walls depends on the manner of bubble nucleations .
even if the domain walls are initially plane - symmetric , they take random shape at the end of the phase transition because of randomness in nucleation of bubbles .
if the wall formed during qcd phase transition remains to be static after the transition , the existence of wall conflicts with cosmological observation @xcite as in the case of axionic domain walls @xcite .
the wall in our model , however , completely splits off after phase transition and the remnants move away by almost light speed .
thus we can avoid the above - mentioned problem ; it can be said that the universe is fiiled with the relativistic gas of walls @xcite .
let us estimate the present mass density of the relics of the wall . at the present time ,
the density @xmath43 is given as @xmath44 where @xmath45 and @xmath46 . here
the parameter @xmath47 @xmath48 is required for the reasm that we take an account of the effect of transformation of the energy of fragments into that of other massive particle .
the constraint that the energy density @xmath49 is lower than the critical density of the present universe @xmath50 yields @xmath51 this constraint can easily be satisfied for a wide range of parameters . the contribution of @xmath49 to the total energy density might explain the missing mass in the universe , which has not yet been identified
. it is dependent on a hidden sector coupled to @xmath2 whether the relics of the wall can provide the dark matter which gathers other matters to generate galaxies .
this is because the fragments of wall remain to move fast if there is no interaction between @xmath2 and other field . in a class of extended inflationary scenario , hidden sectors which violate the weak equivalence principle
have been considered @xcite . in our context
, we have no compelling ground to determine the property of the hidden sector . since the stage of our investigation in this paper is primitive , we will examine the property of other fields coupled to @xmath2 and gravity in a future work including numerical simulations of walls .
in the rest of the present paper , we investigate the connection to the oscillating universe model @xcite . in the extended model we treated ,
@xmath52 was not a smooth function , for a simple analysis .
of course , this is not an essential thing . from now on ,
we suppose @xmath3 takes the form of the trigonometric function ( fig . 1 ) , since we want to consider a coherent oscillation of the field @xmath2 .
our analysis below need only a rough sketch of @xmath3 ; dimensional analysis can lead to a sufficient estimation .
suppose that there exists a non - vanishing ( positive ) cosmological constant in the present universe .
then an effective interaction of the scalar @xmath2 appears , provided that the value of the cosmological term is of the same order of the thermal energy or more . in this time
, we consider a spatially homogeneous , oscillating field .
using an expansion of the effective interaction with respect to small @xmath2 , we obtain the period of oscillation as @xmath53 where @xmath54 is the present cosmological constant .
recent astronomical observation revealed very large , periodic structure with a period of about @xmath55 million lightyears in the universe @xcite .
morikwa and serveral authors have claimed that the structure is an illusion induced from an error in estimation of the distance scale by means of measurement of galaxies redshift @xcite .
an assumption that the expansion rate of the scale factor is oscillating leads to such an artificial pattern .
morikawa took a model in which the oscillation of a coherent scalar field bring about the oscillation of `` hubble constant '' through modified einstein equations . here
we try to identify such a scalar field with @xmath2 in our model . the period ( [ ( 13 ) ] ) becomes @xmath55 million years when we take @xmath54 as @xmath56 according to another group of authors in ref .
@xcite , a finite cosmologlcal constant of order of @xmath57 is demanded by observations of distant galaxies .
if we take such an allowed value for @xmath54 in ( [ ( 14 ) ] ) , @xmath58 if we further require a sufficient amplitude of oscillating hubble parameter which can explain the observation , @xmath15 must be larger than @xmath59 , according to steinhardt in ref .
@xcite .
if we choose @xmath60 , we must set @xmath61 according to ( [ ( 15 ) ] ) .
these values for parameters give rise to no contradiction with the previous constraints ( [ ( 9 ) ] ) .
the constraint ( [ ( 12 ) ] ) is also cleared @xmath62 .
this value of @xmath47 means that a finite fraction of energy density produced by the domain walls remains in the form of relativistic gas .
it can never be justified that our naive model describes the true nature of the gravity theory .
the time variation of the gravitational constant and other constant is strictly constrained by observation ; even if the constraints are able to be satisfied , fine - tuning is necessary , for instance , in the phase of oscillations .
the important point we want to claim is that a general modification of gravity could explain different problems in cosmology in the same time , such as dark matter problem and problems in large scale structure of the universe .
accordingly , it is interesting and significant to investigate the detail of the consequence of the model and to make more complicated models which have more relevance to cosmological puzzles and no conflicts with observation .
we also have to investigate the possibility that the first - order transition should be identified to unknown transition , for instance , in relation to sub - quark dynamics .
we wish to study the effect of cosmic expansion and gravity on the production of walls as well as the numerical simulation including the bubble nucleation and the wall intersection . after completing this work ,
the present author became aware of the paper @xcite .
they also considered the wall which disappears after a certain phase transition .
moreover they claimed that the wall can provide the seed for generating galaxies .
while their motivation in cosmology is very interesting , their analysis depends crucially on a peculiar model .
we think that we must consider models which are related with many cosmological aspects and can be checked by observations . |
nonequilibrium growth models leading naturally to self - organized fractal structures , such as diffusion limited aggregation ( dla ) @xcite , are of continuing interest due to their relevance for many important physical processes including dielectric breakdown @xcite , electrochemical deposition @xcite , and laplacian flow @xcite .
a powerful method , namely iterated stochastic conformal mapping @xcite , has been already successfully applied to generate and analyze dla @xcite and laplacian @xcite growth patterns in two dimensions .
this has provided an alternative way to address many of the important open questions related to pattern formation in dla in two dimensions , one of these being the existence of minimal fields for growth at the boundary of the growing cluster . in previous work
@xcite we studied the properties induced by a _ fixed _ , material dependent critical field @xmath10 for growth , and showed that in the presence of such a threshold all clusters ultimately become pinned and , in addition , this simple constraint has remarkable consequences for the resulting patterns the rich , branched structure of dla is replaced by a much lower dimensional shape consisting of a few surviving branches . in this paper
we address a similar , but significantly more important question because of its relationship to the multifractal spectrum of dla , that of what happens when there exists a critical field for growth on the boundary of the cluster , field decaying like @xmath0 as the cluster increases in size .
we shall call this model the `` @xmath11-model '' . as we will show below , as @xmath11 decreases from @xmath3 toward a critical value @xmath12 ( which corresponds to the most singular possible behavior for the laplacian field , that occuring at the tip of a line )
, there is a continuous transition from dla toward lower dimensional shapes for which the multifractal spectrum is necessarily different from that of dla .
we study this transition in terms of the fractal dimension @xmath5 of the emerging patterns , and we derive analytic expressions for the behavior of @xmath5 in the range @xmath8 where the dla universality class breaks down and @xmath9 close to the pinning transition .
witten and sander @xcite have shown that the growth probability at any point @xmath13 on the boundary of a dla cluster of length @xmath14 is given by the harmonic measure @xmath15 , where @xmath16 obeys laplace s equation @xmath17 subject to the boundary conditions @xmath18 on the ( evolving ) boundary of the cluster and @xmath19 as @xmath20 ( corresponding to a uniform flux of particles far away from the cluster ) .
the model we study is a variant of the two - dimensional dla growth model described above in which growth is disallowed at points on the cluster boundary where the probability for growth is smaller than a critical value @xmath0 , where @xmath1 ( the exact meaning will be defined later ) is the radius of the @xmath21 particle cluster , i.e. , @xmath22|\nabla v ( s')|ds ' } } , & |\nabla v| > r_n^{-\gamma } , \cr 0 , & |\nabla v| < r_n^{-\gamma},\cr \end{cases}\ ] ] where @xmath23 is the multifractal exponent at point @xmath24 on the cluster boundary , @xmath14 is the length of the boundary , and the step function @xmath25 ensures that only those regions of the cluster boundary obeying @xmath26 contribute to the normalization integral .
estimates of this integral will be very important in our analysis of the fractal dimension of the growing `` @xmath11-cluster '' .
since we need to calculate the harmonic measure on a freely evolving interface , this is handled by using conformal mapping techniques @xcite .
the method was presented in great detail in refs .
@xcite , and thus here we will just briefly review the main results . the basic idea is to follow the evolution of the conformal mapping @xmath27 of the exterior of the unit circle in a mathematical @xmath28plane onto the complement of the cluster of @xmath29 particles in the physical @xmath30plane rather than directly the evolution of the cluster s boundary .
the initial condition is chosen to be @xmath31 .
the process of adding a new `` particle '' of constant shape and linear scale @xmath32 to the cluster of @xmath33 `` particles '' at a position @xmath13 which is chosen randomly according to the harmonic measure is described via a function @xmath34 , where @xmath35 -1 \right \ } ^a \nonumber \\ & & \phi_{\lambda,\theta } ( \omega ) = e^{i \theta } \phi_{\lambda,0}(e^{-i \theta } \omega ) \
, , \label{eq - f}\end{aligned}\ ] ] which conformally maps the unit circle to the unit circle with a bump of linear size @xmath36 localized at the angular position @xmath37 @xcite .
the shape of the bump depends on the parameter @xmath38 .
following the analysis in @xcite , we have used @xmath39 througout this paper , as we believe the large scale asymptotic properties will not be affected by the microscopic shape of the added bump . the conformal map for an @xmath29-particle cluster @xmath40 can be built by adding one `` particle '' to an @xmath33-particle cluster @xmath41 , resulting in the recursive dynamics @xmath42 which can be solved in terms of iterations of the elementary bump map @xmath43 , @xmath44 in eqs .
[ recursive ] and [ comp ] the angle @xmath45 $ ] at step @xmath29 is randomly chosen since the harmonic measure on the real cluster translates to a uniform measure on the unit circle in the mathematical plane , @xmath46 and @xmath47 is chosen in order to ensure that the size of the bump in the physical @xmath30 plane is @xmath32 . since @xmath32 is a natural length scale in the problem ,
in can be scaled out by measuring all the lengths in terms of it .
we re - emphasize here that although the composition eq . [ comp ] appears at first sight to be a standard iteration of stochastic maps , this is not so because the order of iterations is inverted the last point of the trajectory is the inner argument in this iteration . as a result
the transition from @xmath41 to @xmath40 is achieved by composing the @xmath29 former maps eq .
[ comp ] starting from a different seed . finally , identifying @xcite the radius @xmath48 of the growing pattern with the coefficient @xmath49 in the laurent expansion of @xmath50 ,
@xmath51 the constraint to grow only at values of @xmath37 which obey eq .
[ growth ] translates into @xmath52'(e^{i\theta})| } > \left({f_1^{(n-1)}}\right)^{-\gamma}~.\ ] ] this constraint is implemented as follows . at step @xmath29
, @xmath53 is choosen from a uniform distribution in @xmath54 $ ] , independent of previous history .
if it obeys the constraint given by eq .
[ constraint ] accept this value of @xmath53 , otherwise repeat until the constraint is obeyed .
for the model defined in sec . ii one would expect the resultant patterns to have fractal shapes which depend on @xmath11 , and in order to characterize these shapes we will focus on the scaling behavior of the first laurent coefficient @xmath55 .
following the arguments in ref .
@xcite , for a given value @xmath11 one expects a scaling law of the form @xmath56 where @xmath5 is the effective fractal dimension of the resulting cluster .
we have simulated the model defined in sec .
ii for a number of values @xmath11 in the range @xmath57 and we have calculated @xmath55 as an average over @xmath58 clusters ( for @xmath59 ) , and , respectively , over @xmath60 clusters ( for @xmath61 ) of size @xmath62 . in fig .
[ fig1 ] we show typical clusters of size @xmath62 for @xmath63 , @xmath64 and @xmath65 , respectively .
typical clusters ( size @xmath66 ) grown with ( a ) @xmath63 , ( b ) @xmath67 , and ( c ) @xmath68 , respectively . ] typical clusters ( size @xmath66 ) grown with ( a ) @xmath63 , ( b ) @xmath67 , and ( c ) @xmath68 , respectively . ] typical clusters ( size @xmath66 ) grown with ( a ) @xmath63 , ( b ) @xmath67 , and ( c ) @xmath68 , respectively . ]
it can be easily seen that the rich , branched structure of the cluster at @xmath63 changes , as @xmath11 decreases toward @xmath68 , into a much lower - dimensional shape with only a few branches surviving .
this can be intuitively understood by considering the effect the pinning probability @xmath69 has on the multifractal spectrum of the cluster . from multifractal
scaling @xcite we know that the interface of a fully developed dla cluster consists of sets of @xmath70 sites with growing probabilities @xmath71 , and we will assume that such a structure is also valid for the clusters grown with pinning probability @xmath69 . as growth proceeds ,
the lowest probability sites ( large @xmath72 ) , normally deep in the fjords , will be pinned first , with the hot tips surviving longest , leading to `` pruning '' of the branches where the tips have a singularity @xmath73 .
as anticipated , for all the values @xmath74 that we have tested the coefficient @xmath55 has a clear power - law dependence on the size @xmath29 , as shown in fig .
[ fig2](a ) . assuming the exponent @xmath75 to be related to a fractal dimension as given by eq .
[ f1_scaling ] , the dependency @xmath5 ( shown in fig .
[ fig2](b ) ) is obtained from a power - law fit to the data .
( a ) @xmath55 as a function of @xmath29 for clusters grown with @xmath76 , respectively ( log - log plot ) . also shown ( dashed lines )
are the limit cases of a dla cluster , @xmath77 , where @xmath78 , and of a line cluster , @xmath79 , respectively .
( b ) the effective fractal dimension @xmath5 obtained from @xmath80 as a function of @xmath11 ( circles ) .
the dotted line is just a guide to the eye .
the solid line is the theoretical prediction ( eq .
[ d_gam ] ) with @xmath81 from ref .
@xcite . ]
( a ) @xmath55 as a function of @xmath29 for clusters grown with @xmath76 , respectively ( log - log plot ) . also shown ( dashed lines )
are the limit cases of a dla cluster , @xmath77 , where @xmath78 , and of a line cluster , @xmath79 , respectively .
( b ) the effective fractal dimension @xmath5 obtained from @xmath80 as a function of @xmath11 ( circles ) .
the dotted line is just a guide to the eye .
the solid line is the theoretical prediction ( eq .
[ d_gam ] ) with @xmath81 from ref .
@xcite . ]
it can be seen that at values of @xmath82 the behavior is close to that of a dla cluster , i.e. , @xmath83 , while for @xmath84 the behavior of the radius @xmath55 tends to @xmath29 , i.e. , @xmath85 , thus the behavior of a growing line . in order to understand
these results theoretically let us begin with a very simple argument based on the assumption that the clusters have a multifractal spectrum in the sense described above , in that the interface consists of sets of @xmath86 sites with growth probabilities @xmath87 . since the constraint will cut - off growth at regions in the cluster with exponents in the range @xmath88 of the multifractal spectrum , we can write down the following equation for the rate of growth of the cluster in the presence of the barrier in terms of the rate of growth of a dla ( no barrier for growth ) cluster @xmath89 where @xmath90 . the enhancement of growth comes from eq .
[ growth ] together with the estimate @xmath91 now , we know that the harmonic measure is concentrated at @xmath92 , and thus for @xmath3 the integral is dominated by the value of the integrand at @xmath92 , while for @xmath8 it is dominated by the value at @xmath11 , and thus @xmath93 where @xmath94 is some constant independent of @xmath21 .
[ near_one ] therefore implies @xmath95 for the case where @xmath8 , i.e. , close to the breakdown in the dla universality class , we may assume that the multifractal spectrum of the cluster is only weakly perturbed from its value in dla and therefore @xmath96 , or @xmath97 this prediction may be tested using the recently computed @xmath98 spectrum @xcite .
the results are shown in fig .
[ fig2](b ) ( solid line ) , and it can be seen that for @xmath99 the theoretical predictions are indeed close to the measured values @xmath5 . the discrepancies at smaller values of @xmath11 can be attributed to the fact that the multifractal spectrum of the cluster is not exactly the one of dla , and when @xmath100 , as we will now discuss , it actually may be expected to deviate significantly from that of a dla . for @xmath9 , the change in the multifractal spectrum from that for dla is significant .
for example , it is known @xcite that for the dla spectrum @xmath7 , while we see that growth continues significantly below this value , with @xmath101 being the asymptotic limit .
our simulations show strong evidence for this limit , as can be seen in fig .
[ fig3](a ) : the average number of attempts for growing clusters of sizes @xmath102 and @xmath66 ( scaled by the actual size of the cluster ) exibits a steep increase as @xmath103 . ( a ) average number of attempts for growing clusters of sizes @xmath102 ( dashed line ) and @xmath66 ( solid line ) , scaled by the actual size of the cluster , respectively , as function of @xmath11 ( b ) the multifractal spectrum @xmath104 from eq . [ tildef ] ( symbols ) in comparison with the multifractal spectrum of dla ( solid line ) from ref . @xcite . ] ( a ) average number of attempts for growing clusters of sizes @xmath102 ( dashed line ) and @xmath66 ( solid line ) , scaled by the actual size of the cluster , respectively , as function of @xmath11 ( b ) the multifractal spectrum @xmath104 from eq .
[ tildef ] ( symbols ) in comparison with the multifractal spectrum of dla ( solid line ) from ref .
@xcite . ]
we shall assume that for these highly pruned `` @xmath11-clusters '' there is a well defined limiting form for the multifractal spectrum , @xmath105 , defined for @xmath106 and obeying @xmath107 , which is a monotonically increasing function of @xmath11 .
defining @xmath108 , we can use a reasoning similar to that used by turkevich and sher @xcite in their estimate of the fractal dimension of dla to estimate @xmath104 .
the idea is that for @xmath9 only the `` hottest '' tips contribute to growth , and thus one can write @xmath109 where the last relation follows from the fact that the integral is dominated by the value of the integrand at @xmath11 . thus , we obtain @xmath110 since from simulations we know the values @xmath5 , eq . [ tildef ]
( which is valid for @xmath111 ) allows the calculation of the upper multifractal exponent @xmath104 . as shown in fig .
[ fig3](b ) , the pinning threshold @xmath112 leads to a shifting of the multifractal spectrum for @xmath113 to the left , i.e. , @xmath114 ( more hot tips , and larger fields at those hot tips , due to pruning ) .
these results can be intuitively understood as a flow of singularities away from @xmath11 ( which acts as an unstable fixed point of the dynamics ) . for any particular value @xmath115 ,
what happens while the cluster evolves is that screening is reduced compared to dla and therefore there is a flow of singularities @xmath116 with @xmath117 .
in addition , new singularities with @xmath118 , can be created .
thus , we would expect that the number of singularities @xmath119 or @xmath120 on the other hand for @xmath121 the singularity flow @xmath116 can only act toward an increase @xmath122 since such points can never grow and thus can only either keep their original singularity or get a higher value of @xmath72 during growth .
using the stochastic conformal mapping techniques we have studied the patterns emerging from laplacian growth with a power - law decaying threshold for growth @xmath0 , @xmath84 .
we have shown that due to the enhancement of growth at the hot tips as @xmath11 decreases the growth evolves from patterns in the dla universality class for @xmath3 to clusters with a lower fractal dimension @xmath5 for @xmath4 due to the enhancement of growth at the hot tips .
we have presented evidence that @xmath6 , corresponding to the singularities at the tip of a purely one - dimensional ( line ) growth pattern , is the lower limit for growth , with all clusters becoming ultimately pinned for @xmath123 . by using multifractal analysis , we have proposed analytic expressions for @xmath5 for both @xmath8 near the breakdown of the dla universality class and near the pinning transition @xmath9 .
finally , we have shown that in the small @xmath11 range the multifractal spectrum of the resulting cluster is significantly changed from that of a dla .
we have suggested that this change may be due to a flow of singularities with @xmath11 acting as an unstable fixed point of the dynamics , but further work will be necessary to fully elucidate this point .
this work has been supported by the petroleum research fund .
we would like to acknowledge the very stimulating discussions , leading to the idea for this paper with tom witten , benny davidovitch and thomas seligman and at the cic , unam in cuernavaca , morelos , mexico .
one of us ( mnp ) would also like to thank the physics department at emory university for hospitality during the period when some of this work was done .
the authors would like to thank prof .
i. procaccia and the authors of ref .
@xcite for making available the data for the multifractal spectrum of dla . | using stochastic conformal mapping techniques we study the patterns emerging from laplacian growth with a power - law decaying threshold for growth @xmath0 ( where @xmath1 is the radius of the @xmath2 particle cluster ) .
for @xmath3 the growth pattern is in the same universality class as diffusion limited aggregation ( dla ) growth , while for @xmath4 the resulting patterns have a lower fractal dimension @xmath5 than a dla cluster due to the enhancement of growth at the hot tips of the developing pattern .
our results indicate that a pinning transition occurs at @xmath6 , significantly smaller than might be expected from the lower bound @xmath7 of multifractal spectrum of dla .
this limiting case shows that the most singular tips in the pruned cluster now correspond to those expected for a purely one - dimensional line . using multifractal analysis ,
analytic expressions are established for @xmath5 both close to the breakdown of dla universality class , i.e. , @xmath8 , and close to the pinning transition , i.e. , @xmath9 . |
big bang nucleosynthesis ( bbn ) , as well known , is one of the solid pillars of the standard cosmological model .
the theory predicts that relevant abundances of light elements , namely @xmath0 , @xmath1 , @xmath2 and @xmath3 , have been produced during the first minutes of the evolution of the universe .
the predictions span about 9 orders of magnitude and are in reasonable agreement with observations .
theoretical calculations are well defined and very precise .
the largest uncertainty arises from the values of cross - sections of the relevant nuclear reactions .
theoretical accuracy is at the level of 0.2% for @xmath2 , 5% for @xmath0 and @xmath1 and 15% for @xmath3 .
however , comparison of theoretical results with observational data is not straightforward because the data are subject to poorly known evolutionary effects and systematic errors .
still , even with these uncertainties , bbn permits to constraint important cosmological parameters and to eliminate many modifications of the standard model , thus allowing to derive restrictions on the properties of elementary particles and , in particular , of neutrinos . in this paper , we briefly review the physics of bbn . in sect .
[ bbn - phys ] we introduce the essential parameters and inputs . in sect .
[ obs ] we summarize the present situation of observational data . in sect .
[ param ] we discuss the determination of cosmological parameters .
the last section is dedicated to bbn bounds on non - standard neutrinos and , specifically , to bbn and neutrino oscillations .
to understand primordial nucleosynthesis , we must follow in detail the histories of nucleons in the early universe .
this is usually done by using numerical codes ( among which the fortran code by wagoner @xcite , updated by kawano @xcite has become a standard tool ) .
however , the main features of big - bang nucleosynthesis can be obtained by simple analytic arguments described in this section ( see @xcite for details ) .
primordial nucleosynthesis occurs at temperatures @xmath4 mev , which are small with respect to nucleon masses . at these temperatures ,
the number of nucleons is simply equal to the initial baryon asymmetry of the universe .
it is useful to describe this quantity in terms of the _ present _ baryon to photon ratio : @xmath5 the parameter @xmath6 is simply related to the baryon density of the universe , being @xmath7 .
the neutron to proton ratio is controlled by the weak processes @xmath8 which interconvert neutron and proton .
when the temperature @xmath9 of the universe is about 1 mev , these reactions are fast enough to maintain neutron and proton in chemical equilibrium . the neutron abundance is thus given by : @xmath10 where @xmath11 mev is the neutron - proton mass difference and @xmath12 is the dimensionless chemical potential of electron neutrinos ( in standard bbn @xmath13 is assumed to be negligible ) .
when the temperature @xmath9 drops below @xmath14 mev , the neutron - proton inter - conversion rate , @xmath15 , becomes smaller than the expansion rate the universe , @xmath16 , where @xmath17 counts the total number of relativistic degrees of freedom of the early universe .
the deviation of @xmath17 from the standard value , @xmath18 , is usually described in terms of an _ equivalent _ number of massless neutrinos @xmath19 according to : @xmath20 for temperatures @xmath21 , chemical equilibrium can no longer be maintained .
the neutron abundance @xmath22 evolves only due to neutron decay , according to @xmath23 , where @xmath24 is the neutron lifetime .
one should note that the `` freeze - out '' temperature @xmath25 scales as @xmath26 and thus is sensitive to the particle content of the early universe .
the larger is @xmath17 ( or equivalently @xmath27 ) , the earlier is the freeze - out of the neutron abundance , at an higher value , and hence , the larger is the @xmath2 abundance produced in bbn .
when the temperature of the universe is equal to @xmath28 mev neutrons and protons start to react each other to build up light nuclei .
the exact value of @xmath29 depends on the baryon to photon ratio @xmath6 .
only two body reaction are indeed important in bbn , such as @xmath30 , @xmath31 , @xmath32 , etc .
( see @xcite ) .
deuterium must be produced in appreciable quantity before the other reactions can proceed at all .
however , due to the large number of photons per baryon , photodissociation of deuterium is not suppressed until the temperature decreases well below the deuterium binding energy @xmath33mev . following @xcite , one can see that the temperature @xmath29 below which deuterium production is favoured scales as @xmath34 .
once deuterium is formed , nucleosynthesis begins and light nuclei are produced rapidly .
essentially all available nucleons are quickly bound into @xmath2 , which is the most tightly bound light nuclear species .
in addition to @xmath2 , substantial amounts of @xmath0 , @xmath1 and @xmath2 are produced .
no heavy elements ( @xmath35 ) are produced , due both to the fact that coulomb - barrier suppression is very significant and to the absence of stable isotopes with @xmath36 and @xmath37 . in fig.1
we show the light element abundances produced during bbn , as calculated by using the kawano code @xcite , for @xmath6 between @xmath38 and @xmath39 .
the calculation of @xmath40he abundance includes small corrections due to radiative processes at zero and finite temperature , non - equilibrium neutrino heating during @xmath41 annihilation , and finite nucleon mass effects @xcite .
theoretical predictions are affected by uncertainties at the level of 0.2% for @xmath2 , 5% for @xmath0 and @xmath1 and 15% for @xmath3 .
these uncertainties are due to uncertainties in the weak rates ( which are `` normalized '' to the measured neutron lifetime @xmath42 , see @xcite for details ) and in the values of the relevant nuclear reaction rates .
they have been estimated by montecarlo or semi - analytical methods @xcite .
recently the nuclear data have been re - analyzed , leading to improved precision in the abundance predictions @xcite .
the abundances of light elements synthesized in the big bang have been subsequently modified through chemical evolution of the astrophysical environments where they are measured .
the observational strategy is to identify sites which have undergone as little chemical processing as possible and rely on empirical methods to infer the primordial abundances .
for example , measurements of deuterium are made in in quasar absorption line systems ( qas ) at high redshift ; if there is a `` ceiling '' to the abundance in different qas then it can be assumed to be the primordial value .
the @xmath43 abundance is measured h ii regions in blue compact galaxies ( bcgs ) which have undergone very little star formation .
its primordial value is inferred either by using the associated nitrogen or oxygen abundance to track the stellar production of helium , or by simply observing the most metal - poor objects .
closer to home , the observed uniform abundance of @xmath3 in the hottest and most metal poor pop ii stars in our galaxy is believed to reflect its primordial value .
( we do not consider @xmath1 whose post - bbn evolution is more complex . )
as observational methods have become more sophisticated , the situation has become more , instead of less , complex .
relevant discrepancies , of a systematic nature , have emerged between different observers . in the following ,
we give a brief summary of the present situation for deuterium and helium-4 ( looking from outside by a non - expert ) .
we refer to @xcite for a more complete and up - to - date discussion .
_ deuterium _ + post - bbn evolution of deuterium ( d ) is simple .
deuterium is burnt in stars producing @xmath1 .
any deuterium measurement provides thus a lower limit for the primordial d abundance and an upper limit for baryon density of the universe . in recent years
, measurements of deuterium have been made in quasar absorption line systems ( qas ) at high redshift .
these systems are presumably not contaminated by stellar processes and thus the observed deuterium should be close to the primordial one .
since deuterium isotope shift corresponds to velocity of only @xmath44 , clearly qas with simple velocity structure are needed for reliable determinations .
moreover , ionization corrections , possible `` interlooper '' etc . further increase systematic uncertainties . in tab .
[ deuterium ] we give the results of recent deuterium determinations in qas @xcite .
an estimate of primordial deuterium abundance can be obtained from the weighted mean of data in tab .
[ deuterium ] .
it should be noted , however , that the dispersion among the different determinations is not consistent with errors in the single measurements ( see @xcite for detailed discussion ) .
we will use , in the following , the value @xmath45 given in @xcite , which is the weighted mean of the log d / h values given by @xcite .
the quoted error is the 1@xmath46 error in the mean , given by the standard deviation of the five log d / h values divided by @xmath47 .
this error is used instead of the usual error in the weighted mean , in order to take into account the `` anomalous '' dispersion of deuterium data .
.deuterium abundance in quasar absorption line systems at high red - shift ( see @xcite for details ) .
[ cols= " < , < , < , < , < , < , < " , ] @xmath48 this system was first analyzed by @xcite with the result @xmath49 .
the quoted value is from @xcite .
[ deuterium ] _ helium-4 _ + as a result of stellar processing , @xmath50 is produced , increasing its abundance above the primordial value , together with `` metals '' , such as nitrogen , oxygen and other elements heavier than @xmath50 , which are not produced in bbn .
the observed @xmath50 abundance provides thus an upper bound to the primordial one , @xmath51 .
helium observations are done in h ii regions in blue compact galaxies ( bcgs ) which have undergone very little star formation ( at present @xmath52 h ii regions have been studied for their helium content ) . in order to infer the primordial value @xmath51 ,
one extrapolates to zero metallicity ( @xmath53 ) the observed relation between helium ( @xmath54 ) and metals ( @xmath55 ) .
this is usually done using by nitrogen ( @xmath56 ) or oxygen ( @xmath57 ) as metallicity tracers .
alternatively , one can simply average helium abundances in most metal poor objects .
the present situation is that independent determinations of @xmath51 have a statistical errors at the level of @xmath58 but differ among each others by about @xmath59 . in particular , by using independent data sets , olive and steigman @xcite and olive et al .
@xcite have obtained @xmath60 , while izotov et al .
@xcite and izotov and thuan @xcite have found @xmath61 .
the discrepancy between different determinations is possibly related to different description of the complex physical processes acting in h ii regions .
several sources of systematic uncertainties may , in fact , affect the helium determination at a level comparable or larger than the reported statistical errors , like e.g. the evaluation of the ionization correction factor ( which is related to how much neutral helium is in the object under scrutiny ) , of the temperature correction factor , of underlying stellar absorption , etc .
( see @xcite for a review ) . as discussed in the next section ,
it is extremely important to have a better determination of the @xmath2 primordial abundances and a reliable estimate of the total ( statistical + systematic ) associated error . for our estimates
, we will use the central values for @xmath51 reported above ( @xmath62 and @xmath63 ) , and the error estimate @xmath64 , which is obtained e.g. in @xcite from the dispersion of the various @xmath51 determinations .
the deuterium abundance @xmath45 can be used to determine the baryon density of the universe . as discussed in @xcite
, the quoted value corresponds to @xmath65 ( in standard bbn ) or , equivalently , to @xmath66 .
the error budget is dominated by the observational uncertainties which are about a factor 3 larger than uncertainties in theoretical prediction .
the obtained value for @xmath67 has to be compared with independent determination of the baryon density of the universe . in particular with the result @xmath68 given in @xcite which is obtained from a combined fit to the cosmic microwave background ( cmb ) and large scale structure ( lss ) data .
the agreement of these two independent determinations is extremely important because they rely on completely different physical phenomena ( which occurred at different time during the evolution of the universe ) .
we note that cmb ( and lss ) are presently more accurate than bbn in determining the baryon density of the universe .
the value of @xmath6 deduced from deuterium can be used , in standard bbn , to predict the abundance of the other elements and to compare with observations .
following @xcite , one obtains @xmath69 , @xmath70 and @xmath71 .
it is evident that there is tension between the quoted values and the observational results .
the `` predicted '' abundance for @xmath2 is higher than the `` high '' helium value of izotov et al .
moreover , the `` predicted '' @xmath3 abundance is a factor 2 - 3 larger with respect to the present observational results @xcite .
the origin of these differences has to be clarified .
they could be due to systematic errors in the measurements or to evolutionary effects ( e.g. @xmath3 depletion ) or they could be a real indication for non - standard effects in bbn . in particular , the present d and @xmath2 data seems to favour an equivalent number of neutrino families @xmath72 . in order to understand the present situation , it is useful to combine the deuterium value @xmath73 , with the `` low '' helium abundance , @xmath74 , or with the `` high '' helium abundance , @xmath75 .
the error @xmath76 is the `` estimated '' systematic error in @xmath2 measurements ( see above ) . if we fit these data in the plane @xmath77 following @xcite , we obtain the bound @xmath78 ( @xmath79 ) in the first case , and @xmath80 ( @xmath79 ) in the second . in both cases ,
the central values are below three , even if the errors are large enough to allow for the standard value @xmath81 .
the described results clearly indicate that a large number of effective neutrinos is disfavoured .
one can conclude , in principle , that an upper bound on the number of _ extra _ neutrinos , @xmath82 , is @xmath83 .
it is clear , however , that the situation is quite delicate .
the error @xmath84 is completely dominated by systematic error in @xmath2 measurements . for this reason , we believe that , at present stage , a more safe upper bound on the number of extra neutrinos is @xmath85 . hopefully in the near future we will be able to derive a stronger limit .
other physical parameters which can be bounded by bbn are the chemical potentials of different neutrino species , @xmath86 where @xmath87 . the possible role of neutrino degeneracy in bbn was noted already in @xcite and then discussed in a number of papers . a non vanishing chemical potential for @xmath88 , @xmath89 or @xmath90 would increase the neutrino contribution to the energy density and can be described as an increase in @xmath27 . an additional ( and dominant ) effect exists for electron neutrinos which directly participate to n - p interconversion reactions .
a non vanishing @xmath91 would shift the equilibrium between neutrons and protons , see eq .
( [ np - eq ] ) , with large effects on light elements production . several analysis have been made of the bbn limits on neutrino chemical potentials . a recent analysis which include also cmb data
@xcite concludes : @xmath92 where @xmath93 are the dimensionless chemical potentials . for further implications and for a discussion of the case in which both @xmath27 and @xmath13 are free to vary see @xcite .
effects of neutrino oscillations on bbn are much different if only active neutrinos are mixed , if only one active and one sterile neutrino are mixed or if we consider the more `` complete '' case of mixing between three active and one sterile neutrino
. if neutrinos are in thermal equilibrium with vanishing chemical potentials , mixing between active neutrinos does not introduce any deviation from standard bbn results .
the situation is more interesting if neutrinos are degenerate . in particular , it was shown recently @xcite that , for the mixing parameters which explain the solar neutrino problem @xcite ( @xmath94 and @xmath95 ) and the atmospheric neutrino anomaly @xcite ( @xmath96 and @xmath97 ) , asymmetries in the muonic and/or tauonic neutrino sectors would produce , through oscillations , an asymmetry into the electronic neutrino sector .
this means that , in presence of oscillations , the restrictive bounds on the chemical potential of electron neutrinos applies to all neutrino flavours .
it is thus possible to obtain the restrictive bound : @xmath98 valid for any flavour @xcite .
there are three possible effects on bbn created by mixing between active and sterile neutrinos .
first is the production of additional neutrino species in the primeval plasma , leading to @xmath99 .
the second effect is a depletion of the number density of electronic neutrinos which results in a higher neutron freezing temperature .
both these effects lead to a larger neutron - to - proton ratio and to more abundant production of primordial deuterium and helium-4 ( for the details see e.g. review @xcite ) .
if mixing between active neutrinos is absent the second effect would manifest itself only in the case of @xmath100-mixing , if we neglect relatively weak depopulation of @xmath101 through the annihilation @xmath102 .
the third effect is a generation of large lepton asymmetry due to oscillations between active and sterile species @xcite .
however , this effect takes place only for very weak mixing , much smaller than the experimental bound and is not discussed in this paper .
the problem of active - sterile neutrino oscillation is quite complex and has been discussed in many papers starting from 1990 ( a large list of references can be found in ref .
the problem was recently re - considered in @xcite both analytically and by solution of the complete system of integro - differential kinetic equations .
earlier derived bounds have been re - analyzed and significantly different results have been found in the resonance case . the results of @xcite are shown in fig .
[ f2 ] . the effect on bbn
is expressed in term of variation of the effective number of neutrinos @xmath103 .
the upper panels are for the case of @xmath104 ( or @xmath105 ) mixing , while the lower panels refer to the case of @xmath106 mixing .
the obtained results clearly depend on the sign of the mass differences . for positive
is positive if sterile neutrino is heavier than active neutrino , in the limit @xmath107 . ]
mass differences , @xmath108 ( left panels ) , sterile neutrino production occurs through non - resonant transitions . for @xmath109 ( right panels )
, one has instead resonant active - sterile transitions which result in much stronger bounds on the neutrino oscillation parameters .
the solid lines in fig.[f2 ] correspond to numerical results , while the red dotted lines corresponds to analytic approximate results .
it is evident that an observational bound on extra neutrinos much better than unity , say @xmath110 , could give very restrictive limits on active - sterile neutrino mixing .
unfortunately , the present observational bound @xmath111 is not accurate enough to put relevant constraints .
it is practically established now that all active neutrinos are mixed with parameters given by the large mixing angle solution to solar neutrino problem ( @xmath94 and @xmath95 ) and by atmospheric neutrino data ( @xmath96 and @xmath97 ) .
existence of fast transitions between @xmath101 , @xmath112 , and @xmath113 may noticeably change bbn bounds on mixing with sterile neutrinos , expecially for small values of mass difference .
in particular , due to oscillations between active neutrinos , a deficit of @xmath112 or @xmath113 would be efficiently transformed into a deficit of @xmath101 , leading to stronger bounds on active - sterile mixing .
the effects of mixing between active neutrinos on the bbn bounds on a possible active - sterile admixture has been investigated in detail in @xcite .
comparison of bbn theoretical results with observational data is not straightforward because the data are subject to poorly known evolutionary effects and systematic errors .
still , even with these uncertainties , bbn permits to constraint important cosmological parameters , like e.g. the baryon density @xmath114 , the effective number neutrino families @xmath27 , the neutrino degeneracy parameters @xmath115 etc .
the present bound on the number of extra neutrinos species @xmath116 is about unity and is not accurate enough to put relevant constraints on active - sterile neutrino mixing .
if this limit could be reduced in the next future , say to @xmath110 , very restrictive limits on active - sterile admixture could be obtained .
b. d. fields and s. sarkar , phys .
d * 66 * , 010001 ( 2002 ) .
g. steigman , arxiv : astro - ph/0307244 .
smith , l.h .
kawano , r.a .
malaney : astrophys .
j. suppl .
* 85 * , 219 ( 1993 ) .
r. e. lopez and m. s. turner : phys .
d * 59 * , 103502 ( 1999 ) + r. e. lopez , m. s. turner and g. gyuk , phys .
d * 56 * , 3191 ( 1997 ) s. esposito et al .
* 81 * , 59 ( 2000 ) g. fiorentini , e. lisi , s. sarkar and f. l. villante : phys
. rev .
d * 58 * , 063506 ( 1998 ) r. h. cyburt , b. d. fields and k. a. olive : new astron .
* 6 * , 215 ( 1996 ) k. m. nollett and s. burles : phys .
d * 61 * , 123505 ( 2000 ) + s. burles , k. m. nollett and m. s. turner : astrophys .
j. * 552 * , l1 ( 2001 ) a. cuoco et al .
: arxiv : astro - ph/0307213 .
s. burles and d. tytler : astrophys . j. * 507 * , 732 ( 1998 ) .
+ d. tytler and s. burles : astro - ph/9606110 s. burles and d. tytler : astrophys .
j. * 499 * , 699 ( 1998 ) + d. tytler , x. m. fan and s. burles : nature * 381 * , 207 ( 1996 ) j. m. omeara et al : astrophys .
j. * 552 * , 718 ( 2001 ) m. pettini and d. v. bowen : astrophys .
j. * 560 * , 41 ( 2001 ) d. kirkman et al .
: arxiv : astro - ph/0302006 .
levshakov et al .
: astrophys .
j. * 565 * , 696 ( 2002 ) k. a. olive and g. steigman : astrophys .
j. suppl .
* 97 * , 49 ( 1995 ) k. a. olive , e. skillman and g. steigman , astrophys .
j. * 483 * , 788 ( 1997 ) y. i. izotov , t. x. thuan and v. a. lipovetsky : astrophys .
j. suppl .
* 108 * , 1 ( 1997 ) .
y. i. izotov and t. x. thuan : astrophys . j. * 500 * , 188 ( 1998 ) .
v. luridiana : arxiv : astro - ph/0209177 .
d. n. spergel et al .
astrophys .
j. suppl .
* 148 * , 175 ( 2003 ) s. g. ryan et al : astrophys .
j. * 530 * , l57 ( 2000 ) + p. bonifacio and p. molaro , mon . not .
roy . astron .
285 * , 847 ( 1997 ) e. lisi , s. sarkar and f. l. villante : phys . rev .
d * 59 * , 123520 ( 1999 ) p. c. de holanda and a. y. smirnov , jcap * 0302 * , 001 ( 2003 ) super - kamiokande coll .
, y. fukuda et al . :
lett * 81 * , 1562 ( 1998 ) + macro coll . , m.ambrosio et a. : phys .
b * 434 * , 451 ( 1998 ) s. h. hansen ey al .
d * 65 * , 023511 ( 2002 ) v. barger et al .
b * 569 * , 123 ( 2003 ) a. d. dolgov et al : nucl .
b * 632 * , 363 ( 2002 ) r. foot , m. j. thomson and r. r. volkas : phys .
d * 53 * , 5349 ( 1996 ) + r. foot and r. r. volkas : phys .
d * 55 * , 5147 ( 1997 ) + n. f. bell , r. foot and r. r. volkas , phys .
d * 58 * , 105010 ( 1998 ) + p. di bari and r. foot , phys .
d * 65 * , 045003 ( 2002 ) | we present a brief review of big bang nucleosynthesis ( bbn ) .
we discuss theoretical and observational uncertainties in deuterium and helium-4 primordial abundances and their implications for the determination of important cosmological parameters .
we present , moreover , some recent results on active - sterile neutrino oscillations in the early universe and on their effects on bbn . |
left - handedness is the most interesting nature of the weak charged current . the charged weak boson @xmath5 couples only to the left - handed fermions in the standard model ( sm ) , and this structure of the sm results in purely left - handed charged currents .
another notable aspect of the weak charged current is the cabibbo - kobayashi - maskawa ( ckm ) mixing in the quark sector @xcite . because of the ckm mixing the strength of the quark charged current varies depending on the flavors involved in a transition , while the left - handedness remains universal apart from tiny radiative corrections .
new physics ( np ) beyond the sm may violate the universality of the charged current .
in particular , np effects are expected to be significant for ckm suppressed transitions if the supposed np has a different flavor structure from the sm . among several
ckm suppressed channels , the @xmath3 transition is most likely to be affected by np in this sense .
moreover , we can examine it with the competent data of @xmath1 factory experiments .
it was pointed out that a mixture of right - handed current ( rhc ) in the @xmath3 charged current explained the discrepancy among magnitudes of the relevant ckm matrix element , @xmath0 , determined by various @xmath1 decay modes @xcite .
recently , belle collaboration updated the data on the pure tauonic @xmath1 decay , @xmath6 @xcite , and the new result seems more consistent with the sm than the previous one @xcite . in this work ,
we revisit the possibility of rhc in the @xmath3 transition taking the new belle data into account . in sec .
[ sec : vub ] , we introduce the @xmath3 rhc and explain how it affects the @xmath0 determination in leptonic and semileptonic @xmath1 decays . then , we show that the present experimental data suggest a sizable @xmath2-violating ( cpv ) rhc in this transition . according to this observation
, we study possible cpv signals in hadronic @xmath1 decays , @xmath7 , @xmath8 and @xmath9 , in the presence of the @xmath3 rhc in sec .
we find that new cpv signals absent in the sm arise in direct @xmath2 asymmetries and the measurement of angles of the unitarity triangle .
comparing these possible signals of the rhc with the relevant experimental data , we show that the @xmath3 cpv rhc is a viable np scenario .
we also compare the @xmath3 rhc induced by squark mixings in the minimal supersymmetric standard model ( mssm ) to the experimental constraint obtained in sec .
[ sec : vub ] . our conclusion is given in sec .
[ conclusion ] .
a set of right - handed quark charged currents appears in the sm once we introduce higher dimensional operators .
the gauge - invariant effective lagrangian containing the lowest dimensional operator responsible for the @xmath3 transition is expressed by @xmath10 where @xmath11 is the higgs doublet , @xmath12 , @xmath13 represents the energy scale of np , and @xmath14 is a dimensionless constant that depends on the details of np .
this lagrangian leads to the desired right - handed interaction as a result of the electroweak symmetry breaking .
combined with the ordinary left - handed interaction , the @xmath3 transition is described by the following modified charged current lagrangian : @xmath15 where @xmath16 is the su(2 ) gauge coupling , @xmath17 , and @xmath18 denotes the left - handed ckm matrix element .
the effective right - handed ckm matrix element @xmath19 is given by @xmath20 where @xmath21 denotes the vacuum expectation value of @xmath11 .
the known magnitude of the @xmath3 charged current corresponds to @xmath22 and the right - handed component induced by np at the tev scale is potentially comparable . in the rest of this section , we treat @xmath18 and @xmath19 as complex parameters and determine their values using experimental data of leptonic and semileptonic @xmath1 meson decay modes along with the unitarity of the left - handed ckm matrix @xmath23 assuming that no sm interactions besides the rhc in eq . are affected by np .
we emphasize that the nonvanishing relative phase between @xmath18 and @xmath19 is a new cpv degree of freedom .
the pure tauonic @xmath1 decay is a theoretically clean mode to determine @xmath0 provided that the @xmath1 meson decay constant @xmath25 is known accurately enough .
the decay rate is written as @xmath26 where @xmath27 is the effective ckm matrix element that is determined by experiments assuming the sm .
since the axial - vector current contributes to this decay mode , one finds @xmath28 in the presence of the rhc in eq . .
the present world average of the branching ratio including the updated belle data is provided by the heavy flavor averaging group ( hfag ) @xcite as @xmath29 . using this value and @xmath30 @xcite
, we obtain @xmath31 the axial - vector part in eq
. does not contribute to the process @xmath32 owing to the parity invariance of the strong interaction .
the contribution of the vector part leads to the following differential decay rate , @xmath33 where @xmath34 represents the momentum transfer , @xmath35 , and @xmath36 , which reduces to the ordinary ckm matrix element in the absence of the right - handed interaction .
the hadronic form factor @xmath37 plays a crucial role in the determination of @xmath27 .
we employ the one evaluated using the light - cone sum rule ( lcsr ) @xcite and its explicit form is given in appendix [ ap : hffp ] for completeness .
the mass of the charged lepton is neglected in eq . .
evaluating the branching ratio with eqs . and , and comparing it to the experimental result @xcite , @xmath38 , we obtain @xmath39 where the error includes 13% theoretical uncertainty in the magnitude of @xmath37 @xcite as well as the experimental errors .
both the vector and axial vector currents give rise to the decay @xmath41 , where @xmath42 denotes a vector meson and we consider two modes with @xmath43 .
the differential decay rate is given as @xmath44 where @xmath45 represents the three momentum of the vector meson in the rest frame of the @xmath1 meson , @xmath46 for @xmath47 considered below , and the helicity amplitudes @xmath48 are expressed in terms of hadronic form factors @xmath49 and @xmath50 : @xmath51 we use lcsr form factors @xcite and they are explicitly given in appendix [ ap : hffv ] .
we can experimentally determine the magnitude of the @xmath3 charged current , represented by @xmath27 , from the branching ratio of @xmath41 .
integrating the differential rate in eq . over the phase space
, we find @xmath52 where @xmath53 for @xmath54 . in our numerical analysis ,
we employ the following experimental results @xcite : @xmath55 and @xmath56 , where the errors include theoretical uncertainties in the form factors .
since significant part of theoretical uncertainties is involved in these errors , neither the theoretical uncertainty in @xmath57 nor the effect of experimental cut in the phase space integration performed to obtain eq . is taken into account in the present analysis .
we note that a precise study of decay distribution in these modes with copious data at the superkekb / belle ii experiment may provide an improved method to probe the rhc @xcite .
both the left- and right- handed currents contribute to the inclusive @xmath3 semileptonic process .
however the interference between them is strongly suppressed because of the small mass of the up quark .
thus , the decay rate is proportional to @xmath59 . in our numerical analysis ,
we use the following result of ggou method @xcite given by hfag @xcite : @xmath60 where the statistic and systematic errors are linearly added in order to take the sizable method dependence into account .
since the above five processes of direct @xmath0 measurement are described by two independent quantities , @xmath61 and @xmath62 , they are insufficient to determine the absolute values of @xmath18 and @xmath19 , and their relative phase . in order to extract more information on @xmath18 and @xmath19 , we utilize the unitarity of the ckm matrix @xmath23 assuming the validity of the sm except for the @xmath19 term in eq . as stated previously .
the unitarity of @xmath23 is conveniently represented by the unitarity triangle in fig .
[ fig : rhoeta ] , where wolfenstein parameterization @xcite is introduced .
measuring @xmath63 by the @xmath1@xmath64 mixings and @xmath65 ( or @xmath66 ) with @xmath2 violation in @xmath67 decays , together with results of kaon and @xmath68 semileptonic decays that give @xmath69 and @xmath70 in wolfenstein parameterization , we can indirectly determine the magnitude and phase of @xmath18 .
unitarity triangle ( solid lines ) .
experimental constrains from the @xmath1@xmath64 mixings and @xmath2 violation in @xmath67 decays are also indicated by the red arc and the green narrow sectors respectively . ]
the mass difference in the @xmath71 meson system due to the @xmath71@xmath72 mixing , denoted by @xmath73 , is dominated by the top quark loop in the sm and proportional to @xmath74 ; and similarly in the @xmath75 system , @xmath76 .
theoretical uncertainties of the relevant hadronic matrix elements are reduced by taking the ratio of @xmath73 and @xmath77 owing to the su(3 ) flavor symmetry : @xmath78 where @xmath79 @xcite represents the su(3 ) breaking effect .
thus , we determine @xmath63 from the present experimental data , @xmath80 and @xmath81 @xcite .
the result is shown as the red arc in fig .
[ fig : rhoeta ] . the time - dependent @xmath2 asymmetries in @xmath82 processes such as @xmath83
give @xmath84 with small theoretical uncertainty in the sm .
this argument does not change in the presence of the @xmath3 rhc .
the combined experimental data @xmath85 @xcite gives @xmath65 with a four - fold ambiguity .
it turns out that only the solution favored in the sm , as depicted in fig .
[ fig : rhoeta ] , is consistent with the rhc .
consequently , one of the apices of the unitarity triangle @xmath86 is uniquely determined ( with errors ) and @xmath87 is evaluated as @xmath88 where @xmath89 and @xmath90 are used @xcite .
allowed region of @xmath91 ( solid lines ) .
the constraints from @xmath0 and the unitarity triangle are combined .
prediction of mssm is also indicated by dashed lines .
( see sec .
[ mssm ] . ) ] combining the results in eqs . , , , , , and
, we obtain a constraint on @xmath19 .
the numerical result is presented in fig .
[ fig : vr ] , where @xmath92 , @xmath93 and @xmath94 allowed regions are indicated by solid lines .
the best fit is given by @xmath95 with @xmath96 .
we obtain @xmath97 in the sm and thus the scenario of the @xmath3 rhc exhibits a similar consistency as the sm among the above experimental results in @xmath0 determination . as seen in fig .
[ fig : vr ] and eq .
, a large relative phase between @xmath19 and @xmath18 is favored .
we examine its implication for @xmath2 violations in hadronic @xmath1 decays in the next section .
as shown in the previous section , the present experimental data related to the @xmath3 transition allow the right - handed @xmath3 current with a large relative phase to the left - handed counter part .
we examine possible consequences of this new complex phase in @xmath2 violations in hadronic @xmath1 decays . the time - dependent @xmath2 asymmetry in @xmath98
is described by the following formula @xcite , @xmath99 where the direct @xmath2 asymmetry @xmath100 and the mixing - induced @xmath2 asymmetry @xmath101 are expressed as @xmath102 and @xmath103 respectively .
the amplitude ratio @xmath104 is defined by @xmath105 and the ratio of the @xmath1@xmath64 mixing coefficients is given as @xmath106 for the @xmath71 case under consideration here .
the isospin analysis is mandatory to extract the information on the weak phase in this process because of the penguin pollution @xcite .
the decay amplitudes of the isospin doublet @xmath107 are expressed in terms of the isospin amplitudes @xmath108 ( @xmath109 ) : @xmath110 @xmath111 @xmath112 we note a simple triangle relation , @xmath113 .
the @xmath114 decay amplitudes bear similar relations to @xmath115 .
the relative phase between @xmath116 and @xmath117 can be determined with a twofold ambiguity as well as their magnitudes by measuring the branching fractions of three decay modes in eqs . , and ; and likewise for @xmath114 and @xmath118 .
the ratio of @xmath98 amplitudes in eq . is expressed in terms of the isospin amplitudes as @xmath119 where @xmath120 , @xmath121 , and they are obtained from the relevant branching fractions as described above .
the amplitudes of @xmath122 are determined by the tree - level @xmath5 boson exchange since the gluon penguin diagram has the nature of @xmath123 . in the sm ,
the @xmath122 amplitudes are governed by the single weak phase of @xmath18 and thus there is no @xmath2 asymmetry in this channel except the small correction due to the electroweak penguin diagrams . in the presence of the right - handed @xmath3 current in eq . , the amplitudes of @xmath122 consist of the left- and right- handed contributions : @xmath124 and @xmath125 .
thus , the large imaginary part of @xmath91 suggested by the analysis in sec .
[ sec : vub ] implies a possibility of @xmath2 violation in the @xmath122 channel .
we neglect the electroweak penguins in the following analysis because effects of the rhc are expected to be larger than them .
the direct @xmath2 asymmetry in @xmath126 , which vanishes in the sm , is written as @xmath127 where the effect of the rhc in the @xmath122 channel is represented by @xmath128 other cpv observables are also affected by the rhc : @xmath129 @xmath130 and @xmath131 where @xmath132 is one of the angles of the unitarity triangle in fig .
[ fig : rhoeta ] and @xmath133 is the counter part of @xmath100 in @xmath134 .
we note that @xmath133 is determined with the time - integrated decay rate of the tagged @xmath134 process .
the experimental data of these observables and relevant @xmath2 averaged branching fractions are summarized in table [ tab : pipi ] .
the phase @xmath132 is extracted from the unitarity triangle construction indicated in fig .
[ fig : rhoeta ] as @xmath135 .
.[tab : pipi]experimental data in @xmath7 , taken from the compilation by hfag @xcite . [ cols= " < ,
< " , ] these experimental data constrain @xmath136 and the allowed region is presented in fig .
[ fig : rhorho ] , in which @xmath137 and @xmath138 are chosen as axes .
it turns out that the triangle dictated by the isospin relation , @xmath139 , and the charge - conjugated one are squashed .
hence , only a two - fold ambiguity remains in the isospin analysis in @xmath140 in contrast to the eight - fold one in @xmath7 .
this reduction of the number of solutions results in a more stringent restriction on @xmath138 as seen in fig .
[ fig : rhorho ] . in other words ,
the possible discrepancy in the @xmath141 determinations between @xmath140 and the unitarity triangle are constrained more strongly . allowed region of the direct @xmath2 asymmetry and the possible phase discrepancy in @xmath140 is shown in the same manner as fig .
[ fig : pipi ] . the prediction of the cpv rhc is presented as well .
the region between two black dashed ovals is @xmath92 and the region surrounded by the blue dotted one is @xmath93 . ]
we evaluate @xmath136 in the presence of the cpv rhc using the rge and the factorization method as in the case of @xmath7 .
we obtain @xmath142 as @xmath143 where an independent strong phase @xmath144 is introduced .
this calculation is described in appendix [ ap : facrhorho ] .
the predicted region of @xmath136 for the allowed @xmath145 shown in fig .
[ fig : vr ] and arbitrary values of @xmath144 is also depicted in fig .
[ fig : rhorho ] .
one of the two experimentally allowed regions , which is consistent with the sm , is also compatible with the scenario of the cpv rhc . in fig .
[ fig : rhorhopv ] , we present the @xmath146 value of @xmath147 assuming @xmath148 as well as its range predicted for the allowed region of @xmath145 in fig . [
fig : vr ] .
the cpv rhc is consistent with one of the two possible solutions that is also favored in the sm .
one may judge from figs .
[ fig : rhorho ] and [ fig : rhorhopv ] that the cpv rhc is incompatible with the experimental data at the @xmath92 level .
however this is not the case because of the theoretical uncertainty in the factorization .
we consider that an uncertainty of a factor of two is likely . the @xmath146 value of @xmath147 assuming @xmath148 and the prediction of the cpv rhc as in fig .
[ fig : pipipv ] . ]
two quark processes @xmath149 and @xmath150 ( and their charge conjugates ) give rise to @xmath151 decays in the sm and the latter is modified by the @xmath3 rhc .
we denote the relevant decay amplitudes in the following manner : @xmath152 @xmath153 @xmath154 @xmath155 where amplitude ratios @xmath156 are defined to be positive .
this decay mode is employed to extract @xmath157 ( or @xmath158 ) of the unitarity triangle in the sm @xcite , in which the right - handed contribution vanishes and @xmath159 .
we stress that @xmath160 in the sm , but this is not the case in the presence of the cpv rhc in general .
thus a direct @xmath2 asymmetry , @xmath161 is induced in addition to a discrepancy between @xmath162 and @xmath163 . among several methods to extract @xmath164 in the sm , we focus on the most powerful one , that is the dalitz plot method @xcite , in which the neutral @xmath165 meson in @xmath151 is identified with its dalitz decay @xmath166 .
we extend the method to the case of @xmath167 in the following .
amplitudes of the dalitz decay are written as @xmath168 where @xmath169 and @xmath170 .
we neglect small meson - antimeson mixing and @xmath2 violation in the neutral @xmath165 meson system in the present work .
then , the differential decay rate of @xmath171 is represented as @xmath172d\phi\,,\end{aligned}\ ] ] where @xmath173 is a phase - space factor .
an illustration of the binning in the dalitz plot method . ] in the dalitz plot method , the phase space is divided into @xmath174 bins as illustrated in fig .
[ fig : bin ] .
the binning is symmetric with respect to the diagonal line defined by @xmath175 , and the @xmath176-th bin ( @xmath177 ) with @xmath178 and the @xmath179-th bin with @xmath180 form a symmetric pair .
the partial decay rate into the @xmath176-th bin is written as @xmath181\,,\end{aligned}\ ] ] and that into the @xmath179-th bin is @xmath182\,,\end{aligned}\ ] ] where @xmath183 @xmath184/ \sqrt{t_it_{-i}}\,,\ ] ] @xmath185/ \sqrt{t_it_{-i}}\,,\ ] ] and we have used @xmath186 and @xmath187 . the dalitz distribution @xmath188 is given by the flavor - tagged neutral @xmath165 meson decay and thus @xmath189 s are known as well as @xmath190 , which is determined by the flavor specific @xmath165 decay in @xmath151 .
we notice that the number of unknown quantities ( @xmath191 , @xmath192 , @xmath193 , @xmath162 and @xmath194 ) is @xmath195 , that of observables ( @xmath196 ) is @xmath197 , and in principle , all the unknown quantities can be determined provided @xmath198 . in particular , we can obtain the direct @xmath2 asymmetry in eq . and the angle discrepancy @xmath199 with the ( extended ) dalitz plot method .
it is possible to improve the analysis by using @xmath191 s and @xmath192 s independently extracted from data at a charm factory @xcite .
the entangled @xmath200 states produced near the threshold exhibit quantum interference that depends on @xmath191 s and @xmath192 s . in ref .
@xcite , experimental data of belle corresponding to @xmath196 , @xmath189 are shown for the optimized binning @xcite with @xmath201 . the result for @xmath191 s and @xmath192 s by cleo collaboration @xcite is also summarized in ref .
@xcite . using these data
, we obtain a constraint on the direct @xmath2 asymmetry @xmath202 and the phase disagreement @xmath203 ( @xmath204 , see eq . below . )
as presented in fig .
[ fig : dk ] .
although the restriction is rather mild at present , we confirm that the extended dalitz plot method does work and expect a better sensitivity in future .
allowed region of the direct @xmath2 asymmetry and the possible phase discrepancy in @xmath9 as in the same manner in fig .
[ fig : pipi ] .
the prediction of the cpv rhc is also shown .
the region between two black dashed ellipses is @xmath92 prediction as denoted and the whole plane is practically allowed at the @xmath93 level . ] in order for a comparison with the allowed region in fig .
[ fig : dk ] , we evaluate the effect of the cpv rhc on @xmath205 and the charge conjugation mode .
their amplitudes are decomposed into the left- and right- handed contributions : @xmath206 and @xmath207 where @xmath208 is the weak phase of the left(right)-handed current and @xmath209 denote strong phases .
it is convenient to introduce an amplitude ratio as @xmath210 where @xmath211 .
then , it is straightforward to obtain the following relations from eqs . , and : @xmath212 and @xmath213 the rge and the factorization approximation gives @xmath214 .
we evaluate @xmath215 in eq . for the allowed value of @xmath91 shown in fig .
[ fig : vr ] and @xmath163 determined by the unitarity triangle taking @xmath216 as a free parameter .
then , we obtain theoretical prediction on @xmath202 and @xmath203 as presented in fig .
[ fig : dk ] .
we find that the scenario of the cpv rhc is disfavored at the @xmath92 level despite the moderate current experimental constraint though it is not excluded at @xmath93 .
this is due to the enhancement of the rhc contribution in the @xmath217 mode shown in eq . compared to those in the @xmath218 and @xmath219 modes in eqs . and .
this notable sensitivity , though it is derived in the factorization approximation , might play an important role in future experiments in order to probe or exclude the cpv rhc . it has been pointed out that the @xmath3 rhc is induced by radiative corrections in the mssm @xcite . the gluino - squark one - loop diagram with simultaneous insertions of the left - right mixing in the ( 3,3 ) component of the down - type squark mass matrix ( @xmath220 ) and that in the ( 1,3 ) component of the up - type squark mass matrix ( @xmath221 ) gives the dominant contribution and one obtains @xmath222 where the dimensionless mass insertion parameters are defined by @xmath223 and @xmath224 , and
the masses of relevant supersymmetric partners are assumed to be common for simplicity and denoted by @xmath225 . in fig .
[ fig : vr ] , we present @xmath91 evaluated with @xmath18 in eq . and
@xmath19 in eq . for @xmath226 and @xmath227 .
we observe that the mssm contribution is consistent with the current experimental bound from @xmath0 determination and the unitarity triangle within @xmath93 though the best fitted values do not seem to be realized . a future experiment like superkekb / belle
ii may find a signal of supersymmetry through the @xmath3 rhc .
we have studied the scenario of @xmath3 right - handed current .
our analysis combining the present experimental results for direct @xmath0 determination with the unitarity triangle suggests a significant cpv rhc in the @xmath3 transition as presented in fig .
[ fig : vr ] . according to this analysis
, we have examined cpv signals in two - body hadronic @xmath1 decays : @xmath7 , @xmath8 and @xmath9 .
the expected signals in these decay modes are new direct @xmath2 asymmetries , deviations of @xmath141 in @xmath228 and that of @xmath164 in @xmath9 ; they are depicted in figs .
[ fig : pipi ] , [ fig : rhorho ] and [ fig : dk ] as well as the present experimental constraints . although the direct @xmath2 asymmetries in @xmath228 are strongly constrained ,
a sizable deviation of @xmath229 in @xmath141 is not excluded . as for @xmath9 , the effect of rhc is enhanced by qcd radiative correction in the factorization approximation .
hence the rather moderate current experimental bound tightly restricts the cpv rhc .
we have found that the consistency of the suggested cpv rhc with the present @xmath9 data is in between the @xmath92 and @xmath93 levels in the factorization approximation .
the prediction of the mssm is also compared to the allowed region of @xmath91 as shown in fig .
[ fig : vr ] . in conclusion ,
the @xmath3 right - handed current is a new physics scenario that is still consistent with the present experimental data .
the suggested large @xmath2 violation gives rise to the new @xmath2 violating signals in hadronic @xmath1 decays and they may be detected in a future @xmath1 factory experiment .
we thank r. watanabe for discussion in the initial stage of this study .
the work of mt is supported in part by jsps kakenhi grant number 25400257 .
we briefly summarize the hadronic form factors used in our numerical analysis in the main text . the hadronic form factor @xmath37 in eq .
is defined as @xmath232 the result of lcsr is concisely parameterized in the following form of pole dominance @xcite : @xmath233 where @xmath234 , @xmath235 gev , @xmath236 and @xmath237 .
the form factors in eqs . and
are parameterized as @xmath239 @xmath240 and @xmath241 the lcsr gives @xcite @xmath242 , @xmath243 , @xmath244 , @xmath245 , @xmath246 , @xmath247 , @xmath248 , @xmath249 , @xmath250 for @xmath251 , and @xmath252 , @xmath253 , @xmath254 , @xmath255 , @xmath256 , @xmath257 , @xmath258 , @xmath249 , @xmath259 for @xmath260 .
in this appendix , we describe the calculation of @xmath261 amplitudes in the factorization method .
the effective four - fermion hamiltonian that contributes to the @xmath122 channel in @xmath7 is decomposed into the left and right pieces as @xmath262 , and @xmath263+\text{h.c.}\,,\ ] ] where @xmath264 and @xmath265 denotes a renormalization scale .
the four - fermion operators are defined by @xmath266 @xmath267 where @xmath268 and @xmath66 are color indices .
wilson coefficients @xmath269 ( @xmath270 ) are obtained by solving a set of renormalization group equations in the leading order @xcite .
the relevant anomalous dimensions are @xmath271 for @xmath272 and @xmath273 respectively . as a result
, we obtain the wilson coefficients at the bottom quark mass scale ( @xmath274 ) : @xmath275^{-6/23}- \left[\frac{\alpha_s(m_b)}{\alpha_s(m_w)}\right]^{12/23 } \right)\simeq -0.27\,,\ ] ] @xmath276^{-6/23}+ \left[\frac{\alpha_s(m_b)}{\alpha_s(m_w)}\right]^{12/23 } \right\}\simeq 1.12\,,\ ] ] @xmath277^{24/23}- \left[\frac{\alpha_s(m_b)}{\alpha_s(m_w)}\right]^{-3/23 } \right\}\simeq 0.34\,,\ ] ] and @xmath278^{-3/23 } \simeq 0.92\,,\ ] ] where @xmath279 @xcite and @xmath280 @xcite are used .
we neglect the gluon penguin operators since they do not contribute to the @xmath122 final state .
the amplitude ratio @xmath142 is conveniently evaluated by calculating @xmath126 amplitudes , @xmath281 , at @xmath282 : @xmath283\ , , \label{eq : pipifacl}\end{aligned}\ ] ] and @xmath284\ , , \label{eq : pipifacr}\end{aligned}\ ] ] where we have used fierz rearrangement and ignored annihilation terms and nonfactorizable contributions .
the matrix elements of the vector currents in eqs . and are expressed by the form factors in eq . and those of the axial - vector currents are given by the pion decay constant .
the ( pseudo)scalar operator in eq . are related to the corresponding ( axial-)vector operator using the equation of motion of the quark fields and thus its matrix element is also written in terms of the form factors ( the decay constant ) .
interestingly , we do not need to specify the values of the form factors and the decay constant since they disappear in the ratio @xmath285 .
hence we obtain @xmath286 \simeq 1.56\frac{v_{ub}^{r*}}{v_{ub}^{l*}}\,,\end{aligned}\ ] ] where @xmath287 denotes the bottom quark pole mass , @xmath288 represents the average current mass of the up and down quarks , and we employ @xmath289 @xcite and @xmath290 @xcite in our numerical calculation . the factorization method described above
should be understood as a crude approximation that provides an order of magnitude .
we consider that an uncertainty of a factor of two remains even in the ratio in eq . .
furthermore it gives no information on the phase shift by the strong interaction , and thus we introduce a strong phase in eq . by hand .
the effective four - fermion hamiltonian for @xmath7 , shown in eqs . , and , also describes the @xmath122 amplitudes in @xmath140 .
the relevant matrix elements are evaluated almost in the same way as in the case of @xmath7 .
we finally obtain @xmath291 the effective hamiltonian for @xmath292 and its charge conjugation is given by @xmath262 and @xmath293+\text{h.c.}\,,\ ] ] where the four - fermion operators are defined by @xmath294 @xmath295 the renormalization of these operators are the same as those in @xmath7 in the leading order and hence the wilson coefficients are also given by eqs . , , and .
using fierz rearrangement and ignoring annihilation terms and nonfactorizable contributions , we evaluate the amplitude ratio @xmath296 as in the case of @xmath7 . eventually , we obtain @xmath297 where @xmath298 denotes the charm quark pole mass @xcite and we have neglected the up and strange quark masses .
the quark pole masses emerge when we utilize the equations of motion of the quark fields in order to evaluate the contribution of the rhc .
we note that the @xmath299 form factors and the @xmath165 meson decay constant appearing in each amplitude @xmath300 cancel out in the ratio of eq . as in @xmath7 .
99 n. cabibbo , phys . rev
. lett .
* 10 * , 531 ( 1963 ) .
m. kobayashi and t. maskawa , prog .
phys . * 49 * , 652 ( 1973 ) . c. -h . chen and s. -h .
nam , phys .
* b666 * , 462 ( 2008 ) .
a. crivellin , phys .
d * 81 * , 031301(r ) ( 2010 ) .
a. buras , k. gemmler , g. ishidori , nucl .
phys . * b843 * , 107 ( 2011 ) . i. adachi _ et al . _
( belle collaboration ) , phys .
* 110 * , 131801 ( 2013 ) .
arxiv:1208.4678 [ hep - ex ] . a question whether new physics resolves the discrepancies among @xmath0 ( and @xmath301 ) determinations is studied in the following paper : a. crivellin and s. pokorski , arxiv:1407.1320 [ hep - ph ] .
y. amhis _ et al . _
( heavy flavor averaging group ) , arxiv:1207.1158 [ hep - ex ] .
see also online update at ` http://www.slac.stanford.edu/xorg/hfag/ ` .
flavor lattice averaging group , ` http://itpwiki.unibe.ch/flag/ ` , 2013 .
p. ball and r. zwicky , phys .
d * 71 * , 014015 ( 2005 ) .
p. ball and r. zwicky , phys .
d * 71 * , 014029 ( 2005 ) .
a. sibidanov _ et al .
_ ( belle collaboration ) , phys .
d * 88 * , 032005 ( 2013 ) .
bernlochner , z. ligeti and s. turczyk , arxiv:1408.2516 [ hep - ph ] .
p. gambino , p. giordano , g. ossola and n. uraltsev , j. high energy phys . 10 ( 2007 ) 058 . l. wolfenstein , phys . rev
. lett . 51 , 1945 ( 1983 ) .
olive _ et al . _
( particle data group ) , chin .
c , * 38 * , 090001 ( 2014 ) .
see also ` http://pdg.lbl.gov/ ` .
ckmfitter group ( j. charles _ et al .
_ ) , eur . phys .
j. c * 41 * , 1 ( 2005 ) .
for updated results , see ` http://ckmfitter.in2p3.fr ` . for details see , for example , i. i. bigi and a. i. sanda , _ cp violation _
( second edition ) , cambridge university press , cambridge , uk ( 2009 ) .
m. gronau and d. london , phys . rev . lett . *
65 * , 3381 ( 1990 ) .
i. dunietz , h. quinn , a. snyder , w. toki and h. j. lipkin , phys .
d. * 43 * , 2193 ( 1991 ) .
p. vanhoefer _ et al .
_ ( belle collaboration ) , phys . rev .
d * 89 * , 072008 ( 2014 ) .
j. zhang _
( belle collaboration ) , phys .
* 91 * , 221801 ( 2003 ) .
b. aubert _ et al .
_ ( babar collaboration ) , phys . rev .
* 102 * , 141802 ( 2009 ) .
a. somov _ et al .
_ ( belle collaboration ) , phys . rev
. lett . * 96 * , 171801 ( 2006 ) .
b. aubert _ et al .
_ ( babar collaboration ) , phys . rev .
d * 76 * , 052007 ( 2007 ) .
b. aubert _ et al .
_ ( babar collaboration ) , phys . rev .
d * 78 * , 071104(r ) ( 2008 ) .
m. gronau and d. wyler , phys .
* b265 * , 172 ( 1991 ) ; d. atwood , i. dunietz and a. soni , phys .
lett . * 78 * , 3257 ( 1997 ) .
a. giri , y. grossman , a. soffer and j. zupan , phys .
d * 68 * , 054018 ( 2003 ) .
h. aihara _ et al . _
( belle collaboration ) , phys .
d * 85 * , 112014 ( 2012 ) .
a. bondar and a. poluektov , eur .
j. c * 55 * , 51 ( 2008 ) .
j. libby _ et al .
_ ( cleo collaboration ) , phys . rev .
d * 82 * , 112006 ( 2010 ) .
a. crivellin and u. nierste , phys .
d * 79 * , 035018 ( 2009 ) .
see , for example , r.d.c .
miller and b.h.j .
mckellar , phys . rep . *
106 * , 169 ( 1984 ) ; g. buchalla , a.j .
buras and m.e .
lautenbacher , rev . mod
. phys . * 68 * , 1125 ( 1996 ) . z.z .
xing , h. zhang and s. zhou , phys .
d * 77 * , 113016 , 2008 .
arxiv:0712.1419 [ hep - ph ] . | we show that the experimental data on @xmath0 in various @xmath1 meson decay modes suggest a possibility of @xmath2-violating right - handed current in the @xmath3 transition .
its consequences in @xmath4 are examined and compared with experimental results in order to clarify possible signals of the @xmath2 violation in these decay modes . as a result
, we find that the @xmath2-violating right - handed current is consistent with current experimental data .
its signal might be discovered by precise @xmath2 measurements in future experiments . |
blazars are the most commonly detected extragalactic vhe @xmath1-ray sources , with steadily increasing numbers in the very high energy regime ( vhe , e@xmath8@xmath9gev ) in the past 15 years of ground - based @xmath1-ray astronomy .
these objects are a subclass of active galactic nuclei ( agns ) with a set of characteristic properties like strong continuum emission extending from the radio all the way to the @xmath1-ray regime , high polarization ( at both optical and radio frequencies ) and rapid variability at all frequencies and on all time scales probed so far .
blazars are thought to be agns with jets which are closely aligned with our line - of - sight .
this type of agn subclass emits a characteristic spectral energy distribution ( sed ) with at least two broad emission components : one peak with a maximum in optical to x - ray band and a second peak located in the @xmath1-ray bands @xcite .
the first peak is commonly thought to be related to the synchrotron emission process in magnetic fields of the jets and the second peak is explained as inverse compton ( ic ) scattering of low energy photons @xcite .
if the low energy photons which undergo the ic process are the synchrotron photons , the process is known as the synchrotron self compton ( ssc ) mechanism @xcite .
alternatively , the origin of the low energy photons can be external to the jet due to external compton scattering ( ec ; @xcite ) .
1es0033 + 595 is a blazar near the galactic plane at cordinates ( j2000 ) r.a.:00:35:52.63 and dec:59:50:04.56 @xcite , belonging to the bl lac type .
it is classified as an extreme high - frequency peaked ( hbl ) object with synchrotron emission peaking near @xmath10hz @xcite .
so far , optical observations of 1es0033 + 595 were not able to resolve the host galaxy to determine a photometric redshift and thus the redshift of the blazar remains uncertain .
a tentative redshift of 0.086 was derived by perlman , and mentioned in falomo & kotilainen ( 1999 ) as a private communication , however to the best of our knowledge the details are not yet published . from the hubble space telescope ( hst ) images the only information that could be derived was the brightness of the nucleus and an upper limit to the brightness of the surrounding nebulosity , from which a lower limit of @xmath11 has been derived by sbarufatti et al .
( 2005 ) . 1es0033 + 595
was first detected as a hard x - ray source by the einstein slew survey in 1992 @xcite . in 1996
it was observed by the hst as part of the snapshot survey of bl lac objects , and was resolved into two point - like sources with a separation of @xmath12 @xcite .
these two objects with nearly identical brightness were explained as multiple images of a gravitationally lensed system .
however , the vla astrometric observations performed in 1997 did not detect a second radio source , ruling out that possibility @xcite .
1es0033 + 595 was observed by the x - ray satellite bepposax in december 1999 .
due to high galactic absorption it could only be detected in the lecs instrument above @xmath13kev and in the pds instrument up to @xmath14 @xmath15kev @xcite .
the source was also detected with the _ integral _ satellite in 2003 in the @xmath16
@xmath17kev energy band with a statistical significance of @xmath18 @xmath2 @xcite .
in addition , during integral observations in 2005 @xcite the source was a factor of 2.4 brighter .
the fact that all three x - ray observations show different flux levels emphasize very well the variable x - ray nature of this bl lac object . with its large field of view and nearly continuous sky coverage the burst alert telescope ( bat ) instrument on - board
the _ swift _ satellite also detected 1es0033 + 595 during the first 22 months ( 2004 dec - 2006 oct ) of observation @xcite , and has been included in the 22 , 58 , and 70-month bat catalogs .
in the he @xmath1-ray range the source has been continously detected by the _ fermi _
large area telescope ( lat ) .
the source was reported for the first time after the first 5.5 months of sky survey observations @xcite . since then it has been part of the _ fermi _ first bright agn catalog , with a spectrum consistent with a flat power law with @xmath19 = 2.00 @xmath20 0.13 and a flux of f(@xmath21mev ) = ( 20.3 @xmath20 5.1)@xmath22 @xcite . in the vhe @xmath1-ray band
this source was first observed for @xmath23h in december 1995 by the whipple observatory .
these observations yielded only upper limits , f(@xmath24gev ) @xmath25 @xmath26 , equivalent to 20% of the crab nebula flux @xcite .
magic observations of this source were motivated by the bepposax observations @xcite .
therein 1es0033 + 595 was one of the most promising candidate tev emitters .
magic observed this object in 2006 and later in 2008 for about @xmath27h . from these observations ,
only a flux upper limit at 95% confidence level was obtained : f(@xmath28 gev ) @xmath29 , 9.7% of the crab nebula flux @xcite .
new observations in 2009 during the commissioning phase of the magic stereoscopic system led to the discovery of the source in the vhe-@xmath1-ray range @xcite , as described in this paper .
observations of 1es0033 + 595 performed in each energy bands are decribed below .
magic is a system of two @xmath30m - dish imaging atmospheric cherenkov telescopes ( iacts ) located at the roque de los muchachos observatory ( @xmath31n , @xmath32w , @xmath33 m a.s.l . ) , in the canary island of la palma . since 2009
the magic telescopes have carried out stereoscopic observations with a sensitivity of @xmath34 of the crab nebula flux , for energies above @xmath35gev in @xmath17h of observations @xcite .
the trigger energy threshold of the system is the lowest among the currently operating iacts , with an accessible energy range between @xmath17gev and several tev .
the magic telescopes observed 1es0033 + 595 from 2009 august 17 until october 14 , for a total observation time of @xmath36h .
these observations were performed during the commissioning phase of the magic stereoscopic system . during this time
, the data taking was carried out using the so - called `` soft stereo trigger mode '' , i.e. using the magic - i trigger system and operating magic - ii in `` slave mode '' .
compared to the standard trigger mode adopted for regular stereoscopic observations ( the so - called `` full stereo trigger mode '' , where the cherenkov events are triggered simultaneously by both telescopes ) , the `` soft stereo trigger mode '' has slightly less sensitivity at the energies below @xmath37 gev . in order to take the non - standard trigger condition into account ,
dedicated monte carlo ( mc ) @xmath1-ray simulations were generated and adopted in this analysis .
the observations of 1es0033 + 595 were carried out in the so - called wobble mode @xcite , in which the pointing direction alternates every @xmath16 min between two positions , offset by @xmath38 in r.a . from the source .
the data were taken at zenith angles ranging between @xmath39 and @xmath40 , which resulted in an analysis energy threshold ( defined as the peak of the mc @xmath1-ray simulated energy distribution for an energy distribution of photon index 3.8 after all analysis cuts ) of @xmath41gev .
after the application of standard quality checks based on the rate of the stereo events and the distributions of basic image parameters , @xmath42h of effective on - time events were selected to derive the results .
the rejected data were affected mainly by non - optimal atmospheric conditions during the data taking .
the data analysis was performed using the standard software package mars @xcite , including the latest standard routines for the stereoscopic analysis @xcite . after the calibration @xcite and the image cleaning of the events , the information from the individual telescopes is combined and the calculation of basic stereo image parameters is performed . for the @xmath1/hadron separation and @xmath1-direction estimation a multivariate method called random forest (
rf , @xcite ) is applied . for the former task ,
the algorithm employs basic image parameters @xcite , timing information @xcite , and stereo parameters @xcite to compute a @xmath1/hadron discriminator called _ hadronness _ by comparison of real ( hadron - dominated ) data with mc @xmath1-ray simulations .
the _ hadronness _
parameter ranges from 0 ( for showers confidently identified as initiated by @xmath1-rays ) to 1 ( for those clearly showing the features of a hadronic cosmic - ray initiated shower ) .
finally , the estimation of the energy of the events is achieved by averaging individual energy estimators for both telescopes based on look - up tables @xcite .
the final analysis cuts applied to the 1es0033 + 595 data were optimized by means of contemporaneous crab nebula data and mc simulations . in computing the significance of the signal coming from the 1es0033 + 595 sky region , single cuts in _ hadronness _ and @xmath43 ( see section [ sec : magicresults ] ) optimized for energies close to the threshold were applied - rays of @xmath44 . ] .
conversely , while deriving the spectrum and the light curve of the source , multiple cuts optimized in logarithmic energy bins were considered .
the _ fermi _ lat ( large area telescope ) is a pair conversion telescope designed to cover the energy band from @xmath16 mev to more than @xmath5 gev .
normally it operates in all - sky survey mode , scanning the entire sky every 3h .
therefore it can provide observations of 1es 0033 + 595 simultaneous with magic .
_ fermi _ data presented in this paper were collected from 2009 august 17 to october 14 . in addition , to make a comparison with the behavior of the source over a wider time interval , we also show the data analyzed from the beginning of the science phase of the _ fermi _ mission , that is 2008 august 4 , to 2011 october 28 .
both sets of data were analyzed with the _ fermi _ science tools package version 09 - 27 - 01 , available from the _ fermi _ science support center , and with the post - launch instrument response functions [ irfs , p7source_v6 , @xcite ] .
only events belonging to the `` source '' class and located in a circular region of interest ( roi ) of @xmath45 radius , centered at the position of 1es 0033 + 595 , were selected .
in addition , we excluded photons with zenith angles @xmath46 to limit contamination from earth limb @xmath1-rays , and data taken when the rocking angle of @xmath47 was greater than @xmath48 to avoid time intervals during which the earth entered the lat field of view .
the data analysis of 1es 0033 + 595 is challenging because this source is located near the galactic plane . as a consequence , to reduce the contamination by the galactic plane diffuse emission , we decided to restrict the study to the 300mev - 300gev energy range where we can profit from the narrower point - spread function to separate the @xmath1-ray emission associated with our source from the intense galactic foreground .
the analysis in the time interval simultaneous with magic observations was performed using an unbinned maximum likelihood method @xcite . instead
, a binned maximum likelihood technique was used for the 38 months data set .
all point sources from the 2fgl catalog @xcite located within @xmath49 of 1es 0033 + 595 , and a background model , were included in the model of the region . the background model used for the analysis includes a galactic diffuse emission component and an isotropic component ( including residual instrumental background ) , modelled with the files gal_2yearp7v6_v0.fits and iso_p7v6source.txt , which are publicly available . in the full energy range analysis ,
all point sources within the @xmath45 radius roi were fitted with their paramenters set free , while sources beyond @xmath45 radius roi had their parameters frozen to the values reported in 2fgl .
the normalizations of the background components were allowed to vary freely .
the kva ( kungliga vetenskapsakademien ) telescopes are located at la palma but operated remotely by the tuorla observatory from finland .
these telescopes are used mainly for optical support observations for the magic telescopes .
specifically , there is a @xmath15 cm telescope which is used for polarimetric observations and a @xmath50 cm telescope which performs simultaneous photometric observations with magic .
furthermore the smaller @xmath50 cm telescope monitors potential vhe @xmath1-ray candidate agns in order to trigger magic observations if one of these selected objects is in a high optical state .
such observations are performed in the r - band and the magnitude of the source is measured from ccd images using differential photometry . during the magic observation of 1es0033 + 595 the average optical flux obtained by kva was r=17.93magnitude , which corresponds to @xmath51mjy . to derive this @xmath52 in the optical band the contribution from a nearby star ( 0.22mjy ) was subtracted from the total measured flux ( nilsson et al .
moreover , the brightness was corrected for galactic absorption by r=2.35mag ( schlegel et al .
the average @xmath52 during the magic observations corresponds to 8.5@xmath20 0.5 @xmath53@xmath54 .
as outlined in figure [ fig:0033_opticl_lc ] , during a 9 year kva survey , the source shows only a marginal flux variability and also during the magic observation window ( bottom panel in figure [ fig:0033_opticl_lc ] ) no significant variability was found .
[ fig : maxi ] distributions of the 1es0033 + 595 signal ( open circle ) and background ( filled circle ) estimation from @xmath42h of magic stereo observations taken between 2009 august 17 and october 14 , above an energy threshold of @xmath55gev .
the region between zero and the vertical dashed line ( at @xmath56 @xmath57 ) represents the signal region . , scaledwidth=53.0% ] [ fig : theta2 ] h of magic stereo observations above the estimated energy threshold of @xmath55gev .
the color scale represents the test statistic value distribution .
the white circle in the lower left indicates the point - spread function ( 68@xmath58 containment ) for this analysis .
, scaledwidth=53.0% ] [ fig : sky ] we extracted a 4 - 10 kev light curve covering the time frame from 2009 august 17 until october 14 ( the magic observation window ) from the monitor of all - sky x - ray image ( maxi ) mission onboard the international space station ( @xcite ) .
this light curve is shown in figure 2 .
fitting the x - ray light curve with a constant flux hypothesis yields a flux of @xmath59photons@xmath60s@xmath61 with a @xmath62 , that corresponds to a probability @xmath63 .
the source shows only marginally significant variability throughout the magic observations .
there are no _ swift _ , _ xmm _ or _ integral _ flux points available which would match the magic observation window .
we have thus used data collected from the asdc sed builder tool during the observation period which are closer in time to the magic observation period .
all x - ray data points selected for this study are shown in table1 .
.the selected x - ray data points in this study .
[ cols= " < , < , < , < , < " , ]
in this paper the first detection of vhe @xmath1-rays from 1es0033 + 595 has been reported . from the 2009 magic data the source
is clearly detected at a significance level of @xmath64 .
the multiwavelength data presented here confirms the typical hbl blazar subclass behavior of 1es0033 + 595 : marginal variability in the optical r - band , a hard spectrum in the _ fermi _ lat regime , and emission of vhe @xmath1-rays .
moreover , the magic detection of 1es0033 + 595 confirms the identification as a likely vhe @xmath1-ray emitter by the costamante & ghisellini ( 2002 ) list . since the redshift of this source is unknown , but crucial for accurate sed modeling , a new estimation ( @xmath65 ) with the empirical approach of prandini et al .
( 2010 ) was performed .
this result is in a good agreement with the lower limit of @xmath11 presented in sbarufatti et al .
( 2005 ) and with empirical findings where the sources with redshift greater than 0.2 are characterized by a photon index of @xmath66 .
finally , a comparison with other hbl objects @xcite shows that the model parameters used here for the sed fitting are compatible with those obtained for other hbl class objects .
considering the large uncertainty in the measured vhe spectrum and the unavailability of simultaneous x - ray data , further work on ebl and igmf using the presented data is difficult .
however , proper simultaneous mwl coverage on this high redshift object would allow to perform these studies .
we would like to thank the instituto de astrofsica de canarias for the excellent working conditions at the observatorio del roque de los muchachos in la palma . the support of the german bmbf and mpg , the italian infn , the swiss national fund snf , and the spanish micinn is gratefully acknowledged .
this work was also supported by the cpan csd2007 - 00042 and multidark csd2009 - 00064 projects of the spanish consolider - ingenio 2010 programme , by grant do02 - 353 of the bulgarian nsf , by grant 127740 of the academy of finland , by the dfg cluster of excellence `` origin and structure of the universe '' , by the dfg collaborative research centers sfb823/c4 and sfb876/c3 , and by the polish mniszw grant 745/n - hess - magic/2010/0 .
+ the _ fermi _ lat collaboration acknowledges generous ongoing support from a number of agencies and institutes that have supported both the development and the operation of the lat as well as scientific data analysis .
these include the national aeronautics and space administration and the department of energy in the united states , the commissariat lenergie atomique and the centre national de la recherche scientifique / institut national de physique nuclaire et de physique des particules in france , the agenzia spaziale italiana and the istituto nazionale di fisica nucleare in italy , the ministry of education , culture , sports , science and technology ( mext ) , high energy accelerator research organization ( kek ) and japan aerospace exploration agency ( jaxa ) in japan , and the k. a. wallenberg foundation , the swedish research council and the swedish national space board in sweden . additional support for science analysis during the operations phase is gratefully acknowledged from the istituto nazionale di astrofisica in italy and the centre national dtudes spatiales in france .
+ 999999 abdo , a. a. , ackermann , m. , ajello , m. , et al .
2009 , the astrophysics journal , 707 , 1310 ackermann , m .
et al . , 2012a , astroparticle physics , 35 , 346 ackermann , m .
et al . , 2012b , the astrophysics journal , 750 , 3 albert , j. , aliu , e. , anderhub , h. , et al . 2008a , the astrophysics journal , 674 , 1037 albert , j. , aliu , e. , anderhub , h. , et al .
2008b , nucl .
instrum . meth . , a594 , 407419 albert , j. , aliu , e. , anderhub , h. , et al . 2008c , nucl .
instrum . meth . , a588 , 424432 aleksi , j. , antonelli , l. a. , antoranz , p. , et al . 2011a , the astrophysics journal , 729 , 115 aleksi , j. , antonelli , l. a. , antoranz , p. , et al .
2011b , the astrophysics journal , 726 , 58 aleksi , j. , alvarez , e. a. , antonelli , l. a. , et al .
2012a , astroparticle physics , 35 , 435 aleksi , j. , alvarez , e. a. , antonelli , l. a. , et al . 2012b , the astrophysics journal , 748 , 46 aliu , e. , anderhub , h. , antonelli , l. a. , et al .
2009 , astroparticle physics , 30 , 293 costamante , l. , ghisellini , g. , giommi , p. , et al .
2001 , astronomy & astrophysics , 371 , 512 costamante , l. & ghisellini , g. 2002 , astronomy & astrophysics , 384 , 56 dermer , c. d. & schlickeiser , r. 1993 , the astrophysics journal , 416 , 458 elvis , m. , plummer , d. , schachter , j. , et .
1992 , the astrophysics journal suppliment , 80 , 257 falomo , r. & kotilainen , j. 1999 , astronomy & astrophysics , 352 , 85 franceschini , a. , rodighiero , g. & vaccari , m. 2008 , astronomy & astrophysics , 487 , 837 fomin , v. p. , stepanian , a. a. , lamb , r. c. , et al .
1994 , astroparticle physics , 2 , 137 giommi , p. , capalbi , m. , fiocchi , m. , et al .
2002 , blazar astrophysics with bepposax and other observatories , 63 den hartog , p. r. , hermsen , w. , kuiper , l. , et al .
2006 , astronomy & astrophysics , 451 , 587 hillas , a. m. 1985 , proc . of 19th icrc , la jolla , 445 horan , d. , badran , h. m. , bond , i. h. , et al .
2004 , the astrophysics journal , 603 , 51 kuiper , l. , hermsen , w. , in t zand , j. j. m. & den hartog , p. r. 2005 , the astronomer s telegram , 662 , 1 li , t .- p . &
ma , y .- q .
1983 , the astrophysics journal , 272 , 317 lombardi , s. , et al .
2011 , proc . of 32nd icrc , beijin , china , august 2011 , arxiv:1109.6195 mankuzhiyil , n. ,
2011 , the astrophysics journal , 733 , 14 mankuzhiyil , n. , et al .
2012 , the astrophysics journal , 753 , 154 mannheim , k. , 1993 astronomy & astrophysics , 269 , 67 maraschi , l. & tavecchio , f. 2003 , the astrophysics journal , 593 , 667 mariotti , m. 2011 , the astronomer s telegram , 3719 , 1 matsuoka , m. et al .
2009 , publ .
japan 61 , 999 mattox , j. r. , bertsch , d. l. , chiang , j. , et al .
1996 , the astrophysics journal , 461 , 396 nieppola , e. , tornikoski , m. & valtaoja , e. 2006 , astronomy & astrophysics , 445 , 441 nilsson , k. , pasanen , m. , takalo , l. o. , et al .
2007 , astronomy & astrophysics , 475 , 199 nolan , p. l. , abdo , a. a. , ackermann , m. , et al .
2012 , the astrophysics journal suppliment , 199 , 31 perlman , e. s. 2000 , american institute of physics conference series , 515 , 53 pivato , g .
2013 , the astrophysics journal , 779 , 179 prandini , e. , mariotti , m. & tavecchio , f. 2011 , arxiv:1111.0913 prandini , e. , bonnoli , g. , maraschi , l. , et al .
2010 , monthly notices of the royal astronomy society , 405 , l76 rector , t. a. , gabuzda , d. c. & stocke , j. t. 2003 , the astronomical journal , 125 , 1060 rees , m. j. , 1967 , monthly notices of the royal astronomy society , 135 , 345 rolke , w. a. , lpez , a. m. & conrad , j. 2005 , nucl .
instrum . meth . , a551 , 493 sbarufatti , b. , treves , a. & falomo , r. 2005 , the astrophysics journal , 635 , 173 scarpa , r. , urry , c. m. , falomo , r. , et al .
1999 , the astrophysics journal , 521 , 134 schlegel , d. j. , finkbeiner , d. p. & davis , m. 1998 , the astrophysics journal , 500 , 525 sikora , m. , moderski , g. & poutanen , j. 1997 , the astrophysics journal , 484 , 108 stecker , f. w. , de jager , o. c. , & salamon , m. h. 1992 , the astrophysics journal letters , 390 , l49 tavecchio , f. , maraschi , l. , & ghisellini , g. 1998 , the astrophysics journal , 509 , 608 tavecchio , f. , ghisellini , g. , ghirlanda , g. , 2010 , monthly notices of the royal astronomy society , 401 , 1570 tueller , j. , baumgartner , w. h. , markwardt , c. b. , et al . 2010 , the astrophysics journal suppliment , 186 , 378 uellenbeck , m. et al . 2012 , aip conference proceedings , 1505 , 494 urry , c.m & padovani , p , 1995 , publicatiobs of the astronomical society of the pacific , 107 , 83 @xmath67 ifae , edifici cn . , campus uab , e-08193 bellaterra , spain + @xmath68 universit di udine , and infn trieste , i-33100 udine , italy + @xmath69 inaf national institute for astrophysics , i-00136 rome , italy + @xmath70 universit di siena , and infn pisa , i-53100 siena , italy + @xmath71 croatian magic consortium , rudjer boskovic institute , university of rijeka and university of split , hr-10000 zagreb , croatia + @xmath72 max - planck - institut fr physik , d-80805 mnchen , germany + @xmath73 universidad complutense , e-28040 madrid , spain + @xmath74 inst . de astrofsica de canarias , e-38200 la laguna , tenerife , spain + @xmath75 university of d , pl-90236 d , poland + @xmath76 deutsches elektronen - synchrotron ( desy ) , d-15738 zeuthen , germany + @xmath77 eth zurich , ch-8093 zurich , switzerland + @xmath78 universitt wrzburg , d-97074 wrzburg , germany + @xmath79 centro de investigaciones energticas , medioambientales y tecnolgicas , e-28040 madrid , spain + @xmath80 technische universitt dortmund , d-44221 dortmund , germany + @xmath81 inst . de astrofsica de andaluca ( csic ) , e-18080 granada , spain + @xmath82 universit di padova and infn , i-35131 padova , italy + @xmath83 universit dellinsubria , como , i-22100 como , italy + @xmath84 unitat de fsica de les radiacions , departament de fsica , and ceres - ieec , universitat autnoma de barcelona , e-08193 bellaterra , spain + @xmath85 institut de cincies de lespai ( ieec - csic ) , e-08193 bellaterra , spain + @xmath86 japanese magic consortium , division of physics and astronomy , kyoto university , japan + @xmath87 finnish magic consortium , tuorla observatory , university of turku and department of physics , university of oulu , finland + @xmath88 inst . for nucl . research and nucl .
energy , bg-1784 sofia , bulgaria + @xmath89 universitat de barcelona ( icc , ieec - ub ) , e-08028 barcelona , spain + @xmath90 universit di pisa , and infn pisa , i-56126 pisa , italy + @xmath91 now at nasa goddard space flight center , greenbelt , md 20771 , usa and department of physics and department of astronomy , university of maryland , college park , md 20742 , usa + @xmath92 now at department of physics and astronomy and the bartol research institute , university of delaware , newak , de 19716 , usa @xmath93 now at ecole polytechnique fdrale de lausanne ( epfl ) , lausanne , switzerland + @xmath94 now at department of physics & astronomy , uc riverside , ca 92521 , usa + @xmath95 now at finnish centre for astronomy with eso ( finca ) , turku , finland + @xmath96 now at : astrophysical sciences division , barc , mumbai 400085 , india + @xmath97 also at inaf - trieste + @xmath98 also at instituto de fisica teorica , uam / csic , e-28049 madrid , spain + @xmath99 now at : stockholm university , oskar klein centre for cosmoparticle physics , se-106 91 stockholm , sweden + @xmath100 now at grappa institute , university of amsterdam , 1098xh amsterdam , netherlands | the number of known very high energy ( vhe ) blazars is @xmath0 , which is very small in comparison to the number of blazars detected in other frequencies .
this situation is a handicap for population studies of blazars , which emit about half of their luminosity in the @xmath1-ray domain .
moreover , vhe blazars , if distant , allow for the study of the environment that the high - energy @xmath1-rays traverse in their path towards the earth , like the extragalactic background light ( ebl ) and the intergalactic magnetic field ( igmf ) , and hence they have a special interest for the astrophysics community .
we present the first vhe detection of 1es0033 + 595 with a statistical significance of 5.5@xmath2 .
the vhe emission of this object is constant throughout the magic observations ( 2009 august and october ) , and can be parameterized with a power law with an integral flux above 150 gev of @xmath3 and a photon index of ( @xmath4 ) .
we model its spectral energy distribution ( sed ) as the result of inverse compton scattering of synchrotron photons . for the study of the sed we used simultaneous optical r - band data from the kva telescope , archival x - ray data by
_ swift _ as well as _ integral _ , and simultaneous high energy ( he , @xmath5mev
@xmath6gev ) @xmath1-ray data from the _ fermi _ lat observatory . using the empirical approach of prandini et al .
( 2010 ) and the _ fermi_-lat and magic spectra for this object , we estimate the redshift of this source to be @xmath7 .
this is a relevant result because this source is possibly one of the ten most distant vhe blazars known to date , and with further ( simultaneous ) observations could play an important role in blazar population studies , as well as future constraints on the ebl and igmf .
galaxies : galaxies : bl lacertae objects : individual ( 1es0033 + 595 ) - gamma rays : galaxies . |
an important goal of the @xmath1-physics program is the precise measurement of the ckm matrix @xcite elements @xmath4 and @xmath5 . the most accurate determinations are obtained from semileptonic decays @xmath6 and @xmath7 , respectively .
generally , two different approaches can be used : the hadronic state @xmath8 can be reconstructed either in specific exclusive modes , or inclusively by summing over all possible hadronic final states .
the inclusive determinations rely on an operator product expansion ( ope ) in inverse powers of the @xmath0-quark mass @xmath9 @xcite . at second order in the expansion ,
two nonperturbative parameters arise , which describe the kinetic energy and the chromomagnetic moment of the @xmath0 quark inside the @xmath1 meson . in the kinetic scheme ,
they are denoted by @xmath10 and @xmath11 , respectively .
two more parameters arise at third order , @xmath12 and @xmath13 . in the kinetic scheme , short- and long - distance contributions
are separated by a hard cutoff @xmath14 and the @xmath0-quark mass and nonperturbative parameters are given at @xmath15 .
the mass @xmath9 and the nonperturbative parameters can be determined from the study of kinematic distributions in semileptonic and radiative penguin @xmath1 decays .
precise measurements of @xmath9 are needed both to reduce the uncertainty of inclusive determinations of @xmath5 , as well as for studying new physics effects in the higgs sector at future experiments . here
, we present recent determinations of @xmath9 and higher - order nonperturbative parameters from charm and charmless semileptonic and radiative penguin @xmath1 decays at the experiment @xcite .
in all analyses presented , @xmath16 decays are tagged by reconstructing one @xmath1 meson ( @xmath17 ) fully in hadronic modes , @xmath18 .
the @xmath19 system is composed of hadrons with a total charge of @xmath20 , @xmath21 , with @xmath22 , @xmath23 , @xmath24 .
we test the kinematic consistency of @xmath17 candidates with two variables , @xmath25 and @xmath26 here , @xmath27 is the invariant mass of the @xmath2 system and @xmath28 and @xmath29 denote the energy and momentum of the @xmath17 candidate in the @xmath30 frame .
we require @xmath31 to be @xmath32 within three standard deviations . in events with multiple @xmath17 candidates
we retain the candidate reconstructed in the mode with highest purity as estimated from the ratio of signal over background for events with @xmath33 on monte carlo simulation ( mc ) . by fully reconstructing one of the @xmath1 mesons in the event , the charge , flavor and momentum of the second @xmath1
can be inferred .
all particles that are not used in the reconstruction of the @xmath17 are assigned to the decay of the signal @xmath1 .
the efficiency to reconstruct a @xmath17 candidate is @xmath34 ( @xmath35 ) for @xmath36 ( @xmath37 ) events .
to @xmath38 for two @xmath39 regions .
the dashed curve shows the signal , the solid grey ( red ) curve the background as given by the fit .
the solid black curve shows the sum of signal and background .
( a ) @xmath40 for charged @xmath1 events and ( b ) @xmath41 for neutral @xmath1 events .
( c ) the measured photon energy spectrum before subtraction of backgrounds .
points show the data spectrum , the shaded histogram shows the @xmath42 backgrounds , where the shape is taken from mc.__,title="fig : " ] we use fits to the @xmath38 distribution to subtract the combinatorial background from @xmath42 events and and the background from continuum ( @xmath43 ) events .
the backgrounds are modeled with a threshold function @xcite and the signal is described by a gaussian function joined with an exponential tail to describe photon energy loss @xcite .
examples from the measurement in radiative penguin decays are shown in fig .
[ btosfig ] ( a ) and ( b ) .
the first and second moments of the photon energy spectrum in radiative penguin decays , @xmath44 , are used for a preliminary determination of @xmath9 and @xmath10 @xcite .
the moments are extracted as a function of the lower cut on the photon energy , @xmath45 , measured in the rest frame of the signal @xmath1 .
the full reconstruction of the second @xmath1 in the event results in an improved signal purity and different systematic uncertainties , but lower statistics , than alternative methods .
the measurement is based on a sample of @xmath46 million @xmath42 pairs .
events with a well reconstructed , high energy photon are selected if the photon is not compatible with originating from the decay of a @xmath47 or @xmath48 , or a @xmath49 decay assuming that the second photon from the @xmath47 decay was lost .
continuum background is suppressed by using a fisher discriminant that makes use of the difference between event topologies in @xmath42 and continuum events .
the @xmath39 spectrum is measured in bins of @xmath50 and is shown in fig .
[ btosfig ] ( c ) .
the region @xmath51 is used to normalize the backgrounds , the largest part of which consists of photons from unreconstructed @xmath47 or @xmath48 decays .
the backgrounds are extrapolated into the signal region , @xmath52 , the shape of the backgrounds is taken from mc .
the signal region contains @xmath53 @xmath54 events over an estimated background of @xmath55 events
. the measured photon spectrum is corrected for efficiency , which varies with @xmath39 , and resolution effects .
first and second central moments , @xmath56 and @xmath57 , as a function of @xmath45 are extracted from the corrected spectrum .
the moments for @xmath58 are used to determine @xmath59 and @xmath60 in the kinetic scheme with a correlation of @xmath61 .
moments of the hadronic mass and lepton energy spectra in @xmath6 decays are used for a preliminary extraction of @xmath9 , the charm - quark mass @xmath62 , nonperturbative parameters and @xmath4 @xcite .
we present moments of the hadronic mass @xmath63 , @xmath64 , which use a larger data set than the previous measurement and new measurements of the mixed hadronic mass - energy moments @xmath65 , @xmath66 .
@xmath67 is defined by @xmath68 , where @xmath69 is the mass and @xmath70 the energy of the inclusive @xmath71 system in the @xmath1 rest frame and @xmath72 .
the mixed moments are expected to yield a more precise determination of higher order nonperturbative parameters .
the moments are extracted as a function of a lower cut on the lepton energy between @xmath73 and @xmath74 in the signal @xmath1 rest frame .
spectrum for @xmath75 in the @xmath1 rest frame .
tag - side backgrounds are indicated by the hatched histogram , signal - side backgrounds by the cross - hatched histogram.__,title="fig:",scaledwidth=53.2% ] spectrum for @xmath75 in the @xmath1 rest frame .
tag - side backgrounds are indicated by the hatched histogram , signal - side backgrounds by the cross - hatched histogram.__,title="fig:",scaledwidth=46.8% ] the measurement is based on a sample of @xmath46 million @xmath42 pairs . after reconstructing the @xmath17 and identifying the lepton , where both electrons and muons are used , the hadronic system is reconstructed from the remaining tracks and neutral energy depositions in the event .
a kinematic fit imposing energy - momentum conservation and the missing energy and momentum in the event to be consistent with coming from one neutrino is performed to improve the resolution in the hadronic variables .
the hadronic mass and mixed hadronic mass - energy spectra are shown in fig .
[ btocfig ] .
the moments are extracted directly from the kinematically fitted hadronic masses and energies and are corrected for the effect of lost particles . the main contribution to
the systematic uncertainties on the moments arise from the impact of the reconstruction efficiencies of neutral particles on the inclusive event reconstruction . and non - semileptonic backgrounds .
( b ) unfolded hadronic mass spectrum in @xmath7 .
the inner error bars show the statistical uncertainties only.__,title="fig:",scaledwidth=50.0% ] and non - semileptonic backgrounds .
( b ) unfolded hadronic mass spectrum in @xmath7 .
the inner error bars show the statistical uncertainties only.__,title="fig:",scaledwidth=50.0% ] a combined fit is performed to hadronic mass moments , lepton energy moments @xcite in @xmath6 decays and photon energy moments in @xmath44 decays @xcite and yields @xmath76 , @xmath77 ( correlation @xmath78 ) , @xmath79 , @xmath80 , @xmath81 and @xmath82 in the kinetic scheme .
moments of the hadronic mass spectrum in @xmath7 decays are used for a preliminary extraction of @xmath9 , @xmath10 and @xmath12 .
their determination in @xmath7 allows for a test of the theoretical framework used for the extraction of @xmath5 in the same channel in which @xmath5 is determined .
the hadronic mass moments are measured with an upper cut on the hadronic mass to reduce experimental uncertainties .
the measurement is based on a sample of @xmath83 million @xmath42 pairs . after reconstructing the @xmath17 and identifying the lepton , where both electrons and muons with a minimum energy of @xmath84 in the @xmath1 rest frame are used , the hadronic system is reconstructed from the remaining tracks and neutral energy depositions in the event .
vetos on identified @xmath85 , reconstructed @xmath86 and partically reconstructed @xmath87 are employed to suppress the dominant background from @xmath6 events .
the remaining @xmath6 and non - semileptonic backgrounds are subtracted by a fit to the hadronic mass spectrum ( fig .
[ btoufig ] ( a ) ) .
the full @xmath88 region contains @xmath89 signal events .
the background - subtracted spectrum is unfolded for detector acceptance , efficiency and resolution effects ( fig .
[ btoufig ] ( b ) ) and the first and second and third central moments are extracted from the unfolded spectrum for @xmath90 : @xmath91 with correlation coefficients @xmath92 , @xmath93 and @xmath94 .
the main systematic uncertainties arise from the control of the @xmath6 background . a fit of these moments to predictions in the kinetic scheme @xcite yields @xmath95 with correlation coefficients @xmath96 , @xmath97 and @xmath98 .
@xmath10 plane in the kinetic scheme .
the dotted ellipse shows the result of the presented @xmath44 analysis @xcite , the dashed ellipse the result from @xmath6 decays @xcite , the black solid ellipse the result from its combination with earlier @xmath44 measurements @xcite and the solid grey ellipse the result from @xmath7 decays.__,title="fig:",scaledwidth=50.0% ] we presented preliminary determinations of the @xmath0-quark mass @xmath9 and nonperturbative parameters from charmed and charmless semileptonic and radiative penguin @xmath1 decays at .
the determination in charmless semileptonic @xmath1 decays has been performed for the first time .
the results for @xmath9 and @xmath10 are summarized in table [ resultstable ] and compared in fig .
[ mbmupisqfig ] .
the determinations in the different channels are consistent within the quoted uncertainties and with earlier determinations @xcite .
99 n. cabibbo , phys .
* 10 * , 531 ( 1963 ) .
m. kobayashi and t. maskawa , prog .
phys . * 49 * , 652 ( 1973 ) .
i. i. y. bigi , m. a. shifman , n. g. uraltsev and a. i. vainshtein , phys .
lett . *
71 * , 496 ( 1993 ) [ arxiv : hep - ph/9304225 ] .
j. chay , h. georgi and b. grinstein , phys .
b * 247 * , 399 ( 1990 ) .
a. v. manohar and m. b. wise , phys . rev .
d * 49 * , 1310 ( 1994 ) [ arxiv : hep - ph/9308246 ] .
b. aubert _ et al . _ [ babar collaboration ] , nucl .
instrum .
a * 479 * , 1 ( 2002 ) [ arxiv : hep - ex/0105044 ] .
argus collaboration , h. albrecht _ et al .
_ , z. phys .
c * 48 * , 543 ( 1990 ) .
t. skwarnicki , desy - f31 - 86 - 02 .
b. aubert _ et al . _ [ babar collaboration ] , arxiv:0711.4889 [ hep - ex ] .
submitted to phys .
d. b. aubert _ et al . _ [ babar collaboration ] , arxiv:0707.2670 [ hep - ex ] .
b. aubert _ et al . _ [ babar collaboration ] , phys .
d * 69 * , 111104 ( 2004 ) [ arxiv : hep - ex/0403030 ] .
b. aubert _ et al . _ [ babar collaboration ] , phys .
d * 72 * , 052004 ( 2005 ) [ arxiv : hep - ex/0508004 ] .
b. aubert _ et al . _ [ babar collaboration ] , phys .
lett . * 97 * , 171803 ( 2006 ) [ arxiv : hep - ex/0607071 ] .
p. gambino , g. ossola and n. uraltsev , jhep * 0509 * , 010 ( 2005 ) [ arxiv : hep - ph/0505091 ] . c. w. bauer , z. ligeti , m. luke , a. v. manohar and m. trott , phys .
d * 70 * , 094017 ( 2004 ) [ arxiv : hep - ph/0408002 ] .
o. buchmller and h. flcher , phys .
d * 73 * , 073008 ( 2006 ) [ arxiv : hep - ph/0507253 ] .
k. abe _ et al . _ [ belle collaboration ] , arxiv : hep - ex/0611047 . | knowing the mass of the @xmath0-quark is essential to the study of the structure and decays of @xmath1 mesons as well as to future tests of the higgs mechanism of mass generation .
we present recent preliminary measurements of the @xmath0-quark mass and related nonperturbative parameters from moments of kinematic distributions in charmed and charmless semileptonic and radiative penguin @xmath1 decays .
their determination from charmless semileptonic @xmath1 decays is the first measurement in this mode .
the data were collected by the detector at the pep - ii asymmetric - energy @xmath2collider at the stanford linear accelerator center at a center - of - momentum energy of @xmath3 . |
the importance of etgs in the context of galaxy formation rests on their evolved nature that makes them the fossil evidence of the process of galaxy evolution .
recent studies at high redshift suggest the existence of a class of objects with very similar morphological , dynamical and stellar population properties to ` bona fide ' , nearby etgs ( see e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
this raises the issue of how significant galaxy re - processing has been during the hubble time , in particular at low redshift ( i.e. in the range 0@xmath101 ) where @xmath11cdm hierarchical models predict that most of the etg stellar mass was assembled @xcite
. the study of the stellar populations in nearby galaxies allows us to trace back with cosmic time the evolutionary history of galaxies .
some recent studies paint a surprising picture for the evolution of etgs showing that the environment plays a key role in their evolution . from the study of a sample of 124 galaxies , @xcite find that massive etgs in low density environments ( ldes ) are on average @xmath122 gyr younger and slightly ( @xmath120.05 - 0.1 dex ) more metal rich than their counterparts in high density environments . similarly , the study of @xcite , based on a sample of @xmath1314000 etgs from the sloan digital sky survey ( sdss ) , suggests that etgs in dense environments are @xmath1220% older than those in ldes . on the other hand
, the environment does not affect @xmath7-elements enhancement .
these results , obtained from the analysis of optical spectra , suggest that the evolutionary process is similar in the cluster and in the field but occurred earlier in more dense environments .
etgs in lde frequently show signatures of relatively recent evolution , as indicated by the presence of disturbed morphologies ( see e.g. * ? ? ?
* ; * ? ? ?
* ) and of kinematic sub - components @xcite .
notes : see text for a detailed explanation of single columns .
the age , metallicity and the @xmath7-enhancement obtained from the lick line - strength indices analysis , are obtained from paper iii .
the activity class is discussed in paper iv .
we used here the following notation : lln = liner ; sy = syfert like emission ; comp .
= spectra can be due to either 1 ) a combination of star formation and a seyfert nucleus , or 2 ) a combination of star formation and liner emission @xcite , in = either faint or no emission lines .
[ tab1 ] the advent of the galaxy ultraviolet explorer ( _ galex _ , @xcite ) has contributed to strengthen the idea that recent star formation is often present in etgs . through _ galex
_ ultraviolet imaging , @xcite found that 30% of massive etgs show ongoing star formation and that this fraction is higher in ldes .
@xcite concluded that etgs in low density environments are less likely to present weak episodes of recent star formation than their high - density counterparts .
_ galex _ studies of etgs with shell structures are of particular relevance , since these features , believed to be ` bona fide ' signatures of recent accretion / merging events , characterize a significant fraction ( @xmath1316.5% ) of field etgs @xcite
. combined analyses of lick line - strength indices ( e.g. mg2 , h@xmath14 , h@xmath15 , h@xmath16 ) and uv colours , revealed the presence of recent star formation in shell galaxies @xcite .
similar results have been recently obtained by @xcite for a sauron galaxy sample .
they determined the uv fundamental plane and suggested that the dominant fraction of the tilt and the scatter of this plane is due to the presence of young stars in preferentially low mass etgs .
this is the fifth paper of a series dedicated to the study of etgs with emission lines ( @xcite hereafter paper i ; @xcite hereafter paper ii ; @xcite hereafter paper iii ; @xcite hereafter paper
our program considers 65 etgs mainly located in groups and low density environments ( ldes ) . in paper iii
, we showed that the etgs of our sample have a large age spread , with luminosity - weighted ages ranging from 1 gyr to a hubble time .
luminosity - weighted ages younger than 5 gyr are always associated with the galaxies residing in the lowest density environments .
we suggest that such young ages are due to ` rejuvenation ' episodes , which do not involve more than 25% of the total galaxy stellar mass . in paper iv we investigated the nature of the ionization mechanism through standard emission - line diagnostic diagrams ( e.g. log [ oiii]/h@xmath17 vs. log [ nii]/h@xmath7 ) .
the majority of our etgs are classified as liners .
galaxies classified as seyferts in our sample tend to have very young ( @xmath185 gyr ) mean ages , supporting that star formation and agn phenomena co - exist ( see e.g. * ? ? ?
* ; * ? ? ?
the gas oxygen abundance , derived through the emission lines , gives systematically lower values than that of stars .
a possible explanation is that the gas has an external origin , e.g. derived from the accretion of a metal poorer companion .
we present in this paper galex fuv and nuv surface photometry of 40 etgs .
the uv wavelengths are particularly sensitive to young populations , and thus ideal to characterize any recent star formation .
the paper is organized as follows . in section 2
we describe the sample .
sections 3 presents the _ galex _ uv observations , the data reduction and the comparison of our results with the literature . in section 4
we present our results .
we discuss the galaxy morphological peculiarities , the shape of the surface brightness profiles , the uv radial colour profiles . in appendix a we summarize , for the individual galaxies , the uv and optical photometric properties , as well as the kinematic and morphological peculiarities . in appendix b we include the @xmath19-band surface photometry obtained for a subset of 14 galaxies from sdss data .
a sample of 40 galaxies is selected from a wider sample of 65 etgs ( see paper i and ii ) . table [ tab1 ] reports the main properties of the sample .
column ( 1 ) gives the galaxy identification ; column ( 2 ) provides the galaxy morphological classification according to rc3 @xcite . only in a few cases
does the rc3 differ from that of the rsa @xcite : ngc 4552 , ngc 5846 , ngc 6958 and ngc 7192 are classified as s0 in rsa and as e in rc3 .
column ( 3 ) gives the galaxy systemic velocity , v@xmath20 , which is lower than @xmath125000 km s@xmath21 in all cases .
for galaxies with a heliocentric systemic velocity lower than 3000 km s@xmath21 we adopt the distances ( column 4 ) provided by @xcite . for larger velocities , distances assume a hubble constant of h@xmath22=75 km s@xmath21mpc@xmath21
. columns ( 5 ) and ( 6 ) provide the optical effective radius , r@xmath23 , i.e the radius of the circular aperture encircling half of the total galaxy optical light , and the average ellipticity , as obtained from rc3 and hypercat . in column ( 7 ) we report the galaxy central velocity dispersion , within an aperture of 1/8 of the effective radius . @xcite has investigated the relation between galaxy mass - to - light ratio and the line - of - sight component of the velocity dispersion .
they provide relations ( their equation 7 and 10 ) which allow a direct transformation from the measured velocity dispersion to galaxy mass . according to the above formulae , the quoted velocity dispersion ,
that ranges from @xmath13130 to 340 km s@xmath21 , indicate that our galaxies cover a large range in masses ( roughly 3 @xmath24 10@xmath25 - 9 @xmath24 10@xmath26 m@xmath27 ) .
the luminosity - weighted ages as derived in paper iii are given in column ( 8) ; they range from [email protected] gyr ( ngc 3489 ) to [email protected] gyr ( ngc 1052 ) .
elliptical galaxies are on average older than lenticulars .
a definite trend of increasing metallicity ( column 9 ) and [ @xmath7/fe ] ( column 10 ) with the velocity dispersion is observed , testifying that the chemical enrichment was more efficient and the duration of the star formation shorter in more massive galaxies .
these two relations do not depend on the galaxy morphological type .
nebular emission lines are commonly found in the inner regions of etgs : in the optical , the fractions of documented detection in different samples are 55%-60% @xcite , 72% ( ellipticals , e ) -85% ( lenticulars , s0 ) @xcite ,
66%(e ) -83% ( s0 ) @xcite , and 52% @xcite . concerning a possible environmental effect on the presence of emission lines in etgs , @xcite , in a sample of 62 etgs in low density environments , found that 44% of the luminous etgs ( m@xmath2917 ) are detected in hi ( i.e. 10 times more frequently than in the virgo cluster ) , 60% of which show emission line ratios typical of star - forming galaxies .
optical emission lines are detected in 89% of our original sample of 65 etgs ( paper iv ) .
for the sample of 40 etgs studied in this paper , the detection drops to 82% .
incidence and strength of line emission do not correlate either with the e / s0 classification , or with the fast / slow rotator classification . by means of optical emission line ratios ( e.g. , [ oiii]@xmath305007/h@xmath17 and [ nii]@xmath306584/h@xmath7 , @xcite )
, we have shown that for the majority of our etgs the emission is `` indistinguishable '' from that of low - ionization nuclear emission - line regions ( liners , @xcite ) in agreement with other studies of etgs ( see e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . the activity class , as derived in paper iv from optical emission line diagnostic diagrams , is given in column 11 of table 1 .
the richness of the environment in which our etgs are located is well described by the parameter @xmath31 ( column 12 of table 1 ) provided by @xcite in the _ nearby galaxy catalog_. it represents the density of galaxies brighter than b=@xmath816 mag in the vicinity of each galaxy in mpc@xmath32 .
the galaxies of our sample are mainly located in low density environments .
the local density around our galaxies varies from @xmath31 @xmath13 0.1 , characteristic of nearly isolated galaxies , to @xmath31 @xmath13 4 , which is characteristic of dense galaxy regions in the virgo cluster .
for comparison , in the @xcite catalogue ngc 1380 and ngc 1389 , fornax cluster members , have values of @xmath31=1.54 and 1.50 respectively . thus our sample , although biased towards low density environments , contains a fraction of galaxies in relatively dense environments .
the uv imaging was obtained with galex @xcite in two ultraviolet bands , fuv ( 1344 1786 ) and nuv ( 1771 2831 ) .
the instrument has a very wide field of view ( 1.25 degrees diameter ) and a spatial resolution @xmath1342 and 53 fwhm in fuv and nuv respectively , sampled with 15@xmath2415 pixels @xcite .
the sample we analyze is composed of 16 etgs observed in our cycle 3 proposal ( i d gi3 - 0087 pi r. rampazzo ) and 24 etgs taken from the public _
galex _ archive ( see table 2 ) .
the exposure times for most of our sample are @xmath12 1500 seconds ( limiting magnitude in fuv / nuv of @xmath12 22.6/22.7 ab mag @xcite .
ngc 1389 and ngc 5363 have exposure times @xmath1210 times longer ( @xmath122.5 mag fainter limit ) . for each galaxy , in figure [ fig1 ]
we show the _ galex _ fuv and nuv images and a false colour composite image .
additional _ galex _ observations with exposure times of about 100 seconds exist for the following sub - set of the paper i plus paper ii samples : ngc 1947 , ngc 5193 , ngc 6721 , ngc 6758 , ngc 7007 and ngc 7377 but were not considered in the present paper , in order to have an homogeneous flux limit .
[ cols="^,^ " , ] cc + + radial colour profiles provide information about the presence of age / metallicity gradients in the stellar populations . in figure
[ fig5 ] we plot the ( fuv - nuv ) colour as a function @xmath33/r@xmath23 , where a is the semi - major axis and r@xmath23 is the optical effective radius .
we stress that , while the uv colour profiles may extend out to several optical effective radii , the ages , metallicities and @xmath7-enhancements , derived in paper iii , refer to apertures whose radius is only a fraction of r@xmath23 . in the upper panels of figure
[ fig5 ] the ( fuv - nuv ) colour profiles of galaxies with a age lower than 4 gyr are shown in the left panel and of galaxies older than 10 gyr in the right panel , using age estimates from paper iii ( see table 1 ) .
mean ages younger than 4 gyr are likely due to recent star formation episodes superimposed on an old stellar population .
we exclude from these panels etgs with rings which we collect in the bottom right panel .
figure [ fig5 ] clearly shows that the ( fuv - nuv ) colour tends to become progressively redder from the centre outwards .
this trend is consistent with the stellar populations becoming progressively more metal poor or younger ( or a combination of both ) with increasing distance from the galaxy centre .
in fact , the nuv light is dominated by turnoff stars . on the other hand ,
the fuv emission of old ` normal ' stellar populations is dominated by pagb stars .
as a population gets younger or more metal poor , the turnoff becomes bluer and more luminous , and emits more in the nuv ; at the same time , the contribution to the fuv from pagb stars diminishes because , in spite of the higher luminosity , the duration of the pagb phase gets much shorter ( i.e. , the fuel decreases ) .
this causes the fuv - nuv colour to become redder .
however , the situation could be more complex if ` anomalous ' sub - populations , such as those suggested to be responsible for the uv - upturn of elliptical galaxies , were present
. several populations could develop hot horizontal branch ( hhb ) and agb manqu stars : very helium - rich stars such as those recently found in globular cluster multiple stellar populations ( e.g. * ? ? ?
* ) ; ultra metal - rich stars @xcite ; or normal metallicity stars with enhanced mass loss @xcite .
such populations could contribute , if not dominate , the fuv light .
the effects of such extreme populations will be the subject of a forthcoming investigation . for populations younger than 1 - 2 gyr ( depending on the metallicity ) this trend is inverted , since the turnoff starts to contribute significantly to the fuv band .
thus the ( fuv - nuv ) colour becomes steeply bluer with decreasing age .
we conclude that the fuv - nuv colour profile can not be used alone to discriminate between negative age or metallicity gradients .
we know however from studies based on absorption line indices that strong negative metallicity gradients are present in etgs .
thus , the ( fuv - nuv ) colour trend may be , at least in part , due to a metallicity effect .
three out of the five rejuvenated galaxies in the top left panel of figure [ fig5 ] ( ngc 1521 , ngc 6958 , ngc 7332 ) display redder colours than the ` old ' galaxies .
we notice that the _ galex _ ( fuv - nuv ) colour is not very sensitive to reddening for milky way - type dust ( an e(b - v)=0.3 implies a @xmath34 m @xmath130.2 ) .
thus , if the red colours were due to a reddening effect , we should have an extinction as high as e(b - v)@xmath350.5 .
it is unlikely that such a strong extinction is present over several galaxy effective radii .
for this reason , we interpret the observed colours as intrinsic .
lick indices do not show evidence for particularly low metallicities in the 5 galaxies . on the contrary ,
the metallicities are quite high in all but ngc 7332 consistent with their high velocity dispersions ( @xmath36200 km s@xmath21 ) .
thus , we suggest that the observed uv colours further support the presence of young luminosity - weighted ages in these galaxies . in the bottom left panel of figure [ fig5 ]
we plot separately the profiles of four galaxies , namely ngc 128 , ngc 2911 , ngc 5638 and ngc 7192 .
their luminosity - weighted ages range from 5.7@xmath28 2.0 ( ngc 7192 and ngc 2911 ) to [email protected] gyr ( ngc 128 ) while their metallicities are solar or twice solar ( ngc 2911 and ngc 7192 ) , the latter with a correspondingly high central velocity dispersion . their ( fuv - nuv ) colour profiles tend to be nearly flat .
three out of four of these galaxies show a very faint uv emission extending out to 2 optical r@xmath23 .
only the profile of ngc 128 extends up to 4 r@xmath23 and it is very similar to that of ngc 7332 ( top left panel ) , although much bluer ( ( fuv - nuv)@xmath131 ) in the first galaxy vs. @xmath132.2 in ngc 7332 .
both galaxies are edge - on s0 , with nearly solar metallicity . in the right bottom panel of figure [ fig5 ] ,
the colour profiles are characterized by the presence of blue rings , which appear as blue peaks in the ( fuv - nuv ) colours .
as discussed at the beginning of this section , an abrupt blueing of the uv colours can be due to very recent star formation .
notice however the presence of both rejuvenated ( ngc 3489 ) and old ( ngc 1533 , ngc 2974 ) nuclei as indicated by the ages derived in paper iii in the r@xmath23/8 region .
we examined the behaviour of the central ( fuv - nuv ) colour vs. the ` activity class ' ( see table 1 , column 11 ) .
although most of our galaxies have a liner nuclei , seyfert , composite and inactive nucleus are present ( see paper iv ) .
we find that the central uv colour is totally uncorrelated with the activity class .
the bluest ( fuv - nuv ) colours are seen in ngc 3258 and ngc 4552 nuclei , classified in paper iv as composite , from diagnostic diagrams @xcite .
ngc 3258 ( [email protected] gyr ) and ngc 4552 ( [email protected] ) could have experienced a recent star formation event .
a similar result is found by @xcite on a large etg sample drown from the sdss dr6 .
their analysis of the radial color gradient of etgs , shows that the so - called blue - cored galaxies , possible rejuvenated etgs , occupy both the star - formation and seyfert regions in the classical [ oiii]/h@xmath17 vs. [ nii]/h@xmath7 btp diagram @xcite .
these authors suggest that most blue - cored etgs have centrally concentrated star formation and a smaller fraction could be associated with seyfert activity . the fuv flux and the hi distributions correlate in star forming regions ( see e.g. * ?
* ; * ? ? ?
* ) . in this context , we investigated in the literature the presence and the morphology of the large scale hi distribution in the present etg sample . only a few objects have been observed . the hi
content of ngc 1052 has been observed with the vla by @xcite .
their figure 5 shows an irregular hi structure , elongated roughly perpendicularly to the galaxy major axis . in the southwest region of ngc 1052
, the hi distribution shows what appears to be a tidal tail , suggesting that the gas may have been acquired about 10@xmath37 years ago .
we notice that in ngc 1052 there is no correlation between the hi distribution and the far uv emission .
similar cases of tail - like hi distributions , decoupled from the far uv emission , have been detected in some shell galaxies , namely ngc 2865 , ngc 5018 and ngc 7135 @xcite .
@xcite show the large scale hi distribution in the field of ngc 1533 , one of our etgs showing a blue outer ring .
the hi distribution is connected to a companion galaxy , ic 2038 .
four outer - hii regions are detected , powered by young , low mass ob associations according to @xcite ( see their figure 7 ) .
their fuv - nuv colour , in the range @xmath38 , is bluer than the knotty regions in the galaxy ring , which we will discuss and model in a forthcoming paper @xcite .
the hi distribution has been investigated also in ngc 4374 @xcite and in ngc 5846 @xcite .
hi is detected in neither , although both have hi rich companions .
the left panels of fig .
[ fig6 ] present uv - optical colour - magnitude relations ( cmrs ) of the 14 galaxies for which sdss data are available ( see table 3 and appendix b ) . magnitudes were computed within d@xmath1 .
no clear correlations are observed in the ( fuv@xmath8nuv ) vs m@xmath39 and ( fuv@xmath8r ) vs m@xmath39 planes , while in the ( nuv@xmath8r ) vs m@xmath39 plane the positive trend is mainly driven by one object ( ngc 3489 ) at relatively faint magnitudes . in this plane ,
our galaxies are consistent with the local universe ( 0@[email protected] ) red sequence derived by @xcite ( ( [email protected])@xmath24 m@xmath39 + 0.75 ) .
we also notice that rejuvenated galaxies are mixed with older etgs in these cmrs . in the right panels of fig.6
, we plot the uv - optical colours derived within @xmath40 versus the central velocity dispersion @xmath41 .
a significant correlation is observed only in the fuv@xmath8nuv vs @xmath9 plane , but in this case the sample is larger ( 39 galaxies ) .
the spearman correlation coefficient is @xmath42 with n@xmath4337 degrees of freedom , indicating that an anti - correlation exists with a probability @xmath44 0.99 .
this result is in agreement with @xcite .
studies based on narrow - band indices have shown that the stellar populations in etgs tend to become progressively more metal rich and older with increasing velocity dispersion ( i.e. , increasing galaxy mass ) ( e.g * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
both effects could be responsible for the observed fuv@xmath8nuv trend since in general this colour becomes bluer for more metal rich and older passive populations ( see discussion in section 4.3 ) .
we have obtained _ galex _ nuv and fuv aperture and surface photometry of 40 etgs whose emission line properties have been presented in paper iii and paper iv . in appendix
b we also provide the sdss @xmath19-band surface photometry of 14 of the above etgs .
we used uv and optical imaging to study mechanisms driving both secular and/or external etg evolution such as bars and recent accretion episodes , respectively , as suggested by the young ages resulting from the line strength indices analysis ( see paper iii ) .
we find the following results : . in general , the ( fuv@xmath8nuv ) radial profiles become redder with galactocentric distance ( see also * ? ? ?
this property is common to both rejuvenated ( @xmath45 4 gyr ) and old etgs .
the trend could be due to age and/or metallicity gradients .
the presence of strong metallicity gradients in these etgs as revealed by the analysis of narrow - band indices in paper iii suggests that the radial ( fuv@xmath8nuv ) profiles are due , at least in part , to radial metallicity variations .
the role of the metallicity could be dominant if , as found by clemens et al 2009 , the age increases with galactocentric distance .
the study of moderate redshift ( 0.4 @xmath46 1.5 ) etgs performed by goods ( great observatory origins deep survey ) @xcite suggest that redshift evolution of the observed color gradients is incompatible with a significant variation in stellar age within each galaxy and that the local observations of radial color gradients mostly correspond to a range of metallicities .
relevant exceptions to what seems the ` normal ' ( fuv - nuv ) colour profile are ring / arm - like galaxies .
the nuv and fuv images of ngc 1533 , ngc 2962 , ngc 2974 , ngc 3489 and ic 5063 show that ring and/or arm - like structures are bluer than the galaxy body .
kinematical models suggest that rings have an internally driven origin and are closely associated with the evolution of the bar .
the gas may be driven both in the centre and in the ring region where accumulates and may activate star formation .
the young luminosity - weighted age of ngc 3489 and its agn - like activity may be explained in the framework of a gas inflow toward the galaxy centre . in paper
iv we found evidence that star formation and agn activity are closely connected in time .
we interpret the abrupt blueing of the uv colour in the ring / armlike structure as due to recent star formation .
detailed models are planned in a subsequent work .
we consider these etgs as possible candidates of ongoing secular evolution .
shell galaxies are a prototypical example of the externally driven evolution of etgs .
out of the seven etgs in the sample with stellar shells detected in the optical , ngc 7135 is the only shell galaxy showing a shell structure in the nuv and fuv images .
@xcite suggest that this galaxy could have had a recent nuclear star formation episode as implied by the low age , likely triggered by the accretion event that formed the shell .
the sersic profile shape parameter @xmath3 shows much variation in uv ( 1 @xmath47 ) as in the optical and near infrared bands . s0s
tend to have values of @xmath48 in both nuv and fuv , as already noticed in the optical .
the sersic index @xmath5 is a sort of watershed as noted in the optical by @xcite .
most etgs with @xmath6 have [ @xmath7/fe ] @xmath35 0.15 , implying a short star - formation time scale .
etgs with n=3@xmath281 have a large spread in [ @xmath7/fe ] , including very low values indicative of a more prolonged star formation .
most of these galaxies are s0s , i.e truly disk galaxies , which , as noticed in paper iii , generally appear more rejuvenated with respect to es .
these galaxies have also , in general , a low central velocity dispersion .
we suggest that the sersic index vs. [ @xmath7/fe ] trend reflects the velocity dispersion ( i.e. mass ) @xmath7-enhancement relation . a significant correlation between the ( fuv@xmath8nuv ) colour and @xmath9
is observed , in agreement with previous studies @xcite .
this trend is likely to be driven by a combined effect of the age-@xmath9 and metallicity-@xmath9 relations found from narrow band indices .
a more detailed investigation on the origin of this relation will be performed in a forthcoming paper .
a detailed analysis of the uv colour vs. line - strength indices and of the etgs spectral energy distribution from uv to optical and mir will be presented in forthcoming papers .
we acknowledge mauro donofrio for his support , valuable discussions and comments . am acknowledges support from italian scientist and scholars in north america foundation ( issnaf ) through an issnaf fellowship in space physics and engineering , sponsored by thales alenia space .
rr , ab , fa , lmb acknowledge the asi - inaf support through contract i/016/07/0 .
_ galex _ is a nasa small explorer , launched in april 2003 .
_ galex _ is operated for nasa by california institute of technology under nasa contract nas-98034 .
this work was partly supported by nasa grant nnx07apo8 g .
this research has made use of saoimage ds9 , developed by smithsonian astrophysical observatory and 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 .
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 .
we acknowledge the usage of the hyperleda database ( http://leda.univ-lyon1.fr ) .
+ _ facilities : _ _ galex _ , sdss adelman - mccarth , j.k . , ag " ueros , m.a . , allam , s.s . , allende prieto , c. , anderson , k. s. j. et al .
2008 , apjs , 175 , 297 andredakis , y.c . ,
peletier , r.f . ,
balcells , m. 1995 , mnras , 275 , 874 annibali , f. , bressan , a. , rampazzo , r. , zeilinger , w.w .
2006 , aa , 445 , 79 : paper ii annibali , f. , bressan , a. , rampazzo , r. , zeilinger , w.w . ,
danese , l. 2007 , aa , 463 , 455 : paper iii annibali , f. , bressan , a. , rampazzo , r. , zeilinger , w.w .
, vega , o. , panuzzo , p. 2010 , aa , submitted : paper iv athanassoula , e. 2009 , in `` galaxies in isolation : exploring nature vs. nurture '' ( 2009arxiv0910.5180a ) baldwin , j. a. , phillips , m. m. , terlevich , r. 1981 , , 93 , 5 bertola , f. , bressan , a . , burstein , d. , buson , l. m. , chiosi , c. , & di serego alighieri , s. 1995 , , 438 , 680 bianchi , l. et al .
2007 , in `` galaxy evolution across the hubble time '' iau sympo .
235 , f. combes and j. palous eds .
, p. 301
bianchi , l. 2009 , ap&ss , 320 , 11 bournaud , f. , combes , f. , semelin , b. 2005 , mnras , 364 , l18 bressan , a. , chiosi , c. , fagotto , f. 1994 , apjs , 94 , 63 brown , r.j.n .
, forbes , d.a . , silva ,
d. , helsdon , s.f .
, ponman , t.j .
2003 , mnras , 341 , 747 .
burstein , d. , bertola , f. , buson , l.m . ,
faber , s.m . , lauer , t.r .
1988 , apj , 328 , 440 buta , r. , combes , f. 1996 , fundamentals of cosmic physics , volume 17 , 95 .
caon , n. , capaccioli , m. , donofrio , m. 1993 , mnras .
265 , 1013 cappellari , m. et al .
2006 , mnras , 366 , 1126 cardelli , j.a . ,
clayton , g.c . ,
mathis , j.s .
1989 , apj , 345 , 245 chapman , s. c. , smail , i. , blain , a. w. , ivison , r. j. 2004 , apj , 614 , 671 clemens , m. s. , bressan , a. , nikolic , b. , alexander , p. , annibali , f. , & rampazzo , r. 2006 , , 370 , 702 clemens , m. s. , bressan , a. , nikolic , b. , rampazzo , r. 2009,mnras , 392 , l35 coenda , v. , donzelli , c.j . , muriel , h. , quintana , h. , infante , l. , lambas , d.g .
2005 , aj , 129 , 1237 colbert , j.w .
, mulchaey , j.s . , zabludoff , a.i .
2001 , aj , 121 , 808 de lucia , g. springel , v. white , s.d .
m. , croton , d. kauffmann , g. 2006 , mnras , 366 , 499 de vaucouleurs , g. , de vaucouleurs , a. , corwin , h.g .
1991 _ third reference catalogue of bright galaxies _ , springer - verlag , new york donas , j. deharveng , j - m . ,
rich , r.m , , yi , s.k . , lee , y - w , et al . 2007 , apjs , 173 , 597 ebeling , h. , white , d. a. , rangarajan , f. v. n. 2006 , mnras , 368 , 65 emsellem , e. , cappellari , m. , krajnovi , d. , van de ven , g. , bacon , r. , et al .
2007 , mnras , 379 , 401 .
ferreras , i. , lisker , t. , pasquali , a. , kaviraj , s. 2009 , mnras , 395 , 554 franx , m. , labb , i. , rudnick , g. , van dokkum , p.g . , daddi , e. et al .
2003 , apj , 587 , l79 gil de paz a. boissier , s. madore , b.f .
, seibert , m.j . , young h. ; boselli , a. et al .
2007 , apjs , 173 , 185 goudfrooij , p. 1999 , in _ star formation in early type galaxies _ , asp conference series 163 , eds .
j. cepa and p. carral , 55 grossi , m. , di serego alighieri , s. , giovanardi , c. et al .
2009 , a&a , 498 , 407 heckman , t. m. 1980 , , 87 , 152 kewley l. j. , dopita , m.a . ,
sutherland , r.s . ,
heisler , c.a . ,
trevena , j. , 2001 , apj , 556 , 121 kewley , l. j. , groves , b. , kauffmann , g. , & heckman , t. 2006 , , 372 , 961 kormendy , j. 2004 , ara&a , 42 , 603 kormendy , j. , fisher , d. b. , cornell , m. e. , & bender , r. 2009 , , 182 , 216 krajnovi ' c , d. , cappellari , m. , emsellem , e. , mcdermid , r.m . , de zeeuw , t.p .
2005 , mnras , 357 , 1113 jedrzejewski , r. , 1987 , mnras , 226 , 747 jeong , h. , bureau , m. , yi , s. , krainovic , d. , davies , r.l .
2007 , mnras , 376 , 1021 jeong , h. , yi , s. , bureau , m. , davies , r. , falcn - barroso , j. et al .
2009 , mnras , 398 , 2028 laurikainen , e. , salo , h. , buta , r. , speltincx , t. , block d. 2006 , aj , 132 , 2634 li , y. , van gorkom , j.h .
2001 in gas & galaxy evolution , j.e .
hibbard , m.p .
rupen , and j.h .
van gorkom eds .
, asp conf .
, vol 240 , pag
. 637 macchetto , f. , pastoriza , m. , caon , n. , sparks , w.b . , giavalisco , m. et al .
1996 , aas 120 , 463 malin , d.f .
, carter , d. 1983 , apj , 274 , 534 marino , a. , iodice , e. , tantalo , r. , piovan , l. , bettoni , d. , et al .
2009 , a&a , 508 , 1235 marino , a. , bianchi , l. , rampazzo , r. et al .
2010 , to appear in
uv universe 2010 " proceedings , st .
petersburg june , 2010 martin , d.c . ,
fanson , j. , schiminovich , d. et al .
2006 , apj , 619 , l1 mcdermid , r.m . , bacon , r. , kuntschner , h. , emsellem , e. shapiro , k.l .
et al . 2006 , new ar , 49 , 521 morrissey , p. , conroy , t. , barlow , t.a . , et al .
2007 , apjs , 173 , 682 neff , s. g. , thilker , d. a. , seibert , m. , gil de paz , a. , bianchi , l. , et al .
2005 , apj , 619 , l91 phillips , m. , jenkins , c. , dopita , m. , sadler , e.m . ,
binette , l. 1986 , aj , 91 , 1062 piotto , g. 2008 , memorie della societa astronomica italiana , 79 , 334 prugniel , ph . , simien , f. 1997 , aa , 321 , 111 rampazzo r. , annibali f. , bressan a. , longhetti m. , padoan f. , zeilinger w.w .
2005 , aa 433 , 497 : paper i rampazzo , r. , marino , a. , tantalo , r. , bettoni , d. , buson , l. m. , chiosi , c. , galletta , g. , gr " utzbauch , r. , rich , r. m. 2007 , mnras , 381 , 245 reduzzi , l. , longhetti , m. , rampazzo , r. 1996 , mnras , 282 , 149 rogers , b. , ferreras , i. lahav , o. , bernardi , m. , kaviraj , s. yi , s.k . 2007 , mnras , 382 , 750 sandage , a.r . ,
brucato , r. 1979 , aj , 84 , 472 sandage , a.r . ,
tammann , g. 1987 , _ a revised shapley ames catalogue of bright galaxies _ , carnegie , washington ( rsa ) sarzi , m. , falc ' on - barroso , j. , davies , r.l . , bacon , r. et al .
2006 , mnras , 366 , 1151 schawinski , k. , kaviraj , s. , khochfar , s. , yoon , s .- j . , yi , s. k. et al .
2007 , apjs , 173 , 512 sersic , j. l. 1968 , atlas de galaxias australes , osservatorio astronomico , cordoba suh , h. , jeong , h. , oh , k. , yi , s. , ferreras , i. , schawinski , k. 2010 , apj , 187 , 374 tal , t. van dokkum , p.g . , nelan , j. , bezanson , r. 2009 , aj , 138 1417 terlevich , e. , diaz , a.i . , terlevich , r. 1990 , revista mexicana de astronomia y astrofisica , 21 , 218 thilker d.a . , bianchi , l. , meurer , g. , gil de paz , a. boissier , s. et al .
2007 , apjs , 173 , 538 thilker , d. a. , et al .
2005 , , 619 , l79 thomas , d. , maraston , c. , bender , r. , & mendes de oliveira , c. 2005 , , 621 , 673 treu , t. , ellis , r.s . , liao , t.x . , van dokkum , p.g . 2005 , apj , 622 , l5 tully , r.b . 1988 , _ nearby galaxy catalogue _ , cambridge university press van gorkom , j.h . ,
knapp g.r . , raimond , e. , faber , s.m . ,
gallagher , j.s .
1986 , aj , 91 , 791 wada , k. 2004 , in coevolution of black holes and galaxies , ed . l.c .
ho , cambridge univ .
press , 186 werk , j.k . , putman , m.e .
, meurer , g.r .
, ryan - weber , e.v .
, kehrig , c. et al .
2010 , aj , 139 , 279 yan , r. , newman , j. a. , faber , s. m. , konidaris , n. , koo , d. , & davis , m. 2006 , , 648 , 281 yi , s. , demarque , p. , & oemler , a. , jr .
1997 , , 486 , 201 yi s.k . , yoon , s .- j . ,
kaviraj , s. , deharveng , j .-
m . , rich , r. m. , et al . 2005 , apj , 619 , l111 zabludoff , a. 2001 in gas & galaxy evolution , j.e . hibbard , m.p .
rupen , and j.h .
van gorkom eds .
, asp conf .
ser . , vol 240 , pag .
we briefly comment here on the _ galex _ and sdss surface brightness profiles of the etgs presented in this paper . we also list morphological and kinematic peculiarities which could be relevant for understanding the galaxy evolution .
for a detailed information about the age , metallicity and @xmath7-enhancements as well as about emission line properties we cross - refer to papers iii and papers iv ( these data are summarized in table 1 ) . the basic information about etgs is provided in papers i , ii ( on line material ) and iv ( appendix a ) .
_ ngc 128 _
this edge - on s0 is a liner .
the ( fuv - nuv ) colour profile is flat within the errors ( average @[email protected] ) .
the average sersic index is @xmath49 = [email protected] .
gas and stars counter rotate in this about 10 gyr old fast rotator .
_ ngc 777 _ the ( fuv - nuv ) colour profile drops from @xmath130.5 in the centre of the galaxy to @xmath131.5 in the outskirts .
this slow rotator , seyfert - like galaxy has a young ( [email protected] ) age in the nucleus .
the average sersic index is @xmath49= [email protected] : the quality of the fit is indeed quite poor .
_ ngc 1052 _
this old ( @xmath1314 gyr ) , fast rotating galaxy is considered a prototypical liner .
a gas vs. gas counter rotation is reported for this galaxy . as in previous galaxy the ( fuv - nuv ) colour profile becomes red going from the centre ( @xmath130.75 ) to the outskirts ( @xmath131.75 ) .
( nuv-@xmath19 ) becomes slightly bluer from @xmath130.55 to 0.5 while ( fuv-@xmath19 ) redder from the centre @xmath136.2 to 6.9 .
the average sersic index is @xmath49= [email protected] .
the large spread in the @xmath3 indices does not depend from the quality of the fit .
appendix a is available in its entirety in the online edition of the journal .
a small portion is shown here for guidance .
figure a1 displays composite sdss images available for 14 etgs in our sample . figure a2 shows the surface brightness profiles in the @xmath19 band and the ( fuv-@xmath19 ) and ( nuv-@xmath19 ) colour profiles . | we present galex far - ultraviolet ( fuv , @xmath0=1538 ) and near - ultraviolet ( nuv , @xmath0=2316 ) surface photometry of 40 early - type galaxies ( etgs ) selected from a wider sample of 65 nearby etgs showing emission lines in their optical spectra .
we derive fuv and nuv surface brightness profiles , ( fuv - nuv ) colour profiles and d@xmath1 integrated magnitudes .
we extend the photometric study to the optical _ r _ band from sdss imaging for 14 of these etgs . in general , the ( fuv - nuv )
radial colour profiles become redder with galactocentric distance in both rejuvenated ( @xmath2 gyr ) and old etgs .
colour profiles of ngc 1533 , ngc 2962 , ngc 2974 , ngc 3489 , and ic 5063 show rings and/or arm - like structures , bluer than the body of the galaxy , suggesting the presence of recent star formation .
although seven of our etgs show shell systems in their optical image , only ngc 7135 displays shells in the uv bands .
we characterize the uv and optical surface brightness profiles , along the major axis , using a sersic law .
the sersic law exponent , @xmath3 , varies from 1 to 16 in the uv bands .
s0 galaxies tend to have lower values of @xmath3 ( @xmath4 ) .
the sersic law exponent @xmath5 seems to be a watershed : etgs with @xmath6 tend to have [ @xmath7/fe ] greater than 0.15 , implying a short star - formation time scale .
we find a significant correlation between the fuv@xmath8nuv colour and central velocity dispersions @xmath9 , with the uv colours getting bluer at larger @xmath9 .
this trend is likely driven by a combined effect of ` downsizing ' and of the mass - metallicity relation .
= -1.3 cm [ firstpage ] galaxies : galaxies : elliptical and lenticular , cd galaxies : photometry galaxies : fundamental parameters galaxies : formation galaxies : evolution |
we gather u.s .
mortality rates from www.cdc.gov@xmath9 and aggregate search data of u.s .
citizens from www.google.com@xmath1 .
the mortality data is annual , and google search data is weekly , so in order to make google search data annual , we take the yearly average . then , we want to measure the change in mortality correlated with change in average search volume , because we want to know if next years mortality rates will increase or decrease , and by how much .
we then take logs of the differenced data since it is a monotone transformation and reduces variance in the data ( the purpose of this paper is to show if correlations exist - further methods can be employed to find how strong they are ) . + .
the line outside of the significant band at @xmath10 shows that there is a significant correlation at lag @xmath10.,width=288 ] from figure 1 suggests a correlation between how mortality moves and how hate moves one year in the future .
the time series plot checks out with figure 1 and suggests that hate should increase at time 6 to 7 , since mortality rate increased from time 5 to 6 . ,
width=288 ] table 1 shows where we found significant cross - correlations and at what lag .
an x denotes that there were no significant lags .
the significant lags at @xmath10 for all people and whites for foreclosure and hate suggests that mortality rates can predict how much people search for those words .
the black female data had very strange results , which may suggest that their mortality data in the period that we have is an outlier compared to how their mortality rates normally behave .
the only positive lag correlation was with black females and `` ship , '' which , in theory , could suggest that when more people are searching for `` ship , '' more black females are dying , but since the black female data produced strange results , it is most likely that their data is unreliable .
more data should be needed to help with this claim .
+ unfortunately there are no other significant lags with a positive value , suggesting that none of these words can help predict mortality rates . on the other hand , since for both `` foreclosure '' and `` hate ''
we found significance at lag @xmath10 , it may suggest psychological or sociological evidence into how humans react to financial failure and death . perhaps more people search for `` hate '' when white people die , but a black person s death has no effect on how often the word `` hate '' is searched for .
.table indicating which lags produced statistically significant results .
= male , and = female .
where applicable , the superscripts indicate the sign of the correlation . [ cols="<,^,^,^,^,^,^,^,^,^",options="header " , ]
the data was very limited which caused many problems .
mortality data was annual , and the latest data point was for 2010 .
google search data only went back to 2004 , so we had 7 points of data , and differencing these gave us only 6 changes .
we are also doing multiple testing at once , so we should account for some sort of test correction to handle the probability of type i error , such as the bonferroni test correction .
there are essentially 81 tests being performed , however , so in order to get some preliminary results we forego these test corrections , as the purpose of this paper is to demonstrate how this method can be used to find correlations and show that there may be significance .
in addition , reducing the number of tests is also possible , since we divided up the analysis into several groups races and sexes . +
this method could be repeated removing outlier data to possibly improve the result .
for example , the black women data acted strangely , which may suggest the data we had for this time period did not follow normal mortality trends .
if we remove black women from the overall mortality data , we may get more accurate conclusions .
+ in addition , in the future our analytical methods are more practical .
for example , 10 years from now we will have more than double the data points we had previously , causing for a huge improvement in accuracy and potential results .
in addition , if more detailed mortality data exists , we can test for correlations between certain words and age groups . as an example
, we could find positive correlation between people searching for `` alcohol '' and deaths for people in the age range of 15 - 25 .
+ although it will probably not be possible , if mortality data was available weekly , it would be tremendous for data analysis , since we will have 52 times as many data points .
we did find significant evidence at negative lags even though our data provided no significant conclusions for positive lag cross - correlations , this opens two new doors .
the first being that maybe we havent found the right word , and given the correct word we can predict mortality rates .
the second being that we did find significant evidence for negative lags , providing psychological or sociological insight as to how mortality rates affect our searching on google .
further work can definitely be done to find correlations with mortality , but even aside from that it provides more evidence as to how our lives affects google search data and how google search data affects our lives .
1 . preis , t. , d. reith , and h. e. stanley .
`` complex dynamics of our economic life on different scales : insights from search engine query data . ''
_ philosophical transactions of the royal society a : mathematical , physical and engineering sciences _ 368.1933 ( 2010 ) : 5707 - 719
centers for disease control and prevention . _ centers for disease control and prevention . web .
27 apr . 2012 .
< http://www.cdc.gov/>. 3 .
`` google insights for search . ''
_ google . _ web .
+ < http://www.google.com / insights / search/>. | inspired by correlations recently discovered between google search data and financial markets@xmath0 , we show correlations between google search data mortality rates .
words with negative connotations may provide for increased mortality rates , while words with positive connotations may provide for decreased mortality rates , and so statistical methods were employed to determine to investigate further .
tatistically significant correlations between google search data and financial markets exist@xmath0 .
the methods used can be used to attempt to find correlations between any time series data and google@xmath1 searches . since
insurance premiums are based on mortality rates , if one can predict mortality rates with a point or interval estimator , an insurance company can re - evaluate a policy every year and update the premium based on what search data may suggest in addition to standard pricing methods . + as google search data and mortality rates are time series data , we can use cross correlations ( defined below ) , to determine if correlations exist and we can show where they exist . using r , we can use the ccf ( ) ( cross - correlation function ) command to easily graph and find significant values depending on a significance level ( we choose a liberal @xmath2 ) . in figure 1 , we graph a cross correlation function of all races both sexes and the word hate as demonstration .
these were the graphs used in the statistical analysis of the project .
figure 2 shows how the significant cross - correlation relates to how the graphs work .
+ nine words were used for testing .
these words are chosen somewhat arbitrarily , but with intentions to hopefully be correlated with mortality .
for example , the word `` hate '' was used and may imply that if more people are hateful ( increased search for hate ) then mortality rates may increase ( due to hate related crimes , for example ) .
+ in addition , we search nine groups of mortality rates . there is a generic group ( all sexes all races ) and then groups divided into male and female , and racial groups divided into white and black .
the reasoning for this is that it will both let us determine if there is any outlier data , and will also let us see if there are correlations inside of a specific group .
it is possible that one race or one sex may react differently to a certain word based on their culture . + as discussed above , we present the cross correlation function : @xmath3-e[x(t])e[y(t+\delta
t)]}{\sqrt{e[x^2(t)]-e[x(t)]^2}\sqrt{e[y^2(t+\delta t)]-e[y(t+\delta t)]^2}}.\ ] ] essentially it checks for relationships between two time series @xmath4 and @xmath5 , but at different times , for example if @xmath6 , it will check for @xmath4 s relationship with @xmath7 s relationship .
the denominator simply scales the numerator so that @xmath8 . |
a bosonic system with @xmath157 modes is a quantum system described by a tensor - product hilbert space @xmath158 and a vector of quadrature operators @xmath159satisfying the commutation relations @xcite@xmath160=2i\mathbf{\omega } ~ , \label{commquad}\]]where @xmath161 is a symplectic form in @xmath162 , i.e.,@xmath163by definition a quantum state @xmath16 of a bosonic system is called gaussian when its wigner phase - space representation is gaussian @xcite . in such a case ,
the state is completely described by the first and second statistical moments . in other words , a gaussian state @xmath16 of @xmath157 bosonic modes
is characterized by a displacement vector @xmath164and a correlation matrix ( cm ) @xmath165where @xmath166 denotes the anticommutator @xcite .
the cm is a @xmath167 real and symmetric matrix which must satisfy the uncertainty principle @xcite@xmath168an important example of gaussian state is the tmsv state for two bosonic modes @xmath169 ( whose number - ket representation is given in the letter ) .
this state has zero mean ( @xmath170 ) and its cm is given by@xmath171where @xmath172 and@xmath173here @xmath174 represents the mean number of thermal photons which are present in each mode .
this number is connected with the two - mode squeezing parameter @xcite by the relation@xmath175parameter @xmath50 completely characterizes the state ( therefore denoted by @xmath176 ) and quantifies the entanglement between the two modes @xmath177 and @xmath178 ( being an entanglement monotone for this class of states ) .
another important example of gaussian state is the ( multimode ) coherent state . for @xmath179 modes this
is given by@xmath180where @xmath181 is a row - vector of amplitudes @xmath182 .
this state has cm equal to the identity and , therefore , is completely characterized by its displacement vector @xmath183 , which is determined by @xmath184 .
starting from the coherent states , we can characterize all the possible states of a bosonic system by introducing the p - representation .
in fact , a generic state @xmath16 of @xmath179 bosonic modes can be decomposed as@xmath185where @xmath186 is a quasi - probability distribution , i.e. , normalized to @xmath71 but generally non - positive ( here we use the compact notation @xmath187 ) . by definition
, a bosonic state @xmath16 is called classical if @xmath188 is positive , i.e. , @xmath189 is a proper probability distribution .
by contrast , the state is called non - classical when @xmath186 is non - positive .
it is clear that a classical state is separable , since eq .
( [ pdecomp ] ) with @xmath189 positive corresponds to a state preparation via local operations and classical communications ( loccs ) .
the borderline between classical and non - classical states is given by the coherent states , for which @xmath189 is a delta function .
note also that the classical states are generally non - gaussian .
in fact , one can have a classical state given by a _ finite ensemble _ of coherent states ( whose @xmath189-representation corresponds to a sum of dirac - deltas ) . in general
we use the formalism @xmath17 to denote a bipartite transmitter or , more simply , a transmitter .
this is a compact notation for characterizing simultaneously a bipartite bosonic system and its state .
more specifically , @xmath7 and @xmath9 , represent the number of modes present in two partitions of the system , called signal ( sub)system @xmath6 and idler ( sub)system @xmath8 .
then , given this bipartite system , @xmath16 represents the corresponding global state .
this notation is very useful when both system and state are variable . by definition ,
a transmitter @xmath17 is classical ( non - classical ) when its state @xmath16 is classical ( non - classical ) , i.e. , @xmath190 and @xmath20 . in the comparison between different transmitters
, one has to fix some of the parameters of the signal ( sub)system @xmath6 .
the basic parameters of @xmath6 are the total number of signal modes @xmath7 ( signal bandwidth ) and the total number of signal photons @xmath191 ( signal energy ) .
a gaussian channel is a completely positive trace - preserving ( cptp ) map @xmath192which transforms gaussian states into gaussian states .
in particular , a one - mode gaussian channel ( i.e. , acting on one - mode bosonic states ) can be easily described in terms of the statistical moments @xmath193 .
this channel corresponds to the transformation holevocan , characatt , alexjens@xmath194where @xmath195 is an @xmath196-vector , while @xmath197 and @xmath198 are @xmath199 real matrices , with @xmath200 and @xmath201important examples of one - mode gaussian channels are the following : ( i ) : : the attenuator channel @xmath202 with loss @xmath203 $ ] and thermal noise @xmath204 .
this channel can be represented by a beam splitter with reflectivity @xmath205 which mixes the input mode with a bath mode prepared in a thermal state with @xmath91 mean photons .
this channel implements the transformation of eq .
( [ gauss_trasf ] ) with@xmath206and @xmath207 .
in particular , when the thermal noise @xmath91 is negligible , the attenuator channel @xmath208 is represented by a beam splitter with reflectivity @xmath205 and a vacuum bath mode .
( ii ) : : the thermal channel @xmath209 adding gaussian noise with variance @xmath210 .
this channel implements the transformation of eq .
( [ gauss_trasf ] ) with@xmath211and @xmath207 .
this kind of channel is suitable to describe the effects of decoherence when the loss is negligible .
this is a typical situation within optical apparatuses which are small in size ( as is the case of our memory reader ) .
it is clear that , for every pair of gaussian channels , @xmath212 and @xmath213 , acting on the same state space , their composition @xmath214 is a also gaussian channel .
if @xmath215 and @xmath213 are gaussian channels acting on two different spaces , their tensor product @xmath216 is also gaussian .
hereafter we call bipartite gaussian channel a gaussian channel which is in the tensor - product form @xmath217 .
the action of a bipartite gaussian channel on a two - mode bosonic state is very simple in terms of its second statistical moments .
in fact , let us consider two bosonic modes , @xmath218 and @xmath219 , in a state @xmath220 with generic cm@xmath221where @xmath222 @xmath223 and @xmath224 are @xmath199 real matrices . at the output of a bipartite gaussian channel @xmath225 , we have the cm@xmath226where the matrices @xmath227
refer to @xmath228 ( acting on the first mode ) , while @xmath229 refer to @xmath230 ( acting on the second mode ) .
* both the channels @xmath231 are dilated , so that we have@xmath232 , \notag \\ & \label{double_dil}\end{aligned}\]]where @xmath233 and @xmath234 are supplementary sets of bosonic modes prepared in vacua , while @xmath235 and @xmath236 are gaussian unitaries .
according to eq .
( [ double_dil ] ) , the input cm @xmath237 is subject to three subsequent operations .
first , it must be dilated to @xmath238 , where @xmath239 and @xmath240 are identity matrices of suitable dimensions .
then , it is transformed via congruence by @xmath241 , where @xmath242 and @xmath243 are the symplectic matrices corresponding to @xmath235 and @xmath244 , respectively .
finally , raws and columns corresponding to @xmath245 and @xmath234 are elided ( trace ) . by setting@xmath246we get@xmath247now , by setting @xmath248and@xmath249we
get exactly eq .
( [ v_out_bip_ch ] ) .
@xmath250 . *
inset b. * thermal - loss model , implementing the bipartite gaussian channel @xmath251 ( see text).,scaledwidth=48.0% ]
let us consider the beam - splitter scheme of fig .
[ bs]a , where an input state @xmath16 of two modes ( signal mode@xmath177 and idler mode @xmath178 ) is transformed into an output state @xmath252 .
hereafter we call this scheme the pure - loss model .
the corresponding transformation is a bipartite gaussian channel @xmath253where the attenuator channel @xmath254 acts on the signal mode , and the identity channel @xmath33 acts on the idler mode . by construction
, @xmath255 is a memoryless channel .
thus , if we consider a transmitter @xmath256 , i.e. , @xmath7 signal modes @xmath47 and @xmath9 idler modes @xmath48 in a multimode state @xmath16 , the input state is transformed into the output state @xmath257where@xmath258this is schematically depicted in fig .
[ scenario1 ] . ) .,scaledwidth=47.0% ] now , encoding a logical bit @xmath259 in the reflectivity of the beam splitter @xmath260 corresponds to encoding the bit into the conditional attenuator channel @xmath261 .
thus , reading a memory cell using an input transmitter @xmath17 , and an output receiver , corresponds to the channel discrimination problem depicted in fig .
the error probability in the decoding of the logical bit , i.e. , @xmath262corresponds to the error probability of discriminating between the two equiprobable channels @xmath263 and @xmath264 .
the minimization of this error probability involves the optimization of both input and output . for a _ fixed _
input @xmath17 , we have two equiprobable output states , @xmath265 and @xmath266 , with@xmath267the optimal measurement for their discrimination is given by the dichotomic povm @xcite@xmath268where @xmath269 is the projector onto the positive part @xmath270 of the helstrom matrix @xmath271 @xcite . using this measurement , the two output states , @xmath265 and @xmath272 , are discriminated with a _ minimum _ error probability which is provided by the helstrom bound , i.e. , @xmath273where @xmath274 is the trace distance between @xmath275 and @xmath266 @xcite . in
particular,@xmath276where @xmath277 is the trace norm .
thus , given two equiprobable channels @xmath278 and a _ fixed _ input transmitter @xmath18 , we can always assume an optimal output detection . under this assumption ( optimal detection ) ,
the channel discrimination problem @xmath279 with fixed transmitter @xmath280 is fully characterized by the conditional error probability@xmath281 _ { \sigma _ { u}=\left ( \mathcal{e}_{u}^{\otimes m}\otimes \mathcal{i}^{\otimes l}\right ) ( \rho ) } .
\notag \\ & & \end{aligned}\]]now , the minimal error probability for discriminating @xmath263 and @xmath282 is given by optimizing the previous quantity over all the input transmitters , i.e.,@xmath283 in general , this error probability tends to zero in the limit of infinite energy @xmath284 ( this happens whenever @xmath279 ) . for this reason , in order to consider a non - trivial quantity , we must fix the signal - energy @xmath11 in the previous minimization .
let us denote by @xmath285 the class of transmitters signalling @xmath11 photons .
then , we can define the conditional error probability@xmath286it is an _
open question _ to find the optimal transmitter within the class @xmath285 , i.e. , realizing the minimization of eq .
( [ optimization2 ] ) .
the central idea of our letter is a direct consequence of this open question .
in fact , strictly connected with this question , there is another fundamental problem , whose resolution sensibly narrows the search for optimal transmitters : within the conditional class @xmath285 , can we find a non - classical transmitter which outperforms any classical transmitter ?
more exactly , given two attenuator channels @xmath287 and fixed signal - energy @xmath11 , can we find any @xmath38 such that @xmath288for every @xmath40 ?
in our letter we solve this problem @xcite .
in particular , we show this is possible for an important physical regime , i.e. , for channels @xmath287 corresponding to high - reflectivities ( as typical of optical memories ) and signals with few photons ( as typical of entanglement sources ) . in general , the pure - loss model of fig .
[ bs]a represents a very good description in the optical range if we assume the use of a good reading apparatus . to complete the analysis and show the robustness of the model with respect to decoherence
, we also consider the presence of thermal noise , as explicitly stated in the letter .
the noisy scenario is the one depicted in fig .
[ bs]b , that we call the thermal - loss model .
this corresponds to the bipartite gaussian channel @xmath289where @xmath290acts on the signal mode , and @xmath291acts on the idler mode .
exactly as before , @xmath292 represents a memoryless channel . as a result
, the state @xmath16 of an input transmitter @xmath256 is transformed into an output state @xmath293 , where@xmath294this transformation is depicted in fig .
[ scenario3 ] .
now , encoding a logical bit @xmath259 in the reflectivity of the beam splitter @xmath260 corresponds to encoding the bit into the gaussian channel @xmath295 .
it follows that the readout of our memory cell corresponds to the channel discrimination problem depicted in fig .
[ scenario4 ] . ) .,scaledwidth=47.0% ] it is clear that the present thermal scenario ( fig . [ scenario4 ] ) can be formally achieved by the previous non - thermal one ( fig .
[ scenario2 ] ) via the replacements@xmath296 in this case , the idlers are subject to a non - trivial decoherence channel @xmath297 , which plays an active role in the discrimination problem . in general , this problem can be formulated as the discrimination of two _ bipartite _ channels , @xmath298 and @xmath299 , using a transmitter @xmath280 and an optimal receiver .
the corresponding error probability is defined by@xmath300 _
{ \sigma _ { u}=\left ( \mathcal{s}_{u}^{\otimes m}\otimes \mathcal{d}^{\otimes
l}\right ) ( \rho ) } .
\notag \\ & & \end{aligned}\]]clearly this problem is more difficult to study . for this reason , our basic question becomes the following : given two bipartite channels @xmath301 and fixed signal - energy @xmath11 , can we find any @xmath38 such that @xmath302for a _ suitable large class _ of @xmath40 ? as stated in the letter ( and explicitly shown afterwards ) , we can give a positive answer to this question too .
this is possible by excluding broadband classical transmitters which are not meaningful for the model ( see the letter for a physical discussion ) . in the following secs .
[ app_class_discr ] and [ app_epr_tra ] we introduce the basic tools that we need to compare classical and non - classical transmitters , in both the models : pure - loss and thermal - loss .
in particular , sec .
[ app_class_discr ] provides one of the central results of the work : the classical discrimination bound , which enables us to bound all the classical transmitters .
then , sec .
[ app_epr_tra ] concerns the study of the non - classical epr transmitter .
at that point we have all the elements to make the comparison , i.e. , replying to the questions in eqs .
( [ q1 ] ) and ( [ q2 ] ) .
this comparison is thoroughly discussed in sec .
in particular , for the pure - loss model , we can also derive analytical results , as shown in sec .
[ app_pure_loss ] .
these results are the threshold energy theorem and the ideal memory theorem .
let us consider the discrimination of two bipartite gaussian channels @xmath303and@xmath304by using a bipartite transmitter @xmath280 which signals @xmath11 photons .
the global input state can be decomposed using the following bipartite @xmath305-representation @xcite @xmath306 where @xmath307 is a vector of amplitudes for the signal @xmath7-mode coherent - state@xmath308and @xmath309 is a vector of amplitudes for the idler @xmath9-mode coherent - state @xmath310the reduced state for the @xmath7 signal modes is given by@xmath311where@xmath312the total ( mean ) number of photons in the @xmath7-mode signal system can be written as@xmath313where@xmath314 now , conditioned on the value of the bit ( @xmath2 or @xmath71 ) , the global output state at the receiver can be written as@xmath315where@xmath316and@xmath317assuming an optimal detection of all the output modes ( signals and idlers ) , the error probability in the channel discrimination ( i.e. , bit decoding ) is given by@xmath318 ~,\]]where @xmath265 and @xmath266 are specified by eq .
( output_sigmau ) for @xmath259 .
the study of this error probability can be greatly simplified in the case of classical transmitters .
in fact , by assuming a classical transmitter @xmath190 , the @xmath189-representation of eq .
( prepres ) is positive . as a consequence
, the error probability can be lower - bounded by a quantity which depends on the signal parameters only , i.e.,@xmath319 in other words , for fixed @xmath320 and @xmath321 , we can write a lower bound which holds for all the classical transmitters emitting signals with bandwidth @xmath7 and energy @xmath11 .
this is the basic result stated in the following theorem .
[ theo_forms ] let us consider the discrimination of two bipartite gaussian channels , @xmath322 and @xmath323 , by using a classical transmitter @xmath19 which signals @xmath11 photons .
the corresponding error probability @xmath324 is lower - bounded ( @xmath325 ) by@xmath326where @xmath327 is the fidelity between @xmath328 and @xmath329 , the two outputs of a single - mode coherent state @xmath330 with @xmath107 mean photons . in particular , in terms of all the parameters , we have@xmath331where @xmath114 and @xmath332 are defined by@xmath333 \geq 1~ , \label{omega_theo}\]]and @xmath334with@xmath335 * proof*. the two possible states , @xmath265 and @xmath266 , describing the whole set of output modes ( signals and idlers ) are given by eq .
( [ output_sigmau ] ) , under the assumption that @xmath336 is positive . in order to lowerbound the error probability@xmath337 ~ , \label{eq_pclass_proof}\]]we upperbound the trace distance @xmath274 .
since @xmath338 is a proper probability distribution , we can use the joint convexity of the trace distance @xcite . using this property , together with the stability of the trace distance under addition of systems @xcite
, we get @xmath339 = \notag \\
\int d^{2m}\boldsymbol{\alpha } \int d^{2l}\boldsymbol{\beta } ~\mathcal{p}(% \boldsymbol{\alpha } , \boldsymbol{\beta } ) ~d\left [ \sigma _ { 0}(\boldsymbol{% \alpha } ) , \sigma _ { 1}(\boldsymbol{\alpha } ) \right ] = \notag \\
\int d^{2m}\boldsymbol{\alpha } ~\mathcal{p}(\boldsymbol{\alpha } ) ~d\left [ \sigma _ { 0}(\boldsymbol{\alpha } ) , \sigma _ { 1}(\boldsymbol{\alpha } ) \right ] ~ , \label{eq_rho0_rho1}\end{gathered}\]]where @xmath186 is the marginal distribution defined in eq .
( [ q_alpha_distribution ] ) . in general
, it is known that @xcite@xmath340for every pair of quantum states @xmath16 and @xmath252 . as a consequence
, we can immediately write@xmath341 \leq \sqrt{1-f[\sigma _ { 0}(\boldsymbol{\alpha } ) , \sigma _ { 1}(% \boldsymbol{\alpha } ) ] } ~ , \label{d1f}\]]where @xmath342 $ ] is the fidelity between the two outputs @xmath343 and @xmath344 defined by eq .
( sigmau_alpha ) .
now we can exploit the multiplicativity of the fidelity under tensor products of density operators .
in other words , we can decompose@xmath345= \notag \\ f\left [ \bigotimes_{k=1}^{m}\mathcal{s}_{0}(\left\vert \alpha _ { k}\right\rangle \left\langle \alpha _ { k}\right\vert ) , \bigotimes_{k=1}^{m}% \mathcal{s}_{1}(\left\vert \alpha _ { k}\right\rangle \left\langle \alpha _ { k}\right\vert ) \right ] = \notag \\ \prod_{k=1}^{m}f\left [ \mathcal{s}_{0}(\left\vert \alpha _ { k}\right\rangle \left\langle \alpha _ { k}\right\vert ) , \mathcal{s}_{1}(\left\vert \alpha _ { k}\right\rangle \left\langle \alpha _ { k}\right\vert ) \right ] : = \prod_{k=1}^{m}f_{k}~ , \label{proof_repl}\end{gathered}\]]where @xmath346 is the fidelity between the two single - mode gaussian states @xmath347 and @xmath348 . in order to compute @xmath346 ,
let us consider the explicit action of @xmath349 on the coherent state @xmath350 , which is a gaussian state with cm @xmath351 and displacement @xmath352@xmath353@xmath354 . at the output of @xmath349
, we get a gaussian state @xmath355 whose statistical moments are proportional to the input ones , i.e. , @xmath356 and @xmath357 , were @xmath358 is given in eq .
( csiu_theo ) .
notice that @xmath359 because @xmath360 must be a bona - fide cm @xcite .
then , by using the formula of ref .
fidformulas , we can compute the analytical expression of the fidelity , which is equal to@xmath361 ~ , \label{fidelity_twogaussian}\]]where @xmath114 and @xmath332 are defined in eqs .
( [ omega_theo ] ) and ( [ lambda_theo ] ) , respectively .
it is trivial to check that @xmath362 and @xmath363 using @xmath359 .
now , using eq .
( fidelity_twogaussian ) into eq .
( [ proof_repl ] ) , we get@xmath364=\omega ^{-m}\exp \left [ -\lambda \sum_{k=1}^{m}\left\vert \alpha _ { k}\right\vert ^{2}\right ] = g(e)~ , \label{fe}\]]where@xmath365 ~,\]]and @xmath366 is defined in eq .
( [ e_def ] ) .
thus , by combining eqs .
( eq_rho0_rho1 ) , ( [ d1f ] ) and ( [ fe ] ) , we get@xmath367where we have introduced the @xmath368$]function@xmath369 since the latter quantity depends only on the real scalar @xmath366 , we can greatly simplify the integral @xmath33 in eq .
( [ integral_i ] ) . for this sake ,
let us introduce the polar coordinates@xmath370where @xmath371 is a vector with generic element @xmath372 , and @xmath373 is a vector of phases ( here @xmath374 ) .
then , we can write @xmath375where @xmath376 is the radial probability distribution@xmath377and@xmath378 _ { \boldsymbol{\alpha } = \sqrt{\mathbf{e}}\exp ( i\boldsymbol{% \theta } ) } ~.\]]the integral can be further simplified by setting @xmath379which introduces the further change of variables@xmath380then , we get@xmath381where@xmath382and@xmath383 _ { e_{m}=e-\sum_{k=1}^{m-1}e_{k}}~.\]]it is clear that the simplification of the integral from eq .
( integral_i ) to eq .
( [ integral_last ] ) can be done not just for @xmath384 , but for a generic integrable function @xmath385 . as a consequence
, we can repeat the same simplification for @xmath386 , which corresponds to write the following expression for the total number of photons @xmath387analytically , it is easy to check that @xmath388 is concave , i.e.,@xmath389~,\]]for every @xmath390 and @xmath391 $ ]
. then , applying jensen s inequality @xcite , we get@xmath392 = \mathcal{b}\left ( n\right ) \notag \\ & = & \sqrt{1-g(n)}~. \label{i_final}\end{aligned}\]]here we can set@xmath393where @xmath394 and @xmath327 is given in eq .
( [ stronger_f ] ) . according to eq .
( [ fidelity_twogaussian ] ) , @xmath327 represents the fidelity between the two states @xmath395 and @xmath396 , i.e. , the two possible outputs of the coherent state @xmath330 . in conclusion , by combining eqs .
( [ integral_i ] ) , ( [ i_final ] ) and ( [ g_f ] ) , we get@xmath397using the latter equation with eq . ( [ eq_pclass_proof ] ) , we obtain the lower bound of eq .
( [ class_gaussians ] ) .
@xmath250 according to the previous theorem , the classical discrimination bound @xmath398 can be _ computed _ by assuming a coherent - state transmitter which signals @xmath7 identical coherent states with @xmath394 mean photons each .
this transmitter can be denoted by @xmath399and is schematically depicted in fig .
[ cohepic ] . despite the computation of @xmath400 can be performed in this simple way , we do not know which classical transmitter is actually able to reach , or approach , the lower bound @xmath400 .
copies of a coherent state @xmath401 with mean photon number @xmath174.,scaledwidth=50.0% ] for the pure - loss model of fig .
[ scenario2 ] , the previous theorem can be greatly simplified .
in particular , we get a classical discrimination bound which does not depend on the bandwidth @xmath7 , but only on the total energy @xmath11 of the signal .
this result is stated in the following corollary ( which corresponds to the first theorem in our letter ) .
[ theo_forms_coro ] let us consider the discrimination of two attenuator channels , @xmath402 and @xmath403 , by using a classical transmitter @xmath19 which signals @xmath11 photons .
then , we have@xmath404where@xmath405 } } { 2}~. \label{c_attenuators}\ ] ] * proof*.
let us set @xmath406 in theorem theo_forms , so that@xmath407\otimes \lbrack \mathcal{n}(0)\circ \mathcal{n}(0 ) ] \notag \\ & = [ \mathcal{i}\circ \mathcal{e}(r_{u})\circ \mathcal{i}]\otimes \lbrack \mathcal{i}\circ \mathcal{i}]=\mathcal{e}_{u}\otimes \mathcal{i}~.\end{aligned}\]]clearly we can write @xmath408 .
this quantity is lower - bounded by using eq .
( [ class_gaussians ] ) where now the fidelity term @xmath409 can be greatly simplified . in fact , because of @xmath410 , here we have @xmath411 , which implies@xmath412as a consequence , we have@xmath413 \right\ } ^{m } \notag \\ & = & \exp \left [ -n\left ( \sqrt{r_{1}}-\sqrt{r_{0}}\right ) ^{2}\right ] ~,\end{aligned}\]]providing the expression of eq .
( [ c_attenuators ] ) .
as explained in the letter , we consider a particular kind of non - classical transmitter , that we call eprtransmitter .
this is given by@xmath414where @xmath415 is the tmsv state described in sec .
[ app_intro ] .
this transmitter is completely characterized by the basic parameters of the emitted signal , i.e. , bandwidth @xmath7 and energy @xmath11 , since @xmath416 .
for this reason , we also use the notation @xmath417 . given this transmitter ,
we now show how to compute the error probability which affects the channel discrimination .
we consider the general case of the thermal - loss model , but it is understood that the results hold for the pure - loss model by setting @xmath406 . given an arbitrary epr transmitter , the corresponding input state @xmath418is transformed into the conditional output state@xmath419where@xmath420the single - copy output state @xmath421 has zero mean and its cm @xmath422 can be easily computed from the one of the tmsv state , which is given in eq .
( [ cm_tmsv ] ) .
in fact , if we write the bipartite channel @xmath423\otimes \mathcal{n}% ( 2\varepsilon ) \]]in terms of @xmath424 matrices by using eqs .
( [ kn_1 ] ) and ( kn_2 ) , we can exploit the result of eq .
( [ v_out_bip_ch ] ) .
thus , we get the output cm@xmath425\mathbf{i } & \sqrt{r_{u}(\mu ^{2}-1)}\mathbf{z } \\ \sqrt{r_{u}(\mu ^{2}-1)}\mathbf{z } & ( \mu + 2\varepsilon ) \mathbf{i}% \end{array}% \right ) ~ , \label{outputcm_r}\]]where@xmath426now , the error probability which affects the channel discrimination @xmath427 is equal to the minimum error probability in the m - copy discrimination between @xmath428 and @xmath429 .
this quantity is upper - bounded by the quantum chernoff bound , i.e.,@xmath430where@xmath431 ^{m}~.\]]to compute @xmath432 we perform the normal - mode decomposition of the gaussian states @xmath428 and @xmath433 , which corresponds to the symplectic decomposition of the conditional cm of eq .
( [ outputcm_r ] ) @xcite .
then , we apply the symplectic formula of ref .
unfortunately , the result is extremely long to be presented here ( but short analytical expressions are provided afterwards in the proofs for the pure - loss model ) . note that an alternative upper - bound is the quantum battacharyya bound , which is given by@xmath434
^{m}\geq \mathcal{q}(m , n)~. \label{batta}\]]this bound is larger but generally easier to compute than the quantum chernoff bound ( see ref .
@xcite for details ) .
in this section we show how to make the comparison between the epr transmitter and the classical transmitters in order to reply to the basic questions of eqs .
( [ q1 ] ) and ( [ q2 ] ) .
our strategy is to derive a sufficient condition for the superiority of the epr transmitter by comparing the bounds derived in the previous sections . on the one hand ,
we know that the error probability of any classical transmitter @xmath435 is lower - bounded by @xmath108 . on the other hand ,
the error probability of the epr transmitter @xmath436 is upper - bound by @xmath437 .
thus , a sufficient condition for @xmath438 is provided by the inequality @xmath439 .
depending on the model , our answer can be analytical or numerical . for the pure - loss model , we have @xmath440 and @xmath441 . since here
the classical discrimination bound is not dependent on the bandwidth , it is sufficient to find a @xmath58 such that @xmath442 . in other words ,
our basic question of eq .
( [ q1 ] ) becomes the following : given two channels @xmath278 and signal - energy @xmath11 , can we find a bandwidth @xmath58 such that @xmath443this clearly implies the existence of an epr transmitter @xmath61such that@xmath444for every @xmath40 . as stated in our letter
this question has a positive answer , and this is provided by our threshold energy theorem .
according to this theorem , for every pair of attenuator channels @xmath287 and above a threshold energy @xmath65 , there is an epr transmitter ( with suitable bandwidth @xmath58 ) which outperforms any classical transmitter .
furthermore , we can prove a stronger result if one of the two channels is just the identity , e.g. , we have @xmath445 . in this case we have @xmath84 and
@xmath58 is a minimum bandwidth , as stated by the ideal memory theorem in our letter .
we give the explicit proofs of these theorems in sec .
[ app_pure_loss ] . in order to quantify numerically the advantage brought by the eprtransmitter
, we introduce the minimum information gain @xmath446=g(m , n)~,\]]where@xmath447is the binary shannon entropy . finding a bandwidth @xmath58 such that @xmath448 clearly implies the validity of eq .
( [ qq1 ] ) . in terms of memory readout
, this quantity lowerbounds the number of bits per cell which are gained by an epr transmitter @xmath61 over any classical transmitter . by using @xmath76 one can perform extensive numerical investigations @xcite . in particular
, one can check that narrowband epr transmitters ( i.e. , with low @xmath58 ) are able to give @xmath70 in the regime of few photons @xmath11 and high - reflectivities ( i.e. , @xmath0 or @xmath1 sufficiently close to @xmath71 ) .
these numerical results are shown and discussed in fig .
2(left ) and fig .
3 of our letter . for the thermal - loss model
, we have @xmath440 and @xmath449 . here the classical discrimination bound @xmath450 depends on the bandwidth @xmath7 and tends monotonically to zero for @xmath85 .
this is evident from the general expression of the fidelity term @xmath409 which is present in eq .
( class_gaussians ) .
in fact , from eq .
( [ stronger_f ] ) , we get@xmath451which decreases to zero for @xmath85 whenever @xmath452 ( as is generally the case for the thermal - loss model ) .
for this reason , let us fix a maximum bandwidth @xmath112 and consider the minimum value@xmath453}\mathcal{c}(m , n)=\mathcal{c}(m^{\ast } , n)~.\]]it is clear that , given two bipartite channels @xmath454 and signal - energy @xmath11 , we have@xmath455for every @xmath19 with signal - bandwidth @xmath456 . in other words , using @xmath457 we can bound all the classical transmitters up to the maximum bandwidth @xmath112 . as a result , our basic question of eq .
( [ q2 ] ) becomes the following : given two bipartite channels @xmath458 and signal - energy @xmath191 , can we find an epr transmitter with suitable bandwidth @xmath58 such that @xmath459this clearly implies the existence of an epr transmitter @xmath61such that@xmath460for every @xmath19 with @xmath456 .
as explicitly discussed in the letter , this question has a positive answer too , and the value of @xmath112 can be so high to include all the classical transmitters which are physically meaningful for the memory model
. however , despite the previous case , here the answer can only be numerical .
in fact , the two bipartite channels @xmath458 are characterized by four parameters @xmath461 and the comparison between transmitters involves further three parameters @xmath462 .
for this reason , the explicit expression of the minimum information gain now depends on seven parameters , i.e.,@xmath463by performing numerical investigations , one can analyze the positivity of @xmath76 , which clearly implies the condition of eq .
( [ qq2 ] ) . according to fig .
2(right ) of our letter , we can achieve remarkable positive gains when we compare narrowband epr transmitters ( low @xmath58 ) with wide sets of classical transmitters ( large @xmath112 ) in the regime of few photons ( low @xmath11 ) and high reflectivities ( @xmath0 or @xmath1 close to @xmath71 ) , and assuming the presence of non - trivial decoherence ( e.g. , @xmath97 ) .
see the letter for physical discussions .
in this section we provide the detailed proofs of the two theorems relative to the pure - loss model : the threshold energy theorem and the ideal memory theorem . given two attenuator channels , @xmath464 and @xmath465 , and given the total signal - energy @xmath11 , the basic problem is to show the existence of an epr transmitter @xmath54 which is able to beat the classical discrimination bound @xmath466 , and , therefore , all the classical transmitters @xmath40 .
the two theorems provide sufficient conditions for the existence of this quantum transmitter . for the sake of completeness we repeat here the statement of the theorem (
this is perfectly equivalent to the statement provided in our letter ) .
let us consider the discrimination of two attenuator channels , @xmath467 and @xmath465 with @xmath62 , by using transmitters with ( finite ) signal energy @xmath468then , there exists an epr transmitter @xmath61 , with suitable bandwidth @xmath58 , such that@xmath469 * * proof . * * for simplicity , in this proof we use the shorthand notation @xmath470 . according to corollary [ theo_forms_coro ] ,
the error probability for an arbitrary classical transmitter @xmath40 is lower - bounded by the classical discrimination bound , i.e. , @xmath471 , where @xmath472 is given in eq .
( [ c_attenuators ] ) .
note that , for every @xmath473 $ ] , we have@xmath474as a consequence , we can write@xmath475where@xmath476 now let us consider an epr transmitter @xmath54
. the corresponding error probability @xmath477 can be upper - bounded via the quantum battacharyya bound @xmath478 defined in eq .
( [ batta ] ) .
this bound can be computed from the conditional cm of eq .
( [ outputcm_r ] ) by setting @xmath406 ( see sec .
[ app_epr_tra ] and ref .
minkopra ) . in this problem , for fixed ( finite ) energy @xmath11 ,
the bound @xmath479 is not monotonic in @xmath7 ( easy to check numerically ) .
however , it is regular in @xmath7 and its asymptotic limit @xmath480 exists .
in particular , the asymptotic battacharyya bound is equal to@xmath481where @xmath482 $ ] is defined by@xmath483clearly we have@xmath484which means that @xmath485 , @xmath486 such that@xmath487 in order to prove the result , let us impose the threshold condition@xmath488now , if eq .
( [ th_app ] ) is satisfied , then we can always take an @xmath489 such that@xmath490then , because of the proposition of eq .
( [ b_eps ] ) , we have that @xmath491 such that@xmath492 for this reason , the next step is to solve eq .
( [ th_app ] ) in order to get a threshold condition on the energy @xmath11 .
it is easy to check that , by using eqs .
( [ c_tilde])-([y_exp ] ) , the condition of eq .
( [ th_app ] ) becomes@xmath493where@xmath494 ~.\end{gathered}\]]since @xmath495 is proportional to the difference between an arithmetic mean and a geometric mean , we have @xmath496 , and @xmath497 if and only if @xmath498 .
thus , if we exclude the singular case @xmath498 , we can write@xmath499 in conclusion , if the threshold condition of eq .
( [ eq_n ] ) is satisfied ( where @xmath500 ) , then there exists a bandwidth @xmath58 such that @xmath501 here we prove the ideal memory theorem which refers to the case of ideal memories , i.e. , having @xmath80 .
this scenario corresponds to discriminating between an attenuator channel @xmath502 and the identity channel @xmath503 . for this reason ,
the quantum chernoff bound @xmath432 has a simple analytical expression , which turns out to be decreasing in @xmath7 ( for fixed energy @xmath11 ) . thanks to this monotony ,
our result can be proven above a minimum bandwidth @xmath58 .
in fact , @xmath432 decreasing in @xmath7 implies that @xmath504-\{1-h[\mathcal{c}(n)]\}$ ] is increasing in @xmath7 , so that optimal gains are monotonically reached by broadband epr transmitters . note that this was not possible to prove for the previous theorem ( see sec .
[ app_proof1 ] ) since , in that case , the quantum battacharyya bound @xmath478 turned out to be non - monotonic in @xmath7 . for the sake of completeness we repeat here the statement of the ideal memory theorem (
this is perfectly equivalent to the statement provided in our letter ) . let us consider the discrimination of an attenuator channel @xmath505 , with @xmath506 , from the identity channel @xmath33 , by using transmitters with signal energy @xmath507then , there exists a minimum bandwidth @xmath58 such that , for every epr transmitter @xmath54 with @xmath83 , we have@xmath508 * proof .
* given the discrimination problem @xmath509 for fixed signal energy @xmath11 , let us consider an eprtransmitter @xmath54 .
its error probability @xmath510 is upper - bounded by the quantum chernoff bound @xmath432 , whose analytical expression is greatly simplified here .
in fact , we have@xmath511where@xmath512~. \label{x_r0}\]]this expression can be computed using the procedure sketched in sec .
app_epr_tra by setting @xmath406 and @xmath129 in the conditional cm of eq .
( [ outputcm_r ] ) . for fixed energy @xmath11
, it is very easy to check analytically that @xmath432 is decreasing in @xmath7 ( strictly decreasing if we exclude the trivial case @xmath513 ) . since @xmath514 is bounded ( @xmath515 $ ] ) and decreasing in @xmath7 , the broadband limit@xmath516exists , and clearly coincides with the infimum , i.e.,@xmath517explicitly , we compute@xmath518by definition of limit , eqs .
( [ q_limit ] ) and ( [ q_inf ] ) mean that @xmath519 , @xmath520 such that @xmath521@xmath522now , since we have @xmath523 for every @xmath7 , it is clear that @xmath485 , @xmath520 such that @xmath524 @xmath525 in order to prove our result , we impose the threshold condition@xmath526where @xmath466 is the classical discrimination bound . according to corollary [ theo_forms_coro ] ,
this is given by@xmath527where@xmath528now , if eq .
( [ qc_diseq0 ] ) is satisfied , then we can always take an @xmath489 such that@xmath529then , because of the proposition of eq .
( [ eq_prop ] ) , we have that @xmath530 such that @xmath524 @xmath531 plotted in the finite @xmath532-plane .
only for @xmath533 we have @xmath534 . *
energy function @xmath535 plotted in the infinite @xmath536-plane ( here restricted to the sector @xmath537 ) . only for @xmath538 @xmath539
we have @xmath540 .
dashed line corresponds to the universal energy threshold @xmath84.,scaledwidth=47.0% ] clearly , the next step is to solve eq .
( [ qc_diseq0 ] ) and get a threshold condition on the energy @xmath11 .
after simple algebra , eq .
( [ qc_diseq0 ] ) can be written as@xmath541where@xmath542and@xmath543note that , in the definition of @xmath544 , we are excluding the singular points @xmath545 .
the function @xmath546 can be easily analyzed on the plane @xmath547\times ( 0,1)$ ] .
in particular , its zero - level @xmath548 is represented by the continuous function @xmath549 which is plotted in fig .
[ enethr](left ) .
the threshold condition of eq .
( [ g_cond ] ) is equivalent to @xmath550as evident from fig .
[ enethr](left ) , the function @xmath549 is decreasing in @xmath551 $ ] with extremal values@xmath552}\bar{y}(x)=e^{-1/4}~,\]]and@xmath553}\bar{y}(x)=% \frac{\sqrt{5}-1}{2}~.\ ] ] equivalently , by using eqs .
( [ x_r0 ] ) and ( [ y_n_form ] ) , we can put the threshold condition of eq .
( [ thr_y ] ) in terms of @xmath0 and @xmath11 .
in particular , eq .
( [ thr_y ] ) takes the form@xmath554where @xmath535 is the decreasing function of @xmath555 which is plotted in fig . [ enethr](right ) .
this function has extremal values@xmath556and@xmath557 in order to have a criterion which is universal , i.e. , @xmath0-independent , we can consider the sufficient condition@xmath558which clearly implies eq .
( [ eq_condn ] ) .
more easily we can consider the slightly - larger condition@xmath507 in conclusion , for every @xmath80 and ( finite ) @xmath81 , there exists a minimum bandwidth @xmath58 such that @xmath524 we have @xmath559.@xmath560
in our derivations we have assumed that the output receiver is able to perform an optimal measurement given by the helstrom s povm of eq .
( [ optimal_povm ] ) . despite this measurement
has an extremely simple formula , it is not straightforward to implement it using linear optics and photodetection .
this problem has been already considered in ref .
@xcite for the case of quantum illumination , where an ingenious receiver design has been proven to harness quantum illumination advantage . unfortunately this kind of design can not be applied directly to our scheme , since quantum reading not only has a different task than quantum illumination but also works in a completely different regime ( high reflectivities , low thermal - noise and narrowband signals ) .
note that the working regime of quantum reading is such that strong correlations survive at the output of the process .
for this reason , we can consider simpler receiver designs than the ones for quantum illumination .
in fact , here we prove that an output bell detection followed by a simple classical processing is a sub - optimal receiver able to provide remarkable advantages , i.e. , comparable to the helstrom s povm . since bell detection is a standard measurement ( involving linear optics and photodetection ) our reading apparatus
can be easily implemented in today s quantum optics labs .
the quantum reading scheme with the sub - optimal receiver is depicted in fig .
[ suboptimalpic ] .
the input modes @xmath169 belong to a tmsv state @xmath562 .
these modes are processed as usual , i.e. , the signal mode @xmath177 is conditionally transformed by the cell ( reflectivity @xmath101 encoding the bit @xmath14 ) while the idler mode @xmath178 is sent directly to the receiver .
their quadratures , @xmath563 and @xmath564 , are transformed according to the input - output relations@xmath565and@xmath566where @xmath567 are the quadratures of the external thermal bath , while @xmath568 , @xmath569 and @xmath570 are quadratures associated with the internal thermal channels . at the receiver the two modes
@xmath571 are combined in a balanced beam - splitter which outputs the modes @xmath572 with quadratures@xmath573and@xmath574here the epr quadratures @xmath575 are measured by using two homodyne detectors .
the corresponding outcome @xmath576 is classically processed via a @xmath577-test , whose output is the value of the bit stored in the memory cell . in general , for arbitrary bandwidth @xmath578 , we have @xmath7 identical copies @xmath579 at the input and , therefore , @xmath580 output variables @xmath581 which are processed by the @xmath123-test .
let us analyze the detection process in detail in order to derive the error - probability which affects the decoding .
first of all we can easily verify that the two classical outputs , @xmath582 and @xmath583 , are identical gaussian variables , so that we can simply use @xmath584 to denote @xmath582 or @xmath585 . clearly the gaussian variable @xmath584 is tested two times for each single - copy state @xmath586 and , therefore , @xmath580 times during the reading time of the logical bit .
this variable has zero mean and variance @xmath587 which is conditioned to the value of the stored bit @xmath259 .
explicitly , we have@xmath588 ~ , \label{v_u}\end{aligned}\]]where @xmath101 is the conditional reflectivity of the cell , @xmath589 is the variance associated with the signal - mode , @xmath590 is the noise - variance of the external thermal bath , and @xmath95 is the noise - variance of the internal thermal channel .
it is clear that the decoding of the logical bit @xmath14 corresponds to the statistical discrimination between two variances @xmath591 and @xmath592 . in other words ,
the variable @xmath584 is subject to the hypothesis test@xmath593it is easy to show that @xmath594 of eq .
( [ v_u ] ) is a decreasing function of the reflectivity @xmath101 as long as @xmath595which is always true in the regime considered here ( i.e. , @xmath95 small and @xmath596 close to @xmath71 ) .
this means that @xmath597 implies @xmath598 , which makes eq .
( [ test1 ] ) a one - tailed test .
equivalently , we can introduce the normalized variable@xmath599with zero mean and variance @xmath600 , and replace eq .
( test1 ) with the one - tailed test@xmath601where@xmath602thus , assuming @xmath597 , the decoding of the logical bit is equivalent to the statistical discrimination between the two hypotheses in eq .
( test2 ) , i.e. , @xmath603 and @xmath604 . for arbitrary bandwidth @xmath7
, we collect @xmath580 independent outcomes @xmath605 and we construct the @xmath123-variable@xmath606then , we select one of the two hypotheses according to the following rule @xmath607where @xmath608 is the @xmath609 quantile of the @xmath123 distribution with @xmath610 degrees of freedom . here
the quantity @xmath124 represents the significance level of the test , corresponding to the asymptotic probability of wrongly rejecting the hypothesis @xmath611 , i.e.,@xmath612this quantity must be fixed and its value characterizes the test .
in particular , for finite @xmath7 , there will be an optimal value of @xmath124 which maximizes the performance of the test .
let us explicitly compute the error probability affecting the classical test . for arbitrary variance @xmath613 ,
the @xmath123 distribution with @xmath610 degrees of freedom is equal to@xmath614where @xmath615 is the gamma function .
the probability of finding @xmath616 bigger than @xmath617 is given by the integral@xmath618where @xmath619 is the incomplete gamma function .
thus , given @xmath611 ( i.e. , @xmath604 ) the probability of accepting @xmath620 ( i.e. , @xmath621 ) is given by by contrast , given @xmath623 ( i.e. , @xmath624 ) the probability of accepting @xmath611 ( i.e. , @xmath625 ) is given by@xmath626from eqs .
( [ ph1 ] ) and ( [ ph2 ] ) we compute the error probability affecting the test , which is given by@xmath627
= \frac{1}{2}+\frac{1}{2}\left [ i\left ( m,1,\mathcal{q}_{1-\varphi } ^{2m-1}\right ) -i\left ( m,1+\sigma , \mathcal{q}_{1-\varphi } ^{2m-1}\right ) % \right ] \notag \\ = p_{test}\left ( r_{0},r_{1},n , m,\bar{n},\varepsilon , \varphi
\right ) ~.\end{gathered}\]]clearly this quantity depends on all the parameters of the model , i.e. , memory reflectivities @xmath24 , energy and bandwidth of the signal @xmath628 , levels of noise @xmath629 and significance level of the test @xmath124 .
this error probability must be compared with the classical discrimination bound @xmath108 . for simplicity
let us first consider the pure - loss model ( @xmath406 ) . in this case
@xmath630 must be compared with @xmath631 of eq .
( c_attenuators ) .
then , for given memory @xmath24 and signal - energy @xmath11 , we have that quantum reading is superior if we find a bandwidth @xmath7 and a significance level @xmath124 such that @xmath632 .
this is equivalent to prove the positivity of the ( sub - optimal ) information gain @xmath633~,\]]which provides a lowerbound to the number of bits per cell which are gained by quantum reading over any classical strategy . in fig .
[ bellpic](left ) we optimize @xmath634 over @xmath7 and @xmath124 for an ideal memory in the few - photon regime .
remarkably , we can find an area where @xmath635 bits per cell .
then , let us consider the thermal - loss model , with @xmath636 . in this case
, @xmath637 must be compared with @xmath638 where the value of @xmath112 is high enough to include all the classical transmitters which are meaningful for the model ( here we take @xmath639 ) . in fig .
[ bellpic](right ) , we consider a memory with @xmath133 and @xmath134 ( high - reflectivity regime ) , which is illuminated by a signal with @xmath135 ( few - photon regime ) .
the numerical optimization over @xmath7 and @xmath124 shows an area where @xmath640 bits per cell .
thus , the remarkable advantages of quantum reading are still evident when the optimal helstrom s povm is replaced by a sub - optimal receiver , consisting of a bell measurement followed by a suitable classical post - processing . this sub - optimal receiver
can be easily realized in today s quantum optics labs , thus making quantum reading a technique within the catch of current technology . for every @xmath179-mode bosonic state @xmath16 ,
we can write the @xmath189-representation @xmath641 , where @xmath642 is normalized to @xmath71 and @xmath643 is a generic multimode coherent state . by definition , @xmath16 is called classical ( non - classical ) if @xmath186 is positive ( non - positive ) @xcite . for a classical state , @xmath188 represents a probability distribution ( implying separability ) . for two bosonic modes , @xmath218 and @xmath219 , with quadratures @xmath644 , @xmath645 , @xmath646 and @xmath647
, one can define the two operators @xmath648 ( relative position ) and @xmath649 ( total momentum ) .
then , the system has epr correlations ( in these operators ) if @xmath650 , where @xmath651 is the variance , and @xmath652 is the standard quantum limit ( @xmath653 in this paper . ) note that a bosonic system with epr correlations is entangled , but the contrary is not necessarily true . in today s quantum optics labs ,
the epr correlations represent the standard type of continuous variable entanglement .
these correlations are usually generated via spontaneous parametric down conversion .
a single - layer cd with speed 1x ( i.e. , 500 rpm ) has data - transfer rate of 1.2 mbit / s .
this data is read by a laser diode at 780 nm and power @xmath658 mw , therefore irradiating about @xmath659 nj per bit .
a higher - speed cd needs a laser with proportionally more power , in such a way that at least @xmath71 nj per bit is irradiated , corresponding to @xmath138 photons per bit . in standard cd - r and dvd - r media ,
the data is recorded in internal layers of organic dye ( e.g. , azo , cyanine , or phthalocyanine ) .
the dye polymer darkens when heated by a high - power laser , with the result of causing less reflection from the metallic layer underneath ( dye - sublimation @xcite ) .
jensen s inequality states that , for every random variable @xmath668 , with values in @xmath669 and expectation value @xmath670 , and for any concave function @xmath671 , we have @xmath672 ( the inequality is reversed for convex functions ) .
more in general , the inequality holds for random variables and concave functions which are defined in measurable subsets @xmath673 . | we consider a simple model of digital memory where each cell is a reflecting medium with two possible reflectivities ( used to store a bit of information ) . adopting this model , we show that a non - classical source of light , possessing einstein - podolsky - rosen correlations , can retrieve more information than any classical source .
this improvement is shown in the regime of high reflectivities and few photons , where the gain of information can be dramatic . as a result
, quantum light can be used to increase storage capacities and data - transfer rates of classical digital memories . in recent years , non - classical states of radiation
have been exploited to achieve marvellous results in quantum information and computation books , books2 . in the language of quantum optics ,
the bosonic states of the electromagnetic field are called classical when they can be expressed as probabilistic mixtures of coherent states .
classical states describe practically all the radiation sources which are used in today s technological applications . by contrast
, a bosonic state is called non - classical when its decomposition in coherent states is non - positive @xcite .
one of the key properties which makes a state non - classical is the quantum entanglement . in the bosonic framework , this is usually present under the form of einstein - podolsky - rosen ( epr ) correlations @xcite , meaning that the position and momentum quadrature operators of two bosonic modes are so correlated to beat the standard quantum limit @xcite .
this is a well - known feature of the two - mode squeezed vacuum ( tmsv ) state @xcite , one of the most important states routinely produced in today quantum optics labs . in this letter
, we show how the use of non - classical light possessing eprcorrelations can _ widely _ improve the readout of information from classical digital memories . to our knowledge , this is the first study which proves and quantifies the advantages of using non - classical light for this fundamental task , being absolutely non - trivial to identify the physical conditions that can effectively disclose these advantages ( as an example , see the recent no - go theorems of ref .
@xcite applied to quantum illumination @xcite ) .
our model of digital memory is simple but can potentially be extended to realistic optical disks , like cds and dvds , or other kinds of memories , like the barcodes .
in fact , we consider a memory where each cell is composed by a reflecting medium with two possible reflectivities , @xmath0 and @xmath1 , used to store a bit of information .
this memory is irradiated by a source of light which is able to resolve every single cell .
the light focussed on , and reflected from , a single cell is then measured by a detector , whose outcome provides the value of the bit stored in that cell . besides the signal modes irradiating the target cell
, we also consider the possible presence of ancillary idler modes which are directly sent to the detector .
the general aim of these modes is to improve the performance of the output measurement by exploiting possible correlations with the signals . adopting this model and fixing the number of photons irradiated over the memory ,
we show that a _ non - classical _ source of light with epr correlations between signals and idlers can retrieve more information than any classical source of light .
in particular , this is proven for high reflectivities ( typical of optical disks ) and few photons irradiated . in this regime
the difference of information can be surprising , i.e. , up to 1 bit per cell ( corresponding to the extreme situation where only quantum light can retrieve information ) .
as we will discuss in the conclusion , the chance of reading digital memories using few photons leads to remarkable consequences , like boosting their storage capacities and data - transfer rates .
encoding bit - value @xmath2 , and @xmath3 encoding bit - value @xmath4 .
* readout of the memory*. in general , a digital reader consists of transmitter and receiver . the transmitter @xmath5 is a bipartite bosonic system , composed by a signal system @xmath6 ( with @xmath7 modes ) and an idler system @xmath8 ( with @xmath9 modes ) , which is given in some global state @xmath10 .
the signal @xmath6 emitted by this source has bandwidth @xmath7 and energy @xmath11 ( total number of photons ) .
the signal is directly shined over the cell , and its reflection @xmath12 is detected together with the idler @xmath8 at the output receiver , where a suitable measurement retrieves the value of the bit up to an error probability @xmath13.,scaledwidth=42.0% ] let us consider a digital memory where each cell can have two possible reflectivities , @xmath0 or @xmath1 , encoding the two values of a logical bit @xmath14 ( see fig . [ qreadpic ] ) .
close to the memory , we have a digital reader , made up of transmitter and receiver , whose goal is retrieving the value of the bit stored in a target cell . in general , we call transmitter a bipartite bosonic system , composed by a signal system @xmath6 with @xmath7 modes and an idler system @xmath8 with @xmath15 modes , and globally given in some state @xmath16 .
this source can be completely specified by the notation @xmath17 . by definition ,
we say that the transmitter @xmath18 is classical(non - classical ) when the corresponding state @xmath16 is classical ( non - classical ) , i.e. , @xmath19and @xmath20 @xcite .
the signal @xmath6 emitted by the transmitter is associated with two basic parameters : the total number of modes @xmath7 , that we call bandwidth of the signal , and the total number of photons @xmath11 , i.e. , its energy .
the signal @xmath6 is directly shined on the target cell , and its reflection @xmath12 is detected together with the idler @xmath8 at the output receiver . here a suitable measurement outputs the value of the bit up to an error probability @xmath13 . repeating the process for each cell of the memory ,
the reader retrieves an average of @xmath21 bits per cell , where @xmath22 is the binary shannon entropy .
the basic mechanism in our model of digital readout is the quantum channel discrimination .
in fact , encoding a logical bit @xmath23 in a pair of reflectivities @xmath24 is equivalent to encoding @xmath14 in a pair of attenuator channels @xmath25 , with linear losses @xmath24 and acting on the signal modes .
the readout of the bit consists in the statistical discrimination between @xmath0 and @xmath1 , which is formally equivalent to the channel discrimination between @xmath26 and @xmath27 .
the error probability affecting the discrimination @xmath28 depends on both transmitter and receiver . for a _ fixed _
transmitter @xmath17 , the pair @xmath25 generates two possible output states at the receiver , @xmath29 and @xmath30 .
these are expressed by @xmath31(\rho ) $ ] , where @xmath32 acts on the signals and the identity @xmath33 on the idlers . by optimizing over the output measurements ,
the minimum error probability which is achievable by the transmitter @xmath18 in the channel discrimination @xmath34 is equal to @xmath35 , where @xmath36 is the trace distance between @xmath29 and @xmath30 .
now the crucial point is the minimization of @xmath37 over the transmitters @xmath18 .
clearly , this optimization must be constrained , by fixing the basic parameters of the signal .
here we consider the most general situation where only the signal energy @xmath11 is fixed .
under this energy constraint , it is unknown the optimal transmitter @xmath18 which minimizes @xmath37 .
for this reason , it is non - trivial to ask the following question : does it exist a non - classical transmitter which outperforms any classical one ?
in other words : given two reflectivities @xmath24 , i.e. , two attenuator channels @xmath25 , and a fixed value @xmath11of the signal energy , can we find any @xmath38 such that @xmath39 for every @xmath40 ? in the following we reply to this basic question , individuating remarkable physical regimes where the answer is positive .
the first step in our derivation is providing a bound which is valid for every classical transmitter ( see appendix for the proof ) .
[ theomain]_let us consider the discrimination of two reflectivities _
@xmath24 _
_ using a classical transmitter _ _ @xmath41 _ _ which signals _
_ @xmath11 _ _ photons .
the corresponding error probability satisfies__@xmath42 according to this theorem , all the classical transmitters @xmath40 irradiating @xmath11 photons on a memory with reflectivities @xmath24 can not beat the classical discrimination bound @xmath43 , i.e. , they can not retrieve more than @xmath44 bits per cell .
clearly , the next step is constructing a non - classical transmitter which can violate this bound .
a possible design is the eprtransmitter , composed by @xmath7 signals and @xmath7 idlers , that are entangled pairwise via two - mode squeezing .
this transmitter has the form @xmath45 , where @xmath46 is a tmsv state entangling signal mode @xmath47 with idler mode @xmath48 . in the number - ket representation @xmath49 , where the squeezing parameter @xmath50 quantifies the signal - idler entanglement .
an arbitrary epr transmitter , composed by @xmath7 copies of @xmath51 , irradiates a signal with bandwidth @xmath7 and energy @xmath52 . as a consequence
, this transmitter can be completely characterized by the basic parameters of the emitted signal , i.e. , we can set @xmath53 .
then , let us consider the discrimination of two reflectivities @xmath24 using an epr transmitter @xmath54 which signals @xmath11 photons
. the corresponding error probability can be upper - bounded via the quantum chernoff bound qcbound , minkopra@xmath55 ^{m } , \label{qcb_qread}\]]where @xmath56(\left\vert \xi \right\rangle \left\langle \xi \right\vert ) $ ] .
in other words , at least @xmath57 bits per cell can be retrieved from the memory .
exploiting eqs .
( [ cb_cread ] ) and ( [ qcb_qread ] ) , our main question simplifies to finding @xmath58 such that @xmath59 . in fact , this implies @xmath60 , i.e. , the existence of an epr transmitter @xmath61 able to outperform any classical transmitter @xmath40 .
this is the result of the following theorem ( see appendix for the proof ) .
_ for every pair of reflectivities _
@xmath24 _ _ with _
_ @xmath62 _ _ , and signal energy _
_ @xmath63__there is a _
_ @xmath58 _ _ such that _ _ @xmath64 . thus we get the central result of the paper : for every memory and above a threshold energy , there is an epr transmitter which outperforms any classical transmitter .
remarkably , the threshold energy @xmath65 turns out to be low ( @xmath66 ) for most of the memories @xmath24 outside the region @xmath67 .
this means that we can have an enhancement in the regime of few photons ( @xmath68 ) .
furthermore , for low energy @xmath11 , the critical bandwidth @xmath58 can be low too .
in other words , in the regime of few photons , narrowband epr transmitters are generally sufficient to overcome every classical transmitter .
to confirm and quantify this analysis , we introduce the minimum information gain @xmath69 $ ] . for given memory @xmath24 and signal - energy @xmath11 ,
this quantity lowerbounds the number of bits per cell which are gained by an eprtransmitter @xmath54 over any classical transmitter @xmath40 @xcite .
numerical investigations ( see fig . [ colorbar ] ) show that narrowband epr transmitters are able to give @xmath70 in the regime of few photons and high reflectivities , corresponding to having @xmath0 or @xmath1 sufficiently close to @xmath71 ( as typical of optical disks ) . in this regime ,
part of the memories display remarkable gains ( @xmath72 ) . over the memory plane @xmath73 .
for a few - photon signal ( @xmath74 ) , we compare a narrowband epr transmitter ( @xmath75 ) with all the classical transmitters . inside the black region ( @xmath67 )
our investigation is inconclusive . outside the black region
, we have @xmath70 . *
right .
* @xmath76 plotted over the plane @xmath24 in the presence of decoherence ( @xmath77 ) .
for a few - photon signal ( @xmath74 ) , we compare a narrowband epr transmitter ( @xmath75 ) with all the classical transmitters @xmath78 having @xmath79.,scaledwidth=47.0% ] thus the enhancement provided by quantum light can be dramatic in the regime of few photons and high reflectivities @xcite . in order to investigate this regime more closely
, we consider the case of ideal memories , defined by @xmath80 . as analytical result
, we have the following theorem ( see appendix for the proof ) . _ for every _ @xmath80 _ _ and _ _ @xmath81 _ _ , there is a minimum bandwidth _ _ @xmath58 _ _ such that _ _ @xmath82 _ _ for every
_ _ @xmath83 .
thus , for ideal memories and signals above @xmath84 photon , there are infinite epr transmitters able to outperform every classical transmitter . for these memories ,
the threshold energy is so low that the regime of few photons can be fully explored .
the gain @xmath76 increases with the bandwidth , so that optimal performances are reached by broadband eprtransmitters ( @xmath85 ) .
however , narrowband epr transmitters are sufficient to give remarkable advantages , even for @xmath86 ( i.e. , using a single tmsv state ) .
this is shown in fig .
[ idealpic ] , where @xmath76 is plotted in terms of @xmath0 and @xmath11 , considering the two extreme cases @xmath86 and @xmath85 . according to fig .
[ idealpic ] , the value of @xmath87 can approach @xmath71 for ideal memories and few photons even if we consider narrowband epr transmitters . versus @xmath0 and @xmath11 .
left picture refers to @xmath86 , right picture to @xmath85 .
( for arbitrary @xmath7 the scenario is intermediate . ) outside the inconclusive black region we have @xmath70 . for @xmath85 the black region
is completely collapsed below @xmath84.,scaledwidth=47.0% ] _ presence of decoherence_. note that the previous analysis does not consider the presence of thermal noise .
actually this is a good approximation in the optical range , where the number of thermal background photons is around @xmath88 at about @xmath89 m and @xmath90 .
however , to complete the analysis , we now show that the quantum effect exists even in the presence of stray photons hitting the upper side of the memory and decoherence within the reader .
the scattering is modelled as white thermal noise with @xmath91 photons per mode entering each memory cell .
numerically we consider @xmath92 corresponding to non - trivial diffusion .
this scenario may occur when the light , transmitted through the cells , is not readily absorbed by the drive ( e.g. , using a bucket detector just above the memory ) but travels for a while diffusing photons which can hit the next cells .
assuming the presence of one photon per mode travelling the optimistic distance of one meter and undergoing rayleigh scattering , we get roughly @xmath93 @xcite .
the internal decoherence is modelled as a thermal channel @xmath94 adding gaussian noise of variance @xmath95 to each signal / reflected mode , and @xmath96 to the each idler mode ( numerically we consider the non - trivial value @xmath97 ) .
now , distinguishing between two reflectivities @xmath24 corresponds to discriminating between two gaussian channels @xmath98 for @xmath23 . here
@xmath99 acts on each signal mode , and contains the attenuator channel @xmath100 with conditional loss @xmath101 and thermal noise @xmath91 . to solve this problem
we can use theorem theomain with the proviso of generalizing the classical discrimination bound . in general
, we have @xmath102 , where @xmath103 is the fidelity between @xmath104 and @xmath105 , the two outputs of a single - mode coherent state @xmath106 with @xmath107 mean photons . here the expression of @xmath108 depends also on the bandwidth @xmath7 of the classical transmitter @xmath19 . since @xmath108 decreases to zero for @xmath109 , our quantum - classical comparison is now restricted to classical transmitters @xmath110 with @xmath7 less than a maximal value @xmath111 . remarkably we find that , in the regime of few photons and high reflectivities , narrowband epr transmitters are able to outperform all the classical transmitters up to an extremely large bandwidth @xmath112 .
this is confirmed by the numerical results of fig .
[ colorbar ] , proving the robustness of the quantum effect @xmath70 in the presence of decoherence .
notice that we can neglect classical transmitters with extremely large bandwidths ( i.e. , with @xmath113 ) since they are not meaningful for the model .
in fact , in a practical setting , the signal is an optical pulse with carrier frequency @xmath114 high enough to completely resolve the target cell .
this pulse has frequency bandwidth @xmath115 and duration @xmath116 . assuming an output detector with response time @xmath117 and reading time @xmath118 , the number of temporal modes which are excited is roughly @xmath119 @xcite . in other words ,
the bandwidth of the signal @xmath7 is the product of its frequency bandwidth @xmath120 and the reading time of the detector @xmath18 .
now , the limit @xmath109 corresponds to @xmath121 ( infinite detector resolution ) or @xmath122(infinite reading time ) . as a result ,
transmitters with too large @xmath7 can be discarded .
_ sub - optimal receiver_. the former results are valid assuming an optimal output detection . here
we show an explicit receiver design which is ( i ) easy to construct experimentally and ( ii ) able to approximate the optimal results .
this sub - optimal receiver consists of a continuous variable bell measurement ( i.e. , a balanced beam - splitter followed by two homodyne detectors ) whose output is classically processed by a suitable @xmath123-test with significance level @xmath124 ( see appendix for details ) . in this case
the information gain @xmath76 can be optimized jointly over the signal bandwidth @xmath7 ( i.e. , the number of input tmsv states ) and the significance level of the output test @xmath124 . as shown in fig .
[ bellpic ] , the superiority of quantum reading is fully preserved both in the absence and presence of decoherence .
optimized over the bandwidth @xmath7 ( i.e. , number of tmsv states ) and the significance level of the test @xmath125 . note that @xmath76 can be higher than @xmath126 bits per cell .
results are shown in the absence of decoherence ( @xmath127 ) considering @xmath128 , @xmath129 and @xmath130 . *
right .
* @xmath76 optimized over @xmath7 and @xmath131 .
results are shown in the presence of decoherence ( @xmath132 ) considering @xmath133 , @xmath134 , @xmath135 and @xmath136.,scaledwidth=46.0% ] _ discussion and conclusion_. the remarkable advantages of quantum reading are related with the regime of few photons , roughly given by @xmath137 photons per bit . note that this is very far from the energy which is used in today s classical readers , roughly given by @xmath138 photons per bit @xcite . in order to fully understand the advantages connected with the few - photon regime ,
let us fix the mean signal power @xmath139 which is irradiated over the cell during the reading time .
this quantity is approximately @xmath140 . here , for fixed power @xmath139
, we can decrease @xmath141 together with @xmath18 .
in other words , the regime of few photons can be identified with the regime of short reading times , i.e. , high data - transfer rates .
thus our results indicate the existence of quantum transmitters which allow reliable _ ultra - fast _ readout of digital memories . besides this result
, we have analogous consequences for the capacity .
in fact , for fixed @xmath139 , we can decrease @xmath11 while increasing @xmath114 .
in other words , the regime of few photons can also be identified with the regime of high frequencies , i.e. , high densities @xcite .
thus we have also proven the existence of quantum transmitters which can read , reliably , _ ultra - dense _ digital memories . the previous physical discussion ( done for fixed signal power @xmath139 )
can be generalized to include the price to pay for generating the light source . given a global initial power @xmath142 ,
we can write @xmath143 , where the conversion factor @xmath144 depends on the source to be generated : typically @xmath145 for classical light , while @xmath146 for non - classical light . then , for fixed @xmath142 , we have reading times @xmath147 and frequencies @xmath148 , where @xmath149 . here
the ratio @xmath150 can still become extremely advantageous for quantum reading thanks to its superiority in the few - photon regime .
for instance , today s cds are classically read using @xmath151 photons per bit , so that @xmath152 @xcite . using spontaneous parametric down conversion in periodically poled ktp waveguides @xcite we can generate epr correlations around @xmath153 nm with @xmath154 . exploiting this quantum source in the few - photon regime , e.g. , using @xmath155
, we can get @xmath156 , thus realizing shorter reading times already with current technology .
clearly this improvement will become dramatic as technology provides cheaper ways to create non - classical light ( see , e.g. , the recent achievements of ref .
@xcite ) . besides these potentialities
, there is another advantage provided by quantum light in the few - photon regime : the non - destructive readout of dye - based optical disks @xcite .
for instance , if we try to increase the capacity of these devices using frequencies above the visible range , the organic layers may undergo a rapid uv - degradation @xcite . in this scenario ,
faint quantum light can retrieve the data safely , whereas classical light can only be destructive . from this perspective
, our work shows new possible directions in the technology of organic memories . _
acknowledgments_. s.p . was supported by a marie curie action of the european community .
s.p . thanks s. l. braunstein , a. aspuru - guzik , s. lloyd , j. h. shapiro , r. nair , r. munoz - tapia , c. ottaviani , n. datta , m. mosheni , j. oppenheim , and c. weedbrook for helpful discussions .
* appendix * in this appendix , we start by providing some introductory notions on bosonic systems and gaussian channels ( see following sec . [ app_intro ] ) .
then , we explicitly connect our memory model with the basic problem of quantum channel discrimination ( see following sec . [ app_connection ] ) .
this connection is explicitly shown in two paradigmatic cases : the basic pure - loss model , where the memory cell is represented by a conditional - loss beam - splitter subject to vacuum noise , and the more general thermal - loss model , where thermal noise is added in order to include non - trivial effects of decoherence . in the subsequent sections we consider both these models and we provide the basic tools for making the quantum - classical comparison . in particular , in sec .
[ app_class_discr ] , we provide the fundamental bound for studying classical transmitters , i.e. , the classical discrimination bound . then , in the next sec .
app_epr_tra , we review the mathematical tools for studying eprtransmitters . by using these elements , we compare epr and classical transmitters in sec .
[ app_comparison ] , where we explain in detail how to achieve the main results presented in our letter .
in particular , for the pure - loss model , we can provide analytical results : the threshold energy and ideal memory theorems , which are proven in sec .
app_pure_loss . finally , in sec .
[ app_bell ] , we show how the advantages of quantum reading persist when the optimal output detection is replaced by an easy - to - implement sub - optimal receiver , consisting of a continuous variable bell measurement followed by a simple classical processing . |
in this work @xcite , we are interested in the effect that axion - photon mixing can have on the polarisation of light coming from distant astronomical sources . in particular , the observations of redshift - dependent large - scale coherent orientations of agn polarisation vectors can , at least qualitatively , even in very simple models , be reproduced as a result of such a mixing of incoming photons with extremely light axion - like particles in external magnetic fields .
these observations , presented in the second edition of this conference , were based on good quality measurements of the linear polarisation for a sample of 355 measured quasars in visible light @xcite .
this has been discussed in terms of axion - photon mixing by several authors , in the case of plane waves @xcite and a prediction from this mixing is an observable circular polarisation comparable to the linear one . here , we present the case in which light is described by wave packets and show that the circular polarisation can be suppressed with respect to the plane wave case .
the mixing of axion - like particles with photons is usually discussed mathematically in terms of infinite plane waves . using that description ,
the stokes parameters can be computed and predictions of the polarisation of light from the interaction can be given ; the main properties of such a mixing being _
dichroism _ and _ birefringence _ ( see @xcite for a review of this case ) .
while dichroism would be an interesting way to produce linear polarisation and , in particular , to explain the observations concerning quasars , birefringence which is linked to the creation of circular polarisation would give a very clear signature of the mixing .
indeed , in this formalism of plane waves , except in extremely specific cases , the _ circular polarisation _ predicted _ _ can be as large as the linear polarisation__. the idea discussed here is to send wave packets into a region of uniform magnetic field and to compute the stokes parameters . before the magnetic field ,
the wave packets have the form : @xmath0 where @xmath1 is the plasma frequency of the medium and @xmath2 controls the initial width of the packet ( in the limit @xmath3 , this reduces to the plane wave case ) .
the main motivation for considering this formalism comes from the measurements of circular polarisation of some of the quasars considered in @xcite . while axion - photon mixing would be an attractive explanation of the observations for linear polarisation ,
preliminary results show that circular polarisation of light from these agn seems to be , in general , _ much smaller _ than the linear polarisation @xcite .
this means that if the creation of circular polarisation was really a smoking gun of alp - photon mixing , no matter how refined the description , these observations would rule out the mixing mechanism and could only be used to constrain the existence of axion - like particles . for these reasons
, it can be interesting to work with wave packets , as new effects will be taken into account , including dispersion , separation of packets and coherence ; effects that might be of importance for the stokes parameters .
note that calculating the propagation of packets of the form is numerically tricky , as the computation of the stokes parameters requires a spatial resolution of the order of the width of the wave packets after a propagation over huge distances in the magnetic field ( we will usually consider one magnetic field zone of 10 mpc @xcite and initial wave packets of width @xmath4 m ) . in the plane transverse to the direction of propagation , we choose a basis of two orthogonal linear polarisations , the same as the one used in the plane wave case , so that we will talk about polarisation parallel or perpendicular to the transverse external magnetic field @xmath5 .
this being done , we next choose the electric fields @xmath6 and @xmath7 both initially described by a function of the form .
then , we propagate these using the equations of motion for the electromagnetic field which take into account the interaction with pseudoscalar particles and find the expressions of the electric fields after a propagation , when axion - photon mixing is at work , inside a step - like magnetic field region . in an external magnetic field , in a strong mixing case here , the axion mass is @xmath8 ev , @xmath9 ev and @xmath10 ev .
the initial width of the wave packet has been chosen @xmath11.,scaledwidth=69.0% ] we can then use the expressions of the stokes parameters which are observables built on intensities that can , for example , be plotted as functions of @xmath12 , the distance travelled inside the magnetic field , for a given propagation time , @xmath13 .
this is what is represented in figure [ fig : packets ] which shows what the two packets look like ( respectively @xmath14 and @xmath15 ) but also the total intensity ( which is just the sum of the two ) and the unnormalised circular polarisation , @xmath16 .
this is for a beam with a central wavelength @xmath17 nm , initially 100% linearly polarised , with its polarisation plane making a 45@xmath18 angle initially with the magnetic field direction ( i.e. @xmath19 ; @xmath20 ) and @xmath21 are the two stokes parameters required to describe fully the linear polarisation of a light beam , while @xmath22 accounts for the circular polarisation . ] ; this angle being , in fact , the most favourable one for the creation of circular polarisation , due to birefringence . note also that the abscissa is @xmath23 , the shift in position with respect to a frame moving a the speed of light @xmath24 ( namely , a maximum at @xmath25 corresponds to @xmath26 ) . from the observational point of view
, there is a macroscopic exposure time over which one should integrate these functions to obtain , finally , the value of the observable stokes parameters , e.g. : @xmath27 from these integrals , we obtain that the wave packet formalism leads to a circular polarisation , @xmath28 , _ lowered _ with respect to plane wave case .
figure [ fig : circpol ] illustrates the plane wave ( @xmath29 ) result : it shows the amount of circular polarisation gained due to axion - photon mixing with different values of the coupling @xmath30 ( @xmath31 being the axion - photon coupling constant ) . in that simpler case , it is known that @xmath28 oscillates between @xmath32 and @xmath33 , whereas in the wave packet case it is shown that there is a damping of these oscillations .
it follows from this observation that _
@xmath22 is no longer expected to be as large as the linear polarisation in general_. .
@xmath34 , @xmath1 and @xmath2 , for the wave packet are the same as in figure [ fig : packets ] , other parameters are : @xmath35 mpc , @xmath36 ev . , scaledwidth=69.0% ]
we have briefly presented axion - photon mixing with the use of wave packets .
the main consequence of this treatment is the net decrease of circular polarisation with respect to what is predicted using plane waves .
hence , the lack of circular polarisation in the light from agn does not rule out the alp - photon mixing .
a. p. would like to thank the iisn for funding and to acknowledge constructive discussions on physical and numerical matters with fredrik sandin and davide mancusi .
99 a. payez _ et al .
_ , in preparation .
d. hutsemkers , a&a * 332 * 410 ( 1998 ) ; d. hutsemkers and h. lamy , a&a * 367 * 381 ( 2001 ) ; d. sluse _
_ , a&a * 433 * 757 ( 2005 ) ; d. hutsemkers _ et al . _ , a&a * 441 * 915 ( 2005 ) .
see , for example , s. das _
_ , jcap * 0506 * 002 ( 2005 ) ; d. hutsemkers _
et al . _
, to appear in the proceedings of the astronomical polarimetry 2008 , july 6 - 11 , 2008 , quebec ( canada ) , p. bastien and n. manset ( eds . ) , asp conf .
series [ arxiv:0809.3088 ] ; c. burrage , a.c . davis and d.j .
shaw , phys .
rev d * 79 * 044028 ( 2009 ) .
a. payez , j.r .
cudell and d. hutsemkers , aip conf .
1038 * 211 ( 2008 ) [ arxiv:0805.3946 ] .
d. hutsemkers _
et al . _ , in preparation .
valle , astron .
j. * 99 * 459 ( 1990 ) . | the measurements of the linear polarisation of visible light from quasars give strong evidence for large - scale coherent orientations of their polarisation vectors in some regions of the sky .
we show that these observations can be explained by the mixing of the photons with very light pseudoscalar ( axion - like ) particles in extragalactic magnetic fields during their propagation .
we present a new treatment in terms of wave packets and discuss the circular polarisation . |
the long range ordering of a compound with antiferromagnetic interactions between spins can be hindered by certain geometric arrangements of the magnetic ions @xcite .
this phenomena of geometric frustration is known to occur for a number of different structures with magnetic ions patterned in specific highly symmetric arrangements in two or three dimensions . for
two - dimensional structure types , both triangular and kagom lattices of magnetic atoms can lead to geometric frustration .
good examples of frustration due to triangular lattices can be found in delafossite - type compounds such as licro@xmath4 @xcite and natio@xmath4 @xcite , while frustration has been extensively studied in kagom lattice compounds such as srcr@xmath6ga@xmath7o@xmath5 @xcite which crystallize in the magnetoplumbite structure .
although a greater degree of lattice complexity has been observed for three - dimensional geometrically frustrated antiferromagnets , the basic building blocks of these 3d lattices are triangles and tetrahedra .
some compounds that are frustrated because of their three - dimensional lattice are the garnet gd@xmath1ga@xmath8o@xmath5 @xcite , the spinel zncr@xmath4o@xmath7 @xcite , the fcc compound k@xmath4ircl@xmath9 @xcite , and the pyrochlore dy@xmath4ti@xmath4o@xmath0 @xcite .
la@xmath0ru@xmath1o@xmath2 has a rhombohedral structure closely related to that of monoclinic la@xmath3ru@xmath4o@xmath5 , a compound previously reported to be a strongly geometrically frustrated antiferromagnet @xcite . despite the structural similarities ,
la@xmath0ru@xmath1o@xmath2 is at most only moderately frustrated , with a frustration index of @xmath10 = -@xmath11 @xmath12 6 .
ru atoms in the @xmath13 plane of both compounds are patterned in a triangular lattice .
each in - plane triangle of ru atoms is capped by a fourth ru atom in a neighboring plane , giving rise to a tetrahedral arrangement of ru atoms in three dimensions , as illustrated in fig .
[ rutetra ] .
there are three distinct ru sites in the crystal structure .
la@xmath0ru@xmath1o@xmath2 was previously found to have a moment of 3.49 @xmath14 , and a curie - weiss constant of -58k which indicates medium strength antiferromagnetic interactions between spins@xcite . in the course of that initial investigation of la@xmath0ru@xmath1o@xmath2 , behavior more complex than that expected for a simple antiferromagnet was observed .
here we report a more detailed study of the magnetic behavior of la@xmath0ru@xmath1o@xmath2 , and show that it has an unusual magnetic phase diagram at low temperatures .
starting materials were dried la@xmath4o@xmath1 ( 99.99% , alfa ) and ruo@xmath4 ( 99.9% , cerac ) .
la@xmath0ru@xmath1o@xmath2 was made by mixing la@xmath4o@xmath1 and ruo@xmath4 in the stoichiometric ratio or with a slight ru excess .
the powders were placed in dense alumina crucibles and heated at 775 , 850 , and 875 @xmath15c for at least two days each with multiple intermediate grindings .
samples were then annealed for about two weeks at 875 @xmath15c with multiple regrindings until judged to be single phase by powder x - ray diffraction .
both ac and dc magnetic susceptibility measurements were performed on the physical property measurement system ( ppms , quantum design ) .
field cooled ( fc ) measurements were done by applying the field , cooling the sample , and then measuring the temperature dependence of the magnetic susceptibility upon heating .
field sweeps were done by ramping the field from 9 t to -9 t to 9 t .
the ac susceptibility measurements were done at a frequency of 10khz with an amplitude of 12oe .
all ac susceptibility measurements made in the presence of a constant dc field were done with field cooling .
la@xmath0ru@xmath1o@xmath2 obeys the curie - weiss behavior at temperatures above 60k for all applied fields studied ( 0 - 9 t ) , as seen in the inset to fig .
[ chidc ] .
there is no difference between the field cooled and zero - field cooled data ( not shown ) , indicating that the low temperature drop in the magnetic susceptibility ( fig .
[ chidc ] , main panel ) represents a transition to a long range ordered antiferromagnetic ground state rather than a spin glass ground state .
although the manner in which the field was applied does not affect the data , the magnitude of the field does .
the maximum in the absolute susceptibility ( @xmath16 = m / h ) is at 14k at an applied field of 1 t , and drops to 9k as the applied field is increased to 9 t .
furthermore , the magnitude of the low temperature dc susceptibility ( @xmath16 = m / h ) increases with increasing applied field in the range of 0 @xmath17 h @xmath17 7 t . for antiferromagnets , increasing the applied field
has no effect on @xmath16 until sufficiently high fields are reached , leading to saturation of the spins and causing @xmath16 to decrease with further increases in h. since there is no simple explanation for the field behavior of @xmath16 , it seems that la@xmath0ru@xmath1o@xmath2 may have multiple types of magnetic ordering . in order to better understand the atypical field behavior of the low temperature susceptibility ,
field sweep measurements were made at a variety of temperatures .
below 15k , two distinct features were present in these m vs. h loops , and at 5k and below , hysteresis was observed .
m vs. h data collected at 2.5k is shown in fig .
[ mvsh ] , with hysteresis clearly visible in both the low field ( @xmath183 t ) and high field ( @xmath185 t ) regimes .
it is therefore apparent that two distinct magnetic phase transitions can be induced in la@xmath0ru@xmath1o@xmath2 by applying a magnetic field .
the two features in the m vs. h data can be more clearly resolved by plotting the differential susceptibility , dm / dh , shown in fig .
the two maxima in the derivative plots mark the location of the low field and high field magnetic phase transitions .
the positive and negative field portions of the plots very accurately mirror each other , confirming the quality of the data .
the close proximity of the low field and high field transitions at 7.5k make it impossible to precisely locate the center of the low field transition . in the lowest temperatures ( 2.5 and 5k ) , the location of both transitions depends on the measurement history . at these temperatures
, there is an upper and a lower phase boundary for both the high field and the low field transition due to the observed hysteresis . a second
, complimentary means of observing the magnetic phase boundaries is measurement of the ac susceptibility , @xmath19 .
measurements of @xmath19 can scan temperature while keeping the applied field fixed , while m vs h scans fix the temperature and vary the field . unlike the dm / dh plots , the @xmath19 data can accurately map the phase boundaries near the zero - field transition temperature .
the results of the ac susceptibility measurements on la@xmath0ru@xmath1o@xmath2 are shown in fig .
[ highac ] .
peaks are observed in two temperature ranges in the @xmath19 scans .
figure [ highac ] shows the higher temperature peak .
it can be seen that the shape of this peak changes from broad and asymmetric to sharp and more symmetric as the field is increased from 0 t to 5 t . above 5 t
the intensity of the peak rapidly decreases .
it can be seen that a low - temperature peak is present for the 3 t data .
a low temperature peak is found only within a narrow window ( @xmath180.25 t ) of applied fields near 3 t ( data not shown ) .
the location of the peak shifts to lower temperatures with increasing field .
the very low temperature of this second ac transition ( @xmath174k ) prevented a detailed determination of the low temperature phase boundary , although one data point could be plotted in the overall magnetic phase diagram . combining
the dm / dh and the @xmath19 data allows the magnetic phase diagram of la@xmath0ru@xmath1o@xmath2 to be drawn ( fig .
two definite magnetic phases are labelled in the diagram .
the dm / dh data clearly define the horizontal field boundaries between the different phases , while the ac susceptibility data provide the best measure of the transition temperatures of the phases .
we note that for ac susceptibility measurements with applied fields of @xmath20 2 t , the transition temperature was taken to be the point of maximum slope in @xmath19 vs. t rather than the peak maximum due to the broad shape of the peak at low fields .
even though the highest field ( h @xmath21 6 t ) measurements of @xmath19 still show a peak maximum , these points were not plotted on the phase diagram due to the broader and weaker character of the peaks .
similarly , the peaks observed in dm / dh plots for t @xmath21 10k were no longer sharp and were not plotted in the phase diagram .
the observation of weak features outside the plotted limits of the magnetic phase diagram reflects the difficulty of defining of sharp boundaries for magnetic phase transition , which often involve continuous and gradual spin rearrangements in addition to exhibiting short range correlations at temperatures just above the onset of long - range order .
since the both the low field ( 3 t ) and high field transition ( 5 t ) involve a jump in the magnetization due to the increasing magnetic field , these transitions should involve spin reorientation , with spins moving from an axis defined by the crystal fields or the antiferromagnetic interactions to an axis defined by the magnetic field . from previously published structural data @xcite ,
it is known that there are two different types of layers in la@xmath0ru@xmath1o@xmath2 .
the thicker layer has a 2.67:1 la : ru ratio while the thinner layer has a 2:1 la : ru ratio and has two inequivalent ru sites
. the different crystal environments may give rise to crystals fields of different strengths on the different ru spins , with the effect that different magnetic field strengths are necessary to dislodge the spins from the crystal field in each type of ru site .
it should be noted that since @xmath22 is 18 k for an applied field of 9 t , the energy scale of the applied field is exceeds that of the antiferromagnetic ordering ( t@xmath23 10 k ) , where @xmath24 is the moment of @xmath25 = 3/2 ru , the @xmath26-factor @xmath12 2 , h is the applied field , and @xmath27 is the boltzmann constant .
even though our data is from a powder sample , it is unlikely that the two transitions represent two different orientations of the grains in our powder since no evidence of preferred orientation was observed in previous structural studies @xcite .
although it is interesting to speculate on the nature of the different magnetic phases , it is difficult to make definite conclusions about the precise nature of the magnetic ordering in the absence of neutron diffraction data or single crystal magnetization measurements .
this research was supported by the nsf solid state chemistry and polymers program , grant no . dmr-9725979 .
thanks go to b. sales for insightful comments on the manuscript .
octahedra ( gray ) showing the triangular arrangement of ru atoms in two dimensions .
ru atoms are at the center of the ruo@xmath9 octahedra while o atoms are at the vertices .
ru atoms in the next higher plane ( shown in black ) cap the ruo@xmath9 octahedra , giving rise to a tetrahedral arrangement of ru sites in three dimensions .
bottom : in - plane view , highlighting the tetrahedral arrangement.,width=672 ] | rhombohedral la@xmath0ru@xmath1o@xmath2 can be considered to be nearly geometrically frustrated due to its close structural similarity to the strongly geometrically frustrated compound la@xmath3ru@xmath4o@xmath5 .
the magnetic ordering of la@xmath0ru@xmath1o@xmath2 was explored using a combination of dc and ac magnetic susceptibility measurements .
the magnetic phase diagram shows two different ordering regimes .
a low field magnetic phase occurs at applied fields of 0 - 3 t and temperatures below 10k , while a high field phase is found at fields of 3 - 5 t and temperatures below 9k . |
asteroseismology of solar - like stars is developing rapidly , driven by the successes of the space telescopes corot @xcite and _ kepler _ @xcite . by studying the oscillation modes of these stars , inferences can be made about their interior structures ( e.g. * ? ? ? * ) . except for the most basic analyses , it is crucial to identify the oscillation modes , that is , the radial order @xmath1 and the spherical degree @xmath2 . in the sun and similar stars ,
mode identification is straightforward thanks to the distinctive pattern of alternating odd and even modes in the power spectrum .
this pattern consists of a regular sequence of @xmath3 modes , interspersed with close pairs of @xmath4 and 2 modes that fall approximately halfway in between .
however , stars significantly hotter than the sun have large linewidths that blur the @xmath5 pairs and make mode identification very difficult . in this letter
we demonstrate a solution to this problem that applies the method proposed by @xcite and @xcite , which uses the _
absolute _ frequencies of the oscillation modes .
for main - sequence stars , the frequencies are well - approximated by the asymptotic relation @xcite , @xmath6 here , @xmath7 is the large separation between modes of the same degree @xmath2 and consecutive order @xmath1 , @xmath8 is the small separation between modes of different degree and @xmath0 is a dimensionless offset , which we discuss in greater detail below . typically , only modes of @xmath9 are observed in intensity due to cancellation over the unresolved stellar disk , although @xmath10 modes can be observed in the highest signal - to - noise targets , such as 11 _ kepler _ subgiants for which frequencies have been determined by @xcite and the solar analogs 16 cyg a and b @xcite .
the asymptotic relation makes it easy to determine the mode degrees for the sun and similar stars . each @xmath4 mode
is separated by @xmath11 from an @xmath12 , and separated by @xmath13 from the @xmath3 mode of the same order .
an example is shown in figure [ fig1]a for the _ kepler _ star kic 6933899 , which has an effective temperature of 5840k .
mode linewidth increases with effective temperature @xcite , reflecting shorter mode lifetimes in hotter stars . in some f - type stars , the linewidths become so large that the pairs of @xmath4 and @xmath12 modes are unresolved and it becomes difficult to distinguish between the blended @xmath5 modes and the @xmath3 modes .
this problem was first observed by corot , in the f5 star hd49933 @xcite and has since been seen in other corot stars @xcite and in the bright f5 star procyon @xcite .
we also see it in many _ kepler _ stars and figure
[ fig1]b shows one example , kic 2837475 ( @xmath14k ) . one way to resolve
this identification problem is to fit both possible mode identifications and compare the relative likelihoods of the two scenarios @xcite .
this relies on the profile of the even-@xmath2 modes being significantly broader and also asymmetric , relative to the @xmath3 modes ( owing to the presence of the smaller amplitude @xmath12 modes at a slightly lower frequency than the @xmath4 modes ) .
the correct scenario should provide a better fit to the power spectrum .
difficulties arise at low signal - to - noise and with short observations , for which the lorentzian mode profiles are not well resolved .
this method was first applied by @xcite , who fitted both scenarios for hd49933 . however , with additional data their preferred mode identification was overturned by @xcite .
other methods have been suggested that utilize the sign of the small separation @xmath15 @xcite . in main sequence stars like the sun
, @xmath15 is known to be positive . however , in many red giants @xmath15 is found to be negative @xcite , so at some point in the evolution the sign must flip @xcite . to further complicate matters , the value of @xmath15 is quite small . at low signal - to - noise , it may be difficult to obtain frequencies precise enough to determine the sign of @xmath15 reliably .
@xcite have suggested that scaling the frequencies of a star with a known mode identification could reveal the correct mode identification in a second star .
this method seeks to use information contained within the value of @xmath0 .
for this to be effective , @xmath0 must vary slowly as a function of stellar parameters .
this is indeed the case , with a tight relationship between @xmath0 and effective temperature , @xmath16 , found both in models @xcite and in observations of sun - like stars @xcite .
thus , @xmath0 promises to be an effective way to determine mode identifications , since the difference in the value of @xmath0 for the two possible scenarios is large ( 0.5 ) .
the existence of a relation between @xmath0 and @xmath16 is not surprising .
the value of @xmath0 is determined by the upper and lower turning points of the acoustic waves ( e.g. * ? ? ?
* ) . as such
, @xmath0 is heavily dependent upon the stellar atmosphere , of which @xmath16 is a significant parameter . due to inadequate modeling of the near - surface layers ,
there is a well - known offset between observed and computed oscillation frequencies in the sun @xcite and also in other stars @xcite .
this offset results in the computed @xmath0 being smaller than observed , typically by @xmath17 as inferred from the displacement of model tracks from observations in the @xmath0 diagram .
the purpose of this letter is to extend the relationship between @xmath0 and @xmath16 to higher temperatures , and thereby make reliable mode identifications in f - type stars .
we used observations of solar - like oscillations in 163 stars taken with the _ kepler _ space telescope between may 2009 and march 2011 ( quarters 18 ) .
each star was observed in _
kepler s _ short - cadence mode ( 58.9 s sampling ) for part of this period .
the time series were prepared from the raw observations as described by @xcite and further corrected to remove outliers and jumps as described by @xcite .
effective temperatures were determined from sdss @xmath18 color - temperature relations by @xcite .
spectroscopic temperatures have also been determined for 77 stars in our sample by @xcite . in almost all cases
the photometric and spectroscopic temperatures agree , except for several stars where the disagreement may be due to differing metallicities or unresolved binaries ( the temperatures for kic3424541 , 3456181 , 4638884 , 6679371 , 7976303 , 8938364 , 9908400 , 10018963 , 10124866 and 10963065 were found to disagree by more than 3@xmath19 ) .
figure [ fig2 ] shows a modified hr diagram of this sample , where we have used large separation instead of luminosity . to measure the value of @xmath0 for each star
, we first determined the frequencies of the @xmath4 modes . where it was possible to resolve the @xmath4 and @xmath12 modes , we measured the frequencies of the @xmath4 modes from the peak in the power spectrum after smoothing .
this was possible for 115 stars .
an example is shown in figure [ fig1]a . where it was not possible to resolve the @xmath4 and @xmath12 modes , and therefore not possible to easily determine the correct mode identification , we determined @xmath0 for both scenarios ( 43 stars ) . in this case
we used the frequencies of the ridge centroids , determined from the peaks of the heavily smoothed power spectrum , as shown in an example f star in figure [ fig1]b .
there were also two cooler stars for which the mode identification was not clear ( kic11401708 and 12555505 ) ; these have low signal - to - noise and the @xmath12 modes are not apparent . for five stars with blended @xmath4 and @xmath12 modes it was still possible to make an unambiguous mode identification ( kic6064910 ,
6766513 , 7668623 , 7800289 and 8026226 ) . in these stars , avoided crossings ` bump ' the @xmath3 modes from their asymptotically expected position @xcite , revealing the correct identification . in these five cases
, the frequencies were obtained from the centroids of the @xmath5 ridge .
the values of @xmath7 and @xmath0 were obtained from a weighted least - squares fit to the @xmath4 frequencies , as described by @xcite .
the weights were given by a gaussian function centered at the frequency of maximum power , @xmath20 , with a fwhm of @xmath21 . by the asymptotic relation , equation [ asymp ] , the gradient of this fit is @xmath7 and the y - intercept is @xmath22 . to confirm the validity of our method of frequency determination , we compared the values of @xmath0 derived from our frequencies to those derived from frequencies that have been determined by more traditional ` peak - bagging ' methods in the 61 stars for which this has been done @xcite .
we found good agreement , with @xmath0 values typically agreeing to within 0.1 , which is approximately the size of the typical uncertainty .
this agreement is of particular importance for hotter stars because using ridge centroids instead of @xmath4 frequencies potentially biases @xmath0 towards slightly lower values .
we found this bias to be negligible within the uncertainties .
in figure [ fig3]a we show the so - called @xmath0 diagram for the 120 _ kepler _ stars whose mode identifications were unambiguous .
the observed values of @xmath0 are clearly offset to the right of the models which , as mentioned above , arises from the improper modeling of the near - surface layers . from this figure
it appears that the offset in @xmath0 may be roughly the same for all stars , which corresponds to a fixed fraction of the large separation , as was also inferred by @xcite . in figure
[ fig4]a we show @xmath0 versus @xmath23 for our sample .
the sun is marked in black by its usual symbol , and the stars with unambiguous mode identifications are colored blue .
the trend of decreasing @xmath0 with increasing @xmath16 is clearly seen in these stars .
for the 43 _ kepler _ stars whose mode identifications are uncertain , the relationship between @xmath0 and @xmath16 can help . in figure
[ fig4]a these stars are shown in gray for both scenarios .
owing to the potential ambiguity in @xmath1 , each scenario is also plotted shifted by @xmath24 .
we also include both scenarios of the f stars procyon @xcite , hd 49933 @xcite and hd 181420 @xcite . for most of the stars with ambiguous mode identifications ,
only one of the scenarios lies along the @xmath0@xmath16 trend defined by the stars with secure identifications and we adopt this as the correct one . for the previously studied f stars , we prefer scenario b of procyon @xcite , scenario b of hd 49933 @xcite and scenario 1 of hd 181420 @xcite . due to the width of the @xmath0@xmath16 relation ,
there are a few stars for which the situation is still somewhat ambiguous ( red circles in figure [ fig4]a ) .
the two scenarios in these stars have values of @xmath0 that fall towards the top and bottom of the relation . for these stars we must use additional information to resolve the ambiguity .
to overcome the spread in the @xmath0@xmath16 relation , the value of @xmath7 is useful . for a given effective temperature , stars with higher masses have smaller values of @xmath7 , as can be seen from the models in figure [ fig2 ] .
models also indicate that higher - mass stars have a smaller @xmath0 ( see figure [ fig3]a and * ? ? ?
it follows that in the @xmath0@xmath16 plane , stars that fall towards the bottom of the trend will be more massive and should therefore have smaller values of @xmath7 than lower - mass stars of similar temperature .
we illustrate this in the @xmath7@xmath0 plane in figure [ fig3]b , where we show stars in which the identification was already obvious , as well as those for which the identification could be readily made from the @xmath0@xmath16 relation .
symbols are colored according to @xmath16 , with the trend of decreasing @xmath16 with increasing @xmath0 quite clear . the gradual decrease in @xmath0 with decreasing @xmath7 along lines of constant temperature
is also visible . by comparing a star
whose identification is still ambiguous with stars of a similar large separation , we can better make a decision on the identification .
we show the example of kic 11290197 , whose two scenarios are circled in figures [ fig3]b and [ fig4]a . by comparing its @xmath16 with stars of a similar @xmath7 , we establish the preferred scenario , which in this case has the lower value of @xmath0 .
other seismic parameters may also be useful .
the method of @xcite , which uses the value of the small separation @xmath15 , agrees for all stars , except one ( kic5431016 ) , which has a low signal - to - noise .
another very useful parameter is mode linewidth ( @xmath25 ) .
as mentioned above , linewidth is strongly correlated with effective temperature , and so there should also be a correlation between linewidth and @xmath0 .
linewidths for 41 of our stars ( including 12 with ambiguous identifications ) were previously measured by @xcite .
using the syd method described in that paper , we have measured linewidths in a further 26 stars with ambiguous identifications ( excluding the five ambiguous stars with the lowest signal - to - noise ) .
figure [ fig4]b shows the relation between @xmath25 and @xmath0 for these 67 stars .
we can quantify the likelihood of one scenario over the other by comparing how far each lies from the various relations .
to do so , we first performed a bayesian linear fit to the @xmath0@xmath16 and @xmath0@xmath26 relations for all stars with @xmath27 k with unambiguous identifications ( i.e. the blue points in figures [ fig4]a and b ) .
for all stars with an uncertain identification , we then calculated the likelihood of obtaining the observed @xmath16 for each of the two possible values of @xmath0 ( @xmath28 and @xmath29 ) , the observed @xmath25 , the parameters of the linear fits , and all respective uncertainties , @xmath0 and @xmath26 ; uncertainties for the linear parameters were determined from the marginal posterior of the linear fit . ] .
full details of this method will be provided in a future paper ( m. gruberbauer et al .
2012 , in prep . ) .
the required integration over the parameter space was carried out using multinest @xcite .
these two likelihood values , @xmath30 and @xmath31 , were then used to calculate the bayes factor ( ratio of the likelihoods ) , and hence the odds ratio and probability of each scenario , @xmath32 and @xmath33 , assuming equal prior probability for both identifications .
we performed this calculation using both photometric and spectroscopic effective temperatures , where available .
our preferred identification is the one with the greatest probability .
we denote the value of @xmath0 for the preferred identification as @xmath34 , and use @xmath35 for the alternate value . in table
[ tbl1 ] we list for each star its measured @xmath7 , the value of @xmath34 and @xmath35 , @xmath25 , both effective temperatures , @xmath36 and @xmath37 , and the probabilities , @xmath38 and @xmath39 , of our preferred scenario . for most of the stars we find strong support for our preferred scenario .
one star ( kic1725815 ) , without a measured @xmath37 , is an outlier in the @xmath0@xmath16 plane .
while the linewidth supports one scenario , @xmath36 supports the alternate scenario .
it is not entirely clear which is correct , although the method of @xcite agrees with the alternate scenario favored by @xmath36 .
spectroscopic measurement of the temperature may help resolve this case .
the @xmath0@xmath16 relation for the final selected identification of all stars in our sample is shown in figure [ fig4]c .
lrrrccccc 1430163 & 85.22 @xmath40 0.39 & 1.08 @xmath40 0.09 & 1.58 @xmath40 0.06 & @xmath41 & 6796 @xmath40 78 & 6520 @xmath40 60 & 99.7 & 99.3 + 1725815 & 55.89 @xmath40 0.20 & 1.51 @xmath40 0.06 & 1.07 @xmath40 0.06 & @xmath42 & 6550 @xmath40 82 & & 73.2 & + 2837475 & 75.97 @xmath40 0.14 & 0.84 @xmath40 0.04 & 1.44 @xmath40 0.06 & @xmath43 & 6688 @xmath40 57 & 6700 @xmath40 60 & 99.9 & 99.9 + 2852862 & 53.46 @xmath40 0.18 & 1.10 @xmath40 0.06 & 1.55 @xmath40 0.05 & @xmath44 & 6417 @xmath40 58 & & 89.7 & + 3424541 & 41.58 @xmath40 0.13 & 1.32 @xmath40 0.12 & 0.81 @xmath40 0.05 & @xmath45 & 6475 @xmath40 66 & 6080 @xmath40 60 & 73.2 & 86.7 + 3456181 & 52.02 @xmath40 0.15 & 1.04 @xmath40 0.05 & 1.45 @xmath40 0.04 & @xmath46 & 6732 @xmath40 91 & 6270 @xmath40 60 & 98.9 & 97.7 + 3643774 & 76.15 @xmath40 0.15 & 1.29 @xmath40 0.06 & 0.75 @xmath40 0.04 & @xmath47 & 6029 @xmath40 104 & & 100.0 & + 3733735 & 92.59 @xmath40 0.41 & 0.88 @xmath40 0.09 & 1.43 @xmath40 0.09 & @xmath48 & 6711 @xmath40 66 & 6715 @xmath40 60 & 99.8 & 99.8 + 3967430 & 88.06 @xmath40 0.25 & 0.93 @xmath40 0.06 & 1.50 @xmath40 0.08 & @xmath49 & 6622 @xmath40 53 & & 99.4 & + 4465529 & 72.70 @xmath40 0.23 & 1.09 @xmath40 0.06 & 1.58 @xmath40 0.03 & @xmath50 & 6345 @xmath40 49 & & 99.7 & + 4586099 & 61.42 @xmath40 0.22 & 1.11 @xmath40 0.06 & 1.53 @xmath40 0.05 & @xmath51 & 6383 @xmath40 58 & 6296 @xmath40 60 & 99.2 & 99.0 + 4638884 & 60.46 @xmath40 0.18 & 1.04 @xmath40 0.06 & 1.35 @xmath40 0.06 & @xmath52 & 6662 @xmath40 57 & 6375 @xmath40 60 & 95.9 & 93.4 + 4931390 & 93.07 @xmath40 0.59 & 0.79 @xmath40 0.13 & 1.39 @xmath40 0.04 & @xmath53 & 6557 @xmath40 51 & & 92.8 & + 5431016 & 48.92 @xmath40 0.21 & 1.21 @xmath40 0.07 & 0.86 @xmath40 0.07 & @xmath54 & 6601 @xmath40 62 & & 93.7 & + 5516982 & 83.80 @xmath40 0.28 & 1.14 @xmath40 0.07 & 1.53 @xmath40 0.06 & @xmath55 & 6444 @xmath40 50 & & 89.3 & + 5773345 & 57.28 @xmath40 0.15 & 1.08 @xmath40 0.05 & 1.52 @xmath40 0.04 & @xmath56 & 6214 @xmath40 61 & 6130 @xmath40 60 & 95.7 & 93.5 + 6508366 & 51.29 @xmath40 0.07 & 1.11 @xmath40 0.02 & 1.48 @xmath40 0.06 & @xmath57 & 6499 @xmath40 46 & 6354 @xmath40 60 & 98.2 & 97.7 + 6679371 & 50.69 @xmath40 0.16 & 0.86 @xmath40 0.05 & 1.34 @xmath40 0.05 & @xmath58 & 6598 @xmath40 59 & 6260 @xmath40 60 & 83.7 & 65.1 + 7103006 & 59.34 @xmath40 0.23 & 1.08 @xmath40 0.07 & 1.49 @xmath40 0.06 & @xmath59 & 6421 @xmath40 51 & 6394 @xmath40 60 & 95.7 & 96.3 + 7206837 & 78.69 @xmath40 0.17 & 1.17 @xmath40 0.04 & 1.60 @xmath40 0.05 & @xmath60 & 6392 @xmath40 59 & 6304 @xmath40 60 & 99.6 & 99.5 + 7282890 & 45.27 @xmath40 0.25 & 0.90 @xmath40 0.09 & 1.43 @xmath40 0.12 & @xmath61 & 6341 @xmath40 47 & 6384 @xmath40 60 & 74.7 & 75.4 + 7529180 & 85.89 @xmath40 0.28 & 0.96 @xmath40 0.07 & 1.52 @xmath40 0.12 & @xmath62 & 6682 @xmath40 58 & 6700 @xmath40 60 & 82.2 & 81.2 + 7771282 & 72.55 @xmath40 0.23 & 1.09 @xmath40 0.06 & 1.52 @xmath40 0.07 & @xmath63 & 6407 @xmath40 74 & & 95.7 & + 7940546 & 58.67 @xmath40 0.14 & 1.09 @xmath40 0.04 & 1.56 @xmath40 0.04 & @xmath64 & 6350 @xmath40 111 & 6264 @xmath40 60 & 94.9 & 94.1 + 8360349 & 41.04 @xmath40 0.15 & 0.97 @xmath40 0.06 & 1.54 @xmath40 0.09 & & 6258 @xmath40 49 & 6340 @xmath40 60 & 57.7 & 67.0 + 8367710 & 55.36 @xmath40 0.14 & 1.16 @xmath40 0.04 & 1.54 @xmath40 0.06 & @xmath65 & 6352 @xmath40 66 & 6500 @xmath40 60 & 83.0 & 86.7 + 8579578 & 49.90 @xmath40 0.15 & 1.08 @xmath40 0.05 & 1.43 @xmath40 0.08 & @xmath66 & 6308 @xmath40 45 & 6380 @xmath40 60 & 91.3 & 93.0 + 9206432 & 84.51 @xmath40 0.23 & 1.05 @xmath40 0.06 & 1.48 @xmath40 0.03 & @xmath67 & 6494 @xmath40 46 & 6608 @xmath40 60 & 99.9 & 100.0 + 9226926 & 73.70 @xmath40 0.32 & 0.83 @xmath40 0.08 & 1.14 @xmath40 0.11 & @xmath68 & 7149 @xmath40 132 & 6892 @xmath40 60 & 86.8 & 83.0 + 9353712 & 51.37 @xmath40 0.17 & 1.12 @xmath40 0.06 & 1.57 @xmath40 0.03 & @xmath69 & 6416 @xmath40 56 & & 96.1 & + 9812850 & 64.59 @xmath40 0.22 & 1.10 @xmath40 0.06 & 1.45 @xmath40 0.04 & @xmath70 & 6407 @xmath40 47 & 6325 @xmath40 60 & 98.1 & 98.0 + 9908400 & 36.50 @xmath40 0.17 & 1.00 @xmath40 0.07 & 1.68 @xmath40 0.06 & @xmath71 & 6000 @xmath40 55 & 6400 @xmath40 60 & 62.1 & 85.0 + 10208303 & 62.32 @xmath40 0.37 & 0.90 @xmath40 0.11 & 1.39 @xmath40 0.08 & & 6665 @xmath40 78 & & 73.8 & + 10709834 & 67.98 @xmath40 0.22 & 1.17 @xmath40 0.06 & 0.74 @xmath40 0.06 & @xmath72 & 6754 @xmath40 56 & 6508 @xmath40 60 & 87.7 & 92.0 + 10730618 & 66.16 @xmath40 0.27 & 1.06 @xmath40 0.08 & 1.53 @xmath40 0.08 & @xmath73 & 6422 @xmath40 54 & & 85.5 & + 10909629 & 49.81 @xmath40 0.27 & 0.94 @xmath40 0.09 & 1.47 @xmath40 0.06 & @xmath74 & 6501 @xmath40 61 & & 82.0 & + 11081729 & 90.03 @xmath40 0.20 & 1.03 @xmath40 0.04 & 1.48 @xmath40 0.06 & @xmath75 & 6605 @xmath40 51 & 6630 @xmath40 60 & 99.7 & 99.8 + 11128126 & 77.36 @xmath40 0.28 & 1.11 @xmath40 0.12 & 0.73 @xmath40 0.07 & & 6250 @xmath40 55 & & 73.2 & + 11253226 & 77.30 @xmath40 0.20 & 0.81 @xmath40 0.05 & 1.34 @xmath40 0.04 & @xmath76 & 6682 @xmath40 51 & 6605 @xmath40 60 & 99.5 & 99.4 + 11290197 & 74.56 @xmath40 0.47 & 0.85 @xmath40 0.12 & 1.29 @xmath40 0.17 & & 6544 @xmath40 63 & & 58.0 & + 11401708 & 40.00 @xmath40 0.14 & 1.31 @xmath40 0.07 & 0.70 @xmath40 0.06 & @xmath77 & 5859 @xmath40 63 & & 100.0 & + 12069127 & 48.47 @xmath40 0.13 & 1.01 @xmath40 0.04 & 1.49 @xmath40 0.02 & @xmath78 & 6476 @xmath40 66 & & 77.3 & + 12555505 & 108.08 @xmath40 0.61 & 1.52 @xmath40 0.12 & 0.84 @xmath40 0.12 & & 5704 @xmath40 66 & & 91.8 & + procyon & 56.20 @xmath40 0.35 & 1.06 @xmath40 0.11 & 1.58 @xmath40 0.08 & @xmath79 & & 6530 @xmath40 50 & & 93.4 + hd 49933 & 85.53 @xmath40 0.18 & 1.06 @xmath40 0.04 & 1.54 @xmath40 0.05 & @xmath80 & & 6750 @xmath40 130 & & 99.9
+ hd 181420 & 75.20 @xmath40 0.32 & 0.92 @xmath40 0.09 & 1.36 @xmath40 0.07 & @xmath81 & & 6580 @xmath40 105 & & 97.5
we have presented a method to effectively determine the correct mode identification in stars for which this has previously been a problem .
these are the f stars with large linewidths that make it difficult to distinguish the @xmath4 and @xmath12 modes .
this method uses the relationship between effective temperature , mode linewidth and @xmath0 to determine what values of @xmath0 are reasonable for the star , and therefore which of the two possible scenarios is most likely correct .
this method provides robust results in the vast majority of cases because the value of @xmath0 implied by each scenario is very distinct , even in low signal - to - noise stars , representing a major improvement over previous methods .
for the few cases that are still ambiguous , additional information , such as the large separation , @xmath7 , can be included to help resolve the matter .
the authors gratefully acknowledge the _ kepler _ science team and all those who have contributed to the _ kepler mission _ for their tireless efforts which have made these results possible .
funding for the _ kepler mission _ is provided by nasa s science mission directorate .
we acknowledge the support of the australian research council .
trw is supported by an australian postgraduate award , a university of sydney merit award , an australian astronomical observatory phd scholarship and a denison merit award .
mg received financial support from an nserc vanier scholarship .
sh acknowledges financial support from the netherlands organisation of scientific research ( nwo ) . | asteroseismology of f - type stars has been hindered by an ambiguity in identification of their oscillation modes .
the regular mode pattern that makes this task trivial in cooler stars is masked by increased linewidths .
the absolute mode frequencies , encapsulated in the asteroseismic variable @xmath0 , can help solve this impasse because the values of @xmath0 implied by the two possible mode identifications are distinct .
we find that the correct @xmath0 can be deduced from the effective temperature and the linewidths and we apply these methods to a sample of solar - like oscillators observed with _ |
a rigorous theoretical framework within which quarkonium production can be studied is provided by the nonrelativistic qcd ( nrqcd ) factorization formalism developed by bodwin , braaten , and lepage @xcite .
this approach is based on nrqcd @xcite , an effective field theory that can be made equivalent to full qcd to any desired order in heavy - quark relative velocity .
the nrqcd factorization formalism is not a model , but rather a rigorous derivation within nrqcd of a factorized form for quarkonium production and decay rates .
a central result of the nrqcd factorization formalism is that inclusive quarkonium production cross sections must have the form of a sum of products of short - distance coefficients and nrqcd matrix elements .
the short - distance coefficients are associated with the production of a heavy quark - antiquark pair in a specific color and angular - momentum state .
they can be calculated using ordinary perturbative techniques . as to the nrqcd matrix elements , they parameterize the effect of long - distance physics such as the hadronization of the quark - antiquark pair .
these can be determined phenomenologically .
the power of the nrqcd formalism stems from the fact that factorization formulas for observables are essentially expansions in the small parameter @xmath3 , where @xmath4 is the average relative velocity of the heavy quark and anti - quark in the quarkonium bound state .
@xmath5 for charmonium , and 0.1 for bottomonium .
nrqcd @xmath4-scaling rules @xcite allow us to estimate the relative sizes of various nrqcd matrix elements .
this information , along with the dependence of the short - distance coefficients on coupling constants , permits us to decide which terms must be retained in expressions for observables so as to reach a given level of accuracy . generally , to leading order , factorization formulas involve only a few matrix elements , so several observables can be related by a small number of parameters .
nrqcd predictions therefore presuppose qcd , factorization , and a reasonably convergent @xmath3 expansion .
if the quantitative predictions of nrqcd fail , one of these three assumptions is failing , most likely the validity of the @xmath3 expansion .
in many instances of @xmath0 production , the most important nrqcd matrix elements are @xmath6 , @xmath7 , @xmath8 , and @xmath9 .
these four quantities parametrize the hadronization into a @xmath0 boundstate of a @xmath10 pair produced initially with the stated quantum numbers ( angular momentum @xmath11 , color quantum number 1 or 8) . previous to the development of the factorization formalism of ref .
@xcite , most @xmath0 production calculations took into account only the hadronization of @xmath10 pairs initially produced in the color - singlet @xmath12 state , as parameterized by @xmath6 .
recently , in regard to the three most important color - octet matrix elements involved in @xmath0 production , it has been noticed that there prevails a certain inconsistency between the values measured at cdf @xcite and those measured in other processes @xcite .
we propose here that inclusive @xmath0 production from @xmath13-decay can be of avail in unravelling this difficulty by serving as a supplementary arena for measuring the contentious matrix elements .
inclusive production of @xmath0 from @xmath13-decay provides two measurable combinations of the matrix elements .
the first one is the helicity - summed rate @xmath14 . the second combination concerns the polarization parameter @xmath2 appearing in the electromagnetic decay rate of @xmath0 to lepton pairs : @xmath15 where the polar angle @xmath16 is defined in the @xmath0 rest frame for which the @xmath17-axis is aligned with the direction of motion of the @xmath0 in the lab .
the short - distance physics for the process @xmath18 is described by the four - quark fermi interactions @xmath19 and @xmath20 , at the @xmath21 scale .
the feynman diagram is shown in fig . 1 .
let us define @xmath22 to be the total four - momentum of the @xmath10 system .
the @xmath23 and @xmath24 quarks are assumed massless .
we first calculate the factorization formula for the spin - summed @xmath0 production rate .
let us define @xmath25 as the relative momentum of the @xmath26 and @xmath27 quarks evaluated in the rest frame of the @xmath10 pair .
the nrqcd matching formalism requires only that we know the algebraic form of the amplitude for small values of @xmath25 .
this is because the hadronization of the heavy quark pair into quarkonium occurs with nonnegligible probability only when @xmath28 .
moreover , this is the regime in which the nrqcd effective lagrangian is valid .
therefore , we expand the feynman amplitude in powers of @xmath25 and keep only pieces up to linear order .
the amplitude for @xmath29 is given by @xmath30 \;\ ; \bar{u}_s\gamma^\mu ( 1 - \gamma_5 ) u_b \nonumber\\ & \times & \bigg [ 2m_c \ ; \ ; \lambda^\mu_i \xi_\sigma^\dagger \ ! \sigma^i \ !
\eta_\tau \ ; - \ ; p^\mu \;\;\xi_\sigma^\dagger \ !
\eta_\tau \ ; + \ ; 2i \lambda^\mu_k\epsilon^{mik } \;\;\xi_\sigma^\dagger \ !
\sigma^i \ !
\bigg ] \bigg ( 1 + o\left ( { { \bf q}^2 \over m^2_c } \right ) \bigg ) \ ; , \end{aligned}\ ] ] where @xmath31 and @xmath32 are non - relativistic two - spinors for the @xmath26 and @xmath27 quarks respectively ; @xmath33 are color indices and @xmath34 and @xmath35 are the individual spins of the charm quarks .
the factors involving @xmath36 and @xmath37 are wilson coefficients which govern the scale - evolution of the four - quark fermi interaction .
@xmath38 ( defined in detail in ref .
@xcite ) is the lorentz boost matrix that takes a three - vector from the @xmath10 rest frame to the lab frame . in the above equation
, the term with the color krnecker deltas contributes to the production of @xmath10 pairs in color - singlet states , and the term with color matrices
@xmath39 contributes to the production of @xmath10 pairs in color - octet states . from eq .
[ bdecayfeynamp - reduced ] , using the matching procedure presented in ref @xcite , we determine the spin - summed rate for @xmath40 in the nrqcd factorization formalism : @xmath41 where , upon summing over the light quarks @xmath23 and @xmath24 , we have used @xmath42 .
we have taken the inital @xmath13-quark to be unpolarized . here
@xmath43 are nrqcd @xmath0 production matrix elements .
we consider only the leading color - singlet piece and the leading color - octet pieces in the relativistic @xmath3-expansion .
our calculation of the coefficients in the above spin - summed factorization formula concurs with the results presented in ref .
@xcite . according to the nrqcd @xmath4-scaling rules , the color - octet matrix elements in eq .
[ totalfacform ] are all expected to be suppressed by @xmath44 with respect to the basic color - singlet matrix element @xmath6 .
thus , at the outset , one would not expect the octet contributions to play a major role in the production of @xmath0 articles .
however , as pointed out in ref .
@xcite , the color - singlet coefficient @xmath45 decreases significantly as it is evolved down from its value of 1 at the scale @xmath46 to its value of roughly 0.40 at the scale @xmath21 . on the other hand ,
the color - octet coefficient @xmath47 increases slightly from 1 at the scale @xmath46 to roughly 1.10 at the scale @xmath21 . for this reason ,
the octet contributions are actually as important as , or more important than , the basic singlet contribution .
in fact , the short - distance coefficients in the color - octet terms are some 50 times larger than those in the singlet term !
in the previous section , we discussed how the serendipitous wilson - coefficient enhancement of the octet contributions enables us to measure the color - octet matrix elements contributing to the helicity - summed rate @xmath48 .
this enhancement can be further exploited to determine the color - octet matrix elements by considering the polarization of the produced @xmath0 s , as measured by the parameter @xmath2 in eq .
[ thetadefined ] .
it can be calculated via the formula @xmath49 where @xmath50 represents the production rate of @xmath0 with helicity @xmath51 . since @xmath2 is a ratio , we need only calculate the relative sizes of the various contributions to the helicity rates .
thus , we need only calculate the effective feynman amplitude squared , evaluated in the @xmath13 rest frame , with the @xmath0 momentum @xmath52 oriented in the positive @xmath17-direction . in this case , the angular momentum of the @xmath0 in the canonical @xmath17-direction corresponds simply to the particle s helicity .
beneke and rothstein @xcite and braaten and chen @xcite have developed techniques for deriving the production rates of quarkonia with specified helicities . applying their methods to the amplitude given in eq .
[ bdecayfeynamp - reduced ] , we obtain @xmath53 \nonumber\\ & + & 3 ( c_+ + c_-)^2 \big\langle{\cal o}^\psi_8({}^3s_1 ) \big\rangle \left [ ( m_b^2 - 4 m_c^2 ) \delta_{\lambda 0 } + 4 m^2 ( 1 - \lambda ) \right ] \nonumber\\ & + & 3 ( c_+ + c_-)^2 \big\langle{\cal o}^\psi_8({}^1s_0 ) \big\rangle
m_b^2 \nonumber\\ & + & 9 ( c_+ + c_-)^2 \frac{\big\langle{\cal o}^\psi_8({}^3p_0 ) \big\rangle}{m_c^2 } \left [ ( m_b^2 - 4 m_c^2 ) ( 1 - \delta_{\lambda 0 } ) + 8 m_c^2 - 4 m_c^2 \lambda \right].\end{aligned}\ ] ] in displaying eq . [ rlambda ] , we have adopted the standard practice of rewriting the matrix element @xmath54 using the relation @xmath55 this latter relation is due to heavy - quark spin symmetry , and is valid up to relative order @xmath3 .
it must be kept in mind , however , that @xmath13-decay does not actually produce @xmath10 states in the @xmath56 configuration at leading order in the coupling constants .
some consequences of angular momentum conservation and of the left - handedness of the charged - current fermi interaction are illustrated in fig .
the thin arrow labelled @xmath57 represents the trajectory of the center of mass of the @xmath10 system .
the other thin arrow represents the trajectory of the @xmath23 or @xmath24 quark .
thick arrows denote intrinsic angular momenta , with @xmath58 denoting the helicity of the @xmath10 system prior to hadronization .
since the charged - current fermi interaction couples only the left - handed parts of the fermion fields , the masslessness of the @xmath23 and @xmath24 quarks ensures that they are emitted with negative helicity .
therefore , due to angular momentum conservation , the helicity outcome @xmath59 is not allowed .
this has certain obvious consequences for the @xmath0 helicity production rates : the basic color - singlet contribution ( parametrized by @xmath6 and involving the direct hadronization of a color - singlet @xmath12 state into a @xmath0 ) is zero for @xmath60 , at lowest order in the relativistic expansion . the same is also true of the @xmath7 contribution , which involves the hadronization of a color - octet @xmath12 state into a @xmath0 via the heavy - quark - spin - preserving @xmath61 emission or absorption of two soft gluons @xcite .
the color - singlet contribution to the helicity rates of the @xmath0 produced in @xmath13-decay was calculated in the color - singlet model by m. wise in ref .
@xcite 16 years ago .
this author presented expressions for the production rates of longitudinal and transverse helicities , given in terms of the color - singlet radial wave function at the origin . the result in ref .
@xcite concurs with the first line of our eq .
[ rlambda ] .
while experimental determinations of helicity - summed @xmath62 have already been carried out @xcite , a measurement of the polarization parameter @xmath2 is not yet available . anticipating the availability of this latter measurement
, it is interesting to determine the range of @xmath2 which is predicted by the nrqcd factorization formalism . using eqs .
[ alphasig ] and [ rlambda ] , we can express @xmath2 in terms of the nrqcd matrix elements . before proceding ,
however , we must first decide which value of @xmath21 to use .
since we are presenting a leading - order calculation , we may in principle choose to use either the pole mass or the @xmath63 mass for @xmath21 , the difference between these two being merely a higher - order - in-@xmath64 effect .
however , it turns out that the leading - order nrqcd prediction of the polarization parameter @xmath2 depends quite strongly on the choice of @xmath21
. therefore we report results for a wide range of @xmath21 , from 4.1 gev to 5.3 gev .
this range includes the values of @xmath65 gev corresponding to the @xmath63 mass , and the values @xmath66 gev corresponding to the pole mass determined on the lattice @xcite .
our range is centered around @xmath67 gev , which is the central value for the pole mass @xcite .
this choice is consistent with the fact that the numerical value that we use for @xmath68 is taken from ref .
@xcite , in which the pole mass was used in the extraction of the matrix element value . for concreteness
, we present in the text below only our results for the choice @xmath67 gev . using @xmath69 , @xmath70 and @xmath71 gev ,
one obtains the leading - order nrqcd prediction for the polarization parameter @xmath72 it must be noted that @xmath2 depends only weakly on @xmath7 and most strongly on @xmath73 .
we now procede to determine the range of @xmath2 which is consistent with existing information on the matrix elements @xmath6 , @xmath7 , @xmath8 , and @xmath74 and with various constraints on linear combinations of these quantities .
we first review this information : * bodwin et al .
@xcite have determined the color - singlet matrix element @xmath6 from decay rates of @xmath0 to lepton pairs to be @xmath75 .
the error reflects theoretical uncertainties due to higher order @xmath64 and @xmath3 corrections .
this phenomenologically extracted value for the color - singlet matrix element is in agreement with the lattice calculation result also presented in ref .
@xcite . * a constraint on the octet matrix elements
is given by the requirement that the theoretical prediction for the spin - summed production rate @xmath14 given in eq .
[ totalfacform ] be consistent with the experimentally measured value of @xmath76 .
this constraint can be expressed as @xmath77 to compute the above experimental value of @xmath78ev , we have used @xmath79 = ( [email protected])% , and @xmath35(averaged over @xmath81 hadrons)=1.54 picoseconds . to compute the short - distance coefficients on the right - hand - side of eq .
[ decayconstraint ] , we have taken @xmath82 gev@xmath83 @xcite and @xmath84 @xcite .
one can attribute an error of roughly 30% to the above equation , to reflect relativistic corrections and uncertainties in the scale used to evaluate the wilson coefficients @xcite .
it must be pointed out that the color - singlet model prediction for spin - summed @xmath0 production obtained by setting all color - octet matrix elements to zero , and using @xmath75 is roughly one - third of the experimental value .
however the nrqcd prediction for the @xmath13-decay width can be made to agree with the experimental measurements for reasonable values of the color - octet matrix elements . *
cho and leibovich @xcite have performed a fit to cdf data over low and high @xmath85 ranges and have found @xmath86 and @xmath87 * from photoproduction data , amundson et al . @xcite
have determined @xmath88 note that the errors quoted in eqs .
[ clmconstraint ] , [ clspconstraint ] , and [ afmbound ] are statistical only , and that the analyses in @xcite and @xcite did not take into account theoretical uncertainties due to next - to - leading order ( nlo ) corrections
. however we expect nlo corrections to be very large ( see @xcite for a discussion of nlo corrections in hadronic @xmath0 production calculations ) .
moreover the photoproduction results of ref .
@xcite may also have significant higher - twist corrections @xcite . * finally , due to @xmath4-scaling arguments
, we expect the color - octet matrix elements to be roughly of order @xmath89 . with this information in mind
, we now determine the range of values of @xmath2 that we can expect .
our method of determining this range consists of scanning " through the three - dimensional parameter space of the color - octet matrix elements , and determining the maximum and minimum value of @xmath2 that occur in the allowed volume .
our scan is subject to the following constraints : we allow @xmath7 to vary in the range @xmath90~{\rm gev}^3 \ ; ; \ ] ] we allow the photoproduction - related linear combination of @xmath91 and @xmath92 , to vary in the range @xmath93~{\rm gev^3}\ ; , \ ] ] where the range we have chosen takes into account theoretical uncertainties due to nlo corrections and possibly significant hight - twist corrections ; we allow the b - decay - related linear combination given in eq .
[ decayconstraint ] to vary @xmath94 about the central value of @xmath95 , _ i.e. _ @xmath96~{\rm gev}^3 \ ; , \ ] ] where the range we have chosen takes into account theoretical uncertainties due to @xmath3 relativistic corrections as mentioned previously .
note that the above constraints are sufficient to insure that the helicity rates in eq .
[ rlambda ] are positive for each value of the helicity @xmath51 . to the above experimental constraints on the allowed parameter space
, we must of course add the theoretical requirement that the absolute values of the octet matrix elements respect @xmath4-scaling rules , _
i.e. _ that they must not be unreasonably larger than @xmath97
. moreover , we must impose on our allowed volume of parameter space that the matrix element @xmath8 always be positive .
no similar constraint need to be applied to the matrix element @xmath98 .
indeed , it can be negative , for the following reason .
the bare matrix element ( which is of course positive definite ) contains a power ultraviolet divergence .
such a power ultraviolet divergence must be proportional to a matrix element of an operator of lower dimension . in the case of @xmath99 ,
the divergence is proportional to @xmath100 @xcite .
this divergence must be subtracted to obtain the renormalized matrix element .
since the piece subtracted is comparable in magnitude to the bare matrix element , the difference can be negative .
on the other hand , since the operator @xmath101 belongs to the set of lowest dimension nrqcd four - fermion operators , the bare matrix element @xmath8 can not have power divergences .
thus the subtractions involved in renormalization can not transform the positive definite bare matrix element into a negative quantity @xcite . due to these considerations
, we therefore impose the additional theoretical constraint @xmath102 the constraints expressed in eqs .
[ clm2 ] , [ afm2 ] , [ decayconstraint2 ] and [ s - constraint ] together define the volume in the three - dimensional parameter space that is allowed by current experimental information and theoretical considerations .
in defining the allowed parameter space , we have not used an experimental constraint on the cdf - related linear combination presented in eq .
[ clspconstraint ] .
interestingly , we find that with the imposition of all the above constraints ( eqs . [ clm2 ] through [ s - constraint ] ) , this combination is limited to take values in the range @xmath103~{\rm gev}^3 \ ; .\ ] ] the central value of 0.066 quoted in eq .
[ clspconstraint ] falls just outside the upper end of this range , indicating a possible inconsistency between the result of ref .
@xcite and the results from other processes considered in the present analysis .
we now present our main result , which is the expected range of @xmath2 : the _ maximum _ value for @xmath2 ( @xmath104 ) is obtained when @xmath105 is at the minimum of its allowed range and when the photoproduction - related combination @xmath106 is at the maximum of its allowed range .
the _ minimum _
value for @xmath2 ( @xmath107 ) occurs in the opposite situation .
so far in our discussion , we have considered only the particular choice @xmath67 gev .
table i shows how our findings depend on @xmath21 .
the main results are the nrqcd predictions for the range of @xmath2 . alongside the full nrqcd results , we display the color - singlet model predictions for @xmath2 .
the error bars on the color - singlet - model predictions reflect @xmath44 relativistic corrections ; the relativistic corrections to the color - singlet result are of order @xmath44 , not @xmath3 as one might expect _ a priori _ , because the @xmath3 corrections to the helicity rates factor and cancel in the ratio @xmath2 . in the other three columns of table i , we treat the quantities appearing at the top of the columns as functions of the variables through which we scan ; we report the ranges of values of these quantities that occur for the allowed parameter space .
@xmath108 & & [ .08,.2 ] & & [ .1,.4 ] & & [ $ -.06$,$-.009 $ ] & \cr & & & & & & & & & & & & & \cr \noalign{\hrule } & & & & & & & & & & & & & \cr & & 4.4 & & $ -.34\pm .03 $ & & [ $ -.28 $ , $ -.11 $ ] & & [ .03,.1 ] & & [ .02,.2 ] & & [ $ -.03$,.003]&\cr & & & & & & & & & & & & & \cr \noalign{\hrule } & & & & & & & & & & & & & \cr & & 4.7 & & $ -.40\pm .04 $ & & [ $ -.35 $ , $ -.08 $ ] & & [ .009,.06 ] & & [ $ 0$,.1 ] & & [ $ -.01$,.006]&\cr & & & & & & & & & & & & & \cr \noalign{\hrule } & & & & & & & & & & & & & \cr & & 5.0 & & $ -.45\pm .05 $ & & [
$ -.42 $ , $ -.09 $ ] & & [ .002,.03 ] & & [ $ 0$,.06 ] & & [ $ -.008$,.006]&\cr & & & & & & & & & & & & & \cr \noalign{\hrule } & & & & & & & & & & & & & \cr & & 5.3 & & $ -.49\pm .05 $ & & [ $ -.47 $ , $ -.12 $ ] & & [ .001,.02 ] & & [ $ 0$,.03 ] & & [ -.005,.005]&\cr & & & & & & & & & & & & & \cr \noalign{\hrule } } } \ ] ] our leading - order results depend strongly on the values taken for @xmath21 and @xmath109 .
as stated previously we have used @xmath110 gev to generate the results given in table i. for some optional choices of @xmath109 we obtain the following ranges of @xmath2 . choosing @xmath111 gev and @xmath112 gev we find that @xmath113 $ ] .
choosing @xmath67 gev and @xmath114 gev we find that @xmath115 $ ] . in general for higher values of @xmath21 the predicted range for @xmath2 enlarges . for increasing values of @xmath109 the minimum value we predict for @xmath2 becomes greater , while the maximum value remains roughly unchanged . a general conclusion that can be drawn by looking at table
i is that the inclusion of octet matrix elements _ raises _ the predicted range for @xmath2 significantly from the color - singlet prediction .
we have proposed that a measurement of the polarization of @xmath0 particles produced in the process @xmath40 can furnish a useful new means of determining the color - octet matrix elements involved in @xmath0 production .
this measurement , which is not yet available , will supplement existing experimental information , which includes extractions from cdf , photoproduction , and the helicity - summed rate @xmath14 .
further experimental information would be of great utility , since there currently prevails a certain inconsistency between the values of the matrix elements measured at cdf and those values extracted from other processes .
we have presented the factorization formula for the helicity - summed rate in eq .
[ totalfacform ] .
rates for specified @xmath0 helicity have been calculated using the methods of beneke and rothstein @xcite and braaten and chen @xcite , and are presented in eq .
[ rlambda ] . the resulting expression for the polarization parameter @xmath2
is given in eq .
[ bigresult ] ( for the choice @xmath116 ) . in anticipation of the measurement of the polarization parameter @xmath2 ,
we have found it interesting to determine the range of @xmath2 which is predicted by the nrqcd factorization formalism , given current experimental information on the nrqcd matrix elements .
we find that , for @xmath116 , @xmath2 is expected to range from @xmath107 to @xmath104 .
our method of determining this range consists of scanning " through the three - dimensional parameter space of the color - octet matrix elements that is allowed by the constraints expressed in eqs .
[ clm2 ] , [ afm2 ] , [ decayconstraint2 ] and [ s - constraint ] . upon inspecting the parameter - space volume
so defined , we find that the linear combination of matrix elements that has been measured at cdf @xcite , namely @xmath117 is limited to take values from 0.01 to 0.06 gev@xmath118 , a situation which is at the threshold of incompatibility with the value of @xmath119 reported in ref . @xcite . in table
i , we have shown how our leading order predictions depend on @xmath21 .
in general , it can be concluded that the inclusion of octet matrix elements raises the predicted range for @xmath2 significantly from the color - singlet prediction , and that the range of @xmath2 broadens as @xmath21 is increased .
it may be that an accurate nrqcd prediction of the polarization parameter ( beyond leading order ) will not be possible for a long time .
this possibility , however , does not degrade the importance of our calculation , since an experimental determination of @xmath2 in conjuction with our results will most certainly serve to tighten the constraints on the possible values of the color - octet matrix elements .
in fact , the nrqcd prediction for @xmath2 is very sensitive to the values of the color - octet matrix elements , and this offers the hope that a measurement of the polarization of @xmath0 produced in @xmath120 will be instrumental in determining @xmath7 , @xmath8 , and @xmath98 .
we would especially like to thank eric braaten for many helpful discussions , and geoffrey t. bodwin for discussions regarding the @xmath13-quark mass .
i.m . and s.f gratefully acknowledge james buck montao for his invaluable help with the computer programming in various numerical analyses .
we also wish to acknowledge the hospitality of the university of wisconsin - madison , ohio state university ( i.m . ) , and the university of texas at austin ( s.f ) .
the work of s.f . was supported in part by the u.s .
department of energy under grant no .
de - fg02 - 95er40896 , in part by the university of wisconsin research committee with funds granted by the wisconsin alumni research foundation .
the work of i.m . was supported by the robert a. welch foundation , by nsf grant phy 9511632 , and by nserc of canada .
the work of o.f.h . and h.n .
was supported by les fonds fcar du qubec and by nserc of canada .
* figure 1*. feynman diagram for the four - quark fermi interactions @xmath121 and @xmath122 .
this interaction describes the short - distance physics for the process @xmath40 .
@xmath22 is the total four - momentum of the @xmath10 system , and @xmath23 , @xmath24 and @xmath13 are the four - momenta of the corresponding quarks . the @xmath10 pair is produced initially with quantum numbers @xmath123 with @xmath124 .
it then hadronizes into a @xmath0 particle with helicity @xmath51 .
* figure 2*. angular momentum conservation in @xmath121 and @xmath122 . because the charged - current fermi - interaction couples only the left - handed parts of the fermion fields , the @xmath23 and @xmath24 quarks ( being massless ) are emitted with negative helicity .
due to angular momentum conservation , therefore , the outcome @xmath125 is not allowed .
as a result , the @xmath6 and @xmath7 contributions to the @xmath0 helicity rate are zero for @xmath60 , at lowest order in @xmath3 . | we present the non - relativistic qcd ( nrqcd ) prediction for the polarization of the @xmath0 produced in @xmath1 , as well as the helicity - summed production rate .
we propose that these observables provide a means of measuring the three most important color - octet nrqcd matrix elements involved in @xmath0 production .
anticipating the measurement of the polarization parameter @xmath2 , we determine its expected range given current experimental bounds on the color - octet matrix elements . * nrqcd matrix elements in polarization * * of j / psi produced from b - decay * sean fleming _ department of physics , university of wisconsin , madison , wi 53706 _ oscar f. hernndez _ laboratoire de physique nuclaire , universit de montral , c.p . 6128 , succ .
a , montral , canada h3c 3p8 _ ivan maksymyk _ theory group , department of physics , university of texas , austin , tx 78712 _
hlne nadeau _ department of physics , mcgill university , montral , canada h3a 2t8 _ |
the compound series _ _ lntpn__o ( @xmath0 : lanthanides , @xmath10 : transition metal , @xmath1 : p or as ) have started to attract considerable attention because of the recent discovery of superconductivity with a transition temperature @xmath11 exceeding 50k in the @xmath0feaso@xmath2f@xmath3 series of compounds @xcite,@xcite,@xcite , being the highest @xmath11 except for cuprate systems . while the recent reports focus on the importance of electronic correlation effects due to @xmath6-electrons close to a magnetic state , the homologous compounds with @xmath0 = ce are attractive candidates for strong correlation effects induced by the @xmath12-electrons .
thus , last year we presented a detailed study of the properties of cerupo and ceospo @xcite , @xcite , @xcite and demonstrated that the former one is a rare example for a ferromagnetic ( fm ) kondo lattice system with a fm ordering temperature @xmath13k and a kondo temperature @xmath14k , while the latter one shows antiferromagnetic order of stable trivalent ce - ions below @xmath15k .
more recently , we have shown that cefepo is a heavy fermion metal close to a fm instability , dominated by the magnetism of the @xmath12-electrons @xcite . therefore , it was natural to look also for the physical properties of the related compound with @xmath10 = co. in this report , we present a first study of the basic physical properties of cecopo using susceptibility @xmath16 , specific heat @xmath17 , and resistivity @xmath18 measurements as well as x - ray absorption near - edge spectroscopy structure ( xanes ) at the ce @xmath19 edge
. the results reveal a fm ordering of the co - ions at @xmath5k in analogy to the fm order at 43k in lacopo @xcite .
an enhanced sommerfeld - coefficient of @xmath8mj / molk@xmath9 and the temperature dependence of the resistivity reveal a finite kondo scale of the @xmath12-moments of order @xmath20k .
in addition , we confirm the reported data of the magnetic properties of polycrystalline lacopo .
polycrystalline samples of cecopo and lacopo were synthesized using a sn - flux method in evacuated quartz tubes as described elsewhere @xcite , @xcite .
several powder x - ray diffraction patterns recorded on a stoe diffractometer in transmission mode using monochromated cu - k@xmath21 radiation ( @xmath22 ) confirmed the formation of @xmath0copo with zrcusias structure type ( space group : @xmath23 ; @xmath24 ) .
the lattice parameters refined by simple least square fitting for cecopo @xmath25 , @xmath26 and lacopo @xmath27 , @xmath28 were found to be in good agreement with the reported structure data @xcite , @xcite . in addition
, energy dispersive x - ray spectra of cecopo and lacopo performed on a scanning electron microscope ( philips xl30 ) with si(li)-x - ray detector show similar strong intensities of the oxygen line which confirm a substantial oxygen content of both phases .
however , they revealed diamagnetic cop@xmath29 @xcite as impurity phase and some oxide phases which are present only with small concentration , so that they could not be observed in the x - ray diffraction patterns .
@xmath16 measurements were performed in the temperature range @xmath30k in a commercial quantum design ( qd ) magnetic property measurement system ( mpms ) , the @xmath10-dependent measurements were performed after zero - field cooling the sample .
the resistivity was determined down to 0.4k using a standard ac four - probe geometry in a qd physical property measurement system ( ppms ) .
the ppms was also used to measure @xmath17 with a standard heat - pulse relaxation technique .
xanes - spectra near the ce @xmath19 edge were taken at 300k in transmission geometry at the exafs beamline a1 of hasylab at desy , germany .
the wavelength selection was realized by means of si(111 ) double crystal monochromator , which allowed an experimental resolution of approximately 2ev ( fwhm ) at the ce @xmath19 threshold of 5723ev .
t. both compounds show similar curie - weiss behavior above 230k with @xmath4 and @xmath31k ( straight line ) .
inset : the magnetization of cecopo presents a small hysteresis at small magnetic fields and a nearly saturated moment at higher fields.,width=377 ] in fig . [ figchi1 ] the inverse susceptibility is shown for both cecopo and lacopo . above
230k both curves present similar curie - weiss behavior with an effective moment of @xmath4 and a weiss temperature of @xmath31k .
our results for lacopo are very similar to those of ref . @xcite but important to analyze the magnetism in cecopo
. the identical behavior of cecopo and lacopo in @xmath32 evidences that in both compounds the main contribution to the magnetism results from the @xmath6-electrons of co. however , in cecopo the ce - ions are not completely unmagnetic as will be discussed later . in the inset of fig .
[ figchi1 ] , the magnetization @xmath33 of cecopo is shown .
@xmath33 presents a small hysteresis at low field and a saturated moment of @xmath34 at @xmath35 t , similar to what was observed in lacopo @xcite . at higher fields
, @xmath33 increases further slightly which might be due to the paramagnetism of the ce - ions . in fig .
[ figchi ] , we present @xmath16 of cecopo as a function of temperature at three different magnetic fields . the transition into a fm ordered state
is visible by a sharp increase at @xmath36k and a strong field dependence of @xmath37 at low @xmath10 .
however , the peak at @xmath38 observed in low fields ( @xmath39 t ) does not correspond to the expectation for a simple ferromagnet .
presently , it is not settled whether this is the result of measuring powder samples of a system with a complex anisotropic behavior or if it points to a more intricate magnetic structure with polarization effects on the ce - ions by the internal magnetic fields of the co - magnetism .
the later scenario is more likely , because the @xmath10-dependence of @xmath37 for lacopo ( line in fig .
[ figchi ] ) , where no @xmath12-electrons are present , behaves as expected for a simple ferromagnet .
of cecopo at various magnetic fields . the magnetic transition into a fm ordered state
is visible at @xmath5k . for comparison ,
@xmath16 for lacopo at @xmath40 t is presented as well ( straight line).,width=377 ] the temperature dependence of the specific heat of cecopo is shown in fig .
[ fighc ] without any subtraction of the phonons .
therefore , at high @xmath10 , @xmath17 converges to the classical limit of @xmath41j / molk .
an anomaly is visible on top of the phonon excitations just below @xmath38 .
this anomaly is rather broad , possibly due to inhomogeneities in our polycrystalline sample . in the inset of fig .
[ fighc ] we have plotted the specific heat as @xmath42 on a logarithmic temperature scale . in this representation
the peak due to the onset of fm order is more pronounced .
below @xmath43k , @xmath42 increases logarithmically and tends to saturate below @xmath44k at @xmath45mj / molk@xmath9 .
this value is strongly enhanced compared to the expectation for a normal fm co - system .
this enhancement can be attributed to @xmath12-correlation effects resulting in the formation of heavy fermions .
the entropy at low @xmath10 was calculated by integrating @xmath42 over temperature .
the entropy gain at 10k corresponds to about 25% of @xmath46 , giving a rough estimation of the kondo energy scale of order @xmath7k . for the calculation
, we have not subtracted any nonmagnetic contribution which for @xmath47k is negligible compared to the large value of @xmath42 . as an example , the contribution of the nonmagnetic larupo to the specific heat at @xmath48k would only correspond to 2% of the here measured specific heat @xcite . due to the fm order of cecopo .
the inset shows the specific heat plotted as @xmath42 on a logarithmic temperature scale . below @xmath43k , @xmath42 of cecopo
increases logarithmically down to lower @xmath10 , presenting an enhanced @xmath45mj / molk@xmath9 at 0.35k.,width=377 ] in fig .
[ figrho ] we present the resistivity data on a pressed powder - pellet of cecopo which was subsequently annealed for 150 hours at 700@xmath49c , improving the conductivity of the sample by a factor of 20 .
however , the absolute value of @xmath50m@xmath51 cm is still rather high ; we believe that this is not intrinsic of cecopo but due to the remaining granularity of the polycrystalline sample . to compare our results with @xmath18 of lacopo from yanagi _ et al .
_ , we have normalized our data to @xmath52 . from fig .
[ figrho ] it is obvious that @xmath18 of cecopo differs significantly from the data of lacopo which decreases linearly with temperature with a small change in slope at the fm ordering temperature .
in contrast , cecopo shows linear metallic behavior only above 100k , followed by a pronounced decrease which is a distinct feature of a kondo lattice system .
however , we can not exclude that this decrease is due to reduced spin - spin scattering because of the fm order at @xmath38 , as both energy scales ( @xmath7k and @xmath5k ) are comparable .
thermopower measurements on cecopo would be desirable to distinguish between both effects .
we have performed xanes measurements at the ce@xmath53 edge to determine the valency of the ce - ions to gain a deeper insight of the relevance of cerium to the magnetic behavior in cecopo . in the inset of fig .
[ figxas ] we have plotted the evolution of the tetragonal unit cell volume @xmath54 for the three @xmath55po series ( @xmath56co , fe , ru ) . the contraction of @xmath54 is typical for the @xmath0 series , and a smooth evolution is expected if ce is in a trivalent state .
therefore , one would expect a stable , magnetic ce@xmath53 in cerupo but a stronger tendency to intermediate - valent ce - ions in cefepo and cecopo @xcite .
this trend is confirmed in our xanes spectra shown in fig .
[ figxas ] .
all three spectra are quite close to each other indicating , that the valency of ce is not changing dramatically when going from ru- to the co - compound . above 5.73kev , the co - compound show a slightly higher intensity compared to cerupo , but far away from the double - peak behavior in the formal pure ce@xmath57-system ceo@xmath29 @xcite . from this
we conclude a predominantly ce@xmath53 for all three systems .
preliminary x - ray magnetic circular dichroism ( xmcd ) measurements on cecopo confirm this result showing that more than 90% of the ce - ions are in the @xmath58 configuration @xcite .
m@xmath51 cm . for comparison , @xmath18 of lacopo taken from yanagi _ et al . _
@xcite is shown ( straight line).,width=377 ]
our results on cecopo evidence three interesting features which shall now be discussed : the interplay of the @xmath6 and @xmath12 magnetism , the comparison with the results obtained on cerupo and cefepo , and the relation to the new high temperature superconductors @xmath0fe@xmath1o@xmath2f@xmath3 .
the presented results point to a fm ordering of the co - ions in cecopo similar to the fm order in lacopo @xcite ; however , the unusual temperature dependence of @xmath16 at @xmath38 together with an increasing magnetization for fields @xmath59 t are not expected for a simple fm and not observed in lacopo . on the other hand , the xanes spectra suggest the ce - ions to be close to trivalent which is in agreement with the enhanced @xmath60 , and the temperature dependence of @xmath18 which are typical for pronounced kondo - interactions .
therefore , we suggest the following scenario for cecopo . the spins of the co - ions ( @xmath6-electrons ) order ferromagnetically at @xmath5k , the resulting internal fields partially polarize the ce - ions ( @xmath12-electrons ) , which however remain also partially fluctuating down to lowest measured temperatures .
one could even imagine that a kondo screening increasing with decreasing temperature lead to a decreasing ce - polarization with temperature . in order to get a deeper insight into the many - body phenomena in this compound
, a more microscopic local probe should be used to study the magnetism .
we therefore , started an investigation of cecopo with xmcd @xcite as well as @xmath61p nmr and @xmath62co nmr @xcite to distinguish between @xmath6 and @xmath12 magnetism .
the doniach picture can be used to classify the ce@xmath10po series with respect to the magnetism of the @xmath12-electrons @xcite . from the inset of fig .
[ figxas ] it is evident that the volume of the unit cell decreases from @xmath10=ru to fe and further to co. with decreasing @xmath54 the magnetism of the ce - ions becomes more and more screened by the conduction electrons because of an increasing kondo energy scale .
accordingly , we observe stable ferromagnetism of the ce - ions in cerupo @xcite which has the largest @xmath63nm@xmath64 .
cefepo has a smaller @xmath65nm@xmath64 and is close but already on the nonmangetic side of a fm instability @xcite .
the volume related suppression of the magnetic order can also be observed in pressure experiments on single crystals of cerupo @xcite .
our conclusion of a larger screening of the @xmath12 electrons as identified by a higher kondo temperature @xmath7k in cecopo is in nice continuation of this systematic trend since this compound has a further reduced volume @xmath66nm@xmath64 compared to cefepo .
edge of cetpo ( t = co , fe , ru ) taken on powdered samples at @xmath67k .
the data are normalized to 1 at 5.755kev .
inset : cell volumes of the tetragonal phosphide oxides @xmath55po @xcite , @xcite.,width=377 ] the superconductivity at @xmath68k in f - doped lafeaso came as a big surprise @xcite and subsequently research on these systems reveals systems with @xmath11 s exceeding 50k in the elements with heavier @xmath0 @xcite , @xcite .
the connection between a vanishing magnetic transition and the simultaneous formation of a superconducting state is reminiscent of the behavior in the cuprates and in the heavy fermion systems , and therefore suggests the superconducting state in these doped rfeaso systems to be of unconventional nature , too . in cecopo
, we could not observe superconductivity down to 350mk , which is not so much surprising taking into account the fm co - state .
the magnetism might be easy to suppress either by doping or pressure , because an ordering temperature of @xmath5k is rather low for co - ordering .
further research has to answer the question whether superconductivity appears in this compound after suppression of the fm order .
in summary , we have shown that cecopo presents ferromagnetic order of the co - ions at @xmath5k .
the @xmath12-electrons are screened due to pronounced kondo - interactions with a kondo energy scale of order @xmath7k revealed by the temperature dependence of the resistivity and an enhanced sommerfeld - coefficient @xmath8mj / molk@xmath9 .
in addition , we confirm the fm order in lacopo at @xmath69k .
the authors thank p. scheppan for energy dispersive x - ray analysis of the samples and n. caroca - canales and r. weise for technical assistance in sample preparation .
h. rosner , e. m. brning , and m. baenitz are acknowledged for valuable discussions . | the ground state properties of cecopo , a homologue of the new high temperature superconductors @xmath0fe@xmath1o@xmath2f@xmath3 , were studied by means of susceptibility , specific heat , and resistivity measurements on polycrystals .
the observation of a well defined curie - weiss behavior above 230k with @xmath4 and a ferromagnetic ordering below @xmath5k is similar to what was observed in lacopo and points to magnetism of the co-@xmath6 electrons .
however , the ce - ions are on the border to magnetism with a kondo scale of @xmath7k and show an enhanced sommerfeld - coefficient of @xmath8mj / molk@xmath9 . , , and ferromagnetic system , zrcusias - type structure , kondo lattice system , heavy - fermion 75.20.hr , 71.20.eh , 75.50.cc |
exactly solvable models play a very useful role in various fields of physics .
they help improving our understanding of physical processes and allow us gain more insight into complicated phenomena that take place in nature @xcite .
needless to recall the usefulness of exactly solvable models such as the harmonic oscillator , the nuclear shell model and the ising model , to mention but a few . from a practical point of view , exactly solvable models serve as a very convenient tool for testing the accuracy of numerical algorithms , often used in the study of problems that can not be analytically solved due to the complexity of the systems under investigation . in nature ,
quantum systems are influenced by their surrounding environment through , in general , complicated coupling interactions , leading them to lose their coherence @xcite .
this refers to as the decoherence process @xcite .
moreover , quantum systems exhibit properties that do not have classical analogous @xcite . of great interest is entanglement , the main ingredient for quantum teleportation and quantum computation @xcite . over the last years
, many proposals have been made for the implementation of quantum information processing .
solid state systems are very promising @xcite and have been the subject of many investigations . in particular , decoherence and entanglement of qubits
coupled to spin environments @xcite attracted much attention @xcite .
thus new exactly solvable models describing the dynamics of qubits in spin baths are highly welcome .
recently , the spin star configuration , initially proposed by bose , has been extensively investigated @xcite .
an exact treatment of the dynamics of two qubits coupled to common spin star bath via @xmath0 interactions is presented in @xcite . in this paper
we propose to investigate analytically the dynamics when the two qubits interact with separate spin star baths .
the paper is organized as follows . in section [ sec2 ]
the model hamiltonian is introduced . in section [ sec3 ]
we present a detailed derivation of the time evolution operator and we investigate the dynamics of the qubits at finite @xmath1 for some particular initial conditions . in section [ sec4 ]
we study the thermodynamic limit , in which the sizes of the spin environments become infinite .
section [ sec5 ] is devoted to the second - order master equation .
we end the paper with a short summary .
the system under study consists of two two - level systems ( e.g. , spin-@xmath2 particles ) each of which is embedded in its own spin star environment composed of @xmath1 spins-@xmath2 .
the central particles interact with each other through a ising interaction ; the corresponding coupling constant is equal to @xmath3 , where the factor 4 is introduced for later convenience .
we shall assume that each qubit couples to its environment via heisenberg @xmath0 interaction whose coupling constant is @xmath4 , which is , in turn , scaled by @xmath5 in order to ensure good thermodynamic behavior .
the spin baths will be denoted by @xmath6 and @xmath7 .
the hamiltonian for the composite system has the form @xmath8 where @xmath9 and @xmath10 here @xmath11 and @xmath12 denote the spin operators corresponding to the central qubits , whereas @xmath13 denotes the spin operator corresponding to the @xmath14 particle within the @xmath15 environment . introducing the total spin operators @xmath16 and @xmath17 of the environments @xmath6 and @xmath7 , respectively , one can rewrite the full hamiltonian as @xmath18 the dynamics of the two - qubit system is fully described by its density matrix @xmath19 obtained , as usual , by tracing the time - dependent total density matrix @xmath20 , describing the composite system , with respect to the environmental degrees of freedom , namely , @xmath21\nonumber \\ & = & { \rm tr}_{b_1+b_2}\bigl[\mathbf
u(t)\rho_{\rm tot}(0)\mathbf u^\dag(t)\bigl],\end{aligned}\ ] ] where @xmath22 and @xmath23 designate the time evolution operator and the initial total density matrix , respectively . at @xmath24 the central qubits are assumed to be uncoupled with the environments ; the latter are assumed to be at infinite temperature .
this means that the initial total density density matrix can be written as @xmath25 here @xmath26 is the initial density matrix of the two - qubit system , and @xmath27 is the unit matrix on the space @xmath28 .
the former can be written as @xmath29 , with @xmath30 for @xmath31 .
similarly , we introduce the basis state vectors @xmath32 of @xmath28 , such that @xmath33 ( @xmath34 for @xmath1 even and @xmath35 for @xmath1 odd ) , and @xmath36 .
the time - dependent reduced density matrix can be expressed as @xmath37 where @xmath38 , and @xmath39 @xcite .
hence , our task reduces to finding the exact form of the matrix elements of the time evolution operator @xmath40 ( @xmath41 ) .
this will be the subject of the next section .
the time evolution operator can be expanded as @xmath42 in order to derive analytical expressions for even and odd powers of the total hamiltonian @xmath43 let us notice that @xmath44 anticommutes with @xmath45 , that is , @xmath46_+=0.\ ] ] this can easily be shown using the following properties for spin-@xmath2 operators : @xmath47 , and @xmath48 .
moreover , it is easily seen that @xmath49 , which simply implies that for @xmath50 , @xmath51 in the standard basis of @xmath52 , it can be shown that powers of @xmath53 and @xmath54 are given by @xmath55 it follows that @xmath56.\end{aligned}\ ] ] where @xmath57\end{aligned}\ ] ] and @xmath58.\end{aligned}\ ] ] using the fact that @xmath59,\\ \sum\limits_{k=0}^{\ell-1 } \binom{2\ell}{2k+1 } x^k y^{\ell - k-1}&= & \frac{1}{2\sqrt{x y}}\bigl [ ( \sqrt{x } + \sqrt{y})^{2 \ell } - ( \sqrt{x } - \sqrt{y})^{2 \ell}\bigl],\end{aligned}\ ] ] one obtains @xmath60 where @xmath61,\\ f^\pm_2=\frac{1}{2}\bigl[\bigl(\sqrt{j_+j_-}+\sqrt{\mathcal j_-\mathcal j_+}\bigl)^{2\ell}\pm\bigl(\sqrt{j_+j_-}-\sqrt{\mathcal j_-\mathcal j_+}\bigl)^{2\ell}\bigl],\\ f^\pm_3=\frac{1}{2}\bigl[\bigl(\sqrt{j_-j_+}+\sqrt{\mathcal j_+\mathcal j_-}\bigl)^{2\ell}\pm\bigl(\sqrt{j_-j_+}-\sqrt{\mathcal j_+\mathcal j_-}\bigl)^{2\ell}\bigl],\\ f^\pm_4=\frac{1}{2}\bigl[\bigl(\sqrt{j_-j_+}+\sqrt{\mathcal j_-\mathcal j_+}\bigl)^{2\ell}\pm\bigl(\sqrt{j_-j_+}-\sqrt{\mathcal j_-\mathcal j_+}\bigl)^{2\ell}\bigl].\end{aligned}\ ] ] inserting equation ( [ powereven ] ) into equation ( [ binom ] ) , yields @xmath62 where @xmath63 the above operators satisfy @xmath64 furthermore , one can show that the matrix elements of @xmath65 are given by @xmath66,\label{powerodd1 } \\ ( h^{2n+1})_{12}&=\mathcal j_+\frac{\alpha}{2\sqrt{n \mathcal j_-\mathcal j_+}}[(\sqrt{\mathcal j_-\mathcal j_+}+\sqrt{j_+j_-})(\mathcal m_2^+)^n \\&+(\sqrt{\mathcal j_-\mathcal j_+}-\sqrt{j_+j_-})(\mathcal m_2 ^ -)^n ] , \\
( h^{2n+1})_{13}&=\ j_+\frac{\alpha}{2\sqrt{n j_- j_+}}[(\sqrt{\mathcal j_+\mathcal j_-}+\sqrt{j_-j_+})(\mathcal m_3^+)^n \\&+(\sqrt { j_-j_+}-\sqrt{\mathcal j_+\mathcal j_-})(\mathcal m_3 ^ -)^n ] , \\ ( u^{2n+1})_{14}&=(\delta/2 ) j_+\mathcal j_+ \frac{(\mathcal m_4^+)^n-(\mathcal m_4 ^ -)^n}{\sqrt{j_-j_+\mathcal j_-\mathcal j_+ } } , \end{aligned}\ ] ] @xmath67 , \\
( h^{2n+1})_{22}&=&-\frac{1}{2}\delta[(\mathcal m_2^+)^n+(\mathcal m_2 ^ -)^n ] , \\ ( h^{2n+1})_{23}&=&-(\delta/2 ) j_+\mathcal j_- \frac{(\mathcal m_3^+)^n-(\mathcal m_3 ^ -)^n}{\sqrt{j_-j_+\mathcal j_+\mathcal j_- } } , \\ ( h^{2n+1})_{24}&=&j_+\frac{\alpha/2}{\sqrt{n j_- j_+}}[(\sqrt{\mathcal j_-\mathcal j_+}+\sqrt { j_- j_+ } ) ( \mathcal m_4^+)^n \\&+&(\sqrt{j_- j_+}-\sqrt{\mathcal j_-\mathcal j_+})(\mathcal m_4 ^ -)^{n } ] , \end{aligned}\ ] ] @xmath68 , \\ ( h^{2n+1})_{32}&=&-(\delta/2)j_-\mathcal j_+\frac{(\mathcal m_2^+)^n-(\mathcal m_1 ^ -)^n}{\sqrt{j_+j_-\mathcal j_-\mathcal j_+ } } , \\ ( h^{2n+1})_{33}&=&-\frac{1}{2}\delta[(\mathcal m_3^+)^n+(\mathcal m_3 ^ -)^n ] , \\ ( h^{2n+1})_{34}&=&\mathcal j_+\frac{\alpha/2}{\sqrt{n \mathcal j_-\mathcal j_+}}[(\sqrt{\mathcal j_-\mathcal j_+}+\sqrt{j_- j_+})(\mathcal m_4^+)^n \\&+&(\sqrt{\mathcal j_-\mathcal j_+}-\sqrt{j_- j_+ } ) ( \mathcal m_4 ^ -)^n ] , \end{aligned}\ ] ] @xmath69 , \\ ( h^{2n+1})_{43}&=&\mathcal j_-\frac{\alpha/2}{\sqrt{n \mathcal j_+\mathcal j_-}}[(\sqrt{\mathcal j_+\mathcal j_-}+\sqrt{j_- j_+})(\mathcal m_3^+)^n \\&+&(\sqrt{\mathcal j_+\mathcal j_-}-\sqrt{j_- j_+ } ) ( \mathcal m_3 ^ -)^n ] , \\ ( h^{2n+1})_{44}&=&\frac{1}{2}\delta[(\mathcal m_4^+)^n+(\mathcal m_4 ^ -)^n].\label{powerodd2}\end{aligned}\ ] ] having in hand the explicit expressions of powers of the total hamiltonian
, it can easily be verified that the elements of the time evolution operator , obtained by inserting equations ( [ power1 ] ) and ( [ powerodd1])-([powerodd2 ] ) into equation ( [ taylor ] ) , are given by @xmath70\bigl\},\\ u_{21}(t)=&-\mathcal j_-\frac{i\alpha/2}{\sqrt{n\mathcal j_+\mathcal j_-}}\bigl\{\frac{\sqrt{j_+j_-}+\sqrt{\mathcal j_+ \mathcal j_-}}{\sqrt { \mathcal m_1^+}}\sin\bigl(t\sqrt { \mathcal m_1^+}\bigl)\nonumber\\&-\frac{\sqrt{j_+j_-}-\sqrt{\mathcal j_+ \mathcal j_-}}{\sqrt { \mathcal m_1 ^ -}}\sin\bigl(t \sqrt { \mathcal m_1 ^ -}\bigl)\bigl\},\\ u_{31}(t)=&- j_-\frac{i\alpha/2}{\sqrt{n j_+j_-}}\bigl\{\frac{\sqrt{j_+j_-}+\sqrt{\mathcal j_+ \mathcal j_-}}{\sqrt { \mathcal m_1^+}}\sin\bigl(t\sqrt { \mathcal m_1^+}\bigl)\nonumber\\&+\frac{\sqrt{j_+j_-}-\sqrt{\mathcal j_+ \mathcal j_-}}{\sqrt { \mathcal m_1 ^ -}}\sin\bigl(t \sqrt { \mathcal m_1 ^ -}\bigl)\bigl\},\end{aligned}\ ] ] @xmath71\bigl\},\\ u_{22}(t)=&\frac{1}{2}\bigl\{\cos\bigl(t \sqrt { \mathcal m_2^+}\bigl)+\cos\bigl(t\sqrt { \mathcal m_1 ^ -}\bigl)+i\delta\bigl[\frac{\sin\bigl(t\sqrt { \mathcal m_2^+}\bigl)}{\sqrt { \mathcal m_2^+}}+\frac{\sin\bigl(t \sqrt { \mathcal m_2 ^ -}\bigl ) } { \sqrt { \mathcal m_2 ^ -}}\bigl]\bigl\},\\ u_{12}(t)=&-\mathcal j_+\frac{i\alpha/2}{\sqrt{n\mathcal j_-\mathcal j_+}}\bigl\{\frac{\sqrt{j_+j_-}+\sqrt{\mathcal j_- \mathcal j_+}}{\sqrt { \mathcal m_2^+}}\sin\bigl(t\sqrt { \mathcal m_2^+}\bigl)\nonumber\\&-\frac{\sqrt{j_+j_-}-\sqrt{\mathcal j_- \mathcal j_+}}{\sqrt { \mathcal m_2 ^ -}}\sin\bigl(t \sqrt { \mathcal m_2 ^ -}\bigl)\bigl\},\\ u_{32}(t)=&j_-\mathcal j_+\frac{1}{2\sqrt{j_+j_-\mathcal j_-\mathcal j_+}}\bigl\{\cos\bigl(t \sqrt { \mathcal m_2^+}\bigl)-\cos\bigl(t\sqrt { \mathcal m_2 ^ -}\bigl)\nonumber\\&+i\delta\bigl[\frac{\sin\bigl(t\sqrt { \mathcal m_2^+}\bigl)}{\sqrt { \mathcal m_2^+}}-\frac{\sin\bigl(t \sqrt { \mathcal m_2 ^ -}\bigl ) } { \sqrt { \mathcal m_2 ^ -}}\bigl]\bigl\},\\ u_{42}(t)=&- j_-\frac{i\alpha/2}{\sqrt{n j_+j_-}}\bigl\{\frac{\sqrt{j_+j_-}+\sqrt{\mathcal j_- \mathcal j_+}}{\sqrt { \mathcal m_2^+}}\sin\bigl(t\sqrt { \mathcal m_2^+}\bigl)\nonumber\\&+\frac{\sqrt{j_+j_-}-\sqrt{\mathcal j_- \mathcal j_+}}{\sqrt { \mathcal m_2 ^ -}}\sin\bigl(t \sqrt { \mathcal m_2 ^ -}\bigl)\bigl\},\\ u_{33}(t)=&\frac{1}{2}\bigl\{\cos\bigl(t \sqrt { \mathcal m_3^+}\bigl)+\cos\bigl(t\sqrt { \mathcal m_3 ^ -}\bigl)+i\delta\bigl[\frac{\sin\bigl(t\sqrt { \mathcal m_3^+}\bigl)}{\sqrt { \mathcal m_3^+}}+\frac{\sin\bigl(t \sqrt { \mathcal m_3 ^ -}\bigl ) } { \sqrt { \mathcal m_3 ^ -}}\bigl]\bigl\},\\ u_{13}(t)=&- j_+\frac{i\alpha/2}{\sqrt{n j_- j_+}}\bigl\{\frac{\sqrt{\mathcal j_+\mathcal j_-}+\sqrt { j_- j_+}}{\sqrt { \mathcal m_3^+}}\sin\bigl(t\sqrt { \mathcal m_3^+}\bigl)\nonumber\\&-\frac{\sqrt{\mathcal j_+\mathcal j_-}-\sqrt { j_- j_+}}{\sqrt { \mathcal m_3 ^ -}}\sin\bigl(t \sqrt { \mathcal m_3 ^ -}\bigl)\bigl\},\\ u_{23}(t)=&j_+\mathcal j_- \frac{1}{2\sqrt{j_- j_+ \mathcal j_+\mathcal j_- } } \bigl\{\cos\bigl(t \sqrt { \mathcal m_3^+}\bigl)-\cos\bigl(t\sqrt { \mathcal m_3 ^ -}\bigl)\nonumber\\&+i\delta\bigl[\frac{\sin\bigl(t\sqrt { \mathcal m_3^+}\bigl)}{\sqrt { \mathcal m_3^+}}-\frac{\sin\bigl(t \sqrt { \mathcal m_3 ^ -}\bigl ) } { \sqrt { \mathcal m_3 ^ -}}\bigl]\bigl\},\\ u_{43}(t)=&- \mathcal j_-\frac{i\alpha/2}{\sqrt{n \mathcal j_+\mathcal j_-}}\bigl\{\frac{\sqrt{\mathcal j_+\mathcal j_-}+\sqrt { j_- j_+}}{\sqrt { \mathcal m_3^+}}\sin\bigl(t\sqrt { \mathcal m_3^+}\bigl)\nonumber\\&+\frac{\sqrt{\mathcal j_+\mathcal j_-}-\sqrt { j_- j_+}}{\sqrt { \mathcal m_3 ^ -}}\sin\bigl(t \sqrt { \mathcal m_3 ^ -}\bigl)\bigl\},\end{aligned}\ ] ] @xmath72\bigl\},\\ u_{24}(t)=&- j_+\frac{i\alpha/2}{\sqrt{n j_- j_+}}\bigl\{\frac{\sqrt{\mathcal j_-\mathcal j_+}+\sqrt { j_- j_+}}{\sqrt { \mathcal m_4^+}}\sin\bigl(t\sqrt { \mathcal m_4^+}\bigl)\nonumber\\&-\frac{\sqrt{\mathcal j_-\mathcal j_+}-\sqrt{j_- j_+}}{\sqrt { \mathcal m_4 ^ -}}\sin\bigl(t \sqrt { \mathcal m_4 ^ -}\bigl)\bigl\},\\ u_{34}(t)=&-\mathcal j_+\frac{i\alpha/2}{\sqrt{n \mathcal j_-\mathcal j_+}}\bigl\{\frac{\sqrt{\mathcal j_-\mathcal j_+}+\sqrt{j_- j_+}}{\sqrt { \mathcal m_4^+}}\sin\bigl(t\sqrt { \mathcal m_4^+}\bigl)\nonumber\\&+\frac{\sqrt{\mathcal j_-\mathcal j_+}-\sqrt{j_- j_+}}{\sqrt { \mathcal m_4 ^ -}}\sin\bigl(t \sqrt { \mathcal m_4 ^ -}\bigl)\bigl\},\\ u_{14}(t)=&j_+\mathcal j_+\frac{1}{2\sqrt{j_- j_+\mathcal j_-\mathcal j_+}}\bigl\{\cos\bigl(t \sqrt { \mathcal m_4^+}\bigl)-\cos\bigl(t\sqrt { \mathcal m_4 ^ -}\bigl)\nonumber\\&-i\delta\bigl[\frac{\sin\bigl(t\sqrt { \mathcal m_4^+}\bigl)}{\sqrt { \mathcal m_4^+}}-\frac{\sin\bigl(t \sqrt { \mathcal m_4 ^ -}\bigl ) } { \sqrt { \mathcal m_4 ^ -}}\bigl]\bigl\}.\end{aligned}\ ] ] it should be noted that the above components of the operator @xmath22 can also be derived by solving the schrdinger equation @xcite @xmath73 for instance , we have @xmath74 this set of differential equation can be solved by introducing the following transformations : @xmath75 the resulting differential equations involve diagonal terms ; they can be solved by taking into account the initial conditions : @xmath76 following the same procedure , it is possible to derive the remaining matrix elements of the time evolution operator .
there exist many measures for entanglement . here
we shall use the concurrence , defined by @xcite @xmath77-\sum_{i=1}^4\sqrt{\lambda_i}\},\ ] ] where the quantities @xmath78 are the eigenvalues of the operator @xmath79 .
the above measure is equal to one for maximally entangled states , and is equal to zero for separable states .
the purity @xmath80 can be used to quantify the decoherence of the central system ; it is equal to @xmath81 for maximally mixed states , and one for pure states .
it turns out that the density matrices corresponding to the initial product states @xmath82 , where @xmath83 , are always diagonal .
furthermore , the numerical simulation shows that if the qubits are prepared in one of the above states , they remain unentangled regardless of the values of @xmath1 and @xmath84 .
the purity decays less with the increase of @xmath84 .
the matrix elements of the reduced density matrices corresponding to the states @xmath85 and @xmath86 are shown in the appendix . the evolution in time of the concurrence and
the purity corresponding to the above maximally entangled states is practically the same ; we only present the results obtained for the singlet state .
it is found that , for fixed @xmath84 , the concurrence and the purity saturate as the number of spins increases .
this naturally suggests the investigation of the case @xmath87 ( see the next section ) . for small values of the coupling constant @xmath84
, the concurrence decays from its initial maximum value @xmath88 , then vanishes at a certain moment of time ( i.e. entanglement sudden death @xcite ) . at long times , and sufficiently large @xmath1 and @xmath84 , the purity and the concurrence converge to certain asymptotic values , which increase with the increase of the strength of interaction . here it should be noted that , in contrast to the case of common spin bath , the singlet state is not decoherence free . this
was expected because the latter state is not eigenvector of the hamiltonian @xmath43 .
nevertheless , we find that decoherence can be reduced with strong coupling between the qubits , in agreement with @xcite .
finally let us remark that , although we only have considered infinite temperature , we can ensure that for long - range antiferromagnetic heisenberg interactions within the baths , low temperatures will have the same effect on decoherence and entanglement of the qubits as strong coupling constants .
in the thermodynamic limit , @xmath87 , the operators @xmath89 converge to the positive real random variable @xmath90 whose probability density function is given by @xmath91 indeed , it has been shown in @xcite that the operator @xmath92 converges to the complex normal random variable @xmath93 with the probability density function @xmath94 expressing @xmath93 in terms of the polar coordinates @xmath90 and @xmath95 , i.e. , @xmath96 , simply gives @xmath97 .
then integrating the corresponding probability density function over the variable @xmath95 from 0 to @xmath98 yields @xmath99 from which ( [ prod5 ] ) follows .
hence we can ascertain that @xmath100 where @xmath101 is some complex - valued function for which the integrals in the right - hand side of equation ( [ sumint ] ) converge . using the above result
, one can express the nonzero elements of the reduced density matrix corresponding to the initial state @xmath102 , in the thermodynamic limit , as @xmath103,\\ \rho_{22}(t)&=&\rho_{33}(t)=\frac{1}{4}[\upsilon_+(t)+\upsilon_-(t)+\xi_+(t)+\xi_-(t)],\\ \rho_{23}(t)&=&-\frac{1}{8}[\upsilon_+(t)+\upsilon_-(t)+\xi_+(t)+\xi_-(t)+2\psi(t)],\end{aligned}\ ] ] where ( we set @xmath104 for the sake of shortness ) @xmath105 unfortunately the above functions can not be evaluated analytically ; one should make recourse to numerical integration .
this task can be significantly simplified by transforming the double integration into single one , which is much easier to carry out . to do that notice that the analysis of the expressions of the functions @xmath106 , @xmath107 , and @xmath108 leads to the evaluation of the probability density functions @xmath109 and @xmath110 corresponding , respectively , to the random variables @xmath111 and @xmath112 ( see @xcite for a similar situation ) .
let us begin with the variable @xmath113 ; its probability density function is simply given by the convolution of @xmath114 with itself : @xmath115 note that the upper limit of the integration over @xmath90 is @xmath113 because the quantity @xmath116 should be positive .
the evaluation of the integral is somewhat lengthy , but elementary ; one finds that @xmath117\mathrm e^{-2\mu^2},\label{probfun1}\ ] ] where @xmath118 designates the error function @xcite .
now consider the variable @xmath112 .
one should be careful when using the definition of the convolution , since , in this case , @xmath119 belongs to the interval @xmath120-\infty,\infty [ $ ] .
we have to distinguish between two cases , namely , @xmath121 and @xmath122 . in the first case @xmath123 , and
hence @xmath124\}\mathrm e^{-2 \eta^2}.\label{pp1}\end{aligned}\ ] ] when @xmath122 , then @xmath125 , which implies that @xmath126\}\mathrm e^{-2 \eta^2}.\label{pp2}\end{aligned}\ ] ] combining ( [ pp1 ] ) and ( [ pp2 ] ) , we obtain the following expression for the probability density function of @xmath119 over the real line : @xmath127\}\mathrm e^{-2 \eta^2}\label{profun2}.\end{aligned}\ ] ] the above functions are depicted in figures [ figure3 ] and [ figure4 ] .
clearly , @xmath110 is an even function of its argument ; it takes its maximum value at the origin , that is , @xmath128 .
the maximum value of @xmath109 occurs at @xmath129 , such that @xmath130 . as a simple application let us prove the following : the moments around origin of the random variables @xmath113 and @xmath119 are given by : @xmath131,\label{mom1}\\ \langle \mu^{2n+1 } \rangle&=\frac{\gamma(\tfrac{3}{2}+1)}{2^n}\bigl[\tfrac{1}{\sqrt{2}}+2^{n } ( 2n+1)\ _ 2f_1\bigl ( \frac{3}{2 } + n,\frac{1}{2 } ; \frac{3}{2 } ; -1\bigl)\bigl],\\ \langle \eta^{2n } \rangle&=\langle \mu^{2n } \rangle - n\sqrt{\pi}\gamma\bigl(\frac{1}{2}+n\bigl),\\ \langle \eta^{2n+1 } \rangle&=0,\label{mom4}\end{aligned}\ ] ] where @xmath132 , and @xmath133 denote the gamma and the hypergeometric functions , respectively . relation ( [ mom4 ] ) is obvious since the function @xmath110 is even .
let us prove ( [ mom1 ] ) .
we have that @xmath134 where @xmath135 to calculate @xmath136 and @xmath137 , introduce the functions of the real variable @xmath138 : @xmath139 @xmath140 the first integral can be easily evaluated : @xmath141 the second integral satisfies @xmath142 integrating by parts the rhs of ( [ newfun ] ) with respect to @xmath113 , and using ( [ firsteq ] ) , yield @xmath143 here we have used the fact that @xmath144 .
let @xmath145 .
then from ( [ redn ] ) we have @xmath146 on the other hand equation ( [ difffn ] ) implies that @xmath147 combining the last two equations yields the following first order differential equation for the function @xmath148 : @xmath149 differentiating both sides of ( [ difff ] ) , and again using ( [ dign ] ) , we obtain @xmath150g_n(x)=0.\ ] ] by setting @xmath151 , and @xmath152 , we obtain @xmath153h_n(y)=0,\ ] ] which should be compared with the hypergeometric equation @xmath154\ _ 2f_1(a , b;c;y)=0.\ ] ] thus @xmath155 it follows that @xmath156 putting @xmath157 yields @xmath158 also , using ( [ redn ] ) , we obtain @xmath159 from which ( [ mom1 ] ) readily follows .
the other moments can be evaluated with a similar method .
the functions ( [ f1])-([f3 ] ) can easily be expressed in terms of the functions @xmath109 and @xmath110 .
for example , we have : @xmath160 it should be noted that in contrast to @xmath90 and @xmath161 , the random variables @xmath119 and @xmath113 are not independent . the function @xmath162 can not be further simplified , and should be evaluated using the double integration over the variables @xmath90 and @xmath161 . nevertheless , using the riemann - lebesgue lemma
, we can infer that @xmath163 in a similar way , the remaining functions tend asymptotically to : @xmath164 notice that @xmath165 independently of the values of @xmath84 .
it follows that the asymptotic density matrix can be expressed as @xmath166 where @xmath167 it is easily seen that @xmath168 the corresponding asymptotic reduced density matrix reads @xmath169 which has a concurrence identically equal to zero . on the contrary , in the limit of strong coupling between the central qubits , @xmath170 consequently , @xmath171 a straightforward calculation shows that @xmath172 in general , since @xmath173 , then @xmath174 this allows us to find the following explicit form of the asymptotic value of the concurrence : @xmath175 the latter can also be rewritten as : @xmath176 the variation of the asymptotic concurrence as a function of @xmath84 is shown in figure [ figure5 ] .
it can be seen that @xmath177 remains zero up to a critical value @xmath178 after which it increases , to tend asymptotically to @xmath2 .
the value of @xmath178 can be evaluated numerically : @xmath179 at the critical point , the density matrix reads @xmath180
under born approximation , the second - order master equation yields the following set of integro - differential equations : @xmath181\ ds,\label{master1}\\ \dot{\tilde{\rho}}_{12}(t)&=-\alpha^2\int\limits_0^t\big(2\tilde{\rho}_{12}(s)\mathrm e^{2i\delta(t - s)}-\tilde{\rho}_{34}(s ) \mathrm
e^{2i\delta(t+s)}\bigr)\ ds,\label{master2}\\ \dot{\tilde{\rho}}_{13}(t)&=-\alpha^2\int\limits_0^t\big(2\tilde{\rho}_{13}(s)\mathrm e^{2i\delta(t - s)}-\tilde{\rho}_{24}(s ) \mathrm
e^{2i\delta(t+s)}\bigr)\ ds,\label{master3}\\ \dot{\tilde{\rho}}_{14}(t)&=-\alpha^2\int\limits_0^t 2\tilde{\rho}_{13}(s ) \cos[2\delta(t - s)]\ ds,\label{master4}\\ \dot{\tilde{\rho}}_{22}(t)&=-\alpha^2\int\limits_0^t\big(2\tilde{\rho}_{22}(s)-\tilde{\rho}_{11}(s ) -\tilde{\rho}_{44}(s)\bigr)\cos[2\delta(t - s)]\ ds,\label{master5}\end{aligned}\ ] ] @xmath182\ ds,\label{master6}\\ \dot{\tilde{\rho}}_{24}(t)&=-\alpha^2\int\limits_0^t\big(2\tilde{\rho}_{24}(s)\mathrm e^{2i\delta(s - t)}-\tilde{\rho}_{13}(s ) \mathrm
e^{-2i\delta(t+s)}\bigr)\ ds,\label{master7}\\ \dot{\tilde{\rho}}_{33}(t)&=-\alpha^2\int\limits_0^t\big(2\tilde{\rho}_{33}(s)-\tilde{\rho}_{11}(s ) -\tilde{\rho}_{44}(s)\bigr)\cos[2\delta(t - s)]\ ds,\label{master8}\\ \dot{\tilde{\rho}}_{34}(t)&=-\alpha^2\int\limits_0^t\big(2\tilde{\rho}_{34}(s)\mathrm e^{2i\delta(s - t)}-\tilde{\rho}_{12}(s ) \mathrm
e^{-2i\delta(t+s)}\bigr)\ ds,\label{master9}\\ \dot{\tilde{\rho}}_{44}(t)&=-\alpha^2\int\limits_0^t\big(2\tilde{\rho}_{44}(s)-\tilde{\rho}_{22}(s ) -\tilde{\rho}_{33}(s)\bigr)\cos[2\delta(t - s)]\ ds\label{master10}.\end{aligned}\ ] ] some of the above equations can be solved under a time - local approximation for which the matrix elements @xmath183 are replaced by @xmath184 .
one can find that ( @xmath84 and @xmath185 given in units of @xmath186 and @xmath4 respectively ) @xmath187\exp\bigl\{\frac{1}{\delta^2}[\cos(2 \delta t)-1]\bigl\}\nonumber \\&\qquad + 2(\rho_{11}^0-\rho_{44}^0)\exp\bigl\{\frac{1}{2\delta^2}[\cos(2 \delta t)-1]\bigr\}\biggr\},\label{ms11 } \\
\tilde{\rho}_{22}(t)&=\frac{1}{4}\biggl\{1+\bigl[-1 + 2(\rho_{22}^0+\rho_{33}^0)\bigr]\exp\bigl\{\frac{1}{\delta^2}[\cos(2 \delta t)-1]\bigl\}\nonumber \\&\qquad + 2(\rho_{22}^0-\rho_{33}^0)\exp\bigl\{\frac{1}{2\delta^2}[\cos(2 \delta t)-1]\bigr\}\biggr\ } , \\
\tilde{\rho}_{33}(t)&=\frac{1}{4}\biggl\{1+\bigl[-1 + 2(\rho_{33}^0+\rho_{22}^0)\bigr]\exp\bigl\{\frac{1}{\delta^2}[\cos(2 \delta t)-1]\bigl\}\nonumber \\&\qquad + 2(\rho_{33}^0-\rho_{22}^0)\exp\bigl\{\frac{1}{2\delta^2}[\cos(2 \delta t)-1]\bigr\}\biggr\},\end{aligned}\ ] ] @xmath188\exp\bigl\{\frac{1}{\delta^2}[\cos(2 \delta t)-1]\bigl\}\nonumber \\&\qquad + 2(\rho_{44}^0-\rho_{11}^0)\exp\bigl\{\frac{1}{2\delta^2}[\cos(2 \delta t)-1]\bigr\}\biggr\ } , \\
\tilde{\rho}_{14}(t)&=\rho_{14}^0\exp\bigl\{\frac{1}{\delta^2}[\cos(2 \delta t)-1]\bigr\},\\ \tilde{\rho}_{23}(t)&=\rho_{23}^0\exp\bigl\{\frac{1}{\delta^2}[\cos(2 \delta t)-1]\bigr\}\label{ms23}.\end{aligned}\ ] ] these solutions describe approximately the dynamics at short times .
in fact , the smaller the coupling constant @xmath84 , the better these solutions are . note that when @xmath189 ( i.e. nonlocal dynamics ) , then @xmath190\bigr\}\to \mathrm e^{-2 t^2/n } , \qquad n=1,2.\ ] ] thus the second order time - local master equation shows that the nonlocal dynamics , or , in general , the short time behavior follow a gaussian decay law .
note that the solutions corresponding to the diagonal elements reproduce their asymptotic limit , namely , @xmath191 .
however , those corresponding to the off - diagonal elements fail to reproduce the steady state , since , for example , equation ( [ ms23 ] ) implies that @xmath192 . to end our discussion
let us remark that equations ( [ master2 ] ) , ( [ master3 ] ) , ( [ master7 ] ) and ( [ master9 ] ) can be analytically solved only when @xmath189 .
for instance ( see figure [ figure7 ] ) , @xmath193.\label{ms12}\end{aligned}\ ] ]
in summary we have investigated the dynamics of two qubits coupled to separate spin star environment via heisenberg @xmath0 interactions .
we have derived the exact form of the time evolution operator and calculated the matrix elements of the reduced density operator .
the analysis of the evolution in time of the concurrence and the purity shows that decoherence can be minimized by allowing the central qubits to strongly interact with each other .
the short - time behavior , studied by deriving the second - order master equation , is found to be gaussian .
the next step may consist in considering more central qubits , and investigate whether the above results still hold .
using trace properties of the lowering and raising operators , it can be shown that the nonzero matrix elements corresponding to the initial maximally entangled states @xmath194 are explicitly given by : @xmath195 those associated with the initial state @xmath196 read : @xmath197 99 baxter r j , 1982 _ exactly solved models in statistical mechanics _ ( academic press , london ) breuer h p and petruccione f 2002 _ the theory of open quantum systems _ ( oxford university press , oxford ) zurek w h 1991 _ phys . today _ * 44 * no
. 10 36 divincenzo d p and loss d 2000 _ j. magn .
magn . matter .
_ * 200 * 202 zurek w h 2003 _ rev
. mod .
_ * 75 * , 715 - 775 nielsen m a and chuang
i l 2000 _ quantum computation and quantum information _
( cambridge university press , cambridge ) bennett c h and wiesner s j 1993_phys .
lett . _ * 69 * 2881 bennett c h and divincenzo d p 2000 _ nature _ * 404 * 247 bennett c h , brassard g , crpeau c , jozsa r , peres a and wootters w k 1993 _ phys . rev . lett . _
* 70 * 1895 bennett c h , divincenzo d p , smolin j a and wootters w k 1996 _ phys .
a * 54 * 3824 lloyd s 1004 _ science _ * 261 * 1569 ( 1993 ) ekert a and josza r 1998 _ philos .
. lond . _ a * 356 * 1769 loss d and divincenzo d p 1998 _ phys . rev . _ a * 57 * 120 burkard g loss d and divincenzo d p 1999 _ phys . rev . _ b * 59 * 2070 prokofev n v and stamp p c e 2000 _ rep
. prog .
phys . _ * 63 * 669 zhang w , dobrovitski v v , al - hassanieh k a , dagotto e and harmon b n 2006 _ phys . rev . _
205313 amico l , fazio r , osterloh a and vedral v 2008 _ rev . mod.phys._ * 80 * , 517 - 576 breuer h p , burgarth d and petruccione f 2004 _ phys.rev .
_ b * 70 * 045323 yuan x z , goan h j and zhu k d 2007 _ phys . rev .
_ b * 75 * jing j and l z 2007 _ phys .
_ b * 75 * 174425 jing j , l z and yang g 2007 _ phys .
_ a*76 * 032322 hamdouni y and petruccione f 2007 _ phys . rev . _ b * 76 * 174306 hamdouni y , fannes m and petruccione f 2006 _ pys .
rev . _ b * 73 * 245323 hamdouni y 2007 _ j. phys
. a : math .
_ * 40 * 11569 von waldenfels w 1990 _ sminaire de probabilit ( starsburg ) _ tome 24 349 - 356 ( springer - verlag , berlin ) wootters w k 1998 _ phys .
lett . _ * 80 * 2245 yu t and eberly j h 2004 _ phys
. rev .
lett . _ * 93 * 140404 hamdouni y 2008 _ preprint _
danos m and rafelski j 1984 _ pocketbook of mathematical functions _ ( verlag harri deutsch , frankfurt ) | exact analytical results for the dynamics of two interacting qubits each of which is embedded in its own spin star bath are presented .
the time evolution of the concurrence and the purity of the two - qubit system is investigated for finite and infinite numbers of environmental spins .
the effect of qubit - qubit interactions on the steady state of the central system is investigated . |
throughout the paper we will take an embedded graph to mean a graph embedded in @xmath4 .
we call a graph @xmath5 _ intrinsically knotted _ if every embedding of the graph contains a non - trivially knotted cycle .
conway and gordon @xcite showed that @xmath0 , the complete graph with seven vertices , is an intrinsically knotted graph .
foisy @xcite showed that @xmath6 is also intrinsically knotted .
a graph @xmath7 is a _ minor _ of another graph @xmath5 if it can be obtained from @xmath5 by contracting or deleting some edges .
if a graph @xmath5 is intrinsically knotted and has no proper minor that is intrinsically knotted , @xmath5 is said to be _ minor minimal intrinsically knotted_. robertson and seymour @xcite proved that there are only finitely many minor minimal intrinsically knotted graphs , but finding the complete set is still an open problem .
a @xmath1 _ move _ is an exchange operation on a graph that removes all edges of a triangle @xmath8 and then adds a new vertex @xmath9 and three new edges @xmath10 and @xmath11 .
its reverse operation is called a @xmath12 _ move _ as follows : since a @xmath1 move preserves intrinsic knottedness @xcite , we will concentrate on triangle - free graphs in this paper .
it is known that @xmath0 and the 13 graphs obtained from @xmath0 by @xmath1 moves , and @xmath6 and the 25 graphs obtained from @xmath6 by @xmath1 moves are minor minimal intrinsically knotted .
johnson , kidwell and michael @xcite showed that intrinsically knotted graphs have at least 21 edges . recently two groups , working independently , showed that @xmath0 and the 13 graphs obtained from @xmath0 by @xmath1 moves are the only intrinsically knotted graphs with 21 edges .
this gives us the complete set of 14 minor minimal intrinsically knotted graphs with 21 edges .
lee , kim , lee and oh @xcite proceeded by showing that the only triangle - free intrinsically knotted graphs with 21 edges are @xmath13 and @xmath14 .
barsotti and mattman @xcite relied on connections with 2-apex graphs .
we say that a graph is _ intrinsically knotted or completely 3-linked _ if every embedding of the graph into the 3-sphere contains a nontrivial knot or a 3-component link each of whose 2-component sublinks is nonsplittable .
hanaki , nikkuni , taniyama and yamazaki @xcite constructed 20 graphs derived from @xmath13 and @xmath14 by @xmath12 moves as in figure [ fig1 ] , and they showed that these graphs are minor minimal intrinsically knotted or completely 3-linked graphs .
furthermore they showed that the six graphs @xmath15 , @xmath16 , @xmath17 , @xmath18 , @xmath19 and @xmath20 are not intrinsically knotted .
this has also been shown by goldberg , mattman and naimi independently @xcite . and
@xmath21 more relative graphs ] we say two graphs @xmath5 and @xmath22 are _ cousins _ of each other if @xmath22 is obtained from @xmath5 by a finite sequence of @xmath1 and @xmath12 moves .
the set of all cousins of @xmath5 is called the @xmath5 _ family_. the @xmath6 family consists of 58 graphs , of which 26 graphs were previously known to be minor minimal intrinsically knotted . in @xcite
they showed that the remaining 32 graphs are also minor minimal intrinsically knotted .
as in figure [ fig2 ] the graph @xmath2 is obtained from @xmath15 by adding the new edge @xmath23 .
the @xmath2 family consists of 110 graphs .
they also showed that all of these graphs are intrinsically knotted , and exactly 33 of them are minor minimal intrinsically knotted . ] by combining the @xmath6 family and the @xmath2 family , all 168 graphs were already known to be intrinsically knotted graphs with 22 edges , and 14 graphs among them are triangle - free .
these triangle - free intrinsically knotted 14 graphs are the four cousins 29 , 31 , 42 and 53 in the @xmath6 family and the ten cousins 43 , 56 , 87 , 89 , 94 , 97 , 99 , 105 , 109 and 110 in the @xmath2 family .
especially , only the two cousins , 94 and 110 , of the @xmath2 family drawn in figure [ fig3 ] have at least two vertices with degree 5 or more . indeed
if we contract the edges @xmath24 and @xmath25 of the cousins 94 and 110 , we get the minor minimal intrinsically knotted graphs @xmath26 and @xmath27 of the @xmath0 family , respectively .
family ] we introduce a previously unknown graph @xmath3 , shown in figure [ fig4 ] , that is a triangle - free intrinsically knotted ( but not minor minimal ) graph with 22 edges having four vertices with degree 5 .
note that @xmath3 does not appear in the @xmath6 family or the @xmath2 family . indeed
if we contract the edge @xmath28 , we get the minor minimal intrinsically knotted graph @xmath29 of the @xmath0 family . ]
the goal of this research is to determine the number of intrinsically knotted graphs with 22 edges .
we reach half the destination . in this paper
we find all triangle - free intrinsically knotted graphs with 22 edges that have at least two vertices with degree 5 or a vertex with degree larger than 5 .
[ thm : main ] there are exactly three triangle - free intrinsically knotted graphs with 22 edges having at least two vertices with degree 5 , the two cousins 94 and 110 of the @xmath2 family , and a previously unknown graph named @xmath3 .
these graphs are shown in figure [ fig3 ] and figure [ fig4 ] .
furthermore , there is no triangle - free intrinsically knotted graph with 22 edges that has a vertex with degree 6 or more .
the remaining sections of the paper are devoted to the proof of theorem [ thm : main ] .
this paper relies on the technical machinery developed in @xcite . in section [ sec : term ] , we review this material .
henceforth , @xmath30 denotes a vertex of maximal degree in @xmath5 .
the proof is divided into three parts according to the degree of @xmath30 . in section
[ sec : deg6 ] , we show that any graph @xmath5 with @xmath31 can not be triangle - free intrinsically knotted . in section [ sec : deg5dist2 ] , we show that the only triangle - free intrinsically knotted graphs with two vertices @xmath30 and @xmath32 of degree 5 with distance at least 2 are cousin 110 of the @xmath2 family and the graph @xmath3 . in section [ sec : deg5dist1 ] , we show that the only triangle - free intrinsically knotted graph with two vertices @xmath30 and @xmath32 of degree 5 with distance 1 is cousin 94 of the @xmath2 family .
henceforth , let @xmath33 denote a triangle - free graph with 22 edges . here
@xmath34 and @xmath35 denote the sets of vertices and edges of @xmath5 respectively . for any two distinct vertices
@xmath30 and @xmath32 , let @xmath36 denote the graph obtained from @xmath5 by deleting two vertices @xmath30 and @xmath32 , and then contracting edges adjacent to vertices of degree 1 or 2 , one by one repeatedly , until no vertices of degree 1 or 2 remain .
removing vertices means deleting interiors of all edges adjacent to these vertices and remaining isolated vertices .
the distance , denoted by dist@xmath37 , between @xmath30 and @xmath32 is the number of edges in the shortest path connecting them .
the degree of @xmath30 is denoted by @xmath38 . to count the number of edges of @xmath39
, we have some notation .
* @xmath40 is the set of edges that are adjacent to @xmath30 .
* @xmath41 for a subset @xmath42 of @xmath34 .
* @xmath43 * @xmath44 * @xmath45 * @xmath46 obviously in @xmath47 for some distinct vertices @xmath30 and @xmath32 , each vertex of @xmath48 has degree 1 .
also each vertex of @xmath49 ( but not of @xmath48 ) and @xmath50 has degree 2 .
therefore to derive @xmath39 all edges adjacent to @xmath51 and @xmath48 are deleted from @xmath5 , followed by contracting one of the remaining two edges adjacent to each vertex of @xmath52 , @xmath53 , @xmath50 and @xmath54 as in figure [ fig5](a ) .
thus we have the following equation counting the number of edges of @xmath39 which is called the _ count equation _ ; @xmath55 for short , write @xmath56 and @xmath57 . provided that @xmath30 and @xmath32 are adjacent vertices ( i.e. dist@xmath58 ) , @xmath59 and @xmath54 are all empty sets because @xmath5 is triangle - free .
note that the derivation of @xmath39 must be handled slightly differently when there is a vertex @xmath60 in @xmath34 such that more than one vertex of @xmath61 is contained in @xmath48 as in figure [ fig5](b ) . in this case
we usually delete or contract more edges even though @xmath60 is not in @xmath54 .
, title="fig : " ] + ( a ) ( b ) the following proposition , which was mentioned in @xcite , gives an important condition for a graph not to be intrinsically knotted .
note that @xmath62 is a triangle - free graph and every vertex has degree 3 .
[ prop : planar ] if @xmath39 satisfies one of the following two conditions , then @xmath5 is not intrinsically knotted .
* @xmath63 , or * @xmath64 and @xmath39 is not homeomorphic to @xmath62 . if @xmath63 , then @xmath39 is a planar graph . also if @xmath64 , then @xmath39 is either a planar graph or homeomorphic to @xmath62 .
it is known that if @xmath39 is planar , then @xmath5 is not intrinsically knotted .
since the intrinsically knotted graphs with 21 edges are known , it is sufficient to consider simple and connected graphs having no vertex of degree 1 or 2 .
our process is to construct all possible such triangle - free graphs @xmath5 with 22 edges , delete two suitable vertices @xmath30 and @xmath32 of @xmath5 , and then count the number of edges of @xmath39 . if @xmath39 has 9 edges or less and is not homeomorphic to @xmath62 , then we can use proposition [ prop : planar ] to show that @xmath5 is not intrinsically knotted .
we introduce more notation .
let @xmath7 be a connected subgraph of @xmath5 .
let @xmath65 denote the set of all edges of @xmath66 , and @xmath67 denote the set of all vertices of @xmath5 that are not a vertex of @xmath7 with degree more than 1 . * @xmath68 for some @xmath69 * @xmath70 for some @xmath71 with @xmath72 * @xmath73 = [ \deg(\overline{v}_1 ) , \cdots , \deg(\overline{v}_n)]$ ] * @xmath74| = \deg(\overline{v}_1)+ \cdots + \deg(\overline{v}_n)$ ] we will use @xmath75 and @xmath76 without referring to @xmath7 , and sometimes we call @xmath75 an _ extra edge_. to visualize @xmath5 , we basically proceed through the following steps .
first choose a proper subgraph @xmath7 of @xmath5 .
draw @xmath65 apart from @xmath7 as in figure [ fig6](a ) .
@xmath74|$ ] indicates the number of vertices of degree 1 of @xmath7 and @xmath65 , and they are merged into vertices of @xmath67 as in figure [ fig6](b ) . the graph in figure [ fig6](b ) is an example representing @xmath77 , @xmath74| = 20 $ ] and @xmath73=[4,4,3,3,3,3]$ ] .
, title="fig : " ] + ( a ) ( b )
recall that @xmath5 is a triangle - free graph with 22 edges , every vertex has degree at least 3 and @xmath30 has the maximal degree among all vertices . in this section
we will show that for some @xmath78 , @xmath63 .
then as a conclusion @xmath5 is not intrinsically knotted by proposition [ prop : planar ] .
obviously @xmath79 because , otherwise , @xmath5 would have at least 24 edges .
if @xmath80 , then @xmath81 by the same reason as above .
let @xmath32 be any vertex in @xmath82 . then @xmath83 and @xmath84 .
note that @xmath85 . by the count equation , @xmath63 in @xmath39 .
now assume that @xmath86 . since @xmath5 has 22 edges , @xmath87 .
assume that @xmath88 .
let @xmath60 be any vertex in @xmath52 .
choose a vertex @xmath32 that has the maximal degree from @xmath89 .
then @xmath90 and @xmath91 since @xmath92 when @xmath93 .
thus @xmath63 .
if @xmath94 or @xmath95 , then there must be a vertex , call it @xmath32 , of degree 4 in @xmath82 . if all the vertices of @xmath96 other than @xmath30 have degree 3 , then @xmath97 and @xmath98 , yielding @xmath63
otherwise , there is a vertex @xmath60 of degree at least 4 in @xmath99 .
consider @xmath100 where @xmath101 .
when @xmath102 , @xmath103 because @xmath104 .
when @xmath94 , the graph consisting of the edges of @xmath105 is @xmath5 itself as in figure [ fig7 ] .
at least one edge in @xmath106 must be adjacent to a vertex of @xmath107 other than @xmath32 .
this implies that @xmath108 so that @xmath103 . ]
in this section we will consider the case that there are two vertices @xmath30 and @xmath32 of degree 5 with distance at least 2 . indeed , if dist@xmath111 , @xmath112 because @xmath105 is disjoint from @xmath113 ( similarly for @xmath114 ) .
this implies that @xmath115 and @xmath116 . since they are disjoint , @xmath117 . from now on
we assume that dist@xmath118 .
let @xmath7 be a subgraph of @xmath5 which consists of the edges of @xmath119 and the associated vertices . before dividing into several cases , we give some notation : @xmath120 , @xmath121 , @xmath122 , @xmath123 and @xmath124 ( similarly for @xmath32 ) as indicated in figure [ fig8 ] .
furthermore we will write @xmath125 , @xmath126 and similarly for the others . by abuse of notation
, we will use the same notation for a set and its size .
for example , @xmath127 will also mean @xmath128 .
obviously we may assume that @xmath129 . ] by counting the number of edges of @xmath7 , @xmath130 .
since @xmath131 , @xmath132 .
thus we have the following inequality ; @xmath133 if @xmath134 , then @xmath135 which does not agree with the inequality above . if @xmath136 , then @xmath63
. thus we consider only @xmath137 or @xmath95 .
now we divide into six cases as follows .
if @xmath137 , we divide into the two cases where @xmath138 and @xmath139 .
we can ignore the case of @xmath140 because of the inequality . if @xmath141 , we have four cases : @xmath142 or @xmath95 . indeed , in the case of @xmath141 , @xmath143 .
furthermore we may assume that @xmath144 ( so @xmath145 ) and @xmath146 ( i.e. @xmath147 ) , otherwise @xmath63 .
cousin 110 of the @xmath2 family appears in this case . note that
@xmath149 and there is no extra edge .
suppose that @xmath150 so that @xmath151 . if @xmath152 , then @xmath153 .
recall that @xmath154 and @xmath155 are the unique elements of the sets @xmath127 and @xmath156 respectively .
since @xmath74|=15 $ ] , the associated 15 edges are adjacent to vertices of @xmath67 . among these 15 edges ,
two edges of @xmath157 and three edges of @xmath113 must be adjacent to different vertices because @xmath5 is triangle - free .
so @xmath67 has at least 5 vertices , and we get @xmath73 = [ 3,3,3,3,3]$ ] .
therefore @xmath158 so that @xmath159 .
if @xmath160 , then @xmath74|=16 $ ] . similarly among the associated 16 edges , one edge of @xmath161 and four edges of @xmath113 are adjacent to different vertices of @xmath67 .
so @xmath67 has at least 5 vertices , and we get @xmath73 = [ 4,3,3,3,3]$ ] . therefore @xmath158 or @xmath162 , yielding @xmath159 .
suppose that @xmath163 so that @xmath164 . if @xmath152 , then @xmath74|=14 $ ] . among the associated 14 edges , three edges of @xmath165 and two edges of @xmath113 are adjacent to different vertices
but this is impossible because each vertex of @xmath67 has degree at least 3 .
if @xmath160 , then @xmath74|=15 $ ] . among the associated 15 edges , three edges of @xmath165 and three edges of @xmath113 are adjacent to different vertices , but this is also impossible .
suppose that @xmath166 so that @xmath167 . if @xmath152 , then @xmath168 .
let @xmath169 be the unique vertex of @xmath170 or @xmath171 .
since no edge of @xmath161 , @xmath157 or @xmath172 can be adjacent to @xmath169 , two edges of @xmath173 must simultaneously be adjacent to @xmath169 to make @xmath174 . if @xmath160 , then @xmath74|=14 $ ] . among the associated 14 edges , three edges of @xmath165 and two edges of @xmath113 are adjacent to different vertices .
but this is impossible because each vertex of @xmath67 has degree at least 3 .
finally suppose that @xmath175 so that @xmath176 .
recall that @xmath177 .
if @xmath152 , then @xmath178 and @xmath74|=12 $ ] . thus @xmath73 = [ 5,4,3]$ ] , @xmath179 $ ] or @xmath180 $ ] .
when @xmath73 = [ 5,4,3]$ ] , @xmath5 is homeomorphic to cousin 110 of the @xmath2 family which is intrinsically knotted but not minor minimal and shown in figure [ fig3 ] . when @xmath73 = [ 4,4,4]$ ] , three edges of each @xmath172 are adjacent to three different vertices of @xmath67 .
therefore @xmath181 so that @xmath182 .
since @xmath183 has vertex @xmath184 of degree 5 , it is not homeomorphic to @xmath62 .
when @xmath73 = [ 3,3,3,3]$ ] , three edges of @xmath165 are adjacent to three different vertices of @xmath67 all of which have degree 3 .
therefore @xmath185 so that @xmath186 .
since @xmath187 has vertex @xmath32 of degree 4 , it is not homeomorphic to @xmath62 . if @xmath160 , then @xmath168
let @xmath169 be the unique vertex of @xmath170 or @xmath171 . since no edge of @xmath161 or @xmath172 can be adjacent to @xmath169 , two edges of @xmath188 are simultaneously adjacent to @xmath169 to make @xmath174 . in this case , @xmath190 .
suppose that @xmath150 so that @xmath151 and @xmath191 . if there is a vertex @xmath192 of degree 5 in the set @xmath193 , then @xmath149 and @xmath74|=15 $ ] . among these 15 edges ,
four edges of @xmath194 are adjacent to different vertices of @xmath67 .
thus we get @xmath73 = [ 5,4,3,3]$ ] , @xmath195 $ ] or @xmath196 $ ] . in any case @xmath197 so that @xmath198 .
if three vertices @xmath192 , @xmath199 and @xmath200 have degree 4 , then @xmath201 .
therefore there is an extra edge @xmath202 and @xmath74|=16 $ ] .
thus we get @xmath73 = [ 5,5,3,3]$ ] , @xmath203 $ ] , @xmath204 $ ] or @xmath205 $ ] .
when @xmath73 = [ 4,3,3,3,3]$ ] , @xmath206 so that @xmath63 .
for the other three cases , @xmath207 .
assume that @xmath208 and @xmath209 are the endpoints of @xmath202 .
see figure [ fig9](a ) .
since @xmath5 is triangle - free , we may assume that three edges of @xmath113 are adjacent to @xmath208 , @xmath210 and @xmath211 .
similarly two edges of each @xmath212 are adjacent to @xmath210 and @xmath211 , and the other edge of each @xmath212 is adjacent to @xmath208 or @xmath209 .
furthermore one edge of each @xmath213 is adjacent to @xmath209 to avoid a triangle .
therefore @xmath214 so that @xmath215 . since @xmath216 has a bigon containing two vertices @xmath30 and @xmath209 , it is not homeomorphic to @xmath62 .
suppose that @xmath163 .
then @xmath74|=14 $ ] or 15 depending on the existence of an extra edge .
thus @xmath67 has five vertices all of degree 3 or else fewer than five vertices . in either case
three edges of @xmath217 and two edges of @xmath113 must be adjacent to different vertices of @xmath67 .
thus @xmath218 or it is impossible to construct @xmath5 .
suppose that @xmath166 so that @xmath167 . if @xmath192 has degree 5 , then @xmath74|=13 $ ] . since four edges of @xmath194 are adjacent to four different vertices of @xmath67 , @xmath73 = [ 4,3,3,3]$ ] .
thus @xmath219 so that @xmath220 .
if @xmath192 has degree 4 , then there is an extra edge @xmath202 and @xmath74|=14 $ ] .
let @xmath208 be a vertex of @xmath67 which is the remaining vertex of @xmath96 .
see figure [ fig9](b ) . since @xmath5 is triangle - free
, only @xmath202 and one edge of @xmath194 can be adjacent to @xmath208 .
let @xmath209 be the other endpoint of @xmath202 .
thus we get @xmath73 = [ 5,3,3,3]$ ] or @xmath221 $ ] .
this implies that two edges of @xmath194 are adjacent to @xmath210 and @xmath211 , and three edges of each @xmath172 are adjacent to @xmath209 , @xmath210 and @xmath211 .
since at least two vertices of @xmath67 have degree 3 , @xmath222 so that @xmath223 .
since @xmath224 has vertex @xmath30 with degree 4 , it is not homeomorphic to @xmath62 . with @xmath189,title="fig : " ] + ( a ) ( b ) suppose that @xmath175 . then @xmath201 so that there is an extra edge @xmath202 and @xmath74|=13 $ ] .
since two edges of @xmath165 can not simultaneously be adjacent to the endpoints of @xmath202 , @xmath67 needs four vertices and so @xmath73 = [ 4,3,3,3]$ ] . since @xmath225 , @xmath186 .
since @xmath187 has vertex @xmath32 with degree 4 , it is not homeomorphic to @xmath62 .
recall that @xmath140 , @xmath227 , @xmath228 ( as mentioned before case 4.1 . ) and @xmath143 .
suppose that @xmath150 or 1 .
note that @xmath229 . since @xmath149 and @xmath230 , at least three edges of @xmath231 must be adjacent to different vertices of @xmath193 , a contradiction to @xmath232 .
suppose that @xmath166 .
since @xmath74|=12 $ ] , we get @xmath73 = [ 5,4,3]$ ] , @xmath179 $ ] or @xmath180 $ ] . when @xmath73 = [ 5,4,3]$ ] or @xmath179 $ ] , @xmath143 and @xmath39 has vertices with degree 4 or 5 .
consider @xmath73 = [ 3,3,3,3]$ ] .
if @xmath233 , then @xmath234 so that @xmath235
. assume that @xmath236 .
so @xmath237 and we have the graph of figure [ fig10](a ) .
then @xmath238 and @xmath239 has vertex @xmath240 with degree 4 . with @xmath226 or @xmath148,title="fig : " ] + ( a ) ( b ) the new graph @xmath3 appears in this case .
note that @xmath138 , @xmath241 , @xmath228 , @xmath143 and possibly there is an extra edge @xmath202 .
suppose that @xmath150 .
note that @xmath242 .
then two edges of @xmath231 must be adjacent to different vertices of @xmath193 and another edge of @xmath231 is @xmath202 .
the same thing happens for @xmath243 so that @xmath5 has the triangle consisting of @xmath202 and two edges of @xmath113 .
suppose that @xmath163 .
then , even though one edge of @xmath231 is @xmath202 , two edges of @xmath231 must be adjacent to @xmath192 , a contradiction .
suppose that @xmath166 .
since @xmath74|=13 $ ] , we get @xmath73 = [ 5,5,3]$ ] , @xmath244 $ ] or @xmath245 $ ] .
when @xmath73 = [ 5,5,3]$ ] or @xmath244 $ ] , @xmath143 and @xmath39 has vertices with degree 4 or 5 .
suppose @xmath73 = [ 4,3,3,3]$ ] . if three vertices of @xmath246 other than @xmath30 and @xmath32 have all degree 3 , then @xmath186 and @xmath187 has vertex @xmath32 with degree 4 , a contradiction .
thus one edge of each @xmath172 is adjacent to @xmath208 of degree 4 .
if @xmath233 , then @xmath247 and @xmath248 contains @xmath208 , yielding @xmath235 . assume that @xmath236 .
since @xmath54 is empty , one edge of @xmath161 is adjacent to @xmath208 .
now the third edge of each @xmath249 , @xmath250 , is adjacent to a vertex of set @xmath156 , otherwise @xmath182 and @xmath183 has a vertex @xmath155 or @xmath251 with degree 4 .
thus we have the graph of figure [ fig10](b ) .
this graph @xmath5 is indeed @xmath3 .
again the graph @xmath3 appears in this case . note that @xmath252 , @xmath153 , @xmath228 , @xmath143 and possibly there are two extra edges @xmath202 and @xmath253 .
suppose that @xmath150 .
obviously extra edges and edges of only @xmath194 and @xmath254 can be adjacent to @xmath255 and @xmath256 . since an extra edge can not be adjacent to both @xmath255 and @xmath256 simultaneously ,
there are two edges of @xmath194 adjacent to @xmath255 and @xmath256 respectively .
therefore @xmath257 and @xmath258 , yielding @xmath215 .
since @xmath216 has vertex @xmath30 with degree 4 , it is not homeomorphic to @xmath62 .
suppose that @xmath163 .
then two extra edges and one edge of @xmath194 must be adjacent to @xmath255 .
see figure [ fig11](a ) . since @xmath255 has degree 4 and @xmath74|=15 $ ] , we get @xmath73 = [ 5,4,3,3]$ ] or @xmath195 $ ] .
this implies that @xmath67 has four vertices including @xmath255 .
then the remaining two edges of @xmath194 and two extra edges must be adjacent to three vertices in @xmath67 other than @xmath255 .
thus @xmath5 has a triangle consisting of two edges of @xmath194 and an extra edge . with @xmath189,title="fig : " ] + ( a ) ( b ) suppose that @xmath166 .
since @xmath74|=14 $ ] , we get @xmath73 = [ 5,5,4]$ ] , @xmath259 $ ] or @xmath221 $ ] . when @xmath73 = [ 5,5,4]$ ] , @xmath143 and @xmath39 has vertices with degree 4 or 5 in @xmath67 .
when @xmath73 = [ 5,3,3,3]$ ] , @xmath260 .
therefore @xmath223 and @xmath224 has vertex @xmath30 with degree 4 .
when @xmath73 = [ 4,4,3,3]$ ] , we assume that @xmath261 by the same reason as before .
recall that @xmath208 and @xmath209 have degree 4 , and @xmath210 and @xmath211 have degree 3 .
see figure [ fig11](b ) .
two edges of each @xmath172 must be adjacent to @xmath208 and @xmath209 respectively .
the remaining two edges of @xmath165 and @xmath262 are adjacent to one vertex , say @xmath210 , otherwise @xmath263 and @xmath264 , yielding @xmath235
. furthermore @xmath265 must be empty , so the remaining edge of @xmath266 is adjacent to @xmath155 .
also @xmath54 must be empty , so one edge of each @xmath213 is adjacent to @xmath208 or @xmath209 .
combining all these conditions , we get the graph @xmath3 which is the same graph as in figure [ fig10](b ) with @xmath267 and @xmath268 exchanged . in this case
@xmath270 , @xmath271 , @xmath143 and possibly there are three extra edges @xmath202 , @xmath253 and @xmath272 .
first suppose that all three edges each of which is from @xmath273 , @xmath274 , are adjacent to one vertex @xmath60 which is neither @xmath30 nor @xmath32 .
this means that @xmath61 contains @xmath48 .
if @xmath275 , then @xmath63 . if @xmath276 , then @xmath5 is the union of two components which are @xmath62 and @xmath277 . since neither component is intrinsically knotted and they meet at only two vertices @xmath30 and @xmath32 , their union @xmath5 can not be intrinsically knotted .
now assume that the three edges mentioned above are not all adjacent to one vertex . recall that @xmath278 as mentioned before case 4.1 .
suppose that @xmath150 .
if one of @xmath193 , say @xmath192 , has degree 5 , then @xmath215 and @xmath216 has vertex @xmath30 with degree 4 .
thus @xmath192 and @xmath199 have degree 4 .
since @xmath74|=17 $ ] , we get @xmath73 = [ 5,5,4,3]$ ] , @xmath279 $ ] , @xmath280 $ ] or @xmath281 $ ] . since @xmath282 and @xmath278 , @xmath67 has four vertices with degree more than 3 .
thus we consider the only case @xmath279 $ ] which induces the graph of figure [ fig12](a ) .
note that at least one edge of @xmath194 is adjacent to a vertex of @xmath171 .
therefore @xmath283 and @xmath284 , yielding @xmath215 . since @xmath216 has vertex @xmath199 with degree 4 , it is not homeomorphic to @xmath62
. suppose that @xmath163 . since @xmath74|=16 $ ] , by reasoning similar to that above , we need only consider the case @xmath73 = [ 4,4,4,4]$ ] .
take a vertex @xmath208 with degree 4 from @xmath285 .
then @xmath286 and @xmath287 , yielding @xmath288 as in figure [ fig12](b ) .
since @xmath289 has vertex @xmath32 with degree 4 , it is not homeomorphic to @xmath62 .
suppose that @xmath166 .
since @xmath74|=15 $ ] , we get @xmath73 = [ 5,5,5]$ ] , @xmath290 $ ] , @xmath195 $ ] or @xmath196 $ ] . since @xmath146 , @xmath73 \neq [ 3,3,3,3,3]$ ] . the existence of three extra edges implies @xmath73 \neq [ 5,5,5]$ ] because @xmath5 is triangle - free .
thus @xmath73 = [ 5,4,3,3]$ ] or @xmath195 $ ] , so that @xmath67 consists of four vertices .
since three edges of each @xmath172 are adjacent to different vertices of @xmath67 and @xmath5 is triangle - free , we have @xmath291 producing the graph of figure [ fig12](c ) .
let @xmath208 be the vertex that the three extra edges are adjacent to .
note that at least one of @xmath292 , @xmath293 , must have degree greater than 3 .
now we consider @xmath5 where two sets of vertices @xmath267 and @xmath268 are exchanged .
then in @xmath5 with the new labelling , @xmath294 .
but we ve already completed the argument for all cases except the case @xmath295 and @xmath166 . with @xmath270,title="fig : " ] + ( a ) ( b ) ( c )
in this section we will consider the case that there are exactly two vertices @xmath30 and @xmath32 of degree 5 and their distance is 1 .
so all the other vertices have degree 3 or 4 .
we may assume that @xmath297 .
let @xmath7 be a subgraph of @xmath5 which consists of the edges of @xmath119 and the associated vertices .
now we divide into five cases according to the number of vertices of @xmath52 . in this case
all eight vertices of @xmath82 and @xmath96 except @xmath30 and @xmath32 have degree 4 .
since @xmath201 and there is only one extra edge , @xmath74|=18 $ ] .
so we get @xmath73 = [ 4,4,4,3,3]$ ] or @xmath299 $ ] .
both cases are impossible because the set @xmath96 has four vertices of degree 4 in @xmath67 .
let @xmath60 be the unique vertex of @xmath52 . since @xmath82 has three vertices of degree 4 , @xmath301 and there are two extra edges @xmath202 and @xmath253 .
therefore @xmath74|=19 $ ] and @xmath73 = [ 4,4,4,4,3]$ ] or @xmath302 $ ] .
the latter case is impossible because @xmath96 has three or four vertices of degree 4 . in the former case
, only @xmath303 has degree 3 .
first assume that @xmath96 has four vertices of degree 4 and so @xmath303 is not a vertex of @xmath96 .
we can choose a vertex @xmath169 from @xmath96 such that none of edges of @xmath304 is adjacent to vertex @xmath60 .
this implies that three edges of @xmath304 must be adjacent to vertices of @xmath82 with degree 4 or @xmath303 .
then @xmath305 and @xmath306 , yielding @xmath307 .
since @xmath308 still has a vertex of degree 4 in @xmath96 , it is not homeomorphic to @xmath62 .
now we may say that @xmath96 has three vertices of degree 4 and a vertex @xmath169 of degree 3 which is indeed @xmath303 .
let @xmath208 be the unique vertex of degree 4 in @xmath67 excluding @xmath96 .
now we divide into three cases .
if @xmath309 contains @xmath60 and @xmath169 as in figure [ fig13](a ) , then @xmath96 contains a vertex @xmath23 such that three edges of @xmath310 are adjacent to @xmath107 .
therefore @xmath311 , but @xmath312 still has vertex @xmath208 of degree 4 . if @xmath309 contains only @xmath60 ( or similarly only @xmath169 ) as in figure [ fig13](b ) , then @xmath313 so that @xmath314 .
but @xmath315 still has a vertex of degree 4 in @xmath82 .
finally assume that @xmath309 does not contain @xmath60 and @xmath169 as in figure [ fig13](c ) .
let @xmath316 and @xmath317 be the vertices of @xmath107 and @xmath318 excluding @xmath309 respectively .
@xmath319 must contains @xmath169 ; otherwise , @xmath319 contains the three vertices of @xmath318 , yielding @xmath320 so that @xmath321 .
but @xmath322 still has vertex @xmath208 of degree 4 .
let @xmath323 and @xmath324 be the other vertices of @xmath107 . since @xmath5 is a triangle - free graph , one edge of @xmath325 and one edge of @xmath326
are adjacent to @xmath169 , contradicting the fact that @xmath169 has degree 3 .
cousin 94 of the @xmath2 family appears in this case . here
@xmath328 and there are three extra edges
. therefore @xmath74|=20 $ ] and @xmath73 = [ 4,4,4,4,4]$ ] or @xmath329 $ ] . in the former case , four vertices of @xmath96 have degree 4 .
let @xmath303 be the last vertex of @xmath67 excluding @xmath96 .
then three extra edges must connect @xmath303 with three different vertices , say @xmath208 , @xmath209 and @xmath210 , in @xmath96 respectively .
let @xmath330 be the non - extra edge of @xmath331 and @xmath60 be the other vertex that @xmath330 is adjacent to .
if @xmath60 has degree 3 , then @xmath332 so that @xmath333 . but @xmath334 still has vertex @xmath30 of degree 4 .
if @xmath60 has degree 4 , then the remaining two edges of @xmath106 must be adjacent to four vertices of @xmath96 .
but they can not be adjacent to @xmath208 , @xmath209 and @xmath210 to avoid producing triangles in @xmath5 .
so they are indeed adjacent to the same vertex @xmath211 which is the remaining vertex in @xmath96 , producing a bigon which is not allowed in @xmath5 .
for the latter case @xmath329 $ ] , obviously @xmath258 .
let @xmath316 , @xmath323 , @xmath324 and @xmath335 be two vertices of degree 3 and the other two vertices of degree 4 of @xmath82 . and
similarly we define @xmath317 , @xmath336 , @xmath337 and @xmath338 , and let @xmath303 and @xmath339 denote the remaining vertices with degree 3 of @xmath67 excluding @xmath96 .
see figure [ fig14](a ) .
first we will show that @xmath319 ( or @xmath340 ) contains both @xmath337 and @xmath338 , and by the same reason @xmath341 ( or @xmath342 ) contains both @xmath324 and @xmath335 .
if @xmath343 contains neither @xmath316 nor @xmath323 , then it must contain three vertices among @xmath324 , @xmath335 , @xmath303 and @xmath339 .
thus @xmath344 so that @xmath345 . and if @xmath343 contains only @xmath316 ( or only @xmath323 ) , then similarly @xmath346 so that @xmath347 . but
@xmath348 still has vertex @xmath338 of degree 4 except when there is an edge connecting @xmath316 and @xmath338 as we expect .
if @xmath343 contains both @xmath316 and @xmath323 , still @xmath349 contains @xmath316 or @xmath323 or both .
this implies that @xmath319 ( or @xmath340 ) contains both @xmath337 and @xmath338 .
now consider the graph @xmath39 .
since @xmath143 , the only exceptional case we handle is when it is homeomorphic to @xmath62 . in @xmath39 , dist@xmath350 and dist@xmath351 , so that dist@xmath352 because of the property of the graph @xmath62 . without loss of generality
we may assume that @xmath39 is the graph in figure [ fig14](b ) .
now we try to find proper locations for the two vertices @xmath323 and @xmath336 in @xmath39 so that we can figure out the original graph @xmath5 by adding @xmath353 .
if @xmath323 lies on the edge connecting @xmath324 and @xmath337 ( or similarly @xmath335 and @xmath338 ) , then @xmath336 must lie between @xmath323 and @xmath324 on the same edge preventing a triangle in @xmath5 . in this case @xmath354 and @xmath355 , so that @xmath356 . but
@xmath357 contains vertex @xmath32 of degree 4 .
if @xmath323 lies on the edge connecting @xmath337 and @xmath339 ( or similarly @xmath338 and @xmath303 ) , then possible locations for @xmath336 are at the four dots shown in figure [ fig15](a ) .
for all four cases , @xmath358 but @xmath359 still contains a triangle whose vertices are @xmath30 , @xmath316 and @xmath323 .
if @xmath323 lies on the edge connecting @xmath324 and @xmath303 ( or similarly @xmath335 and @xmath339 ) , then @xmath336 must lie on the edge connecting @xmath323 and @xmath324 .
then similarly @xmath347 but @xmath348 contains a triangle whose vertices are @xmath30 , @xmath317 and @xmath336 .
finally we only need to consider the remaining case that both @xmath323 and @xmath336 lie on the edge connecting @xmath303 and @xmath339 as in figure [ fig15](b ) , although there is the possibility of switching their locations .
cousin 94 of the @xmath2 family appears in either case . in this case
@xmath96 has at most one vertex of degree 3 ; otherwise , @xmath97 and @xmath98 , yielding @xmath63 . since @xmath360 and there are four extra edges , @xmath74|=21 $ ] and @xmath73 = [ 4,4,4,3,3,3]$ ] or @xmath361 $ ] . here
the latter case is impossible because @xmath53 has at most one vertex . in the former case @xmath96
contains three vertices of degree 4 and a vertex of degree 3 .
let @xmath303 and @xmath339 be the vertices of degree 3 in @xmath67 excluding @xmath96 .
there are at least two edges among the four extra edges each of which connects a vertex of @xmath318 and one of @xmath303 or @xmath339 .
let @xmath316 and @xmath323 be the vertices of @xmath318 associated to two extra edges . if @xmath362 , then @xmath363 so that @xmath364 .
thus @xmath316 and @xmath323 are different and let @xmath324 be the other vertex of @xmath318 . in this case
@xmath365 so that @xmath366 . except for the case in figure [ fig16 ]
, @xmath367 has a vertex of degree 4 which is @xmath323 or @xmath324 . for the exceptional case , similarly @xmath368 but @xmath369 contains a triangle whose vertices are @xmath32 , @xmath316 and @xmath324 .
thus it is not homeomorphic to @xmath62 . in this case
@xmath53 must be the empty set ; otherwise , @xmath98 so that @xmath63 . since @xmath372 and there are five extra edges , @xmath74|=22 $ ] and @xmath73 = [ 4,4,4,4,3,3]$ ] or @xmath373 $ ] .
the latter case is impossible because the set @xmath53 is empty . in the former case
all four vertices of @xmath96 except @xmath30 have degree 4 .
let @xmath303 and @xmath339 be the vertices of degree 3 in @xmath67 which are obviously not in @xmath96 .
there is a vertex @xmath60 in @xmath96 so that @xmath106 has an extra edge connecting @xmath60 and one of @xmath303 or @xmath339 .
then @xmath374 so that @xmath103 . aaaaaa j. barsotti and t. w. mattman , _ intrinsically knotted graphs with 21 edges _ , preprint arxiv:1303.6911 .
p. blain , g. bowlin , t. fleming , j. foisy , j. hendricks , and j. lacombe , _ some results on intrinsically knotted graphs _ , j. knot theory ramif . * 16 * ( 2007 ) 749760 .
j. conway and c. mca .
gordon , _ knots and links in spatial graphs _ ,
j. graph theory * 7 * ( 1985 ) 445453 .
e. flapan and r. naimi , _ the @xmath12 moves does not preserve intrinsic knottedness _ , osaka j. math .
* 45 * ( 2008 ) 107111 .
j. foisy , _ intrinsically knotted graphs _ , j. graph theory * 39 * ( 2002 ) 178187 .
n. goldberg , t. mattman , and r. naimi , _ many , many more intrinsically knotted graphs _ , preprint arxiv:1109.1632v1 .
r. hanaki , r. nikkuni , k. taniyama , and a. yamazaki , _ on intrinsically knotted or completely @xmath95-linked graphs _ , pacific j. math .
* 252 * ( 2011 ) 407425 .
b. johnson , m. kidwell , and t. michael , _ intrinsically knotted graphs have at least @xmath375 edges _ , j. knot theory ramif .
* 19 * ( 2010 ) 14231429 .
h. kim , t. mattman , and s. oh , _
bipartite intrinsically knotted graphs with 22 edges _
, preprint .
t. kohara and s. suzuki , _ some remarks on knots and links in spaital graphs _ , knots 90 ( osaka , 1990 ) ( 1992 ) 435445 .
m. lee , h. kim , h. j. lee , and s. oh , _
exactly fourteen intrinsically knotted graphs have 21 edges _ , preprint arxiv:1207.7157 .
r. motwani , a. raghunathan , and h. saran , _ constructive results from graph minors ; linkless embeddings _ , proc .
29th annual symposium on foundations of computer science , ieee ( 1988 ) 398409 .
m. ozawa and y. tsutsumi , _ primitive spatial graphs and graph minors _ , rev . mat .
* 20 * ( 2007 ) 391406 .
n. robertson and p. seymour , _ graph minors xx , wagner s conjecture _ , j. combin .
theory ser .
b * 92 * ( 2004 ) 325357 .
h. sachs , _ on spatial representations of finite graphs _ , colloq .
janos bolyai * 37 * ( 1984 ) 649662 . | a graph is called intrinsically knotted if every embedding of the graph contains a knotted cycle .
johnson , kidwell and michael showed that intrinsically knotted graphs have at least 21 edges .
recently lee , kim , lee and oh , and , independently , barsotti and mattman , showed that @xmath0 and the 13 graphs obtained from @xmath0 by @xmath1 moves are the only intrinsically knotted graphs with 21 edges . in this paper
we present the following results : there are exactly three triangle - free intrinsically knotted graphs with 22 edges having at least two vertices of degree 5 .
two are the cousins 94 and 110 of the @xmath2 family and the third is a previously unknown graph named @xmath3 .
these graphs are shown in figure [ fig3 ] and [ fig4 ] .
furthermore , there is no triangle - free intrinsically knotted graph with 22 edges that has a vertex with degree larger than 5 . |
in recent years there had been considerable activity , both experimental and theoretical , in the field of critical phenomena of antiferromagnets with heisenberg and rkky interactions @xcite -@xcite .
this increase of interest @xcite-@xcite was stimulated by new experiments in high @xmath0 superconductors and heavy fermion compounds .
in particular , the critical spin fluctuations are supposed to be responsible for non - fermi - liquid low temperature behavior of the heat capacity and resistivity in @xmath1 and @xmath2 @xcite,@xcite in the vicinity of the quantum critical point . besides , the fermi - type resonating valence bond excitations were introduced by @xcite,@xcite in the framework of the heisenberg model to describe unusual magnetic properties of high @xmath3 cuprates @xcite and ce - based heavy fermion compounds @xcite,@xcite . in this case
the critical spin fluctuations were shown to be important in the spin liquid formation mechanism .
however , the scaling behavior of kinetic coefficients in the presence of spin - liquid fluctuations can be differ significantly from the predictions of the dynamic scaling theory @xcite .
in this paper we present a microscopic description of the scaling behavior of spin diffusion and relaxation coefficients in isotropic heisenberg antiferromagnet in the critical region above the neel temperature .
the scaling dimension of kinetic coefficients was predicted by halperin and hohenberg @xcite-@xcite .
they proposed the dynamic scaling hypothesis based on the assumption that the critical exponents of kinetic coefficients are the same on both sides of the phase transition point .
the microscopic treatment of the spin diffusion in the paramagnetic phase of ferromagnet was proposed by s.v.maleev @xcite - @xcite . within the framework of this approach
the approximations which are necessary for validity of the dynamic scaling hypothesis were established and momentum and frequency dependence of the spin diffusion coefficient was analyzed .
we will follow maleev s approach to derive the scaling behavior of kinetic coefficients for heisenberg antiferromagnet . the qualitative picture of the critical phenomena expressed in terms of wave number @xmath4 and inverse coherence length @xmath5 is well known @xcite-@xcite .
there are two regions : `` hydrodynamic '' regime deals with the longitudinal fluctuations @xmath6 of the order parameter - the staggered magnetization @xmath7 and `` critical '' one considers the fluctuations with @xmath8 ( the momentum @xmath9 corresponds to small deviation of the momentum from the af vector ) .
however , there is an additional conserved quantity in afm which is the total magnetization @xmath10 .
the fluctuations of this parameter do not lead to any singularities in spin correlators .
nevertheless , there is a region of longitudinal fluctuations of @xmath11 @xmath12 which is also can be called as `` hydrodynamic '' .
the goal of our work is to investigate the spin correlation functions behavior in the paramagnetic phase of af and to establish the relations between the kinetic coefficients in two fluctuation regions of the phase diagram .
the dynamic of magnetization fluctuations is described by the van hove microscopic diffusion equation in the long wave limit @xmath13 where @xmath14 is the spin diffusion coefficient .
this equation is derived from the conservation low for the total magnetic moment , since the operator @xmath15 commutes with the hamiltonian .
the fluctuation behavior of nonconserved order parameter is radically different .
according to the general postulates of the thermodynamics of weakly nonequilibrium processes @xcite the rate of change @xmath16 is proportional to the conjugate thermodynamical force @xmath17 here @xmath18 - is the static susceptibility , and the kinetic coefficient @xmath19 satisfies the condition @xmath20 .
the relaxation function determined by the equation ( [ 0.1 ] ) is nearly uniform , therefore the gradient terms for relaxation processes are omitted .
in contrast to relaxation the first nonzero contribution to the spin diffusion equation is proportional to @xmath21 .
although the average value of @xmath15 is equal to zero on the both sides of the phase transition , the fluctuations of total magnetization around zero take place .
however , unlike the fm case the diffusion mode is not the `` critical '' mode for af transition .
thus , we are interested in the dynamical susceptibility of a cubic heisenberg antiferromagnet in zero magnetic field above the neel temperature , @xmath22 and the dipolar interaction @xcite is neglected . as is known the susceptibility
is related to the retarded green s function by the following equality : @xmath23 where @xmath24 is lande splitting factor , @xmath25 is bohr magneton and @xmath26 > , \label{0.4}\ ] ] @xmath27 @xmath28 using the equations ( [ 0.1 ] , [ 0.2 ] ) one can obtain an expression for the retarded green functions in the fluctuation region in the following form : @xmath29 @xmath30 here @xmath18 is the static spin correlation function . the equations ( [ 0.5 ] , [ 0.6 ] )
correspond to diffusion and relaxation regimes respectively . when @xmath31 ( @xmath32 is the ginzburg number which characterizes the conditions when the landau theory is valid ) the fluctuations become important .
the dynamic properties in this region can be described by halperin - hohenberg scaling hypothesis . according to this theory the equation for the dynamic susceptibility
can be expressed in terms of the scaling function @xmath33 : @xmath34 thus , the dynamic critical exponent @xmath35 , which characterizes the critical fluctuations energy scale @xmath36 , is connected with the static critical exponent @xmath37 defined by the expression @xmath38 .
we assume the static behavior of susceptibility in the form @xmath39 .
we also neglect the fisher index @xmath40 which determines the `` anomalous dimension '' @xcite .
this approximation is valid in 3d case @xcite . to derive the scaling properties of af one have to introduce two types of the scaling functions : @xmath41 and @xmath42 @xmath43 in principal , expressions ( [ 0.7 ] , [ 0.8 ] ) determine the kinetic coefficients @xmath14 and @xmath19 in terms of coherence length .
moreover , the renormalization group analyses @xcite,@xcite has shown that the kinetic coefficients are singular in the fluctuation region of af . the theory under consideration is a variant of the mode - mode coupling theory proposed by kawasaki @xcite ( see also @xcite-@xcite ) . we extend to af systems the dynamic scaling method offered by maleev for ferromagnets . as was mentioned above , the aim of our paper is to derive the scaling behavior and momentum and frequency dependence of @xmath44 functions in the fluctuation region of af .
assumptions which are crucial for the microscopic substantiations of dynamic scaling hypothesis will be also formulated .
we consider the dynamic susceptibility of cubic heisenberg antiferromagnet in the fluctuation region above the neel temperature . the equations ( [ 0.5 ] , [ 0.6 ] )
can be rewritten in more general form : @xmath45 there is a simple expression for the spin - diffusion coefficient in the hydrodynamical regime @xmath46 the generalized kinetic coefficient in the relaxation region is @xmath47 .
the different limits @xmath48 in the expressions ( [ 0.5 ] , [ 0.6 ] ) significantly depend on the relations between @xmath4 and @xmath49 .
these limits are completely analogous to those considered in the fermi liquid theory @xcite . in the treatment
below we consider the the quasistatic limit of @xmath50 , @xmath51 .
as it was shown by maleev , it is possible to determine the kinetic coefficients in terms of the kubo functions for the operators @xmath52 and @xmath53 , where @xmath54 , @xmath55 and @xmath56,\ ] ] @xmath57>.\ ] ] the formal expression ( [ 1.3 ] ) is exact , and it takes into account the nonlinear character of the relaxation forces . in the case of purely exchange interaction , @xmath58 and @xmath59 . the kubo function in the denominator vanishes at @xmath60 , and we arrive at the expression @xmath61 .
however , in studying the dispersion of the kinetic coefficients , i.e. its dependence on @xmath4 , @xmath49 , we can not , generally speaking , neglect this function in the denominator . using the properties of spin operators
, it is not difficult to obtain the relations between different retarded spin green s functions @xmath62 , @xmath63 and @xmath64 in the paramagnetic phase .
these relations are result of dispersion relations @xcite :
@xmath65 @xmath66 \label{1.4}\ ] ] in particular , it is seen from equation ( [ 1.4 ] ) that the function @xmath67 possesses the same symmetry properties as spin - spin correlation function . the spin current operators taken in the form @xmath68 can be connected with the kubo functions for matsubara frequencies @xmath69 with the use of equation of motion : @xmath70 \epsilon_{\alpha\beta\gamma}s^\beta_{{\bf p}+{\bf k } } s^\gamma_{-{\bf p } } \label{1.5}\ ] ] the spin - current correlators can be calculated diagrammatically and then the explicit equations ( [ 1.1]-[1.3 ] ) for the kinetic coefficients can be obtained . when deriving eq .
( [ 1.6 ] ) , we have confined ourselves to the lowest terms of the expansion in powers of @xmath71 and have put @xmath72 , where @xmath73 . since , we are interested in the correction terms of order @xmath74 and @xmath75 ; we consistently neglect the @xmath76 and @xmath77 corrections , because @xmath78 .
thus the problem of derivation of the kinetic coefficients is reduced to the analysis of four - spin correlators with current vertices .
the later problem can be solved by analytical continuation of the temperature feynman s diagram to the upper half - plane of complex variable @xmath49 the feynman diagrams corresponding to the spin - current correlation function are shown on figure 1 .
the `` bare '' poles of the green s functions ( [ 0.5 ] , [ 0.6 ] ) lie on imaginary axis . introducing fictitious quasi - particles `` diffusons '' and `` relaxons ''
we obtain the self - consistent equations for kinetic coefficients and determine their scaling dimensions .
the static susceptibility in the fluctuation region obeys the ornstein - zernike law @xmath79 ( here a is a constant ( @xmath80 ) , @xmath81 ) .
there is no singularities of static susceptibility in the diffusion region @xmath12 , where @xmath82 . in sec .
iii the relations between kinetic coefficients are established and scaling dimensions of kinetic coefficient are analyzed .
iv is devoted to a momentum and frequency dependence of spin diffusion and spin relaxation constants .
we begin with analyzing the set of diagrams which can not be separated into two parts by cutting one line of interaction , and include these diagrams into the irreducible self - energy part of the spin - current correlation function . using the definition of @xmath83 and properties of the functions
@xmath84 we obtain an expression for generalized kinetic coefficient in terms of irreducible self - energy parts : @xmath85 [ 1+g_0^{-1}\frac{{\cal r}^r_{\dot{s}s}({\bf k},\omega)\gamma({\bf k},\omega ) } { i\omega(-i\omega + g_0^{-1}(k)\gamma({\bf k},\omega))}]^{-1 } \label{2.1}\ ] ] the equation ( [ 2.1 ] ) can be derived both from analysis of diagrammatic series for the spin - current correlators @xcite , and directly from larkin equation @xcite,@xcite .
below we use the following notations for retarded spin green s functions : @xmath86 , and @xmath87 - are the irreducible self - energies . to obtain the graphical representation for the irreducible part @xmath88 one have to replace the
`` dressed '' vertex in fig.1 by the irreducible one .
the estimation of @xmath89 in the framework of mean - field theory @xcite , @xcite results in : @xmath90 moreover , due to analytical properties of spin correlators , @xmath91 .
we assume that the expression for @xmath89 contains the small parameter @xmath92 in a limit of small @xmath49 not only in critical region and neglect this contribution below .
therefore , the generalized kinetic coefficient @xmath83 can be rewritten in terms of irreducible self - energies only : @xmath93 the general set of diagrams for the irreducible self - energy part @xmath94 in imaginary time can be classified by the number of intermediate states .
first , we confine ourselves to the diagrams with two - particles intermediate states ( fig.2 ) @xmath95 @xmath96 replacing the sum over momentum @xmath97 by the integration one should use the `` cutoff '' @xmath98 as an upper limit of integration .
this means that the integrations are restricted by the region close to the appropriate singularities ( small @xmath97 and small @xmath99 for @xmath100 in the vicinity of af vector @xmath101 ) . from the point of view of the energy dependence , the vertices @xmath102 being the three - point functions are the analytical functions of all three frequencies @xmath49 and , in each of the frequencies they have a cut along the real axis ; they have no singularities in the physical region of variables @xmath49 @xcite . due to these properties it is possible to separate the vertex into a static part and dynamic contribution , disappearing in the limit @xmath103 @xcite .
we concentrate below on the static vertex . as it can be seen from the fig.2(a )
, the static vertices correspond to the longitudinal processes of `` diffuson '' - `` diffuson '' or `` relaxon''- relaxon " pair creation and annihilation , since in this case only the modes of the same origin couple together . on the other hand , the static green function @xmath104 is not sensitive to the direction of the momentum . in the other words ,
the corresponding lines in the diagrams are not directed .
for this reason , the processes of diffuson and relaxon scattering contain the same vertices as the processes of pair creation and annihilation .
however , for such a scattering processes in the limit @xmath105 the ward identity holds @xcite,@xcite ( fig.3 ) : @xmath106 as was mentioned , the region of integration in the second term is concentrated near the points @xmath107 . the contribution from the staggered magnetization fluctuations into the spin diffusion can be calculated by means of the substitution @xmath108 and , by use of the property @xmath109 .
let us consider the diagram fig.2(b ) .
the external momentum of this diagram can be taken equal to @xmath110 .
therefore , we have two different coupling modes , `` diffuson '' and `` relaxon '' .
thus , this diagram corresponds to the process of `` diffuson '' - `` relaxon '' pair creation . for this reason we can not use the ward identity to analyze this contribution
however , the `` bare '' vertex ( fig.3 ) has the scaling dimension @xmath111 besides , in the magnetic brillouin zone the points 0 and @xmath112 are equivalent .
taking into account the direction independence of interacting modes in fluctuating region one can assume that the processes of multiple scattering on the static fields do not influence the scaling dimension of the static vertex .
using the properties of vertices it is not difficult to perform the summation over @xmath113 and the analytical continuation in @xmath49 @xcite . as a result
we obtain the following expressions for kinetic coefficients @xmath114 \label{2.5}\ ] ] @xmath115 \label{2.6}\ ] ] there index @xmath116 stands for the processes with only two - particle intermediate states taken into account .
the equations generalize the corresponding equation ( [ 2.6 ] ) are obtained by maleev @xcite for ferromagnets . in that case
it is possible to retain only the first term in expression ( [ 2.5 ] ) , since one deals with a single mode regime
. equations ( [ 2.5],[2.6 ] ) can be rewritten in a different form .
taking @xmath117 and @xmath118 and integrating by parts over @xmath113 one can obtain the following result : @xmath119 \label{2.7}\ ] ] @xmath120 these expressions are in fact a generalization of equations derived by maleev @xcite for the case of two coupling modes .
as it was mentioned above , the main contribution to the integrals ( [ 2.5],[2.6 ] ) appears from the vicinity of scaling functions singularities ( [ 0.8 ] ) .
besides , the characteristic energies of critical fluctuations satisfy the condition @xmath121 due to the `` critical slowing down '' .
thus , we can replace the hyperbolic functions ( [ 2.5 ] -[2.7 ] ) by their arguments . calculating integrals over the frequencies and extracting the scaling dimension of kinetic coefficients we obtain the relation between spin diffusion coefficient and the relaxation constant .
@xmath122 it is interesting to note that the expressions ( [ 2.9 ] ) can be derived by direct substitution of retarded green functions ( [ 0.5],[0.6 ] ) to the expressions ( [ 2.5],[2.7 ] ) . after integration over the frequencies we obtain the equations on kinetic coefficients which contain only integration over momentum of the different powers of the static correlator @xmath18 . as a result
the first term is determined by the two - diffuson intermediate state and the second term corresponds to two - relaxon intermediate state .
the integrals ( [ 2.6 ] - [ 2.8 ] ) can be calculated in the same way .
the equation which determines the relaxation constant in terms of spin diffusion coefficient can be written in following form : @xmath123 the coefficients @xmath124 in eqs .
( [ 2.9],[2.10 ] ) depend on the shape of a scaling function and , strictly speaking , can not be calculated in the framework of method used .
solution of self - consistent system of _ algebraic _ equations leads to following scaling dimension of kinetic coefficient ] @xmath125 behavior of kinetic coefficient agrees with predictions of dynamic scaling theory @xcite-@xcite and with renormalization group analysis @xcite@xcite .
therefore , the kinetic coefficients of af are singular in fluctuating region , and spin diffusion is determined by the intermediate relaxation processes . a correction in ( [ 2.7 ] )
corresponding to the influence of self - diffusion is proportional to @xmath126 .
hence , these processes can be omitted in the critical region .
calculation of dynamic critical exponent z ( see eq.[0.7 ] ) results in @xmath127 .
the simple physical assumptions explaining the spin diffusion and relaxation in the fluctuation region are based on the following argument : in the vicinity of the phase transition point @xmath128 the regions of short range order are formed .
the characteristic length scale of the ordered regions is @xmath129 .
the excitations in these regions are the af magnons with sound like dispersion law .
estimations of the spin diffusion coefficient results in @xmath130 , where @xmath131 is a characteristic diffusion time and @xmath132 is a `` sound '' velocity @xcite , therefore @xmath133 . due to the dynamic scaling hypothesis which postulates the invariance of the dynamic scaling exponent @xmath35 ,
the characteristic scaling behavior of relaxation constant is @xmath134 .
although the kinetic coefficients are singular , the characteristic relaxation times of staggered magnetization and spin diffusion become infinite .
this result guarantees the existence of macroscopic ordered regions , corresponding to the partial equilibrium of the system @xcite .
it should be emphasized that we did not use any assumptions about the elementary excitations spectrum in ordered phase when deriving the expressions ( [ 2.9 ] ) .
our study is based only on a concept of total magnetization conservation and order parameter non - conservation .
here we study the kinetic coefficient as a functions of finite but small @xmath135 and @xmath49 . for this sake
we use the relations between spin correlation functions ( [ 1.3 ] ) and kubo functions ( [ 2.1 ] ) .
it can be seen that the corrections due to momentum and frequency dependence of kinetic coefficient are determined both by the nonlinear character of the relaxation forces and the momentum and frequency dependence of irreducible self - energy parts . according to the estimation ( [ 2.1a ] )
one can derive the frequency and momentum dependence of spin diffusion and relaxation constants only in terms of linear response theory .
therefore , we can neglect the nonlinear character of relaxation forces . first , we consider the static renormalizations of kinetic coefficients .
it is not difficult to obtain the usual expansions for spin diffusion and relaxation constants in series of @xmath136 , @xmath137 : @xmath138\ ] ] @xmath139\ ] ] as is known @xcite , these expansion can be done due to the singularities of static spin correlation functions in points @xmath140 , where @xmath141 are integer .
the coefficients @xmath142 , @xmath143 are determined by the shape of static correlation function .
we analyze the energy dependence of kinetic coefficients . using equation ( [ 2.2 ] )
we obtain the following expressions for the real and imaginary parts of the generalized coefficient @xmath144 : @xmath145 since @xmath146 is the odd function of @xmath49 and @xmath147 is the even function of @xmath49 the first nonzero term in the energy expansion of kinetic coefficients is proportional to @xmath148 .
let us introduce the effective generalized kinetic coefficient @xmath149 by the definition : @xmath150 the expression for the coefficient @xmath149 is completely analogous to the definition of effective mass in the theory of quantum liquids .
the renormalization constant @xmath151 is similar to the fermi - liquid @xmath151 factor . @xmath152 calculating @xmath151 in fluctuation region results in the following expressions for renormalization constant : @xmath153 where the coefficients @xmath154 also can be expressed in terms of the integrals of different powers of static correlator @xmath18 . using the definition ( [ 3.2 ] ) for small , but nonzero frequencies
one can obtain the expansions of _ real _ spin diffusion coefficient @xmath155 and relaxation constant @xmath156 : is characteristic fluctuations energy , @xmath127 ] @xmath158\ ] ] @xmath159 \label{3.5}\ ] ] it should be noted that we do not intend to describe the kinetic coefficients properties in the region @xmath160 , @xmath161 .
the diffusion and relaxation picture in this region becomes very complicated and scarcely amenable to detailed analysis at the present time .
for this reason we shall neglect the irregular corrections to kinetic coefficients resulting from the generation of poles chain and drift of the poles toward the new cut @xcite,@xcite . in the region of @xmath135 , @xmath49 under consideration , all these singularities are small corrections , and therefore , the phenomena associated with them are no special interest .
now we turn to discussion of the role of many - particles @xmath162 intermediate states .
as it was mentioned above , we are interested only in regular contributions of such diagrams to the kinetic coefficients : @xmath163 @xmath164 here the functions @xmath84 correspond both to `` diffusons '' and `` relaxons '' .
the integration over frequency is still restricted by the regions close to singularities of scaling function . in the case of @xmath165 the expression ( [ 3.4 ] ) simply transforms into ( [ 2.7],[2.8 ] ) .
the vertices @xmath166 are symmetric functions of the momenta of intermediate particles . for @xmath117
the generalized ward identities similar to thous of ref .
@xcite hold for @xmath166 ; therefore , @xmath166 can be expressed as a sum of derivatives of the ordinary n - particle vertices of static scaling theory .
a `` dimensional '' estimate exists for the latter @xcite : @xmath167 in the limit @xmath50 . as a result ,
the diagrams with m - particle intermediate states in the channel of `` diffusons '' and `` relaxons '' creation do not change the scaling dimension of irreducible self - energies . as to the vertices behavior around the af vector @xmath112
, there is a direct mapping of the properties of the diagrams with two - particle intermediate states onto those with many - particle intermediate states .
therefore , we can confine ourselves only by the diagrams with two - particle intermediate states .
another diagrams only renormalize the numerical constants , and the latter can not be determined in a framework of our method .
the energy dependence of vertices also can not change the scaling behavior of kinetic coefficients @xcite .
we investigated the scaling behavior of generalized kinetic coefficients in 3d isotropic heisenberg antiferromagnets . in a framework of analysis based on the modified mode - mode coupling theory
the critical exponents are derived .
necessary approximations in microscopic approach to satisfy the dynamic scaling hypothesis are formulated . in particular , it is shown that the main sequence of feynman diagrams contain the processes with two - particles intermediate states .
besides , the vertices of static scaling theory are proved to give the main contribution to quasiparticles scattering processes . the momentum and frequency dependence of spin diffusion and relaxation coefficients is found in the `` pole '' approximation .
a definition of the generalized effective coefficient is introduced in a full analogy with that used in the theory of quantum liquids for a definition of effective mass .
taking into account the renormalization processes responsible for multiple scattering of `` diffusons '' and `` relaxons '' we derived the asymptotics of scaling function in the region of small frequencies and momenta , @xmath168 , @xmath169 .
our investigation of the critical dynamic is based on the static and dynamic scaling hypotheses .
we took into account only two coupling modes .
one of them is considered to be responsible for fluctuations of the total magnetization and another deals with the fluctuations of order parameter - staggered magnetization .
existence of two modes is determined by the conservation s laws of the heisenberg model .
for this reason , our results depend only weakly on the specific properties of af and we expect that all they are valid for any system with nonconserved order parameter when the additional conserved quantity exists
. the scaling behavior of more complicated systems , e.g. ce - based heavy fermion compounds with almost integer valence , can be significantly different from critical dynamic of heisenberg magnets . in these systems an additional coupling modes interacting with paramagnons can appear due to , e.g. , the spin liquid correlations , proposed in ref.@xcite .
an additional modes in kondo lattices can be also connected with itinerant - magnetic excitations .
kinetic coefficients of such systems can be measured by the methods of of inelastic neutron scattering to verify a spin liquid correlations existence .
the methods developed in this paper can be useful for the systems close to the quantum critical point @xcite,@xcite @xcite when the temperatures of af ordering are close to zero , or , even negative .
our method also should be valid for anisotropy magnets , antiferrimagnets and systems with dipolar interactions .
we think that the diagrammatic approach for investigation of kinetic coefficient in the vicinity of neel point has some advantages in comparison with other methods @xcite-@xcite , @xcite . this approach can be applied to the unrenormalizible hamiltonians and to systems with nonlocal coupling modes @xcite .
we would like to thank d.n.aristov , yu.kagan , a.v.lazuta , s.v.maleev and v.l.pokrovskii for valuable discussions .
this work was supported by russian foundation for basic research ( 95 - 02 - 04250a ) , international association intas ( 93 - 2834 ) and dutch organization for basic research nwo ( 07 - 30 - 002 ) .
99 s.chakravarty , b.i.halperin , and d.r.nelson , phys.rev * b39 * , 2344 ( 1989 ) d.p.arovas and a.auerbach , phys.rev . * b38 * , 316 ( 1988 ) .
a. chubukov , phys .
rev . b*44 * , 12 318 ( 1991 ) .
h.monien , d.pines and c.p.slichter , phys.rev . *
b44 * , 11 120 ( 1990 ) .
k.ueda,t.moriya and y.takahashi .
_ electron properties and mechanisms of high - t@xmath170 superconductors_. north - holland , amsterdam , 1992 .
a.millis , phys.rev .
* b * 48 , 7183 ( 1993 ) .
s.kambe at al . ,
j.phys.soc of japan * 65 * , 3294 ( 1996 ) .
a.rosch , a.schr@xmath171der , o.stockert and h.v.l@xmath171hneysen .
phys.rev . lett.*79 * , 159 ( 1997 ) , o.stockert , h.v .
l@xmath171hneysen , a.rosch , n.pyka and m.loewenhaupt .
cond - mat/9802086 .
anderson , mater .
, * 8 * , 153 ( 1973 ) .
g. baskaran , z. zou , and p. w. anderson , solid state comm .
* 63 * , 973 .
kagan , k.a .
kikoin , and n.v .
prokofev , physica b * 182 * , 201 ( 1992 ) k.a.kikoin , m.n.kiselev , a.s.mishchenko , sov . jetp letters * 60 * , 600 ( 1994 ) .
physica * b 206&207 * , 129 ( 1995 ) k.a.kikoin , m.n.kiselev , a.s.mishchenko , sov.phys.jetp .
* 111 * , 729 ( 1997 ) .
halperin , p.c .
hohenberg , phys.rev . *
177 * , 952 ( 1969 ) , phys.rev .
* 188 * , 898 ( 1969),rev . mod .
* 49 * , 435 ( 1977 ) .
halperin , p.c .
hohenberg , and e.d.siggia , phys.rev . *
b13 * , 1299 ( 1976 ) .
s.v.maleev , sov.phys.jetp * 38 * , 613 ( 1974 ) .
s.v.maleev , sov.phys.jetp * 39 * , 889 ( 1974 ) , sov.physics rev . a*8 * , 323 ( 1987 ) .
a.z.patashinskii , v.l.pokrovskii .
_ fluctuation theory of phase transition_. pergamon press , 1979 r.freedman and g.f.mazenko , phys.rev.lett * 34 * , 1571 , ( 1975 ) .
k.kawasaki , in _ phase transitions and critical phenomena _ , ed .
c.domb and m.s.green , academic new york , vol 5a , 1976 .
s.w.lovesey , e.engahl , a.cuccoli and v.tognetti .
* 6 * , 7099 ( 1994 ) , j.phys.cond.matt .
* 6 * , 7553 ( 1994 ) e.frey,f.schvabl and s.thoma phys.rev .
* b 40 * , 7199 ( 1989 ) s.w.lovesey , e.engahl , a.cuccoli and v.tognetti . j.phys.cond.matt .
* 7 * , 2615 ( 1995 ) a.a.abrikosov , l.p .
gorkov , i.e. dzyaloshinski .
_ methods of quantum field theory in statistical physics .
prentice - hall_. new jersey 1963 .
l.kadanoff and p.martin , ann .
* 24*,419 ( 1963 ) .
a. izyumov , yu.n .
_ statistical mechanics of magnetically ordered systems_. plenum , new york , 1988 .
v.g.vaks , a.i.larkin , s.a.pikin . sov.phys.jetp
* 24 * , 240 ( 1967 ) , sov.phys.jetp * 26 * , 188 ( 1968 ) . g.m .
eliashberg sov.phys.jetp * 14 * , 886 ( 1962 ) , sov.phys.jetp * 15 * , 1151 ( 1962 ) , s.v.maleev , sov.theor.mat.phys . * 4 * , 86 ( 1970 ) .
a.a.migdal , sov.phys.jetp * 28 * , 1036 ( 1969 ) , a.m.polyakov , sov.phys.jetp * 30 * , 151 ( 1970 ) .
e.m.lifshits , l.p.pitaevskii , _ physical kinetics_. pergamon press , 1981 .
j.hertz , phys.rev .
* b * 14 , 1165 ( 1975 ) , t.moriya . _
spin fluctuations in itinerant electron magnetism _
( springer , berlin , 1985 ) | a theory is proposed for kinetic effects in isotropic heisenberg antiferromagnets at temperatures above the neel point . a metod based on the analysis of a set of feynman diagrams for the kinetic coefficients
is developed for studying the critical dynamics .
the scaling behavior of the generalized coefficient of spin diffusion and relaxation constant in the paramagnetic phase is studied in terms of the approximation of coupling modes .
it is shown that the kinetic coefficients in an antiferromagnetic system are singular in the fluctuation region .
the corresponding critical indices for diffusion and relaxation processes are calculated .
the scaling dimensionality of the kinetic coefficients agrees with the predictions of dynamic scaling theory and a renormalization group analysis .
the proposed theory can be used to study the momentum and frequency dependence of the kinetic parameters , and to determine the form of the scaling functions .
the role of nonlocal correlations and spin - liquid effects in magnetic systems is briefly discussed .
+ + pacs nos:75.40.gb , 75.50.ee + |
analysis on metric spaces with no a priory smooth structure has been rapidly developed resently .
this development is closely related to some generalizations of the differentiability .
important examples of such generalizations and even an axiomatics of so - called `` pseudo - gradients '' can be found in @xcite and respectively in @xcite . in almost all above - mentioned books and papers
the generalized differentiations involve an induced linear structure that makes possible to use the classical differentiations in the linear normed spaces .
a new _ intrinsic _ approach to the introduction of the `` smooth '' structure by means of the construction of `` tangent spaces '' for general metric spaces was proposed by o. martio and by the first author of the present paper in @xcite . in the present paper
we prove that for every starlike set @xmath0 with a center @xmath1 all tangent spaces to @xmath2 at the point @xmath1 are isometric to the smallest closed cone which includes @xmath2 and has the vertex @xmath1 . a partial converse to this result
is also obtained .
important particular cases @xmath3 and @xmath4 are considered in details .
the results of the paper were partly published in the preprint from @xcite .
for convenience we recall the main notions from @xcite .
let @xmath5 be a metric space and let @xmath1 be a point of @xmath6 . fix a sequence @xmath7 of positive real numbers @xmath8 which tend to zero . in what follows this sequence @xmath7
is called a _ normalizing sequence_.
let us denote by @xmath9 the set of all sequences of points from @xmath6 .
[ 1:d1.1 ] two sequences @xmath10 , @xmath11 and @xmath12 , are mutually stable ( with respect to a normalizing sequence @xmath13 ) if there is a finite limit @xmath14 we shall say that a family @xmath15 is _ self - stable _
@xmath7 ) if every two @xmath16 are mutually stable .
a family @xmath17 is _ maximal self - stable _ if @xmath18 is self - stable and for an arbitrary @xmath19 either @xmath20 or there is @xmath21 such that @xmath22 and @xmath23 are not mutually stable .
a standard application of zorn s lemma leads to the following [ 1:p1.2 ] let @xmath24 be a metric space and let @xmath25 .
then for every normalizing sequence @xmath26 there exists a maximal self - stable family @xmath27 such that @xmath28 .
note that the condition @xmath29 implies the equality @xmath30 for every @xmath31 .
consider a function @xmath32 where @xmath33 is defined by .
obviously , @xmath34 is symmetric and nonnegative .
moreover , the triangle inequality for the original metric @xmath35 implies @xmath36 for all @xmath37 . hence @xmath38 is a pseudometric space .
[ 1:d1.3 ] the pretangent space to the space @xmath6 at the point @xmath1 w.r.t .
@xmath7 is the metric identification of the pseudometric space @xmath39 .
since the notion of pretangent space is basic for the present paper , we remind this metric identification construction .
define a relation @xmath40 on @xmath41 by @xmath42 if and only if @xmath43 .
then @xmath44 is an equivalence relation .
let us denote by @xmath45 the set of equivalence classes in @xmath41 under the equivalence relation @xmath40 .
it follows from general properties of pseudometric spaces , see , for example , ( * ? ? ?
* chapter 4 , th .
15 ) , that if @xmath46 is defined on @xmath47 by @xmath48 where @xmath49 and @xmath50 , then @xmath46 is the well - defined metric on @xmath47 .
the metric identification of @xmath38 is , by definition , the metric space @xmath51 .
remark that @xmath52 because the constant sequence @xmath53 belongs to @xmath54 , see proposition [ 1:p1.2 ] .
let @xmath55 be an infinite , strictly increasing sequence of natural numbers .
let us denote by @xmath56 a subsequence @xmath57 of the normalizing sequence @xmath13 and let @xmath58 for every @xmath59 .
it is clear that if @xmath22 and @xmath60 are mutually stable w.r.t .
@xmath7 , then @xmath61 and @xmath62 are mutually stable w.r.t . @xmath63 and @xmath64 if @xmath65 is a maximal self - stable ( w.r.t .
@xmath7 ) family , then , by zorn s lemma , there exists a maximal self - stable ( w.r.t .
@xmath63 ) family @xmath66 such that @xmath67 denote by @xmath68 the mapping from @xmath65 to @xmath69 with @xmath70 for all @xmath71 .
if follows from that after the metric identifications @xmath68 passes to an isometric embedding @xmath72 : @xmath73 under which the diagram @xmath74 is commutative . here
@xmath75 and @xmath76 are metric identification mappings , @xmath77 , @xmath78 .
let @xmath6 and @xmath79 be two metric spaces . recall that a map @xmath80 is called an _
isometry _ if @xmath81 is distance - preserving and onto .
[ 1:d1.4 ] a pretangent @xmath82 is tangent if @xmath83 : @xmath84 is an isometry for every @xmath63 .
note that the property to be tangent does not depend on the choice of @xmath85 in , see proposition [ 1:p1.5 ] in the present paper .
let @xmath6 be a metric space with a marked point @xmath86 @xmath87 a normalizing sequence , @xmath88 a maximal self - stable family and @xmath89 the corresponding pretangent space .
[ 8:d2.14 ] the pretangent space @xmath90 lies in a tangent space if there is a maximal self - stable family @xmath91 such that holds and if @xmath92 the metric identification of @xmath93 , is tangent .
let @xmath5 be a metric space with a marked point @xmath1 , let @xmath79 and @xmath94 be subspaces of @xmath6 such that @xmath95 and let @xmath96 be a normalizing sequence .
[ 6:d5.1 ] the subspaces @xmath79 and @xmath94 are _ tangent equivalent _ at the point @xmath1 w.r.t .
@xmath7 if for every @xmath97 and for every @xmath98 with finite limits @xmath99 there exist @xmath100 and @xmath101 such that @xmath102 we shall say that @xmath79 and @xmath94 are _ strongly tangent equivalent _ at @xmath1 if @xmath79 and @xmath94 are tangent equivalent at @xmath1 for all normalizing sequences @xmath7 .
let @xmath2 be a set in a linear topological space @xmath6 over @xmath103 .
the set @xmath2 is termed starlike with a center @xmath1 if @xmath104=\{x\in x : x = a+t(b - a),\quad t\in[0,1]\}\subseteq a\ ] ] for all @xmath105 .
moreover , @xmath2 is a _ cone _ with the _ vertex _
@xmath1 if the ray @xmath106lies in @xmath2 for every @xmath107 . for nonvoid sets @xmath108 and @xmath25 define @xmath109 as the intersection of all closed ( closed convex ) cones @xmath110 with the common vertex @xmath1 . now we are ready to formulate the first result of the paper . [ t:1.7 ]
let @xmath108 be a set with a marked point @xmath1 .
if @xmath6 is starlike with the center @xmath1 , then for each tangent space @xmath111 there is an isometry @xmath112 where @xmath113 , see , and , moreover , every pretangent space @xmath111 lies in some tangent space @xmath114 .
this theorem can be rewritten in a slightly more general form .
[ t:1.11 ] let @xmath108 be a set with a marked point @xmath1 .
suppose that @xmath6 is strongly tangent equivalent ( at the point @xmath1 ) to a starlike set with the center @xmath1 .
then all pretangent spaces to @xmath6 at the point @xmath1 lie in tangent spaces and there is a closed cone @xmath115 with a vertex @xmath116 such that for every tangent space @xmath111 there exists an isometry @xmath117 , where @xmath113 , see .
theorem [ t:1.7 ] admits a partial converse .
let @xmath118 be a ray with a vertex @xmath1 , let @xmath119 and let @xmath120 . consider the two - sided angular sector @xmath121 where , as usual , @xmath122 write @xmath123 i.e. , a positive number @xmath124 belongs to @xmath125 if and only if the sphere @xmath126 with the center @xmath1 and the radius @xmath124 and the sector @xmath127 have a nonvoid intersection . in what follows
we will use a porosity of the set @xmath125 , so recall a definition .
[ 9:d1 ] let @xmath128 and let @xmath129 .
the right - side porosity of @xmath2 at the point @xmath130 is the quantity @xmath131where @xmath132 is the length of the longest interval in @xmath133 \backslash a.$ ] [ t:1.9 ] let @xmath108 be a set with a marked point @xmath1 .
suppose that all pretangent spaces to @xmath6 at the point @xmath1 lie in tangent spaces and there is a closed cone @xmath115 with a vertex @xmath116 such that for every tangent space @xmath111 there exists an isometry @xmath134 , where @xmath113 , see .
then for every ray @xmath135 with the vertex @xmath1 we have either @xmath136 where @xmath137 is the right - side porosity of @xmath125 at the point @xmath138 .
since every convex set @xmath6 is starlike , theorem [ t:1.7 ] implies the following [ 8:t2.15 ] let @xmath79 be a convex subset of @xmath139 with a marked point @xmath1 and let @xmath87 be a normalizing sequence .
the following statements hold for every pretangent space @xmath140 .
\(i ) if the space @xmath141 is tangent , then @xmath141 and @xmath142 are isometric .
\(ii ) if @xmath141 is pretangent , then @xmath143 lies in some tangent @xmath144 . for @xmath145 or @xmath146 all pretangent spaces
@xmath111 are tangent , see section 3 of the present paper , but for an arbitrary convex @xmath108 pretangent spaces can cease to be tangent .
recall that convex set @xmath6 is termed a convex body if @xmath147 .
[ p:1.12 ] let @xmath6 be a convex body in the plane and let @xmath148 .
then for every normalizing sequence @xmath7 there is a maximal self - stable family @xmath54 such that the corresponding space @xmath90 is not tangent .
in this section we collect some results related to pretangent and tangent spaces of general metric spaces .
[ 1:p1.5 ] let @xmath6 be a metric space with a marked point @xmath1 , @xmath149 a normalizing sequence and @xmath88 a maximal self - stable family with the corresponding pretangent space @xmath89 .
the following statements are equivalent .
\(i ) @xmath150 is tangent .
\(ii ) for every subsequence @xmath151 of the sequence @xmath152 the family @xmath153 is maximal self - stable w.r.t . @xmath154 .
\(iii ) a function @xmath155 is surjective for every @xmath156 .
\(iv ) a function @xmath157 is surjective for every @xmath158 . for the proof see ( * ? ?
* proposition 1.2 ) or ( * ? ? ?
* proposition 1.5 ) .
let @xmath159 .
for a normalizing sequence @xmath7 we define a family @xmath160_y=[\tilde f]_{y,\tilde r}$ ] by the rule @xmath161_y)\leftrightarrow((\tilde y\in\tilde y)\&(\exists\,\tilde x\in\tilde f:\tilde d_{\tilde r}(\tilde x,\tilde y)=0)).\ ] ] [ 6:p5.2 ] let @xmath79 and @xmath94 be subspaces of a metric space @xmath6 and let @xmath162 be a normalizing sequence .
suppose that @xmath79 and @xmath94 are tangent equivalent ( w.r.t .
@xmath7 ) at a point @xmath95 . then following statements hold for every maximal self - stable ( in @xmath163 ) family @xmath164 . * the family @xmath165_y$ ] is maximal self - stable ( in @xmath166 ) and we have the equalities @xmath167_y]_z=\tilde z_{a,\tilde r}=[\tilde z_{a,\tilde r}]_z.\ ] ] * if @xmath168 and @xmath169 are metric identifications of @xmath164 and , respectively , of @xmath170_y$ ] , then the mapping @xmath171_y\in\omega_{a,\tilde r}^y\ ] ] is an isometry .
furthermore , if @xmath172 is tangent , then @xmath173 also is tangent .
the following lemma is a partial generalization of proposition [ 6:p5.2 ] ( i ) .
[ l:2.3 ] let @xmath94 and @xmath79 be subspaces of a metric space @xmath5 , @xmath174 , @xmath7 a normalizing sequence , @xmath175 and @xmath176 maximal self - stable families such that @xmath177_{y,\tilde r}.\ ] ] suppose @xmath79 and @xmath94 are strongly tangent equivalent at the point @xmath1 .
then the equality @xmath178_{z,\tilde r'}\ ] ] holds for every subsequence @xmath179 of the sequence @xmath162 .
let @xmath180 be a subsequence of @xmath96 .
we first note that and imply the equality @xmath181_{z,\tilde r}.\ ] ] consequently , if @xmath182 belongs to the set in the left - hand side of , then there is @xmath183 such that @xmath184 hence @xmath185 the last equality means that @xmath186 belongs to the set in the right - hand side of .
conversely , if @xmath187_{z,\tilde r'},\ ] ] then holds with some @xmath188 .
hence , by , there is @xmath189 such that @xmath190 let us define @xmath191 by the rule @xmath192 limit relations and imply that @xmath193 .
moreover , by the second equality in , we have @xmath194 .
consequently , by , we have @xmath195 [ p:2.4 ] let @xmath79 and @xmath94 be subspaces of a metric space @xmath6 and let @xmath1 be a point in @xmath196 .
suppose that @xmath79 and @xmath94 are strongly tangent equivalent at the point @xmath1 and that each pretangent space @xmath140 lies in some tangent space @xmath197 .
then each pretangent space @xmath172 lies in some tangent space @xmath198 .
let @xmath172 be a pretangent space to @xmath94 at the point @xmath1 and let @xmath164 be the corresponding maximal ( in @xmath163 ) , self - stable family . write @xmath199_{y,\tilde r}.\ ] ] then , by proposition [ 6:p5.2 ] ( i ) , @xmath176 is maximal ( in @xmath166 ) , self - stable family and , by the supposition , there are a subsequence @xmath200 of @xmath7 and a maximal self - stable family @xmath201 such that @xmath202 and @xmath197 , the metric identification of @xmath203 , is tangent . for this @xmath179 consider the family @xmath204 by lemma [ l:2.3 ] we have equality .
write @xmath205_{z,\tilde r'}. $ ] it follows from and that @xmath206 and , moreover , proposition [ 6:p5.2 ] implies that @xmath198 , the metric identification of @xmath207 , is tangent .
let @xmath79 be a subspace of a metric space @xmath5 .
for @xmath208 and @xmath209 we denote by @xmath210 the sphere ( in the subspace @xmath79 ) with the center @xmath1 and the radius @xmath124 .
similarly for @xmath211 and @xmath209 define @xmath212 write @xmath213 [ 6:t5.4 ] let @xmath79 and @xmath94 be subspaces of a metric space @xmath5 and let @xmath95 .
then @xmath79 and @xmath94 are strongly tangent equivalent at the point @xmath1 if and only if the equality @xmath214 holds .
[ c:2.5 * ] let @xmath79 be a dense subset of a metric space @xmath6 .
then @xmath6 and @xmath79 are strongly tangent equivalent at every point @xmath208 .
[ p:2.6 ] let @xmath5 be a metric space with a marked point @xmath1 , @xmath215 a normalizing sequence and let @xmath216 be a maximal self - stable family .
then for every @xmath217 and every @xmath218 with @xmath219 there is @xmath220 such that the double inequality @xmath221 holds for all natural numbers @xmath222 . a simple proof is omitted here .
[ 4:l2.8 ] let @xmath5 be a metric space with a marked point @xmath1 , @xmath223 a normalizing sequence and @xmath88 a maximal self - stable family and @xmath224 a sequence of isometries @xmath225 with @xmath226 for all @xmath227 .
then the family @xmath228is a maximal self - stable family and , in addition , the pseudometric spaces @xmath229 , @xmath230 are isometric .
moreover , @xmath90 and @xmath231 , metric identifications of @xmath232 and , respectively , of @xmath233 , are simultaneously tangent or not . since @xmath234every two @xmath235 and @xmath236 are mutually stable if @xmath237 and @xmath238 are mutually stable and the mapping @xmath239is an isometry .
it is clear that @xmath240 .
hence it suffices to show that @xmath241 is maximal self - stable .
suppose there is @xmath242 such that @xmath243 but @xmath244 and @xmath245 are mutually stable for all @xmath246 .
it is easy to see that @xmath247where @xmath248 is the inverse isometry of the isometry @xmath249 and that @xmath245 and @xmath250 are mutually stable for each @xmath251 .
hence @xmath252 is not a maximal self - stable family , contrary to the condition of the lemma .
suppose that @xmath253 is tangent .
let @xmath254 be a subsequence of @xmath255 then , by proposition [ 1:p1.5 ] , the family @xmath256 is maximal self - stable .
the first part of the proof implies that the family @xmath257 is also maximal self - stable .
consequently , by proposition [ 1:p1.5 ] , @xmath258 is tangent .
in this section we describe tangent spaces to @xmath259 and @xmath260 .
[ 4:e2.1 ] let @xmath261 or @xmath262 and let @xmath263 we claim that each pretangent space @xmath264 ( to @xmath6 at the point @xmath138 ) is tangent and isometric to @xmath5 for all normalizing sequences @xmath7 .
consider the more difficult case @xmath261 .
[ 4:p2.4 ] let @xmath265 be maximal self - stable family and let @xmath266 be an element of @xmath267 such that @xmath268 the following statements are true .
\(i ) for every @xmath269 there is a finite limit @xmath270 and , conversely , if @xmath271 and this limit is finite , then @xmath272 .
\(ii ) for every two @xmath273 and @xmath274 from @xmath275 the equality @xmath276 holds if and only if @xmath277
\(iii ) the pretangent space @xmath278 which corresponds to @xmath275 is isometric to @xmath279 and tangent .
\(i ) if @xmath280 , then there are finite limits @xmath281for the case where @xmath282 we obtain @xmath283 because @xmath284 .
suppose @xmath285 then we have @xmath286write for every @xmath287 @xmath288where , as usual , @xmath289 then , it follows from , the limit @xmath290 exists if and only if there is the limit @xmath291 .
if the last limit does not exist , then there are two infinite sequences @xmath292 and @xmath293 of natural numbers such that @xmath294for all @xmath295 .
consequently , we obtain @xmath296 and similarly we have @xmath297thus we have the equality@xmath298which implies that @xmath299it is shown that for every @xmath300 there is a finite limit @xmath301 .
conversely , let @xmath302 and @xmath303we must show that for every @xmath304 there is a finite limit @xmath305 , i.e. @xmath245 and @xmath306 are mutually stable w.r.t .
@xmath87 . since @xmath307
, we have a finite limit @xmath308 hence @xmath309where constants @xmath310 are defined by and , respectively , by .
\(ii ) statement ( ii ) follows from .
\(iii ) statement ( i ) implies that the sequence @xmath311 with @xmath312 belongs to @xmath275 .
if we take @xmath313 instead @xmath314 in and , then we obtain the mapping @xmath315 where @xmath316it is easy to see that there is a unique mapping @xmath317 @xmath318 such that the diagram @xmath319{e , t}{p } \arrow{ese , b}{f } \node[2]{\omega_{0,\tilde r}}\arrow{s , r}{\psi } \\\node[3]{\mathbb r } \end{diagram } \ ] ] is commutative , where @xmath75 is the metric identification mapping , see . relations imply that @xmath320 is an isometry .
it remains to prove that @xmath89 is tangent .
let @xmath321 be a strictly increasing , infinite sequence of natural numbers and let @xmath322 be the corresponding subsequence of the normalizing sequence @xmath87 . if @xmath323 is a maximal self - stable family such that @xmath324 then , by statement ( i ) , for every @xmath325 there is a finite limit @xmath326define @xmath327 by the rule @xmath328 a simple calculation shows that @xmath329hence , by statement ( i ) , @xmath306 belongs to @xmath275 .
using proposition [ 1:p1.5 ] we see that @xmath278 is tangent .
[ 4:e2.5 ] let @xmath330 be the set of all complex numbers with the usual metric @xmath331 and the marked point @xmath138 and let @xmath332 be a normalizing sequence .
[ 4:p2.6 ] let @xmath275 be a maximal self - stable family with the corresponding pretangent space @xmath278 .
then @xmath278 is tangent and isometric to @xmath333 .
the proof is divided into four lemmas .
[ 4:l2.7 ] let @xmath273 and @xmath334 be elements of @xmath275 such that @xmath335then following statements are equivalent for every @xmath336 \(a ) @xmath244 belongs to @xmath337 \(b ) there are finite limits @xmath338 the implication ( a)@xmath339(b ) is trivial .
suppose that ( b ) holds with @xmath340 and with @xmath341 .
we must prove that there is a finite limit @xmath342write@xmath343since @xmath344 and the first relation in holds , there are finite limits @xmath345consequently we have the limit relations@xmath346@xmath347 @xmath348@xmath349and must prove the existence of @xmath350it is clear from , that , without loss of generality , it is sufficient to take @xmath351 for all @xmath352 .
moreover , evidently holds if @xmath353 hence we may also put @xmath354note that implies@xmath355since@xmath356the first relation in and imply that there exists @xmath357 and , in addition , if follows from that @xmath358see remark [ 4:r2.8 ] below . similarly using , and
last two relations from we see that there are limits @xmath359the remaining relations from imply the existence of @xmath360since there are limits and @xmath361 and @xmath362limit exists if and only if there is the limit @xmath363using and we obtain@xmath364and @xmath365for @xmath366 it implies the existence of . [ 4:r2.8 ] menger s notion of betweenness is well known for the metric spaces , see , for example , @xcite . for points
@xmath367 belonging to a pseudometric space @xmath368 , we may say that @xmath130 lies between @xmath369 and @xmath370 if @xmath371suppose in lemma [ 4:l2.7 ] we have @xmath372 then inequality does not hold if and only if some point from the set @xmath373 lies between two other points of this set ( in the pseudometric space @xmath374 ) .
[ 4:r2.9 ] for the future it is useful to note that if @xmath351 and @xmath375 $ ] for all @xmath227 , then the sequences@xmath376see , , are convergent .
indeed , the function@xmath377 \ni t\longmapsto \cos t\in \left [ -1,1\right]\]]is a homeomorphism .
hence @xmath378 is convergent because there is @xmath379 .
it implies the convergence of @xmath380 .
moreover , it follows from that@xmath381this relation and the existence of limits , imply the convergence of the sequences @xmath382 and @xmath383 .
consequently @xmath384 and @xmath385 are also convergent .
[ 4:l2.9 ] let @xmath330 be the set of all complex numbers with the usual metric @xmath331 and with the marked point @xmath138 and let @xmath386 be a normalizing sequence .
let @xmath387 be a maximal self - stable family for which holds with some @xmath344 . then there is a maximal self - stable family @xmath388 such that : \(i ) @xmath389 and @xmath390 are isometric ; \(ii ) there are @xmath391 and @xmath392 in @xmath393 for which the inequality @xmath394holds and @xmath395\]]for all @xmath227 .
let @xmath396 and @xmath397 be elements of @xmath275 for which holds . consider the sequence @xmath398 of the isometries @xmath399 then we have @xmath400we may assume that @xmath401for all @xmath227 . define a new sequence @xmath402 of isometries @xmath403 by the rule @xmath404using lemma [ 4:l2.8 ] we see that the family @xmath405satisfies all desirable conditions with @xmath406 [ 4:l2.10 ]
let @xmath330 be the set of all complex numbers with the usual metric @xmath331 and @xmath407 a normalizing sequence and @xmath275 a maximal self - stable family for which the conditions of lemma [ 4:l2.7 ] are satisfied . if @xmath408 is a pretangent spaces which corresponds @xmath409 then @xmath410 is tangent and isometric to @xmath333 .
let @xmath245 and @xmath306 be two elements of @xmath390 for which holds . by lemma [ 4:l2.9 ]
there exists a maximal self - stable family @xmath411 which is isometric to @xmath275 and contains @xmath412 and @xmath413 such that holds and @xmath414 \label{4:eq22.22}\]]for all @xmath415 relations imply that for every @xmath416 the sequence@xmath417is convergent , see remark [ 4:r2.9 ] . write@xmath418for every @xmath419 . in particular , we have @xmath420we claim that the function@xmath421is distance - preserving and onto . ( it immediately implies that @xmath422 , the metric identification of @xmath423 , and @xmath333 are isometric , so @xmath278 also is isometric to @xmath424 ) indeed , if @xmath425 , then @xmath426consequently it is sufficient to show that for every @xmath427 there is @xmath428 such that @xmath429 write @xmath430it is clear that @xmath431thus it is enough to prove that @xmath432 . it follows from that@xmath433hence@xmath434and similarly we have @xmath435 therefore , by lemma [ 4:l2.7 ] , @xmath436 it remains to show that @xmath437 is tangent , because , by lemma [ 4:l2.8 ] , in this case @xmath438 is also tangent .
let @xmath439 be a subsequence of @xmath87 and let @xmath440 be maximal self - stable family such that @xmath441then @xmath442 satisfies all conditions of the lemma which is being proved .
hence , similarly , we can define a function@xmath443for @xmath444 let @xmath445 be the metric identification of @xmath446 and let @xmath447 @xmath448 be isometries such that the diagrams @xmath449{e , t}{f } \arrow{se , t}{p}\node[2]{\mathbb c}\arrow{sw , r}{is } \\\node[2]{\omega^*_{0,\tilde r } } \end{diagram } \qquad\text{and}\qquad \begin{diagram } \node{\tilde x^*_{0,\tilde r'}}\arrow[2]{e , t}{f ' } \arrow{se , t}{p'}\node[2]{\mathbb c}\arrow{sw , r}{is ' } \\\node[2]{\omega^*_{0,\tilde r ' } } \end{diagram}\ ] ] are commutative .
similarly we can define an isometric embedding @xmath450 such that the diagram @xmath451{e , t}{\text{in}_{\tilde r ' } } \arrow[2]{s , l}{p}\node[2]{\tilde x^*_{0,\tilde r ' } } \arrow[2]{s , r}{p'}\\ \\ \node{\omega^*_{0,\tilde r}}\arrow[2]{e , b}{em'}\node[2]{\omega^*_{0,\tilde r ' } } \end{diagram}\ ] ] is commutative .
we claim that the following diagram @xmath452{e , t}{\text{in}_{\tilde r ' } } \arrow{se , t}{f } \arrow[2]{s , l}{p}\node[2]{\tilde x^*_{0,\tilde r ' } } \arrow{sw , l}{f ' } \arrow[2]{s , r}{p'}\\ \node[2]{\mathbb c } \arrow{sw , r}{is } \arrow{se , b}{is'}\\ \node{\omega^*_{0,\tilde r}}\arrow[2]{e , b}{em'}\node[2]{\omega^*_{0,\tilde r ' } } \end{diagram}\ ] ] also is commutative . to prove the commutativity of it is sufficient to show that @xmath453for every @xmath454 let @xmath370 be a point of @xmath455 since @xmath81 is a surjection , there is @xmath456 such that @xmath457 hence , using the commutativity of diagrams and and the equality @xmath458 we obtain@xmath459@xmath460consequently holds .
the diagram@xmath461{s , l}{p } \arrow{ese , t}{f}\\ \node[3]{\mathbb c } \arrow{ese , t}{is'}\\ \node{\omega^*_{a,\tilde r } } \arrow[3]{e , t}{em ' } \node[4]{\omega^*_{a,\tilde r ' } } \end{diagram}\ ] ] also is commutative , because is commutative .
since @xmath81 and @xmath462 are surjections , @xmath463 is surjective .
hence , by proposition [ 1:p1.5 ] , @xmath464 is tangent .
the following lemma shows that if @xmath330 , then every maximal self - stable @xmath275 satisfies the conditions of lemma [ 4:l2.7 ] .
it is a final part of the proof of proposition [ 4:p2.6 ] .
[ 4:l2.13 ] let @xmath330 be the set of all complex numbers with the usual metric @xmath465 let be @xmath466 a normalizing sequence and let @xmath275 be a maximal self - stable family .
then there are @xmath467 such that holds , i.e.,@xmath468 suppose that the equality @xmath469holds for all @xmath344
. consider first the simplest case where @xmath470 for all @xmath471 the last equality implies @xmath472for all @xmath473 .
consequently @xmath474 is a self - stable family , so @xmath275 is not maximal self - stable , contrary to the conditions . hence there is @xmath475 such that @xmath476 without loos of generality we may suppose that @xmath477indeed , passing , if necessary , to an isometric @xmath389 , see lemma [ 4:l2.8 ] , we may put @xmath478 for all @xmath227 .
next , since the family @xmath479is maximal self - stable if @xmath480 we can take @xmath481 instead of @xmath482 .
moreover , since @xmath483we see that @xmath484 and , consequently , @xmath485the equalities and imply that @xmath486for all @xmath487 there exist only the following three possibilities under which holds : @xmath488write for @xmath489 @xmath490we claim that the limit relation @xmath491holds for every @xmath492 . indeed , fix @xmath489 , and suppose that the first equality from holds .
( note that is trivial if @xmath493 or @xmath494 . )
let us denote by @xmath495 an arbitrary limit point of the sequence @xmath496 to prove it is sufficient to show that @xmath497 and @xmath498 are situated on the `` real axis''.,width=340 ] the fist equality in and definition imply that @xmath499hence @xmath500 belongs to the intersection of the circumferences @xmath501 since this intersection has the unique element @xmath502 see fig . 1 , we obtain .
similarly we have @xmath503 if @xmath504 and @xmath505for the case @xmath506 , i.e. , @xmath507 for all possible cases .
to complete the proof let us consider the family @xmath508since holds for all @xmath272 , it is easy to prove that @xmath509_x\supseteq \tilde{x}_{0,\tilde r}$ ] where the operation @xmath510_x$ ] was defined in . note that @xmath511_x$ ] is self - stable and that @xmath512 for every @xmath287 .
hence @xmath513 contrary to the supposition about the maximality of @xmath514
examples [ 4:e2.1 ] and [ 4:e2.5 ] are some particular cases of starlike sets on the plane . the next our goal is to prove theorem [ t:1.7 ] which describes tangent spaces to arbitrary starlike subsets of @xmath260 .
[ t:4.1 ] let @xmath515 be a starlike set with a center @xmath1 and let @xmath7 be a normalizing sequence .
then all pretangent spaces to @xmath79 at the point @xmath1 lie in tangent spaces and for each tangent space @xmath140 there is an isometry @xmath516 before proving the theorem we consider its particular cases .
if @xmath79 is an one - point set , then all rays are also one - point and so @xmath517 .
moreover , it is easy to see that , in this case , all pretangent spaces are tangent and one - point .
hence the theorem is valid if @xmath518 . for the case
when each three point of @xmath79 are collinear , desirable isometry was , in fact , constructed in the proof of proposition [ 4:p2.4 ] , see diagram .
if @xmath519 and @xmath520 then the theorem follows from proposition [ 4:p2.6 ] .
consequently it is enough examine the case where @xmath521 and @xmath79 contains at least three noncollinear points .
[ 8:l2.16 ] let @xmath522 be a closed cone with a vertex @xmath138 .
let @xmath523 be a marked point of @xmath524 @xmath525 a normalizing sequence and @xmath275 a maximal self - stable family .
suppose @xmath6 contains at least three noncollinear points and there exist @xmath526 and @xmath527 from @xmath275 such that @xmath528 holds and sequences @xmath529are convergent .
then the following statements hold .
suppose that @xmath536since @xmath245 and @xmath244 are mutually stable , there is a limit @xmath537where@xmath538similarly we have @xmath539where@xmath540consider the system @xmath541the inequality implies that@xmath542hence , by cramer s rule , we obtain from @xmath543@xmath544consequently , the existence of limits and implies the existence of limit . note also that the elements of the sequence @xmath545 are some points of @xmath6 .
hence we have because @xmath6 contains all its limit points as a closed set .
now suppose that @xmath533 and relations , hold . since @xmath275 is maximal self - stable , to prove @xmath546 it is sufficient to show that that @xmath244 and @xmath547 are mutually stable for each @xmath548 let @xmath549 be an arbitrary element of @xmath275 .
statement ( i ) implies that there is @xmath550 such that @xmath551 hence , by , @xmath552i.e .
, @xmath244 and @xmath549 are mutually stable .
let @xmath7 be a normalizing sequence and let @xmath560 be a maximal self - stable family with a corresponding pretangent space @xmath561 .
by lemma [ 8:l2.17 ] we may suppose , passing , if necessary , to a subsequence of @xmath7 , that there exist @xmath562 and @xmath563 in @xmath275 such that holds .
since the sequences@xmath564are bounded , there is a subsequences @xmath565 of sequence @xmath87 such that @xmath566are convergent .
let @xmath567 be a maximal self - stable family such that @xmath568write @xmath569 for the metric identifications of of @xmath556 .
we claim that @xmath570 is tangent . indeed , since all suppositions of lemma [ 8:l2.16 ] are valid , a limit@xmath571exists for every @xmath572 and the mapping @xmath573is distance - preserving and onto . hence there is an isometry @xmath574 such that the diagram @xmath575{e , t}{f ' } \arrow{se , b}{p ' } \node[2]{x}\arrow{sw , b}{is'}\\ \node[2]{\omega_{0,\tilde r ' } } \end{diagram}\ ] ] is commutative .
in particular we have @xmath576 .
similarly , for every infinite subsequence @xmath577 of @xmath151 and for every maximal self - stable @xmath578there are a distance - preserving surjection @xmath579 and an isometry @xmath580 with the commutative diagram @xmath581{e , t}{f''}\arrow{se , b}{p '' } \node[2]{x } \arrow{sw , b}{is '' } \\
\node[2]{\omega_{0,\tilde r '' } } \end{diagram}\ ] ] and with @xmath582 , where @xmath583 is the metric identification mapping of the pseudometric space @xmath584 as in the case of diagram we can find an isometric embedding @xmath585 such that @xmath586 where @xmath587 for @xmath588 . repeating the proof of the commutativity of we
see that the diagram @xmath589{e , t}{\text{in}\,r '' } \arrow{se , t}{f ' } \arrow[2]{s , l}{p ' } \node[2]{\tilde x_{0,r '' } } \arrow{sw , t}{f '' } \arrow[2]{s , r}{p''}\\ \node[2]{x } \arrow{sw , t}{is ' } \arrow{se ,
t}{is''}\\ \node{\omega_{0,\tilde r ' } } \arrow[2]{e , t}{em '' } \node[2]{\omega_{0,\tilde r '' } } \end{diagram}\ ] ] is also commutative .
hence @xmath590 is surjective because @xmath591 and @xmath592 are surjections .
consequently , by proposition [ 1:p1.5 ] , @xmath593 is tangent and , as was shown above , for the isometry @xmath594 we have @xmath595 where @xmath596 .
thus , by definition , @xmath278 lies in the tangent space @xmath597 suppose now that @xmath438 , the metric identification of @xmath598 , is tangent . to prove the existence of an isometry @xmath599 consider , as in the first part of the present proof , a maximal self - stable family @xmath598 such that inclusion holds and diagram is commutative for function . since @xmath600 is an isometry , the commutativity of implies @xmath601 where @xmath602 is the inverse function of @xmath600 .
then combining the last equality with we obtain the commutative diagram @xmath603{e , t}{in_{\tilde r ' } } \arrow{s , l}{p } \node[2]{\tilde x_{0 , \tilde r'}}\arrow[2]{e , t}{f ' } \arrow{s , r}{p ' } \node[2]{x}\\ \node{\omega_{0 , \tilde r } } \arrow[2]{e , t}{em ' } \node[2]{\omega_{0 , \tilde r ' } } \arrow{ene , b}{(is')^{-1 } } \end{diagram}\ ] ] write @xmath604 for tangent spaces the mapping @xmath605 is an isometry , so @xmath606 is an isometry as a superposition of two isometries .
note that the commutativity of diagram implies the equality @xmath607 for @xmath608 .
let us denote by @xmath612 the smallest ( but not necessarily closed ) cone with the vertex @xmath1 and such that @xmath613 then we have the equality @xmath614 where @xmath615 is the closure of @xmath612 in @xmath260 .
indeed , it is easy to prove that @xmath615 is a closed cone
. consequently , the inclusion @xmath616 holds . on the other hand
@xmath617 is a cone .
hence @xmath618 .
it implies @xmath619 and follows . since @xmath623 it is sufficient , by proposition [ 6:t5.4 ] , to prove that @xmath624 where @xmath625 if equality does not hold then there exists @xmath626 such that @xmath627 equality implies that @xmath628 where @xmath629 .
since @xmath630 is a compact subset of @xmath260 , it follows from that there is a finite @xmath631-net @xmath632 for the set @xmath630 .
the starlikeness of @xmath79 implies that there is @xmath633 such that the implication @xmath634 is true for every point @xmath635 where @xmath636 are rays starting from @xmath1 and passing through @xmath637 , see .
hence for all @xmath6380,\gamma[\,$ ] we have the inequality @xmath639 contrary to . without loss of generality we may put @xmath523 .
let @xmath640 be a maximal self - stable family and let @xmath140 be the corresponding pretangent space .
write @xmath641 .
then by lemma [ l:4.6 ] the sets @xmath6 and @xmath79 are strongly tangent equivalent at the point @xmath1 .
it follows from proposition [ 8:p2.18 ] that every pretangent space @xmath111 lies in some tangent @xmath642 .
consequently , using proposition [ p:2.4 ] , we have that @xmath140 lies in some tangent space @xmath197 .
suppose now that @xmath111 is tangent .
write @xmath643_x$ ] . then statement ( ii ) of proposition [ 6:eq5.2 ] implies that @xmath111 , the metric identification of @xmath216 , also is tangent .
hence , by proposition [ 8:p2.18 ] , there is an isometry @xmath644 with @xmath645 where @xmath75 is the projection of @xmath54 on @xmath90 .
statement ( ii ) of proposition [ 6:p5.2 ] implies that the mapping @xmath646_x\in\omega_{a,\tilde r}^x\ ] ] is an isometry .
it is easy to see that the mapping @xmath647 is an isometry with the desirable properties .
it follows from the first part of theorem [ t:1.7 ] that we must only to prove the equality @xmath648 for convex sets @xmath515 and @xmath208 . to prove this , note that the cone @xmath612 is convex for the convex @xmath79 , see , for example , ( *
* chapter i , 2 , corollary 2.6.3 ) and that @xmath615 also is convex ( * ? ? ?
* chapter ii , 6 , theorem 6.1 ) .
moreover , as has been stated in the proof of lemma [ l:4.6 ] , the closure of every cone is a cone .
hence @xmath649 this inclusion and equality imply the inclusion @xmath650 .
since the reverse inclusion @xmath651 is trivial , we obtain .
let @xmath94 be a starlike set with the center @xmath1 such that @xmath94 and @xmath6 are strongly tangent equivalent at the point @xmath1 . to prove the theorem under consideration we can repeat the proof of theorem [ t:4.1 ] using @xmath94 instead of @xmath617 , @xmath6 instead of @xmath79 and theorem [ t:4.1 ] in place of proposition [ 8:p2.18 ] . to prove note that @xmath653 because @xmath654 .
on the other hand , definition [ 9:d1 ] implies the equality @xmath655 for every @xmath656
. consequently , @xmath657 applying the well - known criterion , see ( * ? ? ?
* proposition 2.1.15 ) , we see that a distance function @xmath658 is closed .
the characteristic property of closed maps , ( * ? ? ? * chapter 1 , 13 , xiv , formula ( 7 ) ) implies @xmath659 moreover , it is easy to see that @xmath660 for every @xmath661 . relations imply the inequality @xmath662 for every @xmath661 .
for example we have @xmath663 letting @xmath664 in the last inequality and in we obtain .
[ l:4.7 ] let @xmath522 be a set with a marked point @xmath1 , @xmath135 a ray with the vertex @xmath1 , @xmath96 a normalizing sequence , @xmath665 a positive constant and let @xmath666 belong to @xmath667 , @xmath668 , @xmath669 .
suppose the following conditions are satisfied : * the family @xmath670 is self - stable w.r.t . @xmath7
; * the point @xmath22 lies between @xmath53 and @xmath671 , i.e. , @xmath672 * for every @xmath120 there is @xmath220 such that @xmath673 for all @xmath222 .
if the conclusion of the lemma does not hold , then there is a strictly increasing sequence @xmath677 of natural numbers such that @xmath678 for all @xmath679 .
condition @xmath680 implies that @xmath681 is a bounded sequence . hence , passing if necessary to a subsequence , we may suppose that @xmath682 is convergent .
condition @xmath683 and the existence of @xmath684 imply that @xmath685 also is convergent and @xmath686 write @xmath687 it follows from that @xmath688 where @xmath689 is the interior of the sector @xmath127 . using and it is easy to show that @xmath690 see fig .
2 . the last inequality contradicts because @xmath691 , @xmath692 and @xmath693 .
firstly , note that without loss of generality we may assume @xmath6 to be closed .
indeed , by corollary [ c:2.5 * ] , the sets @xmath6 and @xmath695 are strongly tangent equivalent for every @xmath25 , so using statement @xmath696 of proposition [ 6:p5.2 ] we see that for @xmath6 and for @xmath695 the supposition of theorem [ t:1.9 ] is true ( or false ) simultaneously .
analogously , using lemma [ l:4.6 * ] we can replace @xmath6 by @xmath695 in the conclusion of theorem [ t:1.9 ] .
suppose there is a ray @xmath118 such that @xmath697 since function @xmath137 is decreasing in @xmath698 , definition [ 9:d1 ] implies that for every @xmath6990,1[$ ] there is @xmath700 with @xmath701 where @xmath702 is the length of the longest interval in @xmath703\setminus r(x , l,\beta_0)$ ] .
consequently , there is a decreasing sequence @xmath704 , with@xmath705 such that @xmath706 write @xmath707r_n , t_n[$ ] for the longest open interval in @xmath708\setminus r(x , l,\beta_0)$ ] . if @xmath709 for some @xmath674 , then does not hold .
thus we may suppose @xmath710 for all @xmath711 .
let @xmath712 be a decreasing sequence of positive numbers with @xmath713 .
write @xmath714 where @xmath715 .
the previous reasoning gives the strict inequality @xmath716 for all @xmath717 . since @xmath6 is closed and @xmath718 is compact , there is a sequence @xmath719 such that @xmath720 for all @xmath717 .
let us obtain now some estimations for @xmath721 and for @xmath722 .
using and the definition of intervals @xmath707r_m , t_m[$ ] and of the porosity of @xmath723 we see that @xmath724 that is @xmath725 and so @xmath726 moreover , since the inclusion @xmath727 holds for all @xmath717 , we obtain the inequality @xmath728 for all @xmath729 . letting
@xmath730 we have @xmath731 passing to a subsequence we may suppose that there exists a limit @xmath732 it is clear that @xmath733 for all @xmath711 .
hence , by , and by , @xmath734 by the construction we have @xmath735 . thus implies @xmath736 or , in an equivalent form , @xmath737 similarly we can rewrite as @xmath738 from this , using , , we obtain @xmath739 and , after simple calculations , @xmath740 note that the inequality @xmath741 holds if @xmath742 indeed , inequality is equivalent to @xmath743 by we have @xmath744 .
consequently , to prove it is sufficient to show @xmath745 that is equivalent to @xmath746 because @xmath747 but @xmath748 , see , so implies .
let us take the sequence @xmath749 as a normalizing sequence .
let @xmath750 be a maximal self - stable family such that @xmath751 and let @xmath752 be the corresponding pretangent spaces .
the supposition of the theorem which is being proved , implies that there is a subsequence @xmath677 of natural numbers such that @xmath752 lies in tangent space @xmath753 , @xmath754 .
replacing @xmath755 by @xmath679 in , we may assume that @xmath756 , i.e. @xmath757 is tangent .
write @xmath758 where @xmath75 is the metric identification mapping of @xmath750 and where @xmath719 was defined by , .
it follows from and from the supposition of the theorem that there is @xmath759 with @xmath760 and such that @xmath761 lies between @xmath762 and @xmath763 .
since all conditions of lemma [ l:4.7 ] are satisfied by the triple @xmath764 , there is @xmath220 such that @xmath765 for all @xmath222 .
lemma [ p:2.6 ] and relation imply that there is @xmath766 such that the double inequality @xmath767 holds for all sufficiently large @xmath755 .
moreover , it follows from , that there is @xmath768 such that @xmath769 and @xmath770 for all @xmath771 .
the left inequality in and the right one in give @xmath772 similarly , from the right inequality in and from the left one in we obtain @xmath773 thus we have @xmath774r_n , t_n[\ ] ] for sufficiently large @xmath755 .
in addition implies @xmath775 relations , contradict the definition of intervals @xmath707r_n , t_n[$ ] .
hence double inequality does not hold and the theorem follows . [ 8:e2.20 ]
let @xmath776\},\ 0<\theta_1-\theta_2\leq\pi\}$ ] be the closed convex cone , @xmath523 a marked point of @xmath6 and @xmath215 a normalizing sequence .
write for all @xmath717 @xmath777 where a number @xmath778 belongs to @xmath707\theta_1,\theta_2[$ ] .
indeed , suppose that @xmath438 is tangent .
write @xmath780 and @xmath781 . by statement @xmath696 of proposition [ 1:p1.5 ]
the families @xmath782 and @xmath783 are also maximal self - stable and corresponding spaces @xmath784 , @xmath785 are tangent . passing , if necessary , to subsequences we may suppose that for every @xmath786 there are limits @xmath787 let @xmath788 be a system of distinct representatives of the factor space @xmath438 , i.e. , for every @xmath789 there is a unique @xmath790 such that @xmath791 .
then , by statement @xmath683 of lemma [ 8:l2.16 ] , the mappings @xmath792 are isometries , where @xmath793 and @xmath794 are defined by .
therefore , we have @xmath795 hence , there is an isometry @xmath796 such that @xmath797 note that @xmath798 and @xmath799 .
consequently , equality contradicts the brouwer theorem on the invariance of the open sets .
the results of this paper were partly produced during the visit of the first author to the mersin university ( turkey ) in february - april 2008 under the support of the tbitak - fellowships for visiting scientists programme . | let @xmath0 be a starlike set with a center @xmath1 .
we prove that every tangent space to @xmath2 at the point @xmath1 is isometric to the smallest closed cone , with the vertex @xmath1 , which includes @xmath2 . a partial converse to this result
is obtained .
the tangent space to convex sets is also discussed . + + * key words : * metric spaces ; tangent spaces to metric spaces ; tangent space to convex sets ; tangent space to starlike sets .
+ + * 2000 mathematic subject classification : * 54e35 |
in @xmath10 decay , it is important to be able to evaluate future measurements of competing observables consistent with the standard model ( sm ) prediction of only a @xmath11 coupling and of only its associated discrete - symmetry violations . for this purpose , without consideration of possible explicit @xmath5 violation , in [ 1 ] plots were given of the values of the helicity parameters in terms of a
@xmath12 @xmath13 single additional lorentz structure " versus effective - mass scales for new physics , @xmath14 , associated with each additional lorentz structure . in this paper , the effects of possible explicit @xmath5 violation are reported . in the present formulation , by explicit @xmath15 violation " , c.f .
sec.2 , we mean an additional complex - coupling , @xmath16 or @xmath17 , associated with a specific single additional lorentz structure , @xmath4 .
the main motivation for the present analysis are the observed @xmath18 and @xmath19 violations in @xmath20 decay .
although these discrete - symmetry violations are empirically well - described by the ckm matrix which describes the linear superposition of the quark mass eigenstates which appears in the phenomenological weak eigenstates , the fundamental origin of these symmetry violations is still unknown .
experimental results should soon be available about whether the ckm formulation is also successful in b - quark decay . in the case of the strong interactions ,
there is the opposite difficulty of a fundamental strong @xmath18 problem which has led to the prediction of the existence of axions , the nambu - goldstone bosons associated with a global @xmath21 symmetry .
these axions have yet to be discovered .
lastly , and perhaps more significantly for t - quark decay , most astrophysics studies of electroweak baryogenesis conclude that additional sources of @xmath18 violation , beyond ckm , in elementary particle physics are necessary to explain the observed baryon - to - photon ratio .
so in spite of the robustness of the standard model and of the ckm formulation , perhaps after all , t - quark decay is not the wrong place to look for @xmath18 and @xmath19 violations .
a first measurement of the longitudinal w boson fraction was reported in [ 2 ] .
a recent working group review of t - quark physics is in [ 3 ] .
a recent review of @xmath18 violation in t - quark physics is in [ 4 ] . besides these references and those listed in [ 1 ] ,
some of the related recent literature is [ 5 - 14 ] .
the present analysis assumes that future measurements of @xmath22 decay will be at least approximately consistent with the sm prediction of only a @xmath23 coupling .
if the sm is correct , one expects that the @xmath24 and @xmath25 moduli and relative phase @xmath26 will be the first quantities to be somewhat precisely determined . as shown by table 1 ,
the @xmath27 moduli are factors of 30 and 100 smaller in the sm .
the helicity parameters appear directly in various polarization and spin - correlation functions for @xmath28 decay such as those obtained in [ 15 ] . by measurement of independent helicity parameters , or from other empirical analyses of spin - correlation and polarization observables ,
it will be possible to test in several independent ways that the r - handed b - quark amplitudes , @xmath27 , are indeed negligible to good precision .
eventually there should also be direct evidence for their existence if the sm is correct .
if the r - handed amplitudes are negligible , then besides @xmath29 it follows that @xmath30 and that @xmath31 .
showing an approximate empirical absence of r - handed amplitudes would also be useful in regard to tests for @xmath15 violation : assuming that the l - handed amplitudes dominate , the @xmath32 helicity parameter satisfies the relation @xmath33 @xmath34[p(b_{l})-\zeta ] -($ ] @xmath35 .
for instance , in the sm the vanishing of the right - hand - side is due to the vanishing of @xmath36 provided that the r - handed amplitudes are negligible .
if @xmath15 violation were to occur , besides normally a non - zero approximate - equality in the above relation , there would normally also be a non - zero @xmath37 if the r - handed amplitudes are negligible . * remarks on the dynamical phase - type ambiguities : * due the dominance of the l - handed amplitudes in the sm , the occurrence of the two dynamical ambiguities [ 1 ] displayed in lower part of table 1 is not surprising because these three chiral combinations only contribute to the l - handed b - quark amplitudes in the @xmath38 limit . since pairwise
the couplings are tensorially independent , the @xmath39 and @xmath40 mixtures can each be tuned by adjusting a purely real @xmath14 to reproduce , with opposite sign , the sm ratio of the two @xmath41 l - handed amplitudes . likewise , if experimental data were to suggest that the r - handed amplitudes are larger than expected , e.g. @xmath42 , this might be due to the presence of additional @xmath43 type couplings .
since the @xmath44 couplings only contribute to the longitudinal helicity w amplitudes , they might be of interest in the case of an unexpected w longitudinal / transverse polarization ratio .
versus the upper part of table 1 , given the small @xmath45 mass , this is the reason that the sign of the @xmath46 amplitude can be switched , without other important changes , by the addition of the @xmath47 coupling .
however , in the case of the @xmath48 phase - type ambiguity , from table 1 there are 3 numerical puzzles versus the sm values . in the upper part
, the @xmath49 amplitude for @xmath50 has about the same value in @xmath51 units , as the @xmath52 amplitude in the sm . as @xmath38 , @xmath53 .
the other numerical puzzle(s ) is the occurrence in the lower part of the table 1 of the same magnitude of the two r - handed b - quark amplitudes @xmath54 for the sm and for the case of @xmath55 . except for the differing partial width , by tuning the magnitude of l - handed amplitude ratio to that of the sm ,
the r - handed amplitude s moduli also become about those of the sm .
with @xmath56 determined as in sec .
3 , for the @xmath57 amplitudes @xmath58 versus for instance @xmath59 .
of course , the row with sm values is from a theory " whereas the row of @xmath50 values is not .
nevertheless , dynamical ssb and compositeness / condensate considerations do continue to stimulate interest [ 15 ] in additional tensorial @xmath60 couplings . in table 1 , due to the additional @xmath61 coupling , the net result is that it is the @xmath62 helicity amplitudes @xmath57 which get an overall sign change . fortunately , a sufficiently precise measurement of the sign of @xmath63(sm ) due to the large interference between the w longitudinal / transverse amplitudes can resolve the @xmath64 and @xmath61 lines of this table .
similarly , sufficiently precise measurements of both @xmath65 and @xmath66 could resolve the analogous dynamical ambiguity in the case of a partially - hidden @xmath15 violation associated with the additional @xmath67 coupling , see figs.8 in sec .
3 . a precise measurement of the partial width
@xmath68 could also be useful .
for 1em @xmath0 , the most general lorentz coupling is @xmath69 where @xmath70 , and @xmath71 @xmath72 in this paper , in consideration of the additional lorentz structures to pure @xmath64 , we consider the @xmath73 or @xmath14 as complex phenomenological parameters . for @xmath74 units with @xmath75 ,
the nominal size of @xmath14 is @xmath76 , see [ 1 ] . in the sm , the ew energy - scale is set from the higgs - field vacuum - expectation - value by the parameter @xmath77 and @xmath78 a specific discrete symmetry implies a definite symmetry relation among the associated helicity amplitudes . in the case of @xmath5 invariance , the respective helicity amplitudes must be purely real , @xmath79 @xmath80 intrinsic and relative signs of the helicity amplitudes are specified in accordance with the standard jacob - wick phase convention . @xmath15
invariance will be violated if either ( i ) there is a fundamental violation of canonical time reversal " invariance , or ( ii ) there are absorptive final - state interactions . in the sm , there are no such final - state interactions at the level of sensitivities considered in the present analysis . to keep this assumption of the absence of final - state interactions " manifest
, we refer to this as @xmath5 invariance , see [ 15 ] and references therein
. if experimental evidence for @xmath15 violation were found , it would be very important to establish whether ( i ) , ( ii ) , or some combination of the two effects was occurring .
for instance , unexpected final - state interactions might be associated with addition t - quark decay modes . to assess future measurements of helicity parameters in regard to @xmath15 violation , figs . 1 - 5 , are for the case of a single additional pure - imaginary coupling , @xmath2 or @xmath81 , associated with a specific additional lorentz structure , @xmath82 . in the sm ,
all the relative phases are either zero or @xmath83 so all of the primed helicity parameters are zero
. the two plots displayed in fig.1 are for dimensional couplings with chiral @xmath86 and non - chiral @xmath87 lorentz structures .
the upper plot displays the @xmath88 helicity parameter versus the effective - mass scale @xmath89 with @xmath90 in @xmath51 units .
this parameter is defined by @xmath91 is the relative phase difference the two helicity amplitudes in ( 5 ) .
the peaks of the curves shown in the upper plot do not however correspond to where @xmath92 . instead , at the peaks @xmath93 .
the lower plot displays the induced effect of the additional coupling on the partial width for @xmath94 .
the sm model limit is at the wings " where @xmath95 for each additional dimensional coupling .
if the r - handed b - quark amplitudes were found not to be negligible , it would be important to consider both @xmath96 or equivalently both @xmath97 and @xmath98 .
fig.2 displays plots of the b - polarimetry interference parameters @xmath99 and @xmath100 versus @xmath89 for the case of a single additional @xmath101 and @xmath102 coupling .
these helicity parameters are defined by @xmath103 where @xmath104 and @xmath105 are the relative phases between the two amplitudes in ( 6 ) . in the sm , the analogous @xmath106 in which the cosine function replaces the sine function are the two @xmath107 helicity parameters between the amplitudes with the largest moduli .
unfortunately , the tree - level values of @xmath108 in the sm are only about @xmath109 .
two dimensional plots of the type @xmath110 and @xmath111 , and of their primed counterparts , have the useful property that the unitarity limit is a circle of radius @xmath112 centered on the origin . in both plots
, the peaks in the curves do correspond respectively to where @xmath113 and @xmath114 , except that for @xmath115 in the lower @xmath116 plot where @xmath117 at the peak .
the drops in the curves for small @xmath118 s is due to the vanishing of the sine of the corresponding relative phase .
curves are omitted in the plots in this paper when the couplings produce approximately zero deviations in the helicity parameter of interest , e.g. this occurs for @xmath119 in both the @xmath99 and @xmath100 helicity parameters . an additional @xmath64 type coupling with a complex phase versus the sm s @xmath122 is equivalent to an additional overall complex factor in the sm s helicity amplitudes .
this will effect the overall partial width @xmath68 , but it does nt effect the other helicity parameters . for a single additional gauge - type coupling @xmath123 or @xmath124 , in fig.3 are plots of the b - polarimetry interference parameters @xmath99 and @xmath100 , and of the partial width for @xmath94 versus pure - imaginary coupling constant @xmath81 .
the @xmath125 value is in @xmath51 units . in the cases of the additional dimensionless , gauge - type couplings ,
the sm model limit is at the origin , @xmath126 .
the peaks for the @xmath124 coupling do correspond to where the associated sine of the relative phase has maximum magnitude ; instead , for the @xmath127 couplings , @xmath128 @xmath129 at the peaks .
the plots in fig.4 show the indirect effects of a single additional pure - imaginary chiral coupling , @xmath2 or @xmath81 , on other helicity parameters . for
the coupling strength ranges listed in the middle table " , the upper plot shows the effects on the probability , @xmath130 , that the emitted @xmath131 is longitudinally " polarized and the effects on the probability , @xmath9 , that the emitted b - quark has left - handed " helicity .
each curve is parametrized by the magnitude of the associated @xmath132 or @xmath133 . on each curve
, the central open circle corresponds to the region with a maximum direct @xmath5 violation signature , e.g. for @xmath119 from fig.1 this is at @xmath134 .
the dark / light solid circles correspond respectively to the ends of the ranges listed in the middle table where the direct signatures fall to about @xmath135 of their maximum values .
similarly the lower plot is for the w - polarimetry interference parameters @xmath136 .
curves are omitted for the remaining moduli parameter @xmath137 , the pre - ssb parameter which characterizes the odd - odd mixture of the @xmath138 and @xmath139 polarizations [ 15 ] , because a single additional pure - imaginary coupling in these ranges produces approximately zero deviations from the pure @xmath64 value of @xmath140 .
the plots in fig.5 show the indirect effects of a single additional pure - imaginary non - chiral coupling on other helicity parameters . versus the middle table given here ,
the curves are labeled as in fig.4 .
it is instructive to compare the above plots with their analogues in [ 1 ] . unlike in the analogous plots in
[ 1 ] , finite @xmath45 effects do not lead to sizable oval shapes " in plots in this paper because interference terms must vanish in intensities arising from the sum of a real @xmath64 amplitude and the pure - imaginary @xmath2 or @xmath81 amplitude . in summary ,
sizable @xmath141 violation signatures can occur for low - effective mass scales ( @xmath6 ) as a consequence of pure - imaginary couplings associated with a specific additional lorentz structure .
however , in most cases , such additional couplings can be more simply excluded by @xmath7 precision measurement of the probabilities @xmath8 and @xmath142 . in most cases , the w - polarimetry interference parameters @xmath143 and
@xmath144 can also be used as indirect tests , or to exclude such additional couplings .
the purpose of this section is to consider the situation when the @xmath15 violation exists in the decay helicity amplitudes , but nevertheless does not significantly show up in the values of the moduli parameters .
based on the notion of a complex effective mass scale parameter @xmath145 where @xmath146 varies with the mass scale @xmath147 , we exploit the dynamical phase - type ambiguities to construct two simple phenomenological models in which this happens . when @xmath148 @xmath149 , the imaginary part of @xmath150 could be interpreted as crudely describing a more detailed / realistic dynamics with a mean lifetime scale @xmath151 of pair - produced particles at a production threshold @xmath152 $ ] . for @xmath153 @xmath154 , fundamental time - reversal violation or a new dynamics might approximately correspond to such a complex @xmath155 . in the case of the @xmath48 ambiguity , over the full @xmath146 range ,
this construction does preserve the magnitudes puzzle , see sec.1 , between the @xmath64 and @xmath61 lines in the lower part of table 1 .
in fig.6 are plots of the signatures for a partially - hidden @xmath5 violation associated with a @xmath47 phase - type ambiguity .
we require @xmath156 to hold when the additional @xmath47 coupling , @xmath157 has a complex effective mass scale parameter @xmath158 where @xmath146 varies with the mass scale @xmath159 . for @xmath160 ,
the resulting relationship is @xmath161 for @xmath162 which correspond respectively to @xmath163 .
this construction maintains the standard model values in the massless b - quark limit for the four moduli parameters , @xmath164 and @xmath68 .
the function @xmath165 is then used for the @xmath47 coupling when @xmath166 .
the sm values for the moduli parameters are essentially unchanged .
the phase choice of @xmath167 , cf .
top line in table 1 , has no consequence since it is a @xmath168 phase difference . for @xmath169 , in fig .
6 the solid curve shows @xmath88 plotted versus @xmath170 .
the dashed curve is for @xmath171 which are degenerate .
the dark rectangles show the standard model values at the @xmath172 endpoint . at the other endpoint @xmath173 , or @xmath174 . from the perspectives of ( i ) measuring the @xmath175 interference parameters and of ( ii ) excluding this type of @xmath15 violation , it is noteworthy that where @xmath176 has the maximum deviation , there is a zero in @xmath171 .
so if the latter parameters were found to be smaller than expected or with the opposite sign than expected , this would be consistent with this type of @xmath15 violation . at the maximum of @xmath176 ,
@xmath177 and the other @xmath15 violation parameters are also maximum .
the curves for these parameters have the same over all shape as @xmath176 but their maxima are small , @xmath178 and @xmath179 . for the other case where @xmath180 , all these @xmath5 violation primed parameters have the opposite overall sign .
the signs of other helicity parameters are not changed . in fig .
7 are plots of the signatures for a partially - hidden @xmath5 violation associated with a @xmath48 phase - type ambiguity . as above for the analogous @xmath47 construction
, the additional @xmath48 coupling @xmath181 now has an effective mass scale parameter @xmath182 in which @xmath146 varies with the mass scale @xmath183 to maintain standard model values in the massless b - quark limit for the moduli parameters @xmath184 and @xmath137 . for @xmath185 ,
we require @xmath186 so for @xmath187 the relationship giving @xmath188 is @xmath189 for @xmath190 which correspond respectively to @xmath191 . for the case
@xmath192 , in fig.8 the upper plot shows by the solid curve @xmath176 versus @xmath193 . by the dashed curve
, it shows @xmath171 . at the endpoint @xmath194 , or @xmath195 . here , as in fig
. 7 , where @xmath176 has the maximum deviation , there is a zero in @xmath171 .
the lower plot shows the indirect effect of such a coupling on the partial width @xmath68 .
while @xmath196 varies , two of the relative phases remain almost fixed , @xmath197 respectively , so only one independent relative phase could be viewed as driving the variation , e.g. @xmath26 varies from @xmath198 to zero . at the maximum of @xmath176 , @xmath199 .
the curve for @xmath200 has the same shape and is also maxmimum at the same position with a value @xmath201 .
@xmath202 remains very small .
for the other case , @xmath180 , each of these @xmath15 violation primed parameters has the opposite overall sign . since sign@xmath203 sign@xmath204sign@xmath205 and sign@xmath206 sign@xmath207sign@xmath205 , all the relative phases change sign for the case @xmath208 . in summary ,
sufficiently precise measurement of the w - interference parameters @xmath209 and @xmath176 can exclude such partially - hidden @xmath15 violation associated with either of the two dynamical phase - type ambiguities .
however , if @xmath210 were found to be smaller than expected or with a negative sign , such a measurement would be consistent with this type of @xmath15 violation .
333 c.a . nelson and a.m.
cohen , eur .
j. * c8 * , 393(1999 ) . in the case of the @xmath47 ambiguity ,
table 2 should list @xmath211 , i.e. with a positive sign .
the cdf collaboration reported that the fraction of longitudinal w bosons is @xmath212 assuming a pure @xmath64 coupling , t. affolder , et.al .
, phys.rev.lett . *
84 * , 216(2000 ) ; the d collaboration reported @xmath213 spin - correlation results in s. snyder , fermilab - conf-99 - 294-e , hep - ex/9910029 , to appear in proc .
eps - hep 99 .
m.beneke , _ et .
al . _ , hep - ph/0003033 , to appear in 1999 cern workshop on sm physics ( and more ) at the lhc " .
d. atwood , s. bar - shalom , g. eilam , and a. soni , hep - ph/0006032 .
reports from a top quark workshop at fermilab include g. mahlon hep - ph/9811219 , hep - ph/9811281 ; s.s .
willenbrock , hep - ph/9905498 .
w. bernreuther , a. brandenburg , and z.g .
si , phys.lett . * b483 * , 99(2000 ) ; w. bernreuther , a. brandenburg , and m. flesch , hep - ph/9812387 . m. fischer , s. groote , j.g .
krner , m.c .
mauser , b. lampe , phys.lett . *
b451 * , 406(1999 ) ; b. lampe , hep - ph/9801346 . b. grzadkowski , z. hioki , phys.rev . *
d61 * , 014013(2000 ) ; hep - ph/0003294 .
g. mahlon and s. parke , phys.lett .
* b476 * , 323(2000 ) ; hep - ph/0001201 ; j. kodaira , t. nasuno , s. parke , hep - ph/9807209 . t. tait , phys . rev . *
d61 * , 0340001(2000 ) ; f. larios and c .- p .
yuan , phys.lett . * b457 * , 334(1999 )
; t. tait and c .- p .
yuan , hep - ph/9710372 .
y. kiyo , j. kodaira , k. morii , t. nasuno , and s. parke , hep - ph/0006021 . j. guasch , w. hollik , j. i. illana , c. schappacher , and j. sola , hep - ph/0003109 .
e. boos , m. dubinin , m. sachwitz , and h.j .
schreiber , hep - ph/0001048 ; a.belyaev and e. boos , hep - ph/0003260 ; a.belyaev , hep - ph/0007058 .
m. jack , a. hoefer , a. leike , and t. riemann , hep - ph/0007046 .
nelson , b.t .
kress , m. lopes , and t.p .
mccauley , phys .
* d56 * , 5928(1997 ) ; * d57 * , 5923(1998 ) .
table 1 : for the standard model and at the ambiguous moduli points , numerical values of the associated helicity amplitudes @xmath214 .
the values for the amplitudes are listed first in @xmath215 units , and second as @xmath216 which removes the effect of the differing partial width , @xmath217 for @xmath94 .
[ @xmath218 ] .
1 : the first * five sets of figures * are for the case of a single additional pure - imaginary coupling , @xmath2 or @xmath81 , associated with a specific additional lorentz structure , @xmath82 .
the two plots displayed here are for dimensional couplings with chiral @xmath219 and non - chiral @xmath101 lorentz structures .
the * upper plot * displays the @xmath176 helicity parameter versus the effective - mass scale @xmath89 with @xmath90 in @xmath51 units . the * lower plot * displays the induced effect of the additional coupling on the partial width for @xmath220 .
the standard model(sm ) limit is at the wings " where @xmath221 .
2 : plots of the b - polarimetry interference parameters @xmath99 and @xmath100 versus @xmath222 for the case of a single additional @xmath101 and @xmath102 coupling .
curves are omitted in the plots in this paper when the couplings produce approximately zero deviations in the helicity parameter of interest .
3 : for a single additional gauge - type coupling @xmath123 or @xmath124 , plots of the b - polarimetry interference parameters @xmath99 and @xmath100 , and of the partial width for @xmath94 versus pure - imaginary coupling constant @xmath81 .
the @xmath125 value is in @xmath51 units .
the sm model limit is at the origin , @xmath126 .
4 : these plots show the indirect effects of a single additional pure - imaginary chiral coupling , @xmath2 or @xmath81 , on other helicity parameters . a dark rectangle denotes the value for the sm .
for the coupling strength ranges listed in the middle table " , the * upper plot * shows the indirect effects on the probabilities @xmath130 and @xmath9 .
each curve is parametrized by the magnitude of the associated @xmath132 or @xmath133 . on each curve
, the central open circle corresponds to the region with a maximum direct @xmath15 violation signature .
the dark / light solid circles correspond respectively to the ends of the ranges listed in the middle table where the direct signatures fall to about @xmath135 of their maximum values . the * lower plot * is for the w - polarimetry interference parameters @xmath223 . fig
5 : these plots show the indirect effects of a single additional pure - imaginary non - chiral coupling on other helicity parameters . versus the middle table given here ,
the curves are labeled as in fig .
the * upper plot * is for the two probabilities @xmath130 and @xmath9 .
the * lower plot * is for @xmath223 .
fig . 6 : plots of the signatures for a partially - hidden @xmath5 violation ( see text ) associated with a @xmath47 phase - type ambiguity .
plotted versus @xmath224 for the case @xmath225 is the solid curve for @xmath176 , and the dashed curve for @xmath171 which are degenerate .
7 : plots of the signatures for a partially - hidden @xmath5 violation ( see text ) associated with a @xmath48 phase - type ambiguity . versus @xmath226 for @xmath225 , the * upper plot * shows by the solid curve @xmath176 , and by the dashed curve @xmath227 .
the * lower plot * shows the indirect effect of such a coupling on the partial width . | to enable an evaluation of future measurements of the helicity parameters for @xmath0 decay in regard to @xmath1 violation , this paper considers the effects of an additional pure - imaginary coupling , @xmath2 or @xmath3 , associated with a specific , single additional lorentz structure , @xmath4 .
sizable @xmath5 violation signatures can occur for low - effective mass scales ( @xmath6 ) , but in most cases can be more simply excluded by @xmath7 precision measurement of the probabilities @xmath8 and @xmath9 .
signatures for excluding the presence of @xmath1 violation associated with the two dynamical phase - type ambiguities are investigated .
= -5 mm * on measurement of helicity parameters in top quark decay * + charles a. nelson and l. j. adler , jr .
+ _ department of physics , state university of new york at binghamton + binghamton , n.y .
13902 - 6016 _ + |
determination of the proton and neutron densities , their root - mean - square radii , @xmath15 and @xmath7 , and the neutron skin thickness , @xmath8 = @xmath7 - @xmath15 , are critical to understanding many of the bulk properties of matter @xcite .
_ have pointed out that there are substantial disagreement between theoretical values of @xmath8 @xcite .
furnstahl s recent analysis of neutron radii in the framework of mean - field models shows that relativistic mean - field models overestimate the values of @xmath8 @xcite .
parity violation electron scattering may provide the experimental data to resolve the differences in the theoretical values @xcite , but is also useful to have alternative methods of obtaining neutron densities .
reliable neutron densities are needed for atomic parity violation experiments @xcite , the analysis of antiprotonic atoms @xcite , in understanding the surface crust of neutron stars @xcite , and in extrapolation to proton - rich or neutron - rich nuclei that is important in nuclear astrophysics @xcite . in this work
, we revisit the analysis of medium energy proton - nucleus elastic scattering data with the goal of obtaining reliable , energy independent neutron densities and the values of @xmath7 and @xmath8 .
the analysis of elastic electron scattering which has resulted in reliable ground state charge densities has been a guiding light for our work @xcite . for a number of years we have used the relativistic impulse approximation ( ria ) in the analysis of proton - nucleus elastic and inelastic scattering @xcite and the ria - kdp(kemmer - duffin - petiau ) @xcite for meson - nucleus elastic scattering @xcite .
these approaches produce relativistic optical potentials which result in good agreement with medium - energy scattering observables .
the input to these calculations are the relativistic densities from quantum hadrodynamics ( qhd ) @xcite and the elementary nn amplitudes from arndt _ et al .
_ @xcite . in recent work
we use the modern eft densities @xcite . the seminal analysis of proton - nucleus elastic scattering data done by ray and hoffmann used both the ria and the non - relativistic kmt approach in their fits to get the observables from 300 mev to 1040 mev @xcite .
unfortunately neither approach produced energy - independent neutron densities .
in addition , some of the values of @xmath8 for @xmath1ca and @xmath2pb were negative , in contradiction to all nuclear structure calculations .
shlomo and schaffer used the results from an analyses of 1 gev proton elastic scattering from @xmath0ca and @xmath1ca to obtain the skin thickness for @xmath0ca , @xmath16ca , @xmath17ca and @xmath1ca , see table 2 in @xcite .
starodubsky and hintz extracted the neutron densities from elastic proton scattering of @xmath18pb at 650 mev and obtained @xmath8 of @xmath19 fm for @xmath2pb @xcite .
however , the energy - independence of the neutron densities in the work of ref .
@xcite or ref .
@xcite was not addressed .
recently karataglidis _ et al .
_ have calculated proton and neutron elastic scattering from @xmath2pb and @xmath0ca targets at three energies 40 mev , 65 mev and 200 mev .
they used a model based on coordinate space nonlocal optical potentials using a full folding nn interactions with various skyrme model ground - state densities @xcite .
for @xmath2pb they found that the skm * model gave the best agreement with proton and neutron elastic scattering data at 40 mev , 65 mev and 200 mev .
based on this the authors suggest that @xmath8 for @xmath0ca is @xmath20 -0.05 fm and @xmath2pb is @xmath20 0.17 fm @xcite . in this paper , a new analysis of proton - nucleus elastic scattering
is used to obtain the neutron density .
this work is motivated by our considerable experience in obtaining high quality global proton - nucleus optical potentials from 20 mev to 1040 mev @xcite .
the new method meshes the global approach with the ria and proves to be successful in obtaining energy independent neutron densities .
the starting point is the ria in its simplest form , which for spin - zero nuclei includes only scalar , vector and tensor terms .
the tensor term is very small and is excluded ; as was done in the ria analysis done by ray and hoffmann @xcite .
we find that substantial progress in extracting the neutron densities from proton - nucleus elastic scattering is made by using a global approach focusing on the energy region where the ria is capable of reproducing experiment very well .
we have obtained values of @xmath8 for @xmath0ca , @xmath1ca , and @xmath2pb which agree with nonrelativistic skyrme models @xcite and relativistic hartree - bogoliubov model extended to include density dependent meson - nucleon couplings @xcite .
our results for @xmath1ca and @xmath2pb are generally not in agreement with relativistic mean - field model , see ref .
[ 6 ] and references therein .
the organization of this paper is as follows .
section [ sec : glob ] describes the global method to obtain the neutron density .
section [ sec : results ] discusses the results and the sensitivity of the extracted neutron density , @xmath7 and @xmath8 to the input used in the fitting procedure .
the values of @xmath8 and the neutron rms radii @xmath7 for @xmath0ca , @xmath1ca , and @xmath2pb for the various tests of the input to the model are given in this section .
the summary and conclusion is in sect .
[ sec : sumcon ] .
the ria nuclear reaction formalism is used as the basis of global fits to medium - energy proton nucleus elastic scattering data . the input to the ria consists of the arndt nn amplitudes @xcite and the point proton density , which is fixed from the charge distribution obtained from electron nucleus scattering .
the neutron vector density , and the scalar proton and neutron densities are parametrized , resulting in good fits of p + a elastic scattering data between 500 mev and 1040 mev .
using the ria as the basis for the global fits is a new approach and our results shows that it is a valid method for extracting neutron densities , @xmath7 and @xmath8 . in the global approach used in this work the form of the ria vector and scalar and
tensor optical potentials are given by @xmath21,\end{aligned}\ ] ] where the fourier transforms of the density form factors are @xmath22 the subscripts s , v , t refer to lorentz scalar , vector ( time - like ) , and tensor quantities .
the superscript n and p refer to neutrons and protons , @xmath23 are the invariant nn amplitudes , and @xmath24 is the kinematical factor required to obtain the invariant nn amplitude in the breit frame @xcite . the value of the upper limit on the momentum transfer , @xmath25 , is determined by the available on shell nn data .
many studies conducted have demonstrated that higher order corrections to this first order ria - dirac optical model approach are negligible at the energies being studied in this paper .
_ for more details of the ria - dirac calculations used for elastic scattering observables @xcite .
elastic p + a data between 500 mev and 1040 mev for @xmath0ca , @xmath1ca and @xmath2pb form the data set .
the quality of the fits are good and the predictions of data not in the data set is used to verify the procedure . however , unlike the usual ria , where we have generally used scalar , vector and tensor densities from the results of relativistic effective field theory ( eft ) calculations , we use the ria as a basis for extracting the vector neutron density as well as the two scalar densities using a global fitting procedure .
next we describe the treatment of the vector and scalar densities .
the point proton density is fixed by using the results from electron scattering as follows , @xmath26 where , @xmath27 here @xmath28 is the experimental charge density , @xmath29 is the proton form factor .
the point proton density , @xmath30 is normalized to @xmath31 and the neutron density @xmath32 is normalized to @xmath33 .
the next step is to consider the model that will be used .
after several years of testing a number of different forms used to parametrize the neutron vector and scalar densities , we have chosen to model the vector neutron density and the scalar proton and neutron densities using the cosh form ( cosh ) used in many of our global fits @xcite .
we have found the cosh form parametrization produced more stable results for @xmath34 and @xmath8 than the three parameter fermi ( 3pf ) or the sum of gaussian ( sog ) parametrizations .
this stability of the cosh was understood from the results from many global fits using the all three forms . in global fits
we used four or five different momentum transfer ranges .
for example , if five momentum transfer ranges are used the fits are done from 0.0 @xmath35 to 1.5 @xmath35 to 0.0 @xmath35 to 3.5 @xmath35 in steps of 0.5 @xmath35 . in future work we will consider other parametrizations .
the form of the cosh model in this work the superscript v stands for volume and s for surface terms , @xmath36 - 1\}}{\{\cosh[r / a]+\cosh[r / a]-2\}}\\[8pt ] f^{s}(r , r , a ) & = & \frac{\{\cosh[r / a]-1\}\{\cosh[r / a]-1\}}{\{\cosh[r / a]+ \cosh[r / a]-2\}^2}.\end{aligned}\ ] ] the vector neutron density is @xmath37 the density @xmath38 is given by @xmath39 and @xmath38 is normalized to @xmath40 .
there are two geometric parameters , @xmath41 and @xmath42 , and the parameter , @xmath43 .
the scalar proton and neutron densities are , @xmath44 and @xmath45 each of these densities contains three parameters , two geometry parameters , @xmath46 , @xmath47 , @xmath48 , @xmath49 and the parameters @xmath50 and @xmath51 .
the 10th parameter , @xmath52 , searched is given by @xmath53 = @xmath52 z. a similar 11th parameter given by @xmath54 = @xmath55
n could also have been searched . however , the searches were more stable if the ratio @xmath56 , _
i.e. _ , the ratio of @xmath57 , was fixed by ratio of the volume integrals per particle of the scalar proton and neutron densities from any one of the eft models given by rusnak _
three different eft densities are used , two are point coupling models , va3 and fz4 , and one is a meson model , ma4 .
the parameter @xmath52 is not sensitive to the eft model chosen and the final difference and the scalar proton and neutron densities are not sensitive to the eft model chosen even though the values of @xmath8 for n @xmath58 z in these models differ widely .
for example , values of @xmath8 for @xmath2pb are 0.332 @xmath59 for va3 , 0.259 @xmath59 for ma4 and 0.160 @xmath59 for fz4 .
as mentioned above we found that the model used in this paper was the best of many different models we tried .
we will investigate other models in future work using this global approach .
it has long been known that using a global approach has been very useful in obtaining nn amplitudes and the nn phase shifts .
this is one of the reasons for using a global approach when the data set used is large and usually correlated . in this work we use elastic p + a data between 500 mev and 1040 mev for @xmath0ca , @xmath1ca and @xmath2pb . for @xmath0ca
there are five energies ; 497.5 mev @xcite , 613 mev @xcite , 650 mev @xcite , 797.5 mev @xcite , and 1040 mev @xcite . for @xmath2pb
there are five energies ; 497.5 mev @xcite , 613 mev @xcite , 650 mev @xcite , 797.5 mev @xcite , and 1040 mev @xcite .
however for @xmath1ca only three energies ; 497.5 mev @xcite , 797.5 mev @xcite , and 1044 mev @xcite are available . in order to make predictions of data not included in the data set we remove one energy . for @xmath0ca and @xmath2pb the 650 mev data has been excluded , and for @xmath1ca 1044 mev data has been excluded .
the quality of the global fits to the data are good and the predictions for data not in the global data sets verify the procedure . in previous dirac phenomenological ( dp )
global fitting , the data sets were cut at 100 degrees in the center of mass or at momentum transfer q at 3.0 @xmath35 , whichever came first . in this work
we do global fits using a variety of momentum transfer values as discussed below .
in addition , as is discussed later , we also test the sensitivity of the input to the model . in the global fitting
we have used five momentum transfer ranges from 0.0 @xmath35 to 1.5 @xmath35 to 0.0 @xmath35 to 3.5 @xmath35 in steps of 0.5 @xmath35 . or four momentum transfer ranges from 0.0 @xmath35 to 2.0 @xmath35 to 0.0 @xmath35 to 3.5 @xmath35 again in steps of 0.5 @xmath35 .
the values of @xmath8 and @xmath7 are expected to change with the value of the momentum transfer range used as the data sets are changed .
we find that the values of @xmath8 and @xmath7 for a given momentum transfer range ( we use five momentum transfer ranges or four momentum transfer ranges ) are clustered in a reasonably well constrained set of values for @xmath8 and @xmath7 , we call these stable results . the momentum transfer range for 0.0 @xmath35 to 1.5
@xmath35 is almost always the outlier , which is understandable as the data set is small and does not have the diffraction structure needed when fitting proton scattering from nuclear targets .
of course in any such global approach the data sets are usually correlated .
this is certainly true as the data sets sets with different momentum transfer ranges do overlap .
thus a statistical analysis to obtain a mean value and a standard deviation can not generally be used .
however , we can determine the range of the values of @xmath8 and @xmath7 for each fit from every momentum transfer range , and obtain the values for @xmath8 and @xmath7 which could be shown as a scatter plot , as is usually done when data sets used are correlated .
we have obtained the range of values for @xmath8 and @xmath7 for every test considered , while these ranges can not be interpreted as resulting from a statistical analysis we found that doing such an analysis gives a useful guide to present the results .
there are a number of features in the fitting procedures which could produce changes in the neutron density , @xmath7 and @xmath8 .
as mentioned above , two ranges of momentum transfer cuts are used . in the following sensitivity tests we use both ranges .
generally the smaller range produces the smallest range of values for @xmath7 and @xmath8 so in most of the figures we show the larger range values for @xmath7 and @xmath8 which gives the most conservative results .
we test the results obtained when the following features of the model used in the fitting procedures are changed .
first the three different eft models used in fixing the 10th parameter are considered @xcite . to investigate the effect of the @xmath29 used in obtaining the point proton density we use two different forms they are identified as @xmath60 from @xcite and @xmath61 from @xcite .
the set chosen for the arndt nn amplitudes is input to the fitting procedure , and we use sets fa00 and sm86 to find the sensitivity of the fits to this input @xcite . the data sets included in the fits have been changed to see if the results change significantly .
finally , we use the three different charge density models used in the analysis of electron - nucleus elastic scattering data ; the three parameter fermi ( 3pf ) , the sum of gaussian ( sog ) and the fourier - bessel ( fb ) @xcite .
thus each target we have done global fits which have used effective field theory densities ma4 , fz4 and va3 , and for each three charge distributions are used .
in addition two different form factors @xmath60 and @xmath61 and two sets of arndt amplitudes , set fa00 and set sm86 . figures 1 - 5 show the results of these tests of the global fitting procedure . the results for the @xmath7 and @xmath8 are given in three tables , found in the appendix ; table i for @xmath0ca , table ii for @xmath1ca and table iii for @xmath2pb .
there are three general cases we use in testing the global fits : case 1 uses arndt amplitudes set fa00 and the @xmath60 form factor ; case 2 uses arndt amplitudes set fa00 and the @xmath61 form factor ; and case 3 uses arndt amplitudes set sm86 and the @xmath60 form factor . in every case , we obtain good global fits for ma4 , fz4 and va3 for every charge distribution used ; sog , 3pf , and fb . as discussed above the momentum cuts ranges used
are 0.0 @xmath35 to 1.5 @xmath35 to 0.0 @xmath35 to 3.5 @xmath35 in steps of 0.5 @xmath35 or we remove the 0.0 @xmath35 to 1.5 @xmath35 set . for each case
the one with five momentum transfer ranges or four momentum transfer ranges we calculate the mean and standard deviation for @xmath7 and @xmath8 for every global fit , this gives us an average over all eft models for a given charge distribution .
this is denoted by @xmath62 in the tables .
this is not to be taken as a statistical error , it is however a convenient way to show the range of the values in a consistent way .
we do this rather than remove every outlier . then for each case we show the combined values of the range for @xmath7 and @xmath8 for that case , this gives the average over all charge distributions as well as over all eft models .
these results are denoted as @xmath63 and @xmath62 in the tables .
finally , we calculate the ranges of @xmath7 and @xmath8 for all three cases combined for each target .
the last calculation , which combines all of the values of @xmath7 and @xmath8 for every test made , gives us the most conservative results . from these final test cases , shown at the bottom of the tables in the appendix
, we get the following values using the five momentum transfer ranges for @xmath0ca , @xmath64 fm and @xmath65 fm ; for @xmath1ca , @xmath66 fm and @xmath67 fm ; for @xmath2pb , @xmath68 fm and @xmath69 fm . as mentioned earlier if we use the four momentum transfer ranges results are generally more clustered and we obtain : for @xmath0ca , @xmath70 fm and @xmath71 fm ; for @xmath1ca , @xmath72 fm and @xmath73 fm ; and for @xmath2pb , @xmath74 fm and @xmath75 fm . the values for @xmath7 and @xmath8 for both of the final case results are quite similar .
however , the authors found that the global fits using the arndt nn amplitude set fa00 and the form factor @xmath60 gave the most stable results _
i.e. _ the values did not vary very much , see tables i , ii and iii in the appendix .
this is especially true for using the sog charge distribution . in this case
the values of @xmath7 and @xmath8 using the five momentum transfer ranges are : for @xmath0ca , @xmath76 fm and @xmath77 fm ; for @xmath1ca , @xmath78 fm and @xmath79 fm ; and for @xmath2pb , @xmath80 fm and @xmath81 fm .
the values for @xmath7 and @xmath8 using the four momentum transfer ranges are : for @xmath0ca , @xmath9 fm and @xmath10 fm ; for @xmath1ca , @xmath11 fm and @xmath12 fm ; and for @xmath2pb , @xmath13 fm and @xmath14 fm .
the results using the form factor @xmath61 are quite similar as well as the case using form factor @xmath60 but with the nn set sm86 .
these values as well as the values using the obtained from the final case values are in agreement with nonrelativistic skyrme models @xcite , the relativistic hartree - bogoliubov model extended to include density dependent meson - nucleon couplings @xcite and the recent analysis of antiproton atoms @xcite .
our results are generally not in agreement with relativistic mean - field models .
next we discuss the results of each test of the model separately .
figure [ fig : diffsog1 ] shows the effect of the two different momentum transfer ranges as well as the difference in the three eft models used to fix parameter 10 .
the filled boxes ( the smaller momentum transfer range ) and diamonds ( the larger momentum transfer range ) are the values from the three different eft models used .
the charge density used in this case was the sog for all targets , the proton form factor @xmath60 , and the set fa00 arndt nn amplitudes .
this figure shows that the results from the two different momentum ranges overlap and that the eft model used does not affect the results .
the figure also shows several theoretical results from refs .
the comparison between theory and range of the values from the fits shows that all theoretical @xmath8 values agree with the results of the global fits for @xmath0ca .
however , the theoretical @xmath8 values for @xmath1ca and @xmath2pb are somewhat larger than the global fits .
ca , @xmath1ca and @xmath2pb from the global fitting procedure shows the @xmath62 models values for the sog charge distribution , set fa00 arndt nn amplitudes , the five momentum transfer ranges .
we compare @xmath60 from ref .
@xcite shown as filled boxes and @xmath61 from ref .
@xcite shown as filled diamonds .
the same theoretical values for @xmath8 shown in fig .
[ fig : diffsog1 ] are also shown.,width=441 ] ca , @xmath1ca and @xmath2pb from the global fitting procedure shows the @xmath62 models values for the sog charge distribution , set fa00 arndt nn amplitudes , the five momentum transfer ranges .
we compare @xmath60 from ref .
@xcite shown as filled boxes and @xmath61 from ref .
@xcite shown as filled diamonds .
the same theoretical values for @xmath8 shown in fig .
[ fig : diffsog1 ] are also shown.,width=441 ] next we consider the results when two different proton form factors , @xmath60 from ref .
@xcite and @xmath61 from ref .
@xcite , are used in obtaining the vector point proton density for a given charge distribution model .
the five momentum transfer ranges , the values for the @xmath62 models for the sog model charge distribution and set fa00 arndt nn amplitudes are used and shown in figure [ fig : diffdpoff1 ] . investigating the sensitivity to the set of arndt nn amplitudes used
is done by comparing sets fa00 and sm86 . the larger momentum transfer range , the averaged eft models for the sog model charge distribution , and the proton form factor @xmath60 are used and shown in figure [ fig : difffa00sm86 ] . in order to check the sensitivity due to the data sets included in the fit we have done fits using only two data sets ( 497.5 mev and 797.5 mev ) for @xmath0ca and @xmath2pb .
the results agree very well with the same case using four data sets as is shown in fig .
[ fig : difftwosets ] .
the values of @xmath8 for the two data set case are for @xmath0ca , @xmath82 fm ; and for @xmath2pb @xmath83 fm .
these results agree very well with the values for the four data set case @xmath0ca , @xmath77 fm and @xmath2pb @xmath81 fm .
this encourages us to extend our global procedure to nuclei that have at least two data sets in the medium energy range .
ca , @xmath1ca and @xmath2pb when different data sets included in the fits , one has four energies and the other has two energies .
the results of the global procedure uses the @xmath62 models and the sog charge distribution , the five momentum transfer ranges , the proton form factor @xmath60 , set fa00 arndt nn amplitudes .
the results for four - energy sets are shown by filled boxes and the results for two - energy data sets by filled diamonds .
the data set for @xmath1ca has only three energies so only has two data sets are used in the fits .
the same theoretical values for @xmath8 shown in fig .
[ fig : diffsog1 ] are also shown.,width=422 ] ca , @xmath1ca and @xmath2pb when different data sets included in the fits , one has four energies and the other has two energies .
the results of the global procedure uses the @xmath62 models and the sog charge distribution , the five momentum transfer ranges , the proton form factor @xmath60 , set fa00 arndt nn amplitudes .
the results for four - energy sets are shown by filled boxes and the results for two - energy data sets by filled diamonds .
the data set for @xmath1ca has only three energies so only has two data sets are used in the fits .
the same theoretical values for @xmath8 shown in fig .
[ fig : diffsog1 ] are also shown.,width=422 ] the sensitivity of @xmath8 and @xmath7 to the three different charge distributions , three parameter fermi ( 3pf ) , sum of gaussian ( sog ) and fourier - bessel ( fb ) obtained from ref .
@xcite which are used in the fitting procedure is shown in fig .
[ fig : diffcd1 ] and given in detail in tables i , ii and iii .
the five momentum transfer ranges , the proton form factor @xmath60 and the set fa00 arndt nn amplitudes are used . for @xmath2pb and @xmath1ca there is little difference in the values of @xmath8 and @xmath7 for the different charge densities .
we note that the rms radii for @xmath2pb and @xmath1ca for these three charge density are about the same @xcite .
two of the three @xmath0ca charge distributions also have almost identical charge rms radii , 3pf 3.482(25)fm , sog and 3.479(3)fm but the rms radius for the fb , 3.450(10)fm , is considerably smaller ( see ref .
the result is that for @xmath0ca the difference in the value of @xmath8 is pronounced . as shown in fig .
[ fig : diffcd1 ] the fb charge distribution has range of @xmath8 values that goes from positive to negative but the range of the @xmath8 values 3pf and sog are all negative . we attribute this to the smaller rms radius for the fb charge distribution .
the values of @xmath15 , @xmath7 and @xmath8 should not depend heavily on the momentum transfer ranges , the set of amplitudes , the form factors , and the charge distribution .
in fact , the only case that does not overlap is the fb case for @xmath0ca .
the results for all three of the charge distributions are given in tables i - iii in the appendix .
ca , @xmath1ca and @xmath2pb due to different charge distribution models .
the global fits use set fa00 arndt nn amplitudes , the five momentum transfer ranges , and the proton from factor @xmath60 .
the results for the @xmath62 sog are shown by filled boxes , the @xmath62 3pf are shown by filled diamonds and the @xmath62 fb are shown by filled circles .
the results for all effective field theory cases and for all three charge distributions _
i.e. _ the results from the final cases are given by the open squares .
the same theoretical values for @xmath8 shown in fig.[fig : diffsog1 ] are also shown.,width=441 ] in this global analysis we obtain many densities from the various cases used in checking the results of the fits .
figures 6 and 7 show the vector point proton density distribution , @xmath3 and @xmath84 and the vector point neutron density distribution , @xmath4 and @xmath85 for @xmath0ca , @xmath1ca , and @xmath2pb . the sog charge distribution and the @xmath60 form factor were used to obtain the vector point proton density .
the neutron density also depends on the charge distribution chosen and the form factor used to obtain the vector point proton density as well as the other input to a given global fit , the nn amplitudes , and the eft density used to fix the ratio of @xmath57 . in figs .
6 and 7 the arndt nn amplitude set fa00 , the sog charge distribution , the momentum transfer range was 0.0 @xmath35 to 3.0 @xmath35 , proton from factor @xmath60 , and all three eft cases , ma4 , fz4 and va3 are used .
it is clear that the neutron densities overlap showing that the use of different eft cases does not have any significant impact on the global fit .
tables of the densities and the parameters are available from the authors @xcite . this same input as in figs 6 - 7 is also used in the figs .
8 - 9 which shows the vector proton point density and the scalar proton @xmath5 and @xmath86 , and in figs
. 10 - 11 for the vector point density and the scalar neutron @xmath6 and @xmath87 .
figures 8 - 11 show that the phenomenology gives sensible scalar densities that are not unphysical .
figures 12 and 13 show the observables for one of the energies included in the global fit for @xmath0ca and a prediction of an energy not included in the global fit .
figure 12 shows the fit for @xmath0ca at 497.5 mev and fig .
13 shows the prediction for the @xmath0ca 650 mev spin observables ; no cross section data is available .
figures 14 and 15 show the observables for one of the energies included in the global fit for @xmath1ca and the results from a prediction of an energy not included in the global fit .
figure 14 shows the fit for @xmath1ca at 497.5 mev and the prediction of q at that energy .
figure 15 shows the prediction for @xmath1ca at 1044 mev .
figures 16 and 17 show the observables for one of the energies included in the global fit for @xmath2pb and the results from a prediction of an energy not included in the global fit .
figure 16 shows the fit for @xmath2pb at 497.5 mev and fig .
17 shows the prediction for the 650 mev observables for this target .
the observables shown in figs
. 12 - 17 all used the arndt nn amplitude set fa00 , the sog charge distribution , the proton from factor @xmath60 , the momentum transfer range from 0.0 @xmath35 to 3.0 @xmath35 , and the eft case ma4 .
while the figures of the observable are small we have magnified them and we find that the heights of the diffractive maxima and the angular positions of the minima and maxima are very well reproduced at each energy with no systematic energy dependent discrepancies .
these are the most critical features of the data which determine rms radii .
in fact , precision fits to @xmath88 and q are not as important for determining neutron radii but our fits are quite reasonable .
the magnified figures are available from the authors @xcite .
ca at 497.5 mev.,width=480 ] ca at 650 mev.,width=480 ] ca at 497.5 mev.,width=480 ] ca at 1044 mev.,width=480 ] pb at 497.5 mev.,width=480 ] pb at 650 mev , width=480 ] we also predict the total neutron cross section and the proton reaction cross section . in fig . [
fig : ncapbxs ] the predicted total neutron cross sections for @xmath0ca and @xmath2pb is compared with the experimental values from finlay _
the predicted proton reaction cross sections for the same two targets are given in fig .
[ fig : reactxs ] with the experimental values @xcite and the predictions are reasonable .
ca and @xmath2pb shown as circles are compared with the experimental values shown as diamonds from ref .
@xcite.,width=489 ] ca and @xmath2pb compared with the experimental values shown as diamonds from ref .
the dashed lines are the results of the edai global fit and the dotted lines are the edad fit 3 from ref .
@xcite.,width=489 ] all of the results confirm that the global approach produces good fits to the data , that predictions are quite acceptable and that the neutron densities are energy independent .
these densities should be useful input to a large number of nuclear reactions .
they provide the empirical values of @xmath7 and @xmath8 that are needed for testing theoretical models .
this paper reports a new method for extracting neutron densities from intermediate energy elastic proton - nucleus scattering observables .
neutron densities are needed for atomic parity violation experiments , the analysis of antiproton atoms , the experiments parity violation elastic electron scattering , and theoretical nonrelativistic and covariant mean - field models .
it is interesting to note that our results for @xmath8 are in agreement with work using antiprotonic atoms , see fig . 4 in ref .
the approach uses a global analysis , similar to the global fits using a dirac phenomenology , but in this case it is based on the relativistic impulse approximation ( ria ) .
the input to the procedure are the vector point proton density distribution , @xmath3 , which is determined from the empirical charge density after unfolding the proton form factor and the arndt nn amplitudes .
the other densities , @xmath4 , @xmath5 , @xmath6 , are parametrized .
the neutron densities for @xmath0ca , @xmath1ca and @xmath2pb obtained from the global fits have been tested for sensitivity to the input .
different proton form factors , different charge densities , different sets of the arndt nn amplitudes , different eft models , different data sets included in the fits , and differences in the ranges of the momentum transfer used in the fits have been investigated . with the exception of the fourier - bessel charge density for @xmath0ca
all of these tests produce values for @xmath7 and @xmath8 that overlap .
the prediction of data sets not used in the fits are well reproduced and the calculated observables are in good agreement with data . in conclusion , we have obtained values of @xmath7 and @xmath8 for @xmath0ca , @xmath1ca , and @xmath2pb .
we have obtain good global fits for ma4 , fz4 and va3 for every charge distribution used ; sog , 3pf , and fb .
both five momentum transfer ranges and four momentum transfer ranges were used .
as discussed earlier , for each case we calculate the mean and standard deviation for @xmath7 and @xmath8 for every global fit , this gives us an average over all eft models for a given charge distribution .
this is denoted by @xmath62 in the tables . as mentioned several times
this is not to be taken as a statistical analysis , but as a way to show the range of the values in a consistent way , they are shown as bars in the figures .
we do this rather than remove outliers .
then for each case we show the combined values of the range for @xmath7 and @xmath8 for that case , this gives the average over all charge distributions as well as over all eft models .
these results are denoted as @xmath63 and @xmath62 in the tables .
we then calculate the ranges of @xmath7 and @xmath8 for all three cases combined for each target . from these final test cases , shown at the bottom of the tables in the appendix , using the five momentum transfer ranges we obtain the following values : for @xmath0ca , @xmath64 fm and @xmath65 fm ; for @xmath1ca , @xmath66 fm and @xmath67 fm ; for @xmath2pb , @xmath68 and @xmath69 fm . in conclusion , the rms neutron radius , @xmath7 and the neutron skin thickness @xmath8 obtained from the global fits that we consider to be the most reliable ranges of our results are given as follows : for @xmath0ca , @xmath9 fm and @xmath10 fm ; for @xmath1ca , @xmath11 fm and @xmath12 fm ; and for @xmath2pb @xmath13 fm and @xmath14 fm .
the authors have found that the global fits are more stable when using the arndt nn amplitude set fa00 rather than the arndt nn amplitude set sm86 .
this might be expected as the fa00 analysis is more recent .
both of the form factors used , @xmath60 and @xmath61 when using fa00 were the most stable , see tables i , ii and iii in the appendix .
the impact on the fits due to the charge distribution input , sog , 3pf and fb , as shown in fig .
[ fig : diffcd1 ] , clearly showed that the cases using the sog and 3pf charge distributions were stable , _
i.e. _ the ranges overlapped , for all targets , however , for @xmath0ca the results for the fb charge distribution gave very different results , see table i. we also found that the fits using arndt nn amplitude set fa00 , the @xmath60 form factor , and the sog charge distribution produced the most stable results .
in this case the values of @xmath7 and @xmath8 using the five momentum transfer ranges are : for @xmath0ca , @xmath76 fm and @xmath77 fm ; for @xmath1ca , @xmath78 fm and @xmath79 fm ; and for @xmath2pb @xmath80 fm and @xmath81 fm .
these values , as well as the values using the most conservative errors and the results that we consider most reliable , are in agreement with nonrelativistic skyrme models @xcite , relativistic hartree - bogoliubov model extended to include density dependent meson - nucleon couplings @xcite and results from a recent analysis of antiprotonic atoms @xcite .
our results for @xmath1ca and @xmath2pb are generally not in agreement with relativistic mean - field models , see ref . 6 and references there in .
the global fit based on the ria is a new tool for obtaining the neutron density .
the values of @xmath7 and @xmath8 obtained are robust .
this quality is verified in several ways as discussed in section iii .
for example , we checked the sensitivity due to the data sets included in the fit by doing global fits using only two data sets ( 497.5 mev and 797.5 mev ) for @xmath0ca and @xmath2pb as well as four data sets .
the values of @xmath8 for the two data set case are : for @xmath0ca @xmath82 fm ; and for @xmath2pb @xmath83 fm .
these results agree well with @xmath8 for the four data set case : for @xmath0ca @xmath77 fm ; and for @xmath2pb @xmath81 fm .
these results , shown in fig . [ fig : difftwosets ] , motivates us to use the global procedure for all nuclei that have at least two data sets in the medium energy range .
this work provides energy independent values for @xmath7 and @xmath8 , in contrast to the energy dependent values obtained by previous studies .
in addition , the results presented in paper show that the expected rms neutron radius and skin thickness for @xmath0ca is accurately reproduced .
the values of @xmath7 and @xmath8 obtained from the global fits that we consider to be the most reliable are given as follows : we plan to extend our work to additional nuclei and will continue to investigate different models and procedures . the goal is to continue to improve quality of neutron densities that result from our global fits .
we thank r. j. furnstahl , c. j. horowitz and b. d. serot for useful discussions .
s. hama also thanks the department of physics , the ohio state university for their support while some of this work was done .
b. c. clark and s. hama thanks the national institute for nuclear theory for their hospitality .
this work was supported in part by the national science foundation under grants no .
phy9800964 and phy0098645 .
r. a. arndt , j. s. hyslop and l. d. roper , phys .
d * 35 * , 127 ( 1987 ) ; richard a. arndt , igor i. strakovsky , ron l. workman , marcello m. pavan , phys .
c * 52 * , 2120 ( 1995 ) ; r.a .
arndt , said program , private communication .
lcc + range ( @xmath89 ) & @xmath7 ( fm ) & @xmath8 ( fm ) + & @xmath90 case 1 & + + & @xmath60 & fa00 + + & sog & @xmath62 + 2.0 - 3.5 & @xmath11 & @xmath12 + 1.5 - 3.5 & @xmath78 & @xmath79 + & 3pf & @xmath62 + 2.0 - 3.5 & @xmath91 & @xmath92 + 1.5 - 3.5 & @xmath93 & @xmath94 + & fb & @xmath62 + 2.0 - 3.5 & @xmath95 & @xmath96 + 1.5 - 3.5 & @xmath97 & @xmath98 + & @xmath63 & @xmath62 + 2.0 - 3.5 & @xmath99 & @xmath100 + 1.5 - 3.5 & @xmath101 & @xmath102 + + & @xmath90 case 2 & + + & @xmath61 & fa00 + + & sog & @xmath62 + 2.0 - 3.5 & @xmath103 & @xmath104 + 1.5 - 3.5 & @xmath105 & @xmath106 + & 3pf & @xmath62 + 2.0 - 3.5 & @xmath107 & @xmath108 + 1.5 - 3.5 & @xmath109 & @xmath110 + & fb & @xmath62 + 2.0 - 3.5 & @xmath111 & @xmath112 + 1.5 - 3.5 & @xmath113 & @xmath114 + & @xmath63 & @xmath62 + 2.0 - 3.5 & @xmath115 & @xmath116 + 1.5 - 3.5 & @xmath117 & @xmath118 + + & @xmath90 case 3 & + + & @xmath60 & sm86 + + & sog & @xmath62 + 2.0 - 3.5 & @xmath119 & @xmath120 + 1.5 - 3.5 & @xmath121 & @xmath122 + & 3pf & @xmath62 + 2.0 - 3.5 & @xmath123 & @xmath124 + 1.5 - 3.5 & @xmath125 & @xmath126 + & fb & @xmath62 + 2.0 - 3.5 & @xmath127 & @xmath128 + 1.5 - 3.5 & @xmath129 & @xmath130 + & @xmath63 & @xmath62 + 2.0 - 3.5 & @xmath131 & @xmath132 + 1.5 - 3.5 & @xmath133 & @xmath134 + + & @xmath90 for all three cases & + + 2.0 - 3.5 & @xmath72 & @xmath73 + 1.5 - 3.5 & @xmath66 & @xmath67 + lcc + range ( @xmath89 ) & @xmath7 ( fm ) & @xmath8 ( fm ) + & @xmath135 case 1 & + + & @xmath60 & fa00 + + & sog & @xmath62 + 2.0 - 3.5 & @xmath13 & @xmath14 + 1.5 - 3.5 & @xmath80 & @xmath81 + & 3pf & @xmath62 + 2.0 - 3.5 & @xmath136 & @xmath137 + 1.5 - 3.5 & @xmath138 & @xmath139 + & fb & @xmath62 + 2.0 - 3.5 & @xmath140 & @xmath141 + 1.5 - 3.5 & @xmath142 & @xmath143 + & @xmath63 & @xmath62 + 2.0 - 3.5 & @xmath144 & @xmath145 + 1.5 - 3.5 & @xmath146 & @xmath147 + + & @xmath135 case 2 & + + & @xmath61 & fa00 + + & sog & @xmath62 + 2.0 - 3.5 & @xmath148 & @xmath149 + 1.5 - 3.5 & @xmath150 & @xmath151 + & 3pf & @xmath62 + 2.0 - 3.5 & @xmath152 & @xmath153 + 1.5 - 3.5 & @xmath154 & @xmath155 + & fb & @xmath62 + 2.0 - 3.5 & @xmath156 & @xmath157 + 1.5 - 3.5 & @xmath158 & @xmath159 + & @xmath63 & @xmath62 + 2.0 - 3.5 & @xmath160 & @xmath161 + 1.5 - 3.5 & @xmath162 & @xmath163 + + & @xmath135 case 3 & + + & @xmath60 & sm86 + + & sog & @xmath62 + 2.0 - 3.5 & @xmath164 & @xmath165 + 1.5 - 3.5 & @xmath166 & @xmath167 + & 3pf & @xmath62 + 2.0 - 3.5 & @xmath168 & @xmath169 + 1.5 - 3.5 & @xmath170 & @xmath171 + & fb & @xmath62 + 2.0 - 3.5 & @xmath172 & @xmath173 + 1.5 - 3.5 & @xmath174 & @xmath175 + & @xmath63 & @xmath62 + 2.0 - 3.5 & @xmath176 & @xmath177 + 1.5 - 3.5 & @xmath178 & @xmath179 + + & @xmath135 for all three cases & + + 2.0 - 3.5 & @xmath74 & @xmath75 + 1.5 - 3.5 & @xmath68 & @xmath69 + | a new method for extracting neutron densities from intermediate energy elastic proton - nucleus scattering observables uses a global dirac phenomenological ( dp ) approach based on the relativistic impulse approximation ( ria ) .
data sets for @xmath0ca , @xmath1ca and @xmath2pb in the energy range from 500 mev to 1040 mev are considered .
the global fits are successful in reproducing the data and in predicting data sets not included in the analysis . using this global approach ,
energy independent neutron densities are obtained .
the vector point proton density distribution , @xmath3 , is determined from the empirical charge density after unfolding the proton form factor .
the other densities , @xmath4 , @xmath5 , @xmath6 , are parametrized .
this work provides energy independent values for the rms neutron radius , @xmath7 and the neutron skin thickness , @xmath8 , in contrast to the energy dependent values obtained by previous studies .
in addition , the results presented in paper show that the expected rms neutron radius and skin thickness for @xmath0ca is accurately reproduced .
the values of @xmath7 and @xmath8 obtained from the global fits that we consider to be the most reliable are given as follows : for @xmath0ca , @xmath9 fm and @xmath10 fm ; for @xmath1ca , @xmath11 fm and @xmath12 fm ; and for @xmath2pb @xmath13 fm and @xmath14 fm .
these values are in reasonable agreement with nonrelativistic skyrme hartree - fock models and with relativistic hartree - bogoliubov models with density - dependent meson - nucleon couplings .
the results from the global fits for @xmath1ca and @xmath2pb are generally not in agreement with the usual relativistic mean - field models . |
many of the f stars cataloged in the _ hipparcos _ as single were recently argued to be in fact unidentified binaries ( suchkov & mcmaster 1999 ) .
the criterion used to isolate binary candidates involves the difference between the absolute magnitude @xmath0 based on the _ hipparcos _ parallax and the absolute magnitude @xmath1 derived from the luminosity index @xmath2 , i.e. , @xmath3 . for the best known nearby single f stars ( @xmath4 pc ) ,
the two magnitudes coincide within @xmath5 mag .
however , most of the unresolved binary stars are anomalously bright for their colors , i.e. , their absolute magnitude , @xmath0 , is substantially brighter than the magnitude derived from @xmath6 : @xmath7 ( for simplicity , we will often be referring to @xmath8 as brightness anomaly and consider a star as anomalously bright if @xmath7 ) . unlike the nearby single stars , many of the distant single f stars , ( @xmath9 pc ) turned out to be anomalously bright as well , similar to the known binary f stars .
the analysis showed that in this case , the brightness anomaly is likely associated with the presence of undetected companions ( suchkov & mcmaster 1999 ) .
these binary candidates were called c binaries ( the follow - up radial - velocity survey of northern c binaries with @xmath10 and @xmath11 pc conducted by griffin & suchkov 2001 has already confirmed that more than @xmath12 of the sample stars actually _ are _ binaries ; orbits for many of these binaries have been determined ) . for unresolved binaries ,
@xmath1 is derived from the color indices originating from a double star combined spectrum . because of that @xmath8 underestimates the luminosity of that star if its components are comparably bright , by up to @xmath13 mag in the case of identical components ( this obviously imposes an upper limit of 0.75 on @xmath8 ) .
however , this seems to be not the only , and perhaps not even the main effect leading to the discrepancy between @xmath1 and @xmath0 .
for one thing , a substantial fraction of binary stars have brightness anomaly well above the indicated upper limit of 0.75 mag .
the stars with @xmath14 , which can not be explained in terms of the combined flux of the unresolved binary components , indicate that along with the combined luminosity something else may impact the absolute magnitude of unresolved binary stars .
more evidence favoring the existence of an additional source for brightness anomaly has come from the study of the relationship between @xmath8 and stellar age ( suchkov 2000 ) .
it turns out that in terms of kinematics and metallicity , anomalously bright binary candidates , c binaries , are , on average , substantially older than the `` truly single '' f stars , i.e. , the stars with normal brightness .
if the anomaly were caused entirely by the luminosity contribution from the undetected secondary , one would have to conclude that binaries with the components having about the same mass live longer or tend to form earlier in the galactic evolution than binaries with disparate components .
but such a scenario does not seem very likely .
a more plausible interpretation of the relationship between @xmath8 and age might be that at least some part of the star s brightness anomaly is associated with a peculiar stellar evolution in the binary system , so that the primary f star in that binary stay within the main sequence longer than a normal single f star .
given the obviously important implications of such a possibility , we have examined a number of relationships between @xmath8 and other parameters of f stars in order to get a deeper insight into the nature of brightness anomaly .
the star sample used in this study originates from the list of @xmath15 _ hipparcos _ stars that have measured colors , with effective temperatures corresponding to the spectral range occupied by f and early g stars ( suchkov & mcmaster 1999 ) .
the data are from hauck & mermilliod ( 1998 ) .
we have computed tangential velocities of these stars from the _ hipparcos _ parallaxes and proper motions .
the _ hipparcos _
parallax has also been utilized to derive absolute magnitude @xmath0 from the johnson @xmath16 magnitude and absolute magnitude @xmath17 from the _ hipparcos _
@xmath18 magnitude .
absolute magnitude @xmath1 has been calculated from the data with the algorithm published by moon ( 1985 ; see also moon & dworetsky 1985 ) ; these calculations involve the dereddened luminosity index , @xmath2 , along with the dereddened metallicity and temperature indices . the algorithm in moon ( 1985 )
has also been used to compute standard @xmath19 from the blanketing - corrected color index @xmath20 , from which effective temperature has been obtained .
metallicity has been calculated from the dereddened color indices , utilizing the calibration given by schuster & nissen ( 1989 ) for f stars .
this calibration is somewhat different from that in carlberg et al .
( 1985 ) , which was used in suchkov ( 2000 ) and suchkov & mcmaster ( 1999 )
. this difference , however , has been found not to impact the results of both the previous and present studies .
the schuster & nissen ( 1989 ) calibration is commonly utilized in current studies , so we will use it in this paper .
age estimates are based on isochrone fitting in the @xmath21 diagram using yale 1996 isochrones ( demarque et al .
age has been used only in terms of mean values obtained by averaging over large groups of stars , and only in differential analysis involving relative ages .
therefore , age accuracy , both in terms of random and systematic errors , is not of much concern for the present study .
the sample has been constrained as follows : @xmath22 ; @xmath23 , k ; @xmath24}$\ } } \leq 0.5 $ ] ; @xmath25 ( @xmath26 is the mean error on @xmath27 for an individual star as given in hauck & mermilliod 1998 ; the sample mean for @xmath26 is @xmath28 times smaller ) ; @xmath29 ( @xmath30 is the mean error on @xmath31 for an individual star ) ; @xmath32 , mas ( error on parallax ) ; @xmath33 , mas ( error on proper motion in ra ) ; @xmath34 , mas ( error on proper motion in dec ) .
brightness anomaly is very conspicuous in unresolved binary stars ( fig .
[ h2541f1 ] ) .
as shown in suchkov & mcmaster ( 1999 ) , the distribution of @xmath8 for the best studied single stars within 25 pc is centered at @xmath35 and is rather narrow , with standard deviation of only 0.15 mag .
the corresponding gaussian ( @xmath36 ) is shown in fig . [ h2541f1 ] . for the unresolved binaries ,
the maximum of the @xmath8 distribution is not only offset from zero by @xmath37 mag but is also much broader . within 200 pc ,
the stars cataloged in the _ hipparcos _ as single have a broad distribution as well , but with the maximum only slightly shifted ( @xmath38 mag ) from that of the single stars within 25 pc .
note that the left tail of the distribution , albeit resembling a segment of a gaussian , is far above the level suggested by @xmath36 , meaning that it does probably not represent the error distribution for @xmath8 .
therefore , the population of anomalously faint stars , @xmath39 , should be as real as that of anomalously bright stars .
as seen below , its distinctive identity is reflected in lowest velocities and highest metallicities among the f stars , which makes it the youngest group of f stars .
therefore , we conclude that in general the distributions of both binary and single stars in fig .
[ h2541f1 ] are most likely dominated by some stellar population physics rather than random or systematic errors . the most striking feature in fig .
[ h2541f1 ] is the positive brightness anomaly of binary stars .
therefore , in the rest of the paper , we will focus primarily on the nature of anomalously bright binary stars .
if the combined flux of unresolved binary components is not the only reason for the enhanced brightness of a binary star , and there is a noticeable contribution from a physical anomaly in the binary s primary star caused by its interaction with the companion , one may expect the brightness anomaly to depend on binary components separation , because the effect of the interaction , whatever it is , is probably stronger in tighter pairs .
we have tested this hypothesis by correlating brightness anomaly and projected separation for two very different , non - overlapping samples of binary f stars within 125 pc , for which the _ hipparcos _ provides angular separation , @xmath40 .
only the stars with @xmath8 in the range @xmath41 to 1.0 have been included , so that the potentially less reliable extreme values of @xmath8 do not impact the results .
the chosen distance limit rejects the most distant stars , thus ensuring better data quality for the remainder of the sample ; at the same time it retains enough stars for statistical tests we are interested in .
the stars in the first sample are binaries discovered by _ hipparcos _
( see mignard et al .
1992 , sderhjelm et al .
1992 for the procedures to detect and measure binary stars on the basis of _ hipparcos _ observations ) . only the binaries with angular separation @xmath42 arcsec have been included ; this constraint rejects only a very small fraction of the _ hipparcos _ binaries and is not of great significance for the test below .
the sample has also been constrained with respect to angular separation error , @xmath43 , to include only the stars with @xmath44 .
the second sample comprises unresolved binaries known prior to _
hipparcos_. with respect to angular separations ,
it has been constrained similar to the sample of the _ hipparcos _ binaries , @xmath42 arcsec .
the pre-_hipparcos _ binaries are much more numerous , which allows us to impose a more stringent constraint on the angular separation accuracy : @xmath45 .
additionally , this sample has been constrained to include only hot stars , @xmath46 k. this should improve chances to detect any separation - dependent evolutionary effects in @xmath8 just because the fraction of evolved stars increases toward higher temperatures ( at cooler temperatures , brightness anomaly in the sample stars may be dominated by the separation - independent effect of the combined flux of binary s unresolved components ) .
projected separation for both samples has been derived from the parallax , @xmath47 , and the angular separation , @xmath40 , given in the _ hipparcos _ catalogue .
the stars of the two samples are presented in fig .
[ h2541f2 ] in fig .
[ h2541f3 ] .
as one can see , they do indeed show a correlation between separation and brightness anomaly , with the slope of the linear regression being non - zero well above the error in both cases ( the slope error is given as a percent of the slope value ) .
it is to be noted that the correlation remains the same when absolute magnitude @xmath0 in the definition of @xmath8 is replaced with absolute magnitude @xmath17 obtained from the _ hipparcos _ @xmath18 magnitude ( for the details of _ hipparcos _ photometry see mignard , froschle , & falin 1992 ; evans et al .
the corresponding regression slopes are : @xmath48% for the _ hipparcos _ binaries , and @xmath49% for the pre-_hipparcos _ binaries .
the correlation in fig .
[ h2541f2 ] and fig .
[ h2541f3 ] reveals that @xmath8 is , on average , larger at smaller component separations .
this suggests that brightness anomaly is somehow associated with the components interaction in a tight binary pair .
the question is then : what is the physics behind this anomaly ? to answer the above question , we have explored whether there is any dependence in @xmath8 on age , because the effects caused by interaction in a close binary may accumulate with time .
first of all , we have correlated @xmath8 and tangential velocity , employing the fact that the statistics from stellar kinematics , such as mean tangential velocity and velocity dispersion used here , are age - sensitive . as seen in fig .
[ h2541f4 ] , there is indeed a significant correlation between @xmath50 and @xmath8 .
it implies that @xmath8 is larger , on average , at older ages .
given that the _ hipparcos _ single stars with large @xmath51 are probable close binaries , we conclude that this result is consistent with increase in brightness anomaly for a tight binary pair as the primary star evolves .
[ h2541f4 ] also reveals that the kinematics of known binary stars is `` younger '' than that of the single stars at the same @xmath8 .
this reflects the generally younger age of the population of the known binary f stars , which was discussed in suchkov ( 2000 ) .
mean metallicity is another age - dependent statistics that can be used to check if there is any relationship between @xmath8 and age .
the correlation between and @xmath8 is shown in fig .
[ h2541f5 ] for the same groups of stars as in fig .
[ h2541f4 ] . here , the linear regression slope is non - zero within more than one sigma for both the binary and single star samples , indicating that @xmath8 is larger at lower metallicities . since lower metallicity is indicative of older age , the correlation in fig .
[ h2541f5 ] is consistent with the inference from the preceding subsection that the stars with larger @xmath8 are , on average , older .
the correlation in fig .
[ h2541f5 ] could have been expected on the basis of the correlation in fig .
[ h2541f4 ] because metallicity is known to correlate with kinematics . at the same time
this correlation is obviously weaker than that in fig .
[ h2541f4 ] . since metallicity is also known to have a large spread at any given age , a possible interpretation of weaker correlation may be that metallicity is coupled with age more loosely than kinematics . the weaker correlation in fig .
[ h2541f5 ] , whatever its reason , argues against the possibility that the correlation in fig .
[ h2541f4 ] results from a metallicity - dependent bias in estimates of @xmath1 , hence @xmath8 .
if such a bias did exist ( say , due to flaws in the algorithm that computes @xmath1 ) and cause all the correlation in fig .
[ h2541f5 ] , one would have expected it to translate into a weaker rather than stronger correlation between @xmath8 and kinematics in fig .
[ h2541f4 ] .
it is to be added that the aforementioned hypothetical bias seems unlikely anyway , because the available evidence , including the results from the binary components separation data and the age
velocity relation ( see below ) , argues against any role of this kind of systematic errors in brightness anomaly .
thus , the interpretation of both fig .
[ h2541f4 ] and fig .
[ h2541f5 ] involving the dependence of brightness anomaly on age appears to be reasonably justified .
so far , we have been referring to brightness anomaly in the sense that the absolute magnitude of a star with this anomaly is brighter than that of a normal single star with the same colors .
but it would be equally correct to say that the star has the colors that are anomalous for a normal star at a given effective temperature and luminosity .
the anomaly in this case refers to the color index @xmath52 . for the main sequence f stars
, @xmath52 provides a measure of balmer discontinuity ( see crawford 1975 and references therein ) . in these stars ,
smaller balmer discontinuity at a given effective temperature ( hence smaller @xmath52 ) means a denser stellar atmosphere , i.e. , larger surface gravity . in this way , @xmath52 actually measures stellar surface gravity ( at a given @xmath53 ) , through which it is related to the star s luminosity .
anomalous @xmath52 in an f star then means anomalous surface gravity , @xmath54 .
specifically , the anomaly resulting in a positive @xmath8 corresponds to the surface gravity that is too high for a given effective temperature , @xmath53 , and luminosity , @xmath55 .
consequently , the stellar mass of such a star is larger than the mass of a normal single star at the same @xmath53 and @xmath55 .
( alternatively , one can say that the luminosity of an anomalously bright star is too high for a given @xmath53 and @xmath54 ) .
this obviously requires that stellar evolution of a normal star is different from that of an anomalously bright star .
so , if @xmath52 of the anomalous stars in our sample is dominated by the effect of enhanced surface gravity , these stars can be predicted to evolve in a non - standard way .
we can verify the latter prediction as follows .
let us take the sample of the _ hipparcos _ `` single '' stars and split it into two subsamples : the first one including only anomalously bright stars , @xmath7 ( c binaries ) , and the second subsample that includes the stars with @xmath39 ; for simplicity , the latter stars will be referred to as `` truly single '' , although we expect a fraction of them to be unevolved and/or wide binaries with disparate components .
now , we can derive and compare the age velocity relations ( avr ) for these two group of stars .
if the same stellar evolution model ( meaning the same set of isochrones used to derive age ) is applicable to all stars , we should obviously get the same avr for both groups , simply because stellar kinematics and stellar evolution are entirely unrelated things . if this proves not to be the case , and the age velocity relations are discrepant , the two groups of stars evolve probably differently .
we have derived ages for both groups of stars based on isochrones from demarque et al .
the isochrones as well as the stars used to calculate the age velocity relations are shown in fig .
[ h2541f6 ] .
the stars have been selected in a narrow metallicity range , @xmath56}$\ } } < -0.1 $ ] , that matches the chemical composition of the isochrones , @xmath57 .
additionally they have been constrained to the distance range @xmath58 pc .
they have been grouped into 1 to 2 gyr age bins , for which mean tangential velocity has been calculated . because of large age errors for the young stars near the zams , ages younger 1 gyr have not been considered .
the resulting age velocity relations for the age range from 1 to 8 gyr are given in fig .
[ h2541f7 ] .
.kinematics of normal ( `` truly single '' ) and anomalously bright ( c binaries ) f stars above and below the line @xmath59 shown in fig .
[ h2541f6 ] . [ cols="<,^,^,^,^ " , ] fig .
[ h2541f7 ] reveals a conspicuous discrepancy between the two age - velocity relations , which will be referred to as the avr discrepancy .
the possibility that kinematic differences between the normal and anomalously bright stars arise from different distance sampling ( the latter stars are , on average , more distant ) was checked and ruled out in suchkov ( 2000 ) .
we have also looked into the possibility that the bias associated with the brighter , on average , absolute magnitudes of anomalously bright stars ( see fig .
[ h2541f6 ] ) plays a role in the discrepancy ( for example , this may happen if the isochrones underestimate the age of far evolved stars ) . to this end
, we have compared the kinematics of the normal and anomalously bright stars with @xmath60}$\ } } < 0.3 $ ] in the two regions of the @xmath21 diagram , above and below the line corresponding to 1.0 mag above the zams shown in fig . [ h2541f6 ] .
as seen in table 1 , anomalously bright stars close to the zams have velocities higher , on average , than those of the normal stars far above the zams .
this rules out the indicated bias as the cause of the avr discrepancy .
therefore , the discrepancy must be real , reflecting some inconsistency in the derived ages . assuming that the isochrones in fig .
[ h2541f6 ] adequately represent the age of the normal stars ( truly single ) , we must then conclude that the age of the stars with large brightness anomaly ( c binaries ) is underestimated .
the amount of the avr discrepancy suggests that the underestimation is , on average , as large as 2 gyr or more .
thus , anomalously bright stars are consistent with being actually older than normal stars at the same positions in the @xmath21 diagram .
there are at least two reasons why isochrone fitting could be inadequate for anomalously bright stars and result in incorrect ages : ( i ) combined flux of unidentified binaries with comparably bright components , and ( ii ) non - standard stellar evolution .
if anomalously bright stars are unidentified binaries with comparably bright components , isochrone fitting results in either overestimated or underestimated age , depending on the position of the star in the @xmath21 diagram and the contribution of the secondary into the star s @xmath16 magnitude . for the stars cataloged in the _ hipparcos _ as single ,
age would be , on average , underestimated ( suchkov & mcmaster 1999 ) .
the following , however , argues against the hypothesis that such binaries dominate our sample of anomalously bright stars .
first , the sample stars are much older ( on average ) than the known unresolved binaries ( suchkov , 2000 ; see also fig .
[ h2541f4 ] and fig .
[ h2541f5 ] ) .
this means that most of these stars are different from normal binaries .
even if they are double systems , in most cases their secondary component is hardly a regular star whose emission is comparable to that of the primary .
second , as mentioned before , even among known binary stars a substantial fraction have brightness anomaly exceeding the maximum value of 0.75 mag , hence it can not be attributed to the combined flux of the binary s two components only ( see fig . [ h2541f1 ] ) .
statistically , the excess is quite significant , so there is little doubt that in many cases the cause of brightness anomaly is different from the light contribution from the star s unresolved companion .
so , it seems unlikely that the avr discrepancy in fig .
[ h2541f7 ] is due only to binaries with comparably bright components .
the above isochrone fitting can also be inadequate if for some reasons the evolution of a star is different from the normal evolution of a single star .
in such a case the position of a star in the @xmath21 diagram would correspond to an age different from the one predicted by standard isochrones .
the age velocity relation based on ages from the standard isochrones would be incorrect for these stars , and this may be the reason for the avr discrepancy revealed by fig .
[ h2541f7 ] . as discussed below , anomalous stellar evolution , unlike the effect of combined flux , offers a way to explain the relationship between age and brightness anomaly ; in particular , it explains why c binaries are , on average , older than the truly single stars .
also anomalous evolution appears to be in line with the recent developments in studies of eclipsing binaries .
this makes it a stronger candidate for being the dominant source of the avr discrepancy in fig .
[ h2541f7 ] .
the combined evidence discussed in this paper suggests then that anomalous stellar evolution occurs apparently in tight binary pairs where the evolution of the primary star is affected by the presence of the companion .
the older ages of anomalously bright stars ( suchkov 2000 ) seem to indicate that these stars stay within the main sequence longer than normal single stars .
such a prolonged main sequence evolution is known to occur in stars with enhanced central mixing in the convective core , and is apparently quite ubiquitous among close binary systems at stellar masses slightly above solar ( see , e.g. , chiosi 1999 , andersen & nordstrm 1999 , gimenz et al .
this suggests that anomalous brightness discussed in this paper reflects non - standard stellar evolution that involves enhanced core mixing .
extreme manifestation of brightness anomaly among known unresolved binary stars seems to indicate that the anomaly is somehow related to binarity .
given that @xmath8 correlates with binary components separation , one may infer that interaction in a tight binary pair somehow induces extra mixing in the convective core of the binary s primary component , additional to whatever mixing occurs in a similar single star .
of course , there are other mechanisms that may impact stellar evolution in a tight binary pair .
one may recall , for example , modified convection in the outer convective envelope of a star in a tight binary pair or mass loss / transfer in very close binary systems .
these mechanisms were discussed in the literature as related to the existence of blue stragglers and other unusual stars in stellar clusters ( pols & marinus 1994 ) , apparent age discrepancy between the components in low mass binaries ( clausen et al .
1999 ) , young age of the known unresolved binary stars ( suchkov 2000 ) , etc
. however , enhanced convective core mixing seems to be the only mechanism capable of extending the main sequence evolution by a significant amount .
it is to be noted that despite the potentially longer than normal lifetime , the actual average main sequence lifetime of the known unresolved binary stars is shorter than that of the single stars ( suchkov 2000 ) .
as hypothesized in the latter paper , this may be related to violent interaction of the binary s stellar components , which effectively removes the binary s primary in the color magnitude diagram from the region occupied by the main sequence f stars ; mechanisms of such an interaction were considered , for instance , in pols & marinus ( 1994 ) . on the other hand , c binaries were found to be indeed significantly older , on average , than the truly single stars , possibly indicating that many of c binaries have substellar secondaries incapable of causing too much damage to the primary f star .
unidentified close binaries may represent a large fraction of the local population of f stars . assuming that `` single '' f stars with @xmath7 are in fact tight binary pairs , the fraction of such binaries is at least @xmath61 % .
thus , close binaries may heavily impact stellar population statistics used to probe the formation and evolution of the galaxy .
therefore , these stars are important not only for better understanding of stellar evolution , but also for developing the adequate picture of the history of the galaxy . in conclusion
, we can summarize our results as follows .
the unresolved binary f stars from the _ hipparcos _ are typically much brighter than predicted on the basis of their colors .
this discrepancy , called here brightness anomaly , inversely correlates with the binary components separations , suggesting that at least part of it is associated with binary components interaction in tight pairs .
brightness anomaly correlates with stellar kinematics in a way that implies higher , on average , anomaly values at older ages .
( iv ) . with ages from the same set of isochrones , the age velocity relations of the normal single stars and binary candidates with brightness anomaly ( c binaries ) are discrepant in a way implying that the actual ages of c binaries are older than predicted .
these results argue that the main sequence evolution of the primary f star in a tight binary pair is different from that of a similar single star , evidently because of the interaction with the secondary . the difference appears to be consistent with extra mixing in the convective core of the primary , presumably induced by that interaction ; however , other mechanisms need to be carefully studied before a definitive conclusion can be reached .
i would like to thank the anonymous referee for helpful comments and suggestions .
it is my pleasure to thank claus leitherer , stefano casertano , and nikolai piskunov for stimulating discussions .
i am grateful to s. lyapustina and g. galas for careful reading the paper and numerous revision suggestions . | binary f stars exhibit large brightness anomaly , which is defined here as the difference between the absolute magnitude from the photometry and the actual absolute magnitude of the star .
we have found that the anomaly inversely correlates with the binary components separation .
there is evidence that the correlation reflects actual population differences between close and wide binary pairs , in which case it indicates that the anomaly is somehow associated with the interaction of binary s components .
the anomaly has also been found to correlate with both kinematics and metallicity .
the sense of the correlations implies that the anomaly increases as the star evolves , suggesting a peculiar evolution of a primary f star in a tight binary pair .
this conclusion has further been supported by the study of the age
velocity relation ( avr ) of f stars that are cataloged in the _ hipparcos _ as single . among these stars ,
those with brightness anomaly were previously shown to be most likely unidentified close binaries . we have found that the avr of these binary candidates is different from that of the `` truly single '' f stars .
the discrepancy between the two avrs indicates that the putative binaries are , on average , older than similar normal single f stars at the same effective temperature and luminosity , which is consistent with the inferred peculiar evolution in close binaries .
it appears that this peculiarity is caused by the impact of the components interaction in a tight pair on stellar evolution , which results in the prolonged main sequence lifetime of the primary f star .
binaries : close stars : evolution |
the phoenix dwarf galaxy was discovered in 1976 by @xcite who described it as a very distant ( @xmath0100 kpc ) globular cluster .
it was later identified as a dwarf irregular galaxy ( dirr ) by @xcite .
these galaxies are hi rich and show obvious signs of recent star formation . despite this classification , phoenix has characteristics of a dwarf spheroid galaxy ( dsph ) and it has been suggested that it belongs to an intermediate class of galaxy between the dwarf spheroidal and dwarf irregular , along with three other dwarf galaxies ( pegasus , pisces ( lgs3 ) and antila ) .
the distance to phoenix has been determined to be @xmath0420 kpc using a variety of different methods and data sets ( for example see @xcite and , more recently , @xcite ) placing phoenix well within the local group .
the hi gas surrounding phoenix was mapped by young et al .
@xcite using the vla .
they identify several regions in the immediate vicinity with a variety of shapes and velocities .
one such cloud , located @xmath05south west of the main stellar body , has been unequivocally associated with phoenix based on the excellent agreement of the radial velocities of the stars , obtained from stellar spectra @xcite from the vlt , and the velocity of the hi cloud .
@xcite recognise that this offset suggests that phoenix could offer a unique opportunity to study the possible mechanisms responsible for gas removal in dwarf galaxies , transforming a gas - rich dirr into a gas - poor dsph . extensive optical data of the central region of phoenix also exists : the wide field planetary camera ( wfpc2 ) onboard _ hst _ has imaged the central field of the galaxy with both f814w and f555w filters ( for example see @xcite and @xcite , hereafter h09 ) .
@xcite use both of these data sets to derive separate star formation histories for the eastern and western sides of phoenix .
they report an asymmetry in the star formation rate across the face of the galaxy , in agreement with @xcite .
the western side of phoenix displays evidence for strong episodes of star formation at 180 and 40 myr ago whilst the eastern side shows evidence for more continual star formation from 250 to 50 myr ago .
they note that this is consistent with the hi location , offset to the west of the galaxy s optical emission .
these epochs of star formation are similar to those for another dirr galaxy : the small magellanic cloud ( smc ) .
@xcite report that the smc had `` bursts '' of star formation at 2.5 , 0.4 and 0.06 gyr in the past .
in fact , the phoenix dwarf galaxy shares many other common features with the smc , including its low metallicity ( z=0.004 for the smc , @xcite ; z=0.0015 for phoenix , h09 ) .
the smc is @xmath0100 times as massive as phoenix and host to an unusually high number of high - mass x - ray binaries ( hmxbs ) @xcite .
these are stellar systems in which a compact object , usually a neutron star , accretes material from a more massive donor - either a supergiant or early - type star , both of ob class .
supergiant systems are a persistent source of x - rays with luminosities , @xmath2 , @xmath3@xmath4 ergs s@xmath5 .
be / x - ray binary systems have two types of outburst associated with their x - ray emission .
type i outbursts have l@xmath6 in the range @xmath3-@xmath7 ergs s@xmath5 and occur periodically , around the time of the periastron passage of the neutron star , type ii outbursts reach higher luminosities , l@xmath8 ergs s@xmath5 and show no correlation with orbital phase @xcite .
a comparison with the milky way suggests we see a factor of @xmath050 more hmxbs in the smc than we would expect to observe , based on the mass ratio of the two galaxies .
the stellar winds of massive stars are known to be line driven @xcite and so a lower metallicity will lead to less mass being lost from the compact object progenitor .
this in turn is expected to affect the number of systems that survive the initial supernova explosionand go on to form hmxbs .
@xcite use simulations to demonstrate that this is indeed the case but are unable to replicate the the number and period distribution of the smc hmxbs with a low metallicity environment alone .
it has been known for some time that hmxbs are strong tracers of recent star formation episodes in a galaxy .
it is currently thought that an increase in star formation , possibly caused by tidal interactions with the milky way and/or the large magellanic cloud ( lmc ) @xcite along with the low metallicity of the smc have given rise to the large number of hmxbs in the smc .
@xcite ( 2003 , hereafter g03 ) have quantified this relationship for sources with x - ray luminosities in excess of @xmath4 ergs s@xmath5 .
however be / x - ray binaries , which are the most numerous subclass of hmxb @xcite , do not reach these luminosities .
one of the reasons for choosing such a high threshold luminosity was to avoid contamination from any low - mass x - ray binaries ( lmxbs ) present in any of the galaxies studied .
@xcite has shown that the number of low - mass x - ray binaries ( lmxbs ) scales linearly with the mass of the host galaxy , therefore the low mass local group dwarf galaxies like phoenix ( @xmath9 m@xmath10 , @xcite ) which are too low mass to expect lmxbs are ideal for probing the hmxb
star formation relation in the low luminosity limit . by studying any population
present we hope to further understand factors that can influence star formation in such diverse environments .
predicting the number of hmxbs can not be done with great accuracy as the effects of all the influencing parameters are not fully understood .
if we assume that the logn - logs relationship reported by g03 for hmxbs remains valid down to x - ray luminosities of @xmath11 ergs s@xmath5 , the typical luminosity of the lower type i outbursts seen in the smc @xcite , equation ( 7 ) of g03 and the average star formation rate ( sfr ) of phoenix ( @xmath12 m@xmath10 yr@xmath5 , calculated from data in h09 ) predicts [email protected] .
however , this apparently precise number could be significantly affected by many factors including the metallicity values .
the same equation predicts @xmath020 x - ray sources with @xmath14 ergs s@xmath5 ( adopting the same value for the sfr as g03 , 0.15 m@xmath10 yr@xmath5 , taken from @xcite ) for the smc .
we currently know of @xmath060 hmxbs in the smc@xcite , of which @xmath050 are x - ray pulsars , and more are being discovered every year ( e.g. sxp1062 , @xcite ) .
using a sfr of 0.2 m@xmath10 yr@xmath5 for the lmc @xcite leads to a predictions of @xmath030 sources in the lmc with this luminosity constraint .
there are currently 36 known hmxbs in the lmc , including several supergiant systems and black hole candidates .
the reason behind the apparent discrepancy in the smc could lie with both the activity cycles of these hmxbs and the details of their star formation histories .
the number of hmxbs in a galaxy will be more strongly linked to the recent star formation episodes , rather than the average star formation of the galaxy over its lifetime .
the sfr of the more recent episodes of star formation ( 40 - 60 myr ago ) in phoenix and the smc are very similar .
the smc varies from @xmath15 to @xmath16 m@xmath10 yr@xmath5arcmin@xmath17 @xcite depending on the area of the smc , whilst the most recent epoch of star formation in phoenix had a sfr of @xmath18 m@xmath10 yr@xmath5arcmin@xmath17 @xcite .
it is only by probing galaxies across a broad spectrum of mass , star formation history , and chemical composition that we can establish both trends in the hmxb populations of these galaxies and understand the significance of these objects as tracers of star formation .
the first x - ray images of the phoenix dwarf galaxy were taken with _
xmm - newton _ during the satellite revolution # 1754 on july 7 , 2009 .
this paper will discuss the data | we report on the first x - ray images of the phoenix dwarf galaxy , taken with _
xmm - newton _ in july 2009 .
this local group dwarf galaxy shares similarities with the small magellanic cloud ( smc ) including a burst of star formation @xmath050 myr ago .
the smc has an abundance of high mass x - ray binaries ( hmxbs ) and so we have investigated the possibility of an hmxb population in phoenix with the intention of furthering the understanding of the hmxb - star formation rate relation .
the data from the combined european photon imaging cameras ( epic ) were used to distinguish between different source classes ( foreground stars , background galaxies , agn and supernova remnants ) using epic hardness ratios and correlations with optical and radio catalogues . of the 81 x - ray sources in the field of view ,
six are foreground stars , four are galaxies and one is an agn .
the remaining sources with optical counterparts have log(@xmath1 ) consistent with agn in the local universe .
further investigation of five sources in the field of view suggests they are all background agn .
their position behind the gas cloud associated with phoenix makes them a possible tool for further probing the metallicity of this region .
we find no evidence for any hmxbs in phoenix at this time .
this rules out the existence of the x - ray persistent supergiant x - ray binary systems .
however the transient nature of the be / x - ray binaries means we can not rule out a population of these sources but can conclude that it is not extensive .
[ firstpage ] galaxies : individual : phoenix dwarf - x - rays : galaxies - x - rays : binaries |
the paper introduces an explicit conjecture relating the degree of the jones polynomial of a knot ( and its parallels ) with slopes of incompressible surfaces in the knot complement .
we give an elementary proof of our conjecture for alternating knots and torus knots , and check it with explicit computations for non - alternating knots with @xmath1 and @xmath2 crossings , and for the @xmath3 pretzel knots .
one side of our conjecture involves the growth rate of the degree @xmath4 ( with respect to @xmath5 ) of the colored jones function @xmath6 $ ] of a knot .
the other side involves the finite set @xmath7 of _ slopes of incompressible , @xmath8-incompressible orientable surfaces _ in the complement of @xmath9 , where the slopes are normalized so that the longitude has slope @xmath10 and the meridian has slope @xmath11 ; @xcite . to formulate our conjecture ,
we need a definition .
recall that @xmath12 is a cluster point of a sequence @xmath13 of real numbers if for every @xmath14 there are infinitely many indices @xmath15 such that @xmath16 .
let @xmath17 denote the set of _ cluster points _ of a sequence @xmath13 .
\(a ) for a knot @xmath9 , define the _ jones slopes _ @xmath18 by : @xmath19 ( b ) let @xmath7 denote the set of boundary slopes of incompressible surfaces of @xmath9 . a priori , the structure and the cardinality of the set @xmath18 is not obvious . on the other hand , it is known that @xmath7 is a finite subset of @xmath20 ; see @xcite .
normal surfaces are of special interest because of their relation with _
exceptional _ dehn surgery , and the @xmath21 character variety and hyperbolic geometry , see for example @xcite .
( the slope conjecture ) for every knot we have @xmath22 before we proceed further , and to get a better intuition about this conjecture , let us give three illustrative examples . for the alternating knot @xmath23
we have : @xmath24 where @xmath4 and @xmath25 are the maximum and the the minimum degree of @xmath26 with respect to @xmath5 . on the other hand , according to @xcite , the newton polygon ( based on the geometric component of the character variety ) has @xmath27 sides and its slopes ( excluding multiplicities ) are @xmath28 the reader may observe that @xmath29 and @xmath30 are quadratic polynomials in @xmath31 and four times the leading terms of @xmath29 and @xmath30 are boundary slopes ( namely @xmath1 and @xmath32 ) , and moreover they agree with @xmath33 and @xmath34 where @xmath35 is the number of positive / negative crossings of @xmath23 .
in addition , as y. kabaya observed ( see @xcite ) , @xmath23 has slopes @xmath36 outside the interval @xmath37=[-8,8]$ ] . for the non - alternating pretzel knot @xmath38 we have : @xmath39= \frac{37}{8 } n^2 + \frac{17}{2 }
n + \e(n ) , \\ \d^*(n ) & = & 5n\end{aligned}\ ] ] where @xmath40 is a periodic sequence of period @xmath41 given by @xmath42 if @xmath43 respectively .
@xmath38 is a montesinos knot and its boundary slopes are given by @xmath44 ( see @xcite and @xcite and compare also with @xcite ) . in this case
, @xmath29 is no longer a quadratic polynomial of @xmath31 .
instead @xmath29 is a quadratic quasi - polynomial with fixed leading term @xmath45 .
moreover , four times this leading term is a slope of the knot .
this number was the motivating example that eventually lead to the results of this paper .
likewise , four times the @xmath46-coefficient of @xmath30 is @xmath10 , which is also a boundary slope of this knot .
the pretzel knots @xmath47 and @xmath48 are mutant , alternating and montesinos .
since they are mutant , their colored jones functions ( thus their jones slopes ) agree ; see @xcite .
since they are alternating , their common jones slopes are @xmath49 and @xmath50 ; see theorem [ thm.1alt ] .
since they are montesinos , hatcher - oertel s algorithm implemented by dunfield ( see @xcite ) implies that their boundary slopes are given by @xmath51 it follows that the set @xmath7 is not invariant under knot mutation . in the previous section
, we took the shortest path to formulate a conjecture relating the degree of the colored jones function of a knot with incompressible surfaces in the knot complement . in this section
we will motivate our conjecture , and add some _ structure _ to it .
let us recall that in 1985 , jones introduced the famous _ jones polynomial _
@xmath52 $ ] of a knot ( or link ) @xmath9 in 3-space ; @xcite .
the jones polynomial of a knot is a laurent polynomial with integer coefficients that tightly encodes information about the topology and the geometry of the knot . unlike the alexander polynomial ,
not much is known about the topological meaning of the coefficients of the jones polynomial , nor about its degree , nor about the countable set of jones polynomials of knots .
this unstructured behavior of the jones polynomial becomes more structured when one fixes a knot @xmath9 and considers a stronger invariant , namely the _ colored jones function _ @xmath26 .
the latter is a sequence of elements of @xmath53 $ ] indexed by @xmath15 which encodes the tqft invariants of a knot colored by the irreducible @xmath54-dimensional representation of @xmath55 , and normalized to be @xmath56 at the unknot ; see @xcite . with these conventions , @xmath57 and
@xmath58 is the jones polynomial of @xmath9 . in many ways ,
the sequence @xmath26 is better behaved , and suitable limits of the sequence @xmath26 have a clear topological or geometric meaning .
let us give three instances of this phenomenon : * a suitable formal power series limit @xmath59 $ ] ( known as the melvin - morton - rozansky conjecture ) equals to @xmath60 and determines the _ alexander polynomial _
@xmath61 of @xmath9 ( see @xcite ) . * an analytic limit @xmath62 for small complex numbers @xmath63 near zero equals to @xmath64 also determines the alexander polynomial of @xmath9 ; see ( * ? ? ?
* thm.2 ) .
* the exponential growth rate of the sequence @xmath65 ( the so - called _ kashaev invariant _ ) of a hyperbolic knot is conjectured to equal to the _ volume _ of @xmath9 , divided by @xmath66 ; see @xcite .
on the other hand , one can easily construct hyperbolic knots with equal jones polynomial but different alexander polynomial and volume .
some already observed structure regarding the colored jones function @xmath26 is that it is @xmath5-holonomic , i.e. , it satisfies a linear recursion relation with coefficients in @xmath67 $ ] ; see @xcite .
the present paper is concerned with another notion of regularity , namely the degree of the colored jones function @xmath26 with respect to @xmath31 . since little
is known about the degree of the jones polynomial of a knot , one might expect that there is little to say about the degree of the colored jones function @xmath26 .
once observed , the regularity of the degree seems obvious as bar - natan suggests ; see ( * ? ? ?
* lem.3.6 ) and @xcite .
moreover , the degree of the colored jones function motivates the introduction of two knot invariants , the jones slopes of a knot ( a finite set of rational numbers ) and the jones period of a knot ( a natural number ) . to formulate our new notion of regularity
, we need to recall what is a quasi - polynomial .
quasi - polynomial _
@xmath68 is a function @xmath69 for some @xmath70 where @xmath71 is a periodic fuction with integral period for @xmath72 ; @xcite .
if @xmath73 is not identically zero , then the degree of @xmath74 is @xmath75 .
we will focus on two numerical invariants of a quasi - polynomial , its period and its slopes .
\(a ) the _ period _ @xmath76 of a quasi - polynomial @xmath68 as above is the common period of @xmath71 .
( b ) the set of _ slopes _ of a quadratic quasi - polynomial @xmath68 is the finite set of twice the rational values of the periodic function @xmath77 .
notice that if @xmath68 is a quasi - polynomial of period @xmath76 , then there exist polynomials @xmath78 such that @xmath79 when @xmath80 , and vice - versa .
notice also that the set of slopes of a quasi - polynomial is always a finite subset of @xmath81 .
quasi - polynomials of period @xmath56 are simply polynomials .
quasi - polynomials appear naturally in counting problems of lattice points in rational convex polytopes ; see for example @xcite .
in fact , if @xmath82 is a rational convex polytope , then the number of lattice points of @xmath83 is the so - called _ ehrhart _ quasi - polynomial of @xmath82 , useful in many enumerative questions @xcite .
the next theorem seems obvious , once observed .
the proof , given in @xcite , uses ideas from _ differential galois theory _ of @xmath84-modules and the key _ skolem - mahler - lech theorem _ from number theory .
let @xmath85 denote the degree of a rational function @xmath86 with respect to @xmath5 .
@xcite if @xmath87 is a @xmath5-holonomic sequence of rational functions , then for large @xmath31 , @xmath88 is a quadratic quasi - polynomial .
moreover , the leading term of @xmath87 satisfies a linear recursion relation with constant coefficients .
the restriction _ for large @xmath31 _ in theorem [ thm.ga ] is necessary , since the sequence @xmath89 is @xmath5-holonomic , and its degree ( given by @xmath90 if @xmath91 , by @xmath46 if @xmath92 is even and by @xmath90 if @xmath92 is odd ) is not a quasi - polynomial . on the other hand , if @xmath92 , the degree is given by the quadratic quasi - polynomial @xmath93 .
if @xmath87 is a @xmath5-holonomic sequence of rational functions and @xmath94 then @xmath95 consider a subsequence @xmath96 of the natural numbers such that @xmath97 since @xmath98 is ( for large @xmath31 ) given by a quasi - polynomial , it follows that @xmath99 for a periodic function @xmath100 and for all @xmath15 .
since @xmath101 takes finitely many values , it follows that there is a subsequence @xmath102 of @xmath96 such that @xmath103 for all @xmath31 , where @xmath104 is a slope of @xmath29 . since @xmath99 for all @xmath31
, it follows that @xmath105 .
i.e. , @xmath106 is a slope of @xmath107 .
conversely , it is easy to see that every slope of @xmath107 is the limit point for a subsequence taken to be an arithmetic progression on which @xmath101 takes a constant value . given a knot @xmath9 , we set @xmath108 combining the @xmath5-holonomicity of the colored jones function @xmath26 of a knot @xmath9 with theorem [ thm.ga ] , it follows that @xmath109 is a quadratic quasi - polynomial for large @xmath31 .
\(a ) the _ jones period _ @xmath110 is the period of @xmath109 .
( b ) the _ jones slopes _
@xmath18 is the finite set of slopes of @xmath109 .
corollary [ cor.jslopes ] implies that for every knot @xmath9 we have : @xmath111 where the right - hand side ( and consequently , the left - hand side , too ) is a finite subset of @xmath81 . for every knot
@xmath9 we have : @xmath111 where @xmath112 is a finite subset of @xmath81 .
the @xmath109 invariant records the growth rates of the ( maximum ) degree of the colored jones function of @xmath9 .
we can also record the minimum degree as follows . if @xmath113 denotes the mirror image of @xmath9 , then @xmath114 .
let us define @xmath115 notice that conjecture [ conj.1 ] applied to @xmath113 implies that @xmath116 the next proposition gives a bound for the jones slopes of a knot @xmath9 in terms of the number @xmath117 of positive / negative crossings of a planar projection . for every knot @xmath9 , every @xmath118 and every @xmath119 we have @xmath120 the reader may compare the above lemma with example [ ex.817 ] .
our next theorem confirms conjecture [ conj.1 ] for all alternating knots .
consider a _ reduced planar projection _ of @xmath9 with @xmath117 crossings of positive / negative sign
. if @xmath9 is alternating , then @xmath121 in addition , the two checkerboard surfaces of @xmath9 ( doubled , if need , to make them orientable ) are incompressible with slopes @xmath122 and @xmath123 .
our final lemma relates the jones slopes and the period of a knot .
if @xmath124 is an integer - valued quadratic quasi - polynomial with period @xmath76 , then for every slope @xmath104 of @xmath124 we have @xmath125 in particular , for every knot @xmath9 we have @xmath126 thus , if a knot has a non - integral jones slope , then it has period bigger than @xmath56 . in this section
we discuss the symmetrized version @xmath127 of @xmath109 : @xmath128 of course , @xmath129 and @xmath130 .
pictorially , we have : @xmath131 as we will see below , the symmetrized degree @xmath127 of the colored jones function has a different flavor , and relates ( at least for alternating knots ) to the signature of the knot .
@xmath132 is the _ span _ of @xmath26 , i.e. , the difference between the maximum and minimum degree of @xmath26 . on the other hand
, @xmath133 is the sum of the minimum and maximum degree of @xmath26 , and appears to be less studied . of course ,
@xmath127 are quadratic quasi - polynomials .
since the colored jones function is multiplicative under connected sum , and reverses @xmath5 to @xmath134 under mirror image , it follows that @xmath135 our next theorem computes the @xmath127 quasi - polynomials of an alternating knot @xmath9 in terms of three basic invariants : the _ signature _ @xmath136 , the _ writhe _ @xmath137 and the _ number of crossings _
@xmath138 of a reduced projection of @xmath9 .
our result follows from elementary linear algebra using the results of kauffman , murasugi and thistlethwaite , @xcite , further simplified by turaev @xcite .
see also @xcite and ( * ? ? ?
* prop.2.1 ) .
\(a ) for all alternating knots @xmath9 we have : @xmath139 ( b ) the jones polynomial @xmath140 determines @xmath138 by : @xmath141 ( c ) the jones polynomial of @xmath9 and its 2-parallel determines @xmath137 and @xmath136 by : @xmath142 part ( c ) of theorem [ thm.2alt ] is sharp .
the jones polynomial of an alternating knot determines the number of crossings , but it does not determine the signature nor the writhe of the knot .
shumakovitch provided us with a table of pairs of alternating knots with up to @xmath143 crossings ( using the thistlethwaite notation ) with equal jones polynomials and unequal signature .
an example of such a pair is with @xmath144 crossings is the knot @xmath145 and its mirror ] @xmath145 has jones polynomial @xmath146 and signature @xmath147 .
since the signature is nonzero , @xmath145 is _ not amphicheiral _ , and yet has _ palindromic _
jones polynomial .
the next colored jones polynomial is given by : @xmath148 and it is far from being palindromic .
this is another example where the pattern of the jones polynomial is blurred , but the pattern of the colored jones function is clearer .
results similar to theorem [ thm.1alt ] and [ thm.2alt ] have been also been obtained independently in @xcite using the jones polynomial of an alternating knot .
let us end this section with two comments
. in this paper @xmath9 is a knot , but without additional effort one can state similar results for a link ( and even a knotted trivalent graph , or quantum spin network ) in 3-space .
in addition , we should point out that there are deeper aspects of stability and integrality of the coefficients of the colored jones function .
we will discuss them in a future publication .
in section [ sec.alt ] we use the kauffman bracket skein module and the work of kauffman , le , murasugi thistlethwaite to give a proof of proposition [ prop.c+- ] and theorems [ thm.1alt ] and [ thm.2alt ] .
this proves our slope conjecture for all alternating knots . in section [ sec.nonalt ]
we give computational evidence for the degree of the colored jones polynomials ( and for the slope conjecture ) of the non - alternating knots with @xmath1 and @xmath2 crossings .
in addition , we verify the slope conjecture for all @xmath3 pretzel knots ( using the fusion state - sum formulas of the colored jones function studied in @xcite ) , and all torus knots ( using morton s formula for the colored jones function of those , @xcite ) .
the reader may note that the pretzel knots @xmath3 are well - known examples of montesinos knots , with non - integral slopes ( for @xmath149 ) , which are _ not _ quasi - alternating , thus the results of @xcite do not apply for this family .
in this section we will discuss some future directions , and pose some questions , problems and conjectures .
the slope conjecture involves knots in 3-space , and relates the degree of the colored jones function to their set of boundary slopes .
the slope conjecture may be extended in three different directions : one may consider ( a ) links in 3-space , ( b ) general 1-cusped manifolds and ( c ) general lie algebras .
these extensions have been considered by the author . in @xcite
we introduce a slope conjecture for knots and arbitrary lie algebras .
one may also consider 1-cusped manifolds with the homology of @xmath150 , i.e. , knot complements in integer homology spheres .
for those , the colored jones polynomial exists ( though it is _ not _ a polynomial , but rather an element of the habiro ring ; see @xcite ) .
the colored jones function is @xmath5-holonomic , and we can use the non - commutative @xmath151-polynomial of this sequence to define the set of slopes .
the following problem appears mysterious and tandalizing .
understand the _
selection principle _ which selects * the jones slope from the set of boundary slopes , * the colored jones function from the vector space of solutions of the linear @xmath5-difference equation .
now , let us post some questions , based on the limited experimental evidence .
our first question concerns the class of knots of jones period @xmath56 .
is it true that a prime knot @xmath9 is alternating if and only if @xmath152 ?
theorem [ thm.1alt ] proves the easy direction of the above conjecture . as a warm - up for our next question ,
several authors have studied the _ diameter _
@xmath153 of the set @xmath7 @xmath154 see for example @xcite .
y. kabaya pointed out to us that there are alternating knots @xmath9 with diameter bigger than twice the number of crossings ; see example [ ex.817 ] .
let @xmath155 denote the _ jones diameter _
@xmath156 proposition [ prop.c+- ] shows that @xmath157 .
moreover , the bound is achieved for alternating knots , and more generally for adequate knots ; see @xcite .
the next question concerns the class of knots of maximal jones diameter .
is it true that a prime knot @xmath9 is adequate if and only if @xmath158 ? the class of alternating knots is included in two natural classes : _ quasi - alternating knots _ , and _ adequate knots_. knot homology ( and its exact triangles ) can tell whether a knot is quasi - alternating or not ( see @xcite ) , but it seems hard to tell whether a knot is alternating or not .
adequate knots appeared in @xcite in relation to the jones polynomial and also in @xcite .
it was pointed out to us by d. futer , e. kalfagianni j. purcell and p. ozsvth that the pretzel knots @xmath3 are not adequate nor quasi - alternating for @xmath159 . in all examples ,
the set @xmath18 consists of a single element , whereas the set @xmath7 can have arbitrarily many elements .
thus , conjecture [ conj.1 ] sees only a small part of the set @xmath7 .
our next conjecture claims that the colored jones function @xmath26 of @xmath9 may see all the elements in @xmath7 . to formulate it ,
recall that @xmath26 is a @xmath5-holonomic sequence , and satisfies a unique , minimal order recursion relation of the form @xmath160 where @xmath161 \in \bq[u , v]$ ] are polynomials with greatest common divisor @xmath56 ; see @xcite .
the 3-variable polynomial @xmath162 is often called the _ non - commutative _ @xmath151-polynomial of @xmath9 .
the aj conjecture of @xcite states that every irreducible factor of @xmath163 is a factor of @xmath164 or is @xmath165-free , and vice - versa . here
@xmath166 denotes the @xmath151-polynomial of @xmath9 and @xmath166 contains all components of the @xmath167 character variety of @xmath9 ( including the abelian one ) .
let @xmath168 denote the slopes of the newton polygon of @xmath151 .
these are the so - called _ visible slopes _ of a knot .
it follows by culler - shalen theory ( see @xcite ) that @xmath169 .
let us define the @xmath5-_newton polytope _
@xmath170 of @xmath9 to be the convex hull of the monomials @xmath171 of @xmath172 .
@xmath173 is a convex polytope in @xmath174 , and we may consider the image of it in @xmath175 under the projection map @xmath176 which maps @xmath177 to @xmath178 ( i.e. , sends the monomial @xmath171 to @xmath179 ) .
the @xmath5-_slopes _ @xmath180 of a knot @xmath9 are the slopes of the projection of @xmath173 to @xmath175 .
show that for every knot @xmath9 we have @xmath181 it is easy to see that for every knot @xmath9 we have @xmath182 .
in fact , this holds for arbitrary @xmath5-holonomic sequences ; see ( * ? ? ?
* prop.1.2 ) for a detailed discussion .
thus , problem [ prob.2 ] implies conjecture [ conj.1 ] .
the aj conjecture motivates problem [ prob.2 ] .
this is discussed in detail in @xcite .
our next problem concerns the symmetrized quasi - polynomial @xmath183 of a knot from .
although @xmath183 is not a concordance invariant , it determines the signature of an alternating knot .
show that @xmath183 determines a knot homology invariant .
in this section we prove proposition [ prop.c+- ] and theorems [ thm.1alt ] and [ thm.2alt ] for an alternating knot @xmath9 , using the kauffman bracket presentation of the colored jones polynomial .
the following lemma of le ( * ? ? ?
* prop.2.1 ) , ( based on well - known properties of the kauffman bracket skein module ) shows that the sequences @xmath25 and @xmath4 have at most quadratic growth rate with respect to @xmath31 .
more precisely , for _ every _ knot @xmath9 we have : @xmath184 this implies that the slopes @xmath104 of the quadratic quasi - polynomial @xmath109 satisfy @xmath185 . replacing @xmath9 by its mirror @xmath113
, it implies the same inequality for the slopes @xmath186 of @xmath187 and concludes the proof of proposition [ prop.c+- ] .
consider a reduced planar projection of an alternating knot @xmath9 with @xmath117 positive / negative crossings .
then , the number of crossings @xmath138 and the writhe @xmath137 of @xmath9 are given by @xmath188 and @xmath189 .
let @xmath136 denote the signature of @xmath9 .
then we can express the minimum and maximum degrees @xmath25 and @xmath4 of @xmath9 in terms of @xmath190 and @xmath136 .
this was shown by kauffman , murasugi and thistlethwaite , @xcite , and further simplified by turaev @xcite .
see also @xcite and ( * ? ? ?
* prop.2.1 ) . with our conventions ,
proposition 2.1 of @xcite states that for all @xmath31 we have : @xmath191 where @xmath192 ( resp .
@xmath193 ) is the number of circles of the @xmath151 ( resp .
@xmath194 ) smoothing of the planar projection . for example , for the right - handed trefoil @xmath195
, we have @xmath196 and @xmath197 @xmath198 and @xmath199 murasugi and turaev observe that @xcite @xmath200 equation implies that @xmath4 is a quadratic polynomial ( i.e. , a quasi - polynomial of period @xmath56 ) with coefficient of @xmath46 equal to @xmath201 , i.e. , with slope @xmath202 .
this concludes the proof of theorem [ thm.1alt ] .
equations , , and prove the first part of theorem [ thm.2alt ] . it remains to show that the two checkerboard surfaces of a reduced projection of an alternating knot @xmath9 have slopes @xmath203 and @xmath204 . observe that if @xmath205 is the slope of a surface @xmath206 ( where @xmath207 is the standard meridian - longitude pair ) and @xmath208 denotes the form in the boundary of a neighborhood of @xmath9 , then @xmath209 , @xmath210 . if @xmath206 is a black surface with slope @xmath205 , then the geometrically @xmath104 and @xmath211 intersect at a point , thus @xmath212 .
in addition , @xmath104 follows the knot @xmath9 as we move towards the crossing , and intersects @xmath106 twice around each positive crossing , and none around each negative crossing .
the result follows .
let @xmath213 denote the 3-dimensional @xmath81-vector space spanned by the functions @xmath214 on the set of alternating knots .
there is an involution @xmath215 on this set , which includes an involution on @xmath213 : @xmath216 on the other hand , @xmath107 , @xmath217 and @xmath218 belong to @xmath213 and @xmath219 thus , @xmath220 is a @xmath81-linear combination of @xmath221 , and @xmath183 is a @xmath81-linear combination of @xmath222 and @xmath223 .
this is precisely the content of theorem [ thm.2alt ] .
let @xmath224 $ ] denote the coefficient of @xmath225 in a polynomial @xmath226 .
equations and imply that for all alternating knots @xmath9 we have : @xmath227= 2 \d^+_k[2]\ ] ] and @xmath228-\d^-_k[1 ] , \qquad w_k=2 \d^-_k[2].\ ] ]
to better present the experimental ( and in some cases , proven ) data presented in the next section , let us give some lemmas on quasi - polynomials .
if @xmath124 is a sequence of numbers , consider the generating series @xmath229 the next well - known lemma characterizes quasi - polynomials
. appears in ( * ? ? ?
* prop.4.4.1 ) and ( * ? ? ?
* lem.3.24 ) .
* prop.4.4.1 ) ( * ? ? ? *
lem.3.24 ) the following are equivalent : ( a ) @xmath124 is a quasi - polynomial of period @xmath76 ( b ) the generating series @xmath230 is a rational function where @xmath231 $ ] , every zero @xmath63 of @xmath232 satisfies @xmath233 ( provided that @xmath234 has been reduced to lowest term ) and @xmath235 .
( c ) for all @xmath236 , @xmath237 where each @xmath238 is a polynomial function of @xmath31 and each @xmath239 satisfies @xmath240 .
moreover , the degree of @xmath238 in is one less than the multiplicity of the root @xmath241 in @xmath232 , provided @xmath234 has been reduced to lowest terms .
( of lemma [ lem.spi ] ) given @xmath104 , there exists an arithmetic progression @xmath254 such that for all natural numbers @xmath31 we have @xmath255 now , let @xmath256 now , @xmath257 takes integer values at all integers .
this implies that @xmath258 ( see @xcite ) .
the result follows .
there are several ways to compute the colored jones function @xmath26 of a knot @xmath9 .
for example , one may use a planar projection and @xmath269-_matrices _ ; see for example , @xcite and also @xcite ) .
alternatively , one may use planar projections and _ shadow formulas _ as discussed at length in @xcite and @xcite . or one may use _ fusion _ quantum spin networks and recoupling theory , discussed in @xcite .
all these approaches gives various useful formulas for @xmath26 presented as a finite sum of a proper @xmath5-hypergeometric summand @xcite .
a careful inspection of the summand allows in several cases to compute the degree of @xmath26 . in this section
we guess the sequence @xmath109 of knots with a small number of crossings , using the following strategy , inspired by conversations with d. zagier . using the knotatlas
@xcite we compute as many values @xmath26 of the colored jones function as we can , and record their degree .
this gives us a table of values of the quadratic quasi - polynomials @xmath4 and @xmath25 .
taking the third difference of this table results into a degree @xmath10 quasi - polynomial , i.e. , a periodic function . at this point
, we make a guess for this periodic function , and the corresponding generating series .
then , we use lemma [ lem.dg ] and our guess for the second different to obtain a formula for @xmath270 and @xmath271 .
the partial fraction decomposition then gives us a formula for @xmath4 and @xmath25 . in some cases , using explicit finite multi - dimensional sum formulas for the colored jones polynomial , one can prove that the guessed formula for @xmath4 and @xmath25 are indeed correct . in this section , we will not bother with proofs .
as an example of our method , we will guess a formula for @xmath29 and @xmath30 for the @xmath38 pretzel knot .
one can actually prove that our guess is correct , using the fusion formulas for the 3-pretzel knots , but we will not bother .
the values of @xmath30 starting with @xmath272 are given by : @xmath273 taking the first difference we get the following values of @xmath274 starting with @xmath272 : @xmath275 appears to be the constant sequence from which we guess that @xmath276 , and correspondingly the generating series is @xmath277 more interesting is the sequence @xmath29 starting with @xmath272 : @xmath278 taking the first , second and third difference we obtain : @xmath279 thus , we guess that @xmath280 is a periodic sequence with period @xmath41 and generating series @xmath281 using lemma [ lem.dg ] three times , we compute @xmath282 taking the partial fraction decomposition , it follows that @xmath283= \frac{37}{8 } n^2 + \frac{17}{2 } n + \e(n),\ ] ] where @xmath40 is a periodic sequence of period @xmath41 given by : @xmath284 in this section we list the quasi - polynomials @xmath109 and @xmath187 of non - alternating knots @xmath9 with @xmath1 and @xmath2 crossings . in the rolfsen table of knots ,
the non - alternating knots with @xmath1 crossings are @xmath285 where @xmath286 , with @xmath2 crossings are @xmath287 where @xmath288 .
let us give a combined table of the non - alternating knots @xmath9 with @xmath1 and @xmath2 crossings , their period @xmath110 , their jones slopes @xmath18 and @xmath289 , and their distinct boundary slopes .
the boundary slopes @xmath7 are computed using the program of @xcite , corrected in @xcite , which computes the boundary slopes of all montesinos knots except @xmath290 which is not a montesinos knots . in all those cases ,
the set of boundary slopes agrees with the slopes of the @xmath151-polynomial of @xcite , once @xmath10 is included .
* @xmath30 and @xmath29 are mono - sloped , i.e. , they are of the form @xmath291 where @xmath40 is a linear quasi - polynomial . * for all knots , @xmath292 is a boundary slope , though not necessarily the largest one . * in the case of the @xmath293 and @xmath294 knots , the degree of @xmath40 is @xmath56 and in all other cases , it is zero . *
the period of all non - alternating knots was greater than @xmath56 .
@xmath295 knots have period @xmath296 , and @xmath38 has period @xmath41 . the period of @xmath3 for odd @xmath297 appears to be @xmath298 , and the number of crossings is @xmath299 .
thus the period can be asymptotically as large as the number of crossings .
* for the case of @xmath300 , the leading coefficient is @xmath301 and for @xmath302 it is @xmath147 . in the data below , we will give the first few values of @xmath30 and @xmath29 , the guessed decomposition of the generating series @xmath304 and @xmath305 of the quasi - polynomials @xmath217 and @xmath107 .
some values of @xmath30 and @xmath29 starting with @xmath272 : @xmath306
@xmath307 @xmath308 where @xmath309 is a 2-periodic sequence .
note that in this example the values of @xmath25 for @xmath310 suffice to prove that @xmath4 is _ not _ a polynomial of @xmath31 .
a triple sum formula for the colored jones polynomial of pretzel knots with 3 pretzels is available , and using it we can compute @xmath4 and @xmath25 for all pretzel knots of the form @xmath3 for odd @xmath74 ; see @xcite .
we will state the result of the computation here .
recall the @xmath265-th difference @xmath351 of a sequence @xmath352 from section [ sub.quasi ] .
when @xmath353 is odd , we have : @xmath354 it follows that @xmath355 where @xmath356 are linear quasi - polynomials .
when @xmath357 is odd we have : @xmath358 it follows that @xmath359 where @xmath356 are linear quasi - polynomials .
notice that the above formulas single out exceptional behavior at @xmath360 . the jones period and the jones slopes
are given by @xmath361 on the other hand , hatcher - oertel and dunfield ( see @xcite and @xcite ) compute the slopes of those montesinos knots @xmath362 compare also with mattman @xcite who computes which of those slopes are visible from the geometric component of the @xmath151-polynomial .
equations and together with the fact that @xmath10 is a boundary slope confirm conjecture [ conj.1 ] for all @xmath3 pretzel knots .
let @xmath363 denote the @xmath178 torus knot for a pair of coprime integers @xmath364 .
since the mirror image of @xmath363 is @xmath365 , we will focus on the case of @xmath366 . with our conventions ,
morton s formula @xcite for the colored jones function is the following : for example , @xmath368 the summand of equation consists of two monomials with exponents quadratic functions of @xmath265 .
a little calculation reveals that the maximum and minimum degree of the colored jones function is given by @xmath369 thus the period @xmath370 is @xmath301 when @xmath371 and @xmath56 when @xmath372 or @xmath373 . since @xmath363 is alternating if and only if @xmath372 or @xmath373 , our results are in accordance to question [ que.1 ] .
the boundary slopes of @xmath363 are @xmath374 ( see for example @xcite ) .
this confirms conjecture [ conj.1 ] for torus knots .
the author wishes to thank k. ichihara , y. kabaya , t. mattman , p. ozsvth , a. shumakovitch , r. van der veen , d. zagier , d. zeilberger and especially m. culler and n. dunfield for numerous enlightening conversations .
the idea of the slope conjecture was conceived during the new york conference on _ interactions between hyperbolic geometry , quantum topology and number theory _ in new york in the summer of 2009 .
the author wishes to thank the organizers , a. champanerkar , o. dasbach , e. kalfagianni , i. kofman , w. neumann and n. stoltzfus for their hospitality .
a. barvinok and j.e .
pommersheim , _ an algorithmic theory of lattice points in polyhedra _ , new perspectives in algebraic combinatorics , math .
* 38 * ( 1999 ) cambridge univ .
press 91147 . | the paper introduces the slope conjecture which relates the degree of the jones polynomial of a knot and its parallels with the slopes of incompressible surfaces in the knot complement .
more precisely , we introduce two knot invariants , the jones slopes ( a finite set of rational numbers ) and the jones period ( a natural number ) of a knot in 3-space . we formulate a number of conjectures for these invariants and verify them by explicit computations for the class of alternating knots , the knots with at most 9 crossings , the torus knots and the @xmath0 pretzel knots . |
the galactic x - ray transient h 1743@xmath0322 was classified as a black hole candidate by white & marshall ( 1984 ) based on its x - ray spectral characteristics .
the source was discovered during its 19771978 outburst with _
heao-1 _ and _ ariel v _
( doxsey et al . 1977 ,
kaluzienski & holt 1977 ) . the next convincing detection of h 1743@xmath0322 was made on 2003 march 21 with _ integral _ ( revnivtsev et al .
2003 ) and later with _ rxte _
( markwardt & swank 2003a ) , earning the source the additional designations igr j17464@xmath03213 and xte j1746@xmath0322 .
markwardt & swank ( 2003b ) pointed out that although the original _
_ detection reported two possible locations , it was very likely that the potentially new transient was actually h 1743@xmath0322 .
following its 2003 re - activation in x - rays , the source was extensively monitored in optical , ir , and radio bands ( see , e.g. , rupen , mioduszewski , & dhawan 2003 ; steeghs et al .
2003 and ref . therein ) .
two discoveries made during the 2003 outburst of h 1743@xmath0322 have provided additional evidence that this source harbors a black hole primary .
first , homan et al . ( 2003 ) reported the detection of quasi - periodic oscillations ( qpos ) at 240 hz in the _ rxte _ x - ray flux .
this frequency is typical of black hole candidates , and is tantalizingly close to the keplerian orbital frequency expected at the innermost stable circular orbit ( isco ) around a schwarzschild black hole ( @xmath6 ) .
moreover , there is evidence for qpos at 160 hz , and therefore in the 2:3 ratio which has recently emerged in a few galactic black hole systems ( gro j1655@xmath040 : strohmayer 2001a , grs 1915@xmath7105 : strohmayer 2001b , xte j1550@xmath0564 : miller et al .
2001 , and possibly xte j1650@xmath0500 : homan et al .
. the apparent 2:3 pairing may be evidence of a parametric epicyclic resonance in black hole accretion disks ( see , e.g. , abramowicz & kluzniak 2003 ) .
however , there is evidence that black hole qpo pairs may drift out of a 2:3 ratio ( miller et al.2001 , remillard et al .
2002 ) , and the 2:3 ratio described by the resonance model for so - called khz qpos in neutron star systems is at odds with data ( miller 2003 ) .
second , radio observations have revealed a jet in h 1743@xmath0322 , moving at @xmath8 ( rupen , mioduszewski , & dhawan 2004 ) . on the basis of this relativistic jet , h 1743@xmath0322
may be termed a `` microquasar '' .
indeed , the x - ray spectroscopic results presented in this paper will further establish close similarities between h 1743@xmath0322 and microquasars like gro j1655@xmath040 and grs 1915@xmath7105 , in addition to their radio jet and x - ray timing similarities .
although highly relativistic jets are not a unique black hole primary signature ( fender et al .
2004 ) , such jets are are more commonly observed in black hole systems than in neutron star systems .
we observed h 1743@xmath0322 simultaneously with _
chandra _ and _ rxte _ on four occasions during the bright phase of its 2003 outburst , as part of our larger multi - wavelength black hole observing program . herein , we report the results of an analysis of the x - ray spectra resulting from these observations .
most importantly , the _ chandra_/hetgs spectra reveal variable he - like fe xxv and h - like fe xxvi resonance absorption lines , which may be due to absorption in an outflow . variable , seyfert - like `` warm - absorber '' outflows have recently been detected in the x - ray spectra of the galactic black holes xte j1650@xmath0500 and gx 339@xmath04 ( miller et al.2004a , miller et al .
we examine the absorption in h 1743@xmath0322 within the context of these recent results .
the results of an analysis of the x - ray timing properties of the _ rxte _ observations will be reported in a companion paper ( homan et al.2004 ) .
the results of radio observations made in partial support of this program will be reported in a separate paper by rupen et al .
( 2004 , in prep . ) , and results from optical and ir observations will be reported in steeghs et al .
( 2004 , in prep . )
we observed h 1743@xmath0322 with _
chandra _ and _ rxte _ on four occasions during the bright phase of its 2003 outburst , as part of ongoing and complementary target of opportunity programs to study galactic black holes in outburst in x - rays .
the _ chandra _
observations were each approximately 50 ksec in duration .
_ rxte _ made observations simultaneous with each of these _ chandra _ observations , with individual exposure times ranging between 0.6 ksec and 25 ksec . the observation dates , starting times , and net exposure times are given in table 1 . at the time of writing
, pcu-2 is the best - calibrated pcu in the _
rxte_/pca , based on simple power - law fits to the crab spectrum . in order to best search for subtle features like broad fe k@xmath9 lines
, we restricted our pca spectral reduction and analysis to pcu-2 .
spectra from the hexte - a cluster were included for spectral reduction and analysis to constrain the high energy continuum .
_ rxte _ data reduction tools in the heasoft version 5.2 suite were used to screen the event and spectral files .
spectra from pcu-2 were extracted from data taken in `` standard 2 '' mode , providing full coverage of the 2.0 - 60.0 kev pca bandpass in 129 channels every 16 seconds .
data from all of the xe gas layers in pcu-2 was combined .
background spectra were made using the tool `` pcabackest '' using the latest `` bright source '' background model .
redistribution matrix files ( rmfs ) and ancillary response files ( arfs ) were made and combined into a single file using the tool `` pcarsp . ''
the background spectra were subtracted from the total spectrum using the tool `` mathpha . ''
it is well - known that fits to pca spectra of the crab with a simple power - law model reveal residuals as large as 1% , but adding 1% systematic errors to pca data is often an over - estimate . using the tool `` grppha ''
, we added 0.6% systematic errors to the spectra from pcu-2 .
the lowest energy bins in the pca are poorly calibrated , and the calibration above 25 kev is uncertain ; we therefore restricted our pcu-2 spectral analysis to the 325 kev range .
the hexte - a spectra were made from standard `` archive '' mode data , which has a nominal time resolution of 32 s and covers the 10250 kev range with 61 channels .
all spectra were background - subtracted and deadtime corrected using the standard procedures .
the calibration of the energy channels below 20 kev is poor , and we therefore restricted our analysis of the hexte spectra to energies above 20 kev .
as some of the spectra are soft and therefore dominated by background above 100 kev , we fixed this energy as an upper bound for consistency .
however , analysis of the spectra up to 250 kev was undertaken for those spectra with sufficient signal . the _ rxte _ spectra were analyzed using xspec version 11.2.0 ( arnaud & dorman 2000 ) .
all errors on fits to the _ rxte _ spectra are 90% confidence errors . in each _ chandra _
observation , the high energy transmission grating spectrometer ( hetgs ) was used to disperse the incident flux onto the acis - s ccd array .
the acis - s array was read - out in `` continuous - clocking '' mode ( nominal frametime : 2.8 msec ) to prevent photon pile - up .
information is collapsed into one dimension to achieve the fast read - out in this mode ; thus , our _ chandra _ observations were purely spectroscopic as imaging information was not preserved .
a 100-column `` gray filter '' window was used around the aimpoint on the acis - s3 chip , which acted to read - out only only one in every ten zeroth - order events .
this step was taken to prevent telemetry saturation and frame drop - out on the s3 chip .
we shifted the location of the aimpoint from the nominal s3 aimpoint .
a y - coordinate offset of 0.33 arcmin was used to shift the wavelength at which ccd chip gaps occur in order to ensure a simple response function in the fe k line region , and a sim - z offset of @xmath07.5 mm was used to push the aimpoint toward the ccd read - out to limit charge transfer inefficiencies and to limit wear and tear on the prime imaging area of the s3 chip .
spectra and lightcurves were extracted from the _ chandra _ event lists using the ciao version 2.3 reduction and analysis suite .
the `` evt1 '' file was filtered to accept only the standard event grades , to accept only events from the nominal good time intervals , and to exclude bad pixels .
the `` destreak '' tool was run to remove the effects of `` streaking '' in the ccds .
although imaging information is not preserved in the mode used for these observations , the ciao hetgs spectral extraction routines are robust for locating the observation aimpoint , defining a wavelength grid , and extracting spectra .
spectra were extracted using the default settings for the tools `` tg_resolve_events '' and `` tgextract . ''
we used the canned redistribution matrix files ( rmfs ) to make ancillary response files ( arfs ) via the `` fullgarf '' tool .
the first - order heg spectra and arfs were added to make a single spectrum and single arf using the tool `` add_grating_spectra '' ; the same addition was done with the first - order meg spectra . in the analysis detailed below , we only claim spectral features which are readily detected in the individual dispersed spectra ( prior to addition ) .
chandra_/hetgs spectra were fit with isis version 1.1.3 ( houck & denicola 2000 ) .
all errors on the absorption line parameters measured from the hetgs spectra are @xmath10 errors . in this paper
, we have restricted our analysis to the first - order heg spectra .
moreover , we fit only phenomenological power - law continua and simple gaussian line functions in a range spanning less than 1 .
the primary reasons for these restrictions are ( 1 ) the low - energy portions of the meg and heg spectra do not reveal any evidence for narrow emission or absorption lines , and ( 2 ) only the heg spectra have sufficient sensitivity through the fe k line range to allow meaningful analysis and constraints . for the benefit of scientists examining these spectra in the future , we note some additional reasons behind the above restrictions , since they are related to nuances of the telescope and the observing mode employed .
h 1743@xmath0322 had a flux roughly equivalent to or greater than 1 crab in the _ rxte_/asm ( 1.512.0 kev ) during the first _
observations , _ despite _ the high column density along the line of sight to this source .
when fit with the spectral shapes that describe the _ rxte _ spectra well , the _ chandra_/hetgs spectra show positive residuals above approximately 8 kev , and also at energies below approximately 2 kev .
these effects are inconsistent with photon pile - up , and inconsistent with the effects expected due to contaminant build - up on the acis chips .
rather , the positive residuals above @xmath118 kev are likely due to incident photons scattering off the instrument support structure and falling into the high energy portion of the dispersed spectra ( which is closest to the zeroth order ) .
the nature of the positive residuals below @xmath112 kev is less certain .
if the standard ciao tools over - correct for charge transfer inefficiency at our offset , such a correction might partially account for the positive residuals .
while these problems do not affect our ability to search for narrow emission and absorption lines through the full hetgs bandpass , we were unable to fit the full hetgs continuum with meaningful models , and we do not report such fits in this work .
we note , however , that in the 2.57.5 kev range , the _ chandra_/hetgs spectra broadly agree with the _ rxte _ spectra .
the positive residuals above @xmath118 kev made it impossible to measure the depth of the ( he - like ) fe xxv and ( h - like ) fe xxvi k edges at 8.8 kev and 9.3 kev ( respectively ) .
balancing the optical depth of these k - shell absorption edges with the optical depth of the fe xxv and fe xxvi absorption lines we detected ( see below ) would have allowed tighter constraints .
nevertheless , a number of important constraints can be made using only the variable fe xxv and fe xxvi absorption lines detected in these spectra , and we report measurements of the line parameters , a model for the absorption , and implications for the accretion inflow and outflow in h 1743@xmath0322 in this work .
we made fits to each of the time - averaged _ rxte _ spectra listed in table 1 with simple , phenomenological spectral models .
initial fits revealed that an equivalent neutral hydrogen column density of @xmath12 was easily within the error bars of each individual measurement .
therefore , we fixed this value in all subsequent fits .
this column is higher than that predicted from radio measurements ( @xmath13 , dickey & lockman 1990 ) ; however , the radio estimates are limited by the coarse spatial resolution of the survey observations . our spectral model consisted of the multicolor disk blackbody ( mcd ; mitsuda et al . 1984 ) and power - law components . as the name implies ,
the mcd model is included to account for soft , thermal , blackbody - like emission thought to arise from an optically - thick but geometrically - thin accretion disk .
the power - law component is likely primarily the result of compton - upscattering of disk photons by a central , hot , electron - dominated `` corona '' .
this region may not be distinct from the base of a jet ; if it is distinct , processes like synchrotron self - compton emission from the base of a jet may also contribute to the hard power - law component .
the results of spectral fits in the 3100 kev band with this absorbed mcd plus power - law are reported in table 2 , and an example spectral fit is shown in figure 1 .
the spectral parameters obtained through our fits are typical of galactic black hole candidates in bright phases . in general , statistically acceptable fits are obtained ( @xmath14 , where @xmath15 is the number of degrees of freedom in the fit ) with this simple model .
( the fits would be statistically improved if the xe edge at approximately 4.8 kev was better described in the detector response . )
the disk temperature varies between @xmath16 kev in our observations , indicating that the disk is likely important throughout the bright phase of the outburst and close to the innermost stable circular orbit ( isco ) .
the broad - band spectral properties indicate that h 1743@xmath0322 was likely in an extended `` very high '' or `` steep power - law '' state in observations 1a , 1b , 1c , 2a , 2b , 2c .
observation 3 occurred in a phase intermediate between the `` high / soft '' an `` very high '' ( `` steep power law '' ) state , but was likely more similar to the `` high / soft '' ( or , `` thermal - dominant '' ) state .
observations 4a , 4b , and 4c likely occurred in the `` high / soft '' ( or , `` thermal - dominant '' ) state .
the states observed in h 1743@xmath0322 differ in subtle ways from canonical black hole states , however , and the reader is referred to homan et al .
2004 ( in prep . ) for a more complete analysis and discussion of the states observed .
the highest flux inferred in the 0.510.0 kev band is @xmath17 , which corresponds to a luminosity of @xmath18 ( see table 2 ) . at present , neither the mass of h 1743@xmath0322 nor the distance to the source are known .
given its galactic center position and rather high column density , however , it is reasonable to use a galactic center distance of 8.5 kpc as an estimate of the distance to the source . at such a distance ,
the source emission was ( just ) sub - eddington for a @xmath19 black hole ( and more comfortably sub - eddington for a @xmath5 black hole ) . since homan et al .
( 2003a ) detected qpos at 240 hz and 160 hz in h 1743@xmath0322 , which may be indicative of black hole spin as the qpos are in a 2:3 ratio ( abramowicz & kluzniak 2003 ) , it is worth addressing whether or not our spectral fits support the inference of a disk extending very close to a potentially spinning black hole .
the mcd model allows a measure of the inner disk radius through its normalization : @xmath20^{2 } \times cos(\theta)$ ] ( where @xmath21 is the inclination of the disk to the line of sight ) . the smallest disk normalization is measured in observation 2b . via the normalization , @xmath22 km for @xmath23 kpc and @xmath24 , and @xmath25 km for @xmath23 kpc and @xmath26 .
the isco around a non - spinning black hole is @xmath27 ( where @xmath28 ; @xmath29 ) . for black hole masses greater than @xmath5 ,
@xmath30 km may imply spin and support any such implications from the qpos previously detected . however , this simple analysis has a few important problems and the inferred radii should be regarded cautiously
. the mcd model does not include an inner boundary condition , and inclination effects and radiative processes in the disk can be important and have not been accounted for in this exercise ( see , e.g. , shimura & takahara 1995 , and merloni , fabian , & ross 2000 )
. moreover , disk continuum spectroscopy is particularly sensitive to the absorption model and to the model for the hard component .
finally , qpo frequencies may not be simple keplerian frequencies .
the broad - band _ rxte _ spectra are remarkable for what they _ do not _ show : first , although we have only fitted the data up to 100 kev ( since the softer spectra have little flux above this energy ) , there is no evidence for a break in the power - law component in any of the spectra . breaks may indicate the electron temperature of the putative corona .
the absence of such an observed break in these spectra may indicate that the coronal electron temperature is generally very high ( @xmath31 kev , or greater ) .
observation 2b provides an illustrative case .
this observed spectrum is reasonably hard and has a high s / n , with h 1743@xmath0322 clearly detected out to 250 kev without a break .
replacing the power - law component in our spectral model with the comptt comptonization model ( in which the coronal electron temperature is a variable , titarchuk 1994 ) , @xmath32 kev is obtained .
the model underestimates the hard flux above 100 kev and is statistically poor ( @xmath33 ) , but indicates that the coronal temperature may be very hot indeed .
second , the spectra are free of any features expected to arise from disk reflection ( see fig . 1 ) .
hard x - ray irradiation of the ( relatively ) cool accretion disk by the corona is expected to produce a characteristic set of features : an fe k@xmath9 emission line , a corresponding fe k edge , and a compton - backscattering hump peaking between 2030 kev .
disk reflection features are an extremely common , if not ubiquitous feature of broad - band galactic black hole x - ray spectra . yet
, there is no evidence for any of these features in any of the broad - band _ rxte _ spectra we obtained . in table 2 , the 95% confidence upper limits on the strength of narrow and broad fe k@xmath9 emission lines are given ( fwhm@xmath34 kev and fwhm@xmath35 kev , as per the doppler broadening expected at the inner disk around a schwarzschild black hole ) . in most cases , the upper limits are stringent ( @xmath36 ev ) .
for a disk at the isco , reflection from a neutral disk is expected to produce an emission line with an equivalent width of @xmath37 ev ( george & fabian 1991 ) . especially in bright , high accretion rate phases , the disks around galactic black holes are likely to be ionized , and ionized disk reflection models predict line equivalent widths that are a few times higher than the neutral case ( see , e.g. , ross , fabian , & young 1999 ) .
we made broad - band reflection fits to observation 2b to further address the nature of the inner accretion flow geometry in h 1743@xmath0322 .
we replaced the simple power - law model with the `` constant density ionized disk model '' ( cdid ross , fabian , & young 1999 ; ballantyne , ross , & fabian 2001 ) .
the cdid model is particularly well - suited to highly ionized reflection , and includes fe k@xmath9 line emission self - consistently .
we `` smeared '' the cdid model by convolving it with the laor relativistic emission line function ( laor 1991 ) .
the disk emissivity ( @xmath38 ; @xmath39 is expected for a standard disk ) , the inner disk radius @xmath40 ( in units of @xmath4 ) , and inclination parameters in the smearing function variables in the spectral model . in all cases where the reflection fraction @xmath41 was allowed to vary ,
values along these lines were measured : log@xmath42 , @xmath43 , @xmath44 , @xmath45 , and @xmath46 ( 95% confidence ) .
such fits were formally acceptable ( @xmath47 ) though clearly not statistically required over the phenomenological mcd plus power - law model . in fits with
@xmath48 fixed , even for @xmath49 ( extreme smearing ) the fit was not acceptable @xmath50 .
fits with higher values of @xmath51 fixed were statistically much worse .
the _ rxte _ variability analysis conducted in this work follows the methods and procedures detailed in homan et al .
( 2004b ) .
the reader is directed to that work for a full discussion of the timing analysis presented here .
an example lightcurve showing strong qpo - like variability on the timescale of @xmath52 s is shown in figure 2 .
a power density spectrum of _ rxte _ observation 1a is shown in figure 3 ; the variability seen in the lightcurve corresponds to the broad @xmath53 hz peak in the power density spectrum .
it is clear that much of the power is concentrated in this feature .
the term `` qpo '' is usually reserverd for features with @xmath54 when fit with a lorentzian ( where @xmath55 ) .
as @xmath56 for the @xmath57 hz feature , it may not be termed a qpo .
however , it is important to note that , like a qpo , its frational rms _ increases _ with increasing energy ( homan et al.2004b ; see figures 2 and 5 ) .
we began by analyzing the combined first - order heg and meg time - averaged spectra from each observation in narrow 2 slices , through the 1.225 range .
the continuum in each slice was fitted with a simple phenomenological power - law model . where required , absorption edge(s ) due to neutral ( atomic ) photoelectric absorption in the ism were included and scaled to @xmath58 ( see table 2 ) .
line features were fit with simple gaussian models .
line identifications are based on the calculations and line lists tabulated by verner , verner , & ferland ( 1996 ) .
column density estimates were derived using the relation : @xmath59 + where @xmath60 is the column density of a given species , @xmath61 is the oscillator strength , @xmath62 is the equivalent width of the line in cm units , and @xmath63 is the wavelength of the line in cm units ( spitzer 1978 ) .
in using this relation , we have assumed that the lines are optically thin and that we are on the linear part of the curve of growth . in observations 1 , 3 , and 4 ,
we clearly detect absorption lines which we identify as he - like fe xxv @xmath64 ( @xmath65 ) and h - like fe xxvi @xmath66 ( @xmath67 ) resonance absorption lines .
the line wavelengths , velocity shifts , line breadths , fluxes , equivalent widths , columns , and equivalent neutral hydrogen columns are listed in table 3 .
the spectra are shown in figure 4 .
there is no clear evidence for any narrow emission or absorption lines in the rest of the spectrum .
there are some line - like features in the region of the neutral si edge which might be he - like si xiii and h - like si xiv ; however , these are more likely due to uncertainties in the instrument response in this region ( the spectrum peaks in this region , and the ccds contain si ) .
we have measured significant variability in the line equivalent widths and other parameters between observations , which demands changing absorption and clearly ties the lines to a region close to ( or within ) h 1743@xmath0322 .
in observation 2 , when the source flux was high and twice as spectrally hard as in the other observations , the fe xxv and fe xxvi lines are absent . in observation 1 , the lines are significantly blue - shifted , suggesting that the absorbing matter was flowing into our line of sight .
the fe xxvi line in observation 3 is also slightly blue - shifted . in all other cases ,
the lines are consistent with no net shift . in observations 1 and 4 ,
the fe xxvi lines are resolved , and have velocity widths ( fwhm ) of @xmath68 km / s and @xmath69 km / s , respectively .
the rest of the lines are not resolved and likely have smaller velocity widths .
we made lightcurves of the dispersed spectrum from each observation in the 0.510.0 kev band .
the lightcurves are presented in figure 5 .
strong variability is present in all observations on the timescale of @xmath52 s , but particularly in the first and second observations .
this variability is confirmed to also be present in the lightcurves of the simultaneous _ rxte _ lightcurves of h 1743@xmath0322 ( see fig . 2 and fig .
a 1012% `` dip '' feature may be present in the third _ chandra _ observation . a preliminary examination of other _ rxte _ observations in the public archive revealed much stronger dipping activity typical of dipping black hole binaries viewed at high inclinations ( see , e.g. , kuulkers et al .
1998 ) . to explore
whether the absorption lines vary within the dip , we made spectra of observation 3 from before the dip and within the dip ( see figure 5 ) , and fit the spectrum in the manner noted above .
again , the only lines apparent are the fe xxv and fe xxvi absorption lines found in the time - averaged spectrum .
the results of fits to the spectrum prior to the dip and within the dip are given in table 4 , and the spectra are shown in figure 6 .
both absorption lines are stronger within the dip . while the statistical significance of the variability in the fe xxvi line is marginal , the variability in the strength of the fe xxv line is clearly significant .
indeed , in observation 3 , the fe xxv line is not clearly detected in the pre - dip spectrum .
next , we investigated whether the absorption lines in observation 1 , and 4 vary on the @xmath52 s flaring variability timescale .
we calculated the mean count rate for observations 1 , 2 , and 4 ( see fig
. 5 ) , produced lists of time intervals for count rates above and below the mean rate , and extracted spectra from the time intervals above and below the mean rates .
those spectra were also fitted in the manner described above ; the spectra are shown in figure 6 . as was the case with the time - averaged spectrum of observation 2 ,
the count - rate selected spectra of observation 2 show no absorption lines . in observation 4
, the strength of the lines does not vary significantly between the high rate and low rate spectra . in observation 1 , however , the absorption lines are marginally stronger in the low rate spectra than in the high rate spectra .
this raises the interesting possibility that the absorption may vary on a timescale of @xmath52 s. to establish this more clearly , we created a list of time intervals in which the source count rate was more than 5 counts / s above the mean rate , and more than 5 counts / s below the mean rate , and extracted spectra from these time intervals .
the resultant spectra are shown in figure 7 .
the fe xxv absorption line is clearly stronger in the low count rate spectrum than in the high count rate spectrum , and the variation is statistically significant ( see table 4 ) . to interpret the iron absorption lines measured in the time - averaged spectra in more detail , we computed line profiles from a spherical wind with a photoionization code .
we used the atomic physics packages from the x - ray illuminated accretion disk models of raymond ( 1993 ) . in this case , since the gas is optically thin , it is only necessary to consider a single slab .
we used the spectral parameters taken from table 5 for each of the 4 observations and specify a density , slab thickness and distance from the central object , to model a 1-d , optically - thin wind .
the code iterates to find a self - consistent temperature and ionization state , then computes equivalent widths assuming a doppler profile with widths ranging from 100 to 2000 @xmath70 .
it assumes ionization equilibrium , which is a good approximation for the densities and dynamical times estimated below in all cases . from the output of the code we find the parameters that match the observed fe xxv and fe xxvi absorption line equivalent widths .
the fe xxv and fe xxvi lines can not originate in exactly the same gas , because the line widths and line shifts differ .
a plausible physical picture might consist of denser , slower clumps containing more fe xxv embedded in a less dense gas where fe xxvi dominates .
however , a modest density contrast of order 2 ( approximately ) is sufficient to shift the balance between fe xxv and fe xxvi , so an average density that produces both lines is meaningful , and we do not have enough measured parameters to justify the additional free parameters of a more complex model ( the line widths are uncertain ) .
we therefore choose densities to match the observed equivalent widths and doppler velocity widths intermediate between those of fe xxv and fe xxvi .
there is a maximum radius at which a slab can plausibly be causing the observed absorption .
the fe xxvi / xxv line flux ratio gives an ionization parameter ( @xmath71 ) , and for any given distance from the central source @xmath72 , a specific density @xmath73 is required . a specific slab thickness , which can not exceed @xmath72 , is required to produce the measured line column densities , so that @xmath74 .
because the density that gives the ionization parameter scales as @xmath75 , the slab thickness scales as @xmath75 ( therefore @xmath76 ) , and there is a maximum radius at which a slab of thickness @xmath72 can reproduce the lines observed .
we call this parameter @xmath77
. the lines can be formed at smaller @xmath72 , and if the lines are formed in a wind with @xmath75 density profile , the density at @xmath77 is correspondingly smaller .
table 5 shows the derived parameters .
we can also compute the mass loss rate in a wind of any velocity , because @xmath78 is constant .
the velocity shifts in table 3 are generally uncertain , so we present mass loss rates for a reference wind speed of 300 @xmath79 .
the table shows that the lack of absorption lines in the second observation can not entirely be attributed to a higher ionizing flux .
although the ionizing flux in this observation is about double that of the first observation , an order of magnitude lower column density of absorbing material is needed to match the upper limits to the equivalent widths .
an interesting result in table 5 is the small value of @xmath77 for the 4th observation , which results primarily from the small ionizing flux .
the lines do not show a significant doppler shift , but they arise in a white dwarf - size region near the central source . in the first and third observations ,
the absorbing regions have a size comparable to a few solar radii , which is similar to the separation of the components in several binary black hole systems . the second interesting aspect of the table is the indication that the observed absorption lines are formed in a wind with a mass loss rate of order @xmath80 comparable to the inferred mass accretion rate . if we assume a distance of 8.5 kpc , take the highest inferred 0.510.0 kev luminosity of @xmath81 and assume an efficiency of 10% in the equation @xmath82 , we obtain @xmath83 . taking the highest measured 0.510.0 kev luminosity to be a lower limit on the true eddington luminosity ,
the highest inferred mass loss in the wind ( @xmath84 ) corresponds to an outflow rate that is 22% of the eddington mass accretion rate ( for a unity filling factor ) .
note , however , that the fraction depends crucially on the filling factor , the energy range used , and the distance assumed .
an outflow of this kind may be analogous to the accretion disk winds from cataclysmic variables in their high states and to t tauri stars in high states ( fu ori stars ) , which have wind mass loss rates of order 10% of their accretion rates .
however , observation 4 shows a much smaller mass loss rate along with its lower luminosity .
the high equivalent neutral hydrogen column density along the line of sight is consistent with a central galactic distance to h 1743@xmath0322 . for distances greater than @xmath853
kpc , the highest 0.510.0 kev unabsorbed flux observed ( see table 2 ) exceeds the eddington limit for a neutron star .
moreover , the parameters obtained from broad - band spectral fits to h 1743@xmath0322 with the phenomenological mcd plus power - law model are typical for galactic black hole systems in bright phases ( for a review , see mcclintock & remillard 2003 , see also homan et al .
2001 ) . these facts
and the presence of 240 hz qpos in h 1743@xmath0322 make it a near certainty that h 1743@xmath0322 harbors a black hole primary . however , the apparent absence of reflection features in the spectrum , even in phases with 1 ) a strong hard component , 2 ) a small inferred inner disk radius , and 3 ) high - frequency qpos , is at odds with the majority of galactic black holes .
it is likely that the low reflection fraction obtained is due to a combination of astrophysical effects ( ionization and disk inclination ) and the limitations of our reflection modeling .
zdziarski et al.(2003 ) have shown that when the disk is fully ionized , the reflected spectrum is nearly featureless , and might therefore be confused with a zero reflection scenario .
high disk ionization tends to smooth - out reflection features ; the dips we have detected in the lightcurve of _ chandra _ observation 3 and in additional public _ rxte _ observations of h 1743@xmath0322 point to a system viewed at high inclination . at high inclinations ,
doppler shifts become very important , and tend to smear - out reflection features . the cdid reflection model
we fit to _ rxte _ observation 2b indicates that the disk in h 1743@xmath0322 is likely highly ionized and seen at a high inclination .
while these constraints are likely robust , it must be acknowledged that even the cdid model has limitations .
it is implemented into xspec as a library of angle averaged spectra .
finally , we note that if any reflection spectrum also passes through the hard x - ray emitting corona , which is likely if a corona hugs a disc viewed at high inclination , then the reflection component will be comptonized and much less detectable ( petrucci et al 2001 ) .
it is possible , then , that the total reflection model the cdid model smeared with a line function ( for which inclination is a fit parameter ) does not address high inclinations correctly . were the combination of ionization and inclination effects more accurately described by our model , it is likely that a high reflection fraction would have been within errors .
we have clearly detected variable , narrow fe xxv and fe xxvi resonance absorption lines in the spectra of h 1743@xmath0322 . in the recent past , similar absorption lines have been detected in the microquasars gro j1655@xmath040 ( ueda et al .
1998 ) , and grs 1915@xmath7105 ( kotani et al.2000 ) with _ asca _ , and in grs 1915@xmath7105 with _
( lee et al . 2002 ) .
unfortunately , each of these prior spectra suffered heavily from photon pile - up .
the extraordinary sensitivity of the h 1743@xmath0322 spectra has allowed us to improve upon previous implications based on fe xxv / xxvi absorption lines in microquasars in two important ways : first , we have measured blue - shifts in the lines detected in observation 1 , providing direct evidence that the highly ionized absorbing geometry may be an outflow ( previously , outflows were only asserted ) .
second , we have resolved the lines in two cases , and in other cases the lines are likely narrower than the instrumental resolution .
the fwhm of the lines is vital for constraining the nature of the absorption .
viable descriptions of the absorption geometry in h 1743@xmath0322 must be able to account for a number of our key findings , including : ( 1 ) the blue - shifts observed in the first _
observation , ( 2 ) the large line flux changes observed between observations ( in particular , the absence of absorption lines in the second _ chandra _ observation ) , and the stronger absorption seen in the dip observed during the third _ chandra _ observation , and ( 3 ) the absence of fe emission lines in the spectra , ( 4 ) the variability of the absorption lines in observation 1 on a timescale of a few hundred seconds , and ( 5 ) the absorption must occur at a radius smaller than the maximum radius at which our photoionization models suggest the lines may be produced .
two geometries satisfy or partially satisfy most of these requirements : cyclic absorption in a disk structure resulting from the high inclination , and absorption in an ( potentially clumpy ) outflowing disk wind .
the absence of fe emission lines in the _ chandra _ and _ rxte _ spectra suggests that the absorbing geometry is not spherical ( else , emission line photons would have scattered into our line of sight ) .
indeed , modeling of the absorbers in gro j1655@xmath040 ( ueda et al.2004 ) and grs 1915@xmath7105 ( kotani et al .
2000 ) also suggested a cylindrical , perhaps pancake - like absorbing geometry .
absorption in a disk structure perhaps a disk atmosphere is consistent with this constraint .
however , this picture relies on relatively large covering and filling factors .
the lack of emission lines in the spectra is also consistent with the absorption occurring in relatively dense clumps in a disk wind , with large covering factors but small filling factors .
a clumpy wind is particularly appealing when one considers that at high mass accetion rates ( such as in the first _ chandra _ observation ) , our models for the absorber indicate that the wind would already be optically - thick at approximately @xmath86 . yet , we clearly see continuum spectral components consistent with viewing the innermost flow without significant obscuration .
moreover , if qpos are tied to keplerian frequencies , the qpos detected ( homan et al .
2004 , in prep . ) indicate that emission from the innermost flow is either seen directly or scattered into our line of sight .
a significant fraction of the absorption needs to occur in clumps with a high covering factor but small filling factor to be consistent with the observed continuum spectral properties and variability properties .
the absence of absorption lines due to a combination of geometric effects and ionizing flux ( as implied by our photoionization models ) in the second _ chandra _ observation is easily explained by the clumpy outflow model .
simply , a clump may not have intersected our line of sight to the central accretion observation within this observation .
it is more difficult to explain the absence of absorption lines if the absorption occurs via cyclic absorption in a disk structure .
the disk structure would have to be dampened or disappear in the second _ chandra _ observation .
however , the ionizing flux is higher in this observation , and it has been shown that irradiation can induce warps in accretion disks ( see , e.g. , maloney , begelman , & pringle 1996 ) . to explain the enhanced absorption line fluxes measured within the dip in the third _ chandra _ observation , it is easy to envision a cool clump absorbing fe xxv more strongly than the hotter surrounding wind .
alternatively , it is also possible that the dip observed is due to an unrelated geometric effect , and that the enhanced absorption line flux seen within the dip occurs because the wind along our line of sight is temporarily shielded and allowed to cool . if the absorption lines are instead due to absorption in a disk structure , a second warp or disk structure may need to be invoked to explain the dip itself . the variability of the fe xxv absorption line flux on timescales of a few hundred seconds in the first _
chandra _ observation can be explained by both models .
indeed , this short timescale line variability is one of the most compelling arguments in favor of cyclic absorption in a disk structure .
the period of the line variability broadly agrees with the keplerian orbital period at @xmath3 for a @xmath5 black hole .
if h 1743@xmath0322 harbors a black hole with @xmath87 , then @xmath3 corresponds to @xmath88 cm , which is less than @xmath77 in observations 1 and 3 ( see table 5 ) .
however , the observed variability can not be termed a qpo , and it is therefore less easily tied to a cyclic disk structure than if it qualified as a true qpo . if the absorption occurs in a disk structure , it would suggest that the variability itself is due to changes in the absorption in our line of sight .
this is at odds with the fact that similar variability is observed in the absence of absorption lines in the second _ chandra _ observation .
if the variability were due to absorption , the variability should be more pronounced at low energies ; in fact , the fractional rms of the peak at @xmath57 hz increases with higher energy ( homan et al .
2004b ; see figures 2 and 5 ) .
although it is less exciting , absorption in a clumpy outflowing wind can likely account for the rapid line flux variability at least as well as cyclic absorption in a disk structure .
simply , the gas ionization and recombination timescale is much faster than the variability timescale observed .
if the variability is not due to a disk structure , but rather just due to fluctuations in the intensity of the central engine , an absorbing clump may be able to rapidly respond to the changing flux . finally , the blue - shifts observed in the first _
observation are more easily explained through an outflowing disk wind , than by absorption in the disk .
the modest blue - shifts of the lines are a small fraction of the speed of light , and similar to the blue - shifted lines observed in gx 339@xmath04 and xte j1650@xmath0500 ( miller et al .
2004a ) , which are consistent with a disk wind .
if the absorption is instead due to a cyclic disk structure , it is rather hard to explain the blue - shifts without significant fine - tuning .
if the structure were a warp , its motion should be tangential to our line of sight , not into our line of sight . instead , the absorption might need to occur in a disk wind flowing outward along the disk , and further postulate that the wind density is modulated by a cyclic disk structure underlying the atmosphere .
it has been shown that accretion disks may drive outflows with a significant component along the disk ( see , e.g. , murray & chiang 1997 , proga 2003 ) . on balance , absorption in an outflowing , clumpy
wind appears to be the more compelling than absorption in a disk structure .
both possibilities require a degree of fine - tuning to satisfy the observational constraints .
a combination of these simple models may also be a viable description of the absorption geometry .
the highly ionized outflow revealed in these _ chandra _ observations of h 1743@xmath0322 is very likely connected to the variable , outflowing , seyfert - like warm absorbers recently discovered in gx 339@xmath04 and xte j1650@xmath0500 ( miller et al .
2004a , miller et al.2004b ) . in those systems ,
the outflowing absorbers are consistent with disk winds or shell ejections .
the absorbers in gx 339@xmath04 and xte j1650@xmath0500 are blue - shifted , and so clearly outflowing . in gx 339@xmath04 , he - like ne ix , mg xi , and o vii absorption lines were detected simultaneously with more strongly blue - shifted ne ii and ne iii lines , suggesting a `` clumpy '' outflow with regions of differing temperature and density .
the fe xxv and fe xxvi lines in h 1743@xmath0322 also need to come from distinct regions , though the contrast needed is much less than for the wind observed in gx 339@xmath04 .
the outflow observed in gx 339@xmath04 was likely detected at a lower fraction of the eddingtion mass accretion rate in that source ; its lower ionization can therefore be attributed to a lower central source flux . moreover , the outflow in gx 339@xmath04 implied an order of magnitude lower mass outflow rate than than the highest rate implied in h 1743@xmath0322 ( perhaps as high as 22% of the eddington inflow rate )
. the outflow in h 1743@xmath0322 may be clumpy , however , and have a low filling factor ; in that case , the outflow rate in h 1743@xmath0322 could be lower by a factor of @xmath89 , or more .
it is plausible that the outflow in h 1743@xmath0322 is a hot , high @xmath90 precursor to the cooler , lower @xmath90 outflows observed in gx 339@xmath04 and xte j1650@xmath0500 ( miller et al . 2004a , 2004b ) .
the mass outflow rate in the fourth _ chandra _ observation of h 1743@xmath0322 is far less than that in the first observation ( see table 5 ) , further indicating an evolution along the lines we suggest here .
the outflows in these black hole systems are likely disk winds , and resemble the warm absorbers revealed in some seyfert galaxies ( see , e.g. , reynolds 1997 , elvis 2000 , morales & fabian 2002 ) . such absorption regions could be tied to rather equatorial disk outflows ( murray & chiang 1997 , proga 2003 ) , rather than to a highly collimated flow perpendicular to the disk .
focused studies of galactic black hole winds in the future may enable us to learn about the evolution seyfert warm absorbers ; their evolution is difficult to study directly given the inherently longer timescales .
it is worth addressing the possibility that a putative disk structure and the outflow are not merely coincident , but potentially causally related .
for instance , a family of models has been discussed wherein jets or a wind can be launched vertically from a disk through a coupling between a spiral density wave in the disk and a rossby vortex ( tagger & pellat 1999 ; varniere , rodriguez , & tagger 2002 ; rodriguez , varniere , tagger , & durouchoux 2002 ) .
it is possible , then , that the @xmath52s variations seen in h 1743@xmath0322 are due to a density structure which acts to drive jets close to the black hole , and that the same wave modulates a disk wind far from the black hole .
it is not yet clear if such a mechanism could be at work in driving both jets and a disk wind , but if so , then the appearance of ionized fe absorption lines in strong jet sources like gro j1655@xmath040 , grs 1915@xmath7105 , and now h 1743@xmath0322 would be explained naturally .
we note that using reduction methods exactly like those detailed for h 1743@xmath0322 , we have detected fe xxv and fe xxvi lines in a 60 ksec _ chandra_/hetgs spectrum of cygnus x-1 obtained on 3 march 2004 , when the source was in a canonical low / hard state ( miller et al.2004c , in prep . ) .
cygnus x-1 is known to produce a compact , steady radio jet in this state ( stirling et al .
highly ionized fe absorption lines have also been detected in the neutron star system gx 13@xmath71 ( sidoli et al .
interestingly , gx 13@xmath71 is also a radio emitter in which a delay between its x - ray spectral hardness and radio brightness has been observed which is very similar to that seen in grs 1915@xmath7105 ( homan et al .
2004a ) , strongly suggesting a compact jet is also at work in this system .
however , simultaneous radio observations of h 1743@xmath0322 with the vla / vlba reveal that the radio source was quite strong during the first and second _ chandra _ observations nearly 10 mjy at ghz frequencies ) but only 2 mjy during the third observation , and less than 1 mjy during the fourth observation ( rupen et al . 2004 , in prep . ) .
thus , strong radio activity ( indicative of a jet ) is seen during the second _ chandra _ observation ( where absorption lines are not observed ) , and absorption lines are clearly observed in the third and fourth _ chandra _ observations when the radio flux was low . at minimum , this suggests that if jets and ionized absorption are causally related , they are not related in a simple way
. both may merely be the effect of a strongly active corona , for instance .
moreover , the detection of highly ionized fe absorption lines in high inclination neutron star binaries may signify that it is unlikely that ionized absorption lines are tied to jet production through a mechanism like aei .
similar lines have also been seen in the persistent emission of the ( edge - on ) dipping neutron star binaries x 1254@xmath0690 ( boirin & parmar 2003 ) , x 1624@xmath0490 ( parmar et al.2002 ) , mxb 1658@xmath029 ( sidoli et al . 2001 ) and 4u 1916@xmath0053 ( boirin et al . 2004 ) .
in contrast to the absorption lines in the black hole systems h 1743@xmath0322 , gx 339@xmath04 , and xte j1650@xmath0500 , the absorption lines in these edge - on neutron star binaries are not significantly blue - shifted .
although the appearance of lines in both black hole and neutron star systems suggests a related physical origin , the lack of blue - shifts in the lines seen in neutron stars may hint at a different physical picture . without significant blue - shifts , it is difficult to clearly tie the lines to an outflow .
the lines may not be due to absorption in an outflow , but might be due to absorption in ambient gas within the system .
the absorption lines in mxb 1658@xmath029 , for instance , are not observed to vary with orbital phase or even during dipping events ( sidoli et al .
2001 ) , which suggests that the absorbing gas is not as close to the compact object as the absorbing gas in black hole systems .
+ we wish to thank cxc director harvey tananbaum and the cxc staff , and jean swank and the _ rxte _ staff , for making these too observations . we thank john houck for assistance with isis .
jmm acknowledges helpful discussions with mike nowak , david huenemoerder , aneta siemiginowska , rob fender , herman marshall , and patrick wojdowski .
jmm gratefully acknowledges support from the nsf through its astronomy and astrophysics postdoctoral fellowship program .
whgl and jh gratefully acknowledge support from nasa .
ds acknowledges support from the sao clay fellowship . this research has made use of the data and resources obtained through the heasarc on - line service , provided by nasa - gsfc .
houck , j. c. , & denicola , l. a. , 2000 , astronomical data analysis software and systems ix , in asp conference proceedings , vol .
216 , eds .
n. manset , c. veillet , & d. crabtree ; isis is available at http://space.mit.edu/cxc/isis lll instrument & start time & net exposure ( ksec ) + cxo / acis@xmath7hetgs & 2003 - 05 - 01t21:41:03 & 48.8 + rxte / pca 1a & 2003 - 05 - 01t17:00:32 & 25.2 + rxte / pca 1b & 2003 - 05 - 02t00:02:24 & 0.8 + rxte / pca 1c & 2003 - 05 - 02t00:54:08 & 3.3 + cxo / acis@xmath7hetgs & 2003 - 05 - 28t04:09:21 & 45.5 + rxte / pca 2a & 2003 - 05 - 28t05:28:48 & 1.5 + rxte / pca 2b & 2003 - 05 - 28t06:44:16 & 16.1 + rxte / pca 2c & 2003 - 05 - 28t14:28:32 & 5.6 + cxo / acis@xmath7hetgs & 2003 - 06 - 23t15:56:10 & 50.0 + rxte / pca 3 & 2003 - 06 - 23t17:05:20 & 13.3 + cxo / acis@xmath7hetgs & 2003 - 07 - 30t15:57:58 & 50.2 + rxte / pca 4a & 2003 - 07 - 30t19:48:48 & 0.6 + rxte / pca 4b & 2003 - 07 - 30t21:22:40 & 0.6 + rxte / pca 4c & 2003 - 07:31t00:32:48 & 0.6 + lllllllll observation & 1a & 1b & 1c & 2a & 2b & 2c & 3 & 4 + @xmath91 & @xmath92 & @xmath92 & @xmath92 & @xmath92 & @xmath92 & @xmath92 & @xmath92 & @xmath92 + @xmath93 ( kev ) & @xmath94 & @xmath94 & @xmath95 & @xmath95 & @xmath96 & @xmath97 & @xmath98 & @xmath99 + @xmath100 & @xmath101 & @xmath102 & @xmath103 & @xmath104 & @xmath105 & @xmath106 & @xmath107 & @xmath108 + @xmath109 & @xmath110 & @xmath111 & @xmath112 & @xmath113 & @xmath114 & @xmath115 & @xmath116 & @xmath117 + @xmath118 & @xmath119 & @xmath120 & @xmath121 & @xmath122 & @xmath123 & @xmath124 & @xmath125 & @xmath126 + @xmath127 & @xmath128 & @xmath129 & @xmath130 & @xmath131 & @xmath132 & @xmath133 & @xmath134 & @xmath135 + @xmath136 ( 3 - 100 ) & @xmath137 & @xmath137 & @xmath137 & @xmath138 & @xmath139 & @xmath140 & @xmath141 & @xmath142 + @xmath143 ( 3 - 100 ) @xmath144 & @xmath145 & @xmath146 & @xmath147 & @xmath148 & @xmath149 & @xmath150 & @xmath151 & @xmath152 + @xmath153 & @xmath154 & @xmath155 & @xmath156 & @xmath157 & @xmath158 & @xmath159 & @xmath160 & @xmath161 + @xmath136 ( 0.5 - 10 ) & @xmath162 & @xmath163 & @xmath164 & @xmath165 & @xmath166 & @xmath167 & @xmath168 & @xmath169 + @xmath143 ( 0.5 - 10 ) @xmath144 & @xmath170 & @xmath171 & @xmath172 & @xmath156 & @xmath173 & @xmath174 & @xmath175 & @xmath161 + fe k@xmath9 ( fwhm@xmath1850 kev ) ( ev ) & @xmath186 & @xmath186 & @xmath187 & @xmath188 & @xmath189 & @xmath190 & @xmath190 & @xmath191 + fe k@xmath9 ( fwhm@xmath1852 kev ) ( ev ) & @xmath191 & @xmath188 & @xmath188 & @xmath192 & @xmath193 & @xmath194 & @xmath195 & @xmath189 + lllllllllllll obs . & & theor . & meas . & shift & & flux & w & n@xmath196 & n@xmath197 + & & ( ) & ( ) & ( km / s ) & ( @xmath198 ) & ( km / s ) & ( @xmath198 ph/@xmath199/s ) & ( m ) & ( @xmath200 ) & ( @xmath201 ) + 1 & & 1.850 & 1.848(1 ) & @xmath202 & @xmath203 & @xmath204 & @xmath205 & @xmath206 & @xmath207 & @xmath208 + 1 & & 1.780 & 1.776(1 ) & @xmath209 & @xmath210 & @xmath1 & @xmath211 & @xmath212 & @xmath213 & @xmath214 + 3 & & 1.850 & 1.851(2 ) & @xmath221 & @xmath222 & @xmath223 & @xmath224 & @xmath225 & @xmath226 & @xmath227 + 3 & & 1.780 & 1.778(1 ) & @xmath228 & @xmath229 & @xmath230 & @xmath231 & @xmath232 & @xmath233 & @xmath234 + 4 & & 1.850 & 1.850(1 ) & @xmath235 & @xmath236 & @xmath237 & @xmath238 & @xmath239 & @xmath240 & @xmath241 + 4 & & 1.780 & 1.781(1 ) & @xmath242 & @xmath243 & @xmath244 & @xmath245 & @xmath246 & @xmath247 & @xmath119 + lllllllllllll obs . & & theor . & meas . & shift & & flux & w & n@xmath196 & n@xmath197 + & & ( ) & ( ) & ( km / s ) & ( @xmath198 ) & ( km / s ) & ( @xmath198 ph/@xmath199/s ) & ( m ) & ( @xmath200 ) & ( @xmath201 ) + 1 ( @xmath248mean ) & & 1.850 & & & @xmath249 & @xmath250 & @xmath251 & @xmath252 & @xmath220 & @xmath217 + 1 ( @xmath253mean ) & & 1.850 & 1.848(1 ) & @xmath202 & @xmath249 & @xmath250 & @xmath254 & @xmath148 & @xmath255 & @xmath256 + 1 ( @xmath248mean ) & & 1.780 & 1.776(1 ) & @xmath209 & @xmath257 & @xmath258 & @xmath134 & @xmath259 & @xmath260 & @xmath151 + 1 ( @xmath253mean ) & & 1.780 & 1.776(1 ) & @xmath209 & @xmath261 & @xmath262 & @xmath128 & @xmath156 & @xmath172 & @xmath263 + 1 ( mean@xmath75 c / s ) & & 1.850 & & & @xmath249 & @xmath250 & @xmath220 & @xmath264 & @xmath265 & @xmath218 + 1 ( mean@xmath05 c / s ) & & 1.850 & 1.849(1 ) & @xmath266 & @xmath249 & @xmath250 & @xmath234 & @xmath133 & @xmath267 & @xmath268 + 1 ( mean@xmath75 c / s ) & & 1.780 & 1.775(1 ) & @xmath269 & @xmath257 & @xmath258 & @xmath128 & @xmath270 & @xmath271 & @xmath145 + 1 ( mean@xmath05 c / s ) & & 1.780 & 1.777(1 ) & @xmath272 & @xmath273 & @xmath274 & @xmath206 & @xmath156 & @xmath172 & @xmath263 + 3 ( norm . ) & & 1.850 & & & @xmath249 & @xmath250 & @xmath218 & @xmath275 & @xmath251 & @xmath217 + 3 ( dip ) & & 1.850 & 1.850(1 ) & @xmath235 & @xmath249 & @xmath250 & @xmath276 & @xmath277 & @xmath206 & @xmath278 + 3 ( norm . ) & & 1.780 & 1.777(1 ) & @xmath272 & @xmath249 & @xmath250 & @xmath279 & @xmath280 & @xmath281 & @xmath226 + 3 ( dip ) & & 1.780 & 1.778(1 ) & @xmath228 & @xmath249 & @xmath250 & @xmath254 & @xmath282 & @xmath283 & @xmath145 + c r r r r r r r r r observation & @xmath284 & @xmath285 & @xmath286 & @xmath287 & @xmath288 + 1 & @xmath289 & @xmath290 & @xmath291 & @xmath292 & @xmath293 + 2 & @xmath294 & & @xmath295 & @xmath296 & + 3 & @xmath297 & @xmath298 & @xmath299 & @xmath300 & @xmath301 + 4 & @xmath302 & @xmath303 & @xmath304 & @xmath305 & @xmath306 + | we observed the bright phase of the 2003 outburst of the galactic black hole candidate h 1743@xmath0322 in x - rays simultaneously with _
chandra _ and _ rxte _ on four occasions .
the _ chandra_/hetgs spectra reveal narrow , variable ( he - like ) fe xxv and ( h - like ) fe xxvi resonance absorption lines . in the first observation
, the fe xxvi line has a fwhm of @xmath1 km / s and a blue - shift of @xmath2 km / s , suggesting that the highly ionized medium is an outflow .
moreover , the fe xxv line is observed to vary significantly on a timescale of a few hundred seconds in the first observation , which corresponds to the keplerian orbital period at approximately @xmath3 ( where @xmath4 ) .
our models for the absorption geometry suggest that a combination of geometric effects and changing ionizing flux are required to account for the large changes in line flux observed between observations , and that the absorption likely occurs at a radius less than @xmath3 for a @xmath5 black hole .
viable models for the absorption geometry include cyclic absorption due to an accretion disk structure , absorption in a clumpy outflowing disk wind , or possibly a combination of these two .
if the wind in h 1743@xmath0322 has unity filling factor , the highest implied mass outflow rate is 20% of the eddington mass accretion rate .
this wind may be a hot precursor to the seyfert - like , outflowing `` warm absorber ''
geometries recently found in the galactic black holes gx 339 - 4 and xte j1650@xmath0500 .
we discuss these findings in the context of ionized fe absorption lines found in the spectra of other galactic sources , and connections to warm absorbers , winds , and jets in other accreting systems .
[ firstpage ] |
strong - interaction measurements at the linear collider ( lc ) will form an important component of the physics programme .
the collider offers the possibility of testing qcd at high energy scales in the experimentally clean , theoretically tractable @xmath0 environment .
in addition , virtual @xmath1 interactions will be delivered free by nature , and a dedicated @xmath1 collider is an additional option , allowing detailed measurements of the relatively poorly understood photon structure .
the benchmark physics main topics are : * precise determination of the strong coupling @xmath2 .
* measurement of the @xmath3 evolution of @xmath2 and constraints on the gut scale . *
measurement of the total @xmath1 cross section and the photon structure function ; these issues are discussed elsewhere @xcite .
the current precision of individual @xmath2 measurements is limited at best to several per cent @xcite .
since the uncertainty on @xmath2 translates directly into an uncertainty on perturbative qcd ( pqcd ) predictions , especially for high - order multijet processes , it would be desirable to achieve much better precision .
in addition , since the weak and electromagnetic couplings are known with much greater relative precision , the error on @xmath2 represents the dominant uncertainty on our ` prediction ' of the scale for grand unification of the strong , weak and electromagnetic forces @xcite
. here we will refer to the conventional yardstick of @xmath2 quoted at the @xmath4 mass scale , @xmath5 , unless explicitly stated otherwise .
several techniques for @xmath5 determination will be available at the lc : the determination of @xmath5 from event ` shape ' observables that are sensitive to the 3-jet nature of the particle flow has been pursued for 2 decades and is generally well understood @xcite . in this method one usually forms a differential distribution , makes corrections for detector and hadronisation effects , and fits a pqcd prediction to the data , allowing @xmath5 to vary .
examples of such observables are the thrust , jet masses and jet rates .
the latest generation of such @xmath5 measurements , from slc and lep , has shown that statistical errors below the 1% level can be obtained with samples of a few tens of thousands of hadronic events . with the current lc design luminosities of a few @xmath6 @xmath7/@xmath8/s at 500 , 800 or 1000 gev
, hundreds of thousands of e@xmath9e@xmath10 @xmath11 @xmath12 events would be produced each year , and a statistical error on @xmath5 below the 0.5% level could be achieved .
detector systematic errors , which relate mainly to uncertainties on the corrections made for acceptance and resolution effects and are observable - dependent , are under control in today s detectors at the @xmath13 = 1 - 4% level @xcite .
if the lc detector is designed to be very hermetic , with good tracking resolution and efficiency , as well as good calorimetric jet energy resolution , all of which are required for the search for new physics processes , it seems reasonable to expect that the detector - related uncertainties can be beaten down to the @xmath13 @xmath14 1% level or better .
e@xmath9e@xmath10 @xmath11 @xmath15 , @xmath16 , or @xmath17 events will present significant backgrounds to @xmath12 events for qcd studies , and the selection of a highly pure @xmath12 event sample will not be quite as straightforward as at the @xmath4 resonance .
the application of kinematic cuts would cause a significant bias to the event - shape distributions , necessitating compensating corrections at the level of 25% @xcite .
more recent studies have shown @xcite that the majority of @xmath16 events can be excluded without bias by using only events produced with right - handed electron beams for the @xmath5 analysis .
furthermore , the application of highly - efficient @xmath18-jet tagging can be used to reduce the @xmath17 contamination to the 1% level .
after statistical subtraction of the remaining backgrounds ( the @xmath15 and @xmath16 event properties have been measured accurately at slc and lepi / ii ) , the residual bias on the event - shape distributions is expected to be under control at the better than 1% level on @xmath5 .
additional corrections must be made for the effects of the smearing of the particle momentum flow caused by hadronisation .
these are traditionally evaluated using monte carlo models .
the models have been well tuned at slc and lep and are widely used for evaluating systematic effects .
the size of the correction factor , and hence the uncertainty , is observable dependent , but the ` best ' observables measured at the @xmath4 have uncertainties as low as @xmath13 @xmath14 1% .
furthermore , one expects the size of these hadronisation effects to diminish with c.m .
energy at least as fast as 1/@xmath19 .
hence 10%-level corrections at the @xmath4 should dwindle to 1%-level corrections at @xmath19 @xmath20 500 gev , and the associated uncertainties will be substantially below the 1% level on @xmath5 .
this has been confirmed by explicit simulations using pythia @xcite .
currently pqcd calculations of event shapes are available complete only up to @xmath21 , although resummed calculations are available for some observables @xcite .
one must therefore estimate the possible bias inherent in measuring @xmath5 using the truncated qcd series .
though not universally accepted , it is customary to estimate this from the dependence of the fitted value on the qcd renormalisation scale , yielding a large and dominant uncertainty of about @xmath13 @xmath14 @xmath226% @xcite .
since the missing terms are @xmath23 , and since @xmath2(500 gev ) is expected to be about 25% smaller than @xmath5 , one expects the uncalculated contributions to be almost a factor of two smaller at the higher energy .
however , translating to the yardstick @xmath5 yields an uncertainty of @xmath225% , only slightly smaller than currently .
therefore , although a 1%-level measurement is possible experimentally , it will not be realised unless @xmath23 contributions are calculated .
there is reasonable expectation that this will be achieved within the next 5 years @xcite .
the dependence of the e@xmath9e@xmath10 @xmath11 @xmath17 production cross section , @xmath25 , on the top - quark mass , @xmath26 , and on @xmath5 are discussed elsewhere @xcite . in order to optimise the precision on the @xmath26 measurement near threshold
it is desirable to input a precise @xmath5 measurement from elsewhere .
furthermore , the current theoretical uncertainty on @xmath25 translates into @xmath13 = @xmath27 .
hence , although extraction of @xmath5 from @xmath25 near threshold may provide a useful ` sanity check ' of qcd in the @xmath17 system , it does not appear currently to offer the prospect of a competitive measurement .
a preliminary study has also been made @xcite of the determination of @xmath5 from @xmath28 @xmath29 @xmath30 _ above _ threshold . for @xmath19 @xmath20 500 gev
the uncertainty on @xmath28 due to @xmath26 is around 0.5% .
the limiting precision on @xmath28 will be given by the uncertainty on the luminosity measurement .
if this is as good as 0.5% @xcite then @xmath5 could be determined with an experimental precision approaching 1% , which would be extremely valuable as a complementary precision measurement from the @xmath17 system .
a giga @xmath4 sample offers two additional options for @xmath5 determination via measurements of the inclusive ratios @xmath31 and @xmath32 .
both are indirectly proportional to @xmath2 , and hence require a very large event sample for a precise measurement .
for example , the current lep data sample of 16 m @xmath4 yields an error of 2.5% on @xmath5 from @xmath31 .
the statistical error could , naively , be pushed to below the @xmath13 = 0.4% level , but systematic errors arising from the hadronic and leptonic event selection will probably limit the precision to 0.8% @xcite .
this would be a very precise , reliable measurement . in the case of
@xmath32 the experimental precision from lep and cleo is already at the 1% level on @xmath5 .
however , there has been considerable debate about the size of the theoretical uncertainties , with estimates as large as 5% @xcite .
if this situation is clarified , and the theoretical uncertainty is small , @xmath32 may offer a further 1%-level @xmath5 measurement .
in the preceeding sections we discussed the expected attainable precision on the yardstick @xmath5 .
translation of the measurements of @xmath2(@xmath19 ) ( @xmath33 ) to @xmath5 requires the assumption that the ` running ' of the coupling is determined by the qcd @xmath34 function .
however , since the logarithmic decrease of @xmath2 with @xmath19 is an essential component of qcd , reflecting the underlying non - abelian dynamics , it is vital also to test this @xmath19-dependence explicitly .
such a test would be particularly interesting if new coloured particles were discovered , since deviations from qcd running would be expected at energies above the threshold for pair - production of the new particles .
furthermore , extrapolation of @xmath2 to very high energies of the order of @xmath35 gev can be combined with corresponding extrapolations of the dimensionless weak and electromagnetic couplings in order to constrain the coupling - unification , or gut , scale @xcite .
hence it would be desirable to measure @xmath2 _ in the same detector , with the same technique , and by applying the same treatment to the data _ at a series of different energies @xmath19 , so as to maximise the lever - arm for constraining the running .
simulated measurements of @xmath2(@xmath19 ) at @xmath19 = 91 , 500 and 800 gev are shown in fig .
[ tdralphas ] , together with existing measurements which span the range @xmath36 gev .
the highest - energy measurements are currently provided by lepii .
the point at @xmath19 = 91 gev is based on the @xmath37 technique , and those at 500 and 800 gev are based on the event shapes technique .
the last two include the current theoretical uncertainty , which yields a total error on each point equivalent to @xmath13 = 4% .
it is clear that the lc data would add significantly to the lever - arm in @xmath19 , and would allow a substantially improved extrapolation to the gut scale . with @xmath38 ;
sample @xmath19 values ( gev ) are indicated.,width=529 ]
limited space allows only a brief mention of several other important topics @xcite : * _ hard gluon radiation in @xmath17 events _ would allow several tests of the strong dynamics of the top quark @xcite : test of the flavour - independence of strong interactions ; limits on anomalous chromo - electric and/or chromo - magnetic dipole moments @xcite ; determination of the running @xmath26 . *
_ soft gluon radiation in @xmath17 events _ is expected to be strongly regulated by the large mass and width of the top quark .
precise measurements of gluon radiation patterns in @xmath39 events would provide additional constraints on the top decay width @xcite .
* _ polarised electron ( and positron ) beams _ can be exploited to test symmetries using multi - jet final states . for polarized @xmath0 annihilation to three hadronic jets one can define @xmath40 , which correlates the electron - beam polarization vector @xmath41 with the normal to the three - jet plane defined by @xmath42 and @xmath43 , the momenta of the two quark jets .
if the jets are ordered by momentum ( flavour ) the triple - product is cp even ( odd ) and t odd .
standard model t - odd contributions of this form are expected @xcite to be immeasurably small , and limits have been set for the @xmath44 system @xcite . at the lc
these observables will provide an additional search - ground for anomalous effects in the @xmath39 system . * _ the difference between the particle multiplicity in heavy- ( @xmath45 ) and light - quark events _ is predicted @xcite to be independent of c.m . energy .
precise measurements have been made at the @xmath4 , but measurements at other energies are statistically limited in precision , rendering a limited test of this important prediction .
high - precision measurements at the lc would add the lever - arm for a powerful test .
* _ colour reconnection and bose - einstein correlations _ are important to study precisely since they may affect the precision with which the masses of heavy particles , such as the @xmath46 and top - quark , can be reconstructed kinematically via their multijet decays @xcite . * _ hadronisation studies and renormalon physics _ can be explored via measurements of event - shape observables over a range of @xmath19 values .
there is a rich programme of qcd studies at the linear collider .
precision measurements of the strong coupling and of the strong dynamics of the @xmath17 system will complement inclusive measurements that will be made at the lhc . for details ,
burrows , ` testing qcd at high energy @xmath0 colliders : @xmath47 tev ' , proc .
workshop on physics and experiments with linear colliders , sept .
8 - 12 1995 , morioka - appi , japan , world scientific 1996 , ed .
a. miyamoto _
et al _ , p. 179 . | i summarise the qcd programme at the high - energy @xmath0 linear collider , as reported in the tesla tdr and linear collider physics resource book . |
let @xmath1 be a number field .
this investigation centres upon the distribution of @xmath1-rational points on conic bundle surfaces @xmath2 .
these are defined to be projective non - singular surfaces @xmath3 defined over @xmath1 , which are equipped with a dominant @xmath1-morphism @xmath4 , all of whose fibres are conics .
a summary of our knowledge concerning the arithmetic of conic bundle surfaces can be found in the article by colliot - thlne @xcite . if @xmath5 denotes the anticanonical divisor of @xmath3 , then the degree of @xmath3
is defined to be the self - intersection number @xmath6 , where @xmath7 is the number of geometric fibres above @xmath8 that are degenerate . when @xmath9 it is known that the hasse principle holds for these surfaces , and furthermore , that such @xmath3 are @xmath1-rational as soon as they possess @xmath1-rational points . when @xmath10 , so that @xmath3 has degree @xmath11 , it has been shown by iskovskikh ( * ? ? ?
* proposition 1 ) that two basic cases arise .
either @xmath5 is not ample , in which case @xmath3 is @xmath1-birational to a generalised chtelet surface , or else @xmath5 is ample , in which case @xmath3 is a non - singular quartic del pezzo surface . our interest lies with the quantitative arithmetic of degree @xmath11 conic bundle surfaces . when @xmath12 and the anticanonical height function @xmath13 is associated , we seek to determine the asymptotic behaviour of the counting function @xmath14 as @xmath15 , for a suitable zariski open subset @xmath16 .
the conjecture that drives our work is due to manin @xcite .
let @xmath17 be the picard group of @xmath3 .
still under the assumption that @xmath18 is non - empty , this conjecture predicts the existence of a positive constant @xmath19 such that @xmath20 as @xmath21 .
peyre @xcite has given a conjectural interpretation of @xmath19 in terms of the geometry of @xmath3 .
until very recently we were not in possession of a single conic bundle surface of degree @xmath11 for which this refined conjecture could be established
. henceforth we will be interested in the case @xmath22 .
in joint work with peyre @xcite , the authors have made a study of generalised chtelet surfaces , ultimately establishing for a family of such surfaces that fail to satisfy weak approximation .
the aim of the present investigation is to produce a satisfactory treatment of a non - singular del pezzo surface of degree @xmath11 with a conic bundle structure which admits a section over @xmath0 .
such surfaces are defined as the intersection of two quadrics in @xmath23 .
when @xmath24 , we may assume that @xmath25 is cut out by the system @xmath26 for quadratic forms
@xmath27 $ ] such that the jacobian matrix @xmath28 has full rank throughout @xmath3 .
let @xmath29 be a norm .
given a point @xmath30 \in { { { \mathbb{p}}_{{\mathbb{q}}}^4}}({\mathbb{q}}),$ ] with @xmath31 such that @xmath32 , we let @xmath33 . then @xmath34 is the anticanonical height metrized by the choice of norm .
any line contained in @xmath3 that is defined over @xmath0 will automatically contribute @xmath35 to @xmath36 , for an appropriate constant @xmath37 .
hence it natural to take @xmath38 to be the open subset formed by deleting the @xmath39 lines from @xmath3 .
the best evidence that we have for in the setting of non - singular surfaces of the shape is due to salberger . in work communicated at the conference `` higher dimensional varieties and rational points '' at budapest in 2001
, he establishes the upper bound @xmath40 for any @xmath41 . here , as throughout our work , we allow the implied constant to depend on the choice of @xmath42 .
the manin conjecture has received a great deal of attention in the context of singular del pezzo surfaces of degree @xmath43 and @xmath11 .
an account of recent progress can be found in the second author s survey @xcite .
there is general agreement among researchers that the level of difficulty in establishing the expected asymptotic formula for del pezzo surfaces increases as the degree decreases or as the singularities become milder . among the non - singular del pezzo surfaces , those of degree at least @xmath44
are all toric and so are handled by the work of batyrev and tschinkel @xcite . in @xcite the first author gave the earliest satisfactory treatment of a non - singular del pezzo surface of degree @xmath45 . as highlighted by swinnerton - dyer ( * ? ? ?
* question 15 ) , it has become something of milestone to establish the manin conjecture for a single non - singular del pezzo surface of degree @xmath43 or @xmath11 . in this paper
we will be concerned with a quartic del pezzo surface @xmath46 of the shape , with @xmath47 in particular @xmath3 contains obvious lines defined over @xmath0 , from which it follows that @xmath3 is @xmath0-rational .
this fact is recorded by colliot - thlne , sansuc and swinnerton - dyer ( * ? ? ? * proposition 2 ) , for example .
if @xmath48 is a line in @xmath3 defined over @xmath0 then it is a simple consequence of the fact that any plane through @xmath48 must cut out a pair of lines @xmath49 on each quadric @xmath50 defining @xmath3 , and the intersection @xmath51 meets @xmath3 in exactly one further point , which is defined over @xmath0 .
let @xmath52 be the norm on @xmath53 given by @xmath54 for every @xmath55 such that @xmath56 \in x$ ] , and furthermore @xmath57 , where @xmath58 is the vector obtained by permuting the variables @xmath59 and leaving @xmath60 fixed .
it would be possible to work instead with the norm @xmath61 , but not without introducing extra technical difficulties that we wish to suppress in the present investigation .
we are now ready to reveal our main result .
we have @xmath62 where @xmath63 is the constant predicted by peyre . during the final preparation of this paper
, the authors have learnt of independent work by fok - shuen leung @xcite on the conic bundle surface given by and .
this sharpens salberger s estimate in , ultimately providing upper and lower bounds for @xmath36 that are of the expected order of magnitude .
our result supersedes this , and confirms the estimate predicted by manin and peyre in for the non - singular quartic del pezzo surface under consideration .
the fact that the picard group has rank @xmath45 , as needed to verify the power of @xmath64 , will be established in due course .
the proof of our theorem is long and complicated .
we therefore dedicate the remainder of this introduction to surveying some of its key ingredients and indicating some obvious lines for further enquiry .
given any non - singular surface defined by the system , it is possible to define a pair of conic bundle morphisms @xmath65 , for @xmath66 .
specifically , for any @xmath67 , one takes @xmath68 , & \mbox{if $ ( x_0,x_2)\neq(0,0)$ , } \cr [ x_3,x_1 ] , & \mbox{if $ ( x_1,x_3)\neq(0,0)$ , } \end{cases}\ ] ] and @xmath69 , & \mbox{if $ ( x_0,x_3)\neq(0,0)$ , } \cr [ x_2,x_1 ] , & \mbox{if $ ( x_1,x_2)\neq(0,0)$. } \end{cases}\ ] ] for a given point @xmath70 of height @xmath71 , it follows from the general theory of height functions that there exists an index @xmath72 such that @xmath73 for some @xmath74 of height @xmath75 .
the idea is now to count rational points of bounded height on the fibres @xmath76 and @xmath77 , uniformly for points @xmath74 of height @xmath75 .
this is the strategy adopted by salberger in his proof of . in the present situation , with the quadratic form ,
the fibres that we need to examine have the shape @xmath78 for coprime @xmath79 .
it is clear that @xmath80 is a non - singular plane conic when the discriminant @xmath81 is non - zero .
the reduction of the counting problem to one involving the family of conics is carried out in
[ s : prelim ] , where we have avoided using the height machinery by doing things in a completely explicit manner .
as is well - known there is a group homomorphism @xmath82 , which to a divisor class @xmath83 associates the intersection number of @xmath84 with a fibre .
the kernel of this map is generated by the `` vertical '' divisors , which up to linear equivalence are the irreducible components of the fibres .
since the non - singular fibres are all linearly equivalent , it follows that @xmath17 has rank @xmath85 , where @xmath86 is the number of split singular fibres above closed points of @xmath87 . in our case
there are three closed points , corresponding to the irreducible factors @xmath88 and @xmath89 of @xmath90 .
since each singular fibre is split it follows that @xmath91 , as previously claimed .
roughly speaking , as one ranges over values of @xmath92\in { { { \mathbb{p}}_{{\mathbb{q}}}^1}}({\mathbb{q}})$ ] of height @xmath75 , one expects there to be about @xmath93 rational points on @xmath94 with height restricted to appropriate intervals .
the preliminary reduction to @xmath92\in~{{{\mathbb{p}}_{{\mathbb{q}}}^1}}({\mathbb{q}})$ ] of height @xmath75 is absolutely pivotal here : it is only through this device that we can cover @xmath95 with a satisfactory number of divisors .
were we charged instead with establishing an upper bound like , our analysis would now be relatively straightforward , thanks to the control over the growth rate of rational points on conics afforded by the second author s joint work with heath - brown ( * ? ? ?
* theorem 6 ) .
a key aspect of this estimate is that it is uniform in the height of the conic , becoming sharper as the discriminant grows larger . in [ s : ham ] we will take advantage of these arguments to eliminate certain awkward ranges for @xmath96 in . obtaining an asymptotic formula
is a far more exacting task . using the large sieve inequality ,
serre @xcite has shown that most plane conics defined over @xmath0 do nt contain rational points .
this phenomenon might pose problems for us , given that we want a uniform asymptotic formula for a zariski dense set of rational points on the fibres .
our choice of surface has been tailored to guarantee that this does nt happen , since the corresponding fibres always contain the rational point @xmath97.\ ] ] in the classical manner we can use this point to parametrise all of the rational points on the conic , which ultimately leads us to evaluate asymptotically the number of points belonging to a @xmath98-dimensional sublattice @xmath99 which are constrained to lie in an appropriate region @xmath100 . both @xmath101 and @xmath102
depend on the parameters @xmath103 and @xmath104 , so this estimate needs to be achieved with a sufficient degree of uniformity .
assuming that @xmath102 has piecewise continuous boundary , we would ideally like to apply the familiar estimate @xmath105 where @xmath106 denotes the perimeter of @xmath102 .
unhappily this estimate is too crude for our purposes .
we will use poisson summation to make the error term explicit .
the reduction of the problem to a lattice point counting problem is carried out in [ sec : par ] and its execution is the subject of [ sec:1][s:2 ] .
the one outstanding task , which is the focus of [ s:3 ] , is to evaluate asymptotically the main term arising from the lattice point counting problem .
it turns out that this involves a sum of the shape @xmath107 for a certain multiplicative arithmetic function @xmath108 that is very similar to the ordinary divisor function @xmath109 .
the problem of determining the average order of the divisor function as it ranges over the values of polynomials has enjoyed considerable attention in the literature . in the setting of binary forms current technology
has limited us to handling forms of degree at most @xmath11 . when @xmath110 daniel @xcite has dealt with the case of irreducible binary quartic forms .
we need to extend this argument to deal with a more general class of arithmetic functions and to binary quartic forms that are no longer irreducible .
we have found it convenient to corral the necessary estimates into a separate investigation @xcite , which is of a more technical nature .
the facts that we will need are recalled in [
s : tech ] .
our main goal in this paper is to outline a general strategy for proving the manin conjecture for non - singular del pezzo surfaces of degree @xmath11 equipped with a conic bundle structure admitting a section over @xmath0 . in order to minimise the length and technical difficulty of the work , which is already considerable , we have chosen to illustrate the approach by selecting a concrete surface to work with .
it is likely that the argument we present can be adapted to handle other conic bundle surfaces .
for example , consider the family of surfaces , with @xmath111 it is easy to check that @xmath3 will be non - singular if and only if @xmath112 and @xmath113 .
as we ve already mentioned , our success with is closely linked to the existence of an obvious rational point on all of the fibres .
this is equivalent to the map @xmath114 being surjective . a necessary and sufficient condition for ensuring this is that the morphism @xmath115 should admit a section over @xmath0 .
this fact has a long history and can be traced back to work of lewis and schinzel @xcite .
even when the conic bundle surface admits a section , however , there remains considerable work to be done in handling a general diagonal form @xmath116 as above .
the corresponding fibres will take the shape @xmath117 for binary forms @xmath118 and @xmath119 .
thus , at the very least , one needs analogues of our investigation @xcite for the case in which @xmath120 factors over @xmath0 as the product of two quadratic forms or splits completely . throughout our work @xmath121 will denote the set of positive integers . if @xmath122 then we write @xmath123 for the greatest common divisor of @xmath96 and @xmath92=ab/\gcd(a , b)$ ] for the least common multiple .
we set @xmath124 for the odd part of the greatest common divisor , and use the symbol @xmath125 to indicate a summation in which all the variables of summation are restricted to odd integers .
furthermore , our work will involve the arithmetic functions @xmath126 finally , in terms of the parameter @xmath127 , we set @xmath128 we will reserve @xmath37 for a generic absolute positive constant , whose value is always effectively computable .
this investigation was undertaken while the second author was visiting the _ universit paris 6 pierre et marie curie _ and the _ universit paris 7 denis diderot_. the hospitality and financial support of these institutions is gratefully acknowledged .
it is a pleasure to thank olivier wittenberg for useful conversations relating to the geometry of conic bundle surfaces and the anonymous referee for numerous pertinent remarks .
while working on this paper the second author was supported by epsrc grant number ` ep / e053262/1 ` .
our work requires a number of auxiliary results , ranging from basic estimates using the geometry of numbers to more sophisticated results concerning the divisor problem for binary quartic forms .
as made clear in the introduction to this paper , our proof of the theorem uses the conic bundle structure in order to focus the effort on a family of curves of low degree and rather low height .
the following result is due to heath - brown ( * ? ? ?
* lemma 3 ) , and deals with the situation for lines in @xmath129 , the rational points on which basically correspond to integer lattices of rank @xmath98 .
[ lem : lattice ] let @xmath130 be a lattice of rank @xmath98 and determinant @xmath131 , and let @xmath100 be any region with piecewise continuous boundary
. then we have @xmath132 our next uniform upper bound is extracted from joint work of the second author with heath - brown ( * ? ? ?
* corollary 2 ) , and handles the case of non - singular plane conics .
[ lem : bhb ] let @xmath133 be a non - singular conic .
assume that the underlying quadratic form has matrix of determinant @xmath134 , and that the @xmath135 minors have greatest common divisor @xmath136 .
then we have @xmath137 taking @xmath138 in lemma [ lem : bhb ] , we retrieve the well - known fact that a non - singular plane conic @xmath133 contains @xmath139 rational points of height @xmath127 . in the setting of polynomials in only one variable
there is a well - known result due to nair @xcite which provides upper bounds for the average order of suitable non - negative arithmetic functions as they ranges over the values of the polynomial .
taking advantage of the authors refinement @xcite of this work we have the following result .
[ lem : sky ] let @xmath41 , let @xmath140 and let @xmath141 .
then for @xmath142 there exists an absolute constant @xmath37 such that @xmath143 define the multiplicative arithmetic function @xmath144 for any prime power @xmath145 .
we have @xmath146 for any @xmath147 .
let @xmath148 denote the sum that is to be estimated .
on replacing @xmath149 by @xmath150 we find that @xmath151 now for any @xmath152 it is clear that the polynomial @xmath153 has degree @xmath154 , that it has discriminant @xmath155 and that it has no fixed prime divisor if @xmath156 or @xmath103 is even .
if @xmath157 and @xmath103 is odd then @xmath98 is a fixed prime divisor of @xmath158 .
but then we may break the sum into two sums according to whether @xmath86 is even or odd and make a corresponding change of variables , absorbing the additional factor @xmath159 into an implied constant .
an application of ( * ? ? ?
* theorem 2 ) now reveals that @xmath160 for @xmath161 , where @xmath162 denotes the number of roots modulo @xmath163 of the congruence @xmath164 .
recall the well - known inequalities @xmath165 for any prime power , and @xmath166 if @xmath167 .
then it follows that @xmath168 for a suitable absolute constant @xmath37 . on noting that @xmath169 for @xmath170 , this therefore concludes the proof of the lemma .
we will also need a version of nair s lemma for binary forms @xmath171 $ ] .
let @xmath172 denote the maximum modulus of its coefficients and set @xmath173 ^ 2\,:\ , \begin{array}{ll } \gcd(n_1,n_2,m)=1 \\ f(n_1,n_2)\equiv 0 \bmod{m}\end{array}\big\},\ ] ] for any @xmath174 .
the following estimate is deduced from ( * ? ? ?
* corollary 1 ) by specialising to the generalised divisor function @xmath175 .
[ lem : nair ] let @xmath41 and let @xmath176 $ ] be a non - zero binary form of degree @xmath177 , with @xmath178 and @xmath179 .
then we have @xmath180 where @xmath181 throughout this section we let @xmath72 denote a generic element from the set @xmath182 .
let @xmath183 $ ] be binary forms , with @xmath184 and @xmath185 , such that @xmath186 are non - proportional and @xmath187 is irreducible over @xmath0 .
let @xmath188 ^ 2 $ ] be a convex region whose boundary is defined by a piecewise continuously differentiable function .
assume that @xmath189 and @xmath190 for every @xmath191 .
the workhorse in this paper is an asymptotic formula for sums akin to @xmath192 where @xmath149 is the divisor function and @xmath193 .
in fact the arguments that appear in our work call for a rather more general type of sum .
suppose that @xmath194 is the dirichlet convolution of the divisor function with a multiplicative arithmetic function @xmath195 that satisfies @xmath196 let @xmath197 ^ 4 $ ] be a region cut out by a finite number of hyperplanes with absolutely bounded coefficients .
for any @xmath198 we define @xmath199 then we will encounter sums of the shape @xmath200 if one takes @xmath201 ^ 4 $ ] and @xmath202 a multiple of @xmath3 then @xmath203 .
moreover , if one takes @xmath204 then one arrives at exactly the sum involving @xmath149 that was mentioned above . for any prime @xmath205 and @xmath206 , let @xmath207 and @xmath208 here , we follow the convention that @xmath209 if and only if @xmath210 . the following asymptotic formula is established in our companion paper ( * ? ? ?
* corollary 3 ) .
[ lem : uniform ] let @xmath41 .
assume that @xmath211 .
then we have @xmath212 where @xmath213 and @xmath214 ^ 4 : \max\{v_1,v_2,v_3\}{\leqslant}v_4{\leqslant}\frac{1}{2}\big\}.\ ] ] we will ultimately apply lemma [ lem : uniform ] with @xmath215 equal to the discriminant @xmath90 of the conic .
the exponent of @xmath216 in the lemma reflects the fact that we are dealing with binary forms with three irreducible factors .
the attentive reader will observe a correlation with our description of the rank of @xmath17 in [ s : intro ] .
we will be interested in applications of lemma [ lem : uniform ] when @xmath217 is the multiplicative arithmetic function defined via @xmath218 it easily follows that @xmath194 , with @xmath219 in particular @xmath220 for any @xmath41 , whence holds .
recall the definition of the quadratic forms @xmath221 from and .
we begin this section by relating @xmath36 to the quantity @xmath222 in which the main difference is that the count is restricted to positive integer solutions .
this is achieved in the following result .
[ lem : red-1 ] we have @xmath223 it follows from the calculation of the lines in [ s : local ] that the condition @xmath224 is equivalent to @xmath225 , for any @xmath30\in x$ ] .
we therefore deduce that @xmath226 it follows from the equation @xmath227 that the condition @xmath32 is equivalent to @xmath228 for any vector @xmath229 in which we are interested .
furthermore , for @xmath229 such that @xmath230 , it is clear from that @xmath231 we proceed to consider the contribution from the vectors @xmath232 for which @xmath233 and @xmath234 let us begin with the case @xmath235 , for @xmath236 .
this hyperplane section leads us to estimate the number of solutions to an equation of the form @xmath237 with @xmath238 and @xmath239 .
an application of lemma [ lem : bhb ] therefore yields a contribution of @xmath240 from this case . turning to the contribution from the case @xmath241 and @xmath242 ,
our task is to estimate the number of primitive vectors @xmath243 such that @xmath244 with @xmath245 .
it is a simple matter to show that there is an overall contribution of @xmath240 from this case . bringing everything together
we have therefore shown that @xmath246 it remains to show that we can restrict attention to positive values of @xmath247 . but this follows on noting that @xmath116 is invariant under sign changes in the components of @xmath229 , and @xmath248 demands that @xmath249 should share the same sign as @xmath250 .
thus the above cardinality is @xmath251 , with @xmath252 given by , and the lemma follows .
the next stage of the argument involves parametrising the solutions to the equation @xmath253 .
it is a simple exercise to show that the set of @xmath254 such that @xmath255 and @xmath256 , is in bijection with the set of @xmath257 such that @xmath258 the relation being given by @xmath259 define @xmath260 on recalling the definition , it is easy to see that @xmath261 where @xmath229 denotes @xmath262 .
given integers @xmath96 , recall the definition of the plane conic @xmath80 .
writing @xmath263 and @xmath264 , it follows that @xmath265 is the number of @xmath266 such that holds , with @xmath267 for @xmath268 we define the quantities @xmath269 where @xmath270 is the conic .
there is an abuse of notation here , in that we are interested in vectors @xmath271 that lie in the affine cone above @xmath94 , rather than points @xmath272\in c_{a , b}({\mathbb{q}})$ ] in which coordinates are chosen so that @xmath273 .
note that we must automatically have @xmath274 in the definition of @xmath275 since @xmath276 .
hence @xmath277 we have @xmath278 and we are now ready to establish the following result .
[ lem : red-2 ] we have @xmath279 where @xmath275 is given by .
fundamental to our argument is the observation @xmath280 if and only if @xmath281 . from this
it easily follows that @xmath282 where @xmath283 to establish the lemma it therefore suffices to show that @xmath284 without loss of generality we may view @xmath285 as being the overall contribution from vectors such that @xmath286 the remaining @xmath43 cases being handled in an identical fashion . arguing as above , we find that @xmath287 the overall contribution to the right hand side from the case @xmath288 , so that @xmath289 , is clearly @xmath290 .
we will estimate the remaining contribution by first fixing a choice of @xmath291 and @xmath103 , and then summing over all available @xmath292 and @xmath293 .
first we note that @xmath291 is necessarily odd in any @xmath11-tuple @xmath294 . for fixed @xmath295
the form @xmath296 defines a positive definite binary quadratic form of discriminant @xmath297 .
in particular the total number of available @xmath298 is bounded by the number of representations @xmath299 of @xmath300 by a complete system of inequivalent forms of discriminant @xmath84 .
it easily follows that @xmath301 with @xmath302 . by the classical theory of binary quadratic forms we have @xmath303 where @xmath304 is the kronecker symbol .
the last equality follows since @xmath305 .
an application of lemma [ lem : sky ] therefore reveals that there is an absolute constant @xmath37 such that @xmath306 where @xmath307 is given by .
this completes the proof of the lemma .
it will greatly facilitate our arguments if we can reduce the range of summation for @xmath96 in the statement of lemma [ lem : red-2 ] .
the main aim of this section is to establish the following result .
[ lem : ab - small ] let @xmath308 and recall the definition of @xmath309 .
then we have the following estimates @xmath310 applying the second estimate in lemma [ lem : ab - small ] with @xmath311 and inserting this into lemmas [ lem : red-1 ] and [ lem : red-2 ] , we easily conclude that @xmath312 achieving a corresponding lower bound for @xmath36 is straightforward
. we will have to work much harder to deduce an asymptotic formula .
the main tool in our proof of this result is lemma [ lem : bhb ] . for the conic @xmath94 defined in we have @xmath313 and @xmath314 , since @xmath315 by assumption .
it therefore follows that @xmath316 thus there are two terms to consider , both of which we must sum over the relevant values of @xmath317 lemma [ lem : nair ] immediately implies that the first term contributes @xmath318 to the two sums in the lemma , thereby allowing us to concentrate on the contribution from the second term .
we will show that @xmath319 and @xmath320 for any @xmath321 . combining these with , one easily verifies that these estimates are enough to complete the proof of the lemma .
it therefore remains to establish and .
beginning with the estimation of the sum in , which we denote by @xmath322 , it will be convenient to set @xmath323 and @xmath324 .
in particular @xmath325 .
we deduce that @xmath326 but then it follows from combining lemma [ lem : sky ] with partial summation that @xmath327 for an absolute constant @xmath37 .
this completes the proof of .
it remains to estimate the sum in , which we also denote by @xmath322 .
it easily follows that @xmath328 on taking dyadic intervals for @xmath329 and @xmath330 , we see that the first sum here is @xmath331 by lemma [ lem : nair ] .
similarly , on setting @xmath323 and @xmath324 as above and taking dyadic intervals for @xmath332 and @xmath333 , we deduce from lemma [ lem : nair ] that the remaining sum is @xmath334 as required for .
this completes the proof of the lemma .
the sort of argument used to prove and will feature quite often in our work . recall the definitions of @xmath335 and @xmath336 .
it is easy to see that @xmath337 and @xmath338 has order of magnitude @xmath64 .
our next task is twofold .
firstly , we would like to be able to restrict attention to values of @xmath96 belonging to the set @xmath339 here @xmath37 is an absolute constant that when appearing as an exponent of @xmath335 , we will view as being sufficiently large to ensure that all of our error terms are satisfactory .
secondly , we wish to show that @xmath275 can be replaced by @xmath309 in lemma [ lem : red-2 ] , with a negligible error .
both of these objectives are achieved in the following result .
[ lem : red-3 ] we have @xmath340 where @xmath309 is given by and @xmath341 is given by . taken together with lemma [ lem : red-2 ] ,
an application of lemma [ lem : ab - small ] easily implies that @xmath342 let us proceed by indicating how to replace @xmath275 by @xmath309 in the summand .
using the fact that @xmath343 it is easy to see that @xmath344 but then it follows that we are interested in values of @xmath345 such that @xmath346 an argument based on the proof of lemma [ lem : ab - small ] reveals that we can replace @xmath275 by @xmath309 , as required . in order to restrict to a summation over @xmath347 we must consider the contribution from @xmath96 such that @xmath348 . in view of lemma [ lem : nair ] and
, one sees that the overall contribution is bounded by @xmath349 let us suppose without loss of generality that @xmath350 in the remaining sum . on writing @xmath351
, we see that the sum over @xmath96 is @xmath352 where @xmath353 is defined via @xmath354 one easily confirms that the proof of lemma [ lem : sky ] goes through with @xmath149 replaced by @xmath353 , whence the argument used to establish applies in this setting too .
thus it follows that the contribution to the main term in from @xmath355 is satisfactory .
this completes the proof of the lemma .
in this section we concern ourselves with estimating @xmath309 , as given by .
recall the definition of @xmath94 .
fundamental to our approach is the observation that for each @xmath268 , the conic @xmath94 always contains the rational point @xmath356 in .
we are therefore in a position to parametrise all of the rational points on the conic by considering the residual intersection with @xmath94 of an arbitrary line through @xmath356 .
define the binary quadratic forms @xmath357 for given @xmath122 such that @xmath315 and @xmath358 .
in particular we will always have @xmath359 .
one easily checks that @xmath360 the main aim of this section is to establish the following result .
[ lem : para-1 ] we have @xmath361 where @xmath362 .
recall the definitions and of @xmath94 and @xmath356 , respectively .
let @xmath363 this is the tangent line to @xmath94 along @xmath356 .
let @xmath364 denote the set of projective lines in @xmath129 that pass through @xmath356 , and let @xmath365 be the corresponding subset that are defined over @xmath0 .
we will write @xmath366 now let @xmath367 denote the open subset formed by deleting @xmath356 from the conic . as is well - known , the sets @xmath95 and @xmath368 are in bijection .
the general element of @xmath365 is given by @xmath369 for @xmath370 such that @xmath371 .
there is clearly a bijection between those lines for which @xmath372 and those for which @xmath373 .
we henceforth fix our attention on the latter . in order to have a point in @xmath368 we must insist that @xmath374 , which is clearly equivalent to the condition @xmath375 .
moreover , there are only @xmath290 points in @xmath368 that correspond to taking @xmath376 . to make the bijection between @xmath95 and @xmath368 completely explicit a routine calculation reveals that a point @xmath272 $ ] is in the intersection @xmath377 if and only if @xmath378 , or else @xmath379 , in the notation of . the first alternative leads us to the point @xmath356 , which is to be ignored .
the second alternative implies that @xmath380 by , in the equation for @xmath381 .
any vector @xmath382 that represents a point in @xmath383 is primitive if and only if @xmath384 .
writing @xmath385 as in the statement of the lemma , we therefore find that @xmath386 for any @xmath272\in ( u\cap l_{s , t})({\mathbb{q}})$ ] .
so far we have recorded an explicit bijection between elements of @xmath95 and @xmath368 . in deriving an expression for @xmath309 in terms of this bijection
we will need to restrict the corresponding values of @xmath387 that are to be considered .
specifically we will only be interested in the values of @xmath388 for which @xmath389 and the corresponding values of @xmath390 lie in the region defined by the inequalities @xmath391 finally , we will want to exclude the possibility that @xmath392 .
since there are only @xmath290 such values of @xmath387 to worry about , this therefore concludes the proof that @xmath393 our final task is to show that @xmath394 in this expression .
suppose that @xmath395 .
the height conditions imply that @xmath396 and @xmath373 .
but then it follows that we must necessarily have @xmath397 .
however we must also have @xmath398 , whence @xmath399 where we have written @xmath400 for ease of notation .
this contradiction establishes the lemma .
lemma [ lem : para-1 ] allows us to translate the underlying problem to one that involves counting primitive lattice points in a complicated region contained in @xmath401 .
before we proceed to consider this region in more detail , it will be necessary to gain a better understanding of the greatest common divisor @xmath385 .
[ lem : la ] let @xmath96 be coprime positive integers such that @xmath358 .
let @xmath387 be coprime integers and let @xmath362 .
then we have @xmath402 , where @xmath403 and @xmath404 in the notation of .
let us write @xmath405 for @xmath406 and @xmath407 odd .
observe that @xmath408 the precise value of @xmath409 will depend intimately on the @xmath98-adic valuations of @xmath410 and @xmath411 .
let us write @xmath412 for @xmath413 and @xmath414 odd .
it follows from that @xmath415 suppose first that @xmath416 .
then @xmath417 thus either @xmath418 , in which case @xmath419 , or else @xmath420 . in the latter case @xmath421 and it follows that @xmath422 differentiating according to whether @xmath423 and @xmath424 , or @xmath425 and @xmath426 , it is easily checked that @xmath427 turning to the odd part @xmath428 of @xmath385 , we deduce from that @xmath429 this completes the proof of the lemma .
note that the last equality in is only possible if @xmath430 since @xmath431 is divisible by @xmath11 if @xmath103 and @xmath104 are both odd .
define the three quadratic forms @xmath432 for any positive @xmath433 .
we may then write @xmath434 in , where @xmath435 . for any @xmath436
we define the region @xmath437 furthermore , we set @xmath438 and @xmath439 bringing together lemmas [ lem : para-1 ] and [ lem : la ] , we may deduce that @xmath440 where @xmath441 is the number of @xmath442 subject to the following conditions : 1 .
[ i:1 ] @xmath389 ; 2 .
[ i:2 ] @xmath443 ; 3 .
[ i:- ] @xmath444 and @xmath445 ; 4 .
[ i:5 ] @xmath446 and @xmath447 ; 5 .
[ i:6 ] @xmath448 and @xmath449 ; and 6 .
[ i:7 ] the @xmath98-adic orders of @xmath387 are determined by @xmath409 via .
here we recall the convention that the symbol @xmath125 implies a restriction to odd parameters in the summation and we note that and are together equivalent to @xmath450 in the notation of .
the way forward should now be clear . for given values of @xmath96 , and appropriate values of @xmath451 and @xmath452 , we must attempt to produce an asymptotic formula for the number of primitive lattice points in a complicated region in @xmath401 .
we will do so using exponential sums .
the first step , however , is to remove the coprimality conditions that go into the definition of @xmath453 .
let @xmath454 and @xmath455 such that @xmath456 .
define the set @xmath457\mid s$ , $ k_2{\lambda}_2 \mid s - at$ and $ \ell \mid t$}\}.\ ] ] then @xmath458 is an integer sublattice of rank @xmath98 , whose determinant is given by the following result .
[ lem : det ] we have @xmath459 writing @xmath460 and @xmath461 it follows that @xmath462 , where @xmath463 is the set of @xmath464 such that @xmath465 and @xmath466 , with @xmath467 equal to @xmath468 .
the proof of the lemma then follows on noting that @xmath469 , since @xmath103 and @xmath104 are coprime .
we proceed to remove the coprimality conditions that appear in .
thus an application of mbius inversion yields @xmath470 where now @xmath471 is the number of @xmath472 such that and hold in the definition of @xmath453 .
we now consider the regions and in more detail .
define @xmath473 and @xmath474 for any positive @xmath433 .
we observe that @xmath475 if and only if @xmath476 we then have the following result .
[ lem : volume ] let @xmath436 and @xmath433 be positive .
then we have @xmath477 furthermore , we have @xmath478\times ( 0,c\sqrt{x/|a^2-b^2| } ] , \quad \mathcal{s}_u \subseteq [ -c , c]\times ( 0,c],\ ] ] for an absolute constant @xmath37 .
the first part of the lemma is self - evident , and so it remains to establish the bounds on @xmath479 and @xmath480 in the second part . for this it will clearly suffice to show that @xmath481 and @xmath482 for any @xmath483 .
suppose first that @xmath484 .
then it follows from the inequality @xmath485 that @xmath486 , whence @xmath487 in this case . if @xmath488 then we have @xmath489 and @xmath490 . in particular
it follows that @xmath491 .
let @xmath492 . if @xmath493 then @xmath494 .
thus @xmath495 and @xmath496 in this case .
alternatively , if @xmath497 , then we deduce that @xmath498 since @xmath499 . taking @xmath500
it therefore follows that @xmath501 in every case , which completes the proof of the lemma .
recall the definition of the set @xmath341 . for any @xmath502 it follows from that we are only interested in values of @xmath503 returning to our expression for @xmath453
, we may now show that there is a negligible contribution from large values of @xmath504 .
let @xmath505 then we have the following result .
we have @xmath506 where the summations are subject to @xmath507 .
lemma [ lem : det ] reveals that @xmath508 is an integer lattice of rank @xmath98 and determinant @xmath509 .
hence it follows from lemmas [ lem : lattice ] and [ lem : volume ] that @xmath510 with @xmath511 we must now estimate the overall contribution from each of the two terms .
the contribution from the first term is bounded by @xmath512 using lemma [ lem : nair ] .
since we are interested in values of @xmath513 , we see that the contribution from the second term is bounded by @xmath514 by the proof of .
this completes the proof of the lemma . in view of this result
we may now proceed under the assumption that @xmath515 in .
we will also run into trouble when estimating the contribution from various error terms when the @xmath98-adic order of @xmath516 in @xmath471 is allowed to be very large .
let @xmath517 be defined as for @xmath471 but with the additional restriction that @xmath518 .
we easily adapt the proof of the preceding result to show that @xmath471 may be replaced by @xmath517 in , with an acceptable error .
we now apply mbius inversion in , which thereby allows us to work with the equality @xmath519 here @xmath520 is given by and @xmath521 denotes the number of @xmath522 such that , and hold in the definition of @xmath453 , with @xmath523 and @xmath524 .
note that it will facilitate our ensuing investigation to restrict the summation to odd values of @xmath525 here . when it comes to estimating @xmath521 asymptotically as @xmath15 we will encounter problems when @xmath525 is large . for given @xmath308 and @xmath526 , we now let @xmath527 denote the number of @xmath528 for which @xmath529 , where @xmath530 is given by
note that @xmath531 in the above notation .
we have the following key result .
[ lem : large - gcd ] suppose @xmath308 and @xmath532 , with @xmath533 and @xmath534 then we have @xmath535 + \frac{yz_1^{o(1)}}{t},\ ] ] where @xmath536 $ ] denotes the integer part of @xmath537 .
the idea is to reintroduce a coprimality condition on the @xmath538 that are to be counted .
since we are only interested in an upper bound we may drop any of the defining conditions in @xmath539 that we care to choose . in this way
it is clear that @xmath540 let @xmath541 and @xmath542 note that the summand is zero unless @xmath543 . making the change of variables @xmath544 with @xmath545 ,
we conclude that the summand is bounded by the number of coprime vectors @xmath464 such that @xmath546 with @xmath547 .
recall the definition of @xmath548 and @xmath549 and let @xmath550 . since @xmath545 it follows that @xmath551 in view of the fact that @xmath552 and holds
, we obtain @xmath553 , with @xmath554 .
in particular @xmath555 .
it now follows that there exists an absolute constant @xmath37 such that @xmath556 for @xmath557 .
writing @xmath558 , we deduce that @xmath559 we note here that @xmath560 let us write @xmath561 for the term that is to be estimated in the statement of the lemma .
we will employ the trivial estimate @xmath562 , for any @xmath563 .
hence in @xmath561 there are clearly at most @xmath564 possible values of @xmath565 and @xmath566 .
furthermore there are also at most @xmath567 possible values of @xmath525 such that @xmath568 . on noting that the summation over @xmath409 contributes @xmath569
, we conclude from the proof of lemma [ lem : para-1 ] that @xmath570 where @xmath571 is given by . here
we have used @xmath572 appearing in @xmath530 to ensure @xmath274 in @xmath571 .
moreover it is useful to note that @xmath573 for @xmath574 .
let us write @xmath575 , respectively @xmath576 , for the overall contribution to the right hand side from @xmath577 such that @xmath578 , respectively @xmath579 . to begin with it follows from lemma [ lem : ab - small ] that @xmath580 it remains to estimate @xmath576 .
we return to , now with @xmath581 and @xmath358 in the summation over @xmath96
. we will write @xmath582 for the contribution to the right hand side from values of @xmath583 , and @xmath584 for the contribution from values of @xmath585 . beginning with the former , we note that @xmath586 if @xmath587 . if on the other hand @xmath588 then an application of gives @xmath589 this is satisfactory for the lemma .
finally we must deal with @xmath584 . for this
we will reverse the roles of the variables @xmath96 and @xmath345 in @xmath590 .
we have @xmath591 on writing @xmath592 , it now follows from lemma [ lem : ab - small ] that @xmath593 this is satisfactory and therefore completes the proof of the lemma .
we would now like to eliminate the contribution to from large @xmath525 .
taking @xmath594 and @xmath595 in lemma [ lem : large - gcd ] we deduce that the overall contribution from @xmath596 is @xmath597 this estimate exhibits a feature common to much of what follows .
the notation @xmath598 means that we allow an arbitrary , but absolutely bounded , power of @xmath335 in the error term .
the term @xmath599 should be thought of as a parameter since we are free to take @xmath600 large enough to nullify the effect of the numerator .
drawing this observation together with , our argument so far has established the following result .
[ lem : small_kl ] we have @xmath601 where @xmath341 is given by and @xmath520 by .
our task in this section is to set the scene for an asymptotic formula for the quantity @xmath602 .
let @xmath603 be the lattice defined in . on writing @xmath604 where @xmath605 is defined by
, it follows from the previous section that @xmath606 we now come to a rather delicate feature of the proof . ignoring the supplementary @xmath98-adic conditions in our new expression for @xmath521
, the obvious next step would be to try and approximate this cardinality with something like the volume of the region in question , divided by the determinant of @xmath101 .
ignoring also the contribution from the error term in this approximation , one would then be led to sum this quantity over all of the remaining parameters .
an absolutely crucial observation here is the following : for some ranges of the parameters @xmath607 we will have @xmath608 in lemma [ lem : small_kl ] , even if the approximation @xmath609 is non - zero .
thus a further reduction on the set of allowable parameters @xmath607 is necessary . for any @xmath610 and @xmath347 ,
define the region @xmath611 we will show how the summation over @xmath607 is necessarily restricted to this set for a suitable choice of @xmath612 .
it follows from the definition of @xmath341 that @xmath613 . taken together with lemma [ lem : volume ]
, the fact that we are interested in integers @xmath614 therefore implies that @xmath615 with @xmath3 given by .
the middle inequality here implies that @xmath616 in which we recall our convention that @xmath600 is used to denote a generic absolute positive constant .
but we also know that @xmath617 and @xmath618 , so that we may take @xmath619 and @xmath620 in the first and third inequality , giving @xmath621 the final inequality that we seek to establish is @xmath622 this is rather more subtle and arises from an inherent symmetry between @xmath607 and @xmath623 , where the latter are non - zero integers such that @xmath624 in particular all of @xmath625 are coprime to @xmath103 and @xmath104 .
it follows from the divisibility information on @xmath516 and @xmath626 that there exist non - zero integers @xmath627 such that @xmath628 and @xmath629 .
the height conditions for @xmath630 imply that @xmath631 furthermore , we have @xmath632 in particular @xmath633 since @xmath96 are coprime .
now we have @xmath634 if and only if @xmath635 and @xmath636 for some non - zero integer @xmath292 .
using we easily deduce that @xmath637 which contradicts the hypotheses in the definition of .
hence @xmath638 in , giving the further inequality @xmath639 whence @xmath640 this therefore implies . bringing our argument together we have therefore shown that the summation over @xmath607 in lemma [ lem : small_kl ] is subject to @xmath641 in fact , in view of the bounds @xmath642 , it is trivial to see that @xmath643 here , using the definitions of @xmath520 and @xmath336 , we have been able to replace @xmath644 by an additional factor of @xmath520 .
the following result refines this somewhat .
[ lem : final_prelim ] we have @xmath645 where @xmath341 is given by , @xmath520 by and @xmath646 is the characteristic function of the region .
we have already shown that the lemma is true with @xmath647 in place of @xmath648 . to establish the lemma it suffices to estimate the overall contribution to the main term from values of @xmath649 for which @xmath650 in the notation of .
this forces @xmath607 to satisfy one of four further inequalities .
let us show how to handle the contribution corresponding to @xmath607 satisfying @xmath651 the remaining cases being handled in an identical manner .
let @xmath652 denote the contribution to @xmath602 arising from @xmath442 for which @xmath653 .
lemma [ lem : large - gcd ] implies that the overall contribution to lemma [ lem : final_prelim ] from the remaining quantity is @xmath654 , which is satisfactory . to estimate @xmath652
, we write @xmath544 with @xmath655 and @xmath656 an odd integer divisible by @xmath525 .
the summation over @xmath409 allows us to assume that @xmath657 .
let us write @xmath658 as in and .
we conclude from lemmas [ lem : lattice ] and [ lem : volume ] that @xmath659 where now @xmath660 is the lattice of @xmath464 for which @xmath661 lemma
[ lem : det ] therefore leads us to an overall contribution of @xmath662 here the summation over @xmath663 is subject to the restriction that they should be odd , with @xmath664 , @xmath665 and holding . for fixed @xmath96 and @xmath48
there are at most @xmath666 values of @xmath667 .
hence the overall contribution from the first term in the above summand is @xmath668 , which is satisfactory .
the remaining contribution is clearly @xmath669 on carrying out the summation over @xmath409 .
we now break the summation over @xmath48 into @xmath670 and @xmath671 , with @xmath520 given by .
we see that the overall contribution from @xmath672 is @xmath673 which is satisfactory .
turning to @xmath670 , we take @xmath674 and so obtain the overall contribution @xmath675 we will find it convenient to proceed under the additional hypothesis @xmath676 underlying this assumption is the symmetry that exists between @xmath677 and @xmath678 such that holds .
thus if fails then it is immediately clear from the definition of @xmath341 that @xmath679 furthermore , translates into @xmath680 which also shares the same structure .
this discussion shows that we may proceed under the assumption that holds in what follows .
we now break the summation over @xmath96 into dyadic intervals , in order to fix attention on the range @xmath681 for @xmath682 interchanging the order of summation becomes @xmath683 where @xmath684 furthermore , the summation over @xmath685 is over odd integers subject to @xmath665 and @xmath686 let @xmath687 be the number of square roots of @xmath688 modulo @xmath689 . for given @xmath690 the conditions @xmath691 and @xmath692 force the vector @xmath693 to lie on one of at most @xmath694 integer sublattices of @xmath695 , each of determinant @xmath696 .
now it is easy to see that @xmath697 and @xmath698 where @xmath699 is the real non - principal character modulo @xmath11 and @xmath7 denotes the sums of two squares function .
it follows that @xmath700 for any @xmath701 .
since @xmath96 lie in a region in @xmath401 of volume @xmath702 , an application of lemma [ lem : lattice ] furnishes the contribution @xmath703 on summing over @xmath642 . here
the inner summation is over @xmath607 such that @xmath704 as follows from and .
thus we easily obtain the overall contribution @xmath705 on summing first over @xmath452 , then over @xmath706 and finally over the @xmath707 choices for @xmath708 .
this completes the proof of the lemma .
let @xmath709 be an arbitrary choice of parameters that appear in the main term in lemma [ lem : final_prelim ] s estimate for @xmath252 .
we now come to our estimation of , where @xmath603 is the lattice defined in and @xmath710 is given by . given that @xmath711 is clearly defined with piecewise continuous boundary , we would like to apply the well - known formula .
however , there are two complications that prevent a routine application of this estimate .
firstly , we will need to take care of the @xmath98-adic conditions apparent in .
secondly we need to deal with the fact that once summed over the remaining parameters , the error term in the above asymptotic formula for @xmath712 does not make a satisfactory overall contribution from the point of view of the main theorem . in spite of these objections we will dedicate the remainder of this section to interpreting the main term in in the present context , delaying our discussion of the error term until the subsequent section .
let us consider the first issue mentioned above , namely the @xmath98-adic conditions on @xmath387 .
we retain the shorthand notation @xmath713 and make the observation that @xmath714 , for any @xmath715 furthermore , we will set @xmath716 for @xmath717 .
since the parameters in @xmath101 are all odd it easily follows from lemma [ lem : det ] that @xmath718 .
it is here that our earlier restriction to odd values of @xmath525 pays dividends .
finally , it will be convenient to define @xmath719 for any @xmath720 .
we will separate our investigation according to the value of @xmath409 .
we suppress the dependence on @xmath721 in the expression for @xmath521 in , replacing it by @xmath722 in order to underline the dependence on @xmath409 .
we have @xmath43 basic possibilities to consider : either @xmath723 or @xmath419 or @xmath724 .
note that ensures that the possibility @xmath725 does not arise .
let us begin by supposing that @xmath723 , which according to is only possible when @xmath726 and @xmath727 , so that @xmath728 .
recall the notation introduced above for @xmath729 .
it follows that @xmath730 in line with we expect the cardinality in the summand to satisfy an asymptotic formula with main term @xmath731 we may now conclude as follows .
[ lem : l0 ] we have @xmath732}}\big)\frac{\operatorname{vol}({\mathcal{r}}_0 ' ) } { \det { \mathsf{\lambda } } } + e^0(b),\ ] ] where @xmath733 we have found it useful to include @xmath734 in the main term for this estimate , to help keep track of the fact that we are only interested in the value of @xmath735 when @xmath726 .
when @xmath419 it follows from that @xmath736 if @xmath726 and @xmath737 if @xmath738 .
hence yields @xmath739 arguing as above , we now have @xmath740 drawing out the obvious main term , as previously , we therefore conclude the proof of the following result .
[ lem : l1 ] we have @xmath741 where @xmath742 assume now that @xmath743 , under which assumption implies that @xmath738 and @xmath744 , with @xmath745 in particular , if @xmath746 then it follows that @xmath747 .
this is then automatically bounded above by @xmath748 .
otherwise we must have @xmath749 and @xmath750 .
moreover , we must force @xmath727 in our considerations .
we deduce from that @xmath722 is the sum of @xmath751 and @xmath752 note the first term is always present , since @xmath753 , but the latter term only appears if @xmath754 .
the following result is now available .
[ lem : l2 ] let @xmath724 .
then we have @xmath755 if @xmath756 and @xmath757 + 2}}\big)\frac{\operatorname{vol}({\mathcal{r}}_0')}{\det { \mathsf{\lambda } } } + e_{\alpha}^{\nu}(b)+e_{\beta}^{\nu}(b),\ ] ] if @xmath754 , where @xmath758
it is now time to establish that the error terms in lemmas [ lem : l0 ] , [ lem : l1 ] and [ lem : l2 ] make a satisfactory overall contribution once summed up over the relevant parameters appearing in lemma [ lem : final_prelim ] . in order not to be encumbered with superfluous notation
let us fix our attention on the error term @xmath759 where @xmath3 is given by .
the outcome of our investigation will be a proof of the following result .
[ lem : exp ] let latexmath:[\ ] ] where if @xmath920 is given by and , then @xmath921 here we have equated @xmath922 with the main term in the various estimates from [ s : asy ] .
moreover , we have used the fact that @xmath923 in controlling the error term .
we may therefore deduce from lemma [ lem : volume ] that @xmath924 where @xmath480 is given by for any positive @xmath433 .
define @xmath925 for any @xmath926 .
let @xmath646 be the characteristic function of the region and let @xmath927 for any @xmath526 .
we are now ready to establish the following result .
[ lem : farine ] we have @xmath928 where @xmath929 is given by and @xmath930 in order to avoid repeating the same argument twice we will simply replace @xmath931 by @xmath932 in what follows , leaving it to the patience of the reader to verify that this is indeed permissible .
bringing together our expression for @xmath933 with lemmas [ lem : det ] , [ lem : final_prelim ] and [ lem : exp ] , we deduce that @xmath252 can be replaced by @xmath934 with an acceptable error , where @xmath520 is given by , @xmath646 is the characteristic function of , @xmath935 and @xmath936 for any @xmath588 .
lemma [ lem : volume ] implies that @xmath937 and the definition of @xmath341 yields @xmath938 .
hence the overall contribution from the above error term is clearly satisfactory . applying lemma [ lem : nair ] one easily checks that the overall contribution to @xmath252 from values of @xmath504 such that @xmath939 is also satisfactory .
we may now conclude that @xmath940 where @xmath941 recall the definition of @xmath942 from .
for any arithmetic function @xmath158 , we have @xmath943 it therefore follows that @xmath944 where @xmath195 is given by .
recall the inequality @xmath923 satisfied by any @xmath347 .
a little thought reveals that @xmath945 in the notation of . bringing everything together
we have so far established the upper and lower bounds @xmath946 we proceed to enlarge the set of allowable @xmath96 slightly , by handling separately the contribution from @xmath96 such that @xmath355 or @xmath947 .
for the former , exactly the same sort of argument employed in lemma [ lem : red-3 ] suffices to show that we obtain an overall contribution @xmath948 which is satisfactory . for the latter , the proof of shows that there is a satisfactory contribution in this case too . arguing in a similar manner it is also possible to restrict attention to @xmath96 for which @xmath949 and to enlarge the set of allowable @xmath96 to include the range @xmath950 .
finally , we break the summation over @xmath96 into those for which @xmath350 and those for which @xmath951 , observing that @xmath952 this therefore allows us to restrict to a summation over the set @xmath953 , which thereby completes the proof of the lemma .
we now have everything in place to complete the proof of the theorem . in what follows
let us write @xmath202 for either of the quantities @xmath954 or @xmath955 . define @xmath956 for any @xmath957 such that @xmath958 . observing that @xmath959
it now follows that @xmath960 hence lemma [ lem : farine ] renders it sufficient to estimate @xmath961 asymptotically .
this is achieved in the following result .
[ lem : sig ] we have @xmath962 where @xmath963 is given by , with @xmath964 and @xmath108 given by .
recall the definition of @xmath929 and define the region @xmath965 this is contained in @xmath966 ^ 4 $ ] . on recalling the definition of @xmath967 for @xmath968 ,
one easily checks that @xmath969 it now follows from lemma [ lem : uniform ] that @xmath970 where @xmath971 is given by .
here we have applied the lemma with the region @xmath972 which has volume @xmath973 .
we can calculate the volume of @xmath971 using the software package ` polymake ` @xcite , the outcome being that @xmath974 .
the statement of the lemma is now obvious . recalling the definition of the function , it is a routine calculation to verify that @xmath975 for @xmath976 , using lemma [ lem : volume ] and the definition of @xmath480 . in particular
it follows that @xmath977 . combining this with lemma [ lem : sig ] and in lemma [ lem : farine ] , and inserting it into lemma [ lem : red-1 ] , we arrive at the final bound @xmath978 where @xmath979 finally , we note that @xmath980 where @xmath981 this therefore completes the proof of the asymptotic formula @xmath982 where @xmath963 is given by .
the purpose of this section is to affirm that the constant in our asymptotic formula for @xmath36 agrees with the peyre s prediction @xcite , as required to complete the proof of the theorem . in general terms
the constant @xmath19 should be a product of three constants @xmath983 .
the constant @xmath984 is a rational number defined in terms of the cone of effective divisors of @xmath3 . to calculate its value
we will apply the calculations undertaken by derenthal in the appendix to this paper .
let @xmath985 and let @xmath986 .
then a straightforward calculation reveals that the @xmath39 lines on @xmath3 are given by @xmath987 in particular all of the lines split over @xmath988 and @xmath3 contains precisely @xmath989 lines that are defined over @xmath0 .
let us write @xmath990 for the galois group of the splitting field . as described in the book by manin @xcite ,
if @xmath991 denotes the class of a hyperplane section , then the geometric picard group @xmath992 is given by @xmath993 where @xmath994 are the classes of a set of @xmath45 lines in @xmath3 that are mutually skew .
identifying divisors with their classes in @xmath992 , it is easily checked that the lines @xmath995 satisfy the desired property . since
@xmath996 are defined over @xmath0 and @xmath997 are defined over @xmath988 , but are conjugate under the action of @xmath998 , so it follows that @xmath999 this retrieves the information @xmath91 that was recorded in the introduction .
it now follows from taking @xmath22 in table @xmath1000 in the appendix that @xmath1001 the constant @xmath1002 is equal to the cardinality of @xmath1003 , where @xmath1004 is the geometric picard group . in the present
setting we have @xmath1005 since @xmath3 is @xmath0-birationally trivial .
we now turn to the value of the constant @xmath1006 . for any place @xmath1007 of @xmath0
let @xmath1008 be the usual @xmath1007-adic density of points on the locally compact space @xmath1009 .
the tamagawa number associated to @xmath3 and @xmath34 is then given by @xmath1010 in the present setting it follows from our calculation of @xmath1011 that @xmath1012 where @xmath699 is the real non - principal character modulo @xmath11 , whence @xmath1013 recall the definitions , of @xmath108 and @xmath1014 , respectively .
we will show that @xmath1015 with @xmath1016 furthermore , we will demonstrate the equality @xmath1017 where @xmath1018 is given by .
subject to the proofs of and , let us now verify that the constant in coincides with that predicted by peyre . bringing together our expression for @xmath1006 with and , we find that @xmath1019 in the notation of .
this therefore confirms the constant in is the same one that is predicted by peyre . in order to prove we will adhere to the method outlined by peyre @xcite .
taking into account the fact that @xmath229 and @xmath1021 represent the same point in @xmath23 , the archimedean density of points on @xmath3 is equal to @xmath1022 where @xmath1023 is the leray form , and we have used the same sort of symmetry arguments apparent in lemma [ lem : red-1 ] .
it will be convenient to parametrise the points via the choice of variables @xmath1024 .
observe that @xmath1025 we will take the leray form @xmath1026 .
recall the definition of @xmath1027 .
we will pass from @xmath1024 to @xmath1028 , with @xmath1029 , via the change of variables @xmath1030 the jacobian of this transformation is calculated to be @xmath1031 on noting that @xmath1032 a modest pause for thought now reveals that holds , as claimed
. it will ease notation if we write @xmath1034 for @xmath1035 throughout this section , for any prime power @xmath1036 .
it remains to calculate the value of @xmath1037 , for any prime @xmath205 , where @xmath1038 denotes the number of @xmath1039 for which @xmath1040 .
the usual way of proceeding at this point would be to try and interpret @xmath1041 as an explicit function of @xmath1042 .
this would then entail a parallel calculation of @xmath963 in , in order to check that the values match .
instead we will try and calculate @xmath1038 by mimicking the steps taken in the proof of the theorem .
recall the definitions , of @xmath248 and @xmath116 .
it is easy to check that @xmath1043 , where @xmath1044 , for any prime @xmath205 , with @xmath1045 we will show that @xmath1046 where @xmath1047 is given by
. this will clearly be enough to establish .
let @xmath229 be any vector counted by @xmath1048 .
we claim that there are there are precisely @xmath1049 choices of @xmath1050 such that @xmath1051 , @xmath1052 and @xmath1053 to see this we may suppose without loss of generality that @xmath1054 . for any of @xmath1049 choices of @xmath1055 , the value of @xmath292
is determined uniquely modulo @xmath1056 via the congruence @xmath1057 .
but then values of @xmath104 and @xmath291 are also determined exactly modulo @xmath1056 , which therefore establishes the claim .
recall the definition of @xmath94 from . by a convenient abuse of notation we will write @xmath1058 to denote that the underlying quadratic polynomial is congruent to zero modulo @xmath1056 .
it now follows that @xmath1059 where the index @xmath72 belongs to @xmath182 .
moreover , in view of the fact that is always a solution of the relevant congruence modulo @xmath1056 we may assume in all of our calculations that the summand is always positive . here
we recall the convention that @xmath1060 if and only if @xmath210 . for given @xmath1061 , with @xmath1062 , and given @xmath1063 , let us define @xmath1064 it will be convenient to define @xmath1065 as for @xmath1066 , but without the condition that @xmath1067 .
a little thought reveals that we have @xmath1068 for each @xmath1069 and @xmath1063 .
our calculation of @xmath1066 will depend intimately on the values of @xmath1070 and whether or not @xmath205 is odd .
we have collected together the necessary information in the following result .
[ lem : final ] let @xmath170 .
then we have @xmath1071 and furthermore , @xmath1072 if @xmath1073 and @xmath1074 let @xmath1075
. then we have @xmath1076 we postpone the proof of this result until later . for the moment we return to the expression for @xmath1048 that we are trying to evaluate . for any prime @xmath205 and @xmath206 recall the definitions and of @xmath1077 and @xmath1014 , respectively .
in particular it is not hard to see that @xmath1078 we have @xmath1079 where we have written @xmath1080 for convenience .
beginning with the case @xmath170 let us write @xmath1081 for @xmath1082 coprime to @xmath205 .
since @xmath1083 it clearly follows that at most one of @xmath1084 can be non - zero .
when they are all zero reveals that the summand here is @xmath1085 . when @xmath1086 and @xmath1087 we know that @xmath1088 is a quadratic residue modulo @xmath205 since @xmath1089 is always non - empty .
but then we may apply to estimate the summand .
this same estimate also covers the case in which @xmath1090 or @xmath1091 is positive . when @xmath170 we therefore deduce that @xmath1092 with @xmath1080 .
recall the definition of @xmath108 for odd prime powers .
it is not hard to see that @xmath1093 it therefore follows that holds when @xmath170 , with @xmath1094 .
it remains to deal with the case @xmath1075 .
arguing as for @xmath170 we write @xmath1095 for odd @xmath1082 . since @xmath1096 it now follows that either @xmath1097 , or else @xmath1098 and @xmath1099 . in the latter case
we have @xmath1100 and @xmath1101 , or @xmath1102 and @xmath1103 . in every case we know that @xmath1104 is non - empty , and in fact has a solution @xmath1105 .
suppose first that @xmath1097 .
then precisely one of @xmath96 must be even , and we automatically have @xmath1106 . but then implies that the innermost summand in our expression for @xmath1107 is @xmath1000 .
when @xmath1100 and @xmath1108 we may apply with @xmath1109 to deduce that the summand is @xmath1110 the same applies when @xmath1102 and @xmath1111 .
putting this all together we deduce that @xmath1112 recall the definition of @xmath108 for even prime powers .
it is now clear that holds when @xmath1075 , with @xmath1113 . throughout our proof of the lemma we can and
will assume that @xmath86 is large compared with @xmath1114 and @xmath409 .
assume that @xmath170 and let @xmath1115 .
it follows from hensel s lemma that @xmath1116 where @xmath1117 is the number of @xmath1118 such that @xmath1067 and @xmath1119 .
the latter can easily be calculated using gauss sums with the outcome that @xmath1120 .
hence is now obvious .
we now turn to the calculation of @xmath1121 , for @xmath1073 , for which we will assume that @xmath1122 so that @xmath1088 can be lifted to a quadratic residue modulo @xmath1056 .
hence @xmath1123 similarly for @xmath1124 .
the contribution to the right hand side from those @xmath390 for which @xmath1125 is @xmath1126 this follows from an argument based on hensel s lemma and an easy analysis of the corresponding congruence modulo @xmath205 .
the contribution from those @xmath390 for which @xmath1127 is @xmath1128 if @xmath1129 and @xmath1130 if @xmath1131 . assuming @xmath1131 we therefore obtain the contribution @xmath1132 but this is just @xmath1133 .
it therefore follows that @xmath1134 recalling , we have @xmath1135 and @xmath1136 if @xmath1131 . in view of the fact that @xmath1137 and @xmath1138 , we easily deduce that @xmath1139 moreover , on noting that @xmath1140 and @xmath1141 , and give @xmath1142 armed with these expression we can conclude that holds , as claimed .
we now examine the case @xmath1075 .
recall the definition of @xmath1143 and the corresponding definition of @xmath1144 .
let @xmath1115 , under which hypothesis we will assume that @xmath1145 our argument will differ according to which of @xmath1128 or @xmath98 it is that @xmath1146 is congruent to modulo @xmath989 . beginning with the case @xmath1147 , it is not hard to see that @xmath390 are all necessarily odd in the definition of @xmath1148 .
writing @xmath1149 and @xmath1150 we deduce that @xmath1151 where we have set @xmath1152 and @xmath1153 .
for any @xmath1154 and any @xmath1155 , let us denote by @xmath1156 the number of @xmath1157 for which @xmath1158 .
an easy lifting argument reveals that @xmath1159 varying @xmath1160 and applying the latter equality to estimate the number of relevant @xmath1161 , we may now conclude that @xmath1162 when @xmath1163 . turning to the case @xmath1164
, we conclude that @xmath345 are both necessarily odd in the definition of @xmath1148 and @xmath293 is even .
writing @xmath1149 and @xmath1165 we deduce that @xmath1166 where @xmath1167 is as above but this time with @xmath1168 .
we now conclude as previously .
altogether we have shown that holds with @xmath1115 , under the assumption that holds .
we proceed to consider @xmath1169 , for @xmath1170 , observing that @xmath1171 we will make the assumption that @xmath1172 for any odd @xmath1154 and any @xmath1173 , let @xmath1174 denote the number of @xmath1157 for which @xmath1175 .
it is an easy exercise to check that @xmath1176 for @xmath1173 .
this is implied by our previous bound for @xmath1156 , for example .
let us consider the contribution from those @xmath390 for which @xmath1177 , appealing to where necessary . for each @xmath1178
such that @xmath1177 and @xmath1179 , our calculation of @xmath1180 shows that there are exactly @xmath11 choices for @xmath293 .
altogether , we therefore obtain the contribution @xmath1181 when @xmath1182 , and @xmath1183 when @xmath1184 .
the contribution from those @xmath390 for which @xmath1185 is clearly @xmath1186 say .
we will show that @xmath1187 together with the contribution from the previous treatment , this will suffice to show that holds when @xmath1170 and @xmath419 , under the hypothesis . turning to the proof of ,
suppose first that @xmath1182 .
then @xmath1188 , by , and @xmath1189 when @xmath1185 we note that for any @xmath1190 such that @xmath1191 and @xmath1192 , there are @xmath11 choices for @xmath293 .
the condition modulo @xmath989 is clearly equivalent to @xmath1193 . in particular , assuming that @xmath1194 , we get an overall contribution of @xmath1195 .
alternatively , the case @xmath1177 contributes the same amount via a similar argument , but this time under the hypothesis @xmath1196 .
together this confirms that @xmath1197 is indeed @xmath1000 .
the case @xmath1198 is easily seen to be impossible , whence @xmath1199 . turning to the case @xmath1200 , for which implies that @xmath1201 , we deduce that both @xmath291 and @xmath293 must be even .
hence it follows that @xmath1202 the argument now goes as for the case @xmath1182 , with a separate analysis of the cases @xmath1185 and @xmath1177 .
each contributes @xmath1000 to @xmath1203 , but only one such case arises according to whether @xmath600 is congruent to @xmath1000 or @xmath43 modulo @xmath11 .
hence @xmath1204 .
finally let us turn to the case @xmath1205 , for which we still have @xmath1201 by .
it will now be convenient to write @xmath1206 to indicate the dependence on @xmath86 in the definition .
now either @xmath1207 are both odd or they are both even .
since @xmath1208 whenever @xmath1205 and @xmath291 is odd , so it follows that the contribution to @xmath1209 from odd @xmath291 is @xmath1210 , by our expression for @xmath1174 . the contribution from even @xmath291
is easily seen to be @xmath1211 .
but then induction on @xmath1114 reveals that the contribution from even @xmath291 is @xmath1212 .
together the two contributions suffice to ensure that holds when @xmath1205 .
the nef cone volume in degrees @xmath43 and @xmath11 for @xmath1213 , let @xmath3 be a non - singular del pezzo surface of degree @xmath177 , defined over a perfect field @xmath1 with an algebraic closure @xmath1214 . we want to study the nef cone volume of @xmath3 , defined by peyre ( *
* definition 2.4 ) to be @xmath1215 where @xmath5 is the anticanonical class , @xmath1216 is the dual cone with respect to the intersection form @xmath1217 of the cone of classes of effective divisors in @xmath1218 , and the volume form is described in ( * ? ? ?
* section 6 ) , ( * ? ? ?
* section 2 ) .
assume @xmath1219 .
then @xmath1220 , where @xmath1221 and @xmath1222 is the absolute galois group . by (
* section 7 ) , @xmath984 depends only on the action of @xmath1223 on the lines on @xmath3 .
more precisely , this action of @xmath1223 factorises through the weyl group @xmath1224 associated to the root system @xmath1225 by ( * ? ? ?
* corollary 7.3 ) , the constant @xmath984 depends only on the conjugacy class of the image @xmath34 of @xmath1223 as a subgroup of @xmath1224 . the conjugacy classes of subgroups of @xmath1224 can be listed using ` magma ` @xcite , for example . by the results of (
* section 7 ) , the computation of @xmath984 corresponding to a given subgroup @xmath34 is completely straightforward .
one determines the sum of elements of each orbit under the action of @xmath34 on the set of classes of lines in @xmath1226 . by (
* corollary 6.4 ) , these sums generate the cone of effective divisors in @xmath1227 . using @xcite , for example ,
one computes the dual of this cone and then the volume @xmath984 .
our goal is to compute @xmath984 corresponding to all conjugacy classes of subgroups @xmath34 of @xmath1224 , but to organise the results in a way that allows us to read off @xmath984 for a given del pezzo surface @xmath3 without determining the corresponding @xmath34 . for @xmath1231
, it turns out that @xmath984 depends only on the rank of @xmath1227 and , if @xmath1232 , the number of @xmath1-rational lines on @xmath3 . even though there are @xmath1233 conjugacy classes of subgroups of @xmath1234 , we get just nine cases that can be found in table [ tab : alpha_deg_4 ] . for @xmath1237 ,
the weyl group @xmath1238 has @xmath1239 conjugacy classes of subgroups . here
, the value of @xmath984 depends not only on @xmath1240 and the number of @xmath1-rational lines on @xmath1241 . in most cases
, it is enough to consider the following additional information on @xmath3 : if there are @xmath1242 orbits of length @xmath1243 , for @xmath1244 , then we denote this orbit structure by @xmath1245 .
note that @xmath1246 is the number of lines on @xmath1241 .
the values of @xmath984 for @xmath3 of degree @xmath43 can be found in table [ tab : alpha_deg_3 ] .
only if this orbit structure is @xmath1247 ( i.e. , there are three orbits of length @xmath43 and three orbits of length @xmath44 ) , this does not yet determine @xmath984 ( marked by @xmath1248 in the last column of table [ tab : alpha_deg_3 ] ) .
in this case , we obtain @xmath984 as follows .
compute the sums of elements of each of these six orbits .
if this gives six different elements of @xmath1226 , then @xmath1249 ; otherwise , these sums give four different elements of @xmath1226 , and @xmath1250 .
browning , an overview of manin s conjecture for del pezzo surfaces . _ analytic number theory a tribute to gauss and dirichlet ( gttingen , 20th june 24th june , 2005 ) _ , 3956 , clay mathematics proceedings * 7 * , ams , 2007 . | an asymptotic formula is established for the number of @xmath0-rational points of bounded height on a non - singular quartic del pezzo surface with a conic bundle structure . |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.