article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
the topic i have been asked to review the future of stellar population studies in the milky way ( mw ) and local group ( lg ) is a vast one , within which i can do no more than a whirlwind tour of various topics . to make the task more manageable
, my goal will be primarily to highlight some facets of the area of _ resolved _ stellar populations that are now , and that will soon be , addressable as stellar populations work enters the industrial age " an age where vast databases and automated analysis of these vast databases can be brought to bear on both newly revealed as well as age - old problems in the field .
to this end , a major portion of this contribution will be to summarize those large - scale surveys about which i am aware ; i apologize in advance for any surveys i have failed to include or whose parameters i have misrepresented .
before starting this whirlwind tour of the industrial age " of resolved lg stellar populations work , it is important to be mindful of the great advances being made in sister subdisciplines , such as cosmology , large scale structure formation , dark matter and the analysis of these matters through large numerical simulations .
for example , it is now well established through such simulations ( and with growing corroboration by observations ) that the growth of structure in a cold dark matter ( cdm ) universe is hierarchical , with small structures merging along filaments to form larger structures , like mw - sized galaxies and lg - sized galaxy associations .
but while hierarchical merging until late times ( even presently ) is now well established as a key element of cdm structure formation models , it is useful to recall that both late infall and hierarchical formation was anticipated as a central aspect of mw formation studies through the stellar populations work of searle & zinn ( 1978 ) , who showed that the outer halo globular clusters exhibit a significant age spread ( inferred through interpretation of horizontal branch morphologies ) .
this foundational , early result demonstrated the power of stellar population studies for unlocking the secrets of galaxy formation .
but with @xmath0cdm a now firmly entrenched and generally successful theory , it is sensible to approach future stellar population work with this context in mind , and , in part , look to stellar population studies for testing and/or refining this widely - accepted thesis .
this is only possible because the advanced state of @xmath0cdm simulations allows them to make a rich variety of predictions about the structure , dynamics and chemistry of cdm structures . but
checks on the theory with resolved stellar population studies are all the more important because , while @xmath0cdm has enjoyed great success in matching observations of structures on the largest scales , several problems still remain in matching the theory to observations on galaxy scales . among these problems on small scales
are ( 1 ) the missing satellites problem " , where the models predict a mass spectrum of galaxy subhalos that are strongly in disagreement with that observed around the mw ( e.g. , klypin et al . 1999 ) ; ( 2 ) the central cusps problem " , where @xmath0cdm predicts the mass density of galaxies to have steeply rising central cusps ( e.g. , navarro et al .
1997 ) for which there is little observational evidence in the mw ( see , e.g. , merrifield 2005 ) and which is belied by the flat central density profiles of dwarf galaxies ; and ( 3 ) problems with the predicted angular momentum distributions in galaxies , wherein the models have difficulty making large , extended disks like that of the mw ( e.g. , abadi et al . 2003 ) .
thus , a current focus for advancing @xmath0cdm theory is attempting to resolve problems on galaxy scales , and clearly there are potent avenues by which stellar populations work in the lg , mw and their satellite systems can contribute not only to progress in understanding hierarchical galaxy formation and evolution , obviously , but also to fine - tuning @xmath0cdm theory .
as one example of how stellar population studies can contribute to fine - tuning @xmath0cdm theory , take the notion of infall along filaments , one of the elements of hierachical galaxy formation vividly seen in the simulations as well as in the observed distribution of galaxies on large scales .
is this something we can find evidence for in the lg ?
infall of dm subhalos along filaments should leave rather specific dynamical fingerprints in terms of preferred shapes , orientations and coincidences of orbits of lg systems and mw satellites , accreted globular clusters and ( halo ) stars ( e.g. , knebe et al .
it has long been known that most of the classical " mw satellites show a relatively strong planar alignment ( e.g. , kunkel & demers 1976 , kunkel 1979 , lynden - bell 1976 , 1982 , majewski 1994 , fusi pecci et al .
1995 , metz et al .
2007 , 2009a ) almost perpendicular to the galactic plane .
this preferential spatial distribution holds up statistically even with the inclusion of the new ultrafaint " satellites revealed by the sloan digital sky survey ( sdss ) , and when the orientation of the sdss footprint - bias " ( i.e. , towards the north galactic cap ) is accounted for ( metz et al .
such spatial alignments could be a signature of a dynamical association of the satellites , but must be checked with full 3-dimensional velocity measurements for the satellites . through painstaking work ,
the proper motions , @xmath1 , of mw satellites are slowly being accumulated , though in most cases with still only marginally significant @xmath2 , at least for ground - based data ( see summary in metz et al .
more recently , proper motions measured using hubble space telescope imaging of the magellanic clouds ( `` mcs '' hereafter ; kallivayalil et al . 2006a , b , piatek et al . 2008 ) , other dsph satellites ( piatek et al .
2003 , 2005 , 2006 , 2007 ) and globular clusters ( e.g. , bedin et al .
2006 , kalirai et al .
2007 ) have shown significantly reduced proper motion uncertainties , despite very small time baselines with improvements to be expected if more epochs can be accumulated with the newly - installed wfc3 .
analysis of the proper motions of these mw satellites has more or less supported the notion of dynamical kinship of at least some satellites , at least in that a number of these systems seem to share similar orbital poles , showing common directions of angular momentum ( majewski et al .
1996 , palma et al .
2002 , metz et al .
the meaning of such alignments and whether they are at odds with , or consistent with , expectations of typical satellite configurations around mw - like galaxies in cdm models has been much debated of late . among the postulated theories , such dynamical correlations could reflect siblinghood among satellites : ( 1 ) through the break - up of once large satellites e.g. , a greater magellanic or fornax system ( kunkel 1979 , lynden - bell 1982 , majewski 1994 , metz et al .
2008 ) ; ( 2 ) through the formation of satellites as tidal dwarfs ( e.g. , kroupa et al .
2005 ) ; ( 3 ) through the infall of satellites onto the mw in groups ( li & helmi 2008 , donghia & lake 2008 , metz et al .
2009b ) ; or ( 4 ) through the preferential infall of satellites along cdm filaments ( kang et al .
2005 , zentner et al .
2005 ) .
according to the simulations , the surviving galaxies of today reflect the most recent infall onto the mw of subhalos now in the mw , those in current satellites were the most distant subhalos at earlier times . earlier material to fall in came from initially closer matter and
is now spread out among the star and star cluster debris of the halo .
thus , comparing the orbits of the satellites to the orbits of halo stars is a comparison of how infall proceeded at these different epochs , and this translates to spatial variations in the dynamics . according to high - resolution numerical simulations ( e.g. , diemand et al . 2005 )
the outermost galactic halo stars and globular clusters should be on radially biased orbits while the inner regions should contain more isotropic orbits .
the predicted kinematics for the late - infalling satellite dwarf galaxies is different , with small anisotropies even in the outer parts .
in addition , the net angular momentum of these components is expected to be small and simply reflect that of the dominant dark matter component at this epoch . to test such predictions
requires measures of halo star and satellite orbital anisotropy ; these should be done _ in situ _ to avoid orbital biases attendant with solar neighborhood samples .
( to this end , it is interesting to note the much more complex inner / outer halo structure suggested by recent analysis of the sdss segue stellar sample with proper motions by carollo et al .
2007 , including the suggestion that the outer halo dominates the _ low _ eccentricity component of their relatively nearby stellar sample . )
this , in turn , requires the measurement of not only the bulk proper motions of the mw satellites , but the much more demanding star - by - star measures of _ distant _ , individual halo stars .
llcl & * accuracy * & * flux .
limit * & * stars * + _ usno _ b1.0 & 200 mas & @xmath3 & 1 billion + tycho-2 & 60 mas & @xmath4 & 2.5 million + & 20 - 70 mas & @xmath5 & 40 million + _ pan - starrs _ & 30 mas & @xmath6 & 2 billion + _ lsst _ & 3 mas & @xmath7 & 10 billion + hipparcos & 1 mas & @xmath4 & 117,955 + hst+wfpc2/stis / acs / wfc3 & 1 mas & @xmath8 & pointed instrument + j - maps & 1 mas & @xmath9 & 40 million + & 0.6 mas & @xmath10 & pointed instrument + hst / fgs & 0.2 mas & @xmath11 & pointed instrument + jasmine & 0.010 mas & @xmath12 & 100 million + & 0.010 mas & @xmath13200 mjy & pointed instrument + gaia & 0.020 mas & @xmath9 & 1 billion + sim - lite ( wide angle mode ) & 0.004 mas & @xmath14 & pointed instrument ( @xmath1310,000 sources lifetime ) + at this dawn of the industrial age " of stellar populations studies we can look forward to astrometric studies and surveys on a vast and more precise scale to the realm of billions of stars and microarcsecond precisions that will help us move forward on such issues .
table 1 gives a sampling of astrometric capabilities ( quoted as mission end " positional accuracies for the surveys and achieved / achievable " accuracies for the pointed instruments _ at the magnitude limited given _ ) either already underway or expected in the next decade ( with table values taken from the project websites or published papers ) .
the projects include large ground - based surveys that probe deeply but at lower relative precision ( usno b1.0 , ucac2 and the future pan - starrs and lsst astrometry ) as well as more focused ground - based facilities that can obtain higher precision for example vlba observations of radio sources ( such as masers ) and the soon - to - be - commissioned one degree imager on the wiyn telescope , with its orthogonal transfer arrays capable of on - chip local corrections for atmosphere - induced , differential image wander . but upcoming space - based astrometric missions including j - maps ( which will provide a new epoch of positions that can be matched to those from hipparcos and tycho ) , jasmine ( a mission to do infrared astrometry of the galactic bulge ) , gaia and sim - lite
will revolutionize the field of galactic kinematics with astrometry reaching to microarcsecond precision ( sim - lite in narrow angle mode ) .
these capabilities will enable us to achieve the scientific goals outlined above , considering that to derive transverse velocities of 10 km s@xmath15 accuracy for mw satellites at @xmath13100 kpc requires @xmath1320
@xmath1as yr@xmath15 proper motions .
this is within range of sim - lite for individual stars and for gaia after the averaging of the proper motions for many @xmath16@xmath13@xmath17 giant stars in each satellite .
however , only sim - lite will be able to measure the proper motions of satellites like leo i , leo ii or canven i , which are beyond 200 kpc and have their brightest members at @xmath16@xmath13@xmath18 .
moreover , only sim - lite will be able to measure precise proper motions for many of the newfound ultrafaint dsphs , which , even though at closer distances , have a paucity of bright giant stars . obviously , for single stars , where one can not take advantage of averaging , the problem is more acute , but gaia will give an unparalleled view of the dynamics of inner halo stars , with sim - lite providing _ in situ _ measures of stars in the outer halo .
sim - lite , along with vlba astrometry on the few lg galaxies with detectable masers ( like ic10 and m33 ) , will allow the orbits of galaxies within the lg and beyond to be derived .
this will make it possible to constrain the matter distribution over mpc scales and test cosmological expectations such as infall along filaments ( shaya et al .
there have been some tantalizing suggestions of late lg infall by the discovery of some hypervelocity dwarf galaxies " near m31 , namely and xii and and xiv .
given most reasonable estimates for the m31 mass , these systems have radial velocities suggesting that the dwarfs are not presently bound to m31 or even the lg ( majewski et al .
2007 , chapman et al . 2007 ) ; m31 would need to be at least two times more massive than previously thought to keep these satellites bound to it .
meanwhile , analysis of the hst - based proper motions of the mcs ( see above ) by besla et al .
( 2007 ) strongly suggests that these galaxies are making their first pass around the mw , which implies that they are among the latest hierarchical accumulations of matter by the mw .
of course , the mcs are dirr type systems , and it is interesting that their hyperbolic orbits are consistent with the overall distributions of dwarf satellites of different types within the lg : it is well known that there is a general density - morphology relationship within the lg , with dsph and de galaxies mostly clustered around m31 and the mw , while most dirr galaxies the mcs among the notable and nettlesome exceptions are typically found in isolated regions of the lg .
but the new mc orbital data suggest ( 1 ) that the mcs violate the density - morphology relationship by being near the mw _ now _ due to a coincidence of fortuitous timing in the dynamical evolution of the mcs and the lg , and ( 2 ) that the mcs may be on the first stages of transformation from a dirr morphology to another class of dwarf system .
such evolution of the morphologies of lg dwarf galaxies should be reflected in the star formation histories ( sfhs ) of these systems . through systematic high resolution imaging
especially with the hubble space telescope combined with stellar population synthesis , the sfhs of these lg dwarf galaxies are being painstakingly assembled ( e.g. , see summary in dolphin et al .
this work shows clear differences in the sfhs among nearby dwarf galaxies , with a clear morphology - density - sfh relation that strongly demonstrates how environment is a major driver in the evolution of small galaxies .
this may occur through the tidal shocking of more frequent gravitational encounters with the large lg spirals that accelerates star formation by fostering bursts and leading to an earlier exhaustion of gas , or it may relate to this gas being stripped out by ram pressure interaction with hot coronal gas around the large systems .
the morphology - density - sfh relation also suggests the possibility , long discussed , that dsph galaxies and dirrs may be evolutionarily related .
such a connection is made via @xmath19-body simulations ( e.g. , mayer et al .
2001a , b , klimentowski et al . 2009 ) that show _ tidal stirring _ to be an active agent transforming late type morphologies e.g. , high surface brightness dwarf disks into dsph or de - like systems when placed inside a mw - sized halo .
this process involves the formation and subsequent buckling of bars in the evolving dwarf , which transforms it from a rotation to pressure - supported system . in this regard
it is interesting that the large mc ( and possibly the small mc ) contains a bar .
if , as these models suggest , dirrs are simply systems that havent yet been dynamically stirred , then the lg dirrs should be on very large lg orbits , or they may be the most recent infalling systems into the lg and onto the large lg spirals ( as e.g. , the mcs ) ; thus dirrs may represent the most pristine " of luminous subhalos . but much work remains to verify this picture , especially given the arguments made by , e.g. , grebel et al .
( 2003 ) that the central surface brightness - luminosity and luminosity - metallicity distributions of dwarf galaxies do not favor an evolutionary connection between dirrs and dsphs , but rather a closer evolutionary connection between dsphs and transitional " type dwarfs ( like the antlia , phoenix and lgs3 systems ) a connection that , nevertheless , still belies environment ( i.e. , tidal forces ) driving the differential evolution of dwarf galaxies .
( on the other hand , brown et al .
2007 show that the dirr leo a is similar to dsphs in having a large velocity dispersion and @xmath20 , i.e. @xmath21 in solar units . ) clearly , more stellar populations work in particular , ever deeper hst cmds on the lg dirrs could help resolve these questions of dwarf evolution .
the acs nearby galaxy survey treasury ( angst ) project is a great help in this regard , taking this detailed cmd work out to 4 mpc ( dalcanton et al .
2009 ) . with the advent of ever larger telescopes , as well as more efficient instruments on existing 10-m class telescopes
, we also have the promise in the coming decade of detailed chemistry of individual stars in these systems , as is now being done on the closer , lg dsphs ( see 7 ) . of course
, more accurate information on the orbits of these systems , and understanding better the overall dynamics of the lg , would be extremely helpful in probing the evolutionary history of dwarf galaxies , beyond the tantalizing example offered by the mcs ( 3 ) .
of course , another prediction of cdm models is that infalling subhalos onto mw - like systems , which continue to the present , will face dynamical shredding and create a web of luminous ( and dark ) streams around these systems ( e.g. , bullock & johnston 2005 ) .
this vision of a mw halo networked by streams has been hinted at for several decades by small , pointed pencil - beam surveys with the conclusion that the _ entire _ halo is made from accreted systems ( see , e.g. , majewski 1993 , 2004 , majewski et al .
1996 ) . but
this vision of the mw really has been born out by deep , wide - field photometric surveys like the two micron all - sky survey ( 2mass ) and the sloan digital sky survey ( sdss ) which have enabled the construction of large portraits of the halo where many of these streams can be seen plainly ( e.g. , newberg et al .
2002 , majewski et al .
2003 , belokurov et al .
2006 , grillmair 2006 )
. continued active mining of these data bases continues to unveil not only new tidal debris streams ( e.g. , rocha - pinto et al .
2004 , sharma et al .
2009 for 2mass , and belokurov et al . 2007b , grillmair 2009 , newberg et al .
2009 for sdss , to name just some examples ) , but of course a plethora of new dsph galaxies of extremely low luminosity ( e.g. , willman et al . 2005 , zucker et al .
2006 , belokurov et al .
2007a , irwin et al .
2007 , grillmair 2009 ) .
these ultrafaint dsphs tend to have very high @xmath20 ratios ( muoz et al .
2006b , simon & geha 2007 ) that point to a common dsph mass scale ( strigari et al .
2008 ) , which may arise from some critical scale in the formation of galaxies or a characteristic scale for the clustering of dm .
but a few ultrafaints show evidence for tidal disruption and dynamical instability ( e.g. , coleman et al . 2007 , carlin et al . 2009 ) and may represent exceptions to the common mass scale ( adn et al .
2009 ) .
.wide field photometric surveys past and future [ cols="<,<,<,<,<,<",options="header " , ] significant advances in the understanding of the chemical evolution of our galaxy and its network of disrupted dsphs can be expected in the next decade , where huge databases of medium to high resolution spectroscopy of galactic stars will be generated with a variety of instruments ( table 3 ) .
obviously , the accumulation of high quality radial velocities for many millions of stars will also shape a very detailed understanding of the dynamics of galactic stellar populations .
not only will the galactic halo and its substructure be thoroughly probed by these upcoming surveys , but so too will the populations of the inner galaxy including those typically hidden by dust obscuration by infrared spectroscopic surveys like apogee ( see schiavon et al .
, this proceedings ) .
of course , @xmath0cdm surveys show galaxies forming _ entirely _ from hierarchical merging in an inside - out " manner , so detailed studies of the inner mw are essential to unlocking the dust - concealed story of its early formation . because of this dust obscuration , explorations at infrared wavelengths hold the most promise for this work .
fortunately , the coming decade brings a host of new infrared capabilities to add to the near - infrared ( nir ) imaging brought by 2mass and other more focused surveys . the ukidss project ( lucas et al . 2008 ) will increase the depth of these nir imaging probes at low galactic latitudes , and also provide second epoch positional data that can be combined with 2mass for proper motions ( e.g. , deacon et al . 2009 ) .
a more focused effort of @xmath22-band astrometry of the galactic disk and bulge is the goal of the japan astrometry satellite mission for infrared exploration ( jasmine ; gouda et al .
. the spitzer space telescope , which has significantly opened up longer wavelength photometric and spectroscopic study of the inner mw , will continue to provide longer wavelength imaging capability even as it enters its more wavelength - limited , warm mission .
in particular , the infrared array camera ( irac ) will continue to map the galactic midplane at 3.6 and 4.5@xmath1 m through the glimpse-360 mission , extending the earlier glimpse i and ii projects ( churchwell et al .
2009 ) , which have been instrumental in uncovering the structure of the central galaxy ( e.g. , benjamin et al .
even more coverage ( wavelength and area ) will be given by the wide - field infrared survey explorer ( wise ) , to be launched in the autumn of 2009 . and
apogee will provide high resolution nir spectroscopy of stars otherwise inaccessible behind tens of magnitudes of optical extinction . despite the diminished effects of infrared studies to the presence of dust
, it can not be completely ignored . continued efforts must be made to understand the galactic extinction law , especially given that it is not universal either by ism density or by position in the galaxy ( e.g. , fitzpatrick & massa 2009 , zasowski et al .
2009 , gao et al .
fortunately , the combination of nir and mir photometry ( e.g. , 2mass / ukidss + glimpse / wise ) gives direct information on the reddening foreground to each star , which will make it possible in the future to develop maps not only of the 3-d distribution of stars in the galactic disk , but also of the dust ( zasowski et al .
, in preparation ) .
i appreciate the assistance of richard j. patterson and jeffrey l. carlin with collecting the data for and creating the tables in this paper .
i acknowledge support from nsf grants ast-0607726 and ast-0807945 , as well as support by the _ sim lite _ key project _ taking measure of the milky way _ under nasa / jpl contract 1228235 . | the last decade has seen enormous progress in understanding the structure of the milky way and neighboring galaxies via the production of large - scale digital surveys of the sky like 2mass and sdss , as well as specialized , counterpart imaging surveys of other local group systems . apart from providing snaphots of galaxy structure , these cartographic " surveys lend insights into the formation and evolution of galaxies when supplemented with additional data ( e.g. , spectroscopy , astrometry ) and when referenced to theoretical models and simulations of galaxy evolution .
these increasingly sophisticated simulations are making ever more specific predictions about the detailed chemistry and dynamics of stellar populations in galaxies .
to fully exploit , test and constrain these theoretical ventures demands similar commitments of observational effort as has been plied into the previous imaging surveys to fill out other dimensions of parameter space with statistically significant intensity .
fortunately the future of large - scale stellar population studies is bright with a number of grand projects on the horizon that collectively will contribute a breathtaking volume of information on individual stars in local group galaxies .
these projects include : ( 1 ) additional imaging surveys , such as pan - starrs , skymapper and lsst , which , apart from providing deep , multicolor imaging , yield time series data useful for revealing variable stars ( including critical standard candles , like rr lyrae variables ) and creating large - scale , deep proper motion catalogs ; ( 2 ) higher accuracy , space - based astrometric missions , such as gaia and sim - lite , which stand to provide critical , high precision dynamical data on stars in the milky way and its satellites ; and ( 3 ) large - scale spectroscopic surveys provided by rave , apogee , hermes , lamost , and the gaia spectrometer , which will yield not only enormous numbers of stellar radial velocities , but extremely comprehensive views of the chemistry of stellar populations .
meanwhile , previously dust - obscured regions of the milky way will continue to be systematically exposed via large infrared surveys underway or on the way , such as the various glimpse surveys from spitzer s irac instrument , ukidss , apogee , jasmine and wise .
= = = 1=1=0pt = 2=2=0pt = 2=2=0pt |
in this appendix we give the detailed expression for the the annihilation cross section shown in compact form in eq.([eq : sigmat ] ) , and calculated from the diagrams ( a)(s ) of fig .
[ fig2_last ] . in the following ,
masses of light particles are neglected , while we assume a common mass @xmath79 for the triplets and their supersymmetric partners . the reduced cross section introduced in eq .
( [ eq : sigmat ] ) is defined as : @xmath180 in terms of the integrated squared amplitude , averaged over the initial triplet state ( hence the factor 1/3 ) and summed over the coannihilating particles , given by : @xmath181+\nonumber\\ & + & \frac{4}{3}\sum_{ab}\left\{tr(t_a^2 t_b^2)\left [ f_{(c)+(d)+(e)}+f_{(n)+(o)}+f_{(p)+(q)+(r ) } \right ] + \right .
\nonumber\\&+ & \left . 8 f_{abc } f_{abc } \left [ f_{(f)}+f_{(m)}+f_{(s)}+f^{\prime}_{(d)}\right ] \right\}\ , , \quad\mbox{where } \label{eq : sigmat_detail}\end{aligned}\ ] ] @xmath182 , \;\;\;\;f_{(n)+(o)}=2 + 2\left[-2+\frac{1}{\beta}\ln\frac{1+\beta}{1-\beta } \right ] , \nonumber\\ & & f_{(p)+(q)+(r)}=4 + 2\left[-2+\frac{1}{\beta}\ln\frac{1+\beta}{1-\beta } \right]\;\;\;\;f_{(f)}=\left [ 1-\frac{5}{6}\beta^2-\frac{(\beta^2 - 1)^2}{2\beta}\ln\frac{1+\beta}{1-\beta}\right ] , \nonumber\\ & & f_{(m)}=\frac{\beta^2}{3}+\left[1+\frac{\beta^2 - 1}{2\beta}\ln\frac{1+\beta}{1-\beta } , \right].\nonumber \label{eq : fterms}\end{aligned}\ ] ] in the last equation we have kept within squared parentheses quantities that vanish for @xmath183 , and the subscripts refer to the contributing feynman diagrams listed in fig
. [ fig2_last ] . in eq .
( [ eq : sigmat_detail ] ) , @xmath184 and @xmath185 are the su(2)@xmath186 u(1 ) group generators for the triplet and doublet representations and @xmath187 are the structure constants . assuming the minimal supersymmetric standard model particle content , the traces are given by : @xmath188 and @xmath189 . since annihilation decouples for @xmath190 , the integral in eq .
( [ eq : sigmat_int ] ) can be approximated by making use of the following low - temperature expansion : @xmath191 ^ 2}\ , t^2 \beta(t)\ ,
\sigma(t)\simeq \frac{1}{2 z^3}\left[b_0+\frac{b_1}{z}+ ... \right ] , \\
\mbox{where } & & b_0= 47 + 32 t_w^2 + \frac{49}{2 } t_w^4 \quad \mbox{and}\quad b_1= -\frac{3}{2}\left ( \frac{98}{3}+32 t_w^2 + 19 t_w^4 \right ) .
\label{eq : expansion}\end{aligned}\ ] ] although for our results we used a numerical integration of eq.([eq : sigmat_int ] ) , we have checked that the above approximation leads to a good fit to the full numerical calculation for @xmath192 10 , an interval that safely includes the range of @xmath84 relevant for the present analysis .
we finally notice that the annihilation amplitude increases sizably in the supersymmetric theory compared to the standard model case , in significant excess of the factor o(2 ) suggested by a naf expectation .
in fact , the value of @xmath193 in eq .
( [ eq : expansion ] ) is almost 8 times larger than the standard model value @xmath194 coming from the diagrams ( c)(f ) .
such an enhancement is mainly due to a larger number of available final states for the diagrams ( i ) and ( l ) , corresponding to the `` contact term '' for scalars and to triplet
striplet annihilation to gauginos , respectively .
# 1#2#3phys .
* b#1 * , # 2 ( # 3 ) # 1#2#3nucl .
. * b#1 * , # 2 ( # 3 ) # 1#2#3phys .
* d#1 * , # 2 ( # 3 ) # 1#2#3phys .
lett . * # 1 * , # 2 ( # 3 ) # 1#2#3rev .
* # 1 * , # 2 ( # 3 ) # 1#2#3mod . phys .
* a#1 * , # 2 ( # 3 ) # 1#2#3phys .
rept . * # 1 * , # 2 ( # 3 ) # 1#2#3science * # 1 * , # 2 ( # 3 ) # 1#2#3astrophys . j. * # 1 * , # 2 ( # 3 ) # 1#2#3eur .
j. * c#1 * , # 2 ( # 3 ) # 1#2#3jhep * # 1 * , # 2 ( # 3 ) # 1#2#3prog .
# 1 * , # 2 ( # 3 ) # 1#2#3int .
j. mod .
* a#1 * , # 2 ( # 3 ) m. fukugita and t. yanagida , .
r. barbieri , d.v .
nanopolous , g. morchio and f. strocchi , phys .
b * 90 * , 91 ( 1980 ) ; m. magg and ch.wetterich , phys .
b * 94 * , 61 ( 1980 ) ; j. schechter and j. w. f. valle , phys .
* d22 * ( 1980 ) 2227 ; t. p. cheng and l. f. li , phys . rev .
d * 22 * , 2860 ( 1980 ) ; r.n .
mohapatra and g. senjanovic , phys .
d * 23 * , 165 ( 1981 ) ; g. lazarides , q. shafi and c. wetterich , nucl .
b * 181 * , 287 ( 1981 ) .
p. odonnell and u. sarkar , ; g. lazarides and q. shafi , ; e.j .
chun and s.k .
kang , ; t. hambye and g. senjanovic , ; w. rodejohann , hep - ph/0403236 ; p. gu and x. bi , hep - ph/0405092 ; s. antusch and s. king , hep - ph/0405093 ; hep - ph/0507333 ; w. guo , hep - ph/0406268 .
gunion , j. grifols , a. mendez , b. kayser and f. olness , ; r. vega and d. dicus , ; j.f .
gunion , r. vega and j. wudka , ; r. godbole , b. mukhopadhyaya and m. nowakowski , ; k. cheung , r. phillips and a. pilaftsis , ; k. huitu , j. maalampi , a. pietila and m. raidal , ; t.g .
rizzo , ; n. lepore , b. thorndyke , h. nadeau and d. london , ; j.f .
gunion , c. loomis and k.t .
pitts , hep - ph/9610237 ; b. dion _ et .
_ , ; a. datta and a. raychaudhuri , ; e. ma , m. raidal and u. sarkar , ; .
e. j. chun , k.y .
lee and s.c .
park , phys .
b * 566 * , 142 ( 2003 ) ; a. akeroyd and a. aoki , . | we consider the minimal supersymmetric triplet seesaw model as the origin of neutrino masses and mixing as well as of the baryon asymmetry of the universe , which is generated through soft leptogenesis employing a cp violating phase and a resonant behavior in the supersymmetry breaking sector . we calculate the full gauge
annihilation cross section for the higgs triplets , including all relevant supersymmetric intermediate and final states , as well as coannihilations with the fermionic superpartners of the triplets .
we find that these gauge annihilation processes strongly suppress the resulting lepton asymmetry . as a consequence of this
, successful leptogenesis can occur only for a triplet mass at the tev scale , where the contribution of soft supersymmetry breaking terms enhances the cp and lepton asymmetry .
this opens up an interesting opportunity for testing the model in future colliders .
leptogenesis is an elegant way to generate the baryon asymmetry of the universe in connection with the origin of the observed neutrino masses and mixing through the seesaw mechanism @xcite .
one way of understanding a tiny neutrino mass is to relate it with the small vacuum expectation value of a higgs triplet @xcite whose decay can also induce the cosmological baryon asymmetry in the presence of at least two higgs triplets @xcite or a right - handed neutrino @xcite as required by the generation of non - trivial cp and lepton asymmetry . in the minimal supersymmetric version with one pair of triplets , there is a new way of leptogenesis ( called `` soft leptogenesis '' ) in which cp phases in the soft terms can contribute to generate the lepton asymmetry @xcite .
soft leptogenesis in the minimal supersymmetric higgs triplet model has been considered first in ref .
@xcite . in this paper
, we revisit this last scenario to provide a careful analysis on the quantities for the lepton and cp asymmetries and their cosmological evolution by considering the full set of boltzmann equations including thermal masses and the temperature supersymmetry breaking effects consistently .
we will also derive a set of simple boltzmann equations from the maxwell - boltzmann approximation taking into account the difference between the bose
einstein and fermi dirac statistics , and show that they provide a fairly good approximation to the full boltzmann equations . the most important effect included in our analysis is the contribution of the gauge annihilation processes , which lead to a significant reduction of the resulting lepton asymmetry for the low higgs triplet mass .
the dynamics of such a system is analyzed in ref .
@xcite for the case of the conventional baryogenesis with heavy higgs bosons in the @xmath0 unification scheme .
our analysis is extended to the lowest possible values of the higgs triplet mass where , as will be shown in the following , the annihilation effect dominates over decays and inverse decays .
another crucial ingredient of soft leptogenesis is the suppression of the asymmetry due to a small difference between boson and fermion statistics at finite temperature .
we find that this effect for low values of the triplet mass becomes subleading compared to that due to soft supersymmetry breaking terms .
let us also note that the annihilation effect becomes irrelevant for a triplet mass higher than about @xmath1 gev @xcite , for which , however , the lepton asymmetry in the soft leptogenesis scenario is also suppressed , as it is inversely proportional to the triplet mass @xcite . as a result , we will conclude that the required baryon asymmetry can be generated only at the multi - tev range of the higgs triplet mass , and thus the model can lead to distinct collider signatures through , in particular , the production and decay of a doubly charged higgs boson @xcite .
this opens up another interesting possibility for generating the neutrino masses and mixing as well as the cosmological baryon asymmetry at the tev scale , which can be tested in future colliders @xcite . in the supersymmetric form of the higgs triplet model @xcite
, one needs to introduce a vector - like pair of @xmath2 and @xmath3 with hypercharge @xmath4 and @xmath5 , allowing for the renormalizable superpotential as follows : @xmath6 where @xmath7 contains the neutrino mass term , @xmath8 .
the soft supersymmetry breaking terms relevant for us are @xmath9 note that we have used the same capital letters to denote the superfields as well as their scalar components .
we will consider the universal boundary condition of soft masses ; @xmath10 and @xmath11 . in the limit @xmath12
, the higgs triplet vacuum expectation value @xmath13 gives the neutrino mass @xmath14 the mass matrix of the scalar triplets is diagonalized by @xmath15 where @xmath16 are the mass eigenstates with the mass - squared values , @xmath17 , and the mass - squared difference , @xmath18 . in terms of the mass eigenstates ,
the lagrangian becomes @xmath19 + h.c .
\nonumber\end{aligned}\ ] ] the heavy particles @xmath20 decay to the leptonic final states , @xmath21 , as well as the higgs final states , @xmath22 and @xmath23 .
thus , the out - of - equilibrium decay @xmath24 can lead to lepton asymmetry of the universe . in order to discuss how to generate a lepton asymmetry in the supersymmetric triplet seesaw model
let us first consider the general case of a charged particle @xmath25 ( @xmath26 ) decaying to a final state @xmath27 ( @xmath28 ) and generating tiny cp asymmetric number densities , @xmath29 and @xmath30 .
the relevant boltzmann equations in the approximation of maxwell boltzmann distributions are @xmath31 \nonumber\\ { d y_x \over d z } & = & - z k \gamma_d \left [ y_x- \sum_k 2 b_k { y_x^{eq}\over y_k^{eq } } y_k \right ] \\ { d y_j \over
d z } & = & 2 z k \gamma_d\left [ \epsilon_j ( y_x - y_x^{eq } ) + b_j ( y_x - 2 { y_x^{eq } \over y_j^{eq } } y_j ) \right ] \nonumber\end{aligned}\ ] ] where @xmath32 s are the number densities in unit of the entropy density @xmath33 as defined by @xmath34 , @xmath35 and @xmath36 . here
, the cp asymmetry @xmath37 in the decay @xmath38 is defined by @xmath39 in eq .
( [ boltzmann ] ) , @xmath40 with the hubble parameter @xmath41 at the temperature @xmath42 , and @xmath43 is the branching ratio of the decay @xmath38 .
for the relativistic degrees of freedom in thermal equilibrium @xmath44 , we will use the supersymmetric standard model value : @xmath45 .
the evolution of the @xmath25 abundance is determined by the decay and inverse decay processes , as well as by the annihilation effect described by the diagrams of fig .
[ fig2_last ] , and are accounted for by the functions @xmath46 and @xmath47 , respectively . note that the triplets are charged under the standard model gauge group and thus have nontrivial gauge annihilation effect which turns out to be essential in determining the final lepton asymmetry .
moreover , as a consequence of unitarity , the relation @xmath48 holds , so that one can drop out the equation for @xmath49 , taking the replacement : @xmath50 in the last of eqs .
( [ boltzmann ] ) . in our model , the heavy particle @xmath25 can be either of the six charged particles ; @xmath51 or @xmath52 .
each of them follows the first boltzmann equation in eq .
( [ boltzmann ] ) where @xmath46 and @xmath47 are given by @xmath53 with @xmath54\nonumber \\ & & + 4 \left [ -3+\left(4-\beta^2+\frac{(\beta^2 - 1)(2-\beta^2)}{\beta}\ln\frac{1+\beta}{1-\beta}\right ) \right ] , \label{eq : sigmat}\end{aligned}\ ] ] where @xmath55 with @xmath56 the weinberg angle , and @xmath57 .
the function @xmath46 is the ratio of the modified bessel functions of the first and second kind which as usual takes into account the decay and inverse decay effects in the maxwell boltzmann limit .
the function @xmath47 accounts for the annihilation cross - section of a triplet component @xmath25 summing all the annihilation processes ; @xmath58 standard model gauge bosons / gauginos and fermions / sfermions where @xmath59 is some triplet component or its fermionic partner .
the separate contribution of each diagram in fig .
[ fig2_last ] is detailed in the appendix .
as far as the standard model part is concerned , our result agrees with that of ref .
@xcite , with one exception : the term proportional to @xmath60 , due to the mixed gauge boson ( @xmath61 ) final state in diagrams ( c f ) of fig . [ fig2_last ] , is missing in ref .
@xcite . however , this difference concerns a subdominant contribution which is expected to have a negligible impact on phenomenology .
decay and inverse - decay amplitudes entering in the boltzmann equations .
the solid lines show the decay amplitude @xmath46 in the maxwell bolzmann limit as given by eq .
( [ eq : gamma_d ] ) , and the corresponding inverse
decay amplitude @xmath62 .
the dotted and dashed curves show the result of a full numerical evaluation of the same amplitudes for fermionic and bosonic final states , respectively .
, scaledwidth=65.0% ] the decay and inverse decay amplitudes in the maxwell boltzmann limit are plotted in fig .
[ fig1_last ] , along with a numerical evaluation of the same quantities in the case of bosonic and fermionic final states , where bose einstein and fermi dirac distributions , as well as thermal masses , are included in the calculation .
we use this latter evaluation when we solve the full boltzmann equations for the lepton asymmetry numerically .
the last figure shows that the boltzmann approximation is well justified as expected for the region of our relevance , @xmath63 .
fractional departure of the triplet comoving density @xmath64 from its equilibrium value @xmath65 , as a function of @xmath66 .
the higher curve shows the result of a calculation where the gauge annihilation effect is neglected , while the lower ones show the same quantity including annihilation for @xmath67=1 tev and @xmath68 8,7,6,5,4,3 from left to right .
all curves are evaluated in the maxwell
boltzmann approximation.,scaledwidth=65.0% ] given @xmath46 and @xmath47 , we can now analyze the thermal evolution of @xmath64 @xcite . in fig .
[ fig3_last ] , we plot the quantity @xmath69 , which quantifies the departure of the triplet density from its equilibrium value . in particular , the higher line shows the result when only the processes of decay and inverse decay to light particles are included in the calculation .
as expected , since @xmath70 , @xmath64 follows closely the equilibrium density @xmath71 with a slight deviation of order @xmath72 .
however , annihilation is indeed important in our case and can not be neglected .
this is shown in the same figure by the lower curves , which represent the departure of the triplet density from its equilibrium value when annihilation is included .
the importance of annihilation can be understood in the following way .
the inverse decay freezes out at @xmath73 for @xmath74 as @xmath75 . on the other hand ,
the thermal averages of the annihilation and decay rate can be compared by considering the following ratio @xcite : @xmath76 where @xmath77 .
thus , the annihilation effect becomes negligible for @xmath78 gev .
but in our case of soft leptogenesis , higher @xmath79 suppresses the lepton asymmetry as @xmath80 , so there is a tension between these two effects , and lower values of @xmath79 turn out to be favored . in fig .
[ fig3_last ] , one can see that , due to annihilation which freezes out at @xmath81 , @xmath64 follows more closely its equilibrium density @xmath71 compared with the previous case , with a deviation which is now of order @xmath82 .
in particular , this implies that the approximation @xmath83 is a good one , since decoupling occurs indeed at high @xmath84 .
nevertheless , in our numerical analysis , we solve the full boltzmann equations where the bose einstein and fermi dirac distributions as well as thermal masses , are included properly . to find out the cosmological lepton asymmetry by the decay of @xmath85
, one needs to calculate @xmath86 with the states @xmath87 and @xmath88 and thus the corresponding cp asymmetry : @xmath89 recall that one can not rely on the the above boltzmann equation ( [ boltzmann ] ) for the mechanism of soft leptogenesis in the supersymmetric limit of @xmath90 , as the cp asymmetries in the bosonic and fermionic final states takes the opposite sign , @xmath91 , so that the total asymmetry in the lepton number density vanishes , @xmath92 .
a non - vanishing lepton asymmetry arises after taking into account the supersymmetry breaking effect at finite temperature @xcite , namely the difference between the bosonic and fermionic statistics given by the bose
einstein and fermi dirac distribution , respectively .
such a thermal supersymmetry breaking effect can be well accounted by a slight modification of the last boltzmann equation of eq .
( [ boltzmann ] ) resulting from the extension of the usual maxwell
boltzmann approximation to the second order , as we will show below .
the dashed curve shows the approximation to @xmath134 in eq .
( [ eq : delta_bf ] ) , while the solid line is the result of a numerical evaluation of the same quantity , which includes the effect of thermal masses and of fermi dirac and bose einstein distributions .
the dotted curve shows the result of a numerical calculation of the thermal average @xmath139 which multiplies the soft supersymmetry breaking term in eq .
( [ eq : lepton_asym]).,scaledwidth=65.0% ] concluding the above discussions , we find that the total lepton asymmetry density @xmath140 follows the approximate boltzmann equation : @xmath141\ ] ] where @xmath142 counts the total number of triplet components generating the lepton asymmetry and @xmath143 . in the above equation , the number @xmath144 takes the minimal value of @xmath74 for @xmath145 as we have the relation @xcite ; @xmath146 as one goes away from the minimum value of @xmath147 with @xmath148 ) gets suppressed .
furthermore , one realizes that the resulting lepton asymmetry is maximized in case of @xmath149 with @xmath150 , in which case the boltzmann equation for the lepton asymmetry takes the simplest form of @xmath151 \label{eq : boltzmann}\ ] ] let us now note that , taking the resonance condition @xmath152 , one finds the maximal value of @xmath153 , which becomes order one for @xmath154 tev .
final lepton asymmetry produced by triplet decay as a function of @xmath79 .
different curves refer to @xmath155 tev from bottom to top.,scaledwidth=65.0% ] it is now easy to find the approximate solution for @xmath156 from eq .
( [ ylapp ] ) with the insertion of @xmath64 given in eq .
( [ yxapp ] ) .
both are found to be a fairly good approximation to the numerical solution of the full boltzmann equations , as expected from our previous discussions .
the results of our numerical calculation are shown in fig .
[ fig6_last ] where we plot the final lepton asymmetry as a function of the triplet mass @xmath79 for @xmath157 from 1 to 5 tev .
when @xmath158 , one needs to recover the contributions of order @xmath159 which were neglected , e.g. , in eqs .
( [ pis ] ) and ( [ epspm ] ) . taking the parameter region , @xmath160
, we keep those contributions in the numerical calculations for the curves in fig .
[ fig6_last ] .
a remark is in order here . for @xmath161 gev ,
the annihilation effect becomes irrelevant and the final asymmetry is determined by the decay and inverse decay effects , i.e. , by the value of @xmath147 only , which confirms the result of ref .
@xcite .
one can see this feature in fig .
[ fig6_last ] , where the lepton asymmetry as a function of @xmath79 changes slope at about @xmath162 gev . below this value
the annihilation effect sets in , and the final asymmetry is strongly suppressed compared to the value one would obtain by extrapolating the curve with the slope for @xmath163 gev . when the trilinear coupling is larger than the triplet mass , @xmath164 1 , besides an enhancement of the cp violating term of eq.([epspm ] ) , one could expect that the additional contribution to the coupling of the triplet particles to scalar final states enhances the total annihilation rate , increasing substantially the value of the parameter @xmath147 compared to the amount given by eq .
( [ eq : k ] ) , which is obtained in the limit @xmath165 . as a consequence of this
, the consequent additional wash out effect could in principle suppress the ensuing lepton asymmetry .
however , this is not the case due to the fact that , as we have already discussed , the annihilation process freezes out later than inverse decays , so the latter play almost no role in the determination of the epoch when lepton asymmetry production can start .
actually , this epoch starts when decays eventually overcome annihilations , so a higher value of @xmath147 can slightly anticipate it , leading so to a higher asymmetry instead than a suppression , although this effect is quite mild .
this is what we observe in the numerical calculation shown in fig .
[ fig6_last ] , where we have assumed as before @xmath150 ( in order to maximize the amount of @xmath166 violation given by eq .
( [ epspm ] ) ) and maximized the @xmath166violating phase ( i.e. , we have assumed re@xmath167 ) .
we also remark that sphaleron interactions are kept in thermal equilibrium even after the electroweak phase transition and freeze out around @xmath168 gev , so that only the lepton asymmetry produced for @xmath169 can be efficiently converted into a baryon asymmetry @xcite . as shown in fig .
[ fig3_last ] , due to the gauge annihilation effect , the lepton asymmetry production is delayed until @xmath170 . for low values of @xmath79 ( @xmath171 a few tev )
one can have @xmath172 , which implies a suppression in the final lepton asymmetry .
this explains the fast rise at low values of @xmath79 of all the curves in fig.[fig6_last ] . on the other hand , the dips observed in the final asymmetry for @xmath173 correspond to the case when the two contributions @xmath132 and @xmath174 in eq .
( [ epspm ] ) are of the same order and cancel .
this dip separates the two regions where @xmath132 or @xmath174 dominates in the determination of the final asymmetry . as shown in fig .
[ fig6_last ] , the cp violating contribution from the soft supersymmetry breaking term @xmath174 in eq .
( [ eq : lepton_asym ] ) can strongly enhance the final lepton asymmetry at low values of @xmath79 . as a result ,
it is evident that the required baryon asymmetry can be reached whenever @xmath175 and @xmath79 are in the multi - tev region . before concluding our work ,
let us remark some experimental consequences of the model at future colliders . as shown above
, successful baryogenesis requires a tev - scale triplet mass and yukawa couplings of the same order , @xmath176 .
thus , all the low - energy lepton flavor violating processes like @xmath177 or @xmath178 are highly suppressed @xcite . on the other hand ,
future accelerators have a potential to produce such higgs triplets , in particular , the peculiar doubly charged component through the drell - yan processes @xcite .
then , various features of the model can be checked by observing the branching ratios of the triplet decay to lepton and higgsino pairs , in particular , @xmath179 , allowing also to study neutrino mass patterns @xcite . in conclusion , we have investigated baryogenesis assuming the minimal supersymmetric higgs triplet model as the origin of neutrino masses and mixings .
this model , with only one pair of triplets , can provide a mechanism for soft leptogenesis employing a cp violating phase and a resonant behavior in the supersymmetry breaking sector .
our analysis shows that the original soft leptogenesis , relying on the supersymmetry breaking effect proportional to the small difference between boson and fermion statistics at finite temperature can not produce the right amount of baryon asymmetry due to the gauge annihilation effect .
in particular , we have calculated the full gauge annihilation cross section including all the relevant supersymmetric intermediate and final states , as well as coannihilations with the fermionic superpartners of the triplets , finding that this effect strongly suppresses the resulting lepton asymmetry . on the other hand
, the contribution of soft supersymmetry breaking terms , particularly a sizable value for the @xmath67 parameter , can enhance the lepton asymmetry to provide successful leptogenesis if the triplet mass is in the tev range . in this case
, the model predictions can be tested in future colliders by searching for a very clean signal , e.g. , from the production and decay of doubly charged higgs bosons . |
a feynman diagram can be understood mathematically as a linear combination of horn - type hypergeometric functions of several variables ( see ref .
@xcite ) : @xmath2 where a horn - type hypergeometric function has the structure @xmath3 with the arguments @xmath4 being , in general , rational functions ( typically , simple ratios ) of kinematic invariants of the original feynman diagram , and the parameters @xmath5 and @xmath6 being linear combinations of the exponents of propagators and the dimension of space - time . the @xmath7 and @xmath8 are called upper and lower parameters , respectively .
these statements follow from the multiple mellin - barnes representation for a dimensionally regularized feynman diagram ( see ref .
@xcite ) , and the assumption that there is a region of variables where every term in the linear combination ( [ fd ] ) is convergent .
the horn - type structure permits the hypergeometric functions appearing in ( [ fd ] ) to be reduced to a set of basis functions with parameters differing from the original ones by integer shifts : @xmath9 where @xmath10 denotes a partial derivative and @xmath11 are rational functions @xcite .
these shifts may be implemented by constructing a set of four differential operators @xmath12 , @xmath13 which respectively change @xmath14 , @xmath15 by @xmath16 : @xmath17 or @xmath18 .
these basic operators are called the step - up and step - down operators for the upper and lower parameters .
a procedure of applying step - up and step - down operators to reduce the original hypergeometric function to a basis set is called a * differential reduction*. in the case when some of the variables are equal to one another , @xmath19 , or belong to the surface of singularities @xmath20 , it is necessary to define a limiting procedure for ( [ dr ] ) .
we will illustrate our approach by considering the reduction of a particular generalized hypergeometric function in section [ generalized ] , and by applying it to a particular class of feynman diagrams in section [ example ] .
section [ hyperlogarithms ] describes the reduction at a singular surface using the @xmath21 expansion and hyperlogarithms .
in this section , we will show how differential reduction may be applied to a generalized hypergeometric function of one variable .
let us recall that the generalized hypergeometric function @xmath22 may be defined in a neighborhood of @xmath23 by the series @xmath24 where @xmath25 is called a pochhammer symbol .
the lists @xmath26 and @xmath27 are called upper and lower parameters of the hypergeometric function , respectively . the hypergeometric function @xmath28 satisfies a differential equation @xmath29 { } _ pf_{p-1}(\vec{a } ; \vec{b } ; z ) = 0 \ ; , \label{diff}\ ] ] where @xmath30 is a differential operator and @xmath31 .
constructing a reduction scheme requires a set of step - up and step - down differential operators for both the upper and lower parameters . in this case
, the universal step - up ( step - down ) operators for the upper ( lower ) parameters have a very simple form : @xmath32 and the inverse operators @xmath33 , and @xmath34 can be constructed in accordance with takayama s algorithm .
@xcite ( see also ref .
@xcite . )
the differential reduction algorithm takes the form of a product of several differential step - up / step - down operators @xmath35 : @xmath36 so that the maximal power of @xmath37 in this expression is @xmath38 .
since the hypergeometric function @xmath39 satisfies a differential equation ( [ diff ] ) of order @xmath40 , it is possible to express all terms containing powers of @xmath41 with @xmath42 in terms of @xmath43 with @xmath44 , multiplied by coefficients that are rational functions of the parameters and the argument @xmath45 . in this way
, any function @xmath46 may be expressed in terms of a basic function @xmath39 and its first @xmath40 derivatives : @xmath47 where @xmath48 are lists of integers and @xmath49 and @xmath50 are polynomials in the parameters @xmath51 and @xmath45 . for some special sets of parameters ,
the result of the reduction ( [ decomposition ] ) takes a simple form .
in particular , when one of the upper parameters is an integer , @xmath52 for further details see ref .
+ r5.0 cm
as an example of how the differential reduction applies to a particular type of feynman diagram , let us consider the @xmath1-loop bubble diagram @xmath53 in fig .
[ example ] , with four massive lines ( with masses @xmath54 , @xmath55 as indicated ) and two sets of massless subloops with @xmath56 and @xmath57 lines , respectively .
it is defined as @xmath58^{\rho_1 } \cdots [ k_{r-1}^2]^{\rho_{r-1 } } [ ( k_1 \!+\ ! \cdots \!+\ !
p)^2]^{\rho_r } } \nonumber \\ & & \hspace{15 mm } \times \frac{1 } { [ k_{r+x}^2-m^2]^{\alpha_1 } [ ( k_r \!+\ ! \cdots \!+\ ! k_{r+x
p)^2-m^2]^{\alpha_2 } } \\ & & \hspace{15 mm } \times \frac{1 } { [ k_r^2]^{\sigma_1 } \cdots [ k_{r+x-1}^2]^{\sigma_x } [ k_{q-1}^2-m^2]^{\beta_1 } [ k_{q}^2-m^2]^{\beta_2 } } \ ; , \nonumber \end{aligned}\ ] ] where @xmath59 and @xmath60 the mellin - barnes representation for this diagram is and @xmath61 . ]
@xmath62^q ( -m^2)^{\tfrac{n}{2}\!-\!\alpha_1\!-\!\alpha_2 } ( -m^2)^{\tfrac{n}{2}(q\!-\!1)\!-\!\rho\!-\!\sigma\!-\!\beta_1\!-\!\beta_2 } } { \gamma(\alpha_1 ) \gamma(\alpha_2 ) \gamma(\beta_1 ) \gamma(\beta_2 ) \gamma\left ( \frac{n}{2 } \right ) } \nonumber \\ & & \hspace{15 mm } \times \left\ { \prod_{i=1}^{x } \frac{\gamma(\tfrac{n}{2}-\sigma_i)}{\gamma(\sigma_i ) } \right\ } \left\ { \prod_{j=1}^{r } \frac{\gamma(\tfrac{n}{2}-\rho_j)}{\gamma(\rho_j ) } \right\ } \frac { \gamma\left(\rho-\tfrac{n}{2}(r-1)\right ) } { \gamma\left(\tfrac{n}{2}r-\rho \right ) } \nonumber \\ & & \hspace{15 mm } \times \int ds \left ( \frac{m^2}{m^2}\right)^s \gamma\left(\sigma\!-\!\tfrac{n}{2}x \!-\!s\right ) \gamma\left(\beta_1\!+\!\beta_2\!+\!\rho\!+\!\sigma\!- \!\tfrac{n}{2}(q\!-\!1)\!-\!s\right ) \nonumber \\ & & \hspace{15 mm } \times \frac{\gamma(\alpha_1\!+\!s ) \gamma(\alpha_2\!+\!s ) \gamma(\alpha_1\!+\!\alpha_2\!-\!\tfrac{n}{2}\!+\!s ) \gamma(\tfrac{n}{2}\!+\!s ) \gamma(\tfrac{n}{2}(q\!-\!2 ) \!-\
! \rho \!-\ !
\sigma \!+\!s ) } { \gamma(\alpha_1\!+\!\alpha_2\!+\!2s ) \gamma(\tfrac{n}{2}(x\!+\!1)\!-\!\sigma\!+\!s ) } \nonumber \\ & & \hspace{15 mm } \times \frac { \gamma(\beta_1\!+\!\rho\!+\!\sigma\!-\!\tfrac{n}{2}(q\!-\!2)\!-\!s ) \gamma(\beta_2\!+\!\rho\!+\!\sigma\!-\!\tfrac{n}{2}(q\!-\!2)\!-\!s ) } { \gamma(\beta_1\!+\!\beta_2\!+\!2 \sigma \!+\ !
2 \rho \!- \ !
n(q\!-\!2 ) \!-\ ! 2 s ) } \ ; , \label{b112200mb}\end{aligned}\ ] ] where we have introduced the notations @xmath63 closing the contour of integration on the left ( see ref .
@xcite for details ) , the result may be expressed as a sum of four hypergeometric functions of type @xmath64 .
the general result is too long to reproduce here . in the case that the exponents of the propagators are all integers , the hypergeometric functions appearing in the result are reducible to the following hypergeometric functions and derivatives thereof : @xmath65 where @xmath66 are arbitrary integers . in the last expression ,
some polynomials are also generated .
further simplification is possible when @xmath67 take particular values .
for example , for @xmath68 and @xmath69 , the first hypergeometric function is reducible to @xmath70 or to @xmath71
in physical applications , the case of of equal masses @xmath72 ( a `` single - scale '' diagram ) is of special interest . and ref .
@xcite for the four - loop case @xmath73 . ] for the diagram in fig .
( [ example ] ) , this case corresponds to @xmath74 , which is a singular point for the differential reduction algorithm .
the question is then how to find a smooth limit at this point .
let us recall @xcite that the hypergeometric function ( [ hypergeometric ] ) converges at @xmath74 if @xmath75 . in this way , if the hypergeometric function on the l.h.s . of eq .
( [ decomposition ] ) is well - defined at @xmath74 , a smooth limit of the differential reduction exists .
it is now a technical problem to rewrite the r.h.s . of eq .
( [ decomposition ] ) in terms of the variable @xmath76 .
it is well - known , however , that for @xmath77 , the hypergeometric function @xmath78 is not expressible in terms of hypergeometric functions of the same type in the neighborhood of @xmath74 ( see ref .
@xcite ) .
one approach to this problem is to construct the all - order @xmath21-expansion . ] in terms of functions which are defined for the entire range @xmath79 .
the problem is then solved at each order in @xmath21 .
the hyperlogarithms @xcite belong to that class .
unfortunately , at present , the necessary theorems on the all - order @xmath21-expansion are proven only for special sets of parameters@xcite .
let us recall some results from ref .
there are bases ( sets of parameters ) for which the all - order @xmath21-expansion of a hypergeometric function has the form ( see ref .
@xcite ) @xmath80 where @xmath81 , @xmath82 , and @xmath1 is integer number , the coefficients @xmath83 are polynomials in the parameters @xmath84 and @xmath85 , @xmath86 is primitive @xmath87-root of unity , @xmath88 are hyperlogarithms , @xmath89 is short - hand for @xmath90 , @xmath91 is a variable related algebraically can take the explicit forms @xmath92 where @xmath1 is an integer . under @xmath93 ,
these variables transform as @xmath94 and @xmath95 . ] to @xmath45 , and @xmath96 is a polynomial .
thus , at each order in the @xmath21-expansion , only hyperlogarithms of a single weight are generated .
the definition of the hyperlogarithm as an iterated integral over any rational function @xmath97 where the @xmath98 are arbitrary numbers , together with the structure of the expansion ( [ expansion ] ) , show that the transformation @xmath93 results in functions of the same structure . in this way
, we can construct the necessary limiting procedure without detailed knowledge about the relationship between the functions @xmath99 and @xmath100 for practical applications to diagrams presently of interest , a few coefficients of the @xmath21-expansion suffice , and these have been implemented in several existing packages @xcite .
99 presentation : s.a .
yost , + http://ilcagenda.linearcollider.org/contributiondisplay.py?contribid=76&sessionid=18&confid=2628 m.yu .
kalmykov , v.v.bytev , b.a .
kniehl , b.f.l .
ward , s.a .
yost , proc .
acat 2008 [ arxiv:0901.4716 ] .
kalmykov , jhep * 0604 * ( 2006 ) 056 ; + m.yu .
kalmykov , b.a .
kniehl , b.f.l .
ward , s.a .
yost , proc . of quarks-08 , [ arxiv:0810.3238 ] .
boos , a.i .
davydychev , theor .
math . phys .
* 89 * ( 1991 ) 1052 .
broadhurst , arxiv : hep - th/9604128 .
a. erdelyi ( ed . ) , _ higher transcendental functions _ , vol.1 ( mcgraw - hill , new york , 1953 ) .
y. schrder , a. vuorinen , jhep * 0506 * ( 2005 ) 051 ; + k.g
. chetyrkin , j.h .
kuhn , p. mastrolia , c. sturm , eur .
j. * c40 * ( 2005 ) 361 ; + k.g .
chetyrkin , m. faisst , c. sturm , m. tentyukov , nucl .
* b742 * ( 2006 ) 208 ; + b.a .
kniehl , a.v .
kotikov , phys .
* b638 * ( 2006 ) 531 ; _ ibid . _
* 642 * ( 2006 ) 68 ; + b.a .
kniehl , a.v .
kotikov , a.i .
onishchenko , o.l .
veretin , phys .
* 97 * ( 2006 ) 042001 .
norlund , acta math .
* 94 * ( 1955 ) 289 ; p.o.m .
olsson , j. math .
* 7 * ( 1966 ) 702 ; + w. bhring , siam j. math . anal .
* 19 * ( 1988 ) 1249 .
skorokhodov , comput . math .
* 41 * ( 2001 ) 1718 ; _ ibid . _
* 44 * ( 2004 ) 1102 ; program .
software * 29 * ( 2003 ) 75 ; _ ibid . _ * 30 * ( 2004 ) 150 ; comput .
* 45 * ( 2005 ) 550 ; + a.i .
bogolubsky , s.l .
skorokhodov , program .
software * 32 * ( 2006 ) 145 .
davydychev , m.yu .
kalmykov , nucl .
* 89 * ( 2000 ) 283 ; nucl .
* b605 * ( 2001 ) 266 .
lappo - danilevsky , _ mmoires sur la thorie des systmes des quations diffrentielles linaires _ , ( chelsea , new york , 1953 ) ; + a.b .
goncharov , math .
* 4 * ( 1997 ) 617 ; math .
lett . * 5 * ( 1998 ) 497 ; + e. remiddi , j.a.m .
vermaseren , int .
j. mod .
phys . * a15 * ( 2000 ) 725 ; + j.m .
borwein , d.m .
bradley , d.j .
broadhurst , p. lisonek , trans .
am . math .
* 353 * ( 2001 ) 907 ; + t. gehrmann , e. remiddi , nucl .
* b601 * ( 2001 ) 248 ; + j. vollinga , s. weinzierl , comput .
commun .
* 167 * ( 2005 ) 177 .
s. moch , p. uwer , s. weinzierl , j. math .
* 43 * ( 2002 ) 3363 ; + s. weinzierl , j. math
. phys .
* 45 * ( 2004 ) 2656 .
kalmykov , b.f.l .
ward , s. yost , jhep * 0702 * ( 2007 ) 040 ; _ ibid .
_ * 0710 * ( 2007 ) 048 ; _ ibid . _ * 0711 * ( 2007 ) 009 ; + s.a .
yost , m.yu .
kalmykov , b.f.l .
ward , proc . of ichep 2008
[ arxiv:0808.2605 ] ; + m.yu .
kalmykov , b. kniehl , nucl .
phys . * b809 * ( 2009 ) 365 .
shu oi , [ arxiv:0810.1829 ( math.qa ) ] .
j. fleischer , a. v. kotikov , o.l .
veretin , nucl . phys . * b547 * ( 1999 ) 343 ; + s. weinzierl , comput .
phys . commun .
* 145 * ( 2002 ) 357 ; + m.yu .
kalmykov , o. veretin , phys .
b * 483 * ( 2000 ) 315 ; + f. jegerlehner , m.yu .
kalmykov , o. veretin , nucl .
* b658 * ( 2003 ) 49 ; + a.i .
davydychev , m.yu .
kalmykov , nucl .
* b699 * ( 2004 ) 3 ; + m.yu .
kalmykov , nucl .
* 135 * ( 2004 ) 280 ; + s. moch , p. uwer , comput .
commun .
* 174 * ( 2006 ) 759 ; + t. huber , d. matre , comput .
commun .
* 175 * ( 2006 ) 122 ; _ ibid . _
* 178 * ( 2008 ) 755 . | we describe the application of differential reduction algorithms for feynman diagram calculation .
we illustrate the procedure in the context of the generalized hypergeometric functions @xmath0 , and give an example for a type of @xmath1-loop bubble diagram . |
there was a time when there appeared to be very little interconnection between the presence and evolution of dust grains on the one hand , and galaxy morphology on the other .
this has now changed so dramatically , that an entire international conference was recently devoted to the _ interrelation _ between these two apparently discrepant disciplines ( see block @xcite ) .
allen ( @xcite ) summed it up thus : _ `` we re now looking at a transition to a possible change in the way we look at galaxies .
sometimes ... we see disks that have a spiral structure that we could nt have dreamt existed from looking at the optical picture ... we ve got a possibility here of applying the morphology to a physical framework , perhaps in a way that none of us could have dreamt of before we had the capability of sweeping the dust away from the galaxy in a figurative sense''_. optically thick dusty domains in galactic disks can completely obscure underlying stellar structures .
the presence of dust and the morphology of a galaxy are inextricably intertwined : indeed , the morphology of a galaxy can completely change once the disks of galaxies are dust penetrated ( eg .
block and wainscoat @xcite ; block , elmegreen and wainscoat @xcite ) .
the distribution of dust grains in the bulges and disks of galaxies may be very widespread ( block et al .
@xcite ) , even extending to the outer parts of galaxies ( lequeux and gulin @xcite ) .
the tracing of dust grains in nature can be very elusive : high levels of dust extinction do _ not _ necessarily imply that the effects of dust attenuation ( ie .
observed reductions in surface brightness profiles ) will also be large , because scattering by dust grains may fill in at least part of the lost surface brightness .
the effective albedo of dust grains in the near - infrared can actually be higher than in the optical ( for a determination of the near - infrared albedo of dust grains in m51 , see fig . 4 in block @xcite ) .
furthermore , large amounts of dust do not necessarily imply red v - k ( k : 2.2@xmath1 ) colours .
the dust column density on the far side of an inclined spiral with relatively _ blue _ v - k colours can be just as high as on the near side , where the v - k colour distribution can be much redder ( elmegreen and block @xcite ) .
the classification of galaxies has traditionally been inferred from photographs / ccd imaging shortward of the 1 micron window , where stellar population ii disks are not yet dust penetrated .
the nicmos and other near - infrared camera arrays offer unparralleled opportunities for deconvolving the population i and ii morphologies , because the opacity at k be it due to absorption or scattering is always low .
the extinction ( absorption+scattering ) optical depth at k is only 10% of that in the v - band ( martin and whittet @xcite ) . from a dynamical viewpoint
, the disk of a spiral galaxy can be separated into two distinct components : the _ gas dominated _ population i disk , and the _ star dominated _ population ii disk .
the former component contains features of spiral structure ( ob associations , hii regions , and cold interstellar hi gas ) , which are naturally fast evolving ; dynamically , it is very active and responsive , because , being characterized by small random motions ( a `` cool disk '' ) , it fuels jeans instability . in contrast , the population ii disk , which is dynamically `` warmer '' because of the larger epicycles , contains the old stellar population betraying the underlying mass distribution - it is the ` backbone ' of the galaxy ( lin @xcite ) .
one might expect even in the absence of appreciable optical depths for the two morphologies to be very different , since the near - infrared light comes predominantly from giant and supergiant stars ( rix and rieke @xcite , frogel et al .
@xcite ) .
it is important to stress that from this physical ( dynamical ) point of view , one therefore requires _ two _ classification schemes
one for the population i disk , and a separate one for the population ii disk .
a near - infrared classification scheme can never _ replace _ an optical one , and vice - versa , because the _ current _ distribution of old stars strongly affects the _ current _ distribution of gas in the population i disk .
the dynamic interplay between the two components via a _
mechanism is crucial , and has been studied extensively ( eg .
bertin and lin @xcite , who term this mechanism a dynamical thermostat ) .
a central aspect here is the _ likely coupling _ of the population i disk with that of the population ii disk via the feedback mechanism . to quote pfenniger et al .
@xcite , _ `` the interesting aspect of this coupling is that the systematic global properties of galaxies are then no longer necessarily determined by the initial conditions of collapse . '' _ we find a duality in spiral structure . a typical turbulent speed associated with the cold gas population
i component is @xmath2 6 km s@xmath7 ; in contrast , the velocity dispersion for old stars would typically be six times larger .
while a gas cloud may be constrained to move in a thin annulus @xmath8 , an old star may wander in an annulus 2 kpc thick ( bertin and lin @xcite ) . to derive a _
coherent _ physical framework for the excitation of spiral structure in galaxies , one must consider the co - existence of the _ two _ dynamical components .
_ there is a fundamental limit in predicting what evolved stellar disks might look like_. the greater the degree of decoupling , the greater is the uncertainty .
the fact that a spiral might be flocculent in the optical is very important , but it is equally important to know whether or not there is a decoupling with a grand design old stellar disk .
no prediction on that issue can , a priori , be made ( block , elmegreen and wainscoat @xcite ) . the theoretical framework to explain the co - existence of completely different morphologies within the same galaxy when it is studied optically and in the near - infrared is beautifully described by bertin and lin ( @xcite ) , following the early pioneering work of lindblad ( @xcite ) . a global mode ( bertin et al .
@xcite , @xcite ) can be imagined as being composed of spiral wavetrains propagating radially in opposite directions , much like a standing wave .
thus a feedback of wavetrains is required from the center .
the return of wavetrains back to the corotation circle is guaranteed by refraction , either by the bulge or because the inner disk is dynamically warmer . in
the _ stellar _ disk , such a feedback can be interrupted by inner lindblad resonance ( ilr ) , which is a location where the stars meet the slower rotating density wave crests in resonance with their epicyclic frequency ( mark @xcite ; lynden - bell and kalnajs @xcite ) . in the _ gaseous _ disk ,
the related resonant absorption is only partial , so that some feedback is guaranteed .
once the above described _ wavecycle _ is set up ( in the absence of a cutoff by ilr ) , a self excited global mode can be generated .
the tightness of the arms in the modal theory comes from the mass distribution and rate of shear .
galaxies with more mass concentration , i.e. , higher overall densities ( including dark matter ) and higher shear , should have more tightly wound arms .
if the disk is very light low @xmath9 where @xmath9 is the disk density the mode can be very tight , and one is in the domain of small epicycles ( formally , the stability parameter q = @xmath10 being close to unity , the value of @xmath11 must also be small . here
@xmath11 is the radial velocity dispersion and @xmath12 is the epicyclic frequency ) .
if one increases the mass of the disk one finds a trend toward more open structures , but soon one runs the risk of a disk that is too heavy and a bar mode should manifest itself . in those spirals with more open spiral arms at k but with no sign of a prominent bar , bertin ( @xcite )
anticipates that the galaxy should be _
gas rich_. abundant gas can shock , dissipate and make some violently unstable _ open _ modes ( see frame d of fig . 4.5 in the book by bertin and lin @xcite ) .
this is one important reason why bertin and lin place ` gas content ' in their theoretical framework for classification as the governing parameter for the trend from ` a ' to ` c ' galaxies ( as indicated in fig . 1 of block et al .
@xcite ) .
the redistribution of angular momentum by large - scale spiral torques will be stronger for stellar arms which are more open ; some authors ( eg . pfenniger et al .
@xcite ) have postulated that such a redistribution may lead to rapid changes in the disk and even modify the properties of the rotation curve .
this is the concept of secular evolution of a galaxy , from an open to a more tightly wound morphology , within one hubble time .
many galaxies show the presence of a significant m=1 component in the near - infrared ( often in the form of a lopsidedness of the spiral ) .
the linear modal theory predicts that m=1 modes should generally be dominated by m=2 modes when available , since the latter are more efficient in transporting angular momentum outwards .
however , modes greater than m=2 are generally suppressed in the stellar disk by ilr ( but see the discussion in section 7 ) . while the disk mass participating in the mode is crucial , the gas - content of the galaxy is important : gas - rich spirals can generate modes greater than m=2 .
it had earlier been predicted ( block et al .
@xcite , bertin and lin @xcite ) that infrared images should show _ an ubiquity of global one and two armed structures in the underlying stellar disk _ and we believe that any classification of dust penetrated disks at k should indicate this number .
in contrast , the dynamics of the cold population i gaseous disk , characterized by different scalelengths , velocity dispersion , thickness , and behaviour at the relevant lindblad resonances , explains why spiral galaxies are optically so often overwhelmed by higher m modes and other more irregular fast evolving features , supported by the cold interstellar gas ( bertin @xcite , @xcite ) .
our sample in table 1 was selected to cover as wide a range of optical hubble types as possible from early type to irregular .
it spans a range in blue absolute magnitude from @xmath13 to @xmath14 , a range in linear diameter from 14 kpc to 97 kpc ( the galaxies at the lower and upper limits of these ranges are ngc 1637 and ngc 309 ) and a wide range in linear arm width . from a sample of several hundred galaxies which one of us ( dlb )
has studied on a uniform linear scale , ngc 309 has some of the linearly widest spiral arms known ( see figure 1 in block ( @xcite ) ) . the near - infrared imaging of many of the spirals was secured at mauna kea and in chile and fully described in block et al . ( @xcite ) . at mauna kea ,
a rockwell hgcdte 256 @xmath15 256 camera array was used , sensitive from 1 to 2.5@xmath1 . at the f:10 focus of the university of hawaii s 2.2 m telescope , with 1:1 re imaging optics
, the scale is 0.37@xmath16/pixel , giving a field of 95@xmath16 @xmath15 95@xmath16 . at the european southern observatory at la silla
, we used the irac2 camera mounted on the eso 2.2 m telescope , in mode c ( 0.49@xmath16/pixel , giving a 2@xmath17 field of view ) .
both at la silla and mauna kea , the k@xmath0 filter ( 2.1@xmath1 ) was always selected .
this filter is similar to k , but shifted 0.1 @xmath1 towards the blue , to reduce the thermal background ( wainscoat and cowie @xcite ) .
the k@xmath0 images of additional galaxies were very kindly made available to us by dr p. grosbl ( ngc 3223 , ngc 5085 , ngc 5247 , ngc 5861 and ngc 7083 ) , dr m. verheijen ( ngc 3893 , ngc 3938 , ngc 3992 and ngc 4051 ) and dr m. thornley ( ngc 3521 and ngc 5055 ) .
the 2-d fast fourier decomposition of all the k@xmath0 images in this study , employed a program developed by puerari ( schrder et al .
@xcite ) . in the decomposition ,
logarithmic spirals of the form @xmath18 are employed .
the amplitude of each fourier component is given by ( schrder et al .
@xcite ) @xmath19 where @xmath20 and @xmath21 are polar coordinates , @xmath22 is the intensity at position @xmath23 , @xmath24 represents the number of arms or modes , and @xmath25 is the variable associated with the pitch angle @xmath26 , defined by @xmath27 .
our analytic fourier spectra corroborate earlier observational indications ( block et al .
@xcite ) that there is indeed an ubiquity of m=1 and m=2 modes , and that three principal archetypes for the evolved stellar disk of such galaxies may be proposed , based on the the pitch angle of the arms at k@xmath0 ( see figure 1 ) .
figure 1 is generated from the actual deprojected images at k@xmath0 of our galaxies where the pitch angles have robustly been determined from the fourier spectra .
it is apparent that a galaxy with a winding angle at k@xmath0 of 18 degrees looks remarkably similar to one with a winding angle of 30 degrees ( see figure 1 ) , and we classify those into one distinct bin .
we choose to designate our three dust penetrated ( dp ) bins as @xmath4 , @xmath5 and @xmath6 , depending on the pitch angle of the dominant m - spiral at k@xmath0 .
those l = lopsided galaxies ( where m=1 is a dominant mode ) are designated l@xmath4 , l@xmath5 and l@xmath6 according to the dust penetrated pitch angle ; e = evensided galaxies ( where m=2 is the dominant fourier mode ) are classified into classes e@xmath4 , e@xmath5 and e@xmath6 , according to our three principal dust penetrated archetypes ( see column 2 of table 1 ) .
it is proposed that higher order modes ( which may exist under special circumstances : see section 7 of this paper ) be classified as h3 ( for m=3 ) and h4 ( for m=4 ) , followed by a class description @xmath4 , @xmath5 or @xmath6 appropriate to the pitch angle of the stellar arms .
the terminology dp highlights the fact that at k@xmath0 , one is looking at dust penetrated images .
it is dust in the gaseous population i disk which so often obscures the underlying population ii disk .
it is because of dust that a galaxy can masquerade as late type c ( for the optical morphology ) but as early type a in the dust penetrated population ii regime ( block and wainscoat @xcite ) .
it is dust which obscures the regular two - armed spiral structure found in many optically flocculent galaxies ( eg . grosbl and patsis @xcite , @xcite , thornley @xcite ) .
optical images , combined with near - infrared ones , together with the use of radiative transfer models highlight the fact that gaseous population i disks often have optical depths @xmath28 and more ( block et al .
@xcite , elmegreen and block @xcite ) which is why there can be no _ prediction _ of what an evolved stellar disk of a optically flocculent galaxy , for example , might look like ( block , elmegreen and wainscoat @xcite ) .
we have avoided using i , ii or iii for our dust penetrated classes and have used @xmath4 , @xmath5 and @xmath6 instead , to prevent confusion with the optical luminosity classification scheme of galaxies ( van den bergh @xcite ) .
while there may appear to be a resemblance of our groups to the hubble bins in terms of the openness or pitch angle of the arms , there is no similarity whatever .
it is possible for a late type sc galaxy ( eg .
ngc 5861 below ) to have a _ stellar disk with tightly wound arms _ with a pitch angle at k@xmath0 of 13@xmath3 , which demarcates it to our tightly wound group @xmath4 class .
conversely , it is possible for two galaxies on opposite ends of the hubble sequence ( a , c ) to _ both _ have evolved disks of dust penetrated group @xmath5 : ngc 718 and ngc 2997 are examples which both belong to group @xmath5 , but have hubble classifications of a and c respectively .
( the near - infrared pitch angles of ngc 718 and ngc 2997 differ by only 1@xmath3 see table 1 . )
it is also possible for an optical irregular to have smooth stellar spiral arms with a well - defined pitch angle at k@xmath0 and to belong to one of these groups ( eg . the companion to m51 , as discussed below ) . -2.0truecm cccccc galaxy & dp class & morph .
type & _ p_@xmath29 & range of fit + _ group @xmath4 _ + ngc 2841 & e@xmath4 & sb & 9@xmath3 & 70@xmath16 , 165@xmath16 + ngc 3223 & e@xmath4 & sb & 9@xmath3 & 15@xmath16 , 48@xmath16 + ngc 3521 & e@xmath4 & sbc & 14@xmath3 & 12@xmath16 , 157@xmath16 + ngc 3992 & e@xmath4 & sbbc & 11@xmath3 & 61@xmath16 , 184@xmath16 + ngc 4622 & e@xmath4/l@xmath4 & sb & 8@xmath3 & 11@xmath16 , 50@xmath16 +
ngc 5055 & e@xmath4 & sbc & 13@xmath3 & 18@xmath16 , 150@xmath16 + ngc 5861 & e@xmath4 & sc & 13@xmath3 & 17@xmath16 , 50@xmath16 + _ group @xmath5 _ + ngc 309 & e@xmath5 & sc & 18@xmath3 & 11@xmath16 , 44@xmath16 + ngc 718 & e@xmath5 & sa & 24@xmath3 & 11@xmath16 , 34@xmath16 + ngc 1637 & e@xmath5/l@xmath6 & sc & 30@xmath3 & 13@xmath16 , 45@xmath16 + ngc 2997 & e@xmath5 & sc & 25@xmath3 & 34@xmath16 , 122@xmath16 + ngc 3893 & e@xmath5 & sc & 21@xmath3 & 12@xmath16 , 88@xmath16 + ngc 3938 & e@xmath5 & sc & 22@xmath3 & 12@xmath16 , 168@xmath16 + ngc 4736 & e@xmath5 & sab & 28@xmath3 & 37@xmath16 , 56@xmath16 + ngc 5247 & e@xmath5 & sc & 30@xmath3 & 30@xmath16 , 90@xmath16 + _ group @xmath6 _ + ngc 4051 & e@xmath6 & sbbc & 49@xmath3 & 16@xmath16 , 180@xmath16 + ngc 5085 & e@xmath6 & sc & 39@xmath3 & 25@xmath16 , 50@xmath16 + ngc 5195 & e@xmath6 & irr & 49@xmath3 & 26@xmath16 , 52@xmath16 + ngc 7083 & e@xmath6 & sb & 36@xmath3 & 15@xmath16 , 40@xmath16 + the range of radii over which the fit was applied are all listed in table 1 .
these were selected to exclude the bulge ( where there is no information on the arms ) and extend to the outer limits of the arms in our images .
it is important to discuss the error bars on the determination of our pitch angles using fourier transform techniques .
we use a fast fourier transform , where the incremental step - size in the spectrum is @xmath30 .
we calculate the spectrum from @xmath31 and select a step - size of @xmath30=0.25 .
the error bars decrease with the degree of tightness of the arm pattern . in ngc 5195 for example with very open stellar arms the spectrum peak occurs at @xmath25=1.75 ( see figure 7 ) .
we therefore compute a pitch angle in table 1 of @xmath26=atan(@xmath24=2/@xmath25=1.75)=48.8@xmath3 .
if the peak was to occur one step - size later , at @xmath25=2.0 , the computed pitch angle would have been @xmath26=45.0@xmath3 .
if the fourier spectrum was to have peaked at @xmath25=1.5 , one step - size earlier , then @xmath26=53.1@xmath3 .
so , for this galaxy , our pitch angle at k@xmath0 of @xmath26=48.8@xmath3 has an error bar of ( @xmath32 ) degrees . on the other hand ,
let us consider ngc 3992 , with its tightly wound arms .
the fourier spectrum peaks at @xmath25=10.25 ( see figure 3 ) , and so the derived pitch angle in table 1 is @xmath26=11.0@xmath3 .
if the peak were to have occurred at @xmath25=10.0 , the pitch angle would have been @xmath26=11.3@xmath3 .
if the spectrum peaked one step - size later at @xmath25=10.5 , the derived pitch angle would be @xmath26=10.8@xmath3 . in this case , the error bar for our listed 11@xmath3 tightly wound arm structure in ngc 3992 is ( @xmath33 ) degrees . in the tightly wound scenario ,
it is perhaps not too meaningful to quote error bars on the pitch angle of only @xmath34 .
spirals with @xmath26=11.0@xmath3 or @xmath26=11.2@xmath3 will only be ` different ' after a number of winding angles or turns , and we know it is difficult to find arms at k@xmath0 which actually wind round more than 360@xmath3 , although such galaxies do exist ( eg .
fig 6b in block et al .
@xcite ) .
pitch angles can either be determined from the peaks in the fourier spectra ( as above ) , or from the slopes in @xmath35 plots , where logarithmic spirals appear as straight lines . in both cases , careful deprojections to face - on are critical .
our experience with estimating pitch angles from @xmath35 plots is that eyeball identification of the straight lines ( arms ) and of their resulting slopes can be very difficult , especially if the arm / interarm contrast in the near - infrared image is not high . on the other hand ,
peaks in fourier spectra are unambiguous . as other authors have advocated in the past ( eg .
considre and athanassoula @xcite ; garca - gmez and athanassoula @xcite ) , the fourier transform is the most powerful method to find periodicity in a distribution , whether it be one dimensional ( such as radial stellar pulsations ) or bi - dimensional ( as for the spiral arms in an image ) . that there are galaxies with very open arm morphologies , with pitch angles as large as @xmath2 40 - 45@xmath3 , is corroborated by these earlier studies .
for example , considre and athanassoula ( @xcite ) analysed blue ( b ) images of 16 galaxies , and found pitch angles as large as 37@xmath3 .
garca - gmez and athanassoula ( @xcite ) analysed the distribution of hii regions of 44 galaxies and also found some very open morphologies .
for the dominant m=2 component , garca - gmez and athanassoula ( @xcite ) determine pitch angles of 43@xmath3 for ngc 5194 and 55@xmath3 for both ngc 3627 and ngc 7741 .
many other examples are provided in their table .
these large pitch angles were only for the main m=2 component ; garca - gmez and athanassoula ( @xcite ) report even larger pitch angles for other fourier components which they see in their spectra .
we have also computed the symmetry number proposed by conselice ( @xcite ) for each of our b and k@xmath0 images : @xmath36 the number is determined by rotating every image through 180@xmath3 about the centre , subtracting that from the original image and normalising .
we did this to illustrate that highly symmetric evolved disks exist for all three of our dust penetrated groups ( see figure 2 ) .
a value of @xmath37=0 corresponds to a completely symmetric galaxy at wavelength @xmath38 ; for a completely asymmetric galaxy , the ratio is unity .
in contrast to the work of conselice , where a homogeneous sample was used , our sample here is drawn from a non - homogeneous one ( different instruments and different observers ) .
care must be exercised when working with images from different instruments : the potential problem is that of noise especially in the near - infrared , where disk features might be of a low contrast and where the sky is bright .
a greater dispersion can be expected , but the quantitative trend observed in figure 2 confirms our earlier qualitative findings in block et al .
( @xcite ) , that disks at k@xmath0 are much more regular and symmetric than seen at shorter wavelengths , such as b. the conselice number does not increase with dust penetrated class .
for example , its value for ngc 5195 is only 0.031 at k@xmath0 , confirming the visual impression of a very symmetric evolved disk ( see fig .
5b in block et el .
while the conselice method might be used to confirm the presence of symmetric disks in all three of our dust penetrated archetypes , it is the pitch angle of the arms at k@xmath0 and not the conselice number which determines to which dust penetrated bin a galaxy should be assigned .
our fourier spectra of these galaxies , with tightly wound stellar arms at k@xmath0 , are illustrated in figures 3 and 4 .
ngc 3223 is an optically multi - armed specimen of hubble type sb and van den bergh ( @xcite ) luminosity class i - ii . in the near - infrared
, however , this galaxy shows two bright , _ grand design _ stellar density waves ( sdws ) ( see grosbl and patsis @xcite , @xcite ) .
that multi - arm galaxies such as ngc 3223 may present long , continuous arms with well - defined pitch angles at k@xmath0 is not uncommon : for the famous optically flocculent prototype ngc 2841 ( elmegreen arm class 3 = ` fragmented arms uniformly distributed around the galactic center ' ; see elmegreen and elmegreen @xcite ) , a remarkable system of four long and regular spirals at k@xmath0 is reported ( see block , elmegreen and wainscoat @xcite and block @xcite ) .
independent studies by grosbl and patsis ( @xcite ) and thornley ( @xcite ) show that ngc 2841 is no exception : many optically flocculent galaxies reveal a regular two - armed spiral structure in the near - infrared .
these include ngc 5055 and ngc 3521 ( thornley @xcite ) .
both ngc 5055 and ngc 3521 are of elmegreen arm class 3 , but the spiral arms of these two optically flocculent galaxies with ` grand design ' m=2 near - infrared counterparts have a pitch angle at k@xmath0 of 13@xmath3 and 14@xmath3 respectively , and both galaxies are assigned to our dust penetrated @xmath4 bin . if , in the midst of stellar density waves ( sdws ) , sdw pressures are less than the average turbulent gas pressure , the competition between stars and gas is won by gas : the galaxy would present an _ optically _ fragmented , flocculent appearance ( block @xcite ) . the star formation history , principally triggered from previous bursts of star formation ( rather than sdw pressures ) , would be stochastic / random . high total gas column densities ( and therefore relatively large optical depths at b or v ) would be predicted .
indeed , the v optical depth in ngc 3223 is @xmath2 2 at certain radii ( see block @xcite ) and it is dust which hides the m=2 near - infrared spirals .
we next turn our attention to ngc 5861 and ngc 3992 .
both are described by sandage and bedke ( @xcite ) as having spiral patterns which rank amongst ` the most regular in the sky ' .
they are exquisite grand design galaxies which pitch angles at k@xmath0 of 13@xmath3 and 11@xmath3 respectively , and both galaxies are assigned to our dust penetrated @xmath4 bin .
it is therefore particularly important to note that _ the evolved stellar disks of some of the most exquisite grand design galaxies may belong to the same dust penetrated class as those of the best known optically flocculent prototypes_. indeed , ngc 5861 is of elmegreen class 12 ( defined in elmegreen and elmegreen @xcite as having ` two long symmetric arms dominating the optical disk ' ) while ngc 5055 and ngc 3521 are of elmegreen arm class 3 almost as optically flocculent as one can find .
ngc 4622 is an sb galaxy which displays dual spiral structure ( byrd et al .
@xcite ) . both the outer pair and the inner arm reveal amplitude modulation .
the inner arm is probably the first confirmed evidence for a leading spiral arm .
based on the sequential structure of the bvi photometry of the leading arm , beautiful evidence has been given ( buta , crocker and byrd @xcite ) in support of the pattern being a counter
rotating density wave . a method to determine
the arm character ( trailing or leading ) in spiral galaxies using the fourier analysis of azimuthal profiles is presented by puerari and dottori ( @xcite ) .
our fourier decomposition of the k@xmath0 image of ngc 4622 confirms a strong m=1 mode which is identified with the inner ( leading ) arm .
the pitch angle at k@xmath0 for that m=1 arm is only 4@xmath3 . in block
( @xcite ) , we argued that the leading arm of ngc 4622 is truly exceptional and requires external driving ( much like a _ damped _ mode of a church bell ) , because the torques associated with leading structures are unfavorable to self excitation .
the rarity of leading arms among grand design galaxies may be an indication that tidal driving is very seldom efficient .
the pair of magnificent m=2 arms winding in the opposite direction have a pitch angle at k@xmath0 of 8@xmath3 , and the galaxy is assigned to our dust penetrated @xmath4 bin .
we had earlier published a plot of log(radius ) versus azimuthal angle for our k@xmath0 image of ngc 2841 , where the long arms appear as straight lines ( see colour plate 2b in block @xcite ) .
the pitch angle of the spiral arms can easily be read off from the slopes of the tilted features ; they are found to be approximately 9@xmath3 .
the tightly wound arms at k@xmath0 of this prototypically flocculent specimen at optical wavelengths suggests that we place it in our dust penetrated @xmath4 classification .
although all the galaxies in figures 3 and 4 show fourier spectra where the _ most dominant _ mode is m=2 so that they all belong to the two - armed , evensided e@xmath4 group , there is nevertheless an ubiquity of the lower - order m=1 mode as well .
the only exception where m=1 is as dominant as m=2 is ngc 4622 , which we have already noted has one presumably leading arm ( m=1 ) and a pair of outer ( presumably trailing ) m=2 arms .
note , too , the presence of higher order [ m=4 ] modes , but with considerably lower amplitude than the m=2 mode , in galaxies such as ngc 3223 , ngc 3521 and ngc 3992 .
the fourier spectra of these galaxies are presented in figures 5 and 6 .
galaxies with a very large range in optical morphology from sa to sc are to be found in this dust penetrated class .
they also span a very wide range of elmegreen arm class in the optical ; from extreme grand design arm class 12 to flocculent arm class 3 .
we firstly turn our attention to the sc galaxy ngc 309 , of luminosity classification i ( van den bergh @xcite ) .
we have already noted that this galaxy has some of the linearly widest population i arms known ( block @xcite ) .
optically , ngc 309 presents a multi armed morphology ( see block and wainscoat @xcite ) and the elmegreen arm class is 9 ( ` two symmetric inner arms ; multiple long and continuous outer arms ' as defined by elmegreen and elmegreen @xcite ) . however , at k@xmath0 it looks like an sba galaxy , betraying ( 1 ) a prominent bar and ( 2 ) two symmetrical arms only , with winding angle just over @xmath39 and a pitch angle of @xmath2 18@xmath3 .
the surface brightness enhancements of the m=2 stellar components at k@xmath0 are approximately 60 % above that of the underlying disk .
normal ( sci ) and barred ( sba ) large scale structures co exist within the same galaxy .
next , we focus our attention on another galaxy whose evolved disk is very similar to that of ngc 309 : ngc 718 .
ngc 718 is optically classified as sai , but its evolved disk belongs to the same group at that for ngc 309 .
the pitch angle of its two spiral arms at k@xmath0 is 24@xmath3 and we assign it to our dust penetrated @xmath5 bin .
another galaxy which belongs to the same dust penetrated class as ngc 309 , is the early type spiral ngc 4736 .
ngc 4736 is classified as type rsab(s ) and is of elmegreen optical arm class 3 .
the galaxy is famous for its inner , ` knotty ring ' of hii regions ( see figure 8 in block et al .
@xcite ) . our k@xmath0 image reveals a pair of logarithmic spiral arms , with a pitch angle at k@xmath0 of 28@xmath3 .
the galaxy is assigned to the dust penetrated @xmath5 bin .
also in this dust penetrated class is the grand design sc(s)i spiral ngc 2997 ( sandage and tammann @xcite ) .
optically , the galaxy possesses two dominant spiral arms , one of which has a prominent western branching into a third arm ( see block et al .
ngc 2997 is of elmegreen arm class 9 . at k@xmath0
the galaxy reveals a small oval distortion and a distinct m=1 component .
the most natural interpretation is that there are two dominant modes , m=1 and m=2 , with negative interference on the northern side .
the pitch angle of the stellar arms at k@xmath0 are determined to be @xmath2 25@xmath3 and we assign ngc 2997 to our dust penetrated @xmath5 bin . not only are sa to sc galaxies all to be found in this dust penetrated class , but the dust penetrated @xmath5 bin again spans almost the entire range of optical elmegreen arm classes : from extreme grand design arm class 12 ( eg
. ngc 718 ) to flocculent arm class 3 ( eg .
ngc 4736 ) . from our fourier spectra for these class @xmath5 population ii disks , we note that although the dominant mode is m=2 , there is an ubiquity of m=1 modes as well , but with lower amplitude .
the m=1 component is nearly as striking as the m=2 component in some of the galaxies such as ngc 3893 and ngc 3938 ( figure 6 ) while m=1 has a much lower amplitude than m=2 in others , such as ngc 5247 .
higher order modes ( also with lower amplitude than m=2 ) are also to be found : ngc 2997 shows a distinct m=3 component at k@xmath0 , while ngc 309 and ngc 1637 reveal significant m=4 modes in the near - infrared ( figure 5 ) .
the fourier spectra for these galaxies are presented in figure 7 .
we firstly turn our attention to ngc 5195 , the companion of ngc 5194 = m51 .
optically , ngc 5195 has been classified as ip - ep ( morgan @xcite ) , irr ii ( sandage @xcite ) , sba(r ) ( spinrad and harlan @xcite ) and sb0/a(r ) ( thronson , rubin and ksir @xcite ) . our k@xmath0 image , radically different from the optical , shows _ no evidence for tidal disruption _ of its old stellar population disk by m51 ( figure 5b in block et al .
@xcite ) . the striking symmetry in the ngc 5195 population ii disk ( see the k@xmath0 image in fig .
5b in block et al .
@xcite and confirmed by the evensided fourier mode m=2 in figure 7 of this paper ) is confirmed by a low conselice number at k@xmath0 .
the bisymmetric structures look more like intrinsic global modes where the m=2 spirals have a pitch angle at k@xmath0 of 49@xmath3 . in view of the exceptional symmetry at k@xmath0 ,
ngc 5195 is likely to have _ just disturbed _
m51 s spiral arms ( oort @xcite ) .
the tidal interaction here is far less dramatic than usually thought , based on the more prominent features of the lighter population i disks . as suggested by block et al .
( @xcite ) , the prominent isophotal _ twisting _ within the bar
itself may be the quantitative signature of the tidal field of m51 on the intrinsic bar structure of ngc 5195 .
isophotal twisting of the central bar regions is not uncommon ( shaw et al .
@xcite ) and might be simply due to the incipient spiral structure , commencing with a sharp turn at the bar tip , as often found for intrinsic modes ( see the discussion of ngc 1300 in bertin @xcite or fig .
10 of benedict et al .
@xcite , for ngc 4314 ) .
members of this dust penetrated @xmath6 group include an early type spiral ( ngc 7083 ) , a late type spiral ( ngc 5085 ) as well as an optical irregular .
ngc 5085 is of elmegreen arm class 2 ( ` fragmented spiral arm pieces with no regular pattern ' ) .
also included is ngc 4051 , of elmegreen arm class 5 ( ` two symmetric short arms in the inner regions ; irregular outer arms ' ) .
the fourier spectra for dust penetrated group @xmath6 galaxies with open stellar arms in figure 7 show that the dominant fourier mode is m=2 .
however , modes of higher order do exist ( eg . m=4 for ngc 5085 ) .
ngc 1637 , for example , draws attention because of its exceptional lopsidedness , with a somewhat squarish shape , although there is a regular bisymmetric spiral close to the nucleus .
it is classified as sc ( sandage 1961 ) , but shows a marked lopsidedness ( an m=1 feature ) in the old population ii disk ( see figure 9b in block et al .
@xcite ) .
if one wishes to draw attention to lopsidedness in galaxies displaying a prominent m=1 feature , we propose that the galaxy still be placed in the @xmath4 , @xmath5 or @xmath6 bin , according to the pitch angle of the m=1 component , but that a prefix l clearly be used to designate the lopsidedness in the evolved disk .
the m=1 feature itself may be prominent , but not dominant , as in ngc 1637 . we have carefully examined the m=1 feature at k@xmath0 of ngc 1637 in a @xmath40 diagram and find a straight line , indicating that the fit of a logarithmic spiral to that m=1 arm is excellent . from the slope , we find the pitch angle of the m=1 arm to be @xmath2 35@xmath3 .
we have analysed the inner bisymmetric spiral by setting an appropriate annulus to emcompass the inner spiral but exclude the outer m=1 arm .
the spiral shows up clearly at k@xmath0 in the spectra , and the pitch angle of the dominant m=2 component ( figure 5 ) is 29@xmath3 . for the regular m=2 bisymmetric spiral ,
a dust penetrated e@xmath5 would be appropriate .
binning according to the asymmetric / lopsided mode of m=1 here , the m=1 spiral belongs to the dust penetrated @xmath6 bin and would be designated l@xmath6 .
the pitch angle of the lopsided m=1 arm is 6@xmath3 greater than for the m=2 mode . in figure 1 , ngc 1637 is included by plotting an m=1 logarithmic spiral with a pitch angle of 35@xmath3 .
no companion for ngc 1637 is found on the palomar observatory sky survey within one degree .
such extreme lopsidedness could be a general feature in some isolated galaxies , for which the persistence of such m=1 asymmetries ( baldwin , lynden - bell and sancisi @xcite ) should be explained . as exemplified by these authors , lopsided prototype galaxies such as ngc 891 and 5457 ,
are indeed lopsided both in the light and in the neutral hi gas . within the modal theory
, one may argue that lopsidedness and one armed structures are all non - linear m=1 modes . in the linear theory
they may be the m=1 equivalent of what is more commonly noted as barred and normal two armed spiral modes ( the former occurring in heavy , fully self gravitating disks and the latter occurring instead in light disks ) .
if this is so , such an extreme lopsidedness in the stellar disk would be more natural if the disk is heavy .
indeed , a _ prominent bar _ is observed in the old stellar disk of ngc 1637 .
while there is an ubiquity of m=1 and m=2 modes in our sample ( figures 37 ) , higher order modes in some instances do exist , as we have already pointed out .
however , in no galaxies in our sample are the higher order modes the _ dominant _ ones . in the k - band study of 45 spirals by seigar and james ( @xcite ) ,
the most common dominant modes were also found to be the lopsided l ( m=1 ) and evensided e ( m=2 ) each of which was found to arise in their sample in about 1/3 of the galaxies .
the sample of seigar and james ( @xcite ) showed a great dearth of spirals with dominant m=3 ( only 4 out of the 45 cases ) , but nevertheless such galaxies do exist and one must be able to classify them in a dust penetrated regime .
seigar and james ( @xcite ) did find that approximately 25% of their total showed a dominant m=4 mode in the k - band .
we would suggest the terminology h3 and h4 for the third and fourth harmonics , to assist with easy visualisation of their evolved disk morphologies .
can modes greater than m=2 develop within the modal theory of spiral structure ?
already in the paper by block et al .
( @xcite ) and by bertin ( @xcite ) it was hinted that : firstly , in a _ gas - rich _ system some higher - m modes should develop , and that might also induce some response in the stellar disk , for the stronger cases .
secondly , non - linearly modes may couple and again give rise to higher - m structures 2 + 1 = 3 and 2 + 2 = 4 ( block et al .
@xcite and bertin , private communication ) .
a variation on this theme is the following : if one takes a strong bisymmetric bar and fourier - analyze the deprojected image , one may indeed find m=4 and even m=6 to be prominent , even when one does not see actually four or six arms .
if the bar is strong enough to drive the disk , an m=4 response in the outer disk may show up .
it is interesting to note that almost all of the m=4 cases cited by seigar and james ( @xcite ) are classified as barred in the optical .
thirdly : another point to consider for the generation of higher - m modes is the underlying rotation curve .
if the shear is _ very _ mild ( some sc s have a large part of their disks in almost solid body rotation ) , then the ilr is less effective ( bertin , private communication ) and m=3 and even m=4 modes might develop .
we suggest that these h3 and h4 galaxies , too , should be binned in the dust penetrated regime , according to the pitch angle of their stellar arms .
all of the h3 and h4 galaxies in seigar and james ( @xcite ) belong to our dust penetrated @xmath4 class , so that a galaxy such as ugc 2303 ( m=3 , with a pitch angle at k of 4@xmath3 ) would be classified here as h3@xmath4 , while a spiral such as ic 357 ( m=4 , with a k pitch angle of 8@xmath3 ) is classified here as h4@xmath4 .
in figure 8 , we present the inverse fourier transforms of six of the galaxies in table 1 , two from each @xmath4 , @xmath5 and @xmath6 dust penetrated classes . after having deprojected the k@xmath0 images and identifying the dominant modes , we next calculated the inverse fourier transform , as follows : we define the variable @xmath41 . then @xmath42 where @xmath43 and @xmath44
@xmath45 is a high frequency filter used by puerari and dottori ( @xcite ) . for the spiral with @xmath46 it has the form @xmath47\ ] ] where @xmath48 is the value of @xmath25
for which the amplitude of the fourier coefficients for a given @xmath24 is maximum .
this filter is also used to smooth the @xmath49 spectra at the interval ends ( see puerari and dottori @xcite ) .
the contour overlays of the inverse fourier transforms in figure 8 indicate the excellent fit of our pitch angles , robustly derived from the fourier spectra . in the top panel of figure 8 ,
two galaxies with tightly wound pitch angles at k@xmath0 , belonging to the @xmath4 class , are presented .
the galaxies are ngc 3223 ( optical hubble type b ) and ngc 5861 ( optical hubble type c ) .
the fit of logarithmic spirals to the arms is excellent ( figure 8) , as was already alluded to by the pioneering empirical rectifications of logarithmic spirals by danver ( @xcite ) . in the middle panel of figure 8 ,
two galaxies ( ngc 718 and ngc 309 ) are presented ; they belong to class @xmath5 ; again the overlay of the inverse fourier transform confirms the robustness of the fourier technique of identifying the dominant mode and binning according to pitch angle at k@xmath0 . note that ngc 718 is of hubble type a ,
while ngc 309 is of hubble type c. finally , contour overlays of the fourier transform for two of the class @xmath6 spirals ( ngc 5085 hubble type c and ngc 7083 hubble type b ) are illustrated in the lower panel of figure 8 .
the question of whether star formation actually traces underlying spiral structure in the population ii disk in other words , whether spiral density waves actually trigger star formation has been elegantly reviewed by elmegreen ( @xcite ) .
the issue is whether the interstellar gas does not form many stars while it is in the interarm regions , but rapidly forms stars when it enters a stellar spiral arm .
since an optically flocculent galaxy can belong to the same dust penetrated bin as that of an optically grand design and both have a dominant m=2 mode in the near - infrared , the implication is that stellar density waves do not principally trigger star formation for if they always did , the morphology of the optical disk would closely mimic its dust penetrated archetype .
only if star formation traced density waves could one _ predict _ what the underlying mass distribution of a galaxy would be and what the dust penetrated class would be , which one can not .
ngc 3521 has a regular two - armed stellar mass distribution ; in the optical , this galaxy is flocculent .
important confirmations of this result are to be found in the literature .
elmegreen and elmegreen ( @xcite ) showed that the star formation rate per unit area , from uv , fir , h@xmath4 , integrated colour , and surface brightness data , is the same for grand design and flocculent galaxies .
a study by mccall and schmidt ( @xcite ) showed that the fraction of galaxies with grand design spirals is the same whether they have type i , type ii , or indeed no recorded supernovae , which implies that this fraction is independent of the supernova rate .
similar inferences have been made for our galaxy .
talbot ( @xcite ) found that the star formation rate per unit molecular mass in the milky way is independent of radius , and wouterloot , brand and henkel ( @xcite ) found the same star formation rate per unit molecular mass for molecular clouds in the far - outer disk of our galaxy .
viewed at arcsecond resolution , it has become clear that interarm dust ( and gas ) can be very widespread , covering much of the optical disk of a spiral galaxy ( eg .
block @xcite ) .
giant molecular associations are to be found in flocculent as well as grand design spirals ( eg .
sakamoto @xcite ) . a unified view of macromolecules , very small grains and large dust particles in the whirlpool galaxy m51 has recently been produced by combining optical and infrared space observatory data ( see block et al .
widespread spirals of macromolecules and dust grains lie not only in the arm , but also in the interarm regions ( see figure 1 in block et al .
@xcite ) . if these spirals are indeed precursors to density wave triggered star formation , a _ partial _ decoupling of the population i and ii disk is suggested ( the reader is referred to the ` smooth red arms ' and ` bar ' in the population ii disk of m51 shown by the photographic work of zwicky @xcite ) .
finally , there are indications from rotation curve analyses that optical hubble type is not correlated with the evolved population ii morphology .
indications are that the distribution of stellar mass in the disk , to which the properties of rotation curves are tied , can be different for the _ same _ hubble type ( see burstein and rubin @xcite ) .
these authors find three principal types of mass distribution , with hubble type a and b classes being found _ among all three types _ more or less equally .
to secure rotation curves , one selects galaxies from a sample which are well inclined to the line - of - sight ; whereas for morphology , a preferentially face - on criterion is essential to delineate and study the arms .
if the arms could clearly be delineated in their sample of ( inclined ) galaxies , it would not be unreasonable to predict that a correlation might be found between our dust penetrated classes and their three principal types of mass distribution .
to derive a coherent physical framework for the excitation of spiral structure in galaxies , one must consider the co - existence of _ two different _ dynamical components : a gas - dominated population i disk ( ob associations , hii regions , cold interstellar hi gas ) and an evolved stellar population ii component .
the hubble classification scheme has as its focus , the morphology of the population i component only . secondly , three principal archetypes @xmath4 , @xmath5 and @xmath6 for the dust penetrated morphology may be proposed , characterised by pitch angles at k@xmath0 of @xmath2 10@xmath3 ( the @xmath4 class ) , @xmath2 25@xmath3 ( the @xmath5 class ) and @xmath2 40@xmath3 ( the @xmath6 bin ) , respectively ( see figure 1 ) .
the ubiquity of low - m features finds a natural interpretation within the modal theory , because higher - m modes are expected to be damped by absorption at the ilr in the evolved stellar disk .
if the shear in a galactic disk is not high , it is possible for higher - m modes to develop as the ilr is then not as effective .
such higher - m modes may also develop for gas - rich galactic disks .
fourthly , and related to the decoupling : we find no correlation between optical elmegreen arm class and pitch angle in the dust penetrated regime ( see figure 9 ) . in our @xmath4 bin , for example ,
ngc 3521 is of elmegreen class 3 , ngc 5861 is of elmegreen class 12 , yet both have pitch angles differing by only one degree at k@xmath0 ( see table 1 ) . in the @xmath5 bin ,
the variation in elmegreen arm class in table 1 is again from flocculent ( 3 ) to grand design ( 12 ) .
figure 9 clearly shows why a completely new classification of disks in the near - infrared domain is called for : two galaxies having completely different elmegreen arm classes can have almost identical pitch angles at k@xmath0 . in figure 10
, we plot a similar graph , but of pitch angle at k@xmath0 versus de vaucouleurs t index .
the @xmath4 dust penetrated archetype can be the resident disk to a vast range of optical hubble or de vaucouleurs t indices .
likewise for the @xmath5 and @xmath6 classes .
although the m=2 mode is a dominant mode in our fourier spectra ( figs .
37 ) , there is often a significant m=1 component as well .
if one wishes to draw attention to the m=1 mode , we have proposed that a prefix l indicate the lopsidedness ; the dust penetrated classes would then be l@xmath4 , l@xmath5 and l@xmath6 .
any of the l archetypes will look like one logarithmic spiral ( instead of two ) in figure 1 , determined by the appropriate pitch angle . in that figure ,
we have plotted one logarithmic spiral for the m=1 mode of ngc 1637 and for the ( leading ) arm in ngc 4622 .
we use the terminology e = evensided for dominant two - armed ( m=2 ) modes , and h3 and h4 for third and fourth order harmonics .
it may also be appropriate , in due course , to introduce a class f = fragmentary , for stellar disks which are patchy and irregular .
it is important that the letter i = irregular not be used , since a few optically flocculent and late type sd _
galaxies , for example , can belong to this group .
an example of an f class would be the flocculent type sd companion to m81 , ngc 2976 ( see section 2.3 in block @xcite ) , which presents a very irregular appearance in the near - infrared no central mass concentration and no coherent stellar morphology at all .
as in the optical scenario , it can not be expected of any classification scheme to include _ every single _ galaxy .
the cosmic tapestry , especially at high z , is too rich and varied .
there is one additional crucial point pertaining to galaxy morphology classification in the high redshift universe : the detection of an ` old ' stellar disk ( ages greater than 1 or 2 gyr ) beyond a redshift of 1.5 is extraordinarily difficult see figure 6 in abraham ( @xcite ) .
as one probes progressively deeper into redshift space , the rest frame of a galaxy systematically shifts to the ultraviolet .
this is the morphological k - correction : one is observing galaxies at bluer rest wavelengths as a function of redshift .
strong surface brightness selection effects bias against the detection of even intermediate - aged stellar populations ( see abraham @xcite ) and the decline in visibility of older stellar populations with redshift is dramatic .
the late type/`irregulars ' reported in many of the higher redshift galaxies in the hubble deep field ( hdf ) describe _ ultraviolet _ morphologies of the _ gaseous _ disks only .
some of these disks _ could possibly be _ partially or fully decoupled from their stellar disks as we find locally for the optical irregular in the system ngc 5195/m51 .
( a promising note is that fourier decompositions can be done on hdf images .
this is provided that the number of components used in the decomposition is not too large ( windhorst , private communication ) since faint galaxy images are so small , even with the hubble space telescope ) .
returning to the f class : while some optical irregulars may belong to this type , the dust penetrated disks of other optical irregulars ( eg . ngc 5195 ) would not . we believe that while modes have been identified in large ensembles of dust penetrated galactic disks in our local low redshift universe , there ofcourse will always be exceptions where there are no indications of any spiral arms or of modes
ngc 2976 is one example .
we have deliberately kept our preliminary dust penetrated classification scheme _
simple_. of course it is possible to use the terminology ( b)e@xmath4 , ( b)e@xmath5 and ( b)e@xmath6 for example , to indicate the presence of a bar in the evolved disk of evensided galaxies and possibly to give a measure of bar strength in terms of the ` equivalent angle ' of seigar and james ( @xcite , see their figure 10 ) .
however , since it is not uncommon for a barred and unbarred morphology to co - exist within the same galaxy , we have only differentiated evolved disks here on the basis of the dominant fourier mode and on the opening or pitch angle of the stellar arms .
it is a starting point .
the caution of mike disney ( @xcite ) rings in our ears : _ `` what would worry me is that everybody will go off now and write their own morphology and we ll have 500 more different k - band morphological categories .
i do nt know how to solve it ; perhaps several people at this conference should go away and agree to do it .
at least to have some starting places for it otherwise it may be as haphazard as it s been in the past in the optical . '' _ in this paper , we have tried to stress the _ fundamental _ rather than _
incidental _ need to develop a near - infrared classification scheme of galaxies . up to now , we have only classified one of the two components : the population i disk .
a central aspect here is the likely coupling of the population i disk with that of the population ii disk via a _
feedback _ mechanism .
having formulated our dust penetrated classification scheme above , we have tested it on an independent sample of 45 face - on galaxies observed in the k - band by seigar and james ( @xcite ) . it is interesting to note that seigar and james did not study galaxies with very open arm morphologies in the near - infrared .
indeed , 44 out of their 45 galaxies have pitch angles at k less than 13@xmath3 , which designate them to our @xmath4 bin .
nevertheless , this strengthens our conclusion even further , that _
one _ specific dust penetrated archetype ( in this instance , class @xmath4 ) may be the resident disk to a further 44 spirals with a wide range in optical morphologies : from early ( a ) to late type ( d ) in their table 1 .
_ `` the features must be significant they must indicate physical properties of the nebulae [ galaxies ] themselves and not chance effects of orientation and also they must be conspicuous enough to be seen in large numbers of nebulae . '' _
we see a duality of spiral structure .
one for the population i disk ; often a radically different one for the population ii disk .
one half of the story has been missing . a feedback between a young population i disk of an optically flocculent galaxy , and an old grand design population ii disk , for example .
finally , the morphology of evolved disks with open stellar arms could be interpreted as favouring the secular evolution of galaxies from an open to a more tightly wound scenario : the redistribution of angular momentum by large - scale spiral torques being more efficient for those stellar arms which are more open ( eg . pfenniger et al .
@xcite ; combes and sellwood , private communications ) .
the deep insight of our referee , dr p. james , from both an observational and theoretical point of view , has been especially helpful .
dlb is deeply grateful for invitations to be a visiting astronomer at the institute of astronomy in hawaii and at the european southern observatory ( garching and chile ) where much of the work was initiated .
it is a great pleasure to have observed at mauna kea with dr a. stockton , and for his unfailing input of k@xmath0 images over the years .
the present fourier study was undertaken while one of us ( ip ) was a distinguished visitor in the department of computational and applied mathematics at the university of the witwatersrand .
the distinguished visitor s program is funded by the anglo - american and de beers chairman s fund educational trust to which ip and dlb wish to record a special note of gratitude .
we thank drs .
p. grosbl , m. verheijen and m. thornley for sharing their k@xmath0 images with us .
dlb is deeply grateful to giuseppe bertin , bruce elmegreen , franoise combes , jerry sellwood and rogier windhorst for important discussions .
abraham , r.g .
, 1998 , les houches preprint , astro - ph 9809131 allen , r.j .
, 1996 , in new extragalactic perspectives in the new south africa , eds .
block , j.m .
greenberg , kluwer , dordrecht , p. 608 baade , w. , 1963 , evolution of stars and galaxies , ed . c. payne - gaposhkin , harvard univ press , new york , p. 19
baldwin , j.e .
, lynden - bell , d. , sancisi , r. , 1980 , mnras 193 , 313 benedict , g.f . ,
higdon , j.l .
, tollestrup , e.v . ,
hahn , j.m . ,
harvey , p.m. , 1992 , aj 103 , 757 bertin , g. , lin , c.c . ,
lowe , s.a . ,
thurstans , r.p .
, 1989a , apj 338 , 78 bertin , g. , lin , c.c . , lowe , s.a . ,
thurstans , r.p . , 1989b , apj 338 , 104 bertin , g. , 1991 , in iau 146 , dynamics of galaxies and their molecular cloud distributions , kluwer , p. 93
bertin , g. , 1993 , pasp 105 , 640 bertin , g. , 1996 , in : new extragalactic perspectives in the new south africa , eds .
block , j.m .
greenberg , kluwer , dordrecht , p. 227
bertin , g. , lin , c.c . , 1996 , spiral structure in galaxies : a density wave theory , mit press , cambridge , mass block , d.l .
1982 , a&a 109 , 336 block , d.l . ,
wainscoat , r.j . , 1991 , nat 353 , 48 block , d.l . , bertin , g. , stockton , a. , grosbl , p. , moorwood , a.f.m . , peletier , r.f . , 1994a , a&a 288 , 365 block , d.l . ,
witt , a.n . ,
grosbl , p. , stockton , a. , 1994b , a&a 288 , 383 block , d.l . , 1996 , in new extragalactic perspectives in the new south africa , eds . d.l .
block , j.m .
greenberg , kluwer , dordrecht , p. 1
block , d.l . ,
elmegreen , b.g . ,
wainscoat , r.j . , 1996 ,
nat 381 , 674 block , d.l .
, elmegreen , b.g . ,
stockton , a. , sauvage , m. , 1997 , apj 486 , l95 burstein , d. , rubin , v. , 1985 , apj 297 , 423 buta , r. , crocker , d.a . , byrd , g.g . , 1992 , aj 103 , 1526 byrd , g.g . , thomasson , m. , donner , k.j . , sundelius , b. , huang , t - y .
, valtonen , m.j . , 1989 ,
45 , 31 conselice , c.j . , 1997 ,
pasp 109 , 1251 considre , s. , athanassoula , e. , 1988 , a&as 76 , 365 danver , c.g .
, 1942 , lund .
ann , vol .
10 disney , m.j .
, 1996 , in new extragalactic perspectives in the new south africa , eds .
block , j.m .
greenberg , kluwer , dordrecht , p. 588
elmegreen , b.g .
, 1995 , in the formation of the milky way , eds .
e. alfaro and a.j .
dalgado , cambridge university press , p. 28
elmegreen , b.g . ,
elmegreen , d.m . , 1986 ,
apj 311 , 554 elmegreen , b.g . ,
elmegreen , d.m . , 1987 ,
apj 314 , 3 elmegreen , b.g . ,
block , d.l .
, 1998 , mnras , in press ( astro - ph 9810152 ) frogel , j.a . ,
quillen , a.c . ,
pogge , r.w . , 1996 , in new extragalactic perspectives in the new south africa , eds .
block , j.m .
greenberg , kluwer , dordrecht , p. 65
garca - gmez , c. , athanassoula , e. , 1993 , a&as 100 , 431 grosbl , p. , patsis , p.a . , 1996 , in new extragalactic perspectives in the new south africa , eds .
block , j.m .
greenberg , kluwer , dordrecht , p. 251
grosbl , p. , patsis , p.a . , 1998 , a&a 336 , 840 hubble , e. , 1936 , the realm of the nebulae , dover publishers lequeux , j. , gulin , m. , 1996 , in new extragalactic perspectives in the new south africa , eds . d.l .
block , j.m .
greenberg , kluwer , dordrecht , p. 422
lin , c.c .
, 1971 , in highlights of astronomy , vol.2 , ed . c. de jager , reidel , dordrecht , 88 lindblad , b. , 1963 , stockholm obs .
ann , no.5 .
bell , d. , kalnajs , a.j . , 1972 ,
mnras 157 , 1 mccall , m.l . ,
schmidt , f.h .
, 1986 , apj 311 , 548 mark , j.w - k . , 1971 ,
usa 68 , 2095 martin , p.g . , whittet , d.g.b . , 1990 , apj 357 , 113 morgan ,
w.w . , 1958 ,
pasp 70 , 364 oort , j.h . , 1970 , in iau 38 , the spiral structure of our galaxy , reidel , 1 pfenniger , d. , martinet , l. , combes , f. , 1996 , in new extragalactic perspectives in the new south africa , eds .
block , j.m .
greenberg , kluwer , dordrecht , p. 291
puerari , i. , dottori , h.a . , 1992 , a&as 93 , 469 puerari , i. , dottori , h. , 1997 , apj 476 , l73 rix , h - w . ,
rieke , m.j . , 1993 , apj 418 , 123 sakamoto , k. , 1996 , apj 471 , 173 sandage , a. , 1961 , hubble atlas of galaxies , washington dc , carnegie institution sandage , a. , tammann , g.a .
, 1987 , a revised shapley - ames catalog of bright galaxies , carnegie institution of washington , publication 635 sandage , a. , bedke , j. , 1994 , the carnegie atlas of galaxies , washington dc , carnegie institution schrder , m.f.s . ,
pastoriza , m.g . ,
kepler , s.o . ,
puerari , i. , 1994 , a&as 108 , 41 seigar , m.s . ,
james , p.a .
, 1998a , mnras 299 , 672 seigar , m.s . ,
james , p.a .
, 1998b , mnras 299 , 685 shaw , m.a . , combes , f. , axon , d.j . ,
wright , g.s .
, 1993 , a&a 273 , 31 spinrad , h. , harlan , e. , 1972 , pasp 85 , 815 talbot , r.j . , 1980 ,
apj 235 , 821 thornley , m.d . , 1996 ,
apj 469 , l45 thronson , h.a . ,
rubin , h. , ksir , a. , 1991 , mnras 252 , 550 van den bergh , s. , 1960 , apj 131 , 215 wainscoat , r.j . ,
cowie , l. , 1992 , aj 103 , 332 wouterloot , j.g.a . , brand , j. , henkel , c. , 1988 , a&a 191 , 323 | to derive a coherent physical framework for the excitation of spiral structure in galaxies , one must consider the co - existence of _ two different _ dynamical components : a gas - dominated population i disk ( ob associations , hii regions , cold interstellar hi gas ) and an evolved stellar population ii component .
the hubble classification scheme has as its focus , the morphology of the population i component only . in the near - infrared
, the morphology of evolved stellar disks indicates a simple classification scheme : the dominant fourier m - mode in the dust penetrated regime , and the associated pitch angle . on the basis of deprojected k@xmath0 ( 2.1@xmath1 ) images , we propose that the evolved stellar disks may be grouped into three principal dust penetrated archetypes : those with tightly wound stellar arms characterised by pitch angles at k@xmath0 of @xmath2 10@xmath3 ( the @xmath4 class ) , an intermediate group with pitch angles of @xmath2 25@xmath3 ( the @xmath5 class ) and thirdly , those with open spirals demarcated by pitch angles at k@xmath0 of @xmath2 40@xmath3 ( the @xmath6 bin ) .
there is no correlation between our dust penetrated classes and optical hubble binning ; the hubble tuning fork does not constrain the morphology of the old stellar population ii disks . any specific dust penetrated archetype
may be the resident disk of _ both _ an early or late type galaxy .
the number of arms and the pitch angle of the arms at k@xmath0 of the early - type ` a ' spiral ngc 718 are almost identical to those for the late - type ` c ' spiral ngc 309 .
we demonstrate that galaxies on opposite ends of the tuning fork can display remarkably similar evolved disk morphologies and belong to the same dust penetrated class .
furthermore , a prototypically flocculent galaxy such as ngc 5055 ( elmegreen arm class 3 ) can have an evolved disk morphology almost identical to that of ngc 5861 , characterised in the optical as having one of the most regular spiral patterns known and of elmegreen class 12 .
both optically flocculent or grand design galaxies can reside within the _ same _ dust penetrated morphological bin . as was suggested by block et al .
( @xcite ) , it is the gas dominated population i component which determines the optical types ( a , b , c ) , decoupled from the population ii .
those l = lopsided galaxies ( where m=1 is a dominant mode ) are designated l@xmath4 , l@xmath5 and l@xmath6 according to the dust penetrated pitch angle ; e = evensided galaxies ( where m=2 is the dominant fourier mode ) are classified into classes e@xmath4 , e@xmath5 and e@xmath6 , according to our three principal dust penetrated archetypes .
the l and e modes are the most common morphologies in our sample , which spans a range of hubble types from early ( a ) to late ( irregular ) .
having formulated our dust penetrated classification scheme here , we have tested it on an independent sample of 45 face - on galaxies observed in the near - infrared by seigar and james ( @xcite , b ) . |
at the electron - proton ( @xmath5 ) hera collider , the study of dis is performed in an unique and optimal way up to values of the squared momentum transfer @xmath0 of @xmath6 gev@xmath2 .
recently , the two hera collaborations h1 and zeus have reported @xcite an excess of events at _ very high _ @xmath0 ( defined in the following as @xmath7 15000 gev@xmath2 ) compared to standard model expectations .
this observation has triggered an important activity on `` beyond the standard model '' theories which might explain the effect @xcite .
the most favoured solution involves the production of a new resonance , which after its decay is kinematically similar to a standard deep inelastic scattering ( dis ) event .
indeed , in the naive quark - parton model ( qpm ) , a parton carrying a fraction @xmath8 of the momentum of the proton scatters elastically against an electron and in the standard model , this interaction is representing a t - channel scattering occuring via the exchange of a gauge boson such as a photon or a @xmath9 . beyond the standard model , the naive qpm picture can represent the formation of an s - channel resonance ( generically called `` leptoquark '' ) subsequently followed by a two - body decay . while in the first interpretation the bjorken @xmath8 variable is one of the two relevant variables to characterize the scattering , in the second case
the invariant mass @xmath10 of the system formed is the physical quantity of interest .
this mass is related to @xmath8 in the naive qpm by @xmath11 , @xmath12 being the squared center of mass energy . in `` real '' interactions , the quantum chromodynamics ( qcd )
effects do not modify much this picture at large @xmath0 , since it is characterized in dis by the dominance of one electron + one partonic jet + one remnant jet in the final state , and is thus similar to the one in which a leptoquark is produced and decays in 2 bodies .
so , when looking at high @xmath0 deep inelastic scattering ( dis ) in an inclusive way it is possible to distinguish between these two scenarios by comparing the event rates in different @xmath8 or @xmath10 intervals : the smooth evolution of the dis cross - section as a function of @xmath8 can be opposed to the appearance of a sharp peak in the invariant mass distribution , characteristic of a leptoquark .
the hera effect is complicated by the fact that the invariant mass distributions of the event samples of the two collaborations differ : at a mass greater than 180 gev , the 7 h1 events appear clustered around a mass of 200 gev , while the 5 zeus events are broadly distributed between 191 and 253 gev .
it is the subject of this paper to understand how significant is this difference and to provide a combined mass spectrum using the complete available information of these events in order to give an interpretation of the effect . at hera
the kinematic reconstruction does not need to rely on the scattered lepton only , since the most important part of the hadronic system is visible in the almost @xmath13 detectors h1 and zeus .
this redundancy allows for an experimental control of the systematic errors and the radiative corrections , hence to determine in a more precise way the usual dis kinematic variables @xmath14 and @xmath0 which are defined as : + @xmath15 + @xmath16 being the 4-vectors of the incident proton and lepton , @xmath17 the scattered lepton one . in this report
we briefly review in section 2 the methods used at hera to determine the kinematics of the high @xmath0 events . in section 3
we characterize the high @xmath0 region and introduce a new method which optimizes the kinematic reconstruction by determining the measurement errors on an event by event basis .
section 4 is devoted to the detailed study of the very high @xmath0 hera events and discusses the mass distributions obtained .
in order to introduce the kinematic methods used at hera in the high @xmath0 region let us start by a few definitions .
the initial electron and proton ( beam ) energies are labeled @xmath18 and @xmath19 . the energy and polar angle axis is defined at hera as the incident proton beam direction .
] of the scattered electron ( or positron ) are @xmath20 and @xmath21 .
after identification of the scattered electron , we can reconstruct the following independent hadronic quantities : @xmath22 , obtained as the sum of the scalar quantities @xmath23 associated to each particle belonging to the hadronic final state , @xmath24 as its total transverse momentum and define the hadronic inclusive angle @xmath25 : @xmath26 @xmath27 are the four - momentum vector components of every energy deposit in the calorimeter originating from any hadronic final state particle .
the corresponding quantities for the scattered electron are @xmath28 out of these variables , it is possible to write the four methods which have been used to determine the invariant mass of the very high @xmath0 events ( the formulae are given in the appendix ) .
- the electron only method ( @xmath29 ) based on e and @xmath21 .
+ - the hadrons only method ( @xmath30 ) based on @xmath22 and @xmath24 @xcite . + - the double angle method ( da ) based on @xmath21 and @xmath25 @xcite . + - the sigma method ( @xmath31 ) based on e , @xmath21 and @xmath31 @xcite .
the @xmath29 method is precise at high @xmath32 but becomes less precise in @xmath14 or @xmath10 at low @xmath32 .
the @xmath30 method is the only one available for charged current events , but is less precise than the three others in neutral currents , so it will be ignored in the following .
the da method is precise at high @xmath0 both at high and low @xmath32 .
the @xmath31 method is similar to the @xmath29 method at high @xmath32 and to the da at low @xmath32 apart from @xmath0 which is less precise .
however it is insensitive to initial state qed radiation ( isr ) in m , @xmath32 and @xmath0 .
see @xcite for a detailed discussion of the properties of these methods and @xcite for their behaviour in presence of isr .
it should be stressed that the quality of any kinematic reconstruction depends mainly on the precision achieved on the observables , such as e , @xmath21 , @xmath22 or @xmath25 .
it is one of the major tasks of the experiments to obtain precise and unbiased quantities , and several techniques have been developped in the hera structure function analyses to achieve these goals .
we refer to the original publications for a description of these studies @xcite , which discuss for instance how the electron energy is calibrated , or how the hadronic final state is measured at low @xmath32 , when the hadronic jets are in the vicinity of the beam pipe .
at _ high _ @xmath0 ( defined in the following as @xmath7 2500 gev@xmath2 ) the kinematic reconstruction is in general more precise than at low @xmath0 and indeed the differences between the results obtained with the methods seen above are small .
however since the number of events drops rapidly with increasing @xmath0 it is still crucial to optimize the reconstruction by making a full use of all the observables of these events . the improvements on the reconstruction at high @xmath0 come mainly from the better measurement of the hadronic final state
this is due to the fact that i ) the individual hadron energy is on average greater ; ii ) the losses in the beam pipe and in the material in front of the calorimeter are in proportion smaller than at low @xmath0 , i.e. @xmath31 , @xmath24 and @xmath25 are less affected by these losses ; iii ) the hadronic final state displays more often a ( single ) collimated jet configuration .
this is due both to kinematics ( events are in average at higher @xmath8 ) , and to the smaller gluon radiation and power corrections ( @xmath33 is smaller ) .
all these characteristics enable a precise measurement of the hadronic angle , and thus are particularly favourable to the da method which indeed becomes more accurate with increasing @xmath0 .
this method has however the drawback to be very sensitive to isr and it is thus difficult to apply to a small number of events .
for instance , the isr of a 2.75 gev photon , if not taken into account produces a shift of 10% on the reconstructed mass and 20% on @xmath0 . in order to overcome this drawback
we introduce a new method which , by making use of the precision on @xmath25 , estimates _ from the data _ on an event by event basis the shift in the measurement of @xmath20 and @xmath31 , denoted in the following @xmath34 with @xmath35 or @xmath31 .
these shifts allow i ) the @xmath31 method to be corrected ; ii ) the presence of an undetected initial state photon ( down to an energy of about 2 gev ) to be identified with high efficiency , and therefore to correct for it .
let us assume that there is no isr ( the specific treatment of isr is discussed in the next section , although the two steps are not separated in the procedure ) , and that @xmath21 and @xmath25 are precisely measured ( implying @xmath37 ) . from energy momentum conservation the two following equations can be derived : @xmath38 under these assumptions @xmath39 and @xmath40 are determined on an event by event basis .
varying @xmath21 and @xmath25 within their typical errors , the @xmath25 resolution improves to 30 mrad .
we use here the r.m.s . , rather than the standard deviation obtained from a gaussian fit to the distribution , to take into account non - gaussian tails which contribute in the propagation of the systematic errors .
] ( 5 and 40 mrad ( r.m.s . ) respectively , at high @xmath0 )
we can obtain the errors arising from the angular measurements on @xmath39 and @xmath40 . a more direct way to see the uncertainties arising from this determination
is illustrated in fig.1 , in which the error on @xmath20 and @xmath31 reconstructed using equations 3 and 4 are compared to their `` true '' value , which is known for the reconstructed simulated events .
for a given event , the error on @xmath41 , which is defined as @xmath42 , is of the same order as the r.m.s .
( denoted @xmath43 ) of the @xmath44 distribution obtained from a large sample ( fig.1b ) , showing that its use will not bring an improvement on an event by event basis . on the other side ,
the relative error on @xmath45 is smaller , about 30% of @xmath46 , and diminishes at high @xmath32 ( compare fig.1e to 1f , and 1b to 1c ) implying that correcting @xmath31 will provide a better kinematic measurement , in particular at high @xmath32 .
the @xmath36 kinematic variables are thus derived from the @xmath22 ones by including the effect of @xmath47 , i.e. @xmath48 the comparison of the @xmath36 and @xmath31 reconstruction of m , @xmath32 and @xmath0 for high @xmath0 events fully simulated in the h1 detector is shown in fig.2 .
the improvement obtained by the recalibration of @xmath31 is clearly visible , allowing this method to be compared favourably to the original ones .
this comparison is shown in fig.3 for high @xmath0 events ( in all the high @xmath0 section , an @xmath49 cut against hard initial state radiation is applied , see next section ) both at high @xmath32 and at low @xmath32 . at high @xmath32 @xmath50 ) , the @xmath29 and the @xmath36 methods are comparable in @xmath10 and @xmath32 and slightly better than the da one ( fig.3a , b ) . in @xmath0 , the da displays the narrowest peak both at high and low @xmath32 ( fig.3c , f ) , but its high sensitivity to isr induces tails in the distribution which renders its r.m.s .
larger than the @xmath36 and @xmath29 ones . at low @xmath32 ( @xmath51 the @xmath29 method has a relatively poor resolution in m , much worse than the @xmath36 one which is slightly better than the da one . in conclusion , for the high @xmath0 events the @xmath36 method is similar or slightly better than the @xmath29 and da method .
the influence of isr is taken into account in the simulation programs used by the hera experiments , but on a small number of events it is difficult to control the migration due to an unseen isr photon when using the @xmath29 and da methods .
both h1 and zeus use an experimental cut on the total @xmath49 of the event ( @xmath52 ) or equivalently on the normalized @xmath49 called hereafter @xmath53 for the electron and for the hadronic final state satisfy : @xmath54 ; @xmath55 . ] and defined as @xmath56 the @xmath49 cuts used in h1 and zeus correspond to good approximation to @xmath57 and prevent an isr photon from carrying away more than about 25% of the incident electron energy .
such a photon can induce a large shift to the reconstructed mass @xcite of @xmath58 for @xmath59 ( i.e. @xmath60 at @xmath61 ) and of 25% on @xmath62 at any @xmath32 .
on @xmath63 the shift is zero since the method is independent of colinear isr for m@xmath64 and @xmath0 .
a more stringent cut on @xmath53 is difficult to implement due to the experimental resolutions on @xmath65 and @xmath66 .
the @xmath36 method allows isr to be identified with high efficiency for @xmath53 as high as @xmath67 .
it makes use of the simple fact that depending on the origin of the @xmath53 shift ( which comes mainly either from isr or from hadronic miscalibration ) , the errors obtained with the @xmath36 method have a completely different pattern : for example if the observed @xmath53 is 1@xmath68 , equations 3 and 4 will give for the isr case @xmath69 and @xmath70 , assuming negligible detector smearing , while for the hadronic miscalibration case , assuming negligible error on the electron energy , we will get @xmath71 and @xmath72 .
these two different types of pattern allow isr to be identified even in the presence of detector smearing , as studied on a complete simulation in the h1 detector , with typically 85% efficiency at high @xmath0 for photon energies
greater than 2 gev .
the exact conditions for isr identification of a 2 gev photons is @xmath73 and @xmath74 .
however , to take into account the detector smearing we use instead : @xmath75 these can be slightly varied depending on the isr identification efficiency requested .
these conditions will `` misidentify '' high @xmath0 non - radiative events in less than 3% of the cases in the h1 detector , implying that in the total sample of isr identified events , the fraction of genuine isr events is higher than the fraction of non - radiative events .
this fraction increases with increasing @xmath0 since the hadronic angle becomes more precise , i.e. the isr identification becomes more efficient . in case of isr identification ,
the equations 3 and 4 are solved again after recalculating @xmath66 and @xmath65 using @xmath76 instead of @xmath77 .
the improvement obtained can be visualized in fig.4 which shows the same distributions as in fig.3 but on the sample of dis high @xmath0 events having radiated an isr photon of 2 gev or more .
the dramatic improvement underlines the importance of controlling experimentally the radiation effect .
furthermore , the soft isr photons which can not be identified below 2 gev have a small effect on @xmath78 ( @xmath79 ) but still a sizeable one on @xmath62 , up to about 7% , or 4% on @xmath59 .
let us remind some characteristics of the excess of events at high @xmath0 observed by the h1 and zeus collaborations in the data collected from 1994 to 1996 @xcite . at @xmath80 gev@xmath2 , 24 events have been observed for an expectation of [email protected] , representing a probability of 0.0074 @xcite .
furthermore some of these events exhibit the additional characteristics either to cluster around a rather precise value of the invariant mass of the produced system ( m@xmath81 gev , for 7 events out of the 12 of h1 , while @xmath82 are expected ) , or to lie at very high invariant mass ( m@xmath83 gev and @xmath84 , for 4 events out of the 12 of zeus , while @xmath85 are expected ) .
actually the messages of these two observations could be different as already pointed out @xcite and we will try here to explore further this point with the use of our new kinematic tool . among the obvious possible reasons for the difference
are the influence of isr which can strongly distort a distribution based on a small number of events , or the effect of a specific miscalibration .
the @xmath36 method having been devised to respond to these problems , we can now study in more details the 11 extreme events mentioned above for which the complete kinematic information has been published by the two collaborations .
we add to this sample 2 events and systematic shifts between @xmath29 and da variables of the 5 other zeus published events .
this assumption has a negligible influence on the conclusions drawn below . ] from zeus in order to include all events which satisfy the condition @xmath80 gev@xmath2 and m@xmath86 gev , since we want to test the significance of the h1 event clustering at masses around 200 gev over the complete very high @xmath0 hera sample . in table 1
are given the m , @xmath32 and @xmath0 of the 13 events at very high @xmath0 and high mass reconstructed with the four different methods discussed above .
the values of m , @xmath32 and @xmath0 are identical to those published for the @xmath29 and da methods .
the @xmath31 values have been released recently by h1 @xcite and can be computed straightforwardly for the zeus events .
the errors are also reproduced exactly except for those of the @xmath29 method to which we conservatively added quadratically an error of 1.5% ( i.e. half of the absolute energy scale uncertainty of 3% given by the two experiments ) to account for potential miscalibrations between the different regions of the detectors . as we will see below the overall energy calibration can be checked to be well under control . for the errors not published by the collaborations ( @xmath36 , @xmath31 for h1 and zeus , and da for h1 ) we used the following prescription , which was checked to be consistent with the other published errors :
they are obtained here using a possible error of @xmath87 mrad and @xmath88 mrad for the electron and hadronic angle respectively .
for the error on the electron ( hadronic ) energy we used @xmath89 ( @xmath90 ) for h1 and @xmath91 ( @xmath90 ) for zeus , which includes for the electron case both resolution effects , dead material corrections and the 1.5% overall error just mentioned .
these two values for the electron energy error have been checked on the data ( see below ) .
additional sizeable errors as published by the collaborations , due for instance to special energy corrections , have been taken into account in two events ( h3,z1 ) .
for the h1 events the errors of an event are similar for each method , with a slight advantage for the @xmath29 method at @xmath32 above 0.5 . for the zeus events the da errors are the smallest , since the error on the electron energy which is somewhat larger than in the h1 case has an influence on the 3 other methods .
this also explains why the zeus collaboration chooses the da method rather than the @xmath29 method favoured by h1 .
also mentioned is the result of the @xmath36 isr identification , and we can see that 2 events ( h5 and z4 ) are classified as radiative , and have been corrected accordingly .
these 2 events indeed show the characteristics of a radiative event , i.e. there exists a value of @xmath92 for which the following equations , which hold exactly if there were no detector effects , are verified in a good approximation : @xmath93 @xmath94 @xmath95 @xmath96 indeed in the case of h5 a photon with an energy ( 2.9 gev)consistent with the @xmath97 derived from the previous equations is observed in the special photon calorimeter located close to the beam pipe and designed to measure photons emitted collinearly to the incident electron , such as isr photons . the isr effect on the mass is sizeable , since @xmath98 = 1.08 for h5 and 1.11 for z4 .
both values are consistent with the determined @xmath99 and @xmath100 of these events ( see tab.1 ) .
the reconstructed kinematics are consistent between the different methods except for the 2 radiative events and for the events displaying a significant difference between @xmath59 and @xmath62 ( events h7,z1,z5 ) , which indeed have the largest @xmath99 and/or @xmath100 . on these small samples of events , and after isr corrections , the averages of the absolute value of the error on the hadronic energy are similar for h1 and zeus ( @xmath101= 6% compared to 5% ) , while on the electron energy the h1 average error is smaller : @xmath102=3% compared to 5% for zeus
. the uncertainty on the electromagnetic absolute energy scale can be estimated _ from the data _ using @xmath103 ( the 2 radiative events are excluded from these means ) which gives @xmath104 for h1 and @xmath105 for zeus , both values in good agreement with the value of 3% given by the collaborations .
the uncertainty on the hadronic absolute energy is obtained similarly and gives @xmath106 for h1 and @xmath107 for zeus , also in acceptable agreement with the values quoted by the collaborations .
after all these consistency checks , the use of the @xmath36 method in the two samples will now allow a consistent mass distribution to be derived from the very high @xmath0 events .
.[tab9 ] _ kinematic properties of the 13 events observed by the h1 and zeus collaborations which have , at least in one method , @xmath108 gev@xmath2 and @xmath109 180 gev . for event z6 , only the da values are accurate , the values of the other methods being extrapolated
. the values of m , @xmath32 and @xmath0 are identical to those of the original papers ( @xmath29 , da , @xmath31 for h1 @xcite , @xmath29 , da for zeus @xcite ) .
the new values of the table are for the @xmath31 ( zeus ) and for the @xmath36 ( h1+zeus ) , and for the non available errors ( see text for more details ) .
also given are @xmath53,@xmath110and @xmath111 , which allow to quantify the size of the errors and to tag the presence of initial state radiation and to correct for it ( marked with a `` * '' ) .
after isr corrections , the ( @xmath53,@xmath110,@xmath111 ) values are for event h5 : ( 1.00,.00,.00 ) and for event z4 : ( 1.00,+.04,.01 ) .
the masses are given in gev , the @xmath0 in @xmath112 gev@xmath2 .
_ [ cols="^,^,^,^,^,^,^,^,^,^,^,^,^,^ " , ] in the following we will not consider any more the event z5 , since it survives the @xmath0 cut only for the da method
. it might be a radiative event which could not be identified due to a large smearing in the hadronic energy .
in any case ( radiative or not ) it would not survive the @xmath113 cut , which is one of the 2 conditions the final sample must satisfy .
note `` en passant '' that its mass is reduced from @xmath62=232 gev to @xmath78=207 gev .
the mass distributions of the remaining 12 events , obtained with the @xmath114 da , @xmath31 and @xmath36 , are displayed in fig.5a , b , d , e .
the event z4 has the highest mass at @xmath115 gev .
it is identified as a radiative event and its mass is reduced by the @xmath36 determination to @xmath116 gev , thereby reducing the scattering appearance of the zeus events .
note that the error quoted on the da mass is much smaller than the effect due to radiation .
for the 5 zeus events the weighted average mass is decreased from m@xmath117=226@xmath118 gev to m@xmath119=216@xmath120 gev . actually with the @xmath36 determination 4 out of
the 5 zeus events lie between 220 and 228 gev , the 5@xmath121 one ( z6 ) being at 185 gev , i.e. at a lower mass than any of the 7 h1 events ( see fig.5b ) .
the average mass of the 7 h1 events is essentially not dependent on the method used : m@xmath119=199@xmath122 gev , m@xmath123=200@xmath124 gev , m@xmath117=201@xmath87 gev .
the 7 h1 events remain clustered between 188 and 211 gev .
thus the h1 and zeus samples are concentrated at significantly different mass values and this splitting can not be accounted for either by isr or by detector effects .
the small number of events involved prevents a definite interpretation of this effect .
however the fact that no event among the 5 of zeus is found in the bin where the 7 h1 events are measured suggests that this specific accumulation is a statistical fluctuation . in fig.5c
is shown the reconstruction of a narrow resonance generated at m=200 gev with the lego @xcite program in the h1 detector using the @xmath36 method , @xmath29 , da and @xmath31 respectively : ( 195.5 ; 7.5 ) , ( 195.3 ; 7.4 ) , ( 195.2 ; 9.6 ) , ( 195.3 ; 7.8 ) .
the bias of about 5 gev can be removed by taking into account the mass of the jet , but is present when using inclusive methods . ] .
the width of the distribution which include experimental and qed / qcd radiation effects , can not accomodate the tails of the measured experimental distributions ( fig.5a , b , d , e ) .
if we make an ideogram from the histogram 5e i.e. if we apply gaussian smearing to each of the 12 events using their @xmath36 mass as a mean , and their error as r.m.s .
, we obtain fig.5f , which also shows the incompatibility with fig.5c . to reconcile the 2 `` peaks '' visible in fig.5f
would require to miscalibrate uniformly the electron energy of the h1 events at least by + 6% and the zeus events by @xmath125% , values completely incompatible with the absolute scale uncertainties found in the previous section .
note however , that the h1 events alone support the narrow resonance hypothesis .
since after the @xmath36 kinematic treatment none of the 12 events migrates outside the very high @xmath0 and mass region , we confirm that the visible excess published by the collaborations at very high @xmath0 is not due to detector ( calibration ) or radiation ( isr ) effects .
we have introduced a new reconstruction method ( `` @xmath36 '' ) which allows the kinematic variables of the high @xmath0 events to be determined in a more precise way .
it uses the kinematic constraints on an event by event basis to calibrate the hadronic energy and to identify and correct for the presence of qed initial state radiation .
this method has been applied to the 12 hera events observed by the h1 and zeus collaborations ( for an expectation of about 5 ) at @xmath80 gev@xmath2 and m@xmath86 gev .
the accumulation of the 7 h1 events around 200@xmath126 gev is confirmed .
the average mass of the 5 zeus events is decreased from the double - angle value of 226@xmath118 gev to 216@xmath120 gev .
however none of the zeus events enter the h1 accumulation region . taking the estimation of the systematic errors at face value
, this suggests that , if the observed excess at very high @xmath0 is due to physics beyond the standard model , it is unlikely to be explained by the decay of a single narrow resonance such as a leptoquark .
the increase of the integrated luminosity in the current and in the future years at hera will clarify the origin of this effect . we would like to thank the two collaborations for the data we have studied in this paper . in particular this work has taken place within the h1 collaboration , and some of the results obtained were the outcome of the efforts of many people of the `` beyond the standard model '' group to get the analysis of the very high @xmath0 events completed .
we also would like to thank j. dainton , r. eichler , j. gayler , d. haidt , b. straub and g. wolf for a careful reading of the manuscript and for their useful remarks .
h1 collab .
, t. ahmed et al .
, nucl . phys . *
b439 * ( 1995 ) 471
. + zeus collab .
, m. derrick et al .
, z. phys . *
c65 * ( 1995 ) , 379 .
+ h1 collab .
, s. aid et al .
, nucl . phys . *
b470 * ( 1996 ) 3 .
+ zeus collab .
, m. derrick et al .
, z. phys . *
c72 * ( 1996 ) 399 . | the kinematic reconstruction of neutral current high @xmath0 events at hera is discussed in detail using as an example the recently published events of the h1 and zeus collaborations at @xmath1 15000 gev@xmath2 and m @xmath3 180 gev , which are more numerous than expected from standard model predictions .
taking into account the complete information of these events , the mass reconstruction is improved and the difference between the average mass of the samples of the two experiments is reduced from 26@xmath410 gev to 17@xmath47 gev , but remains different enough to render unlikely an interpretation of the excess observed by the two collaborations as originating from the decay of a single narrow resonance . 15.5 cm -1.0 in -38pt plus 2 mm minus 2 mm desy 97 - 136 + july 1997 + * some properties of the very high @xmath0 + events of hera * ursula bassler , gregorio bernardi + laboratoire de physique nuclaire et des hautes energies + universit paris 6 - 7 , 4 place jussieu , 75252 paris , france + _ e - mail : [email protected] ; [email protected]_ |
by following the stellar dynamics at the center of our galaxy , we have now the most well - established evidence for the existence of a smbh .
the close examination of the keplerian orbits of the so - called s - stars ( also called s0-stars , where the letter `` s '' stands simply for source ) has revealed the nature of the central dark object located at the galactic center . by following s2 ( s02 ) ,
the mass of sgra@xmath2 was estimated to be about @xmath3 within a volume with radius no larger than 6.25 light - hours @xcite . more recent data based on 16 years of observations set the mass of the central smbh to @xmath4 @xcite .
massive black holes in a lower range of masses may exist in smaller stellar systems such as globular clusters .
these are called intermediate - mass black holes ( imbhs ) because their masses range between @xmath5 and @xmath6 , assuming that they follow the observed correlations between smbhs and their host stellar environments .
nevertheless , the existence of imbhs has never been confirmed , though we have some evidence that could favor their existence ( see * ? ? ?
* ; * ? ? ?
* and references therein ) .
if we wanted to apply the same technique to detect imbhs in globular clusters as we use for smbhs in galactic centers , we would need ultra - precise astronomy , since the sphere of influence of an imbh is @xmath7 few arc seconds .
for instance , for a @xmath8 imbh , the influence radius is of @xmath9 assuming a central velocity dispersion of @xmath10 and a distance of @xmath11 kpc .
the number of stars enclosed in that volume is only a few .
currently , with adaptive optics , one can aspire optimistically to have a couple of measurements of velocities if the target is about @xmath11 kpc away on a time scale of about 10 yrs .
the measures depend on a number of factors , such as the required availability of a bright reference star in order to have a good astrometric reference system . also , the sensitivity limits correspond to a k - band magnitude of @xmath7 15 , ( b- ms stars at 8 kpc , like e.g. s2 in our galactic center ) .
this means that , in order to detect an imbh or , at least , a massive dark object in a globular cluster center by following the stellar dynamics around it , one has to have recourse to the very large telescope interferometer and to one of the next - generation instruments , the vsi or gravity @xcite . in this case
we can hope to improve the astrometric accuracy by a factor of @xmath12 . only in that scenario
would we be in the position of following closely the kinematics around a potential imbh so as to determine its mass .
gw astronomy could contribute to imbh detection . in the past years
, the field has reached a milestone with the construction of an international network of gw interferometers that have achieved or are close to achieving their design sensitivity . moreover
, the first - generation ground - based detectors ligo and virgo will undergo major technical upgrades in the next five years that will increase the volume of the observable universe by a factor of a thousand.[multiblock footnote omitted ] the data that will be taken by the advanced interferometers are expected to transform the field from gw detection to gw astrophysics .
the availability of accurate waveform models for the full bbh coalescence in order to construct templates for match - filtering is crucial in the gw searches for compact binaries .
the construction of this kind of templates has recently been made possible thanks to the combination of post - newtonian calculations of the bbh inspiral and numerical relativity simulations of the merger and ringdown .
two approaches to this problem are the effective - one - body techniques @xcite and the phenomenological matching of pn and nr results @xcite . in this paper
we use the results of the latter .
the structure of this paper is as follows : in section [ sec : astro ] we expand the astrophysical context to this problem and give a description of the different efforts made to address the evolution of a bbh in a stellar cluster from its birth to its final coalescence . in section [ sec : wfmodel ]
we introduce the techniques used in the data analysis of the waveform modeling of bbh coalescences , present our hybrid waveform model , and use the model to compute and discuss expected signal - to - noise ratios in present and future gw detectors .
the use of the new waveform model allows us to give an improved estimate for the number of events one can expect for several physical configurations in advanced ligo and the einstein telescope , which we present in section [ sec : events ] .
we conclude with a summary of our results and future prospects of our work in section [ sec : concl ] .
the aim of this section is not to give a detailed explanation of the processes of formation of imbhs and bbhs , but a general introduction of the two different scenarios that play a role in the formation of bbhs . up to now
, the imbh formation process which has drawn the most attention is that of a young cluster in which the most massive stars sink down to the center due to mass segregation .
there , a high - density stellar region builds and stars start to physically collide .
one of them gains more and more mass and forms a runaway star whose mass is much larger than that of any other star in the system , a very massive star ( vms ) .
later , that runaway star may collapse and form an imbh @xcite . in particular , @xcite described in detail the requirements from the point of view of the host cluster to form an imbh in the center of such a system
. the cluster can not have `` too hard '' binaries , the time to reach core - collapse must be shorter than 3 myr and the environmental velocity dispersion can not be much larger than @xmath13 . under these conditions
the authors find that the mass of the vms formed is @xmath14 .
the later evolution of the vms is not well understood , nor are the necessary conditions that it not evolve into a super - massive star ( sms ) ( see for instance * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and the references in their work ) in this particular scenario , nor are the factors that could limit the mass of such an object such that it not collapse into an imbh .
the process depends on a number of factors and assumptions , such as e.g. the role of metallicity , winds ( see e.g. * ? ? ? * though it is rather unclear how to extrapolate the results they obtain , which are limited to stars with masses of maximum @xmath15 to the masses found in the runaway scenario , which are typically at least one order of magnitude larger ) and the collisions on to the runaway star from a certain mass upwards . on the other hand , @xcite investigated the growth of a runaway particle by coupling direct @xmath16body simulations with smooth particle hydrodynamics , ( sph ) analyzed the evolution of the star and found that stellar winds would not inhibit the formation of a very massive star .
more recently , @xcite considered the effects of stellar evolution on the runaway collision product by analyzing the succession of collisions from a dynamical evolution .
for their low - metallicity models , the final remnant of the merger tree is expected to explode as a supernova , and in their high - metallicity models the possibility of forming an imbh is negligible , finishing with a mass of 1014 @xmath17 at the onset of carbon burning . but as a matter of fact , these stars develop an extended envelope , so that the probability of further collisions is higher .
@xcite did not change the masses in the dynamical simulation accordingly . in any case ,
self - consistent direct - summation @xmath18-body simulations with evolution of the runaway process are called in to investigate the final outcome .
henceforth we will assume that imbhs do form ; in that case , the formation of a binary of them ( bbh ) in a cluster can be theoretically explained in two different ways . _
( i ) the double - cluster channel : _ in this scenario , two clusters born in a cluster of clusters such as those found in the antenn galaxy are gravitationally bound and doomed to collide ( see * ? ? ?
* for a detailed explanation of the process and their references ) .
when this happens , the imbhs sink down to the center of the resulting merged stellar system due to dynamical friction .
they form a bbh whose semi - major axis continues to shrink due to slingshot ejections of stars coming from the stellar system . in each of the processes
, a star removes a small fraction of the energy and angular momentum of the bbh , which becomes harder . at later stages in the evolution of the bbh
, gw radiation takes over efficiently and the orbit starts to circularize , though one can expect these systems to have a residual eccentricity when entering the lisa band @xcite .
this detector will typically be able to see systems of binaries of imbhs out to a few gpc .
for this channel and volume , the authors estimated an event rate of @xmath19 . _
( ii ) the single - cluster channel : _ @xcite added a fraction of primordial binaries to the initial configuration in the scenario of formation of a runaway star in a stellar cluster . in their simulations
they find that not one , but two very massive stars form in rich clusters with a binary fraction of 10% .
@xcite investigated the possibility of emission of gws by such a bbh and estimated that lisa and advanced ligo can detect tens of them depending on the distribution of cluster masses and their densities .
more recently , @xcite recalculated the rate of @xcite for the case of the proposed einstein telescope could see and quoted a few to a few thousand events of comparable - mass bbh mergers of the single - cluster channel . in the case of the double - cluster channel , the cluster , which is in rotation
, results from the merger of the two initial clusters and may develop a triaxiality sufficient to produce sufficient centrophilic orbits .
these `` boxy '' orbits , as seen by @xcite , are typical of systems that do not possess symmetry around any of their axes .
in contrast to loop orbits , a characteristic of spherically symmetric or axisymmetric systems , `` boxy '' orbits bring stars arbitrarily close to the center of the system , since it oscillates independently along the three different axes .
therefore , such stars , due to the fact of being potential sling - shots , can feed the process of shrinkage of the bbh semi - major axes by removing energy and angular momentum from it after a strong interaction . in the strong triaxial systems of @xcite ,
the rotation caused in the process of merger creates an unstable structure in the form of a bar . within the bar the angular momentum
will not be conserved and thus the bbh loss - cone is full due to stars on centrophilic orbits , independently of the number of stars @xmath20 . on the other hand , for bbh systems in clusters , the role of brownian motion has an impact in the loss - cone ( see * ? ? ?
* ) . in the models of @xcite ,
the initial conditions are a realistic parabolic merger of two stellar clusters .
the resulting merged cluster does not show the strong axisymmetry of @xcite . in the simulations we address for the results of this work
, the bbh ( of imbhs ) is not stalling , in spite of the reduced number of centrophilic orbits due to the architecture of the stellar system . in figure [ fig.evoltriax0p10p8 ]
we show the role of the cluster symmetry explicitly by depicting the evolution of the triaxiality of the cluster formed as a result of the merger of the two clusters for our fiducial model in the case of the double - cluster channel ( which is the reference model of * ? ? ?
after a merger which is the result of a parabolic orbit , the final system is oblate rather than prolate ; i.e. @xmath21 , where @xmath22 , @xmath23 and @xmath24 are the cluster axes . at the outskirts
the resulting merged cluster is flatter and at the center the binary of imbhs makes it rather spherical .
@xcite addressed the single - cluster channel scenario after the formation of the imbhs and used additional simulations to further evolve the bbh .
they used scattering experiments of three bodies including relativistic precession to 1st post - newtonian order , as well as radiation reaction caused by gw , so that they did not have to integrate every single star in the cluster to understand the posterior evolution of the bbh . in their work , between the strong encounters , @xmath22 and @xmath25 of the bbh
were evolved by resorting to the quadrupolar formul of @xcite .
the bbh will have completely circularized when it reaches the frequencies probed by advanced ligo and the et , because the emission of gws takes over the dynamics of the system . while the emission of gws is present all the time from the very first moment in which the bbh is formed , the amplitude and frequency of the waves is initially so low that no present or planned detector would be able to register any information from the system .
only when the semi - major axis shrinks sufficiently does the frequency increase enough to `` enter '' the lisa band , which we assume starts at @xmath26 hz .
the bbh then crosses the entire detector window during its inspiral phase , as we can see in figure [ fig.lisa ] .
we depict the evolution of a bbh of mass @xmath27 .
the reason for this particular choice of masses is to give the reader a point of reference to understand the whole picture .
recently @xcite included the effect of the rotation of the host cluster and addressed the dynamical evolution of the global system with a bbh of that mass .
the authors have shown that lisa will see the system of figure [ fig.lisa ] with a median snr of few tens .
the fact that the system merges outside its band prevents lisa from observing the loudest part of the bbh coalescence . in order to follow the system at this early stage of its evolution _ in the lisa band _ , a simple post - newtonian approach suffices for modeling the gw radiation .
we are far enough from the highly relativistic regime and only the inspiral phase of the bbh coalescence is visible to the space antenna , with a rather low snr compared to the posterior phases in the evolution of the bbh . as the binary system depicted in figure [ fig.lisa ]
leaves the lisa band and enters the strong field regime , higher order post - newtonian corrections and eventually input from numerical relativity simulations need to be considered in order to model the gw waveform .
three reference frequencies in the evolution of a compact bbh that approaches its merger are the innermost stable circular orbit ( @xmath28 ) of a test particle orbiting a schwarzschild black hole , the light - ring frequency ( @xmath29 ) corresponding to the smallest unstable orbit of a photon orbiting a kerr black hole and the fundamental ringdown frequency ( @xmath30 ) of the decay of the quasi - normal modes computed by @xcite . for the binary system shown in figure [ fig.lisa ] ,
the values of these three frequencies are @xmath31 , @xmath32 and @xmath33 .
should such a binary exist at a distance of 100 mpc ( the choice for this number is based solely on the fact that is is easily scalable ) , and if it were to be detected with advanced ligo , it should produce a sky - averaged snr of @xmath34 , assuming a low frequency cut - off of 10 hz . to that total snr ,
the contribution of parts of the inspiral happening before the system reaches the characteristic frequencies @xmath28 , @xmath29 and @xmath30 would be 0% , 37% and 95% respectively . figure [ fig.percsnr ] illustrates the same percentages for binaries with total masses between 100 and @xmath35 .
it is immediately noticed that for the binaries of imbhs of interest to this study , most of the snr that these binaries will produce in advanced ligo comes from the last stages of the bbh coalescence .
we can estimate the time that the binary system takes to evolve from @xmath36 , a frequency where the bbh can be seen by lisa , to the lower cut - off frequency of 10 hz of advanced ligo or of 1 hz of the et . a lower order approximation based on the newtonian quadrupole formula
@xcite leads to the following expression for the evolution of the frequency in terms of the chirp mass @xmath37 and frequency of the system @xmath38 we find a delay of only 0.2 yrs ( 80 days ) for a bbh with total mass @xmath39 to go from 0.01 hz to the beginning of the advanced ligo band and almost similar numbers to the beginning of the et band ( the evolution of the system is extraordinarily quick in the late inspiral phase , which explains the fast evolution from 1 to 10 hz ) . in view of these figures ,
lisa could be used as an `` alarm '' to prepare ground - based detectors to register the final coalescence in detail , the death of the bbh as such , by adjusting their _ sweet spots _ ( the most sensitive part of the detector ) to the particular bbh .
the high accuracy of parameter estimation during the inspiral phase of which lisa is capable could be combined with the information obtained from the large - snr triggers that the bbh merger and ringdown will produce in advanced ligo or et and thereby achieve a more complete characterization of the system .
accurate theoretical modeling of the gravitational radiation @xmath40 emitted by a bbh is key to improving bbh detectability and parameter estimation . while post - newtonian ( pn ) theory is valid to model the early inspiral phase of the bbh evolution in the lisa band , an exact description of the merger and ringdown stages is only possible via numerical relativity ( nr ) calculations .
several approaches have been proposed to match pn and nr in order to obtain a full waveform for the entire coalescence , most notably for non - spinning configurations in @xcite and recently introducing spins for non - precessing configurations in @xcite . for the purpose of the snr and horizon distance calculations shown in this paper ,
we have chosen a new procedure for the construction of hybrid pn - nr waveforms in the frequency domain proposed by @xcite .
the construction procedure developed for matching pn and nr data is sketched in figure [ fig : hybconstr ] ; further details on the fitting procedure can be found in @xcite .
in essence , the model consists of a phenomenological fit to hybrid pn+nr waveforms for spinning , non - precessing bbh systems with comparable masses ( @xmath41 ) .
this is compatible with the simulations of @xcite , which find pairs of vmss with mass ratios close to 1 .
the phenomenological waveforms are parameterized by three physical parameters : the total mass of the binary @xmath42 , the symmetric mass ratio @xmath43 and the spin parameter @xmath44 , where @xmath45 , @xmath46 and @xmath47 is the spin angular momentum of the @xmath48th black hole .
only the dominant mode @xmath49 of the gravitational radiation enters in the model ; the effects of the higher modes , which become increasingly significant as @xmath43 decreases , are neglected . for comparable - mass scenarios , this restriction does not substantially affect our results .
the waveform model used in our calculations is valid for binary systems with _ aligned _ spins only , which represents a first step towards the incorporation of the spins of the bhs in full inspiral - merger - ringdown ( imr ) models .
a discussion including precessing systems with arbitrary spins to this date still requires work regarding availability of nr simulations and development appropriate techniques to match precessing merger data with pn .
work in this regard is ongoing . for the results presented in this article we choose to focus on three configurations : ( i ) equal - mass , non - spinning , ( ii ) equal - mass , equal - value spins aligned with the direction of the total angular momentum and magnitude @xmath50 and ( iii ) mass ratio @xmath51 ( @xmath52 ) , non - spinning .
provided with the full imr waveforms given by our model , we are interested in assessing the detectability of these systems by current and future gw observatories .
the snr of a model waveform with respect to the output stream of the detector is the quantity typically quoted to signify the detectability of a signal .
the snr @xmath53 produced by a gw signal @xmath40 in a detector can be computed as ( see e.g. * ? ? ?
* ; * ? ? ?
* ) @xmath54 where @xmath55 is the fourier transform of the strain @xmath40 . in the last equation @xmath56 represents the one - sided noise spectral density of the detector . figure
[ fig : sfvsh ] provides a graphical representation of the detectability of several bbhs by current and future generations of ground - based gw detectors .
displayed are the design sensitivity of current initial ligo and virgo ( sensitivities that have been met or approximately met during the s5/vsr1 data taking ) , the proposed noise curves of advanced virgo and two possible configurations of advanced ligo ( broadband or `` base '' and optimized for @xmath57 bbhs ) and the designed noise budget for the einstein telescope in its broadband , `` base '' and `` xylophone '' @xcite configurations .
the hybrid waveforms constructed according to the model of fig .
[ fig : hybconstr ] for the three chosen configurations has been conveniently scaled to represent bbh systems with total mass @xmath58 and @xmath59 . as the right - hand - side of equation [ eq : snrdef ] suggests , plotting the quantity @xmath60 versus @xmath61 allows for direct visual comparison of the importance of each of the stages of the bbh coalescence .
the three frequencies @xmath62 and the lorentzian ringdown frequency @xmath63 are marked on top of the configuration ( i ) waveforms with solid circles , squares and open squares respectively .
one can immediately appreciate that systems with total mass above @xmath64 fall almost completely below the 40 hz `` seismic wall '' of the initial ligo detectors ; however they will become very interesting sources for the second generation of gw interferometers and the proposed einstein telescope .
indeed , as we show in section [ sec : astro ] , they will also be seen by the future space - borne lisa .
additionally it is easy to appreciate why the hang - up configuration ( iii ) will produce larger snrs and therefore will be seen to further distances , for it merges at higher frequencies with respect to a non - spinning configuration with the same @xmath43 .
in contrast , for equal spin values , the snr decreases with smaller symmetric mass ratios ( compare configurations ( i ) and ( iii ) ) . in figure [
fig : snrall ] we compute the snr expected for these sources in each of the above - mentioned detectors as a function of the redshifted total mass of the system @xmath65 , for optimally - oriented and -located sources at a distance of 6.68 gpc ( z=1 ) .
we have cross - checked our snr results with those computed by @xcite for initial ligo at 100 mpc , obtaining snr values within 1% of those quoted by them .
a direct comparison with @xcite shows a disagreement of @xmath66 in the computed snrs for advanced ligo and the et , which might be attributed to the different waveforms and cosmological model used .
unsurprisingly , the snrs calculated for the third generation of ground - based detectors surpass the expectations for initial and advanced virgo and ligo at all masses . at @xmath67 , snrs of the order of 0.1 are expected for current ligo and virgo interferometers for binaries with total mass up to a few hundred @xmath68 .
these values would scale to snrs of @xmath69 at the closer distances typically surveyed by the initial interferometers , which are of the order of tens to few hundreds of mpc .
signals with single - interferometer snrs below the commonly - used threshold of snr=5.5 are most likely to be missed , therefore imbh binary coalescences are of limited interest for the first - generation detectors and even for their current , _ enhanced _
counterparts , which are in turn most sensitive to neutron star binaries and stellar - mass back hole binaries .
advanced ligo and virgo will be able to measure averaged snrs of the order of @xmath70 at @xmath67 , with a maximal response to bbh systems with total mass in the range of 300 to @xmath71 .
therefore , should binaries of imbhs exist in our neighboring universe , the second - generation of gw interferometers should be able to detect them , for they will be loud enough to stand above the detector noise . as we will see later
, the non - negligible imbh binary coalescence rate for advanced ligo indicates that these sources should be taken into account in future match - filtered searchers for inspiral binaries . due to the moderate snr of the signals , a potential detection might be sufficient to confirm the existence of binaries of imbhs but not enough to determine all its parameters , such as mass ratio and spins , with sufficient accuracy ; for the advanced detectors , @xcite estimate an accuracy of @xmath72 in the reduced mass @xmath73 using pn templates with negligible spins at snr@xmath12 .
when they take the spins into account , @xmath74 increases by @xmath75 .
it is expected that the use of full imr templates will improve these figures , but the final accuracy will always be limited by the slightly - above - threshold snrs expected in advanced ligo and virgo .
the einstein telescope will instead measure snr values within the @xmath76 range at @xmath77 , and it is expected to be sensitive to binaries with total masses of the order of @xmath8 , a significantly larger range than that surveyed by advanced ligo and virgo .
it is noticeable how the et xylophone configuration increases the detectability of binaries with masses above @xmath78 with respect to the broadband et configuration .
this is due to its improved sensitivity precisely at frequencies in the range of @xmath79 hz , which is where systems of mass above thousands of solar masses accumulate most of their snr ( see figure [ fig : sfvsh ] ) .
the large - snr events that binaries of imbhs would produce in the et are of great importance for astrophysics ; the reason being that as the snr increases , the accuracy in parameter estimation also does . in the limit of large snr ,
the inverse of the fisher information matrix @xmath80 is an estimator of the errors in the recovered parameters . at the relative large snrs potentially produced by imbh binaries in the et
, the possibility of extracting their mass ratios and spins to high accuracies would be revolutionary for characterizing the imbhs populations .
as for lisa , bbhs with masses of hundreds of solar masses will be seen with a moderate snr ( see * ? ? ?
* for a detailed study of the parameter extraction )
it is only at masses above tens of thousands of solar masses that lisa will start taking over the ground - based observatories .
although the space antenna will be most sensitive to bbh binaries with masses in the range of @xmath81 , the possibility that it can act as a complementary observatory for the einstein telescope for imbh binaries is very promising .
parameter accuracy studies for imbhs in lisa are already available using the inspiral part of the coalescence ( including also non - negligible eccentricities ; see * ? ? ?
* ) , and indicate that masses and sky positions will be recovered with a high accuracy level . in order to complete the characterization of imbhs with the information given by the second and third generations of ground - based detectors , a comprehensive study of parameter recovery taking the bbh coalescence into account is very much desirable .
@xcite estimated for the first time the event rate for _ intermediate - mass _ mergers of imbhs ( i.e. typically stellar black holes merging with imbhs ) in clusters by calculating the luminosity distance for the inspiral , merger and ringdown @xcite out to which these three stages can be detected with an snr larger than 10 . in his approach , the maximum distance for the detector was 3 gpc ( @xmath82 ) , with no cosmological corrections .
the event rate was calculated as @xmath83 . in this equation @xmath84
is the number density of globular clusters , which was taken to be @xmath85 , as in the work of @xcite .
the rate of coalescence of stellar - mass compact objects with the imbh is @xmath86 and @xmath87 is the mass distribution of massive enough black holes in clusters .
obviously , @xmath88 .
@xcite uses the estimation of @xcite for it and finds that a few per year should be detectable during the last phase of their inspiral .
two years later , @xcite revisited the problem using matched filtering for the parameter estimation , an updated curve for the sensitivity of the detector and restricted post - newtonian waveforms to calculate an analytical expression for the luminosity distance @xmath89 ; his estimation is a detection rate for binaries of about 1 per myr in a mass - range of @xmath90 .
subsequently , @xcite calculated the number of events that initial and advanced ligo ( and lisa ) could see from the single - cluster channel . in their estimation
, they assume that the vmss formed in the runaway scenario do not merge into one , but evolve separately and eventually each form an individual imbh , following the numerical results of the monte carlo experiments of @xcite .
they derive a generalized form of the event rate which can be summarized as follows @xmath91 in this expression , @xmath92 is the star formation rate ( sfr ) per unit of comoving volume per unit of local time ; @xmath93 is the fraction of mass that goes into the massive clusters of interest ; @xmath94 is the fraction of massive clusters which form imbhs ; @xmath95 is the relation between local and observed time ; @xmath96 is the change of comoving volume with redshift ; and @xmath97 is the distribution function of clusters over individual cluster mass @xmath98 and total star - forming mass in clusters @xmath99 ; @xmath100 is the maximum redshift that the detector is capable of observing . in order to compute the integral above , an estimation of the observable volume of each individual detector
is required .
a commonly - used measure of the reach of a detector is the horizon distance @xmath101 , defined as the distance at which a detector measures an @xmath102 for an optimally - oriented and optimally - located binary , i.e. an overhead , face - on orbit .
suboptimally located and oriented sources are detected with @xmath102 at closer distances . however , in order to compute merger rates , an _
average _ distance over all possible orientations is more meaningful .
the orientation - averaged distance represents the cube root of the total volume to which a detector is sensitive , assuming uniformly - distributed sources , and is 2.26 times smaller than the horizon distance @xmath101 @xcite .
moreover , at the distances that advanced ligo and the et are expected to survey , a certain cosmological model needs to be assumed .
we adopt the standard @xmath103cdm universe with parameters given by the first five years of the wmap sky survey @xcite .
these are @xmath104 , @xmath105 , @xmath106 , @xmath107 and @xmath108 .
using the full imr waveforms described in sec .
[ sec : wfmodel ] and the corresponding redshift function @xmath109 for the @xmath103cdm model , we compute the orientation - averaged distance @xmath110 for non - spinning systems with symmetric mass ratio @xmath111 and for an equal - mass system with spins aligned in the direction of the angular moment and total spin @xmath50 ( hang - up configuration ) . computing @xmath112 is essentially equivalent to inverting eq .
[ eq : snrdef ] for @xmath113 for @xmath114 using the relation @xmath109 given by the cosmological model .
the results for @xmath112 for our three configurations can be seen in fig .
[ fig.hordistplot ] for advanced ligo and et .
the limit for the integral in @xmath115 is thus given by the observable volume of the detector of interest , quantified using the orientation - averaged distance .
the maximum redshift @xmath116 is in turn a function of the mass of the binary system , its configuration mass ratio , spin and the particular waveform model used in the calculation of the horizon distance .
we present our values for @xmath116 in fig .
[ fig : zmax ] , which essentially shows the same data as fig . [ fig.hordistplot ] but expressed in terms of redshift versus total mass of the binary . as we see in fig .
[ fig.hordistplot ] , the maximum values for the orientation - averaged distance for et are as large as @xmath117 for configuration ( ii ) at @xmath118 .
this implies that the et will be able to probe the different proposed scenarios to produce the first generation of black hole seeds , as pointed out by @xcite .
however , at these large cosmological distances , the stellar formation rate is unknown and the validity of the rate integral can not be stated .
we therefore set a maximum cutoff value of @xmath119 in the computation of eq .
[ eq : integralrates ] .
the final value of @xmath120 that we have used in the computation of the rates is show in figure [ fig : zmax ] for our three particular physical configurations .
regarding the term in the @xmath121 integral in [ eq : integralrates ] , three different parameterizations of the sfr are available in the literature , i.e. @xmath122 as summarized by equations 4 , 5 and 6 of @xcite .
the three models are similar for distances up to @xmath123 , where sfr peaks , differing from there on . for the results shown here ,
we have compared the three of them . as for the distribution of cluster masses ,
the factor can be approximated as @xmath124 in order to compute the integral in the denominator , we take @xmath125 , following the power law form observed by @xcite for young star clusters in the antenn .
the validity of assuming the same law for the larger volume of the universe surveyed by advanced ligo or the et is , however , a generalization not based on direct observations .
thus , we should take this premise with care . by assuming an efficiency factor of @xmath126 , based in the results of @xcite
, we can set the values for the limits in @xmath127 according to the masses of the imbh binaries of interest . in our case , taking the standard definition of imbh into account , this is @xmath128 .
the integral can now be expressed as @xmath129 where @xmath130 is the range of total mass of the bbh that we are considering , @xmath131 is any of the three considered stellar formation rates of @xcite and @xmath132 is the factor that relates the stellar formation rate function in different cosmologies with respect to the einstein - de sitter universe .
regarding the values of the terms @xmath94 and @xmath93 , it is unfortunately very little what we know about the initial cluster conditions required to form an imbh binary .
these factors have therefore large uncertainties . following the existing literature
, we leave the fraction of massive clusters that form imbhs , @xmath94 , as a parameter and set it to @xmath133 as an example . nevertheless , as proven in the simulations of @xcite
, it could be _ as large as @xmath134_. as for the fraction of mass going into massive clusters , @xmath93 , previous works have also tentatively set it to @xmath133 .
nevertheless , the results of @xcite seem to indicate that this value might be too optimistic .
@xcite estimates an empirical cluster formation efficiency by mass of @xmath135 , where @xmath136 is the initial globular cluster population and @xmath137 is the initial reservoir available in the protogalaxy .
he then infers a universal value of @xmath138 after evaluating a sample of 97 giant ellipticals , brightest cluster galaxies and faint dwarfs .
he finds identical results for the population ii spheroid of the milky way .
therefore we have chosen to set our @xmath93 to @xmath139 , noting that , in any case , the final rates scale trivially with both @xmath94 and @xmath93 . we have evaluated the integral [ eq : integralrates ] for the three stellar formation rates we consider and the three binary configurations ( i ) , ( ii ) and ( iii ) for advanced ligo and the et .
our results are summarized in table [ tab : eventrates ] .
we find that the event rate does not depend strongly on the assumed stellar formation rate ; the differences are of the order of @xmath140 for advanced ligo and @xmath141 for the et . in the case of advanced ligo , the insensitivity to the stellar formation model is easily explained by the fact that the three models of @xcite are very similar until @xmath142 ; our integration in @xmath115 stops at @xmath143 ( see red curves in fig .
[ fig : zmax ] ) , therefore ignoring contributions at higher @xmath115 .
for the et , the differences among @xmath131 between @xmath123 and 5 are very much attenuated by the rapid decrease in comoving volume at high redshift . in order to assess the effect of stopping the integration at @xmath144 , we have also computed the rates for the et without imposing this condition . the differences amount to @xmath145
, therefore we conclude that the rates estimations are not very sensitive to very high-@xmath115 sfrs .
nonetheless , we should stress again that the lack of stellar formation data at these large cosmological distances makes assumptions at @xmath146 largely speculative and thus we restrict to presenting the rates calculated with @xmath116 clamped at a value of 5 .
given the inherent uncertainties present in our approach , which can amount to a few orders of magnitude , the factors associated to the choice of sfr do not represent a major source of error .
we therefore quote the results found for @xmath147 only , corresponding to the star formation rate that increases up to @xmath148 and keeps constant afterwards .
.event rates formed in the single - cluster channel for imbh binaries potentially observable by advanced ligo and the et per year .
we take @xmath149 , @xmath150 as standard scaling values[tab : eventrates ] [ cols="^,^,^,^",options="header " , ] the event rates do , however , strongly depend on the spins and mass ratio of the binary . as expected , `` loud '' configurations like the hang - up case increase the event rate by a factor of @xmath151 in the case of advanced ligo .
this effect is more notable the larger the total spin of the binary is ; the correct determination of the spin distribution of imbh binary systems would be extremely useful to further quantify the impact on the total rates .
smaller mass ratios decrease the event rates , the difference between @xmath152 and @xmath153 being also approximately of a factor of 3 .
the differences are not so extreme in the case of the et , due to the fact that we are cutting off @xmath116 at a value of @xmath154 and , thus , neglecting contributions at higher redshift which might amount to another @xmath155 , as discussed above .
even so , it is evident that the expected rates increase with the total spin @xmath156 and the symmetric mass ratio @xmath43 .
the exact computation of the total rates for all possible physical configurations of the binary would necessitate further integration on the mass ratios and spins of the system .
at present it is not clear what that distribution might be , therefore we simply summarize our results for all configurations under consideration in table [ tab : eventrates ] .
we find event rates of the order of @xmath157 for advanced ligo and of the order of @xmath158 for the et .
these rates assume formation of imbh binaries in the single - cluster channel .
for advanced ligo and a non - spinning configuration , @xcite found a expected detection rate of @xmath159 ( @xmath160 in our units ) , adopting a reach of 2 gpc ( @xmath161 ) .
this estimation , calculated assuming a ringdown - only search , underestimates the volume of the universe that advanced ligo would observe using full imr templates , as we show in fig .
[ fig.hordistplot ] .
our results for non - spinning systems are compatible with this observation ; we quote rates larger than those of @xcite by a factor of 3 but otherwise of a compatible order of magnitude . in the case of the et ,
@xcite quote a rate of @xmath162 ( @xmath163 in our units ) for non - spinning configurations , a value of the same order of magnitude than the one found by us .
the fact that @xcite use a different family of waveforms to model the coalescence and a fitted formula for the the averaged distance based on eobnr waveforms , together with the slightly different values in the integration limits explain the discrepancies in the exact figures .
therefore , our new results for the rates of imbh binary coalescence for non - spinning systems agree reasonably well with previous works that used similar methods but different waveform and detector models . in addition , we have now quantified the effect that the spins of the bhs will have on the expected rates and have found it to be of a factor @xmath151 for total spins of the binary as high as @xmath164 thus far we have concentrated on the single - cluster channel scenario . @xcite
give a prescription to calculate an estimate of the event rates for the double - cluster channel .
this was based in the fact that the only difference , in terms of the event calculation , between both astrophysical scenarios involves , firstly , that in the double - cluster channel there is one single imbh in one cluster and , secondly , that these two clusters must collide so that the imbhs form a bbh when they sink to the center due to dynamical friction . as explained in section 4 of @xcite ,
the connection between the event rate estimation of the two channels is @xmath165 where @xmath166 is the event rate of the double - cluster channel , @xmath167 of the single - channel and @xmath168 is the probability for two clusters to collide in the scenario of @xcite .
they find that the unknown parameter @xmath168 could have values in the range @xmath169 $ ] . the total event rate assuming that both formation channels are possible would be @xmath170 assigning parameter @xmath168 its pessimistic and optimistic limits , we can compute the lower and upper limit of the event rates for advanced ligo and the et , assuming contributions of the two channels . taking into account
the double - cluster channel increases the rates upper limit by a factor of 2 .
for instance , for the equal - mass , non - spinning case , the values are @xmath171\,\left(\frac{g}{0.1}\right ) \left(\frac{g_{\rm cl}}{1/400}\right)\,{\rm yr}^{-1}\\ \gamma_{\rm et}^{\rm total } & \in [ ( 0)\,23.7,\,43.0]\,\left(\frac{g}{0.1}\right ) \left(\frac{g_{\rm cl}}{1/400}\right)\,{\rm yr}^{-1}\end{aligned}\ ] ] note that the results are quoted in terms of @xmath94 and @xmath93 ; the already - mentioned uncertainty in their values could increase these rates by at least one order of magnitude .
these event rates are encouraging to address the problem of detection and characterization of systems of imbh binaries with ground - based gw observatories . on the other hand
, one should bear in mind that the existence of imbhs altogether has not yet been corroborated , so that the pessimistic estimate is still somewhat `` optimistic '' .
this is why we have added a @xmath172 in the previous rates as the absolute lower limit .
even though we do not have any evidence of imbhs so far , a number of theoretical works have addressed their formation in dense stellar clusters .
if we were to follow the same techniques that have led us to discover the smbh in our own galaxy , we would need the very large telescope interferometer and next - generation instruments , such as the vsi or gravity , which should be operative in the next @xmath12 yrs . an alternative or even complementary way of discovering imbhs is via their emission of gws when they are in a bbh system .
the identification and characterization of these systems rely on accurate waveform modeling of their gw emission , which has been made possible by the success of numerical relativity in simulating the last orbits of the bbh coalescence and the coupling of these results to analytical post - newtonian calculations of the inspiral phase .
we use a pn - nr hybrid waveform model of the bbh coalescence based on a construction procedure in the frequency domain ( see * ? ? ?
* for details ) .
using this hybrid waveform , we have estimated the snr corresponding to the current and advanced ligo and virgo detectors , the proposed et and the space - based lisa at a distance of @xmath67 , i.e. 6.68 gpc .
the results indicate that imbh binaries will produce snrs sufficient for detection in advanced ligo and virgo and notably larger snrs in the et , thus making them interesting sources to follow up on .
eventual observations of imbh binaries with future ground - based detectors could be complementary to those of lisa , which is expected to detect these systems with moderate snrs and to be more sensitive to smbh binaries instead .
more remarkably , in principle , if lisa and the et are operative at the same time , they could complement each other and be used to track a particular event .
furthermore , we have revisited the event rate of bbhs for various detectors and find encouraging results , within the inherent uncertainties of the approach .
our estimations are consistent with previous works , and additionally we have quantified the effect of the total spin of the binary in the expected event rates .
we have estimated the distance to which advanced ligo and the et will be able to see binaries of imbhs .
this quantity depends strongly on the mass ratio and spins of the binary . for advanced ligo , equal - mass , non
- spinning configurations of observed total mass @xmath173@xmath174 can be seen up to @xmath175 . if the spins are aligned with the total angular momentum and significant ( @xmath176 ) , the reach increases to @xmath177 for observed total masses of @xmath178
. the et will be able to explore even more remote distances , reaching to @xmath179 and further .
our present knowledge of stellar formation at such large redshifts is incomplete , therefore we have computed the event rates for the et integrating only until @xmath144 .
we have compared three star formation models and three different configurations of the binary .
the effect of the particular formation model is not very significant ; remarkably , the physical configuration does however influence the final rates .
we provide the results for three physical configurations , taking into account both single- and double - cluster channels in the binary formation . for a fully correct calculation of the event rate integrated over all possible configurations , more detailed knowledge of the distribution of spins and mass ratios of imbh binaries formed in globular cluster
is required .
advanced ground - based detectors are designed to be able to operate in different modes so that their sensitivity can be tuned to various kinds of astrophysical objects .
considering the importance of an eventual detection of a bbh , the design of an optimized advanced ligo configuration for systems with @xmath180 would be desirable to increase the possibility of observing such a system . in case an imbh binary coalescence were detected , the recovery and study of the physical parameters of the system could serve to test general relativity and prove or reject alternative theories , such as scalar - tensor type or massive graviton theories .
direct _ identification of an imbh with gws will be a revolutionary event not only due to the uncertainty that surrounds their existence and their potential role in testing general relativity .
the information encoded in the detection will provide us with a detailed description of the environment of the bbh / imbh itself .
we are indebted to c. cutler , b. sathyaprakash and b. krishnan for discussions about the work .
we thank c. reisswig and d. pollney for providing us with their cauchy - extrapolated data and f. ohme for helping us to implement the post - newtonian formalism used for the waveform modeling .
we are grateful to e. robinson and d. vanecek for comments on the manuscript .
pas work was partially supported by the dlr ( deutsches zentrum fr luft- und raumfahrt ) .
ls has been partially supported by daad grant a/06/12630 and thanks the imprs for gravitational wave astronomy for support . | even though the existence of intermediate - mass black holes ( imbhs , black holes with masses ranging between @xmath0 ) has not yet been corroborated observationally , these objects are of high interest for astrophysics .
our understanding of the formation and evolution of supermassive black holes ( smbhs ) , as well as galaxy evolution modeling and cosmography would dramatically change if an imbh were to be observed . from a point of view of traditional photon - based astronomy , which relies on the monitoring of innermost stellar kinematics , the _ direct _ detection of an imbh seems to be rather far in the future .
however , the prospect of the detection and characterization of an imbh has good chances in lower - frequency gravitational - wave ( gw ) astrophysics using ground - based detectors such as ligo , virgo and the future einstein telescope ( et ) .
we present an analysis of the signal of a system of a binary of imbhs ( bbh from now onwards ) based on a waveform model obtained with numerical relativity simulations coupled with post - newtonian calculations at the highest available order .
imbh binaries with total masses between @xmath1 would produce significant signal - to - noise ratios ( snrs ) in advanced ligo and virgo and the et .
we have computed the expected event rate of imbh binary coalescences for different configurations of the binary , finding interesting values that depend on the spin of the imbhs .
the prospects for imbh detection and characterization with ground - based gw observatories would not only provide us with a robust test of general relativity , but would also corroborate the existence of these systems .
such detections should allow astrophysicists to probe the stellar environments of imbhs and their formation processes . |
flares on the sun result from catastrophic releases of energy previously stored in strong tangled surface magnetic fields .
these outbursts are most conspicuous at high energies , and under the right circumstances ( or wrong , depending on your perspective ) can profoundly impact the heliosphere and earth .
this is one reason why flares have drawn intense interest from solar observers and theorists for many decades ; and indeed have spawned a vibrant space weather community in recent years .
a less appreciated aspect of solar flares involves the stellar cousins , where extraordinarily large x - ray ( and even white - light ) outbursts have been recorded , in some cases dwarfing by a considerable margin the largest known flares on our own sun , albeit a middle - aged and magnetically rather uninspired star .
the most extreme of these are `` superflares , '' which at high energies can exceed the quiescent coronal x - ray emission of a star by factors of a hundred or more .
interest in such rare , but potentially catastrophic , events concerns how they might affect , for example , survival of primitive lifeforms struggling for existence on a semi - habitable world orbiting a young , magnetically active star . until recently
, reports of superflares mainly involved interpretations of historical materials , such as long - term astronomical plate collections , and disconnected isolated reports or inferences ( schaefer et al .
now , however , the _ kepler _ mission has provided more substantive statistics on superflares although limited to white - light counterparts from a wide variety of stars , normal and otherwise ( e.g. , maehara et al . 2012 ;
shibayama et al .
2013 ; candelaresi et al .
further , the _ swift _ gamma - ray burst satellite has witnessed a handful of the most extreme events , radiating in hard x - rays , thanks to its broad view of the high - energy sky and rapid pointing response ( e.g. , osten et al .
2010 ; drake et al .
the _ swift _ alerts also have promoted supporting multi - spectral observations of flares , especially at radio frequencies that capture non - thermal emissions from accelerated particles ( e.g. , fender et al .
2015 ) . on the other hand
, there have been few cases of such outbursts recorded at wavelengths where the powerful diagnostics of high - resolution spectroscopy can be brought to bear , namely the ultraviolet .
the lack of good examples mostly is because these rare events are , well , rare .
serendipitously , however , a recent _ hst _ cosmic origins spectrograph ( cos ) campaign on young ( @xmath250 myr ) solar analog ek draconis ( hd129333 ; g1.5 v ) , well - described in the previous literature ( e.g. , knig et al .
2005 , and references therein ) , caught a large hour - long fuv outburst , in `` hot lines '' like the 1550 doublet ( t@xmath6 k ) and the extremely hot [ ] 1354 coronal forbidden line ( @xmath7 k ) .
the ek dra flare , viewed with the excellent spectral resolution , broad wavelength coverage , and high sensitivity of cos , provides a unique opportunity to explore properties of such events , to inform interpretations of the physical processes involved .
moreover , the behavior of the fuv spectrum outside the flare , over the several days of intermittent attention by _ hst , _ opens a window onto the `` quiescent '' properties of a hyperactive chromosphere and corona , again with unprecedented clarity .
there were two previous _ hst _ spectroscopic programs of note on ek dra . in 1996 ,
saar & bookbinder ( 1998 ) carried out high time resolution measurements with the low - resolution fuv grating of goddard high - resolution spectrograph ( ghrs ) over a continuous interval of about six hours ( about 10% of the spin period ) .
the authors identified several factor of 2 - 3 impulsive flare - like bursts in 1400 and 1550 when the light curves were binned into 30 s intervals .
the identifiable hot - line emission enhancements occupied about 8% of the timeline .
the remainder was attributed to `` quiescent '' periods . a decade and a half later , in 2010 , ayres & france ( 2010 ) targeted ek dra with newly commissioned cos as part of a larger `` snapshot '' ( partial - orbit filler observations ) survey of fuv coronal forbidden lines , such as [ ] @xmath81354 ( @xmath9 k ) , in more than a dozen late - type stars . the objective was to measure doppler profiles of high - energy coronal lines at fuv resolution , some 20 times better than currently possible in the soft x - ray band , the more natural but spectroscopically challenged
home of coronal transitions .
the cos snap observation of ek dra covered only a short time span , about 20 minutes , compared with the earlier ghrs pointing .
but , it had the important advantage of achieving moderately high spectral resolution ( @xmath10 km s@xmath1 ) in the integrated spectrum , as well as comparably high time resolution ( 30 s ) for flux measurements of key species like 1335 ( @xmath11 k ) , , and [ ] collected in the single g130 m grating setting feasible for the brief partial - orbit visit . the doublet , in particular , exhibited twin factor of @xmath22 enhancements over the mere 20 minute interval , remarkably without comparable responses at lower temperatures ( ) or higher ( ) , contrary to expectations for normal flares , which tend to impact all the atmospheric temperature layers more - or - less simultaneously .
equally remarkable , the doublet features displayed quite large _
redshifts _ , about 20 km s@xmath1 , relative to narrow chromospheric features ( which set the empirical zero - point velocity for the cos exposures )
. redshifts of the hot lines are well documented on the sun , attributed to nearly radial downflows of subcoronal material , but the average shifts in a disk - integrated spectrum might be only a few km s@xmath1 or less ( peter 2006 ) .
ayres & france interpreted the large redshifts of on ek dra as a signature of a scaled up version of solar `` coronal rain : '' previously hot multi - mk gas suspended high in magnetic loops rapidly cooling and draining back down to the lower atmosphere , perhaps in the case of ek dra in a catastrophic way ( what one might call an _ anti - flare : _ impulsive cooling rather than heating ) .
the follow - up _ hst _ program , described here , partnered the high sensitivity of cos with the excellent wavelength precision of stis to explore alternative explanations for the cos redshifts , and thus also the coronal rain hypothesis : ( 1 ) subtle distortions in the cos wavelength scales ; and ( 2 ) `` doppler imaging '' effects , whereby active patches on the stellar surface might appear in the integrated spectrum as a redshift if , say , a bright region happened to be near the receding limb ( where the projected rotational velocity could be as much as @xmath12 km s@xmath1 ) .
a secondary objective was to carry out a more extensive time - domain study of fuv variability , along the lines of the saar & bookbinder work , but with the added dimension of time - resolved spectroscopy possible with high - sensitivity , time - tagged cos .
a related goal was to extend the previous minimal snap wavelength coverage ( g130m / detector side a : 1290 - 1430 ) , adding g160m / sides a and b , especially to reach the important 1550 doublet and @xmath81640 at the longer wavelengths ; and with stis to probe below the cos g130m / a shortwavelength cutoff ( side b was deactivated because ly@xmath5 @xmath81215 violated detector bright limits ) , including especially ly@xmath5 itself .
the characteristics and operational capabilities of cos have been described in a number of previous publications , especially green et al .
( 2012 ) ; and companion uv spectrograph stis , by woodgate et al .
( 1998 ) and kimble et al .
( 1998 ) .
the new _ hst _ program was carried out over two days in late - march 2012 ( the stellar rotation period is 2.8 d ) .
ek dra was in the _ hst _ continuous viewing zone ( cvz ) at the time , by design , which allowed each of the six allocated 96 minute orbits to be used fully , without any loss to earth occultations .
figure 1 is a schematic timeline of the program , also summarized in table 1 .
-13 mm -5 mm cccccc lb3e34010 & apr-22 09:18 & 1160 & psa & g130m-1291 & 3,4 + oboq01010 & mar-27 16:41 & 8880 & pht & e140m-1425 & + lboq02010 & mar-27 19:57 & 1000 & psa & g160m-1577 & 3,4,1,2 + lboq02020 & mar-27 20:41 & 1780 & psa & g130m-1291 & 3,4,1,2 + lboq03010 & mar-28 05:43 & 1000 & psa & g160m-1577 & 3,4,1,2 + lboq03020 & mar-28 06:48 & 1780 & psa & g130m-1291 & 3,4,1,2 + lboq04010 & mar-28 20:33 & 1000 & psa & g160m-1577 & 3,4,1,2 + lboq04020 & mar-28 21:16 & 1780 & psa & g130m-1291 & 3,4,1,2 + lboq05010 & mar-29 10:25 & 1000 & psa & g160m-1577 & 3,4,1,2 + lboq05020 & mar-29 11:08 & 1780 & psa & g130m-1291 & 3,4,1,2 + ocd839010 & may-06 10:10 & 600 & pht & e230h-2713 & the first visit , consisting of two orbits , featured a deep stis exposure with the medium - resolution fuv echelle ( e140m-1425 ) for 8.8 ks , split into four equal - length sub - exposures ( to mitigate instrumental drifts due to telescope `` breathing '' ) .
the initial target acquisition was with the ccd in visible light through the f25nd3 filter .
following that , a peak - up ( accurate target centering via a raster search ) was performed with the @xmath13 slit in dispersed visible light ( ccd with g430m-3936 grating setting ) . then , the stis fuv exposures were taken , through the @xmath14 photometric aperture for maximum throughput , although in principle retaining high accuracy in the wavelength scales thanks to the narrow - slit peak - up . immediately following the stis segment , although in a separate visit , was the first of four one - orbit cos pointings , spaced evenly over roughly half a spin cycle so all sides of the possibly inhomogeneous surface activity would be visible .
the four cos visits ostensibly were identical and occupied a single cvz orbit each .
the cos visits began with an nuv imaging target acquisition , using mirror - a and the bright object aperture ( boa ) . following that , a 1000 s observation was taken with g160m-1577 ( 1387 - 1557 on detector side b ; 1578 - 1749 on side a ) , divided into equal length sub - exposures , one at each of the standard four `` fp - pos '' positions ( small grating rotations to suppress fixed - pattern noise ) .
then , a 1780 s integration , also split into four equal fp - pos sub - exposures , was taken with g130m-1291 ( 1292 - 1432 on side a ; side b turned off , as in the earlier snap program , owing to overbright ly@xmath5 ) .
all the cos observations were done in time - tag mode , which delivers a time - stamped record of photon @xmath15 positions in the 2-d detector format ( @xmath16 corresponds to wavelength ; @xmath17 to cross - dispersion spatial coordinate ) .
these photon event lists can be processed in a number of ways to provide time histories of fluxes and/or line profiles .
the particular g130 m and g160 m settings were chosen because they overlap at the important 1400 doublet , so that it would be recorded in all the observations .
unfortunately , the key [ ] coronal forbidden line was captured only in the g130 m segments ; and the important subcoronal doublet only in the g160ms .
the initial stis observation and first of the cos visits were intended to be carried out as a pair , as close together as practical , to serve as a velocity cross - validation .
although cos exposures have an internal wavelength reference ( lamp flash ) , the calibration spectrum falls well above the stellar stripe , on a part of the detector where uncompensated geometrical distortions are different than where the stellar spectrum is located .
further , the specific @xmath17-position of the stellar spectrum , and thus the local geometrical distortions it experiences , depends on the acquisition method .
however , because the same strategy was used for all four cos visits , any wavelength issues identified in the comparison of the initial cos visit with stis should apply equally well to the three subsequent cos pointings . at the same time , while stis has exquisite velocity precision and can record s / n@xmath220 profiles of the brightest lines ( including ly@xmath5 , forbidden to cos ) , it lacks the sensitivity to resolve shapes of the bright lines on short time scales , or dig down to fainter features like the key ] 1400 density diagnostic ( see , e.g. , cook et al .
1995 , and references therein ) .
the stis spectra of ek dra were processed and combined using protocols developed for the _ hst _ advanced spectral library project ( astralayres / astral/ ] ) , which includes a post - pipeline correction for subtle wavelength distortions .
the four independent sub - exposures were co - added , but without the usual cross - correlation registration , because the relatively faint spectra lacked narrow features of suitable strength for the purpose .
figure 2 compares representative intervals of the average stis fuv echelle spectrum of ek dra to low - activity solar twin @xmath5 centauri a ( g2 v ) , also observed by stis with the same setup ( see ayres 2015 ) . at first glance ,
the two spectra are surprisingly alike , despite the enormous gulf between the stars in coronal activity ( the @xmath18 of ek dra is about a thousand times that of @xmath5 cen ) .
for example , the narrow chromospheric lines of , , and are quite similar in appearance between the hyperactive star and low - activity counterpart . even the hot lines like have similar emission cores .
-5 mm closer examination , however , reveals several differences .
the most noticeable is the exceptionally broad ly@xmath5 profile of ek dra ( although somewhat emphasized by the normalization approach : more distant ek dra [ @xmath19 pc ] is more heavily absorbed by interstellar than close - by @xmath5 cen [ 1.3 pc ] , so is missing much of its intrinsic narrow peak ) . a second ,
in fact blatant , difference is the bright [ ] coronal forbidden line of ek dra , which is absent in low - activity @xmath5 cen ( the weak , narrow feature at 1354 is a chromospheric line ) . on the other hand , the striking difference at the @xmath81640 balmer-@xmath5 line mainly comes about because low - activity @xmath5 cen is dominated by a chromospheric blend on the blueward side of ; rather than high - excitation itself as in ek dra .
the prominent photospheric continuum of @xmath5 cen a in the @xmath81640 panel is an artifact of the scaling procedure : the line - to - continuum contrast is much smaller in the low - activity star compared with hyperactive ek dra . if the normalization instead had been according to the bolometric fluxes , the 1640 photospheric continuum levels of the two stars would be similar , but the emission of ek dra would dwarf that of @xmath5 cen a. ( indeed , a similar dwarfing would occur in all the panels with an @xmath20 normalization . ) more subtle , but significant , differences are the broad low - intensity wings of the ek dra hot lines : , , , and ( @xmath21 k ) .
there are some additional differences seen in the 1353.7 panel , having to do with pumping of the and lines by higher excitation species , leaving @xmath81355 much weaker by comparison in the active star .
-10 mm figure 3 is a schematic of the cos fuv spectrum covered by the 2012 program .
the lamp - flash calibrations ( barely visible at this intensity stretch ) can be seen 100 pixels above the main stellar stripe .
the two g160 m segments were observed simultaneously .
the longwavelength end of g130 m on detector side a overlaps the shortwavelength portion of g160 m on side b at the key 1400 resonance doublet ( and ] density - diagnostic multiplet ) ; side b of g130 m was disallowed , as in the 2010 snap program , owing to intense chromospheric ly@xmath5 .
the broad features ( in cross - dispersion @xmath17 coordinate ) at 1300 are the resonance triplet , dominated by sky - glow through the 2.5@xmath22-diameter cos aperture .
the other bright features are @xmath81335 , @xmath81400 ( distinct pair ) , and @xmath81550 ( close pair ) .
lccccc line-1 & @xmath23 & 20 & 1335.703 & 0.05 & 3000 + line-2 & @xmath24 & 10 & 1354.080 & 0.10 & 1000 + line-3 & @xmath25+@xmath26 & 20 & 1402.773 & 0.05 & 1000 + con - a & @xmath27+@xmath28 & 20 & & & + bkg - a & @xmath29 & & & & + line-1 & @xmath25+@xmath26 & 20 & 1402.773 & 0.05 & 1000 + line-2 & @xmath30 & 20 & 1548.204 & 0.05 & 2000 + line-3 & @xmath31 & 20 & 1640.404 & 0.05 & 1000 + con - b & @xmath32 & 20 & & & + bkg - b & @xmath33 & & & & + con - a & @xmath34 & 20 & & & + bkg - a & @xmath35 & & & & as mentioned earlier , the initial stis and cos visits were intended to be as close as possible to enable a cross - comparison of the two uv instruments .
thanks to diligent work by the _ hst _ schedulers , the first cos exposure ( g160 m ) began a mere 40 minutes after the final stis readout , normally close enough to ensure comparable stellar conditions over the three orbits of contiguous stis and cos attention .
indeed , the stis fuv emission line intensities were stable over the 2.5 hours of near - continuous e140 m exposure .
as luck would have it , however , less than an hour later at the beginning of the first cos g160 m sequence , ek dra apparently was in outburst , experiencing a giant fuv flare .
the star had brightened up in by an order of magnitude , nearly triggering the cos detector bright limits , a remarkable performance for a nearly 8th - magnitude g dwarf .
figure 4a illustrates time - resolved line fluxes and profiles from the first cos observation ( visit 2 ) , with grating g160 m .
the top panel displays the total spectral image ( all four fp - pos event lists combined ) , with specific integration bandpasses marked for the doublet ( blue ) , ( red ) , and ( green ) ; as well as two line - free continuum bands ( `` b : '' orange dashed ; `` a : '' orange / black dashed ) ; and two pairs of flanking background swaths ( black dashed , and black - yellow dashed ) , one for each detector segment .
the parameters defining these measurements , and the subsequently described spectral profile extractions , are listed in table 2 . the middle panel of fig .
4a compares the time - resolved line and continuum count rates , with the same color - coding .
the individual points represent counts accumulated in sliding time bins to achieve a specified signal - to - noise ratio ( s / n)@xmath36 : see table 2 .
the two continuum bands ( orange , and black outlined orange , circles ) were scaled down a factor of 4 for display purposes , while the background count rates ( black and yellow dots ; taken from each full fp - pos time interval ) were scaled upward by 100 times relative to the continuum bands ( taking into account the continuum integration wavelengths and cross - dispersion width : see table 2 ) to illustrate the possible influence of the background for the worst - case scenario ( broad wavelength bands ) .
the background corrections for the continuum bands in this instance are negligible , but still were compensated .
the background contributions for the tiny line spectral footprints are entirely negligible , and were not corrected .
the lower panels illustrate spectrally resolved profiles of three bright features
@xmath81402 , @xmath81640 , and @xmath81548 each accumulated in time intervals corresponding to a fixed number of counts ( @xmath37 : see table 2 ) .
@xmath81402 was chosen over stronger @xmath81393 , because the latter is affected by bad pixels in one of the fp - pos segments of g160m / b .
the specific accumulation intervals are highlighted by small ticks in the middle panel time sequences . in each of the bottom spectral panels ,
the time - resolved profiles are the thin colored curves .
the background yellow shaded envelope represents a profile of the same transition from a spectrum averaged over visits 3 - 5 ( see below ) , scaled to the maximum and minimum of the collection of line peaks for the particular visit .
the upper and lower outlines of the envelopes are black dotted .
remarkably , despite the large flux changes during the post - flare decline caught by visit 2 , the profiles of all three features are nearly identical to the scaled averages ; showing the characteristic redshift of the line core , and the high - velocity extended wings , especially on the redward side of line center .
figure 4b is similar , for the g130 m side a spectrum in the second part of visit 2 . here ,
the integrated count rates and extracted profiles are for , [ ] , and .
the two continuum bands ( now smaller , to avoid the more numerous chromospheric emissions at these wavelengths ) were combined into a single continuum flux , scaled down a factor of two in the figure for display .
the background count rates , again adjusted to the total @xmath38 continuum bandpass , were scaled upward a factor of 25 relative to the displayed continuum values .
all the fluxes continue to decline during this interval , but again the time - resolved line profiles are self - similar to the visits 3 - 5 averages .
figure 5a illustrates the g160 m sequence of visit 3 , about 10 hours after the initial flare segment .
the continuum and background count rates are scaled as in fig .
all the fluxes , especially the continuum bands , have fallen compared with visit 2 , and now the background corrections are more significant .
the line profiles span a more compressed range , essentially identical to the visits 3 - 5 average .
figure 5b depicts the g130 m exposures of visit 3 .
the continuum and background fluxes are scaled as in fig .
again , there is little variability in the integrated fluxes or the time - resolved profiles .
the analogous diagrams for later visits 4 and 5 are very similar to those of visit 3 , and are not illustrated .
figure 6 is an expanded view of the flare in , which was recorded in both cos grating settings . as in figs .
4 and 5 , the symbols represent sliding time bins corresponding to ( s / n)@xmath39 .
accordingly , the bins become wider in time , and the points more separated , as the count rates decline during the later stages of the event .
the cos values were normalized to the average of visits 3 - 5 , which were much quieter than the visit 2 flare . the g160m / b and g130m / a time series were separately normalized to the corresponding quiescent averages , to avoid possible differences in sensitivity with the different grating settings on the different detector segments .
the stis points from visit 1 , immediately prior to the flare rise , were based on calibrated fluxes integrated over the doublet components in each of the four independent e140 m sub - exposures , divided by the integrated flux , measured in the same way , from an average cos spectrum constructed from the quiescent visits 3 - 5 ( as detailed later ) .
the flare evolution was modeled by twin exponentials : a rapid initial decline with an e - folding time constant of 7 minutes , on top of a slower decay with a 1.9 hour time scale .
such bimodal decays are common among stellar x - ray flares , for reasons that are not completely understood ( see examples and discussion by tsang et al .
the lower panel of fig .
6 provides a broader view of the behavior during the entire _ hst _ program , showing explicitly the relatively quiet intervals prior to and following the flare event .
-2 mm figure 7 is a multi - spectral view of the ek dra flare decay , and subsequent quiescent intervals .
error bars to the left of the vertical dotted line indicate the stis average levels and standard deviations ( over the four sub - exposures ) .
the cos time series to the right are the same as illustrated in figs . 4 and 5 , but now including visits 4 and 5 , and normalized to the visits 3 - 5 average .
note the small flare decay that affects all the features in the second part of visit 3 , together with a number of and transients like those seen in the saar & bookbinder time series and the 2010 snap program .
also note the two bursts ( at @xmath403.6 hr and 3.7 hr ) in the g160 m blue and red continuum bands ( marked by the corresponding colors ) , which remarkably are not repeated in any of the other activity tracers .
this phenomenon is discussed in more detail later ( 3.4 ) .
-15 mm although the large fuv flare seemingly ruined the carefully orchestrated stis - cos velocity cross - calibration , the bright side is that this was the first such significant event on a sunlike star to have been captured with the un - matched spectroscopic and time - resolution capabilities of an instrument like cos ; a unique opportunity to explore properties of the outburst beyond the usual simple intensity evolution ( with modest temperature discrimination ) delivered by , say , an x - ray telescope . and , as described above , following the large flare ek dra settled into an period of apparent quiescence to the extent that a star with a thousand times the x - ray luminosity of the sun can be considered `` quiet '' for the next three cos visits ( spanning a day and a half , corresponding to about half a rotation period ) .
the similarity of the time - averaged fuv spectra of these three visits , both among themselves and compared with the initial stis pointing , ultimately provided the desired velocity cross - validation ( as described later in 3.1 ) .
so , all was not lost .
furthermore , the lack of significant fuv profile variability of the hot lines between stis visit 1 and cos visits 3 - 5 , implies that `` doppler imaging '' was not an important factor during the 2012 _ hst _ campaign .
in fact , the constancy of the hot - line fluxes over the more than half a rotational cycle also indicates that the surface distribution of fuv - emitting regions was relatively uniform , at least in that epoch . at higher energies , in contrast , gdel et al .
( 1995 ) reported factor of @xmath22 systematic variations in _
rosat _ all - sky survey x - ray fluxes , in step with the rotation period , and a suggestion of correlated variability at microwave frequencies as well . on the sun ,
albeit perhaps not the best guide to hyperactive ek dra , the fuv - bright plage regions surrounding sunspot groups are much more spatially pervasive than the dark spots themselves , and more so than the compact coronal magnetic loop systems associated with the active region as well .
ek dra is reported to have a faint , m - type companion in a highly eccentric 45 year orbit ( knig et al .
based on the authors ephemeris , the secondary would have been just past maximum elongation in 2012 , nearly due south of the primary with a separation of @xmath41 .
this object would be at the inner edge of the cos psa , and potentially visible as a slightly displaced spectrum ( unfortunately on top of the main spectrum in the @xmath42 direction , given the position angle of the cos focal plane at the time of the 2012 program ) , but entirely outside the stis @xmath14 slot . however , there is no obvious line doubling due to a second spectrum visible in any of the cos fuv grating images ; although the predicted separation between the two is only + 0.35 ( ek dra b to red of a ) , and the bright primary lines are broad .
the secondary also was not clearly visible in ( stacked versions ) of either the stis or cos acquisition images of the 2012 campaign .
serendipitously , however , the faint companion was recovered in a subsequent stis pointing on ek dra , in an 2014 ism snap program ( s. redfield p.i .
; see table 1 ) .
the partial orbit visit had a more conservative ccd acquisition : a total exposure time of 2.3 s through the f25nd3 aperture , significantly deeper than the 0.3 s of the 2012 program with the same filter .
the secondary was about 1% of the primary s brightness , at a position angle of 172@xmath43 and separation of 0.75@xmath22 , just as predicted by the ephemeris of knig et al .
the @xmath44 magnitude of the secondary would be about + 13.5 , taking into account that the broad - band ccd response favors red stars over yellow stars .
the primary / secondary @xmath45 is what was suggested by knig et al .
this is important , because a youthful red dwarf could be quite active , and its hot lines , if bright enough , potentially could affect the profiles of the primary features ( with the unfortunate location of b on top of the a spectrum in 2012 ) .
however , scaling fuv fluxes of the most active dme flare stars , like au microscopii and ad leonis , to the estimated @xmath44 magnitude of the secondary predicts peak flux densities , say at , of only a few percent of the primary s , which would be of little concern . a negligible impact by the red star , at least on the quiescent fuv spectrum of ek dra , is supported by the fact that the cos visits 3 - 5 spectra are so similar to stis visit 1 , which entirely excluded the secondary by virtue of the small stis aperture . as for the large fuv flare ,
if it were from the red star , the associated soft x - ray luminosity ( see below ) would have been quasi - bolometric ( i.e. , @xmath46 ) ; not impossible , but very unlikely .
the @xmath22 known examples , such as the quasi - bolometric flare on dg cvn reported by drake et al .
( 2014 ) , have come from the _ swift _ burst alert telescope , and soft - x - ray follow - up by that satellite .
the survey burst alert situation is much more likely to capture extremely rare events because it is viewing a whole sky full of objects , all the time , rather than pointed at a single target for a limited period .
furthermore , the enhanced narrow peaks of the fuv hot lines during the flare were nearly identical in shape and velocity to the quiescent profiles of the g star seen outside the outburst ; rather than shifted by the @xmath2 + 75 km s@xmath1 to the red expected if the fuv lines were instead from the secondary star .
the preponderance of evidence is that the large fuv flare was from the g - type primary .
a main objective of the 2012 ek dra campaign was to test the reality of the large @xmath10 km s@xmath1 doppler shifts of recorded in the brief cos snapshot two years earlier .
the core of that test was to validate the cos wavelengths against the highly precise scales of stis .
in particular , the e140m-1425 echelle setting captures the entire fuv region ( 1150 - 1700 ) in a single shot , so there is no issue trying to tie together multiple spectral segments ( e.g. , g130 m and g160 m of cos , including the individual fp - pos steps , each with its own potential calibration issues ) .
further , the stis mama cameras have hard - wired physical pixels , whereas the cross - delay - line detectors of cos have floating pixels defined by the @xmath15 impact point of the electron cloud associated with a photon event .
the implication for stis is that a given pixel always experiences whatever geometrical distortions are present locally , and these can be accounted to some degree in the wavelength dispersion solution . with cos ,
conversely , the geometrical distortions are sampled quasi - randomly by whatever distribution of impact points happen to fall at a particular location .
perhaps more important , the stis wavelength calibration light ( from onboard hollow - cathode emission lamps ) follows the same optical path as an external stellar point source , falling on the same echelle order pixels as the stellar spectrum , and consequently experiences the same geometrical distortions .
thus , the stis wavelength calibration spectra encode significant information concerning these local geometrical distortions , which then can be accommodated in a hybrid dispersion solution , such as described in the astral protocols mentioned earlier . for cos ,
on the other hand , the lamp - flash spectrum falls well above the spectral stripe ( see , e.g. , fig .
3 ) , on a portion of the detector where the geometrical distortions are completely different . consequently ,
the cos lamp - flashes are less relevant for confirming the pipeline dispersion solutions .
to test the wavelength fidelity of cos , the three independent combinations of grating setting and detector segment were considered separately .
reference spectra were taken from the calcos pipeline `` x1d '' files , which include a summation over the four fp - pos exposures , eliminating bad pixels and suppressing fixed pattern noise such as grid - wire shadows .
the visits 3 - 5 spectra for each grating setting were registered to visit 3 by cross - correlation , and co - added to increase s / n . in all cases , extracted lamp - flash lines from the individual fp - pos exposures were very stable in velocity , demonstrating the repeatability of the calcos pipeline wavelength assignments ( although not necessarily their accuracy ) .
figure 8 compares selected profiles of bright lines from the various cos grating / detector - segment combinations against the stis e140 m spectrum from visit 1 , convolved with the cos line - spread function .
the zero point of the stis wavelength scale ( as in fig .
2 ) was derived from the average velocity of @xmath81304 , @xmath81306 , and @xmath81351 , the brightest low - excitation narrow chromospheric lines recorded by stis .
the ground - state @xmath81302 resonance line was avoided owing to a blue asymmetry imposed by redward interstellar atomic oxygen absorption ( see fig .
note also that the stis @xmath81670 resonance line is truncated in its red wing by a gap between adjacent echelle orders .
the derived chromospheric radial velocity was @xmath47 km s@xmath1 , compared with the photospheric value of @xmath48 km s@xmath1 expected in 2012 near orbital phase 0.5 ( knig et al .
this indicates that part of the velocity correction , amounting to @xmath49 km s@xmath1 , could be a zero - point error in the stis wavelength scale for this observation , assuming that the fuv chromospheric velocity measured by stis , and the optical photospheric velocities measured by knig et al .
, refer to the same frame , and further that the optical measurements are freer of systematics . aside from the possible zero - point correction ,
the stis wavelength scales have high internal precision .
-11 mm + 2.5 mm -8 mm the cos spectra are depicted in two ways : ( 1 ) with the original pipeline wavelengths , adjusted only for the assumed stellar radial velocity ( @xmath48 km s@xmath1 ) ; and ( 2 ) for a simple modification of the cos wavelength scales to force better global agreement with stis .
the latter corrections , expressed as equivalent velocities , are of the form : @xmath50 the value @xmath51 km s@xmath1 was used for all the grating / segment combinations .
the @xmath52 coefficients and reference wavelengths , @xmath53 , are as follows : @xmath540.112 km s@xmath1 @xmath1 and 1340 for g130m / a ; @xmath540.054 km s@xmath1 @xmath1 and 1400 for g160m / b ; and @xmath540.046 km s@xmath1 @xmath1 and 1640 for g160m / a . here
, the negative @xmath52 coefficients are compensating for systematic redshifts of the assigned wavelengths that become progressively larger with wavelength .
for example , the @xmath81393 feature recorded by g130m / a ( as in the earlier ayres & france cos snapshot ) , would have a total instrumental redshift of 10 km s@xmath1 , about half of the velocities ( @xmath10 km s@xmath1 ) measured in that study .
however , a direct comparison between the two is subject to unknown systematic errors , not the least due to the different acquisition techniques , so the difference should not be taken too literally .
the corrected cos spectra , in general , agree well with ground - truth stis , although there is at least one important exception
@xmath81393 at the extreme blue edge of g160m / b where the linear velocity model appears to fall short .
the cited grating - dependent velocity corrections were applied to the time - resolved cos profiles illustrated earlier in figs . 4 and 5 .
nevertheless , the fundamental velocity measurements of the brightest lines , at least for the quiescent periods of ek dra , were derived from the more precise stis spectra , as described below . as mentioned earlier
, there are interstellar absorptions present in several of the ground - state resonance lines of the stis fuv spectrum of ek dra .
these offer the opportunity to check the stis zero - point velocity deduced from narrow chromospheric emissions .
the ism features are redshifted with respect to ek dra , with an average heliocentric velocity of about @xmath542 km s@xmath1 . according to the maps of redfield & linsky ( 2008 ) ,
ek dra falls at the edge of the local interstellar cloud ( lic ) and also at the edge of the smaller north galactic pole ( ngp ) cloud . depending on the velocity vector assumed for the lic ( e.g. , gry & jenkins 2014 ) ,
the predicted heliocentric velocity in the direction of ek dra falls in the range @xmath54(1 - 3 ) km s@xmath1 ; while for the ngp cloud , the projected velocity is close to 0 km s@xmath1 : the galactic coordinates of ek dra are nearly perpendicular to both directions of motion .
it also is possible , and perhaps likely , that both clouds are contributing to the ism absorptions in the ek dra spectrum .
weighing in on the matter is a recent stis nuv high - resolution echelle ( e230h-2713 ) spectrum of ek dra , taken in the 2014 ism snap program mentioned earlier : a 600 s exposure through the @xmath14 slot , without a peak - up ( see table 1 ) .
the spectrum covers the important resonance lines at 2796 and 2803 , which have ism absorptions on the red sides of their broad chromospheric emission cores .
the ism features fall at a velocity of @xmath55 km s@xmath1 heliocentric .
however , cross - correlation against the same e230h-2713 bandpass from @xmath5 cen a suggests that there might be a @xmath56 km s@xmath1 systematic error in the ek dra nuv photospheric spectrum , which would revise the ism velocity to about @xmath56 km s@xmath1 .
the fuv and nuv interstellar features are illustrated in figure 9 .
the agreement between the two independent sets of ism velocities provides additional confidence in the zero - point velocity calibration of the stis fuv spectra .
tables 3 - 5 list representative measurements from stis visit 1 ( table 3 ) , the cos visits 3 - 5 averages for both g130 m and g160 m ( table 4 ) , and the cos visit 2 flare ( table 5 : g160 m only ) .
the stis entries include important bright lines below the shortwavelength cutoff of cos g130m / a at 1290 , such as ly@xmath5 , @xmath81206 , and the resonance doublet at 1240 .
stis also recorded detailed profiles of the 1305 triplet lines , which fall at the blue end of g130m / a but are corrupted by atomic oxygen skyglow in most of the cos pointings .
the cos tables include weaker features that were not detectable in the stis spectrum , especially the key density - diagnostic multiplet of ] in the 1400 region .
the same g160 m features in table 4 also were measured in the integrated g160 m flare spectrum of visit 2 , and reported in table 5 , to compare with the more quiescent intervals following the outburst . for cos ,
the measured doppler shifts are affected by residual errors in the simple correction of the g130 m and g160 m wavelength scales , so should be treated with caution .
however , the absolute fluxes and derived line widths , say of narrow components ( nc ) and broad components ( bc ) in a bimodal fit , should be more reliable , since neither depends sensitively on accurate wavelengths .
any systematic doppler shifts between nc and bc within a given spectral feature should be reliable as well , because the ( corrected ) cos wavelengths have sufficient precision over small wavelength separations .
llccccc & 1175_m _ & int & ( 1175.7 ) & ( 2.2 ) & @xmath57 & 0.00 + & 1206.500 & gau2 & @xmath58 & @xmath59 & @xmath60 & 0.00 + & 1206.500 & gau2 & @xmath61 & @xmath62 & @xmath63 & + & 1215.339 & gau1s & @xmath64 & @xmath65 & @xmath66 & @xmath67 + & 1215.670 & gau2 & @xmath68 & @xmath69 & @xmath70 & 0.43 + & 1215.670 & gau2 & @xmath71 & @xmath72 & @xmath73 & + & 1215.670 & gau1 & @xmath74 & @xmath75 & @xmath76 & 0.29 + & 1215.670 & gau1x & @xmath77 & & & + & 1215.670 & int & ( 1215.6 ) & ( 5.0 ) & @xmath78 & 0.43 + & 1238.810 & gau2d & @xmath79 & @xmath80 & @xmath81 & 0.00 + & 1238.810 & gau2d & @xmath82 & @xmath83 & @xmath84 & [ @xmath85 + & 1302.168 & gau2r & @xmath86 & @xmath87 & @xmath88 & 0.00 + & 1302.168 & gau2r & @xmath89 & @xmath90 & @xmath91 & + & 1304.858 & gau1 & @xmath92 & @xmath93 & @xmath94 & 0.00 + & 1306.029 & gau1 & @xmath95 & @xmath96 & @xmath94 & 0.00 + & 1334.532 & gau2r & @xmath97 & @xmath98 & @xmath99 & 0.02 + & 1334.532 & gau2r & @xmath100 & @xmath101 & @xmath102 & + & 1335.702 & gau2 & @xmath103 & @xmath104 & @xmath105 & 0.02 + & 1335.702 & gau2 & @xmath106 & @xmath107 & @xmath108 & + & 1334_m _ & int & ( 1335.2 ) & ( 2.8 ) & @xmath109 & 0.02 + @xmath110 & 1351.656 & gau1 & @xmath111 & @xmath112 & @xmath113 & 0.03 + & 1354.08 & gau1 & @xmath114 & @xmath115 & @xmath116 & 0.03 + & 1355.598 & gau1 & @xmath117 & @xmath118 & @xmath119 & 0.03 + & 1355.844 & gau1 & @xmath120 & @xmath118 & @xmath121 & 0.00 + & 1393.760 & gau2d & @xmath122 & @xmath123 & @xmath124 & 0.02 + & 1393.760 & gau2d & @xmath125 & @xmath126 & @xmath127 & [ @xmath128 + & 1548.204 & gau2d & @xmath129 & @xmath130 & @xmath131 & 0.03 + & 1548.204 & gau2d & @xmath106 & @xmath132 & @xmath133 & [ @xmath134 + & 1640.404 & gau2 & @xmath135 & @xmath136 & @xmath137 & 0.11 + & 1640.404 & gau2 & @xmath138 & @xmath139 & @xmath140 & + & gau2 & @xmath141 & @xmath142 & @xmath143 & + & gau2 & @xmath144 & @xmath145 & @xmath146 & + & 2796.352 & gau1 & @xmath147 & @xmath148 & @xmath149 & 13.2 + & 2803.531 & gau1 & @xmath150 & @xmath151 & @xmath152 & 13.2 + & 2796.352 & gau1s & @xmath153 & @xmath154 & @xmath155 & @xmath156 + & 2803.531 & gau1s & @xmath157 & @xmath158 & @xmath159 & @xmath160 llccccc & 1334.532 & gau2r & @xmath161 & @xmath162 & @xmath163 & 0.03 + & 1334.532 & gau2r & @xmath164 & @xmath165 & @xmath166 & + & 1335.702 & gau2 & @xmath167 & @xmath168 & @xmath169 & 0.03 + & 1335.702 & gau2 & @xmath170 & @xmath171 & @xmath172 & + & 1334_m _ & int & ( 1335.2 ) & ( 3.2 ) & @xmath109 & 0.03 + & 1351.656 & gau1 & @xmath173 & @xmath174 & @xmath175 & 0.04 + & 1354.08 & gau1 & @xmath176 & @xmath177 & @xmath178 & 0.04 + & 1355.598 & gau1 & @xmath179 & @xmath180 & @xmath181 & 0.04 + & 1355.844 & gau1 & @xmath182 & @xmath180 & @xmath183 & 0.00 + & 1371.296 & gau1 & @xmath184 & @xmath185 & @xmath186 & 0.05 + & 1393.760 & gau2d & @xmath187 & @xmath188 & @xmath189 & 0.06 + & 1393.760 & gau2d & @xmath190 & @xmath191 & @xmath192 & [ @xmath193 + & 1401.157 & gau7 & @xmath194 & @xmath195 & @xmath196 & 0.07 + & 1397.202 & gau7 & & & @xmath197 & + & 1399.769 & gau7 & & & @xmath198 & + & 1404.780 & gau7 & & & @xmath199 & + & 1404.808 & gau7 & & & @xmath200 & + & 1406.051 & gau7 & & & @xmath201 & + & 1407.379 & gau7 & & & @xmath202 & + & 1393.760 & gau2d & @xmath203 & @xmath204 & @xmath205 & 0.06 + & 1393.760 & gau2d & @xmath206 & @xmath207 & @xmath208 & [ @xmath209 + & 1401.157 & gau7 & @xmath210 & @xmath211 & @xmath212 & 0.07 + & 1397.202 & gau7 & & & @xmath197 & + & 1399.769 & gau7 & & & @xmath213 & + & 1404.780 & gau7 & & & @xmath214 & + & 1404.808 & gau7 & & & @xmath200 & + & 1406.051 & gau7 & & & @xmath201 & + & 1407.379 & gau7 & & & @xmath215 & + & 1526.707 & gau1 & @xmath216 & @xmath217 & @xmath218 & 0.10 + & 1533.431 & gau1 & @xmath219 & @xmath98 & @xmath220 & 0.10 + & 1548.204 & gau2d & @xmath221 & @xmath168 & @xmath222 & 0.14 + & 1548.204 & gau2d & @xmath64 & @xmath223 & @xmath224 & [ @xmath225 + & 1640.404 & gau2 & @xmath226 & @xmath217 & @xmath227 & 0.15 + & 1640.404 & gau2 & @xmath228 & @xmath229 & @xmath99 & + & 1657_m _ & int & ( 1657.1 ) & ( 3.0 ) & @xmath230 & 0.21 + & 1670.787 & gau2 & @xmath231 & @xmath232 & @xmath233 & 0.18 + & 1670.787 & gau2 & @xmath234 & @xmath235 & @xmath236 & llccccc & 1393.760 & gau2d & @xmath237 & @xmath238 & @xmath239 & 0.39 + & 1393.760 & gau2d & @xmath240 & @xmath241 & @xmath242 & [ @xmath243 + & 1401.157 & gau7 & @xmath244 & @xmath245 & @xmath246 & 0.39 + & 1397.202 & gau7 & & & @xmath247 & + & 1399.769 & gau7 & & & @xmath248 & + & 1404.780 & gau7 & & & @xmath249 & + & 1404.808 & gau7 & & & @xmath250 & + & 1406.051 & gau7 & & & @xmath251 & + & 1407.379 & gau7 & & & @xmath252 & + & 1526.707 & gau1 & @xmath253 & @xmath254 & @xmath255 & 0.53 + & 1533.431 & gau1 & @xmath256 & @xmath257 & @xmath88 & 0.53 + & 1548.204 & gau2d & @xmath258 & @xmath259 & @xmath260 & 0.58 + & 1548.204 & gau2d & @xmath261 & @xmath262 & @xmath263 & [ @xmath264 + & 1640.404 & gau2 & @xmath265 & @xmath217 & @xmath266 & 0.69 + & 1640.404 & gau2 & @xmath267 & @xmath268 & @xmath269 & + & 1657_m _ & int & ( 1657.1 ) & ( 3.0 ) & @xmath270 & 0.74 + & 1670.787 & gau2 & @xmath271 & @xmath118 & @xmath272 & 0.77 + & 1670.787 & gau2 & @xmath273 & @xmath274 & @xmath275 & reference ( `` laboratory '' ) wavelengths ( in vacuum ) were taken from ayres ( 2015 ) , except for the solar flare line [ ] , which was not included in the @xmath5 cen study .
the nist atomic spectra database ( asd : kramida et al .
2013 ) wavelength ( 1354.08 ) was adopted for this key coronal forbidden transition , but the quality of the measurement is rated poor . recently ,
young et al .
( 2015 ) proposed a revised wavelength of [email protected] , based on measurements of [ ] in solar flares .
if correct , it would shift the [ ] velocities in tables 3 - 5 by @xmath277 km s@xmath1 to the blue .
a variety of fitting strategies were employed , following the lead of the ayres ( 2015 ) study of stis fuv echelle spectra of @xmath5 cen a and b. the particular measurement scenario is flagged explicitly in col . 3 of the tables , as follows .
`` int '' refers to a numerical flux integration over the velocity bandpass listed in col .
the value in col .
7 is a derived continuum flux density , assumed constant over the spectral feature or multiplet , based on filtering the surrounding intensities to suppress emission lines .
`` gau1 '' refers to a single gaussian fit .
`` gau1s '' is a special case in which a sloping continuum was allowed : col .
7 now contains the zero point and gradient of a linear continuum fit : @xmath278 , where @xmath279 is the monochromatic continuum flux density and @xmath280 is the transition wavelength .
`` gau1x '' is another specialized fit , applied to the ism absorption core of ly@xmath5 .
the fitting function is of the form @xmath281 , where @xmath282 and @xmath283 is the `` doppler width . ''
compared with the normal gaussian ( @xmath284 ) , the steep - sided flat - bottomed gau1x profile is more appropriate for the saturated ly@xmath5 absorption core .
however , the fwhm and total flux have little meaning , especially because the imposed continuum level ( at the apparent peak of ly@xmath5 ) is rather arbitrary , so the derived values are not reported in the table ; only the velocity shift is significant .
`` gau2 '' is a bimodal gaussian fit of the narrow / broad variety , used specifically for the + + collective filtered profiles , but also for other isolated features such as @xmath81206 .
for the filtered hot - line average , the component fluxes are narrow / total and broad / total .
`` gau2r '' refers to a double - gaussian fit applied to ostensibly unrelated features ( say , stellar emission and superimposed interstellar absorption ) , for which the second component is negative in flux to mimic a central reversal or ism absorption . in these cases
, the width of the principal component should not be taken too literally , although the velocity shift should be reliable .
`` gau7 '' is a 7-component gaussian model covering combined multiplets of similar transitions , namely ] + ] , for which the widths of the features were constrained to be the same , and the doppler shifts as well ( i.e. , the wavelength differences between the components were fixed at the reference values , but a shift of the whole transition array was allowed ) .
the derived collective velocity shift and average profile width , together with the associated uncertainties ( and the continuum flux ) are listed for the principal transition . for the remaining multiplet members ,
only the unique flux values are provided .
finally , `` gau2d '' refers to a specialized constrained bimodal gaussian fit applied to the hot - line doublets , such as @xmath81393+@xmath81402 .
figure 10 illustrates such fits to the ( un - smoothed ) ek dra hot lines from the co - added stis spectrum .
this bimodal strategy for decomposing the complex hot - line profile shapes of cool stars has a long history , as discussed recently by ayres ( 2015 ) in the context of low - activity @xmath5 cen a ( and companion b [ k1 v ] ) , and for the earlier cos pointing on ek dra itself by ayres & france ( 2010 ) .
the upper paired panels of fig .
10 illustrate the application to the , , and resonance doublets : principal components to the left , secondary to the right ( but with one - half the @xmath17-axis scale ) . under optically - thin conditions
, the weaker components would be half the flux densities of the stronger ones , and thus should appear similar in these scaled diagrams .
that expectation is met for , but becomes progressively out of sync for and .
the fitting procedure assumed that the doublet members were self - similar : nc of the same width , bc of the same width ; doppler shifts of the nc the same , and similarly for the bc ( although possibly different between narrow and broad ) ; and the same flux ratios , narrow / broad . only the relative total fluxes between the doublet members
was allowed to float , as well as the wavelength difference between the two transitions , to allow for small errors in the laboratory values , or subtle residual distortions in the stis wavelength scales .
-3 mm for this gau2d case , the first set of parameters listed ( in tables 3 - 5 ) is for the nc of the constrained doublet fit , and col .
7 holds the continuum flux as usual .
the col .
2 wavelength is for the principal member of the doublet .
the second set of parameters listed is for the bc .
however , highlighted by square brackets , col .
7 now has the doublet flux ratio , weaker / stronger , which should be close to 0.5 in the optically thin limit ( for the specific hot - line doublets here ) ; as well as the deviation of the fitted wavelength difference from the reference value , expressed in km s@xmath1 .
these latter shifts are small for the stis spectrum , but can be substantial in several of the cos cases , mainly reflecting residual errors in the simple correction of the cos wavelengths . the bottom panel of fig .
10 illustrates a gau2 bimodal fit to a filtered sum of the three hot - line doublets , each member ( six altogether ) scaled to the same integrated flux over the interval @xmath285 km s@xmath1 .
the scaled fluxes were combined on a common velocity grid , applying an olympic filter ( throwing out the highest and lowest of the six values ) for the central @xmath286 km s@xmath1 zone , and a more aggressive filtering ( eliminating the three highest values ) for the outer intervals where the influence of weak emission blends on the broad component is strongest . as with the individual bimodal
fits , the hybrid profile has a relatively narrow nc , about 50 km s@xmath1 fwhm , and is redshifted by 5 km s@xmath1 , similar to those of low - activity g dwarfs like @xmath5 cen a ; but a very broad bc , about three times wider than the nc , and more redshifted ( 9 km s@xmath1 ) .
the enhanced broadening and redshift of ek dra s bc are the main distinguishing characteristics compared with low - activity counterparts like @xmath5 cen a ( e.g. , fig . 2 ) .
uncertainties on the fitted parameters were estimated by a monte carlo approach : a reference profile ( the derived fit itself ) was perturbed by random realizations of the smoothed photometric error , and then refitted .
the standard deviations of the derived parameters over many such trials were taken as a fair gauge of the measurement uncertainties devolving from the photometric noise component .
recall the two bursts ( at @xmath403.6 hr and 3.7 hr ) in the g160 m blue and red continuum bands , which remarkably were not mirrored in any of the other activity tracers .
these isolated bursts are distinctly bluer ( indicating a hotter continuum ) than the more equal count rates of the two separated continuum bands later on , in quiet visits 3 - 5 .
this continuum - only behavior is challenging to understand , and seems to have few , if any , precedents among existing records of stellar flare events .
this likely is because there are only a handful of stellar flare campaigns simultaneously at uv and soft x - ray energies .
one example is mitra - kraev et al .
( 2005 ) , who were able to use the optical monitor on _ xmm - newton , _ with broad - band fuv and nuv filters , to jointly record a number of uv / x - ray events on several dme stars .
the authors found that the uv rise typically preceded the soft x - rays by several minutes , which usually is the case for white - light flares ( wlf ) on the sun .
there is perhaps only a single example in their several time series where a uv burst was seen during a flare decay without an x - ray counterpart ( their fig . 1 ) .
on the sun , wlfs are difficult to separate from the glare of the photosphere , but if one is seen , it usually will have an extensive supporting context , especially at high energies .
fang & ding ( 1995 ) divided solar wlfs into two classes : type i , or `` prompt , '' events which are coincident with the hard x - ray ( hxr ) burst , and apparently result from direct excitation of the chromospheric gas by the flare s energetic particle beams ; and the less frequent type ii , or `` delayed , '' bursts , which occur several minutes after the hxr impulsive phase , and appear to be deeper seated , in the photosphere itself .
fang & ding did not speculate on a source for the type iis .
perhaps the delayed fuv continuum brightenings on ek dra are related to the solar wlf type iis .
one idea , motivated by the hot - line redshifts during the flare decay , is that the bursts are a response to the ballistic splash - down of the cooling gas onto the photospheric layers ; something analogous to accretion streams in t - tauri stars .
ayres et al .
( 2003 ) carried out a search for coronal forbidden lines in stis fuv spectra of the late - type objects with sufficiently deep exposures available at the time . in the hyper - active members of the sample ,
like rs canum venaticorum binaries and dme flare stars , the intrinsically faint [ ] @xmath81354 feature often was visible .
furthermore , the [ ] line strength was found to be linearly correlated with the _ rosat _ soft x - ray flux . in the moderate - to - slow rotators ( in terms of @xmath287 ) , the average fwhm of the coronal iron line was about 110 km s@xmath1 , somewhat larger than the expected thermal width of @xmath288 km s@xmath1 at the formation temperature of 10 mk . in several of the fast rotators , there was a suggestion of @xmath289 enhanced broadening beyond that expected from @xmath287 alone .
the `` super - rotational '' broadening was attributed to highly extended hot coronal structures .
in ek dra , the cos visits 3 - 5 average profile of [ ] does not display obviously extended red wings , like the lower temperature subcoronal features , although certainly the coronal forbidden line is very broad , approximately mid - way in width between the hot - line nc and bc . nevertheless , the fwhm@xmath290 km s@xmath1 is close enough to the low-@xmath287 average noted above that it would be difficult to make a case for super - rotational broadening .
the cos velocity of [ ] in the visits 3 - 5 average spectrum is @xmath291 km s@xmath1 for the nist reference wavelength , although with the young et al .
( 2015 ) revised value , the shift would decrease to about 0 ( @xmath292 km s@xmath1 from the @xmath293 of the reference wavelength ) . at the same time , the stis velocity of [ ] would shift to @xmath294 km s@xmath1 ( @xmath295 km s@xmath1 , where the larger @xmath296 partly is due to the reference wavelength , and partly to the low s / n of the measurement ) . because [ ] correlates well with the _ rosat _ flux for hyper - active coronal sources , the measured intensity during the pre- and post - flare intervals can be exploited to estimate the quiescent x - ray flux of ek dra during the 2012 campaign , noting that the existing _ rosat _ survey @xmath297 erg s@xmath1 ( 0.1 - 2.4 kev ) reported by gdel et al .
( 1995 ) is not exactly simultaneous .
the scaling relation of ayres et al .
( 2003 ) can be expressed as : @xmath298 the stellar bolometric flux , @xmath299 erg @xmath300 s@xmath1 , can be estimated from @xmath44 and a bolometric correction derived from @xmath301 ( see ayres et al .
the bolometric luminosity , @xmath302 erg s@xmath1 , follows from the distance ( 34 pc ) , and is similar to @xmath303 .
the [ ] flux ratio , based on the stis / cos average @xmath304 erg @xmath300 s@xmath1 , is @xmath305 : in the upper tier among the coronal objects considered by ayres et al . ( 2003 ) .
finally , the inferred x - ray luminosity is @xmath306 erg s@xmath1 , about twice the _ rosat _ survey value , but certainly consistent within the uncertainties of the [ ] scaling law , and the temporal variability of @xmath18 noted by gdel et al .
( 1995 ) .
figure 11 is a correlation diagram pitting various diagnostics against the doublet , which was recorded in every cos observation . as with the previous diagram(s ) , the individual flux values from all the visits were normalized to the corresponding averages over visits 3 - 5 .
the upper distributions of symbols are contributed by the large flare , while the clump at ( 1,1 ) is dominated by the quiescent intervals , which nevertheless display low - level variability including a small flare decay .
the straight lines are power - law fits to the individual distributions .
the derived slopes range from 1.44 and 1.38 for [ ] and , respectively , to slightly above unity for ( 1.10 ) , and slightly below for chromospheric ( 0.89 ) .
these slopes mirror similar power laws seen between , say , x - rays and in diverse samples of solar - type dwarfs ( e.g. , ayres et al .
1995 ) , and in this instance must reflect the time - evolution of the post - flare cooling process .
figure 12 compares doublet profiles from the cos g160 m exposures of the visit 2 flare decay , the visits 3 - 5 average cos g160 m spectrum , and a multi - epoch stis average for the quiet solar twin @xmath5 cen a ( ayres 2015 ) ; all including bimodal constrained gaussian fits .
the quiescent profiles of ek dra show redshifts of both components , less for the nc ( @xmath307 km s@xmath1 , for the average of the stis doublet members ) , more for the bc ( @xmath308 km s@xmath1 , for the stis doublet average ) .
this compares with a somewhat smaller redshift of both @xmath5 cen a components ( @xmath309 km s@xmath1 for each ) .
the width of the ek dra nc in the quiet spectrum ( 50 km s@xmath1 ) is somewhat larger than in @xmath5 cen a ( 36 km s@xmath1 ) , but the bc is much broader ( 180 km s@xmath1 vs. 66 km s@xmath1 ) . in the flare ,
remarkably , the nc has a similar redshift and fwhm ( @xmath291 km s@xmath1 and 45 km s@xmath1 , respectively ) to the `` quiet '' average , and indeed to the low - activity solar twin @xmath5 cen a. at the same time , the flare bc not only is broader than in the quiet cases ( @xmath310 km s@xmath1 ) , but is substantially more redshifted ( at about + 40 km s@xmath1 ) .
the larger bc redshifts during the flare are notable but not unusual : similar behavior , with similar velocity shifts , has been seen previously in stis spectra of flares on the dme stars au mic ( robinson et al .
2001 ) and ad leo ( hawley et al .
2003 ) , and appears to be a signature of the post - flare cooling process .
what perhaps is more remarkable is that despite the very large difference in emission levels of the two solar analogs ( expressed , say , in terms of surface fluxes ) , the `` quiet '' profiles are as similar as they are , not only in the bimodal nature of the line shape asymmetries , but also in the magnitudes of the velocity shifts and line widths . -5 mm -5 mm lcccc theoretical limits & 0.160.42 & 0.32.5 & 1.02 & 0.13 + electron density range ( @xmath4 ) & @xmath311 & @xmath311 & + quiet average : cos g130 m & @xmath312 & @xmath313 & @xmath314 & @xmath315 + quiet average : cos g160 m & @xmath316 & @xmath317 & @xmath318 & @xmath319 + flare : cos g160 m & @xmath320 & @xmath321 & @xmath322 & @xmath323 + quiet average : stis e140 m & @xmath324 & @xmath325 & @xmath326 & @xmath327 + flower & nussbaumer ( 1975 ) originally proposed that flux ratios within the ] 1400 intercombination multiplet are sensitive to density over the range @xmath328 @xmath4 thought to be characteristic of solar conditions quiescent at the low end , flaring on the high side at the o@xmath329 ionization equilibrium peak temperature of @xmath330 k. there is a long history of applications to solar and stellar plasmas ( see , e.g. , cook et al .
1995 , and references therein ) , continuing to the present ( e.g. , keenan et al . 2009 ; ayres 2015 ) .
figure 13 illustrates the ] region from the initial visit 2 flare interval ( g160 m , only ) ; the quiet averages ( visits 3 - 5 ) from both g160 m and g130 m ; and the @xmath5 cen a stis spectrum .
the logarithmic presentation emphasizes the wider , more redshifted broad components of the ek dra profiles compared with low - activity @xmath5 cen .
the 7-component gaussian fits ( five ] , two ] transitions ) take into account the small contamination of ] @xmath81404 by ] @xmath81404 , scaled from ] @xmath81406 ( one of the fitted features ) according to plasma emission models ( see ayres 2015 ) .
table 6 summarizes ] diagnostic ratios constructed from the cos measurements in tables 4 and 5 , and including values from the multi - epoch average @xmath5 cen a stis spectrum .
the nomenclature for the ratios is that of ayres ( 2015 ) : the numerators in the traditional density - sensitive @xmath331 and @xmath332 were replaced by @xmath333 to improve s / n with these generally faint emissions : @xmath334 and @xmath335 share a common upper level and their ratio is very close to unity according to theory .
the hybrid numerator ratios are designated @xmath336 .
the ratio @xmath337 also is composed of transitions that share a common upper level , with a theoretical value of 0.13 ( keenan et al .
however , in this case there is no advantage to form a hybrid flux between @xmath338 and @xmath339 because the former transition usually is too low in s / n , if detected at all , in normal astrophysical sources . the @xmath5 cen @xmath340 and @xmath337 ratios are close to the theoretical expectations .
this also is the case for ek dra , although the uncertainties are larger , especially for marginally detected @xmath338 .
note , however , that the diagnostic ratios @xmath341 and @xmath342 are significantly elevated in ek dra compared with @xmath5 cen .
this is a clear indication of enhanced electron densities in the more active star .
the @xmath5 cen ratios fall near the low - density limits of the respective @xmath343 s , indicating @xmath344 just below @xmath345 @xmath4 ; while the ek dra ratios both are consistent with densities in the middle range , near or above @xmath346 @xmath4 , an order of magnitude larger .
curiously , the inferred densities are not much different between the more quiet periods of the ek dra cos campaign and the large fuv flare decay at the beginning ; especially for @xmath342 , which has the largest grasp of the two diagnostic ratios ( i.e. , largest variation in @xmath343 over the range of density sensitivity ) .
the ] densities inferred here are about a factor of 2 higher than those estimated by gdel et al .
( 1995 ) from emission - measure modeling of the _ rosat _ survey pulse - height spectra , at least for the softer ( 2 mk ) component of their 2-temperature fit ( this is the component that showed rotational modulations , as mentioned earlier ; whereas the harder , more dominant , @xmath347 mk one showed essentially no phase - correlated variability ) .
nevertheless , the agreement must be tempered by the fact that the subcoronal gas is only one - tenth the temperature of even the cooler coronal plasma component , and likely is not co - located .
the high densities during the flare decay in visit 2 suggest short cooling times ( @xmath348 ) , at least in the hot - line temperature range ( @xmath349 k ) where the radiative cooling curve peaks .
this seemingly conflicts with the hours - long progression of the fuv decay .
a similar dichotomy confronts interpretations of long - duration flares on the sun and other late - type stars .
the usual explanation is that the long duration events are analogous to two - ribbon ( 2-r ) flares on the sun ( e.g. , gdel 2004 ) , which involve a sequential cascade of excitation through large arcade - like magnetic structures , rather than an impulsive burst in a single coronal loop .
such 2-r events tend to be quite hot , exceeding 10 mk in peak temperature , with coronal electron densities up to about @xmath346 @xmath4 ( e.g. , gdel 2004 ) .
importantly , the decay of the flare plasma in a 2-r event can be dominated by conductive energy exchange with the lower cooler layers , rather than direct radiative losses at the super - hot coronal temperatures ( see aschwanden et al .
thus , the bright fuv emissions and subsonic plasma draining seen in the hot lines of ek dra during the large fuv flare decay likely are a byproduct of these cooling processes playing out in the much more energetic setting of the hyper - active young solar analog .
this brings to mind , as intimated by ayres & france , the dynamic `` coronal rain '' phenomenon associated with post - flare loops on the sun , as well as solar active regions in general ( see , e.g. , antolin & rouppe van der voort 2012 , and references therein ) . because of the unique nature of the time - resolved fuv spectroscopy of the serendipitous ek dra flare decay , it is important to place the event within the context of the nearly 100 large stellar flares captured in the soft x - ray band by historical and contemporary high - energy observatories ( as documented in the review by gdel 2004 ) .
aschwanden et al .
( 2008 ) have fitted various combinations of the derived emission measures ( em ) , x - ray luminosities ( @xmath18 ) at the flare temperature peak ( @xmath350 ) , and decay times ( @xmath351 ) with scaling relations approximated as power laws .
inverting their eq . 12 for the dependence of decay time on flare peak temperature yields an estimate of the ( here unknown ) @xmath350 given the ( known ) @xmath351 : @xmath352 given the apparent , albeit fuv , flare decay time of about 7 ks , the peak x - ray temperature would be about 60 mk . using this value in eq .
13 of aschwanden et al .
yields @xmath353 erg s@xmath1 , which is an order of magnitude larger than the `` quiescent '' @xmath18 inferred from scaling the contemporaneous [ ] flux from the non - flare intervals . supporting that estimate
is the fact that the peak enhancement of seen in cos visit 2 is about a factor of 10 , and the [ ] proxy for the high - temperature soft x - ray flux although recorded later in visit 2 , not simultaneously with parallels the trajectory during the flare decay , suggesting that [ ] would have been enhanced by a similar large factor , and so too the soft x - ray flux . the inferred values of flare peak temperature , x - ray luminosity , decay time , and density ( again with the caveat that the latter is from the cooler fuv plasma ) place the ek dra event in the upper - middle range of the documented stellar x - ray flares , as summarized by aschwanden et al . ; perhaps typical of a young active sunlike star , but more accessible in the case of ek dra by its proximity : several times closer to earth than the members of the similar - age young clusters pleiades and alpha persei .
the joint stis / cos study of ek dra was adversely impacted by the unfortuitously - timed large fuv flare ( interrupting the intended instrumental cross - calibration ) , but nevertheless the event itself provided valuable insight into the post - flare cooling process , mainly thanks to the high spectral resolution achieved for a variety of key chromospheric and higher temperature features .
the apparent quiescent periods following the flare also contributed significant insight , not only with regard to the velocity cross - calibration of the two uv spectrographs , but also in the lack of phase - correlated profile variations caused by rotational modulations . to summarize : ( 1 ) the fuv redshifts appear to be a characteristic of the normal surface activity of the star , rather than an accident of doppler imaging ; ( 2 ) the hot - line profiles did not change qualitatively during the flare outburst ( ten - times enhancement of ) , except for the larger redshifts of the broad components ; and ( 3 ) ] -derived densities were not very different between the quiescent periods and the fuv flare , despite the fact that the high densities themselves are characteristic of large stellar soft x - ray events ( see , e.g. , benz & gdel 2010 ) .
all of these properties suggest that the normal surface conditions on the star are not far removed from those in a big flare ; in other words , the stellar corona might be dominated by flaring structures , with a distribution weighted toward the smaller events , as on the sun , giving the appearance of quiescence at times , when exactly the opposite is true ( see gdel 2004 for a review of the `` flares all the way down '' coronal scenario ; what might be described as a _ `` flare - ona '' _ ) .
one notable outcome of the time - resolved flare spectroscopy was the lack of strongly blueshifted emission , which naively might be anticipated from the catastrophic energy release , and indeed such outbursts on the sun are violently eruptive .
however , as has been noted often in the flare literature , during the explosive heating of the lower atmosphere in an event , most of the outwardly expanding material undoubtedly is too hot to be visible in the fuv lines .
only the subsequent decay phase would contribute strongly to the fuv features , as the super - hot coronal plasma in the flaring arcades conductively redistributes heat into the lower atmosphere , where it is radiated away partially at fuv temperatures and inspires the subsonic plasma draining phenomenon manifested in the hot - line redshifts .
unfortunately , the initial impulsive phase of the ek dra flare was not captured , and the key diagnostic of the hot - up / cool - down scenario [ ] was recorded only during the later phases of the event decay ( although note in fig . 4b the apparent small blueshifts of the coronal forbidden line compared with the average visits 3 - 5 behavior , which would be more exaggerated with the young et al
. reference wavelength ) .
it also must be acknowledged that the peculiarities of the ek dra flare decay should be viewed with some caution , because rarely has such an event been recorded in as much detail as provided by the cos part of the _ hst _ campaign : the oddities of the emission doppler shifts might well be associated with an unusual spatial orientation of the event , for example . in the end , highly exaggerated stellar incidents like the ek dra fuv flare dutifully cause us to look back to the sun in an effort to identify associations with solar high - energy transient phenomena , such as the coronal rain mentioned earlier .
nevertheless , drawing such conclusions in general is tricky given the rather tepid activity levels displayed by our local star , even at its worst .
probably a good thing for habitability of planet earth , but a continual challenge for the solar - stellar connection .
this work was supported by grant go-12566 from the space telescope science institute , based on observations from _ hubble space telescope _ collected at stsci , operated by the associated universities for research in astronomy , under contract to nasa .
this research also made use of public databases hosted by simbad , maintained by cds , strasbourg , france .
antolin , p. , & rouppe van der voort , l. 2012 , , 745 , 152 aschwanden , m. j. , stern , r. a. , & gdel , m. 2008 , , 672 , 659 ayres , t. r. 2015 , , 149 , 58 ayres , t. r. , brown , a. , harper , g. m. , et al .
2003 , , 583 , 963 ayres , t. r. , brown , a. , & harper , g. m. 2005 , , 627 , l53 ayres , t. r. , fleming , t. a. , simon , t. et al . 1995 , , 96 , 223 ayres , t. , & france , k. 2010 , , 723 , l38 benz , a. o. , & gdel , m. 2010 , , 48 , 241 candelaresi , s. , hillier , a. , maehara , h. , brandenburg , a. , & shibata , k. 2014 , , 792 , 67 cook , j. w. , keenan , f. p. , dufton , p. l. , et al .
1995 , , 444 , 936 drake , s. , osten , r. , page , k. l. , et al .
2014 , the astronomer s telegram , 6121 , 1 fang , c. , & ding , m. d. 1995 , , 110 , 99 fender , r. p. , anderson , g. e. , osten , r. , et al .
2015 , , 446 , l66 flower , d. r. , & nussbaumer , h. 1975 , , 45 , 145 green , j. c. , froning , c. s. , osterman , s. , et al .
2012 , , 744 , 60 gry , c. , & jenkins , e. b. 2014 , , 567 , 58 gdel , m. 2004 , , 12 , 71 gdel , m. , schmitt , j. h. m. m. , benz , a. o. , & elias , n. m. , ii 1995 , , 301 , 201 hawley , s. l. , allred , j. c. , johns - krull , c. m. , et al .
2003 , , 597 , 535 keenan , f. p. , crockett , p. j. , aggarwal , k. m. , jess , d. b. , & mathioudakis , m. 2009 , , 495 , 359 kimble , r. a. , woodgate , b. e. , bowers , c. w. , et al .
1998 , , 492 , l83 knig , b. , guenther , e. w. , woitas , j. , & hatzes , a. p. 2005
, , 435 , 215 kramida , a. , ralchenko , yu . , reader , j. , and nist asd team 2013 , nist atomic spectra database ( ver .
5.1 ) , [ online ] .
available : http://physics.nist.gov/asd [ 2014 , august 21 ] .
national institute of standards and technology , gaithersburg , md .
maehara , h. , shibayama , t. , notsu , s. , et al .
2012 , , 485 , 478 mitra - kraev , u. , harra , l. k. , gdel , m. , et al .
2005 , , 431 , 679 osten , r. a. , godet , o. , drake , s. , et al .
2010 , , 721 , 785 peter , h. 2006 , , 449 , 759 redfield , s. , & linsky , j. l. 2008 , , 673 , 283 robinson , r. d. , linsky , j. l. , woodgate , b. e. , & timothy , j. g. 2001 , , 554 , 368 saar , s. h. , & bookbinder , j. a. 1998 , in tenth cambridge workshop on cool stars , stellar systems , and the sun , ( eds ) r. a. donahue and j. a. bookbinder , asp conf . ser . 154 , 1560 schaefer , b. e. 1989 , , 337 , 927 schaefer , b. e. , king , j. r. , & deliyannis , c. p. 2000 , , 529 , 1026 shibayama , t. , maehara , h. , notsu , s. , et al . 2013 , , 209 , 5 tsang , b. t. h. , pun , c. s. j. , di stefano , r. , li , k. l. , & kong , a. k. h. 2012 , , 754 , 107 woodgate , b. e. , kimble , r. a. , bowers , c. w. , et al .
1998 , , 110 , 1183 young , p. r. , tian , h. , & jaeggli , s. 2015 , , 799 , 218 | ek draconis ( hd129333 : g1.5 v ) is a well - known young ( 50 myr ) solar analog . in 2012 , _ hubble space telescope
_ returned to ek dra to follow up a far - ultraviolet ( fuv ) snapshot visit by cosmic origins spectrograph ( cos ) two years earlier .
the brief snap pointing had found surprisingly redshifted , impulsively variable subcoronal `` hot - line '' emission of 1400 ( @xmath0 k ) .
serendipitously , the 2012 follow - on program witnessed one of the largest fuv flares ever recorded on a sunlike star , which again displayed strong redshifts ( downflows ) of 3040 km s@xmath1 , even after compensating for small systematics in the cos velocity scales , uncovered through a cross - calibration by space telescope imaging spectrograph ( stis ) .
the ( now reduced , but still substantial ) @xmath210 km s@xmath1 hot - line redshifts outside the flaring interval did not vary with rotational phase , so can not be caused by `` doppler imaging '' ( bright surface patches near a receding limb ) .
density diagnostic ] 1400 multiplet line ratios of ek dra suggest @xmath3 @xmath4 , an order of magnitude larger than in low - activity solar twin @xmath5 centauri a , but typical of densities inferred in large stellar soft x - ray events . the self - similar fuv hot - line profiles between the flare decay and the subsequent more quiet periods , and the unchanging but high densities ,
reinforce a long - standing idea that the coronae of hyperactive dwarfs are flaring all the time , in a scale - free way ; a _ flare - ona _ if you will . in this picture
, the subsonic hot - line downflows probably are a byproduct of the post - flare cooling process , something like `` coronal rain '' on the sun .
all in all , the new stis / cos program documents a complex , energetic , dynamic outer atmosphere of the young sunlike star . |
the concept of feedback due to outflows is a potential key ingredient in understanding the coevolution of galaxies and black holes ( silk & rees 1998 , fabian 1999 , wyithe & loeb 2003 ) .
recent analytic estimates and simulations demonstrate the importance of feedback from winds / outflows for instance in cosmic downsizing ( scannapieco et al .
2005 ) , in fixing the @xmath0 relation ( di matteo et al .
2005 ) , and in determining galaxy colors ( springel et al . 2005 ) by regulating star formation in the host galaxy .
powerful gaseous outflows in active galactic nuclei ( agn ) deposit mass , energy and metals in the interstellar medium of the galaxy and the intergalactic medium or intracluster medium ( colbert et al .
1996 , churazov et al .
2001 , moll et al . 2007 ) .
agn winds also play a potentially important role in unified models of agn ( elvis 2000 , 2006 ) .
observational evidence for outflows in agn exists on various scales and in all wavebands from the radio ( morganti et al .
2005 , gallimore et al .
2006 ) , to the ir ( rodriguez - ardila et al .
2006 ) , optical ( das et al . 2005 ) , uv ( sulentic et al .
2007 , rodriguez hidalgo et al .
2007 , crenshaw & kraemer 2007 ) , and x - ray band ( chelouche & netzer 2005 , krongold et al .
2007 ) ; see veilleux et al .
( 2005 ) for a review .
there are several lines of evidence that outflows are particularly strong in narrow - line seyfert 1 ( nls1 ) galaxies .
nls1 galaxies are a subclass of agn with extreme emission - line and continuum properties which appear to be in part driven by their high eddington ratios , low black hole masses , and other parameters ( e.g. , osterbrock & pogge 1985 , boroson 2002 , grupe 2004 ; see komossa 2008 for a review ) .
the high eddington ratios likely lead to strong , radiation - pressure driven outflows . as a class of objects with low black hole masses , high accretion rates and strong winds , the location of nls1 galaxies on the @xmath0 plane is of special interest ( mathur et al .
2001 ; see our sect .
komossa & xu ( 2007 ) have shown that nls1 galaxies follow the same @xmath0 relation as normal and broad - line seyfert 1 ( bls1 ) galaxies , if the width of the [ sii]6716,6731 emission lines is used as a surrogate for stellar velocity dispersion .
the width of [ oiii]5007 ( after removal of asymmetric blue wings ) is still a good proxy for stellar velocity dispersion in bls1 and nls1 galaxies , with one important exception .
a subset of nls1 galaxies deviates systematically from the @xmath0 relation , and all of these are characterized by high [ oiii ] _
blueshifts_. at the same time , the [ sii]-based measurements of velocity dispersion of the same objects still place them _ on _ the @xmath0 relation ( fig .
1 ) . therefore
, the [ oiii ] lines of these particular galaxies were not suitable to estimate @xmath2 .
however , the independent question is raised as to which mechanism drives these [ oiii ] line blueshifts , and what we can learn from them regarding the nature of these systems and perhaps the evolutionary state of nls1 galaxies .
that is the topic of this study .
a measurable difference in blueshifts and widths of different nlr emission lines has been recognized early ( review by osterbrock 1991 ) .
the phenomenon of strong [ oiii ] blueshifts exceeding one several hundred kms@xmath1 ( so - called blue outliers ) , is more rare .
it has been detected early in individual objects ( phillips 1976 ) but has only been studied systematically recently ( zamanov et al .
2002 , marziani et al .
2003 , aoki et al .
2005 , boroson 2005 , bian et al .
these studies have shown that blue outliers have high eddington ratios and small balmer line widths ( all of them with fwhm(h@xmath3 ) @xmath4 4000 kms@xmath1 , and most of them with fwhm(h@xmath3 ) @xmath4 2000 kms@xmath1 ; marziani et al .
it has remained uncertain whether there is ( bian et al .
2005 ) or is not ( aoki et al .
2005 ) a direct correlation between [ oiii ] blueshift and eddington ratio .
common to most studies is the presence of a strong correlation between [ oiii ] blueshift and [ oiii ] line width .
this correlation is also known among iron coronal lines of bls1 galaxies ( penston et al .
1984 , erkens et al .
the reason is not yet well understood .
outflows are thought to play a role in explaining the phenomenon . except boroson ( 2005 )
all sample studies concentrated on the [ oiii]-h@xmath3 region of optical spectra of blue outliers in order to examine the phenomenon , while for the first time we include information from all the detected optical nlr lines .
we report the detection of strong correlations and discuss consequences for the nature of blue outliers and for dynamical nlr models .
this paper is organized as follows . in sect .
2 we describe the sample selection and provide details on the data analysis .
results on trends and correlations are presented in sect .
3 which are then discussed in sect .
4 . a summary and conclusions are given in sect .
5 . some individual objects turn out to be remarkable .
notes on them are provided in an appendix .
we use a cosmology with @xmath5=70 kms@xmath1mpc@xmath1 , @xmath6=0.3 and @xmath7=0.7 throughout this paper .
the nine nls1 galaxies of the present work were drawn from the sample of xu et al .
( 2007 , 2008 ) . the original sample selection and standard data reduction procedures were described in detail in that work . in brief , the sample consists of nls1 galaxies from the catalogue of vron - cetty & vron ( 2003 ) and a comparison sample of bls1 galaxies from boroson ( 2003 ) at redshift @xmath8 , which have sloan digital sky survey ( sdss ) dr3 ( abazajian et al . 2005 ) spectra available and which have detectable low - ionization lines ( presence of [ sii]6716,6731 , with s / n@xmath95 ) . the bls1 and nls1 galaxy samples have similar redshift and absolute magnitude distributions .
xu et al . ( 2007 ) corrected the sdss spectra for galactic extinction , decomposed the continuum into host galaxy and agn components , and then subtracted the host galaxy contribution and the feii complexes from the spectra .
emission line profiles of the galaxies were fit with gaussians using the iraf package specfit ( kriss 1994 )
. measured fwhms were corrected for instrumental broadening .
re - classification after spectral emission - line fitting led to 39 bls1 and 55 nls1 galaxies in the sample .
we focus here on the nine galaxies with the highest blueshifts of [ oiii ] ( tab .
1 ) which were identified by komossa & xu ( 2007 ; kx07 hereafter ) .
results from the complete nls1 ( and bls1 ) sample are shown for comparison purposes , and in order to identify trends across the whole bls1 nls1 blue - outlier population .
when we report measurements of optical feii strength ( feii4570 ) , this is the integrated flux of the feii emission complex between the rest wavelengths 4434 and 4684 .
we refer to the flux sum of the two sulphur lines , [ sii]6716 and [ sii]6731 , as [ sii]6725 .
the balmer lines were decomposed into three components , narrow core ( h@xmath10 ; fwhm fixed to that determined for [ sii]6716,6731 ) , and two broad components .
no physical meaning is ascribed to the two separate broad components ; they merely serve as a mathematical description ( see xu et al .
2008 for alternative lorentzian fits ) .
the final width of the broad - line emission , h@xmath11 , is determined as the fwhm of the sum of the two gaussians . with the exception of [ oiii ]
, forbidden lines are well represented by single gauss profiles . the total [ oiii ] emission - line profile , [ oiii]@xmath12 , was decomposed into two gaussian components , a narrow core ( [ oiii]@xmath13 ) and broad base .
we distinguish between two types of [ oiii ] spectral complexity : ( 1 ) the presence of such a broad base which tends to be blue - asymmetric ( e.g. , heckman et al .
1981 ) , and is referred to as `` blue wing '' , and ( 2 ) systematic blueshifts of the whole core of [ oiii ] .
objects which show this latter phenomenon are called `` blue outliers '' ( zamanov et al .
2002)_c}>$]150 kms@xmath1 in order to refer to an object as [ oiii ] blue outlier ; while zamanov et al .
( 2002 ) use this terminology for objects with @xmath14}>$]250 kms@xmath1 , where [ oiii ] outflow velocity is measured relative to h@xmath3 . ] .
measurements of the fwhm and blueshift of [ oiii ] reported in this work refer to the core of the emission - line , unless noted otherwise . ideally , measurements of the velocity shift of [ oiii ] should be done relative to the galaxy restframe , as defined by stellar absorption lines from the host galaxy . however , these features are generally weak or absent in our galaxies ( and in nls1 galaxies in general ) .
we therefore measure velocity shifts of [ oiii ] and of all other lines relative to [ sii ] .
we use positive velocity values to refer to blueshifts , negative ones for redshift .
the shifts of h@xmath3 and other low - ionization lines ( [ oi]6300 , nii[6584 ] , [ oii]3727 ) agree well with [ sii ] , while other high - ionization lines ( [ neiii]3861 , [ nev]3426 , and iron coronal lines ) are characterized by high blueshifts ( sect .
if lines were too faint to be fit by a gaussian of free width , we used fixed width instead ( fixed to fwhm([sii ] ) for low - ionization lines or to fwhm([oiii]@xmath13 ) for high - ionization lines ) and determined the central wavelength to measure the blueshift .
results are reported in tab .
1 . since the redshifts provided by the sdss pipeline are determined based on all strong detected emission lines and are therefore influenced by the blueshifted [ oiii ] , we have re - measured redshifts based on h@xmath10 .
it is these redshifts that are given in tab .
we have measured the average [ oiii]@xmath12/[sii]6725 ratios of our bls1 and nls1 galaxy sample .
we find that , in terms of this ratio , their nlrs are similar , with @xmath4[oiii]@xmath12/[sii]6725@xmath9=5 in bls1 galaxies , and @xmath4[oiii]@xmath12/[sii]6725@xmath9=4 in nls1 galaxies .
notes on individual objects are given in the appendix .
here , we briefly comment on the detection of [ fex]6375 in two blue outliers . in pg1244 + 026 [ fex ]
is blended with [ oi]6365 . for decomposition , the width of [ oi]6365 was fixed to that of [ oi]6300 .
the two - component fit then reproduces the expected ratio of [ oi]6300/[oi]6365 @xmath15 3 .
[ fex ] is blueshifted by 640 kms@xmath1 . in rxj0135@xmath160043 , [ fex ]
overlaps with atmospheric o@xmath17 which is imperfectly corrected for .
therefore , the width of [ fex ] was fixed to that of [ oiii ] .
this gives an outflow velocity of 990 kms@xmath1 . in order to see
how much the fitting procedure affects measurements of [ oiii ] fluxes , asymmetries and blueshifts , we have repeated the spectral analysis several times under different conditions and assumptions : we have fit the host - galaxy - corrected , feii subtracted spectrum and for comparison the original sdss spectrum without any of these corrections .
we have re - fit the [ oiii ] profile in a number of different ways , and with different gaussian decompositions .
the [ oiii ] profile was fit using the red half of the profile only , or different fractions of the total profile , or only the upper 30% of the profile , or with a single and a double gaussian profile .
fitting the profile with one single gaussian component maximizes the blueshift in objects with significant [ oiii ] blue wings . in all other cases , uncertainties in line blueshifts due to non - subtraction of a host galaxy contribution and due to different line fitting procedures
are typically much less than 50 kms@xmath1 .
the blueshift independently determined from the weaker line [ oiii]4959 also agrees with [ oiii]5007 within better than 50 kms@xmath1 .
four of our galaxies are also in the sample of boroson ( 2005 ) .
his [ oiii ] blueshift measurements agree with ours within typically better than @xmath1810 kms@xmath1 .
uncertainties in fwhm of the core of [ oiii ] are typically 30% for [ oiii ] lines with extra blue wing ( smaller in others ) , and are due to uncertainties in decomposition .
we note in passing that the frequency of [ oiii ] blue wings in blue outliers is neither particularly high nor low ( see also boroson 2005 ) .
in order to facilitate a systematic discussion of nlr models of blue outliers ( sect . 4.3 ) , we have investigated whether the non - shifted low - ionization lines have a ( faint ) high - ionization [ oiii ] counterpart and vice versa , other high - ionization lines show blueshifts , too ( sect .
3.3.1 ) , but they are too faint for multi - component line decompositions . among the non - shifted emission lines , [ sii ]
is the strongest unblended line , and we therefore focus on [ sii ] for the corresponding estimates in this section . ] .
we did this representatively for the three galaxies with the highest [ oiii ] blueshifts ( sbs0919 + 515 , sdssj11555 + 0107 and rxj01354@xmath160043 ) , and for the galaxy with the highest s / n ( pg1244 + 026 ) .
we have examined the following three questions : \(1 ) is there any [ oiii ] emission at zero blueshift present in the spectrum ( which could originate from the inner nlr ; or be counterpart to the low - ionization emission lines from the outer nlr ) ? in order to test this , we have enforced an extra gaussian line contribution to [ oiii ] at zero blueshift ( or at @xmath19 } = 50 $ ] kms@xmath1 , the average value of our nls1 sample excluding blue outliers ) and of fixed fwhm([oiii]@xmath20)=fwhm([sii ] ) .
we find that , if present , it must be very weak .
the average ratio [ oiii]@xmath12/[sii]6725 in our bls1 sample is 5 , while in our nls1 sample it is 4 .
this value of 4 is much higher than the upper limit on the extra component which was fit to the observed spectra : typically , [ oiii]@xmath20/[sii]6725 @xmath21 , and always , [ oiii]@xmath20/[sii]6725 @xmath22 .
since the ratio of [ oiii]/[sii ] of nearby seyfert galaxies generally decreases in dependence of radius ( in models and observations ; komossa & schulz 1997 , bennert et al . 2006 ) and since there must be an [ oiii ] contribution from the outer [ sii ] emitting clouds , the limits derived on any non - blueshifted [ oiii ] contribution from the _ inner _ nlr are very tight .
\(2 ) does the blueshifted [ oiii ] have a blueshifted h@xmath3 counterpart ?
i.e. , how much highly blueshifted h@xmath3 could be ` hidden ' in the h@xmath3 profile ? in order to check this , we have re - fit h@xmath3 adding an additional gaussian to describe its profile with parameters of this extra component fixed at @xmath23_c}$ ] , fwhm(h@xmath24)=fwhm([oiii]@xmath13 ) and an intensity ratio [ oiii]@xmath13/h@xmath24=10 .
we find that such a weak component could generally be hidden in the h@xmath3 profile .
\(3 ) does the blueshifted [ oiii ] have a blueshifted low - ionization counterpart in [ sii ] ?
this is not the case . for typical
_ spatially averaged _ ratios of [ oiii]/[sii]6725 @xmath154 of our nls1 sample , blueshifted [ sii ] lines would have generally been detectable , but are absent .
very faint blueshifted [ sii ] , as it may arise in the _ inner _ nlr , can not be excluded with the present data .
in summary , we find ( 1 ) little evidence for zero - blueshift [ oiii ] emission ( the amount of emission consistent with the spectra likely originates from the outer parts of the nlr which also emits in [ sii ] and other low - ionization lines ) ; ( 2 ) an extra blueshifted component in h@xmath3 the expected counterpart to blueshifted [ oiii] can be hidden in the h@xmath3 profile ; and ( 3 ) there is no detectable highly blueshifted component in the low - ionization lines .
black hole masses were estimated by using the radius(@xmath25)-luminosity(@xmath26 ) relation of bls1 galaxies as reported by kaspi et al .
( 2005 ) and the width of h@xmath3 .
the luminosities at 5100 were taken from xu et al .
( 2008 ) and are based on sdss g * and r * magnitudes corrected for galactic extinction .
since we lack multi - wavelength spectral energy distributions , the bolometric luminosity @xmath27 was estimated using a standard bolometric correction of @xmath28 ( kaspi et al . 2000 ) .
@xmath29 was calculated from the black hole mass , according to @xmath30 m@xmath31/m@xmath32 erg / s .
typical errors in individual bh mass estimates can be as large as 0.5 dex , and arise from the use of single - epoch data , and uncertainties in h@xmath3 decomposition and in host galaxy contribution .
the nlr density was derived from the density - sensitive [ sii ] intensity ratio , [ sii]6716/[sii]6731 ( xu et al . 2007 ) .
we have checked for ( 1 ) trends of outflow velocity with other ( atomic ) parameters within each single spectrum , and ( 2 ) for trends across our sample of blue outliers ( do all blue outliers have high eddington ratios , high black hole masses , etc . ? ) , and ( 3 ) for trends across our whole sample of bls1 and nls1 galaxies .
we detect a strong correlation between emission - line blueshift and ionization potential ( fig .
[ ionpot ] ) .
while the low - ionization forbidden lines of [ oi ] , [ oii ] and [ nii ] , and the balmer lines h@xmath33 and h@xmath3 are at velocities very similar to [ sii ] , high - ionization lines on the other hand show strong blueshifts which increase with ionization potential .
two galaxies ( sdssj115533.50 + 010730.4 and pg1244 + 026 ) show extreme blueshifts in [ fe x ] , on the order of 600 - 1000 kms@xmath1 .
there is also a correlation between outflow velocity and critical density of the individual line transitions ( fig .
[ ionpot ] ) , but it is less tight in the sense that [ oi]6300 , which has low ionization potential but high critical density , does not follow the trend . we have correlated the [ oiii ] blueshifts ( of the blue outliers , and of our sample as a whole ) with various other line parameters ( line widths , line ratios ) and with galaxy properties ( absolute magnitude , black hole mass , eddington ratio , nlr density ) . the strongest correlation is the one between [ oiii ] blueshift and [ oiii ] line width ( spearman rank correlation coefficient @xmath34=0.6 for the whole nls1 sample and even higher for just the blue outliers ; fig .
[ ionpot ] ) . within the nls1 sample ,
there is a trend that blue outliers preferentially avoid low eddington ratios @xmath35 and low feii / h@xmath3 ratios .
most blue outliers have small ratios of [ oiii]/h@xmath36 .
we do not find trends of outflow velocity with nlr density @xmath37 , black hole mass , absolute magnitude @xmath38 , and the width of the broad component of h@xmath3 ( fig .
[ divcorr ] ) . however , while trends are absent within the nls1 population itself , it is interesting to note that the presence of blue outliers amplifies trends which become apparent when the bls1 galaxies are added to the correlation plots ( fig .
[ divcorr ] ) . among our sample
, blue outliers only occur in nls1 galaxies . at a velocity @xmath19 } \approxgt 150 $ ] kms@xmath1 , we have a blue outlier fraction of 16% .
if we use @xmath19 } \approxgt 250 $ ] kms@xmath1 instead , as in zamanov et al .
( 2002 ) , the fraction of blue outliers among nls1 galaxies is 5% .
investigating the location of different types of galaxies on the @xmath0 plane , and their potential evolution across the plane , is of great interest in the context of galaxy formation and evolution models . in agn ,
stellar velocity dispersion is often difficult to measure , and the width of [ oiii ] has become a convenient proxy for stellar velocity dispersion ( e.g. , terlevich et al . 1990 , whittle 1992 , nelson & whittle 1996 , nelson 2000 , shields et al .
2003 , boroson 2003 , greene & ho 2005 , netzer and trakhtenbrot 2007 , salviander et al . 2007 ) after removing [ oiii ] blue wings and excluding galaxies with powerful kiloparsec - scale linear radio sources .
the scatter in the relation is larger than in the original @xmath39 relation ( ferrarese & merritt 2000 , gebhardt et al .
2000 ) , and indicates secondary influences on the gas kinematics ( e.g. , nelson & whittle 1996 , rice et al .
2006 ) .
mathur et al . ( 2001 ) pointed out the importance of studying the locus of nls1 galaxies on the @xmath0 plane , which was the focus of a number of subsequent studies ( wang & lu 2001 , wandel 2002 , botte et al .
2004 , 2005 , bian et al .
2004 , grupe & mathur 2004 , barth et al .
2005 , greene & ho 2005 , mathur & grupe 2005a , b , zhou et al .
2006 , ryan et al .
2007 , watson et al .
2007 , komossa & xu 2007 ) .
any such study would involve one important step : the distinction between true outliers from the @xmath0 relation on the one hand , and apparent outliers on the other hand .
apparent outliers would only appear to be off - set from the @xmath0 relation because either the choice of line width as a measure of stellar velocity dispersion was unsuitable , or the choice of line and continuum parameters as a measure of black hole mass was unsuitable .
kx07 have shown that those nls1 galaxies of their sample which deviate significantly from the @xmath0 relation of normal ( and bls1 ) galaxies are all characterized by high [ oiii ] blueshifts .
while these [ oiii ] lines were therefore not suited as surrogate for stellar velocity dispersion , the [ sii]-based measurements of velocity dispersion of the very same objects still located them on the @xmath0 relation . almost all galaxies which deviate most strongly from the @xmath0 relation ( see our fig .
1 ) show @xmath14 } > 150 $ ] kms@xmath1 .
these six objects , plus three additional ones which also show @xmath14 } > 150 $ ] kms@xmath1 define the nine blue outliers that are the target of the present study . in order to measure systematically the deviation of a galaxy from the @xmath39 relation ,
we define the quantity @xmath40 ( as in kx07 ) , where @xmath41 is the observed emission - line velocity dispersion , and @xmath42 is the stellar velocity dispersion predicted from the @xmath39 relation of ferrarese & ford ( 2005 ) .
we find a strong correlation between [ oiii ] blueshift @xmath14}$ ] and @xmath43 ( fig .
this finding demonstrates that [ oiii ] velocity shift systematically affects the deviation of an object from the @xmath0 relation ( see also boroson 2005 ) .
we further find that a correlation between @xmath44 and @xmath43 ( e.g. , mathur & grupe 2005 , greene & ho 2005 ; see also netzer & trakhtenbrot 2007 ) is only present in our sample if we include blue outliers ; when they are removed from the sample , no correlation remains ( last panel of our fig .
[ divcorr ] , see also kx07 ) .
two of our nls1 blue outliers have independent bh mass estimates . from x - ray variability ,
czerny et al . ( 2001 ) estimated @xmath45 for pg1244 + 062 which agrees well with our estimate from applying the @xmath25-@xmath26 relation ( @xmath46 ) . for rx01354@xmath160043 we directly measured @xmath47 from stellar absorption lines ( see appendix ) .
the value agrees well with @xmath48}$ ] and puts rxj01354@xmath160043 almost perfectly on the @xmath49-@xmath47 relation ( fig .
1 ) . these findings demonstrate the importance of measuring [ oiii ] blueshifts , and removing objects with high blueshifts from a sample before putting it on the @xmath50}$ ] relation .
the independent question arises as to what may cause the blueshifts of these [ oiii ] outliers .
the phenomenon of slightly different blueshifts and line widths of nlr emission lines has been recognized since the early days of agn spectroscopy and is generally traced back to a certain stratification of the nlr in the sense that high - ionization lines are produced preferentially at small distances from the core , while low - ionization lines are preferentially produced at larger radii ( e.g. , osterbrock 1991 ) .
the presence of a strong [ oiii ] blueshift on the order of several hundred kms@xmath1 was noticed early in the prototype nls1 galaxy izw1 .
also note systematic blueshifts in high - ionization blr emission lines in quasar samples ( e.g. , gaskell 1982 , sulentic et al .
( e.g. , phillips 1976 , vron - cetty et al .
the galaxy also shows very blueshifted ir coronal lines ( schinnerer et al .
1998 ) and blueshifted uv high - ionization broad lines ( laor et al . 1997 ) .
several recent studies systematically examined the phenomenon of blue outliers in larger samples of type 1 seyfert galaxies ( zamanov et al .
2002 , marziani et al .
2003 , aoki et al .
2005 , boroson 2005 , bian et al . 2005 ) .
according to these studies blue outliers have high eddington ratios and small blr balmer line widths . however , not all sources with high eddington ratios are blue outliers . while bian et al .
( 2005 ) further reported a correlation between blueshift and eddington ratio of their 7 blue outliers , aoki et al .
( 2005 ) did not find such a correlation for 16 objects .
a strong correlation between [ oiii ] blueshift and [ oiii ] line width is often seen , and is usually interpreted as evidence for outflows .
zamanov et al .
( 2002 ) noticed that most of their blue outliers also show high blueshifts in civ1549 ( but this phenomenon is not exclusive to blue outliers ; sulentic et al .
aoki et al .
( 2005 ) reported a trend that blue outliers have high black hole masses ( @xmath51 m@xmath32 ) .
all previous studies of samples of blue outliers focused on the [ oiii]-h@xmath3 spectral region in order to explore the phenomenon , with the exception of boroson ( 2005 ) .
although boroson ( 2005 ) measured the positions of other nlr lines in order to establish a systemic reference for the [ oiii ] line properties , we have , for the first time , measured the widths and strengths of all the optical nlr lines in blue outliers in order to study the nature of blue outliers and explore the dynamics of the nlr . for each galaxy ,
we find a strong correlation between emission line blueshift and ionization potential of the line - emitting ion . coincidentally ,
higher ionization potentials of the ions in question also come with higher critical densities of the forbidden line transitions observed from the respective ions .
therefore , a correlation between outflow velocity and ionization potential would generally imply a correlation with critical density , and vice versa , raising the question which of the two is the fundamental correlation ; density stratification or ionization stratification .
an important exception to the above rule is [ oi]6300 , which has zero ionization potential , while its critical density ( @xmath52 @xmath53 ) is relatively high .
[ oi ] is detected in several of our nls1 galaxies , and we find that [ oi ] follows the trend in ionization potential , but not in critical density , arguing that the former correlation is the underlying one ( fig .
[ ionpot ] ) left . ] .
we confirm the previously known correlation between [ oiii ] blueshift and line width .
blue outliers only occur among the nls1 galaxies of our sample ( i.e. , sources with high eddington ratios and small balmer line widths ) , but within the nls1 population , blue outliers do not show a strong correlation with eddington ratio even though they preferentially avoid low ratios .
no correlation of outflow velocity with black hole mass , absolute magnitude , and with the width of the broad component of h@xmath3 is found . like in other samples ,
the number of objects is small , and larger samples of blue outliers are needed to confirm the weak trends .
it has occasionally been suggested that [ oiii ] in blue outliers appears broadened by orientation effects in the sense that we look face - on on the central engine , and into an outflow ( zamanov et al .
2002 , marziani et al .
2003 , boroson 2005 ) .
if that scenario is correct , we can use blue outliers as test - beds for orientation - dependent models of nls1 galaxies which allow for the possibility that the h@xmath3 line of nls1 galaxies is narrowed due to viewing angle effects . if , in blue outliers , the blr was in a plane and if we viewed it face - on , blue outliers should have the smallest h@xmath3 widths among nls1 galaxies ( here , we temporarily classify nls1 galaxies independently of their h@xmath3 width , but only use the ratios [ oiii]/h@xmath3 and feii / h@xmath3 which still makes all of our blue outliers nls1 galaxies : in all cases [ oiii]/h@xmath54 , and feii / h@xmath55 ) .
however , we do not find any trend for small widths of the broad balmer lines among the objects of our sample ( see fig .
[ divcorr ] ) .
the [ oiii ] lines of several objects still show the phenomenon of blue wings , arguing against the previous suggestion that the classical [ oiii ] emission in blue outliers is absent and we actually only see the blue wing .
[ however , occasionally , the presence of a strong blue wing could mimic a blue outlier , if the wing is not spectroscopically resolved from the core of the line ( e.g. , grandi 1977 , holt et al .
2003 ) ] .
there is good evidence that the motion of nlr clouds is strongly influenced by the bulge gravitational potential ( e.g. , vron 1981 , whittle et al .
1992 , nelson & whittle 1996 ; see the prev .
4.1 ) . at the same time
, there is also evidence for radial motions in nlrs based for instance on blue asymmetries of [ oiii ] profiles ( e.g. , de robertis & osterbrock 1984 , veilleux 1991 , whittle 1992 , vron - cetty et al .
2001 ) , on correlations of blue wing blueshift with the eddington ratio and other arguments ( xu et al .
2007 ) , and on spatially resolved [ oiii ] velocity shifts ( e.g. , schulz 1990 , das et al . 2005 ) .
the existence of [ oiii ] blue wings can formally be equally well interpreted in terms of inflows or outflows , plus selective obscuration .
generally , there is a preference for the outflow interpretation .
regarding theoretical models for outflows , a lot of work has focussed on magnetocentrifugal winds and radiatively driven outflows from the accretion disk region ( see knigl 2006 , everett 2007 , proga 2007 for reviews ) . these models
have successfully been applied to bal flows and the blr emission lines , are supported by recent observations ( young et al .
2007 ) , and have recently been extended to spatial scales typical of the very inner nlr ( proga et al . 2008 ) .
however , little is known about the formation or continuation of such winds on much larger scales on the order of 100 pc
kpc characteristic for nlrs .
mechanisms suggested to be relevant in the nlr ( and clr ) include radiation pressure acting on gas ( binette et al .
1997 , das et al . 2007 ) and on dust grains embedded in the clouds ( binette 1998 , dopita et al .
2002 ) , and the entrainment of nlr clouds in hot winds ( krolik & vrtilek 1984 , schiano 1986 , mathews & veilleux 1989 , smith 1993 , everett & murray 2007 ) .
the presence of collimated outflows of radio plasma in form of jets and their local interaction with nlr clouds has been observed directly ( e.g. , falcke et al .
spatially resolved imaging spectroscopy of nearby agn has produced several good examples of spatial coincidences between radio jets and line blueshifts ( e.g. , riffel et al .
2006 , morganti et al . 2007 ) , while in other cases the jet - cloud interaction is very weak and does not significantly affect the local nlr velocity field ( e.g. , cecil et al .
2002 , das et al . 2005 ) . in some cases ,
jets are absent but emission lines are still blueshifted ( barbosa et al .
2008 , in prep ) .
in agn with strong starburst activity winds could also be starburst driven ( e.g. , rupke et al . 2005 ) . regarding nls1 galaxies , their high eddington ratios make the presence of radiation - pressure driven _ winds _
( on the accretion disk scale ) very likely .
less certain and understood is the efficiency of _ jet _ launching under high - accretion rate conditions .
again , these mechanisms refer to the innermost agn region , not to larger scales
. blue outliers with their extreme velocity shifts place particularly tight constraints on models for agn outflows on large , i.e. nlr , scales .
we discuss several nlr models for blue outliers in turn .
zamanov et al .
( 2002 ) and marziani et al .
( 2003 ) suggested that blue outliers posses a very compact nlr , perhaps the result of the youth of these nls1 galaxies which did not yet develop a full nlr .
their idea was based on the assumption that blue outliers would only posses [ oiii ] emission and lack other narrow emission lines .
however , we do find a classical nlr in terms of the presence of low - ionization lines .
further , the fact that their [ sii ] line widths put the blue outliers on the same @xmath0 relation with bls1 and normal galaxies strongly indicates that their outer nlr is a quiescent classical nlr similar to that of other bls1 and nls1 galaxies .
the strong blueshifts we detect in the high - ionization component raises the possibility that we see two independent components ; a classical nlr , plus an independent outflow component which could be due to a disk wind , a wind from the torus or in form of a lowly ionized warm absorber .
however , this interpretation is very unlikely , because the non - blueshifted [ oiii ] counterpart to the low - ionization lines , as it would be expected from _ the inner part _ of a classical nlr , is weak or absent .
this fact leads us almost inevitably to the third and fourth possibility .
the fact that we do not detect a classical inner nlr , in form of a second , non - blueshifted [ oiii ] emission component , implies that processes directly modify / disturb at least the inner nlr ; perhaps in form of radio - jet cloud interaction .
however , the strong dependence of emission line velocity shift on ionization potential that we detect ( fig . [ ionpot ] ) indicates that we see a highly stratified , photoionized medium , not a local interaction due to shocks and locally disturbed velocity fields .
the phenomenologically most straight - forward solution is one in which the ( clr and ) nlr clouds themselves follow a decelerated outflow .
this scenario requires an efficient driving mechanism and perhaps an efficient deceleration mechanism .
one possibility is direct acceleration of the nlr clouds by radiation pressure acting on gas or dust ; another is the presence of a hot wind which entrains the nlr clouds ( see sect .
if the wind forms a decelerating outflow , the entrained clouds of the clr/ inner nlr would have the highest outflow velocities , respectively , which would decrease in dependence of core distance and would leave the outer , low - ionization part of the nlr unaffected .
zamanov et al .
( 2002 ) reported a linkage of [ oiii ] blue - outlierness with broad - line civ blueshift .
could the same wind persist from the disk , to the high - ionization blr [ but leave the bulk of the blr unaffected ; by blowing perpendicular ? ] , and on to the clr and inner nlr ? in this picture , the nlrs of bls1 galaxies , nls1s , and blue outliers would be intrinsically similar , and the motion of nlr clouds generally dominated by the bulge gravitational potential . which emission line regions partake in an outflow would depend on the efficiency / operating distance of the wind
. it could be present in all agn , but would generally only affect the clr ( in agn with blueshifted iron coronal lines but quiescent nlrs ) , or level off even before reaching the clr . while phenomenologically successful , the question is raised as to the origin of the wind / outflow on the one hand , and the entrainment mechanism and cloud stability against disruptive instabilities ( mathews & veilleux 1989 , schiano et al .
1995 ) on the other hand .
we discuss several possibilities in turn .
[ [ cloud - acceleration - by - radiation - pressure - acting - on - dust ] ] cloud acceleration by radiation pressure acting on dust + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + radiation pressure acting on dust grains embedded in the gas clouds is an efficient way of cloud acceleration ( e.g. , binette 1998 , dopita et al .
2002 , fabian et al .
2006 ) . in blue outliers ,
this mechanism is likely not at work , because we do not expect dust in the high - ionization blr ( we are assuming here that the high - ionization part of the blr partakes in the outflow , motivated by the results of zamanov et al . (
2002 ) described in sect .
. some dust might be present in the clr ( even a small admixture of dust would still lead to efficient radiative acceleration , and generally still predict sufficiently strong gas - phase iron lines ; binette 1998 ) . while , observationally , dust - rich nlrs could be present in individual blue outliers ( e.g. , rxj0135@xmath160043 ) , most of them have very blue optical continua arguing against the presence of an excess of dusty gas along the line of sight .
these arguments make cloud acceleration by radiation pressure acting on dust grains an unlikely scenario .
[ [ cloud - entrainment - in - collimated - radio - plasma ] ] cloud entrainment in collimated radio plasma
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + outflowing radio plasma in form of jets is known to be present and to reach large distances from the nucleus in agn , so could plausibly affect several emission - line regions on its way .
the phenomenon of cloud entrainment in radio plasma has been directly observed and studied in starforming regions ( ostriker et al .
2001 , stojimirovic et al .
jet - cloud interaction can result in a variety of phenomena , including cloud entrainment , jet deflection and disruption and cloud destruction ( e.g. , saxton et al .
2005 , krause 2007 and references therein ) . of interest here is entrainment ( blandford & knigl 1979 , schiano et al .
1995 , fedorenko et al .
one key problem is cloud longevity against various instabilities .
fedorenko et al .
( 1996 ) argued for magnetic nlr confinement which would then allow for nlr cloud entrainment in radio jets .
independent of theoretical considerations the question is raised if the blue outliers of our sample all harbor powerful radio jets .
relatively little is known about the radio properties of nls1 galaxies in general . on average , they tend to be less radio - loud than bls1 galaxies ( zhou et al .
2006 , komossa et al . 2006 ) and share some similarities with compact steep spectrum radio sources ( komossa et al . 2006 ) .
four of the blue outliers of our sample have first radio detections .
these imply radio powers of @xmath56= 10@xmath57810@xmath58 whz@xmath1 ( tab .
1 ) at 1.4 ghz ; similar to those agn which nelson & whittle ( 1996 ) find to be off - set from the @xmath39 relation .
we do not have information on the radio morphology of the blue outliers .
sources are unresolved with first , with one remarkable exception : the radio emission of rxj0135@xmath160043 is extended by @xmath5910 kpc ( or double ; see appendix ) .
spatially resolved radio observations of the sources are required to search for the presence of jets and to study the radio properties in more detail . [
[ thermal - winds ] ] thermal winds + + + + + + + + + + + + + variants of thermal wind models have been studied in order to explain the kinematics of ionized absorbers ( chelouche & netzer 2005 ) , and of nlrs .
_ isothermal _ parker wind models of everett & murray ( 2007 ) , originally computed in order to model spatially resolved nlr velocity gradients of ngc4151 ( das et al .
2005 ) , would have roughly the right properties to explain our average blue outlier velocity shifts and radial velocity changes .
a range in nlr cloud column densities would lead to a spread in cloud velocities ( see eqn .
( 10 ) of everett & murray 2007 ) , thereby perhaps explaining the observed line broadening .
however , as also shown by everett & murray ( 2007 ) , realistic models including photoionization heating have temperature gradients , and adiabatic cooling decelerates the winds too quickly .
an extra heating source of unknown nature would be needed in order to keep the wind isothermal .
models of das et al .
( 2007 ) , based on radiative acceleration of nlr clouds , implied that nlr clouds do not decelerate quickly enough in order to explain nlr velocity gradients of ngc1068 observed with hst ; models work after introducing drag forces from an ambient medium .
detailed modeling of the present data would likely involve several of the above model ingredients .
such modelling is beyond the scope of this paper .
[ [ high - eddington - ratios - and - orientation - effects ] ] high eddington ratios and orientation effects + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + common to blue outliers is their high eddington ratios .
could high @xmath44 be the wind driving mechanism ?
additional orientation effects ( as also discussed by marziani et al .
2003 , boroson 2005 ) are still needed in order to explain why not all agn with high @xmath44 are blue outliers . near face - on orientation
would have us look more down the flow in blue outliers , thereby enhancing the stratification , broadening , and blueshift effect .
it is interesting to point out similarities between the phenomenon of blue outliers in nls1 galaxies and in radio galaxies ( see also holt et al .
several radio galaxies show a similar phenomenon of high [ oiii ] core blueshifts and line broadening ( e.g. , tadhunter et al .
2001 , marziani et al .
2003 , holt et al .
2006 , stockton et al . 2007 ;
see gupta et al .
2005 for a related phenomenon in the uv ) .
jet - cloud interaction in the ( inner ) nlr is the favored interpretation of most of these sources .
these objects may represent the early stages of radio - source evolution ( tadhunter et al .
radio galaxies with blue outliers in [ oiii ] are typically compact flat spectrum sources , are absorbed , and are luminous in the infrared .
for some ( e.g. , pks1549@xmath1679 and pks0736 + 01 ) there is evidence that we have a near face - on view ( tadhunter et al .
2001 , marziani et al .
as pointed out before , the strong stratification we see in nls1 blue outliers argues against local jet - cloud interactions in those galaxies , but would be consistent with nlr clouds entrained in the outflowing radio plasma .
the blue outliers among the radio galaxies show signs of recent mergers . regarding nls1 galaxies in general
, there is no evidence that the majority of them underwent recent mergers or has an excess of companion galaxies ( krongold et al .
2001 , ryan et al .
little is known about the host galaxies of our blue outlier sample , in particular .
iras11598 - 0112 indeed is a merger with prominent tidal tails ( veilleux et al .
2002 ) , ultraluminous in the infrared . inspecting the sdss images of the other galaxies of our sample
we do not find strongly disturbed galaxy images indicating ongoing mergers , but this needs to be confirmed with deeper imaging .
one galaxy , rxj0135@xmath160043 , shows indications of an off - center nucleus .
that effect could be caused by interaction , be mimicked by dust , or have another origin .
recent simulations of merging black holes predict black hole recoil velocities due to emission of gravitational wave radiation up to several thousand kms@xmath1 ( e.g. , campanelli et al . 2007 ; review by pretorius 2007 ) . potentially , high relative outflow velocities of agn emission line regions vs the host galaxy can arise if the recoiled black hole keeps its blr and inner nlr ( bonning et al .
applied to blue outliers , if they harbored recoiled bhs , their whole blr ( plus the high - ionization nlr ) should show high blueshifts , while the ` remnant ' nlr would still appear in low - ionization lines . in that case
, we expect the broad component of the balmer lines to exhibit the highest blueshifts of all emission lines ; which is , however , not observed .
shifts in broad h@xmath3 are less than those in [ oiii ] .
the sdss data base is well suited for a systematic search for more blue outliers .
zhou et al . ( 2006 ) mention in passing the presence of several extreme ones among their nls1 galaxy sample but do not discuss them further .
spatially resolved optical spectroscopy will allow us to measure directly line widths , outflow velocities , etc .
in dependence of the core distance .
spectroscopy in the ir and uv will tell whether trends ( a correlation with ionization potential ) persist in ir coronal lines and high - ionization uv broad lines .
x - ray measurements of blue outliers will be useful to search for signs of high - velocity ionized outflows , and to see whether blue outliers again stick out in agn correlation space when adding their x - ray properties ( x - ray steepness , variability ) to correlation analyses .
( high - resolution ) radio observations of all galaxies will facilitate further comparison between blue outliers in radio galaxies and nls1 galaxies and will tell whether radio jets are present in nls1 blue outliers .
radio observations have the potential to confirm the pole - on hypothesis of blue outliers , if relativistic beaming is detected .
imaging with hst will reveal whether the host galaxies of blue outliers show signs of recent mergers .
in particular , galaxy merger simulations predict strong outflows in the final merger phase ( e.g. , springel et al .
imaging will allow us to test whether blue outliers are in such a phase .
if the face - on interpretation of blue outliers is correct , they are also useful test - beds for the question , whether the width of broad h@xmath3 is systematically affected by orientation .
increasing the sample size will allow us to test more stringently whether the broad component of h@xmath3 is systematically narrower in blue outliers .
if so , this would imply that blr clouds are arranged in a plane , and we would underestimate systematically the bh masses in these objects , and perhaps nls1 galaxies in general . if , instead , their blr is spherical , we would not see systematically narrower h@xmath3 in objects viewed face - on , and would not have to worry about correctness of bh mass estimates .
our preliminary results indicate that the latter is the case . on the theoretical side ,
the question is raised as to which winds can operate across long distances spanning the high - ionization blr , clr and a substantial part of the nlr , predict the observed gradient in cloud velocities ( @xmath591000 kms@xmath1 of high - ionization emission - line clouds close to the nucleus , and several hundred kms@xmath1 further out on typical nlr scales while the outer nlr is mostly unaffected ) , and ensure longevity of the emission - line clouds .
in [ oiii ] blue outliers with their extreme velocity shifts the effects of secondary influences on the nlr kinematics are enhanced or dominate completely .
their study is therefore of great relevance for ( 1 ) scrutinizing the usefulness and limitations of [ oiii ] width as a proxy for stellar velocity dispersion ; ( 2 ) understanding the origin and dynamics of the nlr ; ( 3 ) investigating driving mechanisms of agn outflows on large scales ; and ( 4 ) examining possible links with results from merger simulations which predict that a substantial fraction of the ism of the merger should be outflowing .
we have systematically studied the optical properties of such agn with high [ oiii ] blueshifts which deviate from the @xmath50}$ ] relation of bls1 and nls1 galaxies , and obtained the following results : @xmath60 all of them have high eddington ratios ( @xmath61=0.51.5 ) and narrow blr balmer lines ( fwhm(h@xmath11)=12001800 kms@xmath1 ) , which makes them nls1 galaxies .
the fraction of blue outliers among our nls1 sample is 16% ( @xmath19 } \approxgt 150 $ ] kms@xmath1 ) , and 5% at the highest outflow velocities ( @xmath19 } \approxgt 250 $ ] kms@xmath1 ) .
while blue outliers do enhance correlations which appear across the whole bls1-nls1 population , we do not find strong correlations of [ oiii ] outflow velocity with the eddington ratio within the blue outlier population itself ; perhaps due to the small sample size .
@xmath60 we do detect a strong correlation between emission - line blueshift and ionization potential , and confirm a strong correlation between [ oiii ] blueshift and [ oiii ] line width .
the presence of a classical quiescent _ outer _ nlr is indicated by the existence of low ionization lines , by [ sii ] line widths which locate the blue outliers on the same @xmath62}$ ] relation as other bls1 and nls1 galaxies , and by inferred nlr densities similar to other nls1 galaxies . on the other hand , zero - blueshift [ oiii ] emission expected from a quiescent _ inner _
nlr is weak or absent .
@xmath60 taken together , these observations place tight constraints on models : we favor a scenario where nlr clouds of blue outliers are entrained in a decelerating wind .
similar , less powerful winds could be present in all agn , but would generally only affect the clr ( in agn with blueshifted iron coronal lines but quiescent nlrs ) , or level off even before reaching the clr only affecting the high - ionization blr . @xmath60 the mechanism that drives and decelerates the wind is speculative at present , but could be linked to the high eddington ratios of the galaxies .
extra orientation effects ( near pole - on views ) , considered previously to explain the correlation of [ oiii ] blueshift with line width of blue outliers , would also explain the strong ionization stratification we detect .
@xmath60 two blue outliers have independent bh mass / stellar velocity dispersion measurements and these place them on or close to the @xmath0 relation of non - active galaxies .
this , together with the fact that the width of broad h@xmath3 does not correlate with [ oiii ] outflow velocity , indicates : if blue outliers are indeed seen more face - on , this fact does not reflect strongly in their h@xmath3 widths , implying that their blr geometry is closer to spherical than to planar . @xmath60
most remarkable among the blue outliers is the galaxy rxj01354@xmath160043 .
unlike other nls1 galaxies its radio emission is extended and possibly double , its optical balmer lines appear to be double - peaked , and its optical spectrum shows strong absorption lines from the host galaxy .
the link between blue outliers in nls1 galaxies and in ( compact ) radio galaxies needs further exploration .
a few sources are included in the sdss - nls1 samples of williams et al .
( 2002 ) and anderson et al . (
the high [ oiii ] blueshifts of sdssj115533.50 + 010730.4 and rxj01354@xmath160043 were reported by bian et al .
boroson ( 2005 ) measured [ oiii ] blueshifts of ngc450#86 , rxj01354@xmath160043 , pg1244 + 026 , and sdss17184 + 5734 . here
, we provide a short summary of the multi - wavelength properties of the blue outliers collected from the literature , and give some comments based on our optical spectral analysis .
galaxies are listed in order of decreasing [ oiii ] blueshift .
this agn is a known x - ray source , first detected with the einstein observatory ( chanan et al .
1981 ) , mostly studied in the x - ray regime ( e.g. , boller et al .
1996 , vaughan et al .
2001 ) , and optically identified as nls1 galaxy by stephens ( 1989 ) .
it shows the highest [ oiii ] blueshift of our sample ( @xmath19}=430 $ ] kms@xmath1 ) .
this agn was detected in x - rays during the rosat all - sky survey ( voges et al .
1999 ) and was first classified as nls1 galaxy by williams et al .
it has the second - highest [ oiii ] velocity shift of our sample ( @xmath19}=330 $ ] kms@xmath1 ) , highly blueshifted [ nev ] , and faint coronal line emission of [ fex ] with a blueshift corresponding to @xmath63 } \approx 1000 $ ] kms@xmath1 , among the highest blueshifts reported for coronal lines to date .
this agn was detected in the x - ray and radio band ( brinkmann et al .
2000 , wadadekar 2004 ) and was optically identified as nls1 galaxy by williams et al .
it is remarkable in several respects .
we find that its optical spectrum appears to be dominated by the host galaxy .
strong absorption lines from higher order balmer lines , ca h&k and naid are detected ( fig .
the optical agn continuum emission is either intrinsically weak , or absorbed ( which would require a peculiar geometry of the absorber , given that the broad balmer lines are present ) .
rxj01354@xmath160043 is detected with galex and displays red uv colors .
its ( narrow ) balmer lines appear double - peaked , with a separation of @xmath59580 kms@xmath1 , or else are affected by residual , exceptionally strong and redshifted , host - galaxy features which is very unlikely , given the strength of h@xmath33 .
the line decomposition shown in fig .
5 assumes the same widths of the narrow core of h@xmath3 and of the two [ nii ] lines , fixed to the width of [ sii ] , and further assumes a fixed [ nii ] line ratio .
the remaining profile was modeled with two gaussians of free parameters and can be well fit with a broad component plus a second , redshifted relatively narrow component .
rxj01354@xmath160043 has a first detection with a radio flux of 2 mjy and there is evidence for extended radio emission ( becker et al .
1997 ) , corresponding to a scale of @xmath5910 kpc .
this is remarkable because few if any nls1 galaxies are known to have widely extended radio emission ( ulvestad et al 1995 , komossa et al .
2006 , yuan et al .
2008 ) . inspecting the first image cut - out
, the source appears to be double . among the emission lines , only the balmer lines appear double - peaked .
rxj01354@xmath160043 is the first blue outlier to show this phenomenon .
explanations include jet - cloud interaction , a bipolar outflow or , speculatively , superposed strong balmer absorption ( which is rarely seen because the required level population is not easy to achieve ; hall 2007 , lu et al . in prep . ) .
we have used the stellar absorption lines from the host galaxy for a measurement of @xmath47 .
we obtain @xmath64 kms@xmath1 which is in good agreement with the [ sii]-based measurement , @xmath48}=106 $ ] kms@xmath1 , and locates rxj01354@xmath160043 ( almost perfectly ) on the @xmath39 relation of non - active galaxies ( fig .
1 ) .
this galaxy is ultraluminous in the infrared ( murphy et al .
1996 , kim & sanders 1998 ) with a single nucleus and prominent tidal tails ( veilleux et al .
its optical spectrum is that of a nls1 galaxy ( moran et al .
it is almost radio - loud with a 1.4 ghz radio index @xmath65 ( komossa et al .
2006 ) and was first detected in x - rays with rosat ( voges et al .
the feii emission in its optical spectrum is not well described by our feii template . in particular , the ` red ' and ` blue ' complexes of feii
do not match each other well , and the source redshift appears to differ from that of feii .
strong emission of unknown nature remains blueward of [ oiii ] .
formally , we can describe it with an extra [ oiii ] component , broad ( fwhm([oiii])=1800 kms@xmath1 ) and highly blueshifted ( @xmath66=1300 kms@xmath1 ) .
if real , it could represent an extreme blue wing , or could imply the presence of an extra starburst / shock - driven component .
the structure needs to be confirmed by independent spectroscopy .
this agn ( green et al . 1986 ) is a well - known x - ray , uv , infrared and radio source ( e.g. , elvis et al .
1986 , kellerman et al .
1989 , sanders et al .
1989 , fiore et al .
1998 , ballantyne et al .
2001 , jimnez - bailn et al .
2005 ) with an optical nls1 spectrum ( miller et al . 1992 , veron - cetty et al . 2001 ) .
it is radio quiet with a 5 ghz radio index of @xmath67 ( kellerman et al .
hayashida ( 2000 ) and czerny et al .
( 2001 ) determined its black hole mass based on x - ray variability ( the power density spectrum ) .
czerny et al .
( 2001 ) report @xmath45 which is consistent with our value of @xmath46 from applying the relation of kaspi et al .
we detect in its optical spectrum [ fex ] emission with an outflow velocity of 640 kms@xmath1 .
dx acknowledges the support of the chinese national science foundation ( nsfc ) under grant nsfc-10503005 , and the support of mpg / mpe .
hz acknowledges support from the alexander von humboldt foundation , from nsfc ( grant nsf-10533050 ) , and from program 973 ( no .
2007cb815405 ) .
lb acknowledges support from conacyt grant j-50296 .
we thank our referee for his / her comments and suggestions , and the members of mpe s new physics of galactic nuclei group and j. sulentic , d. proga and d. merritt for discussions .
this research has made use of the sdss data base , 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 .
funding for the sdss and sdssii 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 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 .
llcccccccccc 092247.03 + 512038.0 & sbs0919 + 515 & 0.161 & 430 & 720 & 1250 & 0.3 & 1.3 & 6.7 & 1.5 & 10 & + 115533.50 + 010730.6 & sdssj11555 + 0107 & 0.198 & 330 & 780 & 1510 & 0.3 & 0.7 & 6.7 & 0.9 & 200 & + 013521.68 @xmath16004402.2 & rxj01354@xmath160043 & 0.099 & 240 & 620 & 1710 & 1.1 & 0.5 & 6.5 & 0.5 & 90 & 22.7 + 011929.06 @xmath16000839.7 & ngc 450#86 & 0.091 & 220 & 380 & 1220 & 0.5 & 0.9 & 6.2 & 0.9 & 200 & + 032606.75 + 011429.9 & sdssj03261 + 0114 & 0.128 & 180 & 530 & 1230 & 0.5 & 0.9 & 6.3 & 1.0 & 340 & + 120226.76 @xmath16012915.3 & iras11598 - 0112 & 0.151 & 170 & 340 & 1460 & 0.5 & 2.7 & 6.8 & 1.1 & 110 & 22.9 + 171829.01 + 573422.4 & sdssj17184 + 5734 & 0.101 & 150 & 470 & 1760 & 0.4 & 0.7 & 6.6 & 0.5 & 30 & + 124635.25 + 022208.8 & pg1244 + 026 & 0.049 & 150 & 300 & 1200 & 0.6 & 0.8 & 6.2 & 0.9 & 400 & 22.1 + 091313.73 + 365817.3 & rxj09132 + 3658 & 0.108 & 150 & 350 & 1680 & 1.0 & 0.5 & 6.5 & 0.5 & 40 & 22.5 | we have studied the properties of seyfert galaxies with high [ oiii]5007 blueshifts ( `` blue outliers '' ) , originally identified because of their strong deviation from the @xmath0 relation of normal , narrow - line seyfert 1 ( nls1 ) and broad - line seyfert 1 ( bls1 ) galaxies .
these blue outliers turn out to be important test - beds for models of the narrow - line region ( nlr ) , for mechanisms of driving large - scale outflows , for links between nls1 galaxies and radio galaxies , and for orientation - dependent nls1 models .
we report the detection of a strong correlation of line blueshift with ionization potential in each galaxy , including the measurement of coronal lines with radial velocities up to 5001000 kms@xmath1 , and we confirm a strong correlation between [ oiii ] blueshift and line width .
all [ oiii ] blue outliers have narrow widths of their broad balmer lines and high eddington ratios . while the presence of non - shifted low - ionization lines signifies the presence of a classical outer quiescent nlr in blue outliers
, we also report the absence of any second , non - blueshifted [ oiii ] component from a classical inner nlr .
these results place tight constraints on nlr models .
we favor a scenario in which the nlr clouds are entrained in a decelerating wind which explains the strong stratification and the absence of a zero - blueshift inner nlr of blue outliers .
the origin of the wind remains speculative at this time ( collimated radio plasma , thermal winds , radiatively accelerated clouds ) .
it is perhaps linked to the high eddington ratios of blue outliers .
similar , less powerful winds could be present in all seyfert galaxies , but would generally only affect the coronal line region ( clr ) , or level off even before reaching the clr .
similarities between blue outliers in nls1 galaxies and ( compact ) radio sources are briefly discussed . |
though consisting of primarily old stars , galactic spheroids ( bulges of sb - sa spirals as well as s0 and elliptical galaxies ) are a major source of stellar feedback in form of mass loss and type ia supernovae ( e.g. , ciotti et al . 1991 ; knapp et al . 1992 ; mannucci et al .
the specific energy of this feedback predicts that it should present primarily in x - ray - emitting hot gas .
indeed , such hot gas has been detected in and around spheroids , which typically contain little cool gas .
however , it has been shown repeatedly that the x - ray - inferred hot gas mass and energy are far less than the empirical predictions from the feedback inputs ( e.g. , @xcite ) .
this missing stellar feedback has most likely escaped into large - scale galactic halos , where the gas becomes too tenuous to be detected in existing x - ray imaging observations ( e.g. , @xcite and references therein ) .
the implication of this scenario is profound , because the injection of the mass and energy into the halos could strongly affect the ecosystem of the galaxies and hence their evolution @xcite .
while the above qualitative picture seems clear , there are key issues that still need to be addressed to understand both the stellar feedback itself and its interplay with the environment .
significant uncertainties are still present in the mass and energy input rates of stars ( 2 ) .
the energy input rate is inferred from observations of sne in a large sample of galaxies of diverse optical and near - ir luminosities as well as types ( e.g. , @xcite ) , assuming a certain explosion mechanical energy deposited into the interstellar space .
the stellar mass loss rate is based on the modeling of the @xmath0 emission from the circumstellar medium of evolved stars ( e.g. , @xcite ) .
these semi - empirical rates , uncertain by a factor of at least @xmath1 for individual galaxies , can in principle be directly constrained by the measured temperature and luminosity of galactic coronae .
indeed , detailed simulations have been conducted for relatively isolated `` field '' spheroids and have been compared with observations @xcite , which have led to a qualitative understanding of the feedback processes and effects on x - ray measurements . in particular
, the feedback model expects that the specific energy should not change substantially from one spheroid to another , which is consistent with little correlation between the measured temperatures and k - band luminosities @xmath2 ( e.g. , @xcite ) .
but the measured temperatures ( @xmath3 1 kev ) are substantially lower than the expected value from simulations .
the measured iron abundances of coronal gas are also typically lower than expected for the ia sn - enriched gas ( 2 ) .
at least part of these discrepancies can be accounted for by various 3-d effects of the ia sn heating , which produces very low - density , hot , fast - moving , and enriched bubbles that hardly radiate . when the gas in these bubbles finally mixed with the material from the stellar mass loss at large radii ,
the x - ray emission becomes too weak and diffuse to be effectively detected @xcite .
the observed x - ray emission thus gives only a biased view of the coronae .
we expect that this bias should be minimal for a corona in the icm , however .
the high external thermal and ram pressures tightly confine such coronae , resulting in a low outflow speed and hence relatively local mixing of the feedback materials .
furthermore , one can better characterize the icm environment from observations , important for a self - consistent modeling of a corona .
therefore , coronae embedded in the icm are ideal sites to better constrain the feedback and its interplay with the environment
. there have also been significant efforts in studying stellar feedback - powered coronae embedded in the icm , mostly focusing on the ram - pressure stripping of hot gas ( e.g. , @xcite ) .
it is shown that the global morphological and integrated properties , such as gas mass and luminosity , are strongly influenced by the environment .
a set of 2d simulations done by @xcite show that galactic coronae can be maintained by stellar feedback in poor clusters while be efficiently stripped in rich ones .
@xcite and @xcite simulated galaxies falling into clusters and demonstrated that a galactic corona reached a cyclic `` stripping replenishment '' dynamics due to the periodic orbital motion of the host galaxy as well as the competing processes such as stripping and stellar feedback .
we focus on modeling coronae in and around intermediate - mass spheroids that are embedded in the icm .
such a system is relatively simple with minimum effects due to the feedback from agns and to the radiative cooling of hot gas .
we expect that the hot gas is in a quasi - steady , subsonic outflow semi - confined by the thermal and ram - pressures of the icm .
this state should be only sensitive to the local properties of the icm ( see 2 for more discussion ) , avoiding large uncertainties in modeling the history of the galactic feedback and the evolution of the environment , as would be needed for a field spheroid @xcite .
the simulations can also be compared with an increasing number of x - ray detections of galactic coronae of such spheroids ( e.g. , @xcite ) , leading to an improved understanding of the feedback itself and its interplay with the environment . in particular
, we examine the dependence of the corona properties on the specific energy of the stellar feedback and on the thermal and ram pressures of the icm and check how measurements ( e.g. , temperature , surface brightness , overall luminosity and morphology ) may be made to infer the parameters that can not directly observed ( e.g. , the icm thermal and ram pressures local to an individual galaxy ) . here
, we will present 2-d simulations only , which allow for an efficient exploration of a large parameter space .
the paper is organized as follows : we briefly describe our numerical model and setup in 2 and present results in 3
; we discuss their implications in 4 ; finally in 5 , we summarize our conclusions .
our model galaxy is composed of a stellar spheroid component and a dark matter halo .
we use the spherical hernquist density profile @xcite to represent the stellar mass distribution : @xmath4 where @xmath5 is the total stellar mass , and @xmath6 is the scale radius
. this density profile results in a gravitational potential @xmath7 the above stellar mass distribution resembles the de vaucouleur s law ; the relation between the half - light radius @xmath8 and the scale radius is @xmath9 .
we characterize the dark matter halo with the nfw profile @xcite , @xmath10 where @xmath11 is the scale radius of the dark halo , and @xmath12 is defined as @xmath13 in which @xmath14 is the critical density of the universe .
the dark halo has a mass @xmath15 within the virial radius @xmath16 , which is defined to have a density that is @xmath17 times the mean matter density of the universe @xmath18 .
we adopt @xmath17 to be @xmath19 .
therefore , we have the relation @xmath20 where @xmath21 is the concentration factor defined as @xmath22 and is related to @xmath15 .
we set @xmath23 @xcite .
thus , for a given cosmology and a given @xmath15 , the mass profile of the dark halo is totally determined .
the stellar mass and energy feedback in spheroids are dominated by the mass loss and ia sne of evolved stars , respectively .
we neglect the energy input from the random motion of stars and hence their ejecta .
the total energy released by type ia sne is @xmath24 where @xmath25 for e / s0 galaxies according to @xcite .
the empirical mass input rate from the stellar mass loss is @xmath26 where @xmath27 according to @xcite .
assuming that the mechanical energy of each sn is @xmath28 , the specific energy of stellar feedback is @xmath29 per particle . to account for the uncertainties in these rates and assumptions , we also sample three different lower values of the specfic energy for comparison with observations ( table 1 ) .
we assume that the energy and mass inputs follow the distribution of the stellar mass .
in addition , each ia sn produces @xmath30 of iron ejecta .
we assume that the iron abundance of the mass loss from stars is solar .
if the ejecta is fully and instantaneously mixed with the mass loss , the expected iron abundance relative to the solar value is then @xmath31 however , in observations , supersolar metallicity is quite rare . @xcite and @xcite have shown that ia sn ejecta may not be efficiently mixed with stellar mass loss on microscopic scale , resulting in a low effective metallicity of the ism . while our focus is on the environmental effect . in our simulation , we set the iron abundance of the input mass to the solar value .
the gas hydrodynamics and metal abundance distribution with both the stellar feedback and radiative cooling can be described by the following equations : @xmath32 ) & = & \rho_g \bf g \it \nonumber \\ { \partial ( \rho_g e ) \over \partial t } + \bf \nabla \cdot \it [ ( \rho_g e + p)\bf v \it ] & = & s_e - n_e n_i \lambda(t , z ) + \rho_g \bf g \cdot v \it \nonumber \\ { \partial ( \rho_g x_{iron } ) \over \partial t } + \bf \nabla \cdot \it ( \rho_g x_{iron } \bf v \it ) & = & s_{iron } .\end{aligned}\ ] ] the first equation is mass conservation law , with @xmath33 denoting the mass density of the coronal gas .
the second is momentum equation .
@xmath34 is gas pressure and @xmath35 is gravitational acceleration .
the third one is energy equation .
@xmath36 stands for the specific energy of the gas , including both thermal and kinetic components .
the second term to the right is cooling rate .
we adopt the cooling curve from @xcite , assuming an optically - thin thermal plasma in collisional ionization equilibrium . for calculation of the radiation in a specific band , we use the mekal model , extracted from the x - ray spectral analysis software xspec .
we use the fourth equation to keep track on the iron mass fraction , which is denoted by @xmath37 .
we conduct our simulations with the flash code @xcite , an eulerian astrophysical hydrodynamics code with the adaptive mesh refinement ( amr ) capability .
the simulated region is fixed in the galaxy - rest frame using cylindrical coordinates , with @xmath38 ranging from @xmath39 to @xmath40 and the radius from @xmath41 to @xmath40 .
the axis of the cylinder is through the center of a simulated spheroid and along the direction of its motion .
the upper and lower boundary conditions are fixed so that the icm flows in and out the simulation region at a constant speed .
this mimics the motion of galaxies through a local cluster environment .
we apply reflection boundary condition at @xmath42 and diode boundary condition , which only allows gas to flow out , to the right side of the simulation region .
compared with the simulation box , which is @xmath40 by @xmath43 , the coronae are only on the order of @xmath44 to @xmath45 across .
data near the outer regions will be excluded in our analysis to avoid any potential artifacts introduced by the assumed outer boundary condition of the simulations . also , the outer region will not be shown in the following images .
we allow the resolution to reach @xmath46 , so that the small coronae can be well resolved .
when a simulation starts , there is no interstellar gas in the galaxy .
as the simulation progresses , the stellar feedback gradually accumulates in and around the spheroid to form a corona , which is characterized by its higher iron abundance . in the mean time , the ram - pressure and turbulent motion strips gas at the outer boundaries of the corona .
we end the simulation when it reaches a statistically quasi - steady state . empirically , the time to reach such a state is @xmath47 gyr , while the time for the icm to pass the simulation region ranges from @xmath48 gyr to @xmath49 gyr .
representative results are all extracted from the simulations after this time .
the presence of a local quasi - steady state is a reasonable assumption for a compact corona .
as a galaxy moves through a cluster , the icm condition can of course change drastically .
but the time scale for such a change is typically much longer than the dynamic time for the corona to adapt the local environment . for a cluster of a characteristic size of @xmath50 mpc and temperature of 2 kev , for example
, the crossing time for a galaxy moving roughly at the sound speed is @xmath51 gyr . in contrast , for a corona of a typical size @xmath52 kpc and temperature @xmath53 kev , the sound crossing time is only @xmath54 myr . even if a corona is totally destroyed at some point(e.g . at the central region of a cluster ) , the re - building time scale @xmath55 , as dicussed above , is still shorter than the environment change time scale .
therefore , the local quasi - steady state is a reasonable assumption for characterizating the environmental impact on galactic coronae .
we have simulated a set of cases to characterize the dependence on key parameters .
table 1 lists our adopted model parameter values .
the different combinations of the icm density ( @xmath56 , the number density of all particles ) and temperature ( @xmath57 ) as well as the mach number ( @xmath58 ) and specific energy ( @xmath59 ) of the model galaxy form a set of 48 cases .
[ t : model_p ] [ cols="<,^,^,^,^ " , ] fig .
[ f : plot_compare ] compares the hydrostatic solutions ( eq . [ e : hs ] ) to the simulations with three different @xmath60 parameters . in those cases with small @xmath60 , such as @xmath61 , in which the gravity dominates over the feedback ,
the hydrostatic solutions give nearly perfect matchs to the simulated pressure profiles . while in a case like @xmath62 ( with @xmath63 ) ,
where the feedback dominates , the deviation of the hydrostatic solution from the simulated profile is apparent . for the cases which we think are plausible ( @xmath64=1.8kev , see
[ ss : discussion_02 ] ) , @xmath60 is about @xmath65 or much less than @xmath65 , and the coronae are largely in hydrostatic state . although the same model galaxy ( in terms of the stellar and dark matter masses ) is adopted for all the simulation cases , the resultant x - ray luminosities can still differ by up to two orders of magnitude , due to the different choices of the mach number , @xmath64 , and/or thermal pressure values of the icm . because the corona temperature is determined by @xmath64 , the gas density is @xmath66 if the corona is in a nearly hydrostatic state .
therefore , the surface brightness of a corona provides a measure of the ambient icm pressure and thus may be used to estimate the line - of - sight position in a cluster . however , the x - ray luminosity of a corona depends on several factors .
[ f : plot_lumin ] shows a clear anti - correlation between the luminosity and specific energy @xmath64 .
this anti - correlation is primarily due to the density decrease with the increase of @xmath64 , although it does not strongly affect the size of a corona .
the icm thermal pressure tends to squeeze the corona , hence enhance its luminosity ( subsonic cases ) .
but this effect is complicated by the presence of the ram - pressure .
as the mach number increases , the lumonosity can decrease because the ram - pressure stripping reduces the overall size ofthe corona .
these dependences on the environment as well as the stellar feedback energetics may naturally explain the observed large dispersion of @xmath67 for spheroids of similar @xmath2
. the complications in the dependences may also account for the lack of a clear observed trend in the icm environment effect on x - ray luminosities of coronae @xcite . in our model of the galactic coronae
, the gas temperature is primarily determined by the specific energy of the stellar feedback and thus should not change significantly with the stellar mass .
this independence on the mass or @xmath2 is consistent with the temperature measurements of the coronae of intermediate - mass spheroids @xcite .
this is in contrast to the correlation between the temperature and @xmath68 for more massive systems such as clusters and groups of galaxies @xcite .
in particular , galaxy clusters show a well - defined scaling law between the temperature and luminosity of the observed hot gas , which is a natural result of the predominant gravitational heating in the self - similar cluster formation .
the scaling law for lower mass systems ( e.g. , groups of galaxies ) is known to be slightly different from that for clusters ( e.g. , showing an entropy floor ) , which is believed to be an imprint of preheating(e.g . , starburst and early agns ) .
correlation between @xmath68 and the temperature of hot gas is observed in massive x - ray - bright elliptical galaxies , especially for central galaxies in groups and clusters ( e.g , , @xcite ) .
but the entropies are found lie below the entropy floor(@xmath69 ) discovered in groups of galaxies .
radiative cooling could account for the low entropy , although how this runaway process may be balanced by the heating due to the mechanical inputs from both stellar and agn feedbacks remains unclear .
we have shown that the low value and the positive radial gradient of the entropy are expected from the distributed feedback in intermediate - mass spheroids , in which the radiative cooling is not important .
therefore , we may conclude that the coronae of intermediate - mass spheroids represent the extreme case in which the stellar feedback plays a dominant role , which means they are produced by stellar mass loss and heated by sne .
we may further constrain the stellar feedback based on the measured temperature of the coronae . though with a relative large dispersion , the measured temperatures are mostly fall in the range of @xmath70 , which is significantly higher than those measured in field spheroids , but is still substantially lower than what is inferred from our 2-d simulations if the canonical specific energy value of the stellar feedback is assumed ( @xmath71 ; see
[ s : setup ] ) .
part of this discrepancy could still be due to the 3-d effects of discrete heating by ia sne , as mentioned in 1 ( and characterized in @xcite ) .
but we expect that such effect should be substantially weaker in the compact coronae embedded in the high pressure icm and that the measured temperature should more faithfully reflect the specific energy of the feedback .
to match the measured temperature range of the coronae requires a specfic energy of @xmath72 , or a factor of @xmath73 lower than the canonical value ( @xmath74 ; [ s : setup ] ) .
this factor is probably still within the uncertainties of the semi - emipirical mass and energy injection rates .
further , the assumed mechanical energy per ia sn could be somewhat ( e.g. , a factor of @xmath1 ) less than @xmath75 . also a considerable fraction of the energy can be used to generate cosmic rays , magnetic field and turbulent motion .
the diversion of the energy into these various forms could significantly reduces the temperature , although the hydrodynamics of the coronae , hence the density and pressure distributions , should not be significantly affected .
the simulated coronae with @xmath76 generally have individual luminosties of a few times @xmath77 , consistent with the observed range of @xmath78 @xcite .
a considerably large value of @xmath64 is not favored , because it decreases the expected luminosity steeply ( fig . [
f : plot_lumin ] ) .
we discuss here the potential impacts of the pressure or density enhancement of the coronae on the fueling of the central smbhs and the evolution of cool gas , if present in the spheroids .
the simulation shows that the central density of a corona is sensitive to the thermal pressure of the surrounding medium . to infer the power of a smbh
, we adopt the bondi accretion rate @xcite : @xmath79 where @xmath80 is the mass of the smbh , while @xmath81 and @xmath82 are the density and sound speed at the center of a corona . assuming the fraction of the accretion energy released is @xmath83 , the power of the smbh can be approximated as @xmath84 the smbh mass can be estimated from its correlation with the spheroid mass @xmath85 @xcite .
[ f : plot_agnpower ] shows the dependence of the power on the icm state , the mach number of spheroid , and the specific energy of the feedback . , green for @xmath86 and blue for @xmath87 . to avoid overlap among the symbols ,
the higher mach number models are shifted to the right a litte bit and the subsonic cases to the left . ] in particular the luminosity for the most plausible @xmath88 cases is about @xmath89 , which falls in the range of the x - ray power of a low luminosity agn .
this indicates that a compact corona built up by stellar feedback and embedded in cluster environment could feed a moderate agn .
the icm pressure also tends to enhance the accretion , consistent with the finding that the galaxies with @xmath90 agns are more centrally concentrated than ones without @xcite .
these galaxies with agns are not dominated by galaxies that have recently entered the clusters .
similar conclusions are also reached in more recent studies , such as the one by @xcite , based on the analysis of a sample of @xmath91 radio galaxies and @xmath92 point sources .
the pressure enhancement could also have significant impacts on cool gas in a galaxy . under high pressure ,
cool gas exists preferentially in molecular form rather than atomic one .
the compression of cold gas because of the icm pressure could further lead to star formation @xcite , depriving the galaxy of the gas further . thus it is expected that galaxies contain less amounts of cool gas in clusters than in the field .
these impacts should affect cool gas not only in spheroids , but in spirals as well , consistent with existing observations @xcite .
we have conducted a range of 2-d hydrodynamic simualtions of galactic coronae that result from gradual energy and mass feedback in stellar spheroids moving in the icm enviornment . we have focused on spheroids that are in the intermediate - mass range ( corresponding to @xmath93 ) so that both the agn feedback and the radiative cooling of the hot gas could largely be neglected .
we explore the dependence of corona properties on the specific energy of the stellar feedback as well as on the ram and thermal pressures of the icm .
our major results and conclusions are as follows : 1 .
x - ray coronae embedded in clusters could be naturally explained by the subsonic outflow driven by the stellar feedback , semi - confined by the ram - pressure and compressed by thermal pressure of the surrounding icm .
the corona temperature depends primarily on the specific energy of the input material in such a way that @xmath94 .
the decrease of the thermal energy due to the climbing of the gravitational potential and the expansion is largely compensated by the distributed heating by ia sne .
this result naturally explains the lack of the correlation between the temperature and k - band luminosity for the spheroids in our considered mass range .
an outflow powered by a distributed feedback also has a positive radial entropy profile , mimicking what may be produced by a `` cooling flow '' .
the coronal gas is typically in an approximate hydrostatic state . as a result
, the density of the corona gas depends strongly on the thermal pressure of the icm , but only weakly on the ram pressure .
therefore , the surface brightness of x - ray emission is a good measurement of the thermal icm pressure , which may be used to estimate the line - of - sight location of a spheroid in a cluster .
the total x - ray luminosity of a corona decreases with the increase of the feedback specific energy .
the thermal pressure tends to increase ( or reduce ) the luminosity in subsonic ( supersonic ) cases .
the semi - confinement of the coronae by the icm allows a good constraint on the energetics of the stellar feedback . to be consistent with the observed x - ray luminosity and temperature ,
the specific energy of the feedback should be @xmath95 , a factor of 2 - 3 smaller than the value inferred from the commonly accepted semi - empirical ia sn and mass - loss rates , assuming the mechanical energy of @xmath96 per sn .
4 . the relatively high pressure of the coronae in the icm may have important implications for understanding the agn activity as well as the cool gas properties in spheroids .
the density increase caused by the icm pressure , for example , could enhance the bondi accretion , which may explain the observed central concentration of agns in clusters .
the high pressure can further compress the gas to form molecular clouds and enhance star formation .
the combination of the enhanced consumption and the ram - pressure striping can naturally lead to the deprivation of gas and subsequent passive evolution of galaxies in clusters .
we thank s .- k .
tang for his help in the initial setting up of the simulations . the software used in this work was in part developed by the doe - supported asc / alliance center for astrophysical thermonuclear flashes at the university of chicago .
simulations were performed at the pittsburgh supercomputing center supported by the nsf . the project is partly supported by nasa through grant nnx10ae85 g .
acreman d.m .
, stevens i.r .
, ponman t.j . & sakelliou i. , 2003 , mnras , 341 , 1333 bekki k. & couch w.j . , 2003 , apj , 596 , 13 boroson b. , kim d.w . & fabbiano g. , 2010 , arxiv:1011.2529v1 david l.p . , jones c. , forman w. , vargas i.m . , & nulsen p. , 2006 ,
apj , 653 , 207 edgar r.g . , 2004 ,
new astron .
, 48 , 843 eke v.r . , navarro j.f . & steinmetz m. , 2001 , apj , 554 , 114 fabian a.c .
, & sanders j.s . , 2009 , arxiv:0909.2577 fryxell b. , olson k. , ricker p. , timmes f.x . , zingale m. , lamb d.q . ,
macneice p. , rosner r. , truran j. w. , & tufo h. , 2000 , apjs , 131 , 273 hart q.n .
, stocke j.t . &
hallman e.j . , 2009 ,
apj , 705 , 854 helsdon s.f . & ponman t.j . , 2003 , mnras , 340 , 485 hernquist l. , 1990 , apj , 356 , 359 jeltema t. e. , binder b. , & mulchaey j. s. , 2009 , apj , 679 , 1162 kim d .- w . , kim e. , fabbiano g. , & trinchieri g. , 2008 , apj , 688 , 931 knapp g.r . , gunn j.e . , & wynn - williams c.g . , 1992 , apj , 399 , 76 li z. , wang q.d .
, irwin j.a . , & chaves t. , 2006 , mnras , 371 , 147 li z. , & wang q.d . , 2007 , apj , 668 , 39 li z. , wang q.d .
, & hameed s. , 2007 , mnras , 376 , 960 machacek m. , dosaj a. , forman w. , jones c. , markevitch m. , vikhlinin a. , warmflash a. , & kraft r. , 2005 , apj , 621 , 663 machacek m. , jones c. , forman w.r .
, & nulsen p. , 2006 ,
apj , 644 , 155 magorrian j. , tremaine s. , richstone d. , bender r. , bower g. , dressler a. , faber s.m .
, gebhardt k. , green r. , grillmair c. , kormendy j. , lauer t. , 1998 , aj , 115 , 2285 mannucci f. , della valle m. , panagia n. , cappellaro e. , cresci g. , maiolino r. , petrosian a. , & turatto m. , 2005 , aa , 433 , 807 martini p. , mulchaey j.s . & kelson d.d . , 2007 , apj , 664 , 761 mathews w.g . & brighenti f. , 2003 , araa , 41 , 191 mulchaey j.s . &
jeltema t.e . , 2010 , apj , 751 , l1 navarro j.f . ,
frenk c.s .
, & white s.d.m . , 1997 , apj , 490 , 493 osullivan e. , ponman t.j . , & collins r.s . , 2003 , mnras , 340 , 1375 sato s. & tawara y. , 1999 , apj , 514 , 765 stevens i.r . , acreman d.m .
& ponman t.j . , 1999 , mnras , 310 , 663 sun m. , jones c. , forman w. , nulsen p.e.j . , donahue m. , & voit g.m . , 2006 , apj , 637 , 81 sun m. , jones c. , forman w. , vikhlinin a. , donahue m. , & voit m. , 2007 , apj 657 , 197 sutherland r.s . & dopita m.a . , 1993 , apjs , 88 , 253 tang s. , & wang q.d .
2010 , mnras , in press tang s. , wang q.d .
, lu y. , & mo h. , 2009 , mnras , 392 , 77 tang s. , wang q.d . , maclow m. , & joung m.r . , 2009 , mnras , 398 , 1468 temi p. , brighenti f. , & mathews w.g . , 2008 , apj , 672 , 244 toniazzo t. & schindler s. , 2001 , mnras , 325 , 509 wang q.d . , 2010 , pnas , 107 , 7168 young j.s .
& scoville n.z . , 1991 , araa , 29 , 581 | recently x - ray observations have shown the common presence of compact galactic coronae around intermediate - mass spheroid galaxies embedded in the intracluster / intragroup medium ( icm ) .
we conduct 2-d hydrodynamic simulations to study the quasi - steady - state properties of such coronae as the natural products of the ongoing distributed stellar feedback semi - confined by the thermal and ram pressures of the icm .
we find that the temperature of a simulated corona depends primarily on the specific energy of the feedback , consistent with the lack of the correlation between the observed hot gas temperature and k - band luminosity of galaxies .
the simulated coronae typically represent subsonic outflows , chiefly because of the semi - confinement . as a result ,
the hot gas density increases with the icm thermal pressure .
the ram pressure , on the other hand , chiefly affects the size and lopsidedness of the coronae .
the density increase could lead to the compression of cool gas clouds , if present , and hence the formation of stars .
the increase also enhances radiative cooling of the hot gas , which may fuel central supermassive black holes , explaining the higher frequency of active galactic nuclei observed in clusters than in the field .
the radiation enhancement is consistent with a substantially higher surface brightness of the x - ray emission detected from coronae in cluster environment .
the total x - ray luminosity of a corona , however , depends on the relative importance of the surrounding thermal and ram pressures .
these environment dependences should at least partly explain the large dispersion in the observed diffuse x - ray luminosities of spheroids with similar stellar properties .
furthermore , we show that an outflow powered by the distributed feedback can naturally produce a positive radial gradient in the hot gas entropy , mimicking a cooling flow .
method : numerical ism : kinematics and dynamics x - ray : galaxies : clusters |
relativistic dirac approaches based on the dirac equation have been very successful in treating nuclear reactions @xcite . because the dirac analyses have proven to be very successful in describing the intermediate - energy proton elastic scatterings from the spherically symmetric nuclei and a few deformed nuclei @xcite , the relativistic approaches based on the dirac equation are expanded to inelastic scatterings and have shown considerable improvements compared to the conventional nonrelativistic calculations based on the schrdinger equation @xcite .
the relativistic effect is taken into account at the level of the kinematics in the nonrelativistic calculations based on the schrdinger equation , because it is no longer negligible for the intermediate - energy proton scattering .
but the relativistic kinematics correction seems not to be enough , so the fully relativistic treatment based on the use of the dirac equation seems to be required for the description of intermediate - energy proton scattering .
it should be noted that one of the merit of the relativistic approach based on the dirac equation instead of using the nonrelativistic approach based on the schrdinger equation is that the spin - orbit potential appears naturally in the dirac approach when the dirac equation is reduced to a schrdinger - like second - order differential equation , while the spin - orbit potential should be inserted by hand in the nonrelativistic schrdinger approach in order to describe the intermediate energy nucleon scattering from the nucleus . in this work we performed the dirac coupled channel analyses for the high - lying excited states of s - d shell nuclei @xmath1 mg and @xmath2 mg that belong to the @xmath3 gamma vibrational band at polarized 800-mev proton inelastic scattering .
dirac phenomenological optical potentials are used , employing the s - v optical potential model @xcite where only scalar and time - like vector potentials are considered .
the woods - saxon shape is used for the geometry of the direct optical potentials , assuming the shape of the potential follows the shape of the nuclear density . in order to accommodate the collective motion of the excited deformed nucleus considering the high - lying excited states of 2@xmath0 gamma vibrational band ,
the first - order vibrational collective model is used to obtain the transition optical potentials @xcite .
the complicated dirac coupled - channel equations are solved phenomenologically to reproduce the experimental data by varying the optical potential and deformation parameters , using a computer program called ecis @xcite .
the effect of the multistep process is investigated by including the channel coupling between two adjacent excited states in addition to the couplings between the ground state and the excited states .
the calculated results are analyzed and compared with the experimental data and the results obtained from the nonrelativistic approaches .
relativistic dirac coupled channel analyses are performed phenomenologically for the high - lying excited states of the 2@xmath0 gamma vibrational band at @xmath1 mg ( p , p@xmath4 ) and @xmath2 mg ( p , p@xmath4 ) by using an optical potential model and the first - order vibrational collective model . because @xmath1 mg and @xmath2 mg are the spin-0 nuclei , only scalar , time - like vector and tensor optical potentials survive @xcite , as in spherically symmetric nuclei @xcite , hence the relevant dirac equation for the elastic scattering from the nucleus is given as @xmath5 \psi(r ) = 0 \label{e1}\ ] ] here
, @xmath6 is a scalar potential , @xmath7 is a time - like vector potential , @xmath8 is a tensor potential , and @xmath9 is the coulomb potential .
however , depending on the model assumed , pseudo - scalar and axial - vector potentials may also be present in the equation when we consider inelastic scattering .
we assume , in the collective model approach used in this work , that the appropriate transition potentials can be obtained by deforming the direct potentials that describe the elastic channel reasonably well @xcite .
the shapes of the deformed potentials are assumed to follow the shape of the deformed nuclear densities and the transition potentials are obtained by assuming that they are proportional to the first - order derivatives of the diagonal potentials @xcite . as direct potentials , the scalar and the time - like vector potentials are used in the calculation .
tensor potentials are always present due to the interaction of the anomalous magnetic moment of the projectile with the charge distribution of the target .
however , the tensor potentials are neglected in this calculation because they have been found to be always very small compared to scalar or vector potentials @xcite . the evidence that the large scalar and vector fields of dirac phenomenology may be related to quark degrees of freedom in the nucleon
can be found in the work of cohen @xmath10 @xmath11 @xcite . in the vibrational model of ecis
, the deformation of the nuclear surface is written using the legendre polynomial expansion method as @xmath12 with @xmath13 being the radius at equilibrium , @xmath14 a deformation parameter and @xmath15 the multipolarity .
the transition potentials for the channel coupling are given by @xmath16 where the subscript @xmath17 refers to the real and the imaginary scalar or vector potential , @xmath18 is the radius parameter of the woods - saxon shape . by assuming that the real and the imaginary deformation parameter @xmath19 are equal for the given potential type , two deformation parameters , @xmath20 and @xmath21 , are determined for each excited state . in order to obtain the dirac coupled channel equations , we expand the dirac wavefunction using upper and lower component and substitute them into the dirac equation . after some calculation@xcite
, we can obtain the coupled equations for the radial upper component , @xmath22 , and lower component @xmath23 of the dirac spinors as follows .
@xmath24g_j & - & [ e-\epsilon_j + m + u_s^0-u_0 ^ 0]f_j + ( u_t^0-\frac{\nu } { 2 m } \frac{\partial v_c}{\partial r})g_j \nonumber \\ &
= & \sum_{\lambda j ' } p^\lambda_{jj'}[(u_s^\lambda - u_0^\lambda ) ( sgn\chi_j)(sgn\chi_{j'})f_{j'}- u_t^\lambda g_{j ' } ] \\
\left[\frac{d}{dr } - \frac{\chi_j}{r}\right]f_j & + & [ e-\epsilon_j - m - u_s^0 - u_0 ^ 0]g_j-(u_t^0-\frac{\nu } { 2 m } \frac{\partial v_c}{\partial r})f_j \nonumber \\ & = & \sum_{\lambda j ' } p^\lambda_{jj'}[(u_s^\lambda + u_0^\lambda ) g_{j'}+ u_t^\lambda f_{j'}]\end{aligned}\ ] ] here , we separate the elastic optical potentials , @xmath25 , the scalar and time - like vector potentials from transition potentials , @xmath26 , and the subscript @xmath15 refers to the multipole order of the particular transition potential .
@xmath27 is the quantum number related to the projectile , @xmath28 is the total energy of the system , @xmath29 is the nuclear energy eigenvalue , @xmath30 is the anomalous magnetic moment of the projectile , and @xmath31 is the geometric coefficient related to @xmath32 and @xmath33 channel of multipole order @xcite .
these complicated dirac coupled channel equations are solved numerically to calculate the scattering observables such as differential cross section and analyzing power by using a computer code , ecis @xcite written by j. raynal , which employs the sequential iteration method .
the experimental data for the differential cross sections and analyzing powers are obtained from ref .
17 for the 800-mev polarized proton inelastic scatterings from the @xmath1 mg and @xmath2 mg nuclei .
the high - lying excited states of the 2@xmath0 gamma vibrational band , the second @xmath34(4.24 mev ) and the second @xmath35(6.01 mev ) states for the @xmath1 mg ( p , p@xmath4 ) , the second @xmath34(2,94 mev ) and the second @xmath35(4.90 mev ) states for the @xmath2 mg ( p , p@xmath4 ) are considered and assumed to be collective vibrational states in the calculation .
one can assign the @xmath36 state , which is an unnatural parity state , to the gamma band for both nuclei , but they are neglected in the calculation because they have large error bars and pretty irregular shape @xcite .
first , the 12 parameters of the diagonal scalar and vector potentials in the woods - saxon shapes are determined by fitting the experimental data for the elastic scattering from @xmath1 mg .
the dirac equations are solved numerically to obtain the best fitting parameters to the experimental data by using the minimum @xmath37 method .
calculated results for the 800-mev p + @xmath1 mg elastic scattering are shown as dash - dot - dot lines in figure 1 and it is found that the observable elastic experimental data are reproduced quite well . mg scattering .
the dash - dot - dot line , dash - dot , dashed and solid lines represent the results of the dirac phenomenological calculation where only the ground state is considered , where the ground and the @xmath38 states are coupled , where the ground , the @xmath39 states are coupled , and where the ground , the @xmath39 and the @xmath38 states are coupled , respectively.,width=377 ] .calculated phenomenological optical potential parameters of the woods - saxon shape for 800-mev proton elastic scattering from @xmath1 mg . [ cols="^,^,^,^,^,^,^,^",options="header " , ] in table 2 , we show the deformation parameters for the excited states of the 2@xmath0 gamma vibrational band in @xmath1 mg and @xmath2 mg and compare them with those obtained by using the nonrelativistic calculations . even though the theoretical bases are quite different , the deformation parameters obtained by using the dirac phenomenological coupled - channel calculation for the excited states of the 2@xmath0 gamma vibrational band in @xmath1 mg and @xmath2 mg are shown to agree pretty well with those obtained by using the nonrelativistic calculations @xcite .
the deformation parameters for the @xmath39 and @xmath38 states of @xmath1 mg are found to be larger than those of @xmath2 mg , indicating that the couplings of the excited states to the ground state are stronger at the @xmath1 mg ( p , p@xmath4 ) than at the @xmath2 mg ( p , p@xmath4 ) , even when the excitation energies for those states are higher at the @xmath1 mg ( p , p@xmath4 ) than those at the @xmath2 mg ( p , p@xmath4 ) .
the potential strengths are changed to -322.8 , 19.68 , 123.9 and -80.75 mev for scalar real and imaginary and vector real and imaginary potentials , respectively , in the @xmath39 state and the @xmath38 states coupled case for @xmath1 mg ( p , p@xmath4 ) , -314.1 , -21.99 , 190.0 and -81.16 mev in the @xmath39 and the @xmath38 states coupled case for @xmath2 mg ( p , p@xmath4 ) .
a relativistic dirac coupled channel calculation using the optical potential model could describe the high - lying excited states of the 2@xmath0 gamma vibrational band for the 800-mev polarized proton inelastic scatterings from the s - d shell nuclei @xmath1 mg and @xmath2 mg reasonably well , showing better agreement with the experimental data compared to the results obtained from the nonrelativistic calculation .
the first - order vibrational collective models are used to describe the high - lying excited collective states of the 2@xmath0 gamma vibrational band in the nuclei .
the deformation parameters obtained by using the dirac phenomenological coupled - channel calculation for the high - lying excited states of the 2@xmath0 gamma vibrational band in @xmath1 mg and @xmath2 mg are found to agree pretty well with the those of the nonrelativistic calculations using the same woods - saxon potential shape .
the deformation parameters for the @xmath39 and the @xmath38 states of @xmath1 mg are turned out to be larger than those of @xmath2 mg , indicating that the couplings of the excited states to the ground state are stronger at the @xmath1 mg ( p , p@xmath4 ) than at the @xmath2 mg ( p , p@xmath4 ) .
it is found that the multistep channel - coupling effect seems not important to describe the excited states of the 2@xmath0 gamma vibrational band for the proton inelastic scatterings from @xmath1 mg and @xmath2 mg .
this work was supported by the research grant of the kongju national university in 2016 and by basic science research program through the national research foundation of korea(nrf ) funded by the ministry of education(2016r1d1a1b01014355 ) .
l. g. arnold , b. c. clark , r. l. mercer , and p. swandt , phys .
c * 23 * , 1949 ( 1981 ) .
j. a. mcneil , j. shepard , and s. j. wallace , phys .
lett * 50 * , 1439 ( 1983 ) ; * 50 * , 1443 ( 1983 ) .
s. shim , ph .
d. thesis , the ohio state university 1989 : l. kurth , b. c. clark , e. d. cooper , s. hama , s. shim , r. l. mercer , l. ray , and g. w. hoffmann , phys .
c * 49 * , 2086 ( 1994 ) . s. shim , b.c .
clark , e.d .
cooper , s. hama , r.l .
mercer , l. ray , j. raynal , and h.s .
sherif , phys . rev .
c * 42 * , 1592 ( 1990 ) .
d. l. pham and r de swiniarski , nuovo cimento a * 107 * , ( 1994 ) 1405 .
r. de swiniarski , d. l. pham , and j. raynal , z. phys .
a - hadrons and nuclei * 343 * , 179 ( 1992 ) .
_ , int . jou . of mod
e * 21 * , 1250098 ( 2012 ) .
s. shim , m. w. kim , b. c. clark , and l. kurth kerr , phys .
c * 59 * , 317 ( 1999 ) .
et al . _ ,
j. korean .
51 * , 271 ( 2007 ) : s. shim et al .
, j. korean .
. soc . * 53 * , 1146 ( 2008 ) .
j. j. kelly , phys . rev .
c * 71 * , 064610 ( 2005 ) .
j. raynal , _ computing as a language of physics _ , ictp international seminar course , 281(iaea , italy , 1972 ) : j. raynal , _ notes on ecis94 _ , note cea - n-2772 , 1994 . c. j. horowitz and b. d. serot , nucl .
a * 368 * , 503 ( 1981 ) .
r. j. furnstahl , c. e. price , and g. e. walker , phys .
c * 36 * , 2590 ( 1987 ) . l. ray and g. w. hoffmann , phys .
c * 31 * , 538 ( 1986 ) .
s. shim and m. w. kim , j. korean .
64 * , 483 ( 2014 ) .
t. d. cohen , r. j. furnstahl , and d. k. griegel , phys .
* 67 * , 961 ( 1991 ) .
g. s. blanpied , b. g. ritchie , m. l. barlett , g. w. hoffmann , j. a. mcgill , m. c. milner , k. w. jones , s. k. nanda , r. de swiniarski , phys .
c * 37 * , 1987 ( 1988 ) .
g. s. blanpied , n. m. hintz , g. s. kyle , m. a. franey , s. j. seestrom - morris , r. k. owen , j. w. palm , d. dehnhard , m. l. barlett , c. j. harvey , g. w. hoffmann , j. a. mcgill , r. p. riljestrand and l. ray , phys .
c * 25 * , 422 ( 1982 ) .
g. s. blanpied , n. m. hintz , g. s. kyle , j. w. palm , r. riljestrand , m. barlett , c. j. harvey , g. w. hoffmann , l. ray and d. g. madland , phys .
c * 20 * , 1490 ( 1979 ) .
s. shim and m. w. kim , sae mulli , * 40 * , 545 ( 2000 ) . | dirac coupled channel calculations are performed phenomenologically for the high - lying excited states that belong to the 2@xmath0 gamma vibrational band at the 800-mev polarized proton inelastic scatterings from the s - d shell nuclei , @xmath1 mg and @xmath2 mg .
optical potential model is used and scalar and time - like vector potentials are considered as direct potentials .
first - order vibrational collective models are used to obtain the transition optical potentials in order to accommodate the high - lying excited vibrational collective states .
the complicated dirac coupled channel equations are solved phenomenologically to reproduce the differential cross section and analyzing power data by varying the optical potential and deformation parameters .
it is found that the relativistic dirac coupled channel calculation could describe the high - lying excited states of the 2@xmath0 gamma vibrational band at the 800-mev polarized proton inelastic scatterings from s - d shell nuclei @xmath1 mg and @xmath2 mg reasonably well , showing better agreement with the experimental data compared to the results obtained from the nonrelativistic calculations .
calculated deformation parameters for the excited states are analyzed and compared with those of nonrelativistic calculations . |
the explosion problem concerns existence and regularity of positive solutions of nonlinear elliptic equations of the form @xmath1 in a domain @xmath2 with the dirichlet boundary conditions : @xmath3 on the boundary @xmath4 .
the nonlinearity @xmath5 is convex and increasing with @xmath6 and @xmath7 two typical examples to keep in mind are @xmath8 and @xmath9 with @xmath10 .
the positive parameter @xmath11 measures the non - dimensional strength of the nonlinearity .
it has been shown in the pioneering works of keener and h. keller @xcite , joseph and lundgren @xcite , and crandall and rabinowitz @xcite that there exists a critical threshold @xmath12 so that ( [ intro - ellipt ] ) admits positive solutions for @xmath13 , while no positive solutions exist for @xmath14 .
the regularity of solutions at @xmath15 is a delicate issue : the linearized problem was studied by brezis and vazquez in @xcite in great detail . in particular , when the domain is a ball , and for the exponential and power nonlinearities mentioned above , the solutions at the critical value @xmath16 are uniformly bounded in dimensions less or equal to @xmath17 and @xmath18 , respectively , while in higher dimensions they are unbounded . for more general nonlinearities @xmath19 and
domains @xmath20 , regularity of solutions at @xmath15 in dimensions @xmath21 has been established by nedev @xcite , and more recently in dimension @xmath22 by cabr @xcite . in the present paper
we consider the non - selfadjoint elliptic problem @xmath23 with a prescribed incompressible flow @xmath24 so that @xmath25 .
extension of the aforementioned results to the case when a flow is present is a natural question in the context of the original motivation for the study of ( [ intro - ellipt ] ) as the explosion problem @xcite .
of particular interest is to understand how the presence of an underlying flow and its features affect the explosion limit .
existence of a critical explosion threshold @xmath26 can be established as a straightforward generalization of existing methods .
we are mostly interested here in the qualitative dependence of @xmath26 on the flow @xmath27 whether a flow may raise or lower the explosion threshold , and in the asymptotic behavior of @xmath26 in the limit of a strong flow .
intuitively , a flow improves mixing and interaction with the boundary hence one may expect that an incompressible flow would always raise the explosion threshold . somewhat surprisingly , this was shown not necessarily to be the case in @xcite .
more precisely , berestycki , kagan , joulin and sivashinsky have considered in @xcite , the problem ( [ elliptic ] ) for a two - dimensional cellular flow and observed numerically that while the explosion threshold increases for flows oscillating on a small scale , it may actually decrease if the flow has large scale variations .
this is because such flows may promote creation of hot spots where the explosion would happen faster than without any flow .
the authors of @xcite have also presented a formal asymptotic analysis and found an effective problem in the limit of the large flow amplitude .
the fully nonlinear problem when the flow itself satisfies a navier - stokes type equation coupled to the explosion problem for temperature has been studied in @xcite using numerics and formal asymptotics .
recently , some rigorous results for the behavior of the solutions to the coupled system in the regime of a strong gravity have been obtained in @xcite .
here we derive several qualitative properties of the explosion threshold @xmath26 in terms of the geometry and the amplitude of the flow @xmath27 . in the following
we always assume that @xmath20 is a smooth bounded domain in @xmath28 , @xmath24 is a @xmath29 divergence - free flow ( @xmath30 in @xmath20 ) .
our first proposition establishes the direct analog of the classical results for ( [ intro - ellipt ] ) and allows us to define the critical parameter @xmath26 .
[ thm1 ] there exists @xmath31 such that ( i ) for every @xmath32 the problem ( [ elliptic ] ) has a unique positive classical solution @xmath33 such that the principal eigenvalue @xmath34 of the linearized operator @xmath35 is positive ; ( ii ) if ( [ elliptic ] ) admits another non - negative solution @xmath36 then @xmath37 ; ( iii ) the function @xmath33 is increasing in @xmath38 ; ( iv ) there exists no classical solution of ( [ elliptic ] ) for @xmath39 .
the proof of this result is very close to that in @xcite we present it below both for the convenience of the reader and since we will use some of the intermediate steps in what follows .
another reason to discuss the proofs for @xmath40 is that some of the basic results in the self - adjoint case @xmath41 rely on the variational characterization of the principal eigenvalue of the linearized operator @xmath42 which we do not have when the flow is present .
the next theorem shows that the possible creation of hot spots can not drop the explosion threshold arbitrarily close to zero , no matter what the incompressible flow @xmath24 is .
[ thm - lowerbd ] for any domain @xmath20 and nonlinearity @xmath5 there exists @xmath43 so that the critical threshold @xmath26 for ( [ elliptic ] ) satisfies @xmath44 for all incompressible flows @xmath24 in @xmath20 .
the constant @xmath45 depends on @xmath20 and the function @xmath46 .
this result does not hold without the restriction that the flow @xmath24 is incompressible we describe in section [ sec : uniform ] examples of flows for which @xmath16 may be as small as one wishes .
the proof of theorem [ thm - lowerbd ] involves the following uniform @xmath0 bound for solutions of the dirichlet problem for elliptic diffusion - advection problems with the constant independent of the incompressible flow .
[ lem - lp - linfty ] let the flow @xmath24 be divergence - free and let @xmath47 be the solution of the elliptic problem @xmath48 with @xmath49 , @xmath50 .
there exists a constant @xmath51 which depends on @xmath52 and the domain @xmath20 but not on the flow @xmath24 , so that @xmath53 .
these results raise several interesting open questions .
first , can one identify the optimal constant @xmath54 in ( [ lbd - lem ] ) ?
more importantly , we would like to pose as an open problem to know whether in all domains with smooth boundaries a flow realizing the best constant in lemma [ lem - lp - linfty ] exists .
the same question pertains to the smallest possible explosion threshold in theorem [ thm - lowerbd ] : does the flow minimizing @xmath26 over all incompressible flows exist in any domain @xmath20 ?
if so , what are its geometric characteristics ? numerical simulations in @xcite indicate that a natural guess that @xmath41 turns out not to be correct for all domains as an incompressible flow may create additional hot spots .
more precisely , it has been numerically computed in @xcite that in a very long rectangle the explosion threshold corresponding to a cellular flow with a certain finite positive amplitude is smaller than that corresponding to @xmath41 .
however , it is not clear whether in the situation when @xmath41 is not a minimizer of @xmath26 , such a minimizer exists at all , or if minimizing flows do not exist .
when it exists , how is it determined ?
let us now fix a flow profile @xmath24 and consider the explosion problem ( [ elliptic ] ) with a strong flow @xmath55 , with a large flow amplitude @xmath56 : @xmath57 we are interested in the behavior of the explosion threshold @xmath58 for ( [ elliptic - strong ] ) in the limit @xmath59 .
let us recall that a function @xmath60 is a first integral of @xmath27 if @xmath61 a.e . in @xmath20 .
[ thm5 ] we have @xmath62 as @xmath63 if and only if @xmath27 has no non - zero first integrals in @xmath64 .
this theorem provides a sharp characterization of the flows capable of preventing an explosion for an arbitrary @xmath11 provided that the flow amplitude @xmath65 is sufficiently large .
the proof uses the ideas from @xcite and @xcite together with some techniques of @xcite .
not surprisingly , the explosion threshold tends to infinity as @xmath63 under the same assumptions as the principal dirichlet eigenvalue of the operator @xmath66 ( see @xcite ) , as both quantities measure the effectiveness of the enhancement of the boundary cooling due to the flow .
finally , in section [ sec : cellflow ] , we consider the effective problem for ( [ elliptic - strong ] ) in the limit @xmath59 for the class of two - dimensional cellular flows . we show that in this limit , the various cells of the flow `` do not talk to each other '' .
the main result of that section is theorem [ cf - fr - thm ] . in particular , when @xmath63 , the explosion threshold @xmath58 is close to the explosion threshold on the `` largest '' cell in @xmath20 .
moreover , the explosion threshold for each of the individual flow cells in the limit @xmath63 has an asymptotic description in terms of the freidlin problem .
we recall that the fast flow asymptotics for the parabolic reaction - diffusion equations in flows without cells has been treated by m. freidlin in @xcite , and our results for the elliptic explosion problem in the special case of one cell flows are what one would expect formally from @xcite .
the most interesting and delicate new ingredient is the independent behavior of the solutions in various cells when @xmath63 .
we mention that unlike in @xcite , our proofs are not probabilistic in nature . actually , as a by - product of the present paper
, one can use our arguments to recover some of the results of @xcite by analytic techniques .
another natural variable coefficients extension of the classical results for ( [ intro - ellipt ] ) is to allow the nonlinearity to be spatially dependent work in this direction has been recently done in @xcite . *
* this work has been carried as the first author was visiting the university of chicago .
it has been supported by asc flash center at the university of chicago .
ak was supported by nsf grant dms-0314129 , an was supported by nsf grant dms-0604600 , lr was supported by nsf grant dms-0604687 .
we begin with the proof of proposition [ thm1 ] which is well known for @xmath41 .
we do not assume in this section that the flow @xmath24 is incompressible .
as we have mentioned , the proof is very close to that in @xcite , with some minor modifications .
let @xmath67 $ ] and @xmath68 be the principal eigenvalue and the normalized positive eigenfunction of the adjoint problem @xmath69\eta \hbox { in $ \omega$}\\ & & \eta=0\hbox { on $ \partial\omega$.}\nonumber\end{aligned}\ ] ] note that @xmath67>0 $ ] as the operator @xmath70 has no zero order term .
[ prop - large - lambda ] the problem ( [ elliptic ] ) admits no non - negative classical solutions for @xmath71/g'(0)$ ]
. * proof .
* since the function @xmath19 is convex and @xmath6 we have @xmath72 .
therefore , any classical solution @xmath73 of ( [ elliptic ] ) satisfies @xmath74 multiplying ( [ elliptic-012 ] ) by @xmath75 and integrating by parts we conclude that @xmath76= \mu_1[u]\int \eta\phi_\lambda.\ ] ] it follows that for a positive solution of ( [ elliptic ] ) to exist we must have @xmath67\ge g'(0)\lambda$ ] and thus no non - negative solution of ( [ elliptic ] ) exists if @xmath71/g'(0)$ ] .
@xmath77 next , we show that for a sufficiently small @xmath11 a positive solution of ( [ elliptic ] ) exists .
let @xmath78 be the expected value of the exit - time , solution of @xmath79 and let @xmath80 [ prop - small - lambda ] there exists a constant @xmath81 which depends only on the nonlinearity @xmath19 but not on the flow @xmath24 so that problem ( [ elliptic ] ) admits a minimal non - negative solution @xmath82 for all @xmath83 .
the proof is by constructing a super - solution and using it to show that a positive solution of ( [ elliptic ] ) exists .
let us recall the following fact .
[ lem - super ] assume that there exists a smooth function @xmath84 satisfying @xmath85 then there exists a classical solution @xmath82 of ( [ elliptic ] ) which is minimal in the sense that for any other non - negative solution @xmath86 of ( [ elliptic ] ) we have @xmath87 for all @xmath88 . *
* we construct an approximating sequence @xmath89 by setting @xmath90 and letting @xmath91 be the smooth solution of @xmath92 the difference @xmath93 satisfies @xmath94 it follows from the maximum principle that @xmath95 and thus @xmath96 .
similarly , we have for the higher differences @xmath97 : @xmath98\ge 0 \hbox { in $ \omega$,}\\ & & w_1=0\hbox { on $ \partial\omega$.}\nonumber\end{aligned}\ ] ] then by induction we conclude that @xmath99 since the function @xmath19 is increasing .
the same induction argument shows that @xmath100 for all @xmath101 .
therefore , the sequence @xmath102 converges to a limit @xmath82 which has to be a solution of ( [ elliptic ] ) and satisfy @xmath103 .
as the sequence @xmath102 does not depend on the choice of the super - solution @xmath104 , the limit @xmath82 is a minimal solution of ( [ elliptic ] ) . @xmath77 * proof of lemma [ prop - small - lambda ] . *
observe that for @xmath11 sufficiently small the function @xmath105 satisfies @xmath106 here @xmath78 is the solution of ( [ eq - tau0 ] ) .
this is true provided that @xmath107 . as the function @xmath19 is increasing , for this inequality to hold it suffices to require that @xmath108 .
this condition is clearly satisfied if @xmath83 with a constant @xmath54 which depends only on the function @xmath19 .
now , existence of a minimal solution to ( [ elliptic ] ) follows from lemma [ lem - super ] .
@xmath77 recall that a solution @xmath82 of ( [ elliptic ] ) is stable if the principal eigenvalue @xmath109 of the linearized operator @xmath110 is positive .
[ prop - stable-0 ] any minimal solution of ( [ elliptic ] ) has @xmath111 .
* let @xmath82 be a minimal solution of ( [ elliptic ] ) and assume that the principal eigenvalue @xmath109 of the problem @xmath112 is negative . consider the function @xmath113 , then we have @xmath114 provided that @xmath115 is sufficiently small and @xmath116 .
this contradicts minimality of @xmath117 .
therefore , we have @xmath111 if @xmath82 is a minimal solution
. @xmath77 [ lem - kzero ] assume that @xmath82 is a solution of ( [ elliptic ] ) such that @xmath118
. then no classical solutions of ( [ elliptic ] ) with @xmath119 exists .
* proof . *
assume that @xmath119 and there exists a function @xmath120 such that @xmath121 let us also denote by @xmath86 the positive eigenfunction of the adjoint problem @xmath122 set @xmath123 with @xmath124 $ ]
. then convexity of @xmath46 implies that @xmath125 for all @xmath124 $ ] .
moreover , we have equality in ( [ 0-etaplus0 ] ) when @xmath126 . differentiating ( [ 0-etaplus0 ] ) with respect to @xmath127 at @xmath126 gives the following inequality for @xmath128 : @xmath129 multiplying ( [ 0-etaplus10 ] ) by the eigenfunction @xmath86 of ( [ stable - eq11 ] ) and integrating we obtain @xmath130=\int \zeta\left[-\delta\psi-\nabla\cdot(u\psi)-\lambda g'(\phi_\lambda)\psi\right]=0.\ ] ] this contradiction shows that no solution of ( [ elliptic ] ) for @xmath119 may exist if @xmath118 .
@xmath77 this also finishes the proof of proposition [ thm1 ] .
the critical threshold @xmath26 is the supremum of all @xmath38 for which a stable solution of ( [ elliptic ] ) exists .
we summarize the upper and lower bounds for @xmath26 in lemmas [ prop - large - lambda ] and [ prop - small - lambda ] as @xmath131 we will use these bounds in the sequel .
uniform @xmath132-bounds for the functions @xmath133 at @xmath15 are difficult to obtain and will be investigated elsewhere @xcite
. however , we have the following universal estimate for @xmath134 which will prove useful later .
[ prop - univ - bd ] for any @xmath135 there exists a constant @xmath136 which depends only on @xmath137 and nonlinearity @xmath19 but not on the domain @xmath20 or the incompressible flow @xmath24 so that the minimal positive solution @xmath33 of ( [ elliptic ] ) satisfies @xmath138 for all @xmath139 .
* proof . *
the proof is based on an idea from @xcite .
fix @xmath140 , let @xmath141 and take any @xmath142 .
we denote by @xmath143 and @xmath144 the corresponding classical solutions of ( [ elliptic ] ) with @xmath145 and @xmath146 , respectively . following @xcite , set @xmath147 it follows from the positivity of the function @xmath19 and ( [ gint - infty ] ) that @xmath148 is an increasing positive function with @xmath149 .
we now define the rescaled inverse function @xmath150 note that , since @xmath141 and @xmath151 we have @xmath152 therefore , the function @xmath153 is well - defined for all @xmath154 , and there exists a constant @xmath155 which depends only on the parameter @xmath135 and the nonlinearity @xmath19 so that @xmath156 for all @xmath154 and all @xmath139 .
recall that @xmath144 is the minimal positive classical solution to ( [ elliptic ] ) with @xmath146 and set @xmath164 .
using concavity of the function @xmath153 and expression ( [ basic - phiprime ] ) we observe that the function @xmath104 satisfies the inequality @xmath165 moreover , as @xmath166 the function @xmath104 obeys the dirichlet boundary conditions @xmath167 on @xmath4 .
therefore , the function @xmath104 is a super - solution for ( [ elliptic ] ) with @xmath145 . employing the same iterative procedure as in the proof of proposition [ prop - small - lambda ] we may then construct a non - negative solution @xmath143 of ( [ elliptic ] ) with @xmath145 which is smaller than @xmath168 .
however , by construction we have @xmath169 and the conclusion of proposition [ prop - univ - bd ] holds .
@xmath77 we prove theorem [ thm - lowerbd ] in this section .
let @xmath170 be the minimal positive solution of ( [ elliptic ] ) : @xmath171 according to ( [ upper - mu ] ) , in order to obtain a uniform lower bound for the explosion threshold @xmath16 , it suffices to bound from above @xmath172 , the supremum of the exit time , defined by ( [ eq - tau0 ] ) and ( [ def - thetau ] ) .
that is , it suffices to prove that there exists a constant @xmath173 so that @xmath174 for all divergence free flows @xmath24 in @xmath20 .
the constant @xmath42 should not depend on the flow @xmath24 .
this bound is an immediate consequence of lemma [ lem - lp - linfty ] .
we write @xmath47 , the solution of ( [ lbd - lem ] ) , as @xmath175 the function @xmath176 satisfies the parabolic initial value problem @xmath177 we will now show that there exists a pair of constants @xmath81 and @xmath178 so that for any incompressible flow @xmath27 and any solution of ( [ lbd-3 ] ) with initial data @xmath179 we have a uniform bound @xmath180 .
the proof is as in @xcite with a slight modification , we present the details for the convenience of the reader .
first , multiplying ( [ lbd-3 ] ) by @xmath86 and integrating by parts we obtain @xmath181 using the poincar inequality in @xmath20 we conclude that there exists a constant @xmath178 so that @xmath182 for any pair of times @xmath183 . on the other hand
, we have , using the poincar inequality again , for all @xmath184 : @xmath185 next , using the hlder inequality , with @xmath186 we obtain : @xmath187 provided that @xmath188 or , equivalently , that @xmath189 .
therefore , we have the following nash - type inequality in @xmath20 : @xmath190 with @xmath191 .
however , incompressibility of the flow , the hopf lemma and the boundary conditions imply that @xmath192 .
it follows that @xmath193 going back to ( [ lbd-5 ] ) we conclude that @xmath194 therefore we have a bound @xmath195 combining this inequality with ( [ lbd-51 ] ) evaluated with @xmath196 , @xmath197 we conclude that @xmath198 with @xmath199 .
consider now the solution operator @xmath200 .
we have shown that @xmath201 the adjoint operator to @xmath202 is simply the solution operator corresponding to the ( also incompressible ) flow @xmath203 .
therefore , we have the dual bound @xmath204 which in turn implies that @xmath205 putting these bounds together we obtain @xmath206 which is ( [ lbd-4 ] ) .
the maximum principle also implies that we have a trivial bound @xmath207 .
interpolating between these two bounds we get the estimate @xmath208 .
now , for any @xmath50 we may choose @xmath209 sufficiently small so that the kernel would be integrable at @xmath210 , and ( [ lbd - tau - psi ] ) would imply that @xmath211 and the constant @xmath81 is independent of the incompressible flow @xmath27 .
this finishes the proof of lemma [ lem - lp - linfty ] and hence that of theorem [ thm - lowerbd ] .
@xmath77 as we have mentioned in the introduction , without the incompressibility constraint the explosion threshold may be arbitrarily small . indeed , according to proposition [ prop - large - lambda ] we have an upper bound @xmath212 . therefore , to see that no uniform in the flow lower bound for @xmath16 in compressible flows exists , it suffices to construct flows @xmath213 such that the principal eigenvalue @xmath214 , as @xmath215 .
such example is provided by the radial flows @xmath216 , say , in two - dimensions : @xmath217 then @xmath218 as @xmath215 this can be seen either from the general theory in @xcite or by an explicit computation .
indeed , setting @xmath219 we obtain a self - adjoint problem for @xmath220 : @xmath221 hence , @xmath222 satisfies the variational principle @xmath223 in addition , we have @xmath224 , as follows from the maximum principle applied to ( [ comp - eig ] ) for a test function of the form @xmath225 , where @xmath226 , @xmath227 for @xmath228 and @xmath229 for @xmath230 we obtain by a straightforward computation @xmath231 with @xmath232 .
therefore , @xmath233 as @xmath234 and hence @xmath235 as well .
in this section we consider the elliptic problem ( [ elliptic ] ) when the advecting flow is strong . accordingly
, we introduce a large parameter @xmath236 and re - write ( [ elliptic ] ) as @xmath237 we are interested in the behavior of the solution @xmath170 of ( [ strong - elliptic ] ) for large @xmath65 , as well as in the dependence of the explosion threshold @xmath16 on the amplitude @xmath65 . with a slight abuse of notation
we will denote here by @xmath58 the explosion threshold of the problem ( [ strong - elliptic ] ) .
our first result shows that , when the flow is strong , solution becomes nearly constant on the flow streamlines , at least in an average sense and for @xmath38 away from @xmath58 .
this is a common phenomenon in diffusion - advection problems : a strong flow induces stratification . [ thm7 ] assume in that @xmath238 on the boundary @xmath4 .
then the solution @xmath82 of ( [ elliptic ] ) is nearly constant on the streamlines of @xmath27 for sufficiently large @xmath65 in the sense that for any @xmath135 there exists @xmath239 so that @xmath240 for all @xmath241 .
* proof . *
first , we multiply ( [ strong - elliptic ] ) by @xmath242 and integrate over @xmath20 .
the emerging integrals over the boundary vanish since @xmath243 on @xmath4 , and we obtain the following estimate : @xmath244= \frac{1}{a}\int(u\cdot\nabla\phi_\lambda)\delta \phi_\lambda = -\frac{1}{a}\int_\omega{\frac{\partial{\phi_\lambda}}{\partial{x_k}}}{\frac{\partial{}}{\partial{x_k } } } \left[u_j{\frac{\partial{\phi_\lambda}}{\partial{x_j}}}\right]\\ & & = -\frac{1}{a}\int_\omega{\frac{\partial{\phi_\lambda}}{\partial{x_k } } } { \frac{\partial{u_j}}{\partial{x_k}}}{\frac{\partial{\phi_\lambda}}{\partial{x_j}}}- \frac{1}{a}\int_\omega{\frac{\partial{\phi_\lambda}}{\partial{x_k } } } u_j\frac{\partial^2\phi_\lambda}{\partial x_j\partial x_k}= -\frac{1}{a}\int_\omega{\frac{\partial{\phi_\lambda}}{\partial{x_k } } } { \frac{\partial{u_j}}{\partial{x_k}}}{\frac{\partial{\phi_\lambda}}{\partial{x_j } } } \le \frac{c}{a}\int|\nabla\phi_\lambda|^2.\end{aligned}\ ] ] this means that the variation along the streamlines is smaller than across .
now , we have to bound the @xmath245-norm of @xmath246 .
however , multiplying ( [ strong - elliptic ] ) by @xmath82 and integrating by parts we see that @xmath247 moreover , proposition [ prop - univ - bd ] shows that @xmath82 satisfies a uniform bound @xmath248 as long as @xmath139 .
therefore , for such @xmath38 we know that @xmath249 for all @xmath250 .
it follows that @xmath251 in that sense solution becomes uniform over the streamlines .
@xmath77 an interesting by - product of the estimate ( [ strong - gradient ] ) is that there are no boundary or internal layers in this problem unlike in the problems with boundary forcing in cellular flows @xcite .
the reason is that @xmath82 is set to be constant on the boundary and the normal component of the flow vanishes on the boundary hence , there is no `` conflict '' between the boundary data and uniformization along the streamlines of the flow .
we prove here theorem [ thm5 ] .
we recall that the assumption that there is no @xmath254 first integral @xmath255 such that @xmath256 almost everywhere is equivalent to the fact that the principal dirichlet eigenvalue @xmath257 of the operator @xmath258 on @xmath20 tends to infinity as @xmath63 @xcite .
next , we show that if @xmath27 has no first integral in @xmath254 then @xmath62 as @xmath63 . the proof is based on the lower bound for @xmath58 in ( [ upper - mu ] ) .
the next lemma is contained in @xcite we provide a proof in the spirit of @xcite . *
* as in the proof of theorem [ thm - lowerbd ] we represent the function @xmath261 using the duhamel formula as @xmath262 with the function @xmath263 that solves the parabolic problem @xmath264 it follows from the theorem 5.3 of @xcite that if @xmath27 has no first integrals in @xmath254 then for any @xmath265 we can find a flow amplitude @xmath266 so that @xmath267 for all @xmath268 . therefore , we have an upper bound @xmath269 we conclude that @xmath270 as @xmath63 . @xmath77 we now consider the explosion problem @xmath271 in a two - dimensional simply connected domain @xmath20
. the flow has the form @xmath272 with a stream - function @xmath273 which we assume to be sufficiently smooth .
we assume that the boundary of the domain @xmath20 is a level set @xmath274 which may contain finitely many saddle critical points of the function @xmath275 thus , the boundary is a union of streamlines of the flow away from the critical points .
we also assume that inside @xmath20 the flow has a cellular structure : the saddles of @xmath275 are all non - degenerate and are connected by the flow separatrices which divide @xmath20 into a finite number of invariant regions , called the flow cells , that we will denote by @xmath276 .
the stream - function @xmath273 has only one critical point @xmath277 inside each of @xmath276 , which is a non - degenerate maximum or minimum .
a prototype example of such flow has the stream - function @xmath278 its cells are squares @xmath279\times[m , m+1]$ ] with integer @xmath280 and @xmath281 , and the domain @xmath20 is a finite union of such squares .
a more general flow of such type is depicted in figure [ fig - omega - cell ] .
the strong flow asymptotics for parabolic reaction - diffusion equations for two - dimensional flows with morse class stream - functions has been considered in @xcite .
this class does not include the cellular flows under our consideration as we allow the stream - function to have several saddles on the level set @xmath274 . nevertheless , the limit problem of @xcite is crucial in the explosion problem in a cellular flow .
the limit problem in @xcite was formulated as a system of reaction - diffusion equations on the reeb graph of the function @xmath275 .
we recall and re - derive these results below in the context of the explosion problem in the one - cell setting , as that is what we will need below .
the single cell is also the situation addressed numerically in @xcite . for a one - cell flow
the reeb graph is simply an interval @xmath282 $ ] , where @xmath283 is the value of @xmath275 at the critical point inside the cell @xmath284 that we assume to be a maximum , and @xmath274 is the boundary of the cell .
we are interested in the behavior of solutions and of the explosion threshold in the limit of a large flow amplitude .
the effective freidlin problem on the interval @xmath285 is to find a function @xmath286 satisfying @xmath287 with the coefficients @xmath288 under our assumptions on the stream - function , the average turnover time @xmath289 is bounded from above and below away from zero : @xmath290 the uniform bound from below by @xmath291 in ( [ cf - fr - t - bds ] ) comes from the fact that the maximum of @xmath273 is a non - degenerate critical point .
the term @xmath292 for small @xmath293 appears in ( [ cf - fr - t - bds ] ) because the boundary may contain non - degenerate saddle points of @xmath275 so that the turnover time blows up as @xmath294 .
the coefficient @xmath295 is positive for @xmath293 and behaves as @xmath296 close to @xmath297 .
in particular we have @xmath298 ( diffusivity vanishes at this point ) , while the drift satisfies @xmath299 for @xmath300 near @xmath283 , and points away from @xmath297 .
therefore , the end - point @xmath297 is inaccessible for the diffusion process corresponding to the left side of ( [ cf - fr-3 ] ) , and one does not need to prescribe the boundary condition at @xmath297 in order for ( [ cf - fr-3 ] ) to be well - posed .
the following proposition defines the explosion threshold for the effective problem
. * proof . *
the proof follows the same steps as in section [ sec : basic ] the only required modification is due to the degeneracies at @xmath304 and @xmath297 .
this can be addressed using the general theory in @xcite and @xcite but in the present case the boundary value problem with a prescribed right side @xmath305 has an explicit unique solution @xmath306 where @xmath307 in particular , we have @xmath308 , so that @xmath309 and we also have @xmath310 so that @xmath311 .
therefore , the mapping @xmath312 is a compact map on @xmath313 $ ] and the krein - rutman theory applies to the operator in the left side of ( [ cf - fr-312 ] ) .
we may then repeat the proof of proposition [ thm1 ] essentially verbatim and conclude that the critical threshold @xmath314 for ( [ cf - fr-3 ] ) exists .
@xmath77 [ [ the - explosion - threshold - for - strong - cellular - flows - the - main - result ] ] the explosion threshold for strong cellular flows : the main result ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ the main result of this section is the following theorem .
we assume that the flow has a cellular structure and satisfies the assumptions outlined at the beginning of this section .
then for each cell @xmath315 one may formulate the corresponding one - cell freidlin problem ( [ cf - fr-3 ] ) for a function @xmath316 , posed now on an interval @xmath317 $ ] , where the outer boundary of @xmath276 is the level set @xmath274 and @xmath318 is the value of the function @xmath275 at the ( unique ) extremal point inside @xmath276 .
for the freidlin problem the dirichlet boundary condition @xmath319 is prescribed at @xmath304 , and the derivative @xmath320 is imposed to be bounded .
this defines the explosion threshold @xmath321 for each cell @xmath276 .
the following theorem shows that in the limit of a large flow the explosion threshold for the whole domain @xmath20 approaches the freidlin explosion threshold for the `` largest '' cell @xmath276 .
[ cf - fr - thm ]
let @xmath58 be the explosion threshold for ( [ cf - fr-1 ] ) and @xmath321 be the threshold for the aforementioned effective problem ( [ cf - fr-3 ] ) posed in the cell @xmath276 of the domain @xmath20 .
then we have @xmath322 for various values of the pclet number @xmath323 for the cellular flow with the stream - function @xmath324 on the the domain @xmath325\times[0,2\pi]$ ] with four cells .
dashed - dotted line : @xmath16 for the whole domain @xmath325\times[0,2\pi]$ ] , solid line the minimum of @xmath16 of the four individual cells .
, height=340 ] the proof of theorem [ cf - fr - thm ] proceeds in several steps .
first , we prove a stratification lemma for solutions of forced advection - diffusion problems in cellular flows .
it shows that solution of the dirichlet problem is small not only on the outer boundary but also on the whole skeleton of separatrices and cells `` do not talk to each other '' .
the second step is to establish the result of theorem [ cf - fr - thm ] for domains consisting of one cell where one just has to show that for one cell the explosion threshold converges in the strong flow limit to that of the freidlin problem .
the last step is to generalize this result to a domain consisting of finitely many cells .
we begin the proof of theorem [ cf - fr - thm ] with the following lemma , of an independent interest .
let @xmath326 be the union of all cell boundaries ( separatrices of the flow ) inside @xmath20 including the outside boundary @xmath4 .
first , we show that solutions of a linear problem with the homogeneous dirichlet data on the outer boundary are small on @xmath326 .
intuitively , this lemma says that once a diffusive particle obeying an sde @xmath333 comes close to the skeleton of separatrices , somewhere inside @xmath20 , it exits the domain @xmath20 after a short time .
this is the phenomenon behind the effective diffusivity @xcite , and front speed and principle eigenvalue enhancement @xcite in cellular flows .
* the proof is in two steps .
first , we show that for any @xmath135 there exists a small @xmath293 and a large @xmath329 so that for any @xmath330 we can find @xmath334 with @xmath335 , such that @xmath336 for all @xmath337 on the streamline @xmath338 inside the cell @xmath276 .
let @xmath339 be the interior of those streamlines . in the second step we consider the water - pipe domain @xmath340 , that is , a narrow tube around the skeleton of separatrices .
using the fact that @xmath341 is small , we apply the maximum principle for narrow domains to conclude that the function @xmath342 is smaller than @xmath137 in all of @xmath343 and not only on its boundary .
as a consequence , @xmath342 is small also on the skeleton @xmath326 .
* step 1 . *
we have a uniform @xmath245-bound for the gradient : @xmath344 with the constant @xmath54 independent of @xmath250 , which follows from multiplying ( [ cf - out-14 ] ) by @xmath327 and integrating by parts , together with the uniform @xmath132-bound for @xmath327 : @xmath345 , which follows from lemma [ lem - lp - linfty ] .
now , take @xmath346 and let @xmath347 be the domain between the two streamlines @xmath348 and @xmath349 inside the cell @xmath276 .
we multiply ( [ cf - out-14 ] ) by @xmath350 and integrate over @xmath351 : @xmath352 we used ( [ cf - out - grad - bd ] ) in the last step . averaging this estimate in @xmath346
we conclude that @xmath353 where @xmath354 is the domain between the streamlines @xmath355 and @xmath356 inside the cell @xmath315 .
the constant @xmath357 may blow - up as @xmath294 but that is not important at the moment .
it follows that there exists a value @xmath358 so that along the streamline @xmath359 we have @xmath360 with a new constant @xmath357 .
therefore , the oscillation of @xmath342 along @xmath361 is small : @xmath362 hence , @xmath327 is close to a constant @xmath363 on the streamline @xmath361 when @xmath65 is sufficiently large . as a consequence of the gradient bound ( [ cf - out - grad - bd ] )
, we have @xmath364 if the cells @xmath276 and @xmath365 have a common piece of the boundary . as the outer boundary @xmath4 , where @xmath366 , is also part of some cell boundaries , it follows that @xmath367 for all cells @xmath276 .
therefore , we have @xmath368 if @xmath369 is sufficiently small and @xmath330 is large enough . * step 2 .
* now , we look at the water - pipe @xmath370 and show that solution is below @xmath137 everywhere in @xmath343 . the function @xmath342 inside @xmath343 satisfies @xmath371 , where @xmath372 is the exit time from the slightly larger domain @xmath373 : @xmath374 now , as in the proof of lemma [ lem - lp - linfty ]
we conclude that there exists a constant @xmath357 such that @xmath375 for all @xmath250 .
the same proof shows that @xmath376 as @xmath377 this happens because the principal eigenvalue of the dirichlet laplacian in @xmath378 tends to infinity as @xmath377 while the constants @xmath379 in the poincar inequality @xmath380 , @xmath381 , satisfy @xmath382 as @xmath377 .
it follows that if we take @xmath293 sufficiently small ( independent of @xmath65 ) then @xmath383 for all @xmath250 .
therefore , we have @xmath331 for all @xmath330 in @xmath343 and , in particular , @xmath331 on @xmath326 .
the proof of lemma [ cf - lem - exit ] is now complete .
we now consider the explosion problem @xmath384 in a domain @xmath20 which consists of just one flow cell .
without loss of generality we assume that the single critical point @xmath385 of @xmath275 inside @xmath20 is a maximum and set @xmath386 .
let us now formulate the version of theorem [ cf - fr - thm ] for a one - cell domain .
[ cf - fr - thm - onecell ] let @xmath20 be a one - cell domain and let @xmath58 be the explosion threshold for ( [ fr-1 ] ) and @xmath387 be the threshold for the aforementioned effective problem ( [ cf - fr-3 ] ) posed on @xmath282 $ ] .
then we have @xmath388 the proof of proposition [ cf - fr - thm - onecell ] is in two steps .
first , passing from the problem on the cell to the freidlin problem we show that the fredlin threshold @xmath387 is not smaller than than @xmath259 .
next , we establish the opposite inequality by starting with a solution to the freidlin problem and constructing a super - solution for ( [ fr-1 ] ) .
the second step is quite straightforward in the case when the boundary of @xmath20 contains no saddles of the flow @xmath27 but is somewhat more technical if @xmath4 contains such fixed points .
we first prove that @xmath389 assume that @xmath390 we will show that then @xmath391 by constructing a solution to the freidlin problem ( [ cf - fr-3 ] ) as the limit of a sequence of problems on @xmath20 .
it follows from ( [ fr - lambda - lambda - bar1 ] ) that there exists @xmath135 and a sequence @xmath392 such that @xmath393 .
therefore , as a consequence of proposition [ prop - univ - bd ] , the minimal positive solutions of @xmath394 are uniformly bounded in @xmath395 : @xmath396 with the constant @xmath81 independent of @xmath281 .
hence , the sequence @xmath102 converges weakly in @xmath397 ( after extracting a subsequence ) and strongly in @xmath398 to a function @xmath104 . as the functions @xmath102 are uniformly bounded and @xmath5 is smooth
, the sequence @xmath399 converges to @xmath400 .
we claim that @xmath104 depends only on the variable @xmath401 and satisfies the effective freidlin problem ( [ cf - fr-3 ] ) .
the first claim follows after we divide ( [ fr-1-n ] ) by @xmath402 and let @xmath403 .
this leads to @xmath404 in the sense of distributions .
it is convenient now to introduce the curvilinear coordinates @xmath405 .
the coordinates are chosen so that @xmath406 , that is , the streamlines of the flow are @xmath407 , and the level lines of the coordinate @xmath408 are orthogonal to the flow lines : @xmath409 .
we normalize @xmath410 so that @xmath411 and the boundary @xmath4 is a level set : @xmath412
. then ( [ cf - unablaphi ] ) implies that @xmath104 depends only on the variable @xmath300 .
the @xmath132-bound in ( [ fr - uniform ] ) implies that @xmath413 .
in addition , we have @xmath414 here @xmath415 is the jacobian of the coordinate change .
note that @xmath416 with some scalar function @xmath417 , so that @xmath418 therefore , we have @xmath419 and thus we have a weighted @xmath420-bound @xmath421 which follows from ( [ fr - uniform ] ) , and hence @xmath286 is continuous for @xmath422 , as @xmath296 for @xmath300 close to @xmath283 .
next , we re - write ( [ fr-1-n ] ) in the curvilinear coordinates : @xmath423 integrating this equation in @xmath410 and passing to the limit @xmath403 we obtain the limit problem for the function @xmath104 : @xmath424 with @xmath425 it remains only to observe that ( [ fr-2-n1 ] ) is nothing but the effective problem ( [ cf - fr-3 ] ) .
indeed , as in ( [ jul - ah ] ) we compute that @xmath426 and @xmath427 the last equality above follows from the fact that @xmath428 here @xmath429 is the interior of the streamline @xmath430 .
it follows that ( [ fr-2-n1 ] ) is , indeed , the effective problem ( [ cf - fr-3 ] ) , so that @xmath286 is a positive solution of ( [ cf - fr-3 ] ) .
therefore , in particular , we have @xmath431 and ( [ fr - lambda - lambda - bar ] ) holds .
we now prove that @xmath432 together with ( [ fr - lambda - lambda - bar ] ) this will complete the proof of proposition [ cf - fr - thm - onecell ] .
this is done as follows : we take any @xmath433 and show that @xmath434 for a sufficiently large @xmath65 by constructing a bounded positive super - solution to ( [ fr-1 ] ) with @xmath145 . however , the singular points of @xmath273 cause technical difficulties in the construction of the sub - solution . hence
, we first consider the special case when @xmath273 has no saddles on @xmath4 .
let @xmath433 and let @xmath286 be the corresponding positive solution of ( [ cf - fr-3 ] ) with some @xmath435 .
an important observation is that there exists @xmath436 so that for all @xmath437 ( that is , not the maximum of the stream - function @xmath273 ) , we have @xmath438 and note that for @xmath300 close to @xmath283 we have @xmath439 and @xmath440 so that the first term above is uniformly bounded in @xmath441 .
we look for a super - solution to ( [ fr-1 ] ) in the form @xmath442 where @xmath443 is smooth and bounded . then the uniform bound ( [ cf - laplace - bd ] ) and similar bounds on other second derivatives of @xmath444 show that if we have @xmath445 in @xmath20 and @xmath446 on @xmath4 , then @xmath447 is a weak super - solution to ( [ fr-1 ] ) with @xmath145 in the sense of @xcite and thus @xmath434 .
we choose @xmath448 as the solution of @xmath449 assume that we can show that @xmath450 ) holds : @xmath451 for @xmath65 sufficiently large .
above we used the fact that @xmath452 and a uniform bound for @xmath104 .
hence , @xmath434 for @xmath65 sufficiently large .
this will prove proposition [ cf - fr - thm - onecell ] in the special case when the domain @xmath20 consists of one cell and the boundary @xmath4 contains no saddles of the stream - function @xmath275 .
it remains to establish ( [ cf - eta - unif - bd ] ) . to this end
consider a cut - off function @xmath453 so that @xmath454 and @xmath455 for @xmath456 and @xmath457 for @xmath458 and split @xmath459 with @xmath460.\ ] ] the small parameter @xmath135 is to be chosen below .
we define , accordingly , the functions @xmath461 , @xmath462 as solutions of @xmath463 so that @xmath464
. as @xmath465 is uniformly bounded in @xmath466 , the function @xmath467 can be bounded using lemma [ lem - lp - linfty ] as @xmath468 with some @xmath469 .
we may further split the function @xmath470 , with the function @xmath471 that solves the ode along the closed streamlines : @xmath472 this equation is solvable because @xmath473dl=0,\ ] ] as this is how the freidlin problem is obtained .
the function @xmath471 satisfies the estimate @xmath474 with some function @xmath475 ( which may tend to infinity as @xmath476 ) .
finally , @xmath477 satisfies @xmath478 once again , lemma [ lem - lp - linfty ] together with the @xmath479 estimate ( [ abcdefg ] ) on @xmath471 above implies that @xmath480 altogether we see that for any @xmath209 we can find find @xmath135 , and then find @xmath481 so that for any @xmath330 @xmath482 this proves proposition [ cf - fr - thm - onecell ] in the special case when the domain @xmath20 consists of one cell and the boundary @xmath4 contains no saddles of the stream - function @xmath275 .
now , we establish the claim of proposition [ cf - fr - thm - onecell ] for domains @xmath20 which consist of one cell but may have saddles of the stream - function on the boundary @xmath4 . in order to avoid dealing with the singular points on the boundary of @xmath20 in the construction a sub - solution we need to consider a slightly smaller domain @xmath483 , with @xmath209 small .
the domain @xmath484 has no saddles on @xmath4 and thus the conclusion of proposition [ cf - fr - thm - onecell ] holds for @xmath484 by what we have shown above .
define @xmath485 as the explosion threshold for the problem in @xmath484 : @xmath486 we also let @xmath487 be the explosion threshold for the corresponding freidlin problem : @xmath488 with @xmath289 and @xmath295 still given by ( [ cf - fr - one - coeff ] ) . as we have mentioned , since @xmath484 has no saddles of @xmath275 on @xmath489 , we have @xmath490 for all @xmath209 , according to the what we have already shown above .
it is clear from the definition of @xmath485 and @xmath487 that @xmath491 and @xmath492 . the next two lemmas show that the passage to the limit @xmath493 is harmless .
the first statement concerns the freidlin thresholds .
* the proof of this lemma is rather straightforward .
it is clear that @xmath492 for all @xmath209 . on the other hand ,
given @xmath495 , for @xmath496 we can find @xmath135 which depends only on @xmath497 so that solution of the following problem exists : @xmath498 for all @xmath209 .
then it is easy to verify that , for a sufficiently small @xmath209 ( and @xmath135 fixed ) , solutions of the iteration process @xmath499 with @xmath500 are increasing in @xmath281 , and satisfy @xmath501 for @xmath502 and @xmath503 for @xmath504 .
thus , the sequence @xmath505 converges as @xmath215 to a bounded solution of ( [ cf - fr-3 ] ) so that @xmath506 .
@xmath77 this lemma is sufficient to show that @xmath511 and thus finish the proof of proposition [ cf - fr - thm - onecell ] for all one - cell domains , as we have already established ( [ fr - lambda - lambda - bar ] ) . in order to see that ( [ cf - lambdabar - l ] ) holds ,
take @xmath495 and find @xmath512 and @xmath481 as in lemma [ lem - leps - l0 ] .
consider any @xmath513 .
then lemma [ lem - barl - eps ] implies that there exists @xmath514 so that @xmath515 .
now , ( [ cf - leps - barleps ] ) implies that we can find @xmath516 so that @xmath517 for all @xmath518 . as @xmath514
we may use lemma [ lem - leps - l0 ] to conclude that @xmath519 for all @xmath520 .
therefore , ( [ cf - lambdabar - l ] ) holds .
this finishes the proof of proposition [ cf - fr - thm - onecell ] . @xmath77
the proof of lemma [ lem - leps - l0 ] is based on the iteration argument and stratification lemma [ cf - lem - exit ] .
we start with @xmath521 for all @xmath522 , @xmath523 and construct a solution of the explosion problem on the whole domain @xmath20 by the iteration procedure .
set @xmath500 and let @xmath102 be the solution of @xmath524 we claim that the sequence @xmath89 is increasing in @xmath281 , pointwise in @xmath337 , for each @xmath65 and there exists @xmath481 so that for @xmath330 it has a uniformly bounded limit @xmath525 which satisfies ( [ fr-1 ] ) .
pointwise monotonicity of @xmath89 in @xmath281 is standard : the difference @xmath526 satisfies @xmath527 hence , we have @xmath528 in @xmath20 and @xmath529 .
let us assume that @xmath530 in @xmath20 .
as the nonlinearity @xmath19 is increasing in @xmath531 , the function @xmath532 is the solution of @xmath533\ge 0 \hbox { in $ \omega$,}\\ & & \eta_1=0\hbox { on $ \partial\omega$.}\nonumber\end{aligned}\ ] ] therefore , we have @xmath534 inside @xmath20 and thus the sequence @xmath89 is increasing in @xmath281 .
we need to show that the sequence @xmath89 is uniformly bounded from above .
we recall that @xmath521 and thus the minimal positive solution of ( [ fr-1-eps ] ) satisfies @xmath535 with the constant @xmath536 independent @xmath115 and @xmath65 .
hence , we may find @xmath135 which depends only on @xmath507 but not on @xmath115 or @xmath65 so that solution of the following problem exists : @xmath537 to see that such @xmath135 exists , let @xmath538 satisfy ( [ fr-1-eps ] ) with @xmath539 and set @xmath540 , then @xmath541 satisfies @xmath542 it is a super - solution for ( [ cf - out-8 ] ) if we ensure that @xmath543 , or , equivalently , @xmath140 is taken so small that we have for all @xmath88 : @xmath544 the function @xmath538 obeys a uniform bound @xmath545 with @xmath546 independent of @xmath115 and @xmath65 .
let @xmath547 , then ( [ cf - g - g ] ) is guaranteed if we have @xmath548 for all @xmath549 $ ] .
a direct computation shows that this is possible if we take @xmath550 under this assumption @xmath541 provides a super - solution for ( [ cf - out-8 ] ) and thus a positive solution of this problem can be constructed by the standard iteration procedure .
[ cf - lem - omepsbd ] fix @xmath173 .
let @xmath553 be solution of @xmath554 then , given any @xmath135 there exist @xmath329 and @xmath508 so that for all solutions of ( [ cf - eq - omeps ] ) we have @xmath555 in @xmath556 for all @xmath330 and @xmath557 .
we choose @xmath173 so that @xmath558 for all @xmath559 , where @xmath552 is the minimal positive solution of ( [ cf - out-8 ] ) .
we may also take @xmath250 sufficiently large , as in lemma [ cf - lem - omepsbd ] .
we claim that then we will have , for all @xmath560 , ( i ) @xmath561 in @xmath556 , and ( ii ) @xmath562 for all @xmath559 .
let us prove this by induction .
the function @xmath563 satisfies @xmath564 our choice of @xmath42 ensures that the right side in ( [ cf - out-11 ] ) is bounded above by @xmath42 .
thus , if the amplitude @xmath65 is sufficiently large we have @xmath565 in the tube @xmath566 and in particular on @xmath489 .
therefore , the difference @xmath567 satisfies @xmath568\ge 0 \hbox { in $ \omega_{\varepsilon}$},\\ & & s_1\ge 0\hbox { on $ \partial\omega_{\varepsilon}$.}\nonumber\end{aligned}\ ] ] it follows that @xmath569 and @xmath570 in @xmath484 so that our claim holds for @xmath571 .
assume now that ( i ) and ( ii ) hold for @xmath572 .
it follows from the induction assumption that @xmath573 and @xmath574 in @xmath566 .
therefore , for @xmath89 we have @xmath575 hence , by lemma [ cf - lem - omepsbd ] we have @xmath561 in @xmath566 . on the other hand ,
the difference @xmath576 inside @xmath577 obeys @xmath578\ge 0 \hbox { in $ \omega_{\varepsilon}$},\\ & & s_m\ge 0\hbox { on $ \partial\omega_{\varepsilon}$,}\nonumber\end{aligned}\ ] ] and thus @xmath579 , so that @xmath580 in @xmath484 .
therefore , the sequence @xmath581 is increasing and uniformly bounded from above . the limit
@xmath104 is a positive solution of ( [ fr-1 ] ) and thus @xmath582 .
this finishes the proof of lemma [ lem - leps - l0 ] .
@xmath77 we now look at the explosion problem @xmath583 in a two - dimensional domain @xmath584 with a cellular flow @xmath27 which may contain more than one cell in @xmath20 and complete the proof of theorem [ cf - fr - thm ] .
let @xmath585 be the explosion threshold for the problem inside each cell : @xmath586 we already know that @xmath587 from proposition [ cf - fr - thm - onecell ] and , of course , @xmath588 for all @xmath589 .
hence , all we need to verify for the proof of theorem [ cf - fr - thm ] is that for any @xmath590 solution of the problem ( [ out-1 ] ) on the whole domain @xmath20 exists .
the proof is quite similar to the last part of the proof of proposition [ cf - fr - thm - onecell ] : we construct the solution of ( [ out-1 ] ) by the iteration procedure .
set @xmath591\ ] ] and take @xmath507 fixed . consider any @xmath592 , start the iteration process with @xmath500 and define @xmath581 as the solution of @xmath593 the proof is quite analogous to that of lemma [ lem - leps - l0 ] except that we use lemma [ cf - lem - exit ] where lemma [ cf - lem - omepsbd ] was used before .
the first increment @xmath596 satisfies @xmath597 hence , we have @xmath528 in @xmath20 and @xmath529 .
let us assume that @xmath598 in @xmath20 .
as the nonlinearity @xmath19 is increasing in @xmath531 , the function @xmath532 is the solution of @xmath599\ge 0 \hbox { in $ \omega$},\\ & & \eta_1=0\hbox { on $ \partial\omega$.}\nonumber\end{aligned}\ ] ] therefore , we have @xmath534 inside @xmath20 and thus the sequence @xmath600 is increasing in @xmath281 .
we need to show that it is uniformly bounded from above . as in the proof of lemma [ lem - leps - l0 ] ,
we can find @xmath135 so that solution of the following explosion problem exists on each cell @xmath276 for all @xmath330 , @xmath601 and @xmath602 : @xmath603 and , moreover , @xmath604 with the constant @xmath546 which does not depend on the flow amplitude @xmath65 . this is shown by exactly the same argument we used to show the existence of a positive solution for ( [ cf - out-8 ] ) .
as @xmath605 are uniformly bounded by @xmath606 , we may choose @xmath173 so that @xmath607 for all @xmath589 and all @xmath88 .
we also take @xmath250 sufficiently large , as in lemma [ cf - lem - exit ] ( but with the right side of ( [ cf - out-14 ] ) replaced by the constant @xmath42 rather than @xmath608 ) .
we claim that the following bounds will be preserved by the iteration procedure : for all @xmath609 , ( i ) @xmath561 on the skeleton of separatrices @xmath326 , and ( ii ) @xmath610 for all @xmath611 .
again , we check this by induction .
the function @xmath612 satisfies @xmath613 the right side in ( [ out-11 ] ) is bounded above by @xmath42 and hence , according to lemma [ cf - lem - exit ] , for the amplitude @xmath65 sufficiently large we have @xmath614 on the skeleton @xmath326 .
therefore , the first increment @xmath615 satisfies @xmath616\ge 0 \hbox { in $ { \cal c}_j$},\\ & & s_1\ge 0\hbox { on $ \partial{\cal c}_j$,}\nonumber\end{aligned}\ ] ] and thus @xmath569 and @xmath617 in @xmath315 .
let us now assume that ( i ) and ( ii ) are true for @xmath618 and show that they hold for @xmath551 the argument is exactly as for @xmath619 . by the induction assumption we have @xmath620 and
thus @xmath551 satisfies @xmath621 hence , by lemma [ cf - lem - exit ] we have @xmath622 on @xmath326 .
now , the difference @xmath623 inside @xmath276 satisfies @xmath624\ge 0 \hbox { in $ { \cal c}_j$},\\ & & s_m\ge 0\hbox { on $ \partial{\cal c}_j$,}\nonumber\end{aligned}\ ] ] and thus @xmath579 , so that @xmath625 in @xmath315 .
therefore , the sequence @xmath626 is increasing and uniformly bounded from above . the limit
@xmath627 is a positive solution of ( [ out-1 ] ) .
this completes the proof of lemma [ lem - iteration ] and thus that of theorem [ cf - fr - thm ] .
@xmath77 a.friedman , the asymptotic behavior of the first real eigenvalue of a second order elliptic operator with a small parameter in the highest derivatives , indiana univ . math . jour . ,
* 22 * , 1973 , 10051015 .
zeldovich , g.i .
barenblatt , v.b .
librovich and g.m .
makhviladze , _ the mathematical theory of combustion and explosions , translated from the russian by donald h. mcneill .
consultants bureau [ plenum ] , new york , 1985 . _ | we consider the explosion problem in an incompressible flow introduced in @xcite .
we use a novel @xmath0 estimate for elliptic advection - diffusion problems to show that the explosion threshold obeys a positive lower bound which is uniform in the advecting flow .
we also identify the flows for which the explosion threshold tends to infinity as their amplitude grows and obtain an effective description of the explosion threshold in the strong flow asymptotics in a two - dimensional one - cell flow . |
during 2004 - 2006 , h.e.s.s .
( high energy stereoscopic system ) performed a survey of the inner part of the galaxy where its excellent capability provided a breakthrough in the field of pulsar wind nebulae ( pwn ) study : for the first time the morphological structure of many middle - age pwne were resolved in the @xmath0-ray band . a detailed spectral and morphological analysis of the archetypal example of this population , psr@xmath1b1823 - 13 and its associated tev nebula hess@xmath1j1825 - 137 revealed for the first time in @xmath0-rays a steepening of the energy spectrum with increasing distance from the pulsar , probably due to the synchrotron radiative cooling of the inverse - compton ( ic ) @xmath0-ray emitting electrons during their propagation ( aharonian et al .
2006b ) . in this scenario ,
the different sizes observed in x - rays and very high energy ( vhe ) @xmath0-rays can be explained by the different cooling timescales for the radiating electron populations ( de jager et al .
however , the question of the unusual large size of this pwn candidate compared to what the dynamical simulations predict still remains ( de jager et al .
2005 ; de jager @xmath9 djannati 2008 ) and is difficult to solve . the confusion is even aggravated because of the non - detection of the associated supernova shell , like it is the case for most of the other middle - aged tev pwn candidates ( gallant et al .
2007 ) .
hess@xmath1j1640 - 465 is one of the sources discovered by hess during this galactic survey in 2004 - 2006 ( aharonian et al .
this source is marginally extended with profile close to a gaussian shape .
the spectrum is well fitted by a simple power - law with an index of @xmath10 2.4 and a total integrated flux above 200 gev of @xmath11 . as it is shown in figure 1 ,
the source is spatially coincident with the snr g338.3 - 0.0 , known from radio observations as a @xmath12 diameter broken shell with a 7 jy flux at 1ghz ( whiteoak @xmath9 green 1996 ) .
a radio bridge extending to the north of the system is visible in the most data and coincident with the bright h@xmath1ii region g338.4 + 0.0 ( whiteoak @xmath9 green 1996 ) , possibly connected to the shell .
the fact that the tev source is center filled and does nt match the shell has favoured the hypothesis that the emission is produced by a pwn so far .
this field has been successively observed by asca ( sugizaki et al .
2001 ) and the swift x - ray telescope ( landi et al .
2006 ) ; both detected a highly absorbed source inside the shell , with compatible positions and non thermal spectra .
+ in 2006 , a dedicated xmm - newton observation of the field of view revealed an extended x - ray object at the position ( @xmath13 , @xmath14(j2000))(funk et al .
this extended source ( xmmuj164045.4 - 463131 ) exhibits a strongly absorbed spectrum with a potential non - thermal spectrum ( index @xmath15 ) centered about the vhe source position .
however , because of the low number of counts of these observations the spectrum was poorly constrained , and the insufficient angular resolution was not able to resolve any compact nebula , nor a pulsar which could confirm the hypothesis of a pwn emission .
we report here on the first chandra observation of this field .
we present a detailed data analysis and discuss the physical implications of this first case of a complete middle - age composite system with vhe pwn emission .
we report here on x - ray observation of the field around the position of hess@xmath1j1640 - 465 taken in may 2007 ( obs id:7591 ) using the chandra x - ray observatory .
data were collected with the advanced ccd imaging spectrometer ( acis ) in timed exposure , very faint mode , providing a temporal resolution of 3.2 s. the target was imaged on i3 chip , with the asca source position @xmath16 from the optical axis .
the other acis chips activated during this observation were s2 , s3 , i0 , i1 , i2 .
this imaging system offers an on - axis spatial resolution of 0.8 arcsec . +
a total of 30 ks of live - time was accumulated . after removing small intervals in
which data were not recorded , the final exposure time for this observation was 26.4 ks .
after standard processing had been carried out at the chandra science center , we analyzed the resultant events list using the ciao(v3.4 ) , ftools(v6.0.4 ) , caldb(v3.3 ) , and xspec(v12.2.1 ) x - ray analysis software packages .
figure 2 shows an exposure - corrected and smoothed image ( using a gaussian with @xmath17 ) in the 2 - 10 kev energy range , encompassing the central part of the field of view of the acis - i array . to produce this figure , we divided the map of raw counts by the exposure map and normalized by the exposure time in order to produce an image in units of @xmath18 .
the morphology of the x - ray emission imaged here is consistent with the one seen by xmm at lower spatial resolution : we confirm the presence of a diffuse nebula seen extending @xmath19 around the position of xmmuj164045.4 - 463131 ( funk et al .
but at the higher spatial resolution of the chandra data , some new morphological features become apparent , in particular the two compact sources s1 and s2 ( see figure 2 and table 1 ) : s1 is coincident with the brighter part of the diffuse emission ( north - east ) and the other ( s2 ) is placed on the edge , on the north - west side of the nebula .
+ figure 3 shows the radial profile along arcs centered on the bright point - source ( s1 ) position along the ne - sw direction .
the surface brightness profile is presented in unit of @xmath20 .
all the point sources of the field have been subtracted except s1 .
the dashed line represents chandra s psf at an energy of 2.5 kev , shown at the position of s1 .
it has been obtained by simulating a psf map using caldbv3.3 and the mkpsf tool and superimposing it on the average background level derived from the dashed circular background regions shown in figure 2 , then projecting it on the radial profile .
it is clear from this figure that the extent of the source s1 is consistent with the point - spread function of the telescope .
we can see the compact nebula component lying asymmetrically from -20 to + 70. the inner core of the emission is consistent with a point source surrounded by extended structures lying mainly in the south - west direction .
the fact that this diffuse component steadily fades with increasing distance from s1 supports the interpretation in which it is the central source of energy of this emission .
+ the brightest inner pwn component ( called hereafter comp pwn ) , is @xmath21 in area .
figure 2 shows in addition marginal evidence of a tail , apparently emanating from s1 .
it lies within a @xmath22 area size to the sw of s1 , with a major axis along the se - nw line ( called extended pwn thereafter ) .
we can note that this faint emission has surprisingly a different elongation axis from that associated with the inner core of the nebula .
+ .corrected coordinates of s1 and s2 point sources . [ cols="^,^,^,^,^ " , ] + the high spin - down power used to derive the upper energy of the injected electron spectrum is needed to reproduce the flux and index in the x - ray band .
using a smaller @xmath23 would decrease the electron upper limit energy which would have the effect of a prematured cut - off in the synchrotron spectrum .
the minimum energy value is strongly constrained by the flux level given the age and power evolution of the pulsar .
its fit value looks reasonable given the known properties of the pulsar wind , and it constrains the radio emission to a low flux which is in agreement with the non - detection of any radio nebula in the field .
the magnetic field found is typical of such middle - age vela like objects , the decreasing index corresponds to a field which varies from an initial value of 35 @xmath24 g to an actual value of 6 @xmath24 g ( 15 kyr later ) .
the shape of magnetic field evolution can affect the spectrum index , but the low degree of details in vhe resolution of the source ( because of hess s poor psf ) , make such a precise shape analysis non pertinent in this paper .
finally , the vhe flux can be reached using the cmbr alone , but in order to reproduce the spectral index , we need the addition of the dust and visible target components . + given the reasonable values of parameters needed to reproduce the chandra and hess spectra and their relatively good compatibility with the known characteristics of the middle - aged vela - like pulsars and pwne , the interpretation of the x - ray and vhe gamma - ray emission as originating from a pulsar - driven nebula appears secure .
this high - resolution chandra acis observation of hess@xmath1j1640 - 165 inside the shell g338.3 - 0.0 , has allowed us to spatially resolve a pwn and its putative pulsar , and perform a spectral analysis .
we found that the spectra are both highly absorbed and non - thermal , and the indices and luminosities resemble those of typical vela - like pulsar systems . using h@xmath1i absorption features
, we were able to constrain the distance of the shell between 8.5 and 13 kpc , this distance range being compatible with the large @xmath2 value derived from x - ray spectral analysis , we concluded that it is likely associated with the shell g338.3 - 0.0 .
we finally discussed a scenario in which the @xmath0-rays originate from the ic scattering and x - ray from the synchrotron emission of the electron populations in the nebula , and show that the vhe source hess@xmath1j1640 - 165 is likely a pwn , probably associated with a vela - like pulsar of @xmath25 .
if this interpretation is true , this object is a nice example of a middle - aged composite system where the total pwn extension ( traced by the tev emission ) fills almost 75 percent of the associated snr shell area , which is quite surprising given that dynamical simulations currently predict a 25 percent .
future search for similar cases will be useful in order to confirm or not the generality of this phenomenon and study its causes in detail . in this context
, the detection of the associated pulsars will be very helpful .
a next step of this study could be the high resolution radio imaging of the source and the search for pulsed emission originating from this putative pulsar position .
the authors would like to acknowledge the support of their host institutions .
b.m.g acknowledges the support of the australian research council through a federation fellowship ( grant ff0561298 ) .
pos acknowledges partial support from nasa contract nas8 - 03060 . broad - band x - ray image of the field surrounding xmmuj164045.4 - 463131 from the acis - i chips .
the image was exposure - corrected and smoothed with a gaussian width @xmath26=2.5 to highlight the extended diffuse emission .
black contours are plotted at levels of 12@xmath27 , 20@xmath27 , 68@xmath27 , 90@xmath27 , and 95@xmath27 of the peak intensity .
the location of the two point - sources s1 and s2 are indicated ( see table 1 ) .
the three solid black ellipses represent the regions a , b and c which encompass the different components of the nebula and will be used as extraction regions to measure the pwn spectra .
the red geometric features show the axis and regions from which the radial brightness profile ( figure 3 ) has been computed .
background regions used to derive radial profile and spectra are indicated as dashed black circles._,scaledwidth=80.0% ] to @xmath28 arcsec , and gives the detail of nebula s inner core brightness evolution .
we show the psf of the telescope determined at 2.5 kev ( dashed line ) , normalized to the peak of emission and scaled vertically to match the data .
s2 has been previously extracted from the data._,scaledwidth=75.0% ] kev band , smoothing angle of 1.5 ) and blue ( [ 2.5 - 4.5 ] kev band , smoothing angle of 3.5 ) with a log intensity scale , and the radio emission from 843 mhz most is shown in green ( we supperimposed radio data contours on the shell for a better visibility ) .
the intrinsic extent of hess@xmath1j1640 - 465 is shown as a white dashed circle . _ ] i 21 cm line temperature brightness intensity profile ( k ) as a function of radial velocity ( @xmath29 ) .
this profile shows the averaged background h@xmath1i emission spectrum near the snr g338.3 - 0.0 .
four individual spectra have been used to create this averaged spectrum : each of them has been integrated in the ( l , b ) plane over a 0.04 @xmath30 0.04 deg@xmath31 area ( shown in figure 8) . _ bottom panels : _ absorption spectra in the line of sight of the snr shell , and h@xmath1ii region g338.40 - 0.23 .
these spectra have been computed by subtracting the averaged background spectum , from the spectrum taken from the snr shell and hii region ( figure 8) .
we can see clear absorption features all over the line of sight for both objects.,scaledwidth=65.0% ] i temperature brightness ( @xmath32 ( k ) ) integrated over the velocity range [ -25;-35 ] km.s@xmath33 , in the field of snr g338.3 - 0.0 .
the map is superimposed with the most radio intensity contours at 843 mhz in green .
we see a clear correlation between the h@xmath1i absorption features ( the darkest regions of the map ) and the contours .
additionally , four black squares show the background regions used to compute the figure 7 top panel averaged profile , and two red squares show the snr and h@xmath1ii regions used to compute the absorbed spectra shown in figure 7 . _ ] j1640 - 465 together with the total pwn x - ray spectrum derived with chandra data are shown .
parameters of this fit are summarized in table 3 . the ic emission ( black line )
is computed taking into account the cmb(dashed line ) , star light ( dotted line ) and dust ( dotted and dashed line ) targets , derived from moskalenko et al.(2006)._,scaledwidth=85.0% ] | we present a chandra x - ray observation of the very high energy @xmath0-ray source hess@xmath1j1640 - 465 .
we identify a point source surrounded by a diffuse emission that fills the extended object previously detected by xmm newton at the centroid of the hess source , within the shell of the radio supernova remnant ( snr ) g338.3 - 0.0 .
the morphology of the diffuse emission strongly resembles that of a pulsar wind nebula ( pwn ) and extends asymmetrically to the south - west of a point - source presented as a potential pulsar .
the spectrum of the putative pulsar and compact nebula are well - characterized by an absorbed power - law model which , for a reasonable @xmath2 value of @xmath3 , exhibit an index of 1.1 and 2.5 respectively , typical of vela - like pwne .
we demonstrate that , given the h@xmath1i absorption features observed along the line of sight , the snr and the h@xmath1ii surrounding region are probably connected and lie between 8 kpc and 13 kpc .
the resulting age of the system is between 10 and 30 kyr . for a 10 kpc distance ( also consistent with the x - ray absorption )
the 2 - 10 kev x - ray luminosities of the putative pulsar and nebula are @xmath4 and @xmath5 ( @xmath6 ) .
both the flux ratio of @xmath7 and the total luminosity of this system predict a pulsar spin - down power around @xmath8 .
we finally consider several reasons for the asymmetries observed in the pwn morphology and discuss the potential association with the hess source in term of a time - dependent one - zone leptonic model . |
the wfpc2 camera on the hubble space telescope ( _ hst _ ) has made it possible to determine the morphologies of galaxies at intermediate redshift and beyond .
it has been known for some time that the photometric and spectral properties of galaxies in intermediate redshift clusters differ from galaxies in nearby clusters ; the population of blue , star - forming galaxies and post - starburst galaxies is larger at intermediate redshift , see e.g. , @xcite , @xcite , @xcite , @xcite , and @xcite .
given the correlation between the spectral and morphological properties of galaxies ( @xcite , @xcite , @xcite ) , we might be able to detect a corresponding evolution of galaxy morphology in intermediate redshift clusters .
however , the evolution of morphology is likely to be more subtle than spectral evolution , since galaxies of the same morphological type can have significantly different star formation rates ( e.g. @xcite ) . in the most ambitious study of this sort with wfpc2 to date
, the morphs group have classified over 1200 galaxies in 10 clusters at 0.37@[email protected] ( @xcite ) .
they find that s0 galaxies are less common than in low redshift clusters and that the ratio of s0 s to e s within a radius of @xmath4600 kpc ( for h@xmath9=50 , q@xmath9=0.5 ) decreases with redshift , falling from 2 in low redshift clusters to less than 0.5 at @xmath10 .
@xcite ( see also @xcite ) have independently classified galaxies in a wfpc2 image of cl0939 + 4713 , one of the less concentrated clusters in the morphs sample .
they find a ratio of s0 s to e s of @xmath42 , quite comparable to a low-@xmath7 reference sample in the coma cluster .
however , they classify 40 - 50% of the galaxies in cl0939 + 4713 as spirals ( s ) , in contrast with 20 - 30% s in a comparable region of the coma cluster .
the sample of galaxies in cl0939 + 4713 is relatively small ( @xmath470 ) , and redshifts are available for less than one third of these .
@xcite present a study of the morphological types in three clusters at @xmath7=0.31 , also using wfpc2 images .
there is substantial overlap between the authors of this paper and the morphs group , and the two groups have attempted to adopt a consistent morphological system . at @xmath7=0.31 , couch et al .
find an excess of s s , with an abundance at small radii ( @xmath4400 kpc for @xmath11=50 , @xmath12=0.5 ) approximately twice that in low-@xmath7 reference clusters .
however , averaged over the three clusters , within 400 kpc , they find a ratio of s0 s to e s at most slightly depressed relative to regions of comparable galaxy surface density in low-@xmath7 clusters .. reference population becomes increasingly dominated by e s , @xcite ] note , however , that the average @xmath7 of the morphs clusters is larger ( 0.46 ) .
@xcite report the morphological types in two more distant clusters at @xmath130.9 .
one cluster , cl0023 + 04 , appears to be composed of two low velocity dispersion groups , and contains predominantly s s .
the other , cl1604 + 43 , with a velocity dispersion of @xmath41200 km s@xmath14 , contains @xmath476% early - types . in the latter case ,
the s0/e ratio is found to be [email protected] .
this result is sensitive to the assumed morphological composition of the foreground / background population , but is evidence that the s0/e ratio does not decline smoothly with @xmath7 .
our approach is complementary to the @xcite , @xcite , and @xcite studies which predominantly describe the galaxy morphologies in cluster cores .
we use mosaics of wfpc2 fields to study the galaxy population in a larger region ( allowing larger galaxy samples per cluster ) , and we have acquired large numbers of spectra of cluster galaxies .
the spectra remove ambiguity about cluster membership and allow us to directly connect the morphological and spectral properties of the galaxies .
our sample of clusters is x - ray selected , with x - ray luminosities exceeding 4@xmath1610@xmath17 erg s@xmath14 in the 0.24.5 kev band . in cl1358 + 62 at @xmath7=0.33 , we have drawn a complete sample of 518 galaxies to a magnitude limit @xmath0=22 from a wfpc2 mosaic image of cl1358 + 62 covering 53 square arcminutes .
spectra for 276 of the 518 galaxies in the morphological sample were previously obtained at the multiple mirror and william herschel telescopes .
the color - magnitude relation of the 194 spectroscopically confirmed cluster members in the _ hst _ mosaic ( 3 are fainter than @xmath0=22 ) has been previously described in @xcite .
the spectroscopic properties of 232 cluster members ( some outside the _ hst _ mosaic ) , as well as the cluster dynamics have been described in @xcite .
our objectives in this paper are fourfold .
( 1 ) we introduce the morphological classification techniques that we will apply to our entire sample of clusters .
( 2 ) we classify the galaxies in cl1358 + 62 at z=0.33 , comparing our classifications with those of an experienced external researcher , alan dressler .
our deep sample with two independent classifications provides a useful assessment of the scatter between wfpc2 visual morphological classifications at intermediate redshifts .
( 3 ) we describe the robust , classifier - independent conclusions and explore the physical implications of the differences between the two sets of classifications . ( 4 ) we connect the spectral and morphological types of the cluster galaxies . the paper is organized in the following fashion . in @xmath18 2 , we describe the photometric catalog from which the morphological sample was drawn . the two set of morphological classifications are discussed in @xmath18 3 .
the morphological composition of the cluster and evidence for morphological evolution are presented in @xmath18 4 .
the connection between the morphological and spectral properties of the galaxies is made in @xmath18 5 .
@xmath18 6 contains a brief discussion and conclusions .
our photometric catalog , from which we draw galaxies for morphological classification , is derived from the _ hst _ f814w mosaic image .
the techniques used to construct the mosaic are described in @xcite .
our use of the mosaic image , instead of individual wfpc2 ccd frames , slightly compromises the accuracy of the photometry , but considerably simplifies the source detection problem by eliminating most of the boundaries . because our goal is to select a sample of galaxies for morphological classification to a consistent magnitude limit , rather than to perform precision photometry , this is a beneficial tradeoff .
we used the sextractor package , described by @xcite , to detect the galaxies and perform the photometry .
we use a detection and analysis threshold of 24.5 mag arcsec@xmath19 , and a zeropoint of 30.546 ( 3600 s exposures , may 1997 wfpc2 synphot update ) to convert from instrumental magnitudes to a cousins @xmath0 magnitude magnitude , which is very close to the natural f814w system referred to a vega zeropoint ( @xcite ) ] .
the adopted zeropoint is the average of the value for the 3 wfpc2 ccds , which vary by @[email protected] magnitudes .
we use the sextractor total magnitude estimator ( see @xcite ) , which is insensitive to the analysis threshold . in order to allow convenient comparisons with the results of the morphs group ( @xcite ) ,
we calculate a conversion of observed @xmath0 magnitudes ( for @xmath200.6 ) to a @xmath21 magnitude in the rest frame .
some of the morphs classifications used wfpc2 f702w filter data , so we also derive a consistent conversion from f702w magnitudes to a rest frame @xmath21
. the intent in @xcite was to work to a consistent limit of @xmath1=-20 , but a transcription error from table 9 of @xcite led to the adoption of a deeper limit of @xmath1@xmath4 - 19 .
( here and throughout we use @xmath1 to refer to a rest frame @xmath21 absolute magnitude . ) in contrast with @xcite , we also apply an evolutionary correction to allow a closer comparison with low @xmath7 clusters .
we adopt the @xcite cosmological model ( h@xmath9=50 km s@xmath14 mpc@xmath14 , q@xmath9=0.5 ) we use four numbers in addition to the distance modulus to convert from the wfpc2 magnitudes to @xmath1 : ( 1 ) the conversion from the vega - referenced wfpc2 natural filter system to cousins filter bands , ( 2 ) the @xmath7-dependent @xmath22 correction to transform the observed cousins @xmath23 and @xmath0 magnitudes to the rest frame , ( 3 ) the estimated rest frame galaxy colors , and ( 4 ) a @xmath7-dependent evolutionary correction .
each of these numbers depends on the spectral energy distribution of the galaxies , so the accuracy of this procedure is limited .
we use these numbers , however , only to choose sample limiting magnitudes appropriately scaled with @xmath7 .
we adopt the approximate expressions described below for these conversions .
we first fit a linear relation to the synthetic transformation of f702w to @xmath23 as a function of @xmath24 ( valid for @[email protected] ) , using results from fig .
10 of @xcite : @xmath25 ( 1a ) we take the @xmath24 colors of galaxies as a function of @xmath7 from @xcite , with a morphological mix of 70% e and 30% sbc to convert ( 1a ) to : @xmath26 ( 1b ) to a good approximation ( 0.1 mag ) , fig
. 9 of @xcite shows : @xmath27 ( 1c ) we fit polynomials to the average @xmath22 corrections of @xcite and @xcite , using 70% e and 30% sbc or sc contributions : @xmath28 ( 2a ) @xmath29 ( 2b ) from @xcite we take : @xmath30 ( 3a ) @xmath31 ( 3b ) from the the early - type galaxy fundamental plane study of @xcite we derive an evolutionary correction : @xmath32 ( 4 ) the corrections are applied in the following fashion : @xmath33 here , @xmath34 is the distance modulus . applying these expressions to cl1358 + 62 , with a distance modulus at @xmath35 of 41.62 , we find that the morphs limit of @xmath1=-20 corresponds to an observed @xmath0=20.5 .
our photometric catalog completeness extends below @xmath36 , but we have limited our morphological sample to @xmath5 to increase the reliability of the morphological classifications . at @xmath5 , we attain a signal - to - noise ratio ( s / n ) that is very similar to the s / n attained at the @xmath36 classification limit adopted by the morphs group for their deeper images ( @xcite ) .
the standard deviation of the sky noise in our ( 3600 s ) cl1358 + 62 image has an equivalent surface brightness of 24.4 mag arcsec@xmath19 , as compared with 25 mag arcsec@xmath19 typical for the ( @xmath412600 s ) morphs images .
a crude scaling relation can be derived by assuming that all galaxies have the same surface brightness , and that our criterion is that classifications of equal reliability require the same number of pixels at the same s / n . for a shallower image , we can effectively bin the image of a brighter galaxy that is @xmath37 times larger @xmath38 pixels to achieve a s / n that is @xmath37 times better . to recover a sky s / n deficit of 0.6 mag ,
we need to set a limit @xmath41.2 mag brighter .
there are 518 galaxies in the morphological catalog after removing stars and image artifacts . in the rest frame
, the filter central wavelength corresponds to @xmath46100 .
we use the f814w images for morphological classification because they are deeper than the f606w images . @xcite
have used either f702w and f814w images for classification at comparable redshifts ( between 0.37 and 0.41 ) .
below , we describe two sets of morphological classifications .
the first set was carried out by the authors of the paper using the techniques described in @xmath183.1 .
we realized that our classifications would be of greater value if we could compare them with morphological classifications from an external expert .
alan dressler ( ad ) kindly agreed to independently classify the entire sample .
dressler is a member of the morphs group , and most importantly , has classified a large reference sample of low-@xmath7 cluster galaxies .
the strength of the ad classifications is that the same experienced classifier has classified the galaxies at low and intermediate redshifts .
consistency has obvious benefits when searching for evolution .
however , it is important to keep in mind the difficulty of classifying these relatively faint and small galaxies from wfpc2 images . the brighter cluster ellipticals and s0 s in cl1358 + 62 ( @xmath39 ) have effective radii ( half - light radii ) , r@xmath40 , of 0.5@xmath41 to 0.6@xmath41 .
the faintest galaxies in the sample ( @xmath4222 ) have r@[email protected]@xmath41 .
the 50% encircled energy radius of wfpc2 stellar images measured from our frames is about 0.17@xmath41 , so the numbers of meaningful information elements are small : typically 20 to 40 .
for this reason , we can not be sure that agreement on a common morphological system is the most significant issue .
even the most experienced classifier may not account for all the systematic differences between the low @xmath7 ( photographic ) and the intermediate @xmath7 wfpc2 images . by comparing the two sets of morphological classifications
we will be able to discern which aspects of the classification are most robust .
the 518 sample galaxies were independently classified along the revised hubble sequence by df , mf and pvd from 96@xmath1696 pixel `` postage - stamps '' drawn from both the mosaic and original individual ccd images .
we referred frequently to @xcite and @xcite , as well as to artificially redshifted digital images drawn from the nearby field galaxy survey ( @xcite ) .
we assigned numerical types as follows : -5 ( elliptical ) , -4 ( elliptical or s0 ) , -2 ( s0 ) , 0 ( s0 or sa ) , 1 ( sa ) , 3 ( sb ) , 5 ( sc ) , 7 ( sd ) , 9 ( sm ) , 10 ( i m ) and 99 ( peculiar or merger ) .
intermediate types ( 2 , 4 , 6 , and 8) were also assigned . for 80% of the galaxies ,
the three independent classifications span a range of three or fewer numerical types ( we consider -4 , -2 and 0 to be adjacent numerical types ) , and agree exactly for 34% of the sample . for 6% of the sample
, the object is considered unclassifiable by one or more of the authors , or else the classifications disagree wildly . in these cases ,
we assign a numerical type of 999 .
the remaining 14% are classified type 1 or later by all the authors , but with a broader range of classifications .
table 1 lists the combination rules used to assign types where the agreement is not exact , including type 15 for indeterminate late types .
following the completion of the df / mf / pvd classifications , ad independently classified the cl1358 + 62 galaxies according to techniques described in @xcite .
ad does not assign galaxies to the df / mf / pvd intermediate types -4 ( e or s0 ) or 0 ( s0 or sa ) in the same fashion , preferring to subdivide these into e / s0 ( -4 ) , s0/e ( -3 ) , s0/sa ( -1 ) and sa / s0 ( 0 ) . when discussing the cluster population in broad terms , ad s types -5 and -4 are combined into e , types -3 , -2 and -1 into s0 , and etc .
df / mf / pvd split the contents of the -4 bin equally into e and s0 , and the 0 bin equally into s0 and sa .
when converting ad s descriptive types into numerical types , we have placed a few galaxies into the merger ( 99 ) bin where an individual catalog object corresponds to an interacting or merging pair of objects .
augustus oemler , another member of the morphs group , independently classified 307 galaxies from the cl1358 + 62 sample ( @xmath0=22 limit ) to check if ad s classifications adhere to the morphs system .
oemler s classifications agree very well with ad s overall , with @xmath410% more s0 s by number , and a corresponding decrease in the numbers of e s .
the number of spirals is identical in both classifications .
like any measurement , classifications will suffer from random and systematic errors .
we can estimate these errors by comparing multiple sets of classifications .
ideally , such a comparison should be based on classifications using independent imaging data . here , both sets of classifiers ( df / mf / pvd and ad ) worked from the same data set , so we may underestimate the errors . nonetheless , this comparison is quite interesting .
figure 1 shows the difference between df / mf / pvd and ad morphological types for individual galaxies as a function of magnitude .
as expected , the differences increase at fainter magnitudes .
to avoid creating artifical gaps in figure 1 , we condensed and shifted the numerical types of the early type galaxies for this presentation : -3 for e s , -2 for e / s0 s and s0/e s , -1 for s0 s , 0 for s0/sa and sa / s0 . for types
sa and later , the `` normal '' types were used .
galaxies which were classified as merger , peculiar , or unclassified by one group were ignored .
figure 2 is a scatter diagram comparing the two sets of classifications .
the scatter is dominated by random differences , but there is some evidence for systematic differences for the early type galaxies .
we return to this point below .
the scatter between the classifications has been measured by calculating the mean absolute deviation , and normalizing it to the mean absolute deviation of a gaussian with an rms of 1 .
this measurement of the scatter is much more robust than the rms of the differences .
the result is shown in figure 3 .
the scatter is @xmath41 for bright galaxies , but increases strongly faintwards of @xmath43 .
the appendix describes how classification errors might systematically bias our population estimates ; unfortunately we can not simply calculate correction factors . in what follows ,
we restrict most of our analysis to the subset of galaxies with @xmath2 .
figure 2 shows a systematic difference in the df / mf / pvd and ad classifications of early type galaxies .
this is not surprising as the division between e and so is a difficult problem in visual classifications .
we might hope that looking at two objectively determined structural parameters for the galaxies , ellipticity and the bulge / total light ratio , might be helpful in resolving this issue .
for example , @xcite compare the ellipticity distributions of the morphs e and s0 intermediate-@xmath7 galaxies with those from an @xcite study of coma cluster galaxies as a consistency check of the morphs classifications .
we have therefore examined the structural properties of the cl1358 + 62 galaxies measured by the medium deep survey ( mds ) group , @xcite .
the mds group has published structural properties for 70% of the cl1358
+ 62 galaxies in our morphological catalog .
the missing 30% are located near frame edges or in one mosaic frame that is not yet included in the mds database .
the mds group has chosen the best fitting of four structural models for the objects in their catalogs : star , disk , bulge , or disk + bulge , accounting for the point spread function of hst .
if a galaxy model is chosen , the mds group fit the ellipticity of the disk , bulge or disk and bulge separately as appropriate .
we focus on two structural parameters derived from these fits : the ratio of bulge to total light , and the weighted ellipticity .
the ratio of bulge to total light follows trivially from the best fit model , or from the fits to the disk + bulge model .
we calculate the weighted ellipticity from the mds disk and bulge ellipticities , weighting by the fractions of light in the disk and bulge . in figures 4 and 5 we plot the distributions of the bulge to total light ratio for the e and s0 galaxies with @xmath2 in the df / mf / pvd and ad samples , respectively . excepting the differences in numbers , the distributions for both e s and s0 s look remarkably similar for the two sets of classifications .
in figures 6 and 7 we plot the distribution of weighted ellipticities for the same e and s0 galaxies ( @xmath2 ) , again showing the df / mf / pvd and ad samples independently .
caution must be exercised when comparing to these to other measures of ellipticity since the mds numbers are corrected for the hst psf . a larger fraction of the ad e s have ellipticities exceeding 0.2 , but the mean ad e ellipticity is 0.22 , only slightly larger than the df / mf / pvd mean of 0.17 .
the ellipticities of the two samples of s0 s are similar : ad finds a mean ellipticity of 0.46 and df / mf / pvd find a mean ellipticity of 0.40 .
the sizable difference in the numbers of e and s0 galaxies found by the two sets of classifiers is not reflected in a large difference between the structural parameters for the two samples of e s or s0 s .
we conclude that although the total number of early - type galaxies is well established , the so to e ratio in cl1358 + 62 is uncertain .
we have learned that the differences between the two sets of classifications rise steeply below @xmath43 , suggesting this as a practical limit for our morphological study .
this is also conveniently close to the effective limit of our spectroscopic completeness .
we have redshifts for 277 , one ( # 1295 ) at @xmath44 and the other ( # 1297 ) at @xmath45 .
we have assigned the measured velocity ( cz=99657 ) to both galaxies .
galaxies # 1481 and # 1483 may have both fallen within the spectrograph slit .
we assign the measured velocity ( cz=99792 ) to # 1483 , which is 1.2 mag brighter .
] of the 518 galaxies in the morphological sample , and 191 of these are cluster members by the criteria given in @xcite : 0.31461 @xmath8 z @xmath8 0.34201 .
the spectroscopic completeness is 89% for the galaxies brighter than @xmath46 ( corresponding to @xmath47 , falling to 59% for @xmath48 , 33% for @xmath49 , and 9% for galaxies with @xmath50 .
table 2 lists positions , @xmath0 magnitudes , both sets of morphological classifications , and radial velocities for galaxies brighter than @xmath0=21
. we may study the morphological composition of the subsample of known members , which is nearly complete to the morphs s depth of @xmath51 , without concern about background galaxy subtraction .
the subsample of known cluster members to @xmath46 ( or @xmath51 , see @xmath18 2 ) contains 138 galaxies .
df / mf / pvd classify 27@xmath154% of these as e , 44@xmath156% as s0 , 29@xmath155% as s , with 1 unclassified galaxy .
( the errors here and in the following discussions account only for the poisson statistics of the number of galaxies per classification bin . )
ad classifies 35@xmath155% as e , 38@xmath155% as s0 , and 27@xmath154% as s. in both cases , the total early type population is @xmath472% .
to @xmath1=-20 , then , the only difference between the two groups of classifiers is the relative numbers of e s and s0 s .
we consider also the complete photometric sample to a depth of @xmath43 ( or @xmath4@xmath52 ) , where a small background correction is necessary .
to @xmath43 our sample contains 298 galaxies , for which we have 236 redshifts ( 79.2% completeness ) .
of the 236 galaxies with redshifts , 65@xmath158 are nonmembers , yielding a foreground / background count of 82@xmath1510 after correcting for the spectroscopic completeness .
we determine the morphological composition of the foreground / background galaxies directly from our spectroscopic sample , which contains 86 foreground / background galaxies .
df / mf / pvd classify 3% e , 13% s0 , 77% s , and 6% mergers .
ad classifies 11% e , 9% s0 , 71% s , and 9% mergers .
we average these classifications , and adopt a background composition of 7% e , 11% s0 , 74% s and 8% mergers . for comparison , @xcite adopt a morphological composition of 10% e , 10% s0 , and 80% s. after correcting for background , to @xmath43 , df / mf / pvd find a population of 25@xmath154% e s , 46@xmath155% s0 s , 29@xmath156% s s , and 0@xmath151% mergers .
ad finds 36@xmath154% e s , 36@xmath155% s0 s , 28@xmath156% s s , and 1@xmath151% mergers .
these results are indistinguishable from those for the brighter spectrosopic sample , with @xmath472% early - types in both sets of classifications .
we search for morphological evolution in cl1358 + 62 by comparing its population with that of equivalent low-@xmath7 clusters .
judging which low-@xmath7 clusters are equivalent is somewhat uncertain , but we take as an approximation low-@xmath7 clusters with a similar number of galaxies within a fixed metric aperture , allowing us to correct for the effects of the morphology - density relation .
we use the nearby cluster catalog of @xcite , reanalyzed and summarized in @xcite as a benchmark .
we use the subset of high concentration clusters ( 10 of 55 ) in the low-@xmath7 sample for comparison , because cl 1358 + 62 was selected for its high x - ray luminosity and has a concentration index [email protected] ( @xcite ) .
the data for the 10 high - concentration clusters are plotted in fig .
12 of @xcite .
there are an average of @xmath463 cluster galaxies within a radius of 1450 kpc in these clusters to @xmath1=-20.4 . in cl1358
+ 62 , there are @xmath4114 cluster members within this radius to @xmath1=-20.4 , so cl1358 + 62 is richer than the average cluster in the low-@xmath7 sample .
however , the low-@xmath7 sample does contain a high - concentration cluster as rich as cl1358 + 62 : the coma cluster .
furthermore , the density difference between cl1358 + 62 and the average low-@xmath7 reference cluster , 0.3 dex , is comparable to the bin size in the morphology - density relation plots in @xcite .
the comparison between the morphological composition of the cl1358 + 62 sample to @xmath1=-20 and the low-@xmath7 reference sample is made in table 3 and figure 8 .
the df / mf / pvd s0/e ratio , [email protected] , differs at only 1.4@xmath53 confidence from the low-@xmath7 reference sample ratio of [email protected] .
ad s ratio , [email protected] , differs from the the low-@xmath7 reference sample ratio at 3.5@xmath53 confidence . despite this difference in the s0/e ratio ,
we stress that both classifications for cl1358 + 62 yield fractions of early types ( e+s0 ) , @xmath472% , and late types ( s ) , @xmath428% , that are identical within the errors to the low-@xmath7 reference sample .
we can therefore draw a robust conclusion from the two sets of morphological classifications : cl1358 + 62 does not contain an elevated population of spiral galaxies compared with low @xmath7 reference clusters .
this contrasts with the results of @xcite and and @xcite for other intermediate @xmath7 clusters .
our work suggests that the early type / spiral classifications are likely to be secure , implying that the populations of intermediate @xmath7 clusters vary significantly , even after accounting for the effects of the low @xmath7 morphology - density relation .
this conclusion about the spiral population in cl1358 + 62 limits the range of physical models for evolution in cl1358 + 62 , even allowing for our uncertainty about the e / s0 classifications . as we mentioned earlier ,
the ad classifications have the strong advantage of the same classifier at low and intermediate @xmath7 .
however , given the different character of the low-@xmath7 photographic images and the intermediate-@xmath7 wfpc2 images , we must acknowledge the possibility of systematic , redshift - dependent classification uncertainties .
figure 1 provides some reason to be cautious about this issue . because we do not understand in detail the reasons for the differences between the two sets of visual classifications
, we can not be positive that the two sets of classifications bound our uncertainties .
however , the best we can do at present is to leave the issue of e / s0 classifications open , and to explore the consequences of both sets of classifications below .
the df / mf / pvd classifications would imply that cluster evolution from @xmath54 to the present does not affect the cluster morphological composition or its morphology - density relation within the observed 1.4 mpc radius aperture .
figure 9 shows this relation for cl1358 + 68 , binning the background subtracted data to i=22 ( @xmath1=-18.5 ) radially about the dominant central galaxy .
since we are looking only for radial trends , using the deeper sample is appropriate here .
the average galaxy density in each of four radial bins ( 01 , 12 , 23 , and 35 arcmin ) is the abscissa for this histogram .
the galaxy densities in figure 9 have been normalized to a @xmath1=-20 limit to allow comparison with the low-@xmath7 reference sample .
we find that the morphology - density relation for cl1358 + 62 is indistinguishable within the errors from the low-@xmath7 relation ( to @xmath4@xmath1=-20 ) shown in figure 3 of @xcite .
the ad classifications suggest an evolutionary mechanism that decreases the fraction of e s , increases the fraction of s0 s , while leaving the fraction of spirals unchanged . assuming that there is no plausible mechanism for directly converting e s into s0 s , we can exclude models that transform the observed @xmath7=0.33 cluster spirals into s0 s without accretion of additional galaxies . in order to transform the @xmath7=0.33 ad morphological mix into the low-@xmath7 reference sample population while accreting the smallest number of galaxies , the cluster population would increase by 50% .
approximately 70% of the accreted galaxies would become s0 s by the present day , and 30% spirals .
while a great deal has been learned about the galaxy population in intermediate redshift clusters from relatively small samples of spectra , we must remember that even present - day clusters of galaxies are a heterogeneous group , differing widely in their degree of virialization . because cluster relaxation may drive galaxy evolution ,
the range in cluster galaxy populations may be large at any redshift .
for this reason , it is desirable to connect the morphological and spectral properties of a large sample of galaxies in each of a number of clusters directly .
we have spectral classifications from @xcite sorting each of the galaxies with spectra into one of the four categories : ( 1 ) absorption lines only , ( 2 ) emission lines present , ( 3 ) emission lines plus strong balmer absorption lines , and ( 4 ) k+a ( also called e+a ) .
category ( 4 ) contains galaxies with the normal absorption lines of e / s0 galaxies plus strong balmer absorption lines .
emission line galaxies have [ oii ] 3727 emission with equivalent width ( ew ) @xmath555 .
if h@xmath56 absorption of @xmath554 ew is detected for an emission line galaxy , the galaxy is classified as emission plus balmer lines .
galaxies with [ ( h@xmath56 + h@xmath57 + h@xmath58)/3 ] ew greater than 4 , but [ oii ] emission with @xmath85 ew , are classified as k+a .
we refer the reader to @xcite for a more complete summary of the spectral properties of these galaxies and a comparison with spectra of low-@xmath7 galaxies .
table 4 and figure 10 summarize the comparison between the spectral and morphological properties for the 191 cluster members in common with @xcite , using the df / mf / pvd classifications . in rough terms , the bulk of the cl1358 + 62 galaxies follow the morphological - spectral correlation expected for bright field galaxies at low redshift : the preponderance of the e and s0 galaxies have pure absorption line spectra , while the fraction of galaxies with emission lines rises for the late type spirals .
@xcite found a similar behavior for galaxies in the morphs sample ; see also @xcite .
we do know , however , that the percentage of galaxies with emission lines and strong balmer absorption lines , @xmath419% , is higher than the @xmath46% content of these galaxies in comparable low @xmath7 clusters ( @xcite , see also @xcite ) .
if we use the ad classifications for the cl1358 + 62 galaxies , these conclusions do not change significantly .
the most interesting difference between the two sets of morphological classifications is that df / mf / pvd classify all the e+a galaxies as types s0 to sb , while ad classifies these galaxies as having a wider range of morphologies from e to sbc .
for cl1358 + 62 , we have acquired a unique data set including a large mosaic of hst fields and extensive spectroscopy that allows us to unambiguously determine cluster membership for galaxies with @xmath59 .
we have directly compared the morphological classifications of two sets of classifiers for the galaxies in cl1358 + 62 .
the two sets of classifiers agree that ( to a limit of @xmath1=-20 ) the fraction of early type galaxies ( and therefore spirals ) in this cluster at @xmath7=0.33 is indistinguishable from the fraction in comparable low-@xmath7 clusters .
in contrast , previous workers , @xcite and @xcite , who also studied wfpc2 images of clusters at [email protected] , found an elevated population of spirals compared with low-@xmath7 reference samples . because our work confirms the reliability of early - type / spiral classifications from intermediate @xmath7 wfpc2 observations ,
we conclude that this is evidence for a dispersion in the evolution of intermediate-@xmath7 clusters .
the two groups of classifiers differ on ratio of e to s0 galaxies in cl1358 + 62 .
df / mf / pvd find a population of s0 galaxies ( s0/[email protected] ) that is within 1.4@xmath53 of the low-@xmath7 reference sample , while ad finds a significantly smaller ratio ( [email protected] ) .
this systematic difference is most likely related to the fact that the transition between s0 s and intermediate luminosity e s is rather gradual .
many of the intermediate luminosity e s are thought to have disks , e.g. @xcite , @xcite , and @xcite .
it may only be possible to resolve this issue by direct model fitting to images at low and intermediate @xmath7 .
even though we conclude that we have not reliably determined the ratio of s0 s to e s among the early types , our work significantly restricts possible evolutionary models .
if we accept the mf / df / pvd classifications , evolution must preserve the fraction of e s , s0 s and s s as well as the morphology - density relation . if the ad classifications are correct , evolution must decrease the fraction of e s and increase the fraction of s0 s while maintaining the fraction of s s . a possible mechanism for driving the evolution of the morphological mix in this latter fashion is accretion of additional galaxies from the spiral - rich infall region that become predominantly s0 s .
the cluster population within an @xmath41 mpc radius must increase by a minimum of 50% from @xmath7=0.5 to the present day in order convert an intermediate @xmath7 population rich in e s to a low-@xmath7 population rich in s0 s
. it will be interesting to see whether such accretion can be produced in simulations of cluster formation .
in most cluster formation scenarios , massive clusters form by the merging of pre - existing massive clusters with ( presumably ) similar populations of early type galaxies . it may therefore be difficult to double the ratio of s0 to e galaxies .
we compare our morphological classifications to our previous spectral classifications and conclude that the morphologies of the spectrally `` active '' galaxies are as might be expected from the low-@xmath7 field population : the galaxies with emission lines are predominently spirals and the k+a ( or e+a ) post - starburst galaxies are typically early type disk galaxies ( s0sb ) .
we wish to acknowledge the generous contributions of alan dressler to this paper , including his independent classifications and insightful comments .
we thank gus oemler for checking the classifications and for thoughtful comments on the manuscript , which led to several improvements .
we thank margaret geller for a critical reading of an earlier version of the manuscript and helpful comments .
our referee , ian smail , and our editor , greg bothun , both made insightful comments that helped us clarify the paper .
the effects of classification errors on the distribution of types can be important .
quantifying these effects is difficult because morphological classification is a subjective procedure , but we can gain some insight by considering simple models for the errors .
we begin by assuming that the numerical type is based on a one dimensional measurement with a simple , constant error .
this simple model would imply that any peaks in the distribution of types would be softened .
if we adopt nominal intrinsic fractions of s : s0:e of 0.20:0.53:0.27 , as is approximately correct for the inner 600 kpc of low redshift clusters ( @xcite ) , the errors will automatically decrease the fraction of s0 s , and enhance the fraction of e s and s s .
we can calculate an upper limit to the loss of s0 s for our nominal 20:53:27 ( s : s0:e ) population with this model for the errors .
we assume that the types are scattered with a mean absolute deviation ( mad ) of 1 , and all intermediate types are divided equally between adjacent types .
the outcome of this experiment is a distribution of 0.27:0.38:0.34 ( s : s0:e ) . in this case , all galaxies scattered beyond the normal type boundaries were assigned to the boundary type ( i.e. e ) .
a mad of 1 may be an overestimate as the difference in the types assigned by the two groups of classifiers is of this order .
the intrinsic errors are @xmath60 smaller if the errors are independent . for these smaller errors
the resulting distribution would be 0.24:0.43:0.33 ( s : s0:e ) . in either case
, the results serve to illustrate that the systematic effects can be very significant .
let us consider a second , more physical model for the visual classification errors . here
, we assume that morphological type is based on two independent variables with continuous distributions : bulge - to - total light fraction ( @xmath61 ) , and asymmetric features due to spiral arms ( @xmath62 ) .
this is very similar to the quantitative classification devised by @xcite .
galaxies with low @xmath62 will be classified as early type ( @xmath63 ) and then divided into e s or s0 s based on whether @xmath61 is above or below a critical value .
it has been argued that most @xmath64 ellipticals have faint disks , e.g. @xcite and @xcite .
similarly , the spiral classification will be based on a combination of @xmath61 and @xmath62 . if the intrinsic distribution of @xmath61 is flat , then errors in @xmath61 will not change the ratio of s0 s to e s . however ,
if the intrinsic distribution of @xmath61 is peaked , the errors will have a systematic effect .
the sign depends on the details of the intrinsic @xmath61 distributions and the size of the errors .
contributing to this uncertainty , the errors can also be asymmetric , if for example , faint extended disks are missed in noisy data .
the situation is similar for errors in @xmath62 , where e s and sa s now share a boundary .
extensive simulations are required to estimate the systematic effects of limited s / n on classification errors , taking into account the the point spread function of wfpc .
we will undertake this effort in a future paper .
dum abraham , r. et al .
1996 , , 107 , 1 .
andreon , s. , davoust , e. & heim , t. 1997 , , 323 , 337 .
andreon , s. 1998 , , 501 , 533 .
andreon , s. et al . 1996 , , 116 , 429 .
bertin , e. & arnouts , s. 1996 , , 117 , 393 .
butcher , h. & , oemler , a .. 1984 , , 285 , 426 .
couch , w. & sharples , r. 1987 , , 229 , 423 .
couch , w. et al .
1998 , , 497 , 188 .
dressler , a. 1980 , , 236 , 351 .
dressler , a. 1987 , nearly normal galaxies , ed .
faber , s. , ( new york : springer - verlag ) , 276 .
dressler , a. & gunn , j.e .
1992 , , 78 , 1 .
dressler , a. et al .
1997 , , 490 , 577 .
dressler , a. et al .
1999 , , 122 , 51 .
fabricant , d. , mcclintock , j. & bautz , m. 1991 , , 381 , 33 .
fisher , d. , fabricant , d. , franx , m. , & van dokkum , p. 1998
, , 498 , 195 .
frei , z. & gunn , j. 1994 , , 108 , 1476 .
geller , m. et al .
1997 , , 114 , 2205 .
griffiths , r. et al .
1994 , , 435 , l19 .
gunn , j. & dressler , a. 1988 , towards understanding galaxies at high redshift , ed .
kron , r. & renzini , a. , ( dordrect : kluwer ) , 227 .
holtzmann , j. et al .
1995 , , 107 , 156 .
jansen , r. et al . 1999 ,
submitted to .
jorgensen , i. & franx , m. 1994 , , 433 , 553 .
kelson , d. et al .
1997 , , 478 , 13 .
kennicutt , r. c. , jr .
1992 , , 388 , 310 .
lubin , l. et al .
1998 , , 116 , 584 .
mcleod , b. & rieke , m. 1995 , , 454 , 611 .
morgan , w. w. & mayall , n. u. 1957 , , 69 , 291 .
morgan , w. w. & osterbrock , d. e. 1969 , , 74 , 515 .
poggianti , b. m. 1997 , , 122 , 399 .
poggianti , b. m. 1999 , , 518 , 576 .
postman , m. et al .
1998 , , 506 , 33 .
ratnatunga , k. , griffiths , r. , & ostrander , e. 1999 , , submitted .
rix , h .- w .
& white , s. 1990 , , 362 , 52 .
sandage , a. 1961 , the hubble atlas of galaxies ( washingon , dc : carnegie institution of washington ) sandage , a. & tammann , g. a. 1987 , a revised shapley - ames catalog of bright galaxies ( washingon , dc : carnegie institution of washington ) scorza , c. et al .
1998 , , 131 , 165 .
shectman , s. et al .
1996 , , 470 , 172 .
smail , i. et al .
1997 , , 110 , 213 .
van dokkum , p. et al .
1998 , , 500 , 714 .
cccc -5 & -5 & -4 & -5 + -5 & -5 & -2 & -4 + -5 & -4 & -4 & -4 + -5 & -4 & -2 & -4 + -4 & -4 & -2 & -4 + -2 & -2 & -5 & -4 + -2 & -2 & -4 & -2 + -2 & -2 & 0 & -2 + -2 & -2 & 1 & 0 + -2 & -2 & 2 & 0 + -2 & 0 & 1 & 0 + -2 & 1 & 1 & 0 + -2 & 1 & 2 & 0 + ccccccc 1 & 2185.23 & 198.20 & 20.85 & 99246 & -4 & -4 + 12 & 4165.23 & 344.75 & 19.79 & 97702 & -2 & -4 + 15 & 3424.60 & 366.51 & 19.24 & 95807 & -2 & 1 + 16 & 2155.44 & 387.90 & 19.36 & 97243 & 1 & 1 + 25 & 968.84 & 444.59 & 20.54 & 129864 & 6 & 5 + 27 & 4576.15 & 456.72 & 19.00 & - & 5 & 5 + 30 & 1479.89 & 477.46 & 19.38 & 98463 & -4 & -5 + 33 & 4231.61 & 489.34 & 19.13 & 143869 & 0 & 0 + 35 & 4095.85 & 492.36 & 20.47 & 98424 & -4 & -3 + 37 & 3526.59 & 495.73 & 20.42 & 96743 & -2 & -2 + 38 & 3377.31 & 502.24 & 18.81 & 99803 & 0 & -2 + 48 & 3302.50 & 558.38 & 20.98 & 164142 & 15 & 99 + 51 & 4062.30 & 568.56 & 20.77 & - & 8 & 4 + 52 & 3024.67 & 571.39 & 16.96 & 48524 & 4 & 3 + 53 & 3081.65 & 574.07 & 20.08 & - & 4 & 5 + 54 & 1525.88 & 580.51 & 15.69 & 25570 & 4 & 5 + 61 & 2357.29 & 602.25 & 20.50 & 144702 & 15 & 7 + 66 & 4121.06 & 625.81 & 20.99 & - & -2 & -2 + 68 & 3470.13 & 635.49 & 18.95 & 96710 & -2 & -2 + 70 & 3598.20 & 642.90 & 20.04 & 80512 & 5 & 5 + 79 & 4072.28 & 697.05 & 18.92 & 98334 & 0 & -4 + 94 & 2580.14 & 725.66 & 19.64 & 97702 & -2 & -2 + 101 & 2602.51 & 739.65 & 20.44 & 151970 & 0 & -5 + 103 & 3652.82 & 740.90 & 20.43 & 95699 & 3 & 5 + 107 & 3499.88 & 757.63 & 20.99 & - & 4 & 5 + 114 & 2736.96 & 794.55 & 18.89 & 31012 & 5 & 6 + 118 & 2286.07 & 807.47 & 19.08 & 97741 & 0 & -1 + 120 & 1458.64 & 807.67 & 20.02 & - & 6 & 5 + 125 & 4464.63 & 814.64 & 20.65 & - & 5 & 5 + 127 & 3372.60 & 818.52 & 19.95 & 99845 & -4 & -5 + 132 & 907.72 & 840.09 & 20.65 & - & -5 & -5 + 140 & 1319.17 & 854.62 & 19.83 & 98991 & -4 & -5 + 156 & 1940.02 & 919.07 & 19.74 & 149319 & 1 & 99 + 159 & 3641.99 & 927.73 & 20.65 & - & 5 & 4 + 162 & 2605.02 & 933.83 & 19.94 & - & 99 & -3 + 178 & 558.31 & 1001.72 & 20.38 & - & 999 & 99 + 180 & 1745.41 & 1004.04 & 18.59 & 96677 & 0 & 1 + 186 & 2657.55 & 1009.86 & 20.83 & - & -5 & -5 + 192 & 3689.81 & 1034.60 & 20.64 & 100272 & -2 & -2 + 197 & 2634.39 & 1041.72 & 19.94 & 144632 & 2 & 2 + 205 & 1501.76 & 1068.61 & 19.61 & 120912 & 5 & 1 + 206 & 3733.06 & 1075.41 & 20.86 & 98397 & 0 & 0 + 222 & 873.00 & 1136.15 & 20.86 & 99237 & -4 & -5 + 237 & 2777.48 & 1171.76 & 19.64 & 97525 & 5 & 3 + 239 & 1674.39 & 1188.96 & 20.92 & 207514 & 7 & 10 + 244 & 3455.89 & 1200.72 & 20.08 & - & 999 & 999 + 248 & 1019.98 & 1207.10 & 20.02 & 99291 & -2 & -2 + 269 & 698.78 & 1258.32 & 20.54 & - & 4 & 6 + 277 & 1244.07 & 1269.37 & 20.59 & - & 2 & 2 + 279 & 999.08 & 1272.41 & 20.44 & 162101 & 4 & 3 + 283 & 1400.74 & 1280.19 & 20.93 & 98430 & -2 & -2 + 285 & 3410.44 & 1288.24 & 20.66 & 100257 & 4 & 4 + 286 & 2933.14 & 1291.02 & 20.34 & 98343 & 6 & 4 + 295 & 1818.09 & 1320.64 & 17.85 & 25606 & 4 & 3 + 299 & 2669.82 & 1334.75 & 18.88 & 100970 & 0 & -2 + 303 & 3046.19 & 1343.52 & 19.71 & 97417 & -2 & -2 + 304 & 4330.18 & 1348.34 & 17.29 & 23213 & 1 & 1 + 314 & 1730.45 & 1364.59 & 20.88 & 97564 & -2 & -2 + 316 & 1147.10 & 1373.25 & 20.47 & 99893 & 0 & 3 + 325 & 946.39 & 1406.99 & 19.25 & 96686 & -5 & -5 + 326 & 3531.82 & 1407.04 & 20.77 & 100281 & -2 & 0 + 327 & 3089.78 & 1412.64 & 19.03 & 99330 & 1 & 1 + 328 & 1340.92 & 1417.67 & 20.94 & 202745 & 0 & 3 + 331 & 3058.10 & 1427.59 & 19.08 & 98481 & -2 & -2 + 333 & 4153.37 & 1437.70 & 18.87 & 98104 & -5 & -5 + 336 & 3083.01 & 1446.02 & 19.79 & 97627 & -2 & -2 + 338 & 2994.05 & 1447.54 & 20.62 & - & 2 & 1 + 352 & 869.48 & 1493.94 & 20.70 & 45442 & 3 & 3 + 362 & 715.38 & 1525.08 & 20.21 & 99597 & 1 & 1 + 368 & 2960.35 & 1547.05 & 19.92 & 99513 & 0 & 0 + 382 & 697.67 & 1576.68 & 20.71 & 149212 & 6 & 5 + 385 & 1150.11 & 1586.52 & 18.25 & 98976 & -4 & -5 + 392 & 3917.25 & 1608.25 & 18.60 & 98541 & 3 & 3 + 393 & 2331.21 & 1609.32 & 20.61 & 99237 & -2 & -2 + 397 & 2423.80 & 1617.24 & 19.02 & 100430 & -2 & -2 + 399 & 765.00 & 1624.56 & 20.76 & 159692 & 0 & -2 + 415 & 2527.55 & 1680.94 & 19.65 & 98682 & -4 & -5 + 433 & 2519.00 & 1709.59 & 21.00 & - & 6 & 4 + 434 & 2712.22 & 1710.19 & 18.44 & 97642 & -5 & -5 + 435 & 1048.97 & 1710.26 & 20.02 & 190882 & 3 & 2 + 442 & 2657.45 & 1722.45 & 20.88 & 98607 & 1 & -2 + 452 & 3575.83 & 1753.56 & 19.99 & 98086 & 1 & 0 + 457 & 2202.26 & 1758.15 & 19.29 & 98688 & -2 & -3 + 468 & 2524.04 & 1780.65 & 20.30 & 170632 & 2 & 1 + 476 & 2164.48 & 1808.28 & 18.87 & 96773 & -4 & -5 + 484 & 2281.67 & 1835.66 & 17.57 & 97168 & -5 & -5 + 496 & 4657.77 & 1869.51 & 20.93 & - & -4 & -5 + 499 & 926.31 & 1877.15 & 19.85 & 160364 & -5 & -5 + 513 & 3385.76 & 1937.13 & 19.49 & 99860 & 0 & -2 + 519 & 2633.79 & 1953.48 & 17.76 & 98691 & -4 & -5 + 520 & 2267.84 & 1953.77 & 19.98 & 98961 & -5 & -5 + 521 & 2598.87 & 1954.62 & 20.06 & 99504 & -2 & -2 + 526 & 1819.56 & 1976.82 & 20.91 & - & -4 & -5 + 529 & 2019.97 & 1981.74 & 20.50 & 118028 & 15 & -3 + 530 & 2604.02 & 1985.11 & 19.63 & 99198 & -2 & -2 + 533 & 3492.41 & 1992.67 & 19.56 & 97120 & -5 & -5 + 541 & 2558.48 & 2016.50 & 20.55 & - & 0 & 1 + 545 & 1341.00 & 2022.51 & 20.10 & 149243 & 0 & -4 + 546 & 4213.68 & 2023.37 & 17.26 & 23768 & 5 & 4 + 550 & 2819.78 & 2034.77 & 20.70 & 100970 & 1 & 0 + 558 & 2046.76 & 2042.20 & 19.76 & 95753 & -2 & -2 + 559 & 4563.94 & 2045.51 & 20.58 & - & 4 & 4 + 561 & 4130.90 & 2050.21 & 19.46 & 207072 & 15 & 99 + 562 & 2886.03 & 2061.80 & 18.43 & 69161 & 4 & 3 + 563 & 2957.63 & 2065.25 & 20.21 & 100760 & 6 & 0 + 564 & 3187.62 & 2070.07 & 19.06 & 97222 & 0 & -2 + 570 & 3067.16 & 2074.98 & 20.66 & 99522 & -2 & -5 + 581 & 2750.58 & 2107.99 & 19.12 & 98421 & -2 & -2 + 591 & 2604.10 & 2133.40 & 18.36 & 97762 &
-2 & -2 + 602 & 2350.33 & 2175.38 & 20.38 & 94464 & -5 & -5 + 606 & 2651.60 & 2182.44 & 19.05 & 96413 & -2 & -4 + 611 & 1909.12 & 2186.08 & 18.89 & 97612 & -5 & -5 + 623 & 4308.50 & 2227.50 & 19.93 & 99776 & 3 & 2 + 624 & 2123.91 & 2231.85 & 19.46 & 52634 & 0 & -3 + 629 & 5040.18 & 2242.21 & 17.88 & 76299 & 4 & 1 + 636 & 2538.86 & 2258.28 & 18.75 & 97342 & -4 & -5 + 639 & 2777.26 & 2264.35 & 20.40 & 98032 & -2 & -2 + 641 & 4305.25 & 2279.18 & 20.95 & 23920 & 6 & 4 + 643 & 2833.92 & 2289.38 & 20.50 & 148539 & -2 & -2 + 647 & 2897.12 & 2311.32 & 20.21 & - & 6 & 5 + 651 & 3624.13 & 2320.14 & 21.00 & 100652 & -2 & -1 + 653 & 3096.27 & 2325.88 & 19.87 & 100251 & -2 & -4 + 655 & 4925.76 & 2334.26 & 20.04 & 98116 & -2 & -2 + 656 & 3032.45 & 2336.55 & 20.17 & 99708 & -2 & -2 + 662 & 2762.21 & 2350.96 & 19.15 & 95205 & 1 & 1 + 663 & 2002.63 & 2352.95 & 20.24 & 83425 & 0 & -5 + 670 & 4792.86 & 2365.01 & 19.75 & 98125 &
-2 & -2 + 679 & 2891.70 & 2395.08 & 19.51 & 96653 & 0 & -5 + 680 & 4456.98 & 2396.53 & 20.87 & 124924 & 2 & 3 + 685 & 4144.73 & 2406.59 & 20.20 & - & -4 & -5 + 698 & 2673.91 & 2444.21 & 20.95 & - & 3 & 4 + 707 & 3017.95 & 2463.86 & 20.54 & - & 0 & -2 + 711 & 2158.68 & 2471.78 & 19.83 & 96536 & -2 & -5 + 712 & 2420.36 & 2476.45 & 19.69 & 99920 & 0 & 1 + 713 & 2213.24 & 2477.51 & 20.55 & - & 5 & 4 + 717 & 2719.31 & 2486.68 & 18.78 & 96452 & -2 & -2 + 719 & 2172.06 & 2489.81 & 20.78 & 99339 & -5 & -5 + 724 & 2281.36 & 2494.96 & 18.85 & 98302 & -5 & -5 + 730 & 1712.08 & 2511.63 & 19.03 & 99830 & 3 & 1 + 732 & 2705.21 & 2518.39 & 20.23 & 96773 & -2 & -2 + 736 & 3434.65 & 2523.63 & 20.51 & 98451 & 0 & 0 + 743 & 2741.21 & 2541.98 & 20.06 & 98272 & -5 & -5 + 745 & 2878.57 & 2542.91 & 17.95 & 97033 & -4 & -5 + 748 & 2835.90 & 2554.58 & 20.25 & 97048 & -5 & -5 + 767 & 2784.63 & 2608.14 & 19.91 & 98302 & -5 & -5 + 768 & 2566.24 & 2608.94 & 18.44 & 101929 & 1 & 1 + 771 & 2851.49 & 2618.35 & 19.79 & 97462 & -5 & -2 + 774 & 3465.80 & 2625.73 & 18.81 & 97342 & -2 & -5 + 775 & 2818.67 & 2625.80 & 19.50 & 96824 & -5 & -5 + 778 & 3368.31 & 2635.15 & 20.05 & 99351 & 1 & 1 + 793 & 4179.71 & 2668.46 & 19.19 & 97036 & -4 & -2 + 794 & 2810.14 & 2668.81 & 19.23 & 98514 & 2 & 1 + 799 & 4028.08 & 2680.53 & 19.29 & 96161 & 0 & -4 + 801 & 1766.29 & 2685.75 & 18.63 & 96038 & 1 & 2 + 803 & 3671.86 & 2687.29 & 18.34 & 99702 & 1 & 0 + 810 & 2801.93 & 2719.27 & 17.11 & 98182 & -5 & -5 + 813 & 3341.70 & 2731.51 & 19.51 & 96383 & -2 & -2 + 822 & 4639.03 & 2748.37 & 19.05 & 96997 & 5 & 6 + 828 & 2521.07 & 2770.37 & 19.91 & 97912 & 6 & 6 + 831 & 2801.07 & 2781.44 & 18.74 & 97672 & -4 & -5 + 838 & 2248.16 & 2789.85 & 20.14 & - & 999 & -5 + 842 & 2906.30 & 2794.44 & 20.39 & 100670 & 999 & 0 + 843 & 4713.68 & 2796.95 & 20.55 & 48385 & 7 & 7 + 846 & 2774.56 & 2803.87 & 19.72 & 96563 & -5 & -5 + 867 & 2612.37 & 2853.73 & 18.37 & 97042 & -4 & -2 + 872 & 3640.91 & 2868.48 & 20.90 & - & -2 & -1 + 877 & 1235.86 & 2875.59 & 20.67 & 121526 & 1 & 0 + 886 & 2138.47 & 2909.08 & 20.20 & 95723 & 8 & 9 + 887 & 1185.86 & 2912.30 & 20.47 & - & 999 & 999 + 888 & 1978.03 & 2915.65 & 19.48 & 96563 & 0 & -2 + 889 & 3904.51 & 2916.67 & 20.36 & 97801 & 0 & 1 + 890 & 3514.41 & 2918.22 & 20.68 & 97564 & -4 & -2 + 898 & 3185.08 & 2942.27 & 20.62 & - & -5 & -1 + 900 & 3332.43 & 2943.17 & 20.63 & 97732 & -2 & -2 + 905 & 1591.54 & 2966.15 & 20.91 & 98451 & -5 & -5 + 909 & 3911.90 & 2973.01 & 20.31 & 187822 & 1 & 1 + 912 & 1895.44 & 2979.65 & 20.09 & 98751 & -5 & -3 + 913 & 2748.21 & 2980.39 & 19.26 & 95816 & -2 & -2 + 922 & 3294.59 & 3021.22 & 19.65 & 98931 & -2 & -2 + 927 & 2019.29 & 3026.93 & 19.25 & 95274 & -5 & -5 + 929 & 3951.51 & 3033.18 & 20.10 & 96263 & 0 & -1 + 933 & 4515.25 & 3035.84 & 19.87 & - & -4 & -5 + 938 & 1691.42 & 3049.42 & 19.25 & 71440 & -4 & 0 + 942 & 2829.86 & 3058.61 & 20.59 & 96428 & -2 & -5 + 943 & 3816.51 & 3058.75 & 19.94 & - & -2 & -2 + 944 & 2420.93 & 3065.24 & 20.94 & - & -5 & -5 + 948 & 4108.78 & 3075.68 & 20.49 & - & -5 & -4 + 961 & 2478.92 & 3095.09 & 19.79 & 99381 & -5 & -5 + 967 & 3606.49 & 3104.33 & 20.10 & - & 4 & 1 + 971 & 3131.00 & 3106.51 & 19.94 & 47565 & 5 & 3 + 975 & 3024.81 & 3120.83 & 20.95 & - & 99 & 99 + 978 & 2832.26 & 3123.45 & 20.24 & - & -4 & -5 + 993 & 3323.02 & 3154.25 & 19.05 & 97735 & 0 & -1 + 1003 & 2815.02 & 3176.78 & 20.76 & 100763 & -5 & -5 + 1012 & 2910.90 & 3204.76 & 20.77 & - & -2 & -4 + 1019 & 1598.82 & 3220.61 & 20.02 & 98718 & 0 & 1 + 1020 & 2661.84 & 3222.75 & 18.46 & 83578 & 3 & 1 + 1027 & 2954.26 & 3244.74 & 20.28 & 96686 & 0 & -2 + 1036 & 4346.84 & 3277.98 & 19.22 & 99165 & -2 & -2 + 1041 & 4453.92 & 3289.61 & 19.78 & 98649 & -2 & -2 + 1042 & 3771.63 & 3293.33 & 20.93 & - & -2 & 1 + 1045 & 4485.09 & 3295.86 & 19.95 & 44686 & 99 & 99 + 1049 & 2469.20 & 3302.45 & 20.58 & - & 0 & 0 + 1062 & 3268.23 & 3338.69 & 18.84 & 98182 & 0 & -1 + 1064 & 4749.89 & 3342.41 & 19.45 & 96182 & 5 & 2 + 1069 & 1366.71 & 3359.61 & 19.95 & 99060 & 0 & -2 + 1075 & 4869.37 & 3367.38 & 19.90 & 62244 & 1 & -2 + 1080 & 2041.26 & 3397.59 & 20.86 & 98421 & -5 & -5 + 1091 & 3315.98 & 3426.51 & 20.51 & 99579 & -2 & 0 + 1100 & 4416.75 & 3462.54 & 20.73 & 159095 & 15 & 99 + 1105 & 2881.63 & 3475.46 & 18.79 & 98323 & -2 & -2 + 1108 & 2836.73 & 3485.88 & 19.42 & 99081 & 1 & -1 + 1121 & 1960.44 & 3512.66 & 20.68 & - & -4 & -5 + 1127 & 2942.66 & 3523.15 & 20.76 & 159362 & 5 & 3 + 1141 & 4517.75 & 3545.75 & 19.93 & 99657 & -2 & -2 + 1146 & 4117.26 & 3561.40 & 18.27 & 100080 & -5 & -5 + 1153 & 2566.37 & 3583.80 & 20.14 & 61935 & 9 & 99 + 1157 & 2487.08 & 3591.79 & 18.64 & 62207 & 9 & 5 + 1158 & 3279.02 & 3592.19 & 20.16 & 99935 & -2 & -2 + 1161 & 2923.10 & 3597.44 & 20.01 & - & -5 & -5 + 1163 & 1348.04 & 3601.03 & 20.12 & 144364 & 1 & 3 + 1177 & 2091.53 & 3633.12 & 20.39 & 216269 & 15 & 1 + 1178 & 2361.82 & 3635.18 & 20.30 & 99234 & -2 & -1 + 1192 & 3212.40 & 3665.38 & 20.39 & 97375 & -5 & -5 + 1200 & 2024.55 & 3679.35 & 20.59 & - & 5 & 7 + 1207 & 3978.09 & 3693.67 & 20.68 & 100221 & -4 & -2 + 1209 & 3395.58 & 3703.43 & 19.10 & 83882 & 1 & 1 + 1211 & 2923.22 & 3706.89 & 20.79 & 99177 & -4 & -5 + 1214 & 3909.03 & 3714.53 & 19.51 & 100323 & 1 & 1 + 1223 & 2310.92 & 3752.63 & 20.49 & - & -5 & -5 + 1231 & 3481.26 & 3785.53 & 20.32 & 99018 & -2 & 0 + 1235 & 2884.78 & 3791.48 & 20.67 & - & -4 & -2 + 1239 & 2759.52 & 3812.84 & 20.00 & 62264 & 3 & 2 + 1240 & 2878.69 & 3813.83 & 20.24 & 122164 & 4 & 7 + 1253 & 995.43 & 3851.25 & 19.37 & 97093 & 2 & 4 + 1255 & 2908.32 & 3863.10 & 20.54 & - & 0 & 3 + 1258 & 2180.86 & 3869.70 & 20.60 & - & -5 & -5 + 1265 & 3358.24 & 3888.19 & 19.60 & 99896 & -2 & -2 + 1272 & 3738.88 & 3907.25 & 20.85 & 99501 & -2 & -1 + 1273 & 2182.60 & 3907.71 & 20.83 & - & -5 & 0 + 1275 & 2223.31 & 3912.98 & 20.08 & 99141 & -2 & -1 + 1285 & 1476.03 & 3937.66 & 19.84 & 188687 & 1 & 1 + 1287 & 2727.83 & 3947.24 & 20.20 & - & -2 & -2 + 1297 & 2476.43 & 3965.12 & 20.88 & 97102 & -2 & -2 + 1298 & 2291.46 & 3965.96 & 20.79 & - & -2 & -2 + 1299 & 1553.40 & 3967.72 & 20.25 & 98964 & 3 & 3 + 1301 & 1428.04 & 3978.07 & 19.40 & 99342 & 0 & -2 + 1303 & 3180.35 & 3982.62 & 18.98 & 96901 & 0 & -2 + 1308 & 1799.52 & 3992.55 & 20.30 & - & 5 & 3 + 1322 & 1469.20 & 4032.65 & 20.90 & - & -2 & -2 + 1330 & 2626.63 & 4050.56 & 19.91 & 97051 & 15 & 10 + 1335 & 3007.32 & 4063.99 & 20.39 & - & -4 & -4 + 1343 & 2119.25 & 4082.59 & 17.77 & 99156 & -5 & -5 + 1357 & 1560.05 & 4129.23 & 19.78 & 97000 & -5 & -2 + 1367 & 2972.68 & 4167.09 & 18.02 & 98299 & -5 & -5 + 1370 & 3118.58 & 4169.24 & 19.71 & 97738 & -2 & -5 + 1372 & 3911.67 & 4178.76 & 20.50 & 98472 & -5 & -5 + 1374 & 3888.45 & 4183.67 & 20.52 & 97699 & -2 & -2 + 1376 & 3489.91 & 4187.61 & 19.91 & 98275 & -2 & -2 + 1377 & 2964.53 & 4192.76 & 19.23 & 97975 & -5 & -5 + 1379 & 3412.68 & 4193.30 & 20.93 & - & 0 & 2 + 1382 & 2988.76 & 4209.19 & 20.23 & - & 3 & 4 + 1387 & 2761.87 & 4222.14 & 19.18 & 99417 & -5 & -5 + 1395 & 848.83 & 4239.90 & 19.39 & 99771 & -2 & 1 + 1405 & 3807.69 & 4288.35 & 20.12 & 99285 & -4 & -5 + 1409 & 3951.25 & 4295.39 & 19.69 & 54964 & -2 & -2 + 1411 & 2500.75 & 4322.96 & 19.74 & - & 999 & -2 + 1413 & 3281.08 & 4328.24 & 19.37 & 99774 & 2 & 2 + 1414 & 1303.46 & 4329.27 & 20.65 & 95615 & 5 & 3 + 1415 & 288.53 & 4331.63 & 18.70 & 98844 & -5 & -5 + 1423 & 3633.52 & 4361.01 & 20.56 & 98799 & -2 & -2 + 1430 & 4193.88 & 4380.15 & 20.61 & 99222 & 999 & 3 + 1431 & 3125.60 & 4381.56 & 20.86 & - & -2 & -4 + 1438 & 574.34 & 4410.17 & 18.72 & 81105 & 2 & 1 + 1439 & 3020.74 & 4422.23 & 19.27 & 99339 & 0 & -5 + 1447 & 755.69 & 4452.72 & 20.97 & 171424 & 3 & 4 + 1454 & 3375.34 & 4469.90 & 20.28 & 100859 & -2 & -3 + 1461 & 4207.84 & 4489.60 & 20.98 & 97279 & -2 & -5 + 1466 & 3188.31 & 4501.44 & 20.62 & - & 5 & 5 + 1481 & 810.82 & 4540.51 & 20.61 & 99791 & -2 & 2 + 1483 & 825.24 & 4542.01 & 19.36 & 99791 & -5 & -5 + 1488 & 661.75 & 4553.49 & 19.55 & 80904 & -2 & -4 + 1524 & 943.52 & 4664.34 & 20.89 & 190098 & 6 & 5 + 1529 & 3450.04 & 4681.31 & 20.81 & 114632 & 15 & 10 + 1540 & 3674.58 & 4731.74 & 18.01 & 21317 & 5 & 6 + 1541 & 2265.25 & 4732.11 & 20.43 & 98778 & -5 & -5 + 1559 & 3242.72 & 4783.73 & 20.72 & 100380 & 1 & 1 + 1563 & 4731.77 & 4795.22 & 20.88 & - & -2 & -1 + 1567 & 768.30 & 4815.43 & 20.88 & 100284 & 99 & 10 + 1568 & 4014.76 & 4819.25 & 19.46 & 36222 & 99 & -5 + 1571 & 2338.11 & 4825.66 & 19.97 & 83803 & 4 & 5 + 1588 & 4185.04 & 4870.66 & 20.95 & 126167 & 0 & -4 + 1594 & 1852.07 & 4887.76 & 19.87 & - & 999 & 999 + 1621 & 1467.74 & 4980.41 & 20.03 & - & -5 & 999 + 1623 & 1937.90 & 4990.12 & 20.86 & 187086 & -2 & -2 + 1634 & 3158.57 & 5037.47 & 19.28 & 31985 & 9 & 5 + 1664 & 2256.63 & 5240.59 & 20.61 & 248908 & 5 & 2 + 1666 & 2243.04 & 5270.84 & 19.15 & 99552 & 1 & 2 + | we describe the morphological composition of a sample of 518 galaxies in the field of cl1358 + 62 at z=0.33 , drawn from a large _ hst _ mosaic covering 53 sq .
arcmin .
the sample is complete to @xmath0=22 , corresponding to @xmath1=-18.5 in the rest frame .
the galaxy morphologies have been independently classified by the authors of this paper and by alan dressler .
dressler s classifications place our work in context with the previous morphs study , and allow us to estimate the scatter between different sets of visual classifications .
we restrict most of our analysis to the brighter part of the sample , @xmath2 ( @xmath3 ) , where the scatter between the two sets of classifications is @xmath41 in morphological type .
the scatter doubles at @xmath5 , presumably due to the lower signal - to - noise and poorer sampling of faint , small galaxy images . to @xmath0=21 the two sets of classifiers agree on the fraction of early type galaxies ( elliptical+s0 ) : 72% .
we conclude that cl1358 + 62 does not contain the large population of spiral galaxies found in other studies of clusters at @xmath6 , and that there is probably a significant spread in the degree of cluster evolution at intermediate redshift .
the two groups of classifiers differ on the relative fraction of s0 and elliptical galaxies .
we show that the distributions of ellipticities and bulge / total light can not resolve this discrepancy .
nonetheless , we can derive significant constraints on physical models for the evolution of the galaxy population in cl1358 + 62 .
the higher ratio of s0 to elliptical galaxies ( 1.6 ) found by df / mf / pvd requires that the evolution preserve the relative fraction of elliptical , s0 and spiral galaxies .
alternately , the lower ratio ( 1.1 ) found by ad requires that the evolution preserve the early - type to spiral ratio while increasing the s0 to elliptical ratio . in the latter case , a possible evolutionary mechanism is accretion of galaxies that predominantly evolve to s0 s between @xmath7=0.33 and the present .
we use our large body of spectra to make the correspondence between spectral and morphological type .
our data follow the pattern seen in the field at low redshift : emission line spectra are more prevalent among the later morphological types .
the 11 identified k+a galaxies ( absorption line spectra with strong balmer lines ) have s0sb morphologies . _
accepted for publication in the astrophysical journal _ |
a _ robot swarm _ @xcite is a system of multiple autonomous mobile robots engaged in some collective task . in hostile environments , it may be desirable to employ large groups of low cost robots to perform various tasks cooperatively .
this approach is more resilient to malfunction and more configurable than a single high cost robot . _
swarm robotics _ , pioneered by c. w. reynolds @xcite , is a novel approach to coordinate a large number of robots .
the idea is inspired by the observation of social insects .
they are known to coordinate their actions to execute a task that is beyond the capability of a unit .
the field of swarm robotics has been enriched by many researchers adopting different approaches for swarm aggregation , navigation , coordination and control .
mobile robots can move in the physical world and interact with each other .
geometric problems are inherent to such multiple cooperative mobile robot systems and have been well studied @xcite .
multiple robot path planning , moving to ( and maintaining ) formation and pattern generation are some important geometric problems in swarm robotics
. multiple robot path planning may deal with problems like finding non - intersecting paths for mobile robots @xcite .
the formation and marching problems require multiple robots to form up and move in a specified pattern .
these problems are also interesting in terms of distributed algorithms .
formation and marching problems may act as useful primitives for larger tasks , like , moving a large object by a group of robots @xcite or distributed sensing @xcite .
pattern generation in cellular robotic systems ( crs ) @xcite is related to pattern formation problem by mobile robots .
this paper addresses a very well known and challenging problem involving robot swarms , namely _ gathering_. the objective is to collect multiple autonomous mobile robots into a point or a small region .
the choice of the point is not fixed in advance .
initially the robots are stationary and in arbitrary positions .
gathering problem is also referred to as _ point formation _ , _ convergence _ , _ homing _ or _ rendezvous _ @xcite .
a pragmatic view of swarm robots asks for distributed environment .
several interesting works have been carried out by researchers @xcite on distributed algorithms for mobile robots . a simple basic model called _ weak model _
@xcite is popular in the literature .
the world of the robots consists of the infinite plane and multiple robots living on it .
the robots were considered dimensionless or points .
all robots are autonomous , homogeneous and perform the same algorithm ( _ look - compute - move _ cycle ) @xcite . in _
look _ state , a robot takes a snapshot of its surroundings , within its range of vision .
it then executes an algorithm for computing the destination in _ compute _ state .
the algorithm is same for all robots . in _ move _ state , the robot moves to the computed destination .
the robots are oblivious ( memoryless ) .
they do not preserve any data computed in the previous cycles .
there is no explicit communication between the robots .
the robots coordinate by means of observing the positions of the other robots on the plane .
a robot is always able to see another robot within its visibility radius or range ( may be infinite ) .
two different models have been used for robots movement @xcite . under the sym model ,
the movement of a robot was considered to be instantaneous , i.e. , when a robot is moving , other robots can not see it . later , that model has been modified to corda model where the movement of a robot is not instantaneous .
a robot in motion is visible .
the corda model is a better representation of the real world .
the robots may or may not be synchronized .
synchronous robots execute their cycles together .
in such a system , all robots get the same view . as a result
, they compute on same data . in the more practical asynchronous model
, there is no such guarantee . by the time a robot completes its computation ,
several of the robots may have moved from the positions based on which the computation is made .
the robots may have a common coordinate system or individual coordinate systems having no common orientation or scale @xcite .
the problem of gathering multiple robots has been studied on the basic model of robotic system with different additional assumptions .
prencipe @xcite observed that gathering is not always possible for asynchronous robots .
however , instead of meeting at a single point , if the robots are asked to move very close to each other , then the problem can be solved by computing the center of gravity of all robots and moving the robots towards it . under the weak model ,
robots do not have any memory .
as they work asynchronously and act independently , once a robot moves towards the center of gravity , the center of gravity also changes . moving towards center of gravity does not gather robots to a single point .
however , this method makes sure that the robots can be brought as close as one may wish .
a possible solution to this problem is to choose a point which , unlike center of gravity , is invariant with respect to robots movement .
one such point in the plane is the point which minimizes the sum of distance between itself and all the robots .
this point is called the _ weber or fermat or torricelli _ point .
it does not change during the robots movement , if the robots move only towards this point .
however , the _ weber _ point is not computable for higher number of robots ( greater than 4 ) @xcite .
thus , this approach also can not be used to solve the gathering problem . if common coordinate system is assumed , instead of individual coordinate systems , then the gathering problem is solvable even with limited visibility @xcite . if the robots are synchronous and their movements are instantaneous , then the gathering problem is solvable even with limited visibility @xcite .
cieliebak et al .
@xcite , show that there are solutions for gathering without synchronicity if the visibility is unlimited .
one approach is by _ multiplicity _ detection .
if multiple robots are at the same point then the point is said to have _ strict multiplicity_. however , the problem is not solvable for _ two _ robots even with _ multiplicity _ detection @xcite .
there are some solutions for _ three _ and _ four _ robots . for more than _ four _ robots , there are two algorithms with restricted sets of initial configurations @xcite . in the first algorithm ,
the robots are initially in a _ bi - angular configuration_. in such a configuration , there exists a point @xmath3 and two angles @xmath4 and @xmath5 such that the angle between any two adjacent robots is either @xmath4 or @xmath5 and the angles alternate .
the second algorithm works if the initial configuration of the robots do not form any regular @xmath0-gon .
prencipe @xcite reported that there exists no deterministic oblivious algorithm that solves the gathering problem in a finite number of cycles for a set of @xmath6 robots .
_ convergence _ of multiple robots is studied by peleg and cohen @xcite and they proposed a gravitational algorithm in fully asynchronous model for convergence of any number of robots .
a dimensionless robot is unrealistic .
czyzowicz et al.,@xcite extend the weak model by replacing the point robots by unit disc robots .
they called these robots as _
fat robots_. the methods of gathering of three and four fat robots have been described by them .
the paper considers partial visibility and presents several procedures to avoid different situations which cause obstacles to gathering .
we proposed a deterministic gathering algorithm for @xmath0 @xmath1 fat robots @xcite .
the robots are assumed to be transparent in order to achieve full visibility .
having _ fat robots _ , we first define the _ gathering pattern _ to be formed by the robots .
therefore , _ gathering pattern _ formation becomes a special case of _ pattern formation _ of mobile robots which is also a challenging problem for mobile robots .
recent works @xcite shows that pattern formation and leader election are equivalent for @xmath2 robots .
however , this work @xcite considered the robots to have common handedness or chirality .
we propose a gathering algorithm which assumes no chirality and use the leader election technique in order to form gathering pattern .
we also show that if leader election is possible then formation of the _ gathering pattern _ is possible even with no chirality .
section [ model ] describes the model used in this paper and presents an overview of the problem .
then we move to the solution approach .
section [ char ] characterizes the geometric configurations of the robots for gathering .
section [ algo ] presents the _ leader election _ and _ gathering _ algorithms .
finally section [ con ] summarizes the contributions of the paper and concludes .
we use the basic structure of _ weak model _ @xcite and add some extra features which extend the model . let @xmath7 be a set of fat robots .
a robot is represented by its center , i.e. , by @xmath8 we mean a robot whose center is @xmath8 .
the following assumptions describe the system of robots deployed on the 2d plane : * robots are autonomous .
* robots execute the cycle ( _ look - compute - wait - move _ ) asynchronously .
* robots are anonymous and homogeneous in the sense that they are not uniquely identifiable , neither with a unique identification number nor with some external distinctive mark ( e.g. , color , flag , etc . ) .
* a robot can sense all other robots irrespective of their configuration . * each robot is represented as a unit disc ( _ fat robots _ ) .
* corda @xcite model is assumed for robots movement . under this model
the movement of the robots is not instantaneous . while in motion
, a robot may be observed by other robots .
* robots have no common orientation or scale .
each robot uses its own local coordinate system ( origin , orientation and distance ) .
a robot has no particular knowledge about the coordinate system of any other robot , nor of a global coordinate system .
* robots can not communicate explicitly .
each robot has a camera or sensor which can take picture or sense over 360 degree .
the robots communicate only by means of observing other robots using the camera or sensor .
a robot can compute the coordinates ( w.r.t .
its own coordinate system ) of other robots by observing through the camera or sensor .
* robots have infinite visibility range .
* robots are assumed to be transparent to achieve full visibility .
* robots are oblivious .
they do not remember the data from the previous cycles .
* initially all robots are stationary .
let us define the _ gathering pattern _ which is to be formed by @xmath9 .
the desired gathering pattern for transparent fat robots is shown in fig .
one robot with center @xmath10 is at the center of the structure .
we call it layer @xmath11
. robots at layer @xmath12 are around layer @xmath11 touching @xmath13 ( @xmath14 is a circle with center at @xmath4 and radius @xmath0 ) .
robots at layer @xmath15 are around layer @xmath12 touching @xmath16 and so on . fig .
1(a ) shows a gathering pattern with @xmath17 layers .
[ gathering_pattern ] _ gathering pattern _ is a circular layered structure of robots with a unique robot centered at @xmath10 .
a robot at layer @xmath18 touches @xmath19 and at least @xmath12 and at most @xmath15 robots at layer @xmath18 .
the inner layers are full
. the outer most layer may not be full but the vacant places in the outer most layers must be contiguous . according to the definition , a unique center is required to form a gathering pattern and the layers are created around the center . finding a unique robot and marking it as the center is not possible for @xmath15 , @xmath17 and @xmath20 robots .
this is due to the symmetric structure ( fig .
2 ) . for 2 or 3 robots
all robots are potential candidates to be the center . for 4 robots
there are two such robots . to identify a unique robot for the center of the gathering pattern ,
at least @xmath21 robots are required . for @xmath21 robots , a robot which touches the rest @xmath20 robots
is treated as the central robot or a robot at layer @xmath11 .
1(b ) shows the desired gathering pattern with minimum robots .
we are given a set , @xmath9 , of separated , stationary transparent fat robots .
the objective is to move the robots so as to build the gathering pattern with the robots in @xmath9 . for gathering multiple robots ,
our approach is to select one robot and assign it a destination .
all other robots remain stationary till the designated robot reaches its destination . at each turn , the robot selected for motion is chosen in such a way , that during the motion if any other robot takes a snapshot and computes , this particular robot would be the only one eligible for movement . to implement this
, we choose a destination carefully .
the robot , nearest to the destination , is allowed to move first . however
, a problem will occur when multiple robots are at same distance from the destination .
a leader election mechanism is required to select a robot among such multiple eligible robots . to overcome this problem
, our approach is to order a set of robots on 2d plane , with respect to the destination .
the robots move one by one , according to this order , towards the destination .
the mutual ordering of the robots in the set is invariant , though the size of the set changes .
a set of robots on 2d plane is given .
our objective is to move them , one by one , to form desire _
gathering pattern_. in order to do so , we create an ordering among the robots and move one by one following the ordering .
this ordering will be computed at different robot sites , who have different coordinate systems , orientations and origins .
thus , we need the ordering algorithm to yield the same result , even if the point set is given a transformation with respect to origin , axes and scale . for this to succeed
, no two robots should have the same view . keeping this in mind
, we proceed to characterize the geometric configurations that is required for a set of robots to be orderable ( formal definition is presented shortly ) . in this section
we represent the robots by points .
let @xmath22 be a non - empty set of points on the 2d plane .
@xmath23 is a line on the plane .
let @xmath24 be the points of @xmath25 ( not all ) , which lie on @xmath23 .
@xmath23 partitions @xmath26 into two subsets @xmath27 and @xmath28 where @xmath27 or @xmath28 is non - empty .
let @xmath29 be the straight line that intersects @xmath23 at point @xmath30 such that @xmath31 and @xmath30 is the middle point of the span of the points on @xmath32 .
a point @xmath33 is said to be a straight mirror image of @xmath34 across @xmath35 , if @xmath36 is a mirror image of @xmath37 across @xmath23 ( figure [ mirrorimage](a ) ) .
[ stmirrorimage ] let @xmath36 be the straight mirror image of @xmath37 across @xmath23 .
@xmath38 is the straight mirror image of @xmath36 across @xmath29 .
@xmath38 is said to be skewed mirror image of @xmath37 across @xmath23 ( figure [ mirrorimage](b ) ) .
[ skmirrorimage ] a set of points @xmath25 on the 2d plane is said to be in straight - symmetric configuration , if there exists a straight line @xmath23 ( on that plane ) not containing all the points of @xmath25 , such that each point in @xmath39 has a straight mirror image in @xmath40 ( figure .
[ symconfig](a ) ) .
the line @xmath23 is called a line of straight symmetry .
[ def - straight - symmetric - conf ] note that each point in @xmath32 is the mirror image of itself .
a set of points @xmath25 on the 2d plane is said to be in skew - symmetric configuration , if there exists a straight line @xmath23 ( on that plane ) not containing all the points of @xmath25 , such that each point in @xmath39 has a skewed mirror image in @xmath40 .
[ symconfig](b ) ) .
the line @xmath23 is called a line of skew symmetry .
[ def - skewed - symmetric - conf ] [ def - symmetric - conf ] a set of points @xmath25 on the 2d plane is said to be in symmetric configuration ( denoted by @xmath41 ) , if it is either a singleton set or in straight symmetric or skew symmetric configuration . for a singleton set , any line passing through the point is a line of symmetry .
[ def - asymmetric - conf ] a set of points @xmath25 , which is not in symmetric configuration is in asymmetric configuration ( denoted by @xmath42 ) .
our requirement does not stop at requiring the algorithm to be robust to changes in the coordinate system .
the positions of the robots also change as the algorithm progresses .
our objective is to order the points ( robots ) in @xmath25 such that when the robots move one by one , according to this order , towards the destination , the mutual ordering of the robots in the set is invariant .
we define an orderable set as follows .
[ def - orderable - set ] a set of points @xmath25 , on the plane , is called an orderable set , if there exists a deterministic algorithm , which produces a unique ordering of the points of @xmath25 , such that the ordering is same irrespective of the choice of origin and coordinate system .
[ lemma - sym - ord ] let @xmath25 be a non - empty , non - singleton set of points .
if @xmath25 is in @xmath43 , then @xmath25 is not orderable .
_ proof : _ let @xmath23 be a line of symmetry ( straight or skewed ) for @xmath25 .
let @xmath44 be the set of points from @xmath25 , lying on @xmath23 .
@xmath23 divides @xmath45 into two halves @xmath27 and @xmath28 .
@xmath46 and @xmath47 are mirror images ( straight or skewed ) of each other .
let @xmath48 be a point in @xmath27 and @xmath49 in @xmath28 , the mirror image of @xmath48 .
consider an arbitrary ordering algorithm @xmath50 .
if we run @xmath50 on @xmath25 with @xmath48 as the origin , it produces an ordering of @xmath25 .
let @xmath51 be the first point from that ordering such that @xmath51 is not on @xmath32 .
on the other hand , if we run @xmath50 on @xmath25 , with @xmath49 as the origin , the symmetry tells us that the ordering obtained will have @xmath52 ( the mirror image of @xmath51 ) as the corresponding first point in the order . since , @xmath51 is not in @xmath32 , @xmath53 .
since the choice of @xmath50 was arbitrary , no algorithm will produce the same order irrespective of the choice of origin .
hence , @xmath25 is not orderable .
[ lemma - los - passes - centersecp ] any line of symmetry of @xmath25 passes through the center of the smallest enclosing circle ( sec ) of @xmath25 and divides the points on sec into two equal mirror images ( straight or skewed ) . in order to check whether a set of points is in @xmath43 or
not , we need to find out if a line of symmetry exists for this set . for this search to be feasible ,
we need to reduce the potential set of candidate lines for line of symmetry . in order to do so ,
first the sec of @xmath25 is computed .
the points on the sec are taken to form a convex polygon , say @xmath54 .
a line of symmetry ( straight or skewed ) of @xmath25 cuts @xmath54 at two points .
thus the line of symmetry contains at most two points from @xmath54 .
[ hull - vertex - symmetry ] let @xmath25 be a set of points in @xmath43 .
the mirror image ( straight or skewed ) of a vertex in @xmath55 is also in @xmath54 . _
proof : _ consider a vertex @xmath37 of @xmath54 .
let @xmath36 be the mirror image of @xmath37 across the line of symmetry @xmath23 .
suppose , @xmath36 is not in @xmath56 .
suppose , the line @xmath57 intersects @xmath58 at @xmath30 and intersects @xmath54 at @xmath59 .
@xmath60 . and
@xmath5 are two points on the 2d plane then the distance between @xmath4 and @xmath5 is represented by @xmath61 @xmath59 should also have a mirror image across @xmath23 , in the same side where @xmath37 lies .
let @xmath62 be the mirror image of @xmath59 .
note that @xmath63 also intersects @xmath23 at point @xmath30 and @xmath64 . if @xmath36 is inside the convex hull then @xmath65 ( figure .
[ hullimage](a ) ) . hence , @xmath66 and @xmath67 .
this implies that @xmath37 is not a hull vertex .
contradiction ! on the other hand , if @xmath36 is outside the convex hull then @xmath68 ( figure .
[ hullimage](b ) ) . hence , @xmath69 and @xmath70 .
this implies that @xmath37 is not a hull vertex .
contradiction ! therefore ,
if @xmath37 is a hull vertex , @xmath36 must also be a hull vertex .
if @xmath25 is in @xmath43 , then @xmath54 is in @xmath43 .
_ proof : _ follows from lemma [ hull - vertex - symmetry ] . [ lemmmastraightsymhull ] if @xmath54 is in @xmath41 , then for any line of straight symmetry @xmath23 of @xmath25 , 1 .
if @xmath23 passes through a vertex @xmath71 of @xmath54 , it bisects the interior angle at @xmath71 , 2 .
if @xmath23 intersects an edge @xmath72 of @xmath54 , at a point other than a vertex , it is the perpendicular bisector of @xmath72 .
_ proof : _ follows from the proof of lemma [ hull - vertex - symmetry ] . [ losks - intersect-1-vertex - edge ] any line of skew symmetry intersects @xmath54 either at two vertices or at two edges . [ losks - intersect-2-vertex - edge ] a skew symmetric polygon has even number of vertices and edges .
[ lem - rectangle ] a pair of edges in a polygon inscribed in a circle is parallel and equal if and only if they are opposite sides of a unique rectangle inscribed in that circle .
_ proof : _ * if : * trivial .
* only if : * suppose @xmath73 and @xmath74 are two edges of a polygon such that @xmath75 and @xmath76 ( fig . [ skew1 ] ) .
@xmath10 is the intersection point of the lines @xmath77 and @xmath78 .
it is easy to see that @xmath79 .
so , @xmath80 and @xmath81 .
this means that the cords , @xmath77 and @xmath78 bisect each other .
hence , @xmath77 and @xmath78 are both diameters of the circle .
therefore , @xmath82 .
this implies that @xmath83 is a rectangle .
[ obs - parallel - edge - equal ] a polygon @xmath54 , inscribed in a circle , is skew symmetric if and only if each edge of the polygon has a parallel edge of equal length . _
proof : _ * if : * suppose @xmath73 and @xmath84 , are parallel and equal edges of @xmath54 ( fig [ skew ] ) .
let us add @xmath85 and @xmath86 by a straight line @xmath23 .
we shall show that @xmath23 is a line of skew symmetry for @xmath54 .
@xmath87 is the skewed mirror image of @xmath88 across @xmath23 .
let @xmath89 be the adjacent edge of @xmath88 .
we add @xmath90 . since , @xmath78 is a diameter of the circumcircle of @xmath54 ( lemma [ lem - rectangle ] ) , @xmath91 degree .
we draw the rectangle @xmath92 . by lemma [ lem - rectangle ]
, @xmath93 is the edge of the polygon which is parallel to and equal in length with @xmath89 . by repeating this argument
it can be shown that , polygonal chains on both sides of @xmath23 are skew symmetric .
* only if : * let @xmath54 be a skew symmetric polygon . @xmath23 is a line of skew symmetry for @xmath54 .
@xmath23 partitions @xmath54 into two halves namely , @xmath94 and @xmath95 .
first consider the case when @xmath23 passes through two vertices of @xmath54 , namely @xmath96 and @xmath97 .
( fig [ skewsymhfig ] ) .
suppose , @xmath98 is the edge incident at @xmath96 in @xmath94 and @xmath99 is the edge incident at @xmath100 in @xmath95 .
since , @xmath23 is a line of skew symmetry , @xmath99 is the skewed mirror image of @xmath98 .
therefore , @xmath101 is denoted by @xmath102 and @xmath103 .
let @xmath104 be the edge adjacent to @xmath98 in @xmath94 and @xmath105 the edge adjacent to @xmath99 in @xmath95 .
similarly , @xmath105 is the mirror image of @xmath104 .
therefore , @xmath106 and @xmath107 . in this manner
, we can find a parallel and equal edge of every edge . now consider the case when @xmath23 intersects two edges of @xmath54 , namely @xmath108 and @xmath109 at point @xmath110 and @xmath111 respectively ( fig [ skewsymhfig ] ) .
if we consider a modified polygon with additional vertices at @xmath110 and @xmath111 , the result follows from the previous case .
suppose , @xmath54 is a skew symmetric polygon .
let @xmath23 be a line intersecting two vertices of @xmath54 , namely @xmath96 and @xmath97 .
let @xmath112 be the interior angle of @xmath54 at vertex @xmath96 and @xmath113 be the interior angle of @xmath54 at vertex @xmath100 .
@xmath23 divides @xmath112 into @xmath114 and @xmath115 and @xmath113 into @xmath116 and @xmath117 ( fig [ skewsymhfig ] ) .
suppose , @xmath98 is an edge incident at @xmath96 in @xmath94 and @xmath99 is an edge incident at @xmath100 in @xmath95 .
@xmath104 is the edge adjacent to @xmath98 in @xmath94 and @xmath105 is the edge adjacent to @xmath99 in @xmath95 .
[ losks - vertex ] for a skew symmetric polygon @xmath54 , @xmath23 is a line of skew symmetry if and only if @xmath118 or @xmath119 .
_ proof : _ * if : * if @xmath118 , @xmath103 . from lemma [ obs - parallel - edge - equal ] ,
there exist an edge @xmath120 such that @xmath121 and @xmath122 . as @xmath54 is convex @xmath121 and @xmath103 implies @xmath120 and @xmath99 are the same .
hence , @xmath101 . using an argument similar to that used in the proof of lemma [ lem - rectangle ]
, it can be shown that , polygonal chains on both sides of @xmath23 are skew symmetric .
similarly , if @xmath119 , polygonal chains on both sides of @xmath23 are skew symmetric .
hence , @xmath23 is a line of skew symmetry . * only if : * follows from the definition of skew symmetric polygon .
let @xmath23 be a line intersecting two edges of @xmath54 , namely @xmath108 and @xmath109 at point @xmath110 and @xmath111 respectively ( fig [ skewsymhfig ] ) .
[ losks - edge ] for a skew symmetric polygon @xmath54 , @xmath23 is a line of skew symmetry if and only if @xmath98 @xmath123 @xmath99 , @xmath124 and @xmath125 . _
proof : _ * if : * @xmath98 @xmath123 @xmath99 , @xmath124 and @xmath125 implies that @xmath126 is the skewed mirror image of @xmath127 and @xmath128 is the skewed mirror image of @xmath129 across @xmath23 . as @xmath101 , using an argument similar to that used in the proof of lemma [ lem - rectangle ] , it can be shown that , polygonal chains on both sides of @xmath23 are skew symmetric .
hence , @xmath23 is a line of skew symmetry . * only if : * follows from the definition of skew symmetric polygon . in order to check whether a set of points @xmath25 is in @xmath43 or
not , we first compute the sec of @xmath25 .
the convex polygon @xmath54 , as described earlier , is also computed . for each vertex and each edge of @xmath54
, we look for a line of symmetry @xmath23 ( straight or skewed ) passing through that vertex or edge .
since , we want the ordering to be the same for any choice of origin and axes , we can only use information which are invariant under these transformations .
examples of such properties are , distances and angles between the points . the distances may be affected by the choice of unit distance , but even then their ratios remain the same .
one possible solution is to select the robot closest to the destination as the leader or the candidate to move .
it also satisfies our extra requirement that it remains the point closest the destination , and hence the leader , as it moves .
robots equidistant from the destination are on the circumference of a circle with the destination as its center .
they also form a convex polygon .
let @xmath130 be a set of robots forming such a convex polygon inscribed in a circle .
for the rest of this paper , by _ convex polygon _ we mean such polygons which are inscribed in a circle .
the center of the circle is called the center of the polygon .
[ def - straight - symmetric - poly ] a convex polygon @xmath130 is _ straight symmetric _ if the set of vertices in @xmath130 is in straight symmetric configuration .
[ def - skew - symmetric - poly ] a convex polygon @xmath130 is _ skew symmetric _ if the set of vertices in @xmath130 is in skew symmetric configuration .
[ def - symmetric - poly ] a convex polygon @xmath130 is symmetric if it is either straight symmetric or skew symmetric .
[ def - asymmetric - poly ] a polygon , which is not symmetric is asymmetric . note : a single point on a circle is a special case .
it is symmetric .
though , any line passing through it is a line of symmetry , we shall only call the line passing through the center of the circle ( and the point itself ) as the line of symmetry .
[ obs - asym - poly ] the set of vertices of an asymmetric polygon is in @xmath42 .
lem - skew - los ] any straight line passing through the center of a skew symmetric polygon , is a line of skew symmetry for that polygon . _
proof : _ let @xmath130 be a skew symmetric polygon .
let @xmath131 and @xmath132 be two parallel and equal length edges of @xmath130 ( figure . [ skew2 ] ) .
let @xmath133 and @xmath134 be the lines passing through @xmath135-@xmath136 and @xmath137-@xmath138 respectively . from lemma
[ lem - rectangle ] , it follows that @xmath133 and @xmath134 pass through the center @xmath10 of @xmath130 and they are lines of skew symmetry of @xmath130 .
now it is sufficient to prove that any line passing through the center and intersecting @xmath139 is a line of skew symmetry for @xmath130 .
without loss of generality let us rotate @xmath133 by some angle , around @xmath10 , keeping it between @xmath133 and @xmath134 .
let @xmath140 be the new position of the line .
following lemma [ losks - edge ] and using an argument similar to that used in the proof of lemma [ lem - rectangle ] , it can be shown that , polygonal chains on both sides of @xmath140 are skew symmetric .
hence , @xmath140 is a line of skew symmetry . for a convex polygon , a line of symmetry @xmath23 ( straight or skewed ) intersects the polygon at two points .
the points can be two vertices or they may lie on two edges or one point may be a vertex and the other point lies on an edge ( figure .
[ symmetric ] ) .
let @xmath35 intersect @xmath130 at @xmath30 and @xmath141 .
suppose @xmath130 has @xmath142 vertices .
the vertices are labeled starting from the vertex next ( clockwise ) to @xmath30 up to the previous vertex of @xmath30 , as @xmath143 . if @xmath30 is a vertex , then @xmath144 . if @xmath30 lies on an edge , @xmath145 .
similarly , the vertices are labeled starting from the vertex next ( clockwise ) to @xmath141 up to the previous vertex of @xmath141 , as @xmath146 . if @xmath141 is a vertex , then @xmath147 . if @xmath141 lies on an edge , @xmath148 .
the following notations are used in the rest of this paper .
* @xmath149 .
* @xmath150 .
* @xmath151 . * @xmath152 .
@xmath153 iff @xmath154 for ( @xmath155 ) .
[ lemma - straight - sym ] let @xmath130 be a straight symmetric polygon .
a straight line @xmath23 is a line of straight symmetry for @xmath130 if and only if @xmath153 or @xmath156 .
_ proof : _ follows from definition [ def - straight - symmetric - conf ] and definition [ def - straight - symmetric - poly ] .
[ lemma - skewed - sym ] let @xmath130 be a skew symmetric polygon and @xmath23 a line of skew symmetry .
@xmath23 divides @xmath130 into two skewed mirror image parts if and only if @xmath157 or @xmath158 .
_ proof : _ follows from definition [ def - skewed - symmetric - conf ] and definition [ def - skew - symmetric - poly ] . [ obs - asymmetric - polygon ] a polygon @xmath130 is asymmetric if and only if all of the following conditions are true 1 .
@xmath159 for @xmath155 .
2 . @xmath160 for @xmath161 and @xmath162 .
3 . @xmath163 for @xmath161 and @xmath162 note : @xmath160 and @xmath164 are equivalent .
theorem 1 can also be written as follows .
* theorem 1.1 * : a polygon @xmath130 is asymmetric if and only if all of the following conditions are true 1 .
@xmath159 for @xmath155 .
2 . @xmath164 for @xmath161 and @xmath162 .
3 . @xmath165 . for @xmath161 and @xmath162 [ lem - asym - ord ]
an asymmetric polygon is orderable .
_ proof : _ let @xmath130 be an asymmetric polygon .
for each vertex @xmath166 ( @xmath167 ) , we compute the tuple @xmath168 and take the lexicographic minimum of the @xmath0 tuples as the ordering of @xmath130 . since @xmath130 is asymmetric , @xmath159 for any @xmath135 and @xmath160 , @xmath169 for any @xmath135 and @xmath137 such that @xmath162 ( theorem [ obs - asymmetric - polygon ] ) .
hence , the ordering is same irrespective of the choice of origin or the coordinate axes .
thus , the set of vertices of an asymmetric polygon is orderable ( definition [ def - orderable - set ] ) . a set of points which are equidistant from the destination , form a convex polygon such that the vertices of the polygon are on the circumference of a circle .
there can be multiple such sets , i.e. , the robots within a set are equidistant from the destination but the robots in different sets are not equidistant from the destination . in such a case , we get multiple concentric circles each of them enclosing a convex polygon . let @xmath170 = @xmath171 , be the @xmath172 convex polygons whose union is the whole set of points .
the vertices of each polygon in @xmath170 are on the circumference of a circle .
these circles are concentric .
the center of the circles is known . generally , the center @xmath10 is the center of the sec formed by the points in @xmath170 .
@xmath10 is also considered as the center of any polygon @xmath173 .
elements of @xmath170 are sorted according to their distances from the center .
any polygon @xmath173 can be extracted by selecting the equidistant points from the center .
the polygon at level @xmath12 ( @xmath174 ) is the closest and the polygon at level @xmath172 ( @xmath175 ) is the farthest .
[ def - sym - pair ] a pair of polygons @xmath176 and @xmath177(denoted by @xmath178 ) in @xmath170 is called a _
symmetric pair _ , if @xmath176 and @xmath177 have a common line of symmetry .
a pair is _ asymmetric _ if it is not symmetric .
[ obs - pair - asym - ord ] if any of the polygons of a pair is asymmetric , then the pair is asymmetric .
[ lem - sym - notord ] a symmetric pair is not orderable . _
proof : _ suppose @xmath178 is a symmetric pair .
@xmath176 and @xmath177 have a common line of symmetry ( straight or skewed ) @xmath23 , which divides @xmath176 and @xmath177 into two equal halves ( straight or skewed mirror image ) .
the union of the vertices of @xmath176 and @xmath177 is divided into two equal halves by @xmath35 .
the union of the vertices of @xmath176 and @xmath177 is in @xmath41 .
hence , @xmath178 is not orderable ( lemma [ lemma - sym - ord ] ) .
suppose , the vertices of @xmath176 in pair @xmath178 are projected radially on the circumference of the enclosing circle of @xmath177 ( @xmath179 ) .
construct a polygon @xmath180 with the full set of vertices on the circle .
[ obs - merg - asym - pair ] if @xmath178 is an asymmetric pair , then the polygon @xmath180 is asymmetric .
[ lem - asym - pair - ord ] an asymmetric pair is orderable . _
proof : _ let @xmath181 be an asymmetric pair .
@xmath180 is an asymmetric polygon ( observation [ obs - merg - asym - pair ] )
. therefore @xmath180 is orderable ( lemma [ lem - asym - ord ] ) .
therefore @xmath182 is orderable .
[ def - sym - calg ] @xmath170 is called symmetric ( or @xmath170 is in @xmath183 ) if all polygons in @xmath170 have a common line of symmetry .
[ lem - all - pair - asym - ord ] if there exists an asymmetric pair in @xmath170 , then @xmath170 is orderable . _
proof : _ let @xmath181 in @xmath170 be the asymmetric pair such that @xmath184 is lexicographically minimum among all pairs in @xmath170 which are asymmetric . if @xmath170 has one pair , i.e. , @xmath170 = @xmath185 and @xmath186 is asymmetric , then @xmath170 is orderable ( lemma [ lem - asym - pair - ord ] ) . consider the case when @xmath170 has more than one pair .
since @xmath182 is asymmetric , it is orderable ( lemma [ lem - asym - pair - ord ] )
. let @xmath187 and @xmath188 be two orderings of @xmath176 and @xmath177 respectively .
let @xmath189 ( @xmath190 ) be any polygon in @xmath170 .
since , @xmath181 is asymmetric , @xmath180 is asymmetric ( observation [ obs - merg - asym - pair ] ) .
hence , @xmath191 is asymmetric and is orderable ( lemma [ lem - asym - pair - ord ] ) .
let @xmath192 be an ordering of @xmath189 .
varying @xmath142 from @xmath11 to @xmath172 ( @xmath190 ) , we get the ordering of polygons @xmath193 in terms of @xmath194 respectively .
hence , we get the ordering of @xmath170 .
[ lem - one - pair - asym - ord ] if @xmath170 contains at least one asymmetric polygon then @xmath170 is orderable . _
proof : _ let @xmath195 be asymmetric . for all @xmath196 , @xmath197 is asymmetric ( observation [ obs - pair - asym - ord ] ) .
hence , following lemma [ lem - all - pair - asym - ord ] , @xmath170 is orderable .
let @xmath176 , @xmath177 , and @xmath189 in @xmath170 be pairwise symmetric .
let @xmath198 , @xmath199 , @xmath200 be the lines of ( same ) symmetry of @xmath178 , @xmath201 and @xmath202 respectively .
our aim is to show that @xmath176 , @xmath177 and @xmath189 have a common line of symmetry . to show this
, we first characterize the polygon on the basis of symmetry they have .
the following lemma states a very interesting property for a convex regular polygon .
[ lemma - sym - regular ] if a polygon @xmath130 is convex and has more than one line of symmetry such that each line of symmetry passes through at least one vertex of @xmath130 , then @xmath130 is regular . and @xmath203 passing through the vertices @xmath204 and @xmath166 respectively.,width=188,height=226 ] _ proof : _ let @xmath205 be a line of symmetry for @xmath130 passing through the vertex @xmath204 of @xmath130 .
let the vertices starting from the vertex next to @xmath204 in clockwise direction be @xmath206 .
the angular distances between the vertices of @xmath130 , starting from @xmath204 in clockwise direction , are denoted by @xmath207 ( fig .
[ mod ] ) .
since @xmath205 is a line of symmetry for @xmath130 , @xmath208 i.e. , @xmath209 suppose there is another line of symmetry passing through the vertex , @xmath166 .
this implies , @xmath210 .
we can represent the series by the following equations : @xmath211 @xmath212 let us consider any angle @xmath213 for @xmath214 . combining equations [ 2nd ] and [ 3rd ]
we get , @xmath215 . substituting @xmath216 in equation @xmath17
we get @xmath217
. therefore , @xmath218 for @xmath219 .
this implies that , @xmath130 is regular .
let us define different types of polygon on the basis of symmetry as follows . *
a * type 0 * polygon is a regular convex symmetric polygon . * a * type 1 * polygon ( fig .
[ type3 ] ) is a convex , symmetric , non - regular polygon with even number of vertices and has exactly two lines of straight symmetry @xmath220 and @xmath221 , such that * * @xmath222 ( @xmath223 ) does not pass through a vertex of the polygon . * * @xmath224 .
+ it does not have any other line of straight symmetry but admits lines of skew symmetry . *
a * type 2 * polygon ( fig . [ type3 ] ) is a convex , symmetric , non - regular polygon with even number of vertices and has exactly one line of straight symmetry passing through either two vertices or two edges . *
a * type 3 * polygon ( fig . [ type3 ] ) is a convex , symmetric , non - regular polygon with odd number of vertices and has exactly one line of straight symmetry passing through a vertex and an edge . *
a * type 4 * polygon ( fig .
[ type3 ] ) is a convex , symmetric , non - regular polygon with odd number ( not prime ) of vertices such that the number of lines of symmetry is more than one but less than the number of vertices .
+ note : the number of lines of symmetry for this polygon is odd .
[ obs - exh - poly ] the above characterization of straight symmetric polygons is exhaustive .
[ obs - sym - type-0 ] if @xmath130 is a type 0 polygon with even number of vertices then any straight line passing through the center of @xmath130 is a line of symmetry for @xmath130 .
[ obs - sym - type-01 ] if @xmath130 is a type 0 polygon with odd number of vertices then any line of symmetry must pass through exactly one vertex of @xmath130 .
[ lem - sym - type-01 ] any straight line passing through the center of a type 1 polygon , is a line of symmetry ( skewed or straight ) for that polygon . _
proof : _ let @xmath130 be a type 1 polygon .
@xmath130 has exactly two lines of straight symmetry say @xmath133 and @xmath134 .
@xmath133 and @xmath134 pass through the center @xmath10 of @xmath130 .
it is sufficient to prove that any line between @xmath133 and @xmath134 is a line of skew symmetry for @xmath130 . without loss of generality
let us rotate @xmath133 by some angle , around @xmath10 , keeping it between @xmath133 and @xmath134 .
let @xmath140 be the new position of the line .
@xmath140 intersects @xmath130 either at two edges or passes through two vertices .
if @xmath140 passes through two edges of @xmath130 , then using lemma [ losks - edge ] and using an argument similar to that used in the proof of lemma [ lem - rectangle ] , it can be shown that , polygonal chains on both sides of @xmath140 are skew symmetric . if @xmath140 passes through two vertices of @xmath130 , then using lemma [ losks - vertex ] and using an argument similar to that used in the proof of lemma [ lem - rectangle ] , it can be shown that , polygonal chains on both sides of @xmath140 are skew symmetric . hence , @xmath140 is a line of skew symmetry .
[ obs - gcd ] let @xmath176 and @xmath177 be two polygons with @xmath225 and @xmath226 vertices respectively .
@xmath225 and @xmath226 are odd numbers . if @xmath227 , @xmath228 has one common line of symmetry , then @xmath227 , @xmath228 has @xmath229 many common lines of symmetry .
let @xmath176 and @xmath177 ( @xmath179 ) be two polygons in @xmath170 with number of vertices @xmath225 and @xmath226 respectively such that * @xmath225 , @xmath226 are odd .
* @xmath225 does not divide @xmath226 or @xmath226 does not divide @xmath225 . *
@xmath230 .
@xmath178 has at least one common line of symmetry .
we construct @xmath180 , and replace both @xmath231 and @xmath177 by @xmath180 in @xmath170 .
following observation [ obs - gcd ] , @xmath180 has @xmath229 many common lines of symmetry .
[ lem - los - gij ] a line of symmetry for @xmath180 is a common line of symmetry for @xmath178 . _
proof : _ the number of lines of symmetry of @xmath180 is @xmath229 , where @xmath225 and @xmath226 are the numbers of vertices of @xmath176 and @xmath177 respectively .
the angle between two adjacent lines of symmetry for @xmath180 is always @xmath232 degrees .
the angle between two adjacent lines of symmetry of @xmath176 is @xmath233 degrees , which divides @xmath232 degrees .
this implies that these lines of symmetry of @xmath180 are also the lines of symmetry for @xmath176 . using similar argument
it can be stated that the lines of symmetry of @xmath180 are also the lines of symmetry for @xmath177 .
hence the result follows .
[ theorem - los - g1n ] a line of symmetry for @xmath234 is a common line of symmetry for @xmath235 . _
proof : _ the statement is true for two concentric polygons ( lemma [ lem - los - gij ] ) .
suppose the result is true for @xmath37 polygons
. now @xmath236 polygon @xmath237 is introduced .
if @xmath237 has even number of vertices then any common line of symmetry for @xmath238 , which will pass through the center of the polygons , is also a line of symmetry for @xmath237 .
hence the result is true .
suppose @xmath237 has odd number of vertices .
let @xmath133 be a line of symmetry for @xmath239 . from lemma [ lem - los - gij ]
, @xmath133 is a common line of symmetry for @xmath174 and @xmath240 .
we merge @xmath239 and @xmath241 to get @xmath242 .
let @xmath134 be a line of symmetry for @xmath242 . from lemma [ lem - los - gij ]
, @xmath134 is a common line of symmetry for @xmath239 and @xmath243 .
again , @xmath134 is a common line of symmetry for @xmath244 and @xmath245 .
proceeding further in this manner it can be shown that if @xmath246 is a line of symmetry for @xmath247 then @xmath246 is a common line of symmetry for @xmath235 . [ obs - type 4 ] if @xmath225 and @xmath226 are odd , @xmath178 has a common line of symmetry .
additionally , if @xmath248 , then @xmath180 is a type 4 polygon .
we repeat this process of merging polygons until every symmetric polygon with odd number of vertices becomes either of type 0 or type 3 or type 4 .
we get a modified version of @xmath170 , and call it @xmath249 .
[ lem - type4 ] let @xmath176 , @xmath177 and @xmath189 be three polygons in @xmath249 having odd number of vertices @xmath225 , @xmath226 and @xmath250 respectively . no two of @xmath225 , @xmath226 and @xmath250 are equal and no one of @xmath225 , @xmath226 and @xmath250 is a multiple of other . if @xmath176 , @xmath177 and @xmath189 are pairwise symmetric then @xmath226 , @xmath226 and @xmath250 are prime to each other .
_ proof : _ suppose @xmath226 , @xmath226 and @xmath250 are not prime to each other . without loss of generality ,
let @xmath230 .
hence , @xmath176 and @xmath177 can be merged to form @xmath180 , which is a type 4 polygon .
this contradicts the construction process of @xmath251 .
[ lem-3-pair - sym ] any three polygons @xmath176 , @xmath177 , and @xmath189 in @xmath249 are pairwise symmetric if and only if @xmath176 , @xmath177 and @xmath189 have a common line of symmetry .
_ proof : _ * if : * trivial . * only if : * since every pair is symmetric each individual polygon of @xmath176 , @xmath177 and @xmath189 must be symmetric .
each polygon is either skew symmetric or type 0/1/2/3/4 polygon . let @xmath252 , @xmath253 and @xmath254 be three common lines of symmetry for @xmath178 , @xmath201 and @xmath255 respectively .
@xmath252 , @xmath253 and @xmath254 pass through the common center ( @xmath10 ) of @xmath176 , @xmath177 and @xmath189 . *
if any of the polygons @xmath176 , @xmath177 and @xmath189 ( without loss of generality let it be @xmath176 ) is a skew symmetric polygon , then any line passing through @xmath10 is a line of symmetry for @xmath176 ( lemma [ lem - skew - los ] ) .
as @xmath253 passes through @xmath10 , it is also a line of symmetry for @xmath176 .
therefore , @xmath253 is a common line of symmetry for @xmath176 , @xmath177 and @xmath189 . *
if any of the polygons @xmath176 , @xmath177 and @xmath189 ( without loss of generality let it be @xmath176 ) is a type 1 polygon , then any line passing through @xmath10 is a line of symmetry for @xmath176 ( lemma [ lem - sym - type-01 ] ) .
so @xmath253 is also a line of symmetry for @xmath176 . *
if any of the polygons @xmath176 , @xmath177 and @xmath189 ( without loss of generality let it be @xmath176 ) is a type 2 or type 3 polygon , then @xmath176 has exactly one line of symmetry .
hence , @xmath256 . so the result follows . *
if any of the polygons @xmath176 , @xmath177 and @xmath189 ( without loss of generality let it be @xmath176 ) is a type 0 polygon with even number of vertices , then any line passing through @xmath10 is a line of symmetry for @xmath176 ( observation [ obs - sym - type-0 ] ) .
so @xmath253 is also a line of symmetry for @xmath176 .
* if each of @xmath176 , @xmath177 and @xmath189 is a type 0 polygon with odd number of vertices or type 4 polygon then following sub cases are possible : * * if any two polygons ( without loss of generality , let them be @xmath176 and @xmath177 ) have equal number of vertices , then all lines of symmetry of @xmath177 are also the lines of symmetry for @xmath176 and vice - versa . hence , @xmath253 is a line of symmetry for @xmath176 . * * suppose , the number of vertices of one polygon ( say @xmath176 ) is a multiple of the number of vertices of another polygon ( say @xmath177 ) .
since , @xmath176 and @xmath177 share a common line of symmetry , every line of symmetry for @xmath177 is also a line of symmetry for @xmath176 .
so @xmath253 is a common line of symmetry for all three .
+ * * now we consider the case when none of the above is true .
let @xmath225 , @xmath226 and @xmath250 be the number of vertices of @xmath176 , @xmath177 and @xmath189 respectively ( figure [ oddprime ] ) .
@xmath225 , @xmath226 and @xmath250 are prime to each other ( lemma [ lem - type4 ] ) .
@xmath252 passes through a vertex @xmath257 of @xmath176 and a vertex @xmath204 of @xmath177 .
first we consider the case when both @xmath257 and @xmath204 lie on the same ray ( @xmath258 ) starting from @xmath10 .
suppose , @xmath252 does not pass through any vertex of @xmath189 .
let @xmath259 be the vertex of @xmath189 which is closest to @xmath260 in the clockwise direction .
@xmath261 makes an angle @xmath262 with @xmath258 at @xmath10 .
we label the vertices of @xmath176 in the clockwise direction starting from @xmath257 and denote them by @xmath263 .
we label the vertices of @xmath177 in the clockwise direction starting from @xmath204 and denote them by @xmath264 .
we label the vertices of @xmath189 in the clockwise direction starting from @xmath259 and denote them by @xmath265 .
suppose , @xmath253 passes through the @xmath266 vertex @xmath267 of @xmath177 and the @xmath268 vertex @xmath269 of @xmath189 .
@xmath254 passes through the vertex @xmath270 of @xmath189 and @xmath271 of @xmath176 .
@xmath272 and @xmath273 make angles @xmath274 and @xmath275 respectively with @xmath258 .
@xmath276 and @xmath277 make angles @xmath278 and @xmath279 respectively with @xmath258 .
+ since , @xmath280 , we get the following equation : @xmath281 since , @xmath282 , @xmath283 + substituting equation @xmath284 from equation @xmath285 , + @xmath286 + since @xmath287 is an integer , @xmath288 must be an integer . @xmath289 and @xmath290 . @xmath291 . since , @xmath225 , @xmath226 and @xmath250 are pairwise relatively prime , @xmath292 . from equation @xmath284 , @xmath293 .
hence , @xmath253 is the line of symmetry for @xmath176 , @xmath177 and @xmath189 .
let us consider the case when @xmath257 and @xmath204 lie at different sides of @xmath10 ( fig [ oddprime1 ] ) .
the equations @xmath284 and @xmath285 will now change . since
, @xmath280 , we get the following equation : @xmath294 since , @xmath282 , @xmath283 substituting equation @xmath295 from equation @xmath296 , @xmath297 from the argument stated previously , the value of @xmath298 is a fraction . if @xmath299 , then it has a fractional part .
so , @xmath300 is not a non - zero integer .
since , @xmath301 is an integer , it must be zero . from equation @xmath295 , @xmath302 .
hence , @xmath253 is a common line of symmetry for @xmath176 , @xmath177 and @xmath189 .
similarly the result follows for other cases such as when @xmath267 and @xmath269 lie on different sides of @xmath10 or @xmath271 and @xmath270 lie on different sides of @xmath10 .
[ theorem - sym - pair - symg ] every pair in @xmath251 is a symmetric pair , if and only if there is a common line of symmetry for all the polygons in @xmath251 . _
proof : _ * if : * trivial . * only if : * we prove this by induction on the number of polygons in @xmath251 .
if @xmath251 contains three polygons then the result follows from lemma [ lem-3-pair - sym ] .
suppose , the statement is true for @xmath303 polygons .
now the @xmath236 polygon say @xmath237 is introduced .
if @xmath237 has even number of vertices then any line of symmetry for @xmath304 is a line of symmetry for @xmath237 .
hence the result is true .
suppose @xmath237 has odd number of vertices .
@xmath305 is a symmetric pair ( given ) . since @xmath306 has a common line of symmetry ( induction hypothesis ) , that same line is also a common line of symmetry for @xmath307 .
similarly , using the induction hypothesis @xmath308 is also a symmetric pair . hence following lemma
[ lem-3-pair - sym ] , @xmath309 , @xmath310 and @xmath311 have a common line of symmetry , say @xmath23 . following theorem [ theorem - los - g1n ] , @xmath23 is a line of symmetry for @xmath312 .
thus the result follows .
[ obs1-gequig ] if @xmath170 is symmetric then every pair in @xmath170 is a symmetric pair .
[ gequig ] @xmath251 is symmetric if and only if @xmath170 is symmetric . _
proof : _ follows from theorem [ theorem - los - g1n ] .
let @xmath25 be a set of points on the 2d plane .
we first identify the sec for @xmath25 . with respect to the center of the sec
we divide the points of @xmath25 into a set of concentric polygons @xmath313 .
[ psym - gsym ] @xmath25 is in @xmath183 iff @xmath313 is in @xmath183 . _
_ proof:__*if : * if @xmath313 is in @xmath183 , all polygons in @xmath170 have a common line of symmetry ( definition [ def - sym - calg ] ) .
hence , the vertices in @xmath314 have a line of symmetry .
the set of vertices in @xmath314 , which is @xmath25 , is in @xmath183 . * only if : * let @xmath23 be a line of symmetry for @xmath25 .
@xmath23 also passes through the center of sec of @xmath25 ( observation [ lemma - los - passes - centersecp ] ) .
however , @xmath25 is the set of vertices in @xmath314 .
@xmath23 also is the line of symmetry of the set of vertices in @xmath315 .
it is easy to note that the mirror image of any vertex of @xmath316 , is a vertex of @xmath317 .
this implies that , @xmath23 is the line of symmetry for all @xmath173 .
therefore all @xmath316 , are symmetric across @xmath23 . hence , @xmath313 is in @xmath183 ( definition [ def - sym - calg ] ) .
@xmath25 is in @xmath318 iff @xmath313 is in @xmath318 .
[ pasym - pord ] @xmath25 is orderable if and only if @xmath25 is in @xmath318 .
_ proof : _ * if : * if @xmath25 is in @xmath318 , @xmath313 is in @xmath318 ( theorem [ psym - gsym ] ) . if @xmath313 is in @xmath318 , there is no common line of symmetry for the polygons in @xmath313 ( definition [ def - sym - calg ] ) .
there exists an asymmetric pair in @xmath313 .
following lemma [ lem - all - pair - asym - ord ] , @xmath319 is orderable .
hence @xmath25 is orderable . * only if : * if @xmath25 is orderable , @xmath25 is in @xmath318 ( lemma 1 ) .
[ cor - pasym - pord ] @xmath25 is orderable if and only if @xmath313 is in @xmath318 .
let @xmath9 be a set of robots . from the previous section
, we note that if @xmath9 is orderable , then leader election is possible from @xmath9 . in this case
the first robot in the ordering becomes the leader . in this section ,
we present the leader election algorithm for a set of robots , @xmath9 .
@xmath9 can be viewed as a set of multiple concentric polygons @xmath320 as described previously .
the leader election algorithm elects leader from @xmath321 .
first we present leader election algorithm for a single polygon @xmath130 .
then we extend the algorithm for the set of concentric polygons @xmath170 .
a polygon may have more than one line of symmetry .
the number of lines of symmetry of @xmath130 is called the degree of symmetry of @xmath130 .
algorithms @xmath322 and @xmath323 are used for checking symmetry in @xmath130 when the numbers of vertices of @xmath130 are odd and even respectively .
@xmath324 @xmath325 _ correctness of check_symmetry_odd(g ) : _ follows from theorem [ obs - asymmetric - polygon ] .
@xmath324 @xmath325 @xmath326 @xmath327 _ correctness of check_symmetry_even(g ) : _ follows from theorem [ obs - asymmetric - polygon ] .
note that if @xmath130 has even number of vertices then every line of symmetry ( straight or skewed ) passes through 2 vertices and is counted twice .
therefore , the algorithm returns the @xmath328 by dividing it by @xmath15 .
next we present a leader election algorithm , when the convex polygon has degree of symmetry one .
_ correctness of elect_leader_sym(g ) : _ follows from theorem [ obs - asymmetric - polygon ]
. [ 1sym_le ] if the degree of symmetry of a convex symmetric polygon is one and the line of symmetry passes through two vertices of the polygon , then leader election is possible . once the leader is elected for a symmetric polygon with degree of symmetry one , the leader can be moved such a way that the new polygon becomes asymmetric .
following algorithm does this task .
@xmath329 = @xmath330 move @xmath329 , @xmath331 distance , to its right side , on the circumference of the circle inscribing @xmath130 return @xmath130 _ correctness of make_symtoasym(g ) : _ whenever the leader @xmath329 moves from its position , @xmath130 becomes asymmetric .
then no other robot executes the algorithm @xmath332 .
therefore , there is no chance that @xmath130 becomes symmetric again .
following algorithm elects leader when @xmath130 is asymmetric .
@xmath333 = lexicographic minimum of @xmath334 @xmath325 @xmath329 = @xmath135 return @xmath329 _ correctness of elect_leader_asym(g ) : _ follows from ( lemma [ lem - asym - ord ] ) .
let @xmath170 be a set of @xmath172 concentric polygons as described previously .
@xmath174 is the nearest to the center .
@xmath175 is the farthest from the center .
algorithm @xmath335 selects a leader from @xmath170 .
the algorithm assumes that @xmath175 is asymmetric .
therefore , @xmath336 is an asymmetric pair ( observation [ obs - pair - asym - ord ] ) .
@xmath335 checks if @xmath174 is asymmetric .
if @xmath174 is asymmetric , then the leader is selected from @xmath174 by algorithm @xmath337 . if @xmath174 is symmetric , then @xmath335 computes @xmath338 . since ,
@xmath175 is asymmetric , @xmath338 is orderable .
leader election is possible from @xmath338 . _
correctness of electleader(@xmath170 ) : _ follows from lemma [ lem - one - pair - asym - ord ] .
first @xmath170 is reconstructed to form @xmath251 .
@xmath339 first @xmath170 is preprocessed to generate @xmath251 .
algorithm @xmath340 selects a leader from @xmath251 , assuming that @xmath175 in @xmath251 is asymmetric .
the algorithm @xmath341 pre - processes @xmath251 to make @xmath175 asymmetric , whenever that is possible .
let @xmath175 in @xmath251 be symmetric .
if any @xmath176 in @xmath251 is asymmetric , then @xmath251 is asymmetric ( lemma [ lem - one - pair - asym - ord ] ) . if there is no asymmetric polygon in @xmath251 , but their exists a @xmath176 such that @xmath176 contains only one vertex @xmath71 , and the vertex is on the line of symmetry of @xmath175 , then @xmath71 is moved slightly ( @xmath331 distance ) form its current position .
@xmath342 becomes asymmetric .
@xmath343 is computed .
the first vertex from @xmath175 , in the ordering of @xmath343 is selected as a leader from @xmath175 .
the leader is moved to a new position ( @xmath331 distance away from the current position ) to make @xmath175 asymmetric .
@xmath344 finds an asymmetric polygon @xmath345 from @xmath251 using binary search on the set @xmath314 .
if @xmath344 can not find any asymmetric polygon @xmath251 then an asymmetric pair @xmath181 is searched using algorithm @xmath346 . if any @xmath178 in @xmath251 is asymmetric , then @xmath251 is asymmetric ( lemma [ lem - all - pair - asym - ord ] ) . in such a case ,
@xmath347 is computed by radially projecting the polygon @xmath180 on @xmath175 .
the first vertex from @xmath175 , in the ordering of @xmath347 is selected as a leader from @xmath175 .
the leader is moved to a new position ( @xmath331 distance away from the current position ) to make @xmath175 asymmetric . if no such pair is found , then @xmath251 is symmetric and hence not orderable .
the algorithm also maintains the number of vertices of @xmath175 to be more than @xmath17 .
this assures that the sec will not change during the motion of the leader from @xmath175 . finally , @xmath341 adjusts @xmath175 to be asymmetric if it is initially symmetric .
@xmath348 [ obs - v - move ] suppose there exists no asymmetric polygon and asymmetric pair , but there exists a polygon @xmath349 such that @xmath176 contains a single vertex @xmath329 .
in this case , the leader @xmath329 selected from @xmath176 is same no matter whether @xmath329 is in motion or @xmath350 distance apart from its position in symmetry .
[ lem - vl - move ] when @xmath329 is moving no other robot will be selected as leader .
_ proof : _
@xmath329 is selected either from @xmath175 or @xmath351 .
if @xmath329 is selected from @xmath175 , then after @xmath329 starts moving , @xmath175 becomes asymmetric .
no robot will execute @xmath352 again .
therefore @xmath329 remains the leader .
if @xmath329 is selected from @xmath351 , then after @xmath329 starts moving , @xmath329 forms the new @xmath351 where @xmath329 is the only robot in @xmath351 .
hence , @xmath329 remains the leader .
_ correctness of makeasym(@xmath175 ) : _ the algorithm always maintains the number of vertices of @xmath175 to be greater than three .
this ensures that the sec would not change during the motion of any robots from @xmath175 .
following lemma [ lem - all - pair - asym - ord ] , lemma [ lem - one - pair - asym - ord ] and observation [ obs - v - move ] , if their is an asymmetric polygon or an asymmetric pair or a polygon with single vertex on the line of symmetry of @xmath251 , then @xmath251 is orderable and leader election is possible . lemma [ lem - vl - move ] assures that when the leader is moving , no other robots will move . in this section ,
the algorithm @xmath353 for gathering @xmath0 @xmath354 transparent fat robots is presented .
initially all robots are assumed to be stationary and separated .
let @xmath9 be a set of such robots .
let @xmath355 be the set of robots already in the gathering pattern .
initially @xmath355 is null .
the algorithm first finds the center @xmath10 of the sec for the robots in @xmath9 .
the robots in @xmath9 are ordered based on their distances from @xmath10 .
the robots equidistant from @xmath10 are on the circumference of a circle and form a convex polygon .
the robots are the vertices of the polygons .
hence , we get a number of concentric polygons having a common center @xmath10 . the robots equidistant from @xmath10 , are said to be in the same level .
let there be @xmath172 levels of distances .
the robots nearest to @xmath10 are at level @xmath12 and the robots farthest from @xmath10 are at level @xmath172 .
let @xmath170 be the set of these @xmath172 concentric polygons .
@xmath170 is preprocessed to generate @xmath251 using @xmath356 .
the gathering algorithm then processes @xmath251 using @xmath341 to make @xmath175 asymmetric .
if @xmath341 reports that @xmath251 is in @xmath41 , then gathering is not possible .
if @xmath341 returns an orderable @xmath170 with @xmath175 asymmetric , then @xmath353 selects a leader robot @xmath357 using @xmath358 and moves @xmath357 to extend the gathering pattern already formed , using algorithm @xmath359 .
@xmath357 moves towards @xmath10 in order to build the desired gathering pattern .
if @xmath10 is not occupied by any other robot , then @xmath357 becomes the central robot of the gathering pattern .
if @xmath10 is occupied by another robot , then the algorithm finds the last layer @xmath360 of the gathering pattern .
if @xmath360 is full , then @xmath357 makes the new layer @xmath18 ( figure .
[ gatheringalgof ] ) .
if @xmath360 is not full , then @xmath357 may slide around @xmath10 ( if required ) and is placed in layer @xmath360 .
when @xmath357 reaches its destination and stops , the robot is removed from @xmath9 and added to @xmath355 . of the _
_ sec__,width=453,height=170 ] remove @xmath357 from @xmath9 add @xmath357 to @xmath355 calculate @xmath361 by the robots in set @xmath9 .
@xmath10 be the center of @xmath361 build the set @xmath170 with the robots in set @xmath9 @xmath251 = @xmath356 the following @xmath362 lemmas shows that the choice of the destination and the robot for movement remain invariant during the motion of the designated robot . remains minimum during its movement , width=188,height=188 ] when a robot is moving towards @xmath10 ( the center of sec ) and/or sliding around @xmath10 , its distance from @xmath10 is still minimum among all the robots in set @xmath9 .
_ proof _ : let @xmath357 be a robot , which is selected for moving towards @xmath10 .
let @xmath363 ( figure .
[ invariant ] ) .
no other robot is inside the circle of radius @xmath364 . according to @xmath353 , only @xmath357 starts moving towards @xmath10 and other robots are stationary .
@xmath357 moves towards @xmath10 and after touching the robots at last layer @xmath360 , it may start sliding around @xmath10 and stops after getting its final position . during the motion of @xmath357
, it remains inside the circle of radius @xmath364 .
therefore , @xmath357 remains at minimum distance from @xmath10 , among all robots in set @xmath9 .
after it reaches its final position , it is removed from set @xmath9 and added to set @xmath355 .
a new nearest robot in set @xmath9 is selected .
no obstacle appears in the path of a mobile robot to its destination . _
proof : _ let @xmath357 be a robot which is selected for moving towards @xmath10 . according to @xmath353 , @xmath357 is nearest to @xmath10 .
let @xmath365 in figure .
[ obstacle ] be the rectangle in which @xmath357 resides during its movement to @xmath10 .
@xmath366 is @xmath15 units .
let rectangles @xmath367 and @xmath368 be of unit width and length equal to the length of rectangle @xmath365 .
it is sufficient to show that , if center of no other robot appears inside the quadrilateral @xmath369 , then the path of @xmath357 towards @xmath10 will be free of obstacles .
if the center of any other robot appears inside the region @xmath370 , then that robot will be nearest to @xmath10 .
however , @xmath357 has been selected for moving .
therefore , @xmath357 is nearest to @xmath10 and no robots appear inside the region @xmath370 . if any robot appears inside the regions @xmath371 or @xmath372 then either the robots touch @xmath357 or is farther from @xmath10 than @xmath357 .
as we consider the robots in set @xmath9 are separated , no other robots will touch @xmath357 . according to @xmath353
, @xmath357 is the only robot to be considered for moving .
therefore , any robot inside the regions @xmath371 and/or @xmath372 is not considered in the path of @xmath357 to its destination .
hence , no obstacle appears in the path of a mobile robot to its destination .
in this paper , we have proposed a deterministic distributed algorithm for gathering @xmath0 @xmath1 autonomous , oblivious , homogeneous , asynchronous , transparent fat robots .
the robots do not have explicit communication capability or common coordinate system .
we assume that initially all robots are stationary and separated . given a set of such robots
, we find a destination which remains invariant during the run time of the algorithm .
one robot is allowed to move at any point of time .
the robot which moves , is selected such a way that , it will be the only eligible robot until it reaches to its destination .
therefore , a leader election technique is required to elect a robot for movement from the set of homogeneous robots .
we also propose the leader election algorithm in order to select an robot for movement .
this paper characterizes all the cases where leader election is not possible . to do so
, we make an ordering with a set of arbitrarily positioned robots .
if ordering is possible with a set of robots , then leader election is also possible and hence the robots can gather deterministically .
the paper characterizes the configurations formed by robots where ordering and hence , gathering is not possible .
this configurations are defined as symmetric configurations .
the gathering algorithm ensures that multiple mobile robots will gather for all asymmetric configurations .
ando h. , oasa y. , suzuki i. , yamashita m. : a distributed memoryless point convergence algorithm for mobile robots with limited visibility , ieee transaction on robotics and automation , 15(5 ) 1999 , pp .
818 - 828 .
dieudonne y. , petit f. , villain v. : brief announcement : leader election vs. pattern formation , podc 2010 , pp .
404 - 405 .
efrima a. , peleg d. : distributed models and algorithms for mobile robot systems , @xmath377 software seminar , 2007 , lncs 4362 , pp .
70 87 .
flocchini p. , prencipe g. , santoro n. , widmayer p. : gathering of autonomous mobile robots with limited visibility , @xmath378 annual symposium on theoretical aspects of computer science , 2001 , lncs 2010 , pp .
247 - 258 .
gan chaudhuri s. , mukhopadhyaya k. : gathering asynchronous transparent fat robots , @xmath379 proc .
international conference of distributed computing and internet technology , lncs 5966 , 2010 , pp .
170 - 175 . katayama y. , tomida y. , imazu h. , inuzuka n. , wada k. : dynamic compass models and gathering algorithms for autonomous mobile robots , @xmath381 international colloquium on structural information and communication complexity , 2007 , lncs 4474 , pp .
274 - 288 . | _ gathering _ of autonomous mobile robots is a well known and challenging research problem for a system of multiple mobile robots .
most of the existing works consider the robots to be dimensionless or points .
algorithm for gathering _ fat robots _
( a robot represented as a unit disc ) has been reported for _ three _ and _ four _ robots .
this paper proposes a distributed algorithm which deterministically gathers @xmath0 @xmath1 asynchronous , fat robots . the robots are assumed to be transparent and they have full visibility .
since the robots are unit discs , they can not overlap ; we define a _ gathering pattern _ for them . the robots are initially considered to be stationary .
a robot is visible in its motion .
the robots do not store past actions .
they are anonymous and can not be distinguished by their appearances and do not have common coordinate system or chirality .
the robots do not communicate through message passing . in the proposed
_ gathering _
algorithm one robot moves at a time towards its destination .
the robot which moves , is selected in such a way that , it will be the only robot eligible to move , until it reaches its destination . in case of a tie ,
this paper proposes a _
leader election _ algorithm which produces an ordering of the robots and the first robot in the ordering becomes the leader .
the ordering is unique in the sense that , each robot , characterized by its location , agrees on the same ordering .
we show that if a set of robots can be ordered then they can gather deterministically .
the paper also characterizes the cases , where ordering is not possible .
this paper also presents an important fact that , if leader election is possible then _ gathering pattern _ formation is possible even with no chirality .
it is known that , pattern formation and leader election are equivalent , for @xmath2 robots when the robots agree on chirality .
this paper proposes an algorithm for formation of the _ gathering pattern _ for transparent fat robots without chirality .
asynchronous , fat robots , gathering , leader election , chirality . |
many realistic turbulent flows are subject to modulated driving forces , as e.g. the atmosphere of the earth driven by the periodic heating of the sun or the pulsed flow through a pipeline .
three dimensional turbulence is characterized by an energy cascade from the outer length scale , where the forcing acts , to the dissipative scale , where most of energy is dissipated , see e.g. @xcite .
the down - cascading of energy from large to small scales takes a characteristic time @xmath1 . in a statistically stationary flow
the energy dissipation rate equals the energy input rate . in a situation with time dependent energy input ,
on the other hand , this statement will only hold on _ average _ , whereas the energy dissipation at a certain time @xmath2 is expected to depend on the energy input at an _ earlier _ time due to the finite time delay of the energy transfer . in a previous work @xcite the effect of an energy input rate modulated in time , @xmath3 with a modulation amplitude @xmath4 and a modulation frequency @xmath5 , has been studied within a variable range mean - field theory @xcite .
the response of the system can be observed in the second order velocity structure function of the flow field at the outer length scale @xmath6 , @xmath7 , which is equivalent to the reynolds number @xmath8 of the flow and the total energy @xmath9 of the system .
here , @xmath10 is the rms of one velocity component and @xmath11 is the viscosity .
the response follows the oscillation of the energy input rate with almost constant modulation amplitude at low frequencies @xmath5 of the energy input rate , whereas the response amplitude strongly decreases ( @xmath12 ) at higher frequencies .
the finite energy transfer time @xmath1 plays a crucial role in this theory .
this time @xmath1 is the average time the energy stays within the system while it is transported by the interaction cascade from the large eddies towards the small eddies , where it is finally dissipated .
this intrinsic time scale of the system is a multiple @xmath13 of order 1 of the large eddy turnover time @xmath14 , corresponding to the sum over the eddy turnover times on all scales . @xmath15
determines the frequency at which the crossover takes place between the regime of constant response amplitude and decreasing amplitude .
in addition , it leads to an oscillating behavior of the system response with driving frequency @xmath5 , where the maxima and minima are at frequencies connected to the inverse of the energy transfer time @xmath1 . in the limit of large frequencies
@xmath5 , the extrema of the response can be estimated to be at frequencies @xmath16 , @xmath17 . as a function of the driving frequency @xmath5 for a modulated energy input rate @xmath18 calculated within the goy shell model ( full dots ) , see section [ ssecgoymode ] .
the modulation amplitude is set to @xmath19 , and the cascade time delay turned out to be @xmath20 .
the stationary reynolds number is , the viscosity @xmath21 , and the large eddy turnover time @xmath22 .
time and length units are set by @xmath11 , @xmath23 and @xmath24 in goy .
our findings are compared with the response amplitude as calculated within the mean - field model with the same @xmath25 and @xmath1 ( dashed lines ) .
( a ) log - log plot of the amplitude @xmath26 versus frequency .
the long - dashed line denotes the low frequency limit of the mean - field theory , @xmath27 , and the dotted line the high frequency limit , @xmath28 .
the arrow denotes @xmath29 . near to this frequency
the crossover takes place in goy .
inset : linear scale plot of the response amplitude .
( b ) log - log plot of the amplitude compensated by the asymptotic amplitude , i.e. , @xmath30 versus frequency . a clear maximum is observed in goy at a frequency near to the maximum of the mean - field amplitude . ] as a function of the driving frequency @xmath5 for a modulated energy input rate @xmath18 calculated within the rewa model ( full dots ) , see section [ ssecrewamode ] .
the modulation amplitude is set to @xmath31 , and the cascade time scale results to be @xmath32 .
the kolmogorov constant is found to be @xmath33 in this simulation instead of @xmath34 .
the stationary reynolds number is @xmath35 , the viscosity @xmath36 , and the large eddy turnover time @xmath37 .
times are measured in units of @xmath38 in rewa .
the result is compared with the response amplitude as calculated within the mean - field model with the same @xmath25 , @xmath1 , and @xmath39 ( dashed lines ) .
( a ) log - log plot of the amplitude @xmath26 versus frequency .
the dotted line is @xmath12 .
the arrow indicates the mean - field crossover frequency @xmath40 .
inset : linear scale plot of the response amplitude .
( b ) log - log plot of the compensated amplitude , i.e. , @xmath30 versus frequency . a clear maximum is observed in rewa at a frequency near to the first maximum of the mean - field amplitude .
the arrow indicates the height of the maximum , i.e. , a deviation from the @xmath0-decay by a factor of 1.4 in rewa . ] and
amplitude of the energy input rate @xmath41 as a function of the driving frequency @xmath5 for a modulated driving force @xmath42 calculated within the goy shell model ( full dots ) , see section [ ssecgoymodf ] . the modulation amplitude is set to @xmath43 , and the cascade time scale
is found to be @xmath44 .
the stationary reynolds number is @xmath45 , the viscosity @xmath21 , and the large eddy turnover time @xmath46 .
the result is compared with the response amplitude as calculated within the mean - field model with the same @xmath25 and @xmath1 ( dashed lines ) .
( a ) log - log plot of the amplitude @xmath26 versus frequency .
the long - dashed line denotes the low frequency limit of the mean - field theory , @xmath47 , and the dotted line the high frequency limit , @xmath48 .
the arrow denotes @xmath49 . near to this frequency
the crossover takes place in goy .
inset : linear scale plot of the response amplitude .
( b ) log - log plot of the compensated amplitude , i.e. , @xmath30 versus frequency .
the dotted line denotes @xmath50 const .
( c ) linear scale plot of the energy input amplitude @xmath41 versus frequency .
the mean - field amplitude as well as the goy amplitude start for low @xmath5 with @xmath51 and merge at @xmath52 for high frequencies .
the goy amplitude shows only the first main maximum . ]
recent experiments on modulated turbulence in a cylinder between two counter rotating disks @xcite revealed evidence for the proposed response maxima . in accordance with the predictions from the mean - field theory @xcite , for small frequencies
a constant response amplitude was measured . for large driving frequencies a @xmath0-decay of the velocity response amplitude
was observed , again in agreement with the prediction from our mean - field approach .
note here , that both the velocity response as well as the energy response are the same up to a factor of 2 , in linear order , cf .
section vi of reference @xcite . in the experiments the amplitude of the driving force rather than that of the energy input rate is modulated .
since the energy input rate is not a controlled quantity any more it can serve to measure the response of the system . of course , also within the mean - field theory we can apply a modulated driving _ force _
, see @xcite . the main features , the @xmath0-decay of the energy response amplitude for high frequencies and the constant response amplitude for low frequencies , pertain .
the response maxima are only slightly shifted in comparison to the case of a controlled and modulated energy input rate . in the case of a modulated driving force , as in the experiments , the energy input rate as a response of the system , also shows maxima in addition to the mentioned mean features .
these are at the same frequencies as the maxima of the total energy response amplitude . in the mean - field approach ,
the ( intermittent ) fluctuations of the energy and , in particular , of the cascade time @xmath1 are not present . in experiments and numerical simulations
these fluctuations are of course present , and they may lead to broader and less pronounced response maxima and minima . therefore , in this paper we shall study the frequency dependence of the response to a modulated energy input rate into a system where turbulent fluctuations are included . in particular , we shall address the question whether the response maxima and minima can still be well identified in the presence of fluctuations .
furthermore , we not only consider a modulated energy input rate , but also discuss the slightly different case of a modulated driving force in order to compare with the above mentioned experiments . an appropriate way to numerically study the problem of modulated turbulence would be a direct numerical simulation of the navier - stokes equation for this specific time - dependent energy input rate .
however , as we need high reynolds numbers to achieve fully developed , isotropic , and homogeneous turbulence and , in addition , need the response of the system as a function of time for a wide range of driving frequencies , the computational demands would be too high .
therefore , we first study the problem within a dynamical cascade model of turbulence , the gledzer - ohkitani - yamada ( goy ) shell model @xcite . with this model large reynolds numbers and enough statistics within a reasonable computing time for each driving frequency can be achieved .
the goy model has been successfully used in a study about decaying and kicked turbulence @xcite .
in addition , to be even closer to a numerical navier - stokes simulation and to distinguish between real effects and artifacts of the turbulence model , we follow another approach .
we calculate the response of the system to a modulated energy input rate within a reduced wave vector set approximation ( rewa ) , @xcite , where the navier - stokes equation is solved on a reduced , geometrically scaling subset of wave vectors .
this method is much closer to the navier - stokes dynamics than the goy - model , as it contains ( i ) much more modes than goy , ( ii ) it solves the navier - stokes equation for those modes and not only a model equation , and ( iii ) it is three dimensional .
our main results are summarized in figs .
[ goymode - fig ] , [ rewamode - fig ] , and [ goymodf - fig ] . in figs.[goymode - fig ] and [ rewamode - fig ] ,
the amplitude @xmath26 of the energy response is shown as a function of the driving frequency for both the goy model ( fig.[goymode - fig ] ) and the rewa simulation ( fig.[rewamode - fig ] ) with a modulated energy input rate .
this is compared with the results of the mean - field model with the corresponding parameters , i.e , the same modulation amplitude @xmath25 and time scale @xmath1 . in fig.[goymodf - fig ]
the results from the goy model solutions are shown for a modulated driving force and compared with the mean - field model .
in all cases we observe a constant amplitude for low driving frequencies and a @xmath0-decay for high frequencies .
this can in particular be observed in the compensated plots ( parts b of all three figures ) , where @xmath26 , compensated by its asymptotic amplitude , i.e. , @xmath30 , is plotted versus frequency .
the @xmath0-decay of the response means , that for fast modulation no response is detectable any more .
the remaining dissipation rate is that of the stationary system itself . in the mean - field approach ,
a sequence of response maxima is present for both types of forcing , starting at a frequency @xmath53 . in the simulations ,
this main maximum can also be observed , although it is weaker and broader , i.e. , it is `` washed out '' by fluctuations .
the higher order maxima and minima are not visible in the simulations , but are completely washed out by fluctuations . on the other hand , we emphasize that the turbulent fluctuations in the goy model are strongly overestimated due to the extreme mode reduction in this model . in the rewa simulation , an artificially large kolmogorov constant @xmath39 is found .
using such large @xmath39 in the mean - field approach also leads to a considerable weakening of the first maximum and a shrinking of the higher order maxima and minima towards very small amplitudes .
these results will be explained and discussed in detail in this paper , which is organized as follows . in the next section we study the modulated turbulence within the goy shell model . before calculating the response of the system to a modulated energy input rate as well as a modulated forcing in section [ ssecgoymode ] and [ ssecgoymodf ] we briefly introduce the model and study its stationary properties in section [ ssecgoystat ] . in section [ secrewa ]
we present our findings on modulated turbulence within the reduced wave vector set approximation .
we summarize our results in section [ secconcl ] .
the goy shell model consists of a set of coupled odes for one - dimensional complex velocity modes @xmath54 @xcite .
these modes @xmath54 correspond to velocity _ differences
_ @xmath55 on scale @xmath56 .
@xmath57 modes are taken into account , @xmath58 , one complex velocity mode per cascade level @xmath59 , defined by the wave numbers @xmath60 which are equally spaced on a logarithmic scale , here , @xmath61 .
the model equations read : @xmath62 where @xmath63 , @xmath64 , @xmath65 , and @xmath66 .
these are the traditional parameters .
we impose boundary conditions on the @xmath54 , i.e. , @xmath67 for @xmath68 or @xmath69 .
we use @xmath70 shells , a viscosity of @xmath71 , and @xmath72 .
the forcing acts on the largest scale , i.e. , the first shell , @xmath73 .
@xmath74 is constant , @xmath75 .
together with @xmath11 and @xmath23 this sets the time and length units as well as the reynolds number . equations ( [ goy - eq ] )
are integrated using a fourth order runge - kutta scheme with adaptive step size @xcite . with the above chosen parameters
the goy dynamics is chaotic @xcite .
the system is forced on large scales while most of the energy is dissipated on small scales .
it reaches a steady state , in which the velocities are stochastically fluctuating . in this sense
the system has similar properties as three dimensional navier - stokes turbulence .
the scaling behavior of structure functions and dissipation has been extensively studied in @xcite .
the deviations from k41 scaling due to intermittency observed in the goy model are very similar to experimental values .
most of the previous studies have been done with 22 or more shells .
as we use here only 14 shells in order to reduce the computational effort we explicitly check some scaling properties of , e.g. the structure functions and the energy spectrum in a simulation with constant , stationary forcing @xmath76 in equations ( [ goy - eq ] ) .
@xmath70 then turns out to be sufficient .
the reynolds number of the system can be defined as follows .
an outer length scale @xmath6 is given by the smallest wave number @xmath77 , @xmath78 .
a typical velocity @xmath79 is the velocity on that scale , @xmath80 .
the average @xmath81 is taken over time . with
these length and velocity scales the reynolds number of the present simulation is @xmath82 .
the simulated time interval is several hundreds of large eddy turnover times @xmath83 .
( full dots ) corresponding to the second order velocity structure function as a function of wave number @xmath84 , averaged over a long time interval ( @xmath85 ) and with a stationary , constant forcing , @xmath76 . between shell 2 and 9 isr - scaling behavior , @xmath86 ( solid line )
is observed , whereas the shells 12 - 14 form the vsr .
@xmath87 is zero by definition .
the dotted line indicates the middle of the crossover region between the isr and the vsr . ] .
the spectrum is obtained by fast fourier transforming ( fft ) an actual time series of the velocity @xmath88 .
about one frequency decade of kolmogorov scaling @xmath89 is observed .
the @xmath90-space the scaling regime of the spectrum is more extended , i.e. , about three decades , as has been already shown in fig.[scalfig ] . ] for the second order structure function we use the following method . in @xcite
it has been suggested to study the scaling of @xmath91 instead of the pure moments of the velocity @xmath92 in order to eliminate the period 2 and period 3 oscillations which are an artifact of the goy model . the second order quantity @xmath93 , corresponding to the second order structure function , is shown in fig.[scalfig ] as a function of wave number index @xmath94 .
one clearly observes an inertial subrange ( isr ) between shell 2 and 9 , where the second order structure function shows scaling with an exponent near to the k41 value 2/3 , i.e. , @xmath95 corresponding to @xmath96 for the structure function . the scaling exponent is not equal to the k41 value , because the model shows intermittency corrections . the higher wave numbers , i.e. , smaller scales , @xmath97 belong to the viscous subrange ( vsr ) , where dissipation takes place . in this range
the viscosity term is dominant and the velocity decays rapidly with @xmath98 , in fact more than exponentially @xcite .
the external forcing of the flow acts on shell @xmath73 , therefore the stirring subrange ( ssr ) contains the first shell only .
the spectrum , which is obtained by fast fourier transforming ( fft ) a time series of the velocity @xmath88 and raising it to the power two , can be compared to the energy spectrum under the assumption of the taylor hypothesis .
this spectrum is shown in fig.[specfig ] for @xmath70 .
we observe about one frequency decade of ( nearly ) kolmogorov scaling , where the energy decays as @xmath89 with frequency @xmath99 . for @xmath70
the kolmogorov scaling range is not yet well developed . for less shells
this region becomes even narrower and the spectrum shows strong peaks at some intrinsic frequencies . when reducing the number of shells even more , the velocities relax to a stationary value without any fluctuations , i.e. , the chaotic behavior of the system is lost
. in conclusion , the goy model with @xmath70 shells exhibits an inertial subrange scaling , although the isr for the frequency spectrum is only narrow , but the spectrum in @xmath90-space has a scaling range of about three orders of magnitude .
this seems acceptable for our goal to study the response of the system to a modulated driving , because we are only interested in global quantities like the total energy @xmath100 but do not need information on scale resolved quantities .
the time scales in the model have been determined as follows . for each shell
@xmath59 an eddy turnover time @xmath101 is defined by @xcite : @xmath102 this is also considered as the time scale for the turbulent energy transfer through the @xmath59th level .
the time scale relevant for the energy loss on level @xmath59 due to viscosity is defined as @xmath103 both time scales are shown in fig.[taufig ] . in the isr between shell 2 and 9 , where the energy transfer times are the relevant time scales for the dynamics , the decrease of the @xmath101 with @xmath59 is near to @xmath104 ( dashed line in fig.[taufig ] ) as expected for the turnover times of eddies of sizes @xmath105 . in this range
the dissipation time scales @xmath106 are much larger than the @xmath101 , meaning that the turbulent energy transfer is much faster than the viscous dissipation , and therefore the dominant process . in the vsr instead @xmath107 , i.e. , on average the energy is dissipated by viscosity before it can be transferred to smaller scales .
( full dots ) and of the viscous dissipation @xmath108 ( full triangles ) as functions of the level number @xmath59 .
the dashed line is a linear fit of @xmath109 , @xmath110 , and gives @xmath111 .
the shown @xmath101 are obtained by averaging over about 1400 large eddy turnover times .
the dotted line indicates the middle of the crossover region between the isr and the vsr . ]
the largest eddy s turnover time is in general defined by the velocity on the outer length scale @xmath6 , i.e. , on the length scale of the forcing , which in this case is @xmath112 .
however , in this model , the time @xmath113 is disturbed due to finite size effects .
therefore , we extrapolate from the turnover times of the other shells .
a linear fit in fig.[taufig ] for @xmath109 with @xmath110 leads to @xmath114 . without intermittency
one would have @xmath115 ; the small deviation corresponds to the intermittent scaling of @xmath116 or an intermittency correction of @xmath117 .
the extrapolation for @xmath73 yields for the large eddy turnover time @xmath118 .
the time scale corresponding to the energy transfer time @xmath1 used in the mean - field model @xcite is the sum over the eddy turnover times of all energy - input and inertial - range shells , i.e. , here @xmath119 .
the factor @xmath120 between the transfer time and the large eddy turnover time is then @xmath121 . until now ,
we have considered a constant forcing @xmath76 .
the resulting energy input rate @xmath122 then fluctuates around its mean value because of the @xmath123-fluctuations . in the mean - field theory @xcite the energy input rate @xmath124 is constant instead . for closer comparison
we also consider another type of forcing in the goy - model : @xmath125 .
this forcing @xmath126 fluctuates as @xmath127
. then the energy input rate is @xmath128 by definition .
the isr - scaling behavior as well as the energy spectrum then turn out to be similar to the previously discussed ones with the constant forcing @xmath76 .
the energy transfer time @xmath1 is slightly larger in this case , namely @xmath129 , and the large eddy turnover time is @xmath130 .
again , the large eddy turnover time is extrapolated from @xmath131 .
this leads to the factor @xmath20 between the total time delay of the energy cascade and the large eddy turnover time . in the following sections we will study the time - dependent cases where either the energy input rate @xmath124 , i.e. , @xmath132 ( section [ ssecgoymode ] ) or the forcing @xmath76 is modulated ( section [ ssecgoymodf ] ) . in this section
we apply a modulated energy input rate to the goy model , i.e. , we set the forcing @xmath133 in equations ( [ goy - eq ] ) as f(t)=_0(1+et ) [ modein - eq ] with a modulation amplitude @xmath19 . then , the resulting energy input rate @xmath124 is @xmath134 and has a prescribed modulation amplitude @xmath135 by definition .
the total energy of the system @xmath136 is calculated for a wide range of driving frequencies @xmath5 in order to study the frequency behavior of the response .
the brackets @xmath137 denote the ensemble average .
this ensemble average is performed as follows . from a long stationary simulation
we collect an ensemble of 1500 starting configurations which we then let evolve according to eqs.([goy - eq ] ) but now including the modulation of the forcing @xmath126 , eq.([modein - eq ] ) , and average over these 1500 time series . to ensure that the different realizations can be considered as statistically independent ,
the time delay between the successive starting configurations for the different realizations is chosen to be about 100 large eddy turnover times .
the adaptive step size routine controlling the runge - kutta integration algorithm for odes does not produce the same time steps for all time series . to overcome this ,
we have calculated also the same number of equidistant time steps by spline interpolation for all time series .
( dotted lines ) and energy content @xmath138 ( solid lines ) for four different modulation frequencies @xmath5 calculated in the goy model .
the energy input rate @xmath124 is modulated with a modulation amplitude of @xmath139 of the constant energy input rate @xmath140 , @xmath19 , according to eq.([eindef - eq ] ) .
also included is the fit according to eq.([fit - eq ] ) for the energy @xmath138 ( dashed lines , indistinguishable from the solid lines ) .
( a ) @xmath141 , ( b ) @xmath142 , ( c ) @xmath143 , ( d ) @xmath144 . ] the oscillating response of the system @xmath145 is then studied in terms of the ratio between the energy @xmath100 with modulated energy input and the energy @xmath146 without modulation , namely @xmath147 @xmath148 and @xmath149 both are averaged over 1500 realizations . in spite of the averaging not only @xmath148 but also @xmath149 still contains ( weak ) fluctuations .
therefore , we write @xmath146 , as @xmath149 is still slightly fluctuating around its mean value .
@xmath148 and @xmath149 contain about the same size of fluctuations .
accordingly , the energy input rate @xmath124 is normalized by its stationary value , @xmath150 . in fig.[4ffufig ] the input rate @xmath151 and the energy @xmath152
are plotted for four different driving frequencies .
for the two low frequencies where @xmath153 , the energy follows the oscillation of the energy input rate with almost constant , but smaller amplitude . for higher frequencies the amplitude of the deviations of the normalized energy from its stationary value 1 strongly decrease , and a phase shift with respect to the energy input becomes visible .
the same behavior of the energy has been observed in the mean - field theory @xcite .
to quantitatively access the frequency behavior of the response amplitude , we calculated time series of the total energy @xmath100 for 85 different driving frequencies varying over almost 3 decades between @xmath154 .
the chosen frequencies are approximately equally spaced on a logarithmic scale .
the normalized energy @xmath152 is fitted by a function of the form @xmath155 with three free parameters : @xmath156 , the amplitude @xmath26 , and the phase shift @xmath157 .
@xmath156 is near to 1 for all frequencies , i.e. , @xmath158 .
the fits ( [ fit - eq ] ) are included in fig.[4ffufig ] as dashed lines but they are mostly indistinguishable from the solid lines for the energy itself .
fig.[goymode - fig ] of section [ secmotivation ] shows the amplitude @xmath26 , resulting from the fitting procedure , as a function of the dimensionless frequency @xmath159 .
@xmath26 is almost constant for low frequencies and has a value of about @xmath160 .
for higher frequencies the amplitude decreases as @xmath12 .
the same features have been observed in the mean - field calculations , see dashed lines .
the long - dashed line in fig.[goymode - fig ] represents the low frequency limit of the mean - field theory , @xmath27 , and the dotted line the high frequency limit , @xmath48 .
the crossover between the regime of constant amplitude and the one of @xmath0-decay of the energy response takes place at @xmath161 , which is indicated by the arrow in fig.[goymode - fig ] . in the mean - field approach
this crossover is always at @xmath162 , independent of the factor @xmath13 between the large eddy turnover time and the total time scale of the energy transfer . in experiments @xcite
the crossover frequency has been used to measure the energy cascade time scale .
the present simulations confirm that this frequency gives the correct order of magnitude for the energy transfer time .
response maxima , as observed in the mean - field model at frequencies connected with the inverse energy transfer time , are difficult to be identified in fig.[goymode - fig]a .
there is some structure visible at @xmath163 and @xmath164 . in fig.[goymode - fig]b , where the amplitude @xmath26 , compensated by the asymptotic amplitude @xmath165 , i.e. , @xmath30 is plotted versus frequency , this structure becomes more evident , and we see a clear maximum at a frequency of about @xmath164 .
this maximum probably corresponds to the mean - field maximum .
of course , the maximum in goy is broadened and weakened due to the large fluctuations , and the higher order maxima and minima are apparently washed out completely . for non - modulated forcing @xmath132 in the goy model .
strong fluctuations are observed .
inset : probability distribution of @xmath166 .
the mean is @xmath167 , and the width fwhm/2=0.46@xmath13 giving about 50% fluctuations . ] as in the mean - field theory no fluctuations are included , the energy cascade time @xmath1 is considered to be constant .
however , in the goy model this assumption is not true , as can clearly be seen in fig.[tauwidthfufig ] . here , a time series of the cascade time @xmath168 is plotted , computed within the goy - shell model with non - modulated forcing @xmath169 .
the inset shows the probability distribution of @xmath166 .
this distribution has its maximum at @xmath170 , almost at the mean cascade time @xmath167 , and a width ( fwhm/2 ) of about @xmath171 .
the width is almost half the size of the mean which indicates that the transfer time fluctuates strongly and therefore we have to expect that the response maxima are more or less washed out .
however , these strong fluctuations are considered as an artifact of the goy - model and not as a feature of real turbulence .
the goy - model contains only one velocity mode per cascade level instead of infinitely many modes in real turbulence .
this one - mode approximation leads to an overestimation of the fluctuation strength . in order to confirm this
, we performed another simulation with more modes per level within the reduced wave vector set approximation ( rewa ) of the navier - stokes equation .
this will be presented in section [ secrewa ] . in this section
we present further results within the goy model based on a non - fluctuating driving force @xmath24 which is regularly modulated as was @xmath140 in the previous section
. this case may be more comparable to the experimental method in reference @xcite , because there the driving force is modulated . what can not be modeled with goy
is the spatial inhomogeneity in the experiments @xcite . in equations ( [ goy - eq ] )
we now apply a forcing @xmath172 with a modulation amplitude of @xmath43 . as in section [ ssecgoymode ]
we calculate the ensemble averaged time series of the energy input rate @xmath173 , see first line of equation ( [ eindef - eq ] ) , and the total energy of the system @xmath100 cf .
equation ( [ edef - eq ] ) , for 89 different frequencies between @xmath174 , again logarithmically equally distributed . the normalized energy @xmath152 and energy input rate @xmath151
are then fitted by a function according to equation ( [ fit - eq ] ) , with the parameters @xmath156 , @xmath26 , @xmath157 , and @xmath175 , @xmath41 , @xmath176 , respectively .
the amplitudes @xmath26 and @xmath41 are plotted in fig.[goymodf - fig ] of section [ secmotivation ] as a function of the dimensionless frequency @xmath177 .
also in this case of a modulated force , the response amplitude is almost constant for small frequencies , namely @xmath47 , and decreases as @xmath0 for high frequencies , see fig.[goymodf - fig]a ( full dots ) .
again , the long - dashed line represents the low frequency limit of the mean - field theory for @xmath26 ( which is @xmath47 in this case ) as well as the dotted line the high frequency limit . as in the previous section , the crossover frequency between the two regimes
is determined by the energy transfer time , i.e. , @xmath178 ( as @xmath121 in this case ) , which is marked by the small arrow in fig.[goymodf - fig]a . the amplitude of the energy input rate @xmath41 starts with a value of about @xmath179 for low frequencies and merges towards @xmath180 for high frequencies , see fig.[goymodf - fig]c .
this indicates that at very large frequencies the velocity is not oscillating any more as it only feels a mean constant force .
the oscillations of the energy input rate are then only a consequence of the oscillation of the driving force @xmath74 . in the mean - field theory
we have observed the same trend for both amplitudes . the corresponding mean - field results are included as dashed lines . both amplitudes @xmath26 and
@xmath41 show a maximum at a frequency near to the crossover frequency @xmath181 . in the compensated plot fig.[goymodf - fig]b , where @xmath30 is plotted as a function of frequency , a clear deviation from the dotted line representing @xmath182 can be observed . at this frequency , the mean - field theory predicts a first maximum for the energy response amplitude ( fig.[goymodf - fig]a ) and a maximum directly followed by a minimum for the energy input rate , see fig.[goymodf - fig]c .
this frequency is connected to the energy transfer time @xmath1 . in the mean - field model
further maxima of the energy response and wiggles of the energy input rate are observed at multiples of this frequency .
however , in the goy model all further maxima and minima are washed out presumably because of the strong fluctuations .
the fluctuations of the energy cascade time scale are found to be similar as in the case of a fluctuating force shown in fig.[tauwidthfufig ] in the previous section .
as was pointed out in the introduction of this chapter a full numerical simulation of the navier - stokes equation for modulated turbulence at high reynolds numbers is still not possible or requires low reynolds numbers .
therefore , we first have considered the goy shell model .
this model correctly describes many features of turbulence , however , due to the one - mode approximation in each cascade level , it contains various artifacts .
namely , it strongly overestimates the strength of the fluctuations .
the aim of this section is to study the problem of modulated turbulence within another model , the reduced wave vector set approximation ( rewa ) @xcite , which is much closer to the navier - stokes equation than the goy model and contains much more modes per cascade level .
of course , as compared to full numerical simulations of the navier - stokes equation , it still contains a mode reduction in order to make the computational effort reasonable for the desired high reynolds numbers .
the present approximation has been introduced and extensively studied in @xcite . here , we use it together with a time - dependent driving . for completeness
we briefly explain the approximation before we present the results with modulated driving .
the velocity field @xmath183 is fourier transformed into plane waves , @xmath184 .
periodic boundary conditions are applied on a periodicity volume @xmath185 .
the wave vectors @xmath186 are given by @xmath187 , with @xmath188 . in order to efficiently deal with the large number of modes involved
, the reduced wave vector set approximation selects a limited number of modes by admitting only a geometrically scaling subset @xmath189 of wave vectors , i.e. , @xmath190 . on this subset @xmath191 the navier - stokes equation for incompressible flow , [ rewa - eq ] u_i(_n^(l))&=&-(_n^(l))^2u_i(_n^(l))+f_i(_n^(l ) ) + -i&m_ijk&(_n^(l))__1,_2k,_1+_2=_n^(l ) u_j(_1)u_k(_2 ) , together with the continuity equation , _
n^(l)(_n^(l))=0 , is solved .
@xmath192 is the coupling matrix , @xmath193 , where @xmath194 is the orthogonal projector to @xmath186 .
the subset @xmath195 consists of a basic subset @xmath196 together with its scaled replicas @xmath197 , @xmath198 . in the present simulation
we take @xmath199 wave vectors @xmath200 , k_0=&&\{(2,2,2),(-1,2,2),(-2,1,1),(3,0,0 ) , + & & ( 4,1,1),(4,-2,1),(-3,3,3),(-5,1,1 ) , + & & ( 4,4,1),(3,3,0),(1,1,10),(-10,5,5 ) + & & + } these wave vectors @xmath201 are chosen such that they span a wide range of length scales , but still dynamically interact to a good degree .
for the navier - stokes equation this means that as many navier - stokes interactions @xmath202 as possible between the wave vectors in @xmath195 are allowed for .
the largest eddies of the order @xmath203 are represented by the wave vectors in @xmath204 , whereas the subsets @xmath205 contain wave vectors of smaller and smaller eddies .
the choice of the smallest eddies , i.e. , the value of @xmath206 , depends on the kinematic viscosity @xmath11 . @xmath206 and @xmath11 are adjusted such that the velocity amplitudes @xmath207 of the smallest eddies are almost zero . in this simulation
, @xmath11 is chosen as @xmath208 and the number of levels as @xmath209 . to maintain the turbulent flow
we apply a forcing as in @xcite : [ rewaforce - eq ] ( , t)= \ { cc _ 0(1+et ) , & k_in + 0 , & k_in . .
the subset @xmath210 of @xmath204 by choice contains the wave vectors with the three smallest lengths .
@xmath210 contains 14 vectors , namely @xmath211 . in reference
@xcite it has been shown that the statistics of the solutions of the equations of motion do not depend on the particular choice of @xmath204 .
this forcing corresponds to the same type of forcing , which has already been applied to the goy model in section [ ssecgoymode ] .
it enforces the energy input rate to be modulated : [ einrewa - eq ] e_in=__n^(l)ku^(_n^(l))(_n^(l))=_0(1+et ) .
eq.([rewa - eq ] ) is a set of @xmath212 coupled odes for the complex mode amplitudes @xmath213 which is numerically solved within the burlisch - stoer integration scheme with adaptive step size @xcite .
length scales are measured in units of @xmath203 and time scales in units of @xmath38 .
a reynolds number can be defined as follows : the wave length @xmath214 of the smallest wave vector gives an external length scale @xmath215 , and a typical velocity on that scale is determined by the rms of one velocity component , @xmath10 .
then , in our case , the reynolds number is @xmath216 because from the simulations we obtain @xmath217 .
the main features of fully developed turbulence as irregular velocity signals , characteristic scaling of structure functions , etc .
are well described within this approximation , as has been shown in @xcite .
the rewa solutions show small scale intermittency , which is produced by the competition between down - scale energy transport and viscous dissipation on the small scales @xcite .
other mechanisms leading to intermittency in turbulence as e.g. nonlocal interactions between wave vectors are underestimated in this approximation @xcite .
the down - scale energy transport in the rewa fluid is less effective than in real turbulence , because in this approximation the larger wave vectors are more and more thinned out @xcite .
this is in contrast to the case of the complete set of wave vectors ( e.g. in full grid simulations ) where the density of states increases @xmath218 , whereas in the reduced wave vector set @xmath195 the number of admitted wave vectors decreases as @xmath219 @xcite . in reference
@xcite it has been shown that this reduced energy transport leads to an overestimation of the taylor reynolds number of the system as well as the kolmogorov constant @xmath39 , defined by @xmath220 , by roughly one order of magnitude for our choice of @xmath57 . in the present simulation
we obtain @xmath33 instead of @xmath221 as in experiments @xcite .
since @xmath222 is the energy density @xmath223 of the fluctuations in the fluid system , the large @xmath39 value indicates that even in the rewa approximation the strength of the fluctuations is highly overestimated .
the large kolmogorov constant will change the relevant time scales in the system , as will be shown in section [ ssecrewamode ] .
the characteristic time scale for the turbulent energy transfer on scale @xmath224 can be estimated as : [ taurewadef - eq ] ( l)= , where @xmath225 denotes the mean wave number on scale @xmath224 , i.e. , it is the mean inverse eddy size in @xmath205 . as in the goy model , eq.([taudissdef - eq ] ) , the time scale of viscous dissipation is @xmath226^{-1}$ ] .
again , in the isr @xmath227 , whereas in the vsr @xmath228 . from a simulation with stationary forcing ,
i.e. , @xmath229 in eq .
( [ rewaforce - eq ] ) , the time delay of the energy down - transport @xmath230 is then estimated by the sum of all @xmath231 in the isr , @xmath232 .
the largest of these @xmath231 , on the largest scale , @xmath233 can be regarded as a large eddy turnover time @xmath14 .
thus , @xmath234 , and the factor between the cascade time scale and the large eddy turnover time is @xmath235 . in rewa .
inset : probability distribution of @xmath236 .
the fluctuations are considerably smaller than in the goy model .
note the different scales in this figure and fig.[tauwidthfufig ] .
the mean is @xmath235 , and the width @xmath237 giving about 2% fluctuations . ] as we have seen in the goy model , fig.[tauwidthfufig ] , the energy transfer time is strongly fluctuating .
we attributed these strong fluctuations to the one - mode per level approximation of the goy model .
fig.[tauwidthrewafig ] shows a time series of the energy transfer time @xmath238 ( use eq.([taurewadef - eq ] ) with @xmath239 ) in the present reduced wave vector set approximation ( rewa ) , and , in the inset , the distribution of this time scale
. clearly , the fluctuations are much weaker than in the goy model ; they are zero in the mean - field approximation .
the distribution is centered around * @xmath240 * with a width ( fwhm/2 ) of about * @xmath241*. the response of the system to a modulated driving force , cf .
eq.([rewaforce - eq ] ) , is calculated now in terms of the total energy of the system [ edefrewa - eq ] e(t)=_l=0^l = l_max_k_l|(,t)|^2 .
the modulation amplitude of the energy input rate @xmath124 ( eq.([einrewa - eq ] ) ) is chosen as @xmath31 .
the average @xmath137 is performed as follows .
we average over 25 to 30 realizations , which are obtained from eqs.([rewa - eq ] ) with different starting values .
the delay between the different starting values is one driving period . for the higher frequencies the period @xmath242 becomes too small to ensure statistical independence of the different realizations .
then , we chose the delay between the successive starting values of the different realizations to be several driving periods such that it is at least @xmath243 .
the response is calculated for 150 ( approximately equally spaced on a logarithmic scale ) frequencies between * @xmath244*. the energy is normalized by @xmath149 , calculated from a stationarily forced solution with @xmath229 and averaged as @xmath148 . the oscillating response @xmath145 of the system
is defined in the same way as for the goy model , see eq.([goyresp - eq ] ) . then , as for the goy calculations , the averaged and normalized signals @xmath152 are fitted with eq.([fit - eq ] ) .
the fit parameter @xmath156 is again near to 1 for all frequencies , @xmath245 . in fig.[4frewafig ]
the time averaged responses and the normalized energy input rates are plotted for four different driving frequencies . also the fits according to eq.([fit - eq ] ) have been included as dashed lines but are indistinguishable from the solid lines for the energy signal itself .
( dotted lines ) and energy @xmath138 ( solid lines ) for four different modulation frequencies @xmath5 as calculated in the rewa simulation .
the energy input rate is modulated with a modulation amplitude of @xmath246 of the constant energy input rate , i.e. , @xmath31 in eq.([rewaforce - eq ] ) .
also included is the fit to the energy data cf .
eq.([fit - eq ] ) as dashed lines but these are in all cases indistinguishable from the solid lines .
the averaged time series of @xmath138 are repeated once for better visibility . ( a )
@xmath247 , ( b ) @xmath248 , ( c ) @xmath249 , ( d ) @xmath250 . for larger @xmath159
the energy is indistinguishable from 1 on this scale .
the crossover to the @xmath0-decay regime is in this simulation at @xmath251 , between the frequencies of b ) and c ) . ]
we observe in fig.[4frewafig ] for rewa the same features as in fig.[4ffufig ] for the goy model and in fig.1 of reference @xcite for the mean - field model .
for the two lower frequencies the response amplitude remains almost constant and is about @xmath160 of the amplitude of the energy input rate , whereas for the two higher frequencies the response amplitude strongly decreases . this trend becomes more clear in fig.[rewamode - fig]a in section [ secmotivation ] , where the amplitude of the response determined from the fit ( [ fit - eq ] ) is shown as a function of the driving frequency ( full dots ) . for low driving frequencies @xmath27 , whereas for high frequencies
the amplitude decreases as @xmath0 .
the crossover between the regime of constant amplitude and the one of @xmath0-decay takes place at @xmath252 , i.e. , at a much smaller frequency than expected from the original case of the mean - field theory in which the crossover was at @xmath253 with @xmath254 .
we understand this as follows . in section [ ssecrewadescribe ]
it was mentioned that the kolmogorov constant in the rewa simulation is @xmath33 instead of @xmath221 as in experiments . in the figures of the mean - field approach @xcite
we have set @xmath254 .
the mean - field solution for a general @xmath39 revealed that the crossover frequency decreases with increasing @xmath39 while the positions of the response maxima are left unchanged .
for @xmath33 the mean - field crossover frequency is at @xmath255 in close agreement to what we observe in the rewa simulations .
the response amplitude calculated from the mean field model with @xmath33 is included in fig.[rewamode - fig]a as dashed line .
apart from the changed crossover frequency we observe that , in the mean - field calculations , the first response maximum at @xmath256 is considerably smaller and broadened as compared to the case with @xmath254 , given as the dashed line in fig.[goymode - fig]a . in agreement with this , our rewa simulations ( with a value @xmath33 ) show a broad maximum in the response amplitude at @xmath257 .
this means it occurs at a similar frequency as the mean field model .
the maximum becomes more clear in the compensated plot , fig.[rewamode - fig]b , where @xmath30 is shown as a function of frequency .
there , we observe a deviation from the @xmath0-decay of the amplitude by a factor 1.4 at the maximum in the rewa simulations .
the mean field maximum has a height of 2.8 .
the subsequent maxima and minima in the mean field model occur at frequencies where the amplitude is already very small ( @xmath258 ) because the crossover to the @xmath0-regime takes place at a much smaller frequency whereas the response maxima stay at the same frequencies as for a smaller @xmath39 .
therefore , the higher order maxima are not visible in the rewa simulations .
the cascade time shows about 2% fluctuations in rewa as shown in fig .
[ tauwidthrewafig ] . however , at small response amplitudes these fluctuations are already large enough to wash out the higher order response maxima . in conclusion ,
the rewa system reproduces qualitatively the features of modulated turbulence as predicted by the mean - field model including the first response maximum .
the latter is considerably weakened due to the large kolmogorov constant in rewa .
another consequence of the large @xmath39 is , that the crossover between constant amplitude and @xmath0-decay is shifted towards smaller frequencies , and therefore the higher order maxima and minima are already at very small amplitudes where the fluctuations in the cascade time scale are finally large enough to wash them out .
we can not clarify , at present , how close the response in direct numerical simulations which lead to an order of magnitude
smaller @xmath39 and thus have much smaller fluctuations will come to the mean - field features , but we expect a clearly visible first maximum at least .
we have simulated the response of modulated turbulence within two numerical models .
namely , we have used the goy shell model and the reduced wave vector set approximation of the navier - stokes equation ( rewa ) .
the results are compared with predictions from a mean - field theory . for a modulated energy input rate
this mean - field theory had predicted a constant response amplitude for low frequencies and a @xmath0-decay for high driving frequencies .
in addition , at certain frequencies connected with the energy cascade time scale , a sequence of maxima and minima of the response amplitude is observed . both numerical models
well reproduce the basic trend , i.e. , the constant amplitude for small @xmath5 and the @xmath0-decrease for large @xmath5 .
the main response maximum can be observed in both numerical models , although it is weakened due to fluctuations .
the higher order maxima and minima as predicted by the mean - field theory can not be identified in the simulations . obviously , they are strongly washed out by fluctuations . in the goy model
the large fluctuations are explicitly visible in a broad cascade time distribution .
we attribute these fluctuations to the strong mode reduction in the model , i.e. , they are an artifact of the model properties and not a feature of real turbulence . in the reduced wave vector set approximation of the navier - stokes equation these fluctuations
are much weaker , and , we believe , more realistic for real turbulence , as more modes are taken into account . however , due to an overestimated kolmogorov constant @xmath39 in rewa the higher order maxima are considerably reduced , and therefore washed out by the fluctuations although being smaller than the goy - fluctuations
. therefore , we believe , that in real turbulence with a realistic kolmogorov constant the first maximum should be clearly observable and possibly also the higher order extrema in the response . thus , the predictions of the mean - field model , which excludes _ all _ fluctuations , might be quite reasonable for real turbulence . to further study the response maxima numerically ,
it is necessary to perform full numerical simulations of the navier - stokes equation , as then all relevant time scales including their fluctuations are reproduced realistically , which turned out to be essential for the observation of the response maxima .
recent experiments on modulated turbulence @xcite revealed evidence for the response maxima . these experiments may be more comparable to the case of a modulated _ force _ instead of a modulated energy input rate .
we have studied this case also within the mean - field model and have found basically the same behavior of the energy response as for a modulated energy input rate .
in addition , the amplitude of the energy input rate showed `` wiggles '' at the same frequencies where the energy response had maxima . in the experiments
the response maxima were measured in the energy input rate , which can be regarded as a response of the system as well in this case .
also the constant amplitude for low driving frequencies and the @xmath0-decay of the velocity response which in leading order is corresponding to a @xmath0-decay of the energy response , as well have been observed in the experiments .
here , we have studied the case of a modulated driving force within the goy shell model .
also in this simulation , the response amplitude behaves basically as in the mean - field model , i.e. , it decreases as @xmath0 .
the energy response amplitude as well as the amplitude of the energy input rate show the main maximum .
due to the above mentioned large fluctuations all higher order maxima are washed out in the goy model .
there are two regimes in the frequency behavior of the response amplitude , namely a constant amplitude at low @xmath5 and a decreasing amplitude at high @xmath5 .
the present simulations give further confidence that the crossover frequency between these two regimes gives the correct order of magnitude of the cascade time scale , i.e. , in experiments it can be used to measure this time scale as suggested in reference @xcite .
both models in the present study were able to reproduce the main features of the frequency behavior of the response amplitude in modulated turbulence as predicted by the mean - field model , however , both also have their shortcomings , which prevent us from correctly predicting the behavior of real turbulence in all quantitative details .
therefore , we believe , that it is worth to further study modulated turbulence numerically as well as experimentally .
* acknowledgments : * the work is part of the research program of the stichting voor fundamenteel onderzoek der materie ( fom ) , which is financially supported by the nederlandse organisatie voor wetenschappelijk onderzoek ( nwo ) .
this research was also supported by the german - israeli foundation ( gif ) and by the european union under contract hprn - ct-2000 - 00162 . | numerical simulations of fully developed turbulence driven by a modulated energy input rate or driving force are performed within two dynamical cascade models , the goy shell model and a reduced wave vector set approximation of the navier - stokes equation ( rewa ) .
the frequency behavior of the system response is studied and compared with predictions from a variable range mean - field theory , which excludes turbulent fluctuations . in agreement with the mean - field approach
we find a constant response amplitude for low driving frequencies and a @xmath0-decay of the amplitude for high frequencies . in the mean - field theory ,
the finite cascade time scale had lead to an oscillating behavior of the response amplitude as a function of the driving frequency . in the simulations of both models
we observe the main maximum .
the higher maxima and minima are completely washed out by fluctuations .
u v |
over the past decade or so , one of the most exciting advances in high energy astrophysics has been the detection of sources at tev energies with ground - based gamma ray facilities ( see weekes 2003 for a recent review ) .
among the sources detected , blazars are arguably the most intriguing .
they represent the only type of active galactic nuclei ( agn ) that has been detected at tev energies ( although a 4-@xmath1 detection of m87 has been reported ; aharonian et al .
2003 ) . to date , there are a total of six firmly established tev blazars . the emission from a blazar is generally thought to be dominated by radiation from a relativistic jet that is directed roughly along the line of sight ( review by urry & padovani 1995 and references therein ) .
relativistic beaming is necessary to keep gamma - ray photons from being significantly attenuated by the surrounding radiation field ( via photon - photon pair production ) .
the spectral energy distribution ( sed ) of tev blazars invariably shows two characteristic peaks in the @xmath2 representation , with one located at x - ray energies and the other at tev energies ( fossati et al .
there seems to be a general correlation between the two sed peaks as the source varies ( e.g. , buckley et al . 1996 ; catanese et al .
1997 ; maraschi et al . 1999 ; petry et al .
2000 ) .
a popular class of models associates the x - ray emission from a tev blazar with synchrotron radiation from highly relativistic electrons in the jet and the tev emission with inverse - compton scattering of the synchrotron photons by the electrons themselves ( i.e. , synchrotron self - compton or ssc for short ; marscher & gear 1985 ; maraschi et al .
1992 ; dermer et al .
1992 ; sikora et al . 1994
; see bttcher 2002 for a recent review ) .
the ssc models can , therefore , naturally account for the observed x - ray
tev correlation .
moreover , they have also enjoyed some success in reproducing the measured seds .
however , the models still face challenges in explaining some of the observed phenomena , such as the presence of `` orphan '' tev flares ( krawczynski et al . 2004 ; cui et al .
2004 ) .
alternatively , the jet might be energetically dominated by the magnetic field and it is the synchrotron radiation from highly relativistic protons that might be responsible for the observed tev gamma rays ( aharonian 2000 ; mcke et al .
other hadronic processes have also been considered , including photo - meson production , neutral pion decay , and synchrotron - pair cascading ( e.g. , mannheim & biermann 1992 ; mcke et al .
2003 ) , but they are thought to be less important in tev blazars ( aharonian 2000 ; mcke et al .
another class of hadronic models invokes @xmath3 processes , for instance , in the collision between the jet and ambient `` clouds '' ( e.g. , dar & laor 1997 ; beall & bednarek 1999 ) or inside the ( dense ) jet ( pohl & schlickeiser 2000 ) . in this case , the gamma - ray emission is mainly attributed to the decay of neutral pions produced in the @xmath3 interactions . in both classes of hadronic models ,
the emission at x - ray and longer wavelengths is still attributed to the synchrotron radiation from relativistic electrons ( and positrons ) in the jet , as in the ssc models . although the hadronic models may also be able to describe the observed sed of tev blazars and accommodate the x - ray
tev correlation , they are generally challenged by the most rapid gamma - ray variabilities observed in tev blazars ( gaidos et al . 1996 ) .
tev blazars are also known to undergo flaring episodes both at x - ray and tev energies .
the flares have been observed over a wide range of timescales , from months down to less than an hour .
the observed x - ray flaring hierarchy seems to imply a scale - invariant physical origin of the flares ( cui 2004 ; xue & cui 2005 ) .
blazar flares are thought to be related to internal shocks in the jet ( rees 1978 ; spada et al .
2001 ) , or to the ejection of relativistic plasma into the jet ( e.g. , bttcher et al . 1997 ; mastichiadis & kirk 1997 ) .
recently , it is suggested that the flares could also be associated with magnetic reconnection events in a magnetically dominated jet ( lyutikov 2003 ) and thus they could be similar to solar flares in this regard .
such a model might offer a natural explanation for the hierarchical flaring phenomenon , again in analogy to solar flares . to make further progress on distinguishing the emission models proposed for tev blazars
, we believe that a large amount of simultaneous or contemporaneous data is critically needed over a wide range of flux , especially in the crucial x - ray and tev bands , for quantifying the sed and spectral variability of a source and for allowing investigations of such important issues as variability timescales , cross - band correlation , spectral variability , spectral hysteresis , etc .
such data are severely lacking at present , despite intense observational efforts over the years . in this paper
, we present results from an intensive multi - wavelength monitoring campaign on mrk 421 .
this source is the first tev blazar discovered ( punch et al .
1992 ) and remains one of the few blazars that can be detected at tev energies nearly all the time with ground - based imaging atmospheric cherenkov telescopes ( iacts ) .
some of the preliminary results have appeared elsewhere ( cui et al . 2004 ) ; they are superseded by those presented in this work .
we have assumed the following values for the various cosmological parameters : @xmath4 , @xmath5 , and @xmath6 .
the corresponding luminosity distance of mrk 421 ( @xmath7 ) is about 129.8 mpc .
from 2003 february to 2004 june , mrk 421 was observed at tev energies with the whipple 10 m telescope ( on mt .
hopkins , az ) during each clear night within the dark moon observing periods .
the typical exposure time of a nightly observation was 28 minutes , corresponding to one observing run , but more runs were taken on occasion , especially near the end of the campaign ( in 2004 april ) when the source was seen to undergo an usually large x - ray outburst as seen by the all - sky monitor ( asm ) on _ rxte_. to achieve simultaneous coverages of the source both in the tev and x - ray bands , we communicated with the _ rxte science operations facility _ to ensure that the _ whipple _ and _ rxte _ observing schedules were matched as closely as possible .
a total of 306 runs were collected in good ( empirically designated as `` a '' or `` b '' ) weather , and roughly 80% of them were taken at zenith angles @xmath830 . the procedure for reducing and analyzing _ whipple _ data has been standardized over the years ( hillas 1985 ; reynolds et al . 1993 ; mohanty et al . 1998 ) .
a detailed description of the current hardware can be found in finley et al .
( 2001 ) . for clarity
, we briefly summarize a few key points that are relevant to this work .
the success of iacts lies in the fact that the cherenkov images of an air shower produced by a gamma - ray primary have different shapes and orientations than those found in an air shower produced by cosmic - ray particles ( mostly protons ) .
the images in gamma - ray events are typically more compact and are more aligned to point towards the position of the source than those in cosmic - ray events . in practice ,
the image shape of an event is characterized by the major and minor axes of the best - fit ellipse to the image .
the orientation of the ellipse is characterized by the parameter `` @xmath9 '' , defined as the angle between the major axis and the line connecting the center of the ellipse to the center of the field - of - view ( fov ) .
we have developed standard selection criteria based primarily on the image shape and orientation parameters that remove over 99% of the cosmic ray events while keeping about half of the gamma - ray events ( see , e.g. , falcone et al .
2004 for more details ) .
the _ whipple _ observations are conducted in one of the two modes : tracking and on / off . in the tracking mode , the telescope tracks the target across the sky so that the source stays at the center of the fov throughout the observation . in the on / off mode , on the other hand , the telescope tracks the target only during the on run ; it is then offset by 30 minutes in right ascension during the off run , and tracks the field as it covers the same range of zenith and azimuthal angles .
the off run provides a direct measurement of the background , although the difference in the sky brightness of the fields between the on and off runs must be taken into account by using a technique known as software padding ( reynolds et al 1993 ) . for tracking observations ,
the background is derived from the @xmath9 histogram of the events that have passed all but the orientation cuts . since real gamma - ray events from a source should all be concentrated at small @xmath9 values ( @xmath10 15 ) for on - axis observations , the background level can be estimated from events with larger @xmath9 values ( 2065 ) , if the ratio of the number of background events in the two ranges of the @xmath9 parameter is known .
this ratio is derived from observations of fields in which there is no evidence for a gamma - ray source ( see , e.g. , falcone et al .
2004 ) .
nearly all of the _ whipple _ observations reported in this work were carried out in the tracking mode .
we followed the standard procedure just described to obtain count rates from individual runs ( taken in the good weather ) and thus the long - term light curves . to examine variability on short timescales ,
we sometimes sub - divided runs into time intervals shorter than the nominal 28 minutes and constructed light curves with correspondingly smaller time bins . to correct for the effects due to changes in the zenith angle and overall throughput of the telescope
, we applied the method developed by lebohec & holder ( 2003 ) to the data .
we should note that for a given season the corrections were made with respect to a ( somewhat arbitrarily chosen ) reference run taken at 30 zenith angle during a clear night . to correct for changes in the telescope throughput across seasons
, we used the measured rates for the crab nebula to further calibrate the light curves . for reference ,
the rate of the crab nebula is about 2.40 and 2.93 @xmath11 for the 2002/2003 and 2003/2004 seasons , respectively .
the spectral analysis was carried out by following method 1 described in mohanty et al .
the technical aspects and difficulties involved in finding matching pairs for the spectral analysis of tracking observations are explained in detail by petry et al .
( 2002 ) and daniel et al .
combined with the fact that the observations were taken only in the tracking mode , poor statistics made it extremely challenging to derive a tev spectrum from observations taken at low fluxes .
for those cases , the tolerance for the parameter cuts was tightened to be just 1.5 standard deviations from the average value of the simulations , as opposed to the usual 2 standard deviations , in order to further reduce the cosmic - ray background ( daniel et al .
the tighter cuts still retained @xmath12 of gamma rays .
the downside is that the effective collection area of the telescope is not as independent of energy as can be ideally hoped for ( mohanty et al .
1998 ) . in coordination with each _
observation , we took a snapshot of mrk 421 at x - ray energies with _ rxte _ ( with a typical exposure time of 23 ks ) .
we should note , however , that not every x - ray observation was accompanied by a simultaneous _ whipple _ observation ( due , e.g. , to poor weather ) and vice versa ( especially near the end of the campaign , when many more _ whipple _
observations were made to monitor the source in an exceptionally bright state ; see fig . 1 ) .
for this work , we only used data from the pca instrument on _ rxte _ , which covers a nominal energy range of 260 kev .
the pca consists of five nearly identical proportional counter units ( pcus ) .
however , only two of the pcus , pcu 0 and pcu 2 , were in use throughout our campaign , due to operational constraints .
pcu 0 has lost its front veto layer , so the data from it are more prone to contamination by events caused by low - energy electrons entering the detector .
the problem is particularly relevant to variability studies of relatively weak sources , such as mrk 421 . for this work , therefore
, we have chosen pcu 2 as our `` standard '' detector for flux normalization and spectral analysis .
we followed cui ( 2004 ) closely in reducing and analyzing the pca data .
briefly , the data were reduced with _
ftools 5.2_. for a given observation , we first filtered data by following the standard procedure for faint sources , which resulted in a list of good time intervals ( gtis ) .
we then simulated background events for the observation by using the latest background model that is appropriate for faint sources .
using the gtis , we proceeded to extract a light curve for each pcu separately .
we repeated the steps to construct the corresponding background light curves from the simulated events .
we then subtracted the background from the total to obtain the light curves of the source . following a similar procedure , for each observation
, we also constructed the x - ray spectrum for each pcu and its associated background spectrum . in this case , however , we only used data from the first xenon layer of each pcu ( which is most accurately calibrated ) , which limits the spectral coverage to roughly 2.525 kev . since few counts were detected at higher energies , the impact of the reduced spectral coverage is very minimal .
the optical data were obtained with the fred lawrence whipple observatory ( flwo ) 1.2 m telescope ( located adjacent to the whipple 10 m gamma - ray telescope on mt .
hopkins ) and with the 0.4 m telescope at the boltwood observatory in stittsville , ontario , canada .
we note that we had no optical coverage of the source during the 2002/2003 _ whipple _ observing season .
the flwo 1.2 m was equipped with 4shooter ccd mosaic with four @xmath13 chips .
each chip covers a @xmath14 fov .
the data were collected during 31 nights , from 2003 december 14 to 2004 february 17 .
a total of 77 images were obtained in the @xmath15 band , 69 in the @xmath16 band , 67 in the @xmath17 band , and 78 in the @xmath18 band , with an exposure time of 30 s for each image .
the preliminary processing of the ccd frames was performed with the standard routines in the iraf ccdproc package .
photometry was extracted using the daophot / allstar package ( stetson 1987 ) .
the fitting radius for profile photometry was varied with the seeing .
the median seeing in our dataset was 4.5 . a detailed description of the applied procedure
is given in
3.2 of mochejska et al .
the derivation of photometry was complicated by the proximity of two very bright stars , hd 95934 ( v=6 mag ) and hd 95976 ( v=7.5 mag ) , at angles of 2.1 and 4 from mrk 421 , respectively .
these angles correspond to distances of 187 and 356 pixels on the images .
the presence of these stars introduces a gradient in the background , approximately in the north - south direction . to estimate the magnitude of this gradient , we examined two @xmath19 pixel regions located at distances of 3040 pixels north and south of mrk 421 .
these regions were chosen to coincide with the annulus of 18 - 45 pixels used by allstar for background determination .
the difference in counts between the two regions is at a level of 1.4% , 0.7% , 0.3% , and 0.2% of the peak value of mrk 421 with bvri filters , respectively , on our best seeing images , and of 3.9% , 1.9% , 1.1% , and 0.6% on images with seeing close to the upper @xmath20 percentile .
thus , it varies with the seeing in bvri by 2.5% , 1.2% , 0.8% , and 0.4% , respectively , all very small compared with the variability amplitudes of the source .
the transformations of instrumental magnitudes to the standard system were derived from observations of 27 stars in 3 landolt ( 1992 ) standard fields , collected on 19 january 2004 .
the systematic errors associated with the calibration are estimated to be 0.02 for bvi and 0.01 for r. the boltwood 0.4 m is equipped with an apogee ap7p ccd camera that uses a back - illuminated site 502a chip .
mrk 421 was observed during the period from 2003 november 8 to 2004 june 11 .
the data were taken with an uncalibrated cousins r filter ( designed by bessell and manufactured by omega ) .
the photometric measurements are differential ( with respect to the comparison star 4 in villata et al .
aperture photometry was performed with custom software .
the aperture used is of 10 in diameter .
data points were obtained from averaging over between four and six 2 minute exposures .
the typical statistical error on the relative photometry of each data point is 0.02 in magnitude .
the seeing - induced errors or other systematic errors were not taken into account . to cross check results from the two sets of measurements , we scaled up the boltwood values ( by adding a constant to the measured differential magnitudes ) so that they agree , on average , with the flwo fluxes for the overlapping time period .
we found that the observed variation patterns agree quite well between the two data sets .
this also implies that any variability caused by systematic effects must be small compared with the intrinsic variability of the source .
we observed mrk 421 frequently with the 26-meter telescope at the university of michigan radio astronomy observatory ( umrao ) and the 13.7-meter metshovi radio telescope at the helsinki university of technology .
the umrao telescope is equipped with transistor - based radiometers operating at center frequencies of 4.8 , 8.0 , and 14.5 ghz ; their bandwidths are 560 , 760 , and 1600 mhz , respectively .
all three frequencies utilize rotating , dual - horn polarimeter feed systems , which permit both total flux and linear polarizations to be measured using an on - off observing technique at 4.8 ghz and an on - on technique ( switching the target source between the two feed horns closely spaced on the sky ) at the other two frequencies .
the latter technique reduces the contribution of tropospheric interference by an order of magnitude . observations of mrk 421 were intermixed with observations of a grid of calibrator sources : 3c 274 is the most frequently used calibrator , except during a period each fall when the sun is within 15@xmath21 of the calibrator .
the flux scale is set by observations of cas a ( e.g. see baars et al .
details of the calibration and analysis techniques are described in aller et al .
( 1985 ) .
the metshovi observations were carried out as part of a long - time monitoring program .
the observations were made with dual - horn receivers and the on - on technique , and they covered the 22 and 37 ghz bands .
the flux measurements have been calibrated against dr 21 ( with the adopted fluxes 19.0 and 17.9 jy at 22 and 37 ghz , respectively ) .
the full description of the receiving system can be found in tersranta et .
the data were obtained during the period from 2003 january 1 to 2004 june 28 .
figure 1 shows the light curves of mrk 421 in the representative bands covered by the campaign .
the source was relatively quiet during the 2002/2003 season , although it clearly varied at x - ray and tev energies .
the largest tev flare occurred around mjd 52728 (= 30 march 2003 ) , reaching a peak count rate of nearly 4 @xmath11 and lasting for about a week .
the tev flare was accompanied by a flare in x - rays , which reached a peak count rate of @xmath065@xmath22 ( or about 24 mcrab ) .
there is no apparent time offset between the x - ray and gamma - ray flares .
mrk 421 became much more active in the 2003/2004 season , with several major flares observed .
in particular , an usually large flare ( or outburst ) took place near the end of the campaign , with the source reaching peak fluxes of @xmath0135 mcrab in x - rays , as seen by the asm / rxte , and @xmath03 crabs in the tev band , respectively . an expanded view of this flaring episode is shown in figure 2 .
the flare lasted for more than two weeks ( from @xmath10 mjd 53104 to roughly mjd 53120 ) , although its exact duration is difficult to quantify due to the presence of a large data gap between mjd 53093 and mjd 53104 .
interestingly , during this giant flare , the tev emission appears to reach the peak much sooner than the x - ray emission . although the x - ray light curve is not as densely sampled as the gamma - ray light curve , the fact that the x - ray measurements were made at the _ same _ time as the corresponding gamma - ray measurements ( see fig .
2 ) makes it quite unlikely that the difference in the rise time between the two bands is caused by some sampling bias .
the light curves also show that mrk 421 varies much less at optical and radio wavelengths .
the variability amplitude in the r band is less than about 0.9 magnitude ; the variability is even less obvious in the radio bands due to relatively large measurement errors .
there is no apparent correlation between the optical and radio bands or between either radio or optical band and the high - energy ( x - ray and tev ) bands . for completeness ,
figure 3 shows ( flwo ) optical and radio light curves for all of the bands covered .
the figure shows highly correlated variability among the optical bands , while the situation is not nearly as clear among the radio bands , due both to large measurement errors and sparse coverages . to quantify the energy dependence of variability amplitudes , we computed the so - called normalized variability amplitude ( nva ) for each spectral band .
the nva is defined as ( edelson et al .
1996 ) : @xmath23 where @xmath24 represents the mean count rate or flux , @xmath25 the standard deviation , and @xmath26 the mean measurement error in a given spectral band . to facilitate comparison of the nvas for different bands
, we only used light curves that are relatively well sampled and cover the entire 2003/2004 season ( see fig . 1 ) .
consequently , the results are obtained only for the following spectral bands : 14.5 ghz , r , x - ray , and gamma - ray .
we first rebinned the selected light curves with the same bin size . to examine variabilities on different timescales ,
two different bin sizes were used : 1 day and 7 days .
we then computed the nvas from the rebinned light curves .
the results are shown in figure 4 .
the variability of mrk 421 shows a general increasing trend toward high energies .
the source is highly variable in the x - ray and tev bands , with the nva reaching up to about 65% ( with 1-day binning ) , but varies much less in the radio and optical bands , with the nva equal to @xmath011% and @xmath016% , respectively , almost independent of the binning schemes used .
there are several caveats in the cross - band comparison .
first of all , the density of data sampling is quite different for different bands .
for instance , the sampling is more sparse in the 14.5 ghz band than in any of the other bands .
however , the under - sampling of the radio data would probably only lead to an _ under - estimation _ of the radio nva .
secondly , the measurement errors for the optical data are probably underestimated , due to possible systematic effects caused by , e.g. , the presence of bright stars in the field .
this would result in an _ over - estimation _ of the optical nva .
finally , the measured optical flux includes contribution from the host galaxy , which is at a @xmath015% level ( nilsson et al .
since it only affects the average flux , not the absolute variability amplitude , the optical nva should be about 15% higher , which is a small correction .
therefore , the conclusion that the nva is comparable in the radio and optical bands seems secure . from the light curves ( fig . 1 ) , we can see that the x - ray and tev variabilities of mrk 421 are roughly correlated , although they are clearly not always in step .
figure 5 shows the x - ray and gamma - ray count rates for all simultaneous measurements .
although a positive correlation between the rates seems apparent , it is only a loose one .
we should note that the dynamical range of the data is quite large ( @xmath030 in both energy bands ) which is important for studying correlative variability of the source . to be more rigorous , we computed the z - transformed discrete correlation function ( zdcf ; alexander 1997 ) from light curves in the two bands .
the zdcf makes use of the fisher s z - transform of the correlation coefficient ( see alexander 1997 for a detailed description ) .
its main advantage over the more commonly used dcf ( edelson & krolik 1988 ) is that it is more efficient in detecting any correlation present .
figure 6 shows the zdcf ( in 1-day bins ) derived from the 2002/2003 data set .
the zdcf seems to peak at a negative lag .
fitting the peak ( in the narrow range of -77 days ) with a gaussian function , we found its centroid at @xmath27 days , which is of marginal significance .
if real , the result would imply that the x - ray variability _ leads _ the gamma - ray variability .
other zdcf peaks are most likely caused by the _
whipple _ observing pattern , such as the quasi - periodic occurrences of the dark moon periods .
similarly , we computed a zdcf for the 2003/2004 data set .
the results are shown in figure 7 . in this case , the main feature is very broad and significantly skewed toward _ positive _ lags .
the feature appears to be a composite of multiple peaks , although large error bars preclude a definitive conclusion .
a positive zdcf peak means that the x - ray emission _
lags _ behind the gamma - ray emission .
we checked the results with different binning schemes and found no significant changes .
the detection of x - ray _ lags _ in the 2003/2004 data set should not come as a total surprise , because we have already seen ( from fig .
2 ) that the x - ray flux rose more slowly than the gamma - ray flux during the 2004 giant flare .
the difference in the rise times can be the cause of the broad zdcf peak that is skewed toward positive lags . to check that
, we excluded the flare from the x - ray and gamma - ray light curves ( by removing all data points after mjd 53100 ; see fig . 1 ) and computed a new zdcf . the results are also shown in fig .
7 ( bottom panel ) for a direct comparison .
there is a narrow peak at around -1 day .
fitting it ( in the range of -73 days ) with a gaussian function yields the centroid at @xmath28 days , which is not inconsistent the measured value for the 2002/2003 data set though it is even less significant statistically .
the features at around + 7 and -15 days ( which are also discernable in the overall zdcf ) are , again , most likely caused by the _
whipple _ observing pattern for the season . despite the complications , it is almost certain , by comparing the two zdcfs , that the x - ray lags in the 2003/2004 data set are indeed associated with the difference between the x - ray and gamma - ray rise times of the giant flare . to investigate the effects of data gaps ( which are present both in the x - ray and gamma - ray data ) on zdcf , we did the following experiment .
we created three light curves : the actual x - ray light curve of mrk 421 in the 2003/2004 season(_lc1 _ ) , an artificial light curve ( _ lc2 _ ) made by shifting _
lc1 _ by + 8 days , and another artificial light curve ( _ lc3 _ ) made by modulating _
lc2 _ with the _
whipple _ sampling pattern .
figure 8 shows the zdcfs between _
lc1 _ and _ lc2 _ and between _ lc1 _ and _ lc3 _ separately . the artificially - introduced 8-day lag is easily recovered in both cases
. we also shifted _
lc1 _ by different amounts and the lags are always recovered correctly .
therefore , the coverage gaps associated with the _ rxte _ and _ whipple _ monitoring campaigns do not wash out intrinsic offsets between the x - ray and gamma - ray light curves , although the effects of the data gaps on the shape of the zdcfs are measurable . examining the x - ray and
tev light curves more closely , we noticed that some of the tev flares have no _ simultaneous _ x - ray counterparts ( or counterparts at long wavelengths ) .
they are , therefore , similar to the reported `` orphan '' tev flare in 1es1959 + 650 ( krawczynski et al .
2004 ) . figure 9 shows an example of the phenomenon .
the tev flare peaks at almost 8 @xmath11 at around mjd 53033.4 when the x - ray flux is low .
it is interesting to note , however , that the source is clearly variable in x - rays during this time period and that the x - ray flux seems to have peaked about 1.5 days before the tev flux .
therefore , the tev flare might not be a true orphan event ; it might simply lag behind its x - ray counterpart .
alternatively , it is also possible that the tev flare is a composite of two sub - flares , with one being the counterpart of the x - ray flare and the other a true , lagging orphan flare .
the sparse sampling of the data prevents us from drawing a definitive conclusion in this regard .
we note the remarkable similarities between the tev flare shown in fig . 9 and the reported `` orphan '' flare in 1es1959 + 650 ( see fig .
4 in krawczynski et al . 2004 ) , including similar variation patterns in x - rays . from the x - ray and tev light curves
, we also detected flares on much shorter timescales .
figure 10 shows examples of sub - hour x - ray flares .
the most rapid x - ray flare detected lasted only for @xmath020 minutes and shows substantial sub - structures , implying variability on even shorter timescales . only on one occasion was a sub - hour x - ray flare detected during a ( longer - duration ) gamma - ray flare .
no counterpart is apparent at tev energies ( see fig .
10 ) , although the error bars on gamma - ray measurements are quite large . we note that , if a strong rapid tev flare , for example , like the one detected by gaidos et al . (
1995 ) , had occurred , we should have easily detected it , given the improved instrumentation . the rapid x - ray flare shown in fig .
10 is relatively weak , with a peak - to - peak amplitude of only about 5% of the average flux level .
the data do not allow any direct comparison on long timescales , due to the short exposure of the x - ray observation .
we first divided the _
observations into 8 groups based on the x - ray count rate of mrk 421 , with an increment of 20 @xmath29 . for this work , we focused on three of the groups : low ( 020 @xmath29 ) , medium ( 4060 @xmath29 ) , and high ( 140160 @xmath29 ) , with the average count rate increasing roughly by a factor of three from low to medium and from medium to high . then , for each observation in a group we searched for an observation at tev energies that was made within an hour
only matched pairs were kept for further analysis .
we ended up with a total of 16 , 9 , and 3 pairs in the low - flux , medium - flux , and high - flux group , respectively .
it turns out that the low - flux group consists of observations taken between 2003 march 8 and may 3 , the medium - flux group between 2004 january 27 and march 26 , and the high - flux group between 2004 april 1620 . for each group , we proceeded to construct a flux - averaged sed at x - ray and tev energies .
the flux - averaged x - ray spectrum can be fitted satisfactorily by a power law with an exponential roll - over ( with reduced @xmath30 ) .
we should note that we added a 1% systematic uncertainty to the x - ray data for spectral analysis ( which is a common practice ) , to take into account any residual calibration uncertainty . also , we fixed the hydrogen column density at @xmath31 ( dickey & lockman , 1990 ) .
the best - fit photon index is about @xmath32 , @xmath33 , and @xmath34 for the low- , medium- , and high - flux group , respectively , and the roll - over energy about @xmath35 , @xmath36 , and @xmath37 kev .
therefore , the x - ray spectrum of mrk 421 hardens toward high fluxes . using the best - fit model we then derived the x - ray sed for each data group . for the gamma - ray spectral
fits , a bin size of 0.1667 in @xmath38 was adopted for the medium- and high - flux groups and a wider bin size of 0.4 for the low - flux group . as shown in fig .
12 , the low - flux sed still has very large error bars .
the first energy bin corresponds to an energy of @xmath0260gev .
the gamma - ray spectra can all be satisfactorily fit by a power law , with a photon index of @xmath39 , @xmath40 , and @xmath41 for the low- , medium- , and high - flux groups , respectively .
the errors bars only include statistical contributions .
for the purpose of comparison with some of the published results , we also fitted the spectra with a cut - off power law ( see , e.g. , krennrich et al . 2002 ) but with the cut - off energy fixed at 4.96 tev .
the photon index is @xmath42 , @xmath43 , and @xmath44 for the low- , medium- , and high - flux groups , respectively . like the x - ray spectrum
, therefore , the tev spectrum also seems to harden toward high fluxes , although the uncertainty here is quite large . finally , we searched for radio and optical observations that fall in one of the groups and computed the average fluxes to complete the sed for the group .
given the fact that mrk 421 did not vary as significantly at these wavelengths , we believe that the derived seds are quite reliable .
figure 12 summarizes the results . we experimented with using a one - zone ssc model ( see krawczynski et al .
2004 for a detailed description of the code , which we revised to use the adopted cosmological parameters ) to fit the measured flux - averaged seds .
briefly , the model calculates the sed of a spherical blob of radius @xmath17 .
the blob moves down the conical jet with a lorentz factor @xmath45 .
the emitting region is filled with relativistic electrons with a broken power - law spectral distribution : @xmath46 , for @xmath47 , and @xmath48 , for @xmath49 ( although the code does not evolve the electron spectrum self - consistently ) .
the model accounts for the attenuation of the very - high - energy @xmath50-rays by the diffuse infrared background ( as modelled by macminn & primack 1996 ) .
we first found an initial `` best fit '' to each sed by visual inspections .
we then performed a systematic grid search around the `` best fit '' that involves the following parameters : magnetic field @xmath15 in the range of 0.0450.45 g , doppler factor @xmath51 in 1020 , @xmath52 in 1.82.2 , @xmath53 in 2.93.7 , @xmath54 in 9.912.2 , @xmath55 in 9.912.2 , and the normalization ( @xmath56 ) of @xmath57 in 0.006750.44325 @xmath58 .
variability constraints ( see the next section ) were taken into account in the choice of some of the parameter ranges .
all other parameters in the model ( e.g. , @xmath17 ) were fixed at the nominal values determined by the visual inspections .
we found roughly where the global @xmath59 minimum lies through a coarse - grid search , and then conducted a finer - grid search through much smaller parameter ranges around the minimum to find the best fit .
figure 11 shows the results for the high - flux group .
it is apparent that the model severely underestimates the radio and optical fluxes .
large deviations are also apparent at tev energies .
we should stress that our grid searches are , by no means , exhaustive .
however , the results should be adequate for revealing gross discrepancies between the model and the data .
to investigate whether the fit could be improved by introducing additional zones ( which are assumed to be independent of each other ) , we introduced a new zone to account for the radio fluxes and another for the optical fluxes . to be consistent with the observed variability
, the additional zones must be placed much further down the jets .
figure 12 shows fits to the seds with such a three - zone ssc model .
this ad - hoc approach does seem to yield a reasonable fit to the data .
in such a scenario , the observed variability at x - ray and tev energies would be associated with zones close to the central black hole , while radio and optical emission are expected to vary on longer timescales further down the jet .
the success of the standard ssc model lies partly in the fact that it provides a natural explanation for the correlation between the x - ray and gamma - ray emission .
however , we found that the correlation is not as tight as one might naively expect from the ssc model .
it seems unlikely that the loose correlation is caused by the choice of spectral bands used in the cross - correlation analysis , given how broad the _ pca _ and _ whipple _ passing bands are .
in fact , both the derived variability amplitudes ( fig . 4 ) and seds ( fig . 12 ) indicate that the x - ray and gamma - ray photons are likely to have originated from the same population of electrons in the context of the ssc model . in the standard ssc model ,
x - ray photons are compton upscattered to produce gamma - ray photons , so an x - ray lead is naturally expected
. however , the characteristic timescale of the ssc process would be much too short to account for the x - ray lead of 12 days ( see fig .
6 ) , although we can not be sure about whether the lead is actually real , due to large uncertainties and possible systematic effects .
what is certain is that the ssc model can not explain the difference between the x - ray and gamma - ray rise times of the 2004 giant flare .
it is conceivable that the flaring episode could have started with an orphan tev flare and followed by a pair of simultaneous x - ray and tev flares .
if this is the case , it would be opposite to the theoretical scenario recently put forth by bttcher ( 2004 ) . in the hadronic models , both x - ray lag and lead
could , in principle , occur , even at the same time , depending on the relative roles of the primary and secondary electrons .
however , it would also seem challenging for these models to account for the different x - ray and gamma - ray rise times of the 2004 giant flare or to quantitatively explain the long x - ray lead ( of 12 days ) if it is real . combined with the measured sed
, rapid x - ray flares pose severe constraints on some of the physical properties of the flaring region , in a relatively model - independent manner , because the x - ray emission from mrk 421 is almost certainly of synchrotron origin .
the most rapid x - ray flare detected during our campaign has a duration of about 20 minutes and reaches a peak amplitude of @xmath015% of the ( local ) non - flaring flux level ( see fig .
therefore , the size of the flaring region must satisfy : @xmath60 , where @xmath61 is the duration of the flare ( @xmath62 s ) , @xmath51 is the doppler factor of the jet ( @xmath63 ) , and @xmath64 is the redshift of mrk 421 ( @xmath65 ) . here and in the remainder of the paper , we use the primed symbols to denote quantities in the co - moving frame of the jet and unprimed ones corresponding quantities in the frame of the observer .
it is interesting to note that the derived upper limit is already comparable to the radius of the last stable orbit around the @xmath66 black hole in mrk 421 ( barth et al .
2003 ) , for @xmath67 . since the peak flux of the flare is a significant fraction of the steady - state flux ,
the size of the flaring region is probably comparable to the lateral extent of the jet ( to within an order of magnitude ) .
if the jet originates from accretion flows , as is often thought to be the case , the result would also represent an upper limit on the inner boundary of the flows .
the decaying time of the most rapid x - ray flare ( about 600 s ) sets a firm upper limit on the synchrotron cooling time of the emitting electrons .
the cooling time is given by ( rybicki & lightman 1979 ) , @xmath68 , where @xmath69 is the electron rest mass , @xmath70 is the thomson cross section , @xmath71 is the strength of the magnetic field in the region , and @xmath72 is the characteristic lorentz factor of those electrons that contribute to the bulk of the observed x - ray emission . by requiring @xmath73 , where @xmath74 is the measured decaying time of the flare ( @xmath75 ) , we derived a lower limit on @xmath71 , @xmath76 , where @xmath77 . from modeling the seds
, we found @xmath78 .
it should be noted that the limits derived are only appropriate for the region that produced the 20-min x - ray flare and that not all flares are necessarily associated with the same region . in our attempt to model the seds ( 3.2 ) , we only require that the model parameters be consistent with variability timescales of hours . results from more sophisticated modeling will be presented elsewhere .
further constraints can be derived from the detected tev flares .
the fact that we detect tev photons from mrk 421 requires that the opacity due to @xmath79 must be sufficiently small near the tev emitting region(s ) .
this requirement leads to ( dondi & ghisellini 1995 ) : @xmath80^{1/(4 + 2\alpha)},\ ] ] where @xmath81 is the luminosity distance , @xmath82 is the tev flux doubling time , @xmath9 and @xmath83 are the local spectral index and the energy flux , respectively , at the target photon frequency , at which the pair production cross section peaks , @xmath84 where @xmath85 is the frequency of the gamma - ray photon . for @xmath86 tev , we have @xmath87 hz . from the high - flux sed ( fig .
12 ) , we found @xmath88 and @xmath89 . given @xmath90 mpc and @xmath91 hour ( see fig .
10 ) , we derived a lower limit on the doppler factor , @xmath92 .
since tev emission is the consequence of inverse - compton scattering of ( synchrotron ) x - ray photons by the electrons themselves in the standard ssc model , a flare at tev energies must be _ preceded _ by an accompanying flare at x - ray energies . while the model might still be able to accommodate the presence of orphan x - ray flares , the presence of orphan tev flares seems problematic . on the other hand ,
the `` orphan '' tev flares in mrk 421 ( as shown fig .
9 ) and in 1es1959 + 650 ( krawczynski et al . 2004 ) might not be orphan events after all . in both cases , the gamma - ray flares seem to be preceded by x - ray flares , which could be attributed to the ssc process although the x - ray lead ( of 12 days ) would seem too long .
alternatively , it might be possible to attribute the x - ray lead to some sort of feedback between different emission regions in the jet ( i.e. , an orphan x - ray flare in one region triggers an orphan tev in another )
. genuine orphan tev flaring would not be easy to understand in the hadronic models either , despite looser coupling between the x - ray and tev emission , unless the injected e / p ratio changes significantly from flare to flare .
any change in the proton population will likely be accompanied by a change in the electron ( primary or secondary ) population , so tev and x - ray flares should both occur as a result .
interestingly , orphan tev flares may be understood in a hybrid scenario in which protons are present in the jet in substantial quantities but are not necessarily dominant compared to the lepton component ( bttcher 2004 ) . in this case ,
an orphan tev flare is associated with ( and _ _ follows ) a pair of simultaneous x - ray and tev flares that originate in the standard ssc process .
the synchrotron ( x - ray ) photons are then reflected off some external cloud and return to the jet ( following a substantial delay with respect to the initial flares ) .
the returned photons interact with the protons in the jet to produce pions and subsequent @xmath93 decay produces tev photons in the flare .
the model was shown to be able to account for the `` orphan '' tev flare observed in 1es1959 + 650 ( bttcher 2004 ) .
it might also explain some of the similar flares in mrk 421 , for example , the one shown in fig .
9 , if it is a composite of two gamma - ray flares ( see discussion in
3.1.3 ) .
the sed of mrk 421 varies greatly both at x - ray and tev energies but only weakly at radio and optical wavelengths .
the much reduced variability at long wavelengths is expected for short timescales , because of long synchrotron cooling times of the radio or optical emitting electrons . in other words
, there is no reason to expect a tight correlation between the low - energy ( radio or optical ) emission and the high - energy ( x - ray or tev ) emission .
of course , the lack of such a correlation could also be due to the fact that the low - energy emission originates in regions further down the jet . between the radio and optical bands
, we found that the nvas of the source are comparable ( see fig .
4 ) , which is somewhat surprising because the synchrotron cooling time of the electrons responsible for the optical emission is expected to be much shorter than that of those for the radio emission . on the other hand , this could be evidence for the presence of different populations of electrons that produce the emission in the radio and optical bands .
the populations may be located in physically separated regions and may have different spectral energy distributions ( e.g. , @xmath94 ) .
we found that both sed peaks move to higher energies as the luminosity of the source increases .
moreover , the x - ray spectrum becomes flatter , which seems to be common among high - frequency - peaked bl lacs ( e.g. , giommi et al .
1990 ; pian et al .
1998 ; xue & cui 2004 ) .
similar spectral hardening has also been see at tev energies ( krennrich et al .
2002 ; aharonian et al .
2002 ) . in our data set ,
we have seen some indication of spectral hardening but we can not be certain because of the low statistics of the data .
the evolution of the sed could be driven by a hardening in the electron energy distribution and/or an increase in the strength of the magnetic field at high luminosities .
in this paper , we have presented a substantial amount of multi - wavelength data that have been collected on mrk 421 as the result of a long - term monitoring campaign .
the large dynamical range of the data has allowed us to carry out some detailed investigations on variability timescales , cross - band correlation , and spectral variability .
the main results from this work are summarize as follows : * we have shown that the emission from mrk 421 varied greatly at both x - ray and gamma - ray energies and that the variabilities in the two energy bands are generally correlated .
this is consistent with some of the earlier results but the dynamical range of the data and statistics are much improved here . equally important is the finding that the correlation is only a fairly loose one ( see figures 57 ) .
the emission clearly does not always vary in step between the two bands .
* we have discovered that the x - ray emission reached the peak days _ after _ the tev emission during the giant flare in april 2004 ( see fig . 1 ) .
such a difference in the rise times for the two energy bands poses a serious challenge to the standard ssc model , as well as the hadronic models .
whether it can be accommodated by some hybrid model remains to be seen .
in addition , there is tentative evidence that the x - ray variability leads the gamma - ray variability by 12 days .
if proven real , such a long x - rad lead would also be difficult for either the ssc or hadronic models to explain quantitatively .
* we have detected x - ray and tev flares on relatively short timescales .
for example , the most rapid x - ray flare detected has the duration of only @xmath020 minutes and a peak amplitude of @xmath015% of the local non - flaring flux level .
although no similarly rapid tev flares were detected during our campaign , we did observe a tev flare with duration of hours .
physical constraints on the flaring regions have been derived from the rapid x - ray and tev variabilities .
* we have seen tev flares in mrk 421 that are similar to the reported `` orphan '' flare in 1es1959 + 650 . genuinely orphan or
not , this presents another serious challenge to the proposed emission models ( see 4.3 ) .
if they are true orphan events , they may be explained by a hybrid model in which both electrons and protons are present in the jet in substantial quantities . on the other hand ,
if the preceding x - ray flare turns out to be the counterpart , the 12 day x - ray lead would challenge the proposed emission models ( see 3.1.3 ) .
* we have derived high - quality seds of mrk 421 from simultaneous or nearly simultaneous observations at tev , x - ray , optical , and radio energies .
the x - ray spectrum clearly hardens toward high fluxes ; the gamma - ray spectrum also appears to evolve similarly ( although we can not be sure due to large uncertainties of the data ) .
a one - zone ssc model fails badly to fit the measured fluxes at the radio and optical wavelengths , and , to a lesser extent , also underestimates fluxes at the highest tev energies .
the introduction of additional zones improves the fit significantly .
* further evidence for the presence of multiple populations of emitting electrons is provided by the comparable radio and optical nvas ( see fig . 4 and discussion in 4.4 ) .
we acknowledge the technical assistance of e. roache and j. melnick .
the veritas collaboration is supported by the u.s .
department of energy , national science foundation ( nsf ) , the smithsonian institution , nserc ( canada ) , pparc ( uk ) , and science foundation ireland
. w. cui and m. baejowski also gratefully acknowledge financial support from nasa and thank peter biermann for useful discussions on the subject and comments on the manuscript .
umrao is supported in part by funds from nsf and from the department of astronomy at university of michigan .
aharonian , f. 2000 , new astronomy , 5 , 377 aharonian , f. , et al .
2002 , a&a , 393 , 89 aharonian , f. , et al .
2003 , a&a , 403 , l1 alexander , t. , 1997 , in proc .
`` astronomical time series '' , eds .
d. maoz , a. sternberg , & e. m. leibowitz , ( dordrecht : kluwer ) , 163 aller , h. d. , aller , m. f. , latimer , g. e. , & hodge , p. e. 1985 , apjs , 59 , 513 baars , j.w.m . ,
genzel , r , pauliny - toth , i.i.k . , & witzel , a. 1977 , a&a , 61 , 99 beall , j. h. , & bednarek , w. 1999 , apj , 510 , 188 bttcher , m. , et al . , 1997 , a&a , 324 , 395 bttcher , m. 2002 , in proc .
`` the gamma - ray universe '' , xxii moriond astrophysics meeting , eds .
a. goldwurm et al . , p. 151 bttcher , m. 2005 , apj , 621 , 176 buckley , j. h. , et al . 1996 , apj , 472 , l9 catanese , m. , et al .
1997 , apj , 487 , l143 cui , w. 2004 , apj , 605 , 662 cui , w. , et al .
2004 , in proc .
`` international symposium on high energy gamma - ray astronomy '' ( gamma-2004 ) , eds .
f.a . aharonian and h. voelk , aip conf .
ser . , 745 , 455 ( astro - ph/0410160 ) daniel , m. , et al . 2005 , apj , 621 , 181 dar , a. , & laor , a. 1997 , apj , 478 , l5 dermer , c. d. , et al .
1992 , a&a , 256 , l27 dicke , j , & lockman , j. 1990 , ara&a .
28 , 215 dondi , l. , & ghisellini , g. 1995 , mnras , 273 , 583 edelson , r. a. , & krolik , j. h. 1988 , apj , 333 , 646 edelson , r. a. , et al .
1996 , apj , 470 , 364 falcone , a. , et al .
2004 , apj , 613 , 710 finley , j. p. , et al .
2001 , in proc .
cosmic ray conf . , 199 fossati , g. , et al .
1998 , mnras , 299 , 433 gaidos , j. a. , et al .
1996 , nature , 383 , 319 giommi , p. , barr , p. , pollock , a. m. t. , garilli , b. , & maccagni , d. 1990 , apj , 356 , 432 hillas , a. m. , et al .
1998 , apj , 503 , 744 krawczynski , h. , et al .
2004 , apj , 601 , 151 krennrich .
f. , et al .
2002 , apj , 575 , l9 lebohec , s. , & holder , j. 2003 , astropart .
19 , 221 landolt , a. 1992 , aj , 104 , 340 lyutikov , m. 2003 , new astr .
47 , 513 mannheim , k. , & biermann , p. l. 1992 , a&a , 253 , l21 maraschi , l. , ghisellini , g. , & celotti , a. , 1992 , apj , 397 , l5 maraschi , l. , et al . 1999 ,
phys . , 11
, 189 marscher , a. p. , & gear , w. k. , 1985 , apj , 298 , 11 mastichiadis , a. , & kirk , j. g. , 1997 , a&a , 320 , 19 macminn , d. , & primack , j. r. 1996 , space sci .
rev . , 75 , 413 mochejska , b. j. , stanek , k. z. , sasselov , d. d. , & szentgyorgyi , a. h. 2002 , aj , 123 , 3460 mohanty , g. , et al .
1998 , astropart .
, 9 , 15 mcke , a. , et al .
2003 , astropart .
phys . , 18 , 593 nilsson , k. , pursimo , t. , takalo , l. o. , sillanp , a. , & pietil , h. 1999 , pasp , 111 , 1223 petry , d. , et al . , 2000 ,
apj , 536 , 742 petry , d. , et al .
, 2002 , apj , 580 , 104 pian , e. , et al .
1998 , apj , 492 , l17 pohl , m. , & schlickeiser , r. 2000 , a&a , 354 , 395 punch , m. , et al .
1992 , nature , 358 , 477 rees , m. j. , 1978 , mnras , 184 , p61 reynolds , p. , et al .
1993 , apj , 404 , 206 rybicki , g. b. , & lightman , a. p. , radiative processes in astrophysics ( new york : john wiley & sons ) sikora , m. , et al .
1994 , apj , 421 , 153 spada , m. , et al .
, 2001 , mnras , 325 , 1559 stetson , p. b. 1987 , pasp , 99 , 191 tersranta , h. , et al . 1998 , a&as , 132 , 305 urry , c. m. , & padovani , p. 1995 , pasp , 107 , 803 villata , m. , raiteri , c. m. , lanteri , l. , sobrito , g. , & cavallone , m. 1998 , a&as , 130 , 305 weekes , t. c. 2003 , proc .
28th icrc ( astro - ph/0312179 ) xue , y. , & cui , w. 2005 , apj , 622 , 160 | we report results from an intensive multi - wavelength monitoring campaign on the tev blazar mrk 421 over the period of 20032004 .
the source was observed simultaneously at tev energies with the whipple 10 m telescope and at x - ray energies with _ rossi x - ray timing explorer _ ( rxte ) during each clear night within the _ whipple _ observing windows . supporting observations were also frequently carried out at optical and radio wavelengths to provide simultaneous or contemporaneous coverages . the large amount of simultaneous data has allowed us to examine the variability of mrk 421 in detail , including cross - band correlation and broad - band spectral variability , over a wide range of flux .
the variabilities are generally correlated between the x - ray and gamma - ray bands , although the correlation appears to be fairly loose .
the light curves show the presence of flares with varying amplitudes on a wide range of timescales both at x - ray and tev energies .
of particular interest is the presence of tev flares that have no coincident counterparts at longer wavelengths , because the phenomenon seems difficult to understand in the context of the proposed emission models for tev blazars .
we have also found that the tev flux reached its peak days _ before _ the x - ray flux did during a giant flare ( or outburst ) in 2004 ( with the peak flux reaching @xmath0135 mcrab in x - rays , as seen by the asm / rxte , and @xmath03 crab in gamma rays ) .
such a difference in the development of the flare presents a further challenge to the leptonic and hadronic emission models alike .
mrk 421 varied much less at optical and radio wavelengths .
surprisingly , the normalized variability amplitude in optical seems to be comparable to that in radio , perhaps suggesting the presence of different populations of emitting electrons in the jet .
the spectral energy distribution of mrk 421 is seen to vary with flux , with the two characteristic peaks moving toward higher energies at higher fluxes .
we have failed to fit the measured seds with a one - zone ssc model ; introducing additional zones greatly improves the fits .
we have derived constraints on the physical properties of the x - ray / gamma - ray flaring regions from the observed variability ( and sed ) of the source .
the implications of the results are discussed . |
motivated by the ads / cft correspondence @xcite , the authors of @xcite constructed a holographic superconductor model , which presents a gravitational description of the ( unconventional ) superconductors .
the schwarzschild - ads black hole provides a temperature to the system . a critical temperature @xmath1 exists , below which the black hole grows scalar hair , which corresponds to the condensate of the dual operator on the boundary .
according to the ads / cft correspondence , the phenomena indicates there is a superconducting phase transition of the dual theory on the boundary @xcite .
many works are carried out to study the topic , such as gauss - bonnet gravity @xcite , born - infeld theory @xcite , or including both of them @xcite . despite the numerical results in the original paper @xcite
, the phase transition can also be investigated analytically with the help of matching method @xcite and sturm - liouville method @xcite .
the @xmath0 gravity is one of modifications to the einstein gravity , which has its advantage to generate the late - time expansion of the cosmic acceleration ( see @xcite for a review ) .
it is hard to find nontrivial black hole solutions with nonconstant ricci scalar for general @xmath0 .
there are even no such solutions in some cases @xcite .
recently , several interesting solutions are obtained for some specific @xmath0 @xcite , even with matters @xcite .
questions naturally arising here are whether we can find a suitable black hole to realize the superconducting phase transition , and how the modifications of @xmath0 gravity affect the propertities of the corresponding holographic superconductors as already done in other higher order gravity theory @xcite . according to the variousness of @xmath0 theory
, we only consider a specific @xmath0 model to address the mentioned issues . following the method @xcite ,
black holes of non - constant ricci scalar with planar topology are obtained .
properties of the black hole and the holographic superconductors on the boundary are investigated in a perturbative manner .
the corrections to the lowest order of @xmath2 , which corresponds to the modifications of @xmath0 to einstein gravity as we will indicate , are obtained explicitly .
the paper is organized as follows . in section 2
, we construct an asymptotic ads black hole with planar topology and investigate the corresponding @xmath0 theory . in section 3 ,
we compute the modified critical temperature @xmath1 and the condensate of the dual operator @xmath3 . in the last section ,
we make a simple conclusion .
in order to realize a superconducting phase transition , nontrivial black holes which are asymptotic to ads spacetime and have the planar topology are needed . in this section ,
we follow the method of @xcite to get such solutions and the corresponding @xmath0 .
action describing the @xmath0 gravity and mini - coupled matter fields in four dimensional spacetime has the following form @xmath4,\end{aligned}\ ] ] where @xmath0 is an analytic function of the ricci scalar .
when @xmath5 , it recovers the einstein gravity with cosmological constant . from the action , the vacuum modified einstein equation can be written as @xmath6 where the prime on @xmath0 means the derivative with respect to @xmath7 .
the equation usually includes higher derivatives of the metric for a general @xmath0 , which makes it difficult to solve .
the general metric of the static and planar symmetric black holes have the following form : @xmath8 as @xcite , solutions to the modified einstein e.q .
( [ einstein ] ) should satisfy the following equations : @xmath9 with definitions as @xmath10 and @xmath11 .
we make an ansatz that @xmath12 has the following form : @xmath13 substituting it into e.q.s ( [ eom6 ] ) , solutions are obtained as follows : @xmath14 where @xmath15 are arbitrary constants . to have schwarzschild-(a)ds black holes , we set @xmath16 and @xmath17 to be a positive number
. then the metric of the spacetime has the form @xmath18 the ricci scalar can be calculated from the metric as follows : @xmath19 it s an analytic function of the radius @xmath20 . in the asymptotic region with large @xmath20 , the ricci scalar can be expanded as a function of @xmath21 @xmath22 the result indicates that the spacetime is asymptotic to ( a)ds . for latter discussion , we only consider the asymptotic ads black holes , which can be obtained by choosing the parameters properly .
the effective cosmological constant and the radius of the asymptotic ads spacetime can be defined as @xmath23 supposing the inverse function of e.q .
( [ ricci ] ) can be obtained , we will get @xmath12 as a function of @xmath7 by substituting @xmath24 into the e.q .
( [ f(r ) ] ) .
then the corresponding @xmath0 can be obtained from the integration of @xmath25 .
actually , e.q .
( [ ricci ] ) indicates that the exact inverse function ca nt be obtained . if @xmath26 in e.q .
( [ f(r ) ] ) , then @xmath27 is a constant .
the solutions of the e.q.s ( [ eom6 ] ) will be @xmath28 which is just the trivial case with constant ricci scalar .
then @xmath0 can take any form as soon as it satisfies the constraint @xmath29 , which is obtained by contracting the e.q .
( [ einstein ] ) . considering @xmath2 as a small parameter
, the problem can be solved in a perturbative manner , and @xmath2 corresponds to the modifications of the @xmath0 gravity to some extent .
as we will see later , it relates to the parameter @xmath30 which truly represents the modifications .
the ricci scalar ( [ ricci ] ) is expanded to the 1st order of @xmath2 as @xmath31 the inverse function is obtained as @xmath32 substituting e.q .
( [ r(r ) ] ) into e.q . ( [ f(r ) ] ) and making a integration , we obtain the corresponding @xmath0 theory to the 1st order of @xmath2 , @xmath33 it is the approximate action , the higher order corrections of which are omitted .
the problem can also be viewed from the other point , which means that if we take ( [ f(r ) ] ) as the precise action , the metric ( [ metric0 ] ) can be checked to satisfy the equation of motion ( [ einstein ] ) to the lowest order of @xmath2 .
now we make some justifications as follows : * compare e.q .
( [ f(r ) ] ) with einstein - hilbert action with negative cosmological constant , we set @xmath34 and @xmath35 ; * from e.q .
( [ ricci ] ) , the limit is calculated as @xmath36 .
so we set @xmath37 and @xmath38 ; * the action should nt involve the integration parameters @xmath39 explicitly , so we define @xmath40 ; * at last , we set @xmath41 for simplicity , which means we focus on a subclass of the solutions obtained above . under these justifications ,
the approximate @xmath0 gravity has the following form , @xmath42 the @xmath0 theory of this type were already studied @xcite and nontrivial solutions with spherical topology were obtained @xcite .
when considering the ads / cft correspondence , the bulk gravity ( [ approximate action ] ) appears as classical dual of the quantum cft on the boundary in large n limit .
from the definition @xmath43 , we know that the parameter @xmath2 really represents the modification of @xmath0 to einstein gravity . the metric changes into @xmath44 note that in the einstein limit @xmath45 or in the asymptotic region @xmath46 , the solution truly goes to ads spacetime .
the structure of the black hole is modified for nonzero @xmath2 , which means the horizons are not degenerate anymore . for small @xmath2 ,
the radius of the outer horizon can be calculated perturbatively .
we assume the radius of the horizon has the following form : @xmath47 substituting it into the equation @xmath48 , the corrected radius is obtained as follows to the 2rd order of @xmath2 , @xmath49 with @xmath50 as a quantum effect , the hawking temperature of black holes is model independent .
it can be obtained either by the method of computing the surface gravity @xmath51 or the euclidean correspondence avoiding the conical singularity . for the metrc of this type
, the temperature can be calculated simply as @xmath52 the result indicates that there are no corrections to the hawking temperature to the 1st order of @xmath2 .
other thermal quantities such as the entropy and energy , and the thermal dynamical laws should also be investigated .
black holes with planar topology in einstein gravity were studied in @xcite .
the form of boundary term and euclidean method were addressed for @xmath0 gravity @xcite .
but for the black holes with nonconstant curvature , it has some subtleties and still needs more investigations .
so the topic is left to the further investigations . in latter discussions , @xmath53 is taken for simplicity .
following the method @xcite and taking the probe limit as @xcite , we discuss the holographic superconductors in the background described by the metric e.q .
( [ metric2 ] ) .
the matter fields are chosen to be maxwell field and charged complex scalar field , which are described by the following lagrangian : @xmath54 the mass of the scalar field is taken to be @xmath55 with @xmath56 defined in e.q .
( [ effective lambda1 ] ) . referring to the symmetry of the background ,
we adopt the following ansatz : @xmath57 the equations of motion for matter fields are obtained as follows : @xmath58 where the prime means the derivative with respect to @xmath20 , and the parameter @xmath59 is defined as @xmath60 .
finiteness requires the electric potential to be zero @xmath61 at the horizon .
the e.q.s ( [ horizon1 ] ) and ( [ eom7 ] ) imply the relation @xmath62 , which is calculated to the 1st order of @xmath2 . in the asymptotic region ,
the solutions behave as @xmath63 where @xmath64 and @xmath65 are the chemical potential and the charge density of the dual field theory @xcite . in 4-dimensions , both of the two terms of @xmath66 are normalizable and
can be viewed as the vacuum expectation values @xcite . referring to the dictionary of holography
, the vacuum expectation values of the dual scalar operators @xmath67 on the boundary can be obtained as @xmath68 variable transformation is taken as @xmath69 , under which the e.q.s ( [ eom7 ] ) and ( [ eom8 ] ) change into @xmath70 where all the primes mean the derivatives with respect to @xmath71 . the asymptotic behavior e.q .
( [ bc1 ] ) changes into @xmath72 we expand the field @xmath73 and @xmath74 near the horizon @xmath75 , @xmath76 with the help of the equations of motion ( [ eom9 ] ) ( [ eom10 ] ) and the relations at the horizon @xmath77 and @xmath78 to the 1st order of @xmath2 , the taylor expansions can be expressed with fewer parameters as @xmath79 in order to match the equations ( [ bc4 ] ) on the boundary and ( [ taylor4 ] ) ( [ taylor5 ] ) near the horizon , the following equations must be satisfied at some intermediate point @xmath80 : @xmath81 here we only consider the case with nonvanishing @xmath82 to illustrate our purpose . actually , a similar discussion can be carried out trivially for @xmath83 . from e.q.s ( [ mc9 ] ) and ( [ mc10 ] )
, we obtain @xmath84 .
\label{mc12}\end{aligned}\ ] ] when there is no condensate on the boundary , which means @xmath85 , @xmath86 must be zero . from e.q .
( [ mc8 ] ) , we obtain @xmath87 using e.q.s ( [ horizon1 ] ) ( [ hawking1 ] ) ( [ mc11 ] ) and ( [ condensate4 ] ) , we obtain the relation @xmath88 where definitions are taken as @xmath89 and @xmath90
. when @xmath91 approaches some critical value @xmath1 , the condensate @xmath82 approaches 0 .
so the critical temperature of the phase transition can be decided from e.q.s ( [ mc12 ] ) and ( [ condensate5 ] ) as @xmath92 we define @xmath93 as the critical temperature without @xmath0 modifications . as discussed in @xcite , the dual theory on the boundary turns into the superconducting phase below the critical temperature @xmath1 .
we plot the result when @xmath2 takes different values . from figure 1
, we see that the modification shifts the curves of the critical temperature as a function of charge density @xmath65
. varies with the parameter @xmath65 , where the matching point is chosen to be @xmath94 .
the curves correspond to @xmath95 from the bottom up separatively .
, width=240 ] from e.q.s ( [ condensate ] ) , ( [ mc12 ] ) , ( [ condensate5 ] ) and ( [ critical temperature2 ] ) , we obtain the condensate of the dual scalar operator on the boundary @xmath96 near the critical temperature @xmath97 , the condensate behaves as @xmath98 with @xmath99.\end{aligned}\ ] ] we see that near the critical temperature it consists with the second order phase transition of superconductor , which is typical experimentally . beside such region ,
the relation is modified by the parameter @xmath2 as e.q .
( [ condensate6 ] ) . when taking the limit @xmath100 , the modified condensate at zero temperature is obtained as @xmath101 in the second line , the e.q .
( [ critical temperature2 ] ) and the definition of @xmath102 are adopted .
the result shows that there are no corrections to the 1st order of @xmath2 when @xmath91 is zero .
the expansion around zero temperature has the following form : @xmath103at \nonumber\\ & & + \mathcal{o}(a^2).\end{aligned}\ ] ] from figure 2 , we know that the presence of @xmath0 makes it harder to form scalar hair either near zero temperature or the critical point
. we point it out again that the result is just referring to our specific model with approximate action ( [ approximate action ] ) , and not universal to the @xmath0 gravity . as a comparison
, we can follow the same method to discuss the topic with the schwarzschild - ads black hole in einstein gravity .
the results are simply as follows : @xmath104 explicitly , they consist with our @xmath0 model taking the limit @xmath105 in e.q.s ( [ critical temperature2 ] ) and ( [ condensate6 ] ) .
in this paper , we investigate the topic of holographic superconductors in @xmath0 gravity .
the background spacetime is an exact planar black hole solution , which is asymptotic to ads spacetime for large radius .
the corresponding @xmath0 ( [ approximate action ] ) is obtained in a perturbative manner for small @xmath2 , which corresponds to the modifications referring to the relation @xmath106 .
thermal quantities and horizon structure are modified to the black hole , for which the perturbative results are obtained as ( [ horizon1 ] ) and ( [ hawking1 ] ) . from e.q .
( [ hawking1 ] ) , we know the correction to the hawking temperature is at least 2rd order . following the method @xcite ,
we compute the corrected critical temperature and condensate explicitly referring to the e.q.s ( [ critical temperature2 ] ) and ( [ condensate6 ] ) separately .
as illustrated in figure 1 , we know the critical temperature of the superconductor is shifted . in figure 2 , we see that it s more difficult to form the scalar hair for a given temperature in such an @xmath0 theory .
the calculations indicate that corrections are taking place even we investigate the topic to lowest order of @xmath2 only .
qiyuan pan , bin wang , e. papantonopoulos , j. de oliveira and a.b .
pavan , _ holographic superconductors with various condensates in einstein - gauss - bonnet gravity _ , phys
d * 81 ( 2010 ) 106007 [ arxiv : hep - th/0912.2475 ] .
ji - liang jing , lian - cheng wang , qi - yuan pan and song - bai chen , _ holographic superconductors in gauss - bonnet gravity with born - infeld electrodynamics _ , phys . rev . * d * 83 ( 2011 ) 066010 [ arxiv : gr - qc/1012.0644 ] . | we study the holographic superconductors in @xmath0 gravity , and show how the critical temperature and condensate of the dual operators depend on the modifications of @xmath0 to einstein gravity . a nontrivial planar black hole which is asymptotic to ads spacetime
is investigated for a specific @xmath0 .
the corrections to the thermal quantities of the black hole , the critical temperature and the condensate of the dual operator are all obtained in a perturbative approach explicitly .
some comments are given on the affections of such modifications to the holographic superconductors . |
in a weakly - interacting massive particle dark matter ( wimp dm ) scenario , a dm particle with mass of @xmath13 gev
tev should have an ( thermally - averaged ) annihilation cross section of @xmath15 in order to reproduce the observed dm abundance due to the thermal production . on the other hand , recently reported excesses of cosmic - ray positron @xcite and electron fluxes @xcite may be interpreted as signatures of annihilating dark matter with fairly large annihilation cross section of order of @xmath16
@xmath17 depending on dm mass @xmath18 , which is typically three orders of magnitude larger than the canonical value quoted above , although constraints from other observations , such as gamma - rays @xcite and neutrinos @xcite are also stringent and might have already excluded some parameter regions .
one way to achieve the `` boost factor '' of @xmath19 is to make the dm annihilation cross section velocity - dependent . in this case
the annihilation cross section in the early universe is not same as that in the galaxy or elsewhere , simply because typical velocity of the dm particle varies from place to place .
hence it is in principle possible that the dm has canonical annihilation cross section at the freezeout epoch in the early universe reproducing the dm abundance observed by wilkinson microwave anisotropy probe ( wmap ) , while explaining the cosmic - ray positron / electron excesses .
a common mechanism would be sommerfeld enhancement of annihilation cross section @xcite .
if a dm interacts with a light particle through which it annihilates , non - perturbative effects enhance the annihilation cross section .
the cross section is enhanced by inverse of the dm velocity , @xmath20 or @xmath21 , in this class of models . in the breit - wigner enhancement scenario on the other hand , dm annihilates through @xmath22-channel resonance where a particle in the intermediate state has a mass close to two times dm mass @xcite . in this case
the dm cross section can scale as @xmath23 at an earlier time or @xmath21 at a later time .
in these models the annihilation cross section increases as the temperature decreases in the early universe , and hence dm continues to inject high energy particles through the cosmic history .
therefore , it is quite non - trivial whether these models satisfy constraints from big - bang nucleosynthesis ( bbn ) and cosmic microwave background ( cmb ) . in the case of velocity - independent annihilation cross section , bounds from bbn @xcite and cmb @xcite were derived in previous works . in this paper
, we extend the analysis to the velocity - dependent annihilation cross section and derive general upper bound on the annihilation cross section .
this paper is organized as follows . in sec .
[ sec : dm ] a simple prescription for treating the velocity - dependence of dm annihilation cross section is described . in sec .
[ sec : constraints ] we present constraints from bbn and cmb and give implications on dm models .
[ sec : conc ] is devoted to conclusions and discussion .
below we briefly give examples of dm with velocity - dependent annihilation cross section .
after that we will explain our unified treatment for describing the cosmological effects from dm annihilation with velocity - dependent annihilation cross section .
a dm particle @xmath24 is assumed to have an interaction with @xmath25 , which may be a scalar or gauge boson with coupling constant @xmath26 , whose mass is much lighter than the dm mass : @xmath27 .
let us consider the dm annihilation process mediated by @xmath25 exchanges .
if the mass of @xmath25 is sufficiently small , the @xmath25-mediated interaction can be regarded as a long - range force and such an annihilation cross section receives an enhancement @xmath22 compared with tree - level perturbative expression @xcite , @xmath28 where @xmath29 is the initial dm velocity in the center of mass frame .
thus the dm annihilation cross section is proportional to @xmath30 for @xmath31 .
this @xmath30 enhancement saturates at @xmath32 .
there is another interesting effect caused by the bound state formation , which resonantly enhances the dm annihilation rate for some specific dm mass @xcite .
it is known that the enhancement is proportional to @xmath21 near the zero - energy resonance , and this @xmath21 behavior also saturates due to the finite width of the bound state . in the breit - wigner enhancement scenario ,
dm particles annihilate through @xmath22-channel particle exchange ( @xmath25 ) , where the mass of @xmath25 , @xmath33 , is close to @xmath34 .
the square amplitude of this @xmath22-channel process is proportional to @xmath35 where @xmath36 and @xmath37 are defined as @xmath38 and @xmath39 ( @xmath40 is the decay width of @xmath25 ) , respectively .
if @xmath36 and @xmath37 are much smaller than unity , we have @xmath41 in the limit @xmath42 $ ] . at smaller velocity ,
it becomes proportional to @xmath21 . finally , in sufficiently small @xmath29 the cross section saturates at a constant value .
some dm models have velocity - dependent annihilation cross section as described above . in order to treat the effects of velocity - dependence , we phenomenologically parametrize the annihilation cross section as @xmath43 where @xmath44 is a constant
, @xmath29 is the ( thermal - averaged ) velocity of dm particle , and @xmath45 is the velocity at the freezeout of dm annihilation in the velocity - independent case .
typically , the freezeout temperature is given by @xmath46 @xcite , which gives @xmath47 ; in our numerical calculations , we take @xmath48 independently of @xmath49 although the freezeout epoch may deviate from @xmath50 given above . in eq . , @xmath51 is a dimensionless parameter which determines the cutoff below which the velocity - dependence disappears . since we are interested in the case that @xmath52 , we recover @xmath53 in the limit @xmath54
. the power law index @xmath49 and the cutoff parameter @xmath51 depend on models .
the sommerfeld enhancement predicts @xmath55 and @xmath56 , while it also predicts @xmath57 in the zero - energy resonance region . in the breit - wigner enhancement
, the dm annihilation cross section reduces to the form ( [ sigmav ] ) with @xmath58 and @xmath59/v_0 ^ 4 $ ] in the limit @xmath60 $ ] . in another limit @xmath61 $ ] , it is of the form with @xmath57 and @xmath62/(2\delta v_0 ^ 2)$ ] , after rescaling @xmath63 . when the annihilation cross section with @xmath58 , the dm annihilation cross section is large enough to reduce the dm number density significantly even below the freezeout temperature .
thus , we consider the cases of @xmath55 and @xmath64 in the following analysis , and it would give conservative bounds on the breit - wigner enhancement . with the annihilation cross section being given ,
the annihilation term in the boltzmann equation , which governs the evolution of the number density of dm @xmath65 , is given by @xmath66_{\rm ann } = - n_{\rm dm}^2 \langle \sigma v\rangle.\end{aligned}\ ] ] in deriving constraints from bbn and cmb , spectra of injected energy per unit time are needed for all the daughter particles . for the particle species @xmath67 ,
such a quantity is given by @xmath68_{\rm ann } = \frac12 n_{\rm dm}^2 \langle \sigma v\rangle \frac{d n_i}{de},\end{aligned}\ ] ] where @xmath69 is the energy spectrum of @xmath67 from the pair annihilation of dm .
the energy spectra of decay products depend on the property of dm ; for a given decay process , we calculate @xmath69 by using pythia package @xcite .
in order for a qualitative understanding of the effects of velocity - dependent cross section , it is instructive to consider the total energy injection @xmath70 in typical cosmic time , which is @xmath71 , with @xmath72 being the expansion rate of the universe . for this purpose , let us define @xmath73 where @xmath74 is the total release of visible energy in one pair - annihilation process of dm , and @xmath75 is the entropy density . because we consider the case that the cosmic expansion is ( almost ) unaffected by the dm annihilation , the quantity @xmath76 is approximately proportional to the amount of injected energy in a comoving volume per hubble time .
numerically , we obtain @xmath77 , \label{deltarho}\end{aligned}\ ] ] where @xmath78 is the cosmic temperature . in addition , for the convenience of the following discussion , we have introduced the enhancement factor @xmath79 we have set the present dm energy density to be consistent with the wmap observation , @xmath80 @xcite . in the standard thermal relic scenario , @xmath81 is fixed once we fix the dm abundance , though it deviates from the canonical value @xmath82 due to the velocity dependence around the dm freezeout epoch , especially in the case of @xmath57 @xcite . in the following we derive upper bound on @xmath81 with fixed dm abundance . in the case of velocity - independent cross section ( where @xmath83 )
, it is evident that the energy injection per comoving volume decreases as @xmath78 decreases .
this is natural since the dm number density decreases as the universe expands . in the case of the velocity - dependent cross section , however , some non - trivial features appear .
first note that the velocity is estimated as @xmath84 where @xmath85 denotes the temperature at the _
kinetic decoupling_.precisely speaking , there appears a dependence on the relativistic degrees of freedom @xmath86 for @xmath87 , as @xmath88 .
we have taken into account this correction .
below this temperature , a dm particle can not maintain kinetic equilibrium with thermal plasma , and hence it propagates freely and loses its momentum only adiabatically by the hubble expansion .
notice that typical kinetic decoupling temperature for wimp dm is much smaller than the freezeout temperature ( we may say @xmath89 kev
mev for wimp dm candidates ) @xcite .
thus it is found that , for @xmath87 and @xmath90 , the energy injection ( [ deltarho ] ) is constant , or increases as @xmath78 decreases as long as the velocity dependence is not saturated . in fig.[fig : rhos ] , we plot @xmath91 as a function of time for @xmath55 ( top ) and @xmath57 ( bottom ) .
red solid lines correspond to @xmath92 mev , and green dashed lines correspond to @xmath93 kev , for @xmath94
@xmath95 from bottom to top .
@xmath92 kev approximately corresponds to a lower bound on the temperature of kinetic decoupling in order not to suppress the density fluctuation for a formation of the lyman-@xmath96 clouds @xcite .
we have taken @xmath97 , @xmath98 tev and @xmath99 .
then , even at @xmath100 0.1 mev , we get an enhancement factor of the order of @xmath101 @xmath102 with @xmath55 ( @xmath57 ) .
this implies that constraints become stronger than the case of usual dm without velocity dependence . in the following
we perform detailed calculations of the effects on bbn and cmb , and derive constraints on the cross section with various choice of @xmath51 and @xmath85 .
( top panel ) and @xmath57 ( bottom panel ) .
red solid lines correspond to @xmath103 mev and green dashed lines correspond to @xmath103 kev , for @xmath104 from bottom to top .
we have taken @xmath105 and @xmath98 tev .
, title="fig : " ] ( top panel ) and @xmath57 ( bottom panel ) .
red solid lines correspond to @xmath103 mev and green dashed lines correspond to @xmath103 kev , for @xmath104 from bottom to top .
we have taken @xmath105 and @xmath98 tev .
, title="fig : " ]
it has been known that injection of high - energy particles which are emitted through the annihilation of long - lived massive particles during / after the big - bang nucleosynthesis epoch ( at a cosmic time @xmath106
@xmath107 sec ) significantly changes the light element abundances @xcite .
however , the effect of the injection highly depends on what particles are injected .
we discuss two possibilities : ( i ) injection of electromagnetic particles and ( ii ) injection of hadronic particles in this section .
the injection of high - energy electromagnetic particles such as photon and electron induces the electromagnetic cascade , which produces a lot of energetic photons .
those photons destroy the background @xmath108he and produce lighter elements such as deuterium ( d ) , tritium ( t ) , @xmath109he , and heavier elements such as @xmath110li nonthermally at @xmath111 sec . in particular there is a striking feature that the @xmath109he to d ratio ( @xmath109he / d ) tends to increase . by comparing to the observed value of @xmath109he / d ,
this gives us the most stringent bound on the annihilation cross section in case of the injection of electromagnetic particles @xcite .
this reaction occurs at the cosmic temperature of @xmath112 mev .
it is notable that this constraint from bbn @xcite is stronger than that on the @xmath113- or @xmath114-distortion from the planck distribution of cmb @xcite . on the other hand ,
the injection scenario of high - energy hadrons such as pion , proton ( @xmath115 ) , neutron ( @xmath49 ) and their antiparticles might be more complicated , but has been understood in detail @xcite .
the emitted high - energy neutron and proton destroy the background @xmath108he and produce d , t , @xmath109he or @xmath110li .
the charged pions , @xmath116 and @xmath117 pairs induce an extra - ordinal interconversion between the background proton and neutron , which makes the neutron to proton ratio ( @xmath118 ) increase
. then this mechanism produces more @xmath108he . in terms of the annihilating dark matter ,
the overproduction of d or the increase of @xmath119 to deuterium ratio ( @xmath120 ) gives us the most stringent constraint on the annihilation cross section @xcite . in the following ,
we perform a detailed calculation of the light - element abundances ; to take account of the injection of hadronic and electromagnetic particles , we follow the procedure given in @xcite . then , comparing the theoretical prediction with the updated observational constraints , we derive precise upper bounds on the annihilation cross section as a function of the dm mass .
next we discuss observational limits on d / h and @xmath109he / d which are adopted in this study . in the previous work @xcite , it was shown that these elements give us more stringent constraints than the others .
the recent observation of the metal - poor qso absorption line system qso q0913@xmath121072 , together with the six previous measurements , leads to value of the primordial deuterium abundance with a sizable dispersion @xcite , @xmath122 compared with data adopted in the previous analyses in refs .
@xcite , the error of @xmath123 has been reduced by about @xmath124 .
we adopt an upper limit on @xmath125 which is recently observed in protosolar clouds @xcite , @xmath126 this value was also used in ref .
@xcite . here
we discuss the case of an electromagnetic annihilation modes into electron and/or photon .
it is notable that the total amount of energies into electromagnetic modes approximately determines the bound , independently of the detail of each mode . in fig .
[ fig : bbn_rad ] we plot the upper bounds on the annihilation cross section obtained from the observational limit on @xmath127he / d , with @xmath55 ( top ) and @xmath57 ( bottom ) for various values of @xmath128 @xmath129 . here
the kinetic decoupling temperature is set to be @xmath130 mev .
the dashed line denotes the canonical annihilation cross section ( @xmath131 ) . in the top panel
, we see that the bounds highly depend on the cutoff parameter @xmath51 when @xmath132 .
this behavior can be understood from the fact that the production of @xmath133 becomes most efficient when @xmath134 ; at such a temperature , the enhancement factor is estimated as @xmath135 , which becomes smaller than @xmath136 with the present choice of parameters .
then , when @xmath137 , the cross section is enhanced purely by the factor of @xmath138 . to allow the canonical value of the annihilation cross section for a few tev mass of dark matter , we need @xmath139 at least . in the case of @xmath57 which is plotted in the bottom panel of fig . [ fig : bbn_rad ] , @xmath140 is much smaller than @xmath51 everywhere in this parameter space .
therefore @xmath138 determines the enhancement of the annihilation cross section , and there exists a simple scaling law for the line of the limits , which means that the upper bound is proportional to @xmath51 .
this feature is slightly different in case of @xmath92 kev . because the inverse of the enhancement factor with @xmath55 is the order of @xmath141 , any constraints with @xmath142 is insensitive to @xmath51 . in case of @xmath57 ,
the constraint is same as the bottom panel of fig .
[ fig : bbn_rad ] because of the same reason .
he / d limit with @xmath55 ( top ) and @xmath57 ( bottom ) for various values of @xmath143@xmath129 . here
dm is assumed to annihilate purely radiatively into electron and/or photon .
the kinetic decoupling temperature is set to be @xmath130 mev .
the dashed line denotes the canonical annihilation cross section ( @xmath144 ) .,title="fig : " ] he / d limit with @xmath55 ( top ) and @xmath57 ( bottom ) for various values of @xmath143@xmath129 . here
dm is assumed to annihilate purely radiatively into electron and/or photon .
the kinetic decoupling temperature is set to be @xmath130 mev .
the dashed line denotes the canonical annihilation cross section ( @xmath144 ) .,title="fig : " ] , but for the kinetic decoupling temperature set to be @xmath130 kev .
the case of @xmath57 is completely same as the bottom panel of fig .
[ fig : bbn_rad ] . ] when we consider the injection of hadronic particles , the limit is completely different from that of the electromagnetic particles .
the constraint on the overproduction of the deuterium due to the @xmath108he destruction often gives the most stringent constraint @xcite . to study the hadronic injection , hereafter
, we assume dm annihilates into a @xmath145-boson pair as a typical hadronic dm annihilation channel ; in such a case , significant amount of hadrons are produced by the subsequent decay of the @xmath145 bosons produced by the dm annihilation .
constraints do not change much for other cases , such as dm annihilation into @xmath146 @xcite . in fig .
[ fig : bbn_haddh ] we plot the upper bound on the annihilation cross section obtained from the observational limit on d / h with @xmath55 ( top ) and @xmath57 ( bottom ) .
the kinetic decoupling temperature is set to be @xmath130 mev .
first let us consider the case of @xmath55 .
because the hadrodissociation processes become most effective at @xmath147 mev , for which the enhancement factor is estimated as @xmath148 , the constraint is determined only by the value of @xmath51 if @xmath149 .
to agree with the canonical annihilation cross section , we have to assume @xmath150 for a few tev mass of dark matter .
if the kinetic decoupling occurs at around 1 kev , the enhancement factor behaves differently from the case of @xmath93 mev because the hadrodissociation occurs before the time of the kinetic decoupling . as is shown in fig .
[ fig : bbn_haddhkev ] , then the enhancement factor is estimated to be @xmath151 at @xmath152 mev , from which we easily find that the constraint is independent of @xmath51 for @xmath153 when we consider @xmath57 , the cutoff parameter determines the upper bound everywhere in the current parameter space for both @xmath154 1 mev and 1 kev .
the result is shown in the bottom panel of fig .
[ fig : bbn_haddh ] as a representative of both cases .
although so far we have discussed the limit from d / h , consideration of other light elements sometimes tighten the constraint .
in particular the limit from @xmath109he / d by the photodissociation of @xmath108he could also give us stronger limits .
note that even in the annihilation into quarks and gluons , the photodissociation occurs because a sizable amount of the electromagnetic particles is also injected as decay products .
for example , the electromagnetic energy corresponds to @xmath155 of the total energy in case of the annihilation into a @xmath145-boson pair @xcite . in refs .
@xcite , it has been shown that the upper bound from @xmath109he / d due to the photodissociation accompanied with the hadronic annihilation is much weaker than that from d / h when the annihilation cross section does not depend on @xmath29 . the bound on @xmath51 from d / h was severer than the one from @xmath109he / d by three or four order of magnitude .
on the other hand in case of the @xmath29-dependent cross section , the situation can be altered since the enhancement factor depends on the temperature .
( top ) and @xmath57 ( bottom ) for various values of @xmath156 @xmath129 . here
dm is assumed to annihilate into a @xmath145-boson pair .
the kinetic decoupling temperature is set to be @xmath130 mev .
the dashed line denotes the canonical annihilation cross section ( @xmath157 ) which gives the right amount of the dark - matter relic density.,title="fig : " ] ( top ) and @xmath57 ( bottom ) for various values of @xmath156 @xmath129 . here
dm is assumed to annihilate into a @xmath145-boson pair .
the kinetic decoupling temperature is set to be @xmath130 mev .
the dashed line denotes the canonical annihilation cross section ( @xmath157 ) which gives the right amount of the dark - matter relic density.,title="fig : " ] , but for the kinetic decoupling temperature set to be @xmath130 kev .
the case of @xmath57 is same as the bottom plot of fig .
[ fig : bbn_haddh ] . ] because the constraint on @xmath109he / d is sensitive to the cosmic history at @xmath112 mev , the enhancement factor is the order of @xmath158 for @xmath159 mev with @xmath55 . for a small cutoff parameter @xmath160 , then the constraint from @xmath109he / d can become stronger than that from d / h for @xmath161 tev .
this feature is shown in the top panel of fig .
[ fig : bbn_hadhe3d ] . in this figure
we plot the constraint only from @xmath109he / d , ignoring the d / h constraint .
notice that large amounts of d are produced in most parameter space as is seen from fig .
[ fig : bbn_haddh ] .
since hadro / photo - dissociations of @xmath162he also create @xmath127he , both d and @xmath127he are produced from the standard bbn and hadro / photo - dissociation processes .
for sufficiently large @xmath81 , both d and @xmath127he are produced by the hadrodissociation of @xmath162he at @xmath163mev , and the constraint comes from the additional photodissociation effects at around @xmath164mev .
these mixed processes complicate the behavior of the lines in fig .
[ fig : bbn_hadhe3d ] .
there is no simple scaling law among lines with respect to the line of @xmath29-independent constraint . on the other hand ,
if we take @xmath57 , the bound from @xmath109he / d is always weaker than that of d / h .
this is clearly shown in the bottom panel of fig .
[ fig : bbn_hadhe3d ] . , but for dm annihilating into a @xmath145-boson pair.,title="fig : " ] , but for dm annihilating into a @xmath145-boson pair.,title="fig : " ] , but for dm annihilating into a @xmath145-boson pair with @xmath55 ( top ) and @xmath57 ( bottom ) .,title="fig : " ] , but for dm annihilating into a @xmath145-boson pair with @xmath55 ( top ) and @xmath57 ( bottom ) .,title="fig : " ] we also consider the case of @xmath93 kev .
results are shown in fig .
[ fig : bbn_hadhe3dkev ] . as in the previous case ,
the constraint is mostly from the abundance of d. in some parameter region , however , @xmath133/d gives the most stringent constraint .
this fact is seen in the case of @xmath55 for @xmath165 tev and @xmath166 . in addition , for @xmath57 ( the bottom panel of fig .
[ fig : bbn_hadhe3dkev ] ) , a simple scaling law ( i.e. , the proportionality of the upper bound on @xmath81 to @xmath51 ) breaks down once @xmath51 becomes smaller than @xmath167 ; for such a small value of @xmath51 , the constraint becomes significantly stringent .
this feature can be understood analytically because , for @xmath160 , @xmath76 starts to increase as a function of @xmath168 after @xmath169 sec ( @xmath170 kev ) .
such a behavior is clearly seen as the dashed line in the bottom panel of fig .
[ fig : rhos ] . before closing this subsection
, we comment on the constraints from the li abundances .
photo / hadro - dissociation processes also modify abundances of @xmath171li and @xmath172li .
however , constraints from observations of @xmath172li / h and @xmath171li / h are weaker than those from d / h and/or @xmath127he / d for annihilating dm @xcite .
energy injection around the recombination epoch affects the cmb anisotropy @xcite .
this is because injected energy can ionize neutral hydrogens and modify the standard recombination history of the universe .
the effect is characterized by the quantity @xmath173 , which represents the fraction of injected electron ( photon ) energy @xmath174 for @xmath175 at the redshift @xmath176 used for ionization of the hydrogen atom at the redshift between @xmath177 and @xmath178 .
the evolution equation of the ionization fraction of the hydrogen atom , @xmath179 , includes the following additional term , @xmath180_{\rm dm } = \int \frac{dz'}{h(z')(1+z')}\frac{n_{\rm dm } ^2(z ' ) \langle \sigma v\rangle}{n_h(z ' ) } \frac{m}{e_{\rm ry}}\frac{d\chi_{\rm ion}^{(f)}(m , z',z)}{dz},\ ] ] where @xmath181ev is the rydberg energy , @xmath182 is the number density of the hydrogen atom and @xmath183.\ ] ] here @xmath184 denotes the final state of the dm annihilation , e.g. , @xmath185 , @xmath186 , etc . ,
and @xmath187 denotes the energy spectrum of the electron ( photon ) generated from the cascade decay of @xmath184 . in the case of @xmath188
, we have @xmath189 . for general final state @xmath184
, it is evaluated by the pythia code @xcite .
we follow the methods described in refs .
@xcite to compute @xmath190 .
this term is included in the recfast code @xcite implemented in the camb code @xcite for calculating the cmb anisotropy .
( dotted ) and @xmath191 ( dashed ) for @xmath98tev and assuming dm annihilation into @xmath192 with velocity - independent annihilation cross section .
also shown are data points from wmap , quad , acbar and cbi . ]
additional energy injections from dm annihilation around the recombination epoch cause ionization of neutral hydrogen atoms .
thus the effect is to slow down the recombination of the universe . as a result
, anisotropies in cmb are dumped at small scales due to the increase in thickness of the last scattering surface .
[ fig : cl ] shows the @xmath193 power spectrum of the cmb temperature anisotropy , with / without dm annihilation effect .
the solid line corresponds to the best - fit @xmath194cdm model without dm annihilation , and dotted line to dm annihilation cross section into @xmath192 with @xmath195 and dashed line to dm annihilation cross section with @xmath196 , while all cosmological parameters are fixed .
here we have taken @xmath98tev with velocity - independent annihilation cross section .
it is seen that dm annihilation effects suppress the @xmath193 spectrum , reflecting the increase in thickness of the last scattering surface .
it is not hard to imagine that this effect has a degeneracy with other cosmological parameters . in particular , the increase of the reionization optical depth
causes similar effects . in order to derive conservative bounds on the dm annihilation cross section
, we must take into account degeneracies between dm annihilation effect and other cosmological parameters .
we have derived 2@xmath197 constraints using a profile likelihood function where the other cosmological parameters including the six standard ones ( @xmath198 in the notation of ref .
@xcite ) and the amplitude of the sunyaev - zeldovich effect are marginalized so that the original likelihood function is maximized for given dm annihilation cross section and mass .
the likelihood surface is scanned by using the cosmomc code @xcite ; in our analysis , we have modified the cosmomc code to take account of the above mentioned effect of energy injection .
the used datasets include wmap @xcite , acbar @xcite , cbi @xcite and quad @xcite .
as opposed to bbn constraints , cmb constraint depends on the injected radiative energy , hence purely leptonic annihilation is more strongly constrained than the hadronic one . the result is presented in fig .
[ fig : cmb ] where we plot upper bounds on the annihilation cross section obtained from cmb anisotropy data as a function of dm mass for @xmath199 @xmath200 .
dm is assumed to annihilate into @xmath192 pair in the top panel and @xmath186 in the bottom panel .
here we have taken @xmath55 and @xmath201 .
we have checked that the results do not change for @xmath57 and/or @xmath202 .
this is because the cmb constraint is sensitive to the annihilation rate at around the recombination epoch , @xmath203 ev , and hence the annihilation cross section is already saturated for most interesting range of @xmath51 for both @xmath55 and @xmath57 . comparing them with bbn constraints , it is found that the cmb constraint is severer for the leptonic annihilation case independently of the parameters . in the case of hadronic annihilation with @xmath204 a few tev ,
the situation is not so simple . for @xmath55 and @xmath93mev
, cmb gives weaker constraint than bbn for @xmath205 , as seen from fig .
[ fig : bbn_haddh ] , but becomes tighter for @xmath206 .
the situation is similar for @xmath93kev .
this is because the bbn constraint from the observation of d / h is sensitive to the annihilation at @xmath163mev , and the annihilation cross section do not saturate at that epoch for small @xmath51 for @xmath55 . on the other hand , for @xmath57 , bbn gives tighter constraint than the cmb for parameter ranges shown in the figures .
therefore , cmb takes complementary role to bbn in constraining the dm annihilation with velocity - dependent annihilation cross section .
dm is assumed to annihilate into @xmath192 pair in the top panel and @xmath186 in the bottom panel .
here we have taken @xmath55 and @xmath207 .
results do not change for @xmath57 and/or @xmath208 .
, title="fig : " ] @xmath200 .
dm is assumed to annihilate into @xmath192 pair in the top panel and @xmath186 in the bottom panel .
here we have taken @xmath55 and @xmath207 .
results do not change for @xmath57 and/or @xmath208 .
, title="fig : " ]
in this paper we have investigated effects of dm annihilation on bbn and cmb , and derived constraints on the dm annihilation rate , particularly focusing on the case where the annihilation cross section has a velocity - dependent structure .
this is partly motivated by the observations of cosmic - ray positron / electron excesses and their explanations by the dm annihilation contribution .
we phenomenologically parametrized the velocity - dependence of the annihilation cross section and the critical velocity at which such an enhancement saturates , and derived general constraints on them .
our constraints are applicable to known velocity - dependent dm annihilation models , such as the sommerfeld and breit - wigner enhancement scenarios .
these results have been plotted in figs .
[ fig : bbn_haddh ] [ fig : cmb ] by changing the parameters and observation .
therefore readers can read off the allowed parameter regions from those figures in accordance with the intended use .
some comments are in order .
if the dm annihilation is helicity - suppressed , the @xmath115-wave process may be the dominant mode , as is often the case with majorana fermion dm . in this case
we obtain @xmath209 : @xmath210 . thus the annihilation cross section becomes smaller as the temperature decreases , until the @xmath22-wave process becomes efficient . for negative @xmath49 ,
the bbn / cmb constraints are weaker than the velocity - independent case . in the sommerfeld enhancement scenario ,
it was pointed out that the dm - dm scattering mediated by light particle exchanges causes observationally relevant effect @xcite , and this also gives significant constraint @xcite .
this work is supported by grant - in - aid for scientific research from the ministry of education , science , sports , and culture ( mext ) , japan , no .
20244037 ( j.h . ) , no .
20540252 ( j.h . ) , no .
22244021 ( j.h .
and t.m . ) , no .
14102004 ( m.k . ) , no .
21111006 ( m.k .
, k.k . and k.n .
) , no.22244030 ( k.k . and k.n . ) , no .
18071001 ( k.k . ) , and no .
22540263 ( t.m . ) , and also by world premier international research center initiative ( wpi initiative ) , mext , japan .
k.k . was also partly supported by the center for the promotion of integrated sciences ( cpis ) of sokendai .
99 o. adriani _ et al . _
[ pamela collaboration ] , nature * 458 * , 607 ( 2009 ) [ arxiv:0810.4995 [ astro - ph ] ] .
j. chang _ et al .
_ , nature * 456 * , 362 ( 2008 ) .
a. a. abdo _ et al .
_ [ the fermi lat collaboration ] , phys .
lett . * 102 * , 181101 ( 2009 ) [ arxiv:0905.0025 [ astro-ph.he ] ] ; m. ackermann _ et al .
_ [ fermi lat collaboration ] , phys . rev .
d * 82 * , 092004 ( 2010 ) [ arxiv:1008.3999 [ astro-ph.he ] ] ; f.
aharonian _ et al .
_ [ h.e.s.s .
collaboration ] , phys .
* 101 * , 261104 ( 2008 ) [ arxiv:0811.3894 [ astro - ph ] ] ; f. aharonian _ et al .
_ [ h.e.s.s .
collaboration ] , astron .
astrophys . * 508 * , 561 ( 2009 ) [ arxiv:0905.0105 [ astro-ph.he ] ] .
s. profumo and t. e. jeltema , jcap * 0907 * , 020 ( 2009 ) [ arxiv:0906.0001 [ astro-ph.co ] ] ; a. v. belikov , d. hooper , phys .
d * 81 * , 043505 ( 2010 ) [ arxiv:0906.2251 [ astro-ph.co ] ] ; m. kawasaki , k. kohri and k. nakayama , phys . rev .
d * 80 * , 023517 ( 2009 ) [ arxiv:0904.3626 [ astro-ph.co ] ] ; m. cirelli , p. panci , p. d. serpico , nucl .
b * 840 * , 284 ( 2010 ) [ arxiv:0912.0663 [ astro-ph.co ] ] .
a. a. abdo _ et al .
_ , astrophys .
j. * 712 * , 147 ( 2010 ) [ arxiv:1001.4531 [ astro-ph.co ] ] ; a. a. abdo _ et al .
_ [ fermi - lat collaboration ] , jcap * 1004 * , 014 ( 2010 ) [ arxiv:1002.4415 [ astro-ph.co ] ] ; a. a. abdo , m. ackermann , m. ajello _ et al . _ , phys . rev .
lett . *
104 * , 091302 ( 2010 ) [ arxiv:1001.4836 [ astro-ph.he ] ] .
j. hisano , m. kawasaki , k. kohri and k. nakayama , phys .
d * 79 * , 043516 ( 2009 ) [ arxiv:0812.0219 [ hep - ph ] ] ; j. hisano , k. nakayama and m. j. s. yang , phys .
b * 678 * , 101 ( 2009 ) [ arxiv:0905.2075 [ hep - ph ] ] .
j. liu , p. -f .
yin , s. -h .
zhu , phys .
d * 79 * , 063522 ( 2009 ) [ arxiv:0812.0964 [ astro - ph ] ] .
d. spolyar , m. buckley , k. freese , d. hooper and h. murayama , arxiv:0905.4764 [ astro-ph.co ] ; m. r. buckley , k. freese , d. hooper , d. spolyar and h. murayama , phys .
d * 81 * , 016006 ( 2010 ) [ arxiv:0907.2385 [ astro-ph.he ] ] ; a. e. erkoca , m. h. reno and i. sarcevic , phys .
d * 82 * , 113006 ( 2010 ) [ arxiv:1009.2068 [ hep - ph ] ] .
r. abbasi , _ et al . _
[ icecube collaboration ] , [ arxiv:1101.3349 [ astro-ph.he ] ] .
j. hisano , s. matsumoto and m. m. nojiri , phys .
lett . * 92 * , 031303 ( 2004 ) [ arxiv : hep - ph/0307216 ] ; j. hisano , s. matsumoto , m. m. nojiri and o. saito , phys .
d * 71 * , 063528 ( 2005 ) [ arxiv : hep - ph/0412403 ] .
j. hisano , s. matsumoto , o. saito and m. senami , phys .
d * 73 * , 055004 ( 2006 ) [ hep - ph/0511118 ] .
n. arkani - hamed , d. p. finkbeiner , t. r. slatyer and n. weiner , phys .
d * 79 * , 015014 ( 2009 ) [ arxiv:0810.0713 [ hep - ph ] ] ; i. cholis , d. p. finkbeiner , l. goodenough , and n. weiner , jcap * 0912 * , 007 ( 2009 ) [ arxiv:0810.5344 [ astro - ph ] ] .
m. h. reno and d. seckel , phys .
d * 37 * , 3441 ( 1988 ) .
j. a. frieman , e. w. kolb and m. s. turner , phys .
d * 41 * , 3080 ( 1990 ) .
k. jedamzik , phys .
d * 70 * , 083510 ( 2004 ) [ arxiv : astro - ph/0405583 ] .
j. hisano , m. kawasaki , k. kohri and k. nakayama , phys .
d * 79 * , 063514 ( 2009 ) [ erratum - ibid .
d * 80 * , 029907 ( 2009 ) ] [ arxiv:0810.1892 [ hep - ph ] ] ; j. hisano , m. kawasaki , k. kohri , t. moroi and k. nakayama , phys .
d * 79 * , 083522 ( 2009 ) [ erratum - ibid .
d * 80 * , 029905 ( 2009 ) ] [ arxiv:0901.3582 [ hep - ph ] ] .
n. padmanabhan and d. p. finkbeiner , phys .
d * 72 * , 023508 ( 2005 ) [ arxiv : astro - ph/0503486 ] .
s. galli , f. iocco , g. bertone and a. melchiorri , phys .
d * 80 * , 023505 ( 2009 ) [ arxiv:0905.0003 [ astro-ph.co ] ] ; g. huetsi , a. hektor and m. raidal , astron .
astrophys .
* 505 * , 999 ( 2009 ) [ arxiv:0906.4550 [ astro-ph.co ] ] ; m. cirelli , f. iocco and p. panci , jcap * 0910 * , 009 ( 2009 ) [ arxiv:0907.0719 [ astro-ph.co ] ] ; t. r. slatyer , n. padmanabhan and d. p. finkbeiner , phys .
d * 80 * , 043526 ( 2009 ) [ arxiv:0906.1197 [ astro-ph.co ] ] ; q. yuan , b. yue , x. bi , x. chen and x. zhang , jcap * 1010 * , 023 ( 2010 ) [ arxiv:0912.2504 [ astro-ph.co ] ] . m. kawasaki , t. moroi and t. yanagida , phys .
b * 370 * , 52 ( 1996 ) [ arxiv : hep - ph/9509399 ] .
j. hisano , k. kohri and m. m. nojiri , phys .
b * 505 * , 169 ( 2001 ) [ arxiv : hep - ph/0011216 ] .
s. profumo , k. sigurdson and m. kamionkowski , phys .
* 97 * , 031301 ( 2006 ) [ arxiv : astro - ph/0603373 ] .
t. bringmann and s. hofmann , jcap * 0407 * , 016 ( 2007 ) [ arxiv : hep - ph/0612238 ] .
a. loeb , m. zaldarriaga , phys .
d * 71 * , 103520 ( 2005 ) [ astro - ph/0504112 ] .
k. jedamzik and m. pospelov , new j. phys .
* 11 * , 105028 ( 2009 ) [ arxiv:0906.2087 [ hep - ph ] ] .
m. kawasaki , k. kohri and t. moroi , phys .
d * 71 * , 083502 ( 2005 ) , [ arxiv : astro - ph/0408426 ] ; phys .
b * 625 * , 7 ( 2005 ) , [ arxiv : astro - ph/0402490 ] .
j. zavala , m. vogelsberger and s. d. m. white , phys .
d * 81 * , 083502 ( 2010 ) [ arxiv:0910.5221 [ astro-ph.co ] ] .
k. jedamzik , phys .
d * 74 * , 103509 ( 2006 ) [ hep - ph/0604251 ] .
x. l. chen and m. kamionkowski , phys .
d * 70 * , 043502 ( 2004 ) [ arxiv : astro - ph/0310473 ] .
t. kanzaki and m. kawasaki , phys . rev .
d * 78 * , 103004 ( 2008 ) [ arxiv:0805.3969 [ astro - ph ] ] .
s. seager , d. d. sasselov and d. scott , astrophys . j. * 523 * , l1 ( 1999 ) [ arxiv : astro - ph/9909275 ] ; astrophys .
j. suppl .
* 128 * , 407 ( 2000 ) [ arxiv : astro - ph/9912182 ] .
a. lewis , a. challinor and a. lasenby , astrophys . j. * 538 * , 473 ( 2000 ) [ arxiv : astro - ph/9911177 ] .
j. dunkley _ et al . _
[ wmap collaboration ] , astrophys .
j. suppl .
* 180 * , 306 ( 2009 ) [ arxiv:0803.0586 [ astro - ph ] ] .
a. lewis and s. bridle , phys .
d * 66 * , 103511 ( 2002 ) [ arxiv : astro - ph/0205436 ] . c. l. reichardt _ et al .
_ , astrophys .
j. * 694 * , 1200 ( 2009 ) [ arxiv:0801.1491 [ astro - ph ] ] . j. l. sievers _ et al . _ ,
arxiv:0901.4540 [ astro-ph.co ] . c. pryke _ et al .
_ [ quad collaboration ] , astrophys .
j. * 692 * , 1247 ( 2009 ) [ arxiv:0805.1944 [ astro - ph ] ] ; m. l. brown _ et al .
_ [ quad collaboration ] , astrophys .
j. * 705 * , 978 ( 2009 ) [ arxiv:0906.1003 [ astro-ph.co ] ] .
j. l. feng , m. kaplinghat and h. b. yu , phys .
lett . * 104 * , 151301 ( 2010 ) [ arxiv:0911.0422 [ hep - ph ] ] ; m. r. buckley and p. j. fox , phys .
d * 81 * , 083522 ( 2010 ) [ arxiv:0911.3898 [ hep - ph ] ] ; c. arina , f. -x .
josse - michaux , n. sahu , phys .
b * 691 * , 219 - 224 ( 2010 ) [ arxiv:1004.0645 [ hep - ph ] ] . j. l. feng , m. kaplinghat and h. b. yu , phys .
d * 82 * , 083525 ( 2010 ) [ arxiv:1005.4678 [ hep - ph ] ] .
d. p. finkbeiner , l. goodenough , t. r. slatyer , m. vogelsberger and n. weiner , [ arxiv:1011.3082 [ hep - ph ] ] . | we derive cosmological constraints on the annihilation cross section of dark matter with velocity - dependent structure , motivated by annihilating dark matter models through sommerfeld or breit - wigner enhancement mechanisms . in models with annihilation cross section increasing with decreasing dark matter velocity ,
big - bang nucleosynthesis and cosmic microwave background give stringent constraints . icrr - report-580 - 2010 - 13 + ipmu 11 - 0015 + kek - th-1442 + ut-11 - 05 + 1.35 cm * cosmological constraints on dark matter models with velocity - dependent annihilation cross section * 1.2 cm junji hisano@xmath0 , masahiro kawasaki@xmath1 , kazunori kohri@xmath2 , takeo moroi@xmath3 , + kazunori nakayama@xmath4 and toyokazu sekiguchi@xmath5 0.4 cm
_ @xmath6department of physics , nagoya university , nagoya 464 - 8602 , japan _ + _
@xmath7institute for cosmic ray research , university of tokyo , kashiwa 277 - 8582 , japan _ + _
@xmath8institute for the physics and mathematics of the universe , university of tokyo , kashiwa 277 - 8568 , japan _ + _
@xmath9cosmophysics group , theory center , ipns , kek , tsukuba , 305 - 0801 , japan _
+ _ @xmath10department of particle and nuclear physics , the graduate university for advanced studies , tsukuba , 305 - 0801 , japan _ + _ @xmath11department of physics , tohoku university , sendai 980 - 8578 , japan _ + _ @xmath12department of physics , university of tokyo , bunkyo - ku , tokyo 113 - 0033 , japan _ |
a key issue for a proper description of nuclear burning processes in the sun is an accurate treatment of chain reaction kinetics in the solar core plasma .
the major nuclear inputs in kinetics simulations are rate parameters ( reactivities ) for reactions forming the @xmath2 chain and carbon - nitrogen - oxygen ( cno ) cycle in the sun .
it is known that standard solar models ( ssms ) rely on a nuclear reaction network operating with thermal reactivities for particles having maxwellian velocity distribution functions . at the same time , a question whether non - maxwellian distortions of particle distributions may appear and affect reaction rates in the solar core has still been poorly studied .
clayton et al .
@xcite analyzed how some non - maxwellian distortions can alter the generation of solar neutrinos . in the context of the @xmath12b neutrino problem ,
the authors introduced a radical _ ad hoc _ assumption that ion distributions have depleted high - energy tails and thus depart from a maxwellian function .
this model _ decreases _ the contribution to reaction rates of fast particles with energies large compared to the thermal energy . although the model was shown to provide a pathway toward a solution of the neutrino problem , later bahcall argued @xcite that the degree of depletion needed is unlikely to occur in the dense core . in the meantime ,
other non - maxwellian phenomena which _ increase _ the contribution of fast particles can manifest in the solar core plasma .
these are suprathermal processes triggered by fast nucleons and lightest nuclei naturally appearing in the plasma .
these particles are predominantly generated in exoergic reactions of the @xmath2 chain and usually have mev energies . when slowing down in the plasma they can undergo in - flight nuclear reactions and thus contribute to total reaction rates .
the level of this suprathermal effect depends on a particular reaction .
for exoergic processes , candidates most appropriate for the effect to manifest are reactions with pronounced resonances at suprathermal energies . in turn , direct reactions with cross sections moderately depending on energy are less sensitive to fast particles . as for endoergic processes ,
most of them can be appreciably influenced by fast particles having energies in excess of reaction threshold @xmath13 .
one may expect therefore that energetic particles provide different impacts on forward ( exoergic ) @xmath14 and reverse ( endoergic ) @xmath15 reactions .
this can lead to a deviation of the relation between forward and reverse reactivities , @xmath16 and @xmath17 , from a standard law for maxwellian plasma @xmath18 in eq .
( [ eq : rev - for ] ) , the quantity @xmath19 is the q - value of the forward reaction , @xmath20 is the plasma temperature in units of energy , @xmath21 and @xmath22 are algebraic combinations of particle mass numbers @xmath23 and temperature - dependent partition functions @xmath24 , respectively ( for details , see @xcite ) .
the violation of this relation caused by non - maxwellian particles in astrophysical plasmas was recently obtained for several main processes of big bang nucleosynthesis @xcite and for some reactions of the solar cno cycle @xcite . in particular , it was shown that 8.674-mev @xmath0 particles produced in the @xmath25 reaction of the @xmath2 chain in the sun crucially affect the balance of @xmath26 reactions , so that the reverse @xmath27 process can even block the forward @xmath28 one , distorting running of the cno cycle .
this particular finding suggests the role of fast reaction - produced ions in the solar core may be unappreciated in ssms , and in a greater or lesser degree it may concern processes of the @xmath2 chain and cno cycle .
furthermore , if we find that suprathermal reactions are capable of altering some element abundances in the core , suprathermal corrections to the fluxes of some solar neutrinos may also be obtained . in this context
, it is worthwhile to note that the cno cycle subdominant for energy generation in the sun is an important source of neutrinos released in the @xmath29-decay of @xmath30n , @xmath31o , and also @xmath7f nuclei .
the cno neutrinos carry valuable information on solar core metallicity @xcite , and may also provide an independent test on the core temperature determined earlier on the basis of sudbury neutrino observatory and super - kamiokande experiments .
all this serves as a clear argument in favor of a further study of non - maxwellian ( suprathermal ) reactions in the solar interior . in this paper , a consistent model to properly describe these processes
is formulated .
the model is based on a formalism of in - flight reaction probability , operates with different approaches for the treatment of fast particle slowing - down in a matter , and takes into account some plasma peculiarities , such as the effect of electron degeneracy on in - flight reaction rates and electron screening of thermal reactions in the core . on
the example of the cno cycle we will demonstrate in detail how the model works and will improve results on @xmath0-particle - induced suprathermal processes published previously @xcite .
let us consider non - relativistic fast particles @xmath32 produced in an exoergic reaction @xmath33 in the core plasma . to examine their effect on some @xmath34 reaction
, one needs to know the particle emission rate @xmath35 , the particle energy loss rate @xmath36 in the matter , and the @xmath34 reaction cross section @xmath37 at thermal as well as suprathermal energies .
the particle emission rate @xmath35 is determined by the following equations @xmath38
@xmath39 @xmath40 in these equations , @xmath41 is the number of particles @xmath32 produced per pair of @xmath42 , @xmath43 is the @xmath44 reaction rate , @xmath45 and @xmath46 are the @xmath44 reactivity and cross section , respectively , @xmath47 ( @xmath48 ) is the number density of species @xmath49 with the velocity distribution function @xmath50 normalized to @xmath47 .
we will use a formalism of in - flight reaction probability to describe the suprathermal @xmath34 reaction .
this probability @xmath51 satisfies the equation @xmath52 the solution of which is @xmath53.\ ] ] it presents the probability that the fast particle @xmath32 undergoes the in - flight @xmath34 reaction while slowing in the plasma from an initial energy @xmath54 down to the thermal energy @xmath55 . in eq .
( [ eq : w ] ) , @xmath56 is the particle energy in the laboratory frame , @xmath57 is the particle mass , @xmath46 is the reaction cross section , @xmath58 is the number density of target nuclei @xmath59 , and @xmath60 is the particle energy loss rate .
a proper choice of this rate is a crucial point for accurate calculations of @xmath51 . in the solar core plasma
fast particles lose energy through coulomb elastic scattering ( coul ) off background charged species , and through nuclear elastic scattering ( nes ) off ambient ions @xmath61 ( neutrons can be ignored ) . accordingly , @xmath62 the nes term in eq .
( [ eq : loss ] ) can be presented in a form @xcite @xmath63 where @xmath64 is the differential cross section for @xmath65 nes ( allowing for the contribution of coulomb - nuclear interference ) , @xmath66 is the cosine of scattering angle in the center - of - mass frame , and @xmath67 ( if @xmath68 ) or 0 ( if @xmath69 ) . in most cases ,
the nes contribution to @xmath60 is subdominant and the majority of particle energy loss comes from coulomb scattering .
its rate can be written as @xmath70 where subscripts @xmath71 and @xmath61 stand for bulk electrons and ions , respectively . to compose a detailed picture of particle thermalization in the sola core plasma
, we will consider different models for coulomb slowing - down process .
the first one is a standard binary - collision model with a debye cut - off described by sivukhin @xcite
. we will refer to it as siv66 . in this model , the partial terms @xmath72 ( @xmath73 ) in eq .
( [ eq : ces ] ) are given by @xmath74 @xmath75 @xmath76 in this equations , @xmath77 and @xmath78 are the number density and the temperature of plasma species @xmath79 , while @xmath80 and @xmath81 ( @xmath82 ) are the mass and the charge number of particle @xmath83 .
the quantity @xmath84 is the coulomb logarithm treated in classical or quantum - mechanical approximations @xmath85 where @xmath86 is the debye shielding length , @xmath87 is the impact parameter for @xmath88 deflection , and @xmath89 is the de broglie wavelength .
they are @xmath90 where @xmath91 and @xmath92 . equations ( [ eq : ces1-a])-([eq : ces1-c ] ) were derived under the assumption of maxwellian velocity distribution functions for all plasma species @xmath79 .
the second model developed by kamelander @xcite is based on the fokker - planck collision theory
. we will refer to it as kam86 .
this model makes it possible to carry out an extended analysis of particle slowing - down in a dense matter .
it operates with maxwellian as well as fermi - dirac distribution functions , taking into account degeneracy of electron component . for
the energy loss rate @xmath72 ( @xmath73 ) one obtains @xcite @xmath93 the quantity @xmath94 is a function of particle velocity @xmath95 .
it is given by @xmath96,\ ] ] where @xmath97 @xmath98 @xmath99 and @xmath100 is the velocity distribution function of plasma species @xmath79 .
if all species are maxwellian , the integrals @xmath101 ( @xmath102 ) are taken analytically and after some algebra eq .
( [ eq : ces2-a ] ) can be reduced to the form of eq .
( [ eq : ces1-a ] ) in which @xmath103 is replaced by @xmath104 @xmath105.\ ] ] under conditions typical of the solar core ( @xmath106 g/@xmath107 and @xmath108 kev ) the electron temperature @xmath109 is close to the fermi energy @xmath110 .
for example , in the inner core @xmath111 and the plasma electrons are in a weakly degenerate state . to incorporate electron degeneracy in our model
, we properly determine the electron distribution function @xmath112 which obeys fermi - dirac statistics @xmath113^{-1}.\ ] ] in eq .
( [ eq : fe - a ] ) , the degeneracy parameter @xmath114 ( @xmath115 with @xmath66 being the chemical potential ) is chosen to satisfy the normalization condition @xmath116 the coulomb logarithm @xmath84 is also corrected to electron degeneracy by using an _ ad hoc _ procedure @xcite of the replacement @xmath109 in eq .
( [ eq : clog - a ] ) by @xmath117 where @xmath118 is the fermi temperature . substituting now the distribution function @xmath112 , eq .
( [ eq : fe - a ] ) into eqs .
( [ eq : ces2-c1])-([eq : ces2-c3 ] ) , one can calculate the energy loss rate @xmath119 , eq .
( [ eq : ces2-a ] ) , through @xmath120 collisions with degenerate electrons .
the third model for energy loss being considered was developed by skupsky @xcite .
we will refer to it as skup77 .
this model is based on a different energy loss mechanism .
it was assumed that a charged particle @xmath32 induces an electric field in a matter , which acts back on this particle and ultimately decreases its kinetic energy .
for this mechanism , the energy loss rate is expressed as @xmath121 where @xmath122 and @xmath123 are the real and imaginary part of the matter dielectric function @xmath124 , respectively .
using the random - phase - approximation ( rpa ) form of the quantum - mechanical dielectric function , skupsky @xcite obtained that the particle energy loss to plasma electrons of arbitrary degeneracy is given by @xmath125 \ln\lambda_\text{rpa},\ ] ] where @xmath126 is the fermi integral to order 1/2 @xmath127 the quantity @xmath128 in eq .
( [ eq : ces3-b ] ) is a generalization of the classical coulomb logarithm .
it has the following form @xmath129^{-1 } dk,\ ] ] where @xmath130 and @xmath131 .
we should note that this formula somewhat differs from @xmath128 presented in @xcite .
equations ( [ eq : ces3-b ] ) and ( [ eq : ces3-c ] ) were derived for the case where the particle velocity @xmath95 is less than the average electron velocity @xmath132 .
it was also assumed that @xmath133 .
et al_. @xcite recently proposed a reduced model for particle energy loss through scattering off plasma electrons
. we will refer to it as evrg13 .
the model conveniently interpolates between limiting cases for @xmath119 based on classical and quantum kinetic equations .
it gives @xmath134 where the function @xmath135 is described by eq .
( [ eq : ces1-b ] ) .
the argument @xmath136 is @xmath137^{1/3}.\ ] ] the approximation formula for @xmath138 has the following form @xmath139 where @xmath140 . for the electron velocity @xmath141 in eq .
( [ eq : ces4-c ] ) we consider the most probable velocity @xmath142 satisfying the equation @xmath143 . with the electron distribution function @xmath112 , eq .
( [ eq : fe - a ] ) , we found that in the solar core @xmath144 .
skup77 and evrg13 operate only with @xmath120 scattering . to obtain the total coulomb energy loss for these models , we have also taken into account @xmath65 scattering for maxwellian ions given by eq .
( [ eq : ces1-a ] ) .
note that , since the above four models are based on different approaches , it becomes possible to give an extended description of particle slowing - down in the solar core plasma .
now we introduce some other informative parameters characterizing fast particles in the solar core .
the particle thermalization range @xmath145 and time @xmath146 in the plasma are @xmath147 @xmath148 since @xmath146 reflects the particle `` lifetime '' in a suprathermal state , the number density of suprathermal particles approximately is @xmath149 , where @xmath35 is the particle emission rate in some @xmath44 reaction , eq .
( [ eq : k - rate ] ) .
one more informative parameter is the effective temperature @xmath150 of non - maxwellian particles .
it can be evaluated by equating the pressure of these particles having a slowing - down distribution function to the pressure of maxwellian particles @xcite .
this procedure gives @xmath151 where @xmath152 is the crossover velocity defined in @xcite .
this temperature determines the average kinetic energy of non - maxwellian particles through a relation @xmath153 .
the quantity of particular importance is the rate of suprathermal @xmath34 reactions induced by fast particles @xmath32 in the solar core . using the in - flight reaction probability @xmath51 defined by eq.([eq : w ] ) , this rate @xmath154 can be evaluated as @xmath155 where @xmath35 is the particle emission rate in the @xmath44 reaction .
equation ( [ eq : rfast ] ) can be presented in a standard form for reaction rate @xmath156 where @xmath157 denotes the effective suprathermal @xmath34 reactivity @xmath158 although @xmath157 is not a conventional reactivity , it reasonably indicates the strength of the suprathermal @xmath34 reaction .
it is possible to generalize eq .
( [ eq : rfast ] ) to the case of suprathermal reactions induced by fast particles @xmath32 having a continuous source energy spectrum @xmath159 .
such particles can be produced , e.g. , in a reaction like @xmath160 .
the generalized form of eq .
( [ eq : rfast ] ) is obtained by folding the in - flight reaction probability @xmath51 over the source energy spectrum .
if this spectrum @xmath159 covers the range of energies @xmath161 , the suprathermal reaction rate is @xmath162^{-1 } \int_{e_\text{m}}^{e_2 } w_{kl}(e'_k \rightarrow e_\text{th } ) s(e'_k ) \,de'_k,\ ] ] where @xmath163 $ ] .
note that for monoenergetic particles with @xmath164 this equation is reduced to eq .
( [ eq : rfast ] ) .
we complete the description of our model by a method of accounting for electron screening of nuclear reactions in the solar core .
as known , electron screening reduces the repulsive potential barrier between reacting ions and thereby enhances reaction cross sections @xmath46 as compared with those @xmath165 for bare nuclei ( for details , see a review paper @xcite and references therein ) . to incorporate this effect in the model
, we use a weak - screening approximation @xcite properly describing electron screening for @xmath166 reactions with @xmath167 @xcite . the respective enhancement factor @xmath168
is given by @xmath169 where @xmath20 is the plasma temperature and @xmath170 is the debye shielding length with a degeneracy correction @xmath171 the quantity @xmath172 accounts for electron degeneracy in the solar core @xcite .
apparently , electron screening does not affect the probability of in - flight reaction @xmath51 , eq .
( [ eq : w ] ) , as this process is induced by energetic particles .
nevertheless , it can still enhance the in - flight reaction rate , eq .
( [ eq : rfast ] ) , due to an increase of the particle emission rate @xmath35 in a screened thermal @xmath44 reaction .
in this section we explore capabilities of our model on the example of suprathermal effects in the solar cno cycle triggered by mev @xmath0 particles produced in some @xmath2 chain reactions .
we consider that plasma electrons and ions have equal temperature @xmath173 .
the radial profiles of temperature and element number densities in the solar core obtained @xcite by running the mesa code @xcite are employed in the present work . the emission rate @xmath174 of fast @xmath0 particles in different regions of the solar core .
] it was recently found @xcite that the major nuclear sources of mev @xmath0 particles in solar core are two reactions of the @xmath2 chain @xmath175 @xmath176 the @xmath0-particle emission rates @xmath174 in these reactions are presented in fig .
[ fig : a - rate ] . shown
are the screened rates based on the corresponding reactivities for bare nuclei taken from the nacre ii compilation @xcite .
the first reaction generates monochromatic @xmath0 particles with energy @xmath177 of 8.674 mev , whereas the second one provides a continuous spectrum of @xmath0 particles with energies up to 4.3 mev .
let us calculate particle slowing - down characteristics introduced in section [ model ] , focusing mainly on more energetic @xmath0 particles from the @xmath25 reaction .
the energy loss rate of a 8.674-mev @xmath0 particle from the @xmath25 reaction in the solar core plasma .
the siv66 and kam86 ( maxw ) models show the results for the maxwellian plasma , while the kam86 model takes into account the effect of electron degeneracy ( see details in text ) . ]
the energy loss rate of an @xmath0 particle as a function of its kinetic energy @xmath178 .
three regions of the solar core are considered : ( a ) inner core at @xmath179 ; ( b ) middle core at @xmath180 ; ( c ) outer core at @xmath181 .
the curves marked with different symbols correspond to the different slowing - down models . ]
figure [ fig : eloss - a ] shows the energy loss rate of a 8.674-mev @xmath0 particle in the solar core plasma at @xmath182 . to correctly clarify the role of electron degeneracy , we present here the results obtained within the _ single _ slowing - down model , eq . [ eq : ces2-a ] , considering first that all plasma species are maxwellian and then taking into account degeneracy of the electron component .
the corresponding curves in fig .
[ fig : eloss - a ] are denoted as kam86 ( maxw ) and kam86 . due to spectral hardening of the fermi - dirac electron distribution ,
that is , an increase of the fraction of high - energy electrons as compared with the maxwellian case , the @xmath0 particle energy loss through @xmath183 collision decreases .
however , the reduction of @xmath184 proves to be at a rather moderate level of not more than several percent as the plasma electrons are weakly degenerate .
the energy loss rate found in the classical siv66 model for maxwellian plasma , eq .
[ eq : ces1-a ] , is also plotted in fig . [
fig : eloss - a ] .
this curve well coincides with that for maxwellian kam86 .
parameters of 8.674-mev @xmath0 particle thermalization in the solar core .
( a ) thermalization range @xmath185 .
( b ) thermalization time @xmath186 . ] figure [ fig : eloss - b ] shows the calculated energy loss rate @xmath187 as a function of @xmath0-particle energy @xmath178 in the inner , middle , and outer core regions . all models being considered
siv66 , kam86 ( maxw ) and kam86 , skup77 , evrg13 were used in these calculations .
it is seen that both siv66 and kam86 models as well as the interpolated evrg13 formula provide comparatively close values for @xmath187 throughout the solar core .
the lowest loss rate is predicted by the kam86 model properly accounting for electron degeneracy .
at the same time , the skup77 model , based on a completely different mechanism for energy loss , leads to a sizable increase of @xmath187 at high energies . in view of this
, we remind the reader that this model was developed for the case where the @xmath0-particle velocity @xmath188 is less than the average electron velocity @xmath132 . for 8.674-mev @xmath0 particles @xmath189 is close to unity , so one should bear in mind that some inaccuracy in the skup77 predictions at high energies is likely to occur . nevertheless , the limiting cases of the lowest and highest particle energy loss correspond to the kam86 and skup77 models , respectively . the @xmath0-particle thermalization range @xmath190 and time @xmath191 , eqs .
( [ eq : th - range ] ) and eqs .
( [ eq : th - time ] ) , for the limiting cases are presented in fig .
[ fig : therm ] .
both parameters increase towards the outer core in accordance with the decrease of @xmath187 seen in fig .
[ fig : eloss - b ] .
the @xmath0-particle thermalization range @xmath192 @xmath66 m , so all @xmath193-dependent plasma parameters @xmath194 entering eq .
( [ eq : w ] ) can be assumed to be constant .
this simplifies calculations of the in - flight reaction probability @xmath195 .
figure [ fig : temps ] shows the effective temperature @xmath196 , eq .
( [ eq : teff ] ) , of reaction - produced @xmath0 particles with the initial energy @xmath197 mev [ the @xmath1 reaction ] and 4.3 mev [ the upper energy for the @xmath198 reaction ] . as seen , @xmath196 exceeds the core temperature @xmath20 by a factor up to 3 orders of magnitude .
the effective temperature @xmath196 of reaction - produced @xmath0 particles in a comparison with the solar core temperature @xmath20 . the initial @xmath0-particle energy @xmath197 mev [ the @xmath1 reaction ] and 4.3 mev [ the upper energy for the @xmath198 reaction ] . ]
thus , the main characteristics of fast @xmath0 particles are evaluated and now one can examine a role which these particles play in the cno cycle .
the first three branches of the cycle are schematically shown in fig .
[ fig : cno ] .
they involve several exoergic forward @xmath199 reactions , whereas corresponding endoergic reverse @xmath200 processes are neglected in the ssm reaction network .
thresholds @xmath201 of the reverse @xmath11 processes are higher than 1 mev , and according to eq .
( [ eq : rev - for ] ) their maxwellian reactivities @xmath202 in the solar core with the temperature of @xmath203 kev are dramatically suppressed by the factor @xmath204 .
this means that maxwellian @xmath11 nuclear flow does not appear in the cno cycle .
if however fast non - maxwellian @xmath0 particles are produced in the matter , the situation may change significantly .
the first three branches of the solar cno cycle .
the dashed arrow shows the endoergic reverse @xmath205 reaction neglected in the ssm reaction network . ]
we will show this on the example of the branch ii ( the cno - ii cycle ) , focusing on the following processes @xmath206 @xmath207 the importance of the forward @xmath28 reaction , eq .
( [ eq : forward ] ) , is that it closes the branch ii and is one of main processes determining nuclear fusion rates in the cno cycle @xcite .
the reverse @xmath11 reaction , eq .
[ eq : reverse ] , neglected in ssm studies is shown by the dashed arrow in fig .
[ fig : cno ] . for this reaction ,
the @xmath0-particle threshold energy @xmath208 is only 1.531 mev , so the process can easily be activated by 8.674-mev @xmath0 particles from the @xmath25 reaction . to calculate its characteristics
, we used experimental data on the reaction cross section which exhibits a complicated resonant behavior ( see , e.g. , @xcite and references therein ) . the probability of the in - flight @xmath3 reaction in the solar core calculated for the different models of particle energy loss . the @xmath0-particle initial energy @xmath209 mev .
] the influence of electron degeneracy and electron screening on the suprathermal @xmath3 reaction rate @xmath210 .
shown is the ratio @xmath211 ( see details in text ) . ]
figure [ fig : rev - prob ] shows the probability @xmath212 , eq .
( [ eq : w ] ) , for a 8.674-mev @xmath0 particle to undergo the in - flight @xmath3 reaction while slowing down in the solar core plasma .
it is seen that all models for particle energy loss lead to a similar dependence of @xmath212 on radius @xmath193 .
as one might expect , the highest and lowest probabilities are provided by the kam86 and skup77 models , respectively .
using eq .
( [ eq : rfast ] ) , one can convert the in - flight reaction probability @xmath212 to the corresponding suprathermal reaction rate @xmath213 both electron degeneracy and electron screening in the plasma can enhance this rate .
the first mechanism affects the in - flight probability @xmath212 while the second one increases the thermal reactivity @xmath214 determining the fast @xmath0-particle production .
their effect is clarified in fig .
[ fig : ds - effect ] .
it shows the ratio @xmath215 of the suprathermal rate @xmath210 ( both mechanisms are included ) to the rate @xmath216 ( both mechanisms are ignored ) .
the rate enhancement by a factor of @xmath217 is observed at @xmath4 . the suprathermal @xmath3 reaction triggered by fast @xmath0 particles in a comparison with the @xmath8 and @xmath9 processes responsible for @xmath7o burn - up in the solar core .
shown are the @xmath218 and @xmath219 reaction rates determined by eqs .
( [ eq : rfast ] ) and ( [ eq : ij - rate ] ) , respectively .
two @xmath0-particle sources are considered .
the solid curves marked with symbols correspond to 8.674-mev @xmath0 particles from the @xmath1 reaction .
symbols `` @xmath220 '' present the illustrative results for @xmath0 particles from the @xmath198 reaction having the upper energy of 4.3 mev . ]
figure [ fig : rates ] shows the rate @xmath210 in a comparison with the rate of the forward @xmath8 reaction . for completeness , the @xmath9 reaction is also shown in this figure .
note that both @xmath221 processes are responsible for @xmath7o burn - up in the cno cycle .
shown are the screened reaction rates .
the @xmath221 reactivities for bare nuclei were taken from a recent compilation @xcite for temperatures above 10@xmath222 k and extrapolated to lower temperatures @xmath223 k typical of the outer core region , taking into account the @xmath20-dependence of @xmath224 obtained in @xcite . in fig .
[ fig : rates ] , the rate @xmath210 provided by 8.674-mev @xmath0 particles from the @xmath25 reaction is shown by solid curves marked with symbols ( they are not resolved well on a logarithmic scale ) . additionally , we present some estimates for another source of fast @xmath0 particles ( see fig . [
fig : a - rate ] ) .
it is the @xmath198 reaction generating @xmath0 particles in the 04.3 mev range .
the most effective energy for this reaction , i.e , the gamow peak energy is @xmath225 @xcite , so in the solar core region @xmath226 varies within 1422 kev . for such deep subbarrier energies reliable data on the @xmath227 @xmath0-particle spectrum
are not available in the literature that greatly complicates an analysis of the @xmath0-particle contribution to the @xmath3 rate .
however , some useful estimates can be done if we consider that all @xmath0 particles from this reaction have the upper energy of 4.3 mev .
although this case is not realistic , it demonstrates the upper limit for the @xmath198 contribution .
the respective results are shown by symbols `` @xmath220 '' ( for the effective temperature @xmath228 of 4.3-mev @xmath0 particles , see fig .
[ fig : temps ] ) .
we consider these estimates as an argument in favor of a further study of the actual role of this reaction .
figures [ fig : rev - prob ] and [ fig : rates ] demonstrate that the non - maxwellian nuclear effects predicted on the basis of a simplified model @xcite can really occur in the solar core .
one can schematically divide the core into three regions . in the inner core at @xmath229
the reverse @xmath3 reaction is weaker than the forward @xmath8 one and does not perturb the cno running . in the narrow shell at @xmath230 these two reactions
have nearly equal rates , so the resultant nuclear flow between @xmath7o and @xmath231n almost vanishes , making the branch ii unclosed ( see fig .
[ fig : cno ] ) . in the outer core at @xmath232
the @xmath3 reaction enhanced by mev @xmath0 particles becomes much stronger than the @xmath8 process .
accordingly , counter - clockwise nuclear flow in the branch ii is redirected at @xmath7o to clockwise flow in the branch iii .
it is worth noting that the region of this distortion covers @xmath233% of the core with the radius @xmath234 .
the influence of the suprathermal @xmath3 reaction on the @xmath7o number density at @xmath235 . ]
one of possible consequences of this phenomenon is an increase of the @xmath7o abundance in the outer core .
indeed , in this region the rate of @xmath7o built - up through the @xmath3 reaction exceeds the rates of two competing @xmath8 and @xmath9 processes .
it follows from fig .
[ fig : cno ] that the @xmath7o number density satisfies a rate equation @xmath236 where @xmath237 s is the mean lifetime of @xmath7f . the last term in eq .
( [ eq:17o - rate ] ) accounts for the suprathermal built - up of @xmath7o .
we estimate its contribution on the example of the rate equation nearly at steady state . in this case , the number density @xmath238 is related to the thermal number density @xmath239 ( obtained by ignoring the last term in eq .
( [ eq:17o - rate ] ) ) as @xmath240 these number densities are shown in fig .
[ fig:17o ] .
it is seen that the suprathermal @xmath3 reaction is capable of essentially enhancing the @xmath7o abundance in the outer core .
the exact degree of this enhancement however may differ from that displayed in fig .
[ fig:17o ] because @xmath7o may burn not in equilibrium .
nevertheless , these estimates suggest that the @xmath7o abundance predicted by ssms is likely to be underestimated in the outer core region .
in this paper , we have formulated the consistent model for the description of non - maxwellian nuclear processes triggered by fast reaction - produced particles in the solar interior .
it is based on the formalism of in - flight reaction probability , operates with different methods for the treatment of particle slowing - down in the matter , and takes into account some peculiarities typical of nuclear processes in dense plasmas .
these are the influence of electron degeneracy both on charged particle energy loss and suprathermal reaction rates , and electron screening of thermal processes in the solar core plasma .
our model extends the previous approach to study suprathermal solar reactions discussed in @xcite and refines its predictions . to explore capabilities of this model
, it has been applied to calculate the main characteristics of non - maxwellian 8.674-mev @xmath0 particles generated in the @xmath1 reaction of the @xmath2 chain , and to examine suprathermal processes in the solar cno cycle induced by them .
we particularly focused on the @xmath3 reaction neglected in ssm studies .
the effect of electron degeneracy and electron screening on its rate in the solar core has been clarified .
it has been shown that they increase the suprathermal @xmath11 rate by @xmath241% at @xmath4 .
the point of particular importance is that at @xmath242 this rate appreciably exceeds the rates of two conventional reactions
@xmath8 and @xmath9 .
the first one closes the cno - ii cycle and both of them burn up @xmath7o in the core .
this distorts running of the cno cycle so that normal branching @xmath5 of nuclear flow ( closing the branch ii and starting the branch iii ) transforms to abnormal sequential flow @xmath6 .
the region of this distortion covers @xmath233% of the solar core volume . as a result
, the @xmath3 reaction causes an enhancement of the @xmath7o abundance in the core as compared with standard estimates . for the steady state case ,
the enhancement factor in the outer core reaches @xmath10 .
rough estimates also suggest that an additional effect here may come from fast @xmath0 particles produced in the @xmath243 reaction of the @xmath2 chain .
furthermore , we expect that suprathermal @xmath11 reactions other than @xmath3 may also alter abundances of cno elements , including those generating solar neutrinos .
this issue however is addressed to a future publication . in closing , the applicability of the model presented is not restricted to solar studies and it can be used for an analysis of non - maxwellian processes in some other stars , if main sources of energetic particles are accurately identified . | a consistent model for the description of non - maxwellian nuclear processes in the solar core triggered by fast reaction - produced particles is formulated .
it essentially extends an approach to study suprathermal solar reactions discussed previously [ phys .
rev .
c * 91 * , 028801 ( 2015 ) ] and refines its predictions .
the model is applied to examine in detail the slowing - down of 8.7-mev @xmath0 particles produced in the @xmath1 reaction of the @xmath2 chain , and to study suprathermal processes in the solar carbon - nitrogen - oxygen ( cno ) cycle induced by them .
the influence of electron degeneracy and electron screening on suprathermal reactions through in - flight reaction probability and fast particle emission rate is clarified . in particular , these effects account for a 20% increase of the @xmath3 reaction rate at @xmath4 .
this new type of correction is important for the suprathermal reaction like @xmath3 as it is recognized to be capable of distorting the cno cycle in the 95% region of the solar core . in this region ,
normal branching @xmath5 of nuclear flow transforms to abnormal sequential flow @xmath6 , and the @xmath3 reaction rate exceeds the rate of @xmath7o burn up through conventional @xmath8 and @xmath9 processes .
it is shown that these factors can enhance the @xmath7o abundance in the core as compared with standard estimates . for the steady state case ,
the abundance enhancement is estimated to be as high as @xmath10 in the outer core region .
a conjecture is made that other cno suprathermal @xmath11 reactions may also alter abundances of cno elements , including those generating solar neutrinos . |
a graded , infinite , directed graph defines a dynamical system called a bratteli - vershik or adic system ( see @xcite and the references given there for background on adic systems ) .
one of the main problems concerning these systems is to determine their ergodic invariant measures , and the basic tool for attacking this problem is the asymptotics of numbers of paths in the graph with certain properties .
a very interesting bratteli - vershik dynamical system is associated to the eulerian two - dimensional graph with vertices at the points of @xmath0 .
the paths in the eulerian graph may be counted in many different combinatorial ways , but all attempts to extend those counts to the multi - dimensional analogues of the eulerian graph have failed up to this moment ; techniques applied to the two - dimensional eulerian graph , such as the recurrence of random walks in two dimensions in @xcite , the coding of paths by permutations in @xcite , or the left - to - right monotonicity in @xcite , are not available in dimensions higher than two . in this paper
we make a first step in the multi - dimensional direction .
we consider a multi - dimensional analogue of the eulerian graph and determine the asymptotics of the numbers of paths with certain properties .
we plan to apply these asymptotics to studying the ergodic measures on these and related systems in future work . as a rather unexpected corollary of our asymptotics
we get a family of new identities between stirling numbers of the first and second kinds , which are classical objects , arise in a variety of combinatorics problems , and for which many general relations have long been known ( see for example @xcite ) .
now we formulate the results of this paper .
denote by @xmath1
, @xmath2 the ordinary basis vectors in @xmath3 .
let @xmath4 .
we define the graph @xmath5 to be the graph with vertices @xmath6 and the following directed multiple edges .
each vertex @xmath7 is connected just to @xmath8 vertices : to vertex @xmath9 by @xmath10 edges if @xmath11 and to @xmath12 by @xmath13 edges .
such a graph models a random walk on the graph @xmath14 consisting of @xmath8 loops based at a single vertex , with a special kind of opposite reinforcement : when the walker takes a step in the direction of @xmath15 , the numbers of edges in all the other directions ( @xmath16 for @xmath17 ) are incremented by 1 ; but when the walker takes a step in the direction of @xmath16 for some @xmath17 , _ only _ the number of edges in the direction of @xmath15 is incremented by @xmath18 ( cf .
@xcite ) . when @xmath19 and @xmath20 , @xmath5 is the standard two - dimensional eulerian graph .
the corresponding two - dimensional eulerian adic system was introduced in @xcite and studied further in @xcite and @xcite ; see also @xcite
. for each @xmath21 denote by @xmath22 the number of paths in the graph @xmath5 from @xmath23 to @xmath24 .
the standard eulerian numbers ( see @xcite ) are the numbers @xmath25 with @xmath19 . in this paper
, we give a formula for the leading term of the growth rate of @xmath26 as @xmath27 tends to infinity .
we show that if @xmath28 , then @xmath29 as @xmath30 and [ introduction prop-1 and i ] [ b(i ) ] b_c_1 , , c_n+1(i_1,
,i_n ) = _ ff _
j=1^m ( c_f(j ) + c_n+1 + j-1 ) , where @xmath31 denotes the set of all maps @xmath32 such that @xmath33 for @xmath34 . as a corollary
we obtain new identities relating stirling numbers of the first and second kinds ; for @xmath35 , @xmath36 , we get [ introductioncoefs ] r+n - k-1 r s_1(n , r+n - k)= + _
m=0^k m+n - k m+1_i=0^r i+n - k+m-1 i(r - i+1 ) ! + s_2(m+1,r - i+1 ) s_1(n , i+n - k+m ) , where @xmath37 are stirling numbers of the first and second kinds : see theorem [ thm s s ] and corollary [ cor - stirlingidentities ] .
we thank r. stanley for telling us about theorem of frobenius ( theorem [ stanley ] ) and the referee for suggestions leading to the improvement of the exposition .
the numbers @xmath22 of paths in the graph @xmath5 from 0 to @xmath24 satisfy the recurrence relation a_c ( ) = ( c_n+1+i_1+ +i_n ) a_c(i - e_n+1 ) +
_ j=1^n(c_j + i_n+1 ) a_c(-e_j ) .
we are interested in the limit [ eq - bc ] b_c(i_1, ,i_n)=_i_n+1 .
[ prop - bform ] let @xmath38 then [ b ] b_c(u ) = _ s_n _ j_1, ,j_n=0^ _ k=1^n ( c_(k ) + j_1+
+j_k ) ( ) ^j_k . to form a path from @xmath39 to @xmath40 in @xmath5
, we need to take one step in the direction of each @xmath41 , @xmath42 , and @xmath27 steps in the direction of @xmath43 . thus each such path will correspond to a permutation @xmath44 of @xmath45 which will determine when the step in the direction of each @xmath46 is taken , together with a partition @xmath47 of @xmath27 into nonnegative integers , which will determine how many consecutive steps are taken in the @xmath43 direction between steps in other directions .
each time that @xmath48 steps are taken in the @xmath43 direction , all the @xmath49 are incremented by 1 ; and each time that a step is taken in a direction other than @xmath43 , @xmath50 is incremented by 1 .
therefore , & a_c ( 1 , , 1,i_n+1 ) + & = _ s_n_j_1+
+ j_n+1=i_n+1 c_n+1^j_1(c_(1)+j_1)(c_n+1 + 1)^j_2(c_(2)+j_1+j_2 ) + & ( c_(n)+j_1+ +j_n)(c_n+1+n)^j_n+1 + & = _ s_n_j_1+
+ j_n+1=i_n+1 ( c_n+1+n)^j_n+1 + & = ( c_n+1+n)^i_n+1 _ s_n_j_1+ + j_n+1=i_n+1 _ k=1^n ^j_k ( c_(k)+j_1+ +j_k ) + & = ( c_n+1+n)^i_n+1 _ s_n_^i_n+1 _ k=1^n ^j_k
( c_(k)+j_1+ +j_k ) .
this reasoning gives formula ( [ b ] ) .
clearly , @xmath51 is a convergent sum .
similarly , existence of the limit in formula ( [ eq - bc ] ) for any @xmath52 can be shown as follows .
let @xmath53 , @xmath31 , be the same as in section [ sec - intro ] . then & a_c(i_1, ,i_n , i_n+1)= + & = _ f f_j_1+ + j_m+1=i_n+1 c_n+1^j_1(c_f(1)+j_1)(c_n+1 + 1)^j_2(c_f(2)+j_1+j_2 ) + & ( c_f(m)+j_1+ +j_m)(c_n+1+m)^j_m+1 + & = ( c_n+1+m)^i_n+1_f f _ ^i_n+1 _ k=1^m ^j_k ( c_f(k)+j_1+ +j_k ) .
dividing by @xmath54 and letting @xmath30 again leaves a sum of convergent series .
thus we get
[ b i ] b_c(i_1, ,i_n ) = _ ff _ j_1, ,j_n=0^ _ k=1^n ( c_f(k ) + j_1+
+j_k ) ( ) ^j_k .
we shall give two formulas for @xmath51 .
in this section we evaluate @xmath51 using a _ two - dimensional _ formula from @xcite . in section [ subsec - derivsofseries ]
another formula is obtained by using derivatives of geometric series . comparing the two formulas we get new identities involving stirling numbers of the first and second kinds ( corollary [ cor - stirlingidentities ] ) .
[ conj ] we have [ c ] b_c(u ) = _ s_n _ k=1^n(c_(k ) + c_n+1 + k-1 ) .
denote by @xmath55 , the elementary symmetric functions in @xmath56 , _ i ( c ) = _ 1j_1< <j_inc_j_1 c_j_i .
denote @xmath57 . clearly , the function @xmath51 is a symmetric polynomial in @xmath56 of the form [ rn ] b_c(u ) = _
i=0^n i!(n - i ) ! _
i , n(c_n+1 ) = n!_n ( c ) _
n , n(c_n+1 ) + + ( n-1)!_n-1 ( c ) _
n-1,n(c_n+1 ) +
+ n ! _ 0 ( c ) _ 0,n(c_n+1 ) , where @xmath58 , are suitable functions in @xmath50 . for example , _ 0,n(c_n+1 ) & = _ j_1, ,j_n=0^ _ k=1^n ( j_1+ +j_k ) ( ) ^j_k , + _
n , n(c_n+1 ) & = _ j_1, ,j_n=0^ _ k=1^n ( ) ^j_k , and so on . for @xmath59 , b_(c_1,c_2,c_3)(u ) & = _ j_1,j_2=0^ + & ( ) ^j_1 ( ) ^j_2 + & = 2 !
2,2(c_3 ) + ( c_1+c_2)_1,2(c_3 ) + 2 ! _
0,2(c_3 ) , where [ al ] _ 2,2(c_3 ) & = _ j_1,j_2=0^()^j_1 ( ) ^j_2 , + _
1,2(c_3 ) & = _ j_1,j_2=0^(2j_1+j_2)()^j_1 ( ) ^j_2 , + _
0,2(c_3 ) & = _ j_1,j_2=0^j_1(j_1+j_2)()^j_1 ( ) ^j_2 .
our goal is to calculate all the coefficients @xmath60 and show that the result gives formula ( [ c ] ) .
to do so let us calculate the number [ eq_secondproduct ] b_c_1,c_2(n ) = _ j_1, ,j_n=0^ _ k=1^n ( c_1 + j_1+ + j_k ) ( ) ^j_k .
the function @xmath61 has the form : b_c_1,c_2(n ) = c_1^n _ n , n(c_2 ) + c_1^n-1 _
n-1,n(c_2 ) + + _ 0,n(c_2 ) , where the functions @xmath62 , @xmath63 , are the _ same _ functions of one variable as in ( [ rn ] ) .
( to see this , note the identical form of the two products in formulas ( [ b ] ) and ( [ eq_secondproduct ] ) , and note that the sum over the symmetric group is absorbed into the coefficients that do not involve @xmath50 . ) we have b_c_1,c_2(2 ) & = _ j_1,j_2=0^(c_1+j_1)(c_1+j_1+j_2)()^j_1 ( ) ^j_2 + & = c_1 ^ 2 _ 2,2(c_2 ) + c_1_1,2(c_2 ) + _ 0,2(c_2 ) , where the functions @xmath64 are the same functions of one variable as in ( [ al ] ) . to calculate the functions @xmath62 and prove theorem [ conj ] it is enough to calculate @xmath61 .
[ lem main ] we have [ 1 ] b_c_1,c_2(n ) = _ k=1^n ( c_1 + c_2 + k-1 ) .
let us use formula ( 1.1 ) of @xcite to evaluate [ eq-1 ] b_c_1,c_2(n)=_i .
formula ( 1.1 ) for @xmath65 has @xmath8 summands and the leading summand is ( c_2+n)^n+i .
this term after division by @xmath66 gives the right hand side of ( [ eq-1 ] ) .
the lemma is proved .
lemma [ lem main ] gives theorem [ conj ] .
formula ( [ c ] ) for @xmath68 easily gives a formula for @xmath69 for any @xmath52 .
[ prop-1 and i ] let @xmath53 , @xmath31 , be the same as in section [ sec - intro ] .
then [ b(i ) ] b_c_1 , , c_n+1(i_1,
,i_n ) = _ ff _
j=1^m ( c_f(j ) + c_n+1 + j-1 ) .
theorem [ conj ] has as corollaries interesting identities involving stirling numbers of the first and second kinds . a a^-k-1_i+j = k-1 a(i , j)q^i+1(q - a)^j+1 + = a^-k-2_i+j = k-1a(i , j)(i+1 ) q^i+1(q - a)^j+2 + + a^-k-2_i+j = k-1a(i , i)(j+1 ) q^i+2(q - a)^j+1 + = a^-k-2_r+s = ka(r , s-1)(r+1 ) q^r+1(q - a)^s+1 + + a^-k-2_r+s = ka(r-1,s)(s+1 ) q^r+1(q - a)^s+1 + = a^-k-2_r+s = ka(r , s ) q^r+1(q - a)^s+1 .
let @xmath78 denote stirling numbers of the second kind , which count the number of partitions of a @xmath76-element set into @xmath53 nonempty subsets , so that & s_2(1,1)=1 , & + & s_2(2,1)=1 , s_2(2,2)=1 , & + & s_2(3,1)=1 , s_2(3,2)=3 , s_2(3,3)=1 .
& we have s_2(k , m ) = s_2(k-1,m-1 ) + m s_2(k-1,m ) .
we will be interested in the polynomials stirling numbers of the first kind are the coefficients in the expansion ( q)_n = _ k=0^n s_1(n ,
k ) q^k , where @xmath80 is the falling factorial ( q)_n = q(q-1)(q-2)(q - n+1 ) .
we have s_1(n , k ) = _ 1i_1< <i_n - kn-1 ( -1)^n - k i_1i_2i_n - k . for natural numbers
@xmath81 , define a polynomial @xmath82 define @xmath83 .
we have @xmath84 and _ n ,
k(q ) = _ m=0^n - k k-1+mms_1(n , k+m)q^m . recall the notations @xmath85 and @xmath86 defined at the beginning of section [ subsec - derivsofseries ] . denote _
n = _ b=1^n11-u_b , i_n^k(q ) = _ 1i_1 < < i_kn ( _ j=1^k ( d_i_j+d_i_j+1+ + d_n ) ) _
n . for example , i^1_n(q ) = ( nd_n + ( n-1)d_n-1 + + d_1)_n . denote @xmath55 , the elementary symmetric functions in @xmath87 as before . the left hand side of formula ( [ new n ] ) equals _
k!(n - k)!_n - k ( c ) i^k_n(q ) .
the right hand side equals _
k!(n - k)!_n - k ( c ) _
n+1,n - k+1(q ) . by comparing coefficients of the symmetric polynomials in these two equations we get
[ i s ] i^k_n(q ) = _ n+1,n - k+1(q ) .
we have _ n+1,n - k+1(q ) - _ n ,
n - k(q ) = ( q - n ) _
n , n - k+1(q ) and i^k_n(q ) = _ j=0^kn - j k - ji^j_n-1(q ) ( ( d_n)^k - j11-u_n ) . hence [ ink ] i^k_n(q ) - i^k_n-1(q ) 11-u_n = _ j=0^k-1n - jk - ji^j_n-1(q ) ( ( d_n)^k - j11-u_n ) .
the left hand side of ( [ ink ] ) is [ eq_stirlingidentities ] _ n , k(q)=_m=1^n - km+k m+1 _ m+1(q , n)_n , k+m(q ) , k=1 , , n . comparing coefficients of @xmath92 of the two sides yields identities involving the stirling numbers @xmath93 and @xmath94 of the first and second kinds , indexed with @xmath95 ( and equalling 0 outside this range ) . for @xmath96 , this says s_1(n , n - k)=s_1(n , n - k)+_m=1^km+n - k m+1 s_1(n , n - k+m ) , and hence [ star ] s_1(n , n - k)=_m=1^k m+n - k m+1(-1)^ms_1(n , n - k+m ) , k=1 , , n. this contains the known formulas s_1(n , n-1)=-n 2 s_1(n , n-2)=n(n-1)(n-2)(3n-1 ) . | we obtain formulas for the growth rate of the numbers of certain paths in a multi - dimensional analogue of the eulerian graph .
corollaries are new identities relating stirling numbers of the first and second kinds . |
what is a relation between the lmmp and some concrete geometrical results such as polyhedral properties of effective and mobile cones for ft varieties , finiteness results for modifications of ft varieties and minimal models , factorization of birational isomorphisms of mori fibrations into sarkisov links and birational rigidity ?
one of the possible explanations for those relations can be found by exploring the geography of log models .
this paper presents this explanation that was originally introduced in @xcite @xcite and partially developed in @xcite .
we always assume the lmmp and the semiampleness conjecture .
one can find a detailed information about those conjectures and which of their variations are needed in section [ section7-lmmp - semi ] .
take a pair @xmath0 where @xmath1 is a proper morphism and @xmath2 is a reduced b - divisor of @xmath1 with the distinct prime components @xmath3 . in this section
we introduce @xmath4 equivalence relations on the _ unit cube _ : @xmath5 s_i \cong [ 0,1]^m.\ ] ] an element @xmath6 will be referred to as a _ boundary _
divisor even when it is actually a b - boundary . for the concepts of the lmmp used in these relations , see section [ section7-lmmp - semi ] .
[ [ model - equivalence . ] ] model equivalence .
+ + + + + + + + + + + + + + + + + + boundaries @xmath7 are _ model equivalent _ and we write @xmath8 if the pairs @xmath9 have the same set of resulting models @xcite .
[ [ weakly - log - canonical - equivalence . ] ] weakly log canonical equivalence .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + boundaries @xmath7 are _ wlc model equivalent _ and we write @xmath10 if both pairs @xmath9 have the same wlc models and the models are numerically equivalent , that is : for any model @xmath11 of @xmath1 , @xmath12 is a wlc model of @xmath13 if and only if @xmath14 is a wlc model of @xmath15 ; and the divisors @xmath16 have the same signatures with respect to the intersection for all curves on @xmath11 . [
[ log - canonical - model - equivalence . ] ] log canonical model equivalence .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + boundaries @xmath7 are _ lc model equivalent _ and we write @xmath17 if the pairs
@xmath9 have the same ( rational ) iitaka fibrations : for a given ( rational ) contraction @xmath18 , @xmath19 is an iitaka fibration for @xmath20 if and only if @xmath19 is an iitaka fibration for @xmath21 , or equivalently , @xmath13 has a ( rational ) iitaka fibration @xmath22 if and only if @xmath15 has a ( rational ) iitaka fibration @xmath23 and the fibrations are naturally isomorphic , that is , there is a commutative diagram @xmath24\ar@{-->}[rd ] & \\
y\ar[rr]^{\cong } & & y ' } \ ] ] with the natural isomorphism @xmath25 .
[ [ mobile - equivalence . ] ] mobile equivalence .
+ + + + + + + + + + + + + + + + + + + boundaries @xmath7 are _ mobile equivalent _ and we write @xmath26 if the positive parts @xmath27 of both b - divisors @xmath28 have the same signatures with respect to the curves on rather high models @xmath11 . on a rather high model
we can take any model @xmath11 over some wlc models of @xmath9 .
[ [ fixed - equivalence . ] ] fixed equivalence .
+ + + + + + + + + + + + + + + + + + boundaries @xmath7 are _ fix equivalent _ and we write @xmath29 if the fixed parts @xmath30 of both
b - divisors @xmath28 have the same signatures of their multiplicities , that is : for any prime b - divisor @xmath31 , @xmath32 note that all the relations above are defined on the whole @xmath33 .
the resulting models obtained by the lmmp are at least projective , and even slt by the slt lmmp .
an abstract definition of resulting models gives a larger class of them .
however , according to the following it is enough to consider only the good ones : projective @xmath34-factorial wlc models , e.g. , slt wlc models .
[ prop - qfact - proj ] the equivalence @xmath35 for projective @xmath34-factorial wlc models is the same as that for all wlc models
. any non-@xmath34-factorial wlc model can be obtained by a crepant contraction of some @xmath34-factorial wlc model that gives a wlc model by the semiampleness .
any @xmath34-factorial nonprojective wlc model can be obtained from a projective and @xmath34-factorial wlc model by a ( generalized ) log flop .
both constructions use the slt lmmp and @xcite .
thus if the pairs have the same projective and @xmath34-factorial , numerically equivalent models , they have the same numerically equivalent wlc models .
the converse statement means that each projective and @xmath34-factorial wlc model of @xmath13 will be also projective and @xmath34-factorial wlc model of @xmath15 when @xmath10 ( cf .
the proof of lemma [ lemma - convex relations ] ) .
[ convention ] the model equivalences @xmath36 are defined for projective and @xmath34-factorial resulting models of pairs .
in fact , we can even use slt wlc models since proposition [ prop - qfact - proj ] holds for slt wlc models too . in
most of constructions below , we use slt wlc resulting models
. however , slt wlc models are not stable for limits of boundaries ( see lemma [ lemma - closed n ] ) .
the following relations hold for model equivalences .
[ prop - equivalence ] we have the following implications : @xmath37 @xmath38 immediate by definition and the semiampleness [ conj - semiample ] . in general , @xmath39 for example , @xmath40 follows from the fact that the small modifications of flopping facets preserve the @xmath41 equivalence relation but not @xmath42 ( see theorem [ thrm - facet types ] , [ thrm - ridge types ] ) .
define the subset of @xmath33 : @xmath43 equivalently , we can also use the condition : @xmath44 ( see numerical kodaira dimension in section [ section7-lmmp - semi ] ) .
moreover , by the semiampleness [ conj - semiample ] , this is equivalent to the nonnegativity of ( invariant ) kodaira dimension ( see kodaira dimension in section [ section7-lmmp - semi ] ) : @xmath45 .
proposition [ prop - equivalence ] implies that @xmath35 is the finest of the model equivalences on @xmath46 .
thus , for the pairs with wlc models , the finiteness and polyhedral properties of all the equivalence relations follow from those of @xmath35 .
all the above equivalence relations , except for @xmath42 , are not interesting outside @xmath46 because they are trivial on @xmath47 .
it is expected that many similar properties hold for @xmath42 on @xmath33 in the case of mori log fibrations @xcite and that plays an important role for the genuine sarkisov program .
recall that @xmath48 is the @xmath49-vector space of b - divisors spanned by @xmath50 : @xmath51 the space @xmath48 contains the cube @xmath33 .
a _ convex polyhedron _ @xmath52 in @xmath48 is a set defined by the intersection of finitely many ( open or closed ) half spaces and hyperplanes in @xmath48 .
a _ polyhedron _ is a finite union of convex ones .
a convex polyhedron @xmath52 is said to be _ open _ in @xmath48 if the half spaces in the intersection are open .
the closure @xmath53 , the interior @xmath54 , and the boundary @xmath55 of a convex polyhedron @xmath52 will be taken in its linear span , the minimal intersection of hyperplanes containing @xmath52 .
@xmath56 of a convex polyhedron @xmath52 is the intersection @xmath57 ( possibly @xmath58 or @xmath53 itself ) for some hyperplane @xmath59 such that @xmath60 .
a _ facet _ of @xmath52 is a maximal face , that is , a face of dimension @xmath61 .
a _ ridge _ is a facet of a facet . for two distinct vectors @xmath62
, @xmath63 denotes the open ray starting from @xmath64 through @xmath65 and @xmath66 denotes the line through @xmath64 and @xmath65 .
@xmath67 $ ] ( @xmath68 , etc ) is defined as the closed ( resp .
half closed , open , etc ) interval between @xmath64 and @xmath65 .
let @xmath52 be a convex polyhedron in @xmath33 .
a facet ( face ) @xmath56 of @xmath52 is _ given by a facet @xmath69 of _
@xmath33 if @xmath70 . the facet ( resp .
face ) @xmath69 is given by equation @xmath71 or @xmath72 .
thus @xmath56 can be given in @xmath53 by one of those equations .
note the following property of the linear function @xmath73 on @xmath48 : if @xmath74 ( respectively @xmath75 ) in @xmath76 and @xmath77 ( respectively @xmath78 ) in some point of @xmath52 , then @xmath74 ( respectively @xmath75 ) on @xmath56 and @xmath77 ( respectively @xmath78 ) in @xmath79 . by a _ geography _ on @xmath46 ,
we mean a finite rational polyhedral decomposition of it , and the subsets @xmath80 of such a decomposition will be referred to as _ classes _ or just _
polyhedrons_. we call a class @xmath81 of maximal dimension , @xmath82 , a _ country _ , a class @xmath56 of codimension 1 a _ facet _ , and a class @xmath83 of codimension 2 a _
ridge_. according to the following , each of the above equivalence relations defines a geography on @xmath46 , but the notations @xmath46 will be reserved only for the _ wlc _ geography given by @xmath35 .
thus we consider @xmath46 as a set with the structure .
let @xmath56 be a minimal face of @xmath33 which contains a polyhedron @xmath52 .
we say that a polyhedron @xmath52 is open in @xmath33 if it is open in @xmath56 , that is , @xmath84 , where @xmath85 is an open polyhedron in @xmath48 .
[ mainthrm - geography]the set @xmath46 is closed convex rational polyhedral and it is decomposed into a finite number of @xmath35 classes @xmath52 .
each class @xmath52 of the wlc geography is a convex rational polyhedron which is open in @xmath33 . for two classes
@xmath86 , @xmath87 holds if and only if @xmath88 is a face of @xmath53 .
let @xmath89 be a projective surface with a contraction @xmath90 of a nonsingular rational curve @xmath91 .
consider the pairs @xmath13 where @xmath92c$ ] .
then by the adjunction formula , we have the equality @xmath93 . if @xmath94 , the pair @xmath95 is a wlc model of itself .
if @xmath96 , @xmath97 is a wlc model of @xmath95 . in total , we have the following geography : example1.eps ( -1,-10)@xmath75 ( 42,-10)@xmath98 ( 96,-10)@xmath99 in this example , we have the following : 1 . @xmath100 ; 2 . there are three @xmath35 classes in @xmath46 : @xmath101 @xmath102 @xmath103c$ ] ; 3 .
there are two @xmath104 classes : @xmath105c$ ] ; 4 .
@xmath106 , but @xmath107 : @xmath108 .
[ example - cremona ] suppose that @xmath109 and @xmath110 are related by a standard quadratic transformation .
choose a general nonsingular prime b - divisor @xmath111 where @xmath112 is a straight line on @xmath113 for @xmath114 .
consider the geography in the unit cube @xmath33 where @xmath115 .
example2.eps ( 10,75)@xmath116 ( 50,50)@xmath117 ( 75,10)@xmath118 ( -23,71)@xmath119 ( -12,95)@xmath120 ( 65,-12)@xmath121 ( 95,-10)@xmath122 in particular , @xmath123 are resulting models for @xmath116 and @xmath118 .
two @xmath124 s are isomorphic under an unnatural isomorphism ( the quadratic transform ) : @xmath116 and @xmath118 are different classes for @xmath42 .
however , @xmath125 and @xmath104 is not convex .
[ lemma - convex relations]@xmath35 is convex in @xmath46 .
the proof of ( * ? ? ? * lemma 2 ) . )
if @xmath10 and @xmath12 is a projective @xmath34-factorial wlc model of @xmath13 , then @xmath126 is a projective @xmath34-factorial wlc model of @xmath15 .
moreover , for any @xmath127 $ ] , @xmath128 is a projective @xmath34-factorial wlc model of @xmath129 .
the models are numerically equivalent to the edge models @xmath130 @xmath126 . thus all
@xmath128 have the same ( generalized ) log flops and wlc models .
note that @xmath131 are linear with respect to @xmath31 and by proposition [ prop - equivalence ] , @xmath132 has the same signatures for all multiplicities with respect to the prime b - divisors .
thus the log flops of @xmath128 corresponding to wlc models are the same for all @xmath127 $ ] ( cf .
corollary [ cor - property - p - e ] ( 1 ) ) .
we say that a _ set _ @xmath52 in a finite dimensional @xmath49-vector space @xmath133 is _ locally conical _ if for any point @xmath134 in @xmath133 , there exists an open neighborhood @xmath135 containing @xmath134 such that the intersection @xmath136 is conical .
in other words , for any @xmath137 , either @xmath138 or @xmath139 . a _ decomposition _ of a set into disjoint subsets @xmath52
is said to be _ locally conical _ if each @xmath52 is locally conical with the same @xmath135 .
as we will see shortly , any wlc geography is a locally conical decomposition of @xmath46 .
[ lemma - locally conical ] let @xmath140 be a closed bounded ( convex ) set in a finite dimensional @xmath49-vector space .
then @xmath140 is locally conical if and only if @xmath140 is polyhedral .
we use induction on @xmath141 .
the case @xmath142 is trivial .
let @xmath143 and assume that the lemma holds in @xmath144 .
the property of being locally conical is preserved under taking hyperplane sections .
therefore by the inductive assumption , all the sections of @xmath140 are polyhedral .
this implies that the set @xmath140 is polyhedral by the following standard criterion @xcite : @xmath140 is polyhedral if and only if all of its @xmath145-dimensional plane sections are polyhedral where @xmath146 .
the converse is clear .
[ lemma - closed n]let @xmath147 be a sequence of boundaries converging to a limit @xmath148 , and @xmath11 be a model of @xmath1 such that @xmath149 is a sequence of wlc models of @xmath150 .
then @xmath12 is a wlc model of @xmath13 .
in particular , @xmath151 .
however , @xmath12 may not be a slt wlc model of @xmath13 even if all models @xmath149 are slt wlc models of @xmath150 .
we call a model @xmath12 in the lemma _ limiting _ from a class @xmath52 if all @xmath152 .
immediate by definition .
first , we establish the finiteness , convexity , and polyhedral property for the closures @xmath53 of the classes in the geography .
note that the closures do not give a decomposition but , by lemma [ lemma - closed n ] , @xmath46 is a finite union of these closures . in particular , @xmath46 is closed .
the convexity of wlc classes follows from lemma [ lemma - convex relations ] . the polyhedral property of @xmath53 follows from the locally conical property of @xmath53 by lemma [ lemma - locally conical ] and its compactness .
the conical property for @xmath53 follows from that of @xmath52 : the closure of a locally conical set is locally conical . but
@xmath52 is locally conical by the stability of wlc models .
this means the following .
take @xmath151 , and let @xmath12 be a _
stable _ wlc model of @xmath13 .
the stability here means that , for all @xmath153 in a neighborhood @xmath135 of @xmath65 in @xmath33 , the pairs @xmath154 are slt initial models of @xmath15 .
such a model @xmath12 can be constructed from a slt wlc model of @xmath13 by additional blowups of lc centers which lie in divisors @xmath155 having @xmath75 multiplicities in @xmath65 .
then , for any boundary @xmath156 , there exists a ( generalized ) log flop @xmath157 such that , for all @xmath158 , the models @xmath159 are slt wlc models of @xmath15 , and the boundaries @xmath153 are wlc equivalent , where @xmath160 denotes the open ray from @xmath65 through @xmath31 .
the existence of @xmath161 follows from ( * ? ? ?
* corolary 9 and addendum 5 ) .
the equivalence of @xmath153 follows from the construction of the flop and from ( * ? ? ? * corollary 11 and addendum 6 ) ( cf .
the proof of lemma [ lemma - convex relations ] above ) .
note that @xmath162 may not be slt , but by the slt lmmp the dlt condition can be replaced by the lc condition everywhere in ( * ? ? ?
* remarks after proposition 1 ) by ( * ? ? ?
* conjecture and heuristic arguments ) .
the finiteness also follows by induction on dimension @xmath46 and the stability
. see ( * ? ? ?
* proposition 3.2.5 ) for details .
these established facts imply all stated properties of @xmath46 , except for the convexity and rationality of @xmath46 .
note that convexity follows from the convexity of positive parts : for all @xmath163 , @xmath164 the slt lmmp assumption is sufficient here .
the proof of the rationality ( of @xmath165 ) and the open property ( of @xmath52 ) uses properties of functions @xmath166 introduced below .
actually , for each @xmath52 there exist finitely many curves @xmath167 on a rather high model and finitely many prime b - divisors @xmath168 such that @xmath52 in @xmath53 can be given by the system of inequalities : @xmath169 in particular , those inequalities determine any country @xmath81 in @xmath46 while @xmath46 itself is given in @xmath33 ( more accurately , in the minimal face containing @xmath46 ) by nonstrict inequalities @xmath170 for some curves @xmath171 and some b - divisors @xmath172 , where the linear functions @xmath173 are extensions of linear functions given by @xmath174 , respectively on the linear span of @xmath46 as it will be explained in the continuation of the proof by the end of the section , and equations ( of the face ) @xmath71 or @xmath72 .
the closure @xmath53 of any other lower dimensional class @xmath52 lies in a face of dimension @xmath175 of some country @xmath81 of ( cf .
corollary [ cor - facet = face ] below ) and thus @xmath52 can be given in @xmath33 by strict inequalities and equations : @xmath176 or @xmath74 .
( to be continued ) on a rather high model @xmath177 of @xmath1 and for a curve @xmath178 on @xmath177 , the function @xmath179 is defined as @xmath180 where @xmath12 is a wlc model of @xmath13 , and @xmath181 is a birational morphism .
a _ rather high _ model satisfies the following universal property : @xmath133 dominates some wlc model for any @xmath151 , and @xmath182 is @xmath49-cartier over @xmath133 ( cf . definition of @xmath183 ) .
such a universal model @xmath177 exists by the finiteness in the first half of the proof of the theorem .
by @xcite , the function is independent of the choice of such a model @xmath177 . for a prime b - divisor @xmath31 of @xmath1 ,
the _ initial _ log discrepancy @xmath184 is defined as @xmath185 for a prime b - divisor @xmath31 of @xmath1 , the _ immobility _ function @xmath186 is defined as @xmath187 where @xmath12 is a wlc model of @xmath13 .
it is called immobility because @xmath188 is the multiplicity in b - divisors @xmath31 of the fixed part @xmath189 of @xmath190 . by the invariant property of log discrepancies of wlc models ,
the function @xmath188 is independent of the choice of the wlc model @xmath12 ( * ? ? ?
* @xmath191 ) .
[ prop - p - e]the functions @xmath166 on @xmath46 are nonnegative continuous rational piecewise linear , and linear on each @xmath192 , where @xmath52 is a class of @xmath46 .
moreover , @xmath188 is convex from above , and @xmath193 is convex from below for rather general ( mobile ) curves , that is , for a curve @xmath194 through a general point of a rather high model @xmath177 of @xmath1 .
the function @xmath193 is nonnegative by the nef property of b - divisor @xmath182 .
the function @xmath188 is nonnegative by definition or because the fixed part @xmath189 is effective . the rationality and the linearity on each @xmath53 follow from the corresponding properties of the intersection pairing and the discrepancy function by lemma [ lemma - closed n ] .
thus the finiteness established in the first half of the proof of theorem [ mainthrm - geography ] implies the continuity and the rational piecewise linearity of the functions .
this includes also the rational property of any maximal linear subset ( cf .
facets in corollary [ cor - eqn for facets ] below ) .
the convex properties of @xmath193 from above and of @xmath188 from below follow from the inequalities : @xmath195 and @xmath196 , where @xmath197 and @xmath198 .
[ cor - property - p - e]for any curve @xmath178 on a rather high model @xmath177 of @xmath1 and for any prime b - divisor @xmath31 of @xmath1 , the functions @xmath166 satisfy the following . 1 .
each function @xmath166 has the constant signature on each class @xmath52 : either @xmath78 or @xmath199 on @xmath52 .
the converse holds also , that is , if the values @xmath166 and @xmath200 have the same signatures , then @xmath10 .
2 . @xmath201 on a class @xmath52 if and only if @xmath178 is contracted by the lc contraction @xmath202 ( cf .
log canonical model in section [ section7-lmmp - semi ] . ) of the pair @xmath13 with @xmath203 .
in particular , the contraction depends only on @xmath52 .
equivalently , @xmath204 on @xmath52 if and only if @xmath205 is a curve on @xmath206 .
@xmath207 on a class @xmath52 if and only if @xmath31 can be blown up on a wlc model of this class , i.e. , @xmath31 is nonexceptional on some wlc model @xmath11 .
equivalently , @xmath208 on a class @xmath52 if and only if @xmath31 is exceptional on any wlc model of @xmath13 for @xmath203 .
\(1 ) by definition @xmath183 is equivalent to the same signatures for all @xmath193 , and so is @xmath41 for @xmath188 .
thus by proposition [ prop - equivalence ] @xmath35 is equivalent to the same signatures for all @xmath166 .
\(2 ) immediate by construction of @xmath209 as a composition of @xmath210 followed by the lc contraction of a wlc model @xmath12 , where @xmath12 is an appropriate wlc model of @xmath13 .
the composition contracts exactly the curves @xmath178 on @xmath177 with @xmath211 ( cf .
* ( ii ) ) ) .
\(3 ) by definition @xmath207 for any wlc model @xmath12 , on which @xmath31 is nonexceptional . on the other hand , if @xmath207 and @xmath31 is exceptional on a wlc model @xmath12 , then @xmath212 and one can blow up @xmath31 to a crepant slt wlc model @xmath162 of @xmath13 with a ( higher ) model @xmath213 of @xmath214 by the slt lmmp . by ( 1 )
this depends only on @xmath52 .
[ cor - facet = face]let @xmath52 be a class of @xmath46 and @xmath56 be its face .
then @xmath215 for some class @xmath85 of @xmath46 .
in particular , @xmath216 .
thus the geography @xmath46 is determined up to finitely many possibilities by its countries @xmath81 and even their closures @xmath217 .
note that @xmath53 is a convex polyhedron by the first half of the proof of theorem [ mainthrm - geography ] .
immediate by proposition [ prop - p - e ] and corollary [ cor - property - p - e ] ( 1 ) .
an elementary property of linear functions is useful here : a nonnegative linear function on a convex polyhedron @xmath218 has the same signature on @xmath219 .
[ cor - eqn for facets]let @xmath52 be a class of @xmath46 , @xmath56 be its facet , and @xmath177 be a rather high model of @xmath1 .
then either @xmath56 is given by a facet of @xmath33 ; there exists a curve @xmath178 on a rather high model @xmath177 of @xmath1 such that @xmath220 and @xmath221 ; or there exists a prime b - divisor @xmath31 of @xmath1 such that @xmath222 and @xmath223 .
moreover , if @xmath224 , then @xmath225 .
however , if @xmath226 , then @xmath227 , and this is possible only for facets given by a facet of @xmath33 .
the same holds for the faces of @xmath52 .
take @xmath76 , and an outer direction @xmath228 from @xmath52 , that is , @xmath229 and @xmath230 .
let @xmath12 be a wlc model of @xmath13 which is a limit of ( slt ) wlc model @xmath126 of @xmath15 for @xmath231 ( e.g. , along @xmath232 ; see lemma [ lemma - closed n ] ) . by the polyhedral property of lc @xcite , the following three cases are only possible : \(i )
@xmath153 is not a b - boundary for any @xmath233 ;
\(ii ) @xmath154 is not lc for any @xmath233 ; \(iii ) for any @xmath233 in a neighborhood of @xmath65 , @xmath153 is a b - boundary and @xmath154 is lc . in the case ( i ) , there exists prime @xmath3 , such that @xmath234 or @xmath235 .
the facet @xmath69 of @xmath33 with the equation @xmath71 or @xmath74 respectively gives @xmath56 ( see the property of function @xmath73 on page 5 above ) . in the case ( ii )
, there exists a prime b - divisor @xmath31 of @xmath1 such that @xmath236 for any @xmath233 .
if we do not assume ( i ) , @xmath31 is exceptional on @xmath237 and , by the linear property of discrepancy , @xmath238 and @xmath239 for any @xmath240 .
thus @xmath207 and , if for every @xmath241 , @xmath242 , then for all @xmath153 near @xmath65 on @xmath232 , @xmath243 and @xmath244 for @xmath233 , a contradiction again by the negation of ( i ) .
hence @xmath245 for some @xmath241 .
this gives required @xmath31 by corollary [ cor - facet = face ] .
the case ( iii ) itself has two subcases : for all @xmath233 in a neighborhood of @xmath65 , ( iii-1 ) @xmath126 is not a wlc model ; or ( iii-2 ) @xmath126 is a wlc model but not of @xmath15 . indeed
, otherwise , for all those @xmath153 , the pairs @xmath126 are wlc models of @xmath15 . by definition
all functions @xmath200 are linear on @xmath232 in the neighborhood of @xmath65 .
thus @xmath200 have the same signatures and all those @xmath153 are @xmath35 , a contradiction with our assumptions . in the case ( iii-1 ) , by ( * ? ? ?
* corollary 9 ) there exists a curve @xmath246 on @xmath11 such that @xmath247 for @xmath233 and @xmath248 .
thus @xmath249 for @xmath240 , and for a curve @xmath178 on a rather high model @xmath250 of @xmath1 with @xmath251 and @xmath252 , again by corollary [ cor - facet = face ] , @xmath220 and @xmath221 .
the same holds for a birational transform of @xmath178 on @xmath177 @xcite . in the case
( iii-2 ) , there exists a b - divisor @xmath31 with @xmath253 . by the linear property of discrepancy and of initial discrepancy , @xmath245 for @xmath240 and @xmath207 .
thus @xmath31 is a required divisor .
the results about @xmath254 follow from corollaries [ cor - property - p - e ] ( 1 ) and [ cor - facet = face ] ; in the case of faces by induction .
let @xmath255 be the minimal face of @xmath33 containing @xmath46 .
then the linear span of @xmath46 is the linear span of @xmath56 by the monotonicity @xmath256 for @xmath257 in @xmath33 ( cf .
proposition [ prop - monotone n ] ) .
moreover , the linear span of @xmath46 can be given in @xmath48 by equations @xmath74 .
in particular , the span is rational .
the rationality of @xmath46 , of the polyhedrons @xmath53 , and the open property of the classes @xmath52 follow from corollaries [ cor - facet = face ] and [ cor - eqn for facets ] .
this gives also required equations and inequalities .
the linear function @xmath258 on the linear span of @xmath52 is determined by the linear function @xmath193 on @xmath52 , in particular , those functions with @xmath221 for a facet @xmath56 of @xmath52 will suffice .
similarly , one can introduce functions @xmath259 .
finally , the interior @xmath219 of a face of @xmath52 given by a facet of @xmath33 , belongs to @xmath52 or is disjoint from it .
if @xmath86 are two classes and @xmath260 is a face of @xmath53 , then @xmath87 .
conversely , if @xmath87 but @xmath260 is not a face of @xmath53 , then by corollary [ cor - facet = face ] , @xmath261 .
moreover , by the convexity of classes and of @xmath46 , we can suppose that @xmath88 contains a face @xmath262 of @xmath53 ( maybe for different @xmath52 ) which is a polyhedron of the same dimension as @xmath85 and with @xmath263 .
hence there exists a facet @xmath264 of @xmath52 which contains an internal point of @xmath85 .
this gives a contradiction with corollary [ cor - eqn for facets ] .
for example , if @xmath56 is given by a facet of @xmath33 with an equation @xmath71 or @xmath72 , then @xmath71 or @xmath72 respectively on @xmath88 , in particular , on @xmath262 , and thus on @xmath53 , a contradiction .
similarly , for other functions .
for first applications of geography , we recall and define natural cones in @xmath48 and @xmath265 .
let @xmath266 be a model of @xmath1 , and @xmath267 be a reduced b - divisor of @xmath1 .
our _ numerical space _ @xmath265 denotes the space of @xmath49-divisors , not necessarily @xmath49-cartier , mudulo the numerical equivalence @xmath268 . nonetheless , the relative weil - picard ( class ) number @xmath269 for proper @xmath1 .
[ [ semiample - cones . ] ] semiample cones .
+ + + + + + + + + + + + + + + + in the numerical space @xmath265 and in the space of b - divisors @xmath48 , respectively , the cones @xmath270 \in \operatorname{n^1}(x / z)\left|\begin{array}{l } d\equiv d ' \text { for some semiample}\\ \text{divisor $ d'$ on } x / z \end{array}\right.\right\}_,\\ \operatorname{s\mathfrak{a}}_s(w / z)&:=\{d\in{\mathfrak d}_s\mid d_w \text { is semiample on } w / z\ } \end{array}\ ] ] are defined .
[ [ nef - cones . ] ] nef cones .
+ + + + + + + + + + in @xmath271 , respectively , the cones @xmath272 \in \operatorname{n^1}(x / z)\mid d \text { is nef on } x / z\},\\ \operatorname{\mathfrak{nef}}_s(w / z)&:=\{d\in{\mathfrak d}_s\mid d_w \text { is nef on } w / z\ } \end{array}\ ] ] are defined and closed . in the projective case
, they are the closures of the ample ( khler ) cones .
the cone @xmath273 is dual to the kleiman - mori cone , the closure of the cone of curves .
thus its polyhedral properties are related to the cone of curves ( the mori theory ) .
however , we consider only cones of divisors ( see corollary [ cor - positive cones ] below ) . [
[ mobile - cones . ] ] mobile cones .
+ + + + + + + + + + + + + in @xmath271 , respectively , the cones @xmath274\in \operatorname{n^1}(x / z)\left|\begin{array}{l } d\equiv d ' \text { for some $ { \mathbb{r}}$-mobile } \\ \text{divisor $ d'$ on } x / z \end{array}\right.\right\}_,\\ \operatorname{\mathfrak{m}}_s(w / z)&:=\{d\in{\mathfrak d}_s\mid d_w \text { is $ { \mathbb{r}}$-mobile on } w / z\ } \end{array}\ ] ] are defined .
[ [ effective - cones . ] ] effective cones .
+ + + + + + + + + + + + + + + + in @xmath271 , respectively , the cones @xmath275 are defined . in general
, the numerical equivalence @xmath268 of divisors does not preserve the semiampleness and mobile properties of divisors while clearly @xmath276 does . thus introducing cones in the numerical space
, we had a choice to assume such a property on the whole class @xmath277 $ ] or on some representative @xmath278 $ ] .
we chose the second one to have a larger cone . according to the choice
, the linear map @xmath279 $ ] agrees the @xmath280 equivalence property with the numerical one ( cf .
the proof of corollary [ cor-0log num cones ] ) .
it is well - known that the above cones , numerical and linear , are convex and satisfy the inclusions : @xmath281 we recall also that : * _ @xmath75-log pair _
@xmath13 is a pair with a proper morphism @xmath1 and a boundary @xmath65 on @xmath89 such that @xmath282 is klt and @xmath283 , or equivalently @xmath284 according to @xcite ; and * ( relative ) _ ft ( fano type ) _
variety @xmath1 is a variety such that there exists a boundary @xmath65 on @xmath89 such that the pair @xmath13 is a klt log fano variety , in particular , @xmath1 is projective . for other equivalent characterizations of relative ft varieties ,
see ( * ? ? ? * lemma - definition 2.8 ) .
the most useful one among them to us is the following .
[ lemma - ft=0log ] let @xmath1 be be an ft variety .
then for any fixed reduced divisor @xmath50 of @xmath89 , there exists an @xmath49-boundary @xmath65 on @xmath89 such that @xmath13 is a @xmath75-log pair and @xmath285 .
moreover , @xmath65 is big .
conversely , if @xmath13 is a @xmath75-log pair with projective @xmath1 and big @xmath65 , e.g. , prime components of @xmath286 generate @xmath265 , then @xmath1 is ft
. see ( * ? ? ?
* lemma - definition 2.8 ) .
the big property for @xmath65 follows from the big property of @xmath287 for ft varieties . to obtain polyhedral properties of cones in @xmath48 from geography of log models ,
we translate polyhedrons in the space .
let @xmath288 be the translation by @xmath289 .
it translates @xmath65 into the origin @xmath290 , the vertex of cones .
geography allows to construct only some polyhedrons in @xmath291 .
however each polyhedron near @xmath65 is conical .
thus to compare the translation of a polyhedron with a cone in a neighborhood @xmath135 of @xmath290 would be sufficient . in what follows
, we say that two sets @xmath292 ( cones or polyhedrons ) coincide in @xmath135 if @xmath293 [ prop-0log cones coincide in u]let @xmath13 be a @xmath75-log pair and put @xmath294 .
then there exists a neighborhood @xmath135 of @xmath75 in @xmath48 in which : @xmath295 where @xmath296 is the closure of a maximal dimensional class @xmath297 of wlc models with @xmath298 , @xmath299 are the classes of wlc models with @xmath298 , and @xmath300 are the closures of classes @xmath301 of wlc models with @xmath237 isomorphic to @xmath89 in codimension @xmath72 .
each divisor @xmath302 has a unique mob+exc decomposition : @xmath303 and @xmath304 , where @xmath305 is @xmath49-mobile and @xmath306 is effective very exceptional with respect to a @xmath72-contraction given by @xmath305 ; the @xmath72-contraction , @xmath306 are unique and @xmath305 is unique up to @xmath280 and @xmath268 , respectively . in particular , in @xmath48 the @xmath49-mobile and @xmath280 effective properties of divisors are equivalent to those for @xmath268 .
moreover , for @xmath307 , @xmath308 and @xmath309 , respectively , where @xmath310 .
first we clarify the statement .
let @xmath311 be a boundary in a class @xmath52 .
if the pair @xmath15 has a wlc model @xmath312 for which @xmath22 is a small birational isomorphism , that is , an isomorphism in codimension @xmath72 , we denote the corresponding class by @xmath313 . in particular , if @xmath298 , we have isomorphism in codimension @xmath314 , or in dimension @xmath315 , and we denote it by @xmath299 . if @xmath89 is nonprojective or / and non @xmath34-factorial , we allow such models .
but if we have to follow convention [ convention ] , we need to replace the models by their small projective @xmath34-factorializations ( * ? ? ?
* corollary 6.7 ) ( cf .
proposition [ prop - qfact - proj ] ) . for @xmath316 , in addition , we suppose that @xmath317 is a small log flop .
thus we can replace @xmath1 by a small projective @xmath34-factorialization .
now we choose a conical neighborhood @xmath135 of @xmath290 such that , for each @xmath318 , the pair @xmath15 with @xmath310 is a klt initial model of itself , including the boundary condition ; @xmath319 is conical for each class @xmath320 ( this is sufficient for the classes @xmath52 with @xmath321 ) .
such a neighborhood exists by our assumptions and by theorem [ mainthrm - geography ] . for any divisor @xmath318 , @xmath322 thus @xmath323 @xmath324 @xmath325 @xmath326 this gives the required equalities and mob+exc decompositions .
indeed , since @xmath327 is lc , the @xmath49-mobile part @xmath305 of @xmath31 is @xmath328 and respectively , the very exceptional part @xmath306 is @xmath329 .
the @xmath49-linear decomposition gives the numerical one because the former is essentially numerical by the lmmp ( see ( * ? ? ?
* definition 3.3 and proposition 3.4 ) ) .
the semiampleness gives the equality @xmath330 .
the closed property and the equalities with closures follow from lemma [ lemma - closed n ] or corollary [ cor - property - p - e ] ( 1 ) . finally , by corollary [ cor - property - p - e ]
( 1 ) , by convexity of @xmath331 and by the open property of classes , @xmath332 . holds for one class @xmath297 ( cf .
* lemma 2 ) ) . to construct all the required models , the klt slt lmmp is sufficient , and for contractions of those models , the semiampleness for klt @xmath34-boundaries is sufficient ( see corollary [ cor - qsamp = rsamp ] ) .
[ cor-0log cones]let @xmath13 be a @xmath75-log pair and @xmath285 be a reduced divisor on @xmath89 .
then the cones @xmath333 in @xmath48 are closed convex rational polyhedral .
each divisor @xmath302 has a unique mob+exc decomposition : @xmath303 and @xmath304 , where @xmath305 is @xmath49-mobile and @xmath306 is effective very exceptional with respect to a @xmath72-contraction given by @xmath305 .
moreover , @xmath268 instead of @xmath280 gives the same cones in @xmath48 .
immediate by proposition [ prop-0log cones coincide in u ] and theorem [ mainthrm - geography ] for @xmath294 . if @xmath334 , we obtain the required polyhedral properties by intersection with the subspace @xmath48 . [ cor-0log num cones]let
@xmath13 be a @xmath75-log pair such that prime components of @xmath294 generate @xmath265 .
then the cones @xmath335 in @xmath265 are closed convex rational polyhedral .
thus @xmath336 is the pseudo - effective cone too . the relations @xmath268 and @xmath280 coincide on @xmath1 .
each divisor @xmath337 has a unique mob+exc decomposition : @xmath338 , where @xmath305 is @xmath49-mobile and @xmath306 is effective very exceptional with respect to a @xmath72-contraction given by @xmath305 . the corollary gives a new result even for @xmath273 when @xmath1 is nonprojective .
the required properties of the linear cones in @xmath48 imply the same properties of the numerical cones in @xmath265 .
indeed , the natural linear map @xmath339\colon { \mathfrak d}_s & \rightarrow \operatorname{n^1}(x / z)\\ d & \mapsto [ d],\text { the numerical class of } d , \end{array}\ ] ] maps the cones @xmath340 @xmath341 since prime components @xmath3 of @xmath50 generate @xmath265 , the map @xmath279 $ ] and the induced maps of cones are surjective , that gives the above implication by corollary [ cor-0log cones ] .
the surjectivity for cones uses the following : @xmath280 can be replaced by @xmath268 in the definition of cones in @xmath48 by the same corollary .
moreover , the coincidence of @xmath268 and @xmath276 holds on @xmath1 .
this is well - known for projective @xmath1 ( * ? ? ?
* lemma 4.1.12 ) ( the proof uses the assumption that @xmath342 ; possibly , conjecture [ conj - semiample ] in the positive characteristic allows us to omit the assumption ) .
the general case uses the rationality of singularities of @xmath89 .
all required models in construction will be ft varieties by lemma [ lemma - ft=0log ] .
semiampleness holds for ft varieties in characteristic @xmath75 : for @xmath34-divisors , it holds by the base freeness of @xcite , that is sufficient for @xmath49-divisors by corollary [ cor - qsamp = rsamp ] ( cf .
* corollary 10 ) ) . by lemma [ lemma - ft=0log ] ,
the big klt slt lmmp is sufficient for the ft varieties .
[ cor - positive cones]let @xmath1 be an ft variety , and @xmath50 be a reduced divisor of @xmath89 .
then the cones @xmath343 in @xmath265 and in @xmath48 , respectively are closed convex rational polyhedral . the relations @xmath268 and @xmath280 coincide on @xmath1 .
each class @xmath277\in\operatorname{eff}(x / z)$ ] ( respectively divisor @xmath302 ) has a unique decomposition mob+exc : @xmath277=[m]+[e]$ ] ( respectively @xmath303 ) , where @xmath305 is @xmath49-mobile and @xmath306 is effective very exceptional with respect to a @xmath72-contraction given by @xmath305 .
immediate by lemma [ lemma - ft=0log ] and corollaries [ cor-0log cones ] , [ cor-0log num cones ] .
we use here the finiteness of the weil - picard number : @xmath269 .
we will give a natural decomposition of the cones @xmath344 in @xmath265 and @xmath345 in @xmath48 in the next section .
the support of a b - divisor @xmath346 is defined as the reduced b - divisor @xmath347 .
so , a reduced b - divisor @xmath348 is identified with its support : @xmath349 .
let @xmath13 be a @xmath75-log pair and @xmath294 .
then the geography of @xmath46 in proposition [ prop-0log cones coincide in u ] _ induces _ a finite decomposition of the cone @xmath350 into open convex rational polyhedral subcones @xmath52 .
the open and polyhedral properties follow from that of classes in @xmath46 by theorem [ mainthrm - geography ] .
note that @xmath65 is an internal point of @xmath33 .
this decomposition gives also a decomposition of @xmath351 for any reduced @xmath352 by corollary [ cor-0log cones ] ; its subcones will be called also classes . to give an internal interpretation of this decomposition , we introduce the following relation . for divisors
@xmath353 , @xmath354 if , for their mob+exc decompositions @xmath355 , the rational @xmath72-contractions @xmath356 given by @xmath357 are the same , and @xmath358
. the relation @xmath280 can be replaced by @xmath268 and used for classes : @xmath277{\sim_{\mathrm{m+e}}}[d']$ ] .
the corresponding _ induced _ classes of @xmath336 in @xmath265 will be denoted by @xmath359 and their contractions by @xmath360 .
the rational @xmath72-contraction depends only on the class @xmath52 and @xmath359 , respectively , and will be referred to as the @xmath31-contraction . if @xmath1 is a birational contraction with respect to @xmath31 , then @xmath361 is small and well - known as a @xmath31-flip of @xmath1 @xcite . by construction , it is also a log flop of @xmath13 in the latter case .
( _ warning _ : @xmath31-contraction is not a contraction with respect to a divisor @xmath31 in the @xmath31-mmp . ) in particular , this gives polyhedral decompositions of cones @xmath336 , @xmath362 in @xmath265 and @xmath363 @xmath364 in @xmath48 for @xmath75-log pairs @xmath13 , when prime components of @xmath294 generate @xmath265 , and for ft varieties @xmath1 .
the decomposition of mobile cones @xmath365 @xmath366 is interesting for applications ( see proposition [ prop - ft mobile cones ] ) rather than their general structure : each class of those cones is a class in @xmath336 or @xmath350 respectively and it is determined by its @xmath72-contraction ( @xmath367 ) .
so , we focus on classes of effective cones .
let @xmath368 be a @xmath72-contraction with projective @xmath11 .
a polarization on @xmath11 is an ample @xmath49-divisor @xmath369 .
the polarization is called _ supported _ in @xmath50 if @xmath370 for some effective divisor @xmath306 on @xmath89 , very exceptional on @xmath237 , where @xmath371 is the pull back ( transform ) of @xmath369 @xcite ; in general , @xmath371 is an @xmath49-weil divisor .
[ cor-0log eff decomposition]let @xmath13 be a @xmath75-log pair and @xmath285 be a reduced divisor on @xmath89 .
then the decomposition of the cone @xmath350 induced by @xmath46 can be given by the mob+exc relation @xmath372 .
there are finitely many rational @xmath72-contractions @xmath373 , and @xmath374 is projective with a polarization supported in @xmath50 .
conversely , any such contraction corresponds to a class @xmath52 .
immediate by proposition [ prop-0log cones coincide in u ] , corollary [ cor-0log cones ] and corollary [ cor - property - p - e ] ( 1 ) . note that , for @xmath375 such that @xmath376 , in the proposition , the contraction condition of @xmath372 means that @xmath377 , and the support condition that @xmath378 . but for any prime divisor @xmath306 exceptional on @xmath89 , @xmath379 by the klt property in @xmath135 .
thus the support condition means that @xmath380 , or equivalently , @xmath381 have the same signatures .
the finiteness of rational @xmath72-contractions @xmath373 follows from the finiteness of geography by theorem [ mainthrm - geography ] . by construction ,
each @xmath374 is projective with a polarization supported in @xmath50 .
the converse is by definition .
[ cor-0log num eff decomposition]let @xmath13 be a @xmath75-log pair such that components of @xmath294 generate @xmath265 .
then the decomposition of the cone @xmath336 induced by @xmath46 can be given by the mob+exc relation @xmath372 .
there are finitely many rational @xmath72-contractions @xmath360 , and @xmath382 is projective .
conversely , any such contraction corresponds to a class @xmath359 .
immediate by the above corollary and corollary [ cor-0log num cones ] .
[ cor - ft eff decomposition]let @xmath1 be an ft variety and @xmath50 be a reduced divisor on @xmath89 .
then the decomposition of cones @xmath383 induced by @xmath46 can be given by the mob+exc relation @xmath372 .
there are finitely many rational @xmath72-contractions @xmath384 , and @xmath385 are projective with a polarization supported in @xmath50 for @xmath386 .
conversely , any such contraction corresponds to a class @xmath387 .
immediate by lemma [ lemma - ft=0log ] and corollary [ cor-0log num eff decomposition ] .
[ cor-0log finite contractions]let @xmath13 be a @xmath75-log pair such that components of @xmath294 generate @xmath265 .
then there are only finitely many rational @xmath72-contractions @xmath388 with projective @xmath11 .
immediate by corollary [ cor-0log num eff decomposition ] .
[ cor - ft finite contractions]let @xmath1 be an ft variety .
then there are only finitely many rational @xmath72-contractions @xmath388 with projective @xmath11 .
immediate by lemma [ lemma - ft=0log ] and corollary [ cor-0log finite contractions ] .
[ thrm - finite wlc big klt]let @xmath13 be a klt log pair such that @xmath389 is big .
then there are only finitely many projective wlc models @xmath390 of @xmath13 .
let @xmath12 be a slt wlc model of @xmath13 .
then by the monotonicity ( * ? ?
* lemma 2.4 ) , @xmath12 is also klt and thus @xmath391 .
let @xmath392 be its lc model . by construction , @xmath392 is also klt and projective .
it is known also that any wlc model @xmath393 of @xmath13 is a generalized log flop of @xmath392 .
generalized means that it can blow up divisors @xmath31 of @xmath394 , but only with log discrepancies @xmath395 .
thus any wlc model @xmath393 of @xmath13 is a @xmath72-contraction @xmath396 of a projective _ terminalization _
@xmath397 where @xmath398 is a crepant blow up of all prime @xmath31 with @xmath395 . since the set of those @xmath31 is finite ( * ? ? ?
* corollary 1.7 ) , such a blow up exists .
the terminalization @xmath398 is ft by lemma [ lemma - ft=0log ] ( for birational contraction @xmath399 any divisor is big ) and the number of @xmath72-contractions into projective models is finite by corollary [ cor - ft finite contractions ] .
[ cor - finite big minimal ] let @xmath1 be a relative variety of general type
. then there are only finitely many projective minimal models of @xmath1 .
immediate by theorem [ thrm - finite wlc big klt ] applied to a resolution of singularities of @xmath1 .
@xmath31-contractions can be constructed by the @xmath31-mmp with the initial model @xmath400 @xcite .
[ cor - dmmp on 0log]let @xmath13 be a @xmath75-log pair with projective @xmath1 and @xmath31 be an @xmath49-cartier divisor on @xmath89 supported in @xmath286 .
then the @xmath31-mmp holds for @xmath400 and , if @xmath31 is pseudo - effective , then the @xmath31-mmp followed by the @xmath31-contraction of a resulting model gives a rational @xmath72-contraction @xmath373 for the class @xmath52 of @xmath31 .
it is well - known that , in this situation , the @xmath31-mmp is equal to the lmmp with an initial model @xmath401 for @xmath402 and @xmath403 of proposition [ prop-0log cones coincide in u ] .
each birational transformation is a rational @xmath72-contraction and they are not isomorphic as log pairs ( monotonicity ) of log discrepancies ) .
existence of transformations ( divisorial contractions and flips ) and their termination hold by the klt slt lmmp . in the pseudo - effective case
, we get a wlc model and the required rational @xmath72-contraction by the semiampleness .
otherwise , @xmath31 will be negative on some fibration and thus not pseudo - effective .
[ cor - num dmmp 0log]let @xmath13 be a @xmath75-log pair with projective @xmath1 and such that the components of @xmath294 generate @xmath265 .
then for any @xmath49-cartier divisor @xmath31 on @xmath89 , the @xmath31-mmp holds for @xmath400 and , if @xmath31 is pseudo - effective , then the @xmath31-mmp followed by the @xmath31-contraction of a resulting model gives a rational @xmath72-contraction @xmath373 for the class @xmath52 given by @xmath31 .
the class @xmath52 can be replaced by numerical @xmath359 .
termination in this case is universally bounded ( the bound depends only on the variety @xmath1 ) .
immediate by corollary [ cor - dmmp on 0log ] .
the boundedness for termination follows from corollary [ cor - ft finite contractions ] , because @xmath1 is ft ( cf .
the following lemma ) .
[ cor - dmmp on ft ]
let @xmath1 be an ft variety .
then the @xmath31-mmp holds for @xmath400 and , if @xmath31 is pseudo - effective , then the @xmath31-mmp followed by the @xmath31-contraction of a resulting model gives a rational @xmath72-contraction @xmath360 for the class @xmath359 of @xmath336 containing @xmath31 .
termination in this case is universally bounded ( the bound depends only on variety @xmath1 ) .
immediate by lemma [ lemma - ft=0log ] and corollary [ cor - num dmmp 0log ] .
theorem [ thrm - finite wlc big klt ] gives the finiteness of projective wlc models for a klt pair @xmath13 of general type .
it is well - known that all wlc models are crepant birationally isomorphic . moreover , if @xmath393 and @xmath404 are two wlc models of @xmath13 , then their natural ( identical on rational functions ) birational isomorphism @xmath405 can be factored into projective(@xmath406 ) crepant minimizations @xmath407 @xmath408 and a small log flop @xmath409 : @xmath410\ar@{-->}[r ] & v'\ar[d]\\ y\ar@{-->}[r ] & y'/z . } \ ] ] a minimal model @xmath411 is maximal for the contraction order @xmath412 for wlc models of @xmath13 : @xmath133 is @xmath34-factorial , projective@xmath413 and blows up all prime @xmath31 with @xmath207 ( cf .
the proof of theorem [ thrm - finite wlc big klt ] ) .
moreover , @xmath414 is a slt wlc model of @xmath13 exactly when @xmath177 is projective . in its turn
, each crepant birational contraction can be factored into a sequence of extremal divisorial and small contractions and their number is bounded by the relative picard number @xmath415 for a projective@xmath416 minimization @xmath417 of the log canonical model . thus the remaining factorization concerns the log flop .
note also that @xmath418 is projective but @xmath177 may be not . if @xmath419 , any klt model , in particular , above minimization is projective@xmath416 . however , if @xmath420 , @xmath177 can be nonprojective and factorization of birational isomorphism @xmath421 or @xmath422 into elementary ( extremal ) flops is out of our grasp .
thus we consider only projective wlc models and focus on minimal ( slt ) models @xmath423 .
then their birational isomorphism @xmath405 gives a small log flop and can be factored into elementary flops .
( note that , for an elementary flop , the flopping locus can be very nonelementary , e.g. , not an irreducible curve in dimension @xmath424 . ) the general case , not necessarily of general type models , will be considered in corollary [ cor - small klt slt wlc models in elementary log flops ] below .
here we have a little bit stronger result .
it is natural to consider the factorization problem for flops of projective @xmath75-log pairs @xmath13 .
however to control polarization in @xmath286 we need to consider big @xmath65 and thus ft varieties by lemma [ lemma - ft=0log ] . note that a @xmath34-factorialization of an ft variety is again an ft variety , and , for any rational @xmath72-contraction @xmath388 with projective @xmath11 , @xmath11 is ft ( * ? ? ?
* lemma 2.8 ) .
[ cor - qfact ft flops]let @xmath425 be two @xmath34-factorial ft varieties .
then any small birational isomorphism @xmath405 can be factored into elementary small birational ( flips , antiflips , or flops ) transformations@xmath416 .
the number of such transformations is bounded ( the bound depends only on @xmath11 ) .
the factorization is a general fact .
for instance , we can use the @xmath31-mmp , where @xmath31 is the birational transform of any polarization from @xmath426 .
each transformation will be an elementary @xmath31-flip .
the number of flips is bounded by the number of rational @xmath72-contractions with projective @xmath382 according to corollary [ cor - ft finite contractions ] .
[ cor - slt flops ] let @xmath423 be two klt slt wlc models of general type . then any ( small ) log flop @xmath427 can be factored into elementary log flops and the number of such flops is bounded ( the bound depends only on @xmath393 ) .
immediate by corollary [ cor - qfact ft flops ] for ft variety @xmath428 .
we conclude the section with results that shed a light on terminology : _ a mobile cone @xmath362 as a polarizations cone .
_ [ prop - small iso mobile cones ] let @xmath1 be a variety with a reduced divisor @xmath50 and @xmath368 be a small isomorphism . then @xmath429 induces an isomorphism @xmath430 that preserves cones @xmath431 . if the isomorphism is a rational @xmath72-contraction of a class @xmath52 in @xmath350 , then under the above identification : @xmath432 . for ft @xmath433 , the last subcone @xmath434 . uses the birational invariance of mob+exc decomposition .
of course , all definitions work and the result holds for general @xmath1 but cones are not necessarily closed nor polyhedral . [ prop - ft mobile cones ] let @xmath1 be an ft variety , and @xmath359 be the classes in @xmath362 with rational @xmath72-contractions @xmath360 .
then @xmath435 in @xmath265 , where @xmath436 is the ample cone .
furthermore , @xmath437 , @xmath438 in @xmath265 .
transform the decomposition into the classes @xmath359 of @xmath439 into their polarization cones : @xmath437 .
the equation is given by the pull back @xmath440 $ ] of the numerical classes of ample divisors @xmath369 on @xmath382 by @xmath441 .
the points of polyhedron @xmath46 fills up the the cube @xmath33 .
[ prop - monotone n]@xmath46 satisfies the following : 1 .
( monotonicity ) if @xmath442 and @xmath257 , then @xmath443 ; 2 . ( the face of geography ) the linear span of @xmath46 is the linear span of a minimal face @xmath444 of @xmath33 containing @xmath46 ; moreover , equations of the span have the form @xmath74 , and are the linear equations for @xmath445 ; 3 .
@xmath446 .
the monotonicity is immediate by the inequality for kodaira dimensions : @xmath447 .
however , the slt lmmp is sufficient . for this
, one can use the stability of wlc models @xcite .
indeed , by the compact and convex property of @xmath46 , there exists a maximal @xmath448 in the direction @xmath449 , that is , @xmath450 $ ] .
if @xmath451 $ ] , the monotonicity holds in the direction . otherwise , let @xmath128 be a slt wlc model of @xmath129 .
then by ( * ? ? ?
* corollary 9 and addendum 5 ) after finitely many log flops of this model , it will become a wlc model such that , for each @xmath452\ ] ] near @xmath31 , @xmath453 has a mori log fibration ( cf .
the proof of theorem [ mainthrm - geography ] ) .
then , since @xmath454 , for a generic curve @xmath194 of the fibration , @xmath455 a contradiction .
the other two properties follow from the monotonicity and definitions . [ [ separatrix ] ] separatrix + + + + + + + + + + so , to find the largest lower bound of @xmath46 , or the minimum divisors ( by the closedness ) of @xmath46 , is sufficient to determine @xmath46 .
such a bound @xmath456 will be called a _
separatrix_. it consists of boundaries @xmath151 which _ separate _
@xmath46 : for any neighborhood @xmath135 of @xmath65 in the face @xmath444 , there exists a boundary @xmath457 , that is , @xmath458 and @xmath459 .
@xmath460 if and only if @xmath151 and is a limit of boundaries @xmath461 with @xmath459 . [ prop - s = closed rational poly]@xmath456 is a closed rational polyhedron ( usually nonconvex ) , and @xmath462 more precisely , a bordering facet of @xmath46 lies either in a facet of @xmath33 ( or of @xmath444 ) , or in @xmath456 , and @xmath456 is the closure of union of those facets . a facet in the proposition is a facet of geography @xmath46 .
( however , the statement holds also for facets of the polyhedron @xmath46 . ) a facet @xmath56 of _ geography _
@xmath46 is _ bordering _ if @xmath463 .
immediate by theorem [ mainthrm - geography ] and proposition [ prop - monotone n ] .
the separatrix plays an important role in birational geometry of uniruled varieties , because it divides mori log fibrations from wlc models ( see corollary [ cor - geoface = cube for general ] below ) . to factorize the birational transformations of those fibrations into more elementary standard transformations ( links ) , we will use a path on @xmath456 connecting corresponding points .
the best one would be a segment on @xmath456 connecting the points but the separatrix is usually not convex .
thus we can use _ geodesics _ on @xmath456 according to the euclidean metric of the ambient space @xmath48 or induced by a projection .
we prefer a natural projection from the _ origin _
@xmath464 of the face of geography
@xmath444 : @xmath465 where @xmath74 for the equations of @xmath46 and @xmath199 otherwise .
[ lemma - nonempty separatrix]@xmath466 if and only if @xmath467 and @xmath468 , or equivalently , @xmath469 and @xmath470 .
immediate by proposition [ prop - s = closed rational poly ] and definition .
let @xmath471 be a projection from @xmath464 onto a hyperplane @xmath369 in @xmath444 given by the equation @xmath472 ( here we discard all @xmath73 corresponding to the equations of @xmath46 ) .
( see diagram [ dia projection ] below . )
separatrix.eps ( -7,-7)@xmath464 ( 25,36)@xmath65 ( 68,70)@xmath31 ( 52,52)@xmath473 ( 35,85)@xmath46 ( 75,24)@xmath369 ( 45,10)@xmath456 ( 45,50)@xmath474 ( 63,70)@xmath474 by lemma [ lemma - nonempty separatrix ] , @xmath475 is well - defined on @xmath456 .
[ prop - image of separatrix]if @xmath466 , the image @xmath476 is a closed convex rational polyhedron .
moreover , @xmath475 is @xmath72-to-@xmath72 on @xmath456 .
immediate by theorem [ mainthrm - geography ] and proposition [ prop - monotone n ] .
the @xmath72-to-@xmath72 property follows from definition .
[ [ geography - of - general - type ] ] geography of general type + + + + + + + + + + + + + + + + + + + + + + + + + it is a geography @xmath46 containing a boundary @xmath65 such that the pair @xmath13 is of general type .
this assumes that the geography @xmath46 is equipped with certain functions , e.g. , @xmath166 of section [ section2-geography ] , @xmath477 etc .
one of those functions is the numerical kodaira dimension @xmath478 .
( for other functions , see section [ section2-geography ] . )
it behaves almost constant .
[ prop - kd nd are constants]for each @xmath479 ( in particular , @xmath480 ) , @xmath481 the same holds for @xmath482 .
immediate by theorem [ mainthrm - geography ] and the following fact on a rather high projective model @xmath177 .
if @xmath483 are nef divisors on @xmath177 , then for any subvariety @xmath484 and any real numbers @xmath485 such that @xmath486 , @xmath487 is big on @xmath488 if and only if @xmath489 is big on @xmath488 , and if @xmath31 is big on @xmath488 , then @xmath487 is big on @xmath488 . thus @xmath490 . for @xmath482 , one can use semiampleness .
[ cor - geoface = cube for general ] let @xmath46 be a geography of general type
. then @xmath491 ( the number of prime components of @xmath50 ) , in particular , @xmath468 and , for any @xmath460 , @xmath492 .
immediate by proposition [ prop - monotone n ] and the open property of general type : if @xmath493 is of general type , then @xmath15 is also of general type for any @xmath494 near @xmath65 . for the kodaira dimension
, this follows from definition , for numerical one from the slt lmmp .
if @xmath495 are two reduced b - divisors on @xmath1 , then there exists a natural inclusion of geographies : @xmath496 and @xmath497 where @xmath498 , the exceptional part of @xmath499 . indeed , by definition @xmath12 is a wlc model of @xmath13 if and only if this holds for @xmath15 .
this correspondence preserves @xmath35 .
the transition from @xmath50 to @xmath500 will be called an _
extension_. it allows us to perturb any wlc model to a much better model as follows . for instance , if @xmath12 is a wlc model of @xmath13 not slt or not of general type , e.g. , for @xmath501 , we can convert it to a polarized wlc model .
indeed , if @xmath500 is sufficiently larger than @xmath50 , the geography @xmath502 is of general type , and @xmath153 belongs to a face of a country @xmath503 .
thus @xmath504 where @xmath505 , has a lc model @xmath506 , and by lemma
[ lemma - closed n ] a limit of such models gives a wlc model @xmath12 of @xmath13 ( might be different from @xmath237 above or / and not slt ; cf . with convention [ convention ] ) with a polarization of @xmath507 .
moreover , as we will see below , for a good choice of @xmath500 , @xmath11 will be @xmath34-factorial . of course
, the variety @xmath206 obtained by a perturbation is not unique .
nonetheless it is useful ( see corollary [ cor - small klt slt wlc models in elementary log flops ] and theorem [ thrm - mori morphism is cte ] below ) .
[ [ generality - conditions ] ] generality conditions + + + + + + + + + + + + + + + + + + + + + in what follows we assume that @xmath50 is a reduced divisor , @xmath508 is a slt initial pair of general type , and components of @xmath50 generate @xmath265 .
it is easy to satisfy these conditions on a log resolution of @xmath508 after an extension of @xmath50 .
each country in such a geography has a unique model .
[ thrm - unique wlc for general geo]under the generality conditions on @xmath508 , let @xmath509 be an internal point of a country @xmath81 in the geography @xmath46 .
then @xmath13 has a unique wlc model @xmath12 .
moreover , the model is a klt slt wlc model of @xmath13 , in particular , @xmath391 , @xmath11 is projective @xmath34-factorial , depending only on @xmath81 , and the model is lc ; components of @xmath510 generate @xmath511 . the key instrument to investigate generic geographies
is the following .
[ lemma - rho of wlc / lc]under the generality conditions on @xmath13 ,
let @xmath393 be a klt slt wlc model of @xmath13 for a boundary @xmath65 in a subclass @xmath512 and @xmath513 be its lc contraction .
then @xmath514 is ft , components of @xmath510 generate @xmath511 , and @xmath515 since the initial model @xmath13 is klt slt and @xmath516 , by the klt slt lmmp there exists a klt slt wlc model @xmath393 of @xmath13 and components of @xmath517 generate @xmath511 .
otherwise , we get a mori log fibration and nothing to prove . by construction of the lc model , @xmath518 is a @xmath75-log pair and by lemma [ lemma - ft=0log ] , @xmath514 is ft .
the contraction exists by the base freeness @xcite , because @xmath519 is big .
the klt slt properties imply also that the image @xmath520 of @xmath52 under the natural linear map @xmath521 has the same codimension : @xmath522 ( * ? ? ?
* lemma 3.3.2 ) . by the open property in theorem [ mainthrm - geography ] ,
@xmath523 when @xmath524 .
on the other hand , the generation property of @xmath510 gives the surjectivity of the natural linear map @xmath525.\ ] ] by construction and the definition of @xmath35 , @xmath520 lies in the kernel of the last map .
thus @xmath526 finally , note that the constructed model @xmath514 can be different from the one in the statement of lemma . however , the difference is in a log flop . by corollary [ cor - qfact ft flops ] , it is a composition of elementary small birational transformations which preserve the required properties .
let @xmath12 be a slt wlc model of @xmath13 .
since the latter pair is klt initial , the former pair is klt and @xmath391 ( * ? ? ?
* lemma 2.4 ) . by our assumptions and proposition [ prop - kd nd are constants ] , the pair @xmath393 has general type and a lc model @xmath527 .
thus by lemma [ lemma - rho of wlc / lc ] and corollary [ cor - geoface = cube for general ] , @xmath528 and @xmath529 , the uniqueness .
the last uniqueness was established only for slt models .
any other wlc model is a rational @xmath72-contraction of @xmath514 and thus coincide with @xmath237 ( see the projective case in corollary [ cor - ft eff decomposition ] ; in general , cf .
the proof of theorem [ thrm - finite wlc big klt ] ) .
models of facets and ridges for a generic geography also behave quite controllable . in dimension
@xmath530 , all facets are bordering and ridges only cube bordering . in dimension
@xmath531 , flopping facets are impossible and the birational modifications of ridges are only divisorial ( see the classification of ridges below ) .
we divide facets into two types according to their location in @xmath46 : _ bordering _ and _ internal_. furthermore , they are divided into two subtypes , respectively : _ cube bordering , ( mori ) fibering _ and _ flopping , divisorial_. let @xmath56 be a facet of @xmath46 under the generality conditions , @xmath532 , and @xmath12 be a projective wlc model of @xmath13 .
first , we describe neighboring classes and properties of their models ( only projective , possibly not @xmath34-factorial ) and then prove them ( see the proof of theorem [ thrm - facet types ] below ) .
[ [ cube - bordering ] ] cube bordering + + + + + + + + + + + + + + the facet @xmath56 is bordering for @xmath46 and its closure @xmath533 lies in a facet of @xmath33 , or satisfies an equation @xmath71 or @xmath72 in @xmath46 . in this case
, @xmath12 can be non klt . if it is klt , the wlc models can be classified as types below .
we are not interested in this because after a perturbation we have internal boundaries ( see the proof of corollary [ cor - small klt slt wlc models in elementary log flops ] below ) .
[ [ mori - fibering ] ] ( mori ) fibering + + + + + + + + + + + + + + + the facet @xmath534 is also bordering for @xmath46 but @xmath535
. the pair @xmath130 @xmath391 , is a klt slt wlc model of @xmath13 and @xmath237 has a mori log fibration @xmath536 given by the lc contraction .
the base @xmath537 is @xmath34-factorial .
such a mori fibration will be called _ polarized _ by the boundary @xmath519 . the model @xmath393 and the fibration are unique for @xmath65 and both morphisms @xmath538 depend only on the facet @xmath56 .
moreover , there exists a unique ( neighboring ) country @xmath81 of @xmath46 such that @xmath533 is its facet . for any @xmath539 ,
@xmath312 is a lc model of @xmath15 , and @xmath393 is limiting from @xmath81 . on the other side ,
if @xmath540 and is sufficiently close to @xmath65 , the mori log fibration @xmath541 is a slt mori log fibration of @xmath15 .
the facet is given by the equation @xmath542 in @xmath543 where @xmath178 is a curve on a rather high model @xmath177 of @xmath1 , contracted on @xmath537 but not on @xmath237 .
the function is @xmath78 on @xmath81 and @xmath244 in the points of @xmath47 near the points of @xmath56 for the linear extension of @xmath544 from @xmath81 .
the functions @xmath545 have the same signatures on @xmath546 .
the next two types are possible respectively only in dimension @xmath547 and @xmath548 .
[ [ flopping ] ] flopping + + + + + + + + the facet is internal for @xmath46 : @xmath549 , and @xmath533 is a facet of two ( neighboring ) countries @xmath550 in @xmath46 .
the pair @xmath13 has exactly three wlc models which are klt of general type and related by an elementary small flop , extraction or contraction .
countries @xmath550 give exactly two slt wlc models @xmath551 @xmath552 @xmath553 @xmath554 of @xmath13 , where @xmath555 @xmath556 are lc models of @xmath15 for @xmath557 , respectively .
those models of @xmath13 are limiting from @xmath550 , respectively .
the third model @xmath558 , of @xmath13 is a lc one .
both contractions @xmath559 are elementary small .
in particular , @xmath537 is not @xmath34-factorial and the third model is not slt .
the facet is given by equations @xmath560 in @xmath561 where @xmath562 are curves on a rather high model @xmath177 of @xmath1 contracted on @xmath537 but not on @xmath563 , respectively ; @xmath564 on @xmath116 and @xmath315 on @xmath117 ; @xmath565 on @xmath116 and @xmath566 on @xmath117 .
the functions @xmath545 have the same signatures on @xmath567 .
[ [ divisorial ] ] divisorial + + + + + + + + + + the facet is also internal for @xmath46 : @xmath549 , and @xmath533 is a facet of two ( neighboring ) countries @xmath550 in @xmath46 .
the pair @xmath13 has exactly two wlc models which are klt of general type and related by an elementary divisorial extraction or contraction of a prime divisor @xmath31 .
the country @xmath116 gives exactly one slt wlc model @xmath568 , of @xmath13 , where @xmath569 is a lc model of @xmath15 for @xmath570 .
the model @xmath571 is limiting from @xmath116 .
the second model @xmath558 , of @xmath13 is the lc one , the contraction @xmath572 of @xmath31 is elementary divisorial . in particular , @xmath537 is @xmath34-factorial , @xmath573 is a slt wlc model , but not a slt wlc model of @xmath13 .
the model @xmath573 is limiting from @xmath117 . the facet @xmath56 is given by the equation @xmath542 in @xmath574 where @xmath178 is a curve on a rather high model @xmath177 of @xmath1 contracted on @xmath537 but not on @xmath575 ; @xmath576 on @xmath577 and @xmath78 on @xmath116 .
similarly , @xmath56 can also be given by the equation @xmath207 in @xmath578 : @xmath242 on @xmath579 and @xmath78 on @xmath117 .
[ thrm - facet types]under the generality conditions on @xmath508 , let @xmath56 be a facet of the geography @xmath46 .
then it has one of four above types : cube bordering , fibering , flopping or divisorial , and behaves accordingly .
( see diagram [ dia4 facets ] below . )
facets.eps ( 49,103)(1 ) ( 72,53)(3,4 ) ( 24,17)(2 ) @xmath580 a bordering facet @xmath56 of @xmath46 has the cube bordering or fibering type by definition , proposition [ prop - s = closed rational poly ] and corollary [ cor - geoface = cube for general ] .
cube bordering type : @xmath56 satisfies an equation @xmath71 or @xmath72 and lies in the corresponding facet of @xmath581 by corollary [ cor - geoface = cube for general ] .
for the boundaries @xmath532 of the other type , @xmath65 is _ internal _ : @xmath535 . by the generality assumption , @xmath13 is a klt slt initial pair of itself and components of @xmath516 generate @xmath265 . in particular
@xmath65 is big .
thus any wlc model @xmath12 of @xmath13 is klt ( * ? ? ?
* lemma 2.4 ) with big @xmath391 and has a lc contraction @xmath582 @xcite . if @xmath514 is projective , it is ft by lemma [ lemma - ft=0log ] .
such a model @xmath393 exists because @xmath151 and it satisfies lemma [ lemma - rho of wlc / lc ] when it is a slt wlc model of @xmath13 .
the slt model can be constructed by the slt lmmp from the above initial model .
fibering type .
suppose that @xmath56 is fibering and @xmath393 is a slt wlc model of @xmath13 . by construction and lemma [ lemma - rho of wlc / lc ] ,
@xmath514 is a mori log fibration .
indeed , it is not isomorphism and an extremal fibration by corollary [ cor - geoface = cube for general ] . the model ( in the projective class ) and
contraction are unique by corollary [ cor - ft eff decomposition ] .
since @xmath237 is @xmath34-factorial and @xmath514 is ft , @xmath537 is @xmath34-factorial too ( cf . @xcite ) .
the boundary @xmath519 is a polarization of @xmath514 .
thus if we perturb @xmath65 to @xmath153 in a positive direction , @xmath583 , then @xmath539 for a unique neighboring country @xmath81 of @xmath46 and @xmath312 is a wlc ( actually lc ) model of @xmath15 with a polarization @xmath584 over @xmath537 . hence the functions @xmath585 have the same signatures on @xmath546 . by lemma
[ lemma - closed n ] , @xmath393 is limiting from @xmath81 , and by theorem [ mainthrm - geography ] , @xmath533 is a facet of @xmath81 .
a perturbation @xmath312 in a negative direction , @xmath540 , gives a slt mori log fibration @xmath541 of @xmath15 . by construction and theorem
[ thrm - unique wlc for general geo ] , a required equation @xmath542 can be given by a curve @xmath178 on @xmath177 with the curve image in a fiber of @xmath514 .
the next two types are possible only in dimension @xmath548 and internal : @xmath549 , by theorem [ mainthrm - geography ] @xmath533 is a facet of two ( neighboring ) countries @xmath550 in @xmath46 , and @xmath480 .
( according to the open property in theorem [ mainthrm - geography ] , @xmath586 . ) by our assumptions and proposition [ prop - kd nd are constants ] , any wlc model @xmath393 of @xmath13 has general type and a lc model @xmath587 with the lc contraction @xmath588 .
let @xmath393 be a slt wlc model of @xmath13 .
note that it is not lc : @xmath589 .
otherwise , any small perturbation @xmath153 of @xmath65 in @xmath33 is @xmath590 : @xmath591 , a contradiction .
hence @xmath592 , or equivalently , @xmath514 is elementary by lemma [ lemma - rho of wlc / lc ] . flopping type .
now we claim that @xmath56 is of flopping type if @xmath11 is small .
it is possible only in dimension @xmath593 . by corollary [ cor - ft eff decomposition ]
the pair @xmath13 has exactly three projective wlc models @xmath594 which are related by an elementary small flop , extraction or contraction : @xmath595\ar[rd ] & & y_2\ar[ld]\\ & t / z . & } \ ] ] ( the cone @xmath596 is a line @xmath49 with three subcones ( classes ) : @xmath597 @xmath598 . )
the boundary @xmath519 is a polarization of @xmath514 .
if we perturb @xmath65 to @xmath153 in a positive direction , @xmath583 , that is , @xmath584 becomes a polarization , then @xmath599 is limiting from @xmath116 if @xmath600 , otherwise @xmath601 is limiting from @xmath117 .
so , the boundary @xmath539 for one of the two neighboring countries @xmath81 of @xmath46 and @xmath312 is a wlc ( actually lc ) model of @xmath15 with a polarization @xmath584 .
the same holds for another neighboring country after a log flop of @xmath514 . by theorem [ mainthrm - geography ]
, @xmath533 is a facet of @xmath116 and of @xmath117 .
the third wlc model of @xmath13 is @xmath573 . in particular , @xmath537 is not @xmath34-factorial and the model is not slt . by construction , the functions @xmath188 have the same signatures on @xmath567 . by construction and theorem [ thrm - unique wlc for general geo ] , required equations @xmath560 can be given by curves @xmath562 on @xmath177 with the curve image in a fiber of @xmath602 , respectively .
divisorial type . in the remaining last type case
, @xmath514 is a divisorial contraction of a unique prime divisor @xmath31 by the extremal and @xmath34-factorial properties .
again by corollary [ cor - ft eff decomposition ] the pair @xmath13 has exactly two projective wlc models which are related by an elementary divisorial extraction or contraction .
( the cone @xmath596 is a line @xmath49 with two subcones ( classes ) : @xmath603,(0,+\infty)$ ] . )
the boundary @xmath519 is a polarization of @xmath514 .
thus if we perturb @xmath65 to @xmath153 in a positive direction , @xmath583 , that is , @xmath584 becomes a polarization , say for @xmath600 , then @xmath599 is limiting from @xmath116 .
the second wlc model of @xmath13 is @xmath573 . in particular ,
@xmath537 is @xmath34-factorial , @xmath573 is a slt wlc model , but not a slt wlc model of @xmath13 : @xmath207 .
the model @xmath573 is limiting from @xmath117 . by construction and theorem [ thrm - unique wlc for general geo ] , required equations @xmath604 can be given respectively by the contracted divisor @xmath31 and by a curve @xmath178 on @xmath177 with the curve image in a fiber of @xmath514 .
[ cor - small klt slt wlc models in elementary log flops]let @xmath423 be two klt slt wlc models . then any small log flop @xmath427 can be factored into elementary log flops .
it is sufficient to assume that the transformation is small and @xmath605 is the birational image of @xmath519 .
we can prove the corollary using elementary log flops of the @xmath31-mmp with @xmath606 where @xmath607 is an effective polarization on @xmath426 @xcite . according to the stability of (
* addendum 5 ) , the @xmath31-mmp of @xmath11 is equivalent to the klt slt lmmp of the initial model @xmath608 for some @xmath609 . we can use a geography instead of the @xmath31-lmmp which is more symmetric . pick up two effective ( generic ) polarizations @xmath610 on @xmath611 respectively .
( in dimension @xmath612 , one can suppose that @xmath610 are prime . )
take a log resolution @xmath508 of @xmath613 where @xmath614 and @xmath615 are the exceptional divisors of @xmath616 .
then for an appropriate klt boundary @xmath6 , both models @xmath393 and @xmath404 are klt slt wlc models of @xmath13 . after a perturbation by polarizations @xmath610 , they will become klt log canonical models @xmath617 respectively of @xmath618 where @xmath619 are perturbations of @xmath65 in @xmath33 .
in other words , there are two classes @xmath620 in the geography @xmath46 such that @xmath621 and @xmath423 are limiting from @xmath620 respectively ; @xmath622 . as above by the stability ( * ? ? ? * addendum 5 ) , the geography is conical near @xmath65 .
hence for a perturbation , @xmath620 are actually countries and the segment @xmath623 $ ] intersects only facets @xmath56 , which are also adjacent to @xmath65 : @xmath624 is limiting for the facets . by theorem [ thrm - facet types ] , this gives a required factorization .
but for this we need to expand @xmath50 to fulfil the generality conditions .
we also need to verify that all intersections @xmath623\cap{\mathfrak f}$ ] have flopping type . in other words , if a prime divisor @xmath625 is not exceptional on @xmath237 , then @xmath625 should not be exceptional on the wlc models of all @xmath626 $ ] and vice versa . by construction , @xmath610 are mobile and big on the divisors of all those wlc models . thus if all divisors in the extension of @xmath50 are also mobile and big ( e.g. , ample on @xmath89 ; cf .
the perturbation in the proof of theorem [ thrm - mori morphism is cte ] ) on those divisors , we secure the required nonexceptionality : each wlc model for @xmath626 $ ] corresponds to an effective perturbation of @xmath519 . on the other hand , for @xmath65 with all @xmath627 or close to @xmath72 and any exceptional on @xmath237 b - divisor @xmath306 , by the klt assumption for @xmath628 the inequality
@xmath629 holds near @xmath65 .
hence @xmath306 will be exceptional on the wlc models of all @xmath626 $ ] .
finally , note that this proof allows us to relax the klt slt lmmp to the klt slt big one .
a birational transformation @xmath630 gives a ( small ) log flop if and only if the transformation is small and @xmath631 is the birational image of @xmath519 ( * ? ? ?
* proposition 2.4 ) .
we divide ridges also into two types according to their location in @xmath46 : _ bordering _ and _ internal_. furthermore , the bordering ones are divided into two subtypes : _ cube bordering _ and _ ( mori ) fibering_. the internal ridges correspond to general type and will be also referred to as _ birational_. in its turn , fibering and birational types have three subtypes each .
essentially they can be divided into types as _ only flopping _ , _ divisorial and flopping _ , etc .
but we prefer just a conventional ordering 2a , 2b , etc : e.g. , 2b means fibering type b. so , let @xmath83 be a ridge of @xmath46 under the generality conditions , @xmath632 , and @xmath12 be a projective wlc model of @xmath13 .
first , we describe the neighboring classes and properties of their models ( only projective , possibly not @xmath34-factorial ) and then prove the statements ( see the proof of theorem [ thrm - ridge types ] below ) .
[ [ cube - bordering-1 ] ] cube bordering + + + + + + + + + + + + + + the ridge @xmath83 is _ bordering _ , that is , @xmath633 , and the closure @xmath634 either lies in a ridge of @xmath33 , or in a facet @xmath56 of @xmath33 . in the former case ,
@xmath83 satisfies two equations @xmath71 or @xmath72 , @xmath635 or @xmath72 in @xmath46 . in the latter case
, @xmath56 satisfies an equation @xmath71 or @xmath72 in @xmath33 and @xmath83 is a ridge of @xmath46 in @xmath56 . in both cases ,
@xmath12 can be non klt .
if it is klt , the wlc models can be classified as types below . for the internal types , the wlc models @xmath12 of @xmath636 , are klt and @xmath391 .
each of them has the lc contraction @xmath637 and @xmath514 is ft .
the base @xmath537 depends only on the ridge @xmath83 .
the pair @xmath393 with projective @xmath11 is a slt wlc model of @xmath13 if and only if @xmath638 and there exist exactly two extremal contractions of @xmath237 over @xmath537 . [
[ fibering - or-2 ] ] fibering or 2 + + + + + + + + + + + + + the ridge @xmath639 is also bordering for @xmath46 but @xmath535 .
equivalently , @xmath640 .
the closure @xmath634 is a facet of @xmath641 ( neighboring ) facets @xmath642 in @xmath46 , and a ridge of @xmath643 countries @xmath644 ; @xmath645 has ( closed ) facets @xmath646 ( see diagram [ dia5 facets2 ] below ) . ridge.eps ( 10,90)@xmath116 ( 37,80)@xmath117 ( 59,45)@xmath645 ( 47,10)@xmath647 ( -5,103)@xmath648 ( 32,103)@xmath649 ( 63,83)@xmath650 ( 72,60)@xmath69 ( 74,40)@xmath651 ( 50,-8)@xmath652 ( 68,8)@xmath653 ( 2,35)@xmath83 ( 9.5,39.5)@xmath474 facets @xmath654 are fibering , and @xmath655 are _ birational _ : flopping or divisorial .
let @xmath656 be mori log fibrations of facets @xmath654 , @xmath657 , be lc models of the birational facets , and @xmath658 , be lc models of countries @xmath645 .
if @xmath653 is divisorial , we suppose that @xmath649 is divisorial ( after interchanging @xmath116 with @xmath647 ) .
each @xmath659 , is flopping , and corresponding @xmath660 are not @xmath34-factorial .
the varieties @xmath661 are @xmath34-factorial .
the varieties @xmath662 are @xmath34-factorial if and only if @xmath663 respectively are divisorial .
more precisely , if @xmath649 is divisorial , then @xmath664 is an elementary divisorial contraction of a prime divisor @xmath120 .
the similar holds for divisorial @xmath653 with contraction @xmath665 of a prime divisor @xmath666 ( possibly the same as b - divisor @xmath120 ) . for @xmath667 , the variety @xmath668 has ( only ) two elementary contractions @xmath669 .
the same holds for @xmath670 if @xmath663 respectively are flopping .
all contractions @xmath671 are small .
the same holds for @xmath672 if @xmath663 respectively are flopping .
each @xmath673 , has a natural contraction to @xmath537 , @xmath674 is extremal , ft and fibered for @xmath675 .
the base @xmath537 is not @xmath34-factorial exactly when @xmath676 is small .
the base @xmath537 is @xmath34-factorial for @xmath677 , in particular , in dimension @xmath678 , and for types 2a with @xmath679 , 2b - c . in all other cases , @xmath537 can be non @xmath34-factorial .
the pair @xmath13 has exactly two wlc models @xmath551 @xmath680 , @xmath681 , @xmath682 , @xmath683 , equipped with a mori log fibration : @xmath684 @xmath685 over @xmath537 , respectively .
the fibrations @xmath656 are _ square _ birational , that is , both contractions @xmath686 are birational if and only if the facets @xmath654 are coplanar . for @xmath679 , 2a is only possible , the pair @xmath13 has exactly one wlc model @xmath571 , the latter is slt and has two mori log fibrations : @xmath687 over @xmath537 , the base @xmath537 is @xmath34-factorial and @xmath688 are mori log fibrations too , @xmath689 are never coplanar . the flopping ( small ) contractions exist only for @xmath547 .
thus , @xmath690 in dimension @xmath531 ( there are no fibering ridges in dimension @xmath691 ) . in dimension @xmath547 , the only nonflopping cases are : 2a with @xmath679 , 2b with @xmath692 and 2c with @xmath693 .
each pair @xmath694 , @xmath695 , @xmath667 , is a klt slt wlc model of @xmath13 , limiting from @xmath645 .
the pairs @xmath696 satisfy exactly the same when respectively @xmath663 are flopping , and they are limiting from @xmath697 , respectively . each pair @xmath698 , is a non @xmath34-factorial nonslt projective wlc model of @xmath13 , limiting from @xmath69 .
the pairs @xmath699 satisfy exactly the same when respectively @xmath663 are flopping .
if @xmath649 is divisorial , @xmath700 is an @xmath34-factorial projective wlc model of @xmath13 but not a slt wlc model of @xmath13 .
the similar holds for @xmath653 .
there are no other projective wlc models of @xmath13 than above .
the closures @xmath701 are given by equations @xmath702 @xmath199 in @xmath703 , respectively where @xmath704 are curves on a rather high model @xmath177 of @xmath1 , contracted on @xmath661 but not on @xmath670 , respectively .
moreover , if @xmath705 is rather general , then @xmath564 on @xmath706 , and @xmath315 on @xmath707 whereas @xmath199 if and only if the fibration @xmath708 factors through @xmath685 : the diagram @xmath709\ar@{-->}[r ] & y_m\ar[d]\\ t_1\ar@{-->}[r ] & t_{m+1 } , } \ ] ] with the natural birational isomorphism @xmath710 and a ( natural ) rational contraction @xmath711 , is commutative . the similar holds for general @xmath712 .
the models @xmath656 are square birational if and only if @xmath713 . moreover , @xmath656 are _ generically isomorphic _ , that is , they are natural isomorphic over the general points of @xmath714 , if and only if the facets @xmath654 are coplanar .
if @xmath649 is divisorial , its closure @xmath715 is given by the equation @xmath716 in @xmath579 : @xmath717 on @xmath718 and @xmath199 on @xmath719 ; @xmath716 on @xmath720 if @xmath721 as b - divisor or @xmath653 is flopping .
if @xmath722 , then @xmath717 on @xmath720 .
the similar holds for @xmath666 if @xmath653 is divisorial .
other equations for the divisorial and flopping facets were given in theorem [ thrm - facet types ] above .
[ [ fibering - a - or-2a ] ] fibering a or 2a + + + + + + + + + + + + + + + + ( the sarkisov link of type iv . ) both @xmath723 .
the pair @xmath571 is a slt wlc model of @xmath13 .
the base @xmath537 is possibly not @xmath34-factorial for @xmath724 .
all intermediate facets @xmath725 , are flopping .
the only possible other slt wlc models @xmath694 of @xmath13 correspond to the countries @xmath726 , and they are log flops of @xmath571 , directed by a polarization on @xmath668 . more precisely , those flops are composition of elementary ones : @xmath727 the total birational transformation of fibrations : @xmath728\ar[d]&y_2\ar@{-->}[r ] & \cdots \ar@{-->}[r]&y_i\ar@{-->}[r]&\cdots\ar@{-->}[r ] & y_{m-1}\ar@{-->}[r ] & y_m\ar[d]\\ t_1\ar[rrrd]^{\not\cong } & & & & & & t_{m+1}\ar[llld]_{\not\cong}\\ & & & t / z . & & & } } \ ] ] [ [ fibering - b - or-2b ] ] fibering b or 2b + + + + + + + + + + + + + + + + ( the sarkisov link of type i and its inverse type iii . ) @xmath729 but @xmath730 .
the base @xmath537 is @xmath34-factorial .
the facet @xmath649 is divisorial with the elementary divisorial contraction @xmath731 and the only possible extremal contraction of @xmath732 is its mori fibration .
all other intermediate facets @xmath733 , are flopping .
the only slt wlc models of @xmath13 are @xmath734 , limiting from the countries @xmath645 , and they are log flops of @xmath735 , directed by a polarization on @xmath668 .
more precisely , those flops are composition of elementary ones : @xmath736 the total birational transformation of fibrations : @xmath737&y_2\ar[l]\ar@{-->}[r]&\cdots \ar@{-->}[r]&y_i\ar@{-->}[r]&\cdots\ar@{-->}[r ] & y_{m-1}\ar@{-->}[r ] & y_m\ar[d]\\ & & & & & & t_{m+1}\ar[llld]_{\not\cong}\\ t_1\ar[rrr]^{\cong } & & & t / z .
& & & } } \ ] ] [ [ fibering - c - or-2c ] ] fibering c or 2c + + + + + + + + + + + + + + + + ( the sarkisov link of type ii . ) @xmath738 .
the base @xmath537 is @xmath34-factorial .
facets @xmath663 are divisorial with elementary divisorial contractions @xmath739 , respectively .
the only possible extremal contractions of @xmath740 are their mori fibrations .
all other intermediate facets @xmath659 , are flopping .
the only slt wlc models of @xmath13 are @xmath741 , which are limiting from the countries @xmath645 , and they are log flops of @xmath735 , directed by a polarization on @xmath668 . more precisely , those flops are composition of elementary ones : @xmath736 the total birational transformation of fibrations : @xmath709 & y_2\ar[l]\ar@{-->}[r ] & \cdots\ar@{-->}[r]&y_i\ar@{-->}[r]&\cdots\ar@{-->}[r ] & y_{m-1}\ar[r ] & y_m\ar[d ] \\ t_1\ar[rrrd]^{\cong } & & & & & & t_{m+1}\ar[llld]_{\cong}\\ & & & t / z . & & & } \ ] ] [ [ internal - or - birational - or-3 ] ] internal , or birational , or 3 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + the ridge @xmath83 is _ internal _ if @xmath742 . in this case , as for internal facets , all wlc models @xmath12 are klt with @xmath391 and of general type with the lc model @xmath743 .
the model @xmath744 of @xmath1 depends only on @xmath83 .
the closure @xmath634 is a facet of @xmath745 ( neighboring ) facets @xmath746 in @xmath46 , and a ridge of @xmath643 countries @xmath644 ; @xmath645 has ( closed ) facets @xmath747 ( see diagrams 6 - 8 below ) .
any two subsequent facets @xmath748 are strictly in a half - plane .
all facets are birational : flopping or divisorial .
let @xmath749 , be lc models of the birational facets @xmath69 and of countries @xmath645 , respectively .
if there is a divisorial facet , we suppose that @xmath649 is divisorial with contraction @xmath731 after shifting indices @xmath750 .
any variety @xmath751 is @xmath34-factorial , and @xmath660 is @xmath34-factorial if and only if @xmath69 is divisorial .
the base @xmath537 is @xmath34-factorial if and only if there are two subsequent divisorial facets @xmath689 with contractions @xmath752 .
all varieties @xmath753 are ft extremal or identical , and @xmath668 are ft with @xmath754 .
the pairs @xmath755 are the nonslt wlc models of @xmath13 .
[ [ birational - a - or-3a ] ] birational a or 3a + + + + + + + + + + + + + + + + + + the weil - picard number @xmath756 all facets @xmath69 are flopping , and all contractions @xmath757 respectively elementary and extremal small
. the pairs @xmath758 @xmath695 , are the only slt wlc models of @xmath13 .
the models @xmath660 and base @xmath537 are not @xmath34-factorial .
there exists a sequence of elementary log flops in the facets @xmath69 : @xmath759 a subsequence @xmath760 of flops is directed with respect to a polarization on @xmath761 if the flopping facets @xmath762 are strictly in a half - plane . in general , the configuration of facets and countries is not symmetric , in particular , facets are not coplanar .
however two facets @xmath763 are coplanar if and only if @xmath764 is a generalized extremal and directed log flop@xmath765 with respect to their polarizations .
otherwise the span of @xmath69 intersects @xmath766 and @xmath767 is a generalized non - extremal and directed log flop@xmath765 with respect to a polarization .
an equation @xmath768 for the closure @xmath769 in @xmath770 , can be given by a curve @xmath771 on a rather high model @xmath177 of @xmath1 .
this allows us to find coplanar facets in terms of linear extensions for functions @xmath772 from certain countries .
the commuting , centrally symmetric , case is possible only for @xmath773 with coplanar pairs @xmath774 and @xmath775 ; both flopping loci are disjoint .
see diagram [ dia6 ] below : @xmath776&&\ar@{--}[ll]\ar@{--}[rd]&\\ y_1\ar@{-->}[r]\ar[rrd]^{\not\cong}&y_2\ar@{-->}[r ] & \cdots\ar@{-->}[r ] & y_{m-1}\ar@{-->}[r]&y_{m}\ar[lld]_{\not\cong}\\ & & t & & } \ ] ] l c r bir-ridge-a.eps ( 25,-20)(@xmath777 ) ( -5,-1)@xmath649 ( 23,70)@xmath69 ( 43,50)@xmath645 ( -8,55)@xmath650 ( 100,14)@xmath778 ( 63,73)@xmath651 ( 90,59)@xmath653 ( 45,12)@xmath83 ( 60,10)@xmath116 ( 6,23)@xmath117 ( 70,30)@xmath647 & & bir-ridge-ab4.eps ( 23,46)@xmath83 ( 85,-8)@xmath648 ( 8,-7)@xmath649 ( -8,105)@xmath650 ( 67,103)@xmath779 ( 40,15)@xmath116 ( 2,50)@xmath117 ( 30,80)@xmath118 ( 64,50)@xmath780 ( -64,-28)(@xmath773 , flopping loci are disjoint ) + [ [ birational - b - or-3b ] ] birational b or 3b + + + + + + + + + + + + + + + + + + the weil - picard number @xmath781 , and the only divisorial facets are @xmath663 with elementary divisorial contractions @xmath739 of a unique prime b - divisor @xmath120 .
all other facets @xmath69 are flopping with elementary small contractions @xmath782 with @xmath783 or @xmath784 , and with extremal divisorial contractions
@xmath674 , @xmath785 .
the pairs @xmath786 @xmath695 , @xmath667 , are the only slt wlc models of @xmath13 . the models @xmath662 are @xmath34-factorial .
all other models @xmath660 and the base @xmath537 are not @xmath34-factorial .
there exists a sequence of an elementary divisorial extraction , elementary log flops in the facets @xmath787 , and of an elementary divisorial contraction : @xmath788 ccc + & & + & & + bir-ridge-b.eps ( 30,-20)(@xmath777 ) ( -11,38)@xmath649 ( -6,59)@xmath650 ( 26,79)@xmath69 ( 44,62)@xmath645 ( 102,40)@xmath653 ( 65,80)@xmath651 ( 90,67)@xmath789 ( 35,-7)@xmath778 ( 52,24)@xmath83 ( 23,17)@xmath116 ( 10,43)@xmath117 ( 75,47)@xmath790 ( 75,20)@xmath647 ( 48,30)@xmath474 & & bir-ridge-ab4.eps ( -55,-30)(@xmath791 ) ( 23,46)@xmath83 ( -8,105)@xmath650 ( 67,103)@xmath779 ( 85,-8)@xmath648 ( 8,-4)@xmath649 ( 40,15)@xmath116 ( 2,50)@xmath117 ( 30,80)@xmath118 ( 65,50)@xmath780 + any subsequence @xmath760 of flops is directed with respect to a polarization on @xmath761 , and the flopping facets @xmath762 are strictly in a half - plane .
moreover , the facets @xmath792 , are in a half - plane and strictly exactly when @xmath663 are not coplanar . in general , the configuration of facets and countries is not symmetric , in particular , pairs of facets @xmath793 and @xmath663 are not coplanar
. however , two facets @xmath793 are coplanar if and only if @xmath794 is a generalized divisorial contraction , that is , for @xmath795 , @xmath796 is negative on @xmath797 , and @xmath798 is @xmath49-cartier . if @xmath799 are not coplanar and the span of @xmath800 intersects @xmath116 ( respectively @xmath647 ) , then @xmath801 ( respectively @xmath802 ) is a generalized extremal and directed log flop@xmath765 with respect to a polarization .
the facets @xmath663 are coplanar if and only if the flopping locus of @xmath803 does not contain @xmath804 .
this is possible only for @xmath805 .
moreover , the commuting , centrally symmetric , case is possible only for @xmath773 with coplanar pairs @xmath774 and @xmath775 ; the flopping locus is disjoint from @xmath120 .
an equation @xmath768 for the closure @xmath769 in @xmath806 , can be given by a curve @xmath771 on a rather high model @xmath177 of @xmath1 .
for closures @xmath807 respectively in @xmath703 , the equations @xmath808 are given by the contractible divisor @xmath120 .
this allows us to find coplanar facets in terms of linear extensions for functions @xmath809 from certain countries .
[ [ birational - c - or-3c ] ] birational c or 3c + + + + + + + + + + + + + + + + + + the weil - picard number @xmath810 , @xmath805 , the only divisorial facets are @xmath811 with elementary divisorial contractions @xmath812 @xmath813 of two prime divisors @xmath814 , respectively .
the other contractions @xmath815 are extremal divisorial with both exceptional divisors @xmath816 .
the contractions @xmath817 corresponding to @xmath818 are elementary small .
the pairs @xmath786 @xmath695 , @xmath819 are the only slt wlc models of @xmath13 . the models @xmath820 are @xmath34-factorial .
all other models @xmath821 , are not @xmath34-factorial .
there exists a sequence of elementary divisorial extractions , elementary log flops in the facets @xmath822 , and of elementary divisorial contractions : @xmath823 any subsequence @xmath760 of flops is directed with respect to a polarization on @xmath761 , and the flopping facets @xmath762 are strictly in a half - plane . moreover , the facets @xmath824 , and @xmath825 respectively , are in a half - plane and strictly exactly when @xmath774 and @xmath663 , respectively , are not coplanar . in general , the configuration of facets and countries is not symmetric , in particular , pairs of facets @xmath774 and @xmath663 are not coplanar .
however , pairs @xmath774 and @xmath663 are coplanar if and only if respectively @xmath826 and @xmath827 .
each case or both cases simultaneously can occur for @xmath828 .
however , for @xmath773 , both pairs are coplanar or none of the pairs are coplanar .
the former is possible only for the commuting centrally symmetric case with disjoint contractible divisors @xmath816 .
an equation @xmath768 for the closure @xmath769 in @xmath806 , can be given by a curve @xmath771 on a rather high model @xmath177 of @xmath1 .
for closures @xmath807 respectively in @xmath703 , the equations @xmath808 are given by the contractible divisor @xmath120 .
for closures @xmath829 respectively in @xmath830 the equations @xmath831 are given by the contractible divisor @xmath832 .
this allows to find coplanar facets in terms of linear extensions for functions @xmath833 from certain countries .
@xmath834\ar[d]&y_4\ar@{-->}[r ] & \cdots\ar@{-->}[r ] & y_{m-2}\ar@{-->}[r]&y_{m-1}\ar[d]\\ y_2\cong t_3 \ar[rrd ] & & & & y_m\cong t_m\ar[lld]\\ & & y_1\cong t_1\cong t_2\cong t & & } \ ] ] lll bir-ridge-c.eps ( 44,19)@xmath83 ( -8,-3)@xmath649 ( -2,60)@xmath650 ( 24,78)@xmath835 ( 39,65)@xmath836 ( 50,80)@xmath651 ( 81,69)@xmath653 ( 100,13)@xmath778 ( 55,5)@xmath116 ( 10,25)@xmath117 ( 77,35)@xmath647 & & bir-ridge-ab4.eps ( -45,-30)(@xmath773 , @xmath837 ) ( 25,46)@xmath83 ( -8,105)@xmath650 ( 67,103)@xmath779 ( 85,-8)@xmath648 ( 8,-4)@xmath649 ( 40,15)@xmath116 ( 3,50)@xmath117 ( 30,80)@xmath118 ( 65,50)@xmath780 [ thrm - ridge types]under the generality conditions on @xmath508 , let @xmath83 be a ridge of the geography @xmath46 .
then it is one of the above types : cube bordering , fibering , or birational , and behaves accordingly .
( see diagram [ dia9 ] below . )
ridges.eps ( 31,103)(1 ) ( 65,58)(3 ) ( 15,25)(2 ) ( 23,36)@xmath474 ( 35,96)@xmath474 ( 57,57)@xmath474 @xmath838 a bordering ridge @xmath83 of @xmath46 has the cube bordering or fibering type by definition , proposition [ prop - s = closed rational poly ] and corollary [ cor - geoface = cube for general ] . indeed , if @xmath839 , it is cube bordering .
otherwise , @xmath640 and @xmath83 is fibering .
cube bordering type : @xmath83 satisfies at least one equation @xmath71 or @xmath72 .
if @xmath83 satisfies another equation @xmath635 or @xmath72 , it lies in the ridge of @xmath33 given by these equations .
otherwise , @xmath840 for a facet @xmath56 of @xmath33 which satisfies the equation @xmath71 or @xmath72 . for the boundaries @xmath632 of the other type ridges , @xmath65 is internal : @xmath535 . as in the proof of theorem [ thrm - facet types ] , any wlc model @xmath12 of @xmath13 is klt with big @xmath391 and has a lc contraction @xmath513 .
if @xmath514 is projective , it is ft by lemma [ lemma - ft=0log ] with @xmath841 and with at most two extremal contractions@xmath765 by lemma [ lemma - rho of wlc / lc ] .
as we see from the following , @xmath638 holds and there exist two extremal contractions of @xmath237 over @xmath537 exactly when @xmath393 is a slt wlc model of @xmath13 .
in particular , @xmath548 .
such a model @xmath393 exists because @xmath151 and it satisfies the assumptions of lemma [ lemma - rho of wlc / lc ] .
the slt model can be constructed by the slt lmmp from the above initial model .
fibering type .
suppose that @xmath83 is fibering .
then by the open property of classes and the last statement in theorem [ mainthrm - geography ] , there exist required countries @xmath645 and facets @xmath800 .
most of stated properties follow from theorem [ thrm - facet types ] and standard facts .
the pair @xmath571 is a wlc model of @xmath13 by lemma
[ lemma - closed n ] , limiting from @xmath116 . by definition
it has a mori log fibration @xmath842 and @xmath843 .
thus @xmath844 . on the other hand ,
since there exists a slt wlc model of @xmath13 over @xmath575 , this model is @xmath571 and @xmath845 by lemma [ lemma - rho of wlc / lc ] .
similarly , @xmath681 is a slt wlc model of @xmath13 too .
moreover , all intermediate facets @xmath846 , are flopping and the models @xmath847 are the only slt wlc models of @xmath13 .
if @xmath679 , @xmath732 has two mori log fibrations @xmath848 , which are the only extremal contractions@xmath765 , and by corollary [ cor - small klt slt wlc models in elementary log flops ] , @xmath571 is the only wlc model of @xmath13 . in the last case , the facets @xmath689 are noncoplanar by corollary [ cor - facet = face ] . for @xmath724 ,
another extremal contraction @xmath849 is birational and flopping by corollary [ cor - small klt slt wlc models in elementary log flops ] . indeed ,
if it is divisorial , @xmath850 and @xmath851 has a mori log fibration @xmath852 , a contradiction .
similarly , all other intermediate facets are flopping .
now the description of all wlc models can be obtained from corollary [ cor - ft eff decomposition ] and theorem [ thrm - facet types ] .
again the pair @xmath571 is a wlc model of @xmath13 , limiting from @xmath116 . by definition
, it has a mori log fibration @xmath853 .
thus @xmath854 and @xmath855 is the only extremal contraction for @xmath732 . by theorem [ thrm - facet types ]
, @xmath649 is divisorial .
otherwise , @xmath575 has another extremal contraction to @xmath856 .
in particular , @xmath724 .
so , the pair @xmath735 is a wlc model of @xmath13 , limiting from @xmath117 , with an elementary divisorial contraction @xmath857 . again
@xmath858 and @xmath735 is a slt model of @xmath13 .
moreover , all intermediate facets @xmath859 , are flopping and the models @xmath734 are the only slt wlc models of @xmath13 . if @xmath692 , @xmath851 has exactly one more contraction@xmath765 , the mori log fibration @xmath860 , and by corollary [ cor - small klt slt wlc models in elementary log flops ] , @xmath735 is the only slt wlc model of @xmath13 .
for @xmath745 , another extremal contraction @xmath861 is birational and flopping by corollary [ cor - small klt slt wlc models in elementary log flops ] . indeed ,
if it is divisorial , @xmath862 and @xmath863 has a mori log fibration @xmath864 , a contradiction .
similarly , all other intermediate facets are flopping .
as above the description of all wlc models can be obtained from corollary [ cor - ft eff decomposition ] and theorem [ thrm - facet types ] .
type 2c . as in type 2b ,
@xmath649 is divisorial and @xmath735 is a slt model of @xmath13 with an elementary divisorial contraction @xmath857 .
similarly , @xmath653 is divisorial and @xmath865 is a slt model of @xmath13 with an elementary divisorial contraction @xmath866 and the subsequent mori log fibration @xmath867 . in particular ,
as above all intermediate facets @xmath868 , are flopping and the models @xmath869 are the only slt wlc models of @xmath13
. the description of all wlc models can be obtained from corollary [ cor - ft eff decomposition ] and theorem [ thrm - facet types ] .
note that flopping facets only possible for @xmath547 .
the properties of equations given by @xmath870 @xmath871 and @xmath872 @xmath873 follow from theorem [ thrm - facet types ] . moreover , if @xmath705 is a sufficiently general curve of a mori log fibration @xmath708 ( from a rather high model @xmath177 of @xmath1 ) , it is not contracted on all models @xmath874 . hence @xmath564 on @xmath706 , and @xmath315 on @xmath707 .
more precisely , @xmath875 on @xmath707 if and only if the birational image of @xmath705 belongs to a fiber of @xmath685 , that is , @xmath708 factors through the last fibration .
the similar holds for a general curve @xmath712 on @xmath685 .
thus the fibrations @xmath656 are square birational to each other if and only if @xmath686 are birational ( the same general curves in fibers ) , or @xmath713 for general @xmath704 .
moreover , @xmath656 are generically isomorphic if and only if the facets @xmath654 are coplanar . indeed ,
if the isomorphism holds then @xmath876 and the natural modification @xmath710 does not touch the general fiber .
thus @xmath877 is linear on @xmath46 near @xmath83 and @xmath654 are coplanar .
conversely , if @xmath654 are coplanar , the convexity of proposition [ prop - p - e ] implies that @xmath878 and @xmath877 is linear on @xmath46 near @xmath83 .
thus the modifications @xmath879 do not touch @xmath705 and @xmath656 are generically isomorphic .
birational type .
suppose that @xmath83 is birational .
then by the open property of classes and the last statement in theorem [ mainthrm - geography ] , there exist required countries @xmath645 and facets @xmath69 .
most of stated properties follow from theorem [ thrm - facet types ] and standard facts .
we discuss the others below . in this case as for internal facets ,
all wlc models @xmath12 are klt with @xmath391 , but now of general type with the lc model @xmath587 by our assumptions and proposition [ prop - kd nd are constants ] .
the facets @xmath748 are in a half - plane by convexity of @xmath645 and strictly by corollary [ cor - facet = face ] . in particular , @xmath745 . type 3a .
since @xmath756 and by lemma [ lemma - rho of wlc / lc ] , for a slt wlc model @xmath393 of @xmath13 , the contraction @xmath514 is small with @xmath638 and two extremal contractions@xmath765 are also elementary small . as above , for some @xmath750 , @xmath880 , limiting from a country
@xmath645 with two flopping facets @xmath748 , corresponding to two extremal contractions@xmath765 by theorem [ thrm - facet types ] .
the rest also follows from the classification of facets in the theorem and corollaries [ cor - ft eff decomposition ] , [ cor - small klt slt wlc models in elementary log flops ] .
note also , that if two facets @xmath763 are coplanar , then @xmath764 is a generalized extremal and directed log flop@xmath765 with respect to a polarization .
indeed , if the facets are coplanar , the one dimensional subspace of @xmath49-cartier divisors in @xmath881 is generated by @xmath882 or @xmath883 , and respectively @xmath884 is ample or antiample@xmath765 .
the converse follows from definition and the description of all projective rational @xmath72-contractions of @xmath514 .
similarly , one can treat the case with the span of @xmath69 intersecting @xmath766 .
an equation @xmath768 for a facet @xmath69 in @xmath806 can be given by a curve @xmath771 on a rather high model @xmath177 of @xmath1 which is contracted on @xmath660 but not on @xmath885 , respectively .
the commuting case can be treated with the convexity of proposition [ prop - p - e ] .
since @xmath781 , in particular , @xmath537 is not @xmath34-factorial , and by lemma [ lemma - rho of wlc / lc ] , for a slt wlc model @xmath393 of @xmath13 , the contraction @xmath514 is not small with @xmath638 but two extremal contractions@xmath765 can be small .
however , applying the @xmath120-lmmp for a prime divisor @xmath120 of @xmath237 exceptional on @xmath537 , we obtain a new slt wlc model of @xmath886 , say , @xmath735 , limiting from @xmath117 and with an elementary divisorial contraction @xmath857 of @xmath120 .
equivalently , @xmath649 is divisorial .
thus @xmath575 is @xmath34-factorial and @xmath571 is a nonslt model of @xmath13 , limiting from @xmath116 .
the second facet @xmath648 of @xmath116 is flopping by theorem [ thrm - facet types ] , because it corresponds to a small extremal contraction @xmath732 .
in particular , @xmath120 is the only prime divisor of @xmath237 exceptional on @xmath537 .
the log flop in @xmath648 gives another @xmath34-factorial nonslt model of @xmath13 : @xmath681 , limiting from @xmath647 .
again by theorem [ thrm - facet types ] , the facet @xmath653 is divisorial with an elementary divisorial extraction @xmath887 of @xmath120 .
moreover , @xmath663 are the only divisorial facets , e.g. , by corollary [ cor - small klt slt wlc models in elementary log flops ] .
the rest follows from the classification of facets and corollaries [ cor - ft eff decomposition ] , [ cor - small klt slt wlc models in elementary log flops ] . by construction and definition ,
@xmath717 on @xmath888 and @xmath199 on the other countries and facets .
thus by the convexity of @xmath889 , the facets @xmath792 , are in a half - plane and strictly exactly when @xmath663 are not coplanar .
the facets @xmath663 are coplanar if and only if the function @xmath889 is linear on the half - plane of @xmath648 , or equivalently , the flop in @xmath648 does not touch @xmath804 generically .
this is possible only for @xmath805 by corollary [ cor - facet = face ] . by construction facets @xmath793
are coplanar if and only if @xmath890 is a generalized divisorial contraction of @xmath120 .
otherwise , @xmath801 or @xmath891 is a generalized extremal and directed log flop@xmath765 with respect to a polarization ( * ? ? ?
* proof - explanation of corollary 1.9 ) .
note that the contraction @xmath797 is not small , but extremal , and @xmath796 is negative on @xmath797 for @xmath892 when @xmath893 is @xmath49-cartier , or equivalently @xmath153 is coplanar with @xmath800 .
in this situation , for @xmath600 ( respectively @xmath894 ) , a generalized log flip with respect to @xmath153 or a generalized directed log flop of @xmath895 with the polarization @xmath896 ( respectively @xmath897 is @xmath801 ( respectively @xmath802 ) ; it is neither a divisorial contraction or a small flip / flop . for @xmath795 , a log flip or
a log flop of @xmath797 is a generalized divisorial contraction @xmath898 .
the commuting case can be treated as above .
since @xmath810 , @xmath537 is @xmath34-factorial . as in type
2c , we can construct a slt wlc model of @xmath13 , say , @xmath899 with @xmath900 , limiting from @xmath118 and with an elementary divisorial contraction @xmath901 of a prime divisor @xmath832 .
thus @xmath650 is divisorial , @xmath902 is @xmath34-factorial and @xmath735 is a limiting from @xmath117 nonslt model of @xmath13 .
however , now by the @xmath34-factorial property of @xmath537 , the second facet @xmath649 of @xmath117 is also divisorial with an elementary divisorial contraction @xmath903 of a prime divisor @xmath120 .
again by theorem [ thrm - facet types ] the facets @xmath904 are divisorial with elementary divisorial contractions @xmath905 of divisors @xmath816 , respectively . in particular , @xmath805 .
all other facets are flopping .
the rest follows from the classification of facets and corollaries [ cor - ft eff decomposition ] , [ cor - small klt slt wlc models in elementary log flops ] . by construction and definition ,
@xmath717 on @xmath888 and @xmath75 on the other countries and facets .
thus by the convexity of @xmath889 , the facets @xmath792 , are in a half - plane and strictly exactly when @xmath663 are not coplanar .
the facets @xmath663 are coplanar if and only if the function @xmath889 is linear on the half - plane of @xmath648 , or equivalently , the extraction @xmath832 in @xmath648 does not touch @xmath804 generically , that is , @xmath827 .
similarly , the facets @xmath906 , are in a half - plane and strictly exactly when @xmath774 are not coplanar .
the facets @xmath774 are coplanar exactly when @xmath907 . if @xmath816 are disjoint , then @xmath773 and the commuting case holds with coplanar pairs @xmath774 and @xmath775 .
conversely , if @xmath773 , then @xmath908 has two extremal divisorial contractions@xmath765 of @xmath816 . if @xmath909 , the commuting case holds with the coplanar pairs .
otherwise , the extremal property implies that @xmath910 and both pairs @xmath774 and @xmath775 are not coplanar .
[ [ central - model ] ] central model + + + + + + + + + + + + + let @xmath83 be a fibering or birational ridge of a geography under the generality conditions , @xmath632 be a boundary with a decomposition @xmath911 , where @xmath912 and @xmath913 is big and @xmath49-mobile over the lc model @xmath537 of @xmath13 .
such a decomposition always exists and , in applications , the decomposition mob+exc is typical .
a klt slt wlc model @xmath393 of @xmath13 is _ central _ with respect to the decomposition @xmath911 if @xmath914 is a weak log fano ( * ? ? ?
* definition 2.5 ) .
since @xmath915 , @xmath237 depends only on @xmath83 when @xmath916 ( and @xmath917 ) for the mob+exc decomposition of @xmath519 .
theorem [ thrm - ridge types ] implies the following properties of a central model @xmath393 : 1 .
@xmath514 is a @xmath34-factorial ft contraction with @xmath638 .
2 . @xmath918 , and all other slt wlc models @xmath919 can be reconstructed from @xmath393 by elementary log flops in both directions : @xmath920 @xmath921 and @xmath922 are sequences of a log flop and log flips with respect to @xmath923 ( the modifications are optional ) , with at most one log flop for both chains . in particular , this restores all other divisorial and fibering contractions over @xmath537 : @xmath924 @xmath925 and all other wlc models of @xmath13 .
thus a modification ( link ) of @xmath575 ( respectively @xmath926 ) into @xmath927 ( respectively @xmath928 ) is factored into log antiflips , a log flop ( optional ) and log flips with respect to @xmath929 .
the pair @xmath914 is a log fano exactly when there are no flops ; flops are possible only in dimension @xmath547 .
a prime divisor @xmath31 of @xmath237 is contractible@xmath765 by @xmath930 if and only if there exists a @xmath34-factorial wlc model @xmath404 of @xmath13 where @xmath31 is exceptional on @xmath426 and with @xmath931 . in particular , there are no such divisors if , for any @xmath34-factorial wlc model @xmath404 of @xmath13 , @xmath932 is terminal ( the last condition can be extended for non @xmath34-cartier @xmath933 , assuming that @xmath932 is terminal when it is terminal for a @xmath34-factorialization ) .
4 . if @xmath393 is @xmath934-lc ( in codimension @xmath935 )
, then @xmath936 is also @xmath934-lc ( respectively , in codimension @xmath935 ) .
for example , this holds if @xmath282 is @xmath934-lc ( respectively , in codimension @xmath935 , @xmath937 and the components of @xmath65 are nonexceptional on @xmath237 ) .
[ cor - central model exists]let @xmath83 be a fibering or birational ridge in @xmath46 under the generality assumptions and @xmath632 .
then for any decomposition @xmath911 as above , @xmath13 has a central model @xmath393 .
for such a decomposition , the model @xmath393 is unique up to a log flop of @xmath914 over @xmath537 .
conversely , any klt weak log fano @xmath914 with the property ( i ) in central model above corresponds to such a ridge . to construct a central model one can use the @xmath938-mmp@xmath765 with any slt wlc model as an initial one . since @xmath938 is @xmath49-mobile , the modifications are small .
note also that @xmath939 by definition of @xmath537 : @xmath940 .
thus a required model is defined up to a log flop of @xmath914 .
conversely , take @xmath941 , @xmath942 where @xmath913 and @xmath943 ( @xmath49-complement ) .
the generality conditions can be satisfied near @xmath65 according to the big and the semiample property of @xmath305 .
note that , for @xmath294 , @xmath944 is possibly non slt .
thus to apply theorem [ thrm - ridge types ] , we can consider its local version near @xmath65 or we can replace @xmath944 by an appropriate log resolution @xmath945 .
the latter uses the klt property of @xmath946 : for some @xmath947 , @xmath948 and @xmath949 is a slt wlc model of @xmath950 corresponding to a ridge .
the following finiteness is a key to birational factorizations .
[ cor - central models bounded]let @xmath934 be a positive real number .
suppose that ( * ? ? ?
* conjecture 1.1 ) holds for @xmath934-lc varieties of dimension @xmath951 .
then the family of central models @xmath514 and thus their links in dimension @xmath951 for @xmath934-lc slt models @xmath393 are bounded over @xmath952 .
immediate by corollary [ cor - central model exists ] and by conjecture 1.1 in @xcite . for links ,
see section [ section6-linkage ] below .
[ example - d=2 links]all possible central models with links in dimension @xmath953 with @xmath954 , e.g. , @xmath955 , and terminal singularities i.e. , nonsingular , ( see links in mori category in section [ section6-linkage ] below ) are as follows : 2a : @xmath956 , a nonsingular quadric ; 2b : @xmath957 , a blowup of the plane @xmath958 ; 2c : @xmath959 , the standard fiber modification , where @xmath960 are @xmath961-fibrations over a nonsingular curve @xmath178 and @xmath962 is a blowup of @xmath963 with blowdown of another curve to @xmath964 .
note that the last model is @xmath965 .
there are much more such links @xmath966 and central models in dimension @xmath424 in the nonsingular case , and thousands are expected for terminal singularities .
[ [ links ] ] links + + + + + let @xmath967 be two extremal contractions ( possibly birational ) of @xmath34-factorial projective@xmath968 birationally equivalent varieties @xmath563 . their _ link _ @xmath969 is a composition ( chain ) of extremal divisorial extractions , contractions and small modifications ( e.g. , flips , flops or antiflips ) over another variety @xmath537 with contractions @xmath970 : @xmath709\ar@{-->}[r]&\cdots\ar@{-->}[r ] & y_2\ar[d]\\ t_1\ar[rd ] & & t_2\ar[ld]\\ & t. & } \ ] ] such a link from @xmath708 into @xmath971 is _ elementary _ if @xmath563 and all intermediate models @xmath514 in the link are @xmath34-factorial ft with @xmath841 , and modifications are elementary in the following order : ( at most 2 ) divisorial extractions , small modifications , ( at most 2 ) divisorial contractions ; all modifications are optional .
the sequences of modifications of fibering 2a - c and birational 3a - c ridges are elementary .
if , in addition , @xmath967 are mori fibrations and all the models in the link are terminal , it is one of sarkisov links of types i , ii , iii , or iv @xcite .
if @xmath972 are two wlc models of @xmath13 for @xmath65 in a fibering ridge @xmath83 connected by generalized log flops as in types 2a - c , then these log flops give an elementary link from @xmath708 into @xmath685 by theorem [ thrm - ridge types ] .
we can take also their inverses but the only new will be for nonsymmetric type 2b .
they are sarkisov links when all varieties @xmath751 are terminal .
those links will be referred to as _ cte _ ( _ centralized terminal elementary _ ) if they have a terminal _
model @xmath514 with @xmath973 , in particular , @xmath974 is big and @xmath49-mobile itself@xmath765 on any wlc model @xmath393 of @xmath13 . in addition , @xmath237 should be terminal for all @xmath34-factorial wlc models @xmath393 of @xmath13 .
e.g. , this is true if a slt model @xmath694 in the link is terminal and @xmath975 or @xmath976 is big@xmath765 on any divisor of @xmath751 ( cf . the mobility below ) .
conversely , the sarkisov links are cte .
the terminal and elementary properties follow from definition .
the central property also follows from definition .
more precisely , if the first small modification is a flip or a flop , @xmath732 is a central model .
otherwise , a central model @xmath668 is a result of maximal chain of antiflips ; if the next modification is a flop , @xmath668 is a weak fano fibration , otherwise , @xmath668 is a fano fibration . in total , the link is factored into a divisorial extraction , antiflips , a flop , flips and a divisorial contraction ( all factors are optional ; cf .
( ii ) in central model ) .
flops are possible only in dimension @xmath547 .
the contractions @xmath514 in a cte or sarkisov link satisfy the _ mobility@xmath765 _ : @xmath977 is @xmath49-mobile and big@xmath765 in codimension @xmath72 .
the latter means that @xmath977 is big@xmath765 and big@xmath765 on any divisor of @xmath237 .
thus @xmath978 for @xmath979 , @xmath977 is nonzero effective@xmath765 modulo @xmath276 .
the boundedness of corollary [ cor - central models bounded ] with @xmath980 holds also for the sarkisov links ( cf .
* proposition 6 ) ) .
[ cor - links / pt bounded in 3](global ) cte links with @xmath981 are bounded in dimension @xmath982 .
immediate by corollary [ cor - central models bounded ] with @xmath980 and @xcite . [
[ mori - category ] ] mori category + + + + + + + + + + + + + let @xmath1 be a variety .
its mori category is the category of minimal resulting models of @xmath1 with birational transformations as morphisms .
so , an object is a projective terminal @xmath34-factorial model @xmath11 of @xmath1 with nef @xmath983 , a minimal model , or a projective terminal @xmath34-factorial model @xmath11 of @xmath1 equipped with a mori fibration @xmath541 .
a factorization of morphisms in the case of minimal models is known ( see corollary [ cor - small klt slt wlc models in elementary log flops ] above ) .
a factorization between mori fibrations was suggested by sarkisov .
[ thrm - mori morphism is cte]a morphism between mori fibrations can be factorized into cte links , or equivalently , into sarkisov links .
however , this is not a factorization of the genuine sarkisov program : the sarkisov s invariants may not decrease .
so , this is not a solution to the sarkisov program
. follows from a linkage ( * ? ? ?
* example 2.11 ) after a perturbation . for simplicity ,
we assume the absolute case : @xmath984 , and omit the notation @xmath416 below .
we suppose also that the dimension @xmath985 .
so , let @xmath986_g\ar@{-->}[r ] & y'\ar[d]^{g ' } \\
t&t ' } \ ] ] be a birational transformation for mori fibrations @xmath987 .
we can convert each mori fibration @xmath514 into a polarized one @xmath518 with an ample boundary @xmath519 .
take a sufficiently ( very ) ample divisor @xmath369 on @xmath537 and put @xmath988 where @xmath31 is a generic reduced irreducible divisor in a linear system @xmath989 @xmath990 , and @xmath991 . by construction @xmath628
is a terminal ( in codimension @xmath612 ) slt model with the lc contraction @xmath429 , the mori fibration .
similarly , @xmath992 has a polarization @xmath993 with @xmath994 .
we can assume also that @xmath995 have a common log resolution @xmath89 on which ( the birational transformations of ) @xmath619 are big , mobile ( even free ) and disjoint . according to ( * ? ? ?
* example 2.11 ) and corollary [ cor - small klt slt wlc models in elementary log flops ] ( see also factorization for a birational transformation of wlc models in section [ section4-finiteness ] ) , there exists a factorization of @xmath996 into links .
the intermediate models @xmath997 in the factorization are slt wlc models of @xmath998 for @xmath999 , a class ( an open interval ) in the separatrix @xmath1000 .
( since each @xmath976 is @xmath49-mobile , we identify @xmath976 with its birational transform @xmath1001 on @xmath751 by writing the same @xmath976 . )
the union @xmath1002 gives a path from @xmath65 to @xmath153 in @xmath456 . by construction
each @xmath751 has only terminal singularities and has a mori fibration @xmath1003 ( possibly after a log flop of @xmath997 ) by corollary [ cor - geoface = cube for general ] .
we can suppose that @xmath1004 for the first interval @xmath1005 and @xmath1006 for the last interval @xmath1007 .
see diagram [ dia10 ] below .
a link from @xmath1003 into @xmath1008 corresponds to the point @xmath1009 of intersection @xmath1010 . note that @xmath1011 are wlc models of @xmath1012 by lemma [ lemma - closed n ] .
the link may not be elementary , in particular , @xmath660 is not necessarily the lc model of @xmath997 , and the picard ranks for intermediate models @xmath1013 over the lc model of @xmath1012 can be rather high .
ccc nonpert-path.eps ( 60,26)@xmath153 ( 60.5,30)@xmath474 ( 7,59)@xmath1005 ( 32,35)@xmath1014 ( -2,87)@xmath65 ( 22,48)@xmath1015 ( 21,44.8)@xmath474 ( 82,50)@xmath1016 ( 75,56)perturbation & & perturbed-path2.eps ( 25,86)@xmath65 ( 84,26)@xmath153 ( 5,-4)@xmath1017 ( 35,78)@xmath1018 ( 40,49)@xmath1019 ( 48,34)@xmath1020 ( 73,37)@xmath1021 ( 35,39)@xmath1014 ( 29,55)@xmath1005 ( 56,31)@xmath1022 ( 31,76)@xmath474 ( 38.5,45)@xmath474 ( 45.5,38.5)@xmath474 ( 74,31)@xmath474 to attain cte links we use a perturbation and theorem [ thrm - ridge types ] . to fulfil the generality conditions for @xmath945
, we add to @xmath1023 generic ( very ample ) divisors @xmath3 so that the conditions hold for a suitable extension @xmath1024 .
note that after a perturbation our new subclasses @xmath1025 and @xmath1009 are embedded into @xmath1026 .
( if @xmath1027 , the subclasses are cube bordering : @xmath1028 . )
now we perturb the whole path @xmath1029 into a path in @xmath456 away from @xmath1030 .
we perturb the path in such a way that it goes only through internal points of fibering facets and ridges .
( this path can be straight under the projection of proposition [ prop - image of separatrix ] . )
thus now we suppose that new intervals @xmath1025 are in fibering facets @xmath69 and new vertexes @xmath1009 in fibering ridges @xmath1031 ( we use the same notations for the perturbed path even though there is no @xmath72-to-@xmath72 correspondence for intervals and vertexes under the perturbation ) . by construction @xmath1032
are the mori log fibrations of @xmath1033 , @xmath1034 are terminal and @xmath751 is terminal for other mori log fibrations of facets @xmath69 . indeed , for the perturbation @xmath1035 of @xmath65 , the pair @xmath1036 has the same contraction @xmath514 because the contraction is extremal ft . the same works for @xmath426 : for the perturbation @xmath1037 in the facet @xmath762 close to @xmath153 , a slt model @xmath1038 is terminal , and so is @xmath1039 .
the terminal property is also stable under perturbation .
each slt wlc model @xmath1040 of @xmath1041 for a ( klt ) boundary @xmath1042 , is terminal because @xmath1041 is terminal and @xmath1043 is non contractible .
( this model may have cn singularities on the lc model . ) similarly , the terminal property holds for intermediate @xmath751 .
moreover , the same holds for any intermediate @xmath34-factorial model of a link from @xmath1003 into @xmath1008 by theorem [ thrm - facet types ] .
indeed , if we have the lc model @xmath1040 for @xmath1043 in a divisorial facet of @xmath46 , with a contracted divisor @xmath306 of the facet , then for pair @xmath1040 , the only cn singularity is @xmath1044 .
but by our assumptions @xmath1045 contains the center , and @xmath488 is terminal .
since @xmath1046 is big and @xmath49-mobile , the link is cte by corollary [ cor - central model exists ] and a required factorization is @xmath1047 .
[ [ polarized - linkage ] ] polarized linkage + + + + + + + + + + + + + + + + + so , we have constructed a chain of links with _
polarizations _ as follows .
there exist a model @xmath945 of @xmath237 , where @xmath1048 is a reduced divisor with only big and mobile prime components , with components of @xmath1049 generating each @xmath1050 , and intervals @xmath1051 such that : all boundaries in the interval @xmath1025 are @xmath35 equivalent with terminal klt wlc models @xmath1052 , of @xmath998 ; actually @xmath976 on @xmath751 is the birational transform of @xmath976 from @xmath89 ; @xmath1053 is the lc contraction of @xmath997 , that is , @xmath1054 is a pull back of an ample divisor from @xmath660 ; and each vertex @xmath1009 gives a cte or sarkisov link between limiting models @xmath1011 ; they can be cn and with lc models @xmath1055 over which the link goes : for @xmath1009 , 2a : @xmath1056 and each @xmath1013 have at most one another elementary small and numerically trivial contractions@xmath1057 with respect to @xmath1058 ; 2b : @xmath1059 , an elementary divisorial extraction @xmath1060 corresponds to the unique _ maximal _
center @xmath1061 ( exceptional divisor ) with cn singularity : @xmath1062 with center @xmath1063 ( a small linear prolongation in direction @xmath1064 gives a unique noncn center ) , and @xmath1065 has at most one another elementary small and numerically trivial contraction@xmath1057 with respect to @xmath1066 ; the same can hold after interchanging @xmath750 and @xmath1067 ; 2c : @xmath1068 with elementary divisorial extractions @xmath1069 as in 2b ; each of those links can be centralized ( see links above ) . a limit to boundaries in @xmath1070 supported in @xmath1023 ( the original path ) gives similar conditions with cn singularities of pairs instead of terminal and semiampleness instead of ampleness on @xmath660 and @xmath1055 .
[ [ rigidity ] ] rigidity + + + + + + + + a mori fibration @xmath514 is _ rigid _ if , for any other mori fibration @xmath992 , any birational isomorphism @xmath405 gives an isomorphism of the fibrations @xmath987 .
the definition and the definitions below work over any @xmath1071 .
however , if @xmath1072 the rigidity is only possible when @xmath1073 .
thus , for @xmath1074 , it is reasonable to use a weaker version : @xmath514 is birationally rigid ( see below ) and any mori fibration @xmath1075 satisfies the above property .
there are other weaker versions of the rigidity .
if it holds over the general points of @xmath1076 we say that @xmath514 is _ generically _ rigid .
for example , so do the conic bundles satisfying sarkisov s criterion below .
a mori fibration @xmath514 ( or over the general point of @xmath537 ) is _ birationally _ rigid if , for any other mori fibration @xmath992 , any birational isomorphism @xmath405 gives a birational isomorphism of the fibrations @xmath514 and @xmath992 , that is , they are square birational under the isomorphism .
the fibration is _ weakly _ rigid if , in addition , any mori fibration @xmath992 square birational to @xmath514 is isomorphic to @xmath514 over the general point of @xmath537 ( possibly under a different birational isomorphism @xmath405 ) .
such an isomorphism allows us to treat the birational isomorphism @xmath405 as a birational automorphism of @xmath514 and vice versa .
del pezzo fibrations of degree @xmath953 and @xmath424 give nontrivial examples of the last rigidity ( see corollary [ cor - delpezzo123rigid ] below ) .
in general , it is expected that a mori fibration @xmath514 is weakly rigid if its general fiber is weakly rigid and @xmath514 is sufficiently twisted , e.g. , the discriminant locus is large or the map of @xmath537 into moduli of fibers has a large degree . moreover , in higher dimensions , a rather general mori fibration @xmath514 is usually rigid@xmath1077 if its general fiber is rigid .
we will treat this with more details and applications elsewhere .
here we only illustrate this in some special cases usually assuming @xmath1072 .
[ cor - conic rigid threshold]suppose ( * ? ? ?
* conjecture 1.1 ) for terminal weak fano varieties of dimension @xmath1078 . in dimension @xmath951 , there exists a rational number @xmath1079 ( _ the birational rigidity threshold for conic bundles of dimension @xmath951 _ ) satisfying the following .
a mori conic bundle @xmath1080 with @xmath1081 is generically rigid if , for some real number @xmath1082 , @xmath1083 is cn and @xmath1084 is pseudo - effective where @xmath178 is the discriminant locus of the conic bundle @xmath1080 .
moreover , for @xmath1085 , we can omit ( * ? ? ? * conjecture 1.1 ) . for @xmath1086 ,
the corollary becomes the sarkisov theorem @xcite @xcite .
for any @xmath1087 with cn @xmath1083 , the condition @xmath1084 is pseudo - effective is equivalent to the condition @xmath1088 by conjecture [ conj - semiample ] . according to zagorskii @xcite in dimension @xmath424 and to sarkisov ( * ? ? ? *
theorem 1.13 ) in higher dimensions , any conic bundle @xmath1080 has a mori conic bundle model @xmath514 ( square birational to @xmath1089 with cn @xmath1083 for any @xmath1090 .
actually , in the corollary one can suppose that @xmath1080 is a morphism which is a mori or standard conic bundle over codimension @xmath72 .
such a weak standard model with cn @xmath1083 can be easily constructed for any conic bundle and any @xmath1091 .
suppose that @xmath1092 is another mori fibration , @xmath89 and @xmath1093 are birational , but not square birational .
note that square birational links preserve all conditions except for the cn property @xmath1083 .
however , if @xmath1094 is pseudo - effective on such a cn model , it is true on any model with the birational transform of @xmath178 plus any exceptional divisors ( cf .
* lemma 2.2 ) ( * ? ? ?
* lemma 4 ) ) . now consider a cte link from @xmath1080 into @xmath1092 which is not square birational , that is , over some @xmath488 with @xmath1095 , there exists a fibering contraction @xmath1096 . according to the big and @xmath49-mobile property of @xmath1097 , @xmath1098 .
hence , for @xmath1099 , by the key formula @xmath1100 ( * ? ? ?
* lemma 2 ) , where @xmath1101 , a contradiction .
now by corollary [ cor - central models bounded ] , the ft varieties @xmath1102 and thus @xmath1103 are bounded generically@xmath1104 .
note that @xmath1105 and they are mori log fibrations@xmath1104 with @xmath1106 .
this implies our improvement for @xmath1107 for all weil reduced divisors @xmath178 on @xmath537 and assuming @xmath1108 .
actually , @xmath1109 .
indeed , the cn property of @xmath1083 and the pseudo - effective property of @xmath1084 for any @xmath1082 implies that of @xmath1110 and of @xmath1111 for @xmath1112 .
[ cor - delpezzo123rigid ] let @xmath1113 be a @xmath424-fold mori fibration of degree @xmath1114 over a curve @xmath178 : @xmath1115 , @xmath178 is a nonsingular curve ( not necessarily complete ) , @xmath1113 is a mori fibration whose general fibers are del pezzo surfaces of degree @xmath951 .
then @xmath1113 is weakly rigid if one of the following holds : 1 .
@xmath1113 has sufficiently many singular fibers over a nonsingular completion @xmath1116 of @xmath178 , that is , there exists a natural number @xmath1117 such that the number of nonsmoothable fibers on the completion of @xmath178 is @xmath1118 ; 2 .
the map @xmath1119 , @xmath1120 the class of the smooth geometric fiber @xmath1121 of @xmath1122 , has a sufficiently large degree , where @xmath1123 is a coarse moduli space of smooth del pezzo surfaces of degree @xmath951 ; 3 . for each mori model @xmath1124 of @xmath1113 over the completion @xmath1116 , @xmath1125 holds ; or 4 . for each mori model @xmath1124 of @xmath1113 ,
@xmath1126 holds , where @xmath1127 denotes the cone of curves on @xmath237 . moreover , for degree @xmath72 , @xmath1113 is generically rigid and , for degree @xmath1128 , any birational automorphism of @xmath1113 is a composition of ( bertini and geiser ) involutions and of isomorphisms over the general point of @xmath178 .
the number in ( i ) is the sum of degrees of points in @xmath1116 , the fibers over which are not smoothable .
the condition ( ii ) is geometrical and , according to the proof below , ( i ) can be treated geometrically too . the condition in ( iii ) for @xmath1124
was introduced by grinenko @xcite and is known as the @xmath1129-_condition _
* conjecture 1.2 ( ii ) ) ) .
for a mori fibration @xmath1130 over a nonsingular projective curve @xmath178 , it is equivalent to the inclusion : @xmath1131\oplus g^*\operatorname{eff}(c)\ ] ] where @xmath1132 and @xmath1133 .
the condition in ( iv ) for @xmath1124 was introduced by pukhlikov @xcite and is known as the @xmath1134-_condition_. the @xmath1134-condition implies the @xmath1129-condition . after a completion we can suppose that @xmath89 and @xmath1135 are projective . by the weak rigidity of minimal del pezzo surfaces of degree @xmath678 ( * ? ? ?
* proposition 3.3 ) , each link over @xmath1136 is square birational and preserves all conditions ( i - iv ) . for @xmath691 , the links do not change @xmath1113 generically over @xmath178 . for @xmath1137 , the only nontrivial fiberwise links are ( bertini and geiser ) involutions and they give the same fibration generically .
but the ( global ) links@xmath1138 are bounded by corollary [ cor - links / pt bounded in 3 ] that contradicts ( i ) and ( ii ) with sufficiently large constants .
the property ( ii ) is geometrical because the degree is geometrical and , actually , we need to consider moduli over an algebraically closed field .
the property ( i ) is geometrical too because the smoothness is geometrical due to park : the smooth compactification is unique for @xmath1139 ( even @xmath1140 ) @xcite . by the mobility of cte links ,
there are no links of @xmath1130 under the conditions ( iii - iv ) .
more precisely , the @xmath1134-condition implies the @xmath1129-condition and the latter contradicts the mobility .
indeed , the link has type 2b or 2c .
let @xmath426 be its central model and @xmath1141 be an anticanonical model of @xmath426 with a small contraction @xmath1142 .
thus there exists a ( natural ) small modification @xmath1143 into a fano @xmath424-fold @xmath1141 .
this gives a decomposition @xmath1144 , that contradicts the @xmath1129-condition , where @xmath305 is a ( big ) @xmath34-mobile divisor , @xmath625 is a fiber of @xmath1130 , and @xmath1145 is a positive rational number .
note that @xmath1146 is an ample divisor on @xmath1141 and @xmath1147 is a weil divisor on @xmath1141 .
[ cor - dp1nonsing rigid = k ] let @xmath1113 be a @xmath424-fold mori fibration of degree @xmath72 over a nonsingular projective curve @xmath178 .
then @xmath1113 is generically rigid if and only if , for each mori model @xmath1130 of @xmath1113 , the @xmath1129-condition holds .
this is a weak version of the @xmath1129-conjecture ( * ? ? ?
* conjecture 4.4 ) for del pezzo fibrations of degree @xmath72 .
immediate by corollary [ cor - delpezzo123rigid ] ( iii ) and [ grin2 , theorem 4.5 ] .
it is not difficult to construct many examples of weakly rigid @xmath424-fold mori fibrations satisfying corollary [ cor - delpezzo123rigid ] ( i - ii ) .
those examples and corollary [ cor - dp1nonsing rigid = k ] allow us to construct many rigid mori fibrations of degree @xmath72 satisfying corollary [ cor - delpezzo123rigid ] ( iii ) .
examples of such fibrations satisfying corollary [ cor - delpezzo123rigid ] ( iv ) for degree @xmath72 or ( iii - iv ) for degrees @xmath1128 are unknown at present , but they are expected ( * ? ? ?
* problem 5.9 ( 3 ) ) .
[ [ questions ] ] questions + + + + + + + + + for what degrees @xmath951 , does the @xmath1129- , or @xmath1134-condition hold , or the mobility not hold for almost all del pezzo fibrations with terminal singularities ? equivalently , for what degrees @xmath951 , does the condition @xmath1148 , or the mobility is _ bounded _
, that is , the del pezzo fibrations @xmath1113 of degree @xmath951 with given property and only with terminal singularities form a bounded family ? the same questions are especially important for mori fibrations . for example , the condition @xmath1149 is unbounded for cubic mori fibrations ( * ? ? ?
* lemma 3.6 ) .
this and corollary [ cor - delpezzo123rigid ] give an unbounded family of terminal gorestein weakly rigid cubic mori fibrations@xmath1150 without the @xmath1134-condition . on the other hand , all nonsingular mori fibrations with @xmath1129-condition of degrees @xmath1151 and rather general of degree @xmath424
are weakly rigid ( * ? ? ?
* theorems 5.6 , 6.4 and 7.1 ) . some of them does not satisfy the @xmath1134-condition . however , it is unknown whether the condition ( iii ) of corollary [ cor - delpezzo123rigid ] holds for some of them .
the proofs of these rigidity results use the maximal singularity method and some elements of the sarkisov program @xcite
@xcite @xcite .
probably , the latter can be replaced by polarized linkage .
maximal singularities play a crucial role in birational geometry .
anyway , our approach improves the maximal singularity method .
[ thrm - nonrigid mori ] let @xmath1080 be a mori fibration .
the following properties are equivalent : 1 .
@xmath1080 is nonrigid@xmath1077 ; 2 .
there exists a mobile@xmath765 linear system @xmath1152 of degree ( possibly a rational number ) @xmath1153 which has a single exceptional maximal b - divisor ; 3 .
there exists a mobile@xmath765 linear system @xmath1152 of degree @xmath1153 which has an exceptional maximal b - divisor ; 4 .
there exists a big mobile@xmath765 linear system @xmath1152 of degree @xmath1153 which has a single exceptional cn b - divisor and is big@xmath765 on all its cn b - divisors ; 5 .
there exists a big mobile@xmath765 linear system @xmath1152 of degree @xmath1153 which has an exceptional cn b - divisor and is big@xmath765 on all its cn b - divisors ; 6 .
@xmath1080 has a model @xmath514 such that @xmath514 is a weak projective terminal @xmath34-factorial fano variety with @xmath638 and with big@xmath765 @xmath977 in codimension @xmath72 ; 7 .
@xmath1080 has a model @xmath514 such that @xmath514 is a terminal fano variety with @xmath1154 ; 8 .
@xmath1080 has a model @xmath514 such that @xmath514 is a weak projective terminal @xmath34-factorial fano variety with @xmath844 and with big@xmath765 @xmath977 in codimension @xmath72 ; and 9 .
@xmath1080 has a model @xmath514 such that @xmath514 is a terminal fano variety with @xmath1155 . in the statement and below @xmath1156
means that , for any weil divisor @xmath1157 , @xmath1158 , that is , @xmath1159 .
in other words , @xmath1160 denotes an integral weil divisor @xmath31 such that @xmath1161 .
( of course , for the integral numbers @xmath951 , @xmath1162 has a canonical choice up to @xmath1163 . ) note that @xmath31 is not unique even if @xmath951 is integral .
the linear system @xmath1164 _ has a cn @xmath1165-divisor _
@xmath306 if , for general @xmath1166 , the cn threshold @xmath1167 of @xmath1168 is @xmath1169 and @xmath1170 .
the cn @xmath1165-divisor is _ maximal _ if @xmath1171 .
note that @xmath1167 is independent of general @xmath31 and each nonexceptional divisor is cn if @xmath1172 .
suppose that @xmath1080 is nonrigid@xmath1077 and @xmath1173 with @xmath1174 is the first cte link .
the link goes over @xmath537 and let @xmath514 be its central model .
the model @xmath514 satisfies ( vi ) and ( viii ) , and ( iv - v ) with @xmath1164 , the birational transformation of @xmath1175 for some @xmath990 .
taking @xmath1164 with the fixed exceptional divisor for @xmath1176 , we get ( iii ) .
the contraction of flopping curves of @xmath514 in ( vi ) ( the anticanonical model ) gives a model in ( vii ) and ( ix ) .
conversely , the @xmath34-factorialization of models in ( vii ) and ( ix ) gives models in ( vi ) and ( viii ) respectively .
thus ( vi ) , ( viii ) and ( vii ) , ( ix ) are equivalent respectively .
note that , ( i ) , ( iv ) and ( vi ) are equivalent ( see the proof of corollary [ cor - central model exists ] ) , and follow from ( ii ) .
similarly , ( v ) and ( viii ) are equivalent and follow from ( iii ) .
a perturbation of the linear system @xmath1177 in ( iii ) allows us to construct a linear system in ( ii ) .
hence ( ii ) and ( iii ) are equivalent .
similarly , ( iv ) and ( v ) are equivalent .
finally , ( ii ) implies ( iv ) by definition .
[ cor - rigid mori ] a mori fibration @xmath1080 is rigid@xmath765 if and only if any mobile@xmath765 linear system @xmath1178 of rational degree @xmath1153 has no exceptional maximal b - divisors .
immediate by theorem [ thrm - nonrigid mori ] .
this implies that rigidity and nonrigidity are constructive conditions ( cf .
@xcite ) . recall that a _ mori - fano variety _ is a mori fibration over a point @xmath981 .
[ cor - rigid mori fano ] suppose that the base field k is algebraically closed and assume ( * ? ? ?
* conjecture 1.1 ) for projective weak fano varieties of dimension @xmath1078
. then the rigid ( respectively nonrigid ) mori - fano varieties of dimension @xmath951 form a constructive set in the coarse moduli of mori - fano varieties of dimension @xmath951 .
it is enough to consider nonrigidity . by theorem [ thrm - nonrigid mori ] , the nonrigid mori - fano varieties are the images of the terminal fano varieties @xmath89 with @xmath1179 .
more precisely , the latter should have a fixed @xmath34-factorialization and a fixed extremal ray on it .
then the image is given by a modifications starting from the ray .
the image gives mori - fano varieties for a constructive subset and is constructive itself .
this section gives an overview of the lmmp and semiampleness and detailed assumptions for the statements of the paper .
we use the basic notions and notations from @xcite @xcite @xcite .
below we recall some of them .
let @xmath1 be a morphism of varieties .
a variety means a geometrically reduced and irreducible normal algebraic variety over some field @xmath1180 , usually , of characteristic 0 ( not necessarily algebraically closed ) .
the role of the last assumption will be explained below . according to the following
we can suppose that @xmath1 is proper .
a model of @xmath1 is another _ proper _ morphism @xmath11 of varieties such that @xmath89 and @xmath237 are birationally isomorphic@xmath416 .
moreover , the birational isomorphisms @xmath1181 is fixed and called _ natural _ for the model @xmath11 . in other words , a _ model _
is such an isomorphism @xmath388 .
a _ natural birational isomorphism _ of models @xmath969 for @xmath1 is a unique birational isomorphism in the commutative diagram : @xmath24\ar@{-->}[rd ] & \\ y_1\ar@{-->}[rr ] & & y_2 } \ ] ] with natural birational isomorphisms @xmath1182 .
similarly , if @xmath925 , @xmath1183 are two proper morphisms of models@xmath416 , their _ ( rational ) morphism _ is a commutative square : @xmath709\ar@{-->}[r]&y_2 \ar[d]\\ t_1\ar@{-->}[r]&t_2 } \ ] ] with the natural birational isomorphism @xmath969 ; the _ natural _ rational morphism @xmath1184 is unique
. two models @xmath1185 are considered as the _ same ( equal ) _ if the natural birational isomorphism is biregular , that is , they are naturally _
isomorphic_. respectively , @xmath967 are _ naturally isomorphic _ if @xmath1186 are biregular isomorphisms . usually , everybody omit the adjective natural " .
the symbol @xmath1187 or @xmath1188 denotes a natural isomorphism and @xmath1189 any other .
for example , a quadratic transformation @xmath1190 gives a model @xmath1191 of @xmath1192 which is unnaturally isomorphic to @xmath1192 ( see example [ example - cremona ] above ) .
there are two subtleties in the model theory .
first , we can consider local and even formal theory , that is , we can replace @xmath1071 by a local or formal germ of @xmath1071 at some scheme point @xmath1193 .
this gives _
local _ and _ formal _ models over @xmath1071 whereas models over a variety @xmath1071 , especially for projective @xmath1071 , are known as _
global_. for simplicity , a reader can suppose that all models are global . however , everything works ditto for local and formal models .
second , that is more important for us , we consider models of pairs .
this leads to log models . a pair @xmath129 consists of a proper morphism @xmath1 and a b - divisor @xmath1194 .
usually , we suppose that @xmath1195 is a b - boundary .
a b - boundary means a finite b-@xmath49-divisor @xmath1196 where the sum runs over a finite set of distinct prime b - divisors @xmath1061 with multiplicities @xmath73 in the unit segment @xmath1197 $ ] of real numbers .
the divisors @xmath1061 are supposed to be fixed .
for example , all @xmath1198 and @xmath1199 s_i \cong [ 0,1]^m,\ ] ] where @xmath1200 is a fixed finite reduced b - divisor .
an element @xmath6 will be referred to as a _ boundary _
divisor even when it is actually a b - boundary .
let @xmath13 be a pair with a boundary @xmath6 .
a log pair @xmath12 is a _ log model _ of @xmath13 if @xmath11 is a model of @xmath1 , with a natural birational isomorphism @xmath1181 , and @xmath1203 is the _ log transformation _ of @xmath65 on @xmath237 , that is , @xmath1204 where @xmath519 is the trace of @xmath65 on @xmath237 and the summation runs over the distinct @xmath1205-exceptional prime divisors @xmath1061 on @xmath237 , not components of @xmath50 .
since @xmath12 is a log pair , for any canonical divisor @xmath983 of @xmath237 , the divisor @xmath1206 is @xmath49-cartier .
note that the boundary @xmath1203 is also the trace on @xmath237 of an _ infinite _ b - boundary @xmath1207 where @xmath1061 are the distinct prime divisors exceptional on @xmath89 and not prime components of @xmath50 . both boundaries @xmath1208 and @xmath1203 depend on @xmath50 .
however , if @xmath50 is a usual divisor ( i.e. , a b - divisor without exceptional components ) on @xmath89 , the divisors @xmath1208 and @xmath1203 are independent of @xmath50 .
it is a lc pair with a boundary .
such a pair @xmath12 is called an _ initial model _ of @xmath13
if @xmath12 is a log model of @xmath13 and the following inequalities hold : for each prime @xmath1209-exceptional divisor @xmath306 of @xmath89 , @xmath1210 and , for each prime component @xmath3 of @xmath50 exceptional on @xmath237 , @xmath1211 . using the initial log discrepancy @xmath1212 defined in section [ section2-geography ]
, we can combine the the two inequalities into one : @xmath1213 .
an initial _ strictly lt ( slt ) _ model is a slt pair with a boundary . recall that the slt property means that the variety is @xmath34-factorial and projective over the base , the pair is lt .
an initial model @xmath12 of @xmath13 is called an initial _ slt _ model of @xmath13 if @xmath12 is a slt pair and the above inequalities are strict for all divisors @xmath306 of @xmath89 and all prime components of @xmath50 exceptional on @xmath237 .
existence of an initial model for some pair @xmath13 is not obvious and related to log resolutions .
if a pair @xmath945 is log nonsingular and the divisor @xmath1214 is nonexceptional , then @xmath13 is initial for any boundary @xmath1215 . for any pair @xmath508 and for any of its projective@xmath416 log resolution @xmath1216 ,
the pair @xmath12 is a slt initial model of @xmath13 for all @xmath6 .
it is an initial model @xmath13 with a nef divisor @xmath389 .
the pair is a _ log canonical ( lc ) model _ if @xmath389 is ample .
an initial model @xmath12 of @xmath13 is called a _ weakly log canonical ( wlc ) model _ of @xmath13 if @xmath12 is a wlc model .
it will be a _ log canonical ( lc ) model _ of @xmath13 if @xmath12 is a lc model .
a lc model of @xmath13 is unique ( up to a natural isomorphism ) if it exists and will be denoted by @xmath392 where @xmath1217 .
a wlc model is called a _ strictly log terminal ( slt ) _ wlc model if it is also slt . respectively , a _ strictly log terminal ( slt ) _ wlc model @xmath12 of @xmath13 is a slt wlc model which is a slt initial model of @xmath13 .
it is an elementary fano log fibration ( * ? ? ?
* definition 1.6 ( v ) ) .
more precisely , it is an initial model @xmath13 with a contraction @xmath1218 such that : ( a ) @xmath1219 ; ( b ) @xmath13 has only @xmath34-factorial and lt singularities ; ( c ) the relative picard number @xmath1220 ; and ( d ) @xmath1221 is relatively ample@xmath765 . such a model without ( b - c ) ( respectively only without ( b ) ) is known as an ( elementary ) fano log fibration .
note that the contraction @xmath1080 is a fixed structure of the mori log fibration .
it is an _ slt _
mori log fibration if , additionally , @xmath744 is projective .
a mori log fibration is _ generic _ if it is considered over the general point of @xmath537 .
an initial model @xmath12 of @xmath13 with a contraction @xmath1222 is called a _
mori log fibration of _
@xmath13 if @xmath12 is a mori log fibration with the contraction @xmath514 .
similarly , one can define a _
fano log fibration of _ @xmath13 .
the model @xmath12 is an _ slt
_ mori log fibration of @xmath13 if @xmath12 is a slt initial model of @xmath13 and @xmath12 is also a slt mori log fibration .
thus , for each slt mori log fibration of @xmath13 , the pair @xmath12 is a slt mori log fibration , but not conversely in general even when @xmath12 is a mori log fibration of @xmath13 . similarly , for a wlc model @xmath12 of @xmath13 ,
a slt wlc model @xmath12 is not the same as a slt model of @xmath13 , in general .
if @xmath1223 and @xmath89 is trm then a slt mori log fibration @xmath13 is simply a _
mori fibration_. respectively , if @xmath1224 and @xmath237 is trm then a slt mori log fibration @xmath1225 is a _ mori fibration _ @xmath11 or a _ mori model _ _ of _ @xmath1 .
it is a wlc model or a mori log fibration ( or even a fano log fibration ) .
a _ resulting model _ of a pair @xmath13 is either a wlc model or a mori log fibration ( or even a fano log fibration ) of @xmath13 . according to @xcite
, any pair @xmath13 can not have both resulting models simultaneously .
however , a pair can have many wlc models or mori ( fano ) log fibrations .
it is expected that every pair @xmath13 with a boundary divisor @xmath65 has at least one of them , that is , a resulting model .
the log minimal model program ( lmmp ) allows us to construct a resulting model .
the _ lmmp _
is a special case of the @xmath31-minimal model program ( @xmath31-mmp ) @xcite .
an initial model for the lmmp is an initial model @xmath13 and @xmath1226 ( usually , it works with @xmath268 instead of @xmath1188 ) .
we suppose also that @xmath1 is projective for reasons explained ( see characteristic @xmath75 below ) .
then a @xmath31-minimal model is a projective wlc model and a nonbirational @xmath31-contraction @xmath1227 is a projective elementary fano log fibration .
the program itself is a chain of birational transformations of the initial model which are supposed to exist and terminate .
each transformation is a ( generalized ) log flip of a projective negative extremal contraction @xcite , but not necessarily elementary ( see below ) .
intermediate models obtained after each transformation are again projective initial models .
the last model under the lmmp is either a projective wlc model or a projective elementary fano log fibration of the initial model .
the last model under the lmmp starting from some initial model of @xmath13 is a resulting model of @xmath13 .
however , the slt lmmp is easier and more common .
an initial model of the _ slt lmmp _ is a slt initial model @xmath13 , in particular , @xmath1 is projective .
each transformation is either a log flip or a divisorial contraction which are supposed to be elementary .
intermediate models obtained after each transformations are again slt initial .
the last model under the lmmp is either a slt wlc model or a slt mori log fibration of the initial model .
the last model under the lmmp starting from some slt initial model of @xmath13 is a slt resulting model of @xmath13 : either a slt wlc model or a slt mori log fibration of @xmath13 .
an _ elementary _ contraction ( extraction ) or _ elementary _ small transformation ( modification ) is respectively a projective contraction ( extraction ) @xmath1228 or a small nonregular projective modification @xmath1229 such that @xmath611 are @xmath34-factorial and @xmath1230 .
actually , for an elementary contraction ( extraction ) , @xmath1231 and , for an elementary small transformation , @xmath1232 .
an _ extremal _ contraction or _
extremal _ transformation ( modification ) is respectively a contraction ( extraction ) @xmath1228 or a transformation @xmath1229 such that @xmath1230 .
actually , for an extremal contraction ( extraction ) , @xmath1231 and , for an extremal transformation , @xmath1232 thus each pair @xmath13 has a resulting model by the lmmp and a slt resulting model of @xmath13 by the slt lmmp , of course , if such a program exists for @xmath13 , including existence of an initial model ( see characteristic @xmath75 below ) .
for such a construction , a _ weak _ termination is enough , that is , some sequence of birational transformations terminates , gives a resulting model .
for example , the ( slt ) lmmp is established over any perfect field in dimension @xmath953 : for any algebraic surface over such a field , the lmmp exists .
[ cor - qsamp = rsamp ] in any fixed dimension @xmath951 , the slt lmmp in dimension @xmath951 and the semiampleness with @xmath34-boundaries imply the semiampleness with @xmath49-boundaries .
moreover , for the klt semiampleness with @xmath49-boundaries , the klt slt lmmp and the klt semiampleness with @xmath34-boundaries are sufficient .
let @xmath13 be a wlc model .
then by conjecture [ conj - semiample ] the @xmath49-divisor @xmath389 defines a _ lc ( iitaka ) contraction _
@xmath1234 onto a normal projective variety @xmath206 , called a _ lc ( iitaka ) model _ of @xmath13 .
in particular , if @xmath1235 is birational , the pair @xmath392 with @xmath1236 is the lc model of @xmath13 ( see weakly log canonical model above ) .
if both the lmmp and conjecture [ conj - semiample ] hold for a pair @xmath13 and the pair has a wlc model @xmath12 , then one can define a rational iitaka contraction @xmath1237 as a composition of the birational transformation @xmath22 and of the iitaka contraction of @xmath1238 . if @xmath282 is lc , the rational iitaka contraction is a @xmath72-contraction . by @xcite , the image @xmath394 and the rational map @xmath1239 depend only on the pair @xmath13 .
thus we can write @xmath394 instead of @xmath1240 . furthermore
, it is known that the rational iitaka contraction @xmath1235 and its image @xmath394 depend only on the @xmath104-class by definition(see corollary [ cor - property - p - e ] ) . in particular , the contraction @xmath1235 is independent of @xmath65 modulo @xmath35 .
the _ positive part _ of @xmath1241 is the b - cartier divisor @xmath1242 where @xmath12 is a wlc model of @xmath13 . by definition @xmath182
is nef and @xmath49-mobile by the semiampleness [ conj - semiample ] .
if @xmath13 does not have a wlc model , that is , @xmath1243 , put @xmath1244 .
@xcite given a pair @xmath13 with @xmath6 , we define the _ relative numerical kodaira dimension _ @xmath1247 as @xmath1248 if the resulting model of @xmath13 is a mori log fibration and as the integer @xmath1249 if @xmath13 has a wlc model @xmath12 . by definition ,
the following are equivalent : ( i ) @xmath44 , ( ii ) there exists a wlc model of @xmath13 , ( iii ) @xmath445 ( see section [ section2-geography ] above ) , ( iv ) @xmath1250 ( if the semiampleness holds [ conj - semiample ] ) .
if this is the case , the integer @xmath478 is independent of the choice of the wlc model @xmath237 @xcite .
* definition 2.2.3 ) let @xmath65 be an @xmath49-boundary on @xmath1 and @xmath11 be a model of @xmath1 such that the log pair @xmath12 is lc . then the _ relative invariant log kodaira dimension
_ @xmath1251 of the pair @xmath13 is defined as the integer : @xmath1252 where @xmath1253 denotes the relative iitaka dimension of @xmath31 . for a @xmath34-divisor @xmath31 ,
this is the same as the usual definition , but for an @xmath49-divisor @xmath31 , see ( * ? ? ?
* definition 2.2.3 ) .
most of the results in this paper , if not all , can be established under the following assumptions : existence of log resolutions , the lmmp and the semiampleness . of course , by the subsection lmmp above we can include log resolutions into the lmmp .
indeed , to construct an ( slt ) initial model of @xmath13 , one can use a log resolution ( see initial model above ) . however , usually it is not useful especially in characteristic @xmath75 , because in that case a log resolution always exists .
moreover , in characteristic @xmath75 some parts of the lmmp are established or announced : the cone theorem , the contraction of an extremal ray and existence of some special flips ( e.g. , slt ) .
according to the proofs , these results require projectivity of @xmath11 .
the main remaining problem is termination .
a _ weak _ termination , that is , some sequence of flips terminates , is sufficient for many results in birational geometry , in particular , for this paper ( cf .
assumptions for our results below ) .
thus we suppose the weak termination in the lmmp .
note also , that in characteristic @xmath75 , the semiampleness holds for any nef divisor of relative ft varieties .
the lmmp with usual termination is established only in dimension @xmath1254 @xcite and the slt lmmp with weak termination in dimension @xmath1255 @xcite .
for the klt big lmmp with weak termination in all dimensions see in @xcite .
the _ big " _ means that @xmath65 is big for an initial model , or equivalently , the same holds for each initial model .
since the lmmp is still not established even in characteristic @xmath75 and some of our results can be use in its proof , we prefer to put explicitly what is needed in each statement even some of assumptions is already proved . in the low dimension applications , we do not do this .
so , in addition to the assumptions in characteristic @xmath75 , we need the following .
note that the big conjecture [ conj - semiample ] means that of for big @xmath389 .
needed assumptions are mentioned after colons and each assumption is sufficient in the dimension of varieties of the statement and the dimension @xmath1256 .
proposition [ prop-0log cones coincide in u ] : the klt slt lmmp and conjecture [ conj - semiample ] for klt pairs with @xmath34-boundaries ; the klt slt big lmmp for any convex closed polyhedral subset of klt big boundaries .
corollary [ cor-0log num cones ] : the klt slt big lmmp .
comment : it looks that @xmath342 is only needed for the coincidence @xmath268 and @xmath280 . for arbitrary characteristic
, one can use conjecture [ conj - semiample ] .
iskovskih , _ on the rationality problem for three - dimensional algebraic varieties fibered over del pezzo surfaces . _ trudy mat .
inst . steklov . 208
( 1995 ) , teor .
chisel , algebra i algebr .
geom . , 128 - 138
. y. kawamata , k. matsuda and k. matsuki , _ introduction to the minimal model program _
, algebraic geometry , sendai ( t. oda , ed . ) , kinokuniya - north- holland , 1987 , adv . stud .
pure math , vol 10 , pp .
283 - 360 .
shokurov , _ anticanonical boundedness for curves .
_ appendix to nikulin v.v .
hyperbolic reflection group methods and algebraic varieties in higher dimensional complex varieties ( trento , june 1994 ) ed .
andreatta m. , and peternell t. berlin : new york : de gruyter , 1996 , 321 - 328 . | an introduction to geography of log models with applications to positive cones of ft varieties and to geometry of minimal models and mori fibrations . |
the interpretation of low amplitude , periodic radial velocity variations in solar - type stars as being induced by orbiting low mass companions was recently put into question .
gray ( 1997 ) and gray & hatzes ( 1997 ) claimed to detect distortions to the line profile bisector ( the locus of midpoints of a stellar absorption line from the core up to the continuum ) in one spectral line ( fe i @xmath0 ) of the star 51 pegasi at the now well known 4.23 day radial velocity period ( mayor & queloz 1995 ; marcy et al .
they stated that the planet hypothesis could not account for variations in the line profiles , and proposed the alternative explanation that 51 peg was undergoing non - radial pulsations .
further observations and analyses ( gray 1998 ; hatzes , cochran & bakker 1998a , b ; brown et al .
1998a , b ) failed to confirm the claimed period in the line bisector , and the planet hypothesis has emerged as the most reasonable explanation of the radial velocity curve . nevertheless , the question of interpretation of variations in the line bisector remains .
is it a given that the planet explanation is immediately excluded should intrinsic variations in the line profiles of 51 peg be detected at the claimed orbital period ?
our purpose in this letter is twofold ; first , to provide a simple model whereby a close - in orbiting companion produces periodic distortions to the line profiles , and second , to explore the observational consequences of this effect , the amplitude of which might be close to current observational limits , and the specific signature of which is set by the orbital inclination of the system and the radius and albedo of the companion .
the amount of light reflected by the planets of our solar system is a tiny fraction of the solar output .
however , for the close - in extrasolar giant planets ( cegps ) discovered to date , this fraction increases substantially due to the proximity of a large planet to the star .
the semi - major axes of these systems , @xmath1 boo , 51 peg , @xmath2 and and @xmath3 cnc , are 0.046 , 0.051 , 0.057 and 0.11 au respectively ( mayor & queloz 1995 ; butler et al .
an estimate of the flux ratio can be achieved as follows .
let @xmath4 denote the distance from the star to the planet , @xmath5 the radius of the planet , @xmath6 the radius of the star , and @xmath7 the angle between the star and the earth as seen from the planet .
we assume that @xmath8 and that each surface element on the planet reflects a fraction @xmath9 of the incident flux uniformly back into the local half - sky ( lambert s law ) .
if @xmath9 does not vary with position , then it is simply the bond albedo .
integration over the sphere yields the ratio of the observed flux of the planet at @xmath10 to that of the star , @xmath11 for either 51 peg b ( @xmath12 au and assuming @xmath13 ( guillot et al .
1996 ) ) or @xmath1 boo b ( @xmath14 au and assuming @xmath15 ) , the result is @xmath16 .
the observed flux ratio will vary with the angle @xmath7 , which is a combination of the orbital inclination @xmath17 and the orbital phase @xmath18 . here
we define the orbital phase to be the one conventionally given by radial velocity measurements ; for a circular orbit ( the only case we consider , as tidal effects circularize the orbits of these extremely close companions ) , @xmath19 is taken to be the time of maximum recessional velocity of the star .
the angle @xmath20 $ ] is then given by @xmath21 .
integration of the intensity over the surface of the sphere viewed at an angle @xmath7 yields the phase dependent observed flux ratio , @xmath22 for a discussion of the albedos and phase functions of solar system objects , see harris ( 1961 ) . note that in the expression for @xmath23 , we have ignored the possibility of occultation ; eclipses would have a much more pronounced effect and are ruled out for @xmath1 boo , 51 peg and @xmath3 cnc ( henry et al .
1997 ; baliunas et al .
1997 ) .
thus cegps at moderate to high inclination and of high reflectivity will produce a photometric modulation of the system which will be accessible to upcoming satellite missions , as discussed by charbonneau ( 1998 ) .
however , photometry does not make use of the spectroscopic separation of the planet and star .
the key to the method we propose here is in analyzing the spectrum of the system . assuming the planet reflects an essentially undistorted spectrum of the star , a copy exists in the observed stellar spectrum with relative amplitude @xmath23 but at a location which is doppler shifted due to the motion of the planet .
the ratio of the amplitude of the radial velocity variations of the planet to that of the star , @xmath24 , is simply the ratio of the mass of the star to the mass of the planet , @xmath25 . from the single line spectroscopic orbit ,
only the minimum mass of the planet , @xmath26 , is determined ( assuming the stellar mass is estimated via other means ) .
thus , only the combination @xmath27 is known . for 51 peg , @xmath28 and @xmath29 , so @xmath30 .
thus there can be a huge difference in velocity amplitude between the stellar and reflected light signal . for a given stellar spectral line , the reflected line will move back and forth as the phase changes . at @xmath31 and @xmath32 the radial velocities of the star and planet
will be the same and the reflected line will lie on top of the stellar line . at @xmath19 and @xmath33 ,
the lines will be maximally separated .
a single spectrum with sufficient signal - to - noise would in principle be able to show both .
we have obtained many high signal - to - noise , high resolution spectra of @xmath1 boo , and are attempting to isolate the doppler shifted reflected spectrum from the orders of magnitude brighter stellar spectrum by making use of the known orbital parameters supplied by radial velocity measurements ( butler et al .
the results of this complementary investigation will be presented in a separate paper . here
we present an alternate method of detection which capitalizes on the high resolution observations and analysis methods that have already been developed for studying line bisectors .
the motion of the reflected line through the stellar line will cause a small distortion which might be detectable by careful measurements of the line bisectors or through other techniques .
we assume that the planet reflects an undistorted spectrum of the primary , leading to a composite spectrum consisting of a reflected component greatly reduced in amplitude and doppler shifted by the orbital motion of the companion .
we model this process for an individual isolated spectral line .
typical line bisectors for solar type stars show a distinctive * * shape caused by granulation in the photosphere ( see e.g. , gray 1992 ) . here
, we have described the lines as simple gaussian profiles of appropriate width and depth for the star in question . while we have calculated more detailed models ( multiple gaussian components to the line )
so as to reproduce the constant bisector * * shape , the effect of this is simply to add a time invariant term to the bisector shape derived from the symmetric gaussian model .
the velocity of the planet relative to the star at phase @xmath18 is simply @xmath34 we have modelled the stellar and planetary line profiles as a function of velocity by @xmath35 where we have explicitly allowed for a different planetary line width ( @xmath36 ) and depth ( @xmath37 ) due to the rotation rates of the planet and star ( see below ) .
the flux ratio is given by equation [ flux ] , thus the observed spectrum of the system is @xmath38 where we have renormalized to the continuum flux level .
the effect of the reflected light component on the observed bisector is shown schematically in figure [ twolines ] .
for each point in the phase , we deduce the bisector by evaluating the midpoint between the two halves of the absorption feature at a number of flux levels , interpolating between sampled points to the desired flux .
we have also characterized these distortions in terms of two parameters conventionally used in bisector analyses , the velocity span and the bisector curvature .
the velocity span measures the velocity difference in the bisector at two flux levels ; we have chosen these levels at two percent below the continuum ( @xmath39 ) and two percent above the line depth ( @xmath40 ) .
the bisector curvature is defined to be the velocity span of the top portion ( between @xmath39 and @xmath41 ) minus the velocity span of the bottom portion ( between @xmath41 and @xmath42 ) of the bisector . for convenience , we assume @xmath43 in our model calculations . the amplitude of the distortion to the bisector ( and hence to the velocity span and curvature ) is linear in @xmath9 .
the results we present below should therefore be scaled for a given choice of the reflectivity .
we have computed simulations for 51 peg with @xmath44 and a stellar line width of @xmath45 .
we assume the reflected light is undistorted , so that @xmath46 and @xmath47 .
we sample the data simulating a spectral resolution @xmath48 comparable to current high - resolution observations . the predicted distortions to the bisector for @xmath49 are shown in the left panel of figure [ bisectors ] for a selection of radial velocity phases .
the deduced velocity span and bisector curvature are shown in the top panels of figures 3 and 4 respectively .
the bisector is distorted for the fraction of the phase where the stellar and planetary lines overlap and the planet is sufficiently illuminated as viewed by the observer . for a near edge - on geometry in the case of 51 peg , this corresponds to roughly 3% of its 4.23 day period , or about 3 hours , centered on @xmath32 .
although the stellar and planetary lines also overlap at @xmath31 , there is less distortion as the planet is less illuminated as viewed by the observer . for lower inclinations ,
the fraction of the period in which there is a significant distortion is greater , but the amplitude of the distortion is less .
we have also modelled this effect for another cegp system , @xmath1 boo ( @xmath50 and @xmath51 , ( butler et al . 1997 ) ) .
observations of ca ii h @xmath52 k lines suggest that the rotational period of the star is the same as the observed orbital period of the companion ( baliunas et al .
1997 ) , implying that the star and planet are tidally locked .
the synchronization timescales for @xmath1 boo are discussed in marcy et al .
if indeed the system is tidally locked , the planet does not see the rotational broadening of the spectral lines of the primary , as there is no relative motion between any point on the surface of the planet and any point on the surface of the star .
thus the planet reflects a rotationally unbroadened spectrum of the primary , with an intrinsic width due to the stellar photospheric convective motions .
an observer then sees this line broadened only by the planetary rotation , which in this tidally locked scenario is small .
thus in the case of @xmath1 boo , we predict relatively narrow planetary lines ( we adopt @xmath53 ) superimposed on the much broader stellar lines ( @xmath54 ) .
the results of our simulated bisector distortions for @xmath49 are shown in the right panel of figure [ bisectors ] for a selection of radial velocity phases .
the deduced velocity span and curvature with phase are shown in the bottom panels of figures [ span ] and [ curvature ] respectively .
the line is significantly distorted for the fraction of the phase where the planetary and stellar lines overlap , which now is substantially increased due to the high degree of rotational broadening experienced by the lines of the star . for an edge - on configuration
, this amounts to roughly 10% , or about 8 hours of the 3.3 day period , centered on @xmath32 .
due to the different line widths , the deduced bisector varies with a much greater amplitude and in a more complex manner than in the case of 51 peg .
the velocity span and curvature do not well characterize the behavior of the bisector .
we advocate that the full bisectors be examined in a search for this effect .
we have shown that the presence of a close - in extrasolar giant planet can distort the observed spectral line profiles of the system through the contribution of reflected light from the companion . for 51 peg , assuming @xmath49 and @xmath55 , the amplitude of the distortion is roughly @xmath56 and @xmath57 ( peak - to - peak ) to the velocity span and curvature , respectively .
if @xmath9 is high , the effect is commensurate with current observational limits : hatzes et al .
( 1998a ) combined measurements of five spectral lines of 51 peg and achieved a scatter of @xmath58 and @xmath59 for the velocity span and curvature .
more recent results ( hatzes et al .
1998b ) achieve a precision of @xmath60 for the velocity span and @xmath61 for the curvature . in the case of @xmath1
boo , assuming that the system is tidally locked and that @xmath49 and @xmath55 , the velocity span and curvature distortions are both about @xmath62 peak - to - peak .
again , for a high value of @xmath9 , these are close to the current upper limits as measured by hatzes & cochran ( 1998 ) .
the amplitude of the velocity span and curvature variations we have stated here rely upon measurement of the line bisector to within 2% of the continuum flux level .
bisector measurements tend to avoid the extreme wings of the line profile as @xmath63 is small and hence the deduced bisector is more easily corrupted by noise . for 51 peg ,
the equal linewidths of the stellar and reflected spectrum mean that much of the signal in the velocity span and curvature is a result of the distortion of the bisector close to the continuum , as shown in figure 2 .
for @xmath1 boo , the different line widths in the tidally locked scenario produce a second trough and peak in the velocity span and curvature which are insensitive to the height to which the bisector is measured , because the distortion occurs away from the continuum . note that confirmation of the predicted distortions does not require accurate measurement of the _ intrinsic _ line bisector , only precise measurement of the temporal variations to it , and
thus it is not threatened by contamination from line blends .
also , isolated lines are not required as distortions induced by the transit of nearby lines can easily be modelled under the assumption that the planet reflects a doppler shifted copy of the stellar spectrum .
the estimate of the reflected light amplitude presented in equation 1 does not consider the underlying source of the albedo , nor any angle dependent scattering effects associated with it .
estimates of the albedos of these cegps are uncertain as the chemistry of these atmospheres is unknown .
the gas giants of the solar system have bond albedos ranging from 0.7 to 0.9 .
the large reflectivity is primarily due the presence of ices and clouds high in the atmosphere , and these could not exist at the much greater effective temperatures of cegps .
dominant sources of reflectivity for cegps are scattering from dust and rayleigh scattering .
detailed modelling of the atmospheric chemistry of these planets indicates that condensates such as @xmath64 may indeed form at the required height ( burrows et al .
recent models by seager & sasselov ( 1998 ) , which explicitly solve the radiative transfer through a model atmosphere for @xmath1 boo b , predict an albedo of less than 1% at 5000 .
although this model is incomplete in that it considers only one dust species , of interstellar size distribution and solar adundance , it seems unlikely that albedos as high as those of the solar system gas giants could be reached by a dusty cegp ; further investigation is called for .
an upper limit on the bisector distortions which is sufficiently strict to disallow a large value of @xmath9 would provide support for such models .
via reflected light , a close - in extrasolar giant planet could produce a distortion to the stellar spectral line bisectors at a level which is approaching the current observational upper limits , given a sufficiently high albedo .
the variation of the bisector distortions with phase has a very distinctive signature which depends on the inclination of the orbit of the planet to the line of sight .
measurement of such periodic bisector variations would constitute a direct detection of the companion , a major advance in the field of extrasolar planet research .
this would yield the inclination , and hence the mass of the companion .
it would also measure the product of the albedo and the square of the radius of the planet , from which a minimum planetary radius may be deduced .
an estimated albedo would yield a value for the planetary radius , which can be combined with the measured mass to calculate such critically important quantities as the surface gravity and average density .
conversely , a firm upper limit on such distortions would place a constraint on the albedos of these planets . either by detection or by an upper limit
, an accurate measurement of the line profile bisector would place an important observational constraint on models of these close - in extrasolar giant planets .
we wish to thank david latham , in whose informative seminar this idea was conceived , and timothy brown , artie hatzes and dimitar sasselov for pragmatic comments .
this work was supported in part by a nsf graduate research fellowship .
99 baliunas , s. a. , henry , g. w. , donahue , r. a. , fekel , f. c. , & soon , w. h. 1997 , , 474 , l119 brown , t. m. , kotak , r. , horner , s. d. , kennelly , e. j. , korzennik , s. , nisenson , p. , & noyes , r. w. 1998a , , 494 , l85 brown , t. m. , kotak , r. , horner , s. d. , kennelly , e. j. , korzennik , s. , nisenson , p. , & noyes , r. w. 1998b , , _ in press _
, astro - ph/9712279 burrows , a. et al .
1997 , in the tenth cambridge workshop on cool stars , stellar systems , and the sun , eds .
donahue & j.a .
bookbinder ( san francisco : asp ) , 27 , _ in press _ , astro - ph/9709278 butler , r. p. , marcy , g. w. , williams , e. , hauser , h. , & shirts , p. 1997
, , 474 , l115 charbonneau , d. 1998 , in planets outside the solar system : theory and observations , nato asi series , ed . j .-
mariotti ( dordrecht : kluwer academic publishers ) , _ in press _ gray , d. f. 1992 , the observations and analysis of stellar photospheres , second edition ( cambridge : cambridge university press ) gray , d. f. 1997 , , 385 , 795 gray , d. f. 1998 , , 391 , 153 gray , d. f. , & hatzes , a. p. 1997
, , 490 , 412 .
guillot , t. , burrows , a. , hubbard , w. b. , lunine , j. i. , & saumon , d. 1996 , , 459 , l35 harris , d. l. 1961 , in the solar system , vol .
3 , planets and satellites , ed . g. p. kuiper and b. m. middlehurst ( chicago : univ .
chicago press ) , 272 hatzes , a. p. & cochran , w. d. 1998 , , 502 , 944 hatzes , a. p. , cochran , w. d. , & bakker , e. j. 1998a , , 391 , 154 hatzes , a. p. , cochran , w. d. , & bakker , e. j. 1998b , , _ in press _ henry , g. w. , baliunas , s. l. , donahue , r. a. , soon , w. h. , & saar , s. h. 1997 , , 474 , 503 marcy , g. w. , butler , r. p. , williams , e. , bildsten , l. , graham , j. r. , ghez , a. m. , & jernigan , j. g. 1997 , , 481 , 926 mayor , m. , & queloz , d. 1995 , , 378 , 355 seager , s. , & sasselov , d. 1998 , , 502 , l157 | we discuss the interpretation of distortions to stellar spectral lines with particular attention to line bisectors in the presence of an orbiting planetary companion .
we present a simple model whereby light _ reflected _ by the companion can cause temporal variations to the observed line profiles .
these distortions have a characteristic signature depending on the inclination angle of the system . for the known close - in extrasolar giant planets ,
the expected amplitude of the effect might not be far from current detection capabilities .
this method could be used to directly detect the presence of the companion , yielding the orbital inclination and hence the planetary mass .
futhermore , a detection would measure a combination of the planetary radius and albedo , from which a minimum radius may be deduced . |
accelerator mass spectrometry ( ams ) @xcite is an established technique for the detection and measurement of trace level of long - lived radionuclides @xcite such as @xmath13c , @xmath14be , @xmath15cl .
it improves over conventional techniques such as neutron activation analysis ( naa ) , inductively - coupled plasma mass spectrometry ( icp - ms ) as well as secondary ion mass spectrometry ( sims ) , and has been successfully applied to a wide variety of disciplines like archeology , geology , environmental science , biomedicine , safeguards of nuclear materials and so on @xcite .
measurement of trace concentrations of naturally - occurring , as well as cosmic - ray or fission induced radioactive isotopes is an important technique with major impact to low background experiments @xcite .
the texono collaboration is pursuing a research program in low energy neutrino and astroparticle physics @xcite .
an important aspect is to devise procedures for measuring such trace radiopurity using the ams techniques .
significant improvements in sensitivities and flexibilities over existing methods can be expected .
results were published already for the measurement of trace @xmath16i @xcite in inorganic crystal and organic liquid scintillator ( ls ) .
the isotope @xmath16i is a long - lived fission fragment , where the corresponding ams procedures are already matured and well - established . in this article .
we report on our efforts and results on potassium ( @xmath17 ) .
the isotope @xmath17 is stable and exists in large abundance in nature @xmath18 and there are no prior attempts to measure trace abundance of it with ams .
these poses new experimental challenges and the results reported here represent advances in the field .
potassium has three naturally - occurring isotopes : @xmath17 , @xmath0 and @xmath19k with isotopic abundances of 93.26% , 0.0117% and 6.73% , respectively . among these ,
@xmath17 and @xmath19k are stable while @xmath0 has a half - live of @xmath20 years , where the dominant decay channels are : @xmath21 and @xmath22 both decays pose severe background problems to low - count - rate experiments where the signals are at the sub - mev energy range , such as the studies of pp and @xmath23be solar neutrinos , the searches of cold dark matter , as well as of neutrino magnetic moments with reactor neutrinos @xcite .
potassium contaminations in the active detector materials will give rise to a continuous @xmath24-decay background up to 1.31 mev , while those in the passive materials in the vicinity of the detector will produce @xmath25-rays at 1.46 mev which , after compton scatterings , can lead to background at the sub - mev range .
the suppression and an efficient measurement technique of ultra - low amount of potassium in a wide spectrum of materials are therefore an issue of crucial significance . as an illustrative example , in order for the borexino and kamland experiments to achieve the goals of studying the low energy @xmath23be solar neutrinos with ls , the required contaminations of @xmath0 should be less than @xmath26 g / g .
while the radioactive isotope of interest to low background experiments is @xmath0 , the ams techniques devised in this work are for the measurement of @xmath17 .
the @xmath17-to-@xmath0 ratio of 7971:1 is constant for naturally - occurring materials .
therefore , a measurement of ultra - low @xmath0 radio - purities is equivalent to one on the level of @xmath17-contaminations , but with a sensitivity enhancement factor of almost @xmath27 .
in addition , @xmath0 measurements by the ams techniques are vulnerable to background contaminations by @xmath28ca , a stable isotope which exists in abundance in nature .
the sensitivities of the various techniques to the measurement of ultra - low @xmath0 contaminations are shown in table [ compare ] .
photon counting of the 1.461 mev @xmath25-line by high - purity germanium detectors ( hpge ) @xcite is limited by the ambient background .
the counting test facility ( ctf ) @xcite at the gran sasso laboratory is a 4 ton ls detector .
it took @xmath2930 days to measure only a single sample of ls , and is therefore of limited flexibilities . to address these drawbacks , the naa techniques for potassium @xcite
were recently developed , where the sensitivities of @xmath30 were achieved . in this paper
, we present results on a complementary approach using the ams techniques . .
comparisons of the various techniques for ultra - low potassium measurements .
[ cols="<,^,^,^,^,^ " , ] the relation established in eq .
[ eq::conversion ] is applicable for calibrations of ams measurements in solid samples following the processing procedures of figure [ lsprocess ] . for organic liquid samples , an extra factor
@xmath18 the survival efficiency of potassium at the solid residuals after vacuum evaporation ( @xmath31 ) @xmath18 has to be evaluated .
this was studied by measuring the @xmath17-yields in solid ppo ( @xmath32 ) and comparing it to those samples with ppo mixed with pc .
the efficiency @xmath31 is given by @xmath33 where @xmath34 are , respectively , the @xmath17-levels in ppo solid , pure pc liquid without ppo , and pc mixed with ppo in known fractions @xmath35 .
the results are summarized in table [ tab::pposcan ] .
the most accurate measurement is one with a high concentration of ppo ( @xmath35=1% ) , giving @xmath36 .
this result demonstrates that most potassium in the ppo+pc solution remained in the solid residuals after evaporation .
measurements with smaller ppo concentrations were still consistent with complete potassium extraction , but with larger uncertainties due to limited statistics above the background .
samples of different masses of pure ppo powder , pure pc liquid , as well as their mixtures as liquid scintillators were measured , following the proper procedures of calibrations and background subtraction discussed above .
the results are summarized in table [ lsresults ] .
consistent results were obtained for the two decades variations in sampling mass , in both @xmath37}$ ] and @xmath38}$ ] . the liquid scintillator sample was commercially mixed .
the measured @xmath0 contamination was @xmath39 ~ = ~ ( 1.3 \pm 0.1 ) \times 10^{-13 } ~
g / g } \ ] ] implying equivalently a @xmath0-activity of @xmath40 .
such commercial ls sample is not yet good enough for the low background experiments .
further purifications are therefore necessary .
the source of contaminations come primarily from the ppo solute , such that it would be more effective to focus the purification program on it .
the best measurement is , expectedly , one on pure pc at the largest sampling mass , where @xmath41 ~ = ~ ( 1.4 \pm 0.1 ) \times 10^{-15 } ~
g / g } ~.\ ] ] the expected @xmath0-decay rate is @xmath42 .
these characteristic features should be valid in general @xmath18 that the contaminations in organic liquid scintillator come primarily from the dye solutes while the liquid solvent are orders of magnitude cleaner .
another two measurements of pc with much smaller sampling masses did not give rise to statistically significant excess of the @xmath37}$ ] ratio , such that only 90% confidence level upper limits are derived @xcite in table [ lsresults ] .
the limits are consistent with the measured results of eq .
[ eq::k40pc ] .
it is instructive to compare the results of tables [ lsresults]&[tab::pposcan ] , which are from the measurements on two different batches of ppo and pc samples from the same supplier .
the @xmath17 contaminations of the two batches of ppo are , respectively , @xmath43 and @xmath44 , while the those for the pc are @xmath45 and @xmath46 .
the differences indicate the range of variations of potassium levels in different production batches among the same materials .
an important parameter to characterize the performance of this measurement technique is the limiting sensitivity .
it represents the upper limits that can be set for samples where no excess of @xmath17-counts above background are observed @xmath18 or alternatively , the minimal contamination levels that will produce a positive measurement .
the measurements of the csi samples do not involve chemical pre - processing .
therefore , the sensitivity is limited by the ion - beam related background measured in the `` blind target '' run discussed in section [ sect::beambkg ] .
the @xmath17 background count rates of @xmath4750 per second at the ams detector can be translated to limiting sensitivities following the same normalization steps of section [ sect::csi ] : @xmath48 _ { limit } } & ~ \sim ~ & \rm { 6 \times 10^{-9 } ~
g / g , ~~ or ~ equivalently } \nonumber \\ \rm { [ ~ \k40 / csi ~ ] _ { limit } } & ~ \sim ~ & \rm { 8 \times 10^{-13 } ~
g / g } ~. \label{eq::limitcsi}\end{aligned}\ ] ] for organic liquid , potassium introduced during the pre - processing procedures is the dominant background
. the limiting sensitivities of @xmath49}$ ] can be derived from the uncertainties of the background blank measurement in eq .
[ eq::bkg ] , giving @xmath50 _ { limit } ~ \sim ~ 2 \times 10^{-9 } ~
g / g ~ . } \ ] ] to convert into [ @xmath17/(org.liq . ) ] units , the relation of eq .
[ eq::conversion ] is used .
the ratio of [ m(sample)/m(cu ) ] is taken from the measurement with the largest liquid sample volume ( 100 ml ) where the extraction of the contaminants was successfully demonstrated .
the sensitivities are @xmath51 _ { limit } } & ~ \sim ~ & \rm { 2 \times 10^{-13 } ~
g / g , ~~ or ~ equivalently } \nonumber \\ \rm { [ ~ \k40 / ( org . liq . ) ~ ] _ { limit } } & ~ \sim ~ & \rm { 3 \times 10^{-17 } ~
g / g } ~. \label{eq::limitorgliq}\end{aligned}\ ] ] the improvement of sensitivities in eqs .
[ eq::limitorgliq]&[eq::limitcsi ] of organic liquid over solid csi powder follows from the [ m(cu)/m(sample ) ] ratio in eq . [ eq::massscaling ] . at a typical m(cu)@xmath2910 mg and the largest liquid extraction volume of 100 ml
[ that is , m(sample)@xmath29100 g ] , an enhancement factor of @xmath52 is readily attained in the evaporation and extraction processes .
in this article , we report the first measurement of ultra - low potassium contaminations in an inorganic salt ( csi ) and in an organic scintillator liquid ( ls = pc+ppo ) using accelerator mass spectrometry . comparable sensitivities to the neutron activation techniques were achieved .
potassium were positively identified in samples of commercially available csi and ls , at the contamination levels of @xmath53 and @xmath54 , respectively .
the radioactivity in the ls is mainly due to the dye solutes .
the limiting sensitivities of the techniques are of the order of @xmath55 g / g of @xmath0 in solid powder and of @xmath56 g / g of @xmath0 in an organic liquid . in the case of organic liquid , the sensitivities are limited by ambient potassium impurities introduced during the chemical processing of the liquid depicted in figure [ lsprocess ]
. further improvement can be made along these directions .
the beam - related background is small in comparison .
a boost in sensitivities can also be made with the use of larger initial liquid volume for evaporation .
the largest sample demonstrated in the present studies is 100 ml .
extensions to treatment of liter - scale sample are expected to be technically similar , such that a limiting sensitivity of @xmath57 on liquid scintillator using ams should be possible .
this work demonstrate the feasibilities of measuring ultra - low impurities of naturally - occurring yet unstable isotopes with ams .
accordingly , other radio - isotopes which are problematic to low - background experiments , such as the @xmath58u and @xmath58th series as well as @xmath59rb , @xmath60cd , @xmath61 in , @xmath62la and @xmath63lu , can also be measured by the ams techniques .
the goals of our future efforts would be to devise schemes to measure some of these isotopes .
extensions to rb will be relatively straight - forward , while the measurements of the @xmath58u and @xmath58th series will involve upgrades of the present accelerator systems .
the authors are grateful to m. lin , s.h .
li , y.d .
chen , a.l .
li , j. li , k.x .
liu , b.f .
ni , h.q . tang , w.z .
tian and z.y .
zhou for helpful discussions , as well as to q.b .
you , y.b . bao , y.m .
hu and the technical staff at ciae for smooth accelerator operation .
this work is supported by contract 10375099 from the national natural science foundation , china .
u. fehn et al . ,
nature * 323 * , 707 ( 1986 ) ; + j.c .
rucklidge , g.c .
wilson and l.r .
kilius , nucl .
instrum . and methods * b 52 * , 507 ( 1990 ) ; + h.e .
gove et al . , nucl .
instrum . and methods * b 52 * , 502 ( 1990 ) . | levels of trace radiopurity in active detector materials is a subject of major concern in low - background experiments . among the radio - isotopes ,
@xmath0 is one of the most abundant and yet whose signatures are difficult to reject .
procedures were devised to measure trace potassium concentrations in the inorganic salt csi as well as in organic liquid scintillator ( ls ) with accelerator mass spectrometry ( ams ) , giving , respectively , the @xmath0-contamination levels of @xmath1 and @xmath2 g / g . measurement flexibilities and sensitivities are improved over conventional methods .
the projected limiting sensitivities if no excess of potassium signals had been observed over background are @xmath3 g / g and @xmath4 g / g for the csi and ls , respectively .
studies of the ls samples indicate that the radioactive contaminations come mainly in the dye solutes , while the base solvents are orders of magnitude cleaner .
the work demonstrate the possibilities of measuring naturally - occurring isotopes with the ams techniques . * measurement of ultra - low potassium contaminations with accelerator mass spectrometry * + 0.5 cm k.j .
dong@xmath5 , h.t .
wong@xmath6 , m. he@xmath7 , s. jiang@xmath7 , + j.z .
qiu@xmath8 , y.j .
guan@xmath9 , s.y .
wu@xmath7 , j. yuan@xmath7 .2 cm @xmath7 department of nuclear physics , china institute of atomic energy , beijing 102413 .
+ @xmath10institute of physics , academia sinica , taipei 11529 .
+ @xmath11armed police force academy , langfang 065000 .
+ @xmath12department of physics , guanxi university , nanning 530004 .
+ 0.5 cm * pacs codes : * 07.75.+h , 92.20.td , 41.75.-i + * keywords : * mass spectrometers , radioactivity , charged - particle beams |
the lyman forest in qso absorption spectra is now generally believed to be caused by absorption by large - scale density fluctuations of moderate amplitude in a warm photo - ionized intergalactic medium ( igm ) .
this is substantially different from the old cloud picture for the ly@xmath19 forest ( see rauch 1998 for a review ) .
it is mainly sustained by cosmological hydrodynamical simulations which successfully reproduce the absorption features in qso spectra ( cen et al .
1994 ; petitjean , mcket & kates 1995 ; zhang , anninos & norman 1995 ; hernquist et al . 1996
; miralda - escud et al .
1996 , theuns et al .
the picture had , however , been suggested before these numerical simulations became available ( bond , szalay & silk 1988 ; bi , brner & chu 1992 ) .
important results of the simulations are a tight correlation between the and the dark matter distribution ( on scales larger then the jeans length of the igm ) and a simple gas temperature - density relation which depends only on the reionization history of the universe ( hui & gnedin 1997 , haehnelt & steinmetz 1998 ) . as a consequence ,
the ly@xmath19 forest can be an important probe for the distribution of dark matter and its evolution over a wide redshift range ( c.f .
weinberg 1999 for a review ) on scales smaller than those accessible by the galaxy distribution or the lensing effects of foreground large scale structures on background galaxies ( c.f .
bartelmann & schneider 1999 ) .
for example , gnedin & hui ( 1998 ) showed that the column density distribution of ly@xmath19 absorption lines depends on the amplitude of the density fluctutions in the igm .
croft et al used the flux power spectrum of qso absorption spectra to constrain the dm power spectrum on scales from @xmath20 to @xmath21 mpc ( croft et al .
1998 , croft et al .
1999 , weinberg et al .
nusser and haehnelt ( 1999 ) ( hereafter nh99 ) proposed an iterative procedure to obtain the real space density and its probability distribution function . here
we apply a modified version of this technique to two hires spectra of the qso q1422 + 231 .
we infer the fluctuation amplitude of the dm density on the effective jeans scale and constrain the normalization constant of the optical depth distribution which depends on the ionization rate @xmath22 , the baryon density @xmath23 and the hubble constant ( @xmath24 ) .
section 2 describes our modified technique to extract moments of the probability distribution ( pdf ) of the density and presents tests on mock spectra . in section 3
we present the application to the data .
section 4 discusses implications for the cosmological density parameter . in section 5
we relate our work to previous work and give an outlook .
section 6 contains our conclusions .
the optical depth in redshift space due to resonant ly@xmath19 scattering is related to the density along the line of sight ( los ) in real space by @xmath25 \dd x , \\ \nonumber \label{tau}\ ] ] where @xmath26 is the effective cross section for resonant line scattering , @xmath27 is the hubble constant at redshift @xmath28 , @xmath29 is the real space coordinate ( in @xmath30 ) , is the voigt profile normalized such that @xmath31 , @xmath32 is the los peculiar velocity , and @xmath33 is the doppler parameter due to thermal / turbulent broadening .
the absorption features in the ly@xmath19 forest are mainly produced by regions of moderate densities where photoheating is the dominant heating source and shock heating is not important .
hydrogen in the igm is highly ionized ( gunn & peterson 1965 , scheuer 1965 ) and the photoionization equilibrium in the expanding igm establishes a tight correlation between neutral and total hydrogen density .
this relation can be approximated by a power law @xmath34 , where the parameter @xmath19 depends on the reionization history .
the possible range for @xmath19 is is @xmath35 with a value close to @xmath36 just after reionization , and decreasing at later times ( hui & gnedin 1997 ) .
numerical simulations have shown that the gas density traces the fluctuations of the dm density on scales larger than the jeans length , so that @xmath37 , where @xmath38 is the density at mean dark matter density . in this relation @xmath39
is the dark matter density smoothed on the jeans length scale below which thermal pressure becomes important .
the jeans length in the linear regime is given by ( comoving ) , @xmath40 where @xmath41 is the sound speed .
however , in the non - linear regime gas can collapse to scales smaller than this and the jeans scale becomes a somewhat ambiguous quantity . here
we define an effective non - linear jeans length , @xmath13 , as the width of a kernel of the form @xmath42^{-2}$ ] , such that the rms fluctuation amplitude of @xmath39 is the same as that of the unsmoothed dark matter density filtered with this kernel ( see section 4 for details ) . on scales larger than the effective jeans length ,
equation ( [ tau ] ) can be written as @xmath43 \dd x , \label{tauw}\ ] ] with @xmath44 where @xmath45 is the baryonic density in terms of the critical density and @xmath22 is the photoionization rate per hydrogen atom .
the doppler parameter in the last equation depends on @xmath46 as @xmath47 .
nh99 have presented a direct lucy - type iterative scheme to recover the optical depth and the corresponding mass and velocity fields in the los from the normalized flux , @xmath48 .
they showed that the density field can be successfully recovered below a threshold value above which the corresponding flux saturates .
nh99 further used mock spectra extracted from an n - body simulation to show that the moments of the mass density can be reasonably well estimated by fitting a log - normal pdf to the low density tail of the pdf of the recovered density . in the next section we present a modified method for estimating the moments of the density distribution .
nh99 provided an estimate for the quantity @xmath49^{\alpha } , \label{lod}\ ] ] which we term the local optical depth .
it is related to the observed optical depth @xmath50 by a convolution with a voigt profile as described in equation ( [ tau ] ) .
the recovered @xmath51 is a good approximation to the true field only in regions with @xmath51 less than a certain value . for large optical depths
, the recovered @xmath51 typically underestimates the true field .
we define a truncated local optical depth @xmath52 as @xmath53 for @xmath54 , and @xmath55 otherwise .
the corresponding truncated moments of @xmath56 can be written in terms of the density pdf as , @xmath57 where @xmath58 is the density contrast and @xmath59 .
we further define @xmath60/\mu_2 $ ] , where @xmath61 and @xmath62 are the average and @xmath1 values of @xmath63 .
nh99 had also shown that the pdf of the dm density smoothed on the scale relevant for the ly@xmath19 forest ( the effective jeans scale ) can be reasonably well approximated by a log - normal distribution ( bi & davidsen 1997 , cf .
sheth 1998 and gaztanaga & croft 1998 for different forms of the pdf ) . for a log - normal density distribution , @xmath64 , the truncated moments in ( [ trunc ] ) can be written as , @xmath65 \label{taumom}\ ] ] where @xmath66 is the value of @xmath67 corresponding to @xmath68 . by expressing @xmath67 in terms of @xmath51 in ( [ taumom ] )
we find @xmath69 .
\label{taumomfin}\ ] ] we have here used the relation @xmath70 which follows from the condition @xmath71 for the log - normal .
the moments of the truncated optical depth distribution depend on four parameters @xmath72 , @xmath62 , @xmath19 and @xmath73 .
the parameter @xmath73 is chosen such that for @xmath74 the local optical depth does not suffer from the biases introduced in saturated regions . as apparent from equation ( [ taumomfin ] ) there are two basic degeneracies leaving us with two independent parameters : @xmath75 in terms of these two parameters , the moments of @xmath52 can finally be written as @xmath76 .
\label{taumomfin}\ ] ] the first two moments , @xmath77 and @xmath78 , are therefore sufficient to determine the parameters @xmath79 and @xmath80 . from these we can then infer the @xmath1 fluctuation amplitude of the dm density and the normalization constant of the optical depth as discussed in section 3.3 .
the use of truncated moments of the local optical depth has a number of advantages .
it is numerically more stable than fitting a parametric form of the pdf to the distribution of the recovered density .
it allows to introduce a cut - off in the form of an optical depth instead of the density which depends on the a priori unknown value of .
finally , it shows the basic degeneracies between @xmath19 , @xmath62 and . in reality
the pdf of the mass density smoothed on an effective jeans scale is expected to be somewhat more positively skewed than a log - normal .
the relations derived above can easily be generalized to other forms of the density pdf .
we were , however , not able to constrain the additional skewness parameter of an edgeworth expansion .
the same was found by nh99 .
we therefore used a log - normal pdf for the analysis of the data .
we assess the possible biases that arise in the determination of and using mock spectra in the next section .
we had two spectra of q1422 + 231 at @xmath81 available for our analysis .
the spectra were observed with the hires spectrograph by ( 1997 ) and ( 1996 ) on the keck i telescope .
the resolution of the spectra is @xmath82 and the pixel size is @xmath83 .
we used the wavelength region from 4742 (ly@xmath84 at the redshift of the qso ) to 5618 (ly@xmath19 at the redshift of the qso ) to avoid confusion due to an overlap of intervening ly@xmath19 and ly@xmath84 absorption .
this corresponds to a redshift range from @xmath85 to @xmath81 and is equivalent to @xmath86 ( comoving ) in an einstein - de sitter universe .
the two spectra were continuum - fitted independently by the observational groups .
the differences in our results should thus give an idea of the error due to continuum - fitting .
two sources of errors are important for our analysis : first the intrinsic error caused by the inversion procedure ; second the cosmic variance error resulting from the use of only one quasar spectrum with limited redshift range .
we quantify these errors and possible systematic biases by analyzing mock spectra .
the mock spectra were produced using a modified version of the method suggested by bi et al .
( 1992 ) [ see also bi & davidsen ( 1997 ) ] .
the mock spectra have the same redshift range , pixel size , resolution and s / n as the observed spectra of q1422 .
we generated random realizations of a gaussian density field with a linear cdm - like power spectrum .
we then assumed a parametric form for the pdf of the non - linear density and obtained the non - linear density field along the los from the linear field by a rank - ordered mapping ( weinberg 1992 ) .
the corresponding non - linear peculiar velocity , @xmath87 , was computed as described in nh99 .
the optical depth was calculated using equation ( [ tauw ] ) with @xmath88 and @xmath89 .
the factor was chosen such that the mean flux decrement in the mock spectrum matches the observed value .
finally , instrumental broadening as well as photon and pixel ( read out ) noise was added . based on numerical simulations
, we expect the true pdf to be more positively skewed than a log - normal .
our method for estimating and assumes , however , that the pdf is log - normal .
it is therefore important to test the sensitivity of the estimated moments to the shape of the true pdf .
we generated mock spectra for a log - normal and for pdfs described by an edgeworth expansion around the log - normal , @xmath90 , \label{edge}\ ] ] where the skewness parameter is defined as @xmath91 .
n - body simulations with gaussian initial conditions have shown this parameter to be positive , independent of the detailed shape of the initial power spectrum ( colombi 1993 , nh99 ) .
we will present detailed results for two different pdfs with @xmath92 and @xmath93 , but we have also tested the technique with @xmath94 . in figure
[ fig : pdf_rho ] we show representations of the pdf with @xmath95 .
the deviations of the pdf of @xmath96 from a log - normal distribution with the same @xmath62 are significant only for @xmath97 .
since only the low density part of the pdf is relevant for the truncated moments , the method is expected to produce reasonable estimates of @xmath62 even if the true pdf is somewhat skewed relative to log - normal .
we have generated mock spectra for both pdfs with and without peculiar velocities to test the effect of redshift distortions .
for each case 70 mock spectra ( i.e. 70 random realizations for the same density power spectrum ) were analyzed .
we have also modified the inversion procedure by introducing a power law relation between the doppler parameter and the local optical depth , @xmath98 .
this has significantly speeded up the convergence of the inversion . to determine @xmath99 we define the function @xmath100 , where @xmath101 is the error and @xmath102 the observed flux in pixel @xmath103 .
@xmath104 is the flux which corresponds the recovered local optical depth .
we choose @xmath99 to be the largest value which gives @xmath105 after a certain number of iterations . in fig .
[ fig : pdf_tau_mock ] the average @xmath106 $ ] of the recovered local optical depth of 70 mock spectra is shown ( for the case with @xmath95 ) in real ( solid curve ) and redshift space ( dashed curve ) .
the dotted line represents the average true pdf .
the error bars are @xmath107 cosmic variance errors calculated from the 70 random realizations .
the spikes in the pdfs of the recovered local optical depth are due to the systematic underestimate in saturated regions .
figure [ fig : pdf_tau_mock ] demonstrates that we have successfully corrected for redshift distortions in regions of low optical depths .
this correction shifts the location of the prominent spike to lower @xmath51 because the redshift distortions tend to enhance the real space density contrast .
we use the location of the spike to choose the cut - off @xmath73 for the truncated moments .
we have calculated the first two moments @xmath108 and @xmath109 for three cutoff values , @xmath110 , @xmath111 and @xmath112 .
the parameters and were determined by solving equations ( [ taumomfin ] ) with broyden s method ( e.g. press et al 1992 ) .
figures ( [ fig : parms_nv ] ) and ( [ fig : parms ] ) show the true _ vs _ the recovered values of the parameters and for four sets of 70 mock spectra ( two input pdfs , with and without peculiar velocities ) .
the tight correlation between the true and recovered values demonstrates robustness of the method .
the estimated values of are unbiased in all four cases while is systematically underestimated ( except in the log - normal case with peculiar velocities set to zero ) . we found similar results for the mock spectra produced with(@xmath113 , @xmath114 ) .true and estimated parameters for the mock q1422 + 231 spectra generated with a log - normal pdf of the mass density
listed are estimated parameters from mock spectra with and without peculiar velocities @xmath115 .
the first entry shows the parameters estimated after correcting for redshift distortions , the second before the correction and the third the parameters from spectra generated with @xmath116 .
the first and second rows are @xmath117 and the @xmath107 error due to the the inversion .
the fourth entry lists the true values of the parameters and the @xmath107 error due to cosmic variance . the transformation from and to is given by @xmath118 and requires a value for @xmath19 .
[ cols="<,^,^,^ " , ] [ tab : real ] for a given value of @xmath19 we can now infer values of @xmath15 and from our estimates of and .
the values of given in table ( [ tab : real ] ) are derived using @xmath119}.\ ] ] since is systematically biased low , the values of in the table should be considered as lower limits .
assuming that the true is biased low by a factor of 2 as suggested by our analysis of the mock spectra with @xmath120 we estimate @xmath121 ( @xmath122 for @xmath88 ) .
we now discuss the inferred value of @xmath123 . for a pdf described by an edgeworth expansion ( equation [ edge ] ) , @xmath15 and @xmath124)^2>^{1/2}$ ]
are related by @xmath125 we have demonstrated in the previous section section that the estimates of are unbiased even for true pdfs with rather large skewness parameter @xmath126 .
the value of @xmath127 is therefore also unbiased .
hence by assuming @xmath128 we somewhat overestimate @xmath15 . for the values of @xmath62 and @xmath126 measured in numerical simulations ( nh99 ) we expect this bias to be less than 15% , significantly smaller than the uncertainty in the value of @xmath19 .
we therefore use the relation @xmath129}-1}.\ ] ] note that the small bias introduced here will also somewhat lower our estimates for @xmath8 in the next section .
possible values of @xmath19 range between 1.56 and 2 but @xmath19 is probably about 1.7 at @xmath130 for plausible reionization histories ( hui & gnedin 1997 , abel & haehnelt 1999 ) . the estimated value of @xmath131 for the rauch et al .
spectrum translates into @xmath132,@xmath133 , and @xmath134 .
the values for the songaila & cowie spectrum are 20 percent higher .
there is a whole family of cdm - like power spectra which is consistent with a variety of observational constraints on the dm clustering amplitude .
if we normalize these power spectra to match the local space density of galaxy clusters ( e.g. , eke , cole & frenk 1996 ) then , for a given @xmath8 , we can predict the fluctuation amplitude on the scale and redshift relevant to the estimate of @xmath15 in the previous section . in the following
we obtain constraints on @xmath8 by comparing our measurement of @xmath15 with such model predictions .
we start with the _ linear _ mass power spectrum as given by bond , eftsathiou , frenk & white ( 1993 ) @xmath135^{2/\nu } } , \label{pk}\ ] ] where the wavenumber @xmath136 is in units of @xmath137 , @xmath138 is a normalizing factor , @xmath139 , @xmath140 , @xmath141 , and @xmath142 . for standard cold dark matter models
the constant @xmath17 is identified with @xmath143 .
the space density of present - day galaxy clusters is reproduced if @xmath138 is chosen such that the linear @xmath1 value , @xmath144 , of the density fluctuations in a sphere of radius @xmath145 is given by @xmath146 for flat and open models , respectively ( eke , cole & frenk 1996 ) .
we then use an analytic transformation from the linear to the _ non - linear _ power spectrum ( peacock & dodds 1996 ; see also jain , mo & white 1995 ) .
in section 2 we introduced an effective jeans scale @xmath13 such that the fluctuation amplitude of the dm density field smoothed on this scale can be computed from the non - linear power spectrum as @xmath147^{-2}.\ ] ] even though @xmath13 should be related to the linear jeans scale , the exact choice of @xmath13 is not obvious and we have left @xmath13 as a free parameter .
the nonlinear @xmath15 depends then on the shape parameter @xmath17 , the background density , @xmath8 , the cosmological constant , @xmath148 , and the effective jeans scale , @xmath13 . in figure ( [ fig : modcomp ] ) we show the model prediction for @xmath15 at @xmath149 as a function of the effective jeans scale @xmath13 for different cosmological parameters .
the value of @xmath15 inferred from the rauch et al .
spectrum is shown as the open symbols with error bars for @xmath150 , @xmath151 and @xmath112 .
three factors contribute to the rather strong @xmath8 dependence of the predicted @xmath15 : the power spectrum normalization , the time evolution of the fluctuation amplitude and the relation between length and velocity scale .
there is a similar but somewhat weaker dependence on @xmath148 . for a given @xmath152
, we define the function @xmath153 ^ 2/\sigma^2_{\rm err}$ ] where the @xmath154 is the total @xmath107 error including the uncertainty in the power spectrum normalization ( [ sig81 ] ) .
figure ( [ fig : modchi_all ] ) is a contour plot of @xmath155 as function of @xmath8 and @xmath13 for flat ( @xmath156 ) and open models ( @xmath12 ) .
if we assume @xmath157 as inferred from the shape of the power spectrum of the observed galaxy distribution at @xmath158 ( peacock & dodds 1994 ) our estimate for @xmath8 can be expressed in a convenient parametric form as follows : @xmath159 for a flat universe , and @xmath160 for an open universe without a cosmological constant .
the errors on @xmath8 are 25% .
the inferred values from the spectrum have the same parametric form but are lower by 15% .
_ the value of _ rauch et al .
have used several hires spectra including the one of q1422 + 231 used here to infer the value of @xmath161 by comparing the flux decrement distribution of the observed spectra to those of mock spectra calculated from hydrodynamical simulations .
this analysis was basically equivalent to a determination of the optical depth constant .
the two quantities are related as @xmath162 rauch et al obtained @xmath163 and @xmath164 for mock spectra produced from a scdm model ( hernquist et al .
1996 ) and @xmath165cdm model ( cen et al . 1994 ) .
the temperatures at @xmath166 were @xmath167 and @xmath168 .
this corresponds to @xmath169 and and @xmath112 , respectively .
no correction for the temperature evolution in the simulations has been made .
this is in reasonable agreement with our estimate of @xmath170 but we would like to stress that our value includes a rather large correction of a factor two for the bias in which we have inferred from our analysis of mock spectra produced with a pdf with @xmath120 .
we therefore do not think that we can currently improve the constraints of rauch et al . on the baryon density and the flux of ionizing photons .
the fact that we obtain a consistent result with a completely different procedure is nevertheless very encouraging . our value is larger than the @xmath171 in the fiducial model of bi & davidsen ( @xmath172 ) .
_ the value of @xmath15 _ gnedin and hui ( 1998 ) used mock spectra calculated from dm simulations to investigate the influence of the fluctuation amplitude on the slope of the column density distribution of qso absorption lines .
they got reasonable agreement with the observed slope if the linear fluctuation amplitude at half the linear jeans scale for @xmath173 was about 2 .
this is larger than our value for the _ non - linear _ @xmath15 .
this indicates that the effective jeans scale is larger than half the linear jeans scale for @xmath173 .
bi & davidsen have assumed @xmath174 in their fiducial model significantly larger than our value @xmath175 for the same value of @xmath19 .
croft et al . ( 1998 , 1999 ) have compared the flux power spectrum of observed absorption spectra to infer the dm power spectrum on scales between @xmath176 and @xmath177 ( comoving , @xmath178 ) by comparison with mock spectra of dm simulations which were calibrated using hydrodynamical simulations .
note that croft et al .
define their length scales to be a factor @xmath179 larger than we do in our definition of the jeans length .
croft et al obtain @xmath180 at @xmath181 , where @xmath182 is the rms fluctuation amplitude of the dm density contributed by a logarithmic interval of @xmath136 .
this measurement was at a scale which corresponds to @xmath183 .
weinberg et al . ( 1999 ) then used the constraints on the dm power spectrum obtained by croft et al .
( 1999 ) to constrain the value of @xmath8 with similar assumptions for the normalization and the shape of the power spectrum as we have made here . for @xmath184 weinberg et al .
1999 obtain @xmath185 ( @xmath186 ) and @xmath187 ( @xmath188 ) .
we obtain somewhat higher values unless the effective jeans scale is larger than the linear jeans scale ( for a temperature of @xmath189 ) by about a factor two .
there are a number of possible improvements on the analysis presented here . as discussed in section 3.4 the true pdf of the density
is expected to be somewhat positively skewed relative to log - normal , but we were not able to constrain the skewness parameter , @xmath126 . to make progress on this issue an improved model of the pdf which e.g. relates @xmath126 to other parameters would be needed .
there are attempts to determine the value of @xmath19 from the correlation of column density and the lower cut - off of the doppler parameter distribution ( schaye et al .
such a determination will obviously improve our constraints on and @xmath15 .
currently we probe the density pdf up to a rather low overdensity of @xmath190 .
this could be extended to higher densities by incorporating higher order lyman series line into the inversion scheme .
this will tighten the constraints on the density pdf and also extend the redshift range of each spectrum available for the inversion .
the main source of uncertainty in our constraints on @xmath8 is the effective jeans lenght scale @xmath13 .
a calibration of @xmath13 by numerical simulations should help to tighten the constraints on @xmath8 .
it should also be possible to extract information on the shape of the power spectrum on small scales from the correlation function of the truncated real - space optical depth .
we estimated the error of @xmath15 due to cosmic variance to be about 20% for one spectrum like q1422 .
this error could be reduced to 5% by applying the method to 20 quasar spectra .
note that the error estimate was based on mock spectra generated with a scdm power spectrum .
it might be different for other power spectra
. it would obviously also be interesting to probe the redshift evolution of the clustering amplitude .
we have recovered the real space optical depth distribution along the los to the qso q1422 + 231 from two different hires spectra of q1422 + 231 using the inversion technique proposed by nusser & haehnelt ( 1999 ) . despite the two spectra having different continuum
fits , the difference in the recovered optical depth distribution is small .
the recovered optical depth distribution is similar to that expected for a log - normal pdf of the density .
we have used a new technique to estimate the fluctuation amplitude of the gas density and the optical depth constant from the truncated optical depth distribution . assuming that the pdf of the gas density is log - normal we showed that the moments of the distribution depend on two parameters @xmath191 $ ] and @xmath192}$ ] .
this demonstrates that there are fundamental degeneracies between fluctuation amplitude of the gas , the power law index @xmath19 , and the optical depth constant for a log - normal pdf .
these degeneracies will be only very weakly broken if the true pdf is somewhat positively skewed relative to log - normal as expected from numerical simulations .
using mock spectra we have shown that robust estimates for the parameters and can be obtained .
the mean of the recovered is unbiased with respect to the true value .
the value of however , is systematically biased low if the true pdf is positively skewed relative to log - normal .
our estimates of and translate into @xmath193 and @xmath194 if we assume @xmath88 , the most likely value for plausible reionization histories .
the error estimates come from our analysis of the mock spectra .
if the igm indeed traces the dm in the way suggested by simulations this is the first measurement of the fluctuation amplitude on an effective jeans scale which corresponds to about two hundred kpc(comoving ) .
the fluctuation amplitude of the dm density on such small scales is of particular interest for the question when the epoch of reionization occurs .
our rather low estimate for the value of @xmath15 strongly argues for a late epoch of reionization .
finally , we have compared our measurement of the fluctuation amplitude to predictions by cdm - like cosmogonies with a power spectrum shape parameter suggested by the clustering behavior of galaxies , and normalized to match the observed space density of present - day clusters of galaxies .
the fluctuation amplitude of such models depends rather strongly on omega .
we obtain @xmath195 ( @xmath196 ) and @xmath197 ( @xmath12 ) .
if we assume that the effective jeans scale is close to the linear jeans scale ( at 15000k ) then the measured fluctuation amplitude is larger than that predicted for an ( @xmath198 , @xmath199 ) universe by a factor of about two .
this argues either for a larger value of @xmath8 and/or a large value of the effective jeans length @xmath13 .
we are grateful to l. cowie , m. rauch , w. sargent and a. songaila for letting us use their hires spectra of q1422 for the analysis presented here .
we would like to thank t. theuns and s. white for comments on the manuscript .
weinberg d.h , miralda - escud j. , hernquist l. , katz n. , miralda - escud j. , , 1998 , in : structure and evolution of the igm from qso absorption lines , eds .
petitjean p. , charlot s. , editions frontieres , p. 133 | the real - space optical depth distribution along the line of sight to the qso q1422 + 231 is recovered from two hires spectra using a modified version of the inversion method proposed by nusser & haehnelt ( 1999 ) .
the first two moments of the truncated optical depth distribution are used to constrain the density fluctuation amplitude of the intergalactic medium ( igm ) assuming that the igm is photoionized by a metagalactic uv background and obeys a temperature - density relation . the fluctuation amplitude and the power law index @xmath0 of the relation between gas and neutral hydrogen ( )
density are degenerate .
the @xmath1 of the igm density at @xmath2 estimated from the first spectrum is @xmath3}-1}$ ] , with @xmath4 for plausible reionization histories .
this corresponds to @xmath5 with @xmath6 .
the values obtained from the second spectrum are higher by @xmath7 .
if the igm density traces the dark matter ( dm ) as suggested by numerical simulations we have measured the fluctuation amplitude of the dm density at an effective jeans scale of about a hundred to two hundred ( comoving ) kpc . for cdm - like power spectra
the amplitude of dark matter fluctuations on these small scales depends on the cosmological density parameter @xmath8 . for power spectra
normalized to reproduce the space density of present - day clusters and with a slope parameter of @xmath9 consistent with the observed galaxy power spectrum , the inferred @xmath8 can be expressed as : @xmath10 for a flat universe , and @xmath11 for a @xmath12 universe .
@xmath13 is the effective jeans scale in ( comoving ) @xmath14 .
based on a suit of detailed mock spectra the 1-@xmath15 error is @xmath16 .
the estimates increase with increasing @xmath17 . for the second spectrum
we obtain @xmath18 lower values .
h # 1 # 1 # 1#1 # 1#1 # 1#1 cosmology : theory , observation , dark matter , large - scale structure of the universe intergalactic medium quasars : absorption lines |
qcd with massless @xmath4- , @xmath5- and @xmath6-quarks has a chiral su(3)@xmath0 su(3)@xmath1 symmetry which is known to be spontaneously broken .
the scalar density of quarks develops a non - vanishing ground state expectation value , the chiral condensate @xmath7 .
the hellmann - feynman theorem suggests that the magnitude of this condensate decreases with increasing baryon density @xcite at a rate determined , to linear order in density , by the pion - nucleon sigma term . at the density of normal nuclear matter , @xmath8 , the condensate changes from its value at @xmath9 by roughly 30 % .
spontaneously broken chiral symmetry is expected to be restored at high temperatures and possibly at very high baryon densities . at that point
the masses of the ( @xmath10 ) vector mesons should become degenerate with those of their chiral partners , the ( @xmath11 ) axial vector mesons .
one expects that in - medium vector meson spectra should show traces of this tendency already at moderate densities .
in particular , if vector mesons experience a sufficiently strong attractive potential in matter , they could form bound states with ordinary nuclei .
such attraction was suggested by in - medium qcd sum rules @xcite and by a simple scaling hypothesis ( `` brown - rho scaling '' , @xcite )
. more detailed analysis @xcite has revealed that the @xmath12 meson is probably not a good signal since its large vacuum width of about 150 mev will be further magnified by in - medium reactions .
in contrast , the @xmath2 meson appears to be a better candidate @xcite .
its small width in free space ( about 8 mev ) might increase by a factor of five or so in nuclear matter , but this is still within reasonable limits so that the @xmath2 meson has a chance to survive as a quasi - particle in the nuclear medium . at the same time
the calculations suggest that its in - medium mass is shifted downward by about 15% at @xmath13 @xcite , an amount consistent with the original qcd sum rule analysis @xcite .
while much of the recent activities have focused on possible in - medium modifications of vector mesons spectra in the context of ultra - relativistic heavy - ion collisions , it is certainly useful to explore such effects under less extreme , better controled conditions : hence the quest for possible @xmath2 meson ( quasi- ) bound states in ordinary nuclei .
similar questions have been raised in refs .
@xcite .
our paper proceeds as follows . in section 2
we present an update of the @xmath14 interaction and the in - medium spectrum of the @xmath2 meson .
section 3 focuses on the @xmath2-nuclear optical potential and its quasi - bound states . in section 4
we explore possible mechanisms for producing such states : the @xmath3 reaction and pion - induced @xmath2 production . concluding remarks follow in section 5 .
the framework of our discussions is an effective lagrangian , based on chiral su(3)@xmath0 su(3)@xmath1 symmetry , which incorporates the octet of pseudoscalar goldstone bosons ( pions , kaons , ... ) , baryons and vector mesons @xcite .
it specifically includes anomalous couplings of the @xmath2 meson related to the wess - zumino term . in ref .
@xcite we have calculated in - medium spectra of @xmath15 and @xmath16 mesons using this framework . for the @xmath2 meson ,
the basic input is the @xmath2-nucleon amplitude in free space , @xmath17 , as a function of energy @xmath18 at zero three - momentum , @xmath19 . restricting the discussion to @xmath19
is justified since @xmath2 mesons bound to nuclei will have small average momenta , @xmath20 . in comparison with our previous results for @xmath21 presented in ref .
@xcite , improvements have recently been made by more rigorous inclusion of constraints from the measured @xmath22 cross section @xcite .
we have also examined uncertainties arising from the incompletely known ( off -shell ) behaviour of @xmath17 at large @xmath18 .
we have performed further calculations of @xmath23 as specified in ref .
@xcite where all the technical details can be found , but using two different input sets , a and b @xcite .
these sets differ in their treatment of @xmath24 processes involving intermediate @xmath25 states , as illustrated in fig .
[ diag5 ] . set a includes the box diagrams , figs .
[ diag5]a and b. set b inorporates in addition the diagram , fig .
[ diag5]c , in which the @xmath25 interaction vertex has large tensor coupling ( @xmath26 ) . fig .
[ fig1a]a shows the real and imaginary parts of @xmath27 calculated with sets a and b , respectively . the imaginary parts are evalutated as described in ref .
@xcite ( choosing the sign of the @xmath28 coupling to be consistent with the constraints from @xmath29 ) , and the real parts are obtained using the dispersion relation ( [ eq2 ] ) .
the differences between the solid and dashed curves ( sets a vs. b ) give a rough impression of the model dependence of @xmath23 at high energy due to uncertainties in the @xmath30 coupled channel dynamics .
0.32 mm ( 400,150)(0,0 ) ( 10,15)(30,15 ) ( 30,15)(70,15 ) ( 70,15)(90,15 ) ( 30,15)(30,55 ) ( 70,15)(70,55 ) ( 30,55)(70,55)4 ( 50,15)(20,0,180)4 ( 10,75)(30,55)33 ( 70,55)(90,75)33 ( 50,0)[](a ) ( 20,40)[]@xmath12 ( 80,40)[]@xmath12 ( 50,65)[]@xmath31 ( 50,40)[]@xmath31 ( 20,76)[]@xmath2 ( 0,10)[]@xmath32 ( 30,15)2 ( 70,15)2 ( 30,55)2 ( 70,55)2 ( 160,15)(180,15 ) ( 180,15)(220,15 ) ( 220,15)(240,15 ) ( 180,55)(220,55 ) ( 180,15)(180,55)4 ( 220,15)(220,55)4 ( 160,75)(180,55)33 ( 220,55)(240,75)33 ( 205,0)[](b ) ( 150,10)[]@xmath32 ( 205,75)[]@xmath12 ( 170,35)[]@xmath31 ( 230,35)[]@xmath31 ( 150,78)[]@xmath2 ( 180,15)2 ( 220,15)2 ( 180,55)2 ( 220,55)2 ( 310,15)(330,15 ) ( 330,15)(420,15 ) ( 420,15)(440,15 ) ( 310,35)(330,15)33 ( 420,15)(440,35)33 ( 375,15)(25,0,180 ) ( 375,0)[](c ) ( 375,50)[]@xmath12 ( 325,40)[]@xmath2 ( 300,10)[]@xmath32 ( 330,15)2 ( 420,15)2 ( 350,15)2 ( 400,15)2 we note that @xmath23 is strongly energy dependent , its real part changing from repulsion at @xmath33 to attraction at @xmath34 . the imaginary part rises quickly at energies @xmath35 .
it represents the sum of the @xmath36 , etc .
reaction channels ( see fig . [
fig1a]b ) .
the @xmath37 channel is actually the prominent one for @xmath38 , while the relatively small @xmath39 reaction width determines @xmath40 at @xmath41 .
our calculated effective scattering length ( with set a ) , @xmath42 suggests a substantial downward shift of the in - medium @xmath2 mass and a moderately increasing width in matter ) are smaller than those in ref .
@xcite where the @xmath43 reaction width had been overestimated .
this has now been corrected . ] .
this downward shift comes in large part from @xmath44-exchange processes such as fig .
1(b ) which , at the same time , have large inelasticities from @xmath45 .
remaining uncertainties concern the constant @xmath46 which enters into the dispersion relation @xmath47 used to derive @xmath48 from the calculated @xmath49 .
vector meson dominance ( vmd ) connects @xmath50 with the thomson limit of the compton scattering amplitude , so that @xmath51 with the nucleon mass @xmath52 and the universal vector coupling constant @xmath53
. deviations from vmd could arise from additional contact terms proportional to @xmath54 in the effective lagrangian , so that the constant @xmath55 remains basically unconstrained . nevertheless , a first order estimate for the @xmath2 meson mass shift in nuclear matter gives , using eq.([eq1 ] ) : @xmath56 which is consistent with the in - medium qcd sum rule analysis @xcite .
the in - medium qcd sum rule does not have quantitative predictive power because of uncertainties related to the treatment of the density dependence of four - quark condensates , but it is still useful for orientation .
it provides a model independent consistency test for the first moment of the @xmath2 meson spectral distribution @xcite , in vacuum as well as in nuclear matter .
the detailed evaluation @xcite of the in - medium @xmath2 meson spectrum involves the self - energy tensor , @xmath57 , of the @xmath2 meson in nuclear matter at density @xmath12 .
for vanishing momentum @xmath58 its longitudinal and transverse parts @xmath59 coincide , and we introduce @xmath60 ( 150,75 ) ( 0,0 ) ( 80,0 ) at @xmath9 , the imaginary part of @xmath61 gives the vacuum decay width of the @xmath2 meson , and its real part is absorbed into the physical free mass @xmath62 .
the density dependent part of @xmath61 involves the @xmath2-nuclear interaction which is treated to leading order in the density @xmath12 .
the in - medium propagator for the @xmath2 meson `` at rest '' , @xmath63 , reduces to : @xmath64 with @xmath65 at @xmath66 and the complex @xmath67 amplitude @xmath23 as given in fig .
[ fig1a](a ) .
note that @xmath68 represents the @xmath69 decay and grows even faster with energy @xmath18 than the @xmath70 phase space .
the spectral distribution @xmath71 at nuclear matter density @xmath72 is shown in fig .
[ fig1b ] in comparison with the vacuum ( @xmath9 ) spectrum .
the solid and dashed curves at @xmath13 reflect the model dependence in the calculations . despite such uncertainties , the following features are clearly visible : * the @xmath2 meson ( unlike the @xmath12 meson ) is expected to persist as a quasi - particle in nuclear matter when it has a small momentum relative to the surrounding medium . * the in - medium @xmath2 meson mass at @xmath13 is shifted downward by about 15 % from the free @xmath62 .
* the predicted in - medium @xmath2 meson decay width ( at resonance ) is @xmath73 mev , about five times as large as the free width , but still about an order of magnitude smaller than @xmath62 .
( 160,75 ) ( 40,0 ) we repeat that our calculated @xmath2 spectrum in matter is perfectly consistent with the qcd sum rule analysis @xcite .
in fact , the first moment of this spectral distribution ( which does not depend on the four - quark condensate but only on the gluon and quark condensates together with the first moment of the quark distribution known from deep inelastic lepton - nucleon scattering ) follows precisely the trend expected from qcd sum rules @xcite .
we use our predicted spectrum as basis for the discussion of possible @xmath2 meson quasi - bound states .
the term `` quasi - bound '' should indicate that such bound states , if existent , will have short lifetimes due to their hadronic decays , but the widths are not overwhelmingly large according to our calculated @xmath74 .
in the local density approximation , the @xmath2 meson self - energy ( [ eq4 ] ) can be translated into a complex , energy dependent local potential : @xmath75.\end{aligned}\ ] ] notably , @xmath76 can reach values of order @xmath77 mev in the center of nuclei and at the energies in question . non
- local terms proportional to @xmath78 are expected to be less important and will be dropped in the present survey calculation .
given that the longitudinal and transverse parts of the self - energy are the same in the limit @xmath79 , all four vector components of the @xmath2 meson field are described by a single wave function @xmath80 which satisfies the klein - gordon equation @xmath81 \phi ( \vec{r}\,)=0.\ ] ] quasi - bound @xmath2-nuclear states are defined by self - consistent solutions of eq .
( [ eq7 ] ) at complex energies @xmath82 for which @xmath83 .
we introduce @xmath84 and refer to @xmath85 as the binding energies of these states .
their total decay widths are given by @xmath86 .[tab1 ] complex eigenenergies , @xmath87 , obtained by solving the klein - gordon equation ( [ eq7 ] ) with the potential ( [ eq6 ] ) and density ( [ eq10 ] ) using set a as input . [ cols="^,^,^,^,^ " , ] for the nuclear density distributions we use two - parameter fermi profiles @xmath88 with @xmath89 fm and @xmath90 fm .
the normalization constant @xmath91 is determined by @xmath92 . in table 1
we summarize the complex eigenenergies found by solving eq.([eq7 ] ) for selected ( somewhat unusual ) light and medium weight nuclei .
the reason for their choice is that these nuclei can be formed in @xmath3 reactions as in the gsi proposal @xcite .
the calculations have been done with spherically symmetric density distributions ( [ eq10 ] ) , so that the quasi - bound @xmath2 meson states are labeled by their principal quantum number and their angular momentum , @xmath93 .
one observes that the attractive potential ( [ eq6 ] ) supports quasi - bound omega meson states even for light nuclei .
the examples given in table 1 should be representative for a wider class of nuclei , and we do not expect that detailed nuclear structure effects such as deformations change this picture qualitatively .
these quasi - bound states are of course short - lived , with full widths that are generally of similar magnitude as their binding energies .
unless these widths are grossly underestimated , one should be able to see spectral strength at energies below the free @xmath2 mass , even if these states may not be resolved individually .
our predictions in table 1 are in qualitative agreement with the findings of ref .
they differ at a quantitative level since our @xmath2-nuclear potential has the important feature that it is strongly energy dependent .
the strength of the attractive real part of our potential tends to grow with increasing energy in the vicinity of @xmath94 , so that it actually supports more quasi - bound states than the potential of ref .
the energies of the lowest @xmath95 states are nevertheless strikingly similar in both calculations .
the primary point of our approach is that we can offer an estimate , based on explicit calculations , for the widths of such states with realistic constraints from the prominent @xmath67 reaction channels .
our intermediate conclusion is the following : if the @xmath2 meson mass experiences a downward shift in nuclear matter by @xmath96 at @xmath13 , then this effect should be visible already in ordinary , even light nuclei by the appearance of quasi - bound states .
hayano et al .
@xcite suggested to produce bound @xmath2 states through transfer reactions of the type @xmath97 . here
the incoming deuteron with momentum @xmath98 picks up a bound proton in the target nucleus @xmath99 to form an outgoing @xmath100 together with an @xmath2 meson attached to the residual @xmath101 nucleus .
the final @xmath100 energy @xmath102 and momentum @xmath103 is measured in forward direction .
the basic idea is to minimize the momentum @xmath104 transfered to the residual system with its bound @xmath2 meson .
energy and momentum conservation implies that such conditions would ideally be met with a deuteron beam of kinetic energy @xmath105 gev to produce the @xmath2 meson at rest . even for the lower deuteron beam energies available at gsi ( @xmath106 gev ) the situation would still be favourable since the momentum transfers would be around @xmath107 gev , well within the range of the typical of nucleon fermi momentum in the nucleus .
( 70,70 ) ( 0,0 ) the framework of our calculation is the distorted wave impulse approximation .
the @xmath3 differential cross section for @xmath2 meson production is written @xmath108 the free cross section @xmath109 has been measured at saturne @xcite .
its value in the lab frame at @xmath110 is approximately @xmath111 .
the task is then to calculate the response function @xmath112 which describes the removal of a proton from the target nucleus and the formation of the final state , with the @xmath2 meson attached to the residual nucleus , and it includes the distorted waves of the incoming deuteron and outgoing @xmath113 . the energy variable is @xmath114 which roughly equals @xmath115 , the energy of the produced @xmath2 meson plus the binding energy of the removed proton .
the calculation of @xmath112 is performed using a nuclear shell model with realistic single particle wave functions for the bound proton .
its generic form is @xcite @xmath116 where the sum is taken over all possible final states that can be reached at given energy transfer @xmath18 . here
@xmath117 is the @xmath2 meson green function related to the klein - gordon equation ( [ eq7 ] ) .
the amplitude @xmath118 involves the bound proton wave function @xmath119^{[j_p]}.\ ] ] it also involves the distorted waves , @xmath120 of the incoming deuteron and @xmath121 of the outgoing @xmath100 .
given the high energies of these particle , we can use the eikonal approximation and write ( with their momenta @xmath98 and @xmath103 chosen along the @xmath122-axis ) : @xmath123 we introduce @xmath124 and the impact parameter @xmath125 and write the distortion factor @xmath126 as : @xmath127 } .\ ] ] we use @xmath128 and @xmath129 for the deuteron - nucleon and @xmath100-nucleon cross section at typical beam energies around 4 gev . carrying out all necessary spin and angular momentum sums ( including the polarization degrees of freedom of the @xmath2 meson ) the response function reduces , after some algebra , to the following expression : @xmath130 where @xmath131 is the proton radial wave function specified in eq.([eq14 ] ) , @xmath132 denotes the number of protons in the specific orbital , and @xmath133 is the remaining radial green function of the @xmath2 meson for given angular momentum @xmath134 , with @xmath135 and @xmath136 .
it is written in terms of the regular solution @xmath137 and the outgoing solution @xmath138 of the klein - gordon equation and applies both for continuum states and for quasi - bound states .
the wave number is @xmath139 , and bound states are expressed by analytic continuation to imaginary @xmath140 .
furthermore , @xmath141 with legendre polynomials @xmath142 .
note that the @xmath3 transfer reaction takes place predominantly at the edges of the nuclear target , as shown in fig.4 .
the distortion factor causes a suppression of the cross section by typically two orders of magnitude .
substituting the response function ( [ eq17 ] ) into the cross section ( [ eq11 ] ) we can now study the effects of @xmath2-nuclear binding on the missing energy spectrum measured in the transfer reaction .
we choose the case @xmath143 as an example , mainly for simplicity of its proton orbital structure and since this choice is referred to in the gsi proposal @xcite .
the result is shown in fig.5 .
the differential cross section is small , but a systematic downward shift of strength as compared to quasi - free @xmath2 meson production should be nevertheless visible , even in light nuclei .
the effect should be further enhanced in heavier nuclei .
the increased @xmath2 meson width in the nuclear environment prohibits a separate identification of isolated bound states , but the overall downward shift of the @xmath100 missing energy spectrum below the free @xmath2 meson threshold would indicate the presence of quasi - bound states if they exist .
a flat background , not shown here , is expected to come primarily from @xmath12 meson production .
the model dependence of the input @xmath67 amplitude is again illustrated by showing results for sets a and b discussed previously .
the qualitative conclusions are evidently not much influenced by the difference between these choices .
( 140,68 ) ( 0,0 ) ( 78,0 ) fig .
5b shows the decompostion of the quasi - bound spectrum into separate ( @xmath6- and @xmath144-state ) orbitals of the @xmath2 meson coupled to different proton - hole states .
another interesting option is the @xmath145 reaction in nuclei @xcite .
for a neutron emitted in forward direction , a pion with lab . kinetic energy @xmath146 incident on a proton at rest produces a zero momentum @xmath2 meson . for an @xmath2 meson bound to a nucleus
, @xmath147 may be sufficient .
consider the case in which the @xmath2 meson energy @xmath148 is measured by observing its decay @xmath149 , whereas the emitted neutron remains unobserved .
we wish to estimate the differential cross section for this process .
the @xmath122-axis of the lab .
frame is chosen to coincide with the incoming pion momentum .
the cross section is written ( 70,70 ) ( 0,0 ) ( 80,60 ) ( 0,0 ) @xmath150 where the integral is taken over the solid angle of the emitted neutron .
the factor @xmath151 is the branching ratio for the dilepton decay channel of the @xmath2 meson , @xmath152 for the free @xmath2 and possibly up to an order of magnitude smaller for the in - medium decay .
cross section for @xmath153 in free space ( see fig .
7 ) is almost isotropic and in the range 1 - 2 mb @xcite .
one just needs to transform this quantity to the lab frame .
it remains to calculate the response function @xmath154 which now depends on the neutron emission angle .
we follow similar steps as in the previous section [ 4.1 ] .
the ( undetected ) energetic neutron in the final state is treated as a plane wave @xmath155 .
the distorted wave of the incident pion is written @xmath156 with the distortion factor @xmath157 } \equiv d(z , b).\ ] ] we use @xmath158 mb . note that the reaction zone is now concentrated in the front hemisphere of the target nucleus ( see fig .
the final result for the @xmath159 response function becomes : @xmath160 ( 80,68 ) ( 0,0 ) where now ( @xmath161 ) : @xmath162 with @xmath163 , @xmath164 , legendre functions @xmath165 and bessel functions @xmath166 .
the remaining integrals are solved numerically .
we show in fig .
[ o - pi - li ] the expected results for @xmath167 as an example , for the case in which the @xmath2 spectrum is detected through @xmath168 pairs .
we use set a for the @xmath67 input amplitude .
the strong downward shift of the spectral weight in the presence of the @xmath2 binding potential is quite apparent , and there are even indications of structures that can be associated with separate quasi - bound orbits of the @xmath2 meson .
however , the rates are very low .
in fact we have used the free @xmath169 branching ratio for normalization , and one expects to loose about an order of magnitude when the in - medium branching ratio is used .
we have investigated the possible formation of quasi - bound @xmath2 meson states in nuclei .
such states are indeed predicted to exist if the @xmath2 meson ( unlike the @xmath12 meson ) survives as a quasiparticle in nuclear matter , and if its in - medium mass is shifted downward by more than 10% from its free mass , @xmath170 gev , as suggested by several models . *
the complex @xmath2-nuclear potential is expected to be strongly energy dependent , its real part changing from repulsion at low energy to attraction ( @xmath171 ) at @xmath172 .
this behaviour is consistent with the qcd sum rule analysis of the first moment of the @xmath2 meson spectral distribution in nuclear matter . *
the estimated widths of @xmath2-nuclear states in a variety of nuclei with @xmath173 are @xmath174 , up to about five times larger than the free @xmath2 decay width
. such widths may prohibit the separate identification of @xmath2-nuclear states in distinct orbitals , but pronounced spectral strength at excitation energies @xmath175 below the quasifree production threshold should be visible .
* we have examined two possible options for producing @xmath2-nuclear states : @xmath3 transfer reactions and pion induced formation .
the @xmath176 process with subsequent detection of the @xmath2 by its decay into @xmath168 is at the borderline of feasibility , with prohibitively low rates .
on the other hand , missing energy spectra in @xmath177 reactions look more promising , with forward differential cross sections at the level of nb/(sr@xmath178mev ) even for the lightest nuclei .
optimal kinematic conditions would be realized for deuteron beam kinetic energies @xmath179 - 10 ) gev .
even at @xmath180 gev the transfered momentum is still well within the range of typical nuclear fermi momenta . * the investigation of spectral distributions of @xmath2 mesons `` at rest '' in normal nuclei would be complementary to the search for medium effects on vector meson spectra in high energy heavy ion collisions .
if such effects are significant at high baryon densities , they should also be seen in less extreme but much better controlled conditions as they are realized in ordinary nuclear systems .
99 e.g. drukarev and e.m .
levin , nucl .
phys . * a 511 * ( 1990 ) 679 ; prog .
* 27 * ( 1991 ) 77 .
cohen , r.j .
furnstahl and d.k .
griegel , phys .
* c 45*(1992)1881 ; phys . rev . * c 45 * ( 1992 ) 1881 .
s. klimt , m. lutz and w. weise , phys .
* b 249 * ( 1990 ) 386 . t. hatsuda and s.h .
lee , phys .
* c 46 * ( 1992 ) r34 .
brown and m. rho , phys .
* 66 * ( 1991 ) 2720 .
m. herrmann , b.l .
friman and w. nrenberg , nucl .
phys . * a 560 * ( 1993 ) 411 ; g. chanfray and p. schuck , nucl .
phys . * a 555 * ( 1993 ) 329 ; g. chanfray , r. rapp and j. wambach , phys .
* 76 * ( 1996 ) 368 .
f. klingl , n. kaiser and w. weise , nucl .
phys . * a 624 * ( 1997 ) 527 ; f. klingl and w. weise , hep - ph/9802211 , acta phys .
polonica ( 1998 ) , in print .
hayano , s. hirenzaki and a. gillitzer , preprint nucl - th/9806012 ( 1998 ) ; r.s .
hayano et al . , gsi - sis proposal # 214 ( 1997 ) .
k. tsushima , d.h .
lu , a.w .
thomas and k. saito , preprint adp-98 - 28/t302 ; nucl - th/9807028 ( 1998 ) , phys . lett .
b ( in print ) .
f. klingl , n. kaiser and w. weise , z. phys . * a 356*(1996)193 ; n. kaiser , t. waas and w. weise , nucl . phys . * a 612 * ( 1997 ) 297 .
b. friman , nucl - th9801053 , proc . of the apctp - workshop on astro - hadron physics , seoul , korea , october 1997 .
f. klingl , dissertation , tu mnchen ( 1998 ) . f. klingl and w. weise , in preparation .
r. wurzinger et al . , phys .
* c 51 * ( 1995 ) r443 .
o. morimatsu and k. yazaki , nucl .
phys . * a 483 * ( 1988 ) 493 ; nucl .
phys . * a 435 * ( 1985 ) 727 ; j. hfner , s. y. lee and h. a. weidenmller , nucl.phys . * a 234*(1974)42 ; c.b .
dover , l. ludeking and g.e .
walker , phys .
* c 22 * ( 1980 ) 2073 .
h. karami et al . , nucl .
* b 154 * ( 1979 ) 503 ; a. baldini , v. flaminio , w.g .
moorhead and d.r.o .
morrison , landolt brnstein * vol .
12a * , ed .
h schopper ( springer , berlin , 1988 ) .
w. schn , h.bokemeyer , w. koenig and v. metag , acta phys .
polonica * b 27 * ( 1996 ) 2959 . | models based on chiral su(3)@xmath0 su(3)@xmath1 symmetry and vector meson dominance suggest an attractive potential for the @xmath2-meson in a nuclear medium .
we discuss the feasibility of producing nuclear bound states of @xmath2-mesons using @xmath3 and pion induced reactions on selected nuclear targets . |
fundamental processes , such as transcription and replication require a transient , local opening of the dna double helix @xcite .
bubbles _ occur spontaneously under physiological conditions @xcite , while complete melting and the separation of the two complementary strands requires temperatures around @xmath2c @xcite .
bubbles have been implicated as an explanation for high cyclization rates in short dna fragments @xcite , but their main interest lies in biology and in the physical mechanism underlying the functioning and control of transcription and replication start sites : the stability of dna is sequence - dependent @xcite and opening is strongly influenced by superhelical stress @xcite and the binding of regulatory proteins @xcite .
in particular , benham and coworkers showed significant correlations between the positions of strongly stress - induced destabilized regions and regulatory sites @xcite .
+ several models exist to describe the internal opening of dna .
the peyrard - bishop - dauxois @xcite and the poland - scheraga ( ps ) models @xcite have already been used to quantify bubble statistics @xcite and for the _ ab initio _ annotation of genomes on the basis of correlations between biological function and thermal melting @xcite . here
, we describe the bubble statistics of random and biological sequences at physiological conditions using ( i ) the zimm - bragg ( zb ) model , an efficient approximation of the ps model @xcite , and ( ii ) the benham model @xcite , a generalization of the zb model accounting for superhelical density but neglecting writhe . after exploring the bubble statistics of unconstrained dna using the zb model
, we introduce an asymptotically exact self - consistent linearization @xcite of the benham model as a precise and convenient tool to study the huge impact of superhelicity on local bubble opening .
the numerical efficiency of the method allows us to investigate the bubble statistics for entire genomes and random sequences of sufficient length ( @xmath3 bp ) to obtain statistically significant results for sequence effects on bubble statistics and positioning . in a final step , we correlate the positions of highly probable bubbles within the genome of _
e. coli _ with the position of transcription start sites ( tss ) and start codons .
the most widely - used model to treat the denaturation of dna chains is the ps model @xcite which offers predictive power for thermal melting and strand dissociation for dna of arbitrary length , strand concentration and a wide range of ionic conditions . for long heterogeneous sequences ,
whose local denaturation is dominated by the quenched sequence disorder @xcite , the related and computationally faster zb model gives surprisingly good results @xcite . in the ps and zb formalisms ,
the free energy of a given configuration is decomposed into formation free energies for closed base - pair steps and free energy penalties for the nucleation of unpaired regions ( or bubbles ) . using periodic boundary conditions , the zb hamiltonian for a circular chain of length @xmath4 can be expressed as @xcite @xmath5 where @xmath6 ( @xmath7 ) if base - pair @xmath8 is open ( closed ) .
@xmath9 ( @xmath10 ) is the nearest - neighbor ( nn ) free energy to form the base - pair step @xmath11 .
@xmath12 $ ] is the loop nucleation penalty and depends on the ps cooperativity @xmath13 , on the interacting self - avoiding loop exponent @xmath14 @xcite and on a typical bubble length @xmath15 @xcite ( @xmath16 for @xmath17 ) .
parameters are taken at physiological conditions for temperature @xmath18 c and salt concentration @xmath19=0.1 $ ] m @xcite .
being formulated as a 1d ising model , the model is easily solved analytically ( numerically ) for homogeneous ( heterogeneous ) sequences using transfer matrix methods .
the partition function of the system is then given by @xcite @xmath20,\ ] ] the individual closing probability by @xmath21.\ ] ] in particular , it is possible to calculate position resolved opening probabilities , @xmath22 , for bubbles containing @xmath23 open bps and beginning at the closed bp @xmath8 , by @xmath24\ ] ] with @xmath25 @xmath26 the global closing probability @xmath27 and the probability per bp @xmath28 to observe a bubble of length @xmath23 are given by @xmath29 and @xmath30 .
basic summing rules on the bubble probabilities imply that @xmath31 , and that @xmath32 , @xmath33 where @xmath34 is the average bubble length and @xmath35 is the probability per bp to observe a bubble of arbitrary length . for homogeneous sequences ,
these equations could be solved easily . in the asymptotic limit @xmath36
, it leads to @xmath37^n , \label{eq : homoz}\\ \langle\theta\rangle & = & \frac{1}{2}\left[1+\frac{g-1}{\sqrt{(g-1)^2 + 4s}}\right ] , \label{eq : homozb}\\ p_l & = & \frac{1}{2}\left[\frac{2}{1+g+\sqrt{(g-1)^2 + 4s}}\right]^{l+2 } \nonumber \\ & & \times\left[\frac{2s+g\left(g-1+\sqrt{(g-1)^2 + 4s}\right)}{\sqrt{(g-1)^2 + 4s}}\right ] \label{eq : homop } \end{aligned}\ ] ] with @xmath38 $ ] and @xmath39 $ ] .
the bubble probability is therefore a decreasing exponential function of @xmath23 for homogeneous sequences . for heterogeneous sequences
, we numerically compute the different desired observables in a @xmath40-algorithm . in a first step ,
we iteratively compute the forward and backward products , @xmath41 and @xmath42 .
the second step consists in applying equations [ eq : zbz ] , [ eq : zbthet ] and [ eq : zbpb ] . in living organisms
, dna is highly topologically constrained into circular domains ( closed - loops or circular molecules ) @xcite .
each closed domain is defined by a topological invariant @xmath43 , the so - called linking number .
@xmath43 represents the algebraic number of turns either strand of the dna makes around the other .
it can be decomposed in two contributions : the twist which is the number of turns of the double - helix around its central axis , and the writhe which is the number of coils of the double - helix . in the majority of living systems ,
the average linking number @xmath43 is below the characteristic linking number value @xmath44 of the corresponding unconstrained linear dna , due to a negative superhelical density @xmath45 imposed by protein machineries , where @xmath46 is the linking number difference . per bp as a function of the relative opening @xmath47 in the asymptotic limit , for different stress values :
@xmath48 ( squares ) , -0.03 ( stars ) , -0.06 ( circles ) and -0.1 ( crosses).,width=321 ] in this article , we consider bubble openings in superhelically constrained circular dna using the benham model for an imposed superhelical density @xmath49 , where the standard thermodynamic description of base - pairing ( @xmath50 ) is coupled with torsional stress energetics @xcite . for each state ,
if one neglects the writhe contribution , the imposed linking difference @xmath51 can be decomposed in three contributions : 1 ) the denaturation of @xmath52 base - pairs relaxes the helicity by @xmath53 where @xmath54 bp / turn is the number of base - pair in a helical turn ; 2 ) the resulting single - strand regions can twist around each other inducing a global over - twist of @xmath55 ; 3 ) then , the bending and twisting of double - stranded parts is put in the residual linking number @xmath56 . therefore , due to the topological invariance of @xmath51 , we get the closure relation @xmath57 for denatured regions , the high flexibility of single - stranded dna allows unpaired strands to interwind . the energy associated with the helical twist @xmath58 ( in rad / bp ) of open base - pair @xmath8 is @xmath59 where the torsional stiffness @xmath60 is known from experiments , @xmath61 .
the individual twist @xmath58 are related to the global over - twist @xmath55 via the relation @xmath62 . for
paired helical regions , it has been experimentally found that superhelical deformations induce an elastic energy , quadratic in the residual linking difference @xmath63 where @xmath64 is the number of open base - pairs and @xmath65 . by integrating over the @xmath52 continuous degrees of freedom @xmath58 , the superhelical stress energetics
is represented by a non - linear effective hamiltonian @xcite : @xmath66 ^ 2 \nonumber \\ & & -\frac{1}{2\beta } \log\left(\left[\frac{2\pi } { \beta c}\right]^{n_o}\frac{4\pi^2 c}{4\pi^2 c+k n_o}\right ) \label{eq : qo } \end{aligned}\ ] ] @xmath67 is minimal for a non - zero number of opening base - pairs which increases as the stress strength increases ( see figure [ fig : qeff ] ) .
+ the total effective hamiltonian is given by @xmath68 .
we fix @xmath69 .
the benham hamiltonian is defined in a superhelical density ( @xmath49)-imposed ensemble defined by equation [ eq : she ] . in this section ,
we briefly discuss similar model but in the torque - imposed ensemble .
+ in this ensemble , a linear dna segment ( length @xmath4 ) is constraint by a weak torque @xmath70 applied on base @xmath4 , the first bp being fixed . for each bp @xmath8
, we define @xmath71 its orientation in the plane perpendicular to the average axis of the double - helix and oriented in the 5 to 3 direction ( for a denatured bp - step , @xmath72 of the benham model ) .
the total hamiltonian of the system is then given by @xcite @xmath73 with @xmath74 and @xmath75 the natural helical twist .
writing @xmath76 , and integrating over the @xmath77 , leads to the effective hamiltonian ( relatively to the situation without torque ) : @xmath78\sum_i \theta_i\ ] ] applying the constant torque @xmath70 is therefore equivalent to adding an external field @xmath79 $ ] in the zb formalism . recently , jeon et al @xcite derived formulas to solve the benham model for homogeneous and random heterogeneous sequences , including the computation of sequence - average bubble properties .
the computation is based on a reorganization of the partition function sum into partial sums for fixed numbers of bubbles . for heterogeneous sequences , fye and benham @xcite proposed an exact @xmath80-algorithm by decomposing the prefactor @xmath81
$ ] in discrete fourier modes , and by using the transfer - matrix method .
benham and coworkers @xcite have also developed an approximate method which first involves a windowing procedure to find the minimum free energy and then consider only the states whose energies do not exceed the minimum one by more than a given threshold . at high threshold values or high negative superhelicity ,
the computation time for this algorithm scales exponentially with the threshold and the superhelical stress .
both schemes are still time demanding for very long sequences . in order to speed up the resolution of the benham model and to access directly to position - dependent opening properties of bps _ and _ bubbles , we develop an efficient variational method @xcite allowing us to use the transfer - matrix solution of the zb model . for long sequences , assuming that fluctuations of @xmath52 are small around the value @xmath82 ,
we can expand @xmath83 around @xmath82 .
the approximated effective hamiltonian then takes the typical zb form : @xmath84 where @xmath85 represents the mean - field of our approximation . if one imposes the superhelical stress @xmath70 ( ie , the torque ) instead of the superhelical density , @xmath86 is related to the effective field @xmath87 generated by the torque ( see above ) .
+ in the following , we employ the benham ensemble of an imposed superhelical density , @xmath49 , and determine @xmath86 ( or @xmath70 ) self - consistently .
self - consistency requires @xmath88 , and is equivalent to solving @xmath89 because the function @xmath90 is monotonic . for homogeneous sequences ,
the general solution of the self - consistency equation [ eq : fh ] can not be computed analytically . however , at low temperatures , weak superhelical densities and in the limit of infinitely long chains , a small perturbation development is valid and leads to analytical expressions for @xmath86 . for infinitely long chains , @xmath67 becomes @xmath91 with @xmath92 , @xmath93 and @xmath94/(2\beta)$ ] .
assuming that the fraction of open base - pairs @xmath95 is small ( @xmath96 ) , @xmath97 is given by @xmath98 ^ 2}{a^2 ( 4\pi^2 c)^2}\right)x+o(x^2)\ ] ] using eq.[eq : homozb ] and noting @xmath99 $ ] , we also have @xmath100 in the limit @xmath101 , inserting @xmath102 in eq.[eq : dh ] and solving eq.[eq : fh ] , leads to @xmath103 this expression is valid until @xmath104 ( @xmath105 ) , i.e. @xmath106)}\right)\equiv \sigma_l\ ] ] for @xmath107 , we could write @xmath108 ( with @xmath109 )
. then , @xmath110 and @xmath111 .
solution of eq.[eq : fh ] leads to @xmath112 figure [ fig : approx ] shows that the numerical solution of eq.[eq : fh ] agrees very well with the two expressions found above ( eq.[eq : h1 ] and [ eq : h2 ] ) .
+ for a homogeneous sequence ( @xmath113).,width=321 ] to compute bp or bubbles properties , formulas [ eq : homoz ] , [ eq : homozb ] and [ eq : homop ] are still available if one replaces @xmath114 by @xmath115 and @xmath116 by @xmath117 . for heterogeneous sequences ,
we use the bisection method coupled to the newton - raphson method @xcite to numerically solve the self - consistency equation , for fixed values of the temperature and of the superhelical density . knowing that an evaluation of the function @xmath118 requires one transfer matrix method computation ( @xmath40 each )
, it takes typically @xmath119 evaluations to determine the root with a relative precision of @xmath120 .
this allows numerically efficient computation of denaturation profiles .
for example , computing the local closing probabilities for the _ e.coli _ genome ( @xmath121 mbps ) takes about @xmath122 seconds on a 2.4 ghz computer with the self - consistent method , whatever the density is . on the same computer
, it would take about @xmath123 s with the exact method ( @xmath80 algorithm ) for any @xmath49 values ( interpolation of data given in ref.@xcite ) , and about @xmath124 s with the approximate method for @xmath125 ( and around 40 times more for @xmath126 ) @xcite .
$ ] for the extended neighborhood of the tss of fig .
[ fig : land ] , computed using the benham s web server @xcite ( a ) or using our formalism ( b).,width=321 ] the self - consistent linearization consists in working in the _ torque_-imposed ensemble and determining the torque self - consistently to better approximate the _ superhelical density_-imposed ensemble .
this representation has the advantage of decoupling the opening of different parts of the molecule and is probably the simplest way to express the destabilizing effect of undertwisting on dna stability . in the thermodynamic limit , i.e. , for very long ( genomic ) sequences and at low temperature ( well below the melting temperature ) , we expect our linearization to be a quasi - exact solution of the benham model due to the asymptotic decrease of fluctuations within the system .
for shorter sequences , however , the non - linearity of the effective superhelical hamiltonian @xmath67 ( see eq.[eq : qo ] ) may significantly couple remote domains along the sequence , leading for example , to the closing of an open domain as one increases the superhelical stress ( as observed in fig.[fig : compa]a ) .
although , the self - consistent linearization neglects such effects , it gives a reliable general picture of the superhelically - stressed destabilization of dna sequences . in figure
[ fig : compa ] , we show a comparison of results obtained from the benham web server @xcite and by our method .
the agreement is excellent and the small deviations are mostly due to the slightly different parametrizations of the zb parts and to the absence of finite size effect in our approach as discussed above .
in the following , we discuss results obtained for random homogeneous and heterogeneous sequences , as well as for some bacterial genomes ( _ e.coli _ , _ t. whipplei _ , _ a. baumanii _ , _
b. subtilis _ and _ s. coelicolor _ , see table [ tab : zs ] ) .
the results shown for random heterogeneous sequences were obtained by compiling profiles from 100 random sequences each containing @xmath127 bp . the melting properties of constrained dnas reflect a balance between the two parts of @xmath128 . at @xmath129
c , @xmath50 opposes local opening the more strongly the higher the gc - content of the sequence . in contrast
, @xmath67 is minimal for a finite number of open base - pairs , which increases as @xmath49 becomes more negative . at biological levels , the free energy of completely closed
dna becomes prohibitively large ( see fig.[fig : qeff ] ) . as a consequence
, superhelicity leads to a notable level of base - pair opening at physiological temperatures , where unconstrained dna exhibits negligible breathing . as a function of temperature for an unconstrained ( stars ) or constrained ( @xmath48 : squares , @xmath130 : circles , @xmath131 : crosses ) random sequence ( gc=0.5).,width=321 ] figure [ fig : thetatm ] shows the overall impact on the opening probability of imposing a superhelical constraint . for comparison , we have also included a melting curve for unconstrained dna . at physiological temperatures ,
the unconstrained dna is very stable whereas the superhelicity significantly contributes to opening of base pairs and bubbles .
however , for intermediate and weak negative stresses , the destabilizing effect of an imposed superhelical density is reversed close to the melting temperature due to the overall stabilizing impact of untwisting on the rest of the dna ( @xmath132 ) , resulting in a slowdown of the melting process via a change of @xmath70 ( or @xmath86 ) with temperature .
this effect has already been pointed out for positively - stressed homogeneous molecules @xcite . for stronger stresses ,
the effective twisting hamiltonian @xmath67 at the melting transition ( @xmath133 ) is still negative and then results in a decrease of the melting temperature . a quantification on this salt- , gc - dependent effect is described below .
$ ] ) at different salt concentrations ( @xmath19\in[0.05,1]$ ] m ) as a function of the superhelical density @xmath49 ( dots ) , respectively to the situation with @xmath48 .
( b ) melting temperature at @xmath48 for random sequences as a function of their gc - content , at @xmath19=0.05 $ ] ( circles ) , 0.1 ( squares ) and 0.3 m ( crosses ) , respectively to the corresponding unconstrained ( zb ) situation .
dashed lines represent the fitted relation given in eq.[eq : tmsiggc].,width=321 ] on figure [ fig : salt ] a , we observe that , under the different considered gc - contents and salt concentrations , the melting temperature of random superhelical dna is a quadratic function of @xmath49 , and only the intercept of this function ( @xmath134 ) depends on the gc and on @xmath19 $ ] ( see figure [ fig : salt ] b ) . from a systematic study of the melting temperature @xmath135 as a function of the gc - content and the salt concentration , we fitted the empirical relation ( in @xmath136 c ) : @xmath137,\textrm{gc},\sigma ) & = & 2.35 -0.54\log[na^+]-3.42\textrm{gc } \nonumber \\ & & -0.93\textrm{gc}^2 + 4.7\sigma-46.8\sigma^2\end{aligned}\ ] ] where @xmath138 is the difference in melting temperature between a constrained and an unconstrained random dna polymers .
c of the global opening probability @xmath139 for superhelically constrained random heterogeneous ( full lines ) and homogeneous ( dashed lines ) dnas under different @xmath49 ( from -0.02 to -0.1 every 0.01 : color scale from light to dark red , @xmath130 : blue line ) and for several bacterial genomes at @xmath130 : _ e. coli _ ( triangles ) , _ t. whipplei _ ( squares ) , _ a. baumanii _ ( diamonds ) , _ b. subtilis _ ( circles ) and _ s. coelicolor _ ( stars ) .
inset : proportion of the different contributions to the linking number difference @xmath51 : relaxation due to denaturation @xmath140 ( i ) , over - twist of the denatured regions @xmath55 ( ii ) and residual linking number @xmath56 ( iii ) ( see eq.[eq : she]).,width=321 ] figure [ fig : stresrandt ] shows the dependance of the opening probability @xmath139 as a function of the gc - content .
results are reported for different levels of superhelical density and at @xmath18 c , in the biological relevant range , i.e the overall degree of opening is small yet strongly increased relative to the unconstrained case ( see also fig.[fig : thetatm ] ) .
we have included results for both homogeneous and heterogeneous systems . in the former case , the employed nn - free energies @xmath141 were determined as composition dependent averages over the tabulated step parameters @xcite .
+ passing a threshold @xcite ( see inset in fig.[fig : stresrandt ] ) depending on the gc - content , strong @xmath49-values allow the opening of many bp along the sequences . for a fixed superhelical stress , as expected , @xmath139 is a decreasing function of the gc - content .
differences between homogeneous and heterogeneous are weak , meaning that sequence heterogeneity self - averages and has only a small effect on the total degree of opening .
+ the inset in figure [ fig : stresrandt ] shows the different contributions to the linking number difference as a function of the superhelical density ( see eq.[eq : she ] ) for random sequences with gc=0.5 .
the over - twist @xmath55 contribution is estimated by integrating over the @xmath52 continuous degrees of freedom @xmath58 @xcite and applying the self - consistent linearization : @xmath142 we observe that the over - twist increases linearly with the opening probability .
while for weak stresses , almost all the superhelical energy is stored in the ( residual ) deformations of the double - helix , for strong stresses , more than 50 % of the imposed superhelical constraint is used to drive the local denaturation of bps .
accounting for the writhe in the model should however decrease the contributions due to bp - denaturation and over - twisting since a part of the constraint would be absorbed by coils of the double - helix . to observe a bubble of length @xmath23 .
( a ) for unconstrained molecules : random sequences of different gc - content ( gc from 0 to 1 every 0.1 : color scale from light to dark red , gc=0.5 : squares ) and the genome of _ e. coli _
( dashed line with crosses , gc=0.51 ) .
inset : average bubble length @xmath34 as a function of the gc - content .
( b ) for superhelically - stressed molecules : a random sequence ( @xmath143 , squares ) , the genome of _
e. coli _ ( crosses ) and
a homogeneous sequence ( with @xmath144 equals to the average nn - free energies in _ e. coli _ , circles ) , and @xmath48 ( dotted and dashed lines ) , @xmath145 ( full lines ) and -0.1 ( dashed lines ) .
, width=321 ] figure [ fig : probzb]a shows the bubble probabilities computed with the zb model for random dnas of different gc - content , with @xmath146 consistent with very small loops ( @xmath147 , see inset in fig.[fig : probzb ] a ) .
we remark an exponential decrease of @xmath28 with very short decay lengths , corresponding to fairly closed molecules .
increasing gc - content stabilizes the dna and reduces average bubble lengths .
even accounting for the weak biological sequence effect apparent in our results , the absolute level of bubble opening in unconstrained dna seems too small for natural breathing to play a direct biological role @xcite .
whereas we obtain similar @xmath34-values as reported for the peyrard - bishop - dauxois ( pbd ) model @xcite , absolute probabilities @xmath28 are far lower ( for example @xmath148 ( zb ) versus @xmath149 ( pbd ) for @xmath150 and gc=0.5 ) , mainly due to the absence of a cooperativity penalty factor preventing bubble formation in the pbd model .
figure [ fig : probzb]b highlights the huge impact of the superhelical stress on the bubble statistics . for physiological levels , the opening probability per bp remains significant even for large bubbles and has increased by many orders of magnitude compared to the unconstrained situation .
for example , for gc=0.5 and @xmath130 , @xmath151 and @xmath152 for superhelical dna , while for the corresponding unconstrained dna , we found @xmath153 and @xmath154 . ( a , b ) and of the average bubble length @xmath34 ( c , d ) for superhelically constrained random heterogeneous ( b , d ) and homogeneous ( a , c ) dnas under different stresses @xmath49 ( from -0.02 to -0.1 every 0.01 : color scale from light to dark red , @xmath130 : blue line ) and for several bacterial genomes at @xmath130 : _ e. coli _ ( triangles ) , _ t. whipplei _ ( squares ) , _ a. baumanii _ ( diamonds ) , _ b. subtilis _ ( circles ) and _ s. coelicolor _ ( stars).,width=321 ]
figure [ fig : stresrand ] shows the evolution of the bubble occurence @xmath155 and of the average bubble length @xmath34 for different @xmath49-values and gc - content . with @xmath156 only being a function of the total number of open base - pairs , bubble sizes in homogeneous systems are determined by a competition between the bubble initiation penalty , @xmath157 , favoring the opening of a small number of large bubbles , and entropy , favoring the opening of a large number of small bubbles . in heterogeneous systems , it is possible to lower the fraction of stable gc - steps in the open domains by denaturing a larger number of smaller bubbles in particularly at - rich regions ( see fig.[fig : gcloc ] ) @xcite .
the comparison in figs .
[ fig : stresrand ] and [ fig : gcloc ] shows that the disorder effect dominates in the present case @xcite with the number of bubbles being maximal around gc=0.5 . for biological superhelicities , @xmath158 for random at- and gc - dnas , while @xmath159 for @xmath160 .
sequence - heterogeneity plays an essential role by lowering the fraction of stable gc - steps in the open domains and leads to a _ localization _ of open base - pairs in ( at - enriched ) stress - induced duplex destabilized regions @xcite , whose length in turn limits the bubble sizes . indeed ,
figure [ fig : gcloc ] shows that bubbles appear mainly in at - enriched regions compared to the background gc - content .
interestingly , for intermediate ( biological ) @xmath49 values , the evolution of the gc - composition of bubbles remains flat over a large range of global gc - content .
( from 0 to -0.1 every 0.01 : color scale from light to dark red , @xmath130 : blue line ) .
the dashed line represents results for homogeneous sequences.,width=321 ] compared to random sequences with identical gc - content , bacterial genomes ( _ e.coli _ , _ t. whipplei _ , _
a. baumanii _ , _
b. subtilis _ and _ s. coelicolor _ ) exhibit higher averages degrees of opening and increased average bubble sizes ( symbols in figs.[fig : stresrandt ] and [ fig : stresrand ] ) .
figure [ fig : probzb]b shows striking discrepancies in the bubble distribution @xmath28 between random and biological sequences , with a significant enrichment of large bubbles . in the latter case , the nearly flat distributions resemble those expected for homogeneous sequences and can be viewed as a signature of large stress - induced destabilized domains , which concentrate the dna breathing into large regions with homogeneous opening profiles .
we also note that the level of opening in the biological domains is comparatively insensitive to the precise level of the superhelical density .
.z - scores computed relatively to the average bubble length for 5 prokaryotic organisms . [ cols="<,^,^,^",options="header " , ] in general , biological sequences are more destabilized than random sequences with a same gc - content , leading to less but longer bubbles .
we estimate these differences by computing , for all genomes , the z - score relatively to @xmath34 ( see tab.[tab : zs ] ) .
the z - score is an estimation of the non - randomness of a specific sample . for the average bubble length @xmath34 ,
it can be computed by @xmath161 where @xmath162 and @xmath163 are the mean and the standard deviation of @xmath34 computed on 100 random sequences with the same gc and length as the corresponding genome .
the z - score represents therefore the distance between a specific point and the mean value obtained for random sequences , given in standard deviation units .
z - scores of 20 for @xmath34 suggests the presence of over - represented long at - rich domains in bacterial genomes which have the ability to easily open under superhelical stress .
interestingly , the organism with the lowest z - score ( _ t . whipplei _ ) was shown to exhibit a random - like behavior relatively to the local melting temperature distribution @xcite .
$ ] ) for the neighborhood of a transcription start site ( bp n@xmath136 259382 ) in the _ e. coli _ genome , in the presence ( b , c , d ) or in the absence ( a ) of an imposed superhelical density @xmath164 ( b ) , -0.06 ( c ) and -0.10 ( d).,width=321 ] positions of strongly stress - induced destabilized regions have been shown to correlate with many regulatory regions @xcite including transcription start sites @xcite or origins of replication @xcite . in this section , we focus on bubble positioning inside the promoter regions of the bacterium _ e. coli_. + the transcription of dna is initiated by the local opening of the double - helix at transcription start sites .
figure [ fig : compa ] illustrates their association with strongly stress - induced destabilized regions .
in addition to position - dependent opening probabilities , our approach allows us to calculate the complete bubble free - energy landscape , @xmath165 .
figure [ fig : land ] shows the effect of superhelicity on @xmath166 for the neighborhood of the same tss in _
e. coli_. the analysis reveals that opening is the result of the meta - stable unwinding of a large bubble and not of enhanced small scale breathing .
we note that knowledge of @xmath166 is essential for modeling the dynamics of bubble nucleation and growth @xcite . )
bubble centers included in the regions [ tss-300,tss+300 ] for the 760 studied tss ( black bars ) or for randomly picked regions ( blue line ) .
( c ) probability distribution function of the distance between highly probable bubble centers and the nearest start codons for the 1158 actually found ( black line ) or randomly situated ( blue line ) bubbles.,width=321 ] figure [ fig : tss ] analyses the statistical relation beween tsss and bubbles induced by superhelical stress for the entire genome of _
e. coli_. figure [ fig : tss]a shows a significant and non - random destabilization around tsss , with a maximal opening around @xmath167 .
the same computation using the zb model shows insignificant and orders - of - magnitude smaller opening probabilities .
however , we find a non - random signal around tss-10 corresponding to the position of the at - rich pribnow box , an essential motif to start transcription in bacteria @xcite .
figure [ fig : tss]b gives the relative positions of highly probable bubbles included in tss neighborhoods .
the centers of these bubbles are mainly localized in the [ tss-200,tss ] region where many transcriptional and promoter factors are recruited and bind to dna @xcite .
conversely , figure [ fig : tss]c confirms that the majority of highly probable bubbles are situated upstream and close to start codons of genes .
actually , these bubbles are composed by around @xmath168 of coding bps , significantly lower than the percentage of coding bps in _
( @xmath169 ) .
we have developed a numerically efficient , self - consistent solution of the benham model of bubble opening in superhelically constrained dna .
in particular , we are able to go beyond the calculation of opening probabilities for base pairs and to address the full , position - dependent bubble statistics for entire genomes .
our results indicate , that negative supercoiling leads to ( meta- ) stable unwinding of bubbles comprising @xmath170 base - pairs . in heterogeneous sequences
, bubbles are strongly localized in at - rich domains with sequence disorder dominating the bubble statistics .
as we have shown , large bubbles open with a significantly larger probability in biological sequences , than in random sequences with identical gc - content . in the case of _
e. coli _ , the most likely bubbles are located directly upstream from transcription start sites , highlighting the biological importance of this now well understood , physical property of dna .
99 c. calladine , h. drew , b. luisi , and a. travers , _ understanding dna ; the molecule and how it works _ ( elsevier academic press , 2004 ) .
b. alberts , d. bray , j. lewis , m. raff , k. roberts , and j.d .
watson , _ molecular biology of the cell _ ( garland science , 2002 ) .
d. kowalski , and m. eddy , proc .
usa * 85 * , 9464 ( 1988 ) .
d. kowalski , d. natale , and m. eddy , embo j. * 8 * , 4335 ( 1989 ) . m. frank - kamenetskii , biopolymers * 10 * , 2623 ( 1971 )
cloutier , and j. widom , mol .
cell * 14 * , 355 ( 2004 ) .
j. yan , and j.f .
marko , phys .
* 93 * , 108108 ( 2004 ) .
j. santalucia , proc .
usa * 95 * , 1460 ( 1998 ) .
benham , j. mol .
biol . * 225 * , 835 ( 1992 ) .
fye , and c.j .
benham , phys .
e * 59 * , 3408 ( 1999 ) .
benham , and r.r.p .
singh , phys .
lett . * 97 * , 059801 ( 2006 ) .
sobell , proc .
usa * 82 * , 5328 ( 1985 ) .
t. ambjrnsson , and r. metzler , phys
. rev .e * 72 * , 030901 ( 2005 ) .
h. wang , m. noordewier , and c.j .
benham , genome res .
* 14 * , 1575 ( 2004 ) .
p. ak , and c.j .
benham , plos comp .
* 1 * , e7 ( 2005 )
. h. wang , and c.j .
benham , plos comp .
biol . * 4 * , e17 ( 2008 ) .
m. peyrard , and a.r .
bishop , phys .
* 62 * , 2755 ( 1989 ) .
t. dauxois , m. peyrard , and a.r .
bishop , phys .
e * 47 * , r44 ( 1993 ) .
d. poland , and h.a .
scheraga , j. chem . phys . * 47 * , 1456 ( 1966 ) .
d. jost , and r. everaers , biophys .
j. * 96 * , 1056 ( 2009 ) .
g. kalosakas , k.o .
rasmussen , a.r .
bishop , c.h .
choi , and a. usheva , europhys .
lett . * 68 * , 127 ( 2004 ) .
van erp , s. cuesta - lopez , j .-
hagmann , and m. peyrard , phys .
. lett . * 95 * , 218104 ( 2005 ) .
s. ares , and g. kalosakas , nano lett .
* 7 * , 307 ( 2007 ) .
y. kafri , d. mukamel , and l. peliti , eur .
b * 27 * , 135 ( 2002 ) .
r. blossey , and e. carlon , phys .
e * 68 * , 061911 ( 2003 ) .
t. garel and c. monthus , j. stat .
mech . , p06004 ( 2005 ) .
b. coluzzi , and e. yeramian , eur . j. phys .
b * 56 * , 349 ( 2007 ) . c. monthus , and t. garel , arxiv : cond - mat/0605448v1 ( 2007 ) .
e. yeramian , s. bonnefoy , and g. langsley , bioinformatics * 18 * , 1 ( 2002 ) .
e. carlon , a. dkhissi , m.l .
malki , and r. blossey , phys .
e * 76 * , 051916 ( 2007 ) .
d. jost , and r. everaers , j. phys . : condens .
matter * 21 * , 034108 ( 2009 ) .
vologodskii , a.v .
lukashin , v.v . , anshelevich , and m. frank - kamenetskii , nucleic acids res . * 6 * , 967 ( 1979 ) .
t. garel , h. orland , and e. yeramian , arxiv : q - bio.bm/0407036 ( 2004 ) .
benham , and c. bi , j. comput .
11 * , 519 ( 2004 ) . c. bi , and c.j .
benham , _ proceedings of the 2003 ieee bioinformatics conference _ , ( ieee computer society , 2003 ) . w.h .
press , s.a .
teukolsky , w.t .
vetterling , and b.p .
flannery , _ numerical recipes in fortran : the art of scientific programming _
( cambridge university press , 1992 ) . c. bi and c.j .
benham , bioinformatics * 20 * , 1477 ( 2004 ) .
benham s web - server : + http://benham.genomecenter.ucdavis.edu .
jeon , j. adamcik , g. dietler , and r. metzler , phys .
* 105 * , 208101 ( 2010 ) .
benham , phys .
e * 53 * , 2984 ( 1996 ) .
t. hwa , e. marinari , k. sneppen , and l .- h .
tand , proc .
usa * 100 * , 4411 ( 2003 ) .
t. ambjrnsson , s.k .
banik , o. krichevsky , and r. metzler , biophys .
j. * 92 * , 2674 ( 2007 ) .
s. gama - castro _ et al _ , nucleic acids res . *
36 * , d120 ( 2008 ) .
pribnow , d. , proc .
usa * 72 * , 784 ( 1975 ) . | we present a general framework to study the thermodynamic denaturation of double - stranded dna under superhelical stress .
we report calculations of position- and size - dependent opening probabilities for bubbles along the sequence .
our results are obtained from transfer - matrix solutions of the zimm - bragg model for unconstrained dna and of a self - consistent linearization of the benham model for superhelical dna .
the numerical efficiency of our method allows for the analysis of entire genomes and of random sequences of corresponding length ( @xmath0 base pairs ) .
we show that , at physiological conditions , opening in superhelical dna is strongly cooperative with average bubble sizes of @xmath1 base pairs ( bp ) , and orders of magnitude higher than in unconstrained dna . in heterogeneous sequences ,
the average degree of base - pair opening is self - averaging , while bubble localization and statistics are dominated by sequence disorder .
compared to random sequences with identical gc - content , genomic dna has a significantly increased probability to open large bubbles under superhelical stress .
these bubbles are frequently located directly upstream of transcription start sites . |
this note strengthens and extends several geometric results of the paper @xcite , joint with dmitri orlov , in which we constructed exceptional sequences of maximal possible length on burniat surfaces with @xmath0 .
the construction was based on certain results about the picard group and effective divisors on burniat surfaces . here
, we extend the results about picard group to burniat surfaces with @xmath1 . we also establish a complete description of the semigroup of effective @xmath2-divisors on burniat surfaces with @xmath3 .
( for the construction of exceptional sequences in @xcite only a small portion of this description was needed . )
finally , we construct an exceptional collection on a nonnormal semistable degeneration of a 1-parameter family of burniat surfaces with @xmath0 .
in this paper , burniat surfaces will be certain smooth surfaces of general type with @xmath4 and @xmath5 with big and nef canonical class @xmath6 which were defined by peters in @xcite following burniat .
they are galois @xmath7-covers of ( weak ) del pezzo surfaces with @xmath8 ramified in certain special configurations of curves .
recall from @xcite that a @xmath7-cover @xmath9 with smooth and projective @xmath10 and @xmath11 is determined by three branch divisors @xmath12 and three invertible sheaves @xmath13 on the base @xmath11 satisfying fundamental relations @xmath14 , @xmath15 , @xmath16 .
these relations imply that @xmath17 , @xmath18 , @xmath19 .
one has @xmath20 , where the @xmath21-algebra @xmath22 is @xmath23 .
the multiplication is determined by three sections in @xmath24 where @xmath25 is a permutation of @xmath26 , i.e. by sections of the sheaves @xmath27 , @xmath28 , @xmath29 vanishing on @xmath30 , @xmath31 , @xmath32 .
burniat surfaces with @xmath0 are defined by taking @xmath11 to be the del pezzo surface of degree 6 , i.e. the blowup of @xmath33 in three noncollinear points , and the divisors @xmath34 , @xmath35 , @xmath36 to be the ones shown in red , blue , and black in the central picture of figure [ fig : burniat ] below .
the divisors @xmath37 for @xmath38 are the @xmath39-curves , and those for @xmath40 are @xmath41-curves , fibers of rulings @xmath42 .
the del pezzo surface also has two contractions to @xmath33 related by a quadratic transformation , and the images of the divisors form a special line configuration on either @xmath33 .
we denote the fibers of the three rulings @xmath43 and the preimages of the hyperplanes from @xmath33 s by @xmath44 . ,
height=144 ] burniat surfaces with @xmath45 , @xmath46 are obtained by considering a special configuration in figure [ fig : burniat ] for which some @xmath47 triples of curves , one from each group @xmath48 , @xmath49 , @xmath50 , meet at common points @xmath51 .
the corresponding burniat surface is the @xmath7-cover of the blowup of @xmath52 at these points . up to symmetry , there are the following cases , see @xcite : 1 .
@xmath53 : @xmath54 ( our shortcut notation for @xmath55 ) .
@xmath56 , nodal case : @xmath54 , @xmath57 .
3 . @xmath56 , non - nodal case : @xmath54 , @xmath58 .
@xmath59 : @xmath60 , @xmath61 , @xmath62 .
@xmath63 : @xmath54 , @xmath57 , @xmath64 , @xmath65 .
we generally denote the divisors upstairs by @xmath66 and the divisors downstairs by @xmath67 for the reasons which will become clear from lemmas [ lem : lattices ] , [ lem : lattices2 - 5 ] .
we denote @xmath68 and @xmath69 is the blowup map at the points @xmath51 .
the exceptional divisors are denoted by @xmath70 .
the curves @xmath37 are the curves on @xmath11 , the curves @xmath71 are their strict preimages under @xmath72 .
( so that @xmath73 in the case ( 1 ) , etc . )
the divisors @xmath74 are the curves ( with reduced structure ) which are the preimages of the latter curves and @xmath70 under @xmath75 .
the surface @xmath76 is the burniat surface with @xmath45 . the building data for the @xmath7-cover @xmath75 consists of three divisors @xmath77 , @xmath78 , @xmath79 .
it does _ not _ include the exceptional divisors @xmath70 , they are not in the ramification locus . one has @xmath80 , @xmath81 , @xmath82 , and @xmath83 . for the canonical class , one has @xmath84 . indeed , from hurwitz formula @xmath85 , where @xmath86 .
therefore , the above identity is equivalent to @xmath87 .
this holds on @xmath68 , and @xmath88 for the surfaces with @xmath89 and @xmath90 ( non - nodal case ) , @xmath91 and @xmath92 are ample . for the remaining cases , including @xmath93 , the divisors @xmath91 and @xmath92 are big , nef , but not ample .
each of the curves @xmath94 ( among @xmath37 ) through two of the points @xmath51 is a @xmath95-curve ( a @xmath96 with square @xmath97 ) on the surface @xmath11 .
( for example , for the nodal case with @xmath56 @xmath98 is such a line ) .
its preimage , a curve @xmath99 on @xmath10 , is also a @xmath95-curve .
one has @xmath100 , and the curve @xmath99 is contracted to a node on the canonical model of @xmath10 .
note that both of the cases with @xmath63 and @xmath101 are nodal .
in this section , we recall two results of @xcite . [
lem : lattices ] the homomorphism @xmath102 defines an isomorphism of integral lattices @xmath103 . under this isomorphism , one has @xmath104 .
this lemma allows one to identify @xmath2-divisors @xmath67 on the del pezzo surface @xmath11 with classes of @xmath2-divisors @xmath66 on @xmath10 up to torsion , equivalently up to numerical equivalence .
this identification preserves the intersection form .
the curves @xmath105 are elliptic curves ( and so are the curves @xmath106 , etc . ) .
moreover , each of them comes with a canonical choice of an origin , denoted @xmath107 , which is the point of intersection with the other curves which has a distinct color , different from the other three points .
( for example , for @xmath108 one has @xmath109 . ) on the elliptic curve @xmath108 one also defines @xmath110 , @xmath111 , @xmath112 .
this gives the 4 points in the 2-torsion group @xmath113 $ ] .
we do the same for @xmath114 , @xmath115 cyclically .
[ thm : picx ] [ @xcite , theorem 1 ] one has the following : 1 . the homomorphism @xmath116 is injective , and the image is the subgroup of index 3 of @xmath117 ) \times ( \bz.p_{00 } + b_0[2 ] ) \times ( \bz.p_{00 } + c_0[2 ] ) \simeq \bz^4 \times \bz_2 ^ 6.\ ] ] consisting of the elements with @xmath118 divisible by 3 . here
, we denote an element of the group @xmath119 $ ] by @xmath120 , etc .
@xmath121 induces an isomorphism @xmath122\times b_0[2]\times c_0[2]$ ] .
3 . the curves @xmath123
, @xmath124 , generate @xmath125 .
this theorem provides one with explicit coordinates for the picard group of a burniat surface @xmath10 , convenient for making computations .
in this section , we extend the results of the previous section to the cases @xmath1 .
first , we show that lemma [ lem : lattices ] holds verbatim if @xmath126 .
[ lem : lattices2 - 5 ] assume @xmath127 .
then the homomorphism @xmath128 defines an isomorphism of integral lattices @xmath129 , and the inverse map is @xmath130 . under this isomorphism , one has @xmath131 .
the proof is similar to that of lemma [ lem : lattices ] .
the map @xmath132 establishes an isomorphism of @xmath133-vector spaces @xmath134 and @xmath135 together with the intersection product because : 1 .
since @xmath136 for @xmath40 and @xmath137 , by noether s formula the two vector spaces have the same dimension .
2 . @xmath138 .
a crucial observation is that @xmath139 sends @xmath140 to integral classes . to see this , it is sufficient to observe that @xmath141 is generated by divisors @xmath67 which are in the ramification locus and thus for which @xmath142 is integral
. consider for example the case of @xmath53 .
one has @xmath143 .
the group @xmath144 is generated by @xmath145 . since @xmath146 ,
the divisor class @xmath147 lies in group spanned by @xmath148 and @xmath144 .
so we are done . in the nodal case @xmath56 , @xmath147
is spanned by @xmath149 and @xmath144 , @xmath150 by @xmath151 and @xmath144 ; exactly the same for the non - nodal case . in the case
@xmath59 , @xmath147 is spanned by @xmath152 and @xmath144 , @xmath150 by @xmath151 and @xmath144 , @xmath153 by @xmath154 and @xmath144 .
therefore , @xmath155 is a sublattice of finite index in @xmath156 .
since the former lattice is unimodular , they must be equal .
one has @xmath157 , so the inverse map is @xmath130 .
i thank stephen coughlan for pointing out that the above proof that @xmath140 is generated by the divisors in the ramification locus does not work in the @xmath63 case . in this case , each of the lines @xmath158 , @xmath40 contains exactly two of the points @xmath159 .
what we can see easily is the following : there exists a free abelian group @xmath160 which can be identified with a subgroup of index 2 in @xmath140 and a subgroup of index 2 in @xmath156 . consider a @xmath2-divisor ( not a divisor class ) on @xmath161 @xmath162 such that the coefficients @xmath163 of @xmath70 are even .
then we can define a canonical lift @xmath164 which is a divisor on @xmath76 , and numerically one has @xmath142 .
note that @xmath67 is linearly equivalent to 0 iff @xmath66 is a torsion . by theorem [ thm : picx ] , for a burniat surface with @xmath0
, we have an identification @xmath165 \times b_0[2 ] \times c_0[2 ] = \bz_2 ^ 2 \times\bz_2 ^ 2 \times\bz_2 ^ 2.\ ] ] it is known ( see @xcite ) that for burniat surfaces with @xmath5 one has @xmath166 with the exception of the case @xmath63 where @xmath167 .
we would like to establish a convenient presentation for the picard group and its torsion for these cases which would be similar to the above .
we define the following vectors , forming a basis in the @xmath168-vector space @xmath169 : @xmath170 , @xmath171 , @xmath172 , @xmath173 , @xmath174 , @xmath175 .
further , for each point @xmath176 we define a vector @xmath177 .
we also define the standard bilinear form @xmath178 : @xmath179 .
[ thm : torsion2 - 5 ] the restriction map @xmath180\times b_0[2]\times c_0[2]$ ] is injective , and the image is identified with the orthogonal complement of the subspace generated by the vectors @xmath181 .
the restrictions of the following divisors to @xmath169 give the subset @xmath182 $ ] : @xmath183 among these , the divisors containing @xmath184 are precisely those for which the vector @xmath185\subset v$ ] satisfies @xmath186 . repeating this verbatim ,
one has the same results for the divisors @xmath187 and vectors @xmath188 .
let @xmath67 be a linear combination of the divisors @xmath189 , @xmath190 , @xmath191 , and the corresponding divisors for @xmath192 $ ] , @xmath113 $ ] .
define the vector @xmath193 to be the sum of the corresponding vectors @xmath194 , etc .
now assume that the vector @xmath195 satisfies the condition @xmath196 for all the points @xmath51 .
then the coefficients of the exceptional divisors @xmath70 in the divisor @xmath197 on @xmath161 are even ( and one can also easily arrange them to be zero since the important part is working modulo 2 ) .
therefore , a lift of @xmath198 to @xmath76 is well defined and is a torsion in @xmath199 .
this shows that the image of the homomorphism @xmath200 contains the space @xmath201 .
but this space already has the correct dimension .
indeed , for @xmath202 the vectors @xmath203 are linearly independent , and for @xmath63 the vectors @xmath204 , @xmath205 , @xmath206 , @xmath207 are linearly dependent ( their sum is zero ) and span a subspace of dimension 3 ; thus the orthogonal complement has dimension 3 as well .
therefore , @xmath208 is a bijection of @xmath209 onto @xmath201 .
[ thm : picx2 - 5 ] one has the following : 1 .
the homomorphism @xmath210 is injective , and the image is the subgroup of index @xmath211 in @xmath212\times b'_0[2]\times c'_0[2]$ ] , where @xmath213 for @xmath214 and @xmath215 for @xmath63 .
2 . @xmath121 induces an isomorphism @xmath216\times b'_0[2]\times c'_0[2]$ ] .
3 . the curves @xmath217 , @xmath124 , generate @xmath218 .
\(2 ) is and ( 1 ) follows from it . for ( 3 ) , note that @xmath219 is generated by the divisors @xmath217 and that the proof of the previous theorem shows that @xmath220 is generated by certain linear combinations of these divisors .
since @xmath132 and @xmath221 provide isomorphisms between the @xmath133-vector spaces @xmath222 and @xmath223 , it is obvious that the cones of effective @xmath133- or @xmath224-divisors on @xmath10 and @xmath11 are naturally identified . in this section
, we would like to prove the following description of the semigroup of effective @xmath2-divisors : [ thm : eff - divisors ] the curves @xmath123 , @xmath124 , generate the semigroup of effective @xmath2-divisors on burniat surface @xmath10 .
we start with several preparatory lemmas .
[ lem : eff - semigroup ] the semigroup of effective @xmath2-divisors on @xmath11 is generated by the @xmath39-curves @xmath225 .
since @xmath91 is ample , the mori - kleiman cone @xmath226 of effective curves in @xmath222 is generated by extremal rays , i.e. the @xmath39-curves @xmath227 . we claim that moreover the semigroup of integral points in @xmath226 is generated by these points , i.e. the polytope @xmath228 is totally generating .
the vertices of this polytope in @xmath229 are @xmath230 , @xmath231 , @xmath232 , @xmath233 , @xmath234 , @xmath235 , and the lattice @xmath236 is generated by them .
it is a prism over a triangular base , and it is totally generating because it can be split into 3 elementary simplices . [
lem : nef - semigroup ] the semigroup of nef @xmath2-divisors on @xmath11 is generated by @xmath237 , @xmath238 , @xmath239 , @xmath240 , and @xmath241 .
again , for the @xmath133-divisors this is obvious by mmp : a divisor @xmath67 is nef iff @xmath242 for @xmath243 , and the extremal nef @xmath67 divisors correspond to contractions @xmath244 with @xmath245 .
another proof : the extremal nef divisors correspond to the faces of the triangular prism from the proof of lemma [ lem : eff - semigroup ] , and there are 5 of them : 3 sides , top , and the bottom .
now let @xmath67 be a nonnegative linear combination @xmath246 and let us assume that @xmath247 ( resp .
@xmath248 ) . since the intersections of @xmath237 ( resp .
@xmath240 ) with the curves @xmath249 above are 0 or 1 , it follows that @xmath250 ( resp .
@xmath251 ) is also nef .
we finish by induction .
we write the divisors @xmath67 in @xmath252 using the symmetric coordinates @xmath253 note that , as in theorem [ thm : picx ] , @xmath252 and can be described either as the subgroup of @xmath254 with coordinates @xmath255 satisfying the congruence @xmath256 , or as the subgroup of @xmath254 with coordinates @xmath257 satisfying the congruence @xmath258 .
[ lem : exceptions ] the function @xmath259 on the set of nef @xmath2-divisors on @xmath11 is strictly positive , with the exception of the following divisors , up to symmetry : 1 .
@xmath260 for @xmath261 , one has @xmath262 2 .
@xmath263 for @xmath261 , one has @xmath264 .
3 . @xmath265 and @xmath266 for @xmath261 , @xmath264 .
4 . @xmath267 and @xmath268 , @xmath264 .
the divisors in @xmath269 are in the linear system @xmath270 , where @xmath271 is a fiber of one of the three rulings @xmath272 .
the divisors in @xmath273 and @xmath274 are obtained from these by adding a section .
the divisors in @xmath275 belong to the linear systems @xmath276 and @xmath277 .
let @xmath67 be a nef @xmath2-divisor .
by lemma [ lem : nef - semigroup ] , we can write @xmath278 with @xmath279 .
let us say @xmath280 .
if @xmath281 then @xmath282 . otherwise , @xmath283 , where @xmath284 , @xmath285 , or @xmath286 . then using the elementary formula @xmath287 we see that @xmath288 . continuing this by induction and adding more @xmath289 s and @xmath290 s
, one easily obtains that @xmath291 with the only exceptions listed above .
starting with @xmath292 instead of @xmath293 works the same .
[ cor : exceptions ] the function @xmath294 on the set of nef @xmath2-divisors on @xmath11 is strictly positive , with the same exceptions as above . indeed , since @xmath295
, one has @xmath296 .
[ lem : d+k ] assume that @xmath297 is a nef divisor on @xmath10 with @xmath298 .
then the divisor @xmath299 is effective .
one has @xmath300 since @xmath301 , this implies that @xmath302 .
we say that an effective divisor @xmath66 on @xmath10 is _ in minimal form _ if @xmath303 for the elliptic curves @xmath304 , and for the curves among those that satisfy @xmath305 , one has @xmath306 in @xmath307 $ ] .
if either of these conditions fails then @xmath308 must also be effective since @xmath249 is then in the base locus of @xmath309 .
a minimal form is obtained by repeating this procedure until it stops or one obtains a divisor of negative degree , in which case @xmath66 obviously was not effective .
we do not claim that a minimal form is unique .
let @xmath66 be an effective divisor on @xmath10 .
we have to show that it belongs to the semigroup @xmath310 .
_ step 1 : one can assume that @xmath66 is in minimal form_. obviously .
_ step 2 . :
the statement is true for @xmath311 .
_ there are finitely many cases here to check .
we checked them using a computer script .
for each of the divisors , putting it in minimal form makes it obvious that it is either in @xmath312 or it is not effective because it has negative degree , with the exception of the following three divisors , in the notations of theorem [ thm : picx ] : @xmath313 , @xmath314 , @xmath315 .
the first two divisors are not effective by ( * ? ? ?
* lemma 5 ) .
the third one is not effective because it is @xmath92 and @xmath316 .
_ step 3 : the statement is true for nef divisors of degree @xmath317 which are _ not _ the exceptions listed in lemma [ lem : exceptions ] . _ one has @xmath318 , so @xmath319 and the condition @xmath320 implies that @xmath66 is effective .
we are going to show that @xmath66 is in the semigroup @xmath312 .
consider the divisor @xmath321 which modulo torsion is identified with the divisor @xmath322 on @xmath11 . by lemmas [ lem : d+k ] and [ lem : eff - semigroup ] , @xmath322 is a positive @xmath2-combination of @xmath323 .
this means that @xmath324 a direct computer check shows that for any torsion @xmath325 the divisor @xmath326 is in @xmath312 for a single curve @xmath327 .
( in fact , for any @xmath328 the divisor @xmath329 is already in @xmath312 . )
thus , @xmath330 _ step 4 : the statement is true for nef divisors in minimal form of degree @xmath317 which _ are _ the exceptions listed in lemma [ lem : exceptions ] .
_ we claim that any such divisor is in @xmath312 , and in particular is effective . for @xmath331
this is again a direct computer check . for @xmath332 ,
the claim is true by induction , as follows : if @xmath66 is of exceptional type ( 1,2 , or 3 ) of lemma [ lem : exceptions ] then @xmath333 has degree @xmath334 and is of the same exceptional type .
this concludes the proof .
note that we proved a little more than what theorem [ thm : eff - divisors ] says .
we also proved that every divisor @xmath66 in minimal form and of degree @xmath335 is effective and is in the semigroup @xmath312 . for burniat surfaces with @xmath1 , a natural question to ask
is whether the semigroup of effective @xmath2-divisors is generated by the preimages of the @xmath39- and @xmath95 curves on @xmath161 .
these include the divisors @xmath217 and @xmath336 but in some cases there are other curves , too .
degenerations of burniat surfaces with @xmath3 were described in @xcite . here , we will concentrate on one particularly nice degeneration depicted in figure [ fig : degn ] .
it is described as follows .
one begins with a one - parameter family @xmath337 of del pezzo surfaces , in which the curves degenerate in the central fiber @xmath338 to a configuration shown in the left panel .
the surface @xmath339 is obtained from @xmath340 by two blowups in the central fiber , along the smooth centers @xmath341 and then ( the strict preimage of ) @xmath342 .
the resulting 3-fold @xmath339 is smooth , the central fiber @xmath343 is reduced and has normal crossings .
this central fiber is shown in the third panel .
the log canonical divisor @xmath344 is relatively big and nef over @xmath345 .
it is a relatively minimal model .
the relative canonical model @xmath346 is obtained from @xmath339 by contracting three curves .
the 3-fold @xmath346 is singular at three points and not @xmath133-factorial .
its central fiber @xmath347 is shown in the last , fourth panel .
the 3-folds @xmath348 and @xmath349 are the corresponding @xmath7-galois covers .
the 3-fold @xmath350 is smooth , and its central fiber @xmath351 is reduced and has normal crossings .
it is a relatively minimal model : @xmath352 is relatively big and nef . the 3-fold @xmath353 is obtained from @xmath350 by contracting three curves .
its canonical divisor @xmath354 is relatively ample .
it is a relative canonical model .
we note that @xmath350 is one of the 6 relative minimal models @xmath355 , @xmath356 , that are related by flops .
[ thm : exc - colln - singular ] then there exists a sequence of line bundles @xmath361 on @xmath362 whose restrictions to any fiber ( including the nonnormal semistable fiber @xmath351 ) form and exceptional collection of line bundles .
it seems to be considerably harder to construct an exceptional collection on the surface @xmath363 , the special fiber in a singular 3-fold @xmath353 . and
perhaps looking for one is not the right thing to do .
a familiar result is that different smooth minimal models @xmath355 related by flops have equivalent derived categories . in the same vein , in our situation the central fibers
@xmath364 , which are reduced reducible semistable varieties , should have the same derived categories .
the collection we construct works the same way for any of them . on the surface @xmath351
, we have 12 cartier divisors @xmath123 , @xmath365 .
the `` internal '' divisors @xmath123 , @xmath40 have two irreducible components each . of the 6 `` boundary '' divisors ,
@xmath366 are irreducible , and @xmath367 , @xmath368 , @xmath369 are reducible .
our notation for the latter divisors is as follows : the curve @xmath370 is a smooth elliptic curve ( on the bottom surface @xmath371 the corresponding curve has 4 ramification points ) , and the curve @xmath372 is isomorphic to @xmath96 ( on the bottom surface the corresponding curve has 2 ramification points ) .
we have natural norm map @xmath377 . indeed ,
every line bundle on the reducible curve @xmath378 can be represented as a cartier divisor @xmath379 , where @xmath380 are nonsingular points .
then we define @xmath381 since the dual graph of the curve @xmath378 is a tree , one has @xmath382 and @xmath383 .
we define a map @xmath386 as the composition of the restriction to @xmath378 and the norm map @xmath387 .
we also have similar morphisms @xmath388 , @xmath389 for the other two reducible curves . for the irreducible curves
@xmath366 the corresponding maps are simply the restriction maps on picard groups .
[ lem : restrictions ] consider the map @xmath390 defined as @xmath391 in the first component and the maps @xmath392 , @xmath388 , @xmath393 in the other components .
then the images of the cartier divisors @xmath123 , @xmath365 are exactly the same as for a smooth burniat surface @xmath359 , @xmath394 .
[ lem : subtract1 ] let @xmath66 be an effective cartier divisor @xmath66 on the surface @xmath351 .
suppose that @xmath398 for @xmath38 .
then the cartier divisor @xmath399 is also effective .
( similarly for @xmath400 . ) for an irreducible divisor this is immediate , so let us do it for the divisor @xmath369 which spans two irreducible components , say @xmath401 of the surface @xmath402 . let @xmath403 , @xmath404 , @xmath405 .
then @xmath406 where the right - hand intersections are computed on the smooth irreducible surfaces .
one has @xmath407 and @xmath408 .
therefore , @xmath409 .
thus , @xmath410 implies that @xmath411
. then @xmath372 must be in the base locus of the linear system @xmath412 on the smooth surface @xmath413 .
let @xmath414 be the multiplicity of @xmath372 in @xmath415 .
then the divisor @xmath416 is effective and does not contain @xmath372 . by what we just proved
, @xmath66 must contain @xmath417 .
thus , it passes through the point @xmath418 and the multiplicity of the curve @xmath419 at @xmath420 is @xmath421 , since @xmath66 is a cartier divisor .
suppose that @xmath66 does not contain the curve @xmath370 .
then @xmath422 , and @xmath423 which provides a contradiction .
we conclude that @xmath66 contains @xmath370 as well , and so @xmath424 is effective .
[ lem : subtract2 ] let @xmath66 be an effective cartier divisor @xmath66 on the surface @xmath351 .
suppose that @xmath425 for @xmath38 but @xmath426 in @xmath427 .
then the cartier divisor @xmath399 is also effective .
( similarly for @xmath400 . ) finally , assume that @xmath435 and @xmath415 does not contain @xmath372 .
by assumption , we have @xmath436 but @xmath437 in @xmath432 .
this implies that @xmath438 is effective and that @xmath66 contains the point @xmath418 .
but then @xmath439 . contradiction .
the proof of ( * ? ? ?
* lemma 5 ) , used verbatim together with the above lemmas [ lem : subtract1 ] , [ lem : subtract2 ] works . crucially , the three `` corners '' @xmath443 , @xmath444 , @xmath445 are smooth points on @xmath351 . by @xcite , for every @xmath394
they restrict to the invertible sheaves @xmath448 on a smooth burniat surface which form an exceptional sequence . by lemma [ lem : restrictions ] ,
the images of @xmath449 under the map @xmath450 are also @xmath451 .
we claim that @xmath452 also form an exceptional collection .
the properties ( 2 ) and ( 3 ) are checked by repeatedly applying ( the analogues of ) lemmas [ lem : subtract1 ] , [ lem : subtract2 ] , [ lem : ao-4.5 ] until @xmath458 ( in which case @xmath66 is obviously not effective ) . in our case
, one has @xmath459 by flatness . since we proved that lemmas [ lem : subtract1 ] , [ lem : subtract2 ] , [ lem : ao-4.5 ] hold for the surface @xmath351 , and
since the cartier divisor @xmath460 is nef , the same exact proof for vanishing of @xmath461 goes through unchanged . the semiorthogonal complement @xmath462 of the full triangulated category generated by the sheaves
@xmath463 is the quite mysterious `` quasiphantom '' . a viable way to understand it could be to understand the degenerate quasiphantom @xmath464 on the semistable degeneration @xmath351 first .
the irreducible components of @xmath351 are three bielliptic surfaces and they are glued nicely . then one could try to understand @xmath462 as a deformation of @xmath465 .
i. bauer and f. catanese , _ burniat surfaces i : fundamental groups and moduli of primary burniat surfaces _ , classification of algebraic varieties , ems ser .
soc . , zrich , 2011 , pp .
mr 2779467 ( 2012f:14074 ) | in this short note , we extend the results of [ alexeev - orlov , 2012 ] about picard groups of burniat surfaces with @xmath0 to the cases of @xmath1 .
we also compute the semigroup of effective divisors on burniat surfaces with @xmath0 .
finally , we construct an exceptional collection on a nonnormal semistable degeneration of a 1-parameter family of burniat surfaces with @xmath0 . |
in the classical mechanics , the phase space volume occupied by dynamical systems is preserved in time .
this is the statement of the liouville theorem , shown by using the hamilton dynamics : @xmath8 if we replace the phase space @xmath9 by the spacial coordinates @xmath0 , then the volume of the real space occupied by particles is preserved in time .
this gives the incompressible hydrodynamics : @xmath10 where @xmath11 is the poisson bracket , and @xmath12 is the stream function of the incompressible fluid which guarantees the continuity of the flow , @xmath13 .
if the flow is incompressible perfect fluid , the @xmath14 is regarded as the stream function . here
we assume the density of the fluid @xmath15 is constant in time .
it is easily understood that if we consider the three dimensional fluid , eq .
( [ 2d velocity ] ) is replaced by @xmath16 where the right hand side @xmath17 is the jacobian @xmath18 , being called nambu bracket at present @xcite . the advective term which describes the temporal change of a physical quantity @xmath19 by floating with the flow ,
is represented in terms of poisson or nambu bracket as @xmath20 therefore , the fluid dynamics can be given as @xmath21 where @xmath22 is the force acting on the fluid .
now the navier - stokes equation can be described in terms of nambu - poisson brackets : @xmath23 where @xmath24 is the pressure and @xmath25 is the viscosity .
this was done by nambu in his last work @xcite .
now we introduce the non - commutativity in space coordinates , and find the additional terms which will appear .
this was carried out in @xcite where the non - commutativity is introduced by replacing the nambu - poisson brackets with the moyal brackets @xcite .
we discuss the two dimensional hydrodynamics in this paper , so the moyal bracket becomes @xmath26_p \xrightarrow{replacement } \frac{1}{i\theta}[a , b]_m,\end{aligned}\ ] ] where @xmath27_m&\equiv&a*b - b*a\\ a*b&\equiv&\exp\left ( \frac{i\theta}{2}\epsilon_{ab } \frac{\partial^2}{\partial y^a \partial z^b}\right ) a(y)b(z)\bigg|_{y , z\rightarrow x}\end{aligned}\ ] ] this replacement is equivalent to consider the space coordinates @xmath0 to be the operators @xmath28 satisfying the commutation relation @xmath29=i \theta$ ] , the @xmath3 is a constant parameter . from this , the uncertainty relation is derived : @xmath30 the three dimensional case is a little more complicated .
see the details in @xcite . after the replacement of the poisson brackets with the moyal products , new terms arise only from the advective term ,
that is @xmath31_m= \{v_i,\varphi\}_p\nonumber\\ & & -\frac{\theta^2}{24 } \left ( \frac{\partial}{\partial y^a}\frac{\partial}{\partial z^b } -\frac{\partial}{\partial y^b}\frac{\partial}{\partial z^a } \right)^3 v_i(y^a , y^b ) \varphi(z^a , z^b)|_{y , z \rightarrow x } + o(\theta^4 ) .
\label{1 - 9}\end{aligned}\ ] ] now , the navier - stokes equation of the hydrodynamics on the 2 dimensional non - commutative space is given by @xcite @xmath32 the @xmath33 is the new force term which reflects the non - commutativity of the space . in the study of hydrodynamics ,
reynolds number , re @xmath34 is important , which is the ratio of the kinetic term to the viscosity term in the equation of motion and it determines the gross behavior of the flow . in the non - commutative hydrodynamics
, another measure appears , which is the ratio of the extra force term to the kinetic term in the equation of motion , _
i.e. _ sm @xmath35 . here , @xmath36 and @xmath37 is the typical velocity and length scale in the problem .
in the perfect fluid without viscosity ( @xmath39 ) , it is easy to understand the behavior of the extra force @xmath33 which is inherent to the non - commutativity .
let s examine the typical flows .
\1 ) one dimensional flow ; if @xmath40 , and @xmath41 , then @xmath42 . to consider the behavior of @xmath33 near the sink , the source , the vortex , and the flow along the corner , the complex notations are useful . define @xmath43 , and their derivatives @xmath44 , @xmath45 , and @xmath46 and @xmath47 , then we have @xmath48 where @xmath49 is the stream function .
let s introduce the complex velocity potential @xmath50 so that its real part is the usual velocity potential and its imaginary part is the stream function
. then , @xmath51 , and @xmath52 .
now we have the following results : \2 ) source or sink ; @xmath53 is a source , and @xmath54 is a sink . ) , and @xmath55 , or @xmath56 .
3 ) vortex ; @xmath57 is clock - wise , and @xmath58 is counter - clockwise ) , and @xmath59 , or @xmath60 .
4 ) flow along the corner with angle @xmath61 ; @xmath62(c : a real constant ) , and @xmath63 . from 2 ) , we understand that the additional force @xmath33 disturbs the fluid to flow into the sink , but accelerates it to flow out from the source . from 3 ) , the force @xmath33 is understood to expand the vortex in the radial direction .
therefore , the minimum size of the particle @xmath64 implied by the non - commutativity gives the indication that the flowing into the small channel becomes difficult due to the finite size effect of the particle .
so , we may hope that the hydrodynamics on the non - commutative space gives the hydrodynamics of the granular materials . in the above examples , the viscous force vanishes , @xmath65 , and hence the behavior of @xmath33
does not be altered , even if introducing the viscosity perturbatively as the expansion from the perfect fluid .
[ stv ] shows the behaviors of the extra force @xmath33 under their example flows .
the magnitude of @xmath33 depends on the size of non - commutative parameter @xmath3 and the parameters @xmath66 in @xmath50 , but the shape of the distribution is the same .
in calculating the equation of motion @xmath67 we use the fractional step method @xcite @xcite , one of the standard numerical methods to solve incompressible navier - stokes equation .
namely , to obtain the data at ( n+1)-th time step from the data at ( n)-th step , @xmath68^{(n)}.\nonumber\end{aligned}\ ] ] applying the equation of continuity to ( [ dif ] ) , the poisson equation @xmath69 is derived and the pressure @xmath70 is obtained by solving ( [ po ] ) by the iteration method . the advective term is calculated by third - order accurate upwind difference method .
third order differential factors in @xmath71 are expressed as @xmath72 here the choice of grid points is changed depending on the distance from the boundaries . in our calculation , the higher order terms of @xmath3 are ignored , by considering that even though @xmath2 terms are the lowest order ones , but are quite important to see the effect of the non - commutativity .
the boundary conditions are taken as follows : on the wall boundary , the velocity @xmath73 is set to be zero ( no - slip ) and the vertical component of gradient of the pressure @xmath74 is zero . at the entrance of the flow at @xmath75
= 0 we fix the fluid velocity to be a special value and @xmath74 = 0 , while at the exit of the flow at @xmath76 , we take the free outflow with @xmath70 = 0 . in this study
, unequally spaced grid is used .
the horizontal line and the vertical line are parametrized by @xmath77 and @xmath78 , respectively , with @xmath79 = 64 cm and @xmath80 = 21 cm , and total grid points is 175 @xmath81 71 .
the flow is assumed to be uniform along the left boundary @xmath82 an obstacle is placed at 15 cm down stream , and so the flow is separated by this obstacle and is guided to the upper or the lower open channels .
we call these open channels as slits
. the upper slit is small ( s ) , while the lower slit is large ( l ) .
the aperture of the small slit ( s ) is @xmath83 = 2 cm and that of the large slit ( l ) is @xmath84 = 8 cm .
the vertical length of the obstacle is @xmath85 = 11 cm .
both ends of the obstacle are semicirculars with the radius of 2 cm .
the minimum size of the grid spacing is 0.167 cm .
the step size of time is @xmath86 [ s ] and it carries on 2000000 steps to calculate 2000 [ s ] flow .
we have considered the flow analysis given above , with the following intention : the hydrodynamics on the non - commutative space may give the minimum area @xmath3 to the size of the constituent particles , so that if @xmath3 is introduced , then the flow rate passing through the small slit ( s ) is reduced and that passing through the large slit ( l ) is increased while keeping the total flow rates constant .
we fix the reynolds number re = 700 .
it is because in the following analysis , the activity of the vortices is important to see the non - commutativity effects .
so , we wish to take a large value of re , but the larger re @xmath87 2000 introduces the turbulent flow , and hence we fix it to 700 . if we consider the water flow , then @xmath88 and @xmath89 . in this case , if @xmath36 is chosen to @xmath90 0.55 cm / s and @xmath37 is chosen to the obstacle length @xmath9111 cm .
in order to compare the flowability through the slits between the usual flow and the non - commutative flow , the flow rates at slit ( s ) and ( l ) are estimated without and with @xmath3 in fig .
[ r0 ] and fig .
[ r2.9 ] , respectively . first look at fig .
[ r0 ] , @xmath92 cm which is the usual flow .
there exists the fluctuation of flow rates in the beginning , being regularly and periodically depicted like a sine curve , but afterwards for @xmath93 [ s ] , the oscillation is damped .
this is a characteristic of the usual flow .
the following figures show the snapshots of the flow at @xmath94495 [ s ] , 528[s ] , 1025 [ s ] and 2000 [ s ] .
cm case.,width=377 ] the velocity field and the vorticity of @xmath64 = 0 cm flow is shown in fig .
[ r0v ] , where the velocity at each points is depicted by the white arrow while the vorticity is represented by the color gradation , that is , blue is negative vorticity ( anti - clockwise rotation ) , red is positive vorticity ( clockwise rotation ) and green is the medium .
while the flow rates at slit ( s ) and ( l ) oscillate , vortices are periodically generated near the slits and the other boundaries , and flow out into the down stream as shown in fig .
[ r0a ] . in the periodical oscillation phases of ( a )
: 495 [ s ] and ( b ) : 528 [ s ] , there appear a number of red regions and a number of blue regions which means the vortices are periodically produced and flow out .
the spiral just behind the obstacle is changed from red at ( a ) to blue at ( b ) .
however at ( c ) : 1025 [ s ] , the blue regions and red regions begin individually connected to form the bands , and the so called two attached eddies " appear behind the obstacle . finally at ( d ) : 2000 [ s ] , the flow becomes stable .
this shows the damping of the oscillation of the flow rates .
now we will show the flow with @xmath95 cm .
[ r2.9 ] which gives the flow rates at slit ( s ) ( brown ) and slit ( l ) ( green ) .
cm case.,width=453 ] this figure shows the behavior of the flow rates at slit ( s ) and ( l ) in case of @xmath64=0.29 cm , or of the non - commutative flow . the oscillation of flow rates is observed , but the activity of the oscillation is weak and irregular compared to the case of @xmath96 , vortices are observed .
the flow rates becomes stable after @xmath941300 [ s ] .
this behavior can be understood from the snapshot figs .
[ r29v ] taken at @xmath94 495 , 850 , 1028 and 2000 [ s ] .
the figures show that the pattern of the vortices are almost stable from the beginning .
there is no active production of vortices just after the obstacle as was observed in the normal flow .
the extra force disturbs the active production of vortices behind the obstacle .
the oscillation arises from the activity of vortices in the down stream . at ( c):1028 [ s ] the vortices in the down stream collapse , but after 1300 [ s ]
, the whole stream becomes stable . at around @xmath94 1143
[ s ] , the flowing vortices bocome staying behind of the obstacle again , where the activity of the vortex can be seen in the snapshot in fig .
[ r29v2 ] at this moment .
= 1143 [ s ] in the case of @xmath95 cm , width=453 ] the averaged difference of the flow rates @xmath97 at the small slit with and without @xmath3 , that is , @xmath98 is depicted in fig .
[ ave ] , where the flow rates are averaged between 0 [ s ] and @xmath99 [ s ] . before @xmath99 = 1000 [ s ] , the flow rate at slit ( s ) becomes smaller for @xmath95 cm than that for @xmath100 , which matches to the intuitive understudy of the size of the fluid particle .
however , the discrepancy disappears after @xmath99 = 1000 [ s ] .
in this paper a two - dimensional simulation is performed for the hydrodynamics on the non - commutative space .
the parameter characterizing the non - commutativity of the space , @xmath3 , gives the uncertainty relation of the space , or the minimum size of the fluid particle @xmath101 .
we have prepared the flow path of 21 cm wide and 64 cm length where the obstacle is placed in the middle of the flow , and so the flow should passes through the small 2 cm slit ( s ) or the large 8 cm slit ( l ) .
the simulation is done for the water with reynolds number re = 700 , where the maximal incoming velocity is 0.55 cm / s .
re = 700 is so chosen as the activity of vortex is large , but is not too large to introduce the turbulent flow .
the results we obtained are as follows : \(1 ) for the usual flow without @xmath3 , the flow rate at the small slit ( s ) oscillates like a sine curve , but the amplitude of the oscillation diminishes slowly after @xmath941200 [ s ] .
\(2 ) for the non - commutative flow with @xmath1020.29 cm , the oscillation of the flow rate is small and irregular .
the fluctuation is stabilized after @xmath941200 [ s ] .
\(3 ) the pattern of the flow is correlated to the activity of the vortices . in the usual flow the oscillation of the flow
late is diminished after the certain time has passed , and this change is related to the occurrence of the two attached eddies " . in the non - commutative flow
this two attached eddies " appears from the beginning , and hence the flow rates do not actively oscillate .
so , the difference between the usual flow and the non - commutative flow comes from the difference of activities of vortices in the down stream .
as is understood from the analysis in the perfect fluid , the extra force @xmath33 , relevant to the non - commutativity , has the tendency to interrupts the flowing out into the sink , and to expand the size of the vortex in the radial direction .
these effects are small for a measure of the particle size @xmath103 the size of the small slit , sm @xmath104 0.29 cm @xmath103 2 cm @xmath105 0.15 , but they become extremely large near the singular positions such as the position of sink and source and the center of the vortex . therefore , it is important to know how the vortices are created from the boundary layers , especially from the boundary layer of the obstacle near the slits in both cases with and without @xmath3 , as well as to know the difference of the interaction of vortices both in the usual flow and in the non - commutative flow . for this
we have to know the vorticity transportation equation with the extra force @xmath33 , being proportional to the non - commutative parameter squared , @xmath2 .
this can be done easily . using the complex notations , the navier - stokes equation with @xmath33 reads @xmath106 \nonumber \\ & & = -2 \partial p + 8i \eta ( \partial \bar{\partial})(\partial \varphi ) - \rho \frac{2}{3 } \theta^2 \partial \varphi \left(\overleftarrow{\partial } \overrightarrow{\bar{\partial } } - \overleftarrow{\bar{\partial } } \overrightarrow{\partial } \right)^3 \varphi.\end{aligned}\ ] ] taking the rotation of this equation , @xmath107 , we have the following equation @xmath108 where @xmath109 is the vorticity defined by @xmath110 if @xmath96 , this is the usual vorticity transportation equation which describes the disscipation by the viscosity @xmath25 . on the non - commutative space , the vorticity transpiration equation
is modified so as to include the @xmath2 term .
the stream function can be expressed in terms of the vorticity by eq .
( [ def of vorticity ] ) as @xmath111 then , the vorticity transportation equation can be written on the non - commutative space as follows : @xmath112 the interaction between vortices is modified naturally in the hydrodynamics on the non - commutative space .
the remained problem to be studied is by using the creation rate and the interaction of vortices to clarify the simulation results obtained in this paper .
a preliminary stage of this paper was described in the thesis by one of the authors ( m.s . )
the authors are grateful to kazuharu bamba , gi - chol cho , takeo inami , katsuya ishii , satoshi iso , takashi koide , satoko komurasaki , shin nakamura , and ko okumura for fruitful discussions .
the flow fields were visualized by `` clef3d '' made by institute of computational fluid dynamics .
99 y. nambu , phys .
rev . * d7 * , 2405 ( 1973 ) .
y. nambu : a talk at international workshop : extra dimensions in the era of the lhc , dec .
12 - 14 ( 2011 ) ; + y. nambu : a talk at international symposium on research frontiers on physics , earth and space science , dec .
17 - 18 ( 1913 ) .
m. saitou , k. bamba and a. sugamoto : ptep 103b03 ( 2014 ) .
j. e. moyal , quantum mechanics as a statistical theory " ,
in the proceedings of cambridge philosophical society , vol * 45 * , 99 ( 1949 ) . g. d. smith : numerical solution of partial differential equations " ( oxford university press , 1985 ) t. kawamura and k. kuwahara : computation of high reynolds number flow around circular cylinder with surface roughness " , aiaa paper 84 - 0340 ( 1984 ) ; + k. kuwahara and t. kawamura : fluid computation and difference calculus " ( in japanese ) ( asakura publishing co. , 2005 ) . m. saitou ,
hydrodynamics on non - commutative space " ( in japanese ) , + ( march .
2016 ) http://hdl.handle.net/10083/59556 | a simulation of the hydrodynamics on the two dimensional non - commutative space is performed , in which the space coordinates @xmath0 are non - commutative , satisfying the commutation relation @xmath1=i \theta$ ] .
the navier - stokes equation has an extra force term which reflects the non - commutativity of the space , being proportional to @xmath2 .
this parameter @xmath3 is related to the minimum size of fluid particles which is implied by the uncertainty principle , @xmath4 . to see the effect of this parameter on the flow , following situation
is considered .
an obstacle placed in the middle of the stream , separates the flow into small slit and large slit , but the flow is joined afterwards in the down stream . for the reynolds number 700 , the behavior of the flows with and without @xmath3 is observed to differ , and the difference is seen to be correlated to the difference of the activity of vortices in the down stream .
the oscillation of the flow rate at the small slit diminishes after the certain time in the usual flow when the two attached eddies " appear . in the non - commutative flow
this two attached eddies appear from the beginning and the behavior of the flows does not fluctuate largely .
the irregularity in the flow existing in the beginning disappears after the certain time . * ocha - pp-343 * + * a simulation of hydrodynamics + on non - commutative space + * _ @xmath5faculty of sciences , + ochanomizu university , tokyo 112 - 8610 , japan
_ + _
@xmath6information technology center , + ochanomizu university , tokyo 112 - 8610 , japan _ + _ @xmath7tokyo bunkyo study center , + the open university of japan , tokyo 112 - 0012 , japan _ |
these routines compute the functions @xmath0 and @xmath1 , which constitute a numerically satisfactory pair of independent solutions of the modified bessel equation of imaginary order : @xmath11 the routine dkia computes @xmath0 and its derivative .
dlia computes @xmath1 and its derivative .
both routines require that a method for computing airy functions of real variable is available .
the routines use algorithm 819 @xcite .
the algorithm combines different methods of evaluation in different regions .
comparison between the different methods ( see accompanying paper @xcite ) , together with the use of the wronskian relation , has been used to determine the accuracy of the algorithm ( see section [ sec2 ] ) .
the algorithm computes either the functions @xmath0 , @xmath1 and their derivatives or scaled functions which can be used in larger regions than the unscaled functions .
the scaled functions are defined in the accompanying paper @xcite ( section 3 , eqs .
( 43 ) and ( 44 ) ) and in the comment lines of the algorithms .
the relative accuracy of the algorithms is ( except close to the zeros of the functions ) better than @xmath7 in the @xmath12-region @xmath13 \times [ -200,200]$ ] , better than @xmath14 in @xmath15 \times [ -500,500]$ ] and close to @xmath9 in @xmath16 \times [ -1500,1500]$ ] . both scaled and unscaled functions can be computed in @xmath17 but only scaled functions can be evaluated in the whole domains @xmath18 and @xmath19 without causing overflows / underflows for typical ieee machines .
see the accompanying paper @xcite , section 3 .
because the scaled and unscaled functions are even functions of @xmath2 , we can restrict the study to @xmath20 . in order to determine the region of applicability of each method ,
we have compared all the methods available with the non - oscillating integral representations in the accompanying paper @xcite , section 2.4 , which are expected to be valid in all the @xmath12 plane except close to @xmath21 .
this is so because , as explained in @xcite , the integration paths become increasingly non - smooth as one approaches the transition line @xmath21 ( in this region , the uniform asymptotic expansions in the accompanying paper , section 2.3 , are the best option ) .
the remaining approaches have little or no overlap in their regions of validity , except for the continued fraction method for the computation of @xmath0 and @xmath22 and the asymptotic expansions for large @xmath3 ; in fact , the validity region of the cf - method completely covers that of the asymptotic expansion ; therefore , the cf method is preferred over the asymptotic expansion for large @xmath3 for the computation of @xmath0 and its derivative .
as we will next describe , for not too large values of @xmath2 and @xmath3 there are at least two alternative methods of computation in any region of the @xmath12 plane .
this is important in order to determine the accuracy reachable by the algorithm .
a second validation of the methods is provided by the wronskian relation @xmath23 which is also satisfied by the scaled functions @xmath24 , @xmath25 , @xmath26 , @xmath27 . of course
, when two different approaches are available within a given accuracy in a same region , we choose the fastest method of computation . by order of speed , from fastest to slowest we can list the different methods of computation described in @xcite as follows : 1 .
series expansions and continued fraction method .
2 . asymptotic expansions .
integral representations .
figures 1,2 show the regions in the plane @xmath12 where the different methods are used in the routines dkia and dlia , respectively . in the first comment lines of the codes ,
the explicit equations for the curves separating the different regions are given .
* figure 2 .
* regions of applicability of each method for the @xmath1 , @xmath28 functions . *
uae * : uniform airy - type asymptotic expansion ; * i * : integral representations ; * s * : power series ; * cf * : for continued fraction and * ae * : asymptotic expansions . both the direct comparison with integral representations as well as the wronskian check show that the unscaled and scaled functions can be computed with an accuracy better than @xmath7 in the region @xmath29\times [ 0,200]$ ] without using integral representations .
this accuracy was found to be in a good compromise with efficiency .
we expect that function values outside this range will not be needed very often .
as the range considered is increased , the accuracy decreases mildly , being better than @xmath30 in @xmath6 \times [ 0,500]$ ] and close to @xmath9 in the full range of computation @xmath10\times [ 0,1500]$ ] . of course , near the zeros of the functions
( there are infinitely many of them in the oscillatory region @xmath31 ) relative accuracy loses meaning and only absolute accuracy makes sense .
figure 3 shows the cpu time spent by the code in the regions @xmath32\times [ 0,l]$ ] ; these cpu times refer to a pentium - ii 350-mhz processor running under debian - linux ; the compiler used was the gnu fortran compiler g77 .
* figure 3 . * cpu times in @xmath33\times [ 0,l]$ ] as a function of @xmath34 .
it is apparent from the figure that different methods with different speeds appear as @xmath34 grows . for small @xmath34 , only series ( and the cf method for @xmath35 )
are needed , which are the fastest methods .
as soon as asymptotic expansions come into play the cpu times tend to increase more rapidly . for @xmath34 larger than @xmath36 the effect of
the use of quadratures becomes prominent .
tables 1 and 2 show the average time spent by each method of computation in microseconds , both for dkia and dlia routines . also , the percentage of use of each method is shown .
these data are shown for two different regions of computation : the `` @xmath7 accuracy region '' ( @xmath8\times [ 0,200]$ ] ) ( table 1 ) and the full range permitted by the routines ( @xmath10\times [ 0,1500]$ ] ) ( table 2 ) .
@xmath37 * table 1 . *
percentage of use and average cpu times of each method in the region @xmath8\times [ 0,200]$ ] .
@xmath38 * table 2 . *
percentage of use and average cpu times of each method in the region @xmath10\times [ 0,1500]$ ] .
the algorithm has been tested in several computers and operating systems ( pentium ii pc under debian linux , pentium iv laptop under windows xp , pentium iv pc under windows xp and red hat linux and solaris 8 workstation ) and several compilers ( g77 and f77 for linux and windows , digital fortan and lahey fortran for windows and the sun fortran 95 ) .
in spite of the many applications of the modified bessel functions of imaginary order , the only previously published algorithm , as far as the authors know , is the code by thompson and barnett @xcite ( tb algorithm from now on ) . this code is intended for the computation of coulomb wave functions of complex order and arguments , which have as a subset modified bessel functions of complex orders .
it is not surprising that , as we later discuss , the tb algorithm has limitations when computing the @xmath0 and @xmath1 functions : this is a more general algorithm than ours and it was specially designed for real parameters , although it can be applied for complex parameters too . apart from this code , there are no available algorithms in the public domain , as can be judged by browsing at gams ( guide to available mathematical software : http://gams.nist.org ) , which are able to compute @xmath0 , @xmath1 and their derivatives .
as figure 4 shows , the tb algorithm ( program coulcc available from cpc library ) is inaccurate for imaginary orders and suffers from overflow and convergence problems .
we have tested the program coulcc @xcite against our algorithms .
figure 4 shows the comparison . at the lighter points the flag for detecting errors in coulcc was different from zero , showing that the algorithm failed to converge or suffered overflows .
the rest of points show discrepancies with our code , which persist for @xmath39 accuracy and lower .
of course , at the lighter points all the accuracy is lost and the problems persist for larger values of @xmath3 and @xmath2 .
* figure 4 .
* regions where the tb algorithm fails due to poor convergence or overflows ( light shaded points ) and points where accuracy is lost ( dark shaded points ) .
the discrepancy is greater than @xmath39 in the left figure and greater than @xmath7 in the right figure .
therefore , our algorithm seems to be the first accurate and efficient code which is capable of computing modified bessel functions of imaginary order with an accuracy close to full double precision in a wide domain .
99 a. gil , j. segura , n.m .
computing solutions of the modified bessel differential equation for imaginary orders and positive arguments .
acm trans . math . soft .
( submitted for publication in this same issue ) . | fortran 77 programs for the computation of modified bessel functions of purely imaginary order are presented .
the codes compute the functions @xmath0 , @xmath1 and their derivatives for real @xmath2 and positive @xmath3 ; these functions are independent solutions of the differential equation @xmath4 .
the code also computes exponentially scaled functions .
the range of computation is @xmath5\times [ -1500,1500]$ ] when scaled functions are considered and it is larger than @xmath6\times [ -400,400]$ ] for standard ieee double precision arithmetic .
the relative accuracy is better than @xmath7 in the range @xmath8\times [ -200,200]$ ] and close to @xmath9 in @xmath10\times [ -1500,1500]$ ] . |
for heavy quarks , their mass @xmath4@xmath5@xmath6@xmath7@xmath8 mev ( critical temperature ) . hence their kinetic equilibration time
is expected to be larger than that for light partons , thus showing better sensitivity to the in - medium interactions . due to a slower thermalization a smaller elliptic flow of @xmath2 mesons , @xmath9 ,
was expected ; however , it was suggested that a sizable @xmath9 could result from that of the light quarks @xcite through a coalescence hadronization mechanism @xcite .
surprisingly , data from the relativistic heavy - ion collider ( rhic ) for single electrons ( @xmath10 ) associated with semileptonic @xmath3 and @xmath2 decays in semi - central au - au collisions exhibited a @xmath11 of up to 10@xmath12 @xcite , indicating substantial collective behavior of charm ( @xmath13 ) quarks consistent with the assumption of a @xmath13-quark @xmath11 similar to the one for light quarks , apart from a @xmath14 shift due to radial - flow effects @xcite .
in addition , the nuclear suppression factor was found to be comparable to the pion one , @[email protected] @xcite .
perturbative qcd ( pqcd ) calculations of radiative energy loss can not explain these findings , even after inclusion of elastic scattering .
based on lattice qcd ( lqcd ) results which suggest resonance structures in the meson - correlation function at moderate temperatures , an effective model for heavy - light quark scattering via @xmath2 and @xmath3 resonances was suggested @xcite .
when implementing such a picture into a langevin simulation the results for semileptonic @xmath10 spectra @xcite are in reasonable agreement with rhic data @xcite .
we employ a fokker - planck approach to evaluate drag and diffusion coefficients for @xmath13 and @xmath16 quarks in the qgp based on elastic scattering with light quarks via @xmath2- and @xmath3-meson resonances with a width @xmath17@xmath18@xmath19-@xmath20 mev ( supplemented by perturbative interactions ) @xcite .
heavy - quark ( hq ) kinetics in the qgp is treated as a relativistic langevin process @xcite , and the medium is modeled by a spatially homogeneous elliptic thermal fireball which expands isentropically . for rhic the details of the fireball and
the parameters used can be found in @xcite ; for lhc we extrapolate to @xmath21@xmath71400 for central @xmath22=5.5 tev pb - pb collisions .
the expansion parameters are adjusted to hydrodynamic simulations , resulting in a total lifetime of @xmath23@xmath76 fm / c at the end of a hadron - gas qgp mixed phase and an inclusive light - quark elliptic flow of @xmath24=7.5% .
the main change with respect to rhic is the expectation of a new `` phase '' in which hadron - like resonances melt at higher temperature . in order to implement this `` melting '' of @xmath2- and @xmath3-mesons above @xmath25=@xmath26@xmath7360 mev
, a `` transition factor '' @xmath27)^{-1}$ ] ( @xmath28=50 mev ) has been introduced into the transport coefficients .
initial hq spectra are computed using pythia with parameters as used by the alice collaboration @xcite , and @xmath13 and @xmath16 quarks are hadronized into @xmath2 and @xmath3 mesons at @xmath6 by coalescence with light quarks @xcite ; `` left - over '' heavy quarks are treated by @xmath29-function fragmentation .
distributions at rhic ( left ) and lhc ( right ) .
circles are the initial distribution , the bands show the ones after interactions in the qgp .
dashed lines are the thermal distributions shown for comparison.,title="fig:",width=240 ] distributions at rhic ( left ) and lhc ( right ) .
circles are the initial distribution , the bands show the ones after interactions in the qgp .
dashed lines are the thermal distributions shown for comparison.,title="fig:",width=240 ]
in fig . [ fig1 ] the invariant @xmath14 distributions for @xmath13 ( red ) and @xmath16 ( blue ) quarks at rhic ( left ) and lhc ( right ) are shown .
it is evident that the charm - quark distributions after rescattering in the sqgp ( band corresponding to the uncertainty in the resonance width ) are close to the thermal distribution for @xmath14@xmath30@xmath31 where most of the yield resides . on the contrary , @xmath16 quarks are predicted to be off - equilibrium ( even at lhc energies ) .
[ fig2 ] summarizes our results for hq diffusion in a qgp in terms of the meson @xmath32 and @xmath33 for @xmath16=7 fm pb - pb collisions at the lhc .
our most important findings are : ( a ) resonance interactions substantially increase ( decrease ) @xmath11 ( @xmath15 ) compared to perturbative interactions ; ( b ) @xmath16 quarks are much less affected than @xmath13 quarks , reducing the effects in the @xmath34 spectra ; ( c ) there is a strong correlation between a large @xmath11 and a small @xmath15 at the quark level , which , however , is partially reversed by coalescence contributions which increase _ both _ @xmath11 and @xmath15 at the meson ( and @xmath34 ) level .
this feature turned out to be important in the prediction of @xmath34 spectra at rhic ; ( d ) the predictions for lhc are quantitatively rather similar to our rhic results @xcite , due to a combination of harder initial hq-@xmath14 spectra ( see circles in fig . [ fig1 ] ) and a decrease in interaction strength in the early phases where non - perturbative resonance scattering is inoperative . therefore we conclude that if at rhic the dominant contribution to hq interactions are hadron - like resonances , at lhc we should observe an @xmath15 and @xmath11 pattern similar to rhic .
= 7 fm , @xmath22=5.5 tev pb - pb collisions : @xmath15 ( left ) and @xmath11 ( right ) for @xmath2 and @xmath3 mesons.,title="fig:",width=240,height=163 ] = 7 fm , @xmath22=5.5 tev pb - pb collisions : @xmath15 ( left ) and @xmath11 ( right ) for @xmath2 and @xmath3 mesons.,title="fig:",width=240,height=163 ] | in ultrarelativistic nuclear collisions heavy quarks are produced out of thermal equilibrium in the very early stage of the reaction and their thermalization time was expected to be considerably larger than that of light quarks . on the other hand
, a strongly - interacting qgp can be envisaged in the heavy quark sector due to the presence of heavy - light hadron - like resonances in the qgp for temperatures up to @xmath0@xmath1 .
we investigate the consequences of such states employing a relativistic langevin simulation .
hadronization is modeled by a coalescence - fragmentation scheme .
we present the predictions for the nuclear modification factor and elliptic flow of @xmath2 and @xmath3 mesons at lhc energies and compare the results with the successful predictions of this model for rhic conditions .
we find similar heavy - quark thermalization effects at lhc and rhic . , , heavy quarks , quark - gluon plasma , collective flow , hadronization . 12.38 mh , 24.85.+p , 25.75 nq , 25.75 ld |
we provide a specific design which admits a strong coupling of a nanomechanical resonator to an lc circuit and simultaneously admits optical coupling ( fig .
[ f : setup ] ) .
specifically we envision a parallel plate capacitor where we replace one of the plates by a set of wires of thickness @xmath22 , width @xmath23 , and separated by a distance @xmath24 , see fig .
[ fig : capacitor](a ) . in order to have an interaction with light we introduce a hole in the capacitor to allow for laser beams to get through ( not shown ) .
this hole has a minimal effect on the capacitance when the membrane area is much larger than the mode cross section of the cavity , consistent with a ( 1 mm)@xmath33 membrane and a 50 @xmath34 m cavity waist .
a dielectric membrane of thickness @xmath25 is inserted into the capacitor a distance @xmath8 from the top of the wires .
the replacement of one of the capacitor plates by the set of wires creates a spatial inhomogeniety of the electric field , which attracts the membrane towards the wires when the capacitor is charged . at the same time
this inhomogeniety also means that the capacitance @xmath35 will depend on the position of the membrane @xmath36 . expanding the capacitance around the equilibrium position
gives rise to an lc - membrane coupling @xmath37 .
this is analogous to the radiation pressure coupling that occurs in the optical domain , i.e. , a @xmath38 type nonlinearity . as such
, we can enhance the coupling strength by providing a classical displacement of the lc circuit s charge , with either a dc or ac voltage bias @xmath39 providing an offset charge , @xmath40 . for simplicity
we restrict ourselves to the case of a dc voltage , though generalization to the ac case is a simple extension of these ideas and allows to frequency match the lc and mechanical systems .
the coupling between the membrane position and the charge fluctuations around the equilibrium @xmath41 is then @xmath42 and is enhanced by the large charge @xmath43 induced on the capacitor , in direct analogy to the similar effect for cavity optomechanics .
the full hamiltonian including electrical and mechanical contributions is : @xmath44 here @xmath45 is the equilibrium membrane position at @xmath46 .
the fixed point of the classical charge @xmath43 at a given bias voltage and the equilibrium displacement @xmath47 are then found from @xmath48 which yields @xmath49 here we have introduced a characteristic length scale @xmath50 defined by @xmath51 , which describes the relative change in the capacitance at the new equilibrium position @xmath47 . around these classical values , we consider the remaining quantum fluctuations @xmath52 .
we change to annihilation and creation operators @xmath53 for the membrane ( lc ) , and find a hamiltonian @xmath54 with @xmath55 .
this corresponds to the model hamiltonian examined in the beginning . assuming a constant value of @xmath56 from @xmath45 to @xmath47 , the coupling constant can be expressed in a more intuitive form @xmath57 , i.e. , through the capacitance change @xmath58 caused by the displacement of the membrane due to the applied voltage @xmath39 .
the solution is a stable point under the condition @xmath59 .
we have neglected a small nonlinear correction , due to the femtometer length scale of the zero - point membrane fluctuations .
devices which enhance this nonlinear coupling are of interest , but beyond the scope of the present work . to obtain quantitative estimates of the feasible coupling constant , we assume the plate electrode to be much larger than the separation of the plate and transverse dimensions of the wires @xmath60 .
we can then find the capacitance for a given position of the membrane by solving for the potential using the finite element method .
we express the capacitance as @xmath61 , where @xmath62 is a dimensionless number of order unity , which describe the deviation from a standard parallel plate capacitor . as an example , we take a sin membrane of dielectric constant @xmath30 and thickness @xmath63 nm inserted into a capacitor with a separation @xmath64 @xmath34 m and dimensions @xmath26 , @xmath27 , and @xmath28 . a simulation with these values
is shown in fig .
[ fig : capacitor ] ( b ) . from this simulation
we extract the values of @xmath65 @xmath34 m for a distance of @xmath66 @xmath34 m . hence if the applied voltage shifts the equilibrium position by @xmath67 nm around @xmath66 @xmath34 m the coupling constant is @xmath68 and the system is in the strong coupling regime if the q values of the lc circuit and membrane exceed 100 . assuming an operating frequency of @xmath69 mhz and an oscillator length @xmath70
fm this displacement only requires an applied voltage on the order of a few volts .
if a larger initial separation is desirable , a similar coupling ( @xmath71 ) could be achieved if the equilibrium distance is shifted from @xmath72 @xmath34 m to @xmath73 @xmath34 m ( @xmath74 @xmath34 m for @xmath73 @xmath34 m ) with an applied voltage of several hundred volts .
the membrane may be efficiently cooled via optomechanical coupling between the radiation pressure force of a cavity field and the position of the central area of the membrane .
the details of this process have been analyzed by a wide variety of groups @xcite .
in essence the effect on the membrane degree of freedom @xmath75 is to induce a damping @xmath76 , which is much greater than the intrinsic damping rate of the membrane @xmath77 , but is limited by the cavity decay rate @xmath78 .
this additional damping only produces moderate additional quantum fluctuations associated with the vacuum noise of the light field ( which can be accounted for by adjusting the temperature of @xmath79 defined below ) . working with the lc circuit with damping @xmath80 , resonant with the membrane ( @xmath81 )
, we may expect an efficient coupling between the optomechanical system and the lc circuit , provided that @xmath82 such that we can get excitations out of the system faster than they leak in .
we use the input - output formalism in the rotating wave approximation to find a full description of this combined mode cooling .
the heisenberg - langevin equations describing this situation are @xmath83 in a strong damping limit ( @xmath84 ) , we can treat the coupled lc resonator as a perturbation and arrive at @xmath85 this equation describes the cooling of @xmath86 through the membrane - light system with a rate @xmath87 . in the continuous cooling limit , we expect to achieve a thermal population in @xmath86 given by @xmath88 where @xmath89 are the original thermal occupation of modes @xmath75 and @xmath86 . typically this population will be dominated by the heating of the lc circuit ( the first term ) since the membrane can have a very large @xmath90 @xcite . in the mode - resolved , strong - coupling limit , with @xmath91 and @xmath92 , each of the two normal modes @xmath93 and @xmath94 have frequencies @xmath95 and a damping rate given given by the average of the two damping rates @xmath96 .
the optomechanical coupling then works independently on each of the two modes .
assuming again the @xmath97 of the membrane to be much higher than the @xmath97 of the lc , a standard argument for optomechanical cooling @xcite leads to a thermal occupation number of @xmath98 . comparing the two limits derived above we see that the minimal thermal occupation is achieved at a cooling laser power and detuning such that @xmath99 , where we obtain a population @xmath100 .
the cooling of the membrane is , however , limited by the cavity decay rate @xmath78 .
hence the cooling limit is the larger of @xmath100 and @xmath101 .
we have neglected optical heating effects , consistent with our assumption of good sideband - resolution ( @xmath102 ) .
the cooling identified above realizes an interface between optical fields and rf excitations of the lc circuits at the single photon level , and we now turn to a possible application of this interface .
often lc circuits are used in sensitive detectors to pick up very small signals @xcite . as we will now show
, the sensitivity in such experiments can be improved by detecting the cooling light leaving the cavity .
this takes advantage of the fact that the homodyne detection of laser light can be quantum noise limited with near - unit quantum efficiency , thus avoiding many of the noise sources present for low frequency signals . to show this we will work in the strong damping limit identified above @xmath103 with the lc circuit tuned into resonance with the membrane ( @xmath104 ) .
again we also assume the damping of the mechanical motion of the membrane to be negligible @xmath105 . in this limit
the membrane and the cavity mediate an effective interaction between the lc mode @xmath86 and the optical cavity input / output modes @xmath106 with the effective cooling rate @xmath107 introduced above . in the rotating wave approximation
this situation is described by the generic equations @xmath108 here , we have introduced an incoming signal to be measured , which is described by @xmath109 . if the signal is from a voltage @xmath39 applied to the system , @xmath110 .
suppose that we want to measure the fourier components of the incoming signals @xmath111 detuned by a frequency @xmath112 with respect to the resonance frequency of the lc circuit @xmath113 within a certain bandwidth @xmath114 .
this can be done by splitting the outgoing signal on a beamsplitter ( @xmath115 , where @xmath116 is the annihilation operator for the other mode incident on the beamsplitter ) and inferring the two quadratures @xmath117 and @xmath118 from a homodyne detection of the @xmath119 ( @xmath120 ) quadrature of the @xmath121 ( @xmath122 ) mode .
the signal - to - noise ratio for , e.g. , a measurement of the amplitude can be defined by @xmath123 . here
@xmath124 describes the noise @xmath125 in the absence of any signal . from the equations of motion we find @xmath126}(2n_d+1)}}\ ] ] here @xmath127 describes the number of thermal excitations in the field used to probe the circuit and we assume that the fields incident on the beamsplitter are of the same type such that @xmath128 .
let us compare our approach to the case where the lc circuit is read out by homodyne detection with an rf amplifier assumed to have a similar number of thermal excitations as the system being measured @xmath129 . disregarding any additional noise added during the amplification , @xmath130 is optimized for @xmath131 and is limited to @xmath132 with a detection bandwidth @xmath133 .
in contrast , with the optical readout , the incoming laser fields can be quantum noise limited with @xmath134 if light is in a coherent state . in this case
we obtain twice the signal - to - noise ratio @xmath130 for @xmath135 .
the optimal sensitivity is thus better with optical detection , even if we assume ideal detection of the fields in both cases .
such an ideal detection is routinely achieved by homodyne detection of optical fields with near unity quantum efficiency , whereas it is hard to achieve for rf fields . for realistic limited detect or efficiencies of rf fields ,
the sensitivity may thus be significantly improved using optical readout .
furthermore the high sensitivity with laser cooling is obtained over a much larger bandwidth which is determined by @xmath136 . in other words ,
if , prior to laser cooling , the lc circuit had a high - q and a narrow bandwidth @xmath80 that is less than the bandwidth @xmath137 required for a particular application , laser cooling allows an increase of the bandwidth with a limited decrease in the sensitivity ( @xmath138 db ) if @xmath139 . using regular rf techniques
, an alternative approach would be to increase the bandwidth by increasing the damping of the circuit , but this would result in a decrease of the sensitivity by a factor of @xmath140 . crucially , since optical fields are shot - noise limited even at room temperature , this measurement setup does not involve cryogenics .
the potential benefits of this approach high quantum efficiency conversion from radio frequency to optical photons , and the corresponding potential for low temperature detection of radio frequency signals are limited by the finite q values for room temperature inductors .
appropriate replacements may be considered , such as crystal resonators or cryogenic superconducting resonators .
an additional benefit of a cryogenic setup is the possibility to enter the quantum strong coupling limit , @xmath141 , at which point the conversion from radio frequency to optical domain can be used as a quantum interface .
however , understanding of these features and improvements requires further investigation .
we gratefully acknowledge useful discussions with koji usami , ole hansen , silvan schmidt , anja boisen , and john lawall .
jmt thanks the nbi for hospitality during his stay .
this research was funded by aro muri award w911nf0910406 , darpa and by the eu project q - essence .
schliesser , a. , kippenberg , t. j. _ adv . in amo phys . _
* 58 , * 207 ( 2010 ) .
aspelmeyer , m. , et al . _ josa b _ * 27 , * a189 ( 2010 ) .
genes , c. , et al . _ adv . in amo phys .
_ * 57 , * 33 ( 2009 ) .
kippenberg , t. j. & vahala , k. j. _ science _ * 321 * , 1172 ( 2008 ) .
favero , i. & karrai , k. _ nature photonics _ * 3 , * 201 ( 2009 ) .
marquardt , f. & girvin , s. m. _ physics _ * 2 , * 40 ( 2009 ) . j. d. thompson , et al . , * 452 * , 72 ( 2008 ) ; a. m. jayich , et al . , * 10 * , 095008 ( 2008 ) . | we explore a method for laser cooling and optical detection of excitations in a lc electrical circuit .
our approach uses a nanomechanical oscillator as a transducer between optical and electronic excitations . an experimentally feasible system with the oscillator
capacitively coupled to the lc and at the same time interacting with light via an optomechanical force is shown to provide strong electro - mechanical coupling .
conditions for improved sensitivity and quantum limited readout of electrical signals with such an `` optical loud speaker '' are outlined .
cooling plays an essential roles in most areas of physics , in part because it reduces detrimental thermal fluctuations . for sensing application , where thermal fluctuations may hide the small signals one is trying to measure , strong coupling of mechanical and electrical oscillators to systems in a pure quantum state , such as light or polarized atomic ensembles
, opens up new possibilities for quantum sensing of fields and forces @xcite .
this in principle allows for enhanced sensitivity of the oscillators , where readout of their state is limited only by quantum fluctuations . in recent years
, dramatic advances in optomechanical coupling and cooling of high quality - factor ( q ) mechanical systems have been made @xcite . in this letter , we propose to extend laser cooling of mechanical objects to electrical circuits . by coupling a high - q inductor - capacitor resonator ( lc ) to a near - resonant nanomechanical membrane @xcite in the radio frequency ( rf ) domain , the electrical circuit can be effectively cooled by the cold mechanical system . since such electrical circuits are used in a wide variety of settings , the reduction of thermal fluctuations in these system will likely find numerous applications .
we also show that the cooling techniques explored here allow for optical readout of electrical signals in the circuit . since light fields are routinely measured with quantum limited precision this allows for high sensitivity broadband detection of weak electric signals .
moreover , light and atomic ensembles can behave as oscillators with a negative mass and thus provide the means to measure fields and forces beyond the standard quantum limit using the power of entanglement @xcite .
finally , our coupling occurs at the level of individual radio frequency photons ; thus , for systems with sufficiently low thermal load , our approach provides a versatile interface for quantum information , allowing for the reliable transfer of quantum states from radio frequency to optical domains and back .
the key idea in this work is to achieve sufficient coupling between a nanomechanical membrane and a high - q lc circuit to induce normal mode splitting @xcite , and then observe the electrical excitations via opto - mechanical coupling between the membrane and a high - q optical cavity .
our suggested method is insertion of the membrane into the fringing field of a capacitor @xcite , as shown in fig .
[ f : setup ] , such that the capacitance depends upon the displacement of the membrane .
we demonstrate that for reasonable component parameters , when combined with a voltage bias and an inductive component to make a resonant electrical circuit near the frequency of a mechanical resonance @xmath0 , the coupling @xmath1 between radio - frequency ( rf ) photons and the membrane phonons can become sufficiently large to induce normal mode splitting , where the resonant response of the system comprises combined electro - mechancial excitations .
, width=288 ] a model hamiltonian describing the coupled electromechanical system ( fig .
[ f : setup ] ) in the limit of well separated , high - q resonant electrical and mechanical modes is h = + + + + ( q x_m ) .
here @xmath2 , @xmath3 , and @xmath4 .
the flux @xmath5 and the membrane momentum @xmath6 are the canonical momenta conjugate to the charge @xmath7 and position @xmath8 .
when the two modes are brought into resonance @xmath9 , the natural canonical variables become normal mode solutions @xmath10 with @xmath11 ; @xmath12 ; @xmath13 ; @xmath14 , with frequencies @xmath15 for @xmath16 as considered below .
this coupled - mode system could be inserted into an optical cavity ( fig .
[ f : setup ] ) , whose mode couples to the membrane position @xmath8 via radiation pressure @xcite .
both normal modes interact with the cavity mode as @xmath17 ; thus , radiation pressure - based detection can be applied independently to each normal mode when the cavity linewidth @xmath18 is narrow enough to resolve the normal mode splitting @xmath19 .
alternatively , both normal modes can be observed simultaneously when @xmath20 . in both cases
the optical system can also cool the combined electro - mechanical system .
in essence this cooling is achieved because the membrane - cavity system acts as a transducer up - converting excitations in the lc circuits to optical frequencies .
this means that the lc circuit will equilibrate with the optical modes which are in the quantum mechanical ground state even at room temperature . at the same time
the interface between rf excitations in the lc circuits and optical photons also allows for detection of rf - electric signals by optical measurement .
since optical measurement can by quantum noise limited this opens up new possibilities for detection weak electrical signals , as we outline below .
separated from a set of wires by a distance @xmath21 .
the wires have thickness @xmath22 and width @xmath23 and are separated by a distance @xmath24 .
a membrane of thickness @xmath25 is separated from the wires by a distance @xmath8 . to find the capacitance we perform fem simulations over a cross section of the capacitor ( right ) .
the gray scale indicates the simulated electric potential with @xmath26 , @xmath27 , @xmath28 , @xmath29 , and @xmath30 .
( b ) ratio of capacitance with the membrane to a parallel plate geometry without membrane , @xmath31 ( full curve , left axis ) and characteristic length @xmath32 ( dashed curve , right axis ) for the same parameters as in ( a).,title="fig:",width=321 ] separated from a set of wires by a distance @xmath21 .
the wires have thickness @xmath22 and width @xmath23 and are separated by a distance @xmath24 .
a membrane of thickness @xmath25 is separated from the wires by a distance @xmath8 . to find the capacitance we perform fem simulations over a cross section of the capacitor ( right ) .
the gray scale indicates the simulated electric potential with @xmath26 , @xmath27 , @xmath28 , @xmath29 , and @xmath30 .
( b ) ratio of capacitance with the membrane to a parallel plate geometry without membrane , @xmath31 ( full curve , left axis ) and characteristic length @xmath32 ( dashed curve , right axis ) for the same parameters as in ( a).,title="fig:",width=321 ] |
the minority game ( mg ) was originated from the el farol bar problem in game theory first conceived by arthur in 1994 @xcite , where a finite population of people try to decide , at the same time , whether to go to the bar on a particular night . since the capacity of the bar
is limited , it can only accommodate a small fraction of all who are interested . if many people choose to go to the bar , it will be crowded , depriving the people of the fun and thereby defying the purpose of going to the bar . in this case , those who choose to stay home are the winners .
however , if many people decide to stay at home then the bar will be empty , so those who choose to go to the bar will have fun and they are the winners . apparently ,
no matter what method each person uses to make a decision , the option taken by majority of people is guaranteed to fail and the winners are those that choose the minority one .
indeed , it can be proved that , for the el farol bar problem there are mixed strategies and a nash - equilibrium solution does exist , in which the option taken by minority wins @xcite .
a variant of the problem was subsequently proposed by challet and zhang , named as an mg problem @xcite , where a player among an odd number of players chooses one of the two options at each time step .
subsequently , the model was studied in a series of works @xcite . in physics , mg has received a great deal of attention from the statistical - mechanics community , especially in terms of problems associated with non - equilibrium phase transitions @xcite . in the current literature
, the setting of mg is that there is a _ single _ resource with players two possible options ( e.g. , in the el farol bar problem there is a single bar and the options of agents are either going to the bar or not ) , and an agent is assumed to react to available global information about the history of the system by taking on an alternative option that is different than the current one it is taking .
the outstanding question remains of the nonlinear dynamics of mg with _
multiple resources_. the purpose of this paper is to present a class of multi - resource mg models . in particular , we assume a complex system with multiple resources and , at any time , an individual agent has @xmath3 _ resources / strategies _ to choose from .
we introduce a parameter @xmath0 , which is the probability that each agent responds based on its available local information by selecting a less crowded resource in an attempt to gain higher payoff .
we call @xmath0 the _ minority - preference probability_. we find , strikingly , as @xmath0 is increased , the phenomenon of grouping emerges , where the resources can be distinctly divided into two groups according to the number of their attendees .
in addition , the number of stable pairs of groups also increases .
we shall show that the grouping phenomenon plays a fundamental role in shaping the fluctuations of the system .
the phenomenon will be demonstrated numerically and explained by a comprehensive analytic theory .
an application to the analysis of empirical data from a real financial market will also be illustrated , where grouping of stocks ( resources ) appears .
our model is not only directly relevant to nonlinear and complex dynamical systems , but also applicable to social and economical systems .
our multi - resource mg model is presented in sec .
[ sec : model ] .
the emergence of grouping phenomenon is demonstrated in sec .
[ sec : numerics ] .
a general theory is developed in sec .
[ sec : theory ] to elucidate the dynamics of the emergence and evolution of the strategy groups .
concluding remarks are presented in sec .
[ sec : conclusion ] .
we consider a complex , evolutionary - game type of dynamical system of @xmath4 interacting agents competing for multiple resources .
each agent will chose one resource in each round of the game . and
, each resource has a limited capacity , i.e. , the number of agents it can accommodate has an upper bound @xmath5 .
there are thus multiple strategies ( @xmath6 , @xmath7 , @xmath8 , @xmath9 , where @xmath9 is the maximum number of resources / strategies ) available to each agent . on average ,
each strategy can accommodate @xmath10 agents , and we consider the simple case of @xmath11 .
let @xmath12 be the number of agents selecting a particular strategy @xmath13 .
if @xmath14 , the corresponding agents win the game and , consequently , @xmath13 is the _ minority strategy_. however , if @xmath15 , the associated resource is too crowded so that the strategy fails and the agents taking it lose the game , which defines the `` majority strategy . ''
the optimal solution to the game dynamics is thus @xmath16 . in a real - world system , it is often difficult or even impossible for each agent to gain global information about the dynamical state of the whole system .
it is therefore useful to introduce the concept of _ local information network _ in our multiple - resource mg model . at each time step , with probability @xmath0 , namely the _ minority - preference probability _ , each agent acts based on local information that it gains by selecting one of the @xmath9 available strategies . in contrast , with probability @xmath1 , an agent acts without the guidance of any local information . for the minority - preference case , agent @xmath17 has @xmath18 neighbors in the networked system . the required information for @xmath17 to react consists of all its neighbors strategies and , among them , the winners of the game , i.e. , those neighboring agents choosing the minority strategies at the last time step .
let @xmath19 be the set of minority strategies for @xmath17 s winning neighbors , where a strategy may appear a number of times , if it has been chosen by different winning neighbors .
with probability @xmath0 , agent @xmath17 will chose one strategy randomly from @xmath20 .
thus , the probability @xmath21 for strategy @xmath13 to be selected is proportional to the times it appears in @xmath20 , i.e. , @xmath22 , where @xmath23 is the number of elements in @xmath20 and @xmath24 the times strategy @xmath13 appears in @xmath20 .
if @xmath20 is empty , @xmath17 will randomly select one from the @xmath9 available strategies . while , for the case that an agent selects a strategy without the guidance of any local information with probability @xmath1 , it will either choose a different strategy randomly from the @xmath9 available ones with mutation probability @xmath25 @xcite , or inherit its strategy from the last time step with probability @xmath26
as a concrete example to illustrate the strategy - grouping phenomenon , we set @xmath27 . figures [ fig : ts](a - c ) show time series of @xmath12 , the number of agents selecting each strategy @xmath13 , for @xmath28 , 0.45 , and 1.0 , respectively . for fig .
[ fig : ts](a ) where @xmath29 , an agent makes no informed decision in that it changes strategy randomly with probability @xmath25 but stays with the original strategy with probability @xmath26 . in this case , @xmath12 s appear random .
for the opposite extreme case of @xmath30 [ fig .
[ fig : ts](c ) ] , each agent makes well informed decisions based on available local information about the strategies used by its neighbors . in this case , the time series are quasiperiodic ( a detailed analysis will be provided in sec .
[ sec : theory ] ) . for the intermediate case of @xmath31 [ fig .
[ fig : ts](b ) ] , agents decisions are partially informed . in this case ,
an examination of the time series points to the occurrence of an interesting grouping behavior : the 5 strategies , in terms of their selection by the agents , are divided into two distinct groups @xmath32 and @xmath33 that contain @xmath34 and @xmath35 resources , respectively . the time series associated with the smaller group exhibit larger fluctuations about its equilibrium . to better characterize the fluctuating behaviors in the time series @xmath12
, we calculate the variance @xmath36 ^ 2\rangle$ ] as a function of the system parameter @xmath0 , where @xmath37 is the expectation values averaged over a long time interval , as shown in fig .
[ fig : variance ] on a logarithmic scale .
we observe a generally increasing behavior in @xmath38 with @xmath0 and , strikingly , a bifurcation - like phenomenon .
in particular , for @xmath39 , where @xmath40 is the bifurcation point , @xmath41s for all strategies assume approximately the same value . however , for @xmath42 , there are two distinct values for @xmath41 , signifying the aforementioned grouping behavior [ fig .
[ fig : ts](b ) ] . from fig .
[ fig : variance ] , we also see that , after the bifurcation , the two branches of @xmath41 are linear ( on a logarithmic scale ) and have approximately the same slope @xmath43 , suggesting the following power - law relation : @xmath44 , for @xmath45 , where @xmath46 and @xmath47 are the intercepts of the two lines in fig .
[ fig : variance ] .
we thus obtain @xmath48 in sec .
[ sec : theory ] , we will develop a theory to explain the relations among the variances of the grouped strategies and to provide formulas for the amplitudes of the time series in fig . [
fig : ts ] and the sizes of the groups ( denoted by @xmath49 and @xmath50 , respectively ) .
specifically , our theory predicts the following ratio between the variances of the two bifurcated branches : @xmath51 which is identical to the numerically observed ratio in eq .
( [ eq : main_ratio ] ) , with the additional prediction that the strategies in the group of smaller size exhibit stronger fluctuations since the corresponding value of @xmath41 is larger .
overall , the emergence of the grouping behavior in multiple - resource mgs , as exemplified in fig .
[ fig : variance ] , resembles a period - doubling like bifurcation .
while period - doubling bifurcations are extremely common in nonlinear dynamical systems , to our knowledge , in complex game systems a clear signature of such a bifurcation had not been reported previously .
a careful examination of the time - series for various @xmath0 has revealed that the strategy - grouping processes has already taken place prior to the bifurcation point @xmath40 in the variance @xmath41 , but all resulted grouping states are unstable .
take as an example the @xmath52-strategy system in fig .
[ fig : ts ] . in principle
, there can be two types of pairing groups : @xmath53 and @xmath54 .
for any grouping state , the following constraint applies : @xmath55 there are in total @xmath56 ( if @xmath9 is even ) or @xmath57 ( if @xmath9 is odd ) possible grouping states for the system with @xmath9 available strategies .
however , the grouping states are not stable for @xmath58 .
what happens is that a strategy can remain in one group but only for a finite amount of time before switching to a different group .
assume that the sizes of the original two pairing groups are @xmath59 and @xmath60 , respectively .
the sizes of the new pair of groups are thus @xmath61 and @xmath62 , as stipulated by eq .
( [ eq : constraint ] ) . associated with switching to a different pair of groups , the amplitudes of the time series @xmath12 for each strategy also change . as the bifurcation parameter @xmath0 is increased , the stabilities of different pairs of grouping states also change . at the bifurcation point @xmath40 ,
one particular pair of groups becomes stable , such as the grouping state @xmath54 in fig .
[ fig : variance ] . the bifurcation - like phenomenon and the emergence of various strategy - grouping states are general for multiple - resource mg game dynamics .
for example , fig .
[ fig : k16 ] shows @xmath41 as a function of @xmath0 for a system with @xmath63 available strategies .
there are in total 8 possible grouping states , ranging from @xmath64 to @xmath65 . as @xmath0 is increased , the grouping states @xmath64 , @xmath66 , @xmath67 , @xmath68 and @xmath69 become stable one after another , as can be seen from the appearance of their corresponding branches in fig .
[ fig : k16 ] . the behavior can be understood theoretically through a stability analysis ( sec .
[ sec : theory ] ) .
another phenomenon revealed by fig .
[ fig : k16 ] is the merging of bifurcated branches .
for example , as @xmath0 is increased through about 0.8 , the grouping states disappear one after another in the reverse order as they initially appeared .
this can also be understood through the stability analysis ( sec .
[ sec : theory ] ) .
here we develop an analytic theory to understand the emergence , characteristics , and evolutions of the strategy groups . in general , for a multiple - resource mg system of @xmath4 agents , as the parameter @xmath0
is increased so that agents become more likely to make informed decision for strategy selection , the available strategies can be divided into pairs of groups .
the example in fig .
[ fig : ts](b ) presents a case where there are two distinct strategy groups @xmath32 and @xmath33 , which contain @xmath49 and @xmath50 strategies , respectively , where @xmath70 . for fig .
[ fig : ts](b ) , we have @xmath71 .
the strategies belonging to the same group are selected by approximately the same number of agents , i.e. , the time series @xmath72 for strategies in the same group are nearly identical . during the time evolution , a strategy @xmath13 can switch iteratively from being a minority strategy [ @xmath73 to being a majority one [ @xmath74 .
in particular , as shown in the schematic map in fig . [
fig : sketchmap ] , for the strategy in group @xmath32 denoted by @xmath75 , and the strategy in group @xmath33 denoted by @xmath76 , if @xmath77 , we will have @xmath78 .
in addition , the time series @xmath12 reveals that the average numbers of agents for strategies @xmath75 and @xmath76 , denoted by @xmath79 and @xmath80 ( the blue dash line and red dot line in fig .
[ fig : sketchmap ] ) , respectively , are not equal to @xmath10 ( the black solid line in fig .
[ fig : sketchmap ] ) .
in fact , we have @xmath81 with @xmath82 , and @xmath83 . here
, @xmath84 is the absolute value of the difference between @xmath85 and @xmath10 . the number of agents @xmath86 [ or @xmath87 typically fluctuates about the equilibrium @xmath79 [ or @xmath88 with amplitude @xmath89 [ or @xmath90 , as shown in the schematic map in fig . [ fig : sketchmap ] .
based on the numerical observations , we can argue that the strategy grouping phenomenon is intimately related to the fluctuations in the time series @xmath12 . assuming the mg system is closed so that the number of agents is a constant , we have @xmath91 thus , for two consecutive time steps , we have , @xmath92 substituting eq .
( [ eq : i1 ] ) into eq .
( [ eq : i2b ] ) , we have @xmath93 from which we obtain the relations between @xmath89 and @xmath94 , @xmath95 and @xmath96 as @xmath97 we see that the fluctuations of the time series are closely related the grouping of the strategies . from the definition of @xmath38 , we can write the variances of @xmath98 and @xmath99 as @xmath100 using eq .
( [ eq : i4 ] ) , we obtain @xmath101 as shown in fig .
[ fig : variance ] , the ratio of the variances of group @xmath32 and @xmath33 from the simulation agree very well with eq .
( [ eq : i5 ] ) .
we develop a mean - field theory to understand the fluctuation patterns of the system . to be concrete , we still treat the case of two distinct groups .
consider strategy @xmath75 that belongs to group @xmath32 and assume that @xmath75 is the majority strategy at time @xmath102 , i.e. , @xmath103 .
according to the mean - field approximation , at the next time step @xmath104 , the number of agents @xmath105 choosing strategy @xmath75 is @xmath106 \nonumber \\ & + & ( n - n_{s_i}^{(0)})(1-p)m\frac{1}{k},\end{aligned}\ ] ] where @xmath107 here , @xmath108 and @xmath109 together are the number of agents abandoning strategy @xmath75 ( or the flow out of strategy @xmath75 ) .
that is , of the @xmath110 agents , @xmath108 agents will act based on local information by selecting a minority strategy different than @xmath75 for the next time step @xmath111 . at the same time , there will be @xmath109 agents acting without local information by choosing randomly one of the other @xmath112 strategies .
the quantity @xmath113 represents the flow into @xmath75 from the remaining @xmath114 agents .
these agents will mutate randomly to switch their strategies to @xmath75 without any local information .
we thus have @xmath115 suppose @xmath116 .
then @xmath75 and the other @xmath117 strategies in @xmath32 are the minority strategy , and the agents selecting those strategies win the game at @xmath104 .
the time series @xmath118 at time @xmath119 can be written as @xmath120 \nonumber \\ & + & ( n - n_{s_i}^{(1)})[p\frac{1}{k_{g_1}}+(1-p)m\frac{1}{k } ] \nonumber\\ & = & n_{s_i}^{(1)}-(w_4+w_5)+(w_6+w_7 ) , \label{eq : nsi2}\end{aligned}\ ] ] where @xmath121 and @xmath122 stand for the flows out of strategy @xmath75 , while @xmath123 and @xmath124 represent the flows into @xmath75 from those @xmath114 agents on other strategies .
we have @xmath125 where @xmath118 is larger than @xmath10 , @xmath75 and all other strategies in group @xmath32 will be the majority strategy again , as at time @xmath126 .
the process @xmath127 thus occurs iteratively . from eqs .
( [ eq : nsi+ ] ) and ( [ eq : nsi++ ] ) , we can get the number of agents for one given strategy at any time @xmath128 .
in particular , denoting @xmath129 , @xmath130 , and @xmath131 , we obtain the iterative dynamics for agents selecting strategy @xmath75 as @xmath132 where @xmath133 and @xmath134 stands for the number of agents at @xmath135 . carrying out the iterative process in eq .
( [ eq : nsib ] ) , we obtain @xmath136 for the case where @xmath72 exhibits stable oscillations , i.e. , the system is in a _ stationary state _ , we have , @xmath137 we thus obtain the values of @xmath138 and @xmath139 as a function of the probability @xmath0 , mutation probability @xmath25 , and the grouping parameter @xmath49 , and @xmath9 : @xmath140 from @xmath12 , we can get the expression of the amplitude of the fluctuation , the mean value @xmath141 , and its difference from @xmath10 as , @xmath142/2=\frac{\gamma_{g_1}}{2(1+\alpha ) } , \\
\nonumber & & \langle{n_{s_i}}\rangle=[n_{s_i}^{(2a)}+n_{s_i}^{(2a+1)}]/2 = \frac{2\beta+\gamma_{g_1}}{2(1-\alpha ) } , \\
\nonumber & & \delta n_{s_i}=|\langle{n_{s_i}}\rangle - n / k|=|\frac{k\gamma_{g_1}-2np}{2k(1-\alpha)}| = \frac{np|k / k_{g_1}-2|}{2k(1-\alpha)}.\end{aligned}\ ] ] in the above derivation , we have assumed that @xmath75 belongs to group @xmath32 , and obtained expressions of the @xmath98 , @xmath89 , @xmath141 , and @xmath95 .
similarly , using eq .
( [ eq : conv ] ) , we can calculate the time series @xmath99 , the number of agents selecting the strategy @xmath76 belonging to group @xmath33 , and the corresponding characterizing quantities . alternatively , following the steps similar to those from eqs .
( [ eq : nsi ] ) to ( [ eq : nsib ] ) , we can write the recurrence formula for @xmath99 as @xmath143 where @xmath144 . for the case of stable strategy
, we get the corresponding number of agents as @xmath145 we find that the values of @xmath12 obtained from eq .
( [ eq : nsj2a+1 ] ) agree with those from eq .
( [ eq : conv ] ) very well ( derivation and data not shown ) .
in addition , the expressions of @xmath94 , @xmath146 , and @xmath96 are identical to those associated with @xmath75 , with the quantity @xmath147 replaced by @xmath148 .
the mean - field theory is ideally suited for fully connected networks .
indeed , results from the theory and direct simulations agree with each other very well , as shown in fig . [
fig : mf ] .
however , in real - world situations , a fully connected topology can not be expected , and the mean - field treatment will no longer be accurate .
for example , we have carried out simulations on square - lattice systems and found noticeable deviations from the mean - field prediction . to remedy this deficiency ,
we develop a modified mean - field analysis for mg dynamics on sparsely homogeneous networks ( e.g. , square lattices or random networks ) . due to the limited number of links in a typical large - scale network , it is possible for a failed agent to be surrounded by agents from the same group ( who will likewise fail the game ) . in this case
, the failed agent has no minority strategy to imitate ( set @xmath20 is empty ) and thus will randomly select one strategy from the @xmath9 available strategies . taking this effect into account
, we can modify the mean - field approximation in eqs .
( [ eq : nsi ] ) and ( [ eq : nsi2 ] ) as @xmath149 with the two modified terms given by @xmath150 where @xmath151 is the probability for one agent in group @xmath152 to be surrounded by agents from the same group .
the quantity @xmath153 stands for the flow from the failed agents in group @xmath154 [ the number is @xmath155 , who react to the information [ the number is @xmath156 but with no winner surrounded to supply the optional minority strategy [ the number is @xmath157 , and thus select @xmath75 with probability @xmath158 .
the quantity @xmath159 represent two factors : ( 1 ) the @xmath160 failed agents in group @xmath33 who should have flow into @xmath75 [ i.e. @xmath123 ] but are held back because they are surrounded by agents in @xmath33 , and ( 2 ) the failed agents in group @xmath33 who are surrounded by agents in @xmath33 and thus select @xmath75 with @xmath158 probability , the number of which is @xmath162 .
apparently , we have @xmath163 . from eq .
( [ eq : mf22 ] ) , we obtain @xmath164 where the parameters are @xmath165 .
\nonumber\end{aligned}\ ] ] the equation set ( [ eq : mfn ] ) represents the modified mean - field description of the time series associated with the stable strategies in the game system supported on sparsely homogeneous networks .
the density of agents in @xmath152 is denoted by @xmath166 . for the case where agents from different groups are well mixed in the network ,
the probability for one given agent in @xmath152 to meet with agents in @xmath167 is @xmath168 ( for @xmath169 ) .
if the average degree of the network is @xmath18 , the probability that one agent from @xmath152 is surrounded by agents from the same group is @xmath170 . from the simulation on the square lattice system where each agent has @xmath171 neighbors
, we observe a quite weak effect of clustering of agents in @xmath32 or @xmath33 , so @xmath172 .
based on the quantity @xmath151 , analytic prediction of the time series in the modified mean - field ( mmf ) theory can be obtained , as shown in fig .
[ fig : mf ] for a square - lattice system .
we observe a good agreement with simulation results .
simulations on homogeneous networks of different values @xmath18 have also been carried out , with results in good agreement with the prediction from the modified mean - field theory . our mean - field treatment yields formula characterizing the stable oscillations associated with the grouping state @xmath173 , which include the variance ratio of the pairing groups [ eq . ( [ eq : i5 ] ) ] and time series @xmath174 [ eqs . ( [ eq : nsi2a+1 ] ) and ( [ eq : mfn ] ) ] .
however , from the simulation result shown in fig .
[ fig : k16 ] , we see that not all the grouping states are stable in the parameter space . as @xmath0 is increased , the strategy - grouping state of the smaller @xmath49 becomes stable , and the corresponding branch appears .
it is therefore useful to analyze the stability of the grouping state . in our treatment
we have assumed @xmath103 and @xmath175 .
then , the necessary condition for the grouping state @xmath176 to become stable is @xmath116 and @xmath177 . using eqs .
( [ eq : nsi2a+1 ] ) and ( [ eq : nsj2a+1 ] ) , we get @xmath178 where @xmath179 and @xmath180 are continuous functions of the parameters @xmath0 and @xmath25 , and @xmath181 . the two inequalities in eq .
( [ eq : stability_1 ] ) are nevertheless equivalent to each other .
figure [ fig : pm ] presents a phase diagram in the parameter space , where the curves of @xmath182 for @xmath183 , @xmath184 , @xmath8 , @xmath185 are shown .
the necessary condition for the strategy - grouping state with @xmath186 to be stable is that the the parameters @xmath0 and @xmath25 are in the upper - right region of the curve @xmath182 .
for certain value of @xmath25 , only when @xmath187 will the state of @xmath186 be stable . while the value of @xmath40 from simulation is different from the theoretical value @xmath188 , our mean - field theory does provide a qualitative explanation for the phenomenon in fig .
[ fig : k16](a ) , where more branches of smaller strategy - grouping states become stable as @xmath0 is increased .
we have also seen in fig .
[ fig : variance ] , and fig .
[ fig : k16](a ) that , as @xmath0 approaches @xmath185 , the bifurcated branches of different grouping state merge together . for the case of even @xmath9
, @xmath12 fluctuates stably in the grouping state with @xmath189 .
while , for the case of odd @xmath9 , @xmath12 fluctuates quasiperiodicly [ see fig .
[ fig : ts](c ) ]
. actually , the grouping state always switches between @xmath190 and @xmath191 , with @xmath192 .
we can understand the instability and merging of grouping states from fig .
[ fig : ts](c ) , and the schematic map in fig .
[ fig : sketchmap ] , as follows . as @xmath0
is increased to @xmath185 , @xmath95 and @xmath96 increase and become comparable to the amplitudes @xmath89 and @xmath94 , respectively .
namely , the attendances @xmath12 of strategies can be very close to @xmath11 . in case that @xmath12 of one strategy does not get cross @xmath5 because of _ noise _ , i.e. , it acts as minority ( or majority ) strategy twice , then , the fluctuation of @xmath12 , as well as the grouping state is changed .
the financial market is a representative multi - resource complex system , in which many stocks are available for investment .
we analyze the fluctuation of the stock price from the empirical data of 27 stocks in the shanghai stock market s steel plate between 2007 and 2010 .
we regard the @xmath2 stocks , which are derived from the iron and steel industry , as constituting a mg system with @xmath193 resources , where the agents selecting the resources correspond to the capitals invested .
this system is open in the sense that capital typically flows in and out , which is the main difference from our closed - system model . in particular , given the time series @xmath194 of the daily closing price of stock @xmath17 , the daily log - return is defined as @xmath195 .
the average return of the @xmath2 stocks at time @xmath128 , denoted by @xmath196 , signifies a global trend of the system at @xmath128 , which is caused by the change in the total mount of the capital in this open , 27-stock system .
however , when we analyze the detrended log - returns @xmath197 , the system resembles a closed system , as our model mg system .
we shall demonstrate that the strategy - grouping phenomenon occurs in this real - world system .
we calculate the pearson parameter @xmath198 of each pair of the detrended log - returns @xmath199 and @xmath200 , which leads to a @xmath201 correlation matrix @xmath202 , as shown in fig .
[ fig : matrix](a ) . in terms of the eigenvector associated with the maximum eigenvalue of matrix @xmath202 ,
we rank the order of the stocks and obtain the matrix @xmath203 , as shown in fig .
[ fig : matrix](b ) . _
the striking behavior is that the matrix is apparently divided into 4 blocks , a manifestation of the grouping phenomenon . _ in particular , the matrix elements @xmath198 among the first 15 stocks and those among the remaining 12 stocks are generally positive , but the cross elements between stocks in the two groups are negative .
it is thus quite natural to classify the first 15 stocks as belonging to group @xmath32 and the remaining 12 to group @xmath33 .
we can then write the matrix in a block form as @xmath204 where the elements of @xmath205 and @xmath206 are positive , and those of @xmath207 and @xmath208 are negative .
the phenomenon is that the @xmath2-stock system has self - organized itself into a @xmath209 grouping state , which is a natural consequence of the mg dynamics in multi - resource complex systems . for one given stock @xmath17 ,
the mean absolute correlation is @xmath210 this parameter reflects the weight of the stock in the system . if @xmath211 , oscillations of stock @xmath17 are contained in the noise floor . in this case , there is no indication as to whether this stock belongs to group @xmath32 or @xmath33 .
the larger the value of @xmath212 , the less ambiguous that the stock belongs to either one of the two groups . from the value of @xmath212 ranked in the same order as in @xmath203
, we can see that the boundary of the two groups is the stock with minimum @xmath212 .
thus @xmath212 can be considered as the characteristic number to distinguish different groups .
we can also reorder the matrix @xmath203 according to @xmath212 within group @xmath32 and @xmath33 , respectively .
this leads to the matrix @xmath213 , as shown in fig .
[ fig : matrix](c ) , further demonstrating the grouping phenomenon .
minority game , since its invention about two decades ago , has become a paradigm to study the social and economical phenomena where a large number of agents attempt to make simultaneous decision by choosing one of the available options @xcite . in the most commonly studied case of a single available resource with players two possible options ,
agents taking the minority option are the guaranteed winners .
various minority game dynamics have also received attention from the physics community due to their high relevance to a number of phenomena in statistical physics .
it has become more and more common in the modern world that multiple resources are available for various social and economical systems .
if the rule still holds that the winning options are minority ones , the questions that naturally arise are what type of collective behaviors can emerge and how they would evolve in the underlying complex system .
our present work aims to address these questions computationally and analytically .
the main contribution and findings of this paper are the following .
firstly , we have constructed a class of spatially extended systems in which any agent interacts with a finite but fixed number of neighbors and can choose either to follow the minority strategy based on information about the neighboring states or to select one randomly from a set of available strategies .
the probability to follow the local minority strategy , or the probability of minority preference , is a key parameter determining the dynamics of the underlying complex system .
secondly , we have carried out extensive numerical simulations and discovered the emergence of a striking collective behavior : as the minority - preference probability is increased through a critical value , the set of available strategies / resources spontaneously break into pairs of groups , where the strategies in the same group are associated with a specific fluctuating behavior of attendance .
this phenomenon of strategy - grouping is completely self - organized , which we conjecture is the hallmark of mg dynamics with multiple resources .
thirdly , we have developed a mean - field theory to explain and predict the emergence and evolution of the strategy - grouping states , with good agreement with the numerics .
fourthly , we have examined a real - world system of a relatively small - scale stock - trading system , and found unequivocal evidence of the grouping phenomenon .
our results suggest grouping of resources as a fundamental type of collective dynamics in multiple - resource mg systems
. other real - world systems for which our model is applicable include , e.g. , hedge - fund portfolios in financial systems , routing issues in computer networks and urban traffic systems .
we expect our model and findings to be not only relevant with statistical - physical systems , but also important to a host of social , economical , and political systems .
additionally , the double - grouping ( or paired grouping ) period - doubling like bifurcation are not the exclusive mode for the grouping of resources .
we have also observed the period@xmath214 double - grouping [ see fig . [
fig : mapping](a)(b ) ] , and the period@xmath214 triplet - grouping bifurcation phenomena [ see fig . [
fig : mapping](c ) ] .
further research about these phenomena are of great interest and is valuable for the understanding of chaotic and bifurcation in social systems .
zgh thanks prof . matteo marsili for helpful discussions .
this work was partially supported by the nsf of china ( grant nos . 10905026 , 10905027 , 11005053 , and 11135001 ) , srfdp no .
20090211120030 , frfcu no .
lzujbky-2010 - 73 , lanzhou dstp no .
2010 - 1 - 129 .
ycl was supported by afosr under grant no .
fa9550 - 10 - 1 - 0083 .
m. anghel , z. toroczkai , k. e. bassler and g. korniss , _ competition - driven network dynamics : emergence of a scale - free leadership structure and collective efficiency _ , phys .
lett . * 92 * , 058701 ( 2004 ) . | the minority game ( mg ) has become a paradigm to probe complex social and economical phenomena where adaptive agents compete for a limited resource , and it finds applications in statistical and nonlinear physics as well . in the traditional mg model , agents are assumed to have access to global information about the past history of the underlying system , and they react by choosing one of the two available options associated with a single resource .
complex systems arising in a modern society , however , can possess many resources so that the number of available strategies / resources can be multiple .
we propose a class of models to investigate mg dynamics with multiple strategies .
in particular , in such a system , at any time an agent can either choose a minority strategy ( say with probability @xmath0 ) based on available local information or simply choose a strategy randomly ( with probability @xmath1 ) .
the parameter @xmath0 thus defines the _ minority - preference probability _ , which is key to the dynamics of the underlying system .
a striking finding is the emergence of strategy - grouping states where a particular number of agents choose a particular subset of strategies .
we develop an analytic theory based on the mean - field framework to understand the `` bifurcation '' to the grouping states and their evolution .
the grouping phenomenon has also been revealed in a real - world example of the subsystem of @xmath2 stocks in the shanghai stock market s steel plate .
our work demonstrates that complex systems following the mg rules can spontaneously self - organize themselves into certain divided states , and our model represents a basic mathematical framework to address this kind of phenomena in social , economical , and even political systems . |
in the study of finite quandles and the knot invariants they define , decomposable quandles are often neglected in favor of indecomposable or connected quandles since for knots , two decomposable quandles with the same orbit decomposition define the same counting invariant : if @xmath0 then @xmath1 the sum of the counting invariants defined by the orbit subquandles , where @xmath2 denotes the cardinality of the set @xmath3 .
this follows from the fact that knot quandles have a single orbit , so the image of any homomorphism @xmath4 from a knot quandle @xmath5 into a finite target quandle @xmath6 must lie inside a single orbit .
the quandle of a multi - component link , however , is not connected , and the counting invariants defined by nonisomorphic decomposable quandles with the same orbit decomposition are generally different for links . the motivation for this paper was to study the counting invariants defined by decomposable quandles for multi - component virtual links .
we started with the case of finite target quandles whose orbits are trivial subquandles , so that all of the interesting structure is in the off - diagonal portion of the quandle s matrix trivial subquandles which have been put together in a non - trivial way .
we call such quandles _ trivial orbit quandles _ or
_ toqs_. our main result says that the linking number of a two - component classical link is determined up to sign by the counting invariants for a family of trivial orbit quandles of a particular form .
the paper is organized as follows : in section [ q ] we review the definitions of quandles , the quandle counting invariant and virtual linking numbers . in section [ l ]
we give our main result and some examples .
in this section we review some definitions .
@xmath7 @xmath8 next , we need we may specify a quandle structure on the finite set @xmath9 by giving an @xmath10 matrix @xmath11 whose entry in row @xmath12 column @xmath13 is @xmath14 where @xmath15 ; that is , @xmath11 is the operation table of @xmath16 where we forget the `` @xmath17 ' 's and just write the subscripts .
quandles and their structure have been studied in many works such as @xcite , @xcite , @xcite and more .
a quandle @xmath18 is _ decomposable _ if we may write @xmath0 where @xmath19 and @xmath20 are both subquandles of @xmath18 .
note that even if a quandle is not decomposable in this sense , we may still be able to write @xmath21 for @xmath22 disjoint subquandles ; see @xcite and @xcite .
a quandle is _ connected _ if there is an element @xmath23 such that every element of @xmath18 is equivalent to a word of the form @xmath24 in @xcite ( see also @xcite ) , a quandle is defined for any oriented tame classical link @xmath25 via a wirtinger - type presentation , which is called the _ knot quandle _ of the link @xmath25 .
specifically , looking in the positive direction of the overcrossing strand @xmath26 , if the undercrossing arc on the right is @xmath27 then the one on the left is @xmath28 .
note that we do not use the orientation of the undercrossing strand .
a @xmath29 the isomorphism type of the knot quandle is invariant under reidemeister moves ; indeed , the quandle axioms are chosen specifically to make this so . if @xmath25 is a knot and @xmath17 is a generator assigned to an arc in @xmath25 , then every generator ( and hence every word ) in the knot quandle is equivalent to a word starting with @xmath17 since we can just proceed around the knot following the specified orientation , picking up quandle operations to build a word , until we reach @xmath17 .
hence , the knot quandle of a single - component link is connected .
while joyce defined the knot quandle only for classical knots and links , in @xcite kauffman observed that the same wirtinger - type definition extends to virtual knots and links by simply ignoring virtual crossings .
the resulting quandle , while not a classifying invariant for virtual knots and links , is nevertheless an invariant of virtual isotopy . given a finite quandle @xmath6 ,
the number of homomorphisms from a knot quandle @xmath30 into @xmath6 is an easily computable link invariant : since @xmath30 is finitely generated , there is a finite set of possible homomorphisms into @xmath6 , so we can simply check which of these satisfy the relations in @xmath30 and hence are homomorphisms . the following is a standard result : let @xmath25 be a link with knot quandle @xmath30 and let @xmath6 be a finite quandle .
then the number of homomorphisms @xmath31 is an invariant of quandle isomorphism and hence is an invariant of virtual isotopy of the link .
indeed , joyce @xcite showed that the knot quandle determines both the knot group and a longitude , and hence determines the knot type of a classical knot up to homeomorphism ( not necessarily orientation preserving ) of pairs @xmath32 .
hence , many of the classical link invariants should be derivable from the knot quandle
. a deeper understanding of how link invariants arise from the knot quandle should provide us with a better understanding both of how other link invariants are related and of links themselves . in the next section we will show how linking number is determined up to sign by the quandle counting invariants associated to a particular kind of non - connected quandle .
we begin with a discussion of linking number for virtual knots . for classical links , the two linking numbers are equal since if @xmath33 crosses over @xmath34 at a positive crossing , it must either cross back over @xmath34 at a negative crossing ( contributing zero to both linking numbers ) or cross under @xmath34 at a positive crossing ( contributing @xmath35 to both linking numbers ) to come back to join itself without any virtual crossings . for virtual links ,
the two linking numbers are completely independent .
indeed , we have for virtual links , the virtual linking numbers have some advantages over the usual classical linking number . for example , if @xmath36 is split , then both @xmath38 and @xmath39 ; this is a stronger condition than @xmath40 , since contributions from two unequal virtual linking numbers may cancel .
as noted in @xcite , the first two quandle axioms are easy to check since the diagonal entries are @xmath45 in order and each column is a permutation of @xmath46 .
thus , we need only check that the third quandle axiom is satisfied .
we do this in cases .
case 1 : @xmath47 all in the same subquandle . in this case , @xmath48 since the subquandles are quandles themselves . for the remainder of this paper ,
let @xmath25 be a two - component virtual link diagram @xmath36 . to find quandle colorings of @xmath25
, we take a choice of color @xmath75 for one arc and attempt to `` push it through '' each crossing , that is , assign to the opposite undercrossing arc the color determined by the quandle coloring condition and the color on the overcrossing arc . since our coloring quandle @xmath42 is a trivial orbit quandle , pushing any color through at a crossing where both strands are from the same component yields the same color on both underarcs since the colors on a component are from the same trivial orbit subquandle .
thus we have : in a quandle coloring of a link by @xmath42 ( or any trivial orbit quandle ) , the colors on the undercrossing strands at a crossing can differ only if the overcrossing and undercrossing strands belong to distinct components .
[ l0 ] in a quandle coloring of @xmath25 by @xmath42 , an inbound undercrossing arc colored @xmath76 pushes through a crossing with an overarc colored @xmath77 to become @xmath78 at a positive crossing and @xmath79 at a negative crossing . pushing through any choice of color for a starting arc
extends the same color along that component of the link until overcrossings from the other component of @xmath25 are reached .
for any pair of starting arcs with one arc on each component , if both starting colors are from the same orbit , then both colors push through unchanged and we have a coloring of the link which is constant on each component . if the colors are from different orbits in @xmath42 , then one component has every arc colored @xmath77 , and a given starting coloring @xmath60 pushes through an overcrossing colored with @xmath77 to become @xmath82 or @xmath83 depending on the crossing sign . continuing pushing through each crossing , we eventually reach our original starting arc .
then we have a valid quandle coloring iff the color thus determined on the starting arc agrees with the original color . in both cases
( starting arcs in the same orbit subquandle or in different orbit subquandles ) , the colors on all of the arcs are determined by the colors on the initial two arcs .
there are @xmath81 such choices .
if we color one arc of @xmath84 with @xmath89 , then in any quandle coloring of @xmath25 every arc of @xmath84 must be colored with @xmath77 , since every element of @xmath42 acts trivially on @xmath77 . in order to have a valid quandle coloring
, the original color on the starting arc must match the color determined by pushing @xmath17 through the full length of the component to return to the starting arc .
lemma [ l0 ] implies that this final color is @xmath90 , so an initial color of @xmath91 on an arc of @xmath86 pushes through to a valid coloring iff @xmath92 .
finally , note that the property of pushing through does not depend on the actual value of @xmath91 , only its orbit , so if one such color yields a valid quandle coloring , then all @xmath88 colors in @xmath85 yield valid quandle colorings .
we have @xmath94 if @xmath25 has colorings with @xmath77 on @xmath84 and colors from @xmath85 on @xmath86 for both @xmath95 ; @xmath96 if @xmath25 has colorings with @xmath77 on @xmath84 and colors in @xmath85 on @xmath86 for @xmath97 or @xmath98 but not both ; and @xmath99 if @xmath25 has no multi - orbit colorings .
note that @xmath100 implies that the virtual linking numbers are unequal and @xmath25 is non - classical .
@xmath94 if and only if each starting coloring choice for a given pair of arcs on the two components pushes through to a valid coloring . by lemma [ l1
] , this can only happen if both virtual linking numbers are zero mod @xmath88 , i. e. , if @xmath107 and @xmath108 conversely , suppose @xmath110 for all @xmath114 .
then by corollary [ c1 ] at least one of the two virtual linking numbers is greater than 0 in absolute value .
if both are greater than 1 in absolute value , then let @xmath119 , contrary to hypothesis .
hence either @xmath111 , @xmath112 , or both @xmath115 .
if @xmath121 then neither linking number is 0 or 1 in absolute value by corollaries [ c1 ] and [ c2 ] .
then @xmath124 says that @xmath88 divides one linking number but not the other by lemma [ l1 ] ; @xmath125 is the largest number which divides both @xmath126 and @xmath127 by corollary [ c4 ]
. if both virtual linking numbers are zero , then corollary [ c1 ] says @xmath130 .
if both are @xmath131 , then by corollary [ c2 ] we have @xmath132 . finally , if @xmath133 ( which equals @xmath127 since @xmath25 is classical ) then the set @xmath134 includes all and only positive integer divisors of @xmath135 , so @xmath136
as a practical observation , we note that corollary [ c3 ] implies that it is not necessary to test an infinite number of quandle counting invariants to determine @xmath138 ; it is sufficient to compute @xmath80 for @xmath139 where @xmath140 is the number of multi - component crossings in @xmath25 .
of course , our interest in theorem [ main ] lies not in finding a new less practical method of computing linking numbers but rather as a connection between quandle - based link invariants and classical link invariants . | we study the quandle counting invariant for a certain family of finite quandles with trivial orbit subquandles .
we show how these invariants determine the linking number of classical two - component links up to sign .
keywords : virtual links , linking number , finite quandles 2000 msc : 57m27 , 55m25 |
the _ fermi _ satellite was launched on 2008 june 11 and has since greatly increased the number of known pulsars emitting in the @xmath0-ray band . the principle instrument aboard
the satellite is the large area telescope ( lat ; @xcite ) , a pair - production telescope which is sensitive to energies in the range 20 mev - 300 gev .
the impact of the mission so far can be seen from the 2nd _ fermi _ pulsar catalog @xcite , which details 117 detections ( at @xmath2 100 mev ) of @xmath0-ray - loud pulsars .
this factor of 20 increase in the number of pulsars which can be studied in the high energy regime makes investigation of the pulsar @xmath0-ray emission process more important than ever .
knowledge of the ` viewing geometry ' of these pulsars can significantly aid this study .
particularly important is the determination of the path traced across the magnetosphere by the observer s line of sight as the pulsar rotates , which can be characterised by two angles ; the angle between the rotation and magnetic axes , @xmath1 , and the angle @xmath3 between the magnetic axis and the line of sight at the closest approach between the two . if these are known for a particular pulsar , a given model of the sky pattern of @xmath0-ray emission ( such as those described in @xcite ) will be able to predict , for example , the shape of that pulsar s intensity profile
this can then be compared with observations , allowing the veracity of the model to be judged .
the determination of the viewing geometry is therefore useful in the study of the pulsar emission mechanism .
a recent study by @xcite highlighted a general problem that @xmath0-ray light - curves can only be meaningfully interpreted when radio data are also considered .
these authors used @xmath0-ray light - curve and radio profile fitting to constrain the viewing geometry in terms of @xmath1 and @xmath4 , the angle between the line of sight and the rotation axis .
they note that if they consider radio - loud pulsars , but ignore their radio profiles in the fitting procedure , the resulting @xmath3 values do not appear to be confined to small values .
this is unlikely to be realistic , given the fact that the radio emission should be significantly more beamed compared to the @xmath0-rays .
they conclude that @xmath0-ray light - curve fitting in general does not lead to unique solutions in @xmath1 and @xmath5 space and that radio observations are crucial to break this degeneracy . the intention of this paper is to derive emission geometries , which can facilitate the interpretation of @xmath0-ray light - curves and therefore potentially impose important constraints on various proposed @xmath0-ray models . using data obtained with the parkes radio telescope as part of the _ fermi _ timing programme @xcite we attempt to place constraints on the viewing geometry for 28 young , @xmath0-ray - detected pulsars , all of which were included in @xcite . as the angles @xmath1 and @xmath3 should be independent of observing frequency [ j0908 ] )
. however , the intrinsic values are frequency - independent and the uncertainty of a few degrees introduced by scattering is in general not the most significant source of uncertainty in the sample considered in this paper . ]
, the results derived here can be applied to observations of these pulsars at any part of the electromagnetic spectrum .
a further point which is worthy of note is that , although there are large uncertainties on most of the @xmath1 values presented in this paper , @xmath3 will be shown to be well constrained for most of the sample .
this is useful for attempts to constrain the @xmath0-ray models as requiring a small value of @xmath3 greatly restricts the region of ( @xmath1 , @xmath5 ) parameter space which is of interest , helping to lift the aforementioned degeneracy
. one method of constraining @xmath1 and @xmath3 is to fit the rotating vector model ( rvm ; @xcite ) to the observed polarisation position angle ( pa ) swing .
the pa of the linearly polarised emission , @xmath6 , is a function of @xmath1 , @xmath3 and the pulse phase , @xmath7 , according to @xmath8 where @xmath9 and @xmath10 are the position angle and pulse phase , respectively , of the steepest part of this curve @xcite .
the steepest gradient of the rvm curve is given by @xmath11 @xcite .
if the observed pa swing shows little curvature , only its gradient is constrained , leading to a constraint on @xmath12/@xmath13 ( see @xmath14 [ sectchi2surfaces ] for details of the fitting procedure ) . in
( @xmath1 , @xmath3 ) space this results in a characteristic ` banana ' shape of the goodness - of - fit , which can be seen in , for example , fig .
[ figj0631 + 1036 ] .
the steepest gradient occurs at @xmath10 , the ` inflection point ' of the pa curve .
the rvm predicts that this coincides with the passage of the line of sight through the ` fiducial plane ' , the plane containing both the rotation and magnetic axes .
however , in practice observations show the inflection point to be delayed relative to the position of the fiducial plane as inferred from the intensity profile ( @xmath15 ) by an amount of rotational phase @xmath16 .
this delay is predicted by relativistic effects known as aberration and retardation ( a / r ) .
the net delay predicted by a / r effects is @xmath17 where @xmath18 is the rotation period of the star , @xmath19 is the speed of light and @xmath20 is the emission height , the distance of the emission region from the centre of the star @xcite .
the observed delay can be shown to be independent of @xmath1 and @xmath3 @xcite .
estimates can be obtained for @xmath10 from rvm fitting ( see @xmath14 [ sectchi2surfaces ] ) and for @xmath15 based on the profile morphology ( see @xmath14 [ sectemissionheightconstraint ] ) . given these estimates @xmath21 can be determined from the data and so eq . [ eqoffset ] can be used to calculate @xmath20 . as discussed by for instance @xcite
, variations in the emission height between different components will cause distortions in the pa curve and hence could affect the determination of @xmath21 .
this effect is probably strongest for pulsars older than those discussed in this paper , hence we assume @xmath20 to be constant across the emission region for a given pulsar .
the potential effects of variations in the emission height will be considered in @xcite . the angles @xmath1 and @xmath3 are also related to another observable , the pulse width of the radio profile .
if the radio beam is defined as the cone bounded by tangents to the last open field lines at the emission height , the half - opening angle of the beam , @xmath22 , can be calculated .
assuming that the magnetic field is dipolar , @xmath23 with @xmath24 , the angular radius of the open - field - line region , given by @xmath25 ( e.g. , @xcite ) .
@xcite showed that the half - opening angle of a conical emission beam centred on the magnetic axis can be related to @xmath26 , the range in pulse phase for which the line of sight samples the open - field - line region .
this relation is dependent on the viewing geometry via the equation @xmath27 a similar method of combining this information has been used by previous authors ( e.g. , @xcite ) . however
, this work differs from those in two important ways .
firstly , we consider uncertainties on the relevant parameters , which allows us to determine errorbars on the fit parameters in a more objective way . secondly , although we use the core - cone model @xcite as a basis for our estimation of @xmath15 , we argue that the conservative nature of these estimates means that our results are unlikely to be inconsistent with the patchy beams suggested by @xcite ( see @xmath14 [ sectemissionheightconstraint ] ) .
the paper is organised as follows : in @xmath14 [ sectmethod ] we describe the observations and methodology used . in @xmath14 [ sectresults ] we present intensity and polarisation profiles and viewing geometry constraints for the individual pulsars and a table with constraints is compiled . in @xmath14 [ sectalphadistribution ]
we discuss the derived distributions of magnetic inclination angles for pulsars exhibiting single - component and multiple - component profiles .
we also compare our results with those obtained from @xmath0-ray modeling by @xcite .
@xcite will discuss the overall distribution of @xmath1 values , and its dependence on assumptions about the radio beam .
the data used were collected as part of the _ fermi _ timing programme at the parkes radio telescope , described in detail by @xcite . a total of 168 pulsars
are observed drawn from a list of pulsars with large - spin - down rates , and hence potentially high energy @xmath0-ray emission , given by @xcite supplemented by a few additional objects .
the programme consists of monthly observations at 1369 mhz ( 20 cm ) with a 256 mhz bandwidth and twice - annual observations of the same pulsars at 3100 mhz ( 10 cm ) and 685 mhz ( 50 cm ) , with bandwidths of 1024 mhz and 40 mhz respectively @xcite . at each frequency
the band is divided into 1024 channels and the pulse phase resolution is 1024 bins per pulse . in this paper
we concentrate on the sample of 28 pulsars which are @xmath0-ray - loud and published in the second fermi pulsar catalogue @xcite .
we use observations taken between 2007 april and 2013 october . for the majority of pulsars ,
the observations at 1369 mhz were used .
some pulsars however are significantly scatter - broadened due to propagation through the interstellar medium at this frequency and so the observations at 3100 mhz were used instead ( see table [ tabderivedparameters ] ) . in the cases of psrs
j10195749 and j14106132 the profile is still severely scatter - broadened at 3100 mhz ; in the latter case we used unpublished parkes data at 6100 mhz .
the data were polarisation calibrated and individual observations of the same pulsar were summed largely following the method described in @xcite , resulting in a single profile for each pulsar .
the pa as a function of phase , @xmath28 , was calculated from the stokes parameters for each pulsar .
any pa values for which the signal - to - noise ratio of the linearly polarised component of the emission was less than 2@xmath29 were discarded .
only the remaining values were used in subsequent fitting of eq .
[ eqpa ] . in order to constrain the geometrical parameters ,
a gridsearch was performed in ( @xmath1 , @xmath3 ) space .
the least-@xmath30 fit between the rvm curve and the data was then determined by optimising the remaining free parameters @xmath10 and @xmath9 .
the obtained constraints on @xmath1 and @xmath3 are sensitive to the pulse phase resolution of the data . decreasing the resolution increases the signal - to - noise ratio per bin . as a result ,
the significance of the linear intensity in bins at the edges of the pulse is increased , allowing the pa to be determined over a wider range of pulse phase .
however , decreasing the resolution can potentially result in distortion of the pa swing ( especially where the gradient is largest ) leading to an unreliable rvm fit . for each pulsar , @xmath30 surfaces
were calculated using 1024 , 512 , 256 and 128 bins per rotation period and the optimum resolution ( see table [ tabderivedparameters ] ) was determined and used for all subsequent processing of that pulsar .
the reduced-@xmath30 values resulting from this fitting process correspond to a surface in ( @xmath1 , @xmath3 ) space , an example of which can be seen in fig . [ figj0631 + 1036 ] .
three contours are shown , corresponding to 2 , 3 and 4 times the global least-@xmath30 value .
these represent 1@xmath29 , 2@xmath29 and 3@xmath29 uncertainties in @xmath1 and @xmath3 .
the error on @xmath10 was calculated by fitting for @xmath1 , @xmath3 and @xmath9 repeatedly as @xmath10 was varied , and determining the value of @xmath10 at which the resulting fit had a reduced-@xmath30 four times that of the global least-@xmath30 fit ( equivalent to a 3@xmath29 error ) , thereby recognising the fact that there are unmodeled features in the observed shape of the pa curve ] .
this range and the most likely value are displayed by the horizontal bar above or below the pa curve in figs .
[ figj0631 + 1036 ] - [ figj1835 - 1106 ] .
previous authors who have used rvm fitting and the a / r effect to determine viewing geometries ( e.g. , @xcite ) have traditionally assumed that the emission region fills the open field lines and hence that @xmath15 must be at the centre of the observed on - pulse region .
however , the emission beam may not be symmetrically illuminated . to account for such an effect , the assumption that @xmath15 is at the centre of the profile was relaxed in this investigation .
instead a range of possible values was determined based on the component positions , using the core - cone model as a basis @xcite and allowing for some components to be missing from the observed profile . the horizontal bar above the profile in figs .
[ figj0631 + 1036 ] - [ figj1835 - 1106 ] displays the chosen range and our preferred value .
alternatively , it is possible that the asymmetry is the consequence of a beam which is populated by randomly distributed ` patches ' @xcite . in this case
the core - cone model will not be valid .
however , as the @xmath15 ranges chosen typically cover a large proportion of the pulse they are likely to be consistent with a patchy beam , and if the patches are truly random we do not expect patchy beams to cause any bias in the derived viewing geometries .
this would therefore imply , for instance , that the derived @xmath1 distribution should still describe the population as a whole . as discussed in @xmath14 [ sectchi2surfaces ] , a range of allowed values for @xmath10
was determined from fitting to the pa curve .
this range , together with the range of @xmath15 determined from the profile shape , allows a range @xmath31 to be established .
the corresponding ranges of @xmath20 and @xmath22 were calculated using eqs .
[ eqoffset ] , [ eqbeamhalfopeningangle ] and [ eqthetapc ] .
finally , if @xmath26 is known , eq . [ eqrhocontours ] provides an additional constraint . in these equations @xmath22
is taken to be the opening angle of a beam corresponding to an emission region confined to the open - field - line region . to account for the possibility of beams which are only partially illuminated we take @xmath26 to be twice the separation ( in phase ) between the determined @xmath15 and the edge of the profile furthest from it .
the assumption we make is therefore that at least one edge of the profile reaches the last open field line and hence , as the open - field - line region is symmetric about the fiducial plane , @xmath26 must account for an equal amount of phase before and after @xmath15 .
there is an uncertainty in the precise location of the pulse edges due to the tapering nature of the emission and the presence of noise .
the points at which the intensity was 10% of the peak intensity were taken as the most likely to correspond to the last open field lines .
this definition makes the determination of the pulse edges independent of the s / n . in most cases
the uncertainties were taken to be the differences between the phases at which the intensity was 20% and 10% of the peak . however , in some cases such choices were clearly not representative of the pulse edge and so the limits were chosen somewhat more subjectively ( see @xmath14 [ sectresults ] ) .
the most likely location and the considered range for each pulse edge are displayed by the horizontal bars immediately below the profile in figs .
[ figj0631 + 1036 ] - [ figj1835 - 1106 ] . as a result of specifying upper and lower limits for each edge , upper and lower values
could also be determined for @xmath26 .
the parameters @xmath10 , @xmath15 and the pulse edges were varied within the allowed ranges and for each combination @xmath22 and @xmath26 were calculated .
. [ eqrhocontours ] was used to derive the corresponding contour in ( @xmath1 , @xmath3 ) space .
the full set of possible contours cover a region of ( @xmath1 , @xmath3 ) space ( the green regions in figs .
[ figj0631 + 1036 ] - [ figj1835 - 1106 ] ) , which represents the possible viewing geometries consistent with the profile width and the offset of the pa curve with respect to the fiducial plane .
such a consideration of the uncertainties on these parameters , which allows a more complete appraisal of the possible viewing geometries for each pulsar , has not been used previously when constraining viewing geometries .
the viewing geometry should then lie within the overlap between this region and the 3@xmath29 limit of the @xmath30 surface from rvm fitting ( indicated in greyscale in figs .
[ figj0631 + 1036 ] - [ figj1835 - 1106 ] , with a key displayed at the right - hand side of the plot ) .
in addition to this , a single contour ( orange curve in the figures ) , termed the `` favoured contour '' , was plotted for 25 of the 28 pulsars ( see @xmath14 [ sectalphadistribution ] ) .
the favoured contour was specified using the optimum @xmath10 value determined from the rvm fit , our preferred value for @xmath15 and the value of @xmath26 derived using this fiducial plane position and the most likely locations of the pulse edges . in some cases a further constraint
can be added to the viewing geometry due to the absence of an interpulse . assuming the presence of an illuminated beam with half - opening angle @xmath22 at each magnetic pole
, we expect to observe an interpulse when the line of sight passes within @xmath22 of both ends of the magnetic axis .
this occurs when the conditions @xmath32 and @xmath33 are both satisfied , where @xmath34 is the angle between the line of sight and the rotation axis .
the lack of an interpulse indicates that the second condition , which can be written as @xmath35 is not satisfied . if the minimum allowed @xmath22 value calculated from eq .
[ eqbeamhalfopeningangle ] is non - zero , the corresponding region of ( @xmath1 , @xmath3 ) space can be excluded , under the given assumptions .
the boundaries of this region are displayed in figs .
[ figj0631 + 1036 ] - [ figj1835 - 1106 ] ( where applicable ) as red diagonal lines .
in this section we present and discuss the constraints to the viewing geometry for each pulsar in the sample . table [ tabderivedparameters ] summarises the wavelength , rotation measure and pulse phase resolution used , the favoured values and allowed ranges determined for @xmath10 and @xmath15 , and the subsequently calculated values of @xmath21 , @xmath26 and @xmath20 for each pulsar .
derived half - opening angles of the beam , @xmath22 , and the allowed ranges of @xmath1 and @xmath3 are presented in table [ tabuncorrectedviewinggeometries ] .
psr j0631@xmath361036 at 20 cm .
the upper panel in the top plot shows the profile .
the black profile represents the total intensity and the red and green profiles give the intensity due to the linearly and circularly polarised components of the emission respectively .
the convention is used that circular polarisation intensity is positive for left - handed polarisation and negative for right - handed polarisation .
the horizontal bars show the preferred position and range of positions which were considered for the fiducial plane ( above the profile ) and the start and end of the pulse ( below the profile ) .
the lower panel shows the pa points with a significance greater than 2@xmath29 superimposed with the global least-@xmath30 rvm fit .
the horizontal bar in this panel shows the position of the inflection point including its 3@xmath29 error .
the bottom plot shows the ( reduced ) @xmath30 surface ( greyscale , with contours corresponding to 1@xmath29 , 2@xmath29 and 3@xmath29 ) overlain with the solutions consistent , within the uncertainties , with the a / r effect in combination with the observed pulse width ( green region ) .
the key at the right - hand side of the plot describes the magnitude of the reduced-@xmath30 .
the ` favoured ' contour , which corresponds to the preferred values of @xmath10 , @xmath15 and the pulse edges , is shown in orange . viewing geometries between the two red ( straight diagonal ) lines are inconsistent with the lack of an observed interpulse .
, title="fig : " ] psr j0631@xmath361036 at 20 cm .
the upper panel in the top plot shows the profile .
the black profile represents the total intensity and the red and green profiles give the intensity due to the linearly and circularly polarised components of the emission respectively .
the convention is used that circular polarisation intensity is positive for left - handed polarisation and negative for right - handed polarisation .
the horizontal bars show the preferred position and range of positions which were considered for the fiducial plane ( above the profile ) and the start and end of the pulse ( below the profile ) .
the lower panel shows the pa points with a significance greater than 2@xmath29 superimposed with the global least-@xmath30 rvm fit .
the horizontal bar in this panel shows the position of the inflection point including its 3@xmath29 error .
the bottom plot shows the ( reduced ) @xmath30 surface ( greyscale , with contours corresponding to 1@xmath29 , 2@xmath29 and 3@xmath29 ) overlain with the solutions consistent , within the uncertainties , with the a / r effect in combination with the observed pulse width ( green region ) .
the key at the right - hand side of the plot describes the magnitude of the reduced-@xmath30 .
the ` favoured ' contour , which corresponds to the preferred values of @xmath10 , @xmath15 and the pulse edges , is shown in orange . viewing geometries between the two red ( straight diagonal ) lines are inconsistent with the lack of an observed interpulse .
, title="fig : " ] the profile for this pulsar is highly symmetrical , which makes it likely that the fiducial plane is close to the centre of the profile .
however , to ensure that the constraint on the viewing geometry is conservative we have allowed the fiducial plane to be located at any point between the two central peaks .
the constraints on the viewing geometry of this pulsar are consistent with the earlier results of @xcite , who found @xmath37 km and @xmath38 . solutions with @xmath39 @xmath40 are excluded by the lack of a visible interpulse .
psr j0659@xmath361414 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j0659@xmath361414 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] this pulsar has a long history of polarization studies over a wide range of frequencies @xcite .
the profile is roughly triangular at our observing frequency .
however , the profile is not symmetric about the peak .
the leading edge shows depolarisation ( which may be indicative of a conal component ) .
the observed component could be the trailing component in a double , as noted for several pulsars by jw06 .
therefore the possibility that the fiducial plane is before the peak was included in the allowed range .
the range of possible locations of the trailing edge was extended to include the trailing component .
our results are on the viewing geometry are consistent with the values derived in @xcite .
psr j07291448 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j07291448 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] this profile is asymmetric at 1.4 ghz , which means there is ambiguity in the position of the fiducial plane . at 3.1 ghz the profile becomes double peaked , with the newly - apparent peak coinciding with the earlier component in this profile ( jw06 ) .
this implies that the fiducial plane is between these two components .
this is reflected in the choice of upper and lower limits on @xmath15 , which correspond to the peaks of the trailing and leading components .
jw06 found @xmath41 km and a corresponding @xmath42 .
they also found @xmath1 to be unconstrained while @xmath43 .
our values are consistent with these results .
we can exclude @xmath44 and @xmath45 which in turn indicates @xmath46 .
psr j07422822 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j07422822 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] polarization profiles of this pulsar have been presented most recently in @xcite , @xcite and @xcite and the pulsar was shown to alternate between two profile states by @xcite .
the profile is complex , with a boxy structure showing several prominent peaks and a smaller trailing component .
the evolution of the profile with frequency is also complex @xcite .
for this reason we assign a wide range to @xmath15 from @xmath47 to @xmath48 . as a consequence
this does not lead to a useful additional constraint in ( @xmath1 , @xmath3 ) space .
we note in passing that the orthogonal polarization mode ( opm ) jump at the trailing edge of the profile at 3.1 ghz seen by @xcite is clearly seen here for the first time at 1.4 ghz , at @xmath49 pulse phase , and has been included in the fit .
psr j08354510 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j08354510 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j08354510 ( vela ) is the pulsar on which the rvm was based @xcite .
it is the brightest pulsar in the sky at this wavelength and emission can be seen down to very faint levels ( 50 db below the peak ) .
multi - frequency observations strongly suggest that the peak seen at 1369 mhz is the core component followed by the trailing cone with the leading cone absent @xcite .
this is the motivation for our choice of @xmath15 .
@xcite claim @xmath50 , @xmath51 whereas @xcite have @xmath52 , @xmath53 .
x - ray observations of the torus by @xcite allowed a determination of @xmath54 to be 64@xmath55 , inconsistent with the @xcite result .
ng & romani suggest a solution with @xmath56 , @xmath57 .
recent @xmath0-ray constraints @xcite also prefer @xmath1 values close to 70@xmath55 .
we note the @xcite solutions are not consistent with our rvm fits .
our result would be strongly affected by what we assume as the overall pulse width .
taking the pulse edges to be at 10% of the peak intensity ( consistent with the rest of the paper ) leads to ( @xmath1 , @xmath3 ) = ( @xmath58 , @xmath59 ) .
this is in good agreement the @xcite constraint and the @xmath0-ray models .
this scenario is used in table [ tabderivedparameters ] and in subsequent analysis .
psr j09084913 at 20 cm . as fig .
1 . the profiles shown are the main pulse ( upper panel ) and the interpulse ( middle panel ) .
both profiles are normalised to the maximum intensity of the main pulse .
the edges of the interpulse were determined according to 10% of the interpulse maximum intensity .
the blue - green shading represents the superposition of the a / r constraints from the main pulse and from the interpulse .
it can be seen that neither of these constraints can exclude any of the displayed region of @xmath1 and @xmath3 .
, title="fig : " ] psr j09084913 at 20 cm . as fig .
the profiles shown are the main pulse ( upper panel ) and the interpulse ( middle panel ) .
both profiles are normalised to the maximum intensity of the main pulse .
the edges of the interpulse were determined according to 10% of the interpulse maximum intensity .
the blue - green shading represents the superposition of the a / r constraints from the main pulse and from the interpulse .
it can be seen that neither of these constraints can exclude any of the displayed region of @xmath1 and @xmath3 .
, title="fig : " ] psr j09084913 at 20 cm . as fig .
1 . the profiles shown are the main pulse ( upper panel ) and the interpulse ( middle panel ) .
both profiles are normalised to the maximum intensity of the main pulse .
the edges of the interpulse were determined according to 10% of the interpulse maximum intensity .
the blue - green shading represents the superposition of the a / r constraints from the main pulse and from the interpulse .
it can be seen that neither of these constraints can exclude any of the displayed region of @xmath1 and @xmath3 .
, title="fig : " ] this pulsar is one of two in the sample for which an interpulse is detected .
@xcite found a remarkably well - constrained @xmath60 using rvm fitting .
however , they also showed that the geometry they determined varied slightly with observing frequency , from ( @xmath1 , @xmath3 ) = ( 96.6@xmath55 , 8.1@xmath55 ) at 1.4 ghz to ( 96.1@xmath55 , 5.9@xmath55 ) at 8.4 ghz .
they attributed this variation to a small amount of scattering by the interstellar medium .
they therefore favour the viewing geometry determined at 8.4 ghz , as this higher frequency will be least affected .
they suggested the fiducial plane to coincide with the centre of the interpulse and hence also with the leading component of the main pulse .
this component has a high spectral index , suggesting an origin close to the magnetic axis . from this
they obtain @xmath61 km and @xmath62 .
we used a range of @xmath15 values to lie between the two components for both the main pulse and the interpulse ( which is too wide to constrain the viewing geometry further ) .
our results agree with those of @xcite at the corresponding frequency .
we therefore adopt the viewing geometry determined at 8.4 ghz by @xcite as the preferred viewing geometry .
psr j09405428 at 20 cm . as fig .
[ figj0631 + 1036 ] . ,
title="fig : " ] psr j09405428 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] the profile for the pulsar is double peaked ( jw06 ) and likely to be a conal double .
the range of @xmath15 was chosen between the two components with a position approximately at the midpoint of this range being favoured .
jw06 found @xmath1 to be unconstrained and @xmath63 , in agreement with our results .
the profile is unusually wide , at some 60@xmath55 .
the relatively large proportion of the pulse period that the line of sight spends within the emission region implies a small @xmath1 value or a large emission height .
the combination of the a / r effect and the @xmath30 surface suggests that the emission height can not be very large , thereby excluding @xmath64 .
psr j10165857 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j10165857 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] this profile is a double as shown also by jw06 .
the pa curve shows a large difference in pa when the two components are compared .
this was found to be relatively well fitted by an opm jump between the two components , which resulted in a @xmath65 times lower reduced-@xmath30 .
the inflection point occurs _ earlier _ than the midpoint between the two components .
this leads us to believe that we are not seeing a conal double , but rather a core component and a trailing cone with the leading cone missing .
we have therefore judged @xmath15 to be located at the peak of the first component .
the small inferred emission height implies that the two axes are close to alignment .
this is reinforced by the large width of the pulse , which is @xmath66 .
psr j10195749 at 10 cm . as the upper plot in fig .
[ figj0631 + 1036 ] . ] this pulsar is highly scattered by the interstellar medium , even at 3100 mhz .
smearing of the pa curve towards later phase means the steepest part of the curve will appear to be earlier than the true position of the inflection point .
also , as the profile is distorted , it is difficult to obtain reliable estimates of the fiducial plane position and the pulse width .
we therefore have not included this pulsar in further analysis .
psr j10285819 at 20 cm . as fig .
[ figj0631 + 1036 ] .
the a / r constraint shown corresponds to the case in which the emission region fills the open - field - line region , in order to be consistent with the other plots presented in this paper .
however , as discussed in the text , it appears likely that the true viewing geometry is at a smaller @xmath3 than allowed by this constraint .
, title="fig : " ] psr j10285819 at 20 cm . as fig .
[ figj0631 + 1036 ] .
the a / r constraint shown corresponds to the case in which the emission region fills the open - field - line region , in order to be consistent with the other plots presented in this paper .
however , as discussed in the text , it appears likely that the true viewing geometry is at a smaller @xmath3 than allowed by this constraint .
, title="fig : " ] this pulse profile is by far the narrowest in the sample .
it was discovered and analysed by @xcite .
the polarisation position angle is constant across the pulse , as can be seen in the figure , meaning the rvm curve for virtually any viewing geometry can be made to fit the data .
there are two explanations for such a narrow profile a nearly diametrical cut across a small emission region , or a grazing cut at the edge of a larger emission region .
the latter is unlikely to be the case for this pulsar , as radius - to - frequency mapping @xcite would cause the pulse width to be very sensitive to frequency .
the pulse profiles in @xcite show no significant change in the width between 1.4 ghz and 3.1 ghz .
the profile is an apparent double .
the fiducial plane would then be at the midpoint between the two peaks .
alternatively , it is possible that either of these is a core component and the other a conal component , with the other conal component remaining undetected . for this reason
the range of possible fiducial plane positions chosen stretches between the two peaks .
the lack of a gradient of the pa curve shows that the a / r effect must be sufficient to move the inflection point outside the on - pulse region .
this then gives a lower limit on @xmath20 and hence @xmath22 .
we find that @xmath67 if the fiducial plane coincides with the trailing peak and @xmath68 if the fiducial plane coincides with the leading peak . under the assumption that the emission region fills the open - field - line region ,
such a limit on @xmath22 , coupled with the small value of @xmath26 , makes the contours virtually independent on @xmath1 such that @xmath69 .
however , to explain the lack of frequency evolution of the pulse width we must assume @xmath3 to be significantly less than @xmath22 . to allow such solutions , the illuminated part of the beam must be at least a factor of 2 smaller than the open - field - line region .
psr j10485832 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j10485832 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] this pulsar is peculiar in the sample in that it shows a component with a very low fractional linear polarization .
the profile is complex and shows a strong variation with frequency @xcite .
the rvm fit for this pulsar is good and the @xmath30 surface alone allows values of @xmath70 and values of @xmath71 and @xmath72 to be excluded .
also , @xmath3 is exceptionally well constrained for a given @xmath1 .
the position of the fiducial plane in the profile is not obvious as there is not a high degree of symmetry . at higher frequencies
the profile becomes more symmetric about @xmath73 , suggesting this is the position of the fiducial plane .
however , a conservative estimate @xmath74 was used , which includes all possibilities around the dominant component , and also allows for the possibility of an undetected leading component ( jw06 ) .
this range allows the constraint to be refined further to @xmath75 , @xmath76 .
psr j10575226 at 20 cm . as fig .
[ figj0631 + 1036 ] .
the profiles shown are the main pulse ( upper panel ) and the interpulse ( middle panel ) .
both profiles are normalised to the maximum intensity of the main pulse .
the edges of the interpulse were determined according to 10% of the interpulse maximum intensity .
the constraint from the a / r effect ( blue region ) was derived using the interpulse only .
however , @xmath1 and @xmath3 are expressed with respect to the main pulse .
, title="fig : " ] psr j10575226 at 20 cm . as fig .
[ figj0631 + 1036 ] .
the profiles shown are the main pulse ( upper panel ) and the interpulse ( middle panel ) .
both profiles are normalised to the maximum intensity of the main pulse .
the edges of the interpulse were determined according to 10% of the interpulse maximum intensity .
the constraint from the a / r effect ( blue region ) was derived using the interpulse only .
however , @xmath1 and @xmath3 are expressed with respect to the main pulse .
, title="fig : " ] psr j10575226 at 20 cm . as fig .
[ figj0631 + 1036 ] .
the profiles shown are the main pulse ( upper panel ) and the interpulse ( middle panel ) .
both profiles are normalised to the maximum intensity of the main pulse .
the edges of the interpulse were determined according to 10% of the interpulse maximum intensity .
the constraint from the a / r effect ( blue region ) was derived using the interpulse only .
however , @xmath1 and @xmath3 are expressed with respect to the main pulse .
, title="fig : " ] this pulsar exhibits an interpulse and it has been argued that the two pulses originate from opposite poles @xcite . an investigation into the viewing geometry by @xcite determined ( @xmath1 , @xmath3 ) @xmath77 ( @xmath78 , 36@xmath55 ) .
in that paper the authors argued that the fiducial plane was between the central spike and the trailing component of the interpulse , and hence was at the leading edge of the main pulse .
independently of this choice they concluded that the main pulse was generated outside what is conventionally thought to be the open - field - line region , with the trailing edge of the main pulse most likely originating a factor of two further from the magnetic axis than the last open field lines .
the rvm fit to the pa curve is good for the main pulse and the leading half of the interpulse .
however , there is an abrupt jump in @xmath6 at the approximate centre of the interpulse , which is accompanied by complete depolarisation .
this deviation is more pronounced than in the data of @xcite due to the increase in s / n .
it is not possible to find an rvm curve which will simultaneously fit the data on both sides of the jump .
despite this , the @xmath30 surface presented here is well constrained and consistent with the viewing geometry of the earlier paper . according to weltevrede & wright the emission region responsible for the main pulse
is not confined to the conventional open - field - line region .
for this reason the constraint from the emission height and pulse width presented here was calculated using the interpulse only .
possible positions of the fiducial plane between the leading and trailing components ( @xmath79 ) were considered , with the favoured value coinciding with the fiducial plane position used in the earlier paper . the resulting constraint ( shown in blue in the figure )
is consistent with the @xmath30 surface .
the favoured contour is consistent with the viewing geometry reported by weltevrede & wright .
psr j11056107 at 20 cm . as fig . [ figj0631 + 1036 ] .
, title="fig : " ] psr j11056107 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] jw06 found @xmath80 and @xmath1 to be unconstrained for this pulsar .
our @xmath30 surface is less well constrained in @xmath3 , although we can exclude @xmath81 at the 3@xmath29 level .
we can also exclude @xmath82 based on the rvm fit alone .
the profile shows a high degree of symmetry at this frequency .
this suggests that the fiducial plane is at the centre of the profile .
however , at 3100 mhz the trailing component is significantly more intense than the leading component ( jw06 ) .
this difference in spectral index might suggest that one of the components is a core component , and for this reason the fiducial plane range was taken as @xmath83 .
combination of this constraint with the @xmath30 surface indicates that @xmath84 .
psr j11126103 at 10 cm . as fig .
[ figj0631 + 1036 ] . , title="fig : " ] psr j11126103 at 10 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] the reduced-@xmath30 of the rvm fit is good for this pulsar .
the @xmath30 surface is correspondingly well constrained in @xmath3 , giving @xmath85 .
the positions of the two peaks were used as the limits on the fiducial plane position .
this only marginally improves the constraint in ( @xmath1 , @xmath3 ) space .
psr j11196127 at 20 cm . as fig .
[ figj0631 + 1036 ] .
in addition , the dashed line describes the total intensity of the double - peaked profile as reported by @xcite .
the pa values at @xmath86 ( marked x ) are from the double - peaked profile and fitting was performed on the concatenated pa curve .
the constraints are shown for the cases with the rrat - like components ( not visible in the profile ) outside ( middle panel ) and inside ( bottom panel ) the open - field - line region .
, title="fig : " ] psr j11196127 at 20 cm . as fig .
[ figj0631 + 1036 ] .
in addition , the dashed line describes the total intensity of the double - peaked profile as reported by @xcite .
the pa values at @xmath86 ( marked x ) are from the double - peaked profile and fitting was performed on the concatenated pa curve .
the constraints are shown for the cases with the rrat - like components ( not visible in the profile ) outside ( middle panel ) and inside ( bottom panel ) the open - field - line region .
, title="fig : " ] psr j11196127 at 20 cm . as fig .
[ figj0631 + 1036 ] .
in addition , the dashed line describes the total intensity of the double - peaked profile as reported by @xcite .
the pa values at @xmath86 ( marked x ) are from the double - peaked profile and fitting was performed on the concatenated pa curve .
the constraints are shown for the cases with the rrat - like components ( not visible in the profile ) outside ( middle panel ) and inside ( bottom panel ) the open - field - line region .
, title="fig : " ] this pulsar usually exhibits a single component as shown by the solid lines in the figure and by jw06 .
however , on one occasion @xcite a second , comparable component was observed , the peak of which was @xmath87 later in phase than the commonly observed peak ( shown as the dashed line in the figure ) .
in the same paper , two rrat - like components were reported , flanking the two main components .
although the pa curve obtained from our data contains points over a smaller range of phase ( @xmath88 ) than that in @xcite , our improved s / n results in a more constrained @xmath30 surface from rvm fitting .
we also included the values from @xcite at later phase , which allowed the constraint to be improved further .
@xcite discussed possible alignments of the double - peaked profile relative to the single - peaked profile and argued that only a coincidence between the peaks of the single profile and the leading component of the double profile is plausible .
if the double - peaked profile was presumed to occur earlier relative to the single - peaked profile the values derived here for the emission height , beam half - opening angle and @xmath1 would all increase .
we consider two situations , one in which the rrat - like components are outside the open - field - line region and the other in which they are inside the open - field - line region . in both cases , the component configuration would have a high degree of mirror symmetry about a pulse phase @xmath89 .
this therefore seems the likely position of the fiducial plane , although the range of allowed values was taken to be between the two peaks of the double profile .
the range of @xmath22 values was the same for the two situations .
the choice of location of the pulse edges depends on whether or not the rrat - like emission was generated on open field lines . in the first situation ( outer components outside the open - field - line region ) the region
@xmath90 was excluded ( middle panel in the figure ) . in the other situation this exclusion extended to @xmath91 ( bottom panel in the figure ) .
it is unclear which situation is correct .
both scenarios are quoted in table [ tabderivedparameters ] .
the favoured contour used in subsequent analysis was that corresponding to the situation in which the rrat - like components are outside the open - field - line region , as this scenario provides the more conservative constraint on the viewing geometry .
however , the favoured geometry is similar in the two situations , so this choice is not critical to the conclusions of this paper .
psr j13576429 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j13576429 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] the profile is symmetric , single and broad with a shallow pa curve resulting in a relatively poor constraint on @xmath3 from rvm fitting alone and we find that @xmath92 .
this is at odds with @xcite who appear however to have used a value of the slope of the pa curve which is much too high .
the symmetry in the profile suggests that the fiducial plane corresponds to the peak ( @xmath73 ) , close to the inflection point of the pa curve .
this means that @xmath22 is likely to be small ( although the large error on @xmath10 allows significantly larger @xmath22 values ) . also , the profile is wide .
these two effects suggest that the axes are close to being aligned . to allow a scenario in which the leading peak of a double - peaked profile is unobserved ( jw06 ) , a conservative fiducial plane range was chosen .
the combination of the two constraints shows that @xmath93 or @xmath94 and that @xmath95 , while the favoured contour suggests that both @xmath1 and @xmath3 are significantly smaller than this .
psr j14106132 at 5 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j14106132 at 5 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] the profile is highly scattered at 20 cm and at 10 cm . for this reason we used archival data taken at 5 cm using the parkes telescope in 2007 november .
the profile exhibits a single component .
a conservative estimate of the fiducial plane position was adopted , which includes virtually the entire pulse .
the pa curve is well fitted by the rvm with only a small error on @xmath10 .
the resulting @xmath30 surface shows that @xmath96 , with an excellent constraint on @xmath3 for a given @xmath1 .
the conservative @xmath15 estimate means that the a / r effect can not constrain the geometry further .
psr j14206048 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j14206048 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] the profile of this pulsar is wide and double ( see also @xcite ) .
the higher s / n of our data allows an improved rvm fit .
the @xmath30 surface in the figure shows that @xmath97 and @xmath98 .
although more constrained , the surface is consistent with that of @xcite . assuming the fiducial plane is between the two peaks allows the viewing geometry to be constrained to @xmath99 , @xmath100 when aberration is taken into account .
psr j15095850 at 20 cm . as fig .
[ figj0631 + 1036 ] .
no constraint from the emission height and pulse width is possible as the nature of the best fit , and therefore @xmath10 , is highly uncertain.,title="fig : " ] psr j15095850 at 20 cm . as fig .
[ figj0631 + 1036 ] .
no constraint from the emission height and pulse width is possible as the nature of the best fit , and therefore @xmath10 , is highly uncertain.,title="fig : " ] the s / n of the profile is low but shows a double component similar to that shown in @xcite .
the percentage linear polarisation was found in that paper to be relatively low for this pulsar and the data prevented the authors constructing a pa curve or an rvm fit .
we find a similarly low percentage linear polarisation , but the increased s / n of our data allows a pa curve to be measured , a first for this pulsar . similarly to psr j10285819 ( @xmath14 [ j1028 ] ) , the limited range over which the pa can be determined , coupled with its shallow gradient , means that the rvm curve for virtually any viewing geometry can be made to fit the data .
we therefore exclude this pulsar from further analysis .
psr j15135908 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j15135908 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] the profile appears as a single component with a very low intensity leading shoulder at pulse phase @xmath101 .
this leading shoulder appears more significant at higher frequencies ( jw06 ) .
the larger degree of circular polarisation in the trailing component is a feature typical of double profiles in young pulsars ( e.g. , jw06 ) .
hence , the fiducial plane position range was chosen as @xmath102 .
however , to allow for the possibility that the leading shoulder does not originate within the open - field - line region , the allowed range for the location of the leading edge was extended to the leading edge of the trailing component .
the measured pa curve is very well fitted by the rvm . however , as the curve is relatively shallow , neither @xmath1 nor @xmath3 can be constrained from the @xmath30 surface alone . the 3@xmath29 error on @xmath10 is larger than the range of phase shown in the figure , indicating that the emission height , and hence @xmath22 , are highly uncertain . as a result ,
the combined constraint from the @xmath30 surface and the a / r effect is unable to constrain @xmath1 and provides a relatively poor @xmath3 constraint , @xmath103 .
psr j15315610 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j15315610 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] the pa curve is shallow , meaning that @xmath3 is relatively poorly constrained from the @xmath30 surface .
the profile exhibits two components , so the limits on the position of the fiducial plane were chosen to coincide with the two peaks and the midpoint between the two peaks was chosen as the favoured value of @xmath15 .
the resulting constraint suggests that @xmath104 but leaves @xmath1 unconstrained .
psr j16484611 at 10 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j16484611 at 10 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] the profile is scattered at 1.4 ghz , hence we show the 3.1 ghz profile here .
the profile is symmetric giving a strong indication that the fiducial plane is at the midpoint between the two peaks .
however , to account for the possibility of a missing component either before or after the observed profile , all phases between the two peaks were included in the range of possible positions of the fiducial plane .
the @xmath30 surface is relatively well constrained , to @xmath105 .
the relatively large error on @xmath10 means that neither @xmath1 nor @xmath3 can be further constrained by the a / r effect .
psr j17024128 at 10 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j17024128 at 10 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] the profile contains a dominant component with a shoulder at its leading edge , @xmath106 before the peak .
this suggests that the profile is an overlapping double .
judging by the profile the fiducial plane is between the two components .
however , as for other pulsars in this sample we allow for a missing leading component ( jw06 ) .
the combined constraint shows that solutions in the region @xmath107 can be excluded .
psr j17094429 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j17094429 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] rvm fitting alone imposes the limit @xmath108 , @xmath109 on the viewing geometry for this pulsar .
the profile is a single component with a high degree of symmetry over a wide range of radio frequencies @xcite .
however , to account for the possibility of the observed component being the trailing side of a double , the chosen range of fiducial plane position was extended considerably towards earlier pulse phase . when combined with the @xmath30 surface
this predicts @xmath110 and @xmath111 .
psr j17183825 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j17183825 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] the profile exhibits three components . the central component could be a core component , with a conal component on either side .
alternatively , the profile could show one side of a core - double - cone configuration , in which case either of the outer components could be the core . to allow for these possibilities , positions of the fiducial plane between the leading and trailing components
were considered .
the midpoint between the peaks of the two outer components , which roughly coincides with the centre of the profile , was taken as the favoured position of the fiducial plane .
the large width of the profile suggests that the axes are relatively aligned and that @xmath112 or @xmath113 .
this pulsar was included in the sample of @xcite and the results obtained here are largely consistent with theirs .
psr j17303350 at 10 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j17303350 at 10 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] the data used here were taken at 10 cm as the profile is scattered at longer wavelengths .
the figure shows that the rvm fit for this pulsar can constrain @xmath114 , but leaves @xmath1 unconstrained .
if the fiducial plane coincided with the single peak in the profile it would clearly be later than the inflection point , which would result in an unphysical negative emission height . furthermore , the observed component is steeper at the trailing edge , which is characteristic of a leading conal component .
this could suggest that any missing component would be later , not earlier , than the observed component .
however , to prevent a negative emission height it seems likely that there is an absent leading component ( similar to the transient trailing component of psr j11196127 , see @xmath14 [ j1119 ] ) . this would allow the fiducial plane to be at a sufficiently early phase to make the offset @xmath21 positive and is therefore included in the allowed range for @xmath15 .
the inferred values of @xmath1 and @xmath3 are highly sensitive to the fiducial plane position and the conservative range chosen here does not allow the viewing geometry to be constrained further than the @xmath30 surface .
@xcite reported a constraint of @xmath115 for this pulsar from rvm fitting .
however , the data they used were taken at 1351 mhz .
hence , scattering will have affected their rvm fit and is likely to be the source of the discrepancy between that result and the constraint presented here .
psr j18012451 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j18012451 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] the profile shows a single component , suggesting the fiducial plane could be close to the peak . alternatively , this may be a trailing conal component ; the profile is slightly steeper at the leading edge , which is characteristic of such components . in this case
the fiducial plane would be close to the leading edge of the pulse .
the inflection point of the best rvm fit is slightly earlier in phase than the peak , although the relatively large 3@xmath29 error on this value allows small positive offsets even if @xmath15 is close to the peak .
the combined constraint from rvm fitting and the emission height limits @xmath116 or @xmath117 and @xmath118 .
psr j18351106 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] psr j18351106 at 20 cm . as fig .
[ figj0631 + 1036 ] .
, title="fig : " ] the @xmath30 surface indicates that @xmath119 .
the profile exhibits a single peak , which could indicate that the fiducial plane is at the observed peak .
however , it was noted by jw06 that the trailing component of a double profile in young pulsars is often the more intense . it is therefore possible that there is a leading component which has not been observed .
the range of fiducial plane positions was chosen to reflect these two possibilities .
this large @xmath15 range , along with the comparatively large error on @xmath10 , leads to a large range of possible @xmath22 values . as a result
the a / r effect does not help constrain @xmath1 .
[ cols="<,^,>,^ , > , < , > , < , > , < , > , < , > , < " , ] lr@ @xmath120 l > r@ - lr@ - lrr psr & & & + & & & & & @xmath3 / @xmath55 + j0631 + 1036 & 16.61 & @xmath121 & 33 & 152 & 10.5 & 2.5 & 97.2 & 10.4 + j0659 + 1414 & 22.89 & @xmath122 & 32 & 160 & 22 & 5 & 120.7 & 16.7 + j07291448 & 19.41 & @xmath123 & 32 & 148 & 2 & 7 & 90 * & 6 * + j07422822 & 13.52 & @xmath124 & 55 & 180 & 7 & 0 & 90 * & 6.5 * + j08354510 & 11.97 & @xmath125 & 40 & 98 & 7.5 & 5 & 74.2 & 7.3 + j09084913 ( mp ) & 21.01 & @xmath126 & 96 & 96.8 & 8.5 & 6.3 & 96.1 @xmath127 & 5.9 @xmath127 + j09084913 ( ip ) & 19.33 & @xmath128 & & & 83.9 @xmath127 & 6.3 @xmath127 + j09405428 & 16.91 & @xmath129 & & 0 & 18 & 143.5 & 12.0 + j10165857 & 10.03 & @xmath130 & & 8.5 & 0 & 163.0 & 3.1 + j10195749 @xmath131 & & & & & + j10285820 @xmath131 & & & & & + j10485832 & 10.19 & @xmath132 & 0 & 50 & 0 & 7.5 & 29.6 & 4.8 + j10575226 ( mp ) & 4.4 & @xmath133 & 68 & 98 & 8 & 48 & 75.6 @xmath134 & 34.7 @xmath134 + j10575226 ( ip ) & 22.36 & @xmath135 & & & 104.4
@xmath134 & 5.9 @xmath134 + j11056107 & 17.49 & @xmath136 & 28 & 140 & 2 & 5 & 90 * & 4 * + j11126103 & 12.5 & @xmath137 & 0 & 180 & 5.5 & 0 & 126.6 & 3.8 + j11196127 ( a ) & 7.63 & @xmath138 & & 22 & 0 & 170.9 & 3.8 + j11196127 ( b ) & 7.63 & @xmath138 & & 19 & 0 & 173.1 & 2.9 + j13576429 & 7.42 & @xmath139 & & 0 & 50 & 6.9 & 4.9 + j14106132 & 5.96 & @xmath140 & 0 & 180 & 0 & 5.5 & 147.0 & 2.9 + j14206048 & 12.37 & @xmath141 & 0 & 33 & 0 & 8.5 & 17.1 & 3.8 + j15095850 @xmath131 & & & & & + j15135908 & 24.15 & @xmath142 & 0 & 180 & 0 & 70 & 13.7 & 15.3 + j15315610 & 16.91 & @xmath143 & 0 & 180 & 43 & 0 & 158.4 & 14.4 + j16484611 & 5.96 & @xmath144 & 0 & 180 & 14 & 0 & 158.9 & 4.1 + j17024128 & 7.84 & @xmath145 & & 13.5 & 0 & 22.8 & 5.3 + j17094429 & 18.97 & @xmath146 & 12 & 50 & 5.5 & 19 & 31.6 & 12.9 + j17183825 & 13.64 & @xmath147 & & 0 & 16 & 23.1 & 5.5 + j17303350 & 8.05 & @xmath148 & 0 & 180 & 10 & 0 & 148.4 & 3.7 + j18012451 & 18.7 & @xmath149 & & 12 & 0 & 146.8 & 7.3 + j18351106 & 23.86 & @xmath150 & 0 & 180 & 0 & 13 & 85.5 & 11.4 +
the @xmath1 distribution of the sample was investigated by considering the favoured @xmath1 values of 25 of the pulsars . these and the associated @xmath3 values are given in table [ tabuncorrectedviewinggeometries ] .
each favoured value was determined by finding the `` crossing point '' , the point along the favoured contour ( @xmath14 [ sectemissionheightconstraint ] ) for which @xmath30 was lowest and within the 3@xmath29 limit . in three out of 28 cases
this was not possible . these were psr j10195749 , for which @xmath15 could not be determined due to scattering - induced distortion of the profile , and psrs j10285819 and j15095850 , for which @xmath10 could not be determined due to the shallow gradient of the respective pa curves . in four cases the favoured contour was inconsistent with the @xmath30 surface .
possible reasons for this are a small error in the choice of the preferred position of the fiducial plane or measurement uncertainties in the position of the inflection point or pulse edges .
alternatively it is possible that both intersections of the last - open - field - lines by the line of sight occur outside the observed pulse , in which case @xmath26 will have been underestimated .
it follows from eq .
[ eqrhocontours ] that a given @xmath22 contour will deform and shift towards larger @xmath151 values . for these pulsars
the favoured contour was closest to the 3@xmath29 limit when @xmath152 , so these pulsars were assumed to be orthogonal rotators . in the table we use the result from @xcite for psr j09084913 and set @xmath153 for the other three pulsars .
the resulting @xmath1 distribution ( fig . [ figoriginaldistribution ] ) shows a pronounced skew towards low @xmath1 values , with an unexpectedly low number of sources with @xmath154 . in a typical sample of young pulsars ,
for which alignment of the axes via magnetic torques has yet to take effect @xcite , the axes would be expected to be randomly orientated , leading to a sinusoidal @xmath1 distribution ( e.g. , @xcite ) . furthermore , our sample of young pulsars are all @xmath0-ray - detected .
@xcite predicted that pulsars are more easily detectable in @xmath0-rays when @xmath1 is large , and therefore we would expect our sample to have a bias towards high @xmath1 values relative to the sinusoidal distribution .
the figure shows that this clearly is not the case .
the kolmogorov - smirnov ( ks ) test allows us to quantify how dissimilar two distributions are @xcite .
a low enough probability indicates that the two distributions are statistically different .
the result of such a test in which the observed @xmath1 distribution was compared to a sinusoidal distribution was 0.037% , indicating to a confidence greater than 3 @xmath29 that the observed values are not drawn from a sinusoidal distribution .
the distribution would be affected by , for example , a systematic bias in the favoured positions of the fiducial planes .
these were judged using the profile shape assuming the core - cone model @xcite .
however , the alternative to this assumption , the ` patchy ' beam model suggested by @xcite , would also not explain the observed bias in @xmath1 . in this model the emission is generated randomly across the polar cap , meaning that there should be no systematic bias in the illumination of the beam with respect to the fiducial plane and hence no systematic bias in @xmath15 . for the possible interpretations of the observed @xmath1 distribution ,
see @xcite .
distribution of most likely @xmath1 values ( mapped into the interval @xmath155 ) .
the sinusoidal distribution which might be expected for a sample of young pulsars is shown as the dashed curve .
a preference for low @xmath1 values relative to the sinusoidal distribution can be clearly seen . ]
@xcite ( jw06 ) noted that in their sample of young pulsars with high rates of rotational energy loss ( comparable to our sample in terms of spin parameters , but comprising only 14 objects ) , double - peaked profiles were often brighter in the trailing components .
here we take this argument a step further , considering the case in which the asymmetry is sufficient that the leading component is unobserved . in @xmath14 [ sectresults ]
we argued that this could apply to psrs j17303350 , j18012451 and j18351106 .
this raises the question of whether all apparent single - component profiles in our sample are in fact such highly asymmetric doubles .
this pronounced version of the jw06 effect would strongly affect the estimated position of the fiducial plane .
the profile appears as a single component and so the fiducial plane would in most cases be wrongly assumed to be at the peak .
however , if in reality there is a missing leading component the true location of the fiducial plane will be close to the leading edge of the observed component . by not accounting for this effect where it is present , we will have overestimated @xmath15 and hence underestimated the offset in rotational phase between the fiducial plane and inflection point , affecting the determination of @xmath1 .
multiple - component profiles will not have been affected in this way , as even if an asymmetry is present both components are still discernible .
the fiducial plane position affects the measured @xmath1 value in two ways .
firstly , the phase difference between the fiducial plane and inflection point is related to the derived emission height and as a consequence the beamwidth via eqs .
[ eqbeamhalfopeningangle ] and [ eqthetapc ] .
a later fiducial plane implies a smaller beamwidth and it can be seen from eq .
[ eqrhocontours ] that if the beamwidth is underestimated @xmath1 will also be underestimated for a given @xmath26 , the pulse longitude range covered by the open field lines .
however , the fiducial plane position also affects the estimated @xmath26 , which is determined by the offset of the pulse edge furthest away from the fiducial plane .
if @xmath26 is underestimated the measured @xmath1 value corresponding to a particular beamwidth will be overestimated .
these two competing effects mean that adjusting the position of the fiducial plane has a complicated effect on the measured @xmath1 value for a given pulsar .
if all apparently single - component pulsars in our sample are subject to the jw06 effect , the systematic misplacement of the fiducial plane could potentially lead to a further skew towards low @xmath1 values in the distribution of the single - component pulsars relative to that of the multiple - component pulsars .
if we assume that the profile morphology is not @xmath1-dependent the two distributions should be intrinsically similar , and hence the additional skew applicable to single - component pulsars would lead to a difference between our derived distributions .
it is therefore desirable to investigate whether such an effect is indeed present in our sample , so that any affected @xmath1 values could then be corrected . to test this hypothesis
we first of all calculated the @xmath1 distribution of the single - component pulsars1414 , j13576429 , j14106132 , j17094429 , j17303350 , j18012451 and j18351106 .
] when the fiducial plane of each was positioned at the profile peak .
this distribution is shown as the solid line in the lower panel of fig .
[ figsinglesversusmultiples ] and can be compared with that derived for the multiple - component pulsars ( upper panel of the same figure ) .
the peak is later than the inflection point for psrs j17303350 and j18012451 , resulting in a negative derived emission height and meaning that the favoured contour can not be determined .
however , the overall constraint on @xmath1 ( see figs . [ figj1730 - 3350 ] and [ figj1801 - 2451 ] ) indicates that both these pulsars have small magnetic inclinations .
for this reason both were assigned a value @xmath156 .
psr j15135908 also yields a negative derived emission height in this situation .
however , @xmath1 is unconstrained for this pulsar ( see fig . [ figj1513 - 5908 ] ) and so no reliable estimate can be made in the absence of the favoured contour .
for this reason , the pulsar was excluded from this part of the analysis . applying the ks test to the alpha distributions derived for the multiple - component pulsars and the single - peaked profiles under the assumption that the jw06 effect is absent for all pulsars results in a probability of 24% .
hence , there is no evidence for the distributions to be different .
the test was repeated using the @xmath1 distribution ( dashed line in the lower panel of fig .
[ figsinglesversusmultiples ] ) resulting when the fiducial plane was placed at the leading edge of the pulse for the pulsars for which there was a suggestion of a missing leading component ( psrs j17303350 , j18012451 and j18351106 ) .
this therefore corresponds to the preferred values as they appear in table [ tabuncorrectedviewinggeometries ] .
the result of the ks test in this case was 57% .
this shows that applying the jw06 effect to these pulsars has a non - significant effect on the distribution .
the test was also repeated with the fiducial plane positioned at the leading edge for all the single - component pulsars ( dotted line in the lower panel of fig .
[ figsinglesversusmultiples ] ) , corresponding to the scenario that all single - peaked profiles have missing leading components .
the ks test result was again 57% .
applying the jw06 effect to some pulsars might increase the similarity between the distributions for single- and multiple - component pulsars , indicating that the effect may be present in some particular cases ( and indeed for psrs j17303350 and j18012451 it appears necessary in order to avoid negative derived emission heights ) .
however , the sample size is too small to draw any significant conclusions . applying the effect to all single pulsars has little effect on the overall @xmath1 distribution .
we therefore take a conservative option , whereby we apply the effect in those cases for which there is evidence , but do not apply the effect to other single - component pulsars .
the values given in table [ tabuncorrectedviewinggeometries ] correspond to this situation .
there are reasons to expect that the jw06 effect is not applicable to all pulsars exhibiting a single - component profile .
firstly , psr j11196127 has a single - peaked profile in almost every observation .
however , @xcite showed a transient component trails the often - seen component .
the two rrat - like components detailed in that paper are believed to frame the two main components , strongly suggesting the fiducial plane to be located after , not before , the ` single ' component .
this then appears to be a case where the jw06 effect is not applicable .
furthermore , the intensity asymmetry discussed in jw06 does not appear for all @xmath0-ray - loud pulsars exhibiting double - component radio profiles .
for example , the profile of psr j11056107 ( @xmath14 [ j1105 ] ) shows two components with approximately equal peak intensities . given that some double - component profiles do not show this asymmetry , it is plausible that there also exist some single profiles which are not subject to the effect and hence are intrinsically single , rather than extremely asymmetric double - component profiles .
( upper panel ) distribution of most likely @xmath1 values of pulsars with profiles exhibiting multiple components .
( lower panel ) distribution of most likely @xmath1 values of the single component pulsars when the fiducial plane is taken to be at the peak of the observed component ( solid line ) , after applying the johnston & weisberg effect to those pulsars ( psrs j17303350 , j18012451 and j18351106 ) for which there is a suggestion of a missing leading component ( dashed line ) and after making the assumption that each pulsar has an unobserved leading component ( dotted line ) . for clarity
the lines have been displayed with slight offsets in the bins where they would otherwise coincide .
, title="fig : " ] ( upper panel ) distribution of most likely @xmath1 values of pulsars with profiles exhibiting multiple components .
( lower panel ) distribution of most likely @xmath1 values of the single component pulsars when the fiducial plane is taken to be at the peak of the observed component ( solid line ) , after applying the johnston & weisberg effect to those pulsars ( psrs j17303350 , j18012451 and j18351106 ) for which there is a suggestion of a missing leading component ( dashed line ) and after making the assumption that each pulsar has an unobserved leading component ( dotted line ) . for clarity
the lines have been displayed with slight offsets in the bins where they would otherwise coincide .
, title="fig : " ] recently , @xcite ( henceforth phg14 ) have derived maps of the ` likelihood ' of a given viewing geometry from @xmath0-ray and radio profiles and hence determined the most likely emission geometry for a large number of @xmath0-ray pulsars based on four @xmath0-ray models : the polar cap @xcite , slot gap @xcite , outer gap @xcite and one - pole caustic model @xcite .
since we also studied @xmath0-ray detected pulsars , phg14 includes most of the pulsars examined in this paper with the exception of psr j15315610 .
the emission geometries derived by phg14 are based on very different information compared to our study ( e.g. radio polarization was not considered ) .
it is therefore interesting to investigate whether our measurements are consistent with the geometries needed to explain @xmath0-ray light curves .
since @xmath3 is confined to a small range for radio - detected pulsars and its precise value can be expected to be highly dependent on details in the way the radio profiles are interpreted , we choose to investigate if there is a correlation between the @xmath1 values presented in this paper and those by phg14 .
there is inherent disagreement between the viewing geometries implied by the various @xmath0-ray models .
for instance , the relatively small emission region of the polar cap model , coupled with the typically wide @xmath0-ray intensity profiles , implies low @xmath1 values .
in contrast to this , the other models propose the emission to be generated in the outer magnetosphere , resulting in larger derived magnetic inclination angles .
this can be seen in fig .
[ figusvspierbattista ] , which shows the @xmath1 values derived by phg14 against the values derived in @xmath14 [ sectresults ] .
none of the four models can explain the large range of @xmath1 values we have found in this paper .
the low @xmath1 values in particular are incompatible with the slot gap , outer gap and one - pole caustic models .
it can also be seen from fig .
[ figusvspierbattista ] that the two sets of values are uncorrelated for all four @xmath0-ray models considered by phg14 , in the sense that a large @xmath1 value derived in this paper does not necessarily correspond to a large @xmath1 value according to phg14 , although the values are subject to large errors . to investigate whether the two sets of values could be considered correlated within these errors a series of trials was performed .
for each trial , two @xmath1 values ( corresponding to the constraints presented in this paper and to those presented by phg14 ) were drawn randomly for each pulsar , presuming a gaussian distribution function with widths determined by the quoted errors .
the spearmann rank - order correlation coefficient was then found between the two resulting sets of values .
none of the four cases showed a significant tendency towards correlation .
there are several reasons why the two sets of values might be uncorrelated .
firstly , it is possible that none of the @xmath0-ray models are a good description of the physics .
if true , this might be somewhat surprising given that the shape of the @xmath0-ray light - curves , at least for the models which place the extended emission region in the outer magnetosphere , is in first order determined by the pulse phase ranges where caustic emission is expected to happen .
one could expect this geometric effect to be at some level independent of the detailed physics .
another possibility is that the absence of a correlation between the @xmath1 values derived in this paper and by phg14 is caused by the fact that the interpretation of the radio data is very different .
for example , phg14 only considered the shape of the radio profile and assumed that the radio emission height follows an empirical relation described by @xcite , whereas in this paper we considered both the radio profile morphology and polarisation allowing us to determine the emission height .
although these methods will be inconsistent up to some level , it remains to be seen whether this difference would have a significant effect on the derived @xmath1 values from joint radio and @xmath0-ray fitting . as shown by phg14 ,
the effect of considering radio data is effectively to down - weight solutions with large @xmath3 values .
since the different methods to interpret the radio data will have a similar effect , it could be expected that the fine details of the radio model only have a minor effect on the resulting @xmath1 values .
how big this effect is can potentially be determined by considering the constraints derived in this paper together with the fits of @xmath0-ray data .
in this paper we have presented constraints to the viewing geometry ( characterised by the magnetic inclination angle , @xmath1 , and the impact parameter of the line of sight relative to the magnetic axis , @xmath3 ) for a sample of @xmath0-ray - loud pulsars .
this constraint has two components .
the first is associated with the goodness of fit between the rotating vector model and the observed pa curve .
the second is associated with the emission beam half - opening - angle derived from the relative offset in pulse phase between the inflection point of the pa curve and the location of the fiducial plane in the profile ( which represents the point of closest approach between the line of sight and the magnetic axis ) . in determining the latter constraint ,
we have taken into account uncertainties on the fiducial plane , inflection point and pulse edges in a systematic way .
this new approach is more conservative and therefore in many cases allows regions of the ( @xmath1 , @xmath3 ) parameter space to be objectively excluded with greater confidence .
in addition to this conservative constraint , we also determined our preferred viewing geometry for each pulsar , allowing a statistical analysis of the results .
we find that the @xmath1 distribution exhibits an unexpected skew towards low values .
these pulsars are all younger than the various observationally estimated timescales for alignment of the magnetic inclination angle due to magnetic torques , suggesting the observed distribution should be the birth distribution for this sample .
if the magnetic axis is randomly orientated with respect to the rotation axis at the birth of a neutron star , it follows that this birth distribution of @xmath1 should be sinusoidal .
further to this , the pulsars investigated in this paper have all been detected in @xmath0-rays .
high energy models predict that pulsars are more easily detectable in @xmath0-rays when the magnetic inclination is large , as the intensity modulation is more pronounced .
this would be expected to introduce a selection effect resulting in an abundance of highly inclined pulsars relative to the sinusoidal distribution .
these considerations make the observed tendency towards low @xmath1 values surprising .
this skew and its possible causes will be investigated in detail in @xcite .
we have also reported evidence for absent leading components in some of the pulse profiles which exhibit only a single component .
an argument can be made that these are extreme cases of the intensity asymmetry noted in double - component profiles of young pulsars by @xcite .
however , the sample size is too small to determine whether this effect applies to all apparently single - component young pulsars .
we have found a lack of correlation between the @xmath1 values derived here from radio intensity and polarisation data and those derived by @xcite from consideration of the radio and @xmath0-ray light curves .
this indicates a possible problem in the interpretation of the radio and/or @xmath0-ray data which requires further investigation .
however , despite this uncertainty in @xmath1 , the tight constraints on @xmath3 which we have presented should prove useful for future attempts to determine the @xmath0-ray emission mechanism .
we wish to thank the referee aris karastergiou for his constructive input .
pulsar research at jbca is supported by a consolidated grant from the uk science and technology facilities council ( stfc ) .
the parkes radio telescope is part of the australia telescope which is funded by the commonwealth of australia for operation as a national facility managed by csiro .
m. , zavlin v. e. , grondin m .- h . , shannon r. , smith d. a. , burgay m. , camilo f. , cohen - tanugi j. , freire p. c. c. , grove j. e. , guillemot l. , johnston s. , keith m. , kramer m. , et al .
2011 , a&a , 533 , a102 | we present radio intensity and polarisation profiles of 28 @xmath0-ray - detected pulsars with the aim of putting constraints on their viewing geometries using data from the parkes telescope .
constraints are formed both from the goodness - of - fit of the position angles to the rotating vector model and from the beam opening angle considering aberration and retardation effects .
uncertainties on the relevant parameters are systematically taken into account in order to produce a more robust constraint , using a new approach .
surprisingly , we find that the distribution of the magnetic inclination angle ( @xmath1 ) in this subset of pulsars peaks at low values , contrary to the predictions of @xmath0-ray models .
we find a lack of correlation between these and a set of @xmath1 values which were derived using @xmath0-ray light curves , suggesting a problem in the interpretation of the data in one or both of these domains .
finally , we also show that the @xmath1 distribution of pulsars with multiple radio components is no different to that of single - component pulsars .
[ firstpage ] pulsars : general polarisation . |
during the last few years it has been established that almost all massive galaxies are hosting a super - massive black hole ( smbh ) in their centers .
it has been found that the smhb mass is well correlated with the host galaxy s spheroidal component mass and with the bulge luminosity ( e.g. , kormendy & richstone 1995 ; magorrian et al.1998 ; mclure & dunlop 2001 ) .
furthermore , a remarkably close correlation of the smbh mass and the stellar velocity dispersion , @xmath15 , is detected for active , as well as for non - active galaxies ( gebhardt et al.2000 ; graham et al.2001 ; merritt & ferrarese 2001 ; mclure & dunlop 2002 ; tremaine et al.2002 ) .
these results , together with indications that luminous quasars reside in massive early type galaxies ( e.g. , mclure et al.1999 ; kukula et al.2001 ; nolan et al.2001 ; percival et al.2001 ; dunlop et al.2003 ) , point to a closely related formation of massive spheroidal galaxies and the formation and growth of central smbh . however , it is still under debate whether the black hole and the host galaxy formed simultaneously or whether one component formed before the other ( e.g. , silk & rees 1998 ; cattaneo et al.1999 ; salucci et al.1999 ; kauffmann & haehnelt 2000 ; haiman & loeb 2001 ; menou et al.2001 ; archibald et al.2002 ; dimatteo et al.2003 ) .
it is of great importance to estimate smbh masses at higher redshifts to investigate whether the relations of smbh mass with fundamental parameters of their harboring galaxies that are found for active and non - active galaxies in the local universe are already established at early cosmic epochs .
recent studies of quasars at moderate ( mclure & jarvis 2002 , @xmath16 ) and high - redshifts ( shields et al.2003 , @xmath17 ; vestergaard 2004 , @xmath18 ) assume that the locally determined radius luminosity relation can be applied up to redshifts of @xmath19 . particularly , black hole masses have been estimated for several high - redshift quasars , providing evidence that even at early cosmic times black holes with several times @xmath20 are already in place ( barth et al.2003 ; willott et al.2003 ; vestergaard 2004 ; warner et al.2004 ) .
the detection of quasars at redshifts @xmath21 ( fan et al.2001 ) and black hole mass estimates well beyond @xmath22 ( bechthold et al.2003 ; netzer 2003 ) challenges black hole growth models to buildup smbhs on time scales of several times @xmath9 myr , because at @xmath23 the age of the universe is less than 0.9 gyr . however , the high - redshift @xmath24 estimates are all based on the iv civ@xmath25 emission line and the assumption of virialized gas motion .
it is obviously important to use more than a single emission line to determine @xmath24 as a fundamental agn parameter .
much effort has been spent to derive methods to estimate the black hole mass , @xmath24 , based on reverberation experiments and assuming virial dominated motion of the broad line region ( blr ) gas ( e.g. , kaspi et al.2000 ; peterson & wandel 1999,2000 ; wandel , peterson , & malkan 1999 ) .
this results in an empirically established radius - luminosity relation , @xmath26 with e.g. @xmath27 ( kaspi et al.2000 ) .
since long - term variability studies are currently not feasible for quasars , alternative methods are suggested employing single epoch emission line profile width measurements to estimate @xmath24 ( vestergaard 2002 ) .
there is still some concern regarding this approach ( e.g. , krolik 2001 ) and that the r - l relation might be flatter than @xmath28 ( e.g. , peterson et al.2000 ; mclure & jarvis 2002 ; corbett et al.2003 ; netzer 2003 ; vestergaard 2002 , 2004 for a more detailed discussion ) .
however , using emission line profiles of h@xmath4 and iv civ@xmath25 to estimate the smbh mass of quasars yields consistent results that are on average within @xmath29%(e.g .
, vestergaard 2002 ; warner et al.2003 ) .
recently , mclure & jarvis ( 2002 ) suggested mgii@xmath30 to measure black hole masses . particularly for quasars at redshifts @xmath12 ,
mgii is of great interest because h@xmath4 is shifted out of the k - band . as a low ionization line mgii
can be used as a well suited replacement of h@xmath1 instead of using the high ionization line iv civ@xmath25 .
in addition , independent methods based on the well established @xmath31 and @xmath32 relations for active and non - active galaxies ( gebhardt et al.2000 ; ferrarese et al.2001 ; laor 2001 ; mclure & dunlop 2001,2002 ; tremaine et al.2002 ) result in consistent black hole mass estimates ( see barth 2003 ; vestergaard 2004 for a more detailed discussion ) . in the following
we present the results of smbh mass estimates for luminous quasars at redshifts @xmath0 that we observed in the near infrared to study the evolution of the feii / mgii emission line ratio ( dietrich et al.2002a,2003a ) .
we analyzed the broad emission line profiles of mgii@xmath30 and h@xmath1 .
for most of the quasars we also have observed or have access to the ultraviolet spectral range that contains the strong iv civ@xmath25 emission line .
hence , we could use almost always at least two emission line profiles to determine @xmath24 and for the first time to compare the obtained smbh masses . in the following we assume @xmath33 kms@xmath34mpc@xmath34 , @xmath35 , @xmath36 . introducing @xmath37 ( netterfield et al.2002 ) instead of @xmath38 ,
the luminosities at the highest redshifts would be @xmath39% larger .
hence , a slight change in luminosity will only have a minor impact on the estimated black hole mass .
we have observed a sample of 15 luminous high - redshift quasars with @xmath40 up to @xmath41 ( table 1 ) .
the spectra were obtained using infrared and optical spectrographs of the cerro tololo inter - american observatory at chile ( osiris , 4 m ) , the eso observatories at la silla ( sofi , 3.5 m ) and paranal ( isaac , 8.2 m ) at chile , the calar alto observatory ( twin , 3.5 m ) at spain , and the w.m.keck observatory ( nirspec , 10 m ) at hawaii ( for more details on the observations see dietrich et al.1999,2002a,2003a ; dietrich & wilhelm - erkens 2000 ) . in order to study the broad ultraviolet feii emission and
the mgii@xmath30 emission line ratio j , h , and k - band spectra were recorded .
for the quasars with @xmath11 we observed in the k - band the rest frame optical wavelength range which covers h@xmath1 and [ oiii]@xmath42 .
furthermore , for 10 of the studied quasars we have ultraviolet spectra of the iv civ@xmath25 emission line ( dietrich et al.2002b ) .
for each of the high - redshift quasars we measured the emission line profile width of h@xmath4 , mgii@xmath30 , and iv civ@xmath25 if covered . to determine a reliable profile width it is necessary to identify and to remove potential contamination by additional emission features .
although the mgii@xmath30 emission line doublet appears to be relatively isolated in quasar spectra , it is embedded in the superposition of many thousands of discrete ultraviolet feii emission lines that form broad emission blends ( e.g. , wills , netzer , & wills 1985 ) .
the careful correction for ultraviolet fe emission yields a more solid measurement of the mgii profile width than assuming a linear pseudo - continuum beneath the mgii emission line .
the contamination of h@xmath4 by optical feii emission is less severe ( e.g. , boroson & green 1992 ) .
iv civ@xmath25 suffers only minor contamination of weak emission lines , like niv]@xmath43 in the blue wing and from feii emission beneath the iv civ line profile ( e.g. , vestergaard & wilkes 2001 ) . to account for these emission line profile contaminations we analyzed the high redshift quasar spectra using a multi - component fit approach ( dietrich et al.2002a,2003a ) .
following wills , netzer , & wills ( 1985 ) we used the following four components to reconstruct each quasar spectrum , ( i ) a power - law continuum ( f@xmath44 ) , ( ii ) balmer continuum emission ( grandi 1982 ; storey & hummer 1995 ) , ( iii ) feii emission templates for the ultraviolet and optical wavelength range ( vestergaard & wilkes 2001 ; boroson & green 1992 ) , and ( iv ) and gaussian profiles to account for individual broad emission lines , such as h@xmath1 , mgii , and iv civ .
the power - law continuum fit to the individual quasar spectra was iteratively determined taking into account the measured strength of the feii emission and balmer continuum emission derived by the fitting process .
the emission line profile widths , expressed as full width at half maximum ( fwhm ) which we measured for h@xmath1 , mgii , and iv civ if are accessible , are given in table 2 , together with the spectral slope @xmath45 of the power - law continuum . since the spectral resolution of the infrared spectra is of the order of @xmath46 to 500 kms@xmath34 ( table 2 ) , a correction of the measured profile width with respect to the instrumental resolution is not necessary . the empirically justified radius - luminosity relation , @xmath26 ( e.g. , @xmath27 , kaspi et al.2000 ) and the assumption of virial dominated motion of the broad line region ( blr ) gas ( peterson & wandel 1999 ; onken & peterson 2002 ) has led to equations which connect the black hole mass @xmath24 with observationally accessible parameters like luminosity and emission line profile width .
we applied the equations given by kaspi et al.(2000 ) , mclure & jarvis ( 2002 ) , vestergaard ( 2002 ) , and warner et al.(2003 ) to estimate the black hole mass for the high - redshift quasars we observed .
the explicit form of the equations we employed , are presented in the appendix for h@xmath1 , mgii@xmath30 , and iv civ@xmath25 .
it should be kept in mind that although iv civ@xmath25 is a high - ionization emission line , in contrast to h@xmath1 and mgii@xmath30 as typical low - ionization lines , there are good reasons to assume that the broad line iv civ emitting gas motion also is gravitationally dominated ( peterson & wandel 2000 ) .
this assumption is based on variability response , the emission line profile shape , and the relation of the profile widths of h@xmath1 and iv civ@xmath47 combined with variability time lags ( onken & peterson 2002 ) .
furthermore , it has been shown that using emission line profiles of h@xmath1 and of iv civ@xmath25 to estimate the smbh mass of quasars yields consistent results on average ( e.g. , vestergaard 2002 ; warner et al.2003 ) .
we applied the equations ( a1 ) to ( a5 ) given in the appendix together with the emission line profile widths of h@xmath4 , mgii , and iv civ and the spectral index @xmath45 of the continuum slope ( table 2 ) , to calculate the mass of the smbh for each of the high - redshift quasars . to estimate the uncertainty of the determined smbh masses the errors of the emission line profile width measurements and of the continuum luminosity
have been propagated . in table 3
the resulting black hole masses for each emission line estimator are presented . in figure 1
the black hole mass is displayed for each emission line as a function of redshift .
the luminous quasars under study have black hole masses in the range of @xmath48 up to @xmath49 .
the smbh mass based on the h@xmath4 profile width are consistent within a factor of @xmath50 , with lower mass estimates provide by the relation suggested by mclure & jarvis ( 2002 ) compared to those of kaspi et al.(2000 ) .
this difference is mainly caused by the different slopes of the applied r
l relation . assuming
a slope of @xmath51 instead of @xmath27 will result in @xmath52 times lower @xmath24 estimates .
utilizing the iv civ emission line profile width , the resulting smbh mass estimates , using equations a3 and a4 as given in the appendix , are consistent within @xmath53% , as expected since the same slope is assumed for the r
l relation ( eqs . a3 and a4 ) .
the scatter of the two smbh estimates based on iv civ are due to the use of the individual spectral index @xmath45 for each quasar that we determined instead of using a uniform continuum slope of @xmath54 for example .
furthermore , to use the continuum luminosity l(1450 ) for each quasar an additional @xmath45-dependent factor is introduced . in spite of these modifications
the resulting black hole mass estimates , using the iv civ profile width , agree well with each other .
close inspection of figure 3 obviously shows that the @xmath24 estimates based on mgii are systematically lower than the ones based on h@xmath1 and iv civ@xmath25 . on average the mgii@xmath30 emission line profile indicates a @xmath5 lower smbh mass than the h@xmath4 and iv civ@xmath25 line profile .
for each of the high - redshift quasars we calculated a mean smbh mass based on h@xmath1 , mgii , and iv civ if available ( table 3 ) .
although the smbh mass estimates based on mgii are lower than those given by h@xmath4 and iv civ , we do not apply possible corrections ( see sect.4.2 ) . a basic assumption of the r
l relation involving the mgii@xmath30 emission line is that the emission regions of mgii and h@xmath1 are nearly at the same distance from the central continuum source . while the comparison of the profile width of both emission lines indicate that @xmath55 is nearly identical with @xmath56 ( mclure & jarvis 2002 ) , corbett et al.(2003 ) find that mgii has generally a broader profile than h@xmath1 . in figure 2
we compare the fwhm of the line profiles of mgii and h@xmath1 for the seven quasars at @xmath11 . with the exception of q2050@xmath57359 and q0105@xmath572634 ( similar width ) and q2348@xmath574025 ( larger mgii profile than h@xmath1 ) the h@xmath4 profile is generally slightly broader than the mgii profile .
the average ratio of the fwhm of the h@xmath1 to mgii@xmath30 line profile amounts to @xmath58 based on the seven @xmath11 quasars we studied .
the most direct measure of the blr radius for mgii is given by reverberation mapping experiments .
so far , almost all monitoring campaigns dedicated to agns have been focused in the optical on the h@xmath4 region and in the ultraviolet on the ly@xmath45
iii ciii]@xmath59 wavelength range .
however , for ngc5548 ( clavel et al.1991 ; peterson et al.1991 ; dietrich & kollatschny 1995 ) and ngc3783 ( reichert et al.1994 ; stirpe et al.1994 ) mgii@xmath30 and h@xmath4 were measured .
ngc3783 showed only minor mgii emission line flux variations during the seven month campaign in 1992 .
the derived emission line flux variation delays for mgii and h@xmath1 are of the same order of @xmath60 days ( reichert et al.1994 ; stirpe et al.1994 ) .
in contrast , ngc5548 underwent quite strong variations for h@xmath1 and mgii during the first year of its monitoring in 1989 .
dietrich & kollatschny ( 1995 ) used the ngc5548 observations taken with iue in the ultraviolet ( clavel et al.1991 ) and from the ground - based monitoring campaign ( peterson et al.1991 ) to estimate a typical radius of the mgii@xmath30 and h@xmath1 emission line regions .
they determined time delays based on the centroid ( @xmath61 ) , as well as on the peak ( @xmath62 ) , of the interpolated cross - correlations function ( iccf ) .
the time delays based on the iccf centroid amount to @xmath63d and @xmath64d , while the peak of the iccf yield delays of @xmath65d and @xmath66d ( dietrich & kollatschny 1995 ) . assuming that these delays represent typical distances of the line emitting gas from the central variable continuum source , in the case of ngc5548
the mgii gas is located @xmath52 times more distant than the h@xmath1 emitting region . in a very simplified model of gravitationally bound keplerian orbits it can be expected that the typical velocity of mgii is about @xmath67 times narrower than for h@xmath1 .
however , this estimate of @xmath68(mgii ) is based on only one seyfert galaxy .
although this result is quite consistent with the ratio of the profile widths of h@xmath4 to mgii that we find for the high - redshift quasars ( figure 2 and table 2 ) , it is necessary to measure @xmath68(mgii ) for more seyfert galaxies and quasars . since a larger distance of the mgii blr emitting region than for h@xmath1 will introduce a factor of @xmath69 , the estimated smbh mass will be slightly underestimated assuming @xmath70 . to compare the smbh mass with the bolometric luminosity , @xmath71 , and to determine the eddington luminosity ratio , @xmath72
, we calculated @xmath71 . to transform measured rest frame luminosities @xmath73 into @xmath71 scaling factors
are provided by e.g. elvis et al.(1994 ) and kaspi et al.(2000 ) .
most recently , vestergaard ( 2004 ) and warner et al.(2004 ) present revised scaling factors for optical and ultraviolet continuum luminosities employing results of multi - frequency studies on the agn spectral energy distributions from x - ray energies to the mid - infrared wavelength range .
they found consistent scaling factors which are slightly smaller ( @xmath29% ) than previously assumed .
we used the relation @xmath74 to calculated the bolometric luminosity ( warner et al.2004 ) for each of the high - redshift quasars under study ( table 1 ) . in figure 3 we plot the average smbh masses as a function of l@xmath75 . at a given l@xmath75
the determined m@xmath76 of the high - redshift quasars cover a range of up to one order of magnitude in mass , similar to comparable studies for quasars at lower redshift ( e.g. , woo & urry 2003 ; vestergaard 2004 ; warner et al.2004 ) . for a fully ionized pure hydrogen gas , i.e.
@xmath77 , the eddington luminosity is given as ( rees 1984 ; peterson 1997 ) @xmath78 assuming a more realistic gas chemical composition containing hydrogen , helium , and some heavier elements will result in an additional factor @xmath79 accounting for the atomic weight per electron ( haiman & loeb 2001 ) .
this will yield a 15% higher eddington luminosity . in the following
we will assume pure hydrogen gas because @xmath80 has only minor impact on @xmath81 .
the eddington ratio @xmath8 is determined for each quasar individually .
the mean black hole masses are used to calculate the eddington luminosity , @xmath81 , for spherical accretion .
together with the bolometric luminosity , the eddington ratio , @xmath72 , is calculated ( table 3 ) .
the determined @xmath72 covers a range of @xmath82 to @xmath83 which is in good agreement with results of studies of quasars sample that are dominated by low and intermediate redshift quasars ( e.g. , woo & urry 2003 ; vestergaard 2004 ; warner et al.2004 ) .
hence , there is no difference regarding the eddington ratio @xmath72 for quasars at high - redshifts compared to those at lower redshift .
in table 3 the smbh mass estimates based on the three emission lines h@xmath1 , mgii@xmath30 , and iv civ@xmath25 are listed for comparison .
similar to previous studies there appears to be an approximate upper limit for @xmath24 at @xmath84 ( e.g. , corbett et al.2003 ; netzer 2003 ; shields et al.2003 ; vestergaard 2004 ; woo & urry 2003 ; warner et al.2003 ) .
only q0105 - 2634 yields an estimate of @xmath85 ( eq.a1 ) and @xmath86 ( eq.a2 ) , respectively , based on the h@xmath1 emission line width . however , the h@xmath1 emission profile of this particular quasar has a rather low signal - to - noise ratio .
furthermore , the estimate @xmath87 based on the r l relation with a slope of @xmath27 . assuming a flatter slope @xmath1 of the r
l relation , as suggested by peterson et al.(2000 ) and mclure & jarvis ( 2002 ) the mass estimate is reduced by a factor of @xmath52 and in this case within the uncertainties in better agreement with an empirical upper limit of @xmath88 which also has been suggested by umemura ( 2003 ) and rocca - volmerange et al.(2004 ) who studied early phases of galaxy formation and related black hole growth . for two quasars at redshift @xmath11 we detected quite prominent [ oiii]@xmath42emission .
it has been suggested that the emission line profile width of [ oiii]@xmath42can be used as a surrogate of the stellar velocity dispersion @xmath15 ( nelson 2000 ) . applying the relation that connects black hole mass with stellar velocity dispersion ( tremaine et al.2002 ) the use of [ oiii]@xmath42 as a smbh mass estimator has been investigated by shields et al.(2003 ) . using the emission line profile width of [ oiii]@xmath89 the estimated the mass of the black holes amount to @xmath90 ( q0256 - 0000 ) and @xmath91 ( q0302 - 0019 ) which are consistent with the masses provided by iv civ , mgii , and h@xmath1 ( table 3 ) .
while the black hole mass estimates based on iv civ@xmath25 and h@xmath4 are quite consistent , the @xmath24 estimate provided by mgii@xmath30 indicates lower masses ( table 3 ) .
although r
l relations based on statistical results found for quasar samples introduce uncertainties of a factor of at least @xmath92 in the case of individual quasars ( e.g. , vestergaard 2002 ) there are some additional possibilities which may cause lower @xmath24 estimates based on mgii@xmath30 . first , the slope @xmath4 of the @xmath93 relationship is still not well defined , varying between @xmath94 ( e.g. , kaspi et al.2000 ; peterson et al.2000 ; mclure & jarvis 2002 ; vestergaard 2002 ; netzer 2003 ) .
a flatter slope in the @xmath95 relation will result in less massive black holes . while in the original analysis by kaspi et al.(2000 ) it is found that @xmath96 , there are indications that the slope is closer to @xmath97 ( peterson et al.2000 ) .
mclure & jarvis ( 2002 ) derived @xmath98 for the r
l relation using mgii .
they determined this flatter slope under the assumption that mgii and h@xmath1 originate at nearly the same blr radius .
hence , they used h@xmath1 reverberation results for seyfert galaxies and quasars to analyze the correlation of r@xmath99(mgii ) with the luminosity @xmath100l@xmath101 .
however , assuming a slightly flatter slope of @xmath51 instead of @xmath27 will decrease the estimated black hole mass based on h@xmath1 by a factor of @xmath52 . using furthermore a slightly steeper slope for mgii ( e.g. , @xmath51 instead of @xmath102 ) will provide on average @xmath103 times higher smbh estimates .
a second aspect as discussed in sect.4.2 is the size of the mgii emitting region .
if mgii originates at slightly larger distances to the center of the continuum emission than h@xmath1 , the estimated mass can be increased by an additional factor .
currently , we suppose that the different smbh estimates provided by iv civ and h@xmath1 compared to those based on mgii are predominantly caused by the different slopes of the r - l relations .
furthermore , it is still an open question whether the r
l relation can be extrapolated to quasar luminosities since the relation is derived for active galactic nuclei with @xmath104 ergs@xmath34 ( e.g. , corbett et al.2003 ) . in the case of quasars
this relation is extended to @xmath105 to @xmath106 times more luminous objects . yet , in spite of the uncertainties regarding the use of the h@xmath4 , mgii , and iv civ emission line profiles to determine the black hole mass , we emphasize that all three emission lines indicate consistently the presence of super - massive black holes with masses of @xmath107 at redshifts up to z=5.1 .
the presence of smbhs at high - redshifts and the close relation of quasar activity and galaxy formation provide valuable constraints on the epoch when these smbhs had to start to grow in their forming host galaxies .
the time span that is necessary to build - up a @xmath108 smbh can be described by an e - folding time for the growth of a single seed black hole ( e.g. , haiman & loeb 2001 ) .
we use the following equation to estimate the e - folding time scale and hence the epoch when the seed black hole started to grow ( salucci et al.1999 ) @xmath109 with @xmath110 as time elapsed from the initial time , @xmath111 , to the observed time , @xmath112 , @xmath113 as seed black hole mass , @xmath114 as eddington ratio , @xmath115 as efficiency to convert mass to energy , and @xmath116 as eddington time scale , with @xmath117yr ( rees 1984 ) .
the eddington time @xmath116 describes the time necessary to radiate at the eddington luminosity the entire rest mass of an object .
in hierarchical models for structure formation the first baryonic objects which collapsed had masses of the order of @xmath118 ( e.g. , silk & rees 1998 ; larson 2000 ; shibata & shapiro 2002 ; bromm & loeb 2003 ) .
this provides upper limit estimates for @xmath113 of black holes .
recent models of early star formation indicate that nearly metal free popiii stars , formed at high - redshifts ( @xmath119 ) , were predominantly very massive with @xmath120 ( e.g. , fryer , woosley , & heger 2001 ; abel , bryan , & norman 2002 ; bromm , coppi , & larson 2002 ) .
the stellar remnants are expected to be of the order of @xmath121 ( e.g. , fryer 1999 ) .
however , some early star formation models even indicate the possibility of black hole remnants with several times @xmath122 ( bond , arnett , & carr 1984 ) . in the following the efficiency @xmath115 to convert mass to energy is assumed to be @xmath7 .
it should be noted that the efficiency @xmath115 can reach values up to @xmath123 for a maximal rotating kerr black hole .
we employed equation ( 2 ) to derive the time @xmath124 which is necessary to accumulate the determined mean black hole mass ( table 3 ) for seed black holes with @xmath125 , @xmath126 , and @xmath127 , respectively .
we find that the time @xmath124 to build - up a smbh is between several 100 myr and @xmath128 gyr , i.e. , almost always @xmath129 is smaller than the age of the universe at the corresponding redshift . to illustrate this in a proper way we used @xmath124 to calculate @xmath130 ,
i.e. the epoch when the smbh , observed at @xmath112 , started to form .
in figure 4 the time @xmath130 is displayed as a function of the observed redshift @xmath131 .
the determined time scales @xmath124 and hence @xmath130 indicate that most of the smbh residing in the observed high - redshift quasars started to grow at redshifts of @xmath132 .
as can be seen for low massive black hole seeds , i.e. , stellar remnants with @xmath133 some black holes had to have started to grow at very early epochs and for q0105@xmath572634 and sdss0388@xmath1340021 the required time @xmath124is even larger than the actual cosmic age at @xmath131 . to avoid too long growing times @xmath129 ,
higher @xmath113 masses appear to be favored . assuming high mass seed black holes with @xmath135 , the smbh in the @xmath11 quasars began to grow at redshifts @xmath136 and earlier .
the luminous quasars at redshifts @xmath13 show a smaller scatter for @xmath130 compared to those at @xmath11 . based on our results the smbh residing in the @xmath137 quasars started at @xmath132
. however , stellar mass black hole seed are still a valid option because a flatter slope of the r
l relation for h@xmath1 with @xmath138 ( peterson et al.2000 ; mclure & jarvis 2002 ) will yield about two times less massive black holes .
hence , the eddington luminosity will be reduced and the eddington ratio will be increased . as a result
the growing time scale @xmath129 will decrease and smbhs can be built - up with stellar mass remnants as black hole seeds as well .
it is remarkable that most of the super - massive black holes of the studied high - redshift quasars are in place at the end of the era of reionization of the universe which is supposed to end at @xmath19 ( e.g. , becker at al.2001 ; fan et al.2002 ) .
particularly , for the quasars at redshifts @xmath13 we want to emphasize that the epoch when these black holes started to grow coincides with the beginning of first intense star formation that marks early phases of the host galaxy formation .
this epoch of major star formation also is indicated by the chemical enrichment history of the quasar gas . to achieve the estimated metallicities of several times solar ( e.g. , dietrich et al.2003b , c )
a major and intense star formation episode had to start at @xmath139 .
further circumstantial evidence for a vigorous star formation phase at those high - redshifts is given by the lack of evolution of the iron/@xmath45 element ratio compared to results for local quasars ( e.g. , barth et al.2003 ; dietrich et al.2002,2003a ; freudling et al.2003 ; iwamuro et al.2002 ; maiolino et al.2003 ; thompson et al.1999 ) .
the presence of super - massive black holes in luminous high - redshift quasars , the estimated time scale required to build - up black holes of several @xmath140 , the gas metallicity of at least solar up to several times solar , and the lack of evolution of the relative iron/@xmath45 element abundance ratio provide evidence that those quasars reside in already mature host galaxies , at least with an evolved massive central component ( hamann et al.2004 ) .
while the massive galaxy is forming , i.e. , major and vigorous star formation activity occurs , the super - massive black hole grows .
it takes about @xmath82 gyr to accumulate a smbh with several times @xmath141 to @xmath142 that is necessary to power quasar activity .
it takes approximately this same amount of time for a stellar population to enrich the gas around quasars to the observed solar or super - solar levels ( e.g. , dietrich et al . 2003b , c ; matteucci & padovani 1993 ; matteucci & recchi 2001 ) .
therefore , by the time most quasars become visible , even at the highest redshifts , a substantially evolved stellar population is harboring the active nucleus .
this conclusion is supported by some theoretical models of joint smbh
host galaxy evolution ( archibald et al.2002 , kawakatu et al.2003 ) .
direct imaging studies of low redshift quasars show clearly that the host galaxies are characterized by at least moderately old , stellar populations on @xmath143kpc scale ( nolan et al.2001 ; dunlop et al.2003 ) .
although there is less imaging data at higher redshifts at least some quasars still have substantial hosts ( kukula et al.2001 ) .
it can be assumed that early phases of galaxy formation that are accompanied by intense and vigorous star formation also mark the era when super - massive black holes are growing from massive seed remnants ( hamann et al.2004 ) .
a joint era of smbh growth and early phases of galaxy formation can be understood in the context of recently suggested galaxy formation models ( e.g. , archibald et al.2002 ; dimatteo et al.2003 ; granato et al.2001,2004 ) .
these models further suggest that the @xmath144 relation evolves over cosmic time scales and is finally established in the local universe ( dimatteo et al.2003 ) . a less steep @xmath31 relation also has been suggested by shields et al.(2003 ) .
this may provide an explanation to avoid huge host galaxy masses ( @xmath145 ) and large stellar velocity dispersions ( @xmath146 kms@xmath34 ) which have never been observed ( e.g. , netzer 2003 ) .
we have analyzed a sample of 15 luminous high - redshift quasars ( @xmath0 ) to estimate the mass of super - massive black holes . for the first time results for @xmath24 based on three different emission lines , iv civ , mgii , and h@xmath1 ,
are compared for high - redshift quasars .
we determine black hole masses in the range of @xmath147 up to @xmath3 similar to prior studies there appears to be an approximate upper limit for @xmath24 at @xmath84 .
the black hole mass estimates based on iv civ and h@xmath4 are quite consistent while mgii indicates by a factor of @xmath148 lower smbh masses . in spite of the uncertainties regarding the use of the h@xmath4 , mgii , and iv civ emission line profiles to determine the black hole mass , all three emission lines indicate consistently the presence of super - massive black holes with masses on average @xmath149 at redshifts up to z=5.1 .
we determine an eddington luminosity ratio of @xmath72 in the range of @xmath82 to @xmath83 .
this is consistent with results of studies of quasars sample that are dominated by low and intermediate redshift quasars .
this indicates that high - redshift quasars are not different regarding the eddington ratio @xmath72 compared to those at lower redshift .
to date the epoch when the smbhs which drive high - redshift quasar activity should have started to grow , we assume logarithmic growth by spherical accretion with a mass to energy conversion efficiency of @xmath7 . generally , the growth - time @xmath129 is of the order of several @xmath9 myr assuming higher mass seed black holes ( @xmath150 ) the smbh in the @xmath11 quasars begun to grow at redshifts @xmath136 and earlier while the quasars with @xmath13 started at @xmath132 .
the presence of super - massive black holes in luminous high - redshift quasars , the estimated time scale required to build - up black holes of several times @xmath140 , the several times solar gas metallicity and the lack of evolution of the relative iron/@xmath45 element abundance ratio provide evidence that luminous quasars at high - redshift reside in already mature host galaxies , at least with an substantially evolved massive central stellar component .
this result indicates that early phases of massive galaxy formation that are accompanied by intense and vigorous star formation also mark the era when super - massive black holes are growing from massive seed remnants .
the estimated black hole growth time scale suggest a delay of about @xmath82 gyr of the quasar activity with respect to preceding intense star formation which is consistent with the observed high gas metallicity .
we dedicate this paper to the memory of richard j. elston , who left us far too early on january 26 , 2004 .
md and fh acknowledge support from nasa grant nag 5 - 3234 and nsf grant ast-99 - 84040 ( university of florida ) .
the equations we applied , given by kaspi et al.(2000 ) , mclure & jarvis ( 2002 ) , vestergaard ( 2002 ) , and warner et al.(2003 ) have been modified for the use of the continuum luminosity @xmath151 that we measured at @xmath152 for each of the high - redshift quasars of this study .
we explicitly present the dependence on the spectral index @xmath45 which transforms the continuum luminosity measured at different wavelengths .
the uniform use of @xmath153 does not introduce an additional uncertainty , because instead of assuming a fixed continuum slope @xmath45 to rescale @xmath154 as in prior studies we used the measured continuum slope @xmath155 for each quasar which we determined for the power - law continuum fit ( @xmath156 , table 2 ) in the multi - component analysis of the quasar spectra ( dietrich et al.2002a,2003a ) .
the continuum slope is measured for a wavelength range of @xmath157 ( @xmath158 quasars ) and @xmath159 ( @xmath160 quasars ) . for the h@xmath1 emission line profile we determined @xmath24 applying the equation given by kaspi et al.(2000 )
( eq.a1 ) and presented by mclure & jarvis ( 2002 ) ( eq.a2 ) .
the main difference is given by the slope of the radius luminosity ( r l ) relation .
the black hole mass is estimated based on the iv civ@xmath25 line profile width applying the following equations ( vestergaard 2002 , eq.a3 ; warner et al.2003 , eq.a4 ) .
warner et al.(2003 ) assumed that the radius of the iv civ blr is half the size of the h@xmath4 emitting region which is well justified by variability studies ( peterson 1993 ) .
abel , t. , bryan , g.l .
, & norman , m.l .
2002 , science , 295 , 93 archibald , e.n . ,
dunlop , j.s . , jimenez , r. , friaa , a.c.s .
, mclure , r.j . , & hughes , d.h .
2002 , , 336 , 353 barth , a.j .
2003 , in coevolution of black holes and galaxies , carnegie observatories astrophysics series , vol.i , ed .
l.c.ho ( cambridge : cambridge univ.press ) , in press barth , a.j . ,
martini , p. , nelson , c.h .
, ho , l.c . 2003 , , 594 , l95 bechthold , j. , siemiginowska , a. , shields , j. , et al .
2003 , , 588 , 119 becker , r.h . , et al .
2001 , , 122 , 2850 bond , j.r . ,
arnett , w.d . , & carr , b.j .
1984 , , 280 , 825 boroson , t.a .
& green , r.f .
1992 , , 80 , 109 bromm , v. coppi , p.s . , &
larson , r.b .
2002 , , 564 , 23 bromm , v. & loeb , a. 2003 , , 596 , 34 bromm , v. , yoshida , n. , & hernquist , l. 2003 , , 596 , l1 cattaneo , a. , haehnelt , m.g . , & rees , m.j . 1999 , , 308 , 77 clavel , j. , reichert , g.a . , alloin , d. , et al . 1991 , , 366 , 64 corbett , e.a . ,
croom , s.m . ,
boyle , b.j .
2003 , , 343 , 705 dietrich , m. , appenzeller , i. , wagner , s.j . ,
et al . 1999 , , 352 , l1 dietrich , m. , appenzeller , i. , vestergaard , m. , & wagner , s.j .
2002a , , 564 , 581 dietrich , m. , hamann , f. , shields , j.c . , et al .
2002b , , 581 , 912 dietrich , m. , hamann , f. , appenzeller , i. , & vestergaard , m. 2003a , , 596 , 817 dietrich , m. , appenzeller , i. , hamann , f. , heidt , j. , jger , k. , vestergaard , m. , & wagner , s.j .
2003b , , 398 , 891 dietrich , m. , hamann , f. , shields , j.c . , constantin , a. , heidt , j. , jger , k. , vestergaard , m. , & wagner , s.j .
2003c , , 589 , 722 dietrich , m. & kollatschny , w. 1995 , , 303 , 405 dietrich , m. & wilhelm - erkens , u. 2000 , , 354 , 17 dimatteo , t. , croft , r.a.c . , springel , v. , & hernquist , l. 2003 , , 593 , 56 dunlop , j.s . ,
mclure , r.j . ,
kukula , m.j . ,
baum , s.a .
, odea , c.p .
, & hughes , d.h .
2003 , , 340 , 1095 elvis , m. , wilkes , b.j . ,
mcdowell , j.c .
, et al . 1994 , , 95 , 1 fan , x. , narayanan , v.k . ,
lupton , r.h .
, et al . 2001 , , 122 , 2833 ferrarese , l. , pogge , r.w . , peterson , b.m . , merritt , d. , wandel , a. , & joseph , c.l .
2001 , , 555 , l79 freudling , w. , corbin , m.r . , & korista , k.t .
2003 , , 587 , l67 fryer , c.l .
1999 , , 522 , 413 fryer , c.l . ,
woosley , s.e . , & heger , a. 2001 , , 550 , 372 gebhardt , k. , et al .
2000 , , 543 , l5 graham , a.w . ,
erwin , p. , caon , n. , & trujillo , i. 2001 , , 563 , l11 granato , g.l . ,
silva , l. , monaco , p. , panuzzo , p. , salucci , p. , dezotti , g. , & danese , l. 2001 , , 324 , 757 granato , g.l .
, dezotti , g. , silva , l. , bressan , a. , & danese , l. 2004 , , 600 , 580 grandi , s.a .
1982 , , 255 , 25 haiman , z. & loeb , a. 2001 , , 552 , 459 hamann , f. , dietrich , m. , sabra , b. , & warner c. 2004 , in origin and evolution of the elements , carnegie observatories astrophysics series , vol.iv , ed .
a.mcwilliams and m.rauch ( cambridge : cambridge univ.press ) , in press iwamuro , f. , et al .
2002 , , 565 , 63 kaspi , s. , smith , p.s . ,
netzer , h. , maoz , d. , jannuzi , b.t . , & giveon , u. 2000 , , 533 , 631 kauffmann , g. & haehnelt , m.g .
2000 , , 311 , 576 kawakatu , n. , umemura , m. , & mori , m. 2003 , , 583 , 85 kormendy , j. & richstone , d. 1995 , , 33 , 581 krolik , j.h .
2001 , , 551 , 72 kukula , m.j . ,
dunlop , j.s . ,
mclure , r.j .
, et al . 2001 , , 326 , 1533 laor , a. 2001 , , 553 , 677 larson , r.b .
2000 , in proc.33d eslab symp .
, star formation from the small to the large scale , ed .
f.favata , a.a.klaas , & a.wilson ( esa sp-445 ; noordwijk : esa ) , ( preprint astro - ph/9912539 ) magorrian , j. , tremaine , s. , & richstone , d. 1998 , , 115 , 2285 maiolino , r. , juarez , y. , mujica , r. , nagar , n.m . , & oliva , e. 2003 , , 596 , l155 matteucci , f. & padovani , p. 1993
, , 419 , 485 matteucci , f. & recchi , s. 2001 , , 558 , 351 mclure , r.j . ,
dunlop , j.s , kukula , m.j . ,
baum , s.a . ,
odea , c.p .
, & hughes , d.h . 1999 , , 308 , 377 mclure , r.j .
& dunlop , j.s .
2001 , , 327 , 199 mclure , r.j . &
dunlop , j.s .
2002 , , 331 , 795 mclure , r.j . &
jarvis , m.j .
2002 , , 337 , 109 menou , k. , haiman , z. , & narayanan , v.k .
2001 , , 558 , 535 merritt , d. & ferrarese , l. 2000 , , merritt , d. & ferrarese , l. 2001 , , 547 , 140 nelson , c.h .
2000 , , 544 , l91 netterfield , c.b .
, et al . 2002 , , 571 , 604 netzer , h. 2003 , , 583 , l5 nolan , l.a . ,
dunlop , j.s . ,
kukula , m.j . ,
hughes , d.h . ,
boroson , t. , & jimenez , r. 2001 , , 323 , 308 onken , c.a . & peterson , b.m .
2002 , , 572 , 746 percival , w.j . ,
miller , l. , mclure , r.j . , & dunlop , j.s .
2001 , , 322 , 843 peterson , b.m .
1993 , , 105 , 247 peterson , b.m .
1997 , in ` an introduction to active galactic nuclei ' , cambridge university press peterson , b.m . ,
balonek , t.j . , barker , e.s .
, et al . 1991 , , 368 , 119 peterson , b.m . & wandel , a. 1999 , , 521 , l95 peterson , b.m .
& wandel , a. 2000 , , 540 , l13 peterson , b.m . ,
mchardy , i.m . ,
wilkes , b.j .
, et al . 2000 , , 542 , 161 rees , m.j .
1984 , , 22 , 471 reichert , g.a .
, rodriguez - pascual , p.m. , alloin , d. , et al .
1994 , , 425 , 582 rocca - volmerange , b. , leborgne , d. , debeuck , c. , fioc , m. , & moy , e. 2004 , , 415 , 931 salucci , p. , szuszkiewicz , e. , monaco , p. , & danese , l. 1999 , , 307 , 637 shibata , m. & shapiro , s.l .
2002 , , 572 , l39 silk , j. & rees , m.j .
1998 , , 331 , l1 shields , g.a . , gebhardt , k. , salviander , s. , et al .
2003 , , 583 , 124 stirpe , g.m . , winge , c. , altieri , b. , et al .
1994 , , 425 , 609 storey , p.j . & hummer , d.g .
1995 , , 272 , 41 thompson , k.l . , hill , g.j .
, & elston , r. 1999 , , 515 , 487 tremaine , s. , et al . 2002 , , 574 , 740 umemura , m. , in coevolution of black holes and galaxies , carnegie observatories astrophysics series , vol.i , ed .
l.c.ho vestergaard , m. & wilkes , b.j .
2001 , , 134 , 1 vestergaard , m. 2002 , , 571 , 733 vestergaard , m. 2004 , , 601 , 676 wandel , a. , peterson , b.m . , &
malkan , m.a .
1999 , , 526 , 579 warner , c. , hamann , f. , & dietrich , m. 2003 , , 596 , 72 warner , c. , hamann , f. , & dietrich , m. 2004 , , in press willott , c.j .
, mclure , r.j . ,
& jarvis , m.j .
2003 , , 587 , l15 wills b.j . ,
netzer , h. , & wills , d. 1985 , , 288 , 94 woo , j .- h . &
urry , c.m .
2002 , , 579 , 530 lcccccccc bri 0019 - 1522 & ctio&09 - 00&300 & 230 & & 4.53 & 46.92&47.56 + br 0103 + 0032 & ctio&09 - 00&300 & 230 & & 4.44 & 46.99&47.63 + q 0103 - 260 & vlt & 12 - 98&750 & & & 3.38 & 46.83&47.47 + & ntt & 10 - 99&&500 & 430 & & & + q 0105 - 2634 & ntt & 10 - 99&&500 & 430 & 3.49 & 47.46&48.10 + pss j0248 + 1802&ctio&09 - 00&300 & 230 & & 4.44 & 47.24&47.88 + q 0256 - 0000 & ntt & 10 - 99&300 & 500 & 430 & 3.38 & 47.14&47.78 +
q 0302 - 0019 & ntt & 09 - 99&300 & 500 & 430 & 3.29 & 47.24&47.88 + sdss 0338 + 0021&vlt & 12 - 98&750 & & & 5.00 & 46.61&47.25 + pc 1158 + 4635 & keck&05 - 00&300 & 520 & & 4.73 & 47.39&48.03 + sdss
1204 - 0021&vlt & 04 - 01&&525 & & 5.09 & 46.79&47.42 + q 2050 - 259 & ctio&09 - 00&&230 & 230 & 3.51 & 47.44&48.08 + pks 2126 - 158 & ca & 08 - 93&300 & & & 3.27 & 47.61&48.25 + & ctio&09 - 00&&230 & 230 & & & + q 2227 - 3928 & ntt & 10 - 99&&500 & 430 & 3.44 & 47.07&47.71 + bri 2237 - 0607 & ctio&09 - 00&300 & 230 & & 4.56 & 47.27&47.91 + q 2348 - 4025 & ntt & 10 - 99&&500 & 430 & 3.31 &
47.27&47.91 + lcccc bri 0019 - 1522 & & @xmath166&@xmath167&@xmath168 + br 0103 + 0032 & & @xmath169&@xmath170&@xmath171 + q 0103 - 260 & @xmath172&@xmath173&@xmath174&@xmath175 + q
0105 - 2634 & @xmath176&@xmath177&&@xmath178 + pss j0248 + 1802&&@xmath179&@xmath180&@xmath181 + q
0256 - 0000 & @xmath182&@xmath183&@xmath184&@xmath185 + q
0302 - 0019 & @xmath186&@xmath187&@xmath188&@xmath189 + sdss 0338 + 0021&&&@xmath190&@xmath191 + pc 1158 + 4635 & & @xmath192&@xmath193&@xmath194 + sdss 1204 - 0021&&@xmath195&&@xmath196 + q 2050 - 259 & @xmath197&@xmath198&&@xmath199 + pks 2126 - 158 & @xmath200&&@xmath201&@xmath202 + q 2227 - 3928 & @xmath203&@xmath204&&@xmath205 + bri 2237 - 0607 & &
@xmath206&@xmath207&@xmath208 + q 2348 - 4025 & @xmath209&@xmath210&&@xmath211 + lccccccc bri 0019 - 1522 & & & @xmath212&@xmath213&@xmath214&@xmath215&1.73 + br 0103 + 0032 & & & @xmath216&@xmath217&@xmath218&@xmath219&3.48 + q 0103 - 260 & @xmath220&@xmath221&@xmath222&@xmath223&@xmath224&@xmath225&0.50 + q 0105 - 2634 & @xmath226&@xmath227 & & & @xmath228 & @xmath229&0.45 + pss j0248 + 1802 & & & @xmath230&@xmath231&@xmath232 & @xmath233&0.91 + q 0256 - 0000 & @xmath234&@xmath235&@xmath236&@xmath237&@xmath238&@xmath239&0.99 + q 0302 - 0019 & @xmath240&@xmath241&@xmath242&@xmath243&@xmath244&@xmath245&2.95 + sdss 0338 + 0021 & & & @xmath246&@xmath247 & & @xmath248&0.56 + pc 1158 + 4635 & & & @xmath249&@xmath250&@xmath251 & @xmath252&1.35 + sdss 1204 - 0021 & & & & & @xmath253 & @xmath254&1.00 +
q 2050 - 259 & @xmath255&@xmath256 & & & @xmath257 & @xmath258&1.38 + pks 2126 - 158 & @xmath259&@xmath260&@xmath261&@xmath262 & & @xmath263&1.01 + q 2227 - 3928 & @xmath264&@xmath265 & & & @xmath266&@xmath267&0.56 + bri 2237 - 0607 & & & @xmath268&@xmath268&@xmath269&@xmath270&2.25 + q 2348 - 4025 & @xmath271&@xmath272 & & & @xmath273 & @xmath219&6.67 + | we investigated a sample of 15 luminous high - redshift quasars ( @xmath0 ) to measure the mass of their super - massive black holes ( smbh ) and compare , for the first time , results based on iv civ , mgii , and h@xmath1 emission lines at high - redshifts . assuming gravitationally bound orbits as dominant broad - line region gas motion , we determine black hole masses in the range of @xmath2 up to @xmath3 .
while the black hole mass estimates based on iv civ and h@xmath4 agree well , mgii typically indicates a factor of @xmath5 lower smbh masses . a flatter slope of the h@xmath1 radius luminosity relation , a possibly steeper slope of the mgii radius luminosity relation , and a slightly larger radius of the mgii blr than for h@xmath1 could relax the discrepancy . in spite of these uncertainties ,
the iv civ , mgii , and h@xmath1 emission lines consistently indicate super - massive black hole masses of several times @xmath6 at redshifts up to z=5.1 . assuming logarithmic growth by spherical accretion with a mass to energy conversion efficiency of @xmath7 and an eddington ratio @xmath8 calculated for each quasar individually , we estimate black hole growth - times of the order of several @xmath9 myr which are smaller than the age of the universe at the corresponding redshift . assuming high - mass seed black holes ( @xmath10 ) the smbhs in the @xmath11 quasars began to grow at redshifts @xmath12 , while for the quasars with @xmath13 they started at @xmath14 .
these estimated time scales for forming smbhs at high redshifts , together with previous studies indicating high quasar metallicities , suggest that the main smbh growth phase occurs roughly contemporaneously with a period of violent and extensive star formation in proto - galactic nuclei . |
inspired by some aspects of string theory and loop quantum gravity , _ fuzziness _ of spacetime can be expressed using the following relation for non - commutativity of coordinate operators[1,2 ] @xmath0=i\theta^{ij}\ ] ] where @xmath1 is a real , antisymmetric matrix , with the dimension of length squared which determines the fundamental cell discretization of spacetime manifold . as a consequence of above relation ,
the notion of point in the spacetime manifold becomes obscure as there is a fundamental uncertainty in measuring the coordinates @xmath2 this finite resolution of the spacetime points especially affects the cosmological dynamics in early stages of the universe evolution . on the other hand ,
inflation has been identified as a great opportunity to test theories of planck scale physics including noncommutative geometry .
essentially , effects of trans - planckian physics should be observable in the cosmic microwave background radiation [ 3 - 9 ] .
for this reason , various attempts to construct noncommutative inflationary models have been done by adopting various different approaches .
these approaches include using relation ( 1 ) for space - space [ 10 ] and space - time [ 11 ] coordinates and constructing a noncommutative field theory on the spacetime manifold by replacing ordinary product of fields by weyl - wigner - moyal @xmath3 .
another way to incorporate effects of high energy physics in inflationary models is using the generalized uncertainty principle ( gup ) which is a manifestation of the existence of a fundamental length scale in the system [ 12 ] .
+ recently a new approach to noncommutative inflation has been proposed by rinaldi [ 13 ] using the coherent state picture of noncommutativity introduced in [ 14 ] .
this model is free from some of the problems that plagued models based on @xmath3 such as unexpected divergencies and uv / ir mixing ( see [ 15 ] for a full review ) .
the key idea in this model is that noncommutativity _ smears _ the initial singularity and as a result there will be an smooth transition between pre and post big bang eras via an accelerated expansion .
it has been shown that noncommutativity eliminates point - like structures in the favor of smeared objects in flat spacetime .
as nicolini _
have shown [ 16 ] ( see also [ 17 ] for some other extensions ) , the effect of smearing is mathematically implemented as a substitution rule : position dirac - delta function is replaced everywhere with a gaussian distribution of minimal width @xmath4 . in this framework , they have chosen the mass density of a static , spherically symmetric , smeared , particle - like gravitational source as follows @xmath5 as they have indicated , the particle mass @xmath6 , instead of being perfectly localized at a point , is diffused throughout a region of linear size @xmath4 . this is due to the intrinsic uncertainty as has been shown in the coordinate commutators ( 1 ) .
recently we have constructed a noncommutative braneworld inflation scenario [ 18 ] based on the idea that initial singularity is smeared in a noncommutative background . within the same streamline ,
the purpose of this letter is to study time evolution of cosmological perturbations in a braneworld inflation scenario in the context of spacetime noncommutativity .
the 5d field equations in the randall - sundrum ( rs ) ii [ 19 ] setup are @xmath7,\ ] ] where @xmath8 is a gaussian normal coordinate orthogonal to the brane ( the brane is localized at @xmath9 ) , @xmath10 is the brane tension , and @xmath11 is the energy - momentum tensor of particles and fields confined to the brane .
the effective field equations on the brane are derived from the gauss - codazzi equations and junction conditions ( using @xmath12-symmetry)[20,21 ] @xmath13 where @xmath14 is the high - energy correction term , which is negligible for @xmath15 , while @xmath16 is the projection of the bulk weyl tensor on the brane .
the general form of the brane energy - momentum tensor for any matter fields ( scalar fields , perfect fluids , kinetic gases , dissipative fluids , etc . ) , including a combination of different fields , can be covariantly given in terms of a chosen 4-velocity @xmath17 as @xmath18 here @xmath19 and @xmath20 are the energy density and isotropic pressure and @xmath21 and @xmath22 are the momentum density and anisotropic stress respectively .
@xmath23 defined as @xmath24 projects into the comoving rest space at each event where @xmath25 is the spacelike unit normal to the brane .
the modified friedmann and raychaudhuri equations in the background are [ 20 ] @xmath26 and @xmath27 respectively . by definition ,
@xmath28 where @xmath29 is the dark radiation energy density . for a matter content consisted of a perfect fluid or a minimally coupled scalar field the total effective energy density , pressure
, momentum density and anisotropic stress can be written as [ 21 ] @xmath30 where superscript @xmath31 denotes the contribution of the bulk weyl tensor which enters the modified friedmann equation as a non - local dark radiation term . using these definitions ,
the modified friedmann and raychaudhuri equations can be rewritten as @xmath32 the tracefree property of @xmath33 in equation ( 5 ) implies that the pressure obeys @xmath34 .
+ the local conservation equations on the brane are [ 21 ] @xmath35 where @xmath36 is the volume expansion rate , which reduces to @xmath37 in the frw background ( @xmath38 is the background hubble rate ) , @xmath39 is the 4-acceleration , and @xmath40 is the covariant derivative in the rest space .
the non - local conservation equations for the dark radiation matter can be expressed as [ 21 ] @xmath41 we now suppose that the initial singularity that leads to rs ii geometry afterwards , is smeared due to spacetime noncommutativity .
a newly proposed model for the similar scenario in the usual 4d universe suggests that one could write the energy density as [ 13,18 ] @xmath42 note that we suppose that the universe enters the rs ii geometry immediately after the initial smeared singularity which is a reasonable assumption ( for instance , from a m - theory perspective of the cyclic universe this assumption seems to be reliable , see ref . [
22 ] ) . using equation ( 20 ) , and setting @xmath43 , the friedmann equation ( 14 ) in noncommutative space could be rewritten as follows @xmath44 where @xmath45 is given by equation ( 10 ) . from equation
( 16 ) one can find the effective noncommutative pressure using equation ( 20 ) as @xmath46 so , the equation of state parameter will be @xmath47 and the speed of sound is @xmath48 using equations ( 10 ) and ( 11 ) we can find the _ effective _ equation of state and speed of sound . to do this end , we note that there are constraints from nucleosynthesis on the value of @xmath49 so that @xmath50 at the time of nucleosynthesis [ 23,24 ] . in this respect , we can neglect this contribution to find @xmath51\times\ ] ] @xmath52 ^{2}\times\ ] ] @xmath53^{-1}\ ] ] which simplifies to the following equation for the high energy regime ( @xmath54 ) @xmath55 similarly , the effective speed of sound in the high energy regime will be @xmath56 figure @xmath57 shows the evolution of the equation of state parameter and the effective equation of state parameter as given by equations ( 23 ) and ( 25 ) respectively . as one can see from this figure , there is a small variation in @xmath58 and @xmath59 around the smeared singularity .
figure @xmath60 shows the evolution of the effective speed of sound .
it is obvious from this figure that in @xmath61 and in the high energy noncommutative regime , @xmath62 is imaginary . in this respect ,
the evolution of the universe in the early , inflationary stage is a phantom evolution .
+ we use these results in the next section to determine time evolution of the cosmological perturbations .
the evolution of cosmological perturbations in the randall - sundrum braneworld scenario has been studied extensively ( see for instance [ 25 ] and references therein ) . to analyze the scalar perturbations in our noncommutative setup ,
we define energy density and expansion perturbations as [ 26 ] using the covariant 3 + 1 analysis developed in [ 27 ] @xmath63 similarly , the perturbations in nonlocal quantities associated with the dark radiation matter are defined as @xmath64 with these definitions , the equations governing on the evolution of perturbations ( equations ( 16)-(19 ) ) will take the following forms @xmath65 \delta \,,\\ { \dot{u } } & = & ( 3w-1)h { u } + \left({4c_{\rm s}^2\over 1+w}\right)\left({{\rho^\varepsilon } \over\rho}\right ) h\delta -\left({4{\rho^\varepsilon } \over3\rho}\right ) z - ad^2 { q}\,,\\ { \dot{q } } & = & ( 3w-1)h { q}-{1\over3a } { u}-{{2\over3 } } a { d^2\pi}+{1\over3 a}\left [ \left({4c_{\rm s}^2\over 1+w}\right){{\rho^\varepsilon}\over\rho}-3(1+w ) { { \rho\over\lambda}}\right]\delta\,.\end{aligned}\ ] ] where @xmath19 , @xmath58 and @xmath62 are given by equations ( 20 ) , ( 23 ) and ( 24 ) respectively . + in general ,
scalar perturbations on the brane can not be predicted by brane observers without additional information from the bulk because there is no equation for @xmath66 in the above set of equations .
nevertheless , it has been shown that on large scales one can neglect the @xmath67 term in equation ( 33 ) .
so , on large scales , the system of equations closes on the brane , and brane observers can predict scalar perturbations from initial conditions intrinsic to the brane without the need to solve the bulk perturbation equations [ 26,27 ] .
+ to solve the above system of equations using the simplification mentioned , we introduce two new variables ; the first is a scalar covariant curvature perturbation variable @xmath68 where @xmath69 is the ricci curvature of the surfaces orthogonal to @xmath17 .
the second variable is a covariant analog of the bardeen metric potential @xmath70 , @xmath71 along each fundamental world - line , covariant curvature perturbation , @xmath72 , is locally conserved @xmath73 with these new variables , the system of equations reduces to @xmath74\phi - \left[(1+w){a^2\kappa^4\rho^2\over 2 h}\right]u + \left[(1+w ) { \kappa^2 \rho\over 4h}\right]c_0\\ \dot{u } & = & -h\left[1 - 3w+{2\kappa^2{\rho^\varepsilon}\over 3h^2}\right]u -{2 { \rho^\varepsilon}\over 3 a^2 h\rho}\left[1+{\rho\over\lambda } - { 6 c_{\rm s}^2h^2\over ( 1+w)\kappa^2\rho}\right]\phi+ \left[{{\rho^\varepsilon}\over 3a^2h\rho}\right ] c_0\,.\end{aligned}\ ] ] if there is no dark radiation in the background , @xmath75 , then @xmath76 where @xmath77 is the number of e - folds . in this case , the above system reduces to a single equation for @xmath78 which is @xmath79\phi = ~{}\left[{(1+w)\kappa^2 \rho\over 4h^2}\right]c_o - \left[{3(1+w)a_o^2\rho^2\over \lambda h^2}\right]e^{2n}u\end{aligned}\ ] ] where @xmath80 is given by ( 39 ) .
we use these results in the next section to study noncommutative modifications of the scalar perturbations dynamics .
now we want to solve equation ( 40 ) using explicit noncommutative forms of @xmath19 , @xmath38 , @xmath58 and @xmath80 given by ( 20 ) , ( 21 ) , ( 23 ) and ( 39 ) respectively . to do this end
, we need to specify the noncommutative form of @xmath77 which has been appeared in equation ( 39 ) . as we have shown in ref .
[ 18 ] , the noncommutative number of e - folds is given by @xmath81\ ] ] @xmath82\ ] ] where @xmath83 denotes the error function . by expanding the error functions in equation ( 41 ) in series , the number of e - folds ( supposing that the universe enters the inflationary phase immediately after the big bang , that is , @xmath84 and @xmath85 ) will be given by @xmath86.\ ] ] now we can integrate equation ( 40 ) to find @xmath87 @xmath88 @xmath89.\ ] ] figure ( 2 ) shows the evolution of @xmath78 for both usual braneworld scenario and our noncommutative setup in the high energy inflation regime ( @xmath54 ) .
one should note that the subsequent evolution of the universe after times greater than a few @xmath4 should be governed by a matter content for particle creation in an expanding universe . ]
different than the one used in equation ( 21 ) ( _ i.e. _ energy density of the initial singularity smeared by noncommutativity ) .
so , the evolution of @xmath58 , @xmath90 and @xmath78 in the low energy regime essentially should be different .
the solution ( 43 ) is valid when there is no dark radiation in the background .
if @xmath91 , then one should solve the system of equations ( 37 ) and ( 38 ) using the explicit form of @xmath49 .
generally the time dependence of @xmath49 for brane observer is not determined .
here we introduce a possible candidate for this quantity : as we have mentioned previously , the constraint from nucleosynthesis on the value of @xmath49 is so that @xmath50 at the time of nucleosynthesis .
based on this constraint , we can assume for instance that @xmath49 is a small fraction of @xmath19 at a given time .
since the time evolution of @xmath19 is determined by ( 20 ) , the time evolution of @xmath49 can be supposed to be @xmath92 where @xmath93 is a small constant and less than @xmath94 .
this form of @xmath95 can be used to solve the system of equations ( 37 ) and ( 38 ) explicitly .
nevertheless , this procedure needs a lot of calculations with very lengthy solutions that we ignore their presentation here .
+ the curvature perturbation defined in metric - based perturbation theory is @xmath96 which reduces to @xmath97 on uniform density ( @xmath98 ) hypersurfaces .
if there is no dark radiation in the background ( @xmath99 ) , the total curvature perturbation on large scale is given by the following differential equation [ 24 ] @xmath100 { \delta\rho^{\varepsilon } \over ( \rho+p)(1+\rho/\lambda)}\,.\ ] ] where @xmath101 is matter perturbation which is zero for adiabattic perturbations . since the time variation of @xmath19 , @xmath38 , @xmath20 , @xmath62 and @xmath102 are given by equations ( 20 ) , ( 21 ) , ( 22 ) , ( 24 ) and ( 39 ) respectively , we can obtain the time evolution of the curvature perturbation explicitly as follows @xmath103 @xmath104 where @xmath105 is the exponential integral defined as @xmath106 .
figure @xmath107 shows the evolution of @xmath108 versus the cosmic time for both commutative and noncommutative brane inflation . for the commutative braneworld inflation
, we have considered the chaotic inflation with @xmath109 and we obtained the form of @xmath110 using its relation with @xmath19 .
the amplitude of perturbations in the commutative regime decays faster than the noncommutative regime with the same parameter values . as a result
we need smaller number of e - folds in the noncommutative regime to have a successful braneworld inflation .
spacetime noncommutativity as a trans - planckian effect , essentially could have some observable effects on the cosmic microwave background radiation . in this respect , it is desirable to study an inflation scenario within a noncommutative background .
recently we have shown the possibility of realization of a non - singular , bouncing , early time cosmology in a noncommutative braneworld scenario [ 18 ] . in that work , using the smeared , coherent state picture of the spacetime noncommutativity , we have constructed a braneworld inflation that has the potential to support the scale invariant scalar perturbations . here , following our previous work ,
we have studied the time evolution of the perturbations in this noncommutative braneworld setup .
we have neglected the contribution of the dark radiation term ( originating in the bulk weyl tensor ) in the background geometry to have a closed set of equations on the brane .
however , the contribution of this term in the evolution of perturbations on the brane are taken into account . in this way , by studying the effective quantities ( such as the effective equation of state and speed of sound ) we have realized the possibility of a phantom evolution in the early , inflationary stage of the universe history .
our analysis of the perturbations on the brane shows that the amplitude of perturbations in the commutative regime decays faster than the noncommutative regime with the same parameter values . as a result
we need smaller number of e - folds in the noncommutative regime to have a successful braneworld inflation . + * acknowledgment * + this work has been supported partially by research institute for astronomy and astrophysics of maragha , iran .
28 m. r. douglas and n. a. nekrasov , _ rev .
phys . _ * 73 * ( 2001 ) 977 - 1029 + r. j. szabo , _ phys .
_ * 378 * ( 2003 ) 207 + n. seiberg and e. witten , _ jhep _ * 9909 * ( 1999 ) 032 + a. connes and m. marcolli , [ arxiv : math.qa/0601054 ] + a. connes , _ j. math
* 41 * ( 2000 ) 3832 + a. konechny and a. schwarz , _ phys .
* 360 * ( 2002 ) 353 + m. chaichian _
et al _ , _ eur .
j. c _ * 29 * ( 2003 ) 413 + a. micu and m. m. sheikh - jabbari , _ jhep _ * 0101 * ( 2001 ) 025 + a. h. chamseddine and a. connes , [ arxiv:1004.0464 ] .
g. veneziano , _ europhys .
* 2 * ( 1986 ) 199 + d. amati , m. ciafaloni and g. veneziano , _ phys .
b _ * 197 * ( 1987 ) 81 + d. amati , m. ciafaloni and g. veneziano , _ int . j. mod . phys . a _ * 3 * ( 1988 ) 1615 + d. amati , m. ciafaloni and g. veneziano , _ nucl . phys .
b _ * 347 * ( 1990 ) 530 + d. j. gross and p. f. mende , _ nucl . phys
. b _ * 303 * ( 1988 ) 407 + d. amati , m. ciafaloni and g. veneziano , _ phys . lett .
b _ * 216 * ( 1989 ) 41 .
s. chu , b. r. greene and g. shiu , _ mod .
a _ * 16 * ( 2001 ) 2231 .
+ f. lizzi , g. mangano , g. miele and m. peloso , _ jhep _ * 0206 * ( 2002 ) 049 .
+ s. f. hassan and m. s. sloth , _ nucl .
b _ * 674 * ( 2003 ) 434 . s. alexander and j. magueijo , proceedings of the xiiird rencontres de blois , ( 2004 ) pp281 , [ arxiv : hep - th/0104093 ] + s. alexander , r. brandenberger and j. magueijo , _ phys .
* 67 * ( 2003 ) 081301 .
+ s. koh and r. h. brandenberger , _ jcap _ * 0706 * ( 2007 ) 021 .
a. smailagic and e. spallucci ,
_ j. phys .
a _ * 37 * ( 2004 ) 1 [ erratum - ibid . a * 37 * ( 2004 ) 7169 ] + a. smailagic and e. spallucci , _ j. phys .
a _ * 36 * ( 2003 ) l467 .
+ a. smailagic and e.spallucci , _ j. phys .
a _ * 36 * ( 2003 ) l517 .
t. g. rizzo , _ jhep _ * 09 * ( 2006 ) 021 + s. ansoldi , p. nicolini , a. smailagic and e. spallucci , _ phys .
b _ * 645 * ( 2007 ) 261 + e. spallucci , a. smailagic , p. nicolini , _ phys . lett .
b _ * 670 * ( 2009 ) 449 + k. nozari and s.h .
mehdipour , _ class .
quantum grav .
_ * 25 * ( 2008 ) 175015 + k. nozari and s. h. mehdipour , _ jhep _ * 0903 * ( 2009 ) 061 .
p. j. steinhardt and n. turok , _ phys .
* 65 * ( 2002 ) 126003 + p.
j. steinhardt and n. turok , _ nucl .
proc . suppl . _
* 124 * ( 2003 ) 38 + j. khoury , p. j. steinhardt and n. turok , _ phys .
* 92 * ( 2004 ) 031302 + n. turok and p.
j. steinhardt , _ phys .
scripta _ t*117 * ( 2005 ) 76 + m. bojowald , r. maartens and p. singh , _ phys .
* 70 * ( 2004 ) 083517 .
r. maartens , _ phys
* 62 * ( 2000 ) 084023 .
+ j. d. barrow and r. maartens , _ phys .
b _ * 532 * ( 2002 ) 153 + r. maartens , d. wands , b. a. bassett and i. p. c. heard , _ phys .
* 62 * ( 2000 ) 041301 .
+ e. j. copeland , a. r. liddle and j. e. lidsey , _ phys .
* 64 * ( 2001 ) 023509 .
+ d. langlois , r. maartens , d. wands , _ phys .
b _ * 489 * ( 2000 ) 259 .
+ s. mukohyama , _ phys .
* 62 * ( 2000 ) 084015 .
+ d. langlois , _ phys .
lett . _ * 86 * ( 2001 ) 2212 . | following our previous work in noncommutative braneworld inflation ( ref . [ 18 ] ) , in this letter we use smeared , coherent state picture of noncommutativity to study evolution of perturbations in a noncommutative braneworld scenario .
we show that in this setup , the early stage of the universe evolution has a phantom evolution with imaginary effective sound speed .
we show also that the amplitude of perturbations in the commutative regime decays faster than the noncommutative regime with the same parameter values , and as a result we need smaller number of e - folds in the noncommutative regime to have a successful braneworld inflation . + * pacs * : 02.40.gh , 11.10.nx , 04.50.-h , 98.80.cq + * key words * : inflation , perturbations , noncommutativity 0 mm 0 mm 0 mm 0 mm 0 mm 165 mm 220 mm .65 cm + and + |
in recent years , the effect of a longitudinal field in the ising antiferromagnet on an anisotropic simple cubic ( sc ) lattice has been discussed .
the experimental example is compound @xmath10 @xcite .
the differential magnetic susceptibility @xmath11 of this compound was analysed as a function of an extra external field ( @xmath12 koe ) and of temperature ( @xmath13 k ; @xmath14 k ) .
the compound is a typical layer - type ferromagnet , with a very weak antiferromagnetic coupling between the @xmath15 layers , where has been established the magnetic phase diagram of the antiferromagnetic structure .
one of the attractive points of investigating the properties of @xmath10 is that as a consequence of the antiferromagnetic interlayer coupling we may obtain quantitative information about the anisotropy and @xmath16 ( exchange coupling ) by investigating the field dependence of the susceptibility at @xmath17 . in previous papers @xcite
it has been reported that the @xmath18 compounds of general formula @xmath19 , where @xmath20 , @xmath21 , @xmath22 , @xmath23 , @xmath24 , @xmath25 , @xmath26 and @xmath27 or @xmath28 , may be considered as consisting of nearly isolated magnetic layers .
other example of the compound with cubic anisotropy are antiferromagnet @xmath29 @xcite , @xmath30 and @xmath31 @xcite .
three dimensional ( 3d ) ising models and ( pseudo- ) ising physical systems have been analysed extensively @xcite .
graim and landau @xcite studied the critical behavior of a spin-@xmath32 ising model on a simple cubic lattice with spatially anisotropic nearest - neighbor coupling using the monte carlo method .
this model on an anisotropic square lattice was investigated by using a modified mean - field theory in which the intrachain is treated exactly and the interactions between chains are considered in the mean - field theory @xcite .
various approximate methods have show this critical bahavior of the curve @xmath33 versus @xmath0 , such as mean field approximation ( mfa ) @xcite , effective - field theory ( eft ) @xcite , mean field renormalization group ( mfrg ) @xcite , effective - field renormalization group ( efrg ) @xcite , monte carlo simulation ( mc ) @xcite , and high - temperature series expansion ( se ) @xcite . for the case of the 3d lattice ,
the theoretical calculations show disagreement between differente methods .
the results obtained by the mfa and eft methods show a reentrant behavior in the phase diagram in low - temperature , i.e. , if @xmath0 is just above @xmath34 , then these are two phase transitions as the temperature is increased . in recent years
, the effect a longitudinal field in the ising antiferromagnetic on an anisotropic square lattice was explored by mc @xcite .
although mc simulations play an important role for the study of phase transitions and critical phenomena , the well - known difficulties arise when one uses standard algorithms ( one - flip algorithms ) @xcite for the study of first - and second - order phase transitions .
this has contributed for the development of alternative mc methods , such as parallel - tempering @xcite , cluster algorithms @xcite , multicanonical algorithms @xcite , and more recently wang - landau method @xcite . in the present paper
we use the mc simulations and effective - field theory in clusters with four spins ( eft-4 ) .
we investigate the first- and second - order phase transition in the plane @xmath5 of the ising superantiferromagnet on an anisotropic simple cubic lattice in the presence of a magnetic field .
standard finite - size scaling techniques were used to estimate the critical temperatures . in section
ii we present the model and formalism .
the numerical results and discussions are given in section iii .
finally , the last section is devoted to conclusions .
the model in this work is the nearest - neighbor ( @xmath35 ) ising antiferromagnetic in a longitudinal field magnetic divided into two equivalent interpenetrating sublattices @xmath36 e @xmath37 , that is described by following hamiltonian @xmath38 where @xmath39 is the @xmath40 component spin-@xmath32 pauli operator at site @xmath41 , @xmath42 is the exchange coupling along the @xmath43 axis , @xmath44 denotes the nearest - neighbor vector along the @xmath43 axis and @xmath0 is the longitudinal magnetic field .
we define the parameters @xmath45 and @xmath46 .
the ordered state for low temperatures and fields is the superantiferromagnetic order ( saf ) , which is characterrized by a parallel spin orientation in horizontal direction and an antiparallel spin orientation of a parallel spin orientation of nearest - neighbors in vertical direction and therefore exhibit nel order within the initial sublattice @xmath36 and @xmath37 ( see figure ( [ cubo2 ] ) ) . if @xmath8 ( @xmath47 ) , the lattice is now composed of independent planes , so the model is exactly solved for @xmath48 , and the critical temperature is obtained by the following relation @xcite @xmath49 where for the particular isotropic case @xmath50 ( @xmath51 )
we have @xmath52 . for @xmath53 , with @xmath54 , @xmath51 we have an ising model with an external magnetic field on an anisotropic square lattic ( [ 1 ] ) , which was already studied by mc @xcite .
accordingly , we improve the understanding of this model by studying it by means of the effective - field theory and monte carlo simulations for the case @xmath55 and @xmath53 . in order to implement the present model to perform mc by the metropolis algorithm ,
the simple cubic lattice of size @xmath56 having @xmath57 sites is decomposed into two sublattices ( @xmath36 and @xmath37 ) with opposite spins , corresponding to the saf ground state .
to meassure the saf order , the appropriate order parameter is defined by @xmath58 , where @xmath59 is the magnetization of the sublattice @xmath60 , @xmath37 and @xmath61 number of spins . the susceptibility related to this order parameter
is defined as follows :
@xmath62 where @xmath63 is the dimension of the lattice . in our simulations
we have considered lattices with periodic boundary conditions . in order to determine the system s behavior in the thermodynamic limit ( @xmath64 ) , which is imposible to implement on account of computational limitation
, we have to use a finite - size scaling theory @xcite . accordingly , to locate the critical temperature for second - order phase transitions , we approximately locate the crossing point of curves for different sizes of the fourth - order cumulant @xmath65 ( binder cumulant ) defined as @xcite @xmath66 where @xmath67 and @xmath68 are the canonical averages of the second and fourth moments of magnetization , respectively . on the other hand ,
the critical temperature @xmath69 can also be obtained by means of the relation @xmath70 , where @xmath71 is the pseudocritical temperature corresponding to the size @xmath56 , and @xmath72 is the correlation critical exponent . for first - order phase transitions the same formula applies by setting @xmath73 .
a more extensive description of the monte carlo method was published elsewhere @xcite and the reader is referred there for further details .
we performed simulations for @xmath55 , for several values of @xmath74 , with a temperature step @xmath75 and runs comprising up to @xmath76 mc steps after equilibration .
the statistical errors of the mc simulations used for the estimation of @xmath77 of a particular @xmath78 , @xmath79 and @xmath6 were found much smaller than the statistical errors coming from the fact that we used .
therefore , the error bars are not shown in our graphs because they are smaller than the symbol sizes . as a starting point ,
the averages of a general function involving spin operator components @xmath80 are obtained by @xcite @xmath81 where the partial trace tr@xmath82 is taken over the set @xmath83 of spin variables ( finite cluster ) specified by the multisite spin hamiltonian @xmath84 and @xmath85 indicates the usual canonical thermal average .
the method treats the effects of the surrounding spins of a finite cluster with @xmath86 spins through a convenient differential operator technique @xcite such that , in contrast to the usual mfa procedure , all relevant self - spin correlations are taken exactly into account .
the interactions within the cluster are exactly treated and the effect of the remaining lattice spins is treated by a given approximation ( here we use the random phase approximation - rpa ) . to treat the model ( [ 1 ] ) by the eft approach , we consider a simple example in cluster of size @xmath87 spins , and the hamiltonian for this cluster is given by @xmath88 where @xmath89 with @xmath90 , @xmath91 , @xmath92 and @xmath93 corresponds to @xmath35 vectors .
substituting eq .
( [ 5 ] ) in ( [ 4 ] ) , we obtain the average magnetizations in sublattices @xmath36 and @xmath37 , respectively , by @xmath94 and@xmath95with@xmath96where @xmath97 and @xmath98 . using the identity @xmath99 , where @xmath100 and @xmath101 are four- dimensional differential operator and vector , respectively , @xmath102 , and also the van der waerden identity for @xmath103 component pauli spin operator , i.e. , @xmath104 , eqs .
( [ 6 ] ) and ( [ 7 ] ) are rewritten as ( @xmath105 or @xmath37 ) @xmath106 with@xmath107and @xmath108where @xmath109 , @xmath110 , @xmath111 , @xmath112 , @xmath113 , @xmath114 , @xmath115 , @xmath116 , @xmath117 , @xmath118 , @xmath119 , @xmath120 , @xmath121 , @xmath122 , @xmath123 and @xmath124 .
the magnetization @xmath125 in eq .
( [ 9 ] ) is expressed in terms of multiple spin correlation functions .
the problem becomes unmanageable when we try to treat exactly all boundary spin - spin correlation function present in eq .
( [ 9 ] ) . here
we use a decoupling procedure that ignores all higher - order spin correlations on both right - hand sides in eq .
( [ 9 ] ) , namely @xmath126 where @xmath127 and @xmath128 @xmath129 .
the approximation ( [ 12 ] ) neglects correlations between different spins but takes relations such as @xmath130 exactly into account , while in the usual mfa all the self- and multi spin correlations are neglected .
we can then rewrite the eq .
( [ 9 ] ) in the form @xmath131 and the expression for the magnetization in sublattice @xmath37 is given by @xmath132 defining the uniform @xmath133 and staggered @xmath134 magnetizations , and using the identity @xmath135 , we obtain @xmath136 and @xmath137 where the expressions for the coefficients @xmath138 and @xmath139 are again omitted here .
we note that is not possible to calculate the firts - order transition line on the basis of only the equation of state , eq .
( [ 15 ] ) , to solve this problem one needs to calculate the free energy for each state ( p and saf ) .
assuming that this equantion of state is obtained by the minimization of a given free energy functional like @xmath140 ( i.e. , @xmath141 ) , then after intergration we obtain @xmath142 , \label{17}\ ] ] where @xmath143 are arbitrary functions which turn out to be irrelevant for searching the second and first - order transitions . to obtain this phase transitions we use maxwell construction , that correspond to the intersection point where the free energy between the phases are equal . in the case of the transitions between the saf ordered and p disordered @xmath144 phases we obtain the point of intersection @xmath145 from eq .
( [ 17 ] ) , i.e , @xmath146 the phase transition temperatures between the p and saf states are found by simultaneously solving three transcendental expressions , eqs .
( [ 15 ] ) , ( [ 16 ] ) and ( [ 18 ] ) . for a second - order transition
, we obtain @xmath147 , while the first - order transition we have @xmath148 , where this value corresponds to the discontinuity of the staggered magnetization at @xmath149 .
we obtained the phase diagrams in the @xmath5 plane for @xmath9 of the model [ 1 ] by using eft-4 and mc simulatons .
both methods confirmed the existence of a tricritical point in a critical frontier separating the saf order with the paramagnetic disorder as shown in figure 2 . for completeness ,
figure 2 also includes the critical frontier obtained by mc simulatons for the particular case @xmath51 , @xmath8 ( square lattice ) , which lacks of first - order criticality @xcite . in what effective - field calculations concerns ,
we obtain the critical frontier , which consists of the transition temperature as a function of the external magnetic field , separating the saf and p phases , by simultaneously solving the three transcendental equations eqs .
( [ 15 ] ) , ( [ 16 ] ) and ( [ 17 ] ) .
the range of ratio @xmath6 determines the limits of second - order @xmath150 and first - order @xmath151 phase - transition frontiers , where @xmath152 is the tricritical point ( tcp ) . from landau theory ,
the tcp is calculated by the condition given by @xmath153 , @xmath154 and @xmath155 , obtaining in eft-4 the following values : @xmath156 and @xmath157 .
the temperature variation of @xmath158 for fixed reduced field @xmath159 present two type of behavior .
the first for low - field the order parameter decrease to zero continuously as the reduced temperature approaches to the critical point , the temperature ( @xmath0 fixed ) where @xmath147 is the second - order phase transition temperatute , @xmath160 . on the other hand ,
when we are at high - fied the temperature at which the order parameter make a jump discontinuity is the first - order phase transitions going to zero discontinuously at the point @xmath161 . for h value greater than the upper limit of these field ( i. e. , @xmath162 ) , the system exhibits no phase transition ( the order parameter is null for all finite temperature , @xmath163 ) + + in order to obtain relevant critical points to get the critical frontier in mc simulations , we estimate the critical temperature for a given value of field by a finite - size scaling analysis .
for instance , for @xmath164 , figure 3 shows how the critical temperature was located around the crossing point of @xmath165 curves for sizes @xmath166 , resulting in @xmath167 .
note that critical temperatures for lower fields are superestimated by the eft-4 technique ( see figure 2 ) .
the limit of the second - order frontier was found by obtaining several @xmath165 curves for different field values for a given lattice size .
consequently , we found the beginning of the first - order criticality after detecting a change in the behavior of the binder cumulant curve above a certain value of field .
this is well exemplified in figure 4 , for two different field values @xmath168 , @xmath169 , for @xmath170 .
figure 4a shows that the binder cumulant curve already presents a clear first - order transition for @xmath171 , because of the sharply negative fluctuated peaks around the pseudocritical temperature .
moreover , figure 4b also suggests this first - order criticality by the aparent delta - dirac form of the susceptivility , for @xmath171 . in order to confirm what is suggested in short lattice sizes ,
we draw upon the fact that around the critical temperature of a phase transition the susceptibility peaks behave as @xmath172 , where @xmath173 ( @xmath63 is the lattice dimension ) , for a first - order criticality , whereas @xmath174 for a second - order one .
consequently , figure 5 shows a finite - size scaling of the susceptibility peak versus @xmath56 , for sizes @xmath175 , @xmath176 , @xmath177 and @xmath178 . for @xmath168
, we estimated @xmath179 , which is ( within the error bar ) an universal value @xcite for the second - order criticality of the three - dimensional ising model .
for @xmath171 , @xmath180 , which is close to @xmath181 , which suggests a first - order transition .
therefore , the tricritical point must be in the interval @xmath182 .
on the other hand , the first - order curve must end at @xmath183 .
+ + we could estimate some points of the first - order frontier by a finite - size scaling method , however , in order to avoid finite - size problems to obtain well behavied curves for the specific heat or the susceptibility around critical points belonging to the first - order frontier , we decided to estimate the corresponding critical temperatures for given fields by computing the free energy versus @xmath184 , from monte carlo data of energy versus @xmath185 , for a given size .
for instance , in figure 6a it is shown the hysteresis effect from energy data around the critical temperature obtained when cooling or heating the system too fast .
this is because of huge intrinsic autocorrelation times in finite - size systems .
we can improve the location of the critical temperature by obtaining the associated free energies of the low- and high - temperature branches@xcite .
consequently , the crossing point of the free energy branches gives a good estimation of the critical temperature , inasmuch the stable phase has the lower free energy , as shown in figure 4b .
+ + the first - order frontier in mc simulations does not present a reentrant behavior as eft-4 frontier does .
so , this reentrance seems to be an artifact of the effective - field approach .
in summary , we investigated the anisotropic three - dimensional nearest - neighbor ising model with competitive interations in an uniform longitudinal field .
we obtained the phase diagram in the plane @xmath5 , where the critical frontier separates the saf order with the paramagnetic disorder . for lower fields ,
the eft-4 approach superestimates the critical temperatures obtained by mc simulations .
a tricritical point was found for @xmath182 , by a finite - size analysis . at low temperatures
eft-4 calculations show a reentrant first - order frontier , which does not appear by mc simulations .
it suggests that improvements in treating correlations , or by increasing the cluster size in the effective - field approach , could correct this reentrant curve . at zero temperature , the critical field
is exactly obtained , so @xmath186 .
our quantitative estimation for the tricritical point could be bettered by using larger lattice sizes with better metropolis monte carlo techniques like parallel tempering .
we thank professor david p. landau for fruitful suggestions in the vi bmsp ( brazilian meeting on simulational physics 2011 ) .
this work was partially supported by cnpq ( edital universal ) and fapeam ( programa primeiros projetos - ppp ) ( brazilian research agencies ) .
( @xmath6 ) .
these frontiers separate the colinear order ( saf ) and the paramagnetic phase ( p ) .
the curve ( a ) was obtianed by mc simulations @xcite for the present model implemented in square lattices .
the solid line corresponds to second - order transitions .
the curves ( b ) and ( c ) are our results obtainded by mc and eft-4 methods , respectively .
the dashed lines corresponds to first - order transitions appeared for the present model implemented in cubic lattices.,width=340,height=340 ] , for the energy .
( b ) the associated free energy obtianed from the monte carlo data in ( a ) , by thermodynamic integration . ,
title="fig:",width=226,height=226 ] , for the energy .
( b ) the associated free energy obtianed from the monte carlo data in ( a ) , by thermodynamic integration .
, title="fig:",width=226,height=226 ] | we have studied the anisotropic three - dimensional nearest - neighbor ising model with competitive interactions in an uniform longitudinal magnetic field @xmath0 .
the model consists of ferromagnetic interaction @xmath1 in the @xmath2 direction and antiferromagnetic interaction @xmath3 in the @xmath4 direction .
we have compared our calculations within a effective - field theory in clusters with four spins ( eft-4 ) in the simple cubic ( sc ) lattice with traditional monte carlo ( mc ) simulations .
the phase diagrams in the @xmath5 plane ( @xmath6 ) were obtained for the particular case @xmath7 ( anisotropic sc ) .
our results indicate second - order frontiers for all values of @xmath0 for the particular case @xmath8 ( square lattice ) , while in case @xmath9 , we observe first- and second - order phase transitions in the low and high temperature limits , respectively , with presence of a tricritical point .
using eft-4 , a reentrant behavior at low temperature was observed in contrast with results of mc . |
clusters of galaxies are a unique probe of the growth and dynamics of structure in the universe .
in particular , active mergers of sub - clusters provide a window to the processes by which massive clusters are assembled . in these systems , the galaxies and associated dark matter are essentially collisionless .
in contrast , the ionized intracluster gas , typically at temperatures of @xmath10 k , is strongly interacting and experiences complex dynamics . in extreme cases , the normally associated dark matter and intracluster gas can be significantly separated .
the bullet cluster ( 1e 065756 ) at @xmath11 , is a massive cluster consisting of two sub - clusters in the process of merging . the smaller sub - cluster or
bullet " has passed through the larger main cluster .
x - ray observations infer a bow shock velocity of @xmath124700 km / s ( markevitch 2006 ) , while simulations of the collision yield a substantially lower speed for the sub - cluster ( springel & farrar 2007 ) .
this collision is perpendicular to the line of sight , providing an ideal system for studying interacting sub - clusters ( clowe et al .
2006 ) .
the mass surface density of the bullet cluster has been measured using weak and strong gravitational lensing of light from background galaxies .
there are significant angular offsets between the peaks of the x - ray surface brightness , which trace the baryonic gas through thermal bremsstrahlung emission , and the peaks of the lensing surface density , which are associated with the majority of the mass .
the combined weak and strong lensing analyses of ( 2006 ) show that the main cluster and sub - cluster are separated from their associated x - ray peaks at @xmath13 and @xmath14 significance respectively .
this result has been recognized as providing direct evidence for the presence of collisionless dark matter in this system ( clowe et al .
2006 ) .
the sunyaev - zeldovich effect ( sze ) provides an independent probe of the intracluster gas . in the sze , a small fraction ( @xmath15 ) of cosmic microwave background ( cmb ) photons
undergo inverse compton scattering from intracluster electrons ( sunyaev & zeldovich 1970 ; birkinshaw 1999 ) .
this process distorts the planck blackbody spectrum of the cmb and produces a signal proportional to the gas pressure integrated along the line of sight . at @xmath16ghz ,
the sze produces a temperature decrement with respect to the unperturbed cmb intensity .
early detections of the sze in the bullet cluster include work by ( 1999 ) and ( 2004 ) . unlike the x - ray surface brightness , the peak sze surface brightness for a given cluster is independent of redshift . therefore , the sze has the potential to be an effective probe of intracluster gas out to the redshifts at which clusters are assembled .
sze measurements of galaxy clusters provide complementary constraints on cluster properties typically derived from x - ray measurements such as central electron density , core radius of the intracluster gas , cluster gas mass , and fraction of the total cluster mass in gas
. since the sze and x - ray signals are proportional to the line - of - sight integral of the electron density and electron density squared respectively , sze results will be less sensitive to clumping of the intracluster gas .
for all comparisons between sz and x - ray data , we assume a @xmath17cdm cosmology , with @xmath18 , @xmath19 , and @xmath20 . in this paper , we present a 1 resolution sze image of the bullet cluster at @xmath16ghz made with the apex - sz instrument .
it is the first reported scientific result from observations with a large array of multiplexed superconducting transition - edge sensor bolometers . in
[ sec : observations ] , we discuss the instrument and observations .
calibration is discussed in [ sec : calibration ] . in [ sec : datareduction ] , we describe the data reduction procedure , and in [ sec : results ] , we present the results of fits to the sze surface brightness with cluster models , including mass - weighted electron temperature and gas mass fraction calculations .
we summarize the conclusions and discuss future work in [ sec : conclusions ] .
apex - sz is a receiver designed specifically for sze galaxy cluster surveys ( schwan et al . 2003 ; dobbs et al . 2006 ; schwan et al .
2009 , in preparation ) .
it is mounted on the 12-meter diameter apex telescope , located on the atacama plateau in northern chile ( gsten et al .
the observing site was chosen for its extremely dry and stable atmospheric conditions .
the mean atmospheric transmittance is frequently better than 95@xmath21 in the apex - sz frequency band at 150 ghz ( peterson et al .
2003 ; chamberlain & bally 1995 ) .
the telescope is capable of round - the - clock observations .
three reimaging mirrors in the cassegrain cabin couple the apex telescope to the focal plane of apex - sz .
we achieve the diffraction limited performance of the telescope across the entire 0.4field of view with a mean measured beam full width half maximum ( fwhm ) of 58 , and a measured beam solid angle of 1.5 arcmin@xmath22 , including measured sidelobes at the @xmath23 db level .
the apex - sz receiver houses a cryogenic focal plane , operating at 0.3 k. the focal plane contains 330 horn - fed absorber - coupled superconducting transition - edge sensor bolometers ( richards 1994 ; lee et al .
1996 ) , with 55 detectors on each of six sub - array wafers . of the 330 detectors ,
280 are read out with the current frequency - domain multiplexed readout hardware .
we measure the median individual pixel noise equivalent power ( nep ) to be @xmath24 w/@xmath25 and the median noise equivalent temperature ( net ) to be @xmath26k@xmath27 .
the measured optical bandwidth of the receiver is 40% narrower than the design goal of 38 ghz , resulting in lower sensitivity than anticipated .
the large field of view of the apex - sz instrument is designed for surveying large areas of sky . in order to efficiently observe a single target
, we use the circular scan pattern illustrated in figure [ fig : scanpattern ] .
the circle center is fixed in az / el coordinates for twenty circular sub - scans , with a total duration of 100 seconds .
this choice has a number of important advantages .
the sky signal is modulated so that it appears in the timestream at frequencies higher than atmospheric drifts and readout @xmath28 noise .
in addition , the circle scan has a moderate continuous acceleration ; the lack of high acceleration turn - arounds makes it possible to achieve a high observing efficiency .
approximately @xmath29 of the total observing time is spent moving the telescope to a new center position before the start of the next scan .
every bolometer maps a @xmath30 sub - field , with a combined map field of @xmath31 every 100 seconds .
marker indicates the source position with respect to the scan pattern .
the center of the circles is constant in azimuth and elevation as the source drifts across the field .
the small disk in the lower left indicates the 58 mean fwhm beam for a single bolometer.,scaledwidth=45.0% ] observations of the bullet cluster were conducted over a period of seven days in august 2007 , when the cluster was visible between the hours of 03:00 and 15:00 local time .
the weather over this period was typical for the site , with precipitable water vapor varying between 0.25 and 1.5 mm , and a median atmospheric transmittance of @xmath32 .
for the analysis in this paper , 235 scans are used , each scan consisting of twenty 5-s circular sub - scans , for a total of 6.4 hours of on - source data .
the response of the receiver to astronomical sources is measured with daily raster scans of mars over every bolometer in the array . for each bolometer ,
the observations provide a primary flux calibration and a high signal - to - noise beam profile from which we determine beam parameters such as size , ellipticity , and position with respect to the array - center pointing .
additional observations of rcw57 and rcw38 are used to monitor gain stability , and frequent observations of bright quasars near the cluster source are used to monitor pointing stability .
the wmap satellite has been used to calibrate the brightness temperature of mars at @xmath33ghz in five measurement periods spanning several years ( hill et al .
the wmap mars temperatures are tied to the cmb dipole moment and are accurate to better than @xmath34 .
the brightness temperature of mars changes significantly ( @xmath35 ) as a function of its orbit and orientation .
we use a version of the rudy model ( rudy et al . 1987 ; muhleman & berge 1991 ) , that has been updated and maintained by bryan butler , to transfer the wmap mars temperature results to the apex - sz frequency band and specific times of our mars observations . after applying a constant scaling factor
, we find the rudy model predictions for the mars brightness temperature to be in excellent agreement with the wmap measurements .
we find that the rudy model brightness temperatures at @xmath33ghz are systematically a factor of @xmath36 higher than those measured by wmap in the five published observation periods .
in contrast , repeating the same exercise with the thermal model developed by ( 1976 , 2007 ) , as implemented in the online jcmt - fluxes program , results in a scaling factor of @xmath37 .
this is consistent with the 10% rescaling of this model called for in ( 2009 ) , but the scaling factor exhibits significantly larger rms scatter than that of the rudy model .
we therefore use the wmap @xmath33ghz calibrated rudy model to compute the mars brightness temperatures at @xmath16ghz for the specific times of our mars observations by reducing the rudy model @xmath16ghz temperatures by a factor of 1.052 .
the rudy model @xmath33ghz to @xmath16ghz frequency scaling factor is @xmath38 at the times of our mars observations , and we adopt the rms scatter in this frequency scaling factor as an estimate of its uncertainty . combining the uncertainties in the wmap
mars calibration , the wmap to rudy model scaling factor at @xmath33ghz , and the rudy model frequency scaling factor , we estimate the uncertainty in mars temperature to be @xmath39 .
the measured signals from the calibrators are corrected for atmospheric opacity , which is measured with a sky dip observation at the beginning and end of each day s observations .
measured zenith transmittance over the observing period ranged between 0.92 and 0.98 , with a median of 0.97 .
based on the observed temporal variability of the opacity , drifts in atmospheric opacity between the sky dip and observation contribute @xmath40 to the overall calibration uncertainty . after correcting for the atmospheric opacity
, we find that the mars temperature measured by apex - sz varies from the model prediction by up to @xmath41 over the course of the observation period .
this gain variation is included as a source of error in the final calibration uncertainty .
the apex - sz observing band center is measured with a fourier transform spectrometer to be @xmath42 ghz . the uncertainty in the band center results in a @xmath43 uncertainty in extrapolation of the mars based calibration to cmb temperature .
the beam shape , including near sidelobes , is characterized by creating a beam map from mars observations , combining the same bolometer channels that are used to make the science maps .
we adjust the calibration and measured beam size for the small ( @xmath44 ) correction due to the @xmath45 angular size of mars .
we estimate a fractional uncertainty in the beam solid angle of @xmath46 .
the apex - sz detectors operate in a state of strong negative electrothermal feedback which results in a linear response to changes in the input optical power .
we have measured the response of the detectors during sky dips between @xmath47 and @xmath48 elevation ( antenna temperature difference @xmath49k ) , and find no significant deviation from the expected linear response to loading .
we therefore conclude that detector non - linearity makes a negligible contribution to the calibration uncertainty . slowly changing errors in telescope pointing result in both a pointing uncertainty and a flux calibration uncertainty due a broadening of the effective beam pattern . to measure pointing errors during our observations , we observe a bright quasar within a few degrees of the bullet cluster every 12 hours , and apply a pointing correction as needed .
the typical rms pointing variations of the apex telescope between quasar observations is @xmath50 .
this pointing uncertainty results in a slightly larger effective beam for the coadded maps than is measured with the individual calibrator maps .
the correction to the flux calibration of the coadded maps due to pointing uncertainty is negligible , particularly for the observation of extended objects such as the bullet cluster .
we estimate the pointing uncertainty in the coadded maps to be @xmath51 in both the ra and dec directions .
the uncertainty in the cmb temperature calibration of the apex - sz maps is summarized in table [ tbl : cal ] .
the combination of all contributions to the calibration uncertainty described above results in an overall point source flux uncertainty of @xmath8 .
ll wmap mars temperature at @xmath33ghz & @xmath52 + rudy model to wmap scaling factor at @xmath33ghz & @xmath52 + @xmath33ghz to @xmath16ghz frequency scaling factor & @xmath53 + frequency band center & @xmath54 + beam solid angle & @xmath55 + atmospheric attenuation & @xmath56 + gain variation & @xmath57 + total & @xmath58 +
the data consist of 280 bolometer timestreams sampled at 100 hz , telescope pointing data interpolated to the same rate , housekeeping thermometry data , bolometer bias and readout configuration data , and other miscellaneous monitoring data .
the fundamental observation unit is a scan comprising twenty 5-s circular sub - scans in az / el coordinates , allowing the source to drift through the field of view ( fov ) , as described in
[ sec : observations ] above .
data reduction consists of cuts to remove poor - quality data , filtering of @xmath28 and correlated noise due to atmospheric fluctuations , and binning the bolometer data into maps .
these steps are described in more detail below .
timestream data are first parsed into individual circular sub - scans .
we reject @xmath59% of the data at the beginning and end of the scan where the telescope deviates from the constant angular velocity circular pattern .
we reject bolometer channels that are optically or electronically unresponsive , or lack high - quality flux calibration data ; typically , 160200 of the 280 bolometer channels remain after these preliminary cuts . the large number of rejected channels is due primarily to low fabrication yield for two of the six bolometer sub - array wafers .
we reject spikes and step - like glitches caused by cosmic rays or electrical interference .
these are infrequent and occur on time scales faster than the detector optical time constant .
we use a simple signal - to - noise cut on the data to reject these , since the timestream is noise dominated even for the @xmath121 mk bullet cluster signal .
step - like glitches are often correlated across many channels in the array , so we reject data from all channels whenever a spike or glitch in @xmath60 of the channels is detected .
these cuts result in a loss of 8% of the remaining data . for each circular sub - scan , we also reject channels that have excessive noise in signal band , resulting in a loss of 19% of the remaining data . after the timestream data cuts , fluctuations in atmospheric emission
produce the dominant signal in the raw bolometer timestreams .
the atmospheric signal is highly correlated across the array , which can be exploited to remove the signal . principal component analysis ( pca )
has been used by some groups to reduce the atmospheric signal ( see , e.g. , scott et al . 2008 ; laurent et al . 2005 ) .
however , the effect of pca filtering on the source is difficult to predict and is a function of the atmospheric conditions .
we have developed an analysis strategy that reduces the atmospheric signal through the application of spatial filters that have a constant and well understood effect on the signals we are attempting to measure .
atmospheric fluctuation power is expected to follow a kolmogorov spatial power spectrum , with most power present on scales larger than the separation between beams as they pass through the atmosphere , resulting in an atmospheric signal that is highly correlated across the array . to reduce these fluctuations , we first remove a polynomial and an elevation dependent airmass opacity model from each channel s timestream , then
remove a first - order two - dimensional spatial polynomial across the array for each time - step .
this algorithm is described in detail in the two following subsections .
this atmospheric fluctuation removal strategy requires that both the spatial extent of the scan pattern and the instantaneous array fov are larger than the source .
the @xmath61 radius circular scan and the 23@xmath62 array fov allow us to recover most of the bullet cluster s flux , but some extended emission is lost as is described in [ sec : results ] .
we observe scan - synchronous signals in the bolometer timestreams due to elevation - dependent atmospheric emission .
the optical path length @xmath63 through the atmosphere is proportional to the cosecant of the elevation angle @xmath64 , @xmath65 .
the change in optical path - length is nearly a linear function of elevation angle over the @xmath61 circular scan radius . in the circular drift scans , this modulation of the elevation - dependent opacity produces an approximately sinusoidal modulation in the bolometer timestream . for each channel in each scan , we simultaneously fit and remove an atmospheric model consisting of this cosecant function plus an order 20 polynomial ( one degree of freedom per circular sub - scan ) to remove slow drifts in the atmospheric opacity and readout @xmath28 noise . this scheme effectively removes the common scan - induced atmospheric signal as well as most of the atmospheric fluctuation power below the frequency of the circular sub - scan ( 0.20 hz ) , while only modestly affecting the central bullet cluster signal . removing the cosecant - plus - polynomial model from the timestream data reduces low - frequency atmospheric fluctuation power , but not higher frequency power corresponding to smaller spatial scales near those where the cluster signal occurs . to reduce these fluctuations ,
the atmosphere can be modeled as a spatially correlated signal across the array pixel positions on the sky with a low - order two - dimensional polynomial function . at each time step
, we fit and remove a low - order two - dimensional polynomial function across the array , similar to the procedure described in sayers ( 2007 ) . the relative gain coefficients for each bolometer channel
are calculated by taking the ratio of each channel s timestream , which is dominated by correlated atmospheric noise , to a median timestream signal generated from all channels . with the favorable atmospheric conditions of these observations , we find that a first order spatial polynomial ( offset and tilt ) is adequate to remove most of the atmospheric signal while preserving the cluster signal
. bolometer channels with excess uncorrelated noise are more easily identified after removing the correlated atmospheric noise component ; we reject these noisy channels , then perform the spatially correlated signal removal a second time . the atmospheric removal algorithms described above act as a high - pass filter .
they suppress signals on scales comparable to the scan length or the focal plane fov .
the cluster emission can be quite extended , and therefore the data reduction filtering process attenuates diffuse flux in the cluster signal and produces small positive sidelobes around the cluster decrement . the data reduction pipeline filters can be tailored , within limits , to meet various scientific objectives .
thus , our primary data products consist of two different high signal - to - noise maps of the cluster . for one map
, we mask a circular region centered on the cluster source prior to fitting the timestream and spatial filters described in [ sec : atmremoval ] , then apply the resulting filter functions to the entire data set , including the source region .
the source - mask procedure prevents the cluster signal within the masked region from influencing the baseline fits , and thus reduces attenuation of the source central decrement and extended emission at the expense of increasing the contribution of low - frequency noise in the map center .
we choose a source - mask radius of @xmath66 as a compromise between attenuation of diffuse emission and increased map noise .
we use the source - masked map to visually interpret the morphology and extended emission in the cluster .
these results are discussed in
[ subsec : tempmap ] .
we also produce a map in which we do not mask the source when applying filters .
the non - source - masked map is used for model parameter estimation because it has higher signal to noise in the central region of the map .
in addition , it is easier to take into account the effects of the data reduction filters , or transfer function , on the underlying sky intensity distribution , which is necessary for comparing the data to the model for parameter estimation .
the fitting procedure and results are described in more detail in [ subsec : fitting ] . for each of the two maps ,
the post - cut , filtered timestream data are binned in angular sky coordinates to create maps . for a given scan ,
a map is created from each bolometer channel , applying the channel s pointing offset and flux calibration .
a coadded scan map is created by combining individual channel maps with minimum variance weighting in each pixel , using the sample variance of the conditioned timestream data in the scan .
the final coadded map is created by combining all scan maps , again with minimum variance weighting in each pixel .
we bin maps at a resolution of 10 to oversample the beam .
the source - masked map that we present in [ sec : results ] is convolved with a 1 fwhm gaussian to smooth noise fluctuations to the angular size of the beam .
however , the radial profiles presented below and the non - source - masked map used for model fitting do not include this additional smoothing .
figures [ fig : bulletcoadd ] and [ fig : bulletcoaddzoom ] show the source - masked temperature map from our observations of the bullet cluster .
the map has a resolution of 85 fwhm which results from the combination of the 58 instrumental beam , the data reduction filters , and a final 1 fwhm gaussian smoothing of the map .
the source - masked map is shown in order to provide a more accurate representation of the extended emission and cluster morphology .
the noise in the central region of the source - masked map is 55 @xmath67 per 85 fwhm resolution element . near the cluster center ,
the emission hints at elongation in the east - west direction , which is along the axis between the main and sub - cluster gas detected in the x - ray , see [ sec : xray ] .
the more extended emission appears to be elongated in the northwest - southeast direction , which is the major axis of the best - fit elliptical @xmath3 model discussed in
[ subsec : fitting ] .
figure [ fig : bulletcoaddzoom ] shows the centroid position of the best - fit elliptical @xmath3 model , and the position of the dust obscured , lensed galaxy detected at 270 ghz by ( 2008 ) . as discussed in
[ sec : source_contributions ] , we see no evidence for emission from this source in our 150 ghz map . radial profiles for the unsmoothed source - masked and non - source - masked maps are shown in figure [ fig : radial_plots ] . the source - masked map has a signal - to - noise of 10 within the central 1 radius , compared to 23 for the non - source - masked map , due to the fact that source - masking allows more large - scale atmospheric fluctuation noise to remain in the map . however , the source - masked map preserves signal on larger spatial scales than the non - source - masked map . in both the source - masked and non - source - masked maps ,
the sky intensity distribution is filtered by the instrument beam and data reduction pipeline described in [ sec : datareduction ] .
we do not renormalize the map amplitudes , since the source is extended and an assumption would need to be made about the shape of the sky - brightness distribution to do so .
however , in order to accurately estimate cluster parameters such as the central temperature decrement , a model for cluster emission must be adopted , and the instrument beam and data reduction filtering must be taken into account . , with color scale adjusted to the limits of the detail region , and a contour interval of 100 @xmath68 .
the @xmath69 marker indicates the centroid position of the best - fit elliptical @xmath3 model , see [ subsec : fitting ] .
the @xmath70 marker indicates the position of the bright , dust obscured , lensed galaxy detected at 270 ghz by ( 2008 ) , see [ sec : source_contributions].,scaledwidth=45.0% ] we fit an elliptical @xmath3 model to the non - source - masked temperature map to allow a straightforward comparison of cluster gas properties derived from our measurements to those derived from x - ray observations . in all analyses here
we assume the cluster gas is well - described by an isothermal @xmath3 model , and is in hydrostatic equilibrium .
these assumptions are unphysical in the case of the bullet cluster , which is a dynamically complex merging system where the gas is separated from the rest of the mass ( clowe et al .
2006 ) . however , we find that with the sensitivity and spatial resolution of the observations , these assumptions yield an adequate description of the observed emission .
we model the three - dimensional radial profile of the electron density with an isothermal @xmath3 model ( cavaliere & fusco - femiano 1978 ) : @xmath71 here , @xmath72 is the central electron number density , @xmath73 is the core radius of the gas distribution , and @xmath3 describes the power - law index at large radii .
the radial surface temperature profile of the sze takes a simple analytic form : @xmath74 where @xmath75 is the central temperature decrement , and @xmath76 is the core radius divided by the angular - diameter distance . a similar form exists for the x - ray surface brightness .
because of the significant ellipticity in the measured sze intensity profile , we generalize the cluster gas model to be a spheroidal rather than spherical function of the spatial coordinates : @xmath77 with @xmath78 @xmath79 here @xmath80 and @xmath81 are angular offsets on the sky in the ra and dec directions , with respect to center positions @xmath82 and @xmath83 .
the axial ratio , @xmath84 , is the ratio between the minor and major axis core radii , @xmath85 is the angle between the major axis and the ra ( @xmath86 ) direction .
@xmath75 is given by the gas pressure integrated along the central line of sight through the cluster : @xmath87 where @xmath88 , @xmath89 describes the frequency dependence of the sze , @xmath90 is the thomson scattering cross - section , and @xmath91 k. for all results in this paper , we use the relativistic sze spectrum @xmath89 provided by ( 2000 ) , and neglect the kinematic effect . at 150
ghz and @xmath92 kev ( see [ sec : xray ] ) , this is a 9% correction to the non - relativistic value . to accurately estimate @xmath3 model parameters for the cluster , the instrument beam and data reduction filters , or transfer function , must be applied to the model before comparing it with the data ( see , e.g. , benson et al . 2003 ; reese et al .
we characterize this transfer function by creating a map from a simulated point source , convolved with the instrument beam , and inserted into a noiseless timestream , similar to the method described in ( 2008 ) .
the point source transfer function map @xmath93 is then convolved with a simulated @xmath3 model cluster map @xmath94 to generate a filtered model map @xmath95 , which is a noiseless simulated apex - sz observation of a @xmath3 model cluster . the filtered model map , @xmath95 , is then differenced with the data map , @xmath96 , and model parameters are estimated by minimizing a @xmath97 statistic .
simulating maps of many different cluster models is required for model parameter fitting .
convolving the cluster model with the point source transfer function map is much faster than processing each model through the reduction pipeline .
we find that the resulting simulated maps from both methods agree sufficiently well to have negligible effect on the parameter estimation results .
we use the unsmoothed non - source - masked map with @xmath98 pixelization described in [ sec : mapmaking ] for all parameter estimation described below , since this map has lower noise and a more easily characterized transfer function than the source - masked map shown in figure [ fig : bulletcoadd ] . diffuse cluster emission
is more attenuated in the non - source - masked map , but this is taken into account using the point source transfer function .
map noise properties are assessed in the spatial frequency domain using jackknife noise maps ( see sayers 2007 ; sayers et al .
2009 ) . to estimate the noise covariance @xmath99
, we assume that the noise is stationary in the map basis . with this assumption ,
the fourier transform of the noise covariance matrix , @xmath100 , is diagonal , and the diagonal elements are equal to the noise map power spectral density ( psd ) . for each of 500 jackknife noise map realizations , we find the two - dimensional fourier transform , then average over all realizations .
this averaged map psd is the experimental estimate of the diagonal elements of @xmath100 .
however , these jackknife maps do not include fluctuations due to the primary cmb anisotropies .
we estimate the cmb signal covariance from the wmap5 best - fit power spectrum ( nolta et al . 2009 ) convolved with the point source transfer function described earlier and add it to the jackknife noise psd to determine the total covariance matrix .
we construct a @xmath97 statistic for the model fit using the transform of the filtered @xmath3 model , @xmath101 , and the transform of the central @xmath102 portion of the data map @xmath103 as : @xmath104 using markov chain monte carlo ( mcmc ) methods , the likelihood , @xmath105 , is sampled in the 7-dimensional model parameter space and integrated to find the marginal likelihood distributions of the @xmath3 model parameters .
the model parameter estimates and uncertainties that we report are the maximum likelihood values and constant - likelihood 68% confidence intervals , respectively , of the marginal likelihood distributions .
the above approach to noise covariance estimation is chosen for its simplicity and because we do not have enough linear combinations of individual scan maps to fully sample the noise covariance matrix using jackknife noise maps .
but , the method relies on several simplifying assumptions , including that the bolometer noise is stationary for each 100-s scan , the timestream noise is uncorrelated from scan to scan , and the map coverage is uniform .
our map coverage is not actually uniform , but we find through simulations of non - uniform gaussian noise maps that the @xmath97 statistic is not significantly affected .
in addition , the validity of the approach is tested by inserting simulated clusters into the real timestream data ; the simulated cluster parameters are accurately recovered within the estimated uncertainties .
results of the @xmath3 model parameter estimation are given in table [ tbl : betamodelfits ] .
due to the degeneracy between the core radius @xmath106 and @xmath3 parameters , we assume a prior probability density on @xmath3 of @xmath107 , which is found from fits to rosat x - ray data by & mitsuda ( 2004 ) . ( 2007 ) find that in hydro / n - body simulations , @xmath3 derived from fits to sze profiles is higher than that from x - ray , with @xmath108 for fits within @xmath109 . we do not account for that factor here due to the significant uncertainty in the x - ray derived @xmath3 value , but we note that our sze data prefer a higher value for @xmath3 than the peak value of the prior . we further discuss this choice of prior in [ subsec : mwte ] .
the best - fit @xmath3 model fits the data well , with a reduced @xmath97 value of 1.008 , and with 7219 degrees of freedom ( dof ) has a probability to exceed ( pte ) of 31.5% .
the difference map between the data map , @xmath96 , and the best - fit filtered @xmath3 model , @xmath95 , shows no evidence of residual cluster structure or point sources .
llll @xmath82 & ra centroid position & @xmath110 ( j2000 ) & @xmath111 + @xmath83 & dec centroid position & @xmath112 ( j2000 ) & @xmath113 + @xmath114 & central temperature decrement & @xmath115k@xmath116 & @xmath117k@xmath116 + @xmath118 & central comptonization@xmath119(@xmath92 kev ) & @xmath120 & @xmath121 + @xmath118 & central comptonization@xmath119(@xmath122 kev ) & @xmath123 & @xmath121 + @xmath124 & core radius & @xmath125 & @xmath126 + @xmath84 & ellipse minor / major core radius ratio & 0.889 & @xmath127 + @xmath128 & ellipse orientation angle & @xmath129 & @xmath130 + @xmath3 & power - law index & 1.15 & @xmath131 + radial profile plots of the best - fit @xmath3 model , @xmath94 , and the filtered @xmath3 model map , @xmath95 , are shown in figure [ fig : radial_plots ] . also plotted for comparison
are the radially binned data from the unsmoothed non - source - masked map @xmath96 , used for model fitting , and the unsmoothed source - masked map , used to visualize extended emission ( without the 1 gaussian smoothing used in figures [ fig : bulletcoadd ] & [ fig : bulletcoaddzoom ] ) .
uncertainties in both sets of radially binned data are highly correlated due to large - spatial - scale correlated noise in the maps .
the coincidence of the non - source - masked data ( @xmath96 , filled circles ) and the filtered best - fit @xmath3 model ( @xmath95 , red solid line ) show that the data and best - fit @xmath3 model are in good agreement .
the source - masked map preserves signal on larger spatial scales than the non - source - masked map , and is useful for visualizing extended emission on larger spatial scales .
but , as expected , even the source - masked map attenuates signal on scales exceeding the @xmath132 radius of the source masking , and thus has a lower signal amplitude when compared with the unfiltered @xmath3 model ( @xmath94 , blue dashed line ) . model . points with error bars are the sze data binned in 1 radial bins , from the non - source - masked map , ( @xmath96 , filled circles ) and the source - masked map ( open circles ) .
the lines show the radial profile of the best - fit @xmath3 model , unfiltered ( @xmath94 , blue dashed line ) and after convolving with the instrument beam and non - source - masked data reduction filters ( @xmath95 , red solid line ) .
the non - source - masked map radial profile is reasonably well fit by the filtered @xmath3 model .
the source - masked map preserves signal on larger spatial scales than the non - source - masked map , but still attenuates signal on scales exceeding the @xmath132 radius of the source masking .
the source - masked data thus have a lower signal amplitude when compared with the unfiltered @xmath3 model , as expected .
see text for details.,scaledwidth=45.0% ] radio sources associated with a galaxy cluster and background ir galaxy sources can have a significant impact on the measurement of the sze emission at 150 ghz .
we interpret the published results of observations of the bullet cluster at other frequencies and conclude that the measured sze decrement is not significantly contaminated .
the most important source of potential confusion is a bright , dust obscured , lensed galaxy in the direction of the bullet cluster recently reported by ( 2008 ) .
this source has a flux density of @xmath133 mjy at an observing frequency of 270 ghz , and is centered at ra @xmath134 , dec @xmath135 ( j2000 ) , @xmath136 to the east of the measured sze centroid position , see figure [ fig : bulletcoaddzoom ] . assuming a spectral index @xmath137 , where @xmath138 , we expect a flux density of @xmath139mjy at @xmath16ghz corresponding to a temperature increment of @xmath140 in the 1.5 arcmin@xmath22 apex - sz beam solid angle .
this lensed source is expected to be the dominant contribution to positive flux in the direction of the cluster and we have repeated the @xmath3 model fit taking it into account .
we first add the source at its measured position with the predicted 150 ghz flux to the sze @xmath3 model and repeat the model fit . as expected , including the point source results in a slightly ( @xmath141 ) deeper decrement ; however , the @xmath97 of the model fit slightly increases .
we next allow the flux of the point source to vary along with the other model parameters and find that values of positive flux are a poorer fit than no source at all .
therefore , we have no evidence for significant emission from this source at 150 ghz . for the results in this paper , we use cluster model parameters derived from fits that do not include this ir source .
the bullet cluster is also associated with a number of relatively compact radio sources and one of the brightest cluster radio halos yet discovered . however , these sources are predicted to produce negligible temperature increments in the apex - sz beam when extrapolated to 150 ghz .
\(2000 ) report the detection of eight radio point sources all of which have steeply falling spectra .
only two of these sources were detected with acta at 8.8 ghz , and they were found to have flux densities of @xmath142 mjy and @xmath143 mjy .
the spectra of these sources , measured between @xmath144 and @xmath145ghz are falling with @xmath146 and @xmath147 , respectively . extrapolating to @xmath16ghz , the flux of these sources
are expected to be @xmath148 and @xmath149mjy , corresponding to cmb temperature increments of @xmath150 and @xmath151 in the 1.5 arcmin@xmath22 apex - sz beam solid angle .
the radio halo in the bullet cluster is very luminous , but has a characteristically steeply falling spectrum .
( 2000 ) measure the flux and spectra for the two main spatial components of the halo . at @xmath145ghz
, they find the two components to have fluxes of @xmath152 and @xmath153mjy , with spectral indices of @xmath154 and @xmath155 , respectively .
extrapolating to @xmath16ghz , the combined flux from the radio halo is expected to be @xmath156mjy .
this emission is spread over an area comparable to the size of the cluster and therefore corresponds to a temperature increment @xmath157 in the apex - sz beam .
x - ray emission in the ionized intracluster gas is dominated by thermal bremsstrahlung .
the x - ray surface brightness can be written @xmath158 where @xmath159 are the electron and ion densities in this gas , @xmath160 is the x - ray cooling function , and the integral is taken along the line of sight .
the x - ray flux is proportional to the line - of - sight integral of the square of the electron density , resulting in emission that is more sensitive to local density concentrations than the sze emission .
the bullet sub - cluster and the bow shock are apparent in the x - ray surface brightness map shown in figure [ fig : bullet - overlay ] .
the sze contour map of the bullet cluster in figures [ fig : bulletcoadd ] & [ fig : bulletcoaddzoom ] is overlaid on an x - ray map and weak lensing surface mass density reconstruction from ( 2006 ) .
the x - ray map is made from xmm data ( observation i d : 0112980201 ) extracted in the [ 0.5 - 2 ] kev band , corresponding to bullet rest frame energies where the x - ray cooling function for hot gas is relatively insensitive to temperature .
the map is smoothed with a 12gaussian kernel .
the sze contours do not resolve the sub - cluster .
however , an elongation of the inner contours to the west suggests that a contribution from it may be detected .
the observed sze map is consistent with expectations , given the 85 resolution of the sze map , the different dependence of the x - ray and sze signals on gas density , and the mass and temperature difference between the two merging components which predict a factor of @xmath161 lower integrated pressure from the sub - cluster .
there is no evidence in the sze contours of a contribution from the lensed sub - mm bright galaxy discussed in
[ sec : source_contributions ] .
.,scaledwidth=45.0% ] the combination of cluster sze and x - ray measurements can be used to place constraints on the thermal structure of the intracluster gas .
the sze intensity is proportional to the product of the electron density and the electron temperature along the line of sight , see equation ( [ eqn : integratedpressure ] ) .
therefore , if the electron density is known from another measurement , the sze can be used to measure a mass - weighted temperature . for simplicity , we assume here that the intracluster gas is isothermal , but a more detailed comparison of the sze surface brightness and projected density could be used to constrain the thermal structure in the cluster .
we perform this calculation with two different descriptions of the intracluster gas density . first , we model the spatial distribution of the intracluster gas as a spherical @xmath3 model following equation ( [ eqn : betamodeldensity ] ) .
we use @xmath3 model parameters from & mitsuda ( 2004 ) , derived from rosat hri ( @xmath162 resolution ) measurements of the inner @xmath61 radius of the bullet cluster : @xmath4 , @xmath163 , and @xmath164 @xmath165 .
we construct an x - ray derived sze surface brightness model from the @xmath3 model electron surface density profile using equation ( [ eqn : integratedpressure ] ) . to account for @xmath3 model uncertainties ,
we incorporate the values and uncertainties for @xmath3 , @xmath166 , and @xmath72 as independent gaussian priors .
we then use the analysis method described in [ subsec : fitting ] to minimize @xmath97 on the difference between the x - ray derived sze model , convolved with the point source transfer function , and the apex - sz non - source - masked data .
the free parameters in the fit are the three @xmath3 model parameters , the mass - weighted electron temperature @xmath167 , and the relative map alignment in ra and dec .
we find @xmath168kev after marginalizing over the other parameters in the fit and including the sze flux calibration uncertainty and the effect of the apex - sz band center frequency uncertainty on the relativistic sze spectrum @xmath169 .
the reduced @xmath97 of the best fit model is @xmath170 with an associated pte of @xmath171 , indicating that the spherical @xmath3 density model and the assumption of isothermality produce an acceptable fit to the data . given the complex morphology of this merging system , the validity of the spherical @xmath3 model is questionable .
we therefore repeat the determination of the mass - weighted temperature by directly comparing x - ray measurements of the projected intracluster gas density with the measured sze signal in order to produce a less model - dependent measurement of the mass - weighted temperature .
we make use of the publicly available electron surface density map derived from chandra x - ray satellite data presented in ( 2006 ) . using the same analysis as above , and marginalizing over the relative map alignment parameters , we find mass - weighted electron temperature @xmath172kev .
the fit to the data is again good , with a reduced @xmath173 and a pte of 20.6% .
this is in excellent agreement with the value @xmath174kev found from the above @xmath3 model analysis . given the complex dynamics in the bullet cluster , there have been several studies of the temperature structure ( e.g. , finoguenov , bhringer , & zhang 2005 ; markevitch 2006 ; andersson , peterson , & madejski 2007 ) .
there have also been several published results for the spectroscopic temperature within annuli about the cluster center .
chandra data was used by ( 2004 ) to determine a spectroscopic x - ray temperature of @xmath175kev within @xmath176mpc of the cluster center . from the analysis of xmm data within an annulus of @xmath177mpc radius , ( 2006 )
find a temperature of @xmath178kev .
analyzing the combination of xmm and rxte data , , madejski , & luli ( 2006 ) , find @xmath179kev within a radius of @xmath180mpc .
the published x - ray spectroscopic temperatures span a range much larger than the stated uncertainties in the measurements . given the complex thermal structure for the cluster , and the presence of gas at temperatures corresponding to energies at or above the upper limits of the chandra and xmm energy response , the variation in the measured x - ray spectroscopic temperature is not surprising .
the mass - weighted temperature found with apex - sz falls near the lowest of the reported x - ray spectroscopic temperatures .
however , we do not expect exact agreement between the mass - weighted and spectroscopic temperatures .
using chandra data for a sample of 13 relaxed clusters , ( 2006 ) find that , due to the presence of thermal structure in the intracluster gas , the x - ray spectroscopic temperature is typically a factor of @xmath181 larger than the x - ray derived mass - weighted electron temperature .
this is consistent with the simulation results of , vikhlinin , & kravtsov ( 2007 ) who find @xmath182 for relaxed clusters and @xmath183 , with a somewhat larger scatter , for unrelaxed systems . naively applying this correction to the published x - ray spectroscopic temperatures , we infer results for mass - weighted temperatures that bracket the apex - sz result . using the sze measurements , we construct a model for the intracluster gas distribution which
, when combined with x - ray measurements , can be used to determine the gas mass , total mass , and therefore gas mass fraction of the cluster .
the gas mass is estimated by integrating a spheroidal model for the cluster gas , following ( 2006 ) .
several assumptions about the model must be made to estimate the gas mass .
we assume that the cluster gas is isothermal in order to convert pressure to density .
we also assume spheroidal symmetry for the gas distribution in order to convert the two - dimensional sze integrated pressure measurement to a three - dimensional gas distribution .
we consider two simple cases , an oblate spheroid generated by rotation about the minor axis and a prolate spheroid generated by rotation about the major axis , where the symmetry axis is in the sky plane . the gas mass , under these assumptions , becomes : @xmath184 where @xmath185 is the nucleon / electron ratio , taken to be 1.16 ( grego et al . 2001 ) .
the factor of eight is due to integrating over only one octant of the spheroid .
the factor @xmath186 is set to unity in the case of oblate spheroidal symmetry , while in the case of prolate spheroidal symmetry , @xmath186 is set to @xmath84 .
the total cluster mass is estimated by assuming hydrostatic equilibrium and integrating the inferred gas distribution ( grego et al .
2000 ) to find : @xmath187 here @xmath188 is the mean molecular weight of the intracluster gas , which is assumed to be 0.62 ( zhang et al .
2006 ) . using equations ( [ eqn : gasmass ] ) & ( [ eqn : hydroeq ] ) , and our model parameters in table [ tbl : betamodelfits ] , we calculate the gas mass , total mass , and gas mass fraction for the cluster . in table
[ tbl : massestimates ] , we give these results .
the gas mass fraction results for a prolate gas distribution model are @xmath189 larger than those for an oblate model , while the total mass and gas mass are @xmath190 larger .
we quote only the oblate spheroidal results .
we calculate our results within two different radii .
the first is the radius of the cluster at which its mean density is equal to 2500 times the critical density at the redshift of the cluster , @xmath191 .
the second radius is 1.42 mpc , which is the same radius used by ( 2006 ) for their gas mass fraction calculation .
this will allow for a more direct comparison to their result , and is also near where our measured sze radial profile has unity signal to noise .
for all results , we assume a @xmath17cdm cosmology , with @xmath18 , @xmath19 , and @xmath192 . the results of this analysis are summarized in table [ tbl : massestimates ] . under the assumption of a @xmath193kev mass - weighted temperature ( the lowest of the published x - ray spectroscopic temperatures and near our mass - weighted temperature results in [ subsec : mwte ] )
, we find gas mass fractions @xmath194 and @xmath195 within @xmath191 and 1.42 mpc , respectively . the fact that the computed gas fraction in the central region significantly exceeds the cosmic average determined by wmap5 ( @xmath196 , dunkley et al .
2009 ) , and a lower value observed in relaxed clusters ( @xmath197 , see , e.g. , vikhlinin et al . 2009 ) , is likely due to deviations of the intracluster gas from isothermal hydrostatic equilibrium . on larger scales , baryon fractions produced for the range of reported x - ray temperatures bracket the published results using x - ray and weak lensing data .
( 2006 ) measure a gas mass fraction @xmath198 by comparing the gas mass calculated from chandra x - ray measurements to weak lensing total mass measurements in a @xmath199 box roughly centered around the cluster .
( 2006 ) , measured a gas mass fraction @xmath200 within a radius of 1.42 mpc . despite the limitations of applying a hydrostatic equilibrium model to this merging cluster , the apex - sz results for the gas mass fraction are in good agreement with previous work .
ccccccc @xmath201 & @xmath202 & 2.77 & 0.739 & @xmath203 & @xmath204 & @xmath205 + @xmath201 & @xmath206 & 5.32 & 1.42 & @xmath207 & @xmath208 & @xmath209 + @xmath210 & @xmath211 & 2.15 & 0.572 & @xmath212 & @xmath213 & @xmath214 + @xmath210 & @xmath215 & 5.32 & 1.42 & @xmath195 & @xmath216 & @xmath217 +
measurements of the sze provide a robust and independent probe of the intracluster gas properties in galaxy clusters .
the apex - sz @xmath16ghz observations detect the bullet system with @xmath2 significance within the central 1 radius of the sze centroid position .
we do not expect to see a resolved signal from the bullet sub - cluster in the @xmath16ghz 85 fwhm resolution sze maps , and no obvious feature , such as a secondary peak , is present .
we expect no significant contamination of the observed sze decrement due to radio sources , and there is no evidence for significant contamination by a known bright lensed dusty galaxy .
we process an elliptical @xmath3 model through the observation transfer function and fit it to the measured temperature decrement map .
we also measure the cluster mass - weighted electron temperature and gas mass fraction with the sze data . combining the apex - sz map with a map of projected electron surface density from chandra x - ray observations ,
we determine the mass - weighted temperature of the cluster gas to be @xmath9kev .
this value is consistent with the lowest x - ray spectroscopic temperatures reported for this cluster and should be less sensitive to the details of the cluster thermal structure .
the derived baryon fraction is also found to be in reasonable agreement with previous x - ray and weak lensing determinations . throughout this work
, we make an assumption of isothermal cluster gas .
clearly , incorporating thermal structure , measured by x - ray observations , in the analysis of the sze data would improve the determination of the gas distribution and gas mass fraction .
ultimately , a more sophisticated analysis could be implemented that combines x - ray , sze , and weak lensing data and relaxes assumptions of hydrostatic equilibrium between the gas and dark matter components of the cluster .
this is particularly important for a detailed understanding of actively merging systems such as the bullet cluster .
we thank the staff at the apex telescope site , led by david rabanus and previously by lars - ke nyman , for their dedicated and exceptional support .
we also thank lbnl engineers john joseph and chinh vu for their work on the readout electronics .
apex - sz is funded by the national science foundation under grant nos .
ast-0138348 & ast-0709497 .
work at lbnl is supported by the director , office of science , office of high energy and nuclear physics , of the u.s .
department of energy under contract no .
de - ac02 - 05ch11231 .
work at mcgill is supported by the natural sciences and engineering research council of canada and the canadian institute for advanced research .
rk acknowledges partial financial support from mpg berkeley - munich fund .
nwh acknowledges support from an alfred p. sloan research fellowship . | we present observations of the sunyaev - zeldovich effect ( sze ) in the bullet cluster ( 1e 065756 ) using the apex - sz instrument at 150@xmath0ghz with a resolution of @xmath1 .
the main results are maps of the sze in this massive , merging galaxy cluster .
the cluster is detected with @xmath2 significance within the central 1 radius of the source position .
the sze map has a broadly similar morphology to that in existing x - ray maps of this system , and we find no evidence for significant contamination of the sze emission by radio or ir sources . in order to make simple quantitative comparisons with cluster gas models derived from x - ray observations , we fit our data to an isothermal elliptical @xmath3 model , despite the inadequacy of such a model for this complex merging system . with
an x - ray derived prior on the power - law index , @xmath4 , we find a core radius @xmath5 , an axial ratio of @xmath6 , and a central temperature decrement of @xmath7 , including a @xmath8 flux calibration uncertainty . combining the apex - sz map with a map of projected electron surface density from chandra x - ray observations , we determine the mass - weighted temperature of the cluster gas to be @xmath9kev , significantly lower than some previously reported x - ray spectroscopic temperatures . under the assumption of an isothermal cluster gas distribution in hydrostatic equilibrium ,
we compute the gas mass fraction for prolate and oblate spheroidal geometries and find it to be consistent with previous results from x - ray and weak lensing observations .
this work is the first result from the apex - sz experiment , and represents the first reported scientific result from observations with a large array of multiplexed superconducting transition - edge sensor bolometers . |
let @xmath20 be a compact connected surface and @xmath21 be either the real line @xmath22 or the circle @xmath3 .
for a closed subset @xmath23 denote by @xmath24 the group of all @xmath18 diffeomorphisms of @xmath20 fixed on @xmath25 .
this group acts from the right on the space @xmath26 by the following rule : if @xmath27 and @xmath28 , then the result of the action of @xmath29 on @xmath30 is the composition map @xmath31 . for @xmath28 let @xmath32 be respectively the _ stabilizer _ and the _ orbit _ of @xmath30 under that action .
endow these spaces with @xmath18 topologies and denote by @xmath33 and @xmath34 the corresponding path components of @xmath35 in @xmath24 and @xmath36 , and by @xmath37 the path component of @xmath38 containing @xmath30 .
we will omit @xmath25 from notation whenever it is empty .
[ def : morsemap ] a smooth map @xmath39 will be called _ morse _ if * all critical points of @xmath30 are non - degenerate and belong to the interior of @xmath20 ; * the restriction of @xmath30 to each connected component of @xmath40 is a constant map . in a series of papers
@xcite the author studied the homotopy types of @xmath41 and @xmath42 for certain classes of smooth mappings @xmath39 .
the results concerning morse maps can be summarized as follows .
[ th : fibration_dmx_of ] @xcite .
let @xmath39 be a morse map with @xmath43 critical points , and @xmath25 be a finite ( possibly empty ) union of regular ( that is containing no critical points ) connected components of certain level sets of @xmath30 .
then the following statements hold .
\(1 ) @xmath44 and this space * has a homotopy type of some ( possibly non - compact ) @xmath45-dimensional cw - complex*. \(2 ) the map @xmath46 defined by @xmath47 is a serre fibration with fiber @xmath36 , i.e. it has a homotopy lifting property for cw - complexes .
suppose either @xmath30 has a critical point of index @xmath48 or @xmath20 is non - orientable .
then @xmath41 is contractible , @xmath49 for @xmath50 , @xmath51 , and for @xmath52 we have the following short exact sequence of fibration @xmath53 : @xmath54 where @xmath55 .
if @xmath30 is * generic * , i.e. each critical level set of @xmath30 contains exactly one critical point , then @xmath42 is * homotopy equivalent to some @xmath56-dimensional torus @xmath57*. \(3 ) the restriction @xmath58 is also a serre fibration with fiber @xmath59 .
suppose either @xmath60 or @xmath61 .
then @xmath33 and @xmath34 are contractible and from the exact sequence of homotopy groups of the fibration @xmath62 we get that @xmath63 for @xmath64 and the boundary map @xmath65 is an isomorphism . in the present note
we obtain exact description of the homotopy types of @xmath42 for all morse maps @xmath39 for the case when @xmath20 is orientable and distinct from @xmath66 and @xmath67 , see theorem [ th : main_result_of_tpg ] below .
let @xmath39 be a morse map .
consider the partition of @xmath20 into connected components of level - sets @xmath68 , @xmath69 .
then the corresponding factor space has a structure of a finite one - dimensional cw - complex and is called the _ kronrod - reeb _ graph of @xmath30 , see e.g. @xcite .
we will denote this graph by @xmath70 .
its vertices correspond to _ critical _
( i.e. containing critical points ) connected components of level - sets of @xmath30 and to the connected components of @xmath40 .
recall , @xcite , that @xmath71 naturally acts on @xmath70 by the following rule .
let @xmath72 .
then @xmath73 , and so @xmath74 for all @xmath69 .
hence @xmath29 permutes connected components of @xmath68 being _ points _ of @xmath70 . in other words
, @xmath29 yields a bijection of @xmath70 which is in fact homeomorphism .
therefore we get an action homomorphism @xmath75 into the group of all homeomorphisms of @xmath70 .
the image of @xmath76 is finite .
let @xmath77 be the group of all automorphisms of @xmath70 induced by isotopic to the identity diffeomorphisms from @xmath71 .
our main result is the following theorem .
[ th : main_result_of_tpg ] suppose @xmath20 is orientable and distinct from @xmath13-sphere @xmath66 and @xmath13-torus @xmath67 .
then for each morse map @xmath39 there exists a free action of the group @xmath78 on a @xmath56-dimensional torus @xmath57 for some @xmath79 such that @xmath42 is homotopy equivalent to the factor space @xmath80 .
this result refines statements ( 1 ) and ( 2 ) of theorem [ th : fibration_dmx_of ] .
a sketch of a proof of theorem [ th : main_result_of_tpg ] will be given in
[ sect : proof : th : main_result_of_tpg ] and [ sect : proof : prop : lambda_is_t_hom ] .
the proof is constructive in the sense that for every morse map @xmath39 one can explicitly describe the corresponding free action of @xmath81 on @xmath57 which gives a homotopy equivalence @xmath82 .
if @xmath83 , then by ( 2 ) of theorem [ th : fibration_dmx_of ] @xmath42 is not aspherical , and so it is not homotopy equivalent to @xmath80 for any free @xmath81-action on any torus @xmath57 .
on the other hand there are some partial results supporting the conjecture that theorem [ th : main_result_of_tpg ] might hold for @xmath84 , see @xcite .
in fact , theorem [ th : main_result_of_tpg ] is true for a larger class of maps @xmath30 having singularities smoothly equivalent to homogeneous polynomials @xmath85 without multiple factors .
but in this case one should extend the notion of kronrod - reeb graph for such maps .
this version of theorem [ th : main_result_of_tpg ] together with detailed proofs will appear elsewhere .
let @xmath81 and @xmath86 be two groups
. then the set @xmath87 of all maps @xmath88 ( not necessarily homomorphisms ) is a group with respect to the point - wise multiplication .
moreover , the group @xmath86 acts on @xmath87 from the right by the following rule : if @xmath89 , and @xmath90 , then the result @xmath91 of the action of @xmath29 on @xmath92 is defined by the formula : @xmath93 , for all @xmath94 .
the semidirect product @xmath95 associated with this action is called the _ wreath product _ of @xmath81 and @xmath86 and is denoted by @xmath96 .
more generally , let @xmath97 be another group and @xmath98 be a homomorphism .
this homomorphism induces a natural right action of @xmath97 on @xmath86 , and so one can define the corresponding semidirect product @xmath99 which will be called the _ wreath product of @xmath81 and @xmath97 over @xmath100 _ and denoted by @xmath101 .
evidently , if @xmath102 is the identity isomorphism , then @xmath103 is the same as @xmath96 . in particular , for @xmath104 let @xmath105 be the natural @xmath106 homomorphism
. then for every group @xmath107 the wreath product @xmath108 will be denoted by @xmath109 thus @xmath110 is the set @xmath111 with the multiplication defined in the following way : if @xmath112 , then @xmath113 , where @xmath114 is given by @xmath115 for @xmath116 . for @xmath79 let @xmath117 be a @xmath56-dimensional torus which can also be regarded as a topological product of @xmath56 circles .
we also assume that @xmath118 is a point .
suppose a finite group @xmath12 freely acts on @xmath57 .
then the factor map @xmath119 is a covering map .
this implies that @xmath120 is aspherical and we have the following short exact sequence : @xmath121 where @xmath122 is the corresponding boundary homomorphism .
[ defn : tprop ] let @xmath123 be a group homomorphism .
we will say that @xmath76 _ arises from a free action on a torus _ if there exists a free action of @xmath12 on @xmath57 for some @xmath79 and the following commutative diagram : @xmath124_{\delta } \ar[rr]^-{\alpha}_-{\cong } & & { \mathcal{s}}\ar[d]^{\lambda}\\ g \ar[rr]^-{\beta}_-{\cong } & & g } \ ] ] in which @xmath92 and @xmath125 are isomorphisms .
[ lm : examples_t_prop ] ( 1 ) a trivial homomorphism @xmath126 , @xmath127 , arises from the free trivial action of the unit group @xmath128 on @xmath129 .
( 2 ) for each @xmath130 the canonical @xmath131-epimorphism @xmath105 arises from the free action of @xmath132 on @xmath133 defined as follows : if @xmath134 and @xmath135 , then @xmath136 .
( 3 ) suppose a homomorphism @xmath137 arises from a free action of @xmath138 on @xmath139 for @xmath140 .
denote @xmath141 then the product homomorphism @xmath142 arises from the free action of @xmath143 on @xmath144 defined as follows : if @xmath145 and @xmath146 , then @xmath147 .
( 4 ) suppose @xmath123 arises from a free action of @xmath12 on @xmath57 and let @xmath127
. then the homomorphism @xmath148 defined by @xmath149 arises from the free @xmath12-action on @xmath150 defined as follows : if @xmath151 , @xmath152 , and @xmath153 , then @xmath154 .
( 5 ) suppose @xmath123 arises from a free action of @xmath12 on @xmath57 and let @xmath104 .
define the following homomorphism @xmath155 by @xmath156 for @xmath157 .
then @xmath158 arises from the free action of @xmath159 on @xmath160 defined as follows : if @xmath161 , @xmath162 , @xmath163 , and @xmath164 , then @xmath165 where all indices are taken modulo @xmath166 .
statements ( 1)-(4 ) are easy .
the proof of ( 5 ) is technical and will be published elsewhere .
suppose @xmath20 is orientable and distinct from @xmath66 and @xmath67 .
let also @xmath39 be a morse map , @xmath167 be the action homomorphism , @xmath168 , and @xmath25 be any ( possibly empty ) collection of connected components of @xmath40 .
notice that each @xmath169 preserves every connected component of @xmath40 with its orientation , and
so it is isotopic in @xmath170 to a diffeomorphism @xmath171 fixed on @xmath40 , see e.g. ( * ? ? ?
* lemma 6.1 ) .
in particular , we obtain that @xmath172 , whence @xmath173 moreover , it is easy to see that for each @xmath174 its image @xmath175 depends only on the isotopy class of @xmath176 $ ] in @xmath177 , and so @xmath76 induces the following epimorphism @xmath178 [ prop : lambda_is_t_hom ] there exists a collection @xmath25 of connected components of @xmath40 satisfying the following conditions : * the boundary map @xmath179 is an isomorphism ; * the homomorphism arises from a free @xmath81-action on some torus @xmath57 .
it follows from ( b ) that we have an isomorphism @xmath180 , which together with ( 1 ) of theorem [ th : fibration_dmx_of ] and ( a ) gives another isomorphism : @xmath181 as both @xmath80 and @xmath42 are aspherical , there exists a homotopy equivalence between these spaces such that the corresponding isomorphism of fundamental groups coincides with @xmath182 .
this proves theorem [ th : main_result_of_tpg ] modulo proposition [ prop : lambda_is_t_hom ] .
let @xmath25 be a connected component of @xmath40 and @xmath185 . then by the boundary map @xmath186 is an isomorphism .
therefore it remains to show that the homomorphism @xmath187 arises from a certain free @xmath81-action on a torus @xmath57 for some @xmath79 .
\1 ) suppose @xmath191 .
this is possible only when @xmath192 , see figure [ fig : function_n01 ] .
we can also assume that @xmath193 and each of the circles @xmath194 , @xmath195 , is a connected component of some level set of @xmath30 .
in this case @xmath70 consists of a unique edge , and therefore @xmath196 .
it follows from the proof of case 2 of lemma 6.3 from @xcite that @xmath197 .
then by ( 1 ) of lemma [ lm : examples_t_prop ] @xmath198 arises from a trivial action of @xmath81 on @xmath118 .
\2 ) suppose @xmath199 . in this case
@xmath183 , @xmath200 , a unique critical point @xmath201 of @xmath30 is a local extreme , and @xmath70 again consists of a unique edge , so @xmath202 , see figure [ fig : function_n01 ] . by morse lemma
we have that @xmath203 in some local representation of @xmath30 at @xmath201 , therefore not loosing generality one can assume that @xmath204 is a unit @xmath13-disk in the complex plane , and the level sets of @xmath30 are the origin @xmath205 and concentric circles @xmath206 , @xmath207 $ ] .
it follows from the proof of case 3 of lemma 6.3 from @xcite that @xmath197 . hence again by ( 1 ) of lemma [ lm : examples_t_prop ] @xmath198 arises from a trivial action of @xmath81 on @xmath118 .
\3 ) now let @xmath208 .
assume that we proved our statement for all morse maps @xmath209 and @xmath210 having less than @xmath43 critical points , and let @xmath39 be a morse map with exactly @xmath43 critical points .
notice that the kronrod - reeb graph @xmath70 of @xmath30 is a finite tree .
let @xmath211 be a vertex @xmath70 of @xmath30 corresponding to @xmath25
. then there exists a unique edge @xmath212 in @xmath70 incident to @xmath211 .
consider another vertex @xmath213 of @xmath214 and let @xmath215 be the connected component of the level set of @xmath30 corresponding to @xmath213 , see figure [ fig : function_n2 ] .
for @xmath216 let @xmath217 be a closed @xmath218-neighbourhood of the point @xmath219 in @xmath21 and @xmath220 be the connected component of @xmath221 containing @xmath215 .
we can choose @xmath218 so small that @xmath222 and the set @xmath223 contains no critical points of @xmath30 .
let @xmath224 be a connected component of @xmath225 containing @xmath25 .
then @xmath224 is a cylinder having no critical points of @xmath30 .
let @xmath226 be a dehn twist preserving @xmath30 and supported in the interior of @xmath224 , see @xcite , and @xmath227 be its isotopy class .
evidently , @xmath228 . *
. then @xmath130 .
* if @xmath231 is such that @xmath232 is divided by @xmath166 , then the class @xmath233 contains a representative fixed on some neighbourhood of @xmath220 . * if @xmath234 , then @xmath233 contains a representative fixed on some neighbourhood of @xmath235 . for each connected component
@xmath236 of @xmath225 put @xmath237 .
then the pair @xmath238 is diffeomorphic either to @xmath239 or to @xmath240 .
as @xmath30 is constant on connected components of @xmath241 , the restriction @xmath242 is morse in the sense of definition [ def : morsemap ] .
let @xmath243 be the kronrod - reeb graph of @xmath244 .
then @xmath243 can be regarded as a subtree of @xmath70 .
denote @xmath245 .
let also @xmath246 be the corresponding action homomorphism , and @xmath247 .
3a ) suppose @xmath249 .
then by ( b ) of lemma [ lm : epimorphism_eta ] every @xmath231 has a representative fixed near @xmath220 .
hence if we denote by @xmath250 all the connected components of @xmath225 distinct from @xmath224 , then @xmath251 for all @xmath252 and @xmath253 .
[ lm : m_1_iso_pi0s ] @xcite there is the following commutative diagram : @xmath254 ^ -{\alpha}_-{\cong } \ar[d]_-{\bar{\lambda } } & & \pi_0{\mathcal{s}}\ar[d]^-{\lambda } \\
\mathop{\times}\limits_{i=1}^{a } { g_{{x}_{i } } } \ar[rr]^-{\beta}_-{\cong } & & { g}}\ ] ] in which @xmath92 and @xmath125 are isomorphisms , and @xmath158 is defined by @xmath255 for @xmath256 and @xmath257 .
hence by ( 3 ) and ( 4 ) of lemma [ lm : examples_t_prop ] @xmath76 arises from a free action on a torus .
let @xmath258 , @xmath252 , and @xmath257 .
choose a representative @xmath259 of @xmath260 such that @xmath261 is fixed on some neighbourhood of @xmath262 , and so it extends by the identity to a diffeomorphism of @xmath20 .
let @xmath231 be the isotopy class of the composition @xmath263 .
put @xmath264 .
it is shown in @xcite that @xmath92 is a group isomorphism .
let @xmath265 , @xmath252 .
notice that @xmath266 is a subtree of @xmath70 and that @xmath267 extends by the identity to an automorphism of all of @xmath70 .
let @xmath268 .
then it is easy to show that @xmath269 , i.e. it is induced by some @xmath253 , so we set @xmath270 .
it is proven in @xcite that @xmath125 is a group isomorphism .
it also easily follows from the definitions that the diagram is commutative .
3b ) suppose @xmath104 .
let @xmath250 be all the connected components of @xmath225 distinct from @xmath224 and _ invariant _ with respect to all @xmath253 , and @xmath271 be the collection of all connected components of @xmath225 being _ not invariant _ under @xmath107 . moreover , there is the following commutative diagram : @xmath282 ^ -{\alpha}_-{\cong } \ar[d]_-{\bar{\lambda } } & & \pi_0{\mathcal{s}}\ar[d]^-{\lambda } \\
\bigl(\mathop{\times}\limits_{i=1}^{a } { g_{{x}_{i}}}\bigr ) \
\times \ \bigl(\bigl(\mathop{\times}\limits_{j=1}^{b}{g_{{y}_{j}}}\bigr ) \
\wr\ { { \mathbb z}}_m\bigr ) \ar[rr]^-{\beta}_-{\cong } & & { g}}\ ] ] in which @xmath92 and @xmath125 are isomorphisms and @xmath158 is defined by @xmath283 for @xmath284 , a map @xmath285 , and @xmath257 .
hence by ( 3 ) and ( 5 ) of lemma [ lm : examples_t_prop ] @xmath76 arises from a free action on a torus .
let @xmath258 , @xmath252 , @xmath286 be any map , and @xmath257 .
choose a representative @xmath259 of @xmath260 fixed on some neighbourhood of @xmath262 .
let also @xmath287 , @xmath288 , @xmath289 , be a representative of @xmath290 fixed on some neighbourhood of @xmath291 , and @xmath292 be a map given by the formula : @xmath293 then @xmath253 and we define @xmath294 to be the isotopy class of @xmath29 .
it is shows in @xcite that @xmath92 is a group isomorphism .
let @xmath265 , @xmath252 , @xmath295 be map , and @xmath296 .
define @xmath297 by the following formula : @xmath298 then @xmath269 and the correspondence @xmath299 is a group isomorphism , see @xcite .
it also easily follows from the definitions that the diagram is commutative .
suppose @xmath20 is orientable and distinct from @xmath66 , @xmath67 , @xmath300 , and @xmath184 .
set @xmath301 and denote @xmath185 .
then @xmath60 whence by the boundary map @xmath302 is an isomorphism .
\(b ) let @xmath306 be the kronrod - reeb graph of @xmath307 , @xmath308 , @xmath309 be the corresponding action homomorphism , and @xmath310 , @xmath140 .
then we have the following commutative diagram : @xmath311
^ -{\alpha}_-{\cong } \ar[d]^{\mathop{\times}\limits_{i=1}^{n}\lambda_i } & & \pi_0{\mathcal{s}}\ar[d]_{\lambda } \\
\mathop{\times}\limits_{i=1}^{n } { g_{i } } \ar[rr]^-{\beta}_-{\cong } & & { g}}\ ] ] in which @xmath92 and @xmath125 are isomorphisms .
surfaces @xmath312 and the isomorphism @xmath92 are defined in ( * ? ? ?
* theorem 1.7 ) .
the isomorphism @xmath125 and diagram are constructed in @xcite .
in fact @xmath92 and @xmath125 are similar to the corresponding isomorphisms from lemma [ lm : m_1_iso_pi0s ] and therefore we skip their descriptions .
since each @xmath304 is either a @xmath13-disk @xmath300 or a cylinder @xmath184 , the homomorphism @xmath313 arises from a free action of @xmath314 on some torus .
then by ( 3 ) of lemma [ lm : examples_t_prop ] so does @xmath187 .
proposition [ prop : lambda_is_t_hom ] is completed . | let @xmath0 be a connected orientable surface , @xmath1 be either the real line @xmath2 or the circle @xmath3 , and @xmath4 be a morse map
. denote by @xmath5 the group of diffeomorphisms of @xmath0 isotopic to the identity .
this group acts from the right on the space of smooth maps @xmath6 and one can define the stabilizer @xmath7 and the orbit @xmath8 of @xmath9 with respect to that action .
in turn , the stabilizer @xmath10 acts on the kronrod - reeb graph @xmath11 of @xmath9 .
denote by @xmath12 the group of all automorphisms of @xmath11 induced by elements from @xmath10 .
it is proved that if @xmath0 is distinct from the @xmath13-sphere and @xmath13-torus then there exists a free action of @xmath12 on a @xmath14-dimensional torus @xmath15 for some @xmath16 such that the orbit @xmath17 ( endowed with @xmath18 topology ) is homotopy equivalent to the factor space @xmath19 .
[ multiblock footnote omitted ] |
stellar archaeology of the most metal - poor stars observed in the halo of our galaxy provides valuable insights on the star formation and chemical enrichment histories at high redshifts .
the observed surface elemental abundances and the metallicity distribution function of galactic halo stars shed light on the nature and properties of the first supernovae ( pop iii sne ) that pollute the interstellar medium ( ism ) with the first metals and dust grains .
galactic halo metal - poor stars have a well defined abundance pattern at @xmath0 < -2.5 $ ] @xcitethat can be reasonably well reproduced by theoretical yields of pop iii core - collapse sne with progenitor masses in the range @xmath2 @xcite .
an exception are the so - called carbon - enhanced extremely metal - poor stars ( cemp , defined as those with @xmath3 > + 1$],@xcite , which comprise a fraction of @xmath4 of metal - poor stars at @xmath0 < -2 $ ] @xcite ) .
interestingly , the observed fraction of c - rich metal - poor stars increases with decreasing iron abundance and , at @xmath0 < -4.5 $ ] , 4 out of the 5 stars currently known are cemp , including the recently discovered smss j031300.36 - 670839.3 ( hereafter smss j031300 ) with [ fe / h ] @xmath5 @xcite .
a number of theoretical scenarios have been proposed to explain the origin of cemp stars , invoking mass transfer in pop iii binary systems @xcite , formation from material enriched in stellar winds of rotating massive and intermediate mass pop iii stars @xcite , or in the ejecta of pop iii sne in which mixing was minimal and fallback was large . in the latter model ,
the small iron abundance thus reflects the fact that in these so - called `` faint '' sne iron was made , but failed to be mixed sufficiently far out to be ejected @xcite .
alternatively , the large c / fe could result from multiple generations of sne , where the ejecta of a `` normal '' sn is combined with that from a low - energy sn in which only the outer layers are ejected @xcite , or from a jet - induced sn explosion @xcite .
the understanding of the origin of cemp stars is complicated by the fact that these do not form a homogeneous group and show various heavy - element abundance patterns : depending on the level of enrichment of neutron capture elements formed by the `` s '' or `` r '' processes , these fall in different sub - classes ( cemp - no , cemp - rs , cemp - s,@xcite ) .
generally , cemp - s and cemp - rs stars have abundance patterns that suggest mass transfer from companion agb stars @xcite ; indeed , most of the stars falling in these sub - classes are members of binary systems @xcite . however , at @xmath0 < -3 $ ] , 90% of cemp are cemp - no stars , which do not show large enhancements of heavy neutron capture elements @xcite . using a homogeneous chemical analysis of 190 metal - poor stars @xcite .
norris et al .
( 2013 ) have recently investigated a sample of cemp - no stars with @xmath0 < - 3 $ ] .
they conclude that the observed chemical abundances are best explained by models which invoke the nucleosynthesis of pop iii sne with mixing and fallback , of rotating massive and intermediate mass stars and of sne with relativistic jets .
similar conclusions have been drawn by keller et al .
( 2014 ) in their discovery paper of smss j031300 , a galactic halo star with only an upper limit on the iron abundance of @xmath0 < -7.1 $ ] . the observed abundance pattern of this extremely iron deficient star is optimally matched by the nucleosynthetic yields of a @xmath6 pop iii sn with a small amount of ejecta mixing due to rayleigh - taylor instabilities , followed by extensive fallback of material onto the black hole remnant @xcite .
they argue against a jet - like explosion as this would lower the [ mg / fe ] below that observed ; in addition , rapidly rotating star models are also disfavoured by the low upper limit on n compared to c @xcite .
a general conclusion of the above analyses is that below [ fe / h ] @xmath7 there appear to be two populations of stars , carbon - normal and carbon - rich , whose observed elemental abundances reflect the nucleosynthetic products of `` normal '' and `` faint '' pop iii sne , respectively . the different composition in the material out
of which these low - mass stars form have stimulated the hypothesis that their formation relies on two different cooling channels ( norris et al .
2013 ) : fine - structure - line cooling and dust cooling .
in particular , c - rich stars form from c - enriched material where gas cooling and fragmentation is dominated by cii ( and eventually oi ) fine structure lines @xcite .
frebel et al .
( 2007 ) have introduced the so - called transition discriminant , @xmath8 } + 0.9 \times 10^{[o / h]})$ ] and predicted that metal - poor stars forming through this mechanism must have @xmath9 @xcite .
nearly all stars satisfy this criterion with the exception of sdss j1029151 + 1729 ( caffau et al .
this is a carbon - normal star with [ fe / h]@xmath10 and a solar - like chemical abundance pattern , leading to a total metallicity of @xmath11 and representing the most chemically pristine object currently known .
the origin of the star has been investigated by schneider et al . ( 2012b ) who suggest that the observed abundance pattern of sdss j1029151 can be well matched by yields of `` normal '' pop iii sne with @xmath12 and @xmath13 ; the mass of dust formed in these sn ejecta is enough to satisfy the minimal conditions to activate dust - induced fragmentation , which require the gas to be pre - enriched above a critical dust - to - gas mass ratio @xmath14 \times 10^{-9}$ ] , with the spread reflecting the dependence on the grain properties @xcite . note that even if dust formation in the first sne is less efficient or if strong dust destruction by the sn reverse shock occurs , grain growth during the collapse of the parent gas cloud is sufficiently rapid to activate dust cooling and fragmentation @xcite . in principle , the values
\times 10^{-9}$ ] and @xmath9 could provide a way to test dust cooling and fine - structure - line cooling against the observations .
this has been recently explored by ji et al .
( 2014 ) who - however - make the apriori assumption that dust is made by silicates . in this way they estimate the critical silicate abundance which corresponds to @xmath16 and compare these values to chemical abundances of metal - poor stars , with [ fe / h ] @xmath17 .
they find that the stars exhibit either high carbon with low silicon abundances or the reverse .
a silicate dust scenario would thus support the hypothesis that the earliest low - mass star formation in the most metal - poor regime may have proceeded by means of two distinct cooling pathways : fine - structure line cooling and dust cooling . here
we intend to investigate the origin of cemp - no stars , and in particular of the recently discovered most iron - deficient member of this class , smss j031300 @xcite . for the first time
, we consider the formation of dust in faint pop iii sn ejecta .
our aim is to verify if dust can form in these c - rich , iron - poor ejecta , and what is the relative role of fine - structure - line cooling and dust cooling in the collapse of material enriched by these explosions .
having this goal in mind , we first compare the observed abundances of smss j031300 with the nucleosynthetic yields of model pop iii sne that range in progenitor mass , internal mixing , and explosion energy ( extent of fallback , section 2 ) .
we perform dust calculation in the ejecta of the selected sn models with a revised version of the bianchi & schneider ( 2007 ) model which includes an improved , non - steady state , molecular network , updated radiative association reactions for co , sio , c@xmath18 and o@xmath18 molecules , and formation / destruction processes between molecules ( marassi et al . 2014a ; marassi et al . 2014b ,
section 3 ) .
we follow the collapse of a gas cloud assuming that the newly formed dust and the metals present in the ejecta are uniformly mixed into the collapsing cloud with the same total metallicity that is observed on the surface of smss j031300 and verify under which conditions the formation of low - mass star is enabled ( section 4 ) .
finally , we discuss the implications of our results for the origin of the first low - mass stars in the universe ( section 5 ) .
our aim here is to identify a set of plausible pop iii sn models which may have polluted the birth cloud of smss j031300 .
the procedure that we follow is to minimize the scatter with the observed abundance pattern adopting the chemical yields produced by a set of pop iii core - collapse sn models with mixing and fallback .
we refer the interested reader to limongi & chieffi ( 2012 ) for a thorough presentation of the models .
similarly to what has been done in schneider et al .
( 2012b ) , we consider a set of models that extends in mass between 13 and 80 @xmath19 , which appear to be consistent with the mass range found by the most recent simulations of first star formation ( hirano et al .
these models have been followed from the pre - main sequence phase up to the onset of the iron core collapse by means of the franec stellar evolutionary code .
the initial composition is set to the pristine big bang nucleosynthesis ( y=0.23 ) and mass loss is switched off during all the evolutionary stages .
the explosive nucleosynthesis has been computed in the framework of the induced explosion @xcite approximation . for each progenitor model
, we identify a combination of mixing and fallback that provides the best - fit to the observed elemental abundance ratios ( or upper limits ) on the surface of smss j031300 .
the observational data points are shown in fig.[fig : fit1d3d ] : black points with arrows show the inferred 1d/3d model atmosphere upper limits and the filled black dots with errorbars are the 1d/3d model atmosphere detections ( see keller et al
. 2014 for a description of the model atmosphere used for the abundance analysis ) . among the three elements that have been detected - c , mg , and ca - we show the elemental abundances relative to ca . the low upper limit on [ n / ca ] disfavour sn progenitors with initial masses in the range @xmath20 \ , m_{\odot}$
in fact , their evolution leads to comparable c and n abundances in the ejecta ( see fig.7 and the discussion in limongi & chieffi 2012 ) , in contrast to the observed pattern in smss j031300 .
hence , we include in the analysis sn progenitors with initial masses equal to @xmath21 .
for each of these models , we have applied the optimal combination of mixing ( set by a minimum and maximum mass coordinates within which the mass of each element is uniformly mixed in the ejecta ) and fallback ( set by the mass - cut ) that minimizes the scatter with the observed abundance pattern . since the elemental abundances inferred from the analysis using 1d or 3d model atmospheres lead to sensible differences in the observed abundances ( or upper limits ) of elements that are critical for the resulting dust mass , such as c , o , and mg , for each progenitor mass we have independently inferred best - fit models using 1d and 3d observed elemental abundances .
the resulting models are shown by the starred points joined with dotted lines in fig.[fig : fit1d3d ] .
it is clear from the figure that while all the models show an overall agreement with the observed abudance pattern , it is hard to simultaneously accomodate the upper limit on [ o / ca ] and the detected [ mg / ca ] ; this is particularly true for low mass progenitors , whose evolution leads to a lower [ mg / o ] in the ejecta .
the best - fit model appears to be the most massive progenitor with @xmath22 : adopting the abundance pattern inferred from 1d atmosphere models , the predicted ejecta composition is compatible with all the data points .
this is not true , however , when the 3d data points are considered ( see the right panel in fig.[fig : fit1d3d ] ) as the more stringent upper limit on [ o / ca ] and the larger [ mg / ca ] can not be reproduced . note that in keller et al .
( 2014 ) , the model that provides the optimal match to the observed abundance pattern is a @xmath23 explosion of a @xmath24 star of primordial composition with a small amount of ejecta mixing and a strong fallback . these properties closely resemble our best - fit model and indeed the predicted [ mg / h ] falls below the observational data point ( see fig . 3 in keller et al .
2014 ) .
table [ tab : masses ] summarizes the main properties of the sn progenitors . for all models ,
we choose the interval of mixing and mass - cut to reproduce the observed [ c / ca ] and maximize the [ mg / ca ] , getting as close as possible to the observed value without exceeding the [ o / ca ] and [ fe / ca ] upper limits .
we find that for all but the most massive model , the ejecta composition is insensitive to the reference set of observed abundances , whether derived with 1d or 3d model atmosphere ( the values quoted in the table refer to the 3d set ) .
conversely , the chemical structure of the @xmath22 progenitor , characterized by a more extended mg shell , leads to [ o / ca ] and [ c / ca ] abundance ratios that are more sensitive to the mixing and fallback prescription adopted . as a result , in table [ tab : masses ]
we show the ejecta composition for both sets of observed abundances and refer to the models as @xmath25 - 1d or @xmath25 - 3d . [ cols="<,^,^,^,^,^,^,^",options="header " , ]
in this section , we test the hypothesis that smss j031300 has formed out of the collapse and fragmentation of a gas cloud that has been pre - enriched by the ejecta of one of the pop iii faint sn studied in the previous section .
we first discuss the expansion of the ejecta into the surrouding ism and then the subsequent recollapse of the star forming cloud , to assess the relative role of fine - structure - line and dust cooling for the formation of smss j031300 .
+ the mixing and fallback model - as first introduced by umeda & nomoto ( 2002 ) - naturally predicts a stratified ejecta : the inner material is assumed to be mixed by the growth of rayleigh - taylor instabilities during shock propagation , while the ejecta is dominated by the external material plus a small fraction of the mixed gas near the mass - cut .
the mixing efficiency is still a matter of debate due to the complexity of the mixing / fallback scenario , which depends on the explosion energy , the gravitational potential , and the asphericity during the collapse .
this is the reason why we have studied dust formation in both perfectly mixed and stratified ejecta . yet , when we compare the observed surface elemental abundance of a stellar fossil , such as smss j031300 , with the yields of a single pop iii sn , we make the implicit assumption that the ejecta material has uniformly enriched the parent star forming cloud .
hence , whatever is the degree of mixing at nucleation , 45 - 80 days after the explosion , or after the passage of the reverse shock , @xmath26 years later , the ejecta material needs to be perfectly mixed into the star forming cloud before its subsequent collapse .
an estimate of the maximum timescale available to fully mix the enriched material can be obtained by computing the total mass and radius of the cloud , m@xmath27 , in which the ejecta needs to be diluted in order to match the observed metallicity on the surface of smss j031300 , @xmath28 @xcite , @xmath29 where @xmath30 and we have assumed an ism with primordial composition . we find that m@xmath27 is an increasing function of the progenitor mass , since more massive progenitors have larger metal yields , and ranges between @xmath31 and @xmath32 .
this gas mass is roughly consistent with the baryonic content of a @xmath33 mini - halo .
the radius of the cloud depends on the density of the ism ( @xmath34 ) : to be consistent with the reverse shock models described in the previous section , we assume values in the range @xmath35 \,gr\ , cm^{-3}$ ] , where the lower limit approximately corresponds to the critical density at @xmath36 .
the corresponding cloud radius is , @xmath37^{1/3}\ ] ] and varies in the range @xmath38 , which is reached by the sn shock in its sedov - taylor phase on a timescale , @xmath39 myr .
following madau , ferrara & rees ( 2001 ) , we can estimate the growth timescale of rayleigh - taylor instabilities as @xmath40 , where @xmath41 is the gravitational acceleration and @xmath42 is the spatial scale to be mixed . using the values computed above for the cloud mass and radius
, we find @xmath43 myr , depending on the ism density .
hence , in less than a few myr after the explosion of a faint pop iii sn , the ejecta material may have already mixed and enriched the gas out of which second generation stars form .
the resulting stellar mass depends on the cooling and fragmentation properties of the collapsing gas , which we investigate in the following section .
+ we perform a one - zone semi - analytic collapse calculation to follow the temperature evolution of the cloud @xcite , setting the abundances of heavy elements and dust species as inferred from the sn dust models described in the previous section .
we refer the interested reader to chiaki et al .
( 2014 ) for a detailed description of the model .
in this collapse calculation , we solve non - equilibrium chemistry of 27 gas - phase species ( h@xmath44 , @xmath45 , h , h@xmath46 , h@xmath18 , d@xmath44 , d , hd , c@xmath44 , c , ch , ch@xmath18 , co@xmath44 , co , co@xmath18 , o@xmath44 , o , oh@xmath44 , oh , h@xmath18o@xmath44 , h@xmath18o , h@xmath47o@xmath44 , o@xmath48 , o@xmath18 , si , sio , and sio@xmath18 ) with a chemical network of 55 reactions .
radiative cooling owing to atoms / ions ( ci , cii , and oi ) and molecules ( h@xmath18 , hd , co , oh , and h@xmath18o ) , and the heating / cooling by h@xmath18 formation / destruction are computed self - consistently at each time .
gas cooling by heat transfer between gas and dust is calculated as the sum over all dust species and size bins .
in addition to the dust mass that is present in the gas cloud at the onset of collapse , we also treat accretion of gas - phase metal species onto dust grains during the collapse ( grain growth , nozawa et al .
2012 ; chiaki et al .
for the purpose of the present analysis , we set the cosmic microwave background ( cmb ) temperature at @xmath49 , corresponding to @xmath50 ( see schneider & omukai 2010 for the effects of the cmb temperature floor on the collapse of clouds at different metallicities ) .
[ fig : nt ] shows the cloud evolution in the temperature - density plane .
each panel represents the results assuming that the cloud has been pre - enriched by the explosion of an individual pop iii faint sn model .
we have selected the results of three different progenitor masses ( the 13 and 15 @xmath51 models lead to very similar results ) adopting dust yields formed in uniformly mixed ejecta ( 3d models , although the results are not affected by the set of reference abundances adopted in the explosion calibration ) .
for each of these , we consider the decreasing dust - to - metal ratios that emerge from the passage of the reverse shock of increasing intensity .
the initial dust - to - gas mass ratio at the onset of collapse , @xmath52 and initial transition discriminant @xmath53 are shown in table [ tab : crit ] , where we have written in bold face only the values that are below the corresponding critical limits .
hence , we would expect metal - line cooling to be effective for all but models 13 and 15 @xmath51 in the no reverse shock case , and dust cooling to be effective for all but models @xmath54 in the two stronger reverse shock cases ( rev2 and rev3 ) .
we can compare these simple expectations with full collapse calculations .
the circles and triangles in the figure mark the states where fragmentation conditions are met @xcite . at densities @xmath55 ,
fragmentation is due to line - cooling . in the 13 and 15@xmath51 models , the dominant coolants are hd and ci . for models with moderate reverse shock destruction ( norev , rev1 ) , @xmath56 and - as a result - h@xmath18 formation on grain surface
is very efficient , enabling the formation of hd molecules . on the other hand , ci cooling efficiency is comparable to ( 3d cases ) or exceeds ( 1d cases , where [ c / h ] is larger than 3d ) hd cooling for rev2 and rev3 models .
for the larger progenitor mass models ( 50 and @xmath57 ) , the dominant coolant at intermediate densities is ci as a result of the lower initial condensation efficiencies . yet , as shown by the diagonal lines in the three panels , the characteristic masses of the fragments formed in this early stage of collapse ( the jeans mass at the density where instability occurs ) are estimated to be @xmath58 .
it is important to note that although sputtering of ac grains release c atoms in the gas phase that may be potentially accreted onto the surviving grains at a later stage of the collapse , grain growth has a non - negligible impact on the results only if @xmath59 at the onset of collapse .
in fact , ac grain growth is prevented by efficient formation of co molecules at @xmath60 @xcite .
this is confirmed by comparing the carbon condensation efficiencies at the onset of collapse , @xmath61 , and a later stage of the collapse , @xmath62 , when the density is large enough for grain growth to occur efficiently ( @xmath63 , see below ) .
the values predicted by each model are shown in table [ tab : crit ] ; as expected , grain growth has non negligible effect only for the 13 and 15 @xmath19 models with reverse shock .
hence , for all but the smallest progenitor mass models , the conditions for dust cooling and fragmentation are largely determined by the dust - to - gas ratio initially present in the cloud as a result of the sn explosion and the associated reverse shock .
[ fig : nt ] shows that dust cooling is efficient at high densities @xmath64@xmath65 triggering fragmentation and potentially leading to the formation of a low - mass star as smss j031300 .
the only two unsuccessful cases are associated to the @xmath66 faint sn explosion with the stronger reverse shock models .
not surprisingly , in these two models the amount of dust that survives the passage of the reverse shock is not enough to keep the initial dust - to - gas mass ratio above the critical one , @xmath15 \times 10^{-9}$ ] ( see table [ tab : crit ] ) .
+ finally , it can be shown that the full collapse calculations confirm that even in unmixed models , which are characterized by smaller dust yields , dust - driven fragmentation conditions are met for all reverse shock cases considered , consistent with the fact that the initial dust - gas ratios are super - critical , varying in the range @xmath67 . following chiaki et al .
( 2014 ) , we can estimate the critical elemental abundances that enable dust - induced fragmentation . for cemp stars enriched by faint pop iii sn explosions ,
carbon grains are the dominant species in the parent cloud ; therefore , the critical dust - to - gas ratio can be easily associated to a critical carbon abundance , @xmath68 @xmath69 where @xmath70 is the material density of carbon grains ( @xmath71 ) and @xmath72 is the characteristic grain radius for gas cooling @xcite .
table [ tab : crit ] shows the cooling radii and the critical carbon abundances normalized to the solar value ( @xmath73 taken from the photospheric value in asplund 2009 ) . since the observed carbon abundance of smss j031300 is @xmath74 = -2.60 $ ] @xcite .
the critical conditions are satisfied by all the models but the @xmath66 rev2 and rev3 .
this is consistent with the full collapse calculations .
hence we conclude that the formation of the cemp star smss j031300 , even with the lowest iron - content currently observed , may have been triggered by dust cooling and fragmentation .
the results presented in the previous section have been obtained using the dust and metal yields predicted for faint pop iii sn explosions tailored to reproduce the observed elemental abundances of smss j031300 .
for the first time , we have studied the process of dust formation in carbon - rich , iron - poor ejecta that characterize these explosions .
we find that carbon can be significantly depleted onto ac grains , with condensation efficiencies which can be as large as 75% .
due to the small abundance of other refractory elements in the ejecta , silicates , alumina or magnetite grains do not form in the ejecta of faint pop iii sn . for most of the models that we have explored , when normalized to the observed [ c / h ] which characterize cemp stars , the predicted dust - to - gas ratios in the parent star - forming cloud are large enough to trigger dust cooling and fragmentation , even without the contribution of grain growth .
if these results can be considered as representative for the general class of cemp - no stars , which represent 90% of cemp with @xmath75 < -3 $ ] , then our study suggests that dust cooling and fragmentation may have been responsible for the formation of these low - mass stars . in a recent work ,
ji et al . (
2014 ) have suggested that the early formation of low - mass stars may follow two alternative pathways : in the dynamic pathway , many sub - clumps are formed through fine - structure line cooling .
dynamical interactions can lead to the ejection of a protostar from the parent cloud , so that a low - mass star can form even in the absence of dust cooling . in the thermal pathway ,
fragmentation due to fine - structure cooling is inhibited by the small abundance of gas - phase elements , and the entire cloud collapses .
a high - density protostellar disk is formed at the center of the cloud , which then fragments and forms low - mass stars due to dust cooling .
observational evidence for these two alternative pathways comes from the distribution of a sample of metal - poor halo stars and dwarf galaxy stars with @xmath0 < -3.5 $ ] in a plane where on the y - axis is the transition discriminant for line - cooling , @xmath76 and on the x - axis is the [ si / h ] ( see their fig .
6 ) ; the latter quantity is used as an indication of the dust - to - gas ratio in the parent cloud , assuming that the dust is mostly made of silicates . when compared with the critical values for fine - structure - line and dust cooling
, they find that stars that have a [ si / h ] too low to form through silicon - based dust cooling are characterized by a large transition discriminant , and satisfy the criterium for fine - structure line cooling and viceversa .
our study questions this interpretation as it shows that cemp stars form in c - rich gas , where silicates do not form and carbon dust is the dominant dust species .
hence , for cemp stars , [ c / h ] is a better proxy for dust cooling than [ si / h ] .
in addition , we also show that fine - structure - line and dust - cooling are not mutually exclusive : the collapsing gas cloud undergoes two distinct phases of fragmentation , one driven by line cooling ( either by metal lines , such as ci , cii or by molecular transitions of h@xmath18 and hd ) at @xmath60 which leads to the formation of @xmath77 sub - clumps ; and the other , driven by dust cooling at @xmath64@xmath65 , which triggers the formation of low - mass stars ( see also tanaka & omukai 2014 ) .
thus , the thermal pathway may be responsible for the formation of both c - rich and c - normal stars , even at lowest [ fe / h ] currently observed , although the dynamic pathway may operate as well .
unfortunately , it is not straightforward to estimate the dust - to - gas ratio of the parent gas cloud from the observed elemental abundances , similar to what is done for the transition discriminant .
in fact , the dust properties ( its composition and size distribution ) depend on the nature of the sn explosions that have dominated the early enrichment and on the level of destruction suffered by the newly formed dust by the passage of the reverse shock , although the latter process may be mitigated by grain growth during the collapse . using the results that we have obtained in the present study as a proxy for the general class of c - rich metal - poor stars and the results obtained by schneider et al .
( 2012b ) and chiaki et al .
( 2014 ) for the stars sdss j1029151 as a proxy for the general class of c - normal metal - poor stars , we can draw the following guidelines : for c - rich stars , carbon condensation efficiencies into dust of @xmath78 translate into critical carbon abundances of @xmath79_{cr } < -3.01 $ ] ; for c - normal stars , the dominant dust compounds are forsterite ( mg@xmath18sio@xmath80 ) and enstatite ( mgsio@xmath47 ) , with mg and si condensation efficiencies which range between @xmath81 and associated critical abundances of @xmath82_{cr } < -4.73 $ ] and @xmath83_{cr } < -4.73 $ ] ( see table 1 in @xcite ) . in figs . 6 and 7 , we compare these predictions with observational data taken from the sample of c - normal and c - rich stars with @xmath0 < -3 $ ] investigated by norris et al .
as shown in caffau et al .
( 2011 ) , sdss j1029151 is the only star , discovered until now , that falls in the forbidden zone for fine - structure line cooling ( indicated by the triangle below @xmath16 in fig.[fig : csicr ] , [ fig : dcr ] ) and no stars appear to fall in the forbidden zone for dust - cooling although there are two c - rich stars , he0557 - 4840 ( norris et al .
2007 ) and he0057 - 5959 ( yong et al . 2012 ) , which appear to be close to the critical range .
if we represent the critical conditions in terms of the dust - to - gas ratio , as in fig .
[ fig : dcr ] , we see that c - normal and c - rich stars do not appear to follow different distributions on the plane , although - as expected - the c - normal stars have systematically lower @xmath76 and lower dust - to - gas - ratios , @xmath84 . more importantly , the two criteria for low - mass star formation do not appear to be mutually exclusive , but rather point to a single pathway for the formation of the first low - mass stars in the universe .
we thank simone bianchi for his kind collaboration .
the research leading to these results has received funding from the european research council under the european union s seventh framework programme ( fp/2007 - 2013)/erc grant agreement n. 306476 .
gc is supported by research fellowships of the japan society for the promotion of science ( jsps ) for young scientists .
ny is grateful for financial support by the grants - in - aid by the ministry of education , science and culture of japan ( 25287050 ) .
we aknowledge financial support from prin miur 2010 - 2011 , project `` the chemical and dynamical evolution of the milky way and local group galaxies '' , prot .
2010ly5n2 t .
andreazza , c. m. , singh , p. d. , 1997 , , 287 , 287 aoki , w. , beers , t. c. , honda , s. , & carollo , d. , 2010 , , 723 , l201 asplund , m. , grevesse , n. , sauval , a. j. , & scott , p. , 2009
, ara&a , 47 , 481 beers , t. c. & christlieb , n. , 2005 , ara&a , 43 , 531 babb , j. f. & dalgarno , a. , 1995 , phys .
a , 51 , 3021 bianchi , s. & schneider , r. , 2007 , , 378 , 973 bromm , v. & loeb , a. , 2003 , nature , 425 , 812 caffau , e. , bonifacio , p. , francois , p. , sbordone , l. , monaco , l. , spite , m. , ludwig , h.g . ,
cayrel , r. et al . , 2011 ,
nature , 477 , 67 cayrel , r. , depagne , e. , spite , m. , hill , v. , spite , f. , francois , p. , plez , b. , beers , t. et al .
, 2004 , a&a , 416 , 1117 campbell , s. w. , lugaro , m. , & karakas , a. i. , 2010 , a&a , 522 , l6 cherchneff , i. , & dwek , e. , 2009 , , 703 , 642 cherchneff , i. , & dwek , e. , 2010 , , 713 , 1 chiaki , g. , nozawa , t. , & yoshida , n. , 2013 , , 765 , l3 chiaki , g. , schneider , r. , nozawa , t. , omukai , k. , limongi , m. , yoshida , n. , chieffi , a. , 2014 , , 439 , 3121 dalgarno , a. , du , m. l. , & you , j. h. , 1990 , apj , 349 , 675 frebel , a. , johnson , j. l. & bromm , v. , 2007 , , 380 , l40 frebel , a. & norris , j. e. , 2013 , planets , stars and stellar systems , springer heger , a. , & woosley , s.e . , 2010 , , 724 , 341 hirano , s. , hosokawa , t. , yoshida , n. , umeda , y. , omukai , k. , chiaki , g. , and yorke , h. w. , 2014 , , 781 , 60 hirschi , r. , 2007 , a&a , 461 , 571 ji , a. p. , frebel , a. , & bromm , v. , 2014 , , 782 , 95 joggerst , c. c. , woosley , s. e. , & heger , a. , 2009 , , 693 , 1780 keller , s. c. , bessel , m. s. , frebel , a. , casey , a.r .
, asplund , m. , jacobson , h. r. , lind , k. , norris , j.e .
et al . , 2014 ,
nature , 506 , 463 limongi , m. & chieffi , a. 2012 , , 199 , 38l limongi , m. & chieffi , a. 2006 , , 647 , 483 limongi , m. & chieffi , a. 2003 , , 592 , 404 lucatello , s. , tsangarides , s. , beers , t. c. , carretta , e. , gratton , r. g. , ryan , s.g .
, 2005 , apj , 625 , 825 madau , p. , ferrara , a. , rees , m. j. , 2001 , , 555 , 92 marassi , s. , schneider , r. , limongi , m. , chieffi , a. 2014 , proceedings of science , _ the life cycle of dust in the universe : observations , theory , and laboratory experiments - lcdu 2013 , 18 - 22 november 2013 taipei , taiwan _ marassi , s. , schneider , r. , limongi , m. , chieffi , a. 2014 , in preparation masseron , t. , johnson , j. a. , plez , b. , van eck , s. primas , f. , goriely , s. , jorissen , a. , 2010 , a&a , 509 , 93 meynet , g. , ekstrm , s. , & maeder , a. 2006 , a&a , 447 , 623 meynet , et al .
2010 , a&a , 521 , a30 norris , j. e. et al .
2013 , , 762 , 28 nozawa , t. & kozasa , t. 2003 , 598 , 785 nozawa , t. , kozasa , t. , habe , a. , dwek , e. , umeda , h. , tominaga , n. , maeda , k. , nomoto , k. 2007 , apj , 666 , 955 nozawa , t. , kozasa , t. , tominaga , n. , maeda , k. , umeda , h. , nomoto , k. , krause , o. , 2010 , 713 , 356 nozawa , t. , kozasa , t. , nomoto , k. 2012 , , 756 , 35l omukai , k. , 2000 , , 534 , 809 omukai , k. , tsuribe , t. , schneider , r. , ferrara , a. , 2005 , , 626 , 627 paquette , j. a. , & jnuth iii , j.a .
2011 , , 737 , l6 sarangi , a. & cherchneff , i. 2013 , , 776 , 107 schneider , r. , ferrara , a. , salvaterra , r. , omukai , k. , & bromm , v. 2003 , nature , 422 , 869 schneider , r. , ferrara , a. , salvaterra r. 2004 , , 351 , 1379 schneider , r. & omukai , k. 2010 , , 402 , 429 schneider , r. , omukai , k. , bianchi , s. , valiante , r. , 2012a , , 419 , 1566 schneider , r. , omukai , k. , limongi m. , ferrara a. , salvaterra r. , chieffi a. , bianchi , s. , 2012b , , 423 , l60 schneider , r. , valiante , r. , ventura , p. , dellagli , f. , di criscienzo , m. , hirashita , h. , kemper , f. 2014 , , in press silvia , d. w. , smith , b. , d. , & shull , j. m. 2010 , apj , 715 , 1575 silvia , d. w. , smith , b. , d. , & shull , j. m. 2012 , apj , 748 , 12 suda , t. , aikawa , m. , machida , m. n. , fujimoto , m. y. , iben , i. j. 2004 , , 611 , 476 tanaka , k. e. i. & omukai , k. , 2014 , , 439 , 1884 todini , p. & ferrara , a. 2001 , , 325 , 726 tominaga , n. , umeda , h. , & nomoto , k. 2007 , , 660 , 516 truelove , j. k. & mckee , c. f. 1999 , apj , 120 , 299 umeda , h. & nomoto , k. 2002 , , 565 , 385 umeda , h. & nomoto , k. 2003 , nature , 422 , 871 umist datatbase for astrochemistry 2012 , http://udfa.ajmarkwick.net woosley , s. e. , 1988 , , 330 , 218 woosley , s.e . , pinto , p. a. , hartmann , d. , 1989 , , 346 , 395 valiante , r. & schneider , r. 2014 , proceedings of science , _ the life cycle of dust in the universe : observations , theory , and laboratory experiments - lcdu 2013 , 18 - 22 november 2013 taipei , taiwan _ yong , d. , norris , j. e. , bessell , m. s. , christlieb , n. , asplund , m. , beers , t.c . , barklem , p. s. , frebel , a. et al .
, 2013a , , 762 , 26 yong , d. , norris , j. e. , bessell , m. s. , christlieb , n. , asplund , m. , beers , t.c . , barklem , p. s. , frebel , a. et al . , 2013b , , 762 , 27 | we investigate the origin of carbon - enhanced metal - poor ( cemp ) stars starting from the recently discovered @xmath0<-7.1 $ ] star smss j031300 ( keller et al . 2014 ) .
we show that the elemental abundances observed on the surface of smss j031300 can be well fit by the yields of faint , metal free , supernovae . using properly calibrated faint supernova explosion models , we study , for the first time , the formation of dust grains in such carbon - rich , iron - poor supernova ejecta .
calculations are performed assuming both unmixed and uniformly mixed ejecta and taking into account the partial destruction by the supernova reverse shock .
we find that , due to the paucity of refractory elements beside carbon , amorphous carbon is the only grain species to form , with carbon condensation efficiencies that range between ( 0.15 - 0.84 ) , resulting in dust yields in the range ( 0.025 - 2.25)m@xmath1 .
we follow the collapse and fragmentation of a star forming cloud enriched by the products of these faint supernova explosions and we explore the role played by fine structure line cooling and dust cooling .
we show that even if grain growth during the collapse has a minor effect of the dust - to - gas ratio , due to c depletion into co molecules at an early stage of the collapse , the formation of cemp low - mass stars , such as smss j031300 , could be triggered by dust cooling and fragmentation .
a comparison between model predictions and observations of a sample of c - normal and c - rich metal - poor stars supports the idea that a single common pathway may be responsible for the formation of the first low - mass stars . |
understanding the geometry and kinematics of the central regions in agn is one of the major goals of twenty - first century astronomy .
yet little is known beyond the necessity of invoking accretion onto a supermassive object in order to account for the enormous source luminosity . given our inability to resolve the central regions of even the nearest agn
we are left with three approaches towards achieving our goal : ( 1 ) ideology : sophisticated models driven entirely by theories describing black hole and associated accretion disk properties ( e.g. , janiuk , .
zycki , & czerny 2000 ; gammie , shapiro , & mckinney 2004 ) , ( 2 ) analogy : comparison with better - resolved galactic objects ( e.g. interacting binary stars ; zamanov & marziani 2002 ; mirabel 2004 ; maccarone , gallo , & fender 2003 ) and ( 3 ) empiricism : detailed studies of phenomenology in large and diverse samples of agn ( see sulentic , marziani & dultzin - hacyan 2000a for references until late 1999 ; marziani et al .
2003a , hereafter m03 ; dietrich et al . 2002 ; kuraszkiewicz et al .
2004 for examples of more recent works ) .
the last approach fell on hard times until the end of the twentieth century in part due to : ( i ) the apparent intractability of agn phenomenology , ( ii ) the failure of the best nebular physics models to reproduce the observed spectra , and ( iii ) the low quality of so much of the spectroscopic data .
thus we now know more about the morphology of the extended radio and optical emission around quasars than the underlying geometry / physics that drives them .
one positive development involves the application of reverberation mapping towards estimation of the broad line region size ( see peterson 1993 for a thorough review ; horne et al . , 2004 for up - to - date expectations ) .
the relevant questions are how to interpret these data and , more importantly , how to select a representative sample of sources for a reverberation mapping study .
another positive development involves growing databases of agn spectra ( most notably from the sloan digital sky survey ( sdss ) ; richards et al . 2002 ; schneider et al . 2002 ) .
although great part of the individual data still suffers from s / n limitations , average or composite spectra of many objects can significantly improve the situation ( vanden berk et al .
2001 ; sulentic et al .
one of the most important questions is _ how _ to bin the data .
clearly unrestricted composites have little value because they hide the intrinsic dispersion in key parameters which provides clues to the underlying physics ( e.g. , francis et al . 1991 ) . in a previous paper ( sulentic et al .
2002 : hereafter paper i ) we argue that composite spectra binned within the eigenvector 1 parameter space provide the most physically useful results .
paper i presented composites of the h@xmath4 spectral region and this paper presents complementary composites for the uv spectral region and especially for .
we have identified a four dimensional parameter space that overcomes some of the problems mentioned above and that offers a new perspective on agn phenomenology ( sulentic et al .
2000a , b ) .
the eigenvector 1 ( e1 ) parameter space provides optimal discrimination / unification of broad - line agn classes ( boroson & green 1992 ; marziani et al .
2001 , hereafter m01 ; boroson 2002 ) .
the adopted name reflects recognition of the first two parameters in the pioneering work of boroson & green ( 1992 ) . e1
as we now define it involves measures of : 1 . full width at half maximum of low ionization ( lil ) broad emission lines [ fwhm ( ) ] , 2 .
equivalent width ratio of optical feii and : = ew ( blend)/ew ( ) , 3 .
soft x - ray photon index ( @xmath5 ) , 4 .
profile centroid velocity displacement ( at fwhm ) for . in simplest terms
these parameters can be said to measure respectively : ( 1 ) velocity dispersion of lil emitting gas , ( 2 ) relative strength of lil features thought to arise in the same region , ( 3 ) strength of a soft x - ray component and ( 4 ) amplitude of systematic motions in the high - ionization line ( hil ) emitting gas . in less simple ( i.e. more model dependent ) terms
we likely have : ( i ) two variables mainly sensitive to accretion rate ( fwhm ( ) and : nicastro 2000 ; m01 ; pounds , done & osborne 1995 ; czerny et al .
2004 ) , ( ii ) three variables sensitive to source orientation ( all with the possible exclusion of ; e.g. , jackson & browne 1991 ; wills et al . 1995 ; marziani et al . 1996
; m01 ; rokaki et al .
2003 ; sulentic et al .
2003 ) . is sensitive to nebular physics ( electron density @xmath6 and column density @xmath7 ) .
an important addition that reflects the striking changes in emission line properties across e1 involves introduction of the population a and b concept .
at present we arbitrarily define pop .
a and b quasars as those with fwhm less than or greater than 4000 .
it is unclear whether the a - b concept reflects two distinct quasar populations or simply the difference in mean properties of sources at opposite ends of the e1 `` main sequence '' .
the `` elbow '' shape of the sequence , and optical / x - ray indications ( sulentic et al .
2000a , b ; komossa & meerschweinchen 2000 ) for a `` zone of avoidance '' near fwhm ( ) @xmath2 4000 , provide some evidence in support of the former interpretation .
if the distinction is real then `` a '' is a largely rq population ( including narrow line seyfert 1 sources as an extremum ) , while `` b '' contains a mix of broader line rq plus most rl sources .
the population a - b distinction will prove useful in this study and will add support for the concept ( 3.2 and 4.3 ) .
if the e1 concept lives up to initial promise , it may be the strongest and most encompassing type-1 agn unification yet discovered .
this possibility warrants strong tests of its strength and generality .
so far most of our effort has centered on the `` optical plane '' of e1 ( m01 ; zamanov et al .
2002 ; boroson 2002 ; marziani et al .
2003a , b ; sulentic et al . 2003 ) .
this focus has been dictated by the available x - ray and uv archival data that provide the other e1 parameters . in the next section we present composite spectra for ( and other uv lines ) in the e1 context which complements work already done for .
a total of about 700 reasonable quality ( s / n @xmath8 45 ) hst spectra for n = 141 different agn ( quasars or bright seyfert 1 s ; n=61 are radio - loud with @xmath9 , where @xmath10 is the specific flux ratio between 4400 and 6 cm ) have been extracted from the hubble archive . the majority of the spectra ( 83 % ) were obtained with the fos camera and the rest with stis .
the fos / bl data ( 40 % ) were recalibrated with the latest iraf package
stpoa , which takes into account some systematic errors in the wavelength calibration .
no such recalibration is available for the fos / rd , where similar errors are known to exist , which ultimately limits the wavelength accuracy of our composites to about 0.51.0 ( see for instance kerber & rosa , 2000 , for details on fos calibration ) .
composite spectra cover the entire rest frame region between 10003000 while individual source spectra will often cover only a part of it .
we applied the following procedures to compute the composite spectra : 1 .
spectra for a specific source covering the same spectral region were averaged together ( if they were of comparable s / n ) .
a single spectrum for each object was created by adding the parts covering different wavelength regions , with slight adjustments of the continuum levels if necessary .
all spectra were corrected for galactic extinction following schlegel et al .
the typical extinction correction @xmath11 was about 0.1 0.2 magnitude .
no correction for the intrinsic extinction was attempted .
the spectra were de - redshifted using the most accurate available measurements ( m03 ) or by using the latest accurate values available in the literature whenever a source was not part of the m03 sample@xmath125007 or the narrow peak of h@xmath4 ( see zamanov et al .
2002 ) ; they are available from the authors on request . ] .
it should be noted that redshifts from vron - cetty & vron ( 2001 ) ; hereafter vcv01 , sometimes differ considerably from our best values ( @xmath13could be as large as 0.02 ) .
spectra were normalized by their own continuum flux in the wavelength range 14301470 , where no strong lines are observed / expected .
composite spectra were generated by median - combining the averaged spectra of individual agn .
this procedure ensures that only the most typical features will be revealed , and that some unusual features like narrow absorption ( often lost in the noise ) , unusually strong ( narrow ) lines , noise spikes , etc .
will be omitted in the composites .
a median also works better than an average when the individual spectra do not cover the same spectral region ( see 4 for details regarding line profiles of the composites ) .
the composites have been binned to 0.5 per resolution element and have a typical s / n@xmath350 ( see table 1 ) .
the composite spectra are presented in figure 1 with an enlargement of the 11002000 region of greatest interest shown in figure 2 .
the latter show details of the strongest lines such as ly , , and .
the lines that we identified and measured ( 3.1 ) are labeled .
we defined spectral types on the basis of source parameter properties in the optical plane of e1 ( paper i ) .
this approach differs from the one employed in analysis of the sdss quasar sample ( richards et al .
2002 ) where sources were binned _ a posteriori _ in terms of the velocity shift between and .
we are able to create useful composites for five e1 bins with fixed fwhm ( ) and ranges , exactly as done in paper i. the values of fwhm ( ) and that provide an optical bin assignment were taken from m03 . in order to maximize the number of sources in each bin we also include additional spectra for sources ( @xmath2 20% of the total ) , not part of the m03 sample .
this supplemental data involves : 1 ) sources with reasonable quality archival data and 2 ) spectra from literature sources with high enough s / n to permit assignment into one of the predefined e1 bins .
details on bin assignation for individual sources that are not part of the m03 sample will be provided in a future work .
table 1 provides parameter measures and statistical properties for the spectral bins .
the format is : column 1 spectral type designation following paper i ; col
. 2 range of fwhm ( ) ; col .
3 range of ; col . 4 number of sources used for the composite spectrum ( number of rl source in parenthesis ) .
each bin typically involves 15 30 sources resulting in a s / n in the continuum typically of 40 60 ( col . 5 ) .
col . 6 and 7 present the average absolute @xmath14-magnitude ( computed from vcv01 ; @xmath15 mpc@xmath16 ; @xmath17 is adopted ) and redshift for sources used in the bin composite followed by the sample standard deviations .
we also present the average equivalent widths ( ew ) of and for the objects of the sample col .
8 and 9 respectively .
there is a clear trend between @xmath18 and @xmath19 , in the sense that pop .
a sources show lower mean redshift and luminosity .
this is a known bias in the m03 dataset .
m03 is a step towards our goal to develop a sample complete to m@xmath20=16.5 . at this time
the radio - loud part of our sample is apparently complete ( the standard @xmath21 completeness test ( e.g. , peterson 1997 ) gives @xmath30.8 compared to 0.3 for rq sources ) .
the radio loud sources tend to be more luminous ( due in part to relativistic beaming ) resulting in selecting more high redshift sources .
the e1 optical parameters _ show no dependence on luminosity _ to at least @xmath22 and @xmath23 16.5 ( sulentic et al .
we will consider later possible uv line vs. source luminosity correlations ( see 3.3 and 4.3 ) .
we also computed the sample standard deviation for the spectra in each bin .
deviations from the composites at 1@xmath24 confidence level are found to be about 20% and show no specific features after being normalized by the composites themselves .
table 2 presents rest - frame ew and fwhm measures for uv lines in the composite spectra .
e1 bin designations are listed at the top of the table .
uncertainty estimates are given in parenthesis following each measure .
ew uncertainties are about 1020% for the strongest lines ( w@xmath2510 ) and approach 50% for the weaker lines .
fwhm uncertainties are generally below 2030% with the level of uncertainty depending on s
/ n of the composites as well as line blending ( e.g. ) in some cases .
all measurements were carried out after and continuum subtraction except that these corrections were not applied to .
the template covering the 1000 2000 region ( where the most interesting lines are ) was created with spectra for i zw 1 ( see marziani et al .
1996 , m03 for details ) .
the template was intensity scaled and broadened in order to best match the observed feii pattern in the composite spectra .
the results are summarized in the last row of table 2 .
figure 3 illustrates the deblending results for two particularly important groups of lines : + and + + .
the first blend is important because is the only high - ionization line that shows no significant baldwin effect ( sulentic et al .
2000a , see also 4.5 ) .
the importance of the second blend involves the fact that the intensity ratio / may be sensitive to density ( n@xmath26 ) and/or may be affected by metallicity trends along e1 ( m01 ; 4.3 ) .
we carried out a deblending procedure by assuming a lorentzian profile for both lines in the blend and allowing for different component intensities and fwhm .
a possible contamination by feiii blends may contribute to some extend to the relatively poor fit of these lines in a3 bin .
we consider several interesting results that can be summarized as follows : * , , , and show a systematic increase in equivalent width as one proceeds from extreme population a ( bin a3 ) towards the b bins .
the most dramatic change is actually seen in , and also , where bin averaged equivalent widths increase by almost a factor of four ( a factor @xmath310 is seen in if individual sources are considered ) . while we adopted civ centroid shift at fwhm as an e1 parameter
other civ measures follow the e1 sequence as well . *
contrary to the above lines shows almost constant equivalent width ( @xmath220 25 ) across the e1 sequence .
other lines show no clear trends within measurement uncertainties .
* the / intensity ratio decreases monotonically by a factor of four as one proceeds from bin a3 to b1@xmath27 ( from 0.8 to 0.2 ) .
see 4 for possible explanations .
* we find a clear positive trend between fwhm and ionization potential ( @xmath28 ) for the principal broad lines .
the correlation coefficients are in the range 0.60.8 .
fwhm typically scales linearly with @xmath28 up to @xmath29 ev and saturates beyond that value .
the scatter in this relation appears to be smaller for a - bins .
b - bins suggest that some lil such as and deviate from the relationship with larger than expected values of fwhm .
the linear part of the relation can be roughly described by the following expression : @xmath30 , where @xmath31 60 80 and v@xmath32 from @xmath331000 to @xmath333000 from bin a3 to b1@xmath27 .
the saturation occurs at about 4000 for a - bins and at about 5000 6000 for b - bins .
the line deserves special attention as the best studied , and easiest to study , hil .
the intensity and relative lack of contamination by satellite lines , motivated its selection as the representative hil in the e1 parameter space .
profile centroid shift at fwhm intensity was chosen as one of the most important e1 parameters ( sulentic et al .
the blueshifted component of the line seems to disappear in sources with fwhm @xmath84000 providing further support for the pop .
a - b distinction .
this fact and other previous works make it clear that both hil and lil can not be treated as single component lines .
a large part of the confusion in past broad line studies stems from this fact .
the simplest possible decomposition of our composite profiles was attempted after the following processing steps .
1 ) the continuum was subtracted by fitting a low - order lagrange polynomial over the region 10002000 .
2 ) a broadened and scaled template was subtracted ( 3.1 , table 2 ) .
emission is not strong in any of the bin composites however a reasonable correspondence between the optical and uv feii emission strengths can be traced in the sense that bin a3 shows the strongest emission .
3 ) contaminating lines near : and blended + were fitted with lorentzians and subtracted from the wings .
4 ) a narrow component was subtracted from the profile ( see below ) .
results of the decomposition are given in figure 4 and table 3 .
results from modelling our composites ( paper i ) are useful here .
we identified three reasonably unambiguous components in the context of e1 and the population a - b hypothesis .
these identifications , and the adopted profile fits , are largely empirical at this stage .
we focus on bins a2/a3 and b1 as representative of the two agn populations or e1 parameter space extrema .
bin a1 is likely a mix of both population a and b sources ( or a transition between the two ) . * * i ) nlr * + * :* we often see an inflection between the blr and nlr components of lil making profile decomposition less ambiguous ( marziani et al , 1996 ) .
bins a2/a3 show evidence for weak emission ( @xmath34 ) in residuals after modelling the broad line components ( paper i ) .
a more significant nlr component is found in the majority of pop b sources @xmath35 with a large scatter ) .
if we assume a fixed ratio for / , then the table 1 averages for ew ( ) imply a very weak contribution in bins a2/a3 .
this ratio is about 0.11 for pop .
b sources which would imply ew ( ) @xmath36 for bins a2/a3 which is lower than what we actually subtracted except in the `` blue outlier '' sources where is extremely weak and blueshifted ( zamanov et al . 2002 ) . +
* :* in our opinion it is clear that a narrow component exists in many , especially pop .
b , sources because we can see a profile inflection ( sulentic & marziani 1999 ) .
the situation for is more ambiguous than for because of the lower intrinsic resolution ( and often s / n ) of the spectra which less often show the inflection that can guide a profile decomposition .
either gaussian or lorentzian fits to the broad profiles of bin a1 , b1 and b1@xmath27 composites show a narrow residual . although one can not be sure that it is from an nlr component , that is the simplest interpretation and is empirically justified by many cases where nlr emission is unambiguously observed .
new evidence for an nlr component in , at least , pop b rl quasars is given in 4.4 .
we therefore argue that an uncertain nlr subtraction is better than no subtraction for the a1 and b - bins .
no nlr component was subtracted from bins a2/a3 .
the nlr profile width was assumed fwhm@xmath37 ( taking into account the fact that is a doublet ) .
photoionization models suggest ew()@xmath380.1 ew ( ) as a reasonable estimate ( i.e. about 2 4 , using the ew ( ) values from table 1 and paper i ) . we should point out that this ew ratio is rather similar to the flux ratio of the lines since the continuum shape does not seem to change significantly along the e1 bins ( fig .
1 ) . without inter - calibration of optical and uv fluxes , as well as knowledge of the nlr extinction properties
( e.g. netzer & laor 1993 ) , such numbers are only indicative .
see sulentic & marziani ( 1999 ) for more quantitative estimates of the effect of a nlr component on measures of fwhm and line shift .
our adopted estimates of ew()/ew ( ) show a large scatter ( 0.0 0.5 ) among individual sources in a1 and b - bins . *
* ii ) blr * + * :* this is the defining broad - line component for type 1 agn .
the population a b distinction is currently defined in terms of sources with fwhm smaller or larger than 4000 .
a2/a3 bin composites are best fit by an unshifted lorentzian function ( vron - cetty et al .
2001 ) while for pop b bins a gaussian model works better .
while the bin composite profiles are unshifted , individual sources can show red or blue shifts of up to 4000 .
profile shifts larger than about 600 are seen only in pop .
b sources ( sulentic et al .
2000a , b ) .
double peaked profiles are also sometimes ( @xmath15% ) seen in bins b1@xmath27 and b1@xmath39 . was defined in paper i for sources with @xmath400.5 and 8000 @xmath41 fwhm()@xmath4112000 ] measures of fwhm can be strongly influenced by the nlr component ( e.g. l / l@xmath42 and bh mass over- and under - estimated , respectively ) . + * :* a strongly blueshifted line or line component is seen only in population a sources .
the red half of the bin a2/a3 composite profiles can be reasonably well fit with an unshifted lorentzian model whose strength is inferred from the amount of flux redward of the adopted source rest frame . in this context extreme a3 sources should have only a weak unshifted component with the bulk of the line flux strongly blueshifted .
we do nt know if this decomposition has any physical meaning , i.e. whether we have two distinct line emitting regions or a single region , producing a complex profile .
there are two admittedly empirical reasons to favor the two - component model : + \1 .
is best fit with an unshifted lorentzian and there is no _ a priori _ physical reason why some hil emission could not be produced in the clouds responsible for the lil emission .
current models do not rule out the possibility of significant civ emission from the lil clouds ( korista et al .
, 1997 ) , although the accurate quantitative predictions require extensive knowledge of the physical conditions there .
the derived width of the best fit lorentzian components are in good agreement with those estimated for ( about 2500 , see paper i )
the fits in figure 4 are consistent with a systematically increasing unshifted component as one proceeds from bin a3 to b1 .
the blueshifted ( bulk of the emission in bins a2/a3 ) would be strongly blue asymmetric and would truncate rather sharply near the rest frame velocity , an interesting challenge for physical models .
this truncation is actually seen in sources like i zw1 and ton 28 where a single component model is more appropriate ( sulentic et al . 2000b , 2001 ) .
we can not distinguish between a gaussian or lorentzian model for the pop .
the choice will affect our estimate of the strength of the underlying vblr .
results from fits lead us to favor , by analogy , a two gaussian model for pop .
a sources involving unshifted blr plus redshifted vblr components . at this point
we can not place much physical meaning on these lorentzian vs. gaussian fits , they simply reflect the relative sharpness of the line peak and the extent of the profile wings . *
* iii ) vblr * + * :* there is no consensus whether or not the vblr is a distinct component ( e.g. corbin 1997 ; snedden & gaskel , 2004 ; korista & goad 2004 ) .
only pop b sources appear to show this `` red shelf '' or component that we model as a separate vblr ( see references in sulentic et al .
we prefer to speak of the vblr as a distinct component for at least two reasons : 1 ) we often see an inflection between the red wing of the best fit blr gaussian and the red feature and 2 ) we have identified sources that only show a vblr component ( sulentic et al .
paper i composites suggest that the best current description of the components involve : i ) a blr with zero mean shift and fwhm @xmath43 as well as 2 ) a vblr with mean redshift of about 5000 and fwhm@xmath3810000 . + * :* no evidence for a significant vblr component is seen in population a bins paralleling the results for .
a double gaussian decomposition is consistent with both blr and vblr components in pop b bins .
fwhm of the components are in rough agreement with those for .
interestingly vblr line shifts appear to be lower ( @xmath44 ) than those modeled in the composites , however this may be a result of the fact that the two - component fitting is not necessarily a unique operation .
+ furthermore line shift values derived for such an ultra - broad component must be considered very uncertain especially given the low s / n of many hst spectra . while we currently separate population a and b sources at fwhm ( ) = 4000
, if the distinction has any meaning then it will likely involve a complex combination of parameters .
there are two possible interpretations of the pop .
a - b concept : i ) two distinct agn populations with different central source geometry and kinematics involving perhaps a critical eddington ratio or ii ) a single `` main sequence '' involving the majority of agn and driven by a gradual change in the ionization parameter / eddington ratio ( see m01 ) .
the `` elbow '' shape of the source occupation in the optical plane of e1 is an argument in favor of the former interpretation ( i.e. it is not a monotonic correlation ) . in any case
the shape suggests that bin a1 will likely blend the parameter space properties of pop .
a and b sources .
figure 4 and tables 1 , 2 support that view .
a few sources in bin a1 ( e.g. ngc 1566 and mrk 493 ) show very strong nlr emission while others little or none .
our profile analysis supports the population a - b idea in the sense that we find line components unique to each ( pop .
a blueshifted blr , and pop b redshifted vblr ) but mutually exclusive .
our results appear to be robust for sources with [email protected] .
richards et al .
( 2002 ) created similar composite spectra using part of the much larger and higher mean redshift ( @xmath45 ) sdss quasar sample .
their spectral resolution is similar to our own but the s / n of individual spectra are usually much lower .
they binned spectra showing into four groups ( each involving almost 200 sources ) based on the centroid shift of .
we lack lil data that would allow us to assign unambiguous population a and b tags to these bins but the two extreme groups ( richards et al .
composite d : most blueshifted and composite a : least shifted ) should , more or less , correspond to our a2/a3 and b1/b2 samples respectively .
if this statement is true then e1 predicts sdss composite a sample should contain many radio - loud sources and composite d very few .
we also note that sdss composite a is more sharply peaked than composite d , consistent with a stronger nlr component .
we applied our profile decomposition procedure to the sdss a and d composites kindly provided by g. richards and the results quantitatively verify the previous statement about their similarity to our lower redshift composites .
the profiles are also consistent with the presence of a growing ( from d to a ) unshifted blr component that further supports our two component fits to the bin a2/a3 composites .
richards et al .
( 2002 ) interpret the overall shift of as the result of `` missing '' flux on the red side of the profile .
we think that the body of evidence , within the e1 context , does not support this interpretation , e.g. the blue wing of civ@xmath121549 in extreme pop .
a sources does not resemble the blue wing of civ@xmath121549 in the broader pop .
b sources .
similarly the blue wing on sdss composite d is steeper than that for composite a. if we relate missing flux to obscuration then it would be surprising to find the strongest soft x - ray emission from sources with the most extreme blueshifted profiles since both should hardly be related .
sdss composites can not be used to help us decide whether there is a continuous `` main sequence '' in e1 or two distinct populations .
caution is needed in interpreting the sdss binning because we do not know what the sdss bin definition means in an e1 context , especially when ( wings ) may be strongly affected by feii emission .
if follows the line shift behavior of large relative shifts can occur in pop .
a because of blueshifts and in pop .
b because of possible redshifts .
if blueshifts ( relative to rest frame and not ) remain a population a phenomenon at high redshift then the sdss results suggest that the fraction of population a sources is larger at high redshift ( 75% vs. our estimate that 60% of quasars are population a ; marziani et al .
this would be consistent with our suggestion that pop .
a sources represent the `` seed '' population of young quasars in the universe ( sulentic et al . 2000a ; mathur , 2000 ; constantin & shields , 2003 ) .
we can also compare rl vs. rq properties using suitably constructed composites .
selection of rq and rl subsamples from a single bin will ensure that we compare quasars with similar physical and observational properties .
an extreme comparison of pop .
a vs. b will not be possible using a single bin .
we would have to compare an extreme pop . a bin with an extreme pop .
instead this will be a comparison between rl sources and rq sources with optical spectra indistinguishable from them .
this comparison offers hope to isolate subtle effects that might be direct manifestations of radio loud activity .
we must use a b bin for this comparison because that is where most rl sources are found .
we choose b1 which is the most densely populated and where rl and rq objects are almost evenly represented .
radio ( 6 cm ) and optical v - band fluxes were derived from data in the vcv01 catalog . a radio - index ( kellermann et al .
1989 ) @xmath46 was computed and all sources with log(@xmath47)@xmath02 are considered radio - loud ( see sulentic et al .
figure 5 shows the rl and rq composites , as well as rl
rq residuals , for both and .
we see evidence for differences : 1 ) most significantly , stronger ( ew@xmath35 ) , nlr emission in the rl sources and 2 ) a possible excess redshifted blr ( or vblr ) emission in rl sources ( this is also present in figure 7 a d composite difference of richards et al . 2002 ) .
the former is not a surprise ( see e.g. xu et al .
1999 ) while the latter , if real , may be an important clue . since we know that some bin b1 rq sources also show nlr emission result 1 ) reflects a lower limit for the mean nlr contribution in bin b1 rl sources .
a comparison shows also that the rl sources from this bin have on average stronger [ oiii ] and components ( typically by about 30% ) than the rq sources .
there is no doubt that rq sources exist with vblr components as strong as the strongest observed among the radio - louds ( e.g. sulentic et al .
2000c ) so further tests of the red residual are needed . in order to explore any luminosity correlation with composite spectra
we again restrict ourselves to bin b1 .
sources were separated into brighter and fainter groups using @xmath48 as the reference value .
this value separates the bin b1 sample into two approximately even groups .
composite spectra ( covering 10002000 ) and differences are shown in fig .
one can see a difference in the broad components of all the strongest high ionization lines ( , , and ) in the sense that the equivalent width is larger for the lower luminosity composite spectrum ( the baldwin effect ?
see 4.5 for a thorough discussion ) .
this difference was not seen in the rl vs. rq comparison .
fainter , lower ionization lines do not show any difference , furthermore they are weak enough that any difference would likely be undetectable .
the use of composite spectra raises an immediate question .
do median composites yield typical quasar spectra for each bin ? for example , is a symmetric profile in a median composite the result of equal numbers of blue and red asymmetric profiles ?
a visual inspection of our sample spectra shows , e.g. for b - bins , about 60% of sources with a more or less symmetric profile , while the remainder show 10 , 20 and 10% , respectively red , blue and indeterminate ( due to low s / n ) asymmetries .
this suggests that median composites represent the most typical quasar spectrum and that the changes between our spectral bins are real .
if our bin populations had @xmath25 100 objects then we would use weighted average composites , but medians appear to be a reasonable approach at this time .
it is also important to point out that bin sizes were chosen to include sources that are statistically similar within the average parameter measurement errors .
another important question is whether or not an ensemble of line profiles with a wide distribution of widths , strengths and asymmetries will yield a representative profile of the same type after combination .
we see that the many of the lines ( especially the strongest ) are better fit by a lorentzian whatever that means physically .
is it possible , for instance , that an ensemble of gaussian profiles will generate a lorentzian - like shape in composite spectra or the converse ?
this might lead to incorrect interpretation , and hence modeling , of line - forming regions . in order to test this possibility we ran a simulation by combining together about 100 artificial line profiles with a wide ( normal ) distribution of widths and intensities , resembling the real situation to some extent .
we performed simulation with both gaussian and lorentzian profiles , creating median and average composites from the individual pseudo - spectra .
the results suggest that profile type is preserved , i.e. a composite of gaussians or lorentzians will preserve their respective profile forms in a composite spectrum .
proper decomposition of emission lines in composite spectra along the e1 sequence offers insights into changes in blr geometry and kinematics .
several results from this and previous papers provide interesting clues .
asymmetric blueshifted hil emission is seen almost exclusively in population a sources . in the most extreme cases a blueshift
is also observed in the unusually weak features ( zamanov et al .
2002 ; marziani et al .
one simple interpretation that immediately suggests itself involves some kind of outflow from an optically thick emitting region into an optically thin nlr . in a scenario where the broad lines arise from an accretion disk
this could involve a high ionization disk outflow or wind .
it would be strongest in population a sources and would gradually weaken or even disappear as one proceeds along the e1 sequence toward population b. along the same sequence an unshifted / symmetric lorentzian hil component ( which may be analogous to the dominant lil component ) may be growing , in exactly the opposite sense of the blueshifted feature .
in fact some theoretical models predict lorentz - like profile wings
( e.g. penston et al . ,
1990 ; dumont & collin - souffrin 1990 ) .
the profile may become more gaussian in the population b region .
we have argued in a pre - e1 context that hil vs. lil line shifts appeared to be completely uncorrelated for rq sources and marginally correlated for the rl minority ( marziani et al .
composite spectra , binned in an e1 context , suggest that civ and may show analogous emission components except for the blueshifted hil feature seen only in pop .
a that are almost always rq agn .
a very broad and redshifted vblr component is only seen in population b sources and may be present in both hil and lil . since nlr - blr - vblr inflections are rarely seen , there may be less spatial / kinematic decoupling between these regions for civ than is seen for , where clear inflections are common .
that is far from proven because we also see lil profiles with little or no inflection between components .
inflections may be driven by kinematic and/or orientation .
there are some indications that the region producing emission is closer to the center that the region producing ( sulentic & marziani 1999 ) .
the vblr may be most simply interpreted as arising at the inner edge of the classical blr in analogy to vblr ( marziani et al .
1993 ; sulentic et al .
the lower redshift of the pop .
b. sources vblr may also indicate that a high ionization wind is still present in the vlbr since the component is still blueshifted with respect to the corresponding component .
so far there is no single picture capable of explaining the overall properties of the broad line emitting region .
currently fashionable models involving line emitting clouds near an optically thick accretion disk suffer a number of difficulties . if the motions of the clouds are @xmath49 then they will collide with the disk and will likely be destroyed on short time scales .
if one adopts an outflow / inflow model then the entire profile will show large red or blue - shifts because , the standard optically thick disk will obscure emission from the far side of a flow , while in reality large shifts in line profiles are rare . alternatively , emission lines originating from a geometrically thin , optically thick disk ( e.g. dumont & collin - souffrin 1990 ) will show extremely small fwhm when observed face - on .
this situation may be relevant for pop .
a sources but latest work suggests that the situation is different for rl and , by analogy , other pop .
b sources . the most jet - aligned superluminal rl sources , which , should have negligible disk rotation contribution to lil fwhm , show fwhm@xmath3830004000 ( rokaki et al . , 2003 ; sulentic et al .
it is however true that core - dominated ( cd ) and superluminal sources with @xmath50 show systematically narrower profiles than other ( frii ) rl agn ( sulentic et al .
cd sources are displaced towards lower fwhm ( ) in the optical e1 plane , which is consistent with orientation predictions . however , even these near pole - on sources lie well above many rq sources with similar , i.e. , they have larger fwhm ( ) for the same ( sulentic et al .
the results for radio loud sources suggest that we need a considerable velocity dispersion in the vertical direction to account for observed lil widths ( fwhm()@xmath51 ) .
the above results might be telling us that the bulk of line emission in rl and population b rq sources does not arise from a standard optically thick accretion disk especially when we observe fwhm lil / hil @xmath3 34000 in sources reasonably interpreted as almost disk face - on .
one candidate for the line emitting region involves the outer self - gravitating part of the disk ( collin & hur , 2001 ; bian & zhao , 2002 ) .
the disk is expected to become gravitationally unstable at some radius , where it will break up into rings or discrete self - gravitating clouds .
the vertical velocity dispersion of such clouds may be amplified by cloud - cloud encounters .
some studies suggest that the disk can become self - gravitating at a distance from the central black hole as small as 100200 , where is the schwarzschild radius ( see for instance collin & hur , 2001 ) .
a detailed theory for a self - gravitating disks is not yet available , however it is expected that the radius of the self - gravitating region will depend on the central black hole mass and eddington ratio ( and two related and very important parameters : viscosity , @xmath52 , and disk opacity ) .
there is no consensus on the correct dependence of on parameters like @xmath53 or eddington ratio @xmath54 ( here we assume that the accretion efficiency ( @xmath55 ) does not change with luminosity , otherwise @xmath56 , where @xmath57 . a very simple approach for exploring this idea involves application of the toomre stability criterion to a standard shakura sunyaev disk ( shakura & sunyaev 1973 ) .
this will give an estimate of the radius where the stability parameter ( q ) becomes approximately 1 .
different models of the self - graviting region often predict different dependence on the accretion parameters because of different initial assumptions about the outer regions of the disk ( i.e. opacity , pressure , etc ) .
bian & zhao ( 2003 ) obtain in a recent work : @xmath58 based on standard accretion disk solutions assuming dominant kramers opacity .
they assume that it is equal to the radius of the blr @xmath59 ( here and below is expressed in ) .
collin & hur ( 2001 ) adopt a numerical approach with realistic opacities and get a complex dependence roughly approximated by @xmath60 for the region log(@xmath53)=7 9 , m@xmath61 .
following these authors , for a bh mass m=10@xmath62 and eddington ratio @xmath381 , @xmath63 , while for m=10@xmath64 and [email protected] , is less than about 500 . on the other hand we can consider `` superluminal '' radio - loud sources with the narrowest rl line profiles : fwhm()@xmath41 3000 , and the narrowest pop . a sources , with fwhm()@xmath41 1000 , also likely to be observed at @xmath65
assuming orbital motion with keplerian angular velocity ( @xmath66 ) , one can write for these apparently face - on sources : @xmath67 where @xmath68 is the vertical velocity dispersion , assumed to be proportional to the @xmath49 velocity by a factor of @xmath69 .
a reasonable guess for @xmath69 is about 0.10.2 .
one can easily show that fwhm of 3000 implies a radius of about 5000 while 1000 implies 500 , values quite close to the size of the self - graviting region assumed above .
even if the agreement is perhaps fortuitous , and the exact dependence is far from certain , it is reasonable to conclude that ( @xmath3 @xmath70 ) may be smaller in pop b sources by a factor @xmath71 , and that this may leave a very `` small '' emitting surface for any standard optically - thick geometrically - thin disk component .
part of the line profile may be produced in the disk if it is illuminated by a geometrically thick inner region or by disk warping ( e.g. bachev , 1999 ) and if it is not extremely hot ( which would completely ionize the lower - ionization species ) .
on the other hand , the presence of a blue - shifted component can be most simply explained as the signature of an outflow from an optically thick disk .
the disk - wind scenario can produce blue - shifted high - ionization lines as already suggested in several works ( e. g. murray & chiang 1997 ; bottorff et al .
if our interpretation is correct this implies that the wind strength increases toward the most extreme population a sources ( bins a2/a3 with strongest ) . population b sources may have a much lower , or may not develop , a wind component . following these considerations we propose a simple blr scenario
( see also fig . 1 from collin & hur , 2001 ) : * a `` hot '' innermost accretion region , located inside some transition radius @xmath72 ) .
our data leave the structure of this region open to speculation , although we may reasonably assume that the continuum region ( or part of a geometrically thick , optically thin advection - dominated accretion flow ( adaf ; see narayan , mahadevan & quataert 1998 for an exhaustive review ) can heat the outer line - emitting region . * an optically thick , geometrically thin disk whose extent decreases by a factor @xmath310 from pop . a to pop b. the heated disk produces a wind that is responsible for the blueshifted component .
the outer region of the disk may emit and other lils as well in pop . a sources .
the disk may become almost negligible in pop b sources because may become as small as 100 200 ( e.g. bachev & strigachev , 2004 ) . * a self - gravitating and fragmented zone where line emission might arise in dense gas clumps .
this region may be the _
entire blr _ for pop .
b sources while it may contribute mainly to lil emission in pop . a objects . within the context of this model + e1
we can eventually explain several observational results from this paper as well as from previous works : * the different vertical velocity dispersion in the self - gravitating zone for pop .
b ( closer ) and pop . a ( farther ) sources accounts for different line widths at @xmath73 ; * after subtraction of an unshifted / symmetric lorentzian component the blueshifted residual shows reasonable persistency in shape the intensity may change but not the overall shape .
the blue component becomes marginal at bin a1 .
this behavior is consistent with the idea that the blue component comes from a disk - generated wind , where the kinematics ( velocity field ) is strongly affected by the detailed properties ( e.g. magnetic field structure and terminal wind velocity ) .
the overall shape of a line arising in a disk wind will be less dependent on the radial distance where the wind originates .
in other words , our model provides an easy explanation why population a sources tend to show decoupled hil
lil components while pop .
b sources show more similar and profiles . in the most extreme cases , there are sources where the entire profile resembles the blue component alone ( e.g. i zw 1 , ton 28 ; sulentic et al .
2000b ) . + leighly ( 2004 ) presented an interesting explanation of the apparent anticorrelation between the strengths of the central and the blue - shifted ( wind ) components .
the wind that gives rise to the blue - shifted component might filter the euv ionizing radiation that would otherwise reach the outer , self - graviting parts ( where the unshifted line component might originates )
. the wind can not however filter effectively the hard x - rays that are reprocessed in lil emission in the same region .
+ an alternative approach is to ascribe the entire profile of civ in a2/a3 bins to an outflow ( wind ) . in such a case the redward emission is seen either because the disk is ( partially ) transparent to civ photons or because of higher inclinations , resulting in significant projection of the keplerian / radial component of the wind velocity onto the line of sight .
both possibilities seem somewhat unlikely , at least for these extreme bins .
* there are trends between @xmath28 and fwhm ( e.g. tytler & fan 1992 ) .
a sources can be explained with a split blr : inner hil and an outer lil emitting regions .
since we expect that gas motion will retain its circular velocity in the wind and will acquire an upward and outward velocity component , the velocity dispersion in the wind will be higher than for a keplerian velocity component in the disk .
lil emission from the outer , fragmented region will therefore be narrower . in pop b sources we expect a `` normal '' blr stratification : hil emitted closer to the central continuum source than lil .
in general , we expect that both hil and lil will be emitted rather close to the central continuum source , accounting for the rather high ionization spectrum .
this outline requires testing with photoionization models .
reverberation mapping indicates continuum - response times of consistent with the one of ( at least in ngc 5548 ; clavel et al .
this result suggests that is primarily produced in the low - ionization emitting part of the blr . under the assumption of nearly constant [ si / c ] abundance ratio , the behavior of the / intensity ratio along the e1 sequence
can then be understood in terms of increasing electron density toward type a3 ( m01 ) in the lil - emitting blr .
the critical of the line is @xmath74 3 , much lower than that of the line , @xmath75 3 .
therefore , for @xmath76 3 , the line should be more and more collisionally suppressed with increasing , while the line should become favored .
we also observe a consistent increase in the strength of the line , with the intensity ratio / reaching @xmath77 for spectral type a3 ( @xmath75 3 ) .
the line is expected to strengthen with increasing electron density ( baldwin et al .
1996 ; korista et al .
however , as pointed out by the referee , chemical abundance may not be constant along the e1 sequence .
we do not have information on the [ si / c ] and [ al / c ] abundance ratios .
intensity ratios involving nitrogen lines suggests that super - solar metallicity is appropriate for extreme population a ( i.e. , a3 ) sources . from table 2
, we have /@xmath78 and /@xmath23 for a3 , while for a1 /@xmath79 , and /@xmath21.5 .
it is intriguing that we are observing again , for spectral type a3 , line ratios that are more frequently found in high-@xmath19 quasars ( hamann & ferland 1999 ) .
highly supersolar nitrogen enrichment may have been produced by star formation with a top - loaded initial mass function ( hamann & ferland 1992 ; 4.5 ) .
it is possible that changes of line ratios from pop .
b to pop . a and along the a - spectral type are due to _ both _ an -increase and a metallicity increase . quantifying these effects
is a challenge for further work .
comparison of rl and rq line profiles ( 3.2 ) in the e1 context should begin with a comparison of population a and b sources .
the differences that we find between pop . a and b hil civ composites are even larger than for lil .
we can say with some confidence that sources showing low ew , blueshifted and blue asymmetric have low probability ( p@xmath401% ) to be radio - loud .
pop.b sources have a reasonably high probability to be rl ( p@xmath350% ; marziani et al .
2003b ) . since not all pop .
b sources are rl we focus ( 3.3 ) on a particular bin ( b1 ) within population b where we can compare properties for rq and rl agn that appear to be spectroscopically indistinguishable .
our rl - rq difference spectrum is shown in figure 5 .
we see evidence for a significantly stronger nlr component in pop .
b rl sources .
this might be explained if a significant part of the narrow line emission is produced in interactions between a radio jet and the ambient medium , forming presumably a narrow - line - bright cocoon around the jet ( e. g. , steffen et al .
1997 ; axon et al .
1998 ) . aside from a possible excess of redshifted broad line gas in the rl sources , pop .
a rl and rq line profiles look very similar as was also found for ( marziani et al .
. the vblr component , which might be related to this redshifted gas , is seen in both rl and rq pop . b agn ( sulentic et al .
2000c ) and , therefore , is not a rl signature .
one is tempted to consider host galaxy morphology as a major difference between the rl and rq sources in bin b1 .
rl sources tend to be hosted in ellipticals ( pagani et al .
2003 ) while a typical pop .
b rq source involves a classical seyfert 1 like ngc 5548 , hosted in a spiral . perhaps the pop .
b emission line spectrum indicates that rl activity is possible as far as physical conditions near the bh are concerned but host morphology dictates whether it happens or not . if it were that simple however we might expect to see many frustrated rl sources involving trapped core emission ( unless it is self quenching under that circumstance ) .
the bin b1 results emphasize the fact that pop .
b rq and rl sources are spectroscopically indistinguishable as far as _ broad _ lines are concerned .
this may be telling us that radio - loudness is not directly dependent on , or strongly influencing , blr structure ( and therefore the structure of the accretion disk , at least at @xmath80 ) .
this makes a circumstantial case for the role of black hole spin , which is unlikely to have any influence on the accretion disk structure outside 10100 where the uv line emission likely originates , at least if the bardeen - petterson effect ( bardeen & petterson , 1975 ) is not in play .
the black hole spin can , however , play a significant role in jet production ( e.g. blandford & znajek , 1977 ) .
the fact that the majority of rl sources occupy b - bins may be associated with a change of the innermost accretion disk structure ( from a thin disk to an adaf ? , rzanska & czerny , 2000 , and references therein ) that is probably triggered by the lower accretion rate ( or by some function of the black hole mass and the accretion rate ) .
if true , the accretion disk structure change ( from thin disk to adaf ) will be a necessary but not sufficient condition for radio - loudness ( i.e. the ultimate trigger will be the black hole spin , subject to the environmental caveat mentioned above ) . in this scenario only b - bin objects which contain a rapidly spinning black hole would be able to produce powerful jets .
measurement of broad line ew s reveals both a trend and the lack of a trend ( see also 3.1 ) : * , and show a systematic ew decrease by factors of 2 4 as one proceeds from bin b1@xmath27 to a3 .
table 2 shows the same trend for other lines but caution is needed when interpreting blended lines and data for lines with ew@xmath415that have estimated uncertainties @xmath2530% at the 1-@xmath24confidence level . *
the line shows no trend with ew ( ) @xmath2 2025 in all spectral bins .
these results essentially describe the so - called `` baldwin effect '' ( hereafter be ; originally defined as an anti - correlation between rest frame ew and source luminosity at 1450 : baldwin 1977 ) .
it is well known that does not follow the be trend ( see sulentic et al .
2000a for references before 1999 ; dietrich et al . 2002
; green et al . 2001 ; croom et al . 2002 ) . while the original be showed a tight correlation involving a few tens of sources , more recent work ( dietrich et al .
2002 ; warner et al .
2003 ) show it to be a very loose correlation that only becomes significant over a wide range in continuum luminosity @xmath81=4248 ( e.g. kinney et al .
1990 ) . considering the large dispersion
it is not surprising that several studies using small samples found no evidence for a be ( see sulentic et al .
2000a for numerical simulations and a thorough discussion of the issue ) .
the be - like trend found between our spectral bins does _ not _ represent a dependence on luminosity .
the e1 sequence is independent of source luminosity ( sulentic et al .
2000b ) and , is more likely governed by the eddington ratio ( m01 ; marziani et al .
in fact pop .
b spectral bins contain more high luminosity sources than pop .
a because rl in our sample are overluminous , and overrepresented , relative to rq sources ( see also boroson & green 1992 ) .
our be trend is also _ not _ bh mass - dependent ( e.g. , warner et al .
2003 ) : the largest bh masses are again found in population b ( m03 ) .
if we compare the average m values for our bins ( from m03b ) with ew ( ) measures from our composite spectra we obtain the following approximate relation : @xmath82 ew ( ) @xmath83 excluding bin b1@xmath27 .
our models predict that b1@xmath27 sources are simply more inclined b1 sources ( m01 ) . including b1@xmath27 yields @xmath84 ew()@xmath85 using @xmath86 bin averages from m03b or @xmath87 ew()@xmath88if bin averages are taken from this paper (
l / m is measured in solar units ) .
f we consider all sources ( rq and rl ) a robust fitting method yields @xmath84 ew()@xmath89 , which becomes @xmath84 ew ( ) @xmath90 for rq sources alone .
parameter dispersion in the baldwin effect ( from ew ( ) @xmath2 20 200 ) is well reproduced by sources spanning the range @xmath3 0.02 1 in @xmath91regardless of the fitting solution details .
highest @xmath91 sources show the lowest ew ( ) and it is a well - established fact that nlsy1s show the lowest ew ( ) ( marziani et al . 1996 ; rogriguez - pascual et al . 1997 ; sulentic et al .
2000a ) and are likely to radiate close to the eddington limit .
many known nlsy1s are relatively low redshift / luminosity sources that tend to blur standard be correlations at the low luminosity end ( see e.g. , brotherton & francis 1999 ) because they show ew ( ) similar to the high redshift quasars ( sulentic et al .
sulentic et al . (
2000a ) suggested that the be might be caused by preferential ( probably driven by selection effects and/or intrinsic evolution ) detection of large @xmath91 sources at high redshift . in a flux
limited sample , where @xmath92 and @xmath19are highly correlated , an `` evolutionary '' ( redshift - dependent ) baldwin effect is expected and is indeed found in the large bright quasar survey ( green et al .
2001 ) . given a sample that is complete in terms of @xmath91 ,
the be may survive with even larger scatter or disappear depending on the relative importance of evolutionary and selection effects ( see dietrich et al .
deeper surveys may detect larger ew ( ) sources at high redshift : see for example the distribution of ew ( ) at @xmath93 in constantin et al .
intriguingly , they detect several sources with ew()@xmath2550 at @xmath94 .
interpretation of the be has been highly controversial with suggestions encompassing orientation effects ( netzer 1985 ) , selection effects ( sulentic et al . 2000a , and references therein ) , black hole mass ( wandel 1999 ; warner et al . 2003 ) , eddington ratio ( sulentic et al .
2000a , shang et al . 2003 ; baskin & laor 2004 ) . a continuum softening with increasing luminosity ( wandel 1999 ) is a less model - dependent explanation .
warner et al . ( 2003 ) suggest that the driving parameter may be the central black hole mass .
these authors computed the bh mass using a modified kaspi et al .
( 2000 ) relation , employing fwhm instead of fwhm and assuming virial motions .
this _ does not demonstrate _ that mass is driving the be because it is assumed to be @xmath95 , and this assumption makes the computation in part circular .
furthermore , we question the validity of estimating bh masses using ( e.g. vestergaard 2002 and warner et al . 2003 ; 2004 ) . while shows average broad line profiles that can be argued to arise in virialized clouds ,
is blue asymmetric and blue shifted in about 60% of sources . in many cases
the shift is significant and in some virtually all of the line flux
is displaced to the blue side of the local rest frame .
it is not at all clear that the clouds ( or wind ? ) producing the hil emission can be assumed to be virialized .
finally failure to correct for nlr emission in the civ profile will contribute to the dispersion in bh mass that might be present .
nlr emission tends to decrease the dispersion blr fwhm measures ( fwhm measures for pop .
b become more like pop a ) .
figure 5 shows that nlr uncorrected fwhm civ measures will underestimate bh mass and overestimate @xmath96 values for rl sources .
our work shows that a exists in , at least , pop .
b sources and that a significant part ( if not all ) of emission in bins a2/a3 arises in a wind
. the terminal wind velocity could be only indirectly related to keplerian ( i.e. virialized ) motions .
the overall velocity dispersion is even less likely to be related to virial motion in near face - on sources , where the accretion disk likely obscures the receding part of the flow .
the line profile in such cases is almost fully blueshifted with respect to the rest frame . while any correlation with mass will be little affected over several orders of magnitude ( warner et al .
2003 ) , any more fundamental correlation with @xmath91 may be entirely washed out ( @xmath91 range is a factor @xmath4150 ; woo & urry 2002 ) . comparing fwhm h@xmath4 and civ in different spectral bins ( pop .
b with and without subtraction ) suggests that errors could be at least a factor @xmath310 for some objects ( since , for example , the fwhm ratio is @xmath2 3 in izw1 ; marziani et al . 1996 ; 2003b ) .
however , a shallow trend between ew ( ) and @xmath97 is noted in a large sample of quasars of redshift @xmath98 ( warner et al .
2004 ) , which seems to support our results for @xmath990.8 .
another circumstantial element supporting a role for the accretion rate @xmath91 in the be involves the constancy of ew ( ) and the increasing intensity ratio / along the sequence from b1@xmath27 to a3 .
the large values of ew ( ) for all spectral types are suggestive of solar or supersolar metallicity ( turner et al .
2003 ; bentz & osmer 2004 ) .
it is not unreasonable to suppose that n may have been enhanced by vigorous formation of massive stars that burn h via the cno cycle in the agn circumnuclear regions .
the largest enrichment may have occurred in relatively young or rejuvenated quasars radiating at large accretion rates as nlsy1s are thought to be ( several nlsy1s hosts may be merging dwarf or barred galaxies ; krongold et al .
2001 ; crenshaw et al .
we computed and analyzed composite uv spectra for 5 different spectral bins along the main sequence of the e1 optical parameter plane .
data were analyzed following paper i , which considered similar composite spectra for the optical region involving , our representative lil .
we measured the fwhm and equivalent width of a number of emission lines focussing on , our representative hil .
our main results include : * we decompose hil into possible nlr , blr and vblr components many of which have lil analogs .
we consider them to be distinct line emitting regions .
type 2 agn show only nlr lil and hil emission .
type 1 agn are know with pure blr , pure vblr and , most often , composite blr+vblr ( with and without nlr ) spectra if they belong to population b. a strong asymmetric blueshifted blr hil component is unique to population a ( bins a2/a3 ) sources .
a reasonably strong and redshifted vblr is unique to hil and lil in population b ( bins b1 and b1@xmath27 ) sources .
* we infer a trend along the e1 sequence using the ratio of the semi - forbidden lines and . under the assumption of a nearly constant chemical abundance along e1
, we find that density increases by about an order of magnitude between bins b1@xmath27 and a3 . however , there are indications that both density and metallicity may increase from extreme b - bins to extreme a - bins .
* we compare sub - samples of rl and rq sources in bin b1 .
we find an excess of nlr and redshifted broad line emission in the rl subsample
. the former represents virtual proof that nlr line emission is present in the spectra of many type 1 agn .
* we advance the following empirically driven model for the structure of the accretion flow : 1 .
an inner hot optically thin ( possibly adaf ) disk , located inside @xmath100 .
2 . a standard thin disk , between @xmath101 and @xmath102 , producing most of the optical continuum .
this is the place where a wind may originate and produce the blue - shifted component .
an outer self - gravitating region at @xmath103 , producing most of the broad - line emission .
+ @xmath101 and @xmath102 are functions of the accretion parameters
. a n eventual discontinuity in their behavior as functions of m and m may account for the existence of different quasar populations ( e.g. pop .
* we suggest that the eddington ratio is the primary driver of the baldwin effect rather than source luminosity .
this suggestion is motivated by the robust be that we find in e1 which is luminosity independent .
this explains why high redshift sources might preferentially show low ew and blueshifted if pop .
a sources are the `` seed '' ( i.e. youngest high accreting lowest mass bh ) agn .
this interpretation is supported by comparison of our low redshift sample with first results from higher redshift sdss quasars .
the latter show an increase in the fraction of sources with a blueshifted hil a defining characteristic of pop . a agn .
rb and jws acknowledge the kind hospitality and the financial support from the osservatorio astronomico di padova where much of this work has been done .
we thank dr .
g. richards for providing his sdss composite spectra in an electronic form .
we are indebted to an anonymous referee for his / her constructive criticism .
axon , d. j. , marconi , a. , capetti , a. , maccetto , f. d. , schreier , e. , & robinson , a. 1998 , , 496 , l75 bachev r. , 1999 , a&a , 348 , 71 bachev r. , strigachev a. , 2004 , an , 325 , no . 4 , 1 baskin , a. & laor , a. 2004 , , 350 , l31 baldwin j. a. , 1977 , apj , 214 , 679 baldwin j. a. , ferland g. j. , korista k. t , et al . , 1996 , apj , 461 , 664 bardeen j.m . , petterson j.a . ,
1975 , apj , 195 , l65 bentz m. c. , osmer p. s. , 2004 , aj 127 , 576 bian w. , zhao y. , 2002 , a&a , 395 , 465 blandford r. d. , znajek r. l. , 1977 , mnras , 179 , 433 bottorff m. , korista k. t. , shlosman i. , blandford r. d. , 1997 , apj , 479 , 200 boroson t.a . , green r.f .
, 1992 , apjs , 80 , 109 boroson , t. a. 2002 , , 565 , 78 brotherton m. s. , francis p. j. , 1999 , in `` quasars and cosmology '' , asp conference series 162 , eds . g. ferland and j. baldwin .
clavel , j. , et al .
1991 , , 366 , 64 collin s. , hur j .- m . , 2001 , a&a , 372 , 50 constantin a. , shields j. c. , hamann f. , et al . , 2002 , apj , 565 , 50 constantin a. , shields u. c. , 2003 , pasp , 115 , 592 corbin m. r. , boroson t. a. , 1996 , apjs , 107 , 69 corbin , m. r. , 1997 , apj , 485 , 517 crenshaw , d. m. , kraemer , s. b. , & gabel , j. r. 2003 , , 126 , 1690 croom s. m. , rhook k. , corbett e. a. , et al . , 2002 , mnras , 337 , 275 czerny , b. , nikolajuk , m. , r ' o. za ' nska , a. , dumont , a .- m . , loska , z. , & zycki , p. t. 2003 , , 412 , 317 czerny b. , rozanska a. , kuraszkiewicz j. , 2004 , astro - ph/0403507 dietrich , m. , hamann , f. , shields , j. c. , constantin , a. , vestergaard , m. , chaffee , f. , foltz , c. b. , & junkkarinen , v. t. 2002 , , 581 , 912 dumont a. m. , collin - souffrin s. , 1990 , a&a , 229 , 313 francis , p. j. , hewett , p. c. , foltz , c. b. , chaffee , f. h. , weymann , r. j. , & morris , s. l. 1991 , , 373 , 465 gammie , c. f. , shapiro , s. l. , & mckinney , j. c. 2004 , , 602 , 312 green p. j. , forster k. , kuraszkiewicz j , 2001 , apj , 556 , 727 hamann , f. & ferland , g. 1992 , , 391 , l53 hamann , f. & ferland , g. 1999 , , 37 , 487 horne , k. , peterson , b. m. , collier , s. j. , & netzer , h. 2004 , , 116 , 465 jackson n. , browne , i. w. a. , 1991 , mnras , 250 , 422 janiuk , a. , .
zycki , p. t. , & czerny , b. 2000 , , 314 , 364 kaspi s , smith p. s. , netzer h. , et al . , 2000 , apj , 533 , 631 kellermann k. i. , sramek r. , schmidt m. , shaffer d. b. , green r. , 1989 , aj , 98 , 1195 kerber , f. , rosa , m. , 2000 , st - ecf newsletter no .
27 , 4 kinney a. l. , rivolo a. r. , koratkar a. p. , 1990
, apj , 357 , 338 komossa , s. , meerschweinchen , j. , 2000 , a&a , 354 , 411 korista k. , baldwin j. , ferland g. , verner d. , 1997 , apjs , 108 , 401 korista k. t. , goad m. r. , 2004 , apj , 606 , 749 krongold y. , dultzin - hacyan d. , marziani p. , 2001
, aj , 121 , 702 kuraszkiewicz , j. k. , green , p. j. , crenshaw , d. m. , dunn , j. , forster , k. , vestergaard , m. , & aldcroft , t. l. 2004 , , 150 , 165 leighly k. , 2004 , astro - ph/0402452 maccarone , t. j. , gallo , e. , & fender , r. 2003 , , 345 , l19 marziani p. , sulentic j. w. , calvani m. , perez e. , moles m. , penston m. v. , 1993 , apj , 410 , 56 marziani , p. , sulentic , j. w. , dultzin - hacyan , d. , calvani , m. , moles , m. , 1996 , apjs , 104 , 37 marziani p. , sulentic j.w .
, zwitter t. , dultzin - hacyan d. , calvani m. , 2001 , apj , 558 , 553 ( m01 ) marziani p. , sulentic j. w. , zamanov r. , calvani m. , dultzin - hacyan d. , bachev r. , zwitter t. , 2003a , apjs , 145 , 199 ( m03 ) marziani p. , zamanov r. , sulentic j. w. , calvani m. , 2003b , mnras , 345 , 1133 marziani , p. , zamanov , r. , sulentic , j. w. , calvani , m. , & dultzin - hacyan , d. 2003c , memorie della societa astronomica italiana , 74 , 492 mathur , s. , 2000 , mnras , 314 , l17 mirabel , i. f. 2004 , arxiv astrophysics e - prints , astro - ph/0404156 murray n. , chiang j. , 1997 , apj , 474 , 91 narayan , r. , mahadevan , r. , & quataert , e. 1998 , theory of black hole accretion disks , edited by marek a. abramowicz , gunnlaugur bjornsson , and james e. pringle .
cambridge university press , 148 netzer h. , 1985 , mnras , 216 , 63 nicastro , f. 2000 , , 530 , l65 pagani , c. , falomo , r. , & treves , a. 2003 , , 596 , 830 penston m. v. , croft s. , basu d. , fuller n. , 1990 , mnras , 244 , 357 peterson , b. m. 1993 , , 105 , 247 peterson , b. m. , 1997 , an introduction to active galactic nuclei , cambridge u press pounds , k. a. , done , c. , & osborne , j. p. 1995
, , 277 , l5 richards g. t. , vanden berk d. e. , reichard t. a. , hall p. b. , schneider d. p. , subbarao m. , thakar a. r. , york d. g. , 2002 , , 124 , 1 rodriguez - pascual p. m. , mas - hesse j. m. , santos - lleo m. , 1997 , a&a , 327 , 72 rokaki e. , lawrence a. , economou f. , mastichiadis a. , 2003 , mnras , 340 , 1298 rzanska a. , czerny b. , 2000 , a&a , 360 , 117 schlegel d. j. , finkbeiner d. p. , davis m. , 1998 , apj , 500 , 525 schneider , d. p. , et al .
2002 , , 123 , 567 shakura n. i. , sunyaev r. a. , 1973 , a&a , 24 , 337 shang , z. , wills , b. j. , robinson , e. l. , wills , d. , laor , a. , xie , b. , & yuan , j. 2003 , , 586 , 52 snedden s. a. , gaskell , c. m. , 2004 , astro - ph/0403174 steffen , w. , gomez , j. l. , raga , a. c. , & williams , r. j. r. 1997 , , 491 , l73 sulentic , j. w. , marziani , p. , 1999 ,
apj , 518 , l9 sulentic , j. w. , calvani , m. , & marziani , p. 2001
, the messenger , 104 , 25 sulentic j.w . ,
marziani p. , dultzin - hacyan d. , 2000a , ar a&a , 38 , 521 sulentic j. w. , zwitter t. , marziani p. , dultzin - hacyan d. , 2000b , apj , 536 , l5 .
sulentic j. w. , marziani p. , zwitter t. , dultzin - hacyan d. , calvani m. 2000c , , 545 , l15 sulentic j. w. , marziani p. , zamanov r. , bachev r. , calvani m. , dultzin - hacyan d. , 2002 , , 566 , l71 ( paper i ) sulentic , j. w. , stirpe , g. m. , marziani , p. , zamanov , r. , calvani , m. , & braito , v. 2004 , arxiv astrophysics e - prints , astro - ph/0405279 sulentic j. w. , zamfir s. , marziani p. , bachev r. , calvani m. , dultzin - hacyan d. , 2003 , apj , 597 , l17 turner t. j. , kraemer s. b. , mushotzky r. f. , et al .
, 2003 , apj , 594 , 128 tytler d. , fan x - m , 1992 , apjs , 79 , 1 vanden berk , d. e. , et al .
2001 , , 122 , 549 vron - cetty , m .-
, vron , p. 2001
, a&a , 374 , 92 ( vcv01 ) vron - cetty m .- p . , vron p. , goncalves a. c. , 2001 , a&a , 372 , 730 wandel a. , 1999 , apj , 527 , 649 warner c. , hamann f. , dietrich m. , 2003 , apj , 596 , 72 wills b. j. , thompson k. l. , han m. , et al . , 1995 ,
apj , 447 , 139 woo j .- h . , urry c. m. , 2002 , apj , 579 , 530 xu c. , livio m. , baum s. , 1999 , aj , 118 , 1169 zamanov r. , & marziani p. , 2002
, apj , 571 , l77 zamanov r. , marziani p. , sulentic j. w. , calvani m. , dultzin - hacyan d. , bachev r. , 2002 , apj , 567 , l9 ccccccccc a3 & @xmath404000 & @xmath01.0 & 10 ( 2 ) & 39 & 23.1 ( 2.1 ) & 0.20 ( 0.20 ) & 20 ( 5 ) & 50 ( 54 ) + a2 & @xmath404000 & 0.5 - 1.0 & 14 ( 1 ) & 35 & 24.4 ( 1.6 ) & 0.26 ( 0.22 ) & 18 ( 9 ) & 67 ( 63 ) + a1 & @xmath404000 & @xmath400.5 & 40 ( 12 ) & 74 & 24.3 ( 2.2 ) & 0.26 ( 0.24 ) & 44 ( 25 ) & 95 ( 89 ) + b1 & 4000 - 8000 & @xmath400.5 & 54 ( 28 ) & 87 & 25.1 ( 1.7 ) & 0.36 ( 0.24 ) & 46 ( 26 ) & 98 ( 95 ) + b1@xmath27 & @xmath08000 & @xmath400.5 & 21 ( 17 ) & 52 & 25.4 ( 1.4 ) & 0.44 ( 0.22 ) & 60 ( 34 ) & 99 ( 98 ) + ovi 1034 & - & - & - & - & 17 ( 4 ) & 3770 ( 500)&21 ( 6)&4350 ( 600)&18 ( 4 ) & 5510 ( 700 ) + ciii 1176 & - & - & 4 ( 3 ) & 3600 ( 2000)&1 ( 1 ) & 5100 ( 2000)&4 ( 3 ) & 5350 ( 2500 ) & - & - + ly@xmath521216&48 ( 15)&2960 ( 700 ) & 92 ( 21)&2960 ( 500)&101(25)&3450 ( 500)&116(21)&3700 ( 400)&93 ( 18)&5420 ( 400 ) + nv 1240 & 26 ( 9)&3870 ( 700 ) & 21 ( 7)&4110 ( 700)&17 ( 6 ) & 3380 ( 600)&28 ( 9)&5080 ( 800)&25 ( 6 ) & 7250 ( 900 ) + siii 1264 & 3 ( 2)&2250 ( 1000 ) & 2 ( 1 ) & 1420 ( 800 ) & - & - & - & - & - & - + siii 1306 & 6 ( 2)&3530 ( 500 ) & 5 ( 1 ) & 2290 ( 300)&3 ( 1 ) & 2290 ( 300)&4 ( 1 ) & 3670 ( 500 ) & 3 ( 1 ) & 4820 ( 1000 ) + cii 1335 & 2 ( 1)&1970 ( 500 ) & 3 ( 1 ) & 1680 ( 300)&1 ( 1 ) & 2240 ( 300)&1 ( 1 ) & 2240 ( 600 ) & 1 ( 1 ) & 3590 ( 1500 ) + siiv 1398 & 16 ( 3)&4500 ( 400 ) & 16 ( 3)&3860 ( 400)&13 ( 3 ) & 4070 ( 400)&11 ( 2)&3640 ( 400 ) & 12 ( 2)&6000 ( 400 ) + niv ] 1486 & - & - & - & - & 2 ( 1 ) & 2610 ( 800)&1 ( 1 ) & 3020 ( 800 ) & - & - + civ 1549 & 23 ( 4)&4160 ( 400 ) & 35 ( 9)&3290 ( 400)&69 ( 16)&3480 ( 400)&83 ( 16)&4260 ( 400)&79 ( 10)&6190 ( 400 ) + heii 1640 & 8 ( 2)&5850 ( 700 ) & 10 ( 2)&4750 ( 500)&11 ( 2 ) & 4570 ( 400)&11 ( 2)&4930 ( 500 ) & 4 ( 1 ) & 4750 ( 600 ) + oiii ] 1663 & 2 ( 1)&2880 ( 800 ) & 5 ( 2 ) & 4140 ( 700)&4 ( 1 ) & 3060 ( 500)&7 ( 2 ) & 4320 ( 700 ) & 5 ( 2 ) & 3600 ( 700 ) + niii ] 1750 & 4 ( 3 ) & 2050 ( 1000 ) & - & - & - & - & 4 ( 3 ) & 4410 ( 1500 ) & - & - + aliii 1860 & 8 ( 4 ) & 4670 ( 1200 ) & 5 ( 3 ) & 3860 ( 1200)&2 ( 1 ) & 4180 ( 1000)&3 ( 2 ) & 4700 ( 2000 ) & 3 ( 2)&5700 ( 2500 ) + siiii ] 1892 & 7 ( 3 ) & 1900 ( 800 ) & 8 ( 2 ) & 1740 ( 700)&6 ( 2 ) & 2530 ( 600)&5 ( 3 ) & 2530 ( 1000 ) & 3 ( 2)&3010 ( 1500 ) + ciii ] 1909 & 10 ( 4)&2350 ( 800 ) & 15 ( 4)&1720 ( 700)&17 ( 4 ) & 2510 ( 500)&20 ( 7)&3140 ( 700 ) & 18 ( 9)&3920 ( 1200 ) + mgii 2798 & 12 ( 8)&2030 ( 700 ) & 14 ( 4)&1500 ( 700)&16 ( 2 ) & 1600 ( 600)&40 ( 8)&3960 ( 600 ) & 44 ( 8)&6540 ( 600 ) + & 22 ( 10)&3000 ( - ) & 16 ( 8)&3000 ( -)&9 ( 5 ) & 3000 ( - ) & 7 ( 5 ) & 5000 ( - ) & 7 ( 5)&9000 ( - ) + a3 & lor & 1547.3 & 2220 & 50 & & blue & 1532 & 4800 & 50 + a2 & lor & 1547.8 & 2600 & 76 & & blue & 1534 & 3900 & 24 + a1 & lor & 1547.7 & 3580 & 100 & & - & & & + b1 & gau & 1548.3 & 4170 & 42 & & gau & 1555.8 & 15590 & 58 + b1@xmath27 & gau & 1547.8 & 4780 & 28 & & gau & 1552.6 & 17480 & 72 + | we present composite uv spectra for low redshift type 1 agn binned to exploit the information content of the eigenvector 1 ( e1 ) parameter space .
composite spectra show high enough s / n and spectral resolution to permit a decomposition of the line profile one of the strongest high - ionization lines ( hil ) and fundamental in defining e1 space .
the simplest decomposition into narrow ( nlr ) , broad ( blr ) and very broad ( vblr ) components suggests that different components have an analog in with two major exceptions .
vblr emission is seen _ only _ in population b ( fwhm()@xmath04000 ) sources .
a blue shifted / asymmetric blr component is seen _ only _ in population a ( fwhm()@xmath14000 ) hil such as .
the blueshifted component is thought to arise in a high ionization wind or outflow .
our analysis suggests that such a wind can only be produced in population a ( almost all radio - quiet ) sources where the accretion rate is relatively high .
we propose a model to account for several differences between low- and high - ionization line profiles .
part of the broad line emission is attributed to a self - gravitating / fragmented region in an accretion disk .
an inner optically thick geometrically thin region of the flow may give rise to a wind / outflow and produce the blueshifted hil spectrum in pop . a sources .
the fragmented region may produce all or most of the broad line emission in population b , which contains radio - quiet and the majority of radio - loud quasars .
comparison between _ broad _ uv lines in radio - loud ( rl ) and radio - quiet ( rq ) sources in a single well populated e1 parameter space bin ( b1 ) shows few significant differences .
clear evidence is found for a significant nlr civ component in most radio - loud sources .
the blr / vblr similarity in bin b1 provides circumstantial evidence in favor of black hole ( bh ) spin , rather than bh mass or accretion rate , as a key trigger in determining whether an object will be rl or rq .
we find a ten - fold decrease in ew with eddington ratio ( decreasing from @xmath21 to @xmath30.01 ) while shows no change .
these trends suggest a luminosity - independent `` baldwin effect '' where the physical driver may be the eddington ratio . |
in this paper , we study the yamabe invariant of compact orbifolds with a finite number of singularities ( i.e. , the orbifold yamabe invariant ) and its relationship to the yamabe invariant of cylindrical manifolds ( i.e. , the cylindrical yamabe invariant ) . in the @xmath14-dimensional case , we also establish a method for estimating the cylindrical yamabe invariant ( including the orbifold yamabe invariant ) from above , by means of conformal geometry and the atiyah - patodi - singer @xmath0-index theory . * [ int].1 .
the orbifold yamabe invariant . *
first , we define the yamabe invariant of such orbifolds , that is , the orbifold yamabe invariant , and then we study some fundamental properties on the orbifold yamabe invariant .
an orbifold @xmath2 under consideration here is a relatively compact smooth manifold outside of a finite number of singular points @xmath17 , and near each point @xmath11 it is locally homeomorphic to the orbit space @xmath18 , where @xmath19 is a finite group acting freely on @xmath20 .
we use the notation @xmath21 and @xmath22 , @xmath23 for the singularities of @xmath2 . throught this paper , we assume that @xmath24 .
there are natural _ orbifold riemannian metrics _
compatible with the orbifold structure .
we denote the space of such metrics by @xmath25 .
let @xmath26 , @xmath27 and @xmath28 denote respectively the scalar curvature , the volume form and the volume corresponding to @xmath29 . to start with , we note that the ( normalized ) einstein - hilbert functional @xmath30 has exactly the same property as in the case of compact manifolds : the set of critical points @xmath31 coincides with the space of _ orbifold einstein metrics _
( see proposition [ eh ] ) .
we also observe that the functional @xmath32 restricted on each _ orbifold conformal class _ @xmath33_{{{\mathrm o}{\mathrm r}{\mathrm b}}}\in { \mathcal c}^{{{\mathrm o}{\mathrm r}{\mathrm b}}}(m)$ ] has the same basic properties as in the case of compact manifolds ( cf .
@xcite ) , where @xmath34 denotes the space of orbifold conformal classes on @xmath2 .
this leads us to the definitions of the _ orbifold yamabe constant _
@xmath35_{{{\mathrm o}{\mathrm r}{\mathrm b}}}}(m)$ ] and the _ orbifold yamabe invariant _
@xmath5 respectively : @xmath36_{{{\mathrm o}{\mathrm r}{\mathrm b}}}}(m):= \inf_{\tilde{g}\in [ g]_{{{\mathrm o}{\mathrm r}{\mathrm b } } } } i(\tilde{g } ) , \ \ \ \ y^{{{\mathrm o}{\mathrm
r}{\mathrm b}}}(m):=\sup_{c\in { \mathcal c}^{{{\mathrm o}{\mathrm r}{\mathrm b}}}(m ) } y_c(m).\ ] ] * [ int].2 . the orbifold and the cylindrical yamabe invariants . * next , we study the relationship between the oribifold yamabe invariant and the corresponding @xmath6-cylindrical yamabe invariant .
consider , for simplicity , the case when a compact orbifold @xmath2 has only one singularity @xmath37 .
we observe that , in the category of smooth manifolds , the open manifold @xmath38 could be considered as the underlying smooth manifold of a _ cylindrical manifold _
@xmath39 with the cylindrical end @xmath40 , where @xmath41 is the corresponding spherical space form equipped with the standard metric @xmath42 of constant curvature one . as an object in the category of riemannian manifolds ,
the orbifold @xmath2 equipped with an orbifold metric @xmath43 is quite different from the cylindrical manifold @xmath39 . even from the viewpoint of conformal geometry , there is no `` cylindrical metric '' within the conformal class @xmath33 $ ] , in general .
however , the following holds ( theorem [ s2_th1 ] ) .
* theorem a. * _ let @xmath2 be a compact orbifold with one singularity @xmath37 . then @xmath44 . here
, @xmath45 denotes the @xmath42-cylindrical yamabe invariant of @xmath15 ( see section 2.6 or ( * ? ? ?
* section 2 ) for the definition of @xmath45 ) . _ using this , we prove the following estimate of the orbifold yamabe invariant ( corollary [ s2_cor1 ] ) .
* theorem b. * _ let @xmath2 be a compact orbifold of @xmath46 with singularities @xmath47 . then _ @xmath48 _ where @xmath49 on the slice @xmath10 of each end @xmath50 , and @xmath51 denotes the order of @xmath52 .
* [ int].3 .
harmonic spinors on cylindrical @xmath14-manifolds .
* our main goal is to obtain an estimate of the @xmath6-cylindrical yamabe invariant for a general slice metric @xmath6 and , via theorem a , the orbifold yamabe invariant , by means of the atiyah - patodi - singer @xmath0-index theory for cylindrical @xmath14-manifolds .
indeed , in section 3 , we will give some estimates for the @xmath6-cylindrical yamabe invariant from above , which can be regarded as a natural generalization of those for the yamabe invariant obtained in gursky - lebrun @xcite .
let @xmath39 be a cylindrical @xmath14-manifold modeled by @xmath53 for a metric @xmath6 on the slice @xmath54 .
it means that on the end @xmath55 , the metric @xmath56 is cylindrical with respect to a product coordinate system @xmath57 .
then we use the notation that @xmath58 .
now , we assume that there exists a spin@xmath59-structure on @xmath39 given by an element @xmath60\ ] ] which is not a torsion class , where @xmath61 denotes the second cohomology with compact support .
let @xmath62 be the @xmath63-line bundle with @xmath64 and with the restriction @xmath65 being `` cylindrical '' . similarly to the case of compact manifolds
, there exists a unique _ @xmath0-harmonic form @xmath66 _ representing the cohomology class @xmath67 .
note that the form @xmath68 does not have compact support unless @xmath69 .
however , there exists a sequence @xmath70 of closed @xmath71-forms with compact support such that @xmath72 = a$ ] and @xmath73 in an appropriate topology .
we choose @xmath74-connections @xmath75 ( with compact support ) on @xmath62 such that @xmath76 .
then , for each @xmath77 , the associated twisted dirac operator @xmath78 is defined , where @xmath79 stands for the plus / minus spin bundle associated to @xmath80 . note that the @xmath0-index of @xmath81 is independent of @xmath77 .
we denote by @xmath82 the total dimension of the negative eigenspaces of the intersection form on @xmath83 .
the following is the main result of section [ s3 ] ( theorem [ s3_t1 ] ) .
* theorem c. * _ let @xmath39 be a cylindrical @xmath14-manifold modeled by @xmath53 with @xmath84 .
assume that the @xmath0-index of @xmath81 is positive . then _
@xmath85}(z\times { { \mathbf r}})\}.\ ] ] the proof of theorem c is subtle .
we modify the argument due to gursky - lebrun @xcite , involving the bochner technique and a generalization of the _ modified scalar curvature _ to the case of noncompact manifolds .
one of the difficulties in the proof is the following : it turns out that a suitable conformal metric @xmath86 ( where @xmath87 ) for this argument never provides a complete metric on @xmath15 ( see lemmas [ s3_l2 ] , [ s3_l4 ] and the estimate ( [ s3_eq6 ] ) in section [ s3 ] ) .
the main technical difficulty arises from this point . in order to overcome the difficulty , we have to estimate , for instance , the function @xmath88 uniformly on the cylindrical end @xmath55 ( see lemma [ s3_l6 ] ) .
we remark that the hodge theorem and the atiyah - singer index theorem still hold for compact orbifolds ( under some modifications ) ( cf .
@xcite , @xcite ) .
hence , if one considers only compact @xmath14-orbifolds and their orbifold yamabe invariants , the arguments for the estimate are more direct with the aid of these theorems .
however , we establish here an estimate of the cylindrical yamabe invariant for cylindrical 4-manifolds modeled by general compact riemannian @xmath89-manifolds .
moreover , the combination of conformal geometry and analysis on cylindrical manifolds is of independent value and leads to new interesting results ( cf .
@xcite , @xcite , @xcite ) .
therefore , we rather deal with conformal geometry and analysis on cylindrical manifolds by means of the @xmath0-hodge theory and the atiyah - patodi - singer @xmath0-index theory .
* [ int].4 .
examples . * here we present particular cylindrical @xmath14-manifolds @xmath90 defined as follows .
let @xmath91 be the anti - canonical @xmath63-line bundle over @xmath92 , and @xmath93 for @xmath94 .
then , let @xmath90 denote the total space of the bundle @xmath95 , which will be naturally the underlying smooth open manifold ( with a connected tame end ) of a cylindrical manifold . for any cylindrical metric @xmath96 on @xmath90 , the cylindrical end of @xmath90
is reprented by @xmath97 with a metric @xmath6 on @xmath98 , where @xmath99 .
note that the one - point compactification @xmath100 of @xmath90 has a natural orbifold structure with one singularity @xmath101 .
the main result of section [ s4 ] is the following estimate ( theorem [ s4_t1 ] ) : * theorem d. * _ with the above understood , let @xmath102 denote the space of metrics of positive scalar curvature on @xmath103 . _ 1 . for @xmath94 and
any metric @xmath104 homotopic to @xmath105 in @xmath102 , @xmath106}((s^{3}/\gamma_{\ell})\times { { \mathbf r}})\}.\ ] ] 2 . for any metric @xmath107 which is sufficiently @xmath108-close to @xmath105 , @xmath109}((s^{3}/\gamma_{\ell})\times { { \mathbf r } } ) & \mbox{if } \ \ \ \ell\geq 2 .
\end{array } \right.\ ] ] in particular , @xmath110 for @xmath111 . *
the plan of the paper . * in section [ s2 ] , we define and study the orbifold yamabe constant / invariant and prove theorems a and b ( theorem [ s2_th1 ] and corollary [ s2_cor1 ] ) . in section [ s3 ] ,
we define and study the modified scalar curvature .
then we review the necessary part of the @xmath0-index theory and prove theorem c ( theorem [ s3_t1 ] ) .
section [ s4 ] is devoted to the above examples and theorem d ( theorem [ s4_t1 ] ) . *
. acknowledgements .
* both authors are grateful to harish seshadri for interesting discussions .
the first author is grateful to the department of mathematics at the university of oregon for kind hospitality .
the first author also would like to thank tosiaki kori and mikio furuta for useful comments on eta invariants and the atiyah - singer index theorem over orbifolds , respectively .
* [ s2].1 . orbifolds with a finite number of singularities : definitions .
* there are several different approaches to and definitions of orbifolds ( cf .
@xcite , @xcite ) .
since we work here with orbifolds of a particular type , i.e. , with only a finite number of singularities , we give suitable definitions ( cf .
we assume here that a finite group @xmath112 of the @xmath1-dimensional orthogonal group @xmath113 inherits the standard @xmath113-action on @xmath114 .
we always assume that @xmath115 .
first we define orbifolds with a finite number of singularities as an object in smooth category .
[ s2-d1 ] let @xmath2 be a locally compact hausdorff space .
we say that @xmath2 is an @xmath1-dimensional _ orbifold with singularities _
@xmath116 if the following conditions are satisfied : 1 . @xmath117 , and @xmath118 is a smooth manifold of dimension @xmath1 .
@xmath23 is the collection of subgroups @xmath52 of @xmath113 .
each group @xmath119 is a nontrivial finite subgroup of @xmath113 acting freely on @xmath20 .
3 . for each @xmath120
, there exist an open neighborhood @xmath121 of @xmath11 and a homeomorphism @xmath122 for some @xmath123 such that the restriction @xmath124 is a diffeomorphism . here
@xmath125 .
we refer to the pair @xmath126 as a _ singular point with the structure group _ @xmath52 and the pair @xmath127 as a _ local uniformization_.
let @xmath128 denote the canonical projection .
[ s2-d1 - 1 ] let @xmath129 be an open neighborhood of @xmath11 and @xmath130 a homeomorphism ( for some @xmath131 ) .
we call the pair @xmath132 a _ compatible local uniformization _ of @xmath126 if the following conditions are satisfied : 1 .
the restriction @xmath133 is a diffeomorphism .
there exists a @xmath52-equivariant diffeomorphism @xmath134 for some @xmath135 ( possibly not onto ) such that @xmath136 where @xmath137 denotes also the canonical projection . to simplify the presentation , we assume , without particular mention , that an orbifold @xmath2 has only one singularity , i.e. , @xmath138 . let @xmath139 be a local uniformization and @xmath140 the canonical projection .
we also always assume that @xmath2 is compact .
now we give the definition of orbifold metrics .
[ s2-d2 ] a riemannian metric @xmath141 is an _ orbifold metric _ if there exists a @xmath112-invariant smooth metric @xmath142 on the disk @xmath143 such that @xmath144 on @xmath145 .
we denote by @xmath25 the space of all orbifold metrics on @xmath2 .
we note that if @xmath29 , then @xmath146 , and hence the scalar curvature @xmath26 extends to a continuous function on @xmath2 .
@xmath147 in the case when @xmath148 , the space @xmath25 of orbifold metrics is defined similarly . *
einstein orbifold metrics .
* let @xmath2 be a compact orbifold with a singularity @xmath37 .
consider the ( normalized ) einstein - hilbert functional @xmath149 as in the case of smooth compact manifolds , we have the following fundamental result since the stokes formula and the divergence theorem still hold over orbifolds .
[ eh ] the set of critical points of @xmath150 coincides with the set of einstein orbifold metrics on @xmath2 .
* [ s2].3 .
orbifold conformal classes .
* we say that two orbifold metrics @xmath151 are _ pointwise conformal _ if there exists a function @xmath152 such that @xmath153 here we used the composition of the maps @xmath154 .
then we define an _ orbifold conformal class of @xmath43 _ as follows : @xmath155_{{{\mathrm o}{\mathrm r}{\mathrm b } } } : = [ g ] \cap \riem^{{{\mathrm o}{\mathrm r}{\mathrm b}}}(m ) = \
{ e^{2f } \cdot g \
\left| \begin{array}{l } f\in c^0(m)\cap c^{\infty}(m\setminus\{\check{p}\ } ) , \\ ( \phi^{-1}\circ \pi)^*f \in c^{\infty}({{\mathbf b}}_\tau(0 ) ) \end{array } \right.\ } .\ ] ] let @xmath34 denote the space of all orbifold conformal classes .
the proofs of the following two lemmas are similar to the case of smooth compact manifolds ( see @xcite ) .
[ lemma1 - 5 ] let @xmath156 be a metric satisfying one of the following conditions : @xmath157 , @xmath158 or @xmath159 everywhere on @xmath2 .
then for another such metric @xmath160_{{{\mathrm
o}{\mathrm r}{\mathrm b}}}$ ] , the sign of @xmath161 is identical with the sign of @xmath26 . [ lemma1 - 6 ] let @xmath156 be a metric with @xmath162 . then for any metric @xmath163_{{{\mathrm o}{\mathrm r}{\mathrm b}}}$ ] with @xmath164 , there exists a constant @xmath165 such that @xmath166 .
an orbifold @xmath2 is called _ good _ if its universal cover @xmath167 is a smooth manifold ( cf .
let @xmath168 be the projection .
let @xmath156 be an einstein metric with @xmath169 .
then any metric @xmath163_{{{\mathrm o}{\mathrm r}{\mathrm b}}}$ ] with @xmath164 is also einstein .
furthermore , if @xmath2 is a good orbifold , then @xmath166 for a constant @xmath165 unless @xmath170 is isometric to @xmath171 , where @xmath172 denotes the standard metric of constant curvature one .
the first statement follows directly from the proof of proposition 1.4 in @xcite combined with the divergence theorem for orbifolds ( see also ( * ? ? ?
* section 2 ) ) .
now assume that the universal cover @xmath167 is a smooth manifold .
let @xmath168 denote the projection .
let @xmath173 be an orbifold metric with @xmath174 on @xmath2 .
then unless @xmath175 , ( * ? ? ?
* theorem a ) implies that @xmath176 on @xmath167 , and hence @xmath177 on @xmath2 .
* [ s2].4 .
orbifold yamabe constants / invariants . * we define the _ orbifold yamabe constant _
@xmath35_{{{\mathrm o}{\mathrm r}{\mathrm b}}}}(m)$ ] of @xmath178_{{{\mathrm o}{\mathrm r}{\mathrm b}}})$ ] as follows : @xmath36_{{{\mathrm o}{\mathrm r}{\mathrm b}}}}(m ) : = \inf_{\widetilde{g}\in [ g]_{{{\mathrm o}{\mathrm r}{\mathrm b } } } } \frac{\int_m r_{\widetilde{g } } d\sigma_{\widetilde{g } } } { \vol_{\widetilde{g}}(m)^{\frac{n-2}{n}}}\ .\ ] ] similarly to the case of smooth compact manifolds , we have aubin s inequality @xmath179_{{{\mathrm
o}{\mathrm r}{\mathrm b}}}}(m ) \leq y(s^n)$ ] ( cf .
. then we define the _ orbifold yamabe invariant of @xmath2 _
@xmath180 now we state some technical facts without proofs .
the following result also follows from the divergence theorem for orbifolds .
[ s2_l1 ] for an orbifold metric @xmath181_{{{\mathrm o}{\mathrm r}{\mathrm b}}}$ ] with @xmath182 , then @xmath183 we consider the yamabe functional @xmath184 for any @xmath185 with @xmath186 and @xmath187 .
we recall that for a compact smooth manifold @xmath188 of dim @xmath189 , any point @xmath190 and any metric @xmath191 , the yamabe constant @xmath192}(n)$ ] satisfies @xmath193}(n)= \inf_{\begin{array}{c}^{u\in c_c^{\infty}(n \setminus \{q\})}\\ ^{u\not\equiv 0}\end{array } } q_h ( u).\ ] ] the orbifold yamabe constant has a similar property .
[ s2_l2 ] the orbifold yamabe constant @xmath35_{{{\mathrm o}{\mathrm r}{\mathrm b}}}}(m)$ ] satisfies @xmath36_{{{\mathrm o}{\mathrm r}{\mathrm b}}}}(m ) = \inf_{\begin{array}{c}^{u\in c_c^{\infty}(m\setminus \{\check{p}\})}\\ ^{u\not\equiv 0}\end{array } } q_g(u)\ .\ ] ] here @xmath194 is the singular point of @xmath2 . *
approximation for orbifold metrics .
* at the singular point @xmath195 , we prove the following approximation , which is an orbifold version of kobayashi s approximation lemma ( ( * ? ? ?
* lemma 3.2 ) ) .
this approximation is a crucial tool linking the orbifold yamabe invariant to the corresponding @xmath6-cylindrical yamabe invariant .
[ s2_p1 ] let @xmath196 be an orbifold metric and @xmath197 the lifting metric with @xmath198 on @xmath199 let @xmath200 denote normal coordinates with respect to @xmath142 on an open neighborhood @xmath201 of @xmath202 , and @xmath203 for small @xmath204 .
then for any small @xmath205 there exists a metric @xmath206 such that : 1 .
@xmath207 on @xmath208 , where @xmath209 , 2 .
@xmath210 is pointwise conformal to the flat metric @xmath211 on @xmath212 for a constant @xmath213 , where @xmath214 denotes the corresponding lifting metric to @xmath215 , 3 .
@xmath216 uniformly on @xmath217 and @xmath218 uniformly on @xmath143 as @xmath219 , 4 .
@xmath220 uniformly on @xmath2 as @xmath219 .
let @xmath221 be the flat metric on @xmath222 .
then @xmath223 since @xmath142 is @xmath112-invariant , the exponential map ( with respect to @xmath142 ) commutes with @xmath112 , and hence the flat metric @xmath224 and the distance function @xmath225 from @xmath202 to @xmath226 are also @xmath112-invariant .
moreover , @xmath227 is a compatible local uniformization .
we denote @xmath228 , and consider the metric @xmath229 then it follows that @xmath230 .
now we use the cut - off function @xmath231 ( given in @xcite ) to construct the following approximation @xmath232 : @xmath233 since @xmath234 and @xmath235 , kobayashi s approximation technique @xcite implies that @xmath232 satisfies the above conditions * ( 1)**(4 ) * on @xmath236 . by the construction , the metrics @xmath142 , @xmath237 and the function @xmath231 are @xmath112-invariant , and hence the metric @xmath232 is also @xmath112-invariant .
this implies that there exists an orbifold metric @xmath238 such that @xmath239 .
then the metric @xmath215 satisfies the above conditions * ( 1)**(4)*. the following lemma is an analogue of the case of smooth compact manifolds ( cf .
@xcite ) .
[ s2_l4 ] let @xmath240 be metrics satisfying @xmath241 then @xmath242_{{{\mathrm o}{\mathrm r}{\mathrm b}}}}(m)\to y_{[g]_{{{\mathrm o}{\mathrm r}{\mathrm b}}}}(m)$ ] as @xmath219 .
* [ s2].6 .
orbifold and @xmath6-cylindrical yamabe invariants .
* we recall briefly the following : the open manifold @xmath217 equipped with an appropriate cylindrical metric can be considered as a cylindrical manifold ( see @xcite or section [ s3 ] ) . for a complete metric @xmath243 , the open manifold @xmath244 is called a _
cylindrical manifold modeled by _ @xmath245 if there exist an open neighborhood @xmath246 of @xmath194 and a coordinate system @xmath247 such that @xmath248 the _ cylindrical yamabe constant _
@xmath249}(m\setminus \{\check{p}\})$ ] is defined by @xmath250}(m\setminus \{\check{p}\ } ) : = \inf_{\begin{array}{c}^{u\in c_c^{\infty}(m\setminus \{\check{p}\})}\\ ^{u\not\equiv 0}\end{array } } q_{\bar{g}}(u ) \ \ \ ( \ \leq y(s^n)\ ) .\ ] ] let @xmath251 be the space of cylindrical metrics modeled by @xmath245 .
then the _ @xmath42-cylindrical yamabe invariant _
@xmath252 is also defined by @xmath253}(m\setminus \{\check{p}\ } ) \ \ \ ( \ \leq y(s^n)\ ) .\ ] ] [ s2_th1 ] let @xmath2 be a compact orbifold with a singularity @xmath254
. then @xmath255 .
theorem [ s2_th1 ] and ( * ? ? ?
* propositions 2.11 , 2.12 , 6.5 ) imply the following result .
[ s2_cor1 ] let @xmath2 be a compact orbifold with singularities @xmath256 @xmath257 .
then @xmath258 proof of theorem [ s2_th1 ] first we prove that @xmath259 . by proposition [ s2_p1 ] and lemma [ s2_l4 ] ,
for any @xmath260 and any metric @xmath196 there exists a metric @xmath261 such that @xmath262 let @xmath263_{{{\mathrm o}{\mathrm r}{\mathrm b}}}$ ] be a cylindrical metric on @xmath264 with @xmath265 on the end @xmath266
. then we use lemma [ s2_l2 ] to show @xmath267_{{{\mathrm o}{\mathrm r}{\mathrm b}}}}(m ) & = & \displaystyle \inf_{\begin{array}{c}^{u\in c_c^{\infty}(m\setminus \{\check{p}\})}\\ ^{u\not\equiv 0}\end{array } } q_{g_{\epsilon}}(u ) = \displaystyle \inf_{\begin{array}{c}^{u\in c_c^{\infty}(m\setminus \{\check{p}\})}\\ ^{u\not\equiv 0}\end{array } } q_{\bar{g}_{\epsilon}}(u ) \\ \\ & = & y^{\cyl}_{[\bar{g}_{\epsilon}]}(m \setminus \{\check{p}\ } ) \leq y^{h_{\gamma}\hbox{-}\cyl}(m \setminus \{\check{p}\ } ) .
\end{array}\ ] ] this implies that @xmath268 .
second we prove that @xmath269 .
we start with an arbitrary cylindrical metric @xmath270 and a cylindrical coordinate system @xmath271 as above .
set @xmath272 , then @xmath273 on the cylinder @xmath274 .
note that on @xmath274 the metric @xmath275 is flat and is extended to a smooth metric @xmath224 on @xmath217 .
this implies that there exists a homeomorphism @xmath276 for a constant @xmath277 such that 1 .
@xmath278 is a diffeomorphism , 2 .
@xmath279 , where @xmath280 are normal coordinates around @xmath194 with respect to @xmath224 , 3 .
the diffeomorphism @xmath281 is lifted to a @xmath112-equivariant diffeomorphism @xmath282 onto its image for a constant @xmath283 . in other words
, @xmath284 is a compatible local uniformization , and the above metric @xmath224 is an orbifold metric on @xmath2 .
> from lemma [ s2_l2 ] , we obtain that @xmath285_{{{\mathrm o}{\mathrm r}{\mathrm b}}}}(m ) = \displaystyle \!\!\!\!\!\ ! \inf_{\begin{array}{c}^{u\in
c_c^{\infty}(m\setminus \{\check{p}\})}\\ ^{u\not\equiv 0}\end{array}}\!\!\ !
q_{g_0}(u ) = \displaystyle \!\!\!\!\!\ ! \inf_{\begin{array}{c}^{u\in c_c^{\infty}(m\setminus \{\check{p}\})}\\ ^{u\not\equiv 0}\end{array}}\!\!\ !
q_{\bar{g}}(u ) = y^{\cyl}_{[\bar{g}]}(m\setminus \{\check{p}\ } ) . \end{array}\ ] ] this implies that @xmath286 .
* [ s3].1 . modified scalar curvature .
* throughout the rest of the paper , we consider only @xmath14-manifolds .
let @xmath15 be an open smooth @xmath14-manifold with tame ends , i.e. , it is diffeomorphic to @xmath287 , where @xmath288 is a compact submanifold of dim @xmath289 with boundary @xmath290 ( possibly disconnected ) , see fig .
[ s3].1 .
( 5,0 ) ( -5,20)@xmath291 ( 40,5)@xmath292 ( 10,45)@xmath293 ( 35,40)@xmath294 cylindrical manifold @xmath39 .
let @xmath295 denote the space of all cylindrical metrics on @xmath15 .
we choose a cylindrical metric @xmath296 . in terms of @xcite
, @xmath297 is a _
cylindrical @xmath14-manifold modeled by _ @xmath53 for @xmath298 if there exists a product coordinate system @xmath57 on the end @xmath299 such that @xmath300 on @xmath301 .
below we use the notation that @xmath302 .
let @xmath303 be a 2-form .
we denote by @xmath304 the conformal laplacian of @xmath305 .
we define the _ modified conformal laplacian @xmath306 of @xmath307 _ by @xmath308 note that the pointwise norm @xmath309 is only of @xmath310 class on @xmath15 , in general .
[ s3_f1](cf .
@xcite ) set @xmath311 for @xmath312
. then @xmath313 for @xmath314 .
( gursky - lebrun @xcite ) for @xmath315 , let @xmath311 be a conformal metric , where @xmath312 .
we call the function @xmath316 the _ modified scalar curvature of @xmath317_. set @xmath318d\sigma_{\bar{g}}$ ] for @xmath319 , where @xmath320 denotes the sobolev space of square - integrable functions on @xmath15 ( with respect to @xmath305 ) up to their first weak derivatives
. then we consider the functional @xmath321 and the bottom of the spectrum of the operator @xmath306 @xmath322 > from now on we assume that the @xmath71-form @xmath323 has a compact support .
[ s3_l1 ] for @xmath324 and @xmath325 as above , @xmath326 .
indeed , @xmath327 since @xmath328 and @xmath325 . on the slice manifold @xmath53 , we consider the _ almost conformal laplacian _ @xmath329 ( see @xcite ) . * convention 1 . * from now on
we always assume that @xmath330 , where @xmath331 denotes the first eigenvalue of @xmath332 .
( ( * ? ? ?
* section 2 ) ) it is easy to see that the condition @xmath333 everywhere on @xmath54 provides a sufficient condition for @xmath334 . on the other hand , if @xmath335 , then @xmath249}(x)\leq 0 $ ] .
moreover , if @xmath336 , then @xmath249}(x)=-\infty$ ] .
[ s3_l2 ] assume that @xmath337
. then there exists a function @xmath338 ( for any @xmath339 ) such that @xmath340 . _ proof .
_ consider the cylinder @xmath341 with @xmath94 and denote @xmath342)$ ] .
note that if there exists a constant @xmath94 satisfying @xmath343 then the standard argument implies the existence of a non - zero minimizer @xmath338 of the functional @xmath344 .
hence it is enough to prove that @xmath345 for a constant @xmath94 .
we choose @xmath346 satisfying @xmath347
. then @xmath348 and @xmath349 on @xmath341 .
hence for any @xmath350 , @xmath351d\sigma_h \\ \\ \!\!&\geq&\!\ ! \displaystyle \int_{[\ell,\infty)}\!\ ! dt \left(\ ! \lambda_h \cdot \int_{z } f^2 d\sigma_h\ !
\right ) \geq \lambda_h \cdot \int_{z\times [ \ell,\infty)}f^2 d\sigma_{\bar{g } } . \
\box \end{array}\ ] ] the next lemma shows that we may always assume the condition of lemma [ s3_l2 ] without loss of generality .
the technique below of changing the given metric conformally within a compact set is known as the _ conformal - rescaling argument _ ( see also the proofs of ( * ? ? ?
* theorem 3 ) and ( * ? ? ? * proposition 7.3 ) ) .
[ s3_l3 ] there exists a metric @xmath352 $ ] with @xmath353 on @xmath354 such that @xmath355 .
in particular , @xmath356 . when @xmath357 , the inequality @xmath358 is trivial .
hence we consider only the case when @xmath359 .
we change the cylindrical metric @xmath305 to another cylindrical metric @xmath360 , where @xmath361 with @xmath362 in particular , @xmath363 on @xmath291 .
we choose @xmath364 sufficiently large .
then the dirichlet first eigenvalue @xmath365 on @xmath291 satisfies that @xmath366 .
hence by the domain monotonicity of the dirichlet eigenvalues , we then obtain that @xmath367 . to complete the proof ,
we let @xmath368 .
* convention 2 . * from now on
we always assume that @xmath358 since @xmath369}(x)$ ] @xmath370}(x)$ ] .
lemmas [ s3_l2 ] , [ s3_l3 ] imply the following lemma , whose proof is similar to the case of compact smooth manifolds ( see @xcite ) .
[ s3_l4 ] there exists a function @xmath371 ( where @xmath339 ) such that for the metric @xmath86 @xmath372 furthermore , these cases are mutually exclusive .
the next result follows from ( * ? ? ?
* proposition 7.1 ) .
[ s3_f2 ] let @xmath373 be two cylindrical metrics which are pointwise conformal .
then @xmath374 .
now for @xmath324 , let @xmath375_{l^{1,2}_{\bar{g}}}\ ( \subset [ \bar{g}])$ ] denote the @xmath376-_conformal class _ of @xmath305 consisting of all metrics @xmath377 , where @xmath378 ( see @xcite ) .
note that the @xmath379-norm @xmath380 of @xmath381 depends only on the conformal class @xmath375 $ ] , and hence set @xmath382}}:=\|\zeta\|_{l^2_{\bar{g}}}$ ] .
then , fact [ s3_f2 ] combined with the argument given in ( * ? ? ?
* corollary 4 ) implies the following assertion .
[ s3_p1 ] let @xmath305 be a cylindrical metric on @xmath15 , and @xmath383 as above .
then one of the following * ( 1)(3 ) * holds : 1 .
there exists a metric @xmath384_{l^{1,2}_{\bar{g}}}$ ] such that @xmath385 on @xmath15 , 2 .
@xmath386 , 3 .
@xmath387 , and there exists a metric @xmath388_{l^{1,2}_{\bar{g}}}$ ] such that @xmath389 on @xmath15 . *
@xmath0-harmonic 2-forms .
* let @xmath39 be a cylindrical @xmath14-manifold modeled by @xmath53 .
a @xmath71-form @xmath390 is an @xmath379-harmonic @xmath71-form on @xmath15 , i.e. , @xmath68 satisfies @xmath391 here @xmath392 stands for the codifferential with respect to the metric @xmath305 .
we consider the space of @xmath379-harmonic @xmath71-forms @xmath393 ( cf .
@xcite ) let @xmath66 be an @xmath379-harmonic 2-form .
then the function @xmath394 decays exponentially on @xmath292 .
[ s3_f4]@xmath395 1 .
( atiyah - patodi - singer @xcite ) there is an isomorphism : @xmath396,\ ] ] where @xmath61 denotes the second cohomology with compact support .
2 . ( dodziuk @xcite ) for any @xmath379-harmonic 2-form @xmath66
, there exists a sequence of closed @xmath71-forms @xmath397 such that @xmath398=[\zeta ] \in h^2(x;{{\mathbf r } } ) , \\ \zeta_j\to
\zeta \ \ \mbox{in the $ l^{k,2}_{\bar{g}}$-topology on $ x$ for all $ k\geq 0 $ as $ j\to \infty$. } \end{array } \right.\ ] ] recall that the sobolev embedding @xmath399 holds for all @xmath400 .
this implies that @xmath401 as @xmath402 .
hence each closed @xmath71-form @xmath403 is _ not harmonic _ but _ almost - harmonic _ for @xmath404 unless @xmath405 . *
@xmath0-harmonic spinors .
* > from now on we assume that @xmath297 is an oriented cylindrical @xmath14-manifold . an element @xmath406 is a _ characteristic element _ if @xmath407 mod 2 ( where @xmath408 is the second stiefel - whitney class )
. we will identify @xmath67 with its image in @xmath409 .
let @xmath410 denote the canonical projection .
we choose a characteristic element @xmath411 $ ] which is not a torsion class .
then there exists a hermitian cylindrical @xmath63-line bundle @xmath80 over @xmath15 such that @xmath412 here @xmath413 is the first chern class of @xmath80 .
let @xmath79 denote the plus / minus spin bundle associated to @xmath80 . here
@xmath414 is the ( virtual ) plus / minus bundle .
then we have , as a hermitian vector bundle , @xmath415 > from ( * ? ? ?
* theorem 2.7 ) , there exists a unique @xmath379-harmonic 2-form @xmath66 such that its cohomology class @xmath416=a\in h^2(x;{{\mathbf r}})$ ] .
in particular , from fact [ s3_f4]-(2 ) , there exists a sequence of closed @xmath71-forms @xmath417 such that @xmath398 = a \in h^2(x;{{\mathbf r } } ) , \\ \zeta_j\to
\zeta \ \ \mbox{in } \ \ l^{3,2}_{\bar{g}}(x)\cap c^{0,\alpha}(x ) \ \ \mbox{as } \ \ j\to\infty . \end{array } \right.\ ] ] let @xmath418 denote the space of @xmath74-connections on @xmath80 . then there exists a sequence of @xmath74-connections @xmath419 with compact support such that @xmath420 we also denote by @xmath421 the associated ( twisted ) dirac operator
. * assumption a1 . *
> from now on we assume that , for a fixed cylindrical metric @xmath422 with @xmath423 , @xmath424 here @xmath425 stands for the @xmath0-index of @xmath81 and @xmath426 for the extended @xmath0-kernel of the adjoint operator @xmath427 ( cf . @xcite ) .
we emphasize that the above index is independent of @xmath77 ( cf .
@xcite ) .
[ s3_p2 ] for any @xmath324 with @xmath58 , let @xmath428 be a non - torsion characteristic element and @xmath66 its @xmath379-harmonic representative . then , under the assumption * a1 * , @xmath250}(x)\leq \min \ { 4\pi\sqrt{2(a^+)^2 } , \ y^{\cyl}_{[h+dt^2]}(z \times { { \mathbf r } } ) \}.\ ] ] here @xmath429 , and @xmath430 stands for the self - dual part of @xmath68 with respect to @xmath305 .
@xmath249}(x)\leq y^{\cyl}_{[h+dt^2]}(z\times{{\mathbf r}})$ ] ( see @xcite ) .
hence it is enough to show that @xmath249}(x)\leq 4\pi\sqrt{2(a^+)^2}$ ] . set @xmath431 in proposition [ s3_p1 ] . as we have seen , the form @xmath323 is not closed in general .
we postpone the proof of the following assertion to the end of this section .
[ s3_c1 ] for any @xmath324 with @xmath58 , let @xmath428 be a non - torsion characteristic element and @xmath432 its @xmath379-harmonic representative . then , under the assumption * a1 * , the following inequality holds : @xmath433}(x ) \leq\|4\sqrt{2}\pi\zeta_j^+\|_{l_{\bar{g}}^2 } = 4\pi\left(2 \int_x |\zeta_j^+|^2_{\bar{g } } d\sigma_{\bar{g } } \right)^{\frac{1}{2}}\quad \textrm{for any}\ j.\ ] ] by taking @xmath434 in ( [ s3_eqq1 ] ) , proposition [ s3_c1 ] implies that @xmath250}(x ) \leq 4\pi\left ( 2 \int_x |\zeta^+|^2_{\bar{g } } d\sigma_{\bar{g } } \right)^{\frac{1}{2 } } = 4\pi\sqrt{2(a^+)^2}.\ ] ] this completes the proof of proposition [ s3_p2 ] . * assumption a2 .
* we assume that @xmath435 .
this implies that @xmath436 for any @xmath324 .
[ s3_t1 ] let @xmath39 be a cylindrical @xmath14-manifold modeled by @xmath53 . then , under the assumptions * a1 * and * a2 * , @xmath437}(z\times { { \mathbf r}})\}.\ ] ] proof let @xmath315 be any cylindrical metric with @xmath58 . for the 2-forms @xmath68 , @xmath403 and the @xmath74-connections @xmath75 as above , fact [ s3_f5 ] and assumption a1 imply that @xmath438 for any @xmath77 .
then , proposition [ s3_p2 ] and assumption a2 give the inequality @xmath439}(x ) \leq 4\pi\sqrt{2(a^+)^2 } = 4\pi\sqrt{2 a^2}. $ ] therefore , @xmath440 .
now we consider the following subspace of @xmath441 @xmath442 [ s3_f5](cf .
@xcite ) let @xmath443 be a @xmath74-connection with compact support @xmath444 .
let @xmath445 be two metrics homotopic in @xmath446 .
for any two cylindrical metrics @xmath447 with @xmath58 and @xmath448 , then @xmath449 then , theorem [ s3_t1 ] combined with fact [ s3_f5 ] implies the following .
[ s3_co1 ] let @xmath450 be a metric homotopic to @xmath6 in @xmath446
. then @xmath451}(z^3\times { { \mathbf r } } ) \}.\ ] ] * [ s3].4
. behavior of @xmath0-harmonic spinors on the cylindrical ends . * here we modify the bochner techinque given in @xcite , @xcite and @xcite to our case .
let @xmath452 be a cylindrical metric with @xmath453 and @xmath454 a conformal metric on @xmath15 with @xmath455 .
then the dirac operators @xmath456 and @xmath457 on @xmath458 are related as follows ( cf .
* proposition 2 ) ) : @xmath459 for any @xmath460 , then we also obtain the harmonic spinor @xmath461 .
now let @xmath462 be a non - zero @xmath379-harmonic plus - spinor .
then @xmath463 is a non - zero harmonic spinor with respect to @xmath464 , and hence the bochner formula gives that @xmath465 here and below `` @xmath466 '' and `` @xmath467 '' stand for the clifford multiplication corresponding to the metrics @xmath305 and @xmath468 , respectively .
then we have on @xmath469 ( for @xmath470 ) : @xmath471d\sigma_{\check{g } } \\ \\
! \displaystyle \int_{x(\ell)}\left [ \frac{1}{4}r_{\check{g}}|\check{\psi}|^2 + \frac{1}{2}\<f_a^+\check{\ccd } \check{\psi } , \check{\psi}\ > \right]d\sigma_{\check{g } } \\ \\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
\displaystyle - \int_{\p x(\ell)}\
< \nabla^{a,\check{g}}_{\check{\nu } } \check{\psi } , \check{\psi}\ > d\sigma_{\check{g}}|_{\p x(\ell ) } \\ \\ \!\!&\geq & \!\!\!\ ! \displaystyle \int_{x(\ell)}\left [ \frac{1}{4 } \left ( r_{\check{g } } -2\sqrt{2 } |f_a^+|_{\check{g } } \right)|\check{\psi}|^2 \right]d\sigma_{\check{g } } \\ \\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \displaystyle - \int_{\p x(\ell)}| \nabla^{a,\check{g}}_{\check{\nu } } \check{\psi}| \cdot| \check{\psi}| d\sigma_{\check{g}}|_{\p x(\ell ) } \ . \end{array}\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\ ] ] here @xmath472 is the outer unit normal vector field on @xmath473 with respect to @xmath468 . since @xmath474 , we have @xmath475 , @xmath476 , we have the estimate : @xmath477 \psi\right| \\ \\ & \leq & \!\!\ ! k \cdot u^{-\frac{5}{2 } } \left ( u^{-1}|\p_tu|\cdot |\psi| + |\nabla^{a,\bar{g}}_{\p_t}{\psi}| + u^{-1 } |du|_{\bar{g}}\cdot |\psi| \right ) \\ \\ & \leq&\!\!\ ! k^{\prime
} \cdot u^{-\frac{5}{2 } } \left ( u^{-1}|du|_{\bar{g}}\cdot |\psi| + |\nabla^{a,\bar{g}}_{\p_t}{\psi}| \right ) , \end{array}\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\ ] ] where @xmath478 is a local orthonormal frame on @xmath473 with respect to @xmath305 .
we use ( [ s3_eq2 ] ) to obtain that @xmath479 [ s3_l5 ] let @xmath480 be a function satisfying @xmath481 on @xmath482 for a fixed @xmath483
. then @xmath484 proof of proposition [ s3_c1 ] set @xmath485 and let @xmath486 be the same function as in lemma [ s3_l4 ] .
suppose that the @xmath487-metric @xmath86 satisfies that @xmath385 .
perturbing @xmath486 on a compact set , we may assume that @xmath486 is smooth and that @xmath488 .
then , the inequality ( [ s3_eq1 ] ) combined with the estimate ( [ s3_eq3 ] ) and lemma [ s3_l5 ] implies that @xmath489d\sigma_{\check{g } } \ .\ ] ] hence the condition @xmath490 contradicts the inequality ( [ s3_eq4 ] ) .
finally , from proposition [ s3_p1 ] , @xmath249}(x)\leq \|2\sqrt{2}f_{a_j}^+\|_{l_{\bar{g}}^2}= \|4\sqrt{2}\pi\zeta_j^+\|_{l_{\bar{g}}^2}$ ] .
now we have to prove lemma [ s3_l5 ] .
there are two independent tasks here to deal with , i.e. , the decay of the function @xmath486 and the pointwise norms @xmath491 , @xmath492 .
the spinor norms are taken care by the following result .
[ s3_f6](@xcite ) let @xmath493 be an @xmath494-harmonic spinor as above .
then the pointwise norms @xmath491 , @xmath492 decay exponentially on the cylinder @xmath55 . to complete the proof of lemma [ s3_l5 ]
, we need the following technical result .
[ s3_l6 ] there exists a positive constant @xmath495 such that @xmath496 on the cylinder @xmath55 .
_ proof of lemma [ s3_l5 ] .
_ by fact [ s3_f6 ] and lemma [ s3_l6 ] , there exist positive constants @xmath497 , @xmath498 , @xmath495 and @xmath499 such that @xmath500 as @xmath501.@xmath147 proof of lemma [ s3_l6 ] recall that the support @xmath502 is compact . hence there exists @xmath503 ( @xmath504 ) such that the restriction of @xmath323 is zero on the cylinder @xmath505 . in particular , the operator @xmath506 on @xmath505 .
recall that @xmath507 and @xmath508 denote respectively the first eigenvalue of the operator @xmath332 and the bottom of the spectrum of @xmath306 .
then , we consider the following operator @xmath509 on @xmath54 @xmath510 let @xmath511 be the eigenvalues and eigenfunctions of @xmath509 , i.e. , @xmath512 and @xmath513 . here
we may assume that @xmath514 on @xmath54 .
since @xmath515 , we obtain that @xmath516 .
let @xmath517 be a product coordinate system associated to @xmath305 .
note that @xmath518 on @xmath519 .
then on the cylinder @xmath519 , @xmath520 hence @xmath521 on @xmath519 .
then the function @xmath522 restricted on @xmath523 can be represented as @xmath524 we set @xmath525 and @xmath526 for @xmath527 .
let @xmath528 ( @xmath529 ) be the multiplicity of the @xmath530-th eigenvalue
. then @xmath531 it then follows that there exists a positive constant @xmath495 such that @xmath532 .
now we set @xmath533 . here
we need the following fact .
[ s3_cs ] with the above understood , then @xmath534 note that if there exists @xmath535 such that @xmath536 , then @xmath537 for sufficiently large @xmath538 .
this contradicts the positivity of @xmath486 everywhere on @xmath15 .
hence @xmath539 and @xmath540 on @xmath54 .
now suppose that @xmath541 is not the first eigenvalue .
note that the function @xmath542 itself is an eigenfunction corresponding to @xmath541 .
the condition @xmath543 implies that @xmath544 , which contradicts that @xmath545 , @xmath539 and @xmath540 on @xmath54 .
therefore , @xmath546 .
note that claim [ s3_cs ] implies the estimate @xmath547 on @xmath523 for a constant @xmath548 .
then the estimates ( [ s3_eq5 ] ) and ( [ s3_eq6 ] ) imply that @xmath496 on @xmath15 for a constant @xmath549 .
this completes the proof of lemma [ s3_l6 ] .
let @xmath550 denote the complex projective plane . in @xcite ( cf .
@xcite ) , lebrun proved that @xmath551 , and that @xmath552 is achieved by a conformal class @xmath553 if and only if @xmath553 coincides with the pullback @xmath554 $ ] of the conformal class @xmath555 $ ] by a diffeomorphism @xmath556 . here
@xmath557 denotes the fubini - study metric on @xmath550 . in this section ,
we give some estimates of @xmath16 for particular cylindrical @xmath14-manifolds @xmath15 , which are generalizations of part of the above result . * [ s4].1 . example 1 .
* let denote by @xmath91 the anti - canonical @xmath63-line bundle over the complex projective line @xmath558 and by @xmath93 the @xmath63-line bundle of degree @xmath94 .
then we choose a hermitian metric @xmath559 on the bundle @xmath560 , and consider the disk and the circle bundles of @xmath560 : @xmath561 here @xmath562 with the generator @xmath563 and the group @xmath564 acts on the sphere @xmath565 by @xmath566 .
we also denote by @xmath90 the total space of @xmath560 .
we consider @xmath90 as an open @xmath14-manifold with a tame end : @xmath567 first we note some necessary facts . 1 .
when @xmath568 , then @xmath569 is diffeomorphic to @xmath570 , and hence @xmath569 is an open @xmath14-manifold with tame end @xmath571 , where @xmath572 is a point in @xmath550 .
when @xmath111 , then the one - point compactification @xmath100 has a natural orbifold structure with @xmath101 .
* example 4.1.27 ) ) there exists a cylindrical metric @xmath573 @xmath574 with @xmath575 such that @xmath576 on @xmath90 .
@xmath577 \cong { { \mathbf z}}$ ] , and @xmath578 for an appropriate orientation of @xmath90
. 1 .
@xmath581}(z_{\ell } \times { { \mathbf r}})\}\ ] ] for @xmath94 and any metric @xmath582 homotopic to @xmath105 in @xmath583 .
2 . for any metric @xmath584 which is sufficiently @xmath108-close to @xmath105 , @xmath585}(z_{\ell } \times { { \mathbf r}})\ \ & \mbox{if } \ \ \ \ell\geq 2 . \end{array } \right.\ ] ] in particular , @xmath110 for @xmath111 . 1 .
the inequality @xmath586 holds only if @xmath587 .
2 . let @xmath2 be a closed @xmath14-manifold .
for any @xmath588 and a positive integer @xmath589 , then @xmath590 3 .
if we replace @xmath591 by @xmath592 , a similar result still holds for @xmath592 .
4 . unfortunately ,
the estimate @xmath593 in theorem [ s3_t1 ] is not effective for @xmath594 ( @xmath588 ) and @xmath595 since @xmath596 for @xmath111 .
proof of theorem [ s4_t1 ] for @xmath597 , let @xmath598 be a metric which is homotopic to @xmath105 in @xmath583 .
by fact [ s4_f1]-(2 ) , it is easy to construct a cylindrical metric @xmath305 on @xmath15 satisfying @xmath599 on @xmath600 , @xmath601 on @xmath602 and @xmath375 = [ h + dt^2]$ ] on @xmath603 .
then , from ( * ? ? ?
* proposition 4.6 , theorem 4.7 ) , we also obtain that @xmath250}(x_{\ell } ) \geq \min \{y_1 , y_2\ } > 0,\ ] ] and hence @xmath604 . here
@xmath605 } ( x_{\ell}(1 ) , \p x_{\ell}(1 ) ; [ \bar{g}|_{\p x_{\ell}(1 ) } ] )
> 0 $ ] denotes the relative yamabe constant of @xmath606)$ ] ( see @xcite ) and @xmath607 } ( z_{\ell } \times [ 1 , \infty ) , z_{\ell } \times \{1\ } ; [ \bar{g}|_{z_{\ell } \times \{1\ } } ] ) \\ \\ & = & y^{\cyl}_{[h + dt^2]}(z_{\ell } \times [ 1 , \infty ) , z_{\ell } \times \{1\ } ; [ \bar{g}|_{z_{\ell } \times \{1\ } } ] ) > 0 \end{array}\ ] ] denotes the relative cylindrical yamabe constant of @xmath608)$ ] ( see ( * ? ? ?
* section 4.1 ) ) , respectively .
> from theorem [ s3_t1 ] , corollary [ s3_co1 ] and fact [ s4_f1]-(3 ) , in order to prove the assertion * ( 1 ) * , it is enough to show that there exists a characteristic element @xmath609 $ ] satisfying the assumption * a1 * for the cylindrical metric @xmath610 in fact [ s4_f1]-(2 ) and @xmath611 .
note that @xmath90 has a natural complex structure .
then let @xmath612 \cong { { \mathbf z}}\ ] ] be the generator satisfying @xmath613 , and set @xmath614 .
it is easy to see that @xmath615 mod @xmath71 .
then , there exist the @xmath616-harmonic @xmath71-form @xmath617 and a sequence of closed @xmath71-forms @xmath618 such that @xmath398=[\zeta ] = a \in h^2(x_{\ell } ; { { \mathbf z } } ) , \\ \zeta_j\to \zeta \ \
\mbox{in $ l^{k,2}_{\bar{g}_0}$-topology for any $ k\geq 1$}. \end{array } \right.\ ] ] now we consider the determinant @xmath63-line bundle @xmath619 and the associated plus / minus spin bundle @xmath620 .
then we may assume that , as hermitian vector bundles , @xmath621 where @xmath622 denotes the canonical projection .
moreover , there exist @xmath74-connections @xmath623 with compact support such that @xmath624 .
now we denote by @xmath625 the corresponding twisted dirac operator .
recall that @xmath575 , and hence @xmath626 .
then the hirzebruch signature formula ( cf .
* example 4.1.9 ) ) combined with @xmath627 gives @xmath628 here @xmath629 stands for the eta invariant of the dirac operator @xmath630 and @xmath631 the spin dirac operator on @xmath632 , respectively .
_ indeed , we have @xmath634 where @xmath635 is the signature defect ( see @xcite or ( * ? ? ?
* ( 4.1.34 ) ) ) .
note that @xmath636 ( see @xcite or ( * ? ? ?
* example 4.1.27 ) ) .
then we have @xmath637 & \\ = & \frac{1}{8\ell } \left ( \ell^2 + 4\ell+4 - 4\ell + 4 -\ell^2 \right ) = 1 . & \ \
\box \end{array}\ ] ] we also note that @xmath638 by the choice of @xmath67 . this completes the proof of the assertion * ( 1)*. by the remark after ( * ? ? ?
* proposition 6.5 ) , @xmath639}(z_{\ell } \times { { \mathbf r } } ) = \frac{y(s^4)}{|\gamma_{\ell}|^{1/2 } } = 8\sqrt{\frac{6}{\ell}}\pi .
\end{array}\ ] ] > from the continuity of yamabe constants ( * ? ? ?
* proposition 3.4 ) , remark 4.1-(4 ) , and that @xmath640}(s^3 \times { { \mathbf r } } ) = y(s^4 ) = 8\sqrt{6}\pi > 12\sqrt{2}\pi$ ] , we obtain that @xmath641}(z_{\ell } \times { { \mathbf r}})\ \ & \mbox{if } \ \ \ \ell\geq 2 \end{array } \right.\ ] ] for any metric @xmath584 which is sufficiently @xmath108-close to @xmath105 .
this completes the proof of the assertion * ( 2)*. [ s4_c2](cf .
* theorem b ) ) set @xmath642 . for any integers @xmath94 and @xmath643 , @xmath644}(z_{\ell}\times { { \mathbf r } } ) \!\ } \end{array}\!\!\!\!\!\!\!\!\!\!\ ] ] for any metric @xmath582 homotopic to @xmath105 in @xmath583 .
proof of corollary [ s4_c2 ] there exists a metric @xmath647 such that @xmath648 and @xmath649 .
then , similarly to the case of @xmath90 , we obtain that @xmath650 for any metric @xmath651 homotopic to @xmath105 in @xmath583 .
note that @xmath652 then we choose @xmath653 . similarly to the case of @xmath90 , we also obtain that @xmath654 , where @xmath655 denotes the corresponding dirac operator on @xmath656 ( see the argument given in @xcite ) .
note that @xmath657 , and hence theorem [ s3_t1 ] implies the estimate ( [ s4_eq1 ] ) . *
example 2 .
* let consider the open @xmath14-manifold @xmath658 with tame end @xmath659 and the connected sum @xmath660 .
we first note that @xmath661 \cong \
c}}\p^2 ; { { \mathbf z } } ) \cong { { \mathbf z}}.\ ] ] let @xmath662 denote a product metric on @xmath663 , where @xmath664 denotes the standard metric on @xmath665 of constant curvature one .
similarly to theorem [ s4_t1 ] , we also obtain the following .
it is not clear for us whether a metric @xmath667 homotopic to @xmath668 in @xmath669 satisfies the inequality @xmath670}((s^2
\times s^1 ) \times { { \mathbf r } } ) > 12\sqrt{2}\pi\ ( \ = y({{\mathbf
c}}\p^2)\ ) .\ ] ] let @xmath671 denote the standard product metric on @xmath672 , where @xmath673 stands for the circle of radius @xmath674
. then we note that ( cf .
@xcite ) @xmath675}(s^2 \times { { \mathbf r}}^2 ) = \lim_{r \nearrow \infty}y^{\cyl}_{[h_r + dt^2]}(s^2 \times s^1 \times { { \mathbf r}}),\ ] ] where @xmath224 denotes the euclidean metric on @xmath676 and @xmath677}(s^2 \times { { \mathbf r}}^2 ) : = \inf_{u \in c^{\infty}_c(s^2 \times { { \mathbf r}}^2)}q_{h_s + g_0}(u)\ ] ] ( cf .
@xcite ) . if @xmath678}(s^2 \times { { \mathbf r}}^2 ) > 12\sqrt{2}\pi$ ] , then the estimate @xmath679 ( coming from theorem [ s3_t1 ] ) is effective for sufficiently large @xmath674 . 30 k. akutagawa , b. botvinnik , the relative yamabe invariant , * comm
. anal .
10 * ( 2002 ) , 925954 .
k. akutagawa , b. botvinnik , manifolds of positive scalar curvature and conformal cobordism theory , * math .
324 * ( 2002 ) , 817840 .
k. akutagawa , b. botvinnik , yamabe metrics on cylindrical manifolds , to appear in * geom .
* k. akutagawa , b. botvinnik , o. kobayashi , h. seshadri , the weyl functional near the yamabe invariant , * j. geom .
13 * ( 2003 ) , 120 .
m. atiyah , v. patodi , i. singer , spectral asymmetry and riemannian geometry .
i , * math .
cambridge philos .
77 * ( 1975 ) , 4369 .
t. aubin , some nonlinear problems in riemannian geometry , * springer monographs in math . * , springer - verlag , 1998 .
a. besse , einstein manifolds , springer - verlag , 1987 .
b. booss - bavnbek , k. wojciechowski , elliptic boundary problems for dirac operators , birkhuser , 1993 .
j. dodziuk , sobolev spaces of differential forms and de rham - hodge isomorphism , * j. diff .
16 * ( 1981 ) , 6373 .
j. dodziuk , @xmath0-harmonic forms on complete manifolds , seminar on differential geometry ( ed . by s .-
yau ) , * ann . of math .
102 * , 291 - 300 , princeton univ . press , 1982 .
j. j. duistermaat , the heat kernel lefschetz fixed point formula for the spin - c dirac operator , * progress in nonlinear differential equations and their applications 18 * , birkhuser , 1996 . j. escobar , uniqueness theorems on conformal deformation of metrics , sobolev inequalities , and an eigenvalue estimate , * comm . pure appl .
43 * ( 1990 ) , 857883 .
m. gursky , four - manifolds with @xmath680 and einstein constants of the sphere , * math
318 * ( 2000 ) , 417431 .
m. gursky , c. lebrun , yamabe invariants and @xmath681 structures , * geom .
8 * ( 1998 ) , 965977 .
o. hijazi , a conformal lower bound for the smallest eigenvalue of the dirac operator and killing spinors , * comm .
104 * ( 1986 ) , 151162 .
n. hitchin , harmonic spinors , * adv . in math .
14 * ( 1974 ) , 155 .
m. kapovich , hyperbolic manifolds and discrete groups , * progress in math .
183 * , birkhuser , 2001 .
j. kazdan , f. warner , scalar curvature and conformal deformation of riemannian structure , * j. diff .
10 * ( 1975 ) , 113134 .
t. kawasaki , the index of elliptic operators over @xmath682-manifolds , * nagoya math . j. 84 * ( 1981 ) , 135157 .
o. kobayashi , scalar curvature of a metric with unit volume , * math .
279 * ( 1987 ) , 253265 .
r. kobayashi , einstein - khler @xmath682-metrics on open satake @xmath682-surfaces with isolated quotient singularities , * math .
272 * ( 1985 ) , 385398 . c. lebrun ,
yamabe constants and the perturbed seiberg - witten equations , * comm .
geom . 5 * ( 1997 ) , 535553 .
j. lott , eigenvalue bounds for the dirac operator , * pacific j. math .
125 * ( 1986 ) , 117126 .
l. nicolaescu , seiberg - witten invariants of lens spaces , * canad . j. math .
53 * ( 2001 ) , 780808 .
l. nicolaescu , notes on seiberg - witten theory , * graduate studies in math .
28 * , amer .
soc . , providence , ri , 2000 .
m. obata , certain conditions for a riemannian manifold to be isometric with a sphere , * j. math .
japan 14 * ( 1962 ) , 333 - 340 .
j. petean , best sobolev constants and manifolds with positive scalar curvature metrics , * ann . global anal .
20 * ( 2001 ) , 231242 .
i. satake , the gauss - bonnet theorem for @xmath682-manifolds , * j. math .
japan 9 * ( 1957 ) , 464492 .
r. schoen , variational theory for the totatl scalar curvature functional for riemannian metrics and related topics , topics in calculus of variations ( montecatini terme , 1987 ) , 120 - 154 , * lecture notes in math .
1365 * , springer - verlag , 1989 .
r. schoen , s .-
yau , conformally flat manifolds , kleinian groups and scalar curvature , * invent .
92 * ( 1988 ) , 4771 . | we study the yamabe invariants of cylindrical manifolds and compact orbifolds with a finite number of singularities , by means of conformal geometry and the atiyah - patodi - singer @xmath0-index theory . for an @xmath1-orbifold @xmath2 with singularities @xmath3 ( where each group @xmath4 is of finite order )
, we define and study the _ orbifold yamabe invariant _ @xmath5 .
we prove that @xmath5 coincides with the corresponding @xmath6-@xmath7 @xmath8 defined by the authors @xcite , where @xmath9 is the standard metric on the slice @xmath10 of each end with infinity @xmath11 . using this ,
we show that @xmath5 is bounded by @xmath12 from above , where @xmath13 . for a cylindrical @xmath14-manifold @xmath15 with a general slice metric @xmath6 on the end
, we also establish a method for estimating the @xmath6-cylindrical yamabe invariant @xmath16 from above , in terms of the geometry and topology of @xmath15 .
we conclude by an explicit estimate of @xmath16 for particular cylindrical @xmath14-manifolds @xmath15 , including that of @xmath5 for @xmath14-orbifolds @xmath2 . |
recent years has seen a tremendous progress in the ability to study the physics of atoms trapped in confined optical lattices .
a significant development has been the tuning of the interactions ( scattering lengths ) between the particles via feshbach resonances.@xcite a further important achievement is the ability to confine just a few atoms in a potential well or trap.@xcite the situation of a few particles confined in this fashion is also encountered in other areas of physics
. one old problem in nuclear physics is for example that of two or more nucleons outside a closed shell .
the recent experimental research in atomic physics that enables a control of interacions as well as the environment represented by the trap has stimulated theoretical studies of this basic quantum - mechanical problem : two or more particles interacting while trapped in some specified potential well . in this paper the investigation is restricted to two particles in either a 3-dimensional h.o .
trap or in a spherical square - well trap . of particular interest in a study of this kind
is of course the two - body interaction between the particles .
it can be represented by some potential model e.g. psudo - potential @xcite or chiral effective field theory potential@xcite .
it is however found that the interactions , under specific asumptions , can be related to scattering data and this is the route taken in this report .
an expression first derived by busch et al @xcite relates the _ energy - spectrum of two point - like particles , interacting in a 3-d harmonic oscillator ( h.o . ) well , to scattering phase - shifts .
the derived formula is restricted to point - like particles , meaning that the range of the interaction has to be short compared to the size of the trap and is then the solution of an exactly solvable model .
the pointlike interaction implies that the relative angular momentum @xmath8 . the formula has been extended to @xmath9 assuming that the range of the interaction is still small .
@xcite _ in the present work , restricted to @xmath8 , the expression for the two - particle energy spectrum as derived in ref .
@xcite is used to find the _ shifts in energy - levels from the non - interacting to the interacting spectrum .
it is shown that these shifts converge to the phase - shift approximation for large h.o .
quantum - numbers @xmath1 and for small as well as large scattering lengths .
calculations , shown in sect .
2 , moreover show that the phase - shift approximation and the exact result for the two - body spectrum obtained by busch et al . give practically indistinguishable results . _ the good agreemnet with the phase - shift approximation for the two - body spectrum may be unique to the h.o . trap . for comparison , in the similar case of two particles interacting in a spherical square well the calculations presented in sect .
3 shows that the phase - shift approximation is viable only in the limit of small ( in units of box - radius ) scattering lengths and large @xmath1-values and it is in general a much worse an approximation than it is for the h.o . if the trap is a spherical square - well ( or box ) the shift in energy spectrum due to the two particle s mutual interaction can , as shown below in sect .
3 , easily be obtained without reference to a potential model.@xcite in the limit of a large size of the box this shift is @xmath10 , the scattering phase - shift at momentum @xmath11 .
( only angular momentum @xmath8 is considered here although most results are also valid for all values of @xmath12 . )
a more elaborate but instructive method of obtaining the energy - shift was used in early attempts of developping a many - body theory of nuclei .
brueckner initially assumed the in - medium ( effective ) two - body interaction to be @xmath13 .
@xcite the initial problem studied was that of an infinite system of nuclear matter for the purpose of calculating binding energy and saturation properties of this system .
the @xmath14 approximation came from the assumption that the reactance matrix @xcite , could be used as an approximation for the in - medium interaction in the many - body problem .
( the diagonal part of this matrix is @xmath13 ) .
the definition of the reactance matrix involves a principal value integration over a _ continuous spectrum .
it was assumed that the numerical summation over the closely spaced level - spectrum of nucleons in a very large ( infinite ) box would give an equivalent result .
this assumption was substantiated by reifman and dewitt@xcite .
it was however soon realized to be incorrect .
several authors ( one of them dewitt ) showed that the correct limit for two particles in a big box would yield an in - medium interaction @xmath10 rather than @xmath14 . @xcite and at low density this might not make any difference but with @xmath15 ( large scattering length ) , a situation often discussed in recent works , it obviously does . ]
_ the reactance matrix is part of scattering theory .
the nuclear matter problem assumes a box , although large but still finite with boundary conditions different from that of scattering theory .
the proofs in the papers refered to above differ , but the essential point is that the integration over the continuous spectrum vs the summation over the discrete spectrum of particles in a box differ , even though the level - spacing in a big box goes towards zero as the box - size increases .
the principal value integration relates to the scattering problem with boundary conditions different from that for a box where the wave - functions are zero at the edge of the box even in the limit of the box being infinitely large .
the exact statement of the results in the referenced papers is that the energy - shift @xmath16 due to the interaction of two - particles confined in a spherical box in the limit when the size of the box approaches infinity is given by @xmath17 where @xmath18 is the spacing between the levels of the unperturbed spectrum and @xmath19 this energy shift implies that the _ diagonal part of the effective interaction in a big box is given by @xmath20 _ in nuclear matter studies ( implying an infinite box ) this has been referred to as the phase - shift approximation a term adopted here .
it was for example used in early calculations on the neutron - gas@xcite .
many - body calculations do in general also require _ off - diagonal elements of the interaction , as in the calculation of the brueckner reaction - matrix . in some early nuclear matter studies ,
these were obtained by assuming @xmath21 to be a separable interaction with the diagonal given by eq .
( [ veff]).@xcite . _
it is of interest to compare the energy - shifts in a _
finite size box with those of the h.o .
results of these calculations are shown in sect .
units of energy and length are here chosen to be @xmath22 and @xmath23 respectively .
the total energy for two non - interacting particles with zero angular momentum in a 3-d h. o. well is then @xmath24 ( @xmath25 ) .
assuming the center = of - mass energy to be @xmath26 the energy of relative motion is @xmath27 . with the particles interacting , @xmath28 and with @xmath29 the energy spectrum
is obtained from @xcite @xmath30}{2\gamma[(3-\eta)/4 ] } \label{busch}\ ] ] where @xmath31 .
this expression for the energy - spectrum of two particles trapped in a h.o .
well has already played an important role in interpreting experimental data in atomic physics .
( e.g. refs.@xcite ) it is cited in numerous recent articles related to nuclear structure.(e.g .
it has been extended to relative angular momenta @xmath9 by suzuki et al @xcite .
it is here recast in a different form to explicitly show the energy-_shifts @xmath32 incurred by the two - body interaction .
one finds @xmath33 .
after substitution in eq .
( [ busch ] ) and using the reflection formula for @xmath34-functions @xcite @xmath35 one finds _ @xmath36 or @xmath37 where @xmath38 with @xmath39 eq .
( [ bush2a ] ) is equivalent to eq .
( [ busch ] ) , first derived by busch et al @xcite although here rewritten in terms of the shift @xmath16 instead of the energy @xmath40 . of particular interest here
is the function @xmath41 .
[ bush1 ] shows a rapid convergence of this function to its asymptotic value , i.e. @xmath42 for increasing values of @xmath43 , and in this limit eq .
( [ bush2a ] ) reduces to eq.([d ] ) , referred to as the phase - shift - approximation .- function in ref.@xcite one finds @xmath44 , indicating an appreciably slower convergence to the asymptotic value than the exact calculation that is shown in fig .
[ bush1 ] . ]
this also occurs , irrespective of the the value of @xmath41 , when @xmath45 and more importantly when @xmath46 in which case @xmath47 , a well - known result.@xcite .
a correction @xmath48 to the phase - shift approximation of @xmath16 ( i.e. for values of @xmath43 when @xmath49 ) is obtained from @xmath50 it is damped with decreasing values of @xmath51 and goes to zero when @xmath52 ( and when @xmath53 ) in accordance with above .
the left part of fig . [ bush2 ] shows the difference between the energy shifts calculated by eq .
( [ bush2a ] ) , and the phaseshift - approximation , eq .
( [ d ] ) , respectively for h.o .
quantum - numbers @xmath54 and @xmath55 .
results are shown as a function of scattering length @xmath56 in units of @xmath57 .
the difference ( error ) is seen to decrease rapidly with increasing @xmath1 and ( as also shown above ) with increasing value of @xmath58 .
the peaks are a result of the competition between the @xmath41 and @xmath51 terms respectively shown in eq .
( [ correct ] ) . the right part of the fig
. [ bush2 ] shows the same differences but in @xmath59 of the exact result again showing the rapid decrease with increasing values of @xmath1 and with @xmath58 .
one finds larger differences for @xmath4 , while appreciably smaller for @xmath60 and @xmath61 .
this is of course related to the function @xmath41 .
the smallest value of this function is for for the smallest value of @xmath62 ( see fig . [ bush1 ] ) , which occurs for @xmath4 @xmath63 , @xmath60 @xmath64 etc .
the largest discrepancy is consequently expected to be found for @xmath4 and for @xmath65 , where the shift is negative .
the comparison between the exact result given by eq .
( [ busch ] and the phase - shift approximation for this worst case is shown in fig .
[ bush4 ] .
it shows practically complete overlap between the two curves even for this worst case .
the correcting factor @xmath41 is also shown in this figure .
it is seen that @xmath66 for negative scattering lengths but approaches the value @xmath67 for positive scattering lengths which is consistent with eq .
( [ z ] ) . for comparison ,
calculations show that @xmath68 when @xmath60 and with complete overlap between exact and the approximate results .
the conclusion is that the phase - shift approximation for the effective interaction applies equally well or even better in the h.o .
case than it does in the square - well case shown below in sect .
3 . again , of course , in a many body environment ( three or more particles ) also off - diagonal elements are needed .
j. rotureau @xcite solves this problem by generating an interaction by a unitary transformation of the h.o .
two - particle spectrum given by eq .
( [ busch ] ) .
a main result of this investigation is that the energy spectrum of two particles interacting in a h.o .
trap can be simply constructed from the free - scattering phase - shifts as follows : @xmath69 where @xmath70 and @xmath71 are the perturbed and un - perturbed levels respectively with @xmath72 and @xmath73 .
as written , this result refers to two particles with point - like interactions and angular momentum @xmath8 .
it can be extended to @xmath9 but is then expected to be less accurate , as is also the case with the original formula eq .
( [ busch ] ) .
it is of interest to compare the above result obtained with a h.o .
trap with that of two particles in a spherical box of finite size as this may better correspond to some experimental setups .
the problem of two particles interacting in a box of _ infinite size was considered by several authors @xcite referred to in the introduction . with the result given by eq .
( [ d ] ) , the phase - shift approximation .
the special interest was , at the time , that of nuclear matter , a system of nucleons in a large but finite enclosure or of free electrons in a metal . _ it is of interest to compare the result obtained with a h.o .
trap with that of two particles in a spherical box of _ finite size .
this may better correspond to some experimental setups . _
the simplest solution @xcite of the problem at hand is to explicitly consider the wave - functions of the two particles in this box .
it is an example of an exactly solvable model in quantum mechanics.@xcite considering only @xmath74-states , the radial wavefunctions of free non - interacting particles are ( easily extended to angular momentum @xmath9 ) @xmath75 with the two particles interacting , the wave - function outside the range of the interaction ( assumed @xmath76 or pointlike ) is : @xmath77 the boundary condition implies that wave - functions vanish at the boundary of the sphere . with units of energy and length chosen to be @xmath78 and well - radius @xmath6
respectively one finds for the non - interacting case @xmath79 and for the interacting case @xmath80 the spacing between unperturbed levels is @xmath81 the energy shift is @xmath82 so that by eq .
( [ ds ] ) @xmath83 showing that for large @xmath1 and/or small @xmath0 @xmath84 which is the phase - shift approximation ( [ d ] ) .
[ bush3 ] shows the corrections @xmath85 to the phase - shift approximation as a function of scattering length @xmath56 and for some indicated values of quantum - number @xmath1 .
the phaseshift is here a function of @xmath56 and @xmath86 by eq .
( [ corr1 ] ) which is solved selfconsistently . a comparison with the analogous result for the h.o . in fig .
[ bush2 ] shows notable differences .
while the largest deviation from the phase - shift approximation in the h.o . case is @xmath87 ( and typically much smaller ) , it is in the square well case seen to be as large as @xmath88 . in the h.o .
the largest error is for small scattering lengths while the error goes to zero for large scattering lengths .
this is opposite to the situation encountered for the square well .
[ bush3 ] does however show the known result that @xmath89 when @xmath90 @xcite .
[ bush10 ] shows the two - body spectra calculated exactly compared with those obtained in the phase - shift approximation .
the agreement between the the two sets of curves increases with increasing @xmath1 as well as increasing @xmath6 , i.e. size of the box but a comparison with fig .
[ bush4 ] shows that the approximation works better in the h.o .
case for the low - lying states .
the validity of the phase - shift approximation , @xmath91 for the energy - shift due to the interaction of two partcles in a trap , where @xmath92 is the free particle scattering phase - shift and de the energy - level spacing of the unperturbed spectrum , has been investigated .
the traps considered were a 3-d h.o . and a spherical infinite square well ( box ) .
the square - well case was treated a long time ago as referenced above , with the main emphasis on the large size limit of the box with the result expressed by eq .
( [ d ] ) , the phase - shift approximation .
somewhat of a surprise was the finding that a not only somewhat similar situation exists for the h.o .
trap but that the approximation is so much more accurate for this trap in the region of trap parameters of interest for finite systems .
eq.([fin ] ) can therefore be considered as an alternative to the formula of busch et al @xcite .
in addition to being simple to use for calculating the two - body spectra in the h.o . it can also very simply be used to extract phase - shifts from experimental data .
the differences between the results for the two respective traps are seen by comparing fig .
[ bush2 ] for the h.o.trap with fig .
[ bush3 ] for the square - well trap .
while in the square well case , the differences ( errors ) by using the phase - shift approximation are as large as @xmath88 , the differences are much smaller and even practically zero in the h.o . case .
the accuracy of the approximation in the h.o .
case is also seen in fig .
[ bush4 ] with curves practically overlapping even in the worst case , @xmath4 . for comparison ,
the scenario is rather different in the square well case shown by fig .
[ bush10 ] where for the low - lying states the approximation is only valid for small values of @xmath93 .
another difference between the h.o .
and the box rsults is that , with @xmath94 the phase - shift approximation becomes perfect in the h.o .
case while this is not so in case of the box . in both cases
the approximation becomes good for small values of the scattering - lengths , which is not surprising as the born approximation would then be applicable .
the results presented here were for zero - range potentials , i.e. for zero effctive range i.e. with @xmath95 .
jonsell @xcite found the finite range - corrections to be small in the h.o .
the small range is of course in relation to the physical size of the system , i.e. the length parameters defined above .
this subject was also the subject of a recent publication.@xcite the question arises of course whether the phase - shift approximation has a more general validity e.g. for two particles in a _ non - spherical atomic nucleus or for particles in a cubical box .
it was however already illustrated above that the situation is rather different in the 3-d h.o .
case compared to the case of a spherical well .
the conceptual differences between scatterings and interactions in the spherical vs cubical box was already pointed out by dewitt in a detailed discussion of this matter .
quote@xcite:_only the spherical box ( with spherical waves ) is suitable for establishing a connection between single scattering processes and discrete spectrum theory .... but the problem of two atoms in a anisotropic h.o .
trap was considered in ref .
@xcite . a related problem that of the relation between the energy spectrum and scattering phases of two particles in a cubical box but with _ periodic boundary conditions shift was however solved by lscher@xcite . but
he problem of two atoms in anisotropic h.o .
trap was considered in ref .
_ _ _ 10 immanuel bloch , jean dalibard and wilhelm zwerger , rev .
* 80 * ( 2008 ) 885 .
m. khl , h .
moritz , t. stferele , k .
gunter , and t. esslinger , phys.rev .
* 94 * ( 2005 ) 080403 .
h. s. khler , arxiv:1110.0039 [ nucl - th ] brueckner , levinson and mahmoud , phys . rev .
* 95 * ( 1954 ) 217 .
roger g. newton , scattering theory of waves and particles .
mcgraw - hill book company alfred reifman and bruce s. dewitt phys .
* 101 * ( 1956 ) 877 .
n. fukuda and r.g .
newton phys .
* 103 * ( 1956 ) 1558 .
* 103 * ( 1956 ) 1565 .
riesenfeld and k.m .
watson phys . rev . *
104 * ( 1956 ) 492 . f. g. fumi , phil . mag .
* 46 * ( 1955 ) 1007 .
k. gottfried , in _ quantum mechanics , w.a benjamin 1966 .
gerald d. mahan , in _ many particle physics springer 2000 .
martin block and martin holthaus , phys .
a * 65 * ( 2002 ) 052102 .
k.a.brueckner , john l. gammel and joseph t.kubis , phys . rev .
* 1118 * ( 1960 ) 1095 .
sood and s.a .
moszkowski , nucl . phys . * 21 * ( 1960 ) 582 .
khler , nucl . phys .
* a415 * ( 1983 ) 37 .
thomas busch , berthold - georg englert , kazimierz rzazewski and martin wilkens , foundation of physics * 28 * ( 1998 ) 549 .
yang , j. rotureau , b. r. barrett and u. van kolck , preprint akira suzuki , yi liang and rajat k. bhaduri , phys . rev . * a80 * ( 2009 ) 033601 .
s. jonsell , few - body systems * 31 * ( 2002 ) 255 .
j. rotureau , eur .
. j. * 67 * ( 2013 ) 153
. m. abramowitz and i. a. stegun ( eds . ) _ handbook of mathematical functions with formulas , graphs and mathematical tables ( dover publications , new york , 1972 ) thomas luu , martin j.savage , achim schwenk and james p. vary , phys . rev.c
* 82 * ( 2010 ) 034003 .
martin lscher , nucl . phys .
* b354 * ( 1991 ) 531 .
t. stferle , h. moritz , k. gnther , m. khl and t. esslinger , phys .
* 96 * ( 2006 ) 030401 . c. ospelkaus , s. ospelkaus , l. humbert , p. ernst , k. sengstock and k. bongs phys.rev
* 97 * ( 2006 ) 120402 .
z. idziaszek and t. calarco , phys .
a * 71 * ( 2005 ) 050701(r ) _ _ _ | the energy - spectrum of two point - like particles interacting in a 3-d isotropic harmonic oscillator ( h.o . )
trap is related to the free scattering phase - shifts @xmath0 of the particles by a formula first published by busch et al .
it is here used to find an expression for the _ shift of the energy levels , caused by the interaction , rather than the perturbed spectrum itself . in the limit of high energy ( large quantum number @xmath1 of the h.o . )
this shift is shown to be given by @xmath2 , also valid in the limit of infinite as well as zero scattering length at all h.o .
energies .
_ numerical investigation shows that the shifts differ from the exact result of busch et al , by less than @xmath3 except for @xmath4 when it can be as large as @xmath5 .
this approximation for the energy - shift is well known from another exactly solvable model , namely that of two particles interacting in a spherical infinite square - well trap ( or box ) of radius @xmath6 in the limit @xmath7 , and/or in the limit of large energy .
it is in this context referred to as the _ phase - shift approximation .
_ it can be ( and has been ) used in ( infinite ) nuclear matter calculations to calculate the two - body effective interaction in situations where in - medium effects can be neglected .
it has also been used in expressing the energy of free electrons in a metal . *
harmonic oscillator trap and the phase - shift approximation * + h. s. khler + _ physics department , university of arizona , tucson , arizona 85721,usa _ + |
sp is grateful to a.j .
bray and h. chate for useful discussions .
skd is grateful to the university grants commission , india , for financial support in the form of a research fellowship .
mcc thanks the school of physical sciences , jnu for hospitality during the stay in which this work was begun .
99 m.c . cross and p.c .
hohenberg , . for example , see y. kuramoto , _ chemical oscillations , waves and turbulence _ , springer - verlag , berlin ( @xmath104 ) ; w. van saarloos , in _ spatiotemporal patterns in nonequilibrium systems _ ( ed . p.e .
cladis and p. palffy - muhoray ) , addison - wesley , reading , ma ( @xmath105 ) .
for example , see h. chate , nonlinearity * 7 * , 185 ( 1994 ) for the @xmath106 cgl equation ; h. chate and p. manneville , for the @xmath9 cgl equation .
k. binder , in _ materials science and technology , vol . 5 : phase transformations of materials _
cahn , p. haasen and e.j .
kramer ) , p. 405 , vch , weinheim ( 1991 ) .
bray , adv .
phys . * 43 * , 357 ( 1994 ) .
t. ohta , d. jasnow and k. kawasaki , .
bray and s. puri , ; h.toyoki , .
mazenko , , .
hagan , .
aranson , l.b .
aranson , l. kramer and a. weber , ; a. weber , l. kramer , i.s .
aranson and l.b .
aranson , .
pargellis , p. finn , j.w .
goodby , p. pannizza , b. yurke and p.e .
cladis , ; b. yurke , a.n .
pargellis , t. kovacs and d.a .
huse , .
bray and k. humayun , .
das , s. puri and m.c .
cross , in preparation .
0.5 cm * figure 1 : * evolution of the cgl equation from a small - amplitude random initial condition .
the evolution pictures were obtained from an euler - discretized version of eq .
( [ cgl ] ) with @xmath107 , implemented on an @xmath108-lattice ( @xmath24 ) .
the discretization mesh sizes were @xmath109 , @xmath110 ; and periodic boundary conditions were imposed in both directions .
the snapshots show regions of constant phase @xmath111 , measured in radians , with the following color coding : @xmath112 $ ] ( black ) ; @xmath113 $ ] ( red ) ; @xmath114 $ ] ( green ) .
the snapshots are labeled by the appropriate evolution times .
+ + * figure 2 : * ( a ) plot of @xmath28 $ ] vs. @xmath29 for @xmath8 and @xmath115 denoted by the specified symbols . the characteristic length scale , @xmath31 ,
is obtained from the square root of the inverse defect density measured directly from snapshots as shown in figure @xmath21 .
the numerical data shown here was obtained as an average over @xmath116 independent runs for @xmath23-lattices ( with @xmath30 ) .
the solid line has a slope of @xmath54 .
+ ( b ) plot of saturation length @xmath32 vs. @xmath59 for a range of @xmath7-values .
the corresponding values of @xmath41 ( as a function of @xmath7 ) are obtained from hagan s solution , cf .
figure @xmath116 of ref .
the solid line denotes the best linear fit to the numerical data .
+ + * figure 3 : * correlation function for the @xmath21-armed spiral solution when @xmath22 ( @xmath117 ) .
we plot @xmath76 vs. @xmath77 for different spiral sizes , @xmath118 denoted by the specified line - types .
these results are obtained from a direct numerical integration of eq .
( [ cr ] ) . + + * figure 4 : * numerical data for the correlation function @xmath98 vs. @xmath82 at @xmath99 for the cases @xmath8 and ( a ) @xmath22 ; ( b ) @xmath119 ; ( c ) @xmath101 .
the numerical data was obtained as an average over @xmath116 independent runs for @xmath23-lattices ( with @xmath30 ) .
the solid line refers to the numerical integration of eq .
( [ cr ] ) with @xmath27 as an adjustable parameter .
subsequently , the @xmath82-axis is scaled so that the point @xmath120 is matched for the numerical data and the analytical expression . | we present results from a comprehensive analytical and numerical study of nonequilibrium dynamics in the 2-dimensional complex ginzburg - landau ( cgl ) equation .
in particular , we use spiral defects to characterize the domain growth law and the evolution morphology
. an asymptotic analysis of the single - spiral correlation function shows a sequence of singularities analogous to those seen for time - dependent ginzburg - landau ( tdgl ) models with @xmath0 symmetry , where @xmath1 is even .
# 1#2#3 phys .
rev .
lett . *
# 1 * , # 2 ( # 3 ) # 1#2#3_ibid .
_ * # 1 * , # 2 ( # 3 ) # 1#2#3 j. stat .
phys . * # 1 * , # 2 ( # 3 ) # 1#2#3phys .
lett . a * # 1 * , # 2 ( # 3 ) # 1#2#3phys .
rev . a * # 1 * , # 2 ( # 3 ) # 1#2#3phys .
rev .
b * # 1 * , # 2 ( # 3 ) # 1#2#3phys .
rev .
e * # 1 * , # 2 ( # 3 ) # 1#2#3j . phys . a * # 1 * , # 2 ( # 3 ) # 1#2#3j .
chem .
phys . * # 1 * , # 2 ( # 3 ) # 1#2#3j .
phys . chem . *
# 1 * , # 2 ( # 3 ) # 1#2#3physica a * # 1 * , # 2 ( # 3 ) # 1#2#3physica d * # 1 * , # 2 ( # 3 ) # 1#2#3pramana j. phys . * # 1 * , # 2 ( # 3 ) # 1#2#3adv . phys . * # 1 * , # 2 ( # 3 ) # 1#2#3rev .
mod . phys . *
# 1 * , # 2 ( # 3 ) # 1#2#3nature * # 1 * , # 2 ( # 3 ) # 1#2#3siam j. appl .
math * # 1 * , # 2 ( # 3 ) # 1#2#3prog .
. phys . * # 1 * , # 2 ( # 3 ) i much recent interest has focused on pattern formation in the complex ginzburg - landau ( cgl ) equation : @xmath2 where @xmath3 is a complex order - parameter field which depends on space ( @xmath4 ) and time ( @xmath5 ) . in eq .
( [ cgl ] ) , @xmath6 and @xmath7 are real parameters .
the cgl equation arises in a range of diverse contexts , as reviewed by cross and hohenberg @xcite .
this universality arises from the fact that the cgl equation provides a generic description of oscillations in a spatially - extended system near a hopf bifurcation @xcite .
the cgl equation exhibits rich dynamical behavior with variation of the parameters @xmath6 and @xmath7 , and the `` phase diagram '' has been investigated ( mostly numerically ) in various studies @xcite . in a large range of parameter space , the emergence and interaction of spiral defect structures characterizes the morphology . in this letter ,
we study the nonequilibrium dynamics of the cgl equation resulting from a small - amplitude random initial condition . in general , this nonequilibrium evolution is referred to as `` phase ordering dynamics '' or `` domain growth '' , and constitutes a well - studied example of far - from - equilibrium statistical physics @xcite .
our analytical understanding of phase ordering systems has depended critically upon modeling the dynamics of defects in these systems ( e.g. , interfaces , vortices , monopoles , etc . ) @xcite-@xcite . in this letter
, we use spiral defect structures to characterize the evolution morphology in the cgl equation .
many novel features emerge in our study , which should be of great relevance for both experiments and subsequent numerical simulations . for simplicity , we will focus on the cgl equation with @xmath8 and dimensionality @xmath9 .
however , the results presented here are also relevant for the cases with @xmath10 and @xmath11 , as the underlying paradigm does not change , i.e. , spirals continue to determine the morphology in large regions of parameter space . following the work of hagan @xcite , aranson _
et al . _
@xcite , and chate and manneville @xcite , let us briefly discuss the phase diagram of the @xmath9 cgl equation with @xmath8 .
the limiting case @xmath12 corresponds to the dynamical xy model , which is well understood .
the appropriate ( point ) defects are vortices , and domain growth is driven by the attraction and annihilation of vortex - antivortex pairs .
the relevant growth law for the characteristic length scale is @xmath13 @xcite ; and the analytic form of the time - dependent correlation function ( which characterizes the evolving morphology ) has been obtained by bray and puri , and ( independently ) toyoki @xcite . without loss of generality , we focus on the case with @xmath14 . for @xmath15 ( @xmath16 @xcite ) , spirals ( which are asymptotically plane - waves ) are linearly stable to fluctuations . for @xmath17 ( @xmath18 @xcite )
, spirals are linearly unstable to fluctuations , but the growing fluctuations are advected away , i.e. , the spiral structure is globally stable . finally , for @xmath19 ,
the spirals are globally unstable and can not exist for extended times @xcite .
our results correspond to the parameter regime with @xmath20 .
figure @xmath21 shows the typical evolution from a small - amplitude random initial condition for the case with @xmath22 .
our numerical simulations were performed by implementing an isotropic euler - discretization of eq .
( [ cgl ] ) on @xmath23-lattices ( @xmath24 for figure 1 ) , with periodic boundary conditions in both directions .
the discretization mesh sizes were @xmath25 and @xmath26 . in figure @xmath21
, we plot constant - phase regions and the relevant color - coding is provided in the figure caption .
the evolving morphology is characterized by spirals and antispirals , and there is a typical length scale @xmath27 , e.g. , inter - spiral spacing or the square root of inverse defect density , which is the definition we will use subsequently . figure 2(a ) plots @xmath28 $ ] vs. @xmath29 for 5 representative values of @xmath7 .
the length - scale data was obtained from 5 independent runs on @xmath23-lattices with @xmath30 .
after an initial transient period , the length scale @xmath31 should saturate to an equilibrium value ( @xmath32 ) because of an effective spiral - antispiral repulsive potential @xcite .
this should be contrasted with the @xmath33 case , where vortices continue to anneal ( at zero temperature ) as @xmath34 . as a matter of fact
, the data for @xmath35 in figure 2(a ) does not exhibit this morphological freezing on the time - scales of our simulation , though signs of the crossover are evident for @xmath36 . to understand this crossover behavior
, we recall the analytical solution for an @xmath37-armed spiral due to hagan @xcite : @xmath38 , \ ] ] where @xmath39 ; and @xmath40 , where @xmath41 is a constant which is determined by @xmath7 @xcite .
the limiting forms of the functions @xmath42 and @xmath43 are @xmath44 where @xmath45 is a constant which is determined by finiteness conditions .
we will focus on the case with @xmath46 , as only @xmath21-armed spirals are stable in the evolution @xcite .
furthermore , we are only interested in distances @xmath47 , where @xmath48 is the defect core .
thus , we consider the spiral form in eq .
( [ hgs ] ) with @xmath49 and @xmath50 ( appropriate for @xmath51 ) .
we expect that spirals behave similarly to vortices for @xmath52 , where @xmath53 .
thus , the early evolution should be analogous to that for the xy model , both in terms of the domain growth law and correlation function . in figure 2(a ) , the solid line has a slope of @xmath54 and the initial growth ( at least for @xmath55 ) appears to be consistent with the behavior for the xy model , i.e. , @xmath56 for @xmath9 .
we also expect the saturation length @xmath32 to scale with @xmath57 .
figure 2(b ) plots @xmath58 vs. @xmath59 for a range of @xmath7-values , and demonstrates that our numerical data is consistent with @xmath60 .
we can also obtain the scaling law for the crossover time and the corresponding numerical results ( not shown here ) are in agreement with it .
next , we consider the correlation function for the evolution morphology shown in figure @xmath21 .
it is obviously relevant to first consider the correlation function for a single spiral of length @xmath27 , as the snapshots in figure @xmath21 can be thought of as consisting of disjoint spirals of size @xmath27 .
( of course , this ignores modulations of the order parameter at spiral - spiral boundaries but we will discuss those later . ) we have approximated the @xmath21-armed single - spiral solution as @xmath61 $ ] .
the correlation function is obtained by considering the correlation between points @xmath62 and @xmath63 and integrating over @xmath62 as follows : @xmath64h(l-|\vec r_{1}+\vec r_{12}|),\end{aligned}\ ] ] where @xmath65 is the spiral volume ; and we have introduced the step function @xmath66 if @xmath67 .
the step function ensures that we do not include points which lie outside the defect of size @xmath27 .
it is convenient to introduce variables @xmath68 ; @xmath69 ; @xmath70 , to obtain @xmath71 h[1-(x^2+r^2 + 2xr \cos\theta)^{1/2}],\end{aligned}\ ] ] where we have used @xmath72 in @xmath9 . thus
, the scaling form of the single - spiral correlation function is @xmath73 . in general
, there is no scaling with the spiral size because of the additional factor @xmath74 .
we recover scaling only in the limit @xmath75 ( @xmath12 ) , which corresponds to the case of a vortex .
essentially , spirals of different sizes are not morphologically equivalent because there is more rotation in the phase as one goes out further from the core .
figure 3 plots @xmath76 vs. @xmath77 for the case with @xmath22 ( @xmath78 ) .
these results are obtained by a direct numerical integration of eq .
( [ cr ] ) .
we consider @xmath79 different values of @xmath27 .
the functional form in figure @xmath80 exhibits near - monotonic behavior for small values of @xmath27 ( i.e. , in the vortex or xy limit ) ; and pronounced oscillatory behavior for larger values of @xmath27 , as is expected from the integral expression .
notice that @xmath81 larger values of @xmath82 correspond to the point @xmath83 lying outside the defect .
the asymptotic behavior of the correlation function in the limit @xmath84 ( though @xmath85 ) is of considerable importance as it determines the tail of the momentum - space structure factor @xcite .
in particular , we are interested in the singular part of the correlation function as @xmath86 . in this limit , we can discard the step function in eq .
( [ cr ] ) as it only provides corrections at the edge of the defect .
the asymptotic analysis of the integral in eq .
( [ cr ] ) involves considerable algebra , which we will report in detail elsewhere . here ,
we confine ourselves to quoting the final result for the singular part of @xmath87 : @xmath88 eq . ( [ csing ] ) is one of the central results of this paper and we would like to briefly discuss its implications .
the leading - order singularity is the same as that for the xy model ( @xmath89 ) , @xmath90 @xcite , as expected .
however , there is also a sequence of sub - dominant singularities proportional to @xmath91 , @xmath92 , etc . , and these become increasingly important as the length scale @xmath27 increases .
these sub - dominant terms in @xmath93 are reminiscent of the leading - order singularities in models with @xmath0 symmetry , where @xmath1 is even @xcite . of course
, in the context of @xmath0 models , these singularities only arise for @xmath94 as there are no topological defects unless this condition is satisfied . in the present context ,
all these terms are already present for @xmath9 .
the implication for the structure - factor tail is a sequence of power - law decays with @xmath95 , where @xmath96 , etc .
thus , though the true asymptotic behavior in @xmath9 is still the generalized porod tail , @xmath97 , it may be difficult to disentangle this from other power - law decays .
finally , figure @xmath79 compares our numerical data for the correlation function with the functional form of the single - spiral correlation function .
recall that the correlation function does not scale with the characteristic length because of the spiral nature of the defects . in figures 4(a)-(c ) , we have plotted numerical data for @xmath98 vs. @xmath82 at @xmath99 , and @xmath100 . for the comparison with eq .
( [ cr ] ) , the length scale @xmath27 is taken to be an adjustable parameter . in each case
, the best - fit value of @xmath27 matches the length scale obtained from the inverse defect density ( see figure 2(a ) ) within 10 percent .
as is seen from figure @xmath79 , the single - spiral correlation function is in good agreement with the numerical data for the multi - spiral morphology upto ( approximately ) the first minimum . as a matter of fact
, the agreement is excellent ( perhaps fortuitously ) for @xmath101 , shown in figure 4(c ) . in the context of phase ordering dynamics
, the gaussian auxiliary field ( gaf ) ansatz @xcite-@xcite has proven particularly useful for the characterization of multi - defect morphologies .
we have critically examined the utility of the gaf ansatz in the present context @xcite and find that it is only reasonable at early times where , in any case , the ordering process is analogous to that for the xy model .
we are presently studying methods of improving the gaf ansatz for the cgl equation and will discuss this elsewhere .
more generally , the utility of the gaf ansatz arises from the summation over phases from many defects , which results in a near - gaussian distribution for the auxiliary field .
however , in the present context , the shocks between spirals effectively isolate one spiral region from the influence of other regions . as a matter of fact ,
the waves from other spirals decay exponentially through the shock and the phase of a point is always dominated by the nearest spiral .
therefore , we expect that the correlation function will be dominated by the single - spiral result in accordance with our numerical results . to summarize : we have undertaken a detailed analytical and numerical study of nonequilibrium dynamics in the cgl equation .
for early times ( @xmath102 ) , the domain growth process is analogous to that for the xy model , which is well understood . at later times ,
distinct effects due to spirals are seen and the evolving system freezes ( in a statistical sense ) into a multi - spiral morphology .
we have undertaken an asymptotic analysis of the correlation function @xmath87 for a single spiral .
it exhibits a sequence of singularities as @xmath103 .
furthermore , this correlation function is in good agreement with the numerical data for multi - spiral morphologies , over an extended range of distances . |
it is conjectured that every co - compact kleinian group contains a surface subgroup .
we show that , for some special examples , much more is true . [ main ]
let @xmath0 be a right - angled , compact coxeter polyhedron in @xmath1 , and let @xmath2 be the group generated by reflections in the faces of @xmath0 .
then every finitely - generated free subgroup of @xmath3 is contained in a surface subgroup of @xmath3 .
_ remarks : _ + 1 .
it is well - known that every such @xmath3 contains a surface subgroup . indeed ,
it was shown in @xcite that the number of `` inequivalent '' surface subgroups of @xmath3 grows factorially with the genus .
lewis bowen has recently applied theorem [ main ] to show that every such @xmath3 contains a sequence of surface subgroups for which the hausdorff dimensions of the limit sets approach two ( see @xcite ) .
given a free subgroup @xmath4 , we look at the convex core @xmath5 , which will be homeomorphic to a handlebody . replacing @xmath6 with a suitable nighborhood in @xmath1 , we can expand the handlebody to make it polyhedral , so that the boundary is a union of copies of the faces of @xmath4 . by expanding further , we can make the induced decomposition of the boundary finer and finer .
if we expand enough , it becomes possible to attach mirrors to certain faces along the boundary ( see figure 1 ) , in such a way that the resulting 3-orbifold is the product of a compact 2-orbifold with an interval . the desired surface group is a finite - index subgroup of the 2-orbifold group .
the first ingredient is the tameness theorem .
let @xmath4 be a free subgroup of @xmath3 .
then by @xcite and @xcite , the ( infinite - volume ) hyperbolic manifold @xmath7 is _ topologically tame _ , i.e. homeomorphic to the interior of a compact 3-manifold
. then work of canary ( @xcite ) implies that @xmath4 is geometrically finite i.e. if @xmath8 is the convex hull of the limit set of @xmath4 , then @xmath9 is compact .
the next step , based on the ideas of @xcite , is to give a polyhedral structure to @xmath9 .
let @xmath10 be the tesselation of @xmath1 by copies of @xmath0 , and let @xmath11 be the _ tiling hull _ of @xmath4 this is the union of all the tiles in @xmath10 which meet @xmath8 .
then @xmath11 is invariant under @xmath4 , and @xmath12 is a compact , irreducible 3-manifold with free fundamental group .
thus @xmath12 is a handlebody @xmath13 .
the tesselation @xmath10 induces a tesselation of @xmath14 .
since all dihedral angles of @xmath0 are @xmath15 , then every pair of adjacent faces in @xmath14 will meet at an angle of either @xmath15 or @xmath16 .
however , if two faces meet at an angle of @xmath16 , then we actually consider them as part of a single face .
thus , every face in @xmath14 can be decomposed as a union @xmath17 , where each @xmath18 is congruent to a face of the original polyhedron @xmath0 . along each @xmath18
, we may attach to @xmath13 a copy of @xmath0 , to obtain a handlebody with convex boundary containing @xmath13 , called the * expansion of w along f*. more generally , we define an * expansion * of @xmath13 to be a handlebody @xmath19 , obtained from @xmath13 by a finite sequence of such operations .
let @xmath20 be the genus of @xmath21 , and represent @xmath21 as @xmath22 , for a planar surface @xmath0 .
let @xmath23 be the boundary curves of @xmath24 .
say that a collection of faces @xmath25 of @xmath26 forms a * face annulus * if the faces can be indexed @xmath27 , where @xmath28 is adjacent to @xmath29 if and only if @xmath30 ( mod n ) , and @xmath31 . the last condition excludes the case of three faces meeting at a vertex .
we may assume , after an isotopy , that each face in @xmath14 meets @xmath40 in a collection of disjoint , properly embedded arcs .
let @xmath41 suppose @xmath42 .
let @xmath43 be the number of faces in @xmath14 which meet @xmath40 in @xmath44 components .
let @xmath45 such that @xmath46 , and let @xmath32 be the expansion of @xmath13 along @xmath4 . note that @xmath47 is a polyhedron @xmath48 ( made up of copies of @xmath0 ) with dihedral angles @xmath15 .
let @xmath49 be the face of @xmath48 which is identified to @xmath4 , and let @xmath50 be the faces in @xmath48 which are adjacent to @xmath49 , in cyclic order .
let @xmath51 , and let @xmath52 be the union of @xmath53 together with all faces in @xmath48 which meet faces in @xmath53 . since @xmath48 is a coxeter polyhedron in @xmath1 , it follows that @xmath54 is an embedded disk .
( of lemma [ gamma ] ) + + * case 1 : * there are four endpoints ( say @xmath66 ) which lie on four distinct sides of @xmath49 .
+ in this case , we let @xmath67 be a properly embedded arc in @xmath53 , disjoint from @xmath68 , which separates @xmath69 and @xmath70 ( see figure 2 ) . for each @xmath71 , let @xmath72 ( resp .
@xmath73 ) be an arc , properly embedded in some @xmath29 , so that one endpoint is on @xmath74 , the other is the point @xmath75 ( resp .
@xmath76 ) , and so that the arcs @xmath77 are all disjoint from each other and from @xmath67 .
let @xmath78 be the component of @xmath79 which is disjoint from @xmath67 .
let @xmath80 .
after an isotopy ( supported in a neighborhood of @xmath78 in @xmath81 ) the arcs @xmath59 satisfy the hypotheses of the lemma .
+ + * case 2 : * suppose that some edge of @xmath4 meets every arc @xmath58 .
+ we repeat the construction from case 1 .
( i.e. pick an arc @xmath67 in @xmath53 disjoint from the @xmath58 s , separating @xmath69 and @xmath70 ; then construct @xmath82 s , @xmath78 s , and @xmath59 s . )
the only difference is that we must arrange that the arcs @xmath83 are not all parallel ( i.e. their union meets at least three distinct sides ) , and that the arcs @xmath84 are not all parallel .
this can be done , since , @xmath48 being a right - angled coxeter polyhedron in @xmath1 , each @xmath85 has at least five edges .
( see figure 3 ) .
now we return to the proof of lemma [ annuli ] .
we obtain a loop @xmath86 in @xmath32 , by replacing each @xmath87 with @xmath88 .
let @xmath89 .
since the face @xmath4 has been removed , and replaced by faces which meet @xmath90 in fewer than @xmath44 components , we have @xmath91 .
similarly , we see that , by enlarging @xmath13 repeatedly , @xmath43 can be reduced until it reaches 0 . by further enlargements , we may assume that @xmath92 .
so we may assume that @xmath38 is connected for each @xmath4 .
let @xmath34 be the union of the faces which meet @xmath35 .
for each face @xmath4 in @xmath93 , let us define the * overlap * of @xmath4 by the formula : @xmath94 since the core curve of @xmath34 is essential in @xmath13 , no point in @xmath14 meets every face in @xmath34 .
thus , if @xmath95 for all @xmath96 , then the @xmath34 s are the disjoint face annuli we are looking for .
let @xmath4 be a face in @xmath34 , let @xmath97 and @xmath98 be the two faces in @xmath34 which are consecutive to @xmath4 , and let @xmath99 .
let @xmath100 and @xmath101 be the components of @xmath102 .
we say that @xmath4 is * good * if one of the @xmath59 s is disjoint from the interior of @xmath93 .
+ + * case 3 : * every face in @xmath93 is good .
+ let @xmath4 be a face in some @xmath34 , and let @xmath103 . by previous assumption , @xmath104 is connected .
let @xmath105 and @xmath106 be the endpoints of @xmath104 . as before ,
let @xmath32 be the enlargement of @xmath13 along @xmath4 , let @xmath107 , and let @xmath49 be the face of @xmath32 which is identified to @xmath4 .
let @xmath50 be the faces adjacent to @xmath49 in @xmath48 , labeled consecutively , so that @xmath108 and @xmath109 .
since @xmath4 is good , then we may assume that none of the faces @xmath110 is glued to a face in @xmath93 .
as in the proof of lemma [ gamma ] , we replace @xmath104 with an appropriate arc @xmath111 . in this case , we choose arcs @xmath112 ( resp .
@xmath113 ) from @xmath105 ( resp .
@xmath106 ) to @xmath114 , so that @xmath112 and @xmath113 each meet only one face of @xmath115 .
we let @xmath116 be the component of @xmath117 contained in @xmath110 ; then we perturb @xmath116 so that it is a properly embedded arc in @xmath118 . see figure 4 .
a complication is that @xmath119 may not be an embedded circle in @xmath48 , and thus there may be pairs of adjacent faces in @xmath48 which meet @xmath116 non - consecutively . in this case
, we perform `` shortcut '' operations on @xmath116 , as indicated in figure 5 .
let @xmath120 .
then we have the required arc @xmath121 , and a new loop @xmath122 .
the number of faces with positive overlap decreases , so eventually we may eliminate them all .
+ + * case 4 : * suppose there is a face @xmath4 in @xmath93 which is not good .
+ here the construction is similar to the construction of case 3 . in this case , we choose @xmath116 to be either of the two components of @xmath123 ; then we push @xmath116 off of @xmath114 ; and then , as in case 3 , we perform shortcuts if possible .
the result is that the face @xmath4 is removed , and replaced with good faces . repeating this operation along all faces which are not good
, we may reduce to case 3 .
returning to the proof of theorem [ main ] , we let @xmath124 be the group generated by @xmath4 , together with the reflections in the lifts to @xmath1 of the faces of the face annuli @xmath125 .
then we claim that @xmath124 is the group of a closed , hyperbolic 2-orbifold .
indeed , let @xmath126 be the orbifold with underlying space @xmath13 , and with mirrors on the faces of @xmath125 .
then @xmath126 is a hyperbolic 3-orbifold with convex boundary , and there is a local isometry @xmath127 , with induced map @xmath128 , so that @xmath129 . since @xmath126 has convex boundary ,
every element in @xmath130 is represented by a closed geodesic , and since @xmath71 takes geodesics to geodesics , it follows that @xmath71 is @xmath131-injective .
note that @xmath126 is equivalent to a product orbifold @xmath132 , where @xmath133 is the 2-orbifold with reflector edges corresponding to one of the components of @xmath134 .
thus @xmath135 is isomorphic to the orbifold fundamental group of @xmath133 .
the orientable double cover of @xmath133 is a 2-orbifold , @xmath136 , where the underlying space is an orientable surface of genus @xmath20 , and the cone points of @xmath137 all have order 2 .
if we identify @xmath124 with @xmath138 , then the loops generating @xmath4 all lift to @xmath137 , and so @xmath139 .
the group @xmath140 has a torsion - free subgroup , of index two ( if the number of cone points is even ) or four ( if the number of cone points is odd ) , containing @xmath4 .
this is the surface subgroup we were looking for
. 2 ian agol , `` tameness of hyperbolic 3-manifolds '' , arxiv : math/0405568 .
+ l. bowen , `` free groups in lattices '' , pre - print .
+ r. d. canary , `` a covering theorem for hyperbolic @xmath141-manifolds and its applications '' , _ topology _ * 35 * ( 1996 ) , no .
3 , 751778 .
+ d. calegari and d. gabai , `` shrinkwrapping and the taming of hyperbolic 3-manifolds '' , _
* 19 * ( 2006 ) , no .
2 , 385446 ( electronic ) . | we show that every finitely - generated free subgroup of a right - angled , co - compact kleinian reflection group is contained in a surface subgroup . |
it is notoriously difficult to determine distances to dwarf elliptical galaxies ( hereafter des , subsuming `` dwarf spheroidals '' , cf .
ferguson & binggeli 1994 ) .
their low gas content rules out hi 21cm - line observations , and their low surface brightness makes optical spectroscopy feasible only for the very brightest members of the class .
most of the available redshifts of des are for galaxies in the virgo cluster ( binggeli et al .
1993 ) and the centaurus cluster ( stein et al .
1997 ) , and reflect peculiar velocities rather than individual distances within the cluster . but in clusters there is no real need for individual distances : des are so abundant and strongly clustered ( see , e.g. , binggeli et al .
1987 for virgo , or thompson & gregory 1993 for coma ) that they must lie near the mean distance of the cluster . in poor groups and the field ,
where the distribution of des is sparse , individual distance information is absolutely indispensable .
( we recall that true `` field '' ( i.e. isolated ) des are apparently very rare : cf .
binggeli et al . 1990 ) . here
the only way to locate a diffuse , dwarf - like system and thus to unveil its physical nature is to estimate its distance . for nearby de candidates , it is possible to resolve the stellar population and derive the distance from a deep colour - magnitude diagram ( cmd , e.g. dacosta et al .
1996 , smecker - hane et al .
1996 , stetson 1997 ) , or from the tip of the red giant branch ( trgb , lee et al .
1993 , caldwell et al .
these methods are costly and time consuming . with hst ,
the cmd method is feasible out to @xmath10 4 mpc , and the trgb method out to @xmath10 12 mpc .
alternatively , the surface brightness - magnitude relation of des ( e.g. binggeli & cameron 1993 , jerjen & binggeli 1997 ) can be used to estimate the distance of a de out to perhaps 50 mpc . however , the uncertainty ( 1@xmath11 ) in the distance modulus for a single galaxy is large , typically @xmath12 0.7 mag , corresponding to a 40% distance error .
young & currie ( 1995 ) claimed that this error can be reduced to @xmath12 0.4 mag ( comparable to the accuracy of the tully - fisher method ) with a similar relation based on the shape of the de luminosity profile , but this claim appears to be incorrect ( binggeli & jerjen 1998 ) .
here we show that the well - known surface brightness fluctuation ( sbf ) method can be used to determine distances to dwarf ellipticals in the intermediate distance range of 110 mpc ( and possibly beyond ) with an accuracy of a few percent .
the sbf method was introduced by tonry & schneider ( 1988 , hereafter ts88 ) to measure distances to bright elliptical ( e ) galaxies .
the method is based on the discrete sampling of a galaxy image with a ccd detector and the analysis of the resulting pixel - to - pixel variance caused by unresolved stars . for a detailed description of the method the reader
is referred to ts88 and jacoby et al.(1992 ) .
application of the sbf method was subsequently extended to bulges of spiral galaxies ( tonry 1991 , luppino & tonry 1993 ) and to globular clusters ( ajhar & tonry 1993 ) .
a massive sbf observing programme is in progress to improve the cosmic distance scale ( tonry et al .
1997 ) .
the first , and up to now only , application of the method to _ dwarf _ ellipticals is due to bothun et al . (
1991 ) who observed the fluctuations in very large , nearly flat , low - surface brightness galaxies ( a special type of des ) in the fornax cluster of galaxies .
we show here that the sbf method is particularly well suited for des at a distance of a few mpc slightly beyond the distance where the galaxies would be resolved into stars .
in this paper we present a pilot sbf analysis for five dwarf ellipticals in the nearby sculptor ( scl ) group , also known as the south polar group .
these galaxies have recently been identified in an extensive survey of the southern scl and centaurusa ( cena ) groups for faint de members ( jerjen et al .
1998a : jbf98a ) .
special care has been taken with the calibration of the fluctuation magnitude @xmath3 , because the stellar population of dwarf ellipticals is quite distinct from that of the brighter ellipticals .
in fact , even among the local dwarf spheroidals there is a wide range of star formation histories .
nevertheless , we find that @xmath3 is sufficiently robust against such variations , at least for stellar systems dominated by old populations , and propose a first estimate of @xmath3 for des
. we will show that the sbf distances derived for the five des ( between 1.7 and 4.4 mpc ) , unambiguously identify these objects as sculptor group members .
moreover , the dwarfs neatly fall into place with the known substructure of the group .
we demonstrate the great potential of the sbf method applied to des as a tool to map the supergalactic structure in the local volume ( out to the virgo cluster ) . with our enlarged scl sample ,
we re - analyse the 3d distribution of these galaxies and confirm earlier findings ( tully & fisher 1987 , binggeli 1989 ) that the sculptor complex has a prolate structure and is part of a much larger cloud of galaxies which includes the local group ( lg ) .
the sbf results for the de sample from our cena group survey will be published elsewhere ( jerjen et al .
1998b : jfb98b ) .
following ct et al.s ( 1997 , hereafter c97 ) successful search of the scl group region for faint new dwarf irregular members , we conducted a similar visual search for faint de candidates using the same plate material .
seven very diffuse objects were identified which had de - like morphology and which were thus suspected to be dwarf spheroidal members of the scl group ( for the details of this survey , see jbf98a ) .
subsequent ccd imaging in the @xmath0 and @xmath1 bands showed two of these objects to be irregulars or background galaxies .
the five remaining galaxies of early - type morphology are the subject of the present sbf analysis .
they are listed in table [ tbl-1 ] with their basic photometric properties from jbf98a .
llcccccccc ngc 59 & ds0 & 00 15 25.1 & @xmath1321 26 38 & 11.90 & 25.9 & 21.0 & 1.07 & 0.09 & 1.04 scl - de1 ( sc22 ) & de & 00 23 51.7 & @xmath1324 42 18 & 16.94 & 19.3 & 25.4 & 0.79 & 0.06 & 0.77 eso294 - 010 & ds0/im&00 26 33.4 & @xmath1341 51 19 & 14.36 & 20.3 & 22.9 & 1.17 & 0.03 & 1.16 eso540 - 030 & de / im & 00 49 21.1 & @xmath1318 04 34 & 15.54 & 21.5 & 24.2 & 0.83 & 0.10 & 0.79 eso540 - 032 & de / im & 00 50 24.5 & @xmath1319 54 23 & 15.36 & 25.0 & 24.4 & 1.08 & 0.09 & 1.05 the photometric parameters span a wide range in total @xmath1 magnitude ( @xmath14 , column 5 ) , effective radius which contains half of the total light ( @xmath15 , column 6 ) , and mean surface brightness within the effective radius ( @xmath16 , column 7 ) . to get the extinction - corrected colour @xmath17 ( column 10 ) the foreground reddening e(@xmath18 )
was determined from the dust maps of schlegel et al .
( 1998 ) and converted with the ratio @xmath19 .
further details about the photometric data can be found in jbf98a .
= 6.5 cm = 6.5 cm + = 6.5 cm = 6.5 cm + = 6.5 cm = 6.5 cm images of the dwarfs are shown in figs .
[ fig1 ] and [ fig2 ] , and the classifications given in column 2 of table [ tbl-1 ] .
there is some morphological variety among early - type dwarfs ( cf . the atlas of virgo cluster dwarfs by sandage & binggeli 1984 ) . at the bright end of the luminosity function
, there is a variant of the pure de which was named ds0 , because it is often distinguished by a s0-like , two - component structure ( see also binggeli & cameron 1991 ) .
ngc 59 is clearly of this type , eso294 - 010 arguably so . at fainter magnitudes ,
des ( dwarf spheroidals ) are often hard to distinguish from smooth irregulars of type imv .
in fact , there may be a true evolutionary transition between gas - rich irregulars and gas - poor des ( e.g. ferguson & binggeli 1994 ) .
dwarfs with a mixed morphology ( appearing too smooth for a plain i m , but too lumpy for a pure de ) were called `` intermediate '' or `` ambiguous '' and classified de / im ( sandage & binggeli 1984 ) .
a well - known local example is the phoenix dwarf system ( van derydt et al .
1991 ) .
three of our objects seem to fall in this class and are classified accordingly .
eso540 - 030 and eso540 - 032 show a sprinkle of resolved stars in their central regions .
these stars may mark the presence of a small population of young stars .
eso294 - 010 , on the other hand , appears slightly lumpy in the centre ( see fig.[fig2 ] ) .
subsequent spectroscopy ( in sec.6 ) showed emission lines , so this is a clear example of a mixed type .
but even the smoothest des probably contain small amounts of intermediate - age or young stars ; this we know from the local dwarfs ( cf .
note also that the colours of our objects ( with @xmath20 , cf .
column 10 of table [ tbl-1 ] ) are rather blue for des , whose @xmath21 values range from 0.5 to 2.3 ( e.g. , evans et al . 1990 ) .
so the morphological deviations from a pure de are simply the manifestation of an underlying variety in the ( unresolved ) stellar populations .
one of the principal results of the present paper is that this variety does not cause difficulties for the application of the sbf method to des , because the fluctuation magnitude @xmath3 turns out to be rather insensitive to variations in the stellar contents within the observational constraints ( sec.5 ) . in the following
we will use the term `` de '' in a loose way , including all of the morphological variants discussed above .
the galaxies were observed in two runs , on september 1115 , 1996 , and august 2830 , 1997 , using the imager at the nasmyth b focus of the 2.3 m anu telescope at siding spring observatory . the detector was a tek 1k@xmath221k thinned ccd with a pixel size of 24@xmath23 m , yielding a scale of 0.6@xmath24 pixel@xmath9 and a 6.7@xmath25 diameter circle field of view .
the ccd gain was 1 e@xmath26/adu and the readout noise was 7 e@xmath26 .
we observed in the kron - cousins @xmath1-band , because our application of the sbf technique differs from previous studies which used @xmath27 or @xmath28 ; we are working on small , low - surface brightness objects so our data are photon - limited by the sky rather than by the galaxy itself .
the sky is significantly fainter in @xmath1 than in @xmath27 relative to the rgb stars which produce the surface brightness fluctuations [ @xmath29=1.5 versus @xmath30=0.8 ] .
another advantage of the @xmath1 filter is to avoid fringing which occur with thinned ccds beyond 7000 .
weather conditions were photometric and the seeing measured between @xmath31 .
for each of our galaxies a series of 4 to 5 exposures were taken each randomly offset by @xmath32 and of 450 to 600 sec duration , giving a total integration time of 1800 to 3000 sec .
flat fields were obtained every night from exposures of the morning and evening sky .
standard stars of graham ( 1982 ) were observed throughout each night for the photometric calibration .
processing of the ccd frames with iraf was carried out in the usual manner .
the overscan region provided the bias level , which was subtracted from the frame .
next , each science frame was flattened with a master twilight - sky flat , constructed from a median of 59 sky flats taken during the same night .
all reduced frames were flat to 0.1% .
the sky was modelled by fitting a plane to selected star - free areas well away from the galaxy .
while sky determinations can be quite difficult for giant ellipticals due to their extended halos and galaxy sizes comparable to the field of view , this is simple in our case .
finally , the sky subtracted images of a galaxy were registered and combined to a master frame in order to increase the signal - to - noise ratio and to remove artifacts , bad pixels , and cosmic rays . in fig.[fig1 ] we show the processed @xmath1-band images for four of our galaxies . @xmath0 and
@xmath1-band images of eso294 - 010 are shown in fig.[fig2 ] .
to prepare the master frames for the sbf analysis we essentially followed the procedure described in ts88 .
first , we employed daophot ( stetson 1987 ) routines to identify point sources which would disturb a 2d - fit of the mean galaxy light distribution . because of the small size of the galaxies and the high galactic latitude , there is only little contamination by foreground stars and background galaxies .
globular clusters are also a minor problem here , because globular clusters are rare in faint dwarf galaxies . however , as mentioned in sec.2 , two of our galaxies , classified as de / im , are contaminated with a few bright stars of their own .
these also had to be removed .
all identified point sources were replaced by a nearby patch of galaxy of the same surface brightness .
isophotes were computed for the cleaned galaxy image by fitting ellipses with variable radius , ellipticity , and position angle . these were used to model the mean galaxy surface brightness distribution , which was then subtracted from the original master frame . to normalise the variance amplitude of the stellar fluctuations across the residual image
, we divided the residual image by the square root of the mean galaxy surface brightness distribution . for each galaxy
the sbf analysis was carried out on two different square subimages ( field 1 and 2 ) with size between 30 to 60 pixels which were selected within the 27.5 @xmath1magarcsec@xmath33 isophote .
the overlap of the fields was kept as small as possible ( @xmath34% ) to get independent distance measurements .
special care was further taken to choose subimages with only few ( @xmath34 ) resolved point sources , galaxies , or other features daophot formerly had identified .
for example , the region of eso294 - 010 ( see sec.6 for more details ) was excised from the analysis .
remaining objects in the subimages were replaced by uncontaminated patches ( i ) randomly selected from the region outside of the subimage area and ( ii ) in the same surface brightness range .
this patching process was employed as alternative to the masking method described in ts88 .
the idea was to replace the few disturbed image parts with patches carrying the fluctuation signal .
as artificial periodicities and poor signal - to - noise data could be introduced in this way it is crucial to take into account criteria ( i ) and ( ii ) .
the number of pixels corrected never exceeded 5% of the total subimage area .
experiments showed that , at this percentage level , the patching method is uncritical for the signal we intend to measure . at this stage of the procedure
, we have images whose main sources of variance are the stellar fluctuations and a photon shot noise that is nearly uniform over the image ( the ccd readout noise is negligible ) .
to disentangle the two components we fourier - transformed the cleaned subimages and analysed their power spectra ( ps ) , which are shown in fig .
[ fig3 ] . having calculated the ps of the point spread function ( psf ) from well isolated stars on the master frame , the observed ps of the galaxy was modelled as a weighted combination of a psf - convolved component and a constant white - noise component : @xmath35 the free parameters @xmath36 and @xmath37 ( the two weights ) were determined by a least - squares fit of equ.1 to the data .
the fit was restricted to wave numbers @xmath38 to exclude the range which is affected by the galaxy subtraction .
the result of this decomposition for every galaxy field is shown in fig .
[ fig3 ] .
= 4.8 cm = 4.8 cm + = 4.8 cm = 4.8 cm + = 4.8 cm = 4.8 cm + = 4.8 cm = 4.8 cm + = 4.8 cm = 4.8 cm + the quantity @xmath36 is directly linked to the apparent fluctuation magnitude @xmath2 which is the luminosity - weighted average stellar luminosity of the stellar population and is approximately the apparent magnitude of a giant star in the galaxy which contributes to the sbf signal : @xmath39 the system ( telescope - filter - detector)-related quantity @xmath40 is the magnitude of a star yielding 1 adu per second on the ccd , and @xmath41 is the exposure time of the observation . from theory ( ts88 ) , we expect @xmath37 to be the variance of the photon shot noise @xmath42 divided by the square root of the number @xmath43 of exposures : @xmath44 where @xmath45 is the flux from the sky ( @xmath46 magarcsec@xmath33 at new moon ) and @xmath47 is the mean surface brightness of the galaxy on the subimage ( in counts ) . that equ.3 is approximately fulfilled can easily be verified from table [ tbl-2 ] where all relevant parameters are listed .
the sky brightness of individual images varied a few percent due to changes in the airglow intensity during the night .
column 6 gives the average sky value we measured for each series of images .
the indicated error corresponds to the uncertainty in the sky flatness of the masterimage .
rcccrrrcccc ngc 59 f1 & 24.30 & 5@xmath22450 & 1.4 & 836 & 4166(8 ) & 34.9(1.7 ) & 2.7(0.1 ) & 27.08(0.06 ) & 0.06 & 27.02(0.06 ) + f2 & & & & 706 & & 33.3(1.8 ) & 3.5(0.1 ) & 27.12(0.06 ) & & 27.06(0.06 ) + scl - de1 f1 & 24.31 & 5@xmath22600 & 1.5 & 101 & 4456(11 ) & 131.1(16.5 ) & 18.5(0.4 ) & 25.96(0.13 ) & 0.04 & 25.92(0.13 ) + f2 & & & & 77 & & 118.8(19.7 ) & 26.5(0.5 ) & 26.08(0.17 ) & & 26.04(0.17 ) + eso294 - 010 f1 & 24.30 & 4@xmath22450 & 1.5 & 329 & 3572(7 ) & 234.3(15.1 ) & 5.9(0.4 ) & 25.00(0.07 ) & 0.02 & 24.98(0.07 ) + f2 & & & & 234 & & 214.0(13.8 ) & 10.7(0.5 ) & 25.11(0.07 ) & & 25.09(0.07 ) + eso540 - 030 f1 & 24.32 & 5@xmath22450 & 1.5 & 193 & 3769(6 ) & 62.0(4.4 ) & 9.7(0.3 ) & 26.47(0.07 ) & 0.06 & 26.41(0.07 ) + f2 & & & & 96 & & 65.8(7.4 ) & 24.5(1.0 ) & 26.41(0.11 ) & & 26.35(0.11 ) + eso540 - 032 f1 & 24.27 & 5@xmath22450 & 1.3 & 123 & 4146(17 ) & 112.9(17.0 ) & 12.8(0.5 ) & 25.77(0.15 ) & 0.06 & 25.71(0.15 ) + f2 & & & & 95 & & 151.0(28.2 ) & 16.3(0.6 ) & 25.46(0.19 ) & & 25.40(0.19 ) + the error of a @xmath36 measurement is dominated by the fitting error and the error in the sky value . while the first accounts for 59% in all cases , the latter effect depends on @xmath47 going from a negligible 1% for the fields of ngc 59 to 18% for eso540 - 032 where the galaxy counts are few .
other possible sources of errors are the psf normalisation and the shape variation of the stellar psf over the ccd area . using different bright stars as template for the psf power spectrum we found the derived @xmath36 of a field changed by only 13% .
column 7 of table [ tbl-2 ] gives the values of @xmath36 together with the combined error from all discussed sources .
assuming further an accuracy in photometry to @xmath48mag ( column 2 ) and a random extinction error at the south galactic pole of @xmath49mag , we get an overall uncertainty for @xmath50 of 0.06 to 0.19mag ( column 11 ) . the error - weighted mean of @xmath50 for each galaxy derived from the two independent fluctuation magitude measurements are given in column 2 of table [ tbl-3 ] .
lrrrrr ngc 59 & 27.04(0.04 ) & 28.21(0.07 ) & 4.39(0.15 ) & @xmath51 & @xmath52scl - de1 & 25.96(0.10 ) & 27.13(0.12 ) & 2.67(0.16 ) & @xmath53 & @xmath54eso294 - 010 & 25.04(0.05 ) & 26.17(0.08 ) & 1.71(0.07 ) & @xmath55 & @xmath56eso540 - 030 & 26.39(0.06 ) & 27.52(0.08 ) & 3.19(0.13 ) & @xmath57 & @xmath58eso540 - 032 & 25.59(0.12 ) & 26.72(0.13 ) & 2.21(0.14 ) & @xmath59 & @xmath60
we chose to use @xmath1-band observations ( instead of the usual @xmath27 or @xmath28 ) to avoid fringing effects on thinned ccds and to take advantage of the relatively darker sky .
however , the drawback with this filter at present is the missing empirical calibration of @xmath61 in this photometric band .
most of the sbf applications have focussed on the @xmath27-band ( e.g. tonry et al . 1989 , 1990 ; tonry 1991 ; tonry et al .
1997 ) and @xmath28-band ( e.g. luppino & tonry 1993 ; pahre & mould 1994 ; jensen et al .
1996 ) . but even working in the @xmath27-band would not help in our case because all galaxies analysed to date are high - surface brightness , giant ellipticals , while our objects are low - surface brightness , dwarf galaxies .
the two classes are quite different in almost all of their properties , including stellar composition , although these differences are by no means well understood ( e.g. , ferguson & binggeli 1994 ) . in particular , @xmath61 strongly depends on the underlying stellar population of a galaxy , and it would be unclear how to adapt empirical results from e galaxies to des . to bypass this problem , we constructed a mean @xmath3 for des by applying stellar population synthesis models to the local , resolved dwarf spheroidals .
a recent comparison between theory and empirical results ( tonry et al .
1997 ) found very good agreement in the @xmath27-band when using the model predictions of worthey ( 1993a , 1993b , 1994 ) .
we therefore used worthey s on - line programworthey/ ] to calculate @xmath61 values for various photometric bands and different stellar mixtures composed of a series of single - burst stellar populations . for this procedure each population requires a set of input parameters : mean metal abundance , age , relative mass weight , and the slope of the imf . to use the program
, we first had to estimate the metal abundance of the des . for this we used the observed [ fe / h]m@xmath62 relation of local group des ( da costa 1994 , 1998 ) .
absolute visual magnitudes for our dwarfs were estimated by combining the extinction - corrected observed @xmath0 magnitudes , the mean @xmath63 colour of 0.75 ( bothun et al .
1989 ) , and an average distance modulus ( m m)=27.0 for the sculptor group ( c97 ) .
the resulting magnitudes , in the range @xmath64 , were converted into abundances according to the relation [ fe / h]@xmath65m@xmath66 ( derived from fig.7 of da costa 1998 ) .
as expected , our des are all metal poor systems , with metallicities in the range @xmath67[fe / h]@xmath68 .
as emphasised by da costa ( 1997 ) , des are not single - burst populations like a globular cluster .
rather , a diverse and complex set of star formation histories ( sfh ) is observed among the local dwarf spheroidals .
their stellar populations range from old ( ursa minor ) and mainly old ( e.g. tucana , leoii ) through intermediate - age episodic ( e.g. carina , leoi ) to intermediate - age continuous ( e.g. fornax ) . on the other hand , phoenix and lsg3
are classified as de / im , because they shows similarities to both dwarf spheroidals and dwarf irregulars .
these systems are dominated by an old metal - poor population with no evidence for _ major _ star formation activities after the initial episode 810 gyr ago . however , both systems have a minor population of young stars , with ages of about 150 myr , which makes these galaxies resemble dwarf irregulars
. this situation would make us suspect that @xmath69 for these galaxies is also diverse .
fortunately , this is not the case .
we will show that @xmath69 does not depend sensitively on the star formation history within the observed range .
this issue will be discussed further in jfb98b .
we used worthey s program to calculate @xmath3 values for a metal poor old ( age @xmath70gyr ) population with three different types of sfhs ( single burst , episodic , continuous ) .
seven sfhs were modelled with a series of star formation bursts , as given in table [ tbl-4 ] .
the bursts were taken to occur at equal intervals of time between the first and the last burst ( columns 24 of table [ tbl-4 ] ) .
the adopted initial metal abundance was @xmath71 or @xmath72 .
the enrichment per burst was taken to be constant ( column 6 ) , where the value of this constant was chosen such that the present - day abundances of the model spanned the range of abundances derived above for our galaxies .
a salpeter imf was adopted throughout . the relative mass weight per burst
was taken to be constant .
the resulting @xmath3 values are listed in column 9 .
lrcclccrc burst13 & 13 & 0 & 13 & @xmath71 & 0 & @xmath71 & 100 & @xmath73 burst11 & 11 & 0 & 11 & @xmath71 & 0 & @xmath71 & 100 & @xmath74 burst9 & 9 & 0 & 9 & @xmath71 & 0 & @xmath71 & 100 & @xmath75 episodical-1.9 & 13 & 1.5 & 8.5 & @xmath72 & 0.033 & @xmath71 & 25 & @xmath76 episodical-1.2 & 13 & 1.5 & 8.5 & @xmath72 & 0.267 & @xmath77 & 25 & @xmath78 continuous-1.9 & 13 & 0.5 & 8.5 & @xmath72 & 0.011 & @xmath71 & 10 & @xmath79 continuous-1.2 & 13 & 0.5 & 8.5 & @xmath72 & 0.089 & @xmath77 & 10 & @xmath80 the modelling reveals @xmath3 to be essentially insensitive to the changes in sfh for these populations older than @xmath81gyr .
all @xmath61 magnitudes are found in the narrow interval @xmath82 with a median value @xmath83 .
this value can be expected to be a good calibration constant for metal poor old populations as observed in ursa minor , tucana , or ngc 205 . if we assume that the morphological similarities between these lg dwarfs and scl - de1 and ngc 59 indicate similar star formation histories , then the median value from the models can be adopted for our two sample galaxies .
the remaining three scl dwarfs show some indication of a small population of young stars as found in phoenix . to estimate @xmath3 for these dwarfs , we analysed the old populations ( column 1 , table [ tbl-4 ] ) but polluted with a 5 and 10 percent ( in mass ) metal - rich intermediate and young population .
this second component was introduced in two different ways : ( 1 ) a constant star formation rate between 1 and 8 gyr simulated with a series of bursts every 1 gyr , ( 2 ) an episodic component with two bursts at 3 and 6 gyr .
each burst got equal mass weight and a metallicity of [ fe / h]@xmath84 ( this is the lowest metallicity allowed in worthey s program for bursts younger than 8 gyr ) . as in the pure `` old '' cases , @xmath3 is insensitive to the underlying star formation history of the pollution , i.e. continuous or episodical .
the important factor which changes @xmath3 ( see fig.4 ) is the mass ratio @xmath85 of the two subpopulations . applying equal weight to the results of all seven considered sfhs yields an average empirical relation @xmath86 . for eso294 - 010
, eso540 - 030 , and eso540 - 032 , we estimate the contribution of the young stars to the total integrated light to be less than 3% .
the typical mass - to - light ratio for the generated synthetic stellar systems is about 1.5 in the @xmath1-band , so we consider an old population with a young star pollution at the 5 percent level as a fair approximation for these three dwarfs .
to determine their distances , we will use the mean value from our models @xmath87 .
we are very grateful for the use of worthey s public program which enabled the work on @xmath3 described above .
we note however that star formation histories of des like fornax or carina could not be explored with this program .
both dwarfs produced a significant fraction of their metal poor ( [ fe / h]@xmath88 ) intermediate population between 4 and 8 gyr ( da costa 1997 ) .
this part of the parameter space , i.e. young ( @xmath89gyr ) and metal poor ( [ fe / h]@xmath90 ) , is not accepted by the model program at present .
as an independent , zero - order check on the intrinsic scatter of @xmath3 for an old metal poor stellar population , we explored the @xmath61[fe / h ] relation for globular clusters ( gc ) in @xmath91 and @xmath27 ( ajhar & tonry 1994 ) for the relevant metallicity range .
our primary interest here is not with the absolute values but the observed intrinsic dispersions . using all 11 gcs in their sample with @xmath67[fe / h]@xmath68 , we derived a mean scatter of @xmath92 and @xmath93 , respectively .
the exclusion of the anomalous cluster @xmath94cen reduces the scatter in the @xmath27 band to @xmath95 . assuming that the results for the @xmath1 band are comparable
, we conclude that ( 1 ) theory _ and _ observations predict a very small variation of @xmath61 in this metallicity range , ( 2 ) the empirical scatter depends only weakly on the photometric passband . for the following discussion , we will adopt @xmath96 ( model ) for our `` de '' classified dwarfs and @xmath87 ( model ) for the intermediate type `` de / im '' . in both cases an error of @xmath97 ( from gcs )
is adopted .
we regard this as a preliminary answer in the search for the empirical calibration of the sbf distance indicator for des .
interestingly , the results from the models are right between the average values of @xmath98 and @xmath99 for metal poor gcs with @xmath100 and @xmath101 , respectively ( ajhar & tonry 1994 ) . the distances to our galaxies
can now be determined in the usual manner : @xmath102)=0.2\cdot(\bar{m}_r^0-\bar{m}_r - a_r-25).\ ] ] the true ( extinction - corrected ) distance moduli , the distances in mpc , and the resulting absolute total magnitudes in @xmath1 and @xmath0 for the five scl dwarfs are listed in table [ tbl-3 ] . also given there are the uncertainties in these quantities .
we recall here the detailed origin of these errors : @xmath103 ( fitting , sky , psf variation ) , @xmath48mag ( calibration ) , @xmath104mag ( modelling ) , and @xmath49mag . the total sbf distance error for a single measurement is 5 to 10% .
the new sbf distances and their relevance for the 3d - structure of the sculptor group are discussed in sec.7 below .
first we report on a spectroscopic confirmation of the distance for one dwarf .
for an independent proof of group membership for at least one of our des , we selected our second brightest galaxy eso294 - 010 for a spectroscopic follow - up .
this galaxy was recently part of a study on the star formation histories of sculptor group dwarf galaxies ( miller 1996 ) .
miller found that eso294 - 010 is a typical early - type dwarf with no evidence for ongoing massive star formation .
no regions were discovered and the total h@xmath6 luminosity is very low . nevertheless , with a sufficiently long integration time we hoped to detect balmer absorption lines in the spectrum which then could be used to measure a redshift .
this galaxy is even more interesting because c97 reported a @xmath105 detection in at a redshift of 4450 but we measured the galaxy to be only 2.4 times more distant than m31 .
long - slit spectroscopy was carried out for eso294 - 010 in september 1997 .
we used the double - beam spectrograph ( dbs ) at the nasmyth a focus of the 2.3 m anu telescope to observe a blue and a red spectrum simultaneously .
the detectors were two site 1752@xmath22532 thinned ccds with a pixel size of 15@xmath23 m , and a scale of 0.9@xmath24 pixel@xmath9 across the dispersion .
the wavelength scales were 1.1 pixel@xmath9 in the blue and 0.55 pixel@xmath9 in the red .
beam splitter and grating angles were chosen to cover the wavelength ranges 35005500 and 60007000 , respectively .
the slit was positioned at the galaxy centre and aligned along the major axis .
the slit length was @xmath106 arcmin with a width of 2@xmath24 .
a series of four 2000 sec exposures were taken .
after each science exposure a ne - ar lamp was observed for the wavelength calibration .
iraf procedures were used for the spectral reduction .
the galaxy spectra were individually wavelength - calibrated and sky subtracted , and each set of blue and red spectra were combined to improve the signal - to - noise ratio .
[ fig5 ] shows slightly smoothed versions of the important parts of the spectra .
no significant balmer absorption lines are visible . however , narrow emission lines at [ ] @xmath1075007and h@xmath6 @xmath1076563are prominent with 5 and 10@xmath11 significance relative to the adjacent continuum rms variations ( the two apparent [ ] lines @xmath1076716 and @xmath1076731 are not significant ) . in view of miller s ( 1996 ) results mentioned above , this finding was unexpected , even though modest amounts of dust and gas in des are not uncommon , e.g. in ngc 185 and ngc 205 ( hodge 1973 , young & lo 1997 and references therein ) , or ngc 59 ( c97 ) .
in fact , there is a hint of lumpiness in the central part of eso294 - 010 , clearly seen in fig .
[ fig2 ] , which made us classify this galaxy _ a posteriori _ as an intermediate - type ds0/im ( see sec.2 ) .
tracking down the source of the emission , we noticed a circular patch 7 arcsec or 0.06 kpc across , located 18 arcsec south of the galaxy centre ( fig .
[ fig2 ] ) .
this patch coincides in position and extent with the position and width of the two emission lines along the spatial axis of the spectra , and is probably a small region .
the agreement of the velocities derived from the individual lines is excellent , with @xmath108 .
we suggest that 5 is a more realistic estimate of the velocity error .
this low velocity is in accord with our low sbf distance for that galaxy , and it will be shown below that it fits nicely into the velocity pattern of other scl group members .
the most important result of our sbf analysis is that all five objects fall in the distance range of previously known scl group members , thus confirming their suspected membership as group dwarfs .
moreover , the new de members help to clarify our view of the complex 3d - structure of the group . as it will turn out in the following , the notion of a scl `` cloud '' of galaxies would be a better description of this supergalactic structure .
the main members of the scl group are the six late - type ( sc , sd , sm ) spirals ngc 45 , 55 , 247 , 253 , 300 , and 7793 .
individual distances to these galaxies , determined from various methods , are listed in table [ tbl-5 ] , column 2 .
the rms uncertainties of the distances are given in parentheses .
most data are taken from puche & carignan ( 1988 , table ii ) .
where more recent measurements are available , we have calculated a mean value from the references given in column 3 .
one scl dwarf , the irregular sdig , which already had a known distance , is also included here . to this list
we can now add our 5 scl early - type dwarfs with their newly determined sbf distances .
lcllrrl ngc 55 & 1.66(0.20 ) & pc88 & 129(3 ) & 98 & 111 & dc91 eso294 - 010 & 1.71(0.07 ) & this study & 117(5 ) & 71 & 80 & this study ngc 300 & 2.10(0.10 ) & fr92 & 144(1 ) & 101 & 114 & eso - lv eso540 - 032 & 2.21(0.14 ) & this study & & & & ngc 247 & 2.48(0.15 ) & t87 , pc88 , f98 & 156(4 ) & 173 & 211 & dc91 scl - de1 & 2.67(0.16 ) & this study & & & & sdig & 2.63(0.80 ) & l77 , h97 & 229(10 ) & 217 & 238 & c97 ngc 253 & 2.77(0.13 ) & pc88 , f98 & 245(5 ) & 246 & 278 & dc91 eso540 - 030 & 3.19(0.13 ) & this study & & & & ngc 7793 & 3.27(0.08 ) & t87 , pc88 & 227(2 ) & 226 & 250 & eso - lv ngc 45 & 4.35(1.40 ) & pc88 & 463(3 ) & 488 & 525 & dc91 ngc 59 & 4.39(0.15 ) &
this study & 362(10 ) & 392 & 432 & c97 velocities ( where available ) are also given in table [ tbl-5 ] .
heliocentric velocities ( column 4 ) are taken from c97 , da costa et al .
( 1991 ) , and lauberts & valentijn ( 1989 ) for all but one galaxy ( eso294 - 010 ) for which the velocity was established in the present study ( see sec.6 ) .
these velocities were transformed into galactocentric velocities ( column 5 ) and velocities relative to the barycentre of the lg ( column 6 ) , applying the apex vectors given by de vaucouleurs et al .
( 1991 ) and karachentsev & macharov ( 1996 ) , respectively .
[ fig6 ] is a pseudo-3d plot of the distribution of the 12 scl galaxies with known distances listed in table [ tbl-5 ] : it shows the sky distribution of these galaxies , with the distance of every galaxy indicated by the symbol size ( the distances and velocities are also given as numbers ) .
we first note the large depth spanned by the galaxies : distances range from 1.7 mpc ( ngc 55 ) to 4.4 mpc ( ngc 45 ) .
in fact , puche & carignan ( 1988 ) excluded the distant ngc 45 from their study because they insisted on a gravitationally bound scl group with a suitably low mass - to - light ratio . here
we relax any dynamical requirements because , judged from the distribution of galaxies , such a cut seems artificial .
this will become clearer below .
we further note the close match in distance of eso294 - 010 with ngc 55 at the near end ( 1.7 mpc ) , and of ngc 59 with ngc 45 at the far end ( 4.4 mpc ) of the scl complex ( despite the large distance error for ngc 45 ) .
it seems very likely that the two dwarfs are bound companions to the major members ngc 55 and 45 .
both pairs are separated by less than 200 kpc in projection and have a radial velocity difference of less than 100 .
eso540 - 030 , eso540 - 032 , and scl - de1 , on the other hand , seem to be associated with ngc 247 and 253 at an intermediate distance of @xmath109 2.5 mpc .
eso540 - 032 could be bound to ngc 247 .
the other two dwarfs are too far off for a single companionship to either giant , but a bound substructure centred on ngc 247 and 253 is strongly suggested .
the close spatial coincidence of our five dwarf objects with the substructure of the scl group traced out by its main members demonstrates two things : ( 1 ) the sbf method to determine distances to early - type dwarfs is reliable and accurate to the claimed level .
( 2 ) the trend of `` field '' des to be satellites of giant galaxies is confirmed , in accord with the morphology - density relation for dwarf galaxies ( binggeli et al .
the only isolated pure ( old - population ) de known so far is the tucana system in the lg ( da costa 1994 , see also fig [ fig9 ] .
below ) . the only galaxy classified here as a pure de , scl - de1 , may also be fairly isolated among our dwarfs .
so far we have not considered the 14 ( ! ) additional dwarf irregular scl group members discovered by c97 , because of a lack of distance information .
however , we can include them in our analysis by using their observed radial velocities as distance indicators via the tight velocity - distance relation which will be discussed below .
dwarf irregulars are known ( and below also shown ) to avoid high - density regions ( cf .
again binggeli et al . 1990 ) and therefore to possess fairly unperturbed velocities .
[ fig7 ] is a pseudo-3d plot of the scl group , like fig .
[ fig6 ] but now with the galactocentric velocity as third dimension .
all c97 irregulars are included here , raising the number of scl galaxies to 23 .
we recognize again the ngc 55 subclump on the near side with eso294 - 010 , and possibly also sc18 as bound companions .
ngc 300 and sdig are slightly more distant .
unfortunately , the three des associated with ngc 247 and 253 ( fig . [ fig5 ] ) have no velocity and hence are lost here in fig . [ fig7 ] .
ddo006 and 226 with 300@xmath110400 must lie behind ngc 247 ( 173 ) . likewise , eso473 - 024 ( 571 ) is probably behind ngc 45/59 . on the other hand
, a possible pair of irregulars appears in the se ( lower left ) corner with ngc 625 ( 333 ) and eso245 - 005 ( 314 ) .
there are some very nearby dwarf irregulars , such as sc24 ( 61 ) and sc2 ( 98 ) , but also some very distant ones , such as eso348 - 009 ( 626 ) and eso347 - 017 ( 674 ) .
overall , the impression is that the dwarf irregulars are not strongly associated with the main scl group members but are scattered over the entire volume in the form of a `` cloud '' .
this is of course again in accord with the morphology - density relation for dwarf galaxies ( binggeli et al .
the whole scl group seems to have the shape of a cigar of length 3 mpc and thickness 1 mpc , which we see pole - on because the lg lies near its end .
a much better view of this can be gained if the cigar is turned to the side by plotting the distances linearly . for this
we have first to transform velocities to distances via a hubble diagram . in fig .
[ fig8 ] we show the hubble diagram for the total of nine galaxies with known distances and velocities , to which we have contributed eso294 - 010 and ngc 59 .
the velocities ( @xmath111 ) are relative to the centre of our galaxy .
the major scl group members show a tight linear relation between velocity and distance .
such a trend was already seen by puche & carignan ( 1988 ) and is now confirmed and amplified by the two new data point from the des ( filled squares )
. a maximum likelihood linear fit to the data yields : @xmath112= 119(\pm7 ) \cdot d [ { \rm mpc}]\,\,-136(\pm14)\,,\ ] ] where @xmath111 is taken to be the independent variable and the distance errors are given in table [ tbl-5 ] .
the mean scatter of the residuals about the line is remarkably small , with @xmath113mpc ( 0.23mpc without ngc 45 ) and @xmath114(27without ngc 45 ) .
large peculiar velocities due to internal group dynamics are clearly absent .
this strongly suggests that our scl galaxies form a `` cloud '' ( e.g. devaucouleurs 1975 , tully 1982 , tully & fisher 1987 ) , meaning a supergalactic structure of higher - than - average galaxy density which , in contrast to a `` group '' , is gravitationally _
unbound_. however , bound substructures , e.g. defined by one giant galaxy plus a swarm of dwarf companions , are often found embedded in such a cloud .
this also seems to be the case here .
the slope of the hubble relation of equ.5 is very steep ( 119mpc@xmath9 ) .
a plausible reason for this deviation from the global hubble law ( @xmath115 10 mpc@xmath9 ) is the gravitational influence of the lg .
the deceleration of the local expansion field ( for @xmath116 4 mpc ) by the lg was nicely recovered by sandage ( 1986 , 1987 ) who used his local galaxy data to put constraints on the mass of the lg . however , an attempt to fit one of sandages template curves for the local hubble diagram ( 1986 , his fig . 2 ) to our data ( fig .
[ fig8 ] ) proved unsuccessful . at present
we do not understand the dynamics underlying the scl cloud velocity field , and a dynamical analysis is beyond the scope of this paper .
because the velocity - distance relation of equ.5 is so tight , we can use it empirically to obtain approximate distances to those scl galaxies which have only velocities , i.e. the dwarf irregulars of c97 .
based on these distances we show , in fig . [ fig9 ] , the distribution of _ all _ scl cloud members in projection on to the supergalactic plane ( sgp ) .
this provides us with a side view of the cigar - shaped scl cloud , as it was described above .
the scatter of the third dimension here , the height above / below the sgp , is quite small ( @xmath117mpc ) .
we have also included lg `` members '' and galaxies at the outer fringes of the lg in this plot ( data taken from van den bergh 1994 ) , to demonstrate that the scl cloud and the lg may actually form _ one _ structure .
this conjecture was raised before by binggeli ( 1989 ) .
there is a marked asymmetry in the distribution of galaxies within a distance of 1.5 mpc . aside from the close companions to our galaxy and m31 , most lg dwarf members and suspected members ( ic1316 , wlm , ngc 6822 , phoenix , tucana , sagdig , ddo210 , pegasus , ic5152 , ... ) form a southern extension of the lg a bridge across to the scl cloud ( see also fig.2 in binggeli 1989 ) .
the whole structure can be described as a prolate ( 6 @xmath22 2 mpc ) cloud of ( unbound ) dwarf irregulars , in which a number of denser groups are embedded , each containing one or more massive galaxies plus a swarm of ( bound ) dwarf companions which tend to be of early type .
certainly our galaxy and m31 with their satellites form such condensations each alone , and together as the lg . in the scl region
we note again the three subclumps mentioned before : centred on ngc 55 ( plus ngc 300 ? ) , ngc 247/253 , and ngc 45 .
ngc 7793 may define a fourth subclump or may be member of the first ( davidge 1998 ) .
we note again the position of our five early - type dwarfs in this map : they tend to mark the condensations .
the dynamics of these scl subsystems with respect to each other , and with respect to the lg in particular , is not clear at present .
we note that the major scl group members are much less massive than m31 or our galaxy , and that m31 and the galaxy lie at the extreme end of the prolate cloud , but we do not know if this is physically significant . finally , it is worth mentioning that the scl - lg cloud is part of a yet larger structure a prolate cloud that stretches out to the comai cluster at @xmath10 15 mpc , in the direction of the virgo cluster .
the feature is indeed called the `` coma - sculptor cloud '' ( tully & fisher 1987 , see plate 14 ) .
we have shown that the surface brightness fluctuation method can be successfully used as a distance indicator for low surface brightness dwarf elliptical galaxies as faint as @xmath118 .
the application to the dwarf ellipticals rather than normal ellipticals has two advantages : ( 1 ) there are many des in the lg and its vicinity , and independent distances to these nearby des can be used to calibrate the absolute fluctuation magnitude @xmath3 , ( 2 ) through hst studies , we have detailed information about the variety of stellar content and star formation history of this galaxy type .
however , most of them are dominated by an old metal poor stellar population . in this situation
, first results from stellar population synthesis models suggest that @xmath3 is relatively insensitive to the sfh .
the internal fluctuation error for our des is 0.04 - 0.12mag which includes the results from comparison of two different fields of the same galaxy .
a small systematic uncertainty for @xmath3 comes from the unknown stellar content of the galaxy .
for example , @xmath3 gets 0.12mag fainter when going from a pure old population ( all stars older than 8 gyr ) to an old population contaminated by 10% of young population .
this trend was taken into account here by applying appropriate calibration constants to the different morphological galaxy types .
the overall distance error we derive is between 4 and 7% .
although a more detailed analysis is required to follow up the population issue , a very satisfactory hint at the reliability of the theoretical values comes from the fact that all des turned out to be close companions of scl group members for which independent distance measurements exist .
these parent galaxies cover the full 3mpc depth of the scl group . based on the results for our five des
, we further reviewed the kinematics and the 3d - distribution of the scl group .
surprisingly , the hubble flow at the group shows no evidence for disturbance due to gravitational interaction between the main group members but is possibly decelerated by the tidal force of the nearby , massive lg leading to a large local hubble constant of @xmath119mpc@xmath9 . taking advantage of the tight empirical velocity - distance relation we determined distances to dwarf irregulars in the scl region and compiled distances from the literature for galaxies in the local volume to illustrate that the lg and scl group are not isolated in space but are part of a large prolate cloud ( 6 @xmath22 2 @xmath22 1 mpc@xmath120 in size ) of dwarf irregulars well defined within the supergalactic plane .
a number of groups are embedded in this structure , each containing one or more massive galaxies accompanied by a population of bound dwarf satellites which tend to be of early type .
finally it is interesting to note that four of our de galaxies were already known prior this study .
it is their membership of the scl group which was uncovered here .
this is a clear signal that the search for nearby dwarf elliptical galaxies should not only be focussed on the discovery of new , very low surface brightness galaxies but also on a careful follow - up of already known galaxies which appear nearby from their morphology . for such follow - up work ,
the sbf method is a powerful tool for measuring the distance of a galaxy . for the future ,
this opens up the possibility to measure accurate distances to all known des in the distance range between 1 and 10mpc without the expensive requirement of resolving the stellar system into stars .
this will allow us to establish background - free de samples , to explore the galaxy content of medium and low - density regions of the nearby universe , and to study the physical properties of des in greater details .
furthermore , the existence ( or non - existence ) of isolated early - type dwarfs can be established as a constraint on the formation and evolution of galaxies .
we are grateful to john tonry and alan dressler for very helpful advice , to the referee who made some important suggestions , and to sylvie beaulieu for doing test observations of ngc 5128 which were used to establish the sbf method on low - surface brightness regions .
two of us ( hj and bb ) thank the swiss national science foundation for financial support .
hj thanks further the freiwillige akademische gesellschaft der universitt basel and the janggen - phn stiftung for their partial financial support . this project made use of the nasa / ipac extragalactic database ( ned ) , which is operated by the jet propulsion laboratory , caltech , under contract with the national aeronautics and space administration . | as part of an ongoing search for dwarf elliptical galaxies ( de ) in the vicinity of the local group ( jerjen et al .
1998a ) , we acquired deep @xmath0 and @xmath1-band images for five de candidates identified on morphological criteria in the sculptor ( scl ) group region .
we carried out a surface brightness fluctuation ( sbf ) analysis on the @xmath1-band images to measure the apparent fluctuation magnitude @xmath2 for each de . using predictions from stellar population synthesis models ( worthey 1994 ) giving @xmath3 values in the narrow range between @xmath4 to @xmath5 , the galaxy distances were determined .
all of these de candidates turned out to be satellites of scl group major members .
a redshift measurement of the de candidate eso294 - 010 yielded an independent confirmation of its group membership : the [ ] and h@xmath6 emission lines from a small region gave a heliocentric velocity of 117(@xmath7 ) , in close agreement with the velocity of its parent galaxy ngc 55 ( v@xmath8 ) .
the precision of the sbf distances ( 5 to 10% ) contributes to delineating the cigar - like distribution of the scl group members , which extend over distances from 1.7 to 4.4 mpc and are concentrated in three , possibly four subclumps .
the hubble diagram for nine scl galaxies , including two of our des , exhibits a tight linear velocity distance relation with a steep slope of 119 mpc@xmath9 .
the results indicate that gravitational interaction among the scl group members plays only a minor role in the dynamics of the group . however , the hubble flow of the entire system appears strongly disturbed by the large masses of our galaxy and m31 leading to the observed shearing motion . from the distances and velocities of 49 galaxies located in the local group and towards the scl group ,
we illustrate the continuity of the galaxy distribution which strongly supports the view that the two groups form a single supergalactic structure . |
for a compact surface @xmath3 the complex of curves @xmath4 introduced by harvey in @xcite , has vertices the isotopy classes of essential , non - boundary - parallel simple closed curves in @xmath5 a collection of vertices spans a simplex exactly when any two of them may be represented by disjoint curves , or equivalently when there is a collection of representatives for all of them , any two of which are disjoint .
analogously , for a @xmath6manifold @xmath7 the disk complex @xmath8 is defined by using the proper isotopy classes of compressing disks for @xmath9 as the vertices .
it was introduced in @xcite , where it was used in the study of mapping class groups of @xmath6manifolds . in @xcite , it was shown to be a quasi - convex subset of @xmath10 by @xmath1 we denote the @xmath6dimensional handlebody of genus @xmath11 recall that a compact connected surface @xmath12 with boundary is properly embedded if @xmath13 and @xmath14 is transverse to @xmath15 a _ compressing disk _ for @xmath14 is a properly embedded disk @xmath16 such that @xmath17 is essential in @xmath18 a properly embedded surface @xmath12 is _ incompressible _ if there are no compressing disks for @xmath18 recall also that a map @xmath19 \rightarrow h_{g}$ ] is a proper isotopy if for all @xmath20 , $ ] @xmath21 is a proper embedding . in this case
we will be saying that @xmath22 and @xmath23 are properly isotopic in @xmath1 and we will use the symbol @xmath24 to indicate isotopy in all cases ( curves , surfaces etc ) .
let @xmath25 be the simplicial complex whose vertices are the proper isotopy classes of compressing disks for @xmath2 and of properly imbedded boundary- parallel incompressible annuli and pairs of pants in @xmath26 for a vertex @xmath27 $ ] which is not a class of compressing disks , it is also required that @xmath14 is isotopic to a surface @xmath28 embedded in @xmath29 via an isotopy @xmath30 \rightarrow h_{g}\ ] ] with @xmath31 @xmath32 and @xmath33 being proper when restricted to @xmath34 a collection of vertices spans a simplex in @xmath25 when any two of them may be represented by disjoint surfaces in @xmath26 note that the class of properly embedded incompressible surfaces in @xmath1 is very rich .
for example , it contains surfaces of arbitrarily high genus ( see @xcite , @xcite ) which are not included as vertices in the complex @xmath25 defined above .
also observe that there exist properly embedded annuli and pairs of pants which are not isotopic to a surface entirely contained in @xmath15 the isotopy classes of such surfaces are also excluded from the vertex set of @xmath35 note that we may regard @xmath36 as a subcomplex of @xmath25 or , by taking boundaries of the representative disks , of @xmath37 note also that the vertices of @xmath25 represented by annuli correspond exactly to the vertices of @xmath38 represented by curves that are essential in @xmath29 but are not meridian boundaries .
we define the complex of annuli @xmath39 to be the subcomplex of @xmath25 spanned by these vertices .
together , the vertices of @xmath40 span a copy of @xmath41 in @xmath42 and we regard @xmath41 as a subcomplex of @xmath43 our goal is to show that for a handlebody @xmath1 of genus @xmath44 the automorphisms of the complex @xmath25 are all geometric , that is , they are induced by homeomorphisms of @xmath26 this can be rephrased by saying that the map @xmath45 is an onto map , where @xmath46 is the group of automorphisms of the complex @xmath47 and @xmath48 is the ( extended ) mapping class group of @xmath49 i.e. the group of isotopy classes of self - homeomorphisms of @xmath26 moreover , we will show ( see theorem [ main_theorem ] below ) that the map @xmath50 is 1 - 1 except when @xmath1 is the handlebody of genus @xmath51 in which case a @xmath52 kernel is present generated by the hyper - elliptic involution .
consisting of non - separating , non - meridian curves , @xmath53 for the proof of this result we perform a close examination of links of vertices in @xmath43 this examination establishes that an automorphism @xmath54 of @xmath25 must map each vertex @xmath55 in @xmath25 to a vertex @xmath56 consisting of surfaces of the same topological type as those in @xmath57 in particular , @xmath54 induces an automorphism of the subcomplex @xmath41 which permits the use of the corresponding result for surfaces ( see @xcite , @xcite ) .
it is a well known result that for genus @xmath58 the complex of curves @xmath41 is a @xmath59hyperbolic metric space in the sense of gromov ( see @xcite,@xcite ) . in the last section
we deduce that the complex @xmath60 is itself a @xmath59hyperbolic metric space in the sense of gromov .
moreover , it follows that @xmath46 does not contain parabolic elements and the hyperbolic isometries of @xmath61 correspond to pseudo - anosov elements of @xmath62 in a recent preprint of m. korkmaz and s. schleimer ( see @xcite ) , it was shown , in a more general context , that @xmath48 and @xmath63 are isomorphic .
apart from this isomorphism , our motivation for constructing the copmplex @xmath25 is the study of the mapping class group of a heegaard splitting in a @xmath6manifold @xmath64 this group ( originally defined for @xmath65 and often called the goeritz mapping class group ) consists of the isotopy classes of orientation preserving homeomorphisms of @xmath9 that preserve the heegaard splitting .
the mapping class group of a heegaard splitting is known to be finitely presented ( see @xcite , @xcite , @xcite ) only for @xmath66 and for a genus @xmath51 heegaard splitting .
we aim to examine the corresponding open questions for @xmath66 and heegaard splittings of genus @xmath67 as well as for certain classes of hyperbolic 3-manifolds . for these purposes ,
the complex @xmath25 is a suitable building block for defining a complex encoding the complexity of the goeritz mapping class group , because @xmath25 contains a copy of the curve complex of the boundary surface @xmath15 a @xmath6dimensional handlebody @xmath1 of genus @xmath0 can be represented as the union of a handle of index @xmath68 ( i.e. a @xmath6ball ) with @xmath0 handles of index @xmath69 ( i.e. @xmath0 copies of @xmath70 $ ] ) . for an essential simple closed curve @xmath71 in @xmath29 we will be writing @xmath72 $ ] for its isotopy class and the corresponding vertex in @xmath37 we will be writing @xmath73 $ ] for the corresponding vertex in @xmath39 where @xmath74 is the annulus corresponding to the curve @xmath75 provided that @xmath71 is not a meridian boundary .
we will be saying that @xmath73 $ ] is an _
annular vertex_. if @xmath71 is a meridian boundary we will be writing @xmath76 $ ] for the corresponding vertex in @xmath77 we will be saying that @xmath76 $ ] is a _ meridian vertex _ and @xmath71 a meridian curve .
a vertex in @xmath78 will be called a _ pants vertex_. consisting of a single non - separating meridian curve , and @xmath79 non - meridian curves , @xmath53 by writing @xmath72 \cap \left [ \beta \right ] = \varnothing$ ] for non - isotopic curves @xmath80 we mean that there exist curves @xmath81 $ ] and @xmath82 $ ] such that @xmath83 by writing @xmath72
\cap \left [ \beta \right ] \neq \varnothing$ ] we mean that for any @xmath81 $ ] and @xmath84 , $ ] @xmath85 by saying that the class @xmath72 $ ] intersects the class @xmath86 $ ] at one point we mean that , in addition to @xmath72 \cap \left [ \beta \right ] \neq \varnothing,$ ] there exist curves @xmath87 $ ] and @xmath84 $ ] which intersect at exactly one point .
the above notation with square brackets will be similarly used for surfaces .
if @xmath14 is an incompressible surface we will denote by @xmath88 \right ) $ ] the link of the vertex @xmath27 $ ] in @xmath89 namely , for each simplex @xmath90 containing @xmath27 $ ] consider the faces of @xmath90 not containing @xmath27 $ ] and take the union over all such @xmath91 we will use the notation @xmath92 to declare that two links are not isomorphic as complexes .
we will also use the classical notation @xmath93 to denote a surface of genus @xmath94 with @xmath95 boundary components .
in this section we will show that every automorphism of @xmath60 must preserve the subcomplexes @xmath39 and @xmath77 in particular , we will show that for @xmath27 \in \mathcal{i}\left ( h_{g}\right ) , $ ] the topological type of the surface @xmath14 determines the link of @xmath27 $ ] in @xmath25 and vice - versa . to do this
we will find topological properties for the link of each topological type of surfaces ( meridians , annuli and pairs of pants ) that distinguish their links .
it is well known that a pants decomposition for @xmath29 is a collection @xmath96 of @xmath97 essential , non - parallel , simple closed curves such that the closure of each component of the complement of these curves is a pair of pants .
the number of pairs of pants is @xmath98 thus , the maximal number of vertices in a simplex of @xmath25 is @xmath99 in other words the dimension of @xmath25 is @xmath100 to see that simplices of dimension @xmath101 actually exist , observe that there exists a pants decomposition @xmath96 so that each @xmath102 is a non - separating , non - meridian curve for all @xmath103 this is displayed in figure [ non_meridian_pants ] for @xmath104 and for @xmath105 see remark [ r5 ] below .
for such a choice of @xmath102 s , all @xmath106 pairs of pants formed by @xmath96 are incompressible surfaces .
apparently , all such pairs of pants give rise to distinct elements in @xmath35 thus , a pants decomposition @xmath96 with all @xmath102 s being non - meridian curves gives rise to @xmath97 annular surfaces @xmath107 these surfaces along with the @xmath106 pairs of pants formed by @xmath96 give rise to a simplex in @xmath25 containing @xmath108 vertices .
we have established the following the dimension of the complex @xmath25 is @xmath109 we next examine the dimension of @xmath110 \right ) $ ] when @xmath16 is a meridian and of @xmath111 \right ) $ ] when @xmath74 is an annular surface . ] ( 22,12 ) ( 102,88)@xmath112 ( 152,35)@xmath113 ( 160,137)@xmath114 ( 233,27)@xmath115 ( 293,88)@xmath116 ( 233,140)@xmath117 [ dim1]if @xmath74 is an annular ( incompressible ) surface then the link of the vertex @xmath73 $ ] in @xmath118 has dimension @xmath119 we first assume that @xmath71 is a separating curve
. then @xmath71 decomposes @xmath29 into surfaces @xmath120 and @xmath121 with @xmath122 and @xmath123 with @xmath71 being isotopic to the boundary of @xmath120 as well as to the boundary of @xmath124 to complete the proof in this case , it suffices to find a pants decomposition for @xmath29 consisting of non - meridian curves and containing the curve @xmath125 for the latter , it suffices to show the following claim : : @xmath120 can be decomposed into @xmath126 pairs of pants so that the boundary curves of each are non - meridian when viewed as curves in @xmath15 the first step is to find pair - wise disjoint non - separating curves @xmath127 in @xmath120 such that @xmath102 does not bound a disk in @xmath128for all @xmath103 to see this , let @xmath129 be two simple non - separating curves in @xmath29 such that the curves @xmath130 bound a pair of pants in @xmath15 as @xmath71 is not the boundary of a meridian in @xmath49 it is clear that @xmath131 can not both be meridian boundaries in @xmath26 assuming @xmath132 is not meridian boundary , we may cut @xmath120 along @xmath132 to obtain a surface @xmath133 by the same argument , we may find a non - separating curve @xmath102 in @xmath134 @xmath135 which is not meridian boundary . apparently , cutting @xmath120 along @xmath136 we obtain a sphere @xmath137 with @xmath138 holes , such that the boundary components of @xmath137 do not bound disks when viewed as curves in @xmath15 we now claim that we may find pair - wise disjoint curves @xmath139 such that @xmath140 does not bound a disk in @xmath128for all @xmath141 to see this , let @xmath142 be two simple closed curves in @xmath137 such that the curves @xmath143 bound a pair of pants and the curves @xmath144 bound a pair of pants as shown in figure 3 . if both @xmath145 bound properly embedded disks in @xmath49 say @xmath146 respectively , then @xmath147 is a properly embedded arc in @xmath1 which separates @xmath148 into two half - disks .
similarly for @xmath149 appropriate unions of these half - disks along @xmath147 establish a contradiction since none of @xmath150 is a meridian boundary .
thus , at least one of @xmath151 say @xmath152 does not bound a disk .
cutting @xmath137 along @xmath153 we obtain a pair of pants and a surface @xmath154 which has the same property as @xmath155 namely , all boundary components of @xmath154 do not bound disks when viewed as curves in @xmath15 by applying the same argument repeatedly , we may find the desired collection of curves @xmath139 none of which is a meridian boundary .
apparently , the collection of curves @xmath139 decomposes @xmath137 into @xmath126 pairs of pants as required .
this completes the proof of the claim and the proof of the lemma in the case @xmath71 is separating .
assume now that @xmath71 is non - separating . using two copies of @xmath71 and a simple arc joining them
we may construct a separating curve @xmath156 which decomposes @xmath29 into surfaces @xmath157 and @xmath158 with @xmath156 being isotopic to the boundary of @xmath157 as well as to the boundary of @xmath159 note that @xmath158 contains @xmath125 then by the above claim we have that @xmath157 can be decomposed into @xmath160 ( incompressible ) pairs of pants by using non - meridian curves @xmath161 @xmath162 contained in @xmath157 together with the cirve @xmath163 by adding the curve @xmath71 we obtain a pants decomposition @xmath164 with all curves being non - meridian .
hence , @xmath73 $ ] is contained in a simplex of maximum dimension , namely , of dimension @xmath101 which shows that the dimension of @xmath111 \right ) $ ] is @xmath119 [ dim2]if @xmath16 is a meridian then the link of the vertex @xmath165 $ ] in @xmath25 has dimension @xmath166 first assume that @xmath167 $ ] is non - separating .
we may find a pants decomposition @xmath168 for @xmath29 such that @xmath102 is non - meridian for all @xmath169 ( see figure [ meridian_pants ] ) .
this collection of curves decomposes @xmath29 into @xmath106 pairs of pants such that exactly two of these have @xmath17 as boundary component and , hence , they are compressible surfaces .
thus , a non - separating meridian @xmath167 $ ] is contained in a simplex with @xmath170 vertices and , hence , the dimension of @xmath171 \right ) $ ] is @xmath172 let now @xmath173 be any pants decomposition with corresponding pairs of pants @xmath174 such that one of them , say @xmath175 has two boundary components isotopic to @xmath176 then the third boundary component of @xmath177 will also be a meridian , thus , another pair of pants distinct from @xmath177 will also be compressible .
this shows that a class @xmath167 $ ] with @xmath16 non - separating meridian can not be contained in a simplex of more that @xmath178 vertices and , thus , @xmath110 \right ) $ ] is equal to @xmath166 if @xmath167 $ ] is separating , it is clear that any decomposition @xmath168 for @xmath2 with @xmath102 being non - meridian for all @xmath179 has the property that exactly two of the corresponding pairs of pants are compressible and we work similarly . [ non_iso]let @xmath167 $ ] be a meridian vertex , @xmath180 $ ] an annular vertex and @xmath181 $ ] a pants vertex .
then the links @xmath110 \right ) , lk\left ( \left [ s_{\alpha}\right ] \right ) $ ] and @xmath182 \right ) $ ] are pair - wise non - isomorphic as complexes . by the previous two lemmata , the links of the vertices @xmath167 $ ] and @xmath73 $ ] have distinct dimensions , hence , it is clear that
@xmath110 \right ) \ncong lk\left ( \left [ s_{\alpha } \right ] \right ) .$ ] it remains to distinguish @xmath182 \right ) $ ] from @xmath110 \right ) $ ] and @xmath183 \right ) .$ ] let @xmath181 $ ] be a vertex in @xmath60 such that @xmath184 is a pair of pants with boundary components @xmath185 the vertices in @xmath182 \right ) $ ] form a cone graph , that is , the vertex @xmath186 $ ] belongs to @xmath187 \right ) $ ] and is connected by an edge with any vertex in @xmath188 \right ) .$ ] we will reach a contradiction by showing that @xmath189 \in lk\left ( \left [ d\right ] \right ) , \exists \left [ r\right ] \in lk\left ( \left [ d\right ] \right ) : \left [ q\right ] \cap \left [ r\right ] \neq \varnothing \tag{$\ast$}\label{starp}\ ] ] and similarly for @xmath111 \right ) .$ ] for , if @xmath190 is a boundary component of a surface representing @xmath191 \in lk\left ( \left [ d\right ] \right ) $ ] then there exists a curve @xmath192 such that @xmath193 and @xmath194 let @xmath195 $ ] be the vertex represented by @xmath196 if @xmath192 is non - meridian and by @xmath197 if @xmath192 is a meridian boundary
. then @xmath195 \in lk\left ( \left [ d\right ] \right ) $ ] is the required vertex which is not connected by an edge with @xmath198 , $ ] thus @xmath110 \right ) $ ] satisfies property @xmath199 similarly , we show that @xmath200 \right ) $ ] also satisfies property @xmath199 [ r5]let @xmath201 be non - separating curves in @xmath202 decomposing @xmath203 into two components which we denote by @xmath204 note that @xmath205 may not be isotopic . to see this , denote by @xmath206 the generators of @xmath207 corresponding to the longitudes of @xmath208
we may choose non - separating curves @xmath80 on @xmath203 which represnt the second powers @xmath209 up to conjugacy .
choose an essential non - separating curve @xmath192 such that @xmath201 are mutually disjoint and non isotopic .
these curves separate @xmath210into two components ( pairs of pants ) @xmath184 and @xmath211 if @xmath205 were isotopic then @xmath212 would be homeomorphic to the product @xmath213 $ ] and any two of the boundary components of @xmath184 would give rise to generators for @xmath214 since neither @xmath215 nor @xmath216 are generators for the free group on @xmath206 it follows that , for this particular choice of @xmath217 the surfaces @xmath205 are not isotopic .
let @xmath45 be the map sending a mapping class @xmath33 to the automorphism it induces on @xmath89 that is , @xmath218 is given by @xmath219 : = \left [ f\left ( s\right ) \right ] .\ ] ] [ main_theorem ] the map @xmath220 is onto for @xmath44 and injective for @xmath221 for @xmath222 @xmath50 has a @xmath223kernel generated by the hyper - elliptic involution .
we will use the following immediate corollary of proposition [ non_iso ] .
[ non_iso_more]automorphisms of @xmath25 preserve all types ( meridian , annular and pants ) of vertices .
we will also need the following [ l1]if @xmath54 @xmath224 and @xmath225 then @xmath226 \right ) = \left [ s\right ] $ ] for any vertex @xmath27 \in \mathcal{i}\left ( m\right ) $ ] except in the case mentioned in remark [ r5 ] , namely , if @xmath105 and @xmath184 is a pair of pants with all boundary components of @xmath227 being separating curves decomposing @xmath203 into @xmath51 components @xmath205 , then either , @xmath228 \right ) = \left [ p\right ] $ ] or , @xmath229
\right ) = \left [ p^{\prime}\right ] .$ ] we have to show that @xmath230 fixes every vertex @xmath181 $ ] where @xmath184 is a pair of pants .
let @xmath181 $ ] be such a vertex in @xmath35 by corollary [ non_iso_more ] it is clear that @xmath228 \right ) $ ] is a vertex @xmath231 $ ] with @xmath232 being a pair of pants .
denote by @xmath233 @xmath234 @xmath235 the boundary components of @xmath184 and , similarly , @xmath236
@xmath237 @xmath238 for @xmath211 if @xmath239 \cap \left [ \alpha_{j_{0}}^{\prime}\right ] \neq \varnothing$ ] for some @xmath240 then the vertex @xmath241 $ ] is connected by an edge with @xmath181 $ ] and is not connected by an edge with @xmath231 .$ ] as @xmath242 $ ] is fixed by @xmath243 it follows that @xmath244 \right ) $ ] can not be equal to @xmath245 .$ ] thus , we may assume that @xmath246 \cap \left [ \alpha_{j}^{\prime}\right ] = \varnothing \mathrm{\ for\ all\ } i , j=1,2,3 .
\tag{$\ast \ast$}\ ] ] consider the following property : @xmath247 if property @xmath248 holds then @xmath249 unless @xmath105 and @xmath233 @xmath234 @xmath235 are all non - separating curves which decompose @xmath203 into @xmath51 pairs of pants ( cf . remark [ r5 ] ) which may or may not be isotopic . thus ,
if property @xmath248 holds then either @xmath244 \right ) = \left [ p\right ] $ ] or the exception in the statement of the lemma occurs .
we examine now the case where @xmath104 and property @xmath248 does not hold . by assumption @xmath250
we may cut @xmath29 along @xmath233 @xmath234 @xmath235 to obtain either * the surface @xmath184 and a surface @xmath251 ( if all @xmath233 @xmath234 @xmath235 are non - separating ) or , * the surface @xmath252 a surface @xmath253 and a surface @xmath254 for some @xmath255 ( if exactly one of @xmath233 @xmath234 @xmath235 is separating and the other two curves are non - isotopic ) or , * the surface @xmath184 and a surface @xmath157 ( if exactly one of @xmath233 @xmath234 @xmath235 is separating and the other two curves are isotopic ) or , * the surface @xmath184 and surfaces @xmath256 for some @xmath257 with @xmath258 ( if all @xmath233 @xmath234 @xmath235 are separating ) note that if @xmath184 is a pair of pants , it is impossible to have exactly two of its boundary curves @xmath233
@xmath234 @xmath235 being separating . in all cases , @xmath232 is contained in a surface of the form @xmath259 for some @xmath260 and @xmath261 mentioned above .
thus , we may find a non - meridian curve @xmath71 in @xmath29 such that @xmath262 \cap \left [ \alpha_{j_{0}}^{\prime}\right ] \neq \varnothing \mathrm{\ for\ some\ } j_{0}\in \{1,2,3\}.\ ] ] then , for the annular surface @xmath74 we have that @xmath263 $ ] is connected by an edge with @xmath181 $ ] and is not connected by an edge with @xmath231 .$ ] as @xmath180 $ ] is fixed by @xmath243 it follows that @xmath229 \right ) $ ] can not be equal to @xmath231 .$ ] this completes the proof of the lemma . [
proof of theorem [ main_theorem]]we will use the corresponding result for surfaces , see @xcite,@xcite , which applies to the boundary of the handlebody @xmath15 we first show that every @xmath230 is geometric . by proposition [ non_iso ] we know that @xmath264 and @xmath265 in particular , @xmath266 the restriction @xmath267 of @xmath54 on @xmath41 induces an automorphism of @xmath41 which by the analogous result for surfaces ( see @xcite,@xcite ) is geometric , that is , there exists a homeomorphism @xmath268 such that @xmath269 as @xmath270 maps @xmath8 to @xmath271 @xmath272 sends meridian boundaries to meridian boundaries .
it follows that @xmath272 extends to a homeomorphism @xmath273 we know that @xmath274 on @xmath41 and we must show that @xmath274 on @xmath43 this follows from lemma [ l1 ] which completes the proof that every @xmath275 is geometric . let @xmath276 since @xmath50 is shown to be onto , there exists a homeomorphism @xmath277 such that @xmath278 \right ) = f.$ ] this implies that @xmath279 and @xmath280 in particular , @xmath54 restricted to @xmath281 induces an automorphism @xmath282 of the complex of curves @xmath37 by @xcite , @xcite , there exists a homeomorphism @xmath283 such that @xmath284 such a homeomorphism is unique unless @xmath105 in which case the map @xmath285 has a @xmath223kernel generated by an involution of @xmath286 however , any homeomorphism of @xmath29 which extends to @xmath1 it does so uniquely ( see , for example , ( * ? ? ?
* theorem 3.7 p.94 ) ) , and therefore the map @xmath287 is injective unless @xmath105 in which case it has a @xmath223kernel .
we first establish hyperbolicity for @xmath43 as far as hyperbolicity is concerned , the 1-skeleton @xmath288 of @xmath25 is relevant .
@xmath289 is endowed with the combinatorial metric so that each edge has length @xmath290 apparently , we have an embedding @xmath291 with @xmath292 where the superscript @xmath293 always denotes @xmath294skeleton . we claim that this embedding is isometric . indeed , if @xmath295 , \left [ \alpha_{2}\right ] $ ] are distinct vertices with distance @xmath296 , \left [ \alpha_{2}\right ] \right ) $ ] in @xmath297 then the distance @xmath298 \right ) , i\left ( \left [ \alpha_{2}\right ] \right ) \right ) $ ] can not be smaller . for , if @xmath299 = i\left ( \left [ \alpha_{1}\right ] \right ) , \left [ s_{1}\right ] , \ldots,\left [ s_{k}\right ] = i\left ( \left [ \alpha_{2}\right ] \right ) $ ] is a sequence of vertices which gives rise to a geodesic in @xmath300 of length less than @xmath296 , \left [ \alpha_{2}\right ] \right ) , $ ] equivalently , @xmath301
\right ) , i\left ( \left [ \alpha_{2}\right ] \right ) \right ) = k < d_{\mathcal{c}}\left ( \left [ \alpha_{1}\right ] , \left [ \alpha_{2}\right ] \right)\ ] ] then for each @xmath302 consider @xmath140 to be any boundary component of @xmath303 it is clear that @xmath140 is disjoint from @xmath304 and @xmath305 therefore , the sequence @xmath306 , \left [ \beta_{1}\right ] , \ldots,\left [ \beta_{k-1}\right ] , \left [ \alpha_{2}\right ] $ ] is a segment in @xmath307 of length @xmath308 with @xmath309 , \left [ \alpha_{2}\right ] \right ) $ ] , a contradiction . for any vertex @xmath181
$ ] in @xmath310 we may find an annular vertex , namely , @xmath311 $ ] where @xmath227 is any component of the boundary of @xmath252 which is connected by an edge with @xmath181 .$ ] thus , @xmath289 is within bounded distance from @xmath312 since @xmath297 is @xmath59hyperbolic in the sense of gromov , so is @xmath313 an element @xmath314 is called pseudo - anosov when it restricts to a pseudo - anosov homeomorphism on @xmath15 the proof of the following proposition is immediate from the corresponding result for surfaces ( see ( * ? ? ? * prop . 4.6 ) ) along with the above mentioned fact that @xmath41 is cobounded in @xmath43 w. harvey , _ boundary structure of the modular group _ , riemann surfaces and related topics , in : proceedings of the 1978 stony brook conference , state univ .
new york , stony brook , n.y .
, 1978 , ann .
97 , princeton univ . press , princeton , n.j . , 1981 , pp .
245 - 251 . | for a genus @xmath0 handlebody @xmath1 a simplicial complex , with vertices being isotopy classes of certain incompressible surfaces in @xmath1 , is constructed and several properties are established . in particular , this complex naturally contains , as a subcomplex , the complex of curves of the surface @xmath2 . as in the classical theory ,
the group of automorphisms of this complex is identified with the mapping class group of the handlebody . |
at present , there are three indications in favour of neutrino oscillations @xcite , namely the solar @xmath2 deficit @xcite , the atmospheric @xmath3 @xcite deficit and the result of the lsnd experiment @xcite hinting at @xmath4 transitions . whereas
in the case of the first two indications several experiments agree on the existence of the effect , the third indication is found only by the lsnd collaboration .
therefore , in many analyses the lsnd result is left out . however
, if all three indications in favour of neutrino oscillations are confirmed , for three mass - squared differences of different orders of magnitude ( @xmath5 or @xmath6 , @xmath7 ev@xmath8 , @xmath9 ev@xmath8 ) one needs a minimum of four neutrinos , three active ones and a sterile one @xcite . in that case
a major revision of our picture of the lepton sector of the elementary particles would be necessary , with a mixing between the active and the sterile neutrinos ; i.e. , @xmath10 if we stick to the minimum of four neutrinos . in eq .
( [ mixing ] ) the left - handed flavour fields are denoted by @xmath11 and the left - handed mass eigenfields by @xmath12 , and the @xmath13 neutrino mixing matrix @xmath14 is assumed to be unitary .
one of the most important issues in the context of 4-neutrino scenarios is the question of the 4-neutrino mass spectrum @xcite .
there are two different spectral classes with very different properties : the first class contains four types and consists of spectra where three neutrino masses are clustered together , whereas the fourth mass is separated from the cluster by the mass gap needed to reproduce the lsnd result ; the second class has two types where two pairs of nearly degenerate masses are separated by the lsnd gap .
these two classes have been dubbed ( 3 + 1 ) and ( 2 + 2)-neutrino mass spectra , respectively @xcite .
the main difference between these two classes is that , if a ( 2 + 2)-spectrum is realized in nature , the transition into the sterile neutrino is a solution of either the solar or the atmospheric neutrino problem , or the sterile neutrino has to take part in both , whereas with a ( 3 + 1)-spectrum it could be only slightly mixed with the active neutrinos and mainly provide a description of the lsnd result .
it has been argued in the literature @xcite that the ( 3 + 1)-spectra are strongly disfavoured by the data , whereas the ( 2 + 2)-spectra are the preferred ones , in agreement with all data showing evidence for neutrino oscillations and also with those where no such evidence has been found .
recently , in refs .
@xcite this statement has been challenged because in the latest lsnd analysis the allowed region in the @xmath0@xmath1 plane , where @xmath1 is the lsnd transition amplitude , has undergone a slight shift towards smaller mass - squared differences , which makes the ( 3 + 1)-spectra somewhat less disfavoured . furthermore , in a 2-neutrino analysis of atmospheric neutrino oscillations the super - kamiokande data prefer @xmath15 conversion over @xmath16 @xcite .
moreover , there is some debate also in the solar neutrino problem whether the @xmath17 transition is disfavoured in comparison with other solutions @xcite , though such a feature seems not to be borne out by a global analysis of the data @xcite . in any case , moving away from pure 2-neutrino considerations in the solar and atmospheric neutrino problems , transitions into active sterile superpositions @xcite give viable solutions to both problems within the ( 2 + 2)-spectral schemes , with features which will be tested in the future @xcite .
the arguments presented in refs .
@xcite , which disfavour the ( 3 + 1)-mass spectra , are based on exclusion curves from short - baseline ( sbl ) experiments , and solar and atmospheric neutrino data enter into this simplified analysis only through inequalities .
the advantage of this approach is that its parameters are confined to the quantities are used instead . ] @xmath18 and the sbl or lsnd mass - squared difference @xmath0 . for definiteness
we assume that the mass separated by the lsnd gap is @xmath19 and , therefore , @xmath20 .
it has turned out that the up - down asymmetry of atmospheric multi - gev @xmath21-like events measured in the super - kamiokande experiment @xcite is very suitable to constrain @xmath22 @xcite , whereas from the solar data it follows that @xmath23 must be small .
the probability of sbl @xmath24 transitions is given by the two - neutrino - like formula @xcite @xmath25 where @xmath26 is the distance between source and detector and @xmath27 is the neutrino energy and @xmath28 the lsnd experiment gives an allowed region in the @xmath0@xmath1 plane . however , the arguments of refs .
@xcite are not based on a well - defined statistical procedure .
therefore , they remain on a semi - quantitative level and do not allow to assess a confidence level ( cl ) which quantifies the degree at which the ( 3 + 1)-spectra are excluded . in this paper
we make a step forward towards such an assessment .
the main points to achieve our goal are the following : * the aim is to arrive , for every sbl mass - squared difference @xmath0 , at a probability distribution solely in terms of @xmath23 and @xmath22 ; a suitable method for this purpose is given by the likelihood function in combination with the bayesian approach ( see , e.g. , refs .
@xcite ) . *
we make full use of the data of the sbl bugey @xcite , cdhs @xcite and karmen @xcite experiments through the likelihood function . * in the spirit of refs .
@xcite , all information pertaining to the atmospheric and solar mass - squared differences is included via inequalities . within our probabilistic framework
we are able to treat inequalities as prior probabilities or with a kind of a maximum likelihood method . * in this way we treat the inequality following from the atmospheric up - down asymmetry and , similarly , we include also the result of the chooz experiment @xcite ; the solar neutrino data allow a simpler treatment in the context of the bugey data @xcite , as described in refs .
@xcite . * eventually , for every given sbl @xmath0 and any cl @xmath29 , we are able to calculate an upper bound @xmath30 on the transition amplitude @xmath1 , and we can compare such bounds @xmath30 with the 90% and 99% cl regions in the @xmath0@xmath1 plane found by the lsnd experiment @xcite .
in the same framework we will also discuss the ( 2 + 2)-spectra .
the plan of the paper is as follows . in section [ uda ]
we rederive the atmospheric up - down inequality @xcite in a form which is suitable for our purpose . in section [ b+l ]
we introduce the likelihood function and the bayesian approach , and describe how to incorporate inequalities ; we apply the methods developed there to the chooz result and the atmospheric up - down inequality . in section [ plots ]
we explain how we calculate bounds on @xmath1 and discuss each of the sbl experiments we use , together with their features which are important in this context .
details of technical nature are deferred to the appendix .
our main result , represented as a plot in the @xmath0@xmath1 plane , is also given in this section .
in section [ 2 + 2 ] we consider the ( 2 + 2)-spectra in the framework of our statistical approach and in section [ concl ] we draw our conclusions .
the most convincing evidence for @xmath3 disappearance in atmospheric neutrino experiments is given by the so - called up - down asymmetry @xmath31 measured by the super - kamiokande collaboration @xcite , where @xmath14 and @xmath32 refer to the number of up - going and down - going @xmath21-like multi - gev events , respectively . quoting the number for fully contained events , after 1289 days of operation
the result @xmath33 was found @xcite .
adding statistical and systematic error in quadrature , one obtains @xmath34 because of the smallness of the systematic error this value is identical with the statistical error .
let us now rederive the atmospheric up - down inequality @xcite . in the following we will not indicate antineutrinos but our arguments will hold for both , neutrinos and antineutrinos . with the assumption that downward - going atmospheric neutrinos do not oscillate with the frequency associated with @xmath35 and that oscillations according to @xmath0 are averaged out , we obtain @xmath36 denoting the number of muon ( electron ) neutrinos and antineutrinos produced in the atmosphere by @xmath37 ( @xmath38 ) , it follows from eq .
( [ pd ] ) that @xmath39 + \frac{1}{2 } n_e a_{\mu;e } \,.\ ] ] for the upward - going neutrinos we have the inequalities @xcite @xmath40 we , therefore , have the inequality @xmath41 note that for the ( 3 + 1)-spectra the amplitude @xmath1 is given by eq .
( [ a ] ) . since @xmath42 ( [ aud ] ) is a monotonously decreasing function in @xmath14 , using eqs .
( [ d ] ) and ( [ u ] ) we obtain the so - called up - down inequality for @xmath21-like atmospheric events @xmath43 in this equation we have defined @xmath44 .
the numerical value @xmath45 can be read off from fig . 3 in ref .
@xcite of the super - kamiokande collaboration . with similar arguments one can also find an upper bound @xmath46 hence
, the up - down asymmetry is confined to the interval @xmath47 some remarks are at order . our inequality ( [ g ] ) is a little different from the analogous inequality in ref .
@xcite , because in the present case we have not eliminated @xmath23 ; this is useful because our aim is to derive a probability distribution in @xmath23 and @xmath22 ( see next section ) .
the ratio @xmath48 has a slight dependence on the atmospheric zenith angle which is neglected here ; however , since @xmath23 is confined to rather small values @xcite by the result of the bugey experiment @xcite , the terms containing @xmath48 are rather unimportant numerically . the assumptions for deriving eqs .
( [ g ] ) and ( [ h ] ) are not exactly fulfilled : for down - going neutrinos with zenith angles around @xmath49 , oscillations according to small sbl mass - squared differences @xmath0 around 0.2 ev@xmath8 are not completely averaged out ; for down - going neutrinos with zenith angles close to @xmath50 , oscillations according to @xmath35 do occur already , if the atmospheric mass - squared difference is large .
we have checked that both effects do not change numerically the bound @xmath51 by more than a few percent in the worst case .
finally , matter effects , which are important for up - going neutrinos , do not affect the inequalities ( [ pu ] ) and thus also not the inequality for @xmath14 .
( [ g ] ) is , therefore , a well - established inequality which restricts mainly the allowed range of @xmath22 . due to
( [ a ] ) , it will be used in the following to constrain the sbl amplitude @xmath1 .
as mentioned in the introduction , a suitable method for the purpose of deriving a probability distribution in the variables @xmath23 and @xmath22 is given by the likelihood function combined with the bayesian approach , which is defined as follows .
suppose one has a series of measurements @xmath52 and @xmath48 parameters @xmath53 to be estimated .
then the bayesian approach allows to construct a `` posterior '' probability density in the parameter space via ( see , e.g. , ref .
@xcite ) @xmath54 in this expression , @xmath55 is the likelihood function and @xmath56 is the prior probability density associated with the parameters @xmath57 , reflecting the state of knowledge of @xmath57 before the measurement . let us suppose now that we consider an observable @xmath58 whose experimental value is @xmath59 and its _ true _ value is @xmath60 .
we assume that the values of measurements of @xmath58 are distributed according to a gaussian distribution @xmath61\ ] ] around the true value .
suppose further that from some theoretical consideration we have the knowledge that the true value @xmath60 is bounded by @xmath62 we conceive the true value @xmath60 , which is otherwise unknown , as a parameter in our scenario and assign to it a prior probability density @xmath63 where @xmath64 denotes the heaviside function .
due to lack of further knowledge , we have assumed a flat prior probability density .
since we are not interested in a posterior probability density in the parameter @xmath60 we perform the integral @xmath65 with @xmath66 where we have made use of the error function defined by @xmath67 the function ( [ lz ] ) represents then the relevant factor in the posterior probability density which takes into account the inequality ( [ priorknow ] ) . to check if we have been lead to a meaningful expression ( [ lz ] ) , we want to discuss the behaviour of this function . following from @xmath68 , we find that for @xmath69 we have @xmath70 to a very good approximation , and for @xmath71 or @xmath72 we have @xmath73 . in this discussion , `` much smaller '' or `` much larger '' is defined in units of @xmath74 .
if we assume that @xmath74 becomes negligibly small , the function @xmath75 approaches a step function with @xmath76 being the height of the step .
the edges of the step are smoothened out by a finite @xmath74 .
thus , we will have a maximal contribution to the posterior probability density for @xmath77 well inside the interval @xmath78 $ ] , whereas for @xmath77 well outside the interval @xmath78 $ ] the function @xmath75 is very close to zero .
thus we can be confident that the inequality ( [ priorknow ] ) is reasonably well taken into account by our procedure .
again we depart from eqs .
( [ lz ] ) and ( [ priorknow ] ) .
now we take into account our lack of knowledge about the true value of @xmath58 by maximizing @xmath79 as a function of @xmath80 $ ] .
it is easy to check that one obtains @xmath81 } l_z \equiv l^m_z = } \nonumber \\ & & \frac{1}{\sqrt{2\pi } \sigma_z } \exp \left\ { -\frac{1}{2 } \left [ \left ( \frac{z_\mathrm{exp}-b}{\sigma_z } \right)^2 \theta ( z_\mathrm{exp}-b ) + \left ( \frac{a - z_\mathrm{exp}}{\sigma_z } \right)^2 \theta ( a - z_\mathrm{exp } ) \right ] \right\ } \ , .
\label{lzm}\end{aligned}\ ] ] in appendix [ relation ] we will discuss how this method is related to the method in the previous section . the chooz experiment is a long - baseline @xmath82 disappearance experiment .
it measures the survival probability @xmath83 , for which we derive the inequality @xmath84 we have used the abbreviation @xmath85 .
we conceive @xmath83 in eq .
( [ choozineq ] ) as the _ true _ survival probability , as opposed to the experimental value @xmath86 with the error @xmath87 @xcite .
thus we have the range @xmath88 making the substitutions @xmath89 the chooz result can be included in the statistical analysis according to both methods described in the previous subsections . for the atmospheric up - down inequality we substitute @xmath90 where @xmath51 , @xmath91 , @xmath92 and @xmath93 are given by eqs .
( [ g ] ) , ( [ h ] ) , ( [ aexp ] ) and ( [ da ] ) , respectively .
notice that , since we have @xmath94 and @xmath95 , the first term in the exponential of the expression ( [ lzm ] ) does not contribute here , and the first term in the expression ( [ lz ] ) is 1 for all practical purposes . the substitutions ( [ chsubst ] ) and ( [ udsubst ] ) allow us to define according to eqs .
( [ lz ] ) , ( [ piz ] ) , ( [ lz ] ) , ( [ lzm ] ) the functions @xmath96 , @xmath97 , @xmath98 , @xmath99 and @xmath100 , @xmath101 , @xmath102 , @xmath103 , referring to the chooz experiment and the atmospheric up - down inequality , respectively .
these functions will be used in the following discussions .
let us now describe how to derive our desired probability distribution .
we concentrate on the four parameters @xmath23 , @xmath22 , @xmath104 and @xmath83 .
we want to stress again that the latter two quantities are conceived as _ true _ values .
therefore , they are parameters within our procedure and we will treat them according to the methods described in section [ b+l ] , in order to arrive at a distribution solely in @xmath23 and @xmath22 .
the likelihood function is given by @xmath105 where the first factor @xmath106 is the product of the likelihood functions of the bugey , cdhs and karmen experiments .
this likelihood function will be discussed in the next section and in appendix [ app ] . for the treatment of parameters other than @xmath23 and @xmath22 , which appear in the fitting procedure to the sbl experiments , see also appendix [ app ] .
the physically allowed region @xmath107 of @xmath23 and @xmath22 is described by the inequalities @xmath108 in order to incorporate it into our procedure , we define a function @xmath109 such that @xmath110 for @xmath111 and 0 otherwise .
this function has thus the task of a prior which confines @xmath112 to the physically meaningful region . adopting the maximum likelihood method of section [ mla ] in order to deal with the parameters @xmath104 and @xmath83 and combining this method with eq .
( [ bayes ] ) , we finally arrive at the desired probability distribution @xmath113 the dependence of @xmath114 and @xmath115 on @xmath23 and @xmath22 comes in through the substitutions ( [ udsubst ] ) and ( [ chsubst ] ) of the boundaries @xmath116 and @xmath117 .
discussing now the bayesian approach described in section [ ba ] , we have the prior probability density @xmath118 we stress that @xmath119 is fulfilled .
this equation tells us that , after integrating over the other variables , the prior probability density for @xmath23 and @xmath22 is uniform . since these are the variables whose distribution we want to calculate , eq . ( [ priorcond ] ) assures us that with our choice of priors we have not introduced a bias in the distribution of @xmath23 and @xmath22 .
according to the bayesian approach in section [ ba ] we perform the integrations @xmath120 finally , we obtain via this method the probability distribution @xmath121 it is important to note that for every @xmath0 we have such a distribution . the sbl mass - squared difference , which is hidden in @xmath106 , is not on the same footing as @xmath23 and @xmath22 in our approach . for every given @xmath0 , we find restrictions on @xmath23 and @xmath22 from experiment is treated in the same way . ] which allow us finally to obtain a bound on the transition amplitude @xmath1 for any given cl . choosing a cl @xmath29 , we find the corresponding bound on @xmath1 by the prescription @xmath122 for instance , if we choose @xmath123 , with eq .
( [ abound ] ) we can calculate a number @xmath124 such that @xmath125 at 99% cl .
the two distributions @xmath126 ( [ pddm ] ) and @xmath127 ( [ pddb ] ) are , of course , different functions of the variables .
the method to obtain bounds on @xmath1 as given by eq .
( [ abound ] ) is the same for both distributions . in the following
, we will comment on how much the bounds on @xmath1 calculated with both distributions differ numerically . as mentioned previously ,
the sbl experiments are treated with the likelihood function @xmath128 which enters eq .
( [ l ] ) . here
we want to discuss some features of @xmath106 and how these features influence the bound on @xmath1 .
let us start with the bugey @xcite and cdhs @xcite experiments .
both are disappearance experiments and , therefore , the relevant sbl survival probabilities are given by @xmath129 where @xmath130 refers to the bugey and @xmath131 to the cdhs experiment .
both experiments have not found evidence for neutrino oscillations . in the case of the bugey experiment
the survival amplitude @xmath132 is bounded by 0.1 or smaller values in the relevant range @xmath133 at 90% cl @xcite .
thus , @xmath23 is either very small or close to 1 . since for solar neutrinos
the inequality @xmath134 holds @xcite , @xmath23 must be small must be small is independent of the actual solution to the solar neutrino problem . ] and in the fit of the bugey data we make the approximation @xmath135 in the survival probability . in principle , the inequality ( [ sun - ineq ] ) should be included in our analysis with the help of one of the methods discussed in sections [ ba ] and [ mla ] , but in view of the smallness of the bugey survival amplitude this is not necessary . in the case of the cdhs experiment , for @xmath136 ev@xmath8 , there is an analogous feature concerning @xmath22 @xcite ; there , the selection of the small @xmath22 is guaranteed by the up - down inequality ( [ g ] ) @xcite . for smaller @xmath0 the cdhs restriction disappears and values of @xmath22 as large as 0.5
are allowed by the up - down inequality .
therefore , in @xmath137 we do not neglect @xmath138 in the survival amplitude because this would not be justified . in the karmen experiment , @xmath139 transitions have not been observed and the karmen exclusion curve @xcite cuts right through the region allowed by lsnd .
therefore , it is important to take into account the karmen result in our analysis . for the details of our fit to the karmen data and the analogous details concerning the bugey and cdhs experiments see appendix [ app ] .
we include the chooz result in our analysis for @xmath140 ev@xmath8 , where the bound on @xmath141 in the chooz plots is a straight line @xcite ; this guarantees that oscillations with the sbl @xmath0 are averaged out and hence eq .
( [ choozineq ] ) is valid .
the chooz result has an effect on the @xmath1 exclusion curve mainly for @xmath0 where the bugey bound on @xmath23 is not so strong , which is the case for @xmath0 around 0.1 ev@xmath8 and @xmath142 ev@xmath8 .
[ plot1 ] represents the main result of this paper . in this figure
we show the regions in the @xmath0@xmath1 plane allowed by lsnd at 90% and 99% cl @xcite and our exclusion curves for @xmath1 with ( 3 + 1)-mass spectra at 90% , 95% and 99% cl .
these exclusion curves have been calculated using the maximum likelihood approach of section [ mla ] for the treatment of the _ true _
values @xmath104 and @xmath83 .
this method gives for most values of @xmath0
slightly weaker bounds than the bayesian approach of section [ ba ] ; however , the numerical difference between the two methods is always smaller than 4% .
we read off from fig .
[ plot1 ] that the 99% cl lsnd region has no overlap with the allowed region to the left of our 95% cl bound .
this shows that the likelihood of ( 3 + 1)-mass spectra is not very high . comparing the 99% cl exclusion curve with the 99% cl lsnd region and confining ourselves to @xmath143 ev@xmath8
, we find overlaps at @xmath144 and approximately between 0.25 and 0.4 ev@xmath8 .
this agrees with the findings in ref .
@xcite , where 90% exclusion curves and the bound on @xmath1 derived in ref .
@xcite are compared with the 99% cl lsnd region .
the result of the inclusion of the chooz result for @xmath140 ev@xmath8 can be seen as a jump in our exclusion curves . from this jump
the effect of the chooz result can be read off for small mass - squared differences .
let us stress once more that in our analysis an exclusion curve for a given cl is the result of a well - defined statistical procedure and is obtained by including all available data other than the lsnd data ; such exclusion curves have a precise statistical meaning ; an overlap of the 99% cl lsnd region with the region to the left of an exclusion curve occurs only with the exclusion curve at 99% cl , but not at 95% cl .
there are claims in the literature that all three indications in favour of neutrino oscillations are compatible with _
neutrinos @xcite .
such claims have been refuted by a combined fit to the data @xcite . in our analysis
we arrive at the same conclusion , because our treatment of the ( 3 + 1)-spectra is also applicable in the case of three neutrinos , where one has ( 2 + 1)-mass spectra , since there is a sbl mass - squared difference @xmath0 and another mass - squared difference much smaller than @xmath0 ; assuming that @xmath145 is the mass separated from the other two by the lsnd mass gap , we have to make the identification @xmath146 ( @xmath147 ) .
the values of @xmath0 and @xmath1 , indicated in the papers of ref .
@xcite as preferred solutions , lie in the region allowed by lsnd at 99% and thus beyond our 95% cl exclusion curve in fig .
[ plot1 ] .
if we want to treat the ( 2 + 2)-neutrino mass spectra with the method discussed in the previous two sections , the situation is more involved . in this case the @xmath24 transition amplitude is given by @xcite @xmath148 and can not be expressed by @xmath149 as in the case of the ( 3 + 1)-spectra ( see eq .
( [ a ] ) ) .
this suggests to perform an analysis with the five parameters @xmath23 , @xmath22 , @xmath1 , @xmath104 and @xmath83 in the case of the ( 2 + 2)-spectra .
let us explore in more detail the relationship between the amplitude @xmath1 and the parameters @xmath23 and @xmath22 . from eq .
( [ a2 ] ) , using the cauchy
schwarz inequality , we readily obtain the inequality @xcite @xmath150 \,.\ ] ] this inequality implies that for every @xmath1 the following region is allowed in the @xmath23@xmath22 plane : @xmath151 for every @xmath1 between 0 and 1 , this is a region in the unit square , confined by two hyperbolas .
let us label the neutrino masses such that @xmath152 and @xmath153 .
then , with the definitions ( [ d2 ] ) , eqs .
( [ d ] ) and ( [ u ] ) hold also for the ( 2 + 2)-mass spectra and we obtain the inequalities @xmath154 and @xmath155 for the atmospheric up - down asymmetry .
furthermore , the survival probability for solar neutrinos is bounded by @xmath156 .
therefore , @xmath23 is close to one is small , then it follows that @xmath157 , which is in disagreement with the result of the homestake experiment ( see first paper in ref .
@xcite ) ; moreover , for small @xmath23 we obtain the same bound on @xmath1 as for the disfavoured ( 3 + 1)-mass spectra . ] and we make the approximation @xmath158 in the bugey survival amplitude .
note that the chooz inequality ( [ choozineq ] ) also holds for the ( 2 + 2)-spectra . in analogy to eq .
( [ abound ] ) in the ( 3 + 1)-case , our aim is to obtain a probability distribution in @xmath1 .
here we confine ourselves to the maximum likelihood method of section [ mla ] for the treatment of inequalities .
in addition to the up - down and chooz inequalities we have to take into account eq .
( [ f ] ) . according to the method of section [ mla ] , we maximize with respect to @xmath104 , @xmath83 , @xmath23 and @xmath22 .
the maximization with respect to the first two variables leads to the functions @xmath159 and @xmath99 , which are defined in section [ ch+ud ] .
the prime on the first function indicates that now it is obtained from @xmath103 by the replacement @xmath160 and @xmath161 .
generalizing the method of section [ mla ] , we define @xmath162 \,,\ ] ] where @xmath163 is given by eq .
( [ lsbl ] ) , but in view of the appearance of @xmath1 in @xmath164 it is interpreted as a function of three variables . finally , we obtain the probability distribution for @xmath1 given by @xmath165 in fig .
[ plot2 ] the 90% cl bound on @xmath1 calculated with the distribution ( [ pm ] ) is plotted . in this figure
the 90% and 99% cl level regions of lsnd are also depicted .
the region to the left of the bound has an overlap area with the lsnd region of 90% cl .
this shows that the ( 2 + 2)-spectra are well compatible with all data .
also shown in fig .
[ plot2 ] are the karmen and bugey exclusion curves .
as expected , for large @xmath0 the solid line of our exclusion curve follows rather well the karmen exclusion curve , whereas for small @xmath0 it tends to follow the bugey curve . due to the inclusion of the chooz result above 0.05 ev@xmath8 ,
the solid line is more restrictive there than the bugey curve , whereas below 0.05 ev@xmath8 both curves are nearly identical .
we have also investigated the exclusion curve of the ( 2 + 2)-spectra within the bayesian approach to inequalities of section [ ba ] . in this approach
we have more freedom to make `` reasonable '' choices for priors than in the ( 3 + 1)-case , and , numerically , the @xmath1 bounds tend to be a little more restrictive than the solid line in fig . [ plot2 ] .
in this paper we have examined the ( 3 + 1 ) and ( 2 + 2)-classes of the 4-neutrino mass spectra within a rigorous statistical analysis . since we do not have sufficient information concerning the final lsnd data @xcite , we have chosen the approach to analyse all other available data and compare our result with the lsnd result in a @xmath0@xmath1 plot .
this approach suggests to extract that information from the solar and atmospheric data which is most relevant for sbl oscillations with respect to @xmath0 .
this extraction is most appropriately done in the form of inequalities involving elements of the neutrino mixing matrix @xcite .
similarly , we have extracted the `` sbl information '' contained in the chooz result . on the other hand , concerning the sbl experiments , we have fully included the data of the bugey , cdhs and karmen experiments . since
our aim has been to incorporate in the statistical analysis the atmospheric up - down inequality , eq .
( [ g ] ) or ( [ g ] ) , and the chooz inequality ( [ choozineq ] ) , we have made use of the likelihood function in combination with the bayesian approach to probability which allows us to derive probability distributions of the parameters which are to be estimated . in this context we have presented two possibilities for including inequalities involving parameters : one way is to treat them in the form of prior probability densities for which a `` reasonable choice '' has to be made ; another way is a kind of maximum likelihood treatment .
numerically , we have compared both methods for exclusion curves in the case of the ( 3 + 1)-spectra and found that the difference is negligible . with the method described in the paper we have obtained for every @xmath0 a probability distribution for the sbl transition amplitude @xmath1 , from which we could derive bounds as a function of @xmath0 on this amplitude for any cl . the results are shown in figs .
[ plot1 ] and [ plot2 ] for the ( 3 + 1 ) and ( 2 + 2)-neutrino mass spectra , respectively . in the latter case
our 90% cl exclusion curve is close to the karmen exclusion curve down to @xmath166 ev@xmath8 ; there it turns off and starts to come close to the exclusion curve given by the bugey data .
thus , for the ( 2 + 2)-spectra our method reproduces more or less what one obtains by naively comparing the karmen and bugey exclusion curves with the region allowed by lsnd .
therefore , these spectra are well compatible with all the data . on the other hand ,
in the case of the ( 3 + 1)-spectra , our 95% cl bound has no overlap with the lsnd region of 99% cl .
we , therefore , come to the conclusion that this spectral class is rather unlikely , even with the recent change in the lsnd region .
thus we strengthen with the method presented here the claims made in refs .
should the lsnd result be confirmed by the miniboone collaboration @xcite , then , in a 4-neutrino scheme , the sterile neutrino should make its appearance either in the solar or atmospheric neutrinos , or both .
we thank c. giunti and s.m .
bilenky for many useful discussions .
we are very grateful to f. dydak and j. wotschack for valuable information about the cdhs experiment and to c. walter for providing us with the latest results of the super - kamiokande up - down asymmetry .
is supported by the austrian academy of sciences .
here we want to elucidate the relationship between the methods for the inclusion of inequalities introduced in sections [ ba ] and [ mla ] . in the bayesian approach to this problem we consider in general integrals of the type @xmath167 = \int_a^b dz l_z(z ) \pi(z ) \,,\ ] ] where @xmath79 is given by eq .
( [ lz ] ) .
a prior @xmath168 is a positive and piecewise continuous function such that @xmath169 . in section [ ba ]
we have used the flat prior ( [ piz ] ) .
if we denote the set of all prior probability densities on the interval @xmath78 $ ] by @xmath170 , then the following proposition holds .
+ * proposition : * for the integral in eq .
( [ int ] ) one has the upper limit @xmath171 = l_z^m \,,\ ] ] where @xmath172 is given by eq .
( [ lzm ] ) . + * proof : * the proof is very straightforward . let us first assume that @xmath173 .
then we have @xmath174 < l_z(b ) \int_a^b dz\ , \pi(z ) = l_z(b ) \,.\ ] ] thus we have found an upper bound on @xmath175 .
if we can find a sequence of @xmath176 of prior probability densities such that @xmath177 = l_z(b ) \,,\ ] ] then the proposition is proven for the case @xmath173 .
such a sequence is easy to give : any sequence @xmath176 , for which @xmath178 for @xmath179 holds , fulfills our purpose .
for @xmath180 we proceed analogously .
if @xmath181 $ ] , the upper bound on @xmath175 is @xmath182 .
the sequence @xmath176 is constructed accordingly with the idea given before .
thus for the three cases @xmath173 , @xmath180 and @xmath181 $ ] we have the upper limits @xmath183 , @xmath184 and @xmath182 , respectively . in summary , we just have obtained the expression ( [ lzm ] ) . q.e.d .
thus the expression @xmath172 is obtained from the bayesian approach as the maximum over all possible priors .
sloppily speaking , the prior in @xmath175 , which gives @xmath172 , is a delta function ( @xmath185 , @xmath186 or @xmath187 ) ; however , in order to choose the correct delta function one has to know the experimental value @xmath77 and , therefore , such a function does not deserve the name `` prior '' anymore . though @xmath188 holds , it is not obvious which method gives the stronger constraint in an actual situation , because one has additional factors in the probability distribution one aims at , coming from additional data , and one has to normalize the combined distribution ( see eq .
( [ bayes ] ) ) . moreover , the bounds @xmath116 and @xmath117 are in general functions of the parameters whose distribution we want to know . in the concrete situation
discussed in section [ plots ] the maximum likelihood method represented by @xmath172 gives a slightly weaker restriction on the transition amplitude @xmath1 for most @xmath0 .
in this appendix we describe how the data of the sbl experiments bugey , cdhs and karmen is included in our analysis .
we use as much information as can be recovered from the publications of the experimental groups to perform a fit to the data .
the fact that we can reproduce the published 90% cl bounds in the case of 2-neutrino oscillations to a good accuracy inspires confidence in our analysis .
the bugey experiment @xcite searches for @xmath82 disappearance at the three distances 15 m , 40 m and 95 m away from a nuclear reactor .
the electron antineutrinos are detected through the reaction @xmath189 . as input data for our analysis we use fig .
17 of ref .
@xcite , where the ratios of the observed events to the number of expected events in case of no oscillations are shown for the three positions in bins of positron kinetic energy in the range 1 mev @xmath190 6 mev . for the analysis we follow eq .
( 9 ) of ref .
@xcite and use the @xmath191-function @xmath192 ^ 2 } { \sigma_{ji}^2 } + \frac{(a_j - 1)^2}{\sigma_a^2 } \right\ } + \frac{(a-1)^2}{\sigma_a^2 } + \frac{b^2}{\sigma_b^2 } \ , .
\label{chi2bugey}\ ] ] here @xmath193 labels the three positions , @xmath194 the positron energy bins and @xmath195 are the numbers of bins at each position . for @xmath196
we take the mean positron energy in bin @xmath197 .
@xmath198 is the ratio of measured to expected events in each bin with its statistical error @xmath199 , both read off from fig .
17 of ref .
@xmath200 is the theoretical prediction for this ratio in the case of oscillations , depending on the oscillation parameters , and we set @xmath201 , where @xmath202 is given in eq .
( [ disapp ] ) .
various systematic uncertainties are taken into account by minimizing the @xmath191 , eq .
( [ chi2bugey ] ) , with respect to the five parameters @xmath203 , @xmath204 and @xmath117 for given oscillation parameters . in ref .
@xcite the values @xmath205 mev@xmath206 and @xmath207 mev are given . to perform the averaging of the survival probability
we estimate the uncertainty in the flight length of the neutrinos because of the size of the production region and the detector to 3 m and we assume that the flux varies with the distance as @xmath208 . in the relevant energy range antineutrino and positron energy
are related by @xmath209 mev to a very good approximation . for the purpose of averaging the survival probability over the energy range in one bin
it is a good approximation to take neutrino flux , detection cross section and efficiency as constant with energy .
the reason for this is that the energy bins are relatively small and only ratios of observed to expected events in each bin are considered .
furthermore , we assume a gaussian resolution function for the positron energy measurement with variance 0.4 mev .
the likelihood function in eq .
( [ lsbl ] ) which contains the information of the bugey experiment is obtained from the @xmath191 of eq .
( [ chi2bugey ] ) by @xcite @xmath210 the 90% cl bound on @xmath23 obtained from this likelihood function alone can be compared to the bugey exclusion curve in the 2-neutrino case with the identification @xmath211 .
the curve obtained by our analysis is shown in fig .
[ plot2 ] and compares well with the originally published one @xcite . in eq .
( [ disapp ] ) contributions of oscillations because of @xmath212 to the sbl disappearance amplitude are neglected .
this approximation may not be exactly fulfilled in the bugey experiment for small @xmath213 and large @xmath212 .
we have calculated the 90% cl bound from bugey by taking into account also oscillations with @xmath214 ev@xmath8 , which is the 99% cl upper bound on @xmath212 @xcite , and find that for @xmath215 ev@xmath8 the effect is smaller than 6% .
the cdhs experiment @xcite searches for @xmath216 disappearance by comparing the number of events in the so - called back and front detectors at the distances @xmath217 m and @xmath218 m , respectively , from the neutrino source .
the neutrinos are detected via muons produced in charged - current interactions .
the data is given in form of the double ratios @xmath219 where @xmath220 is the number of observed events in the back ( front ) detector and @xmath221 and @xmath222 are the corresponding quantities expected for no oscillations calculated by monte carlo .
the ratios ( [ rcorr ] ) are given in 15 bins of `` projected range in iron '' .
this is the distance traveled by the muon in the detector ( consisting of iron ) projected onto the detector axis , which has an angle of @xmath223 relative to the neutrino beam axis .
we calculate the range in iron @xmath224 of a muon with energy @xmath225 by integrating eq .
( 23.1 ) of ref .
the muon energy intervals @xmath226 $ ] corresponding to the intervals of projected range ( @xmath227 ) for bin @xmath194 given in table 1 of ref .
@xcite are obtained by applying the relation @xmath228 , where @xmath229 is the average scattering angle of the muons @xcite .
we estimate the number of events in bin @xmath194 and at position @xmath230back or front using @xmath231 here @xmath232 is the detector mass at position @xmath233 . taking into account the size of the decay tunnel ( 52 m ) and the length relevant for detection in the back / front detector ( 72 m/22 m ) @xcite we have @xmath234 m and @xmath235 m. the disappearance probability @xmath236 depending on the oscillation parameters
is given in eq .
( [ disapp ] ) and @xmath237 is the cross section for the deep inelastic scattering process @xmath238 ( see , e.g. , ref .
the neutrino flux @xmath239 in the relevant neutrino energy range @xmath240 gev is proportional to @xmath241 @xcite .
finally , we obtain for the theoretical prediction for the double ratios ( [ rcorr ] ) in bin @xmath194 @xmath242 and we define the cdhs likelihood function by @xmath243\,.\ ] ] assuming total correlation between any two bins , the covariance matrix is given by @xmath244 . @xmath245 and
its statistical error @xmath246 are read off from table 1 of ref .
@xcite . the overall systematic error in the ratio of event rates in the two detectors
was estimated to @xmath247% in ref .
@xcite .
the 90% cl bound on @xmath248 obtained by our analysis is shown in fig .
[ plot3 ] .
it is very similar to the bound published by the cdhs collaboration @xcite .
there are minor differences for small mass squared differences , which could have some effect for our bound on @xmath1 in the region 0.2 ev@xmath249 ev@xmath8 : our bound disappears at @xmath250 ev@xmath8 , whereas the cdhs bound extends down to approximately 0.24 ev@xmath8 .
the latest data of the @xmath251 oscillation search in the karmen experiment is presented in ref .
@xcite . analysing the data taken from february 1997 to march 2000 they find a total of 11 candidate events , in good agreement with the expected number of background events for no oscillations of @xmath252 . for our analysis
we use the data resulting from the detection process @xmath189 .
the positron spectrum @xmath253 expected for @xmath254 and @xmath255 ev@xmath8 we take from fig .
2(a ) of ref .
@xcite . to estimate the number of events in a positron energy interval @xmath256 $ ] resulting from neutrino oscillations we follow eq .
( b1 ) of ref .
@xcite : @xmath257 where the oscillation probability @xmath258 is given in eq .
( [ pmue ] ) .
for @xmath259 we take @xmath260 m @xcite and antineutrino energy and positron kinetic energy are related by @xmath209 mev . the normalization factor @xmath261
is fixed by requiring that for the total positron energy range @xmath262 mev , @xmath263 mev , full mixing ( @xmath264 ) and @xmath265 ev@xmath8 the number of events resulting from oscillations is 2442 ( see table 1 of ref .
@xcite ) . from fig .
2(b ) of ref .
@xcite we read off for each of the 9 positron energy bins the number of observed events @xmath266 and the number of background events expected for no oscillations @xmath267
. then we construct the likelihood function by using the poisson distribution : @xmath268 here @xmath269 is calculated from eq .
( [ nosc ] ) by choosing @xmath270 and @xmath271 according to the bin @xmath194 .
the 90% cl bound on @xmath1 for a given @xmath0 obtained from the probability distribution implied by eq .
( [ lkarmen ] ) in the bayesian approach is shown in fig .
[ plot2 ] .
our bound is very close to the one presented in ref .
@xcite .
et al . _ ,
j. * 496 * , 505 ( 1998 ) ; k.s .
et al . _ ,
kamiokande coll .
, phys . rev .
77 * , 1683 ( 1996 ) ; w. hampel _
_ , gallex coll .
b * 447 * , 127 ( 1999 ) ; d.n .
_ , sage coll .
* 83 * , 4686 ( 1999 ) ; y. fukuda _ et al .
_ , super - kamiokande coll .
lett . * 81 * , 1158 ( 1998 ) y. fukuda _
et al . _ ,
kamiokande coll .
b * 335 * , 237 ( 1994 ) ; r. becker - szendy _ et al .
_ , imb coll .
b ( proc . suppl . ) * 38 * , 331 ( 1995 ) ; w.w.m .
et al . _ ,
soudan coll .
b * 449 * , 137 ( 1999 ) ; m. ambrosio _
et al . _ ,
macro coll .
b * 434 * , 451 ( 1998 ) c. athanassopoulos _ et al .
_ , lsnd coll .
lett . * 77 * , 3082 ( 1996 ) ; _ ibid _ * 81 * , 1774 ( 1998 ) j.t .
peltoniemi , d. tommasini and j.w.f .
valle , phys .
b * 298 * , 383 ( 1993 ) ; j.t .
peltoniemi and j.w.f .
valle , nucl .
b * 406 * , 409 ( 1993 ) ; d.o .
caldwell and r.n .
mohapatra , phys .
d * 48 * , 3259 ( 1993 ) ; e. ma and p. roy , _ ibid _ * 52 * , r4780 ( 1995 ) ; e.j .
et al . _ ,
b * 357 * , 608 ( 1995 ) ; j.j .
gomez - cadenas and m.c .
gonzalez - garcia , z. phys .
c * 71 * , 443 ( 1996 ) ; e. ma , mod .
a * 11 * , 1893 ( 1996 ) ; s. goswami , phys .
d * 55 * , 2931 ( 1997 ) s.m .
bilkenky , c. giunti and w. grimus , in _ proceedings of neutrino 96 _ , helsinki , finland , 1319 june 1996 , p. 174
, edited by k. enqvist , k. huitu and j. maalampi ( world scientific , singapore 1997 ) ; s.m .
bilenky , c. giunti and w. grimus , eur .
j. c * 1 * , 247 ( 1998 ) d. dooling , c. giunti , , k. kang and c.w .
kim , phys .
d * 61 * , 073011 ( 2000 ) ; c. giunti , m.c .
gonzalez - garcia and c. pea - garay , phys .
d * 62 * , 013005 ( 2000 ) ; m.c .
gonzalez - garcia and c. pea - garay , hep - ph/0011245 g. cowan , _ statistical data analysis _ ( oxford university press , new york 1998 )
; b.p . roe , _ probability and statistics in experimental physics _
( springer , new york 1992 ) ; g. cowan , in _ present and future cp measurements _ , p. 98 , hep - ph/0102159 k. eitel and b. zeitnitz , karmen coll .
, in _ proceedings of neutrino 98 _ , takayama , japan , 4 - 9 june 1998 , nucl .
. suppl . *
77 * , 212 ( 1999 ) k. eitel , karmen coll . , talk given at _
neutrino 2000 _ , 1521 june 2000 , sudbury , canada , hep - ex/0008002 r.p .
thun and s. mckee , phys .
b * 439 * , 123 ( 1998 ) ; g. barenboim and f. scheck , phys .
b * 440 * , 332 ( 1998 ) ; t. ohlsson and h. snellman , phys .
d * 60 * , 093007 ( 1999 ) ; g. conforto _
b * 447 * , 122 ( 1999 ) ; o. haug , a. faessler and j.d .
vergados , hep - ph/0005068 | we examine the ( 3 + 1)-class of 4-neutrino mass spectra within a rigorous statistical analysis based on the bayesian approach to probability .
the data of the bugey , cdhs and karmen experiments are combined by using a likelihood function .
our statistical approach allows us to incorporate solar and atmospheric neutrino data and also the result of the chooz experiment via inequalities which involve elements of the neutrino mixing matrix and are derived from these data . for any short - baseline @xmath0
we calculate a bound on the lsnd transition amplitude @xmath1 and find that , in the @xmath0@xmath1 plane , there is no overlap between the 99% cl region allowed by the latest lsnd analysis and the region allowed by our bound on @xmath1 at 95% cl ; there are some small overlap regions if we take the bound at 99% cl .
therefore , we conclude that , with the existing data , the ( 3 + 1)-neutrino mass spectra are not very likely .
however , treating the ( 2 + 2)-spectra with our method , we find that they are well compatible with all data . |
let @xmath5 , with @xmath6 and @xmath7 real variables .
then for @xmath8 the riemann zeta function is defined by @xmath9 it follows immediately from eqn .
( 1 ) , that for any @xmath7 @xmath10 it was shown by riemann@xcite that @xmath1 can be analytically continued to a function which is meromorphic in the complex plane , that its only divergence is a simple pole at @xmath11 = 1 , and that it has no zeroes on the half - plane @xmath8 .
the riemann hypothesis@xcite ( rh ) states that the only zeroes of @xmath1 which do not lie on the real axis lie on the critical line @xmath12 . it has resisted rigorous proof for over 150 years , and is now widely considered to be the most important unsolved problem in mathematics.@xcite the significance of the rh for physics has been shown by many authors.@xcite the gram points@xcite are points on the critical line for which * im*@xmath0 but * re*@xmath13 . as stated by h. m. edwards,@xcite
to locate the gram points computationally is quite easy . "
the reason for this is that the spacing between neighboring gram points varies very smoothly , in sharp contrast to the spacing between neighboring zeroes on the critical line .
specifically , @xmath14 to show how rapidly and uniformly this converges as @xmath15 increases , we display 1 minus the ratio @xmath16 in fig
if we use the geometric mean @xmath17 , the convergence is even faster . because of this good behavior , up until now interest in the gram points has been focused primarily on their utility for locating the critical zeroes of @xmath1 . 1 minus the ratio between the actual gap between two consecutive gram points and the large @xmath15 approximation of eqn .
( 3 ) , as a function of the gram point number , @xmath15 .
log - log plot.,width=326 ] there is a special subset of the gram points which lie on contour lines * im*@xmath0 , which run from @xmath18 to @xmath19 without passing through any zero of @xmath20.@xcite these lines divide @xmath20 into strips which run roughly parallel to the real axis .
the contour lines * im*@xmath0 and * re*@xmath0 are highly constrained by the cauchy - riemann equations .
an extensive discussion of their properties , including many illustrations , has been given by arias de reyna.@xcite in the work presented here , the author used the computer program of collins to study these lines.@xcite it is not practical to find double precision values for the locations of the gram points directly from the output of collins computer program . finding approximate values from the graphical output of this program , the accurate values used for fig .
1 were taken from a list supplied to the author by michael rubinstein.@xcite as shown by arias de reyna,@xcite two * im*@xmath0 contour lines which pass through special gram points can not intersect when @xmath21 .
however , there is no general proof that contour lines of * im*@xmath0 can not intersect , although no such intersections are known to exist .
a proof of this would immediately imply that all the zeroes of @xmath1 are simple , an unsettled issue which has been of interest for a long time .
if it is possible for two * im*@xmath0 contour lines to intersect , then the strips bounded by the contour lines running through the special gram points might not be well defined . in order for two of these * im*@xmath0 contour lines to intersect at a point @xmath22 , two conditions must be satisfied simultaneously .
the first condition is that @xmath23 , where @xmath24 is the derivative of @xmath1 .
the second condition is that * im*@xmath25 .
it is known@xcite that the rh requires that there be no zeroes of @xmath24 for @xmath26 .
therefore , the strips are well defined if the rh is true and in addition all the zeroes of @xmath1 are simple . in this work we will study the properties of the special gram points and their associated contour lines . using data up to a height @xmath3 , we will find numerical evidence of some remarkable behavior which is a result of the way the smooth , monotonic variation in the spacings between neighboring gram points , as illustrated in fig
. 1 , fits together with the strips bounded by the * im*@xmath0 contour lines .
the average width of these strips does not depend on the height @xmath7 .
the author suspects that this behavior is related in some way to dyson s conjecture@xcite about the connection of the rh with quasicrystals .
we define the riemann - siegel phase @xmath27 by @xmath28 then the contour lines at the top and the bottom of each strip have @xmath29 = 1 , _ i.e. _ @xmath30 is an integer multiple of @xmath31 . counting the crossings of the critical line , @xmath11 = 1/2 , by these contour lines passing through the special gram points , we plot the number of strips as a function of the height @xmath7 for the first 1102 strips in fig . 2 .
height of the bottom of a strip on the critical line @xmath6 = 1/2 as a function of strip number , @xmath32 , for the first 1102 strips.,width=326 ] the linear least squares fit to the data is @xmath33 where the numbers in parentheses are the statistical errors in the last significant figure .
since there is a nontrivial distribution of strip widths , there is a small amount of jitter of the data around the fitting line .
however , there is absolutely no indication of any curvature in the fit .
the bottom of the first strip crosses the critical line at a height of @xmath7 = 9.6669080561 ... .
it is therefore somewhat mysterious that the y - intercept of the fitting line is consistent ( within the statistical error ) with a value of zero .
if one fits the heights of the tops of the strips instead , one finds ( unsurprisingly ) that the slope of the fitting line is essentially unchanged , but the y - intercept is now found to be 9.07(5 ) . for large positive @xmath6 eqn .
( 1 ) can be approximated by its first two terms . under this condition
@xmath34 thus the strip boundaries for large positive @xmath6 and @xmath35 will be @xmath36 where the strip number , @xmath32 , is a positive integer .
the numerical value of @xmath37 is 9.06472028 ... . assuming the rh is correct
, it seems a reasonable conjecture that the strips remain essentially horizontal for any value of @xmath7 when @xmath38 , which implies that the slope of the least squares fit for the heights of the bottom of each strip will be independent of @xmath6 .
the author sees no reason , however , why the y - intercept of this fit should be independent of @xmath6 .
in fact , it appears that for @xmath39 this y - intercept becomes clearly greater than zero .
the reader should note that , since the sum on the right hand side of eqn .
( 1 ) does not even converge for @xmath40 , it is rather surprising that the data taken on the critical line , shown in fig .
2 , are well fit by a straight line with a slope of @xmath41 .
based on the analysis of berry and keating,@xcite for example , one might have expected to see oscillations about this line coming from the higher order terms of the sum .
somewhat similar ideas have been discussed by steuding and wegert.@xcite we now examine in detail the departures from the straight - line behavior , to see if there is anything resembling berry - keating oscillations .
we do this by subtracting @xmath42 from the actual height of the @xmath32-th special gram point .
the results for various ranges of @xmath32 up to 1102 are shown in fig . 3 through fig .
deviation of the position of the @xmath32-th special gram point from the average behavior , as a function of strip number , for the first 70 strips.,width=326 ] deviation of the position of the @xmath32-th special gram point from the average behavior , as a function of strip number , for strip numbers 70 to 140.,width=326 ] deviation of the position of the @xmath32-th special gram point from the average behavior , as a function of strip number , for strip numbers 140 to 280.,width=326 ] deviation of the position of the @xmath32-th special gram point from the average behavior , as a function of strip number , for strip numbers 280 to 560.,width=326 ] deviation of the position of the @xmath32-th special gram point from the average behavior , as a function of strip number , for strip numbers 560 to 1102.,width=326 ] note that all of the points in the entire data set from 1 to 1102 have deviations in the range -2 to 2 , and that the typical size of the deviations ( _ i.e. _ the variance of the distribution of deviations ) appears to be fairly insensitive to @xmath32 .
it is clear , however , that the points are not located randomly .
there are obvious sets of nested arches which appear near certain values of @xmath32 .
this demonstrates rather extensive correlations exist in those regions where the arches are present .
it turns out that the heights where the nested arches appear are given by the simple expression @xmath43 where @xmath44 and @xmath45 are ( positive ) integers with no common prime factors . in units of the strip number , @xmath32
, this expression has the form @xmath46 the most prominent arches in figs .
3 to 7 are centered at locations given by @xmath47 , for @xmath44 = 4 , 5 , 6 , 7 , 8 , 9 and 10 .
sets of secondary arches are found at values of @xmath48 . in fig .
6 and fig .
7 there are also somewhat less defined arches which correspond to @xmath49 and perhaps @xmath50 .
the nested arches above and below the main arches are due to strips having one more or one less zero than they
ought " to have .
consistent with this idea , the vertical spacings between the nested arches for @xmath51 and @xmath49 are about 1/2 and 1/3 , respectively , of the vertical spacings of the @xmath52 arches . in addition , all of these vertical spacings are decreasing slowly as @xmath7 increases . for the @xmath52 case eqn .
( 8) is obtained by requiring that @xmath53 divided by the spacing between consecutive gram points at height @xmath7 be equal to @xmath44 .
this is a kind of resonance effect between the average height of a strip and the spacing between gram points at height @xmath7 .
similarly , the expression for larger values of @xmath45 may be thought of as higher order resonances .
it would , of course , be helpful to have an explicit analytical derivation of eqn .
at this stage , we can only speculate about the behavior of these arches for very large values of the height @xmath7 .
it seems reasonable to guess that the @xmath52 and @xmath51 arches will continue to be present for large values of @xmath7 .
it is not clear to the author what will happen at large @xmath7 for larger values of @xmath45 .
number of critical zeroes in a strip versus strip number , for the first 1102 strips .
the @xmath54-axis is scaled logarithmically.,width=326 ] the number of critical zeroes in a strip versus strip height is shown in fig .
we see that the average number of zeroes increases logarithmically with strip number .
it has been known for many years that the density of the critical zeroes increases approximately as @xmath55 , defined in eqn .
the average density of critical zeroes as a function of @xmath7 is actually known to a much greater precision than this , because it is identical to the average density of gram points .
this follows directly from the fact that the rh and the assumption that all zeroes are simple implies that the number of zeroes in any strip must be equal to the number of gram points , counting the special gram point on the bottom edge but omitting the one on the top edge .
( number of zeroes on the strip)/(strip width ) versus strip number , for the first 1102 strips .
the @xmath54-axis is scaled logarithmically.,width=326 ] due to eqn .
( 3 ) the width of each strip on the critical line @xmath40 is determined to high accuracy by the height @xmath7 and the number of zeroes in the strip . in fig .
9 we display the function ( number of zeroes on the strip ) divided by ( strip width ) versus the strip number .
it is thus no surprise that these data lie on a straight line whose slope is determined by eqn . 3 , as shown in fig .
strip width on the critical line @xmath40 as a function of strip number , for the first 1102 strips .
the @xmath54-axis is scaled logarithmically.,width=326 ] it is more revealing to plot the strip width versus strip number , shown in fig .
we see that the data points sit close to lines which are hyperbolic .
line " contains all the points corresponding to a particular number of zeroes in a strip .
the spacing between these lines is decreasing with @xmath32 logarithmically .
there is an observable tendency for the range of strip widths to increase as @xmath32 increases , so it is not obvious what the large @xmath32 behavior will be . in order to understand the behavior of the deviations of the points from the lines , we perform an analysis similar to the one of figs . 3 to 7 . in fig . 11 to fig . 15
, we plot the deviations of the data from the straight line of fig .
note that the vertical scales of figs .
11 to 15 are all different .
the average deviation is decreasing approximately logarithmically as @xmath32 increases .
this decrease compensates for the decrease in the spacing between the hyperbolic lines as the number of zeroes per strip increases .
therefore we expect that the qualitative behavior of fig . 10 , lines "
whose width is much narrower than the spacing between them , remains valid indefinitely as @xmath32 increases .
deviation of the number of zeroes on strip @xmath32 divided by the width of strip @xmath32 from a straight line fit , as a function of strip number , for the first 70 strips.,width=326 ] deviation of the number of zeroes on strip @xmath32 divided by the width of strip @xmath32 from a straight line fit , as a function of strip number , for strip numbers 70 to 140.,width=326 ] deviation of the number of zeroes on strip @xmath32 divided by the width of strip @xmath32 from a straight line fit , as a function of strip number , for strip numbers 140 to 280.,width=326 ] deviation of the number of zeroes on strip @xmath32 divided by the width of strip @xmath32 from a straight line fit , as a function of strip number , for strip numbers 280 to 560.,width=326 ] deviation of the number of zeroes on strip @xmath32 divided by the width of strip @xmath32 from a straight line fit , as a function of strip number , for strip numbers 560 to 1102.,width=326 ] ( number of the primary zero @xmath56)/(number of zeroes ) versus strip number , for the first 1102 strips.,width=326 ] due to the requirements of the cauchy - riemann equations , in each strip there is one special zero , which we will call the primary zero .
each primary zero has the property that the contour with phase @xmath30 = 0 going out of it extends to @xmath57 , at a height @xmath58 for all the other critical zeroes , the contour with phase @xmath59 goes to @xmath60 .
one can now ask the question where is the primary zero located in the strip ?
" it seems obvious , by reason of symmetry , that the average position of the primary zero should be at the center of the strip . however , there is no symmetry reason why the probability distribution for the primary zero should be uniform . in fig .
16 we display the values for the function ( number of the primary zero ( counting from the bottom of the strip ) @xmath61 ) divided by ( number of zeroes in the strip ) versus the strip number .
the subtraction of 0.5 in the numerator is necessary so that this function has the value 0.5 when the primary zero is the middle zero .
the linear least squares fit to the data of fig .
16 shows that the average position of the primary zero is indeed at the center of the strip .
remarkably , one sees that the width of this probability distribution seems to be independent of the height .
this observation is quantitatively confirmed by calculating the variance for subsets of the data , which always gives a result close to 0.014 , independent of the range of @xmath7 used .
if this probability distribution remains nontrivial ( _ i.e. _ neither becoming uniform nor collapsing ) in the limit @xmath62 , then we must conclude that all of the zeroes in a strip are a collective entity , and that the positions of these zeroes are highly correlated with each other .
in this work we have done an analysis of some properties of the special gram points of the riemann zeta function . we have uncovered some previously unknown facets of the behavior of this remarkable function along its critical line .
the most remarkable of these occur when the ratio between @xmath41 , the average strip width , and the spacing between gram points passes through integers or rational numbers of small denominator .
this seems to be some kind of a resonance effect , but its origin is not clear at this point . the author thanks stephen wolfram for making the wolfram cdf player 8 available as a free public download .
he also thanks peter sarnak and elliott lieb for helpful conversations .
he thanks michael rubinstein for a list of the locations of the gram points , which made the work described here possible . , edited by p. borwein , s. choi , b. rooney and a. weirathmueller ( springer , new york , 2008 ) .
j. b. conrey , notices amer . math
. soc . * 50 * , 341 ( 2003 ) . m. v. berry and j. p. keating , siam rev . *
41 * , 236 ( 1999 ) .
g. sierra and p. k. townsend , phys .
* 101 * , 110201 ( 2008 ) .
r. mack _ et al .
a * 82 * , 032119 ( 2010 ) .
d. shumayer and d. a. w. hutchinson , rev .
phys . * 83 * , 307 ( 2011 ) .
y. v. fyodorov , g. a. hiary and j. p. keating , phys .
lett . * 108 * , 170601 ( 2012 ) . h. m. edwards , _
riemann s zeta function _ , ( academic press , new york , 1974 ) . h. m. edwards , _ op .
_ , p. 126 .
j. arias de reyna , arxiv : math/0309433v1 ( 2003 ) .
r. fisch , arxiv:1210.3618v1 ( 2012 ) .
b. collins , _ contour plots of the zeta function _ from the wolfram demonstrations project , http://demonstrations.wolfram.com/contourplotsofthezetafunction/ ( 2009 ) . in manual mode , this program will accept decimal fractions as input values .
m. o. rubinstein , unpublished .
n. levinson and h. l. montgomery , acta math . * 133 * , 49 ( 1974 ) .
f. j. dyson , notices amer .
soc . * 56 * , 212 ( 2009 ) .
j. steuding and e. wegert , experimental math .
* 21 * , 235 ( 2012 ) . | we have studied some properties of the special gram points of the riemann zeta function which lie on contour lines * im*@xmath0 which do not contain zeroes of @xmath1 .
we find that certain functions of these points , which all lie on the critical line * re*@xmath2 = 1/2 , are correlated in remarkable and unexpected ways .
we have data up to a height of @xmath3 , where @xmath4 . |
external biasing forces are often applied to enhance sampling in regions of phase space which would otherwise be rarely observed @xcite .
these perturbing forces are applied in both time - dependent , as in single - molecule pulling @xcite and steered molecular dynamics @xcite experiments , or time - indepedent forms , as in umbrella sampling simulations @xcite .
usually , the direction of bias is chosen to follow the reaction coordinate for an interesting process , such as a large conformational change in a biological macromolecule . with judicious reweighing of biased ensemble properties ,
previous workers have reconstructed the potential of mean force ( pmf ) along this bias coordinate @xcite . the ability to reconstruct pmfs in multiple dimensions and along multiple alternative degrees of freedom
would significantly expand the data analysis tool kit for biased experiments .
although the appropriate choice of reaction coordinate is fundamental to kinetic calculations , transition state theory , and the accurate separation of distinct microstates , its selection is limited by prior knowledge of the system
. it would be useful to compute pmfs along alternate potential reaction coordinates , in one or several dimensions , after biased data is collected along an initial , possibly arbitary choice .
in addition , by using multidimensional pmfs , the mutual entropy @xcite can be used quantify and compare correlations between modes .
here , i develop methods to perform these calculations . they are applicable to both computer simulations , in which system properties are completely known , or laboratory experiments in which several properties can be simultaneously measured .
the pmf , @xmath0 , along a single coordinate , @xmath1 , is defined as @xmath2 } = \frac { \int \delta [ x - x(\bm{r } ) ] e^ { - \beta \ { h_0(\bm{r } ) \ } } \ , d\bm{r } } { \int e^ { - \beta \ { h_0(\bm{r ' } ) \ } } \ , d\bm{r } ' } \label{eq : pmf_1d}\end{aligned}\ ] ] where @xmath3 is an arbitary constant , @xmath4 is @xmath5 , and @xmath6 is the unperturbed system energy as a function of the phase space position @xmath7 . in two dimensions ,
@xmath0 becomes @xmath8 and @xmath9 $ ] is replaced by @xmath9 \delta [ y - y(\bm{r})]$ ] .
generalizing to multiple dimensions , @xmath0 becomes @xmath10 and the delta function is replaced by @xmath9 \prod_{i=1}^{n } \delta [ y_i - y_i(\bm{r})]$ ] , where @xmath11 is the number of dimensions .
if the pmf encompasses all system dimensions , it can simply be referred to as the potential .
multidimensional pmfs can be reconstructed from time - dependent biasing experiments by applying the nonequilibrium work relation @xcite in a manner analogous to hummer and szabo @xcite . in a single - molecule pulling experiment with the hamiltonian @xmath12 $ ] , where @xmath13 $ ] is the time - dependent external bias , the biased probability density function is @xmath14 \ } } } { \int e^ { - \beta \ { h_0(\bm{r ' } ) + v[\bm{r'},0 ] \ } } \ , d\bm{r } ' } = \langle \delta ( \bm{r } - \bm{r}(t ) ) e^ { - \beta w_t } \rangle \label{eq : phase_space_fe}\end{aligned}\ ] ] where @xmath15 is the accumulated work , @xmath16 @xcite . to calculate the pmf along the pulling coordinate , hummer and szabo consider the special case that the perturbation depends only on the biasing coordinate and time , @xmath17 $ ] .
they then multiply equation [ eq : phase_space_fe ] by @xmath18 $ ] , integrate over all phase space , and use the definition in equation [ eq : pmf_1d ] to obtain @xmath19 \right ) } \rangle \label{eq : pullcoord_fe}\end{aligned}\ ] ] if we instead multiply equation [ eq : phase_space_fe ] by a two - dimensional delta function , we then obtain the two - dimensional pmf . similarly , multiplying by a multidimensional delta function will lead to a multidimensional pmf .
( although this is possible in principle , high dimensionality will lead to greater difficulty with sampling . ) in the two - dimensional case , the pmf is @xmath20 \right ) } \rangle \label{eq : twod_fe}\end{aligned}\ ] ] in any particular time slice , the position will be more highly sampled near the minima of the bias potential . to improve accuracy over the entire range of @xmath1 ,
data from different time slices can be combined by adapting the weighted histogram analysis method ( wham ) @xcite .
after unbiasing with respect to the accumulated work as well as the time - dependent potential @xcite , the resultant pmf is @xmath21
the usefulness of reconstructing @xmath22 from experiments biased along @xmath1 is expected to be related to their degree of correlation .
one standard measure for the correlation of two coordinates with known probability density , @xmath23 , is the mutual entropy @xcite , @xmath24 where @xmath25 and @xmath26 are the marginal probabilities along @xmath1 and @xmath22 . as a reference ,
let @xmath27 be a probability density estimate based on a histogram of sampling along @xmath22 without any correction for external bias along @xmath1 .
let the estimate of the pmf using @xmath27 , @xmath28 , be known as the nave estimate .
if @xmath1 and @xmath22 are completely independent , @xmath29 , @xmath30 , and the mutual entropy is zero . in this case
, a nave estimate of the pmf along y should be as good as a reconstruction from equation [ eq : time_slice ] .
if @xmath1 and @xmath22 are highly correlated , then the mutual entropy should be high .
the nave estimate will fail , making the utility of equation [ eq : time_slice ] more evident .
equation [ eq : time_slice ] was tested on series of two - dimensional potential surfaces consisting of a linear combination of @xmath31 and @xmath32 + 1.25 \times 10^{-7}(2x+3y)^2(2x+3y-180)^2 \label{eq : ub}\end{aligned}\ ] ] according to @xmath33 where @xmath34 is in units of @xmath35 and positions @xmath1 and @xmath22 are in units of @xmath36 .
@xmath37 and @xmath38 both have minima at ( 0,0 ) and ( 45,30 ) , and @xmath37 has additional minima at ( 0,30 ) and ( 45,0 ) .
part c of figures [ fig : u0 ] , [ fig : p30 ] , [ fig : d75 ] , and [ fig : p100 ] show the potential @xmath39 at @xmath40 of 0 , 0.30 , 0.75 , and 1.00 , respectively .
pmf reconstruction on the surface with @xmath41 from umbrella sampling .
* a*. pmfs along the @xmath22 coordinate .
the blue line is the theoretical surface , the red asterisks are the biased reconstruction , and green dots are the nave reconstruction . * b*. contour plot of the reconstructed potential . * c*. contour plot of the theoretical potential .
the scale for both b and c and indicated by the side color bar .
* d*. pmfs along the @xmath1 coordinate .
annotation is the same as in * a*. ] pmf reconstruction on the surface with @xmath42 from time - dependent biasing experiments .
annotation is the same as in figure [ fig : u0 ] . ]
pmf reconstruction on the surface with @xmath43 from diffusion experiments .
annotation is the same as in figure [ fig : u0 ] , except there are no wham reconstructions .
] pmf reconstruction on the surface with @xmath44 from time - dependent biasing experiments .
annotation is the same as in figure [ fig : u0 ] . ] in @xmath37 ( fig .
[ fig : p100]c ) , @xmath1 and @xmath22 are completely independent , while in @xmath38 ( fig .
[ fig : u0]c ) , they are highly correlated .
this is quantified by the mutual entropy ( fig .
[ fig : mutentropy ] ) , which was calculated by assuming a boltzmann distribution for @xmath1 and @xmath22 , @xmath45 , and numerically integrating equation [ eq : mut_entropy ] between @xmath46 and @xmath47 via adaptive simpson quadrature .
entropy as a function of the coupling parameter @xmath40 .
the mutual entropy , @xmath48 , is designated by connected squares .
the joint entropy , @xmath49 , and the marginal entropies , @xmath50 and @xmath51 , are labeled and marked by connected circles . ] as expected , mutual entropy decreases as the potential becomes more like @xmath37 .
the joint entropy , @xmath52 , and the marginal entropies , @xmath53 and @xmath54 , are mostly stable as a function of @xmath40 .
these magnitude of these factors can not be directly compared to the mutual entropy because they are offset by an arbitrary constant , whereas mutual entropy is invariant to scale .
brownian dynamics simulations @xcite were run on the series of potential surfaces .
the progress of a dynamical variable is given by @xmath55 , where @xmath56 is the force , @xmath57 is the temperature ( 300 k ) , @xmath58 is the diffusion constant ( 1200 @xmath59 ) , @xmath60 is the time step ( 1 ms ) , and r is a normally distributed random variable . after 1000 steps of equilibration , umbrella sampling and pulling simulations
were run for 100 iterations of 10000 time steps .
in addition , 10 unbiased diffusion simulations were run for 100000 time steps .
umbrella sampling was performed with a harmonic bias , @xmath61 , setting the spring constant , @xmath62 , at @xmath63 and the bias center , @xmath64 , at one of 100 evenly spaced positions between -5 and 50 nm .
pulling simulations were performed with the time - dependent perturbation , @xmath65 = 1/2 k_s ( b(t ) - x(\bm{r}))^2 $ ] , with @xmath62 set to @xmath66 . a periodic biasing program @xcite , @xmath67 , which has one period per trajectory , was used . accumulated work was numerically integrated by @xmath68(b_j - b_{j-1})$ ] , where @xmath69 is the position of the bias center at time step @xmath70 .
theoretical surfaces , @xmath71 and @xmath72 , were calculated by assuming a boltzmann distribution , @xmath45 , and numerically integrating @xmath23 over the other coordinate ( with limits of @xmath46 and @xmath47 ) using adaptive simpson quadrature to obtain a marginal probability density .
pmfs are then obtained by taking natural logorithms and multiplying by @xmath73 . in time - dependent biasing experiments ,
reconstructed surfaces for @xmath1 were calculated using equation [ eq : pullcoord_fe ] , two - dimensional surfaces from equation [ eq : time_slice ] , and surfaces along @xmath22 by integrating over @xmath1 ; from the estimate of @xmath23 , the sum is taken over @xmath1 to yield an unnormalized @xmath26 , from which @xmath74 is calculated .
pmfs from umbrella sampling were calculated as described by roux @xcite .
minima of all pmfs were set to zero .
errors were calculated by the root mean square deviation from the theoretical surface at 50 points between @xmath75 and @xmath76 , ranges which include the zeroes of @xmath37 and @xmath38 . the root mean square error ( rmse ) along @xmath22 is defined as @xmath77 ^ 2}$ ] .
when @xmath1 and @xmath22 are completely independent , the nave estimate for @xmath78 is as good as the estimate from equation [ eq : time_slice ] ( fig .
[ fig : p100]a ) .
if @xmath1 and @xmath22 are highly correlated , however , the nave estimate fails and a wham - informed reconstruction method is necessary ( fig . [
fig : u0 ] ) .
there is no simple dependence , however , between error and mutual entropy ( fig .
[ fig : error ] ) , as other factors of the free energy surface come into play .
root mean square error for @xmath22 pmf estimates as a function of mutual entropy .
reconstruction errors for pulling experiments are indicated by squares and for umbrella sampling by circles .
errors from unbiased diffusion on the surface are indicated by diamonds .
errors in wham - reconstructed surfaces are connected by solid lines , while the nave estimate errors are not connected . ] for example , due to the highly diffusive nature of the dynamics and the large timestep necessary for fast sampling , sampling is erroneously improved near the boundaries of small low - energy wells .
this leads to systematic free energy underestimates in all simulations , even those without any external bias ( i.e. fig .
[ fig : d75 ] ) .
although there are errors in wham - informed reconstructions from simulations with external bias , they are no worse than errors from long simulations with free diffusion ( fig . [
fig : error ] ) . on certain free energy surfaces , phase space sampling along the @xmath22 dimension may be limited by trapping in local minima .
this is especially true if @xmath1 and @xmath22 are relatively independent and biasing along @xmath1 does not enhance conformational sampling along @xmath22 .
sampling is generally limited in high - energy regions of phase space ( see figs . [
fig : u0 ] , [ fig : p30 ] , [ fig : d75 ] , and [ fig : p100 ] ) . in these regions ,
the most accurate two - dimensional pmf reconstruction would require biasing along both @xmath1 and @xmath22 , which can be accounted for by a simple modification of equation [ eq : time_slice ] . in general , accurate pmf reconstruction in higher dimensions
may require biasing along coordinates that closely correspond to these dimensions .
close correspondence may be indicated by high mutual entropy ; this aspect of analysis will be left to further study .
root mean square error for @xmath22 pmf estimates as a function of number of simulations , @xmath11 .
error bars indicate the mean and standard deviations of rmse from five reconstructions using @xmath11 simulations .
reconstruction errors for pulling experiments are indicated by squares and for umbrella sampling by circles .
unbiased simulations were run for 10 times as long and are therefore spaced 10 times farther apart on the x axis . ] on the toy two - dimensional surfaces studied , pmf reconstructions along @xmath22 using equation [ eq : time_slice ] actually converge more quickly than equilibrium umbrella sampling and reconstruction from unbiased experiments ( fig . [
fig : convergence ] ) .
while time - dependent biasing experiments will sample the whole span of the bias coordinate _ de facto _ , umbrella sampling will inherently require a certain number of simulations at different bias centers to attain this broad range .
similarly , unbiased simulations will need to be run for longer in order to reach more distal regions of phase space . generally , it is preferable to run a greater number of nonequilibrium experiments than a single long equilibrium experiment because the outcome of each individual trajectory can be highly dependent on initial conditions .
these sampling problems described on two - dimensional surfaces are likely to be enhanced in systems with higher complexity .
the main caveat with this method , as with all methods based on the nonequilibrium work relation , is that with greater free energy differences , the likelihood of observing a trajectories with negative dissipative work ( work that is less than the free energy difference ) is increasingly unlikely .
pulling experiments are poised to find broader utility in studies of nanoscale systems , particularly when free energy differences between states are not prohibitively large .
i hope this work will find wide application in computational and laboratory experiments and stimulate further research in the area .
matlab source code to perform the described analysis is available at http://mccammon.ucsd.edu/~dminh / software/.
i would like to thank r. amaro , c.e .
chang , j. gullingsrud , and j.a .
mccammon for helpful discussions .
i am funded by the nih molecular biophysics training grant at ucsd .
mccammon group resources are supported by the nsf ( mcb-0506593 ) , nih ( gm31749 ) , hhmi , ctbp , nbcr , w.m .
keck foundation , and accelrys , inc . | external biasing forces are often applied to enhance sampling in regions of phase space which would otherwise be rarely observed . while the typical goal of these experiments is to calculate the potential of mean force ( pmf ) along the biasing coordinate , here
i present a method to construct pmfs in multiple dimensions and along arbitary alternative degrees of freedom .
a protocol for multidimensional pmf reconstruction from nonequilibrium single - molecule pulling experiments is introduced and tested on a series of two dimensional potential surfaces with varying levels of correlation . reconstruction accuracy and convergence from several methods - this new protocol , equilibrium umbrella sampling , and free diffusion -
are compared , and nonequilibrium pulling is found to be the most efficient . to facilitate the use of this method
, the source code for this analysis is made freely available .
department of chemistry and biochemistry , center for theoretical biological physics , and howard hughes medical institute , university of california at san diego , san diego , california 92093 |
there are two types of classical relativistic direct interaction theories ( rdit ) , namely , manifestly poincar - invariant four - dimensional formulations and three - dimensional approaches @xcite . applying each of them to the description of real physical @xmath0-particle systems we come across the problem of space - time interpretation of some formal constructions . in four - dimensional approaches difficulties
are caused by the presence of redundant variables specific to many - time formalisms and in three - dimensional ones they are caused by the so - called no - interaction theorem . in the relativistic hamiltonian mechanics , this theorem forbids identification ( in the interaction zone ) of canonical variables @xmath1 with physical ( covariant ) position coordinates @xmath2 ( @xmath3 ; @xmath4 ) which transform as a space part of the events in the minkowski space - time @xcite . in the lagrangian formulation @xcite the no - interaction theorem asserts that the interaction lagrangian depends inevitably on all higher time derivatives @xmath5 , @xmath6 , of the physical particle coordinates .
this theorem is essentially related to the use of individual particle variables ( pv ) in both the hamiltonian and lagrangian approaches . in papers
@xcite the following approach to the construction of rdit was proposed .
the starting point is the 3-dimensional poincar - invariant lagrangian formalism @xcite with using the physical coordinates @xmath2 and with allowing the connection to field theories via the fokker - type formalism @xcite .
these aspects of the lagrangian description along with the possibility to use the nther theorem for obtaining the conservation laws @xcite are the main advantages of this approach from the point of view of the physical interpretation of the theory .
the next step is a transition to the canonical hamiltonian formalism admitting the quantization of the theory @xcite .
another aspect of rdit is the problem of how to separate the motion of the system as a whole from the internal motion of its constituents .
this problem can be solved in a manner similar to that of the non - relativistic mechanics case by introducing center - of - mass variables ( cmv ) .
relativistic generalization of the non - relativistic notion of the center - of - mass is ambiguous .
the most frequently used are , firstly , the center - of - inertia ( covariant center - of - mass ) whose coordinates @xmath7 are transformed by the poincar group as 3-coordinates of the minkowski - space points , and , secondly , the center - of - spin ( or canonical center - of - mass ) with the canonical coordinates @xmath8 @xcite .
external variables @xmath8 and @xmath9 describing the motion of a system as a whole , along with the corresponding set @xmath10 , @xmath11 of @xmath12 internal variables are often used in the relativistic hamiltonian mechanics @xcite-@xcite . in the relativistic lagrangian formalism
it is natural to choose the coordinates of the center - of - inertia @xmath13 as external position variables and to formulate the problem of constructing the relativistic lagrangian mechanics of an @xmath0-particle system in terms of cmv .
this set also contains @xmath14 internal variables @xmath15 @xcite-@xcite .
the internal variables @xmath15 may be introduced by postulating their transformation properties with respect to the poincar group .
the corresponding relations which are written down in section 2 constitute the prerequisites which enable us to find a connection between cmv and pv and , accordingly , between the relativistic lagrangians in terms of these two classes of variables .
an important feature of cmv is the existence of the poincar - invariant interaction lagrangians which depend only on a finite number of time derivatives ( section 3 ) .
the transition to the hamiltonian formalism ( section 4 ) leads to the well - known bakamjian - thomas model supplemented by the relations between the canonical variables and the physical particle positions . in sections 5 and 6
the first - order quasi - relativistic approximation ( up to @xmath16 terms ) is considered from both the classical and the quantum mechanical viewpoints .
relativistic corrections to the energy spectrum of a two - particle system with coulomb - like interactions are calculated as an illustration .
since the external variables @xmath7 describe the motion of an @xmath0-particle system as a whole , we require their transformation properties with respect to the poincar group to be the same as those of the physical coordinates @xmath2 of a single point particle . in the instant form of dynamics
@xcite they are @xcite : @xmath18 here @xmath19 , @xmath20 , @xmath21 , and @xmath22 are generators of time and space translations , space rotations , and boosts , respectively , and @xmath23 is the total time derivation operator . as usual , the dot denotes the time derivation ( e.g. , @xmath24 are the components of the external velocity @xmath25 ) .
we postulate further that similar to the case of non - relativistic mechanics the internal cmv @xmath15 are the components of the translationally invariant 3-vectors @xmath26 @xmath27 which do not depend explicitly on time : @xmath28 the transformation of these variables with respect to boosts is determined by the poincar group structure .
it can be shown that the commutation relations of the poincar algebra @xmath29 , along with equations ( [ t_r])([t_r ] ) , lead to the following conditions @xcite : @xmath30 the expressions for generators @xmath31 of the poincar group in terms of cmv may be obtained from equations ( [ t_r])([tr ] ) immediately @xcite .
the evolution of a particle system is described by the solutions of the euler - lagrange - ostrogradsky equations @xmath32 following from the variational principle @xmath33 .
the conditions of poincar invariance of equations ( [ el ] ) may be expressed by the system of first - order differential equations for the lagrange function : @xmath34 where the auxiliary functions @xmath35 have to satisfy the compatibility conditions : @xmath36 with @xmath37 being the structure constant tensor of the poincar group @xmath17 .
for this group the relations ( [ cc ] ) admit the following set of functions @xmath35 : @xmath38 in which case the conditions ( [ pi ] ) of poincar invariance take on the form : @xmath39 the structure of equations ( [ cc ] ) is quite different from the structure of the poincar invariance conditions in terms of pv @xmath2 @xcite .
whereas the interaction lagrangian as a function of the particle variables is defined on @xmath40 , equations ( [ cc ] ) admit lagrangians that depend on cmv and their time derivatives of finite orders .
the simplest structure of @xmath41 is given by @xmath42 here @xmath43 and @xmath44 is an arbitrary function of scalar products of the vectors @xmath45\,,\qquad { \mbox{\boldmath$\beta$}}_b = \gamma{\mbox{\boldmath$\dot\alpha$}}_b - c^{-1}\gamma^2[{\bf\ddot r } { \mbox{\boldmath$\alpha$}}_b]\,,\\ { \mbox{\boldmath$\gamma$}}&=&\gamma{\bf\ddot r } + c^{-1}\gamma^2[{\bf\dot r}{\bf\ddot r}]+c^{-2}\gamma({\bf\dot r}{\bf\ddot r } ) { \bf\dot r}\,.\end{aligned}\ ] ] contrary to @xmath46 and @xmath47 , the @xmath48 dependence of @xmath41 is not necessary .
the lagrangian ( [ l ] ) depending on a finite number of cmv can be made equivalent to the lagrangian defined on an infinite - dimensional set of pv @xmath49 , only if the cmv are functions of all these pv .
the set of equations ( [ t_r])([tr ] ) with generators @xmath50 , @xmath20 , @xmath21 , and @xmath22 , determined in terms of pv @xcite , may be regarded as the basis for determining these functions @xcite . using the nther theorem for lagrangians with higher derivatives @xcite-@xcite we found ten integrals of motion @xmath51 corresponding to the poincar - invariance of the theory , based on lagrangian ( [ l ] ) . from their analysis
we obtain @xcite the following equation of external motion @xmath52 = { \bf k}_0\ ] ] in the rest frame of reference ( @xmath53 , @xmath54 , @xmath55 , @xmath56 ) .
it should be noted that the equation @xmath57 = 0\,,\ ] ] being the consequence of ( [ em ] ) , was obtained in @xcite for classical particles with a spin ; it is a three - dimensional form of the well - known mathisson equation .
spin * s * is equal to the internal angular momentum of the system .
solving ( [ em ] ) in a coordinate system with the origin at point @xmath58 , where @xmath59 is the value of integral @xmath60 in the rest frame and the o@xmath61 axis is directed along spin @xmath62 , we see that point @xmath63 describes a circular motion in a plane orthogonal to the spin : @xmath64 here , frequency @xmath65 ; @xmath66 and @xmath67 are constants of the integration . taking into account the usual non - relativistic limit procedure , in accordance with which the center of mass is at rest as measured in the rest frame , we must set @xmath68 .
indeed , as @xmath69 ( @xmath70 ) , the limits of the trigonometric functions in equation ( [ circ ] ) do not exist .
this means that in the set of solutions of equations ( [ el ] ) for the lagrangian ( [ l ] ) we keep only those which satisfy the standard free - particle equation @xmath71 .
there are several ways to obtain the hamiltonian formulation corresponding to the lagrangians with higher derivatives and with supplementary conditions , like @xmath71 @xcite . here
we use an approach close to that of @xcite .
let us perform a change of the variables @xmath72 , where @xmath73}{m(e+mc^2)}\,,\\ { \bf q}_b&=&\gamma{\mbox{\boldmath$\rho$}}_b - \frac{c^{-2}\gamma^2}{\gamma + 1}{\bf\dot r}({\bf\dot r}{\mbox{\boldmath$\rho$}}_b)\,.\end{aligned}\ ] ] relation ( [ qr ] ) is a definition of the canonical center - of - mass @xcite . in terms of new variables ,
the lagrangian ( [ l ] ) has the form : @xmath74 where @xmath75 contains only the terms which do not contribute to the equations of motion if the external equation of motion @xmath71 is satisfied .
therefore , the term @xmath75 may be omitted in equation ( [ l_td ] ) and , thus , the transformation to the hamiltonian description may be performed in the usual manner on the basis of lagrangian @xmath76 . using the legendre transformation we construct hamiltonian @xmath77 corresponding to lagrangian @xmath76 : @xmath78 where @xmath79 we note that the external momentum variable @xmath13 is a motion integral ( the total momentum of the system ) .
expression ( [ hbk ] ) for @xmath77 reproduces the particle - system hamiltonian of the bakamjian - thomas model @xcite .
the advantage of our way of constructing this model is the possibility to express the internal hamiltonian @xmath80 in terms of the interaction lagrangian which , in turn , may be related to some field models @xcite . on the other hand , in our approach
the canonical variables @xmath81 , and @xmath82 have a definite relation to the lagrangian cmv @xmath83 and their time derivatives @xmath84 and , therefore , to the individual particle variables @xmath85 , at least in the approximations in the parameter @xmath16 .
let us consider the main steps of our approach in the first - order approximation in @xmath16 ( i.e. quasi - relativistic or post - newtonian approximation ) .
we use the following sufficiently general form of a quasi - relativistic two - particle lagrangian with static non - relativistic potential @xmath86 @xcite : @xmath87 where @xmath88 , @xmath66 and @xmath89 are arbitrary constants and @xmath90 is an arbitrary function . for interactions which correspond to the linear field theories
, we have @xmath91 and constants @xmath66 and @xmath89 depend on the tensor structure of interaction @xcite .
for example , we have @xmath92 for a vector interaction , and @xmath93 for a scalar interaction .
further we consider a specific case of interactions which correspond to the non - relativistic long - range potential @xmath94 where @xmath67 is a constant of the interaction .
we suppose that the galileo - invariant function @xmath90 has the form : @xmath95 where @xmath96 is the constant .
let us note five significant cases covered by the interaction described in a quasi - relativistic approximation by the lagrangian of type ( [ lqr ] ) where functions ( [ u(r ) ] ) and ( [ u_1 ] ) enter : * @xmath97 , and @xmath98 : the darwin lagrangian which describes an electromagnetic interaction of two point charges ; * @xmath99 , and @xmath100 : the einstein - infeld - hoffmann lagrangian which describes a gravitational interaction of two point masses ; * @xmath101 , and @xmath102 : a gravitational interaction of charged bodies ( the baaski lagrangian @xcite ) ; * @xmath103 , and @xmath104 : a gravitational interaction in the ppn - formalism @xcite which classifies suitable theories of gravitation ( i.e. those which do not contradict the four classical experiments , satisfy the conditions of the lorentz - invariance , and admit @xmath0-particle lagrangian functions ) in the first - order approximation in @xmath16 ; * @xmath105 , and @xmath98 : the lagrangian of the quasi - relativistic approximation of the whitehead - shield theory of gravitation @xcite where the massless tensor field of the second rank as an intermediate field is assumed .
the correspondence between cmv and pv for a two - particle system in the first - order quasi - relativistic approximation has the following form @xcite : @xmath106 where @xmath107 .
using these relations in the lagrangian ( [ lqr ] ) leads to the following two - particle quasi - relativistic lagrangian in terms of cmv : @xmath108\bigr]\right)+\frac{1}{2c^2}\left\ { \left({\bf\dot r}^2-\left(\frac{\mu}{m } + 2a\right){\mbox{\boldmath$\dot\rho$}}^2\right)u(\rho ) + \right.\nonumber\\ & + & \left.\left(\left(\frac{\mu}{m}-2b\right)({\mbox{\boldmath$\rho$}}{\mbox{\boldmath$\dot\rho$}})^2 - [ { \bf\dot r}{\mbox{\boldmath$\rho$}}]^2\right ) \frac{1}{\rho}\frac{{{\rm
d}}u}{{{\rm d}}\rho}\right\ } - \frac{1}{c^2}u_1(\rho)\,.\end{aligned}\ ] ] this expression agrees with the corresponding approximation of the lagrangian ( [ l ] ) . in the quasi - relativistic approximation the canonical external variable ( [ qr ] ) becomes @xmath109}{m(e+mc^2)}\right|_{{\bf\ddot r}=0}\equiv\nonumber\\ & \equiv&{\bf r } + \frac{\mu}{2mc^2}\left[{\bf\dot r}[{\mbox{\boldmath$\rho$}}{\mbox{\boldmath$\dot\rho$}}]\right]\,.\end{aligned}\ ] ] the total momentum @xmath110 is canonically conjugated to * q*. quasi - relativistic inner canonical variables are defined by the relations @xmath111\right]}{2c^2 } + \frac{1}{\mu c^2}\frac{\partial\lambda}{\partial { \mbox{\boldmath$\dot\rho$}}}\,,\\ { \mbox{\boldmath$\pi$}}&=&\left(1-\frac{{\bf\dot r}^2}{2c^2}\right ) \frac{\partial l}{\partial { \mbox{\boldmath$\dot\rho$ } } } - \frac{1}{c^2}\frac{\partial\lambda}{\partial{\mbox{\boldmath$\rho$}}}\,,\end{aligned}\ ] ] where @xmath112 is an arbitrary galileo - invariant function .
choosing this function in the form @xmath113 we obtain the hamiltonian function depending on the square of the inner momentum @xmath114 and therefore being convenient for quantization : @xmath115 here @xmath116 and @xmath117
starting with the hamiltonian function ( [ h ] ) we write down the stationary schrdinger equation : @xmath118 the total wave function @xmath119 of the system is equal to the product @xmath120 of the external wave function @xmath121 and the inner one @xmath122 .
we consider the states with sharply defined total momentum @xmath123 , so the external wave function is then a plane wave , @xmath124 , where the wave vector @xmath125 . due to the separation of the external and internal variables in equation ( [ se ] ) , we obtain the following wave equation : @xmath126 where @xmath127 here @xmath128 and @xmath123 are eigenvalues of the non - relativistic hamiltonian and total momentum operators , respectively .
the internal hamiltonian operator is defined by the formula @xmath129 where the first term is some hermitian operator which corresponds to the classical function @xmath130 .
we use the following sufficiently general quantization rule : @xmath131\hat\pi_j + \lambda\hat\pi^2(a(q))\,,\ ] ] where the inner momentum operator is chosen in the usual form : @xmath132 this rule preserves the rotational invariance of the inner hamiltonian and depends on one parameter @xmath133 .
the value @xmath134 corresponds to the symmetrization @xmath135 . to illustrate our approach we consider the coulomb - like potential ( see equations ( [ u(r ) ] ) and ( [ u_1 ] ) ) . in this case the functions @xmath136 and @xmath137 take the form : @xmath138 here @xmath139 and the parameters @xmath66 , @xmath89 , @xmath96 , and @xmath67 are defined by the type of interaction ( see section 5 ) . after some calculations we arrive at the following hamiltonian operator : @xmath140 where @xmath141 denotes a well - known transcendental number instead of the value of the inner momentum .
we shall study the discrete spectrum of @xmath77 , when @xmath142 .
the use of methods of the perturbation theory ( e.g. , see @xcite ) allows us to obtain quasi - relativistic corrections to the non - relativistic bohr expression @xmath143 where @xmath144 is the main quantum number .
let us rewrite the hamiltonian operator ( [ h ] ) in the form @xmath145 where @xmath146 is a non - relativistic energy operator possessing spectrum ( [ e0 ] ) . according to the stationary perturbation theory @xcite , the quasi - relativistic term @xmath147 is equal to the matrix element @xmath148 where @xmath149 are the eigenfunctions of operator ( [ h_0 ] ) and @xmath150 is a quasi - relativistic perturbation of the non - relativistic energy operator . while calculating ( [ me ] ) we take into consideration that @xmath149 are the eigenfunctions of @xmath151 .
we also use the average values of the following functions of @xmath152 @xcite : @xmath153 therefore , the quasi - relativistic correction to the energy spectrum is given by @xmath154 we note a dependence of the obtained spectrum on the quantization parameter @xmath155 . for the case of an electromagnetic interaction ,
when the parameters @xmath92 , @xmath98 ( see section 5 ) , spectrum agrees with the results of the conventional quantum electrodynamics provided that @xmath134 .
let us consider a hydrogen - like atom .
then the constant of interaction @xmath156 , where @xmath157 is the charge of the nucleus .
since the mass of the nucleus is much greater than the rest mass of electron @xmath158 , we have @xmath159 and , therefore , the ratio @xmath160 . here
@xmath161 is the total mass of the atom .
introducing these parameters into equations ( [ a ] ) and ( [ v ] ) we obtain @xmath162 and @xmath163 . according to equation ( [ e1 ] ) , the quasi - relativistic correction is equal to @xmath164 it is in good agreement with the quasi - relativistic correction to the hydrogen - like spectrum which follows from the dirac equation ( see @xcite ) .
now let us consider parapositronium .
the mass of an electron is equal to the mass of a positron and , therefore , @xmath165 and the ratio @xmath166 .
whence the constants @xmath167 and @xmath168 . putting these parameters into equations ( [ e1 ] ) we re -
obtain an expression for the quasi - relativistic splitting of the parapositronium spectrum which is conventionally found by using the breit equation @xcite : @xmath169
the transformation of particle variables into center - of - mass ones for a class of two - particle quasi - relativistic lagrangians with interactions having field - theoretical analogues has been carried out . as a result
, the motion of a system as a whole was separated from its inner motion .
a set of canonical variables in terms of which the hamiltonian has a quadratic dependence on the inner momentum variable is found .
the problem is quantized .
the quasi - relativistic splitting of the discrete energy spectrum is obtained . as expected ,
the degeneracy in the orbital quantum number is eliminated . in the specific case of an electromagnetic interaction between particles , the quasi - relativistic corrections to the non - relativistic discrete spectrum
are in agreement with the corresponding approximation of the spectrum of a hydrogen - like atom obtained by using the dirac equation ( modulus of the spin of the electron which was not taken into account ) . for parapositronium ( where the summary spin is equal to zero )
the spectrum obtained within the frame of the cmv - formalism coincides with the one derived by using the breit equation .
99 gaida r. p. quasi - relativistic systems of interacting particles //
fiz.elem .
chast . & atom .
yadra , 1982 , vol . 13 , no 2 , p. 427493 ( in russian ; transl .
in english : sov . j. part .
nucl . , 1982 , vol . 13 , p. 179 ) .
currie d. g. , jordan j. f. , sudarshan e. c. g. relativistic invariance and hamiltonian theories of interacting particles // rev .
phys . , 1963 , vol . 35 , no 2 , p. 350375 .
gaida r. p. , kluchkovsky yu .
b. , and tretyak v. i. lagrangian classical relativistic mechanics of a system of directly interacting particles .
. mat .
1980 , vol .
44 , no 2 , p. 194208
( in russian ; transl . in english : theor . math . phys . , 1980 ,
44 , p. 687 ) .
gaida r. p. , kluchkovsky yu .
b. , and tretyak v. i. lagrangian classical relativistic mechanics of a system of directly interacting particles .
. mat .
1980 , vol . 45 , no 2 , p. 180198
( in russian ; transl . in english
: theor .
phys . , 1980 , vol .
45 , p. 963 ) .
gaida r. p. , kluchkovsky yu .
b. , and tretyak v. i. three - dimensional lagrangian approach to the classical relativistic dynamics of directly interacting particles .
in : constraint s theory and relativistic dynamics
. g. longhi , l. lusanna , editors .
world scientific publ . , 1987 , p. 210241 .
gaida r. p. and tretyak v. i. single - time form of the fokker - type relativistic dynamics //
acta phys . pol .
b , 1980 , vol .
11 , no 7 , p. 502522 .
gaida r. p. and tretyak v. i. direct interaction lagrangians and hamiltonian description of a particle system in various forms of relativistic dynamics./ preprint of the institute for theor.phys .
, itp8287p , kiev 1982 , 38 p. ( in russian ) . pryce m. h. l. the mass - centre in the restricted theory of relativity and its connection with the quantum theory of elementary particles // proc .
london , 1948 , vol .
a195 , no 1040 , p. 6281 . fleming g. n. //
covariant position operators , spin , and locality phys .
, 1965 , vol .
b137 , p. 188197
bakamjian b. and thomas l. h. relativistic particle dynamics .
ii // phys .
, 1953 , vol .
92 , no 5 , p. 13001310 .
sokolov s. n. coordinates in relativistic hamiltonian mechanics // teor
. mat .
fiz . , 1985 , vol .
62 , no 2 , p. 210221
( in russian ) .
sokolov s. n. relativistic addition of direct interactions in point form of dynamics // teor
. mat .
, 1978 , vol .
36 , no 2 , p. 193207
( in russian ) .
duviryak a. a. and kluchkovsky yu .
b. space - time interpretation of relativistic hamiltonian mechanics of particle system // ukr .
fiz.zh . , 1992 , vol .
37 , no 2 , p. 313320 ( in ukrainian ) .
gaida r. p. , tretyak v. i. , and yaremko yu .
g. center - of - mass variables in the relativistic lagrangian dynamics of a system of particles // teor . mat .
1994 , vol . 101 , no 3 , p. 402416 ( in russian ; transl . in english
: theor . math .
phys . , 1994 , vol . 101 , no 3 , p. 14331453 ) .
jaremko yu .
g. relativistic lagrangian dynamics of @xmath0-particle system in terms of centre - of - mass coordinates./ preprint of the institute of applied problems of mechanics and mathematics , lviv 1988 , no 14 - 88 , 34 p. ( in russian ) .
yaremko yu .
g. quasi - relativistic two - body problem in lagrangian centre - of - mass variables./ preprint of the institute for theor.phys . , itp9063u , kiev 1990 , 33 p.
( in ukrainian ) .
gaida r. p. , tretyak v. i. , and yaremko yu .
g. center - of - mass variables in the relativistic lagrangian mechanics of particle system // ukr .
, 1991 , vol .
36 , no 12 , p. 18071818
( in ukrainian ) .
weyssenhoff j. , raabe a. relativistic dynamics of spin - fluids and spin - particles //
acta phys . pol .
, 1947 , vol .
9 , no 1 , p. 718 .
damour t. and schafer g. redefinition of position variables and the reduction of higher - order lagrangians // j. math .
phys . , 1991 , vol . 32 , no 1 , p. 127 - 134 .
llosa j. , vives j. hamiltonian formalism for nonlocal lagrangians // j. math .
phys . , 1994 , vol .
35 , no 6 , p. 28562877 .
woodcock h. w. , havas p. approximately relativistic lagrangians for classical interacting point particles // phys .
d , 1972 , vol . 6 , no 12 , p. 34223444 .
baaski s. lagrange function for the motion of charged particles in general relativity theory //
acta phys . pol .
, 1957 , vol .
16 , no 6 , p. 423433 .
kluchkovsky yu .
b. approximate lorentz - invariance and ppn - formalism in the theory of gravitation // math . methods and phys .- mech .
fields , no 16 , kiev , nauk . dumka , 1982 , p. 107110
( in russian ) . whitrow g. j. , morduch g. e. relativistic theories of gravitation . in : vistas in astronomy , vol
london , pergamon press , 1965 , p. 167 .
messia a. quantum mechanics .
moscow , nauka , 1979 ( in russian ) .
berestetskii v. b. , lifshitz e. m. , and pitaevskii l. p. quantum electrodynamics .
moscow , nauka , 1989 ( in russian ) . | collective center - of - mass variables are introduced in the lagrangian formalism of the relativistic classical mechanics of directly interacting particles .
it is shown that the transition to the hamiltonian formalism leads to the bakamjian - thomas model .
the quantum - mechanical system consisting of two spinless particles is investigated .
quasi - relativistic corrections to the discrete energy spectrum are calculated for some coulomb - like interactions having field theoretical analogues .
quasi - relativistic mechanics , center - of - mass variables , schrdinger equation 03.20.+i |
successive merging and accretion of many comparatively small stellar systems like dwarf galaxies , as well as intergalactic gas , is an important part of the theoretical framework in which large ( disk ) galaxies are predicted to have formed ( e.g. , white & rees 1978 ; steinmetz & muller 1995 ) .
significantly , studies of resolved stars in dwarf companions ultimately allow one to explore two paired issues , namely giant galaxy assembly as well as the star formation history of dwarfs , which describes the gradual conversion of gas into stars , followed by the expulsion of remaining gas by stellar winds and the resultant heating and cooling of the interstellar medium .
since baade ( 1944a , 1944b ) first succeeded in resolving the brightest stars of several dwarf elliptical ( de ) satellites of the giant spiral galaxy andromeda , those galaxies have been regarded as simple , old stellar systems , with a stellar content resembling that of galactic globular clusters ( population ii ) . from the mid-1990s onward
however the impressive improvement in the quality of colour - magnitude diagrams ( cmds ) for nearby de galaxies provided by the hst has revealed that those simple " galaxies often display varied , and in some cases , complex star formation histories ( sfhs ) ( da costa 1998 ; mateo 1998 ) .
however , it remains unclear whether the andromeda de companions ( ngc 205 , ngc 147 and ngc 185 ) have had significant star formation episodes since the primeval event ( see da costa 1998 for a recent review ) .
the existence of a dozen luminous blue stars ( baade 1944a ) and other population i features ( dust clouds , hi gas , sn remnant ) at the centre of ngc 185 has been an intriguing feature of this galaxy for several decades because such galaxy building blocks had been alledged to have been depleted at earlier formation epochs in elliptical galaxies .
this recent star formation epoch is confined to its central 150 @xmath0 90pc , where the youngest , 100myr old , population is found ( hodge 1963 ; lee , freedman & madore 1993 ; martinez - delgado , aparicio & gallart 1999 , hereafter md99 ) . in ngc 147
however , there are no signs of recent star formation ( in the past 1gyr ) or gas at its centre ( han et al .
1997 ; young & lo 1997 ; sage , welch & mitchell 1998 ) .
thus , this difference between ngc 185 and ngc 147 is intriguing because otherwise they are similar in many of their properties ( e.g. type , mass , size ) .
similarly , ngc 205 has been regarded as an unusual de galaxy it has gas ( johnson & gottesman 1983 ; young & lo 1997 ) and a young central stellar population ( baade 1944b ; lee 1996 ) , and many of the brightest objects have recently been resolved into groups of uv - bright stars ( cappellari et al .
so , why do the recent ( @xmath12 1gyr ) sfhs of andromeda s de companions differ ?
a natural way to probe this issue is to search for an intermediate - age ( 1 - 10 gyr ) population in these galaxies .
the first evidence of such a population in these galaxies was the presence of stars above the tip of the red giant branch ( trgb ) in cmds from ground - based observations of their inner regions ( ngc 205 : lee 1996 ; ngc 185 : lee et al .
1993 , martinez - delgado & aparicio 1998 , hereafter mda98 ; and ngc 147 : davidge 1994 ) and have been interpreted as intermediate - age agb stars by several authors ( davidge 1994 ; lee et al .
1993 ; lee 1996 ) .
however , ground - based observations can lead to severe central crowding , making rgb stars appear brighter than they are , mimicking an intermediate - age agb population ( martinez - delgado & aparicio 1997 ) .
two other paired issues are the origin and possible stripping of the gas from which young stars are born in such galaxies .
one could view the concentration of young stars lying in the central part of ngc 185 and ngc 205 as a hint that new stars could originate from material ejected by dying stars and that this process would only be efficient enough at the center of the galaxies .
this would suggest an internal origin for the gas in these systems , as was argued by faber & gallagher ( 1976 ) .
however , young & lo ( 1997 ) find different scenarios for the three andromeda de satellites : no gas in ngc 147 ( m@xmath13/l@xmath14m@xmath9/l@xmath9)/l@xmath15 data is taken from from grebel , gallagher & harbeck ( 2003 ; their table 1 ) ] ; gas with a different angular momentum per unit mass to the stars in ngc 205 ( m@xmath13/l@xmath16m@xmath9/l@xmath9 ) ; and gas kinematically compatible with that of the stellar component in ngc 185 ( m@xmath13/l@xmath17m@xmath9/l@xmath9 ) .
although several theories ( e.g. external origin : knapp et al . 1985 ; ram - pressure stripping : lin & faber 1983 ; or different initial conditions
( e.g. , gas number density and gas mass ) : hensler & burkert 1989 ) have been proposed to explain this puzzling situation in dwarf galaxies , the exact origin of the current inner region differences remains a mystery . in the present paper ,
we draw on a new colour magnitude diagram based analysis of resolved stellar populations in ngc 205 and ngc 185 to take a new look at their ages and abundances .
we re - analyze ngc 147 data for comparative purposes .
the layout of the present paper is as follows .
the data and data reduction are described in sec .
[ obs_datared ] .
extinction is assessed briefly in sec .
[ sec_extinction ] and an overview of the stellar content based on an inspection of our colour magnitude diagrams is given in sec .
[ cmd_description ] .
we infer the distance modulus and [ fe / h ] in sec .
[ sec_dist_metal ] .
abundance distributions are probed in sec .
[ secabundance_distrib ] , and we make a comparison with previous studies and a control field in sec .
[ tails ] . in sec .
[ redclump ] , we briefly discuss the possible nature of the red clump / bump feature in ngc 185 . results concerning young - to - intermediate age stars are reported in sec .
[ youngstars ] .
in addition , as ngc 205 offers an opportunity to study a bona fide nucleated dwarf satellite in our neighbourhood we report several newly studied properties of its central compact star cluster and comment briefly on ngc 205 in the context of tidal stripping ( sec . [ nc_sbfitting ] )
finally , we discuss briefly the nature of ngc 205 in sec . [ stripping ] and
summarize our main conclusions in sec .
[ conclusions ] .
in the present study of ngc 185 and ngc 205 ( and ngc 147 ) , calibrated science- and data - quality images have been retrieved from the space telescope science institute hst data archive .
i ( f814w)- and v ( f555w)-band frames from selected hst / wfpc2 pointings has been examined . for our study of the nuclear star cluster in ngc 205
we retrieved v(f555w)- and r ( f675w)-band frames .
source information is given in table [ obs_log ] ; and relative positions of wfpc2 footprints can be seen in fig .
[ chart_ngc185 ] . for the data reduction , the following strategy has been applied .
photometry of the stars was determined using the hstphot ( dolphin 2000a ) point spread function ( psf)-fitting photometry package which is designed for optimal reduction and analysis of wfpc2 data .
stellar magnitudes are calibrated by hstphot using the charge transfer efficiency and zero point magnitude corrections derived by dolphin ( 2000b ) and are transformed into the v- and i - band passes .
lastly , artificial star tests identical to those described in butler , martinez - delgado & brandner ( 2004 ) have been performed using hstphot .
[ figcompltnss1 ] plots the completeness function against magnitude for each pointing : the completeness functions are the number of artificial stars recovered divided by the number of input stars .
a completeness plateau generally occurs for the brightest stars because their signal - to - noise is sufficiently high and crowding affects these stars equally .
a completeness of 100% is typically not reached because of bad pixels . in those frames containing one or more prominent star clusters ,
an over - detection of bright stars at bright magnitudes occurs due to the so - called bin - shifting effect
some faint artificial stars have blended with stars from the star cluster(s ) , and have been detected as brighter stars . for the final photometry list used in this study , objects have been selected if flagged as valid stars for both band - passes and have s / n @xmath5 5 ; @xmath18 @xmath12 7 ; -0.5 @xmath12 sharpness @xmath12 0.5 ; and @xmath19 @xmath12 0.25mag . the number of stars in each field after selection is given in table [ obs_log ] .
knowledge of the extinction towards stars is a basic necessity for determining their distances , masses , chemical compositions and ages . for an upper limit on the extinction from m31 s dust clouds in front of ngc 205 , plus foreground extinction due to the galaxy
, we recall that extinction in several outer disk fields in m31 has been estimated to be of the order of e(b - v)= 0.13mag ( e.g. , hodge & lee 1988 ; bellazzini et al .
2003 / bcf03 ) with a dispersion in their values of about 25% . for an estimate of the foreground extinction alone , towards m31/ngc 205 , we adopt e(b - v ) = 0.086mag from the schlegel , finkbeiner & davis ( 1998 ; sf98 ) dust maps .
as ngc 205 is behind m31 ( e.g. sec .
[ sec_dist_results ] ) with a projected location at the far outer edge of m31 , one might presume e(b - v ) @xmath12 0.13mag , i.e. , smaller than values measured using outer disk stars of m31 . as a reasonable compromise
, we adopt the average of extinction data for the m31 outer disk and the sfd98 value , namely e(b - v ) = 0.11mag , and assume the galactic reddening law with a@xmath15 = 3.1@xmath0e(b - v ) and a@xmath20/a@xmath15 = 0.48 ( cardelli , clayton & mathias 1989 ) .
a further issue is the internal extinction in ngc 205 .
about a dozen dust patches in the central 1 - 2@xmath21 have been catalogued , the largest of which points toward differential extinction values of up to about 0.4mag in v - band ( hodge 1973 ) .
those patches occupy about 5% of the f4 field , and their effect on results in the present paper is minor as determined by repeated tests in which photometry from 5% of each field was dimmed , conservatively , by 0.3mag the resulting systematic change in rgb / faint - agb colours is up to 0.02mag or 0.01mag in distance modulus with a negligible change in [ fe / h ] .
however , there remains the possibility of systematic errors with a similar order of magnitude due to a larger area being affected by a smaller level of extinction ( e.g. 0.05mag ) . for ngc 185 and
ngc 147 , foreground extinction estimates from sfd98 are e(b - v ) = 0.184mag and 0.173mag respectively
. however , ngc 185 and ngc 147 are located closer to the galactic plane ( b@xmath4 - 14.5@xmath22 ) than ngc 205 ( b@xmath4 - 21@xmath22 ) , and so , it is important to be aware of the possibility of significant inaccuracy in sfd98 extinction estimates towards low galactic latitudes that has yet to be established ( for @xmath23b@xmath23 @xmath12 10@xmath22 at least ; see sfd98 ) .
it is however rather encouraging that the above - mentioned extinction estimate from sfd98 for ngc 185 is in very good agreement with the result of a direct measurement of the stellar colour excess , i.e. e(b - v ) = [email protected] ( lee et al .
extinction data from sfd98 is given for each galaxy in table [ extinction1 ] .
figs . [ cmds1 ] and [ cmds2 ]
show the ( i , v - i ) cmd at each pointing in ngc 185 and ngc 205 respectively .
they display the brightest 5 - 6mag of the rgb / faint - agb sequence and give information on the stellar content therein . in ngc 185 , the fields in general look very similar .
the most prominent feature is the rgb - like structure , the upper two magnitudes of which is very similar to that found by mda98 in a low spatial resolution study of ngc 185 .
this structure is typical of a galaxy that clearly comprises a mixed stellar population : it is the locus of old and intermediate - age rgb stars , as well as faint agb stars .
there is a prominent red clump - like feature at a similar v - band magnitude to the red - end of the horizontal branch ( hb ) described below ; and we discuss the clump later in sec .
[ redclump ] . a significant number of bright red stars occur above the i - band discontinuity that marks the top of the rgb : they could be intermediate - age agb stars covering a moderate range of ages and metallicities , and are discussed further in sec .
[ youngstars ] .
there is clear evidence for blue and red horizontal branch ( hb ) stars at 0.2 @xmath2 v - i ( mag ) @xmath2 0.6 , characteristic of an ancient population of metal poor stars with [ fe / h ] @xmath2 -1.5dex .
this is the first direct evidence for such a population in this galaxy ( also see geisler et al .
there are stars surrounding the hb at v - i @xmath12 0.5mag , for example , that could comprise main sequence stars , blue loop stars as well as a foreground population of stars ; and those will be discussed further using isochrones in sec .
[ youngstars ] . in ngc 205 , somewhat like ngc 185 , there is a prominent sequence of rgb / faint - agb stars , similar to that presented by lee ( 1996 ) . that upper sequence does not appear to have a homogeneous colour distribution at a given i - band magnitude , which we detail in sec .
[ secabundance_distrib ] .
we fail to detect evidence for , or against , a hb in the central region due to the magnitude limit of the photometry .
however , there is an increased density of stars at about i = 24.5 - 25.5 , v - i @xmath20.6mag , the location where a hb might be expected in outer fields ( f2 and f1 ) , because crowding - noise is less there , but contamination from the stars in m31 s halo is more likely , as detailed later below .
there is also a scatter of stars redward of the rgb / faint - agb ( v - i @xmath8 3 ) ; they could be very metal - rich stars that belong to m31 and are briefly mentioned later in sec . [ tails ] .
lastly , the ngc 205 cmds contain several bright , blue stars ( at i@xmath223 , ( v - i)@xmath12 1 ) that are probably mostly main sequence and blue loop stars ; they are examined further in sec .
[ youngstars ] . for an analysis of
how much the ngc 205 cmds are contaminated by m31 stars , we used a b - band isopleth map of ngc 205 ( with an elliptically averaged surface brightness profile of m31 subtracted ; fig . 9 of ferguson et al .
2002 ) and minor axis surface brightness data for m31 from walterbos & kennicutt ( 1988 ; wk88 ) .
it follows that 16 - 40% of the brightest stars ( probably field stars , upper ms / blue loop stars and upper agb / rgb stars ) across the f1 field belong to m31 . adopting that level of contamination
, we would implicitly assume the same age and chemical composition for ngc 205 and the halo of m31 , and we therefore made a different check : applying our selection criteria , detailed in sec .
[ obs_datared ] , to photometry from bellazzini et al .
( 2003 ) for a nearby field distant from ngc 205 and lies between ngc 205 and m31 . ] ( g58 ) , we found 1193 stars at i @xmath12 22.5mag . scaling that number for ngc 205 , where the m31 halo is approximately 1.5mag arcsec@xmath24 fainter ,
the estimated contamination is in good agreement with the previous estimate of 16 - 40% ; and can be applied for agb / rgb stars at i @xmath12 22.5mag .
changing now to the f2 field , the contamination level should be below 1% , which is negligible , and even less for the remaining fields which are closer to the nucleus of ngc 205 .
additional evidence that contamination from m31 stars decreases towards ngc 205 comes from the blue plume or upper main sequence population in m31 : i.e. , compare the blue plume in fields g58 , g64 and g108 from bcf03 .
we note that about half a dozen bright , blue stars occur at v - i @xmath12 1mag and 20@xmath12 i ( mag ) @xmath12 21 in each ngc 205 field
they may well be foreground stars , even though inspection of nearby ( control ) fields , g58 and g64 ( see fig .
5 & 6 in bcf03 ) , argues for a negligible surface density of foreground stars .
alternately , they might be members of ngc 205 , but that interpretation is disfavoured in sec .
[ youngstars ] . for ngc 185 and ngc 147 ,
the surface density of ( unobscured ) field stars ( and point - like field galaxies ) along the line of sight is taken to be 65@xmath68 and 70@xmath68 stars respectively per wfpc2 pointing at i@xmath2 22.2mag and dominated by stars redder than about r - i = 0.4mag ( battinelli & demers 2004a , 2004b ) .
their surface density is such that they are unlikely to have a significant effect on our distance / abundance measurements later .
as we consider the population of ms and blue - loop stars later , we note that some contamination by field stars at r - i @xmath12 0.4mag is possible , but the level should be negligible ( @xmath12@xmath12 few percent of the field surface densities given above ) on average based on an inspection of the cmds from the battinelli & demers wide - field study . the number of contaminating field galaxies was estimated to be 23@xmath65 per wfpc2 pointing at 20 @xmath12 i ( mag ) @xmath12 22 from the medium deep survey ( griffiths et al .
1994 ) , which is an unlikely source of significant contamination that was not assessed by the artificial star tests .
in the absence of star - to - star abundance information from spectroscopic measurements , popular ways to measure distances and to infer [ fe / h ] values for ancient stars in local group galaxies have included the rgb method , the horizontal branch ( hb ) v - band magnitude , rr lyrae stars , and the turn - off magnitude of main sequence stars .
we consider the first two of these .
such diagnostic tools have , necessarily , been calibrated using ancient stellar populations ( galactic globular clusters ; ggcs ) ; and are thus best applied to a single stellar population of similar ancient age ( e.g. 13gyr ) . clearly , our photometry does not sample such simple stellar populations but for ngc 185 and ngc 147 there is evidence that their stellar populations are mostly ancient , with a small intermediate - age component
see md99 ( ngc 185 ) and han et al .
( 1997 ; ngc 147 ) . for ngc 205 however we are less certain because of the possibility of a significant fraction of intermediate - age stars . on the reliability of the absolute magnitude - abundance relation for the hb , on which the hb method relies
, we recall the relation is controversial in the literature ( e.g. see caputo 1997 , and references therein ) due to the dependence of m@xmath25(rr lyrae ) on the composition and evolutionary phase of the variable star .
that leads to a possible range in [ fe / h ] of the order of 0.2dex for a given absolute v - band magnitude . despite this caveat
, there is an advantage to using both the hb magnitude method together with the rgb method , namely that in stellar systems dominated by ancient stars with similar abundance levels , one would expect an agreement , within uncertainties , between the two methods . based upon the hypothesis that the rgb is dominated by ancient stars with similar abundance levels to those of the hb stars
, we will make a comparison of the inferred abundance estimates . in the present paper , when we refer to stellar abundance ( or metallicity ) , we are referring to [ fe / h ] as inferred from the colour ( ( v - i)@xmath26 ) at a certain absolute i - band magnitude for ancient stellar populations . in this way , we are making an inference of the metallicity of ancient rgb stars , using an approach that can be applied to other stellar systems .
a key concern however is the so - called age - metallicity degeneracy in stellar populations comprising ancient and intermediate - age stars .
the trouble is that evolved ( i.e. , post - main sequence ) stars that are between about one and a few times 10@xmath27yr old evolve along the blue side of the rgb causing the inferred median ( or mean ) colour to be bluer than that of the ancient rgb stars .
we use the median colour as opposed to the mean value as the median is a more reliable tracer of the predominant colour when the colour distribution is skewed otherwise the median and mean values are identical . however , whichever of these two statistics is used , an inferred ( mean / median ) [ fe / h ] value will under - estimate the actual [ fe / h ] of ancient rgb stars , and so our photometric abundance is taken as a lower limit as is done implicitly in other studies that determine photometric abundances ( see grebel , gallagher & harbeck 2003 ) .
photometric abundances are useful because they are available for a wide range of systems , thereby allowing them to be compared .
for example , see grebel , gallagher & harbeck 2003 , their section 2.1 , for a useful comparison of nearby dwarf galaxy properties . for the rgb - tip indicator of distance and [ fe / h ] ,
one requires a knowledge of the tip i - band magnitude of the rgb , i@xmath28 , and its uncertainty . in the case of ngc 205 , we estimated i@xmath28 by applying a slightly modified version of the edge - detection method described by sakai et al .
i.e. , we evaluated the edge - detection response at fine intervals ( @xmath1210@xmath29mag ) followed by smoothing using a 0.05mag - wide moving - average window and recorded the peak response in a restricted magnitude range ( i@xmath1221mag ) .
that was repeated a few hundred times in a bootstrap way to measure the average tip magnitude and its uncertainty .
for the inner two fields combined , we found i@xmath28 = [email protected] where the uncertainty includes a conservative error to account for the probable slight bias introduced by the smoothing process .
for ngc 185 and ngc 147 , we could not accurately identify the rgb tip magnitude using the edge - detection approach using individual fields as there are too few stars , making the rgb tip poorly defined . using the ensemble data for ngc 185 , but restricting the edge detection to i @xmath12 20.5 , we find i@xmath28 = 20.4 @xmath60.1mag which is close to the value from nowotny et al .
( 2003 ; nk03 ) one obtains i@xmath28 = 20.31mag for ngc 185 from their data . however , as their detection criterion is much better , owing to significantly more stars , we adopt their value , which agrees with previous estimates in the literature ( lee et al .
1993 ; mda98 ; ngc 185 ) . for ngc 147
, we deduce that there are too few stars at i@xmath12 21mag to allow an accurate measurement using the edge detection method ; and so , we adopt i@xmath28 = 20.70mag taken from the nk03 incidentally , that i@xmath28 value is in good agreement with the estimate by han et al .
( 1997 ) that was based on visual inspection of the i - band luminosity , using the same wfpc2 dataset retrieved for the present paper .
the ( median ) colour of the rgb / faint - agb sequence at m@xmath30 could then be measured .
derived ( median ) ( v - i)@xmath31 data as well as the inferred [ fe / h ] and distance modulus values , using the rgb-[fe / h ] relation ( lee et al .
1993 ; see below ) , and the rgb magnitude and colour ranges detailed below are given in table [ distance_comp2 ] . for the v(hb )
method of [ fe / h ] and distance estimation , the relevant relations are ( a ) the m@xmath15[fe / h ] relation for the hb at the location of the rr lyrae star instability strip , i.e. , m@xmath15 = 0.17[fe / h ] + 0.82 ( lee et al . 1993 ) and ( b ) the empirical calibration relating the de - reddened rgb v - i colour at m@xmath32 = -3.5mag ( ( v - i)@xmath33 ) to [ fe / h ] i.e. , [ fe / h ] = -12.64 + 12.61 ( v - i)@xmath34 - 3.33 ( v - i)@xmath35 ( lee et al . 1993 ) ; the relation is valid for -2.2@xmath12[fe / h]@xmath12 - 0.7dex .
we recall that ( m - m)@xmath26 = i@xmath28 + bc@xmath20 - m@xmath36 - a@xmath20 , where bc@xmath20 is the bolometric correction to the i - band magnitude , and m@xmath36 is the bolometric magnitude of the rgb tip magnitude . following da costa & armandroff ( 1990 ) , we take bc@xmath20 = 0.881 - 0.243(v - i)@xmath28 , and m@xmath37 = -0.19[fe / h ] - 3.81 .
we obtained median rgb colours and uncertainty values using the previously mentioned bootstrap approach in the m@xmath20 range -3.3 to -3.7mag ( ( v - i)@xmath34 ) and -4.1 to -3.95mag ( ( v - i)@xmath28 ) .
similarly , we measured the hb v - band magnitude and its uncertainty for stars at 24.5 @xmath12 v(mag ) @xmath12 25.7 and 0.2 @xmath12 v - i ( mag ) @xmath12 0.7 .
we then solved for the distance modulus and [ fe / h ] in an iterative way that is approximately self - consistent because both relations used are calibrated against ancient simple stellar populations ( galactic globular clusters ) the systematic error arising from the [ fe / h ] matching tolerance is @xmath60.025dex . in order to estimate the ( random ) uncertainty in the derived [ fe / h ] value , the iterative procedure was re - run using v(hb ) + @xmath38 and again using v(hb ) - @xmath38
the final [ fe / h ] value , the derived distance modulus , the v(hb ) magnitude and their uncertainties are given in table [ distance_comp ] for ngc 185 and ngc 147 .
such measurements for the f1 field in ngc 205 are precluded by a poorly defined hb . comparing our distance ( and [ fe / h ] ) data from the two methods we applied ,
there is a good agreement for ngc 185 . as we adopted the trgb i - band magnitude data from nk03 , it is not surprising that the derived distance modulus is in excellent agreement with their value , 24.04mag , also determined using colour information .
our [ fe / h ] lower limit for ngc 185 , [email protected] , derived using the v(hb ) method ( table [ distance_comp ] ) , is consistent with that from mda98 , as well as their report of an outward radial decrease in [ fe / h ] ; but is insufficient to confirm it .
however , there is a marked discrepancy between the nk03 finding of -0.89dex for ngc 185 and our value that can be explained by a mis - match between the foreground extinction adopted by nk03 ( a@xmath15=0.48mag ) and our value ( a@xmath15=0.57mag , from sfd98 ) : that extinction difference causes their value to be systematically more metal - rich than our finding , by about 0.2dex . for ngc 147
the v(hb ) method provides a distance modulus estimate that is markedly bigger by 0.1mag than the value from the rgb method which relies on the trgb magnitude from nk03 and is significantly different to their high signal - to - noise estimate of 24.44mag based on the rgb tip magnitude method ; and could be hinting that the dominant population on the rgb / faint - agb is not ancient and / or is more metal - rich than the hb stars .
as expected , our [ fe / h ] data from the rgb method is in fine agreement with that of han et al .
( 1997 ) and agrees well with the finding of nk03 ( -1.11dex ) and mould , kristian & da costa ( 1983 ) , each of whom also used colour information .
our distance modulus calculation for ngc 205 leads to ( m - m)@[email protected] ( for e(b - v ) = 0.11mag ) which would place ngc 205 farther behind m31 than previously estimated in the literature , e.g. grebel et al .
( 2003 ) and references therein .
lastly , our inferred [ fe / h ] lower limit is [email protected] from the rgb method , which is markedly more metal - poor than the mould , kristian & da costa ( 1984 ) inference that was based on rgb colour information from seeing - limited data ( [ fe / h ] @xmath5 [email protected] ) .
just like the discrepancy we found for ngc 185 between our [ fe / h ] value nk03 result , mentioned above , a probable cause is a systematic under - estimation of foreground extinction by about ( a@xmath15 =) 0.1mag by mould et al .
( they assumed a@xmath15 = 0.23mag ) .
lastly , it is worth remembering that if future reddening maps with higher spatial resolution than that of the sfd98 maps ( 6.1arcmin ) find that sfd98 underestimates extinction at our wfpc2 pointings ( [email protected] ) , our inferred photometric abundance data should be shifted to lower abundances values ( e.g. , using the lee et al .
1993 [ fe / h]@xmath39rgb relation ) .
in the previous section , one of the relations we used to infer abundances is the often employed empirical rgb-[fe / h ] relation that is calibrated using ancient rgb stars , to infer a photometric abundance for the stars in each field . in the present section ,
we examine the distribution of abundances in each field , and in order to be able to check the effect of a wide age range in a homogeneous way , we use model rgb-[fe / h ] relations by employing a code to generate synthetic cmds . as one does not know the actual age distribution of the stars , we adopt a single , mid - range age ( 10gyr ) , as worthey et al .
( 2004 ) did , and an adequately wide abundance range 0.0001 @xmath40 z @xmath40 0.009 ( @xmath4 -2.3 to -0.30dex ) = log(z / z@xmath9 ) where z@xmath9 = 0.018 . ] spreading age uniformly in a wide range ( e.g. 5 - 15gyr ) causes the inferred abundance distribution to widen by about @xmath60.2dex , as expected ( see worthey 1994 ; his table 6 ) .
incidentally , that age dependence is why the abundances derived later using the synthetic 10gyr models ( see table [ abund_summary ] ) produce [ fe / h ] values that are systematically more metal - rich by about 0.1dex than the mean data derived in sec .
[ sec_dist_metal ] using ggc stars that are a few times 10@xmath27yr older ( see tables [ distance_comp2 ] and [ distance_comp ] ) .
the basic idea here is to assign an abundance value to measured giant star colours from a synthetic cmd ( 10gyr , with random z values from 0.0001 to 0.009 ) in which realistic photometric scatter is added giant stars were selected such that -3.9 @xmath12 m@xmath20 ( mag ) @xmath12 -3.3 and 0.7 @xmath12(v - i)@xmath26 ( mag ) @xmath12 4 . only distributions for two fields in each galaxy are plotted for the sake of clarity ( see fig . [ abundance_distrib ] ) , but for all fields we give in table [ abund_summary ] the number of stars per distribution as well as ( a ) the fraction of stars that are old and metal - poor and or young and metal - rich ( [ fe / h]@xmath12 -1.4dex ) and ( b ) the fraction that is metal - rich ( -1.4 @xmath12[fe / h]@xmath12 -0.2dex ) ; the abundance separation at -1.4dex is an ad hoc value roughly mid - way in the range -0.5 to -2.5dex . for a check of the dependence of the inferred metal - rich tail in ngc 205 on a significant error in extinction ( 20% , for instance ) , owing to its proximity to m31 , we made a visual inspection of the distribution of stars in the ( m@xmath41,v - k)@xmath26 cmd , shown in fig .
[ mk_v_k_cmd ] , where the k@xmath42-band photometry is from davidge ( 2003 ) .
that photometry came from observations under 0.7arcsec seeing at r @xmath12 1.33@xmath21 ( e.g. , see his fig .
[ mk_v_k_cmd ] ) ; k@xmath42-band completeness is above 75% at k@xmath42 @xmath2 18.2mag and @xmath43 @xmath12 0.35mag ) @xmath2 -6mag only , due to a substantial increase in k - band incompleteness at fainter magnitudes . ] .
accordingly , with the help of the wide dereddened colour baseline ( v - k)@xmath26 , it follows from visual inspection that even with the sizeable hypothetical extinction error , the metal - rich component of the rgb is probably at least as rich as -0.7dex .
another issue is whether there are radial changes in the shapes of inferred distributions , for which one can make use of the skewness parameter listed in table [ abund_summary ] .
there is no clear evidence for ngc 185 , but ngc 205 becomes systematically redder towards its inner regions . that finding is unlikely to be affected by strong differential extinction , which affects only a small fraction ( @xmath4 5% ) of the central field
; but there remains the possibility of a smaller levels of differential extinction ( e.g. 0.05mag ) over a wider area .
lastly however , as abundance levels could be higher at successive epochs of star formation there remains the possibility of a correlated spatial variation in stellar ages that might act together to give an rgb / faint - agb colour roughly independent of position .
on the issue of metal - poor stars , it is hard to make a firm statement regarding the accuracy of a metal - poor tail reaching to [ fe / h]@xmath12 -1.5dex without spectroscopic information and or evidence for a hb , simply because stellar ages are unknown . for ngc 185
however we know that at least some of the stars in the metal - poor tail should be ancient because of the presence of a hb reaching to colours as blue as v - i @xmath4 0.1 - 0.2mag .
another issue is the similarity of the distributions . in the case of ngc 205
, we determined whether the shape of the control field and ngc 205 distributions differ .
we did this by determining the mean skewness and its uncertainty based on the bootstrap approach described earlier in sec .
[ sec_dist_results ] the derived data is given in table [ abund_summary ] .
based on an application of the student s t - statistic to the bootstrap samples , we find that the skewness in the control field ( g58 ) differs significantly from that in any of the ngc 205 fields . as the skewness in ngc 205 is unlikely to be the result of significant contamination from m31 ( see sec .
[ cmd_description ] ) , we infer that the rgb / faint - agb sequence in ngc 205 is significantly redder than that in the control field ( g58 ) .
the rgb / faint - agb sequence in ngc 205 is also skewed to redder colours than the sequence in ngc 147 or ngc 185 .
an exact interpretation for the rgb / faint - agb sequence ngc 205 is out of reach without modeling using data with a much fainter magnitude limit , but the skewness data hints that a significant fraction of the rgb / faint - agb stars occur in a relatively narrow age and or abundance interval compared to the control field .
finally , we note that skewness varies markedly with radial position in ngc 205 , in the sense that the inner region stars are generally redder than the outer region stars ; and it is unlikely to be fully attributeable to contamination from m31 .
from the poorly defined red - end of the bright agb , we infer z@xmath4 0.1z@xmath9 at ( v - i)@xmath26 @xmath4 3.5mag , with higher abundances likely for redder colours based on synthetic cmds . for a comparison
, the work of gallagher , hunter & mould ( 1984 ) suggests that the abundances of the most metal - rich stars in ngc 185 may be less than 1/3 solar ( i.e. , z @xmath2 1/3z@xmath9 , = 0.006 ) based on the n@xmath44/s@xmath44 ratio for one planetary emission nebula near the centre of ngc 185 .
similarly , the mean oxygen abundance ( 12 + log ( o / h ) = 8.2 ) from several planetary nebulae ( pne ) ( richer & mc call 1995 ) is consistent with a system whose most metal - rich stars have about one third of the solar abundance .
consequently , there is rough agreement with our cmd - based finding . in ngc 205 ,
the reddest pair of agb stars reach to ( v - i)@[email protected] corresponding to z@xmath4 0.0065 , about one third of the solar value .
consequently , abundances in ngc 205 appear to reach to richer values than in ngc 185 , a fact that is supported by the richer mean oxygen abundance for ngc 205 ( 12 + log ( o / h ) = 8.6 ; richer & mc call 1995 ) .
further on the subject of metal - rich stars , there are some very red stars ( v - i @xmath83 ) below the i - band rgb tip magnitude in ngc 185 and ngc 205 . whether those stars belong to ngc 205 is unclear , but such stars occur in m31 fields .
for example , bcf03 found them in wfpc2 photometry of m31 halo fields and noted that such stars would be highly metal - rich , if they are members of m31 .
examination of such stars with high abundance levels is largely precluded by incompleteness and model uncertainties .
stellar evolution theory predicts that the luminosity of the rgb and agb bumps in the colour magnitude plane depends on stellar abundance and age .
such theory also predicts a red hb in intermediate - age to ancient stars . in the present section we consider the red bump / clump ( rc ) and the prominent hb in ngc 185 . for the mean rc magnitude in each band - pass , we fitted gaussian function to the rc a few hundred times in a bootstrap way ( see sec .
[ sec_dist_results ] ) at v - i = 1 - 1.6mag , binned at 0.05mag intervals . the mean v(rc ) , i(rc ) and associated standard errors are 25.28 @xmath60.02mag and 24.20 @xmath6 0.01mag respectively .
consequently , we find that there is no v - band offset between the clump and the hb that we determined at 0.2 @xmath12 v - i ( mag ) @xmath12 0.7 , within uncertainties the offset is 0.01 @xmath6 0.15mag .
it is unclear whether the rc actually comprises rgb / faint - agb stars , a red hb , or both .
however , we may examine the implications of the third option using the model @xmath3v@xmath45 diagram from alves & sarajedini ( 1999 ) ( their fig .
6 ) . consequently , based on visual inspection of their fig .
6 we infer that [ fe / h ] would be about -1.5dex at @xmath3v(bump- hb ) = 0 for ancient stars , with a higher abundance level for intermediate - age stars , regardless of whether the v - band magnitude of the red hb is the same for all ages or one considers the theoretical age - dependence adopted by alves & sarajedini ( 1999 ) .
fig [ ngc205_v_vicmd_isochr ] and fig [ ngc185_v_vicmd_isochr ] show the ( v , v - i ) cmds for ngc 205 and ngc 185 respectively , in which an arbitrary selection of stellar isochrones from bertelli et al .
( 1994 ) covering young and old ages are overlaid as a visual guide .
the presence of the brightest stars at v - i @xmath2 0.5mag in the ngc 205 cmd is curious as one might not suspect a foreground population because of the negligible surface density of such young blue stars in the nearby ( control ) fields g58 and g64 ( see fig . 5 & 6 of b03 ) .
on the other hand they occur in each field with no significant radial trend in the dwarf galaxy .
additionally , from a visual inspection there is an ( apparent ) paucity of blue stars , e.g. , v - [email protected] , at v@xmath4 21.5 - 23mag that is hard to explain by a conspiracy of incompleteness functions , and is inconsistent with a main sequence / blue loop population .
we take the opinion that the bright , blue stars are not the upper part of a main sequence / blue loop stellar population , but rather that they are probably foreground stars . for the fainter , blue stars at ( v @xmath2 23 ,
v - i @xmath2 0.5mag ) in ngc 205 , which are probably dominated by main sequence and blue loop stars , we made a visual comparison with several sets of isochrones , and deduce that star formation was probably still active about 1 - 3 @xmath010@xmath1yr ago .
that finding fits broadly with the discovery that many of the objects identified as bright stars by hodge ( 1973 ) are in fact clusters of bright uv sources , two of which have stellar ages of 0.5@xmath010@xmath1 and 10@xmath1yr respectively ( cappellari et al .
this confirms earlier reports of recent episodes of star formation in the nuclear region , based on iue observations with a 10@xmath46 @xmath0 20@xmath46 aperture ( wilcots et al .
furthermore , complementary information comes from near - ir studies of agb peak brightness and j - k colours of the blue agb sequence that find there are agb stars near the centre of ngc 205 that formed less than 10@xmath1yr ago ( davidge 2003 ; his fig 5 ) .
lastly , we note that with an orbital period of 10@xmath1yr ( cepa & beckman 1988 ) , it is possible that a past disk crossing triggered the recent star formation activity .
in the case of ngc 185 , there are about two dozen blue stars ( v - i @xmath21mag ) at v @xmath4 20.5 - 23mag in the ensemble cmd presented in fig [ ngc185_v_vicmd_isochr ] . a comparison with models ( see sec . [ cmd_description ] )
indicates that the bulk of them are probably foreground stars owing to ngc 185 s low galactic latitude , but some red supergiants from ngc 185 may be present as noted by nowothny et al .
( 2003 ) and martinez - delgado & aparicio ( 1998 ) .
another point against them as members of ngc 185 is that a significant number of fainter stars would be needed if the stars were actually ms and blue loop stars .
however , the presence of a significant population of faint blue stars at v@xmath824mag ( excluding the hb of ancient stars at v@xmath425 - 25.5mag ) argues for a real population of ms and blue loop stars .
guided by the isochrones in fig [ ngc185_v_vicmd_isochr ] , one infers that star formation was probably still active about 4@xmath010@xmath1yr ago .
we have not tried to derive the star formation history of the faint blue stars at v@xmath824mag due to significant incompleteness at such faint magnitudes . however , although we detect the faint population of blue stars at v@xmath8 24.5mag based on visual inspection of the ensemble ( v , v - i ) cmd , the model produced by md99 ( their fig .
16 ) , based on a synthetic cmd analysis , appears to be markedly fainter .
thus , the mean sfr density for the last 10@xmath27yr is probably above the md99 estimate , i.e. , @xmath5 2.6@xmath010@xmath47m@xmath9yr@xmath48 pc@xmath24 at r @xmath12 2@xmath42 .
next , we probe whether the mean age of young agb stars varies radially with respect to older stars . for that reason we list in table [ agb_rgbratios ]
the star counts and count ratios for bright , young agb stars ( i@xmath12 i@xmath28 ) and stars on the upper rgb / faint - agb sequence , defined in colour magnitude space by boxes a and b respectively in fig .
[ cmds1 ] for ngc 185 ; we shifted the boxes appropriately for ngc 205 and ngc 147 with regard to distance and extinction differences .
for ngc 205 , it follows that a real increase occurs in the mean ages of bright , young agb stars with respect to older stars towards the nuclear region , even if a conservative contamination level of 40% in the f1 field is included .
simulating conservative systematic errors of 50% in e(v - i ) and / or 0.1mag , for example , in the distance modulus causes a shift of a few percent in bright , young agb star counts , but the conclusion remains unchanged . in the ngc 185 ( and ngc 147 ) data , we find no statistically significant field - to - field changes in the mean ages of bright , young agb stars with respect to older stars .
due to a relatively short crossing time in and between the wfpc2 fields , the agb and older stars could be well - mixed ( e.g. conservatively , take an rms velocity of 1kms@xmath48 ) , a case that would be compatible with the apparent absence of field - to - field variations .
we can make a stab at answering the question using table [ agb_rgbratios ] which lists the number of agb stars that are relatively young ( i.e. , are above the rgb i - band tip magnitude in ( i , v - i ) cmds ) , normalised by the number of fainter ( and mostly older ) stars , selected from suitable colour magnitude boxes ( see fig . [ cmds1 ] ) .
in fact , we are using ( i ) the fact that the brightness of a young agb star of a given chemical composition is a function of its age , meaning that brighter agb stars are younger and ( ii ) and that the population of ancient stars , which is probably significant , is fainter than the rgb i - band tip magnitude .
we assume that there is a negligible contamination from ( a ) rgb stars near the rgb tip magnitude that might arise due to photometric scatter and ( b ) the number of mira variable stars occuring near the tip of the rgb that are in their bright state . as the fractional contributions of faint - agb and rgb stars
is unknown , we consider the available low time resolution ( wide - age bins ) information offered by our star count data .
consequently , we infer that the contribution of stars that formed in the past several 10@xmath49yr is significantly greater in ngc 205 than in ngc 185 or ngc 147 . for ngc 147
the contribution is significantly greater than that in ngc 185 .
ngc 205 offers an opportunity to study a bona fide nucleated dwarf satellite in our neighbourhood .
we perform a detailed assessment of a number of the nuclear cluster ( nc ) properties , namely colour , size and intensity . for a determination of an nc s structural properties
there are a number of useful studies to refer to .
for example , see bker et al .
2004 and references therein ( also see walcher et al . 2005 ) .
we fitted a separate analytical profile to the nc and to the underlying galaxy surface brightness profile that we model as a srsic profile ; such a profile is regarded as plausible based on galaxy morphology studies and has the form i@xmath50 ( r ) = i@xmath51@xmath50 exp[(r / r@xmath52)@xmath53 , where a srsic index n = 1 represents an exponential profile and n = 4 is a de vaucouleurs law .
as the radial coverage of the central wfpc2 field makes an accurate fit to the galaxy s profile hard , we fitted the srsic profile to the r - band surface brightness profile from peletier ( 1993 ) by trial and error at r = 20@xmath46 to 250@xmath46 .
for the v - band data , that peletier profile was simply shifted to match the v - band wfpc2/pc1 data at r = 9@xmath46 . at such distances from the nc ,
the light contribution from the nc is negligible , as is the effect of the hst wfpc2 psf on the fitting . to model the intrinsic cluster profile
, we used the nuker - law profile ( lauer et al .
1995 ) that is implemented in galfit ( peng et al .
the ` best ' model parameters are those that minimize the deviation from the data in a @xmath18 sense , and is determined by minimizing residuals between the model and the original image .
we note that the nuker model is a fine choice for the present study as it provides the best fit to the cluster data in a reduced-@xmath18 sense , so that reliable magnitudes and colours may be derived .
we used galfit in a circular aperture ( r @xmath12 4@xmath46 ) , centred on the nc , that is dominated by emission from the nc ; the underlying galaxy profile had been subtracted beforehand .
significantly , the galfit software takes the input instrumental psf into account by convolving the analytical cluster profile with it .
we made a psf for each band - pass using isolated stars , modeled by a moffat function with power index of 2.5 .
a technical point is that the width of a wfpc2 stellar psf for a given filter depends on the spectral type(s ) of the reference stars ; however , as the cluster is well - resolved a mismatch between the spectral energy distribution of the psf and the nc has a negligible effect on the derived intrinsic profile of the cluster .
circular symmetry was not assumed during cluster profile fitting which resulted in a minor - to - major axis ratio of about 0.92 , similar to q=0.90 @xmath60.05 from capellari et al .
( 1999 ) , q = 0.89 measured by heath jones et al .
( 1996 ) using pc1/f555w data , all based on data of comparable angular resolution that resolve the nc well . the actual r - band profile , the fitted and derived intrinsic profiles as well as the peletier ( 1993 ) profile are included in fig . [ ngc205_cntr_cluster ] .
in order to quantify the cluster size , we used the so - called effective radius , r@xmath54 , the radius of an area that contains half of the cluster light in projection ( see carlson & holtzman 2001 ) ; and deduced it for each band - pass from the derived intrinsic profile of the nc . to check how much the measured r@xmath54 ( and intensity )
depends upon small errors in the wings of the fitted nc profile , we masked 15% of the pc1 pixels , randomly selected , and fitted the nuker profile .
after repeating this several times , we determined the mean and standard deviation for each of the affected parameters ( m@xmath55 ; r@xmath56 ; logi@xmath56 ) ( see table [ ncproperties ] ) . taking the derived ( dereddened ) colour and size at face value
, we conclude that the nc s blue and red stellar components have similar compactness , as judged by the fwhm0.007@xmath46 ( r ) and [email protected]@xmath46 ( v ) . ]
parameter , but there is an excess of blue light with respect to red in the cluster s outer region the measured nc s ( apparent )
effective radius is smaller by a factor of 2.4 in r - band than in v - band .
some tentative supporting evidence for a blue excess comes from cappellari et al .
( 1999 ) who noted that the nc looks more elongated in the uv than at longer wavelengths ( e.g v - band ) . however , caution is needed regarding our interpretation because the possibility of significant differential extinction , on a length - scale similar to the cluster size , remains .
we find that like the ncs in late - type spirals , which have 10@xmath10 - 10@xmath1l@xmath57 ( deduced from bker et al .
2002 ) , and unlike ggcs , the nc in ngc 205 is quite bright , with 10@xmath58l@xmath59 ( 10@xmath60 ) where m@xmath11 = 4.31mag and m@xmath61 is given in table [ ncproperties ] .
lastly , it is worth bearing mind that the nc may be a multi - component population , as suggested by the ( apparent ) blue excess in its outer region , possibly a merged pair of globular clusters ( but also read the walcher et al .
( 2005 ) discourse on possible nc formation scenerios ) . in that case
the youngest / bluest component would probably be younger than the mean age derived from the integrated v - r colour ( see the next section ) , and could therefore be less than a few times 10@xmath1yr old .
clues regarding mean age can be drawn from the nc s colour .
mean age can also be estimated from the clusters luminosity and mass - to - light ratio ; however both of these require a mass estimate , whose significance would be uncertain .
the nc in ngc 205 is bluer than catalogued milky way globular clusters ( harris 1996 ) , whose bluest clusters reach to v - r @xmath4 0.4mag .
consequently , the nc s colour hints that the nc is not ancient , a supposition that is supported by spectra from da costa & mould ( 1988 ) , especially strong hydrogen lines , that are similar to those seen in the spectra of young globular clusters in the magellanic clouds . a further way to get a handle on the mean
nc age is through a comparison of the cluster colour with model predictions , again for a single burst of star formation .
based on models of an instantaneous burst of star formation that produces a 10@xmath10m@xmath9 stellar cluster ; an upper mass cutoff of 30 or 100m@xmath9 ; and z = 0.001 - 0.040dex ] , a ( dereddened ) v - r index above 0.3mag would only be expected for an approximately 0.5gyr - old cluster progressing to values as blue at ( v - r)@xmath26 @xmath4 - 0.1mag at younger ages for a wide abundance range , 0.04 @xmath40 z @xmath40 0.001 , ( leitherer et al . 1999 ) . consequently ,
taking the measured mean colour at face value and assuming a single epoch of star formation , one would argue for a young system up to a few times 10@xmath62yr old .
two key issues will be examined briefly here .
they are : ( a ) is it significant that the nc in ngc 205 is two - to - three times smaller than those in the virgo sample of geha , guhathakurta & van der marel ( 2002 ) ? , and ( b ) are typical galactic globular clusters akin to ncs ? on the first question ,
it is interesting that the nc in six bright de galaxies in the virgo cluster examined by geha et al .
have v - band effective radii of 8 - 13pc , larger than the nc in ngc 205 ( r@xmath63 @xmath4 4pc ) .
however , they sampled the bright end of the de luminosity function .
consequently , future studies extending to the ncs of fainter virgo des should aim to state whether or not the observed sizes are ubiquitous in the virgo cluster . on the link ,
if any , between nuclear star clusters and galactic globular clusters , which are non - central objects , it is of at least academic interest to make a brief comparison .
we note that the median r@xmath54 of ggcs is almost 3pc , with the bulk of them below 10pc ( see bker et al .
2004 ; fig .
so , although similar to some ggcs in terms of size , ggcs are ancient unlike the nc in ngc 205 which may be up to a few times 10@xmath1yr old as surmised in the previous sub - section .
consequently , the fact that the average size of the nc in ngc 205 ( @xmath4 3pc ) happens to be compatible with a massive ggc might be hinting that nc size is not strongly age dependent .
that ngc 205 may have experienced interactions with m31 has been suggested in the literature ( e.g. , velocity dispersion : bender , paquest & nieto 1991 ; twisted outer isophotes : choi , guhathakurta & johnston 2002 ) .
interactions can cause tidal stripping of stellar matter that could produce stellar sub - structure in so - called streams .
the most prominent sub - structure around m31 belongs to its giant stellar stream ( ibata et al .
2001 ; ferguson et al .
recent kinematical evidence plus modeling suggests that it probably does not include ngc 205 ( ibata et al .
2004 ) . our distance modulus measurement for ngc 205 , ( m - m)@xmath7 = [email protected] ( when e(b - v ) = 0.11mag ) , would support that result by placing ngc 205 at d @xmath4 890kpc , putting it farther behind m31 than previously estimated .
we presented the first detailed study of ngc 205 and ngc 185 using wfpc2 data .
our key results are : * there is an ancient stellar sub - population in ngc 185 with [ fe / h ] @xmath2 -1.5dex , based on the presence of a well - defined horizontal branch . *
we inferred the median [ fe / h ] of ancient stars ( @xmath510gyr ) in ngc 185 from colour information to be @xmath5 [email protected] , assuming e(b - v ) = 0.184mag . * in ngc 185 , we found no v - band offset between the red bump / clump and the hb v - band magnitude . from a comparison with theory ,
that finding suggests that ancient stars have [ fe / h ] @xmath4 -1.5dex , with a higher abundance level for intermediate - age stars ( 1 - 10gyr ) .
* we find that star formation was probably still active about 4@xmath010@xmath1yr ago .
* whether there is a horizontal branch in ngc 205 is still an open question as the photometry s limiting magnitude is insufficient to detect central hb stars if they are present .
* we determined that the rgb / faint - agb at each pointing in ngc 205 is significantly skewed to redder colours than that of our control field .
that is probably the result of a narrower range in stellar abundances and or age for a significant fraction of the stars in ngc 205 than in the control field located on the outskirts of m31 .
* we inferred the median [ fe / h ] of ancient stars in ngc 205 from colour information to be @xmath5 [email protected] dex , assuming e(b - v ) = 0.11mag . *
our analysis indicates that star formation activity in ngc 205 was probably still active less than 3@xmath010@xmath1yr ago .
such recent star formation activity could have been triggered by an interaction with m31 .
* we derived a new distance estimate using the rgb i - band tip magnitude method , obtaining ( m - m)@xmath26 = [email protected] , adopting e(b - v ) = 0.11mag .
* several properties of the nuclear star cluster in ngc 205 have been examined : in terms of size and intensity , it is like a galactic globular cluster or a nuclear cluster in a late - type spiral galaxy ; but from a comparison with models , the blue colour hints that its stellar population is young , up a few times 10@xmath1yr old . the ( apparent ) v-/r - band effective radii indicates an excess of blue light with respect to red light in the cluster s outer region . + like the ncs in late - type spirals , which have 10@xmath10 - 10@xmath1l@xmath57 ( deduced from bker et al .
2002 ) , and unlike ggcs , the nc in ngc 205 is quite bright , with 10@xmath58l@xmath64 . in the introductory section we raised key questions about the similarities and differences between the three m31 dwarf companions .
although an investigation of whether the galaxies had significant star formation episodes since the primeval occurrence is precluded by the low temporal resolution afforded by our cmds and also by the bright limiting magnitude of our cmds , we can compare the galaxies in other ways , namely via the abundances of certain sub - populations ; the most recent epoch of star formation ; and the contribution of intermediate - age stars to the total population .
ngcs 205 , 185 and 147 do not show marked differences in their inferred median photometric abundances ( from colour information ) for ancient stars , which correspond to intermediate or richer values ( [ fe / h ] @xmath5 -1.1dex ) .
however , both ngc 147 and ngc 185 have a sub - population of ancient stars with [ fe / h ] @xmath2 -1.5dex , based on the presence of a hb in each of them .
the difference between these [ fe / h ] estimates ( i.e. , from the rgb colour and hb methods ) could be explained if each contains an ancient and or intermediate - age population of predominantly metal - rich stars with [ fe / h ] @xmath8 - 1.5dex .
alternatively , the discrepancy may be resolved to some extent if extinction has been under - estimated .
for example , based on the rgb colour[fe / h ] relation from lee et al .
( 1993 ) , an under - estimation of our ngc 185 e(b - v ) data by 0.04mag , for example , leads to a more metal - rich abundance ( [ fe / h ] ) by about 0.2dex .
accordingly , we conclude that the ancient stellar populations of the three galaxies appear to have broadly similar [ fe / h ] values ; but these need to be scrutinized by detailed future star formation history analyses and spectroscopy - based studies . on the issue of the most recent epoch of star formation activity ,
ngc 185 and ngc 147 differ markedly , in the sense that star formation occured more recently in ngc 185 ( @xmath4 4@xmath010@xmath1yr ago ) than in ngc 147 ( @xmath8 1gyr ago ; han et al .
this is a restatement of the known crisis in explaining the differences in the ngc 147/ngc 185 pair ( mateo 1998 ) .
for example , if gas origin in dwarf galaxies is largely internal ( via dying stars ) , and they are still capable of gas accumulation at the present epoch , there is the suggestion ( mateo 1998 ; his sec .
4 ) from the lack of gas and recent star formation activity that ngc 147 has avoided accumulating gas since the last star formation episode(s ) . for the intermediate - age population we probed whether their contribution varied significantly among the three galaxies .
we inferred from an agb rgb star count analysis that the contribution from stars that formed in the past several 10@xmath49yr is significantly greater in ngc 205 than in ngc 185 or ngc 147 .
for ngc 147 the contribution is significantly greater than that in ngc 185 .
as ngc 147 and ngc 185 are similar in terms of mass and size , an implication is that ngc 147 has had a markedly higher star formation rate at intermediate - age epochs , compared to that at older epochs , than in ngc 185 .
t. davidge and m. bellazzini are thanked for the photometry data they provided , as is f. ferraro for providing globular cluster rgb ridge - line data .
it is a pleasure to thank j. walcher for useful comments on an early draft .
d. gouliermis is thanked for a pertinent comment on young star clusters and the anonymous referee is warmly thanked for a careful reading of the manuscript and valuable comments .
djb devotes this work to the memory of his brother martin mary de burgo butler .
alves , d. r. , sarajedini , a. , 1999 , , 511 , 225 aparicio , a. , gallart , c. , bertelli , g. , 1997 , , 114 , 680 baade , w. , 1944a , , 100 , 147 baade , w. , 1944b , , 100 , 137 battinelli , p. & demers , s. , 2004a , , 418 , 33 battinelli , p. , & demers , s. , 2004b , , 417 , 479 bellazzini , cacciari c. , federici , l. , fusi pecci , f. , rich m. , 2003 , , 405 , 867 bender , r. , paquet , a. , nieto , j .- l . , 1991 . , 246 , 349 bertelli , g. , bressan , a. , chiosi , c. , fagotto , f. , nasi e. , 1994 , , 106 , 275 bker , t. , sarzi , m. , mclaughlin , d. e. , van der marel , r. p. , rix , h - w
et al . , 2004 , , 127 , 105 bker , t. , laine , s. , van der marel , r. p. , sarzi , m. , rix , h - w .
et al . , , 123 , 1389 butler , d. j. , mart ' inez - delgado , d. , brandner , w. , 2004 , , 127 , 1472 cappellari , m. , bertola , f. , burstein , d. , buson , l. m. , greggio , l. , renzini , a. , 1999 , , 515 , 17 caputo , f. , 1997 , , 284 , 994 carretta , e. , gratton , r. g. , 1997 , , 121 , 95 cepi , j. , beckman , j. e. , 1988 , , 200 , 21 choi , p. i. , guhathakurta , p. & johnston , k. v. , , 124 , 310 da costa , g. s. & mould , j. r. , 1988 , apj , 334 , 159 da costa , g. s. , armandroff , t. e. , 1990 , , 100,162 da costa , g. s. , 1998 , in aparicio , a. , herrero , a. & sanchez , f. , eds , stellar astrophysics for the local group , cambridge university press , cambridge , p351 davidge , t. j. , 1994 , , 108 , 2123 davidge , t. j. , 2003 , , 597 , 289 davies , r. d. , 1975 , , 170 , 45 dolphin , a. , 2000a , , 112 , 1383 dolphin , a. , 2000b , , 112 , 1397 faber , s. m. , gallagher , j. s. , 1976 , , 204 , 365 ferrarese , l. , mould , j. r. , kennicutt , r. c. , jr . ,
huchra , j. , ford , h. c. , et al . , 2000 , , 529 , 745 ferguson , a. m. n. , irwin , m. j. , ibata , r. a. , lewis , g. f. , tanvir , n. r. , et al .
, 2002 , , 124 , 1452 ferraro , f. , et al . , 1999 , , 118 , 1738 gallagher , j. s. iii , hunter , d. a. , mould , j. , 1984 , , 284 , 544 geha , m. , guhathakurta , p. , van der marel , r. p. , 2002
, aj , 124 , 3073 geisler , d. , armandroff , t. , da costa , g. , lee , m. g. , sarajedini , a. , 1999 , iaus , 192 , 231 grebel , e. , gallagher , j. s. , harbeck , d. , 2003 , , 125 , 1926 griffith , g. , ratnatunda , k. u. , neuschafer , l. w. , casertano , s. , i m , m. , et al .
, 1994 , , 437 , 67 harris , w. e. , 1996 , , 112 , 1487 han , m. , hoessel , j. g. , gallagher , j. s. iii , holtsman , j. , stetson , p. b. , 1997 , , 113 , 1001 jones , d. h. , mould , f. r. , watson , a. m. , grillmair , c. , gallagher , j. iii , et al .
, 1996 , , 496 , 742 hensler , g. , burkert , a. , 1990 , , 170 , 231 hodge , p. w. , 1963 , , 68 , 691 hodge , p. w. , 1973 , , 182 , 671 hodge , p. w. , lee , m. g. , 1988 , , 329 , 651 ibata , r. , lewis , g. f. , irwin , m. , totten , e. , quinn , t. , 2001 , , 551 , 294 ibata , r. , chapman , s. , ferguson , a. m. , irwin , m. , lewis , g. , et al .
, 2004 , , 351 , 117 johnson , d. w. , & gottesman , s. , 1983 , , 275 , 549 knapp , g. r. , turner , e. l. , cunniffe , p. e. , 1985 , , 90 , 3 lauer , t. r. , ajhar , e. a. , byun , y - i . , dressler , a. , faber , s. m. , et al . , 1995 , , 110 , 2622 lee , m. g. , freedman , w. l. , madore , b. f. , 1993 , , 106 , 964 lee , m. g. , 1996 , , 112 , 1438 leitherer , c. , schaerer , d. , goldader , j. d. , delgado , r. m. g. , robert , c. , et al . , 1999 , , 123 , 3 lin , d. n. c. , faber , s. m. , 1983 , , 266 , 21 mould , j. , kristian , j. , da costa , g. s. , 1984 , , 278 , 575 martinez - delgado , d. , aparicio , a. , 1997 , , 480 , 107 martinez - delgado , d. , aparicio , a. , 1998 , , 115 , 1462 martinez - delgado , d. , aparicio , a. , gallart , c. , 1999 , aj , 118 , 2229 mateo , m. , 1998 , , 36 , 435 mould , j. r. , kristian , j. , da costa , g. s. , 1990 , , 100,162 nowotny , w. , kerschbaum , f. , olofsson , h. , schwarz , h. e. , 2003 , , 403 , 93 peletier , r. f. , 1993 , , 271 , 51 peng , c. , ho , l. c. , impey , c. d. , rix , h - w .
, 2002 , , 124 , 266 richer , m. g. & mccall , m. l. , 1995 , 445 , 642 sage , l. j. , welch , g. a. , mitchell , g. f. , 1998 , 499 , 209 sakai , s. , madore , b. f. , freedman , w. l. , 1996 , 461 , 713 schlegel , d. j. , finkbeiner , d. p. & davis , m. , 1998 , , 500 , 525 steinmetz , m. & muller , e. , 1995 , , 276 , 549 walterbos , r. a. m , kennicutt , r. c. jr . , 1988 , , 198 , 61 walcher c. j. , van der marel r. , mclaughlin d. , rix h .- w . , bker t. , et al . , 2005 , , 618 , 237 white , s. w. & rees , m. , 1978 , , 183 , 341 wilcots e. m. , hodge , p. , eskridge , p. b. , bertola , f. , buson , l. , 1990 , 364 , , 87 worthey , guy , 1994 , , 95 , 107 worthey , g. , mateo , m. , alonso - garca , j. , espaa , a. , 2004 , , 116 , 295 young , l. m. , lo , k. y. , 1997 , 476 , 127 english , j. , freeman , k. c. , 1999 , , 118 , 752 lllccc ( 0 ) & ( 1 ) & ( 2 ) & ( 3 ) & ( 4 ) & ( 5 ) + ngc 185 & & & & & + f4 & 1.6 & v / i & 2/3 & 1300/2800 & 22857 + f3 & 2.1 & v / i & 2/3 & 1300/2800 & 22833 + f2 & 3.3 & v / i & 2/3 & 1300/2800 & 20747 + f1 & 4.1 & v / i & 2/3 & 1300/2800 & 16899 + ngc 205 & & & & + f4 & 0.3 & v / i & 2/3 & 1300/2700 & 8957 + f3 & 0.8 & v / i & 2/3 & 1300/2700 & 15718 + f2 & 2.6 & v / i & 2/3 & 1300/2700 & 24073 + f1 & 8.1 & v / i & 2/3 & 1300/2700 & 19053 + ngc 205 & 0.5 & v / r & 4/3 & 400/50 & - + + ( 0 ) galaxy name ; ( 1 ) angular separation of the field taken from mast , the multi - mission archive at stsci , from the galaxy s center to the coordinates listed in the data pointings table ( nominally the middle of the field of view ) ;
( 3 ) & ( 4 ) selected filters and frame integration time ; ( 5 ) number of stars detected ; ` - ' means that stellar photometry was not performed .
[ obs_log ] ccccc ( 0 ) & ( 1 ) & ( 2 ) & ( 3 ) & ( 4 ) ngc 185 & 0.570 & 0.273 & 120.79 & -14.48 ngc 205 & 0.34@xmath65 & 0.16@xmath65 & 120.72 & -21.14 ngc 147 & 0.535 & 0.257 & 119.82 & -14.25 + extinction values from schlegel , finkbeiner & davis ( 1998 ) ; @xmath66 : galactic coordinates are from the simbad database ; @xmath65 : ( see sec .
[ sec_extinction ] for further explanations ) .
[ extinction1 ] lcccc ( 0 ) & ( 1 ) & ( 2 ) & ( 3 ) & ( 4 ) 185 & 4.1 & [email protected] & 24.10 @xmath6 0.04 & -1.12 @xmath6 0.08 & 3.3 & [email protected] & 24.06 @xmath6 0.04 & -1.20 @xmath6 0.06 & 2.1 & [email protected] & 24.09 @xmath6 0.04 & -1.11@xmath6 0.05 & 1.6 & [email protected] & 24.06 @xmath6 0.03 & -1.00@xmath6 0.04 & mean@xmath67 & [email protected] & 24.08 @xmath6 0.02 & -1.11@xmath6 0.08 205 & 8.1 & [email protected] & 24.77 @xmath6 0.03 & -1.04@xmath6 0.03 & 2.6 & [email protected] & 24.76 @xmath6 0.03 & -1.05 @xmath6 0.03 & 0.8 & [email protected] & 24.76 @xmath6 0.03 & -1.03@xmath6 0.03 & 0.3 & [email protected] & 24.74 @xmath6 0.05 & -1.11 @xmath6 0.03 & mean@xmath67 & [email protected] & 24.76 @xmath6 0.01 & -1.06@xmath6 0.04147 & 3.0 & [email protected] & 24.47 @xmath60.03 & -1.12 @xmath6 0.04 & 0.2 & [email protected] & [email protected] & -1.14 @xmath6 0.08 & mean@xmath67 & [email protected] & [email protected] & -1.13 @xmath6 0.01 + columns : as for table [ distance_comp ] , except for ( 2 ) dereddened median colour of the rgb / faint agb sequence at the absolute i - band tip magnitude of the rgb ( 3 ) derived dereddened distance modulus ; @xmath67 mean ( and standard deviation ) of all values ; error values ( r ) are uncertainties in v - i@xmath31 , or the effect of the error in v - i@xmath68 ( [ fe / h ] ) , or the effect of the error in v - i@xmath31 and i@xmath28 ( ( m - m)@xmath26 ) .
[ distance_comp2 ] lcccc ( 0 ) & ( 1 ) & ( 2 ) & ( 3 ) & ( 4 ) 185 & 4.1 & [email protected] & 24.18 @xmath6 0.00 & -1.39 @xmath6 0.03 & 3.3 & [email protected] & 24.04 @xmath6 0.02 & -1.14 @xmath6 0.06 & 2.1 & [email protected] & 24.02 @xmath6 0.00 & -1.02@xmath6 0.02 & 1.6 & [email protected] & 24.11 @xmath6 0.02 & -1.04@xmath6 0.05 & mean@xmath67 & [email protected] & 24.09 @xmath6 0.07 & -1.15@xmath6 0.17 147 & 3.0 & [email protected] & [email protected] & -1.10 @xmath6 0.03 & 0.2 & - & 24.38@[email protected] & - 0.98@xmath69 @xmath6 0.02 & mean@xmath67 & - & 24.38@xmath6 0.01 & -1.04 @xmath6 0.08 + columns : ( 0 ) galaxy i d ; ( 1 ) angular separation of the field from the galaxy s centre ; ( 2 ) measured median horizontal branch v - band magnitude ; ( 3 ) derived dereddened distance modulus ; ( 4 ) inferred lower limit for the logarithm of the iron - to - hydrogen abundance relative to the solar value , for ancient rgb stars .
@xmath67 mean ( and standard deviation ) of all values ; error values ( r ) are uncertainties in v(hb ) , or indicate the effect of that error on ( m - m)@xmath26 or [ fe / h ] . the systematic error estimate for [ fe / h ] is 0.025dex
; it is half of the tolerance used to end the iterative search for [ fe / h ] and ( m - m)@xmath26 .
formally , the corresponding systematic error in ( m - m)@xmath26 , determined by applying m@xmath15 = 0.17[fe / h ] + 0.82 from lee et al .
( 1993 ) is 0.005mag . - : means the hb is poorly defined ; @xmath69 : adopted the v(hb ) value from the outer wfpc2 field in ngc 147.[distance_comp ] respectively , in order of increasing isochrone age .
the main sequence ( ms ) and blue - loop region ( bl ) of the isochrones are marked .
see sec .
[ youngstars ] for further explanations.,width=491,height=396 ] @xmath70 .
we determined the uncertainty in each median [ fe / h ] value using the bootstrap technique outlined in sec .
[ sec_dist_metal ] , which turns out to be negligible .
see sec .
[ secabundance_distrib ] for further explanations.,width=491,height=472 ] , ( v - k)@xmath7 and m@xmath20 , ( v - i)@xmath7 cmds for the inner - most pointing using ( m - m)@xmath26 = 24.76mag and e(v - i ) = 0.18mag .
stars brighter than the rgb tip magnitude in i - band ( dots / tiny squares ) are separated from fainter stars ( open squares ) for the purpose of presentation .
globular cluster fiducial ridgelines ( solid ) and stellar evolution tracks for the adopted z = 0.001 and 0.004 padova models ( dashed ) are overlaid .
globular cluster [ fe / h ] data is on the caretta & gratton ( 1997 ) scale and is listed in table 1 of ferraro et al .
the [ fe / h ] z relation used is [ fe / h]= 1.024logz + 1.739 , valid for 10 to 15.8gyr - old stars and -2 . @xmath12
[ fe / h ] ( dex ) @xmath12 -0.5 ( bertelli et al .
1994 ; their eqn .
see sec .
[ secabundance_distrib ] for further details.,width=340,height=604 ] lcccc 185 & & & & + a & 15@xmath64 & 24@xmath65 & 47@xmath67 & 47@xmath67 b & 1105@xmath633 & 1668@xmath641 & 2193@xmath647 & 2409@xmath649 a / b ( 185 ) & 0.013(4 ) & 0.015(3 ) & 0.021(4 ) & 0.020(3 ) 205 & & & & a & 72@xmath68 & 224@xmath615 & 272@xmath616 & 219@xmath615 b & 1706@xmath641 & 3554@xmath660 & 3122@xmath656 & 2034@xmath645 a / b ( 205 ) & 0.042(6 ) & 0.063(5 ) & 0.087(7 ) & 0.11(1 ) 147 & & & & + a & 69@xmath68 & 175@xmath613 & & b & 1890@xmath643 & 7122@xmath684 & & a / b ( 147 ) & 0.037(5 ) & 0.025(2 ) & & + a : brightyoung agb , b : upper rgb / faint - agb . the star counts ( n ) and ratios ( corrected for a mean incompleteness of about 5% at i @xmath2 23mag ) . star count error is @xmath6 @xmath70 .
the error in the last digit of each a / b value is given in parentheses .
box boundaries are defined in fig .
[ cmds1 ] .
[ agb_rgbratios ] lcccc ngc 185 & & & & @xmath3r ( @xmath21 ) & 4.1 & 3.3 & 2.1 & 1.6 no . of stars & 254 & 398 & 691 & 740 [ fe / h ] ( dex ) & -0.96 & -0.98 & -0.93 & -0.86 skewness & [email protected] & [email protected] & [email protected] & [email protected] f@xmath71 & [email protected] & [email protected] & [email protected] & [email protected] f@xmath72 & [email protected] & [email protected] & [email protected] & [email protected] ngc 205 & & & & + @xmath3r ( @xmath21 ) & 8.1 & 2.6 & 0.8 & 0.3 no . of stars & 386 & 950 & 814 & 585 [ fe / h ] ( dex ) & -0.96 & -0.91 & -0.89 & -0.88 skewness & [email protected] & -0.85@xmath6 0.01 & [email protected] & [email protected] f@xmath71 & [email protected] & [email protected] & 0.13@xmath6 0.02 & 0.14@xmath6 0.02 f@xmath72 & 0.84@xmath6 0.08 & 0.84@xmath6 0.06 & 0.87@xmath6 0.06 & 0.86@xmath6 0.07 ngc205/g58@xmath69 & & & & @xmath3r ( @xmath21 ) & 11.2@xmath73 & - & - & - no . of stars & 63 & - & - & - [ fe / h ] ( dex )
& -0.94 & & & skewness & [email protected] & & & f@xmath71 & [email protected] & & & f@xmath72 & 0.92@xmath6 0.24 & & & ngc 147 & & & & + @xmath3r ( @xmath21 ) & 3.0 & 0.2 & - & - no . of stars & 399 & 1410 & - & - [ fe / h ] ( dex ) & -0.91 & -0.84 & & skewness & [email protected] & [email protected] & & f@xmath71 & [email protected] & [email protected] & & f@xmath72 & 0.94@xmath6 0.1 & 0.93@xmath6 0.05 & & + ` - ' means no wfpc2 data available .
the [ fe / h ] data refer to median values .
f@xmath71 and f@xmath72 refer to the fraction of stars that are metal - poor or young ( [ fe / h]@xmath12 -1.4dex ) and the fraction that are metal - rich ( -1.4 @xmath12[fe / h]@xmath12 -0.2dex ) .
@xmath69 we applied our photometry selection criteria to photometry from bcf03 . @xmath73 angular separation from the centre of ngc 205 .
star count error is @xmath6@xmath70 . for each histogram
, skewness is defined here as 1/n(@xmath74((x@xmath75 - @xmath76)/@xmath77)@xmath78 , where x is the abundance data , and @xmath77 is its standard devation .
[ abund_summary ] lcccccccccc ( 1 ) & ( 2 ) & ( 3 ) & ( 4 ) & ( 5 ) & ( 6 ) & ( 7 ) & ( 8) & ( 9 ) & ( 10 ) & ( 11 ) 205 & 0.89 & 19.22 & 18.82 & 14.49 & 14.32 & -10.59 & -10.69 & 0.95/0.43 & 4.3/1.9 & 4.23/4.64 & ( 0.02 ) & & & ( 0.04 ) & ( 0.01 ) & ( 0.13 ) & ( 0.07 ) & ( 0.02)/(0.03 ) & ( 0.2)/(0.25 ) & ( 0.01)/(0.05 ) + notes : col .
( 2 ) : adopted galaxy distance ( see sec . [ sec_dist_metal ] ) .
( 3 ) and ( 4 ) : observed ( not inclination - corrected ) peak surface brightness of the galaxy disk underlying the nc . col . ( 5 ) and ( 6 ) : apparent v- and i - band magnitude of nc .
( 7 ) and ( 8) : absolute magnitudes of the nc corrected for foreground reddening .
uncertainties are given in parentheses and arise from fitting errors , a distance uncertainty of 0.05mag and an adopted 10% error in foreground extinction .
( 9 ) angular effective r - band radius and uncertainty as discussed in sec .
[ nc_sbfitting ] .
( 10 ) : effective r - band radius in parsecs .
( 11 ) : logarithm of effective r - band intensity of the nc , logi@xmath79 = 0.4(4.31 - m@xmath61 ) - log(2@xmath80r@xmath81 ) , and logi@xmath82 = 0.4(4.83 - m@xmath83 ) - log(2@xmath80r@xmath84 ) with r@xmath54 in parsecs .
[ ncproperties ] | we present the first detailed analysis of resolved stellar populations in the dwarf galaxies ngc 185 and ngc 205 based on archival v- and i - band wfpc2 pointings . for ngc 185
we deduce from the brightest main sequence and blue loop stars that star formation was probably still active about 4@xmath010@xmath1yr ago ; and have three key abundance - related results : ( 1 ) we identify ancient stars with [ fe / h ] @xmath2 -1.5dex by a well - defined horizontal branch ( hb ) .
( 2 ) we find a prominent clump / bump - like feature along the red giant branch / faint asymptotic giant branch ( rgb / faint agb ) with the same mean v - band magnitude as in the hb , within uncertainties ( i.e. , @xmath3v(bump - hb ) = 0 ) ; from a comparison with theory , the implication is that ancient stars have [ fe / h ] @xmath4 -1.5dex , with a higher abundance level for intermediate - age stars . ( 3 ) from colour information we infer that median [ fe / h ] @xmath5 [email protected] for ancient stars ( assuming e(b - v ) = 0.18mag ) . for ngc 205 , we record a new distance modulus , ( m - m)@xmath7 = [email protected] , taking e(b - v ) = 0.11mag , based on the red giant branch ( rgb ) tip magnitude method in i - band .
we find that stars were probably still forming less than 3@xmath010@xmath1yr ago in ngc 205 , which is compatible with star formation triggered by an interaction with m31 .
three key abundance - related results for ngc 205 are : ( 1 ) the red giant / faint - asymptotic giant branch is significantly skewed to redder values than that of a control field in the outskirts of m31 ; it probably results from a relatively narrow metallicity and or age range for a significant fraction of the dwarf s stars .
( 2 ) from a comparison with models , the most metal - rich rgb stars reach to [ fe / h ] @xmath8 -0.7dex ( @xmath80.2z@xmath9 ) .
( 3 ) for ancient stars we infer from colour information that median [ fe / h ] @xmath5 [email protected] ( for e(b - v ) = 0.11mag ) .
we briefly compare the stellar populations of ngc 205 , ngc 185 and ngc 147 .
finally , we study several v- and r - band structural properties of the nuclear star cluster in ngc 205 for the first time : the apparent v-/r - band effective radii indicate a blue excess in the cluster s outer region . in terms of size , the cluster is like a typical galactic globular cluster or a nuclear cluster in a nearby late - type spiral galaxy ; but it is quite bright ( 10@xmath10l@xmath11 ) , unlike an ancient globular cluster , and by matching with models , the blue colour hints that its stellar population is young , up to a few times 10@xmath1yr old . |
fuzzy models have been proposed , @xcite-@xcite , as a potential method of doing `` lattice '' field theory .
the basic idea is to take a classical phase space of finite volume , quantize it and thus obtain a space with a finite number of degrees of freedom .
things are of course a little more complicated but essentially this idea works when the phase space is a co - adjoint orbit , the simplest such example being the two sphere @xmath1 , with the resulting quantized space known as the fuzzy sphere @xcite . field theory models on the fuzzy sphere then possess only a finite number of modes .
the simplest field theory model of a scalar field with @xmath0 interaction was proposed in @xcite and has proved to be an excellent testing ground for the idea of using fuzzy spaces for doing ` lattice ' studies . a potential problem with using these fuzzy spaces to get a finite approximation to field theories and
thus do ` lattice ' physics has emerged due to the phenomenon of uv / ir mixing @xcite .
the problem was discussed in the moyal plane in @xcite where the fields possess an infinite number of modes , and consequently a regularization procedure is needed .
this phenomenon appears to be generic for field theories in non - commutative spaces . in a recent article vaidya
@xcite pointed out that this uv / ir mixing phenomenon is present in the @xmath0 model of @xcite on the fuzzy sphere .
his work was followed by that of chu , madore and steinacker,@xcite , who calculated explicitly the one - loop contribution to the two point vertex function and found that in the commutative limit , the non - planar diagram retains a residual finite contribution over and above the expected commutative term .
the additional term , can be seen as a nonlocal rotationally invariant contribution to the effective action . though this term is non - singular for the fuzzy sphere they showed that in the planar limit it incorporates the uv / ir mixing singularity of the moyal plane .
the implications of this result are very serious for the program of using the matrix model approximations to continuum field theories to study the non - perturbative continuum behaviour .
one implication is that the scalar action considered by these authors can not be the correct fuzzy action for the ` lattice ' program outlined above .
we therefore return to the problem and study how serious it in fact is and whether it has a natural solution .
we find that indeed it has a quite natural solution and that the problem disappears when the interaction term in the matrix action is `` normal ordered '' , i.e. when the appropriate subtractions associated with wick contractions of the @xmath0 term are included in the action .
the action ( [ normalorderedaction ] ) with normal ordered vertex is therefore the correct starting point for the fuzzy lattice physics program .
the paper is organized as follows : we begin by briefly reviewing the @xmath2-product , that realizes matrix multiplication at the level of functions on the fuzzy sphere .
we then review the one - loop calculation of the two - point function , repeating the calculation using the @xmath2-product .
next we discuss the 4-point vertex functions and demonstrate that the problem anomalous contributions in that commutative limit does not arise here .
we further show , in fact , that this problem is localized to the case of a propagator returning to the same vertex that it leaves , i.e. that it is associated with tadpole diagrams .
we finally propose our solution for the elimination of these unwanted contributions normal ordering the @xmath0 vertex .
the article concludes with a discussion and conclusions where we define the matrix model which has as limiting theory the standard continuum model .
at the level of the fundamental representation the fuzzy sphere @xcite can be defined as the orbit the adjoint action of @xmath3 on a rank one projection operator , @xmath4 , in a 2-dimensional hilbert space , @xcite @xcite , with @xmath5 , @xmath6 and @xmath7 . at a point on the sphere parameterized by the unit vector @xmath8 in @xmath9 , obtained by rotating the north pole by @xmath10
, we have @xmath11 here @xmath12 is the projector at the north pole , @xmath13 . at the level of the @xmath14-dimensional representation we construct a rank one projector @xmath15 as the @xmath16-fold symmetric tensor product of @xmath17 .
associated with every @xmath18 matrix is a function on the fuzzy sphere , defined by @xmath19 and an associative @xmath2-product between two such functions is given by @xmath20 in terms of derivatives @xmath21 with @xmath22 , where @xmath23 and @xmath24 so @xmath25 and @xmath26 ( indices @xmath27 are raised and lowered with the euclidean metric , @xmath28 , in @xmath9 ) .
the action of the @xmath3 generators on a function on the fuzzy sphere can be written as @xmath29\right),\ ] ] where @xmath30 are the generators in the @xmath14-dimensional matrix representation and @xmath31 are the generators in differential form .
the @xmath32 dimensional matrices can be expanded in terms of an ortho - normal basis of matrices @xmath33 normalized so that @xmath34 the image of these matrices under the map to functions ( [ maptofunctions ] ) we denote @xmath35 .
these functions are proportional to the usual spherical harmonics @xmath36 with @xmath37 and are polynomials in @xmath38 of order @xmath39 , @xcite . for a general function @xmath40 which is the image of the matrix @xmath41 the relation ( [ tnorm ] ) takes the form @xmath42 where the rotationally invariant operator @xmath43 has eigenvalues @xmath44 .
( [ f_lfreln ] ) defines the relation between functions which we denote with a subscript @xmath16 and those without this subscript .
the operator @xmath45 was introduced in @xcite , where an alternative expression for it can be found .
using the fact that @xmath46 the integral of the @xmath2-product in equation ( [ inttrace ] ) can be related to the integral over an ordinary commutative product by decomposing the functions @xmath47 and @xmath48 into harmonics , ( note they only contain @xmath49 ) , and rescaling each component by an @xmath39-dependent factor @xmath50 , then @xmath51 where the pairs @xmath52 , @xmath40 and @xmath53 , @xmath54 are related as in ( [ f_lfreln ] ) . the nave matrix action for a real scalar @xmath0 theory on @xmath55 at level @xmath16 is @xmath56&=&{4\pi\over ( l+1)}{\hbox{tr}}\left\{{1\over 2}[l_a,\phi][l_a,\phi]+ { 1\over 2}r\phi^2
+ { \lambda\over 4!}\phi^4\right\}\\ & = & \int_{s^2}d^3{\bf n}\left\ { { 1\over 2}\left({{\cal l}}_a \varphi_l * { { \cal l}}_a \varphi_l + r \varphi_l * \varphi_l \right ) + { \lambda\over 4!}(\varphi_l*\varphi_l*\varphi_l*\varphi_l)\right\ } , \nonumber\\\end{aligned}\ ] ] where @xmath57 and @xmath58 , so that @xmath59 is a real field .
the parameters @xmath60 and @xmath61 are real with @xmath61 positive .
the coherent state representation of the free - field propagator between two points @xmath38 and @xmath62 on the fuzzy sphere is @xmath63 it is symmetric under interchange of its arguments , @xmath64 because @xmath65 , and can be expressed in terms of legendre polynomials : @xmath66 in the continuum limit ( @xmath67 ) the massless free - field propagator reads ( omitting the @xmath68 term ) @xmath69 note that the propagator in ( [ fuzzyg ] ) is not defined with a @xmath2-star product , since the two points @xmath38 and @xmath62 are independent and in general distinct .
it was pointed out in @xcite and @xcite that there is a subtlety in the one - loop correction to the two - point function for this theory . for the commutative theory
the only infinities arise in the tadpole diagram and are accounted for by a mass renormalization , but the fuzzy model ( or fuzzy regularization ) introduces new momentum dependent terms into the effective action even at one - loop .
these are not just wave - function renormalizations but are new momentum dependent two - point interactions that do not disappear in the continuum @xmath67 limit .
these terms were interpreted in @xcite as being due to uv / ir mixing .
their presence however means that the fuzzy action ( [ fuzzys ] ) does not reduce to the usual @xmath0 theory in the continuum limit .
it is therefore unsatisfactory as a matrix approximation of the commutative model .
the problem arises because in non - commutative @xmath0 there are two different diagrams at one loop which contribute to the quadratic term in the effective action , a planar diagram and a non - planar diagram , see fig 1 .
using cyclic symmetry of the vertex it is easy to see that the 8 ways of connecting the legs in the planar diagram contribute @xmath70 $ ] to the effective action where @xmath71=\int d^2{{\bf n}}\sum_{l , m}\left\ { { \varphi_l({{\bf n } } ) * { \overline y}^l_{l , m}({{\bf n } } ) * y^l_{l , m}({{\bf n}})*\varphi_l({{\bf n}})\over l(l+1)+r}\right\},\ ] ] and @xmath72 . similarly the 4 ways of connecting the legs in the non - planar diagram contribute @xmath73 $ ] where @xmath74 = \int d^2{{\bf n}}\sum_{l , m } \left\{{\varphi_l({{\bf n } } ) * { \overline y}^l_{l , m}({{\bf n}})*\varphi_l({{\bf n } } ) * y^l_{l , m}({{\bf n } } ) \over l(l+1)+r}\right\},\ ] ] the planar diagram can be evaluated rather straightforwardly by noting that @xmath75 so that @xmath76 and the final @xmath2 can be eliminated by replacing @xmath77 with @xmath78 .
so we have @xmath71={\hbox{$\bigcirc\kern-3.6pt{\bf\bullet}$\kern 1.3pt}}(l , r)\int d^2{\bf n}{\kern 5pt } \varphi^2({\bf n}),\ ] ] where @xmath79 is a logarithmically divergent function in the limit of large @xmath16 .
the difference between the planar and non - planar diagrams is @xmath80:=i_p[\varphi]-i_{np}[\varphi]= \int d^2{{\bf n}}\sum_{l , m } \left\{{\varphi_l({{\bf n } } ) * { \overline y}^l_{l , m}({{\bf n } } ) * [ y^l_{l , m}({{\bf n}}),\varphi_l({{\bf n } } ) ] _ * \over l(l+1)+r}\right\},\ ] ] where @xmath81_*:=f_l*g_l - g_l*f_l$ ] . using the trace formula ( [ inttrace ] ) for the integral , and the fact that @xmath82 , cyclicity of the trace allows the sum and integral to be re - arranged as @xmath83={1\over 2 } \int d^2{{\bf n}}\sum_{l , m } \left\{{[\varphi_l({{\bf n}}),{\overline y}^l_{l , m}({{\bf n } } ) ] _ * * [ y^l_{l , m}({{\bf n}}),\varphi_l({{\bf n } } ) ] _ * \over l(l+1)+r}\right\}.\ ] ] the explicit expression for the @xmath2-product in ( [ star ] ) shows that the commutators start at order @xmath84 in the large @xmath16-limit ,
so one might expect that @xmath85 would be of order @xmath86 and hence vanish as @xmath67 .
but the problem is that the derivatives in the @xmath2-product introduce factors of momenta into the numerator , which are summed to @xmath16 , rendering the result of order one .
expanding @xmath59 in spherical harmonics , @xmath87 , shows that there is a momentum dependent contribution to the quadratic term in the effective action .
for example if we restrict ourselves to fields @xmath59 that are linear functions of @xmath38 , i.e. fields with angular momentum @xmath68 and @xmath88 we see from ( [ star ] ) that the derivatives involving @xmath59 terminate at @xmath89 , so we can write @xmath90 _ * = { 2i\over l}(\partial_a \varphi_l ) j^{ab } ( \partial_b{\overline y}^l_{l , m})= -{2\over l}(\partial_a\varphi_l)({{\cal l}}_a{\overline y}^l_{l , m}),\ ] ] and we get @xmath91={2\over l^2 } \int d^2{{\bf n}}\sum_{l , m } { \{(\partial_a\varphi_l ) ( { { \cal l}}_a{\overline y}^l_{l , m})\ } * \{({{\cal l}}_b y^l_{l , m})(\partial_b\varphi_l)\ } \over l(l+1)+r}.\ ] ] since @xmath92 are constants , in accordance with equations ( [ tint ] ) and ( [ tnorm ] ) the remaining @xmath2 in this expression can be replaced with an ordinary product provided we replace @xmath93 with @xmath94 . thus @xmath91=-{2\over l^2 } \int d^2{{\bf n}}\sum_{l , m } { ( \partial_a\varphi_l)({{\cal l}}_a{\overline y}_{l , m } ) ( { { \cal l}}_b y_{l , m } ) ( \partial_b\varphi_l ) \over l(l+1)+r}.\ ] ] the problem now reduces to the evaluation of @xmath95 its exact form can be determined by noting that @xmath96 so @xmath97 from which we deduce that @xmath98 finally , noting that @xmath92 only involves @xmath88 we can write @xmath99 , with @xmath100 expanded in terms of the ordinary @xmath101 so that we obtain @xmath91=\int d^2{{{\bf n}}}{{\cal l}}_a\varphi { 1\over l(l+2)}\sum_{l=1}^l \left({l(l+1)(2l+1)\over l(l+1)+r}\right){{\cal l}}_a\varphi,\ ] ] in particular we have @xmath102={2\over l(l+2)}\sum_{l=1}^l \left({l(l+1)(2l+1)\over l(l+1)+r}\right),\ ] ] which is clearly finite as @xmath67 .
the coefficient @xmath103 was derived previously using other methods in @xcite . for higher @xmath39
the @xmath2-products in the commutators in ( [ deltaicom ] ) involve terms up to @xmath104 which look as though they should be of order @xmath105 , but again the derivatives in the numerator conspire to give a finite answer .
the full expression can be obtained by noting that the linear operator @xmath106 is rotationally invariant and has eigenvectors @xmath107 and @xmath108 independent eigenvalues @xmath109 given by @xmath110\ ] ] using expressions from varshalovich et al @xcite ( equations ( 24 ) on page 46 and equation ( 12 ) page 236 ) we obtain an intermediate expression in terms of wigner 6j - symbols which can be summed to give @xmath111 using an explicit expression for the 6j - symbols ( equation 6 , page 294 of @xcite ) allows us to express the eigenvalues in terms of the eigenvalues of @xmath112 so that we find @xmath113 we can therefore express operator @xmath114 in the form @xmath115 where @xmath116 is defined by @xmath117 and @xmath118 .
finally we find the contribution to the effective action of the difference of planar and non - planar diagrams can be written as @xmath119=\int d^2 { \bf
n } \varphi_l({\bf n}){\cal q}_l({{\cal l}}^2)\varphi_l({\bf n}),\ ] ] where the operator @xmath120 is given by @xmath121 the eigenvalues of this operator agree with those found in @xcite .
this operator itself can clearly be expressed as the trace of a function of @xmath122 . it is argued in @xcite that the one - loop contribution to the two - point function for external momentum @xmath39 is given , to a very good approximation , by @xmath123=2\sum_{k=1}^l{1\over k}+o(1/l).\ ] ] we can recover this estimate by noting that @xmath124 are eigenfunctions of our operator @xmath125 and using the expansions of @xcite .
in this section we argue that the problem described in the last section is specific to the 2-point function and does not affect the 4-point function at one loop .
the one - loop four - point function has two vertices and in the non - commutative case these can be either planar or non - planar , giving rise to four distinct contributions to the quartic term in the effective action .
the four distinct diagrams are given in figures 2 and 3 .
stripping off irrelevant factors these diagrams are : @xmath126:&=&\int d^2{{\bf n}}'d^2{{\bf
\sum_{l',m'}\sum_{l'',m''}\left\ { { \bigl(\varphi_l*{\overline y}^l_{l',m'}*y^l_{l'',m''}*\varphi_l\bigr)_{{{\bf n } } ' } \over l'(l'+1)+r}\right.\nonumber\\ & & \kern 70pt\left.\times{\bigl(\varphi_l*{\overline y}^l_{l'',m''}*y^l_{l',m'}*\varphi_l\bigr)_{{{\bf n } } '' } \over l''(l''+1)+r}\right\}\nonumber\\\end{aligned}\ ] ] @xmath127:&=&\int d^2{{\bf n}}'d^2{{\bf
\sum_{l',m'}\sum_{l'',m''}\left\ { { \bigl(\varphi_l*{\overline y}^l_{l',m'}*y^l_{l'',m''}*\varphi_l\bigr)_{{{\bf n } } ' } \over l'(l'+1)+r}\right.\nonumber\\ & & \kern 70pt\left.\times{\bigl(\varphi_l*{\overline y}^l_{l',m'}*y^l_{l'',m''}*\varphi_l\bigr)_{{{\bf n } } '' } \over l''(l''+1)+r}\right\}\nonumber\\\end{aligned}\ ] ] @xmath128:&=&\int d^2{{\bf n}}'d^2{{\bf
\sum_{l',m'}\sum_{l'',m''}\left\ { { \bigl(\varphi_l*{\overline y}^l_{l',m'}*\varphi_l*y^l_{l'',m''}\bigr)_{{{\bf n } } ' } \over l'(l'+1)+r}\right.\nonumber\\ & & \kern 70pt\left.\times{\bigl(\varphi_l*{\overline y}^l_{l'',m''}*y^l_{l',m'}*\varphi_l\bigr)_{{{\bf n } } '' } \over l''(l''+1)+r}\right\}\nonumber\\\end{aligned}\ ] ] and @xmath129:&=&\int d^2{{\bf n}}'d^2{{\bf n } } '' \sum_{l',m'}\sum_{l'',m''}\left\ { { \bigl(\varphi_l*{\overline y}^l_{l',m'}*\varphi_l*y^l_{l'',m''}\bigr)_{{{\bf n } } ' } \over l'(l'+1)+r}\right.\nonumber\\ & & \kern 70pt\left.\times{\bigl(\varphi_l*{\overline y}^l_{l'',m''}*\varphi_l*y^l_{l',m'}\bigr)_{{{\bf n } } '' } \over l''(l''+1)+r}\right\}.\nonumber\\\end{aligned}\ ] ] surprisingly diagrams with the interchange in the second factor of @xmath130 ( which we indicate with a bar on the diagram label ) are generally not equal to their unbarred counterpart , even though this corresponds to a spatial twist of @xmath131 in third dimension out of the plane of the paper .
however only the above four are in fact distinct , the one new contribution being @xmath132 , the others are not new but related those given by the identities @xmath133 and @xmath134 .
our aim is to show that these diagrams have no anomalous contributions i.e. that their limiting form for @xmath67 is that of the commutative theory .
we begin by arguing that the large @xmath16 limit of the planar contribution @xmath135 is the commutative one . for this
we look at the contribution to @xmath136 comming from the process @xmath137 which after simplifying yields : @xmath138 where we have defined @xmath139 because of the triangular inequalities @xmath140 it is first established that @xmath141 for a positive constant @xmath142 , depending on the external momenta .
this established that the @xmath67 limit of @xmath135 is finite and we can therefore interchange the process of taking the large @xmath16 limit and the summations .
the limiting form of @xmath135 is then obtained by noting that for @xmath67 we have @xmath143 then substituting this into ( [ ldependentpart ] ) we obtain the leading asymptotic , which is readily seen to be the commutative expression . the contribution to the diagram @xmath132 can be obtained from that for @xmath135 ( [ i_pp_result ] ) but the corresponding @xmath144 acquires the additional phase factor @xmath145 .
this procedure can be followed for the remaining diagrams , to establish that they also reduce to the commutative limit .
however , this is not very insightful so we shall take a different tact .
we shall show that the difference of diagrams vanishes as @xmath67 , and so establish that they all have the same limit , which from the above we know is the commutative limit .
consider , therefore , difference between ( [ pp ] ) and ( [ np ] ) , @xmath146:&=&i_{p , p}[\varphi]-i_{n , p}[\varphi]= \int d^2{{\bf n}}'d^2{{\bf n } } '' \sum_{l',m'}\sum_{l'',m''}\nonumber\\ & & \kern-40pt\left\{{\bigl(\varphi_l*{\overline y}^l_{l',m'}*[y^l_{l'',m''},\varphi_l]_*\bigr)_{{{\bf n } } ' } \bigl(\varphi_l*{\overline y}^l_{l'',m''}*y^l_{l',m'}*\varphi_l\bigr)_{{{\bf n } } '' } \over \bigl(l'(l'+1)+r\bigr)\bigl ( l''(l''+1)+r\bigr)}\right\}.\nonumber\\\end{aligned}\ ] ] again one can re - arrange traces and sums to express this as @xmath147:&= & { 1\over 2}\int d^2{{\bf n}}'d^2{{\bf n } } '' \sum_{l',m'}\sum_{l'',m''}\nonumber\\ & & \kern-40pt\left\{{\bigl([\varphi_l,{\overline y}^l_{l',m'}]_**[y^l_{l'',m''},\varphi_l]_*\bigr)_{{{\bf n } } ' } \bigl(\varphi_l*{\overline y}^l_{l'',m''}*y^l_{l',m'}*\varphi_l\bigr)_{{{\bf n } } '' } \over \bigl(l'(l'+1)+r\bigr)\bigl ( l''(l''+1)+r\bigr)}\right\}.\nonumber\\\end{aligned}\ ] ] navely one might infer from the commutators to that this difference is of order @xmath86 , but the analysis of the two point function shows that one must be more careful .
in fact we shall show that , for the four - point function , this expectation is indeed correct .
the crucial difference from the two - point function is that neither of the propagators here closes on itself .
expanding the commutators to first order , as was done for the 2-point function , gives @xmath148&\approx & { 2\over l^2}\int d^2{{\bf n}}'d^2{{\bf
n } } '' \left\{\bigl({{\cal l}}_a\varphi_l({{\bf n}}')\bigr)j^{ab}({{\bf n}}')\bigl({{\cal l}}'_b{{\cal g}_l}({{\bf n}}',{{\bf n}}'')\bigr)\right.\nonumber\\ & & \kern90pt\times\left.\bigl({{\cal l}}_d\varphi_l({{\bf n}}')\bigr)j^{dc}({{\bf n}}')\bigl({{\cal l}}'_c { { \cal g}_l}({{\bf n}}',{{\bf n}}'')\bigr ) \varphi^2_l({{\bf n}}'')\right\},\nonumber\\\end{aligned}\ ] ] plus corrections down by further factors of @xmath84 .
the simplest way to study the structure of @xmath149 $ ] as @xmath67 is to write the propagators using ( [ approxfuzzyg ] ) and observe that @xmath150 where @xmath151 is the derivative of the legendre polynomial .
any singularity in the integrand in ( [ fourpointdiff ] ) is coming from terms with @xmath152 and @xmath153 near @xmath16 and most of the support of the legendre polynomials for such high orders is concentrated within a region of width @xmath84 at the edges of the interval @xmath154 .
this allows us to set @xmath155 in the argument of the @xmath59 terms in ( [ fourpointdiff ] ) . since @xmath156 we can approximate the sum in ( [ gprime ] ) by @xmath157
finally we split the double integral up into centre - of - mass co - ordinates @xmath158 and relative co - ordinates @xmath159 .
the relative co - ordinates are integrated over a set of measure @xmath86 on the fuzzy sphere and in this region @xmath160 .
this implies that ( [ fourpointdiff ] ) is approximately @xmath147&\approx & -{l^2\over 2}\int d^2{{\bf n}}_1
\bigl({{\cal l}}_a\varphi^2_l({{\bf n}}_1)\bigr)j^{ab}({{\bf n}}_1 ) \bigl({{\cal l}}_d\varphi^2_l({{\bf n}}_1)\bigr)j^{dc}({{\bf n}}_1)\nonumber\\ & & \kern 50pt\times\int_{1/l^2 } d^2{{\bf n}}_2 ( { { \bf n}}_1\times{{\bf n}}_2)_b({{\bf n}}_1\times{{\bf n}}_2)_c,\end{aligned}\ ] ] where the integral over @xmath159 is over a region of area @xmath86 in which @xmath161 .
the second integral above is therefore of order @xmath162 hence the whole expression is of order @xmath86 and so vanishes in the continuum limit , @xmath67 .
thus , in the @xmath67 limit , there is no difference between the planar and non - planar contributions to the four - point coupling at one - loop .
a similar conclusion holds for the differences ( [ pp ] ) and ( [ pbarp ] ) , ( [ pp ] ) and ( [ nn ] ) , and between ( [ np ] ) and ( [ nn ] ) .
we have have only examined the first term in the @xmath2-star product expansion , but the result can be similarly established for higher derivatives as well , since the crucial observation was that all derivatives of green functions only involve green functions with different arguments and only the double integral ever allows these points to co - incide .
it is relatively easy to extend the above analysis to higher vertex functions and higher loop orders .
one can , in fact , see that , for this two dimensional theory on the fuzzy sphere , as long as the free propagator entering the diagram does not return to the same vertex from which it departs the diagram will limit to the commutative on and there will be no anomalous contribution . in other words for this theory anomalous contributions are restricted to tadpole diagrams .
we have seen that the nave non - commutative action ( [ inttrace ] ) does not reproduce standard @xmath0 theory in the @xmath67 limit , but instead gives a continuum theory with a more complicated momentum dependence in the quadratic term .
however the only source of this deviation from the standard theory is the tadpole diagrams and this allows it to be removed in a standard manner , by using a normal ordered vertex . by a normal ordered vertex we mean one that has subtracted all possible self contractions with a suitable propagator . to this end
we define a modified bare action in matrix form , @xmath163&=&{4\pi\over ( l+1)}{\hbox{tr}}\left\{{1\over 2}[l_a,\phi][l_a,\phi ] + { 1\over 2 } t \phi^2 + { \lambda\over 4!}:\phi^4:\right\},\end{aligned}\ ] ] where the normal ordered interaction can be written @xmath164^\dagger[\phi , \hat y_{lm}]\over l(l+1)+t}\right ) \right\}.\nonumber\\\end{aligned}\ ] ] the middle term is the usual tadpole subtraction that renders the two dimensional theory finite and corresponds to normal ordering the commutative vertex .
the last term is the additional , manifestly positive , noncommutative subtraction that is necessary to obtain the correct commutative limit . a more transparent form of this new action is given by @xmath163={4\pi\over ( l+1)}{\hbox{tr}}\left\{{1\over 2}\phi\left(\hat{{{\cal l}}}^2-{\lambda\over 2}{\cal q}_l(\hat{{{\cal l}}}^2 ) + t-{\lambda\over 2}{\hbox{$\bigcirc\kern-3.6pt{\bf\bullet}$\kern 1.3pt}}(l , t)\right)\phi + { \lambda\over 4!}\phi^4\right\ } , \label{normalorderedaction}\end{aligned}\ ] ] where @xmath165 $ ]
, @xmath125 is defined in ( [ calq_defn ] ) and @xmath166 in ( [ commutative_tadpole ] ) .
note that the first terms can be interpreted as a momentum dependent wavefunction renormalization since @xmath167 is a power series in @xmath168 which starts at order @xmath168 and therefore we could have written @xmath169 this theory , ( [ normalorderedaction ] ) , is therefore the correct matrix model that represents a lattice regularization of commutative theory .
note that we have further replaced the parameter @xmath60 with @xmath170 , as the two are different . in a purely commutative theory
the replacement @xmath171 establishes the relationship between the two parameters .
this can be iterated to generate all tadpole contributions .
the new quadratic terms exactly cancel all the unwanted momentum dependent quadratic terms in the effective action arising from non - planar diagrams in the non - commutative theory and the continuum limit of this theory is the standard @xmath0 theory in 2 dimensions .
curiously , for this new model the direct large @xmath16 limit of the action does not give the commutative action but rather gives a non - local action .
however this non - locality is precisely what is need for the full quantum field theory to reproduce the commutative limit . for a complex scalar field on the fuzzy sphere
there are two potential vertices : @xmath172 and @xmath173 .
the first has only a planar tadpole and so will have no anomalous contribution , however the second has a non - planar and therefore anomalous tadpole . again
the normal ordering prescription removes the unwanted contribution .
in fact for any two dimensional @xmath0 model with global @xmath174 symmetry normal ordering the vertex will guarantee the commutative limit is recovered .
we have here only established the correct model for the two dimensional case . a case that warrants further attention
is that of @xmath0 for a four dimensional fuzzy space .
the simplest example is to take the case @xmath175 . in this case
it is easily seen from the above techniques that the problems are more severe and that there are additional residual non - local differences for the two and four - point functions .
this is due to the divergences that appear in the commutative limit .
it therefore appears to be difficult to establish precisely what fuzzy model will reproduce the commutative limit .
we leave this question open for the moment but hope to return to it in the near future . as a final comment we observe that the normal ordering prescription proposed here removes the uv / ir mixing problems for the two dimensional @xmath0 model on the moyal plane .
this is readily seen since the this latter model can be recovered as a particular scaling limit of the model considered here @xcite . *
acknowledgments * it is a pleasure to thank sachin vaidya for discussions which led to this work .
we also benefited from comments especially of a.p .
balachandran , oliver jahn and alan stern .
this work was supported by the joint conacyt - nsf grant e120.0462/2000 and by conacyt grant 30422-e .
the work of p.p . was partially supported by vega project 1/7069/20 .
h. grosse , c. klimk and p. prenajder , `` towards finite quantum field theory in noncommutative geometry , '' int .
j. theor .
* 35 * , ( 1996 ) 231 , [ hep - th/9505175 ] ; h.grosse and a.strohmaier , _ lett .
phys . _ * 48 * , ( 1999 ) 163 , _ hep - th/9902138_. h. grosse , c. klimk and p. prenajder , _ comm .
phys . _ * 178 * , ( 1996 ) 507 ; h. grosse and p. prenajder , _ lett .
_ * 46 * , ( 1998 ) 61 .
balachandran and s. vaidya , _ int .
_ * a16 * , ( 2001 ) 17 , hep - th/9910129 ; a. p. balachandran , t. r. govindarajan and b. ydri , hep - th/9911087 ; a. p. balachandran , t. r. govindarajan and b. ydri , _ mod .
lett . _ * a15 * , ( 2000 ) 1279 , hep - th/0006216 ; a. p. balachandran , x. martin and d. oconnor , and _ int .
phys . _ * a 16 * ( 2001 ) 2577 - 2594 , hep - th/0007030 . s. baez , a. p. balachandran , s. vaidya and b. ydri , _ comm .
phys . _ * 208 * , ( 2000 ) 787 , hep - th/9811169 .
g. alexanian , a.p .
balachandran , g.immirzi and b.ydri , _ j. of geom . and phys .
_ ( in press ) , hep - th/0103023 .
j. madore , `` the fuzzy sphere , '' class
. quant .
* 9 * , ( 1992 ) 69 .
s. minwalla , m. van raamsdonk and n. seiberg , `` noncommutative perturbative dynamics , '' jhep * 0002 * , ( 2000 ) 020 , [ hep - th/9912072 ] .
y. kinar , g. lifschytz , j. sonnenschein , `` uv / ir connection , a matrix perspective '' , hep - th/0105089 ; a. matusis , l. susskind and n. toumbas , `` the ir / uv connection in the non - commutative gauge theories , '' jhep * 0012 * , ( 2000 ) 002 , [ hep - th/0002075 ] ; f. r. ruiz , `` gauge - fixing independence of ir divergences in non - commutative u(1 ) , perturbative tachyonic instabilities and supersymmetry , '' phys .
b * 502 * , ( 2001 ) 274 , [ hep - th/0012171 ] ; k. landsteiner , e. lopez and m. h. tytgat , `` instability of non - commutative sym theories at finite temperature , '' hep - th/0104133 ; c. p. martin and f. r. ruiz , `` paramagnetic dominance , the sign of the beta function and uv / ir mixing in noncommutative u(1 ) , _ nucl .
phys . _ * b597 * ( 2001 ) 197 , _ hep - th/0007131 _ ; i. ya .
arefeva , d.m . belov and a.s .
koshelev , `` a note on uv / ir for noncommutative complex scalar field '' _ hep - th/0001215 _ ; l. griguolo and m. pietroni , `` hard noncommutative loops resummation '' _ hep - th/0102070 _ ; wilson renormalization group and the noncommutative ir / uv connection '' * jhep 0105 * ( 2001 ) 032 , _
hep - th/0104217_. s. vaidya , _ perturbative dynamics on fuzzy @xmath1 and @xmath176 _ , _ phys .
lett . _ * b512 * ( 2001 ) 403 - 411 , hep - th/0102212 chong - sun chu , john madore and harold steinacker , _ scaling limits of the fuzzy sphere at one loop _ , hep - th/0106205 d. a. varshalovich , a. n. moskalev and v. k. khersonsky , `` quantum theory of angular momentum : irreducible tensors , spherical harmonics , vector coupling coefficients , 3nj symbols , '' _ singapore : world scientific ( 1988)_. | we demonstrate that the uv / ir mixing problems found recently for a scalar @xmath0 theory on the fuzzy sphere are localized to tadpole diagrams and can be overcome by a suitable modification of the action .
this modification is equivalent to normal ordering the @xmath0 vertex . in the limit of the commutative sphere
, the perturbation theory of this modified action matches that of the commutative theory . |
evolution equations play an important role in both inclusive @xcite and exclusive @xcite hard processes .
they describe the dependence of parton distribution functions of dis and of the parton wave functions on the renormalization parameter @xmath5 .
the main ingredients of these equations are the perturbatively calculable evolution kernels .
the two - loop dglap evolution kernels @xmath0 were obtained in @xcite , and a more complicated two - loop kernel @xmath1 of the brodsky lepage ( bl ) evolution equation for meson wave functions was calculated in @xcite . a three - loop calculation of these kernels
looks like being a tremendous problem .
recently , the results of very complicated calculations of the first few elements of the three - loop anomalous dimension of composite operators in dis @xmath6 have been presented in @xcite . lately
, these results were applied to improve the qcd analysis of dis experimental data @xcite . despite this important technical and phenomenological progress
one may feel insufficiency because of the absence of the whole kernel in this order or , even at least , some parts of it .
of course , the kernel admitting a physical probability interpretation is a more general and convenient object from the theoretical point of view .
a method of estimating anomalous dimensions of composite operators in the limit of a large number of flavours , @xmath7 , has been suggested by j. gracey @xcite . it is based on conformal properties of the theory at the critical point @xmath8 corresponding to the non - trivial zero @xmath9 of the @xmath10-dimensional @xmath11-function @xmath12 .
the generating function has been constructed to obtain the leading large-@xmath7 asymptotics of the anomalous dimension @xmath13 to any order @xmath14 of perturbation theory ( pt ) . in this letter ,
i propose a method of calculating certain classes of multiloop diagrams directly for the dglap kernel @xmath0 as well as for the bl kernel @xmath1 in the scheme rather than their moments as in @xcite .
these diagrams contain insertion of chains of one - loop self - energy parts ( renormalon chains ) into the one - loop diagrams ( see fig.1 a , b ) for kernels .
the kernels @xmath15 are obtained for any order @xmath14 of pt by analysing of these dressed by chain " diagrams .
then , the kernel @xmath16 ( @xmath17 ) for diagrams with the totally dressed propagator ( , in the all loops " approximation ) is calculated ; this kernel appears to be a generating function to obtain the partial kernels @xmath15 .
moreover , this kernel @xmath16 is the analytic function in variable of a coupling constant @xmath18 .
the insertions of the renormalon chains lead to converged pt series , and the usual renormalon singularities do nt appear in the renormgroup function calculations .
this was found first for @xmath11- and @xmath19-functions in @xcite , the following development of this approach is presented in @xcite .
it is to be stressed that the method of obtaining @xmath20 or @xmath17 does not depend on the nature of self - energy insertions and does not appeal to the value of parameters @xmath21 or @xmath22 ( for qcd case ) associated with different loops .
another distinctive feature is that the pt - improved evolution kernels are calculated in the direct and standard way by using rather elementary methods .
this is possible due to a simple algebraic structure of the counterterms of the diagrams considered . in this letter , we mainly present technical results on resummation of the above mentioned counterterms within the framework of the @xmath2_6 $ ] model and , in part , in qcd . the solution to the bl evolution equation for a large number of flavors
@xmath7 is also briefly discussed . to develop the diagrammatic analysis of multiloop evolution kernels ,
let us first consider a toy model describing @xmath23 , the charged quark " fields , and @xmath24 , the gluon " one , in @xmath25 space - time dimensions based on the lagrangian @xmath26 . the number of quark " flavors , @xmath7 , will be considered as an arbitrary free parameter associated with the quark " loops
this theory has much in common with qcd : it is renormalizable and its @xmath11-function has a structure similar to that of @xmath27 , @xmath28 moreover , the one - loop evolution kernels @xmath29 and @xmath30 , following from the simplest triangular diagram in figs.1a and 1b , turn out to be proportional to the corresponding expressions for the same diagrams in qcd ap_0(z)=a(1-z ) , av_0(x , y)= a ( ( y > x)+ ( > ) ) , where @xmath31 .
the similarity of the kernels in both the theories extends to the two - loop level @xcite .
the advantage of the @xmath2_6 $ ] model is that the relevant feynman integrals are simple , and it is much easier to study the structure of multiloop expressions . in general , calculation of the evolution kernels
@xmath32 @xcite and @xmath33 @xcite is quite analogous to that of the anomalous dimensions .
the major modification is the change @xmath34 of the vertex factor corresponding to the composite operator , @xmath35 being the relevant momentum related to the incoming quark " line , see figs .
1a , b . in these figs .
, @xmath36 ( @xmath37 or @xmath36 ) is the external momentum of the diagrams ; @xmath14 is a light - like vector ( @xmath38 ) introduced to pick out the symmetric traceless composite operator ; @xmath39 .
detailed examples of similar calculations can be found in @xcite .
in this section , the method will be demonstrated by using well - understood @xcite quark - loop insertions as an example .
let us consider the one - loop triangular diagram @xmath40 for kernels in fig.1c with the chain of self - energy part insertions @xmath41 into the gluon line .
expressions for @xmath0 or @xmath1 in terms of the renormalization constant @xmath42 for the diagram @xmath40 in the scheme ( see @xcite and @xcite ) are given by z_=1 - r ( ) , p =- a_a(z^(1 ) _ ) = a_a(_1 r ( ) ) .
z here , @xmath43 is the incomplete bphz @xmath44-operation ; @xmath45 and @xmath10 is the space - time dimension ; @xmath46 picks out poles in @xmath47 , where @xmath48 projects a simple pole , and @xmath49 is the coefficient of the simple pole in the expansion of @xmath42 . for @xmath41 , being the one - loop insertions , expression ( [ z ] ) can be simplified following the definition of the @xmath43-operation as _ 1 r()_1 r(g_i g_i)=_1 p_(n)=(n+1)a _ 1 , pn where @xmath50 denotes the outer triangle " part of the diagram @xmath40 without the chain of bubbles , and rg_i(k^2 ) & = & ( 1 - _ 1 ) g_i(k^2 ) = - ( _ ( ) ( ) ^-_(0 ) ) rg + _ ( ) & = & c ( ) b(2- , 2- ) .
ghere @xmath51 is the euler b - function , @xmath52 is the one - loop anomalous dimension of the gluon field ( at @xmath53 ) , and the constant @xmath54 reflects the concrete choice of scheme where every loop integral is multiplied by the scheme factor @xmath55
. note , the function @xmath56 in ( [ g ] ) will play an important role in our consideration . substituting ( [ rg ] ) into ( [ pn ] ) and performing a direct calculation ( see , , @xcite ) ,
one arrives at the expression pnz p_(n)^(1)(z)=(n+1 ) a p_0(z ) ( -a n_f _ ) ^n _ 1 .
it is instructive to consider the properties of the first few kernels @xmath57 .
their expressions are presented in table 1 for @xmath58 . to obtain them , the form 2.0 program
@xcite has been used . , see @xcite .
note that the contents of table 1 is limited here only by place . ]
the lessons are : i ) the perturbation theory for kernels @xmath59 looks like being improved in comparison with the corresponding anomalous dimensions @xmath13 .
indeed , the term in @xmath59 leading in @xmath60 is @xmath61 , , it is factorial suppressed in @xmath14 , while for the corresponding contribution to @xmath62 is not suppressed ; ii ) one can see factorial suppression in @xmath14 of all other logarithmic terms in @xmath59 ; iii ) the @xmath14-bubble chain generates @xmath63 ( the riemann zeta - function ) in the non - logarithmic term in @xmath59 ; this may be traced back to the expansion of the euler b - function as their possible origin .
these properties give hint about a possible resummation of the @xmath59-series .
+ indeed , the closed expression for the sum of partial kernels @xmath64 exist , t1 p^(1)(z ; a ) = a_n=0^p_(n)^(1)(z ) = a p_0(z ) z^-a(1-a ) ( ) , a = a n_f _ and the kernel @xmath65 is the generating function for @xmath59 .
the proof of this assertion is presented in appendix a. this result possesses several remarkable properties : * the @xmath65 becomes the dominant part of the total perturbative kernel @xmath0 when + @xmath3 .
below , the result for the kernel @xmath66 will be completed by taking into account similar corrections to the quark leg " ( see fig.1d ) whose contribution is proportional to @xmath67 . * the analytic properties of @xmath66 in the variable @xmath68
are completely determined by the one - loop anomalous dimension " @xmath69 in @xmath10 dimensions ( see eqs.([rg ] , [ g ] ) ) .
the singularities of @xmath66 correspond to zeros of the function @xmath70 in @xmath69 .
of course , @xmath69 provides the scheme dependence in @xmath66 due to the factor @xmath71 in the definition ( [ g ] ) . *
the kernel @xmath66 is an analytic function in variable @xmath68 , except for singularities at the points + @xmath72 @xmath73 , where the function has simple poles .
the nearest singularity appears at @xmath74 , , at @xmath75 , determining the range of convergence of the pt series . in a similar way
, one can derive an expression for the sum of diagrams in fig.1 d connected with the anomalous dimension of the quark field @xmath76 obtained in the all quark - loop " approximation .
collecting the results of resummation in the main approximation in @xmath68 , which correspond to the diagrams in figs.1c and d , we arrive at the final expression final2 p^(1)(z ; a ) - ( 1-z)_^(1)(a ) = a ( ) .
integration produces the following expression for the anomalous dimension @xmath77 of the composite operator gamma ( n , a)&= & _ 0 ^ 1 z^n ( p^(1)(z ; a ) - ( 1-z)_(a ) ) dz = + & = & a _ . formula ( [ gamma ] ) can be obtained by another method applied in @xcite to the qcd case ( see eq.([mq1 ] ) below ) .
note that the anomalous dimension @xmath78 corresponds to parton momentum . the equality @xmath79 , following from the r.h.s . of ( [ gamma ] ) , is expected
; it signals that the parton momentum is conserved .
consider now the class of diagrams in fig .
2 . there are closed expressions for sums of partial kernels t2a p^(2a)(z ; b)&=&a _ m=0^p^(2)_(m)(z)= a p_0(z ) ( ( ) ^-b ) , , + ^(2a)(z;b)&=&a_m=0^(2-_0,n ) p^(2)_(m)(z ) = a p_0(z ) ( 2 ( ) ^-b -1 ) , + mc , t2b + p^(2b)(z ; b)&=&a _ m=0^ ( m+1 ) p^(2)_(m)(z)= a p_0(z ) ( 1+b ) ( ( ) ^-b ) , , t2c + ( 1-z ) ^(2 ) _ & = & a ( 1-z ) ( ) , ; t2d the functions @xmath80 appear as generating functions for the corresponding partial kernels . here
@xmath81 is the partial kernel with @xmath82 insertions into one of the quark lines ( see fig.2a ) ; @xmath83 is the one - loop anomalous dimension of the quark field in @xmath10-dimension ; for this model @xmath84 , @xmath85 , @xmath86 ; mc denotes a mirror - conjugate diagram .
the first eq.([t2a ] ) corresponds to the diagrams in fig .
2a where the chain of quark self - energy parts is substituted only into the left quark line of the triangle .
to prove it , one has to repeat the way similar to proof in appendix a. equation ( [ t2b ] ) corresponds to the sum of diagrams in fig.2a and its mc diagrams .
this result will be used to restore the corresponding kernel @xmath1 in the next section .
the analytic properties of the functions @xmath87 and @xmath88 in the parameter @xmath89 are the same as for the kernel @xmath90 ; they are determined by the function @xmath91 .
equation ( [ t2c ] ) corresponds to substitutions of the chains into the both quark lines of the triangle in fig.2b .
at least , eq.([t2d ] ) corresponds to contributions to the anomalous dimension of the quark field from the diagram in fig.2c .
the contribution @xmath92 will be suppressed in the parameter @xmath89 in comparison with @xmath93 , if @xmath3 , , @xmath94 .
note , however , that the @xmath95-moments of the kernels @xmath96 @xmath97 decrease in @xmath95 slowler than @xmath98 corresponding to the kernel @xmath99 . therefore , at sufficiently large @xmath95 @xmath100 for any fixed parameters @xmath68 and @xmath89 .
here , some partial results of the bubble resummation for the bl kernels @xmath17 are presented . we obtain them as a byproduct " of our previous results for the kernel dglap @xmath20 .
the results obtained may be used , in particular , to check the regular calculation of the kernel @xmath17 in high orders of pt .
note , the diagrams for the bl kernel differ from the dglap diagrams only by the exclusive " kinematics of the input momentum , compare the diagrams in fig.1a and 1b .
so , one can repeat the proof similar to that in appendix a for this case .
there is another far more elegant way to use exact relations between @xmath17 and @xmath20 kernels for triangular diagrams which were established for any order of perturbation theory in @xcite .
these relations work for both the @xmath2_6 $ ] model and qcd .
i quote these propositions without proofs .
let the diagram in fig.1c have a contribution to the dglap kernel in the form @xmath111 @xmath112 ; then its contribution to the bl kernel is v1 v(x , y ) = c ( y > x ) ^_0 dz + ( y - x ) c , where @xmath113 . substituting eq .
( [ final2 ] ) for @xmath114 into relation ( [ v1 ] ) we immediately derive the expression for @xmath115 v1a v^(1)(x , y ; a ) = c a ( ) .
the later expression may independently be verified by other relations by reducing any @xmath17 to @xmath20 @xcite + ( @xmath116 reduction ) reduct v^(1)(x , y ; a ) = c(y >
x)f(;a ) - ( y - x ) c(a ) p^(1)(z ; a ) & & + p^(1)(z ; a)= ( 1-z ) f(z;a)- ( 1-z ) c , & & indeed , substituting ( [ v1a ] ) into ( [ reduct ] ) we return to the same eq.([final2 ] ) for @xmath114 . moreover , the first term of the taylor expansion of @xmath117 in @xmath68 coincides with the results of the two - loop calculation in @xcite .
the contribution @xmath118 should dominate for @xmath3 in the whole kernel @xmath17 , and moreover , the function @xmath115 possesses an important symmetry
. really , the function @xmath119 is symmetric under the change @xmath120 , @xmath121 .
this symmetry allows us to obtain the eigenfunctions @xmath122 of the equation ev & & _ 0 ^ 1 v^(1)(x , y ; a)_n(y ) dy= ( n;a ) _
n(x ) , + & & _
n(y ) ~(y)^d_(a ) - 1 c_n^(d_(a ) - 12 ) ( y- ) , d_(a ) = d_a/2 - 1 , d_a=6 - 2a , solv1 @xmath123 is the effective dimension of the scalar field when the anomalous dimension is taking into account , and @xmath124 are the gegenbauer polynomials of an order of @xmath125 .
this form of eigenfunctions @xmath126 as well as the @xmath120 symmetry of the function @xmath127 are the consequences of conformal symmetry conservation ( see @xcite for details ) for the sum of diagrams under consideration . equation ( [ ev ] ) is tightly connected with the bl evolution equation in a special case when the @xmath11-function @xmath128 . in this case , the bl equation is simplified , the variables are separated , and the evolution equation may be reduced to ( [ ev ] ) .
the partial solution to the bl equation turns out to be proportional to @xmath129 in ( [ solv1 ] ) . for the diagrams in fig .
2a plus its mc diagrams , the relation between @xmath130 and @xmath131 is ( see @xcite ) v2 ^(2a)(x , y ) = c(y > x ) 12 . substituting ( [ t2b ] ) into ( [ v2 ] ) one arrives at the expression for @xmath132 ^(2a)(x , y ) = c a(y > x ) \ { ( ) - } , which satisfies the same check as in the previous case . of course , this formula does not cover the diagrams in fig .
2b , where both the quark lines are dressed .
the assertions about diagram contributions to @xmath20 and @xmath17 , similar to the presented above , are also valid for the qcd diagrams .
their proof does not contain essentially new elements , but looks rather cumbersome .
the complete results for the qcd evolution kernels will be presented in a subsequent paper . here , at first , the qcd results for triangular diagrams in fig . 1 c , d in the feynman gauge are discussed .
based on the proof in appendix a in the qcd version , one can derive the result for the sum of diagrams , in fig .
1c , in qcd pqcd p^(1c)(z ; a ) = c_f 2(1-a)^2 z^-a - c_f(1-z ) ( 1(1-a ) -1 ) . here , @xmath133
; @xmath134 is twice the contribution to a one - loop @xmath10dimensional anomalous dimension of the gluon field and @xmath135 ; @xmath136 is the contribution to the standard anomalous dimension ; @xmath137 .
it should be emphasized again that the form of exp.([pqcd ] ) does not depend on the nature of self - energy insertions into the gluon line .
one can use it for the resummation of both the quark @xmath138 and gluon ( @xmath139 ) loops . substituting into the general formula ( [ pqcd ] ) ,
the well - known expressions for @xmath134 from the quark or gluon ( the ghost loop is also added ) loops piq _ q ( ) &
= & -8 n_f t_r b(d/2,d/2 ) c ( ) , + _ g ( ) & = & b(d/2 - 1,d/2 - 1 ) ( ) c ( ) , pig one obtains the expression @xmath140 for the quark loop insertions pq p^(1c)_q(z ; ) & = & c_f 2z^- + & & - c_f(1-z ) ( 1(d_q/2 - 1 ) -1 ) where @xmath141 , @xmath142 , @xmath143 is the scheme factor , + and the expression @xmath144 for the gluon
loop insertions pg p^(1c)_g(z ; ) & = & c_f 2z^- + & & - c_f(1-z ) ( 1(d_g/2 - 1)-1 ) , where @xmath145 . the expression ( [ pg ] ) for @xmath146
has no any singularity in the parameter @xmath147 due to the asymptotic freedom .
consider eq.([pq ] ) for @xmath148 in detail . by adding the contributions from the diagrams in fig.1d to ( [ pq ] )
, one can find that the second term cancels in part and the final expression turns out to be the expected plus form " ( see , @xcite ) pq1 p^(1c , d)_q(z ; ) = c_f 2 ( z^- ) _ + , ^1_0 p^(1c , d)_q(z ; ) dz= _ q^(1c , d)(n=0 , ) = 0 due to the current conservation .
the analytic properties of @xmath149 in @xmath150 are the same as for its scalar analogy @xmath66 ( see ( [ t1 ] ) in sect.2 ) ; they are determined by the behavior of the function @xmath151 in @xmath150 , see eq.([piq ] ) .
the nearest singularity of @xmath140 in @xmath150 appears at @xmath152 .
the moments of eq .
( [ pq1 ] ) agree with the corresponding part of the generating function in @xcite mq1 _ q^(1c , d)(n , ) = -c_f , see the first term in eq.(14 ) in @xcite , and ref .
@xcite ( note that our moments differ in sign from the definition of the anomalous dimension of composite operators used there ) . to complete the qcd calculations of the kernel @xmath99
, we need the contribution from the last diagram in fig.1e with the chain in the gluon line that is inserted into the composite operator .
it can be obtained in a similar way , as in the previous qcd
calculations for eq.([pqcd ] ) pqcd2 p^(1e)(z ; a ) = c_f 2 ( ) _ + , and the expression automatically has the plus form " , see @xcite for details . collecting eq.([pq1 ] ) and eq.([pqcd2 ] ) , one easily arrives at the complete qcd expression for @xmath153 final p_q^(1)(z ; ) = c_f 2 _ + .
the contribution @xmath153 is gauge invariant . the moments of @xmath153 again agree with the complete generating function obtained in @xcite ( see eq.(14 ) there ) . to obtain the contribution from both kinds of insertions into the gluon line one would take certain substitutions in ( [ final ] ) : @xmath154 and @xmath155 , here @xmath156 . of course , for this case a final result must be gauge dependent .
the method for calculating certain classes of multiloop diagrams for the kernel @xmath0 of the non - singlet dglap evolution equation is presented .
these multiloop diagrams appear due to the insertion of chains of one - loop self - energy parts ( renormalon chains ) into the lines of the one - loop diagrams for the kernel .
closed expressions @xmath157 are found for sums of all the diagrams which belong to two of the diagram classes , see sect . 2 - 3 .
the corresponding proofs are based on a simple algebraic structure of the counterterms for the diagrams under consideration ; this is outlined in appendix a. moreover , the kernels @xmath158 are generating functions for the partial kernels @xmath159 which correspond to the @xmath14-bubble insertion .
the contribution @xmath160 ( corresponding to one of the diagram classes ) would dominate in the kernel for @xmath3 .
the analytic properties of the function @xmath158 in the variable @xmath161 are briefly discussed .
the expressions for partial kernels @xmath162 for the diagrams of a mixed class " , in any order of perturbation theory can also be obtained by using the form program .
the contributions @xmath163 to the brodsky
lepage kernel are obtained in sect . 4 for the same classes of diagrams as a byproduct " of the previous technique . when @xmath3 , a special solution to the brodsky lepage equation is derived .
we emphasize that the method of calculating the evolution kernels @xmath164 or @xmath165 does not depend on the nature of self - energy insertions and does not appeal to the value of parameters @xmath21 or @xmath22 ( for qcd case ) associated with different loops .
the method and results are exemplified with a simple @xmath2_6 $ ] model .
it is clear that all the results obtained above in the framework of the scalar model have a wider meaning and can be applied to the qcd case . here , in sect .
5 , some qcd results are presented too ; in particular , the kernel @xmath166 , that corresponds to the diagram dressed by the main quark - loop chain , is derived .
the anomalous dimension @xmath167 corresponding to this kernel agrees with the generating function obtained earlier by gracey @xcite .
the contribution @xmath168 from the diagrams with the gluon - loop chain is derived in the same way .
the same can be obtained for the mixed case when both kinds of bubbles ( quark and gluon ) are involved .
the completed qcd case will be considered in detail in a subsequent paper . * acknowledgements * the author is grateful to dr
. l. avdeev , dr .
m. kalmykov and dr .
a. bakulev for their help in programming the calculations by form , dr .
a. grozin and dr .
r. ruskov for fruitful discussions of the results , to dr .
n. stefanis , dr .
g. korchemsky and the referee of this letter for careful reading of the manuscript and useful remarks .
this investigation has been supported in part by the russian foundation for fundamental research ( rffr ) 96 - 02 - 17631 and intas 93 - 1180 ext
here , formular ( [ t1 ] ) is proved ; all others resummation formulae in the text above can be obtained in a similar manner .
+ * proof*. one could split the sum in square brackets , eq.([pnz ] ) , into two parts qt1 & & ( n+1 ) _
j=0^n ( f())^j = s_(n+1 ) ( ) + , + & & s_(n+1 ) ( ) .
snhere , we use the notation : @xmath169 , @xmath170 is the sum including all combinatoric factors " of the l.h.s . of ( [ qt1 ] ) , and the last term in the r.h.s . of ( [ qt1 ] ) may be associated with the @xmath171 term in sum . the term @xmath170 in the r.h.s . of ( [ qt1 ] )
does not contribute to the pole part , by virtue of lemma : @xmath172 .
the expression for @xmath59 can be derived by substituting decomposition ( [ qt1 ] ) into ( [ pnz ] ) and using the above mentioned lemma p_(n)^(1)(z ) & = & a p_0(z ) ( -a)^n _ 1 = + & = & a p_0(z ) ( a)^n _ 1 = a p_0(z ) . \ { } |_=0 .
pnz1now , let us take a sum of @xmath59 over all @xmath14 using exp.([pnz1 ] ) ; the differentials are naturally collected into the shift argument operator " @xmath173 shift p^(1)(z ; a ) = a _
n p_(n)^(1)(z ) = ap_0(z ) \ { ( a _ ) .
|_=0 } . finally ,
performing the shift of argument in the squar bracket in ( [ shift ] ) , we arrive at the expression pta p^(1)(z ; a)=a_n=0^ p_(n)^(1)(z ) = a p_0(z ) z^-a(1-a ) ( ) , and the partial kernels @xmath59 appear in the taylor expansion of @xmath66 in @xmath68 by construction let us consider the expansion of every element of the sum in @xmath170 lem1 in powers of @xmath47 up to @xmath177 .
any power @xmath178 of this expansion is accompanied by powers @xmath179 , where @xmath180 , and a coefficient that does not depend on @xmath181 .
therefore , the expansion of @xmath182 as a whole in the power series @xmath178 will generate coefficients of the powers which are composed only of the elements proportional to @xmath183 .
all these elements are equal to @xmath184 for @xmath185 in virtue of the identity [ a7 ] _
j=0^n+1j^l ( -)^n+1-j n+1 j = 0 , l < n+1 .
therefore , the power expansion of @xmath182 can be started with the power higher than @xmath186 .
note , the similar trick with the identity ( [ a7 ] ) had been used in @xcite .
the proof implies an obvious generalization of the elements ( [ lem1 ] ) that can be constructed as superpositions of @xmath40-functions depending on @xmath181 only through the arguments like @xmath187 .
gribov and l.n .
lipatov , sov.j . nucl.phys .
* 15 * ( 1972 ) 438 ; 675 ; l.n .
lipatov , sov.j .
. phys . * 20 * ( 1975 ) 94 ; y.l .
dokshitser , jetp * 46 * ( 1977 ) 641 ; g. altarelli and g. parisi , nucl .
* b 126 * ( 1977 ) 298 .
brodsky , and g.p .
lepage , phys .
* b 87 * ( 1979 ) 359 ; phys . rev .
* d 22 * ( 1980 ) 2157 .
e.g. floratos , r. lacaze and c.kounnas , phys .
* b 98 * ( 1981 ) 89 ; 285 .
dittes and a.v .
radyushkin , phys .
* b 134 * ( 1984 ) 359 ; m.h .
sarmadi , phys .
* b 143 * ( 1984 ) 471 ; s.v . mikhailov and a.v.radyushkin , _
evolution kernel for the pion wave function : two loop qcd calculation in feynman gauge."_. dubna preprint jinr p2 - 83 - 721 ( 1983 ) .
larin , t. van ritbergen , j.a.m .
vermaseren , nucl.phys .
* b 427 * ( 1994 ) 41 ; + s.a .
larin , p. nogueira , t .
van ritbergen , j.a.m .
vermaseren , _ the three loop qcd calculation of the moments of deep inelastic structure functions .
_ nikhef-96 - 010 , hep - ph/9605317 j. a. gracey , phys .
* b 322 * ( 1994 ) 141 ; j. a. gracey , _ renormalization group functions of qcd in large-@xmath7 _ , talk presented at third international conference on the renormalization group , jinr , dubna , russia , 26 - 31 august , 1996 , hep - th/9609164 .
vladimirov , theor.math .
* 36 * ( 1978 ) 732 ; * 43 * ( 1980 ) 417 .
vermaseren , symbolic manipulation with form , version 2 , tutorial and reference manual ( computer algebra nederland , amsterdam,1991 ) . l. v. avdeev et .
al . , phys . lett . * b 336 * ( 1994 ) 560 ; l. v. avdeev , comp.phys.commun .
* 98 * ( 1996 ) 15 ; l.v .
avdeev , j. fleischer , m. yu .
kalmykov , m. tentyukov , _ towards automatic analytic evaluation of massive feynman diagrams " _ , hep - ph/9610467 .
dittes , d. mller , d. robaschik , b. geyer and j. horejsi , phys .
* b 209 * ( 1988 ) 325 .
m. makeenko , sov .
. phys . * 33 * ( 1981 ) 440 .
p. gosdzinsky and n. kivel ,
_ `` resummation of @xmath188 corrections to the photon - meson transition form factor @xmath189 '' _ , hep - ph/9707367 . | the contributions to non - singlet evolution kernels @xmath0 , for the dglap equation , and @xmath1 , for the brodsky lepage equation , are calculated for certain classes of diagrams including the renormalon chains .
closed expressions are obtained for the sums of contributions associated with these diagram classes .
calculations are performed in the @xmath2_6 $ ] model and qcd in the scheme .
the contribution of one of the classes of diagrams dominates for a large number of flavours @xmath3 . for the latter case ,
a simple solution to the brodsky lepage evolution equation is obtained .
# 1@xmath4 -1 cm -10 mm -10 mm + s.v .
mikhailov + _ joint institute for nuclear research , bogoliubov lab . of theoretical physics , + 141980 , moscow region , dubna , russia _
+ pacs : 12.38.cy , 12.38.-t , 13.60.hb + keywords : dglap evolution kernel , anomalous dimensions , multiloop calculation |
lmc x-4 is a 13.5s pulsar orbiting a 20@xmath6 o7 iii - v companion every 1.4 days ( kelley 1983 ; ilovaisky 1984 ) .
the system exhibits a long term cycle with a roughly 30-day period ( lang 1981 ; ilovaisky 1984 ) similar to the 35-day period of her x-1 .
although the high - mass companion of lmc x-4 implies the presence of a wind , lmc x-4 has many properties in common with the disk - fed system her x-1 . in both systems
the long term periods are roughly 20 times the orbital periods .
both systems are fully eclipsing ( tananbaum 1972 ; li 1978 ) ; show a power - law spectrum with a soft x - ray excess " in the 0.5 - 10 kev band ( dennerl 1989 ) ; optical light curves with a power spectrum that shows power at the sum of the orbital and long term frequencies , but not in the difference , implying that the long - term variations are due to precession of an accretion disk ( ilovaisky 1984 ) ; and correlation between times of low hard x - ray luminosity and episodes of spindown ( dennerl 1991 ; wilson 1994 ) .
the optical lightcurves of lmc x-4 show changes in peak - to - peak amplitude of up to 40% relative to the mean over the 30-day period whereas the optical lightcurves of her x-1 show less than 20% change in peak - to - peak amplitude over the 35-day cycle . for lmc x-4
the x - ray flux varies by a factor of 60 between the high and low states of the 30-day period ; however , the flux of scattered x - rays , measured during x - ray eclipses , remains unchanged throughout the 30 day cycle ( woo 1995 ) , implying that the low state is not caused by a decrease in x - ray luminosity but by attenuation in intervening matter .
lmc x-4 also displays flares which typically occur once a day ( dennerl 1989 ; levine 1991 ; woo 1996 ) .
the x - ray luminosity during the flares rises to @xmath75 times the eddington luminosity and the spectrum changes from a power law to a very hot thermal spectrum .
there are several mechanisms that could cause the uv continuum or lines to vary with the 13.5 second neutron star rotation period . in vela
x-1 it has been shown that the siiv and nv p cygni lines of vela x-1 vary with the 283 second pulsar period , presumably as a result of time - dependent photoionization of the stellar wind by the x - rays ( boroson 1996a ) .
the uv continuum could pulsate as a result of x - ray heating of either the normal star or the accretion disk , as is the case in her x-1 ( boroson 1996b ) . here
we present coordinated ultraviolet and x - ray observations of lmc x-4 taken with the ghrs on hst and with the gis and sis on asca .
the high temporal resolution of the ghrs allowed us to search for a manifestation of the pulsations in the uv and the high spectral resolution allows us to study the geometry of the system as reflected in the line profiles .
some results from an asca observation covering a full binary orbit taken an year earlier are also presented .
we compare the hst data as well as archival iue data with models predicting uv continuum emission from the x - ray heated disk and star that has been successfullly applied to the her x-1 system .
we interpret the dramatic changes with orbital phase observed in the ultraviolet spectra in terms of the effects of x - ray photonionization on the stellar wind of the normal companion .
the observations and analysis are described in section 2 and our interpretation is discussed in section 3 .
figure 1 shows the location of our observation in comparison with the one - day averaged lightcurves obtained from the all sky monitor on xte and the same data folded with the long term ephemeris and period of dennerl ( 1992 ) .
this indicates that our simultaneous hst and asca observations occurred during the high state of the 30-day cycle , corresponding to a phase coverage of @xmath8=0.20 - 0.22 .
the asca lightcurves and hst coverage are plotted on figure 2 .
hst took eight seperate observations , but as the target was nearly in hst s continuous viewing zone the gaps between the observations are small .
a journal of the observations is given in table 1 .
hst captured an eclipse egress that did not have simultaneous x - ray coverage .
the hst observations were taken with the ghrs g160 m grating centered alternately at 1240 ( n v ) and 1550 ( c iv ) in the rapid mode with a time resolution of 0.5s .
our use of the ghrs rapid mode prevented us from over - sampling the spectrum , checking for bad counts , and correcting for the doppler shift due to the spacecraft orbit .
the observed flux in the nv @xmath9 line measured with iue ( van der klis 1982 ) ranges from @xmath10ergs @xmath4 @xmath11 at minimum ( @xmath12 ) to @xmath13 ergs @xmath4 s@xmath2 at maximum ( @xmath14 . the fluxes measured from our observations are listed in table 1 .
the 1996 asca observations caught an eclipse ingress and two pre - eclipse dips that did not have simultaneous uv coverage .
the 1994 asca observations covered more than a full binary orbit including an eclipse .
detailed studies of the asca observations will appear elsewhere ( boroson 1997b ) .
no flares occurred during either the 1994 or 1996 asca observations although each covered more than a binary orbit and flares are reported to occur roughly once a day .
it is possible that flares ( which last roughly 30 minutes ) occurred during the data gaps ( roughly 40 - 50 minutes ) due to earth occultation . a model involving x - ray heating of the disk and star as previously applied to her x-1 ( cheng , vrtilek , & raymond 1995 ) was adapted for use with lmc x-4 .
changes to the model from that described in cheng , vrtilek , & raymond are inclusion of both gravity darkening and limb darkening effects ( important for the massive companion in lmc x-4 ) as well as a different reddening curve towards the lmc ( nandy 1981 ) .
we use a distance to the source of 50 kpc and an e@xmath15 of 0.05 .
the companion star is modelled as a 20 @xmath6 o7iii star with an effective temperature of 35,000k and iue spectra of stars of varying temperature are used to determine the spectral shape at different points on the star and disk surface .
predictions of the model from the ultraviolet and optical b band are shown in figure 3 .
we are using archival iue data for comparison , hence we restrict ourselves to the region near c iv since the presence of strong geocoronal lyman alpha prevents accurate determination of the iue flux near n v. in addition our model , which is constructed to predict continuum emission , consistently overpredicts the flux near n v in this system ; this is likely due to the presence of a stellar wind not included in the model which results in strong absorption at n v. because the 30.25d period has a large uncertainty the long term phase of the iue observations can not be determined . the ultraviolet data including the current observations and all archival iue observations fall within extremes given by a lower limit to @xmath0 of 3.2 @xmath16 yr@xmath2 and an upper limit of 1.0 @xmath17 yr@xmath2 .
the best fit curve to the hst continuum data ( shown as filled triangles ) corresponds to an @xmath0 of 4.0 @xmath1810@xmath19 yr@xmath2 .
there is uncertainty in the flux level , both for the model and the data .
our hst continuum determination is based on a very narrow wavelength band since the observations were centered on lines and the total available bandpass was only 37@xmath20 for the spectral resolution we required ; the hst data are absolute flux calibrated to only 10% ; there is a difference in resolution between iue and ghrs . as for the model ,
the uncertainties in the reddening and stellar temperature can shift the flux .
the fluxes near c iv and n v show a similar flux vs. orbit trend .
the relative errors in flux are likely to be much smaller than the overall flux normalization .
the change in peak - to - peak amplitude of the optical flux from the 30-day low to the 30 day high predicted by the model is consistent with that observed ( ilovaisky 1984 ) .
figure 3b shows the extremes over the 30.25 day cycle for b magnitude variations and figure 3c shows the model predictions at the x - ray on and off states for a given @xmath0 reproducing the variations shown by ilovaisky 1984 in their figure 1 . during binary eclipse the total uv and optical flux is due to the companion star . during the time of maximum uv flux ( binary phase 0.75 ) for the highest @xmath0 the contribution from the disk is 8% and from the heated star 10% .
a comprehensive presentation of the fits to individual iue spectra and archival optical data will appear in a later paper ( preciado 1997 ) .
the x - ray spectra obtained by asca are consistent with a power law of photon index @xmath21 = [email protected] , a [email protected] kev blackbody , and an iron line at [email protected] kev with an equivalent width of 99@xmath525ev , typical for this source over the energy range observed .
the 2 - 10 kev flux is 2.9 @xmath3ergs s@xmath2 @xmath4 with about 10% changes from the mean during the period of observations .
the luminosity obtained from the @xmath0 measured with the uv continuum fits , using l@xmath22 = 0.5gm@xmath0/r , is consistent with the luminosity implied by this flux .
figure 4 shows the n v and c iv line profiles at four orbital phases each ; these are different for the two lines since they could not be observed simultaneously at this spectral resolution .
these two lines are the strongest in the uv spectrum and are known from iue observations to be strongly and weakly phase dependent , respectively .
the c iv profiles show little evolution from binary phase 0.15 - 0.49 . at binary phase 0.11 , the observation closest to the x - ray eclipse ( binary phase 0.0 ) , we see broad absorption in the n v doublets .
the maximum velocity at this phase ( 1150 km / s ) is reasonable for a terminal velocity for the wind however it is likely that by phase 0.11 some ionization has already taken place and terminal velocities of up to 1300 km / s are possible ( boroson 1997a ) . as we progress in phase towards seeing more of the x - ray source
we find that the broad absorption disappears and what remains are relatively narrow absorption profiles .
the width of the residual absorption left at phase 0.41 in n v is consistent with the width of the optical lines which hutchings attributed to the photosphere .
comparison with low - resolution iue archival data suggests that the variations are phase dependent effects , presumably arising in the geometry - dependent interaction between the x - ray source and stellar wind .
interstellar absorption lines from sii@xmath23 are visible in our exposures in the wavelength region near nv@xmath24 .
the velocities of these absorption lines ( heliocentric redshifts are consistent with zero within the errors of our measurement ) suggest that they arise in gas within the galaxy , and not in the lmc ( where we would expect velocities of the order 280 - 320 km s@xmath2 ; bomans 1996 ) .
the narrow civ absorption lines are superimposed on broader absorption features , which we interpret as arising in the photosphere of the o star in the lmc x-4 system . as a result of the overlap , the velocity and especially the equivalent widths of the interstellar civ lines are uncertain .
the heliocentric redshifts of the civ lines are also consistent with zero within the uncertainties of our measurement implying that the civ absorber is also within the galaxy .
there is no evidence for a civ absorber associated with lmc x-4 , which might be expected if the x - ray source photoionizes surrounding interstellar gas ( mccray , wright , and hatchett 1977 ) .
the nv line at @xmath25 shows what appears to be a narrow feature at the base of the absorption due to the blue doublet component .
however , observations at higher signal - to - noise ratio are needed to establish that this results from an interstellar feature and not merely a fluctuation in the photospheric line profile
. the pulse periods that we measured from asca data of 1994 ( [email protected] ) and 1996 ( [email protected] ) are consistent with spin - down over the past 10 years ( figure 5 ) . while earlier observations are scant they indicate that sometime between 10 and 20 years ago lmc x-4 went through a period of spin - up . such rapid , erratic , changes from spin - up to spin - down are expected from equilibrium rotators ( chakrabarty 1997 ) , systems in which the spin period equals the kepler period near the inner boundary of the accretion disk .
we searched for pulsations in the uv spectrum at the 13.5 second pulsar period using an analysis of variance ( anova ) method ( davies 1990 , 1991 ) .
this approach , which involves binning the data over trial periods , is well - suited to data with gaps or with non - uniform readout times . to detect pulsations from the accretion disk ,
we need to subtract the changes in light - travel time from the orbiting pulsar to the orbiting hst from the uniform readout times of the ghrs .
thus we report only the anova method and not a power - spectral search to find pulsations from material moving with the neutron star . from the 4 hst orbits we find a 5@xmath26 limit of 1.8% for the fractional peak - to - peak pulse amplitude in the continuum centered at 1240 surrounding the nv line . for the continuum surrounding civ ,
the limit is 2.7% .
our limits for pulsation in the p cygni absorption troughs are 12.4% in the nv line and 7% in the civ line .
a model that incorporates x - ray heating of the companion star and the accretion disk provides good fits to the continuum uv emission from lmc x-4 .
the value of @xmath0 derived from ghrs observations is consistent with that from x - ray flux measured during simultaneous observations .
owing to the size and temperature of the companion the major contribution to the ultraviolet and optical flux is the unheated companion star . at maximum light ( binary phases 0.25 and 0.75 )
the contribution of the unheated star is 82% ; heating of the primary contributes 10% and the disk 8% .
although we captured no flares during our observations the x - ray heating model can accomodate flares by using unsteady accretion from the stellar wind . during
flares the intensity can increase by factors of up to 20 for times ranging from @xmath720s to 45 minutes , resulting in super eddington luminosities ( dennerl 1989 ; woo 1995 ) .
the necessary @xmath0 is 3.2 @xmath27 yr@xmath2 and implies a range of a factor of 100 in @xmath0 for lmc x-4 .
by contrast the flaring states of z - source lmxbs such as sco x-1 and cyg x-2 , where the flaring state is also associated with super - eddington accretion , are produced with changes of only a factor of 2 - 4 in @xmath0 ; this is because sco x-1 and cyg x-2 are disk fed systems normally accreting at just under the eddington limit .
it is possible that the greater range and overall greater @xmath0 required for lmc x-4 is due to some intrinsic difference between x - ray binaries in the magellanic clouds compared with systems within our own galaxy .
the mean luminosity of those sources in the clouds which have massive ob - type companions similar to lmc x-4 is 50 times that of the counterparts in our galaxy ( van paradijs & mcclintock 1995 ) .
it has been suggested that this is due to the lower abundances of metals in the clouds .
one linking mechanism is the effect of x - ray heating of gas as it falls toward a compact object : this depends strongly on the atomic number z via the photoelectric cross section ( @xmath28 ) .
for spherical accretion , such heating can seriously impede the accretion flow and thereby reduce the limiting luminosity to a value far below the eddington limit ; so the lmc sources may be more luminous because their low - z accretion flow is less impeded by heating .
a second metallicity dependent effect takes place in wind - fed systems .
the accretion rate depends sensitively on the velocity of the stellar wind ; all available evidence indicates that the terminal wind velocity decreases with z. such behavior is predicted by successful theories of radiation - driven stellar winds ( kudrizki & hummer 1990 ; kudritzki 1991 ) . in this case
the lower metallicity in the lmc means a lower terminal velocity , a higher @xmath0 and a more luminous x - ray source .
the dramatic orbital variations shown by the n v profiles can be interpreted in terms of x - ray photoionization of the stellar wind from the companion .
the narrow absorption lines shown in figure 4d can be attributed to the surface of the companion star and the broad lines to the wind . in this scenario ,
the x - ray source ionizes nearly the entire stellar wind that is not in the shadow of the companion star , so that when the neutron star is in front of the normal star , the wind absorption disappears and mainly the photospheric absorption lines are visible .
it is only near phose 0.0 , i.e. during x - ray eclipse , that the broad lines that reveal the high wind velocities become visible .
detailed fits to the line profiles with a model that takes into account the structure of the wind and its influence on spectral features will be presented in a later paper ( boroson 1997a ) .
the pulse amplitude in the x - rays is lower in lmc x-4 ( the pulsed fraction of lmc x-4 is 10% in quiescence ) than in her x-1 and vela x-1 . in her x-1
the x - ray pulsed fraction is close to 50% and uv continuum pulsations have an amplitude of @xmath29% ( boroson 1996 ) . if lmc x-4 resembles her x-1 than we would expect uv continuum pulsations with an amplitude of @xmath30% , which is below our detection limit .
the narrow wavelength regions surrounding the lines provided too low a count rate to detect pulsations similar to those seen in other systems . in vela x-1 the pulsed fraction for 2 - 10 kev x - ray is 0.3 , and , while no continuum pulsations were detected in the uv using fos observations , 3% pulsations in the p - cygni lines were detected .
p cygni line pulsations , seen in vela x-1 , may be smeared by light travel time in lmc x-4 , owing to a more extended x - ray photoionized region .
the stis , recently installed on hst , extends in several significant ways the ultraviolet capabilities that became available with the ghrs : with the echelle grating it is possible to sample continuously a broad region ( 600@xmath20 ) of the spectrum at greater spectral resolution than with the ghrs .
observations with the stis that cover binary phase 0.0 can confirm our interpretation of the line profiles , provide improved measures of the continuum , and enable a more sensitive search for uv manifestation of the x - ray pulses . until recently ,
long - term approximately - periodic variability attributed to disk precession has been known for only three systems : her x-1 , lmc x-4 , and ss433 .
xte observations suggest similar long - term behavior in smc x-1 and cyg x-2 ( levine 1996 ; wijnands , kuulkers , & smale 1996 ) .
the interpretation of these periods in terms of disk precession has been questioned by several authors including , e.g. , kondo 1983 , who showed that precession of a disk controlled by the gravitational fields of the neutron and companion stars is untenable : any induced precession disppears rapidly because of differential precession within the disk .
recently iping and petterson ( 1990 ) proposed that the behavior attributed to precession is maintained by the influence of the x - ray emission on the structure of the disk .
accordingly , in these systems , the tendency of disks to undergo periodic changes in their orbital orientations is more appropriately termed radiation - driven warping . iping and petterson s primary result , that strong central illumination can maintain disk warping , has been obtained analytically by pringle ( 1996 ) and maloney , begelman , & pringle ( 1996 ) .
the success of our x - ray heated star plus warped disk model in fitting the optical , uv , and x - ray lightcurves for both the orbital and long - term periods of her x-1 and lmc x-4 supports an interpretation in terms of radiation - driven warping . 0.2 in based on observations with the nasa / esa _ hubble space telescope _ , obtained at the space telescope science institute , which is operated by the association of universities for research in astronomy , inc . , under nasa contract go-05874.01 - 94a .
we are grateful for quick - look results provided by the asm / rxte team .
sdv and bb were supported in part by nasa ( nag5 - 2532 , nagw-2685 ) , and nsf ( dge-9350074 ) .
* references * 0.1 in = -20pt bomans , d.j . ,
deboer , k.s .
, koornneef , j , & grebel , e.k . .
1996 , a&a 313 , 101 boroson , b. , mccray , r. , kallman , t.r . , &
nagase , f. 1994 , apj , 465 , 940 .
boroson , b. , mccray , r. , kallman , t. , & nagase , f. 1996a , apj , 465 , 940 .
boroson , b. , vrtilek , s.d . ,
mccray , r. , kallman , t. , & nagase , f. 1996b , apj , 473 1079 .
boroson , b. , vrtilek , s.d . ,
kallman , t. , & mccray , r. 1997a , in preparation .
boroson , b. , vrtilek , s.d .
, xu , c. , kelley , r. , & stahle , c. 1997b , in preparation .
chakrabarty , d. 1997 , apj , 474 , 414 .
cheng , f.h .
, vrtilek , s.d . , & raymond , j.c .
1995 , apj , 452 , 825 .
davies , s.r .
1990 , mnras , 244 , 93 .
davies , s.r .
1991 , mnras , 251 , 64p .
dennerl , k. 1989 , phd thesis at max planck institute for ... (d89 ) .
dennerl , k. 1989 , in 23rd eslab symp . ,
j. hunt & b. battrick ( noordwijk : esa - estec ) , 39 .
dennerl , k. 1992 , in lecture notes in physics , 416 , new aspects of magellanic cloud research , ed .
b. baschek , g. klare , & j. lequeux ( berlin : springer ) , 74 .
heemskerk , m.h.m . , & van paradijs , j. 1989 , a&a , 223 , 154 .
ilovaisky , s.a .
, chevalier , c. , motch , c. , pakull , m. , van paradijs , j. , & lub , j. 1984 , a&a , 140 , 251 .
iping , r.c .
, & petterson , j.a . 1990 , a&a , 239 , 221 .
kelley , r.l . ,
jernigan , j.g . , levine , a. , petro , l.d .
, & rappaport , s. 1983 , apj , 264 , 568 .
kondo , y. , van flandern , t.c .
, & wolff , c.l .
1983 , apj , 273 , 716 .
lang , f. l. , 1981 , apj , 246 , l21 .
levine , a. m. , bradt , h. , cui , w. , jernigan , j. g. , morgan , e. h. , remillard , r. , shirey , r. e. , & smith , d. a. 1996 , apj , 469 , l33 .
levine , a. , rappaport , s. , putney , a. , corbet , r. , & nagase , f. 1991 , apj , 381 , 101 .
li , f. 1978 , nature , 271 , 38 .
maloney , p. r. , begelman , m. c. , & pringle , j. e. 1996 , apj , 472 , 582 .
mccray , r. , wright , c. , and hatchett , s. 1977 , apj , 211 , l29 .
nandy , k , morgan , d.h . , willis , a.j . ,
wilson , r. , & gondhalekar , p.m. 1981 , mnras , 196 , 955 .
preciado , m. , boroson , b. , vrtilek , s.d . , & raymond , j.c .
1997 , in preparation .
pringle , j.e .
1996 , mnras , 281 , 357 .
tananbaum , h. , et al .
1972 , apj 174 , l143 .
van der klis , m. 1982 , a&a , 106 , 339 .
van paradijs , j. & mcclintock , j.e . , 1995 , in x - ray binaries , eds .
lewin , j. van paradijs , & e.p.j .
van den heuvel , cambridge university press : cambridge .
wijnands , r. a. d. , kuulkers , e. , & smale , a. p. 1996
, apj , 473 , l45 .
wilson , r.b . ,
finger , m.h . ,
pendleton , g.n . ,
brigg , m. , & bildsten , l. 1994 , in the evolution of x - ray binaries , eds .
holt , & c. s. day , aip press : new york .
woo , j.w . ,
clark , g.w .
, & levine , a.m. 1995 , apj , 449 , 880 .
woo , j.w . ,
clark , g.w . ,
levine , a.m. , corbet , r.h .
, & nagase , f. 1996 apj 467 , 811 * figure captions * 0.1 in * figure 1 . * ( a ) one day averages of the flux observed from lmc x-4 with the all sky monitor on board the rossi x - ray timing explorer .
( quick - look results were provided by the asm / rxte team . )
the arrow indicates the time of the simultaneous hst / asca observations , and the horizontal bar represents one 30.25 day interval .
( b ) the data from ( a ) ( light lines ) with the lightcurve from ( c ) ( dark lines ) superposed .
( c ) the data from ( a ) folded with the ephemeris for the long - term period provided by dennerl 1992 ( p@xmath31 = 30.25+/-.03d and @xmath32 = jd 2,448,226.0 ) .
arrow indicates the phase of the simultaneous hst / asca observations .
the errors are 1@xmath26 from counting statistics .
0.1 in * figure 2 . *
( a ) the exposure times of the hst observations which covered the binary phases 0.08 - 0.49 ( start time = jd 2,450,228.44 ) .
the first observation is n v and is alternated with c iv .
( b ) asca lightcurves ( 0.5 - 10 kev ) : observations covered the binary phases 0.75 - 1.84 ( start time = jd 2,450,227.92 ) determined from the ephemeris of woo ( 1996 ) with p@xmath33 = 1.40840249@xmath34d and @xmath35 = jd 2,446,[email protected] . 0.1 in * figure 3 . * ( a ) average uv continuum flux near c iv _ vs. _ binary phase .
the solid squares represent the ghrs observations .
the open triangles are archival iue data .
the dashed line represents a 30.25 day phase when the x - ray flux is in the low state ; the solid line a 30.25d phase during an x - ray high state ; and the dotted line represents the best fit continuum for the ghrs observations .
( b ) the b magnitude predictions of the model for the three states depicted in 3a .
( c ) the b magnitude predictions for a given @xmath0 during the x - ray on and off states .
( b ) and ( c ) should be compared with figure 1 in ilovaisky 1984 . 0.1 in * figure 4 . * the ghrs observations of n v and c iv . 0.1 in * figure 5 . * the pulse period history of lmc x-4 . * table 1 : log of ghrs observations * [ cols="^,^,^,^,^,^ " , ] @xmath36using the ephemeris of woo ( 1996 ) with the nth eclipse in jd given by a0 + a1*n + a2*n@xmath37 where a0 = jd 2,446,[email protected] , a1 = 1.40840249@xmath38 days , and a2 = -1.45e-9 . @xmath35
= a0 and @xmath39 = a1 + a2*n .
+ @xmath37flux near n v ( [ 1226 - 1234@xmath20]+[1255 - 1259@xmath20 ] ) .
+ @xmath41flux near c iv ( [ 1532 - 1544@xmath20]+[1556 - 1567@xmath20 ] ) . | we present first results from simultaneous ultraviolet ( hst / ghrs ) and x - ray ( asca ) observations of the 13.5s pulsar lmc x-4 taken in 1996 may .
the asca observations covered 1.12 binary orbits ( 1.58 days ) and the hst observations were centered on this for roughly 0.4 orbital phase coverage ( 0.56 days ) .
the ghrs data are the highest resolution ( both temporal and spectral ) ultraviolet spectra ever taken of lmc x-4 . with generally - accepted parameters for the source ,
fits to the uv continuum using a model that incorporates x - ray heating of the companion star and the accretion disk yields a mass accretion rate @xmath0 = 4.0 @xmath1 yr@xmath2 ; the x - ray luminosity implied by this value is consistent with the x - ray flux measured during simultaneous observations ( 3.2 @xmath3 ergs @xmath4 s@xmath2 ) .
the model accurately predicts observed b magnitude and ultraviolet variations over both orbital and long - term periods .
the ultraviolet p - cygni lines show dramatic changes with orbital phase with strong broad absorption near x - ray eclipse and narrow absorption when the x - ray source is in the line - of - sight .
we interpret this as a result of x - ray photoionization of the stellar wind ; when the neutron star is in front of the normal star , the wind absorption disappears and mainly the photospheric absorption lines are visible .
the x - ray pulse period measured during our observations , [email protected] , is consistent with steady spin - down over the past 10 years .
no pulsations were detected in the ultraviolet observations with upper limits to the pulsed fraction around nv and civ of 1.8% and 2.7% in the continuum and 12.4% and 7% in the absorption troughs . 9.0 in 6.4 in -.50 in + 0.2 in plus 3pt minus 2pt |
near - field wireless power transfer ( wpt ) has drawn significant interests recently due to its high efficiency for delivering power to electric loads without the need for any wire .
inductive coupling ( ic ) @xcite is the conventional method to realize near - field wpt for short - range applications typically in a couple of centimeters .
the wireless power consortium ( wpc ) that developed the `` qi '' standard @xcite is the main industrial organization for commercializing wireless charging based on ic .
recently , magnetic resonant coupling ( mrc ) @xcite has been applied to significantly enhance the transfer efficiency as well as range for wpt compared to ic , thus opening a broader avenue for practical applications . in mrc
enabled wpt ( mrc - wpt ) , compensators each being a capacitor of variable capacity are embedded in the electric circuits of the power transmitter and receiver to tune their oscillating frequencies to be the same as the frequency of the input voltage / current source to achieve resonance . alternatively , resonators each of which constitutes a simple rlc circuit resonating at the source frequency can be deployed in close proximity of the coils of the off - resonance power transmitter and receiver to help efficiently transfer power between them . with mrc , the total reactive power consumption in the system is effectively minimized due to resonance and thus high power transfer efficiency is achieved over longer distance than ic @xcite .
the preliminary experiments in @xcite show that mrc enables a single transmitter to transfer @xmath0 watts of power wirelessly with @xmath1@xmath2 efficiency to a single receiver at a distance of about @xmath3 meters . formed after the merging between alliance for wireless power ( aw4p ) that developed the `` rezence '' specification and power matters alliance ( pma ) , airfuel alliance is the main industrial organization for promoting wireless charging based on mrc @xcite .
the rezence specification advocates a superior charging range , the capability of charging multiple devices at the same time , and the use of two - way communication via e.g. bluetooth between the charger unit and devices for real - time charging control .
these features make rezence and its future extensions a promising technology for wireless charging systems . in the current rezence specification , one transmitter with a single coil
is used in the power transmitting unit , i.e. , only the single - input multiple - output ( simo ) mrc - wpt is considered , as shown in fig .
[ fig : wpttable](a ) .
although this centralized wpt system performs well when the receivers are all sufficiently close to as well as perfectly aligned with the transmitter , the power delivered to a receiver decays significantly as it moves more distant away from the transmitter or misaligns with its orientation .
this thus motivates distributed wpt where the single centralized transmitter coil is divided into multiple coils each with smaller size ( radius ) and these coils are placed in different locations to cover a given region , as shown in fig .
[ fig : wpttable](b ) . by coordinating the transmissions of multiple coils via jointly allocating their source currents , in @xcite
it has been shown that their induced magnetic fields can be directed more efficiently toward one or more receivers at the same time , thus achieving a _
magnetic beamforming _ gain in a manner analogous to multi - antenna beamforming in the far - field wireless communication and power transfer @xcite .
in addition , distributed wpt significantly shortens the distance from each receiver to its nearest transmitter(s ) compared to centralized wpt , thus achieves more uniform charging performance in the region .
the optimal magnetic beamforming design in multiple - input single - output ( miso ) mrc - wpt systems has been investigated in @xcite .
specifically , @xcite has formulated a convex optimization problem to jointly optimize the source currents at all transmitters to maximize the wpt efficiency subject to that the deliverable power to the receiver load is fixed . on the other hand
, @xcite has jointly optimized the transmitter currents to maximize the deliverable power to a single load by considering a sum - power constraint for all transmitters as well as practical peak voltage and current constraints at each individual transmitter . recently
, selective wpt has also been proposed for simo mrc - wpt systems in @xcite .
this technique delivers power to only one selected receiver ( i.e. , receiver selection ) at each time to eliminate the magnetic cross - coupling effect among different receivers and hence achieve more balanced power transfer to them , assuming that their natural frequencies are set well separated from each other .
alternatively , @xcite has proposed to jointly optimize the load resistance of all receivers in a simo mrc - wpt system to manage the magnetic cross - coupling effect and further alleviate the near - far issue by delivering balanced power to all receivers regardless of their distances to the power transmitter .
the selective wpt technique has also been utilized in miso mrc - wpt systems where one transmitter is selected at each time ( i.e. , transmitter selection ) to deliver wireless power to a single receiver @xcite . in general
, selective wpt requires a simpler control mechanism than magnetic beamforming , while its performance is also limited since only one pair of transmitter and receiver is allowed for power transfer at each time .
in contrast , magnetic beamforming enables multiple transmitters to send power to one or more receivers simultaneously by properly assigning the currents in the transmitters and/or resistance values in the receivers , thus in general achieving better performance than transmitter / receiver selection .
the studies in @xcite have shown promising directions to improve the efficiency as well as performance fairness in mrc - wpt systems , but all of which have assumed that the transmitters and receivers are at given locations in a target region . in practice , a wireless device ( e.g. , mobile phone ) is desired to be freely located in any position in the region ( e.g. , above a charging table ) when it is being charged , for more convenience of its user . in this case , given fixed locations of the transmitters in a msio mrc - wpt system , the deliverable power at a single receiver can fluctuate significantly over its locations .
such power fluctuation degrades the quality of service , since the power requirement of the receiver load may not be satisfied at all locations in the region , even when magnetic beamforming is applied to dynamically adjust the magnetic fields according to the instantaneous location of the receiver .
this thus motivates our work in this paper to optimize the transmitter locations and magnetic beamforming jointly in a miso mrc - wpt system to achieve the maximum and yet uniformly deliverable power in the region .
the main results of this paper are summarized as follows : * first , we formulate the magnetic beamforming problem for a miso mrc - wpt system with distributed transmitters to maximize the deliverable power to a single receiver subject to a given transmitters sum - power constraint , by assuming that the power transmitters and receiver are all at fixed locations .
we derive the closed - form solution to the magnetic beamforming problem as a function of the mutual inductances between the transmitters and the receiver .
our solution shows that the optimal current allocated to each transmitter is proportional to the mutual inductance between its coil and that of the receiver .
for the special case when the transmitters are sufficiently separated from each other , we show that the optimal magnetic beamforming reduces to the simple transmitter selection scheme @xcite where all power is allocated to the single transmitter that has the highest mutual inductance with the receiver . * to demonstrate the performance gain of magnetic beamforming , we compare it to an uncoordinated wpt system with equal current allocation over all transmitters @xcite , as well as the transmitter selection scheme .
furthermore , we compare the performance of distributed wpt with magnetic beamforming versus centralized wpt subject to the same total size of transmitter coils .
* next , with the optimal magnetic beamforming solution derived , we formulate the node placement problem to jointly optimize the transmitter locations to maximize the minimum deliverable power to the receiver over a given one - dimensional ( 1d ) target region , i.e. , a line of finite length where the receiver can be located in any point in the line . although the formulated problem is non - convex , we propose an iterative algorithm for solving it approximately by leveraging the fact that the transmitters should be symmetrically located over the mid - point of the target line to maximize the minimum deliverable power .
we present extensive simulation results to verify the effectiveness of our proposed transmitter location optimization algorithm in improving both the minimum deliverable power as well as the average deliverable power over the target line as compared to a heuristic design that uniformly locates the transmitters .
* at last , we extend the node placement problem to the general two - dimensional ( 2d ) region case , i.e. , a disk in 2d with a finite radius . using an example of five transmitters ,
we show that the design approach for the 1d case can be similarly applied to obtain the optimal locations of the transmitters under the 2d setup with magnetic beamforming to maximize the minimum deliverable power over the target disk by exploiting its circular symmetry .
the rest of this paper is organized as follows .
section ii introduces the system model .
section iii formulates the magnetic beamforming problem and presents its optimal solution .
section iv formulates the node placement problem for the 1d target region case , and presents an iterative algorithm for solving it .
section v presents simulation results for the 1d case .
section vi extends the node placement problem to the 2d target region case with an example of five transmitters .
finally , we conclude the paper in section vii .
in this paper , we consider a miso mrc - wpt system with @xmath4 identical single - coil transmitters , indexed by @xmath5 , @xmath6 , and a single - coil receiver , indexed by @xmath7 for convenience .
it is assumed that all the transmitters and receiver are each equipped with a bluetooth communication module to enable information exchange among them to achieve coordinated wpt @xcite .
each transmitter @xmath5 is connected to a stable energy source supplying sinusoidal voltage over time given by @xmath8 , where @xmath9 is a complex number denoting the steady state voltage in phasor form and @xmath10 denotes its angular frequency .
note that @xmath11 represents the real part of a complex number . on the other hand
, the receiver is connected to an electric load , e.g. , the battery of a mobile phone .
let @xmath12 , where @xmath13 , with @xmath14 , denotes the steady state current in transmitter @xmath5 .
this current produces a time - varying magnetic flux in the transmitter coil , which passes through the receiver coil and induces time - varying current in it .
we denote @xmath15 , with @xmath16 , as the steady state current at the receiver . for the time being , we consider the case of 1d region for wpt , which is a straight line of finite length @xmath17 , with @xmath18 .
specifically , the receiver can move horizontally along the line with its x - coordinate satisfying @xmath19 , which lies in the @xmath20 plane with @xmath21 , @xmath22 , as shown in fig .
[ fig : miso - linear ] . note that @xmath23 denotes the absolute value of a real / complex number .
the transmitters are installed at fixed locations along the line that is in parallel with the target line in the @xmath24 plane with @xmath25 .
this 1d target line model is mainly used for the purpose of exposition , while it may be also applicable to practical scenarios such as a magnetic train that moves over segments of tracks each equipped with a number of distributed wireless power transmitters for charging the train .
let @xmath26 with @xmath27 ( @xmath28 with @xmath29 ) denote the location of transmitter @xmath5 ( receiver ) over the x - axis . in this paper , we consider that @xmath26 s are symmetric over @xmath30 . that such symmetric structure of the transmitters maximizes the minimum power deliverable to the receiver over the target line .
] hence , we consider the following two cases for the symmetric deployment of the transmitters .
* case @xmath31 : @xmath32 is even . in this case , let @xmath33 and we set @xmath34 , with @xmath35 , @xmath36 , as shown in fig .
[ fig : miso - linear](a ) .
* case @xmath3 : @xmath32 is odd . in this case , let @xmath37 and we set @xmath34 , with @xmath35 , @xmath36 , and @xmath38 , as shown in fig .
[ fig : miso - linear](b ) .
let @xmath39 ( @xmath40 ) , @xmath41 ( @xmath42 ) , and @xmath43 ( @xmath44 ) denote the parasitic resistance , the self - inductance , and the capacity of the compensator in each transmitter ( receiver ) , respectively .
denote @xmath45 as the resistance of the load at the receiver .
accordingly , we use @xmath46 to represent the total ohmic resistance of the receiver . by assuming that the coil of each of the transmitters as well as the receiver consists of multiple closely winded turns of round - shaped wire , we obtain @xmath47 and @xmath48 , where @xmath49 ( @xmath50 ) , @xmath51 ( @xmath52 ) , @xmath53 ( @xmath54 ) , and @xmath55 ( @xmath56 ) are the average radius of the coil of each transmitter ( receiver ) , the radius of the wire used to make the coil , the ohmic resistivity of the wire , and the number of turns of the coil , respectively . furthermore , we obtain @xmath57 and @xmath58 , where @xmath59n / a@xmath60 is the magnetic permeability of the air @xcite . the capacities of compensators are then chosen such that the natural frequencies of the transmitters and receiver become the same as the source angular frequency @xmath61 @xcite , i.e. , we set @xmath62 and @xmath63 .
let @xmath64 and @xmath65 be real numbers denoting the mutual inductance between the coils of transmitters @xmath5 and @xmath66 , with @xmath67 , as well as that between transmitter @xmath5 and the receiver , respectively . based on the so - called conway s mutual inductance formula @xcite
, we obtain @xmath68 where @xmath69 , @xmath70 , and @xmath71 is the bessel function of the first kind of order @xmath72 with @xmath73 denoting the the factorial of a positive integer .
the integration terms in ( [ eq : h_nk ] ) and ( [ eq : h_n0 ] ) can be computed numerically , while there are no closed - form analytical expressions for them . in practice
, the transmitters and receiver commonly use small coils for wpt ; therefore , @xmath65 in ( [ eq : h_n0 ] ) can be simplified as follows .
[ lemma : mutualinductance ] if @xmath74 , we have @xmath75 where @xmath76 is a constant with the given coil parameters .
please see appendix a. to validate the accuracy of the proposed approximation in ( [ eq : hn0-simplified ] ) , we consider the following setup . particularly , we consider case @xmath3 in fig . [
fig : miso - linear](b ) with @xmath77 identical transmitters , @xmath78 m , and variable @xmath79 , where the physical and electrical characteristics of the coils in the transmitters and receiver are given in tables [ tab : physical - charac ] and [ tab : electrical - charac ] ( see section [ sec : numerical example ] ) , respectively .
we assume that the transmitters are uniformly located over @xmath80 m , with @xmath81 m , @xmath82 m , and @xmath83
[ fig : mutual_inductance_veification](a ) and [ fig : mutual_inductance_veification](b ) compare the actual and approximated values of the mutual inductance between transmitter @xmath31 and the receiver , @xmath84 , versus the receiver s x - coordinate @xmath28 under fixed height @xmath85 cm and @xmath86 cm , respectively .
it is observed that the approximation is tight in general ; whereas there are small discrepancies at @xmath87 .
it is also observed that the discrepancies decrease when @xmath79 increases .
note that the similar result can be obtained for the mutual inductance between other transmitters and the receiver , while the peak value of the mutual inductance shifts over the x - axis accordingly , i.e. , it moves from @xmath88 to @xmath89 when transmitter @xmath5 is considered instead of transmitter @xmath31 . in this paper , for simplicity we will use the approximation in ( [ eq : hn0-simplified ] ) to design the node placement for the transmitters in sections [ sec : node placement optimization ] and [ sec : node placement optimization 2d ] , while the actual value in ( [ eq : h_n0 ] ) is used for all simulations to achieve the best accuracy .
[ t ! ] by applying kirchhoff s circuit laws to the electric circuits of the transmitters and receiver in our considered mrc - wpt system shown in fig .
[ fig : electric - circuit ] , we obtain @xmath90 accordingly , the average power drawn from the energy source at transmitter @xmath5 , denoted by @xmath91 , and that delivered to the load at the receiver , denoted by @xmath92 , are obtained as @xmath93 where @xmath94 denotes the conjugate of a complex number . furthermore , with the result given in ( [ eq : p_n ] ) , the sum - power drawn from all transmitters is obtained as @xmath95 from ( [ eq : p_n ] ) and ( [ eq : p_sum ] ) , it follows that the power consumption of each individual transmitter depends on all the mutual inductances between the transmitters and the receiver , @xmath65 s , as well as those between any pair of transmitters , @xmath96 s , while their sum - power consumed depends on @xmath65 s only . from ( [ eq : p_0 ] ) and ( [ eq : p_sum ] ) , it also follows that the real - part currents @xmath97 s and the imaginary - part currents @xmath98 s contribute in the same way to @xmath92 or @xmath99 .
consequently , in this paper , we set @xmath100 , @xmath101 , and focus on designing @xmath97 s without loss of generality .
moreover , due to the fact that each @xmath65 is a function of both @xmath28 and @xmath26 with given @xmath79 ( see , e.g. , ( [ eq : hn0-simplified ] ) ) , we re - express @xmath92 and @xmath99 in ( [ eq : p_0 ] ) and ( [ eq : p_sum ] ) as functions of @xmath28 , @xmath26 s , and @xmath97 s as @xmath102 next , we introduce four metrics to evaluate the performance of the mrc - wpt system considered in this paper , which are the average value , the minimum value , the maximum value , and the min - max ratio of the deliverable power to the receiver load over the target line ( or target region in general ) , defined as @xmath103 note that both the transmitter currents @xmath97 s and the transmitter locations @xmath26 s can have an influence on each of the above performance metrics for the mrc - wpt system ; therefore , we need to design them jointly to optimize each corresponding performance in general . in practice , it is desirable to have both large @xmath104 and @xmath105 to maximize the wpt efficiency , and yet have acceptably high @xmath106 and @xmath107 to achieve uniform performance over the target region . however , in general , there are trade - offs in achieving these objectives at the same time , e.g. , maximizing @xmath105 versus @xmath106 . in the rest of this paper , we first design the magnetic beamforming via adjusting @xmath97 s by assuming fixed locations of the transmitters and receiver ( @xmath26 s and @xmath28 ) to maximize the deliverable power subject to a given sum - power constraint for all transmitters .
next , with the obtained optimal magnetic beamforming solution , we optimize the transmitter locations @xmath26 s to maximize the minimum power deliverable to the receiver over the target region , i.e. , @xmath106 given in ( [ eq : p0min ] ) , for both the cases of 1d and 2d target regions , respectively .
in this section , we first present the results on the magnetic beamforming optimization .
we then use a numerical example to demonstrate the performance advantage of optimal distributed magnetic beamforming .
assume that @xmath26 s and @xmath28 are given , and hence the mutual inductance values @xmath65 s are known .
we formulate the magnetic beamforming problem for designing the transmitter currents @xmath97 s to maximize the deliverable power to the receiver load , @xmath108 given in ( [ eq : fun_p_0 ] ) , subject to a maximum sum - power constraint at all transmitters , denoted by @xmath109 , as follows .
@xmath110 ( p1 ) can be shown to be a non - convex quadratically constrained quadratic programming ( qcqp ) problem @xcite , since its objective is to maximize a convex quadratic function in ( [ eq : p1_obj ] ) . however , we obtain the optimal solution to ( p1 ) in the following proposition . [
proposition : optimal_current ] the optimal solution to ( p1 ) is given by @xmath111 , @xmath101 , with @xmath112 please see appendix b. from ( [ eq : u_n ] ) , it follows that the current allocated to each transmitter @xmath5 is proportional to the mutual inductance between its coil and that of the receiver , @xmath65 .
moreover , it can be seen that when there exists an @xmath5 such that @xmath113 , @xmath114 , then @xmath115 .
this means that all transmit power is allocated to transmitter @xmath5 which has the dominant mutual inductance magnitude with the receiver ( e.g. , when the receiver is directly above transmitter @xmath5 and more far apart from its adjacent transmitters ) , i.e. , the transmitter selection technique @xcite is optimal . to implement the optimal magnetic beamforming solution in practice , each transmitter @xmath5 needs to estimate the mutual inductance between its coil and that of the receiver , @xmath65 , in real time @xcite , and send it to a central controller via e.g. the bluetooth communication considered in the rezence specification @xcite . given the information received from all transmitters , the central controller computes the optimal transmitter currents @xmath111 s and sends them to the individual transmitters for implementing distributed magnetic beamforming . as shown in fig .
[ fig : electric - circuit ] , it is more convenient to use voltage source than current source at the transmitters in practice .
from ( [ eq : in ] ) , we can easily obtain the optimal source voltages @xmath116 s that generate the optimal currents @xmath117 s for practical implementation .
next , by substituting @xmath118 , @xmath101 , in ( [ eq : fun_p_0 ] ) , the power delivered to the load with optimal magnetic beamforming is given by @xmath119 from ( [ eq : p_0^ * ] ) , it follows that the deliverable power with optimal magnetic beamforming is a function of @xmath120 s , hence invariant to the signs of individual @xmath65 s .
this is expected since optimal magnetic beamforming constructively combines the magnetic flux generated by individual transmitters at the receiver .
@xmath49/@xmath50 ( cm ) & number of turns @xmath121/@xmath122 & material of wire & radius of wire @xmath51/@xmath52 ( mm ) & resistivity of wire @xmath53/@xmath54 ( @xmath123/m ) + transmitter & @xmath124 & @xmath125 & copper & 0.1 & @xmath126 + receiver & @xmath127 & @xmath128 & copper & 0.1 & @xmath126 + [ tab : physical - charac ] @xmath129/@xmath130 ( @xmath123 ) & self - inductance @xmath131/@xmath132 ( mh ) & series compensator @xmath133/@xmath134 ( ff ) + transmitter & @xmath135 & @xmath136 & @xmath137 + receiver & @xmath138 & @xmath139 & @xmath140 + [ tab : electrical - charac ] we consider an mrc - wpt system with @xmath141 identical transmitters and a single receiver that is connected a load with resistance @xmath142 .
the physical and electrical characteristics of coils in the transmitters and receiver are given in tables [ tab : physical - charac ] and [ tab : electrical - charac ] , respectively .
we set @xmath85 cm , @xmath143 m , @xmath144rad / sec @xcite , and @xmath145w . in this example
, we assume that transmitters are uniformly located over @xmath80 m , with @xmath146 m , @xmath147 m , and @xmath148 . for performance comparison
, we also consider the uncoordinated wpt with equal current allocation over all transmitters , as well as the transmitter selection technique which only selects the transmitter with the largest mutual inductance ( squared ) value with the receiver for wpt with the full transmit power , @xmath149 .
[ t ! ] [ t ! ] fig .
[ fig : p0_3tx_x0 ] compares the deliverable load power @xmath92 given in ( [ eq : fun_p_0 ] ) versus the receiver location @xmath28 by three schemes : equal ( transmitter ) current with uniform ( transmitter ) location ( ecul ) , optimal ( transmitter ) current with uniform ( transmitter ) location ( ocul ) , and transmitter selection with uniform ( transmitter ) location ( tsul ) .
it is observed that ecul in general delivers higher power to the load than ocul and tsul , and also achieves a larger minimum power over the receiver location @xmath28 .
it is also observed that the three schemes all tend to deliver more power to the load when the receiver is in close proximity of one of the transmitters at @xmath87 , @xmath150 m , and @xmath151 m .
furthermore , it is observed that tsul performs quite close to ocul except in the middle areas between any two adjacent transmitters , where the minimum deliverable power occurs .
this observation is expected since when the receiver is in the middle of any two adjacent transmitters , optimal magnetic beamforming with both transmitters delivering power to the receiver load achieves a more pronounced combining gain as compared to the transmitter selection with only one of the two transmitters selected for wpt .
next , we show the performance of centralized wpt , where a single transmitter is located at @xmath152 which sends wireless power to a receiver moving along the target line . for this centralized transmitter case , we set @xmath153 turns and @xmath154 mm , where the radius of its coil is @xmath141 times larger than that of each transmitter in the case of distributed wpt for fair comparison .
[ fig : p0_1tx_x0 ] plots @xmath92 for centralized wpt versus @xmath28 .
it is observed that the deliverable power to the load is zero at @xmath155 m , while its global and local maximums occur at @xmath87 and @xmath156 m , respectively .
note that from ( [ eq : h_n0 ] ) , it follows @xmath157 at @xmath155 m ; as a result , by setting @xmath157 in ( [ eq : p_0 ] ) , we have @xmath158 , regardless of the transmit current . the details of performance comparison between distributed versus centralized wpt in terms of the four metrics introduced in section [ sec : system - setup ] ( see ( [ eq : p0ave])([eq : min - max - ratio ] ) ) are given in table [ tab : quantitivecomparision ] .
@xmath106 ( w ) & @xmath105 ( w ) & @xmath107 ( % ) + & ocul & @xmath159 & @xmath160 & @xmath161 & @xmath162 + & ecul & @xmath163 & @xmath164 & @xmath165 & @xmath166 + & tsul & @xmath167 & @xmath168 & @xmath161 & @xmath169 + & @xmath170 & @xmath7 & @xmath171 & @xmath7 + it is observed that distributed wpt with ocul and tsul achieves similar @xmath105 and slightly better @xmath172 compared to centralized wpt . however , in terms of @xmath106 and the min - max load power ratio @xmath107 , distributed wpt achieves significant improvement over centralized wpt .
although distributed wpt with ocul achieves the highest @xmath107 of @xmath173 , it is still far from the ideal uniform power profile with @xmath174 . to further improve this performance , in the next section
, we will formulate the node placement problem to design the transmitter locations to maximize the minimum deliverable power to the load over the target line jointly with the optimal magnetic beamforming .
it is worth pointing out that the transmitter locations can be optimized with magnetic beamforming to improve other performance metrics such as maximizing the average load power , maximizing the min - max ratio of the load power , etc . , which will lead to different optimal transmitter locations in general .
we leave other possible node placement problem formulations to our future work .
in this section , we first present the node placement optimization problem for the 1d target region case , and then propose an iterative algorithm to solve it . let @xmath175 denote the minimum deliverable power to the load over the target line in the 1d case .
the node placement problem is formulated as @xmath176 with @xmath177 given in ( [ eq : p_0^ * ] ) .
first , it can be easily shown by contradiction that the optimal solution @xmath26 s to ( p2 ) must be symmetric over @xmath30 , as shown in fig .
[ fig : miso - linear ] . with symmetric transmitter locations
, then it follows that the load power profile is also symmetric over @xmath30 ; as a result , the constraint ( [ eq : p2_c1 ] ) only needs to be considered over @xmath178 . with the above observations , we simplify ( p2 ) for the cases of even and odd @xmath32 , respectively , as follows .
when @xmath32 is even , we have @xmath179 where @xmath180 is defined as @xmath181 note that since it can be easily verified that the constraint in ( [ eq : p2_c1 ] ) is infeasible regardless of @xmath28 when @xmath182 , we define @xmath183 for @xmath182 in ( [ eq : const1:p2-e ] ) for convenience . on the other hand ,
when @xmath32 is odd , we have @xmath184 ( p2@xmath185evenn ) and ( p2@xmath185oddn ) are both non - convex optimization problems due to the constraints in ( [ eq : const1:p2-e ] ) and ( [ eq : const1:p2-o ] ) , respectively .
thus , it is difficult to solve them optimally . in the following ,
we propose an iterative algorithm to obtain approximate solutions for them . in this subsection , we focus on the problem ( p2@xmath185evenn ) for the even @xmath32 case , while the proposed algorithm can be similarly applied for ( p2@xmath185oddn ) in the odd @xmath32 case . in ( p2@xmath185evenn ) , we need to find the largest @xmath175 , @xmath186 , under which the problem is feasible over all possible transmitter ( one - sided ) locations @xmath187 s . to this end , we apply the bisection method to find the largest @xmath175 by using the fact that if ( p2@xmath185evenn ) is not feasible for a certain @xmath188 , @xmath189 , then it can not be feasible for @xmath190 . similarly ,
if ( p2@xmath185evenn ) is feasible for @xmath188 , then it must be feasible for @xmath191 .
the detail of our proposed algorithm is given in the following .
initialize @xmath192 and @xmath193 . at each iteration , we first set @xmath194 , and test the feasibility of ( p2@xmath185evenn ) given @xmath175 by considering the following feasibility problem .
@xmath195 if ( p2f@xmath185evenn ) is feasible , we save its solution as @xmath196 , @xmath36 , and update @xmath197 to search for larger values of @xmath175 in the next iteration .
otherwise , if ( p2f@xmath185evenn ) is infeasible , we update @xmath198 to search for smaller values of @xmath175 in the next iteration .
we stop the search when @xmath199 , where @xmath200 is a small constant controlling the algorithm accuracy .
it can be easily shown that the algorithm converges after about @xmath201 iterations .
after convergence , we return @xmath196 as the solution to ( p2@xmath185evenn ) , and set @xmath202 , @xmath101 , as the solution to ( p2 ) for the even @xmath32 case .
next , we focus on solving the feasibility problem ( p2f@xmath185evenn ) at each iteration . since ( p2f@xmath185evenn ) is non - convex , we use the following gradient based method to search for a feasible solution to this problem in an iterative manner .
initialize @xmath203 , @xmath36 . at each iteration @xmath204 , given @xmath187 s
, we check whether the constraint ( [ eq : const1:p2-e ] ) holds or not .
if the constraint holds , we return @xmath187 s as a feasible solution to ( p2f@xmath185evenn ) and stop the search ; otherwise , we update @xmath187 s as follows .
first , we find @xmath205 , which can be numerically obtained with given @xmath187 s .
next , by setting @xmath206 in the left hand side ( lhs ) of ( [ eq : const1:p2-e ] ) , which gives the minimum deliverable power to the load @xmath92 over the target line , @xmath106 , with the given transmitter locations @xmath187 s , the derivative of @xmath106 with respect to each @xmath187 is computed as @xmath207 accordingly , we set @xmath208 if @xmath209 , or @xmath210 otherwise , @xmath36 , with @xmath211 denoting a small step size .
it can be easily verified that the above update helps increase @xmath106 if @xmath212 is chosen to be sufficiently small .
we repeat the above procedure for a maximum number of iterations , denoted by @xmath213 , after which we stop the search and return that ( p2f@xmath185evenn ) is infeasible since the constraint ( [ eq : const1:p2-e ] ) still does not hold with all @xmath187 s derived . in practice ,
the performance of the gradient - based search for the feasible solution to ( p2f@xmath185evenn ) depends on the initial values of @xmath187 s as the search in general converges to a local maximum of the lhs function of ( [ eq : const1:p2-e ] ) . to improve the accuracy of the search ,
if it fails to find a feasible solution to ( p2f@xmath185evenn ) after @xmath214 iterations , then we repeat the search with a new initial point given by @xmath215 , @xmath36 , with randomly generated @xmath216 which is uniformly distributed over @xmath217 $ ] .
the maximum number for the set of randomly generated initial points is limited by @xmath218 , and we decide ( p2f@xmath185evenn ) is infeasible if we fail to find a feasible solution to ( p2f@xmath185evenn ) with all @xmath219 sets of initial points generated . in general , a larger @xmath219 helps improve the overall accuracy of the bisection search , but at the cost of more computational complexity . to summarize , the complete algorithm to solve ( p2@xmath185evenn ) is given in table [ algorithm:1 ] , denoted by algorithm 1 . ' '' '' * algorithm @xmath31 * ' '' '' * initialize @xmath200 , @xmath220 , @xmath221 , @xmath222 , @xmath192 , and @xmath193 . * * while @xmath223 do * : * * set @xmath224 .
* * set @xmath225 , @xmath226 , @xmath227 , and @xmath228 , @xmath36 . * * * * while * @xmath225 , @xmath229 , and @xmath230 : * * * * given @xmath187 s , check the constraint ( [ eq : const1:p2-e ] ) . *
if * it holds , * then * set @xmath231 and go to step 3 ) ; * otherwise * , find the derivatives @xmath232 s as in ( [ eq : gradient ] ) and set @xmath208 if @xmath233 , or @xmath210 otherwise , @xmath234 . * * * * set @xmath235 . * * * * * if * @xmath236 and @xmath230 , * then * reset the initial points as @xmath215 , @xmath36 . set @xmath237 and @xmath226 . * * * if * @xmath231 , * then * set @xmath238 , @xmath36 , and @xmath197 ; * otherwise * set @xmath239 . *
return @xmath240 s as the solution to ( p2@xmath185evenn ) . ' '' '' [ algorithm:1 ]
in this section , we present further simulation results to evaluate the performance of the proposed transmitter node placement algorithm , i.e. , algorithm @xmath31 .
we consider the same system setup as that in section [ sec : numerical example ] , with @xmath141 identical transmitters . since @xmath32 is odd here ,
we modify algorithm @xmath31 for the even @xmath32 case to apply it for our considered system setup with @xmath141 transmitters .
we set @xmath241 , @xmath242 , @xmath243 , and @xmath244 .
first , fig .
[ fig : optimallocaion ] shows the optimized ( transmitter ) locations ( ol ) , i.e. , @xmath245 s given by algorithm @xmath31 , versus the uniform ( transmitter ) locations ( ul ) .
it is observed that for ol , except the transmitter that is located below the center of the target line ( @xmath30 ) , the other four transmitters all move closer to the center compared to ul .
[ t ! ] [ t ! ] @xmath106 ( w ) & @xmath105 ( w ) & @xmath107 ( % ) + ocol & @xmath246 & @xmath247 & @xmath161 & @xmath248 + ecol & @xmath249 & @xmath250 & @xmath251 & @xmath252 + ocul & @xmath159 & @xmath160 & @xmath161 & @xmath162 + [ tab : quantitivecomparision2 ] next , fig .
[ fig : nodepalacement_3tx ] compares the deliverable power to the load , @xmath92 given in ( [ eq : fun_p_0 ] ) , versus the receiver location @xmath28 , under three schemes : optimal ( transmitter ) current with optimized ( transmitter ) location ( ocol ) , equal ( transmitter ) current with optimized ( transmitter ) location ( ecol ) , and optimal ( transmitter ) current with uniform ( transmitter ) location ( ocul ) .
it is observed that ocol with both optimized transmitter locations and optimal magnetic beamforming improves the minimum deliverable power significantly over the other two schemes with only optimized transmitter locations or optimal magnetic beamforming .
in fact , ocol achieves the best performance in terms of all metrics , where the details are given in table [ tab : quantitivecomparision2 ] . besides , fig .
[ fig : p0min_d ] plots the minimum load power @xmath106 given in ( [ eq : p0min ] ) versus the target line length @xmath253 , under the three schemes .
it is observed that ocol consistently achieves better performance than the other two schemes , although the gain decreases when @xmath253 is small or large .
this can be explained as follows .
when @xmath253 is small , the mutual inductance between the receiver and different transmitters is less sensitive to their locations , which implies that the gain of transmitter placement optimization is small . in this case , from ( [ eq : u_n ] ) , it follows that the transmitter currents tend to be all equal , hence the magnetic beamforming gain over the equal current allocation is also negligible .
similarly , when @xmath253 is large , the distance between transmitters is large for both ul and ol designs , since there are only five transmitters available to cover the target line . in this case , the magnetic coupling between the transmitters is small , hence they can be treated as independent transmitters . as shown in fig .
[ fig : p0_1tx_x0 ] , using a single transmitter for wpt can not provide any magnetic beamforming gain . as a result ,
both transmitter location and current optimization do not yield notable performance gains .
last , we consider the practical case where the total length of coil wires for manufacturing all @xmath32 transmitters is fixed as @xmath254 m , and thus the radius of each individual transmitter coil shrinks as @xmath32 increases .
accordingly , we set the transmitter coil radius as @xmath255 in mm and keep the number of the turns fixed as @xmath153 regardless of @xmath32 .
the other parameters of the coils are assumed to be the same as in section [ sec : numerical example ] .
[ fig : p0min_n ] plots the minimum load power @xmath106 over the number of transmitters @xmath32 , with each of the three schemes .
it is observed that for all three schemes , @xmath106 first increases and then decreases with @xmath32 .
this implies that using either a small number of transmitters each with larger coil or a large number of transmitters each with smaller coil is both inefficient in maximizing the minimum deliverable power .
note that for the case of @xmath256 , i.e. , centralized wpt , @xmath257 , which is in accordance with the result in fig .
[ fig : p0_1tx_x0 ] . [ t ! ] [ t ! ]
in this section , we extend the node placement optimization to the 2d target region case .
we assume that the receiver can move horizontally within a disk of radius @xmath258 which has a fixed height of @xmath21 and the center point at @xmath259 , while all transmitters are placed in the z - plane with @xmath25 in parallel to the target region .
let @xmath260 , with @xmath261 , ( @xmath262 , with @xmath263 ) denote the location of transmitter @xmath5 ( receiver ) . in this case ,
the mutual inductance expressions given in ( [ eq : h_nk ] ) and ( [ eq : h_n0 ] ) as well as the approximation given in ( [ eq : hn0-simplified ] ) can be modified by setting @xmath264 and @xmath265 .
accordingly , the transmitters sum power and the deliverable power to the receiver load given in ( [ eq : p_sum ] ) and ( [ eq : p_0 ] ) can be re - expressed as functions of @xmath262 , @xmath260 s , and @xmath97 s as @xmath266 and @xmath267 , respectively
. define @xmath268 , which is a convex set over @xmath28 and @xmath269 .
the four performance metrics introduced for the 1d case , i.e. , the average value , the minimum value , the maximum value , and the min - max ratio of the load power given in ( [ eq : p0ave])([eq : min - max - ratio ] ) , can be redefined as @xmath270 moreover , with the optimal transmitter currents given in ( [ eq : u_n ] ) for magnetic beamforming , the deliverable power to the load in ( [ eq : p_0^ * ] ) can be rewritten as @xmath271 . similar to ( p2 ) for the 1d case , we formulate the node placement problem to maximize the minimum deliverable power to the receiver over @xmath272 in the 2d target region case as @xmath273 similar to the 1d case , it can be verified that the optimal transmitter locations in ( p3 ) must be _ circularly symmetric _ over a disk region @xmath274 . in general ,
multiple circularly symmetric structures may exist for a disk target region with different values of @xmath32 , as shown in fig .
[ fig : symstructure ] for the system of @xmath141 transmitters , where in total three circularly symmetric structures exist . denote @xmath275 as the number of all circularly symmetric structures for a given @xmath32 .
for each structure @xmath276 , @xmath277 , we first simplify ( p3 ) by exploiting the symmetry in the structure , and then solve it using a similar algorithm like algorithm @xmath31 for the 1d case .
let @xmath278 and @xmath279 denote the optimized transmitter locations and the resulting minimum load power for structure @xmath276 , respectively .
the optimal solution to ( p3 ) is thus given by @xmath280 , where @xmath281 .
next , we illustrate the above procedure for the case of @xmath141 transmitters , while the approach is general and can be applied to the cases with other @xmath32 values . for structure @xmath31 shown in fig .
[ fig : symstructure](a ) , ( p3 ) is simplified as @xmath282 where @xmath283 , @xmath284 , and @xmath285 , with @xmath286 ( the regions of @xmath287 for structures @xmath3 and @xmath288 are shown in figs .
[ fig : symstructure](b ) and [ fig : symstructure](c ) , respectively ) . in ( p3@xmath1855tx@xmath185s1 ) , @xmath289 is the single decision variable ( with @xmath175 as an auxiliary variable ) , hence algorithm @xmath31 can be easily modified to solve this problem .
let @xmath290 denote the obtained solution to ( p3@xmath1855tx@xmath185s1 ) .
accordingly , we set @xmath291 , @xmath284 , for structure @xmath31 .
similarly , we can simplify ( p3 ) for structure @xmath3 and @xmath288 , but we need to jointly optimize two decision variables @xmath289 and @xmath292 for these two structures , as shown in fig . [ fig : symstructure](b ) and fig .
[ fig : symstructure](c ) , respectively .
the details are omitted for brevity . to illustrate the performance of joint magnetic beamforming and transmitter location optimization in the 2d target region case
, we consider the same system parameters as in section [ sec : numerical example ] for the 1d target line , which is now replaced by a disk target region of radius @xmath293 m , i.e. , @xmath294 m in diameter , where the target region area ( @xmath295m@xmath60 ) is about ten times larger than the sum - area of all transmitter coils ( @xmath296m@xmath60 ) . as shown in figs .
[ fig : symstructure ] , three circularly symmetric structures exist for the system of @xmath141 transmitters . after obtaining the optimized transmitter locations for these structures , we have @xmath297 m with @xmath298 for structure @xmath31 . for structure @xmath3 , we similarly obtain @xmath299 m and @xmath300 . for structure @xmath288
, we obtain @xmath301 m , @xmath302 m and @xmath303 .
since @xmath304 , it follows that structure @xmath31 has the best performance in terms of maximizing the minimum deliverable power to the load over the 2d target region .
@xmath106 ( w ) & @xmath105 ( w ) & @xmath107 ( % ) + @xmath31 & @xmath305 & @xmath306 & @xmath161 & @xmath307 + @xmath3 & @xmath308 & @xmath309 & @xmath161 & @xmath310 + @xmath288 & @xmath311 & @xmath312 & @xmath313 & @xmath314 + [ tab : quantitivecomparision3 ] figs .
[ fig:2d - p0](a ) , [ fig:2d - p0](b ) , and [ fig:2d - p0](c ) show the deliverable load power distribution over the 2d disk region by structures @xmath31 , @xmath3 , and @xmath288 , respectively , with the optimized transmitter locations in each structure case and optimal magnetic beamforming .
the detailed performance comparison among the three structures is also summarized in table [ tab : quantitivecomparision3 ] , from which it is observed that the minimum deliverable power achieved by structure @xmath31 is much larger than those of the other two structures .
in this paper , we have studied the node placement optimization for a miso mrc - wpt system with optimal distributed magnetic beamforming .
first , we propose the optimal magnetic beamforming solution to jointly assign the currents at different transmitters subject to their sum - power constraint with given locations of the transmitters and receiver .
we show that although distributed wpt with optimal magnetic beamforming achieves better performance than centralized wpt , the resulting load power profile still fluctuates over a given target region considerably . to tackle this issue
, we formulate a node placement problem to jointly optimize the transmitter locations to maximize the minimum power delivered to the load over a 1d target region .
we propose an efficient algorithm for solving this problem based on bisection method and gradient - based search , which is shown by simulation to be able to improve the load power distribution significantly .
finally , we extend our design approach to the general case of 2d target region and show that significant performance gain can also be achieved in this case . in this paper , for simplicity we assume identical transmitter coils of equal size , while the performance of wpt may be further improved if the sizes of transmitter coils can be optimized jointly with the transmitter locations , an interesting problem worthy of further investigation . in ( [ eq : h_n0 ] ) , we can express @xmath315 , with @xmath316 .
given @xmath74 , we have @xmath317 over @xmath318 , since its maximum value over @xmath319 is @xmath320 , with @xmath321 , which decreases to zero as @xmath322 and @xmath322 for @xmath323 .
hence , we can simplify ( [ eq : h_n0 ] ) as @xmath324 where @xmath325 .
next , let @xmath326 , where @xmath327 denotes a real number and @xmath328 represents the laplace transformer . specifically , we have @xmath329 it is known that for any real function @xmath330 , with @xmath331 denoting its laplace transform , we have @xmath332 , @xmath333 and so on . for ( p1 ) , the optimal current solution @xmath337 s to ( p1 ) can be obtained by leveraging the karush - kuhn - tucker ( kkt ) conditions of the optimization problem @xcite .
let @xmath338 denote the dual variable corresponding to the constraint ( [ eq : p1_c1 ] ) .
the lagrangian of ( p1 ) is then written as @xmath339 the karush - kuhn - tucker ( kkt ) conditions of ( p1 ) are thus given by @xmath340 where ( [ eq : primalfes ] ) and ( [ eq : dualfes ] ) are the feasibility conditions for the primal and dual solutions , respectively , ( [ eq : kkt - zero - deriv ] ) is due to the fact that the gradient of the lagrangian with respect to the optimal primal solution @xmath97 s must vanish , and ( [ eq : kkt - compl - slackn ] ) stands for the complimentary slackness . to solve the set of equations in ( [ eq : primalfes])([eq : kkt - compl - slackn ] ) , we consider two possible cases as follows .
@xmath341 case 1 : @xmath342 .
it can be verified that any set of @xmath97 s satisfying @xmath343 and @xmath344 can satisfy the kkt conditions ( [ eq : primalfes])([eq : kkt - compl - slackn ] ) in this case .
however , the resulting @xmath97 s will make the objective function of ( p1 ) in ( [ eq : p1_obj ] ) equal to zero , which can not be the optimal value of ( p1 ) ; as a result , this case can not lead to the optimal solution to ( p1 ) .
@xmath341 case 2 : @xmath345 . from ( [ eq : kkt - zero - deriv ] ) , it follows that @xmath346 , @xmath114 .
moreover , from ( [ eq : kkt - compl - slackn ] ) , it follows that @xmath347 .
accordingly , we obtain @xmath348 , @xmath101 , and @xmath349 , where @xmath350 the obtained @xmath97 s and @xmath351 satisfy the kkt conditions ( [ eq : primalfes])([eq : kkt - compl - slackn ] ) . note that except the above set of primal and dual solutions to ( p1 ) , @xmath97 s and @xmath351 , given in case @xmath3 , there is no other solution that satisfies the kkt conditions ( [ eq : primalfes])([eq : kkt - compl - slackn ] ) .
thus , we can conclude that the solution obtained in case @xmath3 is indeed the optimal solution to ( p1 ) because the kkt conditions are necessary ( albeit not necessarily sufficient ) for the optimality of a non - convex optimization problem , which is the case of ( p1 ) . the proof is thus completed .
a. kurs , a. karalis , r. moffatt , j. d. joannopoulos , p. fisher , and m. soljacic , `` wireless power transfer via strongly coupled magnetic resonances , '' _ science _ ,
83 - 86 , july 2007 .
y. li , j. li , k. wang , w. chen , and x. yang , `` a maximum efficiency point tracking control scheme for wireless power transfer systems using magnetic resonant coupling , '' _ ieee trans .
power electron .
_ , vol . 30 , no . 7 , pp . 3998 - 4008 , july 2015 . j. shin , s. shin , y. kim , s. ahn , s. lee , g. jung , s. jeon , and d. cho , `` design and implementation of shaped magnetic - resonance - based wireless power transfer system for roadway - powered moving electric vehicles , '' _ ieee trans .
1179 - 1192 , mar .
2014 .
y. zhang , t. lu , z. zhao , f. he , k. chen , and l. yuan , `` selective wireless power transfer to multiple loads using receivers of different resonant frequencies , '' _ ieee trans .
power electron .
_ , vol . 30 , no .
11 , pp . 6001 - 6005 , nov . 2015 .
kim , d. ha , w. j. chappell , and p. p. irazoqui , `` selective wireless power transfer for smart power distribution in a miniature - sized multiple - receiver system , '' _ ieee trans .
1853 - 1862 , mar .
m. r. v. moghadam and r. zhang , `` multiuser wireless power transfer via magnetic resonant coupling : performance analysis , charging control , and power region characterization , '' _ ieee trans .
signal inf . process .
72 - 83 , mar .
2016 .
r. tseng , b. novak , s. shevde , and k. grajski , `` introduction to the alliance for wireless power loosely - coupled wireless power transfer system specification version 1.0 , '' in _ proc .
wireless power transfer ( wpt ) _ , pp .
79 - 83 , may 2013 . | in multiple - input single - output ( miso ) wireless power transfer ( wpt ) via magnetic resonant coupling ( mrc ) , multiple transmitters are deployed to enhance the efficiency of power transfer to the electric load at a single receiver by jointly optimizing their source currents to constructively combine the induced magnetic fields at the receiver , known as _ magnetic beamforming_. in practice , since the receiver is desired to be freely located in a target region for wireless charging , its received power can fluctuate significantly over locations even with adaptive magnetic beamforming applied . to achieve uniform coverage , the transmitters need to be optimally placed such that a minimum charging power can be achieved for the receiver regardless of its location in the region , which motivates this paper .
first , we drive the optimal magnetic beamforming solution in closed - form for a distributed miso wpt system with fixed locations of the transmitters and receiver to maximize the deliverable power to the receiver subject to a given sum - power constraint at all transmitters . with
the optimal magnetic beamforming derived , we then jointly optimize the locations of all transmitters to maximize the minimum power deliverable to the receiver over a given one - dimensional ( 1d ) region .
although the problem is non - convex , we propose an iterative algorithm for solving it efficiently .
extensive simulation results are provided which show the significant performance gains by the proposed design with optimized transmitter locations and magnetic beamforming as compared to other benchmark schemes with non - adaptive or heuristic currents allocation and transmitters placement .
last , we extend our approach to the general two - dimensional ( 2d ) region case , and highlight the key insights for practical design .
wireless power transfer , magnetic resonant coupling , magnetic beamforming , node placement optimization , uniform coverage . [ section ] [ section ] [ section ] [ section ] [ section ] [ section ] [ section ] [ section ] |
in this paper we study a family of graphs that generalises the class of cayley graphs .
we call the new graphs two - sided cayley graphs and explore some of their properties . for a subset @xmath8 of a group @xmath1 such that the identity @xmath9 and @xmath10 ( where @xmath11 ) , the _ cayley graph
_ @xmath12 is the graph with vertex set @xmath1 such that a vertex pair @xmath13 is an arc if and only if @xmath14 . the conditions @xmath15 and @xmath10 ensure that @xmath16 may be viewed as a simple undirected graph , and that the group @xmath1 acts by right multiplication as a vertex - transitive group of automorphisms @xcite .
cayley graphs were introduced by arthur cayley in 1878 to explain the concept of an abstract group given by a set of generators @xcite .
there are many meaningful applications of cayley graphs in molecular biology , computer science and coding theory @xcite .
because of their symmetry properties , cayley graphs are used as models for many interconnection networks .
moreover various generalisations of cayley graphs have been introduced in the literature as prototypes of transitive graphs with some degree of success : for instance , quasi - cayley graphs by gauyacq @xcite , various kinds of groupoid graphs by mwambene @xcite , group action graphs by annexstein et al @xcite , and general semigroup graphs by kelarev and the second author @xcite .
one of the advantages of using cayley graphs as models of networks is that their vertex - transitivity makes it possible to implement the same routing and communication scheme at each node of the network . at each vertex @xmath3 of @xmath17
the edges from @xmath3 go to the vertices @xmath18 for @xmath19 .
thus we may view @xmath8 as a ` connection subset ' @xmath20 of permutations of @xmath1 ( where @xmath21 is the permutation @xmath22 ) that determines adjacency locally at each vertex .
a two - sided cayley graph also possesses this local connection property , and is more general in the sense that the connection subset of permutations may act _ on both sides _ , that is to say , we consider permutations of the form @xmath23 for certain @xmath24 .
[ def:2sc ] for non - empty subsets @xmath0 of a group @xmath1 , we define the _ two - sided cayley graph _ @xmath2 as the ( directed ) graph with vertex set @xmath1 with an arc from @xmath3 to @xmath4 if and only if @xmath5 for some @xmath6 and @xmath7 .
the _ connection set _ of @xmath2 is defined as the set @xmath25 .
we note that the edges of @xmath2 from a vertex @xmath3 go to the vertices @xmath26 , for @xmath27 .
if the adjacency relation is symmetric , in the sense that @xmath28 is an arc if and only if @xmath29 is an arc , then @xmath30 will be regarded as an undirected graph .
conditions for this to occur and also for the graph to contain no loops or multiple edges , that is to say , conditions for @xmath30 to be a simple undirected graph , can be given in terms of properties of @xmath31 and @xmath32 as follows .
[ def:2scprop ] a pair @xmath33 of subsets of a group @xmath1 with identity element @xmath34 has the _ @xmath35s - cayley property _
if both @xmath36 are non - empty , and the following conditions all hold . * @xmath37 for each @xmath38 ; * @xmath39 for each @xmath38 ; * @xmath40 for each @xmath38 . [ thm:2scsimple ]
let @xmath36 be nonempty subsets of a group @xmath1 .
then @xmath2 is a simple undirected graph if and only if @xmath33 has the @xmath35s - cayley property .
if this is the case then @xmath2 has valency @xmath41 .
we prove theorem [ thm:2scsimple ] in section [ sec : props ] by studying separately the significance of each of the conditions in definition [ def:2scprop ] ( lemma [ lem:2scsimple ] ) . if both of the sets @xmath0 are inverse - closed , that is to say , if @xmath42 and @xmath43 , then the conditions of definition [ def:2scprop ] are somewhat simpler .
however it is possible for a pair of subsets to have the @xmath35s - cayley property but not both be inverse - closed , as demonstrated by example [ ex0 ] .
we also show in lemma [ lem:2scsimple ] that , for @xmath33 with the @xmath35s - cayley property , each element @xmath44 of @xmath45 is a _ derangement _ of @xmath1 , that is to say , @xmath44 has no fixed points in @xmath1 .
proposition [ prop : derangement ] provides a set of conditions on the connection set @xmath45 which hold for @xmath33 with the @xmath35s - cayley property .
[ rem:2sc ] \(a ) each cayley graph is a two - sided cayley graph , namely if @xmath15 and @xmath10 , then @xmath46 , and it is easy to check that @xmath47 has the 2s - cayley property .
there are sometimes other possibilities for the subsets @xmath36 giving the same cayley graph .
we describe such a graph in example [ ex1 ] .
\(b ) as mentioned above each cayley graph is vertex transitive .
this is not the case for two - sided cayley graphs , indicating in particular that these graphs form a strictly larger family of graphs : we present in example [ ex2 ] a two - sided cayley graph with two non - isomorphic connected components .
\(c ) a cayley graph @xmath17 is connected if and only if @xmath8 generates @xmath1 .
theorem [ thm : conn ] gives an analogue of this connectivity criterion for two - sided cayley graphs in the case where @xmath36 are inverse - closed subsets . by a _ word _ in a subset @xmath31 of a group @xmath1 , we mean a string @xmath48 with each @xmath49 ; the integer @xmath50 is called the _ length _ of @xmath51 , denoted @xmath52 , and we often identify @xmath51 with its _ evaluation _ in @xmath1 ( the element of @xmath1 obtained by multiplying together the @xmath53 in the given order ) .
[ thm : conn ] let @xmath36 be inverse - closed subsets of a group @xmath1 with the @xmath35s - cayley property , and let @xmath54 .
then @xmath55 is connected if and only if * @xmath56 , and there exist words @xmath51 in @xmath31 and @xmath57 in @xmath32 , with lengths of opposite parity , such that @xmath58 in @xmath1 .
further , if @xmath56 , but condition @xmath59 does not hold , then @xmath55 is disconnected with exactly two connected components .
[ rem : conn ] @xmath60 the exceptional case of theorem [ thm : conn ] does arise , see the family of graphs in example [ ex3 ] .
moreover , the proof of theorem [ thm : conn ] in section [ sec : props ] shows that , if @xmath61 but condition @xmath59 does not hold then , for each @xmath38 , and each expression @xmath62 , where @xmath51 and @xmath57 are words in @xmath31 and @xmath32 respectively , the parity of the sum @xmath63 is independent of the words @xmath64 , and depends only on @xmath65 .
let @xmath66 where @xmath67 .
we show that the connected components of @xmath55 are the sets @xmath68 , for @xmath69 .
in particular @xmath70 and @xmath71 .
@xmath72 it is not clear how to modify the proof to cover the cases where not both of @xmath36 are inverse - closed .
[ probconn ] find necessary and sufficient conditions for @xmath2 to be connected , where @xmath0 have the @xmath73-cayley property but are not both inverse - closed .
we now mention several generic isomorphisms between two - sided cayley graphs for a group @xmath1 , that are induced from automorphisms of @xmath1 .
this result is proved in section [ sec : iso ] .
[ thm : iso ] let @xmath36 be non - empty subsets of a group @xmath1 , let @xmath74 , and @xmath75 . 1 .
if at least one of the following pairs has the @xmath35s - cayley property , then they all do : @xmath76 .
2 . let @xmath54 and suppose that @xmath33 has the @xmath35s - cayley property .
then @xmath77 the smallest vertex transitive graph that is not a cayley graph is the petersen graph @xmath78 on @xmath79 vertices , and we wondered whether @xmath78 could be a two - sided cayley graph for some group of order @xmath79 .
however , the following result about connected two - sided cayley graphs of prime valency shows that @xmath78 can not be a two - sided cayley graph , since @xmath78 has valency 3 and is not a cayley graph , and the only groups of order 10 are abelian or dihedral .
[ cor : primval ] let @xmath36 be subsets of a group @xmath1 with the @xmath35s - cayley property , and suppose that @xmath54 is connected and regular of prime valency @xmath80 . then either @xmath16 is a cayley graph , or @xmath80 is odd and the following all hold , up to interchanging @xmath31 and @xmath32 . 1 .
@xmath81 , where @xmath82 lies in a @xmath1-conjugacy class @xmath83 of non - central elements ; 2 .
@xmath84 , @xmath85 , and @xmath86 .
moreover if @xmath1 is abelian or if @xmath1 is dihedral of twice - odd order , then @xmath16 is a cayley graph .
[ q1 ] find other natural classes of two - sided cayley graphs that can be guaranteed to be cayley graphs .
example [ ex2 ] demonstrates that not all two - sided cayley graphs are cayley graphs .
we next give sufficient conditions on @xmath36 for @xmath2 to be a cayley graph for @xmath1 .
it has been known since at least 1958 that a simple undirected graph is a cayley graph if and only if its automorphism group has a subgroup acting regularly on the vertices ( see sabidussi ( * ? ? ? * lemma 4 ) ) .
a permutation group is _ regular _ if it is transitive , and only the identity fixes a point . for any group @xmath1 ,
right regular representation _ and _ left regular representation _ give two regular subgroups of @xmath87 , each isomorphic to @xmath1 , namely @xmath88 and @xmath89 .
[ twosidediscayley ] let @xmath36 be subsets of a group @xmath1 with the @xmath35s - cayley property , and let @xmath54 .
1 . @xmath90 if and only if @xmath91 for each @xmath38 ; here @xmath92 .
@xmath93 if and only if @xmath94 for each @xmath38 ; here @xmath95 .
[ rem : iscayley ] \(a ) theorem [ twosidediscayley ] is proved in section [ sec : cay ] , and we derive from it , in proposition [ twosidediscayley1 ] , two sufficient conditions for @xmath2 to be a cayley graph for @xmath1 in terms of certain factorisations of the group @xmath1 .
\(b ) the conditions given in theorem [ twosidediscayley ] are far from necessary .
we give in example [ ex4 ] a family of examples of two - sided cayey graphs , all of which are cayley graphs , where none of the conditions of theorem [ twosidediscayley ] or of proposition [ twosidediscayley1 ] hold .
\(c ) a more general study of the permutations @xmath96 which act as automorphisms of a two - sided cayley graph leads to a sufficient condition for such a graph to be vertex - transitive , namely that @xmath1 factorises as @xmath97 ( proposition [ vt ] ) . studying the subclass of two - sided cayley graphs @xmath54 which are guaranteed to be vertex - transitive by the condition @xmath97 may be profitable .
in particular the @xmath73-cayley property becomes much simpler ( proposition [ simple ] ) .
however we do not know if all such graphs are cayley graphs , or if there are vertex - transitive two - sided cayley graphs for which @xmath98 .
[ prob : simple ] determine whether or not all two - sided cayley graphs @xmath54 for which @xmath97 are cayley graphs .
[ prob : nas ] find necessary and sufficient conditions on a two - sided cayley graph to be vertex - transitive .
in this section we give several examples of two - sided cayley graphs which exhibit various interesting properties .
first we give an example of a two - sided cayley graph which is a simple graph but for which the defining subsets @xmath36 are not inverse - closed .
[ ex0]let @xmath1 be the dihedral group @xmath99 of order @xmath100 , and let @xmath101 and @xmath102 .
we note that @xmath31 is not inverse - closed , but @xmath43 .
we check below that @xmath33 has the 2s - cayley property , and so by theorem [ thm:2scsimple ] , @xmath103 is a simple undirected two - sided cayley graph of valency @xmath104 .
also the condition of theorem [ twosidediscayley ] ( b ) holds so the left multiplication action of @xmath105 is a subgroup of automorphisms , proving that @xmath16 is a cayley graph , in fact @xmath106 .
also @xmath16 is the lexicographic product @xmath107 $ ] of a cycle of length @xmath108 and two isolated vertices @xmath109 .
we verify the condition of definition [ def:2scprop ] for @xmath103 .
note that @xmath43 .
for @xmath110 we have @xmath111 and so @xmath112 . similarly for @xmath113 , @xmath114 .
this proves condition ( 1 ) .
next we note that , for each @xmath38 , @xmath115 equals @xmath31 or @xmath116 and each is disjoint from @xmath32 , so ( 2 ) holds .
finally @xmath117 is @xmath1-invariant and meets @xmath118 in @xmath119 , so ( 3 ) holds .
next we give the cayley graph example referred to in remark [ rem:2sc](a ) .
[ ex1]let @xmath120 , @xmath121 and @xmath122 .
it is easy to check that @xmath33 satisfies conditions @xmath123 of definition [ def:2scprop ] , so @xmath33 has the 2s - cayley property , and by theorem [ thm:2scsimple ] , @xmath124 is a simple undirected two - sided cayley graph of valency @xmath104 . a straightforward computation verifies that @xmath16 is isomorphic to the complete graph @xmath125 with the edges of the following matching removed : @xmath126 .
also @xmath16 is the cayley graph @xmath127 ( see theorem [ twosidediscayley ] and remark [ rem : iscayley](a ) ) .
this representation of @xmath127 as a two - sided cayley graph is in addition to the one in remark [ rem:2sc ] , namely @xmath128 .
note also that @xmath129 , see theorem [ twosidediscayley ] .
now we present a two - sided cayley graph with non - isomorphic connected components , promised in remark [ rem:2sc](b ) .
[ ex2]let @xmath1 be the dihedral group @xmath99 of order @xmath100 , and let @xmath130 and @xmath131 .
it is easy to check that @xmath33 has the 2s - cayley property , and so by theorem [ thm:2scsimple ] , @xmath2 is a simple undirected two - sided cayley graph of valency @xmath132 .
in addition , @xmath133 , so @xmath2 is disconnected , by theorem [ thm : conn ] .
in fact @xmath2 has two non - isomorphic components of orders 4 and 8 , as shown in figure [ pic1 ] , and in particular this graph is not vertex - transitive .
the next family of examples demonstrates that disconnected two - sided cayley graphs @xmath2 exist with @xmath134 ( see remark [ rem : conn ] ) .
[ ex3 ] let @xmath1 be the dihedral group @xmath135 of order @xmath136 , and let @xmath137 and @xmath138 .
it is easy to check that @xmath33 has the 2s - cayley property , and so by theorem [ thm:2scsimple ] , @xmath124 is a simple undirected two - sided cayley graph of valency @xmath35 , so each connected component of @xmath16 is a cycle .
an easy computation confirms that if @xmath139 is odd then @xmath140 is connected , while if @xmath139 is even then @xmath141 has two connected components .
we now see how this confirms the conclusions of theorem [ thm : conn ] . for all @xmath139 , @xmath142 .
let @xmath143 be words in @xmath36 respectively such that @xmath58 .
in particular @xmath144 , and hence @xmath51 evaluates to @xmath34 in @xmath1 , so @xmath52 is even .
thus also @xmath145 in @xmath1 , and there exists such a word @xmath57 of odd length if and only if @xmath139 is odd , so by theorem [ thm : conn ] , @xmath16 is connected if and only if @xmath139 is odd .
suppose that @xmath139 is even , and note that the identity @xmath34 is joined to @xmath146 and @xmath147 . in this case the two components of @xmath16 are as described in remark [ rem : conn ] , for example , the component @xmath148 containing @xmath34 consists of the elements @xmath149 for all @xmath150 . finally we give the examples referred to in remark [ rem : iscayley ] ( b ) . to construct these graphs we use groups with at least one non - normal subgroup .
a _ hamiltonian group _ is a non - abelian group in which all subgroups are normal .
richard dedekind investigated finite hamiltonian groups in 1895 , and came close to a classification of them @xcite .
the classification was completed by baer @xcite .
they are precisely groups of the form @xmath151 , where @xmath152 is the quaternion group of order @xmath153 and @xmath78 is any abelian group containing no element of order @xmath104 .
dedekind named these groups in honour of william hamilton , the discoverer of quaternions , and groups in which all subgroups are normal ( that is , abelian or hamiltonian groups ) are sometimes called _
dedekind groups _ , or _
quasi - hamiltonian groups_. these are the only groups we need to avoid in example [ ex4 ] .
[ ex4]let @xmath154 and @xmath155 be non - abelian , non - hamiltonian groups with identity elements @xmath156 and @xmath157 , and set @xmath158 . for @xmath159 , let @xmath160 be a non - normal subgroup of @xmath161 , and set @xmath162 .
let @xmath54 , where @xmath163 and @xmath164 .
[ lem : ex4 ] let @xmath165 be as in example @xmath166
. then 1 .
@xmath33 has the @xmath73-cayley property , but does not satisfy any of the properties on @xmath36 in theorem @xmath167 or proposition @xmath168 .
the map @xmath169 induces an isomorphism from @xmath16 to @xmath170 .
the sets @xmath36 are inverse - closed , so definition [ def:2scprop](1 ) holds .
the other two conditions definition [ def:2scprop](2 ) and ( 3 ) hold because @xmath171 while @xmath172 .
thus @xmath33 has the @xmath73-cayley property . if @xmath173 with @xmath174 , then @xmath175 ( such an element @xmath176 exists since @xmath177 is not normal in @xmath154 ) .
similarly if @xmath178 with @xmath179 , then @xmath180 , so neither condition of theorem @xmath167 holds .
the factors on the left hand side of condition @xmath181 in proposition [ twosidediscayley1 ] are @xmath182 and @xmath183 , and their product is not equal to @xmath1 .
similarly condition @xmath184 does not hold .
this proves part ( a ) .
each arc of @xmath16 is of the form @xmath185 , for some @xmath186 .
this is mapped by @xmath187 to the pair @xmath188 , which is an arc of @xmath170 , and conversely each arc of @xmath170 is the image under @xmath187 of an arc of @xmath16 .
in this section we prove theorems [ thm:2scsimple ] , and [ thm : conn ] .
we begin by showing the significance of each of the conditions in definition [ def:2scprop ] .
recall the definition of @xmath44 given before definition [ def:2sc ] .
we will say that @xmath103 contains _ multiple arcs _ if , for some @xmath74 , there are at least two arcs of @xmath16 from @xmath3 to @xmath4 .
[ lem:2scsimple ] let @xmath36 be nonempty subsets of a group @xmath1 , and let @xmath103 .
* definition @xmath189 holds if and only if the adjacency relation of @xmath16 is symmetric .
* definition @xmath190 holds if and only if each @xmath44 ( for @xmath191 ) is a derangement , and this holds if and only if @xmath16 has no loops .
* definition @xmath192 holds if and only if @xmath16 has no multiple arcs .
in particular the conclusion of theorem @xmath193 is valid .
\(a ) suppose first that the adjacency relation of @xmath16 is symmetric .
then we have the following equivalent conditions on elements @xmath74 . [ cols= " < , < " , ] now we prove theorem [ cor : primval ] , which implies that the petersen graph is not a two - sided cayley graph for any group of order 10 .
* proof of theorem [ cor : primval ] .
* let @xmath54 , where @xmath36 are subsets of a group @xmath1 with the @xmath35s - cayley property , and suppose that @xmath16 is connected and regular of prime valency @xmath80 , that is , each vertex lies on exactly @xmath80 edges .
suppose moreover that @xmath16 is not a cayley graph .
then @xmath80 is an odd prime , since each connected graph of valency @xmath35 is a cycle , and hence a cayley graph . by theorem [ thm:2scsimple ]
, the valency of @xmath16 satisfies @xmath194 , and so @xmath195 .
also , by theorem [ thm : iso](b ) , @xmath196 , and so , up to isomorphism , we may assume that @xmath85 and @xmath197 .
let @xmath81 .
if @xmath198 , then each element @xmath65 is adjacent precisely to the vertices @xmath199 , for @xmath6 , and hence @xmath92 , a contradiction .
hence @xmath82 lies in a @xmath1-conjugacy class @xmath83 of non - central elements , and ( a ) holds .
note that @xmath86 by definition [ def:2scprop ] ( 2 ) .
since @xmath16 is connected , each @xmath38 can be reached by a path from @xmath34 , and by the definition of adjacency and using definition [ def:2scprop ] ( 1 ) , we find that @xmath200 , for some @xmath201 .
hence @xmath202 , where @xmath203 . now by definition [ def:2scprop ] ( 1 ) with @xmath204 , we have @xmath205 , and hence @xmath206 .
suppose that @xmath207 .
then @xmath208 has index @xmath35 in @xmath1 . consider the bijection @xmath209 given by @xmath210 for @xmath211 .
then @xmath212 induces an isomorphism from @xmath16 to @xmath213 where @xmath214 ( since , for all @xmath211 , an arc @xmath215 is mapped to an arc @xmath216 of @xmath217 , while an arc @xmath218 of @xmath16 is mapped to an arc @xmath219 of @xmath217 ) .
this contradiction shows that @xmath220 , so part ( b ) is proved .
if @xmath1 is abelian then there are no non - central elements , so ( a ) fails and therefore each @xmath16 is a cayley graph .
suppose now that @xmath221 with @xmath139 odd .
let @xmath222 , and note that the set of involutions in @xmath1 forms a single conjugacy class and equals @xmath223 .
let @xmath224 be as above .
since @xmath225 , it follows that @xmath31 must contain some element of @xmath223 , and by theoprem [ thm : iso ] we may assume that @xmath226 .
since @xmath86 , the element @xmath82 belongs to @xmath227 , and in particular @xmath228 is odd ( dividing @xmath139 , recall that @xmath229 ) . by definition [ def:2scprop](1 ) ,
@xmath230 , and hence @xmath231 , for each @xmath38 .
now @xmath232 or @xmath233 according as @xmath234 or @xmath235 .
thus @xmath236 .
for each involution @xmath237 ( for example , @xmath238 ) , this implies that @xmath31 contains @xmath239 for each @xmath240 ( recall that @xmath228 is odd ) .
simlarly if @xmath241 then @xmath242 , and hence also @xmath243 : so in this case we again find that @xmath31 contains @xmath244 for each @xmath245 .
thus @xmath31 is a union of cosets of the subgroup @xmath246 , and hence @xmath85 is divisible by @xmath228 .
it follows that @xmath247 , @xmath248 , and hence that
@xmath249 so @xmath250 .
an easy calculations shows that @xmath16 is the cayley graph @xmath251 .
this contradiction completes the proof .
we begin by proving theorem [ twosidediscayley ] , which gives sufficient conditions on @xmath36 for a two - sided cayley graph @xmath2 to be a cayley graph .
recall , for a group @xmath1 , the regular permutation groups in @xmath87 , @xmath88 and @xmath89 , each isomorphic to @xmath1 . * proof of theorem [ twosidediscayley ] .
* let @xmath36 be subsets of a group @xmath1 with the @xmath35s - cayley property , and let @xmath54 . by definition [ def:2scprop](1 )
, each arc of @xmath16 has the form @xmath252 for some @xmath253 .
suppose first that @xmath91 for each @xmath38 .
then @xmath254 for some @xmath255 .
the image of the arc @xmath256 under an element @xmath257 is the pair @xmath258 , for some @xmath259 ( using @xmath260 ) , and this is again an arc . hence @xmath261 is an automorphism of @xmath16 , and it follows that @xmath262 contains the regular subgroup @xmath263 .
thus , by sabidussi s theorem @xcite , @xmath16 is a caley graph for @xmath1 . indeed , since each vertex @xmath3 is adjacent precisely to the vertices in @xmath264 it follows that @xmath92 .
conversely if @xmath90 , then the set of vertices adjacent to @xmath265 is equal to the image under @xmath266 of the set @xmath267 of vertices adjacent to @xmath34 , and this is the set @xmath264 .
thus @xmath268 for each @xmath3 . here
@xmath269 .
a similar proof shows that , the condition @xmath94 , for each @xmath38 , holds if and only if @xmath262 contains the regular subgroup @xmath105 .
( in this case @xmath270 under the isomorphism @xmath271 for @xmath38 . )
now we give sufficient conditions for @xmath272 to be a cayley graph , as mentioned in remark [ rem : iscayley ] . for any subset @xmath208 of a group @xmath1 ,
the _ normalizer _ of @xmath208 in @xmath1 is @xmath273 .
it is a subgroup of @xmath1 .
[ twosidediscayley1 ] let @xmath36 be subsets of a group @xmath1 with the @xmath35s - cayley property , and let @xmath54 .
then @xmath55 is a cayley graph if at least one of the following conditions holds : @xmath274 @xmath275 in particular , if @xmath276 then @xmath92 , and if @xmath277 then @xmath95 .
let @xmath38 .
we show that @xmath91 if condition ( @xmath278 ) holds . from @xmath181
we have @xmath279 where @xmath280 and @xmath281 .
therefore , @xmath282 and then , by theorem [ twosidediscayley ] , @xmath92 .
a similar argument shows that condition @xmath184 implies that @xmath94 for each @xmath65 in @xmath1 . then by theorem [ twosidediscayley ] , @xmath270 .
[ vt ] let @xmath36 be non - empty subsets of a group @xmath1 , and let @xmath54
. then @xmath262 contains @xmath285 and @xmath286 .
in particular , @xmath287 acts vertex - transitively on @xmath16 if and only if @xmath97 .
we note that @xmath288 is a subgroup and is equal to @xmath289 , with similar comments for @xmath290 .
let @xmath291 , and let @xmath292 be an arc of @xmath16 .
the image of @xmath256 under @xmath96 is the pair @xmath293 , and this is an arc of @xmath16 since @xmath294 , for some @xmath295 and @xmath255 , by the definition of @xmath287 .
let @xmath296 .
then the image of @xmath256 under @xmath297 is @xmath298 , and this is an arc of @xmath16 since @xmath299 .
let @xmath38 .
suppose first that @xmath97 .
then @xmath279 for some @xmath300 , and @xmath291 maps @xmath34 to @xmath65 so @xmath287 is vertex - transitive .
conversely suppose that @xmath287 is vertex - transitive , so some @xmath291 maps @xmath34 to @xmath65 .
then @xmath301 .
the first author acknowledges support of an endeavour award of the australian government for his study leave during which the research for this paper began .
he thanks the school of mathematics and statistics , university of western australia , for their warm hospitality during his visiting appointment and for the facilities and help provided .
in addition he thanks the iranian elite national foundation ( bonyad melli nokhbegan ) for financial support and the department of mathematical sciences , sharif university of technology , for their hospitality during his postdoctoral appointment and for the facilities .
lowell w. beineke and robin j. wilson ( eds . ) , _ topics in algebraic graph theory _ , en- cyclopedia of mathematics and its applications , vol .
102 , cambridge university press , cambridge , 2004 . mr mr2125091 ( 2005m:05002 ) gea hahn and gert sabidussi ( eds . ) , _ graph symmetry _ , nato advanced science insti- tutes series c : mathematical and physical sciences , vol .
497 , dordrecht , kluwer aca- demic publishers group , 1997 , algebraic methods and applications .
mr mr1468785 ( 98d:05002 ) j. kececioglu and d. sankoff , exact and approximation algorithms for sorting by reversals , with application to genome rearrangement , _ algorithmica _ 13 ( 1995 ) , no .
12 , 180210 . mr mr1304314 ( 95j:68125 ) s. lakshmivarahan , jung sing jwo , and s. k. dhall , symmetry in interconnection networks based on cayley graphs of permutation groups : a survey , _ parallel comput .
_ 19 ( 1993 ) , no .
4 , 361407 . mr mr1216119 ( 94j:68240 ) | we introduce a family of graphs that generalises the class of cayley graphs . for non - empty subsets @xmath0 of a group @xmath1 , the two - sided cayley graph @xmath2 is the directed graph with vertex set @xmath1 and an arc from @xmath3 to @xmath4 if and only if @xmath5 for some @xmath6 and @xmath7 .
thus , in common with cayley graphs , two - sided cayley graphs may be useful to model networks as the same routing and communication scheme can be implemented at each vertex .
we determine when two - sided cayley graphs are simple undirected graphs , and give sufficient conditions for them to be connected , vertex - transitive , or cayley graphs .
several open problems are posed .
many examples are given , including one on 12 vertices with connected components of sizes 4 and 8 .
+ * cayley graph , vertex - transitive graph , group .
* .
* two - sided cayley graphs * + + + |
the quasifuchsian space @xmath4 of once punctured tori can be embedded in @xmath14 by the complex fenchel - nielsen coordinates ( c.f .
@xcite ) . by varying the complex twist @xmath15 and keeping the complex length @xmath16 being fixed as a positive real value @xmath11
, we can define the _ linear slice _
@xmath17 of @xmath4 . in this paper
we investigate the global properties of @xmath9 realized in the complex plane .
to state our results , recall that @xmath9 has a component containing the open interval @xmath18 which was studied in @xcite and also in @xcite . in this paper
we call this component the _ standard component _ and the others _ non - standard_. we will show there exists some positive constant @xmath19 such that for any @xmath11 satisfying @xmath20 , @xmath9 coincides with the standard component .
there exists some positive constant @xmath21 such that for any @xmath11 satisfying @xmath22 , @xmath9 contains non - standard components . in section 7 , we also consider the scaling property of @xmath9 .
linear slice has an asymptotic scaling constant .
see figure 1 for theorem 5.1 and figure 2 and 3 for theorem 6.1 and corollary 7.3 .
the parameters used in the figures are explained in 4.1 .
let us describe some historical background of our subject .
a marked quasifuchsian punctured torus group @xmath1 is a free marked two generator discrete subgroup of @xmath23 such that the commutator of the generators is parabolic , and the regular set @xmath24 consists of two non - empty simply connected invariant components @xmath25 .
quasifuchsian space @xmath4 is the space of marked quasifuchsian punctured torus groups modulo conjugation in @xmath23 .
the convex core @xmath26 has two boundary components @xmath27 each of which is a once - punctured torus and admits an intrinsic hyperbolic structure making it a pleated surface . in their seminal paper
@xcite l. keen and c. series defined the _ bers - maskit slice _
@xmath28 for a fixed measured lamination @xmath29 and @xmath30 , as the subset of @xmath4 on which the bending lamination of @xmath31 and @xmath29 belong to the same projective class and the length of @xmath29 in @xmath31 is equal to @xmath11 . by using their theory of pleating coordinates
, they showed that @xmath32 is simply connected .
j. parker and j. parkkonen also studied these slices for the case that @xmath29 is a rational lamination ( they call them the @xmath16-_slices _ ) , and considered a generalization of i. kra s plumbing construction and degeneration of @xmath28 to the maskit slice @xmath33 ( c.f . @xcite ) . the first author and j. parkkonen further studied @xmath28 ; they showed that the boundary of @xmath28 is a jordan curve which is cusped at a countable dense set of points ( c.f .
@xcite ) . in this papaer
we would like to study the outside of @xmath28 in @xmath9 and its scaling property .
this paper is organized as follows . in section 2
we will review the basic notions of the quasifuchsian space @xmath4 of once punctured tori and its pleating varieties following @xcite .
the complex fenchel - nielsen coordinates of @xmath4 will be introduced in section 3 , and we will define the main subject of this paper , the _ linear slice _
@xmath9 of @xmath4 in section 4 . in sections 5 and 6
we will study connected components of @xmath9 and prove our main theorems . and in the last section we will observe the asymptotic self similarity of @xmath9 .
we are grateful to caroline series for showing us the preprint @xcite of otal , and raquel diaz for explaining us her idea in section 6 .
it was fruitful for us to discuss with them in nara in january 2000 ; in practice this work was almost done during their stay in japan .
we also wish to thank hideki miyachi for enjoyable conversations with him on the topic in section 7 , and kentaro ito and sara maloni for telling their interests in our paper recently .
the first author was partially supported by grant - in - aid for scientific research(c ) ( 19540194 ) , ministry of education , science and culture of japan .
let @xmath34 be an oriented once - punctured torus . any pair of simple closed loops on @xmath34 that intersect exactly once are free generators of @xmath35 .
let @xmath36 be such an ordered pair of free generators , chosen so that their commutator @xmath37 represents a positively oriented loop around the puncture .
the ordered pair @xmath36 is called a _
marking_. a _ punctured torus group _ is a discrete subgroup @xmath39 that is the image of a faithful representation @xmath40 of @xmath35 such that the image of the loop around the puncture is parabolic . if @xmath36 is a marking of @xmath34 , and if @xmath41 , then the commutator @xmath42 is parabolic and the ordered pair @xmath43 is called a _ marking _ of @xmath1 .
the group @xmath1 is _ quasifuchsian _ if the regular set @xmath44 consists of two non - empty simply connected invariant components @xmath45 .
the limit set @xmath46 is topologically a circle . _ quasifuchsian space
_ @xmath4 is the space of marked quasifuchsian punctured torus groups modulo conjugation in @xmath47 ; it has a holomorphic structure induced from the natural holomorphic structure of @xmath48 .
let @xmath49 be the set of @xmath47-conjugacy classes of representations @xmath40 of @xmath35 such that the image of the loop around the puncture is parabolic .
considering the compact open topology on @xmath49 , minsky showed that the closure of @xmath4 in @xmath49 is equal to @xmath50 , the set of punctured torus groups modulo conjugation in @xmath47 ( c.f .
@xcite ) . _ fuchsian space _
@xmath38 is the subset of @xmath4 such that the components @xmath25 are round disks .
it is canonically isomorphic to the teichmller space of marked conformal structures on @xmath34 .
the quotients @xmath51 are punctured tori with conformal structures , and hence also with orientations inherited from @xmath52 ; we assume that the orientations of @xmath53 and @xmath34 agree whereas those of @xmath54 and @xmath34 are opposite .
a point @xmath55 represents an equivalence class of marked groups in @xmath56 .
we choose once and for all a triple of distinct points in @xmath52 and let @xmath57 denote the representative normalized so that the repelling and attracting fixed points of @xmath58 and the fixed point of @xmath59 are equal to the fixed triple points in @xmath52 . if it is clear from the context , for readability , we suppress the dependence on @xmath60 .
denote by @xmath61 , the set of free unoriented homotopy classes of simple closed non - boundary parallel curves on @xmath34 .
as is well known , this set can be naturally identified with @xmath62 .
one way to see this is as follows ; let @xmath63 denote the integer lattice @xmath64 .
topologically @xmath34 is the quotient of the punctured plane @xmath65 by the natural action of @xmath66 by the horizontal and vertical translations . a straight line of rational slope in @xmath67 projects onto a simple closed curve on the marked punctured torus @xmath68 , and the projection of all lines of the same rational slope with the same orientation are homotopic .
we denote the unoriented homotopy class obtained by projecting the line of slope @xmath69 by @xmath70 $ ] .
relative to our choice of marking , @xmath70 $ ] is in the homology class of @xmath71 or @xmath72 on @xmath73 , where @xmath74 are projections of the horizontal and vertical lines corresponding to @xmath75 respectively . setting @xmath76
, we obtain that the map @xmath77 defined by @xmath78 $ ] which is well - defined and bijective .
the reason for the choice of convention that @xmath70 $ ] corresponds to @xmath71 , is the following ; if we identify the teichmller space @xmath79 of once punctured tori with the upper half plane @xmath80 , then one can easily compute that the boundary point @xmath81 is the point where the extremal length of curves in the class @xmath70 $ ] has shrunk to zero .
suppose that @xmath83 is a quasifuchsian punctured torus group , marked as usual by generators @xmath84 .
we denote the unique geodesic in the homotopy class of @xmath85)$ ] in @xmath86 by @xmath87 . in particular , for @xmath55 , @xmath88 represents the corresponding geodesic in @xmath89 . for each @xmath90
, we can find an explicit word @xmath82 in the marked generators @xmath91 of @xmath35 representing @xmath70 $ ] as follows .
the words are generated from the initial data @xmath92 by the formula @xmath93 whenever @xmath94 and @xmath95 .
we denote by @xmath96 the corresponding special word in @xmath97 .
we are now ready to discuss the convex hull boundary and the pleating locus .
let @xmath55 and let @xmath98 be the corresponding marked quasifuchsian group with the regular set and the limit set @xmath99 respectively .
the @xmath100-manifold @xmath101 is homeomorphic to @xmath102 .
the surfaces @xmath103 at infinity form the boundary @xmath104 .
let @xmath105 be the boundary of the hyperbolic convex hull of @xmath46 in @xmath106 ; it is clearly invariant under the action of @xmath1 .
the nearest point retraction @xmath107 , defined by mapping @xmath108 to the unique point of contact with @xmath105 of the largest horoball in @xmath106 centered at @xmath109 with interior disjoint from @xmath105 , can easily be modified to a @xmath1-equivariant homeomorphism .
we denote two connected components of @xmath105 corresponding to @xmath110 by @xmath111 respectively .
thus each component @xmath112 is topologically a punctured torus .
( in the special case in which @xmath1 is fuchsian , @xmath105 is a flat plane whose two sides serve as a substitute for the two components @xmath113 . )
@xmath114 are pleated surfaces in @xmath86
. more precisely , there are complete hyperbolic surfaces @xmath115 , each homeomorphic to @xmath34 , and maps @xmath116 , such that every point in @xmath117 is in the interior of some geodesic arc which is mapped by @xmath118 to a geodesic arc in @xmath119 , and such that @xmath118 induce isomorphisms @xmath120 .
further , @xmath118 are isometries onto their images with the path metric induced from @xmath106 ( c.f .
@xcite ) . the _
bending _ or _ pleating locus _ of @xmath121 consists of those points of @xmath115 contained in the interior of one and only one geodesic arc which is mapped by @xmath118 to a geodesic arc in @xmath119 . for @xmath1 non - fuchsian ,
the pleating loci are geodesic laminations , meaning that they are unions of pairwise disjoint simple closed geodesics on @xmath117 .
we denote these laminations by @xmath122 , and usually identify such a lamination with its image under @xmath118 in @xmath119 .
a geodesic lamination is called _ rational _ if it consists entirely of closed leaves .
we concentrate on the special case in which at least one of the pleating loci is rational in this sense .
since the maximum number of pairwise disjoint simple closed curves on a punctured torus is one , such a lamination consists of a single simple closed geodesic and is therefore of the form @xmath88 for some @xmath90 .
given @xmath90 , we set @xmath123 we call these sets the @xmath124-_pleating varieties_. for any @xmath90 , consider the trace @xmath125 of the special word @xmath82 associated to @xmath124 defined in [ sec : special word ] . for @xmath126
, we may consider the function @xmath127 as a rational function on @xmath49 .
we define the _ hyperbolic locus _ of @xmath128 to be the set @xmath129 .
the _ complex translation length _ @xmath130 of @xmath131 is given by the equation @xmath132 where @xmath133 is the trace of @xmath134 and we choose the sign so that @xmath135 .
the complex length is invariant under conjugation by mbius transformations and has the following geometric interpretation , provided @xmath134 is not parabolic ; let @xmath109 be a point on the axis @xmath136 of @xmath134 and let @xmath137 be a vector normal to @xmath138 at @xmath109 . then @xmath139 is the hyperbolic distance between @xmath109 and @xmath140 and @xmath141 is the angle @xmath142 between @xmath143 and the parallel transport of @xmath137 to @xmath140 , measured facing the attracting fixed point @xmath144 of @xmath134 . in particular , if @xmath134 is loxodromic then @xmath145 and if @xmath134 is purely hyperbolic then in addition @xmath146 . for @xmath55 and @xmath147
, we denote the element in the group @xmath97 representing @xmath148 by @xmath149 .
because the trace is a conjugation invariant , the complex translation length @xmath150 depends only on @xmath60 and is independent of the normalization of @xmath97 .
we want to define the complex length @xmath151 as a holomorphic function on @xmath4 with values in @xmath152 , not @xmath153 .
to do this , we choose the branch that is real valued on @xmath38 . since @xmath154 on @xmath4 this choice uniquely determines a holomorphic function @xmath155 . from now on ,
the term `` complex length '' will always refer to this branch .
we remark that @xmath156 is the hyperbolic length of @xmath148 in @xmath157 .
the complex fenchel - nielsen parameters were introduced in @xcite as a generalization to @xmath4 of the classical fenchel - nielsen coordinates for fuchsian groups . here
we briefly summarize the main points as applied to the case of a punctured torus @xmath34 .
let @xmath158 be a marked quasifuchsian punctured torus group constructed from a pair of marked generators @xmath74 of @xmath35 as described in [ sec : normalization ] .
the complex fenchel - nielsen coordinates @xmath159 for @xmath160 are obtained as follows ; the parameter @xmath161 is the complex translation length of the generator @xmath162 , or equivalently the complex length @xmath163 .
the twist parameter @xmath164 measures the complex shear when the axis @xmath165 is identified with the axis @xmath166 by @xmath167 .
more precisely , if the common perpendicular @xmath168 to @xmath165 and @xmath166 meets these axes in points @xmath169 respectively , then @xmath170 is the _ signed _ distance from @xmath171 to @xmath172 and @xmath173 is the angle between @xmath168 and the parallel translate of @xmath174 along @xmath166 to @xmath171 , measured facing towards the attracting fixed point of @xmath58 .
the conventions for measuring the signed distance and the angle are explained in more detail in @xcite . as shown in @xcite ,
given the parameters @xmath175 , and fixed a normalization , one can explicitly write down the matrix generators for a marked two generator group @xmath176 in which the commutator @xmath177 $ ] is parabolic as follows : @xmath178 @xmath179 this group may or may not be discrete
. the matrix coefficients of @xmath1 depend holomorphically on the parameters .
the construction thus defines a holomorphic embedding of @xmath180 into a subset of @xmath181 , in which fuchsian space @xmath38 is identified with the image of @xmath182 .
we want to lift this to an embedding into @xmath5 . in [ sec : cxlenlox ] we discussed how to lift the length function @xmath183 on @xmath4 to a holomorphic function on @xmath152 .
we can similarly lift the twist parameter @xmath184 by specifying that it will be real valued on @xmath38 . on @xmath38 ,
the real valued parameters @xmath175 reduce to the classical fenchel - nielsen parameters @xmath185 defined by the above construction with @xmath183 the hyperbolic translation length @xmath186 of @xmath58 and @xmath184 the twist parameter @xmath187 .
clearly , the complex fenchel - nielsen coordinates can be made relative to any marking @xmath188 of @xmath1 .
as described in detail in section 5 of @xcite , for fixed @xmath189 and @xmath190 , the complex fenchel - nielsen coordinates relative to @xmath188 determines a pleated surface @xmath191 .
we review this process .
write @xmath192 for the set of all lifts of the simple closed curve @xmath148 corresponding to @xmath193 to @xmath194 . since @xmath148 is simple , @xmath192
consists of a set of pairwise disjoint geodesics in @xmath194 , namely the axis of @xmath193 and all of its conjugates .
these axes in @xmath192 partition @xmath194 into pieces @xmath195 .
the map @xmath196 is defined in such a way that @xmath196 is an isometry on each axis in @xmath192 and on each closed piece @xmath195 .
let @xmath197 and let @xmath198 be an oriented geodesic from @xmath109 to @xmath199 .
let @xmath200 be the pieces cut in order by @xmath198 , that meet along axes @xmath201 .
orient @xmath202 so that , in @xmath194 , @xmath203 lies to the left of @xmath202 and @xmath195 to the right .
let @xmath204 and let @xmath205 be tangent vectors to @xmath206 and @xmath207 at @xmath208 , oriented in the direction inherited from @xmath198 , so that @xmath209 points out of @xmath210 and @xmath211 points into @xmath212 .
let @xmath213 be the projections of @xmath205 onto the directions orthogonal to the image of the bending axis at @xmath208 .
then @xmath214 is the angle from @xmath209 to @xmath211 measured facing along @xmath215 .
we embed @xmath194 in the hyperbolic ball model @xmath216 of @xmath106 as the equatorial plane such that the origins in @xmath194 and in @xmath216 coincide .
we arrange that the axes of @xmath193 and @xmath217 in @xmath218 lie in the boundary of a piece @xmath219 contained in @xmath194 .
we then choose @xmath196 to be the identity on @xmath219 .
we set @xmath220 for the image of the pleated surface in @xmath216 .
then @xmath196 induces the group isomorphism @xmath221 satisfying that @xmath222 for @xmath223 and @xmath224 .
the next proposition explains the relation between @xmath196 and the bending locus of @xmath225 for @xmath55 .
[ prop : homeo ] for @xmath55 , let @xmath226 be the complex fenchel - nielsen coordinates relative to marked generators @xmath227 of @xmath97 , and let @xmath148 be the simple closed curve corresponding to @xmath193 .
assume that @xmath193 is purely hyperbolic and let @xmath191 be the pleated surface defined above .
then @xmath196 is a homeomorphism if and only if @xmath228 .
first suppose that @xmath196 is a homeomorphism .
then the boundary of @xmath229 is @xmath230 .
@xmath229 divides @xmath106 into two domains ; one of which is convex , hence contains @xmath231 .
moreover @xmath229 contains the axis of @xmath193 and all of its conjugates in @xmath97 , and the complement of them consist of totally geodesic pieces .
therefore it is one of the component of @xmath232 and from the bending construction in the above argument , it should be equal to @xmath225 ( c.f .
section 7.1 in @xcite ) .
next suppose that @xmath228 .
then @xmath225 consists of the axis of @xmath193 and all of its conjugates in @xmath97 , and totally geodesic pieces .
the stabilizer subgroup of each totally geodesic piece is conjugate to the fuchsian subgroup @xmath233 .
therefore we can construct the pleated surface satisfying that @xmath234 , which implies that @xmath196 is a homeomorphism .
let @xmath235 be the complex fenchel - nielsen coordinates relative to marked generators @xmath227 of @xmath1 , and let @xmath148 be the simple closed curve corresponding to @xmath193 .
assume that @xmath193 is purely hyperbolic and let @xmath11 be the hyperbolic length of @xmath148 in @xmath119 .
we denote the slice @xmath236 by @xmath7 and call it the _ rational quakebend plane_. clearly , @xmath7 meets @xmath38 along the earthquake path ( c.f .
the quakebend parameter @xmath15 is a holomorphic coordinate on @xmath7 . on @xmath7 ,
the quakebend parameter @xmath15 and @xmath237 are related by @xmath238 on @xmath7 , @xmath237 is a holomorphic function of @xmath15 , branched at @xmath239 .
( see figure 5.1 in @xcite ) .
when @xmath240 is real , @xmath10 is contained in the strip @xmath241 from the argument in [ sec : genqbs ] .
@xmath237 takes the right half strip @xmath242 conformally onto the right half plane @xmath243 minus the interval @xmath244 $ ] where the interval @xmath245 in the imaginary axis is folded at the origin by @xmath237 and its image is @xmath246 $ ] .
we remark that @xmath10 is also periodic under the action of the dehn twist @xmath247 , and symmetric under the holomorphic involution @xmath248 .
for @xmath249 , a marked group @xmath250 modulo conjugation in @xmath47 is uniquely determined by @xmath251 and @xmath252 .
in fact , ignoring marking , @xmath97 modulo conjugation in @xmath47 is determined only by @xmath253 and @xmath254 ( more precisely , the pair @xmath255 determines a marked group @xmath256 or @xmath257 modulo conjugation in @xmath47 ) . as an application of the jorgensen s theory on the combinatorial structure of the ford domain of a punctured torus group
, there is an algorithm roughly answering whether @xmath97 is a geometrically finite discrete group or not from the data @xmath255 ( c.f .
especially fixing @xmath258 , then we can use this algorithm to draw the picture of @xmath259 we call this set the _
discrete locus_. let @xmath7 be the rational quakebend plane in the complex fenchel - nielsen coordinates relative to the corresponding marked generators @xmath260 of @xmath1 where we assume that @xmath261 is real .
the _ linear slice _
@xmath9 in the right half plane @xmath243 , which is the image of @xmath10 under @xmath254 . because @xmath4 is open in @xmath5 in complex fenchel - nielsen coordinates and @xmath254 is a open map on @xmath7 , @xmath9 is open in @xmath243 .
then from the definition of @xmath262 and @xmath9 , @xmath9 is a subset of the interior of @xmath262 .
the interior of @xmath262 is equal to @xmath9 .
it is enough to show that any point @xmath263 of @xmath262 , not contained in @xmath9 is a boundary point of @xmath262 .
first suppose that @xmath264 is not a free group .
then some word , say @xmath265 is trivial in @xmath264 .
then applying the jorgensen s inequality for the subgroup @xmath266 generated by @xmath267 and @xmath268 $ ] , we can see that if we take a small neighborhood @xmath269 of @xmath263 , for any point @xmath60 of @xmath269 except @xmath263 , @xmath270 is not discrete which means that @xmath97 is also indiscrete .
therefore @xmath263 is an isolated point of @xmath262 .
next suppose that @xmath264 is free , hence geometrically finite non - quasifuchsian punctured torus group
. then it must be a cusp ( c.f .
hence there is some word , say @xmath271 which is parabolic in @xmath264 .
since @xmath272 is a holomorphic function of @xmath254 , it is a open map , hence there is a path in the @xmath254-plane starting from @xmath263 such that on this path @xmath267 is elliptic .
therefore this path is outside of @xmath262 .
this implies that @xmath263 is a boundary point of @xmath262 . from this result
, we can see @xmath9 as the interior @xmath262 and study them experimentally .
figures at the end of this paper drown by the second author show computer - generated linear slices , revealing some global properties .
the black region corresponds to the discrete locus . in the first picture ,
@xmath253 is fixed at @xmath273 and @xmath254 ranges in the square of width @xmath274 centered at @xmath275 so that we see the familiar picture of the maskit slice . by setting @xmath276
, we get the second picture . in figure 2 and 3 ,
the value of @xmath253 is fixed at @xmath277 and @xmath278 respectively while changing the ranges of @xmath254 .
the width of the squares are 16 , 32 , 128 and 128 , 2560 , 12800 respectively .
we can clearly see the `` rough self similarity '' of the pictures between figures 3 and 5 and between figures 6 and 8 , with which we will discuss in section 7 . for any @xmath30
, @xmath9 has a component containing an open interval @xmath18 .
there exists a component in @xmath10 containing @xmath279 the real line which is periodic under the action of the dehn twist @xmath280 and symmetric under @xmath248 ( c.f .
see also @xcite ) .
then its image under @xmath254 is the required component .
this component is called the _ bm - slice _ in @xcite and also called the _ @xmath16-slice _ in @xcite . in this paper
we call this component of @xmath9 the _ standard component _ , and call the other components the _ non - standard components _ if they exist .
because of the existence of the standard component which contains the critical value of @xmath254 , if there is a non - standard component , it is a conformal image of a component of @xmath10 under the map @xmath254 .
therefore we can consider that @xmath9 describes the picture of @xmath10 .
next result shows that topologically every component is a disk .
each component of @xmath9 is simply connected .
this is a consequence of a result of mcmullen @xcite that @xmath4 is disk convex in @xmath49 , that is , for any continuous map @xmath281 whose restriction to the unit disk @xmath282 is holomorphic , @xmath283 implies @xmath284 .
( the maskit slice ) + if we consider the limiting case where @xmath285 , we can no longer consider the complex fenchel - nielsen coordinates . but by using @xmath254 we can realize the part of the boundary of @xmath4 defined by the condition that @xmath58 is parabolic .
then the standard component defined above corresponds to the so - called maskit slice @xmath286 .
[ thm : non - existence ] there is some positive constant @xmath19 such that for any @xmath11 satisfying @xmath287 , @xmath9 coincides with the standard component .
this is an immediate consequence of the following result due to j. p. otal @xcite .
corollaire 9.1 in @xcite ) + [ thm : otal ] there exists a positive constant @xmath19 such that for a marked quasifuchsian punctured torus group @xmath97 and @xmath288 representing a simple closed geodesic @xmath148 in @xmath157 , if @xmath193 is purely hyperbolic and the hyperbolic length @xmath289 of @xmath148 is less than @xmath19 , then @xmath148 is a bending locus of @xmath232 . following the proof of proposition 9 in @xcite , we will give a proof of theorem [ thm : otal ] to estimate @xmath19 in the next subsection 5.2 .
suppose that @xmath148 is not the bending locus of @xmath232 .
then the pleated surface @xmath196 is not a homeomorphism by proposition [ prop : homeo ] .
let @xmath290 be a fuchsian subgroup @xmath233 where @xmath0 is a marking of @xmath97 .
denote the totally geodesic plane whose boundary @xmath291 contains @xmath292 by @xmath293 .
let @xmath294 be the convex hull of @xmath292 in @xmath293 .
then @xmath295 is the convex hull of @xmath296 in @xmath297 .
now we have a following claim . [ prop : intersection ] if the pleated surface @xmath196 is not a homeomorphism , then there exists @xmath298 such that @xmath294 and @xmath295 intersect transversally in the axis of @xmath193 . to show this proposition , we need two lemmas . [ lem : group ] for @xmath298 , @xmath299 is trivial or cyclic subgroup generated by @xmath300 .
[ lem : limit set ] for @xmath298 , @xmath301 is empty or fixed points of @xmath300 .
@xmath290 and @xmath302 are fuchsian subgroups of a quasifuchsian group @xmath97 . from a theorem of suskind ( see theorem 3.14 in @xcite ) , @xmath303 hence it concludes the proof by using lemma [ lem : group ] .
now we can show proposition [ prop : intersection ] .
since we assume that @xmath196 is not a homeomorphism , there exists @xmath304 such that the interior of @xmath294 and the interior of @xmath295 intersect transversally .
then from lemma [ lem : limit set ] , @xmath301 is empty
. therefore the axis of @xmath193 cuts @xmath295 transversally .
now we assume that @xmath19 is smaller than the margulis constant .
then the interior of @xmath295 cut the margulis tube @xmath305 with radius @xmath306 along the axis of @xmath193 transversally .
hence now we have a geodesic disk @xmath307 on @xmath295 .
the hyperbolic area of @xmath282 is bigger than @xmath308 .
if @xmath295 intersects the axis of @xmath193 orthogonally , then @xmath282 is a hyperbolic disk of radius @xmath306 , hence the hyperbolic area of it is @xmath308 .
if @xmath295 intersects the axis of @xmath193 not orthogonally , then @xmath282 contains a hyperbolic disk of radius @xmath306 , hence the hyperbolic area of @xmath282 is bigger than @xmath308 .
now we can give a proof of theorem [ thm : otal ] .
by the margulis lemma , @xmath282 projects into the image of @xmath295 in @xmath157 injectively , whereas the image of @xmath295 in @xmath119 has its hyperbolic area @xmath309 since it is the isometric image of a punctured cylinder .
therefore if the hyperbolic length of @xmath148 in @xmath157 is sufficiently small , we can take a radius @xmath306 of the margulis tube @xmath305 satisfying @xmath310 , which is a contradiction .
this concludes the theorem .
following @xcite , we have a formula of the radius of a margulis tube .
theorem in section 3 of @xcite ) + for @xmath55 , assume that @xmath288 representing a simple closed geodesic @xmath148 in @xmath157 , which is purely hyperbolic . if the hyperbolic length @xmath311 of @xmath148 satisfies @xmath312 , then there is a margulis tube with radius @xmath306 satisfying @xmath313 the inequality @xmath314 and the above formula give us a lower bound of @xmath19 . @xmath315 .
[ thm : existence ] there is some positive constant @xmath21 such that for any @xmath11 satisfying @xmath316 , @xmath9 contains non - standard components .
to prove this theorem , we use the earle slice @xmath317 of punctured tori studied in @xcite .
this idea is due to raquel diaz .
we review notations of @xmath317 ( c.f .
earle slice _
@xmath317 of @xmath4 is the set of @xmath250 satisfying the following symmetry ; there exists an elliptic element of order 2 such that @xmath318 . then @xmath317 is a holomorphic slice of @xmath4 and considering the conformal structure of @xmath319 , it is naturally isomorphic to the teichmller space of punctured tori .
any element of @xmath317 can be represented by the following matrices in @xmath48 of the form @xmath320 , where @xmath321 the complex parameter @xmath322 gives a holomorphic embedding of @xmath317 into the right half plane @xmath243 and we assume that @xmath317 is embedded in @xmath243 . then @xmath317 contains the positive real line @xmath323 which is the fuchsian locus @xmath324 of @xmath317 . put @xmath325 . to show theorem [ thm : existence ]
, we need lemmas .
there is a positive constant @xmath326 such that for any @xmath327 , the hyperbolic locus @xmath328 of @xmath329 satisfies @xmath330 we remark that @xmath331 .
then we can check our claim by direct calculation .
there is a positive constant @xmath332 such that the @xmath333-pleating variety @xmath334 satisfies @xmath335 in @xcite , it is shown that @xmath336 is equal to two components of @xmath337 terminating to the unique critical point of @xmath125 on @xmath338 ( c.f .
theorem 5.1 in @xcite ) .
then we can check our claim by direct calculation .
@xcite ) there is a positive constant @xmath339 such that @xmath340 now we can prove theorem [ thm : existence ] .
there is a positive constant @xmath21 such that for any @xmath341 , there is @xmath342 such that the word @xmath343 is purely hyperbolic and @xmath344 , but @xmath322 is not contained in @xmath334 .
this concludes the theorem . to estimate @xmath21
, we need to know the size of the round disk contained in @xmath317 tangent to the boundary @xmath345 of @xmath317 at the origin ( see @xcite ) . comparing with the results in section 5 and 6
, we have the following conjecture supported by numerical experiences by the second author .
there exists a unique @xmath19 such that @xmath9 coincides with the standard component for any @xmath346 , while @xmath9 contains infinitely many non - standard components for any @xmath347 .
in the final section we will study the self - similar phenomena of @xmath9 which we can observe from figures of @xmath9 in this paper .
first we remark that @xmath9 has analytic automorphisms coming from dehn twists .
the following trace identity is well known : @xmath354 divide both sides by @xmath355 and put @xmath356 .
then we have @xmath357 moreover put @xmath358 , then @xmath359 finally put @xmath360 , then @xmath361 since @xmath58 is purely hyperbolic , the linear fractional transformation @xmath362 is also purely hyperbolic , hence all points besides the repelling fixed point of @xmath58 converge to the attracting fixed point of @xmath58 , @xmath363 . from the above arguments
, @xmath364 converges to this point .
linear slices with ranges 16(left ) , 32(center ) , 128(right),title="fig:",width=132 ] linear slices with ranges 16(left ) , 32(center ) , 128(right),title="fig:",width=132 ] linear slices with ranges 16(left ) , 32(center ) , 128(right),title="fig:",width=132 ] linear slices with ranges 128(left ) , 2560(center ) , 12800(right),title="fig:",width=132 ] linear slices with ranges 128(left ) , 2560(center ) , 12800(right),title="fig:",width=132 ] linear slices with ranges 128(left ) , 2560(center ) , 12800(right),title="fig:",width=132 ] h. akiyoshi , m. sakuma , m. wada and y. yamashita , punctured torus groups and 2-bridge knot groups ( 1 ) , lecture notes in mathematics 1909 , springer - verlag ( 2007 ) .
r. d. canary , d. b. a. epstein and p. green .
notes on notes of thurston . in d. b.
a. epstein , editor , _ analytical and geometric aspects of hyperbolic space _ , lms lecture notes 111 , 392 .
cambridge university press ( 1987 ) .
l. keen and c. series . how to bend pairs of punctured tori . in j.
dodziuk and l. keen , editors , _ lipas= legacy , proceedings of the bers colloquium 1995 _ , contemp .
211 , 359388 .
ams ( 1997 ) .
l. keen and c. series , pleating invariants for punctured torus groups , topology 43 , ( 2004 ) , 447491 .
y. komori on the boundary of the earle slice for punctured torus groups , kleinian groups and hyperbolic 3-manifolds , lond .
. notes . 299 ( 2003 ) , 293 - 304 .
y. komori and j. parkkonen , on the shape of bers - maskit slices , ann
32 ( 2007 ) , 179 - 198 .
y. komori and c. series , pleating coordinates for the earle embedding , ann .
des sciences de toulouse vol .
x , ( 2001 ) , 69105 . c. kourouniotis .
complex length coordinates for quasi - fuchsian groups .
, 41(1 ) ( 1994 ) , 173188 .
r. meyerhoff , a lower bound for the volume of hyperbolic 3-manifolds , can.j.math .
xxxix , no.5 ( 1987 ) , 1038 - 1056 . k. matsuzaki and m. taniguchi , hyperbolic manifolds and kleinian groups , oxford university press ( 1998 ) . c. mcmullen , complex earthquakes and teichmller theory , j. amer .
soc . 11 ( 1998 ) , 283 - 320 .
y. minsky , the classification of punctured - torus groups , ann . of math .
149 ( 1999 ) , 559 - 626 .
otal , sur le coeur convexe varit hyperbolique de dimension 3 , preprint .
j. parker and j. parkkonen , coordinates for quasi - fuchsian punctured torus groups , geometry and topology monographs 1 ( 1998 ) , 451 - 478 . s. p. tan .
complex fenchel - nielsen coordinates for quasifuchsian structures .
, 5(2 ) ( 1004 ) , 239251 .
w. thurston , the geometry and topology of 3-manifolds , princeton university lecture notes ( 1982 ) . | after fixing a marking @xmath0 of a quasifuchsian punctured torus group @xmath1 , the complex length @xmath2 and the complex twist @xmath3 parameters define a holomorphic embedding of the quasifuchsian space @xmath4 of punctured tori into @xmath5 .
it is called the complex fenchel - nielsen coordinates of @xmath4 . for @xmath6 ,
let @xmath7 be the affine subspace of @xmath5 defined by the linear equation @xmath8 .
then we can consider the _ linear slice _
@xmath9 of @xmath4 by @xmath10 which is a holomorphic slice of @xmath4 . for any positive real value @xmath11
, @xmath9 always contains the so called _ bers - maskit slice _
@xmath12 defined in @xcite . in this paper we show that if @xmath11 is sufficiently small , then @xmath9 coincides with @xmath12 whereas @xmath9 has other components besides @xmath13 when @xmath11 is sufficiently large .
we also observe the scaling property of @xmath9 . |
recent new observations of hadrons are in many cases associated with the presence of heavy quarks . in this regard , a systematic study of heavy baryons is important .
so far there are two @xmath21 resonances , @xmath1 and @xmath2 , whose quantum numbers are @xmath22 and @xmath7 , respectively @xcite .
they are both @xmath0-wave bottom baryons , which were studied by capstick and isgur in 1986 using the relativistic quark model @xcite , and their predicted masses are in very good agreement with the lhcb and cdf results @xcite . in the following 30 years
, various phenomenological models have been applied to study these excited bottom baryons , including the constituent quark model @xcite , the relativistic quark model @xcite , the color hyperfine interaction @xcite , the heavy quark effective theory @xcite , the chiral unitary model @xcite , the heavy meson effective theory with @xmath23 corrections @xcite , the heavy quark symmetry in multihadron systems @xcite , the quark pair creation model @xcite , the faddeev approach @xcite , the relativistic flux tube model @xcite , and the lattice qcd @xcite ; see a recent review article for more details @xcite . in this paper
we shall study @xmath0-wave bottom baryons using the method of qcd sum rule @xcite in the framework of heavy quark effective theory ( hqet ) @xcite , which has been successful applied to study heavy mesons containing a single heavy quark @xcite , and heavy baryons also containing a single heavy quark @xcite .
there are also some studies using the method of qcd sum rules but not in hqet @xcite . during the calculations
, we shall take the @xmath24 corrections ( @xmath25 is the heavy quark mass ) into account , and extract the chromomagnetic splitting . until now
, @xmath1 and @xmath2 are the only two @xmath0-wave bottom baryons well observed in the experiments @xcite .
this is quite different from charmed baryons @xcite : the four @xmath0-wave charmed baryons , @xmath26 , @xmath27 , @xmath28 and @xmath29 , are well observed and complete two @xmath4 @xmath30 multiplets ; there are some other candidates , like @xmath31 , @xmath32 and @xmath33 , etc
. however , future experiments , such as lhcb and belleii , are able to search for more bottom baryons , as well as more charmed baryons .
hence , it is a suitable time to carry out this study and write this paper , after we have studied @xmath0-wave charmed baryons in ref .
@xcite which already contains much information . in this paper
, we shall use the same procedures and study @xmath0-wave bottom baryons .
this paper is organized as follows . in sec .
[ sec : sumrule ] , we briefly introduce the interpolating fields for @xmath0-wave bottom baryons , and briefly introduce how to use them to perform qcd sum rule analyses at the leading order and at the @xmath24 order .
the procedures are identical to what we have done in ref .
@xcite for @xmath0-wave charmed baryons . in this paper
we just need to redo numerical analyses using the @xmath34 quark mass instead of the @xmath35 quark mass .
these analyses are done in sec .
[ sec : numerical ] , where we use the bottom baryon doublet @xmath36 $ ] as an example . in sec .
[ sec : summary ] we discuss the results and offer a summary .
@xmath0-wave heavy baryons have been systematically classified in ref .
@xcite , @xmath0-wave charmed baryon interpolating fields have been systematically constructed in ref .
@xcite , and @xmath0-wave bottom baryon fields can be easily constructed by replacing the @xmath35 quark field by the @xmath34 quark field .
the results are briefly shown in fig . [
fig : pwave ] , which is taken from ref .
@xcite and slightly modified . for the notation the reader is advised to consult the discussion in ref .
@xcite .
in this paper we also use the same notation , @xmath37 , to denote the @xmath0-wave bottom baryon field belonging to the baryon multiplet @xmath38 $ ] , where @xmath39 , @xmath0 , and @xmath40 are the total angular momentum , parity and @xmath4 flavor representation ( @xmath5 or @xmath14 ) of the bottom baryons ; @xmath41 and @xmath42 are the total angular momentum and spin angular momentum of the light components ; @xmath43 and @xmath44 denote @xmath43type ( @xmath45 and @xmath46 ) and @xmath44type ( @xmath47 and @xmath48 ) , respectively , where @xmath49 is the orbital angular momentum between the two light quarks , and @xmath50 is the orbital angular momentum between the bottom quark and the two - light - quark system .
their relations are @xmath51 and @xmath52 , where @xmath53 is the spin of the bottom quark .
the explicit forms of @xmath37 have been given in eqs .
( 8)(21 ) of ref .
@xcite , but @xmath54 and @xmath55 therein contain both @xmath56 and @xmath57 components . in this paper
we give the pure @xmath57 fields @xmath58 \mathbb{c } \gamma_t^{\nu } q^b - q^{at } \mathbb{c } \gamma_t^{\nu } [ \mathcal{d}_t^{\mu } q^b ] \big ) \times \gamma^{\alpha_1\alpha_2,\mu\nu } h_v^c \ , , \\ & & j^{\alpha_1\alpha_2}_{5/2,-,\mathbf{6}_f,2,1,\lambda } \\ \nonumber & & = i \epsilon_{abc } \big ( [ \mathcal{d}_t^{\mu } q^{at } ] \mathbb{c } \gamma_t^{\nu } q^b + q^{at } \mathbb{c } \gamma_t^{\nu } [ \mathcal{d}_t^{\mu } q^b ] \big ) \times \gamma^{\alpha_1\alpha_2,\mu\nu } h_v^c \ , , \end{aligned}\ ] ] where @xmath59 is the projection operator : @xmath60 the notations used in these equations are : @xmath61 , @xmath62 and @xmath63 are color indices , and @xmath64 is the totally antisymmetric tensor ; the superscript @xmath65 represents the transpose of the dirac indices only ; @xmath66 is the charge - conjugation operator ; @xmath67 is the light quark field at location @xmath68 , and it can be either @xmath69 or @xmath70 or @xmath71 ; @xmath72 is the heavy quark field , and we have used the fierz transformation to move it to the rightmost place ; @xmath73 is the gauge - covariant derivative ; @xmath74 is the velocity of the heavy quark ; @xmath75 , @xmath76 , and @xmath77 .
we can use these @xmath0-wave bottom baryon interpolating fields to perform qcd sum rule analyses .
the heavy baryon belonging to the multiplet @xmath38 $ ] has the mass : @xmath78 where @xmath25 is the heavy quark mass , @xmath79 is the sum rule result evaluated at the leading order , and @xmath80 is the sum rule result evaluated at the @xmath24 order : @xmath81 here , the coefficient @xmath82 is defined to be @xmath83 it is directly related to the baryon mass splitting within the same multiplet , caused by the chromomagnetic interaction .
the other coefficient is @xmath84^{3/\beta_0 } \approx 0.8 $ ] , where @xmath85 .
the ( analytical ) formulae for @xmath86 , @xmath87 and @xmath88 have been obtained in ref .
@xcite . in this paper
we just need to use these equations to redo numerical analyses but using the @xmath34 quark mass .
there are two free parameters in these equations : the borel mass @xmath65 and the threshold value @xmath89 . following the procedures used in ref .
@xcite , we can obtain a borel window @xmath90 for a fixed threshold value @xmath89 .
we shall not repeat these procedures again .
the other parameter , @xmath89 , can be fixed by requiring that its dependence of the mass prediction be the weakest , which will be done in the next section .
to perform numerical analyses , we use the following values for various condensates and parameters @xcite : @xmath91 besides them , another important parameter is the @xmath34 quark mass , for which we use the @xmath92 mass @xmath93 gev @xcite .
we shall see that this value gives reasonable results . as an example
, we use the bottom baryon doublet @xmath94 $ ] to perform qcd sum rule analyses .
it contains two bottom baryons , @xmath95 and @xmath96 , and the relevant interpolating field is @xmath97 c \gamma_t^\nu d^b - u^{at } c \gamma_t^\nu [ \mathcal{d}_t^{\mu } d^b ] \big ) \sigma_t^{\mu\nu } h_v^c \ , .\end{aligned}\ ] ] we can use this current to perform qcd sum rule analyses , and calculate @xmath98 , @xmath99 and @xmath100 .
this has been done and the results are listed in eq .
( a6 ) of ref .
@xcite , which can be used to further evaluate masses of @xmath95 and @xmath96 through eqs .
( [ eq : mass ] ) and ( [ eq : masscorrection ] ) , as shown below .
c + c + firstly , we take @xmath101 gev and obtain a borel window @xmath102 gev @xmath103 gev producing the following numerical results : @xmath104 whose values correspond to @xmath105 gev and @xmath101 gev . using eqs .
( [ eq : mass ] ) and ( [ eq : masscorrection ] ) , we can further obtain @xmath106 } & = & 11 \mbox { mev } \ , , \end{aligned}\ ] ] where @xmath107 are the masses of the two bottom baryons belonging to the baryon doublet @xmath94 $ ] , and @xmath108 } \equiv m_{\lambda_b(3/2 ^ - ) } - m_{\lambda_b(1/2 ^ -)}$ ] is their mass splitting .
we show variations of @xmath109 and @xmath108}$ ] with respect to the borel mass @xmath65 in the top panel of figs . [
fig : rho3f11 ] and [ fig : rho3f11split ] .
secondly , we change the threshold value @xmath89 and redo these processes .
our criterion is to require that the @xmath89 dependence of the mass prediction be the weakest .
we show variations of @xmath109 and @xmath108}$ ] with respect to the threshold value @xmath89 in the bottom panel of figs .
[ fig : rho3f11 ] and [ fig : rho3f11split ] .
we find that there are non - vanishing borel windows as long as @xmath110 gev , and the @xmath89 dependence is weak and acceptable in the region @xmath111 gev@xmath112 gev . we note that this dependence is the weakest around @xmath113 gev , but we prefer to choosing a suitable region where we can know the related uncertainty .
the results for @xmath114 gev are also shown , and for such cases , we choose the borel mass @xmath65 when the pc ( pole contribution , defined in eq .
( 34 ) of ref .
@xcite ) is around 20% .
finally , we choose @xmath111 gev@xmath112 gev and @xmath102 gev @xmath103 gev as our working regions , and obtain the following numerical results for the baryon doublet @xmath94 $ ] : @xmath115 } & = & 11 \pm 4 \mbox { mev } \ , , \end{aligned}\ ] ] whose central values correspond to @xmath105 gev and @xmath101 gev , and the uncertainties are due to the borel mass @xmath65 , the threshold value @xmath89 , the bottom quark mass @xmath116 and the quark and gluon condensates .
our result for the mass splitting within the same doublet , @xmath108}$ ] , does not depend much on the charm quark mass and the threshold value , so they are produced quite well , with much less theoretical uncertainties , compared to their masses , @xmath109 and @xmath117 .
these results are consistent with experimental masses of @xmath1 ( @xmath118 ) and @xmath2 ( @xmath119 ) as well as their difference @xcite : @xmath120 this suggests that the use of the @xmath92 mass @xmath93 gev @xcite is reasonable . while , if we use the @xmath121 mass @xmath122 gev @xcite , we would obtain significantly lower masses .
c + we also use the bottom baryon doublet @xmath123 $ ] to perform qcd sum rule analyses .
we show the variation of @xmath109 with respect to the borel mass @xmath65 and the threshold value @xmath89 in fig .
[ fig : lambda3f10 ] .
we find that there are non - vanishing borel windows as long as @xmath124 gev , but in this region the @xmath89 dependence is not so weak . we still choose @xmath125 gev @xmath126 gev and @xmath127 gev @xmath128 gev as our working regions , but note that the previous results obtained using the doublet @xmath94 $ ] seem to be more stable ( reliable ) , because the @xmath43mode signal is clearer than the @xmath44one .
in fact the lower panel of fig . [
fig : lambda3f10 ] , which is monotonically increasing , shows that the @xmath44mode is not well observed . this does not , however , say that the @xmath44mode mass is larger than the @xmath43mode mass in a reliable manner , though the current choice of the borel window suggests it .
we obtain the following numerical results @xmath129 } & = & 5 \pm 2 \mbox { mev } \ , , \end{aligned}\ ] ] whose central values correspond to @xmath130 gev and @xmath131 gev .
we note again that the mass splitting within the same doublet is reproduced quite well with much less ( theoretical ) uncertainty than the absolute values .
these results are also consistent with masses of @xmath1 ( @xmath118 ) and @xmath2 ( @xmath119 ) as well as their difference @xcite .
there are two baryons , @xmath1 and @xmath2 , which are the only p - wave bottom baryons well observed in experiments @xcite , and there are no other well established candidates . to evaluate masses of other baryon multiplets ,
we follow the previous procedures used for @xmath94 $ ] as well as for charmed baryons @xcite : 1 .
we use the bottom baryon singlet @xmath133 $ ] to perform qcd sum rule analyses .
we show the variation of @xmath134 with respect to the borel mass @xmath65 and the threshold value @xmath89 in fig .
[ fig : lambda6f01 ] .
we find that there are non - vanishing borel windows as long as @xmath135 gev , and the @xmath89 dependence is weak and acceptable in the region
@xmath136 gev@xmath137 gev . accordingly , we choose @xmath136 gev @xmath138 gev and @xmath139 gev @xmath140 gev as our working regions , and obtain the following numerical result @xmath141 where the central value corresponds to @xmath142 gev and @xmath143 gev .
+ c + 2 .
we use the bottom baryon doublet @xmath144 $ ] to perform qcd sum rule analyses .
we show the variations of @xmath134 with respect to the borel mass @xmath65 and the threshold value @xmath89 in fig .
[ fig : rho6f10 ] .
we find that there are non - vanishing borel windows as long as @xmath145 gev , and the @xmath89 dependence is weak and acceptable in the region @xmath136 gev@xmath137 gev .
hence , we choose @xmath136 gev @xmath138 gev and @xmath146 gev @xmath147 gev as our working regions , and obtain the following numerical results @xmath148 } & = & 4\pm2 \mbox { mev } \
, , \end{aligned}\ ] ] whose central values correspond to @xmath149 gev and @xmath143 gev .
we note again that the mass splitting within the same doublet is reproduced quite well with much less ( theoretical ) uncertainty .
+ c + 3 .
we use the bottom baryon doublet @xmath150 $ ] to perform qcd sum rule analyses .
we show the variation of @xmath134 with respect to the borel mass @xmath65 and the threshold value @xmath89 in fig .
[ fig : lambda6f21 ] .
we find that there are non - vanishing borel windows as long as @xmath151 gev , and the @xmath89 dependence is weak and acceptable in the region
@xmath136 gev@xmath137 gev . accordingly , we choose @xmath136 gev @xmath138 gev and @xmath152 gev @xmath153 gev as our working regions , and obtain the following numerical results @xmath154 } & = & 15 \pm 7 \mbox { mev } \ , , \end{aligned}\ ] ]
whose central values correspond to @xmath149 gev and @xmath143 gev .
we note again that the mass splitting within the same doublet is reproduced quite well with much less ( theoretical ) uncertainty .
+ c + 4 . for flavor partners of @xmath21 and @xmath18 , we use @xmath155 = 0.5 gev , following the relations obtained / used for charmed baryons in ref . @xcite .
5 . other bottom baryon multiplets @xmath156 $ ] , @xmath157 $ ] , and @xmath158 $ ] do not give useful qcd sum rule results , which we shall not discuss to avoid confusions .
we summarize these results obtained using the baryon multiplets @xmath3 $ ] , @xmath13 $ ] , @xmath15 $ ] , @xmath16 $ ] and @xmath17 $ ] in table [ tab : results1 ] . [ cols="^,^,^,^,^,^,^,^,^,^ " , ] [ tab : results1 ]
in this paper we followed ref .
@xcite and studied the @xmath0-wave bottom baryons using the method of qcd sum rule in the framework of hqet .
we note that we just need to redo numerical analyses using the @xmath34 quark mass .
the results are summarized in table [ tab : results1 ] for baryon multiplets @xmath3 $ ] , @xmath13 $ ] , @xmath15 $ ] , @xmath16 $ ] and @xmath17 $ ] .
our results suggest that the two observed states @xmath1 ( @xmath6 ) and @xmath2 ( @xmath119 ) @xcite can be well described by the baryon doublet @xmath3 $ ] .
they are the bottom partners of @xmath26 and @xmath27 @xcite , and also belong to the @xmath4 @xmath5 multiplets of @xmath6 and @xmath7 .
our results suggest that their @xmath4 flavor partners , @xmath8 and @xmath9 , have masses @xmath10 gev and @xmath11 gev , respectively , with a mass splitting @xmath12 mev .
the qcd sum rule results obtained by using baryon doublet @xmath13 $ ] are similar and also consistent with the experimental data @xcite . here we obtain that the @xmath43 mode excitation , @xmath3 $ ] , is lower than the @xmath44 one , @xmath13 $ ] .
this is consistent with our previous analysis of the charm sector @xcite , in contrast with the quark model expectation @xcite .
thus , further studies are needed to clarify their nature .
particularly , it would be useful to study decays and productions of these heavy baryon excitations .
we also studied the @xmath4 @xmath14 multiplets by using the baryon multiplets @xmath15 $ ] , @xmath16 $ ] and @xmath17 $ ] .
our results suggest that the @xmath0-wave bottom baryons @xmath18 , @xmath19 and @xmath20 have ( averaged ) masses about 6.0 gev , 6.2 gev and 6.4 gev , respectively . while checking the status of observed bottom baryons listed in pdg , we notice that progress have been made in the experimental observations .
especially , in the past years , experiments like lhcb and cdf announced the observations of higher radial excitations of bottom baryon @xcite .
we believe that higher radial and orbital excitations will be released in future experiments . with the running of lhcb at 13 tev , experimental exploration to higher radial and orbital excitations of bottom baryons will be a penitential and intriguing research topic .
the present theoretical results of the mass of @xmath0-wave bottom baryons can provide valuable information . in near future
, the joint efforts from experimentalist and theorist will be helpful to identify more and more bottom baryons .
we thank cheng - ping shen and ke - wei wei for useful discussions and careful reading of the manuscript .
thanks profs .
luciano maiani , rinaldo baldini and pierluigi campana for their hospitality during his stay at infn - lnf .
this project is supported by the national natural science foundation of china under grants no .
11205011 , no . 11475015 , no .
11375024 , no . 11222547 , no .
11175073 , and no .
11261130311 , the ministry of education of china ( srfdp under grant no . 20120211110002 and the fundamental research funds for the central universities ) . k. a. olive _ et al . _
[ particle data group ] , review of particle physics , chin .
c * 38 * , 090001 ( 2014 ) .
s. capstick and n. isgur , baryons in a relativized quark model with chromodynamics , phys .
d * 34 * , 2809 ( 1986 ) .
r. aaij _ et al . _
[ lhcb collaboration ] , observation of excited @xmath159 baryons , phys .
lett . * 109 * , 172003 ( 2012 ) .
t. a. aaltonen _ et al . _
[ cdf collaboration ] , evidence for a bottom baryon resonance @xmath160 in cdf data , phys .
d * 88 * , 071101 ( 2013 ) .
h. garcilazo , j. vijande and a. valcarce , faddeev study of heavy baryon spectroscopy , j. phys .
g * 34 * , 961 ( 2007 ) .
g. ortega , d. r. entem and f. fernandez , quark model description of the @xmath161 as a molecular @xmath162 state and the possible existence of the @xmath163 , phys .
b * 718 * , 1381 ( 2013 ) .
t. yoshida , e. hiyama , a. hosaka , m. oka and k. sadato , spectrum of heavy baryons in the quark model , arxiv:1510.01067 [ hep - ph ] .
d. ebert , r. n. faustov and v. o. galkin , masses of excited heavy baryons in the relativistic quark model , phys .
b * 659 * , 612 ( 2008 ) .
m. karliner , b. keren - zur , h. j. lipkin and j. l. rosner , the quark model and @xmath62 baryons , annals phys . * 324 * , 2 ( 2009 ) .
m. karliner and j. l. rosner , prospects for observing the lowest - lying odd - parity @xmath164 and @xmath18 baryons , arxiv:1506.01702 [ hep - ph ] .
w. roberts and m. pervin , heavy baryons in a quark model , int .
j. mod .
a * 23 * , 2817 ( 2008 ) . c. garcia - recio , j. nieves , o. romanets , l. l. salcedo and l. tolos , odd parity bottom - flavored baryon resonances , phys .
d * 87 * , 034032 ( 2013 ) . c. w. xiao ,
j. nieves and e. oset , combining heavy quark spin and local hidden gauge symmetries in the dynamical generation of hidden charm baryons , phys .
d * 88 * , 056012 ( 2013 ) .
w. h. liang , c. w. xiao and e. oset , baryon states with open beauty in the extended local hidden gauge approach , phys .
d * 89 * , 054023 ( 2014 ) .
j. m. torres - rincon , l. tolos and o. romanets , open bottom states and the @xmath165-meson propagation in hadronic matter , phys .
d * 89 * , 074042 ( 2014 ) .
j. x. lu ,
y. zhou , h. x. chen , j. j. xie and l. s. geng , dynamically generated @xmath166 singly charmed and bottom heavy baryons , phys .
d * 92 * , 014036 ( 2015 ) . c. garcia - recio , c. hidalgo - duque ,
j. nieves , l. l. salcedo and l. tolos , compositeness of the strange , charm and beauty odd parity @xmath167 states , arxiv:1506.04235 [ hep - ph ] .
s. yasui and k. sudoh , probing gluon dynamics by charm and bottom mesons in nuclear matter in heavy - meson effective theory with 1/m corrections , phys .
c * 89 * , 015201 ( 2014 ) .
y. yamaguchi , s. ohkoda , a. hosaka , t. hyodo and s. yasui , heavy quark symmetry in multihadron systems , phys .
d * 91 * , 034034 ( 2015 ) . c. mu , x. wang , x. l. chen , x. liu and s. l. zhu , dipion decays of heavy baryons , chin .
c * 38 * , 113101 ( 2014 ) .
a. valcarce , h. garcilazo and j. vijande , heavy baryon spectroscopy with relativistic kinematics , phys .
b * 733 * , 288 ( 2014 ) .
b. chen , k. w. wei and a. zhang , assignments of @xmath168 and @xmath169 baryons in the heavy quark - light diquark picture , eur .
phys . j. a * 51 * , 82 ( 2015 ) .
t. burch , heavy hadrons on @xmath170 and @xmath171 improved clover - wilson lattices , arxiv:1502.00675 [ hep - lat ] . v. crede and w. roberts , progress towards understanding baryon resonances , rept .
phys . * 76 * , 076301 ( 2013 ) .
m. a. shifman , a. i. vainshtein and v. i. zakharov , qcd and resonance physics : theoretical foundations , , nucl .
b * 147 * , 385 ( 1979 ) . l. j. reinders , h. rubinstein and s. yazaki , hadron properties from qcd sum rules , phys .
rept . *
127 * , 1 ( 1985 ) .
b. grinstein , the static quark effective theory , nucl .
b * 339 * , 253 ( 1990 ) .
e. eichten and b. r. hill , an effective field theory for the calculation of matrix elements involving heavy quarks , phys .
b * 234 * , 511 ( 1990 ) .
a. f. falk , h. georgi , b. grinstein and m. b. wise , heavy meson form - factors from qcd , nucl .
b * 343 * , 1 ( 1990 ) .
e. bagan , p. ball , v. m. braun and h. g. dosch , qcd sum rules in the effective heavy quark theory , phys .
b * 278 * , 457 ( 1992 ) .
m. neubert , heavy meson form - factors from qcd sum rules , phys .
d * 45 * , 2451 ( 1992 ) .
m. neubert , heavy quark symmetry , phys .
rept . * 245 * , 259 ( 1994 ) .
d. j. broadhurst and a. g. grozin , operator product expansion in static quark effective field theory : large perturbative correction , phys .
b * 274 * , 421 ( 1992 ) .
p. ball and v. m. braun , next - to - leading order corrections to meson masses in the heavy quark effective theory , phys .
d * 49 * , 2472 ( 1994 ) .
t. huang and c. w. luo , light quark dependence of the isgur - wise function from qcd sum rules , phys .
d * 50 * , 5775 ( 1994 ) .
y. b. dai , c. s. huang , m. q. huang and c. liu , qcd sum rules for masses of excited heavy mesons , phys .
b * 390 * , 350 ( 1997 ) .
y. b. dai , c. s. huang and h. y. jin , bethe - salpeter wave functions and transition amplitudes for heavy mesons , z. phys .
c * 60 * , 527 ( 1993 ) .
y. b. dai , c. s. huang and m. q. huang , @xmath172 order corrections to masses of excited heavy mesons from qcd sum rules , phys .
d * 55 * , 5719 ( 1997 ) .
p. colangelo , f. de fazio and n. paver , universal tau(1/2)(y ) isgur - wise function at the next - to - leading order in qcd sum rules , phys .
d * 58 * , 116005 ( 1998 ) .
y. b. dai , c. s. huang , c. liu and s. l. zhu , understanding the @xmath173 and @xmath174 with sum rules in hqet , phys .
d * 68 * , 114011 ( 2003 ) .
d. zhou , e. l. cui , h. x. chen , l. s. geng , x. liu and s. l. zhu , the d - wave heavy - light mesons from qcd sum rules , phys .
d * 90 * , 114035 ( 2014 ) .
e. v. shuryak , hadrons containing a heavy quark and qcd sum rules , nucl .
b * 198 * , 83 ( 1982 ) .
a. g. grozin and o. i. yakovlev , baryonic currents and their correlators in the heavy quark effective theory , phys .
b * 285 * , 254 ( 1992 ) .
e. bagan , m. chabab , h. g. dosch and s. narison , baryon sum rules in the heavy quark effective theory , phys .
b * 301 * , 243 ( 1993 ) .
y. b. dai , c. s. huang , c. liu and c. d. lu , 1/m corrections to heavy baryon masses in the heavy quark effective theory sum rules , phys .
b * 371 * , 99 ( 1996 ) .
y. b. dai , c. s. huang , m. q. huang and c. liu , qcd sum rule analysis for the @xmath175 semileptonic decay , phys .
b * 387 * , 379 ( 1996 ) .
s. groote , j. g. korner and o. i. yakovlev , qcd sum rules for heavy baryons at next - to - leading order in alpha - s , phys .
d * 55 * , 3016 ( 1997 )
. s. l. zhu , strong and electromagnetic decays of p - wave heavy baryons @xmath176 , @xmath177 , phys .
d * 61 * , 114019 ( 2000 ) .
j. p. lee , c. liu and h. s. song , qcd sum rule analysis of excited @xmath178 mass parameter , phys .
b * 476 * , 303 ( 2000 ) .
c. s. huang , a. l. zhang and s. l. zhu , excited heavy baryon masses in hqet qcd sum rules , phys .
b * 492 * , 288 ( 2000 ) .
d. w. wang and m. q. huang , excited heavy baryon masses to order @xmath179 from qcd sum rules , phys .
d * 68 * , 034019 ( 2003 ) .
x. liu , h. x. chen , y. r. liu , a. hosaka and s. l. zhu , bottom baryons , phys .
d * 77 * , 014031 ( 2008 ) .
e. bagan , m. chabab , h. g. dosch and s. narison , spectra of heavy baryons from qcd spectral sum rules , phys .
b * 287 * , 176 ( 1992 ) .
e. bagan , m. chabab , h. g. dosch and s. narison , the heavy baryons @xmath164 @xmath18 from qcd spectral sum rules , phys .
b * 278 * , 367 ( 1992 ) .
f. o. duraes and m. nielsen , qcd sum rules study of @xmath180 and @xmath181 baryons , phys .
b * 658 * , 40 ( 2007 ) .
z. g. wang , analysis of @xmath182 and @xmath183 with qcd sum rules , eur .
j. c * 54 * , 231 ( 2008 ) .
h. y. cheng , charmed baryons circa 2015 , arxiv:1508.07233 [ hep - ph ] .
h. x. chen , w. chen , q. mao , a. hosaka , x. liu and s. l. zhu , p - wave charmed baryons from qcd sum rules , phys . rev .
d * 91 * , 054034 ( 2015 ) . c. chen , x. l. chen , x. liu , w. z. deng and s. l. zhu , strong decays of charmed baryons , phys .
d * 75 * , 094017 ( 2007 ) .
k. c. yang , w. y. p. hwang , e. m. henley and l. s. kisslinger , qcd sum rules and neutron proton mass difference , phys .
d * 47 * , 3001 ( 1993 ) .
w. y. p. hwang and k. c. yang , phys .
d * 49 * , 460 ( 1994 ) .
s. narison , qcd as a theory of hadrons ( from partons to confinement ) , camb .
monogr . part .
phys . nucl .
cosmol .
* 17 * , 1 ( 2002 ) .
v. gimenez , v. lubicz , f. mescia , v. porretti and j. reyes , operator product expansion and quark condensate from lattice qcd in coordinate space , eur .
j. c * 41 * , 535 ( 2005 ) .
m. jamin , flavour - symmetry breaking of the quark condensate and chiral corrections to the gell - mann - oakes - renner relation , phys .
b * 538 * , 71 ( 2002 ) .
b. l. ioffe and k. n. zyablyuk , gluon condensate in charmonium sum rules with 3-loop corrections , eur .
j. c * 27 * , 229 ( 2003 ) .
a. a. ovchinnikov and a. a. pivovarov , qcd sum rule calculation of the quark gluon condensate , sov .
j. nucl .
phys . * 48 * , 721 ( 1988 ) [ yad .
fiz . * 48 * , 1135 ( 1988 ) ] .
colangelo and a. khodjamirian , _ `` at the frontier of particle physics / handbook of qcd '' _ ( world scientific , singapore , 2001 ) , volume 3 , 1495 .
y. b. dai and s. l. zhu , decay widths of @xmath184 and @xmath185 to the leading order of @xmath172 in hqet , phys . rev .
d * 58 * , 074009 ( 1998 ) .
l. a. copley , n. isgur and g. karl , phys .
d * 20 * , 768 ( 1979 ) [ phys .
d * 23 * , 817 ( 1981 ) ] . | we study the @xmath0-wave bottom baryons using the method of qcd sum rule and heavy quark effective theory .
our results suggest that @xmath1 and @xmath2 can be well described by the baryon doublet @xmath3 $ ] , and they belong to the @xmath4 @xmath5 multiplets of @xmath6 and @xmath7 .
their @xmath4 flavor partners , @xmath8 and @xmath9 , have masses @xmath10 gev and @xmath11 gev , respectively , with mass splitting @xmath12 mev .
the results obtained using baryon doublet @xmath13 $ ] are similar and also consistent with the experimental data .
we also study the @xmath4 @xmath14 multiplets by using the baryon multiplets @xmath15 $ ] , @xmath16 $ ] and @xmath17 $ ] , and our results suggest that the @xmath0-wave bottom baryons @xmath18 , @xmath19 and @xmath20 have ( averaged ) masses about 6.0 gev , 6.2 gev and 6.4 gev , respectively . |
in cerium , ytterbium , uranium and other anomalous rare - earth and actinide compounds , the interplay between kondo effect and magnetism leads to the formation of various interesting phenomena which still attract large attention .
@xcite both effects depend strongly on the hybridization between @xmath0 and conduction electrons , which in turn significantly influence the level of localization of the @xmath0-electrons .
the extent of the localization is sensitive to various external parameters , such as temperature and pressure , but most importantly to the spatial extension of the orbitals .
actually , in the case of cerium compounds , 4@xmath0-electrons are usually well localized , while in the case of uranium and other actinide compounds , 5@xmath0-electrons can be either localized or itinerant , or in - between , depending on the studied system .
moreover , in the case of 5@xmath0-electrons , the experimental data do not clearly distinguish between a localized 5@xmath7 configuration and a mixed - valence regime .
thus , there is always a challenge to decide which is the best framework for discussing actinide materials .
the nature of the electronic structure of actinide metals has been studied since a long time and extensively reviewed .
@xcite the difference between the @xmath8- and @xmath4-electrons leads to different magnetic properties of rare - earth and actinide compounds .
@xcite in the case of cerium kondo compounds , a competition between the kondo effect on each ce atom and the magnetic ordering of the ce magnetic moments through the rkky interaction has been successfully described by the so - called `` doniach diagram '' .
@xcite in this diagram , both the nel temperature @xmath9 ( or the curie temperature @xmath10 ) and the kondo temperature @xmath11 are obtained as functions of the kondo exchange interaction constant , @xmath5 . in most of the cerium compounds , the magnetic ordering temperatures , @xmath9 or @xmath10 , are rather low , typically of order 5 to 10 k. with increasing pressure ( i.e. with increasing @xmath5 ) ordering temperatures pass through a maximum and then tend to zero at the quantum critical point , above which the systems are non magnetic heavy fermion ones . the situation in actinide compounds is more complex , and it is now established that some of them exhibit a coexistence of magnetic order and kondo effect . indeed , this phenomenon was observed in several uranium compounds , like ute,@xcite ucu@xmath12sb@xmath13,@xcite or uco@xmath14sb@xmath13@xcite in which a ferromagnetic order with large curie temperatures ( equal , respectively , to @xmath15 k , 113 k and 64.5 k ) and a logarithmic kondo - type decrease of the resistivity above @xmath10 has been experimentally detected . a similar behavior has been observed in the neptunium compounds ,
npnisi@xmath13 @xcite and np@xmath13pdga@xmath16,@xcite with curie temperatures equal to , respectively , @xmath1751.5 k and 62.5 k. the origin of this fundamentally different behaviors lies in the fact that 5@xmath0-electrons are generally less localized than the 4@xmath0-electrons and have tendency to partial delocalization . in the latter case ,
5@xmath0-electrons have dual , localized and delocalized , nature which leads to reduction of magnetic moments .
both phenomena have been experimentally detected .
for example , magnetic moments observed in ute are substantially smaller than the free - ion values for either the @xmath1 or the @xmath18 configurations .
@xcite on the same side , in the series of uranium monochalcogenides , us lies closest to the itinerant side for the @xmath4-electrons , use is in the middle and the @xmath4-electrons are more localized in ute , as evidenced by magnetization measurements.@xcite moreover , the curie temperature of ute is passing through a maximum and is then decreasing with applied pressure , which is interpreted as a weak delocalization of the @xmath4-electrons under pressure.@xcite the dual nature of the @xmath4-electrons , assuming two localized @xmath4-electrons and one delocalized electron , has been also studied by zwicknagl _
@xcite who have accounted , by using band calculations , for the mass enhancement factor experimentally observed in some uranium compounds .
another important difference between magnetic cerium and uranium compounds lies in the value of the @xmath4-electron spins @xmath19 which is always larger than 1/2 in uranium systems ( to avoid any confusion , in the following @xmath19 designs always the spin of the @xmath0-localized electrons and @xmath20 the spin of the conduction electrons ) . for spins @xmath19 larger than 1/2
, kondo effect is more complex and depends on the number of screening channels : the underscreened kondo impurity problem and more generally the multichannel impurity problem have been studied extensively @xcite and solved exactly by the bethe ansatz method .
@xcite it has been clearly established that the large spin @xmath19 of the @xmath0-electrons can be completely screened at @xmath21 by spins of conduction electrons only if the number of screening channels @xmath22 is equal to @xmath23 . if this is not the case , the problem is more complicated as shown by nozieres and blandin .
@xcite here we are interested by the case of the underscreened kondo effect , which occurs when @xmath22 is smaller than @xmath23 ; in this case , in contrast to the regular kondo impurity case , the spin is only partially screened at low temperatures , and this leads to a reduced effective spin @xmath24 . for a spin @xmath2 , this happens if there is only one active screening channel , i.e. , when only one conduction band is present near the fermi level .
indeed , for the systems studied here , this is an oversimplification : generally all screening channels are not equivalent , thus , it is natural to consider that only one channel is coupled more strongly to the local spin .
this channel will dominate the behavior , but other screening channels might play a role at lower temperatures , resulting in a two - stage kondo screening with two kondo temperatures.@xcite very recently , the underscreened impurity problem has been widely discussed in relation with experiments performed on quantum dots coupled to ferromagnetic leads,@xcite and in molecular quantum dots ( so called molecular transistors).@xcite in a concentrated kondo system , these reduced effective spins @xmath25 interact ferro- or antiferromagnetically through rkky exchange interaction , leading to magnetic ordering of the reduced moments .
thus coexistence of magnetic ordering and kondo effect is expected to occur more easily in the underscreened case than in the standard @xmath26 kondo lattice .
the first attempt to describe the coexistence of ferromagnetism and kondo effect in uranium compounds has been performed with the help of an underscreened kondo lattice ( ukl ) model which considered localized @xmath0-spins @xmath19=1 to describe a 5@xmath27 configuration of uranium ions.@xcite this model describes the kondo interaction , @xmath5 , between localized @xmath19=1 spins and @xmath3 spins of conduction electrons , and an inter - site ferromagnetic exchange interaction between the @xmath0-spins , @xmath6 .
the mean - field treatment of the ukl model gives an analog of the doniach phase diagram , and qualitatively accounts for the coexistence of ferromagnetism and kondo effect .
however , this model is based on the assumption of localized 5@xmath0-electrons .
of course , this assumption makes a constraint on systems which can be described by the ukl , because , as we already discussed , many metallic actinides do not have fully localized electrons . to improve the model for these compounds
, one needs to include in the ukl a possibility for 5@xmath0-electron delocalization .
this is the main goal of the present study .
the article is organized as follows : we start by considering , in section [ sec : sw ] , the underscreened anderson lattice ( ual ) model , in which the charge transfer is present from the beginning , and we transform it by using the schrieffer - wolff transformation@xcite for @xmath28 , which allows for an effective @xmath0-band term , in addition to the exchange kondo interaction between the @xmath2 @xmath0-spins and the @xmath3 spins of the conduction electrons . then , in section [ sec : mf ] we present the mean field treatment of the model and in section [ sec : rc ] we compute the curie and kondo temperatures as a function of the different parameters and in particular of the @xmath0-bandwidth .
we obtain new phase diagrams which could account for the pressure dependence of uranium systems such as ute compound .
the standard model to describe the physics of the heavy fermion compounds is the periodic anderson lattice model whose hamiltonian can be written as : @xmath29 the first term describes a conduction @xmath30-electron band : @xmath31 where @xmath32 creates a conduction quasiparticle with spin @xmath33 and momentum @xmath34 , and @xmath35 is the energy of conduction electrons .
the term @xmath36 includes all local energy terms of the @xmath0-electrons and it is given by : @xmath37~ , \ ] ] where @xmath38 is the energy of the two - fold degenerate @xmath0-level and @xmath39 is the number operator for @xmath0-electrons on lattice site @xmath40 , orbital @xmath41 and spin @xmath33 , @xmath42 and @xmath43 are the coulomb repulsions among electrons in the same and in the different orbitals , respectively , and @xmath44 is the hund s coupling constant . for @xmath45 electrons per site ,
the ground state of @xmath36 is the triplet state with @xmath2 .
we assume here that this triplet state is much lower in energy than the singlet states .
this is achieved if @xmath46 is much smaller than @xmath47 and than @xmath42 .
we study the sw transformation in this limit , assuming that 2 electrons on the same site can only be coupled in the @xmath2 state . both subsystems , localized @xmath0-electrons and conduction band , are coupled via a hybridization term , @xmath48 where @xmath49 and @xmath50 are creation and annihilation operators for @xmath0-electrons , carrying spin and orbital indexes , @xmath33 and @xmath51 , respectively . in the schrieffer - wolff transformation ,
the @xmath52 hybridization term is eliminated by a canonical transformation .
thus , using the classical method explained in ref .
, we start the procedure by writing the hamiltonian as : @xmath53 where @xmath54 .
then , we look at the scattering of an initial state @xmath55 to a final state @xmath56 , through an intermediate state @xmath57 , where the states @xmath55 , @xmath56 and @xmath57 are eigenstates of @xmath58 and @xmath59 , @xmath60 and @xmath61 are their eigenvalues , respectively .
the sw transformation consists in replacing the @xmath62 term of the hamiltonian by an effective interaction which is of second order in the hybridization parameter @xmath63 of the starting hamiltonian @xmath62 .
the detailed description of the calculations can be found in ref . and .
the resulting effective hamiltonian is given by : @xmath64 where : @xmath65 the sw transformation has been initially performed for the case of one @xmath8-electron in the @xmath66 configuration . here , we will present a derivation of the sw transformation in the case of a @xmath1 configuration , corresponding to a @xmath0-spin @xmath2 , and to the so - called `` underscreened kondo lattice '' model where the @xmath2 spins are coupled to a non - degenerate conduction band .
we study the sw transformation for 2 @xmath0-electrons per site and allowing fluctuations of the number of @xmath0-electrons between 1 and 2 .
several interactions are generated by the sw transformation.@xcite below we present the derivation of the most relevant terms , leading to both local and intersite effective interactions , namely the kondo interaction and the effective hopping of @xmath0-electrons .
first , we derive the @xmath67 exchange hamiltonian for @xmath2 localized @xmath0-spins .
the corresponding eigenstates of @xmath58 are , therefore , given by : @xmath68 and the corresponding eigenvalues are @xmath69 the sw transformation leads to the standard kondo - like @xmath67 exchange hamiltonian , but here with @xmath0-spins @xmath2 , and it is given by : @xmath70\end{aligned}\ ] ] where the different components of the spin @xmath2 read : @xmath71 and the corresponding exchange integral is : @xmath72 this exchange integral @xmath73 can be easily simplified , because @xmath35 can be restricted to values very close to the fermi energy .
then , the difference between @xmath34 and @xmath74 can be neglected in both the values of @xmath35 and of @xmath75 .
we also assume that the mixing parameter does not depend on the orbital index @xmath41 .
thus , @xmath73 can be approximated in the following by : @xmath76 where @xmath77 is the fermi level and @xmath78 is the value of @xmath63 at the fermi level .
gives the definition of the kondo exchange interaction , @xmath5 that we will use in the following .
it is also interesting to notice that the kondo effect is large when the energy @xmath79 is very close to the fermi energy .
in fact the denominator in eq . is the energy difference between the ground state energy @xmath80 of two @xmath0-electrons in @xmath55 and @xmath56 states and the energy of the intermediate state @xmath57 with one @xmath0-electron : @xmath81 . among the different terms emerging from the sw transformation we consider in detail those that correspond to a non - local interaction involving 2 different sites @xmath40 and @xmath82 . in this case , the relevant initial and final states @xmath55 and @xmath56 are two - sites states with a total occupation of three @xmath0-electrons , allowing charge fluctuations between sites @xmath40 and @xmath82 .
consequently , an effective @xmath0 bandwidth is obtained in the @xmath83 order in @xmath84 .
thus , we derive an effective band hamiltonian , @xmath85 , as the sum of three terms which arise from the schrieffer - wolff transformation and where the sum over @xmath40 and @xmath82 refers to different sites : @xmath86 where : @xmath87 @xmath88 @xmath89 these three terms can be simplified in the mean field approximation by introducing average occupation numbers @xmath90 . for simplicity , we will neglect all interorbital transfer terms . with these assumptions , we can see that the @xmath85 terms can be considered as effective @xmath0-hopping terms between @xmath40 and @xmath82 sites with a spin dependent hopping ( see next section ) .
if the @xmath34 dependence of @xmath84 is neglected , as is often assumed , one can deduce from eqs .
, and that there is no intersite hopping .
thus , the @xmath34 dependence of @xmath84 is at the origin of the effective @xmath0-bandwidth and has to be taken into account .
this @xmath34 dependence is due to non - local hybridization between @xmath0 and @xmath30 electrons , which has no influence on the kondo interaction , but has to be taken into account for intersite interactions . in the following we write the coefficient which enters in the expression of @xmath85 as : @xmath91 where the function @xmath92 includes the @xmath34-dependence of @xmath93 , while the @xmath94-dependence of @xmath35 is not essential here , but could be also included . finally , the resulting transformed hamiltonian contains two terms : @xmath95 , that gives the kondo exchange interaction for spin @xmath2 and the important and new term @xmath85 , which can be considered as an effective band term for the @xmath4-electrons .
thus , in addition to the kondo interaction , we have derived here a term which leads to a finite @xmath0-bandwidth ; this newly introduced term will be shown as giving a better description of uranium and actinide compounds where the @xmath4-electrons are less localized than the @xmath8-electrons in rare earth compounds . besides this effective band term , there are other intersite interactions which lead to rkky exchange , but they arise only to @xmath96 order in hybridization @xmath84 .
we do not compute all these terms , but , instead , we will introduce them phenomenologically as an additional intersite exchange parameter of the model , @xmath6 .
combining all terms obtained in the preceding section , we can write the new effective hamiltonian as : @xmath97 the heisenberg interaction @xmath6 is considered here as a ferromagnetic exchange between nearest neighbors only . in fact
rkky interactions are long range and oscillating , but since our aim is to study the coexistence of ferromagnetism and kondo effect , we consider only ferromagnetic interactions . the mean field approach has been previously described in ref . .
for the kondo part @xmath95 , it is based on a generalization of a functional integration approach described by yoshimori and sakurai for the single impurity case @xcite and by lacroix and cyrot for the @xmath26 kondo lattice case.@xcite here we introduce the following mean field parameters : the average occupation numbers @xmath90 , @xmath98 and the `` kondo '' parameter @xmath99 . we restrict ourselves to uniform solutions in which all these quantities are site and orbital independent , i.e. , @xmath100 , @xmath101 and @xmath102 .
thus , @xmath85 can be written as : @xmath103 the effective band dispersion depends on spin but not on the orbital indexes . then : @xmath104 with @xmath105~ , \nonumber \\\end{aligned}\ ] ] and where @xmath92 is the dispersion relation for the @xmath0-band .
we assume , for simplicity , that the @xmath0-band dispersion is similar to the conduction electron dispersion , i. e. , @xmath106 .
we also assume that the @xmath0-band should be narrower than the conduction band , so @xmath107 .
parameter @xmath108 can be included in the local energy @xmath38 , and @xmath109 is a multiplicative coefficient that can be absorbed in the definition of @xmath110 .
then we have : @xmath111 with @xmath112~,\end{aligned}\ ] ] , @xmath113 , @xmath114 , @xmath115 and @xmath28.,scaledwidth=40.0% ] -electrons @xmath116 , magnetization of @xmath30-electrons @xmath117 , and kondo parameters @xmath118 for both spin directions for cases ( b ) and ( c ) ( see the text for definitions ) , with @xmath114 , @xmath115 and @xmath28.,scaledwidth=40.0% ] the total hamiltonian can be now written in the mean field approach as follows : @xmath119 where we have : @xmath120\nonumber \\ & \quad+\frac{j_k}{2}n ( \lambda_{\uparrow } + \lambda_{\downarrow})^2-\frac{j_h}{2}zn ( m^f)^2 \nonumber \\ & \quad - j_k n m^c m^f~,\end{aligned}\ ] ] with @xmath121 , @xmath122 and @xmath123 .
the diagonalization of the hamiltonian gives one pure @xmath0-band , @xmath124 given by : @xmath125 and two hybridized bands @xmath126 given by : @xmath127~,\end{aligned}\ ] ] with @xmath128 ^ 2 + 8(\lambda_{\sigma})^2}~.\end{aligned}\ ] ] on fig .
[ figdos ] , we present a typical band structure resulting from the three bands @xmath129 and @xmath130 for @xmath113 and @xmath114 . in all figures presented here , @xmath131 , where @xmath132 is the number of nearest neighbors in a simple cubic lattice . in our calculations ,
the values of @xmath5 and @xmath6 are defined in units of the half bandwidth @xmath133 of the conduction band .
one can see on fig .
[ figdos ] the important effect of the finite @xmath0-bandwidth : the band structure is very different from that without any @xmath0-bandwidth used in ref . .
in this section , we present numerical results obtained from this model , using the general method described in detail in ref .
: we derive the green functions and we calculate self - consistently the magnetization @xmath116 for the @xmath0-electrons , the magnetization @xmath117 for the conduction electrons and the two spin - dependent @xmath118 parameters which describe the kondo effect , by imposing constraints on the total number of @xmath0-electrons and conduction electrons , @xmath28 , and @xmath134 , respectively . having solved the self - consistent equations , we study various properties of the model .
the curie and kondo temperatures are defined , within this mean field approach , as the temperatures at which respectively the magnetizations or the @xmath118 parameters tend to zero .
-electrons @xmath116 , magnetization of @xmath30-electrons @xmath117 , and kondo parameter @xmath135 for cases ( a ) , ( b ) and ( c ) , with @xmath114 , @xmath115 and @xmath28 . here
@xmath136.,scaledwidth=40.0% ] as already mentioned in the previous section , the half @xmath0-bandwidth derived from the schrieffer - wolff transformation is spin - dependent and it is given by @xmath137 , eq . .
@xmath137 can be rewritten as : indicating that the effective bandwidth and the magnetization are correlated .
in fact it can be easily checked that the bandwidth for @xmath139-spin increases with magnetization while it is the opposite for @xmath140-spin : this is consistent with the double exchange process in the ferromagnetic phase , which favors itinerancy of the conduction electrons with spin parallel to the localized moment , because of intra - atomic hund s coupling . here
, however , we would like to explore the parameter dependence of the effective bandwidth including different possibilities for the relative variation of the kondo coupling , @xmath5 , and the @xmath0-bandwidth @xmath141 . in order to do that
, we considered also the following definitions of @xmath141 : * case ( a ) : a constant bandwidth : @xmath142 ; * case ( b ) : a bandwidth @xmath141 proportional to the kondo coupling constant : @xmath143 ; in this way we can take into account the effect of pressure on both the bandwidth and the kondo coupling , since both are sensitive to the increase of hybridization under pressure .
* case ( c ) : a bandwidth directly obtained from the sw transformation . from eqs . -
, we get a spin dependent bandwidth : @xmath144 . on fig .
[ figq012p20 ] , we present a plot of the temperature variation of the kondo correlations @xmath135 and @xmath147 , and also the @xmath0 and @xmath30 magnetizations , @xmath116 and @xmath117 , for cases ( b ) and ( c ) .
the parameters are @xmath115 and @xmath114 .
upper plot is for case ( b ) with @xmath148 , while lower plot is for case ( c ) with @xmath149 .
the two magnetization curves clearly show a second order magnetic phase transition at @xmath10 , below which @xmath117 and @xmath116 are always antiparallel , as expected because @xmath5 is an antiferromagnetic coupling . at low temperatures , kondo effect and ferromagnetism coexist together , and , due to the breakdown of spin symmetry at @xmath10 , @xmath135 and @xmath147 become slightly different in the ferromagnetic phase .
we define the kondo temperature as the temperature where @xmath135 and @xmath147 vanish .
the fact that the kondo parameter vanishes at a particular temperature is a well known artefact of the mean field approximation .
actually @xmath11 is a crossover temperature , associated with the onset of kondo screening . in all cases ,
the kondo temperature @xmath11 is larger than the curie temperature @xmath10 and we never found situations where @xmath150 : once ferromagnetism is established , kondo effect does not appear below @xmath10 ; it is blocked by the effective internal magnetic field . to investigate the effect of the pressure on the kondo and curie temperatures , @xmath11 and @xmath10 , respectively , we computed these characteristic temperatures for various values of @xmath5 for fixed values of exchange interaction @xmath6 and number of conduction electrons @xmath151 .
3a,3b and 3c are obtained for three different characterizations of the f - bandwidth , cases ( a ) , ( b ) and ( c ) , respectively .
we notice that the temperatures @xmath11 and @xmath10 are obtained as the temperatures at which the mean field parameters ( @xmath0 and @xmath30 magnetizations and the kondo parameters @xmath152 ) vanish . in the three cases we note
that the kondo temperature @xmath11 becomes non - zero only above a critical value @xmath153 which varies from case to case . in all cases ,
once non - zero , the kondo temperature rapidly increases for larger values of @xmath5 .
the curie temperature , @xmath10 , is non - zero above a given @xmath5 value in case ( a ) , below a given @xmath5 value for case ( b ) and is non - zero for all studied values of @xmath5 for case ( c ) .
the reason for these different behaviors is easy to understand . in case
a ) the f - bandwidth is constant and the system needs a finite value of @xmath5 to get magnetic ordering because f - electrons are itinerant even at small @xmath5 . in case
b ) the f - bandwidth increases linearly with @xmath5 , so , for low values of @xmath5 the f - electrons are localized and they are magnetic even for @xmath154 ; thus as soon as @xmath5 is different from zero , magnetic ordering occurs .
when increasing @xmath5 the f - bandwidth also increases , and magnetism is destroyed because of the itinerant character of the f - electrons .
finally in case c ) the f - bandwidth depends on both @xmath5 and magnetization , and the dependence is different for up and down spin electrons;it can be seen on figure 3 that this complex dependence of the bandwidth leads to small variation of the curie temperature , with a weak maximum .
however a crossing point , at which @xmath10 and @xmath11 are equal is obtained in all cases . concerning the kondo effect , in all cases ( a ) , ( b ) and ( c )
a peculiar behavior has been obtained for values of @xmath5 just above this crossing point : at the temperature @xmath155 indicated on fig .
[ figtctk]a ) , [ figtctk]b ) and [ figtctk]c ) the kondo parameters vanish , being non - zero only between @xmath155 and @xmath11 . to better understand this behavior ,
we have plotted on fig .
[ fig4 ] , @xmath116 , @xmath117 and @xmath135 for @xmath115 , @xmath114 but for values of @xmath5 near the crossing of @xmath10 and @xmath11 , i.e. , @xmath156 for case ( a ) , @xmath157 for case ( b ) and @xmath158 for case ( c ) .
it appears clearly that , with decreasing temperature , the kondo effect occurs first , then there is a coexistence of kondo effect and ferromagnetism , and finally the kondo effect disappears to yield only a strong ferromagnetism at extremely low temperatures .
this behavior can be interpreted in the following way : kondo effect for a spin @xmath2 can not be complete , as explained in the introduction .
thus if exchange is large enough , the ordering of the remaining @xmath0-moments occurs in the kondo phase .
however , at lower temperature , when these magnetic moments are large , they act as an internal magnetic field that destroys the kondo effect .
it should be pointed out that there is at present no experimental evidence in favor or in contrast of such an effect in actinide compounds at very low temperature .
another interesting result that can be pointed out is the decrease of the curie temperature for large @xmath5 above the intersection point particularly in case ( b ) , but also within a small range of value of @xmath5 in case ( c ) .
this decrease can probably be considered as resulting from the `` delocalization '' of the 5@xmath0-electrons .
let us also remark that @xmath5 increases with increasing pressure and that the two figs .
[ figtctk]b ) and [ figtctk]c ) can give a description of the experimentally observed variation of @xmath10 with pressure in ute compound , which is passing through a maximum and then decreasing with applied pressure.@xcite to summarize our paper , the present work improves the previous @xmath2 ukl model of ref . by explicitly including the effect of a weak delocalization of the @xmath4-electrons .
within this improved model , we have obtained new phenomena in the region where @xmath10 and @xmath11 are of the same order of magnitude : a possible disappearance of the kondo effect at low temperature , which is a direct consequence of the underscreened kondo effect , and a maximum of @xmath10 as a function of @xmath5 .
it is worth to note that , in our model , the delocalization of the @xmath4-electrons increases when @xmath5 increases , i.e. when pressure is applied ; then the magnetization decreases and in the same way the curie temperature .
therefore , the change in the curie temperature at large @xmath5 is more influenced by delocalization than by competition between magnetism and kondo effect .
this is a different result compared with the case of cerium compounds , where the magnetization is destroyed by kondo effect , i.e. by the screening of the magnetic moment . in the underscreened @xmath2 kondo lattice , because kondo screening can never be complete , the kondo effect alone does not destroy ferromagnetism . to conclude
, we have shown that our model includes two effects which are essential to describe the @xmath4-electron compounds : the small delocalization of the @xmath4-electrons , and the @xmath2 spins found in uranium or neptunium compounds .
the first effect works against magnetism , while the second one favors magnetism .
the competition between these two effects leads to complex phase diagrams which can improve the description of some actinide compounds and explain in particular the maximum of @xmath10 observed experimentally in ute compound with increasing pressure . | magnetic properties of uranium and neptunium compounds showing the coexistence of kondo screening effect and ferromagnetic order are investigated within the anderson lattice hamiltonian with a two - fold degenerate @xmath0-level in each site , corresponding to @xmath1 electronic configuration with @xmath2 spins .
a derivation of the schrieffer - wolff transformation is presented and the resulting hamiltonian has an effective @xmath0-band term , in addition to the regular exchange kondo interaction between the @xmath2 @xmath0-spins and the @xmath3 spins of the conduction electrons .
the obtained effective kondo lattice model can describe both the kondo regime and a weak delocalization of @xmath4-electron . within this model
we compute the kondo and curie temperatures as a function of model parameters , namely the kondo exchange interaction constant @xmath5 , the magnetic intersite exchange interaction @xmath6 and the effective @xmath0-bandwidth .
we deduce , therefore , a phase diagram of the model which yields the coexistence of kondo effect and ferromagnetic ordering and also accounts for the pressure dependence of the curie temperature of uranium compounds such as ute . |
the standard model ( sm ) of the strong and electroweak interactions is the present paradigm of particle physics .
its validity has been tested to a level better than one per mille at particle accelerators @xcite .
nevertheless , there are arguments against the sm being the fundamental model of particle interactions @xcite , giving rise to the investigation of competing alternative or extended models , which can be tested at high - energy colliders , such as the large hadron collider ( lhc ) @xcite , or a @xmath10 linear collider ( lc ) @xcite .
one of the most promising possibilities for physics beyond the sm is the incorporation of supersymmetry ( susy ) , which leads to a renormalizable field theory with precisely calculable predictions to be tested in future experiments .
the simplest supersymmetric extension of the sm is the minimal supersymmetric standard model ( mssm ) @xcite . up to now the major effort on the computation of susy radiative corrections
has been put into the computation of virtual susy effects in observables that involve only sm external particles , or into the calculation of loop effects in the extended higgs sector of the mssm . in this context , if the masses of the extra non - standard particles are very large as compared to the sm electroweak scale , the effects of these particles decouple , leaving the sm as a low - energy effective theory .
this means that if the extra particles are too heavy we could not discern between the sm and the mssm by just looking at the low - energy end of the spectrum , since the only trace of the mssm would be a light higgs boson ( @xmath11 ) @xcite , whose properties would not differ from the sm one .
but for the case of direct production of susy particles , one also needs a detailed knowledge of the higher - order effects for the processes with these susy particles in the external states .
in contrast to the case of virtual susy effects in sm green s functions , there is a great variety of additional electroweak mssm processes , viz .
those involving higgs bosons and/or ( r - odd ) sparticles in the external legs , for which the decoupling limit can not be applied . in this case
several kinds of non - decoupling effects may appear which grow with the mass of the sparticles @xcite .
these effects can be very important , as they could provide the clue to discovering susy physics in the colliders .
this was amply demonstrated in the past for decay processes @xcite , and also very recently for production cross - sections in hadron colliders@xcite , in both cases exploiting the susy threshold corrections in the top quark and higgs boson sector . in the present study , however , we will face not only threshold effects , but also a new type of non - decoupling ( so - called universal ) contributions .
both types of non - decoupling effects have to be considered for a complete study of sfermion decays in the colliders .
the lhc will be able to discover new particles with masses up to @xmath12 .
provided they are not too heavy , the lc will be able to make precision measurements of their properties .
for example , at a @xmath13 lc with a total integrated luminosity of @xmath14 a measurement of the top - squark mass and the top - squark mixing angle can be performed with a precision of @xmath15 and @xmath16 respectively @xcite . for an adequate analysis ,
precise theoretical predictions are required , going beyond the born approximation .
these studies of purely supersymmetric processes at the quantum level are necessary not only to refine the prediction of the corresponding observables but also because 1 ) the quantum corrections may severely affect the physical production of the supersymmetric particles , and 2 ) in addition they probe the underlying susy nature present in the model , that is : the relation between the gauge couplings of the sm gauge bosons and the yukawa coupling of its susy partners ( charginos and neutralinos ) . beyond leading order this relation receives corrections which are non - decoupling .
a number of studies have already addressed this issue , for production as well as for decay processes . for squark and gluino production in hadron collisions ,
the nlo qcd corrections are available @xcite ; for squark - pair production in @xmath10 collisions , the nlo qcd are also known , together with the yukawa corrections @xcite . concerning the subsequent squark decays into charginos / neutralinos ,
the qcd corrections were presented in @xcite , was studied in @xcite . here
we will assume @xmath17 .
] , whereas the yukawa corrections were given in @xcite . in this
last work large corrections were found .
they were derived , however , in the _ higgsino _ approximation for the chargino ; hence , a full computation is required to consolidate the significance of the loop effects .
we have performed a complete one - loop computation of the electroweak radiative corrections to the partial decay widths of sfermions into fermions and charginos / neutralinos , @xmath18 we present the structure of the corrections in detail , and illustrate their main features and their significance in representative numerical examples .
explicit results are displayed for all kind of sfermions .
first results of this study were presented in ref.@xcite . in processes with exclusively sm particles in external states , it is possible to divide the one - loop contributions into sm - like and non - sm - like subclasses .
this separate treatment is often used in the literature , and it is useful since it allows to make the computation in small steps , checking each sector individually . as a distinctive feature of the radiative corrections to processes with supersymmetric particles in the external legs ,
this separability is lost . in such kind of processes
the ultraviolet ( uv ) divergences of diagrams with virtual sm particles cancel the uv divergences of diagrams with non - sm particles .
any partial computation would yield uv - divergent and thus meaningless results .
for this reason we have to compute the entire set of one - loop contributions for our processes , with the proper counterterms involving the renormalization of almost the full mssm lagrangian . as a direct consequence ,
many non - decoupling effects appear .
moreover , we remark that since we have sparticles in the external legs a consistent calculation of the loop integrals requires the use of dimensional reduction in order that the regularization procedure preserves supersymmetry @xcite .
section [ sec : renor ] contains the renormalization of the sfermion sector ( section [ sec : renorsf ] ) , the chargino / neutralino sector ( section [ sec : renorcn ] ) , and their interaction ( section [ sec : renorlagrangian ] ) . the numerical analysis in done in section
[ sec : numeric ] , including the combination of the electroweak effects with the qcd ones , and the computation of the one - loop corrected branching ratios .
finally section [ sec : conclu ] is devoted to the conclusions .
it is our aim to complement the on - shell scheme of the sm to include the susy particles .
the renormalization of the sm is done according to ref .
@xcite apart from some sign conventions .
however , when extending this renormalization framework to embrace the whole mssm we will treat the field renormalization of the supersymmetric particles in a different way that will be described at due point .
the so - called @xmath19-scheme is used , in which the input parameters for the gauge sector are chosen to be the fine structure constant @xmath19 ( defined in the thomson limit ) and the pole masses of the weak gauge bosons @xmath20 , @xmath21 . the electroweak mixing angle is defined on - shell : , @xmath22 , @xmath23 , etc . ]
@xmath24 .
the higgs sector of the mssm has received a lot of attention in the literature @xcite .
here we follow ref.@xcite .
in fact , the only ingredient of the higgs sector needed at one - loop order for renormalization of the sfermion and the chargino - neutralino sectors is the renormalization of @xmath25 .
the counterterm is determined by the condition for the counterterms of the two vacuum expectation values , @xmath26 together with the absence of mixing between the on - shell @xmath27 higgs boson and the @xmath28 weak gauge boson , which gives @xmath29 our choice of @xmath30 is based on simplicity .
it is known that all definitions of @xmath30 not directly related to a physical observable are subject to some gauge dependence and/or induce large variations of the corrections with the parameters , free of these problems , see e.g. ref .
@xcite . as a drawback , however , one has to compute the process - dependent corrections .
see also @xcite for a recent review on this subject . ] . in our case
the dependence is small and , with this simple and well tested definition , we can avoid introducing process - dependent corrections in a framework , like ours , which is already quite cumbersome .
needless to say , the physical observables we are addressing , like decay rates and branching ratios , are completely insensitive to our particular choice of @xmath30 .
besides parameter renormalization , we introduce field - renormalization constants for each particle . exploiting the freedom in the treatment of field renormalization ,
they have been chosen in a way to get fairly simple expressions for the physical observables under study . throughout this paper
we will be using the 3th family squarks as a generic fermion - sfermion notation .
the same relations hold for sleptons , by changing the corresponding charges appropriately .
we denote the two sfermion - mass eigenvalues by @xmath31 , with @xmath32 . the sfermion - mixing angle @xmath33 is defined by the transformation relating the weak - interaction ( @xmath34 ) and the mass eigenstate ( @xmath35 ) sfermion bases : @xmath36 by this basis transformation , the sfermion mass matrix , @xmath37 becomes diagonal : @xmath38 .
@xmath39 is the soft - susy - breaking mass parameter of the @xmath40 doublet due to @xmath40 gauge invariance . ] , whereas @xmath41 is the soft - susy - breaking mass parameter of the singlet . @xmath42 and @xmath43 are the usual third component of the isospin and the electric charge respectively .
the mixing parameters in the non - diagonal entries read @xmath44 our aim is to compute the radiative corrections in an on - shell renormalization scheme ; hence , the input parameters are physical observables ( i.e. the physical masses @xmath45 , ) rather than formal parameters in the lagrangian ( i.e. the soft - susy - breaking parameters @xmath46 , in eq .
( [ eq : sbottommatrix ] ) ) .
specifically , we use the following set of independent parameters for the squark sector : @xmath47 the value of the other stop mass , @xmath48 , is derived from this set of input parameters .
the sbottom and stop trilinear soft - susy - breaking terms @xmath49 and @xmath50 are fixed at the tree - level by the previous parameters as follows : @xmath51 we impose the approximate ( necessary ) condition @xmath52 where @xmath53 is of the order of the average squark masses for @xmath54 , to avoid colour - breaking minima in the mssm higgs potential@xcite .
of course the relation ( [ eq : abt ] ) receives one - loop corrections ; however , since these parameters do not enter the tree - level expressions , these effects translate into two - loop corrections to the process under study .
the bound ( [ eq : necessary ] ) translates into a stringent constraint to the sbottom - quark mixing angle for moderate and large values of @xmath55 : with an approximate limit @xmath56 from the negative output of the chargino search at lep , the condition ( [ eq : necessary ] ) can only be satisfied by a cancellation of the two terms in ( [ eq : abt ] ) which is easily spoiled when @xmath57 is varied . however , the right hand side of eq .
( [ eq : necessary ] ) is not rigorous @xcite ; so we will present results also when this bound is not satisfied , but we will clearly mark these regions . with the use of the bound ( [ eq : necessary ] ) also the squark - squark - higgs - boson couplings are restricted .
this is a welcome feature , since these couplings can in general be very large , eventually spoiling perturbativity .
the slepton sector follows the same procedure , but only one sneutrino is present . thus the input parameters in eq .
( [ eq : inputb ] ) are reduced to the charged slepton masses and mixing angle .
the definition of the on - shell renormalization scheme is driven by the input parameters ( [ eq : inputb ] ) .
the two bottom - squarks and the lightest top - squark are defined to be on - shell , whereas the heaviest top - squark mass receives quantum corrections .
the mixing angle renormalization must also be given . since there is no unique concept of an _ on - shell angle _
, a practical definition is given , which is general enough to be used in any sfermion observable .
other definitions of an _ on - shell _ scheme are also conceivable .
one could think , for example , of a concept having all the sfermions defined on - shell , so that the input parameters would be the four masses and one mixing angle .
such a scheme , however , is problematic .
first of all , this input parameter set is not complete , and one needs to give also the sign of the angle to determine completely the parameters of the sfermion section .
second , the expression derived for the counterterm of the mixing angle is not well defined when the angle is zero ( no mixing ) .
admittedly , the zero - mixing - angle case could be thought of as an academic limit ; it is , however , a very useful scenario for the study of sfermions at colliders , and has widely been used accordingly .
therefore , a renormalization framework which permits a consistent treatment of the zero - mixing - angle limit is desirable .
this on - shell sfermion - sector renormalization was already introduced in @xcite . here
we expand this renormalization framework to include the relation between bottom - squark and top - squark counterterms , and provide a thorough discussion .
for each squark type @xmath58 we introduce a set of field - renormalization constants as follows , @xmath59 where we have attached a superscript @xmath60 to the bare fields , and @xmath61 is the usual kronecker delta . as for parameter renormalization
, we introduce counterterms for each independent parameter in eq .
( [ eq : inputb ] ) : @xmath62 the bare fields in the interaction basis are related to the bare fields in the mass basis as @xmath63 the counterterms to the soft - susy - breaking parameters in eq .
( [ eq : sbottommatrix ] ) can be found by @xmath64 the bottom - squarks and the lightest top - squark are defined to be on - shell , the residue of the renormalized propagators is taken to be @xmath65 and we require no - mixing between the sfermions , that is @xmath66 the mixing angles in the sfermion sector do not receive radiative corrections .
the corresponding counterterm is fixed by means of the non - diagonal field - renormalization constants of eq .
( [ eq : sqwavecounter ] ) as follows : @xmath67 the set of equations ( [ eq : onshellsfermions ] ) , ( [ eq : deltatheta ] ) defines all the counterterms in the sfermion sector .
the one - loop electroweak feynman diagrams contributing to the sfermions self - energies are shown schematically in fig .
[ diags : self]a .
the heaviest top - squark is not defined on - shell , and therefore its mass receives radiative corrections . in order to find them we make use of eq .
( [ eq : sqdmweak ] ) .
since the bottom - squarks are defined to be on - shell , the counterterm to the soft - susy - breaking squared mass @xmath68 is found to be @xmath69 which is used in the top - squark matrix @xmath70 the fermion mass - counterterms @xmath71 are computed with the help of the feynman diagrams of fig . [ diags : self]b . with these results ,
the counterterm for the @xmath72 mass is found to be @xmath73 note that this renormalization prescription breaks down for @xmath74 . in this case
we have that @xmath75 , whose mass is not related to the bottom - squark ; then it would be better to use a renormalization prescription where @xmath76 is the input parameter .
of course , this renormalization prescription would break down in turn at @xmath77 .
the one - loop on - shell mass for the @xmath72 is then given by @xmath78 in section [ sec : numeric ] we give numerical values for the correction to the top - squark mass .
this sector contains six particle masses , but only three free parameters are available for an independent renormalization . as a consequence
, we are not allowed to impose on - shell conditions for all the particle masses . for the independent input parameters ,
we choose : the masses of the two charginos and the mass of the lightest neutralino , @xmath79 although the tree - level chargino - neutralino sector is well known , we give here a short description , in order to set our conventions .
we start by constructing the following set of weyl spinors : @xmath80 the mass lagrangian in this basis reads @xmath81 where we have defined @xmath82 with @xmath83 and @xmath84 the @xmath40 and @xmath85 soft - susy - breaking gaugino masses .
the four - component mass - eigenstate fields are related to the ones in ( [ eq : inosweak ] ) by @xmath86 where @xmath87 , @xmath88 and @xmath89 are in general complex matrices that diagonalize the mass - matrices ( [ eq : mmassacplus ] ) : @xmath90 the renormalization framework is as follows : we introduce the renormalization constants for each parameter in the mass matrices ( [ eq : mmassacplus ] ) .
the counterterms @xmath91 , @xmath92 , @xmath93 and @xmath94 are fixed from the conditions on the gauge and higgs sectors of the theory , and we must provide renormalization prescriptions for @xmath95 , @xmath96 and @xmath97 .
the mixing matrices @xmath87 , @xmath88 and @xmath89 are defined to diagonalize the renormalized mass matrices .
the bare mass lagrangian is then @xmath98 with @xmath99 note that these counterterm matrices are non - diagonal .
this renormalization corresponds to the following relation between the bare weak- and mass - eigenstates fields : @xmath100 note that no counterterms for the mixing matrices are introduced .
as far as field renormalization is concerned , we introduce different field - renormalization constants for each chargino and neutralino : @xmath101 with the chirality projectors @xmath102 .
the renormalization framework is thus complete or the @xmath103 to perform the renormalization .
we will take the self - energies and the corresponding renormalization constants to be that of the @xmath103 .
one has , evidently , @xmath104 and @xmath105 .
now we must supply renormalization conditions for each parameter in eq .
( [ eq : inputinos ] ) and wave - function renormalization constant .
we will fix @xmath95 and @xmath96 by requiring that the one - loop renormalized chargino masses are the on - shell masses .
@xmath97 is fixed by requiring that the lightest renormalized neutralino mass is the on - shell mass .
the wave - function renormalization constants are fixed by requiring that the renormalized propagator has residue one . using this framework , we are able to relate the counterterms of the fundamental parameters to the mass - counterterms @xmath106 of the charginos . similarly to @xcite we find : @xmath107 .
\label{eq : defcountcha } \end{aligned}\ ] ] the mass counterterms @xmath106 are fixed using the on - shell scheme relation , in the convention of @xcite ( but with opposite sign for @xmath108 ) , @xmath109 where @xmath110 denote the one - loop unrenormalized left- , right - handed and scalar components of the self - energy for the @xmath111th - chargino .
the wave function renormalization constants are @xmath112\ , .
\label{eq : dsrc } \end{aligned}\ ] ] the one - loop diagrams contributing to the chargino / neutralino self - energies are shown schematically in fig .
[ diags : self]c . by solving the equations ( [ eq : defcountcha ] ) we find the counterterms of the independent mass parameters of the chargino mass matrix ( [ eq : mmassacplus ] ) : @xmath113 @xmath97
is determined from the lightest neutralino mass , inverting the relation @xmath114 where the neutralino - mass counterterm @xmath115 is fixed by the on - shell condition for @xmath116 , in analogy to ( [ eq : defdeltami ] ) .
the result is then @xmath117 it is a non - trivial check that with the counterterms determined in eqs .
( [ eq : defcountcha ] ) and ( [ eq : defcountneut ] ) , the one - loop masses for the remaining neutralinos , computed as the pole masses , are uv - finite .
the one - loop on - shell neutralino masses read @xmath118 where now the parameters of eq .
( [ eq : mmassaneut ] ) and the masses and mixing matrices computed in ( [ eq : defuvn ] ) have to be regarded as _ renormalized _ quantities .
the choice of the lightest neutralino to fix the counterterm @xmath97 in ( [ eq : defcountneut ] ) is only efficient if it has a substantial _ bino _ component .
if @xmath119 then @xmath120 , and the extraction of @xmath97 from ( [ eq : soldm1 ] ) would amplify the radiative corrections artificially . in this case
it would be better to extract @xmath97 from the @xmath19th neutralino , such that @xmath121 is large .
this is , however , not relevant for the scenarios which are discussed in this work .
notice also that our renormalization procedure makes use of positive - definite mass eigenvalues for charginos and neutralinos , which require the introduction of some purely - imaginary non - zero elements in the @xmath89-matrix ( [ eq : defuvn ] ) . had we chosen a real @xmath89-matrix , with some negative eigenvalues ,
the various renormalization conditions would be plagued with the explicit sign of the corresponding eigenvalue ( see e.g. @xcite ) . at one - loop , also mixing self - energies between the different neutralinos and charginos
are generated , which we write as follows : @xmath122 with @xmath123 denoting the renormalized two - point functions . for the neutralinos ,
the renormalized self - energies ( [ eq : mixingneut ] ) are related to the unrenormalized ones according to @xmath124 analogous expressions hold for the @xmath125 charginos , replacing @xmath126 in eq .
( [ eq : mixingneut ] ) , the renormalized @xmath125 chargino self - energies being given by @xmath127 the one - loop mixing self - energies also contribute to the chargino and neutralino masses ; their contribution is , however , of higher order in perturbation theory , and we do not take it into account in the mass spectrum .
the contribution of these mixing self - energies to the one - loop decay form factors can be written as follows .
if @xmath128 is the amputated one - particle irreducible 3-point green s function for the creation of a @xmath129 ( represented by a spinor @xmath130 ) , the full one - loop process amplitude reads : @xmath131 where the external mixing wave function factors are @xmath132 the same expression is valid for the creation of a @xmath133 ( anti-@xmath134 , using @xmath135 ) changing the indices @xmath136 , @xmath137 .
( anti-@xmath133 ) are @xmath138 , @xmath139 . ]
admittedly , the choice of the two masses as input parameters is not ideal . for each pair of on - shell masses
@xmath140 , @xmath141 there exists up to a fourth - fold discrete ambiguity in the determination of the underlying parameters @xmath83 and @xmath142 .
there are , however , ways to determine uniquely these parameters . for example , assuming one knows both masses , either from a threshold scan in a lc or from the lhc , one can measure at the lc the production cross - section @xmath143 with different polarizations of the initial state electron and positron @xcite .
one finds in this way experimental values for the mixing matrix elements @xmath144 and @xmath145 which disentangle the ambiguity .
one uses then the derived @xmath83 and @xmath142 parameters to compute the renormalized @xmath87 and @xmath88 matrices , which are used for the computation of the radiative corrections .
a comment is in order regarding other renormalization prescriptions .
note that , at variance with ref .
@xcite , our renormalization prescription does not introduce counterterms for the mixing matrices @xmath87 , @xmath88 and @xmath89
. one can , in fact , introduce these counterterms and fix them in different ways .
for example one could take the point of view that the mixing matrices are functions of the parameters in the mass matrix : @xmath146 and then one computes the counterterms as functions of the counterterms of the mass matrix : @xmath147 the problem with this approach is that , while the analytic form of the chargino functions @xmath148 and @xmath149 are known , the neutralino function @xmath150 is usually computed numerically , and then the computation of @xmath151 is not possible .
we have checked that this renormalization framework gives exactly the same results as the non - introduction of mixing matrix counterterms for the one - loop partial decay widths of sfermions into charginos .
the authors of ref .
@xcite take a different approach , introducing independent renormalization conditions for the counterterms of the mixing matrices . in ref .
@xcite the counterterms to the @xmath87 , @xmath88 and @xmath89 matrices are related to those of the mixing self - energies . when comparing the results presented here with the ones of ref .
@xcite , one should therefore take into account that the meaning of the renormalized parameters @xmath83 , @xmath142 , @xmath84 ( and the mixing matrices ) is not the same .
when comparing physical quantities ( such as pole masses ) , the results should be equivalent at one - loop order . using the notation introduced in the above sections ,
the tree - level interaction lagrangian between fermion - sfermion-(chargino or neutralino ) reads @xcite ) , and in the neutralino couplings .
the change in the couplings allows for a joint presentation of the chargino and neutralino expressions . ]
@xmath152 here we have adopted a compact notation , where @xmath153 is either @xmath154 or its @xmath40 partner for @xmath155 being a neutralino or a chargino , respectively .
roman characters @xmath156 are reserved for sfermion indices and @xmath157 for chargino indices ; greek indices @xmath158 denote neutralinos ; roman indices @xmath159 indicate either a chargino or a neutralino .
for example , the top - squark interactions with charginos are obtained by replacing @xmath160 , @xmath161 , @xmath162 , @xmath163 .
the coupling matrices that encode the dynamics are given by @xmath164 with @xmath165 and @xmath166 the weak hypercharges of the left - handed @xmath40 doublet and right - handed singlet fermion , and @xmath167 and @xmath168 are the yukawa couplings normalized to the @xmath40 gauge coupling constant @xmath169 .
as far as vertex renormalization is concerned , the vertex counterterms are already determined by the renormalization procedure described above . introducing the one - loop counterterms analogously to @xcite we obtain the following counterterm lagrangian @xcite @xmath170 + \dz_{\sfr}^{a}\right\ } { \cal l}_{\chi_r \sfra f ' } + { \cal l}_{\chi_r \sfrb f ' } \dz_{\sfr}^{ba}\nonumber\\ & & + \bigg\{{g}\sfra^{*}\bar{\chi_r}\left[\apaf \frac{1}{2 } \left(\dzrchir+\dzlf\right ) \pl + \amaf \frac{1}{2 } \left(\dzlchir+\dzrf\right)\pr\right]\,f'\nonumber\\ & & + { g}\sfra^ { * } \bar{\chi}_r \left(\delta\apaf \pl+\delta \amaf \pr\right)\,f ' + \mbox { h.c.}\bigg\}\,\,,\,\,(b\neq a)\,\,,\nonumber \\ \delta \apit & = & \delta\rot\vo^*-\left(\lt\delta\rtt+\delta \lt\rtt \right)\vt^*\ , , \nonumber\\ \delta\amit & = & - ( \lb\delta\rot + \rot \delta \lb ) \ut\ , , \nonumber\\ \delta\apat & = & \frac{1}{\sqrt{2 } } \left ( \delta\rot\left(\nt^*+\yl\tw\no^*\right ) + \rot
\yl\no^*\delta\tw + \sqrt{2}\left(\lt\delta\rtt+\rtt\delta\lt\right)\nf^ * \right)\ , , \nonumber\\ \delta\amat & = & \frac{1}{\sqrt{2 } } \left ( \sqrt{2}(\lt\delta\rot+\rot\delta\lt)\nf -\yrt(\tw\delta\rtt+\rtt\delta\tw)\no \right)\ , , \nonumber\\ \delta\apib & = & \delta\rob\uo^*-(\lb\delta\rtb+\rtb\delta\lb)\ut^*\ , , \nonumber\\ \delta\amib & = & -(\lt\delta\rob+\rob\delta\lt)\vt\ , , \nonumber\\ \delta\apab & = & -\frac{1}{\sqrt{2 } } \left ( \delta\rob\left(\nt^*-\yl\tw\no^*\right ) -\rob \yl\no^*\delta\tw -\sqrt{2}(\lb\delta\rtb+\rtb\delta\lb)\nth^ * \right)\ , , \nonumber\\ \delta\amab & = & -\frac{1}{\sqrt{2 } } \left ( -\sqrt{2}(\lb\delta\rob+\rob\delta\lb)\nth + \yrb(\tw\delta\rtb+\rtb\delta\tw)\no \right)\ , .\nonumber\\ \frac{\delta \lb}{\lb}&=&\dmb-\frac{1}{2 } \dmws-\dcosb\,\,,\,\ , \frac{\delta \lt}{\lt}=\dmt-\frac{1}{2 } \dmws-\dsinb\,\,,\,\,\nonumber\\ \delta{t_w}&=&\frac{1}{2 s_w c_w}\left(\frac{\delta m_z^2}{m_z^2}-\frac{\delta m_w^2}{m_w^2 } \right ) , \label{eq : counterl}\end{aligned}\ ] ] where @xmath171 , @xmath172 are the charge and mass counterterms for the mssm , as given in @xcite , and the counterterms for the mixing matrices have been defined in eq .
( [ eq : bareweakmass ] ) . the renormalized amplitude for the decay @xmath173 can then be written at the one - loop level as follows , @xmath174 v_{\chi_r}\,\,,\nonumber\\ -i t_{a r}^{\rm loop}&=&i { g}\bar{u}_{f ' } \left[\cpaf\pr + \cmaf\pl\right ] v_{\chi_r}\,\,;\,\ , \cpmaf=\delta\lapmafc+\lapmaf{}^{\sigma } + \lapmaf{}^{\rm 1pi}\,\ , .
\label{eq : renormamp } \end{aligned}\ ] ] it contains , besides the counterterms @xmath175 from ( [ eq : counterl ] ) , the one - loop contributions @xmath176 to the one - particle - irreducible three - point vertex functions shown in fig . [ diags:3point]a , and the quantities @xmath177 corresponding to the higher - order terms from the two - point functions in the one - loop expansion of the general expression ( [ eq : externalz ] ) , explicitly given by @xmath178 where @xmath179 has been defined in ( [ eq : neutmix ] ) .
we are now ready to compute the partial decay widths .
the tree - level expressions read @xmath180\,\ , , \label{eq : treleevelgamma}\end{aligned}\ ] ] with @xmath181[x^2-(y+z)^2]}$ ] . due to the presence of photon loops , the one - loop partial decay width computed using the amplitude ( [ eq : renormamp ] ) is infrared divergent ; hence , bremsstrahlung of real photons has to be added to cancel this divergence .
we therefore include in our results the radiative partial decay width @xmath182 , including both the soft and the hard photon part . , which is obviously zero at tree - level .
] the corresponding feynman diagrams are shown in fig .
[ diags:3point]b .
this yields finally the complete one - loop electroweak correction , @xmath183 \times \nonumber \\ & & \left[(\msfas - m_r^{2}-\mfps ) ( |\apaf|^2+|\amaf|^2 ) -4 m_r \mfp { \rm re}(\apaf \amafc)\right]^{-1}\,\ , .
\label{eq : deltadef}\end{aligned}\ ] ] the loop computation itself is rather tedious , since there is a huge number of diagrams to compute .
this is better done by means of automatized tools .
the computation of the loop diagrams has been performed by using the computer algebra systems _ feynarts 3 _ and _ formcalc 2.2 _ @xcite .
we have produced a set of computer algebra programs that compute the one - loop diagrams ( and the bremsstrahlung corrections ) , which are then plugged into a _
code for the numerical evaluation with the help of the one - loop routines _ looptools 1.2 _
a number of checks have been made on the results .
the uv and infra - red finiteness of the result , relying on the relations between the different sectors of the model , is a non - trivial check .
we also have recovered results already available in the literature ; for instance , we used our set of programs to reproduce the strong corrections of @xcite , and , using the _ higgsino _ approximation , we could also reproduce the results of @xcite .
moreover we also checked that , when using the @xmath184-scheme , the one - loop corrections to neutralino and chargino masses reproduce those of @xcite .
although we consider the chargino and neutralino masses as input parameters , in our numerical study we treat them in a slightly different way .
we choose a set of renormalized input parameters @xmath185 , and apply ( [ eq : mmassaneut ] ) , ( [ eq : defuvn ] ) to obtain the one - loop renormalized masses .
of course , if susy would be discovered the procedure will be the other way around , that is , the mssm parameters will be computed from the various observables measured , for example , from the chargino production cross - section and asymmetries at the lc @xcite . for a consistent treatment , the one - loop expressions for these observables will have to be used @xcite .
we note that there exists a certain combination of contributions in the one - loop amplitude ( [ eq : renormamp ] ) that does not depend on the sfermion flavour .
these contributions can be expressed formally as corrections to the coupling matrices @xmath186 with @xmath187 unfortunately the full contributions to the expressions ( [ eq : universal ] ) are divergent .
the only consistent subset of corrections which makes all the expressions in ( [ eq : universal ] ) finite is the subset of fermion and sfermion loops contributing to the self - energies of the gauge bosons , higgs bosons , charginos and neutralinos . with this restriction
, we can define _ effective coupling matrices _
@xmath188 where @xmath189 , @xmath190 , @xmath191 are given by the expressions ( [ eq : universal ] ) taking into account only loops of fermions and sfermions
. we will refer to these corrections as _
universal corrections_. they are the equivalent of the _ super - oblique corrections _ of ref.@xcite
. these effective coupling matrices present a very interesting feature .
since the divergences in the sm sector of the model ( gauge and higgs sectors ) cancel the divergences of the chargino - neutralino sector , the associated logarithmic terms ( @xmath192 , @xmath193 being the arbitrary mass parameter of dimensional reduced integrals ) must be combined . as a result
, a non - decoupling term @xmath194 appears in the final expression . here
@xmath195 represents a generic susy mass , and @xmath196 a generic sm mass .
we have checked explicitly this effect .
we have computed analytically the electron - selectron contributions to the @xmath197 and @xmath198 matrices ( [ eq : universal ] ) , assuming zero mixing angle in the selectron sector ( @xmath199 ) , we have identified the leading terms in the approximation @xmath200 , and analytically canceled the divergences and the @xmath201 terms ; finally , we have kept only the terms logarithmic in the slepton masses .
the result reads as follows : @xmath202\,\ , , \nonumber\\ \delta\uchait^{(f)}&=&\commonnumfactor\log\left(\frac{m^2_{\tilde e_l}}{\mxs}\right)\ , \uchaio\,\frac{\mw\,(m\,\cbta + \mu\,\sbta ) } { 3\,\sqrt{2}\,(m^2-\mu^2)\,\mchasomenysmchast^2 } \times \nonumber\\ & \times & \left((m^2-\mu^2)^2 + 4\,m^2\,\mws + 4\,\mu^2\,\mws + 2\,\mwf + 2\,\mwf\,\cfbt + 8\,m\,\mu\,\mws\,\stbt\right)\,\ , , \nonumber\\ \delta\vchaio^{(f)}&=&\commonnumfactor\log\left(\frac{m^2_{\tilde e_l}}{\mxs}\right)\,\bigg [ \frac{\vchaio^3}{6 } - \vchait \frac{\sqrt{2}\,\mw\,(\mu\,\cbta + m\,\sbta)}{3\,(m^2-\mu^2)\,\mchasomenysmchast^2 } \left(m^4 - m^2\,\mu^2 + \right.\nonumber\\ & & \left .
+ 3\,m^2\,\mws + \mu^2\,\mws + \mwf + \mwf\,\cfbt + ( \mu^2-m^2)\,\mis + 4\,m\,\mu\,\mws\,\stbt\right)\bigg]\,\ , , \nonumber\\ \delta\vchait^{(f)}&=&\commonnumfactor\log\left(\frac{m^2_{\tilde e_l}}{\mxs}\right)\,\vchaio\ , \frac{\mw\,(\mu\,\cbta + m\,\sbta ) } { 3\,\sqrt{2}\,(m^2-\mu^2)\,\mchasomenysmchast^2 } \times \nonumber\\ & \times & \left((m^2-\mu^2)^2 + 4\,m^2\,\mws + 4\,\mu^2\,\mws + 2\,\mwf + 2\,\mwf\,\cfbt + 8\,m\,\mu\,\mws\,\stbt\right),\nonumber\\ \label{eq : logterms}\end{aligned}\ ] ] @xmath203 being the soft - susy - breaking mass of the @xmath204 doublet ( [ eq : sbottommatrix ] ) , whereas @xmath205 is a sm mass . in the on - shell scheme for the sm electroweak theory
we define parameters at very different scales , basically @xmath206 and @xmath207 .
these wide - ranging scales enter the structure of the counterterms see the last formula in eq.([eq : universal]) and so must appear in eq.([eq : logterms ] ) too . as a result
the leading log in the various terms of this equation will vary accordingly .
for simplicity in the notation we have factorized @xmath208 as an overall factor . in some cases
this factor can be very big , @xmath209 ; it comes from the electron - selectron contribution to the chargino - neutralino self - energies .
its non - decoupling behaviour is logarithmic in the heavy ( susy ) mass and it can be explained from renormalization group arguments relating the supersymmetric gauge couplings at the sm and susy scales ( see below ) . this term is similar to the logarithmic part of the universal effects from the sm gauge bosons , which is related to the renormalization of the ordinary gauge couplings and of course can also be explained by renormalization group arguments in this case involving the internal sm scales @xmath20 and @xmath210 .
equivalent expressions can be found for the quark - squark sector .
we have checked numerically that the expressions ( [ eq : logterms ] ) approximate well the logarithmic term in the full expression ( [ eq : effectivegen ] ) .
we have computed numerically ( [ eq : effectivegen ] ) for different input parameters ( including @xmath211 ) for selectron masses in the range @xmath212 , we have fitted this numerical results to a simple function @xmath213 .
the results show a correlation factor close to @xmath214 in better than @xmath215 .
the simple expressions ( [ eq : logterms ] ) approximate the coefficient of the logarithmic term in better than @xmath216 .
however , the expressions ( [ eq : logterms ] ) do not reproduce the full result ( [ eq : effectivegen ] ) due to the presence of important non - logarithmic terms . upon adding up the three slepton generations
, the contributions to ( [ eq : effectivegen ] ) can be typically of the order of @xmath217 for @xmath218 , whereas the approximation ( [ eq : logterms ] ) gives @xmath219 under the same conditions .
we want to stress that this is a physical , measurable , effect . by measuring the two chargino masses ( @xmath220 , @xmath221 )
, one can extract the susy parameters @xmath83 and @xmath222 , and by assuming cp - conservation one can obtain the renormalized mixing matrices @xmath87 and @xmath88 for each sign of @xmath142 . on the other hand , one can extract the value of the mixing matrices in a polarized @xmath10 linear collider @xcite
. however , the extracted values of the mixing matrices are , in a first approximation , the ones of eq .
( [ eq : effectivegen ] ) , which can deviate from the on - shell ones eq .
( [ eq : defuvn]) at the several percent level . in this way , even if some ( or all ) of the sfermions have masses beyond the reach of an @xmath10 linear collider , one can get information of their mass scale by means of the effective coupling matrices .
in fact , the larger the mass , the larger is the correction .
of course , the experimental value of the cross - section @xmath223 will have to be compared with the full one - loop computed cross - section @xcite , since the rest of the one - loop corrections can be as large the contribution of the effective coupling matrices ( [ eq : effectivegen ] ) .
the ultimate reason for these non - decoupling effects lies in the breaking of susy , which affects the susy relation between the gauge - boson and gaugino couplings ( or higgs - boson and higgsino couplings ) . as a simple example
, susy implies that the @xmath224 coupling must be equal to the @xmath225 coupling . for broken susy
, this equality is lost , and the deviation of the @xmath225-coupling from the @xmath224-coupling grows with the scale of the susy breaking @xcite .
one can understand the appearance of the non - decoupling effects by renormalization group arguments . in an energy scale
much larger than any susy mass scale ( @xmath226 ) the theory is supersymmetric , and the effective @xmath224 gauge coupling ( @xmath227 ) is equal to the effective @xmath225 yukawa coupling ( @xmath228 ) , and their renormalization group equations ( rge ) are the same . if some hierarchy exists in the susy sector ( say , for definiteness @xmath229 ) , at the scale @xmath230 the squarks decouple from the running of @xmath19 and @xmath231 .
quarks , on the other hand , decouple from @xmath231 but not from @xmath19 .
therefore , at scales @xmath232 , the coupling @xmath231 is frozen at the squark mass scale @xmath233 as far as quark / squark contributions are concerned due to the lepton / slepton contributions . here
we leave the slepton contributions out of the discussion for the sake of clarity . ] .
therefore , the comparison between the two couplings gives , at one - loop order : @xmath234 where the qed @xmath235-function does not include the contribution from squarks .
since we are using an on - shell renormalization scheme , we are comparing in fact @xmath236 , and since the quarks decouple from @xmath19 at @xmath237 , we end up with a correction proportional to @xmath238 .
we have explicitly checked this fact using susy quantum electrodynamics as a toy model .
the complete electroweak model looks much cumbersome , since various quantities are fixed at different scales ( e.g. the masses of the gauge bosons @xmath20 , @xmath21 are fixed at their respective pole values , and the electromagnetic constant is fixed in the compton limit @xmath239 ) , therefore different pieces of the corrections carry different scales @xmath43 in the arguments of the logarithms . in figs .
[ fig : ueff ] and [ fig : veff ] we show contour plots in the @xmath240 plane of the relative correction to the elements of the mixing matrices @xmath87 and @xmath88 ( [ eq : effectivegen ] ) . for this figure
we have chosen a sfermion spectrum around @xmath241 , namely : @xmath242 where we have assumed a common mass for all the charged sleptons and down - type quarks , and considered mixing only in the stop sector .
the thick black lines in figs .
[ fig : ueff]-[fig : veff ] correspond to spurious divergences in the relative corrections .
the ones corresponding to @xmath243 are divergences in the corrections themselves , since our renormalization prescription fails in this case see eq .
( [ eq : soldm2dmu ] ) .
this divergence appears also explicitly in the approximate expression ( [ eq : logterms ] ) as the inverse of @xmath244 .
the other divergences correspond to lines where the renormalized mixing matrix elements are zero .
these correspond to @xmath245 ( @xmath246 ) for @xmath247 , @xmath248 ( @xmath249 , @xmath250 ) .
corrections as large as @xmath251 can only be found in the vicinity of these divergence lines .
however , there exist large regions of the @xmath252 plane where the corrections are larger than @xmath253 , @xmath254 , or even @xmath255 .
in this section we tackle the ( cumbersome ) numerical analysis of the corrections to the various @xmath256 partial decay channels according to the following plan .
first of all we assess the relevance of the universal corrections defined in sect .
[ sec : universal ] .
next we focus on the non - universal corrections , with especial emphasis on the non - decoupling effects from gauginos and higgs particles , followed by an exhaustive analysis of the corrections to the various partial decay widths as a function of the most relevant parameters .
after that , we briefly concentrate on the strong corrections .
finally , we combine the universal , non - universal , and strong corrections to the partial decay widths to evaluate the impact on the branching ratios for all the decay channels , which are in practice the true observables . as for the presentation of the results themselves
, we will use the following criteria : the first and second generation of sfermions have very similar properties , so only the results for the first generation are presented .
we will present separate results for the third squark generation ( @xmath257 ) , since the large yukawa couplings and masses make this generation behave in a special way .
the third generation sleptons ( @xmath258 ) have also large yukawa couplings , but their effect is small , unless is very large .
we will refrain to show explicit results for @xmath258 , if these results are similar to the first generation ones .
we have used the following default set of central parameters for our numerical evaluation : @xmath259 we point out that the amount of splitting chosen in the sbottom sector is not critical for the numerical results , and has nothing to do with preserving the vacuum condition ( [ eq : necessary ] ) because we are assuming zero sbottom mixing angle in the first eq.([eq : abt ] ) .
furthermore , the negative sign for the stop mixing angle is related to the chosen sign for @xmath142 and the desired relation @xmath260 via the second eq.([eq : abt ] ) .
the other sm parameters have been taken from ref.@xcite .
for simplicity , we will be using the grand unification ( gut ) relation between the electroweak soft - susy - breaking gaugino masses @xmath261 unless stated otherwise .
the computed values of the heaviest up - type sfermions are given in table [ tab : sup2masses ] . in table
[ tab : massbr ] we show the tree - level chargino and neutralino masses , as well as the tree - level branching ratios of sfermions decaying into charginos and neutralinos , and the one - loop corrections to the neutralino masses .
given the large number of decay channels , a presentation including the corrections to each individual final state would be tedious .
moreover , a large number of channels have a very small branching ratio , and higher - order terms are not of phenomenological interest . for these reasons
most of the discussion will be devoted to the total decay widths of sfermions into charginos and neutralinos , in particular to the relative correction @xmath262 with @xmath263 or @xmath264 .
we will not show results for processes whose branching ratio are less that 10% in all of the explored parameter space .
as we have said , we start our numerical analysis by testing the non - decoupling effects directly associated to the universal corrections discussed in section [ sec : universal ] .
the main aim of this section is to assess the numerical impact of these non - decoupling effects . to this end
we present in fig .
[ fig : unilep ] the universal corrections to the partial decay widths of the first generation of sleptons ( @xmath265 ) as a function of a common mass for all squarks .
since these corrections are universal , and the lepton masses can be safely neglected , they are the same for the other generations of sleptons . looking at the right end of the plots ( @xmath266 ) , the logarithmic scale of the plots makes evident the non - decoupling of squarks by means of a logarithmic term @xmath267 equivalent to that of eqs .
( [ eq : logterms ] ) .
the corrections are also non - negligible for light squark masses .
we see that for squark masses below @xmath268 the corrections reach a @xmath269 value for most of the decay channels , or even larger than @xmath270 for the selectron decay into neutralinos .
we turn now our attention to the squark decays .
[ fig : unisq ] shows the universal corrections to the squark partial decay widths as a function of a common mass for all sleptons .
we show the corrections for down - type and up - type quarks of the first and third generation .
again , the logarithmic behaviour from eq .
( [ eq : logterms ] ) is evident in this figure .
the logarithmic regime is attained already for slepton masses of order @xmath241 .
the universal corrections are seen to be positive for all squark decays , ranging between @xmath255 and @xmath270 for slepton masses below @xmath241 .
the non - universal corrections comprise the set of full corrections excluding the universal corrections of section [ sec : universal ] . for a given sfermion decay
, these corrections do not depend on the parameters of the sfermions of different type or generation .
that is , e.g. for a @xmath271 decay , the dependence on the lepton / slepton parameters and the first and second generation quark / squark parameters appears only in the universal corrections analyzed in the previous section , and the non - universal corrections depend only on the gauge , higgs , chargino / neutralino and stop / sbottom sectors .
the non - decoupling of gauginos was already discussed in @xcite . there , the qcd corrections to the squark decays were computed , and an explicit non - decoupling gaugino term ( of the form @xmath272 ) was found .
the origin of this non - decoupling effect is similar to that of the sfermionic ones in ( [ eq : logterms ] ) , namely the gaugino uv - divergences cancel the gauge boson ones , so that the logarithms associated with the divergence must compensate between susy and non - susy particles , leading to a @xmath273 behaviour which can again be explained from simple renormalization group arguments .
the electroweak corrections present several peculiarities that prevent from obtaining a simple analytical result of these logarithms .
first of all , the non - decoupling particles are part of the final state of the process , thus if their masses are large the decay will be phase - space closed .
second , the complicated structure of the electroweak sector involves the mixing of gauginos and higgsinos .
one can not compute simply the gauge gaugino or the higgs - boson higgsino corrections , unless very special limits are taken @xcite .
thus , a numerical approach is preferred in this case . in this subsection ( [ sec : nouninondec ] ) we give up the gut assumption between the gaugino masses , and will let the soft - susy - breaking mass parameters @xmath83 and @xmath84 vary independently .
in this way we can afford having a light neutralino ( when @xmath84 is light ) , while the @xmath274 gauginos are heavy , and analyze the non - decoupling effects of the @xmath274 gauginos in the neutralino decays . on the other hand , by maintaining @xmath83 light , but taking @xmath84 to be heavy , one obtains a heavy @xmath85 neutralino whose non - decoupling effects can be studied in the chargino sector . in fig . [
fig : nondecm2 ] we show the non - universal corrections to the slepton and squark decays into neutralinos as a function of the soft - susy - breaking @xmath274 gaugino mass parameter @xmath83 in the range @xmath275 , keeping @xmath84 fixed to a light value @xmath276 . for @xmath277
the corrections show a rich structure that will be analyzed below .
only the results for the first generation of sfermions are shown ; although the results for top- and bottom - squarks are slightly different , the same conclusion follows . shown
are the corrections for each individual decay channel with a branching ratio larger than @xmath270 . for @xmath278 most of the decays
have already reached the logarithmic regime , so the figure shows mainly straight lines .
the corrections can have both signs , and range between @xmath253 and @xmath279 at @xmath278 .
the slopes of the curves are small which means that , although there exist a non - decoupling effect , it is very small and of no phenomenological interest for @xmath280 .
a similar situation is found in the non - decoupling effect of the @xmath85 gaugino with respect to the decays into charginos in fig .
[ fig : nondecm1 ] . in this case
the slopes of the different curves are even smaller . the non - decoupling effects from higgs particles can be seen in fig .
[ fig : nondecmhplus ] .
the figure makes clear that the corrections grow as a @xmath281 .
the effect is , of course , much larger in the third generation squark decays , but is also visible in the @xmath282 and @xmath283 decays .
we will present here the behaviour of the non - universal electroweak corrections as a function of parameters relevant for the chargino / neutralino sector and the sfermion sector , and .
in general the corrections present a rich structure , due to the fact that every single parameter controls different aspects of the decay under study .
let s take for example the higgsino mass parameter @xmath142 .
when changing this parameter , the neutralino and chargino masses change , and some decay channels open or close : when this happens , threshold divergences appear in the rest of the channels . at the same time , for @xmath284 our renormalization prescription breaks down ( cf .
eq.([eq : soldm2dmu ] ) ) , and divergent corrections appear .
moreover the @xmath142 parameter enters the coupling of the sfermion to higgs bosons in two ways : directly , in the expression of the feynman rules , and indirectly , in the determination of the soft - susy - breaking trilinear couplings ( [ eq : abt ] ) . as a result
the variation of the corrections with respect to @xmath142 will exhibit a complicated evolution pattern spotted with the spurious divergences associated with the renormalization framework . in figs .
[ fig : mulept ] and [ fig : muquark ] we display the corrections for the partial decay widths into neutralinos and charginos for sleptons and squarks respectively . in all plots
we can see the divergences at @xmath285 .
we also see similar structures in several plots .
these similarities make clear the @xmath274 structure of the theory , for example in the corrections to the partial decays of @xmath286 , @xmath283 , @xmath287 , @xmath288 and @xmath289 into charginos .
the corresponding decays into neutralinos deviate among the different sfermions , since the @xmath85 charge enters the game .
the presence of large yukawa couplings also alters the general behaviour of the corrections .
the corrections to the top- and bottom - squark decays exhibit indeed a very different structure .
differences in the @xmath290 decays would only be visible in the @xmath291 decay channel , which has a branching ratio below @xmath216 ( see table [ tab : massbr ] ) . figs .
[ fig : mlept ] and [ fig : mquark ] show the evolution with the soft - susy - breaking gaugino mass parameter @xmath83 .
again we observe a very similar structure for the sleptons and the first generation squarks . for @xmath292 , and away from the divergence region , the corrections are small ( less than @xmath269 ) in most channels . only the decay mode @xmath293 has corrections around @xmath270 . for @xmath294 the corrections can be moderate , up to @xmath269 for the sleptons and first generation squarks , and @xmath270 for the bottom - squarks .
note also the divergence appearing in the corrections to the @xmath295 decays into @xmath125 when the value of the masses approach the phase space limit .
this divergence arises from the coulomb singularity due to soft - photon exchange between the two final state particles @xcite .
a consistent description of the decay width in this mass regime needs of a proper description of slowly moving final - state charged particles @xcite .
we now turn our view to the parameters of the sfermion sector .
we start with the sfermion mixing angle .
its main effect is not in the corrections themselves , but in the tree - level decay amplitudes . in fig .
[ fig : theta ] we show the variation of the relative corrections as a function of relevant mixing angles . for the @xmath290
decays we see spikes of large corrections for the chargino channels .
these spikes reflect the fact that near @xmath296 ( @xmath297 ) the @xmath298 ( @xmath287 ) has a tiny branching ratio to charginos ; it is basically a @xmath299 , which ( at the tree - level ) couples only to higgsino - type charginos with a small yukawa coupling . at one - loop ,
however , the @xmath299 does effectively couple to charginos both through the one - loop conversion @xmath300 and through genuine vertex diagrams .
these contributions can be of the same size as the tree - level contributions , giving large corrections .
for the sfermions of the first two generations these spikes are more pronounced . since their yukawa couplings are negligible , the one - loop effective coupling is larger than the tree - level one .
the third generation squarks have a very different behaviour . to understand fig .
[ fig : theta]b one has to bear in mind that the heaviest squark mass varies with the squark mixing angle . as @xmath301 varies the @xmath72 mass crosses a series of thresholds .
the corrections to the @xmath302 decays on the other hand have a smooth behaviour , similar to the @xmath290 decays but with the spikes softened by the larger yukawa coupling .
we also show in fig .
[ fig : theta]c the variation of the bottom - squark decays as a function of the @xmath271 mixing angle @xmath301 .
although the shape of the figure might be reminiscent of what would be obtained by the use of the finite threshold corrections to the bottom - quark yukawa coupling @xcite , this is not the leading contribution to the corrections .
in fact , we have checked that the use of the finite threshold corrections to the bottom - quark yukawa coupling reproduce quite well the shape of the corrections to the @xmath303 ( @xmath304 ) partial decay widths .
however , there exist finite terms from other contributions , changing the overall value of the corrections . for the @xmath305 ( @xmath306 )
the finite threshold corrections fail to give an approximate description of the full result .
the evolution with the sfermion mass parameters themselves can be seen in figs .
[ fig : msflight]-[fig : msquarkheavy ] . in fig .
[ fig : msflight ] we vary one of the sfermion masses , by maintaining the splitting between @xmath307 , @xmath308 as in eq .
( [ eq : inputpars ] ) . in figs .
[ fig : msflight]a and b we show the non - universal corrections to the selectron and sneutrino partial decay widths . above @xmath309
the corrections follow the sudakov double - log form @xmath310 @xcite .
this kind of electroweak corrections appears in any observable in which the process energy is much larger than the electroweak mass scale . for comparison
the universal contributions are shown in figs .
[ fig : msflight]c and d as a function of a common slepton mass .
while the universal effects dominate over the sudakov terms in the relevant region where the sfermion masses lie below @xmath311 , the opposite holds once these masses become larger .
it is interesting to further explore the behaviour of the corrections in the mass region below @xmath241 also for the squark decays . in fig .
[ fig : msquarkheavy ] we show the corrections for top- and bottom - squark decays as a function of the masses , in a mass range @xmath312 . in this figure
the splitting between the bottom - squarks is fixed at @xmath313 , whereas the top - squark mass is left free .
several thresholds are seen in the figures .
the value of the corrections behaves smoothly between the threshold points , but it is clear that the value of the corrections depends strongly on the exact correlation between the several mssm masses .
we come finally to the @xmath25 parameter . for the first two sfermion generations
@xmath25 only enters the corrections through the expressions of the masses , since the yukawa couplings are negligible . as a consequence the evolution with @xmath25 is flat above @xmath314 , when the masses acquire their asymptotic values . for the third generation the couplings of fermions and sfermions with higgs bosons and higgsinos loops enter the game .
if we stick to the input parameters in ( [ eq : inputpars ] ) the trilinear soft - susy - breaking couplings @xmath315 acquire large values
( [ eq : abt ] ) and the sfermion - higgs couplings become large , and even non - perturbative .
note however , that large @xmath315 values would generate charge and colour breaking vacuum
( [ eq : necessary ] ) . as long as the @xmath315 are consistent with the ( necessary ) condition ( [ eq : necessary ] ) the corrections remain perturbative .
this is not to say that scenarios with @xmath316 are not possible , but that , for a given @xmath25 , the input parameters @xmath315 , @xmath317 and @xmath318 are constrained so that eq .
( [ eq : necessary ] ) is satisfied . in particular , note that a zero mixing angle in the sbottom sector is not possible for large @xmath25 . to probe the effect of @xmath25 in the corrections
we must choose a different parameter set that satisfies the bound ( [ eq : necessary ] ) all over the explored @xmath25 . to this end
we choose the soft - susy - breaking masses and trilinear couplings as input parameters , so that they reproduce the masses and angles of ( [ eq : inputpars ] ) , that is we adjust the values of the parameters in the sfermion mass matrices ( [ eq : sbottommatrix ] ) as follows @xmath319 with a common value for first and second generation squarks ( @xmath320 , @xmath321 , @xmath322 ) , and for all sleptons ( @xmath323 ) .
the rest of the parameters are set as in eq .
( [ eq : inputpars ] ) ( viz .
@xmath324 , @xmath325 , @xmath326 ) . by choosing this parameter set , the masses and mixing angles will change with @xmath25 . in fig .
[ fig : deltatb ] we show the corrections to the third generation sfermion decays as a function of @xmath25 .
we see that a large value of @xmath25 may increase the absolute value of the corrections , but they stay below few ten percent for the allowed range @xmath327 .
the qcd corrections to the partial decays widths of sfermion decays into charginos and neutralinos were computed in @xcite .
we have performed an independent computation , and have verified the analytical and numerical results of ref.@xcite .
we will need these partial results to compute the full one - loop branching ratios of the sfermion decays .
the qcd corrections can be very large in certain parts of the parameter space .
this applies especially for squarks decaying into higgsino - type charginos or neutralinos .
these large corrections can be absorbed by an adequate resummation of the leading effects .
the latter are of two types : the running of the light - quark masses up to the scale of the sfermion masses produces large negative corrections ; the finite susy threshold corrections to the bottom - quark yukawa coupling can also be large ; they grow with @xmath25 , and its sign is opposite to @xmath142 see the extensive literature @xcite on the subject . the authors of ref .
@xcite demonstrated that , for the case of @xmath328 coupling , these corrections can be exactly resummed to all orders of perturbation theory by using the effective bottom quark yukawa coupling according to @xmath329 where @xmath330 is the running quark mass , and @xmath331 is the finite threshold correction . by using the effective bottom quark mass in eq .
( [ eq : mbdef ] ) we are able to absorb a large part of the corrections into the effective couplings , yielding an improved partial decay width @xmath332 where @xmath333 is the one - loop expansion of @xmath334 .
the previous equation defines the _ remainder _ of the one - loop contributions , namely what is left after subtracting the one - loop part of the resummed threshold corrections from the full one - loop result .
it reads @xmath335 the improved qcd correction factor is then given by @xmath336 in fig .
[ fig : qcdcorr ] we show an example .
we have plotted the one - loop and the improved qcd corrections to the bottom - squark partial decay widths into neutralinos as a function of @xmath25 , using the input parameters ( [ eq : inputssb ] ) and a gluino mass @xmath337 .
the strong coupling constant is evaluated at the mass scale of the decaying particle , using the normalization @xmath338 .
the growing in absolute value of the one - loop qcd corrections is due to the increase of the bottom - squark mass splitting ( entering @xmath331 ) with @xmath25 .
for @xmath339 they already pass the @xmath340 value . on the other hand @xmath341
is well behaved in all the @xmath25 range , and it is seen not to be negligible at all in some cases .
the large negative corrections visible in fig .
[ fig : qcdcorr ] have a twofold origin .
first , the running of the bottom quark mass provides large negative corrections due to _ standard _ qcd renormalization group effects .
second , the sign of the qcd contributions to the threshold corrections is opposite to that of @xmath142 ( since @xmath342 in eq .
( [ eq : mbdef ] ) ) .
therefore both kinds of contributions reinforce mutually in the originally chosen scenario @xmath343 . in the alternative scenario ( @xmath344 )
the two contributions partially cancel each other , giving a smaller total ( negative ) correction .
thus in the @xmath344 scenario the susy - qcd effects actually prevent the decay rates from being too much suppressed by the gluonic corrections .
one has to be careful in the use of the effective bottom - quark mass ( [ eq : mbdef ] ) in the case of positive corrections ( @xmath344 ) , since in this case the effective yukawa coupling grows and can become non - perturbative for large values of @xmath222 and @xmath25 . in fig .
[ fig : qcdcorr]d we show a scenario with @xmath344 .
the input parameters are those of eq .
( [ eq : inputssb ] ) , but changing the sign of @xmath142 and the soft - susy - breaking trilinear couplings @xmath315 .
although the corrections are still negative , we clearly see the change of trend for large @xmath345 ; from this point onwards the ( positive ) threshold corrections are comparable to the effects of the running bottom quark mass . at @xmath346
the two leading contributions nearly compensate each other , giving a total negative correction below @xmath347 .
so at large @xmath25 some channels become essentially free of huge negative corrections , whereas in the original @xmath343 scenario these channels were reduced by more than @xmath348 due to quantum effects ! in ref .
@xcite the numerical results for the qcd corrections to the partial sbottom decay widths were much smaller than those presented in fig .
[ fig : qcdcorr ] .
the reason is that in ref .
@xcite the following set of parameters was used : a low value of @xmath25 ( @xmath349 ) , small splitting between the two sbottom masses ( @xmath350 ) , and a small sbottom mixing angle ( @xmath351 ) . under these conditions the susy threshold corrections to the bottom yukawa coupling ( @xmath331 ) are suppressed @xcite . moreover , the choice of @xmath344 in ref .
@xcite would mean that the finite threshold effects partially compensate for the renormalization group running of the bottom mass , thus decreasing the value of the corrections . at moderate or large values of ( @xmath352 ) , however , the rest of the conditions can not be fulfilled . if the sbottom mixing angle were small ( @xmath351 )
, then the soft - susy - breaking trilinear coupling would be large ( @xmath353 ) , eventually spoiling the vacua condition ( [ eq : necessary ] ) . even if one drops the condition ( [ eq : necessary ] ) , then the weak corrections would blow up due to the non - perturbativity of the higgs - squark - squark trilinear coupling entering the three - point function corrections in fig .
[ diags:3point ] .
we conclude , therefore , that sticking to the necessary condition ( [ eq : necessary ] ) provides a natural suppression of the higgs - squark - squark trilinear coupling , while maintaining the perturbativity of the weak corrections over the whole @xmath25 range . at the same time , it means that for @xmath354 the two sbottom masses have significant splitting , and @xmath355 , providing large corrections due to the finite threshold effects .
we remark that whereas the finite threshold corrections ( @xmath331 ) can be regarded as _ non - decoupling _ effects in the study of the higgs sector of the mssm , this is not so for the sfermion decays under consideration ( even though we have kept the _ non - decoupling _ denomination also in our case ) .
indeed , the non - decoupling property applies when _ all _ the susy parameters ( @xmath142 , @xmath49 , @xmath356 , @xmath357 ) are simultaneously scaled . however , being the sbottom quark mass itself the process energy , the scaling of the susy parameters entails a simultaneous scaling of the process energy , and therefore the @xmath331 corrections can not be considered here as genuine non - decoupling quantum effects .
moreover , in an scenario of large @xmath142 ( @xmath358 ) the light chargino is basically a gaugino , and so its coupling to quarks and squarks is essentially a gauge coupling , not a yukawa coupling . as a result
the sbottom decay into light charginos / neutralinos is not sensitive to @xmath331 in this limit . at the end of the day
we analyze the higher - order effects on the branching ratios .
notice that in practice only the branching ratios are directly accessible from the measurements of the various cross - sections at the colliders .
therefore the @xmath359 are the true observables in this kind of analysis , and we will compute them from the corresponding corrections to the partial decay widths that we have considered in the previous sections .
in particular we have to include the qcd and the threshold effects . in the following analysis
we also take into account the ( generally small ) effects of the shift of the masses in the phase - space factor .
specifically we show the following quantities : * tree - level branching ratio ( @xmath360 ) ; * qcd - correction to the branching ratio as discussed in section [ sec : qcd ] , including the qcd corrections to the @xmath72 mass : @xmath361 @xmath362 being defined in ( [ eq : impqcd ] ) ; * total correction to the branching ratios ( qcd and ew ) , including the qcd and ew corrections to the @xmath72 mass , and the ew corrections to the neutralino masses @xmath363 with @xmath364 as defined in ( [ eq : totalcn ] ) . the corrections to the branching ratios are usually smaller than those to the partial decay widths . the limiting case being when only one decay channel exists , and has obviously no corrections to the branching ratio . only results for the bottom- and top - squarks will be shown .
the corrections to the slepton and first- and second - generation squarks are tiny , for two reasons : usually a single decay channel is dominant and the corrections to the partial decay widths are small .
in addition the corrections will be shown only on those portions of the parameter space in which the only possible decay channels are the fermionic ones .
the bosonic decay channels @xmath365 can be dominant when they are open , and the quantum corrections to these channels must be taken into account to compute the corrections to the branching ratios @xcite . in fig .
[ fig : brm ] we show the tree - level branching ratios of the bottom- and top - squarks as a function of the @xmath40 soft - susy - breaking gaugino mass parameter @xmath83 , as well as the absolute corrections to the branching ratios of squarks decaying into neutralinos .
although the qcd corrections are usually the largest ones , the ew and qcd corrections can be of the same order in certain scenarios .
for example in fig .
[ fig : brm]c the qcd corrected @xmath366 is @xmath367 larger than the tree - level one for @xmath368 , but the ew corrections compensate most of this correction , and the final correction is less than @xmath216 .
the effects of the mass - shifts are in general very small , except near the threshold regions , where a given channel is permitted according to the tree - level masses prediction , but it is closed when one uses the one - loop prediction for the masses .
this is the case in fig .
[ fig : brm]d .
the decay channel @xmath369 is open up to @xmath370 according to the tree - level prediction for the heaviest top - squark mass .
however the negative corrections to @xmath76 table [ tab : sup2masses] enforce this channel to get closed for lighter @xmath83 values , concretely , at @xmath371 including only the qcd corrections and at @xmath372 including the full corrections . in fig .
[ fig : brmsb2 ] we show the branching ratios and its corrections as a function of the lightest bottom - squark mass ( keeping the splitting between the two bottom - squarks at @xmath373 ) .
the lines representing the corrections end at the points where the bosonic decay channels open .
we see again that the ew corrections to the @xmath305 branching ratio can be of the same order as the qcd ones . for @xmath374
the @xmath375 qcd contributions are almost compensated by the ew ones . for the @xmath303 ,
the ew corrections represent a small shift to the qcd - corrected branching ratios . for the top - squark
the qcd corrections can be larger ( @xmath376 ) , and the branching ratio also suffers an important shift from the ew sector .
the effects of the mixing angles are shown in fig .
[ fig : brosq ] .
the corrections to the bottom - squark branching ratio show a simple structure .
the corrections from the weak sector are comparable to that of the qcd sector for practically any value of the mixing angle .
the @xmath271 branching ratios and corrections are shown in figs .
[ fig : brosq]b and d. the effect of the ew corrections is clearly visible above the qcd - corrected branching ratios .
we come finally to analyze the @xmath25 dependence , which is shown in fig .
[ fig : brotb ] . in this case
we will use again the soft - susy - breaking mass parameters in eq .
( [ eq : inputssb ] ) to compute the physical masses and mixing angles in the sfermion sector . the main effect of @xmath25 on the tree - level branching ratios is due to the change in the sfermion masses themselves , providing the opening of the bosonic channels .
again , the weak corrections to the bottom - squark branching ratios are seen to be a tiny addition to the qcd induced ones .
opposite to that , the weak corrections to the top - squark decay branching ratios are larger , especially in the lightest top - squark channels . in this case
they can be comparable to the qcd induced corrections .
we have presented a consistent and complete one - loop on - shell renormalization scheme for the sfermion and the chargino - neutralino sectors of the mssm .
this scheme is suitable for the computation of one - loop electroweak corrections to observables relevant to the next generation of colliders .
we have applied this scheme to compute the full electroweak corrections to the partial decay widths of sfermions into charginos and neutralinos . as a summary of the results we show in table [ tab : brcorr ] the corrected branching ratios for all possible individual sfermion decays for the ( low ) input parameter set ( [ eq : inputpars ] ) .
for squarks we show both : the branching ratio including only qcd effects , and the fully corrected branching ratio .
one can compare these results with the former tree - level results in table [ tab : massbr ] to assess the importance of the corrections in this particular parameter set .
a close comparison shows that , in fact , the ew corrections can have an effect as large as the qcd corrections alone .
we can compare , for example , the leading branching ratio of the lightest up - squark ( @xmath377 : @xmath378 , @xmath379 , @xmath380 ) or the lightest sbottom ( @xmath381 : @xmath382 , @xmath383 , @xmath384 ) .
the corrections show a rich and complicated structure when the mssm parameters are varied .
nevertheless , we have been able to provide physical explanations of our results .
indeed , we have identified sources of non - decoupling effects in these radiative corrections . in contradistinction to the standard model case , none of the mssm particles decouples from these corrections the non - decoupling effects being logarithmic in the heavy masses . as a consequence all particles of the mssm
must be taken into account in any computation involving loop diagrams with external fermion - sfermion - chargino / neutralino couplings . in some cases , however , the term multiplying this logarithm is small .
furthermore , we have identified a class of universal corrections to the fermion - sfermion - chargino / neutralino couplings , that can be treated as _ effective coupling matrices _ for the chargino / neutralino sector .
these universal corrections consist of the fermion - sfermion contributions to the self - energies of the gauge bosons , higgs bosons , charginos and neutralinos .
the _ effective coupling matrices _ absorb the non - decoupling effects of sfermions .
explicit analytic expressions for these corrections have been given in a simple example , but the general analysis has been performed numerically . these corrections can be large ( @xmath385 for sfermion masses around @xmath241 ) and grow logarithmically with the sfermion masses . a physical explanation of this effect using renormalization group arguments has been given .
the bulk of the non - universal corrections grow as a logarithm squared of the particle decay mass , due to the sudakov - type double logarithms of the electroweak corrections . for sfermion masses around @xmath386 , relevant for a @xmath387 @xmath10 linear collider such as tesla , the non - universal electroweak corrections to the slepton and first- and second - generation squark partial decay widths are small . for top- and bottom
- squarks the corrections are larger due to the large yukawa couplings .
threshold_-like corrections to the quark yukawa couplings are important at large for 3rd generation sfermions , and they must be resummed to obtain meaningful results .
the corrections remain always in the perturbative regime as long as the soft - susy - breaking trilinear couplings are not too large .
this condition is , however , granted if the vacuum does not break the charge and/or colour symmetry . in particular , we point out that this vacuum condition can not be preserved at large @xmath25 by degenerating the sfermion masses and/or assuming vanishing sfermion mixing angles . therefore , since @xmath388 is phenomenologically ruled out , one can compensate the @xmath389 term in eq .
( [ eq : abt ] ) by an appropriate choice of the other parameters .
for instance , at @xmath390 the vacuum condition can be preserved with the following set of masses and mixing angles : @xmath391 , @xmath392 @xmath393 and @xmath394 .
the corrections can be significantly larger for individual decay channels that have small branching ratios . therefore large corrections are washed out in the total decay widths @xmath395 and @xmath396 eq .
( [ eq : totalcn ] ) .
we have combined the qcd corrections with the electroweak effects for the top- and bottom - squark partial decay widths , and have evaluated the full one - loop branching ratios of these supersymmetric particles in the case that the only open decay channels are the chargino / neutralino ones . in performing the computation of the corrections to the decay rates and branching ratios
we have also taken into account the corrections to the squark and neutralino masses themselves .
an specially interesting case appears when these mass shifts provide the opening ( closing ) of channels that would be closed ( open ) according to the naive tree - level prediction .
since the overall corrections can be very large for higgsino - type charginos / neutralinos , we have made use of the resummed expressions for the two leading quantum contributions to the bottom - quark yukawa coupling , namely the running quark mass and the finite threshold supersymmetric effects .
the upshot of our analysis should be emphasized : the ew corrections can be of the same order of magnitude as the qcd effects , and therefore a consistent treatment of the sfermion decays beyond leading order in the mssm demands to include the ew quantum contributions on the same footing as the qcd ones .
the calculations have been done using the qcm cluster of the dfg forschergruppe `` quantenfeldtheorie , computeralgebra und monte - carlo simulation '' .
we are thankful to t. hahn for his help regarding the computer algebra system .
j.g . is thankful to d. stkinger , a. vicini , g. rodrigo , m. melles and m. spira for useful discussions .
this collaboration is part of the network `` physics at colliders '' of the european union under contract hprn - ct-2000 - 00149 .
the work of j.g . is supported by the european union under contract no .
hpmf - ct-1999 - 00150 .
the work of j.s . has been supported in part by mecyt and feder under project fpa2001 - 3598 .
j. drees , talk at the _ xx international symposium on lepton and photon interactions at high energies _ , rome , italy , 23rd-28th july 2001 , ` hep - ex/0110077 ` ; + s. mele , _ combined fit to the electroweak data and constraints on sm _ , talk at the _ international europhysics conference on high energy physics _ , budapest , hungary , july 12 - 18 , 2001 ; + the lep collaborations , the lep ew working group , the sld heavy flavour and elecroweak groups , _ a combination of preliminary electroweak measurements and constraints on the standard model _ , ` hep - ex/0112021 ` .
see e.g. m. carena _ et al .
_ , _ report of the tevatron higgs working group _ , http://xxx.lanl.gov/abs/hep-ph/0010338[hep-ph/0010338 ] .
atlas collaboration , _ atlas technical design report _
cern / lhcc/99 - 14/15 ; + cms collaboration , _ cms technical proposal _
cern / lhcc/94 - 38 ; + f. gianotti , proceedings of the _ ivth international symposium on radiative corrections ( radcor 98 ) _ ,
, world scientific 1999 , ed .
j. sol . _
2nd joint ecfa / desy study on physics and detectors for a linear electron - positron collider _
, ` http://www.desy.de/conferences/ecfa-desy-lc98.html ` ; _ tesla technical design report _ ` desy 2001 - 011 ` , part iii : _ physics at an @xmath10 linear collider _ , r. heuer , d. miller , f. richard , p. zerwas editors , ` http://tesla.desy.de/ ` , ` hep - ph/0106315 ` .
h. p. nilles , _ phys .
* 110 * ( 1984 ) 1 ; + h. e. haber , g. l. kane , _ phys . rept . _
* 117 * ( 1985 ) 75 ; + a. b. lahanas , d. v. nanopoulos , _ phys .
* 145 * ( 1987 ) 1 ; + s. ferrara , ed . ,
_ supersymmetry _ , vol . 1 - 2
. north holland / world scientific , singapore , 1987 .
a. dobado , m. j. herrero , s. pearanda , _ eur .
phys . j. _
* c7 * ( 1999 ) 313 , http://xxx.lanl.gov/abs/hep-ph/9710313[hep-ph/9710313 ] ; _ ibid . _
* c12 * ( 2000 ) 673 , http://xxx.lanl.gov/abs/hep-ph/9903211[hep-ph/9903211 ] ; _ ibid . _ * c17 * ( 2000 ) 487 , http://xxx.lanl.gov/abs/hep-ph/0002134[hep-ph/0002134 ] . m. carena , m. quirs , c. e. m. wagner , _ nucl .
_ * b461 * ( 1996 ) 407 , http://xxx.lanl.gov/abs/hep-ph/9508343[hep-ph/9508343 ] ; + h. e. haber , r. hempfling , a. h. hoang , _
z. phys . _
* c75 * ( 1997 ) 539 , http://xxx.lanl.gov/abs/hep-ph/9609331[hep-ph/9609331 ] ; + s. heinemeyer , w. hollik , g. weiglein , _ phys . rev . _
* d58 * ( 1998 ) 091701 , http://xxx.lanl.gov/abs/hep-ph/9803277 [ hep - ph/9803277 ] ; _ phys .
lett . _ * b440 * ( 1998 ) 296 , http://xxx.lanl.gov/abs/hep-ph/9807423[hep-ph/9807423 ] ; _ eur .
phys . j. _
* c9 * ( 1999 ) 343 , http://xxx.lanl.gov/abs/hep-ph/9812472[hep-ph/9812472 ] ; + j. r. espinosa , r. zhang , _ jhep _ * 0003 * ( 2000 ) 026 , http://xxx.lanl.gov/abs/hep-ph/9912236[hep-ph/9912236 ] ; _ nucl .
_ * b586 * ( 2000 ) 3 , http://xxx.lanl.gov/abs/hep-ph/0003246[hep-ph/0003246 ] ; + m. carena _ et al .
_ , _ nucl .
phys . _ * b580 * ( 2000 ) 29 , http://xxx.lanl.gov/abs/hep-ph/0001002[hep-ph/0001002 ] ; + g. degrassi , p. slavich , f. zwirner , _ nucl . phys . _
* b611 * ( 2001 ) 403 , ` hep - ph/0105096 ` .
l.j . hall , r. rattazzi , u. sarid , _ phys .
* d50 * ( 1994 ) 7048 , ` hep - ph/9306309 ` ; + m. carena , m. olechowski , s. pokorski , c.e.m .
wagner , _ nucl .
_ * b426 * ( 1994 ) 269 , ` hep - ph/9402253 ` ; + k. hikasa , y. nakamura , _
z. phys .
_ * c70 * ( 1996 ) 139 , _ erratum : _ _ ibid . _
* c71 * ( 1996 ) 356 , hep - ph/9501382 .
a. djouadi , w. hollik , c. jnger , _ phys .
_ * d55 * ( 1997 ) 6975 , http://xxx.lanl.gov/abs/hep-ph/9609419 [ hep - ph/9609419 ] .
e. katz , l. randall , s. su , _ nucl .
_ * b536 * ( 1998 ) 3 , hep - ph/9801416 .
j. a. coarasa _
_ , _ eur .
phys . j. _
* c2 * ( 1998 ) 373 , http://xxx.lanl.gov/abs/hep-ph/9607485[hep-ph/9607485 ] . j. a. coarasa _ et al .
_ , _ phys . lett .
_ * b425 * ( 1998 ) 329 , ` hep - ph/9711472 ` ; + j. guasch , r. a. jimnez , j. sol , _ phys .
lett . _ * b360 * ( 1995 ) 47 , ` hep - ph/9507461 ` ; + j. a. coarasa , r. a. jimnez , j. sol , _ phys
. lett . _ * b389 * ( 1996 ) 312 , ` hep - ph/9511402 ` .
m. berggren , r. kernen , h. nowak , a. sopczak , proceedings of _ international workshop on linear colliders ( lcws 99 ) _ , sitges , barcelona , spain , 28 apr - 5 may 1999 , universitat autnoma de barcelona 2000 , eds .
e. fernndez , a. pacheco , http://xxx.lanl.gov/abs/hep-ph/9911345[hep-ph/9911345 ] .
w. beenakker , r. hpker , m. spira , p. m. zerwas , _ nucl .
_ * b492 * ( 1997 ) 51 , http://xxx.lanl.gov/abs/hep-ph/9610490[hep-ph/9610490 ] ; + w. beenakker _ et al .
_ , _ nucl .
_ * b515 * ( 1998 ) 3 , http://xxx.lanl.gov/abs/hep-ph/9710451[hep-ph/9710451 ] ; + t. plehn , _ phys .
* b488 * ( 2000 ) 359 , http://xxx.lanl.gov/abs/hep-ph/0006182[hep-ph/0006182 ] . h. eberl , a. bartl , w. majerotto , _ nucl . phys . _ * b472 * ( 1996 ) 481 , http://xxx.lanl.gov/abs/hep-ph/9603206[hep-ph/9603206 ] ; + h. eberl , s. kraml , w. majerotto , _ jhep _ * 05 * ( 1999 ) 016 , http://xxx.lanl.gov/abs/hep-ph/9903413[hep-ph/9903413 ] . s. kraml _ et al .
_ , _ phys .
lett . _ * b386 * ( 1996 ) 175 , http://xxx.lanl.gov/abs/hep-ph/9605412[hep-ph/9605412 ]
. w. beenakker , r. hopker , t. plehn , p. m. zerwas , _
z. phys . _ * c75 * ( 1997 ) 349 , http://xxx.lanl.gov/abs/hep-ph/9610313[hep-ph/9610313 ] . j. guasch , w. hollik , j. sol , _ phys .
lett . _ * b437 * ( 1998 ) 88 , http://xxx.lanl.gov/abs/hep-ph/9802329[hep-ph/9802329 ] . j. guasch , w. hollik , j. sol , contribution to the _
2nd joint ecfa / desy study on physics and detectors for a linear electron - positron collider _ , p.565 , r. heuer , f. richard and p. zerwas
editors , ` lc - th-2000 - 013 ` , ` hep - ph/0001254 ` .
j. guasch , w. hollik , j. sol , _ phys .
lett . _ * b510 * ( 2001 ) 211 , hep - ph/0101086 .
m. bhm , h. spiesberger , w. hollik , _ fortsch .
* 34 * ( 1986 ) 687 ; + w. hollik , _ fortschr . phys . _
* 38 * ( 1990 ) 165 .
p. chankowski , s. pokorski , j. rosiek , _ nucl .
* b423 * ( 1994 ) 437 http://xxx.lanl.gov/abs/hep-ph/9303309[hep-ph/9303309 ] ; + a. dabelstein , _
z. phys . _
* c67 * ( 1995 ) 495 , http://xxx.lanl.gov/abs/hep-ph/9409375[hep-ph/9409375 ] ; _ nucl . phys . _ * b 456 * ( 1995 ) 25 , http://xxx.lanl.gov/abs/hep-ph/9503443[hep-ph/9503443 ] ; + h.e . haber , r. hempfling , _ _ phys . rev .
_ _ * 66 * ( 1991 ) 1815 ; + y. okada , m. yamaguchi , t. yanagida , _ _ prog . theor .
_ _ * 85 * ( 1991 ) 1 ; + j. ellis , g. ridolfi , f. zwirner , _ _ phys .
_ _ * b257 * ( 1991 ) 83 ; _ ibid .
_ * b262 * ( 1991 ) 477 ; + r. barbieri , m. frigeni , _ _ phys .
_ _ * b258 * ( 1991 ) 395 ; + j.a .
bagger , k. matchev , d.m .
pierce , r. zhang , _ _ nucl .
_ _ * b491 * ( 1997 ) 3 , ` hep - ph/9606211 ` .
j. m. frre , d. r. t. jones , s. raby , _ nucl .
phys . _ * b222 * ( 1983 ) 11 ; + m. claudson , l. j. hall , i. hinchliffe , _ nucl .
_ * b228 * ( 1983 ) 501 ; + c. kounnas , a. b. lahanas , d. v. nanopoulos , m. quirs , _ nucl .
_ * b236 * ( 1984 ) 438 ; + j. f. gunion , h. e. haber , m. sher , _ nucl .
* b306 * ( 1988 ) 1 .
d. pierce , a. papadopoulos , _ phys .
* d50 * ( 1994 ) 565 , http://xxx.lanl.gov/abs/hep-ph/9312248[hep-ph/9312248 ] ; + _ nucl .
phys . _ * b430 * ( 1994 ) 278 , http://xxx.lanl.gov/abs/hep-ph/9403240[hep-ph/9403240 ] . j. guasch , j. sol , _ z. phys . _
* c74 * ( 1997 ) 337 , http://xxx.lanl.gov/abs/hep-ph/9603441[hep-ph/9603441 ] .
j. l. kneur , g. moultaka , _ phys .
* d59 * ( 1999 ) 015005 , http://xxx.lanl.gov/abs/hep-ph/9807336[hep-ph/9807336 ] ; + _ phys .
_ * d61 * ( 2000 ) 095003 , http://xxx.lanl.gov/abs/hep-ph/9907360[hep-ph/9907360 ] ; + s. y. choi _
_ , _ eur .
phys . j. _
* c14 * ( 2000 ) 535 , http://xxx.lanl.gov/abs/hep-ph/0002033[hep-ph/0002033 ] .
h. eberl , m. kincel , w. majerotto , y. yamada , _ phys .
* d64 * ( 2001 ) 115013 , hep - ph/0104109 .
t. fritzsche , w. hollik , eur .
j. c * 24 * ( 2002 ) 619 , ` hep - ph/0203159 ` .
d. garcia , j. sol , _ mod .
_ * a9 * ( 1994 ) 211 ; + p.
h. chankowski _ et al .
_ , _ nucl .
phys . _ * b417 * ( 1994 ) 101 .
j. kblbeck , m. bhm , a. denner , _ comput .
_ * 60 * ( 1990 ) 165 ; + t. hahn , _ comput .
_ * 140 * ( 2001 ) 418 , hep - ph/0012260 ; + t. hahn , c. schappacher , _ comput .
* 143 * ( 2002 ) 54 , hep - ph/0105349 .
t. hahn , m. prez - victoria , _ comput .
* 118 * ( 1999 ) 153 , http://xxx.lanl.gov/abs/hep-ph/9807565[hep-ph/9807565 ] ; + t. hahn , _ feynarts _ , _ formcalc _ and _ looptools _ user s guides , available from http://www.feynarts.de[`http://www.feynarts.de ` ] ; + g. j. van oldenborgh , _ comput .
commun . _
* 66 * ( 1991 ) 1 .
m. a. diaz , s. f. king , d. a. ross , _ _ nucl .
_ _ * b529 * ( 1998 ) 23 http://xxx.lanl.gov/abs/hep-ph/9711307[hep-ph/9711307 ] ; + s. kiyoura , m. m. nojiri , d. m. pierce , y. yamada , _ phys .
rev . _ * d58 * ( 1998 ) 075002 http://xxx.lanl.gov/abs/hep-ph/9803210[hep-ph/9803210 ] ; + t. blank , w. hollik , http://xxx.lanl.gov/abs/hep-ph/0011092[hep-ph/0011092 ] .
d. e. groom _ et al .
. phys . j. _ * c15 * ( 2000 ) 1 .
a. sommerfeld : atombau und spektrallinien .
braunschweig : vieweg , 1939 .
a. d. sakharov , _ zh . eksp .
fiz . _ * 18 * ( 1948 ) 631 [ _ sov .
usp . _ * 34 * ( 1991 ) 375 ] .
d. m. pierce , j. a. bagger , k. matchev , r. zhang , _ nucl .
phys . _ * b491 * ( 1997 ) 3 , ` hep - ph/9606211 ` ; + k. s. babu , c. kolda , _ phys .
_ * b451 * ( 1999 ) 77 , ` hep - ph/9811308 ` ; + f. borzumati , g. r. farrar , n. polonsky , s. thomas , _ nucl .
phys . _ * b555 * ( 1999 ) 53 , ` hep - ph/9902443 ` ; + j. guasch , w. hollik , s. pearanda , _ phys .
_ * b515 * ( 2001 ) 367 , hep - ph/0106027 .
m. carena , d. garcia , u. nierste , c.e.m .
wagner , _ nucl .
* b577 * ( 2000 ) 88 , ` hep - ph/9912516 ` .
p. ciafaloni , d. comelli , _ phys .
* b446 * ( 1999 ) 278 , hep - ph/9809321 ; + v. s. fadin , l. n. lipatov , a. d. martin , m. melles , _ phys .
rev . _ * d61 * ( 2000 ) 094002 , ` hep - ph/9910338 ` ; + m. beccaria , m. melles , f. m. renard , c. verzegnassi , _ phys .
_ * d65 * ( 2002 ) 093007 , ` hep - ph/0112273 ` .
a. bartl _ et al .
_ , _ phys .
* b435 * ( 1998 ) 118 , http://xxx.lanl.gov/abs/hep-ph/9804265[hep-ph/9804265 ] ; + a. bartl _ et al .
_ , _ phys .
* b419 * ( 1998 ) 243 , http://xxx.lanl.gov/abs/hep-ph/9710286[hep-ph/9710286 ] ; _ phys .
_ * d59 * ( 1999 ) 115007 ; + | we present a definition of an on - shell renormalization scheme for the sfermion and chargino - neutralino sector of the minimal supersymmetric standard model ( mssm ) . then , apply this renormalization framework to the interaction between charginos / neutralinos and sfermions .
a kind of universal corrections is identified , which allow to define effective chargino / neutralino coupling matrices . in turn , these interactions generate ( universal ) non - decoupling terms that grow as the logarithm of the heavy mass . therefore the full mssm spectrum must be taken into account in the computation of radiative corrections to observables involving these interactions . as an application we analyze the full one - loop electroweak radiative corrections to the partial decay widths @xmath0 and @xmath1 for all sfermion flavours and generations .
these are combined with the qcd corrections to compute the corrected branching ratios of sfermions .
it turns out that the electroweak corrections can have an important impact on the partial decay widths , as well as the branching ratios , in wide regions of the parameter space .
the precise value of the corrections is strongly dependent on the correlation between the different particle masses .
* fermionic decays of sfermions : a complete discussion at one - loop order * jaume guasch@xmath2 , wolfgang hollik@xmath3 , joan sol@xmath4 _ @xmath5 institut fr theoretische physik , universitt karlsruhe , kaiserstrae 12 , + d-76128 karlsruhe , germany + @xmath6 theory group ltp , paul scherrer institut , ch-5232 villigen psi , switzerland + @xmath7 max - planck - institut fr physik , fhringer ring 6 , d-80805 mnchen , germany + @xmath8 departament destructura i constituents de la matria , universitat de barcelona , diagonal 647 , e-08028 , barcelona , catalonia , spain + @xmath9 institut de fsica daltes energies , universitat autnoma de barcelona , e-08193 , bellaterra , barcelona , catalonia , spain _ |
anomalous physical properties and unconventional superconductivity around the magnetic instability have been the subjects of intensive current interests .
let us first look at fig .
[ fig1 ] , showing a phase diagram around the magnetic instability or quantum critical ( qc ) point which separates the paramagnetic fermi liquid(fl ) phase and the magnetic ordered phase with the ordering vector @xmath1 at @xmath2 . substances belonging to this area have been called nearly and weakly ferro- and antiferromagnetic metals .
it seems that the magnetism in this area was called first attention by the discovery of weak itinerant ferromagnets such as zrzn@xmath3 and sc@xmath4 in @xcite .
the curie - weiss(cw ) magnetic susceptibility observed in these systems posed a serious challenge to theorists , urging them to consider a new mechanism without local moments . for nearly ferromagnetic metals , on the other hand , attention was attracted by a prediction based on the paramagnon theory that the effective mass in the fl regime diverges logarithmically as one approaches the qc point @xcite .
since then the theory of spin fluctuations has been developed to deal with the problems in the entire area of the phase diagram shown in fig .
the self - consistent quantum mechanical theory of the coupled modes of spin fluctuations or the self - consistent renormalization ( scr ) theory not only succeeded in explaining the curie - weiss susceptibility of @xmath5 but also predicted various singular behaviors of the physical quantities in the qc regime , which have been confirmed by subsequent experimental investigations @xcite .
the discovery of superconductivity of the materials in this area , heavy electron systems @xcite , organic conductors @xcite and in particular high-@xmath0 cuprates @xcite has renewed and enormously enhanced the interests in this area . upon discovery of the high temperature superconductivity of the cuprates the attention of investigators
was first concentrated on the anomalous normal state properties or the so - called non - fermi liquid properties of these substances where strongly correlated quasi 2-dimensional(2d ) electrons in each cuo@xmath3 layer were considered to play a predominant role in conduction @xcite .
the spin fluctuation theory was extended to cover 2d systems and it was found that the results in the qc and cw regimes were consistent with the observed non - fl behaviors of the cuprates @xcite .
the experimental results were successfully analyzed in terms of the parametrized scr theory and the parameter values were estimated . in the next step
the spin fluctuation - induced superconductivity was studied by using the spin fluctuations thus determined quantitatively . using both weak coupling @xcite and
then strong coupling theories @xcite the superconductivity of @xmath6-@xmath7 symmetry was derived with the values of the transition temperature @xmath0 consistent with the experimental results .
the @xmath6-wave symmetry of the superconducting order parameter of the cuprates was confirmed by experimental investigations in the following years @xcite .
subsequent studies on the relation between @xmath0 and the spin fluctuation parameters showed that the value of @xmath0 was roughly proportional to the parameter @xmath8 indicating the energy spread of the spin fluctuations @xcite this relation may be extended to certain 3d systems and it was shown that the plots of experimental values of @xmath0 against @xmath8 for cuprates and heavy electron systems came around a straight line @xcite .
this result suggests the common origin of superconductivity in these systems .
theoretical studies using the hubbard and the @xmath6-@xmath9 models have been advanced not only for the cuprates @xcite but also for the 2d organic superconductors @xcite by using the fluctuation exchange ( flex ) approximation which may be regarded as the simplest self - consistent theory for the coupled modes of spin fluctuations .
the results turned out to be successful indicating that the hubbard model with proper transfer matrices and a proper electron occupation is a good model for these systems . a phase diagram is then calculated within the same approximation in a parameter space of the hubbard model containing the parameter values for both the cuprates and the organic superconductors .
it was found that both of them belong to a continuous region of a superconducting phase , indicating their common physical origin @xcite .
the theoretical approaches discussed so far are the one around the magnetic instability or qc point , which belongs to the intermediate coupling regime .
as will be discussed later in this article there have been growing pieces of evidence to indicate that the high-@xmath0 cuprates are in the intermediate coupling regime and the spin fluctuation theory is reasonably applied to the main part of the problem . at this point
we would like to mention the pseudo - gap phenomena observed in the under - hole - doped cuprates as a still remaining outstanding subject in the high-@xmath0 problems .
although the physics of doped mott insulators in the entire concentration range , including the pseudogap phenomena , seems still to be worked out theoretically , it is fortunate that at least a substantial part of the sc phase of the cuprates is around the qc regime and is described essentially in terms of the theories of spin fluctuations . in the following sections we summarize very briefly the developments in the theory of spin fluctuations as applied to magnetism and superconductivity and
discuss to what extent we now understand these problems .
the dynamical susceptibility around the magnetic instability for the state with the ordering vector @xmath1 may be expressed for small @xmath10 and @xmath11 as follows : @xmath12 where @xmath13 and @xmath14 for @xmath15 and @xmath16 , respectively . for non - interacting systems
@xmath17 , @xmath18 and @xmath19 are calculated from the given band structure @xcite and rpa result is simply given by @xmath20 , @xmath21 being the on - site interaction constant and the susceptibility is defined without a factor @xmath22 . in the self - consistent renormalization theory the renormalized values of @xmath17 and @xmath18 usually stay constant around the magnetic instability while the temperature dependence of @xmath23 is strikingly renormalized by the mode - mode coupling effects , i.e. , from the sommerfeld expansion to the quantum critical and curie - weiss behaviors . leaving derivations of the theory for ref .
@xcite we show here only the equations for @xmath24 .
@xmath25 where @xmath26 is the mode - mode coupling constant for those with the wave vectors around @xmath1 . eqs .
( [ eq:1 ] ) and ( [ eq:2 ] ) should be solved self - consistently for @xmath23 . for convenience
we introduce here a reduced inverse susceptibility and the following parameters : @xmath27 where @xmath28 is the ordered moment in @xmath29 per magnetic atom , @xmath30 is the effective brillouin zone boundary vector , @xmath31 dimensionality and @xmath32 is the volume per magnetic atom . the dynamical susceptibility is now written as @xmath33 with @xmath34 and @xmath35 , \nonumber \end{aligned}\ ] ] with @xmath36 we summarize the results of this theory as follows @xcite : 1 .
@xmath37 obeys the curie - weiss law with the origin different from the traditional local moment mechanism .
the quantum critical behaviors are obtained as shown in table [ tab1 ] .
the 3d and 2d results were obtained in 1970 s and 1990 s , respectively .
many of the results were confirmed by experimental investigations from 1970 s to 1990 s .
recent renormalization group studies of the same problem lead to the same qc indices as the above results @xcite .
the curie and nel temperatures of 3d magnets are given by @xmath38 these results were confirmed qualitatively(chemical and physical pressure dependence ) and quantitatively by experimental investigations since 1970 s .
expressions for various physical quantities with the 4 scr parameters are given and are applicable to a wide range of temperature including both qc and cw regimes .
.quantum critical behaviors of physical quantities . [ cols="^,^,^,^,^ " , ] for early investigations on transition metals and their compounds we refer to @xcite .
recently non - fermi liquid properties of certain heavy @xmath39-electron systems are investigated intensively and are analyzed in terms of the qc spin fluctuation of 3d or 2d character depending on substances @xcite .
from the beginning of investigations the anomalous or non - fermi liquid properties in the normal state of the cuprates were represented by the @xmath40-linear electrical resistivity in a wide temperature range , anomalous temperature dependences of the hall coefficient @xmath41 and nmr @xmath42 , @xmath41 and @xmath43 showing the curie - weiss behaviors , and the @xmath44-linear relaxation rate of the optical conductivity , etc . in an early stage of investigations all of these properties except for the hall coefficient were successfully analyzed in a consistent manner in terms of the above discussed spin fluctuation theory .
the phenomenological parameters of the theory were determined from analyses of the resistivity and @xmath42 . by using
thus determined dynamical susceptibility a parameter free comparison between theory and experiment were performed successfully on the optical conductivity of yba@xmath3cu@xmath4o@xmath45 @xcite .
the hall effect was studied recently by using the kubo - eliashberg formalism on the hubbard model with appropriate transfer matrices and electron occupations @xcite .
the rrpa ( renormalized random phase approximation ) or flex approximation was used for the spin fluctuations and the importance of the vertex corrections in dealing with strongly anisotropic scatterings was emphasized .
the results for @xmath41 were consistent with experimental results both for the hole- and electron - doped cuprates ; @xmath41 in the former is positive and shows a curie - weiss behavior while the latter decreases with lowering temperature from a positive value at high temperature and changes sign , the decrement showing a curie - weiss behavior . in the second stage of investigations
the pseudo - gap phenomena observed in under - hole - doped cuprates have called attention of investigators .
this may be regarded as low temperature corrections to the non - fl behaviors as discussed in the first stage .
we will discuss this problem later in this article and now move to the problem of superconductivity induced by the spin fluctuations .
since the anomalous normal state properties were explained by the spin fluctuation theory and the dynamical susceptibilities were estimated quantitatively from analyses of the experimental results , next problem is naturally to see if the same spin fluctuations can explain the high @xmath0 superconductivity of the cuprates .
it was shown in earlier studies by using a weak coupling theory @xcite and a strong coupling theory @xcite that the antiferromagnetic spin fluctuations can give rise to a @xmath6-wave superconductivity .
the spin fluctuation induced superconductivity of the cuprates were studied first by using the weak coupling theory @xcite and then the strong coupling theory @xcite which is more appropriate .
the results showed the @xmath6-@xmath7 symmetry of the order parameter and values of @xmath0 consistent with the observed values .
after the unconventional @xmath6-wave symmetry of the order parameter was confirmed experimentally the relations between the spin fluctuation parameters and @xmath0 were investigated .
the most remarkable result is that @xmath0 is roughly proportional to @xmath8 , the energy spread of the spin fluctuations .
dependences of @xmath0 on the other parameters , @xmath46 , @xmath47 , and doping concentration are relatively weak . the effect of dimensionality was also studied . in favorable cases
the values of @xmath0 can be comparable in 3d and 2d systems and are roughly proportional to @xmath8 , although the dependences of @xmath0 on the other parameters are much more significant in 3d than in 2d systems and 2d seems to be generally more favorable than 3d for superconductivity @xcite . in the same context flex approximation
was applied to a nearly half - filled hubbard model with a simple cubic lattice and the calculated @xmath0 was found to be extremely small @xcite . a more recent study of a spatially anisotropic hubbard model with a transfer parameter interpolating between 2d and 3d systems
show that @xmath0 varies only slowly on the 2d side of the parameter values and then decreases rapidly as one approaches the cubic limit @xcite .
we show in fig .
[ fig2 ] the plots of experimental values of @xmath48 against @xmath8 for cuprates and heavy electron superconductors @xcite .
the plots come around a straight line , suggesting the common origin of superconductivity in these groups of systems .
, the characteristic temperature indicating the energy spread of spin fluctuation.,scaledwidth=60.0% ] fully microscopic calculations of spin fluctuation - induced superconductivity were carried out on the hubbard and the @xmath6-@xmath9 models by using the flex approximation @xcite , the 3rd order perturbation theory @xcite , and variational monte carlo method @xcite .
the results may be summarized as follows @xcite : 1 . superconductivity of @xmath6-@xmath7 symmetry is obtained in a fairly large range of intermediate values for @xmath49 .
the calculated values of @xmath0 are of reasonable magnitude compared with experiment .
similar results are obtained both for the hubbard and @xmath6-@xmath9 models .
3 . for the superconducting state ,
the neutron resonance peak observed in ybco is explained by the flex calculations and by rpa calculations assuming the bcs ground state .
also a peak - dip - hump structure in the one electron spectral density as observed by angle - resolved photoemission experiment is explained .
4 . observed differences between the phase diagrams for the electron - doped and hole - doped cuprates are explained at least qualitatively ; a substantially larger concentration range of af phase and smaller values of @xmath0 in the former as compared with the latter @xcite . the calculated ratio of the sc gap and @xmath50 for the former is about half of that in the latter in accord with measured results .
this difference may be related with the van hove singular points which are close to the hot spots , where scatterings due to af spin fluctuations is particularly strong , in the hole - doped systems but not in the electron - doped systems .
it may be worth while to remark that the flex approximation is the simplest possible approach for self - consistently renormalized spin fluctuations and the above success strongly favors the spin fluctuation mechanism for the superconductivity in the cuprates .
it is also remarkable that the hubbard model in the intermediate coupling regime seems to be a good model for the cuprates if one chooses the transfer matrices so as to reproduce the observed fermi surface for each substance .
an organic system @xmath51-(bedt - ttf)@xmath3cun(cn)@xmath3cl is an antiferromagnetic insulator at ambient pressure . under applied pressure
it undergoes a weak first order transition , at @xmath52 , into a metallic state which shows superconductivity below @xmath53 . to a good approximation this substance
is considered to be described by a half - filled hubbard model consisting of antibonding dimer orbitals arranged in a square lattice . as for the transfer integrals we may take @xmath54 between the nearest neighbors and @xmath55 between the second neighbors in one of the diagonal directions .
this model was studied with the flex approximation and the results showed the superconducting order parameter of @xmath6-@xmath7 symmetry and reasonable values for @xmath0 @xcite . since the approach here is just the same as the one discussed in the preceding section for the cuprates , the common success naturally indicates the same origin of superconductivity in these systems .
the above results are of particular interest since the organic compound is found in the metallic side of the mott transition while the high-@xmath0 cuprates are doped mott insulators .
the former is clearly in the intermediate coupling regime and existing approaches from the strong coupling limit does not seem to work . in order to see the situations more clearly a phase diagram was calculated in a parameter space of the hubbard model taking the following quantities for the 3 axes of the parameter space : @xmath49 , @xmath56 and @xmath57 , @xmath58 being the number of electrons per site @xcite .
the flex approximation was employed and the results were extrapolated to @xmath2 with the use of pad approximants .
the calculated phase diagram is shown in fig .
we see that the superconducting states of cuprates and those of the organic compounds are found in the same connected region of a superconducting phase in this phase diagram .
superconducting ( sc ) and antiferromagnetic ( afm ) instability surfaces are indicated .
the boundary between antiferromagnetic metal(afm ) and insulator(afi ) should depend on the impurity potential and is sketched here just to show general idea of its location.,scaledwidth=60.0% ]
among the mechanisms of superconductivity in strongly correlated electron systems the one for the heavy electron systems is considered mainly due to the spin fluctuations , although additional contributions of orbital and charge fluctuations are under investigation @xcite . for the 2d organic compounds the only mechanism explicitly worked out
so far seems to be the spin fluctuation mechanism . for the high-@xmath0 cuprates , on the other hand ,
many different mechanisms have been proposed .
in addition to the approaches as discussed in the above , various approaches from the strong coupling limit based on the @xmath61-@xmath62 model have been pursued extensively @xcite since the @xmath61-@xmath62 model is an approximation to the hubbard model from the strong coupling limit it may be pertinent here to discuss on the phase diagrams of the hubbard model and see where in the parameter space the high-@xmath0 cuprates are really located .
let us first look at a phase diagram of a half - filled hubbard model including the magnetic qc point , an antiferromagnetic phase and a metal to insulator transition , fig .
appropriate ways of approaches for different sections of the phase diagram are also indicated in the figure .
we first emphasize that in the ground state the hartree - fock(hf)-rpa is correct in the strong coupling regime where the ordered moment is enough polarized so that the correlation effect is insignificant ; two electrons with opposite spins seldom encounter with each other .
the spin wave dispersion calculated by rpa in the antiferromagnetic ground state agrees , in the limit of small @xmath63 , precisely with the one calculated from the heisenberg model with the anderson kinetic superexchange interactions @xcite .
since the hf - rpa is known to be correct in the weak coupling limit it should make a fair interpolation between the strong and weak coupling limits in the ground state . in the intermediate coupling regime , where the electron - electron correlations are most significant
, it is a reasonable approach to start from hf - rpa and to make many body corrections .
the self - consistent theory of renormalized spin fluctuations may be an example of possible approaches . at finite temperatures ,
however , hf - rpa is not at all a good approximation in any regimes of present interest . for the mott insulator phase with large @xmath49 the local moment model with the anderson kinetic superexchange ( @xmath63 expansion ) is valid while for weak and intermediate @xmath49 just covering the antiferromagnetic instability the scr spin fluctuation theory is considered to be an adequate approach . for the intermediate regime between these two including the mott transition no really satisfactory approach at finite temperatures is known so far .
we note that 2d organic superconductors are on the metallic side of the first order mott transition and is considered to be located around the hidden antiferromagnetic instability and thus are expected to be well treated by the spin fluctuation theory .
next we discuss doped mott insulators .
if we neglect the impurity potential the ground state of a doped mott ( antiferromagnetic ) insulator is an antiferromagnetic metal . on doping we first have small fermi surfaces which expand with increasing doping concentration at the expense of reduced gap area .
then we have a large fermi surface with small gap area and finally have a paramagnetric metal through the qc point where the af gap just vanishes . as is seen in fig .
[ fig5 ] the magnetic instability or qc point should make a continuous line in a @xmath49 against n plane extending from the half - filled ( @xmath64 ) case with relatively small @xmath49 toward larger @xmath65 and @xmath66 .
the spin fluctuation theory is an approach along the qc line from the side of weaker coupling and is expected to cover at least the intermediate coupling regime . as a matter of fact the values for @xmath49 used in the above calculations are around @xmath67 and
thus @xmath21 is about @xmath68 times the band width .
this is consistent with the value estimated from photoemission experiments @xcite .
according to recent analyses of the spin wave energy dispersion in antiferromagnetic la@xmath3cuo@xmath69 , a parent of high-@xmath0 cuprates , as measured by neutron scattering studies @xcite the dispersion in the entire brillouin zone is well reproduced by an rpa calculation on the hubbard model with @xmath70 @xcite . in order to get a reasonable fit with the local moment model one needs to consider the higher order terms in the @xmath63 expansion including the ring exchange terms @xcite .
this fact seems to evidence the intermediate coupling nature of the cuprates and supply another support for the applicability of the spin fluctuation theory to the high-@xmath0 problems and explain the reason for its success .
on the other hand the approaches based on the @xmath61-@xmath62 model with a few terms of anderson superexchange interactions might not even cover the parent insulator state .
applicability of this approach as an approximation to the hubbard model may also be limited to a very low doping concentration range . for larger doping concentrations it may take us into a different world from the one described by the hubbard model .
pseudogap phenomena were first observed in nmr @xmath42 measurements on underdoped ybco[51 , 52 ] as a pseudo - gap in spin excitations .
the measurements were then extended to transport properties , one electron spectral density as observed by arpes and tunnelling experiments and various theoretical mechanisms were proposed [ 53 ] . according to the flex calculations @xmath0 generally tends to increase with lowering doping concentration until it reaches the af phase boundary calculated introducing weak 3-dimensionality .
this behavior is consistent with the experimental results on the electron - doped cuprates . for the hole - under - doped cuprates
, however , experimental results seem to suggest that there is some additional mechanism to suppress both the af and sc orderings , giving rise to the pseudo - gap at the same time
. among various possibilities proposed so far one natural direction from the present point of view may be to consider superconducting fluctuations(@xmath9-@xmath9 channel ) in addition to the af spin fluctuations(@xmath9-@xmath71 channel ) .
the sc fluctuation is expected to be particularly significant in 2d systems with short coherence lengths . here
we refer to an investigation reported recently[54 - 56 ] .
the sc fluctuations are taken into account within the @xmath61-matrix approximation together with the flex spin fluctuations and the equations are solved self - consistently with still further approximation of expanding the @xmath61-matrix in @xmath72 and @xmath44 near the sc critical point .
the results include many attractive features : ( 1 ) in the under - hole - doped regime @xmath0 decreases from the flex value and the value of @xmath0 decreases with lowering doping concentration .
the af phase tends to be suppressed at the same time .
( 2 ) pseudogap behaviors in the one electron spectral density and reduction of the nmr relaxation rate @xmath43 are obtained .
( 3 ) observed pseudo - gap behaviors in the transport properties : electrical resistivity , hall coefficient , magneto - resistance , thermo - electric power , and the nernst coefficient are all explained consistently .
( 4 ) calculated effects are significant in the hole - underdoped cuprates but not in the electron - doped cuprates in accord with experiment . in order to make this scenario convincing it
is desired to extend the theory to still lower doping concentrations and clarify the phase diagram near the af insulator phase .
we have summarized the present status of the theories of the spin fluctuation mechanism in explaining anomalous or non - fermi liquid behaviors and unconventional superconductivity in strongly correlated electron systems around the magnetic quantum critical point , referring to the high-@xmath0 cuprates , 2d organic compounds and heavy electron systems . so far
the spin fluctuation theories seem to be successful in explaining at least essential parts of the problems , indicating in particular that the spin fluctuation is the common origin of superconductivity in these systems . as for the high-@xmath0
cuprates at least the central part of the sc phase seems to be approached by the spin fluctuation theory .
it was fortunate that the high-@xmath0 cuprates are in the intermediate coupling regime and are close to the magnetic instabilities .
it still remains to describe all the hole - underdoped regime lying between the af insulator phase and the optimal concentration regime , where the pseudo - gap phenomena are still controversial indicating the need for another mechanism in addition to the antiferromagnetic spin fluctuations .
the author wishes to thank k. ueda for critical reading of the manuscript and h. kondo for discussion and help in preparing the manuscript .
b. t. matthias , r. m. bozorth , _ phys .
rev . _ * 100 * , 604 ( 1958 ) [ zrzn@xmath3 ] ; b. t. matthias , a. m. clogston , h. j. williams , e. corenzwit , r. c. sherwood , _ phys . rev .
lett . _ * 7 * , 7 ( 1961 ) [ sc@xmath4 in ] .
see for a recent review : n. d. mathur , f. m. grosche , s. r. julian , i. r. walker , d. m. freye , r. k. w. haselwimmer , g. g. lonzarich , _ nature _ * 394 * , 39 ( 1998 ) ; see also : z. fisk , d. w. hess , c. j. pehtick , d. pines , j. l. smith , j. d. thompson , j. d. willis , _ science _ * 239 * , 33 ( 1988 ) .
k. ueda , t. moriya , y. takahashi , in _ electronic properties and mechanisms of high temperature superconductors _ , edited by t. oguchi et al . ,
north holland , amsterdam , p.145 ( 1992 ) ; _ j. phys .
solids _ * 53 * , 1515 ( 1992 ) .
recently @xmath6-@xmath74 symmetry of the order parameter is proposed base on the thermal conductivity measurement : k. kuroki , t. kimura , r. arita , y. tanaka , y. matsuda , _ phys .
b _ * 65 * , 100516(r ) ( 2002 ) . | we summarize the present status of the theories of spin fluctuations in dealing with the anomalous or non - fermi liquid behavior and unconventional superconductivity in strongly correlated electron systems around their magnetic instabilities or quantum critical points .
arguments are given to indicate that the spin fluctuation mechanisms is the common origin of superconductivity in heavy electron systems , 2-dimensional organic conductors and high-@xmath0 cuprates . |
for this analysis , electron candidates are selected in the momentum range from 1.5 to 3.5 @xmath17 in the rest frame with a solid angle defined by the electromagnetic calorimeter acceptance.the inclusive electron spectrum for charmless semileptonic @xmath6 decays , measured in the rest frame in the momentum range of 2.32.6 @xmath17 , is used to extract @xmath7 . to suppress low - multiplicity qed processes and continuum processes consisting of nonresonant @xmath18 production @xmath19
at least three charged tracks per event are required and a cut on the ratio of fox - wolfram moments @xmath20 @xcite is applied .
the missing momentum four - vector @xmath21 , where all momenta are measured in the laboratory frame and @xmath22 refers to the four - momentum of the initial state of the colliding beams , can be used to select semileptonic events : @xmath23 is requested to be larger than 1 @xmath17 , to point into the detector fiducial volume and the angle between the electron candidate and the missing momentum is required to be greater than @xmath24 .
candidate electrons are rejected if , when paired with an opposite - sign electron , the invariant mass of the pair is consistent with the @xmath25 mass ( @xmath26 @xmath16 ) . for the selection criteria described above
, the detection efficiency for charmless semileptonic decays in the electron momentum interval of @xmath27 @xmath17 ranges from @xmath28 to @xmath29 .
the raw spectrum of the highest momentum electron after the subtraction of continuum background ( determined from resonance data sample ) is shown in fig .
[ e1_spectrum]a . also shown are the mc predictions of the expected signal from @xmath30 decays and background contributions from all other processes .
the result of the subtraction of all backgrounds is shown in fig .
[ e1_spectrum]b . for a given interval in the electron momentum
, the inclusive partial branching ratio is calculated according to @xmath31 here @xmath32 refers to the number of electrons detected -peak and @xmath33 refers to the fitted continuum background in a specified momentum interval , @xmath34 is the background from @xmath35 decays , @xmath36 is the total efficiency for detecting a signal electron from @xmath37 decays ( including bremsstrahlung in detector material ) , and @xmath38 accounts for the distortion of the electron spectrum due to final - state radiation . as the overall normalization
the total number of produced @xmath35 events is used , @xmath39.the systematic error introduced by the efficiency estimation for the signal events is 5% .
the uncertainty in the continuum background subtraction is 5% .
the error coming from the background modeling translates to a relative error of 3% .
variations of the colliding beam energy introduce a systematic error in the @xmath40 background subtraction ( 5% ) . the total systematic error on the partial branching ratio measurement
is @xmath41.the fully corrected differential branching ratio as a function of the electron momentum is shown in fig .
[ e1_spectrum]c . integrating over the interval from 2.3 to 2.6 @xmath17
, we get : @xmath42 to determine the charmless semileptonic branching fraction @xmath43 from the partial branching fraction @xmath44 , the fraction @xmath45 of the spectrum that falls into the momentum interval @xmath46 is needed .
the cleo collaboration has recently used the measurement of the inclusive photon spectrum from @xmath47 transitions @xcite to derive @xmath45 for @xmath48 transition .
they quote a value of @xmath49 for the interval @xmath46 from 2.3 to 2.6 @xmath17 . relying on the cleo measurement
, the result presented here translates into a total branching ratio of @xmath50 .
we extract @xmath0 from the measured inclusive charmless semileptonic branching fraction with the relation in @xcite and the average @xmath6 lifetime of @xmath51 @xcite and find @xmath52 here the first error is the combined statistical and systematic error , and the second refers to the uncertainty on the extraction of @xmath0 from relation in @xcite .
the third one is taken from the cleo analysis and is related to the experimental determination of @xmath53 .
the last error accounts for uncertainties related to assumption that @xmath54 spectrum can be used for shape function modeling in @xmath48 transition .
this analysis is based on events in which one of the @xmath6 meson decays in a fully reconstructed hadronic final state ( @xmath55 ) and the other one is identified as decaying semileptonically by the presence of an electron or a muon . the full reconstruction of one of the two @xmath6 mesons reduces the overall efficiency , but allows to reconstruct both the neutrino and the hadronic system ( @xmath56 ) , to determine the flavour and to separate charged and neutral @xmath6 mesons . in order to reduce systematic uncertainties due to efficiency determination
we extract the branching ratio @xmath57 after measuring the number of events with one identified lepton.to fully reconstruct a large sample of @xmath6 mesons , hadronic @xmath6 decays of the type @xmath58 are selected .
@xmath59 represents a collection of hadrons with a total charge of @xmath60 , composed of @xmath61 , where @xmath62 , @xmath63 , and @xmath64 .
the kinematic consistency of a @xmath55 candidate with a @xmath6 meson decay is checked using two variables , the beam energy - substituted mass @xmath65 and the energy difference , @xmath66 . here
@xmath67 refers to the total energy in the center of mass frame , and @xmath68 and @xmath69 denote the momentum and energy of the @xmath55 candidate in the same frame , respectively . in events with more than one reconstructed @xmath6 decay
, the decay mode with highest purity is selected .
semileptonic decays , @xmath70 , recoiling against the @xmath55candidate are identified by an electron or muon candidate with a minimum momentum ( @xmath71 ) greater than @xmath72 in the rest frame .
correlation between the charge of the prompt leptons and the flavor of the @xmath55 is imposed ( @xmath73 mixing rate is used to extract the prompt lepton yield in case of neutral candidates ) .
the hadron system @xmath56 in the decay @xmath5 is made of charged tracks and neutral energy depositions in the calorimeter that are not associated with the @xmath55 candidate and not identified as a lepton .
the mass of the hadronic system is determined by a kinematic fit that imposes four - momentum conservation , the equality of the masses of the two @xmath6 mesons , and forces @xmath74.the selection of @xmath3 decays is tightened by requiring exactly one charged lepton with @xmath75 , charge conservation ( @xmath76 ) , and missing mass consistent with zero ( @xmath77 ) .
these criteria improve the resolution in @xmath2 and suppress the dominant @xmath4 decays , many of which contain additional neutrinos or undetected @xmath78 .
we suppress the @xmath79 background with a partial reconstruction in which only the slow pion from the @xmath80 decay and the lepton are reconstructed .
we veto events with charged or neutral kaons in the @xmath56 system to reduce the background from @xmath4 decays .
the impact of the selection criteria on the @xmath2 distribution is illustrated on mc in fig .
[ fig : mxeff]a .
we determine @xmath81 from @xmath82 , the observed number of @xmath83 events , and @xmath84 , the number of events with at least one charged lepton : @xmath85 here @xmath86 is the efficiency for selecting @xmath3 decays with all analysis requirements , @xmath87 is the fraction of signal events with @xmath88 , @xmath89 corrects for the difference in the efficiency due to the lepton momentum cut for @xmath5 and @xmath3 decays , and @xmath90 accounts for a possible efficiency difference in the @xmath55 reconstruction in events with @xmath5 and @xmath3 decays .
we derive @xmath84 from a fit to the distribution shown in fig .
[ fig : mxeff]b .
the fit uses an empirical description @xcite of the combinatorial background from continuum and events , together with a narrow signal @xcite peaked at the @xmath6 meson mass .
the residual background in @xmath84 from misidentified leptons and semileptonic charm decays amounts to @xmath91 and has been subtracted .
we obtain @xmath82 from the @xmath2 distribution with a @xmath92 fit to the sum of three contributions : signal , background @xmath93 from @xmath4 , and a background of less than @xmath94 from other sources ( misidentified leptons , secondary @xmath95 and charm decays ) . in each bin of the @xmath2 distribution ,
the combinatorial @xmath55 background is subtracted on the basis of a fit to the @xmath96 distribution .
figure [ fig : fitdata]a shows the @xmath2 distribution with the results of the fit superimposed .
the fit reproduces well the data having a @xmath97 . in the fit , the first bin is chosen to contain all events with @xmath2 less than @xmath98 while the other bins are chosen in order to separate the contribution from each resonant @xmath99 mode .
the @xmath2 cut , set at @xmath98 , has been optimized minimizing the total error .
figure [ fig : fitdata]b shows the @xmath2 distribution after background subtraction with finer binning .
table [ breakdown ] summarizes the results of fits with different requirements on @xmath2 , for electrons and muons , for neutral and charged @xmath55 candidates , and for different ranges of the @xmath55 purity , @xmath100 .
the results are all consistent within the uncorrelated statistical errors.we have performed extensive studies to determine systematic uncertainties .
we use events with charged and neutral kaons in the recoil of the @xmath55 candidate as a control sample to assess that the background from @xmath4 events is properly described .
the relative systematic error ( @xmath101 ) due to the selection criteria related to the reconstruction of particles in the event is @xmath102 .
the uncertainty of the @xmath55 combinatorial background subtraction is estimated by varying the signal shape function ( @xmath103 ) .
the impact of the binning is studied by changing the binning for @xmath104 ( @xmath105 ) .
the branching fractions of @xmath106 and of inclusive and exclusive @xmath107 mesons decays are varied within the world average uncertainties @xcite ( @xmath108 ) .
the limited amount of simulated events causes an uncertainty @xmath109 .
the uncertainty in the hadronization of the final state of @xmath3events is determined by measuring @xmath110 in bins of charged and neutral multiplicities and performing the fit using only the nonresonant signal model instead of the hybrid model(@xmath111 ) .
we also vary the branching fractions for charmless semileptonic @xmath6 decays by one standard deviation @xcite(@xmath112 ) .
the fraction of signal events with contents is varied by @xmath113 for the exclusive component and by @xmath114 for the inclusive one @xcite(@xmath115 ) . in the determination of @xmath116 and @xmath117
we allow the nonperturbative parameters to vary according to @xmath118 and @xmath119 , obtained by scaling the results in @xcite to @xmath120 in order to match the nonresonant mc generator @xcite .
we take into account the correlation of @xmath121 between @xmath122 and @xmath123(@xmath124).we combine the errors related to the detector and the signal and background modeling errors quadratically into the systematic error and obtain @xmath125 , where the errors are statistical , systematic , and theoretical related to the efficiency determination and extrapolation to the full @xmath2range respectively . combining the ratio @xmath81 with the measured inclusive semileptonic branching fraction of @xmath126 @xcite
, we obtain @xmath127 . using the relation in @xcite and the average @xmath6 lifetime of @xmath128
@xcite we find @xmath129 the first error is statistical , the second refers to the experimental systematic uncertainty , the third gives the theoretical uncertainty on the extrapolation of @xmath81 to the full @xmath2 range , and the last error combines quadratically the perturbative and nonperturbative uncertainties in the extraction of from the total decay rate .
.fit results for several data samples . [ cols="<,>,>,>,>",options="header " , ] [ breakdown ]
two different approaches for the extraction of ckm matrix element have been presented .
the analysis based on @xmath2 gives currently the most precise determination of .
this is primarily due to specific advantages of this technique : large phase - space acceptance and high purity of the sample ( signal over background ratio @xmath130 ) .
the two results are consistent and they are in agreement with previous inclusive measurements @xcite .
bigi , m. shifman , and n.g .
uraltsev , annu .
nucl . part .
sci . * 47 * , 591 ( 1997 ) ; j. e. duboscq _ et al .
_ [ cleo collaboration ] , phys . rev . lett . *
76 * , 3898 ( 1996 ) .
d. scora and n. isgur , phys .
d * 52 * , 2783 ( 1995 ) .
goity and w. roberts , phys . rev . *
d51 * , 3459 ( 1995 ) .
f. de fazio and m. neubert , jhep * 9906 * , 017 ( 1999 ) . t. sjstrand , comput .
. commun . * 82 * , 74 ( 1994 ) .
. fox and s. wolfram , phys .
41 * , 1581 ( 1978 ) .
the cleo collaboration , a. bornheim _ et al_. , hep - ex/0202019 ( 2002 ) .
b. aubert _ et al .
_ [ babar collaboration ] , phys .
d * 67 * , 031101 ( 2003 ) .
r. barate _
[ aleph collaboration ] , eur .
j. c * 6 * , 555 ( 1999 ) ; m. acciarri _ et al .
_ [ l3 collaboration ] , phys . lett .
b * 436 * , 174 ( 1998 ) ; p. abreu _ et al .
_ [ delphi collaboration ] , phys .
b * 478 * , 14 ( 2000 ) ; b. h. behrens _ et al .
_ [ cleo collaboration ] , phys . rev .
d * 61 * , 052001 ( 2000 ) ; g. abbiendi _ et al .
_ [ opal collaboration ] , eur .
j. c * 21 * , 399 ( 2001 ) ; a. bornheim _ et al . _
[ cleo collaboration ] , phys .
* 88 * , 231803 ( 2002 ) . | we present two inclusive measurements of cabibbo - kobayashi - maskawa ( ckm ) matrix element @xmath0 : one uses the lepton energy spectrum ( @xmath1 ) and the other uses the invariant mass of the hadronic system ( @xmath2 ) to discriminate signal ( @xmath3 ) and background ( @xmath4 ) events in @xmath5 transitions . both analyses are based on data samples collected by the detector at the pep - ii asymmetric - energy @xmath6 factory at slac . the element of the ckm matrix @xcite plays a central role in tests of the unitarity of this matrix : it s extraction , based on tree level decays , gives results that are independent of new physics contributions .
we report the determination of from two different measurements of the inclusive charmless semileptonic @xmath6 branching fraction , @xmath7 , using @xmath1 spectrum ( endpoint ) and the @xmath2 spectrum on the recoil of fully reconstructed @xmath6 mesons respectively.the selection of @xmath3 events is hampered by the presence of a large @xmath4background : @xmath1 and @xmath2 spectra are used to discriminate the two different transitions .
the endpoint analysis is sensitive to approximately 10% of the @xmath1 spectrum while the acceptance for the @xmath2 approach is larger : @xmath8 of the @xmath2 spectrum is selected by analysis cuts .
the extrapolation of the measured rates to the full phase space introduces theoretical uncertainties@xcite .
results also depend on the shape function ( sf ) modeling of @xmath9 quark fermi motion inside the @xmath6 meson.both measurements are based on data recorded by the detector @xcite at the 2 asymmetric - energy storage ring at slac .
the endpoint analysis data sample consists of about 23 million pairs ( 21 ) collected at the resonance ( on - peak ) , with an additional sample of 2.6 recorded about 40@xmath10below the peak ( off - peak ) , while the @xmath2 measurement uses a data sample of about 88 million pairs ( 82 ) on - peak .
monte carlo ( mc ) simulations of the detector based on 4 @xcite are used to optimize selection criteria and to determine signal efficiencies and background shapes . to simulate @xmath4 transitions
three models are employed : @xmath11 decay is modeled following a parametrization of form factors based on hqet @xcite ; for @xmath12 decays and higher mass charm meson states @xmath13 the isgw2 model @xcite is used ; nonresonant decays , @xmath14 , are modeled according to a prescription by goity and roberts @xcite . in the endpoint analysis the mc simulation of @xmath3 events
is based on the isgw2 model : the hadrons @xmath15 are represented by single particles or resonances with masses up to 1.5@xmath16 and nonresonant contributions are not included . in the @xmath2analysis @xmath3 transitions
are simulated with an hybrid model which is a mixture of resonant and nonresonant components .
the fermi motion of the @xmath9 quark inside the @xmath6 meson is implemented in the nonresonant component using the sf parameterization described in @xcite , and the fragmentation is handled by @xcite . |
in his seminal paper `` the future of data analysis '' ( tukey , @xcite ) , john w. tukey writes : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` it will still be true that there will be aspects of data analysis well called technology , but there will also be the hallmarks of stimulating science : intellectual adventure , demanding calls upon insight , and a need to find out ` how things really are ' by investigation and the confrontation of insights with experience '' ( p. 63 ) .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ fast forward to 2013 in the age of information technology , these words of tukey ring as true as fifty years ago , but with a new twist : the ubiquitous and massive data today were impossible to imagine in 1962 . from the point of view of science , information technology and data are a blessing , and a curse .
the reasons for them to be a blessing are many and obvious .
the reasons for it to be a curse are less obvious .
one of them is well articulated recently by two prominent biologists in an editorial casadevall and fang ( @xcite ) in _ infection and immunity _ ( of the american society for microbiology ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` although scientists have always comforted themselves with the thought that science is self - correcting , the immediacy and rapidity with which knowledge disseminates today means that incorrect information can have a profound impact before any corrective process can take place '' ( p. 893 ) . _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` a recent study analyzed the cause of retraction for 788 retracted papers and found that error and fraud were responsible for 545 ( 69% ) and 197 ( 25% ) cases , respectively , while the cause was unknown in 46 ( 5.8% ) cases ( 31 ) '' ( p. 893 ) .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the study referred is steen ( @xcite ) in the _ journal of medical ethics_. of the 788 retracted papers from pubmed from 2000 to 2010 , 69% are marked as `` errors '' on the retraction records .
statistical analyses are likely to be involved in these errors .
casadevall and fang go on to call for `` enhanced training in probability and statistics , '' among other remedies including `` reembracing philosophy . ''
more often than not , modern scientific findings rely on statistical analyses of high - dimensional data , and reproducibility is imperative for any scientific discovery .
scientific reproducibility therefore is a responsibility of statisticians . at a minimum
, reproducibility manifests itself in the stability of statistical results relative to `` reasonable '' perturbations to data and to the method or model used .
reproducibility of scientific conclusions is closely related to their reliability .
it is receiving much well - deserved attention lately in the scientific community ( e.g. , ioannidis , @xcite ; kraft et al .
, @xcite , casadevall and fang , @xcite ; nosek et al . , @xcite ) and in the media ( e.g. , naik , @xcite ; booth , @xcite ) . drawing a scientific conclusion involves multiple steps .
first , data are collected by one laboratory or one group , ideally with a clear hypothesis in the mind of the experimenter or scientist . in the age of information technology , however , more and more massive amounts of data are collected for fishing expeditions to `` discover '' scientific facts .
these expeditions involve running computer codes on data for data cleaning and analysis ( modeling and validation ) .
before these facts become `` knowledge , '' they have to be reproduced or replicated through new sets of data by the same group or preferably by other groups .
given a fixed set of data , donoho et al .
( @xcite ) discuss reproducible research in computational hormonic analysis with implications on computer - code or computing - environment reproducibility in computational sciences including statistics .
fonio et al .
( @xcite ) discuss replicability between laboratories as an important screening mechanism for discoveries .
reproducibility could have multitudes of meaning to different people .
one articulation on the meanings of reproducibility , replication , and repeatability can be found in stodden ( @xcite ) . in this paper
, we advocate for more involvement of statisticians in science and for an enhanced emphasis on stability within the statistical framework .
stability has been of a great concern in statistics .
for example , in the words of hampel et al .
( @xcite ) , `` robustness theories can be viewed as stability theories of statistical inference '' ( p. 8) . even in low - dimensional linear regression models
, collinearity is known to cause instability of ols or problem for individual parameter estimates so that significance testing for these estimates becomes unreliable . here
we demonstrate the importance of statistics for understanding our brain ; we describe our methodological work on estimation stability that helps interpret models reliably in neuroscience ; and we articulate how our solving neuroscience problems motivates theoretical work on stability and robust statistics in high - dimensional regression models .
in other words , we tell an interwinding story of scientific investigation and statistical developments . the rest of the paper is organized as follows . in section [ sec : fmri ] , we cover our `` intellectual adventure '' into neursocience , in collaboration with the gallant lab at uc berkeley , to understand human visual pathway via fmri brain signals invoked by natural stimuli ( images or movies ) ( cf .
kay et al . , @xcite , naselaris et al . ,
@xcite , kay and gallant , @xcite , naselaris et al . , @xcite ) .
in particular , we describe how our statistical encoding and decoding models are the backbones of `` mind - reading computers , '' as one of the 50 best inventions of 2011 by the time magazine ( nishimoto et al . , @xcite ) .
in order to find out `` how things really are , '' we argue that reliable interpretation needs stability .
we define stability relative to a data perturbation scheme . in section [ sec : stability ] , we briefly review the vast literature on different data perturbation schemes such as jacknife , subsampling , and bootstrap .
( we note that data perturbation in general means not only taking subsets of data units from a given data set , but also sampling from an underlying distribution or replicating the experiment for a new set of data . ) in section [ sec : es - cv ] , we review an estimation stability ( es ) measure taken from lim and yu ( @xcite ) for regression feature selection . combining es with cv as in lim and yu ( @xcite )
gives rise to a smoothing parameter selector es - cv for lasso ( or other regularization methods ) .
when we apply es - cv to the movie - fmri data , we obtain a 60% reduction of the model size or the number of features selected at a negligible loss of 1.3% in terms of prediction accuracy .
subsequently , the es - cv - lasso models are both sparse and more reliable hence better suited for interpretation due to their stability and simplicity .
the stability considerations in our neuroscience endeavors have prompted us to connect with the concept of stability from the robust statistics point of view . in el karoui
( @xcite ) , we obtain very interesting theoretical results in high - dimensional regression models with @xmath0 predictors and @xmath1 samples , shedding light on how sample variability in the design matrix meets heavier tail error distributions when @xmath3 is approximately a constant in @xmath4 or in the random matrix regime .
we describe these results in an important special case in section [ sec : robust ] .
in particular , we see that when @xmath5 and @xmath6 or so , the ordinary least squares ( ols ) estimator is better than the least absolute deviation ( lad ) estimator when the error distribution is double exponential .
we conclude in section [ sec : conclusions ] .
neuroscience holds the key to understanding how our mind works .
modern neuroscience is invigorated by massive and multi - modal forms of data enabled by advances in technology ( cf .
atkil , martone and van essen , @xcite ) .
building mathematical / statistical models on this data , computational neuroscience is at the frontier of neuroscience .
the gallant lab at uc berkeley is a leading neuroscience lab specializing in understanding the visual pathway , and is a long - term collaborator with the author s research group .
it pioneered the use of natural stimuli in experiments to invoke brain signals , in contrast to synthetic signals such as white noise and moving bars or checker boards as previously done .
simply put , the human visual pathway works as follows .
visual signals are recorded by retina and through the relay center lgn they are transmitted to primary visual cortex areas v1 , on to v2 and v4 , on the `` what '' pathway ( in contrast to the `` where '' pathway ) ( cf . goodale and milner , @xcite ) .
computational vision neuroscience aims at modeling two related tasks carried out by the brain ( cf .
dayan and abbott , @xcite ) through two kinds of models .
the first kind , the encoding model , predicts brain signals from visual stimuli , while the second kind , the decoding model recovers visual stimuli from brain signals .
often , decoding models are built upon encoding models and hence indirectly validate the former , but they are important in their own right . in the september issue of _ current biology _ , our joint paper with the gallant lab , nishimoto et al .
( @xcite ) invents a decoding ( or movie reconstruction ) algorithm to reconstruct movies from fmri brain signals .
this work has received intensive and extensive coverage by the media including the economist s oct .
29th 2011 issue ( `` reading the brain : mind - goggling '' ) and the national public radio in their program `` forum with michael krasny '' on tue , sept .
27 , 2011 at 9:30 am ( `` reconstructing the mind s eye '' ) .
as mentioned earlier , it was selected by the time magazine as one of the best 50 inventions of 2011 and dubbed as `` mind - reading computers '' on the cover page of the time s invention issue .
_ what is really behind the movie reconstruction algorithm ? _ _ can we learn something from it about how brain works ?
_ the movie reconstruction algorithm consists of statistical encoding and decoding models , both of which employ regularization .
the former are sparse models so they are concise enough to be viewed and are built via lasso@xmath7cv for each voxel separately . however , as is well - known lasso@xmath7cv results are not stable or reliable enough for scientific interpretation due to the @xmath8 regularization and the emphasis of cv on prediction performance .
so lasso@xmath7cv is not estimation stable .
the decoding model uses the estimated encoding model for each voxel and tiknohov regularization or ridge in covariance estimation to pull information across different voxels over v1 , v2 and v4 ( nishimoto et al . , @xcite ) .
then an empirical prior for clips of short videos is used from movie trailers and youtube to induce posterior weights on video clips in the empirical prior database .
tiknohov or ridge regularization concerns itself with the estimation of the covariance between voxels that is not of interest for interpretation .
the encoding phase is the focus here from now on.=-1 v1 is a primary visual cortex area and the best understood area in the visual cortex .
hubel and wiesel received a nobel prize in physiology or medicine in 1981 for two major scientific discoveries .
one is hubel and wiesel ( @xcite ) that uses cat physiology data to show , roughly speaking , that simple v1 neuron cells act like gabor filters or as angled edge detectors .
later , using solely image data , olshausen and field ( @xcite ) showed that image patches can be sparsely represented on gabor - like basis image patches .
the appearance of gabor filters in both places is likely not a coincidence , due to the fact that our brain has evolved to represent the natural world .
these gabor filters have different locations , frequencies and orientations .
previous work from the gallant lab has built a filter - bank of such gabor filters and successfully used them to design encoding models with single neuron signals in v1 invoked by static natural image stimuli ( kay et al .
, @xcite , naselaris et al . ,
@xcite ) . in nishimoto
( @xcite ) , we use fmri brain signals observed over 2700 voxels in different areas of the visual cortex .
fmri signals are indirect and non - invasive measures of neural activities in the brain and have good spatial coverage and temporal resolution in seconds .
each voxel is roughly a cube of 1 mm by 1 mm by 1 mm and contains hundreds of thousands of neurons .
leveraging the success of gabor - filter based models for single neuron brain signals , for a given image , a vector of features is extracted by 2-d wavelet filters .
this feature vector has been used to build encoding models for fmri brain signals in kay et al .
( @xcite ) and naselaris et al .
( @xcite ) .
invoked by clips of videos / movies , fmri signals from three subjects are collected with the same experimental set - up . to model fmri signals invoked by movies ,
a 3-dim motion - energy gabor filter bank has been built in nishimoto et al .
( @xcite ) to extract a feature vector of dimension of 26k .
linear models are then built on these features at the observed time point and lagged time points . at present sparse linear regression models are favorites of the gallant lab through lasso or @xmath9-l2boost .
these sparse models give similar prediction performance on validation data as neural nets and kernel machines on image - fmri data ; they correspond well to the neuroscience knowledge on v1 ; and they are easier to interpret than neural net and kernel machine models that include all features or variables . for each subject , following a rigorous protocol in the gallant lab , the movie data ( how many frames per second ? ) consists of three batches : training , test and validation .
the training data is used to fit a sparse encoding model via lasso or e - l2boost and the test data is used to select the smoothing parameter by cv .
these data are averages of two or three replicates .
that is , the same movie is played to one subject two or three times and the resulted fmri signals are called replicates .
then a completed encoding determined model is used to predict the fmri signals in the validation data ( with 10@xmath10 replicates ) and the prediction performance is measured by the correlation between the predicted fmri signals and observed fmri signals , for each voxel and for each subject .
good prediction performance is observed for such encoding models ( cf .
figure [ fig : scatter - hist ] ) .
prediction and movie reconstruction are good steps to validate the encoding model in order to understand the human visual pathway .
but the science lies in finding the features that might drive a voxel , or to use tukey s words , finding out `` how things really are . ''
it is often the case that the number of data units is easily different from what is in collected data .
there are some hard resource constraints such as that human subjects can not lie inside an fmri machine for too long and it also costs money to use the fmri machine . but whether the data collected is for 2 hours as in the data or 1 hours 50 min or 2 hours and 10 min is a judgement call by the experimenter given the constraints .
consequently , scientific conclusions , or in our case , candidates for driving features , should be stable relative to removing a small proportion of data units , which is one form of reasonable or appropriate data perturbation , or reproducible without a small proportion of the data units . with a smaller set of data ,
a more conservative scientific conclusion is often reached , which is deemed worthwhile for the sake of more reliable results .
statistics is not the only field that uses mathematics to describe phenomena in the natural world .
other such fields include numerical analysis , dynamical systems and pde and ode .
concepts of stability are central in all of them , implying the importance of stability in quantitative methods or models when applied to real world problems .
the necessity for a procedure to be robust to data perturbation is a very natural idea , easily explainable to a child .
data perturbation has had a long history in statistics , and it has at least three main forms : jacknife , sub - sampling and bootstrap .
huber ( @xcite ) writes in
john w. tukeys contribution to robust statistics : `` [ tukey ] preferred to rely on the actual batch of data at hand rather than on a hypothetical underlying population of which it might be a sample '' ( p. 1643 ) .
all three main forms of data perturbation rely on an `` actual batch of data '' even though their theoretical analyses do assume hypothetical underlying populations of which data is a sample .
they all have had long histories .
jacknife can be traced back at least to quenouille ( @xcite ) where jacknife was used to estimate the bias of an estimator .
tukey ( @xcite ) , an abstract in the annals of mathematical statistics , has been regarded as a key development because of his use of jacknife for variance estimation .
miller ( @xcite ) is an excellent early review on jacknife with extensions to regression and time series situations .
hinkley ( @xcite ) proposes weighted jacknife for unbalanced data for which wu ( @xcite ) provides a theoretical study .
knsch ( @xcite ) develops jacknife further for time series .
sub - sampling on the other hand was started three years earlier than jacknife by mahalanobis ( @xcite ) .
hartigan ( @xcite ) buids a framework for confidence interval estimation based on subsampling .
carlstein ( @xcite ) applies subsampling ( which he called subseries ) to the time series context .
politis and romano ( @xcite ) study subsampling for general weakly dependent processes .
cross - validation ( cv ) has a more recent start in allen ( @xcite ) and stone ( @xcite ) .
it gives an estimated prediction error that can be used to select a particular model in a class of models or along a path of regularized models .
it has been wildly popular for modern data problems , especially for high - dimensional data and machine learning methods .
hall ( @xcite ) and li ( @xcite ) are examples of theoretical analyses of cv .
( @xcite ) bootstrap is widely used and it can be viewed as simplified jacknife or subsampling .
examples of early theoretical studies of bootstrap are bickel and freedman ( @xcite ) and beran ( @xcite ) for the i.i.d .
case , and knsch ( @xcite ) for time series .
much more on these three data perturbation schemes can be found in books , for example , by efron and tibshirani ( @xcite ) , shao and tu ( @xcite ) and politis , romano and wolf ( @xcite ) .
if we look into the literature of probability theory , the mathematical foundation of statistics , we see 5 that a perturbation argument is central to limiting law results such as the central limit theorem ( clt ) .
the clt has been the bedrock for classical statistical theory .
one proof of the clt that is composed of two steps and is well exposited in terence tao s lecture notes available at his website ( tao , @xcite ) .
given a normalized sum of i.i.d .
random variables , the first step proves the universality of a limiting law through a perturbation argument or the lindebergs swapping trick . that is , one proves that a perturbation in the ( normalized ) sum by a random variable with matching first and second moments does not change the ( normalized ) sum distribution .
the second step finds the limit law by way of solving an ode.=-1 recent generalizations to obtain other universal limiting distributions can be found in chatterjee ( @xcite ) for wigner law under non - gaussian assumptions and in suidan ( @xcite ) for last passage percolation .
it is not hard to see that the cornerstone of theoretical high - dimensional statistics , concentration results , also assumes stability - type conditions . in learning theory ,
stability is closely related to good generalization performance ( devroye and wagner , @xcite , kearns and ron , @xcite , bousquet and elisseeff , @xcite , kutin and niyogi , @xcite , mukherjee et al .
, @xcite , shalev - shwartz et al . , @xcite ) . to further our discussion on stability , we would like to explain what we mean by statistical stability .
we say statistical stability holds if statistical conclusions are robust or stable to appropriate perturbations to data .
that is , statistical stability is well defined relative to a particular aim and a particular perturbation to data ( or model ) .
for example , aim could be estimation , prediction or limiting law . it is not difficult to have statisticians to agree on what are appropriate data perturbations when data units are i.i.d . or exchangeable in general , in which case subsampling or bootstrap are appropriate
. when data units are dependent , transformations of the original data are necessary to arrive at modified data that are close to i.i.d . or exchangeable , such as in parametric bootstrap in linear models or block - bootstrap in time series .
when subsampling is carried out , the reduced sample size in the subsample does have an effect on the detectable difference , say between treatment and control .
if the difference size is large , this reduction on sample size would be negligible .
when the difference size is small , we might not detect the difference with a reduced sample size , leading to a more conservative scientific conclusion .
because of the utmost importance of reproducibility for science , i believe that this conservatism is acceptable and may even be desirable in the current scientific environment of over - claims .
for the fmri problem , let us recall that for each voxel , lasso or e - l2boost is used to fit the mean function in the encoding model with cv to choose the smoothing parameter .
different model selection criteria have been known to be unstable .
breiman ( @xcite ) compares predictive stability among forward selection , two versions of garotte and ridge and their stability increases in that order .
he goes on to propose averaging unstable estimators over different perturbed data sets in order to stabilize unstable estimators .
such estimators are prediction driven , however , and they are not sparse and thereby not suitable for interpretation . in place of bootstrap for prediction error estimation as in efron ( @xcite ) , zhang ( @xcite ) uses multi - fold cross - validation while shao ( @xcite ) uses m out of n bootstrap samples with @xmath11
. they then select models with this estimated prediction error , and provide theoretical results for low dimensional or fixed p linear models .
heuristically , the m out of n bootstrap in shao ( @xcite ) is needed because the model selection procedure is a discrete ( or set ) valued estimator for the true model predictor set and hence non - smooth ( cf .
bickel , gtze , and van zwet , @xcite ) .
the lasso ( tibshirani , @xcite ) is a modern model selection method for linear regression and very popular in high - dimensions : @xmath12 where @xmath13 is the response vector and @xmath14 is the design matrix .
that is , there are @xmath1 data units and @xmath0 predictors . for each @xmath15 , there is a unique @xmath8 norm for its solution that we can use to index the solution as @xmath16 where @xmath17 cross - validation ( cv ) is used most of the time to select @xmath15 or @xmath18 , but lasso@xmath7cv is unstable relative to bootstrap or subsampling perturbations when predictors are correlated ( cf .
meinshausen and bhlmann , @xcite , bach , @xcite ) . using bootstrap in a different manner than shao ( @xcite ) , bach ( @xcite ) proposes bolasso to improve lasso s model selection consistency property by taking the smallest intersecting model of selected models over different bootstrap samples .
for particular smoothing parameter sequences , the bolasso selector is shown by bach ( @xcite ) to be model selection consistent for the low dimensional case without the irrepresentable condition needed for lasso ( cf .
meinshausen and bhlmann , @xcite , zhao and yu , @xcite ; zou , @xcite ; wainwright , @xcite ) .
meinshausen and bhlmann ( @xcite ) also weaken the irrepresentable condition for model selection consistency of a stability selection criterion built on top of lasso .
they bring perturbations to a lasso path through a random scalar vector in the lasso @xmath8 penalty , resulting in many random lasso paths .
a threshold parameter is needed to distinguish important features based on these random paths .
they do not consider the problem of selecting one smoothing parameter value for lasso as in lim and yu ( @xcite ) .
we would like to seek a specific model along the lasso path to interpret and hence selects a specific @xmath15 or @xmath18 .
it is well known that cv does not provide a good interpretable model because lasso@xmath7cv is unstable .
lim and yu ( @xcite ) propose a stability - based criterion that is termed estimation stability ( es ) .
they use the cross - validation data perturbation scheme .
that is , @xmath1 data units are randomly partitioned into @xmath19 blocks of pseudo data sets of size @xmath20 or subsamples where @xmath21 is the floor function or the largest integer that is smaller than or equal to @xmath22 . ]
given a smoothing parameter @xmath15 , a lasso estimate @xmath23 is obtained for the @xmath24th block @xmath25 . since the @xmath8 norm is a meaningful quantity to line up the @xmath19 different estimates , lim and yu ( @xcite ) to line up the different solutions , but not a good idea to use the ratio of @xmath15 and its maximum value for each pseudo data set .
] use it , denoted as @xmath18 below , to line up these estimates to form an estimate @xmath26 for the mean regression function and an approximate delete-@xmath27 jacknife estimator for the variance of @xmath26 : @xmath28 the last expression is only an approximate delete-@xmath27 jacknife variance estimator unless @xmath29 when all the subsamples of size @xmath30 are used . define the ( estimation )
statistical stability measure as @xmath31 where @xmath32 . for nonlinear regression functions ,
es can still be applied if we take an average of the estimated regression functions .
note that es aims at estimation stability , while cv aims at prediction stability .
in fact , es is the reciprocal of a test statistic for testing @xmath33 since @xmath32 is a test statistic for @xmath34 , @xmath35 is also a test statistic .
@xmath36 is a scaled version of the reciprocal @xmath37 . to combat the high noise situation where es
would not have a well - defined minimum , lim and yu ( @xcite ) combine es with cv to propose the _ es - cv selection criterion _ for smoothing parameter @xmath18 : _ choose the largest @xmath18 that minimizes es @xmath38 and is smaller or equal to the cv selection .
_ es - cv is applicable to smoothing parameter selection in lasso , and other regularization methods such as tikhonov or ridge regularization ( see , for example , tikhonov , @xcite , markovich , @xcite , hoerl , @xcite , hoerl and kennard , @xcite ) .
es - cv is well suited for parallel computation as cv and incurs only a negligible computation overhead because @xmath26 are already computed for cv .
moreover , simulation studies in lim and yu ( @xcite ) indicate that , when compared with lasso@xmath7cv , es - cv applied to lasso gains dramatically in terms of false discovery rate while it loses only somewhat in terms of true discovery rate .
the features or predictors in the movie - fmri problem are 3-d gabor wavelet filters , and each of them is characterized by a ( discretized ) spatial location on the image , a ( discretized ) frequency of the filter , a ( discretized ) orientation of the filter , and 4 ( discrete ) time - lags on the corresponding image that the 2-d filter is acting on . for the results comparing cv and es - cv in figure [ fig : feature - locations ] , we have a sample size @xmath39 and use a reduced set of @xmath40 features or predictors , corresponding to a coarser set of filter frequencies than what is used in nishimoto et al .
( @xcite ) with @xmath41 predictors .
[ fig : scatter - hist ] we apply both cv and es - cv to select the smoothing parameters in lasso ( or e - l2boost ) . for three voxels ( and a particular subject ) , for the simplicity of display , we show the locations of the selected features ( regardless of their frequencies , orientations and time - lags ) in figure [ fig : feature - locations ] . for these three voxels ,
es - cv maintains almost the same prediction correlation performances as cv ( 0.70 vs. 0.72 ) while es - cv selects many fewer and more concentrated locations than cv .
figure [ fig : scatter - hist ] shows the comparison results across 2088 voxels in the visual cortex that are selected for their high snrs .
it is composed of four sub - plots .
the upper two plots compare prediction correlation performance of the models built via lasso with cv and es - cv on validation data . for each model fitted on training data and each voxel , predicted responses over the validation data are calculated .
its correlation with the observed response vector is the `` prediction correlation '' displayed in figure [ fig : scatter - hist ] .
the lower two plots compare the sparsity properties of the models or model size . because of the definition of es - cv
, it is expected that the es - cv model are always smaller than or equal to the cv model .
the sparsity advantage of es - cv is apparent with a huge overall reduction of 60% on the number of selected features and a minimum loss of overall prediction accuracy by only 1.3% .
the average size of the es - cv models is 24.3 predictors , while that for the cv models is 58.8 predictors ; the average prediction correlation performance of the es - cv models is 0.499 , while that for the cv models is 0.506 .
robust statistics also deals with stability , relative to model perturbation . in the preface of his book `` robust statistics , '' huber ( @xcite ) states : `` primarily , we are concerned with _ distributional robustness _ : the shape of the true underlying distribution deviates slightly from the assumed model . ''
hampel , rousseeuw , ronchetti and stahel ( @xcite ) write : `` overall , and in analogy with , for example , the stability aspects of differential equations or of numerical computations , robustness theories can be viewed as stability theories of statistical inference '' ( p. 8) .
tukey ( @xcite ) has generally been regarded as the first paper on robust statistics .
fundamental contributions were made by huber ( @xcite ) on m - estimation of location parameters , hampel ( @xcite ) on `` break - down '' point and influence curve .
further important contributions can be found , for example , in andrews et al .
( @xcite ) and bickel ( @xcite ) on one - step huber estimator , and in portnoy ( @xcite ) for m - estimation in the dependent case.=-1 for most statisticians , robust statistics in linear regression is associated with studying estimation problems when the errors have heavier tail distributions than the gaussian distribution . in the fmri problem , we fit mean functions with an l2 loss .
what if the `` errors '' have heavier tails than gaussian tails ? for the l1 loss is commonly used in robust statistics to deal with heavier tail errors in regression , we may wonder whether the l1 loss would add more stability to the fmri problem .
in fact , for high - dimensional data such as in our fmri problem , removing some data units could severely change the outcomes of our model because of feature dependence .
this phenomenon is also seen in simulated data from linear models with gaussian errors in high - dimensions .
_ how does sample to sample variability interact with heavy tail errors in high - dimensions ? _ in our recent work el karoui et al .
( @xcite ) , we seek insights into this question through analytical work .
we are able to see interactions between sample variability and double - exponential tail errors in a high - dimensional linear regression model .
that is , let us assume the following linear regression model @xmath42 where @xmath43 an m - estimator with respect to loss function @xmath44 is given as @xmath45 we consider the random - matrix high - dimensional regime : @xmath46 due to rotation invariance , wlog , we can assume @xmath47 and @xmath48 .
we cite below a result from el karoui et al .
( @xcite ) for the important special case of @xmath47 : [ res1 ] under the aforementioned assumptions , let @xmath49 , then @xmath50 is distributed as @xmath51 where @xmath52 , and @xmath53 as @xmath54 and @xmath55 .
denote @xmath56 where @xmath57 and independent of @xmath9 , and let @xmath58 .\ ] ] then @xmath59 satisfies the following system of equations together with some nonnegative @xmath60 : @xmath61^\prime\bigr\ } & = & 1 - \kappa , \\
e \bigl\ { \bigl [ \hat{z}_\varepsilon- \operatorname{prox}_c ( \hat{z}_\varepsilon ) \bigr]^2\bigr\ } & = & \kappa r_\rho^2 ( \kappa ) .\end{aligned}\ ] ] in our limiting result , the norm of an m - estimator stabilizes .
it is most interesting to mention that in the proof a `` leave - one - out '' trick is used both row - wise and column - wise such that one by one rows are deleted and similarly columns are deleted . the estimators with deletions
are then compared to the estimator with no deletion .
this is in effect a perturbation argument and reminiscent of the `` swapping trick '' for proving the clt as discussed before .
our analytical derivations involve prox functions , which are reminiscent of the second step in proving normality in the clt .
this is because a prox function is a form of derivative , and not dissimilar to the derivative appearing in the ode derivation of the analytical form of the limiting distribution ( e.g normal distribution ) in the clt . in the case of i.i.d .
double - exponential errors , el karoui et al .
( @xcite ) numerically solve the two equations in result [ res1 ] to show that when @xmath62 , @xmath63 loss fitting ( ols ) is better than @xmath8 loss fitting ( lad ) in terms of mse or variance .
they also show that the numerical results match very well with simulation or monte carlo results . at a high level
, we may view that @xmath64 holds the key to this interesting phenomenon .
being a weighted convolution of @xmath65 and @xmath9 , it embeds the interaction between sample variability ( expressed in @xmath65 ) and error variability ( expressed in @xmath9 ) and this interaction is captured in the optimal loss function ( cf .
el karoui et al . , @xcite ) .
in other words , @xmath64 acts more like double exponential when the influence of standard normal @xmath65 in @xmath64 is not dominant ( or when @xmath66 or so as we discover when we solve the equations ) and in this case , the optimal loss function is closer to lad loss . in cases when @xmath67
, it acts more like gaussian noise , leading to the better performance of ols ( because the optimal loss is closer to ls ) .
moreover , for double exponential errors , the m - estimator lad is an mle and we are in a high - dimensional situation .
it is well - known that mle does not work in high - dimensions .
remedies have been found through penalized mle where a bias is introduced to reduce variance and consequently reduce the mse .
in contrast , when @xmath68 , the better estimator ols is also unbiased , but has a smaller variance nevertheless .
the variance reduction is achieved through a better loss function ls than the lad and because of a concentration of quadratic forms of the design matrix .
this concentration does not hold for fixed orthogonal designs , however . a follow - up work ( bean et al . , @xcite )
addresses the question of obtaining the optimal loss function .
it is current research regarding the performance of estimators from penalized ols and penalized lad when the error distribution is double - exponential .
preliminary results indicate that similar phenomena occur in non - sparse cases .
furthermore , simulations with design matrix from an fmri experiment and double - exponential error show the same phenomenon , that is , when @xmath69 or so , ols is better than lad
. this provides some insurance for using l2 loss function in the fmri project .
it is worth noting that el karoui et al .
( @xcite ) contains results for more general settings .
in this paper , we cover three problems facing statisticians at the 21st century : figuring out how vision works with fmri data , developing a smoothing parameter selection method for lasso , and connecting perturbation in the case of high - dimensional data with classical robust statistics through analytical work .
these three problems are tied together by stability .
stability is well defined if we describe the data perturbation scheme for which stability is desirable , and such schemes include bootstrap , subsampling , and cross - validation .
moreover , we briefly review results in the probability literature to explain that stability is driving limiting results such as the central limit theorem , which is a foundation for classical asymptotic statistics . using these three problems as backdrop , we make four points .
firstly , statistical stability considerations can effectively aid the pursuit for interpretable and reliable scientific models , especially in high - dimensions .
stability in a broad sense includes replication , repeatability , and different data perturbation schemes .
secondly , stability is a general principle on which to build statistical methods for different purposes .
thirdly , the meaning of stability needs articulation in high - dimensions because it could be brought about by sample variability and/or heavy tails in the errors of a linear regression model .
last but not least , emphasis should be placed on the stability aspects of statistical inference and conclusions , in the referee process of scientific and applied statistics papers and in our current statistics curriculum .
statistical stability in the age of massive data is an important area for research and action because high - dimensions provide ample opportunities for instability to reveal itself to challenge reproducibility of scientific findings .
as we began this article with words of tukey , it seems fitting to end also with his words : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ `` what of the future ?
the future of data analysis can involve great progress , the overcoming of real difficulties , and the provision of a great service to all fields of science and technology .
will it ? that remains to us , to our willingness to take up the rocky road of real problems in preferences to the smooth road of unreal assumptions , arbitrary criteria , and abstract results without real attachments . who is for the challenge ? '' tukey ( p. 64 , @xcite ) .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
this paper is based on the 2012 tukey lecture of the bernoulli society delivered by the author at the 8th world congress of probability and statistics in istanbul on july 9 , 2012 . for their scientific influence and friendship ,
the author is indebted to her teachers / mentors / colleagues , the late professor lucien le cam , professor terry speed , the late professor leo breiman , professor peter bickel , and professor peter bhlmann .
this paper is invited for the special issue of _ bernoulli _ commemorating the 300th anniversary of the publication of jakob bernoulli s ars conjectandi in 1712 .
the author would like to thank yuval benjamini for his help on generating the results in the figures .
she would also like to thank two referees for their detailed and insightful comments , and yoav benjamini and victoria stodden for helpful discussions .
partial supports are gratefully acknowledged by nsf grants ses-0835531 ( cdi ) and dms-11 - 07000 , aro grant w911nf-11 - 1 - 0114 , and the nsf science and technology center on science of information through grant ccf-0939370 . | reproducibility is imperative for any scientific discovery .
more often than not , modern scientific findings rely on statistical analysis of high - dimensional data . at a minimum
, reproducibility manifests itself in stability of statistical results relative to `` reasonable '' perturbations to data and to the model used .
jacknife , bootstrap , and cross - validation are based on perturbations to data , while robust statistics methods deal with perturbations to models . in this article , a case
is made for the importance of stability in statistics .
firstly , we motivate the necessity of stability for interpretable and reliable encoding models from brain fmri signals .
secondly , we find strong evidence in the literature to demonstrate the central role of stability in statistical inference , such as sensitivity analysis and effect detection .
thirdly , a smoothing parameter selector based on estimation stability ( es ) , es - cv , is proposed for lasso , in order to bring stability to bear on cross - validation ( cv ) .
es - cv is then utilized in the encoding models to reduce the number of predictors by 60% with almost no loss ( 1.3% ) of prediction performance across over 2,000 voxels .
last , a novel `` stability '' argument is seen to drive new results that shed light on the intriguing interactions between sample to sample variability and heavier tail error distribution ( e.g. , double - exponential ) in high - dimensional regression models with @xmath0 predictors and @xmath1 independent samples . in particular , when @xmath2 and the error distribution is double - exponential , the ordinary least squares ( ols ) is a better estimator than the least absolute deviation ( lad ) estimator . |
one of the simplest implementations of such a mechanism is the littlest higgs ( lh ) model @xcite based on @xmath5\otimes[su(2)_2\otimes u(1)_2 ] \ .\ ] ] at @xmath6 tev the initial @xmath7 global symmetry spontaneously breaks down to an @xmath8 subgroup in the direction @xmath9 where @xmath10 is the identity matrix .
after symmetry breaking at the energy scale @xmath4 , the dynamics near @xmath11 is described by the non - linear sigma field @xmath12 , where @xmath13 are the pseudo - nambu - goldstone bosons ( pngb ) associated with the 14 generators @xmath14 of the broken symmetry .
an @xmath15 ^ 2 $ ] subgroup of the @xmath7 is weakly gauged .
gauging each of the two @xmath16 leaves a different @xmath17 subgroup unbroken , i.e. unless both @xmath16 are broken there will be a preserved @xmath17 symmetry and the higgs field will be an exact massless nambu - goldstone field .
thus any loop contribution to higgs mass must involve couplings from both copies of @xmath18 . at one loop level
the leading contribution is only logarithmically divergent under this requirement .
this mechanism that protects the higgs mass from quadratic divergneces is often referred as `` collective symmetry breaking '' .
@xmath11 breaks the full gauge group to the diagonal sm electroweak group @xmath16 at energy scale @xmath4 .
four pngb s give tev scale masses to @xmath19 and @xmath20 . @xmath21 and @xmath20 mix and form mass eigenstates @xmath1 and @xmath22 in analog to the sm photon and @xmath23 boson .
other pngb fields group into a doublet identified as the sm higgs and a weak triplet @xmath24 @xmath25 where @xmath26 and @xmath27 in the higgs doublet are eaten by sm weak bosons.@xcite all the new particles are massive .
the new tev scale particles @xmath28 couple to the higgs field and cancel the quadratic radiative corrections to the higgs mass arising from their sm counterparts . however , the tight constraints from precision electroweak data disfavor the lh model at a natural symmetry breaking scale @xmath291 tev .
phenomenological constraints on lh parameters@xcite push the lower boundary of new physics up to about 10 tev , but the naturalness principle sets all dimensionless couplings to @xmath30 and requires the energy scale to be around 1 tev .
thus the lh model needs an energy scale higher than the natural value to stay consistent with electroweak results .
this tension is often referred to as the ` little hierarchy ' problem . to address this issue
, cheng and low proposed that an additional discrete@xcite @xmath0-parity can be imposed to relax@xcite the confrontation between theory and experimental constraints .
similar to the matter parity in supersymmetry , @xmath0-parity is introduced as a global discrete parameter .
it exchanges @xmath31 $ ] and @xmath32 $ ] .
@xmath33 transforms under @xmath0-parity as @xmath34 with @xmath35 .
all sm particles and the lh heavy top quark @xmath36 are assigned @xmath0-even in the little higgs model with @xmath0-parity ( lht ) .
all other heavy particles are assigned @xmath0-odd . in the fermion sector , each sm fermion is extended into a pair of su(2 ) doublets @xmath37 and @xmath38 that transform under @xmath39 and @xmath40 .
@xmath0-parity interchanges @xmath37 and @xmath38 .
their @xmath0-even combination is associated with the sm fermion , while the @xmath0-odd combination is the heavy partner to the sm particle .
interaction terms @xmath41 give mass to @xmath0-odd fermions @xmath42 where @xmath43 gev is the higgs vacuum expectation value in the sm and @xmath44 is a free parameter . for illustration
, we take @xmath45 throughout this paper . in the interaction term @xmath46
is the su(2 ) fermion doublet embedded into the su(5 ) multiplet : @xmath47 , @xmath48 and @xmath49 .
details are given in ref.@xcite .
there is a special treatment in the top sector besides the heavy @xmath50-odd weak _ doublet _ @xmath51 of the third generation . in lh the large top quark loop correction
demands an additional vector - like weak _ singlet _ @xmath36 quark to stabilize the higgs mass .
it is assigned even @xmath0-parity in lht and its @xmath0-odd partner is introduced as another top - like heavy quark @xmath52 . in the bosonic sector , the doublet ( triplet ) higgs @xmath53 ( @xmath24 ) is even ( odd ) under @xmath54-parity .
the @xmath54-even combinations of the gauge fields are the sm @xmath55 gauge bosons @xmath56 and @xmath57 hypercharge gauge boson @xmath58 ; the @xmath54-odd combinations are @xmath54-parity partners(@xmath1 , @xmath59 , @xmath22 ) of the sm gauge bosons .
the masses are given as @xmath60,~ m_{z_h } \simeq m_{w_h}=gf \left[1-\frac{v_{sm}^2}{8f^2}+\cdots\right ] \ .\end{aligned}\ ] ] the lightest @xmath54-odd particle is @xmath1 , which could be a dark matter candidate if @xmath54-parity was strictly conserved .
the new @xmath0-odd bosons have masses around a few hundred gev .
the new fermions have higher masses near 1 tev .
a typical mass spectrum of heavy particles in lht is shown in table [ tab : mass_spectrum ] .
.characteristic masses of the heavy partners of the sm particles .
here we take the scale @xmath611.5 tev , @xmath45 , and the top quark and higgs boson masses to be @xmath62 and @xmath63 gev , respectively .
the dependences of particle masses on @xmath4 are plotted in fig.[fig : mass ] . [ cols="^,^,^",options="header " , ] fig.[fig : totw ] gives the total decay widths of the massive quarks and gauge bosons .
besides the dominant two - body decay modes , many three - body decay channels may not be negligible for the reason that the interactions of the longitudinal polarization of the gauge bosons are enhanced and the bosonic couplings are large .
due to the large mass gap between the sm particles and the heavy partners , and among the heavy particles themselves , three - body decays are usually accessible ( for instance @xmath64 at tev mass ) in strong contrast to the restrictive three - body decay channels @xmath65@xcite and @xmath66@xcite of the top quark in sm .
as the energy scale @xmath4 increases , the phase space of many three - body channels opens up , and their branching fractions become experimentally relevant at higher @xmath4 values .
the @xmath4 dependence of significant three - body decays is plotted in fig.[fig : bdec ] and fig.[fig : fdec ] .
the three - body channels can provide a good testing ground for the detailed structure of the lht interactions .
the particle table , feynman rules , tables of parameters , and event simulations of the lht model have been coded in a public package calchep lht@xcite available at http://hep.pa.msu.edu/lht for the phenomenology of the lht model .
we make use of this convenient tool to calculate various @xmath0-parity preserving multi - body decay channels . in our analysis
we fix the parameter @xmath45 , the sm higgs mass @xmath67200 gev , and take 1.5 @xmath68 2.5 tev as a compromise between naturalness and the electroweak precision constraints.@xcite we do not include photons among the daughter particles because of their suppressed coupling .
= 200 gev for heavy higgs boson widths . in the figure
the symbol @xmath69 denotes @xmath0-odd partners of the sm quarks and leptons . ]
the lightest @xmath0-odd particle is the heavy photon @xmath1 .
exact @xmath0-parity conservation would require that normally @xmath1 be a final decay product from any @xmath0-odd heavy particle . on the other hand , when @xmath0-parity is violated by the anomaly interaction@xcite the @xmath1 will decay rapidly into @xmath2 and @xmath3 .
the final products of the @xmath1 decay can be detected , and the event kinematics can thereby be fully reconstructed .
the identification of three - particle decay modes becomes feasible .
branching fractions of the leading decay channels are plotted versus @xmath4 in fig.[fig : bdec ] .
we retain the channels with a fraction above 0.1% .
it is interesting to see what new channels are available : * we start with the @xmath0-odd neutral @xmath22 .
the two - body decay mode @xmath70 dominates . a fermionic final state is not kinematically allowed
however , the three - boson phase space is open and there is a substantial branching fraction at the level of 10% for @xmath71 .
* @xmath72 is the dominating two - body mode in @xmath73 decay .
similar to @xmath22 decay , @xmath73 is less massive than heavy fermions and any fermionic final state is kinematically disallowed .
the heavy mass @xmath74 allows both @xmath75 and @xmath76 decays at the level of a few percent . * the decay of the singly charged @xmath77 is mainly dominated by the two - body mode @xmath78 .
three - body channels @xmath75 and @xmath79 also give significant contribution to the total width . *
the neutral component @xmath80 of the triplet scalar boson is a complex field .
it is decomposed into the real part ( a scalar @xmath81 ) and the imaginary part ( a pseudoscalar @xmath82 ) of approximately equal masses .
the different spatial parities imply different decay modes .
@xmath83 note the role swap @xmath84 when @xmath85 . both @xmath81 and @xmath82 decay to @xmath86 as well .
* the doubly charged scalar boson @xmath87 can not decay into @xmath88 because of the common @xmath89 mass . at low higgs mass @xmath90120 gev
there are no two - body decays of @xmath87 ; however @xmath91 emerges at higher higgs mass when @xmath80 becomes much more massive than heavy weak gauge bosons .
the virtual process @xmath92 gives the overall leading three - body decay @xmath93 .
the four - body decay channels ( @xmath94 or @xmath95 ) are smaller but can still be of comparable size to the two / three body - decay modes because of the limited width of leading decay modes .
there are many @xmath0-odd fermions @xmath96 of different flavors @xmath97 with tev scale masses .
@xmath50-parity demands at least one heavy boson in the final state . the general pattern of decay channels according to descending branching fractions are + @xmath98 or , + @xmath99 + where @xmath100 are the sm doublet .
the relevant feynman diagrams are shown in fig .
5 . -odd fermion .
@xmath101 refer to different flavors of a sm doublet . ]
[ fig : feyn ] the @xmath36 quark of even @xmath0-parity decays readily through @xmath102 mixing into @xmath103 , @xmath104 , @xmath105 as dominating two - body modes .
the three - body mode @xmath106 occurs as a rare process , in analog to the rare decay modes of the fourth - generation quark as studied in ref.@xcite .
in addition , being @xmath0-even @xmath36 has a rare mode of decaying into two heavy photons .
note that some three - body channels that can cascade from the primary on - shell two - body decay modes are not shown in our plots , mainly because their rates depend very much on the mass cuts in separating out the resonance components ; for example , @xmath107 where @xmath108 can be the resonance of @xmath109 . these channels with intermediate resonance would nevertheless be very important in determining the resonance masses .
little higgs phenomenology at the lhc has been investigated recently in a number of studies@xcite but not for the situation where @xmath1 decays .
the key feature is that the new heavy quarks can also be produced in proton - proton collisions , either from gluon fusion or quark interactions .
most @xmath0-odd quarks are pair produced with the exception that the @xmath0-even @xmath36 particle can be produced along with a sm quark . the produced heavy quarks decay quickly into less massive sm particles and @xmath0-odd bosons that subsequently decay into sm counterparts and heavy photons .
the dominant two - body decay channels @xmath110 transform each heavy quark into a sm quark that may form a jet and one @xmath0-odd gauge boson . @xmath59 and @xmath22 decay into sm gauge / higgs bosons and @xmath1 .
@xmath1 decays through @xmath0-violating wzw interation into @xmath2 , @xmath3 , resulting in an overall @xmath50-even final state . as shown in fig.[fig : feyn ] , the leading three - body decay processes of @xmath0-odd quarks will either add a @xmath111 or @xmath23 boson to the daughter particles , while @xmath112 gives an additional @xmath23 boson in the final state . in the bosonic sector , the heavy gauge bosons have significant decay rates to three - body final states .
the contributions of the three - body decays visibly depend on the energy scale @xmath4 .
as @xmath4 increases to higher @xmath4 values , the three - body phase space opens up faster than the two - body phase space , and the three - body branching fractions steadily increase as the mass gap widens between the heavy particles and sm particles .
the little higgs model with @xmath0-parity ( lht ) is an interesting extension of the little higgs framework .
it alleviates the tension of the little hierarchy " problem and it is also a phenomenologically rich model , giving rise to testable new physics at the tev scale . in the lht model
one can expect that the lhc will produce a large amount of heavy quarks beyond the sm via the strong interaction , and also substantial numbers of new heavy leptons and new heavy gauge bosons by drell - yan - like processes .
their decay patterns can go beyond the usual dominant two - body modes and include contributions from various measurable three - body modes .
since @xmath0-parity is broken by anomaly , the lightest @xmath0-odd particle @xmath1 will decay into detectable @xmath2 or @xmath3 .
as @xmath1 appears as a decay product of all lht processes , the @xmath50-parity - violating decays allow reconstruction of the full event configurations and thereby comprehensive physics tests of the little higgs model at the lhc .
we have studied the multi - body decays of the heavy particles in the lht model that can be produced at the lhc .
detailed analyses of these multi - body channels may be useful in revealing the new symmetry and its interactions at the tev scale .
we thank christopher t. hill and richard j. hill for very helpful discussions .
this research was supported in part by the u.s.department of energy under grant nos .
de - fg02 - 95er40896 and de - fg02 - 84er40173 , and in part by the wisconsin alumni research foundation .
the decay amplitude @xmath113 can be derived from eq .
( [ eq : pswzw ] ) @xmath114 where the levi civita symbol is contracted with vectors .
the momentum term @xmath115 comes from two ways of contracting the @xmath23 field .
it antisymmetrizes the momentum part and the levi civita antisymmetrizes the polarization part .
the combined product is overall symmetric as expected for the boson decay .
we choose @xmath116 in the rest frame of @xmath20 along the @xmath117 direction , and @xmath118 .
notice that transverse - transverse ( tt ) modes vanish , as well as the longitudinal - longitudinal ( ll ) mode .
the only surviving modes are lt or tl .
the relevant vectors in the lt mode are @xmath119 the levi civita symbol becomes the determinant of the above arrays , @xmath120 .
@xmath121 the last factor two counts both lt and tl modes .
thus we obtain @xmath122 the factor @xmath123 in the above decay width comes from the combinatorics of the two identical @xmath23 bosons .
after some algebra , we derive the final expressions eqs .
( [ eq : bzz ] , [ eq : bww ] ) of @xmath124 , and the similar one @xmath125 . the threshold dependence agrees with that in ref .
note that the overall factor of 2 difference between @xmath126 and @xmath2 comes from identical particle effect .
xx r. barbieri and a. strumia , arxiv : hep - ph/0007265 .
n. arkani - hamed , a. g. cohen and h. georgi , phys .
b * 513 * , 232 ( 2001 ) [ arxiv : hep - ph/0105239 ] . for reviews ,
see , for example , m. schmaltz and d. tucker - smith , arxiv : hep - ph/0502182 ; n. arkani - hamed , a. g. cohen , e. katz and a. e. nelson , jhep * 0207 * , 034 ( 2002 ) [ arxiv : hep - ph/0206021 ] .
t. han , h. e. logan , b. mcelrath and l. t. wang , phys .
d * 67 * , 095004 ( 2003 ) [ arxiv : hep - ph/0301040 ] . c. csaki , j. hubisz , g. d. kribs , p. meade and j. terning , phys .
d * 67 * , 115002 ( 2003 ) [ arxiv : hep - ph/0211124 ] .
j. l. hewett , f. j. petriello and t. g. rizzo , jhep * 0310 * , 062 ( 2003 ) [ arxiv : hep - ph/0211218 ] . c. csaki , j. hubisz , g. d. kribs , p. meade and j. terning , phys .
d * 68 * , 035009 ( 2003 ) [ arxiv : hep - ph/0303236 ] .
m. c. chen and s. dawson , phys .
d * 70 * , 015003 ( 2004 ) [ arxiv : hep - ph/0311032 ] .
w. kilian and j. reuter , phys .
d * 70 * , 015004 ( 2004 ) [ arxiv : hep - ph/0311095 ] .
z. han and w. skiba , phys .
d * 71 * , 075009 ( 2005 ) [ arxiv : hep - ph/0412166 ] .
h. c. cheng and i. low , jhep * 0309 * , 051 ( 2003 ) [ arxiv : hep - ph/0308199 ] .
i. low , jhep * 0410 * , 067 ( 2004 ) [ arxiv : hep - ph/0409025 ] . j. hubisz and p. meade , phys .
d * 71 * , 035016 ( 2005 ) [ arxiv : hep - ph/0411264 ] .
j. wess and b. zumino , phys .
b * 37 * , 95 ( 1971 ) .
e. witten , nucl .
b * 223 * , 422 ( 1983 ) .
a. belyaev , c. r. chen , k. tobe and c. p. yuan , effects of phys .
d * 74 * , 115020 ( 2006 ) [ arxiv : hep - ph/0609179 ] .
v. d. barger , w. y. keung and t. g. rizzo , phys .
d * 40 * , 2274 ( 1989 ) .
g. mahlon and s. j. parke , phys .
b * 347 * , 394 ( 1995 ) [ arxiv : hep - ph/9412250 ] .
e. jenkins , phys .
d * 56 * , 458 ( 1997 ) [ arxiv : hep - ph/9612211 ] .
v. d. barger and w. y. keung , phys .
b * 202 * , 393 ( 1988 ) .
q. h. cao and c. r. chen , arxiv:0707.0877 [ hep - ph ] .
k. kong and s. c. park , arxiv : hep - ph/0703057 .
d. choudhury and d. k. ghosh , arxiv : hep - ph/0612299 . s. matsumoto , m. m. nojiri and d. nomura , phys . rev .
d * 75 * , 055006 ( 2007 ) [ arxiv : hep - ph/0612249 ] .
m. carena , j. hubisz , m. perelstein and p. verdier , arxiv : hep - ph/0610156 .
f. ledroit , arxiv : hep - ex/0610005 .
a. freitas and d. wyler , jhep * 0611 * , 061 ( 2006 ) [ arxiv : hep - ph/0609103 ] .
k. cheung , c. s. kim , k. y. lee and j. song , phys . rev .
d * 74 * , 115013 ( 2006 ) [ arxiv : hep - ph/0608259 ] . c. s. chen , k. cheung and t. c. yuan , phys .
b * 644 * , 158 ( 2007 ) [ arxiv : hep - ph/0605314 ] .
m. perelstein , arxiv : hep - ph/0512128 ; h. c. cheng , i. low and l. t. wang , phys .
d * 74 * , 055001 ( 2006 ) [ arxiv : hep - ph/0510225 ] .
a. hektor , m. kadastik , m. muntel , m. raidal and l. rebane , arxiv:0705.1495 [ hep - ph ] .
t. han , h. e. logan and l. t. wang , jhep * 0601 * , 099 ( 2006 ) [ arxiv : hep - ph/0506313 ] .
t. han , h. e. logan , b. mcelrath and l. t. wang , phys .
b * 563 * , 191 ( 2003 ) [ erratum - ibid .
b * 603 * , 257 ( 2004 ) ] [ arxiv : hep - ph/0302188 ] .
d. chang , w. y. keung and s. c. lee , phys .
d * 38 * ( 1988 ) 850 .
solid lines ( blue ) and long dashed lines ( red ) show two- and three - body channels , respectively . due to the limited width of the two - body mode of @xmath87
, the leading four - body modes also reach high branching franctions , as shown in short dashed lines ( grey ) . ] .
cases of parent particles @xmath127 ( or similarly @xmath128 ) , @xmath129 ( or @xmath130 ) , @xmath131 , @xmath36 , @xmath132 , are shown in the composite graphs .
the branching fractions are calculated at @xmath133=200 gev when higgs bosons are involved . ] | @xmath0-parity in the little higgs model could be violated by anomalies that allow the lightest @xmath0-odd @xmath1 to decay into @xmath2 and @xmath3 .
we analyze these anomaly induced decays and the two - particle and the three - particle decay modes of other heavy quarks and bosons in this model which yield unique large hadron collider ( lhc ) signals with fully reconstructable events .
@xmath0-odd quarks in the little higgs model are nearly degenerate in mass and they decay by almost identical processes ; however , members of the heavy higgs triplet follow distinct decay modes .
the branching fractions of three - body decays increase with the global symmetry - breaking energy scale @xmath4 and are found to be at the level of a few percent in heavy quark decays while they can reach up to 10% for heavy bosons .
the higgs mass in the standard model ( sm ) receives large radiative corrections from the short - distance physics at the cutoff scale .
fine - tuning in the higgs sector becomes an eminent problem , especially when the sm predictions are confronted with precision electroweak data.@xcite in order to naturally alleviate the quadratic divergent contributions , new particles are expected to exist with tev scale masses .
the little higgs mechanism@xcite makes use of the light mass property of the pseudo - nambu
goldstone boson ( pngb ) to protect the higgs mass from the one - loop quadratic divergence : its mass receives one - loop radiative corrections from the new tev scale particles , which cancel the corrections from standard model fermion and boson loops . |
the lifting problem for commuting subnormals ( lpcs ) asks for necessary and sufficient conditions for a pair of subnormal operators on hilbert space to admit commuting normal extensions .
it is well known that the commutativity of the pair is necessary but not sufficient ( @xcite , @xcite , lu2 , @xcite ) , and it has recently been shown that the joint hyponormality of the pair is necessary but not sufficient @xcite .
abstract solutions of lpcs were given in ( * ? ? ?
* theorem 3.1 ) and ( * ? ? ? * theorem 2.7 ) , while concrete , necessary conditions , albeit not sufficient , for the lifting were found in ( * ? ? ?
* theorem 3.3 ) and ( * ? ? ? * theorem 2.10 ) ) in the case of @xmath9-variable weighted shifts .
in ( @xcite , @xcite , @xcite , @xcite , @xcite , @xcite and @xcite ) we have shown that many of the nontrivial aspects of lpcs are best detected within the class @xmath0 of commuting pairs of subnormal operators ; we thus focus our attention on this class .
more generally , we will denote the class of subnormal pairs by @xmath10 , and for each integer @xmath11 the class of @xmath12-hyponormal pairs in @xmath0 by @xmath13 . clearly , @xmath14 ;
the main results in @xcite and @xcite show that these inclusions are all proper .
the constructions in @xcite and cuyo2 have shed light on structural and spectral properties of multivariable weighted shifts , and have brought about some new phenomena in joint spectral theory .
more recently , we have made use of the tools and techniques in those papers and in @xcite and @xcite to approach lpcs from a new angle : to what extent the subnormality of the powers of a @xmath9-variable weighted shift can detect the subnormality of the pair .
in @xcite we discovered a large class of @xmath9-variable weighted shifts @xmath15 for which the subnormality of @xmath16 and @xmath17 does imply the subnormality of @xmath4 .
this is the class @xmath18 ( see definition [ tc ] below ) . in this paper
we study the subnormality of @xmath9-variable weighted shifts @xmath19 . since a @xmath9-variable weighted shift is subnormal if and only if its weight moments are the moments of a probability measure , known as the berger - gellar - wallen measure ( or briefly berger measure ) , the search for necessary and sufficient conditions leads to the following concrete problem .
[ tensorcore]let @xmath20 and assume @xmath4 is hyponormal . additionally , assume that @xmath21 is subnormal , with berger measure @xmath22 .
find necessary and sufficient conditions on the rest of the weight data to guarantee the subnormality of @xmath4 .
problem [ tensorcore ] is a special instance of the reconstruction - of - the measure problem , which we now describe .
given @xmath23 , the @xmath24-th row and the @xmath25-th column of the weight diagram have their own berger measures , @xmath26 and @xmath27 , respectively .
solving lpcs in this case amounts to finding a measure @xmath28 on @xmath29 which interpolates @xmath30 . without loss of generality
we can assume that @xmath31 and @xmath32all @xmath33 ( * ? ? ?
* theorem 3.3 ) .
moreover , @xmath26 must equal @xmath34 ( the marginal measure of @xmath35 ) , where @xmath36 ; in fact , @xmath37 , where @xmath38 is the disintegration of @xmath28 by vertical slices ( * ? ? ?
* theorem 3.1 ) ; and similarly for @xmath27 . from this perspective , lpcs consists of `` compatibly gluing together '' the measures @xmath26 and @xmath27 on @xmath39 to produce a measure @xmath28 on @xmath40 which satisfies the required properties to be the berger measure of @xmath4 .
we claim this can be done explicitly for @xmath19 .
note that @xmath41 and @xmath42 .
our main result is theorem [ thm1 ] , which provides a complete solution to problem [ tensorcore ] : @xmath43 is subnormal if and only if measures @xmath44 and @xmath45 given by ( eqpsi ) and ( [ eqphi ] ) , respectively , are positive . as an application
, we give a concrete condition for the subnormality of flat @xmath9-variable weighted shifts ( proposition [ ex1 ] ) .
let @xmath46 be a complex hilbert space and let @xmath47 denote the algebra of bounded linear operators on @xmath46 .
we say that @xmath48 is _ normal _ if @xmath49 and _ subnormal _ if @xmath50 , where @xmath51 is normal and @xmath52 @xmath53 .
an operator @xmath54 such that @xmath55 is said to be _
hyponormal_. for @xmath56
let @xmath57:=st - ts$ ] .
we say that an @xmath58-tuple @xmath59 of operators on @xmath46 is ( jointly ) _ hyponormal _ if the operator matrix @xmath60:=}\left ( \begin{array}{llll } \lbrack t_{1}^{\ast } , t_{1 } ] & [ t_{2}^{\ast } , t_{1 } ] & \cdots & [ t_{n}^{\ast } , t_{1 } ] \\ \lbrack t_{1}^{\ast } , t_{2 } ] & [ t_{2}^{\ast } , t_{2 } ] & \cdots & [ t_{n}^{\ast } , t_{2 } ] \\
\text { \thinspace \thinspace \quad } \vdots & \text { \thinspace \thinspace \quad } \vdots & \ddots & \text { \thinspace \thinspace \quad } \vdots \\
\lbrack t_{1}^{\ast } , t_{n } ] & [ t_{2}^{\ast } , t_{n } ] & \cdots & [ t_{n}^{\ast } , t_{n}]\end{array}\right)\]]is positive on the direct sum of @xmath58 copies of @xmath46 ( cf .
@xcite , @xcite ) .
the @xmath58-tuple @xmath4 is said to be _ normal _ if @xmath4 is commuting and each @xmath61 is normal , and @xmath4 is _ subnormal _ if @xmath4 is the restriction of a normal @xmath58-tuple to a common invariant subspace .
the bram - halmos criterion for subnormality states that an operator @xmath48 is subnormal if and only if @xmath62for all finite collections @xmath63 ( bra , @xcite ) . using choleski s algorithm for operator matrices , it is easy to see this is equivalent to the @xmath12-tuple @xmath64 is hyponormal for all @xmath11 . for @xmath65 a bounded sequence of positive real numbers ( called _ weights _ ) , let @xmath66 be the associated unilateral weighted shift , defined by @xmath67all @xmath68 , where @xmath69 is the canonical orthonormal basis in @xmath70 for notational convenience , we will write sometimes @xmath71 " for @xmath72 . in particular , @xmath73 and @xmath74 for a weighted shift @xmath72 , _ the moments of @xmath75 _ are given as @xmath76it is easy to see that @xmath72 is never normal , and that it is hyponormal if and only if @xmath77 . similarly , consider double - indexed positive bounded sequences @xmath78 , @xmath79 and let @xmath80 be the hilbert space of square - summable complex sequences indexed by @xmath81 .
( recall that @xmath80 is canonically isometrically isomorphic to @xmath82 . )
we define the @xmath9-variable weighted shift @xmath83 by @xmath84where @xmath85 and @xmath86 .
clearly , @xmath87 in an entirely similar way one can define multivariable weighted shifts .
trivially , a pair of unilateral weighted shifts @xmath72 and @xmath88 gives rise to a @xmath9-variable weighted shift @xmath89 , if we let @xmath90 and @xmath91(all @xmath92 ) . in this case
, @xmath4 is subnormal ( resp .
hyponormal ) if and only if so are @xmath6 and @xmath7 ; in fact , under the canonical identification of @xmath80 and @xmath93 , @xmath94 and @xmath95 , and @xmath4 is also doubly commuting . for this reason ,
we do not focus attention on shifts of this type , and use them only when the above mentioned triviality is desirable or needed . given @xmath96 ,
the moment of @xmath97 of order @xmath98 is @xmath99 ( we remark that , due to the commutativity condition ( [ commuting ] ) , @xmath100 can be computed using any nondecreasing path from @xmath101 to @xmath102 . )
we now recall a well known characterization of subnormality for multivariable weighted shifts @xcite , due to c. berger ( cf .
* iii.8.16 ) ) and independently established by gellar and wallen @xcite ) in the single variable case : @xmath103 admits a commuting normal extension if and only if there is a probability measure @xmath28 ( which we call the berger measure of @xmath4 ) defined on the @xmath9-dimensional rectangle @xmath104\times \lbrack 0,a_{2}]$ ] ( where @xmath105 ) such that @xmath106 for all @xmath107 . in the single variable case ,
if @xmath72 is subnormal with berger measure @xmath108 and @xmath109 , and if we let @xmath110 denote the invariant subspace obtained by removing the first @xmath111 vectors in the canonical orthonormal basis of @xmath112 , then the berger measure of @xmath113 is @xmath114 ; alternatively , if @xmath115 is defined by @xmath116then @xmath117 we now formally define the class @xmath18 .
first , we need some notation : @xmath118 and @xmath119 .
[ tc ] ( i ) the _ core _ of a @xmath9-variable weighted shift @xmath4 is @xmath120 ; ( ii ) @xmath4 is said to be of _ tensor form _ if @xmath121 .
@xmath122(when @xmath4 is subnormal , this is equivalent to requiring that the berger measure be a cartesian product @xmath22 ) ; ( iii ) @xmath123 .
we now consider @xmath9-variable weighted shifts such as the one given by figure 1(ii ) , where @xmath124 is subnormal with berger measure @xmath125 and @xmath126 is subnormal with berger measure @xmath127 .
@xmath128further , let @xmath129 ( resp .
@xmath130 ) be subnormal with berger measure @xmath131 ( resp .
@xmath132 ) . by ( [ sam ] ) , and without loss of generality , we will always assume that @xmath133 and @xmath134 .
we recall several notions introduced in @xcite and @xcite : ( i ) given a probability measure @xmath28 on @xmath135 , with @xmath136 , the _ extremal measure _ @xmath137 ( which is also a probability measure ) on @xmath138 is given by @xmath139 ; and ( ii ) given a measure @xmath28 on @xmath138 , the _ marginal measure _
@xmath140 @xmath141resp .
@xmath142 is given by @xmath143 @xmath141resp .
@xmath144 , where @xmath145 @xmath141resp .
@xmath146 is the canonical projection onto @xmath147 @xmath141resp .
@xmath148 .
thus , @xmath149 , for every @xmath150 @xmath141resp .
@xmath151 , @xmath152 .
observe that if @xmath28 is a probability measure , then so are @xmath140 and @xmath153 . for a measure @xmath28 with @xmath154 , we write @xmath155 . for example , @xmath156and @xmath157 . finally , for an arbitrary @xmath9-variable weighted shift @xmath4 , we shall let @xmath158 denote the restriction of @xmath4 to @xmath159 where @xmath160(resp .
@xmath161 is the subspace of @xmath80 spanned by the canonical orthonormal basis associated with indices @xmath162 , where @xmath163 and @xmath164 @xmath165 and @xmath166 , respectively .
in particular , we simply denote @xmath167 and @xmath168 .
it follows that @xmath169 , the core of @xmath4 .
assume that @xmath3 is subnormal , with berger measure @xmath22 .
we let @xmath170and @xmath171where @xmath172 is the berger measure of the subnormal shift @xmath173 .
trivially , @xmath44 and @xmath45 are measures , but they may or may not be _ positive _ measures .
the following result is a very special case of the reconstruction - of - the - measure problem .
[ backext](subnormal backward extension of a @xmath9-variable weighted shift @xcite ) consider the @xmath9-variable weighted shift whose weight diagram is given in figure 1(i ) .
assume that @xmath174 is subnormal , with associated measure @xmath175 , and that @xmath176 is subnormal with associated measure @xmath177 .
then @xmath4 is subnormal if and only if(i ) @xmath178;(ii ) @xmath179;(iii ) @xmath180 .
moreover , if @xmath181 then @xmath182 . in the case
when @xmath4 is subnormal , the berger measure @xmath28 of @xmath4 is given by @xmath183 [ pro1]let @xmath184 be the @xmath9-variable weighted shift whose weight diagram is given in figure 1(ii ) .
then @xmath185 if and only if @xmath44 is a positive measure . in this case ,
the berger measure of @xmath186 is @xmath187 @xmath188 if @xmath189 then @xmath190 with berger measure @xmath191 .
note that @xmath192 . by lemma [ backext](iii ) ,
if we think of @xmath186 as the backward extension of @xmath193 ( in the @xmath194 direction ) , we then have@xmath195thus , @xmath44 is a positive measure.@xmath196 if @xmath44 is a positive measure then @xmath197 is a well defined and positive measure . by a direct calculation , we can see that @xmath198where , for notational convenience , we set @xmath199 .
therefore , @xmath200 interpolates all moments of @xmath186 , so @xmath189 and @xmath201 .
we now have : [ thm1]let @xmath184 be the @xmath9-variable weighted shift whose weight diagram is given in figure 1(ii ) .
then @xmath202 if and only if @xmath44 and @xmath45 are positive measures .
@xmath196 it suffices to find a probability measure @xmath28 satisfying@xmath203 let @xmath204clearly , @xmath28 is well defined .
observe that@xmath205since we are assuming that @xmath44 and @xmath45 are positive measures , it follows from [ eq2 ] that @xmath28 is also positive .
furthermore , observe that @xmath206and@xmath207thus,@xmath208therefore , @xmath28 is a probability measure . next , observe that@xmath209and , for @xmath210,@xmath211thus , if @xmath212 and @xmath210 , we have @xmath213if @xmath214 and @xmath215 , we have @xmath216finally , if @xmath214 and @xmath210 , we have @xmath217where , for notational convenience , we set @xmath199 and @xmath218 .
thus,@xmath219therefore , @xmath28 interpolates all moments of @xmath4 , so @xmath4 must be subnormal , with berger measure @xmath28 . @xmath188 assume that @xmath4 is subnormal with berger measure @xmath28 .
then @xmath186 is also subnormal , and by proposition [ pro1 ] we can see that @xmath44 is a positive measure .
we then have @xmath220since @xmath221 we get @xmath222.\end{aligned}\]]it follows that @xmath223now recall that @xmath224 , so from lemma [ backext](iii ) we obtain@xmath225thus , @xmath45 is a positive measure , as desired .
the proof of theorem [ thm1 ] gives a concrete formula for the berger measure of @xmath4 , namely @xmath226 in proposition [ pro1 ] and theorem [ thm1 ] we noted that @xmath44 ( resp .
@xmath45 ) is a linear combination of @xmath172 and @xmath132 ( resp .
@xmath125 , @xmath227 and @xmath228 ) , where @xmath172 and @xmath132 @xmath141resp .
@xmath125 , @xmath227 and @xmath229 are the berger measures of the subnormal @xmath230-variable weighted shifts in the vertical ( resp .
horizontal ) slices of @xmath4 .
thus , the following conjecture for @xmath9-variable weighted shifts seems natural .
let @xmath184 be the @xmath9-variable weighted shift whose weight diagram is given by figure 1(i ) .
then the subnormality of @xmath4 is determined by a countable collection of inequalities @xmath231 , where each measure @xmath232 is a linear combination of berger measures associated to the @xmath230-variable weighted shifts in vertical or horizontal slices of @xmath4 .
we can now give a concrete condition for the subnormality of flat @xmath9-variable weighted shifts @xmath233 recall that @xmath15 is called _ horizontally flat _ if @xmath234 for all @xmath235 , and _ vertically flat _ if @xmath236 for all @xmath235 @xcite .
if @xmath4 is horizontally and vertically flat , then @xmath4 is simply called _ flat _ ( see figure 2 ) .
it is straightforward to prove that @xmath4 is flat if and only if @xmath19 , with @xmath131 and @xmath132 @xmath230-atomic . without loss of generality , we can always assume that @xmath237 .
now recall that for @xmath238 , @xmath239 is subnormal with berger measure @xmath240 . finally , we know from ( * ? ? ?
* theorem 3.3 ) and ( * ? ? ?
* section 5 ) that if @xmath4 is flat and subnormal then @xmath125 and @xmath127 have the form @xmath241\rho $ \\ $ \eta _ { y}=\ell \delta _ { 0}+m\delta _ { b^{2}}+[1-(\ell + m)]\sigma $ , \end{tabular } \label{eta}\]]where @xmath242 @xmath243 , @xmath244 , and @xmath245 and @xmath246 are probability measures with @xmath247 , @xmath248 .
[ ex1 ] consider the @xmath9-variable weighted shift @xmath249 whose weight diagram is given in figure 2 .
the following statements are equivalent.(i ) @xmath250;(ii ) @xmath44 and @xmath45 are positive measures;(iii ) @xmath251 and @xmath252moreover , when @xmath4 is subnormal , its berger measure is given as @xmath253 \(i ) @xmath254 ( ii ) : this is straightforward from theorem [ thm1].(ii ) @xmath254 ( iii ) : in ( [ eqpsi ] ) and ( [ eqphi ] ) , observe that @xmath255 , @xmath256 , @xmath257 , @xmath258 and @xmath259 .
@xmath122thus , we have @xmath260}{y_{0}^{2}}\sigma _ { 1}-a^{2}\delta _ { b^{2 } } \label{eq5}\end{aligned}\]](where
@xmath261 ) and@xmath262since we are assuming that @xmath44 and @xmath45 are positive measures , it follows from ( [ eq5 ] ) and ( [ eq6 ] ) that @xmath263as desired.(iii ) @xmath254 ( i ) : let @xmath264
. then @xmath265 is well defined , and by the formula for @xmath44 ( given in ( [ eq5 ] ) ) and the condition @xmath266 , we see at once that @xmath267 .
furthermore , @xmath265 is the berger measure of @xmath186 , so that @xmath186 is subnormal .
we now wish to apply lemma backext .
note that@xmath268so that@xmath269we now see that the conditions ( i ) , ( ii ) and ( iii ) in lemma [ backext ] are satisfied , and therefore @xmath270 . | let @xmath0 denote the class of commuting pairs of subnormal operators on hilbert space , and let @xmath1 is of tensor form@xmath2 , where @xmath3 is the core of @xmath4 .
we obtain a concrete necessary and sufficient condition for the subnormality of @xmath5 in terms of @xmath3 , the marginal measures of @xmath6 and @xmath7 , and the weight @xmath8 . |
the idea of cosmic crystallography was introduced by lehoucq et al .
@xcite , and applied by them to several einstein - de sitter ( eds ) models with closed ( i. e. , compact and boundless ) spatial sections of euclidean geometry and nontrivial topology .
the simplest of these cases has the three - torus @xmath0 as spatial section - model e1 in lachize - rey and luminet @xcite review article on cosmic topology .
this model provides the sharpest example of the crystallographic effect of the topology : if @xmath1 is based on a cube of side @xmath2 as fundamental polyhedron ( fp ) then the distances between images of a single source are in the form @xmath3 with @xmath4 integers , so that a plot of the distribution of distances @xmath5 vs. @xmath6 provides neat peaks in the integral values @xmath7 in the other five eds models with closed spatial sections the peaks tend to be in smaller numbers or less sharp , because the distances between some of the images of a single source depend on the latter s position inside the fp .
this is characteristic of locally but not globally homogeneous models @xcite , and can be seen in the study of model e4 by fagundes and gausmann @xcite : when @xmath8 mpc is of the order of magnitude of the observable universe s radius , no peaks showed up in the distribution @xmath9 these studies are based on computer simulated catalogs , and can not yet be compared with real catalogs , given the present limitations of the latter .
the same absence of obvious peaks is true for a model with a closed hyperbolic manifold @xmath10 as spatial section , and hence the metric of friedmann s open model .
this is illustrated in figs . 1 and 2 , with @xmath10 the second of best s @xcite three manifolds with an icosahedron as fp , and
listed as number @xmath11 in the census of closed , orientable hyperbolic manifolds in weeks s computer program snappea 2.5.3 @xcite .
the figure shows plots of @xmath5 against the distances in megaparsecs , assuming the density parameter @xmath12 and hubble s constant @xmath13 km sec@xmath14mpc@xmath15 and a radius of @xmath16 mpc , equivalent to maximum redshift @xmath17 which we take to be the position of the surface of last scattering ( sls ) of the cosmic microwave background .
figures 1 and 2 show the plots for the distributions for the compact model and the infinite model , respectively .
although there is more wiggling in the former , the general shape is the same for both .
so we tried a further step in the crystallography method , which is to map the _ differences _ between the distributions for the compact and infinite cases . in sec .
2 we explain our method and present the results , and in sec .
3 the latter are discussed .
to fix our notation let the metric of the hyperbolic friedmann model be written as @xmath18.\ ] ] to simulate a random distribution of sources in the fp , we first looked for such a distribution in the ball enclosed by the icosahedron s circumscribing sphere , with radius @xmath19 this was done by dividing the ball into 13 shells of thickness @xmath20 with volume @xmath21 where @xmath22 is the medium radius of the shell ; and also dividing the solid angle space into ten zones , with @xmath23 in each zone , where @xmath24 and @xmath25 the zone s medium value of @xmath26 then a number ( @xmath27 ) of pseudorandom positions in the ball were chosen by computer , with the condition that the occupations in each shell and zone were proportional to @xmath28 and @xmath29 respectively .
finally a program routine excluded from the obtained points those that were outside the icosahedron .
the @xmath30 remaining ones were considered the positions of the sources in the fp .
the radius of the sls in normalized units is @xmath31 for the assumed values of @xmath32 and @xmath33 . in order to fill this space without gaps we needed @xmath34 copies of the icosahedron , besides the fp itself .
they were produced by the @xmath35 face - pairing generators @xmath36 ( one - letter words ) of the covering group @xmath37 , @xmath38 two - letter , and @xmath39three - letter words in the@xmath40 s .
this set of icosahedra cover hyperbolic space up to a radius @xmath41 , slightly larger than @xmath42 .
the @xmath34 operators were then applied to each of the @xmath30 sources , the result being accepted as an image if it lay inside the sls ( @xmath43 ) .
the total of potential images ( they are potential in the sense that , because of evolution and other factors , many of them may not be observable ) thus obtained , including the sources , was @xmath44
. then we looked for the present ( or comoving ) distances between each pair of these images , by multiplying the normalized distances by the curvature radius of the universe ( see , for example , @xcite ) , @xmath45 mpc .
a list was prepared by grouping the numbers of occurrence of distances in bins of @xmath46 mpc width .
the corresponding plot is shown in fig . 1 .
we proceeded to obtain a similar distribution in friedmann s open model , with the same parameters as above , but with the same number of pseudorandom sources as there are images , i. e. , @xmath44 sources .
a list similar to the one above mentioned was plotted in fig .
2 . as indicated in the introduction , we could not see a significant difference between these plots .
finally we subtracted the values of the second list from those of the first .
3 is a plot of these differences .
one notices wild oscillations in the scale of the bin width , modulated by a broad pattern , on the scale of @xmath47 we fitted this plot with a five - term fourier series of the form @xmath48 where @xmath49 nearest integer to @xmath50 mpc ) , @xmath51 258 , and \{@xmath52 } = \{96.6 , 134.4 , -211.7 , 127.0 , -89.4}.
a _ clifford translation _ of a metric space @xmath53 is an isometry @xmath54 of @xmath55 such that _ _ distance__@xmath56 is independent of point @xmath57 .
wolf @xcite . in a recent paper , gomero et al .
@xcite link the existence of sharp peaks in cosmic crystallography to that of elements of @xmath58 which are clifford translations of the universal covering space of @xmath59 .
( see also lehoucq et al .
@xcite . ) but in fig . 1 of @xcite we see a significant peak that may not be produced by a clifford translation .
the matter deserves further investigation . in the case of the hyperbolic space @xmath11 treated here
, there seems indeed to be no sharp peaks , as predicted by @xcite and @xcite .
( it should be interesting to look for them in the case of a highly symmetrical manifold , like the hyperbolic dodecahedron space @xcite . )
this is why we proceeded to look for the differences in the distribution of distances between this model and friedmann s open model , whose plot does show a significant pattern . of course we would not expect to be able to infer the correct cosmic topology from a single plot , even one with sharp peaks .
the topology of the universe will probably be revealed slowly and progressively , as more and better data are obtained , and several topology searching methods are applied to them .
e. g. thanks conselho nacional de desenvolvimento cientfico e tecnolgico ( cnpq - brazil ) for a doctorate scholarship and financial help for participation in this symposium .
h. v. f. thanks cnpq for partial financial support , and fundao de amparo pesquisa do estado de so paulo ( fapesp ) for a grant to attend this event .
figure 1 . a plot of the number of occurences of the distances in the abscissa , in bins of @xmath46 mpc , for model with the finite , multiply connected manifold @xmath11 as spatial section@xmath60 with a simulated catalog of @xmath61sources and @xmath44 potential images . | we try to apply the cosmic crystallography method of lehoucq , lachieze - rey , and luminet to a universe model with closed spatial section of negative curvature . but the sharp peaks predicted for einstein - de sitter closed models do not appear in our hyperbolic example .
so we turn to a variant of that method , by subtracting from the distribution of distances between images in the closed model , the similar distribution in friedmann s open model .
the result is a plot with much oscillation in small scales , modulated by a long wavelength quasi - sinusoidal pattern . |
mercury has a peculiar rotation : three rotation periods correspond to two revolution periods .
this spin - orbit resonance leads to interesting physical phenomena such as the longitudinal librations .
the librations are caused by the non - spherical mass distribution of mercury , on which the sun exerts a gravitational torque .
the difference between the orbital and the rotation periods leads to a varying torque along the orbit since the orientation of the long axis of mercury changes with respect to the direction to the sun .
eccentricity further contributes to the variability by causing changes in the distance to the sun and in the orbital speed .
the main libration has a period of 87.97 days , equal to mercury s orbital ( annual ) period , and an amplitude of @xmath0 arcsec @xcite .
in addition , there are smaller amplitude librations at harmonic ( semi - annual , ter - annual , ... ) frequencies .
those librations depend on the interior structure , notably the presence and the size of a liquid core inside the planet . by measuring the librations
, we can infer knowledge about the interior structure .
for example , by measuring mercury s 88d libration amplitude , @xcite concluded that mercury has a large liquid core . since a magnetic field has been detected by the mariner 10 spacecraft ( e.g. * ? ? ? * ) , it is thought that mercury may have a solid inner core inside its liquid core . @xcite and
@xcite have investigated the consequence of the addition of a solid inner core on the rotation dynamics of mercury .
recently , @xcite showed that , if the inner core is larger than about 1000 km , the difference on the 88d libration amplitude may be non - negligible , and of the same order as the present uncertainty , about 1.5 as .
another forced libration results from planetary perturbations .
the periodic force arising from the gravitational interaction of a planetary body with mercury causes a perturbation of mercury s orbital motion , changing its position relative to the sun and thus altering the solar torque acting on its equatorial bulge .
this is an indirect effect of the planets on the rotation of mercury .
these long - period forced librations induced by the planetary perturbations have been predicted by e.g. @xcite , @xcite and @xcite .
they have periods commensurate with the orbital revolution of the planets involved and are expected to have small amplitudes unless their period is close to the period of a free libration in which case a near - resonant amplification can occur . in the absence of an inner core , there is only one such mode , the free mantle libration .
this mode describes an oscillation of the axis of minimum moment of inertia about the mercury - sun line at perihelion @xcite .
the period of the free mantle libration depends on the moments of inertia of mercury , and is approximately 12 years .
this is very close to jupiter s perturbation on mercury s orbit at 11.86y ; @xcite and @xcite have shown that a forced libration of 20 arcsec or more can be generated , the exact amplitude depending on the moments of inertia . besides this 11.86y forced libration ,
at least 4 other long - period forced librations have amplitudes larger than the arcsecond level .
the previous studies on the long - period forced librations assumed no mantle - inner core coupling . adding an inner core has two effects on the free libration : first , as shown by @xcite , @xcite and @xcite , it can lengthen the period of the free libration of the mantle since the motion of the mantle is locked to that of the inner core for this mode . as a result , interior models with a large inner core may no longer have a free period close to jupiter s orbital period .
second , the presence of the inner core adds a second free libration , closely approximated by the free gravitational oscillation between the mantle and inner core , and thus the possibility of additional resonances at other orbit perturbation frequencies . in this study , we investigate how a non - spherical inner core coupled to the mantle and the outer core may influence the long - period forced librations . since
planetary perturbation periods may be close to the period of the two free modes , the long - period librations may be resonantly enhanced and this may lead to a rotation angle that substantially differs from a model where there is no solid inner core .
the presence of the solid inner core affects the rotation state of mercury and may result in a signature that is detectable in the observations , in which case it must be taken into account when analyzing the data . in our rotation model
, we also take into account the dissipation since this effect reduces the libration amplitudes and introduces phase lags .
we include viscous and electromagnetic coupling at the core - mantle and inner - outer core boundaries , as well as the effect of viscous deformation within the inner core .
the signature of the parameters responsible for the dissipation on the libration is discussed . in the theory section
, we derive equations for the amplitude of the long - period forced librations of the mantle and of the inner core .
the equations are given for cases with and without dissipation .
we then numerically evaluate these libration equations on the basis of recent interior models of mercury s @xcite ( section 1.7 ) .
the results are compared for different interior models . in the results section ( section 2 ) , we show that the amplitude of the long - period librations are of the order of a few arcsec , and much larger if the forcing period approaches the period of one of the free modes . in the last section ( section 3 ) , we compare predictions of libration models with and without mantle - inner core coupling with the earth - based radar observations of the rotation rate of mercury of @xcite in order to determine whether the size of the inner core can be determined on the basis of the currently available radar data .
we assume an equatorial flattened bi - axial model of mercury with a silicate shell composed of the mantle and the crust ( we use the symbol @xmath1 for the shell ) , a fluid outer core ( @xmath2 ) and the solid inner part of the core ( @xmath3 ) .
if the mantle and the inner core have a different rotation , their principal axes of inertia will be misaligned and there will be an effect of the gravitational and pressure coupling between these layers ( see for example @xcite ) .
the librational motion of the mantle and the solid inner core can be described by considering the change in angular momentum of these layers as a result of the external torque of the sun and the internal torques . for the libration equation , we assume that the mantle and inner core are rigid solids as the effect of elastic deformations has been shown to be below the observational detection limit @xcite .
we then have the equations of motion : @xmath4 the rotation angle of the mantle @xmath5 describes the orientation of the axis of minimum moment of inertia of the mantle @xmath6 relative to the intersection between the ecliptic and the orbital plane at j2000 .
similarly , @xmath7 is the rotation angle of the inner core .
@xmath8 is the true anomaly , @xmath9 the longitude of the pericenter , @xmath10 the distance between the mass centers of mercury and the sun , @xmath11 are the principal moments of inertia of the inner core and @xmath12 the mantle moments of inertia .
@xmath13 is the mean motion of mercury , @xmath14 the mass of the sun and @xmath15 the gravitational constant .
the factors @xmath16 and @xmath17 describe the strengths of the gravitational and pressure torques on the mantle and inner core , respectively @xcite .
these factors are defined by @xmath18 and @xmath19 , where @xmath20 and @xmath21 are the principal moments of inertia of the bottom part of the fluid core ( a layer between the inner core - outer core boundary ( icb ) and the smallest sphere that can be included in the fluid core , see fig . 2 of @xcite ) while @xmath22 and @xmath23 are related to the rest of the fluid core .
the terms proportional to @xmath24 and @xmath25 capture the solar gravitational torques on the mantle and inner core , respectively .
the additional terms arise from the pressure torques on the boundaries between the outer core and mantle and between the inner core and outer core .
it can be shown that in the limit of no inner core , the expression of @xmath16 reduces to @xmath26 , and we retrieve the classical equation of a planet with two layers .
@xmath27 is the gravitational - pressure coupling constant between the mantle and the inner core .
if the inner and outer parts of the core have uniform density @xmath28 ( @xmath29 for the fluid outer core , @xmath30 for the solid inner core and @xmath31 for the silicate shell ) , @xmath27 is defined by ( e.g. * ? ? ?
* ) @xmath32 \ , , \label{eq_k}\ ] ] where @xmath33 , @xmath34 and @xmath35 are the geometrical equatorial flattenings at the top of the mantle , core - mantle boundary ( cmb ) and icb , respectively .
if an ellipsoidal surface of constant density at a given radius has its three principal semi - axes defined by @xmath36 , the geometrical flattening in the equatorial plane is defined by @xmath37 . for the computation of the longitudinal librations , the polar flattening may be neglected . when radial density variations in both the fluid and solid cores are taken into account ,
the expression for @xmath27 is more complicated and given in @xcite .
since the effect of the small obliquity on the longitudinal librations is below the observational detection limit , the obliquity of mercury is assumed to be @xmath38 ( its true value is @xmath39 , @xcite ) so that mercury s equator and orbit are the same plane .
previous studies of the effect of the inner core on mercury s rotation focused on the amplitude of the 88d librations and considered a kepler orbit , in which the orbital elements are constant with time , except for the true and the mean anomalies . in order to derive differential equations that use a small angle , the rotation angle @xmath5
is usually related to the mantle libration angle @xmath40 with the relation @xmath41 , where @xmath42 is the mean anomaly of the orbital motion and @xmath43 is a chosen initial time .
a similar expression is used to relate @xmath7 and @xmath44 , the libration angles of the inner core . when long - period librations are considered , the longitude of the pericenter @xmath9 is no longer constant with time while the true anomaly @xmath45 and the distance between mercury and the sun @xmath46 have a quasiperiodic evolution because of the planetary perturbations that affect the orbit of mercury .
following @xcite we define the mantle libration angle @xmath40 and an inner core libration angle @xmath44 by @xmath47 where the angle @xmath48 is defined by @xmath49 .
all these angles are functions of time .
the angle @xmath48 can be written as a sum of the mean rotation and a frequency decomposition @xmath50 where @xmath51 , @xmath52 and @xmath53 are the amplitude , angular frequency and phase of the different planetary perturbations on mercury s orbit .
the origin of the angles @xmath5 and @xmath7 is defined by a fixed line with respect to an inertial frame , while the origin of the @xmath40 and @xmath44 angles is moving because of the planetary perturbations ( see figure 1 in @xcite ) .
only @xmath5 can be observed although the libration angle of the mantle @xmath40 is more convenient mathematically because it is a small quantity .
to compute the long - period forced librations , the differential equations ( [ eq_psm ] ) and ( [ eq_pss ] ) are expanded as two functions of the mean anomaly and of the eccentricity .
since we study the long - period librations , we discard the terms containing the mean anomaly of mercury in the right hand side of the equations .
however terms containing the second time derivative of the mean anomaly and the longitude of the pericenter ( @xmath54 ) remain in the left hand side of the equations @xcite .
the angles @xmath40 and @xmath44 are usually small , so that the small angle approximation can also be used . we get two linearized equations for @xmath40 and @xmath44 that contain the planetary perturbations on the orbit of mercury @xmath55 where the frequencies @xmath56 and @xmath57 are given by @xmath58 and where @xmath59 is a function of the eccentricity ( @xmath60 ) . in the linearized equations ( [ eq_gamf ] ) and ( [ eq_gasf ] ) , the planetary perturbations are not included as a torque acting on the mantle and the inner core ( this direct effect is negligible because of the large distance between mercury and the other planets ) but appear as an indirect effect , through the definition of the @xmath61 angles .
this additional term , coming from the time derivative of the angle @xmath5 , acts like a forcing in the differential equations . in the limit of no inner core , @xmath62 and eq . ( [ eq_omm2 ] )
becomes the usual expression of the free mantle libration frequency @xmath63 using the jpl de405/de406 ephemerides @xcite over a period of about 1000 years before and after j2000 and performing a frequency analysis , we get the frequencies @xmath52 , amplitudes @xmath51 and phases @xmath53 of the main planetary perturbations on mercury s orbit ( see table 1 ) .
we use the frequency mapping famous software ( ftp://ftp.obs-nice.fr/pub/mignard/famous ) .
[ cols="^,^,^,^,^ " , ] table 3 : summary of the 9 interior models classes .
crust thickness is 50 km and crust density is 2900 kg/@xmath64 for all models .
see @xcite and @xcite for more details on these models .
the spherical interior model is then transformed to a bi - axial model with equatorial flattening .
we assume that mercury has a rigid bi - axial mantle and a core that is in hydrostatic equilibrium with respect to the mantle .
we specify the geometrical flattening at the top and bottom of the mantle ( @xmath33 and @xmath34 ) , and we compute the flattening as a function of radius in the core in response to the flattened mantle , following the method described in the study of @xcite .
flattening versus depth in the core is calculated under the assumption of hydrostatic equilibrium . in the following sections , we consider that @xmath65 , @xmath66 and @xmath67 .
once the ratio between @xmath33 and @xmath34 is chosen , the flattening of the whole planet is adjusted to match the observational constraints on @xmath68 obtained from measurements by messenger @xcite .
we use the mean value of @xmath69 without assuming an error for all calculations . from the resulting core flattening with depth ,
the coupling strength s @xmath27 , @xmath16 and @xmath17 can be computed @xcite . with increasing inner core radius , the parameters @xmath17 and @xmath27 increase while @xmath16 decreases .
the difference between the parameter values for the 3 flattening assumptions is less than 2% for @xmath16 , between 14% and 30% for @xmath17 , and up to more than 50% for @xmath27 .
, @xmath70 ) as a function of the inner core size .
the different curves represent the 9 different interior models classes .
the dashed curves represent the 4 models with the hot temperature profile .
the light gray areas show the spread of the free libration periods resulting from the flattening assumption , for the ts class of interior model . in this gray area ,
the flattening assumptions varies from @xmath71 ( lower limit ) to @xmath66 ( upper limit ) .
the gray area is almost invisible around @xmath72 .
the horizontal black lines show the 6 main periods of the planetary perturbations .
the orange dashed curve does not extend below 1350 km because for this particular mantle mineralogy ( mc ) and the hot temperature profile , it is not possible to find interior models with a small inner core that fulfill the mass and radius constraints.,height=264 ] the first free mode @xmath72 ( from eq .
( [ eq_om1 ] ) ) represents the free libration of the combined mantle and inner core .
its period increases with inner core size ( fig .
[ fig_periods ] ) . without an inner core ,
this mode reduces to the mantle free libration period @xmath73 ( eq . ( [ eq_omm ] ) ) and its period is between @xmath74 and @xmath75y . for @xmath65 ,
the period @xmath72 is between @xmath74 and @xmath76y , approaching the value of about 18y for an almost fully solid mercury .
it stays very close to @xmath73 for small inner cores but significantly differs from the mantle libration period for inner cores larger than about 1000 km .
the effect of the flattening assumption on the first free period is very small : for @xmath66 , @xmath72 varies between @xmath74 and @xmath77y and between @xmath74 and @xmath78y for @xmath71 .
we plot in fig .
[ fig_periods ] a gray area covered by the different assumptions for the flattening , but this area is almost invisible .
the period of the second free mode @xmath70 ( mantle - inner core gravitational mode ) covers a very large range , depending on the mantle mineralogy , the temperature profile and the flattening assumption . without mantle - inner core coupling ,
this mode does not exist . for @xmath65 , @xmath70 is between @xmath79 and @xmath80y for interior models with a cold temperature profile and between about @xmath81 and @xmath82y for interior models with a hot temperature profile .
in contrast to @xmath72 , @xmath70 is significantly affected by the flattening assumption .
the free period @xmath70 varies between 3.6 and 12.8y for @xmath66 and between 2.5 and 9.4y for @xmath71 .
hence , @xmath70 increases with decreasing cmb flattening . in fig .
[ fig_periods ] , the light gray area shows the large spread in the free libration periods resulting from the different flattening assumptions , for one class of interior models .
the period @xmath70 decreases with inner core size .
this is because @xmath70 depends mainly on the strength of the mantle - inner core gravitational coupling ( constant @xmath27 ) , which depends primarily on the density contrast between the fluid and solid core at the icb . in order to satisfy the constraints on mercury s mass
, models with large ( pure fe ) inner core must have a larger weight fraction of sulfur in the fluid core and thus a larger density contrast at the icb : the larger @xmath27 then leads to a shorter @xmath70 period . as it can be seen on fig .
[ fig_periods ] , different interior density models result in vastly different @xmath70 periods .
there is a clear separation between the hot and cold mantle models ; for a given inner core size , hot models have a much lower sulfur fraction in the fluid core , and thus have a smaller @xmath27 and a longer @xmath70 period .
the hot and cold temperature are plausible temperature profiles for the mantle of mercury ; any intermediary temperature profile would produce a @xmath70 curve that lies in between that of the hot and cold models of fig . 1
however , as other temperature profiles are possible , the range of variation of @xmath70 can be appreciably larger .
lastly , we note that the determination of @xmath70 crucially depends on the radial variations in density in the core @xcite .
taking instead the density of the fluid and solid cores to be uniform and equal to their mean values , the range of @xmath70 values on fig .
[ fig_periods ] would be much narrower , from 2.8y to 5.2y .
when a planetary period is near one of the free periods , the libration amplitude at that frequency is resonantly amplified . whether a large amplification by resonance occurs ( i.e. intersection of @xmath72 or @xmath70 with a planetary orbit period , see the horizontal lines in fig .
[ fig_periods ] ) depends on the interior structure of mercury , and the equatorial flattening with depth . for an inner core radius smaller than about 1000 km ,
the period @xmath72 is very close to the orbital perturbation by jupiter at @xmath83y .
a large inner core moves the free libration period @xmath72 further away from the @xmath83y perturbation period , and would result in a smaller long - period forced libration amplitude .
however , the free period @xmath72 could be very close to the orbital perturbation by saturn ( @xmath84y ) if the inner core radius were approximately @xmath85 km .
the free mode period @xmath70 may be resonant with the @xmath86y , @xmath87y , @xmath88y and the @xmath89y orbital perturbations . for hot interior models ,
the free period @xmath70 may be close to the @xmath83y perturbation period for small inner cores . as a function of the inner core size for 9 different classes of interior models , using damping parameters of @xmath90/y and no viscous inner core deformation ( @xmath91 ) .
the different colors show the 6 main perturbing frequencies .
we use dashed lines for the 4 classes of hot interior models .
the outer core flattening is @xmath92 .
the thicker parts of the curves show the interior models for which the amplitude of the 88d libration falls within the 1-sigma interval of @xcite .
the amplitude of the 88d libration is about 38.5 as . ]
since the forcing frequencies are known from the orbital theory ( table 1 ) , we can compute the librations of the mantle and of the inner core for each of these frequencies .
the amplitudes computed in this section included viscous coupling at both the cmb and the icb ( @xmath90/y ) , no electromagnetic coupling and assumed a rigid inner core ( @xmath91 ) .
the level of amplification for each of the long - period forced librations differs for different interior models of mercury .
this is because each of these models have different coupling parameters @xmath27 , @xmath16 and @xmath17 , which affects the frequencies @xmath93 on which the amplitude depends ( see equation [ eq_psiman ] ) .
[ fig_amplit ] shows the absolute value of the 6 largest amplitudes of the long - period forced librations of the mantle @xmath5 as a function of the inner core size for the case @xmath94 .
a large amplification by resonance occurs when , for a specific inner core radius , the period of one of the free modes is close to the period of an orbital perturbation .
the main long - period forced libration is the one due to jupiter s perturbation on mercury s orbit at 11.86y period ( blue curves in fig .
[ fig_amplit ] ) . as in the case without inner core mantle coupling ( see figure 5 of @xcite ) , a large amplification of this libration occurs for inner cores that are not too large . for an inner core radius smaller than 1200 km ,
the effect of the inner core on the eigenperiod @xmath72 is weak ( fig .
[ fig_periods ] ) , and the amplitude of the 11.86y libration depends mainly on the ratio @xmath95 which determines the free mantle libration period ( see eq .
( [ eq_omm ] ) ) . for hot interior models with small inner cores
, the second free period @xmath70 can also be close to the 11.86y period .
however because the resonance factor associated with @xmath70 ( @xmath96 coefficient in eq .
( [ eq_psimandec ] ) ) is typically about 10 times smaller than the resonance factor associated with @xmath72 , the amplification in that case is much smaller .
long - period forced librations at five other frequencies may also have an amplitude larger than a few arcsec : at 3.95y , 5.66y , 5.9y , 6.57y and 14.7y periods . if mercury does not have an inner core , the amplitude of these librations are not amplified and always remain below 5 as . since the difference between the free periods and these forcing periods changes faster with changing inner core radius than for the 11.86y libration ( see fig .
[ fig_periods ] ) and since the amplitude of the planetary perturbation @xmath51 is small , these resonances occur over a narrower range of inner core radii . as the inner core radius increases from 50 to 600 km , for the cold interior models , the 5.66y libration amplitude ( black curves ) varies from minus a few arcsec to @xmath97 at the point when @xmath70 is precisely equal to 5.66y .
once @xmath70 crosses 5.66y , the resonance factor changes sign ( because of @xmath98 in eq .
( [ eq_psimandec ] ) ) and the mantle libration amplitude jumps to @xmath99 .
a further increase in inner core radius takes @xmath70 further away from the resonance and the mantle libration decreases back to minus a few arcsec .
the quantity plotted in fig [ fig_amplit ] is the absolute value of this amplitude , so it is remains positive .
however , note that the mantle libration amplitude passes through zero for a specific inner core radius .
this marks the location where the planetary forcing on the mantle is equal and opposite to the gravitational torque by the inner core .
this does not occur precisely at @xmath70 = 5.66y , but when @xmath70 is slightly offset ( and smaller ) .
this effect is more visible for the 5.66y period than for other orbital periods because it has the largest orbital perturbation ( see table 1 ) . for other orbital frequencies , since the orbital perturbations @xmath51 are much smaller , the amplitudes far from the resonant frequencies are very small ( close to zero ) and the change of sign happens for libration amplitude smaller than 1 as , outside the graph limits . in order to get the flattening of the mantle
, we use the mean value of @xmath69 from @xcite without assuming an error .
a different choice of @xmath69 within its @xmath100 error bound largely affects the amplitude and location of the resonance because @xmath69 determines the amplitude of the planetary flattening , and therefore it affects the values for @xmath27 , @xmath16 , and @xmath17 . because the free period @xmath70 is very sensitive to the flattening assumption , a small change in the core equatorial flattening induces a change in the position of the resonance and the level of amplification .
the choice of the @xmath101 ratio has the strongest effect on the resonance of the 3.95y , 5.9y and 6.57y periods . for the other orbital periods ,
the choice of the interior model is more important . as fig .
[ fig_amplit ] illustrates , the amplitudes of the long - period forced librations can not be predicted with any reasonable precision because of their strong dependence on interior parameters of mercury for which we do not have good constraints .
nevertheless , some general observations can be made .
for inner cores smaller than about 1200 km , the 11.86y libration is expected to have a forced libration amplitude of at least about 10 as .
its amplitude decreases to about 1 as for interior models with very large inner cores .
detection of this forced libration might thus be used to distinguish between very large inner core or not .
if a large amplitude of the 14.7y libration is detected with a precision of a few arcseconds , then it means that the inner core is large ( radius between 1300 and 1600 km ) .
the amplitude of the 3.95y libration might be large only if the inner core is very large .
however , this is a very narrow resonance .
finally , a very narrow resonant amplification of the 3 other planetary frequencies is possible for large or small inner core radius depending on whether the temperature of the mantle is hot , mild or cold . the interior models in fig . [ fig_amplit ]
cover a large range of @xmath95 values , and therefore a large range of 88d libration amplitudes .
however a recent fit of the 88d libration amplitude to the rotation data @xcite gives a 1-sigma interval for this amplitude of @xmath102 $ ] as .
models that agree with the 1 sigma interval of the 88d libration are represented by thick lines in the figure [ fig_amplit ] .
if we extend the uncertainty interval to a 3-sigma interval , then the number of allowed interior models is much larger : almost all the models in the figure are allowed .
figure [ fig_amplit2 ] shows the amplitudes of the angular rotation velocity variations @xmath103 ( eq . [ eq_psiman ] ) for the same 6 orbital perturbation frequencies and for the different interior models .
for 9 different classes of interior models and for the 6 main orbital frequencies .
the configuration is the same as for fig .
[ fig_amplit ] .
the amplitude of the 88d angular velocity is of the order of 1000 as / y ( out of the boundaries of the figure ) .
the horizontal dashed line shows the current uncertainty on the radar observations.,height=245 ] resonant amplifications are present at the same frequencies as the libration angle for specific interior models ( see fig .
[ fig_amplit ] ) .
their amplitude is usually smaller than 80 as / y , except for a very small range of interior models .
calculations for different choices of @xmath104 and @xmath105 than the nominal values @xmath90/y show that , provided @xmath104 and @xmath105 are not larger than approximately @xmath106/y , the only difference with the results presented in the previous section is a reduction of the peak amplitude of libration at the resonance periods . as discussed in section 1.5 , the addition of em coupling at the cmb should not alter the amplitude of @xmath104 significantly , whereas em coupling at the icb could lead to an amplitude of @xmath105 of the order of @xmath106/y for reasonable assumptions about the magnetic field strength
. therefore , except at the precise location of resonances , the results for the nominal case presented above are not largely affected by the nature and amplitude of coupling at the cmb and icb .
the maximal amplitude of the peak may be approximated by eq .
( [ eq_psimimax ] ) .
the damping parameter @xmath104 has a larger effect than the damping parameter @xmath105 on the libration amplitude of the mantle and of the inner core .
this effect is also observed in the maximal amplitude for @xmath5 .
this can be understood from the expression ( eq .
( [ eq_psimimax ] ) ) of the maximal amplitude for @xmath5 : in the denominator , the coefficient in front of the damping parameter @xmath104 is much smaller than the one in front of @xmath105 .
dissipation can significantly change the phase of the libration but since the electromagnetic and viscous couplings are usually smaller than the gravitational coupling , this only happens if a planetary perturbation frequency is very close to the free mode frequency . if the differences between the frequencies @xmath107 and @xmath108 are larger than about @xmath109 rad / y , the phase angle @xmath110 ( eq . ( [ eq_psiphase ] ) ) reduces to @xmath53 or to @xmath111 ( the value depends on the sign of the frequency differences ) with an error of less than @xmath112 for the nominal value of the dissipation parameters .
the widths of the resonance peaks depend mostly on the amplitude of the planetary perturbations @xmath51 and only very slightly on the damping parameters
. calculations with different choices of the inner core viscous relaxation time @xmath113 show that , provided @xmath113 is larger than about 25 years , the main difference in the results is a reduction of the peak amplitude of librations at the resonance .
thus , except for inner core radii where the free period @xmath72 or @xmath70 almost matches a planetary forcing period , our results in the previous section are barely sensitive to the precise value of @xmath113 . however ,
if @xmath113 is of the order of about 25y or less , then the amplitude of the long - period librations are reduced for a much broader range of inner core radii .
if @xmath113 is less than a few years , then the amplitudes of the long - period librations are almost reduced to the original orbital perturbation @xmath51 .
the phases of the long - period librations are also largely changed .
the amplitude of the 88d forced libration ( 38.5 as , @xcite ) has approximately the same order of magnitude as the long - period forced librations ( see fig . [ fig_amplit ] ) .
this is not the case for the amplitudes of the angular rotation velocity variations @xmath103 ( see fig .
[ fig_amplit2 ] ) : the 88d amplitude is about 1000 as / y , about 10 to 100 times larger than the long - period amplitudes .
this is due to the temporal derivative that leads to more prominent short - period effects in the rotation rate than in the libration angle .
the long - period librations will be more difficult to detect in spin radar data than in orientation data .
these amplitudes may also be compared to the present uncertainty on the observations of the rotation state of mercury .
the average precision on the rotation rates obtained with earth - based radar data @xcite is of the order of 180 as / y .
[ fig_amplit2 ] shows that a velocity amplitude larger than 180 as / y only occurs under a very specific range of inner core radii ( except in the case of the 11.86y libration period for which the signature may be detectable over a larger range of inner core radii ) .
given the current observational precision , it follows from the calculated libration velocities , that a large difference between models with and without inner core coupling can only be seen if mercury happens to be close to a resonance .
the libration angle can also be measured directly .
a measurement of the libration amplitude was recently provided by @xcite using messenger data with an uncertainty on the 88d libration amplitude of about 5 as .
the expected precision on the libration angle with the bepicolombo spacecraft is about 1 arcsecond @xcite .
[ fig_amplit ] shows that libration signatures for a large fraction of interior models would exceed a few arcseconds and would therefore be detectable .
therefore , the influence of the inner core in the long - period forced librations has a better chance to be detected on the basis of future observations of the libration angle rather than the rotation rate . in order to gain pertinent inferences about the interior structure of mercury
, one would need arcsecond precision observations of the libration angle over a time interval long enough to identify the planetary periods , i.e. , 5 - 15y .
clearly , predictions of the long - period forced librations from our model depend on many parameters that are not well known such as the inner core size , the interior density structure , etc . different combination of these parameters can generate long - period librations of equivalent amplitudes .
nevertheless , despite the non - uniqueness , the constraint of matching predicted with observed long - period librations would reduce the possible parameter space and thus provide valuable information on mercury s interior .
the 88d libration amplitude of the inner core @xmath44 is much smaller than the 88d libration amplitude of the mantle @xcite .
the amplitude of the 88d libration is between @xmath114 and @xmath115 as , depending on the interior model and the flattening assumption . at semi - annual period
, the inner core libration is below 0.2 as @xcite .
however , the long - period librations of the inner core may also be largely amplified , for the same frequencies as the mantle libration since the denominators in the equations for the libration amplitude are the same for the inner core and for the mantle ( eqs .
( [ eq_gaman ] ) and ( [ eq_gasan ] ) ) .
although inner core librations could in principle be detected through their effect on the degree - two , order - two gravitational coefficient @xmath69 , the signal is probably too weak to be observed in the near future by either messenger or bepicolombo .
in this section , we investigate whether including the coupling between the mantle and inner core can improve the fit to the radar data of @xcite and determine whether it is possible to constrain the size of the inner core from the measurements of the rotation rate .
earth - based radar measurements of the instantaneous spin state of mercury have been made at 35 different epochs between 2002 and 2012 @xcite . from the correlation of the radar echo signals at two earth stations
, @xcite obtained 35 instantaneous spin rate values .
currently , the averaged uncertainty on these rotation rates is about 180 as / y . since two polarizations have been recorded , two data points and the associated uncertainties are available at each epoch . + we fit the observed rotation rates with a rotation model for mercury that includes the short - period forced librations ( annual , semi - annual , etc . , see section [ sec88 ] ) and the long - period forced librations .
the model also includes dissipation at the cmb and icb ( damping parameters @xmath90/y ) .
we assume that the two free librations have not been excited recently ( e.g. by an internal forcing , @xcite ) and have attenuated to a negligible amplitude . adding a free libration
would also complicate the analysis by introducing additional parameters to estimate and by a possible overlap with the 11.86y libration . in order to compare the solution without an inner core with a situation with an inner core
, we use as an estimator of the goodness of the fit the reduced @xmath116 ( sum of the squares of the differences between the model and the data expressed in units of the uncertainty , divided by the number of degrees of freedom ) .
we seek a global minimum of the reduced @xmath116 for all the interior models . as a function of the annual libration amplitude for the ec interior model class with a cold temperature profile .
the inner core radius increases from right to the left .
the assumption for the flattening is @xmath92 .
the black ( dashed ) line shows the reduced @xmath116 using a rotation model with ( without ) inner core coupling.,height=188 ] fig .
[ fig_chi2 ] shows that the main effect on the reduced @xmath116 is the 88d libration amplitude .
the reduced @xmath116 has a strong dependence on the 88d libration amplitude , both for interior models that take into account the mantle - inner core coupling and for those that do not .
additionally we see that the long - period forced librations increases the reduced @xmath116 for very specific interior models . to match the constraints on mercury s mass , models with large inner core have smaller outer core radius ; this leads to an increase in @xmath117 and thus a decrease in @xmath95 and in the 88d amplitude .
therefore the interior models with small inner cores have 88d libration amplitude larger than 39 as in fig . 4 while models with large inner cores have a 88d libration amplitude smaller then 35 as . in general , the inclusion of mantle - inner core coupling in the rotation model does not improve the best fit ( fig .
[ fig_chi2 ] ) .
the global minimal reduced @xmath116 is about @xmath118 , which is about the same as for the situation without mantle - inner core coupling .
the effects of a small inner core ( radius smaller than about 500 km ) on the rotation model are usually so small that the rotation models with and without mantle - inner core coupling produce almost the same @xmath116 .
however if a resonance between one of the free modes and one of the planetary perturbation periods occurs for some particular interior model , the amplitude of the long - period libration becomes very large and the fit is usually degraded .
thus , based on the reduced @xmath116 analysis , we conclude that the radar observations suggest that mercury s librations do not contain large , resonantly amplified , long - period forced librations .
other flattening assumptions change the position of the resonances , but do not globally decrease the reduced @xmath116 . as a function of the inner core radius for the ec interior model and a cold temperature profile
( see @xcite for details ) .
the assumption for the flattening is @xmath119 and @xmath120 .
the black ( dashed ) line shows the reduced @xmath116 using a rotation model with ( without ) mantle - inner core coupling .
we see that a global minimum for the reduced @xmath116 is when the inner core radius is @xmath121 km , indicating the best fit model.,height=188 ] some interior models have been found that have a reduced @xmath116 smaller than 0.4 when inner core coupling is included .
for example , the reduced @xmath116 is @xmath122 ( see fig . [ fig_chi2glomin ] ) if @xmath92/2 , @xmath123 km and using a cold mantle with a low density ( ec model of @xcite ) .
this is due to a resonance between the 5.66y perturbation and the free period @xmath70 .
for some specific inner core radii close to 1186 km , an amplitude of 23 as for the 5.66y libration allows for a better fit to the observations . however , the improvement in the fit is extremely sensitive to the precise amplification of the 5.66y libration : a small increase in the amplification leads to a large reduced @xmath116 and degrades the fit dramatically .
if we explore the confidence interval for @xmath69 and the other parameters of our model , it is also possible to find very sharp resonances that can , for very specific interior model and inner core radius , slightly decrease the reduced @xmath116 .
the effect of the inner core can therefore provide a slightly better fit to the data for specific interior structure models .
however , the improvement is only marginal and it can not be reasonably argued that they provide a better match to the libration observations .
the uncertainty on the available data is presently too large to be sensitive to the effect of mantle - inner core coupling and by extension , to the size of the inner core . additionally , the time period covered by the available data is too short to separate the different libration amplitudes . since a dramatic worsening of
the fit may occur at resonances , we can only conclude that a very large amplification is most likely not present in mercury s librations . based on this
, combinations of interior parameters that would yield a free mode period very close to one of the main long - period librations must be rejected .
we presented a theory to compute the amplitude and phase of the long - period forced librations of mercury .
these long - period librations are caused by the planetary perturbations on the orbit of mercury .
our model takes into account the internal coupling that occurs between the solid inner core , fluid outer core and mantle .
this includes gravitational and pressure torques between these layers , as well as viscous and electromagnetic couplings at the cmb and icb .
our model also includes viscous deformation of the inner core .
previous studies that have investigated the long - period forced librations of mercury did not take into account the coupling dynamics of the inner core with the rest of the planet .
the inclusion of the inner core leads to two eigen modes that have a large range of possible periods .
their precise periods depend on the interior structure , including the size of the inner core .
large amplification of a long - period forced libration occurs if its period is close to one of the free mode periods .
our results show that for a large set of models , long - period forced librations may have an amplitude well above 5 arcsec . for a small inner core radius ,
the first free period @xmath72 is close to the jupiter perturbation period at @xmath83y while for a large inner core , it can be closer to saturn orbital perturbation at @xmath84y
. the second free period may be resonant with several orbital perturbations : @xmath86y , @xmath87y , @xmath88y , @xmath89y and even the @xmath83y perturbation .
except very close to resonance periods , the long - period forced librations are not sensitive to the strength and nature of the coupling at the cmb and icb if the damping parameters @xmath104 and @xmath105 are smaller than @xmath106/y .
our results are also insensitive to viscous deformation within the inner core , except for the maximal amplitude of the resonances peaks , provided they occur on a characteristic timescale longer than 25y .
we also tested whether the inclusion of an inner core can provide a better fit to the earth based radar observations of mercury s rotation rate .
because the long - period forced librations have amplitudes that are below the current uncertainty on these observations and because the time interval they cover is short compared to the main planetary forcing periods , the data are not very sensitive to these long - period forced librations . therefore , it is not possible to detect the dynamical influence of the inner core on mercury s forced librations on the basis of the currently available observations , other than the fact that very large forced libration amplitude at planetary periods are not suggested by observations .
the amplitude of the long - period forced librations predicted by our model are sufficiently large to be measurable by spacecraft observations of the libration angle of mercury .
this is provided the libration angle is observed for a sufficiently long time window and with a precision of a few arcseconds .
depending on their precision , future measurements of long - period forced librations might then allow us to place useful constraints on the interior structure of mercury . as we have shown in our study
, the combination of parameters ( inner core size , interior profiles of flattening and density , etc ) that can generate a given libration is non - unique .
nevertheless , matching predicted and observed long - period forced librations would reduce the possible parameter space and thus provide valuable information on mercury s interior .
this work was financially supported by the belgian prodex program managed by the european space agency in collaboration with the belgian federal science policy office .
md is currently supported by a discovery grant from nserc / crsng .
we acknowledge j.l .
margot , s. peale and an anonymous reviewer for their useful comments which helped us to improve the paper . , b. j. , johnson , c. l. , korth , h. , purucker , m. e. , winslow , r. m. , slavin , j. a. , solomon , s. c. , mcnutt , r. l. , raines , j. m. , zurbuchen , t. h. , 2011 .
the global magnetic field of mercury from messenger orbital observations .
science 333 , 18591862 .
buffett , b. a. , 1998 .
free oscillations in the length of day : inferences on physical properties near the core - mantle boundary . in : gurnis , m. , wysession , m. e. , knittle , e. , buffett , b. a. ( eds . ) , the core - mantle boundary region .
28 of geodynamics series .
agu geophysical monograph , washington , dc , pp . 153165 .
l . , peale , s. j. , solomon , s. c. , hauck , ii , s. a. , ghigo , f. d. , jurgens , r. f. , yseboodt , m. , giorgini , j. d. , padovan , s. , campbell , d. b. , 2012 .
mercury s moment of inertia from spin and gravity data .
journal of geophysical research ( planets ) 117 , e00l09. , d. e. , zuber , m. t. , phillips , r. j. , solomon , s. c. , hauck , s. a. , lemoine , f. g. , mazarico , e. , neumann , g. a. , peale , s. j. , margot , j .-
l . , johnson , c. l. , torrence , m. h. , perry , m. e. , rowlands , d. d. , goossens , s. , head , j. w. , taylor , a. h. , 2012 .
gravity field and internal structure of mercury from messenger .
science 336 , 214217 . , a. , oberst , j. , preusker , f. , gwinner , k. , peale , s. j. , margot , j. l. , zuber , m. t. , solomon , s. c. , 2012 .
a technique for measurements of physical librations from orbiting spacecraft : application to mercury . in : european planetary science congress 2012 , held 23 - 28 september , 2012 in madrid , spain .
pp . epsc2012 - 254 . , o. , tarits , p. , vacher , p. , rivoldini , a. , van hoolst , t. , 2009 . composition and formation of mercury :
constraints from future electrical conductivity measurements .
planetary and space science 57 , 296305 . | the planetary perturbations on mercury s orbit lead to long - period forced librations of mercury s mantle .
these librations have previously been studied for a planet with two layers : a mantle and a liquid core . here , we calculate how the presence of a solid inner core in the liquid outer core influences the long - period forced librations .
mantle - inner core coupling affects the long - period libration dynamics mainly by changing the free libration : first , it lengthens the period of the free libration of the mantle , and second , it adds a second free libration , closely related to the free gravitational oscillation between the mantle and inner core .
the two free librations have periods between 2.5 and 18 y depending on the internal structure .
we show that large amplitude long - period librations of 10 s of arcsec are generated when the period of a planetary forcing approaches one of the two free libration periods .
these amplitudes are sufficiently large to be detectable by spacecraft measurements of the libration of mercury .
the amplitudes of the angular velocity of mercury s mantle at planetary forcing periods are also amplified by the resonances , but remain much smaller than the current precision of earth - based radar observations unless the period is very close to a free libration period .
the inclusion of mantle - inner core coupling in the rotation model does not significantly improve the fit to the radar observations .
this implies that it is not yet possible to determine the size of the inner core of mercury on the basis of available observations of mercury s rotation rate .
future observations of the long - period librations may be used to constrain the interior structure of mercury , including the size of its inner core .
mercury , planetary dynamics , resonances , spin - orbit , mercury , interior |
we would like to thank jeppe andersen for help with the interpolation ( and extrapolation ) code .
rst would like to thank mandy cooper - sarkar and andreas vogt for discussions , and the royal society for the award of a university research fellowship .
gw would like to thank the science and technology facilities council for the award of a responsive research associate position .
the ippp gratefully acknowledges financial support from the science and technology facilities council .
+ e.b .
zijlstra and w.l .
van neerven , phys . lett .
* b272 * ( 1991 ) 127 ; ibid * b273 * ( 1991 ) 476 ; ibid * b297 * ( 1992 ) 377 ; nucl . phys . * b383 * ( 1992 ) 525 ; + s. moch , j.a.m . vermaseren and a. vogt ,
phys . lett . *
b606 * ( 2005 ) 123 ; + j.a.m .
vermaseren , a. vogt and s. moch , nucl . phys . * b724 * ( 2005 ) 3
. r.v .
harlander , w.b .
kilgore , phys .
* 88 * ( 2002 ) 201801 ; + c. anastasiou and k. melnikov , nucl .
b646 * ( 2002 ) 220 ; + v. ravindran , j. smith and w.l .
van neerven , nucl . phys .
* b665 * ( 2003 ) 325 ; + o. brein , a. djouadi and r.v .
harlander , phys .
lett * b579 * ( 2004 ) 149 ; + r.v .
harlander and w.b .
kilgore , phys .
* d68 * ( 2003 ) 013001 .
t. gehrmann , `` status of nnlo jet calculations '' , in `` proceedings of frif workshop on first principles non - perturbative qcd of hadron jets '' , lpthe , paris , france , 1214 jan 2006 , pp t004 ; and references therein .
s. glazov , proceedings of `` 13th international workshop on deep inelastic scattering ( dis 2005 ) '' , madison , wisconsin , 27 april1 may 2005 , p. 237 ; + m. dittmar _
et al . _ ,
`` working group i : parton distributions : summary report for the hera - lhc workshop proceedings '' , p. 54 , hep - ph/0511119 , hep - ph/0601012 . | we present a new set of parton distributions obtained at nnlo .
these differ from the previous sets available at nnlo due to improvements in the theoretical treatment .
in particular we include a full treatment of heavy flavours in the region near the quark mass . in this way ,
an essentially complete set of nnlo partons is presented for the first time .
the improved treatment leads to a significant change in the gluon and heavy quark distributions , and a larger value of the qcd coupling at nnlo , @xmath0 .
indirectly this also leads to a change in the light partons at small @xmath1 and modifications of our predictions for @xmath2 and @xmath3 production at the lhc . as well as the best - fit set of partons
, we also provide 30 additional sets representing the uncertainties of the partons obtained using the hessian approach . plus
2 mm minus 2 mm 23.0 cm 17.0 cm -1.0 in -42pt ippp/07/23 + dcpt/07/46 + 11th july 2007 + * update of parton distributions at nnlo * a.d .
martin@xmath4 , w.j .
stirling@xmath4 , r.s . thorne@xmath5 and g. watt@xmath6 + @xmath4 institute for particle physics phenomenology , university of durham , dh1 3le , uk + @xmath6 department of physics and astronomy , university college london , wc1e 6bt ,
uk the highest level of precision so far achieved for cross sections calculated in fixed - order perturbative qcd is next - to - next - to - leading order ( nnlo ) .
the cross sections are known at nnlo for many processes involving incoming protons ( or antiprotons ) , i.e. coefficient functions exist for deep - inelastic scattering @xcite , for hadro - production of electroweak particles such as @xmath2 and @xmath3 bosons @xcite ( as a function of rapidity @xcite ) , and even higgs bosons @xcite ( in both the standard and supersymmetric models ) .
the appropriate parton distributions to use with these hard cross sections are those which have been evolved using nnlo evolution kernels .
approximations to these @xcite based on exact moments @xcite and small-@xmath1 limits @xcite have been available for a number of years , and correspondingly approximate nnlo parton distributions have been extracted from global fits @xcite , or fits to dis data @xcite .
the complete calculation of the splitting functions was published in 2004 @xcite ( the results being numerically very similar to the previous estimates ) and this allowed improved nnlo parton distributions to be obtained @xcite .
nevertheless , some additional , sometimes unmentioned , approximations still exist in the published sets of nnlo partons .
one of the most significant of these is the treatment of heavy flavour . at nnlo
the partons become discontinuous at flavour transition points ( conventionally chosen to be at @xmath7 , with @xmath8 ) , due to the non - vanishing of transition matrix elements @xcite if the flavour number is allowed to vary .
there is no reason why any parton should be exactly zero at a particular scale , and it is an accident of the @xmath9 renormalisation and factorisation scheme that the structure of the perturbative heavy flavour is so simple that one can evolve from zero at a given scale even at nlo .
this is not true , for example , in the ms renormalisation scheme or even in the dis factorisation scheme ( an all orders heavy flavour prescription in this scheme is presented in @xcite ) where an @xmath1-dependent input at @xmath7 is required at nlo .
these discontinuities in nnlo partons have so far been ignored , as explained in the appendices of @xcite . in principle
it is possible to avoid this problem of discontinuities at the transition points by simply remaining in a fixed flavour number scheme .
however , in practice the heavy flavour coefficient functions appropriate for this choice are only known to nlo so this is not an option .
a complete nnlo set of partons must have a properly defined treatment of heavy flavour .
none of the previous sets satisfy this requirement .
also , the nnlo parton distributions should be obtained by a comparison to data using cross sections calculated at the same order .
these considerations make it necessary to update the nnlo parton distributions .
compared to the previous version , mrst2004 @xcite , there are two main changes in the theory : an implementation of a new variable flavour number scheme ( vfns ) for the heavy quark @xmath10 flavours @xcite ( based on the nlo schemes in @xcite and @xcite ) which maintains the continuity ( up to n@xmath11lo corrections which are very small ) of both neutral and charged current structure functions by introducing discontinuities in coefficient functions which counter those in the parton distributions ; and the inclusion of nnlo corrections @xcite to the drell yan cross sections .
the most important change compared to the previous nnlo partons , which already used the exact splitting functions @xcite , is the new vfns which leads to a significant change in the gluon and heavy quark distributions .
moreover , because we knew that the previous nnlo treatments were approximate , we did not provide uncertainties along with the central values .
now that the nnlo procedure is essentially complete we rectify this .
the size of the uncertainties due to the experimental errors on the data fitted , which is obtained using the hessian approach @xcite , is similar to that at nlo @xcite .
there is more work to do in order to estimate the theoretical uncertainty , which is certainly important in some regions @xcite .
= 2.74 in the procedure that we use to obtain the partons , and their uncertainties , is very similar to previous analyses , with the data used in the fit being essentially the same as in the mrst2004 analysis .
we provide 15 different eigenvector sets of partons , with an `` up '' and `` down '' set for each .
each of the 30 parton sets corresponds to a @xmath12 of 50 compared to the best fit , this corresponding to an approximate @xmath13 confidence - level for the uncertainty of the partons . by a factor of @xmath14 , because the increase in @xmath15 about the minimum is not completely quadratic for every eigenvector . ]
these partons can be used to estimate the uncertainty of physical quantities in exactly the manner explained for the nlo partons in @xcite .
the central values for the light quarks are similar to the previous nnlo analysis , though the evolution with @xmath16 is a little different in order to obtain the best fit with the different heavy flavour procedure .
this leads to the most marked change being at small @xmath1 and high @xmath16 , and we will return to this . at lower @xmath16
it is most illuminating to instead examine the difference between the nnlo quarks and the quarks at nlo @xcite .
the change in the up quark distribution when going from nlo to nnlo is shown in fig .
[ fig1 ] . at small
@xmath1 the effect of the coefficient functions , particularly @xmath17 , is important and the difference between the nlo and the nnlo distribution is greater than the uncertainty in each . at large @xmath1
the coefficient functions are again important , i.e. @xmath18 and the difference between nlo and nnlo is again larger than the uncertainty in each as seen in fig . [ fig1 ] . at small
@xmath1 the effect of the splitting functions is also important , particularly due to @xmath19 , which has a positive @xmath20 contribution .
this affects the gluon distribution via the fit to @xmath21 , and the nnlo gluon is smaller at very low @xmath1 than the nlo gluon , as shown in fig .
[ fig3 ] .
the major change in the partons comes about due to the improved treatment of the heavy quarks , @xmath8 . as before
, we assume that the heavy quark distributions are generated by perturbative evolution , i.e. by @xmath22 transitions , where @xmath23 is the singlet light quark distribution . at nnlo ,
heavy flavour no longer evolves from zero at @xmath7 .
rather the distributions , @xmath8 , have an input value given by the convolution @xmath24 where , in practice , the heavy flavour distribution starts from a negative value at low @xmath1 , since the main contribution is from the matrix element at small @xmath1 which is @xmath25 plus less singular contributions , and the singlet matrix element has the same small-@xmath1 behaviour up to a colour factor of @xmath26 .
these negative small-@xmath1 matrix elements are combined with the positive high- and moderate-@xmath1 parton distributions in the convolution to give negative heavy flavour contributions .
the resulting discontinuity at @xmath7 in the heavy flavour distributions is by no means insignificant , see fig .
[ fig4 ] .
the discontinuities make the use of a general - mass variable - flavour scheme essential , otherwise the structure functions would have sizable discontinuities at nnlo @xcite . , but there is little sensitivity to this choice since the @xmath27-dependent terms in the matrix elements account for evolution ( without a complete resummation , of course ) . if one made the transition at @xmath28 the input for the heavy partons would be smaller , although it would not be exactly zero anywhere . this would be more complicated to implement and the discontinuities in the other partons would change as well . ] alternatively , ignoring the discontinuities leads to large errors in the heavy flavour partons .
the discontinuities in the light partons are much less significant , being at most a few percent for the gluon , the change quickly being overshadowed by evolution , and are very small indeed for light quarks .
the increased evolution from the nnlo splitting function allows the nnlo charm distribution to catch up partially with respect to that at nlo , which starts from zero at @xmath29 , but it always lags a little behind at higher @xmath16 .
( similarly the correct nnlo charm distribution is smaller than the approximate mrst2004 nnlo distribution which turned on from zero . )
however , the added effect of the coefficient function at nnlo ( estimated at low @xmath16 in @xcite ) is to raise the nnlo value of @xmath30 above the nlo value .
the evolution of @xmath30 at nnlo compared to nlo is shown in fig .
[ fig5 ] .
the increase at low @xmath16 and the decrease which persists to high @xmath16 , means unambiguously that @xmath30 tends to be flatter in @xmath16 at nnlo than nlo .
note that this comparison is made after a refit to data so that the corresponding evolution for light quarks must be such as to fit the hera inclusive data on @xmath31 . in detail
this means that the light quarks at nnlo now actually need to evolve slightly more quickly than at nlo ( or than in the previous approximate nnlo treatment ) to make up for the decrease in the evolution for the heavy flavour .
this correction in the charm procedure then automatically affects the gluon , since it has to compensate for the change in evolution of @xmath30 .
the comparison with the low @xmath16 mrst2004 nnlo gluon is shown in the left of fig .
[ fig6 ] , while the right of fig .
[ fig6 ] shows the ratio after a long evolution length for both the gluon and the up quark ( the effect on the down quark being much the same ) .
we see that the change in the gluon , and on the light quarks at very high @xmath16 , is greater than the uncertainty in some places .
the correct heavy flavour treatment is vital .
our other theoretical improvement at nnlo is also non - trivial .
the nnlo corrections to the drell yan cross section are significant @xcite .
there is an enhancement at high @xmath32 due to large logarithms , which is similar to the @xmath33 enhancement in structure functions .
the nlo correction is large and the nnlo corrections are @xmath34 or more , as seen in fig .
[ fig7 ] . the quality of the fit to e866 drell yan production @xcite in proton proton collisions is @xmath35 at nlo and @xmath36 at nnlo .
the scatter of points is large and a @xmath37 is approaching the best possible with a smooth distribution . the quality of the fit is good , as seen in fig .
[ fig8 ] . it is worse for proton deuterium data .
the positive correction at nnlo requires the data normalisation to be @xmath38 ( @xmath39 at nlo ) , there being little freedom since both the sea quarks for @xmath40 and the valence quarks are already well determined by structure function data . the normalisation uncertainty on the data is @xmath41 , and a change of @xmath42 is large but acceptable .
the one remaining set of data for which we can not use nnlo coefficient functions is the tevatron high-@xmath43 inclusive jet data @xcite , since the calculation is not complete .
however , the systematic uncertainties on these data are of order @xmath34 , which is representative of the size of the nlo corrections , and indications @xcite are that the nnlo corrections are somewhat less than this ( the coupling being rather small at these very high scales , of course ) .
whether one includes these data or not can lead to changes in the high-@xmath1 gluon of over @xmath44 , so a potential error of probably less than @xmath34 from using only the nlo coefficient functions seems trivial in comparison , and we keep these data in the fit . the quality of the full fit for the set of data used is @xmath45 at nlo and @xmath46 at nnlo .
the nnlo fit is consistently better than the nlo fit .
there is most improvement in the description of the high-@xmath1 structure function data , where the nnlo coefficient function leads to a quicker evolution at low @xmath16 .
most dis data sets , and the tevatron jet data , have a slightly better fit quality at nnlo . the exception is the e866 drell yan data , as already noted . there is a tendency for the nnlo @xmath47 to increase with the improved theoretical treatment .
an exactly analogous nlo fit yields @xmath48 , i.e. much the same as in @xcite since the analysis is extremely similar . in comparison , at nnlo
@xmath49 , whereas in @xcite @xmath50 .
the difference can be traced mainly to the slowing of the heavy flavour evolution using the full treatment , leading to an increased coupling as well as a modified gluon distribution to compensate .
overall , although the fit is generally good , particularly at nnlo , there is some room for improvement , and the data would prefer a little more gluon at both high and moderate @xmath1 .
we examine the effect of the change in the nnlo partons on the predictions for @xmath2 and @xmath3 production at the lhc and the tevatron .
the results , compared to those using the nnlo partons of mrst2004 , are shown in table [ tab1 ] , where we use @xmath51 and @xmath52 .
the uncertainties on these cross sections are roughly @xmath53(expt . ) .
the results are largely unchanged for the tevatron cross sections , but there is an increase of @xmath42 in the lhc cross sections using the updated partons .
this is not difficult to understand .
we have already seen the difference in the gluon distribution in fig .
[ fig6 ] .
the larger gluon in the region @xmath54 , and the larger coupling , are required to drive the evolution of the light quarks more quickly to match the small-@xmath1 hera data when the new procedure produces a flatter @xmath30 .
this results in the larger light quark distributions at high @xmath16 for @xmath55 , as seen in the right of fig .
[ fig6 ] , and a 6@xmath56 increase in the lhc cross sections , which are dominated by light quark
antiquark annihilation with @xmath57 .
the tevatron probes the distributions at @xmath58 resulting in little change in the predictions .
ratios of heavy boson cross sections are also essentially unchanged . the 6@xmath56 change in the predictions of the @xmath59 rates at the lhc should not be regarded as an uncertainty it is a consequence of correcting an over - simplified treatment of heavy flavours in previous nnlo analyses .
we also note that this change is for different reasons than the recent change in predictions for vector boson cross sections at the lhc observed in going from the cteq6.1 to cteq6.5 parton distributions @xcite .
the latter was for nlo partons and came about due to the implementation of a general - mass vfns instead of a zero - mass vfns .
our change is due to an improvement in the general - mass vfns , the most important modification being the treatment of parton discontinuities , which is a feature which only appears at nnlo .
.total @xmath2 and @xmath3 cross sections multiplied by leptonic branching ratios at the tevatron and the lhc , calculated at nnlo using the updated nnlo parton distributions .
the predictions using the 2004 nnlo sets are shown in brackets.[tab1 ] [ cols="<,<,<",options="header " , ] we compare with the only other publicised nnlo partons available , those of alekhin @xcite .
we have a much larger @xmath47 , i.e. @xmath60 , as compared to @xmath61 and @xmath62 of refs .
@xcite at a given order , even for a given renormalisation scheme .
the differences are formally of higher order , but may be significant at the level of @xmath63 or so for an extraction of @xmath47 @xcite .
different parton analyses tend to use different definitions .
the details of our procedure can be found in @xcite with extension to nnlo . ] . there is not much difference in the high-@xmath1 valence quarks , except that explained by the difference in @xmath47 .
there are differences in the low-@xmath1 sea quarks but these are dominated by differences in the flavour treatments of @xmath64 and @xmath65 .
the difference between our and alekhin s gluon distribution at small @xmath1 is seen in fig .
[ fig9 ] , and is much bigger than the uncertainties .
this is due to the different heavy flavour treatments , which we do not believe to be consistent in @xcite , and which we have already shown to be important , as well as to differences in the data fitted and in the value of @xmath47 .
the gluons also differ a great deal at high @xmath1 , where , in our analysis , they are determined by the tevatron jet data @xcite ( the comparison now being excellent @xcite ) . in @xcite the gluon is unconstrained here , and should presumably have a large uncertainty .
in the @xmath9 scheme the gluon is more important for tevatron jets at high @xmath1 at nnlo than at nlo because the high-@xmath1 quarks are automatically significantly smaller , as seen in fig .
[ fig1 ] . in summary
, we have new theoretical corrections in our global analysis , and have obtained nnlo partons with uncertainties for the first time . these nnlo partons ( denoted mrst2006 )
can be found at http://durpdg.dur.ac.uk/hepdata/mrs , where , in order to have a precise treatment in the heavy quark transition region , there has been a major upgrade of the interpolation code , which now also allows a reliable extrapolation outside of the grid ( now defined for @xmath66 and @xmath67 ) for high @xmath16 and/or low @xmath1 .
the nnlo fit improves on that at nlo , as well as correcting our previous nnlo analysis . with respect to the latter , the value of @xmath47 at nnlo moves upwards significantly , and all partons change by a significant amount , particularly at small @xmath1 . as we have seen
, this has implications for predictions for processes at the lhc . for the future
, there are more new data to be included : neutrino deep - inelastic scattering data from nutev @xcite and chorus @xcite , hera jets @xcite , updated tevatron high-@xmath43 jets @xcite , new cdf lepton - asymmetry data in different @xmath43 bins @xcite , new heavy flavour data from hera @xcite , and a full treatment of nutev dimuon data @xcite .
there will also soon be averaged hera structure function data @xcite .
this will lead us to produce fully updated nlo and nnlo partons for the lhc complete with uncertainties both experimental and theoretical in a relatively short timescale .
however , until this major update can be finalised , the nnlo partons outlined in this note will serve as by far the most theoretically self - consistent , and most stringently constrained set currently available at nnlo . |
two - dimensional ( 2d ) strongly correlated electron systems have attracted much attention in condensed matter physics . for bis(ethylenedithio)-tetrathiafulvalene ( bedt - ttf ) salts ,
which are 2d strongly correlated organic conductors with various molecular arrangements,@xcite many theoretical scientists use the extended hubbard model in order to understand the role played by long - range coulomb interactions .
it should be noted that not only electron - electron interactions but also electron - phonon couplings can be significant because the crystal structure is often altered at a phase transition .
these materials exhibit many interesting phenomena such as metal - insulator ( m - i ) transitions with charge ordering , the realization of a spin liquid state,@xcite superconductivity,@xcite and so on .
in particular , transitions to states with a charge order ( co ) have been investigated theoretically@xcite and experimentally.@xcite @xmath0-(bedt - ttf)@xmath1rbzn(scn)@xmath2 ( called @xmath0-rbzn for simplicity hereafter ) is a representative compound which undergoes a co transition accompanied with a structural deformation by lowering temperature.@xcite the ground state of the @xmath0-rbzn salt is an insulator with the horizontal - stripe co along @xmath3 bonds ( hco-@xmath3 ) shown in fig .
[ modelf](b ) , where sites @xmath4 and @xmath5 are hole - rich , as confirmed by the x - ray structural analysis,@xcite @xmath6c - nmr measurements@xcite and polarized raman and ir spectroscopy.@xcite the importance of long - range electron - electron interactions is well recognized and the mechanism for stabilizing the hco-@xmath3 has been argued mainly on the basis of the low - symmetry structure . because the lattice distortions are coupled with the electron system , giving rise to the first - order transition sensitive to the crystal structure , electron - phonon interactions are also important .
quite recently , iwai @xmath7 @xmath8 . have observed photoinduced melting of co in the @xmath0-rbzn salt and in @xmath9-(bedt - ttf)@xmath1i@xmath10 ( called @xmath9-i@xmath10 for simplicity hereafter ) by femtosecond reflection spectroscopy.@xcite the @xmath0-rbzn and @xmath9-i@xmath10 salts show large and small molecular rearrangements , respectively , at the m - i transition .
their photoinduced dynamics are qualitatively different : the @xmath0-rbzn salt shows local melting of co and ultrafast recovery of co irrespective of temperature and excitation intensity , while the @xmath9-i@xmath10 salt shows critical slowing down .
thus , it is important to show how electron - phonon interactions are significant in the @xmath0-rbzn salt .
this paper is organized as follows . after a brief explanation of the model in the next section ,
we present the exact - diagonalization results in sec.[numerical ] for the hole - hole correlation functions , the hole densities , and the modulations of transfer integrals , from which we propose that the low - symmetry structure of the @xmath0-rbzn salt at low temperature is reproduced by introducing electron - lattice couplings in the model based on the high - symmetry structure at high temperature . then , in sec.[discussion ] , we discuss the numerical results on the basis of a perturbation theory from the strong - coupling limit .
a brief summary is given in sec.[summary ] .
-rbzn salt : ( a ) high - symmetry structure at high temperature , and ( b ) low - symmetry structure at low temperature with horizontal - stripe co along @xmath3 bonds .
the thin and thick circles represent the hole - poor and the hole - rich sites , respectively .
, width=302 ] we start with the following extended hubbard model at 3/4-filling with electron - phonon couplings of transfer - modulation type , @xmath11 c^\dag_{i,\sigma } c_{j,\sigma } + u \sum_{i } n_{i,\uparrow}n_{i,\downarrow } \nonumber \\ & & + v_c \sum_{\langle i , j \rangle_c } n_i n_j + v_p \sum_{\langle i , j \rangle_p } n_i n_j % - \sum_{i } \mu_i n_i % \nonumber \\ % & & + \sum_{\mu = c , p1,p4 } \frac{k_\mu}{2 } u_\mu^2 \ ; , \label{model}\end{aligned}\ ] ] where @xmath12 creates an electron with spin @xmath13 at site @xmath14 , @xmath15 , and @xmath16 .
@xmath17 , @xmath18 or @xmath19 is the transfer integral for the bond between the @xmath14-th site and its nearest - neighbor @xmath20-th site along the @xmath21 , @xmath22 or @xmath23 bond . @xmath24 and @xmath25 denote the nearest - neighbor pairs @xmath14 and @xmath20 along the @xmath21 bond and the @xmath26 bond , respectively .
@xmath27 represents the on - site coulomb interaction , and @xmath28 is the intersite coulomb interaction between the @xmath14-th site and the @xmath20-th site on the @xmath21(@xmath26)-bond .
@xmath29 is the @xmath14-th molecular translation or rotation explained later from the equilibrium position in the high temperature phase . @xmath30 and @xmath31 are the electron - phonon coupling strength and the elastic coefficient , respectively . for simplicity ,
we perform variable transformations as @xmath32
we show energy values in the unit of electron volt ( ev ) in the following . for electron - phonon couplings ,
we consider three kinds of molecular displacements : translations in the @xmath21-direction ( @xmath33 ) , those in the @xmath34-direction ( @xmath35 ) , and rotations ( @xmath36 ) as discussed in terms of elevation angles by watanabe @xmath7 @xmath8.@xcite we suppose that translations in the @xmath21(@xmath34)-direction contribute to modulations of the transfer integrals on the @xmath37-(@xmath22- ) and @xmath38-(@xmath39-)bonds .
we also assume that molecular rotations produce differences between the transfer integral on the @xmath40-bond and that on the @xmath23-bond .
then , the modulated transfer integral on each bond reads @xmath41 the signs here are so determined that @xmath42 corresponds to a deviation from the high - temperature crystal structure toward the low - temperature one of the @xmath0-rbzn salt .
in this section , we adopt @xmath43 and @xmath44 , @xmath45.@xcite typical values for transfer integrals @xmath46 in bedt - ttf salts are estimated from the extended hckel calculation.@xcite we suppose here from the experimental data@xcite that @xmath47 , @xmath48 , and @xmath49 correspond to the high - symmetry structure at high temperature , where the @xmath22 and @xmath23 bonds are equivalent .
we use the exact - diagonalization method for electrons , regard phonons as classical variables , and apply the hellmann - feynman theorem to impose the self - consistency condition on the phonons ( @xmath50 , @xmath51 , @xmath52 ) determined by @xmath53 we vary mainly coupling strengths , @xmath54 , @xmath55 , and @xmath56 . at the beginning , we clarify the ground - state properties of the model ( [ model ] ) in the absence of lattice distortions . in this case , the exact - diagonalization studies of the 4@xmath574-site cluster show that the hole densities are uniform ( @xmath58 ) in any combination of @xmath44 and @xmath59 ( not shown ) .
degenerate co states are mixed owing to the finite - size effect .
then , we have calculated correlation functions in this cluster .
figure [ corre ] shows the hole - hole correlation functions without electron - phonon couplings on the @xmath60 plane .
@xmath61 denotes the hole - hole correlation function along the @xmath62-bond between the @xmath14-th and its neighbor ( its second - neighbor ) site .
@xmath63 being nearly equal to @xmath64 means that @xmath65 is equally mixed with @xmath66 with small @xmath67 . on the other hand ,
very small @xmath68 represents that @xmath69 is mixed with @xmath70 .
namely , @xmath71 is close to 0 ( 0.5 ) when the charge disproportionation between the @xmath14-th and the @xmath20-th sites is large ( small ) . for large @xmath44 ,
e.g. , for ( @xmath44 , @xmath59)=(0.35 , 0.01 ) , @xmath72 , @xmath73 , @xmath74 , and @xmath75 .
this result means that the diagonal - stripe co is the largest at this point among all hole - hole correlations . on the other hand , for small @xmath44 ,
e.g. , for ( @xmath44 , @xmath59)=(0.01 , 0.35 ) , @xmath76 , @xmath77 , @xmath78 , and @xmath79 . here , the vertical - stripe co is the largest correlation . in the ground state with comparable @xmath44 and @xmath59 , the diagonal- and the vertical - stripe cos coexist . without electron - phonon couplings .
all phases are uniform , but the dominant hole - hole correlations are different .
the vertical - stripe co and the diagonal - stripe co coexist in phase i. the vertical - stripe co is the largest correlation in phase i@xmath80i , while the diagonal - stripe co is the largest correlation in phase i@xmath80i@xmath80i . , width=264 ] regarding the steep variation in the contour plots of the correlation functions in fig .
[ corre ] as a phase boundary , we obtain the ground - state phase diagram on the @xmath44-@xmath59 plane in fig . [ phase - diagram ] .
the phase diagram consists of three uniform phases with different hole - hole correlations ; the coexistent phase ( i ) , the phase with dominant vertical - stripe co correlation ( i@xmath80i ) , and the phase with dominant diagonal - stripe co correlation ( i@xmath80i@xmath80i ) .
this phase diagram is consistent with the previous work.@xcite the most important thing here is that we can not find the horizontal - stripe co phase in any combination of @xmath44 and @xmath59 , which is observed experimentally in the @xmath0-rbzn salt at low temperature .
below we mainly use @xmath81 , @xmath82 in phase i of fig . [ phase - diagram ] , where this ratio of @xmath83 is regarded as appropriate for the @xmath0-rbzn salt.@xcite to check whether these values themselves are appropriate , we have calculated the optical conductivity spectra by means of the continued fraction expansion to show them in fig .
@xcite it is found that for both polarizations @xmath84 they have a broad peak around @xmath85 ev and the conductivity for @xmath86 is larger than that for @xmath87 , which are consistent with the experimental findings.@xcite therefore , these interaction strengths are quite reasonable . ) on the 4@xmath574-site cluster for electric fields , @xmath86 ( thin line ) and @xmath87 ( thick line ) .
we use @xmath43 , @xmath81 and @xmath82 .
the @xmath67-functions appearing in the continued fraction expansion are broadened with width @xmath88=0.01 .
, width=264 ] each of the three kinds of electron - phonon couplings is studied at the particular point of @xmath81 and @xmath82 by the exact diagonalization . because we need to obtain the displacements consistently with bond densities
, we used the smaller system of the 12 sites hereafter and compared some results with those of the 16-site cluster .
the conclusion is found to be unchanged and , in fact , understood from the perturbation theory from the strong - coupling limit , as discussed later .
first , we see the effect of molecular rotations ( @xmath56 ) . the @xmath56-dependence of the modulations of transfer integrals and the hole densities is shown in fig .
[ all](a ) with @xmath89 fixed . for small @xmath56 ,
as all the phonons are undistorted and the hole densities are uniform ( @xmath90 ) , the ground state remains the coexistent state of the vertical - stripe co and the diagonal - stripe co. with increasing @xmath56 , the system discontinuously changes at a critical point @xmath91 , to the co state of broken symmetry .
this co pattern is the hco-@xmath3 ( @xmath92 , @xmath93 @xmath94 @xmath95 , @xmath96 ) , which agrees with the experimental findings.@xcite therefore , the coupling with molecular rotations ( @xmath56 ) plays an important role to realize this hco state .
it is noted that @xmath52 increases almost linearly after the critical point where the hco-@xmath3 state is stable .
this discontinuous change at the critical point would be caused by the finite - size effect : the undistorted state is stabilized by quantum tunneling between different co states , but its energy gain would vanish in the thermodynamic limit .
in fact , @xmath52 rises up linearly by infinitesimal @xmath56 in mean - field calculations,@xcite so we guess that the critical point approaches zero in the thermodynamic limit .
next , we consider the effect of translations in the @xmath21-direction ( @xmath54 ) . for large @xmath54 , they also stabilize the hco-@xmath3 state as shown in fig .
[ all](b ) . in the case of @xmath54 , the charge disproportionation between the hole - rich sites (
@xmath92 , @xmath93 ) and the hole - poor sites ( @xmath95 , @xmath96 ) is smaller than the case of @xmath56 , but the translations in the @xmath21-direction further stabilize the hco-@xmath3 .
the situation regarding @xmath55 is quite different from the cases of @xmath56 and @xmath97 .
the @xmath55-dependence of the transfer modulations and the hole densities at four sites is shown in fig .
[ all](c ) . in the presence of only @xmath51 , which represents translations in the @xmath34-direction ,
the co pattern is still a horizontal - type but the holes are localized on the @xmath98 and @xmath99 bonds .
thus , the experimental findings are not reproduced if only @xmath55 exists . from figs .
[ all](a)-[all](c ) , we conclude that the effects of molecular translations in the @xmath21-direction ( @xmath54 ) and molecular rotations ( @xmath56 ) are stronger than that of translations in the @xmath34-direction ( @xmath55 ) in this salt .
we also come to the same conclusion in the other phases of fig .
[ phase - diagram ] . in fig .
[ variousv ] , we perform the same calculations ( except for the parameters shown ) as in fig .
[ all](a ) for phases i@xmath80i and i@xmath80i@xmath80i .
the @xmath56-induced hco-@xmath3 state is stable in both of phases i@xmath80i and i@xmath80i@xmath80i when @xmath56 is large enough .
therefore , this characteristic phonon - induced feature is widely obtained on the @xmath44-@xmath59 plane , including @xmath100 ( irrespective of hole - hole correlations in the uniform ground state without electron - phonon couplings ) .
finally , we simultaneously take the three kinds of displacements into account .
considering a situation with thermal- or pressure - induced expansion or compression , we vary intersite coulomb interactions . in fig .
[ strng - cplng ] , we show the modulations of transfer integrals and the hole densities as a function of nearest - neighbor coulomb interactions . for small @xmath101 , only @xmath51 is finite with @xmath50 and @xmath52 being zero .
this situation is quite similar to fig .
[ all](c ) , and hence the most stable state is the hco-@xmath98&@xmath99 .
as @xmath101 approaches unity , which is a realistic value for @xmath0-rbzn , the experimentally observed hco-@xmath3 state becomes more stable than the hco-@xmath98&@xmath99 by gradually increasing @xmath50 and @xmath52 and rapidly decreasing @xmath51 . in the vicinity of @xmath102 for @xmath103 ,
@xmath104 , and @xmath105 , @xmath50 and @xmath52 approximately correspond to the low - symmetry structure of @xmath0-rbzn at low temperature . from the x - ray structure analysis,@xcite the molecular translations in the @xmath21- and @xmath34-directions are found to be @xmath106 and @xmath107 , and the corresponding distortions are @xmath108 and @xmath109 .
they give @xmath110 and @xmath111 .
the present coupling strengths @xmath112 and @xmath113 lead to @xmath114 and @xmath115 .
a rough estimation of phonon frequencies by @xmath116 with @xmath117 being the reduced mass of two bedt - ttf molecules shows @xmath118 and @xmath119 , which are quite reasonable values.@xcite as @xmath101 further increases beyond two , all modulations approach zero .
in particular , @xmath51 decreases monotonically .
we explain these behaviors with a perturbation theory from the strong - coupling limit in the following section .
thus , electron - phonon interactions are crucial to stabilize the horizontal - stripe co and to realize the low - symmetry structure of @xmath0-rbzn at low temperature . , @xmath120 and @xmath121 .
coupling strengths are fixed at @xmath122 , @xmath104 and @xmath105 .
the solid ( dashed - dotted ) line in the upper panel represents the magnitude of the molecular translations in the @xmath21(@xmath34)-direction in the low - symmetry structure of @xmath0-rbzn at low temperature.@xcite the dashed line represents the magnitude of the molecular rotations in this structure .
, width=302 ] all these results are consistent with mean - field studies of the same model.@xcite
in this section , we consider why the hco-@xmath3 is the most stabilized by the electron - phonon couplings .
it is easily understood in the strong - coupling limit .
we assume that , in this limit , the holes are perfectly localized on the @xmath3 bonds , namely @xmath123 as shown in fig .
[ modelf](b ) . in the ordinary perturbation theory ,
the second- and third - order contributions to the energy per site read @xmath124 where @xmath125 is the @xmath126 spin operator at the @xmath14-th site on the @xmath3 bond . in the strong - coupling limit
, the one - dimensional half - filled chain is formed along the @xmath3 bonds and @xmath127 is the exact ground - state energy of the isotropic @xmath126 heisenberg chain . according to the third term in eq .
( [ 2nd - perturb ] ) , there is energy gain from spin fluctuations , so that the ground - state energy of the hco-@xmath3 state is further lowered by increasing @xmath128 . on the other hand , because all terms in eq .
( [ 3rd - perturb ] ) are positive , @xmath129 represents energy loss .
when @xmath128 becomes large , @xmath130 becomes small to reduce the energy loss . from the relations ( [ formation ] ) , @xmath131 becomes large and @xmath132 becomes small .
equations ( [ 2nd - perturb ] ) and ( [ 3rd - perturb ] ) can be rewritten as @xmath133 where @xmath34 , @xmath134 and @xmath135 ( @xmath136 ) are independent of @xmath137 , and @xmath138 .
note that we ignore @xmath139 in deriving the above equations , and the coefficients @xmath135 are positive under the condition @xmath140 .
we can infer from eq .
( [ apprx2 ] ) that @xmath52 increases linearly as a function of @xmath56 when the ground state is the hco-@xmath3 state .
because of the high symmetry with respect to @xmath141 , eqs .
( [ apprx2 ] ) and ( [ apprx3 ] ) are independent of @xmath51 . once the distortions @xmath52 and @xmath50 are substantially large , however , the distortion @xmath51 is numerically found to stabilize the hco-@xmath3 further . before closing this section
, we make a brief comment on the spinless fermion case.@xcite in the @xmath142 limit of the model([model ] ) , the hco state becomes unstable even if the modulations of the transfer integrals are introduced because the energy gain from spin fluctuations is absent .
thus we demonstrate that cooperative effects of electron correlations and electron - phonon couplings are important in the @xmath0-rbzn salt .
the stability of the hco state relative to the diagonal co state suggested in the spinless fermion case without electron - phonon couplings@xcite might be caused by the absence of phase factors , which are present if the @xmath142 limit is naively taken .
we have investigated the cooperative effects of electron correlations and lattice distortions on the charge ordering in the @xmath0-rbzn salt . by means of the exact - diagonalization method for systems with up to 16 sites
, we have calculated the hole - hole correlation functions , the hole densities and the modulations of transfer integrals to clarify the role of electron - phonon couplings in this salt . in the absence of electron - phonon interactions ,
there appear three uniform phases characterized by quantum tunneling between the vertical - stripe cos , that between the diagonal - stripe cos , and that between the vertical- and the diagonal - stripe cos .
it is found that all of these phases are changed into the horizontal - stripe co by the introduction of electron - phonon couplings relevant to the @xmath0-rbzn salt .
we can mostly reproduce the low - symmetry structure of @xmath0-rbzn at low temperature by using the model based on the high - symmetry structure at high temperature and by choosing the coupling strengths appropriately .
we conclude that the structural deformation assists the horizontal - stripe co that is experimentally found.@xcite in particular , the effects of molecular translations in the @xmath21-direction and molecular rotations are found to be stronger than that of translations in the @xmath34-direction .
thus , electron - phonon couplings are significant to stabilize the hco-@xmath3 and to realize the low - symmetry structure of @xmath0-rbzn .
otherwise the long - range coulomb interactions favored a different co pattern . with the help of the perturbation theory from the strong - coupling limit
, we easily understand the mechanism for stabilizing the hco-@xmath3 state and the linearly increasing @xmath52 and @xmath50 modulations .
the authors are grateful to s. iwai for showing his data prior to publication and y. yamashita for fruitful discussions .
this work was supported by the next generation supercomputing project , nano science program , and grants - in - aid from the ministry of education , culture , sports , science and technology , japan .
99 t. ishiguro , k. yamaji and g. saito , _ organic superconductors _
( springer , heidelberg , 1998 ) .
j. m. williams , j. r. ferraro , r. j. thorn , k. d. carlson , u. geiser , h. h. wang , a. m. kini and m .- h .
whangbo , _ organic superconductors ( including fullerenes : synthesis , structure , properties , and theory ) _ ( prince hall , englewood criffs , nj , 1992 ) .
y. shimizu , k. miyagawa , k. kanoda , m. maesato and g. saito , phys .
lett . * 91 * , 107001 ( 2003 ) .
k. kanoda , physica c * 282 * -*287 * , 299 ( 1997 ) .
m. kino and h. fukuyama , j. phys .
65 * , 2158 ( 1996 ) .
h. seo , j. phys .
. jpn . * 69 * , 805 ( 2000 ) .
s. mazumdar , r. t. clay and d. k. campbell , phys .
b * 62 * , 13400 ( 2000 ) .
r. t. clay , s. mazumdar and d. k. campbell , j. phys . soc .
jpn * 71 * , 1816 ( 2002 ) .
h. seo , j. merino , h. yoshioka and m. ogata , j. phys .
jpn . * 75 * , 051009 ( 2006 ) ; and references therein . h. mori , s. tanaka and t. mori , phys . rev .
b * 57 * , 12023 ( 1998 ) .
h. mori , s. tanaka , t. mori , a. kobayashi and h. kobayashi , bull .
jpn * 71 * , 797 ( 1998 ) . m. watanabe , y. noda , y. nogami and h. mori , j. phys .
73 * , 116 ( 2004 ) .
k. miyagawa , a. kawamoto and k. kanoda , phys .
b * 62 * , r7679 ( 2000 ) .
r. chiba , h. yamamoto , k. hiraki , t. takahashi and t. nakamura , j. phys .
solids * 62 * , 389 ( 2001 ) .
h. tajima , s. kyoda , h. mori and s. tanaka , synth . met . * 120 * , 757 ( 2001 ) .
k. yamamoto , k. yakushi , k. miyagawa , k. kanoda and a. kawamoto , phys .
b * 65 * , 85110 ( 2002 ) .
s. iwai , k. yamamoto , a. kashiwazaki , f. hiramatsu , h. nakaya , y. kawakami , k. yakushi , h. okamoto , h. mori and y. nishio , phys , rev . lett .
* 98 * , 097402 ( 2007 ) l. ducasse , f. fritsch and f. caster , synth
* 85 * , 1627 ( 1997 ) .
y. imamura , s. ten - no , k. yonemitsu and y. tanimura , j. chem .
phys . * 111 * , 5986 ( 1999 ) .
t. mori , bull .
soc . jpn .
* 73 * , 2243 ( 2000 ) . t. mori , bull .
. jpn . * 71 * , 2509 ( 1998 ) ; t. mori , h. mori , and s. tanaka , bull .
. jpn . * 72 * , 179 ( 1999 ) ; t. mori , bull .
. jpn . * 72 * , 2011 ( 1999 ) .
j. merino , h. seo and m. ogata , phys .
b * 71 * , 125111 ( 2005 ) .
t. mori , j. phys .
jpn . * 72 * , 1469 ( 2003 ) .
e. dagotto , rev .
mod . phys . * 66 * , 763 ( 1994 ) .
h. tajima , s. kyoden , h. mori and s. tanaka , phys .
b * 62 * , 9378 ( 2000 ) .
y. tanaka and k. yonemitsu , j. phys .
. jpn . * 76 * , no . 5 ( 2007 ) . c. hotta , n. furukawa , a. nakagawa and k. kubo , j. phys .
* 75 * , 123704 ( 2006 ) . | combined effects of electron correlations and lattice distortions are investigated on the charge ordering in @xmath0-(bedt - ttf)@xmath1rbzn(scn)@xmath2 theoretically in a two - dimensional 3/4-filled extended hubbard model with electron - lattice couplings . it is known that this material undergoes a phase transition from a high - symmetry metallic state to a low - symmetry insulating state with a horizontal - stripe charge order ( co ) by lowering temperature . by means of the exact - diagonalization method , we show that electron - phonon interactions are crucial to stabilize the horizontal - stripe co and to realize the low - symmetry crystal structure . |
the shade area is the region of quantity @xmath17 that can be assessable to the resonance region .
2 . the @xmath12 evolution of the spin asymmetry function @xmath0 in the resonance region ; the dot - dashed
, short - dashed and solid lines correspond to @xmath0 @xcite at @xmath2 , @xmath53 and @xmath54 gev@xmath4 respectively .
the dash - line is the asymmetry @xmath65 in the scaling limit@xcite , and the data are from ref .
@xcite , which the circle(triangle ) represents @xmath0 at @xmath55(@xmath54 ) gev@xmath4 .
the function @xmath79 as a function of @xmath77 at @xmath2 .
the dashed lines are extensions of @xmath79 to @xmath81 see text .
the @xmath12 dependence of the integral @xmath109 the dashed line corresponds to the contribution from the resonance region only , and the solid line is the total result , see text . | in this paper , we show what we can learn from the cebaf experiments on spin - structure functions , and the transition from the drell - hearn - gerasimov sum rule in the real photon limit to the spin dependent sum rules in the deep inelastic scattering , and how the asymmetry @xmath0 approaches the scaling limit in the resonance region . the spin structure function in the resonance region alone can not determine the spin - dependent sum rule due to the kinematic restriction of the resonance region .
the integral @xmath1 is estimated from @xmath2 to @xmath3 gev@xmath4 .
the result shows that there is a region where both contributions from the baryon resonances and the deep inelastic scattering are important , thus provides important information on the high twist effects on the spin dependent sum rule .
pacs numbers : 13.60.hb , 11.50.li , 12.40.aa,14.20gk the polarized lepton nucleon scattering is determined by two spin structure functions @xmath5 , which can be written as@xcite @xmath6\ ] ] and @xmath7\ ] ] where @xmath8 and @xmath9 are the cross sections for photon scattering with spin of the photon parallel and antiparallel to the spin of the longitudinally - polarized nucleon with mass @xmath10 , @xmath11 represents the interference between the transverse and the longitudinal cross section , and k is the photon flux .
the parton model tells us that the spin dependent structure functions @xmath5 should have simple scaling behaviors in the large @xmath12 limit ; @xmath13 and the corresponding spin dependent sum rules@xcite @xmath14 provide us an important information on the spin - flavor correlations of the nucleon .
however , the average @xmath12 of the polarized lepton - nucleon scattering experiments may not be high enough to reach the scaling limit , and the @xmath12 dependence of the spin structure function becomes significant .
recent studies@xcite have shown that the @xmath12 dependence of the spin structure function should be taken into account in order to explain smc@xcite and slac@xcite measurements on the spin structure function of the nucleon along with the emc@xcite data .
theoretically , such @xmath12 dependence that comes from the high twist effects have been the subject of many studies@xcite .
the information on the spin structure function in the low @xmath12 region would be more interesting since the high twist effects become stronger , this is where the spin structure function in the resonance region plays an important role .
on the other hand , it has been shown@xcite that @xmath12 dependence of the spin structure function arises naturally due to the transition from a negative drell - hearn - gerasimov ( dhg ) sum rule@xcite to the positive spin - dependent sum rule in the large @xmath12 limit .
the investigations in the explicit quark model show@xcite that the dhg sum rule is dominated by the spin structure function in the resonance region , in particular , by the process @xmath15 , which has been confirmed by the data@xcite .
how this transition happens in the small @xmath12 region also provides an important test ground for theories , since the explicit quark model@xcite and chiral perturbation theory@xcite have given different predictions .
the focus of this paper is to examine the @xmath12 dependence of the spin structure function in the resonance region .
our investigation is based on the earlier studies of the @xmath12 extensions of the dhg sum rule@xcite , and we further replace the photon energy @xmath16 by variable @xmath17 which is analogue to the bjorken scaling variable in the deep inelastic scattering . we shall show that the resonance region alone is not sufficient to provide an accurate estimate of the dhg sum rule due to the kinematic restriction . the evolution of the spin asymmetry @xmath0 , which is defined as @xmath18 will also be studied in this framework .
a negative dhg sum rule at the real photon limit requires that the spin asymmetry function @xmath19 be negative at the real photon limit .
this leads to a transition from a negative @xmath0 at the real photon limit to the positive @xmath0 at large @xmath12 , the data@xcite tell us that such a transition should happen at a relative small @xmath12 .
if this is indeed to be the case , the @xmath0 in the resonance region will show how the transition occur .
furthermore , because the dhg sum rule is dominated by the exclusive process @xmath15 , one should expect a dramatic @xmath12 dependence in the small @xmath12 region .
thus , there might be a region that both contributions from the resonance region and deep inelastic scattering are important , where the @xmath12 dependence of the structure function is dominated by the leading high twist effects .
our investigation shows that this is possible for @xmath20 gev@xmath4 . in the next section
, we will establish the connection between the dhg sum rule and the integral of the spin structure function .
the @xmath12 evolution of the asymmetry @xmath0 will also be shown .
eqs .
[ 21 ] and [ 22 ] show that the dhg sum rule integral @xmath21 used in refs .
@xcite is not related to the spin dependent sum rule @xmath22 at large @xmath12 since the photon flux @xmath23 in eqs . [ 21 ] and [ 22 ] is defined as the equivalent photon energy .
therefore , one should estimate the integral @xmath24 , which is related the dhg sum rule in the real photon limit and the spin dependent sum rules in the deep inelastic scattering region .
furthermore , one has to go beyond the resonance region to give an accurate estimate of the integral @xmath25 .
this will be given in section 3 .
finally , the conclusion will be given in section 4 . in order to describe
the strong @xmath12 dependence around @xmath2 of the spin structure function due to the pion photo and electroproduction that dominate the dhg sum rule , it is more appropriate to define the variable @xmath17 as @xmath26 where @xmath27 is the threshold energy for the pion - electroproduction , which is related to the pion mass @xmath28 and the nucleon mass @xmath10 by @xmath29 the variable @xmath17 defined in eq . [ 1 ] is exactly the same as variable @xmath30 in ref .
@xcite , which gives a better description of the strong @xmath12 dependence of the structure function @xmath31 near threshold .
furthermore , the variable @xmath17 naturally leads to the bjorken scaling variable @xmath32 in the large @xmath12 limit , and makes it possible to write the dhg integral @xmath33 in terms of the integral @xmath22 at the real photon limit ; @xmath34 where the photon flux @xmath23 in eq .
[ 21 ] is the photon energy @xmath16 at the real photon limit .
the dhg sum rule gives @xmath35 where @xmath36 is the anomalous magnetic moments of the nucleon .
substitute @xmath37 and @xmath38 for both protons and neutrons into eq .
[ 6 ] , we have @xmath39 replacing the photon energy @xmath16 by the variable @xmath17 at low @xmath12 also shows kinematic restrictions of the resonance region ; rewriting the photon energy as @xmath40 where @xmath41 is the mass of final state when nucleon @xmath10 at rest absorbs the incoming photon with energy @xmath16 , the resonance region corresponds to the final state mass @xmath41 from the threshold @xmath42 to @xmath43 gev which is also the region that can be reached in the future cebaf experiments@xcite . in fig . 1
, we show the range of the variable @xmath17 that is accessible to the resonance region ; at @xmath2 , the resonance region covers @xmath44 , while @xmath17 is only limited to @xmath45 at @xmath46 gev@xmath4 .
this raises a serious question whether the resonance region , or the information from the cebaf experiments , alone would be sufficient to provide an accurate estimate of the spin dependent integral @xmath47 even at moderate @xmath12 .
however , the studies of the spin structure function in the resonance region could provide us very useful information on transition of the spin structure function from the real photon limit to the scaling limit in the deep inelastic scattering .
in particular , the data on the spin asymmetry function @xmath0 in the resonance region will reveal how it reaches the scaling limit , if @xmath0 has little dependence on @xmath12 after @xmath48 gev@xmath4 .
a significant amount of the pion and @xmath49 electroproduction data has been collected in the resonance region@xcite , this enable us to extract the transverse photo coupling amplitudes @xmath50 and @xmath51 for the most prominent resonances in the range @xmath52 gev@xmath4 . for the other relative weak resonances , we used the single quark transition model to calculate their photo coupling constants .
therefore , combining the contribution from the resonances and the background pion production , one can obtain the spin structure function in the resonance region , ( for further detail , see @xcite ) . in the figure 2
, we show that asymmetry @xmath0 of proton in the resonance region at @xmath2 , @xmath53 and @xmath54 gev@xmath4 , and the data at @xmath55 and @xmath54 gev@xmath4@xcite are also shown .
the asymmetry is mostly negative in the region @xmath56 in the real photon limit , since it is dominated by the process @xmath57 .
moreover , the cancellation between the transitions of spin flip and orbital angular momentum flip for the resonance @xmath58 and @xmath59@xcite also makes helicity amplitudes @xmath60 vanishes for these resonances , which leads to a negative @xmath19 for the higher resonances . as @xmath12 increases ,
the resonance @xmath61 is shifting to larger x , and the magnetic spin flip transition becomes dominant , thus the asymmetry @xmath0 becomes positive in the small x region , and changes the sign and approaches to unity as @xmath62 .
this tells us that the asymmetry @xmath0 approaches to the scaling limit in the small @xmath17 region first , and the range of variable @xmath17 that asymmetry @xmath0 scales increases as @xmath12 increases . at the scaling limit
, the valence quark model@xcite predicts @xmath63 where @xmath64 is the ratio between the structure functions of neutrons and protons , which give a very good description of the emc data@xcite in the large @xmath17 region . comparing with the asymmetry @xmath65 in the scaling limit , which is also shown in fig .
2 , we find that the @xmath0 in the resonance region does not reach the scaling limit at @xmath55 gev@xmath4 , and it is much closer to the scaling limit at @xmath66 gev@xmath4 for @xmath67 .
the future experiments at cebaf@xcite would certainly provide the answer in this regard .
there are several problems in estimate the @xmath12 dependence of the integral @xmath68 .
first ; there is little experimental information available on the longitudinal photon coupling @xmath69 in the resonance region , we can only estimate the integral @xmath70 where @xmath71 is the ratio between the longitudinal and transverse cross sections , and @xmath23 is defined as the equivalent photon energy .
this integral equals @xmath72 at the real photon limit and the large @xmath12 limit .
soffer and teryaev@xcite pointed out that the strong @xmath12 dependence of the spin structure function @xmath73 in the small @xmath12 region might be due to the contamination of the function @xmath74 , this means that the quantity @xmath11 in eq . [ 21 ] might play an important role in the small @xmath12 region .
this might be related to the threshold effect around @xmath2 , thus one expected it vanish very quickly as @xmath12 increases .
the quantitative calculation of @xmath11 can be done in the framework of valence quark model , and this is in progress .
second , the kinematic restriction shows that the spin structure function in the resonance region alone might not be sufficient to estimate the integral @xmath75 , fig .
3 shows the quantity @xmath76 as a function of @xmath77 at @xmath2 .
the dhg sum rule should be extracted at @xmath78 , and if the function @xmath79 reaches constant in the small @xmath77 region , it means that the integral @xmath80 can be neglected , thus an accurate dhg sum rule can be obtained experimentally .
unfortunately , fig .
3 shows that @xmath77 is not small enough to warrant such extraction , and @xmath79 has not become constant in the small @xmath77 region .
therefore , there is a theoretical uncertainty in the estimate of the dhg sum rule due to the kinematic restrictions of the resonance region , which has not been discussed in the literature@xcite .
the dhg could be obtained by averaging two extreme cases ; extrapolating @xmath81 from the function @xmath79 at small @xmath77 region assuming a smooth behavior of the function @xmath79 near @xmath78 , or neglecting the contribution from the outside of the resonance region , and the difference between these two cases represents the theoretical uncertainty . following this procedure ,
we find @xmath82 this result is in good agreement with the theoretical prediction for protons in eq .
[ 7 ] , and the theoretical uncertainty is quite significant .
therefore , one can generally write the integral for the spin dependent structure function @xmath75 as @xmath83 where @xmath84 represents the contribution from the resonance region between @xmath85 and 1 , and @xmath86 comes from the outside resonance region . the quantity @xmath87 in eq .
[ 8 ] is @xmath88 with @xmath89 gev .
thus , @xmath87 approaches to unity as @xmath12 increases , and consequently the contribution from the resonance region decreases .
our result shows that the asymmetry @xmath0 outside the resonance region may not reach the scaling limit for @xmath90 until @xmath91 gev@xmath4 , where the asymmetry data in deep inelastic scattering@xcite could be used .
fig .
4 shows our estimate the @xmath12 dependence of the integral @xmath25 .
the qualitative behavior of @xmath84 is consistent with the results in refs .
@xcite , there is a strong @xmath12 dependence of @xmath84 in the region @xmath92 gev@xmath4 due to the transition @xmath15 .
@xmath84 becomes positive at @xmath93 gev@xmath4 and reaches maximum at @xmath94 gev@xmath4 .
however , @xmath84 is not the whole story , one has to add @xmath86 to give an accurate estimate of @xmath25 . since the asymmetry @xmath0 in the resonance region shows a strong @xmath12 dependence in @xmath95 gev@xmath4 , we only add @xmath86 after @xmath96 gev@xmath4 , in which @xmath65 is taken from the emc data@xcite , the structure function @xmath31 is obtained from the parametrization of the experimental data by donnachie and landshoff@xcite , and quantity @xmath97 comes from the global fit of the experimental data by whitlow _
et al_@xcite .
the result shows that @xmath86 gives an dominant contribution , while @xmath84 only accounts for 15 to 20 percent of @xmath25 in the region @xmath98 gev@xmath4 .
this is where we believe that one could obtain a reliable information of the high twist effects .
if we write the @xmath12 dependence of @xmath25 above @xmath99 gev@xmath4 as @xmath100 our result in @xmath101 gev@xmath4 region gives @xmath102 and @xmath103 in particular , the parameter @xmath104 is consistent with recent analysis of the smc and emc data by close and roberts@xcite in which they conclude @xmath105 as well as the qcd sum rule prediction by balitsky _
et al_@xcite used in the analysis by ellis and karliner@xcite .
we have presented a framework to study the @xmath12 dependence of the spin structure function in the resonance region .
at the real photon and large @xmath12 limit , the spin dependent sum rule @xmath75 defined in eq .
[ 25 ] is related the dhg sum rule and the sum rules in the deep inelastic scattering by @xmath106 which also equals to the integral @xmath22 in these limits .
we show that the data from the resonance region alone is not enough to determine the integral of the spin structure function , even the dhg sum rule in the real photon limit .
however , the spin structure function in the resonance region does provides important insights into the transition of the spin structure function as the @xmath12 increases .
the dhg sum rule is a threshold effect .
it decreases very quickly as @xmath12 increases due to its strong connection with the process @xmath15 , which is proportional to @xmath107 at large @xmath12@xcite , thus it is not the hight twist effect that generates the leading @xmath108 corrections to the spin structure functions in the deep inelastic scattering region .
further theoretical and experimental studies of the quantity @xmath11 is needed , it will not only provide information on the spin structure function @xmath74 in the small @xmath12 region , but also allow more precise estimation on the @xmath12 dependence of the spin structure function @xmath73 .
this investigation is in progress .
further information on the asymmetry @xmath0 in the small @xmath17 region around @xmath2 is desirable to give a precise estimate of the dhg sum rule , it may also tell us how the transition from negative to positive @xmath0 happens quantitatively in the small @xmath17 region .
we show that there is a region that both contributions from resonance and the nonresonance regions are important , and the effects of the dhg sum rule is small .
therefore , one could obtain the leading high twist effects reliably .
indeed , our result of the @xmath12 dependence of @xmath25 is quite consistent with the recent theoretical analysis .
the future experiments at cebaf@xcite will certainly provide us more information in this area .
discussions with v. burkert are gratefully acknowledged .
this work is supported by the nsf grant phy-9023586 .
99 b. l. ioffe , v.a .
khoze and l. n. lipatov , hard processes , north - holland , new york ( 1983 ) .
j. d. bjorken , phys . rev . * 148 * , 1467(1966 ) ; phys . rev .
* d1 * , 1376(1974 ) .
j. ellis and r. l. jaffe , phys . rev .
* d9 * , 1444(1974 ) ; * d10 * , 1669(1974 ) .
f.e . close and r. g. roberts , phys .
lett . *
b316 * , 165(1993 ) .
j. ellis and m. karliner invited talk at 13th international conference on particles and nuclei , panic93 , 28 june-2 july 1993 , perugia , italy , cern preprint cern - th .
7022/93 .
zhenping li , phys . rev . *
d48 * , r3945(1993 ) .
spin muon collaboration , b. adeva _
et al . _ , phys . lett . *
302b * , 533(1993 ) .
slac - e142 collaboration , p.l .
anthony _
et al . _ ,
phys .
rev .
lett . * 71 * , 959(1993 ) . j. ashman _
et al _ , phys
. lett . *
206b * , 364(1988 ) ; nucl . phys . *
b328 * , 1(1990 ) .
i.i .
balitsky , v. m. braun and a. v. kolesnichenko , phys . lett . *
b242 * , 245(1990 ) .
x. ji and p. unrau , preprint mit - ctp-2232 , hep - ph/9308263 . s. d. drell and a. c. hearn , phys
. rev .
lett . * 162 * , 1520(1966);s . b. gerasimov , yad . fiz .
* 2 * , 839(1965 ) [ sov
. j. nucl
. phys .
* 2 * , 589(1966 ) ] .
m. anselmino , b. l. ioffe , and e. leader , yad . fiz . * 49 * , 214(1989)[sov . j. nucl
. phys . * 49 * , 136(1989 ) ] .
zhenping li , phys . rev . * d47 * , 1854(1993 )
. v. burkert and zhujun li , phys
. rev . * d47 * , 46(1993 )
. v. burkert and b. l. ioffe , phys
. lett . *
296b * , 223(1992 ) .
v.burkert and zhujun li , ao , computer code , unpublised .
v. bernard , n. kaiser and ulf - g .
meissner , phys . rev .
* d48 * , 3062(1993 ) .
a. donnachie and p.v .
landshoff , university of manchester preprint m / c - th 93/11 , damtp 93 - 23 , nucl .
phys . *
b244 * , 322(1984 ) . g. baum _
et al _ , phys . rev .
lett .
* 45 * , 2000(1980 ) .
f. foster and g. hughes , rep . prog
. phys . * 46 * , 1445(1983 ) .
f.e . close , nucl .
phys . *
b80 * , 269(1974 ) ; invited talk at the 6th ictp workshop , trieste , italy , may 3 - 7 , 1993 .
ral preprint ral-93 - 034 .
f. e. close and f. gilman , phys
. lett . *
b38 * , 541(1972 ) . f. e. close and zhenping li , phys . rev . * d42 * , 2194(1990 ) , _ ibid . _ * d42 * , 2207(1990 ) . v. burkert _ et al _ , measurement of polarized structure functions in inelastic electron proton scattering using the cebaf large acceptance spectrometer , " report no .
cebaf - pr-91 - 023(unpublished ) .
j. soffer and o. teryaev , phys .
rev .
lett . * 70 * , 3373(1993 ) .
i. karliner , phys . rev . *
d7 * , 2712(1973 ) ; r. workman and r. arndt , phys .
rev . * d45 * , 1789(1992 ) .
l.w .
whitlow _
et al _ phys
. lett . *
b250 * , 193(1990 ) . |
the construction of useful confidence sets is one of the more challenging problems in nonparametric function estimation .
there are two main interrelated issues which need to be considered together , coverage probability and the expected size of the confidence set . for a fixed parameter space
it is often possible to construct confidence sets which have guaranteed coverage probability over the parameter space while controlling the maximum expected size . however such minimax statements are often thought to be too conservative , and a more natural goal is to have the expected size of the confidence set reflect in some sense the difficulty of estimating the particular underlying function .
these issues are well illustrated by considering confidence intervals for the value of a function at a fixed point .
let @xmath0 be an observation from the white noise model @xmath1 where @xmath2 is standard brownian motion and @xmath3 belongs to some parameter space @xmath4 .
suppose that we wish to construct a confidence interval for @xmath5 at some point @xmath6 .
let @xmath7 be a confidence interval for @xmath8 based on observing the process @xmath0 , and let @xmath9 denote the length of the confidence interval .
the minimax point of view can then be expressed by the following : subject to the constraint on the coverage probability @xmath10 , minimize the maximum expected length @xmath11 . as an example it is common to consider the lipschitz classes @xmath12 $ } \bigr\},\qquad \mbox{if $ 0<\beta\le1$}\ ] ] and for @xmath13 @xmath14 $ } \bigr\},\ ] ] where @xmath15 is the largest integer less than @xmath16 and @xmath17 .
for these classes it easily follows from results of @xcite , @xcite and @xcite that the minimax expected length of confidence intervals , which have guaranteed coverage of @xmath18 over @xmath19 , is of order @xmath20 .
it should , however , be stressed that confidence intervals which achieve such an expected length rely on the knowledge of the particular smoothness parameters @xmath16 and @xmath21 , which are not known in most applications .
unfortunately , @xcite and @xcite have shown that the natural goal of constructing an adaptive confidence interval which has a given coverage probability and has expected length that is simultaneously close to these minimax expected lengths for a range of smoothness parameters is not in general attainable .
more specifically suppose that a confidence interval has guaranteed coverage probability of @xmath22 over @xmath19 .
then for any @xmath5 in the interior of @xmath23 the expected length for this @xmath5 must also be of order @xmath24 .
in other words the minimax rate describes the actual rate for all functions in the class other than those on the boundary of the set .
for example , in the case that a confidence interval has guaranteed coverage probability of @xmath25 over the lipschitz class @xmath26 , then even if the underlying function has two derivatives , and the first derivative smaller than @xmath21 , the confidence interval for @xmath27 _ must _ still have expected length of order @xmath28 even though one would hope that an adaptive confidence interval would have a much shorter length of order @xmath29 . despite these very negative results there are some settings where some degree of adaptation has been shown to be possible . in particular under certain shape constraints @xcite constructed confidence bands which have a guaranteed coverage probability of at least @xmath30 over the collection of all monotone densities and which have maximum expected length of order @xmath31 for those monotone densities which are in @xmath19 for a particular choice of @xmath16 where @xmath32 .
this construction relies on the selection of a tuning parameter and is thus not adaptive .
dmbgen ( @xcite ) , however , does provide adaptive confidence bands with optimal rates for both isotonic and convex functions under supremum norm loss on arbitrary compact subintervals .
these results are , however , still framed in terms of the maximum length over particular large parameter spaces , and the existence of such intervals raises the question of exactly how much adaption is possible .
it is this question that is the focus of the present paper . rather than considering the maximum expected length over large collections of functions , we study the problem of adaptation to each and every function in the parameter space .
we examine this problem in detail for two commonly used collections of functions that have shape constraints , namely the collection of convex functions and the collection of monotone functions .
we focus on these parameter spaces as it is for such shape constrained problems for which there is some hope for adaptation . within this context
we consider the problem of constructing a confidence interval for the value of a function at a fixed point under both the white noise with drift model given in ( [ w.model ] ) as well as a nonparametric regression model .
we show that within the class of convex functions and the class of monotone functions , it is indeed possible to _ adapt to each individual function _ , and not just to the minimax expected length over different parameter spaces in a collection .
the notion of adaptivity to a single function is also discussed in lepski , mammen and spokoiny ( @xcite ) and @xcite for the related point estimation problem but in these contexts a logarithmic penalty of the noise level must be paid , and thus the notion of adaptivity is somewhat different.=1 this result is achieved in two steps .
first we study the problem of minimizing the expected length of a confidence interval , assuming that the data is generated from a particular function @xmath5 in the parameter space , subject to the constraint that the confidence interval has guaranteed coverage probability over the entire parameter space .
the solution to this problem gives a benchmark for the expected length which depends on the function @xmath5 considered .
it gives a bound on the expected length of any adaptive interval because if the expected length is smaller than this bound for _ any _ particular function , the confidence interval can not have the desired coverage probability . in applications
it is more useful to express the benchmark in terms of a local modulus of continuity , an analytic quantity that can be easily calculated for individual functions . in situations where adaptation is not possible ,
this local modulus of continuity does not vary significantly from function to function .
such is the case in the settings considered in @xcite .
however , in the context of convex or monotone functions , the resulting benchmark does vary significantly , and this opens up the possibility for adaptation in those settings .
our second step is to actually construct adaptive confidence intervals .
this is done separately for monotone functions and convex functions , with similar results .
for example , an adaptive confidence interval is constructed which is shown to have expected length uniformly within an absolute constant factor of the benchmark for every convex function , while maintaining coverage probability over the collection of all convex functions . in other words ,
this confidence interval has smallest expected length , up to a universal constant factor , for each and every convex function within the class of all confidence intervals which guarantee a @xmath18 coverage probability over all convex functions .
a similar result is established for a confidence interval designed for monotone functions .
the rest of the paper is organized as follows . in section [ lowerbound ] the benchmark for the expected length at each monotone function or each convex function
is established under the constraint that the interval has a given level of coverage probability over the collection of monotone functions or the collection of convex functions .
section [ procedure.sec ] constructs data driven confidence intervals for both monotone functions and convex functions and shows that these confidence intervals maintain coverage probability and have expected length within an absolute constant factor of the benchmark given in section [ lowerbound ] for each monotone function and convex function .
section [ regression.sec ] considers the nonparametric regression model , and section [ discussion.sec ] discusses connections of our results with other work in the literature .
proofs are given in section [ proof.sec ] .
as mentioned in the , the focus in this paper is the construction of confidence intervals which have expected length that adapts to the unknown function .
the evaluation of these procedures depends on lower bounds which are given here in terms of a local modulus of continuity first introduced by @xcite in the context of point estimation of convex functions under mean squared error loss .
these lower bounds provide a natural benchmark for our problems .
we focus in this paper on estimating the function @xmath5 at @xmath33 since estimation at other points away from the boundary is similar . for a given function class @xmath34 , write @xmath35 for the collection of all confidence intervals which cover @xmath36 with guaranteed coverage probability of @xmath37 for all functions in @xmath34 . for a given confidence interval @xmath7 , denote by @xmath9 the length of @xmath7 and @xmath38 the expected length of @xmath7 at a given function @xmath5 .
the minimum expected length at @xmath5 of all confidence intervals with guaranteed coverage probability of @xmath37 over @xmath34 is then given by=-1 @xmath39=0 a natural goal is to construct a confidence interval with expected length close to the minimum @xmath40 for every @xmath41 while maintaining the coverage probability over @xmath34 . however although @xmath40 is a natural benchmark for the expected length of confidence intervals , it is not easy to evaluate exactly . instead as a first step toward our goal , we provide a lower bound for the benchmark @xmath40 in terms of a local modulus of continuity @xmath42 introduced by @xcite . the local modulus is a quantity that is more easily computable and techniques for its analysis are similar to those given in @xcite and @xcite where a global modulus of continuity was introduced in the study of minimax theory for estimating linear functionals .
see the examples in section [ example.sec ] . for a parameter space @xmath43 and function @xmath44 ,
the local modulus of continuity is defined by @xmath45 where @xmath46 is the @xmath47 function norm .
the following theorem gives a lower bound for the minimum expected length @xmath40 in terms of the local modulus of continuity @xmath42 . in this theorem and throughout the paper we write @xmath48 for the cumulative distribution function and @xmath49 for the density function of a standard normal density and set @xmath50 .
[ el.bound.thm ] suppose @xmath34 is a nonempty convex set .
let @xmath51 and @xmath52 .
then for confidence intervals based on ( [ w.model ] ) , @xmath53 in particular , @xmath54 the lower bounds given in theorem [ el.bound.thm ] can be viewed as benchmarks for the evaluation of the expected length of confidence intervals when the true function is @xmath5 for confidence intervals which have guaranteed coverage probability over all of @xmath34 .
the bound depends on the underlying true function @xmath5 as well as the parameter space @xmath34 .
the bounds from theorem [ el.bound.thm ] are general . in some settings
they can be used to rule out the possibility of adaptation , whereas in other settings they provide bounds on how much adaptation is possible . in particular
the result ruling out adaptation over lipschitz classes mentioned in the easily follows from this theorem .
for example , consider the lipschitz class @xmath23 and suppose that @xmath5 is in the interior of @xmath19 .
straightforward calculations similar to those given in section [ example.sec ] show that @xmath55 now consider two lipschitz classes @xmath56 and @xmath57 with @xmath58 .
a fully adaptive confidence interval in this setting would have guaranteed coverage of @xmath37 over @xmath59 and maximum expected length over @xmath60 of order @xmath61 for @xmath62 and @xmath63 .
however , it follows from theorem [ el.bound.thm ] and ( [ lip.modulus ] ) that for all confidence intervals with coverage probability of @xmath37 over @xmath64 , for every @xmath65 with @xmath66 , @xmath67 for some constant @xmath68 not depending on @xmath5 . in particular
this holds for all @xmath69 and hence @xmath70 therefore it is not possible to have confidence intervals with adaptive expected length over two lipschitz classes with different smoothness parameters . in the present paper theorem [ el.bound.thm ]
will be used to provide benchmarks in the setting of shape constraints .
denote by @xmath71 and @xmath72 , respectively , the collection of all monotonically nondecreasing functions and the collection of all convex functions on @xmath73 $ ] .
we shall now show that in these cases the modulus and the associated lower bounds vary significantly from function to function .
we now turn to the application of the lower bound given in theorem [ el.bound.thm ] in the case of monotone functions and convex functions . here
we shall evaluate the lower bound for four particular families of functions yielding different rates at which the expected length decreases to zero as the noise level decreases in contrast to the situation just described where the parameter space did not have an order constraint .
two of the functions will be both monotonically nondecreasing and convex . in this case
the lower bound can also be quite different depending on whether we assume the knowledge that @xmath5 is convex or monotonically nondecreasing .
the key quantity that is needed in any application of theorem [ el.bound.thm ] is the local modulus .
we follow the same approach as given in @xcite where a global modulus of continuity is considered for minimax estimation . in each case , for a given function @xmath5 , we first minimize the @xmath74 norm between a function @xmath75 and the function @xmath5 subject to the constraint that @xmath76 for some given value @xmath77 . from here
it is easy to invert and thus maximize @xmath78 given a constraint on the @xmath74 norm between @xmath5 and @xmath79 .
[ linear.ex ] as a first example consider the linear function @xmath80 where @xmath81 is a constant .
this function is both monotonically nondecreasing and convex .
first consider the collection of monotonically nondecreasing functions @xmath71 .
we shall treat separately the case @xmath82 and the case @xmath83 . for the moment we shall take @xmath84 .
suppose that @xmath85 . in this case
@xmath86 and a function @xmath79 that minimizes @xmath87 subject to the constraint that @xmath88 is given by @xmath89 if @xmath90 , @xmath91 if @xmath92 , and @xmath93 if @xmath94 , where @xmath95 satisfies @xmath96 .
the assumption that @xmath97 guarantees @xmath98 .
we then have @xmath99 .
it follows that if @xmath100 @xmath101 and consequently for @xmath102 , if @xmath84 @xmath103 in the case that @xmath83 a function @xmath79 that minimizes @xmath104 subject to the constraint that @xmath105 is given by @xmath106 if @xmath90 , @xmath91 if @xmath107 . in this case
it is easy to check that @xmath108 and hence @xmath109 and hence @xmath110 we now consider the bound for the length of the confidence interval for @xmath111 belonging to the collection of convex functions . in this case
we do not need to treat the cases @xmath82 and @xmath83 separately . the function @xmath79 that minimize @xmath87 subject to the constraint that @xmath79 is convex and @xmath112 is given by @xmath113 if @xmath114 and @xmath115 if @xmath116 . in this case @xmath117 .
it then immediately follows that @xmath118 and so @xmath119 it is important to note that for @xmath82 the minimum expected lengths @xmath120 and @xmath121 are different , one of order @xmath122 and another of order @xmath123 , although the function @xmath111 is the same .
it is also interesting to note that the expected length of the confidence for monotone functions is an increasing function of @xmath124 whereas the expected length of the confidence for convex functions does not depend on @xmath124 .
since we shall show that these bounds are achievable within a constant factor it follows that the minimum expected length of the confidence interval when @xmath111 is the true function depends strongly on whether we specify that the underlying collection of functions is convex or monotone .
plots illustrating shapes of functions @xmath111 and a least favorable function @xmath79 are shown as below in figure [ fig : ex1].=1 and a least favorable function @xmath79 in example [ linear.ex ] with the constraints @xmath112 . ]
[ ex2 ] as a second example which is also both monotonically nondecreasing and convex consider the function @xmath125 where @xmath126 and @xmath127 and @xmath128 are constants .
we consider the cases @xmath129 and @xmath130 separately .
when @xmath129 the function is piecewise linear with the change of slope at @xmath33 .
in this case suppose @xmath131 .
a monotonically nondecreasing function @xmath132 that minimize @xmath133 subject to the constraint that @xmath134 is given by @xmath135 if @xmath116 , @xmath136 if @xmath137 , and @xmath135 if @xmath138 , where @xmath95 satisfies @xmath139 .
the constraint @xmath140 is to guarantee that such a @xmath95 exists with @xmath98 .
then we have @xmath141 , and it follows that if @xmath142 , @xmath143 and consequently for @xmath144 , @xmath145 we can also give a lower bound on the expected length for this same function for confidence intervals which guarantee coverage over the class of convex functions .
suppose @xmath146 . here
we need to find the convex @xmath147 that minimizes @xmath148 subject to the constraints that @xmath149 .
it is given by @xmath150 if @xmath151 , @xmath152 if @xmath153 and @xmath150 if @xmath154 .
then @xmath155 and it follows that if @xmath156 , @xmath157 hence , for @xmath158 , @xmath159 we now turn to the case where @xmath130 .
suppose @xmath160 . in this case
the monotonically nondecreasing function @xmath79 that minimizes @xmath133 subject to the constraints that @xmath134 is given by @xmath135 if @xmath116 , @xmath136 if @xmath137 and @xmath135 if @xmath138 , where @xmath95 satisfies @xmath139 .
as before the condition @xmath160 guarantees that @xmath95 exists with @xmath161 . in this case
@xmath162 for some constant @xmath163 and @xmath164 .
it follows that if @xmath165 , then @xmath166 hence , @xmath167 for a bound on the expected length of this same function for confidence intervals with coverage guaranteed over the collection of convex functions , we suppose @xmath168 . in this case
the convex function @xmath147 that minimizes @xmath148 subject to the constraints that @xmath149 , is given by @xmath169 , @xmath170 , if @xmath171 and @xmath150 otherwise , where @xmath172 and @xmath173 are the intersection points of @xmath174 and the line @xmath175 .
then the function @xmath147 with slope @xmath176 that minimize @xmath148 would be the least favorable function .
it follows that , if @xmath177 , @xmath178 and consequently for @xmath179 , @xmath180 where @xmath181 is a constant depending on @xmath182 only .
it is interesting to note that in this example the rates of convergence for @xmath183 and @xmath184 are the same for the case @xmath129 , and are different when @xmath130 .
plots illustrating shapes of functions @xmath5 and a least favorable function @xmath79 are shown as below in figure [ fig : ex2 ] . and a least favorable function @xmath79 in example [ ex2 ] with the constraints @xmath185 . ]
next we consider a function which is monotonically nondecreasing but not convex .
[ ex3 ] let @xmath186 for some constant @xmath187 and @xmath188 or @xmath189 for @xmath190 suppose that @xmath191 . in this case a function @xmath79 that minimizes @xmath192 subject to the constraint that @xmath76 is given by @xmath193 if @xmath90 , @xmath91 if @xmath92 and @xmath194 if @xmath94 , where @xmath95 satisfies @xmath195 .
as before the condition @xmath191 guarantees that @xmath95 exists with @xmath161 .
then @xmath196 , and it follows that when @xmath197 , @xmath198 hence for @xmath199 , @xmath200 and once again it is clear that the rate at which the expected length decreases to zero depends strongly on the value of @xmath182 .
as a final example we consider a function which is convex but not monotonically nondecreasing .
[ ex4 ] let @xmath201 .
suppose that @xmath202 . in this case
the function @xmath79 that minimizes @xmath192 subject to the constraint that @xmath76 is given by @xmath203 if @xmath204 , @xmath205 if @xmath206 and @xmath193 otherwise . then @xmath207 and it follows that when @xmath208 , @xmath209 hence for @xmath210 , @xmath211 a similar minimization problem is solved in dmbgen ( @xcite ) .
plots illustrating shapes of functions @xmath5 and a least favorable function @xmath79 for both examples [ ex3 ] and [ ex4 ] are shown in figure [ fig : ex3 - 4 ] . and a least favorable function @xmath79 in examples [ ex3 ] and [ ex4 ] with the constraints @xmath185 . ]
in this section we both construct and give an analysis of adaptive confidence intervals for monotone functions and convex functions .
the procedures are easily implementable .
we consider the class of monotonically nondecreasing functions and the class of convex functions .
concave functions and monotonically nonincreasing functions can be handled similarly .
the construction is split into two steps . in the first step a countable collection of confidence intervals
is created each of which has guaranteed coverage probability .
these intervals are based on a collection of pairs of linear estimators . for each interval one of the estimators
has nonnegative bias and the other nonpositive bias .
the one - sided control of the bias of these estimators is a key special feature in these problems and an important part of what makes it possible to adapt to every individual function .
moreover for each function @xmath5 this collection has at least one interval with expected length within a constant factor of the local modulus bound given in theorem [ el.bound.thm ] .
the second step is to select from this collection a particular interval . in the case of monotonically nondecreasing functions we take for each @xmath212 ,
pairs of estimators @xmath213 and @xmath214 .
then for estimating @xmath36 it is easy to check that @xmath215 has nonnegative and monotonically nonincreasing biases while @xmath216 have nonpositive and monotonically nondecreasing biases .
the one - sided control of the biases of these estimators over the class of all monotonically nondecreasing functions easily allows for the construction of a confidence interval . for that we shall need the standard deviation of @xmath215 and @xmath216 . in order to give a unified treatment in both the monotone and convex case
it is useful to establish a common notation .
here we shall set @xmath217 .
it is then easy to check that both @xmath215 and @xmath216 have a standard deviation of @xmath218 .
it is then also easy to see that for each @xmath219 , the confidence interval @xmath220 given by @xmath221\ ] ] has guaranteed coverage of @xmath37 .
we should , however , note that in ( [ mcij ] ) the left endpoint of the interval may be larger than the right endpoint in which case we adopt the convention that the confidence interval is just the empty set .
the length of this confidence interval is then @xmath222 . in the case of convex functions for @xmath223 , let @xmath224 and let @xmath225 .
the following lemma shows that for convex functions @xmath226 have nonnegative and monotonically nonincreasing biases and that @xmath227 have nonpositive and monotonically nondecreasing biases .
[ con.lem ] for any convex function @xmath5 , @xmath228 it is also easy to check that the standard deviation of @xmath226 is equal to @xmath229 where @xmath217 and that @xmath230 has a standard deviation of @xmath231 .
it then follows from the signs of the biases of @xmath232 and @xmath233 that for any given @xmath234 , @xmath235\ ] ] gives a confidence interval with coverage probability of at least @xmath236 .
we should also note once again that the left endpoint of the interval may be larger than the right endpoint in which case the confidence interval is taken to be the empty set , and so in this case the length of this confidence interval is @xmath237 .
these results , for which a more formal proof is given in section [ proof.sec ] are summarized in the following proposition .
[ m.cij.prop ] for every @xmath219 , the confidence interval @xmath238 defined in ( [ mcij ] ) has coverage probability of at least @xmath239 for all monotonically nondecreasing functions @xmath240 , and for every @xmath223 , the confidence interval @xmath241 defined in ( [ cisj ] ) has coverage probability of at least @xmath239 for all convex functions @xmath242 .
the second stage in the construction is that of selecting from these collections of intervals the one to be used .
first note that one should not select the shortest interval since the collections defined in ( [ mcij ] ) and ( [ cisj ] ) will always contain one which corresponds to the empty set .
a more sensible goal is to try to select the interval with the smallest expected length or at least one which has expected length close to the smallest expected length .
the approach we take here is to choose an interval for which the expected length is of the same order of magnitude as the standard deviation of the length .
such an interval will always have expected length close to the shortest expected length . for the case of monotonically
nondecreasing functions the selection of the interval from the countable collection in ( [ mcij ] ) can be done by creating another collection of estimators which can be used to estimate the expected length of the intervals .
more specifically set @xmath243 .
then for @xmath212 , @xmath244 s are independent of each other and both @xmath215 and @xmath216 are independent of @xmath245 for every @xmath246 .
we should note that the estimators @xmath244 are similar to @xmath247 in that they are both differences of averages of @xmath0 to the left and right of the origin and thus estimate the average local change of the function . however @xmath248 are not independent for different @xmath234 whereas the @xmath244 are independent .
it is thus natural to view the @xmath249 as a surrogate for @xmath250 with the technical advantage that they are independent .
the selection of a @xmath234 for which @xmath244 has expected value close to @xmath229 will then result in a confidence interval @xmath238 close to the one with the smallest expected length .
the independence properties of the @xmath244 allows us to guarantee a @xmath18 coverage probability while making this selection .
more specifically the construction proceeds as follows .
let @xmath251 and define the final confidence interval by @xmath252 before we turn to the analysis of this procedure we also introduce here a related confidence procedure in the convex case . here
rather than introducing an independent estimate of the difference between the two estimators used in constructing the confidence interval , we proceed more directly .
the basic idea is similar , but the dependence between the estimates of @xmath234 and the confidence interval constructed from this estimate requires that we adjust the original coverage level of our @xmath241 .
more specifically let @xmath253 .
when the expected value of @xmath254 is the same order as @xmath229 , the confidence interval @xmath241 will then be close to the one with the smallest expected length .
our estimate of @xmath234 is given by an empirical version , namely @xmath255 although this estimate can be used to select the appropriate @xmath241 to use , as just mentioned , care also needs to be taken to make sure that the resulting selected interval maintains the required coverage probability . the analysis given below shows that a choice of @xmath256 in the construction of the original collection of intervals guarantees an overall coverage probability of @xmath257 .
thus in the case of convex functions , we define our interval by @xmath258 in this section we present the properties of the confidence intervals @xmath259 and @xmath260 defined by ( [ m_ci ] ) and ( [ convex.ci ] ) focusing on the coverage and the expected length of these intervals .
we begin with the confidence interval @xmath259 . in this case
it is easy to check the coverage probability of @xmath259 by the independence of the interval @xmath238 and @xmath245 for every @xmath124 satisfying @xmath261 .
the key to the analysis of the expected length is the introduction of @xmath262 where @xmath263 the analysis of the expected length relies on showing that @xmath264 is highly concentrated around @xmath262 .
the concentration of @xmath264 around @xmath262 then provides a bound on the expected length of @xmath265 .
these results , for which a proof is given in section [ proof.sec ] are summarized in the following theorem .
[ m.ci-el.thm ] let @xmath266 .
the confidence interval @xmath259 defined in ( [ m_ci ] ) has coverage probability of at least @xmath267 for all monotonically nondecreasing functions @xmath240 and satisfies @xmath268 where @xmath269 is a constant and can be taken to be @xmath270 for all @xmath271 .
[ rem1 ] the constant @xmath269 in theorem [ m.ci-el.thm ] depends on the upper limit of @xmath257 .
@xmath269 can be smaller if the upper limit on @xmath257 is reduced .
for example , for common choices of @xmath272 or 0.01 , @xmath273 for @xmath274 , and @xmath275 for @xmath276 .
theorem [ m.ci-el.thm ] shows that the coverage probability is attained and also provides an upper bound on the expected length in terms of @xmath277 . in order to establish that this expected length is within a constant factor of the lower bound given in theorem [ el.bound.thm ]
, we need to provide a lower bound for @xmath278 in terms of @xmath279 .
this connection is given in the following theorem .
[ mel_lb ] let @xmath280 and let @xmath281
. then @xmath282 combining theorems [ m.ci-el.thm ] and [ mel_lb ] , we have @xmath283 for all monotonically nondecreasing functions @xmath240 , where @xmath284 is a constant depending on @xmath257 only . for example , @xmath284 can be taken to be @xmath285 for @xmath272 and @xmath286 for @xmath287 . hence , the confidence interval @xmath259 is uniformly within a constant factor of the benchmark @xmath288 for all monotonically nondecreasing functions @xmath5 and all confidence level @xmath289 .
we now turn to an analysis of the properties of the confidence interval @xmath260 defined in ( [ convex.ci ] ) .
the key to this analysis is the introduction of @xmath290 where @xmath291 the analysis of both the coverage probability and the expected length relies on showing that @xmath264 is highly concentrated around @xmath290 .
the probability of not covering @xmath36 can be bounded by @xmath292 \\[-8pt ] \nonumber & & { } + \sum _ { l=-2}^{2 } p\bigl ( f(0 ) \notin \mathrm{ci}_{j_*^c + l}\bigr).\end{aligned}\ ] ] the first two terms are controlled by the high concentration of @xmath293 around @xmath290 , and the last term is controlled by proposition [ m.cij.prop ] which bounds the coverage probability of any given @xmath234 .
the concentration of @xmath264 around @xmath290 also allows control on the expected length of @xmath260 which leads to the following theorem .
[ convex.ci-el.thm ] let @xmath266 .
the confidence interval @xmath260 defined in ( [ convex.ci ] ) has coverage probability of at least @xmath267 for all convex @xmath5 and satisfies @xmath294 where @xmath269 is a constant and can be taken to be @xmath295 for all @xmath271 .
[ rem2 ] the constant @xmath269 in theorem [ convex.ci-el.thm ] depends on the upper limit of @xmath257 .
@xmath269 can be smaller if the upper limit on @xmath257 is reduced .
for example , for common choices of @xmath272 or 0.01 , @xmath296 for @xmath274 , and @xmath275 for @xmath276 .
theorem [ convex.ci-el.thm ] shows that the coverage probability is attained and also provides an upper bound on the expected length in terms of @xmath297 . as was the case for monotone functions , in order to to establish that this expected length for convex functions is within a constant factor of the lower bound given in theorem [ el.bound.thm ] , we need to provide a lower bound for @xmath298 in terms of @xmath299 .
this connection is given in the following theorem .
[ el.bound2.thm ] let @xmath280 and let @xmath300 .
then @xmath301 theorems [ convex.ci-el.thm ] and [ el.bound2.thm ] together yield @xmath302 for all convex functions @xmath242 , where @xmath303 is a constant depending on @xmath257 only . for example , @xmath303 can be taken to be @xmath304 for @xmath272 and @xmath305 for @xmath287 .
hence , the confidence interval @xmath260 is uniformly within a constant factor of the benchmark @xmath306 for all convex functions @xmath5 and all confidence levels @xmath289 .
we have so far focused on the white noise model . the theory presented in the earlier sections
can also easily be extended to nonparametric regression .
consider the regression model @xmath307 where @xmath308 and @xmath309 and where for notational convenience we index the observations from @xmath310 to @xmath311 .
note that the noise level @xmath312 can be accurately estimated easily , as in hall , kay and titterington ( @xcite ) or @xcite .
see also @xcite .
we shall thus assume it is known in this section .
then under the assumption that @xmath5 is convex or monotone , we wish to provide a confidence interval for @xmath36 .
let @xmath313 .
for @xmath314 define the local average estimators @xmath315 we should note that the indexing scheme is the reverse of that given for the white noise with drift process . here
estimators @xmath316 ( or @xmath317 ) with small values of @xmath234 have smaller bias ( or larger bias ) and larger variance than those with larger values of @xmath234 . as in the white noise model
it is easy to check that @xmath316 has nonnegative bias and @xmath317 has nonpositive bias .
simple calculations show that the variance of @xmath318 and @xmath319 are both @xmath320 , where @xmath321 .
it is also important to introduce @xmath322 as in the white noise case , where @xmath323 .
it is easy to check that @xmath324 , @xmath325 s are independent with each other , and both @xmath318 and @xmath326 are independent with @xmath327 for every @xmath328 .
it then follows that @xmath329 $ ] has guaranteed coverage probability of at least @xmath267 over all monotonically nondecreasing functions .
now set @xmath330 and define the confidence interval to be @xmath331 the properties of this confidence interval can then be analyzed in the same way as before and can be shown to be similar to those for the white noise model .
in particular , the following result holds .
[ mel_reg.thm ] let @xmath332 .
the confidence interval @xmath333 defined in ( [ reg.mci ] ) has coverage probability of at least @xmath267 for all monotone functions @xmath5 and satisfies @xmath334 for all monotonically nondecreasing functions @xmath240 , where @xmath335 is a constant depending on @xmath257 only . as in the monotone case , set @xmath313 .
for @xmath336 define the local average estimators @xmath337 we should note that this indexing scheme is the reverse of that given for the white noise with drift process . here estimators @xmath338 with small values of @xmath234 have smaller bias and larger variance than those with larger values of @xmath234 . as in the white noise model it is easy to check that @xmath338 has nonnegative bias .
it is also important to introduce an estimate which has a similar variance but is guaranteed to have nonpositive bias .
the key step is to introduce @xmath339 as an estimate of the bias of @xmath338 .
the following lemma gives the required properties of @xmath338 and @xmath254 .
[ reg.bias.lem ] for any convex function @xmath5 , @xmath340 from ( [ reg.bias.bound2 ] ) it is clear that the biases of the estimators @xmath341 are nonnegative and monotonically nondecreasing . in addition straightforward calculations using both ( [ reg.bias.bound1 ] ) and ( [ reg.bias.bound2 ] ) show that the estimators @xmath342 have a nonpositive and monotonically nonincreasing biases .
simple calculations show that the variance of @xmath343 is @xmath344 .
it then follows that @xmath345 $ ] has coverage over all convex functions .
now set @xmath346 and define the confidence interval to be @xmath347 this confidence interval shares similar properties as the one for the white noise model . in particular , the following result holds .
[ cel_reg.thm ] let @xmath332 .
the confidence interval @xmath260 defined in ( [ reg.cci ] ) has coverage probability of at least @xmath267 for all convex function @xmath5 and satisfies @xmath348 for all convex function @xmath242 , where @xmath349 is a constant depending on @xmath257 only .
the major emphasis of the paper has been to show that with shape constraints it is possible to construct confidence intervals that have expected length that adapts to individual functions . in this section
we shall discuss briefly the maximum expected lengths of our procedures over lipschitz classes that are either monotone or convex in a way that is similar to that provided in dmbgen ( @xcite ) for the maximum width of a confidence band .
we shall also explain how our results can be extended to the problem of estimating the value of @xmath5 at points other than @xmath33 .
although the focus of the present paper has been on the construction of a confidence interval with the expected length adaptive to each individual convex or monotone function , these results do yield immediately adaptive minimax results for the expected length in the conventional sense .
define @xmath350 the following results are direct consequence of theorems [ m.ci-el.thm ] and [ convex.ci-el.thm ] .
the confidence interval @xmath259 defined in ( [ m_ci ] ) satisfies @xmath351 simultaneously for all @xmath352 and @xmath353 , for some absolute constant @xmath335 .
the confidence interval @xmath260 defined in ( [ convex.ci ] ) satisfies @xmath354 simultaneously for all @xmath355 and @xmath353 , for some absolute constant @xmath349 .
we should note that these ranges of lipschitz classes are the only ones of interest in these cases . in particular
suppose that @xmath7 is a confidence interval with guaranteed coverage over the class of monotonically nondecreasing functions .
then for any @xmath13 the class @xmath19 includes the linear function @xmath356 . as shown in example [ linear.ex ] in section [ example.sec ] , @xmath103 hence , @xmath357 \\[-8pt ] \nonumber & = & \sup_{k}\biggl ( 1 - \frac{1}{\sqrt{2\pi } z_{\alpha } } \biggr ) ( 3k)^{{1}/{3}}z^{2/3}_{\alpha}n^{-{1/3 } } = \infty.\end{aligned}\ ] ] a similar results holds for convex functions assumed to belong to @xmath19 with @xmath358 . on the other hand suppose
@xmath5 is convex and assumed to belong to @xmath19 with @xmath359 .
then from the assumption that @xmath5 is in @xmath19 it follows that @xmath360 .
convexity then shows that @xmath361 and the maximum expected length over this class is given above .
the focus of the present paper has been on the problem of estimating the value of @xmath36 .
the basic development is similar for any other point @xmath362 in the interior of the interval @xmath363 $ ] unless @xmath362 is near to the boundary .
more specifically for any @xmath364 we can consider estimators @xmath365 and @xmath366 where @xmath367 for monotone functions and @xmath368 where @xmath369 for convex functions .
the basic theory is the same as before . for monotonically nondecreasing functions ,
the confidence interval @xmath238 is replaced by @xmath370\ ] ] and the choice of @xmath264 is given by @xmath371 where @xmath372 .
the final confidence interval is defined by @xmath373 for convex functions , the confidence interval @xmath241 is replaced by @xmath374,\ ] ] and @xmath264 is chosen to be @xmath375 define the final confidence interval by @xmath376 the modulus of continuity defined in ( [ l.modulus ] ) is replaced by @xmath377 the earlier analysis then yields @xmath378 and @xmath379 where we now have @xmath380 finally we should note that at the boundary the construction of a confidence interval must be unbounded .
for example any honest confidence interval for @xmath381 must be of the form @xmath382 ; otherwise it can not have guaranteed coverage probability .
we prove the main results in this section .
we shall omit the proofs for theorems [ mel_reg.thm ] and [ cel_reg.thm ] as they are analogous to those for the corresponding results in the white noise model .
set @xmath383 .
now note that @xmath384 is convex in @xmath385 for all @xmath386 .
hence @xmath387 is also convex with @xmath388
. for @xmath389 set @xmath390 , and it follows that @xmath391 .
equation ( [ con.bias.eq ] ) follows from the fact that @xmath392 for @xmath393 , and equation ( [ con.eq ] ) follows from the fact that @xmath394 . for any convex function @xmath5 ,
let @xmath395 .
then @xmath396 is convex , increasing in @xmath397 and @xmath398 .
convexity of @xmath399 yields that for @xmath400 , @xmath401 note that @xmath402 and @xmath403 so @xmath404 is equivalent to @xmath405 which is the same as @xmath406 now note that for @xmath407 and @xmath408 , @xmath409 and consequently @xmath410 then ( [ aaaaa ] ) follows by taking @xmath411 and @xmath412 and then summing over @xmath413 .
suppose that @xmath422 where it is known that @xmath423 $ ] .
the confidence interval for @xmath424 which has guaranteed coverage over the interval @xmath423 $ ] and which minimizes the expected length when @xmath425 is given by @xmath426.\ ] ] now for @xmath441 , let @xmath442 .
let @xmath434 be this collection of @xmath443 .
now for the process @xmath444 there is a sufficient statistic @xmath445 given by @xmath446 note that @xmath445 has a normal distribution @xmath447 or more specifically @xmath448 . for monotone functions
, we have @xmath452 where @xmath453 is a standard normal random variable . because @xmath454 and @xmath455 , we have @xmath456 for convex functions , let @xmath457 .
it follows from lemma [ con.lem ] that @xmath458 , and hence we have @xmath459 note that @xmath460 because both @xmath215 and @xmath216 are independent of @xmath245 for @xmath246 , and the event @xmath461 depends only on @xmath245 for @xmath246 , then by proposition [ m.cij.prop ] we have @xmath462 we now turn to the upper bound for the expected length .
note that for @xmath463 , @xmath464 , and so we have @xmath465 it follows from @xmath466 that @xmath467 , and hence we have @xmath468 thus @xmath469 set @xmath470 for @xmath471 .
then it is easy to see that @xmath472 thus @xmath473 the right - hand side is increasing in @xmath257 . through numerical calculations
, we can see that , for @xmath474 , @xmath475 thus , by equation ( [ b8 ] ) , we have @xmath476 note that if @xmath477 , then @xmath478 and hence there is a @xmath479 such that we have either @xmath480 or @xmath481 . if @xmath482 , let @xmath483 and if @xmath484 , let @xmath485 then we have @xmath486 if @xmath487 , let
@xmath488 then we have @xmath489 it then follows that @xmath490 and so @xmath491 to bound the coverage probability note that @xmath500 \\[-8pt ] \nonumber & & { } + \sum_{k = -2}^{2 } p\bigl(f(0 ) \notin \mathrm{ci}_{j_*^c + k}\bigr).\end{aligned}\ ] ] it then follows from equation ( [ too.early ] ) that @xmath501 for all @xmath266 .
it is easy to verify directly that for all @xmath502 , @xmath503 .
furthermore , it is easy to see that for @xmath494 , @xmath504 and so @xmath505 hence , @xmath506 we now turn to the upper bound for the expected length .
note that @xmath510 hence @xmath511 \\[-8pt ] \nonumber & & { } \times\biggl(p\bigl ( \hat j \le j_*^c\bigr ) + \sum _ { k=1}^{\infty } 2^{k/2 } p\bigl(\hat j = j_*^c + k\bigr ) \biggr).\end{aligned}\ ] ] set @xmath512 for @xmath471 .
then it is easy to see that @xmath513 it then follows from ( [ too.far ] ) that @xmath514 the right - hand side is clearly increasing in @xmath257 .
direct numerical calculations show that for @xmath474 , @xmath515 it then follows directly from ( [ el ] ) that @xmath516 note that if @xmath492 , then @xmath517 , and hence there is a @xmath518 satisfying @xmath519 such that @xmath520 , where @xmath521 .
let @xmath79 be defined by @xmath522 there is also a @xmath79 as in the proof of lemma 5 in our other paper with @xmath523 for which @xmath524 if @xmath525 , then let @xmath526 , and then we have @xmath527 it then follows that @xmath528 and so @xmath529 | adaptive confidence intervals for regression functions are constructed under shape constraints of monotonicity and convexity .
a natural benchmark is established for the minimum expected length of confidence intervals at a given function in terms of an analytic quantity , the local modulus of continuity . this bound depends not only on the function but also the assumed function class .
these benchmarks show that the constructed confidence intervals have near minimum expected length for each individual function , while maintaining a given coverage probability for functions within the class .
such adaptivity is much stronger than adaptive minimaxity over a collection of large parameter spaces . , |
recent studies on the passage of dna through narrow channels , apart from being very interesting in their own right , have been motivated by the exciting possibility of developing a practical technique to characterize and sequence dna .
these single molecule experiments were pioneered by kasianowicz , brandin , branton and deamer @xcite using a protein called @xmath7-hemolysin as a pore embedded in a lipid membrane . when voltage is applied across the membrane , ion current can be detected .
when the dna chain is inside and blocking the channel , the current is suppressed . by measuring the blockage time
one can gather information about the voltage driven dynamics of dna translocation .
a large amount of exciting data have been accumulated through such experiments @xcite .
recently , another group of experiments was performed using solid state nano - pores for voltage - driven dna translocation @xcite . on the theoretical side ,
much effort was placed in understanding the entropic barrier associated with translocation of a long polymer @xcite .
the electrostatic barrier associated with charged dna penetration through the low dielectric constant membrane has been recently discussed theoretically @xcite .
another interesting question is about friction : whether friction on the small piece of dna passing through the narrow channel is larger or smaller than the friction experienced by the large dna coils outside the membrane . to this end ,
lubensky and nelson @xcite , assuming the channel friction dominance , were able to account for the observed bimodal distribution of the passage times for single - stranded dna . in their model
, each base preferentially tilts towards one end of the single stranded dna chain ( @xmath8 or @xmath9
. see figure [ fig : dnaporecases ] . ) .
each peak in the distribution of passage times therefore corresponds to a particular end entering the channel first , the channel being just wide enough for a single strand to pass through .
more recently , kantor and kardar @xcite and then storm et al @xcite argued that for sufficiently long dna there must be a crossover to the regime of domination of the out - of - channel friction , in which case translocation should become subdiffusive in character ( with displacement growing slower than @xmath10 with time @xmath11 ) .
a new spin is added to the story by the experiment by meller and coworkers @xcite .
these authors devised an experiment involving dna having a string of identical bases ( adenine ) in the single - stranded portion and a hairpin loop at one end ( figure [ fig : dnaporecases ] ) held in place by bonding of complementary bases .
like double - stranded dna , the hairpin can not enter the channel ( a transmembrane pore of @xmath7-hemolysin ) .
the hairpin therefore constrains the dna to enter the pore with its single - stranded end , as well as preventing the entire dna from crossing the membrane . in their experiment , the dna , driven by an applied voltage , enters the pore with its single - stranded end .
thereafter , once current is blocked by the dna , the voltage is either switched off , in which case the dna diffuses freely ( non - driven ) , or the sign of the voltage is flipped , in which case the dna is pushed back . moreover , by making two dna samples , with the hairpin loop at opposite ends , it is possible to observe dna sliding away from the pore in two opposite directions along the dna contour , and the observation suggests that dna escapes in one direction faster than in the other . in the experiment @xcite , by measuring the so - called `` survival probability '' @xmath12 , which is the probability that a dna molecule will stay in the pore as a function of the waiting time , it was determined that the voltage - free dynamics of the @xmath9 threaded molecules is about two times slower than the corresponding diffusion of @xmath8 threaded molecules having the same sequence .
importantly , in both cases the dna was threaded from the same side of the pore ( called the @xmath13-side of @xmath7-hl ) . to delineate the underlying mechanism responsible for the observed dynamics , the authors of the work @xcite performed all - atom molecular dynamics simulations , which independently confirmed the experimental results for driven dna .
the simulations also showed that the confinement of the dna bases in the @xmath7-hl pore results in an even stronger ( compared to a free dna ) tilt of the bases with respect to the dna backbone towards the @xmath8 end .
authors of the work @xcite phenomenologically interpret their data by assigning two different diffusion constants for the two separate experiments in which the same dna is placed in the channel in two possible orientations .
this interpretation is justified by the fact that the interactions between the dna bases and the pore are different in these two cases ( perhaps via different barrier heights within the framework of a sawtooth potential landscape discussed below ) .
there is a temptation to summarize the experimental findings of the work @xcite in one sentence ( although no one made this mistake , including @xcite ) : dna diffuses in one direction faster than in the other .
indeed , the observed asymmetry of dynamics is consistent with the tilt of the nucleotides with respect to the main dna chain .
this asymmetry then seems easy to understand if the analogy is made with petting a cat along or against the grain of its fur ; the cat responds very differently in the two cases ( presumably because it experiences very different friction ) .
another , possibly even more obvious , analogy would be carrying a christmas tree top first or base first through a narrow door ; one again encounters very different resistance in the two cases .
the point is that such analogies and interpretations are only possible for the driven dna motion , particularly for the system far into the non - linear regime ( in terms of force - velocity relation ) , whereas for the portion of experiments in the work @xcite involving freely diffusing dna such analogies and interpretations would be _ wrong _ ; it is not surprising then that the authors of the work @xcite did not use such analogies and interpretation for the freely diffusing dna .
indeed , for free diffusion , the friction coefficient ( averaged over the scale well exceeding a single base ) moving in one direction and in the opposite direction must be the same , as follows from the onsager symmetry relation , and the assumption of asymmetric friction would be a grave mistake . although no one actually made this mistake , including @xcite , it is worth emphasizing why an assumption of asymmetric friction would be a mistake . indeed , if we only imagine that dna ( not driven by any applied voltage ! ) diffuses in one direction faster than in the other , then we can easily build a _
perpetuum mobile _
( see figure [ fig : perpetuum_mobile ] ) moving indefinitely long through time at the expense of thermal energy from the thermal bath , which is , of course , impossible .
in other words , freely diffusing dna , when it is already in the pore , in contrast to a ( heavily driven ! ) christmas tree through a door , must have the same friction coefficients when the dna moves in either direction .
what is nice is that the experimental findings and their interpretation in the work @xcite are in fact in perfect agreement with this thermodynamic analysis . in order to make this reconciliation very clear
, we immediately refer to the symmetry analysis in figure [ fig : dnaporecases ] .
notice that the pore itself is asymmetric ( its crystallographic structure is known @xcite ) , the dna backbone is also asymmetric ( from @xmath9 end to @xmath8 end ) , and the loopy end creates further asymmetry .
this gives four possible orientations of the pore and the dna with the loop : two possibilities arise from two different mutual orientations of the dna backbone with respect to the pore ( indicated by the numbers 1 and 2 in figure [ fig : dnaporecases ] ) , and for each of these two orientations there are two possibilities to place the blocking loop ( indicated by the letters a and b in figure [ fig : dnaporecases ] ) .
this symmetry analysis , as shown in figure [ fig : dnaporecases ] , is reminiscent of the symmetry analysis in the paper @xcite , except we have no electric field , but instead have loops at the dna ends .
we can now say that in any one of the arrangements , from 1a , 1b , 2a or 2b , the dna must experience the same friction moving up or down the pore ; friction going up equals friction going down . at the same time
, the friction in configurations 1a or 1b can be different from friction in configurations 2a or 2b , and they are likely to be different .
that is why the work @xcite assigns two different diffusion constants to the two dna - pore mutual configurations ( 1a and 2a ) .
by contrast , the loop itself likely has no effect on the friction or diffusion coefficient , so we expect that the diffusion coefficient should be the same for configurations 1a and 1b ( same goes for 2a and 2b ) . in other words
, there should be two distinct diffusion coefficients , not four .
we shall argue in this work that , nevertheless , there will be four different diffusion times corresponding to the four configurations in figure [ fig : dnaporecases ] . to explain our approach , it is convenient to adopt a terminology in which , instead of considering diffusion of the dna chain , we consider diffusion of the passage point along the dna contour . following lubensky and nelson @xcite
, we consider a simple model in which asymmetry is presented in the underlying potential landscape . for simplicity , we model it with a sawtooth profile .
the two orientations of the asymmetric potential ( relative to the boundary conditions ) correspond to the two possible placements of the blocking loop for a given orientation between dna and pore ( e.g. 1a and 1b in figure [ fig : dnaporecases ] ) . like lubensky and nelson @xcite
, we focus on the first passage time , which is the time it takes for the initially fully-`plugged ' dna to completely ` unplug ' from the pore . in other words
, it is the time needed for the diffusing particle ( or a random walker ) to arrive for the first time on the open end of the dna , or to one end of the @xmath14 interval , provided that a reflecting boundary condition is imposed at the opposite end .
we would like to emphasize the fundamental difference between asymmetric diffusion , which is prohibited by thermodynamics , and symmetric diffusion over the asymmetric potential landscape .
it is well known , and we show it explicitly in appendix [ sec : steadystate ] , that stationary diffusion remains symmetric despite the asymmetry of the underlying potential landscape , thus making nonfunctional the _ perpetuum mobile _ design of figure [ fig : perpetuum_mobile ] . in this paper , we compute the mean first passage times ( mfpt ) corresponding to cases 1a and 1b in figure [ fig : dnaporecases ] ( or to cases 2a and 2b ) .
we consider the brownian motion of a particle diffusing classically in an asymmetric sawtooth potential @xmath15 and in the inverted or reversed version of the potential ( figure [ fig : sawtooth ] ) .
this model neglects the entropic barrier ( of order @xmath16 ) presented by the dna coils on both sides of the pore @xcite , but through the consideration of subdiffusion it does take into account the extra friction created by those coils @xcite . from the results , we discuss when the difference between the two times is significant .
( note that since we know very little about the details of the interactions between the dna bases and the pore , we can not determine if case a in figure [ fig : sawtooth ] corresponds to case 1a in figure [ fig : dnaporecases ] , and case b in figure [ fig : sawtooth ] corresponds to case 1b in figure [ fig : dnaporecases ] , or if it is the other way around . ) since dna translocation is ultimately not classical diffusion , but rather subdiffusion @xcite , we consider also the first passage times for the subdiffusion in the presence of an asymmetric potential . in general , the first passage time for subdiffusion was recently a matter of considerable interest and dispute in the literature @xcite .
it is now understood @xcite that the _ mean _ first passage time diverges for subdiffusion , because a subdiffusing walker tends to remain too long on the place that it once reached .
accordingly , we look at the probability distribution for the first passage times ( dfpt ) , and concentrate on its tail at long times . we found that this tail is very different for the two potentials , and the difference turns out to be expressed through corresponding mean first passage times for classical diffusion . with this knowledge
, we construct an average first passage time from a subset of passage events and show that this average scales as @xmath6 .
the result also exhibits the asymmetry between cases 1a and 1b ( or 2a and 2b ) just as in the case of classical diffusion . from our discussion
we make the prediction that the observed passage times for the four possible mutual orientations of the pore and the dna will all be different .
for treating classical or normal diffusion , one often starts with the fokker - planck ( fpe ) equation ( in this context also frequently called smoluchowsky equation ) @xmath17 giving the time - evolution of the probability density @xmath18 . here
@xmath19 is the usual diffusion constant and we have set @xmath20 .
the fpe yields the boltzmann distribution for @xmath21 in the steady - state , as well as giving the linear relation between the mean - squared displacement and time in the absence of external forces . in calculations involving the first passage time
, it would be convenient to consider the equivalent problem of first passage to either @xmath22 or @xmath23 , where the potential @xmath15 for @xmath24 is as illustrated in figure [ fig : sawtooth ] , while the potential for @xmath25 is @xmath15 for positive @xmath26 reflected about the vertical axis . with this picture , the probability for the particle to still be ` alive ' at time @xmath11 ,
also called the survival probability @xmath12 ( and measured in experiment @xcite ) , is given by @xmath27 .
the distribution of first passage times @xmath28 is calculated from @xmath12 via @xmath29 .
this gives the following expression for the mean first passage time @xmath30 @xcite : @xmath31 where @xmath32 is the initial position of the particle , @xmath33 .
it can be shown that @xmath30 satisfies an ordinary differential equation @xcite ( derived in appendix [ sec : equation_for_tau ] ) .
the solution of this differential equation for a sawtooth potential @xmath15 is outlined in appendix [ sec : solution ] . for the particle initially located at the origin ( @xmath34 ) , the mean first passage time to reach @xmath22
is given by , @xmath35 for the potential in figure [ fig : sawtooth]a , and @xmath36 for the potential in figure [ fig : sawtooth]b .
here , we have defined the coefficients @xmath37 expression ( [ tauclassic1 ] ) can be obtained from ( [ tauclassic0 ] ) by flipping the sign of @xmath38 . from the results ( [ tauclassic0 ] ) and ( [ tauclassic1 ] ) ,
it is clear that @xmath39 .
physically , the inequality @xmath40 may be obvious for the case @xmath41 in figure [ fig : sawtooth ] , in which a particle has to surmount a single barrier in order to get to @xmath42 in case b , while there is no barrier in case a. in general , for a given @xmath3 , the potential in figure [ fig : sawtooth]a involves @xmath43 barriers , while the potential in figure [ fig : sawtooth]b involves @xmath3 barriers .
in fact , it is easy to show that in the limit @xmath44 ( @xmath45 in more conventional units ) we have @xmath46 , where the arguments indicate the number of teeth in the sawtooth potentials . for the long dna , when @xmath47 or @xmath48 , the leading terms in both @xmath49 ( [ tauclassic0 ] ) and @xmath50 ( [ tauclassic1 ] ) are proportional to @xmath4 , as one would expect for diffusion times .
to this leading order , first passage times @xmath49 and @xmath50 obey the symmetry in diffusion and are the same .
it is in the subleading terms ( proportional to @xmath51 ) that the two times differ .
let us stress that the difference between @xmath49 and @xmath50 , which is of order of @xmath52 in a relative sense , is entirely due to the boundary conditions and the situation at the ends of the diffusion region .
anomalous diffusion is characterized by the occurrence of a mean square displacement of the form @xmath1 , where @xmath53 in subdiffusion ; traditionally @xcite , this is written in the form @xmath54 where @xmath55 is a generalized diffusion constant and @xmath56 is the gamma function . for @xmath57
one recovers the usual result for classical diffusion .
it can be shown @xcite that this form for the mean square displacement can be obtained from a generalized version of equation ( [ fpe ] ) called the fractional fokker - planck equation ( ffpe ) .
this equation is described in appendix [ sec : fractional_ffpe ] .
although up to this point we have ignored the interactions of the dna bases outside the pore , it seems reasonable to speculate that their effect is to slow down the translocation .
thus one might be able take these interactions into account phenomenologically by positing a value of @xmath58 corresponding to the subdiffusive domain @xmath59 .
( reference @xcite lists possible sources of waiting time distributions leading to anomalous diffusion ) .
it is shown below and in references @xcite that the mfpt does not exist for subdiffusion .
this leads us to consider the probability distributions themselves .
the method of laplace transforms can be used to solve for the transform of the survival probability @xcite but one is left with the very difficult task of obtaining the inverse transform , even for the case of a sawtooth with @xmath41 .
however , it is shown in appendix [ sec : dfpt_mfpt ] that the long - time limit of the survival probability and the first passage time distribution scales as some power of @xmath11 and that they are simply related to the expression for the mfpt in the context of classical diffusion as follows @xmath60 @xmath61 here @xmath62 is the same expression as the mfpt in classical diffusion ( [ tauclassic0 ] ) ( or ( [ tauclassic1 ] ) ) , but containing a generalized diffusion constant .
the long - time limit is reached when @xmath63 .
the relationships ( [ soft ] ) and ( [ foft ] ) ultimately arise from the almost identical expressions for the solution @xmath18 in classical diffusion and in subdiffusion .
the two solutions differ only in the time dependence , which is an exponential for classical diffusion .
because the expectation time of a first passage is infinite , any meaningful experiment , real or computational alike , must be based on some protocol rendering the observation time finite .
we argue that in essence such a protocol is always reduced to discarding the events which fail to come to completion within some specified time @xmath5 ; in other words , only those passage events that each complete within some time @xmath5 are counted . the rest of the events that do not end by time @xmath5 are terminated and discarded .
conditional _ probability distribution of the first passage events that get counted under such a protocol is then given by @xmath64 for such an experiment , there exists a perfectly defined and finite average first passage time .
this _ conditional _
average , for large @xmath5 , is @xmath65 so far , the time @xmath5 should be long enough , but otherwise arbitrary .
now we argue that the time @xmath5 must be chosen such that roughly about half of the passage events at a given @xmath3 get discarded .
this requirement seems reasonable , for if one discards a much smaller fraction @xmath5 becomes too large and the measurements get inefficiently slow ; if one discards a much larger fraction @xmath5 becomes too small and the tail of the distribution does not get sampled properly .
thus , assuming half of the events discarded , @xmath5 becomes of order @xmath66 , just at the boundary of the validity of the asymptotics . substituting this into ( [ conditionalaverage ] ) , one obtains a scaling of @xmath67 for the average first passage time
of course , this scaling is not unexpected for subdiffusion with an average displacement going like @xmath68 .
furthermore , due to the appearance of the classical diffusion times @xmath49 and @xmath50 ( which take the place of @xmath69 depending on the potential ) in the average first passage time we just defined , the asymmetry of the first passage time is once again present in this case .
the ratio of the mfpts in classical diffusion , expressions ( [ tauclassic0 ] ) and ( [ tauclassic1 ] ) , is plotted in figure [ fig : tauratioclassic ] for a few realistic values of @xmath3 and @xmath38 .
we see that for @xmath38 equal to a few @xmath70 , the difference becomes small ( @xmath71 ) for @xmath72 . for @xmath73
, corresponding to the length of ssdna used in the experiments by meller and coworkers @xcite , and for @xmath74 , the fractional difference in mfpts is about @xmath75 .
we emphasize again that one can not use the results of the comparison between these two times ( cases 1a and 1b in figure [ fig : dnaporecases ] ) and apply it to the experimental results in @xcite ( cases 1a and 2a in figure [ fig : dnaporecases ] ) . due to the asymmetry of the pore , the @xmath9 and the @xmath8 threading of dna through one end of the pore ( the so - called @xmath13 side )
can not be readily reduced to cases 1a and 1b in figure [ fig : sawtooth ] .
having established the difference in average first passage times for the two asymmetric potentials , let us now turn to the scaling of the first passage times with @xmath3 . for large @xmath3 ,
the scaling result @xmath4 found earlier is well known for classical diffusion or brownian dynamics .
however , this is in conflict with the equilibration time of a polymer with @xmath3 monomers in the absence of a pore and membrane , which already scales with @xmath3 to some power larger than @xmath76 for rouse dynamics of self - avoiding chains @xcite .
this suggests that a correct description of polymer translocation should be made in the context of subdiffusion , where the scaling of the average first passage time is to power @xmath77 of the classical result , although we do not give a prediction for the value of @xmath58 itself because the interactions involving the dna / polymer located outside the pore were not treated explicitly .
the scaling @xmath6 is also not surprising if one takes the relation @xmath78 and puts @xmath79 , but it does not rule out the argument made above regarding @xmath5 , only that it gives a reasonable and somewhat expected answer . moreover , using similar arguments , the result @xmath6 is consistent with numerical simulations made by chuang , kantor and kardar @xcite for diffusive dynamics of self - avoiding chains in two dimensions .
they found that the average of the first passage time scales as @xmath80 , where @xmath81 in two dimensions .
they also argued , assuming that the translocation coordinate goes like @xmath82 at short times , that @xmath83 . eliminating @xmath84 ,
their formulas imply that @xmath85 . to summarize , based on our calculations for the mean first passage times in asymmetric sawtooth potentials and experiments by meller and coworkers @xcite
, we expect that the average first passage times for the four cases indicated in figure [ fig : dnaporecases ] are all different .
the expression for the tail of the first passage time distribution in subdiffusion is of the form @xmath86 , where @xmath62 is the formula for the mean first passage time in classical diffusion .
because the power of @xmath11 in the distribution is less than two , the mean first passage time diverges . by constructing an average from the first passage times less than time @xmath5 such that approximately half of the passages get rejected
, we find an average that scales as @xmath67 .
this work was inspired by an interesting seminar talk given by amit meller .
we also acknowledge useful subsequent discussions with him .
rcl acknowledges the support of a doctoral dissertation fellowship from the university of minnesota graduate school .
we also wish to thank the minnesota supercomputing institute for the use of their facilities .
this work was supported in part by the mrsec program of the national science foundation under award number dmr-0212302 .
a generalization of the fpe describing anomalous diffusion is given by the fractional fpe @xcite @xmath87 equivalently , @xmath88 where the fokker - planck operator is defined as @xmath89 here @xmath90 is a generalized diffusion coefficient and @xmath15 is an external potential .
we have also set @xmath20 and the einstein relation is implicit .
the riemann - liouville fractional operator is defined through @xmath91 one can easily check that the ffpe reduces to the fpe or diffusion equation for @xmath57 .
given the initial distribution @xmath33 , the solution to equation ( [ ffpe1 ] ) is given by the bilinear expansion @xcite @xmath92 the functions @xmath93 and @xmath94 appear in the separation of variables ansatz @xmath95 .
the product function @xmath96 satisfies the ffpe .
note that the coordinate dependence comes through the eigenfunctions @xmath97 or @xmath98 , which are the same as for regular diffusion , satisfying the ( eigenvalue ) equations @xmath99 @xmath100 @xmath101 however , as to the time dependence , which for classical diffusion is described by exponentials ( @xmath102 ) , for subdiffusion it must satisfy the equation @xmath103 one can check that the following series definition of the mittag - leffler function @xmath104 satisfies equation ( [ mittagdiffeq ] ) @xmath105 this function is a natural extension of the exponential function , to which it degenerates for @xmath57 . by taking the laplace transform of both sides of equation ( [ mittagdiffeq ] ) ,
one obtains an alternative definition of the mittag - leffler function @xmath106 ( the subscript in the constant @xmath107 has been dropped . )
the long - time limit of the mittag - leffler function corresponds to the small @xmath108 limit of the laplace transform . expanding ( [ mittaglaplace ] ) in a series for small @xmath108 , @xmath109 taking the inverse transform , one obtains the long - time behaviour of the mittag - leffler function @xmath110 for @xmath111 , the laplace transform ( [ mittaglaplace ] ) becomes @xmath112 , the inverse transform of which is an exponential . for @xmath113 close to @xmath114
, we expect a long time interval in which the behaviour of the mittag - leffler function @xmath115 behaves like an exponential ; at much longer times the behaviour changes to a power law . the crossover is expected to happen when @xmath116 , or at about @xmath117 .
recall that the mfpt can be calculated from ( equation ( [ mfpt ] ) ) @xmath118 to derive an ordinary differential equation satisfied by @xmath30 , apply the operator @xmath119 to equation ( [ mfpt2 ] ) and use the eigenfunction expansion solution ( [ eigensol ] ) for @xmath18 , @xmath120 dxdt\nonumber\end{aligned}\ ] ] in the last two steps , the eigenvalue equations and the second version of the ffpe ( equation ( [ ffpe2 ] ) ) was used . using the initial condition @xmath33 and the definition of the fractional operator ,
after some algebra one obtains @xmath121\end{aligned}\ ] ] or @xmath122\end{aligned}\ ] ] for @xmath57 , corresponding to classical diffusion , the survival probability @xmath123 decays exponentially and the term with the integral goes to zero , yielding the familiar result of @xmath124 for the right - hand - side @xcite . for @xmath125 , @xmath123 goes like @xmath126 ( see ( [ s - series ] ) ) and the term with the integral goes like @xmath127 .
the right - hand - side diverges , which hints at the non - existence of the mfpt for subdiffusion @xcite .
from the previous section , the differential equation satisfied by the mfpt in the context of classical diffusion is ( temporarily putting back @xmath128 ) @xmath129 we solve for @xmath130 in this equation for a sawtooth potential ( case a , figure [ fig : sawtooth ] ) subject to the boundary conditions @xmath131 and @xmath132 , and the continuity of @xmath133 and @xmath134 in @xmath14 . in what follows we
let @xmath135 .
the solution , for @xmath26 between @xmath136 and @xmath137 where @xmath138 is an integer between @xmath139 and @xmath3 ( inclusive ) , is given by @xmath140 the coefficients @xmath141 and @xmath142 are given by @xmath143 \ ; \nonumber \\ b_m & = & \frac{d}{v^2}+n\left(\frac{a}{v}-(1 - 1/\xi)\frac{d}{v^2}\right)+ \nonumber \\ & + & \frac{d}{v^2}\frac{(\xi-1)^2}{\xi}\frac{(n - m)(n+m-1)}{2 } \ .\end{aligned}\ ] ] the mfpt for a particle initially located at @xmath144 is given by @xmath145 . to obtain the solution for case b in figure [ fig : sawtooth ]
, we may flip @xmath146 ( @xmath147 ) and the sign of @xmath148 in the expressions above .
since the mfpt does not exist for subdiffusion , one would want to calculate the distributions instead . in the long time
limit , using ( [ eigensol ] ) and ( [ mittaglimit ] ) , @xmath149 @xmath150 @xmath151 again , these results indicate that the mfpt diverges for @xmath125 .
it is also interesting to note that all the eigenfunctions @xmath97 , not just the ground state , enter in the expressions . to make sense of the expression multiplying @xmath152 in ( [ f - series ] )
write down the corresponding solution for classical diffusion under the same potential and the same value for the diffusion coefficient @xmath153 ( note exponential instead of mittag - leffler function ) .
the survival probability is given by @xmath154 while the mfpt is given by @xmath155 which is identical to the coefficient of @xmath156 in ( [ f - series ] ) .
in this section we determine the steady state current @xmath157 given fixed concentrations @xmath158 and @xmath159 at the boundaries .
let the potential @xmath15 satisfy @xmath160 , but is otherwise arbitrary .
the classical diffusion equation is given by @xmath161 where @xmath162 ( see equation ( [ fpe ] ) ) . in the steady state ,
@xmath163 , which implies that @xmath157 is spatially uniform . integrating @xmath164 and utilizing the boundary conditions , one obtains @xmath165 this expression is identical with fick s law with an effective diffusion constant of @xmath166 .
l. pontryagin , a. andronov , and a. vitt , zh .
. fiz . * 3 * , 165 ( 1933 ) ; translated and reprinted in _ noise in nonlinear dynamical systems , vol . 1 _ , edited by f. moss and p. v. e. mcclintock ( cambridge univ .
press , cambridge , in press ) , p. 329 . | motivated by experiments in which single - stranded dna with a short hairpin loop at one end undergoes unforced diffusion through a narrow pore , we study the first passage times for a particle , executing one - dimensional brownian motion in an asymmetric sawtooth potential , to exit one of the boundaries .
we consider the first passage times for the case of classical diffusion , characterized by a mean - square displacement of the form @xmath0 , and for the case of anomalous diffusion or subdiffusion , characterized by a mean - square displacement of the form @xmath1 with @xmath2 . in the context of classical diffusion ,
we obtain an expression for the mean first passage time and show that this quantity changes when the direction of the sawtooth is reversed or , equivalently , when the reflecting and absorbing boundaries are exchanged .
we discuss at which numbers of ` teeth ' @xmath3 ( or number of dna nucleotides ) and at which heights of the sawtooth potential this difference becomes significant . for large @xmath3
, it is well known that the mean first passage time scales as @xmath4 . in the context of subdiffusion , the mean first passage time
does not exist .
therefore we obtain instead the distribution of first passage times in the limit of long times .
we show that the prefactor in the power relation for this distribution is simply the expression for the mean first passage time in classical diffusion .
we also describe a hypothetical experiment to calculate the average of the first passage times for a fraction of passage events that each end within some time @xmath5 .
we show that this average first passage time scales as @xmath6 in subdiffusion . |
the atomic system is coupled to the quantized multimode radiation field . within the dipole approximation
, the hamiltonian that describes the dynamics of the atomic ensemble and the radiation field reads @xmath83 .
the first term represents the atomic energy , with @xmath84 being the energy difference between the ground ( @xmath85 ) and the three degenerate @xmath15 manifold states ( @xmath86 , @xmath87 and @xmath0 for @xmath88 and @xmath89 , respectively ) and the vector transition operator for the @xmath10-th atom ( located at @xmath90 ) being defined as @xmath91 .
the second term of the hamiltonian represents the radiation field , and here @xmath92 is the energy of a photon with momentum @xmath93 and polarization @xmath25 and @xmath94 is the annihilation operator of such a photon ( @xmath95 = \delta_{\mathbf{q}\mathbf{q}'}\delta_{\lambda\lambda'}$ ] ) .
the last term represents the coupling between the two systems , with @xmath96 and the coupling coefficient being given by @xmath97 , where @xmath98 is the quantization volume , @xmath99 the modulus of the transition dipole moment between the ground and the three degenerate excited states , and @xmath100 the unit polarization vector of the photon ( @xmath101 ) . within the dipole and born - markov approximations , and tracing out the environment , one obtains the quantum master equation for the atomic density matrix @xmath102 , @xmath16+{\cal d}(\rho)$ ] , which describes the dynamics of the atomic degrees of freedom @xcite .
the many - body hamiltonian that describes the coherent evolution of the system reads @xmath103 , characterized by the coefficient matrix @xmath104 with @xmath105 and @xmath106 being defined in the main text [ eqs .
( 5 ) and ( 6 ) , respectively ] and @xmath107\label{eq : cc}\end{aligned}\ ] ] for @xmath108 .
the dissipator @xmath109 reads @xmath110 , where @xmath111 with @xmath112 and @xmath113 being defined in the main text [ eqs .
( 7 ) and ( 8) , respectively ] and @xmath114\label{eq : ccp}\end{aligned}\ ] ] it is clear from the form of eqs .
( [ eq : v ] ) and ( [ eq : diss1 ] ) that the dynamics of the internal level @xmath115 is disconnected from the other two .
we will consider thus only the dynamics on the subspace formed by the states @xmath116 .
we briefly review the ewald summation technique @xcite introduced originally to cope with the evaluation of the electrostatic potential energy of a gas of electrons interacting through coulomb force . here
we wish to evaluate the discrete sum ( [ eq : wk ] ) .
considering two - dimensional lattices with unit cell defined by two lattice vectors @xmath117 , a vector @xmath118 connecting two atoms @xmath119 and @xmath120 can be written as @xmath121 where @xmath122 is the lattice translation vector , @xmath123 and @xmath124 denotes the position of the @xmath119-th atom in the unit cell .
we also define the reciprocal lattice vectors in the usual way as @xmath125 , where @xmath126 are the primitive reciprocal lattice vectors .
let us consider an algebraically decaying function of the atomic separation @xmath127 and say we want to evaluate the sum @xmath128 the ewald s trick consists of separating the contributions of @xmath129 to the sum to the short ( s ) and long ( l ) ranged part by introducing a regulating function @xmath130 through the identity @xmath131 so that @xmath132 where the function @xmath130 is chosen such that @xmath134 decays exponentially fast as @xmath135 and ( [ eq : fself ] ) is finite . due to the fast decay of @xmath136 ,
the short - range part can be readily obtained by evaluating the sum in ( [ eq : fs ] ) in real space .
we are thus left with the evaluation of ( [ eq : fl ] ) .
we start with the evaluation of the integral in ( [ eq : fl transf ] ) with the function @xmath144 @xmath145 where we have chosen @xmath146 , @xmath147 is an arbitrary control parameter , @xmath148 is the polar angle of @xmath149 and @xmath150 is the @xmath151-th bessel function of the first kind .
next , we split the integration as @xmath152 where @xmath153 finally , we substitute the asymptotic expression for the bessel function ( for large @xmath154 ) j_n(r ) ( r -_n ) , [ eq : bessel approx ] where @xmath155 , to @xmath156 so that _ 1,n_c^i_0 ^ - i_0^c , where i_a^b = _ c^dr ( r ) ( r-_n ) .
[ eq : iab ] importantly , we note , that only the term @xmath157 can give rise to the divergence in ( [ eq : vk pot ] ) , which we now analyze in detail .
substituting the identity = _
0^t t^- e^-t r and using = we get this amounts to evaluation of the integral ii & = & _ 0^t t^- _ 0^r ( 1-e^-r ) e^-t r ( r - _ n ) + & = & _ 0^t _ 0^r ^i r + c.c . + & = & _ 0^t ( - ) + c.c .
, [ eq : ii ] where @xmath158 . the final integration in ( [ eq : ii ] ) can be done by methods of contour integration along the contour depicted in fig .
[ fig : contour ] , i.e. _ c = _
c_1 + _ r + _ c_2 + _ .
[ eq : cont int ] using the property of the branch cut , the integration along @xmath159 becomes _ c_2 & = & _ r^z ( - ) + & = & _ r^z ( - ) + & = & _ r^z ( - ) + & = & _ ^r dz ( - ) .
omitting the c.c .
term , we express the leftmost part of ( [ eq : ii ] ) as ii_- & = & _ 0^z ( - ) + & = & _ 0^z g(z ) = _ 0^z - .
it can be immediately seen , that @xmath160 since @xmath161 in the limit of large @xmath162 .
next , since @xmath163 , assuming @xmath164 and substituting @xmath165 , we get from the estimation lemma | _
| |g(z)| 22 0 .
we are thus left with ii_- & = & _ c = i ( f(z_0 ) - f(z_1 ) ) [ eq : ii minus ] + & = & i , where we have defined the function @xmath166 and the poles f(z ) & = & + z_0 & = & i + z_1 & = & -+ i .
next , we note that this result can be further simplified in specific cases : for @xmath167 and @xmath142 respectively , we get the following options & n=0 : & _ 0 = ( ) = ; ( ) = + & n=2 : & _ 2 = ( ) = ; ( ) = - + & > 0:&(i ) = ( ( i ) ) = ; ( ( i ) ) = + & < 0:&(i ) = - ( ( i ) ) = - ; ( ( i ) ) = . [ eq : na ]
lets first consider the @xmath168 case . using ( [ eq : na ] ) , the relation ( [ eq : ii expansion ] ) simplifies to & > 0 : & ii_n=0 ( ) = - ( ( ( + i ) ) - ( ( + i ) ) ) + & < 0 : & ii_n=0 ( ) = - ( ( ( + i ) ) - ( ( + i ) ) ) . and similar expression is obtained for @xmath169 .
those results can be combined to a single expression ii_n ( ) = ( ) - ( ( ( + i ) ) - ( ( + i ) ) ) , [ eq : ii n=0 ] where @xmath170 is the heaviside step function . remarkably , the function ( [ eq : ii n=0 ] ) is finite in the limit @xmath171 , but diverges as @xmath172 , which constitutes the main analytic result of this section .
_ remark : _ the discontinuity and one - sided divergence could be seen already in ( [ eq : ii ] ) by setting @xmath173 .
this would result in the diverging integral in the vicinity of the origin , namely the part @xmath174 in ( [ eq : cont int ] ) , which , according to the estimation lemma , scales as @xmath175 .
in the previous section we have found that ( [ eq : fs]),([eq : fself ] ) can be evaluated numerically , while ( [ eq : fl ] ) can be written as @xmath176 .
\label{eq:1nv}\ ] ] here , the expressions @xmath177 and @xmath178 are finite and can by obtained numerically . in principle similar decomposition can be obtained when considering the functions ( [ eq:2nv]),([eq:3nv ] ) , however we found it more convenient to evaluate the corresponding element ( [ eq : vk direct ] ) by direct summation in real space .
the final subtlety is related to the fact , that for @xmath179 the self - interaction term ( [ eq : fself ] ) is not defined for @xmath180 as it depends on the direction in which the origin is approached ( which is due to the dependence of @xmath180 on the polar angle of @xmath181 ) .
this problem can be circumvented by considering @xmath182 instead in which case the self - interaction term vanishes . here
, it is still possible to evaluate the integral in ( [ eq : iab ] ) symbolically , which we then use in ( [ eq:1nv ] ) .
66ifxundefined [ 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 \doibase http://dx.doi.org/10.1016/s0003-4916(02)00018-0 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.80.1083 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.82.3045 [ * * , ( ) ] link:\doibase 10.1038/ncomms2531 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.45.494 [ * * , ( ) ] link:\doibase 10.1103/physrevb.23.5632 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.48.1559 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.50.1395 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.61.2015 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.88.180401 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.95.226801 [ * * , ( ) ] link:\doibase 10.1126/science.1137201 [ * * , ( ) ] link:\doibase 10.1126/science.1137430 [ * * , ( ) ] link:\doibase 10.1126/science.1148047 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.100.156401 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.106.236804 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.107.146803 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.106.236802 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.106.236803 [ * * , ( ) ] link:\doibase 10.1103/physrevb.86.201101 [ * * , ( ) ] link:\doibase 10.1103/physreva.86.053618 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.110.185301 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.115.026802 [ * * , ( ) ] link:\doibase 10.1103/physreva.92.033617 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevb.90.165102 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.113.156402 [ * * , ( ) ] link:\doibase 10.1103/physrevb.93.041102 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.116.216802 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.025303 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.266804 [ * * , ( ) ] link:\doibase 10.1103/physrevb.87.081106 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.110.185302 [ * * , ( ) ] link:\doibase 10.1103/physreva.91.053617 [ * * , ( ) ] link:\doibase 10.1103/physreva.92.033609 [ * * , ( ) ] link:\doibase 10.1103/physreva.91.033838 [ * * , ( ) ] link:\doibase 10.1103/physreva.92.063822 [ * * , ( ) ] link:\doibase 10.1103/physreva.81.053821 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrevlett.113.133602 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.113.223601 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevlett.116.183601 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.71.3697 [ * * , ( ) ] link:\doibase 10.1103/physrevb.48.11851 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.110503 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevlett.110.143602 [ * * , ( ) ] link:\doibase 10.1080/00268976.2013.800604 [ * * , ( ) ] link:\doibase 10.1038/ncomms6391 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physreva.47.1336 [ * * , ( ) ] @noop * * , ( ) , , , and , eds . , @noop _ _ ( , , ) @noop _ _ , vol .
( , ) link:\doibase 10.1103/physrevb.15.1959 [ * * , ( ) ] , ed . , @noop _ _ ( , ) link:\doibase 10.1143/jpsj.74.1674 [ * * , ( ) ] link:\doibase 10.1103/physreva.10.1096 [ * * , ( ) ] @noop * * , ( ) @noop ( ) @noop * * , ( ) link:\doibase 10.1038/nphoton.2013.274 [ * * , ( ) ] | we investigate the existence of topological phases in a dense two - dimensional atomic lattice gas .
the coupling of the atoms to the radiation field gives rise to dissipation and a non - trivial coherent long - range exchange interaction whose form goes beyond a simple power - law .
the far - field terms of the potential which are particularly relevant for atomic separations comparable to the atomic transition wavelength can give rise to energy spectra with one - sided divergences in the brillouin zone .
the long - ranged character of the interactions has another important consequence : it can break of the standard bulk - boundary relation in topological insulators .
we show that topological properties such as the transport of an excitation along the edge of the lattice are robust with respect to the presence of lattice defects and dissipation .
the latter is of particular relevance as dissipation and coherent interactions are inevitably connected in our setting . _
introduction .
_ recently , the pursuit of topological phases in quantum many - body systems has been the focus of intense research .
the potential application of these topological states for robust quantum computation @xcite is one of the driving forces for this increased interest .
so - called topological insulators are usually characterized by bulk bands separated by a gap and the presence of gapless edge states whose properties are topologically protected against local perturbations such as external disorder or noise @xcite .
paradigmatic examples of these include the integer and fractional quantum hall effects , which were initially realized on two - dimensional electron gases subject to strong magnetic fields @xcite .
since their discovery , several lattice models that do not require external magnetic fields have been proposed and some realized experimentally @xcite . due to the high degree of experimental control that is achievable nowadays , cold atoms and molecules
have been proposed as platforms for the exploration of novel topological phases of quantum matter @xcite . in particular ,
many - body systems that display long - range interactions @xcite such as polar molecules @xcite , atoms with large magnetic dipoles @xcite , and rydberg atoms @xcite have been shown to feature topologically non - trivial flat bands and fractional quantum hall states .
perpendicular to the lattice and the off - resonant coupling of @xmath0 to an auxiliary state @xmath1 with @xmath2 . *
( b ) : * hamiltonian coefficients @xmath3 [ eq . ( [ eq : a ] ) , solid line ] and @xmath4 [ eq . ( [ eq : b ] ) , dashed line ] vs. @xmath5 . * ( c ) * and * ( d ) : * real space configurations and first brillouin zone for sl and hl , respectively .
the red circle represents the points of divergence and discontinuity of the spectrum . *
( e ) * and * ( f ) : * dispersion relations for the sl and the hl , respectively , which exhibit the existence of a one - sided divergence at @xmath6.,title="fig : " ] + in this paper , we explore the topological properties of a two - dimensional lattice of atoms where long - range interactions arise intrinsically via coherent light scattering between internal atomic states @xcite .
we consider the full interaction going thus beyond the usually employed near - field approximation which is of relevance for atomic separations comparable to the atomic transition wavelength .
this scenario is currently studied in various contexts , e.g. , the exploration of collective light scattering and super- and sub - radiant decay @xcite .
we study two different lattice geometries ( a square and a honeycomb lattice ) and find that this simple system can support topologically non - trivial phases .
the often neglected far - field terms of the interactions lead to one - sided divergences in the single - particle spectrum .
we explore the consequences of the long - range character of the interaction on the relation between the topological properties of the bulk and the boundary of a finite size system .
remarkably , we find that the standard bulk - boundary relation @xcite , which is well established for short - range interactions , does not generally hold .
furthermore , we find that neither the presence of lattice defects nor dissipation destroy the topological properties of the system .
the latter is of particular importance , as the considered system is inevitably open : both dissipation and coherent exchange interactions originate from the coupling of the atoms to the radiation field . _
the system .
_ we consider a two - dimensional optical lattice with @xmath7 sites lying in the @xmath8 plane ( lattice spacing @xmath9 ) , where each site is occupied by a single atom .
the @xmath10-th atom is located at position @xmath11 with internal levels @xmath12 ( the ground state ) and @xmath13 ( the excited states ) corresponding to the @xmath14 and @xmath15 total angular momentum manifolds , respectively [ see fig . [
fig : scheme](a ) ] .
this level structure is naturally available in a variety of systems such as alkaline - earth - metal atoms @xcite , dysprosium atoms @xcite , polar molecules @xcite or rydberg systems @xcite . the coupling of the atoms to the quantized multimode radiation field results in an effective long - range exchange interaction and collective dissipation @xcite .
within the dipole and born - markov approximations , the dynamics of the atomic system is described by the master equation @xmath16+{\cal d}(\rho)$ ] , with @xmath17 here , @xmath18 and @xmath19 note , that we have used here the fact that the dynamics of the @xmath20 level and the @xmath21 manifold are decoupled , see appendix a for details .
the coefficients in eqs .
( [ eq : w ] ) and ( [ eq : gamma ] ) read @xmath22\label{eq : a}\\ b_{jl}&=&\frac{3\gamma}{8}\left[\frac{\cos{\kappa_{jl}}}{\kappa_{jl}}-3\left(\frac{\sin{\kappa_{jl}}}{\kappa_{jl}^2 } + \frac{\cos{\kappa_{jl}}}{\kappa_{jl}^3}\right)\right]\label{eq : b}\\ a'_{jl}&=&\frac{3\gamma}{4}\left[\frac{\sin{\kappa_{jl}}}{\kappa_{jl}}-\frac{\cos{\kappa_{jl}}}{\kappa_{jl}^2 } + \frac{\sin{\kappa_{jl}}}{\kappa_{jl}^3}\right]\label{eq : ap}\\ b'_{jl}&=&\frac{3\gamma}{8}\left[-\frac{\sin{\kappa_{jl}}}{\kappa_{jl}}-3\left(\frac{\cos{\kappa_{jl}}}{\kappa_{jl}^2 } -\frac{\sin{\kappa_{jl}}}{\kappa_{jl}^3}\right)\right],\label{eq : bp}\end{aligned}\ ] ] where @xmath23 is the single atom decay rate and @xmath24 , with @xmath25 being the wavelength of the transition from the ground to the excited state manifold [ see fig .
[ fig : scheme](a ) ] , @xmath26 , and @xmath27 is the polar angle between the @xmath10-th and @xmath28-th atom .
the many - body hamiltonian ( [ eq : h ] ) conserves the number of excitations in the system and describes their exchange among the atoms .
the coefficients in this hamiltonian [ eqs .
( [ eq : a ] ) and ( [ eq : b ] ) ] decay as @xmath29 for short distances ( @xmath30 ) and as @xmath31 for @xmath32 [ see fig . [
fig : scheme](b ) ] .
we will study the topological properties of the system in an intermediate regime , where the full potential needs to be considered .
in particular , in the remainder of the paper we will fix the ratio @xmath33 @xcite ( note that the effects we show are not constrained to this specific value ) and systematically investigate two specific lattice geometries , the square lattice ( sl ) and the honeycomb lattice ( hl ) [ see figs .
[ fig : scheme](c ) and ( d ) , respectively ] .
we will proceed by analyzing first the band structure of the hamiltonian ( [ eq : h ] ) and then studying the effect of the dissipation ( [ eq : diss ] ) on the edge transport in the topologically non - trivial phases . in its current formulation ,
the hamiltonian ( [ eq : h ] ) is symmetric under time reversal and hence is topologically trivial . in order to possibly find topologically non - trivial phases we consider two ways of breaking the time reversal symmetry .
first , we lift the degeneracy of the states @xmath34 , which can be achieved by means of a magnetic field such that the two states energies are shifted by @xmath35 [ see fig . [
fig : scheme](a ) ] .
secondly , we couple off - resonantly the state @xmath0 via a microwave field to an auxiliary hyperfine state , @xmath36 @xcite . up to second order in @xmath37 , where @xmath38 and @xmath39 are the rabi frequency and the detuning of the microwave field , respectively ,
the @xmath0 state gets dressed and the potential ( [ eq : w ] ) is modified as @xmath40 where @xmath41 is the kronecker delta symbol .
_ band structure and divergences .
_ to obtain the band structure we express ( [ eq : h ] ) in the reciprocal space as @xmath42 where @xmath43 and @xmath44 with @xmath45 being the relative distance between any pair of atoms .
note that the potential ( [ eq : w - mod ] ) ( due to the terms that decay as @xmath31 ) makes the sum in eq .
( [ eq : wk ] ) ) not convergent . potentials with similar features have been shown to lead to interesting physical effects , such as supersonic spreading of the correlations @xcite .
such potentials are also encountered in self - gravitating systems @xcite or in electrons in solids subject to the coulomb force @xcite . to evaluate the sum ( [ eq : wk ] ) we employ the ewald summation technique @xcite , see appendix b for details . as a result
, we find that the potential ( [ eq : wk ] ) features one sided divergences in the spectrum occurring for wavevectors lying on a circle of radius @xmath46 in reciprocal space .
more specifically , the potential diverges as @xmath47 , while @xmath48 remains finite ( see appendix b ) .
this situation is depicted in fig .
[ fig : scheme](e ) ( sl ) and fig .
[ fig : scheme](f ) ( hl ) .
note that in this work we only study cases where the divergence falls inside the brillouin zone , @xmath49 for sl and @xmath50 for hl . _ chern numbers and edge states .
_ in the language of differential geometry , the topological properties are studied in terms of differentiable fibre bundles assuming a differentiable hamiltonian map @xmath51 , mapping the brillouin zone ( represented as a two - dimensional torus @xmath52 ) to some target space @xmath53 @xcite . here
the topology is characterized by the chern number defined as @xmath54 where @xmath55 , @xmath56 is the berry connection , @xmath57 is an eigenstate of the hamiltonian ( [ eq : hk ] ) , @xmath58 , and @xmath59 . in the present case
the assumption of differentiability is not satisfied due to the discontinuity of the potential ( [ eq : wk ] ) .
however , formally it is still possible to evaluate the chern number using the algorithm of ref .
@xcite , where special care is taken to avoid the points in the brillouin zone where @xmath60 diverges .
this corresponds to evaluating ( [ eq : c ] ) with an effectively bounded hamiltonian , which in turn yields integer values of @xmath61 .
note also that the divergence appears as a consequence of considering the potential in the thermodynamic limit , a situation that is ill - defined as the finite propagation time of the radiation modes mediating the exchange of photons in the system is neglected @xcite .
the resulting phase diagram is shown in fig .
[ fig : chern](a ) ( sl ) and fig .
[ fig : chern](b ) ( hl ) .
we note that in the sl case it is required that @xmath62 in order to access the topologically non - trivial region and that in hl several topologically distinct regions can be accessed by tuning @xmath63 alone ( i.e. , @xmath64 ) .
plane for the sl * ( a ) * and as a function of @xmath63 for the hl * ( b)*. the numbers in the square brackets indicate the chern numbers and are ordered from the lowest to highest energy bands . * ( c ) * and * ( d ) : * the spectrum in the infinite strip geometry for the sl ( @xmath65 atoms , @xmath66 , @xmath67 ) and hl ( @xmath68 atoms , @xmath69 , bearded zigzag boundary conditions @xcite ) .
note that with these boundary conditions @xmath70 denotes the spacing between neighboring unit lattice sites in the @xmath71 direction .
the corresponding points in the phase diagrams are denoted by the red cross ( a ) and red line ( b ) , respectively .
the size of the circles indicate the weight of the given eigenmode on one ( red ) or the other ( blue ) edge of the strip.,title="fig : " ] + a hallmark of the non - trivial topology of the bulk is the appearance of edge states in finite - size systems . considering an infinite strip keeping the system finite in the @xmath71 direction we plot the corresponding band structure in fig .
[ fig : chern](c ) ( sl ) and fig .
[ fig : chern](d ) ( hl ) , where the edge states appear in the bulk gaps . here
, the size of the filled circles corresponds to the weight of the given eigenmode on one ( red ) or the other ( blue ) edge of the infinite strip .
we note that in the sl case the edge states are nearly degenerate and energetically hidden @xcite making it difficult to access them experimentally . in the hl case , relatively well separated bands form .
moreover , it is easy to verify that in the example shown indeed the hall conductivity or , equivalently , the sum of the chern numbers of the filled bands below a band gap , determines the net number of edge states crossing that gap @xcite . _
finite system : quasi - momentum and driven - dissipative dynamics .
_ one important consequence of the long - ranged character of the interactions in the context of topological systems is that the bulk - edge correspondence , well established for systems with short - range potentials does not necessarily hold , as we exemplify in the following . in fourier space ,
the edge modes can be seen as modes spanning between different energy bands . in a finite system , to obtain a similar measure of whether any given mode has a chiral edge conductivity associated with it , we define a _ quasi - momentum _ following a similar approach to @xcite : when an excitation hops from site to site , it accumulates a phase . if this phase of hopping is constant around the edge of the lattice , the hopping occurs around the edge in a given direction .
if the hopping phase is random between different lattice sites , however , there is no preferred direction of hopping .
we therefore define the _ quasi - momentum _ @xmath72 as the average phase difference between neighbouring lattice sites around the edge of the lattice : @xmath73,\ ] ] where @xmath74 is the number of atoms in the outermost edge of the lattice . in fig .
3(a ) we show the quasi - momentum spectrum for a hl of an hexagonal shape with a total of 486 atoms .
we have overlayed the band structure results for an infinite strip with the same number of atoms in the @xmath71 direction for comparison .
( d ) ] .
the blue ( red ) dots represent states with less ( more ) than 95% spatial support on the edge of the lattice ( see text for details ) . *
( i - iii ) * show the spatial configuration of the indicated eigenstates . *
( b ) * and * ( c ) : * excitation dynamics and the steady state ( with defects and dissipation ) under the driving of a single atom on the edge ( indicated by the star in the leftmost panels ) with detuning @xmath75 ( b ) and @xmath76 ( c).,title="fig : " ] + another defining characteristic of an edge state is that the vast majority of its weight is on the physical edge of the system .
writing an eigenstate of the system as @xmath77 , where @xmath78 and @xmath10 index the internal degrees of freedom and the spatial position , respectively , we quantify the weight of the eigenstate on the edge of the system as @xmath79 .
the blue and red dots in fig . [ fig : finite](a ) correspond to states where @xmath80 and @xmath81 , respectively . the spatial configuration of two possible edge states from the middle and lowest bulk gap is shown in figs . [
fig : finite](i ) and ( ii ) [ which we denote as states ( i ) and ( ii ) in the following ] . a bulk state is shown for comparison in fig . [
fig : finite](iii ) . here , the size of the red circles correspond to the weight of the state at a given site . in order to further characterize the edge states in the finite system , we model the dynamics of an excitation in the lattice under the hamiltonian ( [ eq : h ] ) and weak driving .
specifically , we drive a single atom on the leftmost corner of the hl [ marked by a star in the leftmost panels of figs .
[ fig : finite](b ) and ( c ) ] with weak driving strength and we choose the detuning to be @xmath75 and @xmath76 [ figs .
[ fig : finite](b ) and ( c ) , respectively ] , so that we resonantly address the states ( i ) and ( ii ) , respectively .
we observe that initially in both ( b ) and ( c ) the excitation dynamics is chiral with clockwise propagation before the steady state is reached .
next , as one of the most appealing features of the edge states is their robustness against local perturbations , we simulate the excitation dynamics including @xmath82 of uniformly distributed lattice defects , i.e. , empty sites due to an imperfect loading of atoms [ fourth panels in figs .
[ fig : finite](b , c ) ] .
we observe a clear difference in the excitation propagation when driving the state ( i ) , where the propagation is clearly interrupted , and the state ( ii ) , where the excitations still populate a significant part of the edge .
this strongly indicates that only the second one , state ( ii ) , is an actual edge state .
this is in contrast to what one might expect from the equilibrium bulk analysis , where the edge state conductivity , proportional to the sum of the chern numbers of the filled lower lying bands , is the same for both states ( i ) and ( ii ) [ see the red line in fig . [
fig : chern](b ) ] . finally , as the mechanism that is responsible for the existence of coherent interactions is also responsible for the dissipation in the system
, we simulate the dynamics above including the dissipative term ( [ eq : diss ] ) .
remarkably , we find that while in the case ( i ) the dissipation makes the excitation decay after only a few sites , when exciting the edge state ( ii ) the excitation becomes sub - radiant , i.e. , it decays with a rate much smaller than the single atom decay rate @xmath23 .
_ conclusions and outlook .
_ we have found the existence of topologically non - trivial phases in a dense atomic two - dimensional lattice system coupled to the radiation field .
we show that one can excite edge states that allow for the transport of an excitation over long distances along the edge of the lattice that are robust to the presence of defects .
moreover , these edge states are remarkably long - lived due to the collective character of the dissipation ( a detailed analysis of this effect will be the aim of further investigations ) .
finally , we have found that , due to the long - range character of the interactions , the bulk - boundary relations , well established for topological insulators with short - range interactions , are not generally valid in our setting .
the authors thank m. marcuzzi and s. weber for useful discussions .
the research leading to these results has received funding from the european research council under the european union s seventh framework programme ( fp/2007 - 2013 ) / erc grant agreement no . 335266 ( escquma ) and the epsrc grant no .
ep / m014266/1 .
r.j .
bettles acknowledges funding from the uk epsrc ( grant no .
ep / l023024/1 ) .
b. olmos was supported by the royal society and epsrc grant dh130145 .
c.s .
adams also acknowledges support from fet - proact project ` rysq ' ( h2020-fetproact-2014 - 640378-rysq ) . |
recent studies have shown that the power spectrum of the stock market fluctuations is inversely proportional to the frequency on some power , which points to self - similarity in time for processes underlying the market @xcite .
our knowledge of the random and/or deterministic character of those processes is however limited .
one rigorous way to sort out the noise from the deterministic components is to examine in details correlations at @xmath8 scales through the so called master equation , i.e. the fokker - planck equation ( and the subsequent langevin equation ) for the probability distribution function ( @xmath9 ) of signal increments @xcite .
this theoretical approach , so called solving the inverse problem , based on @xmath10 statistical principles @xcite , is often the first step in sorting out the @xmath11 model(s ) . in this paper
we derive fpe , directly from the experimental data of two financial indices and two exchange rates series , in terms of a drift @xmath6 and a diffusion @xmath7 coefficient .
we would like to emphasize that the method is model independent .
the technique allows examination of long and short time scales _ on the same footing_. the so found analytical form of both drift @xmath6 and diffusion @xmath7 coefficients has a simple physical interpretation , reflecting the influence of the deterministic and random forces on the examined market dynamics processes . placed into a langevin equation
, they could allow for some first step forecasting .
we consider the daily closing price @xmath12 of two major financial indices , nikkei 225 for japan and nasdaq composite for us , and daily exchange rates involving currencies of japan , us and europe , @xmath0/@xmath1 and @xmath1/@xmath2 from january 1 , 1985 to may 31 , 2002 . data series of nikkei 225 ( 4282 data points ) and nasdaq composite ( 4395 data points ) and are downloaded from the yahoo web site ( @xmath13 ) .
the exchange rates of @xmath0/@xmath1 and @xmath1/@xmath2 are downloaded from @xmath14 and both consists of 4401 data points each .
data are plotted in fig .
1(a - d ) .
the @xmath1/@xmath2 case was studied in @xcite for the 1992 - 1993 years .
see also [ 6 ] , [ 8 - 10 ] and [ 11 ] for some related work and results on such time series signals , some on high frequency data , and for different time spans .
/@xmath1 and ( d ) @xmath1/@xmath2 exchange rates for the period from jan . 01 , 1985 till may 31 , 2002,title="fig : " ] /@xmath1 and ( d ) @xmath1/@xmath2 exchange rates for the period from jan . 01 , 1985 till may 31 , 2002,title="fig : " ] /@xmath1 and ( d ) @xmath1/@xmath2 exchange rates for the period from jan .
01 , 1985 till may 31 , 2002,title="fig : " ] /@xmath1 and ( d ) @xmath1/@xmath2 exchange rates for the period from jan .
01 , 1985 till may 31 , 2002,title="fig : " ]
to examine the fluctuations of the time series at different time delays ( or time lags ) @xmath15 we study the distribution of the increments @xmath16 .
therefore , we can analyze the fluctuations at long and short time scales on the same footing .
results for the probability distribution functions ( pdf ) @xmath17 are plotted in fig .
2(a - d ) . note that while the pdf of one day time delays ( circles ) for all time series studied have similar shapes , the pdf for longer time delays shows fat tails as in @xcite of the same type for nikkei 225 , @xmath0/@xmath1 and @xmath1/@xmath2 , but is different from the pdf for nasdaq . of ( a ) nikkei 225 , ( b ) nasdaq , ( c ) @xmath0/@xmath1 and ( d ) @xmath1/@xmath2 from jan .
01 , 1985 till may 31 , 2002 for different delay times .
each pdf is displaced vertically to enhance the tail behavior ; symbols and the time lags @xmath15 are in the insets .
the discretisation step of the histogram is ( a ) 200 , ( b ) 27 , ( c ) 0.1 and ( d ) 0.008 respectively , title="fig : " ] of ( a ) nikkei 225 , ( b ) nasdaq , ( c ) @xmath0/@xmath1 and ( d ) @xmath1/@xmath2 from jan . 01 , 1985 till may 31 , 2002 for different delay times .
each pdf is displaced vertically to enhance the tail behavior ; symbols and the time lags @xmath15 are in the insets .
the discretisation step of the histogram is ( a ) 200 , ( b ) 27 , ( c ) 0.1 and ( d ) 0.008 respectively , title="fig : " ] of ( a ) nikkei 225 , ( b ) nasdaq , ( c ) @xmath0/@xmath1 and ( d ) @xmath1/@xmath2 from jan . 01 , 1985 till may 31 , 2002 for different delay times .
each pdf is displaced vertically to enhance the tail behavior ; symbols and the time lags @xmath15 are in the insets .
the discretisation step of the histogram is ( a ) 200 , ( b ) 27 , ( c ) 0.1 and ( d ) 0.008 respectively , title="fig : " ] of ( a ) nikkei 225 , ( b ) nasdaq , ( c ) @xmath0/@xmath1 and ( d ) @xmath1/@xmath2 from jan . 01 , 1985 till may 31 , 2002 for different delay times .
each pdf is displaced vertically to enhance the tail behavior ; symbols and the time lags @xmath15 are in the insets .
the discretisation step of the histogram is ( a ) 200 , ( b ) 27 , ( c ) 0.1 and ( d ) 0.008 respectively , title="fig : " ] more information about the correlations present in the time series is given by joint pdf s , that depend on @xmath18 variables , i.e. @xmath19 .
we started to address this issue by determining the properties of the joint pdf for @xmath20 , i.e. @xmath21 . the symmetrically tilted character of the joint pdf contour levels ( fig .
3(a - c ) ) around an inertia axis with slope 1/2 points out to the statistical dependence , i.e. a correlation , between the increments in all examined time series . of ( a ) nikkei 225 , ( b ) nasdaq closing price signal and ( c ) @xmath0/@xmath1 and ( d ) @xmath1/@xmath2 exchange rates for @xmath22 and @xmath23 .
contour levels correspond to @xmath24 from center to border , title="fig : " ] of ( a ) nikkei 225 , ( b ) nasdaq closing price signal and ( c ) @xmath0/@xmath1 and ( d ) @xmath1/@xmath2 exchange rates for @xmath22 and @xmath23 .
contour levels correspond to @xmath24 from center to border , title="fig : " ] of ( a ) nikkei 225 , ( b ) nasdaq closing price signal and ( c ) @xmath0/@xmath1 and ( d ) @xmath1/@xmath2 exchange rates for @xmath22 and @xmath23 .
contour levels correspond to @xmath24 from center to border , title="fig : " ] of ( a ) nikkei 225 , ( b ) nasdaq closing price signal and ( c ) @xmath0/@xmath1 and ( d ) @xmath1/@xmath2 exchange rates for @xmath22 and @xmath23 .
contour levels correspond to @xmath24 from center to border , title="fig : " ] the conditional probability function is @xmath25 for @xmath26 . for any @xmath27 @xmath28 @xmath29 @xmath28 @xmath30 ,
the chapman - kolmogorov equation is a necessary condition of a markov process , one without memory but governed by probabilistic conditions @xmath31 the chapman - kolmogorov equation when formulated in @xmath32 form yields a master equation , which can take the form of a fokker - p1anck equation @xcite . for @xmath33 , @xmath34p(\delta x,\tau )
\label{efp}\ ] ] in terms of a drift @xmath6(@xmath35,@xmath36 ) and a diffusion coefficient @xmath7(@xmath35,@xmath36 ) ( thus values of @xmath36 represent @xmath37 , @xmath38 ) .
the coefficient functional dependence can be estimated directly from the moments @xmath39 ( known as kramers - moyal coefficients ) of the conditional probability distributions : @xmath40 @xmath41 for @xmath42 .
the functional dependence of the drift and diffusion coefficients @xmath6 and @xmath7 for the normalized increments @xmath35 is well represented by a line and a parabola , respectively .
the values of the polynomial coefficients are summarized in table 1 and fig .
4 . and
@xmath7 for the pdf evolution equation ( 3 ) ; @xmath35 is normalized with respect to the value of the standard deviation @xmath43 of the pdf increments at delay time 32 days : ( a , b ) nikkei 225 and ( c , d ) nasdaq closing price signal , ( e , f ) @xmath0/@xmath1 and ( g , h ) @xmath1/@xmath2 exchange rates , title="fig : " ] and @xmath7 for the pdf evolution equation ( 3 ) ; @xmath35 is normalized with respect to the value of the standard deviation @xmath43 of the pdf increments at delay time 32 days : ( a , b ) nikkei 225 and ( c , d ) nasdaq closing price signal , ( e , f ) @xmath0/@xmath1 and ( g , h ) @xmath1/@xmath2 exchange rates , title="fig : " ] and @xmath7 for the pdf evolution equation ( 3 ) ; @xmath35 is normalized with respect to the value of the standard deviation @xmath43 of the pdf increments at delay time 32 days : ( a , b ) nikkei 225 and ( c , d ) nasdaq closing price signal , ( e , f ) @xmath0/@xmath1 and ( g , h ) @xmath1/@xmath2 exchange rates , title="fig : " ] and @xmath7 for the pdf evolution equation ( 3 ) ; @xmath35 is normalized with respect to the value of the standard deviation @xmath43 of the pdf increments at delay time 32 days : ( a , b ) nikkei 225 and ( c , d ) nasdaq closing price signal , ( e , f ) @xmath0/@xmath1 and ( g , h ) @xmath1/@xmath2 exchange rates , title="fig : " ] and @xmath7 for the pdf evolution equation ( 3 ) ; @xmath35 is normalized with respect to the value of the standard deviation @xmath43 of the pdf increments at delay time 32 days : ( a , b ) nikkei 225 and ( c , d ) nasdaq closing price signal , ( e , f ) @xmath0/@xmath1 and ( g , h ) @xmath1/@xmath2 exchange rates , title="fig : " ] and @xmath7 for the pdf evolution equation ( 3 ) ; @xmath35 is normalized with respect to the value of the standard deviation @xmath43 of the pdf increments at delay time 32 days : ( a , b ) nikkei 225 and ( c , d ) nasdaq closing price signal , ( e , f ) @xmath0/@xmath1 and ( g , h ) @xmath1/@xmath2 exchange rates , title="fig : " ] and @xmath7 for the pdf evolution equation ( 3 ) ; @xmath35 is normalized with respect to the value of the standard deviation @xmath43 of the pdf increments at delay time 32 days : ( a , b ) nikkei 225 and ( c , d ) nasdaq closing price signal , ( e , f ) @xmath0/@xmath1 and ( g , h ) @xmath1/@xmath2 exchange rates , title="fig : " ] and @xmath7 for the pdf evolution equation ( 3 ) ; @xmath35 is normalized with respect to the value of the standard deviation @xmath43 of the pdf increments at delay time 32 days : ( a , b ) nikkei 225 and ( c , d ) nasdaq closing price signal , ( e , f ) @xmath0/@xmath1 and ( g , h ) @xmath1/@xmath2 exchange rates , title="fig : " ] for two values of @xmath15 , @xmath44 days , @xmath45 day for nasdaq .
contour levels correspond to @xmath46=-0.5,-1.0,-1.5,-2.0,-2.5 from center to border ; data ( solid line ) and solution of the chapman kolmogorov equation integration ( dotted line ) ; ( b ) and ( c ) data ( circles ) and solution of the chapman kolmogorov equation integration ( plusses ) for the corresponding pdf at @xmath47 = -50 and + 50,title="fig : " ] for two values of @xmath15 , @xmath44 days , @xmath45 day for nasdaq .
contour levels correspond to @xmath46=-0.5,-1.0,-1.5,-2.0,-2.5 from center to border ; data ( solid line ) and solution of the chapman kolmogorov equation integration ( dotted line ) ; ( b ) and ( c ) data ( circles ) and solution of the chapman kolmogorov equation integration ( plusses ) for the corresponding pdf at @xmath47 = -50 and + 50,title="fig : " ] for two values of @xmath15 , @xmath44 days , @xmath45 day for nasdaq .
contour levels correspond to @xmath46=-0.5,-1.0,-1.5,-2.0,-2.5 from center to border ; data ( solid line ) and solution of the chapman kolmogorov equation integration ( dotted line ) ; ( b ) and ( c ) data ( circles ) and solution of the chapman kolmogorov equation integration ( plusses ) for the corresponding pdf at @xmath47 = -50 and + 50,title="fig : " ] .values of the polynomial coefficients defining the linear and quadratic dependence of the drift and diffusion coefficients @xmath48 and @xmath49 for the fpe ( 3 ) of the normalized data series ; @xmath43 represents the normalization constant equal to the standard deviation of the @xmath15=32 days pdf [ cols="<,^,^,^,^,^,^,^",options="header " , ] [ tab1 ] the leading coefficient ( @xmath50 ) of the linear @xmath6 dependence has approximately the same values for all studied signals , thus the same deterministic noise ( drift coefficient ) .
note that the leading term ( @xmath51 ) of the functional dependence of diffusion coefficient of the nasdaq closing price signal is about twice the leading , i.e. second order coefficient , of the other three series of interest .
this can be interpreted as if the stochastic component ( diffusion coefficient ) of the dynamics of nasdaq is twice larger than the stochastic components of nikkei 225 , @xmath0/@xmath1 and @xmath1/@xmath2 .
a possible reason for such a behavior may be related to the transaction procedure on the nasdaq
. our numerical result agrees with that of ref .
@xcite if a factor of ten is corrected in the latter ref . for @xmath51 . the validity of the chapman - kolmogorov equation has also been verified .
a comparison of the directly evaluated conditional pdf with the numerical integration result ( 2 ) indicates that both pdf s are statistically identical .
the more pronounced peak for the nasdaq is recovered ( see fig .
an analytical form for the pdf s has been obtained by other authors @xcite but with models different from more classical ones @xcite .
the present study of the evolution of japan and us stock as well as foreign currency exchange markets has allowed us to point out the existence of deterministic and stochastic influences .
our results confirm those for high frequency ( 1 year long ) data @xcite .
the markovian nature of the process governing the pdf evolution is confirmed for such long range data as in @xcite for high frequency data .
we found that the stochastic component ( expressed through the diffusion coefficient ) for nasdaq is substantially larger ( twice ) than for nikkei 225 , @xmath0/@xmath1 and @xmath1/@xmath2 .
this could be attributed to the electronic nature of executing transactions on nasdaq , therefore to different stochastic forces for the market dynamics .
silva ac , yakovenko vm ( 2002 ) comparison between the probability distribution of returns in the heston model and empirical data for stock indices cond- mat/ 0211050 ; dragulescu aa , yakovenko vm ( 2002 ) probability distribution of returns in the heston model with stochastic volatility , cond - mat/0203046 | the evolution of the probability distributions of japan and us major market indices , nikkei 225 and nasdaq composite index , and @xmath0/@xmath1 and @xmath1/@xmath2 currency exchange rates is described by means of the fokker - planck equation ( fpe ) . in order to distinguish and quantify the deterministic and random influences on these financial time series
we perform a statistical analysis of their increments @xmath3 distribution functions for different time lags @xmath4 . from the probability distribution functions at various @xmath4 , the fokker - planck equation for @xmath5
is explicitly derived .
it is written in terms of a drift and a diffusion coefficient .
the kramers - moyal coefficients , are estimated and found to have a simple analytical form , thus leading to a simple physical interpretation for both drift @xmath6 and diffusion @xmath7 coefficients .
the markov nature of the indices and exchange rates is shown and an apparent difference in the nasdaq @xmath7 is pointed out . * key words . * econophysics ; probability distribution functions ; fokker - planck equation ; stock market indices ; currency exchange rates |
from the recent neutrino oscillation experiments it becomes affirmative that neutrinos have masses . the present and near future experiments enter into the stage of precision tests for masses and lepton mixing angles .
in this situation the investigation of the @xmath0 violation effects in the lepton sector has become more and more important . by taking account of the possible leptonic @xmath0 violating phases for majorana neutrinos
, we have obtained the constraints on the lepton mixing angles from the neutrinoless double beta decay ( @xmath4)@xcite@xcite , the @xmath5-@xmath6 conversion and the k decay , @xmath7@xcite . in this paper
, we propose graphical representations of the @xmath0 violating phases which appear in those lepton number violating processes . by using those representations , we derive the allowed regions on the leptonic @xmath0 violating phases from @xmath4 without using any constraints on the mixing angles .
we also try to determine the magnitude of the @xmath0 violating phases by combining the constraints on the neutrino masses and mixing matrix elements from the recent super kamiokande atmospheric neutrino experiment @xcite , the solar neutrino experiment @xcite@xcite@xcite@xcite@xcite , the recent chooz reactor experiment @xcite , and the future kamland reactor experiment @xcite with those from the lepton number violating processes such as @xmath4 .
the amplitudes of those three lepton number violating processes are , in the absence of right - handed weak couplings , proportional to the `` averaged '' masses @xmath8 , @xmath9 and @xmath10 . the `` averaged '' mass @xmath8 defined from @xmath4
is given @xcite by @xmath11 similarly , the `` averaged '' masses @xmath9 defined from @xmath5-@xmath6 conversion and @xmath10 defined from the lepton number violating k decay , @xmath7 are given @xcite @xcite by @xmath12 respectively .
the @xmath0 violating effects are included in the `` averaged '' masses @xmath8 , @xmath9 and @xmath10 defined in eqs.([eq761 ] ) @xmath13 ( [ eq771 ] ) . here
@xmath14 is the maki - nakagawa - sakata ( mns ) left - handed lepton mixing matrix which combines the weak eigenstate neutrino ( @xmath15 and @xmath16 ) to the mass eigenstate neutrino with mass @xmath17 ( @xmath18=1,2 and 3 ) .
the @xmath19 takes the following form in the standard representation @xcite : @xmath20 here @xmath21 , @xmath22 ( @xmath23 ) .
three @xmath0 violating phases , @xmath24 , @xmath25 and @xmath26 appear in @xmath19 for majorana neutrinos @xcite . in this paper
we introduce the graphical representations of the complex masses @xmath27 and @xmath28 .
then , using these representations , we derive the constraints on the @xmath0 violating phases which appear in the lepton number violating processes .
this article is organized as follows . in section 2
we introduce the graphical representations of the complex masses and the @xmath0 violating phases . in section 3
we present constraints on the @xmath0 violating phases from @xmath4 .
constraints from @xmath4 and the neutrino oscillation experiments are discussed in section 4 .
section 5 is devoted to summary .
we now rewrite the complex mass @xmath29 by using the phase convention in eq.([eq772 ] ) as @xmath30 here we have defined the complex masses @xmath31 by [ eq000131 ] @xmath32 we also rewrite the complex mass @xmath33 by using the above @xmath31 as follows : @xmath34 here we have used the unitarity constraint that @xmath35 . furthermore , using @xmath36 , @xmath37 , @xmath38 , @xmath39 and @xmath40 with @xmath41 and @xmath42 , we obtain [ eq013102 ] @xmath43 therefore , the @xmath8 , @xmath44 , and @xmath10 defined in eqs.([eq761 ] ) @xmath13 ( [ eq771 ] ) are reexpressed by the absolute values of averaged complex masses as [ eq013103 ] @xmath45 here the averaged complex masses @xmath46 , @xmath47 , and @xmath48 are defined by @xmath49 now let us introduce graphical representations of the complex value of the @xmath46 , @xmath47 and @xmath48 in a complex mass plane in order to investigate the magnitude of the @xmath0 violating phases in them .
the @xmath46 is the `` averaged '' complex mass of the masses @xmath31 weighted by three mixing elements @xmath50 with the unitarity constraint @xmath51 . therefore , the position of @xmath46 in a complex mass plane is within the triangle formed by the three mass points @xmath31 if the magnitudes of @xmath52 are unknown , which is shown in fig
. 1(a - i ) .
hereafter we refer this triangle as the complex - mass triangle .
this triangle is different from that defined by fogli et al .
@xcite in the sense that ours incorporates the @xmath0 violating phases and masses .
the three mixing elements @xmath50 indicate the division ratios for the three portions of each side of the triangle which are divided by the parallel lines to the side lines of the triangle passing through the @xmath53 ( fig .
1(a - ii ) ) .
the @xmath0 violating phases @xmath54 and @xmath55 represent the rotation angles of @xmath56 and @xmath57 around the origin , respectively .
likewise , the constraints on the positions of @xmath47 and @xmath48 are depicted in figs.1(b ) and 1(c ) .
the position of @xmath47 is given as fig .
the position of @xmath48 in a complex mass plane is within the triangle formed by the three mass points @xmath58 , @xmath59 , and @xmath60 which is shown in fig .
among the lepton number violation processes such as @xmath61 , the @xmath1 conversion and the k decay , @xmath62 , the @xmath4 gives us most restrictive constraints on the @xmath0 violating phases .
therefore , hereafter we concentrate on the @xmath4 and derive constraints on the @xmath0 violating phases from the experimental upper bound on @xmath63 ( we denote it @xmath64 , i.e. , @xmath65 ) .
since @xmath66 , the present experimental upper bound on @xmath63 obtained from the @xmath4 forms the circle in the complex plane and this circle must include the point @xmath46 inside of it .
namely , the allowed region for @xmath46 is the intersection of the inside of the circle of radius @xmath67 around the origin and the inside of the complex - mass triangle which was discussed in section 2 . in the case of @xmath68
, we can obtain the constraints on the @xmath0 violating phases from the allowed region for @xmath46 without using any constraints on the mixing elements @xmath69 as follows . in order to obtain the conditions for the allowed @xmath46 , it is more convenient to survey the forbidden regions for @xmath46 .
it is easily understood from fig 2(a ) that the complex - mass triangle does not overlap with the circle @xmath67 only if the following conditions are satisfied for all @xmath70 and @xmath18 .
@xmath71 here @xmath72 is defined by @xmath73 .
therefore , the allowed region for @xmath46 is the area where , at least , one of the inequalities of eq .
( [ eq020301 ] ) is violated .
since we have @xmath74 , @xmath75 , and @xmath76 , with @xmath54 and @xmath77 in the interval of @xmath78 , we find that majorana @xmath0 violating phases @xmath24 and @xmath79 , must satisfy the following conditions : @xmath80 the allowed region in the @xmath54 vs @xmath77 plane obtained from eq.([eq110580 ] ) is depicted in fig .
eq.([eq110580 ] ) is also useful in the case where the three neutrino masses are almost degenerate and @xmath81 . in this case ,
eq.([eq110580 ] ) reduces to @xmath82 with @xmath83 and the allowed region fig .
2(b ) to fig .
now , we consider the constraints on @xmath84 from the chooz reactor experiment , the recent super - kamiokande atmospheric neutrino experiment , solar neutrino experiments and the future kamland reactor experiment . then , by combining these constraints on @xmath84 with one from @xmath85 we derive the possible constraints on the @xmath0 violating phases by using our graphical representation . in the following discussions we consider three cases for the neutrino mass hierarchy , i.e. , case(a ) : two quasi - degenerate neutrino with @xmath86 , case(b ) : two quasi - degenerate neutrino with @xmath87 and case(c ) : three quasi - degenerate neutrino with @xmath88 . in the case , the oscillation probability for reactor neutrinos in the three - generation model , @xmath89 is given by @xmath90 if @xmath91 the present chooz experiment gives a severe restriction on the mixing angle : @xmath92 in this case , since @xmath93 we obtain @xmath94 on the other hand , the oscillation probability for the atmospheric neutrinos , @xmath95 is @xmath96 the atmospheric @xmath97 deficit in the super kamiokande experiment indicates that @xmath98 namely we obtain @xmath99 from these two constraints , eqs.([eq111 ] ) and ( [ eq222 ] ) , we obtain @xmath100 eq.([eq110604 ] ) imposes the restriction on @xmath101 .
therefore , when combined with the allowed region in the complex mass plane discussed in the section 2 , the position of @xmath46 in our graphical representation is restricted by the chooz and super kamiokande experiments as shown in fig .
3 . we also have the constraints on the mixing angle from the solar neutrino experiments .
they give several separate allowed regions for the position of @xmath46 in our graphical representation as shown in fig .
whether the mixing angle for solar neutrinos is large or small can be determined by the future kamland reactor experiment @xcite .
the future kamland experiment will also lead to the constraint on the @xmath102 and @xmath103 .
since the kamland experiment has the chance to observe a lower order mass difference , @xmath104 we ca nt neglect the term depend on @xmath105 so we rewrite eq.([eq110609 ] ) as follows : @xmath106 \nonumber \\ & = & 1-\left[\frac{2s_3 ^ 2c_3 ^ 2}{\sin^2\left(\frac{\delta m_{12}^2l}{4e}\right ) } + 4 s_1 ^ 2c_1 ^ 2c_3 ^ 4\right ] \sin^2\left(\frac{\delta m_{12}^2l}{4e}\right).\nonumber \\ & \equiv & 1-\xi^2 \sin^2\left(\frac{\delta m_{12}^2l}{4e}\right ) .
\label{eq020701}\end{aligned}\ ] ] here we have used the following conditions , @xmath107 because of their frequent oscillations . let us combine eq.([eq020701 ] ) with the constraint given in eq.([eq110604 ] ) which is obtained from the chooz and super kamiokande experiments .
then the kamland experiment will give the constraints on @xmath102 and @xmath103 , which will restrict the allowed region for the position of @xmath46 as shown in fig .
4 . now , with use of our graphical representation , we proceed to discuss the main subject in this paper : if we have non zero value of @xmath108 , how can we determine the magnitude of majorana @xmath0 phases @xmath109 ?
first we discuss the simple case in which @xmath101 is approximately zero and the large mixing angle solution(lma ) , @xmath110 , is adopted for the solar neutrino problem . in this case , we have @xmath111 given the values of latexmath:[$\langle m_\nu \rangle_{ee},m_1,m_2 , phase @xmath24 is easily obtained from the graphical representation of eq.([eq110605 ] ) .
it goes from fig.1(a - ii ) that the complex - mass triangle gets degenerate to a straight line @xmath113 for @xmath114 case and that the position of @xmath115 moves along the circle with a radius of @xmath116 from the point @xmath117 for changing @xmath24 . on the other hand ,
the measurement of the @xmath108 restricts @xmath46 on the circle with a radius of @xmath108 from the origin .
therefore , the @xmath24 is determined by the intersection of the above two circles as shown in fig.5 . applying the cosine formula to @xmath118 in fig.5 , we find @xmath119 therefore , we obtain @xmath120 it goes from eq.([eq110616 ] ) with the use of @xmath121 that @xmath122 has the lower and upper limits as @xmath123 , which is shown in fig.6 with the definitions of @xmath124 on the other hand , for the case where @xmath125 and the small mixing angle solution(sma ) is adopted for the solar neutrino problem , i.e. , @xmath126 we can not obtain any information about @xmath24 , since we have @xmath127 for @xmath128 or @xmath129 for @xmath130 .
second we consider the case where @xmath131 and the lma solution , @xmath110 , is adopted for the solar neutrino problem . in this case
, we have @xmath132 the graphical representation of eq.([eq022201 ] ) is shown in fig.7 . in fig.7(a ) we consider the case in which the circle of radius @xmath133 around the point @xmath134 ( which we refer as @xmath135 or @xmath136 ) intersects with the circles of radius @xmath137 around the origin at the points @xmath138 and @xmath139 .
we find that @xmath54 is ranging from the argument of @xmath140 to that of @xmath141 as seen in fig.7(a ) .
the relation between @xmath24 and @xmath142 is also derived from eq.([eq022201 ] ) : for fixed @xmath54 , the @xmath142 has two solution @xmath143 and @xmath144 which are determined by the points @xmath145 and @xmath146 as shown in fig.7(b ) . here
the @xmath145 and @xmath146 are the intersections of the circle of radius @xmath147 around the point @xmath148 ( which we refer as @xmath149 ) with the circle of radius @xmath122 around the origin since @xmath150 from eq.([eq022201 ] ) .
thus we obtain the relation between @xmath24 and @xmath142 .
we depict this relation in fig .
the other cases may occur but they can be treated analogously . in this case
, the chooz experiment and the atmospheric neutrino deficit experiment indicates @xmath151 as is seen in fig .
9 . therefore , we can discuss this case with the same way as the case(a ) only by replacing @xmath152 , @xmath102 , and @xmath24 with @xmath153 , @xmath101 , and @xmath154 , respectively .
we assume that all three neutrino masses are almost degenerate , then we have @xmath156 the constraints on the @xmath0 violating phases from this @xmath46 is obtained from the similar discussions as in fig .
7 with only taking @xmath157 in it .
it should be noted that for the case where @xmath158 and @xmath101 is approximately zero , we find @xmath159 which is the same result as one obtained from eq.([eq110616 ] ) with replacing @xmath160 with @xmath161 .
we find from eq.([eq555 ] ) that the following lower limit of @xmath162 is obtained for the large mixing angle solution(lma ) , @xmath163 , of the solar neutrino problem , @xmath164 where the lower limit is realized at @xmath165 .
we have introduced graphical representations of the complex masses , @xmath46 , @xmath47 and @xmath48 whose absolute magnitudes are experimentally observable `` averaged '' masses , @xmath8 , @xmath9 and @xmath10 of the lepton number violation processes such as neutrinoless double beta decay , the @xmath1 conversion and the k decay , @xmath62 , respectively . by using those graphical representations ,
we have investigated how to determine the magnitude of the @xmath0 violating phases from the analysis of the neutrinoless double beta decay .
first we have discussed without using any constraint on the mixing elements @xmath50 and obtained the constraints on the majorana @xmath0 violating phases , eqs.([eq110580 ] ) and ( [ eq110590 ] ) if @xmath166 , from which the allowed region in the @xmath167 plane have been derived and shown in fig .
2 . of course
, we have no constraint on the majorana @xmath0 phases if @xmath168 .
still , eq.([eq110590 ] ) is useful if the three neutrino masses are almost degenerate and @xmath169 .
next by using the constraints on the mixing elements @xmath170 obtained from the recent super kamiokande atmospheric neutrino experiment , the solar neutrino experiment , the recent chooz reactor experiment , and the future kamland reactor experiment , we have further discussed the possible constraints on the majorana @xmath0 violating phases for three cases for the neutrino mass hierarchy , i.e. , case(a ) : two quasi - degenerate neutrino with @xmath86 , case(b ) : two quasi - degenerate neutrino with @xmath87 and case(c ) : three quasi - degenerate neutrino with @xmath88 . in the case(a ) , we have obtained the expression of @xmath171 , eq.([eq110616 ] ) , in terms of @xmath172 , @xmath102 , and @xmath103 for the simple case where @xmath173 with the use of the large mixing angle solution(lma ) for the solar neutrino problem .
the @xmath174 relation is shown in fig .
5 . using @xmath175 , we have found that the @xmath108 has the lower and upper limits as given in eq.([eq110626 ] ) .
we have also obtained the relation between @xmath176 for @xmath131 case which is shown in figs . 7 and 8 .
we can discuss the case(b ) using the same way as the case(a ) by replacing @xmath152 , @xmath102 , and @xmath24 with @xmath153 , @xmath101 , and @xmath154 , respectively . in the case(c ) , we have obtained the expression of @xmath162 given in eq.([eq555 ] ) for the simple case where @xmath158 and @xmath173 with the use of the lma solution for the solar neutrino problem . from this relation
we have found that the lower limit of @xmath162 is given by @xmath177 for the lma solution .
99 t. fukuyama , k. matsuda , and h. nishiura , phys .
* d57 * , 5844 ( 1998 ) . t. fukuyama , k. matsuda , and h. nishiura , mod .
a13 * , 2279 ( 1998 ) .
h. nishiura , k. matsuda , and t. fukuyama , mod .
* a14 * , 433 ( 1999 ) . t. kajita , talk at neutrino 98 ( takayama , japan , june 1998 ) .
hirata et al .
lett . * 63 * , 6 ( 1989 ) ; n. hata , s. bludman , and p. langacker , phys .
* d49 * , 3622 ( 1994 ) ; n. hata and p. langacker , phys . rev . * d50 * , 632 ( 1994 ) .
r. davis jr .
, proc . of the 6th international workshop on neutrino telescope , venezia ,
march 1994 , edited by m.baldo ceolin .
p. anselmann et al . , phys . lett . * b285 * , 376 ( 1992 ) , ibid * b357 * , 237 ( 1995 ) .
abazov et al . , phys .
* 67 * , 3332 ( 1991 ) ; j.n .
abdurashitov et al . , phys . lett . *
b328 * , 234 ( 1994 ) .
fogli , e. lisi , and d. montanino , phys . rev . *
d54 * , 2048 ( 1996 ) .
m. apollonio et al , hep - ex/9907037 and phys
b338 * , 383 ( 1998 ) .
g. gratta , hep - ex/9905011 .
m. doi , t. kotani , h. nishiura , k. okuda , and e. takasugi , phys .
* 102b * , 323 ( 1981 ) ; m. doi , t. kotani , and e. takasugi , prog .
* 83 * , 1 ( 1985 ) ; w.c .
haxton and g.j .
stophenson jr . , prog .
* 12 * , 409 ( 1984 ) .
m. doi , t. kotani , and e. takasugi , prog .
* 83 * ( 1985 ) .
h. nishiura , phys . lett . *
157b * , 442 ( 1985 ) .
bilenky , j. hosek , and s.t .
petcov , phys .
94b * , 495 ( 1980 ) ; m. doi , t. kotani , h. nishiura , k. okuda , and e. takasugi , phys . lett .
* 102b * , 323 ( 1981 ) .
b. achkar et al . , nucl .
b434 * , 503 ( 1995 ) .
k. winter , nucl .
b38 * ( proc . suppl . ) , 211 ( 1995 ) .
l. dilella , nucl . phys .
* b31 * ( proc
. suppl . ) , 319 ( 1993 ) .
particle data group , r.m .
barnet et al . , phys . rev .
* d54 * , 1 ( 1997 ) .
fig.1 graphical representations of the @xmath0 violating phases and the complex masses @xmath46 , @xmath47 and @xmath48 defined in eqs.([eq2013])-([eq2015 ] ) .
( a - i ) the complex - mass triangle for @xmath46 is formed by the three points @xmath31 defined in eqs.([eq2004])-([eq2006 ] ) .
the allowed position of @xmath46 is in the intersection ( shaded area ) of the inside of this triangle and the inside of the circle of radius @xmath67 around the origin .
( a - ii ) the relations between the position of @xmath46 and @xmath178 components of mns mixing matrix .
( b ) the position of @xmath47 .
the position of @xmath47 is at the vertex of the parallelogram of which the other vertexes are at @xmath58 , @xmath179 , and @xmath180 .
namely , rotate @xmath56 clockwise by @xmath181 around @xmath58 and scale down by @xmath182 . from this point
extend the line parallel to the side of @xmath183 by @xmath184 , then we obtain the position of @xmath47 . ( c ) the complex - mass triangle for @xmath48 ( thick lines ) .
the allowed position of @xmath48 is within the triangle formed by the three points @xmath58 , @xmath59 , and @xmath60 .
fig.2 the restrictions of cp violating phases @xmath54 and @xmath185 from @xmath4 with arguments independent of @xmath84 .
( a ) the allowed region of @xmath46 is inside of the complex - mass triangle overlapped with the inside of the circle of radius @xmath186 .
the case where the conditions @xmath187 are satisfied for all @xmath70 and @xmath18 is excluded since the triangle and the circle can not overlap each other .
here we define @xmath188 , @xmath189 , @xmath190 , and , therefore , @xmath191 .
( b ) the allowed region ( shaded area ) in the @xmath54 vs @xmath77 plane for @xmath192 case .
( c ) the special case of fig.2 ( b ) for the case in which three neutrinos have almost degenerate masses with @xmath193 . here
@xmath83 .
fig.3 the allowed region ( shaded area ) of @xmath46 for the case ( a ) from the chooz , the atmospheric @xmath97 deficit and the solar neutrino experiments .
the chooz and the atmospheric @xmath97 deficit experiments restrict the position of @xmath46 in the neighborhood of the side @xmath113 .
the large mixing angle ( lma ) and the small mixing angle ( sma ) solutions for the solar neutrino problem give separate allowed regions .
fig.4 constraints from kamland experiments .
the contours of p@xmath194 in our complex - mass triangle .
( @xmath195 is defined in eq.(4.7 ) . )
they are plotted at the interval of @xmath196 of @xmath197 for the typical values of @xmath198 .
( a ) ; @xmath199 .
( b ) ; @xmath200 .
( c ) ; @xmath201 .
fig.5 the determination of @xmath24 for @xmath114 in the case(a ) .
the @xmath24 is determined from the point @xmath149 which is the intersection of two circles ; the circle of radius @xmath122 around the origin and that of radius of @xmath202 around @xmath134 ( which we refer as @xmath135 ) .
the line @xmath203 is parallel to the line @xmath204 .
here @xmath205 , @xmath206 , @xmath207 .
fig.6 the relation between @xmath208 and @xmath122 which is obtained from eq.(4.11 ) .
the solid line is for the case , @xmath209 . for @xmath210 case
, the relation has a band structure ( shaded region ) .
here we define @xmath211 and @xmath212 .
fig.7 the constraint on @xmath24 and @xmath142 for @xmath213 in the case(a ) .
( a ) the @xmath54 is ranging from the argument of @xmath140 ( angle between @xmath140 and the horizontal axis ) to that of @xmath141 . here
@xmath214 and @xmath207 . in this diagram
we consider the case where the circle of radius @xmath133 around @xmath135 intersects with the circles of radius @xmath137 at @xmath138 and @xmath139 .
( b ) the @xmath142 has two solution @xmath143 and @xmath144 for fixed @xmath54 since @xmath150 .
the dotted line is the circle of radius @xmath147 around @xmath149 which intersects the circle of radius @xmath122 around the origin at @xmath145 and @xmath146 . here
we refer the point @xmath148 as @xmath149 .
fig.9 the allowed region of @xmath46 for the case ( b ) from the chooz and the super kamiokande experiments .
the position of @xmath46 is restricted in the shaded area which is in the neighborhood of the edge @xmath221 . | we examine the constraints on the mns lepton mixing matrix from the present and future experimental data of the neutrino oscillation and lepton number violation processes . we introduce a graphical representation of the @xmath0 violation phases which appear in the lepton number violation processes such as neutrinoless double beta decay , the @xmath1 conversion , and the k decay , @xmath2
using this graphical representation , we derive the constraints on the @xmath0 violation phases in the lepton sector .
# 1#2 # 1#2#1#2 # 1#2#3 # 1@xmath3#1 |
the discovery of neutrino oscillations by the super - kamiokande ( super - k ) experiment in 1998 @xcite opened a new window to explore physics beyond the standard model ( bsm ) .
evidence of neutrino oscillations is one of the most convincing experimental proofs known today for the existence of bsm physics at work .
the mixing parameters of neutrinos were found to be remarkably different from those of quarks , which suggests the presence of an unknown flavor symmetry waiting to be explored .
the extremely small masses of neutrinos compared with those of their charged partners lead to the preferred scenario of a seesaw mechanism @xcite , in which small neutrino masses are a reflection of the ultra - high energy scale of bsm physics .
furthermore , a theoretical framework called leptogenesis points to the intriguing possibility that @xmath0 asymmetries related to flavor mixing among the three generations of neutrinos may have played an important role in creating the observed matter - antimatter asymmetry in the universe @xcite .
this makes a study of the full picture of neutrino masses and mixings and the measurement of the @xmath0 asymmetry in the neutrino sector among the most important and urgent subjects in today s elementary particle physics world .
@xmath0 asymmetry in the neutrino sector arising from the presence of the phase which corresponds to the kobayashi - maskawa phase @xcite in the quark sector , can only be seen if all the three mixing angles governing neutrino oscillations differ from zero .
the super - k detector has successfully measured all three angles .
the angle @xmath16 was first measured in atmospheric neutrino observations @xcite , @xmath17 was constrained in solar neutrino observations @xcite ( together with another water cherenkov detector sno @xcite ) , and the evidence of non - zero @xmath18 was found by t2k @xcite which used super - k as the far detector . in 2013 , t2k established @xmath19 oscillation with 7.3@xmath13 significance , leading the way towards @xmath0 violation measurements in neutrinos @xcite in combination with precise measurements of @xmath18 by reactor neutrino experiments @xcite .
the highly successful super - k program indicates that hyper - kamiokande ( hyper - k ) is well placed to discover @xmath0 violation . in this paper ,
the physics potential of a long baseline neutrino experiment using the hyper - kamiokande detector and a neutrino beam from the j - parc proton synchrotron is presented .
the hyper - k detector is designed as a next generation underground water cherenkov detector that serves as a far detector of a long baseline neutrino oscillation experiment for the j - parc neutrino beam and as a detector capable of observing proton decays , atmospheric and solar neutrinos , and neutrinos from other astrophysical origins .
the baseline design of hyper - k is based on the well - proven technologies employed and tested at super - k .
hyper - k consists of two cylindrical tanks lying side - by - side , the outer dimensions of each tank being @xmath20 the total ( fiducial ) mass of the detector is 0.99 ( 0.56 ) million metric tons , which is about 20 ( 25 ) times larger than that of super - k .
a proposed location for hyper - k is about 8 km south of super - k ( and 295 km away from j - parc ) and 1,750 meters water equivalent ( or 648 m of rock ) deep .
the inner detector region is viewed by 99,000 20-inch pmts , corresponding to the pmt density of @xmath21 photo - cathode coverage ( the same as the second phase of super - k ) .
the schematic view of the hyper - k detector is illustrated in fig .
[ fig : hk - schematic - view ] .
in addition to the long baseline neutrino oscillation experiment that is the main focus of this paper , hyper - k will provide a rich program in a wide range of science @xcite .
the scope of the project includes observation of atmospheric and solar neutrinos , proton decays , and neutrinos from other astrophysical origins .
the physics potential of hyper - k is summarized in table [ tab : intro : phys ] . [ cols="<,<,<",options="header " , ] [ tab:23sensitivity ] the result shown above is obtained with @xmath14 and @xmath22 as free parameters as well as @xmath23 and @xmath5 , with a nominal parameters shown in table [ tab : oscparam ] .
the use of the @xmath24 sample in addition to @xmath25 enables us to also precisely measure @xmath14 and @xmath22 .
figure [ fig : theta23 - 0.50 ] shows the 90% cl allowed regions for the true value of @xmath26 together with the 90% cl contour by t2k @xmath12 disappearance measurement @xcite .
hyper - k will be able to provide a precise measurement of @xmath14 and @xmath22 .
figure [ fig : theta23 - 0.45 ] shows the 90% cl allowed regions on the @xmath14-@xmath22 plane , for the true values of @xmath27 and @xmath28 ev@xmath29 . for the determination of @xmath16 , a @xmath12 disappearance measurement provides precise measurement of @xmath30 .
however , when @xmath31 , there are two possible solutions ( @xmath16 and @xmath32 ) which give the same @xmath30 .
this is known as the octant degeneracy . as seen from eq .
[ eq : cpv - oscprob ] , @xmath11 appearance measurement can determine @xmath33 .
in addition , the reactor experiments provide almost pure measurement of @xmath23 . thus , the combination of those complimentary measurements is known to be able to resolve this degeneracy if @xmath16 is sufficiently away from @xmath34 @xcite . as shown in fig . [
fig : theta23 - 0.45 ] , with a constraint on @xmath23 from the reactor experiments , hyper - k measurements can resolve the octant degeneracy and precisely determine @xmath14 .
the expected precision of @xmath22 and @xmath14 for true @xmath35 with reactor constraint on @xmath23 is summarized in table [ tab:23sensitivity ] .
discrimination of the wrong hierarchy hypothesis as a function of the assumed true value of @xmath36 .
bottom : the discrimination between the wrong octant for each value of @xmath36 .
the uncertainty from @xmath5 is represented by the thickness of the band .
vertical dashed lines indicate 90% confidence intervals of @xmath14 from the t2k measurement in 2014 @xcite . , title="fig:",scaledwidth=60.0% ] discrimination of the wrong hierarchy hypothesis as a function of the assumed true value of @xmath36 .
bottom : the discrimination between the wrong octant for each value of @xmath36 .
the uncertainty from @xmath5 is represented by the thickness of the band .
vertical dashed lines indicate 90% confidence intervals of @xmath14 from the t2k measurement in 2014 @xcite . , title="fig:",scaledwidth=60.0% ] atmospheric neutrinos can provide an independent and complementary information to the accelerator beam program on the study of neutrino oscillation .
for example , through the matter effect inside the earth , a large statistics sample of atmospheric neutrinos by hyper - k will have a good sensitivity to the mass hierarchy and @xmath16 octant . assuming a 10 year exposure , hyper - k s sensitivity to the mass hierarchy and the octant of @xmath16 by atmospheric neutrino data are shown in fig .
[ fig : atmsens ] . depending on the true value of @xmath16 the sensitivity changes considerably , but for all currently allowed values of this parameter the mass hierarchy sensitivity exceeds @xmath6 independent of the assumed hierarchy . if @xmath16 is non - maximal , the atmospheric neutrino data can be used to discriminate the octant at @xmath6 if @xmath37 or @xmath38 . in the previous sections ,
the mass hierarchy is assumed to be known prior to the hyper - k measurements .
this is a reasonable assumption considering the increased opportunities , thanks to a large value of @xmath18 , of ongoing and proposed projects for mass hierarchy determination .
however , even if the mass hierarchy is unknown before the start of experiment , hyper - k itself will be able to determine it with the atmospheric neutrino measurements .
values for accelerator and atmospheric neutrino measurements assuming that the mass hierarchy is unknown .
the true mass hierarchy is normal hierarchy and the true value of @xmath39 .
bottom : by combining the two measurements , the sensitivity can be enhanced . in this example
study , the @xmath40 is simply added .
[ fig : lbl - atm],title="fig:",scaledwidth=60.0% ] values for accelerator and atmospheric neutrino measurements assuming that the mass hierarchy is unknown .
the true mass hierarchy is normal hierarchy and the true value of @xmath39 .
bottom : by combining the two measurements , the sensitivity can be enhanced . in this example
study , the @xmath40 is simply added .
[ fig : lbl - atm],title="fig:",scaledwidth=60.0% ] because hyper - k will observe both accelerator and atmospheric neutrinos with the same detector , the physics capability of the project can be enhanced by combining two complementary measurements . as a demonstration of such a capability
, a study has been done by simply adding @xmath40 from two measurements , although in a real experiment a more sophisticated analysis is expected . assuming the true mass hierarchy of normal hierarchy and the true value of @xmath39 , the values of expected @xmath40 as a function of @xmath5 for each of the accelerator and atmospheric neutrino measurements , _ without _ assumption of the prior mass hierarchy knowledge , are shown in the top plot of fig .
[ fig : lbl - atm ] . for the accelerator neutrino measurement
, there is a second minimum near @xmath41 because of a degeneracy with mass hierarchy assumptions . on the other hand
, the atmospheric neutrino measurement can discriminate the mass hierarchy , but the sensitivity to the @xmath0 violating phase @xmath5 is worse than the accelerator measurement . by adding the information from both measurements , as shown in the bottom plot of fig .
[ fig : lbl - atm ] , the fake solution can be eliminated and a precise measurement of @xmath5 will be possible .
the sensitivity to leptonic @xmath0 asymmetry of a long baseline experiment using a neutrino beam directed from j - parc to the hyper - kamiokande detector has been studied based on a full simulation of beamline and detector . with an integrated beam power of 7.5 mw@xmath110@xmath2 sec
, the value of @xmath5 can be determined to better than 19@xmath42 for all values of @xmath5 and @xmath0 violation in the lepton sector can be observed with more than 3 @xmath13 ( 5 @xmath13 ) significance for 76% ( 58% ) of the possible values of @xmath5 .
using both @xmath11 appearance and @xmath12 disappearance data , a precise measurement of @xmath14 will be possible .
the expected 1@xmath13 uncertainty is 0.015(0.006 ) for @xmath15 .
this work was supported by mext grant - in - aid for scientific research on innovative areas number 25105004 , titled `` unification and development of the neutrino science frontier . ''
in addition , participation of individual researchers has been further supported by funds from jsps , japan ; the european union erc-207282 , h2020 rise - ga644294-jennifer and h2020 rise - ga641540-skplus ; rsf , rfbr and mes , russia ; jsps and rfbr under the japan - russia research cooperative program . | hyper - kamiokande will be a next generation underground water cherenkov detector with a total ( fiducial ) mass of 0.99 ( 0.56 ) million metric tons , approximately 20 ( 25 ) times larger than that of super - kamiokande .
one of the main goals of hyper - kamiokande is the study of @xmath0 asymmetry in the lepton sector using accelerator neutrino and anti - neutrino beams . in this paper ,
the physics potential of a long baseline neutrino experiment using the hyper - kamiokande detector and a neutrino beam from the j - parc proton synchrotron is presented .
the analysis uses the framework and systematic uncertainties derived from the ongoing t2k experiment . with a total exposure of 7.5 mw @xmath1 10@xmath2 sec integrated proton beam power ( corresponding to @xmath3 protons on target with a 30 gev proton beam ) to a @xmath4-degree off - axis neutrino beam , it is expected that the leptonic @xmath0 phase @xmath5 can be determined to better than 19 degrees for all possible values of @xmath5 , and @xmath0 violation can be established with a statistical significance of more than @xmath6 ( @xmath7 ) for @xmath8 ( @xmath9 ) of the @xmath10 parameter space . using both @xmath11 appearance and @xmath12 disappearance data ,
the expected 1@xmath13 uncertainty of @xmath14 is 0.015(0.006 ) for @xmath15 . |
many algorithms , such as spectral clustering , kernel principal component analysis or more generally kernel - based methods , are based on estimating eigenvalues and eigenvectors of integral operators defined by a kernel function , from a given random sample . to set the context from a statistical point of view ,
let @xmath0 be an unknown probability distribution on a compact space @xmath1 and let @xmath2 be a kernel on @xmath1 .
the goal is to estimate the integral operator @xmath3 from an i.i.d .
random sample drawn according to @xmath4 + a first study on the relationship between the spectral properties of a kernel matrix and the corresponding integral operator can be found in @xcite for the case of a symmetric square integrable kernel @xmath5 they prove that the ordered spectrum of the kernel matrix @xmath6 converges to the ordered spectrum of the kernel integral operator @xmath7 .
connections between this empirical matrix and its continuous counterpart have been subject of much research , for example in the framework of kernel - pca ( e.g. @xcite , @xcite , @xcite ) and spectral clustering ( e.g. @xcite ) . in @xcite , the authors study the connection between the spectral properties of the empirical kernel matrix @xmath8 and those of the corresponding integral operator @xmath7 by introducing two extension operators on the ( same ) reproducing kernel hilbert space defined by @xmath2 , that have the same spectrum ( and related eigenfunctions ) as @xmath9 and @xmath7 respectively .
in such a way they overcome the difficulty of dealing with objects ( @xmath10 and @xmath7 ) operating in different spaces . +
the integral operator @xmath7 is related to the gram operator @xmath11 where @xmath12 denotes the reproducing kernel hilbert space defined by the kernel @xmath2 and @xmath13 the corresponding feature map .
+ the main objective of this paper is to estimate gram operators on ( infinite - dimensional ) hilbert spaces .
some bounds on the deviation of the empirical gram operator from the true gram operator in separable hilbert spaces can be found in @xcite in the case of gaussian random vectors .
+ let us introduce some notation .
we denote by @xmath14 a separable hilbert space and by @xmath15 a ( possibly unknown ) probability distribution on @xmath14 . remark that with the above notation @xmath16 our goal is to estimate the gram operator @xmath17 defined as @xmath18 from an i.i.d . sample drawn according to @xmath19 our approach consists in first considering the finite - dimensional setting where @xmath20 is a random vector in @xmath21 and then in generalizing the results to the infinite - dimensional case of separable hilbert spaces . to be able to go from finite to infinite dimension
we will establish dimension - free inequalities . to be more precise , we consider the related problem of estimating the quadratic form @xmath22 which rewrites explicitly as @xmath23 in the finite - dimensional setting we construct an estimator of the quadratic form @xmath24 and we provide non - asymptotic dimension - free bounds for the approximation error that hold under weak moment assumptions .
observe that in the finite - dimensional case the quadratic form @xmath24 can be seen as the quadratic form associated to the gram matrix @xmath25 observe also that the study of the gram matrix is of interest in the case of a non - centered criterion and it coincides , in the case of centered data ( i.e. @xmath26=0 $ ] ) , with the study of the covariance matrix @xmath27 ) ( x - \mathbb e[x])^{\top}\right].\ ] ] many theoretical results have been proposed on the estimation of covariance matrices , e.g. @xcite , @xcite , @xcite .
these results follow from the study of random matrix theory and use as an estimator of @xmath28 the matrix obtained by replacing the unknown probability distribution @xmath29 with the sample distribution @xmath30 in @xcite the non - commutative khintchine inequality is used to obtain bounds on the sample covariance matrix of a bounded random vector .
non - asymptotic results are obtained in @xcite as a consequence of the analysis of random matrices with independent rows , while in @xcite the author uses an extension of the bernstein inequality to matrices .
however , such an empirical estimator becomes less efficient when the data have a long tail distribution . in
@xcite the author presents a different estimator based on the geometrical median which is more robust that the classical empirical one .
+ we first present a way to construct a robust estimator of the gram matrix @xmath28 in finite dimension and then we extend the results to the infinite - dimensional case .
the paper is organized as follows .
section [ sec1 ] deals with the finite - dimensional case .
we provide a new robust estimator of the gram matrix @xmath28 and we use a pac - bayesian approach to obtain non - asymptotic dimension - independent bounds of its approximation error . in section [ sec2 ]
we extend the results to the infinite - dimensional case , taking advantage of the fact that they are independent of the dimension of the ambient space . in section [ sec_emp ]
we propose some empirical results to show the performance of our estimator . in appendix [ emp_g ]
we compare from a theoretical point of view the behavior of our robust estimator to the one of the classical empirical estimator .
finally in appendix [ sec3 ] we extend the results to estimate the expectation of a symmetric random matrix and we consider the problem of estimating the covariance matrix in the case when the expectation is unknown .
let @xmath31 be an unknown probability distribution on @xmath21 and let @xmath32 be a random vector of law @xmath19 we denote by @xmath33 the expectation with respect to @xmath29 .
our goal is to estimate the quadratic form @xmath34 , \quad \theta \in \mathbb r^d,\ ] ] ( that computes the energy in the direction @xmath35 ) from an i.i.d . sample @xmath36 drawn according to @xmath19 observe that @xmath24 can be seen as the quadratic form associated to the gram matrix @xmath37.\ ] ] indeed to recover the gram matrix @xmath28 from the above quadratic form it is sufficient to use the polarization identity @xmath38\\ & = \frac{1}{4 } \left [ n(e_i+e_j ) - n(e_i - e_j ) \right]\end{aligned}\ ] ] where @xmath39 is the canonical basis of @xmath40 a classical empirical estimator of the quadratic form @xmath24 is @xmath41 obtained by replacing the unknown probability distribution @xmath29 with the sample distribution .
however , as shown in @xcite , if the distribution of @xmath42 has a heavy tail for some values of @xmath35 , the quality of the approximation provided by the classical empirical estimator can be improved , using some @xmath43-estimator with a suitable influence function and a scale parameter depending on the sample size .
thus , for any @xmath44 and any @xmath45 , we consider @xmath46 where the function @xmath47 , defined as @xmath48 is symmetric non - decreasing and bounded and satisfies @xmath49 introduce @xmath50 and observe that , since the function @xmath51 is continuous , @xmath52 as soon as @xmath53 moreover , since the function @xmath54 is close to the identity in a neighborhood of the origin , @xmath55 and therefore it is natural to consider as an estimator of @xmath24 a quantity related to @xmath56 , for a suitable value of @xmath57 we consider the family of ( robust ) estimators @xmath58 and we observe that , since @xmath59 is homogeneous of degree @xmath60 in @xmath35 , @xmath61 in the following we will use a pac - bayesian approach linked to gaussian perturbations of the parameter @xmath35 to first construct a confidence region for @xmath24 and then define and study a robust estimator by choosing a suitable value @xmath62 for the parameter @xmath63 . + given @xmath44 , we consider the family of gaussian perturbation @xmath64 of mean the true value @xmath35 and covariance matrix @xmath65 where @xmath66 is a free parameter .
+ let @xmath67 be a finite set of possible values of the couple of parameters @xmath68 and @xmath69 its cardinality .
let us introduce @xmath70^{1/4 } \quad \text{and } \quad \kappa = \sup_{\substack{\theta \in \mathbb{r}^d \\ \mathbb e [ \langle \theta , x \rangle^2 ] > 0 } } \frac{\displaystyle\mathbb e \bigl [ \langle \theta , x \rangle^4 \bigr]}{\mathbb e \bigl [ \langle \theta , x \rangle^2 \bigr]^2}\end{aligned}\ ] ] assuming that these two quantities are finite . and
@xmath71 by their upper bounds . ] for any @xmath72 and @xmath73 , we put @xmath74 where @xmath75 [ prop0 ] with probability at least @xmath76 , for any @xmath77 , any @xmath72 , @xmath78 where @xmath79 and @xmath80 are non - increasing functions defined as @xmath81\\ \phi_{\theta,+}(t ) & = t \ , \left ( 1 + \frac{\gamma + \delta \lambda \lvert \theta \rvert^2 / t}{1 - \mu - \gamma - 2 \delta \lambda \lvert\theta \rvert^2 / t } \right)^{-1 } \mathds{1 } \bigg [ \xi+ \mu+\gamma + 2 \delta \lambda \lvert \theta\rvert^2 / t < 1 \bigg].\end{aligned}\ ] ] for the proof we refer to section [ proof_prop0 ] .
+ observe that since those functions depend on @xmath35 only through @xmath82 , if @xmath35 is such that @xmath83 it is natural to omit the dependence of @xmath35 and write @xmath84 and @xmath85 . in the following
we will omit the dependence of @xmath35 of the functions defined in proposition [ prop0 ] , so that we write @xmath84 and @xmath85 instead of @xmath79 and @xmath80 .
[ prop1 ] let @xmath86 be any energy level .
we consider the set @xmath87 and the bound @xmath88 with probability at least @xmath89 , for any @xmath77 , any @xmath72 , @xmath90.\ ] ] we observe that , for any @xmath91 , if @xmath92 then @xmath93 since , by definition @xmath94 and similarly if @xmath95 , then @xmath96 thus , according to the definition of @xmath97 in equation , we get @xmath98 and moreover @xmath99 which concludes the proof . from now on we fix the finite set @xmath100 of all possible values of the couple @xmath68 as @xmath101 where @xmath102 , with @xmath103 , and @xmath104}.\end{aligned}\ ] ] we put @xmath105\ ] ] and we define our robust estimator as @xmath106 [ prop2 ] let us fix a threshold @xmath107 and set the value of the parameter @xmath108 to @xmath109 .
introduce @xmath110 where @xmath111
$ ] denotes the trace of the gram matrix , and @xmath112 \zeta _ * ( \max \ { t , \sigma \ } ) \leq \sqrt{n } \\
+ \infty & \text { otherwise . } \end{cases}\ ] ] with probability at least @xmath89 , for any @xmath77 , @xmath113.\ ] ] for the proof we refer to section [ proof_prop2 ] .
remark that equation of the proof provides a bound for any choice of the parameter @xmath114 and that we reported here only the numerical value of the bound when @xmath115 for the sake of simplicity .
+ assuming any reasonable bound on the sample size we can bound the logarithmic factor @xmath116 hidden in @xmath117 with a relatively small constant . in particular , if we choose @xmath118 , we get @xmath119 .
observe that the bound @xmath120 does not depend explicitly on the dimension @xmath121 of the ambient space .
more specifically , the dimension has been replaced by the entropy term @xmath122 moreover , we do not need to know the exact values of @xmath71 and @xmath123 to compute the estimator and evaluate the bound , it is sufficient to know upper bounds instead .
indeed , if we use those upper bounds to define our estimator , the above result is still true with @xmath71 and @xmath123 replaced by their upper bounds .
+ we also observe that in order to have a meaningful ( finite ) bound we can choose the threshold @xmath124 such that @xmath125 so that @xmath126 for any @xmath127 , assuming that @xmath128 .
more precisely , using the inequality @xmath129 , we see that equation holds when @xmath130 with this choice the threshold @xmath124 decays to zero at speed @xmath131 as the sample size grows to infinity .
remark that the estimator @xmath132 is not necessarily a quadratic form .
we conclude this section by introducing and studying a quadratic estimator of @xmath133 + we observe that proposition [ prop0 ] provides a confidence region for @xmath24 .
define @xmath134 where we recall that @xmath135 and @xmath100 is defined in equation . according to proposition [ prop0 ] , with probability at least @xmath76 , for any @xmath44 , @xmath136 from a theoretical point of view we can consider as an estimator of @xmath137 any quadratic form belonging to the confidence interval @xmath138 $ ] for any @xmath35 .
such a quadratic form exists with probability at least @xmath89 according to equation .
however , from an algorithmic point of view , we would like to impose these constraints only for a finite number of directions @xmath35 .
in particular , in the following we are going to study the properties of a symmetric matrix @xmath139 that satisfies @xmath140 and @xmath141 where @xmath142 is any finite @xmath143-net of the unit sphere @xmath144 , meaning that @xmath145 the matrix @xmath139 can be computed using a convex optimization algorithm as described in @xcite or @xcite .
+ from now on let @xmath1460,s_4 ^ 2]$ ] be a threshold such that @xmath147 .
next proposition provides the analogous for the quadratic form @xmath148 of the dimension - free bound presented in proposition [ prop2 ] for @xmath149 .
[ prop3 ] with the same notation as in proposition [ prop2 ] , with probability at least @xmath150 for any @xmath151 , @xmath152 since for any @xmath153 there is @xmath154 such that @xmath155 we have @xmath156 let us put @xmath157 we observe that , with probability at least @xmath150 @xmath158 where @xmath85 and @xmath84 are defined in proposition [ prop0 ] and depend on @xmath35 only through @xmath82 .
indeed , in the event of probability at least @xmath76 described in equation , @xmath159 since equation also holds for @xmath137 , and in the same way we get @xmath160 which proves equation . according to corollary [ lem1a ] in section [ appx ] we conclude the proof . according to equation , for any @xmath153 @xmath161 where @xmath162 is such that @xmath163
there is still some improvement to bring , since at this stage we are not sure that @xmath139 is non - negative .
decomposing @xmath139 into its positive and negative parts and writing @xmath164 , we can consider as an estimator of @xmath28 the positive semi - definite symmetric matrix @xmath165 as shown in the following proposition .
[ prop4 ] with probability at least @xmath89 , for any @xmath166 , @xmath167 where @xmath120 is defined in proposition [ prop2 ] .
let us put as before @xmath157 for any @xmath166 , there is @xmath154 such that @xmath168 , so that , according to equation , @xmath169 then we deduce that @xmath170 therefore , for any @xmath166 , @xmath171 combining the above equation with proposition [ prop3 ] we conclude the proof .
in this section we extend the results obtained in the previous section to the infinite - dimensional setting .
+ let @xmath14 be a separable hilbert space and let @xmath172 be an unknown probability distribution on @xmath173 we consider the gram operator @xmath17 defined by @xmath174 and we assume @xmath175 where @xmath176 denotes a random vector of law @xmath19 in analogy to the previous section we denote by @xmath137 the quadratic form associated with the gram operator @xmath177 so that @xmath178 we consider @xmath179 an increasing sequence of subspaces of @xmath14 of finite dimension such that @xmath180 and we endow each space @xmath181 with the probability @xmath182 which arises from the disintegration of @xmath29 , meaning that , if @xmath183 is the orthogonal projector on @xmath181 , @xmath184 .
we denote by @xmath185 the quadratic form in @xmath181 associated with the probability distribution @xmath182 and we observe that , for any @xmath186 , we have @xmath187 in the following we consider i.i.d .
samples of size @xmath188 in @xmath14 drawn according to @xmath19 according to proposition [ prop0 ] , the event @xmath189 is such that @xmath190 . since @xmath191 , by the continuity of measure , @xmath192 this means that with probability at least @xmath89 , for any @xmath193 and any @xmath72 , @xmath194 consequently , since @xmath195 , for any @xmath196 , the following result holds .
[ prop_hspace ] with probability at least @xmath89 , for any @xmath196 , @xmath197 where @xmath198 if we do not want to go to the limit , we can use the explicit bound @xmath199 this bound depends on @xmath200 .
we will see in the following another bound that goes uniformly to zero for any @xmath201 when @xmath2 tends to infinity . in the same way , proceeding as already done in the previous section we state the analogous of proposition [ prop2 ] .
[ prop1.40fr ] let @xmath202 be known constants and put @xmath203 where @xmath204 .
define @xmath205 and consider , according to equation , the estimator @xmath206 for any @xmath196 , define @xmath207 by choosing any accumulation point of the sequence @xmath208 .
define the bound @xmath209 where @xmath146 0 , s_4 ^ 2]$ ] is some energy level such that @xmath210 \zeta _ * ( \sigma ) \leq \sqrt{n}.\ ] ] with probability at least @xmath89 , for any @xmath186 , @xmath211.\ ] ] this is a consequence of the fact that @xmath212 and of the continuity of @xmath120 .
as already discuss at the end of proposition [ prop2 ] , any reasonable bound on the sample size @xmath188 allows bounding the logarithmic factor @xmath117 by a relatively small constant .
in particular , putting @xmath118 , we get @xmath119 . in the following
we construct an estimator of the gram operator @xmath177 .
let @xmath213 be an i.i.d .
sample drawn according to @xmath19 define @xmath214 and @xmath215 let @xmath216 be a @xmath143-net of @xmath217 ( where @xmath218 denotes the unit sphere in @xmath14 ) and remark that @xmath216 is finite because @xmath219 .
we compute a linear operator @xmath220 such that @xmath221 and @xmath222 observe that @xmath223 plays the same role as the symmetric matrix @xmath224 in the finite - dimensional setting .
we consider as an estimator of @xmath225 the operator @xmath226 where @xmath227 is the orthogonal projector on @xmath228 .
let us decompose @xmath229 in its positive and negative parts and write @xmath230 [ propq ] for any threshold @xmath86 such that @xmath107 and @xmath231 , with probability at least @xmath89 , for any @xmath201 and for any @xmath2 , @xmath232 \bigl\lvert \max \bigl\ { \langle \theta , \mathcal{q}_+ \theta \rangle_{\mathcal h } , \sigma \bigr\ } & - \max \bigl\ { \langle \pi_k \theta , \mathcal{g } \pi_k \theta \rangle_{\mathcal h } , \sigma \bigr\ } \bigr\rvert \\ & \leq 2 \max \bigl\ { \langle \theta , \mathcal{q}_+\theta \rangle_{\mathcal h } , \sigma \bigr\ } b _ * \bigl ( \min \ { \langle \theta , \mathcal{q}_+ \theta \rangle_{\mathcal h } , s_4 ^ 2 \ } \bigr ) + 7 \delta \sqrt { \mathbf{tr}(\mathcal{g}^2)}.\end{aligned}\ ] ] for the proof
we refer to section [ proof_propq ] .
let us consider @xmath233 an orthonormal basis of eigenvectors of @xmath225 such that the corresponding sequence of eigenvalues @xmath234 is non - increasing .
[ prop1.41fm ] consider some threshold @xmath86 such that @xmath107 and @xmath231 .
with probability at least @xmath89 , for any @xmath201 and for any @xmath2 , @xmath235 where @xmath236 it is enough to observe that @xmath237 and , for any @xmath201 , we have @xmath238 indeed , @xmath239 so that @xmath240 and in the same way @xmath241 remark that we can use this result to bound @xmath242 , using the inequality @xmath243 let us mention to conclude this section another way to extend the results from finite dimension to separable hilbert spaces .
this second method consists in defining directly a gaussian perturbation of the parameter in the hilbert space . indeed , taking a basis
, we can identify any separable hilbert space @xmath244 with the sequence space @xmath245 .
we can take as our prior parameter distribution @xmath246 the law on @xmath247 of a sequence of independent one - dimensional gaussian random variables with mean @xmath248 and variance @xmath249 .
for any @xmath250 , we can then define the posterior @xmath251 as @xmath252 .
one can prove that @xmath253 moreover , when @xmath254 , @xmath255 exists @xmath251-almost surely and is a gaussian process indexed by @xmath256 , with mean @xmath257 = \langle \theta , x \rangle_{\mathcal{h } } , \qquad x \in \mathcal{h},\ ] ] and covariance @xmath258 - \mathbb{e } \bigl [ w(x ) \bigr ] \mathbb{e } \bigl [ w(y ) \bigr ] = \frac{1}{\beta } \langle x , y \rangle_{\mathcal{h } } , \qquad x , y \in \mathcal{h}.\ ] ] using these properties , we can directly generalize to @xmath259 the pac - bayesian bounds stated in @xmath260 in the previous section .
nevertheless , we thought it was interesting to study more explicitly how the infinite dimension case could be approximated by a finite dimension estimator , as done in this section .
we present some empirical results which show the performance of our robust estimator .
we use here a simplified construction that does not lead exactly to the estimator @xmath139 , for which we have proved theoretical results in the previous section , but should nonetheless exhibit the same kind of behavior . +
we use an iterative scheme based on the polarization formula to estimate the coefficients of the gram matrix in an orthonormal basis of eigenvectors and then update this basis iteratively to a basis of eigenvectors of the current estimate .
let @xmath36 be a sample drawn according to the probability distribution @xmath29 and let @xmath261 let @xmath262 and define @xmath263 as the solution of @xmath264=0.\ ] ] in practice we compute @xmath265 using the newton algorithm .
we observe that , when @xmath266 and @xmath63 is suitably chosen , @xmath263 is an approximation of the estimator @xmath149 of the quadratic form @xmath24 .
+ define @xmath267 as the solution obtained when the parameter @xmath63 is set to @xmath268}\ ] ] where @xmath269 , @xmath270 and @xmath271 . + according to @xcite , this value of the scale parameter @xmath63 should be close to optimal for the estimation of a single expectation from an empirical sample distribution .
let @xmath272 be the eigenvalues of the empirical gram matrix @xmath273 , that will be our starting point , and let @xmath274 be a corresponding orthonormal basis of eigenvectors .
we decompose the empirical gram matrix as @xmath275 where @xmath276 is the orthogonal matrix whose columns are the eigenvectors of @xmath273 and @xmath277 is the diagonal matrix @xmath278 .
we observe that , by the polarization formula , @xmath279 , \quad i , j = 1 , \dots , d,\ ] ] where @xmath280 , with @xmath281 , is approximated by @xmath282 taking notation , for any @xmath283 matrix @xmath284 , we define @xmath285 as the @xmath286 matrix of entries @xmath287.\ ] ] let @xmath288 be the matrix whose @xmath289-th row is the vector @xmath290 , so that @xmath291 we update the gram matrix estimate to @xmath292 then we iterate the update scheme , decomposing @xmath293 as @xmath294 where @xmath295 and @xmath296 is a diagonal matrix and computing @xmath297 the inductive update step is more generally the following .
assuming we have constructed @xmath298 , we decompose it as @xmath299 where @xmath300 and @xmath301 is a diagonal matrix and we define the new updated estimator of @xmath28 as @xmath302 we stop this iterative estimation scheme when @xmath303 falls under a given threshold . in the following numerical experiment we more simply performed four updates .
we take as our robust estimator of @xmath28 the last update @xmath298 .
we now present an example of the performance of this estimator , for some i.i.d .
sample of size @xmath304 in @xmath305 drawn according to the gaussian mixture distribution @xmath306 where @xmath307 and @xmath308[0pt]{\large 0 } \\ & & & \ddots & \\ & & \hspace{10pt}\raisebox{8pt}[0pt][0pt]{\large 0 } & & 0.01 \end{array } \right].\ ] ] the gram matrix of @xmath309 is equal to @xmath310[0pt]{\large 0 } \\ & & & \ddots & \\ & & \hspace{10pt}\raisebox{8pt}[0pt][0pt]{\large 0 } & & 0.8095 \end{array } \right].\ ] ] this example illustrates a favorable situation where the performance of the robust estimator is particularly striking when compared to the empirical gram matrix . as it can be seen on the expression of the sample distribution as well as on the configuration plots below ,
this is a situation of intermittent high variance : the sample is a mixture of a rare high variance signal and a frequent low variance more structured signal .
+ we tested the algorithm on 500 different samples drawn according to the gaussian mixture distribution defined above .
random sample configurations are presented in figure [ fig1 ] .
+ figure [ fig2 ] shows that the robust estimator @xmath139 significantly improves the error in terms of square of the frobenius norm when compared to the empirical estimator @xmath311 .
the red solid line represents the empirical quantile function of the errors of the robust estimator , whereas the blue dotted line represents the quantiles of @xmath312 . +
this quantile function is obtained by sorting the 500 empirical errors in increasing order . + the mean of the square distances @xmath313 on 500 trials is @xmath314 , where the indicated mean estimator and confidence interval is the non - asymptotic confidence interval given by proposition 2.4 of @xcite at confidence level @xmath315 . in the case of the empirical estimator ,
the mean is @xmath316 .
the empirical standard deviations are respectively @xmath317 and @xmath318 .
so we see that in this case the robust estimator reliably decreases the error by a factor larger than @xmath317 and also produces errors with a much smaller standard deviation from sample to sample .
+ in appendix [ emp_g ] we show from a theoretical point of view that the two estimators @xmath139 and @xmath311 behave in a similar way in light tail situations .
, the blue dotted line represents the distances @xmath319 .
, height=188 ]
in this section we propose the proofs of the results presented in the previous sections .
more precisely , section [ proof_prop0 ] refers to proposition [ prop0 ] ( on page ) , section [ proof_prop2 ] refers to proposition [ prop2 ] ( on page ) and section [ proof_propq ] to proposition [ propq ] ( on page ) .
finally section [ appx ] is a ( needed ) technical section .
the proof of proposition [ prop0 ] requires a sequence of preliminary results . +
our approach relies on perturbing the parameter @xmath35 with the gaussian perturbation @xmath320 where @xmath66 is a free parameter .
we have @xmath321 let @xmath322 be a random variable drawn according to @xmath323 therefore , @xmath324 is one - dimensional gaussian random variable with mean @xmath325 and variance @xmath326 .
consequently @xmath327 ^ 2 + \mathbf{var}[\langle w , x\rangle ] = \langle \theta , x \rangle^2 + \frac{\lvert x \rvert^2}{\beta}.\ ] ] accordingly we get @xmath328 .
\end{aligned}\ ] ] in order to pull the expectation with respect to @xmath251 out of the influence function @xmath54 , with a minimal loss of accuracy , we introduce the function @xmath329 where @xmath330 $ ] is such that @xmath331 and @xmath332 is defined by the condition @xmath333 using the explicit expression of the first and second derivative of @xmath54 , we get @xmath334 and @xmath335 + \frac{1 + 2 \sqrt{2}}{2 } .$ ] for any @xmath336 , @xmath337 we first prove that @xmath338 .
the inequality is trivial for @xmath339 since @xmath340 for @xmath341,$ ] performing a taylor expansion at @xmath342 we obtain that @xmath343 since @xmath344 for @xmath345 .
finally we observe that , for any @xmath346 , @xmath347 let us now show that @xmath348 for @xmath349 , we have already seen that the inequality is satisfied since @xmath340 moreover we observe that the function @xmath350 is such that @xmath351 and also @xmath352 performing a taylor expansion at @xmath342 we get @xmath353 since for any @xmath354 $ ] , @xmath355 we deduce that @xmath356 in particular , @xmath357 .
recalling that @xmath358 is an increasing function while @xmath359 is constant on the interval @xmath360 , we conclude the proof .
next lemma allows us to pull the expectation with respect to @xmath251 out of the function @xmath359 .
[ l1 ] let @xmath361 be a measurable space .
for any @xmath362 and any @xmath363 , @xmath364 where by definition @xmath365 moreover , @xmath366 to prove equation we observe that performing a taylor expansion of the function @xmath359 at @xmath367 @xmath368 \geq \chi(z ) + \bigl ( h(\theta ) - z \bigr ) \chi'(z ) + \inf \chi '' \ \frac { \bigl ( h(\theta ) - z \bigr)^2}{2},\ ] ] so that , recalling that @xmath369 , we get @xmath370 \ ,
\mathrm{d } \rho ( \theta ) & \geq \chi \bigl ( \int h \ , \mathrm{d } \rho\bigr ) - \frac{1}{8 } \int \left ( h(\theta ) - \int h(\theta ) \mathrm{d } \rho \right)^2 \mathrm{d } \rho(\theta)\\ & = \chi \bigl ( \int h \ , \mathrm{d } \rho\bigr ) - \frac{1}{8 } \mathbf{var } \bigl ( h \ , \mathrm{d } \rho \bigr).\end{aligned}\ ] ] combining equation with the fact that @xmath338 , for any @xmath336 , we obtain that @xmath371 moreover , since @xmath372 we conclude the proof .
applying this result to our problem we obtain @xmath373 \bigr\},\end{gathered}\ ] ] where , putting @xmath374 , @xmath375 and denoting by @xmath376 a centered gaussian random variable , @xmath377 & = \mathbf{var } \bigl [ ( m + w)^2 \bigr ] \\ & = 4 m^2 \sigma^2 + 2 \sigma^4 = \frac{4 \langle \theta , x \rangle^2 \lvert x \rvert^2 } { \beta } + \frac{2 \lvert x \rvert^4}{\beta^2}.\end{aligned}\ ] ] let us remark that , for any @xmath378 and @xmath379 , @xmath380 since @xmath381 .
therefore , taking the expectation with respect to @xmath284 of this inequality and remarking that @xmath284 and @xmath382 have the same probability distribution we get @xmath383 .\ ] ] thus in our context we put @xmath384 , @xmath385 and @xmath386 and we obtain @xmath387 [ lem1.5 ] for any positive constants @xmath388 and any @xmath389 , @xmath390 for any positive real constants @xmath391 , @xmath392 since the function @xmath393 is non - decreasing for @xmath394 it follows that @xmath395.\ ] ] applying this inequality to @xmath396 $ ] and reminding that @xmath397 , we conclude the proof . as a consequence ,
choosing @xmath398 @xmath399 and @xmath400 , we get @xmath401 \ , \mathrm{d } \pi_{\theta}(\theta ' ) , \end{gathered}\ ] ] where @xmath402 + we observe that the above inequality allows to compare @xmath403 with the expectation with respect to the gaussian perturbation @xmath251 . in terms of the empirical criterion @xmath404 we have proved that @xmath405 \ , \mathrm{d } \pi_{\theta}(\theta ' ) .
\end{gathered}\ ] ] we are now ready to use the following general purpose pac - bayesian inequality .
[ pac]let @xmath406 be a prior probability distribution on @xmath260 and let @xmath407 $ ] be a measurable function with @xmath408 .
with probability at least @xmath409 for any posterior distribution @xmath410 , @xmath411 \ , \mathrm{d } \rho(\theta ' ) + \frac{\mathcal{k}(\rho , \nu ) + \log \bigl ( \epsilon^{-1 } \bigr)}{n},\ ] ] where @xmath412 is the kullback divergence of @xmath413 with respect to @xmath414 . by convention , a non measurable event is said to happen with probability at least @xmath415 when it includes a measurable event of probability non - smaller than @xmath415 . for the proof we refer to @xcite or @xcite . in our context
, we consider as prior distribution @xmath416 and the family of posterior distributions consisting in the gaussian perturbations @xmath417 so that the kullback divergence is given by @xmath418 we observe that since the result holds for any choice of the posterior , it allows us to obtain uniform results in @xmath419 more precisely , we apply the above pac - bayesian inequality to @xmath420,\ ] ] where @xmath421 . using the fact that @xmath422 , we get that , with probability at least @xmath409 for any @xmath423 @xmath424 \mathrm d \pi_{\theta } ( \theta ' ) \\ & \hspace{45 mm } + \frac{\beta \|\theta\|^2}{2n } + \frac { \log(\epsilon^{-1})}{n}\\ & = \mathbb e \biggl [ \langle \theta , x \rangle^2 - \lambda+ \frac{1}{2 } \biggl ( \bigl ( \langle \theta , x \rangle^2 - \lambda \bigr)^2 + \frac{4 \langle \theta , x \rangle^2 \lvert x \rvert^2}{\beta } + \frac{2 \lvert x \rvert^4}{\beta^2 } \biggr ) \\ & \hskip45 mm + \frac{c \lvert x \rvert^2}{\beta } \biggl ( \langle \theta , x \rangle^2 + \frac{3 \lvert x \rvert^2}{2 \beta } \biggr ) \ , \biggr ] + \frac{\beta
\lvert \theta \rvert^2}{2 n } + \frac{\log(\epsilon^{-1})}{n}.\end{aligned}\ ] ] to obtain the last line we have used the fact that @xmath425 & = \frac{4 \langle \theta , x \rangle^2 \lvert x \rvert^2}{\beta } + \frac{2 \lvert x \rvert^4}{\beta}. \end{aligned}\ ] ] let us recall the definition of @xmath426 and @xmath71 introduced in equation .
we have defined @xmath427^{1/4 } \quad \text{and } \quad \kappa = \sup_{\substack{\theta \in \mathbb{r}^d \\ \mathbb e [ \langle \theta , x \rangle^2 ] > 0 } } \frac{\displaystyle\mathbb e \bigl
[ \langle \theta , x \rangle^4 \bigr]}{\mathbb e \bigl [ \langle \theta , x \rangle^2 \bigr]^2}.\ ] ] using the cauchy - schwarz inequality , since @xmath428 & \leq \kappa n(\theta)^2\\ \mathbb e[\langle \theta , x\rangle^2 \|x\|^2 ] & \leq \kappa^{1/2 } s_4 ^ 2 n(\theta),\end{aligned}\ ] ] we get that , with probability at least @xmath409 for any @xmath423 @xmath429 ^ 2 + \biggl [ 1 + ( \kappa -1 ) \lambda + \frac{(2+c ) \kappa^{1/2 } s_4 ^ 2}{\beta } \ , \biggr ] \bigl [ n(\theta ) -
\lambda \bigr ] \\
+ \frac{(\kappa-1 ) \lambda^2}{2 } + \frac{(2+c ) \kappa^{1/2 } s_4 ^ 2 \lambda}{\beta } + \frac{(2 + 3c ) s_4 ^ 4}{2 \beta^2 } + \frac{\beta \lvert \theta \rvert^2}{2n } + \frac{\log(\epsilon^{-1})}{n}. \end{gathered}\ ] ] according to the ( compact ) notation introduced in equation , the above inequality rewrites as @xmath430 similarly , observing that @xmath431 we obtain a lower bound for the empirical criterion @xmath432 namely , with probability at least @xmath415 , for any @xmath77 , any @xmath72 , @xmath433 we now combine the two bounds above to get the confidence region for @xmath24 defined in proposition [ prop0 ] .
assume that both equation and equation hold for any @xmath44 , an event that happens with probability at least @xmath434 + let us introduce @xmath435 and @xmath436 \ , z - \gamma - \tau(\theta ) , \qquad z \in \mathbb{r}.\ ] ] we observe that @xmath437 , and consequently @xmath438 for any @xmath439 .
we consider the case when @xmath440 , meaning that @xmath441 in this case , the second degree polynomial @xmath442 has two distinct real roots , @xmath443 and @xmath444 , where @xmath445 ^ 2 - 4 \xi \bigl [ \gamma + \tau(\theta ) \bigr]}}{2 \xi } , \qquad \sigma\in \{1,- 1\}.\ ] ] since equation can also be written as @xmath446 $ ] , we get @xmath447 ^ 2 } { 1 - \mu - \gamma - 2 \tau(\theta ) } + \bigl [ 1 - \mu - \tau(\theta ) \bigr ] \frac{\gamma + \tau(\theta)}{1 - \mu - \gamma - 2 \tau(\theta ) } - \gamma - \tau(\theta ) = 0 , \end{gathered}\ ] ] which implies @xmath448 therefore , since according to equation , for any @xmath449 $ ] , @xmath450 it holds that @xmath451 \cap \bigl ] z_{-1 } , z_{+1 } \bigr [ = \emptyset.\ ] ] observing that @xmath452 , it follows that @xmath453 this proves that , for any @xmath77 satisfying equation , @xmath454 which rewrites as @xmath455 \leq \frac{\lambda}{\widehat{\alpha}(\theta)^2}.\ ] ] moreover , this inequality is trivially true when condition is not satisfied , because its left - hand side is equal to zero and its right - hand side is non - negative . + proving the second part of the proposition requires a new argument and not a mere update of signs in the proof of the first part .
although it may seem at first sight that we are just aiming at a reverse inequality , the situation is more subtle than that .
let us first remark that in the case when @xmath456 is not satisfied , the bound @xmath457 is trivially satisfied because the left - hand size is equal to zero . in the case when equation is true , it holds that @xmath458 , so that @xmath459 = 0 $ ] , and therefore , according to equation , @xmath460 where @xmath461 + since condition can also be written as @xmath462 , it implies that the second order polynomial @xmath463 has two roots and that @xmath464 is on the right of its largest root , which is larger than @xmath60 . on the other hand , we observe that , under condition , putting @xmath465 , we get @xmath466}{1 + \mu - \gamma + \widehat{\tau}(\theta ) } + \gamma + \widehat{\tau}(\theta ) = 0.\ ] ] therefore , when condition is satisfied , @xmath467 which rewrites as equation .
we first observe that , according to proposition [ prop1 ] , with probability at least @xmath89 , @xmath468 = \inf_{(\lambda , \beta ) \in \lambda } b_{\lambda , \beta } \left [ \|\theta\|^{-2}\widetilde n_{\lambda}(\theta ) \right]\end{aligned}\ ] ] since , by definition , @xmath469 are the values which minimize @xmath470.$ ] according to equation , since @xmath471 and @xmath97 is a decreasing function , we get @xmath472 } \biggr).\ ] ] with the same notation as in proposition [ prop1 ] , we introduce the subset @xmath473 of @xmath474 defined as @xmath475 \xi + \mu + \gamma + 4 \delta \lambda / \max \ { t , \sigma \ } & < 1 , \\ \mu + \gamma + 2 \delta \lambda / \max \ { t , \sigma \ } & \leq 1/2 , \\ \text { \rm and } 2 \gamma + \delta \lambda / \max \ { t , \sigma \ } & \leq 1/2 ~~ \bigr\ } \end{aligned}\ ] ] and the function @xmath476 [ lem1.14 ] for any @xmath72 and any @xmath477 ,
@xmath478 we first observe that when @xmath479 then @xmath480 and hence the two inequalities are trivial .
we now assume @xmath481 and we put @xmath482 we prove the second inequality first . since @xmath483 , we have @xmath484 in order to prove the first inequality , we first check that @xmath485 we start observing that , since @xmath486 , \sigma \right\ } \geq \max \ { t , \sigma \ } / [ 1 + b_{\lambda , \beta}(t ) ] , \ ] ] then @xmath487 \frac{\delta \lambda } { \max \ { t , \sigma \ } } \\ & = \xi + \mu + \gamma + 2 \tau + 2 \tau b_{\lambda , \beta}(t).\end{aligned}\ ] ] moreover , as @xmath488 , we get @xmath489 so that @xmath490 , \sigma \ } \leq \xi + \mu + \gamma + 4 \tau < 1,\ ] ] which proves that , indeed , @xmath491 therefore @xmath492 } \\ & = \frac{\left(\gamma + \tau \right ) \left ( 1 + \tau / ( 1 - \mu - \gamma - 2 \tau ) \right)}{1 - \mu - \gamma - 2 \tau - 2 \tau b_{\lambda , \beta}(t ) } , \end{aligned}\ ] ] where in the last line we have used the definition of @xmath493 observing that @xmath494 we obtain @xmath495 considering that @xmath496 since when @xmath481 , it is true that @xmath497 and @xmath498 we conclude that @xmath499 applying the above lemma to our problem we get that , with probability at least @xmath89 , for any @xmath77 , @xmath500.\ ] ] let us recall the definition of the finite set @xmath100 given in .
let @xmath103 and @xmath501 we define @xmath502 where @xmath503}. \end{aligned}\ ] ] we introduce the explicit bound @xmath504 and @xmath505 \zeta ( \max \ { t , \sigma \ } ) \leq \sqrt{n } \\ + \infty & \text { otherwise . } \end{cases}\ ] ] for any @xmath477 , we have @xmath506 we recall that the function @xmath507 is non - increasing so that @xmath508 moreover , since @xmath509 it is sufficient to prove the result for @xmath510 $ ] . + as equation is trivial when @xmath511 we may assume that @xmath512 , so that @xmath513 in particular , by considering only the second term in the definition of @xmath514 we obtain that @xmath515 which implies @xmath516 therefore , since @xmath517,\ ] ] there exists @xmath518 for which @xmath519 we recall that by equation @xmath520 and we observe that @xmath521 defined as @xmath522 are the desired values which optimize @xmath523 we also remark that , by equation , @xmath524 thus , evaluating @xmath525 in @xmath526 we obtain that @xmath527 \frac{\lambda _ * } { \lambda_{\widehat{\jmath } } } \\
\shoveright { + \sqrt{\frac{(2+c ) \kappa^{1/2 } s_4 ^ 2}{2 n \max \
{ t , \sigma \ } } } \left ( \frac{\beta_*}{\beta_{\widehat{\jmath } } } + \frac{\beta_{\widehat{\jmath}}}{\beta_*}\right ) } \\
\leq \sqrt{\frac{2(\kappa-1)}{n } \left[\frac{(2 + 3c ) s_4 ^ 2}{4 ( 2 + c ) k^{1/2 } \max \ { t , \sigma \ } } + \log(k / \epsilon ) \right ] } \cosh \biggl [ \log\left ( \frac{\lambda_{\widehat{\jmath}}}{\lambda _ * } \right ) \biggr ] \\ + \sqrt{\frac{2 ( 2+c ) \kappa^{1/2 } s_4 ^ 2}{n \max \ { t , \sigma\ } } } \cosh \biggl [ \log\left ( \frac{\beta_{\widehat{\jmath}}}{\beta_*}\right ) \biggr].\end{gathered}\ ] ] by equation we get @xmath528 } \cosh \left ( \frac{a}{4 } \right )
\\ + \sqrt{\frac{2 ( 2+c ) \kappa^{1/2 } s_4 ^ 2}{n \max \ { t , \sigma \ } } } \cosh \left ( \frac{a}{2}\right ) .\end{gathered}\ ] ] we also observe that @xmath529 \leq 4n^{-1/2 } \zeta(t),\ ] ] since by definition @xmath530 in the same way , observing that @xmath531 we obtain @xmath532 \gamma_{\widehat{\jmath } } + 4 \delta_{\widehat{\jmath } } \lambda_{\widehat{\jmath } } / \max \ { t , \sigma\ } \\ & \leq \bigl[6 + ( \kappa-1)^{-1 } \bigr]n^{-1/2 } \zeta ( \max \ { t , \sigma\})\end{aligned}\ ] ] and similarly , @xmath533 & \leq 6n^{-1/2 } \zeta ( \max \ { t , \sigma\ } ) ,
\\ 2 \bigl [ 2\gamma_{\widehat{\jmath } } + \delta_{\widehat{\jmath } } \lambda_{\widehat{\jmath } } / \max \ { t , \sigma \ } \bigr ] & \leq 4 n^{-1/2 } \zeta ( \max \ { t , \sigma\}).\end{aligned}\ ] ] this implies that , whenever @xmath512 , then @xmath534 we have then proved that @xmath535 applying the above lemma to equation and observing that , for any @xmath77 , @xmath536
\leq s_4 ^ 2,\ ] ] we obtain that , with probability at least @xmath89 , for any @xmath77 , @xmath537.\ ] ] since by the cauchy - schwarz inequality @xmath538 we get @xmath539 choosing @xmath540 and computing explicitly the constants we conclude the proof .
we observe that it is sufficient to prove that with probability at least @xmath76 @xmath541 where @xmath542 and @xmath543 indeed , if equation holds true , according to corollary [ lem1a ] , @xmath544 which is the analogous , in the infinite - dimensional setting , of proposition [ prop3 ] .
thus , following the proof of proposition [ prop4 ] we obtain the desired bounds .
+ let us now prove equation .
observe that , for any @xmath545 , @xmath546 where @xmath547 is the closest point in @xmath142 to @xmath548 .
since @xmath549 , with probability at least @xmath415 , for any @xmath72 , @xmath550.\ ] ] let us now remark that for any @xmath551 $ ] , we have @xmath552 , so that @xmath553 . therefore @xmath554
\bigr\ } + \eta \\
\leq \lvert \pi_{v_k } \theta \rvert_{\mathcal h}^2 \
\phi_+^{-1 } \circ \phi_-^{-1 } \bigl\ { n \bigl [ \lvert \pi_{v_k } \theta \rvert_{\mathcal h}^{-1 } \bigl ( \lvert \pi_{v_k } \theta \rvert_{\mathcal h } \xi + \bigl ( \pi_k - \pi_{v_k } \bigr ) \theta \bigr ) \bigr ] \bigr\ } + \eta \\ \leq \phi_+^{-1 } \circ \phi_-^{-1 } \bigl\ { n \bigl [ \lvert \pi_{v_k } \theta \rvert_{\mathcal h } \xi + \bigl ( \pi_k - \pi_{v_k } \bigr ) \theta \bigr ) \bigr ] \bigr\ } + \eta \\ \leq \phi_+^{-1 } \circ \phi_-^{-1 } \bigl ( n \bigl ( \pi_k \theta \bigr ) + \eta \bigr ) + \eta . \end{gathered}\ ] ] indeed , @xmath555 and this is a difference of two vectors belonging to the unit ball . in the same way @xmath556 \bigr\ } - \eta \\ \quad \
\geq \lvert \pi_{v_k } \theta \rvert_{\mathcal h}^2 \
\phi_- \circ \phi_+ \bigl\ { n \bigl [ \lvert \pi_k \theta \rvert_{\mathcal h}^{-1 } \bigl ( \lvert \pi_{v_k } \theta \rvert_{\mathcal h }
\xi + \bigl ( \pi_k - \pi_{v_k } \bigr ) \theta \bigr ) \bigr ] \bigr\ } - \eta \\
\geq \phi_- \circ \phi_+ \bigl ( n \bigl ( \pi_k \theta \bigr ) - \eta \bigr ) - \eta\end{gathered}\ ] ] which proves equation . in all this section we use the same notation as in section [ sec1 ] .
let @xmath1460,s_4 ^ 2]$ ] be such that @xmath557 where @xmath558 is defined in proposition [ prop2 ] .
[ lem0a ] the function @xmath559 where @xmath120 is defined in proposition [ prop2 ] , is non - decreasing for any @xmath560 if @xmath561 , then @xmath562 , so that @xmath563 is obviously non - decreasing .
otherwise , @xmath564 so that @xmath565 therefore the function @xmath566 is of the form @xmath567 where @xmath568 , @xmath569 @xmath570 , and the constants @xmath571 , and @xmath572 are positive .
let @xmath573 and observe that @xmath574 and that @xmath575 .
therefore @xmath576 for any @xmath577 , and @xmath578 showing that @xmath566 is non - decreasing .
[ lem2a ] for any @xmath579 such that , for any @xmath72 , @xmath580 and any threshold @xmath86 such that @xmath581 and @xmath107 , we have @xmath582 by symmetry of @xmath108 and @xmath583 , equation is a consequence of equation . + * step 1 .
* we will prove that @xmath584 where @xmath585 is defined in equation . + * case 1 . *
assume that @xmath586 and remark that , since @xmath85 is non - decreasing and @xmath94 , @xmath587 according to equation , where @xmath97 is defined in equation . therefore in this case , @xmath588 but when @xmath589 , @xmath590 because @xmath591 is a non - increasing function of @xmath592 , thus equation implies that @xmath593 since @xmath594 , equation holds true .
+ * case 2 .
* assume now that we are not in * case 1 * , implying that @xmath595 in this case @xmath596\end{gathered}\ ] ] according to equation .
moreover , continuing the above chain of inequalities @xmath597
\\ = \phi_+ \bigl ( \max \ { b- \eta , \sigma \ } \bigr ) \bigl [ 1 - b_{\lambda , \beta}(a + \eta ) \bigr ] \\
\quad \geq \max \ { b - \eta , \sigma \ } \frac{1 - b_{\lambda , \beta}(a + \eta ) } { 1 + b_{\lambda , \beta}(\max\ { b- \eta , \sigma \ } ) } \\
\geq \max \ { b-\eta , \sigma \ } \frac { 1 - b_{\lambda , \beta}(a + \eta ) } { 1 + b_{\lambda , \beta}(a+\eta)}. \end{gathered}\ ] ] therefore @xmath598 according to lemma [ lem1.14 ] .
this concludes the proof of * step 1*. + * step 2 * taking the infimum in @xmath72 in equation , according to equation , we obtain that @xmath599 we can then use the fact that @xmath600 is non - decreasing ( proved in lemma [ lem0a ] ) to deduce that @xmath601 since there is nothing to prove when already @xmath602 .
remark that @xmath603 and that in the same way @xmath604 .
this proves that @xmath605 by symmetry , we can then exchange @xmath108 and @xmath583 to prove the same bounds for @xmath606 , and therefore also for the absolute value of this quantity , which ends the proof of the lemma . as a consequence the following result holds .
[ lem1a ] for any @xmath579 such that , for any @xmath72 , @xmath580 and any threshold @xmath86 such that @xmath581 and @xmath107 , we have @xmath607 this is a consequence of the previous lemma , of the fact that @xmath608 , and of the fact that @xmath609 .
the main goal of section [ sec1 ] is to estimate the gram matrix @xmath610 , where @xmath32 is a random vector of unknown law @xmath611 , from an i.i.d .
sample @xmath612 drawn according to @xmath19 we have constructed a robust estimator of the gram matrix and in section [ sec_emp ] we have shown empirically its performance in the case of a gaussian mixture distribution . in this section we show from a theoretical point of view that the classical empirical estimator @xmath613 behaves similarly to our robust estimator in light tail situations , while it may perform worse otherwise .
+ as already done in section [ sec1 ] , we consider the quadratic form @xmath614 and we denote by @xmath615 the quadratic form associated to the empirical gram matrix @xmath311 .
according to the notation introduced in section [ sec1 ] , let @xmath103 and let @xmath616 where @xmath617 and @xmath204 .
let us put @xmath618 and let us introduce @xmath619 where @xmath620 is defined in equation as @xmath621 at the end of the section we mention some assumptions under which it is possible to give a non - random bound for @xmath622 .
the following proposition , compared with the result obtained for the robust estimator @xmath623 , presented in proposition [ prop2 ] , shows that the different behavior of the two estimators @xmath624 and @xmath625 can appear only for heavy tail data distributions .
[ prop2.8eg ] consider any threshold @xmath86 such that @xmath626 .
with probability at least @xmath89 , for any @xmath166 , @xmath627_+ \bigl [ 1 - b _ * \bigl ( n(\theta ) \bigr ) \bigr]_+}.\ ] ] where @xmath120 is defined in proposition [ prop2 ] . before proving the above proposition
we observe that , also in this case , the bound does not depend explicitly on the dimension @xmath121 of the ambient space and thus the result can be extended to any infinite - dimensional hilbert space .
let @xmath67 be the finite set defined in equation .
we use as a tool the family of estimators @xmath628 introduced in equation , where @xmath629 is defined in equation .
let us put @xmath630 we divide the proof into 4 steps .
+ * step 1 . *
the first step consists in linking the empirical estimator @xmath625 with @xmath631 + we claim that , with probability at least @xmath409 for any @xmath166 , any @xmath632 such that @xmath633 , @xmath634_+^{-1},\ ] ] with the convention that @xmath635 moreover , with probability at least @xmath409 for any @xmath166 , any @xmath72 , such that @xmath636 , @xmath637 we first observe that , according to the definition of @xmath629 , for any threshold @xmath86 , @xmath638 \leq r \bigl ( \lambda^{1/2 } \widetilde{n}_{\lambda}(\theta)^{-1/2 } \theta \bigr ) = r_{\lambda } \bigl ( \widehat{\alpha}(\theta ) \ , \theta \bigr ) \leq 0,\ ] ] where we have used the fact that the function @xmath54 , introduce in equation , is non - decreasing .
moreover @xmath639 as soon as @xmath640 and this holds true , according to proposition [ prop0 ] , with probability at least @xmath415 , for any @xmath166 and any @xmath72 such that @xmath633 . indeed , by proposition [ prop0 ] , with probability at least @xmath415 , for any @xmath166 , any @xmath632 @xmath641 defining @xmath642 we get @xmath643.\end{gathered}\ ] ] in the same way , with probability at least @xmath415 , for any @xmath166 , any @xmath72 such that @xmath633 , we obtain @xmath644.\ ] ] we remark that the derivative of @xmath645 is @xmath646\\ \displaystyle \frac{\frac{z^2}{2}}{1+z+\frac{z^2}{2 } } & \text{if } \ z \in [ -1,0]\\ \displaystyle\frac{\frac{z^2}{2}}{1-z+\frac{z^2}{2 } } & \text{if } \ z \in [ 0,1 ] , \end{cases}\ ] ] showing that @xmath647 , and therefore that @xmath645 is a non - decreasing function satisfying @xmath648 applying equation to equation we obtain @xmath649 where we have used the fact that @xmath650 since , by the cauchy - schwarz inequality , @xmath651 we get @xmath652 which proves the first inequality .
similarly , since @xmath645 in non - decreasing , we obtain that , with probability at least @xmath415 , for any @xmath166 , any @xmath72 such that @xmath653 , @xmath654 where the last inequality follows from equation . +
* this is an intermediate step .
we claim that , with probability at least @xmath89 , for any @xmath166 , any @xmath72 , any @xmath655 , @xmath656_+^{-1 } \\ \max \bigl\ { \bar{n}(\theta ) , \sigma \bigr\ } & \leq \phi_-^{-1 } \bigl ( \max \ { n(\theta ) , \sigma \ } \bigr)\biggl [ 1 - \tau_{\lambda } \bigl ( \bar{n}(\theta ) \bigl [ 1 - \tau_{\lambda}(\sigma ) \bigr]_+ \bigr ) \biggr]_+^{-1 } \\ \bar{n}(\theta ) & \geq \biggl ( 1 - \frac{\lambda^2}{3 } \biggr)_+ \phi_+\bigl ( n(\theta ) \bigr ) \end{aligned}\ ] ] where @xmath85 and @xmath84 are defined in proposition [ prop0 ] .
+ we consider the threshold @xmath657 where we have used the fact that , by definition , @xmath658 for any @xmath659 we assume that we are in the intersection of the two events of proposition [ prop0 ] , which holds true with probability at least @xmath660 so that @xmath661 according to * step 1 * , choosing as a threshold @xmath662 we get @xmath663_+^{-1},\end{aligned}\ ] ] ( where @xmath664 is still defined with respect to @xmath124 ) , so that , according to equation , @xmath665_+^{-1}.\ ] ] as a consequence , recalling the definition of @xmath666 we have @xmath667_+^{-1}.\ ] ] thus , observing that @xmath668_+^{-1},\ ] ] we obtain the first inequality . to prove the second inequality , we use equation once to see that @xmath669_+\geq \bar{n}(\theta ) \bigl [ 1 - \tau_{\lambda } ( \sigma ) \bigr]_+,\ ] ] and we use it again to get @xmath670_+^{-1}\\ & \leq \phi_-^{-1 } \bigl ( \max \ { n(\theta ) , \sigma \ } \bigr ) \bigl [ 1 - \tau_{\lambda } \bigl ( \bar{n}(\theta ) [ 1 - \tau_{\lambda}(\sigma ) ] _ + \bigr ) \bigr]_+^{-1}.\end{aligned}\ ] ] to complete the proof of the second inequality , it is enough to remark that @xmath671_+ \bigr ) \bigr]_+^{-1}.\ ] ] to prove the last inequality , it is sufficient to remark that @xmath672 by proposition [ prop0 ] and hence , when @xmath673 @xmath674 on the other hand , when @xmath675 this inequality is also obviously satisfied . + * step 3 . * we now prove that , with probability at least @xmath89 , for any @xmath166 , any @xmath72 , any @xmath655 , @xmath676_+ \bigl [ 1 - b_{\lambda , \beta } \bigl ( n(\theta ) \bigr ) \bigr]_+ } , \\ 1 - \frac{\max \ { \bar{n}(\theta ) , \sigma \ } } { \max \ { n(\theta ) , \sigma \ } } & \leq b_{\lambda , \beta } \bigl(n(\theta ) \bigr ) + \frac{\lambda^2}{3},\end{aligned}\ ] ] where @xmath97 is defined in equation and @xmath585 in equation .
+ we observe that , according to * step 2 * , @xmath677_+ ^{-1 } \\ & \leq \frac{\max \ { n(\theta ) , \sigma \ } } { \bigl [ 1 - \tau_{\lambda } \bigl ( n(\theta ) \bigr ) \bigr]_+ \bigl [ 1 -b_{\lambda , \beta } \bigl ( n(\theta )
\bigr ) \bigr]_+ } , \end{aligned}\ ] ] which implies @xmath678_+ } + \frac { \tau_{\lambda } \bigl ( n(\theta ) \bigr)}{\bigl [ 1 - \tau_{\lambda } \bigl ( n(\theta ) \bigr ) \bigr]_+ \bigl [ 1 -b_{\lambda , \beta } \bigl ( n(\theta )
\bigr ) \bigr]_+ } .\ ] ] applying lemma [ lem1.14 ] we obtain the first inequality .
+ to prove the second inequality we observe that , using again * step 2 * , @xmath679^{-1},\end{aligned}\ ] ] where we have used the fact that @xmath680 as shown in equation .
thus we conclude that @xmath681 + * step 4 . * from * step 3 * we deduce that @xmath682_+ \bigl [ 1 -b_{\lambda , \beta } \bigl ( n(\theta )
\bigr ) \bigr]_+ } .\ ] ] to conclude the proof it is sufficient to apply * step 3 * to @xmath683 defined in equation . indeed , by equation , for any @xmath684 @xmath685 and , by equation
, we have @xmath686 we conclude this section by mentioning assumptions under which it is possible to give a non - random bound for @xmath687 defined in equation .
let us assume that , for some exponent @xmath688 and some positive constants @xmath689 and @xmath690 , @xmath691 \leq 1.\ ] ] in this case , with probability at least @xmath415 , @xmath692 where we recall that @xmath693 . $ ] + to give a point of comparison , in the centered gaussian case where @xmath694 is a gaussian vector , we have , for any @xmath695 @xmath696 \biggr ] = 1,\ ] ] where @xmath697 are the eigenvalues of @xmath698 therefore , with probability at least @xmath415 , @xmath699 moreover we observe that @xmath700 in order to replace equation with some polynomial assumptions we need to replace @xmath622 by @xmath701 indeed , by the bienaym chebyshev inequality , we get that , with probability at least @xmath415 , @xmath702 + \biggl ( \frac { \mathbb e \bigl [ \lvert x \rvert^{12 } \bigr]}{n \epsilon } \biggr)^{1/2 } \biggr)^{1/6 } \leq \bigl ( 1 + ( n \epsilon)^{-1/2 } \bigr)^{1/6 } \ , \mathbb e \bigl [ \lvert x \rvert^{12 } \bigr]^{1/12}\ ] ] and hence , with probability at least @xmath703 , @xmath704^{1/12}.\ ] ] also in the case where we consider this new quantity @xmath705 we obtain a bound of the form of proposition [ prop2.8eg ] .
more precisely , the following proposition holds true .
let @xmath706 and let us put @xmath707 with probability at least @xmath89 , for any @xmath166 , @xmath708_+}.\ ] ] we observe that another way to take advantage of equation is to write @xmath709 thus , putting @xmath710 we get that , for any @xmath166 , @xmath711 the same reasoning used to prove * step 2 * of proposition [ prop2.8eg ] shows that , with probability at least @xmath415 , for any @xmath166 , any @xmath72 , any @xmath655 , @xmath712 .\ ] ] as a consequence , with probability at least @xmath89 , for any @xmath166 , any @xmath72 , @xmath713_+}.\ ] ]
in this section we come back to the finite - dimensional framework and we consider the problem of estimating the expectation of a symmetric random matrix .
we will use these results to estimate the covariance matrix in the case of unknown expectation .
let @xmath714 be a symmetric random matrix of size @xmath715 as already observed for the gram matrix , the expectation of @xmath716 can be recovered via the polarization identity from the quadratic form @xmath717 , \qquad \theta \in \mathbb r^d,\ ] ] where the expectation is taken with respect to the unknown probability distribution of @xmath716 on the space of symmetric matrices of size @xmath715 observe that , if we decompose @xmath716 in its positive and negative parts @xmath718 the quadratic form @xmath719 rewrites as @xmath720 - \mathbb e \left [ \theta^{\top}a_-\theta\right ] = n_{a_+}(\theta ) - n_{a_-}(\theta ) . \end{aligned}\ ] ] thus in the following we will consider the case of a symmetric positive semi - definite random matrix of size @xmath715 + from now on let @xmath714 be a symmetric positive semi - definite random matrix of size @xmath121 and let @xmath29 be a probability distribution on the space of symmetric positive semi - definite random matrices of size @xmath715 our goal is to estimate @xmath721 , \quad \theta \in \mathbb r^d,\ ] ] from an i.i.d .
sample @xmath722 of symmetric positive semi - definite matrices drawn according to @xmath19 we observe that the quadratic form @xmath24 rewrites as @xmath723\ ] ] where @xmath724 denotes the square root of @xmath725 + the construction of the ( robust ) estimator @xmath149 follows the one already done in the case of the gram matrix with the necessary adjustments . for any @xmath45 and for any @xmath423
we consider the empirical criterion @xmath726 where the influence function @xmath54 is defined as in equation , and we perturb the parameter @xmath35 with the gaussian perturbation @xmath727 of mean @xmath35 and covariance matrix @xmath728 where @xmath66 is a free real parameter .
we consider the family of estimators @xmath729 where @xmath730 .
let us put @xmath731^{1/4 } \quad \text{and } \quad \kappa = \sup_{\substack{\theta\in \mathbb r^d\\ \mathbb e [ \lvert a^{1/2 } \theta \rvert^2 ] > 0 } } \frac{\displaystyle\mathbb e \bigl [ \lvert a^{1/2 } \theta \rvert^4 \bigr ] } { \displaystyle\mathbb e \bigl [ \lvert a^{1/2 } \theta \rvert^2 \bigr]^{2}}.\ ] ] the finite set @xmath67 of possible values of the couple of parameters @xmath68 is defined as @xmath502 where @xmath732 with @xmath114 , and @xmath733}.\end{aligned}\ ] ] we recall that @xmath734 is defined in equation as @xmath735 .
+ according to equation , let @xmath736 and put @xmath737 $ ] .
define the estimator @xmath624 as @xmath738 [ props1 ] let @xmath7390,s_4 ^ 2]$ ] be some energy level . with probability at least @xmath89 , for any @xmath77 , @xmath537,\ ] ] where @xmath120 is defined as @xmath112 \zeta _ * ( \max \ { t , \sigma \ } ) \leq \sqrt{n } \\
+ \infty & \text { otherwise } \end{cases}\ ] ] and @xmath740 } { \kappa^{1/2 } \mathbb e \bigl [ \lvert a \rvert_{\infty}^2\bigr]^{1/2 } t } + \log(k ) + \log(\epsilon^{-1 } ) \biggr ) } + \sqrt { \frac{98.5 \kappa^{1/2 } \mathbb e \bigl [ \lvert a \rvert_{\infty}^2 \bigr]^{1/2 } } { t}}.\ ] ] to prove proposition [ props1 ] we use many results already proved in the case of the gram matrix ( with the necessary adjustments ) . +
we first observe that , denoting by @xmath743 a gaussian random vector with mean @xmath744 and covariance matrix @xmath745 we have @xmath746 & = \sum_{i=1}^d \mathbb e \bigl ( \langle w , e_i\rangle^2 \bigr)\\ & = \| a^{1/2 } \theta\|^2+\frac{\mathbf{tr}(a)}{\beta}\end{aligned}\ ] ] where @xmath39 is the canonical basis of @xmath21 ( and @xmath747 is a one - dimensional gaussian random variable with mean @xmath748 and variance @xmath749 ) .
therefore the empirical criterion rewrites as @xmath750.\ ] ] we now use lemma [ l1 ] to pull the expectation outside the influence function @xmath54 .
we decompose @xmath716 into @xmath751 , where @xmath752 and @xmath753 and we observe that @xmath754 has the same distribution as @xmath755 , where @xmath756 so that @xmath757 = \mathbf{var } \biggl ( \sum_{i=1}^d \bigl ( ( u^{\top } \theta)_i + w_i \bigr)^2 \lambda_i \biggr )
\\ = \sum_{i=1}^d \lambda_i^2 \mathbf{var } \bigl [ \bigl ( ( u^{\top } \theta)_i + w_i \bigr)^2 \bigr ] = \sum_{i=1}^d \biggl ( \frac{2}{\beta^2 } + \frac{4}{\beta } ( u^{\top } \theta))_i^2 \biggr ) \lambda_i^2 \\ = \frac{2}{\beta^2 } \mathbf{tr}(a^2 ) + \frac{4}{\beta } \lvert a \theta \rvert^2.\end{gathered}\ ] ] as a consequence we get @xmath758\end{gathered}\ ] ] where the function @xmath359 is defined in equation .
we then apply equation with @xmath759 @xmath760 @xmath761 and @xmath762 to obtain @xmath763\end{gathered}\ ] ] and we conclude , by lemma [ lem1.5 ] , that @xmath764 \ , \mathrm{d } \pi_{\theta}(\theta'),\end{gathered}\ ] ] where @xmath765
we then apply the pac - bayesian inequality ( proposition [ pac ] ) to @xmath766\ ] ] where @xmath767 and we choose as posterior distributions the family of gaussian perturbations @xmath768 we obtain that , with probability at least @xmath769 , for any @xmath44 , @xmath770 \mathrm d \pi_{\theta } ( \theta ' ) \\
\shoveright{+ \frac{\beta \|\theta\|^2}{2n } + \frac { \log(\epsilon^{-1})}{n } } \\
= \mathbb{e } \biggl [ \lvert a^{1/2 } \theta \rvert^2 - \lambda + \frac{1}{2 } \bigl ( \lvert a^{1/2 } \theta \rvert^2 - \lambda \bigr)^2 + \frac{(c+2)\lvert a \theta \rvert^2}{\beta } + \frac{(2 + 3c ) \mathbf{tr}(a^2)}{2 \beta^2 } \biggr ] \\
+ \frac{\beta \lvert \theta \rvert^2}{2 n } + \frac{\log(\epsilon^{-1})}{n}. \end{gathered}\ ] ] using the cauchy - schwarz inequality remark that @xmath771 \leq \mathbb e \bigl [ \lvert a \rvert_{\infty } \lvert a^{1/2 } \theta \rvert^2 \big ] \leq \mathbb e \bigl [ \lvert a \rvert_{\infty}^2 \bigr]^{1/2 } \mathbb e
\bigl [ \lvert a^{1/2 } \theta \rvert^4 \bigr]^{1/2 } \leq \mathbb e \bigl [ \lvert a \rvert_{\infty}^2 \bigr]^{1/2 } \kappa^{1/2 } n(\theta)\ ] ] where @xmath71 is defined in equation .
thus @xmath429 ^ 2 + \biggl [ 1 + ( \kappa -1 ) \lambda + \frac{(2+c ) \kappa^{1/2 } \mathbb e \bigl [ \lvert a \rvert_{\infty}^2 \bigr]^{1/2 } } { \beta } \ , \biggr ] \bigl [ n(\theta ) -
\lambda \bigr ] \\
+ \frac{(\kappa-1 ) \lambda^2}{2 } + \frac{(2+c ) \kappa^{1/2 } \mathbb e \bigl [ \lvert a \rvert_{\infty}^2 \bigr]^{1/2 } \lambda}{\beta } + \frac{(2 + 3c ) \mathbb e \bigl [ \mathbf{tr}(a^2 ) \bigr]}{2 \beta^2 } + \frac{\beta \lvert \theta \rvert^2}{2n } + \frac{\log(\epsilon^{-1})}{n}. \end{gathered}\ ] ] this is the analogous of equation .
thus , proceeding as already done in the case of the gram matrix we conclude the proof .
remark that to obtain the above result we have used the fact that @xmath771 \leq \mathbb e \bigl [ \lvert a \rvert_{\infty}^2 \bigr]^{1/2 } \kappa^{1/2 } n(\theta).\ ] ] however , if we use any upper bound of the form @xmath771 \leq f(\mathbb e[a ] ) n(\theta)\ ] ] propositon [ props1 ] holds replacing @xmath772^{1/2 } \kappa^{1/2}$ ] with @xmath773)$ ] in the definition of @xmath558 .
similarly we can replace @xmath774 $ ] by an upper bound . + we observe in particular that @xmath775^{1/2}}{\kappa^{1/2 } } \leq \frac { \mathbb e \bigl [ \mathbf{tr}(a^2 ) \bigr]}{\kappa^{1/2 } \mathbb e \bigl [ \lvert a \rvert_{\infty}^2 \bigr]^{1/2 } } \leq \mathbb e \bigl [ \mathbf{tr}(a ) \bigr ] = \mathbf{tr } \bigl [ \mathbb{e } ( a ) \bigr].\ ] ] indeed , to see the first inequality it is sufficient to observe that @xmath776
. moreover we have that @xmath777 \leq \mathbb e \bigl [ \lvert a \rvert_{\infty } \mathbf{tr}(a ) \bigr ] \leq \mathbb e \bigl [ \lvert a \rvert_{\infty}^2 \bigr]^{1/2 } \mathbb e \bigl [ \mathbf{tr}(a)^2 \bigr]^{1/2},\ ] ] and , denoting by @xmath778 an orthonormal basis of @xmath779 @xmath780 = \sum _ { \substack { 1 \leq i \leq d , \\ 1 \leq j \leq d } } \mathbb e \bigl [ \lvert a^{1/2 } e_i \rvert^2 \lvert a^{1/2 } e_j \rvert^2 \bigr ] \\ \leq \sum _ { \substack { 1 \leq i \leq d , \\ 1 \leq j \leq d } } \mathbb e \bigl [ \lvert a^{1/2 } e_i \rvert^4 \bigr]^{1/2 } \mathbb e \bigl [ \lvert a^{1/2 } e_j \rvert^4 \bigr]^{1/2 } \\
\leq \kappa \sum _ { \substack { 1 \leq i \leq d , \\ 1 \leq j \leq d } } \mathbb e \bigl [ \lvert a^{1/2 } e_i \rvert^2 \bigr ] \mathbb e \bigl [ \lvert a^{1/2 } e_j \rvert^2 \bigr ] = \kappa \mathbb e \bigl [ \mathbf{tr}(a ) \bigr]^2 . \end{gathered}\ ] ] this implies that we can bound @xmath558 in proposition [ props1 ] by @xmath781 } { t } + \log(k ) + \log(\epsilon^{-1 } ) \biggr ) } + \sqrt { \frac{98.5\ \kappa \ \mathbb e [ \mathbf{tr}(a ) ] } { t}}.\ ] ] we conclude this section observing that , since the entropy terms are dominated by @xmath782 $ ] , the result can be generalized to the case where @xmath716 is a random symmetric positive semi - definite operator in a infinite - dimensional hilbert space with the only additional assumption that @xmath782 < + \infty.$ ] let @xmath32 be a random vector distributed according to the unknown probability measure @xmath783 the covariance matrix of @xmath20 is defined by @xmath784 ) ( x- \mathbb e[x])^{\top } \right]\ ] ] and our goal is to estimate , uniformly in @xmath785 the quadratic form @xmath786 \rangle^2 \right ] , \quad \theta \in \mathbb r^d,\ ] ] from an i.i.d .
sample @xmath787 drawn according to @xmath19 we can not use the results we have proved for the gram matrix , since the quadratic form @xmath137 depends on the unknown quantity @xmath26 $ ] .
however we can find a workaround , using the results of the previous section about symmetric random matrices .
indeed , we do not need to estimate @xmath26 $ ] in order to estimate @xmath137 but it is sufficient to observe that the quadratic form @xmath137 can be written as @xmath788\ ] ] where @xmath789 is an independent copy of @xmath790 more generally , given @xmath791 we may consider @xmath792 independent copies @xmath793 of @xmath20 and the random matrix @xmath794 so that @xmath795 = \mathbb e \bigl [ \ , \theta^{\top } \!\ ! a \
, \theta \ , \bigr].\ ] ] we will discuss later how to choose @xmath792 .
in the following we use a robust block estimate which consists in dividing the sample @xmath796 in blocks of size @xmath792 and then in considering the original sample as a `` new '' sample of @xmath797 symmetric matrices @xmath798 ( of independent copies of @xmath716 ) defined as @xmath799 that thus correspond to the empirical covariance estimates on each block .
we can use the results of the previous section to define a robust estimator of @xmath24 .
+ let us introduce @xmath800 } { \mathbb e \bigl [ \lvert a^{1/2 } \theta \rvert^2 \bigr]^2 } , \\ \text{and } \kappa & = \sup_{\substack { \theta \in \mathbb{r}^d\\\mathbb e \bigl [ \langle \theta , x - \mathbb e ( x ) \rangle^2 \bigr]>0 } } \frac{\mathbb e \bigl [ \langle \theta , x - \mathbb e ( x ) \rangle^4 \bigr]}{\mathbb e \bigl [ \langle \theta , x - \mathbb e ( x ) \rangle^2 \bigr]^2}. \end{aligned}\ ] ] replacing @xmath20 with @xmath803 $ ] we may assume during the proof that @xmath804 = 0 $ ] .
it holds that @xmath805 = \mathbb e \biggl [ \biggl ( \frac{1}{q(q-1 ) } \sum_{1 \leq j
< k \leq q } \langle \theta , x^{(j ) } - x^{(k ) } \rangle^2 \biggr)^2 \biggr ] \\ = \frac{1}{q^2(q-1)^2 } \sum_{\substack { 1 \leq j < k \leq q\\ 1 \leq s < t \leq q } } \mathbb e \bigl [ \langle \theta , x^{(j ) } - x^{(k ) } \rangle^2 \langle \theta , x^{(s ) } - x^{(t ) } \rangle^2 \bigr]\end{gathered}\ ] ] recalling the definition of covariance , we have @xmath805 = \frac{1}{q^2(q-1)^2 } \biggl\ { \sum_{\substack { 1 \leq j < k \leq q\\ 1 \leq s < t \leq q } } \mathbb e \bigl [ \langle \theta , x^{(j ) } - x^{(k ) } \rangle^2 \bigr ] \mathbb e \bigl [ \langle \theta , x^{(s ) } - x^{(t ) } \rangle^2 \bigr ] \\ + \sum_{1 \leq j
< k \leq q } \mathbb e \bigl [ \langle \theta , x^{(j ) } - x^{(k ) } \rangle^4 \bigr ] - \mathbb e \bigl [ \langle \theta , x^{(j ) } - x^{(k ) } \rangle^2 \bigr]^2 \\ + \sum_{\substack { 1 \leq j < k \leq q\\ 1 \leq s < t
\leq q\\ \lvert \ { j , k \ }
\cap \{s , t \ } \rvert = 1 } } \biggl ( \mathbb e \bigl [ \langle \theta , x^{(j ) } - x^{(k ) } \rangle^2 \langle \theta , x^{(s ) } - x^{(t ) } \rangle^2 \bigr ] \\ - \mathbb e \bigl [ \langle \theta , x^{(j ) } - x^{(k ) } \rangle^2 \bigr ] \mathbb e \bigl [ \langle \theta , x^{(s ) } - x^{(t ) } \rangle^2 \bigr ] \biggr ) \biggr\ } \\
= \frac{1}{4 } \mathbb e \bigl [ \langle \theta , x^{(2 ) } - x^{(1 ) } \rangle^2 \bigr]^2 + \frac{1}{2q(q-1 ) } \mathbb e \bigl [ \langle \theta , x^{(2 ) } - x^{(1 ) } \rangle^4 \bigr ] - \mathbb e \bigl [ \langle \theta , x^{(2 ) } - x^{(1 ) } \rangle^2 \bigr]^2 \\
+ \frac{q-2}{q(q-1 ) } \biggl ( \mathbb e \bigl [ \langle \theta , x^{(1 ) } - x^{(2 ) } \rangle^2 \langle \theta , x^{(1 ) } - x^{(3 ) } \rangle^2 \bigr ] - \mathbb e \bigl [ \langle \theta , x^{(1 ) } - x^{(2 ) } \rangle^2 \bigr]^2 \biggr).\end{gathered}\ ] ] define @xmath806 and observe that @xmath807 ^ 2 & = 4 n(\theta)^2 , \\ \mathbb e \bigl [ ( w_1 - w_2)^4 \bigr ] & = \mathbb e \bigl [ w_1 ^ 4 \bigr ] + 6 \mathbb e \bigl [ w_1 ^ 2 \bigr ] \mathbb e \bigl [ w_2 ^ 2 \bigr ] + \mathbb e \bigl [ w_2 ^ 4 \bigr ] \\ & = 2 \mathbb e \bigl [ w_1 ^ 4 \bigr ] + 6 \mathbb e \bigl [ w_1 ^ 2 \bigr]^2 \leq ( 2 \kappa + 6 )
n(\theta)^2 , \\
\mathbb e \bigl[(w_1 - w_2)^2 ( w_1 - w_3)^2 \bigr ] & = \mathbb e \bigl [ w_1 ^ 4 \bigr ] + 3 \mathbb e \bigl [ w_2 ^ 2]^2 \leq ( \kappa + 3 ) n(\theta)^2.\end{aligned}\ ] ] therefore @xmath808 \leq \biggl ( 1 + \frac{(q-2)(\kappa-1)}{q(q-1 ) } + \frac{(\kappa + 1)}{q(q-1 ) } \biggr ) n(\theta)^2 = \biggl ( 1 + \frac{\tau_q(\kappa)}{q } \biggr ) n(\theta)^2,\ ] ] and hence @xmath809 , since @xmath810 = n(\theta)$ ] .
the result follows from proposition [ props1 ] , using the definition of @xmath558 given in equation , where we replace @xmath71 by @xmath816 and @xmath188 by @xmath797 . according to lemma [ lem137 ]
we conclude the proof . here
we have used the upper bound for the entropy factor defined in terms of @xmath817= \mathbf{tr } ( \sigma)$ ] , as mentioned in the remarks following proposition [ props1 ] .
we can improve somehow the constants by evaluating more carefully @xmath818 $ ] and @xmath819 $ ] .
[ lem2 ] it holds that @xmath820 & \leq \biggl ( 1 - \frac{q-2}{q(q-1 ) } \biggr ) \lvert \sigma \rvert_{\infty } n(\theta ) + \frac{1}{q } \biggl ( \kappa + \frac{1}{q-1 } \biggr ) \mathbf{tr}(\sigma ) n(\theta ) \\
\mathbb e \bigl [ \mathbf{tr } \bigl ( a^2 \bigr ) \bigr ] & \leq \biggl ( 1 - \frac{q-2}{q(q-1 ) } \biggr ) \mathbf{tr } \bigl ( \sigma^2 \bigr ) + \frac{1}{q } \biggl ( \kappa + \frac{1}{q-1 } \biggr ) \mathbf{tr}(\sigma)^2.\end{aligned}\ ] ] replacing @xmath20 with @xmath803 $ ] we may assume that @xmath804 = 0 $ ] . recall that @xmath821 \leq \kappa \mathbb e [ \| x \|^2 ] ^2 = \kappa \mathbf{tr}(\sigma)^2\ ] ] and @xmath822 = \mathbf{tr}(\sigma^2).$ ] we observe that @xmath771 = \mathbb e \biggl [ \frac{1}{q^2(q-1)^2 } \sum_{\substack{1 \leq j < k \leq q \\ 1 \leq s < t \leq q } } \langle \theta , x^{(j ) } - x^{(k ) } \rangle \langle x^{(j ) } - x^{(k ) } , x^{(s ) } - x^{(t ) } \rangle \langle x^{(s ) } - x^{(t ) } , \theta \rangle \biggr]\ ] ] and @xmath823 \\ & = 4 \mathbb e \bigl [ \langle \theta , x^{(1 ) } \rangle \langle x^{(1 ) } , x^{(2 ) } \rangle \langle x^{(2 ) } , \theta \rangle \bigr ] = 4 \lvert \sigma \theta \rvert^2 \leq 4 \lvert \sigma \rvert_{\infty } n(\theta ) , \\
\mathbb e \bigl [ \langle \theta , x^{(1 ) } - x^{(2 ) } \rangle & \langle x^{(1 ) } - x^{(2 ) } , x^{(1 ) } - x^{(3 ) } \rangle \langle x^{(1 ) } - x^{(3 ) } , \theta \rangle \bigr ] \\ & = \mathbb e \bigl [ \langle \theta , x^{(1 ) } \rangle^2 \lvert x^{(1 ) } \rvert^2 \bigr ] + 3 \mathbb e \bigl [ \langle \theta , x^{(1 ) } \rangle \langle x^{(1 ) } , x^{(2 ) } \rangle \langle x^{(2 ) } , \theta \rangle \bigr ] \\ & \leq \kappa \mathbf{tr}(\sigma ) n(\theta ) + 3 \lvert \sigma \rvert_{\infty } n(\theta ) ,
\\ \mathbb e \bigl [ \langle \theta , x^{(1 ) } - x^{(2 ) } \rangle & \langle x^{(1 ) } - x^{(2 ) } , x^{(1 ) } - x^{(2 ) } \rangle \langle x^{(1 ) } - x^{(2 ) } , \theta \rangle \bigr ] \\ & = 2 \mathbb e \bigl [ \langle \theta , x^{(1 ) } \rangle^2 \lvert x^{(1 ) } \rvert^2 \bigr ] + 2 \mathbb e \bigl [ \langle \theta , x^{(1 ) } \rangle^2 \bigr ] \mathbb e \bigl [ \lvert x^{(1 ) } \rvert^2 \bigr ] \\ & \quad + 4 \mathbb e \bigl [ \langle \theta , x^{(1 ) } \rangle \langle x^{(1 ) } , x^{(2 ) } \rangle \langle x^{(2 ) } , \theta \rangle \bigr ] \\ & \leq 2(\kappa+1 ) \mathbf{tr}(\sigma ) n(\theta ) + 4 \lvert \sigma \rvert_{\infty } n(\theta),\end{aligned}\ ] ] which proves the first inequality . in the same way ,
@xmath824 = \mathbb e \biggl [ \frac{1}{q^2(q-1)^2 } \sum_{\substack{1 \leq j < k \leq q \\ 1 \leq s < t
\leq q } } \langle x^{(j ) } - x^{(k ) } , x^{(s ) } - x^{(t ) } \rangle^2 \biggr],\ ] ] and @xmath825 & = 4 \mathbb e \bigl [ \langle x^{(1 ) } , x^{(2 ) } \rangle^2 \bigr ] \\ & = 4 \mathbf{tr } \bigl ( \sigma^2 \bigr ) , \\ \mathbb e \bigl [ \langle x^{(1 ) } - x^{(2 ) } , x^{(1 ) } - x^{(3 ) } \rangle^2 \bigr ] & = \mathbb e \bigl [ \lvert x^{(1 ) } \rvert^4 \bigr ] + 3 \mathbb e \bigl [ \langle x^{(1 ) } , x^{(2 ) } \rangle^2 \bigr ] \\ & \leq \kappa \mathbf{tr}(\sigma)^2 + 3 \mathbf{tr } \bigl ( \sigma^2 \bigr ) , \\ \mathbb e \bigl [ \langle x^{(1 ) } - x^{(2 ) } , x^{(1 ) } - x^{(2 ) } \rangle^2 \bigr ] & = 2 \mathbb e \bigl [ \lvert x^{(1 ) } \rvert^4 \bigr ] + 2 \mathbb e \bigl [ \lvert x^{(1 ) } \rvert^2 \bigr]^2 + 4 \mathbb e \bigl [ \langle x^{(1 ) } , x^{(2 ) } \rangle^2 \bigr ] \\ & \leq 2 ( \kappa + 1 ) \mathbf{tr}(\sigma)^2 + 4 \mathbf{tr } \bigl ( \sigma^2 \bigr).,\end{aligned}\ ] ] which concludes the proof . using these tighter bounds , we can improve @xmath826 to @xmath827}{\bigl [ \bigl ( 1 - \frac{q-2}{q(q-1 ) } \bigr ) \lvert \sigma \rvert_{\infty } + \frac{1}{q } \bigl ( \kappa + \frac{1}{q-1 } \bigr ) \mathbf{tr}(\sigma ) \bigr ] t } \\ \shoveright { + \log(k ) + \log(\epsilon^{-1 } ) \biggr ) \biggr]^{1/2 } } \\ + \sqrt { \frac{98.5 \bigl [ q \bigl ( 1 - \frac{q-2}{q(q-1 ) } \bigr ) \lvert \sigma \rvert_{\infty } + \bigl ( \kappa + \frac{1}{q-1 } \bigr ) \mathbf{tr}(\sigma ) \bigr ] } { t}}.\end{gathered}\ ] ] therefore , in the case when @xmath828 we have @xmath829 \leq \frac{1}{q}\biggl ( \kappa+1 + \frac{2}{q(q-1 ) }
\biggr ) \mathbf{tr}(\sigma ) n(\theta)\ ] ] and hence , recalling that @xmath830 , we can take @xmath831 if we compare the above result with the bound obtained in proposition [ prop2 ] for the gram matrix estimator , we see that the first appearance of @xmath71 in the definition of @xmath826 has been replaced with @xmath832 and that the second appearance of @xmath71 has been replaced with @xmath833 thus , when @xmath834 , that is not a very strong hypothesis , we can take at least @xmath835 , and obtain an improved bound for the estimation of @xmath836 that is not much larger than the bound for the estimation of the centered gram matrix , that requires the knowledge of @xmath837 , since the difference between the two bounds is just a matter of replacing @xmath71 with @xmath838 . shawe - taylor , j. , williams , c. and cristianini , c. and kandola , j. ( 2005 ) _ on the eigenspectrum of the gram matrix and its relationship to the operator eigenspectrum_. eds . ) : alt 2002 , lnai 2533 , pages 2340 , springer - verlag zwald , l. , bousquet , o. and blanchard , g. ( 2004 ) _ statistical properties of kernel principal component analysis_. in learning theory , volume 3120 of lecture notes in comput .
, pages 594608 .
springer , berlin | in this paper we investigate the question of estimating the gram operator by a robust estimator from an i.i.d . sample in a separable hilbert space and we present uniform bounds that hold under weak moment assumptions .
the approach consists in first obtaining non - asymptotic dimension - free bounds in finite - dimensional spaces using some pac - bayesian inequalities related to gaussian perturbations of the parameter and then in generalizing the results in a separable hilbert space .
we show both from a theoretical point of view and with the help of some simulations that such a robust estimator improves the behavior of the classical empirical one in the case of heavy tail data distributions . |
stability properties of core collapses in massive progenitor stars before supernova ( sn ) explosions ( goldreich & weber 1980 gw hereafter ; goldreich et al . 1996
; lai 2000 ; lai & goldreich 2000 ; blondin et al .
2003 ; murphy et al . 2004 ; burrows et al . 2006 , 2007
; lou & cao 2008 ; cao & lou 2009 ) have come to focus after three decades , because of the realization after numerous unsuccessful one - dimensional sn simulations that the breakdown of spherical symmetry inevitably occurs and plays a key role in sn explosions ( e.g. burrows 2000 ) .
numerical simulations ( e.g. bruenn 1989a , b ) indicate that pre - sn stellar core collapses may be approximately described by a homologous process , first analyzed by gw for a conventional polytropic equation of state ( eos ) @xmath11 where @xmath12 and @xmath13 are the pressure and density and both @xmath14 and @xmath0 are constant . for perturbations obeying identical eos of the background core collapse , gw explored linear stability properties of such stellar collapses and concluded that these collapses are stable .
yahil ( 1983 ) extended homologous collapses to polytropic exponent @xmath15 , noting the presence of an outer supersonic envelope besides the inner core collapse .
lai ( 2000 ) performed perturbation analysis to these extended solutions and claimed that perturbations are stable for @xmath16 .
lou & cao ( 2008 ) substantially extended the @xmath0 self - similar solutions , including those of gw , using a general polytropic eos with a temporally and radially variable @xmath14 being conserved along streamlines .
we obtained a broad family of homologous core collapses allowing @xmath14 as an arbitrary function of the independent self - similar variable .
the specific entropy of an ideal or perfect gas is @xmath17 where @xmath18 is the boltzmann constant and @xmath1 is the adiabatic exponent with @xmath2 . here
@xmath19 determines the structure and evolution of core collapses while @xmath1 controls specific entropy perturbation properties .
the earlier isentropic assumption requires @xmath20 ( e.g. gw ) .
numerical simulations ( e.g. bethe et al .
1979 ; bruenn 1985 , 1989a , 1989b ; woosley et al . 1993 , 2002 ) for the structure of massive stars and sn explosions support variable radial distributions of specific entropy .
thus the model of lou & cao ( 2008 ) allows an arbitrary radial profile of specific entropy . using this general polytropic model
, we conducted a non - radial adiabatic perturbation analysis and classified different perturbation modes parallel to stellar oscillations ( see unno et al .
1979 and cao & lou 2009 ) and concluded that in addition to internal gravity g@xmath8modes for convective instabilities , sufficiently high - order g@xmath10modes also become unstable .
cao & lou ( 2009 ) demonstrated that the criterion of using the sign of the brunt - v@xmath4is@xmath4l@xmath4 buoyancy frequency squared @xmath21 ( see definition [ defbv ] ) for the existence of g@xmath6modes remains valid in self - similar dynamic core collapses . in a conventional polytropic process
, isentropic perturbations with @xmath20 make all g@xmath6modes disappear , i.e. @xmath22 . thus previous perturbation analyses ( gw ; lai 2000 ; lai & goldreich 2000 ) considered only the stability of acoustic modes .
in addition to p@xmath6modes and f@xmath6modes , cao & lou ( 2009 ) found g@xmath6mode instabilities in stellar core collapses and speculated possible consequences for neutron star kicks etc . from the existence criterion of @xmath23 for g@xmath6modes , we realize that in a conventional polytropic core collapse , non - radial adiabatic perturbations with @xmath2 should allow g@xmath6modes .
such nonisentropic assumption has been invoked decades ago ( e.g. ledoux 1965 ; unno et al .
1979 ) and applied to helioseismology and white dwarf oscillations ( e.g. shibahashi et al . 1988 ) .
for star formation , mckee & holliman ( 1999 ) also studied nonisentropic radial perturbations in molecular clouds .
they used a thermal free energy to discuss cloud stability and gave critical cloud masses for locally and globally adiabatic models .
these early results motivate us to perform such a perturbation analysis for stellar core collapses and study the existence and stability of g@xmath6modes in addition to other modes .
the main thrust of this letter is to show such g@xmath6mode instabilities and speculate consequences for stellar core collapses and sn explosions .
the governing equations include conservations of mass and momentum , poisson equation for self - gravity and eos .
the gas is ideal and the background is a @xmath0 conventional polytrope while the adiabatic perturbation with @xmath2 conserves the specific entropy along streamlines . the background collapse is spherically symmetric and as in gw , we introduce a dimensionless independent variable @xmath24 where @xmath25^{1/2}$ ] is the jeans length with a central mass density @xmath26 and @xmath27 is the gravitational constant .
physical variables are cast into the following forms where the first term of each variable is the dynamic background and the last term is the first - order perturbation : @xmath28\ , \label{p2}\\ p=\frac{\kappa^3}{(\pi g)^2}a(t)^{-4}f^4(x)[1 + \beta_1(x,\ \theta,\ \varphi)\tau(t)]\ , \label{p3}\\ \phi=\frac{4}{3}\left(\frac{\kappa^3}{\pi g}\right)^{1/2}a(t)^{-1}[\psi(x)+\psi_1(x,\ \theta,\ \varphi)\tau(t)]\ .\label{p4}\end{aligned}\ ] ] here the time - dependent temporal factor @xmath29 is @xmath30&\quad % ( \lambda\neq0)\ .\ ] ] where the free - fall timescale @xmath31 is defined by @xmath32^{-1/2}\ % \qquad{\rm % ( for\ all\ \lambda)}\ , \nonumber\\ % & = & ( 9\lambda/2)^{1/2}t\qquad{\rm ( for\ \lambda\neq0)}\ .\ ] ] and @xmath33 is a dimensionless ` frequency ' . we do not use stream function for velocity perturbations to allow for vorticities .
the dynamic collapse background is that of gw . defining a dimensionless collapse parameter @xmath34 ( yahil 1983 ) as @xmath35 where the subscript @xmath36 indicates background variables and @xmath37 is the free - fall velocity , we obtain the background density profile determined by the equation for @xmath38 , namely @xmath39+f^3=\lambda\ , \ ] ] with boundary conditions @xmath40 and @xmath41 .
other background variables are then readily determined
. solutions of eq ( [ core ] ) are sensible for core collapses of massive stars when density vanishes at a moving boundary @xmath42 . for this purpose
, the variable @xmath43 should be negative and the time reversal operation must be taken in the governing equations .
we assume adiabatic perturbations with the exponent @xmath1 different from @xmath0 . in our model , we define @xmath44 and @xmath45 $ ] for convenient analysis . after the linearization of hydrodynamic equations including conservations of mass and momentum , poisson equation for self - gravity and the adiabatic eos , the angular variation factors of @xmath46 , @xmath47 , @xmath48 and @xmath49 can be separated out involving the spherical harmonics @xmath50 .
we further let @xmath46 take the consistent form of @xmath51 where the modified transverse gradient operator is @xmath52 as @xmath53 decouples from other perturbation equations , we suppress it in the following analysis . eliminating @xmath47 and @xmath48
, we obtain a set of ordinary differential equations ( odes ) : @xmath54 where the same notation is used for the @xmath55dependent factor of @xmath49 .
odes @xmath56 for non - radial perturbations reduce to those of gw by setting @xmath57 and introducing a stream function for velocity perturbations .
regular boundary conditions are imposed , namely @xmath58 the last of the outer boundary conditions requires a zero lagrangian pressure . with these boundary conditions , we can prove the orthogonality of eigenfunctions , i.e. @xmath59 where superscripts @xmath60 and @xmath61 indicate eigenfunctions of different eigenvalues .
meanwhile , the variational principle ( chandrasekhar 1964 ) can be formulated for this eigenvalue problem . to solve this eigenvalue problem , we use the following numerical procedure . after solving the background physical variables by using the fourth - order explicit runge - kutta scheme , we discretize odes ( [ eq1])@xmath6([eq3 ] ) with a proper mesh . then the inverse iteration method ( e.g. wilkinson 1965 )
is applied to determine eigenvalues and eigenfunctions . to enhance the efficiency of the inverse iteration method , a relaxation scheme is also implemented .
we first test our numerical code to solve @xmath62 isentropic cases and confirm the results of cao & lou ( 2009 ) to guarantee the correctness and reliability of the code .
this check also confirms the errors of p@xmath6mode eigenvalues in gw .
the spectra of eigenvalues versus the mode degree @xmath63 contain characteristic features of oscillations . a typical spectrum of @xmath64 modes is shown in fig .
[ figure1 ] . in these cases ,
all eigenvalues @xmath65 are negative .
we use two dashed lines to separate three distinct classes of modes : p@xmath6modes for @xmath66 , a unique branch of f@xmath6modes for @xmath67 and g@xmath10modes for @xmath68 ( see cowling 1941 for mode classifications of stellar oscillations and unno et al .
1979 for properties of different modes . ) .
the p@xmath6modes are pressure - driven and gravity - modified acoustic oscillations .
the f - modes are essentially trapped surface lamb waves ( lamb 1932 ; unno et al . 1979 ; lou 1990 , 1991 ) and decay exponentially inwards .
the g@xmath6modes are driven by buoyancy and are trapped deep inside the core .
the g@xmath10modes are one subclass of g@xmath6modes with @xmath64 . for a given @xmath63 , the absolute value of eigenvalue @xmath69 decreases as the radial order increases . in that limit
, eigenvalue @xmath65 approaches @xmath36 .
there is no @xmath70 g@xmath10mode .
the other subclass g@xmath8modes exist for @xmath71 with similar characteristics of g@xmath10modes , except that the eigenvalues @xmath65 are positive and decrease towards zero as the radial order increases . by numerical explorations ,
we find similar behaviours of each perturbation mode in the core collapsing background as those in stellar oscillations .
as p@xmath6modes have been studied by gw for @xmath57 , we emphatically show g@xmath10mode eigenfunctions and their eigenvalues in fig .
[ figure3 ] . versus spherical harmonic degree @xmath63 are shown with the parameter @xmath72 and the adiabatic exponent of perturbations @xmath73 .
all eigenvalues lie in the @xmath74 regime .
two dotted lines divide the region into three distinct zones for eigenvalues of different modes .
there is a sign difference between the definitions of the eigenvalue @xmath65 here and in stellar oscillations ( see figures of cox 1976 for a comparison).,scaledwidth=50.0% ] modes with @xmath75 nodes from bottom to top panels with eigenvalues @xmath76 @xmath77 , respectively , are presented here .
the solid and dashed curves represent @xmath78 and @xmath79 , respectively .
factors @xmath80 and @xmath81 are multiplied for compact presentations.,scaledwidth=50.0% ] between p@xmath6modes and g@xmath10modes exists the unique branch of f@xmath6modes for @xmath67 .
eigenvalues of f@xmath6modes separate the p@xmath6 and g@xmath6modes .
such f@xmath6modes are characterized by eigenfunctions of density and radial velocity perturbations without nodes .
they are acoustic in nature and relate to the surface lamb waves ( lamb 1932 ; lou 1990 , 1991 ) . in stellar oscillations ,
the existence criterion for both types of g@xmath6modes depends on the square of the brunt - v@xmath3is@xmath4l@xmath4 buoyancy frequency @xmath21 defined by @xmath82 where @xmath83 and @xmath84 are for a hydrostatic equilibrium , and @xmath1 is the adiabatic exponent of perturbation . since our numerical explorations reveal that if an eigenvalue @xmath65 with its eigenfunction can be obtained for @xmath85 for the hydrostatic limit , its counterpart can also be determined for @xmath86 in a continuous manner .
thus , the existence criterion for g@xmath6modes in stellar oscillations can be applied in our analysis by using partial derivatives instead of derivatives with respect to the radius . for @xmath71 ( @xmath87 ) , g@xmath8modes appear , whereas for @xmath64 ( @xmath88 ) , g@xmath10modes manifest . for @xmath57 ( @xmath22 ) , both g@xmath6modes are suppressed .
we have readily computed the p@xmath6mode eigenfunctions with @xmath89 and they appear qualitatively similar to those shown in gw and in cao & lou ( 2009 ) . we demonstrate mathematically that a g@xmath10mode does not change to a g@xmath8mode or _
vice versa _ as @xmath34 increases from @xmath36 to @xmath90 , i.e. for @xmath91 , there is no @xmath92 eigenvalue for all @xmath93 . by setting @xmath92 in odes ( [ eq1])@xmath6([eq3 ] ) and after straightforward manipulations ,
we derive @xmath94d^3r=0\ , \end{aligned}\ ] ] which requires @xmath95 . thus no non - trivial eigenfunctions can be found .
consequently , the eigenvalue curve @xmath96 for each perturbation mode does not intersect the @xmath92 line .
we show the variation trends of eigenvalues @xmath65 as @xmath34 increases from @xmath36 to @xmath90 in fig .
[ figure4 ] for typical eigenvalues @xmath65 of the three lowest p@xmath6modes and g@xmath6modes for @xmath97 and of f@xmath6mode for @xmath98 as functions of @xmath34 .
[ figure4 ] shows that eigenvalues @xmath65 larger than @xmath99 approach this limiting value as @xmath100 ( see gw ) .
[ figure5 ] illustrates the variation of eigenvalues @xmath65 for the first two lowest orders of p@xmath6modes and g@xmath10modes as well as that of f@xmath6modes all for @xmath72 and @xmath98 as the adiabatic exponent @xmath1 varies from @xmath101 to @xmath102 .
variations of the three lowest acoustic p@xmath6modes and internal gravity g@xmath10modes of @xmath97 and of surface f@xmath6modes for @xmath98 with the collapse parameter @xmath34 characterizing different self - similar dynamic backgrounds .
the adiabatic exponent of perturbations is @xmath103.,scaledwidth=50.0% ] variations of the two lowest acoustic p@xmath6modes and internal gravity g@xmath10modes and of surface f@xmath6modes for @xmath98 with the adiabatic exponent @xmath1 varying from @xmath101 to @xmath102 .
the collapse parameter is @xmath72 for the conventional polytropic stellar core collapse of @xmath0.,scaledwidth=50.0% ] a key question for perturbation analysis is the stability of a core collapse .
the inequality @xmath87 for stellar perturbations is the schwarzschild criterion for convection . in our dynamic core collapse background ,
we emphasize that this inequality is only sufficient but not necessary for instability .
the temporal factor @xmath29 for @xmath104 bears a power law @xmath105 with @xmath106^{1/2}$ ] . for a time reversal operation
, gw showed that the compression of a collapse is responsible for an amplification of @xmath107 .
thus the determinant under the square root in @xmath108 decides the mode stability . for @xmath109 so that @xmath108 has two real roots , the mode is unstable for one of the two roots . for @xmath110 so that @xmath108 has a pair of complex conjugate roots , a perturbation oscillates stably in addition to the compression amplification of core collapse . by this new criterion for instability , we find that all g@xmath8modes and sufficiently high - order g@xmath10modes are unstable
. it can be shown by a local analysis that the definition of @xmath5 remains valid for dynamical core collapses , such that g@xmath8modes lead to convective instabilities .
it also announces that g@xmath10mode instabilities are uniquely associated with self - similar dynamic core collapse .
we compare our results with earlier analyses ( gw ; lai 2000 ; cao & lou 2009 ) .
first , in contrast to oscillations in a static star , we examine non - radial adiabatic perturbations in a self - similar conventional polytropic core collapse .
the adiabatic exponent @xmath1 differs from the background @xmath0 , whereas earlier stability analyses of the dynamic background are restricted to @xmath57 .
secondly , for a conventional polytropic core collapse , we classify different adiabatic perturbation modes , including p@xmath6modes , g@xmath6modes and f@xmath6modes .
gw and lai ( 2000 ) studied only p@xmath6mode perturbations as their isentropic eos makes g@xmath6modes vanish .
cao & lou ( 2009 ) obtained g@xmath6modes by a general polytropic eos with a variable specific entropy distribution , even though the perturbation eos remains the same as the background eos . in comparison
, our model here describes non - radial adiabatic perturbations with @xmath89 .
this leads to @xmath111 so that g@xmath6modes may exist .
we indeed confirm this by numerical explorations .
for various efficiencies of heat transport and radiative losses , the case of @xmath20 should be a special and rare situation .
therefore g@xmath6modes should exist in stellar core collapses in general . for core - collapse sn explosions ,
such g@xmath6mode instabilities should bear physical consequences .
these instabilities complement those revealed by cao & lou ( 2009 ) .
they occur during the pre - sn core collapse stage and should also influence the formation of proto - neutron stars and subsequent emergence and evolution of rebound shocks .
at least , such instabilities lead to early breakdown of spherical symmetry before the emergence of rebound shocks ( e.g. lou & wang 2006 , 2007 ; wang & lou 2008 ; hu & lou 2009 ) . so far , most proposed instabilities occur either in a massive progenitor star prior to the core collapse ( e.g. goldreich et al . 1996 ) or after the core rebound ( e.g. blondin et al .
2003 ; burrows et al .
2006 , 2007 ) .
such instabilities may affect the formation , motion and evolution of a proto - neutron star or a pulsar as speculated by cao & lou ( 2009 ) .
moreover , our perturbation analysis serves to link perturbations in a massive progenitor star and those after the core rebound .
for example , the @xmath112-mechanism ( e.g. goldreich et al . 1996 ) may provide seed perturbations during a core collapse while such perturbations during a core collapse may stimulate those after the core rebound .
we emphasize that while p@xmath6modes , f@xmath6modes and low - order g@xmath10modes are stable , if radiative losses and diffusive processes are involved instead of the adiabatic approximation , these modes might become overstable .
in this letter , we examine 3d adiabatic perturbations in a self - similar conventional polytropic collapsing core with @xmath89 .
the gas is non - isentropic in the sense that the adiabatic index @xmath1 of perturbations differs from that of the background @xmath0 .
as @xmath1 determines the specific entropy perturbation conserved along streamlines , the non - isentropic process actually involves a nonzero buoyancy frequency @xmath5 giving rise to g@xmath6mode perturbations . in comparison , perturbation analysis of cao & lou ( 2009 ) emphasizes a general polytropic core collapse with a variable specific entropy ( lou & cao 2008 ) and 3d adiabatic perturbations of @xmath57 . by imposing proper boundary conditions
, we solve the perturbation eigenvalue problem and derive distinct modes : acoustic p@xmath6modes , surface f@xmath6modes , internal gravity g@xmath8 and g@xmath10modes which are classified by their series of eigenvalues @xmath65 and eigenfunctions in reference to stellar oscillations . in parallel ,
g@xmath6modes involve two types , viz .
g@xmath10 and g@xmath8modes ; their existence depends on the square of the brunt - v@xmath3is@xmath4l@xmath4 buoyancy frequency @xmath113 : the former requires @xmath88 while the latter needs @xmath114 . for adiabatic perturbations ,
g@xmath10modes correspond to @xmath64 while g@xmath8modes occur for @xmath71 . stability properties of these modes in stellar core collapses are examined .
the instability criterion shifts from @xmath115 in a static lane - emden polytropic sphere of @xmath85 to @xmath109 for @xmath93 core collapses .
consequently , p@xmath6modes , f@xmath6modes and sufficiently low - order g@xmath10modes oscillate stably .
the g@xmath8modes are unstable leading to convections .
sufficiently high - order g@xmath10modes which would have been stable in a static polytrope now become unstable for @xmath93 gw dynamic core collapse . the specific radial order that g@xmath10modes
become unstable depends on the parameter pair of @xmath34 and @xmath1 .
we speculate that such inevitable g@xmath6mode instabilities may offer valuable clues to sn simulations and that the formation , motion and evolution of proto - neutron stars can be nontrivially influenced ( cao & lou 2009 ) .
these perturbations again lead to instabilities before the core rebound and they serve as seeds of later fluctuations .
their excitations can come from oscillations of the progenitor star before the onset of core collapse . according to our numerical exploration and typical stellar parameters with @xmath116
, unstable g@xmath10modes appear with fairly high radial orders .
this implies that the central mass blob can be quite small and a sn might even break the core into pieces without forming a ns .
this research was supported in part by tsinghua centre for astrophysics , by the national natural science foundation of china grants 10373009 , 10533020 and j0630317 at tsinghua university , and by the srfdp 20050003088 and 200800030071 , the yangtze endowment and the national undergraduate innovation training project from the ministry of education at tsinghua university .
99 bethe h. e. , et al . , 1979 ,
phys . , a324 , 487 bruenn s. w. , 1985 , apj , 58 , 771 bruenn s. w. , 1989a , apj , 340 , 955 bruenn s. w. , 1989b , apj , 341 , 385 burrows a. , 2000 , nat , 403 , 727 burrows a. , livne e. , dessart l. , ott c. d. , murphy j. , 2006 , apj , 640 , 878 burrows a. , dessart l. , ott c. d. , livne e. , 2007 , phys . rep . , 442 , 23 cao y. , lou y .- q . , 2009 , mnras in press arxiv:0908.3225 chandrasekhar s. , 1964 , apj , 139 , 664 cowling t. g. , 1941 , mnras , 101 , 367 goldreich p. , lai d. , sahrling m. , 1996 , in unsovled problems in astrophysics , ed
. j. n. bahcall & j. p. ostriker , princeton university press , princeton goldreich p. , weber s. v. , 1980 , apj , 238 , 991 hu r. y. , lou y .- q .
, 2009 , mnras , 396 , 878 lai d. , 2000 , apj , 540 , 946 lai d. , goldreich p. , 2000
, apj , 535 , 402 lamb h. , 1932 , hydrodynamics , 6th ed .
cambridge university press , london ledoux p. , 1965
, in aller l. h. , mclaughlin d. b. , eds , stellar structure , univ .
chicago press , chicago , p. 499
lou y .- q . , 1990 , apj , 361 , 527 lou y .- q . , 1991 , apj , 367 , 367 lou y .- q . , cao y. , 2008 , mnras , 384 , 611 lou y .- q . ,
wang w .- g .
, 2006 , mnras , 372 , 885 lou y .- q . ,
wang w .- g . , 2007 , mnras , 378 , l54 mckee c. f. , holliman j. h. ii , 1999 , apj , 522 , 313 murphy j. w. , burrows a. , heger a. , 2004 , apj , 615 , 460 shibahashi h. , sekii t. , kawaler s. , 1988 , in atmospheric diagnostics of stellar evolution : chemical peculiarity , mass loss , and explosion , springer , berlin , p. 86 | we perform a non - radial adiabatic perturbation analysis on homologous conventional polytropic stellar core collapses .
the core collapse features a polytropic exponent @xmath0 relativistic gas under self - gravity of spherical symmetry while three - dimensional perturbations involve an adiabatic exponent @xmath1 with @xmath2 such that the brunt - v@xmath3is@xmath4l@xmath4 buoyancy frequency @xmath5 does not vanish . with proper boundary conditions , we derive eigenvalues and eigenfunctions for different modes of oscillations . in reference to stellar oscillations and earlier results ,
we examine behaviours of different modes and the criterion for instabilities .
the acoustic p@xmath6modes and surface f@xmath6modes remain stable . for @xmath7 ,
convective instabilities appear as unstable internal gravity g@xmath8modes . for @xmath9 , sufficiently low - order internal gravity g@xmath10modes are stable , whereas sufficiently high - order g@xmath10modes , which would have been stable in a static star , become unstable during self - similar core collapses . for supernova explosions , physical consequences of such inevitable g@xmath6mode instabilities
are speculated .
hydrodynamics instabilities stars : neutron stars : oscillations ( including pulsations ) supernovae : general waves |
the use of phase fresnel lenses ( pfl s ) offer a mechanism to achieve superb imaging of astrophysical objects in the hard x - ray and gamma - ray energy regimes @xcite . in principle
, pfl s can concentrate nearly the entire incident photon flux , modulo absorption effects , with diffraction - limited imaging performance .
the impact of absorption is energy and material dependent , but can be almost negligible at higher photon energies .
the performance of these diffraction optics is obtained via tailoring the fresnel profile of each zone to yield the appropriate phase at the primary focal point .
however , pfl s have long focal lengths and are chromatic ; the excellent imaging is available over a narrow energy range . in order to demonstrate the imaging capabilities of these optics ,
we have fabricated ground - testable pfl s in silicon .
pfl s are a natural extension of fresnel - diffractive optics . as opposed to fresnel zone plates ( fzp ) , where alternating half - period or fresnel zones are completely blocked , and phase - reversal zone plates , where the blocked zone in a fzp is replaced by an amount of material to retard the phase by @xmath4 , the entire profile of each fresnel zone in a pfl has the appropriate amount of material to achieve the proper phase shift at the primary focus . in practice , the exact profile of a pfl can be approximated by a multiple - step structure as shown in figure 1 which illustrates the first four fresnel zones of a pfl with 4 steps approximating the ideal pfl profile .
the location of the radial step transitions is given by @xmath5 where @xmath6 is the location of the first fresnel ridge ( @xmath7 is the focal length and @xmath8 the photon wavelength ) and @xmath9 is the number of steps / fresnel ridge with a step index of @xmath10 .
this choice leads to annuli , as defined by each step , of constant area on the pfl .
each contributes equally , ignoring absorption , to the intensity at a given focal point .
this configuration of the stepped - phase profile is denoted as a regular - stepped pfl . for an exact pfl profile
, all the irradiance appears in the first order ( @xmath11 ) focus . in a stepped - profile approximation
, some energy appears in negative and higher order foci . ignoring absorptive effects
, the impact on the lens efficiency of approximating the exact pfl profile by a step - wise function is given by @xcite @xmath12}^2 & ~~{for~ ( n-1)/p = m \in integer } \\ & = 0 & ~~{otherwise } \nonumber\end{aligned}\ ] ] where @xmath13 is the relative intensity into the @xmath14 order focus for @xmath9 steps in each fresnel zone .
as @xmath9 increases , the indices with non - zero intensities of both the real and virtual higher order foci are pushed to higher @xmath15 with the relative intensities into these higher orders decreasing . in the limiting case where @xmath16 ,
the profile is exact for the pfl with the relative intensity in the 1st order ( @xmath11 ) going to 100% ( and the indices of the higher order foci being sent to @xmath17 ) .
more practically , a stepped - pfl with @xmath18 per fresnel zone has 95% efficiency focussing into the 1st order focal point , _ sans _ absorption . the material needed in a pfl to retard the phase
also will attenuate the flux of incident photons .
the index of refraction of matter can be expressed as @xmath19 and is related to atomic scattering factors @xcite .
thus for a photon of wavelength @xmath8 , a material of thickness @xmath20 will retard the phase by @xmath21 while attenuating the intensity by @xmath22 where @xmath23 .
the attenuating effects of the material in a fresnel zone of a stepped - pfl can be calculated by determining the amplitude of the waveform traversing each step of the pfl profile .
if @xmath24 is the material thickness of the @xmath25 step in a particular fresnel zone , the phase will be retarded by @xmath26 .
as shown in figure 1 , the @xmath25 step retards the phase between @xmath27 and @xmath28 , and the amplitude can be expressed as @xcite @xmath29 where @xmath30 is a normalization constant and @xmath31 summing over all @xmath9 steps leads to determining the intensity at the primary focus @xmath32 ^ 2 + \biggl[\sum_{i=1}^p\alpha_i^i \biggr]^2\biggr)\end{aligned}\ ] ] note that circular symmetry is assumed for the pfl , and this calculation is for a single fresnel zone .
if a pfl contains a total of @xmath33 individual fresnel zones with identical , in phase , profiles , the irradiance at the focus would be increased by @xmath34 as each fresnel zone has the same area on the pfl .
this formulation holds for any step spacing , regular or irregular , as long as a sufficiently small scale exists where the phase thickness is effectively constant .
choosing an energy of 8 kev , the efficiency of a @xmath18 regular - stepped pfl , including absorption , is 82.3% in silicon
. if absorption is ignored , the efficiency is 95% which is exactly that as determined from equation 1 for @xmath11 .
for a pfl with diameter @xmath35 , minimum fresnel ridge spacing @xmath36 , focusing at a photon wavelength @xmath8 , the focal length is given by @xmath37 where @xmath7 is in meters for @xmath36 in @xmath38 m , @xmath35 in cm , and @xmath39 in kev in equation 6 .
using the representative values of @xmath40 m , @xmath41 cm , and @xmath42 kev ( cu k-@xmath43 ) , the focal length would be 800 meters which is rather long for a ground - test . at nasa goddard space flight center
, a 600-meter interferometry testbed is available for testing of pfl optics .
the nominal configuration has an optics station 150 meters from an x - ray source and a detector station 450 meters from the optics station . assuming the x - ray emission is isotropic within the field - of - view of the optics , the effective focal length of an optic focussing at the detector station would be @xmath44 meters .
this sets the value of the focal length of a pfl for incorporation into this test beam configuration . using @xmath45 meters , @xmath46 cm , and @xmath42 kev ,
this leads to a minimum fresnel ridge spacing of @xmath47 m which is the natural scale size for micro - fabrication in silicon .
the fresnel ridge height needed to retard the phase by @xmath48 is given by @xmath49 where @xmath8 is the photon wavelength and @xmath50 is the real part of the index of refraction . for silicon ,
@xmath51 $ ] @xmath38 m or 20.5 @xmath38 m at 8 kev .
the gray - scale lithographic fabrication process has been employed at the university of maryland to create pfl structures in silicon wafers @xcite .
this implementation of the gray - scale process employs a lithographic mask that uses small , variable - transmission pixels ( gray levels ) that create , via projection photolithography , a designed , 3-dimensional structure in a photoresist spun on a silicon wafer .
this pattern is then transferred into the silicon via deep - reactive ion etching ( drie ) .
the developed ground - test pfl s have been fabricated using silicon - on - insulator ( soi ) wafers in order to minimize the thickness of the required mechanical substrate under the pfl s and thus maximize the x - ray photon transmission .
the sandwiched oxide layer forms a natural etch stop to remove the silicon substrate directly under each pfl while leaving the surrounding material for mechanical stability .
the unprocessed soi wafer was 100 mm in diameter with 70 @xmath38 m of silicon , 2 @xmath38 m oxide , and 500 @xmath38 m silicon forming the soi wafer structure .
a prototype silicon pfl fabricated using the gray - scale lithographic process is shown in figure 2 . .
copyright 2004 ieee.,height=226 ] ccccc pfl designation & diameter & @xmath36 & # of ridges & # steps / ridge + x3 & 2.99 @xmath52 & 24 @xmath53 & 32 & 16 + x4 & 2.99 @xmath52 & 24 @xmath53 & 32 & 16 + x5 & 2.99 @xmath52 & 24 @xmath53 & 32 & 8 + x6 & 4.72 @xmath52 & 15 @xmath53 & 80 & 8 + table 1 lists the four pfl designs that have been included in this ground - test fabrication .
note that this pfl fabrication incorporated a design to produce @xmath54 , as opposed to @xmath48 thick fresnel optics .
this was chosen for this initial fabrication to effectively double the minimum ridge spacing , p@xmath55 , for a set focal length , pfl diameter , and design energy .
although this will increase absorption losses , the relaxation of the @xmath36 requirement eased the constraints on the device fabrication .
the four pfl s , along with several test structures , were grouped to form a _ die _ which is compact in spatial extent .
twelve of these _ dice _ in a @xmath56 array were fabricated on the 100 mm soi wafer via a step - and - repeat process .
the goal of this fabrication was to produce a sample of pfl s for testing in the 600 m beam line , and the process was not optimized for yield . in order to identify the optimal pfl s for testing
, an optical inspection rejected those with obvious defects .
this rejected 15 out of the possible 48 pfl s .
the remaining pfl s were scanned via an optical profilometer ( veeco , wyko nt1100 ) to determine the accuracy of the fabricated profiles .
for the 3 mm diameter pfl s , the first and last 5 fresnel ridges were scanned and compared to the design profile .
for the 5 mm pfl , the 5 ridges near the half radius were also scanned and compared . using an analysis similar to that presented in equation 4 , albeit ignoring absorption and using a phasor formalism
, the efficiency of each scanned pfl was estimated from the profiles obtained from the profilometer measurements .
note that the profiles are measured along a chosen radial path and circular symmetry was assumed .
figure 3 illustrates the profile measurements and a comparison to the design profile for a 3 mm diameter pfl ( x3 ) for the regions near the center of the device ( leftmost plot ) and near the edge ( rightmost plot ) .
ccccccc & & & + & center & @xmath57 & center & @xmath57 & center & @xmath57 + average & 75.6% & 59.0% & 72.5% & 54.8% & 68.5% & 55.3% + maximum & 77.2% & 67.5% & 80.9% & 64.8% & 74.8% & 63.4% + minimum & 69.6% & 48.0% & 52.6% & 41.1% & 59.1% & 36.5% + cccc & + & center & @xmath57 & @xmath58 + average & 61.6% & 55.1% & 32.4% + maximum & 65.2% & 61.6% & 36.1% + minimum & 35.8% & 54.8% & 21.1% + table 2 lists the maximum , minimum , and average efficiency for the different fabricated pfl s based upon the profile measurements .
the values for the maxima and minima quoted for a pfl are that for a specific lens , i.e. the ensemble of measurements for a specific design were used to determine the appropriate designation .
the quoted efficiencies do not take into account absorptive losses due to either the fresnel profile or the @xmath59 m substrate .
assuming an 8 step / fresnel ridge profile and 8 kev , the reduction in collection efficiency is approximately 14% , i.e. @xmath60 , due to the phase - retarding material in the stepped - fresnel profile and 30% , i.e. @xmath61 , due to the @xmath62 m silicon substrate .
note that the effects of attenuation can be significantly reduced by fabricating pfl s designed for higher photon energies .
the data represented in table 2 demonstrate that , as indicated from profile measurements , stepped - profile pfl s micro - fabricated in silicon have efficiencies significantly larger than that for the simpler zone plates and phase - reversal zone plates .
the data also illustrate that efficiencies determined from the finer pitch fresnel zones are reduced as compared to the larger pitch center fresnel zones .
this is due to the fact that it is more difficult to accurately fabricate zones with higher aspect ratios , defined as the ratio of fresnel ridge height to ridge pitch .
a significant contribution to this effect is due to the aspect - ratio dependence of the etching process ; it is more difficult to remove silicon from narrow ridge regions as shown in figure 3 .
work has progressed on designing appropriate compensation in the lithographic mask and this technique has been demonstrated in the fabrication of a second - generation of fresnel lens structures that exhibit a much reduced aspect - ratio dependence of the pfl profiles @xcite .
we have fabricated ground - test pfl s in silicon using gray - scale lithography .
we have determined the imaging performance of these devices via analysis of the measured profiles of the fabricated optics .
these results indicate that the efficiencies , although less than ideal , are a significant improvement over the theoretical maximum that can be obtained with zone plates and phase - reversal zone plates .
we plan on introducing these devices into the 600 m test beam to demonstrate their imaging capability and verify the anticipated efficiency determination via _ in situ _ x - ray measurements .
this material is based upon work supported by the national aeronautics and space administration under grant apra04 - 0000 - 0087 issued through the science mission directorate office and by goddard space flight center through the director s discretionary fund .
g. skinner : astronomy and astrophysics , * 375 * , 691 ( 2001 ) g. skinner : astronomy and astrophysics , * 383 * , 352 ( 2002 ) h. dammann , optik * 31 * , 95 ( 1970 ) b.l .
henke , e.m gullikson , & j.c .
davis : at .
data and nucl . data
tables , * 54 * , 181 ( 1993 ) j. kirz , j. opt .
amer . * 64 * , 301 ( 1974 ) b. morgan , c.m .
waits , j. krizmanic , & r. ghodssi , jour .
micro - electro - mechanical - systems ( jmems ) * 13 * , 113 ( 2004 ) b. morgan , c.m .
waits , & r. ghodssi , microelectronic engineering * 77 * , 850 ( 2005 ) d. jerius , t.j .
gaetz , & m. karovska , proc .
spie * 5165 * , 433 ( 2004 ) | diffractive / refractive optics , such as phase fresnel lenses ( pfl s ) , offer the potential to achieve excellent imaging performance in the x - ray and gamma - ray photon regimes . in principle , the angular resolution obtained with these devices can be diffraction limited .
furthermore , improvements in signal sensitivity can be achieved as virtually the entire flux incident on a lens can be concentrated onto a small detector area . in order to verify experimentally the imaging performance ,
we have fabricated pfl s in silicon using gray - scale lithography to produce the required fresnel profile .
these devices are to be evaluated in the recently constructed 600-meter x - ray interferometry testbed at nasa / gsfc .
profile measurements of the fresnel structures in fabricated pfl s have been performed and have been used to obtain initial characterization of the expected pfl imaging efficiencies . _
@xmath0universities space research association + @xmath1nasa goddard space flight center , greenbelt , maryland 20771 usa + @xmath2dept . of electrical and computer engineering ,
university of maryland , college park , maryland 20742 usa + @xmath3cesr , 9 , avenue du colonel - roche 31028 toulouse , france _ |
m31 contains hundreds of x - ray sources in a relatively small field .
precision measurement of their positions allows identification of optical and radio counterparts .
long - term monitoring of these sources provides variability information on timescales which are not probed by single observations .
this information can help determine the nature of the x - ray sources .
for example , x - ray binaries containing high and low mass secondaries have somewhat different variability , and snr are not expected to show any variability at all . by using the x - ray variability and luminosity information to determine the nature of the sources one can establish links to the stellar populations in which the sources reside .
these links include the effects of star formation on the x - ray source population and the effects of galaxy evolution on x - ray source production .
several surveys of m31 have been completed in x - rays , finding hundreds of sources , a large fraction of which are variable .
most surveys have concentrated on the central bulge region , which contains most of the bright x - ray sources .
these studies began with @xcite , who used einstein data to catalog 69 objects brighter than @xmath09@xmath810@xmath9 in the m31 bulge and northern m31 disk .
@xcite found 2 variable x - ray point sources in the einstein data of m31 .
later @xcite performed a deeper survey of more than half the m31 disk by combining all einstein data of m31 .
they found 108 sources brighter than @xmath05@xmath810@xmath9 including fourteen additional variable sources .
the central 34 arcmin of m31 was surveyed with the rosat hri @xcite , revealing 86 sources brighter than @xmath010@xmath9 . by comparison with previous einstein observations
, they found nearly half of the sources in the bulge to be variable .
two more rosat surveys were completed with the pspc @xcite .
these surveys together covered most of the disk ( 10.7 deg@xmath10 ) and revealed 560 x - ray sources down to a detection limit of @xmath05@xmath810@xmath11 ; they found 34 sources varied in the 1 year between observations .
recently , m31 has been studied with the _ chandra _ x - ray observatory and xmm-_newton_. the improved resolution and sensitivity have led to additional interesting observations . for example , _ chandra _ observations have revealed several new x - ray transients @xcite , as have xmm observations @xcite .
@xcite discussed xmm-_newton _ and _ chandra _ observations of three of these which were discovered in the year 2000 . using xmm-_newton _ observations
, @xcite showed that the variability properties of one of the brighter sources in m31 indicate that it is a stellar mass black hole binary .
@xcite performed a survey of the central @xmath017@xmath12 of m31 with the _ chandra _ acis - i , finding 204 x - ray sources down to a detection limit of @xmath02@xmath810@xmath11 .
about half of the sources were variable on timescales of months , and 13 sources were transients .
@xcite used hrc - i data of the nuclear region to detect 142 sources brighter than @xmath02@xmath810@xmath11 , finding nearly half of the bright sources to be variable on timescales of @xmath1310 hours . @xcite
performed a deep xmm-_newton _ survey of the northern half in the disk , finding that the m31 disk is deficient in bright x - ray sources .
further xmm-_newton _ observations have discovered diffuse soft x - ray emission associated with the northern disk @xcite . @xcite
surveyed three widely separated portions of the m31 disk with acis - i , finding possible differences between the x - ray source populations in these different regions .
@xcite found that m31 globular clusters can be more x - ray luminous than those of the galaxy and suggested that this was due in part to the larger number of m31 globular clusters ( gcs ) rather than a difference in the shape of the luminosity function ( lf ) .
finally , @xcite have completed a survey for supersoft x - ray sources ( ssss ) and quasisoft sources ( qsss ) in four regions of m31 , finding 33 such objects .
we have obtained _ chandra _ hrc - i data covering most of the m31 optical disk .
these data provide the first opportunity to perform a large area survey of m31 with _
chandra _ , including regular information about the state of the detected sources over a period of two and a half years . while the sensitivity and coverage ( 0.9 deg@xmath10 ) are not as extensive as that of @xcite which covered an area of 10.7 deg@xmath10 , the time baseline is well sampled .
we have also obtained 3-epoch hst ( wfpc2 ) images of 2 newly discovered x - ray transients in order to search for their optical counterparts . among the deepest ground
based images of m31 are those that were obtained as part of the local group survey project @xcite ; we analyze unpublished sections of these data in order to search for new optical counterparts . in this paper
we use the _ chandra _ data to create an x - ray source catalog covering most of m31 and to measure long - term variability in the x - ray emission from these sources , and we use the newly obtained optical data to search for long wavelength counterparts . in
2 we present the x - ray data , source list and lightcurves . in 3 , we discuss the x - ray results , including lfs and variability studies . in
4 we describe the optical data used to search for counterparts to the x - ray sources . in 5 we discuss the results of the search for counterparts .
section 6 describes our detailed analysis of 2 x - ray transient sources detected in the optical with hst . finally ,
7 provides a summary of our conclusions .
the data for this project are originally part of a survey program to look for x - ray transients in m31 . nearly every month from november of 1999 to february of 2001 ,
_ chandra _ took hrc - i images of 5 fields covering most of m31 .
observations were then made every few months until june of 2002 .
each image was shallow ( @xmath01 ks ) but sufficient to detect any strong x - ray transients in the observed epoch .
individual epochs of these data were of limited use for survey purposes because of their short exposure times , but herein we stack the data into a deeper 17 ks hrc - i mosaic of m31 .
the observation i d ( obsid ) numbers , dates , exposure times and pointing coordinates of all of the observations are given in table [ obslist ] .
an exposure map of the stacked data is shown in figure [ expmap ] , where the thick white outline marks the region of the data where the 6@xmath14 sensitivity is at least 1.3@xmath810@xmath15(0.9 deg@xmath10 ) .
our lfs are complete to this luminosity ( see 3.1 ) .
we focused on this region in our analysis to provide a constant area of known sensitivity limit .
the combined , exposure corrected , background subtracted source image is shown in figure [ sim ] .
the image shows the majority of the sources are near the center of m31 .
the areas near center of the galaxy with the highest exposure have a @xmath16 detection limit of @xmath01.5 @xmath8 10@xmath9 .
in order to combine the data , they were first aligned using the ciao script _ align_evt _ , which corrects for small errors in the aspect solutions of different exposures by aligning the detected sources in the field .
the images of the bulge contained @xmath05 sources suitable for this purpose , which allowed alignment to an rms accuracy of @xmath17 .
unfortunately , in the disk , where there are fewer bright sources , this technique was less successful . in these outer regions ,
the fields were aligned using the _ chandra _ aspect solution , which is accurate to @xmath01@xmath18 . in the uncrowded outer regions , we binned the data to a resolution of 2@xmath18 .
this binning provided better detection of faint objects by removing the effects of the less precise alignment between exposures in the disk fields . in the central 18@xmath19 @xmath8 18@xmath19 of the galaxy , where the alignment was better , we binned the data to 1@xmath18 resolution in order to better match the instrumental point spread function ( psf ) . finally , we combined the data into 3 data sets using the task _
merge_all_. one set contained the data for the northern half of the galaxy , another contained the southern half and the last contained the center .
we searched for sources in the three data sets using the ciao task _
wavdetect_. we ran this task searching for sources on 4 size scales : 1 , 2 , 4 and 8 pixels .
the pixels in the merged images were 1@xmath18 in the central 18@xmath19 @xmath8 18@xmath19 and 2@xmath18 outside of this region . by searching on several scales _
wavdetect _ is able to overcome the large changes in the size of of the _ chandra _ psf from about 0.5@xmath18 near the center of the field to over 10@xmath18 in the outer regions of the field .
one hundred sixty - six sources were detected above our 3.5@xmath14 detection threshold .
their short names , positions , detection counts , signal - to - noise ratios ( @xmath20 ) , mean x - ray luminosities ( @xmath21 ; 0.110 kev ) , @xmath22 values for a fit to a constant source , counterparts or previous x - ray detections , and references for those counterparts and previous detections are provided in table [ slist ] .
detailed descriptions of all columns , including a definition of the source naming convention ( r1 , r2 , n1 , s2 , etc . ) are given in the notes to table [ slist ] .
to convert our measured mean x - ray fluxes to luminosities , we assumed a distance to m31 of 780 kpc , a photon index @xmath23 , and absorption @xmath24 @xmath25 .
these assumptions provided a constant conversion factor of 2.5 @xmath8 10@xmath26 erg cm@xmath10 ct@xmath28 from hrc flux in ct @xmath25 s@xmath28 to luminosity ( 0.110 kev ) .
we consulted previous surveys of the x - ray source background to set a limit on the number of possible background sources in our sample .
rosat observations of the lockman hole @xcite show that there are @xmath1340 sources per square degree with luminosities greater than 4@xmath810@xmath9 .
our large - area disk lf quickly falls off below this luminosity due to completeness ( cf .
therefore we estimate that our entire catalog contains @xmath2940 background and/or foreground sources . near the center of m31
, our data contain an area of 0.01 deg@xmath10 complete at 6@xmath14 to @xmath02.5 @xmath8 10@xmath9 ( see the area of highest exposure near the center of m31 in figure [ expmap ] ) .
this area should contain @xmath0100 background sources per deg@xmath10 @xcite , or 1 source in our most sensitive area , where we detect 41 sources .
therefore contamination near the galaxy center is very small . within the central 10 arcmin of m31 ,
all but one of the non - transient sources in our list has been previously detected in the x - ray band .
this region has been well studied with _
chandra _ already @xcite .
source r2 - 68 appears to be a clean new detection , with 19 counts and a s / n ration of 4.6 .
while we do not detect any variations in the hrc snapshots of this source ( @xmath30 ) , the fact that it is not detected in the much deeper acis @xcite and 50ks hrc exposures @xcite indicates it is variable .
as shown in figure [ varsen ] and discussed in 3.2 , our sensitivity to variability in the fainter sources of our survey , such as r2 - 68 , is limited . outside of the central region
, we detect several new objects .
of the new objects , s1 - 76 , s1 - 82 , and n1 - 84 are well detected , with @xmath20 of 9.3 , 8.4 , and 4.8 respectively .
these objects are all brighter than 10@xmath9 ; therefore their non - detection in the rosat surveys ( which reached 5@xmath810@xmath11 ) indicates that they could be variable on timescales as long as a decade .
while we do not detect variability in these four sources between the individual hrc - i snapshots , the counting rates are low enough that variations of a factor of two may be undetected .
in addition , there are several new sources that varied significantly during our survey . among these variable sources
are s1 - 79 , which is a strong transient with a peak luminosity of 1.5@xmath810@xmath31 and peak @xmath20 of 35 , and s1 - 1 , which is variable on year long timescales with a @xmath32 test for a steady source giving @xmath22 of 1.4 ( probability = 13% ) . by combining the data from our fields
, we were able to construct lightcurves covering more than 2 years for all of the objects we detected .
initial lightcurves for the objects were measured using the ciao task _ lightcurve _ , which measures the number of counts in a square aperture around the detected object in each epoch .
we measured source fluxes in each of the 17 epochs detailed in column 3 of table [ obslist ] .
the lightcurves were measured using boxes with 0.7@xmath18 sides within 3@xmath18 of the nucleus where the sources are separated by only 1@xmath18 . this box size
was increased to 2@xmath18 in the central 1@xmath19 , where the sources are generally separated by more than 2@xmath18 .
the box size was increased to 8@xmath18 in the central 18@xmath19 @xmath8 18@xmath19 ( outside of the central 1@xmath19 ) , where the sources are generally separated by more than 8@xmath18 . outside of this region ,
the sources are generally separated by over an arcminute . because these sources are typically located farther off - axis , where the psf is larger
, we used a 16@xmath18 box to measure them .
the average radius of a circle that encloses 90% of the energy at 1.49 kev in the hrc - i is 1@xmath18 at 1@xmath19 off - axis , 4@xmath18 at 6@xmath19 off - axis , and 8@xmath18 at 8@xmath19 off - axis .
these box sizes therefore insured that we measured most of the source counts in off - axis areas where crowding of the sources was not an issue .
the 8@xmath18 box size typically contained only @xmath01 background count per ks timebin , while the 16@xmath18 boxes typically contained @xmath04 background counts per ks timebin .
the background subtracted count rates from _ lightcurve _ were used as a starting point for our lightcurve measurement for each source .
we converted the _ lightcurve _ output from units of counts to flux units using exposure maps which take into account the aspect histograms and instrumental flat field .
as the instrument s effective area depends on the source spectrum , we made the exposure maps assuming a typical m31 source spectrum of an absorbed power law with an index of 1.7 and an absorption column density of @xmath33 @xmath25 .
these maps were created for each epoch of the lightcurve to measure the effective exposure for each object in each epoch individually .
the _ lightcurve _
output was then converted to flux units using the effective exposure of each object in each observation .
finally , the fluxes were multiplied by our conversion factor of 2.5 @xmath8 10@xmath26 erg cm@xmath10 ct@xmath28 to estimate the source luminosity .
table [ slist ] lists the @xmath22 fit of each lightcurve to a constant flux at its mean value .
the objects have a wide range of variability .
sources with @xmath22 values of greater than 1.47 are discussed as variables in 3.2 .
several of those with the highest @xmath22 values ( @xmath3410 ) are transients , discussed in detail in 3.2 and 6 .
the 2@xmath18 box size used for the central arcmin of m31 was appropriate for measuring the lightcurves for all of the central sources except for the well - known 3 bright sources in the nucleus of the galaxy @xcite .
our _ wavdetect _ analysis detected these sources as the single ( extended ) source r1 - 9 .
the lightcurve of these three sources blended together is shown in the top panel of figure [ nucleus ] .
the combined lightcurve reveals variability of the nuclear region by a factor of @xmath010 on timescales of about 6 months .
we attempted to create 3 separate lightcurves for the 3 known objects by dividing the blended nuclear region into 3 parts .
we made lightcurves for each of these parts using a boxsize of 0.7@xmath18 .
while the northernmost of these three is closest to the nucleus @xcite , it is unclear if the source is associated with the central massive black hole .
the 3 lightcurves are shown in figure [ nucleus ] , but the lightcurves appear to be influenced by one another , revealing the spatial limitations of our data set .
even so , we see that the part closest to the nucleus , corresponding to cxom31 j004244.3 + 411608 , shows the least variability .
it appears as constant and faint throughout the observations .
the highest level of variability among the three central sources is shown by the bright , soft source immediately to the south ( cxom31 j004244.3 + 411607 ) which was initially ( and incorrectly ) associated with the central black hole @xcite .
the lightcurve of the source furthest to the south ( cxom31 j004244.3 + 411605 ) shows less variability but mirrors that of its neighbor to the north , suggesting that the two are not completely resolved .
in addition to looking for long - term variability in the nuclear region , we were also able to look for short - term variability using the long exposure of @xcite , where 3 sources are clearly resolved .
the lightcurves of the three objects showed no variability on this timescale .
they all had @xmath22 values of less than 1 when fit to a constant flux , including the soft source which is so highly variable on longer timescales ( @xmath22 = 18.4 ) . in order to better constrain our lightcurves for the objects in the central region we used the data from @xcite which covers 18@xmath19 @xmath8 18@xmath19 about the center .
this deep observation provided excellent @xmath20 for the november 2001 data points in our lightcurves .
objects near the center of the galaxy therefore show very small errors for their fluxes during that epoch ( cf .
figure [ vars ] ) . farther out in the m31 disk ,
the density of bright x - ray sources is small .
this low density of sources allowed larger spatial binning .
this binning was especially useful in the outer parts of the hrc fields because the effective exposure is lower in these outer portions and the _ chandra _ psf is significantly broader on the outskirts of the field . aside from the lower resolution
, the lightcurves were measured the same way as described for the central region .
the lightcurves for the variable objects in our sample are shown in figures [ vars ] , [ trans ] , [ stars ] , and [ 143148 ] .
these sources will be discussed further in 3.2 and 6 .
in addition to our hrc analysis of the lf and variability of the x - ray sources detected in our survey , we analyzed acis observations of the x - ray transients r2 - 67 and r3 - 16 , the two transients for which we found counterparts in our hst followup data .
we applied two analysis techniques to attempt to recover the spectra of r2 - 67 from the acis observations .
first , we used the pile - up model of @xcite as coded in isis 1.0.50 @xcite , ciao 3.0/sherpa @xcite and xspec v11.2 @xcite .
second , we extracted counts only from the wings of the psf which are not piled - up due to their lower counting rates .
each technique has limitations as described in the appendix .
fitting the spectrum of r3 - 16 was more straight - forward , as it was not piled - up .
the spectrum was only fit using ciao 3.0/sherpa .
in all cases we corrected the instrumental response for the contamination build up on the acis detectors , and we limited our analysis to the 0.30.7 kev range where the background is low and the calibration is secure .
counts were grouped into bins containing @xmath3415 counts to allow standard @xmath22 statistics , and error ranges are 68% as determined from @xmath22 contours .
the details of the fitting procedures are provided in the appendix .
the fitting results of r2 - 67 are discussed in detail in 6.3 , and those of r3 - 16 are discussed in 6.4 .
in order to look for differences between the disk and bulge source populations , we measured the lf of the source population within 5 arcmin of the nucleus ( the bulge ) and outside 7 arcmin of the nucleus ( the disk ) . to facilitate comparisons to previous work ( i.e. , @xcite ) , we excluded gcs from our disk sample , and we generated lfs for the bulge both excluding and including globular clusters within 5 arcmin of the nucleus .
we also provide the globular cluster luminosity function ( gclf ; cf .
figure [ gcdlfs ] ) .
we always excluded 3 sources likely to be foreground stars and 2 sources associated with m32 ( see table [ slist ] ) .
x - ray fluxes were converted to luminosities using the conversion described in 2.2 .
the differential luminosity functions ( dlfs ) of the disk and bulge ( cf . fig .
[ dlfs ] ) show some interesting differences .
the lfs of the bulge with and without gcs
are statistically equivalent ( cf . table [ bplfits ] ) .
the difference in completeness between the disk and bulge samples is evident .
the bulge sample contains a large number of sources at @xmath02.5@xmath810@xmath9 , faintward of which the dlf falls off steeply , revealing the depth of the data . in the disk ,
the dlf falls off in a similar fashion at @xmath04@xmath810@xmath9 , reflecting the shallower depth in the disk observations .
this difference in completeness is due to the variable psf of _ chandra _ , which is smaller in the bulge region therefore allowing fainter sources to be detected .
we created @xmath20 histograms for our source list , revealing a peak in the number of sources with 6@xmath14 detections .
we therefore consider our sample complete for detections of 6@xmath14 and higher .
a source near the center of the bulge with a luminosity of 2.5@xmath810@xmath9 will be detected at 6@xmath14 , while a luminosity of @xmath03.5 @xmath35 is required for a 6@xmath14 detection in the disk .
the increase in the _ chandra _ psf with off - axis angle and corresponding drop in sensitivity is somewhat mitigated by the fact that the field - of - view of the observations overlaps at the largest off - axis angles , doubling the exposure time in these regions .
the @xmath36 detection limit is @xmath37 in these high exposure regions , but @xmath38 in nearby non - overlapping areas . in the non - overlapping areas farthest off - axis , the @xmath36 detection limit is @xmath39 . for the remainder of the analysis of the disk lf , we only considered sources with @xmath40 , and for the remainder of the analysis of the bulge lf , we only considered sources with @xmath41 .
interestingly , the lack of disk sources with luminosities @xmath3410@xmath15 is not as pronounced as seen in the xmm survey of the northern disk @xcite . while there is certainly a decrease in the number of bright sources with distance from the center of m31 , our study hints at an additional , more subtle effect .
the southern half of the disk contains a large fraction of the most luminous disk sources .
these sources are shown in figure [ hlll ] , which shows objects with luminosities @xmath42 as crosses in the right panel , and objects with luminosities @xmath43 as circles in the center panel .
the effect discovered by @xcite is apparent : there are very few bright sources in the northern disk
. however , the same panel shows several bright sources in the southern disk .
there is also a hint of this effect in the data set of @xcite who compared several regions of the disk .
their field 2 , which lies in the southern disk , contains most of the bright sources in their sample as well , but this field was also located closest to the galaxy center . of the 9 bright sources south of 41 degrees declination ,
tests to a steady source find @xmath44 only for 3 of them ; the rest have @xmath45 .
the cumulative luminosity functions ( clfs ) of the disk and bulge are shown in figure [ clfs ] , and the results of the broken power - law fits are provided in table [ bplfits ] .
the bulge clf appears qualitatively more complex than that of the disk , but it is adequately fit by a broken power - law with a break at about 7.0@xmath46 @xmath8 10@xmath15 .
this break is higher than the typical values of @xmath02.1@xmath47 seen in previous surveys ( e.g. @xcite ) .
our luminosities are 0.110 kev while the luminosities of those surveys were for narrower energy ranges ( e.g. , 0.37 kev ; @xcite ) ; our wider energy range accounts for only @xmath040% of the discrepancy , assuming a typical spectrum ( @xmath48 , @xmath49 @xmath25 ) .
our break luminosity is more easily compared to that measured by @xcite which was measured with luminosities of the same energy range by applying the maximum likelihood technique to deeper ( 50 ks ) hrc data .
our clf break measurements agree at the @xmath01@xmath14 level with his measurement of @xmath50 .
our maximum likelihood fit broken power - law has a sharp break , with slope indices of [email protected] and [email protected] above and below the break , respectively .
these values for the slopes are also consistent with those measured by @xcite .
the measured values for these parameters were statistically unaffected by the presence of gcs in the sample .
the fit is shown on the clf in the bottom panel of figure [ clfs ] ; this clf contains no gcs .
monte carlo tests show that 50% of samples taken from such a broken power - law distribution provide better fits than our sample .
the disk sample also shows a possible broken power - law consistent with previous observations @xcite . a maximum likelihood fit to these data gives slopes of [email protected] below the break and [email protected] above the break with the break at @xmath5110@xmath15 .
this fit is shown in the top panel of figure [ clfs ] .
this distribution is comparable to the inner disk sample of @xcite .
monte carlo tests show that 56% of samples taken from such a broken power - law distribution provide better fits than our sample . the best - fit single power law ( cf .
[ clfs ] top panel ; table [ plfits ] ) has slope [email protected] .
monte carlo tests show that 71% of samples taken from such a power - law distribution provide a better fit than our sample . while the broken power - law fit is better , both fits are adequate .
our most conservative clf for the disk only includes sources brighter than 1.3@xmath810@xmath15 . above this luminosity ,
the entire region of our survey is complete .
this sample is well - fit by a single power - law with index [email protected] , also shown in figure [ clfs ] .
monte carlo tests show that only 7% of samples taken from such a power - law distribution provide a better fit than our sample .
this index is similar to the value of [email protected] found for the northern disk by @xcite to a faint limit of 10@xmath9 .
the higher break luminosity , and slightly flatter clf below the break of the bulge is consistent with the brightest sources being in the bulge .
the steeper clf in the disk was also seen by @xcite ; however , their sample did not extend more than 9 arcmin from the galaxy center .
in contrast , this disk sample contains sources from 7 to 72 arcmin from the nucleus , showing that the steep clf extends far out into the disk and the bulge contains most of the bright sources .
our disk clf can be compared to that seen in the spiral galaxy ngc 6946 in a survey by @xcite with a similar sensitivity limit ( @xmath010@xmath15 ) .
their sample is clearly disk dominated , as the source distribution traces the spiral arms of the galaxy .
@xcite found that the clf of ngc 6946 is a well - behaved power - law with slope [email protected] .
the slope in ngc 6946 is consistent with the recent conclusion by @xcite , from a sample of x - ray point sources in 32 galaxies , that the clfs in late - type spiral galaxies have slopes of 0.50.8 .
a steeper slope is seen in the m31 disk in both the large ( [email protected] ) and the most conservative ( [email protected] ) samples .
the slope difference may reflect the difference in star formation rates of the galaxies , which are @xmath04 and @xmath131 m@xmath52 yr@xmath28 in ngc 6946 @xcite and m31 @xcite , respectively .
the lower current star formation rate in the m31 disk may not replenish its short - lived , high - luminosity x - ray sources .
such a process has been shown to be responsible for steeper clfs in x - ray population models @xcite .
it is interesting to note that the disk clf for our most conservative sample is similar to the typical slopes found in elliptical galaxies ( @xmath01.4 ) in the @xcite sample .
the only one of these early - type galaxies with a measured star formation rate is ngc 5128 ( 1.7 m@xmath52 yr@xmath28 ) , which has a clf slope of 1.28 .
because elliptical galaxies typically have little or no current star formation , the rate measured for ngc 5128 may be taken as an upper limit for the other early - type galaxies in the @xcite sample .
then the clf slope and star formation rate of the m31 disk are typical of what is seen in the early - type galaxies of their sample .
@xcite were not able to remove gc sources from their sample or break down the sources into bulge and disk populations .
ideally , we would like to compare the m31 disk - only sample to disk - only samples of more distant galaxies . assuming that the @xcite disk galaxy samples are dominated by disk sources , the slope of the x - ray clf of the m31 disk bears a stronger resemblance to those of typical ellipticals than to those of typical spirals .
the gclfs of our sample are shown in figure [ gcdlfs ] .
the figure shows the dlf of all gcs in the sample as well as a break - down of the sample into gcs near the center of m31 and farther out in the m31 disk .
the outer gc sample contains the brightest objects in the survey .
a power - law fit to the total gclf down to 1.3@xmath810@xmath15 yields a slope of [email protected] ( cf .
table [ plfits ] ) .
only 0.4% of 10000 monte carlo tests provided a better fit to this slope than our gc sample .
it should be noted that our lfs for m31 could be affected by the long baseline and short exposure times of our data set .
for example , a typical transient is likely to be detected at near its peak luminosity in one ( or a few ) exposures , and undetected in the majority of the remaining exposures .
in the extreme case where it is detected at its peak in a single exposure , and undetected in all other exposures , the mean luminosity ( used in constructing the lf ) is 1/17 of peak . given that the duty cycle of transients is likely @xmath01% , this extreme example overestimates the mean luminosity by a factor of @xmath06 .
the more typical observational mode , consisting of a single long exposure , may contain only a single ( or few ) transient(s ) at intermediate luminosity , but our survey detected over a dozen transients .
this bias will be most severe for sources in the bulge , where the majority of the transients are located . to test the effect of the transients on the clf of the bulge
, we created a bulge lf with the transient sources removed .
the results of a broken power - law fit to this clf is also given in table [ bplfits ] , and is consistent with the fits to bulge samples both including and excluding gcs .
any source for which a fit to a constant source at the mean flux yields @xmath53 has a 90% probability of being variable in our 17 epoch survey . when fitting a model to a data set with 16 degrees of freedom ( @xmath54 )
, a value of @xmath55 leaves a 10% chance that the model is the correct representation of the data set ( see @xcite for details ) .
there are 44 objects with @xmath53 in our sample .
the lightcurves for these objects are shown in figures [ vars ] , [ trans ] , [ stars ] , and [ 143148 ] . among these 44
are 1 ( r1 - 9 ) quasisoft source ( qss ) and 1 ( r2 - 12 ) supersoft source ( sss ) , as diagnosed by @xcite , and 9 transient candidate objects .
there are also 8 transient sources with @xmath56 ; one of these ( r3 - 115 ) is also a sss .
we defined transient candidates as those objects whose luminosities reached 1.5@xmath810@xmath15 for at least one epoch , but less than 6 epochs . these objects were also required to have luminosities consistent with 0 in at least 2 epochs .
all of the timebins where the luminosity was consistent with 0 were combined to measure upper limits of these objects during quiescence .
we also required a 1@xmath14 upper limit below 2.5@xmath810@xmath9 during the combined quiescent epochs . the peak luminosities and 1@xmath14 upper limits of the quiescent luminosities for the transient candidates
are listed in table [ tlist ] .
some of the candidates have low @xmath22 values due to large errors in the unbinned quiescent epochs .
a few of these transient candidates have been detected in previous surveys .
object r3 - 125 was detected by @xcite at a luminosity of @xmath010@xmath15 , but this object did not appear in acis observations with a detection limit of @xmath57 , confirming its transient nature .
objects r3 - 126 and n1 - 85 have been seen in previous surveys and are known to be repeating transients @xcite .
object r3 - 115 has been recently identified as a supersoft source ( sss ) by @xcite ; this is the only transient source in our catalog corresponding to an sss in their catalog .
the position of n1 - 85 is coincident with a known snr candidate . while the high variability rules out the possibility of this source being the snr itself , the source could be associated with the snr .
objects r1 - 5 , r2 - 16 , r2 - 3 , r2 - 63 , and r3 - 16 have been classified as transients in previous surveys @xcite ; however , they were active too often during our survey to meet our criteria .
figure [ active ] shows a histogram of the number of active transients for each epoch of the survey . on average
there are 2@xmath51 active transients in m31 during any given epoch .
a new ( or recurrent ) transient becomes active about once every 12 months .
it is interesting to note that half of our transient candidates did not fit our variability criteria , highlighting the fact that the fraction of variables we quote truly is a lower limit .
this fact is made clear in figure [ varsen ] , which shows the fraction of variables , and the fractional change in flux necessary to show a 1@xmath14 change , as a function of luminosity in our data set .
forty of the 44 variables had mean luminosities greater than 7@xmath810@xmath9 , where a flux change of more than 40% could be detected .
half of our sources were below this luminosity . therefore ,
about half of the sources for which our data set is sensitive to variability at the @xmath040% level ( @xmath58 7@xmath810@xmath9 ) are variable .
it is entirely possible that half of the sources with mean luminosities @xmath597@xmath810@xmath9 are also variable at the @xmath040% level .
the x - ray transient population of m31 is discussed in further detail in 6 , including a discussion of the x - ray spectral properties of 2 x - ray transient sources , r2 - 67 and r3 - 16 , as measured with acis .
during the course of our survey , wide - field data from the local group survey ( lgs ) covering most of the m31 disk became available @xcite .
in addition , new optical data were obtained by hst through coordinated observations to search for optical counterparts of five new bright x - ray transient events .
we report on the detection of 2 counterparts in 6.3 and 6.4 , and will report on the 3 non - detections in a future paper . in this section
we detail our analyses of the lgs and hst data and report the results of our search for potential optical counterparts to all the hrc x - ray sources .
new optical data used for this project were generously supplied by the ctio / kpno local group survey ( lgs ) collaboration ( @xcite ; + http://www.lowell.edu/users/massey/lgsurvey ) which is acquiring 1 arcsec resolution , photometric data with the 8 k @xmath8 8 k mosaic cameras on the 4-m telescopes at kpno and ctio , entirely covering ten local group galaxies in @xmath60 , and narrow - band h@xmath61 , [ s ii ] ( @xmath626717 , 6731 ) and [ o iii ] ( @xmath635007 ) .
the lgs is working on their own , more rigorous , calibration of these data leading towards a complete @xmath60 catalog of stars .
however , for the purposes of this paper , we have simply used photometry from the literature to perform a rough calibration . the analysis used in this paper
is described in full detail , including tests of the photometry routine , in @xcite . in short , the data consisted of 7 fields from the mosaic camera on the kpno 4-m telescope .
observation dates are shown along with the hst observation dates in tables [ r267obs ] and [ r316obs ] .
these fields cover most of the active portions of the m31 disk , but do not cover the bulge . in order to look for bright , blue stellar counterparts to the x - ray sources
, the johnson @xmath64 and @xmath6 broadband images were analyzed .
the dithered frames in each filter were stacked and reduced using the daophot ii and allstar packages @xcite , and the zero points for each field were determined using published photometry from previous surveys of @xcite and @xcite . optical data for the x - ray transients r2 - 67 and r3 - 16 were obtained within weeks of the x - ray outbursts through the f336w ( @xmath2-band equivalent ) filter of the wfpc2 camera aboard the _ hubble space telescope _ ( hst ) .
eight exposures of 500 seconds each were taken at 3 separate visits for each of the 2 fields observed .
observation dates and exposure times are listed in tables [ r267obs ] and [ r316obs ] . these 8 images from each visit were analyzed using the automated photometry package hstphot @xcite , which is optimized for the processing and photometric measurement of undersampled ccd images like those of wfpc2 .
the package masks out all known bad pixels in the field as well as hot pixels flagged by their deviations from the measured hst psf .
the images are then corrected for minor misalignments between exposures and combined .
the combined images are searched for all source detections with a @xmath20 greater than 4 .
the quality of the psf fit to each source is measured to distinguish blends , and the wfpc2 magnitudes are measured by correcting for the charge transfer efficiency and applying the photometric calibration of @xcite .
we applied this analysis method to the 8 exposures from the 3 different epochs for each transient observed .
the combined final images of each epoch for each object are shown in figure [ hstims ] .
the detections of the uv counterparts of the x - ray transients via hstphot was therefore objective and provided the f336w ( @xmath2-band equivalent ) magnitudes and errors listed in tables [ r267obs ] and [ r316obs ] .
the implications of these measurements are discussed in 6.3 and 6.4 .
the range in optical magnitudes to be expected for counterparts of x - ray sources in m31 can be estimated by scaling the optical magnitudes of the counterparts of galactic sources .
galactic high - mass x - ray binaries ( hmxbs ) typically contain stars of spectral type o8b3 @xcite .
these stars have -5@xmath29m@xmath65@xmath29 - 2 , or 19.8@xmath29@[email protected] when scaled to the distance ( 780 kpc ) and reddening ( @xmath66 ) of m31 .
galactic low - mass x - ray binaries ( lmxbs ) typically have -5@xmath29m@xmath65@xmath295 @xcite , or 19.8@xmath29@[email protected] when scaled to m31 .
these estimates suggest that we may find optical counterparts for hmxbs and bright lmxbs by searching currently available wide - field surveys , which go to a depth of @xmath67 ( e.g. @xcite ) outside of the bulge .
only 17 stellar counterpart candidates have been found for our sample .
the small number of high - mass stellar counterparts in the disk is consistent with the disk xlf , which suggests that the m31 disk contains very few young bright x - ray sources ( cf .
only such young sources are likely to have short - lived , high - mass stellar companions .
the small number of optical counterpart candidates found in the bulge is consistent with the effects of crowding in ground - based surveys of the bulge , which limit the depth to @xmath68 ( see upper limits for bulge sources in table [ lgslist ] ) .
finding counterparts for most x - ray binaries in the bulge will require deeper , higher resolution , optical data .
some hmxbs are transient x - ray sources .
the broad - band optical flux of these objects is dominated by the high - mass donor star and varies by @xmath00.4 mag ( e.g. @xcite ) .
even this low - amplitude variability in the broad - band optical flux may not be correlated with the variability of the x - ray flux ( e.g. @xcite ) .
the optical magnitudes of these sources are therefore given by the magnitudes of the high - mass companions , which are typically be stars with spectral type o8.5b2 @xcite .
these stars have optical magnitudes of -4.8@xmath29m@xmath65@xmath29 - 2.4 , or 20.0@xmath29@[email protected] when scaled to m31 .
some lmxbs are transient x - ray sources .
cyg , one of the most optically luminous galactic x - ray transients , had an outburst with @xmath6=12.7 , @xmath69 , @xmath70 and a reddening of @xmath71 @xcite at a distance of 3 kpc @xcite . scaling to m31 , we would expect @xmath2=22.4 , @xmath64=22.7 , and @xmath6=22.1 .
an example of an optically fainter galactic x - ray transient lmxb is the short - period system a0620 - 00 .
such short - period systems are less luminous than long - period systems such as v404 cyg @xcite .
the outburst of a0620 - 00 had @xmath6=11.2 , @xmath72 , @xmath73 and @xmath74 @xcite at a 1.2 kpc distance @xcite . scaling to m31
, we would expect @xmath2=23.4 , @xmath64=24.3 , and @xmath6=24.4 .
assuming these two galactic examples bracket the typical optical luminosity range for x - ray outbursts in lmxbs , the expected range in the visual magnitude of lmxbs in outburst in m31 is 22.1@xmath29@[email protected] . while bright low - mass transients should be detectable in outburst , they are @xmath755 mag fainter , and therefore undetectable , in quiescence . in addition ,
typical gcs in m31 have @xmath6 magnitudes of 1519 @xcite , so that gc counterparts are easily identified .
there are also numerous catalogs of emission line sources in m31 that include potentially x - ray bright objects , such as supernova remnants
. some of these optical counterparts may also be identified by cross - correlating the positions in the optical catalogs with the x - ray source positions .
we searched the literature for previous detections of our sources in x - rays and at longer wavelengths .
the vast majority of these sources have no known counterparts at longer wavelengths , as has historically been the case for x - ray sources in m31 .
all but 7 of these sources have been previously detected in x - rays while only 55 ( out of 166 ) of them have been detected outside the x - ray band .
twenty - six of these are known gcs .
seventeen are stars from recent ground - based broad band surveys and the local group survey ( cf .
one is a radio source classified as a bl lac candidate .
three are coincident with snr candidates .
one more is a radio source likely to be an snr .
six are classified as planetary nebulae ( pne ) , but are more likely snr since they are x - ray bright , and one is an emission line object of unknown nature that is optically bright . shifting our x - ray positions by 14@xmath18 and applying an identical search for counterparts yields 0 globular clusters , 3 stars , 3 pne , and 0 snrs .
the 14@xmath18 shift was larger than the psf to avoid any real counterparts , but was not so large as to change the surface density of m31 objects .
the number of counterparts found for the shifted positions therefore indicates the expected number of random coincidences between x - ray and optical sources in our sample . in some cases , the variability of an object was helpful for determining the validity of possible counterparts .
for example , r1 - 15 and r1 - 2 have coordinates consistent with the coordinates of planetary nebulae ford 17 and ford 13 respectively . on the other hand , these sources are unlikely to be planetary nebulae because they are luminous ( @xmath76 ) , and they are unlikely to be snrs because they are significantly variable .
these sources are therefore more mysterious in origin than we may have suspected in the absence of the long - term variability information .
four other objects r1 - 24 , r1 - 23 , r1 - 26 , r3 - 7 are likely to be snrs since they have been classified as pne and also have constant strong x - ray flux . by comparing our x - ray positions to optical data from the local group survey ( lgs ) obtained with the mosaic camera on the kpno 4-m at 1@xmath18 resolution @xcite
, we were able to find several new optical counterpart candidates .
these candidates are listed in table [ lgslist ] .
the regions currently surveyed include @xmath075% of the disk , but very little of the bulge , where most of the bright x - ray sources lie .
crowding in ground - based images of the m31 bulge severely limits searches for counterparts , so that comparisons to future lgs data from the bulge is unlikely to yield many new counterparts .
four of these new candidates have appropriate colors and magnitudes to be foreground stars ( s1 - 74,s1 - 45,n1 - 82,n1 - 59 ) .
typically , galactic foreground stars in the direction of m31 have colors of @xmath77 with the highest number having colors of @xmath78 @xcite .
many m31 stars have these colors as well , so that the only way to be sure a star redder than @xmath79 is in m31 is with spectroscopy . because such spectra are not available to us , we assume the stars most likely to be foreground are bright red stars ( @xmath80 , v@xmath1318 ) . of these foreground candidates , objects s1 - 74 and n1 - 82 were too faint to obtain a reliable measurement of variability , and s1 - 45 is unlikely to be variable , with a @xmath22 value of 1.25 . object n1 - 59 is especially interesting , as it is a transient candidate .
on the other hand , there were also 11 counterparts with colors and magnitudes typical of m31 upper main sequence or giant stars ( s1 - 75 , s1 - 78 , s1 - 64 , r3 - 28 , r2 - 8 , r2 - 67 , r3 - 115 , r3 - 16 , r3 - 13 , r3 - 7 , n1 - 81 ) .
object s1 - 75 is associated with a bl lac candidate .
objects r2 - 8 , and r3 - 115 have colors appropriate for m31 red giants .
the x - ray transients r2 - 67 and r3 - 16 are discussed in detail in 6.3 and 6.4 .
objects s1 - 78 , s1 - 64 , r3 - 28 , r3 - 13 , r3 - 7 and n1 - 81 have the colors and magnitudes of evolved and/or reddened m31 upper main sequence stars .
the counterpart candidates most likely to be in m31 are those with b - [email protected] ( s1 - 64 , r3 - 28 , r3 - 13 and r3 - 7 ) .
although these candidates are all slightly @xmath751@xmath18 from their possible x - ray counterparts , their x - ray positions have errors ( given by _ wavdetect _ ) of 0.6@xmath18 , 0.5@xmath18 , 0.5@xmath18 , and 0.4@xmath18 respectively , and the lgs coordinates are accurate to about 1@xmath18 , allowing the possibility that these are the true optical counterparts the majority of these possible stellar counterparts have too few x - ray photons per 1 ks exposure to generate meaningful lightcurves , but s1 - 75 , s1 - 64 , and n1 - 59 are brighter .
the x - ray lightcurves for these 3 are shown in figure [ stars ] .
tests to a constant source find @xmath81 , and 0.68 for s1 - 75 , s1 - 64 , and n1 - 59 respectively .
variability is therefore detected only in s1 - 75 , the known bl lac candidate .
there were also several x - ray sources in regions observed by the lgs that did not correspond to any point sources in the lgs optical images .
for these objects , we obtained lower limits for their @xmath6 band magnitudes by measuring the completeness of the area of interest as a function of @xmath6 magnitude .
the lower limit was chosen as the v magnitude where the mean completeness of the lgs data fell below 50% .
these limits are provided in table [ lgslist ] .
x - ray transient sources located in the bulge ( within 7@xmath19 of the nucleus ) are labeled with a ( b ) in table [ tlist ] .
these are likely to be lmxb transients because most stellar population and interstellar medium studies of the m31 bulge suggest that it is dominated by old stars and contains very few , if any , young stars ( e.g. @xcite ) .
indeed , the central 44@xmath18 contains no stars earlier than b5 @xcite , and only an upper - limit to the star formation rate of the bulge has been measured from the far - uv luminosity ( @xmath82 ; @xcite ) .
in addition , with deep hrc imaging taken during the outburst of r2 - 67 , @xcite found no evidence for the existence of x - ray pulsars in the m31 bulge , suggesting it lacks hmxbs .
while the possibility of a high mass secondary in the bulge is low , it is not zero .
the bulge is extremely crowded , and the highest resolution surveys have not covered the entire region or analyzed spectra of all bright blue stars .
in addition , the discovery of molecular clouds @xcite and supernova remnants @xcite in the m31 bulge allow the possibility for some recent star formation .
the lgs optical counterpart candidates for r2 - 8 , r2 - 67 and r3 - 115 have absolute @xmath6 magnitudes ( neglecting absorption ) of m@xmath83 and @xmath84 , respectively , consistent with high mass ( e.g. , be binary ) counterparts .
the colors measured for these objects from the ground - based data are not completely reliable , as the @xmath64 and @xmath6 observations were separated by nearly 1 year and be stars are typically variable @xcite .
however , we note that apart from times of major outbursts , the scale of variability is small so that these colors should be reasonable estimates .
in addition , the ground - based data are very crowded in these regions which often causes the optical luminosity to be over - estimated . within the disk the star formation rate is higher and there are numerous high mass stars , so transients within the disk may have high mass secondaries .
hmxb transients often contain be secondaries and pulsing neutron star primaries @xcite , and are therefore called be transients . only one optical counterpart was found for a disk transient : n1 - 59 in the northern disk .
this source has a very bright optical counterpart candidate , @xmath85 .
if this object is indeed the optical counterpart and the source is in m31 , then m@xmath86 ( or even brighter if we include the likely @xmath01 mag of reddening ) , which makes it at least as bright as an early o star .
be stars typically have spectral types between o8.5b2 @xcite and optical magnitudes of @xmath87 m@xmath88 @xcite .
the counterpart candidate for n1 - 59 is therefore likely a foreground object .
the relative numbers of x - ray binaries containing black hole vs. neutron star primaries within the m31 bulge may be estimated from the relative numbers of transients vs. persistent sources .
we follow the same argument that has been used to estimate this ratio within the galaxy @xcite .
the argument goes as follows . within the bulge ,
the binary population is dominated by low - mass ( @xmath131 @xmath89 ) systems because the secondaries are old . within the galaxy , persistent lmxbs
are observed to often contain neutron star primaries . on the other hand ,
a large fraction of the transient low - mass systems are observed to contain black hole ( bh ) primaries @xcite .
the duty cycle of these transient black hole x - ray novae ( bhxne ) is believed to be @xmath01% @xcite .
thus the ratio of persistent to transient sources , divided by the duty cycle of the transients , yields an estimate of the ratio of number of neutron star containing lmxbs to the number of black hole containing lmxbs . within the galaxy
, it appears that there are approximately equal numbers of neutron star and black hole containing lmxbs .
this result is surprising , as evolutionary calculations predict that neutron star containing systems should be @xmath0100 times more common than those containing black holes @xcite . within the bulge of m31
, there are @xmath0100 persistent lmxbs . in any individual snapshot , there is on average @xmath131 new bright transient within the bulge ( e.g. , figure [ active ] ) .
the majority of these are detected in more than one snapshot , and therefore have decay times consistent with bhxne @xcite .
if the duty cycles of the bhxne in m31 are the same as in the galaxy ( and there is no reason to suspect otherwise ) , then the problem observed in the galaxy is also seen in the m31 bulge : the number of black hole containing lmxbs is approximately equal to the number of neutron star containing lmxbs .
source r2 - 67 was observed several times with the acis detectors . herein , we limit our discussion of the spectra to obsid 1585 and obsid 2897 .
spectral analysis of these observations was performed to estimate the x - ray luminosity of this transient , as well as its optical extinction .
these values were critical to our determination of its orbital period . the other acis detections will be discussed in detail in future publications . given the long decay time of this transient ,
the x - ray luminosity from obsid 1585 is close enough in time to the measurement of the optical luminosity from hst that the luminosities can be used in the correlation of @xcite .
analysis of this observation is complicated because the source was bright and piled - up ( cf .
appendix ) . during obsib 2897
the transient had faded and pileup was negligible . when the transient is faint , spectral fitting is more straightforward , but given the possibility that the spectrum may be time variable , it is desirable to measure it contemporaneously with the hst @xmath2-band measurements . the counting rate during 2001 nov .
19 ( obsid 1585 ) , at 0.18 , was high enough to give a pileup fraction of @xmath030% ( as fit by sherpa and isis ) . in order to apply the pileup model ,
we follow the prescription as described by @xcite . as this method
is relatively new , we include a detailed discussion of our use of it in the appendix .
isis quickly converges to a pileup model with a grade migration parameter @xmath90 , with a fixed psf fraction of 0.95 . using these pileup parameters
, we fit an absorbed disk blackbody model to the spectrum . the best fit absorption value was @xmath91 @xmath25 , corresponding to @xmath92 @xcite and @xmath93 for a standard extinction law with r@xmath94 @xcite .
the best fit disk parameters were @xmath95 kev and @xmath96 , where @xmath97 is the temperature of the inner edge of the accretion disk , @xmath98 is the normalization parameter , @xmath99 is the distance to the source , and @xmath100 is the inclination angle of the disk . assuming a distance of 780 kpc , @xmath98 provides a lower limit to the radius of the inner edge of the accretion disk , @xmath101 km .
these parameters provide a fit with @xmath22 = 1.2 ( probability=0.2 ) . the observed flux is 1.3 @xmath8 10@xmath102 2s ( 0.3 - 7.0 kev ) , and the modeled emitted luminosity is @xmath103 ( 0.3 - 7.0 kev ) .
the @xmath104 km corresponds to @xmath105 for a 10 @xmath89 black hole , and @xmath106 for the best fit values assuming cos(@xmath100 ) = 1 .
these are reasonable values for an accretion disk around a @xmath010 solar mass black hole during this bright stage of the outburst . by 2002 jan .
8 , the flux had decayed sufficiently that pileup was no longer a problem .
we extracted 177 counts from a @xmath107 radius , and found that a simple power - law with ism absorption models fit the data well ( @xmath108 ) .
the observed 0.37.0 kev flux was @xmath109 2s , and the modeled emitted 0.37.0 kev luminosity at 780 kpc was @xmath110 .
the best fit value of @xmath111 cm@xmath10 is consistent with that due to the galaxy alone , and corresponds to @xmath112 @xcite and @xmath113 for a standard extinction law with r@xmath94 @xcite .
this measurement is consistent with the measurement from obsid 1585 of @xmath114 cm@xmath10 .
when the two measurements are combined , the best estimate of the absorption to r2 - 67 is @xmath115 cm@xmath10 , which corresponds to @xmath116 .
we note that the best fit slope is substantially harder than that found when the source was bright ( see appendix for a power - law fit to the bright state data ) .
the harder slope is consistent with the source having entered the ` low - hard ' or ` intermediate ' state from the ` high - soft ' state in obsid 1585 @xcite .
as these transitions are believed to occur at luminosities between 10% and a few % of eddington , this transition is consistent with the source having a mass of @xmath010 .
two hst observations taken @xmath01 month apart , both when the x - ray source was bright , reveal a @xmath2-band source with f336w = [email protected] mag .
a third observation several months later , when the x - ray source was faint , did not detect the counterpart and set a limit of f336w @xmath75 22.8 mag .
the disappearance of the optical counterpart in concert with the x - ray decline confirms this identification of the optical counterpart of the x - ray transient .
the observations are shown in figures [ hstims ] and [ 143148 ] and summarized in table [ r267obs ] .
fortuitously , most of the hst images contained an x - ray bright gc , which we used to register with our _ chandra_-hrc mosaic to the hst images .
x - ray position errors were estimated by dividing the full - width at half maximum ( fwhm ) of the x - ray source , as measured with the iraf task _ imexamine _
, by the square - root of the number of counts in the x - ray detection .
the final position errors were dominated by the x - ray position errors ; the optical position errors were always @xmath290.1@xmath18 .
the first 2 hst observations of r2 - 67 were aligned with our x - ray image using the gc bol 148 .
the final hst observation of r2 - 67 was aligned using gc bol 144 .
our _ chandra _ positions of the gc sources had errors of 0.15@xmath18 and 0.08@xmath18 , respectively , and the x - ray position of the r2 - 67 had an error of 0.2@xmath18 .
therefore , in figure [ hstims ] , the 0.8@xmath18 error circles marked on the first two hst images of r2 - 67 , and the 0.6@xmath18 error circle marked on the third hst image of r2 - 67 , show the 3@xmath14 position errors of our data .
interestingly , a candidate was also found in the lgs data for r2 - 67 .
this candidate has @[email protected] and @[email protected] .
the counterpart for r2 - 67 seen from the ground was particularly unexpected because the @xmath64-band images were taken october 6 , 2000 , when the x - ray source was quiescent , and the @xmath6-band images were taken september 18 , 2001 , a few months before we detected the transient but during a 3 month gap in our x - ray monitoring
. the lgs object may be the counterpart or a chance superposition of a different star along the line of sight , while the @xmath2-band transient seen by hst is clearly the optical glow of the x - ray nova .
the magnitude of the lgs star from the ground is likely a lower limit as crowding often causes the brightness to be overestimated .
the apparently persistent nature of the lgs candidate ( i.e. it was detected when we believe the x - ray source was dim ) and the clear transient nature of the hst counterpart calls into question the validity of the lgs candidate .
however , the lgs candidate is well - detected in the @xmath6-band , with @xmath20 of 9 , but the detection is less robust in the @xmath64-band , with @xmath20 of 4 .
the x - ray source is not a pulsar @xcite and is therefore unlikely to have a high - mass ( be ) companion which would be persistent optically . while the lgs and on - state hst magnitudes are approximately equal to those expected for a slightly evolved @xmath0b3 star in m31 with @xmath117 , such an object would have been detected in all 3 hst observations .
the lgs photometry allows the possibility that this optical detection is an interloping foreground main - sequence star with f336w @xmath3422.8 mag and @xmath118 .
this possibility offers the simplest explanation for the candidate s non - detection in the third hst image .
the detection of this counterpart candidate in the lgs data underscores the confusion that crowding causes in ground - based images of the m31 bulge .
it is possible to estimate the orbital period for the transient using the period dependence within the correlation between x - ray and optical luminosity found by @xcite . in order to do so ,
estimates of the x - ray luminosity and the @xmath6-band absolute magnitude are needed .
while the former were directly available from the hrc and acis observations , the latter were estimated from the hst ( f336w ) measurements and corrected for the interstellar absorption to the transient , as estimated from the acis data ( cf .
6.1 ) and converted to @xmath120 using the correlation between between @xmath121 and @xmath120 ( i.e. , @xcite ) . plugging the contemporaneous optical and x - ray luminosities into the @xcite relation we can estimate the orbital period of r2 - 67 .
the @xmath122 indicated by combining our x - ray measurements of the absorption toward r2 - 67 ( cf . 6.3.3 ) is @xmath123 .
this @xmath122 implies an absolute m@xmath124 .
we assume an intrinsic @xmath125 , typical for observations for lmxbs in the galaxy @xcite .
these assumptions supply an estimate for m@xmath126 . along with the observed @xmath127
, this m@xmath65 implies an orbital period @xmath128 days .
this is not unreasonable compared to bhxne within the galaxy .
for example v404 cyg has @xmath129 days @xcite , and grs1915 + 105 has @xmath130 days @xcite . a second way to estimate
the orbital period is from the decay time of the outburst .
@xcite developed a model for the outburst of bhxne disks , assuming that irradiation determines the disk temperature profile during the outburst and that the outburst can not end until the irradiation allows the outer edge of the disk to cool below the hydrogen recombination temperature .
this model predicts longer decay times for larger disks and approximately linear decay curves for systems with orbital periods longer than @xmath01 day .
shorter orbital period systems are predicted to produce exponential decay curves . while the data herein do not constrain the shape of the lightcurve ( cf .
[ 143148 ] ) , they do provide a decay time estimate of @xmath00.2 yr .
equation 23 of @xcite describes the time variable mass transfer rate for such a large disk in outburst . assuming @xmath131 , where @xmath132 is the central accretion rate and @xmath133 is the accretion efficiency ( @xmath00.1 ) allows us to re - write this as @xmath134 where @xmath135 is the disk - averaged kinetic viscosity , the constant @xmath136 ( cgs units ) , @xmath137 is the mass of the hot zone and @xmath138 is the time in seconds after the start of the outburst decay . applying our observation that @xmath127 when @xmath139 to equation ( 1 ) allows us to write @xmath140 using the value in equation ( 2 ) as the first term in equation ( 1 ) , applying our observation of the decay time of @xmath141 years , and assuming that the luminosity is @xmath00 at this point ( justified since it is @xmath142 at @xmath143 years ) lets us write @xmath144 equation ( 3 ) lets us determine the viscosity , @xmath145 cm@xmath10 s , a value characteristic of the large disk implied by the @xmath146 day period .
the @xcite alpha - prescription states that @xmath147 where @xmath61 is the angular momentum transport efficiency and @xmath148 is the disk thickness . in the thin disk approximation , @xmath149 ( see @xcite , p. 74 ) . assuming @xmath150 , the outer disk radius , where the disk cools below the ionization temperature , has the value of @xmath151 where @xmath152 is the bh mass in 10 @xmath89 units , and @xmath153 is the local disk temperature in units of @xmath154 k. this is consistent with the binary separation , @xmath155 calculated from the @xmath146 day period with kepler s law .
the similarity of these values suggests that , in addition to the @xmath156 ratio , the outburst decay time is also consistent with that of a bhxn with a @xmath146 day orbital period .
another highly variable source , r3 - 16 , whose lightcurve is also shown in figure [ 143148 ] , has a bright uv counterpart of unknown nature .
hst images of the optical counterpart during 3 epochs are provided in figure [ hstims ] .
the second and third hst observations of source r3 - 16 were aligned with our x - ray mosaic using the gc bol 148 .
the error on the x - ray position of r3 - 16 was 0.4@xmath18 , and that of bol 148 was 0.15@xmath18 .
the 3@xmath14 error circle for this source was therefore 1.3@xmath18 , and it is shown in figure [ hstims ] .
the first hst observation of r3 - 16 did not contain any x - ray bright gc sources . for this observation
, we had to rely on the original coordinate system assigned by the hst pipeline for our alignment .
we added 1@xmath18 to the radius of our error circle in this case to account for uncertainty in the registration between the _ chandra _ and hst coordinates .
therefore , the error circle in the first image has a 2.3@xmath18 radius .
the f336w magnitudes from hst observations and @xmath157 magnitudes from ground - based observations are provided in table [ r316obs ] .
the first 2 hst observations were taken near the peak of the x - ray outburst , and the third during the decline .
the hst f336w magnitudes are @xmath021 and appear variable , while the lgs data yield @[email protected] and @[email protected] . discerning the nature of object r3 - 16 is difficult without any high - quality optical spectral information .
the object was classified as an emission line object of unknown nature by @xcite . however , the x - ray spectrum and optical size of the object provide some new hints about its nature .
we extracted the spectrum from a january 8 , 2002 acis observation ( obsid 2897 ) and found that it is well fit with an absorbed power - law with @xmath121 = [email protected] @xmath8 10@xmath158 @xmath25 and slope @xmath159 ( @xmath160 , probability = 74% ) .
fits of the radial profile of r3 - 16 s optical counterpart candidate in the hst images to the hst psf ( described in detail in @xcite ) yield @xmath161 for all observations .
this statistic is normalized to have a median value of 1 for single stars . in addition ,
sharpness measurements for the object all 3 epochs yield values @xmath162 .
tests of the hstphot software suggest normal stars have @xmath163 and sharpness between -0.3 and 0.3 @xcite .
these measurements suggest that the object is extended ; such @xmath164 and sharpness values are typical of unresolved binaries . on the pc chip of wfpc2 , where we measure 7 point sources to have a fwhm of [email protected]@xmath18 , r3 - 16
has a fwhm of [email protected]@xmath18 , implying an intrinsic size of [email protected]@xmath18 or [email protected] parsec if located in m31 .
an independent measurement on the wf3 chip , where 10 point sources have a fwhm of [email protected]@xmath18 , has a fwhm of [email protected]@xmath18 , implying an intrinsic size of [email protected]@xmath18 , about 50% larger than the result from the pc .
the width of the the same object in the lgs ground based @xmath2-band images , where we measure 10 nearby point sources to have a fwhm of [email protected]@xmath18 , is [email protected]@xmath18 , implying an intrinsic width of the object of [email protected]@xmath18 or [email protected] parsec if located in m31 .
the discrepancies between the implied sizes of the object in ground - based and hst images are difficult to reconcile .
the larger size , like the brighter magnitude , measured in the ground - based images may be due to crowding in the m31 bulge . in any case , the object is not likely to be a single star , though it may be a blended foreground binary with a separation of @xmath130.01 pc .
the x - ray spectrum , optical size , brightness , variability and emission line properties of r3 - 16 allow the possibility that it is a background active galactic nucleus ( agn ) .
such agn have been seen through several other local group galaxies at optical magnitudes similar to those measured for r3 - 16 ( e.g. , @xcite ) .
this possibility is also consistent with the measured @xmath121 which is higher than the galactic value to m31 of @xmath165 @xmath25 , as the light of a background agn would be heavily absorbed by m31 . the increase in apparent angular size with increasing pixel size
would then be a result of integrating more diffuse emission in the larger pixels .
alternatively , r3 - 16 could be a cv at a distance of @xmath131 kpc .
this possibility is consistent with the optical brightness and colors .
the x - ray luminosity would then be @xmath1310@xmath166 during the outburst , and @xmath1310@xmath167 during quiescence .
this range is typical of that seen in cvs ( @xcite , @xcite ) .
the @xmath00.3 mag drop in the @xmath2-band along with the drop in the x - ray flux does not argue for nor against the cv hypothesis , as the expected relation between this relatively small ( for a cv outburst ) @xmath2-band change and the x - ray flux is unclear . the acis spectrum measured during outburst is harder than that typically seen in cvs in outburst @xcite , but there are some cvs which show hard spectra during outburst @xcite .
the acis spectrum also shows absorption above the value expected from the galaxy alone , which may be somewhat unusual but not unheard of for cvs .
the 1 kpc distance should be taken as an upper limit because , at the galactic latitude of m31 ( @xmath168 ) , an object at that distance would be 360 pc below the galactic plane , which is at the extreme range for cvs ( @xcite , section 9.5 ) .
if r3 - 16 is a cv , the extended , persistent optical emission could be a nova shell from an earlier ( unseen ) nova eruption of this cv .
similar to the images of r3 - 16 , many nova remnants are only slightly extended in ground - based and hst images @xcite .
a high quality optical spectrum could uncover the true nature of this source .
we have combined 17 epochs of snapshot observations covering most of the m31 disk with the _ chandra_-hrc .
these data have provided detections of 166 discrete x - ray sources .
all but 7 of these have been previously detected .
comparison of the lf of the bulge sources to that of the disk sources reveals significant differences in shape .
the slope of the disk lf is comparable to that of elliptical galaxies .
this similarity is consistent with the link between star formation rate and lf slope , as the star formation rate in the m31 disk is rather low .
analysis of the spatial distribution of sources shows that most of the brightest sources in the disk lie in the southwestern half of the disk .
we have found candidate counterparts for 55 of the 166 sources at longer wavelengths in previous surveys .
these counterparts come in a variety of types , including snrs , globular clusters , stars and an extended optical source of unknown nature .
fifteen stellar counterpart candidates were detected in recent wide - field m31 data taken by the local group survey @xcite .
counterparts were detected for 2 x - ray transients using data from hst .
analysis of one counterpart ( r2 - 67 ) found it to be an optical transient .
the properties of this system are consistent with a bhxn in m31 .
the ratio of the optical to x - ray flux yields an estimate of the period of this system of @xmath146 days .
this period is consistent with the x - ray decline rate .
the optical properties of the other transient ( r3 - 16 ) are difficult to understand , but it may be a foreground blend , a cv with an associated nova remnant , or a background agn .
the long - term lightcurves of these sources suggest that at least 44 of them varied significantly over the course of these observations , which cover a baseline of about 2.5 years . from the lightcurves , we have selected 17 good transient candidates , and we have determined that at any given time there are 2@xmath51 active x - ray transients in m31 .
the frequency of occurrence for these bright transient events suggests that @xmath131% of the bright x - ray sources in the m31 bulge are new transients .
if these sources are black hole containing lmxbs with duty cycles of @xmath01% , the ratio of black hole to neutron star primaries in lmxbs in m31 is @xmath01 , comparable to the ratio seen in the galaxy but greater than expectations from evolutionary calculations . finally , it is unfortunate that we have been unsuccessful at finding counterparts for more than 100 sources .
many of these counterparts are located in the extremely crowded bulge and will require very high angular resolution optical data to recover .
these objects are likely to be faint in the optical ( v@xmath3420 ) , and they may be heavily absorbed . currently undetected counterparts in the disk are likely even fainter or heavily absorbed .
spectral x - ray data of the disk , unavailable with the hrc data , will help to discern whether intrinsic faintness or absorption are the cause for the lack of detection of counterparts at longer wavelengths .
we thank phil kaaret for allowing us the use of his 50 ks observation of the m31 nucleus to help constrain our lightcurves in that region and for helping to measure the model fits .
we thank the lgs team @xcite for supplying the optical data to search for new optical counterparts for the x - ray sources .
support for this work was provided by nasa through grant number go-9087 from the space telescope science institute and through grant number go-3103x from the _ chandra _ x - ray center .
mrg acknowledges support from nasa ltsa grant nag5 - 10889 .
finally we thank our referee , phil charles , for his many helpful and insightful comments which greatly helped to improve this paper .
, p. r. , & robinson , d. k. 2003 , data reduction and error analysis for the physical sciences ( data reduction and error analysis for the physical sciences , 3rd ed .
, by philip r. bevington , and keith d. robinson .
boston , ma : mcgraw - hill , isbn 0071199268 , 2003 . ) , m. r. , kong , a. , primini , f. a. , barmby , p. , di stefano , r. , mcclintock , j. e. , & murray , s. s. 2001b , in two years of science with chandra , abstracts from the symposium held in washington , dc , 5 - 7
september , 2001 .
, l. o. , & dickel , j. r. 2001 , in young supernova remnants : eleventh astrophysics conference .
aip conference proceedings , vol .
college park , maryland , 16 - 18 oct , 2000 .
edited by stephen s. holt , and una hwang .
american institute of physics , 2001 , p.433 - 436 , 433 , y. , & lewin , w. h. g. 1995 , in x - ray binaries ( cambridge astrophysics series , cambridge , ma : cambridge university press , |c1995 , edited by lewin , walter h.g . ; van paradijs , jan ; van den heuvel , edward p.j . ) , f. , & van den heuvel , e. p. j. 1995 , in x - ray binaries ( cambridge astrophysics series , cambridge , ma : cambridge university press , |c1995 , edited by lewin , walter h.g . ; van paradijs , jan ; van den heuvel , edward p.j . )
cccccc 243 & 1999 - 11 - 30 & 1 & 0:40:27.00 & 40:40:12.0 & 1163.613 + 255 & 1999 - 11 - 30 & 1 & 0:42:08.00 & 40:55:17.0 & 1269.114 + 267 & 1999 - 11 - 30 & 1 & 0:42:44.40 & 41:16:08.3 & 1270.082 + 279 & 1999 - 11 - 30 & 1 & 0:44:07.00 & 41:43:16.0 & 2683.351 + 291 & 1999 - 11 - 30 & 1 & 0:45:20.00 & 41:49:47.0 & 1270.183 + 268 & 1999 - 12 - 23 & 2 & 0:42:44.40 & 41:16:08.3 & 5178.063 + 245 & 2000 - 01 - 19 & 3 & 0:40:27.00 & 40:40:12.0 & 1159.204 + 257 & 2000 - 01 - 19 & 3 & 0:42:08.00 & 40:55:17.0 & 1207.168 + 269 & 2000 - 01 - 19 & 3 & 0:42:44.40 & 41:16:08.3 & 1205.910 + 281 & 2000 - 01 - 19 & 3 & 0:44:07.00 & 41:43:16.0 & 1205.035 + 293 & 2000 - 01 - 19 & 3 & 0:45:20.00 & 41:49:47.0 & 1208.001 + 246 & 2000 - 02 - 13 & 4 & 0:40:27.00 & 40:40:12.0 & 1417.755 + 258 & 2000 - 02 - 13 & 4 & 0:42:08.00 & 40:55:17.0 & 1474.374 + 270 & 2000 - 02 - 13 & 4 & 0:42:44.40 & 41:16:08.3 & 1469.659 + 282 & 2000 - 02 - 13 & 4 & 0:44:07.00 & 41:43:16.0 & 1473.856 + 294 & 2000 - 02 - 13 & 4 & 0:45:20.00 & 41:49:47.0 & 1467.671 + 247 & 2000 - 03 - 08 & 5 & 0:40:27.00 & 40:40:12.0 & 2162.940 + 259 & 2000 - 03 - 08 & 5 & 0:42:08.00 & 40:55:17.0 & 2472.312 + 271 & 2000 - 03 - 08 & 5 & 0:42:44.40 & 41:16:08.3 & 2461.785 + 283 & 2000 - 03 - 08 & 5 & 0:44:07.00 & 41:43:16.0 & 2476.503 + 295 & 2000 - 03 - 08 & 5 & 0:45:20.00 & 41:49:47.0 & 2455.864 + 248 & 2000 - 05 - 26 & 6 & 0:40:27.00 & 40:40:12.0 & 1156.204 + 260 & 2000 - 05 - 26 & 6 & 0:42:40.80 & 40:51:54.0 & 1209.234 + 272 & 2000 - 05 - 26 & 6 & 0:42:44.40 & 41:16:08.3 & 1202.670 + 284 & 2000 - 05 - 26 & 6 & 0:44:07.00 & 41:43:16.0 & 1211.161 + 296 & 2000 - 05 - 26 & 6 & 0:45:20.00 & 41:49:47.0 & 1204.171 + 249 & 2000 - 06 - 21 & 7 & 0:40:27.00 & 40:40:12.0 & 1189.396 + 261 & 2000 - 06 - 21 & 7 & 0:42:40.80 & 40:51:54.0 & 1186.803 + 273 & 2000 - 06 - 21 & 7 & 0:42:44.40 & 41:16:08.3 & 1190.727 + 285 & 2000 - 06 - 21 & 7 & 0:44:07.00 & 41:43:16.0 & 1186.797 + 297 & 2000 - 06 - 21 & 7 & 0:45:20.00 & 41:49:47.0 & 893.869 + 251 & 2000 - 08 - 18 & 8 & 0:40:27.00 & 40:40:12.0 & 977.143 + 263 & 2000 - 08 - 18 & 8 & 0:42:40.80 & 40:51:54.0 & 1184.931 + 275 & 2000 - 08 - 18 & 8 & 0:42:44.40 & 41:16:08.3 & 1188.885 + 287 & 2000 - 08 - 18 & 8 & 0:44:07.00 & 41:43:16.0 & 1186.985 + 299 & 2000 - 08 - 18 & 8 & 0:45:20.00 & 41:49:47.0 & 1170.571 + 252 & 2000 - 09 - 11 & 9 & 0:40:27.00 & 40:40:12.0 & 1189.792 + 264 & 2000 - 09 - 11 & 9 & 0:42:40.80 & 40:51:54.0 & 1189.081 + 276 & 2000 - 09 - 11 & 9 & 0:42:44.40 & 41:16:08.3 & 1182.794 + 288 & 2000 - 09 - 11 & 9 & 0:44:07.00 & 41:43:16.0 & 1193.216 + 300 & 2000 - 09 - 11 & 9 & 0:45:20.00 & 41:49:47.0 & 1150.291 + 253 & 2000 - 10 - 12 & 10 & 0:40:27.00 & 40:40:12.0 & 871.491 + 265 & 2000 - 10 - 12 & 10 & 0:42:40.80 & 40:51:54.0 & 1189.515 + 277 & 2000 - 10 - 12 & 10 & 0:42:44.40 & 41:16:08.3 & 1187.209 + 289 & 2000 - 10 - 12 & 10 & 0:44:07.00 & 41:43:16.0 & 1190.742 + 301 & 2000 - 10 - 12 & 10 & 0:45:20.00 & 41:49:47.0 & 1162.660 + 254 & 2000 - 11 - 17 & 11 & 0:40:27.00 & 40:40:12.0 & 1182.351 + 266 & 2000 - 11 - 17 & 11 & 0:42:40.80 & 40:51:54.0 & 1180.444 + 278 & 2000 - 11 - 17 & 11 & 0:42:44.40 & 41:16:08.3 & 1182.300 + 290 & 2000 - 11 - 17 & 11 & 0:44:07.00 & 41:43:16.0 & 1176.379 + 302 & 2000 - 11 - 17 & 11 & 0:45:20.00 & 41:49:47.0 & 979.562 + 1565 & 2001 - 01 - 31 & 12 & 0:40:27.00 & 40:40:12.0 & 964.220 + 1567 & 2001 - 02 - 01 & 12 & 0:42:08.00 & 40:55:17.0 & 1172.113 + 1569 & 2001 - 02 - 01 & 12 & 0:42:44.40 & 41:16:08.3 & 1171.771 + 1571 & 2001 - 02 - 01 & 12 & 0:44:07.00 & 41:43:16.0 & 1187.686 + 1573 & 2001 - 02 - 01 & 12 & 0:45:20.00 & 41:49:47.0 & 1186.826 + 1566 & 2001 - 06 - 10 & 13 & 0:40:27.00 & 40:40:12.0 & 871.495 + 1568 & 2001 - 06 - 10 & 13 & 0:42:08.00 & 40:55:17.0 & 1183.653 + 1570 & 2001 - 06 - 10 & 13 & 0:42:44.40 & 41:16:08.3 & 1184.236 + 1572 & 2001 - 06 - 10 & 13 & 0:44:07.00 & 41:43:16.0 & 1184.442 + 1574 & 2001 - 06 - 10 & 13 & 0:45:20.00 & 41:49:47.0 & 1194.797 + 2886 & 2001 - 09 - 08 & 14 & 0:40:27.10 & 40:40:12.0 & 867.664 + 2890 & 2001 - 09 - 08 & 14 & 0:42:07.90 & 40:55:15.6 & 1182.953 + 2903 & 2001 - 09 - 08 & 14 & 0:42:44.40 & 41:43:15.6 & 1188.012 + 2907 & 2001 - 09 - 08 & 14 & 0:44:07.00 & 41:43:15.6 & 1179.696 + 2911 & 2001 - 09 - 08 & 14 & 0:45:19.90 & 41:49:48.0 & 1178.814 + 2887 & 2001 - 11 - 19 & 15 & 0:40:27.10 & 40:40:12.0 & 982.551 + 2891 & 2001 - 11 - 19 & 15 & 0:42:07.90 & 40:55:15.6 & 1173.808 + 2904 & 2001 - 11 - 19 & 15 & 0:42:44.40 & 41:16:08.3 & 1181.646 + 2908 & 2001 - 11 - 19 & 15 & 0:44:07.00 & 41:43:15.6 & 1172.898 + 2912 & 2001 - 11 - 19 & 15 & 0:45:19.90 & 41:49:48.0 & 1161.823 + 2888 & 2002 - 01 - 16 & 16 & 0:40:27.10 & 40:40:12.0 & 975.687 + 2892 & 2002 - 01 - 16 & 16 & 0:42:07.90 & 40:55:15.6 & 1083.539 + 2905 & 2002 - 01 - 16 & 16 & 0:42:44.40 & 41:16:08.3 & 1094.222 + 2909 & 2002 - 01 - 16 & 16 & 0:44:07.00 & 41:43:15.6 & 1084.598 + 2913 & 2002 - 01 - 16 & 16 & 0:44:28.40 & 41:56:28.1 & 1097.045 + 2889 & 2002 - 06 - 02 & 17 & 0:40:27.10 & 40:40:12.0 & 980.472 + 2893 & 2002 - 06 - 02 & 17 & 0:42:07.90 & 40:55:15.6 & 1190.200 + 2906 & 2002 - 06 - 02 & 17 & 0:42:44.40 & 41:16:08.3 & 1187.762 + 2910 & 2002 - 06 - 02 & 17 & 0:44:07.00 & 41:43:15.6 & 1196.402 + 2914 & 2002 - 06 - 02 & 17 & 0:45:19.90 & 41:49:48.0 & 1190.132 + [ obslist ] ccccccccr s1 - 74 & 0:39:56.34 & 40:41:00.9 & 51 & 7.0 & [email protected] & 0.242 & foreground ( [email protected] ) & new ( * ) + s1 - 75 & 0:40:13.77 & 40:50:05.1 & 2215 & 118.0 & [email protected] & 1.968 & b3 0037 + 405 & 23 ( bll ) + s1 - 76 & 0:40:14.34 & 40:33:41.5 & 65 & 9.3 & [email protected] & 0.263 & new & new ( x ) + s1 - 77 & 0:40:20.29 & 40:43:58.5 & 1436 & 169.9 & [email protected] & 1.086 & bol 5 & 6 ( gc ) + s1 - 78 & 0:40:22.71 & 40:36:10.5 & 42 & 6.9 & [email protected] & 0.253 & lgsj004022.7 + 403610 & new ( * ) + s2 - 63 & 0:40:24.20 & 40:29:48.5 & 32 & 5.2 & [email protected] & 0.356 & rx j004024.5 + 402946 & 1 ( x ) + s1 - 79 & 0:40:55.16 & 40:56:03.5 & 850 & 27.4 & [email protected] & 18.805 & new & new ( x ) + s1 - 45 & 0:41:18.55 & 40:52:00.5 & 136 & 8.2 & [email protected] & 1.251 & foreground ( [email protected] ) & qss ( * ) + s1 - 31 & 0:41:25.78 & 40:58:47.5 & 171 & 11.7 &
[email protected] & 1.024 & rx j004125.9 + 405842 & 1 ( x ) + s1 - 34 & 0:41:25.93 & 40:53:25.5 & 59 & 5.0 & [email protected] & 0.437 & lgsj004126.2 + 405327 & new ( * ) + s1 - 28 & 0:41:37.60 & 41:01:09.8 & 39 & 3.8 & [email protected] & 0.285 & rx j004138.3 + 410106 & 1 ( x ) + s1 - 16 & 0:41:49.81 & 41:01:05.8 & 20 & 3.7 & [email protected] & 0.146 & rx j004149.8 + 410109 & 1 ( x ) + s1 - 80 & 0:41:54.63 & 40:56:47.4 & 150 & 24.0 & [email protected] & 7.158 & 1rxs j004154.1 + 405648 ( atel97 ) & 16 ( x ) + s1 - 64 & 0:42:02.92 & 40:46:11.4 & 137 & 9.3 & [email protected] & 0.695 & lgsj004203.0 + 404613 & new ( * ) + r3 - 125 & 0:42:05.71 & 41:13:29.7 & 65 & 5.3 & [email protected] & 0.620 & [ pfj93 ] 3 & 2 ( x ) + s1 - 12 & 0:42:06.12 & 41:02:47.5 & 78 & 5.6 & [email protected] & 0.469 & bol d42 & 6 ( gc ) + s1 - 11 & 0:42:07.16 & 41:00:17.9 & 50 & 5.8 & [email protected] & 0.273 & bol d44 & 6 ( gc ) + r3 - 61 & 0:42:07.74 & 41:18:15.4 & 338 & 34.2 & [email protected] & 0.953 & c84 10 & 3 ( x ) + s1 - 9 & 0:42:07.89 & 41:04:34.6 & 82 & 5.3 & [email protected] & 0.408 & rx j004208.2 + 410438 & 1 ( x ) + r3 - 60 & 0:42:09.11 & 41:20:49.0 & 71 & 6.8 & [email protected] & 0.547 & cxom31 j004209.0 + 412048 & 15 ( x ) + r3 - 59 & 0:42:09.53 & 41:17:45.3 & 80 & 9.6 & [email protected] & 0.350 & mita140 & 14 ( gc ) + r3 - 58 & 0:42:10.32 & 41:15:10.1 & 52 & 7.3 & [email protected] & 0.795 & [ pfj93 ] 6 & 2 ( x ) + r3 - 57 & 0:42:11.01 & 41:12:47.7 & 34 & 4.9 & [email protected] & 0.139 & cxom31 j004210.9 + 411248 & 15 ( x ) + r3 - 56 & 0:42:11.79 & 41:10:48.6 & 61 & 5.7 & [email protected] & 0.285 & [ pfj93 ] 7 & 2 ( x ) + r3 - 55 & 0:42:11.92 & 41:16:48.9 & 32 & 5.5 & [email protected] & 0.177 & cxom31j004211.9 + 411648 & 15 ( x ) + s1 - 58 & 0:42:12.12 & 40:53:35.6 & 25 & 4.7 & [email protected] & 0.212 & mla 462 & 17 ( x ) + r3 - 54 & 0:42:12.15 & 41:17:59.1 & 140 & 17.8 & [email protected] & 1.704 & mita153 & 14 ( gc ) + r3 - 52 & 0:42:13.14 & 41:18:36.7 & 647 & 60.4 & [email protected] & 4.630 & c84 13 & 3 ( x ) + s1 - 81 & 0:42:13.77 & 40:39:25.0 & 83 & 4.7 & [email protected] & 0.665 & rx j004213.5 + 403917 & 1 ( x ) + r3 - 50 & 0:42:15.13 & 41:12:34.6 & 342 & 37.7 & [email protected] & 3.871 & [ pfj93 ] 10 & 2 ( x ) + r3 - 48 & 0:42:15.37 & 41:20:31.7 & 77 & 9.9 & [email protected] & 0.206 & [ pfj93 ] 11 & 2 ( x ) + r3 - 47 & 0:42:15.65 & 41:17:21.0 & 149 & 19.5 & [email protected] & 0.988 & c84 16 & 3 ( x ) + s1 - 7 & 0:42:15.86 & 41:01:14.7 & 1214 & 66.6 & [email protected] & 2.675 & mita159 & 14 ( gc ) + s1 - 6 & 0:42:16.55 & 40:55:52.3 & 199 & 25.8 & [email protected] & 0.442 & rx j004216.4 + 405555 & 1 ( x ) + r3 - 46 & 0:42:17.06 & 41:15:08.8 & 43 & 9.0 & [email protected] & 5.710 & cxom31 j004216.9 + 411508 & 15 ( x ) + r3 - 45 & 0:42:18.32 & 41:12:24.1 & 643 & 65.8 & [email protected] & 1.398 & c84 21 & 3 ( x ) + r3 - 44 & 0:42:18.62 & 41:14:01.9 & 574 & 77.8 & [email protected] & 1.307 & mita164 & 14 ( gc ) + s1 - 82 & 0:42:20.82 & 40:51:36.3 & 27 & 5.0 & [email protected] & 0.552 & new & new ( x ) + r3 - 42 & 0:42:21.48 & 41:16:01.6 & 406 & 66.6 & [email protected] & 1.350 & c84 23 & 3 ( x ) + r3 - 41 & 0:42:21.54 & 41:14:18.5 & 46 & 8.4 & [email protected] & 0.557 & cxom31 j004221.5 + 411419 & 15 ( x ) + s1 - 5 & 0:42:22.20 & 40:59:25.6 & 234 & 24.2 & [email protected] & 0.448 & rx j004222.1 + 405926 & 1 ( x ) + r3 - 40 & 0:42:22.41 & 41:13:34.2 & 293 & 44.4 & [email protected] & 1.429 & c84 25 & 3 ( x ) + r3 - 39 & 0:42:22.94 & 41:15:35.4 & 884 & 125.9 & [email protected] & 44.079 & c84 26 & 3 ( x ) + r3 - 38 & 0:42:23.15 & 41:14:07.4 & 141 & 26.8 & [email protected] & 1.193 & c84 27 & 3 ( x ) + r2 - 45 & 0:42:25.11 & 41:13:40.7 & 109 & 19.7 & [email protected] & 1.287 & c84 28 & 3 ( x ) + r3 - 87 & 0:42:25.72 & 41:25:50.3 & 81 & 9.7 & [email protected] & 0.155 & [ pfj93 ] 24 & 2 ( x ) + r2 - 36 & 0:42:26.03 & 41:19:15.3 & 155 & 26.7 & [email protected] & 0.962 & mita174 & 14 ( gc ) + r3 - 36 & 0:42:28.20 & 41:10:00.7 & 310 & 34.2 & [email protected] & 0.638 & cxom31 j004228.1 + 410959 & 15 ( x ) + r2 - 35 & 0:42:28.28 & 41:12:23.2 & 816 & 113.7 & [email protected] & 1.878 & c84 30 & 3 ( x ) + r3 - 111 & 0:42:28.92 & 41:04:35.8 & 608 & 23.3 & [email protected] & 2.483 & c84 31 & 3 ( x ) + r2 - 34 & 0:42:31.13 & 41:16:21.9 & 734 & 133.3 & [email protected] & 0.526 & c84 36 & 3 ( x ) + r2 - 33 & 0:42:31.24 & 41:19:39.2 & 156 & 30.8 & [email protected] & 2.727 & mita192 & 14 ( gc ) + r2 - 32 & 0:42:32.07 & 41:13:14.5 & 411 & 71.9 & [email protected] & 3.280 & c84 39 & 3 ( x ) + r2 - 68 & 0:42:32.63 & 41:17:02.9 & 19 & 4.6 & [email protected] & 0.149 & new & new ( x ) + r2 - 31 & 0:42:32.74 & 41:13:11.0 & 65 & 13.6 & [email protected] & 0.801 & rx j004232.7 + 411318 & 1 ( x ) + r2 - 30 & 0:42:33.87 & 41:16:20.2 & 106 & 21.4 & [email protected] & 0.264 & [ pfj93 ] 32 & 2 ( x ) + r2 - 29 & 0:42:34.43 & 41:18:09.6 & 40 & 9.5 & [email protected] & 1.149 & cxom31 j004234.4 + 411809 & 15 ( x ) + n1 - 74 & 0:42:34.53 & 41:32:47.7 & 159 & 5.5 & [email protected] & 1.207 & bol 116 & 6 ( gc ) + r2 - 28 & 0:42:34.75 & 41:15:23.5 & 69 & 14.6 & [email protected] & 2.119 & cxom31 j004234.7 + 411523 & 15 ( x ) + s1 - 1 & 0:42:34.97 & 40:57:21.5 & 79 & 10.5 & [email protected] & 1.399 & new & new ( x ) + s1 - 83 & 0:42:35.03 & 40:48:38.0 & 160 & 14.6 & [email protected] & 0.866 & bol d63 & 6 ( gc ) + r2 - 27 & 0:42:35.18 & 41:20:05.8 & 114 & 23.9 & [email protected] & 0.567 & [ pfj93 ] 34 & 2 ( x ) + r2 - 42 & 0:42:36.64 & 41:13:50.4 & 31 & 7.1 & [email protected] & 0.154 & qss & qss ( x ) + r2 - 26 & 0:42:38.57 & 41:16:03.9 & 5579 & 589.9 & [email protected] & 7.392 & [ pfj93 ] 35 & 2 ( x ) + r2 - 25 & 0:42:39.52 & 41:14:28.7 & 97 & 22.1 & [email protected] & 0.999 & tf 42 & 3 ( x ) + r1 - 34 & 0:42:39.58 & 41:16:14.6 & 26 & 5.9 & [email protected] & 42.880 & tf 43 & 3 ( x ) + s1 - 84 & 0:42:39.70 & 40:43:18.7 & 114 & 5.8 & [email protected] & 0.359 & rx j004238.8 + 404317 & 1 ( x ) + r1 - 15 & 0:42:39.98 & 41:15:47.7 & 424 & 66.7 & [email protected] & 2.101 & ford m 31 17 & 5 ( pn ) + r2 - 24 & 0:42:40.20 & 41:18:45.6 & 74 & 17.5 & [email protected] & 0.431 & cxom31 j004240.1 + 411845 & 15 ( x ) + r2 - 22 & 0:42:40.64 & 41:13:27.5 & 173 & 33.7 & [email protected] & 2.230 & [ pfj93 ] 38 & 2 ( x ) + r3 - 34 & 0:42:40.65 & 41:10:32.8 & 26 & 4.5 & [email protected] & 0.296 & mita212 & 14 ( gc ) + s1 - 85 & 0:42:40.66 & 40:51:17.5 & 161 & 25.8 & [email protected] & 7.419 & m32-t1 & 22 ( x ) + r1 - 32 & 0:42:41.43 & 41:15:24.4 & 50 & 10.9 & [email protected] & 1.328 & mita213 & 14 ( gc ) + r3 - 31 & 0:42:41.65 & 41:21:06.7 & 30 & 4.5 & [email protected] & 0.408 & cxom31 j004241.6 + 412106 & 15 ( x ) + r1 - 31 & 0:42:42.07 & 41:15:32.4 & 23 & 5.4 & [email protected] & 0.164 & cxom31 j004242.0 + 411532 & 15 ( x ) + r2 - 69 & 0:42:42.16 & 41:14:24.4 & 28 & 6.0 & [email protected] & 0.830 & new & new ( x ) + r1 - 5 & 0:42:42.16 & 41:16:08.4 & 940 & 129.2 & [email protected] & 46.434 & cxom31 j004242.1 + 411608 & 15 ( x ) + r2 - 21 & 0:42:42.33 & 41:14:45.6 & 180 & 38.0 & [email protected] & 0.256 & [ pfj93 ] 39 & 2 ( x ) + s1 - 32 & 0:42:42.45 & 40:51:53.0 & 844 & 71.8 & [email protected] & 1.091 & bbk2000 - 7442 ( m32 ) & 7 ( * ) + r1 - 14 & 0:42:42.47 & 41:15:53.8 & 224 & 35.0 & [email protected] & 0.490 & [ pfj93 ] 41 & 2 ( x ) + r1 - 30 & 0:42:42.60 & 41:16:59.8 & 31 & 6.1 & [email protected] & 0.136 & [ pfj93 ] 40 & 2 ( x ) + r1 - 13 & 0:42:42.98 & 41:15:43.3 & 199 & 42.3 & [email protected] & 0.592 & [ pfj93 ] 42 & 2 ( x ) + r1 - 35 & 0:42:43.12 & 41:16:04.1 & 39 & 8.2 & [email protected] & 0.817 & sss near sn 1885a & 18 ( sn ) + r1 - 24 & 0:42:43.29 & 41:16:40.1 & 36 & 7.0 & [email protected] & 0.215 & ford m 31 322 & 5 ( pn ) + r2 - 19 & 0:42:43.30 & 41:13:19.7 & 92 & 20.4 & [email protected] & 0.726 & qss & qss ( x ) + r1 - 12 & 0:42:43.73 & 41:16:32.7 & 170 & 30.7 & [email protected] & 1.574 & [ pfj93 ] 43 & 2 ( x ) + r1 - 28 & 0:42:43.77 & 41:15:14.2 & 40 & 8.3 & [email protected] & 1.195 & cxom31 j004243.7 + 411514 & 15 ( x ) + r1 - 23 & 0:42:43.84 & 41:16:04.1 & 45 & 8.2 & [email protected] & 0.566 & ford m 31 8 & 5 ( pn ) + r1 - 11 & 0:42:43.86 & 41:16:29.8 & 88 & 18.6 & [email protected] & 0.707 & [ pfj93 ] 43 & 2 ( x ) + r1 - 22 & 0:42:44.22 & 41:16:14.7 & 43 & 6.9 & [email protected] & 0.230 & bfs98-m31 - 53 & 8 ( * ) + r1 - 9 & 0:42:44.36 & 41:16:07.5 & 470 & 62.2 & [email protected] & 8.308 & sss & sss ( x ) + r3 - 30 & 0:42:44.39 & 41:11:58.7 & 39 & 9.2 & [email protected] & 0.289 & cxom31 j004244.4 + 411157 & 15 ( x ) + r1 - 8 & 0:42:44.66 & 41:16:18.7 & 80 & 13.2 & [email protected] & 0.458 & [ pfj93 ] 47 & 2 ( x ) + r3 - 29 & 0:42:44.84 & 41:11:38.2 & 412 & 62.4 & [email protected] & 2.189 & c84 67 & 3 ( x ) + r2 - 18 & 0:42:44.89 & 41:17:40.2 & 66 & 13.3 & [email protected] & 0.505 & cxom31 j004244.8 + 411739 & 15 ( x ) + r2 - 17 & 0:42:45.09 & 41:14:07.4 & 36 & 9.8 & [email protected] & 0.182 & [ pfj93 ] 46 & 2 ( x ) + r1 - 4 & 0:42:45.10 & 41:16:21.9 & 331 & 49.9 & [email protected] & 4.790 & [ pfj93 ] 47 & 2 ( x ) + r1 - 26 & 0:42:45.11 & 41:15:23.4 & 44 & 9.7 & [email protected] & 0.304 & ford m 31 21 & 5 ( pn ) + r2 - 16 & 0:42:45.19 & 41:17:22.5 & 111 & 24.0 & [email protected] & 2.575 & cxom31 j004245.2 + 411722 & 15 ( x ) + r1 - 20 & 0:42:45.20 & 41:16:11.2 & 56 & 10.2 & [email protected] & 0.160 & bfs98-m31 - 419 & 8 ( * ) + r1 - 7 & 0:42:45.59 & 41:16:08.5 & 52 & 9.5 & [email protected] & 0.395 & cxom31 j004245.5 + 411608 & 15 ( x ) + r1 - 19 & 0:42:46.00 & 41:16:19.9 & 37 & 8.0 & [email protected] & 0.518 & cxom31 j004245.9 + 411619 & 15 ( x ) + r1 - 18 & 0:42:46.18 & 41:15:43.1 & 29 & 6.7 & [email protected] & 0.208 & cxom31 j004246.1 + 411543 & 15 ( x ) + r3 - 28 & 0:42:46.95 & 41:21:21.5 & 32 & 5.1 & [email protected] & 0.213 & lgsj004246.9 + 412120 & new ( * ) + r1 - 3 & 0:42:46.97 & 41:16:15.9 & 318 & 42.5 & [email protected] & 2.157 & [ pfj93 ] 49 & 2 ( x ) + r1 - 2 & 0:42:47.16 & 41:16:28.6 & 1004 & 150.4 & [email protected] & 8.826 & ford m 31 13 & 5 ( pn ) + r3 - 27 & 0:42:47.27 & 41:11:57.6 & 33 & 7.8 & [email protected] & 0.210 & cxom31j004247.2 + 411157 & 15 ( x ) + r1 - 17 & 0:42:47.84 & 41:16:23.0 & 30 & 6.9 & [email protected] & 0.327 & cxom31j004247.8 + 411623 & 15 ( x ) + r1 - 6 & 0:42:47.88 & 41:15:33.2 & 348 & 57.1 & [email protected] & 3.205 & [ pfj93 ] 52 & 2 ( x ) + r1 - 25 & 0:42:47.90 & 41:15:49.8 & 18 & 4.7 & [email protected] & 0.180 & sss & sss ( x ) + r1 - 1 & 0:42:48.52 & 41:15:21.4 & 1089 & 175.2 & [email protected] & 1.527 & [ pfj93 ] 54 & 2 ( x ) + r3 - 25 & 0:42:48.55 & 41:25:22.7 & 179 & 16.6 & [email protected] & 2.768 & rx j004248.7 + 412522 & 1 ( x ) + r1 - 16 & 0:42:48.69 & 41:16:24.9 & 46 & 8.9 & [email protected] & 0.417 & cxom31 j004248.7 + 411624 & 15 ( x ) + r2 - 14 & 0:42:49.20 & 41:18:15.9 & 84 & 19.6 & [email protected] & 0.268 & [ pfj93 ] 55 & 2 ( x ) + n1 - 75 & 0:42:51.98 & 41:31:07.7 & 2436 & 54.1 & [email protected] & 1.233 & mita225 & 14 ( gc ) + r2 - 13 & 0:42:52.51 & 41:18:54.8 & 1029 & 165.4 & [email protected] & 1.071 & [ pfj93 ] 57 & 2 ( x ) + r2 - 12 & 0:42:52.52 & 41:15:40.2 & 928 & 159.9 & [email protected] & 3.345 & sss & sss ( x ) + r3 - 69 & 0:42:53.53 & 41:25:55.3 & 45 & 5.7 & [email protected] & 0.197 & ddb 1 - 13 & 11 ( snr ) + r2 - 11 & 0:42:54.92 & 41:16:03.5 & 973 & 169.3 & [email protected] & 0.650 & [ pfj93 ] 60 & 2 ( x ) + r2 - 10 & 0:42:55.14 & 41:18:36.7 & 68 & 16.1 & [email protected] & 0.357 & cxom31 j004255.1 + 411836 & 15 ( x ) + r3 - 23 & 0:42:55.25 & 41:25:55.2 & 102 & 12.0 & [email protected] & 0.304 & [ pfj93 ] 61 & 2 ( x ) + r2 - 9 & 0:42:55.58 & 41:18:35.3 & 48 & 11.9 & [email protected] & 0.352 & bol 138 & 6 ( gc ) + r2 - 8 & 0:42:56.89 & 41:18:44.0 & 42 & 10.4 & [email protected] & 1.121 & lgsj004257.0 + 411844 & new ( * ) + r3 - 22 & 0:42:57.90 & 41:11:04.9 & 414 & 51.1 & [email protected] & 1.638 & rx j004257.7 + 411103 & 1 ( x ) + r2 - 7 & 0:42:58.33 & 41:15:29.2 & 155 & 31.6 & [email protected] & 1.431 & [ pfj93 ] 65 & 2 ( x ) + r2 - 63 & 0:42:59.28 & 41:16:43.5 & 43 & 10.3 & [email protected] & 0.689 & sss & sss ( x ) + r2 - 6 & 0:42:59.64 & 41:19:19.7 & 490 & 79.9 & [email protected] & 0.664 & bol 143 & 6 ( gc ) + r2 - 5 & 0:42:59.86 & 41:16:05.9 & 394 & 78.4 & [email protected] & 3.506 & bol 144 & 6 ( gc ) + n1 - 76 & 0:43:00.87 & 41:30:13.3 & 423 & 8.7 & [email protected] & 0.350 & [ pfj93 ] 69 & 2 ( x ) + r2 - 37 & 0:43:01.07 & 41:13:51.4 & 30 & 7.0 & [email protected] & 0.256 & [ pfj93 ] 68 & 2 ( x ) + r2 - 4 & 0:43:02.93 & 41:15:22.9 & 412 & 71.6 & [email protected] & 5.895 & bol 146 & 6 ( gc ) + r3 - 20 & 0:43:03.08 & 41:10:16.1 & 79 & 10.0 & [email protected] & 0.408 & [ pfj93 ] 71 & 2 ( x ) + r2 - 3 & 0:43:03.23 & 41:15:28.2 & 345 & 62.8 & [email protected] & 6.468 & cxom31 j004303.2 + 411528 & 15 ( x ) + r3 - 19 & 0:43:03.34 & 41:21:22.0 & 198 & 24.3 & [email protected] & 0.217 & mita240 & 14 ( gc ) + r2 - 2 & 0:43:03.85 & 41:18:05.3 & 314 & 54.1 & [email protected] & 3.745 & bol 148 & 6 ( gc ) + r2 - 1 & 0:43:04.25 & 41:16:01.4 & 59 & 12.6 & [email protected] & 0.544 & cxom31 j004304.2 + 411601 & 15 ( x ) + r2 - 67 & 0:43:05.66 & 41:17:03.3 & 222 & 43.0 & [email protected] & 517.664 & lgsj004305.6 + 411703 & new ( * ) + r3 - 67 & 0:43:06.66 & 41:19:16.2 & 27 & 4.6 & [email protected] & 0.110 & cxom31 j004306.8 + 411912 & 15 ( x ) + r3 - 115 & 0:43:07.14 & 41:18:09.5 & 20 & 4.8 & [email protected] & 1.010 & lgsj004306.9 + 411809 & sss ( * ) + r3 - 18 & 0:43:07.52 & 41:20:19.8 & 49 & 7.5 & [email protected] & 0.528 & mita246 & 14 ( gc ) + r3 - 17 & 0:43:08.63 & 41:12:50.1 & 81 & 12.1 & [email protected] & 0.312 & [ pfj93 ] 74 & 2 ( x ) + r3 - 16 & 0:43:09.78 & 41:19:01.2 & 92 & 13.2 & [email protected] & 15.871 & [ wsb85 ] s1 4 & 20 ( ? ) + r3 - 15 & 0:43:10.62 & 41:14:51.7 & 1248 & 150.5 & [email protected] & 2.681 & mita251 & 14 ( gc ) + r3 - 13 & 0:43:13.16 & 41:18:14.1 & 28 & 5.1 & [email protected] & 0.187 & lgsj004318.1 + 411814 & new ( * ) + r3 - 112 & 0:43:14.37 & 41:07:21.8 & 711 & 15.5 & [email protected] & 2.852 & mita257 & 14 ( gc ) + r3 - 9 & 0:43:16.13 & 41:18:41.4 & 56 & 9.2 & [email protected] & 0.363 & cxom31 j004316.1 + 411841 & 15 ( x ) + r3 - 8 & 0:43:18.83 & 41:20:17.4 & 67 & 6.9 & [email protected] & 1.149 & c84 101 & 3 ( x ) + r3 - 126 & 0:43:19.52 & 41:17:56.7 & 98 & 11.3 & [email protected] & 2.628 & xmmu j004319.4 + 411759 & 19 ( x ) + r3 - 7 & 0:43:21.02 & 41:17:50.8 & 105 & 13.3 & [email protected] & 0.925 & ford m 31 209 & 5 ( pn ) + r3 - 63 & 0:43:27.95 & 41:18:31.0 & 199 & 12.8 & [email protected] & 0.196 & ddb 1 - 15 & 11 ( snr ) + r3 - 103 & 0:43:29.28 & 41:07:49.8 & 250 & 11.1 & [email protected] & 0.223 & [ pfj93 ] 81 & 2 ( x ) + r3 - 3 & 0:43:32.50 & 41:10:38.9 & 53 & 7.4 & [email protected] & 0.356 & rx j004331.9 + 411038 & 1 ( x ) + r3 - 2 & 0:43:34.31 & 41:13:26.6 & 262 & 23.8 & [email protected] & 2.120 & c84 105 & 3 ( x ) + r3 - 1 & 0:43:37.27 & 41:14:43.8 & 405 & 32.3 & [email protected] & 0.355 & mita299 & 14 ( gc ) + n1 - 77 & 0:43:44.56 & 41:24:17.1 & 281 & 8.1 & [email protected] & 0.954 & 2e 0040.9 + 4108 & 3 ( x ) + n1 - 78 & 0:43:45.64 & 41:36:57.4 & 47 & 6.8 & [email protected] & 0.570 & bol 193 & 6 ( gc ) + n1 - 79 & 0:43:53.56 & 41:16:51.0 & 200 & 19.1 & [email protected] & 0.448 & d31j004353.8 + 411655.9 & 13 ( * ) + n1 - 80 & 0:44:02.73 & 41:39:26.0 & 31 & 4.9 & [email protected] & 0.378 & rx j004402.4 + 413926 & 1 ( x ) + n1 - 81 & 0:44:22.64 & 41:45:06.7 & 134 & 12.6 & [email protected] & 0.804 & lgsj004422.6 + 414507 & new ( * ) + n1 - 82 & 0:44:25.56 & 41:36:35.2 & 27 & 4.3 & [email protected] & 0.774 & foreground ( [email protected] ) & new ( * ) + n1 - 83 & 0:44:38.02 & 41:45:14.4 & 132 & 11.1 & [email protected] & 0.766 & [ b90 ] 265 & 9 ( radio ) + n1 - 84 & 0:44:48.88 & 41:47:27.9 & 71 & 4.8 & [email protected] & 0.142 & new & new ( x ) + n1 - 61 & 0:45:11.18 & 41:45:56.4 & 97 & 4.7 & [email protected] & 0.282 & rx j004510.9 + 414557 & 1 ( x ) + n1 - 85 & 0:45:28.05 & 41:54:10.9 & 101 & 4.1 & [email protected] & 2.281 & snr90 & 12 ( snr ) + n1 - 17 & 0:45:45.61 & 41:39:41.6 & 5289 & 118.4 & [email protected] & 11.743 & bol 375 & 6 ( gc ) + n1 - 59 & 0:45:45.76 & 41:50:30.1 & 54 & 8.6 & [email protected] & 0.681 & lgsj004545.9 + 415030 & new ( * ) + [ slist ] ccccc disk & [email protected] & 2.6@xmath169 & 1.5@xmath170 & 0.44 + bulge ( w / o gcs ) & [email protected] & 7.0@xmath171 & 1.7@xmath172 & 0.50 + bulge ( w/ gcs ) & [email protected] & 7.1@xmath173 & 1.9@xmath172 & 0.47 + bulge ( w / o transients ) & [email protected] & 7.0@xmath174 & 1.8@xmath172 & 0.35 + [ bplfits ] lcccc s1 - 79 & [email protected] & 1,3,4,5 & @xmath592.33 ( 7 ) & @xmath7566 + s1 - 80 & [email protected] & 1,17 & @xmath591.34 ( 11 ) & @xmath75124 + r3 - 125 & [email protected] & 9,11 & @xmath590.38 ( 11 ) & @xmath7542 + r3 - 46 ( @xmath64 ) & [email protected] & 8,9 & @xmath591.57 ( 9 ) & @xmath7523 + s1 - 82 & [email protected] & 1 & @xmath591.83 ( 10 ) & @xmath759 + r2 - 29 ( @xmath64 ) & [email protected] & 7 & @xmath590.43 ( 12 ) & @xmath7590 + r2 - 28 ( @xmath64 ) & [email protected] & 2,3 & @xmath591.60 ( 11 ) & @xmath7510 + s1 - 1 & [email protected] & 14,15,16 & @xmath592.44 ( 9 ) & @xmath7511 + r1 - 34 ( @xmath64 ) & [email protected] & 14,15 & @xmath592.10 ( 10 ) & @xmath7520 + s1 - 85 & [email protected] & 9,10 & @xmath591.80 ( 9 ) & @xmath7582 + r2 - 69 ( @xmath64 ) & [email protected] & 10 & @xmath590.66 ( 10 ) & @xmath7544 + r2 - 8 ( @xmath64 ) & [email protected] & 9,10,11 & @xmath590.18 ( 10 ) & @xmath75132 + r2 - 67 ( @xmath64 ) & [email protected] & 14,15,16 & @xmath590.81 ( 10 ) & @xmath75421 + r3 - 115 ( @xmath64 ) & [email protected] & 16 & @xmath591.40 ( 8) & @xmath7523 + r3 - 126 & [email protected] & 6,7 & @xmath590.11 ( 8) & @xmath7552 + n1 - 85 & [email protected] & 14 & @xmath592.02 ( 11 ) & @xmath7534 + n1 - 59 & [email protected] & 10 & @xmath591.16 ( 10 ) & @xmath7529 + [ tlist ] cccc s1 - 74 & @xmath292 & & @xmath2914.7 + s1 - 75 & 0.48 & [email protected] & [email protected] + s1 - 76 & & & @xmath7520.7 + s1 - 78 & 0.85 & [email protected] & [email protected] + s2 - 63 & & & @xmath7520.9 + s1 - 45 & @xmath292 & & @xmath2914.7 + s1 - 34 & & & @xmath7521.3 + s1 - 64 & 1.6 & [email protected] & [email protected] + r2 - 24 & & & @xmath7519.5 + r3 - 31 & & & @xmath7520.35 + r1 - 30 & & & @xmath7519.5 + r2 - 18 & & & @xmath7519.5 + r2 - 16 & & & @xmath7519.5 + r3 - 28 & 1.3 & [email protected] & [email protected] + r1 - 2 & & & @xmath7519.5 + r3 - 25 & & & @xmath7520.75 + r2 - 14 & & & @xmath7519.5 + r2 - 13 & & & @xmath7519.5 + r3 - 69 & & & @xmath7520.75 + r2 - 10 & & & @xmath7519.5 + r3 - 23 & & & @xmath7520.75 + r2 - 8 & 1.21 & [email protected] & [email protected] + n1 - 76 & & & @xmath7520.9 + r2 - 67 & 0.54 & [email protected] & [email protected] + r3 - 67 & & & @xmath7520.1 + r3 - 115 & 1.8 & [email protected] & [email protected] + r3 - 16 & 0.73 & [email protected] & [email protected] + r3 - 13 & 1.04 & [email protected] & [email protected] + r3 - 9 & & & @xmath7520.1 + r3 - 8 & & & @xmath7520.1 + r3 - 126 & & & @xmath7520.1 + r3 - 7 & 1.51 & [email protected] & [email protected] + r3 - 63 & & & @xmath7520.55 + n1 - 77 & & & @xmath7520.45 + n1 - 79 & & & @xmath7520.7 + n1 - 80 & & & @xmath7521.4 + n1 - 81 & 0.49 & [email protected] & [email protected] + n1 - 82 & 1.32 & [email protected] & [email protected] + n1 - 83 & & & @xmath7521.7 + n1 - 84 & & & @xmath7521.6 + n1 - 61 & & & @xmath7521.5 + n1 - 59 & 1.50 & [email protected] & [email protected] + [ lgslist ] ccccc 2000 - 10 - 06 & @xmath64 & 300 & [email protected] + 2001 - 09 - 18 & @xmath6 & 300 & [email protected] + 2001 - 11 - 05 & f336w ( u ) & 4000 & [email protected] + 2001 - 12 - 02 & f336w ( u ) & 4000 & [email protected] + 2002 - 02 - 04 & f336w ( u ) & 4000 & @xmath17622.8 + [ r267obs ] ccccc 2000 - 10 - 06 & @xmath64 & 300 & [email protected] + 2001 - 08 - 27 & f336w ( u ) & 4000 & [email protected] + 2001 - 09 - 18 & @xmath6 & 300 & [email protected] + 2001 - 12 - 02 & f336w ( u ) & 4000 & [email protected] + 2002 - 01 - 08 & f336w ( u ) & 4000 & [email protected] + [ r316obs ] in fitting the acis - i spectrum of r2 - 67 from obsid 1585 using a pileup model , we limit ourselves to the 0.37.0 kev range in order to exclude background counts , and we include the ` afterglow ' events , which account for 92 of the 866 total good events .
we do not find a clear change in acis grades in radial averaged profiles but note that this is likely due to our much lower counting rate than that seen from gx 13 + 1 @xcite .
given the high luminosity of the source during the outburst , we should expect the source to show a thermal spectrum in the shape of a disk blackbody @xcite rather than a power - law .
however , fits to both shapes while neglecting pileup may yield useful comparisons to fits including pileup
. fits to a disk blackbody yield @xmath177 ( probability @xmath178 ) , @xmath179 kev , @xmath180 cm@xmath10
. fits to a power - law find @xmath177 , @xmath181 , @xmath182 cm@xmath10 .
both these fits show a clear excess of counts between 2 and 3 kev , which is indicative of pileup @xcite .
a simple blackbody is a very poor representation of the data , yielding at best @xmath183 .
because the pileup model was first developed within isis and later incorporated into sherpa and xspec , we used isis to determine the pileup parameters . allowing the pileup and spectral model parameters to vary ( but freezing the psf fraction to 0.95 )
, isis quickly converges to a model with 30% pileup , a grade migration parameter @xmath90 . with the pileup parameters frozen at the best values
, isis finds spectral parameters of @xmath184 kev , @xmath185 , and @xmath186@xmath25 with @xmath22 = 1.2 ( probability=0.2 ) , where @xmath97 is the temperature of the inner edge of the accretion disk , @xmath98 is the normalization parameter , @xmath99 is the distance to the source , and @xmath100 is the inclination angle of the disk .
the @xmath32 contours are shown in figure [ obs1585 ] .
the observed flux is 1.3 @xmath8 10@xmath102 2s ( 0.3 - 7.0 kev ) , and the modeled emitted luminosity is @xmath103 ( 0.3 - 7.0 kev ) .
we note that when pile - up is included , we are not able to rule - out a power - law model solely on the basis of @xmath32 .
a power - law fit gives a slightly lower @xmath187 , a slope @xmath188 , a much higher @xmath189 cm@xmath10 , but has an acceptable @xmath190 .
while this model has the same observed flux , the steep slope and higher @xmath121 predicts a higher emitted luminosity of @xmath191 .
we discount the power - law model for three reasons .
first , it is not the thermal form expected at this high flux .
second , the @xmath121 is inconsistent with that measured for the same source many months later .
third , this high @xmath121 predicts @xmath192 and the unreasonable high absolute magnitude of m@xmath193 for the hst counterpart
. fits with xspec produced similar results , but were more sensitive to the initial guesses for the spectral parameters and were insensitive to the value of @xmath194 .
we therefore fixed @xmath194 to 0.99 and and the psf fraction to 0.95 .
the best fit disk black body finds @xmath195 , @xmath196 cm@xmath10 , both of which overlap with the isis determined values
. fits with sherpa and the pileup model produced spectral parameters consistent with those determined by isis .
sherpa found a pileup grade migration parameter @xmath197 and a pileup fraction of 30% .
the best fit disk spectral parameters were @xmath198 kev and @xmath199 @xmath200@xmath25 , with @xmath22 = 1.3 ( probability=0.1 ) .
one check on the results of the pileup model fits which has been used previously is to extract the spectrum of the source from only with wings of the psf , where the counting rate is low enough that pileup can be ignored ( @xcite on m81 ) .
while this can give a qualitative measure , the absolute value of the parameters determined this way must be treated with care because hard photons are preferentially scattered into the wings of the psf and the current effective area tools ( i.e. mkarf ) do not take this into account ( i.e. @xcite ) .
we tried ignoring pileup and removing the central pixel only and using the surrounding 8 pixels ( leaving 450 counts ) , and removing the central 9 pixels and using the surrounding 40 ( leaving 107 counts ) .
the first method yields @xmath201 ( prob = 14% ) and shows a significant excess ( @xmath045% above model ) of counts between 2 and 3 kev .
the best fit values are t@xmath202 kev and @xmath203 - 2 .
the second method yields @xmath204 ( prob = 5% ) and similar excess between 2 and 3 kev .
the best fit values are t@xmath205 kev and @xmath206 - 2
. removing the central pixel(s ) appears to decrease the amount of pileup , as evidenced by the slight decrease in the excess ( above model ) of counts between 2 and 3 kev .
however , the fitted temperatures are much higher than and completely inconsistent with the temperature found using all the data and accounting for pileup .
it is unclear to us if these erroneously high temperatures are due to the scattering of hard photons into the wings or due to ignoring the effects of pileup , since both have the effect of hardening the fitted spectrum . | we have obtained 17 epochs of _ chandra _ high resolution camera ( hrc ) snapshot images , each covering most of the m31 disk .
the data cover a total baseline of @xmath02.5 years and contain a mean effective exposure of 17 ks .
we have measured the mean fluxes and long - term lightcurves for 166 objects detected in these data .
at least 25% of the sources show significant variability .
the cumulative luminosity function ( clf ) of the disk sources is well - fit by a power - law with a slope comparable to those observed in typical _ elliptical _ galaxies .
the clf of the bulge is a broken power law similar to measurements made by previous surveys .
we note several sources in the southwestern disk with @xmath1 .
we cross - correlate all of our sources with published optical and radio catalogs , as well as new optical data , finding counterpart candidates for 55 sources .
in addition , 17 sources are likely x - ray transients .
we analyze follow - up hst wfpc2 data of two x - ray transients , finding f336w ( @xmath2-band equivalent ) counterparts . in both cases ,
the counterparts are variable . in one case ,
the optical counterpart is transient with f336w @xmath3 mag .
the x - ray and optical properties of this object are consistent with a @xmath010 solar mass black hole x - ray nova with an orbital period of 23@xmath4 days . in the other case
, the optical counterpart varies between f336w = [email protected] mag and f336w = [email protected] mag .
ground - based and hst observations show this object is bright ( @xmath6 = [email protected] ) and slightly extended .
finally , the frequency of bright x - ray transients in the m31 bulge suggests that the ratio of neutron star to black hole primaries in low - mass x - ray binaries ( @xmath7 ) is @xmath01 . |
a ferroic is a material which exhibits spontaneous and switchable ordering of electric polarization or magnetization or elastic strain .
materials exhibiting more than one of such orderings termed ` multiferroics'- have recently become the focus of much research@xcite .
most of the biferroics investigated in recent years are @xmath3 oxides with perovskite structure .
the @xmath4ness or the zero occupancy of tranistion metal @xmath5 cation is known chemically to favor ferroelectricity@xcite .
hence , the availability of transition metal _ d_-electrons in the perovskite oxides necessary for magnetism , reduces the tendency for off - centering ferroelectric distortions@xcite making multiferroics relatively rare .
how the ordering of @xmath6electronic spins of the @xmath5 cation influence ferroelectric or other competing structural instabilities has not yet been explored and understood .
the coupling between magnetic ordering and structural instabilities is expected to involve interesting physics and is of direct relevance to technological applications@xcite such as multiple state memory elements and novel memory media .
there are at least three families of high temperature biferroic materials .
bi- based perovskite oxides , like bimno@xmath7 @xcite , bicro@xmath7@xcite , and bifeo@xmath7@xcite are known to be promising biferroics .
ferroelectricity in these materials arises from the stereochemical activity of the @xmath8 lone pair electrons of bi .
hexagonal rare earth manganates lnmno@xmath7 and inmno@xmath7 are biferroics which exhibit improper or geometric ferroelectricity@xcite .
rare earth chromites lncro@xmath7 @xcite ( with ln = ho , er , tm , yb , lu or y ) have been recently shown to be biferroic ; ycro@xmath7 has been shown@xcite to exhibit canted antiferromagnetic behavior below 140 k and a ferroelectric transition around 473 k. similarly , lucro@xmath7 becomes a canted antiferromagnet below 115 k , and is ferroelectric below 488 k @xcite .
the absence of any ferroelctricity in lacro@xmath7 , has been attributed to the large size of the la@xmath9 ion , in comparison with y@xmath9 .
however , small values of polarization reported for these materials ( about 2@xmath10c/@xmath11 for ycro@xmath7@xcite and 6@xmath10c/@xmath11 of bifeo@xmath7@xcite ) inspite of large a - cation off - centering distortions remain a puzzle .
more recently@xcite , a new concept of ` local non - centrosymmetry ' in ycro@xmath7 has been proposed to account for the small value of polarization observed .
perovskite oxides are known to have many competing structural instabilities@xcite and this competition is further enriched by the magnetic instabilities .
the coupling and competition between various magnetic and structural ordering can be partly responsible for weak ferroelectricity or the possibility of local non - centrosymmetry .
our goal here is to investigate this issue through determination of various structural instabilities for different magnetic orderings , with a focus on rare earth chromites .
we present results of detailed electronic structure and frozen phonon ( at the @xmath1 and r points ) calculations for a set of five materials ( lucro@xmath7 , ycro@xmath7 , lacro@xmath7 , bicro@xmath7 and yfeo@xmath7 ) in the cubic phase with three different magnetic orderings ( para- , ferro- and antiferromagnetic ) . in section
ii , we briefly describe the methods used in calculations here . in section iii
, we report results for structural energetics , the electronic density of states ( dos ) and the born effective charges ( bec ) for the cubic phase of these materials with different magnetic orderings . in section iv
, we report results for structural instabilities in chromites and compare them with those in a related compound yfeo@xmath7 .
since the many - electron correlations are important in magnetic oxides , we estimate their effects on the structural instabilities through use of the hubbard parameter u @xcite . using our results for structural instabilities , we show how certain nonpolar structural instabilities can cooperatively stabilize ferromagnetism . our work reveals how the structural instabilities of these biferroic oxides depend on the size of a - cation ( with the same b - cation ) , and a change in b - cation .
we interpret the results using arguments based on superexchange@xcite , and the well - known goodenough - kanamori rules@xcite .
finally , we summarize in section v. direction , to represent g - type antiferroamgnetic ordering . ]
our calculations are based on first - principles pseudopotential based density functional theory within a generalized gradient approximation ( gga)@xcite as implemented in the pwscf package@xcite . the interaction between ions and electrons
is approximated with ultrasoft pseudopotentials@xcite treating explicitly 11 electrons [ ( n-1)s@xmath12 ( n-1)p@xmath13 ( n-1)d@xmath14 and ns@xmath12 ] in the valence shell of lu(n=6 ) , la(n=6 ) and y(n=5 ) .
we consider 6 valence electrons for oxygen [ 2s@xmath122p@xmath16 and 14 electrons for cr [ 3s@xmath123p@xmath133d@xmath174s@xmath14 ] .
we used a plane wave basis with kinetic energy cut off of 25 ryd ( 150 ryd ) to represent wave functions ( density ) . for cubic structures
, we sample the brillouin zone using a 5 @xmath19 5 @xmath19 5 monkhorst pack mesh@xcite , and a denser mesh ( 6 @xmath19 6 @xmath19 6 ) and higher energy cut - off ( 30 ryd ) for energy differences .
phonon frequencies calculated with these larger parameters do not differ much from those calculated with a lower cut - off .
we perform spin polarised calculations by initializing different spins on neighbouring magnetic ions ; for paramagnetic ordering we initialise a zero spin on cr ions .
to represent antiferromagnetic ordering , the unit cell is doubled along the @xmath20 direction ( see fig[per ] ) . to investigate structural instabilities in the prototype cubic structure , we determine its dynamical matrix using frozen phonon calculations .
we use a finite difference form of the first derivative to compute an element of the force constant matrix : @xmath21 where f@xmath22 is the hellman - feynman force acting on the _ i@xmath23 _ atom in @xmath24 direction , and , u@xmath25 the displacement of the _
j@xmath23 _ atom in @xmath26 direction with respect to the equilibrium structure .
we used @xmath27=0.04 @xmath28 , about 1@xmath29 of the lattice constant .
the dynamical matrix is then calculated from the force constant matrix , @xmath30 whose eigenvalues correspond to the square of the phonon frequencies ( @xmath31 ) . in periodic systems ,
the dynamical charge tensor or born effective charge tensor can be defined@xcite as the coefficient of proportionality between the macroscopic polarization created in direction @xmath26 and a rigid displacement of the sublattice of atoms _ j _ in direction @xmath24 , @xmath32 @xmath33 being the unit cell volume .
the polarization is determined using the berry phase formalism @xcite as implemented in the pwscf package .
.[lattice]lattice constants of various oxides in the cubic perovskite structure with experimental unit cell volumes . [ cols="^,^,^,^,^ " , ]
in order to represent g - type antiferromagnetic ordering which has been shown to be most favourable energetically , we use a unit cell doubled along the @xmath20 direction .
we determine structural instabilities in this structure with different magnetic orderings .
a single unit cell has 10 atoms which results in 30 phonon branches : 3 accoustic ( which have zero frequency at k=(0,0,0 ) ) and 27 optical , some of which are triply degenerate .
we are interested mainly in optical modes with imaginary phonon frequencies corresponding to instabilities in the structure .
doubling the unit cell along the @xmath20 direction , gives us access to zone boundary phonon modes ( r - point ) which form the dominant structural instabilities in this structure , along with the zone - center modes .
in the paramagnetic phase , both ycro@xmath7 and lucro@xmath7 exhibit a zone center instability at 116.5 and 144.8 @xmath34 respectively , which is a polar mode ( with @xmath35 symmetry ) involving mainly the off centering of a - cation .
this instability , however , is absent in lacro@xmath7 .
we find two instabilities in the fm phase : @xmath35 and @xmath36 modes .
the non - polar @xmath36 mode involves oxygen displacements only , and is strongly unstable in the fm phase .
the @xmath35 mode involves the a - cation ( rare earth ion ) moving in a direction opposite to that of the oxygen cage and cr atom resulting in a ferroelectric polar structural distortion .
note that the cr atom moves in the same direction as the oxygen ion , in contrast to the behavior of ti ion in batio@xmath7@xcite and pbtio@xmath7@xcite , but similar to the behavior of mn in bimno@xmath7@xcite and cr in bicro@xmath7 @xcite . with afm spin arrangement , for lucro@xmath7 , we find three triply degenerate instabilities , at 339.5 @xmath37 , 145 @xmath37 and the weakest at around 60 @xmath37 .
for ycro@xmath7 , the corresponding instabilities are at 309 @xmath37 , 140 @xmath37 and 80 @xmath37 respectively .
the strongest instability ( at around 300 @xmath37 for the two materials ) has r@xmath38 symmetry and corresponds to rotation of the corner connected oxygen octahedra .
the next instability is the ferroelectric @xmath35 mode ( around 140 @xmath37 ) . the weakest instability ( 60 @xmath37 for lucro@xmath7 and 80 @xmath37 for ycro@xmath7 ) has r@xmath39 symmetry and involves displacement of the a - cations ( lu and y for our case ) and small oxygen displacements ; these are antiparallel in neighbouring unit cells . in lacro@xmath7
, we find only two triply degenerate instabilities . the first instability at around 220 @xmath37 corresponding to the oxygen rotations ( the r@xmath38 mode ) and the second close to 18 @xmath37 having @xmath35 symmetry ( see fig [ modes ] ) .
point phonon modes : @xmath35 and @xmath36 modes ] for the rare earth chromites studied in this paper , unstable r - point modes in the afm phase do not change significantly with magnetic ordering . @xmath1 point instabilities in contrast depend strongly on magnetic ordering . only the high frequency phonons are affected in going from para- to antiferromagnetic phase and all @xmath35 and @xmath40 modes are softer in the fm phase .
the @xmath36 mode which brings about a significant change in the cr - o - cr as well as o - cr - o bond angle shows a spectacular change with magnetic ordering .
this mode , highly unstable in the ferromagnetic phase ( at around 200 cm @xmath41 ) becomes stable at around 50 @xmath34 with para- and antiferromagnetic orderings .
this behavior , although not as pronounced , is also seen for this mode in lacro@xmath7 . the r@xmath38 instability , also involving a change in the cr - o - cr bond angle , is not affected by the change in magnetic ordering possibly because the o - cr - o bond angle still remains unchanged .
another significant change is observed for the r@xmath38 mode , which involves a movement of the two b - cations ( cr in our case ) in opposite directions .
after introduction of spin - polarization , the r@xmath38 mode ( close to 250 @xmath34 for pm phase ) becomes more stable at around 400 @xmath34 for the fm and afm phases .
we compare our results with the bi - based biferroic chromite , bicro@xmath7 and find a similar behavior of the @xmath36 mode here as well .
we thus attribute this behavior to the b cation ( cr ion for the chromites ) and expect it to be their general behavior . to interpret the general trend in phonon frequencies ( see fig .
[ phonon ] ) , the following rules apply : \1 .
the modes which involve a change in the cr - o - cr bond angle ( as well as o - cr - o bond angle ) are more stable in the afm phases and are relatively less , or unstable , in the fm phase .
this behaviour is seen in the @xmath36 and @xmath35 modes for lucro@xmath7 .
\2 . secondly , the modes which involve a change in the cr - o bond length tend to harden with the introduction of spin polarization , as observed for the r@xmath38 and r@xmath42 modes . with different magnetic orderings .
insets show similar curves for lacro@xmath7 and bicro@xmath7.,width=302 ] in order to study the effect of change in b - cation ( see fig [ yfecr ] ) , we compare instabilities in yfeo@xmath7 with ycro@xmath7 .
we find that a change in magnetic ordering has an opposite effect on the unstable modes involving a change in fe - o - fe bond angle .
a spectacular change is the r@xmath43 mode , an oxygen breathing mode , which softens in the fm phase as compared to the pm and afm phases .
the @xmath36 mode also shows a different behaviour , showing a stabilization with fm ordering .
these differences are due to the filled e@xmath44 orbitals in fe@xmath9 , which are unoccupied in cr@xmath9 , leading to a much stronger superexchange interaction mediated by the e@xmath44 orbitals , and have a different geometry dependence
. and yfeo@xmath7.,width=321 ] the spectacular change in the frequency of the @xmath36 mode with fm ordering , prompts us to discuss whether ferromagnetism can be stabilized in these chromites by varying the cr - o - cr bond angle .
since the @xmath36 mode is unstable only in the fm phase , we want to study the effect of freezing in a distortion of this mode .
we have calculated the total energy as a function of @xmath36 displacements in the @xmath20 and @xmath45 directions , for the fm and afm phases ( see fig [ cross_over ] ) . for lucro@xmath7
, we observe a crossover , at a displacement of 0.407 @xmath28 , beyond which the fm phase is energetically favoured . the cr - o - cr bond angle , at the crossover point is found to be 153@xmath46 , which is significantly different from , the value suggested by goodenough and kanamori for ferromagnetic superexchange interaction ( 130@xmath46 ) in 1951 .
on examining the density of states beyond the crossover point ( see fig . [ crossoverdos ] ) , we find more significant hybridization between cr and oxygens for the fm phase . secondly , the fm phase so stabilized is found to be insulating . mode for fm and afm orderings in lucro@xmath7 .
fm becomes more stable than the afm state for rhombohedral distortions greater than 0.407 @xmath28.,width=302 ] displacement of 0.407 @xmath28 ( see fig .
[ cross_over ] , where the fm state is more stable).,width=245 ] as mentioned earlier in the paper , the lda+u method has been successfully applied to describe the electronic structure of sytems containing localized _
d _ and _ f _ electrons where lda sometimes leads to incorrect results@xcite , and recently it has been applied to obtain structural parameters that are in better agreement with experimental results than lda or gga . @xcite . in this work
, we use a value of u = 3.0 ev , adapted from work on full structural optimization of bicro@xmath7 @xcite . with the introduction of correlations through u parameter , we find that the modes in fm phase do not change significantly . our results for ycro@xmath7 ( see fig [ ycr_plus_u ] ) bear that only @xmath36 , r@xmath43 and r@xmath42 modes are noticeably affected .
correlations lead to softening of the r@xmath43 and r@xmath42 modes by 50 @xmath34 , and tend to harden the @xmath36 mode in the pm phase . in an afm spin arrangement , these r - point modes are hardened by approximately 20 @xmath34 and the @xmath36 instability is softened . with different magnetic orderings .
for each mode , data on the left of the vertical dashed line represents estimates with gga and the data on the right represents estimates with gga+u ( u=3.0 ev).,width=283 ]
in conclusion , we have determined structural instabilities of lacro@xmath7 , lucro@xmath7 , ycro@xmath7 and bicro@xmath7 in their cubic perovskite structures with different magnetic orderings .
our finding that the g - type antiferromagnetic ordering is most stable can be explained with superexchange arguments .
ferroelectric structural instabilities in the cubic structures involve a - cation ( lu or y ) displacements , as indicated by the eigenvectors of the ferroelectric @xmath35 modes and an anomalous bec of the a - cation .
we find that certain phonon frequencies depend sensitively on magnetic ordering : the modes involving a change in bond - angle are stable ( harder ) with the antiferro- and paramagnetic ordering than in the fm state ; on the other hand , the modes involving a change in cr - o bond length are softer in the paramagnetic phase and comparable in the fm and afm states .
the @xmath36 oxygen mode brings about a significant change in the cr - o - cr bond angle and is highly unstable in the fm phase , and corresponding structural distortion leads to stabilization of ferromagnetic ordering in these chromites . among the competing structural instabilities the antiferrodistortive instability ( @xmath47 mode )
is the strongest .
electron correlations are found to have little effect on the unstable phonon modes , but result in a slight change in a few of the stable phonon modes in the pm phase .
we note that the effects of magnetic ordering on structural instabilities are quite different ( in fact , opposite sometimes ) in yfeo@xmath7 with respect to chromites , and a more detailed study is required to understand such couplings in ferrites .
origin of small polarization and/or local non - centrosymmetry@xcite is probably from the relatively weak ferroelectric instabilities and their competition with various structural magnetic instabilities , and our work should be useful in formulating a phenomenological analysis of the same .
nirat ray thanks jncasr for summer research fellowship programme and joydeep bhattacharjee for discussions .
uvw is thankful to professor c n r rao for stimulating discussions and encouragement for this work and acknowledges use of central computing facility and financial support from the centre for computational materials science at jncasr .
99 n. a. spaldin , phys .
world , april 2004 ; t. kimura et al nature * 426 * , 55 ( 2003 ) ; n. a. spaldin and m. fiebig , science * 309 * , 391 ( 2005 ) .
n.a . hill , j. phys .
b * 104 * , 6684 - 6709 ( 2000 ) .
wood , v. e. ; austin , a. e. _ magnetoelectric interaction phenomena in crystals _ ; freeman , a. j. , schmid , h. , eds . ; gordon and breach : langhorne , pa , 1975 .
hill and k.m .
rabe , phys .
b * 59 * , 8759 ( 1999 ) ; a. moreira dos santos et al , phys . rev .
b * 66 * , 064425 ( 2002 ) .
n.a . hill , pio bttig and claude daul , j. phy chem b * 106 * , 3383 - 3388 ( 2002 ) .
et al . _ ,
science * 299 * , ( 5613 ) , 1719 ( 2003 ) .
j. b. neaton et al , phys .
b * 71 * , 014113 ( 2005 ) .
b. b. vanaken , thomas t.m .
palstra1 , a. filippetti and n. a. spaldin , nat . mat . * 3 * , 164 ( 2004 ) . c. r. serrao
_ et al . _ , j. appl
. phys . * 100 * , 076104 ( 2006 ) . c. j. fennie and k. m. rabe , phys .
b * 72 * , 100103(r ) ( 2005 ) .
m. feibig , j. magn .
mater . * 290 - 291 * , 883 ( 2005 ) . c. r. serrao _ et al .
_ , phys rev b * 72 * , 220101 ( 2005 ) .
j. r. sahu _ et al .
_ , j. mater
17 * , 42 - 44 ( 2007 ) .
k. ramesha _
et al . _ , j. phys .
matter * 19 * , 102202 doi:10.1088/0953 - 8984/19/10/102202 ph ghosez , e. cockayne , u.v . waghmare and k.m .
rabe , phys .
b * 60 * , 836 ( 1999 ) . v. i. anisimov , f. aryasetiawan and a. i. lichtenstein , j. phys .
matter * 9 * , 767 ( 1997 ) .
p. w. anderson , phys .
rev . * 79 * , 350 - 356 ( 1950 ) .
j. b. goodenough , phys . rev . * 100 * 564 ( 1955 ) .
y. wang and j. p. perdew , phys .
b * 44 * , 13298 ( 1991 ) . s. baroni , a. dal corso , s. de gironcoli , and p. giannozzi , 2001 , + http://www.pwscf.org .
d. vanderbilt , phys .
b * 41 * , 7892 ( 1990 ) .
monkhorst and j.d .
pack , phys .
b * 13 * , 5188 ( 1976 ) ; * 16 * , 1748 ( 1977 ) .
ghosez , j. p. michenaud , and x. gonze , phys .
b * 58 * , 6224 ( 1998 ) .
r. resta , _ berry phase in electronic wave functions _ , available at : + http://ale2ts.ts.infm.it:6163/126resta/publ/notes_trois.ps.gz .
m. eibschitz , acta . cryst . * 19 * , 337 ( 1965 ) .
alessio filippetti and nicola a. hill , phys .
85 * , 5166 - 5169 ( 2000 ) .
p. gopal , n. a. spaldin and u. v. waghmare , phys .
b * 70 * , 205104 ( 2004 ) .
stoner , proc .
* 165 * , 372 ( 1938 ) .
r. e. cohen , nature * 358 * , 136138 ( 1992 ) . o. bengone , m. alouani , j. hugel , and p. blchl , comput .
* 24 * , 192 ( 2002 ) .
a. kokalji , xcrysden a new program for displaying crystalline structures and electron densities , j. mol .
graphics modelling * 17 * , 176 ( 1999 ) . | we use first - principles density functional theory - based calculations to investigate structural instabilities in the high symmetry cubic perovskite structure of rare - earth ( r @xmath0 la , y , lu ) and bi - based biferroic chromites , focusing on @xmath1 and @xmath2 point phonons of states with para- , ferro- , and antiferromagnetic ordering .
we find that ( a ) the structure with g - type antiferromagnetic ordering is most stable , ( b ) the most dominant structural instabilities in these oxides are the ones associated with rotations of oxygen octahedra , and ( c ) structural instabilities involving changes in cr - o - cr bond angle depend sensitively on the changes in magnetic ordering . the dependence of structural instabilities on magnetic ordering can be understood in terms of how super - exchange interactions depend on the cr - o - cr bond angles and cr - o bond lengths . we demonstrate how adequate buckling of cr - o - cr chains can favour ferromagnetism . born effective charges ( bec ) calculated using the berry phase expression are found to be anomalously large for the a - cations , indicating their chemical relevance to ferroelectric distortions . |
in feb . 1999 , a new x - ray transient was independently discovered by _ bepposax _
( sax j1819.32525 ; in t zand et al . 1999 ) and _ rxte _ ( xte j1819254 ; markwardt , swank , & marshall 1999a ) .
its position is consistent with that of the variable star v4641 sgr ( note that v4641 sgr had been misidentified as gm sgr [ see iau circular 7277 ] ) .
its 210 kev flux varied between @xmath11 and 80 mcrab ( in t zand et al .
1999 ; markwardt et al .
on 15 sept .
1999 , the source rapidly increased in the optical ( stubbings 1999 ) and with the _ rxte _ all - sky monitor ( asm ) its 212 kev intensity was observed to increase very rapidly ( within 7 hours ) from 1.6 to 12.2 crab ( smith , levine , & morgan 1999a , b ) .
subsequent asm measurements showed that within two hours of this flare , the flux had declined down to a level only marginally detectable with the asm ( @xmath150 mcrab ; smith et al .
1999c ) . optical and infrared spectra taken during this bright x - ray event show emission lines ( ayani & peiris 1999 ; liller 1999 ; djorgovski et al .
1999 ; charles , shahbaz , & geballe 1999 ) , reminiscent of accretion of matter onto a compact object , demonstrating that v4641 sgr is indeed the optical counter part . on 16 sept . 1999 ,
vla observations were taken and a radio source was discovered ( hjellming , rupen , & mioduszewski 1999a ) at a position consistent with v4641 sgr .
follow up vla and atca observations showed that its flux was rapidly declining on time scales of hours to days ( gaensler et al .
1999 ; hjellming , rupen , & mioduszewski 1999b ; hjellming et al .
the vla observations also showed that it was resolved , demonstrating the presence of ejecta ( hjellming et al .
1999b , c ) . on 15 sept .
1999 , a short ( 3000 s ) pointing was taken with the _ rxte _ proportional counter array ( pca ) .
a rapidly variable source was observed with a flaring and a quiescent episode ( markwardt , swank , & morgan 1999b ) .
no pulsations or quasi - periodic oscillations ( qpos ) were detected but red noise below 30 hz was present . here , we discuss in more detail the rapid x - ray variability as observed with the pca during this observation .
several public too pca observations were scheduled between 15 and 18 sept .
1999 for a total of @xmath033 ksec on - source time . however , only the first 1500 s of the first observation ( on 15 sept .
1999 21:1822.43 utc ; see also markwardt et al .
1999b ) showed very high fluxes and strong variability .
we limited our analysis to these first 1500 s in order to study the x - ray variability of the source . during the second 1500 s of this observation and during the later observations ,
v4641 sgr could still be detected with the pca , but at a very low flux level ( @xmath1100 counts s@xmath2 for 5 detectors on ; 260 kev ) and no significant variability could be detected ( upper limits of 5%40% rms [ 222.1 kev ; 0.01100 hz ; depending on total on - source time and source count rate during the individual observations ] on band limited noise similar to that detected during the first @xmath0900 s of the first observation [ see below ] ) . during the pca observations ,
data were accumulated in several different modes which were simultaneously active . here , we used data obtained with the binned modes b_250us_1m_0_249_h ( one photon energy channel [ effective range of 260 kev ] and 244 @xmath3s time resolution ) and b_4m_8b_0_49_h ( eight channels covering 222.1 kev and 3.9 ms time resolution ) .
these data were used to calculate 128 s ffts to create the power spectra and the cross spectra , the 3.9 ms data were used to create light curves , a hardness - intensity diagram , and a hardness curve .
the power spectra were fitted with a function consisting of a constant ( the dead - time modified poisson level ) and a broken power law ( the band - limited noise below @xmath0100 hz ) .
a lorentzian was used to determine upper limits ( 95% confidence level ) on the rms amplitude of qpos above 100 hz , assuming a qpo width of 50 hz . to correct for the small dead - time effects on the lags
, we subtracted the average 50125 hz cross - vector from the cross spectra ( van der klis et al .
1987 ) .
the 222.1 kev light curve of the first 1500 s of the first pca observation is shown in figure [ fig : lc]_a _ , showing the flaring behavior of v4641 sgr between 300 and 900 s from the start .
after 1000 s , the source enters a quiescent state , with one last flare between 1010 and 1060 s ( fig.[fig : lc]_c _ ) .
after that ( and also in the second 1500 s of this observations , and in the other observations ) it remained in this state at very low count rates ( see also markwardt et al .
1999b ) . during the flaring episode , it varied rapidly in count rate . on time
scales of 510 minutes it varied in luminosity by a factor of up to 500 . within one second
, it sometimes increased and then decreased by more than a factor of 4 ( figs .
[ fig : lc]_b _ and _ d _ ) .
the strong variability is also evident from the power spectra obtained from the first 896 s ( see fig . [
fig : powercrossspectra]_a _ ) .
strong band - limited noise ( 47.2%@xmath40.8% rms amplitude ; 0.01100 hz ; 222.1 kev ) can clearly be seen .
its shape fits a broken power law with the break at [email protected] hz and an index below and above the break of [email protected] and [email protected] , respectively .
the rms amplitude decreases from @xmath054% to @xmath035% as a function of energy , increasing from @xmath04 kev to @xmath020 kev ( fig .
[ fig : rmsenergy]_a _ ) . by subtracting the 3.9 ms data ( 222.1 kev ) from the 244 @xmath3s data ( 260 kev ) ,
we could make a power spectrum for the data above 22.1 kev .
the rms amplitude of the noise was even less , 30.6%@xmath40.5% , above 22.1 kev ( effective range of 22.160 kev ) . to show the decrease of amplitude below 20 kev more clearly , we excluded this point from the figure .
the noise also has significant hard phase lags between photons with energies of 29.7 kev and those with energies of 9.722.1 kev ( fig .
[ fig : powercrossspectra]_b _ ) . the lags increased from being consistent with zero below 0.1 hz to @xmath00.1 rad at @xmath00.03 hz . above this frequency , the lags decreased again and above about 10 hz the lags were consistent with zero again .
we studied the energy dependence of the lags for four frequencies intervals : 0.020.1 hz ( fig .
[ fig : rmsenergy]_b _ ) , 0.11.0 hz ( fig .
[ fig : rmsenergy]_b _ ) , 1.010.0 hz ( fig .
[ fig : rmsenergy]_b _ ) , and 10.050.0 hz ( not shown , the lags were always consistent with zero ) . although the lags in the range 0.020.1 hz were barely significant ( fig . [ fig : rmsenergy]_b _ ) a clear trend with energy was present . in the power spectrum obtained from the 244 @xmath3s data ,
no qpos above 100 hz were detected ( upper limits of 2% rms ) .
however , these data cover the whole _ rxte
_ energy band .
therefore , such qpos could have been present but undetectable if their strength depended strongly on energy , similar to the khz qpos observed in many neutron star and some black hole systems ( van der klis 1998 , 1999 ; remillard et al .
1999 ; homan , wijnands , & van der klis 1999 ) . from the 3.9 ms data
, we made a hardness - intensity diagram ( fig .
[ fig : hid]_a _ ) and a hardness curve ( _ b _ ) with a time resolution of 2 s. for comparison , we also plotted the 222.1 kev light curve ( fig .
[ fig : hid]_c _ ) . as hardness we used the count rate ratio between 9.722.1 kev and 29.7 kev and as intensity the 222.1 kev count rate , but plotted on a logarithmic scale to show the variations more clearly . from figure [ fig : hid]_a _ , it is clear that during the flaring episode the hardness decreased when the count rate increased .
when the transition to the quiescent state occurred the source became softer .
during the last flare ( in the quiescent state ; fig .
[ fig : lc]_c _ ) , the overall hardness was less than for the flares during the flaring episode .
we have presented the x - ray variability of v4641 sgr during its 1999 sept .
15 bright event .
it shows strong variations by a factor of 4 on time scales of seconds to variations by a factor of @xmath0500 on time scales of minutes .
these strong variations are also evident in the power spectrum by the presence of strong ( 30%55% rms amplitude , depending on energy ) band - limited noise . assuming the most likely distance of 0.51.0 kpc ( hjellming et al .
1999c ; obtained by performing an hi absorption experiment against the radio counter part ) and using the maximum flux during our pca observation ( markwardt et al .
1999b ) , the maximum intrinsic luminosity ( 260 kev ) was 0.31.0 @xmath5 erg s@xmath2 ( note that it was at considerable higher luminosities during the asm measurements ) .
this luminosity , the strong variability on short time scales , and the presence of optical and infrared emission lines ( see [ intro ] ) suggest that the x - rays are produced by accretion onto a compact object .
the exact nature of the compact object is difficult to determine .
the likely intrinsic luminosity and the hardness of the spectrum ( markwardt et al . 1999b
; see smith et al .
1999c for a more detailed analysis ) strongly indicate that accretion onto a white dwarf , a nova explosion , and thermonuclear burning on a white - dwarf surface can not account for the x - ray emission and its properties .
compared to the intrinsically brightest low magnetic field neutron star and black hole systems , the luminosity of v4641 sgr during our pca observations was relatively low ( the brightest systems have an luminosities of @xmath6 @xmath7 erg s@xmath2 ) , which suggests that also the accretion rate was relatively low . at such low accretion rates , the rapid variability ( van der klis 1995 ; wijnands & van der klis 1999 ) , including the phase lags ( ford et al .
1999 ) , and the spectrum ( e.g. , barret & vedrenne 1994 ) of the low magnetic field neutron star and the black hole systems are very similar ( van der klis 1994 ) , making it difficult to determine the exact nature of the compact object in v4641 sgr . although the very strong variability suggests a black hole primary in this system ( usually such strong variability is only observed for the black - hole systems ) , it can not be excluded that some neutron star systems can also exhibit such strong variability .
the decrease of the strength of the variability in v4641 sgr with energy is similar to what has been observed in several black - hole systems in their low state ( e.g. , nowak et al . 1999a ; nowak , wilms , & dove 1999b ) , but the neutron star systems have not been studied in enough detail in this respect to allow detailed comparisons between the different types of systems .
it was suggested that the nature of the compact objects in x - ray transients could be determined by studying and comparing the x - ray emission properties of those systems in their quiescent state ( see , e.g. , rutledge et al .
the luminosity for v4641 sgr in its quiescent episode during the first pca observation was 0.52.2 @xmath8 erg s@xmath2 but monitoring observations with the pca of the galactic - center region showed that during the last 7 months the luminosity had dropped occasionally to 0.62.4 @xmath9 erg s@xmath2 ( see markwardt et al .
1999b for the fluxes used ) . however
, these luminosities do not give more insight in the nature of the compact object .
both the observed or derived upper limits on the luminosities for neutron star and black hole x - ray transients in quiescence are consistent with the values detected for v4641 sgr in its quiescent episode .
more detailed studies in quiescence are needed to determine the lowest observed luminosity in quiescence .
if v4641 sgr contains a black hole then it is expected that the luminosity should drop sometimes significantly below 10@xmath10 erg s@xmath2 , as observed in other black - hole transients in quiescence .
thus , the properties of v4641 sgr as observed with the pca are very similar to those observed for an accreting compact object which accretes matter from its companion star at a low rate .
although a neutron star primary can not be excluded , the strong variability and the low intrinsic luminosity make an interpretation of v4641 sgr as a black - hole candidate in the low state most favorable .
the most promising way to determine the exact nature is to dynamically ( during the quiescent state ; from the optical line spectrum of the companion star ) constrain the mass of the primary .
if the primary is truly a black hole , then several features have been observed for v4641 sgr which are not commonly observed in black hole systems in their low state .
for example , its very strong variability have only been observed in a few other systems ( e.g. , grs 1915 + 105 : greiner , morgan , & remillard 1996 ; belloni et al . 1997 ; taam , chen , & swank 1997 ; gs 2023 + 338 : terada et al .
1994 ; oosterbroek et al .
but a major difference is that the intrinsic luminosities of those sources were much higher ( about @xmath11 erg s@xmath2 ; see , e.g. , belloni et al .
1997 ; terada et al .
1994 ) than for v4641 sgr .
another unusual property of v4641 sgr is the very short time span of its bright x - ray event ( @xmath12 hours ) . in this respect ,
v4641 sgr is similar to the recently discovered transient ci cam ( smith et al .
. however , several differences are also present .
the outburst of ci cam was on slightly longer time scales ( days ) and much more smooth ( belloni et al . 1999 ) than the event observed for v4641 sgr , which exhibited very strong variability .
so , although v4641 sgr resembles several sources in some of its behavior , it differs from each of them significantly in other respects .
reanalysis of the asm archive revealed several similar short lived events for v4641 sgr as the sept .
15 event ( smith et al .
1999c ) , which went previously unnoticed .
15 event was directly noticed because ( _ a _ ) more attention to v4641 sgr was paid because of its sudden increase in the optical ( stubbings 1999 ) and ( _ b _ ) this event was brighter ( 230 times ) than the others .
the short life times of the events and the strong flux fluctuations indicate that the accretion is very unstable and highly irregular and not much accretion takes place .
the fact that several events went unnoticed suggest that many sources with similar events also fail to get noticed , especially when they are at a greater distance than v4641 sgr and have therefore lower fluxes .
the exact number of such sources in our galaxy is difficult to estimate because of the uncertainties in the distances and the recurrence time scales of their outbursts .
however , if v4641 sgr harbors a black hole , it is clear that a sizeable number of the black holes in our galaxy could be present in v4641 sgr like systems . | we report on the rapid x - ray variability of the variable star and x - ray transient v4641 sagittarii ( sax j1819.32525 ; xte j1819254 ) as observed on 15 sept .
1999 by the proportional counter array ( pca ) on board the _ rossi x - ray timing explorer _ ( _ rxte _ ) . during the first @xmath0900 s of the first pca observation , v4641 sgr showed very strong x - ray fluctuations by a factor of 4 on time scales of seconds to about 500 on time scales of minutes .
the spectrum of the source during this flaring episode became harder when the count rate decreased .
after this flaring episode , v4641 sgr entered a quiescent state in which it remained for the rest of this , and the subsequent pca observations .
the x - ray spectrum was considerably softer in this quiescent state than during the flaring episode .
the intrinsic x - ray luminosity ( both during the flaring episode and the quiescent state ) and the rapid x - ray variability do not strongly constrain the nature of the compact object ( neutron star or black hole ) in the system , although a black hole seems to be more likely .
the very short duration of the bright x - ray phase of v4641 sgr and its likely close distance suggest that many similar objects could be present in our galaxy , most of which are not noticed when they are in x - ray outburst due to the short duration of these outbursts .
a considerable number of the black holes present in our galaxy might be contained in systems similar to v4641 sgr . |
we would like to thank alan bray , lawrence krauss , glenn starkman and rainer trebin for useful comments and are grateful to the isaac newton institute for their hospitality and for providing a unique environment where this work was initiated .
tv is grateful to the rosenbaum foundation for support during his stay at the isaac newton institute .
the work of twk was supported by the u. s. department of energy grant de - fg05 - 85er40226 and the vanderbilt university college of arts and sciences .
999 analogous defects occur in condensed matter systems where they are known as `` semidefects '' .
see r. kutka , h - r .
trebin and m. kiemes , j. phys .
france * 50 * , 861 ( 1989 ) .
y. nambu , nucl . phys . *
b130 * , 505 ( 1977 ) .
h. b. nielsen and p. olesen , nucl .
b61 * , 45 ( 1973 ) .
n. s. manton , phys .
* d28 * , 2019 ( 1983 ) .
t. vachaspati , phys . rev
. lett . * 68 * , 1977 ( 1992 ) ; * 69 * , 216(e ) ( 1992 ) ; nucl . phys . *
b397 * , 648 ( 1993 ) .
t. vachaspati and m. barriola , phys .
lett . * 69 * , 1867 ( 1992 ) .
m. barriola , t. vachaspati and m. bucher , phys . rev .
* d50 * , 2819 ( 1994 ) .
t. p. cheng and l. f. li , `` gauge theory of elementary particle physics '' , oxford university press ( 1991 ) .
g. t hooft , nucl .
b79 * , 276 ( 1974 ) ; a.m. polyakov , jetp lett . * 20 * , 194 ( 1974 ) . if @xmath21 is absent , the cubic term in the potential is allowed and is constrained to be small by the experimental results on the @xmath70 parameter .
t. w. kephart , t.j .
weiler , and t. c. yuan , nucl .
* b330 * , 705 ( 1990 ) ; t. farris , t. w. kephart , t.j .
weiler , and t. c. yuan , phys .
68 * , 564 ( 1992 ) . h. georgi and s. l. glashow , phys .
. lett . * 28 * , 1494 ( 1972 ) .
j. preskill and a. vilenkin , phys . rev . *
d47 * , 2324 ( 1993 ) .
e. witten , nucl .
b249 * , 557 ( 1985 ) .
m. earnshaw and w. perkins , phys
b328 * , 337 ( 1994 ) .
if the @xmath7 field forms a bosonic condensate on the string , there will also be a contribution to the electric current from the bosonic charge carriers on the string .
this additional current will contribute to the energy and quantum numbers on linked loops of string as discussed in @xcite .
t. vachaspati and g. b. field , phys .
lett . * 73 * , 373 ( 1994 ) .
j. garriga and t. vachaspati , nucl . phys .
* b * , to be published ( 1995 ) .
m. james , l. perivolaropoulos and t. vachaspati , nucl .
* b395 * , 534 ( 1993 ) .
the stability of @xmath11strings with a bosonic condensate was analyzed in a slightly different model in t. vachaspati and r. watkins , phys .
lett . * b318 * , 163 ( 1993 ) . in the present case
, we expect the stability to be even stronger because the @xmath7 field directly contributes to the mass of the @xmath64 bosons and instabilities due to @xmath64 condensation are suppressed .
there is also the possibility that , at a critical dilution of monopoles , the strings will percolate and a few very large strings will form .
chudnovsky , g.b .
field , d.n .
spergel and a. vilenkin , phys . rev .
* d34 * , 944 ( 1986 ) ; a. vilenkin and g.b .
field , nature * 326 * , 772 ( 1987 ) ; a. vilenkin , nature * 332 * , 610 ( 1988 ) ; e. chudnovsky and a. vilenkin , phys . rev .
lett . * 61 * , 1043 ( 1988 ) .
a. vilenkin and e.p.s .
shellard , `` cosmic strings and other topological defects '' , cambridge university press ( 1994 ) .
t. vachaspati , phys .
b265 * , 258 ( 1991 ) .
m. aryal , a. vilenkin and t. vachaspati , phys . lett .
* b194 * , 25 ( 1987 ) .
k. schwarz , talk at isaac newton institute , october 1994 .
a similar scenario for baryogenesis was considered by m. barriola , phys .
* d51 * , r300 ( 1995 ) .
krauss and t. vachaspati , in progress .
1 * a schematic diagram of the vacuum expectation values of @xmath3 and @xmath7 as functions of temperature . when the temperature falls to @xmath80 , @xmath7 gets a vev . at @xmath72 , @xmath3 starts acquiring a vev and the vev of @xmath7 starts turning off and vanishes at @xmath69 . | we propose a criterion to classify hybrid defects occuring in field theoretic models such as the standard electroweak model .
this criterion leads us to consider the minimal extension of the electroweak model in which electroweak magnetic monopoles and @xmath0-strings are topological .
we briefly discuss the cosmology of such defects .
@xmath1 1.5 truecm * topological incarnations of electroweak defects * thomas w. kephart _ department of physics and astronomy , vanderbilt university _ _ nashville , tn 37235 .
_ tanmay vachaspati _ physics department , case western reserve university _ _ cleveland , oh 44106 .
_ \1 .
truecm the standard electroweak model is based on an @xmath2 symmetry group with the order parameter @xmath3 being an @xmath4 doublet .
when @xmath3 acquires a vacuum expectation value ( vev ) , the symmetry breaks down to the electromagnetic @xmath5 .
given this symmetry breaking scheme , one finds that the vacuum manifold is a three sphere which has trivial first and second homotopy groups .
this means that there are no topological monopoles and strings in the electroweak model
. however , these topological considerations are blind to any _ confined defects _ @xcite that may be present in the model .
it is known that the standard model of the electroweak interactions contains confined magnetic monopoles @xcite .
these magnetic monopoles are confined by nielsen - olesen strings @xcite which carry magnetic flux of the @xmath0 boson @xcite .
the existence of the @xmath0-strings can be viewed as being due to the embedding of topological defects within the standard model @xcite . but this viewpoint does not apply to the monopole and a somewhat different reasoning must be used to understand the existence of magnetic monopoles in the standard model .
how might we detect confined defects from the topology of the model ?
a topological scheme which is sensitive to confined defects follows once we realize that when an order parameter acquires a vev , composite operators ( suitably normal ordered ) that can be built from the order parameter can also acquire vevs .
the vev of these composite operators yields different symmetry breaking patterns with vacuum manifolds which have different topological properties .
if the vacuum manifold resulting from the vev of a composite operator has non - trivial first or second homotopy groups , we can get confined strings or confined monopoles .
this scheme for classifying confined defects can be illustrated in the case of the standard model . here , when @xmath3 acquires a vev , the composite operator @xmath6 also acquires a vev .
( we have used standard notation such as given in @xcite and normal ordered with respect to the @xmath3 ground state vacuum ) .
the field @xmath7 transforms in the adjoint representation of @xmath4 and is a singlet under @xmath8 .
so the vev of @xmath7 breaks @xmath4 to @xmath9 and the vacuum manifold is @xmath10 with a non - trivial second homotopy group .
therefore there are magnetic monopoles present in the model .
in other words , nambu s electroweak magnetic monopole is precisely a hedgehog in the @xmath7 field @xcite .
the reason the monopoles are confined is that the relation ( [ chi ] ) for a hedgehog configuration of @xmath7 can not be globally inverted to give a non - vanishing field configuration for @xmath3 .
therefore there is a line emanating from the monopole along which @xmath3 vanishes .
this is the location of the @xmath11string attached to the monopole .
the @xmath11string is _ not _ a dirac string that returns the electromagnetic flux ; it is a string that confines the magnetic monopole without having anything to do with the electromagnetic flux .
even higher dimensional representation operators can be constructed from @xmath3 and these will yield other confined defects .
for example , the 5 dimensional representation of @xmath4 constructed using @xmath12 will break @xmath13 to @xmath14 yielding an even richer array of confined topological defects such as alice strings bounded by walls . here
we will not discuss this very interesting issue further but restrict our attention to electroweak magnetic monopoles and @xmath11strings .
having realised that the monopoles are hedgehogs in the composite real scalar field @xmath7 , we can promote the monopoles to being truly topological objects by promoting @xmath7 to be a fundamental field .
so the lagrangian we will consider is : @xmath15 where , @xmath16 is the gradient part of the bosonic sector of the electroweak lagrangian , @xmath17 is the fermionic part of the lagrangian , @xmath18 ( @xmath19 ) and , @xmath20 if we impose an additional @xmath21 symmetry on the lagrangian under @xmath22 the symmetry is @xmath23 and we must set @xmath24 .
in what follows , we shall only consider this case and henceforth ignore the last ( cubic ) term in the potential @xcite .
in this case , the vev of @xmath3 breaks the symmetry down to @xmath25 .
note that the leptons and quarks do not couple to @xmath7 since we have imposed the @xmath21 symmetry on @xmath26 .
this is because there is no lorentz invariant dimension four operator that can be constructed using the left - handed fermion @xmath27 doublets and the @xmath28 field which is a triplet and still be invariant under the @xmath21 transformation .
therefore @xmath17 is identical to the fermionic lagrangian of the standard electroweak model . at non - zero temperatures , the coefficients in the potential
will get thermal corrections .
the lowest order corrections give : @xmath29 where @xmath30 \label{beta2}\ ] ] and @xmath31 \quad .
\label{beta3}\ ] ] here @xmath32 and @xmath33 are the @xmath4 and @xmath8 gauge coupling constants , respectively ; @xmath34 and @xmath35 are the yukawa couplings between the higgs triplet and doublet , and the matter fermions ; @xmath36 is 3 for quarks and 1 for leptons .
@xmath37 is bounded from below if @xmath38 and @xmath39 we make the simplifying but unnecessary assumption that the full symmetry of the theory ( here @xmath40 ) is restored at high temperature .
this requires @xmath41 spontaneous symmetry breaking is signaled when @xmath42 drops below either @xmath43 or @xmath44 .
now we would like to find the minima of the potential for different temperatures and , in that way , find the symmetry breaking pattern .
there are four different phases that can occur : phase 1 : @xmath45 ; phase 2 : @xmath46 ; phase 3 : @xmath47 ; phase 4 : @xmath48 .
as shown in a similar model with a complex singlet instead of the triplet @xcite , there is a range of parameters in which the different phases occur in succession as we go from very high to low temperatures .
we have explicitly checked the existence of such parameters for the model presented here .
for example , one such set is : @xmath49 @xmath50 where @xmath51 and @xmath52 are measured in units of 246 gev - the electroweak scale . in this case
we find that phases 1 , 2 , 3 and 4 follow in succession as we cool from very high to low temperatures . in fig .
1 we schematically show the vevs of @xmath3 and @xmath7 as functions of temperature .
this then yields the following symmetry breaking pattern : @xmath53 where in the last stage the vev of @xmath7 vanishes and this restores the @xmath21 symmetry .
the group @xmath9 is a subgroup of @xmath4 .
the generator @xmath54 of the electromagnetic group @xmath5 is a linear combination of @xmath55 - the generator of @xmath9 and @xmath56 - the generator of @xmath8 : @xmath57 the first stage of symmetry breaking now produces topological magnetic monopoles in the way they would be produced in the georgi - glashow model @xcite .
the flux emanating from these monopoles is of the @xmath9 magnetic field . at the next symmetry breaking , a fraction of this flux becomes massive and gets squeezed into flux tubes - these are the @xmath11strings attached to the monopoles - while the remaining flux is massless electromagnetic flux - this is the magnetic flux emanating from electroweak monopoles . during this second symmetry breaking a distribution of closed loops of @xmath11string
will also be produced .
note that these strings can break by the formation of monopole pairs but this process is exponentially suppressed since the monopoles are assumed to be very heavy compared to the electroweak scale .
the breaking of the @xmath21 symmetry group in the second stage produces domain walls but these will dissolve in the last stage when the @xmath21 is restored .
this scheme can readily be put in a grand unification model because triplets occur quite naturally .
for example , in minimal @xmath58 grand unification , the breaking to @xmath59 is accomplished with a higgs @xmath60 ; twelve of these are eaten as the @xmath61 and @xmath56 bosons become massive ; of the remaining twelve higgses three are in an @xmath4 triplet that can be identified with the @xmath7 field .
we can now discuss the structure of the defects in this model and the possible cosmological consequences .
the monopoles that are produced in the first phase transition are hedgehogs purely in the @xmath7 field and have a mass : @xmath62 where , @xmath63 is the @xmath64 mass in phase 2 .
when the second symmetry breaking occurs , the monopoles get connected by strings . at this stage
, the @xmath64 boson mass gets a contribution from the vev of @xmath3 and is : @xmath65 therefore the mass of the monopole also changes as given by eqn .
( [ mmass1 ] ) .
the strings are topological since they arise from the breaking of @xmath66 to @xmath5 .
apart from instanton processes resulting in the nucleation of monopole antimonopole pairs along the string @xcite , the strings are stable .
the second symmetry breaking also yields domain walls .
we expect that the angle between the two vectors @xmath7 and @xmath67 varies as we traverse one of these walls though , depending on the parameters of the model , one of the two fields @xmath7 or @xmath3 could vanish in the interior of the wall .
such walls are known to exist in condensed matter systems such as @xmath68he . here
we shall not consider the domain walls in greater detail since they will go away once the @xmath21 symmetry is restored in the last stage . in the last transition occuring at @xmath69 , the vev of @xmath7 vanishes .
( this feature of the model means that it is unconstrained by the experimental limits on the @xmath70 parameter . ) at temperatures below @xmath69 , the strings are nearly identical to the @xmath11strings discussed in the literature .
there is however one difference which could be crucial - the strings can have a bosonic condensate @xcite of @xmath7 on them depending on the parameters of the model . to see this note that @xmath3 vanishes at the center of the string and
so it is energetically favorable for @xmath7 to be non - vanishing in this region .
this means that the region in the vicinity of the strings resembles the phase where both @xmath3 and @xmath7 have non - zero vevs - a phase where the strings are topologically stable .
then there are two parameter dependent possibilities : ( i ) the @xmath0-strings are unstable at temperatures below @xmath69 , and , ( ii ) the @xmath0-strings are stabilized at temperatures below @xmath69 by the @xmath7 condensate
. both of these possibilities may have interesting cosmological consequences .
it is important to remember that the standard model @xmath0-string is superconducting @xcite due to fermion zero modes @xcite and can carry a maximum electric current of about @xmath71 amps . in our case , the fermionic part of the lagrangian in eqn .
( [ lag ] ) is unaffected by the presence of the @xmath7 field and the dirac equations for leptons and quarks are identical to those in the standard model .
therefore zero modes for these particles will also exist in the extended model @xmath11string and these strings too will be able to carry up to @xmath71 amps of electric current . due to the presence of the quark zero modes , linked and
twisted @xmath11strings will carry baryon number exactly as they do in the standard model @xcite .
now let us discuss the first possibility where the monopoles get connected by strings at @xmath72 and then the strings , being unstable , break up and decay by the nucleation of monopole pairs at a temperature below @xmath69 .
this scenario might be interesting in the context of baryogenesis from string decay .
the monopoles that form at @xmath69 are specified by their location , velocity , charge and , most importantly for us , by a @xmath9 phase .
( it is well known that the phase of a single monopole is arbitrary since we are free to perform @xmath9 rotations .
but in a situation where there are two or more monopoles , the _ relative _ phase of the monopoles can not be changed by such a global transformation . ) at @xmath72 , @xmath3 gets a vev and then the @xmath3 field configuration of a monopole and an antimonopole will have the following asymptotic forms : @xmath73 where , @xmath74 and @xmath75 are spherical angular coordinates centered on the monopole and antimonopole respectively and the phase @xmath76 is the arbitrary @xmath77 phase .
these monopoles will get connected by @xmath11strings below @xmath72 and , as shown in @xcite , the monopoles plus string field configuration carries baryon number . at the time the monopoles get connected by strings
, we expect their relative phases ( @xmath76 ) to be uncorrelated and so each connected monopole pair will carry of order @xmath78 baryon number .
therefore the magnitude of the local baryon number density at @xmath72 is of the order of the monopole density : @xmath79 if the strings are unstable at low temperatures , they will break up into small segments below @xmath80 and these small segments will ultimately release their energy into radiation . as this decay happens relatively late
, the strings are out of thermal equilibrium and each decay leads to the generation of baryon ( or antibaryon ) number .
a net amount of baryon number will be produced if the cp violation effects of the km matrix are included in the string decay .
the strength of the cp violation in the monopole - string system is not yet known but , as the effect is non - perturbative , it need not be small .
this subject deserves further investigation .
a more exciting possibility is that the strings are stable even at low temperatures due to the @xmath7 condensate @xcite .
then the strings do not break up into small segments below @xmath80 .
the length distribution of strings is expected to be exponentially suppressed since on following a particular string - treated as a brownian walk - there is a certain non - vanishing probability that it will encounter a monopole and terminate .
but the length scale above which the exponential fall off sets in can be quite large since it is governed by the monopole number density and the string density at @xmath72 .
if the monopoles are quite dilute when the strings form , we expect a nearly scale invariant distribution of loops to be formed together with a few open string segments @xcite .
the evolution of a superconducting string tangle has been discussed in several contexts in @xcite and the results are reviewed in @xcite .
the short string segments and loops will collapse and decay into radiation with the production of baryons .
but , in the early universe , the longer string segments which carry currents will be frozen in the ambient plasma provided their curvature scale is larger than a critical scale .
such strings will be stretched with the hubble expansion and will become even longer .
on curvature scales less than the critical scale , the strings are free to oscillate
. such sections of strings will cut across the random magnetic field that is also expected to be produced at @xmath72 @xcite and will result in an oscillating induced current on the string @xcite . by the epoch of structure formation , most of the strings would have decayed but we can expect a few stray , frozen - in loops to survive .
these stray loops would then get trapped in protogalaxies and in protostars .
the ensuing galactic and stellar turbulence would generate more string in a way reminiscent of the generation of vortex lines in superfluids .
( there too a few stray lines result in a fully developed tangle in the presence of flow @xcite . )
the energy emission from superconducting strings ( formed at the electroweak scale ) in the galaxy was discussed in @xcite and it was found that the luminosity due to all the strings would be @xmath81 ergs / s .
this is significant but a characteristic signature of radiation from strings would further assist in observations .
such a characteristic seems possible in the case of electroweak strings due to the role that @xmath11strings can play in baryon number violating processes .
consider , for example , the scenario where a tangle of electroweak strings is present in a galaxy .
( similar phenomena would also occur in stars . ) on curvature scales smaller than @xmath82 cm , the strings are not frozen into the galactic plasma .
loops on this scale will cut across the galactic magnetic field and collapse on a time scale @xmath83 sec where @xmath84 cm / s is the terminal velocity of the strings moving through the galactic plasma @xcite . during this process
an electric field is induced along the string and hence @xmath85 .
the baryon number anomaly equation then gives us the rate of baryon number production during loop collapse @xcite .
order of magnitude estimates tell us that such a collapsing loop releases about @xmath86 baryons where @xmath87 is the magnetic flux that the string cuts across and @xmath88 is the @xmath11flux in the string .
while the baryon number production due to a single loop is quite a slow process , the baryon number produced by all the loops in the galaxy can be significant - about @xmath89 baryons in @xmath90 secs .
roughly half the baryons will be in the form of antimatter and since the strings are neutral , would be in the form of antineutrons which then decay into antiprotons , electrons and neutrinos
. this fact may eventually be useful in the search for electroweak strings in our galaxy @xcite . |
most of the solar system phenomena such as the mercury precession can be successfully explained in the context of einstein gravity , but when we intend to study beyond the solar system or strong gravity regimes , this theory encounters some problems .
accelerated expansion of our universe is one of the main challenging subjects of cosmologists perlmutter1,perlmutter2,perlmutter3 , whereas einstein gravity can not properly explain it . in order to interpret this expansion ,
some various candidates have been proposed by many authors . among various alternatives ,
one may point out the modified gravities like lovelock gravity lovelock1,lovelock2,dehghani1,dehghani2 , brane world cosmology brax1,brax2,brax3 , scalar - tensor theories brans1,brans2,brans3,brans4,brans5,harada1,harada2,harada3,harada4,harada5 and @xmath5 gravity souza1,souza2,souza3,souza4,souza5,souza6,cooney1,cooney2,cooney3,cooney4,cooney5 . from the cosmological point of view , the dark energy is an unknown form of energy which is hypothesized to permeate the spacetime , tending to accelerate the expansion of the universe @xcite .
the dark energy is the most accepted hypothesis to explain the observations since the 1990s indicating that the universe is expanding at an accelerating rate .
another model of these interesting candidates is related to the cosmological constant . in this theory , the cosmological constant may add to the einstein lagrangian @xcite for explaining such acceleration .
the cosmological constant is the simplest possible form of dark energy , which leads to the current standard model of cosmology known as the @xmath6-cdm model and provides a good fit to many cosmological observations . from the astrophysical point of view , stars reach the equilibrium state due to the balance between gravitational force and internal pressure . therefore , in order to study the structure of stars , we should obtain a suitable hydrostatic equilibrium equation ( hee ) and then we can study the structure properties of stars .
the neutron and quark stars are in the category of celestial objects with high mass , high density and small radius , and therefore , they are called compact objects . due to this fact , we need to take into account the effects of general relativity ( spacetime curvature ) for investigating their structures . the first hee of stars in @xmath3-dimensional einstein gravity has been studied by tolman , oppenheimer and volkoff ( tov ) @xcite .
also , the physical characteristics of stars using tov equation have been investigated by many authors silbar1,silbar2,silbar3,silbar4,silbar5,silbar6,silbar7,silbar8 .
recently , survey of compact objects in different gravity has attracted the attention of researchers , in particular , the models of modified gravity which involve higher curvature terms or a scalar field @xcite . on the other hand ,
if one is interested to study the structure and evolution of stars in different theories of gravity with various sources , the new hee must be obtained .
therefore , in recent years , the generalizations and modifications of tov equation have been investigated in various theories of gravity , such as : dilaton gravity @xcite , gravity s rainbow @xcite , massive gravity @xcite , @xmath7 and @xmath8 gravities astashenok2015aastashenok2015b , momeni1,momeni2,momeni3,momeni4 and p ost - newtonian theory @xcite ( see also refs .
boyadjiev , vinayaraj1,vinayaraj2,vinayaraj3,astashenok2013a , astashenok2013b , astashenok2013c , astashenok2013d , lemos1,lemos2,lemos3,lemos4,green1,green2 for more details ) .
the outline of our paper is as follows . in the next section ,
we consider a spherical symmetric metric and obtain the hee in the einstein-@xmath6 gravity for arbitrary dimensions ( @xmath9 ) .
then , we investigate stability and energy conditions for the equation of state of neutron star matter in @xmath3-dimensions . also , we study effects the cosmological constant on maximum mass and its radius of the neutron stars .
next , we examine effects of the cosmological constant on other properties of neutron stars such as ; schwarzschild radius , average density , compactness , buchdahl - bondi bound and dynamical stability .
we finish our paper with some closing remarks .
as first step , we regard @xmath0dimensional einstein@xmath10 gravity to obtain its related hee . the action of the einstein gravity with the cosmological constant in arbitrary dimensions , @xmath11 , is given by @xmath12where @xmath13 and @xmath1 are , respectively , the ricci scalar and the cosmological constant in @xmath0dimensions and @xmath14 is the action of matter field . varying the action ( [ actionen ] ) with respect to the metric tensor @xmath15 , the equation of motion
can be written as @xmath16where @xmath17 , @xmath18 is the gravitational constant in @xmath11-dimensions and define as @xmath19 , where @xmath20 denotes the four dimensional gravitational constant and @xmath21 is the volume of extra dimensions . also , the constant @xmath22 is the speed of light in the vacuum .
in addition , @xmath23 and @xmath24 are the symmetric ricci tensor and energy - momentum tensors , respectively . here
, we want to obtain the static solutions of eq .
( [ eqen ] ) . for this purpose , we assume a spherical symmetric spacetime in the following form @xmath25where @xmath26 and @xmath27 are the unknown metric functions of radial coordinates and @xmath28 is the line element of unit @xmath29-dimensional sphere @xmath30 on the other hand , the energy - momentum tensor for a perfect fluid is @xmath31where @xmath32 and @xmath33 are density and pressure of the fluid which are measured by local observer , respectively , and @xmath34 is the fluid four - velocity . using eqs .
( [ eqen ] ) and ( [ emtensoren ] ) and the metric introduced in eq .
( [ metric ] ) , we can obtain the components of energy - momentum for @xmath11-dimensions as follows @xmath35 we consider the metric ( [ metric ] ) and eq .
( [ 4dim ] ) for the perfect fluid and obtain the components of eq .
( [ eqen ] ) in the following forms @xmath36 f , \label{33}\end{aligned}\]]where @xmath37 and also , @xmath38 , @xmath39 , @xmath32 and @xmath33 are functions of @xmath40 .
we note that the prime and double prime are , respectively , the first and second derivatives with respect to @xmath41 . using eqs .
( [ 1 ] ) - ( [ 33 ] ) and after some calculations , we obtain @xmath42 now , we obtain @xmath43 from eq .
( [ 2 ] ) as follows @xmath44 f}{rgd_{2}}. \label{diff(r)}\ ] ] then , we calculate @xmath27 by using eq .
( [ 1 ] ) with the following form @xmath45where @xmath46 and @xmath47 is the gamma function , in which @xmath48 , @xmath49 and @xmath50 . by considering eqs .
( [ diff(r ) ] ) and ( [ 4g(r ) ] ) , and inserting them in eq .
( [ extraeq ] ) , we can extract the hee of einstein-@xmath6 gravity in @xmath11-dimensions as @xmath51 } { r\left [ -\lambda r^{d_{1}}+d_{1}\left ( \frac{\gamma ( \frac{d_{1}}{2% } ) } { 2\pi ^{^{d_{1}/2}}}c^{2}k_{d}m-\frac{d_{2}}{2}r^{d_{3}}\right ) \right ] } % \left ( c^{2}\epsilon + p\right ) , \label{tov}\]]where for @xmath3-dimensional case ( @xmath52 ) , eq .
( [ tov ] ) reduces to the following equation @xmath53 } { c^{2}r\left [ 6gm - c^{2}r\left ( \lambda r^{2}+3\right ) % \right ] } \left ( c^{2}\epsilon + p\right ) .
\label{4tov}\ ] ] also , eq . ( [ 4tov ] ) leads to primitive tov equation for vanishing @xmath54 @xcite .
it is evident that for @xmath55 , eq . ( [ tov ] ) reduces to the hee obtained by diaz @xcite . as a result
, we have obtained a global form of hee for arbitrary dimensions ( @xmath9 ) in eq .
( [ tov ] ) .
obtaining equation of state for neutron stars enables one to study its properties .
the interior region of the neutron star is a mixture of neutrons , protons , electrons and muons .
these constituents are in charge neutrality and beta equilibrium conditions ( beta - stable matter ) shapiro .
recently , the microscopic constrained variational calculations based on the cluster expansion of the energy functional has been employed to obtain the equation of state of neutron star matter bordbar1,bordbar2,hendi2015 .
the two - nucleon potentials which are used in these studies are the modern argonne av@xmath2 @xcite and charged dependent reid-@xmath56 @xcite .
one of the advantages of these methods is that there is no need for any free parameter in formalism , and they result into a good convergence . also , the results in this method are more accurate comparing to other semi - empirical parabolic approximation methods .
this is due to a microscopic computation of asymmetry energy which is carried on for the asymmetric nuclear matter calculations .
in fact , it is necessary to have a microscopic calculation with the modern nucleon - nucleon potentials which is isospin projection ( @xmath57 ) dependant @xcite . here , we are interested in using the lowest order constrained variational ( locv ) method with the av@xmath2 potential @xcite for calculating the modern equation of state for neutron star matter and investigating some physical properties of neutron star structure .
we use the locv method to calculate the energy of our system .
the locv method is a useful tool for determination of the properties of neutron , nuclear and asymmetric nuclear matter at finite and zero temperature modarres , modarresi , howes , owen , modarresii .
it is notable that , the locv method is a fully self - consistent formalism and it does not bring any free parameters into calculation .
it employs a normalization constraint to keep the higher order term as small as possible @xcite .
the functional minimization procedure represents an enormous computational simplification over unconstrained methods that attempt to go beyond lowest order .
a suitable trial many - body wave function is @xmath58where @xmath59 is the uncorrelated ground - state wave function of @xmath60 independent neutrons , and @xmath61 is a proper @xmath60-body correlation function .
applying jastrow approximation @xcite , @xmath61 is rewritten by @xmath62where @xmath63 is a symmetrizing operator .
also , we consider a cluster expansion of the energy functional up to the two - body term@xmath64)=\frac{1}{n}\frac{\langle \psi |h|\psi \rangle } { \langle \psi |\psi \rangle } = e_{1}+e_{2},\ ] ] the energy per particle up to the two - body term is @xmath64)=e_{1}+e_{2},\]]where @xmath65 and @xmath66 are one - body and two - body energy term , respectively .
the operator @xmath67 has been given in ref .
bordbar044310 ( see refs .
@xcite for more details ) .
our obtained equation of state of neutron star matter is shown in fig .
[ fig1 ] .
@xmath68 in order to obtain more details for equation of state of neutron star matter introduced in this paper , we study both stability and energy conditions in the following subsections . in order to investigate the stability of the introduced equation of state for a physically acceptable model
, one expects that , the speed of sound ( @xmath69 ) be less than the speed of light ( @xmath22 ) @xcite .
this stability condition can be written in the following explicit form @xmath70 by using the above condition and fiq .
[ fig1 ] , and comparing them with diagrams related to sound speed - radius relationship plotted in ref .
bordbar2006 ( see fig . 2 in ref .
@xcite , for more details ) , it is evident that this equation of state satisfies the inequality @xmath71 .
therefore , the introduced equation of state is stable .
one of interesting properties of the mentioned equations of state is related to its central density .
it is notable that the obtained central density of this equations of state is about @xmath72 and this density is larger than the normal nuclear density @xmath73 @xcite .
now , we want to study the energy conditions . in this regard
, we investigate null energy condition ( nec ) , weak energy condition ( wec ) , strong energy condition ( sec ) and dominant energy condition ( dec ) at the center of the neutron stars .
these energy conditions can be written in the following explicit forms @xmath74 where @xmath75 and @xmath76 are the pressure and density at the center of neutron star ( @xmath77 ) , respectively .
considering fig .
[ fig1 ] and the above conditions ( [ 11])-([44 ] ) , our results are presented in table [ tab11 ] .
[ cols="^,^,^,^,^,^",options="header " , ] based on table [ tab3 ] , considering the cosmological constant as a positive quantity , the maximum mass of neutron star decreases as @xmath6 increases .
in other words , the maximum mass of neutron stars with the positive values of cosmological constant are in the range @xmath78 . also , when we consider the cosmological constant to be very small ( @xmath79 ) , the maximum mass reduces to results obtained in the einstein gravity @xcite . in other words , for @xmath80 this constant does not affect the maximum mass and other properties .
so , considering the obtained value for the cosmological constant from the cosmological perspective ( @xmath81 ) , this quantity does not affect the structure of neutron stars .
we plot diagrams related to the maximum mass versus central mass density ( radius ) in figs .
[ fig2 ] and [ fig3 ] .
the variation of maximum mass versus the cosmological constant is also shown in fig .
[ fig4 ] for @xmath82 .
@xmath83 @xmath84 as we can see , mass of neutron star versus radius changes when the @xmath6 increases .
indeed , one of interesting results of this paper is related to change in diagrams of mass versus radius , so that by increasing @xmath1 , these diagrams change from ordinary shapes of the neutron stars to those of quark stars ( see fig .
[ fig3 ] ) . in other words , by increasing @xmath1 , we encounter with a change of state ( phase transition ) of neutron star to the quasi quark star .
on the other hand , we consider negative values of the cosmological constant and plot the density and pressure versus radius of neutron star .
as one can see , the maximum and minimum values belong to the center and surface of neutron star , respectively ( see figs . [ fig44 ] and [ fig444 ] ) , but the behavior of diagrams related to mass - radius relation of these stars are not logical ( see fig .
[ fig5 ] ) . in other words , einstein gravity with the negative values of the cosmological constant are not a suitable theory for explaining the neutron stars with the mentioned equation of state .
our results show that , by decreasing the cosmological constant ( @xmath85 ) , the maximum mass and also the radius of this star are not modified .
in other words , the cosmological constant does not significantly affect the maximum and radius of neutron stars for @xmath86 . as a result of this paper , when the value of the cosmological constant is about @xmath87 , this constant
does not play a sensible role in the structure of neutron stars , but by considering larger values for it ( about @xmath88 ) , the maximum mass and its radius are modified ( reduced ) .
@xmath89 @xmath90 @xmath91 @xmath92 in the following , we investigate other properties of neutron stars by considering the cosmological constant such as ; the schwarzschild radius , average density , compactness , buchdahl - bondi bound and dynamical stability .
now , we obtain the schwarzschild radius for the obtained masses in einstein-@xmath54 gravity in the following form @xmath93 ^{1/3}}{\lambda c}-\frac{c}{\left [ \left ( 3gm+% \sqrt{\frac{c^{4}}{\lambda } + 9g^{2}m^{2}}\right ) \lambda ^{2}c\right ] ^{1/3}}% , \]]where @xmath94 is the schwarzschild radius , @xmath95 @xmath96 and @xmath97 @xmath98 .
it is notable that , by applying the cosmological constant to the einstein gravity , the schwarzschild radius modified and also , considering @xmath99 in eq .
( [ rsch4lambda ] ) , it reduces to the schwarzschild radius in einstein gravity as schwarzschild @xmath100 to find the schwarzschild radius of neutron stars , we use of eq .
( rsch4lambda ) .
the results show that increasing the cosmological constant and decreasing the maximum mass of neutron stars , the schwarzschild radius decreases ( see table [ tab3 ] for more details ) .
the average density of the neutron stars in @xmath101dimensions can be written as the following form @xmath102where the results for various the cosmological constant are presented in table [ tab3 ] .
the compactness of a spherical object may be defined as the ratio of the schwarzschild radius to the radius of object as @xmath103which may be indicated as the strength of gravity .
we obtain the values of @xmath104 in the einstein-@xmath6 gravity and find that increasing @xmath6 leads to decreasing the strength of gravity .
using the equation ( [ 4g(r ) ] ) , we obtain the gravitational redshift in @xmath3-dimensions for this gravity as @xmath105 as one can see , decreasing @xmath6 leads to increasing the gravitational redshift ( see table [ tab3 ] ) .
the maximum value of the gravitational redshift is about @xmath106 .
according to table [ tab3 ] , one finds variation of @xmath1 changes the gravitational redshift , drastically .
this is a way for adjusting the values of @xmath1 based on observational evidences . according to buchdahl - bondi bound @xcite , for a static spherically symmetric perfect fluid sphere , maximum permitted @xmath107 ratio should be @xmath108 . for the einstein-@xmath6 gravity
this bound modified as @xcite @xmath109where @xmath110 . using the above condition
, we investigate the stability condition of neutron stars in einstein-@xmath6 gravity .
the calculations are presented in table [ tab3 ] .
the results show that obtained maximum mass of neutron stars for all values obtained in table tab3 are stable .
the dynamical stability of the stellar model against the infinitesimal radial adiabatic perturbation was introduced by chandrasekhar in ref .
next , this stability condition was developed and applied to astrophysical cases by many authors bardeen1,bardeen2,bardeen3,bardeen4 .
the adiabatic index ( @xmath111 ) is defined as @xmath112 it is notable that , in order to have the dynamical stability , @xmath111 should be larger than @xmath113 ( @xmath114 ) everywhere within the isotropic stars .
so , we plot a diagram related to @xmath111 versus radius for different values of the cosmological constant in fig .
adiabatici . as
one can see , this stellar model is stable against the radial adiabatic infinitesimal perturbations . on the other hand , in order to present more investigations for the inside of neutron stars in einstein-@xmath6 gravity , we plot density and pressure versus radius of this star in figs . [ density ] and
[ pressure ] .
these figures show that density ( pressure ) values have a maximum and a minimum in the center and surface of neutron star , respectively , as expected .
@xmath115 @xmath116 @xmath117
in the first part of this paper , we have considered a spherical symmetric metric and extracted the hydrostatic equilibrium equation of stars for @xmath0dimensional einstein-@xmath1 gravity ( @xmath9 ) .
then , we have investigated the maximum mass of neutron stars by using the modern equations of state of neutron star matter derived from microscopic calculations .
the results showed that the cosmological constant affects the maximum mass of neutron stars .
in other words , considering the @xmath118 , the maximum mass for neutron stars decreases when @xmath6 increases . in this case ,
maximum mass of neutron stars is in the range @xmath4 .
one of the interesting results obtained in this paper is the behavior of diagrams of mass versus radius of the neutron stars . by increasing @xmath1
, these diagrams showed a transition from the neutron star to the quasi quark star . on the other hand , considering @xmath119 , the behavior of diagrams related to mass versus radius ( central mass density ) of neutron stars is not logical , so the einstein gravity with the negative values of cosmological constant can not explain the existence of neutron stars . in order to investigate the interior of neutron stars in einstein-@xmath6 gravity in more details ,
we have plotted density and pressure versus radius of this star .
the results showed that the density and pressure were maximum and minimum of values at the central and surface of neutron stars , respectively .
our results showed that , when the approximate value of the cosmological constant was @xmath87 , this constant did not play a sensitive role in the structure of neutron stars , but by taking larger values for it ( about @xmath88 ) , the maximum mass and its radius were reduced .
another interesting result of this paper is related to the effect of cosmological constant on the so - called gravity strength .
we have shown that when the positive values of the cosmological constant increases , the strength of gravity decreases , and this effect leads to decreasing the maximum mass of neutron stars .
finally , we have studied the stability condition for these stars using buchdahl - bondi bound and found that the obtained maximum mass of neutron stars for all values obtained in table [ tab3 ] were stable .
then , considering adiabatic index , we showed that these stars were stable against the radial adiabatic infinitesimal perturbations .
we would like to thank the anonymous referee for his / her valuable comments .
we also wish to thank shiraz university research council .
ghb wishes to thank the center for excellence in astronomy and astrophysics ( ceaa - riaam ) for financial support .
be acknowledges s. panahiyan for helpful discussions .
this work has been supported financially by research institute for astronomy and astrophysics of maragha . | regarding a @xmath0dimensional spherically symmetric line element in the context of einstein-@xmath1 gravity , the hydrostatic equilibrium equation of stars is obtained .
then , by using the lowest order constrained variational ( locv ) method with the av@xmath2 potential and employing microscopic many body calculations in the modern equation of state , the structure properties of neutron stars are investigated .
regardless of cosmological point of view and considering arbitrary positive and negative values of the cosmological constant , the maximum mass of the neutron stars and their corresponding radius in @xmath3-dimensions are computed .
the results show that there is an upper limit for the maximum mass of neutron star for positive cosmological constant ( @xmath4 ) . on the other hand
, it is shown that the einstein gravity can not explain the structure of neutron star with negative @xmath1 .
other properties of neutron stars such as ; the schwarzschild radius , average density , compactness and buchdahl- bondi bound are studied .
in addition , by using the buchdahl - bondi bound for neutron stars , stability of these stars is investigated .
finally , the dynamical stability is investigated and shown that the neutron stars follow the dynamical stability in this gravity . |
we begin by defining the random process that will be the object of our study .
let @xmath19 be i.i.d .
random variables taking the values @xmath1 , @xmath2 and @xmath3 with probability @xmath4 , @xmath20 and @xmath21 , respectively , with @xmath8 and @xmath22 $ ] .
let @xmath23 with @xmath24 be the corresponding _ random walk _ on @xmath25 , defined by @xmath26 that is , @xmath27 is the step at time @xmath28 and @xmath29 is the position at time @xmath28 . let @xmath30 be i.i.d .
random variables taking the values @xmath31 ( black ) and @xmath32 ( white ) with probability @xmath9 each
. we will refer to @xmath33 as the _ random coloring _ of @xmath0 , that is , @xmath34 is the color of site @xmath35 . the pair @xmath36 is referred to as the _ random walk in random scenery _ associated with @xmath37 and @xmath33 .
let @xmath38 be the sequence of colors observed along the walk .
we will refer to @xmath39 as the _ random color record_. this random process , which takes values in the set @xmath40 and has full support on @xmath41 , will be our main object of study . because the walk may return to sites it has visited before and see the same color , @xmath39 has intricate dependencies .
an overview of the ergodic properties of @xmath39 is given in @xcite .
we will use the symbol @xmath42 to denote the joint probability law of @xmath37 and @xmath33 .
the question that we will address in this paper is whether or not there exists a version @xmath43 of the conditional probability @xmath44 such that the map @xmath45 is everywhere continuous on @xmath41 .
it will turn out that the answer depends on the choice of @xmath46 and @xmath4 . in @xcite , we considered the pair @xmath47 and identified the structure of the set of points of discontinuity for the analogue of the conditional probability in the last display .
however , @xmath47 is much easier to analyze than @xmath39 , because knowledge of @xmath37 and @xmath39 fixes the coloring on the support of @xmath37 .
consequently , the structure of the set of points of discontinuity for @xmath47 is very different from that for @xmath39 .
the same continuity question arises for the two - sided version of @xmath39 where time is indexed by @xmath0 , that is , the random walk is extended to negative times by putting @xmath48 and @xmath49 , @xmath50 , with @xmath27 the step at time @xmath50 . in the present paper , we will restrict ourselves to the one - sided version .
the continuity question has been addressed in the literature for a variety of random processes .
typical examples include gibbs random fields that are subjected to some transformation , such as projection onto a lower - dimensional subspace or evolution under a random dynamics .
it turns out that even simple transformations can create discontinuities and thereby destroy the gibbs property . for a recent overview , see @xcite .
our main result , described in section [ s1.4 ] below , is a contribution to this area . in this section
, we view the conditional probability distribution of @xmath51 given @xmath52 as a map from @xmath53 to the set of probability measures on @xmath54 ( as opposed to a map from @xmath41 to this set ) .
our question about continuity of conditional probabilities will be formulated in terms of so - called _
bad configurations_. [ badconf ] let @xmath42 denote any probability measure on @xmath41 with full support .
a configuration @xmath55 is said to be a bad configuration if there is a @xmath56 such that for all @xmath57 there are @xmath58 and @xmath59 , with @xmath60 and @xmath61 on @xmath62 , such that @xmath63 in words , a configuration @xmath64 is bad when , no matter how large we take @xmath65 , by tampering with @xmath64 inside @xmath66 for some @xmath67 while keeping it fixed inside @xmath62 , we can affect the conditional probability distribution of @xmath51 in a nontrivial way .
typically , @xmath68 depends on @xmath64 , while @xmath28 depends on @xmath65 . a configuration that is not bad
is called a _
good configuration_. the bad configurations are the discontinuity points of the conditional probability distribution of @xmath51 , as made precise by the following proposition ( see @xcite , proposition 6 , and @xcite , theorem 1.2 ) .
[ bad = disc ] let @xmath69 denote the set of bad configurations for @xmath51 . for any version @xmath43 of the conditional probability @xmath70 , the set @xmath69 is contained in the set of discontinuity points for the map @xmath71 .
there is a version @xmath43 of the conditional probability @xmath70 such that @xmath69 is equal to the set of discontinuity points for the map @xmath45 . for the random color record , a _ naive guess _
is that all configurations are bad when @xmath14 because the random walk is recurrent , while all configurations are good when @xmath72 $ ] because the random walk is transient .
indeed , in the recurrent case we obtain new information about @xmath51 at infinitely many times , corresponding to the return times of the random walk to the origin , while in the transient case no such information is obtained after a finite time .
however , we will see that this naive guess is wrong .
before we state our main result , let us make an _ educated guess _ : * ( eg1 ) @xmath73 \
\forall \eps\in[0,1 ) \dvtx \b=\omega$ ] . *
( eg2 ) @xmath74 \
\forall \eps\in[0,1 ) \dvtx \b=\varnothing$ ] .
the explanation behind this is as follows .
_ fully biased .
_ suppose that @xmath75 .
then @xmath76 where we use that , for any @xmath46 and @xmath77 , @xmath78 and @xmath79 are independent . hence , the color seen at time @xmath1 only depends on the color seen at time @xmath80 , so that @xmath81 . (
note that if @xmath10 , then @xmath39 is i.i.d . )
_ monotonicity .
_ for fixed @xmath4 , we expect monotonicity in @xmath46 : if a configuration is bad for some @xmath82 , then it should be bad for all @xmath83 also . intuitively , the random walk with parameters @xmath84 is exponentially more likely to return to @xmath1 after time @xmath65 than the random walk with parameters @xmath12 , and therefore we expect that it is easier to affect the color at @xmath1 for @xmath84 than for @xmath12 .
_ critical value .
_ for a configuration to be good , we expect that the random walk must have a strictly positive speed conditional on the color record .
indeed , only then do we expect that it is exponentially unlikely to influence the color at @xmath1 by changing the color record after time @xmath65 .
to compute the threshold value for @xmath46 above which the random walk has a strictly positive speed , let us consider the monochromatic configuration `` all black . ''
the probability for the random walk with parameters @xmath12 to behave up to time @xmath28 like a random walk with parameters @xmath85 , with @xmath86 $ ] and @xmath87 , is @xmath88 where @xmath89\end{aligned}\ ] ] is the relative entropy of the step distribution @xmath85 with respect to the step distribution @xmath12 .
the probability for the random coloring to be black all the way up to site @xmath90 is @xmath91 the total probability is therefore @xmath92 the question is : for fixed @xmath12 and @xmath93 , does the lowest cost occur for @xmath94 or for @xmath95 ?
now , it is easily checked that @xmath96 is strictly convex and has a derivative at @xmath94 that is strictly positive if and only if @xmath97 , irrespective of the value of @xmath4 and @xmath68 .
hence , zero drift has the lowest cost when @xmath98 $ ] , while strictly positive drift has the lowest cost when @xmath99 $ ] .
this explains ( eg1 ) and ( eg2 ) .
( 10,10)(0,-2 ) ( 0,0)(10,0)10 ( 0,0)(0,10)10 ( 5,0)(5,4)(5,8 ) ( 0,8)(4,8)(8,8 ) ( -2,4)(1,0)4 ( -2,4)(1,-1)3.8 ( -2,4)(1,1)1.8 ( 10.5,4)(-1,0)4 ( 10.5,4)(-1,-1)3.8 ( 10.5,4)(-1,1)2.2 ( 5,11)(0,-1)2 ( -4,2)(2,-1)3.5 ( 11,3.7)@xmath81 ( -5.8,3.7)@xmath100 ( -9.7,1.6)@xmath101 ( 4.75,11.5 ) ? ( 10.5,-0.2)@xmath46 ( -0.2,10.5)@xmath77 ( -0.25,-1.3)@xmath9 ( 4.75,-1.3)@xmath16 ( 7.8,-1.3)@xmath80 ( -1,7.8)@xmath80 ( -1,-0.6)@xmath1 ( 0,0 ) ( 8,0)(8,4)(8,7.8 ) ( 0,0)(2,0)(4.8,0 ) ( 5.2,0)(7,0)(8,0 ) ( 0,0)(0,4)(0,7.8 ) we are now ready to state our main result and compare it with the educated guess made in section [ s1.3 ] ( see figure [ figbad ] ) .
[ phtr ] ( ) there exists a neighborhood of @xmath13 in the @xmath102-plane for which @xmath81 .
this neighborhood can be taken to contain the line segment @xmath103 \times\{0\}$ ] with @xmath104 .
( ) if @xmath105 and @xmath10 , then @xmath106 .
( ) if @xmath14 and @xmath10 , then @xmath107 .
theorem [ phtr](ii ) and ( iii ) prove ( eg1 ) for @xmath108 and @xmath10 , except for @xmath14 and @xmath10 , where ( eg1 ) fails .
we will see that this failure comes from parity restrictions .
theorem [ phtr](i ) proves ( eg2 ) in a neighborhood of @xmath13 in the @xmath102-plane .
we already have seen that @xmath81 when @xmath75 and @xmath8 .
note that theorem [ phtr](ii ) and ( iii ) disprove monotonicity in @xmath46 for @xmath10 .
we believe this monotonicity to fail only at @xmath14 and @xmath10 . to appreciate why in theorem [ phtr](i ) we are not able to prove the full range of ( eg2 ) , note that to prove that a configuration is good
we must show that the color at @xmath1 can not be affected by _ any _ tampering of the color record far away from @xmath1 .
in contrast , to prove that a configuration is bad it suffices to exhibit _ just two _ tamperings that affect the color at @xmath1 .
in essence , the conditions on @xmath46 and @xmath4 in theorem [ phtr](i ) guarantee that the random walk has such a large drift that it moves away from the origin no matter what the color record is .
we close with ( see figure [ figbad ] ) the following conjecture .
[ phtrconj ] is true .
theorem [ phtr ] is proved in sections [ s2][s5 ] : ( i ) in section [ s2 ] , ( ii ) in section [ s4 ] and ( iii ) in section [ s5 ] .
it seems that for @xmath108 and @xmath109 the argument needed to prove that all configurations are bad is much more involved . in section [ s3 ] , we suggest an approach to handle this problem , which will be pursued in future work . the examples alluded to at the end of section [ s1.1 ] typically have both good and bad configurations . on the other hand , we believe that our process @xmath39 has all good or all bad configurations , except at the point @xmath110 and possibly on the line segment @xmath111 .
a simple example with such a dichotomy , due to rob van den berg , is the following .
let @xmath112 be an i.i.d .
@xmath113-valued process with the 1 s having density @xmath114 .
let @xmath115 , @xmath50 . clearly , if @xmath14 , then @xmath116 is also i.i.d . , and hence all configurations are good . however , if @xmath117 , then it is straightforward to show that all configurations are bad .
see @xcite , proposition 3.3 .
in this section , we prove theorem [ phtr](i ) .
the proof is based on lemmas [ l1][l3 ] in section [ s2.1 ] , which are proved in sections [ s2.2][s2.4 ] , respectively . a key ingredient of these lemmas is control of the _ cut times _ for the walk , that is , times at which the past and the future of the walk have disjoint supports . throughout the paper , we abbreviate @xmath118 for @xmath119 with @xmath120 . for @xmath121 with @xmath120 , abbreviate @xmath122 the main ingredient in the proof of theorem [ phtr](i )
will be an estimate of the number of cut times along @xmath123 .
[ cuttimes ] for @xmath58 , a time @xmath124 with @xmath125 is a cut time for @xmath123 if and only if @xmath126 this definition takes into account only cut times corresponding to locations on or to the right of the origin .
let @xmath127 denote the set of cut times for @xmath123 .
our first lemma reads as follows .
[ l1 ] for @xmath124 , let @xmath128 be any event in the @xmath129-algebra of the walk and the color record up to time @xmath130 .
then @xmath131 for all @xmath132 with @xmath133 and all @xmath134 such that @xmath135 .
we next define @xmath136 our second and third lemma read as follows .
[ l2 ] if @xmath137 , then @xmath81 .
[ l3 ] @xmath138 for @xmath12 in a neighborhood of @xmath13 containing the line segment @xmath103 \times\{0\}$ ] .
note that lemma [ l3 ] yields the exponential decay of @xmath139 , which is much more than is needed in lemma [ l2 ] .
note that lemmas [ l2 ] and [ l3 ] imply theorem [ phtr](i ) .
lemma [ l1 ] states that , conditioned on the occurrence of a cut time at time @xmath130 , the color record after time @xmath130 does not affect the probability of any event that is fully determined by the walk and the color record up to time @xmath130 .
lemma [ l2 ] gives the following sufficient criterion for the nonexistence of bad configurations : for any set of times up to time @xmath65 of cardinality at least @xmath140 , the probability that the walk up to time @xmath141 has no cut times in this set , even when conditioned on the color record up to time @xmath28 , decays faster than @xmath142 as @xmath143 , uniformly in @xmath28 and in the color record that is being conditioned on .
lemma [ l3 ] states that for @xmath46 and @xmath4 in the appropriate range , the above criterion is satisfied .
a key formula in the proof of lemmas [ l1][l3 ] is the following .
let @xmath144 denote the range of @xmath145 ( i.e. , the cardinality of its support ) , and write @xmath146 to denote that @xmath145 and @xmath147 are _ compatible _
( i.e. , there exists a coloring of @xmath0 for which @xmath145 generates @xmath147 ) .
below we abbreviate @xmath148 by @xmath149 .
[ decomp ] for all @xmath58 , @xmath150 the factor @xmath151 arises because if @xmath146 , then @xmath147 fixes the coloring on the support of @xmath145 .
write @xmath152 with ( use proposition [ decomp ] ) @xmath153 abbreviate @xmath154 for @xmath155 , etc .
note that if @xmath156 , then we have @xmath157 and @xmath158 .
it follows that @xmath159 with ( shift @xmath160 back to the origin ) @xmath161.\ ] ] likewise , we have @xmath162 the common factor @xmath163 cancels out and so @xmath164 only depends on @xmath165 . therefore , as long as @xmath135 , we have the equality in ( [ ekids ] ) . since @xmath166 for all large @xmath65 , we will assume that all the values of @xmath65 arising in the proof below satisfy this . for @xmath58 and @xmath147 and @xmath167 ,
define @xmath168 we will show that if @xmath169 , then latexmath:[\[\label{mainest } \lim_{m\to\infty } \sup_{n\geq m } \mathop{\mathop{\max } _ { { y_1^n,\bar y_1^n } } } _ { { y_1^{m-1}=\bar y_1^{m-1 } } } @xmath81 by definition [ badconf ] .
in what follows , we @xmath171 and abbreviate @xmath172 .
define @xmath173 using lemma [ l1 ] , we will show that latexmath:[\[\label{stringimpl } choice of @xmath147 and @xmath167 subject to ( [ fixy ] ) ( with the corresponding @xmath175 and @xmath176 ) . together with @xmath137 , ( [ stringimplagain ] ) will prove lemma [ l2 ] .
write @xmath177 .
we will show that @xmath178 and @xmath179 are incompatible .
indeed , by the definition of @xmath175 , we have @xmath180 define @xmath181 , @xmath182 , with the convention that @xmath183 .
estimate , writing @xmath184 to denote the first cut time for @xmath123 in @xmath31 , @xmath185\\ & & \qquad = \sum_{i=1}^{m-|a| } \p ( ct_n\cap b_{i-1 } = \varnothing\mid b_i\in ct_n , y_1^n = y_1^n ) \\ & & \hphantom{\sum_{i=1}^{m-|a| } } \qquad \quad { } \times [ \p(b_i\in ct_n
\mid y_1^n = y_1^n ) - \p(b_i\in ct_n \mid y_1^n = \bar y_1^n ) ] \\ & & \qquad < -2f(m ) \sum_{i=1}^{m-|a| } \p ( ct_n\cap b_{i-1 } = \varnothing\mid b_i\in ct_n , y_1^n = y_1^n ) \\ & & \qquad \leq-2f(m ) \sum_{i=1}^{m-|a| } \p ( b_i\in ct_n , ct_n\cap b_{i-1 } = \varnothing\mid y_1^n = y_1^n ) \\ & & \qquad = - 2f(m ) [ 1-\p(b\cap ct_n = \varnothing\mid y_1^n = y_1^n ) ] , \end{aligned}\ ] ] where in the third line we have used lemma [ l1 ] .
this inequality can be rewritten as @xmath186 by ( [ fdef ] ) , the right - hand side is at most @xmath187 when @xmath179 , which gives a contradiction because @xmath178 .
write @xmath188 using ( [ fdef ] ) in combination with ( [ stringimpl ] ) , we may estimate @xmath189 let @xmath190 denote the elements of @xmath175 in increasing order , and define @xmath191 , @xmath192 , with the convention that @xmath193 .
then , using lemma [ l1 ] , we have @xmath194 & & \hphantom{\sum_{i=1}^{|a| } \bigl[}{}- \p\bigl ( y_0=b , \mathit{fct}_n(a)=a_i \mid y_1^n = \bar y_1^n \bigr ) \bigr]\\[-1pt ] & = & \sum_{i=1}^{|a| } [ \p ( y_0=b , ct_n\cap a_{i-1 } = \varnothing\mid a_i \in ct_n , y_1^n = y_1^n ) \p(a_i\in ct_n \mid y_1^n = y_1^n)\\[-1pt ] & & \hphantom{\sum_{i=1}^{|a| } [ } { } - \p ( y_0=b , ct_n\cap a_{i-1 } = \varnothing\mid a_i\in ct_n , y_1^n=\bar y_1^n ) \\ & & \hspace*{188pt}\hphantom{\sum_{i=1}^{|a| } [ - \p } { } \times \p(a_i\in ct_n
\mid y_1^n= \bar y_1^n ) ] \\[-3.5pt ] & = & \sum_{i=1}^{|a| } \p ( y_0=b , ct_n\cap a_{i-1 } = \varnothing\mid a_i\in ct_n , y_1^n = y_1^n ) d_i,\end{aligned}\ ] ] where @xmath195 in the third line , we have used the fact that @xmath196 ( the @xmath129-algebra generated by @xmath197 ) on the event @xmath198 , so that lemma [ l1 ] applies .
the definition of the set @xmath175 implies that @xmath199 for all @xmath200 .
hence , by using lemma [ l1 ] once more , we obtain @xmath201 & \leq&\sum_{i=1}^{|a| } \ind\{d_i\geq0\ } \p ( ct_n\cap a_{i-1 } = \varnothing\mid a_i\in ct_n , y_1^n = y_1^n ) d_i\\ & = & \sum_{i=1}^{|a| } \p ( ct_n\cap a_{i-1 } = \varnothing\mid a_i \in ct_n , y_1^n = y_1^n ) d_i\\ & & { } + \sum_{i=1}^{|a| } \ind\{d_i<0\ } \p ( ct_n\cap a_{i-1 } = \varnothing\mid a_i \in ct_n , y_1^n = y_1^n ) ( -d_i)\\ & \leq&\sum_{i=1}^{|a| } [ \p ( a_i\in ct_n , ct_n\cap a_{i-1 } = \varnothing\mid y_1^n = y_1^n ) \\ & & \hphantom{\sum_{i=1}^{|a| } [ } { } - \p ( a_i\in ct_n , ct_n\cap a_{i-1 } = \varnothing\mid y_1^n= \bar y_1^n ) ] + 2f(m)|a|\\ & = & \p(ct_n\cap a \neq\varnothing\mid y_1^n = y_1^n ) - \p(ct_n\cap a \neq\varnothing \mid y_1^n= \bar y_1^n ) + 2f(m)|a|\\ & \leq & f(m ) + 2f(m)m.\end{aligned}\ ] ] thus ,
we find that @xmath202 , where the upper bound does not depend on the choice of configurations made in ( [ fixy ] ) . exchanging @xmath147 and @xmath167 , we obtain the same bound for @xmath203 .
hence , we have proved ( [ stringimplagain ] ) . for simplicity , we will only consider @xmath65-values that are a multiple of @xmath204 . the proof is easily adapted to intermediate @xmath65-values .
we first state the following fairly straightforward lemma , where we note that @xmath205 .
[ easy ] for @xmath121 with @xmath120 , @xmath206 note that each cut time @xmath130 corresponds to a cut point @xmath160 , and so the set @xmath207 of cut times corresponds to a set @xmath208 of cut points . on the event @xmath209 , the interval @xmath210 is fully covered by @xmath211 . for each @xmath212 , we look at the steps of the random walk entering or exiting @xmath213 from the right : * if @xmath214 , then during the time interval @xmath215 there is at least one step exiting @xmath213 to the right . * if @xmath216 , then during the time interval @xmath215 there are at least two steps exiting @xmath213 to the right and one step entering @xmath213 from the right ( since there must be a return to @xmath213 from the right ) .
since each step refers to a single point @xmath213 only , and @xmath211 goes along at most @xmath65 edges ( and exactly @xmath65 edges when @xmath10 ) , we get that @xmath217 hence , @xmath218 . still on the event @xmath209 , the cut times corresponding to @xmath219 occur before time @xmath220 , and so @xmath221 hence , @xmath222 , and so ( [ tcspeed ] ) is proved .
for @xmath223 such that @xmath224 , we have @xmath225 therefore , by ( [ tcspeed ] ) , @xmath226 \\[-8pt ] \hspace*{20pt } & & { } \cup\biggl\{s_n \leq\frac{2m}{3}\biggr\}. \nonumber\end{aligned}\ ] ] in this subsection , we obtain upper bounds on the probabilities of the two events on the right - hand side of ( [ lastvisit ] ) when conditioned on @xmath227
. the upper bounds will appear in ( [ 1stterm ] ) and ( [ 2ndterm ] ) below . in section [ s2.4.2 ]
, we use these estimates to finish the proof of lemma [ l3 ] . write @xmath228 \\[-8pt ] & & \qquad = \sum_{k = m}^{n-1 } \p \biggl(s_{k}=\frac{2m}{3 } , s_{k+1}^n > \frac{2m}{3 } \bigm| y_1^n = y_1^n \biggr ) = \sum_{k = m}^{n-1 } \frac{n_k}{d_k } , % \nonumber\end{aligned}\ ] ] with ( recall proposition [ decomp ] ) @xmath229 estimate @xmath230 here , the bound arises by noting that @xmath231 and estimating @xmath232 .
thus , shifting @xmath160 back to the origin , we get @xmath233 with @xmath234 next , estimate @xmath235 here , the bound arises by restricting @xmath236 to the event @xmath237 noting that @xmath238 on this event , and inserting @xmath239 .
thus , shifting @xmath160 back to the origin , we get @xmath240 combining the upper bound on @xmath241 in ( [ nkub ] ) with the lower bound on @xmath242 in ( [ dklb ] ) , and canceling out the common factor @xmath163 , we arrive at @xmath243 \\[-8pt ] & & \qquad\leq\frac{\p(s_k= { 2m}/{3},s_1^k \sim y_1^k)}{\e ( ( 1/2 ) ^{r(s_1^k ) } \ind\{s_1^k \leq s_k\ } \ind\{s_1^k \sim y_1^k\ } ) } .
\nonumber\end{aligned}\ ] ] note that this bound is uniform in @xmath28 .
the numerator of ( [ ndk ] ) is bounded from above by @xmath244 , while the denominator of ( [ ndk ] ) is bounded from below by @xmath245 , where we note that @xmath246 for all @xmath165 on the event @xmath247 .
hence , by ( [ decomp2 ] ) , we have @xmath248 \\[-8pt ] & & \qquad \leq\sum_{k = m}^{n-1 } \frac{\p(s_k = { 2m}/{3 } ) } { ( { p(1-\eps)}/{2})^k}. \nonumber\end{aligned}\ ] ] the bound in ( [ 1stterm ] ) controls the first term in the right - hand side of ( [ lastvisit ] ) . since @xmath249 ,
we have @xmath250 provided @xmath28 is even ( which is necessary when @xmath10 because we have assumed that @xmath251 is even ) .
here , the constant @xmath252 comes from an elementary large deviation estimate , for which we must assume that @xmath253 the bound in ( [ 2ndterm ] ) controls the second term in the right - hand side of ( [ lastvisit ] ) .
in this section , we finally complete the proof of lemma [ l3 ] .
combining ( [ 1stterm])([2ndterm ] ) and recalling ( [ fdef ] ) and ( [ lastvisit ] ) , we obtain the estimate @xmath254 since there exists a @xmath255 such that , for @xmath256 , @xmath257 we see that @xmath258 as soon as @xmath259 note that ( [ driftcond ] ) holds for @xmath12 in a neighborhood of @xmath13 containing the line segment @xmath103 \times\{0\}$ ] . by cramer s theorem of large deviation theory ( see , e.g. , @xcite , chapter i ) , the left - hand side of ( [ actbd ] ) equals @xmath260 with @xmath261,\ ] ] where @xmath262 is the moment - generating function of the increments of @xmath263 . due to the strict convexity of @xmath264 ,
the supremum is attained at the unique @xmath265 solving the equation @xmath266 where we note that @xmath267 because of ( [ driftcond ] ) .
for the special case where @xmath268 , an easy calculation gives @xmath269 implying that @xmath270 with @xmath271^{5/6}[1/6(1-p)]^{1/6}$ ] .
hence , the inequality in ( [ actbd ] ) reduces to @xmath272 , which is equivalent to @xmath273 with @xmath274 .
the same formulas ( [ ipespdef])([barlam ] ) show that ( [ actbd ] ) holds in a neighborhood of @xmath13 .
throughout the remainder of this paper [ with the sole exceptions of section [ s5.1 ] and the claim of independence immediately prior to ( [ lim1 ] ) ] , we use @xmath275 , @xmath276 and @xmath277 to represent specific sequences rather than random sequences .
this abuse of notation will nowhere cause harm . in this section
, we prove theorem [ phtr](ii ) .
the proof is based on the following observations valid for a random walk that can not pause ( @xmath10 ) . on a color record of the type @xmath278^m$ ] , @xmath279 , the walk can not turn . indeed , a turn forces the same color to appear in the color record two units of time apart . any color record @xmath280 up to time @xmath281 can be seen in a unique way along a stretch of coloring of the type @xmath278^m$ ] with @xmath282 .
indeed , on such a stretch each site has a @xmath32-neighbor and a @xmath31-neighbor , so once the starting or ending point of the walk is fixed it is fully determined by @xmath280 .
we prove theorem [ phtr](ii ) by showing the following claim : * for any @xmath275 , @xmath105 and @xmath281 , we can find @xmath283 and @xmath284 such that @xmath285 where @xmath286 denotes the concatenation operation . in view of definition [ badconf ] , this claim will imply that @xmath275 is bad .
fix @xmath281 .
we begin with the choice of @xmath287 . for @xmath288 ,
let @xmath289^m \mathit{wbb } [ \mathit{wwbb}]^{2 m } \mathit{wbb } [ \mathit{wwbb}]^{2m+1}\nonumber \\[-8pt ] \\[-8pt ] & & { } \cdots \mathit{wbb } [ \mathit{\mathit{wwbb}}]^{2m+l-1 } \mathit{wbb } [ \mathit{wwbb}]^{2m+l}. % \nonumber\end{aligned}\ ] ] the interest in this color record relies on three facts : for @xmath290 , on the color record @xmath278^{2m+l}$ ] the walk can not turn [ see ( i ) above ] . on @xmath287 , the isolated @xmath32 s at the beginning of the @xmath291 s play the role of _ pivots
_ , since the walk can only turn there as is easily checked .
we call @xmath292 the pivot @xmath32 seen at time @xmath293 ( this is the first pivot ) and @xmath294 , @xmath295 , the subsequent pivots seen at times @xmath296 = k(2k+8m+1 ) + 5 m , \qquad k=1,\ldots , l.\ ] ] since the length of the color record @xmath278^{2m+l}$ ] increases with @xmath297 , if the walk does not turn on pivot @xmath294 , then it can not turn on any later pivot . indeed , going straight through @xmath294 means that the coloring has an isolated @xmath32 surrounded by two @xmath31 s , and this color stretch is impossible to cross at any later time with any color record of the type @xmath278^m$ ] , @xmath279 .
the first color record @xmath278^m$ ] serves to prevent @xmath292 from being in the coloring seen by the walk up to time @xmath220 , because the walk can not turn between time @xmath65 and time @xmath293 [ see ( i ) above ] .
the total time is @xmath298 the above three facts imply that the behavior of the walk from time @xmath65 to time @xmath28 ( i.e. , the increments @xmath299 ) , leading to @xmath287 as its color record , can be characterized by the first pivot @xmath294 , if any , where the walk makes no turn .
there are @xmath300 possibilities , including the ones where there is a turn at every pivot or at no pivot .
this characterization is up to a 2-fold symmetry in the direction of the last step of the walk , which can be either upwards or downwards ( this is the same symmetry as @xmath301 ) .
note that , except for the case where the walk makes no turn from time @xmath65 to time @xmath28 , the behavior of the walk from time @xmath80 to time @xmath65 ( i.e. , the increments @xmath302 ) is fully determined ( up to the 2-fold symmetry ) by @xmath303 [ see ( ii ) above ] .
this is because @xmath304 is increasing , so that @xmath305 .
our goal will be to prove that for large @xmath306 the walk , conditioned on @xmath307 , with a high probability turns on every pivot and ends by moving upwards . to that end
, we define the following events for the walk up to time @xmath28 : * @xmath308the walk turns on pivots @xmath309 and does not turn on pivots @xmath310 ( `` last turn on @xmath297 '' ) , @xmath290 . * @xmath311 ( `` no turn '' ) . * @xmath312 ( `` end upwards '' ) . * @xmath313 ( `` end downwards '' ) . using these events ,
we may write @xmath314 \\[-8pt ] \hspace*{20pt}&&{}+ \sum_{l=0}^l [ \p ( lt_l , eu \mid y_1^{m-1 } \vee \bar y_m^n ) + \p ( lt_l , ed \mid y_1^{m-1 } \vee\bar y_m^n ) ] .
% \nonumber\end{aligned}\ ] ] now , on the event @xmath315 , the length of the coloring seen by the walk from time @xmath80 to time @xmath28 is @xmath316 = ( l - l+1)(2l+2l+8m+3).\ ] ] only two walks from time @xmath65 to time @xmath28 are in @xmath315 and these are reflections of each other ( one in @xmath317 and one in @xmath318 ) .
for either of these two walks , we have that @xmath319 , where @xmath320 = ( 1 + 8 m ) 1\{l\mbox { odd}\ } + 2l.\ ] ] it is easily checked that any walk in @xmath321 ends a distance at least @xmath322 above any walk in @xmath323 , with ( see figure [ figzigzag ] ) @xmath324 hence we have , using the fact that all walks in @xmath315 visit the same number of colors , @xmath325 since @xmath326 ( because @xmath327 ) and @xmath328 , it follows that for @xmath306 large the probability of @xmath329 is negligible with respect to the probability of @xmath330 uniformly in @xmath297 .
( 10,10)(0,-2 ) ( 0,0 ) ( 2,.6 ) ( -4,1.2 ) ( 6,1.8 ) ( -8,2.4 ) ( 10,3 ) ( -.4,-1.2)@xmath65 ( 1.5,-.6)@xmath331 ( -4.6,0)@xmath332 ( 5.5,.6)@xmath333 ( -8.5,3.1)@xmath334 ( 8.6,3.8)@xmath335 ( 0,0)(1,.3)(2,.6 ) ( 2,.6)(-1,.9)(-4,1.2 ) ( -4,1.2)(1,1.5)(6,1.8 ) ( 6,1.8)(5,1.85)(4,1.9 ) ( -6,2.2)(-7,2.3)(-8,2.4 ) ( -8,2.4)(1,2.7)(10,3 ) ( 10,3)(12,3.1)(14,3.2 ) ( 4,1.9)(-1,2.05)(-6,2.2 ) the same reasoning gives the inequality @xmath336 \\[-8pt ] \hspace*{20pt } & & \qquad \leq \biggl(\frac{p}{1-p } \biggr)^{u(l+1)+5 m } \biggl(\frac 12 \biggr)^{t(l+1)-t(l ) } \p ( lt_{l+1},eu \mid y_1^{m-1 } \vee\bar y_m^n )
. % \nonumber\end{aligned}\ ] ] indeed , any walk in @xmath337 covers @xmath338 more sites than any walk in @xmath339 , while it is not hard to see that it makes at most @xmath340 more steps to the right . since @xmath341 and @xmath342 as @xmath343 , and @xmath344 ( because @xmath345 ) , we find that @xmath346 decreases exponentially in @xmath297 for @xmath297 large .
hence the largest value @xmath347 dominates .
similar estimates allow us to neglect probabilities containing the event @xmath348 . combining ( [ 1split])([ineqkk1 ] )
, we obtain that , for fixed @xmath65 , @xmath349 which immediately yields that , for fixed @xmath65 , @xmath350,\hspace*{-40pt}\ ] ] where the error @xmath351 tends to zero as @xmath352 .
the key point of ( [ limids1 ] ) is that @xmath353 forces the coloring around the origin to look like @xmath354 .
more specifically , @xmath355 tells us the coloring on a large region relative to @xmath356 and , after this , @xmath280 determines the walk from time @xmath80 to time @xmath65 ( relative to @xmath78 ) .
since @xmath78 is independent of @xmath357 , we therefore have @xmath358 equations ( [ limids1 ] ) and ( [ lim1 ] ) tell us that for large @xmath28 , @xmath359 will be very close to @xmath46 or @xmath360 . the idea
now will be to modify the extension far away so that an `` opposite '' type of structure is forced upon us and thereby reverse the @xmath46 and @xmath360 above .
we next move to the choice of @xmath361 .
we take @xmath362^m \mathit{wbb } [ \mathit{wwbb}]^{2 m } \mathit{wbb } [ \mathit{wwbb}]^{2m+1 } \nonumber \\[-8pt ] \\[-8pt ] & & { } \cdots \mathit{wbb } [ \mathit{wwbb}]^{2m+l-1 } [ \mathit{wwbb}]^{2m+l}. % \nonumber\end{aligned}\ ] ]
the difference with @xmath287 in ( [ yblocks ] ) is that we removed the last pivot @xmath363 and the 2 b s following it ( so that @xmath364 ) .
the same computations as before give @xmath365 \\[-8pt ] & & \qquad = \p ( c_0=b \mid lt_{l-1},eu , y_1^{m-1 } \vee\widetilde y_m^n ) [ 1+o(1 ) ] .
\nonumber\end{aligned}\ ] ] now @xmath366 forces the walk to do the exact opposite up to time @xmath367 to what @xmath368 forced it to do , because there is one turn less and the walk still ends upwards .
therefore , by symmetry , the walk from time @xmath80 to time @xmath220 must also do the exact opposite , and so we conclude that , for @xmath369 , @xmath370 \\[-8pt ] & & \qquad \longleftrightarrow\quad \p ( c_0=b \mid lt_{l-1},eu , y_1^{m-1 } \vee\widetilde y_m^n ) = 1-q .
% \nonumber\end{aligned}\ ] ] combining ( [ limids1 ] ) and ( [ limids2])([symq ] ) , we obtain the claim in ( [ limbartil ] ) .
in this section , we prove theorem [ phtr](iii ) .
we will prove that if @xmath14 and @xmath10 , then @xmath371\\[-8pt ] \hspace*{30pt}y_1^\infty&= & \mathit{bbwbb } [ \mathit{wwbb } ] \mathit{wbb } [ \mathit{wwbb}]^2 \mathit{wbb } [ \mathit{wwbb}]^3 \cdots\ \mbox{is good}. % \nonumber\end{aligned}\ ] ] ( in the second line , @xmath372 is put at the beginning to ensure that the first @xmath32 may be a pivot . ) in this subsection , @xmath227 and @xmath373 denote random sequences , and we switch back to specific sequences only in the last display .
write @xmath374 with @xmath375 \\[-8pt ] d(n ) & : = & \p ( y_0^{n-1}=b^n ) = \sum_{i , j\in\n } \biggl(\frac12\biggr)^{i+j+1 } p(n , i , j ) , % \nonumber\end{aligned}\ ] ] where @xmath376 is the probability that simple random walk ( with @xmath14 and @xmath10 ) starting from @xmath1 stays between @xmath377 and @xmath378 ( inclusive ) prior to time @xmath28 . to see the second equality in ( [ nndndef ] ) , let @xmath379 be the event that there is a @xmath31 at the origin , and the first @xmath32 to the right and to the left of the origin
are located at @xmath200 and @xmath380 , respectively .
then @xmath381 which is easily seen to be the claimed sum .
the first equality in ( [ nndndef ] ) is handled similarly .
trivially , @xmath382 for all @xmath383 , and therefore @xmath384 next , using proposition 21.1 in @xcite , we easily deduce that @xmath385^{n-1 } \cases{\displaystyle c_i^\mathrm{even } , & \quad as $ n \to\infty$ through $ n$ even,\cr\displaystyle c_i^\mathrm{odd } , & \quad as $ n \to\infty$ through $ n$ odd , } % \ ] ] where @xmath386 means that the ratio of the two sides tends to 1 , and @xmath387 from this it follows that @xmath388 combining ( [ nndndef])([ratlim ] ) , we get @xmath389 , that is , @xmath390 on the other hand , an extension of @xmath391 with @xmath392 as in section [ s4 ] gives @xmath393\nonumber \\[-8pt ] \\[-8pt ] \hspace*{20pt } & = & \tfrac12 [ 1+o(1 ) ] \nonumber\end{aligned}\ ] ] [ recall ( [ ineqkk1])([lim1 ] ) ] . combining ( [ limb1 ] ) and ( [ limb2 ] ) , we get the first claim in ( [ badgood ] ) .
pick @xmath288 and @xmath394 .
then @xmath395 \mathit{wbb } [ \mathit{wwbb}]^2 \cdots \mathit{wbb } [ \mathit{wwbb}]^l.\ ] ] as in section [ s4 ] , a turn on a white pivot forces turns on all previous white pivots .
therefore a walk compatible with @xmath280 having at least one turn is characterized by the index @xmath396 of its last pivot @xmath397 .
the time of the @xmath130th pivot is @xmath398 $ ] .
conditioning on @xmath399 still leaves us the freedom to choose @xmath400 and @xmath401 . since @xmath14
, it is easily checked that , conditioned on @xmath399 ( and even on the last pivot ) , @xmath78 and @xmath402 are independent fair coin flips .
there are @xmath403 compatible walks with no turn and @xmath404 compatible walks with at least one turn . since @xmath14 , all these walks have the same probability , but the walks with no turn have a larger cost for the coloring .
let @xmath348 and @xmath405^c$ ] denote the event that the walk makes no turn , respectively , at least one turn .
we claim that @xmath406 to see how this comes about , recall proposition [ decomp ] , which says that for an arbitrary walk @xmath407 and an arbitrary extension @xmath287 , @xmath408 ( the notation @xmath409 denotes the walk obtained by appending the second walk to the end of the first walk . )
note that any compatible walk up to time @xmath220 ends either at the right end of the range or at the left end of the range .
let @xmath410 $ ] and @xmath411 $ ] denote compatible walks with no turn , respectively , at least one turn , either both ending at the right end of the range or both ending at the left end of the range
. then @xmath412 \vee\bar s_1^{n - m+1 } ) \geq r(s_1^{m-1}[1 ] \vee\bar s_1^{n - m+1})$ ] .
moreover , for any @xmath413 and @xmath392 , if @xmath410 \vee\bar s_1^{n - m+1 } \sim y_1^{m-1 } \vee\bar y_m^n$ ] , then also @xmath411 \vee\bar s_1^{n - m+1 } \sim y_1^{m-1 } \vee\bar y_m^n$ ] . hence , @xmath414 , y_1^{m-1 } \vee\bar y_m^n ) \leq\p ( s_1^{m-1}[1 ] , y_1^{m-1 } \vee\bar y_m^n ) .\ ] ] summing over @xmath410 $ ] and @xmath411 $ ] , we obtain ( [ ptnt ] ) .
next , on the event @xmath415 , @xmath416 is fully determined by @xmath78 and @xmath417 .
therefore , by symmetry , @xmath418 hence , uniformly in @xmath287 , @xmath419 since @xmath352 as @xmath143 , the second claim in ( [ badgood ] ) follows .
in this section , we explain a strategy for proving that @xmath100 when @xmath420 and @xmath109 .
it seems that this case is much more delicate than the case @xmath97 and @xmath268 treated in sections [ s4][s5 ] .
this strategy will be pursued in future work . for @xmath279
, we use the notation @xmath421 , @xmath422 , @xmath423^m$ ] etc . to abbreviate @xmath424
fix any configuration @xmath425 . to try to prove that @xmath426 is bad
, we do the following : \(1 ) for @xmath427 with @xmath428 , we consider the two color records from time @xmath65 to time @xmath429 defined by @xmath430^kw , \qquad\bar y_m^{m+kk}(w ) : = [ \mathit{bw}^{k-1}]^kb.\ ] ] ( 2 ) we _ expect _ that , for any @xmath431 and @xmath109 , @xmath432 in view of definition [ badconf ] , this would imply that @xmath275 is a bad configuration , as desired .
the idea behind the above strategy is that @xmath433 forces the walk to hit many white sites at sparse times from time @xmath65 onwards . in order to achieve this
, the walk can either move out to infinity , in which case the coloring must contain many long black intervals , or the walk can hang around the origin , in which case the coloring must contain a single white site close to the origin with two long black intervals on either side . since the drift of the random walk is not too large , the best option is to hang around the origin .
the single white site , at or next to the origin , is enough for the walk to generate any ( ! )
color record @xmath280 prior to time @xmath65 , because the pausing probability is strictly positive .
as a result , the conditional probability to see a black origin given @xmath434 is closer to @xmath80 than given @xmath435 . with the latter conditioning , the role of @xmath31 and @xmath32
is reversed , and the effect of the conditioning is to have the origin lie in a region containing a single black site separating two long white intervals , so that the conditional probability to see a black origin is closer to @xmath1 . the task is to control the conditional probability @xmath436 . for that purpose , mark the positions of the walk at the times @xmath437 , @xmath438 , that correspond to the isolated @xmath32 s in @xmath439 . by the definition of @xmath433 , two subsequent @xmath32 s either correspond to the same white site or to two white sites that are separated by a single interval of black sites of length at least @xmath80.=1 on the event @xmath440 , let @xmath292 be the white site visited at time @xmath65 .
relative to this site , all the white sites in @xmath33 can be labeled @xmath441 , with @xmath442 the first white site on the left of @xmath292 , @xmath443 the first white site on the right of @xmath292 , etc .
( see figure [ figcol ] ) .
let @xmath444 denote the black interval between @xmath445 and @xmath446 .
@xmath447 and @xmath448 are the indices of the left - most and right - most white sites visited by the walk between times @xmath65 and @xmath429 .
is the white site seen at time @xmath65 in @xmath449 . ]
the above representation allows to obtain an explicit ( although complex ) formula for the conditional probability @xmath450 involving classical simple random walk quantities .
let @xmath451 denote the event that @xmath444 is visited between times @xmath65 and @xmath429 .
then the key fact that needs to be proved is the following : @xmath452 from ( [ keyfact ] ) , we are able to prove the desired result ( [ star ] ) , but the argument needed to prove ( [ keyfact ] ) is long and we are still working on trying to complete it .
sbastien blachre acknowledges travel support from the esf - program `` random dynamics in spatially extended systems '' ( 20022007 ) .
frank den hollander and jeffrey e. steif are grateful for hospitality at the mittag - leffler institute in stockholm in the spring of 2009 , when a part of the research on this paper was carried out .
we finally thank the referee for a thorough reading and comments on the paper . | in this paper , we consider a random walk and a random color scenery on @xmath0 .
the increments of the walk and the colors of the scenery are assumed to be i.i.d . and to be independent of each other .
we are interested in the random process of colors seen by the walk in the course of time .
bad configurations for this random process are the discontinuity points of the conditional probability distribution for the color seen at time zero given the colors seen at all later times .
we focus on the case where the random walk has increments @xmath1 , @xmath2 or @xmath3 with probability @xmath4 , @xmath5 and @xmath6 , respectively , with @xmath7 $ ] and @xmath8 , and where the scenery assigns the color black or white to the sites of @xmath0 with probability @xmath9 each .
we show that , remarkably , the set of bad configurations exhibits a crossover : for @xmath10 and @xmath11 all configurations are bad , while for @xmath12 in an open neighborhood of @xmath13 all configurations are good .
in addition , we show that for @xmath10 and @xmath14 both bad and good configurations exist .
we conjecture that for all @xmath15 the crossover value is unique and equals @xmath16 .
finally , we suggest an approach to handle the seemingly more difficult case where @xmath17 and @xmath18 , which will be pursued in future work . , + . . |
starburst galaxies are hosts of intense star formation activity with star formation rates ( sfr ) more than a factor of 3 compared to the milky way ( sfr@xmath310m@xmath4yr@xmath5 ) .
the stellar population of the starbursts , especially the number of massive stars and the slope of the initial mass function ( imf ) , directly affect the hardness of the ionization field in the interstellar medium ( ism ) .
therefore , ratios of emission lines of different excitation levels are powerful tools to study the properties of the ionization field and to infer the stellar population of the starbursts .
infrared emission - line ratios are of particular interest and have been explored extensively with the infrared space observatory ( i.e. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , as the active star forming regions in the disks of starburst galaxies and circumnuclear regions are often optically obscured by dust . as a result ,
the emitted radiation , originating from a massive starburst and/or an accreting active nucleus , is emerging in the mid - infrared ( mid - ir ) ( see * ? ? ?
* ; * ? ? ?
* ) and far - ir .
the advent of the spitzer space telescope with its improved sensitivity opens new avenues in this direction , enabling detailed studies of large samples of quiescent and star forming galaxies @xcite , as well as more complex dust enshrouded infrared luminous sources which often host an active galactic nucleus ( agn ) @xcite . estimating the hardness of the ionization field plays a key role in determining the dominant power source at the centers of galaxies .
nuclear starbursts and accreting super - massive black holes ( smbh ) are often intertwined and coexist , as best demonstrated in the case of ultra luminous infrared galaxies ( ulirgs)(see e.g. * ? ? ?
* ; * ? ? ?
* and references therein ) .
the radiation of an agn , produced from the accretion of gas and stars onto the smbh , has a much harder spectrum compared to the one produced by the massive young stellar populations .
therefore , the hardness of the radiation field has been a key parameter addressed by previous diagnostic methods to determine the energy source in galaxies ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in the study of the ionization field of galaxies with massive starbursts , the case of low - metallicity blue compact dwarfs ( bcds )
is particularly interesting .
bcds are galaxies characterized by their blue optical colors ( m@xmath6@xmath7 - 18mag ) , small linear sizes ( @xmath71kpc ) and low luminosities @xcite .
even though bcds are defined based on their morphological properties , they have been ubiquitously found to have low oxygen abundances as measured from the ionized gas in hii regions .
it is well established that the hardness of the stellar radiation increases with decreasing metallicity ( e.g. * ? ? ?
as a result , the ionization field of bcds is generally much harder compared to a typical starburst .
it is also known that the mass of a super massive black hole ( smbh ) is correlated with the mass of the bulge of the host galaxy @xcite while only a fraction of the most massive smbh are actively emitting radiation as a result of mass accretion @xcite .
bcds in general are not massive enough to host such an active smbh , with possibly only a few exceptions ( see * ? ? ?
* ; * ? ? ?
consequently , the excitation mechanism for the hard radiation field in bcds is likely significantly different from the one responsible for the radiation field produced by an agn .
furthermore , unlike starburst galaxies which are often obscured by dust , an analysis of a large sample of bcds suggests that in bcds nearly always the young massive stars which ionize the interstellar medium are easily detected and their dust extinction measured from the optical balmer lines is typically fairly low @xcite .
there might be exceptions indicated from the mid - ir observations , with sbs0335 - 052e as the most extreme case .
however , this is rather a unique galaxy and deviates from the usual trends seen in bcds for different reasons as discussed in detail by @xcite .
understanding the differences in the excitation mechanism between bcds and agns can provide insightful clues to decipher the power source in a starburst galaxy also hosting an agn .
identifying those differences from the mid - ir spectral features is one of the main motivations behind the present work .
bcds are often much fainter than typical starburst galaxies in the infrared .
up until the launch of the _ spitzer _
space telescope @xcite , studies of starburst galaxies in the ir have included only a handful of bcds @xcite . the superb sensitivity of the _ spitzer _ allows us to observe a large sample of bcds ( see * ? ? ?
* ; * ? ? ?
* ) . in this paper
, we provide the ionized gas diagnostic analysis of the low metallicity bcds accessible to the high - resolution modules of the infrared spectrograph ( irs , * ? ? ?
in addition , we contrast the bcd sample with typical starbursts and agns , to probe the excitation mechanism in these types of objects and understand where the difference comes from . in 2 , we introduce the sample and the data reduction method . in 3 , we study the high - ionization emission lines , and propose a new diagnostic diagram to determine the energy source in a galaxy . in 4 , we focus on other infrared emission lines and compare various excitation diagnostic diagrams of starbursts and bcds .
our findings are summarized in 5 .
we have compiled a sample of 12 bcds and observed them in starring mode with the high - resolution modules ( short - high , sh : 4.7@[email protected]@xmath8 and long - high ,
lh : 11.1@[email protected]@xmath8 ) of the irs as part of the instrument team s guaranteed time observation ( gto ) programs .
the galaxies were drawn from the sample of @xcite .
we exclude tol1214 - 277 as it is too faint and its high - resolution spectrum has low signal - to - noise ratio .
we also exclude cg0752 , cg0563 , and cg0598 as they do not show typical bcd signatures .
the characteristics of individual objects in the sample and analysis based on their _ spitzer _ low - resolution spectra are presented in @xcite .
the high - resolution module slits were centered on the nucleus of each galaxy , as was the case of the low - resolution observations @xcite .
the data were processed by the _ spitzer
_ science center ( ssc ) data reduction pipeline version 13.2 .
individual pointings to each nod position of the slit were co - added .
the median of the combined images were extracted using the full slit extraction method of the _ irs _ data analysis package smart @xcite . then the extracted spectra were flux calibrated by multiplying them with a relative spectral response function ( rsrf ) , which was created from the _ irs _ standard star , @xmath10 dra , for which accurate templates were available @xcite .
note that emission from the sky background has not been subtracted from our spectra due to the lack of dedicated sky observations , thus discontinuity exists in some source spectra between sh and lh , however , this will not affect the results of our analysis since our conclusions are based on ratios of emission lines observed in the same module . in the appendix
, we include close up plots of strong and weak emission lines visible in the irs high - resolution spectra of bcds . in order to measure the line strengths ,
a gaussian profile is fitted to the lines above a local continuum fitted by an one - order polynomial .
the [ oiv]25.89@xmath0 m and [ feii]25.99@xmath0 m are close to each other and they are fit together using a two - gaussian profile with fixed rest - frame wavelengths . the measured line strengths for the bcd sample are listed in table [ tabline ] .
the errors are estimated based on the error propagation of the gaussian fit assuming the error of the flux as the root mean square ( @xmath11 ) of the continuum fit .
note that this does not take into account all the systematic uncertainties , especially the flux calibration .
we estimate these to be 5% and choose it as our minimum uncertainty .
the upper - limits are obtained by measuring the flux of a gaussian with a height three times the local @xmath11 , and with a fwhm equal to the instrumental resolution .
no extinction correction has been performed in our analysis .
the corrections are typically minor for mid - ir fine - structure line ratios , except for ratios involving lines found close to the silicate features , such as the [ siv]10.51@xmath0 m line . in the analysis of diagrams involving the [ siv ] line ( see figure [ sratio_neratio ] ) , we provide arrows indicating the effects of an a@xmath12=30 mag of extinction , assuming a uniform dust screen and a @xcite extinction curve ( see * ? ? ?
in order to study the ionization field of starbursts over a broad range of physical properties , we also examine a sample of 24 local starbursts observed in staring mode with the irs high - resolution modules .
the details of the sample , their high - resolution spectra and line measurements are described in @xcite .
it should be emphasized that even though we broadly use the term `` starburst '' when referring to them , some may also have a weak agn component .
in addition , we created a comparison sample of typical agns to investigate how the difference in the excitation mechanism of the gas , namely the radiation produced by an accretion disk versus massive young star formation , is reflected in the mid - ir emission lines .
it is not our intention to have an agn sample that is complete , but rather one that is representative of typical agns .
hence , we choose bona - fide agns , with no ambiguity in their classification , where the nuclear emission is known to dominate the integrated spectra of their host galaxies .
we select from the type 1 sources ( quasars and seyfert 1s ) used in @xcite for which irs high - resolution observations are available .
we also examine the strength of the polycyclic aromatic hydrocarbon ( pah ) emission in their mid - ir spectra and exclude sources with strong pah features , since this would suggest a significant contribution from star forming regions in the disk of the host galaxy . among those sources with weak pah emission , we choose 10 agns that show strong mid - ir emission lines , in order to obtain the best s / n possible in our measurements .
the agn irs spectra were extracted and the emission lines measured in the same way as the bcds ( see 2.1 ) .
furthermore , whenever applicable , we also compare the properties of emission lines in bcds with those in low - redshift ulirgs as well as hii regions found in our galaxy , and in the large and small magellanic clouds ( lmc / smc ) . the sample and line measurements of the low - redshift ulirgs are taken from @xcite , and the hii region data from @xcite and @xcite .
the irs high - resolution spectra of bcds are rich in fine - structure emission lines .
all bcds display strong [ neiii]15.56@xmath0 m , [ siii]18.71@xmath0 m , and [ siv]10.51@xmath0 m emission .
the [ neii]12.81@xmath0 m line is well detected in about half of the bcd sample , but it is not seen in sbs0335 - 052e , tol65 , and viizw403 . in the comparison samples of starbursts and agns , all these lines are strong and we discuss their relative flux ratios in 4 .
the presence of high - ionization emission lines is of great interest , since their excitation often requires extreme conditions .
one particular case is the [ oiv]25.89@xmath0 m line which has an ionization potential ( ip ) of 54ev , just above the heii edge , and therefore it can not be readily generated in the hii regions surrounding main - sequence stars .
however , the [ oiv ] line had already been detected in several starburst galaxies with iso ( see * ? ? ?
a possible explanation for this was that the [ oiv ] emission is due to a dust enshrouded agn as the hard radiation field produced by the agn can easily provide the energy needed to ionize o@xmath13 .
however , @xcite showed that at least in some starburst systems ( such as the outer regions of m82 ) , the [ oiv ] can not be attributed to a weak agn .
instead it must be due to either very hot stars , such as the wolf - rayet ( wr ) stars ( see also * ? ? ? * ) or to the ionizing shocks associated with the starburst activity . with the superb sensitivity of the _ spitzer _
space telescope , we now find that the [ oiv ] line is fairly common in our bcd sample as well as the comparison sample of starburst galaxies .
we detected [ oiv ] emission in 8 of the 12 bcds , and in 21 of the 24 starbursts ( see * ? ? ? * ) .
this is very interesting since , as we will discuss in the following sessions , line ratios involving [ oiv ] provide a good diagnostic tool for identifying the presence of an agn ( see also * ? ? ?
* ; * ? ? ?
25.89@xmath0m/[siii]33.48@xmath0 m as a function of [ neiii]15.56@xmath0m/[neii]12.81@xmath0 m , for bcds ( diamonds ) , starbursts ( squares ) , agns ( crosses ) , and ulirgs ( plus signs ) .
a circle around a symbol indicates that [ nev]14.32@xmath0 m emission has been detected , while a filled symbol implies that this galaxy also has been classified as a wolf - rayet ( wr ) in the optical ( see 3.4 ) .
the error bars of the line ratios in this and all subsequent plots are similar to the symbol size .
the galaxies appear to be localized on two branches .
one branch consists mainly of agns , or starbursts / ulirgs hosting an agn , while the other mainly contains bcds or pure starburst galaxies .
two dashed lines are drawn as a guide to identify the two branches .
[ os3_neratio ] ] in figure [ os3_neratio ] , we plot the line ratios of [ oiv]25.89@xmath0m/[siii]33.48@xmath0 m as a function of [ neiii]15.56@xmath0m/[neii]12.81@xmath0 m for our bcds , starburst galaxies agns and ulirgs .
we choose to use the line ratio of [ oiv]25.89@xmath0 m and [ siii]33.48@xmath0 m because of their large difference in the ionization potential , and the absence of aperture effects , as both lines reside in the same irs long - high module ( see also figure 3 in @xcite ) .
we use the [ neiii]/[neii ] line ratio as the other variable , as it is commonly used as an indicator of the strength of the euv radiation field ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
since both [ neiii ] and [ neii ] lines are of the same element , the ratio is independent of the gas abundance .
we clearly see that the [ oiv]/[siii ] ratio separates agns from bcds , with agns having in general a factor of 10 higher [ oiv]/[siii ] ratios than bcds , even though they have similar [ neiii]/[neii ] ratios .
starburst galaxies display smaller [ neiii]/[neii ] ratios than agns and bcds , but their values of the [ oiv]/[siii ] ratio vary . in particular , it appears that starburst galaxies populate two branches with a common origin ( the dashed lines in figure [ os3_neratio ] ) , one pointing towards the locus of agns and the other to the bcds . among the starburst galaxies ,
mrk266 , ngc1365 , ngc2623 , ngc4194 , ngc4945 , ngc520 , ngc660 , ngc253 , ngc1614 , ngc1097 , ngc3079 , ngc4676 , and ngc7252 are located in the upper branch reaching to the agn locus , while ngc1222 , ngc2146 , ngc3256 , ngc3310 , ngc3628 , ngc4088 , and ngc7714 are in the lower branch that ties to the bcds .
the galaxies ic342 , ngc3556 , and ngc4818 have no detected [ oiv ] emission and their upper limits are indicated with arrows on the plot .
given that the gas in agns and bcds is excited by distinctly different physical mechanisms , the fact that they are found at the corresponding ends of the two branches of the diagram , also suggests that the excitation mechanisms leading to the [ oiv ] line emission in starbursts galaxies populating the two branches is different .
this could be understood if the mid - ir emission of starbursts in the upper branch is due to some agn component , while in starbursts of the lower branch is only due to the reprocessing of pure stellar radiation .
therefore , such a diagram can be used as a diagnostic tool to determine the origin of the dust enshrouded energy source in star forming galaxies .
interestingly , among the starbursts located in the upper branch , six also have detected [ nev]14.32@xmath0 m emission ( see * ? ? ?
* ) , which has an ip of 97ev and is a clear agn indicator .
five galaxies ( ngc520 , ngc253 , ngc1614 , ngc4676 and ngc7252 ) which do not have a confirmed agn signature so far , are located in the lower left corner of the plot , where the two branches merge . since at this area of the plot classification is more ambiguous , it is possible that a weak agn component could be present , even though its [ nev ] emission is too weak to be detected .
unfortunately , in most of the ulirgs with available irs high resolution spectroscopy , their [ siii]33.48@xmath0 m emission lines are redshifted out of the wavelength coverage of the irs ( see * ? ? ?
* ) . among the 10 ulirgs
that have both [ oiv]25.89@xmath0 m and [ siii]33.48@xmath0 m detected , 9 are located on the upper branch and they all have detected [ nev ] emission .
iras09022 - 3615 is the only ulirg with no [ nev ] detection and it is located between the two branches .
this is again consistent with the assertion that star forming galaxies with an agn component do follow the upper branch .
the lower branch of the diagram seems to be populated by sources in which the ionizing radiation is due only to stellar photospheres .
in addition to the bcds , the pure starburst galaxies ngc7714 , ngc3310 , and ngc1222 ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) are located in the lower branch before the two branches merge .
moving along the branch towards higher values of the [ neiii]/[neii ] ratio , the metallicity overall decreases ( see 4 ) .
the harder radiation field produced by the low metallicity environment strongly increases the [ neiii]/[neii ] ratio ( by up to a factor of 100 ) , but increases only moderately the [ oiv]/[siii ] ratio ( by a factor of 10 ) .
this was also noticed by @xcite , who indicated that low - metallicity dwarfs exhibit relatively stronger [ oiv ] than low - excitation starbursts , but the effect is not very pronounced compared to the large excitation difference as measured by the [ neiii]/[neii ] .
one galaxy with a detected [ nev ] emission ( ngc3628 , see * ? ? ?
* ) is on the lower branch .
we note though that its location is also at the far left where the two branches merge , and as a result it does not contradict our bifurcation scheme .
25.89@xmath0m/[siii]33.48@xmath0 m line ratio as a function of the [ siv]10.51@xmath0m/[siii]18.71@xmath0 m for our samples .
the symbols are the same as in figure [ os3_neratio ] .
note that the bifurcation into two branches , indicated again with the dashed lines , is also visible .
[ os3_sratio ] ] as with the [ neiii]/[neii ] ratio , the ratio of [ siv]10.51@xmath0m/[siii]18.71@xmath0 m has also been widely used as a probe of the hardness of the ionization field . when plotting the [ oiv]25.89@xmath0m/[siii]33.48@xmath0 m as a function of the [ siv]/[siii ] line ratio , we observe a similar bifurcation signature ( see figure [ os3_sratio ] ) .
the sources located on the upper and lower branches in figure [ os3_sratio ] are the same to the ones found on the upper and lower branches in figure [ os3_neratio ] .
the separation of the two branches though is not as clear as in figure [ os3_neratio ] , mainly because of the narrower range of the [ siv]/[siii ] values for our sample than the range of the [ neiii]/[neii ] ratios .
in addition , in several starbursts ( 8 out of the 24 galaxies ) no [ siv ] emission was detected , thus making a clear definition of the merging points of the two branches challenging .
close examination of the [ oiv]/[siii ] vs. [ neiii]/[neii ] diagram presented in the previous section suggests that it can be used as an agn - starburst diagnostic , similar to the log([nii]/h@xmath1 ) vs. log([oiii]/h@xmath2 ) diagram ( e.g. the bpt diagram , * ? ? ?
* ) , which has been used traditionally to distinguish agns from starbursts in the optical @xcite . in both diagrams , star forming galaxies
are distributed along two separate branches , each having one end respectively anchored by the location of agns and bcds . in figure
[ diagram ] , we plot on the bpt diagram @xmath1440,000 galaxies from the sloan digital sky survey ( sdss , see * ? ? ?
strong agns have high [ nii]/h@xmath1 and [ oiii]/h@xmath2 ratios and they are found at the tip of the right branch .
low - metallicity bcds have high [ oiii]/h@xmath2 but low [ nii]/h@xmath1 ratios and they are located at the tip of the left branch .
the natural separation of galaxies into two branches indicates again a different excitation mechanism .
the blue dotted line , empirically defined by @xcite , separates the two branches .
these authors classify galaxies below the line as star forming galaxies , and those above it as agns .
the red dashed line is taken from @xcite , and demarcates the maximum position that can be obtained by pure photo - ionization models .
galaxies located above the line require an additional excitation mechanism , such as an agn , or strong shocks .
the general classification scheme using the two separation lines @xcite considers objects above kewley s line as agn dominated sources , between kewley s and kauffmann s line as composite agn and starburst sources , and below the kauffmann s line as star forming galaxies . /h@xmath1 ) vs. log([oiii]/h@xmath2 ) diagram in which we include @xmath1440,000 sdss galaxies , marked as black dots .
the red dashed line and the blue dotted line are taken from @xcite and @xcite and they are used to separate agns , to the top right of the plot , from starburst galaxies to the bottom left .
we include on the diagram our sample of bcds and starbursts as well as the star - forming galaxies from @xcite ( d06 ) .
note that some galaxies in our starburst sample and in d06 also host an active nucleus and as a result are found in the agn locus .
galaxies located at the upper and lower branch of figure [ ne32_o4s3_dale ] are plotted in orange and cyan respectively ( see text ) .
[ diagram ] ] to quantitatively check the analogy between our [ oiv ] diagram ( figure [ os3_neratio ] ) and the log([nii]/h@xmath1 ) vs. log([oiii]/h@xmath2 ) diagram used in the optical , we have collected from the literature existing optical data for our starbursts and bcds .
special care was taken to match the optical apertures to the 11 wide _ spitzer _
irs - lh slit , in which the [ oiv ] line is located .
this is especially important for the starburst galaxies as many of them are extended .
the most appropriate optical datasets for the majority of galaxies were the integrated spectra obtained by @xcite from which we retrieved the optical spectra data for 13 starbursts and 8 bcds . .
galaxies above kewley s line in figure [ diagram ] are red , between kewley s and kauffmann s line are green , and below kauffmann s line are blue .
the dashed lines are the same as in figure 1 .
the dotted line is equation ( 1 ) , defined empirically as the separatrix between the two branches .
[ ne32_o4s3_dale ] ] to expand our dataset , we also use the data for star - forming galaxies from @xcite , which have matchable _ spitzer _ high - resolution mapping spectra extracted with an aperture of 23@xmath915 , and optical spectra taken with an aperture of 20@xmath920 .
there are 19 objects with both detections in [ oiv]25.89@xmath0 m and optical lines used in the bpt diagram , which are displayed in the figure 2 of @xcite . in our figures [ diagram ] and [ ne32_o4s3_dale ] , we place our samples with optical data , as well as the star - forming galaxies from @xcite , in the corresponding optical log([oiii]/h@xmath2 ) vs. log([nii]/h@xmath1 ) and infrared [ oiv]/[siii ] vs. [ neiii]/[neii ] diagrams.the colors of the data symbols in each diagram are assigned by their corresponding classification in the other diagram ( see figure captions ) . in figure
[ ne32_o4s3_dale ] , the galaxies found above kewley s line in figure [ diagram ] are red , between kewley s and kauffmann s line are green , and below kauffmann s line are blue .
we found that most optically identified strong agns ( marked with red @xmath9 ) and starbursts ( blue squares ) are well separated in the upper and lower branches , especially for values of the [ neiii]/[neii ] ratio greater than @xmath15 .
there are four galaxies optically classified as starbursts ( blue squares ) located in the upper branch in figure [ ne32_o4s3_dale ] .
however , we note that their location in the optical diagram in figure [ diagram ] is near the locus where the two branches merge , an area where classification of individual objects is ambiguous . in the optical diagram of figure [ diagram ] ,
the data points are colored based on their classification in the infrared diagram of figure [ ne32_o4s3_dale ] .
more specifically , we empirically define a line separating the upper and lower branch , having the functional form of : @xmath16/[siii]})=0.08\times log(\rm{[neiii]/[neii]})+0.014\ ] ] galaxies located on the upper branch above the line are plotted in figure [ diagram ] in orange , and below the line in cyan .
this demonstrates that the upper and lower branch of the infrared diagram correspond well to the right and left branch of the optical .
we should stress once more that the optical spectra are collected from various sources and have been obtained with different apertures .
this likely introduces uncertainties in the separatrix mentioned above , which can not be easily quantified . even for datasets where optical and infrared apertures are matched , wavelength dependent
extinction will result in discrepancies between the different classifications , in particular for dusty sources , since the infrared can probe into regions which are optically thick .
therefore , it is not surprising to have some outliers that do not correspond to the same regions in both diagrams .
for example , galaxies located between kewley s and kauffmann s line ( green objects in figure [ ne32_o4s3_dale ] ) may lie either on the upper or lower branch of the infrared diagram .
despite these uncertainties , the overall consistency of the two diagrams is very encouraging , convincing us that the [ oiv]/[siii ] vs. [ neiii]/[neii ] diagnostic along with equation ( 1 ) can be used to classify the energy source of nuclear emission in galaxies , particularly for those with high dust content .
the ratio of the flux of [ oiv]25.89@xmath0 m line with the flux of lower - ionization line has been already proposed to diagnose the dominant source of power in ulirgs @xcite . in the genzel et al .
study , the ratio of [ oiv]25.89@xmath0 m to the [ neii]12.81@xmath0 m ( or [ siii]33.48@xmath0 m ) was plotted against the strength of the 7.7@xmath0 m pah .
the pah features are strong in starbursts and weak in agns , as the pah molecules are easily destroyed by the hard radiation of the accretion disk or outshined by the hot dust continuum emission from the agn torus . in figure
[ eqpah_os3 ] , we plot the [ oiv]/[siii ] ratio as a function of the equivalent width ( ew ) of the 6.2@xmath0 m pah feature .
the pah ews for the galaxies in our sample are collected from @xcite , and @xcite . as was the case with figure [ diagram ] colors of the galaxies in figure [ eqpah_os3 ]
are based on their mid - infrared classification in figure [ ne32_o4s3_dale ] .
those found on the upper branch above the dotted separatrix line are marked in orange , while those on the lower branch are in cyan .
we find that the two color groups define two separate trends , with the agns residing at the tip of one group and the bcds at the tip of the other .
the upper trend is similar to the trend found in the traditional genzel et al .
diagram ( indicated with the upper dashed line ) , while the other trend clearly falls below it .
this suggests that the method to diagnose the fraction of agn and starburst power in the mid - ir for a given source suggested by @xcite , can only be applied to sources which are part of the upper trend in figure [ eqpah_os3 ] .
the upper and lower dashed lines can be represented as equation ( 2 ) and ( 3 ) respectively : @xmath16/[siii]})-log(6.0)=0.19/(log(\rm{ew(6.2)})-log(0.7))\ ] ] @xmath16/[siii]})-log(0.35)=0.51/(log(\rm{ew(6.2)})-log(0.7))\ ] ] 25.89@xmath0m/[siii]33.48@xmath0 m ratio as a function of the 6.2@xmath0 m pah ew .
starburst galaxies ( squares ) , bcds ( diamonds ) and agns ( crosses ) located at the upper and lower branch of figure [ ne32_o4s3_dale ] are plotted in orange and cyan respectively .
filled symbols are again wr galaxies .
two dashed lines are drawn to indicate the trend of the upper and lower branches .
[ eqpah_os3 ] ] @xcite have also proposed a diagnostic diagram based on the [ siii]33.48@xmath0m/[siii]34.81@xmath0 m line ratios as a function of the [ neiii]15.56@xmath0m/[neii]12.81@xmath0 m ratios to distinguish agn from starburst emission .
their method relies on the fact that [ siii ] , having an ip of only 8.2 ev , is a low - ionization emission line and a strong coolant of x - ray - irradiated gas @xcite .
agns usually have stronger [ siii ] emission than starburst galaxies and diagnostic diagrams involving [ siii ] present an advantage in cases of faint sources where [ oiv ] is not easily detected . in figure
[ ssi_neratio ] , we plot the galaxies from our sample on the [ siii]/[siii ] vs [ neiii]/[neii ] diagram , similarly to the figure 9 of @xcite .
the galaxies are divided into four regions , and are classified accordingly in a statistical sense : galaxies located in regions i and ii should be classified as an agn with a @xmath17 confidence interval of 83%-97% and 73%-88% respectively , and galaxies found in regions iii and iv should be classified as a star - forming systems with a @xmath17 confidence interval of 84%-93% and 91%-98% , respectively . in figure
[ ssi_neratio ] , the galaxies are also colored the same way as in figure [ diagram ] and figure [ eqpah_os3 ] .
we find that agns and star - forming galaxies situated on the upper branch are preferentially located in regions i and ii , while bcds and star forming galaxies on the lower branch are found in regions iii and iv .
this suggests that the two methods overall agree .
33.48@xmath0m/[siii]34.81@xmath0 m as a function of [ neiii]15.56@xmath0m/[neii]12.81@xmath0 m proposed by dale et al .
the diagram is divided into four regions ( i - iv ) .
galaxies located in regions i and ii should be classified as agn with a @xmath18 confidence interval of 83%-97% and 73%-88% respectively , and galaxies found in regions iii and iv should be classified as star - forming systems with a @xmath17 confidence interval of 84%-93% and 91%-98% , respectively .
the boundaries of the four regions are indicated by lines of the same slope but different offset : log([neiii]/[neii])=8.4log([siii]/[siii]+@xmath19 , where @xmath19=(@xmath203.3,@xmath201.2,@xmath212.5 ) for the dotted , solid and dashed lines .
we include the bcds ( diamonds ) , starbursts ( squares ) and agn ( crosses ) of our sample in the diagram .
the coloring scheme follows the one of figures 3 and 5 , so that galaxies located at the upper and lower branch of figure [ ne32_o4s3_dale ] are indicated here in orange and cyan respectively .
starbursts with a [ nev]14.32@xmath0 m detection are indicated with a circle around their corresponding symbols .
filled symbols are wr galaxies .
[ ssi_neratio ] ] as mentioned in the previous section , the [ oiv]25.89@xmath0 m emission in galaxies can not originate from the hii regions around main - sequence stars . instead , it has to be associated either with an agn component , or with wr stars ( crowther 1999 ) and/or shock excitation @xcite .
we have already shown in 3.2 and 3.3 that the [ oiv ] diagram can successfully identify galaxies where an agn contributes to their mid - ir spectrum as they are placed on the upper branch of the figure [ os3_neratio ] . in this section
, we present a preliminary analysis on the nature of the [ oiv ] emission in the starburst galaxies and bcds of the lower branch , focusing on the interpretation of models which include wr stars and shocks . more detailed study on this subject will be presented in an upcoming paper .
some of the sources in our sample are classified as wolf - rayet galaxies .
they show features of wr stars in their optical spectra , most commonly the broad heii@xmath22 emission which has an ip of @xmath23ev , similar to [ oiv ] . therefore , identifying correlations between wr features and [ oiv ] emission could , in principle , reveal the contribution of those stars .
we cross - matched our bcds with the wr galaxy catalog presented in @xcite and found that seven bcds ( izw18 , sbs0335 - 052e , um461 , iizw40 , ngc1569 , mrk1450 , and ngc1140 ) as well as two starbursts ( ngc1614 and ngc7714 ) have reported wr features . unfortunately , this wr galaxy sample is small , and there are no consistent measurements of the optical wr features from the literature .
thus a direct correlation of the wr strength and the [ oiv ] line emission could not be established . in figure [ os3_neratio ]
we have marked these wr galaxies @xcite as filled symbols .
it appears that at least the detections of wr signatures are not closely associated with high [ oiv]/[siii ] ratios , instead , they appear to be more correlated with high [ neiii]/[neii ] ratios .
another way to probe the contribution of the wr stars to the [ oiv ] emission is by photo - ionization modeling .
we perform a tentative exploration of the contribution of those stars to the presence of the o@xmath13 ion in our lower branch galaxies with starburst 99 ( leitherer 1999 ) and cloudy ( ferland 2001 ) .
we use starburst 99 to model the stellar radiation field of an instantaneous burst of star formation aging from 1 to 10 myrs .
all stars have a metallicity of 1/3 z@xmath4 and the follow a salpeter imf . the output from starburst 99
is used as input to cloudy in order to estimate the flux of the mid - ir emission lines .
the models predict that the presence of considerable [ oiv ] emission is directly related to the appearance of the wolf - rayet phases , becoming strong around 4 myrs after the onset of the burst and fading to undetectable levels 2 myrs later .
on the other hand , the [ neiii ] line is related to the presence of massive main sequence o stars but also becomes very weak after 6 myrs .
both [ neii ] and [ siii ] lines are still strong after 10 myrs .
consequently the [ oiv]/[siii ] and [ neiii]/[neii ] line ratios peak around 4 - 5 myrs after the onset of the burst with values of @xmath140.6 and 40 respectively .
the value of 0.6 for the [ oiv]/[siii ] line ratio agrees with the highest value observed in the galaxies of the lower branch in figure [ os3_neratio ] , but the predicted [ neiii]/[neii ] ratio is about a factor of 4 higher than the observed value .
clearly a more systematic study exploring thoroughly the parameter space of the models and comparing them to larger sample of galaxies with mid - ir measurements is required to further investigate this discrepancy .
we also study the effects of shocks on the production of [ oiv ] utilizing infrared lines which are strong in regions of shocks , such as the [ feii]25.99@xmath0 m ( see * ? ? ?
the infrared [ feii ] lines have been suggested to trace well of the supernova activity ( * ? ? ?
* see also ohalloran et al .
the [ feii ] ionic abundances in supernova remnants are orders of magnitudes larger than the typical values found in orion - like hii regions ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
this is primarily due to shock destruction of grains .
since iron is normally depleted onto the grains , the abundance of iron in the ism is significantly increased in supernova remnants .
furthermore , the cooler and denser gas behind a fast shock generates more low - ionized fe@xmath24 compared to what is found in typical hii regions . following the analysis of @xcite to evaluate the effects of shocks in ngc6240 , we plot in figure [ o4s3_fe2o4 ] the ratio of [ feii]25.99@xmath0m/[oiv]25.89@xmath0 m as a function of [ oiv]25.89@xmath0m/[siii]33.48@xmath0 m .
also plotted are the measurements of the supernovae remnant rcw103 @xcite .
most bcds and starbursts in our sample show an anti - correlation on the diagram , with no obvious indication of enhanced [ feii ] emission
. however , one galaxy , izw18 , appears to be an outlier compared to other bcds with a clear enhancement of the [ feii ] emission relative to the [ oiv ] emission .
this enhanced [ feii ] emission could indicate a significant contribution of shocks in izw18 .
the influence of shocks to the observed [ oiv ] emission can be better constrained with detailed shock models ( e.g. * ? ? ?
* ) but we defer this work to a future paper on this topic .
overall , our preliminary study described in this section suggests that there is no strong indication that shocks play an important role in the mid - ir emission of starbursts and bcds .
hence , wr stars are the most likely culprit of the [ oiv ] emission seen in the lower - branch galaxies of the figure [ os3_neratio ] .
we do note though the difficulty to simultaneously fit the [ oiv]/[siii ] and [ neiii]/[neii ] ratios with current photo - ionization models .
it appears that the models require mechanisms to significantly suppress [ neiii]/[neii ] ratios for a fixed [ oiv ] emission .
this issue is discussed more in 4 .
25.89@xmath0m/[siii]33.48@xmath0 m as a function of the [ feii]25.99@xmath0m/[oiv]25.89@xmath0 m ratio for the bcds ( diamonds ) and starbursts ( squares ) .
galaxies located on the upper and lower branch of figure [ ne32_o4s3_dale ] are plotted in orange and cyan respectively .
the large asterisk indicated the supernovae remnant rcw03 @xcite , while the filled symbols are wr galaxies .
[ o4s3_fe2o4 ] ] as we have discussed earlier , bcds display a significantly harder radiation field than typical starbursts with high [ oiv]/[siii ] and [ neiii]/[neii ] ratios .
however , no [ nev]14.32@xmath0 m emission has been detected in any of them .
it is known that [ nev ] , with an ip of 97ev , serves as an unambiguous indicator of an agn and the 7 out of our 24 starburst galaxies where
[ nev]14.32@xmath0 m was detected , all host an agn ( see table 1 of * ? ? ?
* ) . in figure [ ne52_neratio ]
, we plot the [ nev]14.32@xmath0m/[neii]12.81@xmath0 m ratio as a function of [ neiii]15.56@xmath0m/[neii]12.81@xmath0 m for bcds , starbursts , agns , and ulirgs and find a similar bifurcation signature to the one in figure [ os3_neratio ] .
this also confirms the bifurcation nature of the [ oiv ] diagram discussed in 3.2 and 3.3 .
a similar bifurcation is also seen in the [ nev]14.32@xmath0m/[neii]12.81@xmath0 m vs. [ siv]10.512@xmath0m/[siii]18.712@xmath0 m diagram ( not plotted ) .
14.32@xmath0m/[neii]12.81@xmath0 m as a function of the [ neiii]15.56@xmath0m/[neii]12.81@xmath0 m ratio for bcds , starbursts , agns , and ulirgs .
the symbols are the same as in figure [ os3_neratio ] .
note that since no [ nev ] was detected in any bcd only upper limits are indicated for the corresponding line flux ratios .
[ ne52_neratio ] ]
in addition to the high - ionization emission lines of [ nev ] and [ oiv ] , other common mid - ir emission lines such as [ siv]10.51@xmath0 m , [ siii]18.71@xmath0 m , [ neii]12.81@xmath0 m , and [ neiii]15.56@xmath0 m have been used in order to probe the ionization field in bcds and starbursts , the [ neiii]15.56@xmath0m/[neii]12.81@xmath0 m vs. [ siv]10.51@xmath0m/[siii]18.71@xmath0 m diagram being such an example ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
both @xcite and bernard - salas et al . ( 2009 ) studied this diagram independently for different samples . here
we combine both and also include the data from our comparison sample of agn , ulirgs , and hii regions .
the study is similar to the one presented by @xcite , who also include samples of various galaxy types .
the advantage of the present work is that the measurements are more consistent as all data were obtained with the _ spitzer _ high - resolution modules . /[neii
] vs. [ siv]/[siii ] for bcds , starbursts , agns , galactic / lmc / smc hii regions and ulirgs .
the symbols of bcds , starbursts , agns , and ulirgs are the same as in figure [ os3_neratio ] .
the hii regions are designated with triangles .
most bcds , starburst galaxies , and hii regions are distributed in a narrow band , indicated with two dashed lines.[sratio_neratio ] ] in figure [ sratio_neratio ] , we plot the [ neiii]15.56@xmath0m/[neii]12.81@xmath0 m and the [ siv]10.51@xmath0m/[siii]18.71@xmath0 m ratios for the starbursts , bcds , agns , hii regions , and low - redshift ulirgs . for the hii regions and the ulirg sample , only sources with detections in all four lines are displayed .
in addition , we include the sh mapping observations of the central positions of ngc5253 @xcite , a typical bcd with metallicity of 1/6 z@xmath25 @xcite .
since wr features have also been found in the center of ngc5253 , we designate it with a filled diamond as other bcds with detected wr features . observing the plot
we note that starbursts populate the lower - left corner of the plot ( see also * ? ? ?
bcds in general have higher values in both line ratios than the starbursts , while those with wr signatures systematically have even higher [ neiii]/[neii ] and [ siv]/[siii ] ratios than those without .
the values of [ neiii]/[neii ] in bcds reach as high as 20 , and [ siv]/[siii ] as high as 10 , values which in some cases are even greater than those seen in agns .
bcds , starbursts as well as giant hii regions all lie in a band stretching over 3 orders of magnitude in [ neiii]/[neii ] and [ siv]/[siii ] .
the band , denoted by the two dashed lines , has a slope of @xmath26 , similar to the slope obtained by @xcite . as discussed by @xcite ,
agns and ulirgs are offset from the band displaying slightly higher [ siv]/[siii ] ratios relative to their [ neiii]/[neii ] ratios .
this is very likely due to the harder radiation field in agns which excites [ siv ] more easily than [ neiii ] .
however , we notice that the offset is not as significant as in the [ oiv ] diagram .
this indicates that the [ neiii]/[neii ] and [ siv]/[siii ] ratios are not very sensitive to the excitation mechanisms , and can not be used as diagnostics of the central energy source in a galaxy @xcite .
the higher [ neiii]/[neii ] and [ siv]/[siii ] ratios in bcds compared to average starburst galaxies indicate a harder radiation field in bcds .
this had already been well established ( e.g. * ? ? ?
* ; * ? ? ?
* and references therein ) .
we note , though , that in our sample of bcds and starbursts there is little overlap in the values of the [ neiii]/[neii ] and [ siv]/[siii ] ratios with a separatrix at [ neiii]/[neii]@xmath27 and [ siv]/[siii]@xmath28 . since our bcds have preferentially sub - solar metallicities=1.2@xmath910@xmath29 from feldman & widing ( 2003 ) , ( s / h)@xmath25=1.4@xmath910@xmath30 from asplund et al .
( 2005 ) and ( o / h)@xmath25=4.9@xmath910@xmath29 from allende prieto et al . (
@xcite , while the galaxies in the starburst sample are more metal rich @xcite , their separation in the diagram suggests that metallicity is likely the dominant parameter determining the [ neiii]/[neii ] and [ siv]/[siii ] ratios
. /[neii ] ratio as a function of the ne abundance relative to hydrogen for bcds , hii regions and starbursts with available abundance measurements .
bottom : similar plot for the [ siv]/[siii ] ratio relative to the the s abundance .
the object with the lowest measured ne and s abundance is izw18 and tol 65 @xcite . in both plots
the vertical dashed line indicates the corresponding solar abundance ( see text ) .
[ abund_ne_neratio ] ] to examine this more directly , we use the metallicity measurements of bcds , hii regions , and starbursts from @xcite , @xcite , and bernard - salas et al .
( 2009 ) , and plot in figure [ abund_ne_neratio ] , the [ neiii]/[neii ] ratios vs. the ne abundances and [ siv]/[siii ] ratios vs. the s abundances .
we observe an anti - correlation between excitation and abundances for both ne and s , even though there is a substantial scatter ( see also * ? ? ?
* ; * ? ? ?
as mentioned earlier , the low value of the [ neiii]/[neii ] ratio in starbursts with high metallicities was noted already using iso data ( see also * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , but its implications are still puzzling .
the low ratios apparently suggest a lack of massive stars in the starbursts , in contradiction to the fact that high - mass stars are clearly observed in nearby starbursts ( e.g. * ? ? ?
* ; * ? ? ?
@xcite attributed this to an aging effect : massive starbursts have a shorter lifetime , thus , the hottest stars may not be the dominant source of ionization in the starbursts .
however , @xcite propose that as high - mass stars spend much of their lives embedded within ultra - compact hii regions ( i.e. * ? ? ?
* ) , the high density and high extinction of the ultra - compact hii regions prevents the mid - infrared nebular lines from forming and escaping .
we note that the starbursts with low [ neiii]/[neii ] ratios have fairly high [ oiv]/[siii ] ratios ( see figure [ os3_neratio ] ) .
this puts additional constraints to the issue of low values of observed [ neiii]/[neii ] ratios . as discussed in 3.4 ,
if wr stars are the main source of the [ oiv ] emission , the observed values of the [ oiv]/[siii ] in starbursts argue strongly against a low mass cutoff of the imf in these starbursts , since the wr phase would be too short to sustain the intensity of the [ oiv ] observed .
it will also places some constraints on the scenario that massive stars spend much of their life in ultra - compact hii regions .
the strong extinction and high density will have to somehow significantly affect [ neiii ] and [ neii ] , but not as much the [ oiv ] emission . in figure [ abund_ne_neratio ]
, we notice that the two extremely low metallicity objects , izw18 and tol65 ( see * ? ? ?
* ) , deviate from the anti - correlation of the emission line ratios vs. the abundances .
it appears as if the anti - correlation starts after a maximum value for [ neiii]/[neii ] and [ siv]/[siii ] at @xmath1420 and @xmath148 respectively .
we should note though that extremely low metallicity galaxies are scarce , and their abundance measurements using the infrared fine - structure lines are more challenging due to the non - detection of their hu@xmath1 12.7@xmath0 m line .
however , similar plots using the oxygen abundance determined from optical spectroscopy also shows the same scatter @xcite .
we have presented our analysis of the high spectral resolution mid - ir observations for a sample of 12 bcds obtained with spitzer / irs .
a direct comparison with samples of starburst galaxies , agn , ulirgs , as well as isolated hii regions , enabled us to extend earlier studies of the ionization and excitation in star forming galaxies towards lower metallicities , and to use the observed mid - ir line emission to explore the excitation mechanisms in these systems .
we find that the [ oiv]25.89@xmath0 m line is commonly detected in the bcds , starbursts and agns .
we have demonstrated that the diagram of [ oiv]25.89@xmath0m/[siii]33.48@xmath0 m as a function of [ neiii]15.56@xmath0m/[neii]12.81@xmath0 m can be useful diagnostic to probe the dominant power source in dust enshrouded systems .
star forming galaxies populate two branches in this diagram , with those harboring an agn component located on the upper branch , and those without on the lower one .
this diagram provides a classification consistent with the log([nii]/h@xmath1 ) vs. log([oiii]/h@xmath2 ) diagram in the optical , especially for [ neiii]/[neii ] values above 0.3 .
we empirically define a line ( equation 1 ) separating galaxies with and without agn contribution , which could be used in conjunction with other diagnostics to better determine the presence of an agn in the infrared .
a similar bifurcation is seen in a [ nev]/[neii ] vs. [ neiii]/[neii ] diagram . in the [ neiii]/[neii ]
vs. [ siv]/[siii ] diagram .
agns and ulirgs are only slightly off the band defined by bcds , starbursts , and hii regions , with higher [ siv]/[siii ] values for given [ neiii]/[neii ] ratios . in an effort to probe the contribution of shock excitation to the observed [ oiv]25.89@xmath0 m flux , we compare it with the [ feii]25.99@xmath0 m line , which is a good tracer of supernovae activity .
we found no evidence of shock contribution , demonstrated by an enhanced [ feii ] emission , in starbursts or bcds .
the only exception is izw18 , which may indicate a significant shock contribution in this low metallicity system .
the most likely contributor to the [ oiv ] flux , both in starbursts and bcds , are wr stars . however , current photo - ionization models predict flux ratios of [ neiii]/[neii ] which are too high for the observed [ oiv]/[siii ] values .
this places an additional constrain on the problem of low [ neiii]/[neii ] ratios already known in high - metallicity starbursts , and will be further investigated in future work .
the authors thank the anonymous referee for useful comments which greatly improved the manuscript .
we also thank john moustakas for kindly providing us the optical data of the sings sample .
support for this work was provided by nasa through contract number 1257184 issued by the jet propulsion laboratory , california institute of technology under nasa contract 1407 .
v.c . acknowledges partial support from the eu tok grant 39965 and fp7-regpot 206469 .
lrrrrrrrrrr haro11 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & @xmath70.076 & @xmath70.554 & [email protected] & [email protected] + ngc1140 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & @xmath70.015 & @xmath70.081 & [email protected] & [email protected] + sbs0335 - 052e & [email protected] & @xmath70.014 & [email protected] & [email protected] & [email protected] & @xmath70.150 & @xmath70.012 & @xmath70.042 & [email protected] & [email protected] + ngc1569 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & @xmath70.069 & @xmath70.460 & [email protected] & [email protected] + iizw40 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & @xmath70.054 & @xmath70.286 & [email protected] & [email protected] + ugc4274 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & @xmath70.044 & @xmath70.100 & [email protected] & [email protected] + izw18 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & @xmath70.010 & @xmath70.035 & [email protected] & [email protected] + viizw403 & [email protected] & @xmath70.015 & [email protected] & [email protected] & [email protected] & [email protected] & @xmath70.015 & @xmath70.040 & @xmath70.030 & @xmath70.030 + mrk1450 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & @xmath70.011 & @xmath70.040 & [email protected] & [email protected] + um461 & [email protected] & [email protected] & [email protected] & [email protected] & @xmath70.153 & [email protected] & @xmath70.015 & @xmath70.064 & @xmath70.045 & @xmath70.045 + tol65 & [email protected] & @xmath70.014 & [email protected] & [email protected] & [email protected] & [email protected] & @xmath70.009 & @xmath70.024 & @xmath70.028 & @xmath70.028 + mrk1499 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] & @xmath70.009 & @xmath70.025 & @xmath70.028 & @xmath70.028 + | we present an analysis of the mid - infrared emission lines for a sample of 12 low metallicity blue compact dwarf ( bcd ) galaxies based on high resolution observations obtained with infrared spectrograph on board the _ spitzer _ space telescope .
we compare our sample with a local sample of typical starburst galaxies and active galactic nuclei ( agns ) , to study the ionization field of starbursts over a broad range of physical parameters and examine its difference from the one produced by agn . the high - ionization line [ oiv]25.89@xmath0 m is detected in most of the bcds , starbursts , and agns in our sample .
we propose a diagnostic diagram of the line ratios [ oiv]25.89@xmath0m/[siii]33.48@xmath0 m as a function of [ neiii]15.56@xmath0m/[neii]12.81@xmath0 m which can be useful in identifying the principal excitation mechanism in a galaxy .
galaxies in this diagram split naturally into two branches .
classic agns as well as starburst galaxies with an agn component populate the upper branch , with stronger agns displaying higher [ neiii]/[neii ] ratios .
bcds and pure starbursts are located in the lower branch .
we find that overall the placement of galaxies on this diagram correlates well with their corresponding locations in the log([nii]/h@xmath1 ) vs. log([oiii]/h@xmath2 ) diagnostic diagram , which has been widely used in the optical .
the two diagrams provide consistent classifications of the excitation mechanism in a galaxy . on the other hand , the diagram of [ neiii]15.56@xmath0m/[neii]12.81@xmath0 m vs. [ siv]10.51@xmath0m/[siii]18.71@xmath0 m is not as efficient in separating agns from bcds and pure starbursts .
our analysis demonstrates that bcds in general do display higher [ neiii]/[neii ] and [ siv]/[siii ] line ratios than starbursts , with some reaching values even higher than those found at the centers of agns . despite their hard radiation field
though , no [ nev]14.32@xmath0 m emission has been detected in the bcds of our sample . |
with the measurement of visible linear polarization of field stars , followed by developments of near and far infrared ( respectively nir and fir ) and submillimetric polarimetry , it is now possible to trace magnetic field structures from the scale of the milky way to that of star forming regions and cores . in the diffuse interstellar medium ( ism ) , the general assumption that elongated dust grains usually have their smallest axis aligned in a direction comparable to that of the local magnetic field has lead to a toroidal representation of the galactic magnetic field .
a similar topology of the magnetic field can also sometimes be observed in other galaxies ( e.g. @xcite ) .
on the smaller scales of molecular clouds , a similar relation exists between grain alignment and magnetic field directions ( see @xcite for an exciting review on the subject ) and many topologies of the magnetic field are suspected to exist . among others , helicoidal , toroidal , poloidal and hourglass configurations have been proposed to interpret observed polarization patterns , taking care of the physical conditions in these environments ( e.g. , orion a , m17 ) .
however , one of the most important limitations of linear polarimetry is that we only get a two - dimensional projected average representation of the magnetic field topology because of the integration on the plane of the sky of the contribution to the polarized emission due to each aligned grain , or its dichroic absorption for visible / nir observations .
this leads to a degeneracy problem since many three dimensional ( 3d ) topologies of the field can produce identical polarization patterns .
a method to get rid of this degeneracy problem is the use of models . on the scale of relatively nearby filamentary molecular clouds and cores ,
this method has been used and allows in some cases to select one field topology among many possible ones ( e.g. , @xcite , @xcite ) .
in addition to models , other diagnostics of magnetic fields can be used to lift the degeneracy due to the plane - of - sky integration .
when column densities are high enough , observations of the zeeman effect can yield the line - of - sight component of the magnetic field strength ( e.g. @xcite ) . finally , by determining the ion - to - neutral molecular line - width ratio in non turbulent regions with a linear flow , it is possible to determine the angle between the field direction and the line of sight ( @xcite , @xcite , @xcite ) . by combining all three methods together , we get the 3d field topology ( e.g. , @xcite for m17 , @xcite for orion a ) .
the observation of zeeman splitting of spectral lines requires bright dense cores , but even in other regions , it is possible to get the 3d geometry of the field even if the field strength is unknown .
these methods can inform us about how magnetic fields can contribute to cloud dynamics , but our ultimate goal is to understand the competition between gravitational , thermal , magnetic , and turbulent forces .
more specifically , magnetic fields and turbulence are both competing to slow down the star formation process , otherwise stars would form much faster than observations show . combining polarization modeling techniques ( e.g. , @xcite ) and predictions for magnetized simulations of turbulence ( e.g. , @xcite , @xcite )
, one can hope to be able to determine the relative contributions of magnetic fields and turbulence . to reach this goal , and to understand how magnetic fields thread the dark filaments , and to put observational constraints on modeling we propose to make comparisons of magnetic fields as seen in and around individual objects .
this approach , namely a multi - scale analysis of the magnetic field , has begun to be done via comparisons of visible , nir , fir and/or submm polarization data in the omc1 ( @xcite ) , omc3 ( @xcite ) and gf 9 regions ( @xcite ) . in these works ,
aspects of the galactic magnetic field extending on tens of degrees are compared with aspects of the clouds probed in the densest regions of the clouds with resolutions @xmath2 , but there is generally a lack of information between these two scales . here
we present observations at an intermediate scale , as much as the presence of field stars allows it , to have a clearer view of the topology of the field at the interface ism / filamentary molecular clouds . from a general point of view
, this approach should provide new insights on the way magnetic fields and clouds are dynamically coupled and shaped . in this paper
we address this problem for the gf 9 filamentary molecular cloud region .
a forthcoming paper will address the more complex situation for the orion a region .
a description of data available about gf 9 is given in section [ gf9region ] .
observations and observational results are presented in section [ obsres ] .
an analysis and a discussion follow in section [ discuss ] , and our conclusions are given in section [ conclu ] .
the gf 9 dark filamentary structures have been catalogued first as b150 by @xcite , l1082 by @xcite and gf 9 by @xcite .
a first estimate of the distance to these condensations as @xmath3 pc was determined by @xcite ( see @xcite ) but an estimate of @xmath4 pc to the extreme class 0 object gf 9 - 2 based on star counts was reported more recently by @xcite .
the first estimate is reported in catalogs or works on relatively large core samples by @xcite , @xcite , @xcite , @xcite , @xcite , while the second one is reported in works by @xcite , @xcite and @xcite .
assuming @xmath4 pc , gf 9 - 2 ( see figure [ skymap ] ) has a luminosity @xmath5 , an envelope mass @xmath6 , a ratio @xmath7 and a bolometric temperature @xmath8k .
no outflow manifestation , nor structures are detected but infall motions are observed ( @xcite , @xcite , @xcite and @xcite ) .
gf 9 - 2 must be particularly young since it has a h@xmath9o maser but no detected outflow ( @xcite , @xcite ) .
two @xmath10 regions , namely gf 9-core and gf 9-fila , have been observed by @xcite , @xcite .
the two areas are shown in figure [ skymap ] by blue boxes identified by c and f. the gf 9-core region is associated with the iras point source class 0 protostar psc 20503 + 6006 , but also with gf 9 - 2 and an object named [ lm99]351 .
jhk photometry of these two regions probe similar masses but the filament region is consistent with a uniform - density cylindrical cloud , while the core region contains a centrally condensed core with the highest extinction region lying to the north of the iras point source .
co , @xmath11co and cs surveys of these two regions give consistent views with these two pictures , but also reveal infall / outflow motions in order to explain cs line broadening in the core region where n@xmath12 @xmath13 ( assuming @xmath14 pc ) . in the filament region , n@xmath15 @xmath13 ( see table 2 in @xcite ) . co and nh@xmath16 observations of three regions with an infrared association are presented by @xcite .
these sources are iras 20520 + 6003 , 20526 + 5958 and 20503 + 6006 , and they are associated respectively with the ldn 1082a , b and c cores .
an analysis of these data by @xcite shows that motions consistent with uniform rotation in dense cores are present in ldn 1082a and ldn 1082b , while not in ldn 1082c .
however recent works by @xcite show a velocity gradient along the major axis in the core in gf 9 - 2 which could be due to a rotation of the core .
@xmath11co observations by @xcite have also been done at 3 positions in gf 9 . referring to @xcite , objects named [ lm99]349 and [ lm99]350 ( see figure [ skymap ] ) have no associated embedded young stellar objects ( eyso ) while [ lm99]351 ( @xmath17 west of gf 9 - 2 ) and [ lm99]352 have associated eyso .
first @xmath18 @xmath19 m fir polarimetry observations from space with iso in gf 9 were done in the core and filament regions by @xcite .
additional measurements at @xmath20 @xmath19 m on 9 stars shining through the same regions of gf 9 are also reported by @xcite .
finally , 850 @xmath19 m polarimetry observations were done at the jcmt by jane s. greaves in 1998 at one position in the filament .
initially , the relatively small heliocentric distance of gf 9 , its filamentary shape , and potentially interesting fir measurements in this region were the major factors which motivated us to conduct visible polarimetry around these filamentary dark structures .
the observations were carried out on the 1.6 m telescope at the observatoire du mont - mgantic ( omm ) , qubec , canada , between 2000 september and 2003 july using a 8.2 arcsecond aperture hole and a broad band red filter ( rg645 : 7660 central wavelength , 2410 fwhm ) .
polarization data were taken with beauty and the beast , a two - channel photoelectric polarimeter , which uses a wollaston prism , a pockels cell , and an additional quarter wave plate .
the data were calibrated for polarization instrumental efficiency , instrumental polarization ( due to the telescope mirrors ) and zero point of position angles using a prism , non polarized and polarized standard stars , respectively . on average , the instrumental polarization was @xmath21 and was subtracted .
the observational errors were calculated from photon statistics and also include uncertainties introduced by the previously mentioned calibrations .
the final uncertainty on individual measurements of the polarization @xmath22 is usually around @xmath23 . for more details on the instrument and the observational method ,
see @xcite .
the jcmt data archive has been used and a treatment of the data taken in 1998 was done with commonly used scuba data reduction routines .
data were taken in relatively strong water vapour absorption ( @xmath24 ) conditions , and we finally got very low signal to noise polarization ratios and no significant results out of noise were found .
most of the stars were chosen with the gsc interface at the canadian astronomy data center ( cadc ) .
the others were selected directly on digitized sky survey ( dss ) red plates .
all stars observed are compiled in table [ pvisdata ] .
column 1 gives the gsc number .
column 2 shows a gf 9 star designation number used in this paper .
columns 3 and 4 give equatorial coordinates at epoch 2000 .
degrees of polarization and equatorial position angles with their uncertainties are given in columns 5 and 6 respectively .
column 7 gives the visible extinction coefficient @xmath25 from the @xcite atlas .
finally , 1 is tabulated in column 8 when @xmath26 ( meaning @xmath27 ) , otherwise 0 is tabulated in this column .
we note that when a star has no gsc number , not much information is generally available about it .
figure [ polmap ] presents a polarization map of stars for which the polarization degree meets the condition @xmath26 .
this map is to be compared to figure [ skymap ] .
the histogram of the degree of polarization is shown in figure [ phisto ] . if we only consider data for which @xmath28 ( data shown with full lines ) , the mean and the standard deviation of the distribution are @xmath29 and @xmath30 , respectively .
the histogram of the position angles is shown in figure [ anghisto ] . if we only consider data for which @xmath27 ( data shown with full lines ) , the mean position angle and the standard deviation of this well peaked distribution are @xmath31 and @xmath32 , respectively .
this means a standard deviation of 0.445 rad , comparable to those found for clouds with embedded clusters ( see myers @xmath33
goodman 1991b ) .
stars s35 , s70 , s75 , s86 and s92 are identified on both maps in figures [ skymap ] and 2 .
stars s35 and s86 are suspected to have variable linear polarization .
these stars are located in the continuity of two filaments but in gaps where extinction appears to be relatively low .
the data relative to these stars are compiled in table [ vardata ] with the respective julian date ( j.d . ) of each measurement appearing in the last column . the position angle of stars s70 , s75 and s92 differ significantly from the mean direction of the other stars ( @xmath31 with @xmath34 ) .
star s70 is a visual binary whose components fit both in the aperture hole used for the observations ; its position angle is perpendicular to this mean position angle .
thus , we may be in the presence of a physical binary system with circumstellar material , or of a single system presenting intrinsic polarization and a background star .
the difference of the position angles of stars s75 and s92 with the average value may be related to their position relative to the clouds as seen on figure [ skymap ] .
( @xmath35 ) is located north - east of iras 20520 - 6003 at the edge of two filamentary structures mainly parallel .
comparatively , star s92 is also located in a dense region ( @xmath36 ) in a small filamentary structure oriented mostly north - south .
given their large extinction , these two sources could either be deeply embedded in their clouds or be young stellar objects surrounded with disks in which multiple scattering can produce a relatively high degree of polarization ( e.g. , @xcite ) . without multiwavelength data it is difficult to distinguish between these possibilities .
three stars in our sample , gf 9 s1 , gf 9 s2 and gf 9 s15 have measured spectral types and visual magnitudes . however we could not get reliable distances from these values .
the @xmath37 values obtained are negative or yield very large distances . similarly using @xmath37 values from @xcite
, we could not get reliable distances . in order to study the variations of the degree of polarization with visible extinction
, we used the @xcite catalog to determine @xmath25 , in the direction of each star for which polarization measurements are available .
this catalog is the first version of the atlas and catalog of dark clouds derived by using the optical database digitized sky survey i ( dss ) , and applying a traditional star - count technique to 1043 plates contained in the dss . using the visible extinction coefficient @xmath25 listed in table [ pvisdata ] , values of @xmath39 against @xmath40 , where @xmath41 , are shown in figure [ pav ] .
for all stars , except gf 9 s35 and gf 9 s86 ( suspected variables ) , data directly observed at this wavelength are shown with crosses . assuming nir data from @xcite and visible data probe the same types of grains , the serkowski law , @xmath42,\ ] ] where @xmath43 , is used to derive the relation @xmath44 .
this relation is used to convert the h - band data to equivalent @xmath45 data , assuming @xmath46 .
these equivalent data are listed in table [ irtovis ] and are shown with diamonds in figure [ pav ] .
as was previously mentionned by @xcite , no saturation of @xmath22 can be seen at about @xmath47 ( or @xmath48 ) , and @xmath49 , as observed in cold dark clouds in taurus ( see @xcite ) .
data in the visible confirm this analysis but over the whole filament and the region surrounding it , and not just in the core region .
thus , while not universal , it is possible in some regions to probe magnetic fields inside dark clouds via polarimetry of background stars as suggested here by data taken in the visible and in the nir .
as can be seen in figure [ polmap ] and in the histogram of position angles in figure [ anghisto ] , the polarization pattern in the vicinity of gf 9 is relatively well oriented on the plane of the sky with a mean direction of @xmath31 and a standard deviation @xmath50 . since the filament has an incurved concentric shape reminiscent of supernova remnants ( @xcite ) , one can ask if a possible object at the origin of this shape could be located somewhere in the region delimited by the black circle traced in figure [ skymap ] .
a possible candidate seems to be the post agb star glmp 1012 ( see figure [ skymap ] ) but given its distance estimate @xmath51 3 kpc ( @xcite ) compared with those of gf 9 , even by considering @xmath52 km s@xmath53 winds , it is difficult to defend the view that this source can be responsible for the shape of the filaments .
however since the problem is tridimensional , a source at the origin of such a shape could also be located outside of the domain defined by this circle .
maybe such a source exists but has not been detected , or the apparent concentric shape of the dense ism is only a projection effect and other mechanisms , such as turbulent fluctuations in the ism or density waves , are at the origin of its formation .
a quick look in the catalogs reveals no potential candidate in a circle a few degrees in diameter around the position r.a.(2000 ) = 20@xmath54 and dec.(2000 ) @xmath55 .
if now one compares the direction of the inferred magnetic field with the orientation of the filamentary structures on the plane of the sky , it seems that the magnetic field has approximately the same orientation over an arc circle which covers about 135@xmath56 , from a position angle ( p.a . )
= 270@xmath56 to 45@xmath56 as measured from the center of the circle in figure [ skymap ] .
when looking at both maps from east to west in figure [ skymap ] and 2 , a slow rotation of the direction of the field is present , the average directions near the core and near the filament regions differ by about 10@xmath56 ( see table [ means ] ) , significantly under the variation in p.a . of the filament over the whole region . moreover from east to west
, there is a great dispersion in the orientation of the filamentary structures , thus it is hard with this two dimensional representation to see a direct impact of the field on the apparent shape of the filaments .
this is in agreement with the fact that in other regions there is no prefered projected angle on the plane of the sky between the directions of the ambient magnetic field and the orientation of the filamentary structures ( e.g. @xcite , @xcite ) .
see also @xcite for a discussion on this subject .
comparisons of the orientation of the magnetic field according to the visible , ir and fir data around and in the core and the filament regions can be done by looking at the maps in figures [ mapcore ] and [ mapfila ] respectively . in the core region ,
when going from visible to ir , to fir data , one can see an anticlockwise rotation of the p.a . by about 63@xmath56 in total . for each of these two regions , the means and dispersions of the degree of polarization and of the p.a . of the visible , ir and fir data
are compiled in table [ means ] . the p.a . for the fir data
corresponds to the observed polarization , rotated by 90@xmath56 to give the direction of the magnetic field . in figures
[ mapcore ] and [ mapfila ] , only fir data for which the p.a .
is approximately the same when using the klc instrumental polarization and the c - off instrumental polarization ( clemens 1999 ) are shown .
the degree of polarization of the fir data is not considered in the present analysis .
table [ means ] also includes similar information for the whole gf 9 region and for the local galactic scale , to be discussed below .
_ comparison of visible and ir data _
first , visible and ir data do not probe the same regions on the plane of the sky .
visible data probe the more diffuse parts around prestellar cores while the ir data probe the denser phases where protostellar cores are present .
secondly we note that in the core region the ir vectors have approximately the same orientation which implies that the ir polarization is not intrinsic to each star but produced by dichroic absorption by aligned grains .
this statistical argument , however , does not apply in the filament region where only one ir measurement with sufficient signal - to - noise ratio is available .
thirdly , in both cases the resolution used with each technique is comparable .
fourthly , the simple fact that the ` ir ' magnetic field has not the same orientation than the ` visible ' one implies that if the grains observed with these two techniques have identical properties , and if their dichroic absorption properties are the same at both ir and visible wavelengths , then the net polarization produced by aligned grains located in the densest parts of the cloud dominates the net polarization produced by aligned grains located in the foreground diffuse ism .
all these facts , in addition to the dependence of @xmath22 with @xmath25 mentionned in section [ pversusav ] , suggest that there is a real rotation of the magnetic field lines in the core region when moving across the plane of the sky .
_ comparison of ir and fir data _
firstly , ir and fir data probe the same regions on the plane of the sky in both the core and the filament regions .
secondly we note that in the core region the fir vectors are very well aligned but not with the same p.a . than the ir vectors . ir data result from a magnetic field oriented at p.a .
@xmath57 while fir data suggest a magnetic field oriented at p.a .
thirdly , the resolution with these two techniques is not the same .
the fir data resolution is @xmath59 , many times the ` pencil ' beam resolution of the ir and visible data .
finally , fir observations probe the radiation emitted by grains mainly located in the densest parts of the cloud while ir observations probe grains located between the stars and the observer .
when looking at figure [ mapcore ] , we see that ir and fir data cover a common area @xmath60 in size . thus if these two observational techniques probe the same column densities but at two different resolutions
, one should be able to explain the rotation of the magnetic field lines as seen at these two wavelengths .
a first hypothesis is that a given magnetic field morphology leads to several orientations according to the resolution at which it is observed ( see @xcite , but also @xcite ) .
however , here we see that ir data are spatially distributed through fir data , thus by using a simple principle of juxtaposition it seems impossible to reproduce the mean orientation of fir vectors seen through @xmath59 beams by combining several ir vectors with very small cross sections . on the other hand , since the mean position angle of the stars observed in the visible is not the same than the one for stars observed in the ir , these ir - observed stars should not be foreground to the clouds , but rather embedded in or background to the clouds . in what follows we review possible explanations for the shift in position angles .
\(1 ) the stars observed in the ir could be located inside the gf 9 core .
if this was the case , the impact of their presence on their immediate environment should be observed .
iras observations of these stars show that their luminosities satisfy @xmath61 ( see section [ gf9region ] ) , which would argue against them being class 0 sources , i.e. , very young .
therefore , we reject the hypothesis that these ir stars are all embedded in the gf 9 core .
\(2 ) the stars observed in the ir could be background to the gf 9 core region .
this is suggested by their position in the @xmath62 diagram shown in figure [ pav ] .
two possibilities can be invoked to explain the difference in position angles : + ( a ) a first and may be naive approach is assume that grains along the line of sight to the gf 9 core are all at the same temperature , and that both the ir and fir observational techniques probe all grains with the same efficiency .
if this is right , the mean position angles observed at both wavelengths should be similar , except for the known 90@xmath63 difference between them . in this scenario , the difference in position angles is explained by a second cloud or filamentary structure located behind the gf 9 core , and in which dust grains would be aligned differently .
+ ( b ) a second and probably more realistic scenario is to emphasize differences related to the intrinsic nature of both observational techniques .
based on point ( 1 ) mentioned above , the assumption of a dust temperature gradient for an externally heated cloud seems reasonable . in this case
, the fir measurements should be more sensitive to `` cold '' dust grains aligned in the densest regions of the core than to `` hot '' dust grains aligned in the envelope and in the diffuse ism .
on the other hand , ir observations should probe all the grains along the line of sight , independently of their temperature .
thus grains aligned differently in the envelope than in the denser core should both contribute to the ir polarization position angle , explaining the offset in position angles . while scenario ( 2b ) does not reject the possibilty of a second cloud or filament evoked in ( 2a ) , it has the benefit to propose a simple and consistent explanation .
thus we prefer scenario ( 2b ) , in which the intermediate position angles of ir vectors shown in figure [ mapcore ] may result from a vectorial addition of both fir and visible vectors covering this region and its vicinity .
_ comparisons of the magnetic fields orientations with the cores elongations _
we compare in table [ offsets ] the orientations of five protostellar cores spread into the filaments with the magnetic field orientations .
orientations of the cores l1082a , b and c were determined with the @xmath64 lines of nh@xmath16 while orientations of the lm cores are based on the visible dss data for l1082c and visible extinction for lm351 .
] at first sight we see that , except for l1082b , these dark condensations have a tendency to be elongated along the filaments as was noticed by @xcite ( see figures [ skymap ] and [ polmap ] ) .
names , positions and p.a . of these cores are shown in table [ offsets ] in columns 1 through 4 respectively . for objects lm351 and l1082c , located in the core region , and for objects lm349 and lm350 , located in the filament region , the offset , @xmath65 , on the plane of the sky between the p.a .
of each protostellar core and the mean magnetic field orientation probed with visible data is shown in column 5 .
the same is done with the ir and fir measurements in columns 6 and 7 respectively .
these offsets were not estimated for l1082b since the magnetic field direction is undersampled around this object . on larger scales ,
the magnetic field is mostly along the minor axis of the cores , as can be seen in figure [ polmap ] .
however , on smaller scales , if we accept that ir measurements can probe the magnetic field in dense regions as it is suggested by the variations of @xmath22 with @xmath66 shown in figure [ pav ] , we see that a rotation of the field is apparent in the core region . on this smaller spatial scale
the field is well aligned in the core region in a direction mainly parallel to the cores major axes .
turbulence models such as those developed by @xcite show that matter may flow onto cores along filaments with the field being stretched along the filaments .
in such a case , the field would tend to be along the major axis of the elongated low - density structures connected with higher density cores .
however , such models show that the polarization pattern probing dust in the low density structures would be chaotic rather than smooth and well aligned .
moreover , @xcite showed recently that a velocity gradient of 2.3 km s@xmath53 pc@xmath53 is present along the major axis of the core gf9 - 2 but none along its minor axis . therefore , if we assume that sources named lm351 , l1082c and gf 9 - 2 are associated to a common spatial volume traced with different tracers , and if we assume that the velocity gradient along the major axis of gf 9 - 2 is due to rotation of the core , this picture suggests that an original poloidal field has been twisted on small scales by the rotation of the core into a toroidal field .
this would be consistent with a magnetic support model .
on the other hand , this effect does not seem to be present in the filament region where the magnetic field probed with ir data is mainly parallel to the small axes of the cores detected in this area , in approximately the same direction than the magnetic field seen at greater scales .
but on the other hand if fir measurements are reliable , the chaotic fir polarization pattern suggests that turbulence plays an active role in the filament region .
finally comparison of both regions suggests a magnetic field interacting differently , and maybe less with its surrounding medium in the filament than it does in the core region , in agreement with the evolutionary stage of these two regions .
additional ir data would be necessary to confirm this point , and the quality of fir measurements should of course be confirmed in both regions .
the mean orientation of the ambient magnetic field can also be compared to the local galactic magnetic field orientation .
figure [ mapgal ] shows a polarization map in galactic coordinates based on measurements compiled by @xcite with data for which @xmath67 .
the region presents a complicated picture because we are looking down a spiral arm and possibly because of the influence of the cygnus star - forming complex . in their works @xcite and @xcite have used a large set of polarization data and studied the inferred magnetic field geometry in this part of the galaxy . at first sight
, the simple picture of a toroidal magnetic field implies that the magnetic field lines should point in the direction toward @xmath68 . in their statistical study @xcite show that this trend is effectively real and are also able to detect the curvature of the local spiral arm .
as mentionned by @xcite , since gf 9 is at a relatively close distance it is probably threaded by the magnetic field in the local spiral arm .
jones compares the orientation of the mean ir vector in the core region with the visible vectors on a greater scale , and shows that this vector points in the direction of the ` vanishing point ' approximately located at @xmath68 .
we can see this vector denoted by ` ir ' in figure [ mapgal ] .
we also show the mean visible vector ` v ' in this map .
it does not point in the same direction than the ` ir ' vector but it is approximately parallel to its nearest neighbours .
all these facts suggest that it is preferable to use data probing the diffuse parts of the ism in the environments of molecular clouds to compare ambient to the cloud and galactic magnetic field s orientations . in order to estimate
if the core region observed in cs by @xcite is magnetically supercritical or magnetically subcritical , we use the @xcite method to estimate the magnetic field strength .
the magnetic field strength in the plane - of - sky is estimated with the expression : @xmath69 where @xmath70 is the polarization p.a .
dispersion in degrees , @xmath71 is the molecular hydrogen density in molecules @xmath13 , @xmath72 is the fwhm line width in km s@xmath53 , and @xmath73 is the velocity dispersion ( see @xcite ) .
we use only the ir data for the core region since we have no way of estimating the turbulent velocity dispersion for the larger area corresponding to the visible data .
the value @xmath74 is taken from table [ means ] using the jones ir data .
this value is provided by a relatively small set of six measurements for which the mean of the uncertainties is @xmath75 .
if instead of a direct average , we use a weighted average for estimating the dispersion in p.a . , we get @xmath76 . finally , working with the p.a .
directly instead of the vectors , the dispersions are @xmath77 and @xmath78 , weighted and unweighted respectively .
all these values are close to each other and reflect the fact that there is little dispersion in p.a . in the cloud .
therefore , we believe that despite the small number of measurements , we nevertheless have a representative value for the dispersion .
the various parameters that we used for the cs core region and our results are shown in table [ strengths ] . once @xmath79 is derived , we estimate @xmath80 , the observed mass to flux ratio @xmath81 normalized to its critical value . this observed normalized mass to flux ratio is subject to geometrical bias , as explained by @xcite . in the case of estimating the total field from the plane - of - sky component of the field ,
two of the three components are known and the correction factor for @xmath80 is then @xmath82 , which corresponds to @xmath83 given in table [ strengths ] .
this correction is statistical in nature since we do not know the appropriate angle for any individual object .
in fact , there may not be a line of sight component to the field .
using @xmath84 for the dispersion in p.a . is the quadratic sum of the dispersion @xmath85 and the mean uncertainty @xmath75 ] to take into account the uncertainties in the measurements would decrease the magnitude of @xmath79 and increase the magnitude of @xmath80 and @xmath87 by about 50% . for comparison ,
the dispersion in p.a . for the visible data around the core region is @xmath88 ( table [ means ] ) . with this value
( even though we are outside the area for which the cs line - widths have been measured ) , the values of @xmath80 and @xmath83 are larger by about 110% .
these values are probably too large because the visible data cover a much larger area than just the core region .
also , a bias exists in the chandrasekhar - fermi method that tends to cause estimated field strentghs to be too large and thus to underestimate values of @xmath80 ( e.g. @xcite ) . what is the effect of the distance @xmath89 on these results ?
the column density @xmath90 is used to get the gas mass , so @xmath91 , where @xmath92 is the radius of the cloud , assumed to be spherical . the density @xmath93 . since @xmath94 , @xmath95 , @xmath96 and @xmath97 .
if the distance is 200 pc instead of 440 pc as assumed for the values given in table [ strengths ] , @xmath98 would be closer to unity by about @xmath99 .
is there additional possible evidence for a subcritical core ?
@xcite identified two other tests to distinguish between ambipolar diffusion and turbulence as supporting mechanisms for star formation
. if ambipolar diffusion is acting , one would expect an hourglass configuration for the magnetic field .
but the distribution of the ir polarization vectors ( see fig [ mapcore ] ) suggests a smooth field , or even a slight curvature in the opposite direction than expected for an hourglass shape . a third test identified by @xcite ,
if ambipolar diffusion is acting one expects @xmath100 with @xmath101 approaching 1/2 for larger densitites ( @xmath102 @xmath13 ) .
the values derived here are more or less compatible with this relation ( see figure 1 in @xcite ) .
in addition , the comparison of the magnetic field orientations with the elongations of the cores seen in table [ offsets ] is not in general compatible with ambipolar diffusion , since in that case the cores would be perpendicular to the direction of the magnetic field .
taking all the above arguments into account , our data suggest that the gf 9 core region is approximately magnetically critical .
ambipolar diffusion does not appear to be supported .
also , turbulence would produce a randomness in the polarization vectors which is not observed .
the picture presented in section [ tosmall ] of an original poloidal field which is transformed into a toroidal field by rotation appears most plausible .
visible linear polarization data taken at the mont - mgantic observatory on 78 stars in the vicinity of the dark filamentary cloud gf 9 were presented and compared with available nir and fir polarization data covering part of the same region .
our main conclusions are : there is no evidence for saturation of polarization as a function of extinction for both the visible and nir data , suggesting that even in the denser regions the same grain properties and grain alignment mechanisms apply , and that the magnetic field is still being probed adequately there .
all these facts put together suggest that _ in gf 9 the magnetic field changes direction with the spatial scale . _
the field differs by @xmath103 in the gf 9 core compared to the surrounding region .
ambipolar diffusion does not appear to be supported given the current evidence .
the effects expected from turbulence are definitely missing .
a global interpretation of the results is that an original poloidal field could have been twisted into a toroidal configuration by a rotating elongated structure during its collapse .
future work should check if the trend in polarization position angles to rotate counterclockwise with increasing wavelength still continues in the submm .
also , as the lack of saturation of polarization as a function of extinction suggests , if dust grains are magnetically aligned there should be no polarization hole in the submm in the core and the filament regions . as an additional benefit ,
submm polarization would afford a higher spatial resolution than the fir data discussed here .
we thank the conseil de recherche en sciences naturelles et en gnie du canada for supporting this research .
the authors thank j. s. greaves for providing information about observations of gf 9 with scubapol at the jcmt , as well as b. malenfant and g. turcotte for their helpful and friendly support during observations at mont - mgantic observatory .
we also thank an anonymous referee for his constructive comments and helpful suggestions .
this work made extensive use of the simbad database at the canadian astronomy data center , which is operated by the dominion astrophysical observatory for the national research council s herzberg institute of astrophysics .
andr , ph . ; ward - thompson , d. and barsony , m. 2000 , protostars and planets iv , p59 arce , h. g. ; goodman , a. a. ; bastien , p. ; manset , n. and sumner , m. 1998 , apj 499 , l93 barnard , e. e. 1927 , atlas of selected regions in the milky way , carnegie institute of washington bastien , p. and mnard , f. 1990 , apj 364 , 232 benson , p. j. @xmath33 myers , p. c. 1989 , apjss , 71 , 89 chandrasekhar , s. and fermi , e. 1953 , apj 118 , 113 ciardi , d. r. ; woodward , c. e. ; clemens , d. p. ; harker , d. e. and rudy , r. j. 2000 , aj 120 , 393 ciardi , d. r. ; woodward , c. e. ; clemens , d. p. ; harker , d. e. and rudy , r. j. 1998 , aj 116 , 349 clemens , d. ; kraemer , k. and ciardi , d. 1999 , workshop on iso plarisaton observations , eds . r.j .
laureijs @xmath33 r. siebenmorgen .
esa - sp 435 , p.7 crutcher , r. m. 2004 , astrophysics and space science , v. 292 , issue 1 , p. 225 crutcher , r. m. 1999 , apj 520 , 706 crutcher , r. m. ; nutter , d. j. ; ward - thompson , d. and kirk , j. m. 2004 , apj 600 , 279 dobashi , k. ; uehara , h. ; kandori , r. ; sakurai , t. ; kaiden , m. ; umemoto , t. and sato , fumio 2005 , pasj 57 , sp1 , 1 dobashi , k. ; bernard , j. p. ; yonekura , y. and fukui , y. 1994 , apjss 95 , 419 fiege , d. j. and pudritz , r. e. 2000 , apj 544 , 830 froebrich , d. 2005 , apjs 156 , issue 2 , pp .
169 - 177 furuya , r. s. ; kitamura , y ; wootten , a. ; claussen , m. j. and kawabe , r. 2003 , apjss , 144 , 71 furuya , r. s. ; kitamura , y ; shinnaga , h. 2005 , poster presented at protostars and planets v , no 8087 goodman , a. a. ; benson , p. j. ; fuller , g. a. and myers , p. c. 1993 , apj 406 , 528 goodman , a. a. ; jones , t. j. ; lada , e. a. and myers , p. c. 1995 , apj 448 , 748 gsten , r. 1994 , in _ the cold universe _ , eds . th .
montmerle et al .
( editions frontires : gif - sur - yvette ) , p.169 heiles , c. ; goodman , a. a. ; mckee , c. f. and zweibel , e. g. 2000 , protostars and planets iv , p279 heiles , c. 2000 , aj 199 , 923 heiles , c. 1996 , apj 462 , 316 heitsch , f. ; zweibel , e. g. ; mac low , m - m . ; li , p. ; norman , m. l. , 2001 , apj 561 , 800 heyer , m. h. ; vrba , f. j. ; snell , r. l. ; schloerb , f. p. , strom , s. e. , goldsmith , p. f. & strom , k. m. 1987 , apj 321 , 855 hilton , j. @xmath33 lahulla , j.f .
1995 , aass 113 , 325 houde , m. ; bastien , p. ; dotson , j. l. ; dowell , c. d. ; hildebrand , r. h. ; peng , r. ; phillips , t. g. ; vaillancourt , j. e. and yoshida , h. 2002 , apj 569 , 803 houde , m. ; bastien , p. ; peng , r. ; phillips , t. g. and yoshida , h. 2000a , apj 536 , 857 houde , m. ; peng , r. ; phillips , t. g. ; bastien , p. and yoshida , h. 2000b , apj 537 , 245 houde , m. ; phillips , thomas g. ; bastien , p. ; peng , r. and yoshida , h. 2001 , apj 547 , 311 houde , m. ; dowell , c. d. ; hildebrand , r. h. ; dotson , j. l. ; vaillancourt , j. e. ; phillips , t. g. ; peng , r. and bastien , p. 2004
, apj 604 , 717 jones , t. j. 2003 , aj , 125 , 3208 lazarian , a. 2003 , jqsrt 79 - 80 , 881 - 902 lee , c. w. @xmath33 myers , p. c. 1999 , apjss 123 , 233 lynds , b. d. 1962 , apjs 7 , 1 manset , n. and bastien , p. 1995 , pasp 107 , 483 matthews , b. c. ; wilson , d. w. and fiege , j. d. 2001 , apj 562 , 400 myers , p. c. , fuller , g. a. , goodman , a. a. , @xmath33 benson , p. j. 1991a , apj , 376 , 561 myers , p. c. and goodman , a. a. 1991b , apj 373 , 509 padoan , p. ; goodman , a. a. ; draine , b. t. ; juvela , m. ; nordlund , . and
rgnvaldsson , . e. 2001 , apj 559 , 1005 preite - martinez , a. 1988 , a@xmath33a 76 , 317 scarrott , s. m. ; rolph , c. d. ; @xmath33 semple , d. p. 1990
, in galactic and interstellar magnetic fields , iau symposium 140 , edited by r. beck and p. p. kronberg ( kluwer , dordrecht ) , p. 245 serkowski , k. , mathewson , d. l. , @xmath33 ford , v. l. 1975 , apj vol .
196 , p. 261 - 290 .
schleuning , d. a. 1998 , apj 493 , 811 schneider , s. @xmath33 elmegreen , b. g. 1979 , apjss 41 , 87 valle , j. p. ; greaves , j. s. and fiege , j. d.2003 , apj 588 , 910 viotti , n. r. 1969 , mem .
soc . astron .
italiana , 40 , 75 walawender , j. m. , zweibel , e. g. @xmath33 heiles , c. 2001 , baas , 198 , 41.05 wiesemeyer , h. 1998 , asp conf .
series , vol .
, 189 wiesemeyer , h. 1997 . the spectral signature of accretion in low mass protostars .
phd dissertation , university of bonn lrrrrrrrrcc no&91&20:47:23.5&+60:12:59.5&[email protected]&[email protected]&1.61&1 + 0424600447 & 40 & 20:47:25.8 & + 60:17:41.2&[email protected]&[email protected]&1.34&1 + no&120&20:47:34.0&+59:51:54.4&[email protected]&[email protected]&2.09&1 + 0424601275&46&20:47:35.4&+60:03:35.1&[email protected]&[email protected]&1.89&0 + no&92&20:47:47.6&+60:01:02.2&[email protected]&[email protected]&2.14&1 + 0424601151&57&20:47:48.0&+60:06:49.4&[email protected]&[email protected]&1.78&1 + 0424600411&31&20:47:50.2&+60:14:40.6&[email protected]&[email protected]&1.39&1 + 0396300254&48&20:48:01.0&+59:57:01.1&[email protected]&[email protected]&1.89&1 + no&98&20:48:01.1&+60:00:47.1&[email protected]&[email protected]&2.13&0 + 0424600967 & 47&20:48:08.4&+60:00:24.6 & [email protected]&[email protected]&1.66&1 + 0424600359 & 39&20:48:11.5 & + 60:18:52.2&[email protected]&[email protected]&1.14&1 + 0424601339 & 36&20:48:17.9 & + 60:06:26.9 & [email protected]&[email protected]&2.22&1 + 0424600585 & 18&20:48:20.5 & + 60:01:15.2 & [email protected] & [email protected]&1.66&1 + 0424600629 & 17&20:48:25.5 & + 60:10:11.9 & [email protected] & [email protected]&1.97&0 + 0424600469 & 11 & 20:48:27.7 & + 60:32:54.5 & [email protected] & [email protected]&0.64&1 + 0424601143&13 & 20:48:30.0 & + 60:24:59.4 & [email protected] & [email protected]&0.98&1 + 0424601171 & 16&20:48:32.5 & + 60:15:08.4 &
[email protected] & [email protected]&1.43&1 + 0424600521 & 10 & 20:48:40.8 & + 60:32:15.1 & [email protected] & [email protected]&0.66&1 + 0424600727 & 35&20:48:44.2 & + 60:09:52.9 & see table [ vardata]&&2.17 & + 0424601053 & 34&20:48:46.5 & + 60:09:42.4 & [email protected]&[email protected]&2.17&1 + 0424600659&54&20:48:47.1&+60:12:49.4&[email protected]&[email protected]&1.73&1 + 0424600961&37&20:48:47.3&+60:01:45.4 & [email protected]&[email protected]&1.24&1 + 0424601093&53&20:48:51.3&+60:13:46.1&[email protected]&[email protected]&1.69&1 + 0424600621 & 15 & 20:48:51.4&+60:15:27.6 & [email protected] & [email protected]&1.38&1 + 0424600109 & 14 & 20:49:03.7 & + 60:18:57.7 & [email protected] & [email protected]&1.27&1 + 0424601169&52&20:49:07.1&+60:05:05.0&[email protected]&[email protected]&1.17&1 + 0424600721&55&20:49:09.6&+60:18:43.4&[email protected]&[email protected]&1.56&1 + 0424600857 & 9 & 20:49:13.6 & + 60:30:03.6 & @xmath1050.11 & [email protected]&0.75&1 + no&90&20:49:14.1&+60:12:44.5&[email protected]&[email protected]&2.19&1 + 0424600123 & 8&20:49:14.4 & + 60:29:21.0 & 2.70@xmath104 0.12 & [email protected]&0.80&1 + 0424601087 & 38&20:49:15.7&+60:16:26.1 & [email protected]&[email protected]&1.97&1 + 0424600963&51&20:49:18.6&+60:05:50.6&[email protected]&[email protected]&1.17&1 + 0424600281&56&20:49:18.7&+60:20:20.9&[email protected]&[email protected]&1.42&1 + 0396300376 & 19&20:49:33.3 & + 59:56:54.5 & [email protected] & [email protected]&0.82&1 + 0424600981 & 7 & 20:49:33.7 & + 60:27:11.5 & [email protected] & [email protected]&0.85&1 + 0396300020 & 20&20:49:41.5 & + 59:56:15.7 & [email protected] & [email protected]&0.82&1 + 0424601133 & 21&20:49:44.1 & + 60:03:12.3 & [email protected] & [email protected]&0.84&1 + 0424601015 & 6 & 20:50:04.3 & + 60:24:04.9&[email protected] & [email protected]&1.35&1 + 0424601131 & 42&20:50:25.7&+60:24:15.8 & [email protected]&[email protected]&1.56&1 + 0424601183&50&20:50:25.8&+60:10:52.6&[email protected]&[email protected]&1.04&1 + 0424601341 & 33&20:50:31.9&+60:13:00.4 & [email protected]&[email protected]&1.49&1 + 0424601127 & 44&20:50:36.8&+60:26:15.6 & [email protected]&[email protected]&1.06&1 + 0424601041 & 32&20:50:37.2 & + 60:15:38.6 & [email protected]&[email protected]&1.84&0 + no&86&20:50:45.3&+60:20:01.4&see table [ vardata]&&2.36 & + 0424600477 & 41&20:50:46.3&+60:23:26.3&[email protected]&[email protected]&1.39&1 + 0396300338 & 22&20:50:48.3 & + 59:58:22.0 & [email protected] & [email protected]&0.77&1 + 0396300620 & 23&20:50:56.2 & + 59:59:30.6 & [email protected] & [email protected]&0.73&1 + 0424600655&29&20:51:07.8&+60:11:24.2&[email protected]&[email protected]&1.34&1 + 0424601145&27&20:51:11.2&+60:06:26.2&[email protected]&[email protected]&1.06&1 + 0424601213&28&20:51:12.6&+60:08:53.9&[email protected]&[email protected]&1.16&1 + 0424601111 & 4&20:51:12.8 & + 60:25:15.5 & [email protected] & [email protected]&1.33&1 + 0424600103 & 5 & 20:51:16.2 & + 60:33:19.9 & [email protected]&[email protected]&1.14&1 + no&70&20:51:32.3&+60:41:41.3&[email protected]&[email protected]&1.02&1 + 0424701143 & 43&20:51:33.6&+60:25:05.5&[email protected]&[email protected]&1.44&1 + 0424701033&45&20:51:34.7&+60:14:03.5 & [email protected]&[email protected]&2.07&1 + 0424701049&30&20:51:43.7&+60:11:50.0&[email protected]&[email protected]&1.59&0 + 0396300164 & 24&20:51:49.2 & + 59:54:15.5 & [email protected] & [email protected]&1.01&0 + 0424701155&26&20:51:51.7&+60:05:31.0&[email protected]&[email protected]&1.13&1 + 0424701211 & 3 & 20:51:53.4 & + 60:25:18.5&[email protected] & [email protected]&1.36&1 + no&85&20:51:55.4&+60:19:00.0&[email protected]&[email protected]&2.29&0 + 0424701127 & 2 & 20:51:57.1 & + 60:22:44.6&[email protected] & [email protected]&1.87&0 + no&71&20:52:01.8&+60:38:13.2&[email protected]&[email protected]&1.13&1 + 0424701215 & 1 & 20:52:02.7&+60:22:27.9&[email protected] & [email protected]&1.87&0 + 0424701115&25&20:52:04.5&+60:03:55.6&[email protected]&[email protected]&1.11&1 + 0424701085&72&20:52:23.2&+60:32:48.4&[email protected]&[email protected]&1.36&1 + no&75&20:53:25.7&+60:16:32.6&[email protected]&[email protected]&3.29&1 + no&68&20:53:42.0&+60:24:07.8&[email protected]&[email protected]&2.39&1 + no&66&20:53:52.0&+60:26:37.7&[email protected]&[email protected]&2.52&1 + no&59&20:53:52.8&+60:35:44.6&[email protected]&[email protected]&1.45&1 + no&74&20:53:55.6&60:15:51.6&[email protected]&[email protected]&3.20&0 + no&69&20:53:55.7&+60:22:55.9&[email protected]&[email protected]&2.72&1 + no&69bis&20:53:59.0&+60:22:50.0&[email protected]&[email protected]&2.72&1 + no&61&20:54:01.3&+60:39:55.3&[email protected]&[email protected]&1.15&1 + no&77&20:54:27.8&+60:11:16.5&[email protected]&[email protected]&1.93&1 + no&65&20:54:28.8&+60:26:53.5&[email protected]&[email protected]&2.32&1 + 0424700062&62&20:54:31.3&+60:32:40.7&[email protected]&[email protected]&1.90&1 + no&64&20:54:53.0&+60:29:35.7&[email protected]&[email protected]&2.28&0 + 0424700686&73&20:56:15.6&+60:17:41.9&[email protected]&[email protected]&1.94&1 + lrrrr 35&[email protected]&[email protected]&2,452,054.64583 + 35&[email protected]&@xmath1068.7&2,452,141.72222 + 86&[email protected]&[email protected]&2,452,830.75347 + 86&[email protected]&[email protected]&2,452,835.78125 + lrrrr c1&[email protected]&@xmath107&2.34 + c2&[email protected]&@xmath108&2.34 + c3&[email protected]&@xmath109&2.60 + c4&[email protected]&@xmath110&2.34 + c5&[email protected]&@xmath111&2.60 + c6&[email protected]&@xmath112&2.34 + f2&[email protected]&@xmath113&2.19 + lrrrr @xmath114 ( @xmath115)&@xmath116 & @xmath117 & @xmath118 & @xmath119 + @xmath120 ( @xmath121 ) & @xmath122&@xmath123&@xmath124&@xmath125 + @xmath126 ( @xmath115)&@xmath127&@xmath128 & & + @xmath129 ( @xmath121 ) & @xmath130&@xmath131 & & + @xmath132 ( @xmath115 ) & ... & ... & & + @xmath133 ( @xmath121 ) & @xmath134&random & & + lrrrrrrrr lm349&20 49 6.9&60 11 48&47&@xmath135&@xmath136&random&f + lm350&20 49 46.5&60 15 40&49&@xmath137&@xmath138&random&f + lm351&20 51 28.0&60 18 33&43&@xmath139&@xmath140&@xmath141&c + lm352&20 53 49.9&60 11 27 & ... & ... & ... & ... & ... + l1082c&20 51 27.4&60 19 00&7&@xmath142&@xmath143&@xmath144&c + l1082a&20 53 29.6&60 14 41 & ... & ... & ... & ... & ... + l1082b&20 53 50.3&60 09 47&48 & ? &
& ... + ll @xmath145 ( h@xmath9 @xmath13 ) & 15000 @xmath104 3000 + @xmath90 ( h@xmath9 @xmath146 ) & 9.9 @xmath147 + @xmath148 & 0.16 + @xmath149 @xmath150 & @xmath151 + @xmath152 & @xmath153 + @xmath154 ( @xmath155 ) & 0.88@xmath104
0.20 + @xmath70 @xmath156 & @xmath157 + @xmath79 ( @xmath158 ) & 170@xmath10456 + @xmath80 & @xmath159 0.44 + @xmath83 & 0.35 + | new visible polarization data combined with existing ir and fir polarization data are used to study how the magnetic field threading the filamentary molecular cloud gf 9 connects to larger structures in its general environment .
we find that when both visible and nir polarization data are plotted as a function of extinction , there is no evidence for a plateau or a saturation effect in the polarization at @xmath0 1.3 as seen in dark clouds in taurus .
this lack of saturation effect suggests that even in the denser parts of gf 9 we are still probing the magnetic field .
the visible polarization is smooth and has a well - defined orientation .
the ir data are also well defined but with a different direction , and the fir data in the core region are well defined and with yet another direction , but are randomly distributed in the filament region . on the scale of a few times the mean radial dimension of the molecular cloud , it is as if the magnetic field were ` blind ' to the spatial distribution of the filaments while on smaller scales within the cloud , in the core region near the iras point source psc 20503 + 6006 , polarimetry shows a rotation of the magnetic field lines in these denser phases .
hence , in spite of the fact that the spatial resolution is not the same in the visible / nir and in the fir data , all the data put together indicate that the field direction changes with the spatial scale . finally , the chandrasekhar and fermi method is used to evaluate the magnetic field strength , indicating that the core region is approximately magnetically critical .
a global interpretation of the results is that in the core region an original poloidal field could have been twisted by a rotating elongated ( core@xmath1envelope ) structure .
there is no evidence for turbulence and ambipolar diffusion does not seem to be effective at the present time . |
the herapdf1.0 parton distribution functions ( pdfs ) were extracted using the combined inclusive cross section data from the h1 and zeus collaborations taken at the hera collider during the hera - i running period 1992 - 1997 @xcite .
these data come from neutral and charged current interactions from both @xmath1 and @xmath2 scattering .
the combination of the h1 and zeus data sets takes into account the full correlated systematic uncertainties of the individual experiments such that the total uncertainty of the combined measurement is typically smaller than @xmath3 , for @xmath4 gev@xmath5 , and reaches @xmath6 , for @xmath7 gev@xmath5 .
these pdfs have been updated to herapdf1.5 ( see lhapdfv6.8.6 , http : projects.hepforge.org / lhapdf ) by including a new preliminary combination of the hera inclusive cross section data from the hera - ii running period 2003 - 2007 @xcite .
these data have greater accuracy at high @xmath8 and high @xmath9 than the hera - i data and thus they serve to decrease the pdf uncertainty in these kinematic regions . in this presentation
we discuss the extension of the herapdf1.5 analysis to include inclusive jet data . the new pdf set which results is called herapdf1.6 @xcite .
the jet data included in herapdf1.6 are : two sets of high-@xmath8 inclusive jet production data from zeus @xcite , low-@xmath8 inclusive jet data from h1 @xcite and normalised high-@xmath8 inclusive jet data from h1 @xcite .
the gluon pdf contributes only indirectly to the inclusive dis cross sections .
however , the qcd processes that give rise to scaling violations in the inclusive cross sections , namely the qcd - compton ( qcdc ) and boson - gluon - fusion ( bgf ) processes , are observed as events with distinct jets in the final state provided that the energy and momentum transfer are large enough .
the cross section for qcdc scattering depends on @xmath10 and the quark pdfs . for hera kinematics , this process dominates the jet cross section at large scales , where the quark densities are well known from the inclusive cross - section data , so that the value of @xmath10 may be extracted without strong correlation to the shape of the gluon pdf .
the cross section for the bgf process depends on @xmath10 and the gluon pdf so that measurements of jet cross sections also provide a direct determination of the gluon density .
perturbative qcd predicts the @xmath8 evolution of the parton distributions , but not the @xmath9 dependence .
parton distributions are extracted by performing a direct numerical integration of the dglap equations at nlo ( nnlo fits are discussed in the presentation of radescu ) .
a parametrised analytic shape for the parton distributions is assumed to be valid at some starting value of @xmath11 . for the herapdf the value @xmath12gev@xmath5 is chosen such that the starting scale is below the charm mass threshold , @xmath13 .
then the dglap equations are used to evolve the parton distributions up to a different @xmath8 value , where they are convoluted with nlo coefficient functions to make predictions for the structure functions .
the heavy quark coefficient functions are calculated in the general - mass variable - flavour - number scheme of @xcite , with recent modifications @xcite .
the heavy quark masses for the central fit were chosen to be @xmath14gev and @xmath15gev and the strong coupling constant was fixed to @xmath16 .
the predictions are then fitted to the combined hera data sets for nc and cc @xmath17 and @xmath18 scattering .
a minimum @xmath8 cut of @xmath19 gev@xmath5 was imposed to remain in the kinematic region where perturbative qcd should be applicable . for the jet data the nlo cross - sections
are made using nlojet++ @xcite but such calculations are too slow to be used in an iterative qcd fit hence the programme fastnlo @xcite is used to make these calculations by convoluting pre - calculated weights with the pdfs and @xmath20 .
the fit parameters are those necessary to specify the input analytic shape . just as for herapdf1.0 and 1.5 the valence quark distributions @xmath21 , @xmath22 , and the @xmath23-type and @xmath24-type anti - quark distributions @xmath25 , @xmath26 ( @xmath27 , @xmath28 )
are parametrised at the input scale @xmath29gev@xmath5 by the generic form @xmath30 the parametrisation for the gluon distribution @xmath31 is extended , in comparison to the herapdf1.5 nlo parametrisation , to include a term @xmath32 , such that the nlo gluon may become negative at low @xmath9 and low @xmath8 ( however it does not do so in the kinematic region of the hera data ) .
a further significant extension of the parametrisation is that the low-@xmath9 valence parameters , @xmath33 and @xmath34 are no longer set equal .
otherwise the choice of central parametrisation is as for the published herapdf1.0(1.5 ) fits .
briefly , the normalisation parameters , @xmath35 , are constrained by the quark number sum - rules and momentum sum - rule .
the @xmath36 parameters @xmath37 and @xmath38 are set equal , @xmath39 , such that there is a single @xmath36 parameter for the sea distributions .
the strange quark distribution is expressed as @xmath9-independent fraction , @xmath40 , of the @xmath24-type sea , @xmath41 at @xmath42 . the value @xmath43 would render the @xmath44 and @xmath24 quark densities the same , but the value @xmath45 is chosen to be consistent with determinations of this fraction using neutrino - induced di - muon production .
the further constraint @xmath46 , together with the requirement @xmath39 , ensures that @xmath47 as @xmath48 .
for the central fit only the @xmath21 pdf has non - zero @xmath49 and @xmath50 parameters .
the experimental uncertainties on the herapdf are determined using the conventional @xmath51 tolerance , @xmath52 , for @xmath53c.l .
however model uncertainties and parametrisation uncertainties are also considered .
the choice of the heavy quark masses is varied in the ranges , @xmath54gev and @xmath55gev .
the choice of @xmath56 is varied in the range , @xmath57 , and the choice of the strangeness fraction is varied in the range , @xmath58 . the difference between the central fit and the fits corresponding to model variations of @xmath59 , @xmath60 , @xmath40 , @xmath56
are added in quadrature , separately for positive and negative deviations , to represent the model uncertainty of the herapdf1.6 set .
parametrisation variations for which the @xmath49 and the @xmath50 parameters for all the pdfs are freed one at a time are considered and variation of the starting scale @xmath42 in the range , @xmath61gev@xmath5 , is also considered as parametrisation variation . the difference between these parametrisation variations and the central fit is stored and an envelope representing the maximal deviation at each @xmath9 value is constructed to represent the parametrisation uncertainty .
[ fig : jetnojetfrac ] compares the herapdf1.6 nlo fit including the jet data ( right ) to a the herapdf1.5f nlo fit which does not include these data ( left ) .
note that the herapdf1.5f fit differs from the publically available herapdf1.5 nlo fits in that it has the same flexible parametrisation for the central fit as herapdf1.6 : there are extra parameters for the gluon pdf and the low - x behaviour of @xmath23-valence and @xmath24-valence are not required to be the same . [ cols="^,^ " , ]
hera jet data have been used in addition to hera inclusive data in order to determine the parton distributions in the proton in an nlo qcd fit .
there is no tension between the jet data and the inclusive data .
the pdf set including the jet data , herapdf1.6 , is similar to the publically available herapdf1.5 set both in pdf central values and in uncertainties .
however , the advantage of using jet data is clearly seen when the strong coupling constant @xmath0 is allowed to be a free parameter of the fits .
the uncertainty on the low-@xmath9 gluon pdf due to the correlation with @xmath0 is much reduced when jet data are used and an accurate value of @xmath0 can be obtained : @xmath62 , excluding scale error .
f. d. aaron _ et al . _
[ h1 and zeus collaboration ] , jhep * 1001 * ( 2010 ) 109 [ arxiv:0911.0884 [ hep - ex ] ] .
see talk of v. radescu , these proceedings ; a.m.cooper-sarkar , arxiv:1012.1438 ; v. radescu , pos ( ichep2010 ) 168 ; h1 and zeus collaborations , h1prelim-10 - 142 , zeus - prel-10 - 018 ; h1 and zeus collaborations , h1prelim-10 - 141 , zeus - prel-10 - 017 | hera jet data , in addition to combined hera inclusive data , are input to the herapdf nlo qcd analysis in order to further constrain the gluon pdf and to allow an accurate evaluation of @xmath0 |
in this paper we describe a new theory for the structure and thermodynamics of ionic fluids based on a generalization of the local molecular field ( lmf ) theory we have successfully applied to nonuniform simple fluids.@xcite a basic step in the lmf theory is the replacement of longer ranged and slowly varying parts of the intermolecular interactions by an appropriately chosen effective single particle potential . the structure and thermodynamics of the resulting reference system with shorter ranged intermolecular interactions in the presence of the effective field
is then related to that of the original system of interest .
this strategy seems particularly appropriate for ionic systems since at long distances the coulomb interaction is weak and very slowly varying , and systems with short ranged interactions are significantly easier to treat by theory or simulations .
however coulomb interactions can be very strong and rapidly varying at short distances .
a key question we address is how to divide the coulomb interaction into `` short '' and `` long '' ranged parts so that the lmf theory can give accurate results .
its answer allows for the first _ controlled _ use of the lmf approximation , and we find exceptionally accurate results , even better than those found earlier for fluids with short ranged interactions .
although the most physically interesting applications of these ideas are probably to nonuniform mixtures of size and charge asymmetric ions , in this initial discussion we consider a uniform _ one component charged hard sphere system _ ( occhs ) where almost all the ideas in the lmf theory can be seen in their simplest form .
@xcite the occhs is made up of ( say positively ) charged `` ions '' comprised of hard spheres with a diameter @xmath1 with embedded positive point charges in the presence of a uniform neutralizing background .
the only nontrivial correlations are between the positive ions and for most purposes we can think of this as a one component system with very long ranged repulsive interactions .
a special case is the _ one component plasma _ ( ocp ) where there is no hard core ( @xmath2 ) .
nothing in the theory makes essential use of the simplifications in the occhs .
generalizations to nonuniform and asymmetric models are straightforward in most cases , and equally good results have been found .
we first discuss the qualitative ideas leading to the lmf equation . these will be further developed and made more precise in our discussion of the occhs .
the simplest application of lmf theory relates the structure and thermodynamics of a nonuniform system of interest with a spherically symmetric pair potential @xmath3 in an external field whose value at any point @xmath4 is @xmath5 to those of a _ reference system _ with a shorter ranged pair interaction @xmath6 in a renormalized effective field @xmath7 .
@xmath8 is supposed to be chosen to take account of the averaged effects of the _ perturbation potential _
@xmath9 , where @xmath10 this separation of the intermolecular interaction @xmath11 into two parts can be done in an infinite number of ways , and any choice of @xmath12 defines a possible associated reference system . however the averaging procedure leading to the simple lmf theory can be expected to give very accurate results only for certain properly chosen _ slowly varying _ @xmath12 .
figure 1 gives examples of separations of the repulsive coulomb potential we will use in this paper , parameterized by an important length scale @xmath0 . as explained in detail in section [ sec : lmfocchs ] , when @xmath0 is chosen larger than some state - dependent minimum size @xmath13 , the coulomb perturbation @xmath12 is sufficiently slowly varying that the lmf theory can give very accurate results .
this is the crucial step in developing a simple and accurate theory for ionic systems .
we will refer to the resulting special reference systems with properly chosen @xmath0 as `` mimic systems . '' in realistic models of ionic solutions
there are always strong short ranged repulsive core interactions that must be dealt with in any quantitative theory or simulation .
the mimic system simply treats the short ranged rapidly varying part of the coulomb potential as an additional core - like contribution that generates a modified core interaction .
as we will see , many properties of the full long ranged system can be very accurately described using those of the short ranged mimic system . the separation of the @xmath14 potential into a short ranged piece @xmath15 and a long ranged piece @xmath16 .
a bigger @xmath17 corresponds to a longer ranged mimic system @xmath6 , and a more slowly varying @xmath18 . here
two relevant @xmath17 values are shown for comparison . ] for simple fluids with short ranged interactions the lmf approach has proved most useful when @xmath11 can be divided into a slowly varying perturbation @xmath18 describing the relatively weak and longer ranged attractive interactions and a short ranged rapidly varying core potential @xmath19 , which accounts for the local excluded volume correlations of the particles .
@xcite a separation with these qualitative features suffices to motivate the derivation of the basic lmf equation that follows . for any given @xmath20 an associated @xmath21
could always found in principle so that the nonuniform singlet density @xmath22)$ ] in the reference system ( denoted by the subscript @xmath23 ) equals that in the full system @xmath24)$ ] .
of course the latter is not known in advance and its determination is one of the main goals of the theory . however ,
if a @xmath18 can be chosen to be slowly varying over the range of excluded volume correlations induced by the short ranged potential @xmath6 , then we can make some physically motivated approximations to derive a self - consistent equation to determine the associated @xmath25 at high densities we expect that short ranged correlations in both systems are controlled by packing effects from the identical repulsive cores , and it seems plausible that @xmath8 can be chosen so that both the singlet densities and the conditional singlet densities in the reference and full systems resemble each other .
that is , when @xmath26)=\rho({\textbf{r}};[\phi]),\label{eq : rho0rho}\ ] ] we also expect that @xmath27 ) \simeq\rho({\textbf{r}}_{1}|{\textbf{r}}_{2};[\phi])\label{eq : crho0crho}\ ] ] holds to a good approximation . here
@xmath28)$ ] is the ( conditional ) density at @xmath29 given that a particle is fixed at @xmath30 directly related to the nonuniform pair correlation function . with this assumption
we can derive an equation for @xmath8 that also turns out to give exact results at very low densities , where pair correlations are not important .
as discussed previously,@xcite by subtracting the balance of forces as described by the exact yvon - born - green hierarchy@xcite for the full and reference systems we find a relation between the associated forces @xmath31=
-\int d{\textbf{r}}_{2}\rho_{0}({\textbf{r}}_{2}|{\textbf{r}}_{1};[\phi_{r } ] ) \nabla_{1}u_{1}(r_{12}).\label{eq : wsb}\ ] ] moreover , if @xmath32 is very slowly varying over the range of short ranged pair correlations , then @xmath33 essentially vanishes in the range of integration where @xmath34)$ ] differs significantly from @xmath35)$ ] in eq [ eq : wsb ] .
then we can replace the former by the latter , and take the gradient outside the integral and integrate eq [ eq : wsb ] . in making this replacement
we have ignored correlations between the particles at @xmath4 and @xmath36 , and in most contexts this would represent a crude and generally inaccurate approximation .
however for _ slowly varying _
@xmath37 we see that this particular use of the ( mean field ) approximation can be very accurate , even at high density .
choosing the constant of integration so that the bulk densities in zero field satisfy @xmath38 we arrive at the simple _ local molecular field _
( lmf ) equation @xcite for the effective field @xmath8 : @xmath39)-\rho_{0}^{b}]u_{1}({\textbf{r}}_{12 } ) , \label{eq : lmf}\ ] ] which is the starting point for our work on nonuniform fluids with both short and long ranged interactions . to solve this self consistent equation we need to determine the nonuniform density @xmath35)$ ] in the presence of the effective field @xmath8 .
the lmf approach does not specify or require a particular way to do this .
however since the intermolecular interactions and the effective field have shorter ranges in the reference system , both theory and simulations of the nonuniform structure are usually easier to carry out than in the full system .
equation [ eq : rho0rho ] then provides the fundamental link between structure in the nonuniform reference and full systems , and from this thermodynamic functions can be determined .
the name `` local molecular field '' is suggested by the direct analogy to the spatially varying effective field introduced in the usual mean or molecular field theory for a nonuniform ising model.@xcite however , while the latter theory is usually viewed as a crude approximation , the derivation sketched above suggests that if a proper choice of a slowly - varying @xmath37 is made , then accurate results should be found from a self - consistent solution of the lmf equation in many cases , provided that an accurate treatment of the density response @xmath22)$ ] to a given @xmath8 is used . in this paper
we use the exact eq [ eq : lowrho ] below at low densities and results of computer simulations at higher densities , so whatever errors remain arise only from the lmf approximation itself .
lmf theory is equally useful for uniform fluids .
@xcite in particular eq [ eq : lmf ] is consistent with the physical idea that in a _
dense _ uniform fluid with @xmath40 the forces associated with the slowly varying @xmath37 from oppositely situated particles essentially _ cancel _ in most relevant configurations.@xcite moreover , any residual effects are strongly damped by the small compressibility at high density , so we expect that the radial distribution functions will satisfy @xmath41 as predicted by eqs [ eq : crho0crho ] and [ eq : lmf ] for @xmath42 . for example
, eq [ eq : gog ] holds to a rather good approximation at high density in the uniform lennard - jones ( lj ) fluid provided that the wca separation @xcite with its relatively slowly varying @xmath37 is used , showing the consistency of the physical picture . this perfect cancellation
argument can at best give reasonable results only for uniform fluids at high density .
however , lmf theory can be applied to a general nonuniform fluid , and by taking such a perspective and using only the basic eq [ eq : rho0rho ] , we can significantly improve on the predictions of eq [ eq : gog ] for pair correlations in uniform fluids .
@xcite corrections to eq [ eq : gog ] can be found by considering the particular external field arising from a fluid particle fixed at the origin , @xmath43 .
the induced density now gives @xmath44)=\rho^{b}g(r).\label{eq : rhog}\ ] ] this exact equation@xcite relating the nonuniform singlet density induced by a fixed particle to the radial distribution function @xmath45 in the uniform fluid plays a key role in the theory below .
there are now net unbalanced forces arising from the fixed particle and eq [ eq : lmf ] predicts a nonzero @xmath8 , which can be used in eq [ eq : rho0rho ] to give a more accurate approximation for @xmath46)$ ] and hence @xmath45 . for the uniform lj fluid this approach accurately determines the small corrections to eq [ eq : gog ] at high density.@xcite moreover at very low densities where eq [ eq : gog ] would be very inaccurate , eq [ eq : lmf ] gives @xmath47 and we obtain the _ exact _ low density limit@xcite for @xmath48 $ ] by using eq [ eq : rho0rho ] and the exact low density limit of the reference system in the field @xmath8 : @xmath26)=\rho^{b}e^{-\beta\phi_{r}({\textbf{r}})}.\label{eq : lowrho}\ ] ] here @xmath49 .
similar accurate results for the nonuniform lj fluid have been found for more general external fields representing hard core solutes of various sizes , for the liquid - vapor interface , and for drying transitions.@xcite thus lmf theory has provided a qualitatively and often quantitatively accurate description of structure , thermodynamics , and phase transitions in fluids with short ranged interactions.@xcite
in this paper we focus on the uniform occhs , where there are @xmath50 positive ions in a volume @xmath51 with a uniform neutralizing background that also penetrates the ions .
the pair potential @xmath3 for the ions in the occhs is usually written as @xmath52 the sum of a hard sphere potential @xmath53 and the pair potential @xmath54 arising from point charges of magnitude @xmath55 , where @xmath56 the separation in eq [ eq : wd+wq ] is a special case of eq [ eq : wdiv ] and more general separations of @xmath11 will prove useful in the lmf theory developed below . in eq [ eq : wq ]
the solvent is crudely represented by a uniform static dielectric constant @xmath57 . in the limit @xmath2 , the occhs reduces to the ocp .
it is convenient to introduce a characteristic length describing the typical distance between neighboring particles .
a standard choice is the _ ion sphere radius _ @xmath58 chosen so that @xmath59 the nearest neighbor spacing is about @xmath60 when the ions are arranged in a simple cubic lattice .
thermodynamic properties in the occhs can then be characterized using two dimensionless parameters , the _ _ ionic strength__@xmath61 which compares the bare coulomb interaction energy between two ions separated by the characteristic distance @xmath58 to @xmath62 , and the _ _ hard sphere packing fraction__@xmath63 note that @xmath64 in the ocp @xmath65 and thermodynamic properties depend only on the single dimensionless parameter @xmath66 pair correlations between the ions in the uniform fluid are most conveniently described in terms of the density change induced by fixing a particle at the origin : @xmath67)\equiv\rho(r;[w])-\rho^{b}=\rho^{b}h(r),\label{eq : hdef}\ ] ] where @xmath68 and @xmath69 is the pair correlation function in the uniform fluid .
the unique consequences of the long ranged interaction in the occhs are most easily seen by taking the fourier transform of eq [ eq : hdef ] and defining the dimensionless _ _ structure factor__@xmath70)=1+\rho^{b}\hat{h}(k),\label{eq : s(k)def}\ ] ] where the caret denotes a fourier transform .
as argued generally in the seminal work of stillinger and lovett@xcite ( sl ) , there should be complete screening at long wavelengths of any induced charge distribution in a conducting ionic fluid .
this constrains the behavior at small wavevectors of the charge - charge correlation function . for the occhs the only nontrivial correlations are between the positive ions and the results of sl reduce to the requirement that @xmath71 has the _ universal _ form @xmath72 independent of any details of the short ranged core potential @xmath73 or any other short ranged interactions that might exist . here
@xmath74 is the _ debye wavevector _ , defined by @xmath75 the exact vanishing of @xmath71 at @xmath76 arises from electrical neutrality ( the `` zeroth '' moment condition ) and the fixed coefficient of the quadratic term is an example of the famous sl _
second moment condition_.@xcite this behavior is distinctly different than that found in any fluid with short ranged interactions , where @xmath77 at @xmath76 is finite , proportional to the compressibility , and the coefficient of @xmath78 depends on the details of the intermolecular interactions .
thermodynamic properties can be found by integration of the correlation functions . in particular , accounting for the background by taking the appropriate limits of the standard result for a two component system,@xcite the excess internal energy ( over the ideal gas ) of the uniform occhs can be exactly written as @xmath79
we want to use the general lmf equation to describe the ion correlation functions in the uniform occhs .
it is clear from the derivation in section [ sec : lmf ] that a proper separation of the interaction potential @xmath80 is required for this self - consistent approach to be accurate . at first glance
it may seem natural to use the separation on the right side of eq [ eq : wd+wq ] , where @xmath19 is taken to be the hard core potential @xmath73 and @xmath37 is the full point charge interaction @xmath81 in eq [ eq : wq ] . however at short distances outside the hard core the coulomb potential can be strong and rapidly varying and such interactions can not be accurately treated by the averaging used in eq [ eq : lmf ] .
indeed in the ocp with no hard core there are arbitrarily large and rapidly varying interactions as @xmath82 .
this limit makes it clear that we should try to separate the point charge coulomb pair interaction @xmath54 itself into a slowly varying part @xmath83 , which we will take as a particularly appropriate @xmath18 to use in the lmf theory , and then combine the remainder @xmath84 with @xmath85 ( and more generally with any other short ranged core interactions that exist ) to give the associated @xmath6 .
this strategy differs from that used in many density functional and integral equation methods , where one first chooses a mathematically convenient or especially simple _ reference potential _ @xmath86 and then treats the remainder @xmath87 as a perturbation , taking advantage of the particular form of the reference system in whatever approximate theories are used .
however , for coulomb systems at least , the existing theories often have large and uncontrolled errors with the usual choices of @xmath88 .
we believe our approach of choosing a slowly varying @xmath37 for use in the lmf theory offers many conceptual and computational advantages , and it connects directly to similar physically motivated work on fluids with short ranged interactions . to that end
we interpret the @xmath14 term in @xmath81 as the electrostatic potential of a unit point charge in vacuum .
the same slowly varying asymptotic behavior would come from any other normalized charge distribution and the `` smearing '' of the point charge would produce a less rapidly varying potential at small @xmath89 .
this suggests using a properly chosen charge distribution to determine @xmath83 or @xmath9 .
consider in particular as in the ewald sum method @xcite a normalized unit _
gaussian charge distribution _
@xmath90,\label{eq : psigmar}\ ] ] where @xmath0 indicates the length scale of the smearing .
the particular advantages of this choice will soon become apparent .
our use of a gaussian charge distribution to determine a slowly varying part of the coulomb pair interaction is simpler than in the ewald sum method , which considers periodic images of ion configurations with embedded screening ( negative ) and compensating ( positive ) gaussian charge distributions . @xcite the electrostatic potential @xmath91 arising from eq [ eq : psigmar ] satisfies poisson s equation @xmath92 , which is easily solved by fourier transform to give @xmath93 , \label{eq : vsigmak}\ ] ] or in @xmath89-space , @xmath94 where erf is the usual error function.@xcite the point charge model corresponds to the limit @xmath95 .
thus we can write @xmath96 and use this identity to extract from the dimensionless coulomb pair interaction @xmath97 a @xmath0-dependent perturbation piece @xmath98 : @xmath99 this perturbation remains finite as @xmath100 with @xmath101 .
as illustrated in figure [ cap : potsept ] , with appropriate choices of @xmath0 we can produce a very slowly varying @xmath102 which from eq [ eq : vsigmak ] also decays very rapidly in @xmath103-space . as argued in section [ sec : lmf ] , these are the qualitative features that would be most appropriate for a perturbation @xmath37 to give accurate results from lmf theory .
the choice of @xmath0 in the gaussian charge distribution permits a _ controlled use _ of the lmf approximation , and as shown below , with proper choices of @xmath0 the lmf theory can give exceptionally accurate results .
any particular choice of @xmath0 in eq [ eq : u1sigma ] then fixes the associated reference system interaction@xcite as @xmath104 the coulomb part @xmath105 of the reference interaction decays very rapidly for @xmath106 for large @xmath89 we have @xmath107.\label{eq : u0asymp}\ ] ] we call the special reference systems that result from optimal choices of @xmath0 as discussed in section [ sec : choiceofsigma ] below _ mimic systems _ , since at high density the local structure in the uniform mimic system as exhibited in @xmath108 very accurately approximates the @xmath45 of the full system as in eq [ eq : gog ]
. this again illustrates the consistency and accuracy of the lmf approach when an appropriate mimic system is used .
equation [ eq : u1sigma ] can also be interpreted physically as the coulomb energy arising from two ions each with a rigid gaussian charge distribution , eq [ eq : psigmar ] , with a width @xmath109 .
more generally , in ionic solutions we can always replace point charges on the ions by rigid charge distributions without changing any physics if we appropriately modify the core interactions as in eq [ eq : u0sigma ] . this can be very useful because the rapidly varying short ranged parts of the coulomb interaction can often be more accurately treated by the same specialized methods used for the other strong core interactions , which must be present in any realistic model of ionic solutions .
we now apply the general lmf eq [ eq : lmf ] to the occhs in the special case where the external field @xmath43 is that resulting from an ion fixed at the origin , given by eq [ eq : wd+wq ] .
this choice allows us to describe uniform fluids , as discussed in section [ sec : uniform ] .
we take advantage of spherical symmetry and use the gaussian charge separation of @xmath3 given in eqs [ eq : u1sigma ] and [ eq : u0sigma ] .
the lmf eq [ eq : lmf ] can be naturally rewritten in terms of the more slowly varying `` perturbation part '' of the effective field @xmath110 if the perfect cancellation argument were exact , then @xmath111 , or @xmath112 corresponding to a fixed mimic particle at the origin , and the resulting induced density in the mimic system would be @xmath113)=\rho^{b}h_{0}(r)$ ] , with @xmath114 the pair correlation function in the uniform mimic system .
the lmf equation corrects this approximation by determining a finite short ranged effective field perturbation @xmath115 which we can picture as arising from a modified solute particle at the origin , @xcite that takes better account of the locally averaged effects of the slowly varying interactions @xmath12 .
taking fourier transforms , and introducing for reasons that will soon become apparent a multiplicative parameter @xmath116 that scales the amplitude of @xmath117 , a scaled version of the lmf equation can be written as @xmath118s_{r}(k ) , \label{eq : lmfalpha}\ ] ] where @xmath119)\label{eq : sr(k)}\ ] ] for now we simply note that the original unscaled lmf equation has @xmath120 and view @xmath116 as a parameter at our disposal .
a similar scaling of the lmf equation for systems with short ranged interactions was discussed earlier.@xcite before giving quantitative results in sections [ sec : lowrhoresults ] and [ sec : highdensity ] below , let us discuss some qualitative features of a self - consistent solution of eq [ eq : lmfalpha ] .
such a solution would determine a short ranged effective field , implying a @xmath121 that is finite as @xmath122 along with the associated @xmath123).$ ] by the fundamental assumption of lmf theory in eq [ eq : rho0rho ] , the latter is the lmf approximation to the full @xmath124)$ ] in the occhs .
in particular @xmath125 in eq [ eq : sr(k ) ] is the lmf approximation to @xmath71 in eq [ eq : s(k)def ] , and when no confusion will result , we will simply write @xmath71 .
@xmath125 should be carefully distinguished from @xmath126)$ ] , which equals the structure factor in the _ uniform _ mimic system with @xmath127 in order that @xmath121 remain finite as @xmath128 in eq [ eq : lmfalpha ] , it is clear that with any choice of @xmath129 the associated @xmath130 must vanish identically .
( in practice it is not easy to ensure this in a self consistent iterative solution of eq [ eq : lmfalpha ] , and we give in the appendix in eq [ eq : lmfk ] an alternate but equivalent version that is numerically more stable . )
thus the lmf theory gives an approximate structure factor @xmath125 that always obeys the exact neutrality condition .
its expansion at small @xmath103 has the general form @xmath131 resembling eq [ eq : ssl ] , but the coefficient @xmath132 of @xmath78 depends on @xmath116 and does not necessarily obey the exact second moment condition .
however , by substituting eq [ eq : sr(k)expand ] into eq [ eq : lmfalpha ] , we see that the exact result @xmath133 is found if @xmath116 is chosen self consistently so that@xcite @xmath134 thus by proper choice of @xmath116 , we can guarantee that the approximate structure factor @xmath125 also obeys the exact second moment condition .
we show below in section [ sec : lowrhoresults ] that with optimal choices of the key parameter @xmath0 , even the unscaled lmf theory with @xmath135 often gives very good numerical results .
however it is conceptually important to realize that the lmf approach can be naturally generalized as in eq [ eq : lmfalpha ] so that the exact second moment condition is satisfied , and this adds essentially no numerical costs to the self consistent solution .
we use eq [ eq : lmfalpha ] along with eq [ eq : alpha2nd ] in most of the work reported below , and usually refer to this generalized approach as the lmf theory . if we want to emphasize that the second moment condition is satisfied , we will refer to the lmf2 theory , and distinguish this from the original unscaled version , which we will call the lmf0 theory . the ability to choose a @xmath0 larger than some @xmath13 allows for a consistent application of lmf theory to ionic fluids , ensuring that the lmf approximation is used only for slowly varying parts of the coulomb interactions .
the choice of @xmath13 determines an effective coulomb core size from the core component @xmath105 of the coulomb interaction .
this may be larger or smaller than the `` physical '' core size @xmath136 , which can be varied independently in the occhs . for strong coupling with @xmath137
, we expect considerable cancellation of the very strong forces from ions at distances larger than the nearest neighbor spacing @xmath58 .
thus the effective core size @xmath138 should be of order unity and essentially independent of @xmath139 at large @xmath139 .
however , for weak couplings with @xmath140 we would expect that any choice of @xmath141 with @xmath142 will be sufficient for the lmf equation to give good results .
this ( conservative ) choice of @xmath13 corresponds to the bjerrum length.@xcite only on length scales @xmath13 much less that the average neighbor separation @xmath58 will even the bare coulomb interactions between ions exceed @xmath62 , which would characterize an effective coulomb core size .
a more detailed argument @xcite suggests that we can take even smaller @xmath0 , including @xmath143 , for @xmath140 and still get accurate results from the lmf theory .
we will see below that these qualitative considerations hold true generally .
in particular , by choosing @xmath0 large enough we can guarantee that @xmath144 is nonzero only at small wavevectors , since the gaussian factor in eq [ eq : lmfalpha ] causes rapid decay for @xmath145 .
the gaussian charge distribution produces this very efficient localization of @xmath144 to small wavevectors , and is much superior in this regard to most other smooth distributions .
this property is very important at high density and strong coupling where @xmath125 can have significant structure at @xmath146 where @xmath58 roughly measures the typical distance between nearest neighbor particles . at those larger wavevectors that characterize short ranged structure in @xmath89-space
, @xmath144 essentially vanishes for any choice of @xmath147 @xmath148 @xmath149 thus for such wavevectors we have @xmath150 from a crude linear response argument .
differences in these functions should show up only at small @xmath103 , where @xmath125 will satisfy the sl moment conditions while @xmath77 remains finite as @xmath151 however at high densities the compressibility in the mimic system is small , so that the absolute differences between @xmath125 and @xmath77 remain small even at small wavevectors , as will be illustrated in figure [ cap : occhsg20.sk ] below . on taking inverse transforms
we then expect that @xmath152 holds true to a very good approximation at high densities , as was qualitatively suggested by the cancellation argument leading to eq [ eq : gog ] .
thus we predict a _
family _ of uniform mimic systems for different @xmath147 @xmath153 all of which should give essentially the same short ranged structure at high density that closely approximates that of the full ionic system .
this is a dramatic example showing that the inverse problem of uniquely determining the intermolecular potential from @xmath114 can be ill - conditioned .
a `` molecular - sized '' choice of @xmath154 for the mimic system is considerably smaller than the typical choices made in ewald sum methods , where @xmath0 is usually taken to be proportional to the simulation system size.@xcite larger @xmath0 values will give equally good results , provided that the resulting mimic system is described accurately .
however generally there is little to gain from such choices , since the lmf theory is already consistent for @xmath154 , and it may be more difficult to treat the longer ranged interactions in mimic systems with larger @xmath0 .
thus optimal choices for @xmath0 are generally near @xmath13 . at lower densities and weak coupling
there is little structure in @xmath125 and @xmath77 at larger wavevectors , and we can choose a much smaller @xmath13 as argued above and still make consistent use of the lmf approximation .
very accurate results for @xmath125 are again found from a self - consistent solution of eq [ eq : lmfalpha ] , but the long wavelength perturbations from @xmath144 are not damped by low compressibility in the mimic system , and @xmath114 can differ noticeably from @xmath155 as @xmath156 our theory reduces correctly to the exact debye - hckel ( dh ) limit.@xcite
at low enough bulk densities , the mimic system s response to @xmath157 can be described using the simple ideal gas boltzmann factor as in eq [ eq : lowrho ] , so that @xmath158)=\rho^{b}[e^{-\beta\phi_{r}({\textbf{r}})}-1 ] .
\label{eq : hbz}\ ] ] this also represents the lmf prediction for the full system s @xmath159 , and requires only that second and higher order virial corrections to the _ mimic system s _ pair correlation function can be ignored . when eq [ eq : hbz ] is substituted in the lmf equation [ eq : lmfalpha ] , a closed equation for @xmath21 results .
a self consistent solution is readily found by iteration , using the equivalent but numerically more stable version of the lmf equation in eq [ eq : lmfk ] .
equation [ eq : hbz ] is the same structural approximation that is used in the poisson - boltzmann ( pb ) theory.@xcite indeed if we set @xmath135 and @xmath95 in the lmf equation [ eq : lmfalpha ] combined with eq [ eq : hbz ] , the results reduce _ exactly _ to those of the standard pb theory .
the pb theory thus results from taking the full coulomb interaction of eq [ eq : wq ] as the perturbation @xmath37 in the lmf equation and using the boltzmann approximation for the density response .
we refer to the low density limit of our theory , where eq [ eq : hbz ] is used in eq [ eq : lmfalpha ] , as the _ mimic poisson - boltzmann _
( mpb ) theory .
the mpb theory differs from the pb theory only by the choice of @xmath0 yielding a consistent mimic system along with a choice of @xmath116 that ensures that the second moment condition is exactly satisfied .
as we will see , these simple modifications greatly improve the accuracy and range of validity of the mpb theory .
consider first the ocp , where there is no length scale in the potential to suggest an intrinsic core size .
a qualitative discussion of the choice of the effective size @xmath13 was given above .
in practice it is easy to determine @xmath13 by solving eq [ eq : lmfalpha ] using successively larger values of @xmath160 for @xmath161 the results vary strongly with @xmath0 and are generally very inaccurate . but for all @xmath162 the lmf theory is consistent and should give very similar predictions for the full system s structure as exhibited in @xmath125 , even though the effective fields @xmath144 and the uniform mimic systems structure factors @xmath77 can still vary strongly with @xmath160 this is illustrated in figure [ cap : ocpdiffsigma ] for the state with @xmath163 , where the convergence of the results for @xmath164 and @xmath165 is shown .
the lmf equation itself would continue to give ( even more ) accurate results for larger @xmath0 if an accurate theory for the structure @xmath166)$ ] induced by a given @xmath8 is used .
however the simple boltzmann approximation used in eq [ eq : hbz ] for this quantity must fail at higher densities where there are significant correlations between mimic system particles .
this sets a @xmath167 above which the results of the mpb theory become inaccurate , very roughly estimated by @xmath168 as for hard sphere fluids . as @xmath139 increases , eventually the @xmath13 needed for the accuracy of the lmf approximation exceeds this @xmath167 and the mpb theory fails .
the internal consistency or inconsistency as @xmath0 is varied is very evident from the mpb theory itself . in practice
we find very good and consistent results for all @xmath169 and the slight differences in the @xmath71 curves in figure [ cap : ocpdiffsigma ] and deteriorating results at larger @xmath0 indicate that we are near the upper limit of @xmath139 where the mpb theory can be trusted .
this represents a surprisingly strong coupling , since the lowest order boltzmann approximation for the structure in eq [ eq : hbz ] is used , and shows the virtues of choosing a mimic system .
as shown in figure [ cap : ocp ] , the results of the mpb theory for @xmath163 are in very good agreement with computer simulations,@xcite and compare very favorably to those of the usual ( nonlinear ) pb theory or the generalized debye - hckel ( gdh ) theory developed by levin and coworkers.@xcite note that the mpb theory , unlike the usual pb approximation , can predict oscillations in both @xmath71 and @xmath45 from the self consistent determination of @xmath170 despite using only the lowest structural approximation , eq [ eq : hbz ] .
we now turn to the occhs .
this has more complex structure because of possible competition between correlations induced by the hard cores and the soft repulsive coulomb interactions .
we follow the usual convention where lengths are measured in terms of the hard core diameter @xmath136 . in figure
[ cap : occhs ] we compare the results of the mpb theory to mc simulations , to results of a density functional treatment , and to the pb and dh theories for a low density state with @xmath171 but with a moderate ionic strength @xmath172 the mpb theory gives excellent results with a molecular - sized @xmath173 , noticeably better than those of the considerably more complicated density functional theory.@xcite the right panel in figure [ cap : occhspblim_2ndmntornot ] illustrates the role of @xmath116 for this state , and shows that the self consistent choice of @xmath174 can ensure that the exact second moment condition is satisfied , though on the scale of the graph the differences between the @xmath125 with @xmath135 are hardly visible .
the left panels show that for much weaker coupling with @xmath175 even the usual pb approximation ( with @xmath95 ) gives good results , indicating that at weak coupling the choice of @xmath0 is not so important .
however , unlike the mpb theory , the pb approximation can satisfy the second moment condition exactly only in the limit @xmath156 where it reduces to the dh approximation.@xcite
at high densities we expect that the local structure in @xmath89-space of even the _ uniform _ mimic system , where all corrections from @xmath176 are ignored , will closely resemble that of the full system as suggested in eq [ eq : gog ] .
this is illustrated in figure [ cap : occhsg20.gr ] for the occhs for high density states with moderately strong couplings .
canonical monte carlo simulations are carried out to obtain the uniform mimic system s correlation function @xmath108,@xcite and these are compared to previous simulations for the full system,@xcite where ewald sum methods were used to account for the long ranged interactions . because of the short ranged interactions @xmath19 in the mimic system , our simulations are completely straightforward and no ewald sums or other special treatment of the periodic boundary conditions are required . again
a molecular - sized choice of @xmath177 of order the nearest neighbor spacing suffices .
also shown on the same graphs are the bare ion potential @xmath178 and the mimic potential @xmath179 . despite the much smaller amplitude of the latter and its much shorter range , the mimic
@xmath108 has a striking resemblance to the full system s @xmath180 and the differences are hardly visible on the scale of the graph . at the strongest couplings , the correlation functions at both densities have a first peak shifted away from contact with the embedded hard sphere .
such a correlation function is very different from the correlation function of a hard sphere fluid , which has its maximum at contact , and shows that the strong short ranged parts of the coulomb interactions can compete with packing effects from hard cores even at high density .
this also emphasizes the importance of having the softer piece @xmath181 outside the hard core in our mimic system potential @xmath179 in eq [ eq : u0sigma ] in order to reproduce correlation functions in the occhs , especially for strong coupling states .
there have been several previous empirical attempts to fit correlation functions for coulomb systems at high density using effective short ranged systems .
the dh limit might suggest that a generalized yukawa fluid could be useful , @xcite but the results were not very accurate and there was no systematic way to determine parameters for the effective potentials .
most relevant to our work are ion reaction field ( rf ) methods , where an effective finite - ranged interaction @xmath182 was originally determined from the electrostatic potential of a positive point charge surrounded by a neutralizing uniform spherical charge distribution with a radius @xmath183 .
@xcite good results for correlation functions were found using the rf method in several applications at high density , though some spurious oscillations were seen near the cutoff @xmath183 .
these were attributed to the discontinuity of the second and higher order derivatives of @xmath182 at @xmath183 and better results were found using a smoother `` charge cloud '' distribution that had discontinuities only in fourth and higher order derivatives at the cutoff .
@xcite our @xmath184 in eq [ eq : u0asymp ] can be similarly interpreted in terms of the potential arising from a positive point charge surrounded by a neutralizing gaussian charge distribution .
all derivatives of @xmath184 are continuous because of the smooth cutoff , and by construction the associated perturbation @xmath185 decays very rapidly in @xmath103-space .
it is the latter property that fundamentally leads to mimic system behavior with a proper choice of @xmath0 .
our work thus provides a conceptual framework for understanding why rf methods can work as well as they do in some cases and how results can be significantly improved , especially at lower densities or in nonuniform environments by using the lmf theory .
figure [ cap : occhsmoresig ] gives a more detailed comparison of the structure of the high density / strong coupling state with @xmath186 and @xmath187 to that of two different mimic systems with @xmath177 and @xmath188 . despite the fact that the ( repulsive ) potential of
the latter is always greater than or equal to that of the former , both mimic systems have very similar correlation functions that agree very well with that of the full system , which can be viewed as the limit @xmath189 .
thus for high density states all mimic systems with @xmath162 have essentially the same short - ranged structure in @xmath89-space .
figure [ cap : ocphighg ] gives similar results for the ocp at very strong coupling strengths @xmath190 and @xmath191 we find excellent agreement with simulations of the full system@xcite using a mimic system with @xmath192 as we would expect , for small enough @xmath0 the good agreement fails , as illustrated by results for @xmath193 figure [ cap : occhsg20.sk ] compares the mimic structure factor @xmath77 and a simple estimate for @xmath194 based on a linear response treatment@xcite of the effects of @xmath195 only at very small @xmath103 as revealed in the inset can any differences be seen .
the linear response treatment turns out give an @xmath125 that satisfies exactly both the zeroth and second moment conditions with @xmath135 , and the results converge to @xmath77 at larger @xmath103 controlled by the factor @xmath196 $ ] arising from our choice of a mimic system .
these features would be found in any more exact treatment and suffice at high densities to give a very accurate estimate of @xmath197 this also suggests that the results from other approximate theories may be improved by use of a good mimic system .
indeed , as already shown in figure [ cap : occhsg20.gr ] , the simplest possible theory where the effects from @xmath176 are ignored completely , already gives excellent results for short ranged correlations in @xmath89-space . with accurate approximations for @xmath71 and @xmath198 in hand ,
it is straightforward to calculate thermodynamic properties by integration .
the simplest of these is the internal energy , given by eq [ eq : energy ] .
one can always solve the lmf equation to obtain @xmath198 and use this integral to compute the internal energy of an ionic system , and we would expect very accurate results . here
we show that because of the great similarity of the local structure of the mimic and full systems in @xmath89- and @xmath103-space at _ high density _ , we can obtain an accurate estimate of the energy in terms of the mimic system s energy and a simple analytic correction without explicitly solving the lmf equation .
equation [ eq : energy ] can be rewritten in @xmath103-space and the coulomb interaction separated into mimic and perturbation parts , so that @xmath199\rho^b\hat{h}(k)\nonumber \\ & & + \frac{1}{2}\frac{1}{(2\pi)^{3}}\int d{\textbf{k}}\frac{4\pi\beta q^{2}}{\epsilon k^{2 } } e^{-\frac{1}{4}(k\sigma)^{2}}\rho^b\hat{h}(k).\label{eq : split1}\end{aligned}\ ] ] in the first term of eq [ eq : split1 ] , because @xmath200 differs from @xmath201 only at small @xmath103 , where the factor @xmath202also approaches zero , we can replace the latter by the former with little error . in the second term
, @xmath203 decays very quickly at higher @xmath103 , so only the small-@xmath103 features of @xmath201 @xmath204 from the sl moment conditions are needed to have an accurate estimate of the integration .
the first term then gives the internal energy of the mimic system @xmath205 ( with a background contribution @xmath206 arising from the use of @xmath207 in the integral rather than @xmath208 ) , while the second term corrects the mimic system s energy and can be integrated analytically to give @xmath209 thus the internal energy can be estimated as @xmath210 results for this approximation are compared to mc results in table [ cap : occhse.high ] .[cap : occhse.high]excess internal energy @xmath211 .
the full mc data are taken from references and . at @xmath212 , both the full and the mimic
systems are near solidification , and the results depend on initial conditions . [ cols="^,^,^,^,^,^,^",options="header " , ] for a variety of high density / strong coupling states .
the gaussian charge distribution is the key to the accuracy of eq [ eq : eest ] , both because of its fast decay in @xmath103-space and its use in revealing an excellent mimic system for the full ionic system .
note that at high densities or ionic strengths , the debye wave vector @xmath74 in eq [ eq : kd ] can be very large , making neutrality the major contribution to @xmath213 .
though the physical interpretation of this term is very different , its limiting value is the same as the self - interaction correction in the ewald sum method .
@xcite a similar correction was used in ion rf methods .
@xcite at high density where the compressibility @xmath214 of the mimic system is very small , we can obtain a good estimate of the internal energy by simply replacing @xmath198 by @xmath215 in eq [ eq : energy ] , where because of the background subtraction , a finite result is found for any short ranged @xmath215 . separating the coulomb interaction as in eq [ eq : split1 ] , and using only the @xmath76 result @xmath216 in the second integral we find analogous to eq [ eq : e1 ] : @xmath217 which agrees with eq [ eq : e1 ] in the relevant limit of large @xmath74 and small @xmath214 .
it is straightforward to apply these idea to charge and size asymmetric primitive models@xcite or to `` simple molten salt '' models@xcite with softer repulsive cores .
we have also used the lmf approximation to look at the ocp near a charged hard wall.@xcite here very accurate results are found both in the weak and strong coupling limits , and the more complicated pair level theory recently introduced@xcite is not required .
details will be presented elsewhere , and we only make a few general remarks here . we have found for size asymmetric primitive models that the simplest choice of a single @xmath0 parameter for all species gives excellent results .
this can be understood as a consequence of stillinger and lovett s fundamental insight that for general ionic mixtures universal consequences of the long ranged coulomb interactions can be seen in the small wavevector behavior of the charge - charge correlation function . by using the smeared charge distributions implied by a proper choice of @xmath0 , we arrive at a smeared charge - charge correlation function that has significant structure only at small wave vectors .
the lmf theory can then reproduce the exact long wavelength behavior found by sl , and the slowly varying smeared coulomb perturbations have little effect on the shorter wavelength correlations induced by the modified ion cores in the mimic system , just as illustrated here for the occhs .
the main complication that arises for such mixtures is that the resulting mimic systems will have short ranged attractive as well as repulsive interactions , but this is required if we want the mimic system s structure to resemble that of the full coulomb system .
however , the lmf theory can then to a very good approximation take care of the `` universal '' long ranged parts of the coulomb interactions , which cause major conceptual and computational problems in most approaches , while leaving a nonuniversal , but surprisingly short ranged problem to be treated by whatever means are available .
simulations are straightforward and some recent theoretical developments for treating systems with strong but short ranged interactions look very promising.@xcite we believe the ideas presented here offer a powerful general perspective , and are actively pursuing their consequences for static and dynamic properties of fluids with both short and long ranged interactions .
we are grateful to michael fisher , kirill katsov , michael klein , and lawrence pratt for helpful comments .
we particularly want to thank j. j. weis for sending us his simulation data for the strong coupling occhs correlation functions .
this work was supported by the national science foundation through grant che-0111104 .
the lmf equation [ eq : lmfalpha ] will produce the desired short ranged @xmath170 with a finite value of @xmath218 only if @xmath125 rigorously vanishes at @xmath76 as in eq [ eq : sr(k)expand ] , which of course is the exact result .
however such self consistent equations are usually solved by iteration and any small errors in an intermediate approximation to @xmath125 at small @xmath103 are greatly amplified .
this can lead to numerical instabilities .
the following rewriting of eq [ eq : lmfalpha ] can alleviate this problem .
we can remove the sensitivity at small @xmath103 by multiplying both sides of eq [ eq : lmfalpha ] by @xmath78 , giving @xmath219s_{r}(k).\label{eq : k^2phi}\ ] ] this equation suffices to determine @xmath121 everywhere except near @xmath220 where @xmath117 is assumed to be regular .
we next formally write an identity involving @xmath121 that remains finite as @xmath221 : @xmath222 where @xmath223 is a ( real ) _ constant _ wavevector .
( more generally , we can multiply both sides by a known real function of @xmath103 that does not vanish as @xmath224 we now add these two equations as written and divide by @xmath225 , which yields our final result @xmath226s_{r}(k)\nonumber \\ & & + \frac{k^{2}}{k^{2}+k^{2}}\beta\rho^{b}\hat{\phi}_{r1}(k).\label{eq : lmfk}\end{aligned}\ ] ] this equation has no divergences at small @xmath103 and is stable when iterated , with a choice of @xmath223 of order @xmath74 .
a converged solution will produce a @xmath125 that vanishes identically at @xmath76 with @xmath227 finite . from eq
[ eq : alpha2nd ] this quantity is in fact of order unity when the second moment condition is satisfied . this way of rewriting equations of this kind
was suggested to us by kirill katsov .
this choice of a reference system for the ocp was first suggested in ceperley , d. m. ; chester , g. v. phys . rev .
a * 1977 * , _ 15 _ , 755 .
however they considered too small @xmath17 to observe mimic system behavior and used integral equation methods rather than the lmf theory to try to correct for the long ranged interactions .
using the debye - hckel approximation for the ocp @xmath228 and eq [ eq : kd ] in eq [ eq : lmfalpha ] ( consistent with @xmath229 ) we have @xmath230 $ ] .
the exact @xmath125 has the same form at small @xmath103 ( in accord with the sl moment conditions ) and generally very different behavior at wavevectors larger than @xmath231 that characterize local structure in @xmath89-space . in order to achieve mimic system behavior with @xmath150 at larger wavevectors we require @xmath232 for all such wavevectors .
since @xmath125 is generally of order unity , it should suffice to choose @xmath17 so that @xmath233\ll 1 $ ] . for large @xmath234
we require @xmath17 of order @xmath58 , but for small @xmath235 even @xmath143 will still satisfy this rough criterion .
the @xmath236 at larger wavevectors can be accurately determined by a fourier transform of the simulated @xmath237 , but at small wavevectors we directly sampled the ensemble average @xmath238 we used @xmath239 and @xmath240 particles for the ocp and up to @xmath241 particles for the high density occhs . | structural and thermodynamic properties of ionic fluids are related to those of a simpler `` mimic '' system with short ranged intermolecular interactions in a spatially varying effective field by use of local molecular field ( lmf ) theory , already successfully applied to nonuniform simple fluids . by consistently using the lmf approximation to describe only the slowly varying part of the coulomb interaction , which we view as arising from a rigid gaussian charge distribution with an appropriately chosen width @xmath0 , exceptionally accurate results can be found . in this paper
we study a uniform system of charged hard spheres in a uniform neutralizing background , where these ideas can be presented in their simplest form . at low densities
the lmf theory reduces to a generalized version of the poisson - boltzmann approximation , but the predicted structure factor satisfies the exact stillinger - lovett moment conditions , and with optimal choice of @xmath0 the lowest order approximation remains accurate for much stronger couplings . at high density and strong couplings
the pair correlation function in the uniform mimic system with short ranged interactions is very similar to that of the full ionic system .
a simple analytic formula can then describe the difference in internal energy between the ionic system and the associated mimic system . |
the role of the range of interactions in phase transitions and critical phenomena has been investigated for many years .
for example , there are rigorous theorems which rule out the possibility of long - range ordered phases of a system with only short - range interactions merwag .
it is also well known that mean - field theory becomes exact in the limit of infinite range interactions stanley .
power - law interactions , which fall as @xmath0 with the increasing separation @xmath1 between the interacting elements , can smoothly interpolate between these two extremes anderson70,fisher71 .
the implications of power - law interactions have been explored for several physical systems ; these include , for example , classical kotliar83 and quantum spin glasses dutta02 , ionic systems pitzer85 , complex networks watts98 , etc .
very recently , non - equilibrium phase transitions with long - range interactions have also received attention hinrichsen07 .
power - law interactions can lead to non - trivial and counterintuitive results .
for example , for the @xmath5-dimensional ising model with ferromagnetic power - law interactions , the critical exponents depend on the range parameter @xmath6 and the conventional short - range critical exponents are recovered for relatively large @xmath6 ( typically for @xmath7 ) fisher71 , jenssen97 .
the ferromagnetic to paramagnetic transition in a @xmath5-dimensional ising model with inverse - square interaction @xmath8 is a special type of kosterlitz - thouless transition kosterlitz driven by topological defects kosterlitz76 . for many systems , heuristic reasoning , which are essentially extensions of peierls original arguments peierls ,
have been remarkably successful in predicting the ( im-)possibility of long - range order .
some of these results were later established using rigorous analytical and numerical methods .
examples include predicitions of the lower - critical dimension of the ising model and the impossibility of the existence of a two - dimensional crystaline order at a finite temperature . in this work , employing peierls-type arguments
, we examine the stability of long - range crystalline order in a d - dimensional system with power - law interactions . in other words ,
we consider a crystal at a finite temperature where the atomic oscillators interact via a harmonic potential with a strength that decays algebraically , i.e. , as @xmath0 , with @xmath1 being the seperation between the oscillators .
it is to be noted that , the power - law form of interaction does not have an inherent length scale associated with it .
this is in sharp contrast to the case where the strength of the interaction between two oscillators separated by a distance @xmath1 decays exponentially as @xmath9
. clearly , in the latter case , @xmath10 sets in a length scale over which the interactions are effective and therefore the melting behaviour turns out to be identical to oscillators with only nearest neighbour interactions . the peierls argument generalised to the power - law interactions suggests that the crystalline order persists for a one - dimensional solid even at a finite temperature , provided the range parameter @xmath2 whereas for @xmath11 the crystalline order is destroyed by infinitesimal thermal fluctuations .
the case @xmath12 corresponds to a marginal situation .
we also comment on the stability of the crystalline structure at zero - temperature when the oscillators are quantum mechanical so that quantum fluctuations tend to destroy the crystalline order .
let us consider a one dimensional crystalline arrangement of harmonic oscillators with long - range interactions in the absence of any thermal or quantum fluctuations .
the potential energy for this system(fig.[crystals ] ) in the harmonic approximation can be expressed as @xcite , @xmath13a)]^{2 } \label{lrpot}\ ] ] where @xmath14 is the effective spring constant , @xmath15 is the interatomic seperation and @xmath16 is the index representing the discrete sites on the crystal .
the equation of motion for the atom at the position @xmath17 is given by @xmath18a)+q([n - m]a)-2q(na ) ] , \label{eqnlr}\ ] ] where @xmath19 is the mass of an oscillator .
we assume the usual plane wave solutions of the type @xmath20 , to obtain the dispersion relation of the many - oscillator system given as @xmath21 in the long wavelength limit , i.e. @xmath22@xmath230 , which is relevant in determining the phase transition behaviour , eqn.([displr ] ) reduces to , @xmath24 provided @xmath25 converges . now
in the special case of nearest neighbour short range interactions , we recover the standard phononic kittel dispersion relation , @xmath26 , where @xmath27 is the speed of sound . in the case of long range interactions , on the other hand
, the interaction strength decays with distance as @xmath28 ( @xmath290 ) , then @xmath30 holds for @xmath31 since the sum in eqn.([displrn ] ) converges . for @xmath32 ,
we get ( see appendix i ) @xmath33 it can be shown that the above dispersion relation holds for higher dimensional cases .
we shall restrict our discussion to the range @xmath34 to estimate the marginal dimension or the range of the stability of a crystalline structure .
we shall now examine the effect of thermal fluctuations on the stability of the above solid using the formalism developed by peierls @xcite . in the real space
, the displacement of the atom at the position @xmath17 can be written as a superposition of the different fourier modes as , @xmath35 @xmath36 , where the time dependence has been ignored . making use of the classical equipartition theorem
, we get @xmath37 the real space displacement , @xmath38 is immediately found to be , @xmath39 where @xmath5 denotes the dimensionality of space .
then @xmath40@xmath41@xmath42@xmath43=@xmath40q@xmath42@xmath44@xmath43 gives a measure of the fluctuation in the position of the atom at @xmath17 . to probe the marginality in the melting behaviour , we shall focus on the fluctuations in the displacement of the atom at the origin ( @xmath45=0 ) .
let us first enlist the form of @xmath40@xmath41@xmath42@xmath43 for solids with short range interactions in different spatial dimensions @xmath46 from the above eqn.([del2sr ] ) , we see that @xmath40@xmath41@xmath42@xmath43 diverges for a one and two dimensional solid in the limit @xmath47 .
moreover , the 2-d solid shows a @xmath48 divergence , which means that the divergence is extremely slow .
this log - divergence immediately establishes the spatial dimension @xmath49 as the marginal dimension . for @xmath50 , on the other hand
, @xmath40@xmath41@xmath42@xmath43 is finite in the limit @xmath47 , pointing to the existence of the crystalline order even at a finite temperature in a three - dimensional solid .
let us now extend the above analysis to solids with long range interactions .
the form of the fluctation @xmath40@xmath41@xmath42@xmath43 is enlisted below : @xmath51 the case with @xmath52 turns out to be the most interesting as we observe that for all values @xmath53 , the crystal remains stable even at a finite temperature . at @xmath54 ,
we observe the first signature of a divergence .
therefore , we conclude that @xmath12 happens to be the marginal case , with fluctuations diverging logarithmically with the system size .
this settles @xmath54 as the marginal case for a one - dimensional solid .
if the long range interaction decays faster than @xmath55 , the interaction is feeble and the solid fails to retain the crystal structure even at an infinitesimally small temperature .
in eqn.([realamp1 ] ) , if we put temperature @xmath4 , the average fluctuation @xmath40@xmath41@xmath42@xmath43 vanishes .
we now have to take care of the quantum fluctuations if the harmonic oscillators are quantum mechanical . in that case
, we should now set the average energy of a mode to be the zero - point energy @xmath56 of the corresponding oscillator .
we obtain @xmath57 the above relations lead to a non - trivial modification in the marginal dimension of a solid with quantum mechanical oscillators , as illustrated below . in a spirit similar to that of the previous section ,
let us now examine the @xmath40@xmath41@xmath42@xmath43 for a quantum solid with short range interactions , @xmath58 interestingly , under the effect of quantum fluctuations , both two and three dimensional crystals sustain the long range crystalline order , whereas the one dimensional solid shows a logarithmic divergence in the limit @xmath47 .
hence in the case of quantum oscillators at @xmath4 , @xmath52 is the marginal dimension whereas in higher dimensions a perfect crystalline order exists at absolute zero .
the above finding when contrasted with the melting behaviour of a classical solid with short - range interacting oscillators , an interesting @xmath59 correspondence , with respect to the marginal dimensionality arises .
let us now assume that the quantum oscillators are also long - range interacting as the classical oscillators of the previous section .
a similar line of arguments leads to @xmath60 it is obvious from the above expressions that the question of any instability does not arise even in @xmath52 if @xmath61 .
therefore a one dimensional long - range quantum solid retains the crystalline order for a finite strength of quantum fluctuations , if the decay of the long - range interactions with the distance between the oscillators is sufficently slow .
when @xmath62 , @xmath63 which corresponds to the first sign of a divergence as @xmath64 , and hence the situation is inferred to be marginal .
the strength of the divergence keeps increasing till @xmath65 , when the @xmath40@xmath41@xmath66@xmath43@xmath67@xmath68 .
we have summarised the different stability criteria obtained till now in table i. spatial&classical&quantum&classical&quantum + dimension & ( t@xmath690)&(t=0)&(t@xmath690)&(t=0 ) + d=1&unstable&unstable&stable(1@xmath70@xmath6@xmath702)&stable(1@xmath70@xmath6@xmath703 ) + & & ( @xmath71 ) & unstable(@xmath6@xmath722)&unstable(@xmath71 ) + & & & & for ( @xmath6@xmath733 ) + d=2&unstable & stable&stable(1@xmath70@xmath6@xmath703)&stable + & ( @xmath74)&&unstable(@xmath71 ) & + & & & for @xmath6@xmath733 & + d=3&stable&stable&stable&stable + + [ table2 ]
in this work , we have analyzed the stability of classical and quantum solids using peierls argument , with a special emphasis on the one dimensional situation .
we just note in the passing that mercury chain salts , in which mercury is intercalated into linear chains in @xmath75 to give a compound @xmath76 ( @xmath770.18 at 300k ) , is a system that closely resembles a one dimensional solid @xcite . in the presence of short range interactions
, we never expect the 1-d solid to exist at any finite temperature ( eqn.[del2sr ] ) .
an infinitesimal thermal disturbance is sufficient to destroy long range order .
the existence of long range interactions among the atomic oscillators , where the interaction strength falls off with distance as @xmath0 with @xmath2 , has a non - trivial effect on the stability of the one - dimensional crystal . using a variant of peierls argument
, we have shown that even a one - dimensional solid can sustain the crystalline order at a finite temperature if the range - parameter @xmath6 is sufficiently small ( @xmath2 ) . for @xmath12
, @xmath40@xmath41@xmath42@xmath43 shows a logarithmic divergence and hence this is the marginal case of stability .
this is similar to what is observed also in the case of a ferromagnetic power - law interacting ising model @xcite.that the inverse - square interaction is the marginal dimension even for melting of a long - range solid is an interesting observation .
it should be mentioned that the situation @xmath54 turns out to be a marginal situation in various other long range interacting systems , for example in fracture models with variable range interactions herrman or in networks with long range links moukarzel .
we have also explored the marginal dimension in the case where the oscillators are quantum mechanical . using a generalised version of peierls argument extended to the quantum case , which to our knowledge is new , we observe an interesting @xmath78 correspondence ( table [ table2 ] ) .
a two - dimensional short - range classical solid is marginally stable against thermal fluctuations whereas a one - dimensional short range quantum solid happens to be marginally stable against quantum fluctuations at @xmath4 .
this reminds us of the well known correspondence between the critical behaviour of a @xmath5-dimensional quantum ising model and @xmath79-dimensional classical ising model sachdev .
however , at an infinitesimally small temperature a one - dimensional quantum solid can not retain a long - range order , which means that the quantum effects are irrelevant at a finite temperature as the zero - point energy is vanishingly small compared to the thermal energy @xmath80 .
a unique correspondence emerges out of our study .
we have shown that for a long - range one - dimensional classical solid , @xmath54 is the marginal range of interaction . in the corresponding quantum case , @xmath81 turns out to be the marginal case .
there seems to exist a quantum - classical correspondence in the marginal range ofinteraction , just as in the case of dimension .
one may also wonder about the existence of a tunable parameter in a quantum solid which may be varied to change the strength of the quantum fluctuations .
eqn.([equic ] ) , suggests that the mass of the atomic oscillator is an appropriate candidate for the above as the more massive the atom , the less are the fluctuations or uncertainty in its position .
we should conclude with the comment that in deriving the marginal dimensions or ranges in above mentioned solids , we have used the peierls argument and its generalisation to the quantum case .
although , peierls argument is fairly successful in predicting the marginal dimension or range in several other situations , it fails to throw any light on the transition temperature ( for dimensions above the marginal dimension ) , nature of transitions or role of defects in the transition .
however as we have shown , it predicts interesting results even in the case of melting of classical and quantum solids .
we thank j.k . bhattacharjee for fruitful discussions .
ad acknowledges useful discussions and collaborations with s. m. bhattacharjee and b. k. chakrabarti in related fields .
this work is partially supported by a kvpy fellowship to dc .
when the sum in eqn.([displrn ] ) is divergent , we recast eqn.([displr ] ) in the form of an integral as , @xmath82 where @xmath83=@xmath84 . now the integral is evaluated as follows(@xmath32 ) , @xmath85 since the second integral is just a constant , we obtain eqn.([lrdispsqrt ] ) . for the special case @xmath6=3 , the integrand in eqn.([int2 ] ) diverges at @xmath86 . in this case
, the dipersion relation turns out to be mermin , @xmath87=@xmath88
. 99 merwag n.d .
mermin , h. wagner , phys rev .
17 * , 1133 ( 1966 ) .
stanley , _ phase transition and critical phenomena _ , oxford university press , ( 1971 ) .
g. yuval and p.w .
anderson , phys .
b * 1 * , 1522 ( 1970 ) .
fisher71 m. e. fisher , s. k. ma and b.g .
nickel , phys .
* 29 * , 917 ( 1972 ) .
kotliar83 g. kotliar , p. w. anderson and d. l. stein , phys . rev .
b * 27 * , 602 ( 1983 ) .
dutta02 a. dutta , phys .
b * 65 * , 224427 ( 2002 ) ; a. dutta , physica a*318 * , 63 ( 2003 ) ; a. dutta and r. loganaygam , phys . rev .
b * 75 * , 052405 ( 2007 ) .
pitzer85 k. s. pitzer , m. c. p de lima and d. r. schreiber , j. phys . chem . * 89 * 1854 ( 1985 ) .
watts98 d. j. watts and and s. h. strogatz , nature ( london ) * 393 * , 440 ( 1998 ) .
h. hinrichsen , cond - mat/0703169 ; u. divakaran and amit dutta , physica a * 384 * , 39 ( 2007 ) .
janssen , phys .
e * 58 * , r2673 ( 1998 ) .
j. m. kosterlitz and d.j .
thouless , j. phys .
c * 6 * , 1181 ( 1976 ) .
j. m. kosterlitz , phys .
lett . , * 37 * , 1577 ( 1976 ) ; j. bhattacharjee , s. chakraverty , j. l. richardson and d. j. scalapino , phys . rev .
24 , 3862 ( 1981 ) ; e. luijten and h. messingfeld , phys .
lett . * 86 * , 5305 ( 2001 ) .
r.peierls,_surprises in theoretical physics _ , princeton university press ( 1979 ) .
ashcroft , n.d .
mermin , _ solid state physics _ , holt saunders international editions , ( 1981 ) .
c. kittel , _ quantum theory of solid _ , wiley ( 2nd edition ) , ( 1987 ) .
p.m. chaikin , t.c .
lubensky , _ principles of condensed matter physics _ , cambridge university press , ( 1998 ) .
hidalgo , y. moreno , f. kun and h.j.herrmann , phys .
rev . e*65 * , 046148 ( 2002 ) .
moukarzel and m.a .
de menezes , phys .
e * 65 * , 056709 ( 2002 ) .
s. sachdev , _ quantum phase transitions _ , cambridge university press , ( 1999 ) ; b.k .
chakrabarti , a. dutta and p. sen , _ quantum ising phases and transitions in transverse ising models _ , m41 , springer - verlag , ( berlin , 1996 ) . | we generalise the celebrated peierls argument to study the stability of a long - range interacting classical solid .
long - range interaction implies that all the atomic oscillators are coupled to each other via a harmonic potential , though the coupling strength decays as a power - law @xmath0 , where @xmath1 is the distance between the oscillators .
we show that for the range parameter @xmath2 , the long - range interaction dominates and the one - dimensional system retains a crystalline order even at a finite temperature whereas for @xmath3 , the long - range crystalline order vanishes even at an infinitesimally small temperature .
we also study the effect of quantum fluctuations on the melting behaviour of a one - dimensional solid at @xmath4 , extending peierls arguments to the case of quantum oscillators . |
when moving from set theory to category theory and higher to @xmath2-category theory , we typically weaken equations at one level to natural isomorphisms at the next . these isomorphisms
are then subject to new coherence equations .
for example , multiplication in a monoid is associative , but the tensor product in a monoidal category is , in general , only associative up to a natural isomorphism .
this `` associator '' natural isomorphism has to satisfy some extra equations that are trivial in the case of the monoid . in a similar way , when we move from compact closed categories to compact closed bicategories , the `` zig - zag '' equations governing the dual get weakened to natural isomorphisms and we need to introduce some new coherence laws . in section [ examples ] , we will give several examples of important mathematical structures and how they arise in relation to compact closed bicategories .
following the examples , in section [ previous ] we give the history of the concept and related work .
next , in section [ definition ] we give the complete definition , which to our knowledge has not appeared elsewhere ; we try to motivate each piece of the definition so that the reader could afterwards reconstruct the definition without the aid of this paper . in section [ hoffnung ]
, we prove that a construction by hoffnung is an instance of a compact closed bicategory .
we prove our assertions about the examples in section [ proofs ] .
most of these are corollaries of the main theorem .
in order to get across some of the flavor of these different compact closed bicategories , we will describe the bicategories as well as some weak monoids and monads in them .
the weak monoids and monads play no role in the rest of the paper , but we have found that comparing and contrasting them helps when trying to develop intuition about the bicategories .
the monoids are variations on the notion of an associative algebra , while the monads are variations on the notion of a category . we will give proofs that each of these bicategories ( with the exception of @xmath3 ) are compact closed in section [ proofs ] .
* a * span * from @xmath4 to @xmath5 in a category @xmath0 is an object @xmath6 in @xmath0 together with an ordered pair of morphisms
@xmath7 + \(x ) at ( 2,0 ) @xmath4 ; ( y ) at ( 0,0 ) @xmath5 ; ( s ) at ( 1,1 ) @xmath6 edge [ - > ] node [ above left ] @xmath8 ( y ) edge [ - > ] node [ above right ] @xmath9 ( x ) ; + if @xmath0 is a category with pullbacks , we can compose spans : + \(x ) at ( 2,0 ) @xmath4 ; ( y ) at ( 0,0 ) @xmath5 ; ( s ) at ( 1,1 ) @xmath6 edge [ - > ] node [ above left ] @xmath8 ( y ) edge [ - > ] node [ above right ] @xmath9 ( x ) ; ( w ) at ( 4,0 ) @xmath10 ; ( t ) at ( 3,1 ) @xmath11 edge [ - > ] node [ above left ] @xmath12 ( x ) edge [ - > ] node [ above right ] @xmath13 ( w ) ; ( p ) at ( 2,2 ) @xmath14 edge [ - > ] node [ above left ] @xmath15 ( s ) edge [ - > ] node [ above right ] @xmath16 ( t ) ; ( 1.6 , 1.4)(2,1)(2.4 , 1.4 ) ; + a * map of spans * between two spans @xmath17 and @xmath18 is a morphism @xmath19 making the following diagram commute : + \(x ) at ( 2,0 ) @xmath4 ; ( y ) at ( 0,0 ) @xmath5 ; ( s ) at ( 1,1 ) @xmath6 edge [ - > ] node [ above left ] @xmath8 ( y ) edge [ - > ] node [ above right ] @xmath9 ( x ) ; ( s ) at ( 1 , -1 ) @xmath20 edge [ - > ] node [ below left ] @xmath21 ( y ) edge [ - > ] node [ below right ] @xmath22 ( x ) edge [ < - ] node [ right ] @xmath23 ( s ) ; + since the pullback is associative only up to a natural isomorphism , the same is true of the composition of two spans , so this construction does not give a 2-category ; however , we do get a bicategory span(@xmath0 ) of objects of @xmath0 , spans in @xmath0 , and maps of spans .
+ if @xmath0 is a category with finite products as well as pullbacks , then the bicategory span(@xmath0 ) is a compact closed bicategory where the tensor product is given by the product in @xmath24 a weak monoid object in span(@xmath0 ) is a categorification of the notion of an associative algebra .
for example , one weak monoid in span(set ) is equivalent to the category of polynomial functors from set to itself ; such functors can be `` added '' using disjoint union , `` multiplied '' using the cartesian product , and `` scaled '' by sets @xcite .
a monad in span(@xmath0 ) is a category internal to @xmath0 @xcite . *
sets , relations , and implications form the compact closed bicategory rel , where the tensor product is given by the product in set .
let @xmath1 be the ring of truth values ; then a weak monoid object @xmath25 in rel is a quantale on the powerset of @xmath25 @xcite , while a monad in rel is a preorder . *
a * 2-rig * is a cocomplete monoidal category where the tensor product distributes over the colimits @xcite , though for the purpose of constructing a compact closed bicategory we only need the tensor product to distribute over finite coproducts . given a symmetric 2-rig @xmath1 , mat(@xmath1 ) is the compact closed bicategory of finitely - generated free @xmath1-modules , where the tensor product is the usual tensor product for matrices .
a weak monoid object in mat(@xmath1 ) is a categorified finite - dimensional associative algebra over @xmath26 a monad in mat(@xmath1 ) is a finite @xmath1-enriched category .
a * finite 2-rig * @xmath6 is only finitely cocomplete , but mat(@xmath6 ) is still compact closed .
kapranov and voevodsky @xcite described a bicategory equivalent to mat(finvect ) and called its objects `` 2-vector spaces '' .
* for a category @xmath0 , let @xmath27 be the category of presheaves on @xmath0 .
the 2-category cocont has * * small categories as objects ; * * cocontinuous functors @xmath28 between the categories of presheaves on the source and target as morphisms ; * * natural transformations as 2-morphisms .
+ we can think of cocontinuous functors as being `` set - linear transformations '' , since they preserve sums .
day and street @xcite proved that cocont is a compact closed 2-category . * recall that a profunctor @xmath29 is a functor @xmath30 ; we can think of profunctors as being rather like matrices , where the set @xmath31 is the `` matrix element '' at row @xmath32 and column @xmath33 .
composition of profunctors is given by taking the coend of the inner coordinates , just as matrix multiplication done by summing over the inner index .
small categories , profunctors , and natural transformations form the compact closed bicategory prof , where the tensor product is the product in cat . a weak monoid object in prof
is a promonoidal category @xcite .
a symmetric monoidal monad in prof is a freyd category , also known as an `` arrow '' in the functional programming community @xcite .
+ cattani and winskel @xcite showed that cocont and prof are equivalent as bicategories .
though they do not explicitly state it , the equivalence they construct is symmetric monoidal ; since symmetric monoidal equivalences preserve the dual , cocont and prof are equivalent as compact closed bicategories .
* so far the examples have been rather algebraic in flavor , but there are topological examples , too .
the category @xmath34 is the compact closed category whose * * objects are @xmath35-dimensional manifolds and * * morphisms are diffeomorphism classes of collared @xmath2-dimensional cobordisms between them , + where the tensor product is disjoint union .
atiyah @xcite introduced the category informally in his paper defining topological quantum field theories . + morton @xcite defined the bicategory @xmath36 whose * * objects are @xmath37-dimensional manifolds , * * morphisms are collared @xmath35-dimensional cobordisms , or `` manifolds with boundary '' , and * * 2-morphisms are diffeomorphism classes of collared @xmath2-dimensional maps of cobordisms , or `` manifolds with corners ''
. + the collars are necessary to preserve the smoothness when composing 1- and 2-morphisms .
schommer - pries proved a purely algebraic characterization of @xmath38 , essentially proving the `` baez - dolan cobordism hypothesis '' for the @xmath39 case @xcite .
we expect that @xmath36 will be found to be compact closed , but we will not prove it in this paper . * in a letter to the author , john baez defined two interesting compact closed bicategories . +
a * directed multigraph * is a finite set @xmath40 of edges and a finite set @xmath41 of vertices equipped with functions @xmath42 mapping each edge to its source and target .
a * resistor network * is a directed multigraph equipped with a function @xmath32 assigning a resistance in @xmath43 to each edge : + \(r ) at ( 0,1 ) @xmath43 ; ( e1 ) at ( 1,1 ) @xmath40 edge [ - > ] node [ below ] @xmath32 ( r ) ; ( v1 ) at ( 2,1 ) @xmath41 ; ( e1.15 ) to node [ above ] @xmath44 ( v1.165 ) ; ( e1.345 ) to node [ below ] @xmath45 ( v1.195 ) ; + there are various choices one could make for a morphism of such networks ; baez defined a morphism of resistor networks to be a pair of functions @xmath46 making this diagram commute : + \(r ) at ( 0,.5 ) @xmath43 ; ( e2 ) at ( 1 , 0 ) @xmath47 edge [ - > ] node [ below ] @xmath48 ( r ) ; ( e1 ) at ( 1,1 ) @xmath40 edge [ - > ] node [ above ] @xmath32 ( r ) edge [ - > ] node [ right ] @xmath49 ( e2 ) ; + @xmath50 + ( e2 ) at ( 1 , 0 ) @xmath47 ; ( e1 ) at ( 1,1 ) @xmath40 edge [ - > ] node [ right ] @xmath49 ( e2 ) ; ( v2 ) at ( 2 , 0 ) @xmath51 edge [ < - ] node [ below ] @xmath44 ( e2 ) ; ( v1 ) at ( 2,1 ) @xmath41 edge [ < - ] node [ above ] @xmath44 ( e1 ) edge [ - > ] node [ right ] @xmath52 ( v2 ) ; + @xmath50 + ( e2 ) at ( 1 , 0 ) @xmath47 ; ( e1 ) at ( 1,1 ) @xmath40 edge [ - > ] node [ right ] @xmath49 ( e2 ) ; ( v2 ) at ( 2 , 0 ) @xmath51 edge [ < - ] node [ below ] @xmath45 ( e2 ) ; ( v1 ) at ( 2,1 ) @xmath41 edge [ < - ] node [ above ] @xmath45 ( e1 ) edge [ - > ] node [ right ] @xmath52 ( v2 ) ; + resistor networks and morphisms between them form a category resnet ; this category has finite limits and colimits .
+ there is a compact closed bicategory cospan(resnet ) with an important compact closed subbicategory circ consisting of cospans whose feet are resistor networks with no edges . a morphism in circ is a * circuit , * a resistor network with chosen sets of input and output vertices across which one can measure a voltage drop .
compact closed categories were first defined by kelly @xcite , and later studied in depth by kelly and laplaza @xcite .
bnabou @xcite defined bicategories and showed that small categories , distributors , and natural transformations form a bicategory dist .
distributors later became more widely known as `` profunctors '' , so we will call that bicategory `` prof '' instead .
later , bnabou defined closed bicategories and showed that prof is closed @xcite .
he defined @xmath41-enriched profunctors when @xmath41 is a cocomplete monoidal or symmetric monoidal closed category , defined @xmath41-prof and proved that any @xmath41-enriched functor , regarded as a @xmath41-profunctor , has a right adjoint .
more applications and details are in his lecture notes @xcite .
kapranov and voevodsky @xcite defined braided semistrict monoidal 2-categories , but their definition left out some necessary axioms .
baez and neuchl @xcite gave an improved definition , but it was still missing a clause ; crans @xcite gave the complete definition .
see baez and langford @xcite and shulman @xcite for details .
gordon , power , and street @xcite defined fully weak tricategories ; a monoidal bicategory is a one - object tricategory .
another name for semistrict monoidal 2-categories is `` gray monoids '' , _
i.e. _ monoid objects in the 2-category gray @xcite . day and street @xcite defined compact closed gray monoids , and appealed to the coherence theorem of gordon , power , and street to extend compact closedness to arbitrary bicategories .
the semistrict approach is somewhat artificial when dealing with most `` naturally occurring '' bicategories , since the associator for composition of 1-morphisms is rarely the identity .
however , this approach does apply to cocont .
katis , sabadini and walters @xcite gave a precise account of the double - entry bookkeeping method _
partita doppia _ in terms of the compact closed bicategory span(rgraph ) ; in a later paper @xcite , they cite a handwritten note by mccrudden for the `` swallowtail '' coherence law we use in this paper .
preller and lambek @xcite generalized compact monoidal categories in a different direction .
they considered a compact monoidal category to be a one - object bicategory satisfying some criteria , and then extend that definition to multiple objects .
the resulting concept of `` compact bicategory '' is _ not _ what is being studied in this paper .
mccrudden @xcite gave the first fully general definitions of braided , sylleptic , and symmetric monoidal bicategories .
schommer - pries @xcite gave the correct notion of a monoidal transformation between monoidal functors between monoidal bicategories .
carboni and walters @xcite proved that @xmath41-prof is a cartesian bicategory .
later , they showed @xcite that prof is equivalent to cocont as a bicategory . together with kelly and wood @xcite , they proved that any cartesian bicategory is symmetric monoidal in the sense of mccrudden .
in this section , we lay out the definition of a compact closed bicategory .
first we give the definition of a bicategory , then start adding structure to it : we introduce the tensor product and monoidal unit ; then we look at the different ways to move objects around each other , giving braided , sylleptic and symmetric monoidal bicategories .
next , we define closed monoidal bicategories by introducing a right pseudoadjoint to tensoring with an object ; and finally we introduce duals for objects in a bicategory .
a * bicategory * @xmath53 consists of 1 . a collection of * objects * 2 . for each pair of objects @xmath54 in @xmath53 , a category @xmath55 ;
the objects of @xmath55 are called * 1-morphisms * , while the morphisms of @xmath55 are called * 2-morphisms*. 3 . for each triple of objects @xmath56 in @xmath53 , a * composition * functor @xmath57 we will leave off the indices and write it as an infix operator .
4 . for each object
@xmath58 in @xmath53 , an object @xmath59 in @xmath60 called the * identity 1-morphism on @xmath58*. we will often write this simply as @xmath58 . 5 . for each quadruple of objects @xmath61 , a natural isomorphism called the * associator for composition * ; if @xmath62 is an object of @xmath63 then @xmath64 6 . for each pair of objects
@xmath54 in @xmath53 , natural isomorphisms called * left and right unitors for composition .
* if @xmath9 is an object of @xmath55 , then @xmath65 such that @xmath66 and @xmath67 satisfy the following coherence laws :
1 . for all @xmath68 in @xmath69 the following diagram , called the * pentagon equation * ,
commutes : @xmath70{((f \circ g ) \circ h ) \circ j } \arrow{sse , t } { { { \stackrel{\mbox{\circle{2}}}{a } } } _ { f\circ g , h , j } } \arrow{sw , t } { { { \stackrel{\mbox{\circle{2}}}{a } } } _ { f , g , h } \circ j }
\\ \node{(f \circ ( g \circ h ) ) \circ j } \arrow[2]{s , l } { { { \stackrel{\mbox{\circle{2}}}{a } } } _ { f , g\circ h , j}}\\ \node[3]{(f \circ g ) \circ ( h \circ j ) } \arrow{ssw , b } { { { \stackrel{\mbox{\circle{2}}}{a } } } _ { f , g , h \circ j}}\\ \node{f \circ ( ( g \circ h ) \circ j ) } \arrow{se , b}{f \circ { { \stackrel{\mbox{\circle{2}}}{a } } } _ { g , h , j}}\\ \node[2]{f \circ ( g \circ ( h \circ j ) ) } \end{diagram}\ ] ] 2 . for all @xmath71 in @xmath72 the following diagram , called the * triangle equation * ,
commutes : @xmath73{e , t } { { { \stackrel{\mbox{\circle{2}}}{a } } } } \arrow{se , b } { { { \stackrel{\mbox{\circle{2}}}{r } } } _ f \circ g } \node[2]{f \circ ( b \circ g ) } \arrow{sw , b}{f \circ { { \stackrel{\mbox{\circle{2}}}{l } } } _ g } \\
\node[2]{f \circ g } \end{diagram}\ ] ] the associator @xmath74 and unitors @xmath75 for composition of 1-morphisms are necessary , but when we are drawing commutative diagrams of 1-morphisms they are very hard to show ; fortunately , by the coherence theorem for bicategories , any consistent choice is equivalent to any other , so we leave them out .
we refer the reader to tom leinster s excellent `` basic bicategories '' @xcite for definitions of * morphisms of bicategories , which we call functors , * transformations between functors , which we call pseudonatural transformations , and * modifications between transformations .
an * equivalence * of objects @xmath54 in a bicategory is a pair of morphisms @xmath76 such that there exists an invertible 2-morphism @xmath77 and an invertible 2-morphism @xmath78 . an * adjoint equivalence * is one in which the morphisms @xmath79 are adjoint . for a given morphism @xmath80
any two choices of data @xmath81 making @xmath9 an adjoint equivalence are canonically isomorphic , so any choice is as good as any other . when @xmath79 form an adjoint equivalence , we write @xmath82 .
any equivalence can be improved to an adjoint equivalence .
we can often take a 2-morphism and `` reverse '' one of its edges .
given objects @xmath83 , morphisms @xmath84 , @xmath85 , @xmath86 , @xmath87 such that @xmath23 is an adjoint equivalence , and a 2-morphism \(a ) at ( 0,0 ) @xmath58 ; ( alpha ) at ( 1,0 ) @xmath88 ; ( c ) at ( .67 , -.5 ) @xmath0 edge [ < - ] node [ below left , l ] @xmath9 ( a ) ; ( d ) at ( 1.33 , -.5 ) @xmath89 edge [ < - ] node [ below , l ] @xmath8 ( c ) ; ( b ) at ( 2,0 ) @xmath90 edge [ < - , out=135 , in=45 ] node [ above , l ] @xmath91 ( a ) edge [ < - ] node [ below right , l ] @xmath23 ( d ) ; we can get a new 2-morphism @xmath92 \(a ) at ( 0,0 ) @xmath58 ; ( alpha ) at ( 1,0 ) @xmath88 ; ( c ) at ( .67 , -.5 ) @xmath0 edge [ < - ] node [ below left , l ] @xmath9 ( a ) ; ( d ) at ( 1.33 , -.5 ) @xmath89 edge [ < - ] node [ below , l ] @xmath8 ( c ) ; ( b ) at ( 2,0 ) @xmath90 edge [ < - , out=135 , in=45 ] node [ above , l ] @xmath91 ( a ) edge [ < - ] node [ above left , l ] @xmath23 ( d ) ; ( d2 ) at ( 2.67 , -.5 ) @xmath89 edge [ < - ] node [ above right , l ] @xmath93 ( b ) edge [ < - ] node [ below , l ] @xmath94 ( d ) edge [ < - , out=-135 , in=-45 ] node [ below ] @xmath8 ( c ) ; at ( 2 , -.3 ) @xmath95 ; at ( 1.65 , -.75 ) @xmath96 ; at ( 3,0 ) @xmath97 ; \(a ) at ( 0,0 ) @xmath58 ; ( alpha ) at ( 1,0 ) @xmath98 ; ( c ) at ( .67 , -.5 ) @xmath0 edge [ < - ] node [ below left , l ] @xmath9 ( a ) ; ( d ) at ( 1.33 , -.5 ) @xmath89 edge [ < - ] node [ below , l ] @xmath8 ( c ) ; ( b ) at ( 2,0 ) @xmath90 edge [ < - , out=135 , in=45 ] node [ above , l ] @xmath91 ( a ) edge [ - > ] node [ below right , l ] @xmath93 ( d ) ; at ( 0,-1.2 ) ; where @xmath99 is the 2-morphism from the equivalence .
we denote such variations of a 2-morphism by adding numeric subscripts ; the number simply records the order in which we introduce them , not any information about the particular variation .
a * monoidal bicategory * @xmath53 is a bicategory in which we can `` multiply '' objects .
it consists of the following : + * a bicategory @xmath53 . * a * tensor product * functor @xmath100 .
this functor involves an invertible `` tensorator '' 2-morphism @xmath101 which we elide in most of the coherence equations below .
the coherence theorem for monoidal bicategories implies that any 2-morphism involving the tensorator is the same no matter how it is inserted ( * ? ? ?
* remark 3.1.6 ) , so like the associator for composition of 1-morphisms , we leave it out .
+ the * stasheff polytopes * @xcite are a series of geometric figures whose vertices enumerate the ways to parenthesize the tensor product of @xmath2 objects , so the number of vertices is given by the catalan numbers ; for each polytope , we have a corresponding @xmath37-morphism of the same shape with directed edges and faces : 1 . the tensor product of one object @xmath58 is the one object @xmath58 itself .
2 . the tensor product of two objects @xmath58 and
@xmath90 is the one object @xmath102 .
there are two ways to parenthesize the product of three objects , so we have an * associator * adjoint equivalence pseudonatural in @xmath103 @xmath104 for moving parentheses from the left pair to the right pair , where @xmath105 is denotes the functor @xmath106 and similarly for @xmath107 4 .
there are five ways to parenthesize the product of four objects , so we have a * pentagonator * invertible modification @xmath108 relating the two different ways of moving parentheses from being clustered at the left to being clustered at the right .
( mnemonic : pink pentagonator . ) + ( 0,3)(2,4)(4 , 2)(2,0)(0,1)cycle ; ( abcd1 ) at ( 0,3 ) @xmath109 ; ( abcd2 ) at ( 2,4 ) @xmath110 edge [ < - ] node [ l , above left ] @xmath111 ( abcd1 ) ; ( abcd3 ) at ( 4,2 ) @xmath112 edge [ < - ] node [ l , above right ] @xmath111 ( abcd2 ) ; ( abcd4 ) at ( 2,0 ) @xmath113 edge [ - > ] node [ l , below right ] @xmath114 ( abcd3 ) ; ( abcd5 ) at ( 0,1 ) @xmath115 edge [ - > ] node [ l , below left ] @xmath111 ( abcd4 ) edge [ < - ] node [ l , left ] @xmath116 ( abcd1 ) ; at ( 1,2 ) ; ; at ( 1.5,2 ) @xmath108 ; 5 .
there are fourteen ways to parenthesize the product of five objects , so we have an * associahedron * equation of modifications with fourteen vertices relating the various ways of getting from the parentheses clustered at the left to clustered at the right . + ( -4.306,-3.532)(-2.391 , -.901)(-2.391,3.949)(-5.127,.19)(-5.127,-2.581)cycle ; ( -4.306,-3.532)(-2.391 , -.901)(2.872,-1.858)(4.306,-3.396)(3.212,-4.9)cycle ; ( 2.872,-1.858)(2.872 , 5.07)(-.135,5.617)(-2.391,3.949)(-2.391,-.901)cycle ; ( 4.306,-3.396)(4.306 , 3.532)(2.872,5.07)(2.872,-1.858)cycle ; ( -2.872,1.858)(-.135 , 5.617)(-2.391,3.949)(-5.127,.19)cycle ; ( 3.212,-4.9)(4.306 , -3.396)(4.306,3.532)(2.391,.901)(2.391,-3.949)cycle ; ( -4.306 , -3.532)(3.212,-4.9)(2.391,-3.949)(-5.127,-2.581)cycle ; ( -2.872,1.858)(2.391 , .901)(4.306,3.532)(2.872,5.07)(-.135,5.617)cycle ; ( -5.127,-2.581)(-5.127 , .19)(-2.872,1.858)(2.391,.901)(2.391,-3.949)cycle ; + the associahedron is a cube with three of its edges bevelled .
it holds in the bicategory @xmath53 , where we have used juxtaposition instead of @xmath117 for brevity and the unmarked 2-morphisms are instances of the pseudonaturality invertible modification for the associator .
( mnemonic for the rectangular invertible modifications : green congruences . ) + ( -2.872 , 1.858)(-.135,5.617)(-2.391,3.949)(-5.127,.19)cycle ; ( 3.212,-4.9)(4.306 , -3.396)(4.306,3.532)(2.391,.901)(2.391,-3.949)cycle ; ( -4.306 , -3.532)(3.212,-4.9)(2.391,-3.949)(-5.127,-2.581)cycle ; ( -2.872 , 1.858)(2.391,.901)(4.306,3.532)(2.872,5.07)(-.135,5.617)cycle ; ( -5.127 , -2.581)(-5.127,.19)(-2.872,1.858)(2.391,.901)(2.391 , -3.949)cycle ; + \(a ) at ( -2.391,3.949 ) @xmath118 ; ( b ) at ( -5.127,.19 ) @xmath119 edge [ - > ] node [ l , above left ] @xmath120 ( a ) ; ( c ) at ( -5.127,-2.581 ) @xmath121 edge [ - > ] node [ l , left ] @xmath122 ( b ) ; ( d ) at ( -4.306,-3.532 ) @xmath123 edge [ - > ] node [ l , left ] @xmath124 ( c ) ; ( e ) at ( 3.212,-4.9 ) @xmath125 edge [ < - ] node [ l , below ] @xmath111 ( d ) ; ( f ) at ( 4.306,-3.396 ) @xmath126 edge [ < - ] node [ l , below right ] @xmath111 ( e ) ; ( g ) at ( 4.306,3.532 ) @xmath127 edge [ < - ] node [ l , right ] @xmath111 ( f ) ; ( h ) at ( 2.872,5.07 ) @xmath128 edge [ - > ] node [ l , above right ] @xmath129 ( g ) ; ( i ) at ( -.135,5.617 ) @xmath130 edge [ - > ] node [ l , above ] @xmath131 ( h ) edge [ < - ] node [ l , above left ] @xmath111 ( a ) ; ( j ) at ( -2.872,1.858 ) @xmath132 edge [ - > ] node [ l , below right ] @xmath133 ( i ) edge [ < - ] node [ l , below right ] @xmath111 ( b ) ; ( k ) at ( 2.391,-3.949 ) @xmath134 edge [ < - ] node [ l , left ] @xmath135 ( e ) edge [ < - ] node [ l , above ] @xmath111 ( c ) ; ( l ) at ( 2.391,.901 )
@xmath136 edge [ < - ] node [ l , left ] @xmath111 ( k ) edge [ < - ] node [ l , above ] @xmath131 ( j ) edge [ - > ] node
[ l , above left ] @xmath131 ( g ) ; at ( -1,-1 ) ; ; at ( -.5,-1 ) @xmath108 ; at ( 1,3 ) ; ; at ( 1.5,3 ) @xmath137 ; at ( 3,-1.5 ) ; ; at ( 3.5,-1.5 ) @xmath108 ; at ( -1,-3.7 ) @xmath138 ; at ( -2.5,3 ) @xmath138 ; + + = + + ( -4.306 , -3.532)(-2.391,-.901)(-2.391,3.949)(-5.127,.19)(-5.127 , -2.581)cycle ; ( -4.306 , -3.532)(-2.391,-.901)(2.872,-1.858)(4.306,-3.396)(3.212 , -4.9)cycle ; ( 2.872 , -1.858)(2.872,5.07)(-.135,5.617)(-2.391,3.949)(-2.391 , -.901)cycle ; ( 4.306 , -3.396)(4.306,3.532)(2.872,5.07)(2.872,-1.858)cycle ; + \(a ) at ( -2.391,3.949 ) @xmath118 ; ( b ) at ( -5.127,.19 ) @xmath119 edge [ - > ] node [ l , above left ] @xmath120 ( a ) ; ( c ) at ( -5.127,-2.581 ) @xmath121 edge [ - > ] node [ l , left ] @xmath122 ( b ) ; ( d ) at ( -4.306,-3.532 ) @xmath123 edge [ - > ] node [ l , left ] @xmath124 ( c ) ; ( e ) at ( 3.212,-4.9 ) @xmath125 edge [ < - ] node [ l , below ] @xmath111 ( d ) ; ( f ) at ( 4.306,-3.396 ) @xmath126 edge [ < - ] node [ l , below right ] @xmath111 ( e ) ; ( g ) at ( 4.306,3.532 ) @xmath127 edge [ < - ] node [ l , right ] @xmath111 ( f ) ; ( h ) at ( 2.872,5.07 ) @xmath128 edge [ - > ] node [ l , above right ] @xmath129 ( g ) ; ( i ) at ( -.135,5.617 ) @xmath130 edge [ - > ] node [ l , above ] @xmath131 ( h ) edge [ < - ] node [ l , above left ] @xmath111 ( a ) ; ( m ) at ( -2.391,-.901 ) @xmath139 edge [ < - ] node [ l , right ] @xmath122 ( d ) edge [ - > ] node [ l , right ] @xmath122 ( a ) ; ( n ) at ( 2.872,-1.858 ) @xmath140 edge [ < - ] node [ l , above ] @xmath111 ( m ) edge [ - > ] node [ l , left ] @xmath111 ( h ) edge [ - > ] node [ l , left ] @xmath141 ( f ) ; at ( -4,-.5 ) @xmath142 ; at ( 0,-3 ) ; ; at ( 0.5,-3 ) @xmath108 ; at ( 0,2 ) ; ; at ( 0.5,2 ) @xmath108 ; at ( 3.5,1 ) @xmath138 ; * just as in any monoid there is an identity element 1 , in every monoidal bicategory there is a * monoidal unit * object @xmath143 .
associated to the monoidal unit are a series of morphisms one of each dimension that express how to `` cancel '' the unit in a product .
each morphism of dimension @xmath144 has two stasheff polytopes of dimension @xmath145 as `` subcells '' , one for parenthesizing @xmath146 objects and the other for parenthesizing the @xmath2 objects left over after cancellation .
there are @xmath146 ways to insert @xmath143 into @xmath2 objects , so there are @xmath146 morphisms of dimension @xmath2
there is one monoidal unit object @xmath143 .
2 . there are two * unitor * adjoint equivalences @xmath147 and @xmath32 that are pseudonatural in @xmath58 .
the stasheff polytopes for two objects and for one object are both points , so the unitors are line segments joining them .
@xmath148 @xmath149 3
. there are three * 2-unitor * invertible modifications @xmath150 and @xmath151 .
the stasheff polytope for three objects is a line segment and the stasheff polytope for two objects is a point , so these modifications are triangles .
( mnemonic : umber unitor . ) + ( 0,2)(3,1)(0 , 0)cycle ; ( iab1 ) at ( 0,2 ) @xmath152 ; ( iab2 ) at ( 3,1 ) @xmath153 edge [ < - ] node [ l , above right ] @xmath111 ( iab1 ) ; ( ab ) at ( 0,0 ) @xmath102 edge [ < - ] node [ l , below right ] @xmath147 ( iab2 ) edge [ < - ] node [ l , left ] @xmath154 ( iab1 ) ; at ( 1,1 ) @xmath155 ; + ( 0,2)(3,1)(0 , 0)cycle ; ( ab ) at ( 0,0 ) @xmath102 ; ( aib2 ) at ( 3,1 ) @xmath156 edge [ - > ] node [ l , below right ] @xmath157 ( ab ) ; ( aib1 ) at ( 0,2 ) @xmath158 edge [ - > ] node [ l , above ] @xmath111 ( aib2 ) edge [ - > ] node [ l , left ] @xmath159 ( ab ) ; at ( 1,1 ) @xmath160 ; + ( 0,2)(3,1)(0 , 0)cycle ; ( abi1 ) at ( 0,2 ) @xmath161 ; ( abi2 ) at ( 3,1 ) @xmath162 edge [ < - ] node [ l , above right ] @xmath111 ( abi1 ) ; ( ab ) at ( 0,0 ) @xmath102 edge [ < - ] node [ l , below right ] @xmath163 ( abi2 ) edge [ < - ] node [ l , left ] @xmath32 ( abi1 ) ; at ( 1,1 ) @xmath164 ; 4 .
there are four equations of modifications .
the stasheff polytope for four objects is a pentagon and the stasheff polytope for three objects is a line segment , so these equations are irregular prisms with seven vertices .
+ ( -1,1)(1,1)(1 , -1)(-1,-1)(-1.1,0)cycle ; ( 1,1)(2,0)(1 , -1)cycle ; + \(a ) at ( -1.1,0 ) @xmath165 ; ( b ) at ( -1,1 ) @xmath166 edge [ - > ] node [ l , left ] @xmath167 ( a ) ; ( c ) at ( 1,1 ) @xmath168 edge [ < - ] node [ l , above ] @xmath111 ( b ) ; ( d ) at ( 1,-1 ) @xmath169 edge [ < - ] node [ l , left ] @xmath111 ( c ) ; ( e ) at ( -1,-1 ) @xmath170 edge [ - > ] node [ l , below ] @xmath171 ( d ) edge [ < - ] node [ l , left ] @xmath111 ( a ) ; ( f ) at ( 2,0 ) @xmath172 edge [ < - ] node [ l , above right ] @xmath173 ( c ) edge [ < - ] node [ l , below right ] @xmath147 ( d ) ; at ( 0,0 ) @xmath174 ; at ( 1.33,0 ) @xmath175 ; at ( 2.5 , 0 ) = ; + ( -1,1)(0,0)(-1.1 , 0)cycle ; ( -1,-1)(0 , 0)(-1.1,0)cycle ; ( -1,1)(0,0)(2 , 0)(1,1)cycle ; ( -1,-1)(0,0)(2 , 0)(1,-1)cycle ; + \(a ) at ( -1.1,0 ) @xmath165 ; ( b ) at ( -1,1 ) @xmath166 edge [ - > ] node [ l , left ] @xmath167 ( a ) ; ( c ) at ( 1,1 ) @xmath168 edge [ < - ] node [ l , above ] @xmath111 ( b ) ; ( d ) at ( 1,-1 ) @xmath169 ; ( e ) at ( -1,-1 ) @xmath170 edge [ - > ] node [ l , below ] @xmath171 ( d ) edge [ < - ] node [ l , left ] @xmath111 ( a ) ; ( f ) at ( 2,0 ) @xmath172 edge [ < - ] node [ l , above right ] @xmath176 ( c ) edge [ < - ] node [ l , below right ] @xmath147 ( d ) ; ( g ) at ( 0,0 ) @xmath177 edge [ < - ] node [ l , above ] @xmath178 ( a ) edge [ < - ] node [ l , above right ] @xmath176 ( b ) edge [ < - ] node [ l , below right ] @xmath147 ( e ) edge [ - > ] node [ l , above ] @xmath111 ( f ) ; at ( -.66,.3 ) @xmath179 ; at ( -.66,-.3 ) @xmath175 ; at ( .5,.5 ) @xmath138 ; at ( .5,-.5 ) @xmath138 ; + ( -1,-1)(-1,1)(1 , 1)(1.1,0)(1,-1)cycle ; ( 1,1)(1.1,0)(2 , 0)cycle ; ( 1.1,0)(1,-1)(2 , 0)cycle ; + \(a ) at ( -1,1 ) @xmath180 ; ( b ) at ( 1,1 ) @xmath181 edge [ < - ] node [ l , above ] @xmath111 ( a ) ; ( c ) at ( 1.1,0 ) @xmath182 edge [ < - ] node [ l , left ] @xmath111 ( b ) ; ( d ) at ( 1,-1 ) @xmath183 edge [ - > ] node [ l , left ] @xmath131 ( c ) ; ( e ) at ( -1,-1 ) @xmath184 edge [ - > ] node [ l , below ] @xmath111 ( d ) edge [ < - ] node [ l , left ] @xmath167 ( a ) ; ( f ) at ( 2,0 ) @xmath185 edge [ < - ] node [ l , above right ] @xmath186 ( b ) edge [ < - ] node [ l , above ] @xmath187 ( c ) edge [ < - ] node [ l , below right ] @xmath188 ( d ) ; at ( 0,0 ) @xmath174 ; at ( 1.2,.3 ) ; ; at ( 1.4,.3 ) @xmath189 ; at ( 1.2,-.3 ) ; ; at ( 1.4,-.3 ) @xmath190 ; at ( 2.5,0 ) = ; + ( -1,1)(0,0)(-1 , -1)cycle ; ( -1,1)(0,0)(2 , 0)(1,1)cycle ; ( 0,0)(2,0)(1 , -1)(-1,-1)cycle ; + \(a ) at ( -1,1 ) @xmath180 ; ( b ) at ( 1,1 ) @xmath181 edge [ < - ] node [ l , above ] @xmath111 ( a ) ; ( d ) at ( 1,-1 ) @xmath183 ; ( e ) at ( -1,-1 ) @xmath184 edge [ - > ] node [ l , below ] @xmath111 ( d ) edge [ < - ] node [ l , left ] @xmath167 ( a ) ; ( f ) at ( 2,0 ) @xmath185 edge [ < - ] node [ l , above right ] @xmath186 ( b ) edge [ < - ] node [ l , below right ] @xmath188 ( d ) ; ( g ) at ( 0,0 ) @xmath191 edge [ < - ] node [ l , above right ] @xmath192 ( a ) edge [ < - ] node [ l , below right ] @xmath193 ( e ) edge [ - > ] node [ l , above ] @xmath111 ( f ) ; at ( -.66,0 ) @xmath194 ; at ( .5,.5 ) @xmath138 ; at ( .5,-.5 ) @xmath138 ; + ( -1,1)(1,1)(1 , -1)(-1,-1)(-1.1,0)cycle ; ( 1,1)(2,0)(1 , -1)cycle ; + \(a ) at ( -1.1,0 ) @xmath195 ; ( b ) at ( -1,1 ) @xmath196 edge [ < - ] node [ l , left ] @xmath111 ( a ) ; ( c ) at ( 1,1 ) @xmath197 edge [ < - ] node [ l , above ] @xmath111 ( b ) ; ( d ) at ( 1,-1 ) @xmath198 edge [ - > ] node [ l , left ] @xmath131 ( c ) ; ( e ) at ( -1,-1 ) @xmath199 edge [ - > ] node [ l , below ] @xmath111 ( d ) edge [ < - ] node [ l , left ] @xmath167 ( a ) ; ( f ) at ( 2,0 ) @xmath200 edge [ < - ] node [ l , above right ] @xmath201 ( c ) edge [ < - ] node [ l , below right ] @xmath202 ( d ) ; at ( 0,0 ) @xmath174 ; at ( 1.33,0 ) @xmath203 ; at ( 2.5 , 0 ) = ; + ( -1,1)(0,0)(-1.1 , 0)cycle ; ( -1,-1)(0 , 0)(-1.1,0)cycle ; ( -1,1)(0,0)(2 , 0)(1,1)cycle ; ( -1,-1)(0,0)(2 , 0)(1,-1)cycle ; + \(a ) at ( -1.1,0 ) @xmath195 ; ( b ) at ( -1,1 ) @xmath196 edge [ < - ] node [ l , left ] @xmath111 ( a ) ; ( c ) at ( 1,1 ) @xmath197 edge [ < - ] node [ l , above ] @xmath111 ( b ) ; ( d ) at ( 1,-1 ) @xmath198 ; ( e ) at ( -1,-1 ) @xmath199 edge [ - > ] node [ l , below ] @xmath111 ( d ) edge [ < - ] node [ l , left ] @xmath167 ( a ) ; ( f ) at ( 2,0 ) @xmath200 edge [ < - ] node [ l , above right ] @xmath201 ( c ) edge [ < - ] node [ l , below right ] @xmath202 ( d ) ; ( g ) at ( 0,0 ) @xmath204 edge [ < - ] node [ l , above ] @xmath205 ( a ) edge [ < - ] node [ l , above right ] @xmath206 ( b ) edge [ < - ] node [ l , below right ] @xmath207 ( e ) edge [ - > ] node [ l , above ] @xmath111 ( f ) ; at ( -.66,.3 ) @xmath208 ; at ( -.66,-.3 ) @xmath209 ; at ( .5,.5 ) @xmath138 ; at ( .5,-.5 ) @xmath138 ; + ( -1,-1)(-1,1)(1 , 1)(1.1,0)(1,-1)cycle ; ( 1,1)(1.1,0)(2 , 0)cycle ; ( 1.1,0)(1,-1)(2 , 0)cycle ; + \(a ) at ( -1,1 ) @xmath210 ; ( b ) at ( 1,1 ) @xmath211 edge [ < - ] node [ l , above ] @xmath111 ( a ) ; ( c ) at ( 1.1,0 ) @xmath212 edge [ - > ] node [ l , left ] @xmath131 ( b ) ; ( d ) at ( 1,-1 ) @xmath213 edge [ - > ] node [ l , left ] @xmath111 ( c ) ; ( e ) at ( -1,-1 ) @xmath214 edge [ - > ] node [ l , below ] @xmath215 ( d ) edge [ - > ] node [ l , left ] @xmath111 ( a ) ; ( f ) at ( 2,0 ) @xmath216 edge [ < - ] node [ l , above right ] @xmath217 ( b ) edge [ < - ] node [ l , above ] @xmath218 ( c ) edge [ < - ] node [ l , below right ] @xmath32 ( d ) ; at ( 0,0 ) @xmath174 ; at ( 1.3,.3 ) @xmath219 ; at ( 1.2,-.3 ) ; ; at ( 1.4,-.3 ) @xmath151 ; at ( 2.5,0 ) = ; + ( -1,1)(0,0)(-1 , -1)cycle ; ( -1,1)(0,0)(2 , 0)(1,1)cycle ; ( 0,0)(2,0)(1 , -1)(-1,-1)cycle ; + \(a ) at ( -1,1 ) @xmath210 ; ( b ) at ( 1,1 ) @xmath211 edge [ < - ] node [ l , above ] @xmath111 ( a ) ; ( d ) at ( 1,-1 ) @xmath213 ; ( e ) at ( -1,-1 ) @xmath214 edge [ - > ] node [ l , below ] @xmath215 ( d ) edge [ - > ] node [ l , left ] @xmath111 ( a ) ; ( f ) at ( 2,0 ) @xmath216 edge [ < - ] node [ l , above right ] @xmath217 ( b ) edge [ < - ] node [ l , below right ] @xmath32 ( d ) ; ( g ) at ( 0,0 ) @xmath220 edge [ < - ] node [ l , above right ] @xmath221 ( a ) edge [ < - ] node [ l , below right ] @xmath32 ( e ) edge [ - > ] node [ l , above ] @xmath111 ( f ) ; at ( -.66,0 ) @xmath222 ; at ( .5,.5 ) @xmath138 ; at ( .5,-.5 ) @xmath138 ; a * braided * monoidal bicategory @xmath53 is a monoidal bicategory in which objects can be moved past each other .
a braided monoidal bicategory consists of the following : * a monoidal bicategory @xmath53 ; * a series of morphisms for `` shuffling '' .
+ a * shuffle * of a list @xmath223 into a list @xmath224 inserts each element of @xmath225 into @xmath226 such that if @xmath227 then @xmath228 appears to the left of @xmath229 .
+ an `` @xmath230-shuffle polytope '' is an @xmath2-dimensional polytope whose vertices are all the different shuffles of an @xmath2-element list into a @xmath231-element list ; there are @xmath232 ways to do this .
general shuffle polytopes were defined by kapronov and voevodsky @xcite . as with the stasheff polytopes
, we have morphisms of the same shape as @xmath233-shuffle polytopes with directed edges and faces .
* * @xmath234 : @xmath235 so this polytope has two vertices , @xmath236 and @xmath237 .
it has a single edge , which we call a `` braiding '' , which encodes how @xmath58 moves past @xmath90 .
it is an adjoint equivalence pseudonatural in @xmath54 . @xmath238 * * @xmath239 and @xmath240 : @xmath241 so whenever the associator is the identity_e.g .
_ in a braided strictly monoidal bicategory
these polytopes are triangles , invertible modifications whose edges are the directed ( 1,1 ) polytope , the braiding .
+ ( 30:2cm)(150:2cm)(270:2cm)cycle ; ( a ) at ( 30:2 cm ) @xmath242 ; ( c ) at ( 150:2 cm ) @xmath243 edge [ - > ] node [ l , above ] @xmath244 ( a ) ; ( e ) at ( 270:2 cm ) @xmath245 edge [ < - ] node [ l , below left ] @xmath246 ( c ) edge [ - > ] node [ l , below right ] @xmath247 ( a ) ; at ( 0,0 ) @xmath248 ; + ( 30:2cm)(150:2cm)(270:2cm)cycle ; ( a ) at ( 30:2 cm ) @xmath249 ; ( c ) at ( 150:2 cm ) @xmath243 edge [ - > ] node [ l , above ] @xmath244 ( a ) ; ( e ) at ( 270:2 cm ) @xmath250 edge [ < - ] node [ l , below left ] @xmath251 ( c ) edge [ - > ] node [ l , below right ] @xmath252 ( a ) ; at ( 0,0 ) @xmath253 ; + when the associator is not the identity , the triangles vertices get replaced with associators , effectively truncating them , and we are left with hexagon invertible modifications .
( mnemonic : blue braiding . ) + ( 0:2cm)(60:2cm)(120:2cm)(180:2cm)(240:2cm)(300:2 cm ) cycle ; ( a ) at ( 0:2 cm ) @xmath254 ; ( b ) at ( 60:2 cm ) @xmath255 edge [ - > ] node [ l , above right ] @xmath111 ( a ) ; ( c ) at ( 120:2 cm ) @xmath216 edge [ - > ] node [ l , above ] @xmath244 ( b ) ; ( d ) at ( 180:2 cm ) @xmath220 edge [ - > ] node [ l , above left ] @xmath111 ( c ) ; ( e ) at ( 240:2 cm ) @xmath256 edge [ < - ] node [ l , below left ] @xmath246 ( d ) ; ( f ) at ( 300:2 cm ) @xmath257 edge [ < - ] node [ l , below ] @xmath111 ( e ) edge [ - > ] node [ l , below right ] @xmath247 ( a ) ; at ( -0.25,0 ) ; ; at ( 0.25,0 ) @xmath1 ; + ( 0:2cm)(60:2cm)(120:2cm)(180:2cm)(240:2cm)(300:2 cm ) cycle ; ( a ) at ( 0:2 cm ) @xmath258 ; ( b ) at ( 60:2 cm ) @xmath259 edge [ - > ] node [ l , above right ] @xmath260 ( a ) ; ( c ) at ( 120:2 cm ) @xmath220 edge [ - > ] node [ l , above ] @xmath244 ( b ) ; ( d ) at ( 180:2 cm ) @xmath216 edge [ - > ] node [ l , above left ] @xmath260 ( c ) ; ( e ) at ( 240:2 cm ) @xmath261 edge [ < - ] node [ l , below left ] @xmath251 ( d ) ; ( f ) at ( 300:2 cm ) @xmath262 edge [ < - ] node [ l , below ] @xmath260 ( e ) edge [ - > ] node [ l , below right ] @xmath252 ( a ) ; at ( -0.25,0 ) ; ; at ( 0.25,0 ) @xmath6 ; * * @xmath263 and @xmath264 : @xmath265 so in a braided strictly monoidal bicategory , these polytopes are tetrahedra whose faces are the @xmath266 polytope . + ( -.551,-.318)(.172 , 6.631)(-5.449,-3.146)cycle ; ( -.551,-.318)(5.828 , -3.167)(-5.449,-3.146)cycle ; ( -.551,-.318)(.172 , 6.631)(5.828,-3.167)cycle ; ( -5.449,-3.146)(.172 , 6.631)(5.828,-3.167)cycle ; + again , when the associator is not the identity , the vertices get truncated , this time being replaced by pentagonators ; as a side - effect , four of the six edges are also beveled .
+ ( 3.11,-1.226)(1.675 , -.585)(1.112,-1.562)(2.548,-2.203)cycle ; ( 3.281,.425)(2.009 , 2.629)(-.144,3.591)(-.235,2.722)(1.675,-.585)(3.11 , -1.226)cycle ; ( -3.234 , -1.214)(-2.132,-.578)(-.235,2.722)(-.144,3.591)(-1.798 , 2.636)(-3.062,.436)cycle ; ( -1.566,-1.557)(-2.132 , -.578)(-.235,2.722)(1.675,-.585)(1.112,-1.562)cycle ; ( 2.548,-2.203)(1.112 , -1.562)(-1.566,-1.557)(-2.668,-2.194)(-.515,-3.155)(.894 , -3.158)cycle ; ( -3.234 , -1.214)(-2.132,-.578)(-1.566,-1.557)(-2.668,-2.194)cycle ; ( 1.519,2.346)(-1.16 , 2.351)(-1.798,2.636)(-.144,3.591)(2.009,2.629)cycle ; ( 3.281,.425)(2.791 , .142)(.894,-3.158)(2.548,-2.203)(3.11,-1.226)cycle ; ( -3.234,-1.214)(-2.668 , -2.194)(-.515,-3.155)(-2.425,.152)(-3.062,.436)cycle ; ( 3.281,.425)(2.009 , 2.629)(1.519,2.346)(2.791,.142)cycle ; ( -3.062,.436)(-1.798 , 2.636)(-1.16,2.351)(-2.425,.152)cycle ; ( 2.791,.142)(1.519 , 2.346)(-1.16,2.351)(-2.425,.152)(-.515,-3.155)(.894 , -3.158)cycle ; + this equation governs shuffling one object @xmath58 into three objects @xmath267 : + + ( 1.519 , 2.346)(-1.16,2.351)(-1.798,2.636)(-.144,3.591)(2.009 , 2.629)cycle ; ( 3.281 , .425)(2.791,.142)(.894,-3.158)(2.548,-2.203)(3.11 , -1.226)cycle ; ( -3.234 , -1.214)(-2.668,-2.194)(-.515,-3.155)(-2.425,.152)(-3.062 , .436)cycle ; ( 3.281 , .425)(2.009,2.629)(1.519,2.346)(2.791,.142)cycle ; ( -3.062 , .436)(-1.798,2.636)(-1.16,2.351)(-2.425,.152)cycle ; ( 2.791 , .142)(1.519,2.346)(-1.16,2.351)(-2.425,.152)(-.515 , -3.155)(.894,-3.158)cycle ; + \(a ) at ( -.144,3.591 ) ; at ( a ) @xmath268 ; ( b ) at ( 2.009,2.629 ) edge [ < - ] node [ l , above right ] @xmath111 ( a ) ; at ( b ) @xmath269 ; ( c ) at ( 3.281,.425 ) edge [ < - ] node [ l , above right ] @xmath244 ( b ) ; at ( c ) @xmath270 ; ( d ) at ( 3.11,-1.226 ) edge [ < - ] node [ l , right ] @xmath111 ( c ) ; at ( d ) @xmath271 ; ( e ) at ( 2.548,-2.203 ) edge [ < - ] node [ l , right ] @xmath111 ( d ) ; at ( e ) @xmath272 ; ( f ) at ( .894,-3.158 ) edge [ - > ] node [ l , below right ] @xmath273 ( e ) ; at ( f ) @xmath274 ; ( g ) at ( -.515,-3.155 ) edge [ - > ] node [ l , above ] @xmath247 ( f ) ; at ( g ) @xmath275 ; ( h ) at ( -2.668,-2.194 ) edge [ - > ] node [ l , below left ] @xmath273 ( g ) ; at ( h ) @xmath276 ; ( i ) at ( -3.234,-1.214 ) edge [ - > ] node [ l , left ] @xmath111 ( h ) ; at ( i ) @xmath277 ; ( j ) at ( -3.062,.436 ) edge [ - > ] node [ l , left ] @xmath167 ( i ) ; at ( j ) @xmath278 ; ( k ) at ( -1.798,2.636 ) edge [ - > ] node [ l , above left ] @xmath279 ( j ) edge [ - > ] node [ l , above left ] @xmath167 ( a ) ; at ( k ) @xmath280 ; ( l ) at ( -1.16,2.351 ) edge [ < - ] node [ l , below left ] @xmath111 ( k ) ; at ( l ) @xmath281 ; ( m ) at ( 1.519,2.346 ) edge [ < - ] node [ l , above ] @xmath111 ( l ) edge [ < - ] node [ l , below , sloped ] @xmath131 ( b ) ; at ( m ) @xmath282 ; ( n ) at ( 2.791,.142 ) edge [ < - ] node [ l , left ] @xmath244 ( m ) edge [ - > ] node [ l , left ] @xmath111 ( f ) edge [ < - ] node [ l , above , sloped ] @xmath283 ( c ) ; at ( n ) @xmath284 ; ( o ) at ( -2.425,.152 ) edge [ < - ] node [ l , below right ] @xmath285 ( l ) edge [ < - ] node [ l , above ] @xmath111 ( j ) edge [ - > ] node [ l , right ] @xmath111 ( g ) ; at ( o ) @xmath286 ; at ( -0.25,3 ) ; ; at ( 0,3 ) @xmath108 ; at ( -2,1.5 ) @xmath138 ; at ( 2.3,1.5 ) @xmath138 ; at ( -2.25,-1.5 ) @xmath287 ; at ( 2.5,-1.5 ) @xmath287 ; at ( -0,0 ) ; ; at ( 0.25,0 ) @xmath1 ; + + = + + ( 3.11 , -1.226)(1.675,-.585)(1.112,-1.562)(2.548,-2.203)cycle ; ( 3.281 , .425)(2.009,2.629)(-.144,3.591)(-.235,2.722)(1.675,-.585)(3.11 , -1.226)cycle ; ( -3.234 , -1.214)(-2.132,-.578)(-.235,2.722)(-.144,3.591)(-1.798 , 2.636)(-3.062,.436)cycle ; ( -1.566 , -1.557)(-2.132,-.578)(-.235,2.722)(1.675,-.585)(1.112 , -1.562)cycle ; ( 2.548 , -2.203)(1.112,-1.562)(-1.566,-1.557)(-2.668,-2.194)(-.515 , -3.155)(.894,-3.158)cycle ; ( -3.234 , -1.214)(-2.132,-.578)(-1.566,-1.557)(-2.668,-2.194)cycle ; + \(a ) at ( -.144,3.591 ) ; at ( a ) @xmath268 ; ( b ) at ( 2.009,2.629 ) edge [ < - ] node [ l , above right ] @xmath111 ( a ) ; at ( b ) @xmath269 ; ( c ) at ( 3.281,.425 ) edge [ < - ] node [ l , above right ] @xmath244 ( b ) ; at ( c ) @xmath270 ; ( d ) at ( 3.11,-1.226 ) edge [ < - ] node [ l , right ] @xmath111 ( c ) ; at ( d ) @xmath271 ; ( e ) at ( 2.548,-2.203 ) edge [ < - ] node [ l , right ] @xmath111 ( d ) ; at ( e ) @xmath272 ; ( f ) at ( .894,-3.158 ) edge [ - > ] node [ l , below right ] @xmath273 ( e ) ; at ( f ) @xmath274 ; ( g ) at ( -.515,-3.155 ) edge [ - > ] node [ l , above ] @xmath247 ( f ) ; at ( g ) @xmath275 ; ( h ) at ( -2.668,-2.194 ) edge [ - > ] node [ l , below left ] @xmath273 ( g ) ; at ( h ) @xmath276 ; ( i ) at ( -3.234,-1.214 ) edge [ - > ] node [ l , left ] @xmath111 ( h ) ; at ( i ) @xmath277 ; ( j ) at ( -3.062,.436 ) edge [ - > ] node [ l , left ] @xmath167 ( i ) ; at ( j ) @xmath278 ; ( k ) at ( -1.798,2.636 ) edge [ - > ] node [ l , above left ] @xmath279 ( j ) edge [ - > ] node [ l , above left ] @xmath167 ( a ) ; at ( k ) @xmath280 ; ( p ) at ( -2.132,-.578 ) edge [ < - ] node [ l , above , sloped ] @xmath288 ( i ) ; at ( p ) @xmath289 ; ( q ) at ( -.235,2.722 ) edge [ - > ] node [ l , below right ] @xmath167 ( p ) edge [ < - ] node [ l , left ] @xmath290 ( a ) ; at ( q ) @xmath291 ; ( r ) at ( 1.675,-.585 ) edge [ < - ] node [ l , below left ] @xmath111 ( q ) edge [ - > ] node [ l , above , sloped ] @xmath292 ( d ) ; at ( r ) @xmath293 ; ( s ) at ( 1.112,-1.562 ) edge [ < - ] node [ l , below right ] @xmath111 ( r ) edge [ - > ] node [ l , above , sloped ] @xmath294 ( e ) ; at ( s ) @xmath295 ; ( t ) at ( -1.566,-1.557 ) edge [ - > ] node [ l , below ] @xmath273 ( s ) edge [ < - ] node [ l , below left ] @xmath111 ( p ) edge [ < - ] node [ l , above , sloped ] @xmath296 ( h ) ; at ( t ) @xmath297 ; at ( -0.25,0 ) ; ; at ( 0.25,0 ) @xmath298 ; at ( 2,1 ) @xmath299 ; at ( -2.25,1 ) ; ; at ( -2,1 ) @xmath300 ; at ( 2.25,-1.25 ) @xmath138 ; at ( -2.5,-1.25 ) @xmath138 ; at ( -0.5,-2.5 ) ; ; at ( 0,-2.5 ) @xmath301 ; + this equation governs shuffling one object @xmath89 into three objects @xmath103 : + + ( 1.519 , 2.346)(-1.16,2.351)(-1.798,2.636)(-.144,3.591)(2.009 , 2.629)cycle ; ( 3.281 , .425)(2.791,.142)(.894,-3.158)(2.548,-2.203)(3.11 , -1.226)cycle ; ( -3.234 , -1.214)(-2.668,-2.194)(-.515,-3.155)(-2.425,.152)(-3.062 , .436)cycle ; ( 3.281 , .425)(2.009,2.629)(1.519,2.346)(2.791,.142)cycle ; ( -3.062 , .436)(-1.798,2.636)(-1.16,2.351)(-2.425,.152)cycle ; ( 2.791 , .142)(1.519,2.346)(-1.16,2.351)(-2.425,.152)(-.515 , -3.155)(.894,-3.158)cycle ; + \(a ) at ( -.144,3.591 ) ; at ( a ) @xmath269 ; ( b ) at ( 2.009,2.629 ) edge [ < - ] node [ l , above right ] @xmath260 ( a ) ; at ( b ) @xmath268 ; ( c ) at ( 3.281,.425 ) edge [ < - ] node [ l , above right ] @xmath244 ( b ) ; at ( c ) @xmath302 ; ( d ) at ( 3.11,-1.226 ) edge [ < - ] node [ l , right ] @xmath260 ( c ) ; at ( d ) @xmath303 ; ( e ) at ( 2.548,-2.203 ) edge [ < - ] node [ l , right ] @xmath260 ( d ) ; at ( e ) @xmath304 ; ( f ) at ( .894,-3.158 ) edge [ - > ] node [ l , below right ] @xmath305 ( e ) ; at ( f ) @xmath306 ; ( g ) at ( -.515,-3.155 ) edge [ - > ] node [ l , above ] @xmath246 ( f ) ; at ( g ) @xmath307 ; ( h ) at ( -2.668,-2.194 ) edge [ - > ] node [ l , below left ] @xmath305 ( g ) ; at ( h ) @xmath308 ; ( i ) at ( -3.234,-1.214 ) edge [ - > ] node [ l , left ] @xmath260 ( h ) ; at ( i ) @xmath309 ; ( j ) at ( -3.062,.436 ) edge [ - > ] node [ l , left ] @xmath310 ( i ) ; at ( j ) @xmath311 ; ( k ) at ( -1.798,2.636 ) edge [ - > ] node [ l , above left ] @xmath312 ( j ) edge [ - > ] node [ l , above left ] @xmath310 ( a ) ; at ( k ) @xmath282 ; ( l ) at ( -1.16,2.351 ) edge [ < - ] node [ l , below left ] @xmath260 ( k ) ; at ( l ) @xmath281 ; ( m ) at ( 1.519,2.346 ) edge [ < - ] node [ l , above ] @xmath260 ( l ) edge [ < - ] node [ l , below , sloped ] @xmath313 ( b ) ; at ( m ) @xmath280 ; ( n ) at ( 2.791,.142 ) edge [ < - ] node [ l , left ] @xmath244 ( m ) edge [ - > ] node [ l , left ] @xmath260 ( f ) edge [ < - ] node [ l , above , sloped ] @xmath314 ( c ) ; at ( n ) @xmath315 ; ( o ) at ( -2.425,.152 ) edge [ < - ] node [ l , below right ] @xmath316 ( l ) edge [ < - ] node [ l , above ] @xmath260 ( j ) edge [ - > ] node [ l , right ] @xmath260 ( g ) ; at ( o ) @xmath317 ; at ( -0.25,3 ) ; ; at ( 0,3 ) @xmath318 ; at ( -2,1.5 ) @xmath138 ; at ( 2.3,1.5 ) @xmath138 ; at ( -2.25,-1.5 ) @xmath319 ; at ( 2.5,-1.5 ) @xmath319 ; at ( -0,0 ) ; ; at ( 0.25,0 ) @xmath6 ; + + = + + ( 3.11 , -1.226)(1.675,-.585)(1.112,-1.562)(2.548,-2.203)cycle ; ( 3.281 , .425)(2.009,2.629)(-.144,3.591)(-.235,2.722)(1.675,-.585)(3.11 , -1.226)cycle ; ( -3.234 , -1.214)(-2.132,-.578)(-.235,2.722)(-.144,3.591)(-1.798 , 2.636)(-3.062,.436)cycle ; ( -1.566 , -1.557)(-2.132,-.578)(-.235,2.722)(1.675,-.585)(1.112 , -1.562)cycle ; ( 2.548 , -2.203)(1.112,-1.562)(-1.566,-1.557)(-2.668,-2.194)(-.515 , -3.155)(.894,-3.158)cycle ; ( -3.234 , -1.214)(-2.132,-.578)(-1.566,-1.557)(-2.668,-2.194)cycle ; + \(a ) at ( -.144,3.591 ) ; at ( a ) @xmath269 ; ( b ) at ( 2.009,2.629 ) edge [ < - ] node [ l , above right ] @xmath260 ( a ) ; at ( b ) @xmath268 ; ( c ) at ( 3.281,.425 ) edge [ < - ] node [ l , above right ] @xmath244 ( b ) ; at ( c ) @xmath302 ; ( d ) at ( 3.11,-1.226 ) edge [ < - ] node [ l , right ] @xmath260 ( c ) ; at ( d ) @xmath303 ; ( e ) at ( 2.548,-2.203 ) edge [ < - ] node [ l , right ] @xmath260 ( d ) ; at ( e ) @xmath304 ; ( f ) at ( .894,-3.158 ) edge [ - > ] node [ l , below right ] @xmath305 ( e ) ; at ( f ) @xmath306 ; ( g ) at ( -.515,-3.155 ) edge [ - > ] node [ l , above ] @xmath246 ( f ) ; at ( g ) @xmath307 ; ( h ) at ( -2.668,-2.194 ) edge [ - > ] node [ l , below left ] @xmath305 ( g ) ; at ( h ) @xmath308 ; ( i ) at ( -3.234,-1.214 ) edge [ - > ] node [ l , left ] @xmath260 ( h ) ; at ( i ) @xmath309 ; ( j ) at ( -3.062,.436 ) edge [ - > ] node [ l , left ] @xmath310 ( i ) ; at ( j ) @xmath311 ; ( k ) at ( -1.798,2.636 ) edge [ - > ] node [ l , above left ] @xmath320 ( j ) edge [ - > ] node [ l , above left ] @xmath310 ( a ) ; at ( k ) @xmath282 ; ( p ) at ( -2.132,-.578 ) edge [ < - ] node [ l , above , sloped ] @xmath321 ( i ) ; at ( p ) @xmath322 ; ( q ) at ( -.235,2.722 ) edge [ - > ] node [ l , below right ] @xmath310 ( p ) edge [ < - ] node [ l , left ] @xmath251 ( a ) ; at ( q ) @xmath323 ; ( r ) at ( 1.675,-.585 ) edge [ < - ] node [ l , below left ] @xmath260 ( q ) edge [ - > ] node [ l , above , sloped ] @xmath324 ( d ) ; at ( r ) @xmath325 ; ( s ) at ( 1.112,-1.562 ) edge [ < - ] node [ l , below right ] @xmath260 ( r ) edge [ - > ] node [ l , above , sloped ] @xmath326 ( e ) ; at ( s ) @xmath327 ; ( t ) at ( -1.566,-1.557 ) edge [ - > ] node [ l , below ] @xmath305 ( s ) edge [ < - ] node [ l , below left ] @xmath260 ( p ) edge [ < - ] node [ l , above , sloped ] @xmath328 ( h ) ; at ( t ) @xmath329 ; at ( -0.25,0 ) ; ; at ( 0.25,0 ) @xmath330 ; at ( 2,1 ) @xmath331 ; at ( -2.25,1 ) ; ; at ( -2,1 ) @xmath332 ; at ( 2.25,-1.25 ) @xmath138 ; at ( -2.5,-1.25 ) @xmath138 ; at ( -0.5,-2.5 ) ; ; at ( 0,-2.5 ) @xmath333 ; * * @xmath334 : @xmath335 in a braided strictly monoidal bicategory , this polytope is composed mostly of ( 2,1 ) triangles , but there is a pair of braidings that commute , so one face is a square .
+ ( -5.635,-.347)(-1.888 , .116)(0,-1.695)cycle ; ( 5.635,-.347)(0 , -1.695)(1.888,.116)cycle ; ( 0,2.083)(-1.888 , .116)(0,-1.695)(1.888,.116)cycle ; ( 5.635,-.347)(-5.635 , -.347)(0,-1.695)cycle ; ( 5.635,-.347)(1.888 , .116)(0,2.083)cycle ; ( 0,2.083)(-1.888 , .116)(-5.635,-.347)cycle ; ( 5.635,-.347)(-5.635 , -.347)(0,2.083)cycle ; + when the associator is not the identity , the six vertices get truncated and six of the edges get beveled .
+ ( 0,-.807)(-.944 , .116)(-1.678,-.488)(-.546,-1.564)cycle ; ( .944,.116)(0 , -.807)(.546,-1.564)(1.678,-.488)cycle ; ( .944,.116)(0 , -.807)(-.944,.116)(0,1.079)cycle ; ( .546,-1.564)(.453 , -1.75)(-.453,-1.75)(-.546,-1.564)(0,-.807)cycle ; ( 0,1.079)(-.944 , .116)(-1.72,.642)(-.588,1.829)cycle ; ( .944,.116)(0 , 1.079)(.588,1.829)(1.72,.642)cycle ; ( 1.678,-.488)(2.075 , -.32)(2.109,.417)(1.72,.642)(.944,.116)cycle ; ( -1.72,.642)(-2.109 , .417)(-2.075,-.32)(-1.678,-.488)(-.944,.116)cycle ; ( .588,1.829)(.492 , 2.021)(-.492,2.021)(-.588,1.829)(0,1.079)cycle ; ( 1.678,-.488)(2.075 , -.32)(2.573,-.48)(2.308,-.977)(.453,-1.75)(.546,-1.564)cycle ; ( -.546,-1.564)(-.453 , -1.75)(-2.308,-.977)(-2.573,-.48)(-2.075,-.32)(-1.678 , -.488)cycle ; ( -.453,-1.75)(-2.308 , -.977)(-1.412,-.174)(1.412,-.174)(2.308,-.977)(.453 , -1.75)cycle ; ( 1.72,.642)(.588 , 1.829)(.492,2.021)(2.405,.899)(2.625,.433)(2.109,.417)cycle ; ( -.588,1.829)(-.492 , 2.021)(-2.405,.899)(-2.625,.433)(-2.109,.417)(-1.72 , .642)cycle ; ( 2.075,-.32)(2.573 , -.48)(2.625,.433)(2.109,.417)cycle ; ( -2.075,-.32)(-2.573 , -.48)(-2.625,.433)(-2.109,.417)cycle ; ( .492,2.021)(-.492 , 2.021)(-2.405,.899)(-1.412,-.174)(1.412,-.174)(2.405 , .899)cycle ; ( 2.625,.433)(2.405 , .899)(1.412,-.174)(2.308,-.977)(2.573,-.48)cycle ; ( -2.625,.433)(-2.405 , .899)(-1.412,-.174)(-2.308,-.977)(-2.573,-.48)cycle ; + this equation governs shuffling two objects @xmath54 into two objects @xmath336 : + + ( -.453 , -1.75)(-2.308,-.977)(-1.412,-.174)(1.412,-.174)(2.308 , -.977)(.453,-1.75)cycle ; ( .492 , 2.021)(-.492,2.021)(-2.405,.899)(-1.412,-.174)(1.412 , -.174)(2.405,.899)cycle ; ( 2.625 , .433)(2.405,.899)(1.412,-.174)(2.308,-.977)(2.573,-.48)cycle ; ( -2.625 , .433)(-2.405,.899)(-1.412,-.174)(-2.308,-.977)(-2.573 , -.48)cycle ; ( a ) at ( -2.573,-.48 ) ; at ( a ) @xmath268 ; + \(b ) at ( -2.625,.433 ) edge [ < - ] node [ l , left ] @xmath111 ( a ) ; at ( b ) @xmath269 ; + \(c ) at ( -2.405,.899 ) edge [ < - ] node [ l , left ] @xmath131 ( b ) ; at ( c ) @xmath282 ; + \(d ) at ( -.492,2.021 ) edge [ < - ] node [ l , above left ] @xmath251 ( c ) ; at ( d ) @xmath337 ; + \(e ) at ( .492,2.021 ) edge [ < - ] node [ l , above ] @xmath260 ( d ) ; at ( e ) @xmath338 ; + \(f ) at ( 2.405,.899 ) edge [ < - ] node [ l , above right ] @xmath252 ( e ) ; at ( f ) @xmath339 ; + \(g ) at ( 2.625,.433 ) edge [ < - ] node [ l , right ] @xmath340 ( f ) ; at ( g ) @xmath341 ; + \(h ) at ( 2.573,-.48 ) edge [ < - ] node [ l , right ] @xmath111 ( g ) ; at ( h ) @xmath342 ; + \(i ) at ( 2.308,-.977 ) edge [ - > ] node [ l , right ] @xmath343 ( h ) ; at ( i ) @xmath344 ; + \(j ) at ( .453,-1.75 ) edge [ - > ] node [ l , below right ] @xmath345 ( i ) ; at ( j ) @xmath346 ; + \(k ) at ( -.453,-1.75 ) edge [ - > ] node [ l , below ] @xmath111 ( j ) ; at ( k ) @xmath347 ; + \(l ) at ( -2.308,-.977 ) edge [ - > ] node [ l , below left ] @xmath290 ( k ) edge [ < - ] node [ l , left ] @xmath313 ( a ) ; at ( l ) @xmath280 ; + \(m ) at ( -1.412,-.174 ) @xmath281 edge [ < - ] node [ l , below right ] @xmath111 ( l ) edge [ < - ] node [ l , above right ] @xmath260 ( c ) ; + \(n ) at ( 1.412,-.174 ) @xmath348 edge [ < - ] node [ l , above ] @xmath244 ( m ) edge [ - > ] node [ l , below left ] @xmath111 ( i ) edge [ - > ] node [ l , above left ] @xmath260 ( f ) ; + at ( -2.2,0 ) @xmath349 ; at ( 0,1 ) @xmath350 ; at ( 0,-1 ) @xmath351 ; at ( 2.2,0 ) @xmath352 ; + + = + + ( 0 , -.807)(-.944,.116)(-1.678,-.488)(-.546,-1.564)cycle ; ( .944,.116)(0 , -.807)(.546,-1.564)(1.678,-.488)cycle ; ( .944,.116)(0 , -.807)(-.944,.116)(0,1.079)cycle ; ( .546 , -1.564)(.453,-1.75)(-.453,-1.75)(-.546,-1.564)(0,-.807)cycle ; ( 0 , 1.079)(-.944,.116)(-1.72,.642)(-.588,1.829)cycle ; ( .944,.116)(0 , 1.079)(.588,1.829)(1.72,.642)cycle ; ( 1.678 , -.488)(2.075,-.32)(2.109,.417)(1.72,.642)(.944,.116)cycle ; ( -1.72 , .642)(-2.109,.417)(-2.075,-.32)(-1.678,-.488)(-.944 , .116)cycle ; ( .588 , 1.829)(.492,2.021)(-.492,2.021)(-.588,1.829)(0,1.079)cycle ; ( 1.678 , -.488)(2.075,-.32)(2.573,-.48)(2.308,-.977)(.453,-1.75)(.546 , -1.564)cycle ; ( -.546 , -1.564)(-.453,-1.75)(-2.308,-.977)(-2.573,-.48)(-2.075 , -.32)(-1.678,-.488)cycle ; ( 1.72 , .642)(.588,1.829)(.492,2.021)(2.405,.899)(2.625,.433)(2.109 , .417)cycle ; ( -.588 , 1.829)(-.492,2.021)(-2.405,.899)(-2.625,.433)(-2.109 , .417)(-1.72,.642)cycle ; ( 2.075 , -.32)(2.573,-.48)(2.625,.433)(2.109,.417)cycle ; ( -2.075 , -.32)(-2.573,-.48)(-2.625,.433)(-2.109,.417)cycle ; + \(a ) at ( -2.573,-.48 ) ; at ( a ) @xmath268 ; + \(b ) at ( -2.625,.433 ) edge [ < - ] node [ l , left ] @xmath111 ( a ) ; at ( b ) @xmath269 ; + \(c ) at ( -2.405,.899 ) edge [ < - ] node [ l , left ] @xmath131 ( b ) ; at ( c ) @xmath282 ; + \(d ) at ( -.492,2.021 ) edge [ < - ] node [ l , above left ] @xmath251 ( c ) ; at ( d ) @xmath337 ; + \(e ) at ( .492,2.021 ) edge [ < - ] node [ l , above ] @xmath260 ( d ) ; at ( e ) @xmath338 ; + \(f ) at ( 2.405,.899 ) edge [ < - ] node [ l , above right ] @xmath252 ( e ) ; at ( f ) @xmath339 ; + \(g ) at ( 2.625,.433 ) edge [ < - ] node [ l , right ] @xmath340 ( f ) ; at ( g ) @xmath341 ; + \(h ) at ( 2.573,-.48 ) edge [ < - ] node [ l , right ] @xmath111 ( g ) ; at ( h ) @xmath342 ; + \(i ) at ( 2.308,-.977 ) edge [ - > ] node [ l , right ] @xmath343 ( h ) ; at ( i ) @xmath344 ; + \(j ) at ( .453,-1.75 ) edge [ - > ] node [ l , below right ] @xmath345 ( i ) ;
at ( j ) @xmath346 ; + \(k ) at ( -.453,-1.75 ) edge [ - > ] node [ l , below ] @xmath111 ( j ) ; at ( k ) @xmath347 ; + \(l ) at ( -2.308,-.977 ) edge [ - > ] node [ l , below left ] @xmath290 ( k ) edge [ < - ] node [ l , left ] @xmath313 ( a ) ; at ( l ) @xmath280 ; + \(o ) at ( -2.075,-.32 ) edge [ < - ] node [ l , below , sloped ] @xmath353 ( a ) ; at ( o ) @xmath354 ; + \(p ) at ( -2.109,.417 ) edge [ < - ] node [ l , left ] @xmath111 ( o ) edge [ < - ] node [ l , above , sloped ] @xmath355 ( b ) ; at ( p ) @xmath269 ; + \(q ) at ( -1.72,.642 ) edge [ < - ] node [ l , above , sloped ] @xmath131 ( p ) ; at ( q ) @xmath356 ; + \(r ) at ( -.944,.116 ) @xmath357 edge [ < - ] node [ l , above right ] @xmath260 ( q ) ; + \(s ) at ( -1.678,-.488 ) edge [ - > ] node [ l , below right ] @xmath111 ( r ) edge [ < - ] node [ l , below , sloped ] @xmath313 ( o ) ; at ( s ) @xmath358 ; + \(t ) at ( -.546,-1.564 ) edge [ < - ] node [ l , above , sloped ] @xmath359 ( s ) edge [ < - ] node [ right ] @xmath313 ( k ) ; at ( t ) @xmath360 ; + \(u ) at ( 0,-.807 ) @xmath361 edge [ < - ] node [ l , above left ] @xmath111 ( t ) edge [ < - ] node [ l , above , sloped ] @xmath362 ( r ) ; + \(v ) at ( .546,-1.564 ) edge [ < - ] node [ l , above right ] @xmath111 ( u ) edge [ - > ] node [ left ] @xmath343 ( j ) ; at ( v ) @xmath363 ; + \(w ) at ( 1.678,-.488 ) edge [ < - ] node [ l , above , sloped ] @xmath364 ( v ) ; at ( w ) @xmath365 ; + \(x ) at ( 2.075,-.32 ) edge [ < - ] node [ l , below , sloped ] @xmath343 ( w ) edge [ - > ] node [ l , below , sloped ] @xmath366 ( h ) ; at ( x ) @xmath367 ; + \(y ) at ( 2.109,.417 ) edge [ - > ] node [ l , right ] @xmath111 ( x ) edge [ - > ] node [ l , above , sloped ] @xmath368 ( g ) ; at ( y ) @xmath369 ; + \(z ) at ( 1.72,.642 ) edge [ - > ] node [ l , above , sloped ] @xmath340 ( y ) ; at ( z ) @xmath370 ; + ( a ) at ( .944,.116 ) @xmath371 edge [ - > ] node [ l , above left ] @xmath260 ( z ) edge [ - > ] node [ l , below left ] @xmath111 ( w ) edge [ < - ] node [ l , above , sloped ] @xmath372 ( u ) ; + ( b ) at ( 0,1.079 ) @xmath373 edge [ - > ] node [ l , below , sloped ] @xmath374 ( a ) edge [ < - ] node [ l , below , sloped ] @xmath375 ( r ) ; + ( c ) at ( -.588,1.829 ) edge [ - > ] node [ l , below left ] @xmath260 ( b ) edge [ < - ] node [ l , below , sloped ] @xmath376 ( q ) edge [ < - ] node [ right ] @xmath131 ( d ) ; at ( c ) @xmath377 ; + ( d ) at ( .588,1.829 ) edge [ < - ] node [ l , below right ] @xmath260 ( b ) edge [ - > ] node [ l , below , sloped ] @xmath378 ( z ) edge [ - > ] node [ left ] @xmath340 ( e ) ; at ( d ) @xmath379 ; + [ font= ] at ( -2.4,0 ) @xmath138 ; at ( -1.7,0 ) ; ; at ( -1.5,0 ) @xmath380 ; at ( -2,-.8 ) ; ; at ( -1.8,-.8 ) @xmath381 ; at ( -2,.9 ) ; ; at ( -1.75,.9 ) @xmath382 ; at ( -0.1,1.6 ) ; ; at ( 0.1,1.6 ) @xmath383 ; at ( -.75,.75 ) @xmath138 ; at ( -.75,-.75 ) @xmath138 ; at ( 0,0 ) @xmath138 ; at ( .75,-.75 ) @xmath138 ; at ( .75,.75 ) @xmath138 ; at ( -0.1,-1.5 ) ; ; at ( 0.1,-1.5 ) @xmath384 ; at ( 1.8,.9 ) ; ; at ( 2.05,.9 ) @xmath385 ; at ( 1.8,-.8 ) ; ; at ( 2 , -.8 ) @xmath386 ; at ( 1.5,0 ) ; ; at ( 1.7,0 ) @xmath387 ; at ( 2.4,0 ) @xmath138 ; * the breen polytope . in a braided monoidal category , the yang - baxter equations hold ; there are two fundamentally distinct proofs of this fact .
+ ( acb ) at ( 5,0 ) acb ; ( abc ) at ( 0,3 ) abc edge [ - > ] node [ l , below left ] @xmath251 ( acb ) ; ( cab ) at ( 10,3 ) cab edge [ < - ] node [ l , below right ] @xmath388 ( acb ) ; ( bac ) at ( 0,9 ) bac edge [ < - ] node [ l , left ] @xmath389 ( abc ) ; ( cba ) at ( 10,9 ) cba edge [ < - ] node [ l , above left ] @xmath244 ( acb ) edge [ < - ] node [ l , right ] @xmath345 ( cab ) ; ( bca ) at ( 5,12 ) bca edge [ < - ] node [ l , above left ] @xmath247
( bac ) edge [ < - ] node [ l , below right ] @xmath244 ( abc ) edge [ - > ] node [ l , above right ] @xmath390 ( cba ) ; + ( acb ) at ( 5,0 ) acb ; ( abc ) at ( 0,3 ) abc edge [ - > ] node [ l , below left ] @xmath251 ( acb ) ; ( cab ) at ( 10,3 ) cab edge [ < - ] node [ l , above ] @xmath244 ( abc ) edge [ < - ] node [ l , below right ] @xmath388 ( acb ) ; ( bac ) at ( 0,9 ) bac edge [ < - ] node [ l , left ] @xmath389 ( abc ) ; ( cba ) at ( 10,9 ) cba edge [ < - ] node [ l , below ] @xmath244 ( bac ) edge [ < - ] node [ l , right ] @xmath345 ( cab ) ; ( bca ) at ( 5,12 ) bca edge [ < - ] node [ l , above left ] @xmath247
( bac ) edge [ - > ] node [ l , above right ] @xmath390 ( cba ) ; + in a braided strictly monoidal bicategory , the two proofs become the front and back face of another coherence law governing the interaction of the ( 2,1)-shuffle polytopes ; when the associator is nontrivial , the vertices get truncated .
that the coherence law is necessary was something of a surprise : kapranov and voevodsky did not include it in their definition of braided semistrict monoidal 2-categories ; breen @xcite corrected the definition .
we therefore call the following coherence law the `` breen polytope '' .
in retrospect , we can see that this is the start of a series of polytopes whose faces are `` permutohedra '' @xcite ; the need for the breen polytope and the rest of the series became clear in batanin s approach to weak @xmath2-categories @xcite .
+ ( 1,2.5)(0,4)(0 , 8)(1,9.5)(4,11.5)(6,11.5)cycle ; ( 9,9.5)(10 , 8)(10,4)(9,2.5)(6,0.5)(4,0.5)cycle ; ( 6,11.5)(9 , 9.5)(4,0.5)(1,2.5)cycle ; ( abc1 ) at ( 1 , 2.5 ) ; at ( abc1 ) @xmath216 ; ( abc2 ) at ( 0 , 4 ) edge [ < - ] node [ below left , l ] @xmath260 ( abc1 ) ; at ( abc2 ) @xmath220 ; ( bac1 ) at ( 0 , 8) edge [ < - ] node [ left , l ] @xmath246 ( abc2 ) ; at ( bac1 ) @xmath256 ; ( bac2 ) at ( 1 , 9.5 ) edge [ < - ] node [ left , l ] @xmath111 ( bac1 ) ; at ( bac2 ) @xmath257 ; ( bca1 ) at ( 4 , 11.5 ) edge [ < - ] node [ above left , l ] @xmath247 ( bac2 ) ; at ( bca1 ) @xmath254 ; ( bca2 ) at ( 6 , 11.5 ) edge [ < - ] node [ above , l ] @xmath260 ( bca1 ) edge [ < - ] node [ below right ] @xmath244 ( abc1 ) ; at ( bca2 ) @xmath255 ; ( cba1 ) at ( 9 , 9.5 ) edge [ < - ] node [ above right , l ] @xmath391 ( bca2 ) ; at ( cba1 ) @xmath392 ; ( cba2 ) at ( 10 , 8) edge [ - > ] node [ right , l ] @xmath260 ( cba1 ) ; at ( cba2 ) @xmath393 ; ( cab1 ) at ( 10 , 4 ) edge [ - > ] node [ right , l ] @xmath345 ( cba2 ) ; at ( cab1 ) @xmath259 ; ( cab2 ) at ( 9 , 2.5 ) edge [ - > ] node [ right , l ] @xmath111 ( cab1 ) ; at ( cab2 ) @xmath258 ; ( acb1 ) at ( 6 , 0.5 ) edge [ - > ] node [ below right , l ] @xmath252 ( cab2 ) ; at ( acb1 ) @xmath262 ; ( acb2 ) at ( 4 , 0.5 ) edge [ - > ] node [ below , l ] @xmath260 ( acb1 ) edge [ - > ] node [ below right , l ] @xmath244 ( cba1 ) edge [ < - ] node [ below left , l ] @xmath251 ( abc1 ) ; at ( acb2 ) @xmath261 ; at ( 2 , 7.5 ) @xmath394 ; at ( 5 , 6 ) @xmath138 ; at ( 8 , 4.5 ) @xmath395 ; + + = + + ( 0,8)(1,9.5)(4 , 11.5)(6,11.5)(9,9.5)(10,8)cycle ; ( 10,4)(9 , 2.5)(6,0.5)(4,0.5)(1,2.5)(0,4)cycle ; ( 0,4)(0,8)(10 , 8)(10,4)cycle ; ( abc1 ) at ( 1 , 2.5 ) ; at ( abc1 ) @xmath216 ; ( abc2 ) at ( 0 , 4 ) edge [ < - ] node [ below left , l ] @xmath260 ( abc1 ) ; at ( abc2 ) @xmath220 ; ( bac1 ) at ( 0 , 8) edge [ < - ] node [ left , l ] @xmath246 ( abc2 ) ; at ( bac1 ) @xmath256 ; ( bac2 ) at ( 1 , 9.5 ) edge [ < - ] node [ left , l ] @xmath111 ( bac1 ) ; at ( bac2 ) @xmath257 ; ( bca1 ) at ( 4 , 11.5 ) edge [ < - ] node [ above left , l ] @xmath247 ( bac2 ) ; at ( bca1 ) @xmath254 ; ( bca2 ) at ( 6 , 11.5 ) edge [ < - ] node [ above , l ] @xmath260 ( bca1 ) ; at ( bca2 ) @xmath255 ; ( cba1 ) at ( 9 , 9.5 ) edge [ < - ] node [ above right , l ] @xmath391 ( bca2 ) ; at ( cba1 ) @xmath392 ; ( cba2 ) at ( 10 , 8) edge [ - > ] node [ right , l ] @xmath260 ( cba1 ) edge [ < - ] node [ below , l ] @xmath244 ( bac1 ) ; at ( cba2 ) @xmath393 ; ( cab1 ) at ( 10 , 4 ) edge [ - > ] node [ right , l ] @xmath345 ( cba2 ) edge [ < - ] node [ below , l ] @xmath244 ( abc2 ) ; at ( cab1 ) @xmath259 ; ( cab2 ) at ( 9 , 2.5 ) edge [ - > ] node [ right , l ] @xmath111 ( cab1 ) ; at ( cab2 ) @xmath258 ; ( acb1 ) at ( 6 , 0.5 ) edge [ - > ] node [ below right , l ] @xmath252 ( cab2 ) ; at ( acb1 ) @xmath262 ; ( acb2 ) at ( 4 , 0.5 ) edge [ - > ] node [ below , l ] @xmath260 ( acb1 ) edge [ < - ] node [ below left , l ] @xmath251 ( abc1 ) ; at ( acb2 ) @xmath261 ; at ( 5 , 10 ) @xmath396 ; at ( 5 , 6 ) @xmath138 ; at ( 5 , 2 ) @xmath397 ; in a * sylleptic * monoidal bicategory , a full twist is not necessarily equal to the identity , but may be only equivalent to it ; this adjoint equivalence is called a * syllepsis*. a sylleptic monoidal bicategory @xmath53 is a braided monoidal bicategory equipped with a syllepsis subject to the following axioms .
+ * an invertible modification ( mnemonic : salmon syllepsis ) + ( 1.5,0 ) ellipse ( 37pt and 22pt ) ; ( ab ) at ( 0,0 ) @xmath398 ; ( ba ) at ( 3,0 ) @xmath399 edge [ < - , out=135 , in=45 ] node [ above , l ] @xmath244 ( ab ) edge [ < - , out=-135 , in=-45 ] node [ below , l ] @xmath400 ( ab ) ; at ( 1.5,0 ) @xmath401 ; * this equation governs the interaction of the syllepsis with the @xmath239 braiding : + ( 0:2cm)(60:2 cm ) arc(0:180:1 cm and 5mm)(180:2 cm ) [ rotate=120 ] arc(0:180:1 cm and 5mm)(180:2 cm ) [ rotate=120 ] arc(0:180:1 cm and 5mm)cycle ; ( a ) at ( 0:2 cm ) ; at ( a ) [ right ] @xmath254 ; ( b ) at ( 60:2 cm ) edge [ - > ] node [ l , above right ] @xmath111 ( a ) ; at ( b ) [ above right ] @xmath255 ; ( c ) at ( 120:2 cm ) edge [ - > , out=45 , in=135 ] node [ l , above ] @xmath244 ( b ) ; at ( c ) [ above left ] @xmath216 ; ( d ) at ( 180:2 cm ) edge [ - > ] node [ l , above left ] @xmath111 ( c ) ; at ( d ) [ left ] @xmath220 ; ( e ) at ( 240:2 cm ) edge [ < - , out=165 , in=255 ] node [ l , below left ] @xmath246 ( d ) ; at ( e ) [ below left ] @xmath256 ; ( f ) at ( 300:2 cm ) edge [ < - ] node [ l , below ] @xmath111 ( e ) edge [ - > , out=15 , in=285 ] node [ l , below right ] @xmath247 ( a ) ; at ( f ) [ below right ] @xmath257 ; at ( 0,0 ) @xmath248 ; at ( 0:3.7 cm ) = ; + ( 0:2cm)(60:2 cm ) arc(0:-180:1 cm and 5mm)(180:2 cm ) [ rotate=120 ] arc(0:-180:1 cm and 5mm)(180:2 cm ) [ rotate=120 ] arc(0:-180:1 cm and 5mm)cycle ; ( 60:2 cm ) arc(0:360:1 cm and 5 mm ) ; ( 180:2 cm ) [ rotate=120 ] arc(0:360:1 cm and 5 mm ) ; ( 300:2 cm ) [ rotate=240 ] arc(0:360:1 cm and 5 mm ) ; ( a ) at ( 0:2 cm ) ; at ( a ) [ right ] @xmath254 ; ( b ) at ( 60:2 cm ) edge [ - > ] node [ l , above right ] @xmath111 ( a ) ; at ( b ) [ above right ] @xmath255 ; ( c ) at ( 120:2 cm ) edge [ - > , out=45 , in=135 ] node [ l , above ] @xmath244 ( b ) edge [ - > , out=-45 , in=-135 ] node [ l , below ] @xmath400 ( b ) ; at ( c ) [ above left ] @xmath216 ; ( d ) at ( 180:2 cm ) edge [ - > ] node [ l , above left ] @xmath111 ( c ) ; at ( d ) [ left ] @xmath220 ; ( e ) at ( 240:2 cm ) edge [ < - , out=165 , in=255 ] node [ l , below left ] @xmath246 ( d ) edge [ < - , out=75 , in=-15 ] node [ l , above right ] @xmath402 ( d ) ; at ( e ) [ below left ] @xmath256 ; ( f ) at ( 300:2 cm ) edge [ < - ] node [ l , below ] @xmath111 ( e ) edge [ - > , out=15 , in=285 ] node [ l , below right ] @xmath247 ( a ) edge [ - > , out=105 , in=195 ] node [ l , above left ] @xmath403 ( a ) ; at ( f ) [ below right ] @xmath257 ; at ( 0,0 ) @xmath404 ; at ( 90:1.732 cm ) @xmath401 ; at ( 210:1.732 cm ) @xmath405 ; at ( 330:1.732 cm ) @xmath406 ; * this equation governs the interaction of the syllepsis with the @xmath240 braiding : + ( 0:2cm)(60:2 cm ) arc(0:180:1 cm and 5mm)(180:2 cm ) [ rotate=120 ] arc(0:180:1 cm and 5mm)(180:2 cm ) [ rotate=120 ] arc(0:180:1 cm and 5mm)cycle ; ( a ) at ( 0:2 cm ) ; at ( a ) [ right ] @xmath258 ; ( b ) at ( 60:2 cm ) edge [ - > ] node [ l , above right ] @xmath260 ( a ) ; at ( b ) [ above right ] @xmath259 ; ( c ) at ( 120:2 cm ) edge [ - > , out=45 , in=135 ] node [ l , above ] @xmath244 ( b ) ; at ( c ) [ above left ] @xmath220 ; ( d ) at ( 180:2 cm ) edge [ - > ] node [ l , above left ] @xmath260 ( c ) ; at ( d ) [ left ] @xmath216 ; ( e ) at ( 240:2 cm ) edge [ < - , out=165 , in=255 ] node [ l , below left ] @xmath251 ( d ) ; at ( e ) [ below left ] @xmath261 ; ( f ) at ( 300:2 cm ) edge [ < - ] node [ l , below ] @xmath260 ( e ) edge [ - > , out=15 , in=285 ] node [ l , below right ] @xmath252 ( a ) ; at ( f ) [ below right ] @xmath262 ; at ( 0,0 ) @xmath253 ; at ( 0:3.7 cm ) = ; + ( 0:2cm)(60:2 cm ) arc(0:-180:1 cm and 5mm)(180:2 cm ) [ rotate=120 ] arc(0:-180:1 cm and 5mm)(180:2 cm ) [ rotate=120 ] arc(0:-180:1 cm and 5mm)cycle ; ( 60:2 cm ) arc(0:360:1 cm and 5 mm ) ; ( 180:2 cm ) [ rotate=120 ] arc(0:360:1 cm and 5 mm ) ; ( 300:2 cm ) [ rotate=240 ] arc(0:360:1 cm and 5 mm ) ; ( a ) at ( 0:2 cm ) ; at ( a ) [ right ] @xmath258 ; ( b ) at ( 60:2 cm ) edge [ - > ] node [ l , above right ] @xmath260 ( a ) ; at ( b ) [ above right ] @xmath259 ; ( c ) at ( 120:2 cm ) edge [ - > , out=45 , in=135 ] node [ l , above ] @xmath244 ( b ) edge [ - > , out=-45 , in=-135 ] node [ l , below ] @xmath400 ( b ) ; at ( c ) [ above left ] @xmath220 ; ( d ) at ( 180:2 cm ) edge [ - > ] node [ l , above left ] @xmath260 ( c ) ; at ( d ) [ left ] @xmath216 ; ( e ) at ( 240:2 cm ) edge [ < - , out=165 , in=255 ] node [ l , below left ] @xmath251 ( d ) edge [ < - , out=75 , in=-15 ] node [ l , above right ] @xmath407 ( d ) ; at ( e ) [ below left ] @xmath261 ; ( f ) at ( 300:2 cm ) edge [ < - ] node [ l , below ] @xmath260 ( e ) edge [ - > , out=15 , in=285 ] node [ l , below right ] @xmath247 ( a ) edge [ - > , out=105 , in=195 ] node [ l , above left ] @xmath408 ( a ) ; at ( f ) [ below right ] @xmath262 ; at ( 0,0 ) @xmath409 ; at ( 90:1.732 cm ) @xmath401 ; at ( 210:1.732 cm ) @xmath410 ; at ( 330:1.732 cm ) @xmath411 ; a * symmetric * monoidal bicategory is a sylleptic monoidal bicategory subject to the following axiom , where the unlabeled green cells are identities : + * for all objects @xmath58 and @xmath90 of @xmath53 , the following equation holds : + + ( 0,4)(6,4)(6 , 0)cycle ; ( 0,4)(0,0)(6 , 0)cycle ; ( ab1 ) at ( 0,4 ) @xmath102 ; ( ba1 ) at ( 6,4 ) @xmath412 edge [ < - ] node [ l , above ] @xmath244 ( ab1 ) ; ( ba2 ) at ( 0,0 ) @xmath412 edge [ < - ] node [ l , left ] @xmath244 ( ab1 ) ; ( ab2 ) at ( 6,0 ) @xmath102 edge [ - > ] node [ l , right ] @xmath244 ( ba1 ) edge [ < - ] node [ l , above right ] 1 ( ab1 ) edge [ < - ] node [ l , below ] @xmath244 ( ba2 ) ; ( v ) at ( 2,1 ) @xmath413 ; ( 1 ) at ( 5,3 ) @xmath414 ; ( =) at ( 7,2 ) = ; + ( 0,4)(6,4)(0 , 0)cycle ; ( 0,0)(6,0)(6 , 4)cycle ; ( ab1 ) at ( 0,4 ) @xmath102 ; ( ba1 ) at ( 6,4 ) @xmath412 edge [ < - ] node [ l , above ] @xmath244 ( ab1 ) ; ( ba2 ) at ( 0,0 ) @xmath412 edge [ - > ] node [ l , above left ] 1 ( ba1 ) edge [ < - ] node [ l , left ] @xmath244 ( ab1 ) ; ( ab2 ) at ( 6,0 ) @xmath102 edge [ - > ] node [ l , right ] @xmath244 ( ba1 ) edge [ < - ] node [ l , below ] @xmath244 ( ba2 ) ; ( v ) at ( 4,1 ) @xmath413 ; ( 1 ) at ( 1,3 ) @xmath414 ; given two bicategories @xmath415 , two functors @xmath416 and @xmath417 are * pseudoadjoint * if for all @xmath418 the categories hom@xmath419 and hom@xmath420 are adjoint equivalent pseudonaturally in @xmath58 and @xmath90 .
monoidal closed bicategories satisfy the obvious weakening of the definition of monoidal closed categories : a monoidal * closed * bicategory is one in which for every object @xmath58 , the functor @xmath421 has a right pseudoadjoint @xmath422 similarly , compact closed bicategories weaken the notion of duality from compact closed categories .
a * compact closed * bicategory is a symmetric monoidal bicategory in which every object has a pseudoadjoint .
this means that every object @xmath58 is equipped with a ( weak ) * dual * , an object @xmath423 equipped with two 1-morphisms @xmath424 ; @xmath50 ; and two `` zig - zag '' 2-isomorphisms ( mnemonic : yellow yanking or xanthic zig - zag ) @xmath425
\(a ) at ( 0,0 ) @xmath58 ; ( b ) at ( 2,1 ) @xmath426 edge [ - > ] node [ below right , l ] @xmath427 ( a ) ; ( c ) at ( 0,2 ) @xmath58 edge [ - > ] node [ left , l ] @xmath58 ( a ) edge [ - > ] node [ above right , l ] @xmath428 ( b ) ; at ( .5,1 ) [ above ] @xmath429 ; at ( .5,1 ) @xmath430 ; ( a.center ) to ( b.center ) to ( c.center ) to ( a.center ) ; @xmath431 ; at ( -1,1 ) @xmath430 ; at ( -1,1 ) @xmath429 ; ; ; ; ; @xmath432 \(a ) at ( 0,0 ) @xmath423 ; ( b ) at ( 2,1 ) @xmath433 edge [ - > ] node [ below right , l ] @xmath434 ( a ) ; ( c ) at ( 0,2 ) @xmath423 edge [ - > ] node [ left , l ] @xmath58 ( a ) edge [ - > ] node [ above right , l ] @xmath435 ( b ) ; at ( .5,1 ) [ above ] @xmath436 ; at ( .5,1 ) @xmath430 ; ( a.center ) to ( b.center ) to ( c.center ) to ( a.center ) ; @xmath431 ; at ( -1,1 ) @xmath430 ; at ( -1,1 ) @xmath436 ; ; ; ; ; such that the following `` swallowtail equation '' holds : \(a ) at ( 2,0 ) @xmath437 ; ( b ) at ( 2,2 ) @xmath438 edge [ - > ] node [ right , l ] @xmath439 ( a ) ; ( c ) at ( 0,3 ) @xmath437 edge [ - > ] node [ below left , l ] @xmath437 ( a ) edge [ - > ] node [ above right , l ] @xmath440 ( b ) ; at ( 1.5,1.5 ) [ above ] @xmath441 ; at ( 1.5,1.5 ) @xmath430 ; ( a.center ) to ( b.center ) to ( c.center ) to ( a.center ) ; \(d ) at ( 4,3 ) @xmath437 edge [ - > ] node [ below right , l ] @xmath437 ( a ) edge [ - > ] node [ above left , l ] @xmath442 ( b ) ; ( a.center ) to ( b.center ) to ( d.center ) to ( a.center ) ; at ( 2.5,1.5 ) [ above ] @xmath443 ; at ( 2.5,1.5 ) @xmath430 ; ( e ) at ( 2,4 ) @xmath143 edge [ - > ] node [ above left , l ] @xmath444 ( c ) edge [ - > ] node [ above right , l ] @xmath444 ( d ) ; at ( 2,3 ) @xmath138 ; ( b.center ) to ( d.center ) to ( e.center ) to ( c.center ) to ( b.center ) ; at ( 5,2.5 ) @xmath97 ; ( f ) at ( 6,0 ) @xmath437 ; ( g ) at ( 6,4 ) @xmath143 edge [ - > ] node [ right , l ] @xmath444 ( f ) ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; at ( 2,1.5 ) [ above ] @xmath445 ; at ( 2,1.5 ) @xmath430 ; at ( 7,1.5 ) @xmath138 ; at ( 12,1.5 ) [ above ] @xmath446 ; at ( 12,1.5 ) @xmath430 ; ( b ) at ( 13.5 , -.5 ) ; ( a ) at ( .5,-.5 ) edge [ very thick , double distance=2pt , bend right=10 ] ( b ) ; we have drawn the diagrams in a strictly monoidal compact closed bicategory for clarity ; when the associator is not the identity , we truncate some corners : \(a ) at ( 30:2 cm ) @xmath447 ; ( b ) at ( 90:2 cm ) @xmath448 edge [ - > ] node [ above right , l ] @xmath428 ( a ) ; ( c ) at ( 150:2 cm ) @xmath58 edge [ - > ] node [ above left , l ] @xmath449 ( b ) ; ( d ) at ( 210:2 cm ) @xmath58 edge [ < - ] node [ left , l ] @xmath58 ( c ) ; ( e ) at ( 270:2 cm ) @xmath450 edge [ - > ] node [ below left , l ] @xmath32 ( d ) ; ( f ) at ( 330:2 cm ) @xmath451 edge [ - > ] node [ below right , l ] @xmath427 ( e ) edge [ < - ] node [ right , l ] @xmath111 ( a ) ; at ( 90:.25 cm ) @xmath429 ; at ( -90:.25 cm ) @xmath430 ; ( a.center ) to ( b.center ) to ( c.center ) to ( d.center ) to ( e.center ) to ( f.center ) to ( a.center ) ; \(a ) at ( 30:2 cm ) @xmath452 ; ( b ) at ( 90:2 cm ) @xmath453 edge [ - > ] node [ above right , l ] @xmath435 ( a ) ; ( c ) at ( 150:2 cm ) @xmath423 edge [ - > ] node [ above left , l ] @xmath454 ( b ) ; ( d ) at ( 210:2 cm ) @xmath423 edge [ < - ] node [ left , l ] @xmath423 ( c ) ; ( e ) at ( 270:2 cm ) @xmath455 edge [ - > ] node [ below left , l ] @xmath147 ( d ) ; ( f ) at ( 330:2 cm ) @xmath456 edge [ - > ] node [ below right , l ] @xmath434 ( e ) edge [ < - ] node [ right , l ] @xmath260 ( a ) ; at ( 90:.25 cm ) @xmath436 ; at ( -90:.25 cm ) @xmath430 ; ( a.center ) to ( b.center ) to ( c.center ) to ( d.center ) to ( e.center ) to ( f.center ) to ( a.center ) ; ( 6,0)(4,3)(8 , 3)cycle ; ( 4,6)(8,6)(8 , 3)(4,3)cycle ; ( 4,6)(8,6)(9,9)(6 , 11)(3,9)cycle ; ( 3,9)(0,11)(3 , 13)(6,11)cycle ; ( 6,11)(9,9)(12 , 11)(9,13)cycle ; ( 6,11)(3,13)(6 , 14)(9,13)cycle ; ( 3,13)(3,16)(6 , 17)(6,14)cycle ; ( 6,17)(6,14)(9 , 13)(9,16)cycle ; ( 0,11)(3,13)(3 , 16)cycle ; ( 12,11)(9,13)(9 , 16)cycle ; \(a ) at ( 6,0 ) @xmath437 ; ( b ) at ( 4,3 ) @xmath457 edge [ - > ] node [ below left , l ] @xmath458 ( a ) ; ( c ) at ( 8,3 ) @xmath459 edge [ < - ] node [ above , l ] @xmath111 ( b ) edge [ - > ] node [ below right , l ] @xmath187 ( a ) ; at ( 6,2 ) [ above ] @xmath460 ; at ( 6,2 ) @xmath430 ; ( d ) at ( 4,6 ) @xmath461 edge [ - > ] node [ left , l ] @xmath462 ( b ) ; ( e ) at ( 8,6 ) @xmath463 edge [ < - ] node [ above , l ] @xmath111 ( d ) edge [ - > ] node [ right ,
l ] @xmath464 ( c ) ; at ( 6,4.5 ) @xmath138 ; ( f ) at ( 3,9 ) @xmath465 edge [ - > ] node [ left , l ] @xmath466 ( d ) ; ( g ) at ( 9,9 ) @xmath467 edge [ - > ] node [ right , l ] @xmath310 ( e ) ; ( h ) at ( 0,11 ) @xmath468 edge [ - > ] node [ below left , l ] @xmath469 ( f ) ; ( i ) at ( 6,11 ) @xmath470 edge [ - > ] node [ below right , l ] @xmath260 ( f ) edge [ - > ] node [ below left , l ] @xmath111 ( g ) ; at ( 6,8 ) [ above ] @xmath471 ; at ( 6,8 ) @xmath430 ; ( j ) at ( 12,11 ) @xmath472 edge [ - > ] node [ below right , l ] @xmath473 ( g ) ; ( k ) at ( 3,13 ) @xmath474 edge [ - > ] node [ below right , l ] @xmath260 ( h ) edge [ - > ] node [ below left , l ] @xmath475 ( i ) ; at ( 3,10.5 ) @xmath138 ; ( l ) at ( 9,13 ) @xmath476 edge [ - > ] node [ below right , l ] @xmath477 ( i ) edge [ - > ] node [ below left , l ] @xmath111 ( j ) ; at ( 9,10.5 ) @xmath138 ; ( m ) at ( 3,16 ) @xmath437 edge [ - > ] node [ above left , l ] @xmath478 ( h ) edge [ - > ] node [ right , l ] @xmath449 ( k ) ; ( a.center ) to ( b.center ) to ( d.center ) to ( f.center ) to ( h.center ) to ( m.center ) to ( m.west ) to [ bend right=90 ] ( a.west ) to ( a.center ) ; ( m.west ) to [ bend right=90 ] node [ left , l ] @xmath437 ( a.west ) ; at ( 2.5,13.5 ) [ above ] @xmath479 ; at ( 2.5,13.5 ) @xmath430 ; ( n ) at ( 9,16 ) @xmath437 edge [ - > ] node [ left , l ] @xmath454 ( l ) edge [ - > ] node [ above right , l ] @xmath480 ( j ) ; ( a.center ) to ( c.center ) to ( e.center ) to ( g.center ) to ( j.center ) to ( n.center ) to ( n.east ) to [ bend left=90 ] ( a.east ) to ( a.center ) ; ( n.east ) to [ bend left=90 ] node [ right , l ] @xmath437 ( a.east ) ; at ( 9.5,13.5 ) [ above ] @xmath481 ; at ( 9.5,13.5 ) @xmath430 ; ( o ) at ( 6,14 ) @xmath482 edge [ - > ] node [ above left , l ] @xmath483 ( k ) edge [ - > ] node [ above right , l ] @xmath484 ( l ) ; at ( 6,13 ) @xmath138 ; ( p ) at ( 6,17 ) @xmath143 edge [ - > ] node [ above , l ] @xmath444 ( m ) edge [ - > ] node [ above , l ] @xmath444 ( n ) edge [ - > ] node [ left , l ] @xmath449 node [ right , l ] @xmath454 ( o ) ; at ( 4.5,15 ) @xmath138 ; at ( 7.5,15 ) @xmath138 ; at ( 1,6 ) [ above ] @xmath441 ; at ( 1,6 ) @xmath430 ; at ( 11,6 ) [ above ] @xmath443 ; at ( 11,6 ) @xmath430 ; at ( 15 , 8.5 ) @xmath97 ; ( a ) at ( 17,0 ) @xmath437 ; ( p ) at ( 17,17 ) @xmath143 edge [ - > ] node [ left , l ] @xmath444 ( a ) ;
at the start of this paper , we saw that spans of sets form a compact closed bicategory .
street @xcite suggested weakening the notion of a map of spans to hold only up to 2-isomorphism , allowing to define spans in bicategories rather than mere categories ; hoffnung @xcite worked out the details . a * span * @xmath485 in a bicategory @xmath90 is a pair of morphisms with the same source : @xmath486 . a * map of spans * @xmath9 is a triple @xmath487 consisting of a morphism @xmath488 and two invertible 2-morphisms @xmath489 : \(s ) at ( 1,2 ) @xmath6 ; ( y ) at ( 0,1 ) @xmath5 edge [ < - ] node [ l , above left ] @xmath12(s ) ; ( x ) at ( 2,1 ) @xmath4 edge [ < - ] node [ l , above right ] @xmath13(s ) ; ( s ) at ( 1,0 ) @xmath20 edge [ - > ] node [ l , below left ] @xmath490(y ) edge [ - > ] node [ l , below right ] @xmath491(x ) edge [ < - ] node [ l , left ] @xmath9 ( s ) ; at ( 0.5,1 ) @xmath492 ; at ( 1.5,1 ) @xmath493 ;
a * map of maps of spans * is a 2-morphism @xmath494 such that @xmath495 and @xmath496 .
maps and maps of maps compose in the obvious ways .
hoffnung showed that any 2-category @xmath11 with finite products and weak pullbacks ( also known as iso - comma objects ) gives rise to a monoidal tricategory span(@xmath11 ) whose * objects are objects of @xmath11 , * morphisms are spans in @xmath11 , * 2-morphisms are maps of spans , and * 3-morphisms are maps of maps of spans ; the tensor product of two spans @xmath497 and @xmath498 is the span @xmath499 here we show that the bicategory @xmath500 whose * objects are objects of @xmath11 , * morphisms are spans in @xmath11 , and * 2-morphisms are 3-isomorphism classes of maps of spans .
forms a compact closed bicategory .
[ lemma ] given spans @xmath501 and @xmath502 in @xmath11 and a morphism @xmath503 in @xmath11 such that @xmath504 the composite span @xmath505 is isomorphic to the span @xmath506 _ proof .
_ we form the composite of the two spans using the weak pullback .
\(a ) at ( 4,0 ) @xmath58 ; ( b ) at ( 2,0 ) @xmath90 ; ( c ) at ( 0,0 ) @xmath0 ; ( r ) at ( 3,1 ) @xmath1 edge [ - > ] node [ above right , l ] @xmath9 ( a ) edge [ - > ] node [ above left , l ] @xmath8 ( b ) ; ( s ) at ( 1,1 ) @xmath6 edge [ - > ] node [ above right , l ] @xmath23 ( b ) edge [ - > ] node [ above left , l ] @xmath91 ( c ) edge [ < - ] node [ below , l ] @xmath2 ( r ) ; ( sr ) at ( 2,2 ) @xmath507 edge [ - > ] node [ above right , l ] @xmath508 ( r ) edge [ - > ] node [ above left , l ] @xmath509 ( s ) ; at ( 2,1.25 ) @xmath510 ; at ( 2,.5 ) @xmath97 ; ( 1.8,1.7)(2,1.5)(2.2,1.7 ) ; by the first part of the universal property of the weak pullback , there is a unique morphism @xmath511 such that the following equation holds .
\(b ) at ( 2,0 ) @xmath90 ; ( s ) at ( 1,1 ) @xmath6 edge [ - > ] node [ above right , l ] @xmath23 ( b ) ; ( r ) at ( 3,1 ) @xmath1 edge [ - > ] node [ above left , l ] @xmath8 ( b ) edge [ - > ] node [ below , l ] @xmath2 ( s ) ; ( sr ) at ( 2,2 ) @xmath507 edge [ - > ] node [ above right , l ] @xmath508 ( r ) edge [ - > ] node [ above left , l ] @xmath509 ( s ) ; at ( 2,1.25 ) @xmath510 ; at ( 2,0.5 ) @xmath97 ; ( 1.8,1.7)(2,1.5)(2.2,1.7 ) ; at ( 2 , 3 ) @xmath1 edge [ - > , dashed ] node [ right , l ] @xmath512 ( sr ) edge [ - > , bend right ] node [ left , l ] @xmath2 ( s ) edge [ - > , bend left ] node [ right , l ] @xmath1 ( r ) ; at ( 1.5 , 2 ) @xmath97 ; at ( 2.5 , 2 ) @xmath97 ; at ( 3.5 , 1.5 ) @xmath97 ; \(b ) at ( 2,0 ) @xmath90 ; ( r ) at ( 3,1 ) @xmath1 edge [ - > ] node [ above left , l ] @xmath8 ( b ) ; ( s ) at ( 1,1 ) @xmath6 edge [ - > ] node [ above right , l ] @xmath23 ( b ) ; at ( 2 , 1.5 ) @xmath97 ; at ( 2 , 3 ) @xmath1 edge [ - > , bend right ] node [ left , l ] @xmath2 ( s ) edge [ - > , bend left ] node [ right , l ] @xmath1 ( r ) ; therefore , the following maps of spans compose to form the identity : \(a ) at ( 0,0 ) @xmath6 ; ( b ) at ( 2,0 ) @xmath1 ; ( c ) at ( 1,-1 ) @xmath1 edge [ - > ] node[l , below left]@xmath2(a ) edge [ - > ] node[l , below right]@xmath1(b ) ; ( d ) at ( 1,0 ) @xmath507 edge [ - > ] node[l , above]@xmath513(a ) edge [ - > ] node[l , above]@xmath514(b ) edge [ - > ] node[l , right]@xmath514(c ) ; ( e ) at ( 1,1 ) @xmath1 edge [ - > ] node[l , above left]@xmath2(a ) edge [ - > ] node[l , above right]@xmath1(b ) edge [ - > ] node[l , right]@xmath12(d ) ; at ( .75,.25 ) ; ; at ( 1.25,.25 ) ; ; at ( 1.25,-.25 ) ; ; at ( .75,-.25 ) @xmath515 ; \(a ) at ( 0,0 ) @xmath6 ; ( b ) at ( 2,0 ) @xmath1 ; ( c ) at ( 1,-1 ) @xmath1 edge [ - > ] node[l , below left]@xmath2(a ) edge [ - > ] node[l , below right]@xmath1(b ) ; ( e ) at ( 1,1 ) @xmath1 edge [ - > ] node[l , above left]@xmath2(a ) edge [ - > ] node[l , above right]@xmath1(b ) edge [ - > ] node[l , right]@xmath1(c ) ; at ( .5,0 ) ; ; at ( 1.5,0 ) ; ; at ( -.5,0 ) @xmath97 ; ( note that @xmath516 is whiskered by @xmath12 , so that 2-morphism is the identity . ) let @xmath517 be the following 2-morphism : at ( -1,1 ) @xmath518 ; ( sr1 ) at ( 1,0 ) @xmath507 ; ( sr2 ) at ( 0,1 ) @xmath507 edge [ - > ] node [ below left , l ] @xmath507 ( sr1 ) ; ( r ) at ( .5 , 1.5 ) @xmath1 edge [ < - ] node [ above left , l ] @xmath508 ( sr2 ) ; ( s ) at ( 2,1 ) @xmath6 edge [ < - ] node [ below right , l ] @xmath509 ( sr1 ) edge [ < - ] node [ below , l ] @xmath2 ( r ) ; ( sr3 ) at ( 1,2 ) @xmath507 edge [ < - ] node [ above left , l ] @xmath12 ( r ) edge [ - > ] node [ above right , l ] @xmath509 ( s ) ; at ( .85,.85 ) @xmath519 ; at ( 1,1.6 ) ; ; and @xmath520 be the following 2-morphism : at ( -1,1 ) @xmath521 ; ( r1 ) at ( 2,1 ) @xmath1 ; ( sr1 ) at ( 1,0 ) @xmath507 edge [ - > ] node [ below right , l ] @xmath508 ( r1 ) ; ( sr2 ) at ( 1,2 ) @xmath507 edge [ - > ] node [ above right , l ] @xmath508 ( r1 ) ; ( r2 ) at ( .5,1.5 ) @xmath1 edge [ - > ] node [ above left , l ] @xmath12 ( sr2 ) edge [ - > ] node [ below , l ] @xmath1 ( r1 ) ; ( sr3 ) at ( 0,1 ) @xmath507 edge [ - > ] node [ above left , l ] @xmath508 ( r2 ) edge [ - > ] node [ below left , l ] @xmath507 ( sr1 ) ; at ( 1,1.6 ) ; ; at ( .85,.85 ) ; ; the following equation holds : ( sr1 ) at ( 1,0 ) @xmath507 ; ( sr2 ) at ( 0,1 ) @xmath507 edge [ - > ] node [ below left , l ] @xmath507 ( sr1 ) ; ( r1 ) at ( .5 , 1.5 ) @xmath1 edge [ < - ] node [ above left , l ] @xmath508 ( sr2 ) ; ( s ) at ( 2,1 ) @xmath6 edge [ < - ] node [ below right , l ] @xmath509 ( sr1 ) edge [ < - ] node [ below , l ] @xmath2 ( r1 ) ; ( sr3 ) at ( 1,2 ) @xmath507 edge [ < - ] node [ above left , l ] @xmath12 ( r1 ) edge [ - > ] node [ above right , l ] @xmath509 ( s ) ; at ( .85,.85 ) @xmath519 ; at ( 1,1.6 ) ; ; ( b ) at ( 3,2 ) @xmath90 edge [ < - ] node [ below right , l ] @xmath23 ( s ) ; ( r2 ) at ( 2,3 ) @xmath1 edge [ < - ] node [ above left , l ] @xmath508 ( sr3 ) edge [ - > ] node [ above right , l ] @xmath8 ( b ) edge [ - > ] node [ right , l ] @xmath2 ( s ) ; at ( 1.65,2 ) @xmath519 ; at ( 2.5,2 ) ; ; at ( 3.5,2 ) @xmath97 ; \(b ) at ( 3,1 ) @xmath90 ; ( s ) at ( 2,0 ) @xmath6 edge [ - > ] node [ below right , l ] @xmath23 ( b ) ; ( r1 ) at ( 2,2 ) @xmath1 edge [ - > ] node [ above right , l ] @xmath8 ( b ) edge [ - > ] node [ right , l ] @xmath2 ( s ) ; ( sr1 ) at ( 1,1 ) @xmath507 edge [ - > ] node [ below left , l ] @xmath509 ( s ) edge [ - > ] node [ below right , l ] @xmath508 ( r1 ) ; at ( 1.65,1 ) @xmath519 ; at ( 2.5,1 ) ; ; ( sr2 ) at ( 1,3 ) @xmath507 edge [ - > ] node [ above right , l ] @xmath508 ( r1 ) ; ( r2 ) at ( .5,2.5 ) @xmath1 edge [ - > ] node [ above left , l ] @xmath12 ( sr2 ) edge [ - > ] node [ below , l ] @xmath1 ( r1 ) ; ( sr3 ) at ( 0,2 ) @xmath507 edge [ - > ] node [ above left , l ] @xmath508 ( r2 ) edge [ - > ] node [ below left , l ] @xmath507 ( sr1 ) ; at ( 1,2.6 ) ; ; at ( .85,1.85 ) ; ; ( note on the left hand side that the @xmath522 in the upper right is whiskered by @xmath523 so that 2-morphism is the identity by the definition of @xmath524 ) therefore , by the second part of the universal property of the weak pullback , there exists a unique invertible 2-morphism @xmath525 in @xmath11 such that @xmath526 and @xmath527 given @xmath528 we can show that the following composition of maps of spans is in the same equivalence class as the identity : \(a ) at ( 0,0 ) @xmath6 ; ( b ) at ( 2,0 ) @xmath1 ; ( c ) at ( 1,-1 ) @xmath507 edge [ - > ] node [ l , below left ] @xmath513 ( a ) edge [ - > ] node [ l , below right ] @xmath514 ( b ) ; ( d ) at ( 1,0 ) @xmath1 edge [ - > ] node [ l , below ] @xmath2 ( a ) edge [ - > ] node [ l , below ] @xmath1 ( b ) edge [ - > ] node [ l , right ] @xmath12 ( c ) ; ( e ) at ( 1,1 ) @xmath507 edge [ - > ] node [ l , above left ] @xmath513 ( a ) edge [ - > ] node [ l , above right ] @xmath514 ( b ) edge [ - > ] node [ l , right ] @xmath514 ( d ) ; at ( .75,-.25 ) ; ; at ( 1.25,.25 ) ; ; at ( 1.25,-.25 ) ; ; at ( .75,.25 ) @xmath515 ; \(a ) at ( 0,0 ) @xmath6 ; ( b ) at ( 2,0 ) @xmath1 ; ( c ) at ( 1,-1 ) @xmath507 edge [ - > ] node [ l , below left ] @xmath513 ( a ) edge [ - > ] node [ l , below right ] @xmath514 ( b ) ; ( e ) at ( 1,1 ) @xmath507 edge [ - > ] node [ l , above left ] @xmath513 ( a ) edge [ - > ] node [ l , above right ] @xmath514 ( b ) edge [ - > ] node [ l , right ] @xmath529 ( c ) ; at ( 1.5,0 ) ; ; at ( .65,0 ) @xmath515 ; at ( -.5,0 ) @xmath97 ; \(a ) at ( 0,0 ) @xmath6 ; ( b ) at ( 2,0 ) @xmath1 ; ( c ) at ( 1,-1 ) @xmath507 edge [ - > ] node [ l , below left ] @xmath513 ( a ) edge [ - > ] node [ l , below right ] @xmath514 ( b ) ; ( e ) at ( 1,1 ) @xmath507 edge [ - > ] node [ l , above left ] @xmath513 ( a ) edge [ - > ] node [ l , above right ] @xmath514 ( b ) edge [ ->,bend right , blue ] node [ l , right , pos=.35 ] @xmath529 ( c ) edge [ ->,bend left , red , dashed ] node [ l , left , pos=.65 ] @xmath507 ( c ) ; at ( 1.1,0.25 ) ; ; at ( .5,0.25 ) @xmath515 ; at ( 1.4,-0.25 ) ; ; at ( .9,-0.25 ) ; ; at ( 1,0 ) @xmath530 ; since both @xmath531 and @xmath532 are in the same equivalence class as the corresponding identities , they are inverses and we have constructed the required isomorphism .
@xmath533 [ span2 ] if @xmath11 is a 2-category with finite products and weak pullbacks , then @xmath500 is a compact closed bicategory .
_ as noted , hoffnung @xcite showed that @xmath534 is a monoidal tricategory .
we refer the reader to hoffnung s paper for the complete definition of a monoidal tricategory , but suffice it to say that it replaces the commuting polyhedra in the above definition of a monoidal bicategory with polyhedra that commute up to a specified 3-morphism , and then adds coherence law polytopes to govern them .
when we mod out by 3-isomorphism classes of maps of spans , these 3-morphisms become trivial , so @xmath500 is a monoidal bicategory .
we need to show that @xmath500 has a syllepsis that satisfies the two syllepsis coherence laws and the symmetry coherence law , to show the existence of zig - zag 2-isomorphisms , and to show that the swallowtail coherence law holds . because products in @xmath11 are products in the underlying category of @xmath11 , the universal property of the product means that the braiding @xmath535 in @xmath11 is its own inverse .
the syllepsis in @xmath11 is the identity , so the coherence laws making @xmath11 into a symmetric monoidal 2-category are trivial .
the braiding @xmath244 in @xmath500 comes from the braiding @xmath535 in @xmath11 : at ( -1,0.5 ) @xmath536 ; ( l ) at ( 0,0 ) @xmath537 ; ( r ) at ( 2,0 ) @xmath538 ; ( c ) at ( 1,1 ) @xmath539 edge [ - > ] node [ l , above left ] @xmath535 ( l ) edge [ - > ] node [ l , above right ] 1 ( r ) ; since @xmath535 is its own inverse , this span is also the adjoint braiding .
the syllepsis in @xmath500 is the identity map of spans , so the coherence laws making @xmath500 into a sylleptic and symmetric monoidal bicategory are all therefore trivial .
the dual of an object is the object itself ; the dual of a span swaps the two legs of the span ; and the dual of a map of spans swaps the invertible 2-morphisms .
the unit @xmath444 is the span @xmath540 where @xmath541 is the diagonal 1-morphism in @xmath11 and @xmath542 is the unique 1-morphism into the terminal object in @xmath11 .
the counit @xmath543 is the reverse of the span @xmath444 . in order to prove that the swallowtail coherence law holds
, we will show that all the 2-morphisms in the swallowtail coherence diagram are * strict isomaps of spans*that is , they are all in the same equivalence class as a map of spans of the form \(s ) at ( 1,2 ) @xmath6 ; ( y ) at ( 0,1 ) @xmath5 edge [ < - ] node [ l , above left ] @xmath12(s ) ; ( x ) at ( 2,1 ) @xmath4 edge [ < - ] node [ l , above right ] @xmath13(s ) ; ( s ) at ( 1,0 ) @xmath20 edge [ - > ] node [ l , below left ] @xmath490(y ) edge [ - > ] node [ l , below right ] @xmath491(x ) edge [ < - ] node [ l , left ] @xmath9 ( s ) ; at ( 0.5,1 ) @xmath544 ; at ( 1.5,1 ) @xmath544 ; where @xmath9 is an isomorphism .
most of the 2-morphisms in the coherence diagram are only adjoint equivalences in @xmath534 , but they become isomorphisms when we move to @xmath545 note that because @xmath11 has finite products , a span @xmath546 is the same as a morphism @xmath547 if @xmath548 is monic , then a strict isomap from the span to itself is necessarily the identity .
when @xmath548 is monic , we say the span is * jointly monic*. the swallowtail coherence law is a map from the span @xmath549 to itself ; since @xmath550 is monic , if we can show that the swallowtail is a strict isomap of spans , then it will necessarily be the identity .
the following 2-morphisms are strict isomaps of spans : 1 .
* unitors for composition . *
the composition of the span @xmath501 with the identity span on @xmath90 is isomorphic to @xmath32 by lemma [ lemma ] , where we let @xmath551 @xmath23 and @xmath91 be the identity on @xmath552 and @xmath553 the derivation for the other unitor for composition is nearly identical .
* associator for composition .
* hoffnung showed ( * ? ? ? * prop .
4.22 ) that the associator for composition is a strict isomap of spans .
* 2-morphisms in pseudonatural transformations for l , r , a. * the left unitor @xmath147 for tensor is a pseudonatural transformation .
a pseudonatural transformation has a diagram that commutes up to a 2-morphism , which in turn is subject to various coherence laws .
+ \(a ) at ( 2,0 ) @xmath90 ; ( b ) at ( 0,0 ) @xmath58 edge [ - > ] node [ l , below ] @xmath9 ( a ) ; ( c ) at ( 2,2 ) @xmath554 edge [ - > ] node [ l , right ] @xmath555 ( a ) ; ( d ) at ( 0,2 ) @xmath556 edge [ - > ] node [ l , left ] @xmath557 ( b ) edge [ - > ] node [ l , above ] @xmath558 ( c ) ; at ( 1,1 ) ; ; + for brevity we write both the product of two objects and the weak pullback of a cospan in @xmath11 with juxtaposition . taking @xmath559 the two composite spans in the square above are + \(a ) at ( 0,0 ) @xmath90 ; ( b ) at ( 2,0 ) @xmath560 ; ( c ) at ( 4,0 ) @xmath561 ; ( d ) at ( 1,1 ) @xmath560 edge [ - > ] node[l , above left]@xmath555(a ) edge [ - > ] node[l , above right]@xmath560(b ) ; ( e ) at ( 3,1 ) @xmath562 edge [ - > ] node[l , above left]@xmath563(b ) edge [ - > ] node[l , above right]@xmath564(c ) edge [ - > ] node[l , below]@xmath563(d ) ; ( f ) at ( 2,2 ) @xmath565 edge [ - > ] node[l , above left]@xmath566(d ) edge [ - > ] node[l , above right]@xmath567(e ) ; at ( 2,1.25 ) @xmath568 ; ( 1.8,1.7)(2,1.5)(2.2,1.7 ) ; + \(a ) at ( 0,0 ) @xmath90 ; ( b ) at ( 2,0 ) @xmath58 ; ( c ) at ( 4,0 ) @xmath561 ; ( d ) at ( 1,1 ) @xmath569 edge [ - > ] node[l , above left]@xmath12(a ) edge [ - > ] node[l , above right]@xmath13(b ) ; ( e ) at ( 3,1 ) @xmath561 edge [ - > ] node[l , above left]@xmath557(b ) edge [ - > ] node[l , above right]@xmath561(c ) edge [ < - ] node[l , below]@xmath570(d ) ; ( f ) at ( 2,2 ) @xmath571 edge [ - > ] node[l , above left]@xmath572(d ) edge [ - > ] node[l , above right]@xmath573(e ) ; at ( 2,1.25 ) @xmath574 ; ( 1.8,1.7)(2,1.5)(2.2,1.7 ) ; + by lemma [ lemma ] , the first span is isomorphic to + \(a ) at ( 0,0 ) @xmath90 ; ( b ) at ( 2,0 ) @xmath561 ; ( c ) at ( 1,1 ) @xmath562 edge [ - > ] node [ l , above left ] @xmath575 ( a ) edge [ - > ] node [ l , above right ] @xmath564 ( b ) ; + because the 2-morphism in the weak pullback is invertible , symmetry and lemma [ lemma ] imply that the second span is isomorphic to + \(a ) at ( 0,0 ) @xmath90 ; ( b ) at ( 2,0 ) @xmath561 ; ( c ) at ( 1,1 ) @xmath569 edge [ - > ] node [ l , above left ] @xmath12 ( a ) edge [ - > ] node [ l , above right ] @xmath570 ( b ) ; + we get a strict isomap of spans + \(a ) at ( 0,0 ) @xmath90 ; ( b ) at ( 2,0 ) @xmath561 ; ( c ) at ( 1,1 ) @xmath562 edge [ - > ] node [ l , above left ] @xmath575 ( a ) edge [ - > ] node [ l , above right ] @xmath564 ( b ) ; ( d ) at ( 1,-1 ) @xmath569 edge [ - > ] node [ l , below left ] @xmath12 ( a ) edge [ - > ] node [ l , below right ] @xmath570 ( b ) edge [ < - ] node [ l , right ] @xmath576 ( c ) ; at ( .5,0 ) ; ; at ( 1.5,0 ) ; ; + the right unitor , pentagonator , and tensorator all follow similarly ; in each case , the spans have the form required by lemma [ lemma ] because the right unitor , pentagonator and tensorator in @xmath11 are all natural transformations rather than pseudonatural transformations .
2-morphisms @xmath577 and @xmath108 * + the 2-unitor @xmath578 in @xmath500 is a map between the spans @xmath579 and @xmath580 .
the latter is the composite span + \(a ) at ( 0,0 ) @xmath398 ; ( b ) at ( 2,0 ) @xmath581 ; ( c ) at ( 4,0 ) @xmath582 ; ( d ) at ( 1,1 ) @xmath581 edge [ - > ] node[l , above left]@xmath583(a ) edge [ - > ] node[l , above right]@xmath581(b ) ; ( e ) at ( 3,1 ) @xmath582 edge [ - > ] node[l , above left]@xmath111(b ) edge [ - > ] node[l , above right]@xmath582(c ) edge [ - > ] node[l , below]@xmath111(d ) ; ( f ) at ( 2,2 ) @xmath584 edge [ - > ] node[l , above left]@xmath585(d ) edge [ - > ] node[l , above right]@xmath586(e ) ; at ( 2,1.25 ) @xmath587 ; ( 1.8,1.7)(2,1.5)(2.2,1.7 ) ; + by lemma [ lemma ] , this span is isomorphic to + \(a ) at ( 0,0 ) @xmath398 ; ( b ) at ( 2,0 ) @xmath582 ; ( c ) at ( 1,1 ) @xmath582 edge [ - > ] node [ l , above left ] @xmath588 ( a ) edge [ - > ] node [ l , above right ] @xmath582 ( b ) ; + because the 2-morphism @xmath578 in @xmath11 is trivial , we get a strict isomap of spans + \(a ) at ( 0,0 ) @xmath398 ; ( b ) at ( 2,0 ) @xmath582 ; ( c ) at ( 1,1 ) @xmath582 edge [ - > ] node [ l , above left ] @xmath588 ( a ) edge [ - > ] node [ l , above right ] @xmath582 ( b ) ; ( d ) at ( 1,-1 ) @xmath582 edge [ - > ] node [ l , below left ] @xmath589 ( a ) edge [ - > ] node [ l , below right ] @xmath582 ( b ) edge [ < - ] node [ l , right , pos=.65 ] @xmath582 ( c ) ; at ( .5,0 ) ; ; at ( 1.5,0 ) ; ; + the 2-morphisms @xmath590 all follow similarly because those 2-morphisms in @xmath11 are also identities .
* 2-morphisms @xmath429 and @xmath436 * + the 2-morphism @xmath429 is a map from the identity on @xmath58 to a morphism we will call @xmath591 first we break up @xmath592 into two parts ; by lemma [ lemma ] , @xmath593 + \(c ) at ( 4,0 ) @xmath594 ; ( d ) at ( 6,0 ) @xmath595 ; ( e ) at ( 8,0 ) @xmath561 ; ( f ) at ( 10,0 ) @xmath58 ; + \(i ) at ( 5,1 ) @xmath595 edge [ - > ] node [ l , above left ] @xmath111 ( c ) edge [ - > ] node [ l , above right ] ( d ) ; ( j ) at ( 7,1 ) @xmath596 edge [ - > ] node [ l , above left ] @xmath597 ( d ) edge [ - > ] node [ l , above right ] @xmath598 ( e ) edge [ - > ] node [ l , above ] @xmath597 ( i ) ; ( k ) at ( 9,1 ) @xmath561 edge [ - > ] node [ l , above left ] ( e ) edge [ - > ] node [ l , above right ] @xmath147 ( f ) edge [ - > ] node [ l , above ] @xmath599 ( j ) ; + is isomorphic to the following span : + \(a ) at ( 0,0 ) @xmath594 ; ( b ) at ( 2,0 ) @xmath58 ; ( c ) at ( 1,1 ) @xmath561 edge [ - > ] node [ l , above left ] @xmath600 ( a ) edge [ - > ] node [ l , above right ] @xmath147 ( b ) ; + again by lemma [ lemma ] , when we compose this isomorphic span with the rest of @xmath601 + \(a ) at ( 0,0 ) @xmath58 ; ( b ) at ( 2,0 ) @xmath602 ; ( c ) at ( 4,0 ) @xmath594 ; ( d ) at ( 6,0 ) @xmath58 ; + \(e ) at ( 1,1 ) @xmath602 edge [ - > ] node [ l , above left ] @xmath32 ( a ) edge [ - > ] node [ l , above right ] ( b ) ; ( f ) at ( 3,1 ) @xmath596 edge [ - > ] node [ l , above left ] @xmath603 ( b ) edge [ - > ] node [ l , above right , pos=.75 ] @xmath604 ( c ) edge [ - > ] node [ l , above ] @xmath603 ( e ) ; ( g ) at ( 5,1 ) @xmath561 edge [ - > ] node [ l , above left , pos=.25 ] @xmath600 ( c ) edge [ - > ] node [ l , above right ] @xmath147 ( d ) edge [ - > ] node [ l , above ] @xmath599 ( f ) ; + the result is isomorphic to the following span : + \(a ) at ( 0,0 ) @xmath58 ; ( b ) at ( 2,0 ) @xmath58 ; ( c ) at ( 1,1 ) @xmath561 edge [ - > ] node [ l , above left ] @xmath605 ( a ) edge [ - > ] node [ l , above right ] @xmath147 ( b ) ; + the 2-morphism @xmath429 is given by the left unitor in @xmath11 followed by the two isomorphisms above .
the 2-morphism @xmath436 is similar . given that all these maps of spans are strict isomaps of jointly monic spans , the swallowtail coherence law is a strict isomap of spans between the morphism @xmath444 and itself , and is therefore the identity.@xmath533
[ span ] when @xmath0 is a category with finite products and pullbacks , the bicategory span(@xmath0 ) of objects of @xmath0 , spans in @xmath606 and maps of spans is compact closed . _ proof .
_ when @xmath0 is a category with finite products and pullbacks , @xmath607 is a special case of theorem [ span2 ] where all the 2-morphisms in the pseudopullbacks are identities . @xmath533
the bicategory rel of sets , relations , and implications , is compact closed .
_ rel is a wide sub - bicategory of span(set)that is , it has all sets as objects , but only some of the 1-morphisms .
relations are the jointly - monic spans , the spans for which there is at most one element whose image under the legs of the span picks out any particular pair of related elements .
implications are maps of jointly - monic spans : a function between sets with at most one element is an implication . since all of the structure morphisms in the definition of a compact closed bicategory are jointly monic and all the 2-morphisms between those structure morphisms exist , rel is compact closed.@xmath533 when @xmath1 is a symmetric 2-rig , the bicategory mat(@xmath1 ) of natural numbers , matrices with entries in ob(@xmath1 ) , and matrices with entries in mor(@xmath1 ) is compact closed .
_ the composite of two matrices is given by the usual matrix product , except that we use the monoidal product in @xmath1 instead of multiplication and the coproduct instead of addition .
the identity matrix has the monoidal unit object in r on the diagonal and the initial object off the diagonal .
since the monoidal product in @xmath1 is in general only associative up to an associator , so is the product of matrices ; similarly , composing with the identity matrix only gives the same answer up to the unitor for the monoidal product in @xmath1 .
the pentagon and triangle equations for composition hold in mat(@xmath1 ) because they hold for the tensor product in @xmath26 mat(@xmath1 ) is symmetric monoidal under the usual tensor product for matrices , which is associative up to the associator in @xmath1 ; the monoidal unit is the object 1 up to the unitor in @xmath1 . just as with the usual notion of matrices , the tensor product of two objects
@xmath608 is @xmath609 .
the pentagonator and unitors are the identity in mat(@xmath1 ) because they are elementwise application of the associator and unitors in @xmath1 , so the associahedra and 2-unitors are trivial .
the symmetry involves a permutation of the rows and columns of the matrix that is its own inverse followed by the symmetry in @xmath1 . since @xmath1 is a symmetric monoidal category , braiding twice in mat(@xmath1 ) gives the identity , which in turn implies that the syllepsis and the coherence laws governing it are trivial . because the associahedra and unitors are trivial , the hexagon braiding modifications are identities and the coherence laws governing the braiding are trivial . the dual in mat(@xmath1 ) is the identity on objects ; on matrices , it is a transposition of the matrix followed by the symmetry in @xmath26 the zig - zag morphism [ cols=">,^ , < " , ] is the same as the identity up to the unitor in @xmath26 the swallowtail coherence law is therefore an inverse unitor , an associator , and a unitor , _ i.e. _ the triangle equation in @xmath1 for the pair of objects @xmath610@xmath533 the bicategory cocont is compact closed .
_ day and street @xcite proved that cocont is compact closed ; here we give a more plebeian proof .
cocont is a 2-category , and every 2-category is a bicategory whose associator and unitors for composition are identities .
recall that * the objects are small categories , * morphisms @xmath611 are cocontinuous functors @xmath612 and * 2-morphisms are natural transformations between functors .
this is a categorification of the compact closed category finvect , where * objects are finite sets and * morphisms are linear transformations between the free vector spaces on the sets .
small categories have replaced finite sets , and cocontinuity has replaced linearity .
cocont is a monoidal bicategory : * @xmath613 where @xmath614 is the cartesian product of categories in cat .
* @xmath143 is the terminal category . *
@xmath615 where @xmath616 is the associator in cat .
* @xmath617 where @xmath618 and @xmath619 are the unitors in cat .
* @xmath108 is an equality because the associator for the cartesian product in cat satisfies the pentagon equation .
* @xmath620 are equalities because the triangle equations for the left and right unitors in cat are equations . * the coherence diagrams
the 14-vertex associahedron and the two 9-vertex diagrams involving the unitors are trivial because @xmath621 are equalities rather than merely isomorphisms .
next , we need to check the braiding .
cocont is a braided monoidal bicategory where * @xmath622 where @xmath623 is the braiding for the cartesian product in cat . *
@xmath1 and @xmath6 are equalities because the braiding in cat satisfies the hexagon equations .
* the polytopes involving @xmath1 and @xmath6 are trivial because @xmath1 and @xmath6 are equalities .
third , we need to check the syllepsis .
* because the braiding in cat is symmetric , cocont s braiding is also symmetric ; * the syllepsis @xmath624 is an equality , and * the polytopes involving @xmath624 are trivial . finally ,
to show it is compact closed , we need to check the existence of duals and to verify the swallowtail coherence law . *
every small category @xmath58 has the dual @xmath625 . *
the counit is @xmath626 where @xmath627 is the coend of @xmath569that is , the coequalizer of the diagram @xmath628 * the unit is @xmath629 where @xmath630 * the 2-morphism @xmath429 is a natural isomorphism @xmath631 it is canonical in the sense that we do not need to make any arbitrary choices to construct it . an elegant proof that it is canonical is the following one , shown to the author by tom leinster .
@xmath632 the first isomorphism trades a coend for an end , which is possible because the coend was in the covariant position ; the second isomorphism is `` currying , '' which is possible because set is cartesian closed .
the third is perhaps the most important ; just as the coend quotients by a relation , the end picks out those elements that already satisfy a relation . in this case , the relation is precisely the commuting square in the definition of a natural transformation between the given functors . finally , the last isomorphism is the ordinary yoneda lemma .
the proof for the 2-morphism @xmath436 is almost identical . @xmath533
the bicategories cospan(resnet ) and circ are compact closed .
_ the coproduct of two resistor networks is given by juxtaposition ; the pushout of a cospan @xmath633 of resistor networks is given by juxtaposition followed by identifying the images of @xmath1 in @xmath6 and @xmath11 .
cospans in resnet are spans in resnet@xmath634 where the coproduct and pushout become product and pullback , so cospan(resnet ) is compact closed as a corollary of theorem [ span ] . since every object is self - dual in cospan(resnet ) , the subcategory circ whose objects are resistor networks with no edges is also compact closed .
thanks to the anonymous referee for an extraordinarily careful reading ; to john baez and mike shulman for explanations of concepts and helpful comments ; to alex hoffnung for his patience in explaining weak pullbacks ; and to the members of the n - category caf and mathoverflow for expository background material and references .
k. asada , arrows are strong monads , _ msfp 10 proceedings of the third acm sigplan workshop on mathematically structured functional programming . _ ( 2002 ) , 3342 . also available at http://www.ipl.t.u-tokyo.ac.jp/~asada/papers/arrstrmnd.pdf[www.ipl.t.u-tokyo.ac.jp/@xmath635asada/papers/arrstrmnd.pdf ] . j. baez and j. dolan , higher - dimensional algebra and topological quantum field theory ,
_ j. math .
phys . _ * 36 * ( 1995 ) , 60736105 .
also available at http://arxiv.org/abs/q-alg/9503002v2[arxiv:q-alg/9503002v2 ] .
j. baez and j. dolan , higher - dimensional algebra iii : n - categories and the algebra of opetopes , _ adv .
_ * 135 * ( 1998 ) , 145206 . also available at http://arxiv.org/abs/q-alg/9702014[arxiv:q-alg/9702014 ] .
a. carboni , g. m. kelly , r. f. c. walters , and r. j. wood , cartesian bicategories ii , _ theory and applications of categories _ , * 19:6 * ( 2008 ) 93124 .
available at http://www.emis.de/journals/tac/volumes/19/6/19-06abs.html .
n. gurski , an algebraic theory of tricategories , phd thesis , university of chicago , ( 2007 ) .
available at http://gauss.math.yale.edu/~mg622/tricats.pdf[http://gauss.math.yale.edu/@xmath635mg622/tricats.pdf ] j. hughes , generalising monads to arrows , _ science of computer programming _ * 37 * ( may 2000 ) 67111 .
preprint available at + http://www.cse.chalmers.se/~rjmh/papers/arrows.pdf[http://www.cse.chalmers.se/@xmath635rjmh/papers/arrows.pdf ] .
b. jacobs , c. heunen and i. hasuo , categorical semantics for arrows , _ journal of functional programming _ * 19:3 - 4 * ( 2009 ) 403438 .
the results appeared in an earlier conference paper c. heunen , b. jacobs , arrows , like monads , are monoids , _ entcs _ * 158 * ( 2006 ) , 219236 .
p. katis , r.f.c .
walters , the compact closed bicategory of left adjoints , _ math .
proc . of the cambridge philosophical society _ * 130:1 * ( 2001 ) 7787 .
also available at http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.23.4825 .
g. m. kelly , basic concepts of enriched category theory , _ lecture notes in mathematics _ , * 64 * ( 1982 ) .
republished in _ reprints in theory and applications of categories _ , * 10 * ( 2005 ) , 1136 .
available at + http://www.tac.mta.ca/tac/reprints/articles/10/tr10abs.html . c. schommer - pries , the classification of two - dimensional extended topological field theories , phd thesis , uc berkeley ( 2003 ) .
available at + http://sites.google.com/site/chrisschommerpriesmath/home/schommer-pries-thesis.pdf .
m. shulman , re : inevitability in mathematics , _
@xmath2-category caf _
available at https://golem.ph.utexas.edu/category/2010/06/inevitability_in_mathematics.html#c033741[https://golem.ph.utexas.edu/category/2010/06/ ] + https://golem.ph.utexas.edu/category/2010/06/inevitability_in_mathematics.html#c033741[inevitability_in_mathematics.html#c033741 ] r. street , functorial calculus in monoidal bicategories , _ applied categorical structures _ * 11 * ( 2003 ) 219227 . also available at + http://www.maths.mq.edu.au/~street/extraord.pdf[http://www.maths.mq.edu.au/@xmath635street/extraord.pdf ] .
r. street , frobenius monads and pseudomonoids , _ j. math . physics _
* 45 ( 10 ) * ( october 2004 ) 39303948 .
also available at http://www.maths.mq.edu.au/~street/frob.pdf[http://www.maths.mq.edu.au/@xmath635street/frob.pdf ] . | a compact closed bicategory is a symmetric monoidal bicategory where every object is equipped with a weak dual .
the unit and counit satisfy the usual `` zig - zag '' identities of a compact closed category only up to natural isomorphism , and the isomorphism is subject to a coherence law .
we give several examples of compact closed bicategories , then review previous work . in particular , day and street defined compact closed bicategories indirectly via gray monoids and then appealed to a coherence theorem to extend the concept to bicategories ; we restate the definition directly , emphasizing the combinatorics . finally , we prove that each of the examples are compact closed .
we prove that given a 2-category @xmath0 with finite products and weak pullbacks , the bicategory of objects of @xmath0 , spans , and isomorphism classes of maps of spans is compact closed .
we also prove that given a cocomplete symmetric monoidal category @xmath1 whose tensor product distributes over its colimits , the bicategory mat(@xmath1 ) of natural numbers , matrices of objects of @xmath1 and matrices of morphisms of @xmath1 is compact closed . as corollaries , the bicategory of spans of sets , the bicategory of relations , and certain bicategories of `` resistor networks '' are all compact closed .
we also give a new proof that the 2-category of small categories , cocontinuous functors between the presheaves on those categories , and natural transformations is compact closed . |
modifications of the internal states of an atom due to its coupling to the electromagnetic ( em ) field are , under standard conditions , highly restricted .
once the proper selection rules are satisfied , the enhancement or inhibition of transition rates can be based on the control of the evolution of the atomic center of mass , and the appropriate selection of the features of the surrounding electromagnetic field .
the latter includes its frequency spectrum , angular spectrum and polarization . in this work ,
we make a systematic study of the feasibility of modifying atomic transition rates by tailoring the electromagnetic field , given different schemes for the center of mass dynamics . in general , previous analysis that had the enhancement or inhibition of internal state transition rates in mind ,
can be classified as those which study atoms confined in electromagnetic cavities @xcite or nearby surfaces @xcite , and studies in which laser light is shaped in open space to create electromagnetic fields with a predetermined structure @xcite .
the first scenario , in general , gives rise to polarization restrictions : the natural modes would be , _
e. g. _ , transverse electric or transverse magnetic em fields satisfying the adequate boundary conditions on the cavity surface . in the last scenario ,
most descriptions usually assume circular or linear polarization with respect to the main propagation axis of a given laser beam . here
we present a general analysis that includes both cases .
we take into account the relevance of the central field description of electrons in atoms , and the fact that em fields can be described as superpositions of modes with well defined total ( orbital plus polarization ) angular momentum .
the accompanying symmetries of both atom and em field are essential to get a proper description of the transition rates under realistic conditions . in all the calculations ,
the role of the atomic center of mass will be emphasized . in section
ii , a general accurate description of the so called scalar and vectorial em modes via their wave vector spectra is presented ; using it guarantees the validity of our forthcoming analysis beyond the limitations of the usual paraxial approximation . in section iii ,
the transition matrix elements of the minimal coupling interaction hamiltonian of a nonrelativistic hydrogen like atom are written in terms of products of a center of mass factor , an internal atomic state factor and an electromagnetic factor which couples the former two terms .
based on this , a modular description of transition rates involving all the degrees of freedom of the system atom - em field is developed . in section iv
, we revisit the general properties of the internal transition factor and the standard classification in terms of the coupling of different atomic multipoles to the em field .
we also show that the developed formalism can be used to study internal transitions which require a relativistic description without modifications on the center of mass and electromagnetic factors . in section
v we stress the main features of the center of mass term when it is described by a quantum wave function .
the limitations of such an approach under realistic conditions are also briefly discussed . in order to compare with standard studies of atomic recoil effects , in section
vi , we apply our formalism when the center of mass states is described in terms of plane waves . in section vii
, we report the explicit calculation of the electromagnetic factor for spherical modes , gaussian modes and propagation invariant modes for a center of mass motion with spherical symmetry .
this could correspond , for instance , to atoms in free space or in a spherical trap . in section viii , an optimization procedure to enhance transition rates by selecting the proper parameters of a laguerre - gaussian beam or of a bessel beam
is explicitly worked out ; the trap that confines the cold atoms is assumed to have circular cylinder symmetry . finally , some conclusions derived from this study are given in section ix .
the electric field @xmath0 of an electromagnetic wave in otherwise free space can always be written as a superposition of plane waves with wave vector @xmath1 , frequency @xmath2 and polarization @xmath3 : @xmath4 in this equation @xmath5 denotes the frequency spectra , the delta function guarantees the fulfillment of the dispersion relation , @xmath6 , and @xmath7 is the vectorial spectrum in @xmath1 space for the polarization @xmath3 ; it is characterized by a set of parameters @xmath8 that are required to specify each electromagnetic mode of frequency @xmath2 .
the magnetic field associated to the em wave is given by @xmath9 with @xmath10 in the coulomb gauge , the electromagnetic vector potential for each mode is @xmath11 the polarization @xmath3 reflects the transverse structure of the electric field @xmath12 , so that , @xmath13 one selection of the resulting two independent polarizations is defined in terms of a preselected vector @xmath14 as follows ( from now on , @xmath15 denotes the unitary operator along @xmath16 ) : @xmath17 \tilde{\psi}^{(te)}_{\kappa;\omega}(k_x , k_y),\label{eq : bmodee}\\ { \mathfrak{e}}^{tm}_{\kappa}(k_x , k_y;\omega)&= & \mathcal{e}_\kappa^{tm}[\hat{\mathbf{k}}(\hat{\mathbf{k } } \cdot { \mathbf{a}})- { \mathbf{a}}]\tilde{\psi}^{(tm)}_{\kappa;\omega}(k_x , k_y).\label{eq : bmodem}\end{aligned}\ ] ] these modes are the well known transverse electric ( @xmath18 ) and transverse magnetic ( @xmath19 ) modes frequently used in cavity electrodynamics ( although they can also be generated in free space @xcite ) .
the vector @xmath20 in such systems is usually chosen as an axis of symmetry of the cavity .
the normalized functions @xmath21 are known in the literature as the angular spectra of the mode , the factor @xmath22 is proportional to the mode amplitude .
specific examples of @xmath21 are given below .
if @xmath20 is parallel to the @xmath23 axis , the beams are said to be vectorial em structured beams with a privileged propagation axis along the @xmath23-axis ; with this selection @xmath24 and @xmath25 here @xmath26 is the azimuthal angle in the wave vector space , @xmath27 .
another selection of polarization basis originates from the paraxial description of light beams . in this case , the so called scalar structured beams are described as follows .
first , a fixed direction is selected @xmath28 , and two other orthogonal directions are identified @xmath29 and @xmath30
. then , @xmath31\tilde{\psi}^{(1)}_{\kappa;\omega}(k_1 , k_2 ) \equiv \mathcal{e}_\kappa^{1}\mathbf{e}^{(1)}\tilde{\psi}^{(1)}_{\kappa;\omega}(k_1 , k_2),\label{eq : smode1}\\ { \mathfrak{e}}^{2}_{\kappa;\omega}(k_1 , k_2;\omega)&= & \mathcal{e}_\kappa^{2}[\hat{\mathbf{a}}_2 - \frac{\mathbf{k}\cdot \hat{\mathbf{a}}_2}{\mathbf{k}\cdot \hat{\mathbf{a}}_3}\hat{\mathbf{a}}_3]\tilde{\psi}^{(2)}_{\kappa;\omega}(k_1 , k_2)\equiv\mathcal{e}_\kappa^{2}\mathbf{e}^{(2)}\tilde{\psi}^{(2)}_{\kappa;\omega}(k_1 , k_2),\label{eq : smode2}\end{aligned}\ ] ] in the paraxial limit @xmath28 usually corresponds to the main direction of propagation of the beam . to be specific ,
let it be the @xmath23-axis , @xmath32 @xmath33 , @xmath34 . the paraxial condition @xmath35 yields @xmath36 , @xmath371 , 2 , so that the component of the field along @xmath28 is usually neglected ; if @xmath38 and @xmath39 , @xmath40 and @xmath41 define the two orthogonal linear polarization directions .
scalar em modes may , under adequate circumstances , be approximately generated using a monochromatic laser beam with the proper polarization that impinges a space light modulator or a structured diffraction plate . to that end , each optical device is respectively programmed or designed to guarantee the appropriate boundary condition that defines each mode at the plane @xmath42 : @xmath43 in eq .
( [ eq : circ ] ) , @xmath44 denotes the interior surface of a circumference of radius @xmath45 in the @xmath46 space .
vectorial em modes can be written ( an generated ) as superpositions of scalar em modes and viceversa by the selection of the proper angular spectra ; taking @xmath47 , @xmath48 -\quad\hat{k}_1[\hat{\mathbf{a}}_2 - ( \hat{k}_2/\hat{k}_3)\hat{\mathbf{a}}_3 ] = \quad\hat{k}_2\mathbf{e}^{(1 ) } - \quad\hat{k}_1\mathbf{e}^{(2)},\nonumber\\ \mathbf{e}^{(tm)}=\hat{\mathbf{k}}(\hat{\mathbf{a}_3 } \cdot \hat{\mathbf{k}})- \hat{\mathbf{a}}_3&= & \hat{k}_1\hat{k}_3[\hat{\mathbf{a}}_1 - ( \hat{k}_1/\hat{k}_3)\hat{\mathbf{a}}_3]+ \hat{k}_2\hat{k}_3[\hat{\mathbf{a}}_2 - ( \hat{k}_2/\hat{k}_3)\hat{\mathbf{a}}_3 ] = \hat{k}_1\hat{k}_3\mathbf{e}^{(1)}+ \hat{k}_2\hat{k}_3 \mathbf{e}^{(2)}. \end{aligned}\ ] ] notice that the superpositions depend on @xmath1 .
quantization of structured modes in terms of vectorial beams @xcite is usually more direct than with scalar modes due to the evident orthogonality between @xmath18 and @xmath19 modes , in contrast to the overlap between the scalar modes defined by eqs .
( [ eq : smode1]-[eq : smode2 ] ) .
the purpose of this work is to present a general formalism to compare , in a quantitative way , the dependence of atomic transition amplitudes for different selections of polarization and angular spectra of a coherent light source . for the selection @xmath49
, explicit results will be given for the following angular spectra : * gauss modes with waists @xmath50 and @xmath51 : @xmath52 * laguerre gaussian modes with waist @xmath53 , topological charge @xmath54 and order @xmath55 : @xmath56 * propagation invariant beams with analytical expressions .
these beams , under idealized conditions , maintain its transverse structure with respect to a given direction that here we take as @xmath57 .
they have the angular spectra structure @xmath58 explicitly : * * in circular cylinder coordinates , they are known as bessel beams @xcite : @xmath59 * * in elliptic coordinates with an ellipticity parameter @xmath60 and order @xmath54 , they are known as mathieu beams @xcite : @xmath61 where @xmath62 and @xmath63 are mathieu functions of order @xmath54 in standard notation @xcite . * * in parabolic coordinates , they are known as weber beams @xcite . for even parity and aperture @xmath64 : @xmath65 odd parity modes are obtained by replacing @xmath66 in the denominator by @xmath67 .
each of these elementary modes can be combined to yield , for instance , even standing transverse bessel modes @xmath68 or traveling weber modes by superposing even and odd modes with a @xmath69 phase factor .
the experimental generation of approximate propagation invariant beams can be implemented by replacing the delta factor by a gaussian function of @xmath70 , centered around @xmath71 , and with a width much smaller than @xmath71 .
the calculation of the normalization factors @xmath72 is discussed in appendix a. their selection is done to allow an equitable comparison between the effects of different modes in the atomic transition rates .
we shall also study the results obtained for vectorial spherical waves . in this case , the vector @xmath14 that defines the polarizations is chosen as the gradient operator in @xmath1 space , @xmath73 and it is applied to the spherical harmonics in that space . spherical electromagnetic waves are described in detail in appendix b.
for simplicity we consider a hydrogen like atom in conditions that do not require a relativistic treatment . in the next section ,
we describe how these restrictions can be relaxed to include multi electron atoms and relativistic effects .
we are interested in the evaluation of the transition probability by which an ultra cold atom , initially described by a wave function that can be written as a superposition of functions that are products of a center of mass function @xmath75 and an internal function @xmath76 with quantum numbers @xmath77 and @xmath78 respectively , @xmath79 changes its state to a wave function with the same structure , that is a superposition of wave functions of the form @xmath80 and simultaneously absorbs a photon described by a mode of the em field @xmath81 , with polarization @xmath3 and structure labeled by @xmath8 . in the former expressions the position of the electron
is given by the coordinate @xmath82 where @xmath83 refers to the atom s center of mass , @xmath84 is the electron mass , @xmath85 is the reduced mass , @xmath86 is the nucleus mass , and @xmath87 is the relative coordinate .
@xmath88 and @xmath89 are eigenfunctions of the corresponding schrdinger equation with initial eigenvalues @xmath90 and @xmath91 , and final eigenvalues @xmath92 and @xmath93 .
we focus mostly in the case that the transition is induced by the interaction hamiltonian @xmath94 with @xmath95 , the electron charge and @xmath96 the relative momentum .
this hamiltonian can be understood as a result of the coupling of the electron current @xmath97 to the electromagnetic field @xmath98 in the approximation where @xmath99 valid whenever @xmath100 ( which is satisfied under standard conditions ) .
corrections to this approximation are described below .
if @xmath101 satisfies coulomb gauge ( @xmath102 ) within first order perturbation theory , the transition amplitude @xmath103 between the elementary wave functions @xmath104 and @xmath105 is given by @xmath106\nonumber \\
& \cdot & \big[q_e\frac{\mu}{m_e}{\bf a}_{\bf \kappa}^{(b)}({\bf r}+\frac{\mu}{m_e}{\bf r}_{rel } ) \big],\label{eq : trans}\end{aligned}\ ] ] in the length form . taking into account eq .
( [ eq : a ] ) , @xmath107 the structure of eqs .
( [ eq : bmodee]-[eq : bmodem ] ) and eqs .
( [ eq : smode1]-[eq : smode2 ] ) let us write the absorption transition amplitudes @xmath103 in terms of the light angular spectrum @xmath108 as follows @xmath109 the corresponding emission amplitudes induced by the presence of the coherent em field @xmath110 satisfy the relation @xmath111 .
( [ eq : para ] ) decomposes the transition amplitude @xmath112 in factors that involve separately the internal degrees of freedom of the atom @xmath113 , and the degrees of freedom of the center of mass @xmath114 .
these factors couple , in a modular form , to a coherent em field via its angular spectra @xmath115 , and its polarization @xmath116 ( which in general depends on @xmath1 in a way determined by @xmath8 and @xmath3 ) .
the non relativistic hamiltonian of two charged particles ( the nucleus and the electron ) interacting with the light electromagnetic field via the minimal coupling scheme , involves , besides the hamiltonian @xmath117 , two terms .
one of them , for a monochromatic wave at resonance @xmath118 , gives rise to the transition amplitude @xmath119 \big[q_e{\bf a}_{\bf
\kappa}^{(i)*}({\bf r}+\frac{\mu}{m_e}{\bf r}_{rel } ) + q_n{\bf a}^{(i)*}_{\bf \kappa}({\bf r}-\frac{\mu}{m_n}{\bf r}_{rel})\big]\nonumber \\ & -&i\frac{q_n\mu}{m_n}\omega_{rel}\int d^3r_{rel}d^3r \big[\psi^*_f({\bf r}_{rel},{\bf r}){\bf r}_{rel}\psi^*_0({\bf r}_{rel},{\bf r})\big ] { \bf a}_{\bf \kappa}^{(i)*}({\bf r}-\frac{\mu}{m_n}{\bf r}_{rel}).\label{eq : hi2}\end{aligned}\ ] ] with @xmath120 the nucleus charge and @xmath121 its position . the first term , in the last equation , results from the difference between the electron momentum and the momentum of the relative coordinate in the electron current - em field coupling , the second and the third represent the coupling between the nucleus current @xmath122 to the electromagnetic field @xmath123 . for atoms trapped in a harmonic potential with characteristic frequency @xmath124 , the center of mass motion can be naturally measured in terms of @xmath125 while the transition frequencies are integer multiples of @xmath126 , thus @xmath127 which corresponds to @xmath128 for a harmonic trap with @xmath129 and a hydrogen atom . for the internal motion @xmath130 and @xmath131 , with @xmath132 the fine structure constant
; so that @xmath133 .
thus , @xmath134 that is , the natural units to measure the matrix element @xmath135 that defines the transition amplitude @xmath136 , are much bigger than those related to @xmath137 since , under standard conditions @xmath138 and @xmath139 , the order of magnitude of the relevant transition rates induced by any term in @xmath140 is expected to be much smaller than its equivalent one induced by @xmath117
. note , however , that transition amplitudes of @xmath140 , eq . ( [ eq : hi2 ] ) , can also be written in a modular form similar to eq .
( [ eq : para ] ) that involves the vector integral @xmath141 and the scalar integrals @xmath142 for the first two summands in eq .
( [ eq : hi2 ] ) , and the scalar integral @xmath114 , eq .
( [ eq : icm ] ) , and the vector integral @xmath143 for the third summand in eq .
( [ eq : hi2 ] ) . so that the main features of the modular evaluation of the transition amplitudes developed in the next pages can also be applied to @xmath144 .
the third term @xmath145 in the minimal coupling hamiltonian @xmath146 is usually interpreted in terms of two photon processes : introducing a complete set of intermediate center of mass and relative coordinate states the corresponding transition amplitudes are @xmath147 .
these transition amplitudes are proportional to @xmath148 and , thus , generally smaller than the relevant ones arising from the @xmath117 hamiltonian .
note that the @xmath145 matrix elements can also be written in a modular form involving the scalar integrals @xmath149 and @xmath114 defined above .
it is important to mention that working with the minimal coupling scheme in the coulomb gauge gives equivalent results @xcite to those obtained in the pzw formalism @xcite .
this formalism is obtained by applying a gauge transformation to a minimal coupling lagrangian .
it makes evident the relevance of the magnetic and electric dipole moments of the neutral atom evaluated with respect to the atomic center of mass ; the electric and the magnetic fields of the photon field appear directly in the resulting interaction hamiltonian which also includes a term quadratic in the em field .
detailed studies of transition probabilities of atomic systems have been performed using the linear electric field part of the interaction hamiltonian of the pzw formalism for a laguerre - gaussian beam in the paraxial approximation @xcite .
notice that the matrix element @xmath150 , eq .
( [ eq : jrel ] ) , contains the information of the final and initial internal states of the atom and corresponds to that worked out in standard atomic physics calculations . from it the matrix element @xmath151 , eq .
( [ eq : jrel2 ] ) , can be calculated . in this section
we revisit key features of these matrix elements that are essential to follow the forthcoming calculations .
whenever the internal state of an atom is properly described by a central field interaction , that is , if @xmath152 ( with @xmath153 the @xmath154 , @xmath54 spherical harmonic ) the internal state quantum numbers are @xmath155 , @xmath154 , @xmath54 , and a relevant relation that makes easier the evaluation of the relative coordinate integrals @xmath113 is @xmath156 which expresses plane waves in terms of spherical waves , in this equation @xmath157 is the spherical bessel function of order @xmath158 . each term in this series is an eigenfunction of the orbital angular momentum operators , @xmath159 and @xmath160 . if @xmath161 were taken as an angular spectra and the proper polarization vectors were used , vectorial spherical electromagnetic modes which carry orbital angular momentum would be obtained as described in appendix b. using eq .
( [ eq : sph ] ) , it results @xmath162 with @xmath163 , @xmath164 the polarization vectors defined in eqs .
( [ eq : pol ] ) , @xmath165 and the angular integral @xmath166 the dependence of @xmath167 on @xmath1 is through its modulus @xmath45 and its orientation @xmath161 as a manifestation of the wigner - eckart theorem .
particular useful forms of @xmath168 correspond to linear superpositions of slater type orbitals ( sto ) @xmath169 . for them ,
the relevant integral can be directly computed @xmath170 @xmath171 for a hydrogenic atom @xmath172 .
since the energy associated to changes in the center of mass state of an atom involve in general energies much less than those of electronic states , close to resonance @xmath173 and the term @xmath174 note that , unless @xmath175 is not too small , @xmath32 @xmath33 , for atoms requiring a relativistic description of their internal state , the series that define @xmath176 can be cut at the zeroth order term @xmath177 .
that is , under non relativistic conditions , retardation effects of the atom - light interaction can be neglected , and the internal atom transitions in terms of static electric and magnetic multipoles is valid . besides , for @xmath178 , @xmath179 rapidly decreases as @xmath158 increases : the highest possible values of @xmath180 can be found for @xmath181 , that is , electric dipole transitions induced by the hamiltonian @xmath117 . for such transitions , the companion spherical harmonic @xmath182 does not depend on @xmath183 .
the next order term in the series of @xmath167 depends on @xmath183 through @xmath184 .
if the quantization axis of the atom coincides with the @xmath23-axis , @xmath185 this term can be related to electric quadrupole transitions and magnetic dipole transitions induced by the hamiltonian @xmath117 .
however the latter would not be properly described unless electron spin effects had been incorporated .
the angular structure of @xmath186 given by @xmath187 reflects the vectorial character of the electromagnetic field through the presence of the @xmath188 factor , and the relevance of angular structure of that field through the orbital angular momentum term @xmath189 .
both of them can be coupled to yield an effective total angular momentum @xmath190 a similar expression can be used for @xmath191 so that the angular integrals @xmath187 have a closed expression in terms of the 3-@xmath192 symbols .
the selection rules for each @xmath158 multipole result from the structure of 3-@xmath192 symbols . for the evaluation of the transition amplitudes related to the hamiltonians @xmath140 , eq.([eq : hi2 ] ) , and @xmath145 , eq .
( [ eq : hi3 ] ) , it is necessary to calculate the internal scalar transition factors @xmath193 , eq .
( [ eq : jrel1 ] ) . using the series expansion given by eq .
( [ eq : sph ] ) , these factors can also be written in a completely analogous form to that of the vector factors @xmath194 .
now the functions @xmath195 and @xmath196 are the relevant ones .
whenever retardation effects can be neglected the standard condition @xmath197 makes that the term @xmath181 in the series expansion of the scalar factor @xmath149 gives a negligible value , and the terms with lowest @xmath158 value that may give a significant contribution correspond to @xmath198 , @xmath199 instead of the @xmath181 term of the electric dipole transitions induced by @xmath117 .
that is , the scalar integral @xmath200 term associated to dipole transitions has the structure @xmath201 .
a similar argument follows for higher multipoles . for multiple electrons atoms ,
most of the equations given above are still valid both in the context of the independent particle model with a central field approximation , and for treatments that yield natural orbitals @xcite after including electronic correlation effects .
a relativistic description of the internal state evolution of a hydrogen like atom can be done in terms of the solution of the dirac equation .
then the effective interaction hamiltonian for the electron - em field coupling could be taken as @xmath202 , whenever the cm motion is not relativistic , with @xmath203 the dirac matrices and @xmath204 the electromagnetic potential evaluated at the position of the electron undergoing the atomic internal transition , eq .
( [ eq : re ] ) .
the plane wave exponential would then be written as a product of an exponential term dependent on the cm coordinate , and the exponential term dependent on the relative coordinate , in complete analogy to eq .
( [ eq : para ] ) .
the explicit calculation of the transition amplitudes regarding the internal motion can be found in ref .
the main advantage of this treatment is that the effects of the electron spin are taken into account both directly and properly .
such a relativistic treatment does not change the scalar factors which couple atomic internal transitions to the electromagnetic field ; they are proportional to @xmath205 .
an important modification is found in the polarization factor that in the non relativistic case is encoded in @xmath206 , and that now , distinguishes electric and magnetic effects through the presence of either electric or magnetic spherical polarization vectors @xcite .
since both the internal and center of mass states are simultaneously coupled to the electromagnetic field , controlling transitions between internal atomic states requires the proper regulation of center of mass transitions . during the last years
, there has been impressive developments on laser cooling and trapping techniques , as well as an acquired capability to implement electric and magnetic fields on design .
this has lead to the possibility of creating atomic samples in which each atom can be properly described by a wave function of precisely the form @xmath208 in particular , this applies to some experiments of few ions in paul and penning traps@xcite and , very recently , to neutral atoms magnetically trapped in electromagnetic cavities @xcite . in the case of ultra cold atomic gases in the degenerate regime , and for the condensate fraction , eq .
( [ eq : super ] ) can be used with @xmath209 determined , e. g. , by the gross - pitaevski equation for boson samples . in most other cases ,
environmental effects can not be ignored and a description in terms of a density matrix for the center of mass state would be the most realistic option . the main reason behind this fact is the small value of the energy difference between the center of mass states of an isolated atom , compared to the interaction energy associated to collisions with other atoms and to the interaction with the background electromagnetic field .
the implementation of semiclassical approaches where the center of mass motion is described in terms of newton equations , while the internal dipole transitions are dealt within a quantum approach has probed to be adequate for the description of recoil effects in most standard situations @xcite .
an alternative could be to take , as starting point , an idealized configuration in which each atom behaves as in empty free space with the center of mass stationary states taken as plane wave ( pw ) functions . then the center of mass integral @xmath207 , eq .
( [ eq : icm ] ) , can be evaluated @xmath210 the corresponding transition amplitudes , eq .
( [ eq : tof ] ) ( which is expected to be dominant ) , eq .
( [ eq : hi2 ] ) and eq .
( [ eq : hi3 ] ) would incorporate elementary recoil effects .
the resulting transition rates should then be introduced in bloch equations where other properties like collision and temperature effects could also be taken into account .
such a calculation is beyond the scope of this work .
nevertheless , in the following section , the consequences of using the center of mass factors @xmath211 in the evaluation of the transition amplitudes is discussed for structured modes . since
ultra cold atoms are confined into a given region of space , in many cases , a reasonable description of the center of mass states is given by harmonic oscillator states centered at a position @xmath212 .
then , the center of mass wave function has an structure @xmath213 and the transition integral @xmath207 , eq .
( [ eq : icm ] ) , takes the form @xmath214 the exponential factor @xmath215 can also be written in terms of the wave vector @xmath1 and the position of the center of the trap @xmath216 using again eq .
( [ eq : sph ] ) .
notice however that for the center of mass coordinate , @xmath217 may be greater than one . for the particular case of an atom confined in a potential
that has either spherical symmetry or cylindrical symmetry around the @xmath23-axis the integral in eq.([eq : icm ] ) can be directly evaluated ( see appendix c ) . in the first case
the dependence on @xmath183 is through the spherical harmonics : @xmath218 finally , if the center of was motion is described in terms of free spherical waves : @xmath219 with @xmath220 the de broglie wavelength of the center of mass .
the vector integrals @xmath221 , eq .
( [ eq : icm2 ] ) , in both cases can be directly written in terms of vectorial spherical harmonics ( appendix b ) .
if the center of mass factor @xmath114 is given by eq .
( [ eq : cmpw ] ) , then @xmath222 for electric dipole ( ed ) transitions , @xmath223 which is directly proportional to the angular spectrum as could be expected .
this equation is a quantum analog in momentum space to the standard dipole interaction term @xmath224 used in most semiclassical descriptions of ed recoil effects . from this equation
we observe that recoil effects are determined directly by the polarization and angular spectrum of the mode under consideration .
this result has direct physical consequences , @xmath33 @xmath225 , a bessel mode @xcite or a laguerre - gauss mode @xcite can induce center of mass rotations around the main propagation axis of the beam , while a weber mode has probed to be adequate for inducing a center of mass motion with parabolic symmetry on cold atomic samples @xcite . for electric quadrupole ( eq ) and magnetic dipole ( md ) transitions the transition amplitude has the structure @xmath226 where the eq or md character of the transition is encoded in the factor @xmath227 which depends on both the mode polarization coupling parameter @xmath228 and the effective angular momentum factor @xmath229 .
in this case , the effective interaction term yielding recoil effects depends on the polarization and products of the angular spectrum of the mode with first order terms of @xmath1 . in configuration space
, these products correspond to first order derivatives of the electric and magnetic fields .
due to its frequent use in actual experiments , we discuss in more detail axial symmetric scalar gaussian modes of waist @xmath230 , @xmath231 for scalar circularly polarized modes , the exact expression of the normalization integral is evaluated in appendix a , @xmath232\big]^{-1},\quad \quad a = \frac{\omega \mathrm{w}_\bot}{c},\label{eq : ngauss}\ ] ] with erfi the imaginary error function @xmath233 the factor @xmath234 is a natural scale of the energy of the mode that could be used for the quantization of the em field . for our study , it just provides a reference useful for comparing the effects of different structured modes with the same em energy on atomic transitions rates . for gaussian modes , different choices of the waist
@xmath230 modify the relevance of the @xmath235 factor , eq.([eq : y1 m ] ) ; that is , for paraxial modes @xmath236
so that the internal transition rates which would be dominant are those with @xmath237 . in the antiparaxial regime @xmath238 , that is @xmath239 , @xmath240 and the internal transition rates which would be dominant are those with @xmath241 .
notice however , that for a gaussian mode the transverse momenta distribution is always centered around @xmath242 discarding them as the optimal choice for achieving the antiparaxial regime .
a better option corresponds to using modes with a @xmath46 distribution that becomes maximum at a @xmath70 value different from zero , like the propagation invariant modes mentioned at section ii .
if the atom is trapped in a spherically symmetric external potential of natural frequency @xmath243 , the structure of the scalar center of mass term @xmath114 in the transition amplitude is given by eq .
( [ eq : cmylma ] ) , while that of vectorial internal transition factor @xmath244 has the structure given by eq .
( [ eq : jplus ] ) . since the product of two spherical harmonics
can always be written as a superposition of spherical harmonics , all the transition amplitudes @xmath103 can then be written in terms of the basic integrals @xmath245 similar arguments apply to the vectorial factor @xmath246 and the scalar factors @xmath149 , so that these basic integrals also determine the transition amplitudes associated to the interaction hamiltonians @xmath140 , eq .
( [ eq : hi2 ] ) and @xmath145 , eq .
( [ eq : hi3 ] ) . in general ,
any integral @xmath247 is relevant for the evaluation of more than one transition amplitude .
for instance , internal electric dipole transitions induced by @xmath117 involve @xmath248 which may be coupled to a center of mass transition with no rotational effects , that is @xmath249 and @xmath250 .
in such a case , the center of mass term involves a term @xmath248 , and the integral @xmath251 should be evaluated .
a very different situation corresponds to an electric quadrupole transition induced also by @xmath117 which involves internal @xmath184 terms and a center of mass transition with a similar contribution ( this transition could be accompanied by a change in the rotational wave numbers of the center of mass motion )
. the coupling of the two spherical harmonics will again have a contribution of the integral @xmath251 , but with different internal and center of mass factors . in the following subsections we elaborate on the dependence of the @xmath247 on the electromagnetic parameters @xmath3 and @xmath8 for different em modes .
electromagnetic fields with spherical symmetry can be written in terms of vector spherical harmonics as described in appendix b. if recoil effects are ignored in a radiative transition , the evaluation of internal transition atomic probabilities ( including both electric and magnetic effects important in the relativistic realm ) can be naturally described in terms of these modes @xcite . in this section ,
contrary to standard calculations , recoil effects have been incorporated in the calculation . the integrals @xmath247 necessary to study the role for a given em spherical mode @xmath252 on the coupled angular momentum changes in the internal and center of mass motion ( as encoded in a @xmath253 with @xmath228 = 0 , @xmath254 1 ) are @xmath255 note that , the vector integrals @xmath221 , eq .
( [ eq : icm2 ] ) , are orthogonal to magnetic vector modes , the terms of the hamiltonian @xmath140 that involve @xmath246 would yield center of mass transitions just in the case that electric spherical modes are used to induce them . using equations that can be found in appendix b
, it can be shown that @xmath256\nonumber\\ \vert { \mathcal{h}}^{(m;{m_{ph}};jm)}_{\ell , m_\ell}\vert & = & \frac{\sqrt{\hbar\omega^3}}{2\pi c } \delta_{\ell , j}\delta_{m_\ell , m+{m_{ph}}}\sqrt{2j+1}\begin{pmatrix}j & 1 & j\cr m+{m_{ph } } & -{m_{ph } } & -m\end{pmatrix},\end{aligned}\ ] ] here , @xmath257 and @xmath258 denote electric and magnetic spherical vector waves .
the @xmath259 factors indicate which spherical waves induce the coupled ( internal and center of mass ) atomic transitions .
taking into account the behavior of the 3j - symbols , in general , @xmath260 this result will be useful for the clear comparison between the integrals @xmath247 for different selections of the electromagnetic modes .
though spherical em modes give a natural scenario for the description of spontaneous internal transition rates , induced atomic transitions using these modes would require an efficient procedure for their selected generation ; this is a non trivial task .
perhaps the simplest implementation could be achieved by the controlled excitation of predetermined em modes within cavities with spherical symmetry . for axial symmetric scalar gaussian modes of waist @xmath230 ,
the integrals of the em field that go along the polarization vectors @xmath261 necessary to evaluate atomic transition amplitudes are @xmath262 while the integrals of the em field that go along the polarization vector @xmath263 are @xmath264 some specific expressions are : @xmath265 these functions are plotted in figure 1 .
it can be observed that it is necessary to consider modes with values of @xmath266 to observe a substantial increment of the transition rates with respect to the result obtained for paraxial beams . in general , the maxima of @xmath267 are achieved for waists @xmath230 smaller than the wavelength . as it could be expected , the functions @xmath268 decay as @xmath230 increases faster than the functions @xmath269 do .
in fact , for @xmath270 @xmath271 @xmath272 if @xmath158 is even , and @xmath273 if @xmath158 is odd . as a consequence for a scalar gaussian mode , taking into account the normalization factor given in eq.([eq : ngauss ] ) ,
all transition rates depend on the energy per unit transverse area of the em mode in the paraxial limit .
we have also evaluated the integrals @xmath274 for transverse electric and transverse magnetic gaussian modes .
the integrands differ from those of scalar beams by factors @xmath275 , @xmath276 and @xmath277 as a consequence of eq .
( [ eq : polarizations ] ) and eq .
( [ eq : hbmk ] ) .
they also decay as @xmath278 as @xmath270 , but they yield maxima with higher values of @xmath279 than the scalar modes .
in fact some of these maxima are beyond the spherical modes limit , eq .
( [ eq : sphh ] ) . in appendix
a , the normalization of propagation invariant modes is worked out in several schemes .
two of them refer to the case where ideal propagation modes are considered .
the third is the closest to the experimental realizations , it corresponds to work with a gaussian superposition of ideal modes around a given @xmath71 : @xmath280 with @xmath281 .
the relevant integrals for scalar propagation invariant modes can then be written as @xmath282 the set of terms @xmath283 is common to all propagation invariant modes , while @xmath284 depends on the details of the angular spectra that characterizes each kind of propagation invariant modes .
figure [ f : piofscalar ] illustrates the absolute value of the integrals @xmath285 for scalar modes as a function of the dominant transverse wave number @xmath71 measured in units of @xmath45 . in fig
[ f : piofvector ] , the vectorial analogues of these integrals are shown ; their evaluation incorporate the terms @xmath286 for @xmath287 , @xmath19 .
though both scalar and vector integrals depend on the waist of the gaussian @xmath278 used in eq .
( [ gauss : piof ] ) , the same reported general structure is observed whenever @xmath281 .
note that there is a substantial increment of the @xmath288 values as the parameter @xmath71 deviates from the paraxial approximation .
these occurs both for scalar and vector beams .
the integrals @xmath289 have a very simple expression for bessel beams : @xmath290 which is a manifestation of conservation of total angular momentum during the transition .
notice that this angular momentum can be originated from polarization terms like those in eq .
( [ eq : polarizations ] ) or from orbital terms related to the angular spectra eq .
( [ def : bessel : as ] ) .
they can also either be transferred to the center of mass or to the internal state of the atom depending the origin of the @xmath291 factor in the integral @xmath289 . for mathieu beams @xmath292
can be directly related to the @xmath293 and @xmath294 mathieu coefficients in the standard notation@xcite .
they achieve a maximum value for @xmath295 for @xmath296 as could be expected since in this limit the elliptic coordinates reduce to circular cylinder coordinates , and mathieu modes eq .
( [ def : mathieu : as ] ) become identical to bessel modes , eq .
( [ def : bessel : as ] ) . for moderate values of @xmath60 and as it increases , the off diagonal terms @xmath292 , @xmath297 also increase .
this means that mathieu modes can be used to create superpositions of internal and center of mass states whenever the corresponding internal , @xmath298 and center of mass , @xmath114 , vector or scalar factors have similar values for the @xmath299 values of interest .
[ f : mathieu ] we now report the results for the absolute value of @xmath300 as a function of the continuous parameter @xmath64 that defines the aperture of a weber mode . in this case
@xmath301 is much less than one , in general , with the exception of odd weber beams with @xmath64 nearby well identified discrete values . as a consequence ,
if a transition involving even weber modes takes place , the resulting atomic states will be a superposition of many orbital angular momentum states . in general ,
due to the intrinsic central field basic structure of the internal atomic states , these means that the recoil effects of the transition on the atom will be better described not as simple rotational .
this is consistent with the fact that weber modes are not eigenfunctions of the orbital angular momentum but carry , as a natural mechanical property , the product of the angular momenta along the @xmath23 axis and the linear momentum along a transverse direction @xmath302 , @xmath303@xcite . for internal dipole transitions induced by @xmath117 , the center of mass factor
@xmath114 will have a maximum overlap with the angular spectrum of a weber beam if @xmath304 this could happen if the initial and final wave functions of the center of mass are solutions of the schrdinger equation of a free particle with boundary conditions with parabolic symmetry .
one important question is whether or not the corresponding transition can be described in terms of a conservation law involving the atomic and the electromagnetic dynamical property @xmath303 .
a first experimental study with that scope was reported in ref .
for atoms in a cylindrical trap with the symmetry axis parallel to the main propagation axis of the light beam , according to the formalism developed in the first sections , the relevant integrals have the structure , @xmath305 \tilde \psi_{\kappa}^{(b)}(\mathbf{k}_\bot)y_{\ell , m_\ell}(\hat{\mathbf{k}}).\label{eq : cbmk}\ ] ] for the @xmath112 matrix element , @xmath306 the parameters @xmath228 , @xmath158 and @xmath307 are directly connected to the internal atom transitions as described in eq .
( [ eq : jplus ] ) , that is @xmath228 keeps the information about the vectorial coupling of the electron current to the em field , @xmath158 = 0 for internal dipole transitions , and @xmath158 = 1 , @xmath307=0,@xmath308 for electric quadrupole and magnetic dipole transitions .
higher multipole internal atomic transitions can be evaluated taking into account that , in terms of @xmath70 and @xmath309 , @xmath310 with @xmath311 the associated legendre functions .
the parameters @xmath312 , @xmath313 , @xmath314,and @xmath314 in eq .
( [ eq : cbmk ] ) , encode the information of the center of mass transition matrix elements , eq .
( [ eq : wfcc]-[eq : wfcc4 ] ) in appendix c. the functions @xmath315 and @xmath316 exhibit an exponential factor @xmath317 for a spherical symmetric trap the equivalent exponential factor is recovered , eq .
( [ eq : ss1 ] ) appendix c. the electromagnetic field structure is determined by the polarization vector @xmath318 and the angular spectrum @xmath319 . for bessel modes the quantum numbers are the polarization @xmath3 , the transverse wave number @xmath71 and the topological charge @xmath320 .
the evaluation of the integrals is direct , for scalar beams @xmath321\mathbb{c}^{\ell , m_\ell}_{\delta m_r,\mu,\alpha_\rho,\alpha_z}(\kappa_\bot)\label{eq : bbcs1}\ ] ] while for vector beams @xmath322\mathbb{c}^{\ell , m_\ell}_{\delta m_r,\mu,\alpha_\rho,\alpha_z}(\kappa_\bot)\label{eq : bbcs2 } \\
\mathcal{c}^{(tm;m_{b},\kappa_\bot)}_{\ell , m_\ell;\delta m_r,\mu,\alpha_\rho,\alpha_z}=
\mathcal{n}^{tm}_{\kappa_\bot,\omega}\big [ \frac{\kappa_\bot c\sqrt{1-\kappa_\bot^2c^2/\omega^2}}{\sqrt{2}\omega i}&\big[&\delta_{m_{ph } , 1}\delta_{m_\ell,1-m_{b } } + \delta_{m_{ph } , -1}\delta_{-m_\ell , m_{b}+1}\big]\nonumber\\ & -&\frac{\kappa_\bot^2\omega^2}{c^2}\delta_{m_\ell ,- m_{b}}\big ] \mathbb{c}^{\ell , m_\ell}_{\delta m_r,\mu,\alpha_\rho,\alpha_z}(\kappa_\bot)\label{eq : bbcs3}\end{aligned}\ ] ] with @xmath323 these analytical expressions allow an optimization procedure of the parameters to achieve enhancements or inhibitions of the transitions rates .
this is directly observed from the delta factors which determine the values of the bessel beam topological charge for a given beam polarization that yield specific changes in the atomic quantum number ; but it also applies for the selection of the @xmath71 parameter as illustrated below .
laguerre- gaussian ( lg ) modes , usually studied in the paraxial regime @xcite have an angular spectra given by eq .
( [ def : laggauss : as ] ) .
the vortex factor @xmath324 leads to similarities and important differences between the integrals related to laguerre gaussian and to bessel modes : while both share the same value of the integration on the @xmath309 variable , lg beams become zero as @xmath325 in a continuous way while for bessel beams the variable @xmath70 takes abruptly the nonzero value @xmath71 .
the angular similarity has as a consequence that both bessel and lg beams can be used to enhance or inhibit transitions involving selected changes in the atomic ( cm and internal ) angular momentum as observed in eq .
( [ eq : bbcs1]-[eq : bbcs3 ] ) , and mentioned in the last section for cm motion with spherical symmetry .
a detailed analysis on such a direction is reported , neglecting retardation effects , for bessel beams in ref .
@xcite and for laguerre gaussian beams in ref . @xcite . in the latter case ,
the interaction electric hamiltonian of the pzw formalism is used , free motion is assumed for the center of mass in the @xmath23 direction , and the particular case @xmath326 of the paraxial laguerre gaussian modes are studied . a quite interesting addition theorem for the factorization of the center or mass and the internal motion adequate for cylindrical symmetry is also introduced in ref .
@xcite . the allowed atomic and center of mass transitions for a given polarization and topological charge of the beams as described by eqs.([eq : bbcs1]-[eq : bbcs3 ] )
is consistent with the reported results in both references when the appropriate limits are taken .
let us illustrate an analytical procedure that can be used to optimize the selection of the @xmath327 ( lg beams ) and @xmath71 ( bessel beams ) parameters to enhance a given transition rate .
lg transitions are determined by the elementary factors , @xmath328 the function @xmath329 becomes zero at @xmath330 , 1 , if @xmath331 , @xmath332 @xmath333 , and the factor @xmath334 achieves its maximum value at @xmath335 . if @xmath336 and @xmath337 ( or @xmath338 and @xmath339 ) the maximum is achieved at @xmath340 ( or @xmath341 ) . as a consequence @xmath342 and
@xmath343 the maximum value of the latter term is one , and it is achieved for an elongated trap with @xmath344 , exhibiting and exponential decay on any direction of the @xmath345 plane , with independence on the values of the positive parameters @xmath332 and @xmath331 . as a consequence ,
transition rates are enhanced if the laguerre gauss beam is highly focused ( @xmath346 as small possible ) and the trap is an elongated one with @xmath347 as small as possible .
this is consistent with earlier observations @xcite that predict that atomic transitions can be enhanced by locating the atom close to a light vortex .
note that for @xmath348 , working with paraxial lg beams yields @xmath349 with the last factor being much smaller than one . for bessel beams , and given values of the parameters @xmath350 , the external parameter @xmath351 can be chosen to achieve the maximum value of the modulus of @xmath352 .
calculations can be performed in an analytical way for the elementary term @xmath353 which is also zero at @xmath354 and @xmath355 and achieves an extreme value for @xmath356 given by @xmath357 the sign is selected according to @xmath358 or @xmath359 . if @xmath360 , the maximum is achieved at @xmath361 .
these expressions give the optimal values of @xmath71 as a function of the exponents @xmath332 and @xmath331 , and the geometry of the trap in terms of the parameter @xmath362 .
in order to design experiments towards the control of atomic transition rates via laser light shaping , it is necessary to perform quantitative equitable calculations that allow the identification of the optimal parameters to inhibit or enhance such rates with respect to standard conditions where , in most cases , gaussian laser beams in the paraxial regime are used . in this work
we have shown that these calculations can be easily performed whenever a prior identification of the internal and center of mass states of interest is done ; this follows from the fact that the structure of radiative atomic transition amplitudes can always be written in terms of : * an internal atomic transition factor .
it has a vectorial character for the dominant interaction hamiltonian @xmath74 that results from the coupling of the electron current density to the electromagnetic field , and a scalar character for other terms in the non relativistic interaction hamiltonian including one term in the coupling of the nucleus current density to the electromagnetic field , and the terms quadratic in the electromagnetic field .
the vectorial factor @xmath194 can be evaluated using different approaches to the description of the atomic dynamical space , for instance , a non relativistic approach as that implicit in eq .
( [ eq : trans ] ) or a fully relativistic approach including the electron spin as the one given in ref . @xcite .
the central field structure of atoms makes that the dependence of the internal atomic transition factors on the electromagnetic wave vector @xmath1 can be naturally expressed in terms of spherical harmonics @xmath363 in all cases . *
an atomic center of mass transition factor that describes elementary recoil effects . under standard conditions ,
the velocities @xmath364 involved in the cm motion are much smaller than those associated to the internal motion @xmath365 ; that reflects on smaller values of the matrix elements of the interaction hamiltonian that couples the electromagnetic field to the nucleus current density than those related to the em field - electron current density coupling .
the atomic center of mass transition factor has a scalar character for the dominant interaction hamiltonian @xmath74 and for the terms quadratic in the electromagnetic field , and a vectorial character for the other terms in the minimal coupling interaction hamiltonian .
the vectorial factor @xmath246 was shown to correspond to the gradient in wave vector space of its scalar analogue @xmath366 .
the center of mass transition factor can play an active role in the control of internal atomic transition rates .
this has already been recognized in ref .
@xcite , where it was shown that the probability that the internal state of an atom acquires orbital angular momentum from a bessel mode is maximum for an atom located close to the symmetry axis of the beam . for a center of mass gaussian packet ,
the relevant parameter was found to be the ratio of the spread of the wave packet to the transversal wavelength of the photon . in the present formalism
this corresponds to taking @xmath367 in eq .
( [ eq : icm ] ) and @xmath368 to maximize the @xmath369 value in eq .
( [ eq : cmylma ] ) . in the case of an idealized cm description in terms of plane
waves the calculation of the elementary recoil effects is direct .
the cm transition factors are necessary to perform , via their incorporation in the corresponding bloch equations , realistic simulations of experiments at room temperature and moderate densities . * a factor that involves the em radiation structure which
is coupled to the internal transitions and recoil terms .
the em radiation is characterized in this term by its angular spectra @xmath370 and its polarization which , in general , also depends on @xmath1 .
the electromagnetic structure is completely encoded in the factor ( c ) that can be optimized to enhance or inhibit the transitions of interest .
the general procedure has been exemplified for atoms with center of mass available states described by ( i ) plane waves , ( ii ) wave functions exhibiting spherical symmetry and ( iii ) wave functions for traps with cylindrical symmetry . in the second case ,
em shaping is determined by the integrals @xmath371 given by eq .
( [ eq : hbmk ] ) ; a quantitative comparative analysis between the transition rates for different em fields was shown to be direct , as illustrated for spherical waves , gaussian beams with different waists and propagation invariant beams with diverse angular spectra .
spherical modes provide an ideal scenario for the description of spontaneous atomic transitions between internal states .
their usage for the control of induced transitions is conditioned by the capability of creating spherical modes with predetermined characteristics .
an upper bound for the important functions @xmath371 for spherical waves was found , eq .
( [ eq : sphh ] ) .
this value is also an upper bound for most of the corresponding integrals obtained for gaussian ( fig . 1 ) and propagation invariant modes ( fig . 2 ) .
an interesting exception corresponds to transverse magnetic gaussian beams inducing @xmath372 transitions ( @xmath373 in the antiparaxial regime , fig .
this study also allowed the recognition of the interactive role of polarization and em angular spectra . for electric quadrupole transitions , induced by @xmath74
, the selection rule @xmath374 allows changes of @xmath375 in the atomic angular momentum .
this yields an implicit requirement of light modes with orbital angular momentum .
these modes can be spherical ( as worked out in section vii.a ) or with other symmetries like bessel modes or laguerre gauss modes
. the plane waves decomposition in terms of spherical modes , eq .
( [ eq : sph ] ) , makes explicit their orbital angular momentum content . by changing the ratio between the transverse and longitudinal components in the dominant wave vectors yielding a structured beam ,
the relevance of predetermined internal states involved in a forbidden transitions can be modified .
this is a direct physical consequence of the behavior of the functions @xmath371 that encode the polarization and angular spectra effects on transition rates . by evaluating these functions ,
it was shown that , in the paraxial limit and for scalar gaussian beams , any multipole rate depends on the em energy per unit transverse area .
we also noticed that measurable increments of the transition rates for gaussian beams with respect to this asymptotic value requires waists of the order of the wavelength .
this is not a trivial task and , since gaussian modes ( as well as laguerre - gauss and hermite - gauss beams ) have a transverse wave vector distribution centered around @xmath242 , they are not necessarily the optimal choice for achieving that enhancement . for propagation invariant modes
, we have noticed that they could be used to enhance transition rates in more favorable conditions since their approximate implementation involves a narrow wave vector distribution centered around @xmath376 .
the relevance of the angular spectrum on recoil effects was studied in general and illustrated for mathieu , bessel and weber beams . using a plane wave description of the center of mass motion
, we showed that the em angular spectrum can be directly imprinted in the atomic recoil effects for dipole transitions .
in particular , our calculations confirm that , via dipole transitions , photons with orbital angular momentum can transfer this dynamical variable to the center of mass @xcite . for electric quadrupole and magnetic dipole transitions ,
the angular spectrum and polarization have diverse and interesting consequences on internal and center of mass transition rates .
for those transitions , the effective interaction term describing recoil effects depends directly on the gradient of the em field , so that semiclassical forces could depend on second derivatives of the electric and magnetic fields .
it was also shown that mathieu beams can be used to entangle atomic internal and center of mass degrees of freedom when the available states of the latter have spherical symmetry .
another important observation is that the formalism developed in this work can be adapted to other atomic systems such as molecules or nanostructures . in such cases , the particle that describes the transition of interest is not necessarily an electron .
one could be interested , for instance , on the rotation or vibration of atomic nuclei . in any case
, the symmetry of the internal available states should be taken into account for the most efficient calculation of the elementary integral containing the em factor @xmath377 .
if the atoms of interest are embedded in a material media , the electromagnetic field can be described using an equation similar to eq .
( [ eq : ef ] ) by replacing the proper dispersion relation ( which may depend on the polarization ) .
evanescent waves would also require a modification on the fourier factor @xmath378 to incorporate an exponential decay in the adequate direction .
then , the general formalism presented in this work would be properly adapted to describe transition rates in those systems .
the quantization of the electromagnetic field in terms of modes is usually based on the orthonormality requirement @xmath379 with the integration performed over the whole space @xmath380 .
for modes with polarization @xmath18 , @xmath19 this condition has a simple expression in terms of an integral of the modulus of the angular spectra , @xmath381 that is not the case for structured scalar modes due to the fact that the polarization vectors @xmath382 and @xmath383 are not orthogonal .
this kind of modes are frequently used in experimental realizations where circularly polarized gaussian beams impinge atomic samples . for scalar localized modes , eq .
( [ eq : tetm ] ) can be taken just as a normalization condition , @xmath384\vert\tilde\psi_{\kappa^\prime}(\mathbf{k}_\bot)\vert^2.\label{eq : circ2}\ ] ] in both eqs .
( [ eq : tetm]-[eq : circ2 ] ) , the integral in the left side must be performed within the circle of radius @xmath45 in the @xmath46 space . for axial symmetric scalar gaussian modes of waist @xmath230 ,
@xmath385 and circular polarization , the exact expression of the normalization integral valid beyond the paraxial approximation is @xmath386 \nonumber \\ & = & \hbar\omega,\end{aligned}\ ] ] with @xmath387 and erfi the imaginary error function . in the case of hermite - gaussian modes of waists
@xmath388 that guarantee the validity of the paraxial approximation , @xmath389 the polarization vectors @xmath382 and @xmath383 are quasi orthogonal , and the integral over the circle of radius @xmath45 can be replaced to that over the whole space
. then the orthonormalization condition is approximately achieved by taking @xmath390 .
propagation invariant beams are non localized waves ( the electric fields do not become zero fast enough as @xmath391)so that the orthonormalization requirement eq .
( [ eq : ortho ] ) gives rise to delta functions on the modulus of @xmath70 .
explicit results for te and tm modes have been reported for bessel@xcite , mathieu@xcite and weber@xcite modes .
an alternative for non localized waves , corresponds to evaluate normalization factors within a finite space region @xmath392 for plane waves this normalization usually considers a parallelepiped of volume @xmath393 .
this method has the advantage of being more intuitive and closer to experimental realizations .
taking this into account , we outline the main steps of this procedure for other propagation invariant beams . in that case , the natural space region for performing the normalization condition is a cylinder with symmetry axis along the main propagation axis , radius @xmath394 and length @xmath395 .
the normalization factor should then be written in terms of the angular spectra given above .
a relevant result is : @xmath396\big[\frac{\sin(k_z^\prime -k_z)l/2}{(k_z^\prime -k_z)l/2}\big]\ ] ] then , given a function of the form @xmath397 for @xmath399 , the last term is non negligible just for @xmath400 and , if we consider two modes with the same frequency , necessarily @xmath401 . for propagation invariant te or tm beams , the orthonormalization integral becomes @xmath402 @xmath403 for @xmath404 , @xmath405 so that , @xmath406 for bessel beams and mathieu beams the integral over @xmath26 is simply @xmath407 . for weber beams , the parameter @xmath64 is continuous and @xmath408 an third alternative valid for propagation invariant modes corresponds to defining them in terms of a gaussian superposition of plane waves centered at @xmath71 instead of the idealized definition given by eq .
( [ def : piof : as ] ) , @xmath409 with @xmath410 . in the limit @xmath411 with finite @xmath71
one recovers the idealized definition .
the elementary integrals necessary to evaluate the normalization factor @xmath412 for bessel , mathieu and weber gaussian beams are those described in the previous paragraph besides the integral @xmath413 a complete basis for transverse electromagnetic fields in free space can be constructed in terms of spherical vectors . in the notation established by eq .
( [ eq : bmodem ] ) , the corresponding modes are defined by ( see , e. g. , @xcite ) @xmath414 in these equations , the superscript specifies the electric @xmath415 and magnetic @xmath416 spherical modes , and @xmath417 with @xmath418 @xmath419 are the spherical harmonics , and @xmath420 and @xmath421 it can be shown @xcite that , @xmath422 @xmath423 with @xmath424 , @xmath425 and @xmath426 . from these expressions
an explicit relation for the basic integrals @xmath427 defined in eq .
( [ eq : sphh ] ) can be found , @xmath428 @xmath429\nonumber \\
+ \delta_{\ell , j-1}\sqrt{\frac{\ell+2}{\ell + 1}}\big[\delta_{\lambda,0}\sqrt{\frac{(\ell -m_\ell+1)(\ell + m_\ell+1)}{4(\ell + 1)^2 - 1 } } & + & \delta_{\lambda,1}\sqrt{\frac{(\ell - m_\ell)(\ell - m_\ell + 1)}{2(4(\ell + 1)^2 - 1)}}+\delta_{\lambda,-1}\sqrt{\frac{(\ell+m_\ell+2)(\ell+m_\ell + 1)}{2(4(\ell + 1)^2 - 1)}}\big]\nonumber\end{aligned}\ ] ] for the particular case of a confining potential that has cylindrical symmetry around the @xmath431-axis we write @xmath432 with @xmath433 , and @xmath434 .
the relevant integrals are @xmath435 @xmath436 @xmath437 and @xmath258 the confluent hypergeometric function .
in particular @xmath438 for the particular case of a confining potential that has spherical symmetry and natural frequency @xmath243 , the above results can be applied taking @xmath444 .
alternatively , we can write @xmath445 where @xmath446 in that case it results convenient to use eq .
( [ eq : sph ] ) , so that @xmath218 the radial integral is @xmath447 and the angular integrals can be written in terms of 3-j symbols yielding the coefficients @xmath448 , @xmath449 99 e. m. purcell _ phys .
rev . _ * 69 * , 681 ( 1946 ) ; d. kleppner _ phys .
lett . _ * 47 * , 233 ( 1981 ) ; p. goy , j. m. raimond , m. gross , and s. haroche , _ phys .
* 50 * , 1903 ( 1983 ) ; d. j. heinzen , j. j. childs , j. e. thomas , and m. s. feld , _ phys .
lett . _ * 58 * , 1320 ( 1987 ) .
a. flores - prez , j. hernndez - hernndez , r. juregui , and k. volke - seplveda , _ opt .
lett . _ * 31 * , 1732 ( 2006 ) ; c. maurer , a. jesacher , s. frhapter , s. bernet and m. ritsch - marte , _ new . j. of phys . _
* 9 * , 78 ( 2007 ) .
v. l. letokhov , v. g. minogin , and b. d. pavlik , _ zh
fiz . _ * 72 * , 1328 ( 1977 ) ; v. l. letokhov and v. g. minogin , _ appl .
_ * 17 * , 99 ( 1978 ) ; v. g. minogin and v. s. letokhov , _ laser light pressure on atoms _ ( gordon and breach , new york , 1987 ) . | a modular systematic analysis of the feasibility of modifying atomic transition rates by tailoring the electromagnetic field of an external coherent light source is presented .
the formalism considers both the center of mass and internal degrees of freedom of the atom , and all properties of the field : frequency , angular spectrum , and polarization .
general features of recoil effects for internal forbidden transitions are discussed . a comparative analysis of different structured light sources
is explicitly worked out .
it includes spherical waves , gaussian beams , laguerre - gaussian beams , and propagation invariant beams with closed analytical expressions .
it is shown that increments in the order of magnitude of the transition rates for gaussian and laguerre - gaussian beams , with respect to those obtained in the paraxial limit , requires waists of the order of the wavelength , while propagation invariant modes may considerably enhance transition rates under more favorable conditions . for transitions that can be naturally described as modifications of the atomic angular momentum , this enhancement is maximal ( within propagation invariant beams ) for bessel modes ,
mathieu modes can be used to entangle the internal and center of mass involved states , and weber beams suppress this kind of transitions unless they have a significant component of odd modes . however ,
if a recoil effect of the transition with an adequate symmetry is allowed , the global transition rate ( center of mass and internal motion ) can also be enhanced using weber modes . the global analysis presented reinforces the idea that a better control of the transitions between internal atomic states requires both a proper control of the available states of the atomic center of mass , and shaping of the background electromagnetic field . |
the spectral representation of the correlation functions of the vector current directly relates to the invariant mass spectrum of dileptons and photons .
additionally in the limit of small frequencies it determines a transport coefficient , in the case of the vector correlation function of light quarks , the electrical conductivity .
+ at temperatures relevant for current heavy ion experiments non - perturbative techniques are mandatory for the determination of these quantities .
perturbative studies of the vector spectral functions @xcite and also the inclusion of non - perturbative aspects through the hard thermal loop resummation scheme @xcite break down , especially in the low invariant mass region , indicated by an infrared divergent euclidean correlator @xcite , leading to an infinite electrical conductivity
. however it could be demonstrated that the spectral function at low invariant masses in fact increases linearly resulting in a finite electrical conductivity of the quark gluon plasma @xcite .
this behavior could also be established by previous work using lattice qcd @xcite .
+ in @xcite we analyzed the vector correlation function at @xmath1 and performed its extrapolation to the continuum limit based on precise data at various lattice sizes , corresponding to different lattice cutoffs .
we then obtained reliable results for the determination of the spectral properties and the extraction of the dilepton rates and transport coefficients . here
we report on a first extension of this work to a wider temperature range and present results on the electrical conductivity in the temperature region @xmath0 .
the desired information is encoded in the large distance region of temporal euclidean correlation functions of the vector current at non - zero temperature and fixed spatial momentum , @xmath2 whereby the current - current correlation functions can be represented in terms of an integral over spectral functions , @xmath3 : @xmath4 here we denote by @xmath5 the sum over the three space - space components of the spectral function and also introduce the vector spectral function @xmath6 .
+ the time - time component of the vector correlation function , @xmath7 , is conserved and can be shown to be proportional to the quark number susceptibility , @xmath8 . the vector correlation functions @xmath9 and @xmath10 therefore differ only by a constant , @xmath11 at high temperature and for large energies corrections to the free field behavior can be calculated perturbatively ; the vector spectral function can be deduced from the calculation of one loop corrections to the leading order results for the thermal dilepton rate @xcite .
this yields , @xmath12 in the free field , infinite temperature limit the spatial part of the spectral function contains a @xmath13-function at the origin .
different from the time - time component , where the @xmath13-function is protected by current conservation , this @xmath13-function is smeared out at finite temperature and the low energy part of @xmath5 is expected to be described by a breit - wigner peak @xcite , @xmath14 the limit @xmath15 is sensitive to transport properties in the thermal medium and in this case it leads to the electrical conductivity @xmath16 the breit - wigner form then yields @xmath17 . note here that in the infinite temperature limit the width of the breit - wigner peak vanishes , while at the same time @xmath18 , @xmath19 and consequently the electrical conductivity is infinite in the non - interacting case .
+ these considerations give rise to a sensible ansatz describing the high temperature vector spectral function : @xmath20 this ansatz depends on four temperature dependent parameters ; the quark number susceptibility @xmath21 , the strength ( @xmath22 ) and width ( @xmath23 ) of the breit - wigner peak and the parameter @xmath24 that parametrizes deviations from a free spectral function at large energies .
+ in the following we will use this ansatz to extract the electrical conductivity and the vector spectral function from the correlator data .
in addition to the vector correlation function itself we will calculate its curvature at the largest euclidean time separation accessible at non - zero temperature , _
i.e. _ at @xmath25 .
the curvature is the second thermal moment of the spectral functions at vanishing momentum , @xmath26 where @xmath27 is chosen to be even as all odd moments vanish .
these thermal moments give the taylor expansion coefficients for the correlation function expanded around the mid - point of the euclidean time interval , @xmath28 in the infinite temperature , free field limit the integral in eq .
[ eq : speccora ] can be evaluated analytically @xcite and one straight forwardly obtains the first three non - vanishing moments for massless quarks @xmath29 note here that all thermal moments , @xmath30 with @xmath31 , will be sensitive to the smeared @xmath13-function contributing to @xmath32 although we expect this contribution to become more and more suppressed in higher order moments @xcite .
+ in the following we will analyze the ratio of @xmath33 and the free field correlator @xmath34 , @xmath35 as well as the ratio of mid - point subtracted correlation functions @xmath36 here we used the notation @xmath37 . note that the curvature of these ratios at the mid - point determines the deviation of ratios of thermal moments from the corresponding free field values .
while the ratios of correlation functions differ in the @xmath38 and @xmath39 channels due to the additional constant contributing to @xmath40 , this constant drops out in the subtracted correlation function , _
i.e. _ , @xmath41 and @xmath42 for @xmath43 .
.parameters for the calculation of vector correlation functions on lattices of size @xmath44 . [ cols="^,^,^,^,^,^,^,^,^,^",options="header " , ] to discuss the consequences of our findings on the vector spectral function we invoke the breit - wigner+continuum ansatz defined in eq.[eq : ansatz ] and fit to the @xmath45 and @xmath46 data respecting also the thermal moments .
additionally we analyze the dependence of the fit parameters on @xmath47 , as , with the data being subject to potentially large lattice effects , we expect large errors originating from the fit - window we choose .
the resulting parameters are summarized in tab.[tab : fit - tdep ] .
+ here it can be seen that the results at @xmath48 and @xmath49 have the largest errors , nevertheless these results should be closest to those in the continuum as only the furthest distance points are comparatively free of lattice effects . without a continuum extrapolation
however it is at this point not possible to cleanly discern the residual deviation due to the lattice cut - off .
decreasing @xmath47 on the other hand leads to visible systematic trends in the parameters , as such the correction term @xmath50 and the width @xmath51 both increase with decreasing @xmath47 , while the electrical conductivity decreases .
additionally the @xmath52 increases with decreasing @xmath47 implying a lower quality fit with lower @xmath47 . in light of the argument above these results
show that indeed the lattice effects influence the fit more strongly with smaller @xmath47 , as they can not be taken care of by the employed ansatz .
consequently we attribute the largest part of the systematic trends observed above to the lattice effects .
+ focusing on the parameters themselves note that the correction factor @xmath24 is the most accurately determined .
clearly it shows an increasing trend with decreasing temperature differing by a factor @xmath53 between @xmath46 and @xmath54 .
however as @xmath24 is coupled to the strong coupling via @xmath55 @xcite , this behavior is expected . the same can be seen for the width @xmath56 and the parameter @xmath57 , they also slightly increase with decreasing temperature .
however here , they remain within errors of each other and the increase between @xmath46 and @xmath54 is small .
naturally this also leads to only small deviations of the electrical conductivity .
note however that the results at @xmath58 differ by @xmath59 from those of @xmath60 , while this deviation is only @xmath61 for @xmath62 . with the @xmath58 lattice being most affected by the systematics of the fit , we assume the largest part of this deviation to find its origins there .
even so we find the electrical conductivity to be to a good degree constant and within @xmath63 across the temperature region evaluated here , whereby the values at the individual temperatures have fit errors around @xmath64 .
+ in the temperature range @xmath0 our results therefore imply that the electrical conductivity depends linearly on temperature as expected from perturbation theory @xcite and at @xmath65 we find the value : @xmath66 whereby the first error quoted corresponds to the temperature effects and the second to the fit errors .
extending our earlier study we presented results on the vector correlation function in the high temperature phase of quenched qcd in the temperature range @xmath0 .
first results show a similar behavior to those extensively analyzed at @xmath67 and even at @xmath54 we find no sizable contribution attributable to the @xmath68-resonance . +
even though these calculations could not yet be extended to include also a continuum extrapolation as in @xcite , we were nevertheless able to extract the relevant spectral properties to calculate the electrical conductivity via an ansatz and found this phenomenologically relevant transport coefficient to depend linearly on temperature with a value of @xmath69 . + given the exact relation between the electrical conductivity and the diffusion constant , @xmath70 , it would be interesting to compare these results with those obtained in the case of charm diffusion @xcite or from the momentum diffusion constant @xmath71 calculable via hqet @xcite and we will do so in an upcoming publication . 99 -2pt h. t. ding .
, a. francis , o. kaczmarek , f. karsch , e. laermann and w. sldner , ; o. kaczmarek , a. francis , for an overview and further references , see for instance : f. gelis , j. p. blaizot and f. gelis , e. braaten and r.d .
pisarski , f. karsch , m. g. mustafa and m. h. thoma , p. b. arnold , g. d. moore and l. g. yaffe , ; ; p. aurenche , f. gelis , g. d. moore and h. zaraket , s. gupta , ; g. aarts et al .
, t. altherr and p. aurenche , g. aarts and j. m. martinez resco , p. petreczky and d. teaney , j. hong and d. teaney , w. florkowski and b.l .
friman , k. g. wilson , m. lscher , s. sint , r. sommer , p. weisz and u. wolff , s. wissel , phd - thesis bielefeld university ( 2006 ) h. t. ding , a. francis , o. kaczmarek , f. karsch , h. satz and w. sldner , hep - lat/1107.0311 a. francis , o. kaczmarek , m. laine , j. langelage , hep - lat/1109.3941 d. banerjee , s. datta , r. gavai , p. majumdar , hep - lat/1109.5738 | extending our recent work @xcite , we report on a calculation of the vector current correlation function for light valence quarks in the deconfined phase of quenched qcd in the temperature range @xmath0 . after performing a systematic analysis of the influence of cut - off effects on light quark
meson correlators using clover improved wilson fermions , we discuss resulting constraints on the electrical conductivity in a quark gluon plasma . |
the bound - state - approach advocated by callan and klebanov(ck)@xcite has been shown to work very well for static properties@xcite of strange baryons . in ck approach ,
strange baryons are described by properly quantized states of the @xmath5-meson(s)-soliton bound system .
et al_.@xcite extended the ck approach further to baryons containing a heavy flavor such as charm or bottom .
in particular , the mass spectra and magnetic moments@xcite for charmed baryons are found to be strikingly close to the predictions of the quark model description . in these calculations , vector meson fields such as @xmath6 , @xmath7 and @xmath8
are eliminated in favor of a combination of a background and corresponding pseudoscalar fields , @xmath5 , @xmath9 and @xmath10 .
this approximation is valid only when vector mesons are sufficiently heavier than corresponding pseudoscalar mesons as in the case of @xmath11 and @xmath12 ( @xmath13 mev , @xmath14 mev : @xmath15 ) . for charmed mesons or bottom flavored mesons , however , vector mesons are only a few percent heavier than corresponding pseudoscalar mesons : @xmath16 mev , @xmath17 mev ( @xmath18 ) , @xmath19 mev , @xmath20 mev ( @xmath21 ) , and @xmath22 mev , @xmath23 mev ( @xmath24 ) .
thus it is needed to treat heavy vector mesons correctly on the same footing as heavy pseudoscalar mesons .
the heavy quark symmetry is a new spin and flavor symmetry of qcd in the limit of infinite heavy quark masses .
as a heavy quark becomes infinitely heavy , the dynamics of a heavy quark in qcd depends only on its velocity and is independent of its mass and spin . this symmetry can be seen in weak semileptonic decays @xcite , mass splittings and partial decay widths @xcite of heavy mesons and heavy baryons , whose masses are much bigger than the qcd scale , @xmath25 .
recently , effective heavy meson lagrangians which have both chiral symmetry and heavy quark symmetry have been constructed by several authors@xcite . also , a lot of works on heavy baryons as skyrmions _
la _ callan - klebanov have been reported@xcite . in a series of papers@xcite , jenkins _
_ investigated the binding of a heavy meson with a soliton using such an effective lagrangian .
et al._@xcite studied the heavy quark symmetry in heavy baryon mass spectra in connection with the berry s phase .
et al._@xcite discussed roles of light vector meson degrees of freedom such as @xmath26 and @xmath11 . in these works , however , only the leading order terms in the inverse of the heavy quark mass have been considered .
also , bound heavy mesons are assumed to sit at the center of the soliton with their wavefunctions taken as @xmath27-functions . as a result
, heavy mesons appear to be too deeply bound and @xmath28 and @xmath29 are degenerate in mass . in order to investigate more realistic cases with hyperfine splittings , one needs to include next to leading order terms in @xmath0 . in ref.@xcite ,
mass corrections are roughly estimated by including mass differences between heavy pseudoscalar mesons and heavy vector mesons , while keeping the @xmath27-function - like wavefunctions .
although it may work well for bottom flavored baryons , we may have some doubts on the validity of such @xmath27-function - like wavefunctions for charmed baryons : the finite mass corrections need to be included in the wavefunctions of heavy mesons , leading to different radial functions , though sharply peaked at the center of the soliton . in this paper
, we attempt to establish a smooth " connection between the ck approach for light baryons and the heavy - meson - soliton bound state approach for heavy baryons by clarifying the above - mentioned problems . in our calculation ,
heavy pseudoscalar mesons and heavy vector mesons are treated on the same footing and the next to leading order terms in @xmath0 , are incorporated properly . in the following section ,
we introduce a simple lagrangian which is relevant to our purpose .
then , a soliton - heavy - meson bound state is found in sec . by solving the equations of motion for the classical eigenmodes of heavy mesons moving in the soliton background . in sec .
iv , we discuss the mass formula for heavy baryons containing a heavy quark .
we also discuss the heavy quark symmetry breaking by the wess - zumino term in the heavy baryon mass spectra .
section v contains summary and conclusion .
in order to avoid any unnecessary complications , we work with a simple lagrangian for the interaction of light goldstone bosons with heavy mesons , which has the @xmath30 chiral symmetry and the heavy quark symmetry in the heavy mass limit .
one may obtain such a lagrangian from the skyrme model lagrangian by trimming away all the higher derivative terms or from the heavy quark effective lagrangian by including the next to leading order terms in @xmath0 . up to a single derivative on the goldstone boson fields , the most general chirally invariant lagrangian density and @xmath26@xcite .
] can be written in a form of @xcite @xmath31 where @xmath32 and @xmath33 are the heavy pseudoscalar and the heavy vector meson doublets and @xmath33 than that of ref.@xcite .
our @xmath34 corresponds to their @xmath35.]with masses @xmath36 and @xmath37 , respectively .
for example , in the case of charmed mesons , we have @xmath38 the lagrangian density for the goldstone boson fields is @xmath39 ^ 2 , \eqno(\mbox{\ref{lag}a})\ ] ] with @xmath40 and @xmath41 being the pion decay constant .
the skyrme term " with a dimensionless parameter @xmath42 is included to stabilize the soliton solution . here
@xmath43 and @xmath44 are the @xmath45 and @xmath46 coupling constants .
the vector and axial vector potentials @xmath47 and @xmath48 are defined in terms of @xmath49 as @xmath50 and the covariant derivative @xmath51 and the field strength @xmath52 are @xmath53 under @xmath30 chiral transformations , the fields transform as @xmath54 where @xmath55 and @xmath56 are global transformations in @xmath57 and @xmath58 respectively and @xmath59 is a special unitary matrix depending on @xmath55 , @xmath56 and the goldstone fields .
furthermore , the lagrangian is invariant under the parity operation @xmath60 here , we have used the fact that pions and heavy mesons ( both pseudoscalar mesons and vector mesons ) carry negative intrinsic parity .
we have four parameters in the lagrangian to be fixed ; the pion decay constant @xmath41 , the skyrme parameter @xmath42 and the coupling constants @xmath43 and @xmath44 . the pion decay constant , @xmath41 , and the skyrme parameter , @xmath42 , are fixed by fitting the nucleon and delta masses in the @xmath61 sector@xcite . as for the heavy meson coupling constant @xmath43 and @xmath44 ,
little has been known except the upper bound@xcite .
thus , we use the heavy quark symmetry as a guide line .
we use the empirical masses for the heavy meson masses , @xmath36 and @xmath37 . in the heavy mass limit , we have @xmath62 with the mass difference being of order @xmath0 at most and the two coupling constants @xmath43 and @xmath44 become related to each other by@xcite @xmath63 due to the heavy - quark spin symmetry .
furthermore , @xmath44 approaches a universal constant @xmath64 due to the heavy - quark flavor symmetry .
the nonrelativistic quark model estimate of @xmath64 ( @xmath65)@xcite is consistent with the experimental value ( @xmath66 0.5 ) measured via the @xmath7 decay width@xcite and the @xmath67 and @xmath68 branching ratios@xcite .
one may determine the two coupling constants @xmath43 and @xmath44 from a low energy chiral theory
. in ref.@xcite , a lagrangian for the interactions of @xmath5 and @xmath6 mesons with pions is derived on the basis of @xmath69 chiral symmetry along the hidden gauge symmetry scheme .
comparing it with our lagrangian , we get @xmath70 , which is very close to the nonrelativistic quark model prediction .
although the @xmath46 term proportional to @xmath44 is missing in ref.@xcite , one can find such a term among the homogeneous solutions of the wess - zumino anomaly equation .
( see ref.@xcite for further details . )
using the vector meson dominance hypothesis and the empirical value on the @xmath71 coupling constant ( @xmath72 ) , we obtain @xmath73 from the chiral lagrangian of ref.@xcite .
the lagrangian density @xmath74 supports a stable @xmath61 soliton solution of hedgehog " type @xmath75 with @xmath76 the above solution carries a nontrivial winding number due to its nontrivial topological structure identified as the baryon number @xmath77 and a finite mass @xmath78 with @xmath79 .
now , our problem is to find the eigenmodes of the heavy mesons moving in the static potentials provided by the @xmath80 soliton configuration ( [ hedgehog ] ) sitting at the origin ; _ viz . _ , @xmath81 with @xmath82 the equations of motion can be read off from the lagrangian ( [ lag ] ) : @xmath83 for the pseudoscalar meson field @xmath32 and @xmath84 for the vector meson fields @xmath33 .
the conjugate momenta to the meson fields @xmath32 and @xmath33 are @xmath85 respectively , and we get similar equations for @xmath86 and @xmath87 . since @xmath88 vanishes identically , the @xmath89 can not be an independent dynamical variable .
we eliminate the complementary @xmath89 field by using eq .
( [ eq_phi * ] ) @xmath90 which results in a set of coupled equations @xmath91 where @xmath92 . in order to express the equations of motion only in terms of @xmath32 and @xmath93 , we use the fact that the @xmath89 field is of order @xmath0 at most ; _ viz .
_ , @xmath94 keeping this leading order term leads us to the equations of motion @xmath95 because of the spin - isospin mixing in the hedgehog configuration of the classical background , the equations of motion ( [ eq_k ] ) and ( [ 14 ] ) are invariant only under the rotation by the grand spin @xmath96 defined by @xmath97 with @xmath98 being the spin ( isospin ) of the heavy mesons and @xmath99 the orbital angular momentum .
thus , eigenmodes are classified by the quantum numbers @xmath100 , @xmath101 and @xmath102 ( the parity , @xmath103 with @xmath104 being the orbital angular momentum ) as @xmath105 where @xmath106 and @xmath107 are the generalized spherical spinor and vector harmonics , respectively , and @xmath108 is an index to label the possible vector spherical harmonics with the same @xmath100 , @xmath101 and @xmath102 . to avoid cumbersome notation
, we will suppress the trivial indices @xmath100 , @xmath101 and @xmath102 of the radial functions as @xmath109 and @xmath110 .
from now on , we will restrict our consideration to @xmath111 states , which are expected to have at least one bound state . since pseudoscalar mesons do not carry spin , we have only one spherical spinor harmonics with @xmath111 : @xmath112 here , @xmath113 is the isospin basis for the heavy meson doublets , _
i.e. _ , @xmath114 for vector mesons with spin 1 , we can construct two different @xmath111 vector spherical harmonics@xcite : _ viz .
_ , @xmath115 putting @xmath116 into the equations of motion ( [ eq_k ] ) and ( [ 14 ] ) , we obtain three coupled differential equations for the radial functions : @xmath117 the wavefunctions are normalized such that each mode carries one corresponding heavy flavor number : @xmath118 + g_{_q } \left [ ( a_1+a_2)|\varphi^*_2|^2 - \sqrt2 a_1 ( \varphi_1^{*\dagger } \varphi^*_2 + \varphi_2^{*\dagger } \varphi^*_1 ) \right ] \right\ } , \label{norm}\end{aligned}\ ] ] where we have kept terms up to the next to leading order in @xmath0 . near the origin , the equations of motion behave asymptotically as @xmath119 they imply that we have three independent solution sets as @xmath120 with @xmath121 and @xmath122 . for sufficiently large @xmath123 ,
the three equations decouple from each other : for example , @xmath124 thus the bound state solutions ( @xmath125 ) are @xmath126 with three constants @xmath127 , @xmath128 and @xmath129 .
the lowest energy bound states are found numerically , and the results are shown in table i and fig . 1 . in table
i , the input parameters are listed together with the numerical results on the lowest bound states . in fig . 1 , we give the radial functions @xmath130 and @xmath131 for the @xmath9 and @xmath7 mesons ( solid curve ) and the @xmath10 and @xmath8 mesons ( dashed curves ) . by comparing the two cases , one can easily check that as the meson mass becomes larger , ( 1 ) the radial function becomes more sharply peaked at the origin and ( 2 ) the role of the vector mesons becomes important so that the radial function @xmath131 becomes comparable to @xmath130 ( see also the ratio @xmath132 ) .
the radial function @xmath133 , though not shown in fig .
1 , is hardly distinguishable from @xmath134 .
this can be understood as follows : due to their heavy masses , heavy mesons are localized in the region @xmath135@xmath136 , where @xmath137 \sim [ - a_1(r ) ] \sim f'(0)+o(r^2 ) , \\
\upsilon(r ) \sim \displaystyle\frac{1}{r } -\textstyle\frac14 f^{\prime2}(0)r+\cdots , \end{array}\label{apppot}\end{aligned}\ ] ] so that the equation of motion for @xmath138 is completely decoupled from those for @xmath139 and @xmath140 .
it would be interesting to compare our radial functions with those of ref.@xcite and ref.@xcite . in ref.@xcite ,
vector mesons are assumed to be sufficiently heavy and the following _ ansatz _ is made : @xmath141 which implies that @xmath142 as @xmath143 for heavy mesons due to eq .
( [ apppot ] ) , we have only to compare @xmath144 with @xmath139 in eq .
( [ comp ] ) . in the heavy mass limit
, both should play equally important roles .
but the ansatz strongly suppresses the role of vector mesons by a factor of @xmath145 , since one obtains @xmath146 in the skyrme - term - stabilized soliton solution .
for example , this factor amounts to 0.56 , 0.25 and 0.09 for the cases of @xmath147(892 mev ) , @xmath148(2010 mev ) and @xmath149(5325 mev ) , respectively .
therefore , the ansatz of eq .
( [ ansatz ] ) is not valid unless the vector meson is much heavier than the corresponding pseudoscalar meson .
the wavefunctions of refs.@xcite are obtained in the heavy mass limit , @xmath150 and can be written in our convention as @xmath151 where the radial function @xmath152 , normalized as @xmath153 , is strongly peaked at the origin .
it implies that @xmath154 these radial functions satisfy the normalization condition of eq .
( [ norm ] ) in the leading order in @xmath0 ; _ viz .
_ , @xmath155 it is interesting to note that the pseudoscalar meson and three vector mesons contribute equally to the bound state .
comparing our numerical results given in table i with the binding energy @xmath156 of refs.@xcite which gives @xmath4 mev with the same input parameters , one can see that the @xmath0 corrections amount to @xmath157 mev in the bottom sector and @xmath158 mev in the charm sector .
this is one of the main results of this work . in ref.@xcite ,
the rms radii of the heavy flavor current in heavy baryons are essentially zero . due to the @xmath0 corrections
, however , we have non - zero finite size rms radii in our calculation , _ viz . _ @xmath159 fm for bottom flavored baryons and @xmath160 fm for charmed baryons .
this implies that the rms radii of heavy flavored baryons become small as the masses become large .
due to this effect , the binding energy is smaller than the one obtained with @xmath27-function - type solutions .
so far we have considered soliton - heavy - meson bound states to the order @xmath161 with @xmath162 being the number of color .
the combined system of the soliton and a bound heavy meson carries a baryon number and a heavy flavor number , but does not have the spin and isospin of a heavy baryon .
up to order @xmath161 , the soliton - heavy - meson bound state should be understood as a mixed state of three degenerate heavy baryons containing a heavy quark @xmath163 ; @xmath28 , @xmath164 and @xmath29 , whose mass is @xmath165 . in order to give the spin and isospin quantum numbers and the hyperfine splittings
, we have to go to the next order in @xmath1 , _
i.e. _ , @xmath166
. this is done by quantizing the zero modes associated with the simultaneous @xmath61 rotation of the combined system . a standard collective coordinate
quantization procedure leads us to the mass formula for a heavy baryon with spin @xmath167 and isospin @xmath168 : @xmath169 here @xmath170 is the moment of inertia of the soliton configuration against the @xmath61 collective rotation : @xmath171 and @xmath2 is the hyperfine splitting constant which can be obtained by directly applying the techniques developed in ref.@xcite : @xmath172 \\ & \hskip 3 mm + \left.\frac13g_{_q}\left [ ( f ' - \displaystyle \frac{\sin\!2f}{r } ) |\varphi^*_2|^2 - \frac{1}{\sqrt2 } \frac{\sin\!f}{r } ( 3\cos f + 1 ) ( \varphi_1^{*\dagger } \varphi^*_2 + \varphi_2^{*\dagger } \phi_1 ) \right ] \right\}. % \\ \end{array}\eqno(\mbox{\ref{mf}b})\ ] ] we note that we have also kept terms of the next to leading order in @xmath0 .
one can easily see that the hyperfine constant @xmath2 is of order @xmath0 .
the leading order terms proportional to @xmath173 vanish identically when the radial functions of eq .
( [ imlimit]a ) are used . according to the formula ,
the masses of heavy baryons containing a single heavy quark have following hyperfine splittings : @xmath174 by eliminating @xmath2 from eq .
( [ masssplit ] ) we have a model independent relation @xmath175 with the experimental values @xmath176(@xmath177 mev ) , @xmath178(@xmath179 mev ) and @xmath180(@xmath181 mev ) , we predict the mass of @xmath182 to be 2493 mev and the averaged mass @xmath183 5836 mev . since @xmath2 is of order @xmath0 , the masses of @xmath28 and of @xmath184 are degenerate in the infinite mass limit as the heavy quark symmetry implies and eq .
( [ mrel ] ) is reduced to @xmath185 as in refs.@xcite
. numerical results ( result i ) on the heavy baryon masses are shown in table ii .
they are in rough agreement with the experimental values .
result ii is obtained by taking the two coupling constants as free parameters .
to fit the experimental masses of @xmath186 and @xmath187 , one should have @xmath188 and @xmath189 , which implies that the heavy quark symmetric relation ( [ rel ] ) is _ strongly _ broken in the charm sector .
note that as far as the two coupling constants are related by eq .
( [ rel ] ) , the hyperfine constant is too small . in order to improve the situation
, one may consider higher order terms in the @xmath0 expansion or higher derivative terms of the pion fields . as a guide line
, we may use the skyrme lagrangian@xcite with the vector mesons included via the hidden gauge symmetry , since in the strangeness sector the heavy quark symmetry becomes no longer a good symmetry but the @xmath69 chiral symmetry becomes rather a good symmetry . among many possible terms which will be discussed below
, the wess - zumino ( wz ) term is known to play the most important role in ck approach@xcite : @xmath190 here , @xmath162 is the number of color , @xmath191 is the @xmath32-meson decay constant and @xmath192 is the topological baryon number current .
although its role fades out in the heavy mass limit , the wz term should not be disregarded in the case of finite quark masses .
in addition , there are other contributions of order @xmath0 to the binding potential .
we introduce a typical @xmath193 potential in the lagrangian as in ck approach , which is next to leading order in the derivative of pion fields and turns out to have non - negligible effects in the strangeness sector : @xmath194 now we discuss the effects of the above terms in detail .
let us write @xmath195 this additional lagrangian modifies the equations of motion for the pseudoscalar meson field @xmath32 as @xmath196 while those for the vector meson field @xmath197 remain the same as eq .
( [ 14 ] ) .
consequently , the radial function of the @xmath198 eigenmodes is altered as @xmath199 where @xmath200 the wz term contributes to the hyperfine splitting constant @xmath2 as @xmath201 and to the normalization condition of eq .
( [ norm ] ) by the same amount .
note that there is no direct contribution from @xmath202 to this quantity .
we begin with the role of @xmath203 in the strangeness sector , where the above model lagrangian does not work well as expected in this sector . in table
iii , we show the numerical results for strange baryons obtained with the input parameters @xmath204 mev , @xmath205 , @xmath206 mev , @xmath207 mev and @xmath208 .
note that the role of @xmath202 in the binding energy is important ( @xmath209 80 mev ) , while its effect on the hyperfine constant @xmath2 is rather small .
the wess - zumino term plays a crucial role in the hyperfine constant @xmath2 , of which more than 80% comes from the wz term . as shown in fig .
2 , the effect of the wz term on the radial wavefunction is also remarkable ; with the wz term , the vector meson contribution to the bound states is much suppressed compared with that of the pseudoscalar meson . in the charm sector the role of the wess - zumino term in the heavy mass limit is weakened as discussed in ref.@xcite .
this results from the fact that the role of vector mesons balances off that of pseudoscalar mesons in the heavy mass limit .
the decoupling of the wz term in the heavy mass limit is originally argued in ref.@xcite . in the chiral limit where pseudoscalar mesons predominate ,
the wz term is entirely expressed in terms of these pseudoscalar mesons . to take into account the wz term in the finite mass region ,
we consider the most characteristic expression of the wz term@xcite constructed in terms of pseudoscalar mesons and an adjustable parameter @xmath210 .
the parameter @xmath210 contains the trace of cancellation between contributions of the vector and of the pseudoscalar mesons and should depend on @xmath0 .
that is , we take @xmath211 with the same @xmath212 and @xmath202 as given by eqs . ( [ l1a],[l1b ] ) .
the parameter @xmath213 has the role of turning on and off the effect of @xmath202 . here , @xmath214 is the @xmath9-meson decay constant @xmath215 , which is known to be 1.8 times larger than the pion decay constant @xmath41 .
although the @xmath202 plays a minor role for the heavy flavors such as charm , we keep it to compare its effects in the charm sector with those in the strangeness sector . in fig .
3 , we present @xmath173 and @xmath2 as a function of the mitigating factor @xmath210 . the role of @xmath202
is shown as narrow stripes , with @xmath216 mev effect on the energy and @xmath217 on the hyperfine constant . however , as we can see in fig .
3 , the dependence of the mass spectrum on the parameter @xmath210 is not negligible . in order to fit the charmed baryon masses ,
we need to have @xmath218 mev and @xmath219 .
then we have @xmath220 also from fig .
3 , one can estimate the mitigating factors as @xmath221 and @xmath222 , which reproduce the above mass spectra and then we have @xmath223 fm .
it implies that the increase in @xmath214 alone is not enough to take fully into account the role of the wess - zumino term in the charm sector .
the dependence of @xmath210 on meson masses can be derived by showing how the wess - zumino term scales out as the mass increases .
we are not in a position to illustrate this dependence yet . however , if we assume the dependence to be inversely proportional to the @xmath224 , we find @xmath225 .
note the coincidence of the @xmath210-factor ( @xmath226 ) with the meson mass ratio , @xmath227 .
in this work , we have investigated the mass spectrum of heavy baryons containing a single heavy quark in the bound state approach of skyrme model . to this end , we have worked with the heavy meson lagrangian of ref.@xcite which includes the @xmath0 order terms . the large binding energy obtained in the infinite mass limit
is lowered by introducing @xmath0 corrections .
the binding energy is changed from @xmath4 mev to @xmath228 mev for @xmath229 mesons and to @xmath230 mev for the @xmath231 .
the effect may be crucial for the loosely bound exotic states such as pentaquark " baryons@xcite .
however , due to the realization of the heavy quark symmetry , the hyperfine splitting constant comes out too small compared with the experimental one .
for example , we get @xmath232 for charmed baryons while it should be @xmath233 to reproduce the experimental masses . to resolve this problem
, we introduce the wz term in a mitigated form , which is known to have a crucial role in the strangeness sector . to reproduce the experimental masses for charmed baryons ,
its strength should be weakened by a factor of @xmath234 .
we would like to thank mannque rho for having brought us this problem and for enlightening discussions .
we also would like to thank h. jung for reading the manuscript and useful comments .
also , y. oh is grateful to n. n. scoccola for useful discussions .
this work is supported in part by the korea science and engineering foundation through the src program and by cnu research and scholarship fund . c. g. callan and i. klebanov , nucl .
b262 * , 365 ( 1985 ) .
k. dannbom , e. m. nyman , and d. o. riska , phys .
b227 * , 291 ( 1989 ) ; n. n. scoccola , phys .
b236 * , 245 ( 1990 ) ; e. m. nyman and d. o. riska , nucl . phys . *
b325 * , 593 ( 1989 ) ; j. kunz and p. j. mulders , phys . lett .
* b231 * , 335 ( 1989 ) ; d .-
min , y. s. koh , y. oh , and h. k. lee , nucl phys . * a530 * , 698 ( 1991 ) ; y. kondo , s. saito , and t. otofuji , phys . lett . *
b236 * , 1 ( 1990 ) .
m. rho , d. o. riska , and n. n. scoccola , phys . lett . * b251 * , 597 ( 1990 ) ; z. phys . *
a341 * , 343 ( 1992 ) ; d. o. riska and n. n. scoccola , phys .
b265 * , 188 ( 1991 ) .
y. oh , d .-
min , m. rho , and n. n. scoccola , nucl .
a534 * , 493 ( 1991 ) .
n. isgur and m. b. wise , phys .
b208 * , 504 ( 1988 ) ; phys . lett . *
b232 * , 113 ( 1989 ) ; h. georgi , nucl . phys
. * b348 * , 293 ( 1991 ) ; n. isgur and m. b. wise , phys . rev . *
d43 * , 819 ( 1991 ) .
n. isgur and m. b. wise , phys .
* 66 * , 1130 ( 1991 ) .
g. burdman and j. f. donoghue , phys . lett . *
b280 * , 287 ( 1992 ) .
m. b. wise , phys . rev . * d45 * , 2118 ( 1992 ) . t .- m .
et al . _ ,
phys . rev . * d46 * , 1148 ( 1992 ) .
e. jenkins , a. v. manohar , and m. b. wise , nucl . phys .
* b396 * , 27 ( 1993 ) ; e. jenkins and a. v. manohar , phys . lett . *
b294 * , 273 ( 1992 ) ; z. guralnik , m. luke , and a. v. manohar , nucl . phys . *
b390 * , 474 ( 1993 ) .
min , y. oh , b .- y . park , and m. rho , seoul natl univ .
preprint snutp-92 - 78 , unpublished .
m. a. nowak , m. rho , and i. zahed , phys .
b303 * , 130 ( 1993 ) . k. s. gupta , m. a. momem , j. schechter , and a. subbaraman , phys .
* d47 * , 4835 ( 1993 ) ; j. schechter and a. subbaraman , phys .
* d48 * , 332 ( 1993 ) . g. s. adkins , c. r. nappi , and e. witten , nucl .
phys . * b228 * , 552 ( 1983 ) .
s. barlag _
, ( accmor collaboration ) phys . lett .
* b278 * , 480 ( 1992 ) .
min , y. oh , b .- y . park , and m. rho , ` heavy quark symmetry and skyrmions ' , to be published .
s. butler _
et al . _ , ( cleo collaboration ) phys
lett . * 69 * , 2041 ( 1992 ) .
n. n. scoccola , d .-
min , h. nadeau , and m. rho , nucl .
a505 * , 497 ( 1989 ) .
we combine first the spin basis and the orbital angular momentum basis to the total spin ( @xmath235 ) basis and then combine the isospin isospin .
then , @xmath236 and @xmath237 correspond to @xmath238 and @xmath239 states , respectively .
one may obtain the vector spherical harmonics in the other ways ; for example , by combining first the isospin basis and the orbital angular momentum basis to @xmath240 basis and then the spin basis , which leads us to @xmath241 they are @xmath242 and @xmath243 states , respectively .
particle data group , phys . rev . *
d45 * , part ii ( 1992 ) .
j. soto and r. tzani , phys .
* b297 * , 358 ( 1992 ) . m. a. nowak , m. rho , and i. zahed , phys .
* d48 * 4370 ( 1993 ) .
h. j. lipkin , phys .
b195 * , 484 ( 1987 ) ; c. gignoux , b. silvestre - brac , and j. m. richard , phys . lett . *
b193 * , 323 ( 1987 ) ; d. o. riska and n. n. scoccola , phys . lett . *
b299 * , 338 ( 1993 ) .
y. oh , b .- y . park , and d .-
min , in preparation .
cccccccc @xmath212 & @xmath202 & @xmath244 & @xmath2 & @xmath245 & @xmath246 & @xmath247 & @xmath248 + off & off & 389 & 0.098 & 1257 & 1433 & 1462 & 0.62 + off & on & 291 & 0.148 & 1160 & 1326 & 1369 & 0.56 + on & off & 191 & 0.717 & 1095 & 1151 & 1361 & 0.41 + on & on & 109 & 0.791 & 1022 & 1063 & 1295 & 0.39 + & & & 1116 & 1192 & 1385 &
+ units < 1cm,1 cm > point at 3 2 x from 0 to 15 , y from 0 to 8 at 7.5 0.5 at 7.75 0 units < 9cm,0.40 cm > point at 0 0 x from 0 to 1.0 , y from 0 to 15.0 bottom ticks in numbered from 0 to 1.0 by 0.5 unlabeled short from 0 to 1.00 by 0.1 / left ticks in numbered from 0 to 15 by 5 unlabeled short from 1 to 14 by 1 / top ticks in unlabeled short from 0.1 to 0.9 by 0.1 / right ticks in unlabeled short from 1 to 14 by 1 / at 0.75 -1 ( . ) .00120 8.6307 .04918 8.4075 .05517 8.3504 .07317 8.2363 .09116 8.1247 .10916 7.9766 .12715 7.7979 .14514 7.5931 .16314 7.3658 .18113 7.1195 .19912 6.8574 .21711 6.5827 .23511 6.2982 .27109 5.7115 .30708 5.1181 .34306 4.5357 .37905 3.9783 .41503 3.4561 .45102 2.9760 .48701 2.5418 .52299 2.1548 .52899 2.0949 .56497 1.7619 .60096 1.4727 .63695 1.2241 .67293 1.0122 .70892 .83314 .74490 .68289 .78089 .55763 .87685 .31985 1.0088 .14495 / at 0.4 4.5 .00120 7.3999 .04918 7.2171 .05517 7.1595 .07317 7.0490 .09116 6.9466 .10916 6.8115 .12715 6.6494 .14514 6.4641 .16314 6.2590 .18113 6.0373 .19912 5.8021 .21711 5.5560 .23511 5.3020 .27109 4.7805 .30708 4.2565 .34306 3.7456 .37905 3.2603 .41503 2.8092 .45102 2.3979 .48701 2.0291 .52299 1.7033 .52899 1.6531 .56497 1.3759 .60096 1.1374 .63695 .93444 .67293 .76331 .70892 .62024
.74490 .50155 .78089 .40378 .87685 .22224 1.0088 .09438 / at 0.25 3.5 .00120
13.786 .04318 13.363 .04918 13.186 .06717 12.949 .08516 12.646 .10316 12.253 .12115 11.788 .13914 11.263 .15714 10.692 .17513 10.086 .19312 9.4557 .24110 7.7340 .27709 6.4795 .31308 5.3152 .34906 4.2743 .38505 3.3738 .42103 2.6170 .45702 1.9973 .49300 1.5014 .52899 1.1129 .56497 .81418 .60096 .58849 .63695 .42062 .67293 .29753 .76889 .11301 .83487 .056235 .90084 .02738 .96681 .01309 1.0328 .00617 / at 0.18 12 .00120 12.923 .04318 12.465 .04918 12.359 .06717 12.131 .08517 11.842 .10316 11.469 .12115 11.027 .13914 10.530 .15714 9.9887 .17513 9.4148 .19312 8.8187 .24110 7.1935 .27709 6.0125 .31308 4.9192 .34906 3.9446
.38505 3.1040 .42103 2.3998 .45702 1.8250 .49300 1.3668 .52899 1.0091 .56497 .73519 .60096 .52905 .63695 .37636 .67293 .26490 .76889 .09914 .83487 .04877 .90084 .02344 .96681 .01105 1.0328 .00513 / at 0.10 9.5 units < 1cm,1 cm > point at 1 2.5 x from 0 to 15 , y from 0 to 8.5 at 7.5 0.5 at 7.75 0 . at 7.75 -0.5 units < 4cm,0.2 cm > point at 0 -15 x from 0 to 1.0 , y from -15 to 15 bottom ticks in numbered from 0 to 1.0 by 0.5 unlabeled short quantity 11 / left ticks in numbered from -15 to 15 by 5 unlabeled short quantity 13 / top ticks in unlabeled short quantity 11 / right ticks in unlabeled short quantity 13 / ( . ) from 0 0 to 1 0 at 0.75 -17.5 at 0.5 -20 .00120 13.801 .04318 13.276 .05518 13.102 .06717 12.962 .07917 12.770 .09116 12.535 .10316 12.264 .11515 11.960 .12715 11.628 .13914 11.272 .17513 10.092 .21112 8.8158 .24710 7.5225 .28309 6.2782 .31907 5.1312 .35506 4.1120 .39104 3.2350 .42703 2.5015 .46302 1.9035 .49900 1.4269 .56497 .81201 .63095 .44361 .69692 .23392 .76290 .11963 .82887 .05960 .89484 .02904 1.0268 .00654 / at 0.3 9 .00120 -12.905 .043183 -12.418 .055178 -12.247 .06717 -12.113 .079169 -11.931 .091164 -11.709 .10316 -11.452 .11515 -11.164 .12715 -10.850 .13914 -10.513 .17513 -9.3987 .21112 -8.1949 .24710 -6.9780 .28309 -5.8100 .31907 -4.7361 .35506 -3.7846 .39104 -2.9682 .42703 -2.2877 .46302 -1.7347 .49900 -1.2955 .56497 -.73180 .63095 -.39652 .69692 -.20718 .76290 -.10489 .82887 -.05166 .89484 -.02486
1.0268 -.00544 / at 0.3 -8.5 units < 4cm,0.2 cm > point at -1.15 -15 x from 0 to 1.0 , y from -15 to 15 bottom ticks in numbered from 0 to 1.0 by 0.5 unlabeled short quantity 11 / left ticks in unlabeled short quantity 13 / top ticks in unlabeled short quantity 11 / right ticks in unlabeled short quantity 13 / from 0 0 to 1 0 at 0.75 -17.5 at 0.5 -20 .00120 9.8172 .04318 9.5371 .04918 9.4243 .05517 9.4095 .06117 9.3851 .06717 9.3524 .07317 9.3124 .08516 9.2136 .09716 9.0927 .10916 8.9527 .12115 8.7955 .13315 8.6231 .14514 8.4372 .15714 8.2392 .16913 8.0307 .20512 7.3559 .24110 6.6351 .27709 5.9004 .31308 5.1787 .34906 4.4908 .38505 3.8516 .42103 3.2704 .45702 2.7517 .49300 2.2963 .52899 1.9020 .59496 1.3238 .66094 .90427 .72691 .60827 .79288 .40407 .85886 .26575 .92483 .17340 .99080 .11246 1.0568 .07260 / at 0.4 5.5 .00120 -6.3616 .04318 -6.2120 .04918 -6.1111 .05517 -6.1027 .06117 -6.0880 .06717 -6.0681 .07317 -6.0436 .08516 -5.9826 .09716 -5.9077 .10916 -5.8205 .12115 -5.7225 .13315 -5.6148 .14514 -5.4983 .15714 -5.3740 .16913 -5.2429 .20512 -4.8166 .24110 -4.3583 .27709 -3.8879 .31308 -3.4225 .34906 -2.9758 .38505 -2.5579 .42103 -2.1756 .45702 -1.8324 .49300 -1.5294 .52899 -1.2661 .59496 -.87815 .66094 -.59579 .72691 -.39665 .79288 -.25983 .85886 -.16789 .92483 -.10725 .99080 -.06787 1.0568 -.04262 / .00120 8.6307 .04918 8.4075 .05517 8.3504 .07317 8.2363 .09116 8.1247 .10916 7.9766 .12715 7.7979 .14514 7.5931 .16314 7.3658 .18113 7.1195 .19912 6.8574 .21711 6.5827 .23511 6.2982 .27109 5.7115 .30708 5.1181 .34306 4.5357 .37905 3.9783 .41503 3.4561 .45102 2.9760 .48701 2.5418 .52299 2.1548 .52899 2.0949 .56497 1.7619 .60096 1.4727 .63695 1.2241 .67293 1.0122 .70892 .83314 .74490 .68289 .78089 .55763 .87685 .31985 1.0088 .14495 / .00120 -7.3999 .04918 -7.2171 .05517 -7.1595 .07317 -7.0490 .09116 -6.9466 .10916 -6.8115 .12715 -6.6494 .14514 -6.4641 .16314 -6.2590 .18113
-6.0373 .19912 -5.8021 .21711 -5.5560 .23511 -5.3020 .27109 -4.7805 .30708 -4.2565 .34306 -3.7456 .37905 -3.2603 .41503 -2.8092 .45102 -2.3979 .48701 -2.0291 .52299 -1.7033 .52899 -1.6531 .56497 -1.3759 .60096 -1.1374 .63695 -.93444 .67293 -.76331 .70892 -.62024 .74490 -.50155 .78089 -.40378 .87685 -.22224
1.0088 -.09438 / at 0.4 -4.5 units < 4cm,0.2 cm > point at -2.3 -15 x from 0 to 1.0 , y from -15 to 15 bottom ticks in numbered from 0 to 1.0 by 0.5 unlabeled short quantity 11 / left ticks in unlabeled short quantity 13 / top ticks in unlabeled short quantity 11 / right ticks in unlabeled short quantity 13 / from 0 0 to 1 0 at 0.75 -17.5 at 0.5 -20 .00120 6.4005 .04218 6.3068 .06117 6.1711 .06717 6.1606 .07317 6.1463 .07916 6.1287 .09116 6.0846 .10316 6.0303 .11515 5.9671 .12715 5.8959 .13914 5.8175 .15114 5.7325 .16314 5.6417 .17513 5.5455 .18713 5.4445 .19912 5.3394 .23511 5.0035 .27109 4.6473 .30708 4.2822 .34306 3.9180 .37905 3.5626 .41503 3.2220 .45102 2.9004 .48701 2.6005 .52299 2.3238 .55898 2.0707 .62495 1.6666 .69092 1.3343 .75690 1.0648 .82287 .84835 .88884 .67556 .95482 .53821 1.0208 .42926 / at 0.5 4 .00120 -2.2141 .04218 -2.1706 .06117 -2.1369 .06717 -2.1337 .07317 -2.1292 .07916 -2.1235 .09116 -2.1093 .10316 -2.0917 .11515 -2.0711 .12715 -2.0477 .13914 -2.0219 .15114 -1.9939 .16314 -1.9639 .17513 -1.9320 .18713 -1.8984 .19912 -1.8633 .23511 -1.7505 .27109 -1.6298 .30708 -1.5049 .34306 -1.3790 .37905 -1.2550 .41503 -1.1349 .45102 -1.0204 .48701 -.91276 .52299 -.81261 .55898 -.72034 .62495 -.57180 .69092 -.44880 .75690 -.34888 .82287 -.26897 .88884 -.20592 .95482
-.15673 1.0208 -.11874 / at 0.5 -2.5 .00120 4.5936 .05517 4.5026 .06117 4.4527 .08516 4.4296 .11515 4.3648 .15114 4.2501 .18713 4.1042 .22311 3.9340 .25910 3.7455 .29508 3.5441 .33107 3.3349 .36705 3.1223 .40304 2.9101 .46901 2.5322 .53499 2.1801 .60096 1.8619 .66693 1.5808 .73291 1.3367 .79888 1.1274 .86485 .94971 .93083 .79984 .99680 .67399 1.0628 .56859 / .00120 -2.8501 .05517 -2.8066 .06117 -2.7553 .08516 -2.7339 .11515 -2.6820 .15114 -2.5931 .18713 -2.4821 .22311 -2.3542 .25910 -2.2142 .29508 -2.0666 .33107 -1.9153 .36705 -1.7636 .40304 -1.6144 .46901 -1.3546 .53499 -1.1197 .60096 -.91403 .66693 -.73841 .73291 -.59130 .79888 -.47002 .86485 -.37131 .93083 -.29185 .99680 -.22847 1.0628 -.17830 / units < 1cm,1 cm > point at 3 2 x from 0 to 15 , y from 0 to 8 at 7.5 0.5 units < 9.0cm,12.0 cm > point at -0 . 0 .
x from 0 to 1.0 , y from 0 to 0.5 bottom ticks in numbered from 0 to 1.0 by 0.5 unlabeled short quantity 21 / right ticks in numbered from 0 to 0.5 by 0.1 unlabeled short quantity 21 / at 0.75 -0.025 at 1.0 0.525 [ l ] at 0 0.525 from 0 0.14 to 1 0.14 at 0.75 0.37 at 0.5 0.115 ( * @xmath249 * ) span < 2pt > 0.000 0.046 0.072 0.125 0.078 0.108 0.250 0.114 0.148 0.375 0.154 0.192 0.500 0.198 0.238 0.625 0.244 0.286 0.750 0.291 0.335 0.875 0.339 0.383 1.000 0.387 0.429 / units < 9.0cm,15 cm > point at -0 .
1.2 x from 0 to 1.0 , y from 1.2 to 1.6 top ticks in unlabeled short from 0.05 to 0.95 by 0.05 / left ticks in numbered from 1.2 to 1.6 by 0.1 unlabeled short quantity 21 / from 0 1.417 to 1 1.417 at 0.25 1.470 at 0.12 1.4 0.000 1.458 1.481 0.125 1.438 1.463 0.250 1.416 1.444 0.375 1.392 1.423 0.500 1.366 1.399 0.625 1.339 1.375 0.750 1.309 1.348 0.875 1.279 1.320 1.000
1.248 1.291 / | we take into account the @xmath0 corrections upto @xmath1 order in the heavy - meson - soliton bound state approach for heavy baryons . with these corrections , the mass spectra of baryons with @xmath2-quark as well as of those with @xmath3-quark
are well reproduced . for charmed baryons
, however , the correction to the mass spectra amounts to about 300 mev , which is not small compared to the leading order binding energy , @xmath4 mev . |
since the pioneering work of shapley ( 1918 ) the properties of the galactic globular cluster system have been seen as valuable tracers of the structure and history of the galactic halo ( van den bergh 1995 ) .
zinn ( 1985 ) , armandroff & zinn ( 1988 ) , and armandroff ( 1989 ) , on the basis of correlations between metallicity and kinematics , divided the globular cluster population of the galaxy into a halo and a disk system , the clusters of which have metallicities of [ fe / h ] @xmath6 and [ fe / h ] @xmath0 , respectively .
this has lead to a model in which the metal - richest globular clusters belong to a kinematically uniform disk system .
this letter presents a new analysis of the kinematical and spatial properties of the highest - metallicity ( [ fe / h ] @xmath0 ) galactic globular clusters which shows that they may not constitute a homogeneous population . instead ,
if these clusters are grouped on the basis of their mass , several subsystems can be identified .
we have taken advantage of the april 1996 version of the _ catalog of parameters for milky way globular clusters _ compiled by harris ( 1996 ) .
this was obtained through the world wide web from the site ( url ) http://www.physics.mcmaster.ca/globular.html .
our kinematical analysis follows the lines of frenk & white ( 1980 , 1982 ) , zinn ( 1985 ) , and armandroff ( 1989 ) , but our approach differs from theirs in that we divide the metal - richest clusters into two subgroups according to cluster mass .
throughout this paper a standard @xmath7 galactocentric coordinate system is adopted ( cf .
mihalas & binney 1981 ) .
the solar motion is taken to be in the direction @xmath8 57 , @xmath9 23 , with a velocity @xmath10 km / s ( zinn 1985 ) .
the three components of the solar velocity relative to the local standard of rest ( lsr ) are then @xmath11 km / s .
the velocity of a globular cluster with respect to the lsr ( @xmath12 ) is computed from the above components of the solar motion and the heliocentric radial velocity of the cluster ( mihalas & binney 1981 ) . the radial velocity @xmath13 of a cluster with respect to a _ stationary _ observer at the location of the sun is @xmath14 where @xmath15 is the angle between the apex of the lsr and the cluster s position on the sky , @xmath16 and the rotational velocity of the lsr is assumed to be 220 km / s .
analysis of the bulk rotation of a system of globular clusters depends on that system s rotation curve . in the case of a constant systematic rotation speed @xmath17 without a global radial motion ( frenk & white 1980 ) @xmath18 where @xmath19 @xmath20 is the line - of - sight peculiar velocity of a cluster , and @xmath21 is the cluster s galactocentric distance .
the galactocentric distance of the sun is taken to be @xmath22 kpc . in the case of rigid - body rotation with a uniform angular velocity of @xmath23 ( kinman 1959 , zinn 1985 ) @xmath24 an unbiased estimate of @xmath17 can be obtained by the following sum over all clusters ( frenk & white 1980 ) : @xmath25 with a similar equation for @xmath26 and @xmath27 .
the line - of - sight velocity dispersion @xmath28 of the sample is determined by inserting @xmath17 or @xmath26 into equations 1 and 3 , respectively , and calculating the dispersion in @xmath20 .
globular cluster masses @xmath29 were calculated from the integrated absolute visual magnitudes ( @xmath30 ) tabulated by harris ( 1996 ) . a mass - to - light ratio of @xmath31 was assumed ( chernoff & djorgovski 1989 ) .
in this section we divide the metal - richest galactic globular clusters , which we take to be those with [ fe / h ] @xmath0 , into two mass groups : a _ high - mass _ group for which @xmath32 ( @xmath33 ) , and a _ lower - mass _ group , containing clusters of mass smaller than this .
the kinematical properties of these two groups are summarised in columns 2 and 3 of table 1 .
the parameters listed include the number of clusters in each group @xmath34 , the rotation speed @xmath17 for the case of a flat rotation curve and the value of @xmath35 for the case of rigid - body rotation , as well as the standard deviations in these values .
also given are the line - of - sight velocity dispersions @xmath28 , the mean height above the galactic plane @xmath36 , and the mean radial distance in the galactic plane @xmath37 for each group .
with the caveat that the number of clusters is relatively modest , the entries in table 1 indicate that the kinematical properties of the high - mass and lower - mass clusters are different . while the high - mass clusters show little or no rotation , the low - mass clusters exhibit rotation with a significance level of @xmath38 .
these results suggest that the high - mass clusters define a spheroidal inner - halo - like subsystem whereas the low - mass clusters represent a disk - like component .
kinematical and spatial data for the lower - mass clusters are plotted in the upper and lower panels , respectively , of figure 1 .
least - squares fits to the velocity data for the entire set of lower - mass clusters are shown as short - dashed lines in the upper panels .
two groups of clusters can be distinguished in the top panels of figure 1 .
one group ( open circles ) exhibits a relatively well - defined relationship between @xmath13 and both @xmath27 and @xmath40 .
the remaining clusters are all within a small @xmath27 range ( @xmath41 ) and are represented by filled circles .
these two groups are also relatively distinct in the lower panels of figure 1 . with the exception of ter 7 which is located at galactocentric
coordinates @xmath42 kpc and probably belongs to the sagittarius dwarf spheroidal galaxy ( da costa & armandroff 1995 ) , the group with small @xmath27 falls within a centrally concentrated , bar - like configuration that is extended in the @xmath43 direction and has @xmath44 ; the clusters in this group will be referred to as `` bar clusters . '' the remaining clusters ( open circles ) fall within a ring around the galactic center having an inner and an outer radius of @xmath45 4 and 6 kpc respectively ; these objects will be referred to as the `` 5-kpc - ring '' clusters .
a bar - like feature in the space distribution of globular clusters was noted by woltjer ( 1975 ) and harris ( 1976 ) , who suggested that it could be an artifact of distance measurement errors for clusters in directions close to that of the galactic center .
frenk & white ( 1980 ) performed monte carlo simulations to assess this possibility .
they found that errors in distance moduli would have to be as large as 1.0 mag to explain the bar . whereas the accuracy of distance modulus measurements have increased since the time of woltjer s ( 1975 ) analysis
, the bar is more prominent in figure 1 than in the data of woltjer , the opposite to what would be expected if the bar were an observational artifact .
note that there is also strong evidence for a bar - like distribution of stars and molecular gas in the galactic bulge ( see the review by freeman 1996 ) .
it is striking that the stellar bar has a very similar orientation , spatial distribution , and rotational velocity to the subsystem of `` bar clusters '' , indicating a common origin .
separate least - squares fits were made to the velocity data for the subset of `` bar '' clusters and for the subset of `` 5-kpc - ring '' clusters .
the results are given in columns 4 and 5 of table 1 , and the fits for the `` 5-kpc - ring '' subsystem are shown as long - dashed lines in figure 1 .
two points deserve emphasis . _
( i ) _ the small range of @xmath27 for the _ bar clusters _ leads to large errors in @xmath35 if one assumes solid - body rotation .
nevertheless , the value of @xmath46 km / s / kpc for these clusters is in excellent agreement with zhao s ( 1994 ) estimated angular velocity of the central stellar bar of @xmath47 km / s / kpc .
the assumption of rigid body rotation might therefore be the physically appropriate model for the `` bar clusters . ''
it is interesting that all the clusters of the bar system have positive @xmath13 values .
this bulk flow may indicate that the bar is a transient phenomenon .
_ ( ii ) _ the _ 5-kpc - ring _
clusters constitute a rapidly rotating subsystem , with only a small velocity dispersion .
these clusters seem to be members of the galactic disk .
the rotation speed at @xmath48 kpc , derived from the solid - body assumption ( upper - right panel of figure 1 ) is 120 km / s , which compares with 164 km / s under the assumption of a flat rotation curve .
in addition , @xmath49 , indicative of a disk with well - organised rotation .
this subgroup of `` 5-kpc - ring '' clusters has kinematics not greatly different from the old open cluster system of the galactic disk , which has a scale height of 375 pc ( janes & phelps 1994 ) and kinematics of @xmath50 km / s and @xmath51 km / s ( scott , friel , & janes 1995 ) .
the difference in the rotation speeds of these two cluster systems may be partly related to the differences in their average age and galactocentric distance . the lower - mass `` 5-kpc - ring '' globular clusters might represent `` missing links '' between the globular clusters of the halo and the open clusters of the galactic disk . as illustrated in figure 3 of hufnagel & smith ( 1994 ) , for example , no old open clusters are known with galactocentric distances of less than 7 kpc .
it may be that old open clusters did not form within 7 kpc of the galactic center ; and that more populous globular - like clusters , like m71 , were instead formed in this region of the galaxy .
alternatively , open clusters that formed at @xmath52 kpc may have been disrupted by interactions with molecular clouds .
the objects more likely to survive this process would have been the higher - mass globular clusters . the positional and kinematical data for those globular clusters having @xmath32 and [ fe / h ] @xmath0 are shown in figure 2 .
the fact that @xmath53 indicates that the high - mass , metal - richest globulars are not members of a true rotating disk system .
the formal solution of @xmath54 km / s is heavily weighted by the cluster ngc 104 ( 47 tuc ) with @xmath55 and @xmath56 km / s ; excluding this object would lead to counter - rotation with @xmath57 km / s for the remaining sample of clusters , further strengthening the conclusion that these clusters are not part of a rotating disk population . rather than being members of the disk
, these clusters may instead belong to a metal - enriched inner - halo component .
the velocity dispersion among these clusters is not as large as for the system of more metal - poor galactic halo globular clusters .
this result is in agreement with frenk & white ( 1980 ) who found that the observed line - of - sight velocity dispersion of the halo cluster population increases significantly with galactocentric distance , indicating a non - isothermal cluster distribution function .
the high - mass relatively - high - metallicity clusters might therefore be part of a more centrally condensed and kinematically cooler inner component of the galactic halo .
formed at times when the interstellar medium in the proto - galaxy had experienced dissipational contraction and the metallicity had increased to relatively high values , these clusters may represent the last remnants from the true epoch of globular cluster formation in the galactic halo .
a statistical investigation of the difference between the low - mass and high - mass clusters depicted in figures 1 and 2 , respectively , is affected by the low number ( 7 ) of high - mass clusters . to take account of this limitation
, monte carlo simulations were performed in which random samples of seven clusters were generated from a population with given average rotational velocity @xmath58 and velocity dispersion @xmath59 similar to that of the low - mass clusters . for each sample
generated , the average rotational velocity @xmath17 was calculated using equation 4 .
the results for a large number of artificial samples were combined to give the probability that the rotation speed for such a sample is @xmath60 km / s , corresponding to rotation that is as slow as observed for the high - mass system of clusters .
the data for each set of seven clusters were generated as follows : first , a value for @xmath40 was chosen , randomly distributed between @xmath61 and + 1 .
the cluster velocities @xmath13 were then generated assuming a normal probability distribution function : @xmath62 ^ 2 } { 2 \ \sigma_{pop}^2 } \right)\ ] ] with values of @xmath58 and @xmath59 chosen to be representative of the low - mass system of clusters . using this method
, we find that there exists a 10% probability that the high - mass clusters are drawn from the same kinematical population as the combined sample of low - mass clusters .
comparing the high - mass clusters and the low - mass bar clusters , we find a 51% probability that these two samples have similar kinematical properties . on the other hand ,
the velocity distributions of high - mass clusters and low - mass ring clusters are significantly different , with a probability of less than @xmath63 that both samples result from the same population . in summary
, we can rule out with a high probability that the high - mass clusters have an average rotational velocity and a velocity dispersion similar to the ring clusters .
they appear to be two distinct subsystems .
the relation of the bar clusters to the high - mass clusters is not clear from the present analysis .
the similar kinematical properties of both subsystems ( assuming a constant rotational velocity ) suggests that they may be drawn from the same or related populations .
however , the more extended spatial distribution of the high - mass clusters indicates that this subsystem is part of the inner halo .
the origin of the bar clusters , on the other hand , is more likely related to the formation of the central stellar bar or bulge .
the formation of the inner halo and galactic bulge / bar could also be coupled .
we propose that the following three subsystems can be distinguished among the highest - metallicity ( [ fe / h ] @xmath0 ) globular clusters of the galaxy : _ i ) a centrally condensed halo subsystem of high - mass clusters , ii ) a bar subsystem of lower - mass clusters , and iii ) a rapidly - rotating disk subsystem of lower - mass clusters , many of which fall within a ring about the galactic center of radius 4 - 6 kpc . _
these subsystems indicate that the population of relatively - high - metallicity globular clusters is more heterogeneous than has hitherto been expected from the concept of a single disk population .
if the high - mass , higher - metallicity globular clusters really are members of the galactic halo as suggested above , then it follows that high - mass ( @xmath64 @xmath65 ) cluster formation only took place in the galactic halo , not in the disk .
this suggests that the maximum mass of star clusters which formed in the galaxy may have decreased with time , perhaps along the lines of the following scenario .
the halo supported massive cluster formation , possibly under starburst - like conditions in massive clouds ( murray & lin 1989 ; harris & pudritz 1994 ; brown et al . 1995 ; burkert et al .
however , as the interstellar medium collapsed into the center and the equatorial plane and the metallicity built up , cluster formation tended to produce the low or intermediate - mass globular clusters ( @xmath66 ) found in the bar or the 4 - 6 kpc ring noted in figure 1 .
as the disk built up further from the inside out , the sites of cluster formation propagated outwards through the disk into regions beyond 7 kpc from the galactic center .
the typical masses of clusters being formed at this time were markedly lower , with open clusters like ngc 6791 and ngc 188 being produced .
armandroff , t. e. , & zinn , r. 1988 , , 96 , 92 armandroff , t. e. 1989 , , 97 , 375 brown , j. h. , burkert , a. , & truran , j. w. 1995 , , 440 , 666 burkert , a. , brown , j. h. , & truran , j. w. 1996 , in the history of the milky way and its satellite system , eds . a. burkert , d. hartmann , & s. r. majewski , in press chernoff , d. f. , & djorgovski , s. 1989 , , 339 , 904 da costa , g. s. , & armandroff , t. e. 1995 , , 109 , 2533 freeman , k. c. 1996 , in iau colloq .
157 , barred galaxies , aspcs 91 , eds .
r. buta , d. a. crocker , & b. g. elmegreen ( san francisco : astr .
soc . pac . ) , 1 frenk , c. s. , & white , s. d. m. 1980 , , 193 , 295 frenk , c. s. , & white , s. d. m. 1982 , , 198 , 173 harris , w. e. 1976 , , 81 , 1095 harris , w. h. , & pudritz , r. e. 1994 , , 429 , 177 harris , w. e. 1996 , the catalog of parameters for milky way globular clusters hufnagel , b. , & smith , g. h. 1994 , , 106 , 1068 janes , k. a. , & phelps , r. j. 1994 , , 108 , 1773 kinman , t. d. 1959 , , 119 , 157 mihalas , d. , & binney , j. 1981 , galactic astronomy ( san francisco : freeman & co ) .
murray , s. d. , & lin , d. n. c. 1989 , , 339 , 933 scott , j. e. , friel , e. d. , & janes , k. a. 1995 , , 109 , 1706 shapley , h. 1918 , , 48 , 154 van den bergh , s. 1995 , , 110 , 1171 woltjer , l. 1975 , , 42 , 109 zhao , h. 1994 , phd thesis , columbia university zinn , r. 1985 , , 293 , 424 lcccc n & 7 & 21 & 11 & 10 @xmath17 [ km / s ] & @xmath67 & @xmath68 & @xmath69 & @xmath70 @xmath23 [ km / s / kpc ] & @xmath71 & @xmath72 & @xmath73 & @xmath74 @xmath28 [ km / s ] ( flat ) & 70.4 & 101.2 & 62.7 & 26.0 @xmath28 [ km / s ] ( rigid ) & 55.7 & 98.6 & 55.3 & 22.8 @xmath36 [ kpc ] & 0.45 & 0.22 & 0.33 & 0.31 @xmath37 [ kpc ] & 3.49 & 3.18 & 2.18 & 4.27 | an analysis of the kinematical and spatial properties of the highest - metallicity globular clusters in the galaxy , having metallicities of [ fe / h ] @xmath0 , indicates that these objects do not comprise a homogeneous population .
three subsystems are identified among these clusters . + _
i ) _ the highest - mass clusters with @xmath1 exhibit a very slow net rotation with a speed of @xmath2 km / s and @xmath3 , indicative of a a centrally condensed , relatively - high - metallicity subsystem . + _
ii ) _ roughly half of the lower - mass clusters appear to be located in an elongated bar - like structure which passes through the galactic center , and has similar properties to the central stellar bar of the milky way .
+ _ iii ) _ the remaining lower - mass clusters exhibit very rapid net rotation , with a rotation speed of @xmath4 km / s and @xmath5 .
these clusters are located in the galactic plane , within a ring of 4 to 6 kpc radial distance from the galactic center .
the highest - mass clusters may have formed during relatively advanced stages of the dissipative evolution of the inner galactic halo .
although the lower - mass bar clusters have kinematical properties which are similar to the highest - mass clusters , their spatial distribution suggests that they may be associated with the formation of the galactic stellar bar or bulge .
the lower - mass ring clusters appear to be real disk objects .
they may represent a stage in cluster formation that was intermediate between that of the halo globular clusters and the oldest extant open clusters . |
for a riemannian metric @xmath2 on a manifold @xmath3 with strictly convex boundary @xmath4 , the x - ray transform @xmath5 of a function @xmath6 is the collection of integrals over all finite length geodesics with endpoints on @xmath4 . for a domain in @xmath7 equipped with the euclidean metric ,
this is nothing more than the radon transform .
a natural inverse problem is to recover @xmath8 from @xmath5 , ie . to find an inversion procedure for @xmath0 .
in the euclidean case , radon found an inversion formula @xcite , and this has been extended to hyperbolic spaces by helgason @xcite and berenstein - casadio tarabusi @xcite .
this problem has been studied more generally for _ simple metrics _ , which are metrics with no conjugate points and such that each geodesic has finite length .
let @xmath9 be the unit tangent bundle of @xmath3 and @xmath10 the canonical projection @xmath11 ; if @xmath12 denotes the inward unit pointing vector field to @xmath4 , we define @xmath13 the inward ( - ) and outward ( + ) boundaries .
then if @xmath14 denotes the geodesic flow on @xmath15 at time @xmath16 and if @xmath2 is simple , we can view @xmath0 as the map @xmath17 where @xmath18 is the length of the geodesic @xmath19 relating boundary points of @xmath3 .
the injectivity of @xmath0 for simple metrics has been proved by mukhometov @xcite and anikonov @xcite ( see also the book by sharafutdinov @xcite ) .
more recently pestov - uhlmann @xcite gave an approximate inversion formula in dimension @xmath20 for simple metrics , which is exact when the curvature of @xmath2 is constant .
krishnan @xcite then showed that the pestov - uhlmann formula can be made exact in a small enough @xmath21-neighborhood of a given simple metric @xmath22 with constant curvature , but the bound is not explicit and depends for instance on the diameter of @xmath23 .
the pestov - uhlmann approach , exploiting the interplay between transport equations on @xmath15 and the fiberwise hilbert transform , was used by the second author to derive further inversion formulas for certain types of weighted x - ray transforms on simple surfaces @xcite .
manifolds with geodesics having infinite length are called _ manifolds with trapping_. for instance any negatively curved riemannian manifold with strictly convex boundary which have non contractible loops have closed geodesics and thus geodesics with infinite lengths . in general
we call @xmath24 the set of points @xmath25 such that @xmath26 belongs to the interior @xmath27 of @xmath15 for all @xmath28 , which means that the geodesic @xmath29 has infinite length .
the trapped set @xmath30 is defined by @xmath31 and consists of geodesics never touching the boundary . when @xmath2 has negative curvature , @xmath32 is a hyperbolic set for the flow in the sense of and @xmath2 has no conjugate points . in the recent work @xcite
, the first author proved that for a class of manifolds including those with negative curvature ( see for the precise assumptions ) , the @xmath33-ray transform @xmath5 can be defined by the expression outside @xmath34 and @xmath0 extends as a bounded and injective operator @xmath35 where @xmath36 is some smooth lebesgue type measure on @xmath37 multiplied by a weight vanishing on @xmath38 .
in fact , the same result holds as a map @xmath39 by the boundedness proved in lemma [ i_0i_1 ] below . for @xmath40 , the odd extension of @xmath5
is defined as the @xmath41 element @xmath42 similarly , we can integrate @xmath43-forms along geodesics and we can define the x - ray transform on @xmath43-forms by @xmath44 where @xmath45 is defined by @xmath46 , this also extends as a bounded operator @xmath47 which is injective on the space of divergence free forms .
we let @xmath48 be the adjoint operator when @xmath49 is viewed as a map @xmath50 .
to state our result we need to define two more operators .
let @xmath51 be the hilbert transform in the fibers of @xmath52 ( i.e. multiplication by @xmath53 on the @xmath54 fourier component in the fiber , with @xmath55 ) , and let @xmath56 be the scattering operator of the flow , defined by @xmath57 then we have the following approximate inversion formula extending pestov - uhlmann formula to the setting of manifolds with trapping . [ inversionintro ]
let @xmath58 be a riemannian surface with strictly convex boundary and assume that the trapped set @xmath32 for the flow is hyperbolic , that @xmath2 has no conjugate points ( these conditions are satisfied in negative curvature ) .
then there exists an operator @xmath59 with smooth integral kernel such that for each @xmath60 @xmath61 where @xmath62 is the hodge star operator and @xmath63 the exterior derivative .
we also have an inversion formula for divergence free @xmath43-forms , see theorem [ inversion ] .
we next prove a upper bound on the @xmath64 norm of @xmath65 in a neighborhood of constant negatively curved metrics , showing that @xmath66 can be inverted by the neumann series , and thus providing an exact inversion formula for @xmath0 . in constant curvature ,
the trapped set @xmath32 is a fractal set with hausdorff dimension @xmath67 when @xmath68 ( see @xcite ) .
[ theo2 ] let @xmath23 be a manifold with strictly convex boundary and constant negative curvature @xmath69 and trapped set @xmath32 , and let @xmath70 .
then for each @xmath71 so that @xmath72 , there is a an explicit constant @xmath73 depending only on @xmath74 such that for all metric @xmath2 on @xmath3 with strictly convex boundary and gauss curvature @xmath75 satisfying latexmath:[\[{\lambda}_1^{2}g_0\leq
g\leq { \lambda}_1^{-2 } g_0 , \quad \kappa(x ) \leq -{\lambda}_2 ^ 2\kappa_0 , \quad the remainder operator @xmath65 obtained in theorem [ inversionintro ] has norm @xmath77 and thus the inversion formula allows to recover @xmath60 from @xmath5 by a convergent series .
when @xmath78 , the constant @xmath73 does not depend on @xmath79 .
the constant @xmath73 is given by @xmath80 where @xmath81 and @xmath82 is the constant given in [ bound1 ] .
this shows that there is an explicit neighborhood of the space of constant curvature metrics where the inversion formula can be made exact .
even when @xmath83 the result is new since our bounds on the size of the neighborhood of constant curvature metrics is independent of the volume and the diameter of @xmath58 , which was not the case in the previous works for simple metrics @xcite .
to prove such a bound , we compute on convex co - compact hyperbolic surfaces an operator closely related to @xmath84 in terms of the laplacian on the surface , where @xmath85 is the x - ray transform with a constant attenuation @xmath86 .
this is done in the appendix . finally
, we remark that is is quite possible that helgason type inversion for @xmath0 on @xmath87 could be applied to deal with the case of constant curvature metrics ( passing to the universal cover ) , but that would involve checking that the inversion formula applies to functions which are periodic by a discrete group , thus not tending to @xmath88 at infinity of @xmath87 , and in any case this would not work in variable curvature , unlike our formula . * acknowledgement .
is partially supported by anr grant anr-13-js01 - 0006 and anr-13-bs01 - 0007 - 01 .
fm is partially supported by nsf grant dms-1514820 .
part of this work was done during the research program inverse problems at ihp , paris , spring 2015 .
let @xmath58 be an oriented surface with strictly convex boundary and let @xmath15 be its unit tangent bundle with canonical projection @xmath89 .
we will assume that @xmath90 notice that these assumptions are satisfied if @xmath2 has negative curvature , see @xcite .
we denote by @xmath33 the geodesic vector field on @xmath15 and @xmath14 its flow at time @xmath16 .
recall that hyperbolicity of the flow means that there is a continuous , flow invariant , decomposition on @xmath32 @xmath91 where @xmath92 and @xmath93 are the stable and unstable subspaces satisfying for each @xmath94 @xmath95 some uniform @xmath96 .
let @xmath97 be the vertical vector field on @xmath15 defined by @xmath98 where @xmath99 is the @xmath100 rotation in the fibers of @xmath15 .
locally , in isothermal coordinates @xmath101 on @xmath3 , the metric can be written as @xmath102 for some smooth function @xmath103 , we have associated coordinates @xmath104 on @xmath15 if @xmath105 parametrizes the fibers elements by @xmath106 ; in these coordinates @xmath97 is simply given by @xmath107 .
the tangent space @xmath108 of @xmath15 is spanned by the vector fields @xmath109 where @xmath110 $ ] ; the vectors @xmath111 span the horizontal space @xmath112 , while @xmath97 spans the vertical space @xmath113 in @xmath108 .
the sasaki metric @xmath114 on @xmath15 is the metric such that @xmath115 is orthonormal .
the liouville measure @xmath116 on @xmath15 is the riemannian measure of the metric @xmath114 and we denote by @xmath117 the measure induced by @xmath114 on the boundary @xmath52 of @xmath15 .
we define @xmath118 and @xmath119 the @xmath120 spaces with respect to @xmath116 and @xmath117 .
there is another natural measure @xmath121 on @xmath52 which is given by @xmath122 where @xmath12 is the inward pointing normal vector field to @xmath4 .
let us also define the incoming ( - ) , outgoing ( + ) and glancing ( 0 ) boundaries of @xmath15 @xmath123 define @xmath124 to be a small extension of @xmath3 with no conjugate points and with strictly convex boundary , and so that each geodesic with initial point @xmath125 goes in finite time to @xmath126 ; the existence of @xmath127 is proved in section 2.1 and lemma 2.3 of @xcite .
we keep the notation @xmath33 for the vector field of the geodesic flow on @xmath128 .
the liouville measure @xmath116 extends to @xmath128 . for each point @xmath129 ,
define the time of escape of @xmath15 in positive ( + ) and negative ( - ) time : @xmath130,\\ \ell_- ( x , v)=\inf\ , \ { t\leq 0 ; \varphi_{t}(x , v)\in sm\}\subset [ -\infty,0 ] \end{gathered}\ ] ] and the incoming ( - ) and outgoing ( + ) tail are defined by @xmath131 we define @xmath132 as well on @xmath128 in the same way , so that they extend those on @xmath15 , and they are closed set in @xmath128 are we keep the same notation @xmath132 for both @xmath15 and @xmath128 .
the trapped set is @xmath133 and is invariant by the flow .
we can define the extended stable bundle @xmath134 over @xmath135 satisfying @xmath136 and @xmath137 for some uniform @xmath96 ; the extended unstable bundle @xmath138 over @xmath139 is defined similarly by taking negative times instead of @xmath28 .
the bundles @xmath140 and @xmath141 are defined by @xmath142 it follows from @xcite that under the assumptions we have ( see ( * ? ? ?
2.4 ) for details ) @xmath143 following @xcite , there is an orthogonal decomposition corresponding to fourier series in the fibers @xmath144 where @xmath145 is the space of @xmath146 sections of a complex line bundle over @xmath147 .
similarly , one has a decomposition on @xmath52 @xmath148 using fourier analysis in the fibers of the circle bundle .
there is a canonical map @xmath149 and in particular @xmath150 is just the pull - back by the projection @xmath151 .
we define the dual map @xmath152 on distributions @xmath153 by @xmath154 where the distribution pairing is @xmath155 when @xmath156 .
we notice that @xmath157 and it is easily checked that @xmath158 are the orthogonal projection from @xmath159 to respectively @xmath160 and @xmath161 .
the hilbert transform in the fibers is defined by using the decomposition : @xmath162 with @xmath55 by convention .
we can extend continuously @xmath51 to @xmath163 by the expression @xmath164 similarly , we define the hilbert transform in the fibers on @xmath52 @xmath165 and its extension to distributions as for @xmath128 .
we note that @xmath51 extends as a bounded operator on @xmath159 and @xmath166 as a bounded operator on @xmath167 . for smooth @xmath168
we have that @xmath169 and the identity extends by continuity to the space of distributions in @xmath170 with wave - front set disjoint from @xmath171 since , by ( * ? ? ?
8.2.4 ) , the restriction map @xmath172 obtained by pull - back through the inclusion map
@xmath173 extends continuously to the space of distributions on @xmath170 with wavefront set not intersecting @xmath171 .
the following relation between hilbert transform and flow was proved by pestov - uhlmann ( * ? ? ?
1.5 ) ( it holds on any surface ) : @xmath174w = x_\perp w_0+(x_\perp w)_0\ ] ] where @xmath175 for @xmath176 .
we can use the odd / even decomposition of distributions with respect to the involution @xmath177 on @xmath128 , @xmath15 and @xmath52 : for instance @xmath33 maps odd distributions to even distributions and conversely , @xmath51 maps odd ( resp .
even ) distributions to odd ( resp .
even ) distributions .
we set @xmath178 and @xmath179 .
we write similarly @xmath180 and @xmath181 for the hilbert transform on ( open sets of ) @xmath52 and the relation also holds with @xmath182 replacing @xmath183 if @xmath184 is even .
taking the odd part of , for any @xmath176 @xmath185 for convenience , we extend @xmath127 to a larger with boundary manifold @xmath186 and extend smoothly @xmath33 to @xmath187 in a way that the flow of @xmath33 in @xmath187 is complete and for each @xmath188 , if @xmath189 for some @xmath190 , then @xmath191 for all @xmath192 ; we refer to ( * ? ? ?
* section 2.1 ) .
let @xmath116 be the liouville measure on @xmath127 , which we extend smoothly to @xmath187 in any fashion .
the invariance of liouville measure by the flow in @xmath128 is satisfied , ie .
@xmath193 , so that @xmath33 is formally skew - adjoint operator when acting on @xmath194 . by (
* section 4.2 ) , under the assumptions , for each @xmath195 , the boundary value problem @xmath196 has a unique solution in @xmath197 given by @xmath198 with @xmath199 the operator @xmath200 defined by is bounded and the expression actually extends as a bounded map @xmath201 as well , see ( * ? ? ?
* proposition 4.2 ) .
it has the following mapping properties @xmath202 where @xmath203 is defined by .
define the _ scattering map _ @xmath204 by @xmath205 by ( * ? ? ?
* lemma 3.4 ) the following operator is unitary @xmath206 the _ x - ray transform _
@xmath207 is defined by @xmath208 from ( * ? ? ?
* lemma 5.1 ) and ( * ? ? ?
2.4 ) , we have that , under the assumptions , it extends as a bounded operator @xmath209 we denote by @xmath210 its odd continuation defined by @xmath211 and similarly we define the even continuation by @xmath212 for @xmath213 .
the operator @xmath214 for all @xmath215 is the adjoint of @xmath216 , it is obtained as follows : for @xmath217 , @xmath218 is the unique @xmath219 solution of @xmath220 by this we mean that @xmath184 has an extension in a small neighborhood @xmath221 of @xmath222 as an invariant distribution and the restriction @xmath223 makes sense as a distribution since , by elliptic regularity , @xmath184 has wave - front set @xmath224 which does not intersect the conormal @xmath225 , as @xmath226 over @xmath37 .
a consequence of this is that the unique @xmath219 solution of @xmath227 , @xmath228 with @xmath229 and @xmath230 is given by @xmath231 since it will be convenient to apply the hilbert transform on @xmath167 , we first show [ l^2 ] under the assumptions , the x - ray transform is bounded as a map @xmath232 there exists @xmath233 small such that the following operator is bounded @xmath234 it suffices to consider @xmath235 as the part on @xmath236 is clearly the same . by proposition 2.4 in @xcite , we have @xmath237 for some @xmath238 if @xmath239 this directly implies that for all @xmath240 , we have @xmath241 . by strict convexity of @xmath4 , there is @xmath233 small such that if @xmath242 , one has @xmath243 for some uniform @xmath244 , see ( * ? ? ? * lemma 4.1.2 ) for example .
we then get @xmath245 using hlder inequality and santalo formula , we get for @xmath246 and @xmath247 , if @xmath248 and @xmath249 @xmath250 where @xmath251 .
this concludes the proof of the first statement .
the boundedness of @xmath252 on @xmath146 is direct from the proof above by taking @xmath253 and using @xmath254 on @xmath37 if @xmath255 for @xmath233 small enough .
finally the x - ray transform on functions @xmath0 and on 1-forms @xmath49 are defined as the bounded operators , for any @xmath247 , @xmath256 and @xmath257 for @xmath258 .
in fact , we can show [ i_0l^2 ] under the assumptions , the x - ray transform on functions and on @xmath43-forms are bounded as maps @xmath259 the operators @xmath260 and @xmath261 extend to @xmath262 as smooth pseudo - differential operators of order @xmath263 by ( * ? ? ?
* propositions 5.7 and 5.9 ) , thus one has by using the standard @xmath264 argument that @xmath265 and @xmath266 are bounded .
this implies that @xmath267 is bounded as a map @xmath268 for all @xmath233 fixed , and the same for @xmath269 .
but we also have @xmath270 bounded as a map @xmath268 if @xmath233 is small enough by lemma [ l^2 ] , and the same for @xmath271 .
the proof is complete .
notice however that when we will use the adjoints of @xmath0 or @xmath49 , this will be always adjoint with respect to the space @xmath272 .
let us define the operator @xmath273 we start by proving the following [ regw ] the operator @xmath65 defined in has a smooth schwartz kernel on @xmath274 . if the gauss curvature @xmath75 is a negative constant , then @xmath275 .
we decompose the operator @xmath276 as @xmath277 with @xmath278 where @xmath233 is smaller than the radius of injectivity of the metric . by @xcite ,
the schwartz kernel of @xmath279 is a distribution on @xmath280 with wavefront set @xmath281 where @xmath282 are defined in , @xmath283 is the conormal bundle to the diagonal @xmath284 of @xmath280 and @xmath285 the wave - front set of the schwartz kernel of @xmath286 is @xmath287 thus by the composition rule of wave - front sets given in ( * ? ? ?
* theorem 8.2.14 ) , @xmath288 since a differential operator does not increase wavefront sets , we deduce that @xmath289 the schwartz kernel of @xmath290 is given by @xmath291 if @xmath32 is the kernel of @xmath292 .
using , the same exact arguments as in proposition 5.7 of @xcite show that the push - forward @xmath291 is smooth outside the submanifold @xmath293 if @xmath2 has no conjugate points .
we are reduced to analyzing @xmath294 and this is very similar to the case of a simple metric studied in ( * ? ? ?
* prop 5.1 ) .
we denote by @xmath295 the jacobi field along the geodesic @xmath296 satisfying @xmath297 and @xmath298 where @xmath299 is the rotation of @xmath300 in the fibers of @xmath15 , and let @xmath301 be the jacobi field satisfying @xmath302 and @xmath303 .
they satisfy @xmath304 where we used the splitting @xmath305 .
the vector field @xmath306 satisfy @xmath307 @xmath308 if @xmath309 and @xmath310 is the gaussian curvature .
we can then write @xmath311 and using the change of variable @xmath312 satisfying @xmath313 , we deduce that the schwartz kernel of @xmath314 is given by @xmath315}(d_g(x , y ) ) \big[\frac{1}{b(x , v , t)}v\big(\frac{a(x , v , t)}{b(x , v , t)}\big)\big]\big|_{tv=\exp_{x}^{-1}(y ) } .\ ] ] clearly this kernel is smooth outside @xmath316 .
we now study the singularity at @xmath317 , which is equivalent to @xmath318 in the @xmath319 coordinates . using @xmath320 and the same for @xmath321
, we make an expansion of @xmath322 at @xmath318 @xmath323 thus we get near @xmath318 @xmath324 which shows that the kernel extends continuously to the diagonal .
let us show that @xmath314 extends smoothly to the diagonal @xmath317 . for this
, note that for each @xmath325 the function @xmath326 is smooth in the variable @xmath327 in @xmath328 $ ] where @xmath329 is a neighborhood of @xmath330 in @xmath15 .
then it suffices to show that for each @xmath331 the taylor expansion at @xmath318 of @xmath326 to any order @xmath332 satisfies @xmath333 where @xmath334 are smooth symmetric tensors of order @xmath54 . for this purpose
, we have @xmath335 as @xmath336 where @xmath337 is the symmetrized covariant derivative . then , writing @xmath338 as @xmath336 , we get for each @xmath339 @xmath340 from this we deduce by a direct induction that @xmath341 is the restriction of an element @xmath342 to @xmath15 .
the same argument applies to @xmath343 . therefore using that @xmath344 if @xmath345 , we see that has the desired expansion , showing that @xmath314 extends smoothly to the diagonal . since @xmath233 was arbitrary , we have that @xmath314 has smooth kernel .
let us show that @xmath275 if @xmath346 is constant .
for @xmath347 with @xmath244 large , we define the operator @xmath348 by @xmath349 due to the exponential damping in @xmath16 , it is easy to see that @xmath350 is bounded if @xmath244 is large enough . in @xcite
it is shown that there exists @xmath351 such that @xmath352 admits an analytic extension to @xmath353 if @xmath354 is large enough depending on @xmath79 .
thus @xmath348 also admits an analytic extension to the same half - plane as a map @xmath355 , and we will show it vanishes if @xmath310 is constant .
since the flow is assumed to have no conjugate points and the trapped set is hyperbolic , there is a uniform lower bound @xmath356 for some @xmath233 , and by using gronwall lemma there exists @xmath244 such that @xmath357 for all @xmath358 . using these estimates and reasoning like above
, we have the converging expression for each @xmath359 @xmath360 as a continuous map of @xmath331 if @xmath361 .
now , by , the functions @xmath362 and @xmath363 are constant in @xmath25 if the curvature @xmath310 is constant , thus @xmath364 for all @xmath365 . by analyticity in @xmath366 , it implies that @xmath367 vanishes when @xmath310 is constant . using that @xmath368 on @xmath369 and @xmath370
if @xmath371 is odd with respect to @xmath177 , we obtain that the @xmath146-adjoint of @xmath65 is given by @xmath372 now , we can show that pestov - uhlmann inversion formula @xcite works also under our assumptions .
[ inversion ] if @xmath58 has strictly convex boundary and assumptions hold , then the following identity holds for each @xmath60 and @xmath373 @xmath374 @xmath375 where @xmath65 is the smoothing operator defined in proposition [ regw ] , @xmath376 is its @xmath146-adjoint , @xmath62 denotes the hodge star operator on @xmath377forms and @xmath378 is defined by .
we follow the proof of ( * ? ? ?
* theorem 5.1 ) .
first using we have as distribution on @xmath27 @xmath379 for any @xmath380 . applying @xmath51 to this identity
, we get by @xmath381 but since @xmath382 as @xmath33 maps odd functions to even functions ( and conversely ) , we obtain @xmath383 the unique @xmath219-solution of @xmath227 , @xmath228 with @xmath229 and @xmath230 is given by @xmath231 .
but @xmath384 by lemma [ i_0l^2 ] and , and @xmath385 by proposition [ regw ] , we apply this to and get @xmath386 notice that we have used that @xmath387 vanishes on @xmath236 since @xmath388 on @xmath37 .
applying @xmath389 we get @xmath390 which gives by using , the identity @xmath391 and @xmath392 . the extension to @xmath393 for @xmath394
is obtained by density and boundedness of each of the operators in the formula on the correct spaces .
notice that a priori @xmath395 if @xmath60 but actually shows it is in @xmath396 .
next we prove the inversion formula for co - exact @xmath43-forms .
let @xmath397 , then since @xmath398 is odd , we get @xmath399 and applying @xmath51 with , this gives @xmath400 by using that @xmath401 on @xmath161 and @xmath402 on @xmath160 .
here @xmath403 and its restriction to @xmath52 is @xmath404 by using lemma [ i_0i_1 ] , thus @xmath405 we apply @xmath406 and obtain @xmath407 by density the same identity holds for @xmath373 .
we notice that a divergence - free @xmath43-form @xmath408 with sobolev regularity @xmath409 can be decomposed under the form @xmath410 where @xmath411 and @xmath412 satisfies @xmath413 and @xmath414 . the formula allows to recover a divergence - free @xmath43 form @xmath415 from its x - ray transform , and the x - ray transform of @xmath416 of @xmath184 at @xmath417 is given purely in terms of the integral of @xmath184 on a curve in @xmath4 with endpoints @xmath331 and @xmath418 and the homology class of the geodesic @xmath419 in @xmath3 , which is a known data .
we now show that in pinched negative curvature the operator @xmath65 has small norm on @xmath420 , allowing to invert @xmath421 by neumann series .
first , we have by ( * ? ? ?
* section 5 ) that @xmath422 and since @xmath423 by ( * ? ? ?
* proposition 4.2 ) if @xmath393 with @xmath424 , we see that the integral @xmath425 is defined for almost every @xmath426 if @xmath393 with @xmath424 , by fubini theorem . in (
* proposition 5.7 ) , it is shown that @xmath260 is the restriction to @xmath3 of an elliptic pseudo - differential operator of order @xmath263 on the extended manifold @xmath262 , thus it maps @xmath420 to @xmath420 for each @xmath427 $ ] .
[ normw ] let @xmath58 be a surface with strictly convex boundary and assume that the gauss curvature @xmath428 of @xmath2 is negative , with @xmath429 .
for each @xmath427 $ ] , the @xmath430 norm of @xmath65 is bounded by @xmath431 let @xmath362 and @xmath363 the functions defined by . if we can prove that @xmath432 uniformly in @xmath358 , then the integral in is convergent for almost every @xmath331 when @xmath433 , for each @xmath434 .
we get in this case that for each @xmath393 with @xmath435 $ ] @xmath436 clearly if @xmath437 , we have @xmath438 and @xmath439 , and both are increasing in @xmath16 . first , by wronskian constancy , we get for all @xmath28 and @xmath129 @xmath440 to simplify notations , we will often drop the @xmath25 dependence below . the function @xmath441 satisfies the riccati equation @xmath442 with @xmath443 and @xmath444 as @xmath445 .
we claim that @xmath446 for all @xmath354 : indeed if it were not the case , there is @xmath233 small and @xmath354 so that @xmath447 .
let @xmath448 be the first time so that this happens , then @xmath449 and thus @xmath450 for all @xmath451 close enough to @xmath452 , leading to a contradiction .
we then get for @xmath453 @xmath454 notice that by sturm comparison theorem , we also have for @xmath455 @xmath456 then , when @xmath457 , @xmath458 exists and is continuous in @xmath25 , moreover for @xmath28 we deduce from @xmath459 in fact we have more generally the same estimate for all @xmath460 : for @xmath461 $ ] @xmath462 where we have denoted @xmath463 .
let us set @xmath464 and @xmath465 .
we have @xmath466 where @xmath467 and @xmath309 .
we notice that @xmath468 . by duhamel formula ,
we obtain @xmath469 which gives @xmath470 we want to give a bound on this quantity .
we see that for @xmath471 , @xmath472 combining with , we deduce the uniform pointwise estimates @xmath473 using and , we obtain @xmath474 and the double integral is a finite constant which , after some calculation turns out to be equal to @xmath475
. we will give an estimate in negative curvature of the norm of the operator @xmath476 by considering the geometry in the universal cover .
we let @xmath477 be the universal cover of @xmath3 , equipped with the pull - back metric @xmath478 of @xmath2 .
since @xmath2 has negative curvature , the manifold @xmath477 is a non - compact simply connected manifold with boundary and the fundamental group @xmath479 of @xmath3 acts properly discontinuously on @xmath477 by isometries with respect to @xmath480 .
denote by @xmath481 the covering map , and by abuse of notation we also call @xmath482 the covering map where @xmath483 is the unit tangent bundle of @xmath484 .
fix a fundamental domain @xmath485 for the action of @xmath486 , then @xmath487 is compact since @xmath3 is compact . if @xmath393 with @xmath488 , let @xmath489 be its lift to @xmath477 ( which is @xmath486-automorphic ) , then we have for almost every @xmath490 and @xmath491 @xmath492 where @xmath493 is the geodesic flow lifted to @xmath483 and @xmath494 .
define the operator @xmath495 on @xmath477 as follows : for @xmath496 with compact support , let @xmath497 which is in @xmath498 by . with the same analysis as in proposition [ defw ]
, we find the integral kernel of @xmath499 @xmath500\big|_{t\tilde{v}={\widetilde}{\exp}_{\tilde{x}}^{-1}(\tilde{x } ' ) } \in c^\infty({\widetilde}{m}{\times}{\widetilde}{m}\setminus { \rm diag } ) \end{gathered}\ ] ] if @xmath501 is the exponential map for @xmath477 and @xmath502 solves the same jacobi equation as but for @xmath503 .
we clearly have @xmath504 if @xmath505 and the integral kernel @xmath506 of @xmath495 is positive , symmetric and satisfies @xmath507 for all @xmath508 . from , we can write for almost every @xmath509 @xmath510 let @xmath511 be supported in @xmath487 and such that @xmath512 if @xmath513 , then @xmath514 and we get for almost every @xmath513 and @xmath515 @xmath516 by using that @xmath517 is an isometry of @xmath478 and the invariance of @xmath506 by @xmath486 mentioned above
. we can then view @xmath518 as an operator acting on @xmath519 ( for @xmath488 ) with integral kernel @xmath520 the sum makes sense as an @xmath521 function since @xmath522 and @xmath523 for all @xmath524 .
in fact , we also know that @xmath518 maps @xmath525 to @xmath525 ( since it is the restriction of a pseudo - differential operator of order @xmath263 on @xmath147 ) thus , since @xmath526 , @xmath527 to estimate its @xmath430 norm for @xmath488 , we can use schur s lemma and since @xmath528 is symmetric , this gives directly the following [ normofpi ] with the notation above , the operator @xmath518 has norm bounded by @xmath529 from now on , we shall work in a neighborhood of a constant curvature metric @xmath22 and estimate the norm @xmath518 in that neighborhood .
let us denote by @xmath87 the hyperbolic space with curvature @xmath263 , realized as the unit disk @xmath530 with metric @xmath531 .
let @xmath23 have constant curvature @xmath532 for some @xmath533 , we can extend @xmath3 to a non - compact complete manifold @xmath534 with constant curvature .
indeed , this can be done as follows .
the connected components @xmath535 of the boundary @xmath4 are circles .
let @xmath536 be the distance from @xmath537 to the unique geodesic @xmath538 homotopic to @xmath537 in @xmath3 , these geodesics form the boundary of the convex core @xmath539 of @xmath3 .
if @xmath540 denotes the length of @xmath538 , then in normal geodesic coordinates @xmath541{\times}({\mathbb{r}}/c_i{\mathbb{z}})_{\alpha}$ ] from @xmath538 in the collar between @xmath537 and @xmath538 , the metric is given by @xmath542 then it suffices to replace the region between @xmath537 and @xmath538 by the full hyperbolic half - cylinder @xmath543 with the metric and we denote by @xmath534 the obtained complete surface .
it is realized as a quotient @xmath544 by a convex co - compact group of isometries @xmath545 of the hyperbolic space @xmath546 with curvature @xmath69 ; this space @xmath547 is just the unit disk @xmath530 with metric @xmath548 .
notice that @xmath486 is also convex a co - compact group of isometries of @xmath549 .
the exponent of convergence of @xmath486 is defined to be @xmath550 where @xmath551 denotes the riemannian distance of @xmath549 and @xmath331 is any point fixed in @xmath552 .
let @xmath553 be the limit set of the group @xmath486 ( ie .
the set of accumulation points in @xmath554 of the orbits @xmath555 in @xmath556 ) . by a result of patterson @xcite and sullivan @xcite , @xmath557 where @xmath558 is the trapped set of the geodesic flow .
the universal cover @xmath477 of @xmath3 is a subset of the open disk @xmath556 with infinitely many boundary components which are pieces of circles intersecting @xmath559 at their endpoints .
the solution of is @xmath560 and the integral kernel of @xmath495 and @xmath518 are thus @xmath561 if @xmath562 , @xmath563 and @xmath564 with @xmath565 a fundamental domain .
notice that the expression for @xmath528 extends to @xmath566 and this defines an operator @xmath567 so that @xmath568 if @xmath60 .
we prove the following [ norml^2pi0 ] let @xmath23 be a manifold with strictly convex boundary and constant negative curvature @xmath69 , let @xmath545 be its fundamental group and @xmath569 the exponent of convergence of @xmath486 as defined in .
let @xmath570 $ ] which satisfy the bound @xmath571 and set @xmath572 , then for all metric @xmath2 on @xmath3 with strictly convex boundary and gauss curvature @xmath75
satisfying @xmath573 the operator @xmath574 of @xmath2 has norm bounded by @xmath575 let @xmath576 so that @xmath577
. first , by the comparison result , we have the pointwise bound on the kernel @xmath506 of @xmath495 in the universal cover @xmath578 @xmath579 where we used @xmath580 and @xmath581 on @xmath477 . using the lower bound @xmath582 for @xmath455
, we get @xmath583 this implies that the kernel @xmath528 of @xmath518 satisfies @xmath584 provided that the sum converges , which is the case if @xmath585 .
let @xmath586 be the metric with gauss curvature @xmath263 on @xmath3 and let @xmath587 be the operator on @xmath3 whose integral kernel with respect to the volume form @xmath588 is @xmath589 .
let @xmath590 equipped with the hyperbolic metric @xmath591 ( with curvature @xmath263 ) and consider the operator @xmath592 on @xmath593 defined by .
as explained above , the manifold @xmath593 can be viewed as a complete extension of the manifold with boundary @xmath594 , and we have for each @xmath595 extended by @xmath88 on @xmath593 @xmath596 the operator norm of @xmath592 on @xmath597 is computed in lemma [ pi0hyp ] and we denote this constant by @xmath598 .
we then get @xmath599 , we use together with the fact that the integral kernels of @xmath518 and @xmath587 operator are positive and that @xmath600 this gives the bound @xmath601 the result then follows from and the value for @xmath602 given by lemma [ pi0hyp ] .
combining proposition [ normw ] with theorem [ norml^2pi0 ] , we obtain theorem [ theo2 ] .
we now illustrate with numerical examples the reconstruction formula proposed in theorem [ inversion ] , i.e. , we will reconstruct functions from knowledge of their ray transform , where the underlying surface with boundary is of one of the following two models : * a quotient of the poincar disk @xmath603 by a schottky group @xmath486 .
* the cylinder @xmath604 with a metric of circular symmetry such that the trapped set is hyperbolic in example @xmath605 , the constancy of the curvature makes the operator @xmath65 vanish identically , so that the inversion is expected to be one - shot . in example @xmath606 , we will implement a partial neumann series to invert the operator @xmath607 .
much of the ideas and implementation draw from the second author s previous implementation @xcite in the case of non - trapping surfaces .
the numerical domain can be viewed as a domain in @xmath608 with an isotropic metric @xmath609 ( except for experiment 5 , where we use a metric of the form @xmath610 ) , and with certain identifications producing nontrivial topology and trapped geodesics .
the underlying grid is cartesian ( subset of a @xmath611 uniform discretization of @xmath612 ^ 2 $ ] with @xmath613 here ) , which simplifies the computation of the operator @xmath614 .
more specifically , in isothermal coordinates @xmath615 with an isotropic metric of the form @xmath609 , the operator @xmath616 appearing in takes the expression @xmath617 where , despite the fact that this formula might not make sense for those @xmath615 in the trapped set , the hyperbolicity of the flow is such that , when discretizing the integral over @xmath618 , numerical imperfections will make every geodesic exit the domain .
the operators @xmath619 , @xmath620 can then conveniently computed by finite differences . on the other hand
, this discretization implies that every pixel will carry a different volume , and that accuracy of reconstructions and resolution may be visually position - dependent . to the authors knowledge , litterature on numerical analysis of ray transforms on manifolds ( covering ,
e.g. , error estimates , appropriate sampling , regularization ) , is very scarce outside the well - understood euclidean case @xcite , an exception being on geodesic and horocyclic transforms on the hyperbolic plane in @xcite , whose homogeneity allows for harmonic analysis .
the computation of geodesics is done in two ways depending on the purpose : * when computing the ray transform of a given function , we solve an ode and integrate the function along the way .
the method used in that case is heun s scheme .
* in the case of quotients of @xmath621 , where geodesics can be given by explicit formulas , we use these formulas to compute geodesic endpoints in a fast way during the backprojection step .
computations are performed with matlab , using gpu in order to speed up calculations . as in @xcite ,
the main bottleneck is the backprojection step , which requires computing @xmath622 geodesic endpoints , with @xmath623 the gridsize and @xmath624 the number of directions used to compute integral over @xmath625 .
this process requires computing whole geodesics , more of whom take longer to exit the domain when trapping increases ( this increase here is quantified by the fractal dimension of the limit set of the corresponding schottky group ) .
accordingly , computations run within minutes on a personal computer with gpu card for the case of cylinders , hours for the case of schottky groups with two generators .
here and below , we denote @xmath626 the unique geodesic passing through @xmath627 with direction @xmath628 .
fix @xmath629 , let @xmath630 and define @xmath631 by @xmath632 the unique hyperbolic translation mapping @xmath331 to @xmath633 and define the surface @xmath634 . a fundamental domain for this surface is delimited by the geodesics @xmath635 and @xmath636 .
this surface has no boundary , so we cut one up using an arc circle which intersects the geodesic @xmath635 at a right angle , which ensures that the boundary is smooth .
this boundary has two connected components .
the forward trapped set consists of , at every point , precisely two directions : the ones whose flow - out lands into either `` point at infinity '' @xmath637 or @xmath638 .
the trapped set is simply the projection of the two oriented geodesics in @xmath639 relating @xmath637 to @xmath638 , and these are two oriented closed geodesics in the quotient .
* experiment 1 : * we pick the function in fig . [
fig : model1 ] ( right ) , compute its ray transform on fig .
[ fig : model1_recons ] ( left ) using @xmath640 boundary points on each top and bottom boundary , and @xmath641 directions for each boundary point uniformly within @xmath642 .
we then apply the inversion formula , discretizing the @xmath618 integrals using @xmath640 directions . the error is visualized on fig .
[ fig : model1_recons ] ( right ) . .
the @xmath331-axis represents a point on the boundary parameterized with a parameter in @xmath643 $ ] , and the @xmath644-axis is the shooting direction with respect to the inner normal , an angle in @xmath645 .
right : pointwise error on reconstruction after applying the formula from theorem [ inversion ] to the data.,title="fig : " ] .
the @xmath331-axis represents a point on the boundary parameterized with a parameter in @xmath643 $ ] , and the @xmath644-axis is the shooting direction with respect to the inner normal , an angle in @xmath645 .
right : pointwise error on reconstruction after applying the formula from theorem [ inversion ] to the data.,title="fig : " ] * observations : * based on the results displayed on fig .
[ fig : model1_recons ] , one can notice artifacts of 10% relative size in certain areas .
these artifacts lie in the region between the support of the function and the points at infinity ( this is especially noticeable once we periodize the reconstruction ) , and we propose two possible reasons for this : * as the computation of forward data does not use an exact computation of the geodesics , the accuracy degrades on the longer geodesics , i.e. the ones that wander near the closed one . * the data may have logarithmic singularities at trapped directions , which upon applying the hilbert transform , may create undesirable oscillations in the vicinity of these angles .
these oscillations then create artifacts upon applying the backprojection @xmath48 , especially at the points which are such that the data is nonzero near their trapped directions ( see fig . [
fig : exp2_2 ] below ) .
this is precisely the case when these points lie between the support of the function and the points at infinity .
we now consider examples of quotients of @xmath603 by schottky groups with two generators .
keeping @xmath629 , @xmath646 and @xmath647 defined as above , we define the surfaces @xmath648 and @xmath649 .
both @xmath650 and @xmath651 can have the same fundamental domain , delimited by the geodesics @xmath652 and @xmath653 , though with different identifications , as represented fig .
[ fig : schottky2 ] . in both examples ,
the trapping is now more complicated : at every point @xmath654 , the trapped directions form a fractal set at the tangent circle @xmath655 . in light of experiment 1 ,
this is expected to have quite a detrimental impact on the inversion .
we study this influence by comparing noise on experiments 2 and 4 , where the same model @xmath650 is chosen for two different values of @xmath331 , yielding different trapping intensities .
we quantify this by the dimension of the limit set @xmath656 $ ] of the group @xmath657 .
we compute a numerical approximation of @xmath658 by running geodesic dynamics on a large number of points inside the domain and measuring their escape rate , i.e. the proportion that remains inside as a function of time . according to ( * ? ? ?
* appendix b ) , this escape rate is of the form @xmath659 , and we use this to approximate @xmath658 by linearly interpolating @xmath660 , possibly after throwing away transient regimes . in experiments 2 - 3 - 4 ,
the influx boundary is discretized uniformly into @xmath661 boundary points and @xmath662 influx directions in the range @xmath663 per boundary point
. * experiment 2 . * on @xmath664 , we pick the function in fig . [
fig : exp2 ] ( left ) , compute its ray transform on fig .
[ fig : exp2 ] ( right ) and apply the inversion formula .
the reconstruction is visualized on fig .
[ fig : exp234 ] ( left ) . for later comparison ,
we have computed @xmath665 following the method described above . * experiment 3 . * on @xmath666 , we pick the function in fig . [
fig : exp3 ] ( left ) , compute its ray transform on fig .
[ fig : exp3 ] ( right ) and apply the inversion formula .
the reconstruction is visualized on fig .
[ fig : exp234 ] ( middle ) . * experiment 4 .
* we now repeat experiment 2 , this time on @xmath667 , whose topology is the same , though the trapping is now more prominent ( we have computed @xmath668 ) .
we pick the function in fig . [
fig : exp4 ] ( left ) , compute its ray transform on fig .
[ fig : exp4 ] ( right ) and apply the inversion formula .
the reconstruction is visualized on fig .
[ fig : exp234 ] ( right ) . * observations . * as in experiment 1 , the same comments hold regarding the influence of trapping : the more trapping , the more singularities in the data , which create oscillations upon applying the hilbert transform ( see fig . [
fig : exp2_2 ] ) , in turn yielding artifacts when applying the backprojection @xmath48 .
comparing experiments 2 and 4 , the difference in trapping can be both seen on the singularities of the forward transforms ( compare figs .
[ fig : exp2 ] , right and [ fig : exp4 ] , right ) , and on higher noise in reconstruction ( compare figs . [
fig : exp234 ] , left and right ) .
let the topological cylinder @xmath604 endowed with the metric @xmath669 , where we have defined @xmath670 the gauss curvature is independent of @xmath331 and given by @xmath671 so we have @xmath672 for every @xmath644 , and we also get @xmath673 the structure of the trapped set is similar to that of experiment 1 , with the set @xmath674 the projection on the base of the only closed geodesic and @xmath675 .
we can bound @xmath2 in terms of the constant curvature metric @xmath22 with @xmath676 via the constants @xmath677 and @xmath678 as in theorem [ norml^2pi0 ] .
the requirement @xmath679 there gives the constraint @xmath680 .
since @xmath681 , we deduce the estimate , with @xmath682 , @xmath683 combining this estimate with the estimate in proposition [ normw ] , we deduce @xmath684 in particular , @xmath685 and thus for @xmath686 small enough , inversion via a neumann series is justified . as in @xcite , when implementing the neumann series , the error operator @xmath687 is not implemented directly as it would be hopeless to satisfy at the discrete level . instead , once @xmath0 and its approximate inverse ( call it @xmath688 , say ) are discretized , we set @xmath689 and iterate the neumann series using this operator
. * experiment 5 .
* we pick the function in fig . [
fig : exp5 ] ( right ) , compute its ray transform on fig .
[ fig : exp5_2 ] ( top - left ) and implement the approximate inversion formula ( visualized fig .
[ fig : exp5_2 ] , top - right ) , followed by two neumann series iterations .
pointwise errors after both one - shot and corrected reconstructions are visualized fig .
[ fig : exp5_2 ] ( bottom ) . in light of these experiments
, we make the following comments : * despite the theoretical exactness and stability of the proposed reconstruction formulas , geometries with trapping pose some numerical challenges , related to ( i ) the singularities it creates in data space , which in turn create oscillations after applying the hilbert transform , yieling reconstruction artifacts after backprojection ; and ( ii ) , the loss of accuracy on the computation of long geodesics which wander near the trapped set before exitting the domain .
* as mentioned before in @xcite , another important challenge in surfaces with nonconstant curvature is to understand how to design an appropriate discretization of the influx boundary whose flow - out samples the manifold in as uniform a way as possible .
* in the case of non - constant curvature , iterative corrections , which require a differentiation step at every iteration , may increase noise at high frequencies ( which , as mentioned earlier , becomes more prominent in the presence of trapping , unlike the case of simple surfaces covered in @xcite ) .
regularization methods should then be considered .
let @xmath691 be a quotient of hyperbolic space by a convex co - compact group @xmath545 and denote by @xmath658 the exponent of the group @xmath486 as defined by .
such a manifold is non - compact but can be compactified into @xmath692 where @xmath693 is the set of discontinuity of @xmath486 ; @xmath694 is the largest open set in the closed unit disk @xmath554 where @xmath486 acts properly discontinuously ( we view @xmath87 as the unit disk @xmath556 in @xmath695 ) .
the laplacian @xmath696 has continuous spectrum @xmath697 on @xmath396 and possibly finitely many eigenvalues in @xmath698 , which appear if and only if the exponent @xmath699 of @xmath486 satisfies @xmath700 , in which case @xmath701 .
for further details on these topics , we refer to the book @xcite . in this appendix , we compute for small @xmath702 the operator on @xmath3 @xmath703 where @xmath14 denotes the geodesic flow at time @xmath16 on the unit tangent bundle @xmath15 . just like in section [ normofpi0 ]
, the operator @xmath690 makes sense and has integral kernel @xmath704 , which can written as a converging sum on a fundamental domain @xmath565 of @xmath486 when @xmath705 @xmath706 for @xmath564 and @xmath707 the covering map , where @xmath708 is the integral kernel of the operator @xmath709 defined by the expression when @xmath710 . as in section [ normofpi0 ] , this kernel is given by @xmath711 which justifies that the sum above converges if @xmath705 .
the expression can also be rewritten as @xmath712 using the symmetry @xmath713 in each fiber @xmath714 .
[ pi0hyp ] on a convex co - compact quotient @xmath691 with exponent of convergence given by @xmath658 , the operator @xmath690 can be written when @xmath715 as @xmath716 where @xmath717 is the euler gamma function , and with the convention that @xmath718 when @xmath719 is an @xmath146-eigenvalue of @xmath720 with @xmath721 .
the operator norm is given by @xmath722 since its schwartz kernel is a function of @xmath551 decaying like @xmath723 , the operator @xmath709 is a function @xmath724 of the hyperbolic laplacian @xmath725 ( see for example ( * ? ? ?
* section 1.8 ) ) .
notice that the series defining @xmath704 is absolutely convergent on compact sets of @xmath726 since @xmath727 if @xmath486 is convex co - compact .
the operator @xmath728 is bounded on @xmath729 if @xmath730 : indeed , this is an operator in the class @xmath731 of the @xmath88-calculus of pseudo - differential operators introduced by mazzeo - melrose @xcite and the @xmath146-boundedness on @xmath87 follows for instance from mazzeo ( * ? ? ?
* theorem 3.25 ) . to compute the function @xmath732
, we use the representation theory of @xmath733 , following @xcite . the ( time reversal )
principal series representations are unitary representations into @xmath734 defined as follows @xmath735 let @xmath736 , then @xmath737 ( if we identify @xmath738 in the canonical way ) and to compute @xmath739 we will use the plancherel formula of harish - chandra @xcite @xmath740 there is a unique @xmath32-invariant vector of norm @xmath43 in @xmath741 where @xmath32 is the action of @xmath742 via the representation @xmath743 , it is given by @xmath744 since @xmath745 is invariant by @xmath746 , to consider the action of the right multiplication by an element @xmath747 on @xmath745 with @xmath748 , it suffices to consider the action of @xmath749 on @xmath750 .
the pull - back by the geodesic flow at time @xmath16 corresponds , in each unitary representation @xmath743 , to the unitary operator on @xmath741 @xmath751 the operator @xmath728 can be written as @xmath752 where @xmath753 is any element which projects to @xmath754 . since the average in the compact group @xmath746 amounts to project onto @xmath750 , the operator @xmath709 is unitarily equivalent to multiplication on the representation space @xmath741 by @xmath755 through the representation @xmath743 . by changing variable and by parity in @xmath331 ,
this term is of the form @xmath756 with @xmath757 notice that @xmath758 and that @xmath756 extends holomorphically in each strip @xmath759 and it can be written under the form @xmath760 for some @xmath732 holomorphic in the region containing @xmath761 and bounded by the parabola @xmath762 . the operator @xmath709 can then be written @xmath763 with @xmath764 the expression of @xmath765 for @xmath433 matches with the formula found by @xcite . in the image of the strip @xmath766 by @xmath767 for @xmath233 small enough @xmath768 for some @xmath769 depending only on @xmath686 .
we can then write , using cauchy formula @xmath770 where @xmath771 is the contour given by the image of @xmath772 by the map @xmath767 , where @xmath233 is chosen small . the contour is parametrized by @xmath773 with @xmath774 and the @xmath775 norm of @xmath776 is @xmath777 showing that the integral converges in @xmath778 norm . now
the resolvent @xmath779 has for integral kernel , viewed on the fundamental domain @xmath487 , @xmath780 when @xmath781 , and the sum is uniformly convergent on compact sets of @xmath782 for @xmath783 in that half - space : the proof is done in @xcite or ( * ? ? ?
* proposition 2.2 ) , and follows from the uniform convergence of poincar series @xmath784 for @xmath785 on compact sets and from the pointwise estimate for @xmath786 ( which follows from the explicit formula for @xmath787 given for instance in @xcite ) @xmath788 if @xmath789 is fixed and @xmath790 depends only on @xmath791 and @xmath686 .
this implies that , for @xmath792 and @xmath793 two relatively compact open sets of @xmath87 so that @xmath794 , there is @xmath244 depending only on the volume of @xmath792 and @xmath793 such that for all @xmath795 @xmath796 now , using that for @xmath434 supported in a compact set @xmath792 we have @xmath797r_{{\mathbb{h}}^2}(r^2)f= ( 1+r^2)\chi r_{{\mathbb{h}}^2}(r^2)f\ ] ] for each @xmath798 supported in a relatively compact set @xmath793 with @xmath799 , we get @xmath800 for @xmath801 where @xmath802 depends only on @xmath803 .
this estimate together with the poincar series convergence implies that for each relatively compact open set @xmath804 , @xmath805 and @xmath806 , @xmath807 converges in the sense of schwartz kernels of bounded operator @xmath808 .
this shows that for @xmath809 @xmath810 where @xmath732 is the holomorphic function defined before .
now the bottom of the spectrum of @xmath696 corresponds to @xmath811 .
let us show that the function @xmath812 is decreasing : we have @xmath813 where @xmath814 is the digamma function , and using its series expansion @xmath815 with @xmath816 the euler constant , we get for @xmath455 @xmath817 is negative .
similarly , the function @xmath818 is increasing in @xmath819 : indeed , one has @xmath820 thus using the series expansion of @xmath821 and factorizing we obtain for @xmath822 @xmath823 from the spectral theorem , this implies that the @xmath824 norm of @xmath690 is obtained by evaluating @xmath732 at either @xmath88 if @xmath825 or at @xmath826 if @xmath700 .
this ends the proof . using a similar analysis ,
the decomposition of @xmath827 into irreducible representations when @xmath486 is co - compact ( see for instance @xcite ) proves that the operator @xmath518 defined in @xcite is given on a compact hyperbolic manifold @xmath828 by the same formula as in lemma [ pi0hyp ] , but setting @xmath433 and replacing @xmath696 by @xmath829 if @xmath830 is the orthogonal projection onto @xmath831 in @xmath396 .
b. fridman , p. kuchment , k. lancaster , s. lissianoi , m. mogilevski , d. ma , i. ponomarev , v. papanicolaou , _ numerical harmonic analysis on the hyperbolic plane _ , applicable analysis , taylor & francis , * 76 * ( 2000 ) , 351362 s. helgason , _ the totally - geodesic radon transform on constant curvature spaces . _ integral geometry and tomography ( arcata , ca , 1989 ) , 141149 , contemp . math .
, 113 , amer .
soc . , providence , ri , 1990 .
j. radon _ uber die bestimmung von funktionen durch ihre integralwerte lngs gewisser mannigfaltigkeiten _
leipzig , math .
* 69 * ( 1917 ) , 262277 ; english translation in deans , s. , the radon transform and some of its applications , krieger 1993 , 204217 . | we give reconstruction formulas inverting the geodesic x - ray transform over functions ( call it @xmath0 ) and solenoidal vector fields on surfaces with negative curvature and strictly convex boundary .
these formulas generalize the pestov - uhlmann formulas in @xcite ( established for simple surfaces ) to cases allowing geodesics with infinite length on surfaces with trapping .
such formulas take the form of fredholm equations , where the analysis of error operators requires deriving new estimates for the normal operator @xmath1 .
numerical examples are provided at the end . |
lep2 runs now at energies above the @xmath3 production threshold and the first @xmath5 pairs are observed . during the run of lep2
some several hundred events will be collected .
this might be enough to provide limits for @xmath1 and @xmath2 vertices , which are absent in the standard model of electroweak interactions at tree - level .
however , physics beyond the standard model might give strong contributions to neutral gauge boson vertices @xcite .
( 40,30 ) requiring only invariance under lorentz transformations the most general @xmath6 vertex , see figure [ triplez ] , with two on - shell @xmath5 bosons is given by @xcite : @xmath7 .
\label{fvertex}\ ] ] in contrast to the @xmath8 vertex , with seven possible anomalous couplings @xcite , bose symmetry forbids more couplings in ( [ fvertex ] ) ) , if one additional @xmath5 boson is off - shell , but their contributions are suppressed by a factor @xmath9 . ] .
the anomalous parameter @xmath10 in ( [ fvertex ] ) leads to violation of @xmath11 and @xmath12 symmetry , but maintains invariance under @xmath0 transformations , while @xmath13 would introduce @xmath0 violation in ( [ fvertex ] ) .
the differential standard model cross - section for @xmath3 production in @xmath14 annihilation is known for almost 20 years @xcite .
also , the effects of anomalous neutral gauge boson couplings in the production of @xmath15 , @xmath3 , and @xmath16 bosons have been studied , see e.g. @xcite
. in @xmath17 production processes at the tevatron @xcite and lep @xcite , limits to the vertex @xmath18 were obtained .
however , they can not be transferred to @xmath5 pair production , since the anomalous couplings in equation ( [ fvertex ] ) are independent of couplings in @xmath17 production .
the differential cross - section for @xmath3 production is presented in this section .
it is obtained for the various spin combinations of final state @xmath5 bosons .
this allows to determine their sensitivity to the anomalous couplings .
the finite width effects of the produced @xmath5 bosons are considered by convoluting the cross - section for @xmath3 production with two breit - wigner functions .
the differential cross - section is then expressed by the sum : @xmath19 with @xmath20 where @xmath21 and @xmath22 stand for the possible @xmath5 boson helicities @xmath23 , @xmath24 , and @xmath25 , which are defined by the vectors : @xmath26 the klln function @xmath27 , the breit - wigner factors @xmath28 and other notations may be inferred from @xcite .
the squared matrix element @xmath29 in ( [ sigmadiff ] ) can be expressed by the standard model part @xmath30 and the anomalous contributions
@xmath31 : @xmath32 the squared amplitude in the standard model is : @xmath33 where the @xmath34 couplings @xmath35 and @xmath36 can again be taken from @xcite .
the functions @xmath37 are : @xmath38 with the mandelstam variables : @xmath39 it is noteworthy that the coupling constant combination @xmath40 gives additional contributions to equation ( [ msm ] ) .
however , these terms cancel each other for all measurable cross - sections and are therefore not reproduced bosons are identical , not all of the various spin combinations are observable by themselves . as an example , the contribution of @xmath41 can not be distinguished from @xmath42 and only their sum can be measured . ] .
the contributions from the anomalous diagrams are : @xmath43 with the coupling constants @xmath44 for the @xmath45 coupling .
the denominators of the @xmath46-channel propagators @xmath47 cancel with the corresponding factors in the anomalous vertex function ( [ fvertex ] ) .
photon and @xmath5 exchange have the same @xmath46-dependence in ( [ mss ] ) . again , as in equation ( [ msm ] ) terms with additional coupling constant combinations appear for the various spin combinations , but they do not contribute to physical observables .
the functions @xmath48 for the @xmath46-channel diagrams squared are : @xmath49)f_5^{v_k}f_5^{v_l}\right\ } , \\
{ \cal g}^{\pm\pm}_{s}(s_1,s_2)&= & \left(s_1-s_2\right)^2\sin^2\theta f_5^{v_k}f_5^{v_l } , \\ { \cal g}^{\pm0}_{s}(s_1,s_2)&= & { \cal g}^{0-}_{s}(s_2,s_1 ) , \\ { \cal g}^{\pm\mp}_{s}(s_1,s_2)&=&0 . \label{00s-}\end{aligned}\ ] ] similarly , the interferences @xmath50 between @xmath51 and @xmath52-channel diagram and the @xmath46-channel diagram can be expressed by : @xmath53 as expected there are no contributions of anomalous couplings to the spin combinations @xmath54 and @xmath55 .
these are spin 2 states and can not be produced by the @xmath46-channel diagrams .
equations ( [ 00s+ ] ) ( [ 00i- ] ) exhibit that the @xmath0 violating couplings proportional to @xmath13do neither interfere with the standard model terms nor with terms proportional to @xmath10 .
a further conclusion is that it is impossible to separate out the effects of the parameters @xmath56 and @xmath57 on the differential cross - section of the process @xmath4 .
however , both parameters imply @xmath0 violation in a @xmath6 vertex . in the limit of on - shell @xmath5 pair production
only the combination with one longitudinally and one transversally polarized @xmath5 receives contributions from anomalous couplings . with the given expressions for the various spin combinations ,
the differential cross - section ( [ sigmasummed ] ) is obtained : @xmath58 with the functions @xmath59 -\lambda(s+\sigma)\sin^2\theta\right\ } , \\
\sum_{v_k , v_l= \gamma , z}\!\!\frac{l_{v_k}l_{v_l}+r_{v_k}r_{v_l}}{4s_1s_2m_z^4}\left\ { \lambda ( 1+\cos^2\theta ) \left [ s\sigma f_4^{v_k}f_4^{v_l } + ( s\sigma-8s_1s_2 ) f_5^{v_k}f_5^{v_l } \right ] \label{asdef } \right.\\ & & \mbox{}\left .
\hphantom{\frac{l_{v_k}l_{v_l}+r_{v_k}r_{v_l}}{4s_1s_2m_z^4 } } + ( s_1-s_2)^2\left [ \lambda\sin^2\theta f_4^{v_k}f_4^{v_l } + 16s_1s_2f_5^{v_k}f_5^{v_l}\right ] \right\},\nonumber\end{aligned}\ ] ] where the abbreviation @xmath60 is used .
the standard model part is described by @xmath61 while the anomalous contributions are contained in @xmath62 and @xmath63 . in the standard model limit the differential cross - section in ( [ comp ] )
is in agreement with the result presented in @xcite .
radiative corrections to ( [ comp ] ) due to initial state radiation , see e.g. @xcite , can be applied in the structure function approach including a lorentz boost of the scattering angle as described similarly in @xmath64 pair production @xcite .
the presented analytical results demonstrate how potential anomalous couplings change the differential @xmath3 production cross - section .
anomalous couplings will have their main effect in the production of a longitudinally and a transversally polarized @xmath5 boson . for on - shell @xmath3 production ,
these are the only spin combinations which are sensitive to an anomalous signal . effects of non - vanishing @xmath13 and @xmath10 on the final states with two transversally or two longitudinally polarized @xmath5 bosons are zero or are suppressed by a factor @xmath65 . therefore , a measurement of the final state spins might be used to increase the ratio of the anomalous signal over the standard model background .
but , since only one type of spin combinations is sensitive to anomalous couplings , a spin analysis can not be used to disentangle the signals from @xmath13 and @xmath10 . however , due to the interference contribution of @xmath10 in ( [ 0pi+ ] ) and ( [ pi+0 ] ) , the anomalous parameters @xmath13 and @xmath10 show different angular distributions and it might be possible to separate out their effects at lep2 . * acknowledgment
* + i would like to thank t. riemann for many useful discussions and suggestions .
in addition , i am grateful to t. hebbeker and p. molnr for drawing my attention to the problem , to j. alcaraz for numerical comparisons and to w. lohmann for providing me with information about the current experimental situation . finally , i am also grateful to j. illana and t. riemann for carefully reading the manuscript . p. mery , m. perrottet , and f. m. renard , _ z. phys . _
* c38 * ( 1988 ) 579591 ; + f. m. renard , _ nucl .
* b196 * ( 1982 ) 93 ; + u. baur and e. l. berger , _ phys . rev .
_ * d41 * ( 1990 ) 1476 ; + u. baur and e. l. berger , _ phys .
_ * d47 * ( 1993 ) 48894904 ; + f. cornet , r. graciani , and j. i. illana , `` bounds on the @xmath66 couplings from hera '' , in _ proc . of the workshop on future physics at hera , 2526 sep 1995 , hamburg , germany _ , ( g. ingelman , a. de roeck , r. klanner , eds . ) , vol . 1 ,
pp . 208218 , desy , hamburg , 1996 .
collaboration , f. abe _ et .
_ , _ phys .
_ * 74 * ( 1995 ) 19411945 ; + d0 collaboration , s. abachi _ et .
_ , _ phys .
* 78 * ( 1997 ) 36403645 ; + d0 collaboration , b. abbott _ et .
_ , _ phys . rev . _
* d57 * ( 1998 ) 38173821 .
collaboration , m. acciarri _ et .
_ , _ phys .
_ * b412 * ( 1997 ) 201 ; + delphi collaboration , p. abreu _ et .
_ , _ phys .
_ * b423 * ( 1998 ) 194206 ; + l3 collaboration , m. acciarri _ et .
al . _ , `` study of anomalous @xmath67 and @xmath68 couplings at lep '' , cern preprint cern - ep-98 - 096 ( 1998 ) .
a. leike , `` semianalytic distributions in four fermion neutral current processes '' , in _ proc . of int .
workshop on perspectives for electroweak interactions in @xmath14 collisions , 58 feb .
1995 , tegernsee , germany _ ( b. kniehl , ed . ) , pp .
121130 , world scientific , singapore , 1995 . | the effects of @xmath0 violating anomalous @xmath1 and @xmath2 vertices in @xmath3 production are determined .
we present the differential cross - section for @xmath4 with dependence on the spins of the @xmath5 bosons .
it is shown that from the different spin combinations those with one longitudinally and one transversally polarized @xmath5 in the final state are the most sensitive to @xmath0 violating anomalous couplings .
tzc desy 98163 + hep - ph/9810522 + october 1998 seeking @xmath0 violating couplings + in @xmath3 production at lep2 + jochen biebel _ desy zeuthen , + platanenallee 6 , d-15738 zeuthen , germany _ |
single molecule magnets ( smm s ) have been under intense study recently , due to their potential uses in magnetic storage and quantum computing.@xcite the materials consist of insulating crystalline arrays of identical smm s 1 - 3 nm in size , each containing two or more magnetic ions .
since the magnetic ions in each smm are surrounded by non - magnetic ligands , the intermolecular magnetic interactions are usually negligible .
although the most commonly studied smm s are the high - spin mn@xmath24 and fe@xmath25,@xcite such smm s contain a variety of ferromagnetic ( fm ) and antiferromagnetic ( afm ) intramolecular interactions , rendering unique fits to a variety of experiments difficult.@xcite in addition , there have been many studies of afm fe@xmath26 ring compounds , where @xmath27 etc.@xcite in these studies , analyses of inelastic neutron diffraction data and the magnetic induction @xmath15 dependence of the low - temperature @xmath7 specific heat @xmath28 and magnetization @xmath29 steps were made , using the isotropic heisenberg near - neighbor exchange interaction , the zeeman interaction , and various near - neighbor spin anisotropy interactions.@xcite however , the rings were so complicated that analyses of the data using those simple models were inaccessible to present day computers.@xcite thus , those authors used either simulations or phenomenological fits to a first - order perturbation expansion with different spin anisotropy values for each global ring spin value.@xcite here we focus on the simpler cases of equal spin @xmath30 magnetic dimers , for which the full spin anisotropy effects can be evaluated analytically , investigated in detail numerically , and compared with experiment .
afm dimers with @xmath31,@xcite and various forms of fe@xmath21 with @xmath20 were studied recently.@xcite several fe@xmath21 dimers and effective @xmath22 dimers of the type [ mn@xmath23@xmath21,@xcite have magnetic interactions weak enough that their effects can be probed at @xmath32 1k with presently available @xmath15 .
a comparison of our results with @xmath11 step data on a fe@xmath21 dimer strongly suggests a substantial presence of local spin anisotropy.@xcite the paper is organized as follows . in section
ii , we present the model in the crystal representation and give exact formulas for the matrix elements .
the general thermodynamics are presented in section iii , along with the universal behavior of the @xmath33 and @xmath34 behavior associated with the energy level crossing . in section iv , we solve the model exactly for @xmath4 , giving analytic expressions for @xmath35 and the specific heat @xmath28 . in section
v , we discuss the exact solution for @xmath10 , present the equations from which the eigenvalues are readily obtained , and give examples of the low-@xmath7 @xmath11 curves . in section
vi , we present examples of the low-@xmath7 @xmath20 @xmath6 curves . in section vii , we present our induction representation results for the eigenstates to first order in the anisotropy energies .
these are used to obtain the asymptotic hartree expressions for the thermodynamic quantities @xmath35 and @xmath28 , and for the inelastic neutron scattering cross - section @xmath36 and the electron paramagnetic resonance susceptibility @xmath9 for arbitrary @xmath0 , that are accurate at sufficiently low @xmath7 and/or large @xmath13 .
in addition , an analytic formula accurate to second order in the four independent anisotropy exchange energies is provided for @xmath37 , the @xmath15 values at which the @xmath38th energy level crossing occurs . in section viii
, we consider fm anisotropic interactions sufficiently strong as to remove the level crossing effects .
finally , in section ix , we summarize our results and discuss them with regard to experiments on fe@xmath21 dimers .
we represent the @xmath30 dimer quantum states , @xmath39 in terms of the global ( total ) spin and magnetic quantum numbers @xmath38 and @xmath40 , where @xmath41 and @xmath42 satisfy @xmath43 and @xmath44 , where @xmath45 , @xmath46 , and we set @xmath47 .
we also have @xmath48 , where @xmath49 and @xmath50 for an arbitrary @xmath15 , we assume the hamiltonian has the form @xmath51 , where @xmath52 contains the heisenberg exchange and zeeman interactions , the gyromagnetic ratio @xmath53 , @xmath54 and @xmath55 is the bohr magneton .
the global axial and azimuthal anisotropy terms @xmath56 and @xmath57 respectively , only involve components of @xmath58 , but have been the main anisotropy terms discussed in the smm literature,@xcite so we have included them for comparison .
such terms have been commonly studied as an effective hamiltonian for a singlet orbital ground state , in which the tensor global spin interaction with a fixed spin quantum number @xmath38 has the form @xmath59 , resulting in the principal axes @xmath60 and @xmath61.@xcite for a dimer , we take @xmath62 to be diagonal in the orientation pictured in fig .
1 . then @xmath63 , @xmath64 , and @xmath65 . taking @xmath66 and @xmath67 still leaves @xmath68 unrestricted .
the single - ion axial and azimuthal anisotropy terms , @xmath69 respectively , arise from spin - orbit interactions of the local crystal field with the individual spins .
these terms have usually been neglected in the smm literature , but have been studied with regard to complexes containing a single magnetic ion , such as ni@xmath70,@xcite and with regard to clusters of larger numbers of identical magnetic ions.@xcite the local axially and azimuthally anisotropic exchange interactions @xmath71 satisfy @xmath72 so we need only include either either @xmath73 or @xmath74 and @xmath75 or @xmath76 , respectively.@xcite that is , if we stick to the hamiltonian @xmath77 , we may incorporate the effects of @xmath74 and @xmath76 by letting @xmath78 , @xmath79 , and @xmath80 , @xmath81 , respectively . since @xmath73 and @xmath75 describe the axial and azimuthal anisotropy each single ion attains from its surrounding environment , they are the physically relevant local anisotropy interactions .
we note that @xmath82 and @xmath83 are symmetric under @xmath84 , whereas @xmath85 and @xmath75 are antisymmetric under @xmath84 , independent of @xmath0 . for the case of fe@xmath21,@xcite a constituent of the high - spin smm fe@xmath25 and the afm fe@xmath26 rings,@xcite the exchange between the fe@xmath86 @xmath20 spins occurs via two oxygen ions , and these four ions essentially lie in the same ( @xmath87 ) plane.@xcite we set the @xmath61 axis parallel to the dimer axis , as pictured in fig .
1 . since the quantization axis is along the dimer axis , which is fixed in a crystal , we denote this representation as the crystal representation .
we generally expect each of the @xmath88 for @xmath89 to satisfy @xmath90 , but there are not generally any other restrictions upon the various magnitudes of the @xmath88 . since all dimers known to date have predominantly afm couplings ( @xmath5 ) , and also because their magnetizations and specific heats are particularly interesting , we shall only consider afm dimers .
in addition , since no studies on unequal - spin dimers have been reported to our knowledge , we shall only treat the equal - spin @xmath30 case .
we note that for equal spin dimers , the group symmetry of the dimer environment is @xmath91 , so that dzaloshinski - moriya interactions do not arise.@xcite hence , our hamiltonian @xmath77 is the most general quadratic anisotropic spin hamiltonian of an equal - spin dimer .
dimer , with two bridging o@xmath92 ions ( o ) .
ligands ( not pictured ) are attached to the fe@xmath86 ions ( fe ) .
the arrows signify spins.,scaledwidth=15.0% ] for @xmath93 , we have @xmath94 @xmath73 and @xmath76 contain the individual spin operators @xmath95 and @xmath96 for @xmath97 . after some clebsch - gordon algebra involving the wigner - eckart theorem , for arbitrary @xmath98 , @xmath99 where @xmath100 and @xmath101/(4s^2 - 1)}.\label{eta}\end{aligned}\ ] ] the generalizations of these results to dimers with general @xmath102 are given in appendix a. for @xmath103 , we require @xmath104 , for which @xmath105 .
we then find @xmath106\alpha_{s , s_1},\label{gss1m}\\ h_{s , s_1}^{m,\sigma'}&=&d_{s+(\sigma'+1)/2,s_1}^md_{s+(3\sigma'+1)/2,s_1}^m,\label{hss1msigma}\\ k_{s , s_1}^{x,\sigma'}&=&c_{s+(\sigma'+1)/2,s_1}^{-\sigma'x-(1+\sigma')/2}c_{s+(3\sigma'+1)/2,s_1}^{-\sigma'x-(3\sigma'+1)/2},\label{kss1xsigma}\\ l_{s , s_1}^x&=&2f_s^x\alpha_{s , s_1},\label{lss1}\\ \noalign{and } \alpha_{s , s_1}&=&\frac{3s(s+1)-4s_1(s_1 + 1)-3}{(2s-1)(2s+3)}.\label{alphass1}\end{aligned}\ ] ] we note that @xmath107 and that eq . ( [ alphass1 ] ) holds for @xmath108 . equations ( [ sipm ] ) and ( [ siz ] ) allow for an exact solution to the most general hamiltonian of arbitrary order in the individual spin operators . in all previous treatments of more complicated spin systems with similar anisotropic interactions
, it was only possible to obtain numerical solutions , and therefore the full anisotropy of the magnetization and specific heat was not calculated.@xcite the operations of @xmath109 , @xmath82 and @xmath85 satisfy the selection rules @xmath110 , @xmath111 .
the local anisotropy interactions @xmath73 and @xmath75 allow transitions satisfying @xmath112 , @xmath113 , and @xmath112 , @xmath114 , respectively , so in the presence of either of these interactions , @xmath38 is no longer a good quantum number , unless @xmath4 .
in order to obtain the thermodynamic properties , we first calculate the canonical partition function , @xmath115 . since @xmath77 is not diagonal in the @xmath116 representation , we must construct the wave function from all possible spin states .
we then write @xmath117 where @xmath118 is constructed from the @xmath119 basis as @xmath120 where @xmath121 and @xmath122 is boltzmann s constant . to evaluate the trace , it is useful to diagonalize the @xmath123 matrix . to do so ,
we let @xmath124 , where @xmath125 is constructed from the new orthonormal basis @xmath126 , and @xmath127 is a unitary matrix of rank @xmath128 . choosing @xmath127 to diagonalize @xmath129 , @xmath130
, we generally obtain @xmath131 and the partition function for a smm dimer , @xmath132 the specific heat @xmath133 is then easily found at all @xmath134 , @xmath135.\label{sh}\end{aligned}\ ] ] the magnetization @xmath136 requires @xmath137 for each @xmath15 . as @xmath138 ,
at most two eigenstates are relevant . for most @xmath13 values ,
only one @xmath139 is important .
but near the reduced @xmath38th level - crossing induction @xmath140 at which @xmath141 , two eigenstates are relevant .
we then obtain for these two energies @xmath142 and @xmath143 is the magnetization of the @xmath38th eigenstate only .
we then expand in powers of @xmath144 , @xmath145 and we find @xmath146&{\rightarrow\atop{k_bt\ll|j|}}&c_v^{\rm peak}\label{peak}\\ c_v^{\rm peak}/k_b&= & \bigl(\frac{c}{\cosh c}\bigr)^2\nonumber\\ & \approx & 0.439229,\label{peakheight } \end{aligned}\ ] ] and @xmath147 where @xmath148 and @xmath149 is the solution to @xmath150 .
the coefficients @xmath151 and @xmath152 . the easiest way to see
this is to first rotate the crystal so the quantization axis is along @xmath153 , as discussed in appendix b. an expression for @xmath154 to second order in the @xmath155 is given in appendix d. we note that @xmath156 depends upon @xmath157 , and the direction of @xmath15 when anisotropic interactions are present .
the @xmath34 double peaks are double schottky anomalies at the reduced induction values @xmath158 .
hence , to @xmath159 , the heights and midpoint slopes of the @xmath18 low-@xmath7 @xmath33 steps are uniform and the same as for the isotropic case , but the step positions and hence their plateaus are not . correspondingly , the heights and positions of the @xmath18 @xmath34 double peaks are uniform and the same as for the isotropic case , but positions of the zeroes are not .
hence , the non - universal level - crossing inductions @xmath37 fully determine the low-@xmath7 thermodynamics of weakly anisotropic afm dimers .
in the next three sections , we consider the special cases of @xmath160 and 5/2 .
then , in section vii and appendix d , we present out general expression for @xmath161 accurate to second order in each of the @xmath88 .
we remark that a double peak in the low-@xmath7 @xmath34 curve has been seen experimentally in a much more complicated fe@xmath162 ring compound , and was attributed to level crossing.@xcite
plots of @xmath163 and @xmath164 versus @xmath165 for the isotropic spin 1/2 dimer were given previously.@xcite for @xmath4 with an arbitrary @xmath15 and @xmath88 for @xmath89 , the rank 4 hamiltonian matrix is block diagonal , since @xmath166 is a good quantum number .
the eigenvalues are given by @xmath167 where @xmath168\nonumber\\ & & -b^2\sin^2\theta[j_b - j_d\cos(2\phi)].\label{halfeigen}\end{aligned}\ ] ] for the special cases @xmath169 for @xmath170 , the @xmath171 satisfy @xmath172 where @xmath173 respectively , and where @xmath174 ( @xmath175 ) corresponds to the upper ( lower ) sign .
the magnetization for @xmath176 is given by @xmath177,\\ \delta_y&=&\exp(-\beta j_{y})[\exp(-\beta j)+\exp(2\beta j_{x})],\\ \delta_z&=&\exp(-\beta j_b)[\exp(-\beta j)+1],\end{aligned}\ ] ] respectively .
when the interactions are written in terms of the less physical @xmath178 , and @xmath179 , then @xmath180 and @xmath181 , so that for @xmath4 , @xmath182 and @xmath179 merely renormalize @xmath183 and @xmath184 .
neither of the single - ion spin anisotropy terms @xmath73 and @xmath75 affect the thermodynamics for @xmath4 .
we note that @xmath185 for each @xmath13 , as expected from @xmath186 . from eqs .
( [ epsilon1 ] ) , ( [ epsilon234 ] ) , ( [ lambdaz ] ) , and ( [ lambdaxy ] ) , the single level crossing induction ( @xmath187 ) for am @xmath4 dimer with @xmath176 occurs at @xmath188 provided that @xmath189 and @xmath190 , respectively . to distinguish the different effects of the global anisotropy interactions @xmath183 and @xmath184 that affect the magnetization of @xmath4 dimers , in figs . 2 and 3 ,
we have respectively plotted the low-@xmath7 @xmath191 versus @xmath165 with @xmath192 and @xmath193 for @xmath194 ( solid ) and @xmath195 ( dashed ) , along with the isotropic case @xmath196 ( dotted ) . from fig .
2 , @xmath197 and @xmath198 causes a greater shift to higher @xmath13 values at the magnetization step with @xmath199 than for @xmath195 , consistent with eq .
( [ bstephalf ] ) .
in addition , @xmath184 finite with @xmath200 has a very different effect upon the anisotropy of the magnetization step , as shown in fig .
although for @xmath201 , @xmath13 at the step is slightly reduced from its isotropic interaction value , for @xmath195 , the magnetization step occurs at a larger @xmath13 .
these results are consistent with eq .
( [ bstephalf ] ) .
the midpoint slopes are universal , in accordance with eq .
( [ slope ] ) . versus @xmath165 at @xmath202 for the afm spin
1/2 dimer with @xmath203 , @xmath198 , with @xmath199 ( solid ) , @xmath204 ( dashed ) , along with the isotropic case @xmath196 ( dotted).,scaledwidth=45.0% ] versus @xmath165 at @xmath202 for the afm spin 1/2 dimer with @xmath205 , @xmath200 , with @xmath199 ( solid ) , @xmath204 ( dashed ) , along with the isotropic case @xmath196 ( dotted).,scaledwidth=45.0% ] the specific heat of an @xmath4 dimer with @xmath169 is @xmath206 where the @xmath207 are given by eq .
( [ di ] ) , and the @xmath208 are given in appendix a. plots at low @xmath7 of @xmath163 versus @xmath165 for @xmath4 dimers with the corresponding global anisotropies @xmath203 and @xmath205 are shown in figs .
4 and 5 , respectively .
we note the universal curve shapes , but non - universal level - crossing positions , in quantitative agreement with eqs .
( [ cvzeroes])-([peakheight ] ) . in fig .
4 , the positions of the maxima and the central minimum in @xmath28 track that of the magnetization step in fig .
2 with the same parameters . with @xmath205 , the behaviors in @xmath28 and @xmath209 for @xmath204 are also very similar . however , there is a slight difference in the behaviors for @xmath199 .
note that @xmath209 ( fig .
3 ) shows a slight reduction for @xmath199 in the induction required for the step , whereas @xmath28 ( fig .
5 ) shows a slight increase in the positions of the peaks .
this detail only appears when @xmath210 , for which the effective temperature is slightly higher than for @xmath198 . versus @xmath165 for the afm spin 1/2 dimer with @xmath211 at @xmath202 with the same curve notation as in fig .
2.,scaledwidth=45.0% ] versus @xmath165 for the afm spin 1/2 dimer with @xmath205 , @xmath200 at @xmath202 with the same curve notation as in fig .
for dimers with @xmath10 , the allowed @xmath38 values are @xmath212 .
the three @xmath187 states are decoupled from the six remaining @xmath213 states .
they satisfy a cubic equation given in appendix a. for @xmath176 with @xmath170 , this cubic equation simplifies to a linear and a quadratic equation , as for the @xmath187 eigenstates of @xmath4 dimers .
the remaining six eigenstates corresponding nominally to @xmath213 are in general all mixed .
the matrix leading to the hexatic equation from which the six eigenvalues can be obtained is given in appendix a. that is sufficient to evaluate the eigenvalues for @xmath10 using symbolic manipulation software . when combined with the three @xmath187 eigenvalues , one
can then use eqs .
( [ sh ] ) and ( [ mag ] ) to obtain the resulting exact magnetization and specific heat at an arbitrary @xmath15 .
the combined nine eigenvalues depend upon all four anisotropy parameters @xmath88 for @xmath89 . to the extent that the eigenvalues can be obtained from the solutions to either linear or quadratic equations ,
the expressions for the @xmath214 are simple , and are given in appendix a. in the global anisotropy case @xmath215 , the first level crossing induction @xmath216 with @xmath176 is identical to @xmath217 , the level crossing with @xmath4 given by eq .
( [ bstephalf ] ) . for comparison ,
we expand the first and second level crossing inductions to first order in each of the @xmath88 for @xmath89 for @xmath169 where @xmath170 , @xmath218 where the upper ( lower ) sign is for @xmath195 ( @xmath219 ) , respectively . from these simple first - order results , it is possible to understand the qualitatively different behavior obtained with local , single - ion , anisotropy from that obtained with global anisotropy .
in the isotropic case @xmath220 , the first and second level crossings occur at @xmath221 and @xmath222 , respectively . for each induction direction , the signs of the @xmath183 and @xmath184 contributions to @xmath223 and @xmath224 are the same , whereas the signs of the @xmath225 and @xmath226 contributions to @xmath223 and @xmath224 are the _
opposite_. in figs .
6 - 10 , we plot @xmath191 versus @xmath165 for five low-@xmath7 cases of afm @xmath10 dimers , taking @xmath202 .
these curves all exhibit the universal features predicted by eqs .
( [ msteps ] ) and ( [ slope ] ) .
the corresponding @xmath163 versus @xmath165 curves , which also obey eqs . ( [ cvzeroes])-([peakheight ] ) , are presented elsewhere.@xcite in these figures , only one of the five anisotropy interactions @xmath88 is non - vanishing , and we take @xmath227 , for @xmath228 and @xmath229 , respectively . unlike the case of @xmath4 dimers , for which @xmath179 merely renormalizes @xmath184 , for @xmath10 each of these interactions leads to distinct anisotropy effects in the low-@xmath7 thermodynamics arising from the @xmath230 . and @xmath231 of @xmath164 versus @xmath165 for the afm spin 1 dimer , with the same curve notation as in fig .
2.,scaledwidth=45.0% ] we first show the results for the global anisotropy interactions . in fig . 6 , we plot @xmath33 for @xmath203 ( and all other @xmath232 ) at @xmath202 for the afm @xmath10 dimer .
@xmath233 is largest and increases monotonically with level - crossing number @xmath38 for @xmath234 , but is independent of @xmath38 for @xmath235 , nearly quantitatively consistent with eqs .
( [ b11z])-([b21xy ] ) . in fig .
7 , the corresponding results for @xmath205 ( and the other @xmath232 ) are shown .
again , @xmath233 is largest and increases monotonically for @xmath235 , and is nearly independent of @xmath38 for @xmath234 , as for @xmath236 , nearly quantitatively consistent with eqs .
( [ b11z])-([b21xy ] ) .
note that @xmath237 is nearly indistinguishable from the isotropic case . and @xmath238 of @xmath239 versus @xmath165 for the afm spin 1 dimer , with the same curve notation as in fig .
3.,scaledwidth=45.0% ] next , in figs . 8 and 9
, we show the corresponding curves for @xmath33 at low @xmath7 for the local anisotropies @xmath240 and @xmath241 , respectively , with the other @xmath232 .
in contrast to the global anisotropies , @xmath242 changes sign for both @xmath243 , nearly quantitatively consistent with eqs .
( [ b11z])-([b21xy ] ) .
we note that @xmath237 is nearly indistinguishable from the isotropic case for @xmath244 shown in fig .
9 . and @xmath245 of @xmath164 versus @xmath165 for the afm spin 1 dimer .
the curve notation is the same as in fig .
2 , except that the isotropic case ( dotted ) curve is for @xmath246.,scaledwidth=45.0% ] and @xmath247 of @xmath164 versus @xmath165 for the afm spin 1 dimer .
the curve notation is the same as in fig .
2 , except that the isotropic case ( dotted ) curve is for @xmath248.,scaledwidth=45.0% ] finally , in fig . 10 , we show the corresponding @xmath33 curves for @xmath249 . with this combination , @xmath233 is independent of @xmath38 for both @xmath243 , but is larger for @xmath235 , nearly quantitatively consistent with eqs .
( [ b11z])-([b21xy ] ) with @xmath250 , respectively . in short ,
the case @xmath10 is sufficient to exhibit the very different behaviors obtained from the single - ion , local spin anisotropy interactions from those obtained from the global spin anisotropy interactions .
as @xmath0 increases beyond 1 , the situation becomes not only more complicated , but also more interesting , as shown in the following . and
@xmath251 of @xmath164 versus @xmath165 for the afm spin 1 dimer .
the curve notation is the same as in fig .
2 , except that the isotropic case ( dotted ) curve is for @xmath252.,scaledwidth=45.0% ]
for @xmath20 , one of the cases of greatest experimental interest , when @xmath73 and @xmath75 are present , none of the allowed @xmath253 values is a true quantum number .
that is , @xmath83 and @xmath75 cause all of the states with nominally odd or even @xmath38 to mix with one another .
for @xmath176 for @xmath254 , this simplifies in the crystal representation , as for @xmath10 , since only states with odd or even @xmath40 values in the appropriately chosen representation can mix . by using symbolic manipulation software ,
it is possible to solve for the exact eigenvalues of the @xmath20 dimer .
however , because the analytic expressions for the eigenvalues are much more complicated than those for @xmath10 presented in appendix a , we shall not attempt to present them , but will instead focus upon their numerical evaluation for specific cases . of @xmath164 versus @xmath255 for the afm spin 5/2 dimer with @xmath231 .
the curve notation is the same as in fig .
2.,scaledwidth=45.0% ] versus @xmath165 for the afm spin 5/2 dimer at @xmath202 with @xmath238 .
the curve notation is the same as in fig .
3.,scaledwidth=45.0% ] to first order in the @xmath88 , the first three level crossings for @xmath176 with @xmath254 are @xmath256 in figs .
11 - 14 , we plot @xmath164 and @xmath34 versus @xmath255 for four low-@xmath7 cases of afm @xmath20 dimers , @xmath203 , @xmath205 , @xmath240 , and @xmath249 , respectively , and the other @xmath232 , taking @xmath202 .
each of these curves exhibit the universal step behavior predicted in eqs .
( [ msteps ] ) and ( [ slope ] ) . corresponding @xmath163 versus @xmath165 curves also exhibit the universal double peak behaviors predicted in eqs .
( [ cvzeroes])-([peakheight ] ) , and are shown elsewhere.@xcite in fig .
15 , examples of @xmath257 at fixed @xmath13 and @xmath258 are shown .
figures 13 - 15 are sufficient to distinguish the more interesting local spin anisotropy effects in afm dimers with higher @xmath0 values from the non - existent or less interesting ones present with @xmath259 , respectively . in figs .
11 - 14 , the solid and dashed curves represent the @xmath199 and @xmath195 cases , and the dotted curve represents the isotropic case , @xmath260 , as in figs .
2 - 10 .
we first examine the global spin anisotropy effects of @xmath82 and @xmath85 in figs .
11 and 12 .
these figures exhibit the same behavior shown for @xmath160 in the corresponding figs . 2 , 6 and 3 , 7 .
note that @xmath261 is largest for @xmath235 with @xmath203 and for @xmath234 with @xmath205 , and increases monotonically with @xmath38 , as for @xmath10 in both cases , nearly quantitatively consistent with eqs .
( [ b52step1])-([b52step3 ] ) .
by contrast , @xmath262 for @xmath235 with @xmath203 and for @xmath234 for @xmath205 are nearly indistinguishable from the isotropic case , also nearly quantitatively consistent with eqs .
( [ b52step1])-([b52step3 ] ) . and @xmath245 of @xmath164 versus @xmath165 for the afm spin 5/2 dimer .
the curve notation is the same as in fig .
8.,scaledwidth=45.0% ] and @xmath251 of @xmath164 versus @xmath165 for the afm spin 5/2 dimer .
the curve notation is the same as in fig .
10.,scaledwidth=45.0% ] in contrast , the local field anisotropy interactions show very different and much more interesting behaviors . in fig .
13 , we present our results for the effects of the single - ion axial anisotropy interaction @xmath73 , eq .
( [ lz ] ) . as in fig . 8 for @xmath10 ,
@xmath263 changes sign with increasing @xmath38 .
@xmath264 , whereas for @xmath265 , @xmath266 and increases monotonically with @xmath38 . for @xmath235 ,
nearly the opposite situation occurs .
@xmath267 is positive for @xmath187 and decreases monotonically with increasing @xmath38 . in both cases,@xmath268
is a minimum for @xmath269 .
these local axial anisotropy effects , consistent with eqs .
( [ b52step1])-([b52step3 ] ) , are very different than the global anisotropy ones pictured in figs . 11 and 12 .
they are also much richer and interesting than the corresponding case for @xmath10 pictured in fig .
8 . in fig .
14 , we present our @xmath164 versus @xmath165 results for the case of the local azimuthally anisotropic exchange interaction , @xmath76 , eq .
( [ hc ] ) , evaluated for afm dimers at @xmath202 with @xmath249 and the remaining @xmath232 .
@xmath270 for this case is obtained from eqs .
( [ b52step1])-([b52step3 ] ) by setting @xmath271 , respectively , and @xmath272 .
@xmath273 is nearly indistinguishable from the isotropic case , as in fig .
10 for @xmath10 , except for some minor curve shape effects far from the step midpoints .
@xmath267 is always positive , as for @xmath10 shown in fig .
10 , but has a minimum at @xmath274 .
this is also in stark contrast to the monotonic global anisotropy behavior for @xmath20 seen in figs .
11 and 12 . both of these behaviors are nearly quantitatively consistent with eqs .
( [ b52step1])-([b52step3 ] ) as modified to include @xmath179 .
although not pictured explicitly , the behavior for @xmath241 with the other @xmath232 is rather like that of the @xmath240 curves pictured in fig . 13 with @xmath275 , differing in ways similar to those differences between the @xmath240 and @xmath241 curves pictured for @xmath10 in figs . 8 and 9 .
as indicated in eqs .
( [ b52step1])-([b52step3 ] ) , the @xmath273 are nearly independent of @xmath226 .
however , the @xmath267 with @xmath241 are nearly three times as large as for @xmath240 case , so that the minimum @xmath276 is also a minimum for @xmath269 .
more details are given in section vii and appendix d. at @xmath258 versus @xmath277 near the second step at @xmath278 and @xmath202 for each @xmath227 with @xmath279 .
curves are labelled with @xmath280 values.,scaledwidth=45.0% ] in addition , the angular dependencies of @xmath281 are different for each of the four @xmath88 we presented for @xmath20 . in fig .
15 , we present the results for @xmath282 versus @xmath277 near the second level crossing at @xmath283 and @xmath202 for each of the four @xmath20 afm magnetization cases pictured in figs .
note that @xmath284 .
the @xmath225 and @xmath183 curves , while rather similar at @xmath285 , are very different at @xmath286 .
hence , @xmath11 depends strongly upon the particular type of spin anisotropy .
since the diagonalization of the hamiltonian matrix is difficult for an arbitrary magnetic field direction combined with an arbitrary combination of spin anisotropy interactions , and must be done separately for each value of @xmath0 , it is useful to consider a perturbation in the relative strengths @xmath155 of the anisotropy interactions .
we nominally assume @xmath287 for @xmath89 . however , to compare with low-@xmath7 magnetization and specific heat experiments at various applied field directions and magnitudes , one can not take the magnitude of the magnetic induction to be small . in order to incorporate the magnetic induction of an arbitrary strength and direction accurately , we therefore rotate the crystal axes @xmath288 to @xmath289 , so that @xmath290 . the rotation matrix and a brief discussion of its ramifications
are given in appendix b. in these rotated coordinates , the zeeman interaction @xmath291 , where @xmath292 , is diagonal .
we therefore denote this representation as the induction representation .
the hamiltonian @xmath293 in this representation is given in appendix b. in the induction representation , we choose the quantum states to be @xmath294 . in the absence of the four anisotropy interactions @xmath88 , @xmath295 is diagonal , @xmath296 the operations of the remaining terms in @xmath293 on the eigenstates @xmath294 are given in appendix c. the first order correction to the energy in the induction representation , @xmath297 , is found to be @xmath298-\frac{j_a}{2}[s(s+1)-1]\nonumber\\ & & + \frac{\tilde{j}_{b , a}^{s , s_1}}{2}[m^2+s(s+1)-1]\nonumber\\ & & + \frac{1}{2}[s(s+1)-3m^2]\nonumber\\ & & \times\bigl(\tilde{j}_{b , a}^{s , s_1}\cos^2\theta + \tilde{j}_{d , e}^{s , s_1}\sin^2\theta\cos(2\phi)\bigr ) , \label{esm1}\end{aligned}\ ] ] where @xmath299 and @xmath300 is given by eq .
( [ alphass1 ] ) .
since the @xmath301 dependence of @xmath302 arises from the term proportional to @xmath303 , it is tempting to think that the thermodynamics with @xmath304 and @xmath199 are equivalent to those with @xmath305 and @xmath195 .
however , as shown explicitly in the following , the @xmath301-independent parts of eq .
( [ esm1 ] ) strongly break this apparent equivalence , causing the @xmath161 for those two cases to differ .
this implies that @xmath183 and @xmath184 are inequivalent , as are @xmath225 and @xmath226 , even to first order in the anisotropy strengths . in the hartree approximation ,
the anisotropy interactions are included to first order only . in this approximation ,
@xmath38 and @xmath40 are still good quantum numbers , so the partition function @xmath306 where @xmath307 .
although it is difficult to perform the summation over the @xmath40 values analytically , it is nevertheless elementary to evaluate @xmath308 numerically for an arbitrary @xmath309 , and @xmath7 from the eigenstate energies .
the magnetization in the hartree approximation is @xmath310 similarly , the specific heat in the hartree approximation is @xmath311.\label{cvapprox}\end{aligned}\ ] ] as a test of the accuracy of this hartree calculation , we have compared the hartree and exact @xmath33 obtained for the @xmath20 dimer with @xmath205 and @xmath199 at various @xmath7 values in fig .
[ magtest ] . the corresponding comparison between the hartree and exact @xmath34 is shown in fig . [ shtest ] .
we see that the curves evaluated using the hartree and the exact expressions for @xmath209 and @xmath28 with @xmath20 are indistinguishable at @xmath202 .
the @xmath28 curves are noticeably different at @xmath312 for @xmath313 , and at @xmath314 they are noticeably different for @xmath315
. corresponding noticeable differences in the @xmath209 curves at the same @xmath13 values appear at @xmath7 values roughly three times as high as in the @xmath28 curves . at very low @xmath7 , @xmath316 ,
the most important states in this perturbative scheme are the minima for each @xmath38 value , @xmath317 , which determine the level crossings in the hartree approximation . as @xmath138
, we can ignore all of the @xmath318 states in eqs .
( [ zapprox])-([cvapprox ] ) .
this two - level approximation is the basis for the universal behavior given by eqs .
( [ cvzeroes])-([slope ] ) , which fits all of the exact curves we presented reasonably well .
versus @xmath165 obtained using the hartree asymptotic form ( dotted ) with the exact calculation ( solid ) , for the @xmath20 afm dimer with @xmath205 , @xmath319 , at @xmath320 , as indicated .
inset : expanded view of the region @xmath321.,scaledwidth=45.0% ] versus @xmath165 obtained using the hartree asymptotic form ( dotted ) with the exact calculation ( solid ) , for the @xmath20 afm dimer with @xmath205 , @xmath319 , at @xmath322 , as indicated.,scaledwidth=45.0% ] two response functions relevant for the study of smm s are the inelastic neutron scattering cross - section @xmath14 and the electron paramagnetic resonance ( epr ) susceptibility @xmath9 in strong magnetic inductions @xmath15 .
the low-@xmath7 inelastic neutron cross - section @xmath36 can be evaluated from @xmath323 where @xmath324 is the vector between the spins of a dimer and @xmath325 is the @xmath326th component of @xmath327 .
it is customary to write the operators in the crystal representation , for which @xmath328 .
@xcite with inelastic neutron scattering , one can probe the dimer with a strong magnetic field at various directions with respect to both the dimer axis and the scattering plane .
to the extent that the eigenstates @xmath126 and their energies @xmath139 can be evaluated exactly , @xmath329 where @xmath330 is the unitary operator that diagonalized @xmath77 . when exact expressions for @xmath36 are tedious to obtain , it is straightforward to obtain it in the hartree approximation , @xmath331 . but to do so , it is easiest to define the axes in the induction representation , for which @xmath332 and the @xmath333 for @xmath334 respectively . in the hartree approximation obtained by setting the @xmath335 , the bare wave functions , @xmath331 for an arbitrary @xmath336 is then @xmath337 where @xmath338 is invariant under the rotation , @xmath324 is the vector separating the dimer spins , @xmath339 is the angle between @xmath340 and @xmath15 , and @xmath341 where the coefficients are given respectively by eqs .
( [ dss1 m ] ) , ( [ asm ] ) , and ( [ css1 m ] ) , and we have suppressed the @xmath342 arguments of the @xmath343 functions for simplicity of presentation . in eq .
( [ sqomegahartree ] ) , the only dependencies upon the anisotropy energies is in the first - order eigenstate energies @xmath344 given by eqs .
( [ e0 ] ) and ( [ esm1 ] ) , respectively .
as described in detail in appendix c , corrections to the wave functions first order in the anisotropy interactions lead to the `` extended hartree '' approximation , @xmath345 , which exhibits additional transitions with strengths first and second order in the anisotropy energies .
@xmath346 contains terms arising from the traceless diagonal and off - diagonal parts of the @xmath347 component matrix , and from the contributions of the diagonal part of the component matrix containing corrections to its trace .
these latter terms are proportional to the same four functions of @xmath348 as in eq .
( [ sqomegahartree ] ) , and can be separated experimentally from the other terms by choosing the scattering plane such that @xmath349 ( or @xmath350 ) for all @xmath351 and by averaging over all azimuthal angles @xmath352 .
then , the transitions from @xmath116 to @xmath353 , where @xmath354 and @xmath355 with leading amplitudes second order in the anisotropy interactions upon depend upon the the local anisotropy interactions @xmath225 and @xmath226 through the three functions @xmath356 with @xmath357 given in appendix d. hence , if these transitions can be identified , a study of the @xmath351 dependencies of their strengths can provide direct measurements of those parameters . then , studying the weak transitions @xmath358 with @xmath359 , @xmath360 and @xmath361 can provide further measurements of the global anisotropy interactions @xmath184 and @xmath184 . in an epr experiment
, one applies a strong magnetic field , plus a weak oscillatory transverse field , leading to the overall induction @xmath362 $ ] , where we assume the oscillatory induction precesses clockwise ( counterclockwise ) for @xmath363.@xcite for a weak transverse induction @xmath364 , one measures the resulting linear response @xmath365 given by @xmath366\bigr),\nonumber\\ & & \\
s_{\sigma}(t)&=&e^{i{\cal h}t}s_{\sigma}(0)e^{-i{\cal h}t } , \end{aligned}\ ] ] where @xmath367 , the global spin raising and lowering operators @xmath368 are defined in the induction representation , as described in appendix b and @xmath308 is the partition function , and @xmath369=ab - ba$ ] is the commutator . in principle , an exact expression for @xmath365 can be obtained by starting in the induction representation with the basis @xmath370 , diagonalizing @xmath77 with the unitary operator @xmath371 , @xmath372 , and the new basis @xmath373 is obtained from @xmath374 , leading to @xmath375 .
one then has @xmath376 where @xmath377 .
it is then elementary to obtain @xmath378 in the hartree approximation valid to first order in the anisotropy parameters , which is well - behaved for all induction values .
the imaginary part of @xmath379 is then @xmath380\nonumber\\ & & [ s(s+1)-m^2-\sigma m]\nonumber\\ & & \times\bigl(\delta(e_{s , s_1}^m - e_{s , s_1}^{m+\sigma}+\omega)\nonumber\\ & & -\delta(e_{s , s_1}^{m+\sigma}-e_{s , s_1}^{m}+\omega)\bigr),\end{aligned}\ ] ] where @xmath344 is given by eqs .
( [ e0 ] ) and ( [ esm1 ] ) and @xmath381 is given by eq .
( [ zapprox ] ) .
the arguments of the @xmath382-function give rise to the resonant frequencies first order in the anisotropy energies , or to the first order resonant magnetic inductions , @xmath383 where @xmath384 and @xmath385 are given by eqs .
( [ tildejba ] ) and ( [ tildejde ] ) , respectively , and we have ignored the processes that give rise to finite transition widths.@xcite by varying the direction and magnitude of @xmath15 , it is possible to obtain sufficient information to fit all of the parameters in the model . at low @xmath7
the @xmath386 states dominate , so that only @xmath387 is allowed . by increasing @xmath15 past the @xmath38th level crossing
, one can probe the effective @xmath38th state of the afm dimer using epr . in appendix c , we extend these results to include the corrections to the wave functions first order in the anisotropy interactions . with these corrections ,
additional epr transitions are present , which can provide additional information useful in experimental identification of the anisotropy interactions . it is shown that there are 10 additional resonant magnetic induction strengths of the forms @xmath388 where the @xmath389 and @xmath390 are listed in appendix c. we note that these additional resonant inductions are first order in the anisotropy interactions , but have amplitudes that are second order in the anisotropy interactions , so that in most cases , they may be difficult to detect , but their detection in certain materials would provide a clear signal of the presence of significantly strong anisotropy interactions .
in addition , resonant inductions with @xmath391 are generally very large .
to illustrate one example of an effect of local spin anisotropy upon the epr transitions , we consider the simple case of an @xmath10 afm dimer with the only non - vanishing anisotropy energy @xmath245 and @xmath392 .
the energy levels for this system are pictured in fig .
18 . in fig .
19 , the epr transition energies versus @xmath165 for the @xmath10 afm dimer with @xmath245 and @xmath392 are shown .
the widths of the lines are proportional to the strengths of the epr matrix elements .
note that the ground state at @xmath393 is approximately @xmath394 , but its energy is slightly negative due to an admixture of this state with the nominal @xmath395 states with even @xmath40 , as shown in appendix a. the epr matrix elements are very small at these @xmath13 values , as the transitions only exist due to the admixture of the wave functions . at higher @xmath13 strengths
, the ground state goes through two level crossings , with the first level crossing being to the lowest energy @xmath187 state , which is nominally @xmath396 , with some mixture of the other @xmath397 states .
the leading transition is to the nominal @xmath398 state .
then , the second level crossing causes the ground state to be the nominal @xmath399 state , which is of course modified by the mixing with the nominal @xmath394 and the other nominal @xmath395 for @xmath400 .
note that near to @xmath401 , there is a level repulsion .
the leading epr transition is to the nominal @xmath402 state , and because of the strong matrix elements , the splitting of the energies due to the level repulsion should be observable in epr experiments . versus @xmath165 for @xmath10 afm dimers with @xmath245 and @xmath392 .
the arrows represent the strongest epr transitions at their corresponding field strengths.,scaledwidth=45.0% ] versus @xmath165 from the ground state for the @xmath10 afm dimer with @xmath245 and @xmath392 .
the widths of the lines are proportional to the strengths of the matrix elements for the transitions.,scaledwidth=45.0% ] we can find an expression for the @xmath403 afm level crossing at the induction @xmath404 to first order in the anisotropy interactions for a general @xmath0 spin dimer by equating @xmath405 to @xmath406 , yielding @xmath407\nonumber\\ & & + 3c_{s , s_1}[j_a\cos^2\theta+j_e\sin^2\theta\cos(2\phi)],\label{bstep}\\ c_{s , s_1}&=&\frac{[3 + 3s-5s^2 - 4s^3 + 4s_1(s_1 + 1)]}{2(2s+1)(2s+3)}.\end{aligned}\ ] ] this expression is consistent with those obtained for @xmath408 given by eqs . ( [ bstephalf ] ) , ( [ b11z])-([b21xy ] ) , and ( [ b52step1])-([b52step3 ] ) . in figs .
we note that @xmath409 contains the @xmath301-independent terms , @xmath410 , which distinguish @xmath183 from @xmath184 and @xmath225 from @xmath226 . in particular
, we note that the single - ion anisotropy interactions behave very differently with increasing step number than do the global anisotropy interactions , especially for large @xmath0 .
for the three cases we studied in detail , for @xmath4 , @xmath411 , so that the local anisotropy terms are irrelevant , for @xmath10 , @xmath412 and @xmath413 have different signs , and for @xmath20 as in fe@xmath21 dimers , the first three @xmath414 coefficients are @xmath415 , @xmath416 , and @xmath417 , respectively , the second being an order of magnitude smaller than the other two , and opposite in sign from the first . for @xmath22 dimers such as [ mn@xmath23@xmath21 ,
the first four @xmath418 are @xmath419 , @xmath420 , @xmath421 , and @xmath422 , which changes sign between @xmath269 and @xmath274 , where its magnitude is a minimum .
this is in sharp contrast to the global anisotropy interactions , for which the analogous coefficient @xmath423 increases monotonically with @xmath38 , independent of @xmath0 .
these differences should be possible to verify experimentally in careful low-@xmath7 experiments at high magnetic fields applied at various directions on single crystals of those @xmath20 fe@xmath21 and @xmath22 [ mn@xmath23@xmath21 dimers for which @xmath424 is sufficiently small . to aid in the analysis of experimental data
, we have extended this perturbative calculation to second order in each of the four principle @xmath88 .
since we expect the @xmath425 in most circumstances , this extension ought to be sufficient to accurately analyze most experimentally important samples .
since the @xmath426 magnitudes can become larger than unity for @xmath20 and especially for @xmath22 , we expect second order corrections to be significant in those cases .
to second order in the anisotropy interactions , the eigenstate energies @xmath427 are given in appendix c. we note that the @xmath428 contain divergences at @xmath429 and @xmath430 , so that near to those values , one would need to modify the perturbation expansion to take proper account of the degeneracies .
hence , the expressions for @xmath428 can not be used in the asymptotic expressions for the thermodynamics , eqs .
( [ zapprox])-([cvapprox ] ) . however , as the @xmath38th afm level crossing occurs approximately at @xmath431 , which is far from any divergences , we can safely use this second order expansion to obtain an expression for the level crossings second order in the anisotropy interaction energies .
we find @xmath432 where @xmath433 is given by eq .
( [ bstep ] ) . the full expression for the second term in eq . ( [ bstep2 ] ) is given in appendix d. from this expression , it is easy to see that for @xmath4 , @xmath434 , where @xmath435 and @xmath436 are given in appendix c. for @xmath199 , @xmath437 , and with @xmath438 , @xmath439 , in agreement with the expansion to second order in the @xmath88 of our exact formulas in eq .
( [ bstephalf ] ) . however , the second order functions have more complicated @xmath301 dependencies than do the first order @xmath404 in eq .
( [ bstep ] ) .
we have also explicitly checked each formula in appendix d for @xmath10 and @xmath440 .
thus , eq . ( [ bstep2 ] ) is a highly accurate expression for the full @xmath301 dependencies of all @xmath441 level crossings of a single crystal dimer of single ion spin @xmath0 . by superposing this non - universal level - crossing formula , eq .
( [ bstep2 ] ) , combined with the universal behavior presented in eqs .
( [ cvzeroes])-([slope ] ) , it is easy to use our results in accurate fits to experimental data at low temperatures and high magnetic induction strengths .
finally , we consider some cases of strong anisotropy interactions , in which one or more of the @xmath88 is comparable to @xmath1 in magnitude . the cases of interest are those for which the anisotropy interactions can remove the level crossing effects in afm dimers that give rise to the @xmath442 @xmath33 steps and @xmath34 double peaks
. these cases generally occur for strong fm anisotropy interactions , but there are some examples in which they can occur with strong afm anisotropy interactions .
we first consider the case of @xmath4 dimers , for which the situation can be analyzed analytically , at least for @xmath176 .
then , we consider the @xmath20 case numerically . for the @xmath4 dimer ,
the energies for @xmath443 are given by eqs .
( [ epsilon1])-([jxy ] ) .
we are interested in examining the cases in which the anisotropy is strong enough to cause the level crossing to disappear .
strong anisotropy with @xmath183 and @xmath184 having the same sign as @xmath1 do not differ significantly from the weak coupling cases , as they do not remove the level crossing , and do not affect significantly the heights of the magnetization step and the number and shapes of the @xmath34 double peaks , but just shift their positions , as for weak anisotropy .
there are then three afm ( @xmath5 ) cases of interest .
these are : ( 1 ) @xmath444 , ( 2 ) @xmath445 , and ( 3 ) @xmath446 .
the @xmath33 and @xmath34 curves for these cases with @xmath201 at the very low-@xmath7 value @xmath447 are shown as the solid , dotted , and dashed curves in fig . 20 and 21 , respectively . for case
( 1 ) , @xmath444 , the same four eigenvalues are obtained for @xmath199 and @xmath195 .
assuming @xmath5 , these are ranked from highest to lowest as @xmath448 note that @xmath449 only at @xmath450 ( where @xmath451 ) , @xmath452 decreases with increasing @xmath453 , and @xmath454 increases with increasing @xmath453 , so the levels get further apart with increasing @xmath453 . in this case , @xmath33 shown as the dotted curve in fig .
20 is broad , even at the very low @xmath7 value plotted .
in addition , @xmath34 , shown as the dotted curve in fig .
19 , exhibits a single schottky - like anomaly , but with a different shape than that of the standard schottky case , which arises from a splitting of the two lowest energies linear in @xmath453 at @xmath450 . in this case , the peak value of @xmath163 is 0.439229 , the same as the uniform double peak value , but the peak position is as @xmath455 , where @xmath456 . versus @xmath165 at @xmath447 with @xmath445 ( solid ) for both @xmath457 , @xmath444 ( dotted ) , and @xmath446 with @xmath199.(dashed ) for the @xmath4 afm dimer.,scaledwidth=45.0% ] versus @xmath165 at @xmath447 with @xmath445 ( solid ) for both @xmath458 , @xmath444 ( dotted ) , and @xmath446 with @xmath199 ( dashed ) for the @xmath4 afm dimer.,scaledwidth=45.0% ] next , we examine case ( 2 ) , @xmath445 , @xmath198 . in this case
the energies are different for the two field directions . for @xmath459
that @xmath460 at @xmath450 , these energies comprise a triply degenerate ( or spin 1 ) ground state at 0 and an excited @xmath461 state . with increasing @xmath453 , @xmath452 decreases linearly , and @xmath454 increases linearly , crossing @xmath454 at @xmath462 , which does not affect the low-@xmath7 thermodynamics . in this case
, @xmath33 is a sharp step at @xmath393 , shown as a solid curve in fig .
20 , and @xmath34 is a conventional spin 1 schottky anomaly at low @xmath453 , as shown by the left solid curve in fig . 21 . from eqs .
( [ case2z3])-([case2z1 ] ) , @xmath463 has a maximum value at 0.637203 at @xmath464 .
note that the larger maximum value than in the other @xmath34 cases arises from the higher degeneracy as @xmath465 . for @xmath466 ,
the energies for case ( 2 ) are @xmath467 at @xmath450 , the ground state is again triply degenerate with energy 0 , and the excited state has energy @xmath424 . with increasing @xmath453 , @xmath452 separates from @xmath468 and @xmath469 , curving below them , and @xmath454 increases quadratically . for this induction direction
there is no level crossing , but there is some @xmath470 anisotropy at finite @xmath7 , because of the quadratic versus linear @xmath453 dependencies .
@xmath163 has the maximum value 0.761802 at @xmath471 , where @xmath472 , which is noticeably different from the case with @xmath199 .
the @xmath33 and @xmath34 curves are the right solid curves in figs .
20 and 21 , respectively .
now we consider case ( 3 ) , @xmath473 , @xmath200 .
again , there are slight differences for @xmath474 and @xmath475 . for @xmath474 ,
the eigenstate energies are @xmath476 at @xmath450 , the ground state is doubly degenerate at energy 0 .
the other two states have energies @xmath424 and @xmath477 , respectively . as @xmath453 increases
, @xmath452 decreases below @xmath468 , and @xmath454 increases monotonically .
hence , there is no level crossing , but at low @xmath7 , two levels are relevant , in a fashion slightly different from a standard schottky specific heat anomaly .
the low-@xmath7 @xmath33 and @xmath34 curves for this induction direction are shown as the dashed curves in figs .
20 and 21 , respectively .
we note that they are rather similar from those of case ( 1 ) , but have the broadest @xmath33 increase and @xmath34 peak of the three cases considered .
@xmath163 has the peak value 0.439229 which is the same as the uniform double - peak value , but its position is at @xmath478 , where @xmath479 . for @xmath480 , we have @xmath481 at @xmath450 , the eigenstates are the same as for the field in the @xmath61 direction , with a doubly degenerate ground state at 0 , and excited states at @xmath424 and @xmath477 . however , the field dependence is a bit different than for the @xmath61 direction . again
the ground state decreases with increasing @xmath453 , but the energy scale of the curvature is @xmath482 instead of @xmath424 .
in addition , @xmath469 can cross @xmath454 at @xmath483 .
this level crossing is irrelevant to the low-@xmath7 behavior , however , but modifies the @xmath33 curve somewhat at finite @xmath7 .
versus @xmath165 at @xmath447 with @xmath484 for the afm @xmath20 dimer . the cases @xmath485 ( solid ) , @xmath486(dashed ) , @xmath487 ( dotted ) , @xmath488 ( dot - dashed ) , @xmath489 ( short dashed ) , and @xmath490 ( short dotted ) are shown.,scaledwidth=45.0% ] we now consider some cases of strong anisotropy interactions in the @xmath20 dimer . in fig
22 , we present low-@xmath7 plots of @xmath33 with @xmath484 , where @xmath491 and 0.5 , as indicated . for the weak coupling case @xmath485 , the steps have uniform height . with increasing @xmath229 ,
the curves shift monotonically to lower @xmath453 values .
slight deviations in the uniformity of the step height are detectable in the first two steps for @xmath486 , but the nonuniformity in the step height is progressively more pronounced for @xmath487 and 0.4 , respectively . however , for @xmath489 , the individual steps have completely disappeared , and are replaced by what appears to be a single , broad step centered at @xmath492 . at @xmath490 , @xmath33 rises to its maximum value with a very steep slope at @xmath450 , not showing any evidence for any level crossings . versus @xmath165 at @xmath447 with @xmath493 for the afm @xmath20 dimer .
the cases @xmath485 ( solid ) , @xmath486 ( dashed ) , @xmath487 ( dotted ) , @xmath488 ( dot - dashed ) , @xmath490 ( short dashed ) , and @xmath494 ( short dotted ) are shown.,scaledwidth=45.0% ] versus @xmath165 at @xmath447 with @xmath495 for the afm @xmath20 dimer .
the cases @xmath485 ( solid ) , @xmath486 ( dashed ) , @xmath487 ( dotted ) , @xmath488 ( dot - dashed ) , and @xmath490 ( short dotted ) are shown.,scaledwidth=45.0% ] next , we investigated the effects of large @xmath225 . in fig .
23 we plotted @xmath164 versus @xmath165 for @xmath493 , with @xmath485 ( solid ) , 0.2 ( dashed ) , 0.3 ( dotted ) , 0.4 ( dot - dashed ) , 0.5 ( short dashed ) , and 0.6 ( short dotted ) .
we see the the effects of strong @xmath225 are very different from those of strong @xmath183 and @xmath184 shown in fig .
22 . instead of the step positions decreasing monotonically with increasing @xmath496 , as @xmath225 increases
, the highest three step positions decrease with increasing @xmath225 , but the lowest two step positions increase with increasing @xmath225 .
we now consider the case of large @xmath179 . in fig .
24 , we plotted @xmath164 versus @xmath497 at @xmath447 for the @xmath20 afm dimer with @xmath495 , where @xmath498 and 0.5 .
these results are shown respectively as the solid , dashed , dotted , dot - dashed , and short dotted curves . unlike the strong global anisotropy case pictured in fig .
22 and the strong local axial anisotropy case pictured in fig .
23 , these curves do not shift significantly to lower @xmath453 values with increasing @xmath499 , but the step shapes are greatly altered . with @xmath488 , the first two steps are hard to discern , but rounded remnants of the three last steps are evident . at @xmath490 ,
the third step is hard to detect , and the remnants of the last two steps are reduced in magnitude .
we remark that the curve with @xmath500 is indistinguishable from the dotted curve for @xmath487 , so that in this case , strong @xmath501 of either sign can grossly alter the @xmath502 step behavior . versus @xmath165 at @xmath447 for the afm @xmath20 dimer .
the cases @xmath484 for @xmath489 ( solid ) and @xmath490 ( dot - dashed ) , @xmath493 for @xmath488 ( solid ) and 0.5 ( short - dashed ) and @xmath503 ( thick dashed ) are shown.,scaledwidth=45.0% ] in fig .
25 , we show @xmath163 versus @xmath497 low-@xmath7 curves for some of the afm @xmath20 cases pictured in fig .
the solid and dotted curves on the left - hand side of the figure correspond to @xmath484 with @xmath489 and 0.5 , respectively , the solid and dot - dashed curves in the central portion of the figure correspond to @xmath504 and 0.5 , respectively , and the dashed curve running throughout the domain pictured corresponds to @xmath503 .
we note that for the global anisotropy case shown , strong anisotropy pushes the @xmath34 peaks to lower @xmath453 , and squeezes them together , so that they tend to overlap . at the limiting case @xmath505 ,
the individual level crossings have been eliminated , and all ten of the peaks are combined into a single schottky anomaly . for the local axial anisotropy interaction , strong fm anisotropic interaction squeeze the @xmath34 together in the middle of the domain , with the limiting case of a single peak at @xmath506 . for the strong azimuthal anisotropic exchange interaction case ,
however , strong fm anisotropy shifts the double peak positions only slightly , and broadens the double peaks , so that only the highest set is clearly discernable , even at the very low @xmath7 value plotted . in each case , the deviation of the peak height from the two - level prediction , eq . ( [ peak ] ) , is substantial , indicating more than two levels are important at low @xmath7 , as for the three - level system with @xmath4 and @xmath445 .
thus , the different types of strong spin anisotropy lead to clearly different low-@xmath7 @xmath34 behaviors for @xmath507 dimers .
in summary , we solved for the low - temperature magnetization and specific heat of equal spin @xmath0 antiferromagnetic dimer single molecule magnets , including the most general set of anisotropic spin exchange interactions quadratic in the spin operators .
the magnetization and specific heat exhibit steps and zeroes , respectively , at the non - universal level - crossing induction values @xmath37 , but the magnetization steps and their midpoint slopes , plus the two peaks surrounding the specific heat zeroes all exhibit universal behavior at sufficiently low temperatures to first order in the anisotropy interaction strengths .
strong anisotropy interactions generally lead to highly non - universal behavior in antiferromagnetic dimers .
local ( or single - ion ) anisotropy interactions lead to low - temperature magnetization step plateaus that have a much richer variation with the magnetic induction @xmath15 than do those obtained from global anisotropy interactions , provided that @xmath507 . for the most general quadratic anisotropic spin interactions at an arbitrary @xmath15 , we derived simple , asymptotic analytic expressions for the low - temperature magnetization , specific heat , inelastic neutron scattering cross - section , and electron paramagnetic response susceptibility , which are accurate for weak anisotropy .
we also derived an accurate expression for the level - crossing induction @xmath37 , enabling fast and accurate fits to experimental data .
there were two low-@xmath7 @xmath33 studies of fe@xmath21 dimers.@xcite for @xmath508-oxalatotetrakis(acetylacetonato)fe@xmath21 , all five peaks in @xmath509 were measured in pulsed magnetic fields @xmath510 .
these evenly spaced peaks indicated little , if any , spin anisotropy effects.@xcite on the other hand , studies of the first 2 - 3 @xmath509 peaks in powdered samples of [ fe(salen)cl]@xmath21 , where salen is @xmath511-ethylenebis(salicylideneiminato ) , were much more interesting.@xcite these data showed a broad first peak at @xmath512 t that was only partially resolvable into two separate peaks , followed by a sharp second peak at @xmath513 t , consistent with local axial anisotropy of strength @xmath514 , as obtained from the derivatives of the curves shown in fig . 13 . from eqs .
( [ b52step1])-([b52step3 ] ) , one could also have @xmath515 .
these amounts might perhaps be combined with a smaller @xmath516 , obtained from the derivatives of the curves pictured in fig .
14 . without a detailed single crystal study with magnetic fields along different crystal directions , it is impossible to determine the relative amounts of @xmath225 and @xmath226 that would best fit the data
however , the existing data on [ fe(salen)cl]@xmath21 appear to be inconsistent with a predominant global anisotropy interaction of either type , as obtained from the derivatives of the curves shown in figs .
11 and 12 .
only single crystal studies could determine if a small amount of such global anisotropy interactions were present in addition to one or more presumably larger local spin anisotropy interactions .
in comparing the two materials cited above , it appears that the interaction of a cl@xmath517 ion neighboring each fe@xmath518 ion leads to strong local ( or single - ion ) anisotropy effects . in order to verify this hypothesis and to elucidate the details of the interactions , further experiments using single crystals in different field orientations on this and related fe@xmath21 dimers with 1 - 3 similarly bonded cl@xmath517 ions
are urged.@xcite we also urge single crystal data on some of the @xmath22 [ mn@xmath23@xmath21 dimers , @xcite , as well as on @xmath4 dimers lacking in predicted local spin anisotropy effects .
@xciteto aid in the fits , we derived simple , useful formulas for the magnetization and specific heat at low temperature and sufficiently large magnetic induction .
more important , we derived accurate analytic formulas for the electron paramagnetic resonance susceptibility @xmath519 and the inelastic neutron scattering @xmath36 , which allow for a precise determination of the various microscopic anisotropy energies . with local anisotropy interactions ,
the total spin @xmath38 is not a good quantum number , potentially modifying our understanding of quantum tunneling processes in single molecule magnets . it might also be possible to fit a variety of experimental results using a smaller , consistent set of model parameters.@xcite we emphasize that the study of dimer single molecule magnets , for which the most general anisotropic quadratic exchange interactions can be solved exactly , may be our best hope for attaining a more fundamental understanding of the underlying physics of more general single molecule magnets .
we thank the max - planck - institut fr physik komplexer systeme , dresden , germany , the university of north dakota , grand forks , nd , usa , and talat s. rahman for their kind hospitality and support .
this work was supported by the netherlands foundation for the fundamental research of matter and by the nsf under contract ner-0304665 .
for completeness , we present the results of the single - ion operators for general dimers consisting of magnetic ions with spins @xmath0 and @xmath520 .
we find @xmath521[s^2-(s_1-s_2)^2]}{(4s^2 - 1)s^2}},\nonumber\\ & & \\
\xi_{s , s_1,s_2}&=&\frac{s_1(s_1 + 1)-s_2(s_2 + 1)}{s(s+1)}.\end{aligned}\ ] ] note that for @xmath30 , @xmath522 and @xmath523 reduces to eq .
( [ eta ] ) for @xmath524 . also , note that these relations preserve the global spin properties of @xmath525 and @xmath526 , @xmath44 and @xmath527 , as for equal spin dimers .
in addition , the spins on asymmetric dimers with @xmath528 would be expected to also have different local anisotropy interactions , so that @xmath73 and @xmath75 would become @xmath529 we first present the numerators of the exact expressions for the specific heat with @xmath4 and @xmath176 for @xmath170 .
we have @xmath530}\nonumber\\ & & + f_{x , y}\sinh(\beta f_{x , y})e^{-\beta j_{x , y}}\nonumber\\ & & \times[(j+j_{x , y})e^{-\beta j}+(j_{x , y}-2j_{y , x})e^{2\beta j_{y , x}}]\nonumber\\ & & + \frac{1}{2}e^{-\beta j_{x , y}}\cosh(\beta f_{x , y})\nonumber\\ & & \times\bigl([(j+j_{x , y})^2+f_{x , y}^2]e^{-\beta j}\nonumber\\ & & + [ ( 2j_{y , x}-j_{x , y})^2+f_{x , y}^2]e^{2\beta j_{y , x}}\bigr),\\ { \cal n}_z&=&f_z^2\cosh(2\beta f_z)+\frac{j^2}{4}e^{-\beta(j+2j_b)}\nonumber\\ & & + \frac{1}{2}\cosh(\beta f_z)\bigl(\delta_z(j_b^2+f_z^2)\nonumber\\ & & \qquad+e^{-\beta(j+j_b)}j(j+2j_b)\bigr)\nonumber\\ & & + f_z\sinh(\beta f_z)\bigl(j_b\delta_z+je^{-\beta(j+j_b)}\bigr).\end{aligned}\ ] ] in the remainder of this appendix , we provide some details of our exact results for @xmath10 . the cubic equation for the three @xmath187 eigenvalues
is given by @xmath531\nonumber\\ & & + ( j_a - j_b)[b^2\cos^2\theta+(j_d - j_e)^2]\nonumber\\ & & -(j_d - j_e)(b^2\sin^2\theta\cos(2\phi).\end{aligned}\ ] ] for @xmath532 , the cubic equation is easily solved to yield @xmath533 for @xmath534 , we have @xmath535,\end{aligned}\ ] ] where @xmath536 ( @xmath537 ) corresponds to the upper ( lower ) sign .
the six eigenvalues for the mixed @xmath213 states satisfy @xmath538 we first define @xmath539 then the @xmath540 are the eigenvalues of the hermitian matrix @xmath541 given by @xmath542 the resulting sixth order polynomial for the @xmath540 is given elsewhere.@xcite we note that for @xmath169 , @xmath541 is block diagonal , breaking up into matrices of ranks two and four .
these cases are discussed in detail elsewhere.@xcite here we only present the simplest data for which the eigenvalues are determined either by linear or by quadratic equations .
when only one of the @xmath543 , the eigenvalues for @xmath201 simplify considerably . for @xmath544 , we have @xmath545 for @xmath210 , we find @xmath546 for @xmath547 , the eigenvalues can also be found analytically , @xmath548 we note that the ground state energy in this case is @xmath549 which explicitly involves mixing of the @xmath550 and the @xmath551 states . at the first level crossing with @xmath201
, we have for @xmath547 and the other @xmath232 , @xmath552^{1/2}\bigr).\label{jastep1}\end{aligned}\ ] ] we note that the first level crossing for @xmath553 for @xmath554 is equivalent to that of @xmath4 , given by eq .
( [ bstephalf ] ) . at the second level crossing ,
simple formulas are only obtained for @xmath199 with one @xmath543 . for @xmath199 and @xmath547 , @xmath544 and @xmath210 ,
respectively , we have @xmath555^{1/2}\bigr)^{1/2}.\end{array}\right.\label{secondstep}\end{aligned}\ ] ]
the rotation from the crystal representation to the induction representation is obtained from @xmath556 leading to @xmath290.@xcite this operation is equivalent to a rotation by @xmath557 about the @xmath61 axis , a rotation by @xmath558 about the transformed @xmath559 axis , and then a rotation by @xmath560 about the transformed @xmath61 axis.@xcite in effect , in using the above rotation matrix , we made the arbitrary choice that the rotated @xmath61 axis lies in the @xmath561 plane . after the above rotation , it is still possible to rotate the crystal by an arbitrary angle @xmath562 about the @xmath563 axis , keeping @xmath564 .
hence , there are in effect an infinite number of equivalent rotations leading to @xmath290 .
the resulting hamiltonian matrix will then have off - diagonal elements that depend upon @xmath562 .
however , all such rotations necessarily lead to the identical , @xmath562-independent , set of eigenvalues of the resulting diagonalized hamiltonian matrix . we have explicitly checked that the above rotation gives the exact cubic expression , eq .
( [ halfeigen ] ) , for the @xmath4 eigenvalues , and also leads to the correct eigenstate energies second order in each of the @xmath88 for @xmath10 .
we also showed explicitly that @xmath562 does not enter the eigenstate energies second order in @xmath183 for arbitrary @xmath565 .
the global axial and azimuthal anisotropy interactions in the rotated frame are @xmath566,\nonumber\\\end{aligned}\ ] ] where @xmath567 is the anticommutator .
the operations of the rotated global anisotropy interactions upon these states may be written as @xmath568\sin^2\theta\bigr)| \varphi_s^m\rangle\nonumber\\ & & -\sin(2\theta)\sum_{\sigma=\pm1}(2m+\sigma)a_s^{\sigma m}|\varphi_s^{m+\sigma}\rangle\nonumber\\ & & + \sin^2\theta\sum_{\sigma=\pm1}f_s^{\sigma m}|\varphi_s^{m+2\sigma}\rangle\bigr ] , \label{hbprime}\end{aligned}\ ] ] and @xmath569|\varphi_s^m\rangle\nonumber\\ & & + 2\sin\theta\sum_{\sigma=\pm1}(2m+\sigma)a_s^{\sigma m}\nonumber\\ & & \times[\cos\theta\cos(2\phi)+i\sigma\sin(2\phi)]|\varphi_s^{m+\sigma}\rangle\nonumber\\ & & + \sum_{\sigma=\pm1}f_s^{\sigma m}\bigl[(1+\cos^2\theta)\cos(2\phi)\nonumber\\ & & + 2i\sigma\cos\theta\sin(2\phi)\bigr]|\varphi_s^{m+2\sigma}\rangle\bigr),\label{hdprime}\end{aligned}\ ] ] where @xmath570 is defined by eq .
( [ fsx ] ) .
we note that in this representation , both @xmath82 and @xmath85 preserve the global spin quantum number @xmath38 , but allow @xmath571 transitions . with regard to the local spin anisotropy terms in the rotated coordinate system ,
we write @xmath572 and @xmath573,\end{aligned}\ ] ] where @xmath574 and@xmath575 the operations of these interactions are given by @xmath576 where @xmath577(1-\alpha_{s , s_1}),\nonumber\\ & & \\
n_{s , s_1}^{m,\sigma'}&=&\sum_{\sigma=\pm1}c_{s+(\sigma'+1)/2,s_1}^{-\sigma\sigma'm-(\sigma'+1)/2 } c_{s+(3\sigma'+1)/2,s_1}^{\sigma\sigma'm+(\sigma'-1)/2},\label{nss1msigma}\\ p_{s , s_1}^{m,\sigma}&=&2a_s^{\sigma m}(2m+\sigma)\alpha_{s , s_1},\\ \noalign{\rm and } r_{s , s_1}^{m,\sigma,\sigma'}&=&c_{s+(\sigma'+1)/2,s_1}^{-m\sigma\sigma'-(\sigma'+1)/2}d_{s+(3\sigma'+1)/2,s_1}^{m+\sigma}\nonumber\\ & & + c_{s+(3\sigma'+1)/2,s_1}^{-m\sigma\sigma'-(\sigma'+1)/2}d_{s+(\sigma'+1)/2,s_1}^m,\label{rss1xsigma}\end{aligned}\ ] ] where @xmath524 , @xmath578 , @xmath579 , @xmath580 , and @xmath581 are given by eqs .
( [ eta ] ) and ( [ gss1m])-([lss1 ] ) , respectively .
we note that for @xmath582 , @xmath583 .
here we outline the derivation and provide the important results for the low-@xmath7 inelastic neutron scattering cross - section , @xmath36 . at low @xmath7 , we set @xmath584 given by eqs .
( [ e0 ] ) and ( [ esm1 ] ) .
since this energy was evaluated in the induction representation , it is easiest to evaluate @xmath585 by choosing the axes @xmath586 , where @xmath170 .
the @xmath587 can be obtained from the @xmath588 using the inverse of the rotation matrix in eq .
( [ rotation ] ) .
then , for @xmath589 in the crystal representation , we have @xmath590 in the induction representation , where @xmath591 then , expanding the wave functions to first order in the anisotropy interactions , @xmath592}|\varphi_{s+2\sigma'}^m\rangle\nonumber\\ & & + \sum_{\sigma,\sigma'=\pm1}\bigl(\frac{{\cal x}_{s , s_1}^{m,\sigma,\sigma'}}{\sigma\gamma b+j[\sigma'(2s+1)+2]}|\varphi_{s+2\sigma'}^{m+\sigma}\rangle\nonumber\\ & & + \frac{{\cal y}_{s , s_1}^{m,\sigma,\sigma'}}{2\sigma\gamma b+j[\sigma'(2s+1)+2]}|\varphi_{s+2\sigma'}^{m+2\sigma}\rangle\bigr)+\ldots,\nonumber\\\end{aligned}\ ] ] where @xmath593,\label{calu}\\ { \cal v}_{s , s_1}^{m,\sigma}(\theta,\phi)&=&-\frac{1}{4}f_s^{\sigma m}\bigl[\tilde{j}_{b , a}^{s , s_1}\sin^2\theta\nonumber\\ & & + \tilde{j}_{d , e}^{s , s_1}(1+\cos^2\theta)\cos(2\phi)\nonumber\\ & & + 2i\sigma\tilde{j}_{d , e}^{s , s_1}\cos\theta\sin(2\phi)\bigr],\\ { \cal w}_{s , s_1}^{m,\sigma'}(\theta,\phi)&=&-\frac{1}{2}h_{s , s_1}^{m,\sigma'}[j_a\cos^2\theta+j_e\sin^2\theta\cos(2\phi)]\nonumber\\ & & + \frac{1}{8}n_{s , s_1}^{m,\sigma'}\sin^2\theta[j_a - j_e\cos(2\phi)],\\ { \cal x}_{s , s_1}^{m,\sigma,\sigma'}(\theta,\phi)&=&-\frac{\sigma\sigma'}{8}r_{s , s_1}^{m,\sigma,\sigma'}\bigl(\sin(2\theta)[j_a - j_e\cos(2\phi)]\nonumber\\ & & \qquad-2i\sigma j_e\sin\theta\sin(2\phi)\bigr),\\ \noalign{\rm and } & & \nonumber\\ { \cal y}_{s , s_1}^{m,\sigma,\sigma'}(\theta,\phi)&=&-\frac{1}{8}k_{s , s_1}^{\sigma m,\sigma'}\bigl[j_a\sin^2\theta\nonumber\\ & & \qquad+j_e\bigl((1+\cos^2\theta)\cos(2\phi)\nonumber\\ & & \qquad+2i\sigma\cos\theta\sin(2\phi)\bigr)\bigr],\label{caly}\end{aligned}\ ] ] where @xmath579 and @xmath580 are given by eqs . ( [ hss1msigma ] ) and ( [ kss1xsigma ] ) , respectively , @xmath594 and @xmath595 are given by eqs .
( [ nss1msigma ] ) and ( [ rss1xsigma ] ) , respectively , and @xmath384 and @xmath385 are given by eqs .
( [ tildejba ] ) and ( [ tildejde ] ) , respectively .
we note that for @xmath582 , @xmath583 .
now , we evaluate the matrix elements @xmath596 including the leading corrections to the wave functions due to the anisotropy interactions .
we then find to first order in the anisotropy interactions , @xmath597,\end{aligned}\ ] ] and @xmath598 is obtained from @xmath599 by multiplying the entire quantity inside the summation over @xmath600 by @xmath601 .
now , it is convenient to break up the contributions from @xmath602 and @xmath603 into their symmetric and antisymmetric parts , with @xmath604 etc .
then , the combined contributions from @xmath605 and @xmath606 arising from the first order perturbations of the wave functions in addition to the hartree approximation are @xmath607 where @xmath608 are the contributions to the @xmath609 through second order in the anisotropy interactions , respectively .
the terms to zeroth order in the anisotropy interactions are given by eqs .
( [ f10])-([f40 ] ) in the text .
the non - vanishing first order terms are @xmath610},\nonumber\\ & & \\ q_4^{\sigma,\sigma'}&=&\sum_{\overline{\sigma}=\pm1}\frac{\overline{\sigma}c_{s+(\sigma'+1 + 2\overline{\sigma}\sigma')/2,s_1}^{\sigma\sigma'm+(\sigma'-1)/2 } h_{s+\sigma'(1-\overline{\sigma})/2,s_1}^{m+\sigma(1-\overline{\sigma})/2,\overline{\sigma}\sigma'}}{2[\sigma'(2s+1)+1+\overline{\sigma}]},\nonumber\\ & & \\ f_1^{(1)}(\theta,\phi)&=&\frac{1}{j}\bigl(j_a-3[j_a\cos^2\theta+j_e\sin^2\theta\cos(2\phi)]\bigr),\nonumber\\\end{aligned}\ ] ] where we suppress the @xmath342 arguments of the @xmath611 functions for simplicity of presentation . the contributions @xmath612 second order in the anisotropy interactions may be written as @xmath613 we note that we have suppressed the @xmath614 dependencies of the @xmath615 for brevity .
these functions are given by @xmath616\bigl(g_{1,1}^{\sigma}\bigr)^2,\\ { \cal a}_1^{2,0}&=&\bigl[f_4^{(2)}(\theta,\phi)+2\alpha_{s , s_1}f_5^{(2)}(\theta,\phi)\nonumber\\ & & + \alpha^2_{s , s_1}f_6^{(2)}(\theta,\phi)\bigr]\bigl(g_{1,2}^{\sigma}\bigr)^2,\\ { \cal a}_1^{1,1}&=&f_3^{(2)}(\theta,\phi)\bigl(g_{1,4}^{\sigma,\sigma'}\bigr)^2,\\ { \cal a}_1^{1,2}&=&f_6^{(2)}(\theta,\phi)\bigl(g_{1,5}^{\sigma,\sigma'}\bigr)^2,\end{aligned}\ ] ] @xmath617+\bigl(g_{2,4}^{\sigma,\sigma',-\sigma'}\bigr)^2f_3^{(2)}(\theta,\phi)\nonumber\\ & & + 2g_{2,1}^{\sigma,\sigma'}g_{2,4}^{\sigma,\sigma',-\sigma'}\bigl[f_2^{(2)}(\theta,\phi)+\alpha_{s , s_1}f_3^{(2)}(\theta,\phi)\bigr],\nonumber\\ & & \\ { \cal a}_2^{2,0}&=&\bigl(g_{2,2}^{\sigma,\sigma'}\bigr)^2\bigl[f_4^{(2)}(\theta,\phi)+2\alpha_{s , s_1}f_5^{(2)}(\theta,\phi)\nonumber\\ & & + \alpha^2_{s , s_1}f_6^{(2)}(\theta,\phi)\bigr]+\bigl(g_{2,5}^{\sigma,\sigma',-\sigma'}\bigr)^2f_6^{(2)}(\theta,\phi)\nonumber\\ & & + 2g_{2,2}^{\sigma,\sigma'}g_{2,5}^{\sigma,\sigma',-\sigma'}\bigl[f_5^{(2)}(\theta,\phi)+\alpha_{s , s_1}f_6^{(2)}(\theta,\phi)\bigr],\nonumber\\ & & \\ { \cal a}_2^{1,1}&=&f_3^{(2)}(\theta,\phi)\bigl(g_{2,4}^{\sigma,\sigma',\sigma'}\bigr)^2,\\ { \cal a}_2^{2,1}&=&f_6^{(2)}(\theta,\phi)\bigl(g_{2,5}^{\sigma,\sigma',\sigma'}\bigr)^2,\end{aligned}\ ] ] @xmath618\bigl(g_{3,1}^{\sigma,\pm\sigma}\bigr)^2,\\ { \cal a}_3^{2\pm1,0}&=&\bigl[f_4^{(2)}(\theta,\phi)+2\alpha_{s , s_1}f_5^{(2)}(\theta,\phi)\nonumber\\ & & + \alpha^2_{s , s_1}f_6^{(2)}(\theta,\phi)\bigr]\bigl(g_{3,2}^{\sigma,\pm\sigma}\bigr)^2,\\ { \cal a}_3^{1\pm1,1}&=&f_3^{(2)}(\theta,\phi)\bigl(g_{3,4}^{\sigma,\sigma',\pm\sigma}\bigr)^2,\\ { \cal a}_3^{1,1}&=&\bigl(g_{3,3}^{\sigma,\sigma'}\bigr)^2f_7^{(2)}(\theta,\phi)+\bigl(g_{3,5}^{\sigma,\sigma',-\sigma}\bigr)^2f_6^{(2)}(\theta,\phi)\nonumber\\ & & + 2g_{3,3}^{\sigma,\sigma'}g_{3,5}^{\sigma,\sigma',-\sigma}f_9^{(2)}(\theta,\phi)\bigr],\\ { \cal a}_3^{3,1}&=&f_6^{(2)}(\theta,\phi)\bigl(g_{3,5}^{\sigma,\sigma',\sigma}\bigr)^2,\end{aligned}\ ] ] @xmath619\nonumber\\ & & + \bigl(g_{4,4}^{\sigma,\sigma',\pm\sigma,-\sigma'}\bigr)^2f_3^{(2)}(\theta,\phi)\nonumber\\ & & + 2g_{4,1}^{\sigma,\sigma',\pm\sigma}g_{4,4}^{\sigma,\sigma',\pm\sigma,-\sigma'}\nonumber\\ & & \times\bigl[f_2^{(2)}(\theta,\phi)+\alpha_{s , s_1}f_3^{(2)}(\theta,\phi)\bigr],\\ { \cal a}_4^{1,0}&=&\bigl(g_{4,2}^{\sigma,\sigma',-\sigma}\bigr)^2\bigl[f_4^{(2)}(\theta,\phi)\nonumber\\ & & + 2\alpha_{s , s_1}f_5^{(2)}(\theta,\phi)+\alpha^2_{s , s_1}f_6^{(2)}(\theta,\phi)\bigr]\nonumber\\ & & + \bigl(g_{4,5}^{\sigma,\sigma',-\sigma,-\sigma'}\bigr)^2f_6^{(2)}(\theta,\phi)\nonumber\\ & & + 2g_{4,2}^{\sigma,\sigma',-\sigma}g_{4,5}^{\sigma,\sigma',-\sigma,-\sigma'}\nonumber\\ & & \times\bigl[f_5^{(2)}(\theta,\phi)+\alpha_{s , s_1}f_6^{(2)}(\theta,\phi)\bigr]\nonumber\\ & & + \bigl(g_{4,3}^{\sigma,\sigma',-\sigma'}\bigr)^2f_7^{(2)}(\theta,\phi)+2g_{4,3}^{\sigma,\sigma',-\sigma'}\nonumber\\ & & \times\bigl(g_{4,2}^{\sigma,\sigma',-\sigma}\bigl[f_8^{(2)}(\theta,\phi)+\alpha_{s , s_1}f_{9}^{(2)}(\theta,\phi)\bigr]\nonumber\\ & & + g_{4,5}^{\sigma,\sigma',-\sigma,-\sigma'}f_{9}^{(2)}(\theta,\phi)\bigr),\nonumber\\ & & \\
{ \cal a}_4^{3,0}&=&\bigl(g_{4,2}^{\sigma,\sigma',\sigma}\bigr)^2\bigl[f_4^{(2)}(\theta,\phi)\nonumber\\ & & + 2\alpha_{s , s_1}f_5^{(2)}(\theta,\phi)+\alpha^2_{s , s_1}f_6^{(2)}(\theta,\phi)\bigr]\nonumber\\ & & + \bigl(g_{4,5}^{\sigma,\sigma',\sigma,-\sigma'}\bigr)^2f_6^{(2)}(\theta,\phi)+2g_{4,2}^{\sigma,\sigma',\sigma}g_{4,5}^{\sigma,\sigma',\sigma,-\sigma'}\nonumber\\ & & \times\bigl[f_5^{(2)}(\theta,\phi)+\alpha_{s , s_1}f_6^{(2)}(\theta,\phi)\bigr],\\ { \cal a}_4^{1\pm1,1}&=&f_3^{(2)}(\theta,\phi)\bigl(g_{4,4}^{\sigma,\sigma',\pm\sigma,\sigma'}\bigr)^2,\\ { \cal a}_4^{1,1}&=&\bigl(g_{4,5}^{\sigma,\sigma',-\sigma,\sigma'}\bigr)^2f_6^{(2)}(\theta,\phi)\nonumber\\ & & + \bigl(g_{4,3}^{\sigma,\sigma',\sigma'}\bigr)^2f_7^{(2)}(\theta,\phi)\nonumber\\ & & + 2g_{4,3}^{\sigma,\sigma',\sigma'}g_{4,5}^{\sigma,\sigma',-\sigma,\sigma'}f_9^{(2)}(\theta,\phi),\\ { \cal a}_4^{3,1}&=&f_6^{(2)}(\theta,\phi)\bigl(g_{4,5}^{\sigma,\sigma',\sigma,\sigma'}\bigr)^2,\end{aligned}\ ] ] where the second order angular functions are given by @xmath620 ^ 2\cos^2\theta\nonumber\\ & & \qquad\qquad+j_d^2\sin^2(2\phi)\bigr),\label{f1}\\ f_2^{(2)}(\theta,\phi)&=&\frac{\sin^2\theta}{j^2}\bigl([j_b - j_d\cos(2\phi)][j_a - j_e\cos(2\phi)]\nonumber\\ & & \qquad\times\cos^2\theta + j_dj_e\sin^2(2\phi)]\bigr),\\ f_3^{(2)}(\theta,\phi)&=&\frac{\sin^2\theta}{j^2}\bigl([j_a-\cos(2\phi)j_e]^2\cos^2\theta\nonumber\\ & & \qquad\qquad+j_e^2\sin^2(2\phi)\bigr),\\ f_4^{(2)}(\theta,\phi)&=&\frac{1}{j^2}\bigl(j_b\sin^2\theta+j_d(1+\cos^2\theta)\cos(2\phi)\bigr)^2\nonumber\\ & & + 4\frac{j_d^2}{j^2}\cos^2\theta\sin^2(2\phi),\\ f_5^{(2)}(\theta,\phi)&=&\frac{1}{j^2}[j_b\sin^2\theta+j_d(1+\cos^2\theta)\cos(2\phi)]\nonumber\\ & & \times[j_a\sin^2\theta+j_e(1+\cos^2\theta)\cos(2\phi)]\nonumber\\ & & + 4\frac{j_dj_e}{j^2}\cos^2\theta\sin^2(2\phi),\\ f_6^{(2)}(\theta,\phi)&=&\frac{1}{j^2}\bigl(j_a\sin^2\theta+j_e(1+\cos^2\theta)\cos(2\phi)\bigr)^2\nonumber\\ & & + 4\frac{j_e^2}{j^2}\cos^2\theta\sin^2(2\phi),\label{f6}\\ f_7^{(2)}(\theta,\phi)&=&\frac{1}{j^2}\bigl(j_a-3[j_a\cos^2\theta+j_e\sin^2\theta\cos(2\phi)]\bigr)^2,\label{f7}\nonumber\\ f_8^{(2)}(\theta,\phi)&=&\frac{1}{j^2}\bigl(j_a-3[j_a\cos^2\theta+j_e\sin^2\theta\cos(2\phi)]\bigr)\nonumber\\ & & \times \bigl(j_b\sin^2\theta+j_d(1+\cos^2\theta)\cos(2\phi)\bigr),\\ f_9^{(2)}(\theta,\phi)&=&\frac{1}{j^2}\bigl(j_a-3[j_a\cos^2\theta+j_e\sin^2\theta\cos(2\phi)]\bigr)\nonumber\\ & & \times \bigl(j_a\sin^2\theta+j_e(1+\cos^2\theta)\cos(2\phi)\bigr),\label{f9}\end{aligned}\ ] ] and the constants are given by @xmath621\bigr)},\nonumber\\ & & \\ g_{1,5}^{\sigma,\sigma'}&=&\frac{k_{s , s_1}^{\sigma m,\sigma'}}{4\bigl(2\sigma\gamma b+j[\sigma'(2s+1)+2]\bigr)},\nonumber\\ & & \\ g_{2,1}^{\sigma,\sigma'}&=&\frac{j\sigma(2m+\sigma)}{2\gamma b}\sum_{\overline{\sigma}=\pm1}\bigl(\overline{\sigma}a_{s-\sigma'(\overline{\sigma}-1)/2}^{\sigma m}\nonumber\\ & & \qquad\times d_{s+(1+\sigma')/2,s_1}^{m+\sigma(\overline{\sigma}+1)/2}\bigr),\\ g_{2,2}^{\sigma,\sigma'}&=&-\frac{j\sigma}{8\gamma b}\sum_{\overline{\sigma}=\pm1}\overline{\sigma } d_{s+(1+\sigma')/2,s_1}^{m+\sigma(\overline{\sigma}+1)}\nonumber\\ & & \qquad\times f_{s-\sigma'(\overline{\sigma}-1)/2,s_1}^{\sigma m},\\ g_{2,3}^{\sigma,\sigma'}&=&\sum_{\overline{\sigma}=\pm1}\biggl(\frac{\overline{\sigma}d_{s+\sigma'(\overline{\sigma}+1)+(\sigma+1)/2,s_1}^m } { 4[\sigma'(2s+1)+2+\sigma\sigma'(1-\overline{\sigma})]}\nonumber\\ & & \qquad\times h_{s+\sigma(1-\overline{\sigma})/2,s_1}^{m,\sigma'}\biggr),\\ g_{2,4}^{\sigma,\sigma',\sigma''}&=&-\frac{1}{4}\sum_{\overline{\sigma}=\pm1}\biggl(\sigma\sigma'\overline{\sigma } d_{s+\sigma'(\overline{\sigma}+1)+(\sigma''+1)/2,s_1}^{m+\sigma(\overline{\sigma}+1)/2}\nonumber\\ & & \times\frac{r_{s+\sigma''(1-\overline{\sigma})/2,s_1}^ { m,\sigma,\sigma'}}{\sigma\gamma b / j+\sigma'(2s+1+\sigma''-\sigma''\overline{\sigma})+2}\biggr),\nonumber\\ & & \\ g_{2,5}^{\sigma,\sigma',\sigma''}&=&-\frac{1}{8}\sum_{\overline{\sigma}=\pm1}\biggl(\overline{\sigma } k_{s+\sigma''(1-\overline{\sigma})/2,s_1}^{\sigma m,\sigma ' } \nonumber\\ & & \times\frac{d_{s+(\overline{\sigma}+1)\sigma'+(\sigma''+1)/2,s_1}^{m+\sigma(\overline{\sigma}+1)}}{2\sigma\gamma b / j+\sigma'(2s+1+\sigma''-\sigma''\overline{\sigma})+2}\biggr),\nonumber\\ & & \\ g_{3,1}^{\sigma,\sigma'}&=&\frac{\sigma'j}{2\gamma b}\bigl((2m+\sigma')a_s^{\sigma(m+\sigma')}a_s^{\sigma'm}\nonumber\\ & & -(2m+2\sigma+\sigma')a_s^{\sigma'(m+\sigma)}a_s^{\sigma m}\bigr),\\ g_{3,2}^{\sigma,\sigma'}&=&-\frac{\sigma j}{8\gamma b}\bigl(a_s^{\sigma'(m+2\sigma)}f_s^{\sigma m}\nonumber\\ & & -a_s^{\sigma'm}f_s^{\sigma(m+\sigma')}\bigr),\\ g_{3,3}^{\sigma,\sigma'}&=&\sum_{\overline{\sigma}=\pm1}\frac{\overline{\sigma}a_{s+(\overline{\sigma}+1)\sigma'}^{\sigma m}h_{s , s_1}^{m-\sigma(\overline{\sigma}-1)/2,\sigma'}}{4[\sigma'(2s+1)+2]},\\ g_{3,4}^{\sigma,\sigma',\sigma''}&=&-\sum_{\overline{\sigma}=\pm1}\biggl(\frac{j\sigma\sigma'\overline{\sigma}a_{s+(\overline{\sigma}+1)\sigma'}^{\sigma''m + \sigma\sigma''(\overline{\sigma}+1)/2}}{4\bigl(\sigma\gamma b+j[\sigma'(2s+1)+2]\bigr)}\nonumber\\ & & \times r_{s , s_1}^{m-\sigma''(\overline{\sigma}-1)/2,\sigma,\sigma'}\biggr),\\ g_{3,5}^{\sigma,\sigma',\sigma''}&=&-\sum_{\overline{\sigma}=\pm1}\biggl(\frac{j\overline{\sigma}a_{s+(\overline{\sigma}+1)\sigma'}^{\sigma''m+ ( 1+\overline{\sigma})\sigma\sigma''}}{8\bigl(2\sigma\gamma b+j[\sigma'(2s+1)+2]\bigr)}\nonumber\\ & & \times k_{s , s_1}^{\sigma m+(1-\overline{\sigma})\sigma\sigma''/2,\sigma'}\biggr),\end{aligned}\ ] ] @xmath622\nonumber\\ & & \times c_{s+(\sigma'+1)/2,s_1}^{-\sigma\sigma'm-\sigma\sigma'\sigma''(\overline{\sigma}+1)/2-(\sigma'+1)/2}\nonumber\\ & & \times a_{s+\sigma'(1-\overline{\sigma})/2}^{\sigma''m+\sigma\sigma''(1-\overline{\sigma})/2},\\ g_{4,2}^{\sigma,\sigma',\sigma''}&=&-\frac{\sigma'\sigma''j}{8\gamma b}\sum_{\overline{\sigma}=\pm1}\overline{\sigma}f_{s+\sigma'(1-\overline{\sigma})/2}^{\sigma m+\sigma\sigma''(1-\overline{\sigma})/2}\nonumber\\ & & \times c_{s+(\sigma'+1)/2,s_1}^{-\sigma'\sigma''m-\sigma\sigma'\sigma''(\overline{\sigma}+1)-(\sigma'+1)/2},\nonumber\\ & & \\
g_{4,3}^{\sigma,\sigma',\sigma''}&=&\sum_{\overline{\sigma}=\pm1}\biggl(\frac{\sigma''\overline{\sigma } h_{s+\sigma''(1-\overline{\sigma})/2,s_1}^{m+\sigma(1-\overline{\sigma})/2,\sigma'}}{4[\sigma'(2s+1)+2+\sigma'\sigma''(1-\overline{\sigma})]}\nonumber\\ & & \times c_{s+(\sigma''+1)/2+\sigma'(\overline{\sigma}+1),s_1}^{-\sigma\sigma''m-(\sigma''+1)/2}\biggr),\\ g_{4,4}^{\sigma,\sigma',\sigma'',\sigma'''}&=&-\frac{1}{4}\sum_{\overline{\sigma}=\pm1}\biggl(\sigma\sigma'\sigma'''\overline{\sigma } r_{s+\sigma'''(1-\overline{\sigma})/2,s_1}^{m+\sigma''(1-\overline{\sigma})/2,\sigma,\sigma'}\nonumber\\ & & \times \frac{c_{s+(\sigma'''+1)/2+\sigma'(\overline{\sigma}+1),s_1}^{-\sigma''\sigma'''m-\sigma\sigma''\sigma'''(\overline{\sigma}+1)/2-(\sigma'''+1)/2}}{\sigma\gamma b / j+\sigma'(2s+1+\sigma'''-\sigma'''\overline{\sigma})+2}\biggr),\nonumber\\ & & \\
g_{4,5}^{\sigma,\sigma',\sigma'',\sigma'''}&=&-\frac{1}{8}\sum_{\overline{\sigma}=\pm1}\biggl(\sigma'''\overline{\sigma}j k_{s+\sigma'''(1-\overline{\sigma})/2,s_1}^{\sigma m+\sigma\sigma''(1-\overline{\sigma})/2,\sigma'}\nonumber\\ & & \times \frac{c_{s+(\sigma'''+1)/2+\sigma'(\overline{\sigma}+1),s_1}^{-\sigma''\sigma'''m-(\sigma'''+1)/2-\sigma\sigma''\sigma'''(\overline{\sigma}+1)}}{2\sigma\gamma b / j+\sigma'(2s+1+\sigma'''-\sigma'''\overline{\sigma})+2}\biggr),\label{g45}\nonumber\\\end{aligned}\ ] ] where we have used the identities @xmath623 , @xmath624 , @xmath625 , @xmath626 , and @xmath627 . for completeness
, we write the contributions to @xmath14 from the off - diagonal and antisymmetric diagonal matrix element terms that are first order in the anisotropy interactions .
these could be measured by collecting data at specific @xmath628 values .
we have @xmath629\biggr\},\end{aligned}\ ] ] @xmath630\nonumber\\ & & + \sin^2({\bm q}\cdot{\bm d})\biggl[\sum_{\sigma'=\pm1}\delta\bigl(\omega+e_{s , s_1}^m - e_{s+\sigma',s_1}^{m}\bigr)\nonumber\\ & & \times\bigl\{\bigl(h_1^{(1)}(\theta,\phi,\phi_{b , q})+\alpha_{s , s_1}h_2^{(1)}(\theta,\phi,\phi_{b , q})\bigr)\nonumber\\ & & \times g_{6,3}^{\sigma'}+h_2^{(1)}(\theta,\phi,\phi_{b , q})g_{6,4}^{\sigma'}\bigr\}\nonumber\\ & & -\sum_{\sigma,\sigma'=\pm1}\delta\bigl(\omega+e_{s , s_1}^m - e_{s+\sigma',s_1}^{m+\sigma}\bigr)\nonumber\\ & & \times\bigl\{\bigl(h_1^{(1)}(\theta,\phi,\phi_{b , q})+\alpha_{s , s_1}h_2^{(1)}(\theta,\phi,\phi_{b , q})\bigr)\nonumber\\ & & \times g_{6,5}^{\sigma,\sigma ' } + h_2^{(1)}(\theta,\phi,\phi_{b , q})g_{6,6}^{\sigma,\sigma'}\bigr\}\biggr]\biggr\},\nonumber\\\end{aligned}\ ] ] and @xmath631\biggr\}.\end{aligned}\ ] ] the angular functions and coefficients are given by @xmath632\nonumber\\ & & + 2\sin\phi_{b , q}j_d\sin\theta\sin(2\phi)\bigr),\\ h_2^{(1)}(\theta,\phi,\phi_{b , q})&=&\frac{1}{j}\bigl(\cos\phi_{b , q}\sin(2\theta)[j_a - j_e\cos(2\phi)]\nonumber\\ & & + 2\sin\phi_{b , q}j_e\sin\theta\sin(2\phi)\bigr),\end{aligned}\ ] ] @xmath633,\\ g_{6,2}^{\sigma}&=&\frac{j}{8\gamma b}\bigl(a_s^{\sigma m}\bigr)^2(2m+\sigma),\\ g_{6,3}^{\sigma'}&=&-\frac{\sigma'j}{8\gamma b}d_{s+(\sigma'+1)/2,s_1}^m\nonumber\\ & & \times\sum_{\sigma,\overline{\sigma}=\pm1}\bigl[(2m\overline{\sigma}+\sigma ) a_{s+\sigma'(\overline{\sigma}+1)/2}^{\sigma\overline{\sigma}m}\nonumber\\ & & \qquad\times c_{s+(\sigma'+1)/2,s_1}^{-\sigma\sigma'm-(1+\overline{\sigma}\sigma')/2}\bigr],\\ g_{6,4}^{\sigma'}&=&\frac{j}{16}d_{s+(\sigma'+1)/2,s_1}^m\sum_{\sigma,\overline{\sigma}=\pm1}\nonumber\\ & & \times\biggl(\frac{\overline{\sigma}c_{s+(1+\sigma'-2\overline{\sigma}\sigma')/2,s_1}^{\sigma\sigma'm-(1-\overline{\sigma}\sigma')/2}}{\sigma\gamma b+j[-\sigma'(2s+1)+\overline{\sigma}-1]}\nonumber\\ & & \times r_{s+\sigma'(\overline{\sigma}+1)/2,s_1}^{m,\overline{\sigma}\sigma,-\overline{\sigma}\sigma'}\biggr),\\ g_{6,5}^{\sigma,\sigma'}&=&-\frac{\sigma'j}{8\gamma b}c_{s+(\sigma'+1)/2,s_1}^{-\sigma\sigma'm-(\sigma'+1)/2}(2m+\sigma)\nonumber\\ & & \times\sum_{\overline{\sigma}=\pm1}\overline{\sigma}d_{s+(\sigma'+1)/2,s_1}^{m+\sigma(\overline{\sigma}+1)/2 } a_{s-\sigma'(\overline{\sigma}-1)/2}^{\sigma m},\nonumber\\ & & \\ g_{6,6}^{\sigma,\sigma'}&=&\frac{j}{16}c_{s+(\sigma'+1)/2,s_1}^{-\sigma\sigma'm-(\sigma'+1)/2}\sum_{\overline{\sigma}=\pm1}\nonumber\\ & & \times\biggl(\frac{\overline{\sigma}r_{s-\sigma'(\overline{\sigma}-1)/2,s_1}^{m+\sigma(1-\overline{\sigma})/2,\overline{\sigma}\sigma,\overline{\sigma}\sigma'}}{\sigma\gamma b+j[\sigma'(2s+1)+1+\overline{\sigma}]}\nonumber\\ & & \qquad\times d_{s+(1+\sigma'+2\overline{\sigma}\sigma')/2,s_1}^{m+\sigma(\overline{\sigma}+1)/2}\biggr),\\ g_{7,1}^{\sigma}&=&\frac{j}{4\gamma b}(2m+\sigma)\bigl(a_s^{\sigma m}\bigr)^2,\nonumber\\ & & \\ g_{7,2}^{\sigma,\sigma'}&=&-\frac{\sigma j}{8\gamma b}c_{s+(\sigma'+1)/2,s_1}^{-\sigma\sigma '
m-(\sigma'+1)/2}\nonumber\\ & & \times\sum_{\overline{\sigma}=\pm1}\biggl(\overline{\sigma}c_{s+(\sigma'+1)/2,s_1}^{\sigma\sigma ' m-(1-\sigma'-2\overline{\sigma}\sigma')/2}\nonumber\\ & & \times f_{s+\sigma'(1-\overline{\sigma})/2}^{\overline{\sigma}\sigma m-(1-\overline{\sigma})/2}\biggr),\\ g_{7,3}^{\sigma,\sigma'}&=&\frac{1}{8}c_{s+(\sigma'+1)/2,s_1}^{-\sigma\sigma ' m-(\sigma'+1)/2}\nonumber\\ & & \times\sum_{\overline{\sigma}=\pm1}\biggl(\overline{\sigma}k_{s+\sigma'(1-\overline{\sigma})/2,s_1}^{\overline{\sigma}\sigma m-(1-\overline{\sigma})/2,\overline{\sigma}\sigma'}\nonumber\\ & & \times\frac{c_{s+(1+\sigma'+2\overline{\sigma}\sigma')/2,s_1}^{-\sigma\sigma ' m-(1+\sigma'+2\overline{\sigma}\sigma')/2}}{2\sigma\gamma b / j+\sigma'(2s+1)+1+\overline{\sigma}}\biggr).\end{aligned}\ ] ] here we evaluate the epr response @xmath634 including the corrections to the induction representation wave functions first order in the anisotropy interactions .
it can be shown that second order perturbations to the wave functions only contribute to the epr response to third ( and higher ) order in the anisotropy interactions , and are hence neglected . from eq .
( [ chiexact ] ) , we expand the exact wave functions @xmath635 to first order in the anisotropy interactions . to this order ,
@xmath38 and @xmath40 are still good quantum numbers , so we rewrite @xmath636 .
we then find @xmath637\nonumber\\ & & \times\bigl(\delta(e_{s , s_1}^m - e_{s',s_1}^{m'}+\omega)\nonumber\\ & & \qquad-\delta(e_{s',s_1}^{m'}-e_{s , s_1}^m+\omega)\bigr)\nonumber\\ & & \times\biggl[\sum_{\overline{\sigma}=\pm1}\delta_{s',s}\delta_{m',m+\overline{\sigma}\sigma}\bigl|z_{0}^{\overline{\sigma}\sigma}\bigr|^2\nonumber\\ & & + \delta_{s',s}\sum_{\sigma'=\pm\sigma}\delta_{m',m+\sigma + \sigma'}\bigl|z_{0}^{\sigma+\sigma'}\bigr|^2\nonumber\\ & & + \delta_{m',m+\sigma}\sum_{\sigma''=\pm\sigma}\delta_{s',s+2\sigma''}\bigl|z_{2\sigma''}^{\sigma}\bigr|^2\nonumber\\ & & + \sum_{\sigma',\sigma''=\pm\sigma}\delta_{s',s+2\sigma''}\bigl(\delta_{m',m+\sigma+\sigma ' } \bigl|z_{2\sigma''}^{\sigma+\sigma'}\bigr|^2\nonumber\\ & & \qquad+\delta_{m',m+\sigma+2\sigma'}\bigl|z_{2\sigma''}^{\sigma+2\sigma'}\bigr|^2\bigr)\biggr],\end{aligned}\ ] ] where the amplitudes @xmath638 are given to first order in the anisotropy energies by @xmath639},\nonumber\\ & & \\
z_{2\sigma''}^{\sigma+\sigma'}&=&\frac{1}{\sigma'\gamma b+j[\sigma''(2s+1)+2]}\nonumber\\ & & \times\bigl(a_{s+2\sigma''}^{\sigma(m+\sigma')}{\cal x}_{s , s_1}^{m,\sigma',\sigma''}\nonumber\\ & & \qquad - a_s^{\sigma m}\bigl({\cal x}_{s+2\sigma'',s_1}^{m+\sigma+\sigma',-\sigma',-\sigma''}\bigr)^{*}\bigr),\nonumber\\ & & \\ z_{2\sigma''}^{\sigma+2\sigma'}&=&\frac{1}{2\sigma'\gamma b+j[\sigma''(2s+1)+2]}\nonumber\\ & & \times\bigl(a_{s+2\sigma''}^{\sigma(m+2\sigma')}{\cal y}_{s , s_1}^{m,\sigma',\sigma''}\nonumber\\ & & \qquad - a_s^{\sigma m}\bigl({\cal y}_{s+2\sigma'',s_1}^{m+\sigma+2\sigma',-\sigma',-\sigma''}\bigr)^{*}\bigr),\nonumber\\\end{aligned}\ ] ] leading to @xmath640 ^ 2}\nonumber\\ & & \times\bigl(\sum_{\overline{\sigma}=\pm1}\overline{\sigma}a_{s+\sigma''(1+\overline{\sigma})}^{\sigma m}h_{s , s_1}^{m+\sigma(1-\overline{\sigma})/2,\sigma''}\bigr)^2,\nonumber\\ & & \\
\bigl|z_{2\sigma''}^{\sigma+\sigma'}\bigr|^2&=&\frac{j^2f_3^{(2)}(\theta,\phi)}{16\bigl(\sigma'\gamma b+j[\sigma''(2s+1)+2]\bigr)^2}\nonumber\\ & & \times\bigl(\sum_{\overline{\sigma}=\pm1}\overline{\sigma}a_{s+\sigma''(1+\overline{\sigma})}^{\sigma m+\sigma\sigma'(1+\overline{\sigma})/2 } r_{s , s_1}^{m+\sigma(1-\overline{\sigma})/2,\sigma',\sigma''}\bigr)^2,\nonumber\\ & & \\
\bigl|z_{2\sigma''}^{\sigma+2\sigma'}\bigr|^2&=&\frac{j^2f_6^{(2)}(\theta,\phi)}{64\bigl(2\sigma'\gamma b+j[\sigma''(2s+1)+2]\bigr)^2}\nonumber\\ & & \times\bigl(\sum_{\overline{\sigma}=\pm1}\overline{\sigma}a_{s+\sigma''(1+\overline{\sigma})}^{\sigma m+\sigma\sigma'(1+\overline{\sigma } ) } k_{s , s_1}^{\sigma'm+\sigma\sigma'(1-\overline{\sigma})/2,\sigma''}\bigr)^2,\nonumber\\\end{aligned}\ ] ] where @xmath344 is given by eqs .
( [ e0 ] ) and ( [ esm1 ] ) , @xmath641 , @xmath570 , @xmath579 , @xmath580 , @xmath300 , and @xmath595 are given by eqs .
( [ asm ] ) , ( [ fsx ] ) , ( [ hss1msigma ] ) , ( [ kss1xsigma ] ) , ( [ alphass1 ] ) , and ( [ rss1xsigma ] ) , respectively , the @xmath642 are given in eqs .
( [ f1])-([f7 ] ) , and we have used the relations following eq .
( [ g45 ] ) .
we note that since @xmath363 for clockwise and counterclockwise oscillatory induction senses , setting @xmath643 is equivalent to setting them equal to @xmath644 . for the three cases with @xmath645 , there are no associated resonant magnetic inductions , and since the epr frequency @xmath646 can not be easily varied in an experiment , these terms are irrelevant . for the 10 remaining cases with @xmath647 , the resonant magnetic induction
is then found to have the form given by eq .
( [ bres ] ) . the term with amplitude @xmath648 has @xmath649 , @xmath650 , and @xmath651 .
the term with amplitude @xmath652 has @xmath653 , @xmath654 , and @xmath655 .
the two terms with amplitudes @xmath656 where @xmath657 have @xmath658 , @xmath659 $ ] , and @xmath660 .
the two terms with amplitudes @xmath661 , where @xmath657 , have @xmath662 , @xmath663 $ ] , and @xmath664 .
the four terms with amplitudes @xmath665 , where @xmath666 , have @xmath667 , @xmath668/(\sigma+2\sigma')]$ ] , and @xmath669 .
in this appendix , we evaluate the corrections to the eigenstate energies second order in the four anisotropy interaction energies @xmath88 for @xmath89 .
the operations of the rotated hamilatonian @xmath293 upon the eigenstates @xmath294 may be written as @xmath670 where @xmath671 and @xmath672 are given by eqs .
( [ e0 ] ) and ( [ esm1 ] ) , respectively , and the @xmath673 to @xmath674 functions are given by eqs .
( [ calu ] ) to ( [ caly ] ) , respectively . from eq .
( [ hprime ] ) , the second order eigenstate energies may be written as @xmath675}\nonumber\\ & & + \sum_{\sigma,\sigma'\pm1}\bigl(\frac{|{\cal x}_{s , s_1}^{m,\sigma,\sigma'}|^2}{j[2+(2s+1)\sigma']+\sigma\gamma b}\nonumber\\ & & + \frac{|{\cal y}_{s , s_1}^{m,\sigma,\sigma'}|^2}{j[2+(2s+1)\sigma']+2\sigma\gamma b}\bigr).\end{aligned}\ ] ] @xmath676\nonumber\\ & & \times\bigl(\cos^2\theta[\tilde{j}_{b , a}^{s , s_1}-\cos(2\phi)\tilde{j}_{d , e}^{s , s_1}]^2\nonumber\\ & & \qquad+\sin^2(2\phi)(\tilde{j}_{d , e}^{s , s_1})^2\bigr),\label{esm2}\\ e_{s , s_1}^{m,(2){\cal v}}&=&-\frac{m}{8\gamma b}[2s(s+1)-2m^2 - 1]\nonumber\\ & & \times\bigl([\sin^2\theta\tilde{j}_{b , a}^{s , s_1}+(1+\cos^2\theta)\cos(2\phi)\tilde{j}_{d , e}^{s , s_1}]^2\nonumber\\ & & \qquad+4\cos^2\theta\sin^2(2\phi)(\tilde{j}_{d , e}^{s , s_1})^2\bigr),\\ e_{s , s_1}^{m,(2){\cal w}}&=&\frac{d_{s , s_1}^m}{16j}\bigl(j_a-3[j_a\cos^2\theta+j_e\sin^2\theta\cos(2\phi)]\bigr)^2,\nonumber\\ & & \label{esm2w}\\ e_{s , s_1}^{m,(2){\cal x}}&=&\frac{f_{s , s_1}^m(\gamma b / j)\sin^2\theta}{2j}\nonumber\\ & & \times\bigl([j_a - j_e\cos(2\phi)]^2\cos^2\theta\nonumber\\ & & \qquad+\sin^2(2\phi)j_e^2\bigr),\label{esm2x}\\ e_{s , s_1}^{m,(2){\cal y}}&=&\frac{g_{s , s_1}^m(\gamma b / j)}{64j}\bigl[\bigl(j_a\sin^2\theta\nonumber\\ & & \qquad+j_e(1+\cos^2\theta)\cos(2\phi)\bigr)^2\nonumber\\ & & \qquad+4\cos^2\theta\sin^2(2\phi)j_e^2\bigr],\label{esm2y}\end{aligned}\ ] ] where @xmath677\eta_{s , s_1}^2\eta_{s-1,s_1}^2}{(2s-1)}\nonumber\\ & & + \eta_{s+2,s_1}^2\eta_{s+1,s_1}^2\nonumber\\ & & \times \frac{[(s+1)^2-m^2][(s+2)^2-m^2]}{(2s+3)},\nonumber\\ f_{s , s_1}^m(x)&=&-\frac{\eta_{s , s_1}^2\eta_{s-1,s_1}^2(s^2-m^2)}{(2s-1)^2-x^2}\nonumber\\ & & \times\bigl((2s-1)[(s-1)(s-2)+m^2]\nonumber\\ & & \qquad - m(2s-3)x\bigr)\nonumber\\ & & + \frac{\eta_{s+2,s_1}^2\eta_{s+1,s_1}^2[(s+1)^2-m^2]}{(2s+3)^2-x^2}\nonumber\\ & & \times\bigl((2s+3)[(s+2)(s+3)+m^2]\nonumber\\ & & \qquad - m(2s+5)x\bigr),\\ g_{s , s_1}^m(x)&=&-\frac{2\eta_{s , s_1}^2\eta_{s-1,s_1}^2}{(2s-1)^2 - 4x^2}\bigl((2s-1)[m^4\nonumber\\ & & + m^2(6s^2 - 18s+11)\nonumber\\ & & + s(s-1)(s-2)(s-3)]\nonumber\\ & & -4mx(2s-3)(s^2 - 3s+1+m^2)\bigr)\\ & & + \frac{2\eta_{s+2,s_1}^2\eta_{s+1,s_1}^2}{(2s+3)^2 - 4x^2}\bigl((2s+3)[m^4\nonumber\\ & & + m^2(6s^2 + 30s+35)\nonumber\\ & & + ( s+1)(s+2)(s+3)(s+4)]\nonumber\\ & & -4mx(2s+5)(s^2 + 5s+5+m^2)\bigr).\end{aligned}\ ] ] there is a remarkable amount of symmetry in the angular dependence of the eigenstate energies .
we note that @xmath678 and @xmath679 have the same forms , differing in the replacements of the interactions @xmath384 and @xmath385 with @xmath225 and @xmath226 , respectively , and with different overall constant functions .
the same comparison can also be made with @xmath680 and @xmath681 .
in addition , we note that there is a remarkable similarity in the @xmath301 dependence of @xmath682 with that of the local spin anisotropy part of @xmath683 given by eq .
( [ bstep ] ) .
the contributions to the @xmath38th level crossing second order in the anisotropy interactions are calculated as indicated in eq .
( [ bstep2 ] ) , and are found to be @xmath684 where the @xmath685 are given by eqs .
( [ f1])-([f7 ] ) , respectively .
the coefficients are given by @xmath686}{s(2s+1)(2s+3)},\\ a_{2,0}(s)&=&3(9+s-23s^2 + 4s^3 + 12s^4),\nonumber\\ a_{2,1}(s)&=&9 - 8s-4s^2,\\ a_3(s , s_1)&=&a_3^{\cal u}(s , s_1)+a_3^{\cal x}(s , s_1),\\ a_3^{\cal u}(s , s_1)&=&\frac{[3 - 3s-3s^2 + 4s_1(s_1 + 1)]^2}{2(2s+3)^2}\nonumber\\ & & -\frac{(s-1)[3 + 3s-3s^2 + 4s_1(s_1 + 1)]^2}{2s(2s+1)^2},\nonumber\\ & & \\ a_3^{\cal x}(s , s_1)&=&\frac{[s(s+2)-4s_1(s_1 + 1)]}{2(s+1)(s+3)(2s+1)^2(2s+3)^2}\nonumber\\ & & \times\frac{a_{3,0}^{\cal x}(s)+4s_1(s_1 + 1)a_{3,1}^{\cal x}(s)}{(2s+5)(3s+1)},\nonumber\\
a_{3,0}^{\cal x}(s)&=&(s+1)(s+3)(51 + 114s\nonumber\\ & & + 209s^2 + 302s^3 + 164s^4 + 24s^5),\\ a_{3,1}^{\cal x}(s)&=&129 + 318s+395s^2\nonumber\\ & & + 442s^3 + 300s^4 + 72s^5,\nonumber\\ a_{4}(s , s_1)&=&\frac{(4s-3)}{8s},\\ a_5(s , s_1)&=&-\frac{3[a_{5,0}(s)+4s_1(s_1 + 1)]}{4s(2s+1)(2s+3)},\\ a_{5,0}(s)&=&3 + 3s-5s^2 - 4s^3,\\ a_6(s , s_1)&=&a_6^{\cal v}(s , s_1)+a_6^{\cal y}(s , s_1),\\ a_6^{\cal v}(s , s_1)&=&\frac{[3 - 3s-3s^2 + 4s_1(s_1 + 1)]^2}{8(2s-1)(2s+3)^2}\nonumber\\ & & -\frac{(s-1)[3 + 3s-3s^2 + 4s_1(s_1 + 1)]^2}{8s(2s-3)(2s+1)^2},\nonumber\\ & & \\ a_6^{\cal y}(s , s_1)&=&\frac{1}{4(2s-3)(2s-1)(2s+1)^2(2s+3)^2}\nonumber\\ & & \times\frac{\sum_{n=0}^2a_{6,n}^{\cal y}(s)[4s_1(s_1 + 1)]^n}{(2s+5)(4s+1)(4s+3)},\nonumber\\ a_{6,0}^{\cal y}(s)&=&s\bigl(-216 + 837s+5052s^2 + 3521s^3\nonumber\\ & & -12414s^4 - 21876s^5 - 7464 s^6\nonumber\\ & & + 6720s^7 + 5376s^8 + 1024s^9\bigr),\\ a_{6,1}^{\cal y}(s)&=&3\bigl(81 + 102s+10s^2\nonumber\\ & & + 196s^3 + 296s^4 + 80s^5\bigr),\\ a_{6,2}^{\cal y}(s)&=&189 + 258s-100s^2 + 152s^3\nonumber\\ & & + 640s^4 + 256s^5,\\ a_7(s , s_1)&=&\frac{[s(s+2)-4s_1(s_1 + 1)]}{4(2s+1)^3(2s+3)^3(2s+5)}\nonumber\\ & & \times\bigl(4s_1(s_1 + 1)(-1 + 38s+60s^2 + 24s^3)\nonumber\\ & & + ( s+1)(3 + 67s+94s^2 + 44s^3 + 8s^4)\bigr).\nonumber\\\end{aligned}\ ] ] by expanding the solutions in the crystal representation to second order in the @xmath88 , we have explicitly checked these expressions for @xmath687 , and for @xmath10 , @xmath440 .
we note that for @xmath4 , only @xmath688 and @xmath689 are non - vanishing .
here we present the expression for @xmath154 to @xmath690 appearing in eqs .
( [ epsilonsb ] ) and ( [ peak])-([slope ] ) in the text .
we have @xmath691 where the @xmath356 are given by eqs .
( [ f1])-([f6 ] ) , and @xmath692\nonumber\\ & & \times[4s_1(s_1 + 1)-(s+1)(s+3)]\bigr),\\ b_4(s , s_1)&=&\frac{(2s-1)}{8s}\theta(s-1),\\ b_5(s , s_1)&=&\frac{(2s-1)}{4s}\alpha_{s , s_1}\theta(s-1),\\ b_6(s , s_1)&=&\frac{(2s-1)}{8s}\alpha^2_{s , s_1}\theta(s-1)+b_6^{\cal y}(s , s_1),\\ b_6^{\cal y}(s , s_1)&=&\frac{s(s-1)}{8(4s^2 - 1)}[(s^2 - 1 - 4s_1(s_1 + 1)]\nonumber\\ & & \times[s(s-2)-4s_1(s_1 + 1)]\nonumber\\ & & -\frac{sb_{6,1}(s)[(s(s+2)-4s_1(s_1 + 1)]}{72(2s+1)(2s+3)^2(2s+5)(4s+3)^2}\nonumber\\ & & \times[(s+1)(s+3)-4s_1(s_1 + 1)],\\ b_{6,1}(s)&=&369 + 1011s+1420s^2 + 1076s^3\nonumber\\ & & + 416s^4 + 64s^5 , \end{aligned}\ ] ] where @xmath693 for @xmath694 , @xmath695 for @xmath696 is the heaviside step function . | we present a model of equal spin @xmath0 dimer single molecule magnets .
the spins within each dimer interact via the heisenberg and the most general set of four quadratic anisotropic spin interactions with respective strengths @xmath1 and @xmath2 , and with the magnetic induction @xmath3 .
we solve the model exactly for @xmath4 , 1 , and 5/2 , and for antiferromagnetic heisenberg couplings ( @xmath5 ) , present @xmath6 curves at low @xmath7 for these cases .
low-@xmath7 @xmath8 curves for @xmath4 and electron paramagnetic susceptibility @xmath9 for @xmath10 are also provided . for weakly anisotropic dimers , the hartree approximation yields rather simple analytic formulas for @xmath11 and @xmath12 at arbitrary @xmath0 that accurately fit the exact solutions at sufficiently low @xmath7 or large @xmath13 .
low-@xmath7 , large-@xmath13 formulas for the inelastic neutron scattering cross - section @xmath14 and @xmath9 with arbitrary @xmath0 and @xmath15 in an extended hartree approximation are also given .
we show that the various anisotropy interactions give rise to new , detectable electron paramagnetic resonances and to new peaks in @xmath16 , which can provide precise determinations of the particular set of microscopic anisotropy interaction strengths . for antiferromagnetic heisenberg couplings ( @xmath5 ) and weak anisotropy interactions ( @xmath17 ) , we provide analytic formulas for the @xmath18 level - crossing magnetic inductions @xmath19 , at which the low-@xmath7 magnetization @xmath11 exhibits steps and the low-@xmath7 specific heat @xmath12 exhibits zeroes , surrounded by double peaks of uniform height .
strong anisotropy interactions drastically alter these behaviors , however .
our results are discussed with regard to existing experiments on @xmath20 fe@xmath21 dimers , suggesting the presence of single - ion anisotropy in one one them , but apparently without any sizeable global spin anisotropy .
further experiments on single crystals of these and some @xmath22 [ mn@xmath23@xmath21 dimers are therefore warranted , and we particularly urge electron paramagnetic resonance and inelastic neutron scattering experiments to be performed . |
we study the finite volume discretization of general linear hyperbolic systems with stiff source terms depending of a relaxation parameter @xmath2 , which admit an asymptotic diffusion limit .
this type of system occurs in many physical applications ( transport of particles , damped waves , electromagnetism , linearized gas dynamic , plasma physics ) or in biology , and poses some numerical difficulties .
the classical godunov - type discretizations ( upwind , rusanov or hll schemes ) are not efficient because the time and spatial steps are constrained by the relaxation parameter @xmath2 @xcite , @xcite , @xcite . to treat this problem s. jin , c. d. levemore @xcite-@xcite using the ideas of a. y. leroux @xcite , introduced the notion of asymptotic preserving schemes ( ap schemes ) which eliminate these constraints . to illustrate the advantage of asymptotic preserving discretizations
, we propose a simple numerical example .
we solve the hyperbolic heat equation @xmath3 with two schemes : the upwind scheme and the asymptotic preserving scheme @xcite .
this model is approached when @xmath2 is small by the following diffusion equation @xmath4 the initial data is given by @xmath5 with @xmath6 a gaussian function and @xmath7 .
the parameters are given by @xmath8 and @xmath9 .
the time discretization is explicit and the time step is the half of the stability limit time step .
the convergence errors are computed using the exact diffusion solution .
.table with numerical error and cpu time associated to the upwind and gosse - toscani schemes . [ cols="^,^,^,^",options="header " , ] the tables ( [ tabp3111])-([tabp32])-([tabp34])-([tabp35])-([tabp36 ] ) give the order of convergence for some meshes and some values of @xmath2 .
the remarks introduced on the convergence results of the asymptotic preserving scheme for the @xmath10 model are valid for this test case .
now we solve the @xmath11 and @xmath12 systems with @xmath13 and @xmath14 for @xmath15 @xcite .
the `` diffusive '' part is approximated with the jl-(b ) nodal scheme .
the time discretization is implicit .
this test case is described in @xcite .
the final time is @xmath16 . of the solution fundamental for the @xmath12 model at the time @xmath17 , right @xmath18 for the fundamental solution for the @xmath11 model at the time @xmath17,title="fig : " ] of the solution fundamental for the @xmath12 model at the time @xmath17 , right @xmath18 for the fundamental solution for the @xmath11 model at the time @xmath17,title="fig : " ] the exact solution is composed of dirac functions with the velocities @xmath19 ( @xmath19 are the eigenvalues of @xmath20 ) and smooth functions between the dirac functions . at
the beginning the smooth functions are non negatives and becomes negatives for large time . for the @xmath12 system ,
the speed wave is @xmath21 and for the @xmath11 system the maximal velocity is approximately @xmath22 .
the numerical results reproduce this behavior , see figure [ pnt2 ] .
this test case is an example of a complicated geometry .
we consider a checker - board with different scattering and absorbing opacities on a lattice core ( see @xcite ) .
it is interesting for neutron transport simulations since is a simplification to a reactor core .
the geometry is given in the figure [ geometry ] .
\times\left[0,7\right]$ ] ] we define @xmath23 the scattering opacity and @xmath24 the absorption opacity . in the black square and
the striped squares @xmath25 and @xmath26 . in the white squares @xmath8 and @xmath27 . the relaxation parameter is defined @xmath28 in the whole domain .
all the unknowns are equal to zero at the time @xmath29 .
we solve the @xmath12 and @xmath11 systems with the additional source term @xmath30 where @xmath31 , @xmath32 and @xmath33 are the matrices associated to the @xmath12 or @xmath11 system and @xmath34 with @xmath35 a kronecker product .
the source @xmath36 in the black square and @xmath37 in the rest of the domain .
+ the @xmath11 systems is solved using the jl-(b ) scheme for the `` diffusive '' part .
we plot the first moment with a logarithmic scale @xmath38 at the final time @xmath39 . model and plot the @xmath40 of the first moment @xmath18 . in the right
, we solve the @xmath11 model and plot the @xmath40 of @xmath18.,title="fig : " ] model and plot the @xmath40 of the first moment @xmath18 . in the right
, we solve the @xmath11 model and plot the @xmath40 of @xmath18.,title="fig : " ] the results for @xmath12 and @xmath11 are given for the first moment in figures [ fond11 ] .
they are the same as those in @xcite-@xcite .
we have studied the discretization on distorted meshes of linear hyperbolic systems with stiff source .
we have proposed a method called `` diffusive - non diffusive '' decomposition which consists to split the hyperbolic system between the hyperbolic heat equation and a other system which is negligible in the diffusion regime . using an asymptotic preserving scheme for the hyperbolic heat equation to discretize the `` diffusive '' part and a classical scheme to discretize the other part
, we obtain an asymptotic preserving method for the complete system . for the approximation of transport equation
, we use this decomposition for the simplified models as @xmath0 or @xmath1 approximations . for the @xmath0 systems the decomposition is natural . since the first and second moments gives the limit regime .
the others moments are close to @xmath2 .
the high order moments are added only to obtain a better approximation in the pure transport regime ( @xmath26 ) . for the @xmath1 models , we remark that the diagonalized model admits a structure very close to the structure of the @xmath0 models .
the `` diffusive - non diffusive '' decomposition gives consistent schemes for all the regimes . if the numerical methods used to discretize the different parts of the decomposition are stable in norm @xmath41 , the method is stable in norm @xmath41 . modifying the schemes for the `` non diffusive '' part we can obtain a semi - implicit scheme with a cfl condition independent of @xmath2 .
however this method is not optimal for the discretization of @xmath1 models , since our numerical method does not preserve the positivity . in the future , it would be interesting to design positive and asymptotic preserving method . c. berthon , j. dubois , b. dubroca , t .- h .
nguyen - bui , r. turpault , _ a free streaming contact preserving scheme for the @xmath43 model _ , advances in applied mathematics and mechanics adv .
mech . , 3 ( 2010 ) , pp .
259 - 285 . c. buet ,
b. desprs , e. franck , _ an asymptotic preserving scheme with the maximum principle for the @xmath43 model on distorded meshes _ ,
sci . , paris , sr .
, vol 350 , n 11 - 12 p. 633
- 638 , juin 2012 c. buet , b. desprs , e. franck , _ asymptotic preserving finite volumes discretization for non - linear moment model on unstructured meshes _ , finite volumes for complex applications vi problems , perspectives , springer proceedings in mathematics volume 4 , 2011 , pp 467 - 474 c. chalons , f. coquel , e. godlewski , p. a. raviart , n. seguin _ godunov - type schemes for hyperbolic systems with parameter - dependent source : the case of euler system with friction _ , math .
models methods appl .
20 , 2109 ( 2010 ) y. coudire , j. p. vila , and p. villedieu .
_ convergence dun schma volumes finis explicite en temps pour les systmes hyperboliques linaires symtriques en domaines borns_. c.r .
paris , sr .
i , math . , 331:95 - 100 , 2000 .
n. crouseilles , m. lemou _ an asymptotic preserving scheme based on a micro - macro decomposition for collisional vlasov equations : diffusion and high - field scaling limits _ , kinetic and related models 4 , 2 ( 2011 ) 441 - 477 e. franck , p. hoch , g. samba , p. navarro _ an asymptotic preserving scheme for @xmath12 model using classical diffusion schemes on unstructured polygonal meshes _ esaim : proceedings , october 2011 , vol .
31 , p. 56 - 75 .
l. gosse _ transient radiative transfer in the grey case : well - balanced and asymptotic - preserving schemes built on cases s elementary solutions _ journal of quantitative spectroscopy and radiative transfer 112 ( 2011 ) 1995 - 2012 .
s. jin , _ asymptotic preserving ( ap ) schemes for multiscale kinetic and hyperbolic equations : a review _ , lecture notes for summer school on methods and models of kinetic theory ( m and mkt ) , porto ercole ( grosseto , italy ) , juin 2010 . | in this paper we propose an asymptotic preserving scheme for a family of friedrichs systems on unstructured meshes based on a decomposition between the hyperbolic heat equation and a linear hyperbolic which not involved in the diffusive regime .
for the hyperbolic heat equation we use asymptotic preserving schemes recently designed in @xcite-@xcite . to discretize the second part we use classical rusanov or upwind schemes . to finish
we apply this method for the discretization of the @xmath0 and @xmath1 models which are widely used in transport codes . |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.