VERSION
FEBRUARY2,2008
APreprinttypesetusingLTEXstyleemulateapjv.11/12/01
THEEVOLUTIONOFTHEHIGHENERGYTAILINTHEQUIESCENTSPECTRUMOFTHESOFT
X–RAYTRANSIENTAQLX-1
S.CAMPANA1,L.STELLA2
DraftversionFebruary2,2008
arXiv:astro-ph/0307218v1 10 Jul 2003ABSTRACT
AmoderatelevelofvariabilityhasbeendetectedinthequiescentluminosityofseveralneutronstarsoftX–raytransients.SpectralvariabilitywasfirstrevealedbyChandraobservationsofAqlX-1inthefourmonthsthatfollowedthe2000X–rayoutburst.ByadoptingthecanonicalmodelforquiescentspectrumofsoftX–raytransients,i.e.anabsorbedneutronstaratmospheremodelplusapowerlawtail,Rutledgeetal.(2002a)concludedthattheobservedspectralvariationscanbeascribedtotemperaturevariationsoftheneutronstaratmosphere.Theseresultscanhardlybereconciledwiththeneutronstarcoolingthatisexpectedtotakeplaceinbetweenoutbursts(afterdeepcrustalheatingintheaccretionphase).HerewereanalysetheChandraspectraofAqlX-1,togetherwithalongBeppoSAXobservationinthesameperiod,andproposeadifferentinterpretationofthespectralvariability:thatthisisduetocorrelatedvariationsofthepowerlawcomponentandthecolumndensity(>5,apartofwhichmightbeintrinsictothesource),whilethetemperatureandfluxoftheneutronstaratmosphericcomponentremainedunchanged.Thislendssupporttotheideathatthepowerlawcomponentarisesfromemissionattheshockbetweenaradiopulsarwindandinflowingmatterfromthecompanionstar.Subjectheadings:accretion,accretiondisks—binaries:close—star:individual(AqlX-1)—stars:neutron
1.INTRODUCTION
ThelargeluminosityswingoftransientX–raybinariesal-lowsthesamplingavarietyofphysicalconditionsthatarein-accessibletoaccretingcompactobjectsinpersistentsources.TheverylowluminositythatcharacterisesthequiescentstateofneutronstarsoftX–raytransients,SXRTs,(LX∼1032−1033ergs−1)opensupthepossibilityofstudyingtheseold,fastspinningneutronstarsindifferentandyetunexploredregimessuchasaccretionontotheneutronstarmagnetosphere(propeller),resumedmillisecondradiopulsaractivity,and/orlow-levelatmosphericemissionfromthecoolingoftheneu-tronstarinbetweentheaccretionintervalsoftheoutbursts(e.g.Campanaetal.1998a;Brownetal.1998;Rutledgeetal.2002b).
InrecentyearsthequiescentpropertiesofahandfulSXRTshavebeenstudiedinsomedetail.ThemainoutcomeoftheseinvestigationsisthatthequiescentX–rayspectraofSXRTsdis-playasoftcomponentplusahard(power-law)componentcon-tributingacomparablefluxinthe0.5–10keVband(Campana2001;Bildsten&Rutledge2000;Wijnands2001).Thesoftcomponenthasbeenfrequentlymodelledwithablackbodymodelof0.1–0.3keVtemperatureandfewkmradius.Espe-ciallypromisingistheideathatthesoftcomponentofSXRTsmaybeproducedfromthecoolingoftheneutronstarheatedduringtherepeatedoutbursts(vanParadijsetal.1987;Stellaetal.1994;Campanaetal1998a).Thetheoryofdeepcrustalheatingbypycnonuclearreactionscompareswellwiththeob-servations(Brownetal.1998;Campanaetal.1998a;Rut-ledgeetal.1999;Colpietal.2001).Inparticular,Rutledgeetal.(1999)fittedneutronstaratmosphericmodelstothesoftcomponentofquiescentspectraofSXRTsandderivedslightlysmallertemperatures(0.1–0.3keV)andlargerradii(10–15km,consistentwiththeneutronstarradius)thanthoseinferredfromsimpleblackbodyfits.
1
Observationally,thehardcomponentiswelldescribedbyapowerlawtail.InthequiescentspectrumofAqlX-1andCenX-4observedbyASCAandBeppoSAXthiscomponentissta-tisticallysignificant(Asaietal.1996,1998;Campanaetal.1998b,2000)withphotonindexinthe1–2range.Thesamepowerlawisneeded(evenifnotstatisticallysignificant)intheanalysisofChandradatainordertoachieveanemittingradiusofthecoolingcomponentconsistentwiththeneutronstarra-dius(otherwisetheinferredradiuswouldbesmaller;Rutledgeetal.2001a,2001b).Thenatureofthishardcomponentisstilluncertain.ModelsrangefromComptonizationtoAdvec-tion/ConvectionDominateAccretionFlow(ADAF/CDAF)toshockemissionfromtheneutronstarthatresumeditsradiopul-saractivityinquiescence.Thelattermodelenvisagesasitua-tionsimilartothatoftheeclipsingradiopulsarPSRB1259–63orofthe‘blackwidow’pulsarPSRB1957+20:ashockattheboundarybetweentherelativisticMHDwindfromtheradiopulsarandthematteroutflowingfromthecompanionstar(Ta-vani&Arons1997;Tavani&Brookshaw1991;Campanaetal.1998a).ForthemodeltoexplaintheobservedluminosityinthehardpowerlawcomponentofquiescentSXRTs,somefewpercentofthepulsarspin-downluminositymustbecon-vertedintoshockemission3.Theshockemissionmodelpre-dictssynchrotronemissionwithpowerlawphotonindexesinthe1.5–2rangeandextendingoverawiderangeoffrequen-cies.IndirectindicationsforthepresenceofthisemissionalsofromUVobservationsofCenX-4withHSTrevealingaflatspectrum(i.e.Γ∼2)whichmatcheswelltheextrapolatedX–raypowerlawcomponent(McClintock&Remillard2000).Powerlawindexesoutsidetheaboverangeareindicationsofstrong(inverse-Compton)cooling.
Inawaysimilartowhatisroutinelydoneintheoptical,apromisingtooltoprobetheX–rayemittingregionsisthrougharougheclipsemappingtechnique(e.g.Horne1985).Chandra
INAF-OsservatorioAstronomicodiBrera,ViaBianchi46,I–23807Merate(Lc),Italy
2INAF-OsservatorioAstronomicodiRoma,ViaFrascati33,I–00040MonteporzioCatone(Roma),Italy
3OngoingdeepsearchesintheradiobandhavenotyetrevealedanysteadyorpulsedemissionfromquiescentSXRTs(Burgayetal.2003);howeverfree-freeabsorptionduetomatterinthebinarysystemmightbeanimportantlimitingfactorinthesesearches(Stellaetal.1994;Burgayetal.2003).
1
2
observationsoftheeclipsingSXRT4U2129+47werethefirsttoexploitthepotentialofthistechniquebylookingattheex-tensionoftheemittingregionsthrougheclipses(Nowak,Heinz&Begelman2002).Duringeclipsesthesoftcomponentgetstotallyeclipsedwhereasthehardcomponentistoofainttoberevealed.Thisimpliesanupperlimitontheemissionsizeof∼<
10%theorbitalseparation(Nowaketal.2002).Theinclina-tionofAqlX-1hasbeenestimated,fromellipsoidalvariationsintheRandIbandlightcurves,tobegreaterthan36degrees(Welsh,Robinson&Young2000).
InthispaperweinvestigateinmoredetailthequiescentspectrumofoneofthebeststudiedSXRTsources:AqlX-1.WetakeadvantageoffourChandraexposures(Rutledgeetal.2002a)andone76kslong(unpublished)BeppoSAXexposure.AllthesedatawerecollectedjustaftertheNovem-ber2000outburst.BasedontheChandradataRutledgeetal.(2002a)claimedthatthesoftcomponentdecreasedby∼50%overthreemonths,thenincreasedby∼35%inonemonth,andthenremainedconstant(<6%change)overthelastmonth.Thevariabilityoftheseobservationswasascribedtoanintrin-sicvariabilityofthesoftcomponenthintingtoaccretionontotheneutronstarsurface.HerewediscussinmoredetailtheseobservationstogetherwiththeBeppoSAXlongexposure,prob-ingtheshockemissionmodel.
InSection2wedealwiththedata.InSection3wedescribethespectralfittingandrelatedresults.Discussionandconclu-sionsarereportedinSection4.
2.DATA
2.1.Chandra
AqlX-1wasobservedbyChandraaftertheNovember2000outburstonfouroccasions(seeTable1).Observationswerecarriedoutwiththebackside-illuminatedACIS-Sdetector(S3)atanoff-axispositionof4′andlimitedread-outareaof1/8(achievingatimeresolutionof0.44s)inordertolimitprob-lemsconnectedtopile-up(seeRutledgeetal.2002a).Asex-presslyrequired,AqlX-1fellonthesamephysicalpixelsinor-dertoavoidproblemswiththeCCDsquantumefficiency(seeRutledgeetal.2002a).FortheanalysisweuseCIAO2.2.1withCALDB2.15(thesearelaterversionsthanwhatusedbyRutledgeetal.2002a).Inallobservations,′′weextractedthesourcecountsfromanelliptical4.5′′×3regioncenteredonsourcewithapositionanglematchingthesource.Backgroundphotonswereextracted′′fromanannularregionwithinnerandouterradiiof10and20′′,respectively.Datawereextracted,usingpsextract,intopulse-invariant(PI)spectra.Wegroupedallthespectratohave(atleast)30photonsperchannel.Wealsocorrectedalltheancillary(arf)fileswiththerecentlyre-leasedcorrarftooltoaccountforthecontinuousdegradationintheACISCCD’squantumefficiency.
2.2.BeppoSAX
WeanalyseddatafromthetwoimaginginstrumentsonboardtheBeppoSAXsatellite:theLowEnergyConcentratorSpec-trometer(LECS;0.1–10keV,Parmaretal.1997)andtheMediumEnergyConcentratorSpectrometer(MECS;1.6–10.5keV,Boellaetal.1997).Non-imaginginstrumentsprovidedonlyupperlimits.OnlytwoofthethreeMECSunitswereop-eratingatthetimeoftheobservations.LECSdatawerecol-lectedonlyduringsatellitenight-timeresultinginshorterexpo-suretimes.ForasummaryoftheobservationsseeTable1.Theobservationtookplaceon2001April14,observingAqlX-1for
anetexposuretimeof76kswiththeMECSand30ksfortheLECS.
ProductswereextractedusingtheFTOOLSpackage(v.5.1).LECSandMECSeventswereextractedfromacircleof4′ra-dius.Thebackgroundwassubtractedusingspectrafromblankskyfilesatthesamedetectorcoordinates(aftercheckingthatthebackgroundoftheobservationwascomparable).Were-binnedtheLECSandMECSspectrainordertohave80countsperspectralbineach.
3.OVERALLSPECTRALANALYSIS
3.1.Spectralmodel
Rutledgeetal.(2002a)analysedthefourChandraspectratogether.Spectraaredifferent,e.g.ahardpower-lawtailhasbeendetectedonlyduringtwoofthefourobservations.Toac-countforthesedifferencesRutledgeetal.(2002a)consideredamodelmadebyanabsorbedatmospheremodelplusapowerlawandletvarysingleparameters(fixedalltheother)tryingtoaccountforthevariations.Theyfoundthatdifferencesinthespectracannotbeexplainedasentirelyduetoeitherachangingpowerlawfluxand/orindexortoavariablecolumndensity.Onthecontrary,differencescanbeacceptably(intermsofχ2statistics)explainedasentirelyduetoa(nonmonotonic)tem-peraturevariabilityofthethermalemissioncomponent.Thesevariationscannotbeexplainedwithinthedeepcrustalheatingmodelandtheauthorssuggestedthatthesemightoriginatefromquiescentaccretionontotheneutronstarsurface.
Asdiscussedintheintroduction,wewanttotestherethehy-pothesisthatquiescentemissionofSXRTs(andAqlX-1inpar-ticular)areproducedbyasoftthermalcomponent,likelyaris-ingfromcoolingoftheneutronstar,plusapowerlawhardtail,arisingfromshockemissionduetoanactivemillisecondpulsar.Intheory,thesoftcomponentissteadyonrelativelyshorttimewhereasthehardcomponentcanlikelyvarydependingonthegeometryanddensitytheoutflowingmatter.Hydrodynamicalsimulations(Brookshaw&Tavani1993)aswellasradioobser-vationsofmillisecondradiopulsars(MSPs)inbinarysystemswithasizeablemasstransfershowcomplexgeometriesand,moreimportantly,variationsfromoneorbitalcycletoanother.TheexampleoftherecentlydiscoveredMSPPSRJ1740–5340(D’Amicoetal.2001;Ferrarioetal.2001)isenlighting.Thisisa3.7msMSPorbitingevery32.5hramainsequencecom-panion.ThesourceislocatedintheglobularclusterNGC6397at2.5kpc.Theradiopulsargetspartiallyandtotallyeclipsedoverawiderangeoforbitalphases.ItemitsX–raysasob-servedbyChandra(Grindlayetal.2001)likelyarisingfromshockemission.Astestifiedbythissource,theMSPiseclipsedforlargepartoftheorbitandvariationsfromorbittoorbitareseen.
ThiscasemotivatesustoconsideraspectralmodelforfittingthequiescentX–rayspectraofAqlX-1madebyasoftthermalcomponentfromtheentireneutronstar,avariablepowerlawcomponent(thestrengthofwhichdependsontheinteractionwiththesurroundingmatter)andavariablecolumndensityduetovariationsintrinsictothesource(overafixedinterstellarcol-umndensity).
3.2.Spectralanalysis
SpectralanalysiswascarriedoutwiththeXSPECsoftware(v.11.2.0).Thespectralmodelweadoptedconsistofa(fixed)
3
coolingspectrum(weuseherethehydrogenatmospheremodelbyG¨ansicke,Braje&Romani2002,hyd
4
isconsistentwiththeentiredataset(χ2red=1.00,for109d.o.f.andwithanullhypothesisprobabilityof49.0%).FittingthesamedatasetwiththebestfitmodelbyRutledgeetal.(2002a)withtheadditionoftheBeppoSAXdata,weobtainaslightlyworsefit(χ2red=1.17,for113d.o.f.andwithanullhypoth-esisprobabilityof11.1%).AnequallygoodfitisprovidedbyaneutronstaratmosphereplusaComptonizationcomponent(COMPTT).Thismodelprovidesahighdegreeoffreedomwithaχ2red=0.99fit(for104d.o.f.andwithanullhypothesisprobabilityof50.5%)obtainedfixingthesoftWientemperaturetoatmospheretemperatureandtheplasmatemperaturekeptthesameinalltheobservations.Weconcludethatthescenariopro-posedis(atleast)equallywellconsistentwiththedata,mean-ingthatashockemissionscenariocanaccountforthespectralvariabilityobservedinAqlX-1.WealsonotethatRutledgeetal.(2002)found32%(rms)variabilityinobservationC4.Intheircasethepowerlawcomponentcontributedonly12%oftheflux.Fromourfitthepowerlawcomponentcontributesto38%ofthetotalflux,soitcaninprincipleaccountforallofthe
short-termvariability.
Despitethelownumberofpoints,spectralparametersde-rivedforthepowerlawindexshowsomecorrelationwiththecolumndensity(interpretedasameasureofthevariablemassaroundthesystem,overafixedinterstellaramount)aswellaswiththepowerlawflux.Thiscorrelationmightbeexpectedintheshockemissionscenario(Tavani&Arons1997).Whatisnowexpectedisthelargevalueofthepowerlawindexinthelastobservations.Thismightthenprovideanindicationofadifferentregimeinthesystem,possiblyunderlyingalargerin-verseComptoncooling.Thehardpartofthespectrumisinfactconsistentalsowithathermalbremsstrahlungspectrum.
Furtherobservationscanshedlightonthisnewinterest-ingfield,namelyvariabilityinthequiescentphaseofSXRTs,whichuptonowhasbeenoftenunconsidered.
WethankananonymousrefereeandG.Ghiselliniforusefulcomments.
REFERENCES
Asai,K.,etal.1996,PASJ48257Asai,K.,etal.1998,PASJ50611
Bildsten,L.,Chakrabarty,D.2001,ApJ557292
Bildsten,L.,Rutledge,R.2000,in“TheNeutronStar-BlackHoleConnection”(NATOASIELOUNDA1999),eds.C.Kouveliotouetal.(astro-ph/0005364)
Boella,G.,etal.1997,A&AS122327
Brookshaw,L.,Tavani,M.1993,ApJ410719
Brown,E.F.,Bildsten,L.,Rutledge,R.E.1998,ApJ504L95Burgay,M.,etal.,2003,ApJ589902
Campana,S.2001,intheproceedingsoftheBolognaconferenceon“X–rayAstronomy1999:StellarEndpoints,AGN,andtheDiffuseX-rayBackground”,eds.N.E.White,G.Malaguti,G.G.C.Palumbo,A.I.P.Conf.Proc.59963
Campana,S.,etal.1997,A&A324941Campana,S.,etal.1998a,A&ARev.8279Campana,S.,etal.1998b,ApJ499L65Campana,S.,etal.2000,A&A358583
Colpi,M.,Geppert,U.,Page,D.,Possenti,A.2001,ApJ,548,L175D’Amico,N.,etal.2001,ApJ,561,L89
Ferrario,F.,Possenti,A.,D’Amico,N.,Sabbi,E.2001,ApJ,561,L93
Fiore,F.Guainazzi,M.&Grandi,P.1999,CookbookforBeppoSAXNFIspectralanalysis,ftp://ftp.asdc.asi.it/pub/sax/doc/softwarev1.2.ps.gzG¨ansicke,B.T.,Braje,T.M.,Romani,R.W.2002,A&A3861001
Grindlay,J.,Heinke,C.O.,Edmonds,P.D.,Murray,S.S.,Cool,A.M.2001,ApJ,563,L53
Horne,K.1985,MNRAS213129
Kong,A.K.H.,etal.2002,ApJ570277
McClintock,J.E.,Remillard,R.A.2000,ApJ531956Nowak,M.,Heinz,S.,Begelman,M.2002,ApJ573778Parmar,A.N.,etal.1997,A&AS122309Piro,L.,etal.2001,ApJ558442
Rutledge,R.E.,Bildsten,L.,Brown,E.F.,Pavlov,G.G.,Zavlin,V.E.,1999ApJS124265
Rutledge,R.E.,Bildsten,L.,Brown,E.F.,Pavlov,G.G.,Zavlin,V.E.2000,ApJ529985
Rutledge,R.E.,Bildsten,L.,Brown,E.F.,Pavlov,G.G.,Zavlin,V.E.2001a,ApJ551921
Rutledge,R.E.,Bildsten,L.,Brown,E.F.,Pavlov,G.G.,Zavlin,V.E.2001b,ApJ5591054
Rutledge,R.E.,Bildsten,L.,Brown,E.F.,Pavlov,G.G.,Zavlin,V.E.,2002a,ApJ577346
Rutledge,R.E.,Bildsten,L.,Brown,E.F.,Pavlov,G.G.,Zavlin,V.E.,Ushomirsky,G.2002b,ApJ580413Stella,L.,etal.1994,ApJ423L47
Tavani,M.,Arons,J.1997,ApJ477439
Tavani,M.,Brookshaw,L.1991,ApJ381L21Titarchuk,L.1994,ApJ434570
Welsh,W.F.,Robinson,E.L.,Young,P.2000,AJ120943
WijnandsR.2001,in“TheHighEnergyUniverseatSharpFocus:ChandraScience”,proceedingsofthe113thMeetingoftheAstronomicalSocietyofthePacific(astro-ph/0107600)
Wilms,J.,Allen,A.,McCray,R.2000,ApJ542914
LOGOF
TABLE1
AQLX-1OBSERVATIONS.
SatelliteSeq.Num.StartTime
Exposure(s)Orbitalphase
φorb
BSAXLECSBSAXMECS2123800112123800112001-04-142001-04-1430390763010.38–0.50(±0.02)0.38–0.50(±0.02)
Note.—aOrbitalphaserelativetominimumlight(inferiorconjunctionofthesecondary),ephemerisfromGarciaetal.(1999).ForChandradatathesearetakenfromRutledgeetal.(2002a).
5
SPECTRALFITOF
TABLE2
CHANDRAANDBEPPOSAXAQLX-1OBSERVATIONS.
Parameter
Value(90%c.l.)Componentflux(10−12cgs)
NH(C1)Powerlaw(C1)
.2
6.1+1−1.2
.5
4.0+0−0.5
6.2(81%)
NH(C3)
Powerlaw(C3)
.1
4.6+3−1.4
.6
3.4+1−1.6
1.5(51%)
6
FIG.2.—Powerlawphotonindexvs.columndensitycorrelationofthefiveAqlX-1observations.Overplottedisthebestlinearfit.Dashedlinesindicatetherangeoverwhichthesynchrotronemissionmodellikelyapplies.Onthehardestandsoftestobservations1σcountourshavebeensuperposed.The1,2,3σcountoursobtainedfittingtheentiresetofdatawithasinglepowerlawandcolumndensityisalsoreported.
因篇幅问题不能全部显示，请点此查看更多更全内容