The physics of accretion-ejection with LOFT

1.5 MB·English

by  P. Casella

The physics of accretion-ejection with
LOFT
5
1
0
2

n
a
White Paper in Support of the Mission Concept of the
J

2
Large Observatory for X-ray Timing
1

]
E
H
.
h Authors
p
-
o P.Casella1,R.Fender2,M.Coriat3,E.Kalemci4,S.Motta2,J.Neilsen5,G.Ponti6,
tr M.Begelman7,T.Belloni8,E.Koerding9,T.J.Maccarone10,P.-O.Petrucci11,J.Rodriguez12,
s
a J.Tomsick13,S.Bhattacharyya14,S.Bianchi15,M.DelSanto16,I.Donnarumma17,P.Gandhi18,
[ J.Homan19,P.Jonker20,M.Kalamkar1,J.Malzac21,S.Markoff22,S.Migliari23,J.Miller24,

1 J.Miller-Jones25,J.Poutanen26,R.Remillard19,D.M.Russell27,P.Uttley22,A.Zdziarski28
v
6
6
7 1 INAF,OsservatorioAstronomicodiRoma,ViaFrascati33,I-00040 15DipartimentodiMatematicaeFisica,UniversitàdegliStudiRoma
2 MonteporzioCatone,Italy Tre,ViadellaVascaNavale84,00146Roma,Italy
0 2 DepartmentofPhysics,UniversityofOxford,KebleRoad,OX1 16IstitutoNazionalediAstrofisica,IASFPalermo,ViaU.LaMalfa
. 3RHOxford,UK 153,90146Palermo,Italy
1 3 UniversityofCapeTown,PrivateBagX3,Rondebosch7701,South 17INAF-IstitutodiAstrofisicaePlanetologiaSpaziale,ViaFossodel
0 Africa Cavaliere100,00133Rome,Italy
5 4 FacultyofEngineeringandNaturalSciences,SabancıUniversity, 18SchoolofPhysics&Astronomy,UniversityofSouthampton,High-
1 Orhanlı-Tuzla,34956,Istanbul,Turkey field,Southampton,SO171BJ,UK
: 5 EinsteinFellow,BostonUniversityDepartmentofAstronomy, 19KavliInstituteforAstrophysicsandSpaceResearch,MIT,70Vas-
v Boston,MA02215,USA sarStreet,Cambridge,MA02139,USA
i 6 Max-Planck-InstitutfürextraterrestrischePhysik,Giessenbach- 20SRON,NetherlandsInstituteforSpaceResearch,Sorbonnelaan2,
X
strasse,85748Garching,Germany 3584CAUtrecht,theNetherlands
r 7 JILA,UniversityofColoradoandNationalInstituteofStandards 21UniversitedeToulouse;UPS-OMP;IRAP;Toulouse,France
a andTechnology440UCB,Boulder,CO80309-0440,USA 22AstronomicalInstituteAntonPannekoek,UniversityofAmsterdam,
8 INAF,OsservatorioAstronomicodiBrera,viaE.Bianchi46,23807 SciencePark904,1098XHAmsterdam,theNetherlands
Merate,Italy 23DepartmentofAstronomyandMeteorology&InstituteofCos-
9 DepartmentofAstrophysics/IMAPP,RadboudUniversityNi- micScience,UniversityofBarcelona,MartíiFranquès1,08028
jmegen,POBox9010,6500GLNijmegen,theNetherlands Barcelona,Spain
10DepartmentofPhysics,TexasTechUniversity,Box41051,Lub- 24DepartmentofAstronomy,UniversityofMichigan,AnnArbor,MI
bock,TX79409,USA 48109,USA
11UJF-Grenoble1/CNRS-INSU,IPAG,UMR5274,38041Grenoble, 25InternationalCentreforRadioAstronomyResearch,CurtinUniver-
France sity,GPOBoxU1987,Perth,WA6845,Australia
12LaboratoireAIM,Irfu/Serviced’Astrophysique,CEA-Saclay, 26AstronomyDivision,DepartmentofPhysics,POBox3000,90014
91191Gif-sur-YvetteCedex,France UniversityofOulu,Finland
13SpaceSciencesLaboratory,7GaussWay,UniversityofCalifornia, 27NewYorkUniversityAbuDhabi,POBox129188,AbuDhabi,
Berkeley,CA94720-7450,USA UAE
14DepartmentofAstronomyandAstrophysics,TataInstituteofFun- 28CentrumAstronomiczneim.M.Kopernika,Bartycka18,00-716
damentalResearch,1HomiBhabhaRoad,Mumbai400005,India Warszawa,Poland
Thephysicsofaccretion-ejectionwithLOFT
Preamble
TheLargeObservatoryforX-rayTiming,LOFT,isdesignedtoperformfastX-raytimingandspectroscopy
with uniquely large throughput (Feroci et al., 2014). LOFT focuses on two fundamental questions of
ESA’s Cosmic Vision Theme “Matter under extreme conditions”: what is the equation of state of ultra-
dense matter in neutron stars? Does matter orbiting close to the event horizon follow the predictions of
generalrelativity? ThesegoalsareelaboratedinthemissionYellowBook(http://sci.esa.int/loft/
53447-loft-yellow-book/)describingtheLOFT missionasproposedinM3,whichcloselyresembles
theLOFT missionnowbeingproposedforM4.
TheextensiveassessmentstudyofLOFT asESA’sM3missioncandidatedemonstratesthehighlevel
ofmaturityandthetechnicalfeasibilityofthemission,aswellasthescientificimportanceofitsunique
core science goals. For this reason, the LOFT development has been continued, aiming at the new M4
launch opportunity, for which the M3 science goals have been confirmed. The unprecedentedly large
effectivearea,largegrasp,andspectroscopiccapabilitiesofLOFT’sinstrumentsmakethemissioncapable
of state-of-the-art science not only for its core science case, but also for many other open questions in
astrophysics.
LOFT’s primary instrument is the Large Area Detector (LAD), a 8.5m2 instrument operating in the
2–30keVenergyrange,whichwillrevolutionisestudiesofGalacticandextragalacticX-raysourcesdown
totheirfundamentaltimescales. ThemissionalsofeaturesaWideFieldMonitor(WFM),whichinthe
2–50keVrangesimultaneouslyobservesmorethanathirdoftheskyatanytime,detectingobjectsdownto
mCrabfluxesandprovidingdatawithexcellenttimingandspectralresolution. Additionally,themissionis
equippedwithanon-boardalertsystemforthedetectionandrapidbroadcastingtothegroundofcelestial
brightandfastoutburstsofX-rays(particularly,Gamma-rayBursts).
This paper is one of twelve White Papers that illustrate the unique potential of LOFT as an X-ray
observatoryinavarietyofastrophysicalfieldsinadditiontothecorescience.
Page2of11
Thephysicsofaccretion-ejectionwithLOFT
1 Summary
Understandingthedynamicsandstabilityofgasflowsaroundcompactobjectssuchasblackholes(BH)remains
oneofthemostpressingproblemsinhigh-energyastrophysics. Afullunderstandingofsuchflowsisnecessary
to build a complete picture of X-ray binaries and active galactic nuclei (AGN). This includes their radiative
output, the formation of (sometimes superluminal) jets and the launch of mass-loaded accretion disk winds.
TheradiationaswellasthekineticoutputofaccretingsupermassiveBHsprobablyplaysanimportantrolein
regulatingthegrowthofthebiggestgalaxiesinourUniverse. TofullyunderstandAGNfeedback,onehasto
understand the interplay between the three main outputs of accreting objects: the radiation, winds, and jets.
Notwithstandingthelargedifferencesbetweenstellar-massandsuper-massiveBHs, thephysicsatplaynear
theBHisexpectedtobesubstantiallythesame,albeitscaledwiththedifferentlengthscales,timescalesand
energy densities at work. Studies of BH accretion and outflows have exploited the differences between the
two BH classes through complementary approaches. The number of photons per characteristic time scale is
typicallyseveralordersofmagnitudehigherinAGN,whichhasallowedustoobservethedetailedphysicsat
playinindividual“events”. Ontheotherhand,stellar-massBHshavemuchshortercharacteristictimescales,
allowing us to observe a large number of cycles, providing a good picture of the average behaviour of each
physicalprocessandcontrolfortheeffectsofunknownparameterssuchasdistance,massorspin. Theongoing
improvementindetectorefficiency,togetherwiththeincreaseintelescopearea,arenowbridgingthegapacross
theBH-massrange,asthefastestcharacteristictimescalesstarttobecomeaccessibleevenforstellar-massBHs.
Manybasicquestionsremainunanswered,withunknownsincludingthegeometryandtheradiativeefficiency
of the accretion flow, the physical mechanism triggering the transitions between different accretion regimes,
the particle composition of jets, the amount of mass carried by the winds, and the launching and powering
mechanismsofjetsandwinds. LOFT willrevolutionisethisfield. ThankstothehugeeffectiveareaoftheLAD,
itsCCD-classspectralresolution,andtotheoutstandingmonitoringperformancesoftheWFM,LOFT will:
measuretheX-rayspectralcontinuumwithunprecedentedaccuracy,allowingustostudythedetailed
•
evolutionoftheaccretiondiskandtheComptonizingmediumonthetimescaleofobservedtransitions: in
afewsecondsduringhard-to-softbrighttransitions,afewhoursduringthesoft-to-hardfainttransitions;
detectX-rayabsorptionandemissionlinesfromwindsandjetswithoutstandingstatistics,allowingusto
•
identifythemomentswhenthewindandthejetturnon/off,andtotracktheevolutionoftheirphysical
properties (including density, ionisation, velocity, baryonic content) on time scales as short as a few
secondsandafewhoursduringthebrightandthefainttransitions,respectively;
identifyandpinpointeachfasttransitionwiththeWFM,toestablishacrucialcomponentoftheupcoming
•
multi-wavelengthlegacy;
carryoutallthesebreakthroughsatthesametimeasunprecedentedmeasurementsofthestructureofthe
•
inneraccretingandemittingregions,tolinkchangesinjet,windorradiativeoutputdirectlytochangesin
thecentralenginewhichdrivesthem.
ItisimportanttonotethatLOFT willperformmostofthesemeasurementswiththeobservationsalreadyplanned
tomatchtheCoreSciencerequirements.
2 Introduction
Accretion onto BHs(and to a comparable degree, neutron stars) isthe most efficient source ofenergy inthe
contemporaryUniverse,poweringbothX-raybinaries(XRBs)andAGN.Theenergyliberatedthroughaccretion
emergesintheformofradiationorkineticenergy,thatcaneitherbecollimated(so-calledjets)oruncollimated
Page3of11
Thephysicsofaccretion-ejectionwithLOFT
(winds). The interplay between these three components is not yet understood, neither in stellar-mass nor in
supermassive BHs. While jet ejection events, for example, can be studied in detail in AGN, as their relative
time-scale(setbytheirgravitationalradius)islonger,theevolutionofanaccretingBHunderchangesofthe
accretion rate can be better-studied in XRBs. The luminosity of accreting stellar-mass BHs can vary by as
muchas8ordersofmagnitudebetweenquiescenceandpeakoutburstwithinacoupleofmonths. Asitchanges
itsluminosity,anXRBevolvesthroughasequenceofaccretionstates. Inparticular,twocanonicalstatesare
seen–a“hard”statecharacterisedbyComptonizedhardX-rayemission,highamplitudeX-rayvariabilityand
acompactradio jet, anda“soft” statecharacterisedbystrong X-raythermalemission, lowamplitudeX-ray
variabilityandanequatorialwind. Transitionsbetweenthesecanonicalstates(Begelman&Armitage,2014)
proceedthroughavarietyofintermediateandextremestates,inwhichtheaccretionenergybudgetredistributes
itselfamongthevariouscomponentsontimescalesasshortasafewseconds,asevidentfromX-rayvariability
(see, e.g.,Belloni,2010). Thesesourcesarethusideallysuitedforstudyingtheformationofjetsandwinds,
whichappearandsubsequentlydisappearduringeachoutburstcyclewithcomplexphenomenologyonabroad
rangeoftimescales. WewillthusdiscussthethreedifferentoutputsofanaccretingBH:radiation,windandthe
jetsinviewofthepossibilitiesLOFT willofferinthefuture. LOFT willforthefirsttimebeabletodetectthe
discwindroutinely,andthusofferinasingleobservationinformationaboutboththeradiativepartaswellasthe
mechanicalenergyinthewind. TheproposedlaunchdateandlifetimeofLOFT correspondnicelytotheperiod
ofinitialsurveyswiththephase1oftheSquareKilometreArray. Duetoitsincreasedsensitivityitwillthenbe
possibletoobtaindensecoverageoftheradiolight-curve,allowingalsoforawellstudiedjetcomponent. Aswe
willarguebelow,LOFT willopenaneweraofresearchintothecouplingbetweenaccretionandejection.
3 Getting close to black holes: unlocking the origin of the most powerful jets and winds
Inwhatfollows,wedetailthebreakthroughsthatLOFT willmakeintacklingthephysicsofblackholejets,
windsandtheradiativeemission. Butfirst,itisimportanttostressthatalltheseresultswillbeobtainedatthe
same time as the unprecedented measurements of dynamics and structure of the inner accretion flow which
form the LOFT core science, aimed at testing and quantifying the effects of strong-field gravity (SFG; see
http://sci.esa.int/loft/53447-loft-yellow-book). Thesemeasurementswillusehigh-throughput
spectral-timingofrapidvariabilitytocarryoutX-rayreverberationmappingonsub-gravitational-radiusscales,
andDopplertomographyofQPOstomeasurethedynamicsoftheaccretionflow. Thissuiteofindependent
techniqueswillprovideaccuratemeasurementsofblackholespin,thediscinnerradiusandthegeometryand
locationoftheX-raypower-lawemittingcorona. Moreover,thesemeasurementswillbeobtainedthroughout
therangeofstatesseeninX-raybinaryoutbursts,notablythemostluminoustransitionalstateswhenthemost
powerfuljetsandwindsswitchonandoffandtheX-rayemissionshowsthestrongestspectralevolution.
Thus,LOFT’sSFGmeasurementswillallowthedetailedchangesinstructureandgeometryoftheinnerflow
andemittingregionsduringstatetransitionstobemappeddirectlyontothepropertiesofthejetorwindand
theemissionspectrum,inrealtimeasthesystemevolves. Energetically,weknowthatthemostpowerfuland
energeticoutputsmusthavetheiroriginintheseinnerregions. ThesamephenomenacanbeseeninAGNand
mustevolveonmuchlongertime-scaleswhichwecannotobservedirectly. LOFT studiesofBHX-raybinaries
will therefore allow an unparalleled insight into the physical causes of accretion-powered jets and winds at
accretionratesuptotheEddingtonlimit,providingthecrucial(andcurrentlymissing)informationformodelling
accretionandfeedbackintheirsupermassivecousins,whichmayultimatelyberesponsibleforre-shapingentire
galaxiesandgalaxyclusters.
Page4of11
Thephysicsofaccretion-ejectionwithLOFT
4 The radiative output: the X-ray continuum
TheX-raycontinuumemissionfromaccretingBHshasbeenstudiedandmodelledfordecades. Inthehardstate,
thespectralenergydistributioniswell-modelledbyapowerlawwithaphotonindexofabout1.7andacutoffat
afewhundredkeV,whileinthesoftstate,thermalemissionfromthegeometricallythinaccretiondiskdominates,
and the power law tail is steeper, with no trace of a cutoff out to at least 1MeV (Grove et al., 1998). These
spectraarebroadlyinterpretedastheresultofComptonizationfromacombinationof(inflowingoroutflowing)
thermalandnon-thermalelectrons(see,e.g.,Markoffetal.,2001,2005;Ibragimovetal.,2005;Poutanen&
Vurm,2009). Despitethewealthofobservationalandtheoreticaleffortsinunderstandingthisemission,many
openquestionsremainonthegeometryandenergydensityoftheComptonizingelectronpopulations,mostly
becauseofthedegeneraciesbetweendifferenttheoreticalmodels. Time-resolvedspectralstudiessuggestthat
theComptonizingmediumisinhomogeneous,withmultipleelectronpopulations(forexampleradiallystratified
intemperature/opticaldepth,oroutflowing)contributingtotheobservedcomplexComptonizedemission. The
statisticsoftheLADdataissuchthatitwillbepossibletodeconvolvethecontinuumfromitsreflectionasa
functionofradius,testingtheoreticalmodelsandmeasuringthephysicalparametersofthemultipleelectron
populations.
4.1 Bright and fast: physical measurements in a few seconds
Asthegeometryandthermalpropertiesoftheaccretionflowareexpectedtobedrasticallydifferentinthehard
andinsoftstates,apromisingwaytodiscriminatebetweenmodelsistoobserveindetailtransitionsbetween
thesetwoaccretionstates. Thebrightestofsuchtransitions,outofthehardstate,areassociatedwiththeejection
ofpowerfultransientrelativisticjets(Fenderetal.,2009),whilethefainttransitionsbacktothehardstateduring
theoutburstdecayareassociatedwiththereappearanceofthecompactsteadyjet. Asthesejetsarepowerful,it
isreasonabletoexpecttheirejectiontohaveasubstantialimpactontheaccretionflow. Nevertheless,duringthe
bright,fasttransitions,thespectralparametersbelow10keVvarylittleandsmoothly,seeminglyunawareofthe
presence/absence/propertiesofthehighlyvariablejet. Strongandnon-monotonicchangesareindeedobserved
in the properties of the hard tail in a few hours or days (Motta et al., 2009), but nothing is known about the
spectralvariabilityontimescalesasshortasthoseoftheobservedX-raytransitions,duringwhichquasi-periodic
oscillations and broad-band noise appear/disappear in a few seconds. The epochs of these rapid transitions
indeedseemtobeassociatedwithrapidgeometricalreconfigurationsoftheaccretionflow(Motta,2014),buta
properphysicalunderstandingofthemhasbeenhamperedsofarbythelimitedstatistics,whichhasprevented
spectralanalysisonsuchshorttimescales. ThelargeeffectiveareaoverabroadenergyrangeoftheLAD/LOFT
willchangethis,makingthespectralevolutionaccessibleonthefasttimescalesofthetransitions. Duringthese
hard-to-softtransitions,thetimingpropertiesareoftenobservedtochangerapidly,withthebroad-bandnoise
andquasi-periodicoscillationsappearing/disappearingontimescalesshorterthan10seconds(Casellaetal.,
2004),sometimesmultipletimes(Mottaetal.,2011). Figure1showstwoexampleofLADspectrawithvery
short exposures. The two spectra simulate the expected spectral variability during such extreme transitions,
asinferredbylongaveragedRXTEobservations. Exposuresasshortas2secondswillbesufficienttodetect
significantdifferencesinthespectra, measuringallspectralparameterswitha5%accuracyorbetter. Atthe
same time, it will be possible to examine on very short time scales the detailed properties of the sub-second
oscillationsdetectedbeforeandaftertransitions. Linkingthemtothespectralresultswillallowustoassociate
themtochangesinspectralparametersandtoestablishtheirconnectionwiththeaccretionandejectionflows.
4.2 Dim and slow: monitoring during outburst decay
Thedim,slowtransitionbacktothehardstateduringtheoutburstdecayisanothercrucialepochtotestmodels
andunderstandaccretionphysics(Maccarone,2003). Therelevanttimescalesherearedifferent,asthetransition
Page5of11
Thephysicsofaccretion-ejectionwithLOFT
Figure1: Simulated 2second ex-
posures LAD/LOFT spectra for
GX 339 4. The black and
−
red spectra come from observa- 1!"0!
tions with and without broad-
bandnoisevariability,respectively.
Datawerefittedwithanabsorbed -1eV)!!,-
k%
dpptiloivusnikssetrnibectfllaheecffackseti-cbobtenose..dnTychoperlrrueesflctepecodtwifooenrrcrloeamlwa-- -2-1otons cm s !.!!121$3)45)3)+
2V (Ph.%,)*/0
ke+!1""
1!"0!
E#n$e%&r’g()y*+ %(,k-eV)
lastsdaysorweeks,andthefluxesareordersofmagnitudefainter. Thecompactsteadyjetisknowntoreappear
during this slow transition, although the actual association between the jet properties and the X-ray spectral
evolution is still debated (Kalemci et al., 2005, 2013; Dinçer et al., 2014). Despite coordinated efforts to
understandtheevolutionoftheinnerdiskradiusasafunctionofluminosityduringthesubsequentdecayphase
(Tomsicketal.,2009;Petruccietal.,2014;Plantetal.,2014),wehaveyettodeterminehowthediskmoves
awayfromtheBHduringdecays. AlsotobeunderstoodisthemarkedX-raysofteningobservedinsomesources
duringthedecay,whichcouldbeduetoanevolutionofthepropertiesoftheComptonizingmedium(geometry,
electrontemperatureanddensity,seedphotons;Sobolewskaetal.,2011)ortothepresenceofasecondsource
ofemission(perhapsajet;Russelletal.,2010)dominatingbelowacriticalluminosity. Thelimitedeffective
areaofpastandcurrentX-raysatellites,andtheconsequentneedforlongexposures,havehamperedsofara
propermonitoringofthistransition. TheLADlargeeffectiveareawillallowustotracktheevolutionofthe
spectralpropertiesthroughoutthetransitionandearlyphasesofthedecaywithexposuresasshortasafewhours.
Fora5mCrabsource,a2ksLADobservationwillprovidebetterconstraintsthana40ksSuzaku exposure(as
reportedbyPetruccietal.,2014)totheparametersofthehardX-raycomponentand–ifthegoalof1.5keVsoft
energylimitisreached–eventothoseofthesoftdiskcomponent. Atsimilarfluxlevels,a5ksLADexposure
willbeenoughtoconstraintheinnerdiskradiusthroughironlinemodelling,whiletheexposuretimeshouldbe
increasedtoaround25kstoobtainthefullevolutiondowntoa0.5mCrabfluxlevel.
4.3 The multi-wavelength legacy: solving degeneracies through multi-wavelength timing
Inrecentyears,therehasbeenatremendousamountofinformationcomingfromwavelengthsotherthanX-rays,
rangingoverthewholeelectromagneticspectrum,fromradioallthewaytotheTeVrange. Cruciallyandperhaps
notsurprisingly,thepropertiesofaccretingBHsintheseotherwavebandshavebeenshowntocorrelatewith
thoseinX-rays. Amongthesemanynewobservationalwindows,hightimeresolutionobservationsatoptical-
infraredwavelengthshaveyieldedtheimportantdiscoveryofvariablenon-thermalemissioncorrelatedwiththe
X-rayfluxontheshortestaccessibletimescales(Malzacetal.,2004;Gandhietal.,2010,e.g.). Dedicatedfast
optical/IRphotometerswithgoodquantumefficiencyallowustotackleoldquestionsfromanewperspective:
thefastesttimescalesofaccretionontostellar-massBHsarebecomingaccessibleforstudyinthesynchrotron
emissionfromthesameelectronpopulationresponsiblefortheX-rayComptonizedemission(e.g.,Veledina
Page6of11
Thephysicsofaccretion-ejectionwithLOFT
et al., 2013). At longer wavelengths we can now study synchrotron emission from the electron population
outflowing in the relativistic jet (Casella et al., 2010). Such observations are crucial in order to solve X-ray
continuumdegeneracies,asdifferentphysicalscenarios–whichwouldotherwisepredictverysimilarX-ray
spectralproperties–predictextremelydifferentOptical/Infrared/X-rayspectraltimingproperties. Ifthefirstkey
resultshavebeenlimitedsofarbythenonoptimalOIRtechnologiesandbytheavailableX-raystatistics,the
nextdecadewillseesignificanttechnologicaladvances. Observationsofrapidvariabilityinallwavebandswill
providerevolutionarynewdataforthestudyofaccretionandejection.
Inparticular, thefieldwillbenefitfromthecombinationofthenextgenerationofOIRdetectors–suchas
theMicrowaveKineticInductanceDetectors(MKIDs),whichareabletomeasuretheenergiesofindividual
optical/IRphotonswithoutusingfiltersorgratings,obtainingsimilarspectralresolutionstothosepossiblewith
X-rayCCDinstruments(e.g.,theARCONScamera;Mazinetal.,2013)–andthelargestatisticalthroughput
offeredbyLOFT.Equallyimportantwillbetheexpectedincreasedavailabilityof4–8mclasstelescopes,on
whichtomountsuchinstruments,whichwillprovidethephotoncountratesneededtostudyrapidvariability.
The sum total of all this will result in a twofold field expansion: (a) physical measurements via correlated
OIR/X-raytimingobservationswillbeaccessibleonthetimescalesofthefastesttransitions,helpingtoreveal
theexpectedgeometricalreconfigurations;and(b)itwillbepossibletoperformsuchmeasurements,nowlimited
tothebrightestfluxesand/orthebrightestsources,atorder-of-magnitudelowerfluxes,allowingfullmonitoring
throughthewholeoutburstevolutionofseveralsources.
5 The mechanical output: winds
AlthoughwehavelearnedagreatdealaboutaccretiondiskwindswhilemonitoringGalacticBHsinthelasttwo
decades(see,e.g,Milleretal.,2006;Doneetal.,2007;Milleretal.,2012),therearemanyremainingquestions
(seealsoTombesietal.,2013,andreferencestherein,forwindsinAGN).Whendowindsfirstappearinoutburst
andwhy;whendotheydisappear? Dotheyinteractwithjetsordisruptjetformation(Neilsen&Lee,2009)?
Weknowthatwindsarereliablydetectedafterthetransitionoutofthehardstateandthedisappearanceofthe
jet (Ponti et al., 2012; Neilsen & Homan, 2012), but the instant of their formation has not been definitively
identified. Thismoment,however,hassweepingimplications: ifwindsoriginateafterthestatetransitionandjet
quenching,theycannotpossiblyplayamajorroleinthoseprocesses. Otherequallycrucialunknownsarethe
actuallaunchingregion,aswellastheirlaunchingmechanism(aretheylaunchedbymagneticprocessesthattie
themtojets?). Also,althoughtheirkineticenergymightbeoverallmuchsmallerthanthejetkineticpowerorthe
radiativeluminosity,theymaycarryalargeamountofmassawayfromtheaccretionflow,possiblyinfluencing
thelong-termevolutionoftheoutburstoreventhatofthebinarysystemandofthespinoftheaccretingobject.
Finally,theverystructureoftheseoutflowshasyettobestudied,aswellastheinfluencethatvariationsinthe
luminosityspectrummayhave.
As the relevant physical processes happen on very fast time scales, time-resolved spectroscopy is the key
toansweringthesequestions. Thankstoitslargeeffectivearea,joinedwithitsCCD-classspectralresolution,
theLADwillrevolutionisestudiesofwindsfromaccretingBHs. Suchwindsareknowntobeubiquitousin
high-inclinationBHXRB.ThecharacteristicsignatureofanaccretiondiskwindistheFexxviabsorptionline,
withanequivalentwidthof.35eV.LOFT willdetectthesestrongabsorptionlinesat3σconfidenceinaslittle
as1–2seconds(roughly1000timesfasterthanChandra). Thiscombinationofsensitivityandtimeresolution
opensupnewavenuesofinquiry,includingabsorptionlinevariabilityasaprobeoftheturbulentstructureof
winds. BycomparingobservedX-rayvariabilitytofluctuationsintheionisationparameter(ξ = L/nR2,wheren
istheparticlenumberdensityandRisthedistancetothesource),LOFT willallowustodisentangletheeffects
ofphoto-ionisationandtheclumpystructureofthewind.
Forrepresentativewindparameters(N = 5 1022cm 2 andlogξ = 4.3),> 6%fractionalRMSvariations
H −
×
intheluminositywillinducedetectablechangesinξ;thus,anyanomalousionisationwillbeusedtomapthe
Page7of11
Thephysicsofaccretion-ejectionwithLOFT
K! K!
K" K"
S XV
Fe XXVI Fe XXV Fe XXVI
Fe XXV Fe XXV
S XV Fe XXV
Fe XXVI
Fe XXVI
Figure2: Left: Simulated1ksLADexposureof4U1630 47(0.3Crab,2–10keVabsorbedflux). TheFexxvandFexxvi
−
linesareclearlyresolvedintoKαandKβ. Right: the78ksXMM-Newton observationforcomparison(DíazTrigoetal.,
2014).
density structure of the wind. Furthermore, simulations show that wind speeds above 450kms 1 will be
−
∼
reliablymeasuredbyLOFT.Thus,theLADwillbeabletotracknotonlythestructureofwinds,butalsotheir
masslossratesasfunctionsoftheaccretionstateandtheaccretionrateforcluestotheirformationphysicsand
theirinfluenceonthediskandthejet.
Complementing the significant advances in our understanding of the structure and variability of winds on
shorttimescales,LOFTwillalsoprovideimportantinformationaboutthelaunchingandquenchingprocesses
for winds. The LAD will detect weak Fe xxvi lines (equivalent width 7.5eV) in 1ks in sources as faint as
2 10 10ergcm 2s 1,i.e.,morethantwoordersofmagnitudebelowthepresentsoftstatedetection. Suchshort
− − −
×
exposuresevenatlowfluxeswillallowmonitoringobservationsduringboththerisingandthedecayingphases
ofBHoutbursts. LOFT willthusreadilyrevealthemomentswhenwindsfirstappearandwhentheyeventually
quench. Again,thesemomentsarecriticalforacompleteunderstandingofwindsfromaccretingBHs: isthe
appearance/disappearanceofwindsaneffectcausedbychangesintheirtemperature(thusintheirtransparency
toX-rays),ordotheyactuallystartandcease? Inordertoanswerthesequestions,bothalargeeffectivearea
andausefulspectralresolutionarecrucial. Indeed,simulationsindicatethatLOFT willbemoresensitivethan
Athena towindabsorptionlinesevenatfluxesaslowasafewmCrab,withsignificantdetectionsinsourcesat
leastafactoroftwofainter. Thisisbecausethelinesarenotintrinsicallynarrow(widthsof 500km/sormore
∼
arefairlytypical),sotheimprovementduetothelargereffectiveareaofLOFT dominatesoverthedifferencein
spectralresolution. Athigherfluxes,theeffectofpileupmightstarttoplayanimportantrole,furtherdecreasing
thethroughputofAthena.
WithtightconstraintsfromtheLADontheionisingradiationfieldthroughouttheseoutbursts,LOFT willbe
abletrackthemasslossrateinwindsonthefastestusefultimescalesfromthemomentoftheirformationto
theirdisappearance.
6 The mechanical output: jets
In accreting BHs, jets radiate predominantly through processes such as synchrotron and inverse Compton
emission,fromtheradiothroughX-raywavebands. Twodifferenttypesofjetareobservedduringanoutburst: a
steady,mildlyrelativisticjetisdetectedinthehardstate;discrete,veryenergeticejectionsofrelativisticmatter
areobservedduringthetransitiontowardthesoftstate,probablyassociatedwithrapidchangesinthespectral
distributionofthehardX-rayemissionandwithdrasticchangesintheX-raytimingproperties. Amongthemost
Page8of11
Thephysicsofaccretion-ejectionwithLOFT
Figure 3: Left: simulated 1ks LAD exposure of 4U 1630 47 ( 2.5Crab, 2–10 keV absorbed flux). Right: the 55ks
− ∼
XMM-Newton burst-modeobservation(DíazTrigoetal.,2013,resultingin1.6ksactualexposure),forcomparison. The
weakestline(thered-shiftedFexxviline)whichwasdetectedbyXMM-Newton ata3.5σlevel,isdetectedbyLOFT at
46σlevel.
hotlycurrentlydebatedquestionsthereare: howmuchoftheaccretionpowergetsreleasedintothesurroundings
viathesecollimatedjets? Whatistheirlaunchingmechanism? Aretheyqualitativelydifferentfromwinds,orare
theyanextreme,focusedversionoftheinnerpartofawind? Whatistheircomposition? Dothejetsthemselves
contributetotheobservedX-rayradiation? Howandwheredojetsaccelerateparticlestoveryhighenergies?
Whatistheirrelationtothevariabilityintheinflow? LOFTwillallowustoinvestigatetheroleofrelativistic
jetsincombinationwiththeevolutionofthecharacteristicsofinflowandwind,helpingustounderstandfurther
whatisphysicallydrivingthevariedbehavioursinthesesources.
6.1 X-ray spectroscopy: the jet baryonic content
Theomnipresenceofjetsinastrophysicalsystemsmakesaproperunderstandingoftheirphysicsparticularly
relevant. Areallthesejetsproducedbythesameprocesses? Sinceoneofthecanonicaljetformationmechanisms
(theBlandford-Znajek,orBZ,mechanism;Blandford&Znajek,1977),involvestherotationpowerofaBH,it
certainlycannotoperateinallsystems. AnalternativeistheBlandford-Paynemechanism(BP;Blandford&
Payne,1982),inwhichtherequisiterotationpowercomesfromarotatingaccretiondisk. Asaconsequence,one
expectsBPjetstobebaryonic,sincetheyaremagneto-centrifugaloutflowsfromagaseousreservoir. BZjets,on
theotherhand,mightbe–atleastintheearlystagesafterlaunch–electromagneticallydominated.
Thusitisparticularlyinterestingthataccretingstellar-massBHsappeartoproducetwodifferentkindsofjets
(steadyandtransient;forreviewsseeFenderetal.,2004,2009). Observationally,itremainsunclearwhether
thesetwotypesofjetscorrespondtothetwomechanismsdescribedabove;asuggestedassociationbetween
thetransientjetsandtheBZmechanismhasbeenhotlydebated(Fenderetal.,2010;Narayan&McClintock,
2012;Russelletal.,2013). Differentprocessesmaylaunchjetsatdifferentscalesorunderdifferentconditions,
butbecauseajet’senergyrequirementsaresensitivetoitsbaryoncontent,theseconsiderationshavebroadand
concreteimplicationsforradio-modefeedbackfromaccretingsystems(Fenderetal.,1999). TheXMM-Newton
detectionofrelativisticallyblue-andred-shiftedX-rayemissionlinesfromtheblack-holecandidate4U1630 47
−
(DíazTrigoetal.,2013)openedanewperspectiveonthesequestions. Withaninferredvelocityof0.66c,these
FeandNilinescoincidedwiththeappearanceofopticallythinradioemissionobservedbyATCA,suggestingthe
formationofafast,hotbaryonicoutflow. AnXMM-Newton/ATCAobservation3weeksearlierdetectedneither
Page9of11
Thephysicsofaccretion-ejectionwithLOFT
radioemissionnorX-rayemissionlines. Ifthisdetectionisconfirmed,4U1630 47thusbecomesthesecond
−
knownsourceofapparentlyheavyjets,togetherwiththe“oddball”SS433(Margonetal.,1979). Inorderto
understandthefullimplicationsoftheseresultsforthejetformationmechanismandenergeticbudget,alarge
searchforbaryonicsignaturesinblackholejetsmustbecarriedout. Thelargeeffectiveareaandgoodenergy
resolution of the LAD will allow tracking of X-ray emissions lines from jets with unprecedented sensitivity
andonshorttimescales. Ifwedetecttheseemissionlines,wewillpreciselymeasuretheirvelocitiesandline
widths. The ratio of the line width to the blue-shift is an important diagnostic of the orientation, expansion,
andgeometryofthejet. PreviousXMM-Newton observationscouldonlyindicatethatthelinesdidnotappear
to be intrinsically narrow. In addition, the line strengths will constrain the thermal properties and emission
measureofthemass-loadedjet. Asthesequantitiesevolve,wewillcorrelatethemwiththeX-rayluminosity,the
X-rayspectralandtimingpropertiesaswellastheradioandinfraredemissionfromthejets. Thevariationsof
thesequantitieswillprovideourfirstempiricalinsightsintotheformationofheavyjetsfromBHs. Figure3
(Left panel) shows a LAD simulated spectrum of jets X-ray emissions lines obtained in 1ks exposure time.
An improvement of a factor 15 in signal to noise ratio is achieved with respect to the corresponding 55ks
longXMM-Newton observationof4U1630 47(Fig.3rightpanel;DíazTrigoetal.,2013). Forsourceswith
−
comparablefluxes(whichexceedthecurrentfluxlimitationsforAthena),exposuresasshortas10secondswill
besufficienttodetectthered-shiftedFeXXVI(thefaintest)linewithLOFT atthe4σlevel. Therefore,tracking
thejetvelocityonminutetimescaleswillbecomepossiblewithLOFT.
7 The multi-wavelength legacy and the role of the WFM
The proposed launch date and lifetime of LOFT correspond to the period of initial surveys with the Square
KilometreArray(SKA;seeskatelescope.org)phaseone(SKA1). IntheongoingSKAScienceReviewProcess,
TransientsandHighEnergyAstrophysicshavebeenidentifiedaskeysciencedriversfortheproject,ensuring
supportforthesedirections. Althoughcurrentlyundergoingadesignre-baseliningprocess,SKA1isenvisaged
tohavethreecomponents,allofwhicharepotentiallyimportantcontributorstoLOFT science. SKA1–Lowwill
operate,likeLOFARandMWAatpresent,atverylowfrequencies,andwillhavetheadvantageofaverylarge
fieldofview(morethan100squaredegrees)andsensitivitytocoherentbursts. Ontheotherhand,SKA1-Mid
and SKA-Survey will operate in the 1 5 GHz frequency range, and will be two orders of magnitude more
−
sensitivethancurrentfacilities,achieving0.72microJyinanhourinthe1.4GHzband(about40microJyin
onesecond). Thus,alsothankstothehighlyflexibleschedulingbuiltin,itwillbepossibletomonitorallactive
X-ray binaries at high cadence. Over the 5–year lifetime of the SKA1 surveys, it is anticipated that there
∼
wouldbemanythousandsofhighS/NradioobservationsofX-raybinaries. Inopticalbands,clearplansexistto
makecontinuoussurveysofthesky,aswell. TheLargeSynopticSurveyTelescope(Ivezicetal.,2008)will
observetheentiresouthernskyeverythreedays,downto24thmagnitude. TheZwickyTransientFactory(see
www.ptf.caltech.edu/ztf)willmonitorthenorthernskywithmorethan300epochsperyear,downtomagnitude
20.5orbetter. Evryscope(Lawetal.,2014)willobserveanentirehemisphereoftheskyeveryminutedownto
magnitude16.5,withone–hourdepthof19thmagnitude. Theseprojects,too,willgivewide–fieldcoverageof
transients. Furthermore,theCherenkovTelescopeArray(CTAActisetal.,2011,seewww.cta-observatory.org)
isthenext-generationhighenergy(from30GeVtomorethan100TeV)gamma-rayfacility,plannedtostartfull
operationby2020,involvingaworldwidecollaborationofmorethan1000scientistsandabout30countries.
WithsensitivityandenergycoverageanorderofmagnitudebetterthanthatofH.E.S.S.,plusafactorofafew
improvementonspatialresolutionandfieldofview,CTAwillberunasanobservatory,allowingproposalsfor
targetedobservations,andincludingalsoaToOmode,withareactiontimeoflessthanaminute. CTAwillbe
thekeyTeVpartnerforLOFT,forco-monitoringofhigh-energyemissionsitesandhelpingtotraceparticle
acceleration(ofbothleptonsandpotentiallybaryons).
Page10of11