Ilic AGN variability · binary supermassive black holes (SMBH) – tricky to detect on...

SpectralvariabilityofAGNDraganaIlić

LukaČ.Popović,Andjelka Kovačević,Alla Shapovalova☨

1.DepartmentofAstronomy,FacultyofMath.- University ofBelgrade,2.AstronomicalObservatoryBelgrade,3.SAORAS

AGNopticalspectralvariability

• EVERYTHINGVARIES!– linefluxandcontinuumflux– broademissionline(BEL)profiles

• insomecasesextreme:AGNevenchangetheirtype– e.g.type1->type2– changinglookAGN

MSEScienceMeeting,Tucson,2019TheAGNspectralvariability

0.2

0.4

0.6

0.8

1

50000 51000 52000 53000 54000

Flux

(Hβ)

Julian Date [2400000+]

NGC 4151

NGC4151Hβlightcurve

NGC3516max&minstate

Shapovalova+2019

wavelengthFlux

Shapovalova+2008

• EVERYTHINGVARIES!– linefluxandcontinuumflux– broademissionline(BEL)profiles

• insomecasesextreme:AGNevenchangetheirtype– e.g.type1->type2– changinglookAGN

MSEScienceMeeting,Tucson,2019TheAGNspectralvariability

0.2

0.4

0.6

0.8

1

50000 51000 52000 53000 54000

Flux

(Hβ)

Julian Date [2400000+]

NGC 4151

NGC4151Hβlightcurve

NGC3516max&minstate

Shapovalova+2019

wavelengthFlux

Shapovalova+2008AGNopticalspectralvariability

AGNanditsBroadLineRegion(BLR)• type1(obscured)AGN- withbroad

emissionlinesà BroadLineRegion

• howdoweknowthereisaBLR?à imaging:VLT- GRAVITY(10μas,Sturm+18),and

futureELTs

à spectroscopystillimportant

• yet,BLRphysicsandgeometrynotfullyknownà e.g.whatisgastemperatureanddensity?(Ilic+12)à isitvirialized tothesupermassiveblackhole?à dowehaveoutflows,inflows?(e.g.Wang+17)à whatistheinclination?(e.g.Afanasiev+18)

MSEScienceMeeting,Tucson,2019TheAGNspectralvariability 4

SDSSmeanquasarVanden Berk+01

JURIK PETER / SHUTTERSTOCK

WhytheBELofAGN?• reverberation:directmeasureoftheunseen

(theBLRsize)andtheSMBHmass

• theeraofmultimessengerastronomy:detectbinarysupermassiveblackholes(SMBH)

– trickytodetectonparsec-scales(e.g.Popović+12,Komossa+03,Comerford+09,Ge+12,Benitez+18)

– fromspectroscopy(Bon+12,16,Liu+16)orperiodicityanalysisofphotometriclight-curves(Graham+09,17)

• theultimateaimofastronomytodayà newtoolstomeasuredistanceswithAGN

– e.g.usingopticallinesoftheBLR(e.g.Watson+11)

MSEScienceMeeting,Tucson,2019TheAGNspectralvariability 5

NASA/CXC/A.HOBART,JOSHBARNES(U.OFHAWAII),JOHNHIBBARD(NRAO)

Conclusions

ReverberationMonitoring(RM)• thereisatime-delay btwn continuumandlineflux

• onlyfor~100sAGN:directmeasureofRBLR

• empiricalRadius-Luminosityrelation(e.g.Kaspi+2000,Peterson+2004,Bentz+2009)

à cangetSMBHmassfromsingleepochobservation

à ifwedirectlymeasureRà Là distance

MSEScienceMeeting,Tucson,2019TheAGNspectralvariability

NGC4151lightcurves

time-lags from CCF

Line

continuu

m

ModifiedJD

Shapovalova+2008

Lyutyi&Cherepashchuk,1972;Blandford&McKee,1982;Gaskell&Sparke,1986

MBH = fRBLRFWHM

2

GBentz+2009

time-delay τ = size RR

RMcampaignsQUANTITY

• SDSSreverberationmonitoring– monitoring849quasars,0.1<z<4.5 (Shenetal.2014,2018)– measuredfor44quasars,z<0.3(Grier+2017)and144

quasars,z<1(Li+2017)

• Subaru/FMOSsurvey(Shultze+2018)– 234quasars,highestredshiftz=4.6

basedonHα(211),Hβ(63)andMgII(4)

• OzDES RMcampaign(King+2015,Hoormann+2019)– 23variableAGNsoutof771monitored,0.1<z<4.5

QUALITY• highfidelityRM(Horne+2012)togetvelocity-resolved

RM(DeRosa+2018,Du+2018)

MSEScienceMeeting,Tucson,2019TheAGNspectralvariability 7

2.5m,SDSS 8.2m,SUBARU 4m,AAO

Du+2018

Grier+2017

Long-termRMcampaign(long=decades)• TypicalSeyfert1s:

NGC5548– 9+years(Shapovalova+2004,Ilić2007,Popović+2008,Bon+2016)NGC4151– 11+ years(Shapovalova+2008,2010a,Ilić+2010,Bon+2012)NGC7469– 20years(Shapovalova+2017)NGC3516– 22years(Shapovalova+2019)

• NarrowLineSeyfert1:Ark564– 11years(Shapovalova+2011,Shapovalova+2012)

• DoublePeakedLineAGNs(DPLs):3C390.3– 13years(Shapovalova+2001,2010b,Popović+2011,Jovanović+2010,Kovačević+2014);Arp102B– 12years(Shapovalova+13,Popović+14,Kovačević+14,Ilić+15,Rakić+17)

• Quasar,abinaryblackholecandidate:E1821+643– 25years(Shapovalova+2016,Kovačević+2017,Kovačević+2018)

8MSEScienceMeeting,Tucson,2019TheAGNspectralvariability

6m+1m,SAO

2.1m,OANSPM

2.1m,GHO

3.5m+2.2m,CAO

Ilić+2017

Huntforparsec-scaleSMBHbinaries• time-domainphotometryofAGN

(seee.g.withCRTS,Graham+2017)à famouscase:binarySMBHcandidate

PG1302-102(Graham+2015)• novelhybridmethodtosearchforperiodic

oscillatorybehaviorà appliescontinuouswavelettransformandcorrelation

coefficientsonGaussian-processedlightcurves(seeKovačević+2018,2019fordetails)

• NGC4151:thefirstspectroscopicallyresolvedsub-parsecorbitofSMBH(Bon+2012)à orbitalperiod15.7yr, masses~107 Msun

MSEScienceMeeting,Tucson,2019TheAGNspectralvariability 9

8 Kovacevic et al.

Figure 3. The best fit with a nonstationary GP mean as a solid line with 95% of Confidence Interval

between dashed lines. The observed photometric obseravtions are given as vertical error bars.

is known a priori. A specific test for that noise colour must be applied, which we describe in more

details and apply on both observed and GP modeled light curve in the following subsection.

3.1. Noise test for the light curve

Red noise process can be interpreted as an AR(1), autoregressive process of the first order, with

positive correlation at unit lag. The pink noise can be modeled by means of the differencing parameter

d(= 0.5) of the Box - Jenkins autoregressive integrated moving average (ARIMA) strategy, taking

on it continuous values (Box & Jenkins 1970). Autoregressive fractionally integrated moving average

(ARFIMA) modelling improves the BoxJenkins approach by implementing the differencing parameter

d to have non-integer values. This allows ARFIMA to fit any long range memory in time series

remarkably (Brockwell & Davis 2002). However, these processes are stationary. If the PG1302-102

light curve is found to be non stationary, then it is different from mentioned noise processes. If the

light curve is stationary further statistical procedures must be applied to verify that the light curve

Kovačević+2019

Bon+2012

Changing-look(CL)AGN• extremevariability:appearanceordisappearance

ofBELswithinafewyears• whatisthecause?

– variableaccretionrate– variableobscuration– tidaldisruptionevent

• LAMOSThasfound21newCLAGN(Yangetal.2018)

• perfectcasestostudytheconnectionbetweenAGNanditshostgalaxy

• importanttounderstandAGNevolution

MSEScienceMeeting,Tucson,2019TheAGNspectralvariability 10

NGC3516(Shapovalova+2019)

Yang+2018

MSEwillmakeadifference• reverberationmappingcampaignof5000quasarsuptoz~3

à robustestimatesoftimelagsà accurateSMBHmassforthelargestsampleofquasarstodate(seeViraja’s talk)

• identifynewchanginglookAGN-- synergywithothermissions(e.g.LSST,Pan-STARRS)

• high-resolutionhigh-performancespectroscopyà velocityresolvedreverberationmappingà mappingofthecentralregionsà resolvethebinarySMBH

MSEScienceMeeting,Tucson,2019TheAGNspectralvariability 11

Popović 2012

11.5m,MSE

Summary

12MSEScienceMeeting,Tucson,2019TheAGNspectralvariability

• spectralmonitoringstillcrucialfortheAGNandemissionlineregioninvestigations

• theopticalspectralvariability canbeusedfor:à reverberation,i.e.measuringtheunseenà theradiusoftheBLR

à themassoftheSMBH

à constraintheradius-luminosityrelation

à discovermorechanging-lookAGN

à huntforSMBHbinaries

à constrainthephysicsandgeometryoftheBLR