PKS 1502+106: a new and distant gamma-ray blazar in outburst discovered by the Fermi Large Area Telescope

1.1 MB·English

by  The Fermi LAT Collaboration

DRAFTVERSIONJANUARY21,2010
PreprinttypesetusingLATEXstyleemulateapjv.04/20/08
PKS1502+106:ANEWANDDISTANTGAMMA-RAYBLAZARINOUTBURST
DISCOVEREDBYTHEFERMILARGEAREATELESCOPE
A.A.ABDO2,3,M.ACKERMANN4,M.AJELLO4,W.B.ATWOOD5,M.AXELSSON6,7,L.BALDINI8,J.BALLET9,G.BARBIELLINI10,11,
D.BASTIERI12,13,B.M.BAUGHMAN14,K.BECHTOL4,R.BELLAZZINI8,B.BERENJI4,E.D.BLOOM4,G.BOGAERT15,
E.BONAMENTE16,17,A.W.BORGLAND4,J.BREGEON8,A.BREZ8,M.BRIGIDA18,19,P.BRUEL15,T.H.BURNETT20,
G.A.CALIANDRO21,R.A.CAMERON4,P.A.CARAVEO22,J.M.CASANDJIAN9,E.CAVAZZUTI23,C.CECCHI16,17,O¨.C¸ELIK24,25,26,
A.CHEKHTMAN2,27,C.C.CHEUNG2,3,J.CHIANG4,S.CIPRINI17,1,R.CLAUS4,J.COHEN-TANUGI28,J.CONRAD29,7,30,S.CUTINI23,
0 C.D.DERMER2,A.DEANGELIS31,F.DEPALMA18,19,S.W.DIGEL4,E.DOCOUTOESILVA4,P.S.DRELL4,R.DUBOIS4,
1 D.DUMORA32,33,C.FARNIER28,C.FAVUZZI18,19,S.J.FEGAN15,E.C.FERRARA24,W.B.FOCKE4,M.FRAILIS31,L.FUHRMANN34,
0 Y.FUKAZAWA35,S.FUNK4,P.FUSCO18,19,F.GARGANO19,D.GASPARRINI23,N.GEHRELS24,36,37,S.GERMANI16,17,B.GIEBELS15,
2 N.GIGLIETTO18,19,F.GIORDANO18,19,M.GIROLETTI38,T.GLANZMAN4,G.GODFREY4,I.A.GRENIER9,M.-H.GRONDIN32,33,
J.E.GROVE2,L.GUILLEMOT34,S.GUIRIEC39,Y.HANABATA35,A.K.HARDING24,M.HAYASHIDA4,E.HAYS24,R.E.HUGHES14,
n G.JO´HANNESSON4,A.S.JOHNSON4,R.P.JOHNSON5,W.N.JOHNSON2,M.KADLER40,25,41,42,T.KAMAE4,H.KATAGIRI35,
a J.KATAOKA43,M.KERR20,J.KNO¨DLSEDER44,M.L.KOCIAN4,F.KUEHN14,M.KUSS8,J.LANDE4,L.LATRONICO8,
J M.LEMOINE-GOUMARD32,33,F.LONGO10,11,F.LOPARCO18,19,B.LOTT32,33,M.N.LOVELLETTE2,P.LUBRANO16,17,
1 G.M.MADEJSKI4,A.MAKEEV2,27,M.MARELLI22,E.MASSARO45,W.MAX-MOERBECK46,M.N.MAZZIOTTA19,
2 W.MCCONVILLE24,37,J.E.MCENERY24,37,C.MEURER29,7,P.F.MICHELSON4,W.MITTHUMSIRI4,T.MIZUNO35,
A.A.MOISEEV25,37,C.MONTE18,19,M.E.MONZANI4,A.MORSELLI47,I.V.MOSKALENKO4,S.MURGIA4,P.L.NOLAN4,
] J.P.NORRIS48,E.NUSS28,T.OHSUGI35,N.OMODEI8,E.ORLANDO49,J.F.ORMES48,M.OZAKI50,D.PANEQUE4,J.H.PANETTA4,
E D.PARENT32,33,V.PAVLIDOU46,T.J.PEARSON46,V.PELASSA28,M.PEPE16,17,M.PESCE-ROLLINS8,F.PIRON28,T.A.PORTER5,
H S.RAINO`18,19,R.RANDO12,13,M.RAZZANO8,S.RAZZAQUE2,3,A.READHEAD46,A.REIMER51,4,O.REIMER51,4,T.REPOSEUR32,33,
J.L.RICHARDS46,S.RITZ5,5,L.S.ROCHESTER4,A.Y.RODRIGUEZ21,R.W.ROMANI4,M.ROTH20,F.RYDE52,7,
h. H.F.-W.SADROZINSKI5,D.SANCHEZ15,A.SANDER14,P.M.SAZPARKINSON5,J.D.SCARGLE53,C.SGRO`8,M.S.SHAW4,
p E.J.SISKIND54,D.A.SMITH32,33,P.D.SMITH14,G.SPANDRE8,P.SPINELLI18,19,M.STEVENSON46,M.S.STRICKMAN2,
- D.J.SUSON55,H.TAJIMA4,H.TAKAHASHI35,T.TANAKA4,J.B.THAYER4,J.G.THAYER4,D.J.THOMPSON24,L.TIBALDO12,13,9,
o O.TIBOLLA56,D.F.TORRES57,21,G.TOSTI16,17,A.TRAMACERE4,58,P.UBERTINI59,Y.UCHIYAMA4,T.L.USHER4,
r V.VASILEIOU25,26,N.VILCHEZ44,V.VITALE47,60,A.P.WAITE4,P.WANG4,B.L.WINER14,K.S.WOOD2,H.YASUDA35,
st T.YLINEN52,61,7,J.A.ZENSUS34,M.ZIEGLER5,
a (THEFERMILATCOLLABORATION),
[ AND
E.ANGELAKIS34,T.HOVATTA62,E.HOVERSTEN36,Y.IKEJIRI35,K.S.KAWABATA63,Y.Y.KOVALEV64,34,YU.A.KOVALEV64,
2 T.P.KRICHBAUM34,M.L.LISTER65,A.LA¨HTEENMA¨KI62,N.MARCHILI34,P.OGLE46,C.PAGANI36,A.B.PUSHKAREV66,34,67,
v K.SAKIMOTO35,M.SASADA35,M.TORNIKOSKI62,M.UEMURA63,M.YAMANAKA35,T.YAMASHITA63
9
DraftversionJanuary21,2010
2
0 ABSTRACT
4
The Large Area Telescope (LAT) on board the Fermi Gamma-ray Space Telescope discovereda rapid (
.
2 5 days duration),high-energy(E > 100MeV) gamma-rayoutburstfroma source identified with the blaza∼r
1 PKS 1502+106(OR 103, S3 1502+10, z=1.839) starting on August 05, 2008 ( 23 UTC, MJD 54683.95),
9 andfollowedbybrightandvariablefluxoverthenextfewmonths.Resultsonthe∼gamma-raylocalizationand
0 identification,aswellasspectralandtemporalbehaviorduringthefirstmonthsoftheFermiall-skysurveyare
:
v reportedhereinconjunctionwithamulti-wavebandcharacterizationasaresultofoneofthefirstFermimulti-
i frequencycampaigns.ThecampaignincludedaSwiftToO(followedupby16-dayobservationsonAugust07-
X
22,MJD54685-54700),VLBA(withintheMOJAVEprogram),OwensValley(OVRO)40m,Effelsberg-100m,
r Metsa¨hovi-14m,RATAN-600andKanata-Hiroshimaradio/opticalobservations. Resultsfromtheanalysisof
a
archival observations by INTEGRAL, XMM-Newton and Spitzer space telescopes are reported for a more
completepictureofthisnewgamma-rayblazar. PKS1502+106isasub-GeVpeaked,powerfulflatspectrum
radioquasar(luminosityatE >100MeV,L ,isabout1.1 1049ergs−1,andblackholemasslikelycloseto
γ
×
109 M⊙),exhibitingmarkedgamma-raybolometricdominance,inparticularduringtheasymmetricoutburst
(L /L 100, and 5-day averaged flux F = 2.91 1.4 10−6 ph cm−2 s−1), which was
γ opt E>100MeV
∼ ± ×
characterized by a factor greater than 3 of flux increase in less than 12 hours. The outburst was observed
simultaneouslyfromopticaltoX-raybands(F0.3−10keV =2.18+−00..1152×10−12ergcm−2s−1,andhardphoton
index 1.5, similar to past values) with a flux increase of less than one order of magnitudewith respectto
∼
pastobservations,andwaslikelycontrolledbyComptonizationofexternal-jetphotonsproducedinthebroad
line region (BLR) in the gamma-ray band. No evidence of a possible blue bump signature was observed
in the optical-UV continuum spectrum, while some hints for a possible 4-day time-lag with respect to the
gamma-rayflarewerefound. Nonetheless,thepropertiesofPKS1502+106andthestrictoptical/UV,X-and
gamma-raycrosscorrelationssuggestthecontributionofthesynchrotronselfCompton(SSC),in-jet,process
shoulddominatefromradiotoX-rays.Thismechanismmayalsoberesponsiblefortheconsistentgamma-ray
variabilityobservedbytheLATonlongertimescales, afterthe ignitionofactivityattheseenergiesprovided
bythe BLR-dissipatedoutburst. Modulationsandsubsequentminor, rapidflare eventswere detected,with a
2 Abdoetal.
generalfluctuationmodebetweenpink-noiseandarandom-walk. Theaveragedgamma-rayspectrumshowed
adeviationfromasimplepower-law,andcanbedescribedbyalog-parabolacurvedmodelpeakingaround0.4-
0.5GeV.ThemaximumenergyofphotonsdetectedfromthesourceinthefirstfourmonthsofLATobservations
was15.8GeV,withnosignificantconsequencesonextragalacticbackgroundlightpredictions.Apossibleradio
counterpartof the gamma-rayoutburstcan be assumed only if a delay of more than 3 months is considered
onthebasisofopacityeffectsatcmandlongerwavelengths. Therotationoftheelectricvectorpositionangle
observedbyVLBAfrom2007to2008couldrepresentaslowfieldorderingandalignmentwithrespecttothe
jet axis, likely a precursorfeature of the ejection of a superluminalradio knotand the high-energyoutburst.
ThisobservingcampaignprovidesmoreinsightintotheconnectionbetweenMeV-GeVflaresandthemoving,
polarizedstructuresobservedbytheVLBI.
Subjectheadings:gamma-rays:observations–quasars:individual:PKS1502+106–quasars:general–galax-
ies: active–galaxies:jets–X-rays:galaxies
1Correspondingauthor:S.Ciprini,[email protected]. Fisica Nucleare, Sezione di Trieste, Gruppo Collegato di Udine, I-33100
2Space Science Division, Naval Research Laboratory, Washington, DC Udine,Italy
20375,USA 32Universite´deBordeaux,Centred’E´tudesNucle´airesBordeauxGradig-
3National Research Council Research Associate, National Academy of
nan,UMR5797,Gradignan,33175,France
Sciences,Washington,DC20001,USA
33CNRS/IN2P3,Centred’E´tudesNucle´airesBordeauxGradignan,UMR
4W.W.HansenExperimentalPhysicsLaboratory,KavliInstituteforParti-
5797,Gradignan,33175,France
cleAstrophysicsandCosmology,DepartmentofPhysicsandSLACNational
34Max-Planck-Institut fu¨r Radioastronomie, Auf dem Hu¨gel 69, 53121
AcceleratorLaboratory,StanfordUniversity,Stanford,CA94305,USA
Bonn,Germany
5SantaCruzInstituteforParticlePhysics,DepartmentofPhysicsandDe-
35Department of Physical Sciences, Hiroshima University, Higashi-
partmentofAstronomyandAstrophysics,UniversityofCaliforniaatSanta
Hiroshima,Hiroshima739-8526,Japan
Cruz,SantaCruz,CA95064,USA
6DepartmentofAstronomy,StockholmUniversity,SE-10691Stockholm, 36DepartmentofAstronomyandAstrophysics, Pennsylvania StateUni-
Sweden versity,UniversityPark,PA16802,USA
7TheOskarKleinCentreforCosmoparticlePhysics,AlbaNova,SE-106 37Department of Physics and Department of Astronomy, University of
91Stockholm,Sweden Maryland,CollegePark,MD20742,USA
8IstitutoNazionalediFisicaNucleare,SezionediPisa,I-56127Pisa,Italy 38INAFIstitutodiRadioastronomia,40129Bologna,Italy
9Laboratoire AIM,CEA-IRFU/CNRS/Universite´ Paris Diderot, Service 39CenterforSpacePlasmaandAeronomicResearch(CSPAR),University
d’Astrophysique,CEASaclay,91191GifsurYvette,France ofAlabamainHuntsville,Huntsville,AL35899,USA
10IstitutoNazionale diFisicaNucleare, SezionediTrieste, I-34127Tri- 40Dr. Remeis-Sternwarte Bamberg, Sternwartstrasse 7, D-96049 Bam-
este,Italy berg,Germany
11DipartimentodiFisica,Universita`diTrieste,I-34127Trieste,Italy 41ErlangenCentreforAstroparticlePhysics,D-91058Erlangen,Germany
12Istituto Nazionale di Fisica Nucleare, Sezione di Padova, I-35131 42Universities Space Research Association (USRA), Columbia, MD
Padova,Italy 21044,USA
13Dipartimento di Fisica “G. Galilei”, Universita` di Padova, I-35131 43Waseda University, 1-104 Totsukamachi, Shinjuku-ku, Tokyo, 169-
Padova,Italy 8050,Japan
14Department of Physics, Center for Cosmology and Astro-Particle 44Centred’E´tudeSpatialedesRayonnements,CNRS/UPS,BP44346,F-
Physics,TheOhioStateUniversity,Columbus,OH43210,USA 30128ToulouseCedex4,France
15Laboratoire Leprince-Ringuet, E´cole polytechnique, CNRS/IN2P3, 45Universita`diRoma“LaSapienza”,I-00185Roma,Italy
Palaiseau,France 46CahillCenterforAstronomyandAstrophysics, CaliforniaInstitute of
16IstitutoNazionalediFisicaNucleare,SezionediPerugia,I-06123Peru- Technology,Pasadena,CA91125,USA
gia,Italy 47IstitutoNazionalediFisicaNucleare,SezionediRoma“TorVergata”,
17DipartimentodiFisica,Universita`degliStudidiPerugia,I-06123Peru- I-00133Roma,Italy
gia,Italy 48DepartmentofPhysicsandAstronomy,UniversityofDenver,Denver,
18DipartimentodiFisica“M.Merlin”dell’Universita` edelPolitecnicodi CO80208,USA
Bari,I-70126Bari,Italy 49Max-PlanckInstitutfu¨rextraterrestrischePhysik,85748Garching,Ger-
19IstitutoNazionalediFisicaNucleare,SezionediBari,70126Bari,Italy many
20DepartmentofPhysics,UniversityofWashington,Seattle,WA98195- 50InstituteofSpaceandAstronauticalScience,JAXA,3-1-1Yoshinodai,
Sagamihara,Kanagawa229-8510,Japan
1560,USA
21Institut de Ciencies de l’Espai (IEEC-CSIC), Campus UAB, 08193 51Institut fu¨r Astro- und Teilchenphysik and Institut fu¨r Theoretische
Barcelona,Spain Physik,Leopold-Franzens-Universita¨tInnsbruck,A-6020Innsbruck,Austria
22INAF-IstitutodiAstrofisicaSpazialeeFisicaCosmica,I-20133Milano, 52Department of Physics, Royal Institute of Technology (KTH), Al-
Italy baNova,SE-10691Stockholm,Sweden
23AgenziaSpazialeItaliana(ASI)ScienceDataCenter,I-00044Frascati 53SpaceSciencesDivision,NASAAmesResearchCenter,MoffettField,
CA94035-1000,USA
(Roma),Italy
54NYCBReal-TimeComputingInc.,Lattingtown,NY11560-1025,USA
24NASAGoddardSpaceFlightCenter,Greenbelt,MD20771,USA
55Department of Chemistry and Physics, Purdue University Calumet,
25CenterforResearchandExplorationinSpaceScienceandTechnology
Hammond,IN46323-2094,USA
(CRESST)andNASAGoddardSpaceFlightCenter,Greenbelt,MD20771, 56Max-Planck-Institutfu¨rKernphysik,D-69029Heidelberg,Germany
USA
26DepartmentofPhysicsandCenterforSpaceSciencesandTechnology, 57Institucio´CatalanadeRecercaiEstudisAvanc¸ats(ICREA),Barcelona,
Spain
UniversityofMarylandBaltimoreCounty,Baltimore,MD21250,USA
27GeorgeMasonUniversity,Fairfax,VA22030,USA 58Consorzio Interuniversitario per la Fisica Spaziale (CIFS), I-10133
Torino,Italy
28LaboratoiredePhysiqueThe´oriqueetAstroparticules,Universite´Mont-
59INAF-IstitutodiAstrofisicaSpazialeeFisicaCosmica,I-00133Roma,
pellier2,CNRS/IN2P3,Montpellier,France
Italy
29Department ofPhysics, Stockholm University, AlbaNova, SE-10691
60Dipartimento di Fisica, Universita` di Roma “Tor Vergata”, I-00133
Stockholm,Sweden
30Royal Swedish Academy of Sciences Research Fellow, funded by a Roma,Italy
grantfromtheK.A.WallenbergFoundation 61School of Pure and Applied Natural Sciences, University of Kalmar,
31Dipartimento diFisica, Universita` diUdine and Istituto Nazionale di SE-39182Kalmar,Sweden
PKS1502+106:anewanddistantgamma-rayblazarinoutburstdiscoveredbytheFermiLAT 3
1. INTRODUCTION This radio blazar was identified in the optical band by
Blake(1970)withapositionrefinementbyArgue&Sullivan
The Large Area Telescope (LAT), on board the Fermi
(1980), while an initial spectroscopic inspection was per-
Gamma-raySpaceTelescope(formerlyGLAST; Ritz2007),
formedbyBurbidge&Strittmatter(1972). Variations> 2.5
wassuccessfullylaunchedbyNASAon2008,June11,from
magwereobservedinitsopticalfluxhistory(Palomar-Quest
Cape Canaveral, Florida, on a Delta II Heavy launch vehi-
andCatalinaSkySurveys,ATel#1661),togetherwithavari-
cle. While still in the commissioning and checkout phase,
able and relatively high degree (up to 20%) of linear polar-
itdiscoveredandmonitoredbright,flaringgamma-rayemis-
ization, pointing out a dominant synchrotron emission with
sionabove100MeVfromasourceidentifiedwiththeblazar
noobserveddilutionbythermalcomponents. Theredshiftof
PKS 1502+106 (historically also known as OR 103 and S3
PKS1502+106,asconfirmedbythegoodS/Nspectrumofthe
1502+10).Thelargefieldofview,effectiveareaandsensitiv-
SloanDigitalSkySurvey(z =1.8385 0.0024athighcon-
ity andthe nominalsurveyobservationalmodemake Fermi-
±
fidence),isinagreementwiththevaluez = 1.839estimated
LAT an unprecedented all-sky monitor of γ-ray flares and
previouslyby Smithetal. (1977). A less remotevalue (z =
sourcevariability(see,e.g. McEnery2006;Thompson 2006;
0.56) is reported in other works (Burbidge&Strittmatter
Lottetal.2007;Atwoodetal.2009).
1972; Wrightetal. 1979; Wilkesetal. 1983), although the
Atthe beginningofAugust2008,PKS 1502+106was the
possible multiple MgII absorption system (pointed out by a
second brightest extragalactic source in the γ-ray sky, ex-
hibiting a sudden high-energy outburst announced in ATel featureshortwardofthe4388A˚ emissionline)wouldbevery
#1650. This outburst successfully triggered the first (un- unusualforalowredshiftobject.
planned) Fermi multi-frequency campaign. Major renewed Serendipitous X-ray data of PKS 1502+106 are available
gamma-rayactivity observedby Fermi in January 2009 was becausethesourceliesabout7′NEofthebrightSeyferttype-
announcedviaATel#1905. 1 galaxy Mkn 841, although only one multifrequency work
PKS1502+106isaluminous,quasar-like(opticallybroad- dedicatedtothisblazar(Georgeetal.1994)hasappearedpre-
lineandflatradiospectrum)AGNdiscoveredduringthe178 viously. Early X-ray observations (ROSAT, ASCA) showed
MHzpencilbeamsurveyfromtheMullardRadioAstronomy low-amplitudevariationson short timescales (factor > 2 on
Observatory,Cambridge,UK,(appearinginalistnotincluded timescales of a year), a flat 0.1-10 keV photon index ΓX
in the 4C catalog; Crowther&Clarke 1966; Williamsetal. between 1.4 and 1.9, and an intrinsic X-ray luminosity of
1967), and was re-observed and characterized as an extra- L2−10keV = 1.2 1046 ergs−1, anda 2-10keV flux inthe
galacticsourcebyboththeAustralianNationalRadioAstron- range4.9 6.54× 10−13 ergcm−2 s−1(Georgeetal.1994;
− ×
omyObservatoryofParkes,NSW,Australia,(Dayetal.1966, Akiyamaetal. 2003; Watanabeetal. 2004). PKS 1502+106
id.: PKS 1502+106), and the Ohio State University (“Big was speculated to be a possible γ-ray source before the
Ear”) Radio Observatory, Delaware, OH, USA, (Fitchetal. LAT discovery because of the superluminal motions of jet
1969, id.: OR 103). The source exhibitedsubstantial radio components(upto 187 15µas/yearLister&Homan2005;
±
fluxvariations(factor> 2),ahighdegreeoflinearpolariza- Anetal. 2004),andthemultiwavebandspectralindexesαrx
tion, a core-dominated, one-sided and curved radio jet with and αox (consistent with other FSRQs detected by EGRET,
a variable, a complex morphologyat VLBI scales (Anetal. Georgeetal. 1994). Only modestintrinsic X-ray absorption
2004; Listeretal. 2009a), and a compact large scale struc- wassuggestedbythiswork,andtheopticalandnear-IRred-
ture. 11VLBAobservationsat15.4GHzperformedbetween dening claimed in Watanabeetal. (2004) is probably due to
Aug. 1997 and Aug. 2007 showed a FWHM major beam thesynchrotronjetdominanceattheselowfrequenciesrather
axis in the range 1.02-1.57 mas, a minor axis beam axis of thanbyabsorptionfrominnernuclearlight.
0.5 mas, and a total flux density in the range 0.88-1.93 Jy A relation involving the misalignment between the pc-
(Listeretal.2009a), andapparentjetspeedof(14.8 1.2)c and kpc-scale radio structure (position angle) and the γ-
(Listeretal.2009b). The22 and37GHz fluxhistory±shows ray emission was postulated as well (Cooperetal. 2007).
severallong-termflares(>1Jyvariations,i.e. 60%ofthe However, only a cumulative 2σ upper limit by EGRET
totalfluxrangespan,ontypicaltimescalesofay∼ear,andpeak of 7 10−8 ph cm−2 s−1 was reported (Phase/Cycle I,
×
fluxeswell above2 Jy), with at least five flares and an aver- combined Viewing Periods: 24.0 to 25.0, i.e. April 02-23,
agetrendthatwasslightlyincreasingfrom1988tomid-2004 1992; Fichteletal.1994),andthesourcewaslikewiseunde-
(Tera¨srantaetal.2005). WMAPfluxesatsimilarfrequencies tected in the following EGRET cycles (Hartmanetal. 1999;
(K, Ka, Q bands) are in agreement with these flux ranges Casandjian&Grenier2008).
(Lo´pez-Caniegoetal. 2007). The Doppler factor estimated In the following we use a ΛCDM (concordance) cos-
fromtheobserved37GHzvariabilityandbrightnesstemper- mology with values given within 1σ of the WMAP results
ature (Hovattaetal. 2009) agrees with the jet speed (14.6c) (Komatsuetal. 2009), namely h = 0.71, Ωm = 0.27 and
θcited=ab4o.v7e◦,.a DopplerfactorDvar = 12and viewingangle ΩMΛpc=−10..73,andaHubbleconstantvalueH0 =100hkms−1
var
InSection2,firstresultsontheγ-rayidentification,theob-
served MeV-GeV outburst and the subsequent four months
62Metsa¨hoviRadioObservatory,HelsinkiUniversityofTechnologyTKK,
of monitoring by the Fermi-LAT are described. In Sec-
FIN-02540Kylmala,Finland
63Hiroshima Astrophysical Science Center, Hiroshima University, tion3,multifrequencyresultsobtainedthroughsimultaneous
Higashi-Hiroshima,Hiroshima739-8526,Japan optical-UV-X-ray observations by Swift (thanks to a 16-day
64AstroSpaceCenteroftheLebedevPhysicalInstitute,117810Moscow, longmonitoringfollowingatriggeredTargetofOpportunity,
Russia ToO), and by radio-opticalobservatories(the 40m dish tele-
65DepartmentofPhysics,PurdueUniversity, WestLafayette, IN47907,
scopeoftheOwensValleyRadioObservatory,theEffelsberg
USA
66CrimeanAstrophysicalObservatory,98409Nauchny,Crimea,Ukraine 100mdishradiotelescope,theringradiotelescopeRATAN-
67PulkovoObservatory,196140St.Petersburg,Russia 600,theVLBAwithintheMOJAVEprogram,andtheKanata
*Correspondence:[email protected]
4 Abdoetal.
therein),andanisotropiccomponentincludingtheextragalac-
ticdiffuseemissionandtheresidualbackgroundfromcosmic
rays.
TheFermi-LATdataofPKS1502+106presentedherewere
obtainedduringthefirstfourmonthsoftheLATsurvey(Aug.-
Dec. 2008). In this period PKS 1502+106 was one of the
most persistently bright, variable sources in the high-energy
skyandalmostcertainlythesourcewiththehighestluminos-
ity. The background contribution within a few degrees was
onlyasmallfractionofthesourcecountrate,withnonearby
sourceconfusion. The time intervalwas sufficientfora fine
determinationofthe averagespectrum,fora first lookatthe
FIG.1.— Leftpanel: LATcountmapcumulatedonanine-month(Aug. mid-timescalevariabilityanddetectionofposteriorflares,for
2008 - Apr. 2009) baseline, weighted and smoothed by the point spread a refined localization, and a first cross comparison with the
function (PSF) such that higher energy photons are mapped to higher in-
tensities. Themapisinarbitraryunitsintheenergyrange0.1-100GeVand othermultifrequencymonitoringdata. Thefirst γ-raydetec-
ina2.5◦×2.5◦regioncenteredonPKS1502+106. Thequalitativecircle tion of PKS 1502+106 by the LAT occurred in July 2008,
sizesofthePSFat200MeV, 2GeVand20GeV areoutlined forreference. when it was confirmed by the high-levelAutomatic Science
Rightpanel: LATsourcelocalization with95%and68%uncertainty radii
Processing pipeline monitoring (ASP; Chiangetal. 2006,
(redcircles)superimposedonanarcmin-scale(R-band)opticalimageshow-
ingalsotheX-raycounterparterrorboxbytheSwift-XRTobservationsand 2007), based on a wavelet-based (pgwave) quick-look de-
theradioposition andintensity contours byVLAofPKS1502+106. The tectiontool(e.g. Damianietal. 1997; Marcuccietal. 2004;
bestLATsourceposition, calculated onthesamenine-month period, with Ciprinietal.2007a)andamaximumlikelihoodanalysis,and
thepointliketoolisRA:226.10179◦,Dec: +10.4927◦,∆=0.0027◦,
with68%and95%LATerrorcirclesof0.0077◦and0.0124◦respectively. bytheLATSourceCatalogalgorithm(mr filter)basedon
waveletanalysisinthePoissonregime(Starck&Pierre1998)
telescopeoftheHigashi-HiroshimaObservatory)aresumma-
and the peak-finding tool sExtractor (Bertin&Arnouts
rized. In addition, past and unpublishedobservationsby the
1996). The rapid and markedly time-asymmetric γ-ray out-
XMM-NewtonandSpitzerspacetelescopesareanalyzedand
burst announcedin ATel #1650and triggeringan unplanned
presentedin Section 4 for a more complete picture. Finally,
Target of Opportunity (ToO) multifrequency campaign was
inSection5and6,discussionandconcludingremarksarere-
seenfromAug.05(about23UTC)untilaboutAug.11,2008
ported.
( 5 days duration), with a fast rise, slower decay, and an
∼
approximatelytwo-daysustainedpeakflux.
2. GAMMA-RAYOBSERVATIONSANDRESULTSBYFERMI-LAT
Some caveats related to the preflight instrument response
2.1. LATobservations
functions(P6 V1),whichoverestimatedtheacceptanceatlow
The LAT instrument is a pair tracker-converter telescope energies,arebrieflydescribedinAbdoetal.(2009b).
comprising a modular array of 16 towers—each with a
tracker based on silicon micro-strip detector technology—
2.2. Gamma-raysourcelocalization,associationand
anda calorimeterbasedona hodoscopicarrayof96CsI(Tl)
identification
crystals, surrounded by an Anti-Coincidence Detector capa-
ble of measuring the directions and energies of cosmic γ- The LAT PSF and sensitivity provides an unprecedented
ray photonswith energiesfrom 20 MeV to > 300 GeV (for angular resolution in gamma-rays (68% containment ra-
details, see, e.g. Bellazzinietal. 2002; Michelson 2007; dius better than 1◦ at 1 GeV, Atwoodetal. 2007;
∼
Atwoodetal.2007,2009;Abdoetal.2009h). Cecchietal.2007;Atwoodetal.2009;Abdoetal.2009c,h),
ThereductionandanalysisofLATdatawasperformedus- making the association and identification processes less dif-
ingtheScienceToolsv.9.8,basedinparticularonaunbinned ficult than in previous experiments. In the case of this very
maximum-likelihoodestimator of the spectral modelparam- bright γ-ray source, we obtained—beyond the good spatial
eters(gtliketool). Eventswere selected using the Instru- “association”—a firm “identification” with PKS 1502+106.
mentResponse Functions(IRFs) P6 V1 DIFFUSE. Thisse- The3-monthbrightsourcelistresults(Abdoetal.2009c,id:
lectionprovidesthecleanestsetofevents(intermsofdirec- 0FGL J1504.4+1030), provided a good initial localization:
tion, energy reconstruction and backgroundrejection) at the RA:226.12◦,Dec: +10.51◦;r = 0.05◦ and√TS = 88.2,
95
costof reducedeffectiveareaatlow energies,andtakesinto (r beingthe radiusof95%confidenceregion,TSthelike-
95
account the differences between front- and back-converting lihood test statistic from the 200 MeV to 100 GeV analy-
events. To minimize contamination by Earth albedo γ-ray sis). Application of the pointlike tool (Burnett 2007;
eventsthathavereconstructeddirectionswithangleswithre- Abdoetal.2009c)onamuchlonger(nine-month,Aug.2008
specttothelocalzenith>105◦ havebeenexcluded. Forthis - Apr. 2009)LAT dataset with very high statistics, provided
objectwithhighGalacticlatitude,eventsareextractedwithin anexcellentlocalization(outlinedinFig.1):RA:226.10179◦,
a 10◦ acceptance cone centered at the PKS 1502+106radio Dec:+10.4927◦,∆=0.0027◦,with68%and95%LATerror
position.Thiscone,substantiallylargerthanthe68%contain- circlesof0.0077◦and0.0124◦ respectively(statisticalonly).
mentangleofthePSF atthelowestenergies,providessuffi- Studiesofbrightsourcelocalizationsindicateasystematicun-
cienteventstoaccuratelyconstrainthediffuseemissioncom- certaintyinthelocalizationof< 30”,thatcanbetakenasan
ponents. The gtlike model includes the PKS 1502+106 estimateofthesystematicswiththistool(moredetailsonthe
pointsource component, two other pointsources from the 3 productionofpointlikedensitymapsandlocalizationare
month catalog (both faint and low-confidence sources with describedin Camiloetal. 2009). Therelevantimprovement
TS 0.9% of the TS value of PKS 1502+106for the same inthelocalizationcarriedoutonanine-monthbaselineisdue
≃
period),acomponentfortheGalacticdiffuseemission(GAL- to the high variability that occurred with this source. These
PROPcode,see,e.g. Moskalenkoetal.2003, andreferences localizationvaluesareinagreementwiththeVLBIradioand
PKS1502+106:anewanddistantgamma-rayblazarinoutburstdiscoveredbytheFermiLAT 5
FIG.2.—Mainpanel:Likelihoodflux(E>100MeV)lightcurveobtainedindailybinsfromAug.02toDec.15,2008.Theoutburststateandthesubsequent
post-flare(alowerandintermediatelevelbrightness,farfromthefainteststateobserved)periodwithsimultaneousmonitoringbySwiftarerepresentedbythe
twohorizontallines.Leftinsetplot:Azoomonthecorrespondingfluxlightcurvearoundtheoutburstperiodobtainedusingfiner,12-hourbins(lowerstatistics).
Rightinsetplot:Thegamma-ray(E>100MeV)photonindexvaluesforthesameperiodusingdailybinsasinthemainpanellightcurve.
optical positions of PKS 1502+106, the VLA contours and
theSwiftXRTerrorbox(Fig. 1). PKS1502+106istheonly
brightVLAradiosource(calibratorsourcelist)locatedwithin
theLAT95%confidencecircle. TheSeyfertgalaxyMkn841
(observedwith a hardX-raycutoff,see Sect.4) ispositioned
welloutsideofthese localizationcircles, asareotherHB89-
catalogquasarsinthisregion.
Beyond the excellent spatial association, the most secure
and distinctive signature for firm identification of this new
gamma-ray source found by Fermi is the observed correla-
tionbetweentheγ-ray,X-rayandoptical-UVvariability(see
sections 3 and 4). This object was also a member of the
pre-launchCGRaBS(Healeyetal.2008,objectid: CGRaBS
J1504+1029)andRoma-BZCAT(Massaroetal.2009,object
id: BZQJ1504+1029)catalogslistingcandidategamma-ray
blazars. Finally, a method based on a “figure of merit” (de-
scribedin Sowards-Emmerdetal. 2003, 2005) forthis LAT FIG.3.— TimeseriesanalysisoftheLATlightcurvepresentedinFig.2:
theperiodogram,firstorderstructurefunction(upperinset)andpowerspec-
sourcepositionprovidesaveryhighlikelihoodofidentifica- trum(lowerinset).Thesefunctionsshowavariabilitywithapowerspectrum
tionwithPKS1502+106. consistentwith1/f1.3fluctuations.Thisindicatesavariabilitymodeplaced
betweenflickeringandshotnoise. Thehorizontaldashedlinerepresentsthe
0.01falsealarmprobabilitythreshold(99%significancethatthesignalde-
2.3. Gamma-raytemporalbehavior
tectionisnotwrong).
The typically bright γ-ray flux and the enduring activity
shown by PKS 1502+106 in γ-rays, allowed a firm detec- ahighflux(F > 2 10−6 phcm−2 s−1)plateau,
E>100MeV
×
tion of the source by the LAT on a daily basis. Fig. 2 or a secondary superposedflare, that extendedfor about2.5
shows the light curve (daily bins, E > 100MeV) extracted days. Duringthis outbursta “harderwhenbrighter”spectral
with the gtlike tool over the first four and half months trendis suggested, despite the photonindexerror dispersion
of LAT all-sky survey. A fast-rising, markedly asymmet- (seethedailyphotonindicesreportedontherightinsetpanel
ric and bright outburst was found, with a factor > 3 of in- ofFig. 2). Thegtlikeperformancesandthe currentIRFs
crease in flux in less than 12 hours. The integrated flux at useddidnotallowtogobelowadailybinningintheextrap-
E > 100MeVaveragedinthe12hbinofthepeakemission olation of the photon index values, and this was possible in
(between Aug. 05 and Aug. 06, 2008, i.e. DOY 218-219) anacceptablewaywithrespecttoamplitudeofthestatistical
wasF =(3.7 0.7) 10−6phcm−2s−1(statis- error only for the high-flux and high photon count statistics
E>100MeV
± ×
ticalonly),asmeasuredwhentheLATinstrumentwasstillin availablearoundtheoutburstepoch.
commissioning and checkoutphase (all-sky nominalmode). Aconsistentlevelofvariability,withacoupleofminorbut
TheemissionfromPKS1502+106thenfadedmoreslowlyin relevant rapid flares, occurred after the major outburst with
the following days, and the entire outburst interval spanned fluctuationsontimescalesofweeks. Renewedactivityandin-
Aug. 05 around 23 UTC until to Aug. 11 around 00 UTC, creased average brightness from the end of Nov. 2008 were
2008 (i.e. DOY, 218.95-224.0, 5 days duration Fig. 2). observed. Tworapidflaresapproachedamaximumpeakflux
Thepeakfluxappearedelevatedf∼orlessthantwodays,rival- of F 2 10−6 ph cm−2 s−1(daily bin estima-
E>100MeV
∼ ×
ingthebrighestapparentfluxofotherextragalacticsourcesat tions),onSept. 06(DOY250),whereasimultaneouscross-
thattime(Section2.1).Thefiner,12-hourbinlightcurve( 8 correlatedopticalflarewasobservedaswell, andonOct.30
∼
Fermiorbits,ensuringstillsimilarexposures)reportedinthe (DOY 304). Visual inspection of the light curve reportedin
right inset panel of Fig. 2, shows the elevated flux held for Fig.2 suggests a period of higher activity beginning in mid
atleast one1.5days, while the slower fadingdecayexhibits Nov.2008(afteraboutDOY320),and,ingeneral,aseriesof
6 Abdoetal.
modulationsoccurringonaboutaone-monthtimescale,with TheLPdescriptionintroducestheadvantageofonlyoneex-
faster fluctuations and rapid flare episodes superposed. The tra parameter (while BPL model adds two parameters) with
powerspectraldensity(PSD)showsapower-lawdependence, respect to the simple power-law model, it allows modeling
P(f) 1/f1.3. Similar time scale dependenceis exhibited of mild spectral curvatures with no abrupt cutoffs, and pro-
∝
by the first order structure function (Fig. 3) and by the au- videsa better phenomenologicalphysicalinterpretation. On
tocorrelationfunction. More detailedvariabilityanalysisfor theotherhandtherecanbestillcaveatswhenusinggtlike
thisandotherblazarsusingalongerdatasetwillbepresented witha brokenpower-lawmodel,inparticularindetermining
inAbdoetal.(2009d). thebreakenergywhenstatistics donotallow a highnumber
of energy bins. It is also plausible that an energy spectrum
2.4. Gamma-rayspectra averagedoveralongperiodoftime,andcontainingdifferent
activitystageswithtimevaryinghardness,mayexhibitanap-
We have analyzed the time-averaged spectra of PKS
parentcurvature. FinallythisdoesnotexcludeBPLmodelif
1502+106 for three intervals: the high-state of the outburst
thespectrumisextractedindifferenttimeintervals.
(DOY, 218.95-224.0, i.e. Aug. 05 - Aug. 11, 2008, about
Theaveragespectrumduringtheoutburststateisconsistent
5 days); the post-flare period characterizedby an intermedi-
atebrightnesslevelandduringwhichsimultaneousSwiftob- withasimplepower-lawmodel,dN/dE EΓ. Theoutburst
∝
servationswereperformed(DOY224.0-235.42,i.e.Aug.11- state shows a rather hard spectrum, suggesting a maximum
Aug. 22, 2008, about 11.4 days); the longer and heteroge- peak in the MeV energy bands (in the νFν representation),
neous period that includes the outburst and the following 4 in agreement with the LP peaks found for the spectra cited
months of Fermi-LAT monitoring(Aug.02 to Dec.07, 2008, above.Theextrapolatedandaveragedfluxesintegratedabove
126 days, where the source displayed different stages of 100MeV,andthespectralfitparametersforallthreeperiods
∼activity and significant variability). Events below 200 MeV areshowninTable1.
wereexcludedfromtheseanalysesbecauseofcalibrationun- 3. SIMULTANEOUSMULTIFREQUENCYOBSERVATIONS
certaintiesatthoseenergies. Anisotropicbackgroundmodel
Becauseofthereasonablyuniformexposureandhighsen-
wasusedasPKS1502+106wasverybrightrelativetoother
sitivity of the LAT, and the broad-band radio-to-gamma-ray
point sources during the period stated above, because is lo-
emission of this kind of AGN, simultaneous multifrequency
catedathighGalacticlatitude,andbecausecheckswithmore
data are very important to the investigation of the physical
complexmodelsprovidednosignificantdifference. Further-
properties of supermassive black holes and relativistic jets,
more,noappreciabledifferenceswereobservedusingdiffer-
beyond the benefit of a firm source identification (Section
entacceptanceconeradiifortheeventextraction.
2.2). Withthisinmind,severalcampaignsonafewselected
The spectra for the post-flare and cumulative 4-month
objects, or ToO list of candidates for flaring sources, were
datasets can be consistent with a log-parabola (LP) model,
preparedpre-launch by the Fermi collaboration (Tosti 2007;
dN/dE E−(α+βlog(E)) (see, e.g. Landauetal. 1986;
∝ Thompson2007). PKS1502+106wasapreviouslyunknown
Inoue&Takahara 1996; Fossatietal. 2000; Massaroetal.
γ-raysource,withnopre-plannedmultifrequencycampaign.
2004; Perlmanetal. 2005). The likelihood ratio test
But following the LAT outburst (reported in ATel #1650), a
(Mattoxetal. 1996) rejects the hypothesisthat the spectrum
ToO campaign was initiated on Aug. 07, 2008. This was
isapowerlaw(nullhypothesis)againsttheonethatthespec-
the first Fermi multifrequency campaign that had not been
trum is curved as a log-parabola model. This model is pre-
plannedpre-launch,andsawtriggersforToOpointingsbyIN-
ferredoverasimplepower-lawmodelatthe11.4sigma sig-
TEGRALandSwift,long-termradiofluxandstructuremoni-
nificancelevel. Brokenpower-law(BPL) fitsshow asimilar
toring,aswellasopticalobservationsbygroundbasedfacili-
improvementover single power-law models, but we find no
ties.
evidencetopreferBPL overthelog-parabolarepresentation.
The fast response ToO pointing by INTEGRAL provided
Forthefulltimeintervalcharacterizedbyveryhighstatistics,
200ksofobservationsduringtheperiodAug.09,01:53UT-
the logarithm of the likelihood increases significantly when
Aug. 11, 15:12 UT, 2008 (revolution 711). However, PKS
allowingβ tovary,andan increaseofthe valueforthe BPL
1502+106 was not detected (preliminarily) by the imager
withrespecttosimplepower-lawmodelofthesameorderof
IBISonboardINTEGRAL.ExtrapolatingtheX-rayfluxob-
the increase for the LP vs. power-law test is not observed.
servedbySwift,thehard-X-rayfluxhadlikelyalreadyfaded
to slightly below the IBIS detection threshold in this epoch
TABLE1 (moredetailswillappearinPianetal.2009).
SUMMARYOFTHEUNBINNEDLIKELIHOODSPECTRALFIT
ABOVE100MEV 3.1. SimultaneousX-rayandUV-opticalobservationsand
resultsbySwift
Interval[MJD(DOY)] Best-fitModelandParameters
The Swift satellite (Gehrelsetal. 2004) performed a ToO
Allobservations Log-parabola
monitoringcampaignofPKS1502+106withdailysnapshots
54682.680(217.680) α=1.94±0.05
54775.580(310.580) β=0.10±0.02 from Aug. 07 to Aug. 22, 2008. This quite long-term ob-
FE>100MeV=6.90±0.34×10−7[phcm−2s−1]servingcampaignbySwift allowedextendeddailysnapshots
Outburst/highstate PowerLaw for about 16 days, using the three instruments onboard: the
54683.955(218.955) Γ=−2.06±0.017 X-ray telescope (XRT) for the 0.2-10 keV energy band, the
54688.985(223.985) FE>100MeV=29.1±1.4×10−7[phcm−2s−1] Ultraviolet/OpticalTelescope(UVOT)formultibandphotom-
Post-flare/lowerstate Log-parabola etry,andtheBurstAlertTelescope(BAT)forthe15-150keV
54689.063(224.063) α=1.87±0.20 hardX-rayband. BATdatawerenotusedbecauseofsource
54775.580(310.580) β=0.18±0.08 confusionproblemswithMkn841whichisaboutafactorof
FE>100MeV=5.32±1.03×10−7[phcm−2s−1]10brighterthanPKS1502+106inthehardX-rayband. The
16 days of observationsby Swift allow for cross-correlation
PKS1502+106:anewanddistantgamma-rayblazarinoutburstdiscoveredbytheFermiLAT 7
2
0
0.
01 Outburst
Counts s keV−1−1 0.×105×10−3−3
2
0−3 Post−flare
1
0−4
1
5×1.5
o
Rati 1
0.5 1 2 5
Energy [keV]
FIG.5.—Swift-XRTcombined0.3-10KeVspectraofPKS1502+106ex-
tractedforthehighstate(MJD:54685-54689)andthesubsequentlowstate
(MJD54690-54701),mappingtheX-raybehaviorsimultaneoustotheLAT
flareandtothepost-flare,relaxingactivityandbrightness.
TABLE2
ANALYSISSUMMARYOFTHESIMULTANEOUSDATAOBTAINEDBYTHE
XRTINSTRUMENTONBOARDSwift.
Obs.id.(date) Best-fitModelandParameters
Allobservations PowerLaw
(MJD54685-54701) ΓX =1.53+−00..0067
texp:52910s χ2r=1.05/80
F0.3−10keV =1.79+−00..0181×10−12ergcm−2s−1
Outburst/highstate PowerLaw
(MJD54685-54689) ΓX =1.54±0.08
texp:27680s χ2r=1.11/52
F0.3−10keV =2.18+−00..1152×10−12ergcm−2s−1
Post-flare/lowerstate PowerLaw
(MJD54690-54701) ΓX =1.45+−00..1121
texp:25230s χ2=0.76/32
F0.3−10keV =1.39+−00..1142×10−12ergcm−2s−1
FIG.4.— Simultaneous gamma-ray and multifrequency light curves ob-
tainedduringthemultiwavelengthcampaignofAugust2008triggeredbythe
high energy outburst discovered by Fermi-LAT.Thefluxabove 100MeV, The X-ray spectrum of each observation segment was fit-
the X-ray flux (0.3-10keV) by Swift-XRT, the six-band fluxes monitored ted with an absorbed power law. Because of the low num-
bySwift-UVOT, the Kanata-TRISPEC differential relative magnitude light ber of events from the source, event were not grouped and
curves(optical∆V andnear-IR∆J bands)andcorrespondingmeasuresof
C-statistics was used, fixing the column density N to the
thelinearpolarizationdegree,andthe15GHzradiolightcurvefromOVRO H
40-marereported. Galactic value NHI = 2.19 1020 cm−2 in that direction
×
(inagreementwithvaluesused,forexample,in Georgeetal.
studiesbetweentheγ-ray,X-rayandUV-opticalbandsduring 1994;Akiyamaetal.2003), andusingz = 1.839. Theerror
boththeactiveflaringstageandthefadingpost-flarestageof onthephotonindexandthe flux(0.3-10keV)is largedueto
PKS1502+106. the low statistics. The backgroundphotonswere selected in
The XRT was set in photon counting mode, and the data a circular region close to the source. No significant photon
were processed by the xrtpipeline with the use of standard index variation was observed between the high and the low
software(HEADASsoftware,v6.4)andstandardfilteringand state,whilethecountrateandfluxdidvary.
screeningcriteria. TheXRTeventsinthe0.3–10keVenergy The Swift Ultraviolet/Optical Telescope (UVOT) photom-
bandwere extractedusing the XRTGRBLC FTOOLfrom cir- etry was done using the publicly available UVOT FTOOLS
cular regions centered on the source position with variable datareductionsuite,andisintheUVOTphotometricsystem
radiidependingon the source intensityand applyingcorrec- describedinPooleetal.(2008). Thephotometricdatapoints
tionforvignetting,Point-SpreadFunctioncorrectionsandhot werecorrectedforGalacticextinctionusingthedustmapsof
pixelsandcolumnswiththeuseofexposuremaps. TheXRT Schlegeletal.(1998)andtheMilkyWayextinctioncurveof
X-ray flux light curve is shown in the second panel of Fig. Pei (1992). These simultaneous multi-band optical and UV
4. The0.3-10keVcountrateofPKS1502+106measuredby datashowanincreaseofabout2magnitudesinallfilterswhen
XRTwasatalevel0.05counts/sec(fromourdata),upfroma comparedwiththepast-yearsarchivalvalues(i.e.,fromabout
levelof0.02counts/sec(archivalpastobservations).The0.3- 19 to 17 in B band). The flux light curvesin the six UVOT
10keVXRT16-daylonglightcurveobtainedinAugust2008 bandsareshowninthethirdandfourthpanelofFig.4. These
(Fig.4) shows an initial count rate of 0.05 counts/sec, and a fluxesappearstobewellcorrelated. Aslightriseinfluxof3
gradualdecaydowntothelevelofabout0.02counts/sec. daysisobservedinalltheUVOTbands,followedbyafading
8 Abdoetal.
similartothefluxdecreaseseeninγ-raysandX-raybands.If The measured flux densities in the MJD 54673–54711
thetimeoftheobservedUVandopticalmaximumisrelated (DoY208-246)timeperiodfita1.69Jyconstant-fluxmodel
to the flare activity at higher energies, this would imply an with χ2/(N 1) = 0.70 (N = 15 data points). This in-
−
interestingtimelagofabout4days. dicates no statistically significant variability in this time pe-
riod. The beginningofa brightradioflare isapparentinthe
3.2. Simultaneousnear-infraredandopticalmonitoring MJD 54762–54809 (DoY 297-344) time period with an in-
creaseofatleast30%overtheearliermeanfluxdensity.
PKS 1502+106 was also monitored in the optical V and
Alessintensivemonitoringat37GHzwascarriedoutwith
near-infraredJbandswithsomephotometricandpolarimetric
the 13.7m radio telescope at Metsa¨hovi Radio Observatory,
snapshots by the TRISPEC istrument attached to the 1.5-m
Helsinki University of Technology, Finland. The flux den-
“KANATA”telescopeattheHigashi-HiroshimaObservatory,
sityscaleissetbyobservationsofDR21,andsources3C84
Japan (Watanabeetal. 2005; Uemuraetal. 2008), within a
and3C274areusedassecondarycalibrators. Adetailedde-
twofoldprogramofopticalfollowupforLATflaringsources
scription on the data reduction and analysis can be foundin
andregularmonitoringofabout20blazars. Imagingrelative
Tera¨srantaetal.(1998).
photometry was performed using some comparison stars in
The PKS 1502+106flare was also followed-upby the Ef-
thesamefield,butduetotheabsenceofanaccuratecalibra-
felsberg100-mradiotelescopewithfourmulti-frequencyra-
tionforthisfieldweprefertoreportonlytherelativemagni-
diospectraobtainedonAugust23,September16,October18,
tudedifference∆magwithrespecttotheminimumlevel(fifth
and December 6, 2008 (within the F-GAMMA project, see
panel Fig. 4). The optical and near-infrared band imaging
Fuhrmannetal. 2007). Each radio spectrum consists of si-
photometryisperformedsimultaneouslyinTRISPECwitha
multaneousmeasurementsatvariousfrequenciesbetween2.6
unitofpolarimetricsequence(consistingofsuccessiveexpo-
and 42GHz. The observationswere performedusing cross-
sures at fourposition anglesof the half-wave plate, where a
scans in azimuth/elevation with the number of sub-scans
setoflinearpolarizationparameters,Q/I,U/I,arederived).
matchingthesource’sbrightnessatthegivenfrequencies.The
These flux observations, performed for a longer interval
datareductionwasdoneusingstandardproceduresdescribed
withrespecttotheFermi-Swiftcampaign(i.e. untilSept. 22,
inFuhrmannetal.(2008);Angelakisetal.(2008).
2008), show a high correlation between the V- and J-band
Otherradioobservationsareavailableatsixfrequenciesbe-
lightcurvesand show an optical decay phase comparableto
tween1and21.7GHzthankstothe600-mringtransitradio
that observed in the UVOT photometric observations (fifth
telescope of the Russian Academy of Sciences RATAN-600
panel of Fig.4). Remarkably, a strong correlation with the
(Korolkov&Parijskij 1979), which observed the source on
LATgamma-raylightcurveisfound,includingthefirst(Sept.
September10,26,andOctober2,2008. Aweightedaverage
04-07, 2008)of the possible minor flaresoccurringafter the
of these three observations, is presented in Fig 7. Previous
initial large outburst. The observation of a flare in the opti-
RATAN-600 data which cover the period between 1997 and
cal(V)andnear-IR(J)bands,simultaneouswithasecondγ-
March 2008 are also shown for comparison. The observing
rayflarehavethetwofoldadvantageofprovidingavalidation
methods,thedataprocessing,andtheamplitudecalibrationis
ofsucha minorLATflareasa realfeaturedisplayedbythis
describedinKovalevetal.(1999).
blazar,and,evenmorecrucially,inconfirmingthefirmidenti-
Allsingle-dishspectraobtainedwiththeEffelsberg100-m
ficationofthenewgamma-raysourceseenbyFermiwiththe
and RATAN-600 radio telescopes are presented in Figure 7.
blazarPKS1502+106.
Here, no indicationof a flare or strong difference/variability
Comparing the Kanata-TRISPEC V-band and J-band col-
between Augustand Septembercan be noted. However, the
ors, the V J color index varies between 2.05 and 1.69
− December2008spectrumshowsthebeginningofabrightra-
(during the Sept.04-07 minor, rapid flare cited above). On
dioflare with a clear spectralsteepeningtowardshigherfre-
theotherhand,thedegreeP oflinearoptical(inV,J bands)
quencies(ν > 10GHz). Thisisingoodagreementwiththe
polarization observed (sixth panel of Fig.4) , remains rather
strong flux density increase seen in the 15 GHz light curve
scattered by errordispersionirrespectivethe flux level, even
duringNovember/December(Fig.7).
during the minor flare mentioned (the maximum degrees
Detailed radio images at sub-milliarcsecond scale of the
recordedduringthe monitoringwere P(V) = 15 3%
max
±
andP(J) =13 4%).
max
±
3.3. Simultaneousradioflux-structuredatabysingle-dish
andVLBIobservations
As part of an ongoing blazar monitoring program, the
Owens Valley Radio Observatory (OVRO) 40-m radio tele-
scopehasobservedPKS1502+106at15GHzapproximately
every two days since mid-2007. Flux densities for the peri-
odsfromJuly26toSeptember3,2008(MJD54673–54711)
and October 23 to December 9, 2008 (MJD 54762–54809)
areshowninFigure6.Fluxdensitiesweremeasuredusingaz-
imuthdoubleswitchingasdescribedinReadheadetal.(1989)
FIG.6.—Longtermradiofluxlightcurveat15GHzobtainedbytheOwens
after peaking up on-source. The relative flux density uncer- ValleyRadioObservatory(OVRO)40mdishradiotelescope(filledcircles),
tainty for this source is dominated by a conservative 1.6% showingtherisingpartofaradiooutburst started inlate November2008,
systematic error with a typical thermal error contribution of i.e. almost4monthsafterthegamma-rayoutburstdetectedbyFermi. The
filldiamondsrepresentthefluxmeasurementsperformedbytheMetsa¨hovi
5mJy. Absolutefluxdensityiscalibratedtoabout5%using
14-mradiotelescopeat37GHz(righty-axisscale),confirmingthestartofa
the Baarsal. (1977) modelfor3C286. Thisabsoluteuncer- radiooutburstatahigherfrequency.ThescaledLATdailylightcurveonthe
taintyisnotincludedintheplottederrors. sameperiodisreportedforcomparison(smalllightgreybars).
PKS1502+106:anewanddistantgamma-rayblazarinoutburstdiscoveredbytheFermiLAT 9
FIG.7.—Variablebroad-bandradiospectraobservedwiththeEffelsberg
100-mandRATAN-600radiotelescopessimultaneoustotheLATdata.His-
toricalRATAN-600data(greyopentriangles)andarchivaldatafromthelit-
eratureuntilMarch2008(greyopencircles)areshowninthebackgroundfor
comparison.
PKS 1502+106 superluminal jet were obtained during three
epochsin2008withFermialreadyinorbit: onJune25,Au-
gust06(duringthemaximumpeakoftheγ-rayoutburst),and
November19. Theseobservationswereperformedaspartof
the MOJAVE monitoring program conducted with the Very
LongBaselineArray(VLBA)atλ = 2cm(Lister&Homan
2005;Listeretal.2009a)andprovidedusefulhighresolution
total intensity and linear polarization images. These VLBA
imagesclosetotheγ-rayflarearereportedandcomparedto
the map obtained one year earlier in Figure 8. The highest
integratedflux densityvaluesince the beginningofthe 2cm
VLBAmonitoringin1997(Kellermannetal.2004)wasmea-
suredonNovember19,2008(DOY324)asF =2.0Jy,
15GHz
withapeakintensityof1.6Jybeam−1. Thesevaluesaresig-
nificantly higher than the typical level of 1.3 Jy reported in
the program (Kovalevetal. 2005; Listeretal. 2009a,b) and
indicate a radio flare happening in the source VLBI core.
The core flux density and brightness temperature raised to
higher values as well which means that the flare happensin
FIG.8.—Totalintensity andlinear polarization images observed bythe
theVLBIcore,asexpected. Thesefindingareingoodagree- VLBAat15GHzaspartofthelargeFermi-supportingMOJAVEprogram.
ment with the single-dish results presented above (Figure 6 Naturally-weightedtotalintensityimagesareshownbyblackcontours. The
and Figure 7). The second relevant feature is the direction contours are in successive powers of two times the base contour level of
1.0mJybeam−1.Electricpolarizationvectordirectionsareindicatedonthe
of the electric vector position angle (EVPA) in the core re-
lefthandsidebybluesticks,withtheirlengthbeingproportionaltothepolar-
gion, which rotated between the 2007 and 2008 epochs by izedintensity. Linearfractionalpolarizationisshownontherighthandside
90 degreesmost probablyindicatingan opacitychange— a overlaidaccordingtothecolorwedge.
precursorofanoutburstintheVLBIcore.
In summary, our single-dish and VLBI radio monitoring
of PKS 1502+106 simultaneous to the Fermi LAT observa- OVROdatabeforeandaftertheoutageaswellasSeptember
tions has revealed (i) no significant radio 15GHz variabil- and October Effelsberg 100-m data are consistent with very
ity during the strong LAT γ-ray flare seen in August 2008, little change in the 15 GHz flux over this time period. The
and(ii) a strongradioflare whichbecomesclearly visibleat threefluxmeasuresat37GHz obtainedattheMetsa¨hovira-
ν > 10GHzduringOctober/November2008(Figure6)with dioobservatoryduringtheOVROoutage,confirmsthistrend
arisephaselastingforatleast20days. Ifthisflaringbehav- with very little variability during this period. However, the
ior is associated with the bright γ-ray flare of August 2008, observed radio flare could also be associated with, e.g., the
a delay of more than three months ( 98 days if the lag morerecent,prominentvariabilityseenintheLATγ-raydata
between the starting days of the γ-ray∼outburst, DoY 218.2, duringNovember/December(Figure2, DoY 320–333). A
∼
andthe15GHzoutburst,DoY316.7,isconsidered)couldbe more detailed analysis of such possible correlations and the
explainedby opacityeffectsin the core regionof the source source’s overall radio/γ-ray behavior seen with LAT and si-
(e.g.,Alleretal.1999). Althoughwecannotexcludethepos- multaneous radio observations over a longer period of time
sibilityofradioactivityat15GHzduringtheOVRO40-mob- willbethesubjectofasubsequentwork.
servationsoutage(September3-Otober232008)thatmight
alsobeassociatedwiththegamma-rayflareofAugust2008, 4. ARCHIVALMULTIWAVEBANDDATA
10 Abdoetal.
A full multiwavelength analysis dedicated on PKS
1502+106isavailableonlyfromGeorgeetal.(1994),where
old archival and broadband radio to X-ray data obtained in
1993-94werepresented.DataandanalysisonPKS1502+106
reportedinotherpapersaremostlylimitedtotheradioregime.
Inordertocompareourmultifrequencyfindingswiththepast,
and to form a more complete characterizationof this blazar,
we briefly present here, for the first time, results from un-
publishedpastobservationsbyINTEGRAL,XMM-Newton,
Swift,andSpitzerspacetelescopesperformedin2001,2005
and2006.
4.1. INTEGRALobservationsin2006
TheskyregioncontainingPKS1502+106wasobservedin
2006byIBIS(83ksec,MJDs53760.4to53762.4,Jan.25-27,
2006),anda new softγ-raysource(IGRJ15039+1022)was
detected with a flux density of 1.6 mCrab in the 18-60 keV FIG.9.— The unique observed mid-IR spectrum in the range 5-14 µm
obtained bythe Spitzer Infrared Spectrograph (IRS)low resolution (R =
energy range (correspondingto 1.2 10−11 erg cm−2 s−1, 60−130)module,inAugust13,2005. ThepositionofredshiftedBrackett
×
see ATel #1652). ThisIBIS source was identifiedwith Mkn emissionlinesandPAHlineareindicated, eveniftheyarenotdetected in
841, a Seyfertgalaxyknownto display a well detectedhigh theIRSspectrumthatisconsistentwithasimplepowerlawmodel. Inaddi-
tiontheopticalspectrumbytheSDSSonApril23,2006,andhighprecision
energycut-offaround100keV(Petruccietal.2002),making
JHKSphotometricfluxmeasurementsoftheKittPeakNationalObservatory
itunlikelytoemitintheγ-raydomain. Theangulardistance (KPNO)2.1mtelescopeofMarch29,2002arealsoreported.
of PKS 1502+106fromIGR J15039+1022, 11′, pointsto
∼ the standard Spitzer IRS pipeline (ver. S17.2). Background
a clear non-detectionduringthisJan. 2006INTEGRALob-
was subtracted using the two nod positions along the slit.
servation, while a 2σ upper limit for PKS 1502+106 of 0.7
The spectra were extracted and flux-calibrated with SPICE
mCrabintherange18-60keV(0.52 10−11ergcm−2 s−1)
× ver. 2.1.2, in a standard, expanding point-source aperture.
isinferred.
The two spectral orders match well at 7 µm, indicating a
well-pointed observation. The mid-IR continuum of PKS
4.2. XMM-NewtonandSwift
1502+106rises to the near infraredand appearsto be rather
Four serendipitous, unpublishedXMM-Newton X-ray ob- featureless, consistent with pure synchrotron emission (a
servations of PKS 1502+106 by the EPIC (MOS detector power law F ν−0.9 in the 5-14 µm range, Fig.9). The
ν
only)cameraareavailable asthe sourcewas in the frameof wave-like devia∝tion of the data can be simply explained as
the target Seyfert galaxy Mkn 841. PKS 1502+106 was al- wavelength-dependent spectrograph slit losses, while the
waysontheborderoftheMOSchipsandoutofthePNframe, red-shifted Brackett emission line series (like the 3.3 µm
andthereforesubjecttolowX-raystatistics, regardlessofits PAH feature) falling in this wavelength range are indicated,
intrinsicbrightness. ThefourX-rayEPIC-MOSobservations but they are not well detected in the spectrum. This IRS
(threein2001andoneperformedinJul.17,2005,seeTable3 spectrumissimilartootherblazarswhichhavebeenobserved
forananalysissummary)donotshowvariationsinthe0.2-10 by Spitzer, including BL Lac and 3C 454.3 (Leipskietal.
keV photon index, while the 0.2-10 keV flux intensity var- 2008; Ogleetal. etal. 2009), while the near IR (J,H,K) flux
ied bya factorofa few (inthe range3.5 6.8 10−13 erg datafromthe KittPeak NationalObservatory(KPNO)2.1m
cm−2s−1in2001,comparabletothelower−states×observedby telescope Watanabeetal. (2004) reported in the same figure
ASCA,andamildlyactivestatewith1.1 10−12 ergcm−2 indicatealowerfluxstate andsteeperH-K (1.65-2.15µm),
s−1inJuly2005,Table3). × S
PKS 1502+106 was also observed twice in the past with TABLE3
theSwift-XRTasafill-intarget(TargetID:36388),showinga ANALYSISSUMMARYOFTHEEPIC-MOSINSTRUMENTOBSERVATIONS
fainterX-rayflux(0.02counts/sec)thanthefluxrecordedin (JAN.2001ANDJUL.2005)ONBOARDOFXMM-NEWTON.
theAugust2008campaignobservations(0.05counts/sec). Obs.id.(date) Best-fitModelandParameters
ObsID0070740101 PowerLaw
4.3. Spitzerobservationsandthemultifrequencybehavioron (Jan13,2001,09:20UTC) ΓX=1.6±0.2
Jul-Aug.2005 χ2=1.69/11
r
In the past, only upper limits in the far-/near-infrared F0.2−10keV =3.5×10−13ergcm−2s−1
ObsID0070740301 PowerLaw
bands by IRAS were available for PKS 1502+106
(Neugebaueretal. 1986, and Figures 9 and 11). PKS (Jan14,2001,00:30UTC) χΓX2==1.12.77/1±70.2
1502+106 was observed serendipitously in the mid-infrared F0r.2−10keV =6.8×10−13ergcm−2s−1
bandforthefirstandonlytimebySpitzeronAugust13,2005,
ObsID0112910201 PowerLaw
09:10-09:18UT(PID117,AORRequestKey5011456).The (Jan13,2001,04:58UTC) ΓX=1.6±0.2
Infrared Spectrograph (IRS, Houcketal. 2004), low resolu- χ2=1.05/8
r
tion(R=60-130)module,recordedthemid-IRspectrumfrom F0.2−10keV =3.6×10−13ergcm−2s−1
5-14 µm (shown in Fig.9 with the optical SDSS spectrum ObsID0205340401 PowerLaw
and near-IR photometric data point). High resolution IRS (Jul17,2005,06:32UTC) ΓX=1.69±0.08
modulespectra were also taken, butcould notbe used since χ2r=0.99/58
there were no accompanyingbackgroundobservations. The F0.2−10keV =11.0×10−13ergcm−2s−1
Short-Low (SL) coadded 2D spectra were reduced using