arXiv:1401.2151v5 [astro-ph.HE] 13 Mar 2014 K s t a ra e t n e pu Bl I 1. 1974 a ca B er h M A e nd t missi n e . d e . l ll azar n er t i s v h arc i L S go ng yb t ts: o n e- a e i L acr A yd gy r o gh . t o ; ti M s ono t m r h g K ob GeV C s t n st f R a NOT M fl A I I a C r . 4 2014 14, h r Kawa . h tt on no on isi o n ts s ( p nd on ea a a er K r w are G i C o odu r ey on e onte a d re er pho s e ethod E . st re m V ec l ti 23 s u f il: yn s a sm ss ac are a ng er op h . S m arra t m ea gby ng c ist . . g e s il T x p u c VL c w t s t , is V ra r . . w R i v o s l Fili l y 24 u ea c y yn fre e ti a d, u t udg ound t e o p x n t r p e w s We e re S on m . it c h re nd li d o s: i s ca t. ec fl ea B b er ec d . s , S lli k bu h o s S r t i h t r c xo ux nd . pon on nd T a Aa h e o ED P W i d c ac i t ls c s pp st g t h l ti aw k Kec ill o oko t m ng yp n on e t missi T ra h a A h i t re ti h o s. i a r nv v i e r s ti i&M fr e ti a e . h i ea o s p o b ng 7 on on l o x, a 5 v st t e i . n e a v ng p f s e hu e v re ec ca 18 M , o t a e i , s f v ra k nko l ac e If ec r n r ou l t -ra isti a sti t im qu m g w 7 ar d S h npho on c T ov ophy e t l s m t t , tis xy d t a on r ( a , a e t era ou h si e t h it . 10 . i sl o u i g r S e A a h t y x l l p a -ra o J l a ee u ce ac o h e e f ms B s ac a K 8 ara s t e yun f- ED) b sissm n - n a . iestates tive wee m l s f ra ar x t . ky b r a are fr , b t M ilit db t d ec c . 6 L o C h t t d ti y si a Si e o l d J :i s: op l i l h z er o -ra azar s re C pho e m e no ,bu o, ho ec ec li s ll o c . 1 . i o e c y c re mt t h o t e h ec n n s m e e er Ri e e (A D s on h er a a ov n - o w 2 y t nu t nd e v m ar m v s a nko r l r r ng i t a 0 a ts m mi m fa t p s far-I u n oud d . er on du e o sst ls on a b fl 3, e h ll a d c l ar era s bdo t i t er c 500. y t d m a dby u Fi b v 1989 y a ux e e ec e h er np on od no n s l ze op s k nd l e i si e s st nu s 17 ou e t ar i 8 e du 30 re a ll r a j o , c nk mitt ob ng b du i k ong e st R d po b gn is ti f m t arr n , F e s 9 l I d s t rce , o 25 a p l (AGN T ca azar Rt u i r m t t , t e cr r ho Y . e l: ;M rec -ra s t do og t a ec a h D o e sit i e lt h , h h o 2 a C 16 22 ng K , t a e l; nd e u. ce ti i eAGN t mi w- e f h e s do s s l G o t gn . p dby e fr h y I o mi r t r 2010b . e . c a OAGH F M e . , , t n . u ob ee on t J - r B ss (e 1 t o a h A x li C er ar l it o b S a mi h h L X-ra . m D R Ne o 1310 no. s ss -ra u n oud m er i n s a t . n t f i 1310 e , ud x j a p nv u c mi x T . si u . dd s . ) t ece ec a eac . o SSC; Ya n st ob t re t ec n l h c re C t K y v e F t ho er ra c o e arr Ri st i o t h er o e t a e n ar n i n li i h w-ac ( ’ a S s uh h y i n ov t n a st o g t er o w A n ov s er t m o v gh er ti nd r ee arr e s ) y race ED m ti o - t qu e f a ze f nd ft u w a e . ra e- v ra h li wa dby g ob i 8 g – 487 we l nb on v h ung er a r d a t lC ng s & m tis ac e T e i n p e h h o li m e yn AGILE d era n l e l C ti s e l . WI .g., ee N w- g e 1 ddu s h e a i ec i . s v t ti 1 a v j lm t c h Tra o missi o h a a re on r ti v c e e h ov ob o e cra , it T p k e E 4 h - o l 10 e nn m h li s t t S on e 2 l ac missi m t t l r G d y C , l a r . J a h yn e e o r wee a n e ob e j Ea , o l 9 r on re S o v i , on p ec 33 Y d o ng ti -GeV e R m st ti i e p s 30 . a d w-e 1 , a Y t ng De ED is t w-e K on v c a r c t . ti i B l l 100 c , a on Pitt a t b M S s , st s ti np on c on o a . e a n ec b , h t n nb an t v a d nu T on er Y p a . e h x o e s ti M - l Y t r n R t s 1996 do nd e y i y li n h ec . i . . n I s e m o . h v ou a a frare e on e s c s . 1310 fer M ,2012 8, . e F vo n AB -ra llit er er o . Ue s Ta ou . t a e ti j s: isti e F ca w t l a po t K r e p e er Cl e b m r r G rce W l m rr t Y ra gy Va eV on nd a on gy uk l f- ec h u t o i r y r e fc g f u t t mi n ov l S 26 i p i uda ound il s o m e ) - s - - - - - 1g d C l . bu a ubb . gh ti a t li a TRACT ts e ( (AGN) o l s azawa on ra a non on o nd R (I , s a k fi u a h ob ti t b nd t yn m x B l ; a l T e R l e a –g ra ac o st w-li i on i 2009 t 2011 m o t i g war T ce ce fa fere i e l st acce dbyg e 8 h on by ng n d ra p 5 s . e a ) 1310 m r r c -ra ra v h n er ; 5 k , s st t ef od c x , a i ound s P l ss h ft d on a er . g e eca 2 a era i , l i i F i 5 a P r v e tim b ng er i e m mm n u ud o -ra nd y c n i o , sl is o H 17 , e d gy o a p rt n . . 487 1 s no i f s ) ;B r t t E ng ( ti m , ls e t ll m , a , ou r i D’A K fl t sim sl SSC y e e t ra . , I e h on n 18 e n ob on o w h 487 C; nd T o du ara are R d w od e s U e ( er r dp f E o st ca . v o re e t w ( rce ca it o ist CMB . t A , )e h J s s e re wer it on h b M e r A e j h a g . a ng st h a t e o e s P l ec e e e K i )a t a M r re ere l tt l re J e i accre e dp a r mm ar y s nu gd ng Ha e e f .g., azar ls e . bo 27 missi e c ti ü i s d . t t o e ho c ti s: . s bu v . h S s; ts w erec ou Kr on nd r c g r ti 30 ( A h o a ar T v ea ( e o e , o i S du er k o . we lm t y f ) g C c er ti g isti n h l f f ,2014 8, y A Si a . , h a a e re is l a db i l g T i s m t a mitt on ce e e e w e h ll no m e ndu on h i c st Ze k ndo w l a l 9 d x i t missi &P gh 00 2012 2010, . c h d a l h ko a s i a , e i on h hb ce mma-ray-l s im o , t on e gh y f st u s . it ee s ti x B a e OV ts er -ra ti ( ) ee t f n b t & tti h -e S e n ov S i ssi a We v r h e ra nd on s e si r t a i t . 30 n a d ong m 11 -e ED) s o h . t st f h i isti r n y l s t -ra a r gh u o o on ng. R t t P ec u o s e x 3 er t on. E i h li a lC od X-ra , n fl z . f r – ng o m O t s u , t t b 12 v e n y u t h er gy . S are c er t 1 17 t c e G dop lis c s ar E a r i e ra h s e er T 1 40 is du Hea o fl o st rre . o on hk s gy l ) n p ls h e , 13 0 . m i , h rre 1994 . are d , y m h s , b k, Verce ( -ra o a a r r is GeV i eAGN S A i m e M j 638. b w o s o p i nv tt e are ls a , ) pon a e e b d. ng l s l l . e cc . t ti t du nd y , a a l h ng t L e missi 2009 on yn on a on h are . lli er e , Lar 2001 E A on ti ti i ee s u r fl h c . e y S v a a e nb on v e s o rre n I s n c h e i are si nd n nd E ; (E 18 e- 28 ll gh l ts a don pho d e ou ec a lt h l i m e ts Der ls ss s a is a g d s on S C ra r db 2008 p C h , n on s -e . a lt rce ;M ec ca k o af e - b e 29 ED t on . i t o on v Y )e h li d T no t n n er er on t is n t e m oud T wee t h i r ef N ec h e m a nd n er h . era is o er s Ar on 32 100 g a t 1 h 1 n i missi st tt p 19 31 e is ere ov l t r o i I s nv ü n f , e ; er gy e , on a c t i k ho re t o ea re , o e isms: ti on c h bu t n yp h G j ll A ti A e , e d ra e m e 14 er smi k r 20 c F e (UV) xp p e m &S v w y s i t R ough e ji h t l )e . ee i . h . ti missi g re on, e od ddu nd missi e s m p bou ca , s b is r e l f C . are ere B Verrecc on i 21 e- referre r a nu er i o S on v c missi s , a o e l S A 17 t n e . c non u e r R i C mi , . lli ;t t -ra mi a tt s t a n ls f m c wre ew h 18, n a S h l w ATAN- o , ti M J l o ce t t on. g h nun G on li r v e h t n w 2003 . b h h n . ra ( . on i o a R e y m t cr are mi eKa ar B c ng er do on i& h e eE e ob R - st o . dbyh F reac b k fl n hu . c er d p L o N ,p y L e e wee ea p fi t e re ux b s fr I h a d g t h wa E m t i lli R g is m on er r t st ts l . wee ng t C e Ta ar a l t n i sl azar o oh im h S o 600, a a dh a )c -ra 1 a L g er p v a a mi 6 i m b v t o ;Si l k n O d , 26 e ng nd t e e h e v , ist a p w s a a f e a s 17 ar 15 i d d – n d y ea e ecc 2002 - 1o d o ca , . . r ng sis sl l 2014 , b o oud y o r r er , , ko d ac on tt ce t d f e h h ac- t 25 f 22 er- h i 18 p o k ra i er ss i f- s ) o e c - r - , , . , A A proofs: manuscript no. 1310

Two types ofradio-loud AGNs giverise to the blazar Theradio source GB 1310 487 [also known as phenomenon: flat-spectrum radioquasars (FSRQs)and GB6 B1310 48441, and CGRaBS J1312 4828, listed in BL Lacertae-type objects (BLLacs). Flat-spectrum radio the Fermi -ray source catalogues as 1FGL J1312.4 4827 are characterizedbyhigh luminosities,prominent (Abdo et al. 2010d)and 2FGL J1312.8 4828 (Nolan et al. 2 h m s broad emission linesin their opticalspectra, and the peakof 2012), radio VLBI position J2000 13 12 43 353644 synchrotron jet emission occurring atmid- orfar-IR wave- 0 22 mas, J2000 48 28 30 94047 0 16 mas (Beasley et al. lengths. Thermal emission, probablyoriginating in theaccretion 2002)] is a flat-spectrum radio source. It was unremarkable disk surrounding thecentral blackhole, may contributea among other faint -raydetectedblazars (the E 100 MeV significant fraction of the optical and UV emission in some flux during the first 11 months of the Fermi mission was FSRQs (Villataet al. 2006;Jolley et al. 2009; Abdo et al. 3 10 8 photons cm 2 s 1, as reported in the 1FGL catalogue; 2010a). BL Lacertae-type objects,onthe other hand, show Abdo et al. 2010d) until it appeared in the daily Fermi sky with mostly featureless opticalspectra dominatedbythe nonthermal a flux of 1 0 10 6 photonscm 2 s 1 on 2009 November 183 continuum producedbyarelativistic jet. Their synchrotron (Sokolovsky et al. 2009). AGILE observations reported twodays emission peak is locatedbetween far-IR and hard-X-ray energies later confirmed the high-flux state of the source (Bulgarelli et al. (Padovani&Giommi 1995;Fossati et al. 1998; Ghisellini et al. 2009). Follow-up observationsin the near-IR (Carrasco et al. 1998). In GeV -rays, BLLacsshowawide distribution of 2009)and optical (Itoh et al. 2009)also found GB 1310 487 in spectralslopes, while FSRQs almost exclusively exhibit soft a high statecompared tohistorical records. The daily average -ray spectra (Abdo et al. 2010c). Itisnot clear whether there is -ray flux remained at 1 0 10 6 photonscm 2 s 1 for more a physical distinction between BLLacs and FSRQs,or if they than aweek (Hays&Escande 2009). representtwo extremes ofacontinuous distribution ofAGN This paper presents multiwavelengthobservations of propertiessuch as blackhole mass (M ), spin, or accretion rate GB 1310 487 before,during, and after its active -ray state, and (Ghisellini et al. 2011). Recently, fiveradio-loud narrow-line suggests possible interpretations of the observed SED evolution. Seyfert 1galaxies (NLSy1s) have beendetected in -rays by In Sect.2we describe the observing techniques and dataanal- Fermi LAT, suggesting the presence ofanew class of -ray- ysis. Sect.3presents anoverview of the observational results. emitting AGNs (Abdo et al. 2009d; D’Ammando et al. 2012). In Sect.4we discuss their implications, and we summarize our Therelationshipbetween NLSy1 and blazarsisunder debate. It findingsin Sect.5.Throughoutthis paper, we adoptthefol- has been suggested that radio-loud NLSy1 harborrela- lowing convention:the spectralindex is defined through the tivistic jets (Foschini 2013; D’Ammando et al. 2013),but unlike energy flux density as afunction offrequency F , the blazarsthey are poweredbyless massive blackholes hostedby photon index ph is defined through the number of incoming spiral galaxies (Yuan et al. 2008; Komberg & Ermash 2013). photons as afunction ofenergy dN(E) dE E ph , and the The presence ofarelativistic jetissupportedbyobservation of two indices are relatedby ph 1 . We usea CDM cos- superluminalmotionsin the -scaleradio jet of theNLSy1 mology, with thefollowing values for thecosmological param- 1 1 SBS 0846 513 (D’Ammando et al. 2012). The observational eters: H0 71 kms Mpc , m 0 27, and 0 73 (see evidence that radio-loud NLSy1 have M smaller than those of Komatsu et al. 2009; Hogg 1999), which correspondsto a lumi- blazars has recentlybeen challengedbyCalderoneet al. (2013). nositydistance of DL 3800 Mpc, an angular-size distance of 1 Some nearby radiogalaxiesincluding Cen A(NGC 5128), Per A DA 1400 Mpc, and a linear scale of 6.9 pc mas atthe source (NGC 1275, 3C 84), and VirA(M87, 3C 274)aredetectedby z 0 638 (see Sect.3.4). Fermi LAT (Abdo et al. 2010j). While part of their -ray luminosity is attributed to inverse-Compton scattering of CMB photons on theextended (kpc-scale) radio lobes of the galaxies 2. Multiwavelength observations (Cheung 2007; Abdo et al. 2010i), contribution fromthecore 2.1. Gamma-ray observations with Fermi/LAT region is also evident (Abdo et al. 2009b, 2010h). Unlike other radiogalaxiesstudiedbyAbdo et al. (2010j), Per A exhibits Fermi Gamma-ray Space Telescope(FGST; hereafter Fermi) episodes ofrapid GeV variability (Donato et al. 2010;Ciprini is anorbiting observatory launched on 2008 June 11 by a 2013). Thecore -ray emission in radiogalaxiesisprobably DeltaII rocket fromthe Cape Canaveral Air Force Station in producedbythe same mechanisms asinblazars,but with less Florida, USA. The main instrument aboard Fermi is theLarge extremerelativistic beaming. Area Telescope(LAT; Atwood et al. 2009; Abdo et al. 2009a; Since early satellite observations established theassociation Ackermann et al. 2012), a pair-conversion telescope designed to of some discrete -ray sources with AGNs, it becameclear that cover theenergy band from 20 MeV togreater than 300 GeV. blazars emit aconsiderable fraction of their total energy output The Fermi LAT is providing a uniquecombination of high sen- above 100 MeV (Swanenburg et al. 1978; Hartman et al. 1999; sitivity and awide fieldof view ofabout 60 . Fermi is operated Mukherjee 2002). Thecurrent generation of space-based -ray in an all-sky survey mode most of the time, which makesitideal telescopesthat use solid-state(silicon) detectorsisrepresented for monitoring AGN variability. by instruments onboard AGILE (Tavani et al. 2009, 2008)and The dataset reportedhere was collectedduring the first Fermi (Atwood et al. 2009), whichopen awindow into theworld 33 months of Fermi science observations from 2008 August 4 ofGeVvariability and spectral behavior of -ray-loud AGNs. In to 2011 June 13 in theenergy range 100 MeV – 100 GeV. The contrast toprevious expectations (Vercelloneet al. 2004), most 33 month time intervalisdivided into subintervals according to of the brightest -rayblazars detectedbyFermi and AGILE were the level of its -ray activity as observedbyFermi LAT (see Ta- already known fromtheEGRET era (Tavani 2011). On the other ble 1). hand, many blazars previously unknown as -ray emitters were 1 ThecorrectB1950 source name, if its declination is expressed as 6 2 1 observed to reachhigh fluxes ( 10 photons cm s at ener- three digits, is 1310 487, while 1310 484 is anunrelatednearby radio gies E 100 MeV) for only a short period of time during a flare. source. In this work, we present a detailed investigation of one suchob- 2 see ject. 3 UT dates are used through the text.

Article number,page 2of 18 K. V. Sokolovsky et al.: Two active states of the narrow-line gamma-ray-loud AGN GB 1310 487

Table 1. Changesin the -ray spectrum between time intervals considered in theanalysis.

Period UT interval 100 MeV–100 GeV flux ph TS N 33 months 2008-08-04 – 2011-06-13 (1 03 0 04) 10 7 2 18 0 02 4415 3566 pre-flare 2008-08-04 – 2009-11-16 (0 34 0 05) 10 7 2 41 0 09 205 476 Flare 1 2009-11-16 – 2009-12-21 (6 94 0 32) 10 7 1 97 0 03 3333 947 Interflare 2009-12-21 – 2010-04-26 (1 37 0 11) 10 7 2 15 0 06 917 592 Flare 2 2010-04-26 – 2010-07-26 (2 83 0 16) 10 7 2 14 0 04 1839 907 post-flare 2010-07-26 – 2011-06-13 (0 44 0 06) 10 7 2 34 0 09 236 422 Columndesignation: Col.1, -ray activity state;Col.2,time interval used for spectral analysis; Col.3,average flux inunits of photons cm 2 s 1;Col.4,photon index: dN(E) dE E ph ;Col.5,Test Statistic(TS) defined in Sect.2.1; and Col.6,number of photons attributed to the source (model dependent).

Fermi LAT datacompriseadatabasecontaining arrival dent time bins. The time bin width was chosen tobe sevendays. times,directions, and energies of individual silicon-tracker Sources with less thanone photon detected in the individual bin events supplementedbyinformation aboutthe spacecraft posi- orwith TS 25 were excluded fromthe sky model for that tion and attitude needed to calculate thee ectiveexposure for bin. The lightcurves were computedbyintegrating the power- acelestial region and time interval of interest. The maximum- law modelin theenergy range 100 MeV – 100 GeV. likelihood method is used to analyze these data by constructing For lightcurves with time bins of fixed widths, thechoice anoptimalmodel of the sky region as acombination of point- of bin width is acompromise between temporal resolution and likeand di use sources having a spectrum associated with each signal-to-noiseratio for the individual bins. For Fermi LAT one of them (Mattox et al. 1996; Abdo et al. 2010d). The sig- an alternative method has recentlybeendeveloped (Lott et al. nificance of source detection is quantifiedbytheTest Statis- 2012), in which the time bin widths are flexibleand chosen to tic(TS) value,determinedbytaking twice the logarithm of the produce bins with constant flux uncertainty. Flux estimates are likelihood ratiobetween the models including the targetsource still produced with the standard LAT analysis tools. In this case (L1)and one including only the background sources (L0): TS we used v9r27p1 and 2(ln L1 ln L0). Theratios L0 and L1 are maximized with respect eventselection and IRFs, forwhich thecurrent version of the to the free parametersin the models. The Monte-Carlo simula- adaptive binning method has beenoptimized (we havechecked tion performedbyMattox et al. (1996)forEGRET confirmed that using the class yields very similar fluxes). At theoretical predictions (Wilks 1938) that foraGeVtelescope, in times of high source flux, the time bins are narrower thandur- most cases, theTS distribution is close to 2. ing lower flux levels, therefore allowing usto study more rapid The unbinned likelihood analysis was performed with variabilityduring these periods. the package4 version v9r21p0. The The lower energy limit of the integral fluxes computed for class events in theenergy range 100 MeV – 100 GeV theadaptivelybinned lightcurve is chosen to minimize the bin were extracted from aregion of interest defined as acircle widths needed to reach the desired relative flux uncertainty for of 15 radius centered attheradioposition ofGB 1310 487. most bins. The derivation of this energy limit, called the op- Acut on zenith angle 100 was applied to reduce timum energy, is presentedbyLott et al. (2012). Because the contamination from Earth-limb -rays,producedbycosmic source is variableand the optimum energy value depends on raysinteracting with the upper atmosphere (Shaw et al. 2003; the flux, we compute the optimum energy with theaverage flux Abdo et al. 2009c). Observatory rocking angles of greater than over the first twoyears ofLAToperation reported in the 2FGL 52 were also excluded. A set of instrument responsefunc- catalogue(Nolan et al. 2012). The optimum energy is found to tions (IRFs) was used in theanalysis. The be E0 283 MeV for this source. We produced two sets of sky model containedpointsources fromthe 2FGL catalogue adaptivelybinned lightcurvesin the 283 MeV – 200 GeV energy (Nolan et al. 2012)within20 fromthe target, as well as range,onewith25% flux uncertainties and another with15% Galactic and isotropic uncertainties. For eachof these uncertainty levels we created a di usecompo- second version of the lightcurve by performing theadaptive bin- nents5. All point-source spectra were modeled with a power law; ning in thereverse-time direction. the photon index was fixed to thecatalogue valueforall sources exceptthe target. The di use-background parameters were not fixed. Theestimated systematic uncertaintyof flux measure- 2.2. Gamma-ray observations with AGILE/GRID ments with LAT using P6_V11 IRFs is 10%at 100 MeV,5% 6 The AGILE -ray satellite(Tavani et al. 2009, 2008)was at 500 MeV, and 20%at 10 GeV and above . launched on 2007 April 23 by a PSLV rocket fromthe Satish The lightcurve of the targetsource was constructedbyap- Dhawan Space Centre atSriharikota, India. AGILE is a mis- plying theaboveanalysis technique to a number of indepen- sion of theItalian Space Agency (ASI) devoted tohigh-energy astrophysics, and is currently the only space mission capa- 4 For documentation of the , see ble of observing cosmic sourcessimultaneously in theen- ergy bands 18–60 keV and 30 MeV –30GeV thanksto its 5 The models are availablefromthe Fermi Science SupportCenter two scientific instruments: the hard X-ray Imager (Super- 6 For newer P7_V6 IRFs used in theadaptive lightcurveanalysis AGILE;Feroci et al. 2007)and theGamma-Ray Imaging Detec- describedbelow, the systematic uncertainties are lower at high ener- tor(GRID;Rappoldi&AGILE Collaboration 2009). During the gies: 10%at 100 MeV,5%at 560 MeV,10%at 10 GeV and above first twoyears of the mission, AGILE wasmainlyoperatedby (Ackermann et al. 2012); however, this di erence is not critical for the performing 2–4 week-long pointedobservations,but following present analysis. the reaction wheelmalfunction in October 2009 it was operated

Article number,page 3of 18 A A proofs: manuscript no. 1310 in a spinning observing mode, surveying a large fraction of the mild spectral changes. The Swift XRT observation obtaineddur- sky eachday. ing the Flare 2 and post-flare intervals (Table 1)resulted in a low The AGILE GRID instrument detected enhanced -ray number of detected photons. The Cash (1979) statistic is applied emission from GB 1310 487 from 2009 November 20 17:00 to fitthis dataset with theabsorbedpower-law model. The Cash (JD 2455156.2) to 2009 November 22 17:00 (JD 2455158.2) statistic is basedona likelihood ratio test and is widelyused for (see Bulgarelli et al. 2009, for preliminary results). Level 1 parameter estimation in photon-counting experiments. The net AGILE GRID data were reanalyzedusing the AGILE Standard count rate in the 0.3–10 keV energy rangechangedbyafactor of Analysis Pipeline(see Pittori et al. 2009; Vercelloneet al. 2010, 1.6 between the twoobservations conductedduring Flare 1 and foradescription of the AGILE datareduction). We used - by afactor of 3.5 over thewhole 33-monthperiod. The X-ray ray events fromthe archive, filteredbymeans of spectral analysis results are presented in Table 2. the pipeline. Counts, exposure, and Galactic back- ground -ray maps were created with a bin size of 0 5 0 5,for E 100 MeV. Since AGILE wasin its spinning observing mode, 2.4. Ultraviolet–optical observations all maps were generated including all events collectedupto50 The Swift Ultraviolet-Optical Telescope (UVOT;Roming et al. o -axis. We rejected all -ray events whosereconstructeddi- 2005) has a diameter of 0.3 m and is equipped with rections form angles with the satellite-Earthvector smaller than a microchannel-plate intensified CCD detector operated in 90 , reducing the -ray Earth limb contamination by excluding photon-counting mode. Swift UVOT observed GB 1310 487 si- regions within 20 fromtheEarth limb. We used the latest ver- multaneously with Swift XRT. Various filters were used at dif- sion ( ) of the Calibration files ( ), which will be ferent epochs ranging fromthe U to M2bands (as detailed publicly availableattheASI Science Data Centre (ASDC) site7, in Table 2), with the best coverageachieved in the U band. and the -raydi useemission model (Giuliani et al. 2004). We Since the objectisvery faint, multiple subexposurestakenduring subsequently ran the AGILE Multi-Source Maximum Likelihood eachobservation were stacked together with the tool Analysis ( ) taskusing aradius ofanalysis of 10 inorder fromthe package. Acustom-made script basedon toobtain the position and the flux of the source. A power-law was employed foraperture photometry (using the spectrum with a photon index 2 1 was assumed in theanal- standard5 aperture diameter) and count rate to magnitudecon- ysis. version taking into accountthecoincidence loss (pile-up)correc- tion (Pooleet al. 2008;Breeveld et al. 2010). TheGalacticred- 2.3. X-ray observations with Swift/XRT dening in the direction of this source is E(B V) 0 013 mag (Schlegel et al. 1998). Using theextinction law of Cardelli et al. The X-ray Telescope(XRT;Burrows et al. 2005) onboard the (1989)and coe cients presentedbyRoming et al. (2009), the Swift satellite(Gehrels et al. 2004) providessimultaneousimag- following extinction values were obtained for the individual ing and spectroscopiccapabilityover the 0.2–10 keV energy bands: AV 0 041, AB 0 053, AU 0 065, and AM2 0 122 range. The source GB 1310 487 was observedbySwift atseven mag. Magnitude-to-flux-density conversion was performedus- epochs during the two -ray activityperiods and in June 2011 ing thecalibration of Pooleet al. (2008). during the low post-flare state. A summaryof the Swift ob- A -like objectisvisible in NordicOptical Telescope servationsispresented in Table 2. Swift XRTwas operated in (NOT) images about 3 southwest of theAGN. This object photon-counting (pc) mode during all observations. The low wouldbe blended with theAGNin UVOT images which lack count rate of the source allows ustoneglectthe pile-up e ect adequateangular resolution. Contribution of this objectto the which is ofconcern for theXRT inpc mode if thecount rate8 is total flux measuredbyUVOT is the likely reason for the dis- 0 6 counts 1. crepancybetween UVOT and NOT U-band measurements dur- The task fromthe package ing the low state ofGB 1310 487. The nearby (Sect.3) was used for the data processing with the standard filtering also contributesto the measured UVOT flux. criteria. To increase the number ofcounts for spectral analy- TheNordicOptical Telescope, a 2.5 minstrumentlocated sis, theresulting event files were combined with to on La Palma, Canary Islands, conducted photometric observa- produce average X-ray spectra for the periods of Flare 1 and tions ofGB 1310 487 with its ALFOSC camera on 2010 July7 Flare 2defined in Table 1. The spectrum for the Flare 1pe- and 2011 May29during the second flare and the post-flare low riod was binned to contain atleast 25 counts per bin toutilize state, respectively. TheVaST9 software (Sokolovsky & Lebedev the 2 minimization technique. Thecombined spectra were an- 2005)was applied for the basicreduction (bias removal, flat- alyzed with . The simpleabsorbedpower-law fielding)and aperture photometryof theNOTimages. A fixed model with theHIcolumndensity fixed to theGalactic value aperture 1 5 indiameter was used for the measurements. 20 2 NH I 0 917 10 cm (obtained from radio21cmmeasure- The source 3UC 277-116569, which served asthecompari- ments by Kalberlaet al. 2005) provided a statistically acceptable son star for theKanata observations (see below), wassatu- fit (reduced 2 1 2 for 12 degrees of freedom) to the 0.3– ratedonNOT i band images and so couldnot be used. Instead, h m s 10 keV spectrum. Leaving NHIfree tovary results in the values SDSS J131240.83 482842.9 ( J2000 13 12 40 84, J2000 22 2 NHI 0 4 0 5 10 cm and ph X ray 1 3. However, this 48 28 42 9, Abazajian et al. 2009;see Fig. 1)was used as model does notimprove the fits. The low photon counts prevent thecomparison star. Its Johnson-Cousinsmagnitudes were com- a more detailed study. puted fromthe SDSS photometryusing conversion formulas of Individual observations obtainedduring the periods of Jordi et al. (2006): U 19 133 0 081, B 19 303 0 013, Flare 1 and Flare 2 were also analyzedusing the same fixed–NHI V 18 822 0 012, R 18 589 0 011, and I 18 177 0 018 model,but no evidence of spectral variability within the periods mag. was found; however, the low photon counts could easilyhide Kanata, a 1.5 mtelescopeattheHigashi-HiroshimaOb- servatory, observed GB 1310 487 in the R and I bands with 7 8 9

Article number,page 4of 18 K. V. Sokolovsky et al.: Two active states of the narrow-line gamma-ray-loud AGN GB 1310 487

Table 2. Swift observations ofGB 1310 487.

ObsID Date(UT) JDExposure UVOT XRT 0.3–10 keV unabs. flux ph X ray 2455... (ks)(mag)(cts s)(10 13 erg cm 2 s 1) Flare 1period 001 2009-11-27 162.77 8.7 M2 21 1(2) 0 022(2) 16 7 3 3088 0 18 002 2009-11-30 165.97 8.1 U 21 2(4), B 21 3, V 20 90013(1) Flare 2period 003 2010-06-25 372.66 4.5 U 21 3, B 21 5, V 20 50010(1) 004 2010-07-03 380.65 4.5 U 20 9(3), B 21 5, V 20 30009(1) 6 6 1 5115 0 23 005 2010-07-07 384.63 4.9 U 21 4(3) 0 009(1) 006 2010-07-11 388.68 4.6 U 21 3(3) 0 010(1) post-flare period 007 2011-06-03 717.12 9.5 U 21 1(2) 0 006(1) 5 5 1 8093 0 34 Columndesignation: Col.1,observation number in the Swift archive omitting the leading 00031547;Cols.2,3,date of observation givenbytheGregorian and Julian Date;Col.4,exposure time inkiloseconds; Col.5,Swift UVOT photometry (here and later in the texttheerror inparentheses correspondsto the last decimal place of the value before the parentheses);Col.6,Swift XRT net count rate in counts s and its uncertainty;Col. 7, 0.3–10 keV unabsorbed flux derived from fitting Swift XRT datawith the power- law model (datasets 1–2 and 3–6 are combined to increase the photon statistics); and Col.8,X-ray spectralindex ( ph X ray)as defined in Sect.1.

Imaging Multi-ObjectSpectrograph (DEIMOS)(Faber et al. 2003) on 2013 April 07 with 1 seeing. Two 600 sinte- grations were obtainedusing the 600 lines per mm (7500 Å blaze11) grating, providing coverage in the 4450–9635Årange with a 100 Å gapbetween the two CCDs. With the 1 0 slit, the spectra havean e ectiveresolution of 3 0 Å. Conditions were good, but not completely photometric;the flux scale might be uncertainbyroughly afactor of 2. Moreover, we obtained 2 180 s g-and R-band images of the object with the two cam- eras on theKeck ILow Resolution Imaging Spectrometer (LRIS) (Okeet al. 1995) on May10under 1 seeing, asshown in Fig- ure 2. TheDEIMOS slit on April 07 was placedonthe bright core of the source, atthe parallacticangle(Filippenko 1982) of Fig. 1. NordicOptical Telescope R-band image of theGB 1310 487 PA 143 (measured from N to E), and theextended wings of region obtained on 2011 May 29. Theexposure timewas 300 s. the host were also included. Thecompanion was 3 o the North is up and east is to the left. TheAGNand comparison star slit. SDSS J131240.83 482842.9 used in theNOTdataanalysis are marked A second Keck II DEIMOSspectrum was obtainedon with the letters “Q” and “C,”respectively. June 10 with a di erent slit position and 0 7 seeing. It has a higher signal-to-noiseratio than the first DEIMOSspectrum; however, it was a ectedbyacosmic-rayhit that prevented accu- theHOWPolinstrument (Kawabataet al. 2008) in the nonpo- rate measurement ofH in the z 0 638 system, and conditions larimetric modeforeight nights during the first and second were not photometricwhen the standard star was being observed. -ray flares. Relative point-spread function (PSF) photometry The two spectra are normalized to epoch1(April 07) using the h m s was conductedusing 3UC 277-116569 ( J2000 13 12 54 09, [O II] 3727 lineat z 0 500. Thecontinuum cannot be used to J2000 48 27 58 2, J2000; R 16 109, I 15 657 cross-calibrate the two spectra because of the significantlyvari- mag; Zacharias et al. 2010)asthecomparison star. Theadopted ableAGNflux contribution;the second-epoch continuumlevel Galacticextinction values were AR 0 035 and AI 0 025 mag appearstohave dropped relative to theemission lines by 1 3. (Schlegel et al. 1998). Thecalibration by Bessell et al. (1998) The two spectra were averaged forfurther analysis. was employed for the magnitude-to-flux conversion. The source GB 1310 487 was assigned aredshift of 0.501 basedona 2007 March 21 1200 s Hobby-Eberly Telescope 2.5. Infrared photometry 10 Low Resolution Spectrograph (HET LRS) observation , which Observationsin the near-IR were carriedout with the 2.1 mtele- showed strong [O II] 3727 at 5592 Åand weak evidence of host scope of theGuillermo Haro Observatory, INAOE, Mexico. The absorption features (Healey et al. 2008;Shaw et al. 2012). This telescope is equipped with the CANICAcamera together with J, spectrum had insu cientsignal-to-noiseratio (S N) to exclude H, and Ks filters. We carriedout di erential photometrybetween weakbroad lines,or to cleanly measure the optical continuum, the object of interest and other objects in the 5 5 field. The leaving the nature of the source uncertain. observationsshowed an increase ofabout one magnitude dur- Thus, we reobserved the source with theKeck10mtele- ing the Flare 2period with respectto the Flare 1 and post-flare scopes. Long-slit spectra were obtained with theKeck II DEep periods (results are summarized in Table 3). Magnitudes are re-

10 Falco et al. (1998) previously reported z 0 313, butitwasindicated 11 The“blaze wavelength” is thewavelength forwhich the grating is as a“marginalmeasurement.” the most e cient.

Article number,page 5of 18 A A proofs: manuscript no. 1310

Table 3. Ground-based photometryofGB 1310 487. LRIS g LRIS R Date JD(UTC) Filter mag Instrument

33.0 33.0 2455... Flare 1period 2009-11-28 164.33668 R 20.58(2)Kanata 30.0 30.0

: : 2009-11-29 165.33620 R 20.86(9)Kanata 28 28 : : 2009-12-05 171.31613 R 20.61(8)Kanata 48 48 2009-11-28 164.34442 I 19.41(2)Kanata 2009-11-29 165.35134 I 19.66(6)Kanata 43.7 13:12:43.4 43.2 43.7 13:12:43.4 43.2 2009-12-05 171.31613 I 18.91 Kanata 2009-12-13 179.33910 I 19.52(1)Kanata 2009-11-22 158.04390 H 15.87(6)OAGH interflare period 2010-03-17 272.95922 H 15.97(5)OAGH Flare 2period 2010-07-07 385.44392 U 21.91(6)NOT 2010-07-07 385.47325 B 22.6(1)NOT 2010-07-07 385.43551 V 21.62(5)NOT 2010-06-03 351.03843 R 20.78 Kanata 2010-06-04 352.05203 R 20.52 Kanata 2010-06-05 353.02920 R 20.64 Kanata 2010-07-07 385.47773 R 20.85(2)NOT Fig. 2. Keck I LRISimages ofGB 1310 487 obtained on 2013 May 10. 2010-07-17 395.07192 R 19.90 Kanata Top panel: the g-and R-band images (8 8 fieldof view; north is 2010-07-19 396.99981 R 20.48 Kanata up and east is to the left). Bottom panel: g-and R-band images after 2010-06-03 351.04981 I 19.98(7)Kanata subtraction ofascaledpointsource fromthe o set core. The boxes 2010-06-05 353.04057 I 19.80(2)Kanata show the location of theDEIMOS slit in the 2013 April 07 observations. 2010-07-07 385.46087 I 19.79(2)NOT Thecirclesindicate the location of theradio-loud AGN. Theseresidual 2010-07-17 395.08708 I 19.27 Kanata images reveal arelatively undisturbed foreground galaxy. 2010-06-15 362.70970 J 16.4(1) OAGH 2010-06-16 363.75472 J 16.6(1) OAGH 2010-06-18 365.69752 J 16.3(1) OAGH ferred to the 2MASS12 survey published photometry. The source 2010-06-19 366.71342 J 15.7(1) OAGH GB 1310 487 itselfwas not detected in the 2MASS survey. The 2010-05-17 333.79362 H 14.59(1) OAGH surveydetection limit is J 15 8, H 15 1, and K 14 3 mag s 2010-05-20 336.83470 H 15.11(5) OAGH (Skrutskieet al. 2006). These valuesmaybeconsidered upper 2010-06-15 362.70186 H 14.61(7) OAGH limits on the object brightness atthe 2MASS observation epoch 2010-06-16 363.74752 H 14.68(3) OAGH of JD 2451248.8408 (1999 March11). 2010-06-19 366.69745 H 14.84(5) OAGH The source GB 1310 487 is listed in theWide-field In- 2010-06-15 362.72145 K 13.8(1) OAGH frared Survey Explorer (WISE; Wright et al. 2010)catalogue s 2010-06-16 363.76233 K 13.78(8) OAGH (Cutri et al. 2012)with thefollowing magnitudesin thefour s 2010-06-19 366.72800 K 13.7(1) OAGH WISE bands: 3.4 m W1 12 302 0 024, 4.6 m W2 s post-flare period 11 254 0 021, 12 m W3 8 596 0 021, and 22 m W4 2011-05-29 711.46544 V 21.56(4)NOT 6 368 0 044. TheIR colors (W1–W2 1 048 0 032, W2– 2011-05-29 711.46135 R 20.79(2)NOT W3 2 658 0 030 mag)areatthe blueedge of theareain 2011-05-29 711.46922 I 19.87(3)NOT thecolor–color diagram occupiedbyblazars and Seyfert galax- 2011-07-31 773.69450 H 15.83(7) OAGH ies (see Fig. 12 in Wright et al. 2010, Fig. 1 in D’Abrusco et al. Columndesignation: Cols.1,2,theGregorian and Julian Date 2012, and Fig. 1 in Massaro et al. 2011), indicating thattheAGN of observation, respectively;Col.3,filter;Col.4,magnitudeand and notthe host galaxy’sstars orwarm dust is responsiblefor its uncertainty; and Col.5,telescope name. most of theIR flux in these bands. WISE observations of this area were conducted on 2010 June 3–8 during Flare 2 (Table 1). vations are performed with a dual-beam (each2.5 full width at 2.6. Radioobservations half-maximumintensity, FWHM)Dicke-switched system using cold sky in the o -source beam asthereference. Additionally, As part ofan ongoing blazar monitoring program, the Owens the source is switchedbetweenbeams to reduce atmospheric Valley Radio Observatory (OVRO) 40 m radio telescope has variations. Theabsolute flux-density scale is calibratedusing ob- observed GB 1310 487 at 15 GHz regularly since theend of servations of 3C 286, adopting the flux density (3.44 Jy)from 2007 (Richards et al. 2011). This monitoring programstudies Baars et al. (1977). This results in a 5%absolute flux-density- over 1500 known and likely -ray-loud blazars, including all scale uncertainty, which is not reflected in the plotted errors. CGRaBS (Healey et al. 2008) sources northof declination 20 . The objects in this program are observed in totalintensity twice Multifrequency radioobservations ofGB 1310 487 were performed with the 100 m E elsberg telescope operatedbythe per week. The minimummeasurement uncertainty is 4 mJy 13 while the typical uncertainty is 3% of the measured flux. Obser- MPIfR . Observations were conducted on 2009 December 1

12 13 Max-Planck-Institut für Radioastronomie

Article number,page 6of 18 K. V. Sokolovsky et al.: Two active states of the narrow-line gamma-ray-loud AGN GB 1310 487 following thereported -ray flare, on 2010 June 28 during the 1e-09 Pre-flare second -ray active phase, and on 2011 June 5. Secondary fo- Flare 1 cus heterodyne receivers operating at 2.64, 4.85, 8.35, 10.45, Inter-flare Flare 2 and 14.60 GHz were used. The observations were conducted ) Post-flare 1 - s

with cross-scans (i.e., the telescope’s responsewasmeasured 2 - 1e-10

while slewing over the source position in azimuth and eleva- m c

tion). The measurements were corrected for(a)pointing o - g r e sets, (b)atmospheric opacity, and (c) elevation-dependent gain (

x u

(see Fuhrmann et al. 2008; Angelakis et al. 2009). The multifre- l f y

quencyobservations were completed within1hrfor eachobserv- g

r 1e-11

ing session. Theabsolute flux-density calibration was done by ne observing standard calibratorssuch as,, 3C 161, 3C 286, E 3C 295, and NGC 7027 (Baars et al. 1977; Ott et al. 1994). TheIRAM 30 mPico Veleta telescope observations at 86.24 and 142.33 GHz took place on 2009 December 7. The observa- 1e-12 tions and data-reduction strategy were similar to thosewith Ef- 0.1 1 10 felsberg; a detaileddescription is givenbyNestoras et al. (2014). Photon energy (GeV) Both theE elsberg 100 m and IRAM 30 mtelescope observa- Fig. 3. Fermi LAT -ray SEDs compared to the power-law models tions were conducted in theframeworkof the F-GAMMA pro- (shown aslinesin this logarithmic plot) derived from unbinned like- gram (Fuhrmann et al. 2007; Angelakis et al. 2008, 2010, 2012; lihood analysis. We stress thatthe power-law models are not fits to the Fuhrmann et al. 2014). binned energy flux values; rather, thesearetwo independent waysto Forcomparison with the latest E elsberg, IRAM, and represent Fermi LAT photon data. The time intervals defined in Table 1 OVROresults, we use datafromthe RATAN-600 576 m are color labeled. ring radio telescope of the Special Astrophysical Observatory (Russian Academyof Sciences);RATAN-600 observations of source with the lower-frequency counterpart. See Tables 1, 2, 3, GB 1310 487 were performed in transit modeatthe southern and the discussion belowfor details. sectorwith the flat reflector quasi-simultaneously at 3.9, 7.7, 11.1, and 21.7 GHz in June 2003 within the frameworkof the The -ray spectra ofGB 1310 487 at various activity states spectralsurvey conductedbyKovalev et al. (1999b, 2002). The listed in Table 1 are presented in Figure 3. The plotted spectral flux-density scale is set using calibrators listedbyBaars et al. binssatisfy thefollowing requirements: TS 50 and or model- (1977); Ott et al. (1994). Details on the RATAN-600 obser- predictednumber of source photons N 8. The flux valueat vations and data processing are discussedbyKovalev et al. eachbin was computedbyfitting the model with source position (1999a). and power-law photon index fixed to the values estimatedover TheNationalRadio Astronomy Observatory’s Very Long theentire period in the 0.1–100 GeV energy range. TheGalactic Baseline Array (VLBA, Napier 1994, 1995) is a system of ten component parameters were fixed, as were all other nontarget 25 m radio telescopes dedicated tovery long baseline interfer- source components of the model. ometry (VLBI) observations forastrophysics, astrometry, and To test if the power law (PL) is an adequateapproxima- geodesy. After publication of thereport on theNovember 2009 tion of the observed -ray spectrum, the PL fit (represented -ray flare (Sokolovsky et al. 2009), GB 1310 487 was added by a straightline if plottedona logarithmic scale) was com- to the MOJAV E 14 program (Lister et al. 2009a). Three epochs of pared to the fit with a log-parabola(LP)function defined as ( ln(E E )) VLBA observations at 15 GHz were obtainedbetween 2009 and dN dE N0(E E0) 0 , where N is the number of pho- 2010. tons with energy E, N0 is the normalization coe cient, E0 is a reference energy, is the spectralslopeat energy E0, and is thecurvature parameter around the peak. For thecombined 3. Results 33 month Fermi LAT dataset, the fit with theassumption ofa 3.1. -ray analysis PL spectrum for the targetsource provides its detection with theTest StatisticTSPL 4415, while theLPspectrumleads The -ray counterpart ofGB 1310 487 waslocalized, inte- to TSLP 4423. These valuesmaybecomparedbydefining, grating 33 months of Fermi LAT monitoring data, to J2000 following Nolan et al. (2012)and in analogy with the source 198 187, J2000 48 472, with a 68% uncertaintyof 0 014 detection TS described in Sect.2.1,thecurvature Test Statistic 50 . This is afactor of five larger than the spacecraft align- TScurve 2(ln LLP ln LPL) TSLP TSPL. The obtainedvalue of ment accuracyof 10 0 003 (Nolan et al. 2012). The -ray TScurve 8 correspondsto a 2 8 di erence, which is lower than position is only0005 18 away fromtheradioposition theTScurve 16 (4 ) threshold appliedbyNolan et al. (2012). ofGB 1310 487. Within the Fermi LAT errorcircle no other We conclude that while there is a hint of spectral curvature, it radio sources are seen with theVLAFIRST 1.4 GHz survey cannot beconsidered significant. (Whiteet al. 1997), whichprovidesthe best combination of sen- The brokenpower-law (BPL) model was also tested, butit sitivity and angular resolution for that region of theradio sky didnot provideastatistically significantimprovement over the todate. Therefore, the positional association of the -ray source PL or theLPmodels (TSBPL 4423, for the best-fit break energy with theradio source GB 1310 487 is firmly established. TheX- Eb 3 GeV, the photon indexes ph 1 2 30 0 04, ph 1 raybrightening observedduring the first and, to a lesser extent, 0 05 0 02 aboveand below the break, respectively). Therefore, the second -ray flarestogether with the near-IR brightening dur- we adoptthe simpler PL model for thefollowing analysis. ing the second -ray flare supportthe identification of the -ray The 33-month -ray lightcurve of the source obtained with 14 Monitoring Of Jets in Active galactic nuclei with VLBAExperi- the seven-daybinning is presented in Figure 4. Two major flar- ments, ing periods are clearlyvisible. The first,brighter flare peaked

Article number,page 7of 18 A A proofs: manuscript no. 1310 around 2009 November 27 (JD 2455163)with theweekly av- 2009 2009.5 2010 2010.5 2011 6 2 1 eraged flux of(1 4 0 1) 10 photonscm s . The peak Fermi 5

) 2 Fermi TS>25

flux averagedover the two-day interval centeredonthat date -1 Fermi 2σ upper limit 6 2 1 s is (1 9 0 2) 10 photonscm s . The source continued to -2 AGILE 4-day be observed at a daily flux of 0 5 10 6 photons cm 2 s 1 1.5

foranother two weeks. The second flare peaked around 2010 ph cm June 17 (JD 2455365)atthe seven-day integrated flux of -6

(0 54 0 07) 10 6 photons cm 2 s 1. The daily flux of (10 0 5 10 6 photonscm 2 s 1 was observed foraboutthree weeks 1 around this date. The two flares demonstrateremarkably con- trasting flux evolution:the first is characterizedbyafast rise GeV 0.1-100 0.5 and slower decay, while the second flare shows a gradual flux risefollowedbya sharpdecay. Following Burbidgeet al. (1974), 7-day flux 7-day Valtaojaet al. (1999), and Gorshkov et al. (2008), we define the 0 flux-variability timescaleas tvar t ln S , where ln S is the di erence in logarithm of the photon flux attwo epochs 4700 5000 5300 5600 separatedbythe time interval t. The observed flux-variability JD-2450000.0 timescale during the onset of Flare 1, as estimated fromthe Fig. 4. Weeklybinned Fermi LAT lightcurve. Blue filled circles are val- seven-daybinned lightcurve(Fig. 4), is tvar 3days. The ues with TS 25, gray filled circles are values with5 TS 25, and timescale of flux decay after Flare 2 is tvar 5days. gray arrowsindicate 2 upper limits for time bins withnosignificant The Fermi LAT lightcurveconstructed with thealternative detections (TS 5). Afour-day integrated AGILE data pointisadded analysis method, theadaptive binning (with25% flux uncer- as anopen box forcomparison. tainty at eachbin), is presented in Figure 5. It confirms all the features visible in theconstant bin-width lightcurve,but also al- 2009 2009.5 2010 2010.5 2011 lows usto investigatefast variabilityduring high-flux statesin 2.5 greater detail. The first flare episode, Flare 1, consists offour 2.0 ) 1 prominentsubflares, each with a timewidthofadayor less. - s

2.0 2

The subflaresshow no obvious asymmetry and the variability - m timescale t for the three point rise of the second and third c var 1.5 1.5

subflares (atJD 2455157.5 and 2455161.5)was estimated as ph 6 0 36 0 20 days and 0 45 0 23 days, respectively. A second -

10 1.0 (

adaptivelybinned lightcurvewas produced in thereverse-time V 1.0 direction, whichgives a similar,but notidentical, time binning. Ge 0.5 200 200 Theresult of the timescaleestimates for this second version of - the lightcurvewas found tobeconsistent with the first analy- 0.5 0.0 0.283 x sis. A similar analysis for the lightcurves with15% uncertainties u 5152 5159 5166 give timescaleestimates ofabout 1day for the most rapidvari- Fl ability. We conclude thattheadaptivelybinned lightcurvesshow 0.0 evidence ofavariability timescale of halfaday with aconserva- 4700 5000 5300 5600 tive upper limit of 1day. For the second and fainter flare epoch JD-2450000.0 the timescalesseen in theadaptive binning are consistent with theestimatefromthe fixed-binned lightcurve described above. Fig. 5. Fermi LAT lightcurveconstructed with theadaptive binning method (Lott et al. 2012). The magnifiedplot of Flare 1 is shown in the Table 1presents spectral analysis results for the di erent - insert. Theenergy rangefor this lightcurve is chosen to minimize the ray activity states of the source: “pre-flare” and “post-flare” peri- uncertaintiesin timeand flux, while the lightcurvewasFig. 4 is given ods representthe low-activity level, “Flare 1”and “Flare 2”rep- in thecommonlyused E 100 MeV energy range. resentthe high-activity state, while during the“interflare” inter- valthe source showed an intermediate -ray flux level. Figure 3 presents the observed Fermi LAT spectrum during these states. 12:00 UT, JD 2454587.0–2454597.0)weobtained a 2 upper Significant evolution of the -ray photon index, ph, is detected 6 2 1 limit of FE 100 MeV 0 31 10 photons cm s . between the di erent flux states (Table 1). Figure 6presents ph as afunction of(E 100 MeV) flux. The harder-when-brighter trend is clearlyvisible. 3.2. X-ray to infrared spectrum Integrating the AGILE observations from 2009 Novem- ber 18 12:00 (JD 2455154.0) to 2009 November 22 12:00 UT Results of theX-ray spectral analysis are presented in Table 2. (JD 2455158.0), we obtain a -ray flux FE 100 MeV (1 65 The obtainedvalues of the X-ray photon index ph X ray are 0 48) 10 6 photons cm 2 s 1, at a significance of TS 6 1. among the hardest reported for blazars (Giommi et al. 2002; This result is in good agreement with the flux value derived Donato et al. 2005;Sikora et al. 2009). Radio-loud NLSy1 have fromthe preliminary analysis by Bulgarelli et al. (2009). Prior ph X ray similar to the ones found inblazars (Paliyaet al. 2013; to the Fermi launch, AGILE observed GB 1310 487 (inpointing Abdo et al. 2009d). However, it cannot beexcluded thattheX- mode) during twoother periods,but didnot detectthe source. ray spectrum with an intrinsic value of ph X ray is artificially During the first period (from 2007 October 24 12:00 UT to 2007 hardenedbyadditional absorbing material along the line of sight November 112:00 UT, JD 2454398.0–2454406.0), the 2 upper (see the discussion ofNOTimaging results below). Future high- 6 2 1 limit was FE 100 MeV 0 28 10 photons cm s , while in quality X-rayobservations are necessary for investigating this the second period (from 2008 April 30 12:00 UT to 2008 May10 possibility.

Article number,page 8of 18 K. V. Sokolovsky et al.: Two active states of the narrow-line gamma-ray-loud AGN GB 1310 487

2.6

2.5

2.4 Pre-flare Post-flare 2.3 ph Γ 2.2 Inter-Flare Flare 2 2.1

2 Flare 1

1.9 0.1 1 10 Photon flux (E>100 MeV) 10-7 cm-2 s-1

Fig. 6. The -ray photon index, ph, as afunction of flux for the time periods defined in Table 1.

TheUVand blue parts of the opticalspectrum are flat, in Fig. 7. Keck II DEIMOSspectrum ofGB 1310 487. The twonarrow- contrast to the steep spectrumseen in the near-IR (iJHK bands). line systems are indicated. Theregion 6900–7000 Å is lost to a gap Also, the observedvariability amplitude is decreasing toward between the two CCDs. The slit position is shown in Figure 2. bluer wavelengths, with the U-band brightness being essentially constant. Table 4. GB 1310 487 emission-line strengths. The H-band flux showed an increase ofabout one magnitude a b a b during the Flare 2period with respectto the Flare 1 and post- Species Rest Flux0 500 EW0 500 Flux0 638 EW0 638 flare periods, in contrast to the behavior seen inother bands. The [O II] 3726 3729 4 24 0 07 32 6097 0 06 4 2 0 5 results of ground-based photometric measurements are summa- H 4340 0 65 0 05 3 1 0 7 cosmic-rayhit rized in Table 3. H 4861 1 69 0 06 6 6 0 8024 0 06 4 1 0 7 [O III] 4959 not detected100 0 06 3 6 0 8 [O III] 5007 1 11 0 05 6 5 0 6261 0 05 12 0 2 2 3.3. Imaging with NOT a Fluxes are inunits of 10 17 erg cm 2 s 1, with1 statistical errors. The overall flux scale is uncertainbyupto afactor of 2, TheNordicOptical Telescope images (Fig. 1) showafuzzy buttherelative fluxes are much more accurate. b Theequivalent extendedobject,probably a galaxy, with a pointsource o set width (EW) in angstroms. 0 6 from its center. Considering thatthere are many galaxies ofcomparable brightness visible in the field, this picture may be interpreted astheAGN(corresponding to the pointsource) 3.4. Keck imaging and spectroscopy shining through anunrelated foreground galaxy. This maybe the source ofconfusion in the AGN’s redshift determination Standard reductions, extractions, and calibrations of the (Sokolovsky et al. 2009; Healey et al. 2008;Falco et al. 1998), DEIMOS data produced the spectrumshown in Figure 7;itisthe and it also explainsthe steepness of the optical-IRSED (the ob- average of twoobservations conductedonApril 07 and June 10, served SED was corrected for Milky Way absorption, but ab- 2013. The strongest line, [O II] 3727, confirms theHETred- sorption in the intervening galaxy may alsobe significant). On shiftidentification at z 0 500, and we also see [O III] and the other hand, it is not uncommon forAGNhost galaxiesto narrow Balmer emission for this system. However, there are ad- have disturbed morphologies, making it appear thattheAGNis ditionallines, mostly in theredhalf. Theserepresent a second o -center. system withnarrowforbidden and Balmer emission, this timeat z 0 638. The line strengths are given in Table 4. The [O II] dou- The galaxy contributes a large fraction of the total optical blets are barely resolved, butthe oxygen and Balmer linewidths flux, when the pointsource is in the low state. If the host galaxy are consistent with the instrumental resolution. The [O III] emis- ofGB 1310 487 is similar to the giant ellipticals studiedby sion at z 0 638 appears resolved with a deconvolved width 1 Sbarufatti et al. (2005), MR 22 9 0 5 mag, its magnitude of 200 kms . Unfortunately, thered limit of the spectrum at z 0 500 shouldbe R 20 (or 0 6 mag fainter at z 0 638; does notinclude theH [N II] lines foreither system. For the Sect.3.4). A typical NLSy1 fromtheVéron-Cetty & Véron z 0 500 system, we cover [O I] 6300, which is weakorab- (2010)catalogue having MV 21 4 mag would appear 1 mag sent. For the z 0 638 system, we cover Mg II 2800, and can fainter than a giant ellipticalin the R band assuming V R place a 3 rest equivalent width limit of 1 0 Å on any broad 0 5 mag (Xanthopoulos 1996). Therefore, the observedgalaxy emission;theH line is marginallydetected for the z 0 638 couldbe the host ofGB 1310 487. The visible o set between system at 4 level. TheratioofH to [O III] (z 0 638) is the pointsource and thecenter ofextended emission could re- small, even if theH flux is treated as an upper limit. Together sult fromthe disturbed morphology of the host, as noted above. with theresolved [O III] this indicates nuclear excitation.

Article number,page 9of 18 A A proofs: manuscript no. 1310

Thus, we clearlyhave two superimposed systems and wish Table 5. Multifrequency radioobservations ofGB 1310 487. to identify which system hosts theradio-loud core (and, by infer- ence, the -ray source). TheKeck ILRISimages confirmthe ba- F F sic structure seen in theNOTimages; the source is extended with (GHz) (Jy)(Jy) (GHz) (Jy)(Jy) a brighter core displaced 0 6 to thewest. Figure 2 shows 8 RATAN-600 2003-06 E elsberg 100 m 2010-06-28 regions around theAGN. TheDEIMOSslitposition on April 07 21.74 0.401 0.104 2.64 0.135 0.002 is markedonthe g frames (left). Atthe bottom we show the im- 11.11 0.222 0.012 4.85 0.107 0.001 ages after removal ofapoint-source PSF (g 23 89, R 22 45 7.69 0.157 0.019 8.35 0.111 0.002 mag)fromthe o set core. Theresiduals showarelatively regu- 3.95 0.221 0.057 10.45 0.112 0.005 lar galaxy having FWHM 1 8, with g 21 95 and R 20 59 E elsberg 100 m 2009-12-01 E elsberg 100 m 2011-06-05 mag. Thecoordinate system was referenced through the SDSS 2.64 0.161 0.001 2.64 0.201 0.006 image of the field, with an estimateduncertainty relative to the 4.85 0.133 0.001 4.85 0.189 0.003 radio frame of 0 2;thecirclesshow the position of theVLBI 8.35 0.130 0.002 8.35 0.207 0.005 source (Sect.1)and haveradii twice this uncertainty. Hence, the 10.45 0.130 0.003 10.45 0.213 0.010 radio source is coincident with the point-like peakof thecom- 14.60 0.121 0.006 bined source. We also find thatthe z 0 500 emission lines are IRAM 30 m 2009-12-07 o set 0 21 0 06 SEalong the slit fromthe z 0 638 system, 86.24 0.282 0.070 toward thecontinuumtail representing theextendedgalaxy. The 142.33 0.206 0.065 deprojectedo setis 0 35 E of theAGNcore. We thus con- Columndesignation: Col.1,thecentral observing frequency; clude thatthe trueAGNredshiftisz 0 638, and we are view- and Cols.2,3,the observed flux density and its uncertainty, re- ing it through an approximately face-on galaxy showing strong spectively. narrow-lineemission. Ourextracted spectrumisweighted toward theAGNcore, 0.35 although it also contains appreciable fromtheforeground OVRO Effelsberg galaxy. Both spectra are dominatedbynarrowforbidden lines, VLBA core yetthere is appreciablecontinuum associated withboth compo- 0.30 nents as well. Theforeground galaxy is probablynot an AGN, but we cannot be certain; withoutthe [N II] H lineratio, we 0.25 are unable to fullydistinguish “LINER”(LowIonization Nu- )

clear Emission-line Region)emission from an HIIregion (e.g., Jy (

0.20 ν

Ho et al. 1997). However, the strong [O II] 3727 and lackof ob- F vious [O I] 6300 argueagainst a power-law ionizing spectrum, suggesting thatthe z 0 500 emission represents star formation 0.15 in theforeground galaxy lacking AGN activity. The positional accuracyof theavailable observations of mul- 0.10 tiwavelengthvariability (Sect.2) is notsu cienttodistinguish between theforeground and background objects discussedhere 0.05 and in Sect.3.3asthe source of high-energy emission. The pro- 4400 4600 4800 5000 5200 5400 5600 5800 6000 6200 posed interpretation thatthe background AGN is the high-energy JD - 2450000.0 source rests on theconsideration thatthe observed fast -ray Fig. 8. Radio lightcurveat 15 GHz obtained with theOVRO 40 m variability (Sect.3.1) is typical ofradio-loud AGNs (which the telescope(points) supplemented with two 14.6 GHz measurements ob- background source is), while there are no firmindications of tained with theE elsberg 100 mtelescope(square). VLBA measure- AGN activity in theforeground galaxy. ments of thecore (componentC0, Table 6)areindicated as diamonds. The two arrowsmark the peaks of the -ray flares observedbyFermi.

3.5. Results of radioobservations The SWcomponentincreased its brightness during the three Theradio spectrum ofGB 1310 487 is generally flat, with a epochs. Comparison with the lightcurve in Figure 8 showsthat wide peak locatedbetween22GHz and 86 GHz (Table 5). The this componentisresponsiblefor most of the flux observed with variability amplitudeat 2.64 GHz is slightly lower compared to single-dish instruments. Thefainter componentlocated to the higher frequencies. The 15 GHz lightcurve ofGB 1310 487 ob- NE is gradually fading. If the SWcomponentisthe 15 GHz core, tained with theOVRO 40 mtelescopeand complementedby the position of the second component aligns nicely with the ori- measurements with theE elsberg 100 m and theVLBA is pre- entation of the kiloparsec-scale jet observed with theVLAat sented in Figure 8. Itshows a period of high activity with two 1.4 GHz by Machalski&Condon (1983). No significant proper separate peaksthatstarted in mid-2010 and is still ongoing. motion couldbe detectedbetween the three 15 GHz MOJAV E The 15 GHz VLBA images (Fig. 9) show two emission re- epochs. The 3 upper limit which canbe placedonproper mo- 1 gionsseparatedby 0 4 mas. Toquantify their parameters we tion is 0 3 mas yr , corresponding to app 11 ( app is in fitthe observedvisibilities with a model consisting of two circu- units of the speedof light)atthe source redshift, which is within lar Gaussian components using the software (Shepherd therange ofapparentjetspeeds occupiedby -ray-bright blazars 1997). The modeling results are presented in Table 6. The uncer- (Lister et al. 2009b;Savolainen et al. 2010). The projected linear taintiesinparameters of the model components were estimated size of the double structure resolved with theVLBA is 2 7pc following Lee et al. (2008), and theresolution limit achieved for 8 1018 cm. The overall 15 GHz VLBI of the each component was computed following Lobanov (2005)and source measuredbyMOJAV E is 2.8–6.5%which is indicative of Kovalev et al. (2005). beamedblazar emission. Theweaklybeamed, high viewing an-

Article number,page 10 of 18 K. V. Sokolovsky et al.: Two active states of the narrow-line gamma-ray-loud AGN GB 1310 487

shouldbe less than10%. Fast variability within the Flare 2pe- riod may also contribute to the uncertainty. The value of q is probably larger for Flare 1 than for Flare 2, judging fromthe lower near-IR flux observedduring Flare 1.

4. Discussion 4.1. -ray luminosity, variability, and spectrum The monochromatic -ray energy flux averagedover the du- ration of the first flare is F 10 10 erg cm 2 s 1. Atthe redshift of the source this correspondsto an isotropic lumi- nosityof 1047 erg s 1. Considering theexpectedbolomet- riccorrection ofafactor ofafew, the -ray luminosityof GB 1310 487 is comparable to thattypicallyobserved in flar- ing -rayblazars (e.g., Tanakaet al. 2011; Abdo et al. 2010a,e) and NLSy1 (D’Ammando et al. 2013). Itisabouttwoorders of magnitude lower than the most extremeGeVflares of 3C 454.3 in November 2010 (Abdo et al. 2011b)and PKS 1622–297 in June 1995 (Mattox et al. 1997). The outstanding -ray flare of 3C 120 in November 1968 had acomparable isotropic lumi- nosityof 1047 erg s 1 (Volobuev et al. 1972). Theexceptional GeV photon flux of 3C 120 was due to therelative proximity Fig. 9. VLBAradio image ofGB 1310 487 obtained on 2010-12-24 of the source (z 0 033;Michel&Huchra 1988)compared to at 15 GHz during thecourse of the MOJAV E program. The image map the brightest -rayblazarsmentioned above. The observed large peak is 0.206 Jybeam 1 and the first contour is 0.15 mJybeam 1. Ad- -ray luminosityofGB 1310 487 is an indirectindication ofa jacent contour levels are separatedbyafactor of 2. Naturally weighted high Doppler boosting factor of the source (Tayloret al. 2007; beamsize is indicated atthe lower-left corner of the image. Green cir- Pushkarev et al. 2009). clesindicate positions and best-fitsizes of the model components pre- The di erence in lightcurve shape,overall duration, and sented in Table 6. shortest observedvariability timescale between the two flares of the source may indicate thattheyoccurred indi erentjet regions Table 6. Parsec-scalecomponents observed at 15 GHz. orwerepoweredbydi erent emission mechanisms as discussed below. Inboth cases, this implies di erencesin theemitting- Comp. Distance FWHMFlux density Tb plasma parameters for the two flares, such astheelectron en- (mas)(mas)(Jy)(K)ergy distribution, magnetic field strength, bulk Lorentzfactor,or 2009-12-26 JD 2455192 9 external photon field strength. Variability timescales of 3days C0 0 39 0 090 0 021 3 10 and shorter are common in GeV blazars (e.g., Mattox et al. 8 C1032 0 11 0 42 0 014 0 004 4 10 1997; Abdo et al. 2010f, 2011b;Sbarrato et al. 2011). The light- 2010-08-06 JD 2455415 travel-timeargument limits the -ray emitting region size r 9 15 C0 0 36 0 144 0 033 6 10 c tvar (1 z) afew 10 cm, where c is the speedof light C1036 0 10 0 40 0 008 0 002 3 108 invacuum, z is the source redshift, and conservatively assuming 2010-12-23 JD 2455554 theDoppler factor [ (1 cos )] 1 afew(wehave no ev- C0 0 40 0 211 0 049 7 109 idence ofextremeDoppler boosting from VLBIand -raydata; C1043 0 12 0 52 0 004 0 002 8 107 Sect.4.2), where is theLorentzfactor, is the bulkvelocityof Columndesignation: Col.1,component name, where C0 is the theemitting blob inunits of the speedof light, and is theangle presumed core and C1 is the decaying jet component; Col.2, between the blob velocity and the line of sight. projecteddistance fromthecore (C0);Col.3,FWHM of the Itisimportantto check thatthe observedharder-when- Gaussian component; Col.4,component flux density; and Col.5, brighter trend in the -ray spectrumisnot related to theexpected observedbrightness temperature. correlation between the flux and the index in the power–law model. If the number densityof photons arriving fromthe source is dN dE N0(E E0) ph (where N0, and E0 are constants), the gle sourcesin MOJAV E tend tobe unpolarized (Lister & Homan integrated photon flux between energies Emax and Emin is 2005). ! ! N E E ph 1 E ph 1 F 0 0 max min 1 E E 3.6. SED during the two flares ph 0 0 The SED ofGB 1310 487 is presented in Figure 10. It hasthe Assuming that 1 and Emax is large, classicaltwo-humped shapewith the high-energy humpdomi- ! 1 N E E ph nating over the synchrotron humpduring the first (brighter) flare F 0 0 min by a Compton dominance factor of q 10. For the Flare 2period ph 1 E0 the value of Compton dominance maybe measured accurately The derivative thanksto simultaneous observations of Fermi LAT and WISE: " !# q 12. The uncertaintyof this measurement is limitedbythe dF 1 E F ln min , 0 accuracyof theabsolutecalibration of the two instruments and d ph ph 1 E0 Article number,page 11 of 18 A A proofs: manuscript no. 1310

Fig. 10. Quasi-simultaneous radio to -ray SED ofGB 1310 487 during the two flaring episodes and the post-flare period coveredbyour multi- wavelengthobservations. The time intervals corresponding to theseevents are defined in Table 1.

if ln(Emin E0) , 1 ( ph 1). Therange of parameters derived inspection of Figure 3. The peak-frequency evolution is di cult from ouranalysis is 1 97–2 41, Emin 100 MeV, E0 toquantifyowing to the insu cient number ofcollected pho- 283 MeV, and 1 04 ln(Emin E0) 1 ( ph 1) 1 03 to tons, which results in the simple PL fit (withnocurvature) be- 0 70, sodF d ph 0. Theexpected correlation betweendF ing a statistically acceptable model for theLATdata. However, and ph due to their mathematical dependence is positive, which the overall SED (Fig. 10) suggests thatthe high-energy emis- is opposite to whatisactuallyobserved. We conclude thatthe sion peak shouldbe located somewhere around theLATband. observedharder-when-brighter trend is real and not related to We canuse theLATspectralindex ( ph) vs. Compton peak fre- the intrinsiccorrelation of the model parameters. IC IC quency ( peak Hz)correlation log10 peak Hz 4 0 ph 31 6 re- IC Previously a harder-when-brighter trend (Fig. 6) has portedbyAbdo et al. (2010b) to estimate that peak Hz changed been seen at GeV energies only in a handful of blazars: from 1022 to1024 Hz between the pre-flare and Flare 1periods. 3C 273, PKS 1502 106, AO 0235 164, and 4C 21.35 by Thechange in IC may result not from acontinuousshift Fermi (Abdo et al. 2010k,f,g; Tanakaet al. 2011); 3C 454.3 peak Hz ofasingle -ray emission peak, but from achange in relative (Ackermann et al. 2010; Vercelloneet al. 2010; Abdo et al. strengths of two emission components peaking at di erent fre- 2011b;Stern &Poutanen 2011)and PKS 1510 089 by quencies, as discussed in Sect.4.9. Fermi and AGILE (Abdo et al. 2010a; D’Ammando et al. 2011); and 3C 279 (Hartman et al. 2001)and PKS 0528 134 (Mukherjee et al. 1996) by EGRET. The same harder-when- 4.2. Jet Doppler factor brighter behaviorwassuggestedbythecombined analysis of relative spectralindex changeas afunction ofrelative flux TheDoppler factor of therelativistic jetin GB 1310 487 maybe change in afewof the brightest FSRQs, low-, and intermediate- constrainedusing two independentlines ofargument: one based peaked BLLacs using the first six months of Fermi data by on therequirementthattheemitting region shouldbe transparent Abdo et al. (2010g). Thecurrent detection presents one of the to its own radiation (since we observe it), the other basedon clearest examples of this spectral behavior. theabsence ofapparent proper motion seenbytheVLBA. The minimum Doppler factor needed to avoid atten- The spectral evolution during the Flare 1–interflare–Flare 2 uation for -raysinteracting with lower energy photons present periods (Fig. 3 and Table 1) maybe qualitatively understood as inside theemitting region maybecalculatedusing Eq. (39) of the gradual decrease in energy of the -ray emission peak. Dur- Finkeet al. (2008), ing Flare 1, the spectrumishard, implying thatthe spectral peak 1 is located above oraround 5 GeV. Consequently, during the in- 2a 1(1 z)2 2a D2 6 2a T L syn terflare period the spectrumissofter, with a hint ofcurvature; 4 1 f 1 m c t 1 theemission peak maybe located around 1–2 GeV. Later,during e var Flare 2, the spectrumisalso soft with a peakpossibly located at where it is assumed thatthe synchrotron flux is well repre- syn a even lower energies. This interpretation is inspiredbythe visual sentedbya power law of index a ( f ), T is the scat-

Article number,page 12 of 18 K. V. Sokolovsky et al.: Two active states of the narrow-line gamma-ray-loud AGN GB 1310 487 tering Thomson cross-section, DL is the luminositydistance contribution is significantmostly for BL Lac-type blazars and 2 to the source, me is theelectron mass, and 1 E (mec ) is non-blazar AGN. Finally, as discussed above, the nearby galaxy the dimensionless energy ofa -ray photon with energy E for and the star-like object,bothprobablyunrelated to the source un- which the optical depthof theemitting region 1 (see also der investigation, may contribute to its total optical flux ifanob- Dondi&Ghisellini 1995). The maximum energy of observed - servation lacks angular resolution to separatecontributions from ray photonsthat canbeattributed to the source is 10 GeV these objects. 4 4 (Fig. 3), so 1 10 GeV (5 11 10 )GeV 2 10 . This means 1 5 11 10 5, and thecorresponding frequency for 1 4.5. Radioproperties this is 6 3 1015 Hz. Fromthe observed SED (Fig. 10), we esti- syn 14 1 2 mate f 1 10 erg s cm and a 2. Taking tvar 3days Theradio loudness parameter, Rradio,defined astheratioof 1 5 GHz flux density, L5 GHz, to the B-band optical flux density, LB, (Sect.3.1)weobtain 1 5. 4 Assuming theangle between the jet axis and the line of sight, is Rradio L5 GHz LB 10 . This is anorder of magnitude larger than typical Rradio values found inquasars (Kellermann et al. , is max, the one thatmaximizestheapparentspeed, app, fora given intrinsic velocity, , we may estimate thecorresponding 1989)and radio-loud NLSy1 galaxies (Doi et al. 2006),butitis q comparable to the largest observedvalues (Singal et al. 2013)15. 2 Doppler factor VLBA 11 (if VLBA app 1). We note Theextremely low opticalluminosity compared to theradio lu- thatif is smaller than max, theactual will be larger than the minosity may either bean intrinsic propertyof this source,or it aboveestimate(see, e.g., Cohen et al. 2007; Kellermann et al. may result from absorption in the intervening galaxy (Sect.3.3, 2007;Marscher 2009 foradiscussion ofrelativistic kinematics 3.4). in application to VLBI). Theradio spectrum ofGB 1310 487 (Table 5) is typical for RecentRadioAstron (Kardashev et al. 2013) Space–VLBI a blazar. Relatively rapid (timescale of months)and coherent observations of high brightness temperaturesin AGNssuggest changes across thecm band suggest thatmost of the observed thattheactualjet flow speed is oftenhigher than the jet pattern radio emission comes from acompact region no more than a speed (Sokolovsky 2013). Theseresults question theapplicabil- few parsecsin size. Comparison of the 15 GHz lightcurve pre- ityof estimates basedonVLBI kinematics. Theavailable lower sented in Figure 8 with the 15 GHz VLBAresults (Table 6) in- limits on thecore brightness temperature, Tb, in GB 1310 487 dicatesthatthecomponentC0 (presumably thecore) is the one (Table 6)areconsistent withnegligibleDoppler boosting within responsiblefor most of the observed single-dish flux densityof the standard assumption of theequipartition inverse-Compton the source. Specifically, C0 is the site of the majorradio flare 11 limited Tb afew 10 K(Readhead 1994). peaking around JD 2455500 (October–November 2010). The presence ofcorrelation between cm-band radio and - ray emission is firmly established for large samples of blazars 4.3. Black hole mass (Ackermann et al. 2011; Arshakian et al. 2012; Linford et al. 2012; Kovalev 2009). The typical -ray radio time delay If we equate the linear size estimated fromthe shortest ob- ranges from 1 month to8monthsin the observer’s frame, servedvariability timescale to the Schwarzschild radius, the 10 with -raysleading radio emission (Pushkarev et al. 2010; corresponding blackhole mass wouldbe M 10 M . León-Tavares et al. 2011). However, for individualsourcesit However, TeV observations of ultra-fast (timescale of minutes) is oftendi cult to establish a statistically significant corre- variability inblazarsPKS 2155 304 (Aharonian et al. 2007; lation because of the limited time spanof simultaneous - Abramowski et al. 2010)and Mrk 501 (Albert et al. 2007) lead ray–radiodatacompared to a typical duration ofradio flares to M estimatesinconsistent with those obtainedbyother (Max-Moerbeck et al. 2012). This could also limit our knowl- methods (Begelman et al. 2008),unless an extremely large edge of the maximum possibleradio -ray time delay. Doppler factor 100 is assumed for the -ray emitting re- In thecase ofGB 1310 487, no clear connection is visible gion (Ghisellini&Tavecchio 2008;Sbarrato et al. 2011). Short between its radio and -ray activity, basedbothontheavailable timescale variability may arisefromthe interaction of small (size single-dish (Fig. 8)and VLBI monitoring data(Table 6). r rs) objects such asstars (Barkov et al. 2012) or BLR clouds (Araudo et al. 2010)with a broad relativistic jet. This should caution us against putting much trust in theabove M estimate. 4.6. Object classification Shaw et al. (2012)classified the opticalspectrum of 4.4. UV, optical, and IR emission GB 1310 487 as aLINER, which is inconsistent with the -ray and radio loudness (Sect.4.5; Giuricin et al. 1988). The TheUV-to-IR behavior of the source maybe understood if the absence of broad lines precludes classification as a . near-IRlightisdominatedbythe synchrotron radiation of the Prominent forbidden emission lines are nottypical of BL Lac- relativistic jet, while in the optical–UV thecontribution of line type objects. Therefore, while being similar toblazarsin its emission and or thermal emission fromthe accretion disk starts high-energy, radio, and IR properties, GB 1310 487 cannot todominate over the synchrotron radiation. The lineand thermal beclassified as aclassical blazar on the basis of its optical emission are not relativisticallybeamed and, therefore, more sta- spectrum. blecompared to the beamed synchrotron jet emission, decreas- As discussed in Sect.3.4,the point-source emission at z ing the variability amplitude in the parts of the SED where their 0 638 observedbyKeck is most likely related to theAGN. Pogge contribution to the totallightiscomparable to that fromthe jet. (2000) defines NLSy1 as having permitted lines only slightly Thermal emission features are observed in SEDs of many FSRQ- stronger than forbidden lines, [O III] H 3, and FWHM(H ) type blazars (e.g., Villataet al. 2006; Hagen-Thorn et al. 2009; Abdo et al. 2010a; D’Ammando et al. 2011). Starlight fromthe 15 Singal et al. (2013) use therest frame luminosity at 5 GHz and host galaxy also contributesto the total optical flux in some - 2500 Årespectively todefine R. This definition of Rshouldbecon- ray loud AGN (Nilsson et al. 2007; Abdo et al. 2011a,c). This sistent within afactor ofafewwith the oneweuse.

Article number,page 13 of 18 A A proofs: manuscript no. 1310

1 2000 kms . Theanomalously strong [O III] 4959, 5007 distance between the lenslocated at redshift zlens and theAGN emission formallydisqualifiesthis source, and tendsto support located at redshift zAGN, and Mlens 12 is M expressed in the units a Seyfert 2 (oranarrow-lineradiogalaxy, considering the ob- of 1012 M . Theaboveamplification factorestimate involves a ject’s radio loudness)classification, which wouldbe di cult to number of simplificationsincluding (i) the simplified lens ge- understand if theradio-and -ray jets align with theEarth’sline ometry, (ii) use of the observed AGN–lensseparation which is of sight. Our S N is too low to allow unambiguous detection of larger than the true one, and (iii) the Paczynski (1986)formula FeIIemission. Broad H , if present, is weaker than a thirdof the referring to thecombined light of two images (we knowfrom narrowcomponent, and there is no evidence of Mg II 2800 Å. observationsthatthe single observed image ofGB 1310 487 is Thus,nobroad-linecomponentisobserved. We also find that muchbrighter than its second undetected image). Taking into ac- CaH&Kareweak, if present, and the 4000 Å break is smaller countthesecaveats, we estimate thattheAGNimage is probably than0.1.Theseaspects suggest appreciable nonthermallumi- amplifiedbyafactor ofafew. nosity for thecore AGN emission. Thus, we tentatively advance Since no second image is visible inoptical Keck and radio the view thatsynchrotron emission fromtheAGNdominatesthe VLBA(this work)and VLA (JVA Ssurvey; King et al. 1999) variable point-source core,butthat a surrounding narrow line data, we assume that its contribution to the source lightcurveat region dominatesthe line flux. other bands, including GeV, is alsonegligible. Theabsence ofan Gürkan et al. (2014) studied theWISE infrared colors of observable second image may indicate that either the lensisnot radio-loud AGN; GB 1310 487 falls in aregion of thecolor di- massiveand the source is still outside its RE or the mass distri- agram occupied mostlybyquasars and broad-lineradiogalax- bution in the lensisasymmetricand the lensed source is close to ies, although some narrow-lineradiogalaxies are alsopresent. afoldorcusp caustic. If the lensisa singular isothermalsphere The source GB 1310 487 is well away fromthe locus of low- (SIS; e.g., Refsdal&Surdej 1994;Meylan et al. 2006)and the excitation radiogalaxies (LERGs)and alsohas a 22 mluminos- lensed AGN is just outside theEinstein radius defined foran SIS ity ( 4 1045 erg s 1) typical of high-excitation radiogalaxies through the lensing galaxy’s velocitydispersion SIS, (HERGs; see Fig. 8 in Gürkan et al. 2014). However, if one uses 2 thecriteria of Jackson &Rawlings (1997)GB 1310 487 would SIS DAGN RE SIS 4 2 qualify as aLERG basedonits opticalspectrum. We note that c Dlens WISE photometryof theAGNmight becontaminatedbythe foreground galaxy. then a single image is formedhaving, inprinciple, an arbitrarily Being a narrow-lineradio-loud AGN, the objectisnot a largeamplification factor A 1 (1 1 u)(e.g., Wu 1994). Tak- 1 member ofcommon types of -ray flaring extragalactic sources ing RE SIS 0 6 we estimate SIS 140 kms and the mass 10 (blazars and NLSy1s). One possibility is thatthe objectisanal- inside RE SIS of 6 10 M (Fort&Mellier 1994). The low ogoustonearby radiogalaxieslike Per A with additional ampli- lensing galaxy mass,necessary toputtheAGNimage outside RE fication due togravitationallensing thatmakes -ray emission (and form a single image), maybereconciled with its brightness from its core detectableat high redshift. The similarity to Per A if the galaxy is undergoing intensive star formation, asindicated is supportedbyits lackof superluminalmotion (Lister et al. by strong emission linesin its spectrum. 2013), low inferred fromSED modeling (Abdo et al. 2010h), Ingeneral,gravitationallenses producing a single magnified and absence ofchangesin VLBIand single-dish radioproperties image ofadistantsource shouldbe more common than lenses that canbeattributed to GeV events (Nagai et al. 2012). producing multiple images. However, most gravitationallens Another possibility is thatthe objectmaybeabona fide searches (like the JVA S-CLASS survey;Browneet al. 2003)are blazar with its optical non-thermal emission swampedbythe designed to identifyonly multiple-image lenses. The BL Lac- host elliptical as proposedbyGiommi et al. (2013)as possible type object AO 0235 164 is an example ofablazar shining counterparts of unassociated Fermi sources. In this case,how- through an intervening galaxy and having a single imageweakly ever,onewouldnot understand the observedvariable optical amplifiedbymacrolensing (Abraham et al. 1993). pointsource. Higher S N spectroscopy with increased wave- A possibilityof microlensing by individual foreground length coveragewouldbe helpfulin characterizing the z 0 638 in the z 0 5galaxy cannot beexcluded. The timescale of such -ray radio AGN. Higher resolution spatialimaging is needed to microlensing events maybeestimated as aratioof theEinstein probe the nature of theforeground (z 0 500) galaxy. radius forasingle star to the proper motion of the lens and is on the order of tens of years. Therefore, microlensing is probably unrelated to the observed fast high-energy variability, butmay 4.7. Gravitationallensing provideasignificant amplification thatisnearly constant over Considering thattheAGNis locatedbehind the visible diskof the duration of our observations. A largeconstant amplification another galaxy (Sect.3.4,4.6), amplification of theAGNlight by due to microlensing could also explain theabsence of the second gravitationallensing is areal possibility. In the simplified case lensed image. ofapointlens, theAGNlightisamplifiedbyafactor of A (u2 2) (u u2 4)(Paczynski 1986; Griest 1991; Wambsganss 4.8. Emission model constraints from the SED 2006), where u is theratioof theAGNlensing-galaxy separation (0 6) to the lensing galaxy’s Einstein radius, The high-energy SED humpdominates over the synchrotron r humpduring the first (brighter) flare by afactor of q 10. This 4GM D is commonlyobserved in FSRQ-type blazars. In the framework R lens lens to AGN 1 1(M )1 2 E c2 D D lens 12 of the leptonic model, the large q suggests thatmost of the ob- lens AGN served -ray flux during the first flare shouldbeattributed to where G is the gravitational constant, M is the lensing galaxy theEC process rather than to the SSC scenario. An SSC model mass, Dlens 1300 Mpc is theangular size distance to the lens, would require a large deviation from equipartition:theemitting- 2 DAGN DA 1400 Mpc is theangular size distance to the particleenergy density wouldneed tobe q times greater than the AGN, Dlens to AGN DAGN (1 zlens) (1 zAGN)Dlens is the magnetic field energy density in the source frame(Sikora et al.

Article number,page 14 of 18 K. V. Sokolovsky et al.: Two active states of the narrow-line gamma-ray-loud AGN GB 1310 487

2009), which is not expected; ECmodels do not require devia- flux. Large-amplitudeGeVvariability makesthecorresponding tion from equipartition to explain large values of q. spectral changes apparent even with the limited photon statistics The dramatic di erence in spectralslopesin theradio and IR of the observations. regionssuggests the presence ofabreak in theelectron energy Another way to explain the observed changesin the -ray distribution. Atthe post-flare state, GB 1310 487 shows a steep spectrumisa varying contribution frommultipleEC components -ray spectrum. If this spectrumisdominatedbythe SSC emis- (e.g., EC on accretion disk and dusty torus photons). This sce- sion (Sect.4.9), it is possible to estimate theLorentzfactor, b, nario, however,gives no predictions aboutthe behavior of the ofelectrons emitting attheenergy distribution break fromthe synchrotron SED component, while the SSC EC explanation is 22 positions of the SSC ( SSC 10 Hz) and synchrotron ( syn able todescribe qualitatively the observed changesin the low- 14 1 2 energy SED hump. 10 Hz) SED peaks (Sikora et al. 2009): b ( SSC syn) 104. This value is anorder of magnitude larger than thosefound The di erence in SED during Flare 1 and Flare 2 maybe indetailed SED modeling of FSRQs, (e.g., Vercelloneet al. qualitatively understood if the two flaring events are triggered 2011; Hayashidaet al. 2012; Dutkaet al. 2013). Di erentSED by di erent physicalmechanisms. The inverse-Compton hump models applied to the same source may lead todi erent esti- brightening not associated withbrightening of the synchrotron mates of (Sokolovsky et al. 2010; Abdo et al. 2010a). emission observed in Flare 1 may result from an increasing am- b bient photon field thatmight becausedbyan increasing accre- This is also thecaseforGB 1310 487 (Table 2). Hadronic tion rate onto thecentralsupermassive blackhole(Paggi et al. models predictthat X-rays are producedbysynchrotron radi- 2011).16 Flux increase inboth synchrotron and inverse-Compton ation of the secondaryultra-relativistic population ofelectrons SED components (as observed in GB 1310 487 during Flare 2), and positrons. To reproduce the hard X-ray spectra observed in combined with a peak energy increase of the two components, GB 1310 487 (Table 2)and in FSRQs, an extremely e cient ac- may result from additional electron acceleration. The inverse- celeration ofrelativistic protons within the inner parts of the out- Compton peak energy increase(with respecttopre- and post- flow is needed, and the jet kinetic power must be orders of mag- flare states) is evident during Flare 2. The synchrotron peak en- nitude larger than theEddington luminosity (Sikora et al. 2009), ergy increase during Flare 2 is not excludedbytheavailable data. making this scenarioveryunlikely. This fits the patternof -ray spectrum changes discussed in Sect.4.1,if the observed -ray emission is acombination 4.9. Interpretation of changes intheSED ofEC emission peaking at higher energies and SSC emission peaking atlower energies. The first flare leadsto the increased According to the leptonic interpretation of blazar SEDs outlined EC flux (relative to the SSC flux), which makesthe overall - in Sect.1,the near-IR flux shouldbe dominatedbythe syn- ray spectrum harder. The second flare,probably causedbyad- chrotron radiation while the observed -ray flux is acombina- ditional electron acceleration, is characterizedbytheenhanced tion ofEC and SSC components. The lackof broad linesin synchrotron (as observed in theIR)and thecorresponding SSC the opticalspectrum observedbyus (Sect.3.4)and Shaw et al. flux, possibly together with theEC component. The SSC com- (2012) suggests thatthe BLRin GB 1310 487 is weakor ob- ponent brightening makesthe overall -ray spectrumsofter com- scured from view. This,however,does not exclude theECsce- pared to Flare 1. The post-flare high-energy hump might bea nario: photons fromthe accretion diskor dusty torusmightserve combination of SSC and EC components peaking atlower ener- astargets for inverse-Compton scattering. The optical–UV spec- giesthanduring Flare 2due to lower acceleration of the under- trumisflatter than the near IR one; however,due to contam- lying electron population. Alternatively, since during the post- ination of the host (or intervening) galaxy and a nearby star flare state the Compton dominance is less than anorder of mag- (Sect.3.3), it is not possible todistinguish accretion-disk emis- nitude, the inverse-Compton hump might be dominatedbythe sion thatmight be presentin this wavelength range. Itmay also SSC component; SSC emission with q 1 (asin Flare 2) might be thatthe accretion disk avoids detection because it emits at be produced (Zacharias&Schlickeiser 2012a,b) if theenergy shorter wavelengths. This wouldbe thecase if thecentral black densityofemitting particlesislarger than that of the magnetic hole mass is smaller than the one typically found inblazars, field, i.e., jet plasma is notin equipartition and Compton losses since the accretion-disk temperature decreases with increasing dominate the particleenergy loss budget (e.g., Potter &Cotter M (Shakura &Sunyaev 1973). Far-IR dataavailable onlydur- 2013). Aconstantsupplyofenergy to theemitting particlesis ing the high-IRstate(Flare 2)arenotsu cientto estimate the needed for this condition tobefulfilled foran extendedperiod possiblecontribution from a dusty torus. Thus, accretion-disk of time(Readhead 1994). Coordinatedbrightening of both syn- and dusty torusluminosities remain as free parametersin this chrotron and Compton emission components during Flare 2 sug- discussion. gests thatthe SSC mechanism is the oneresponsiblefor theen- TheEC component peaks at higher energiesthan the SSC hanced Compton emission. component for thefollowing reasons. First, the typical energy Itispossible thatthe lower-energy Compton emission is an- of seed photons for theEC process (IR,optical,orUVcorre- other EC component,notthe SSC component. Forexample, the sponding todusty torus, BLR,or accretion disk asthe dominat- high-energy EC component responsiblefor Flare 1 might beas- ing source ofexternal radiation) shouldbe higher than the typical sociated with accretion-disk photons, while the lower-energy EC energy of synchrotron photons. The synchrotron emission peak component couldbeassociated with lower-energy photons pro- is located in thefar-IR, assuggestedbythe observed steepnear- duced in the dusty torus. Since the SSC emission hasthe same

IRspectrum. Second, due to relativisticaberration, most external 16 photons illuminate the synchrotron-emitting plasma blob head The increased accretion rate should result in accretion-diskbrighten- ing. However, we may still not detectthe accretion disk for the same on. Therefore, theexternal photons are additionallyblueshifted reasons we do not detectitin the quiescentstate:itiseither too faint or in the reference frame of the plasma blob. Achange in relative emits atshorter wavelengths. Finally, we cannot exclude the possibility strengthof theEC and SSC humpsmay explain the observed thatthe singleavailableUVdata pointin the M2band obtainedduring -ray spectrum evolution, with the harder spectrum correspond- Flare 1 (Table 2, Fig. 10) has no contribution fromthe accretion-disk ing togreater contribution of theEC componentto the total GeV emission.

Article number,page 15 of 18 A A proofs: manuscript no. 1310 beaming pattern assynchrotron emission (e.g., Finke 2013), the and comments that helped improve this paper. The Fermi LAT Collaboration ac- SSC Compton dominance is independent of . The observed EC knowledges generous ongoing support from a number ofagencies and institutes radiation intensityhas a stronger dependence on thandoes that have supportedboth the development and the operation of theLATas well asscientific dataanalysis. These include theNational Aeronautics and Space the synchrotron radiation (e.g., Georganopoulos et al. 2001). An Administration (NASA) and theDepartment ofEnergy in theUnited States, the increase in would enhance both synchrotron and EC radi- Commissariat à l’EnergieAtomiqueand the Centre National de la Recherche ation, and theEC emission wouldbeenhancedbya larger Scientifique Institut National de PhysiqueNucléaire et de Physique desPar- factor. This would explain the large value of q during Flare 2 ticulesin France, theAgenzia SpazialeItalianaand theIstituto Nazionale di Fisica Nucleare in Italy, the MinistryofEducation, Culture, Sports, Science and withoutthe necessityofresorting to nonstationary SSC mod- Technology (MEXT), High Energy Accelerator Research Organization (KEK) els. The SSC peak in this scenario is assumed tobeat even and Japan Aerospace Exploration Agency (JAXA) in Japan, and theK. A. Wal- lower energiesthan the low-energy EC peak to account for the lenberg Foundation, and the Swedish Research Council as well asthe Swedish spectral change between Flare 2 and pre post-flare states. This NationalSpace Board in Sweden. Additionalsupport for science analysis during “SSC low-energy-EC high-energy-EC" scenario is more com- the operations phase is gratefully acknowledged fromtheIstituto Nazionale di Astrofisica in Italy and the Centre National d’ÉtudesSpatialesin France. We plex than the“SSC EC" scenario, so we consider theformer to acknowledge the use of public datafromthe Swift dataarchiveattheHigh be less likely. Additional detailed SED modeling is needed to Energy AstrophysicsScience Archive Research Center (HEASARC),provided test theabove scenarios numerically. by NASA’s Goddard Space FlightCenter. Based inpart on observations with The intermediate interflare state is probably acombination of the 100 mtelescope of the MPIfR (Max-Planck-Institut für Radioastronomie) and theIRAM 30 mtelescope. IRAMissupportedbyINSU CNRS (France), the decaying Flare 1 and rising Flare 2. This may indicate that ei- MPG(Germany)and IGN (Spain). TheOVRO 40 mmonitoring programis ther two independent emission zones are responsiblefor the two supported inpart by NASA grants NNX08AW31Gand NNG06GG1G, and by flares,or that a singleemission region is propagating through a NSF grant AST-0808050. This researchhasmade use of datafromthe MO- medium, withproperties gradually changing fromthosewhich JAV E database thatismaintainedbythe MOJAV E team (Lister et al. 2009a). The data presentedherein were obtained inpart with ALFOSC, which is pro- caused Flare 1 to thosecorresponding to Flare 2. videdbytheInstitutodeAstrofisica deAndalucia (IAA) under a joint agree- ThefactthatFlare 1has no counterpartin the 15 GHz radio ment with theUniversityof Copenhagen and NOTSA. The MOJAV E project lightcurve(ifatypical value ofafewmonths for the -ray radio is supported under NASA-Fermi grant NNX08AV 67G. Some of the data pre- delay is assumed;see Sect.4.5) suggests thattheregion responsi- sentedherein were obtained attheW. M. Keck Observatory, which is operated as a scientific partnership among the CaliforniaInstitute ofTechnology, theUni- blefor Flare 1 is locatedupstream of theVLBIcore region. This versityof California, and NASA. TheObservatory wasmade possible by the region is probablyheavily self-absorbed and does not contribute generous financialsupport of theW. M. Keck Foundation. We thank O. Fox, to the observed flux density atthis frequency. Flare 2occurred P. Kelly, I. Shivvers, and W. Zheng forassistance with some of theKeckob- during therise of the majorradio flare;this maybean indication servations. The near-IR observations were carriedout with the 2.1 mtelescope thatthe flaring region that dominates IR and high-energy emis- of theGuillermo Haro Observatory, INAOE, Mexico. F.K.S. and K.V.S. were partly supported for this research through a stipend fromtheInternationalMax sion maybe located close to the 15 GHz radio core, assuggested Planck Research School (IMPRS)forAstronomy and Astrophysics attheUni- for other blazars (e.g., Jorstad et al. 2010;Schinzel et al. 2012; versities of Bonn and Cologne. I.N. and R.S. are members of theInternational Wehrleet al. 2012). Max Planck Research School (IMPRS)forAstronomy and Astrophysics atthe Universities of Bonn and Cologne. F.K.S. acknowledgessupport by theNASA Fermi Guest Investigator program,grant NXX12A075G. K.V.S., Y.A.K., and 5. Conclusions Y.Y.K. were supported inpart by the Russian Foundation for Basic Research (Projects 11-02-00368 and 13-02-12103), the basicresearchprogram “Active We summarize the majorconclusions of this study as follows. processesingalacticand extragalactic objects” of the PhysicalSciences Division of the Russian Academyof Sciences, and the MinistryofEducation and Science of the Russian Federation (agreement No. 8405). Y.Y.K. was also supportedby 1. Identification of the flaring -ray source with theradio theDynasty Foundation. RATAN-600 operations were carriedout with the fi- source GB 1310 487 is firmly established through positional nancialsupport of the MinistryofEducation and Science of the Russian Federa- coincidence and simultaneousmultiwavelengthobservations tion (contract 14.518.11.7054). K.V.S. wassupportedbythe Science Education of the flux variability. Complexof theLebedev Physical Inst. (UNK-FIAN). A.B.P. wassupportedby 2. Significant changesin the -ray photon index with flux were the“Non-stationaryprocessesin theUniverse” Program of the Presidium of the Russian Academyof Sciences. A.V.F. and S.B.C. are grateful for the support of observed, showing the harder-when-brighter trend. Itmay re- NASA Fermi grant NNX12AF12GA, NSF grant AST-1211916, the Christopher sult from achanging relativecontribution ofEC and SSC R. Redlich Fund, and Gary and Cynthia Bengier. This researchhasmade use emission to the total -ray flux in the Fermi LAT band. of theNASA IPAC ExtragalacticDatabase (NED) which is operatedbythe Jet 3. The bright near-IR flare does not correspond to the brightest Propulsion Laboratory, CaliforniaInstitute ofTechnology, under contract with NASA. We alsoused NASA’s Astrophysics Data System. K.V.S. thanksMaria -ray flare. This maybean indication that di erentmecha- Mogilen for her help inpreparing this manuscript. nisms are driving the twoobserved flares. 4. Fromtheabsence ofVLBA proper motion and the opacity argument, we constrain the source Doppler factor: 1 5 11. References 5. No clear association couldbeestablishedbetween the -ray Abazajian, K. N., Adelman-McCarthy, J. K., Agüeros, M. A., et al. 2009, ApJS, variability and changesin radio flux and parsec-scale struc- 182, 543 ture. Simultaneous radio -rayobservations over a longer Abdo, A. A., Ackermann, M., Agudo, I., et al. 2010a, ApJ, 721, 1425 time baselineareneeded to test the possibility thatsome - Abdo, A. A., Ackermann, M., Agudo, I., et al. 2010b, ApJ, 716, 30 Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010c, ApJ, 715, 429 ray events are associated with radio flares. Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010d, ApJS, 188, 405 6. In the optical band, the objectisa blend ofa -ray radio-loud Abdo, A. A., Ackermann, M., Ajello, M., et al. 2011a, ApJ, 727, 129 narrow-lineAGNat z 0 638 with anunrelated emission- Abdo, A. A., Ackermann, M., Ajello, M., et al. 2011b, ApJ, 733, L26 line galaxy at z 0 500. TheAGNis not a member ofcom- Abdo, A. A., Ackermann, M., Ajello, M., et al. 2009a, Astroparticle Physics, 32, 193 mon types of -ray flaring AGNs (blazars and NLSy1s). Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010e, ApJ, 722, 520 7. TheAGNradiation is probably amplifiedbyafactor ofafew Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010f, ApJ, 710, 810 because of gravitationallensing. Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010g, ApJ, 710, 1271 Abdo, A. A., Ackermann, M., Ajello, M., et al. 2009b, ApJ, 707, 55 Acknowledgements. We thank Sara Cutini, Marco Ajello, DenisBastieri, Boris Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010h, ApJ, 719, 1433 Komberg, Seth Digel, Luca Latronico and theanonymous referee for discussions Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010i, Science, 328, 725

Article number,page 16 of 18 K. V. Sokolovsky et al.: Two active states of the narrow-line gamma-ray-loud AGN GB 1310 487

Abdo, A. A., Ackermann, M., Ajello, M., et al. 2009c, Phys. Rev. D, 80, 122004 Fuhrmann, L., Zensus, J. A., Krichbaum, T. P., Angelakis, E., &Readhead, Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010j, ApJ, 720, 912 A. C. S. 2007, in American Institute of PhysicsConference Series, Vol. 921, Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010k, ApJ, 714, L73 The First GLAST Symposium, ed. S. Ritz, P. Michelson, &C. A. Meegan, Abdo, A. A., Ackermann, M., Ajello, M., et al. 2009d, ApJ, 707, L142 249–251 Abdo, A. A., Ackermann, M., Ajello, M., et al. 2011c, ApJ, 736, 131 Gehrels, N., Chincarini, G., Giommi, P., et al. 2004, ApJ, 611, 1005 Abraham, R. G., Crawford, C. S., Merrifield, M. R., Hutchings, J. B., & Georganopoulos, M., Kirk, J. G., &Mastichiadis, A. 2001, ApJ, 561, 111 McHardy, I. M. 1993, ApJ, 415, 101 Ghisellini, G., Celotti, A., Fossati, G., Maraschi, L., &Comastri, A. 1998, MN- Abramowski, A., Acero, F., Aharonian, F., et al. 2010, A&A, 520, A83 RAS, 301, 451 Ackermann, M., Ajello, M., Albert, A., et al. 2012, ApJS, 203, 4 Ghisellini, G. &Maraschi, L. 1989, ApJ, 340, 181 Ackermann, M., Ajello, M., Allafort, A., et al. 2011, ApJ, 741, 30 Ghisellini, G. & Tavecchio, F. 2008, MNRAS, 386, L28 Ackermann, M., Ajello, M., Baldini, L., et al. 2010, ApJ, 721, 1383 Ghisellini, G. & Tavecchio, F. 2009, MNRAS, 397, 985 Aharonian, F., Akhperjanian, A. G., Bazer-Bachi, A. R., et al. 2007, ApJ, 664, Ghisellini, G., Tavecchio, F., Foschini, L., & Ghirlanda, G. 2011, MNRAS, 627 L71 Giommi, P., Capalbi, M., Fiocchi, M., et al. 2002, in Blazar Astrophysics Albert, J., Aliu, E., Anderhub, H., et al. 2007, ApJ, 669, 862 with BeppoSAX and Other Observatories, ed. P. Giommi, E. Massaro, & Angelakis, E., Fuhrmann, L., Marchili, N., Krichbaum, T. P., & Zensus, J. A. G. Palumbo, 63 2008, Mem. Soc. Astron. Italiana, 79, 1042 Giommi, P., Padovani, P., &Polenta, G. 2013, MNRAS, 431, 1914 Angelakis, E., Fuhrmann, L., Nestoras, I., et al. 2012, Journal of PhysicsCon- Giuliani, A., Chen, A., Mereghetti, S., et al. 2004, Mem. SAItSuppl., 5, 135 ference Series, 372, 012007 Giuricin, G., Mardirossian, F., &Mezzetti, M. 1988, A&A, 203, 39 Angelakis, E., Fuhrmann, L., Nestoras, I., et al. 2010, ArXiv:1006.5610 Gorshkov, A. G., Konnikova, V. K., &Mingaliev, M. G. 2008, Astronomy Re- Angelakis, E., Kraus, A., Readhead, A. C. S., et al. 2009, A&A, 501, 801 ports, 52, 278 Araudo, A. T., Bosch-Ramon, V., &Romero, G. E. 2010, A&A, 522, A97 Griest, K. 1991, ApJ, 366, 412 Arshakian, T. G., León-Tavares, J., Böttcher, M., et al. 2012, A&A, 537, A32 Gürkan, G., Hardcastle, M. J., &Jarvis, M. J. 2014, MNRAS, 438, 1149 Atwood, W. B., Abdo, A. A., Ackermann, M., et al. 2009, ApJ, 697, 1071 Hagen-Thorn, V. A., Efimova, N. V., Larionov, V. M., et al. 2009, Astronomy Baars, J. W. M., Genzel, R., Pauliny-Toth, I. I. K., & Witzel, A. 1977, A&A, 61, Reports, 53, 510 99 Hartman, R. C., Bertsch, D. L., Bloom, S. D., et al. 1999, ApJS, 123, 79 Barkov, M. V., Aharonian, F. A., Bogovalov, S. V., Kelner, S. R., & Khangulyan, Hartman, R. C., Böttcher, M., Aldering, G., et al. 2001, ApJ, 553, 683 D. 2012, ApJ, 749, 119 Hayashida, M., Madejski, G. M., Nalewajko, K., et al. 2012, ApJ, 754, 114 Beasley, A. J., Gordon, D., Peck, A. B., et al. 2002, ApJS, 141, 13 Hays, E. & Escande, L. 2009, TheAstronomer’s Telegram, 2316, 1 Begelman, M. C., Fabian, A. C., &Rees, M. J. 2008, MNRAS, 384, L19 Healey, S. E., Romani, R. W., Cotter, G., et al. 2008, ApJS, 175, 97 Bessell, M. S., Castelli, F., &Plez, B. 1998, A&A, 333, 231 Ho, L. C., Filippenko, A. V., &Sargent, W. L. W. 1997, ApJS, 112, 315 Boettcher, M. 2010, arXiv:1006.5048 Hogg, D. W. 1999, ArXiv Astrophysics e-prints Boettcher, M. 2012, ArXiv:1205.0539 Itoh, R., Yamanaka, M., Sasada, M., et al. 2009, TheAstronomer’s Telegram, Breeveld, A. A., Curran, P. A., Hoversten, E. A., et al. 2010, MNRAS, 406, 1687 2320, 1 Browne, I. W. A., Wilkinson, P. N., Jackson, N. J. F., et al. 2003, MNRAS, 341, Jackson, N. &Rawlings, S. 1997, MNRAS, 286, 241 13 Jolley, E. J. D., Kuncic, Z., Bicknell, G. V., & Wagner, S. 2009, MNRAS, 400, Bulgarelli, A., Gianotti, F., Trifoglio, M., et al. 2009, TheAstronomer’s Tele- 1521 gram, 2310, 1 Jones, T. W., O’dell, S. L., &Stein, W. A. 1974, ApJ, 188, 353 Burbidge, G. R., Jones, T. W., & Odell, S. L. 1974, ApJ, 193, 43 Jordi, K., Grebel, E. K., & Ammon, K. 2006, A&A, 460, 339 Burrows, D. N., Hill, J. E., Nousek, J. A., et al. 2005, Space Sci. Rev., 120, 165 Jorstad, S. G., Marscher, A. P., Larionov, V. M., et al. 2010, ApJ, 715, 362 Calderone, G., Ghisellini, G., Colpi, M., & Dotti, M. 2013, MNRAS Kalberla, P. M. W., Burton, W. B., Hartmann, D., et al. 2005, A&A, 440, 775 Cardelli, J. A., Clayton, G. C., &Mathis, J. S. 1989, ApJ, 345, 245 Kardashev, N. S., Khartov, V. V., Abramov, V. V., et al. 2013, Astronomy Re- Carrasco, L., Porras, A., Recillas, E., &Carramiñana, A. 2009, TheAs- ports, 57, 153 tronomer’s Telegram, 2311, 1 Kawabata, K. S., Nagae, O., Chiyonobu, S., et al. 2008, in Societyof Photo- Cash, W. 1979, ApJ, 228, 939 Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 7014, So- Celotti, A. & Ghisellini, G. 2008, MNRAS, 385, 283 cietyof Photo-Optical Instrumentation Engineers (SPIE) Conference Series Cheung, C. C. 2007, in American Institute of PhysicsConference Series, Vol. Kellermann, K. I., Kovalev, Y. Y., Lister, M. L., et al. 2007, Ap&SS, 311, 231 921, The First GLAST Symposium, ed. S. Ritz, P. Michelson, &C. A. Mee- Kellermann, K. I., Sramek, R., Schmidt, M., Sha er, D. B., & Green, R. 1989, gan, 325–326 AJ, 98, 1195 Ciprini, S. 2013, TheAstronomer’s Telegram, 4753, 1 King, L. J., Browne, I. W. A., Marlow, D. R., Patnaik, A. R., & Wilkinson, P. N. Cohen, M. H., Lister, M. L., Homan, D. C., et al. 2007, ApJ, 658, 232 1999, MNRAS, 307, 225 Cutri, R. M., Wright, E. L., Conrow, T., et al. 2012, Explanatory Supplementto Komatsu, E., Dunkley, J., Nolta, M. R., et al. 2009, ApJS, 180, 330 theWISEAll-Sky Data Release Products, Tech. rep. Komberg, B. V. & Ermash, A. A. 2013, ArXiv:1302.2942 D’Abrusco, R., Massaro, F., Ajello, M., et al. 2012, ApJ, 748, 68 Kovalev, Y. Y. 2009, ApJ, 707, L56 D’Ammando, F., Orienti, M., Finke, J., et al. 2012, MNRAS, 426, 317 Kovalev, Y. Y., Kellermann, K. I., Lister, M. L., et al. 2005, AJ, 130, 2473 D’Ammando, F., Raiteri, C. M., Villata, M., et al. 2011, A&A, 529, A145 Kovalev, Y. Y., Kovalev, Y. A., Nizhelsky, N. A., &Bogdantsov, A. B. 2002, D’Ammando, F., Tosti, G., Orienti, M., Finke, J., & on behalf of the Fermi Large PASA, 19, 83 Area Telescope Collaboration. 2013, ArXiv:1303.3030 Kovalev, Y. Y., Nizhelsky, N. A., Kovalev, Y. A., et al. 1999a, A&AS, 139, 545 Dermer, C. D. &Schlickeiser, R. 2002, ApJ, 575, 667 Kovalev, Y. Y., Nizhelsky, N. A., Kovalev, Y. A., et al. 1999b, in IAU Sympo- Doi, A., Nagai, H., Asada, K., et al. 2006, PASJ, 58, 829 sium, Vol. 194, Activity in Galaxies and Related Phenomena, ed. Y. Terzian, Donato, D., Sambruna, R. M., & Gliozzi, M. 2005, A&A, 433, 1163 E. Khachikian, & D. Weedman, 177 Donato, D., Wood, D., &Cheung, C. C. 2010, TheAstronomer’s Telegram, 2737, Lee, S.-S., Lobanov, A. P., Krichbaum, T. P., et al. 2008, AJ, 136, 159 1 León-Tavares, J., Valtaoja, E., Tornikoski, M., Lähteenmäki, A., & Nieppola, E. Dondi, L. & Ghisellini, G. 1995, MNRAS, 273, 583 2011, A&A, 532, A146 Dutka, M. S., Ojha, R., Pottschmidt, K., et al. 2013, ApJ, 779, 174 Linford, J. D., Taylor, G. B., Romani, R. W., et al. 2012, ApJ, 744, 177 Faber, S. M., Phillips, A. C., Kibrick, R. I., et al. 2003, in Societyof Photo- Lister, M. L., Aller, H. D., Aller, M. F., et al. 2009a, AJ, 137, 3718 Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 4841, So- Lister, M. L., Aller, M. F., Aller, H. D., et al. 2013, AJ, 146, 120 cietyof Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Lister, M. L. & Homan, D. C. 2005, AJ, 130, 1389 ed. M. Iye & A. F. M. Moorwood, 1657–1669 Lister, M. L., Homan, D. C., Kadler, M., et al. 2009b, ApJ, 696, L22 Falco, E. E., Kochanek, C. S., &Munoz, J. A. 1998, ApJ, 494, 47 Lobanov, A. P. 2005, arXiv:astro-ph 0503225 Feroci, M., Costa, E., So tta, P., et al. 2007, Nuclear Instruments and Methods Lott, B., Escande, L., Larsson, S., &Ballet, J. 2012, A&A, 544, A6 in PhysicsResearch A, 581, 728 Machalski, J. &Condon, J. J. 1983, AJ, 88, 1591 Filippenko, A. V. 1982, PASP, 94, 715 Marscher, A. P. 2009, ArXiv e-prints Finke, J. D. 2013, ApJ, 763, 134 Marscher, A. P. & Travis, J. P. 1996, A&AS, 120, C537 Finke, J. D., Dermer, C. D., &Böttcher, M. 2008, ApJ, 686, 181 Massaro, F., D’Abrusco, R., Ajello, M., Grindlay, J. E., &Smith, H. A. 2011, Fort, B. &Mellier, Y. 1994, A&A Rev.,5,239 ApJ, 740, L48 Foschini, L. 2013, ArXiv:1301.5785 Mattox, J. R., Bertsch, D. L., Chiang, J., et al. 1996, ApJ, 461, 396 Fossati, G., Maraschi, L., Celotti, A., Comastri, A., & Ghisellini, G. 1998, MN- Mattox, J. R., Wagner, S. J., Malkan, M., et al. 1997, ApJ, 476, 692 RAS, 299, 433 Max-Moerbeck, W., Richards, J. L., Pavlidou, V., et al. 2012, ArXiv:1205.0276 Fuhrmann, L., Angelakis, E., Zensus, J. A., et al. 2014, A&A inpreparation Meylan, G., Jetzer, P., North, P., et al., eds. 2006, Gravitational Lensing:Strong, Fuhrmann, L., Krichbaum, T. P., Witzel, A., et al. 2008, A&A, 490, 1019 Weak and Micro

Article number,page 17 of 18 A A proofs: manuscript no. 1310

Michel, A. & Huchra, J. 1988, PASP, 100, 1423 Wambsganss, J. 2006, ArXiv Astrophysics e-prints Mücke, A. &Protheroe, R. J. 2001, Astroparticle Physics, 15, 121 Wehrle, A. E., Marscher, A. P., Jorstad, S. G., et al. 2012, ApJ, 758, 72 Mücke, A., Protheroe, R. J., Engel, R., Rachen, J. P., &Stanev, T. 2003, As- White, R. L., Becker, R. H., Helfand, D. J., & Gregg, M. D. 1997, ApJ, 475, 479 troparticle Physics, 18, 593 Wilks, S. 1938, Ann. Math. Stat., 9, 60 Mukherjee, R. 2002, Bulletinof theAstronomicalSocietyofIndia, 30, 73 Wright, E. L., Eisenhardt, P. R. M., Mainzer, A. K., et al. 2010, AJ, 140, 1868 Mukherjee, R., Dingus, B. L., Gear, W. K., et al. 1996, ApJ, 470, 831 Wu, X.-P. 1994, A&A, 286, 748 Nagai, H., Orienti, M., Kino, M., et al. 2012, MNRAS, 423, L122 Xanthopoulos, E. 1996, MNRAS, 280, 6 Napier, P. J. 1994, in IAU Symposium, Vol. 158, Very High Angular Resolution Yuan, W., Zhou, H. Y., Komossa, S., et al. 2008, ApJ, 685, 801 Zacharias, M. &Schlickeiser, R. 2012a, MNRAS, 420, 84 Imaging, ed. J. G. Robertson & W. J. Tango, 117 Zacharias, M. &Schlickeiser, R. 2012b, ApJ, 761, 110 Napier, P. J. 1995, in AstronomicalSocietyof the Pacific Conference Series, Zacharias, N., Finch, C., Girard, T., et al. 2010, AJ, 139, 2184 Vol. 82, Very Long BaselineInterferometry and theVLBA, ed. J. A. Zensus, P. J. Diamond, &P. J. Napier,59 Nestoras, I., Fuhrmann, L., Angelakis, E., et al. 2014, A&A inpreparation 1 Nilsson, K., Pasanen, M., Takalo, L. O., et al. 2007, A&A, 475, 199 Max-Planck-Institut für Radioastronomie, Auf dem Hügel 69, D- Nolan, P. L., Abdo, A. A., Ackermann, M., et al. 2012, ApJS, 199, 31 53121 Bonn, Germany Oke, J. B., Cohen, J. G., Carr, M., et al. 1995, PASP, 107, 375 2 Astro Space Center ofLebedev Physical Institute, Profsoyuznaya Ott, M., Witzel, A., Quirrenbach, A., et al. 1994, A&A, 284, 331 Str.8432, 117997 Moscow, Russia Paczynski, B. 1986, ApJ, 304, 1 3 Sternberg Astronomical Institute, Moscow StateUniversity, Univer- Padovani, P. & Giommi, P. 1995, ApJ, 444, 567 sitetskii pr. 13, 119992 Moscow, Russia Paggi, A., Cavaliere, A., Vittorini, V., D’Ammando, F., & Tavani, M. 2011, ApJ, 4 Department of Physics and Astronomy, UniversityofNewMexico, 736, 128 AlbuquerqueNM, 87131, USA Paliya, V. S., Stalin, C. S., Shukla, A., &Sahayanathan, S. 2013, ApJ, 768, 52 5 Pittori, C., Verrecchia, F., Chen, A. W., et al. 2009, A&A, 506, 1563 HiroshimaAstrophysicalScience Center, HiroshimaUniversity, 1- Pogge, R. W. 2000, New A Rev., 44, 381 3-1, Kagamiyama, Higashi-Hiroshima, Hiroshima 739-8526, Japan Poole, T. S., Breeveld, A. A., Page, M. J., et al. 2008, MNRAS, 383, 627 6 INAF IASF–Bologna, ViaGobetti 101, I-40129 Bologna, Italy Potter, W. J. &Cotter, G. 2013, MNRAS, 431, 1840 7 Instituto Nacional deAstrofisica, Optica y Electronica, Tonantzintla, Pushkarev, A. B., Kovalev, Y. Y., & Lister, M. L. 2010, ApJ, 722, L7 Puebla, C.P. 72860 Mexico Pushkarev, A. B., Kovalev, Y. Y., Lister, M. L., &Savolainen, T. 2009, A&A, 8 Department ofAstronomy, Universityof California, Berkeley, CA 507, L33 94720-3411, USA Rappoldi, A. & AGILE Collaboration. 2009, Nuclear Instruments and Methods 9 in PhysicsResearch A, 610, 291 NASA Goddard Space FlightCenter, Greenbelt, MD 20771, USA 10 Readhead, A. C. S. 1994, ApJ, 426, 51 NationalResearch Council Research Associate, National Academy Refsdal, S. &Surdej, J. 1994, Reports on Progress in Physics, 57, 117 of Sciences, Washington, DC 20001, resident at NavalResearch Richards, J. L., Max-Moerbeck, W., Pavlidou, V., et al. 2011, ApJS, 194, 29 Laboratory, Washington, DC 20375, USA Roming, P. W. A., Kennedy, T. E., Mason, K. O., et al. 2005, Space Sci. Rev., 11 Dip. diFisica, Universitá degli Studi diPerugia, ViaA. Pascoli, I- 120, 95 060123 Perugia, Italy Roming, P. W. A., Koch, T. S., Oates, S. R., et al. 2009, ApJ, 690, 163 12 INFN – Sezione diPerugia, ViaA. Pascoli, I-06123 Perugia, Italy Savolainen, T., Homan, D. C., Hovatta, T., et al. 2010, A&A, 512, A24 13 Sbarrato, T., Foschini, L., Ghisellini, G., & Tavecchio, F. 2011, Advancesin INAF-IRA Bologna, ViaGobetti 101, Bologna, Italy Space Research, 48, 998 14 Université Bordeaux 1, CNRS IN2p3, Centre d’Études Nucléaires Sbarufatti, B., Treves, A., &Falomo, R. 2005, ApJ, 635, 173 de Bordeaux Gradignan, 33175 Gradignan, France Schinzel, F. K., Lobanov, A. P., Taylor, G. B., et al. 2012, A&A, 537, A70 15 Laboratoire Leprince-Ringuet, École polytechnique, CNRS IN2P3, Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500, 525 Palaiseau, France Shakura, N. I. &Sunyaev, R. A. 1973, A&A, 24, 337 16 Shaw, M. S., Romani, R. W., Cotter, G., et al. 2012, ApJ, 748, 49 U.S. NavalResearch Laboratory, Code 7653, 4555 Overlook Ave. Shaw, S. E., Westmore, M. J., Bird, A. J., et al. 2003, A&A, 398, 391 SW, Washington, DC, 20375-5352, USA Shepherd, M. C. 1997, in AstronomicalSocietyof the Pacific Conference Series, 17 Department of PhysicalSciences, HiroshimaUniversity, Higashi- Vol. 125, Astronomical DataAnalysis Software and Systems VI, ed. G. Hunt Hiroshima, Hiroshima 739-8526, Japan & H. Payne,77 18 Department of Physics, Stanford University, Stanford, CA 94305, Sikora, M. 2011, in IAU Symposium, Vol. 275, IAU Symposium, ed. USA G. E. Romero, R. A. Sunyaev, & T. Belloni, 59–67 19 Department ofAstronomy, Stockholm University, SE-106 91 Stock- Sikora, M., Begelman, M. C., &Rees, M. J. 1994, ApJ, 421, 153 holm, Sweden Sikora, M., Stawarz, Ł., Moderski, R., Nalewajko, K., &Madejski, G. M. 2009, 20 ApJ, 704, 38 TheOskar Klein Centre for Cosmoparticle Physics, AlbaNova, SE- Singal, J., Petrosian, V., Stawarz, Ł., & Lawrence, A. 2013, ApJ, 764, 43 106 91 Stockholm, Sweden Skrutskie, M. F., Cutri, R. M., Stiening, R., et al. 2006, AJ, 131, 1163 21 Department of Physics, Stockholm University, AlbaNova, SE-106 Sokolovsky, K. & Lebedev, A. 2005, in12th Young Scientists’ Conference on 91 Stockholm, Sweden Astronomy and Space Physics, ed. A. Simon & A. Golovin, 79 22 Department of Physics, PurdueUniversity, 525 Northwestern Ave, Sokolovsky, K. V. 2013, arXiv:1303.5451 West Lafayette, IN 47907, USA Sokolovsky, K. V., Healey, S. E., Schinzel, F., & Kovalev, Y. Y. 2009, TheAs- 23 Univ. Bordeaux, CENBG, UMR 5797, F-33170 Gradignan, France tronomer’s Telegram, 2306, 1 24 CNRS, IN2P3, CENBG, UMR 5797, F-33170 Gradignan, France Sokolovsky, K. V., Kovalev, Y. Y., Lobanov, A. P., et al. 2010, ArXiv:1006.3084 25 Stern, B. E. &Poutanen, J. 2011, MNRAS, 417, L11 Cahill Center forAstronomy and Astrophysics, CaliforniaInstitute Swanenburg, B. N., Hermsen, W., Bennett, K., et al. 1978, Nature, 275, 298 ofTechnology, 1200 E. California Blvd., Pasadena, CA 91101, USA Tanaka, Y. T., Stawarz, Ł., Thompson, D. J., et al. 2011, ApJ, 733, 19 26 ASI–ASDC, ViaG. Galilei, I-00044 Frascati (Roma), Italy Tavani, M. 2011, Nuclear Instruments and Methodsin PhysicsResearch A, 630, 27 NordicOptical Telescope, Apdo. de Correos 474, 38700 Santa Cruz 7 de la Palma, Spain Tavani, M., Barbiellini, G., Argan, A., et al. 2009, A&A, 502, 995 28 Pulkovo Observatory, Pulkovskoe Chaussee 65 1, 196140 St. Peters- Tavani, M., Barbiellini, G., Argan, A., et al. 2008, Nuclear Instruments and Meth- burg, Russia odsin PhysicsResearch A, 588, 52 29 Taylor, G. B., Healey, S. E., Helmboldt, J. F., et al. 2007, ApJ, 671, 1355 Crimean Astrophysical Observatory, 98409 Nauchny, Crimea, Valtaoja, E., Lähteenmäki, A., Teräsranta, H., & Lainela, M. 1999, ApJS, 120, Ukraine 95 30 Instituto Radioastronomía Milimétrica„ AvenidaDivina Pastora 7, Vercellone, S., D’Ammando, F., Vittorini, V., et al. 2010, ApJ, 712, 405 Local 20, E 18012, Granada, Spain Vercellone, S., Soldi, S., Chen, A. W., & Tavani, M. 2004, MNRAS, 353, 890 31 Department ofAstronomy, Kyoto University, Kitashirakawa- Vercellone, S., Striani, E., Vittorini, V., et al. 2011, ApJ, 736, L38 Oiwake-cho, Sakyo-ku, Kyoto 606-8502, Japan Véron-Cetty, M.-P. & Véron, P. 2010, A&A, 518, A10 32 INAF IASF–Palermo, ViaU. La Malfa 153, I-90146 Palermo, Italy Villata, M., Raiteri, C. M., Balonek, T. J., et al. 2006, A&A, 453, 817 33 Volobuev, S. A., Gal’per, A. M., Kirillov-Ugryumov, V. G., Luchkov, B. I., & Kwasan Observatory, Kyoto University, Ohmine-cho Kita Kazan, Ozerov, Y. V. 1972, Soviet Ast., 15, 879 Yamashina-ku, Kyoto, 607-8471, Japan

Article number,page 18 of 18