Linear Polarization as a Probe of Gamma Ray Flaring Blazar Jets
aa r X i v : . [ a s t r o - ph . H E ] D ec Linear Polarization as a Probe of Gamma Ray Flaring Blazar Jets
M. F. Aller, H. D. Aller, P. A. Hughes
University of Michigan, Ann Arbor, MI, 48109-1042, USA
We describe and present initial results from a
Fermi cycle 2 program designed to monitor the behavior of thecentimeter-band linear polarization and total flux density emitted by gamma-ray-bright blazars during flaring.The goal of the program is to identify changes in the magnetic field structure in the radio jet associated withgamma-ray flaring and ultimately to test whether gamma-ray flaring is associated with the onset of shocks inthe radio jet. Light curves illustrating radio band variability patterns are shown for sample program sources.
1. Overview of Project
As part of a
Fermi cycle 2 program, we are obtain-ing monitoring observations at three centimeter bandfrequencies (14.5, 8.0 and 4.8 GHz) with the Uni-versity of Michigan 26-m radio telescope (UMRAO)of the source-integrated total flux density and linearpolarization of approximately 30 radio- and γ -ray-bright sources currently or expected to be in γ -ray-flare phase. The goal of our work is to identify jetconditions responsible for the generation of the γ -rayemission and to test the hypothesis that shocks playa role in the production of this emission as suggestedby earlier work [1]. A causal connection between ac-tivity in the radio jet and γ -ray flaring first proposedbased on EGRET data [2, 3, 4] is supported by early Fermi results. For example, statistical studies of thesources in the 15 GHz MOJAVE VLBA sample [5]have found that the members detected during the firstthree months of
Fermi operation [6] are more core-dominated, have higher brightness temperatures andDoppler factors, and are in a more active radio statethan non-detected MOJAVE sources [7, 8].While there has been increasing evidence suggestingthat the γ -ray emission arises near the radio core [2],a region believed to correspond to either the τ =1 sur-face or to a standing shock [9] and not at a site near tothe black hole/accretion disk, a number of importantquestions remain. These include the trigger for theemission of γ -rays, the position of the emission site(upstream or downstream of the radio core) and thenature of the emitting region itself (standing shock,propagating shock, or ‘blob’ with a chaotic magneticfield where turbulence produces the very high energyelectrons). A plausible scenario for the broadbandfeatures seen as outbursts or flares is that instabilitiesdevelop naturally within the collimated relativistic jetoutflows giving rise to shocks [10]. With the passageof a shock, there is a compression of the magnetic fieldwithin the emitting region and an increase in the de-gree of order of the magnetic field. The observationalsignature of such an event is a swing in the electricvector position angle (hereafter EVPA; an orientationorthogonal to the magnetic field direction in a trans-parent source) and an increase in the fractional linear Table I Program Sources0048 −
097 NRAO 190 0917+449 1510 −
089 2022 − −
234 1127 −
145 1633+382 BL LacDA 55 0528+134 3C 273 3C 345 2201+1710215+015 0716+714 3C 279 1717+178 CTA 1023C 66A 0727 −
115 1308+326 OT 081 3C 454.30235+164 0805 −
077 1502+106 1849+6703C 84 OJ 287 1508 −
055 1908 − polarization.
2. The Source Sample
Our program sources are listed in Table 1. In addi-tion to blazars, our source sample includes the radiogalaxy 3C 84 (NGC 1275); γ -ray emission has been de-tected in this source which appears to be associatedwith compact regions within the radio jet [11]. Selec-tion was based on the following criteria: 1) inclusionin the high confidence list in the LAT catalogue basedon the first three months of Fermi ’s operation [6], oridentified as γ -ray flaring in subsequent Astronomers’Telegrams; 2) 15 GHz total flux density of at least 1Jy in early 2009; and 3) membership in the MOJAVE15 GHz survey [5]. While AGN light curves regularlyshow peaks and troughs that can last for months orlonger, we expect these objects to be highly likely toexhibit bright flaring states during cycle 2. A mini-mum radio band total flux density of 1 Jy was chosento ensure good signal-to-noise in the multifrequencypolarimetry data. Typical fractional linear polariza-tions in blazars are of order a few percent, and theradio band spectra of our targets are characteristi-cally flat or inverted (14.5 GHz total flux density > eConf C091122
1. 2. 3. J AN S KY S S 0.0 0.1 0.2 J AN S KY S P 120. 0. 60. D E G R EE S X OR 103 = 4.8 GHz = 8.0 GHz = 14.5 GHz UMRAO Dec 9, 2009
Figure 1: From bottom to top, daily averages of total flux density S, polarized flux P, and electric vector position angle χ illustrating recent variability at radio band in the high redshift (z=1.839) QSO 1502+106 (OR 103). Observations at14.5, 8.0, and 4.8 GHz are denoted by crosses, circles, and triangles respectively. the inner jet.As new γ -ray-flaring radio-bright AGN are iden-tified, they are added to the UMRAO program ifthey meet our criteria for inclusion in the monitor-ing program; examples of newly added sources areNRAO 190, a bright EGRET source [12] in whichrecent Fermi -detected flaring was reported [13], and3C 345 a well-known bright blazar surprisingly notdetected by EGRET. Additionally, we are observingon a time-available basis a few sources which were ex-ceptionally bright during the EGRET era but whichare currently relatively quiescent in both the radio and γ -ray bands. Included in this group is NRAO 530 [14].Several of the sources listed in Table 1 are new to ourcore program; others have been observed by UMRAOfor many years. Before selecting our sources, we carried out ex-ploratory radio band measurements of a larger groupin 2009 March to evaluate the current radio band vari-ability state and to measure the flux levels for all po-tential program sources. Monitoring of our samplecommenced in 2009 August. Throughout cycle 2 weexpect to obtain two observations weekly at 14.5 GHz(the frequency at which the variability amplitude ishighest in AGN), and one observation weekly at each of 8.0 and 4.8 GHz for flaring sources. This samplingrate is matched to the expected variability timescaleand duty cycle in the radio band based on historicalmeasurements obtained over decades by UMRAO [15].However, the sampling rate will be increased if war-ranted, e.g. during some phases in the BL Lac object0716+714 which has historically exhibited large am-plitude, very rapid radio band variations. Each dailymeasurement consists of a series of on-off measure-ments over a 30-40 minute time span; these measure-ments are interleaved with observations of a calibra-tor every 1-2 hours to monitor the antenna gain andpointing, and to verify the instrumental polarization.As a result of these requirements, only 20-25 programsources can be observed in each 24 hour run, and themost active will be selected from the sample.
3. Early Program Results
The first stage of our program, to obtain data ex-hibiting resolved linear polarization variability tem-porally associated with time periods of γ -ray flaring,is in progress. During the first two months of theprogram we have detected several polarization events,and we are in the process of examining these and oursubsequent data to identify characteristic patterns inthe light curves. Resolved flares potentially suitable eConf C091122
009 Fermi Symposium, Washington, D.C., Nov. 2-5
10. 20. J AN S KY S S 0. 2. 4. 6. P E RC E N T P 20. 40. 60. 80. 100. 120. 140. D E G R EE S X
3C 279 = 4.8 GHz = 8.0 GHz = 14.5 GHz UMRAO Nov 30, 2009
Figure 2: From bottom to top, daily averages of total flux density, fractional linear polarization, and electric vectorposition angle at three frequencies illustrating the recent variability at radio band in 3C 279. Note the changes inpolarization in 2009 April-May which are resolved in these single dish observations. The variations in this sourceduring the mid 1980s were successfully modeled with a transverse shock [17]. for detailed modeling have been identified in 0727-115,0805-077, 3C 279, 1502+106, and 1510-089.We are finding that the timescales of the outburstsin polarized flux and periods of ordered temporalchanges in EVPA are relatively short, typically sev-eral weeks to a few months in duration. As an exam-ple, we show in Figure 1 the light curves for 1502+106(OR 103), a relatively new addition to our core pro-gram, which is highly active at γ -ray band. The MO-JAVE 15 GHz source structure in this source is rela-tively simple, and much of the polarized emission isassociated with a single source region. A study of γ -ray flaring in August 2008 used two MOJAVE imagesseparated by several months to identify a change inmagnetic field orientation [16], but the sampling ofthe imaging data did not permit tracing the change instructure. During the subsequent period covered byour monitoring data, there is a resolved outburst inpolarized flux (middle panel) and a systematic swingin EVPA (top panel) which commenced in 2009 Febru-ary. These variations track at 14.5 and 8.0 GHz; butthe data follow a different path at 4.8 GHz. The multi-frequency total flux density data (bottom panel) showa spectrum characteristic of a self-absorbed source;thus the emission at 4.8 GHz most likely arises in asomewhat different physical region of the source, fur-ther out from the inner region probed by the higher frequencies. In this flare the position angle swing andthe development of the linearly polarized outburst arecharacteristic of a shock event.Figure 2 shows results for a structurally complexQSO, 3C 279, where the contributions from individ-ual core and jet components are blended in the totalflux density light curve. During a series of events inthe mid-1980s the variations apparent in the UMRAOdata for this source were successfully modeled assum-ing a transverse shock [17]; and the jet parametersderived in that work will be used as a starting pointin the new analysis.Some sources exhibiting γ -ray flaring have notshown large amplitude changes in either total flux den-sity or in polarized flux in our observing band (e.g.NRAO 190). We do not know yet whether this isdue to frequency-dependent time delays across bands,the masking of variability in the UMRAO source-integrated measurements due to competing, indepen-dently evolving source components, or whether morethan one scenario is required to explain the origin ofthe high energy emission.Two very bright EGRET sources on the Fermi
LATlist, 0528+134 and NRAO 530, have not exhibitedhigh-amplitude flaring in either the γ -ray or in the ra-dio band since the launch of Fermi . We show the longterm light curves for NRAO 530 in Figure 3 which il- eConf C091122 J AN S KY S S 0.0 0.2 0.4 0.6 0.8 J AN S KY S P 120. 0. 60. 120. D E G R EE S X NRAO 530 = 4.8 GHz = 8.0 GHz = 14.5 GHz UMRAO Dec 9, 2009
Figure 3: From bottom to top, long term monthly averages of the total flux density, polarized flux, and electric vectorposition angle in the QSO NRAO 530, a source which was exceptionally bright in both the γ -ray and in the radio bandduring the EGRET era in the 1990s, but which is currently exhibiting only low level variability in both of these bands. lustrates the relatively low flux level in the radio bandsince the launch of Fermi . The presence of low fluxesin both bands is consistent with the expected behav-ior assuming that the activity is broad band and thatthe same particles are responsible for the high and thelow energy emission.In the next phase of our work, we will carry outshock modeling of resolved radio band events exhibit-ing the shock signature. This will allow us to set limitson the physical conditions in the radio jet during γ -ray flaring. A set of resolved radio band flares willbe selected for detailed shock-in-jet modeling follow-ing the procedures in our previous analyses [10, 17]which use multifrequency linear polarization and to-tal flux density observations as constraints. While ourearlier work assumed a specific shock geometry (trans-verse shocks), the new models will employ a transfercode that admits arbitrary shock orientation, result-ing in extra degrees of freedom which must be con-strained. The complementary VLBA imaging datafrom the MOJAVE and BU programs will be used tolimit the allowable range of orientations. The model-ing is expected to yield information on shock strength,the character of the jet flow, and the low energy cut-off in the spectrum of the radiating particles, all ofwhich are key input parameters in jet emission mod-els designed to explain the origin of the γ -ray emission[18]. Acknowledgments
This work is supported by NASA
Fermi grantNNX09AU16G. The operation of UMRAO is madepossible by funds from the University of Michigan As-tronomy Department and by a series of grants fromthe NSF, most recently AST-0607523. This researchhas made use of data from the MOJAVE database andfrom the
Fermi website.
References [1] S. Jorstad, et al. “Multiepoch Very Long BaselineArray Observations of EGRET-Detected Quasarsand BL Lacertae Objects: Connection BetweenSuperluminal Ejections and Gamma-Ray Flaresin Blazars”, ApJ, 556, 738, 2001.[2] S. Jorstad, et al. “Multiepoch Very Long BaselineArray Observations of EGRET-Detected Quasarsand BL Lacertae Objects: Superluminal Motionof Gamma-Ray Bright Blazars”, ApJS, 134, 181,2001.[3] E. Valtaoja & H. Ter¨asranta,“Gamma radiationfrom radio shocks in AGN jets”, A&A, 297, L13,1995. eConf C091122
009 Fermi Symposium, Washington, D.C., Nov. 2-55[4] M.F. Aller, H.D. Aller, & P.A. Hughes, “TheRadio-Gamma-Ray Connection: the Radio Prop-erties of Gamma-Ray-Bright Blazars” in IAUSymp. 175, ed. R.D. Ekers, C. Fanti, & L.Padrielli, 283, 1996.[5] M.L. Lister, et al. “MOJAVE: Monitoring of Jetsin Active Galactic Nuclei with VLBA Experi-ments. V. Multi-Epoch Images”, AJ, 137, 3718,2009.[6] A.A. Abdo, et al. “Bright Active Galactic Nu-clei Source List from the First Three Months ofthe FERMI Large Area Telescope All-Sky Sur-vey” ApJ, 700, 597, 2009.[7] Y.Y. Kovalev, et al. “The Relation Between AGNGamma-Ray Emission and Parsec-Scale RadioJets”, ApJ Letters, 696, 17L, 2009.[8] T. Savolainen, et al. “Relativistic beaming andgamma-ray brightness of blazars”, A&A. Letters,submitted (astro-ph 0911.4924).[9] A.P. Marscher, “Jets in Active Galactic Nu-clei”, in The Jet Paradigm from Microquasars toQuasars, ed. T. Belloni (astro-ph 0909.2576).[10] P.A. Hughes, H.D. Aller, & M.F. Aller, “Syn-chrotron Emission from Shocked Relativistic Jets.II. - A Model for the Centimeter Wave Band Qui-escent and Burst Emission From BL Lacertae”,ApJ, 341, 68, 1989.[11] A.A. Abdo, et al. “Fermi Discovery of Gamma- Ray Emission from NGC 1275”, ApJ, 699, 31,2009.[12] T.A. McGlynn, et al., “A Gamma-Ray Flare inNRAO 190”, ApJ, 481, 625, 1997.[13] S. Ciprini for the Fermi LAT Collaboration,“Fermi LAT Detection of a GeV Flare fromBlazar NRAO 190 (PKS 0440-00)”Atel No. 2049,2009.[14] G.C. Bower, et al. “A Dramatic Millimeter Wave-length Flare in the Gamma-Ray Blazar NRAO530”, ApJ, 484, 118, 1997.[15] H.D. Aller, M.F. Aller, G.E. Latimer, and P.E.Hodge, “Spectra and Linear Polarizations of Ex-tragalactic Variable Sources at Centimeter Wave-length”, ApJS, 59, 513, 1985.[16] A.A. Abdo et al., “A New and Distant Gamma-Ray Blazar in Outburst Discovered by theFERMI Large Area Telescope”, ApJ, submitted.[17] P.A. Hughes, H.D.Aller, & M.F.Aller, “Syn-chrotron Emission from Shocked Relativistic Jets.III. Models for the Centimeter Wave Band Qui-escent and Burst Emission from 3C 279 andOT 081”, ApJ, 374, 57, 1991.[18] C.D. Dermer et al., “Gamma-Ray Studies ofBlazars: Synchro-Compton Analysis of Flat Spec-trum Radio Quasars”, ApJ, 692, 32, 2009.