Kinematics of the Parsec-Scale Relativistic Jet in Quasar 3C 279: 1991-1997

Please download to get full document.

View again

of 30
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Description
We present results of long-term high-frequency VLBI monitoring of the relativistic jet in 3C279, consisting of 18 epochs at 22 GHz from 1991 to 1997 and 10 epochs at 43 GHz from 1995 to 1997. Three major results of this study are: apparent speeds
Transcript
    a  r   X   i  v  :  a  s   t  r  o  -  p   h   /   0   0   0   8   4   5   8  v   1   2   9   A  u  g   2   0   0   0 Draft version June 21, 2011 Preprint typeset using L A TEX style emulateapj v. 04/03/99 KINEMATICS OF THE PARSEC-SCALE RELATIVISTIC JET IN QUASAR 3C279: 1991 – 1997 A. E. Wehrle 1 , B. G. Piner 1 , 2 , S. C. Unwin 1 , A. C. Zook 3 , W. Xu 4 , A. P. Marscher 5 , H.Ter¨asranta 6 , & E. Valtaoja 7 , 8 ABSTRACTWe present results of long-term high-frequency VLBI monitoring of the relativistic jet in 3C279,consisting of 18 epochs at 22 GHz from 1991 to 1997 and 10 epochs at 43 GHz from 1995 to 1997.Three major results of this study are: apparent speeds measured for six superluminal components rangefrom 4.8 to 7.5 c  ( H  0 =70 km s − 1 Mpc − 1 ,  q  0 =0.1), variations in the total radio flux are due primarilyto changes in the VLBI core flux, and the uniform-sphere brightness temperature of the VLBI core is ∼ 1 × 10 13 K at 22 GHz after 1995, one of the highest direct estimates of a brightness temperature. If thevariability brightness temperature measured for 3C279 by L¨ahteenm¨aki & Valtaoja is an actual valueand not a lower limit, then the rest-frame brightness temperature of 3C279 is quite high and limited byinverse Compton effects rather than equipartition.The parsec-scale morphology of 3C279 consists of a bright, compact VLBI core, a jet component (C4)that moved from  ∼ 2 mas to  ∼ 3.5 mas from the core during the course of our monitoring, and an inner jet that extends from the core to a stationary component, C5, at ∼ 1 mas from the core. Component C4followed a curved path, and we reconstruct its three-dimensional trajectory using polynomial fits to itsposition versus time. Component C5 faded with time, possibly due to a previous interaction with C4similar to interactions seen in simulations by G´omez et al. Components in the inner jet are relativelyshort-lived, and fade by the time they reach ∼ 1 mas from the core. The components have different speedsand position angles from each other, but these differences do not match the differences predicted by theprecession model of Abraham & Carrara. Although VLBI components were born about six months priorto each of the two observed  γ  -ray high states, the sparseness of the  γ  -ray data prevents a statisticalanalysis of possible correlations. Subject headings:  quasars: individual: (3C279) — galaxies: jets — galaxies: active — radiationmechanisms: non-thermal — radio continuum: galaxies 1.  INTRODUCTION The quasar 3C279 ( z  = 0 . 536) is one of the archetypalsuperluminal radio sources (Cotton et al. 1979). At  γ  -rayenergies, the light curve of 3C279 has been sampled inter-mittently since the launch of the Compton Observatory in1991; it is one of the brightest EGRET quasars (Hartmanet al. 1999). 3C279 is also well known as an optically vio-lent variable (OVV), with large and rapid outbursts (Webbet al. 1990). Strong variability on timescales shorter thanone day is observed in high-energy bands (Wehrle et al.1998; Lawson, McHardy, & Marscher 1999).Correlations between the variability seen over the entireelectromagnetic spectrum have proved elusive. Variabil-ity occurs on a variety of timescales, especially at highenergies, and time sampling has been adequate to trackthe variations only at radio, millimeter, and X-ray bands,and as far as practicable in optical bands. There are strongtheoretical motivations for the search for correlations. Thetwo-humped overall spectral energy distribution is mostnaturally explained as a combination of synchrotron ra-diation for the radio through optical-uv region, and in-verse Compton emission at higher energies (Maraschi et al.1994; Wehrle 1999). The synchrotron and inverse Comp-ton emission is generally thought to be associated witha jet of relativistic electrons; however, the source of theseed photons for inverse Compton scattering is a matterof considerable debate (e.g. Maraschi, Ghisellini, & Celotti1992; Sikora, Begelman, & Rees 1994). Distinguishing thediffering mechanisms involves a full understanding of thetime correlations in the different energy bands. A con-sensus is emerging that for GeV-peaked blazars, the seedphotons upscattered to x-ray and  γ  -ray energies srcinateoutside the jet (e.g., in the accretion disk or broad line re-gion clouds) with a minor contribution from synchrotronphotons (Urry 1999). A full understanding may have toawait the next generation of   γ  -ray satellite observatories.In the radio regime, the variability timescale is longer,and flux monitoring at 4.8, 8.4 and 14.5 GHz, completewith polarizationdata, has been obtained at the Universityof Michigan Radio Observatory (e.g. Aller et al. 1985). 1 Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA 91109; Ann.E.Wehrle@jpl.nasa.gov,B.G.Piner@jpl.nasa.gov; Stephen.C.Unwin@jpl.nasa.gov 2 Department of Physics and Astronomy, Whittier College, 13406 E. Philadelphia Street, Whittier, CA 90608; gpiner@whittier.edu 3 Department of Physics and Astronomy, Pomona College, Claremont, CA 91711; azook@pomona.edu 4 Infrared Processing and Analysis Center, Jet Propulsion Laboratory, California Institute of Technology 100-22, Pasadena CA 91125 5 Institute for Astrophysical Research, Boston University, 725 Commonwealth Avenue, Boston MA 02215; marscher@bu.edu 6 Mets¨ahovi Radio Observatory, Mets¨ahovintie 114, 02540 Kylm¨al¨a, Finland; hte@kurp.hut.fi 7 Tuorla Observatory, V¨ais¨al¨antie 20, FIN-21500 Piikki¨o, Finland 8 Department of Physics, Turku University, FIN-20100 Turku, Finland; valtaoja@deneb.astro.utu.fi 1  2Monitoring at 22 and 37 GHz has been done at Mets¨ahoviObservatory (Ter¨asranta et al. 1992; Ter¨asranta et al.1998). Less frequent monitoring has been performed atSEST (90 and 230 GHz) (Tornikoski et al. 1996) and atthe JCMT (230 GHz) (Marscher et al. 1999).The time variability of 3C279’sVLBI structure has beenstudied by several groups, beginning with the earliest daysof the VLBI technique itself (Knight et al. 1971; Whitneyet al. 1971; Cohen et al. 1971). Most of these observa-tions were made using the ad-hoc US and European VLBINetworks, with observations at intervals of about 1 year(Unwin et al. 1989 [hereafter U89]; Carrara et al. 1993[hereafter C93]; Abraham & Carrara 1998). With the ad-vent of the NRAO Very Long Baseline Array 9 (VLBA),more frequent monitoring began in 1991, with an emphasison higher radio frequencies (22 and 43 GHz). This paperpresents results from this more frequent monitoring; pre-liminary results from this monitoring have been presentedby Wehrle, Unwin, & Zook (1994), Wehrle et al. (1996),and Unwin et al. (1998). Polarization-sensitive VLBI im-ages of 3C279 have also been made by Lepp¨anen, Zensus,& Diamond (1995), Cawthorne & Gabuzda (1996), Lis-ter, Marscher, & Gear (1998), Lister & Smith (2000), andHoman & Wardle (1999) (who detect a significant compo-nent of circular polarization). Space VLBI observations at5 and 1.6 GHz have been performed with the VLBI SpaceObservatory Programme (VSOP) since 1998; first resultsare reported by Piner et al. (2000a).The highest angular resolution achieved on 3C279 is50  µ as, at 86 GHz (Rantakyr¨o et al. 1998), and VLBIfringes have been detected up to frequencies of 215 GHz(Krichbaum et al. 1997). In images from 1990 and 1992,Rantakyr¨o et al. (1998) showed a narrow string of com-ponents within about 1 mas of the core. Rantakyr¨o et al.speculate that most of the “missing” flux lies in a moreextended jet which is resolved out by their 50  µ as beam.In this paper, we present the results of a long-term VLBImonitoring campaign on 3C279. The data comprise VLBIimages at 22 GHz over the period 1991 - 1997 (a total of 18 epochs) and at 43 GHz over the period 1995 - 1997 (10epochs). In Section 2 of this paper we present the VLBIobservation series, explaining how the data were collected,calibrated, and analyzed. Section 3 presents the VLBI im-ages. Section 4 discusses the superluminal motion visiblein the image sequence, and shows that different regions of the jet show qualitatively different evolution. Section 5analyzes the flux density and spectral evolution of the ra-dio core and components in the VLBI jet. We present ourconclusions in Section 6. In a subsequent paper (Pineret al. 2000b) we will combine synchrotron self-Comptonmodels with our VLBI data and X-ray data to furtherconstrain the jet kinematics. Throughout the paper weassume  H  0 =70 km s − 1 Mpc − 1 and  q  0 =0.1, and compo-nent speeds measured by others have been expressed inthese terms. With these assumptions, 1 mas correspondsto a linear distance of 5.8 pc, and a proper motion of 1mas yr − 1 corresponds to an apparent speed of 29 c . 2.  VLBI OBSERVATIONS We have observed 3C279 at 22 GHz since the mid1980’s, and at 43 GHz since 1995. Our first experimentsused the Global VLBI Network which was composed of non-identical antennas at various observatories. The datathrough 1994 were recorded in Mark II mode with 2 MHzbandwidth, followed by correlation at the Caltech/JPLBlockII Correlator. The Global Networkusually had threeobservingsessions per year of which two (at most) included22 GHz. During those sessions, seven antennas were ableto mount 22 GHz receivers. In 1991, we added the firstantennas in the new NRAO VLBA. By the mid 1990’s, weused the VLBA alone with 32 MHz bandwidth recorded ontapes correlated at the VLBA Correlator in Socorro, NewMexico. We added 43 GHz to our monitoring starting in1995. Some of the later maps were made in “snapshot”mode while others were made with full ( u,v ) tracks. Mostof the full-track observations since 1995 were done with al-ternating scans at 22 and 43 GHz, and sometimes a lowerfrequency. Results from this monitoring program prior to1991 are discussed by U89 and C93. The data obtainedduring the more frequent monitoring since 1991 are dis-cussed in this paper; these VLBI observations are listed inTable 1. Since the end of our monitoring program in 1997,3C279 has been part of a VLBA polarization monitoringprogram described by Marchenko et al. (1999).Images from some of the epochs listed in Table 1 haveappeared in various conference proceedings (e.g. Wehrleet al. 1994; Wehrle et al. 1996; Unwin et al. 1998). Inmany cases, we have re-analyzed the srcinal data, andobtained significantly improved images. The biggest im-provement was in correcting station-based calibration er-rors (and deletion of bad data in some cases); the in-teractive self-calibration, display, and imaging packageDIFMAP (Shepherd, Pearson, & Taylor 1994) was the keyto realizing these improvements. For some of the epochslisted in Table 1, observations at lower frequencies weremade as well, but since these observations contribute lit-tle to following source structure changes (because of theirlower resolution), discussion of these observations will bedeferred until the discussion of the broad-band spectrumin Piner et al. (2000b). Beginning in 1995 many of ourobservations recorded dual circular polarization. In thispaper, we discuss only the total intensity images formedfrom these observations.The data were fringe-fitted in AIPS, then exported tothe Caltech DIFMAP package (Shepherd et al. 1994) foramplitude and phase calibration, editing, and mapping.The 22 GHz data required particular attention to ampli-tude calibration because water vapor in the earth’s atmo-sphere absorbs at this frequency. Normal self-calibrationdoes not take care of this problem if it is cloudy at mostantennas because there are insufficient crossing points inthe ( u,v ) plane for sources that are nearly equatorial (like3C279); moreover, the problem is worse for antennas thatarelocated far from the mainland arraysuch as Saint Croix(which is nearly at sea level) or Mauna Kea (which ob-serves at low elevation angles). We compared the 22 GHzmonitoring data from Mets¨ahovi (where only the data ob-tained with dry observing conditions are accepted) withthe flux density measured on the shortest VLBI baselines,and applied an initial scaling factor of order 1.1 - 1.3 to 9 The National Radio Astronomy Observatory is a facility of the National Science Foundation operated under cooperative agreement byAssociated Universities, Inc.  3 Table 1VLBI Observations Experiment Bandwidth Obs. Time c Frequencies d Epoch Name VLBA Antennas a Other antennas b (MHz) (minutes) (GHz) Polarization1991 Jun 24 GU2B Fd,Kp,La,Nl,Pt Eb,Gb,Hs,Mc,Nt,On,Ov,Y1 2 973 22 LCP1992 Jun 14 GW6B Br,Fd,Kp,La,Nl,Ov,Pt Eb,Gb,Hs,Mc,Mh,On,Y1 2 703 22 LCP1992 Nov 10 GW6C Br,Hn,Kp,La,Nl,Ov,Pt Gb,Mc,Mh,Nt,On,Y1 2 485 22 LCP1993 Feb 17 GW008 Br,Hn,Kp,Ov,Pt,Sc Eb,Mc,Mh,Y1 2 521 22 LCP1993 Nov 8 BM030 Br,Hn,Mk,Nl,Ov,Pt ... 2 115 22 LCP1994 Mar 2 GW011A Br,Fd,Hn,Kp,La,Mk,Nl,Ov,Sc Eb,Gb,Mc,Mh,Nt,On,Y1 14 755 22 LCP1994 Jun 12 BM032A Hn,Mk,Nl,Ov,Pt,Sc ... 2 96 22 LCP1994 Sep 21 GW011B Br,Fd,Hn,Kp,La,Nl,Ov,Pt,Sc Y1 14 433 22 LCP1995 Jan 4 BB025 All ... 16 32 22 Dual1995 Feb 25 BM038 All ... 16 25, 45 22, 43 Dual1995 Mar 19 GW013B All ... 16 297, 297 22, 43 Dual1996 Jan 7 GW013C All ... 32 311, 311 22, 43 LCP1996 May 4 BM063 All ... 32 58 43 Dual1996 May 13 BW026 All ... 16 278 22 Dual1996 Jun 9 BW026B All ... 16 278, 278 22, 43 Dual1996 Nov 24 BM072 All ... 8 27 43 Dual1997 Jan 15 BW026D All ... 16 256, 256 22, 43 Dual1997 Mar 29 BW031A Br,Fd,Hn,Kp,La,Mk,Ov,Pt,Sc ... 16 190, 190 22, 43 Dual1997 Jul 16 BW031B All ... 16 190, 190 22, 43 Dual1997 Nov 16 BW031C Br,Fd,Kp,Mk,Nl,Ov,Pt,Sc ... 16 168, 167 22, 43 Dual a Br = Brewster, WA; Fd = Fort Davis, TX; Hn = Hancock, NH; Kp = Kitt Peak, AZ; La = Los Alamos, NM; Mk = Mauna Kea, HI; Nl =North Liberty, IA; Ov = Owens Valley, CA; Pt = Pie Town, NM; Sc = St. Croix, US Virgin Islands. b Antenna locations and sizes are as follows: Eb = Effelsberg, Germany, 100 m; Gb = Green Bank, WV, 43 m; Hs = Haystack, MA, 37 m; Mc= Medicina, Italy, 32 m; Mh = Mets¨ahovi, Finland, 14 m; Mp = Maryland Point, MD, 26 m; Nt = Noto, Italy, 32 m; On = Onsala, Sweden,20 m; Ov = Owens Valley, CA, 40m; Y1 = one antenna of the VLA, Socorro, NM, 25 m. c Two numbers indicate time on source at 22 and 43 GHz respectively. d Lower observed frequencies are not listed here since they are not discussed in this paper. antenna gains for stations affected by cloudy weather. Weestimate that 3C279 has about 1 Jy in 22 GHz emissionwhich is too diffuse to be sampled by the shortest spacingsin the ( u,v ) plane. In most cases, we chose an epoch withgood weather to map and model fit, then used the inputmodel to initiate the mapping-self-calibration sequence foradjacent epochs with bad weather. Data from antennasobtained during snow or rain were flagged after we foundthat they had significant adverse effects on the images.The amplitude calibration of the 43 GHz data was com-pared with the 37 GHz monitoring flux densities fromMets¨ahovi. In nearly all cases, the fluxes agreed to within20%; discrepant antennas were scaled accordingly. Self-calibration enables us to make reliable images for the pur-pose of tracking changes in the source structure; however,the scale factors applied render the overall flux scale some-what uncertain. This limits our ability to track the fluxdensity evolution of individual components at better thanabout a 10% level. 3.  VLBI IMAGING RESULTS Figure 1 shows the eighteen 22 GHz images of 3C279from the epochs listed in Table 1, and Figure 2 shows theten 43 GHz images. The images are shown with uniformweighting (uvweight=2,0 in DIFMAP) to maximize theresolution. Even though this produces an image with lowerdynamic range, the high resolution is important for dis-tinguishing components in the inner milliarcsecond. Theparameters of these images are listed in Table 2. Model-fit Gaussian positions are marked with asterisks on theimages. The model-fitting results are discussed in  §  4.1.The parsec-scale morphology of 3C279 during the years1991-1997 consists of the bright compact core, a brightsecondary component at a position angle of   − 114 ◦ whichmoves outward from about 2 mas to about 3.5 mas fromthe core during the observed time range, and an inner jetthat extends from the core out to about 1 mas. We identifythe bright secondary component with the component C4seen previously in our monitoring (U89; C93), and subse-quently by many other authors. In the earlier images, C4is connected to the inner jet emission, but in the later im-ages it is clearly separated, with a gap in emission betweenthe 1 mas point and C4. The region interior to 1 mas iscomplex, with multiple components forming, moving out,and fading on timescales of several years. Component C4is resolved and has significant internal structure, with mul-tiple Gaussians often required to model it (see § 4.1). Thehigher dynamic range images (e.g. the 22 GHz imagesfrom 1993 Feb 17 and 1994 Mar 2) show extended, diffuseemission to the southwest of the main jet. This extended,diffuse emission has a position angle of about  − 140 ◦ to − 150 ◦ , similar to that of the larger-scale VLBI jet seen atlower frequencies (Piner et al. 2000a), and the kiloparsec-scale jet seen with the VLA and MERLIN (de Pater &Perley 1983; Pilbratt, Booth, & Porcas 1987; Akujor et al.1994). No counterjet is detected, and the limit placed onthe jet/counterjet brightness ratio is about 100:1 at thedistance of C4. 4.  MOTION OF JET COMPONENTS 4.1.  Identification of Components by Model Fitting  We used the modelfit routine in DIFMAP to fit Gaus-sians to the visibility data for each epoch. These Gaussianmodels are listed in Table 3. Our procedure was to re-place all CLEAN components with a collection of circularGaussians, letting the circular Gaussians become ellipti-  4   Fig. 1.—  22 GHz uniformly weighted images of 3C279 from the 18 epochs listed in Table 1. The axes are labeled in milliarcseconds. Parameters of the images are given in Table 2.Model-fit Gaussian positions are marked with asterisks. The model-fit Gaussians are identified in Table 3.   5   FIG. 1.— Continued 
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x