Unexpected High Brightness Temperature 140 PC from the Core in the Jet of 3C 120
Mar Roca-Sogorb, Jose L. Gomez, Ivan Agudo, Alan P. Marscher, Svetlana G. Jorstad
aa r X i v : . [ a s t r o - ph . C O ] F e b Draft version October 31, 2018
Preprint typeset using L A TEX style emulateapj v. 11/10/09
UNEXPECTED HIGH BRIGHTNESS TEMPERATURE 140 PC FROM THE CORE IN THE JET OF 3C 120
Mar Roca-Sogorb , Jos´e L. G´omez , Iv´an Agudo , Alan P. Marscher and Svetlana G. Jorstad Draft version October 31, 2018
ABSTRACTWe present 1.7, 5, 15, 22 and 43 GHz polarimetric multi–epoch VLBA observations of the radiogalaxy 3C 120. The higher frequency observations reveal a new component, not visible before April2007, located 80 mas from the core (which corresponds to a deprojected distance of 140 pc), witha brightness temperature about 600 times higher than expected at such distances. This component(hereafter C80) is observed to remain stationary and to undergo small changes in its brightnesstemperature during more than two years of observations. A helical shocked jet model – and perhapssome flow acceleration – may explain the unusually high T b of C80, but it seems unlikely that thiscorresponds to the usual shock that emerges from the core and travels downstream to the locationof C80. It appears that some other intrinsic process in the jet, capable of providing a local burst inparticle and/or magnetic field energy, may be responsible for the enhanced brightness temperatureobserved in C80, its sudden appearance in April 2007, and apparent stationarity. Subject headings: galaxies: active — galaxies: individual (3C 120) — galaxies: jets — polarization —radio continuum: galaxies INTRODUCTION
3C 120 is an active and relatively nearby (z=0.033)radio galaxy with a blazar–like one–sided superlumi-nal radio jet that has proven to be an excellent lab-oratory for studying the physics of relativistic jets inactive galactic nuclei (e.g., Walker et al. 1987, 2001;G´omez et al. 1998, 1999, 2000, 2001, 2008; Homan et al.2001; Marscher et al. 2002, 2007; Jorstad et al. 2005,2007; Chatterjee et al. 2009; Marshall et al. 2009). Pre-vious observations using the Very Long Baseline Array(VLBA) at high frequencies (15, 22 and 43 GHz) haverevealed a very rich inner jet structure, containing mul-tiple superluminal components as well as evidence forstationary features suggestive of a helical pattern viewedin projection (Walker et al. 2001; Hardee et al. 2005). OBSERVATIONS AND DATA REDUCTION
We present VLBA observations taken in November2007, as part of a multi–frequency program to map therotation measure in 3C 120 at all accessible VLBA scales,and in February 2001, when the VLBA was used as partof the ground array for HALCA observations of 3C 120at 5 GHz.The November 2007 VLBA observations were per-formed at 43, 22, 15, [4.6–5.1] and [1.35–1.75] GHz indual polarization, with 9 antennas of the VLBA (SaintCroix was down for maintenance). The highest frequencyobservations were performed with 32 MHz continuousbandwidth centered at the standard 43, 22 and 15 GHzfrequencies. The 4 and 20 cm receivers were split intofour 8 MHz bandwidths to maximize possible detectionof low Faraday rotation measure.Reduction of the data was performed with the AIPSsoftware in the usual manner (e.g., Leppanen et al. Instituto de Astrof´ısica de Andaluc´ıa, CSIC, Apartado3004, 18080 Granada, Spain. [email protected]; [email protected];[email protected] Institute for Astrophysical Research, Boston University,725 Commonwealth Avenue, Boston, MA 02215, [email protected]; [email protected] − ◦ -10 ◦ . After the initial reduction, the datawere edited, self–calibrated, and imaged both in totaland polarized intensity with a combination of AIPS andDIFMAP (Pearson et al. 1994). HIGH BRIGHTNESS TEMPERATURE IN 3C 120
Our high frequency (15–43 GHz) VLBA observationsduring November 2007 (Fig. 1) reveal a component (here-after C80) located 80 mas from the core (deprojectedto >
140 pc for a viewing angle < ◦ ; G´omez et al.2000). This is an unusually large distance for detect-ing emission at these high frequencies –in fact, none ofthe previous VLBI observations of 3C 120 (starting in1982) have ever reported emission at this distance at5 GHz or higher frequencies. We have remapped ourprevious 15, 22 and 43 GHz VLBA data taken from1996 to 2001(G´omez et al. 1998, 1999, 2000, 2001, 2008),covering more than 30 epochs, to check whether wemissed it in our previous analysis, but find no indica-tion for emission at the region of C80. However, remap-ping of the 15 GHz data published in the MOJAVEdatabase (containing also data from other programs; see for more in-formation) revealed the first detection of C80 in April2007. After this epoch C80 appears in all 15 GHz images,as shown in Fig. 2. This sequence of images shows nosignificant motions for C80 during the nearly two yearscovered by the observations. However, during 2007 thecomponent is observed to increase in flux density and toremain quite compact. Later on, C80 becomes more ex-tended, elongating in the south–west direction without Roca-Sogorb et al. C80
Linearly Polarized Intensity (mJy/beam)
Relative Right Ascension (mas)
C12
Fig. 1.—
VLBA images of 3C 120 in 2007 November 7 (2007.85) at 43 ( top ), 22 ( middle ) and 15 ( bottom ) GHz. Ten logarithmic contoursare plotted for the total intensity images between the first contour at 0.18, 0.16, and 0.08%, and the last contour at 90% of the peakbrightness of 1.11, 1.27, and 1.12 Jy beam − at 43, 22 and 15 GHz, respectively. Color shows the linearly polarized intensity, and barsindicate the direction of the electric polarization vector. A convolving beam of 0.63 × × × ◦ , was used at 43, 22 and 15 GHz, respectively. significant changes in its flux.Our lower frequency images at 1.7 and 5 GHz (seeFig. 3) show that the region located around C80 corre-sponds to a double structure, with another componentat ∼
90 mas (hereafter C90) located at the southern-most side of the jet, after which the jet extends to thenorthwest direction. This contrasts to what is found inour previous 2001 image at 5 GHz (see Fig. 3), whichshows extended emission located at ∼ b (seeFig. 4). The T b along the jet is observed to decline withdistance from the core following the r − . proportional-ity found by Walker et al. (1987). Component C80 has abrightness temperature of 5 × K, which is about 600times larger than the expected value of ∼ × K atsuch large distance from the core. 8 × K is also thetypical detection threshold for VLBA observations at 5GHz. Hence, the fact that C80 has not been detectedin any of the previous 5 GHz VLBA images implies anincrease in its brightness temperature by at least a fac-tor of 600. This unusually high T b explains why C80has become visible even at the highest VLBA observingfrequencies (see Fig. 1).Figures 1, 2 and 3 show the electric vector positionangle (EVPA) to be aligned with the local direction ofthe jet in C80 for observations after 2007, in contrastto what is found for the remainder of the jet (see alsoWalker et al. 2001), and to that shown in the 5 GHzimage taken in 2001 (see Fig. 3). Maps of the rotationmeasure at different frequency intervals during November2007 (G´omez et al., in preparation) show values of theorder of 10 rad m − for C80, small enough to marginally rotate only the EVPAs at 1.7 GHz. Hence, we can con-clude that the observed magnetic field in C80 is perpen-dicular to the local intensity structure for observationsafter 2007. The degree of polarization of C80 is ∼ α ∼ − ν ∝ ν α ), which is similar to the values found for the rest ofthe optically thin jet. DISCUSSION
3C 120 has been extensively observed at 1.7 GHz byWalker et al. (2001), showing a variety of moving knotsand a side–to–side structure suggestive of a helical pat-tern seen in projection (see also Hardee et al. 2005), inwhich the helical twisted flow along the southern sideof the jet is more closely aligned with the line of sight.Walker et al. (2001) identified a component located at81 mas from the core that appeared to be stationary(between 1982 and 1997), and could correspond to oneof the southernmost components produced by the en-hanced differential Doppler boosting. We are thereforetempted to identify this with component C80. However,our low frequency observations show that at the locationof the C80 component the emission structure changedsignificantly between 2001 and 2007, and that the south-ernmost emission in 2007 corresponds to C90, insteadof C80 (see Fig. 3). Hence, C90 would be associatedwith a jet region flowing at a smaller viewing angle, andit is therefore very unlikely that C80 would correspondto another bend in the jet, given the estimated helicalwavelengths (Hardee et al. 2005). Furthermore, a bendin the jet would lead to an increase in T b by a factorof ( δ new /δ old ) n , where δ is the Doppler factor and n is2 − α for the case of continuous jet or 3 − α for a movinginhomogeneity (Readhead 1994). In our case of a bendin the jet and an estimated spectral index of α = − n = 3; therefore to account for the 600 increase inT b it is required an increase in δ by a factor of ∼ C80
Linearly Polarized Intensity (mJy/beam)Relative Right Ascension (mas)
Fig. 2.—
Multi–epoch VLBA images of 3C 120 at 15 GHz fromour 2007.85 observations and the MOJAVE program. Vertical mapseparation is proportional to the time difference between successiveepochs of observation, shown for each image. Total intensity isplotted in contours at 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, and1024 mJy beam − . Polarized intensity is shown in color scale andbars (of unit length) indicate the electric vector position angle. Acommon convolving beam of 1.18 × ◦ was used for allimages, and is shown in the lower left corner. All of the imageshave been registered through comparison of the positions of severalcomponents. respect to the estimated mean Doppler factor δ old = 2 . γ old = 5 . θ old = 20 ◦ ; Jorstad et al. 2005). This involvesan unlikely acceleration of the jet from a Lorentz factorof 5.3 to ∼ η , so that themagnetic field is scaled up as B → B/η and the electronenergy density as N → N η − ( γ +2) / , where γ is theelectron energy spectral index (e.g., Hughes et al. 1989;G´omez et al. 1993). This yields an increase in the opti- C80 C90
Linearly Polarized Intensity (mJy/beam)
Relative Right Ascension (mas)
Fig. 3.—
VLBA observations of 3C 120 at 5 GHz in 2001 Febru-ary 21 ( top ), and 2007 November 30 at 5 GHz ( middle ), and 1.7GHz ( bottom ). Total intensity is plotted in contours at 0.1, 0.24,0.48, 0.9, 1.8, 3.6, 7.2, 14.4, 28.8, 57.6 and 90% of the peak bright-ness of 0.95 ( top ), 0.97 ( middle ), 0.84 ( bottom ) Jy beam − . Po-larized intensity is shown in color scale and bars (of unit length)indicate the electric vector position angle. A common convolvingbeam of 3.58 × ◦ was used for the images at 5 GHz,and 13.70 × ◦ at 1.7 GHz. T b ( K ) Distance from the core (mas)
Fig. 4.—
Observed brightness temperatures for different compo-nents along the jet in the 5 GHz images in February 2001 (black)and November 2007 (red). The blue line represents the expecteddecline with distance from the core. cally thin specific intensity of synchrotron radiation by afactor of η − (5 γ +7) / , or equivalently η (5 α − / . The mostlikely scenario would then involve a shocked helical jet,in which the effects of the shock wave compression andthe differential Doppler boosting would add to producean increase in the brightness temperature by a factor of( δ new /δ old ) − α η (5 α − / . Note that for a moving shockwe use n = 3 − α , on the assumption that the radiatingfluid is moving close enough to the shock speed. It ispossible to obtain an upper limit to η by maximizing thecontribution from the Doppler boosting considering that Roca-Sogorb et al. R e l a t i v e D e c li na t i on ( m a s ) Relative Right Ascension (mas)
Fig. 5.—
Position of component C80 at different epochs. Anelliptical Gaussian brightness distribution has been used to model–fit the C80 component in the VLBA images at 15 GHz from 2007.30to 2009.51 (see Fig. 2). Labels (from 1 to 14) indicate the epochs ofobservation in chronological order. The errors in the position havebeen calculated as the standard deviation of the position obtainedby using task
MFIT in AIPS and model–fitting with DIFMAP. in C80 the jet points directly towards the observer, butmaintains the same Lorentz factor of 5.3. The factor of600 increase in T b for C80 would then require a relativelyweak shock with η ≤ .
87. This is in fact a too conser-vative value, since as mentioned previously C90 is at asmaller viewing angle than is C80, so that the jet cannotpoint directly towards the observer at C80. If the jet in-stead bends to a viewing angle of 5 ◦ (10 ◦ ) then η = 0 . η = 0 . ∼
12 mas (C12; see Fig. 1) – which is one of the mostintense ever observed in 3C 120 – has η ∼ .
35. The un-usually high T b of C80 could therefore be explained bya combination of jet bending and a moving shock –andperhaps also some jet flow acceleration and/or unusu-ally large particle acceleration– but it seems very unlikelythat it corresponds to the usual shock that appears nearthe core and moves downstream to the location of C80:as simulated by G´omez et al. (1994), a component mov-ing through a helical jet would progressively increase influx as it approaches the bend, accompanied by a rota-tion of its EVPA. An increase in the flux density of C80is indeed observed during 2007, but not later. Some mo-tion of C80 would also be expected as it approaches themost favorably oriented jet region corresponding to C90,which contrasts with the quasi-stationarity of C80 shownin Fig. 5. Therefore, a shock moving through a helicaljet cannot account entirely for the observed properties ofC80.It appears that a strong, stationary shock generatedin situ, at the location of C80, is needed. This canbe a standing shock, produced perhaps by a steep de-crease in the external pressure. As has been proposedto explain the flaring HST-1 knot in the M87 jet byStawarz et al. (2006), the brightening of C80 in April 2007 may mark the arrival of excess particles and pho-tons produced by the active nucleus in the past. For ajet flow Lorentz factor similar to that measured for thecomponents, we can estimate that the core flare shouldhave taken place near 1975. Indeed, the light curve of3C 120 shows a period of very high activity at this epoch,corresponding to the maximum observed centimeter–wave flux since monitoring began in the mid–1960s (see ). It seems, how-ever, difficult to explain why such a large injection ofenergy into the jet did not result in a shock that re-mained bright as it moved down the jet, visible as anintense moving component in any of the following VLBIobservations.If component C80 corresponds to a fixed planar shockit is possible to estimate the minimum Lorentz factorchange required to explain the observed brightness tem-perature change: T b,new T b,old = (cid:18) δ new δ old (cid:19) − α (cid:18) η new η old (cid:19) α − > η = γ (8 γ − γ + 9) − / & ( √ γ ) − , and γ isthe upstream flow velocity (Hughes et al. 1989). If wesimplify by assuming that the jet points directly towardsthe observer we can write (cid:18) δ new δ old (cid:19) − α (cid:18) η new η old (cid:19) α − ∼ (cid:18) γ new γ old (cid:19) − α (cid:18) γ new γ old (cid:19) − α and thus T b,new T b,old ≈ (cid:18) γ new γ old (cid:19) > γ new > . γ old , that is, a flow ac-celeration from the estimated value of 5.3 to > ∼
10 radm − ) found for the C80–C90 region suggest that suchinteraction with the external medium is probably nottaking place.The observations made by Walker et al. (2001), cover-ing 1982 to 1997, and those presented in this work (2001igh brightness temperature in 3C 120 5and 2007), show evidence that, although the region lo-cated at ∼ ∼
86 masin 2001, and this in turn to the C90 component seen in2007. This motion of the bent jet region can be explainedin the framework of a slowly moving helical pattern, assimulated by Hardee et al. (2005). The estimated upperlimit of ∼ − ( ∼ c ; Walker et al. 2001)for the pattern speed of the helix is consistent with theobservations between 2001 and 2009.Although the helical jet model can explain the observedproperties of C90, none of the proposed models providesa complete explanation for the unusually high T b of C80,its sudden appearance in April 2007, and its apparentstationarity. It appears that some other intrinsic processin the jet, capable of providing a local burst in particleand/or magnetic field energy, may be responsible for theenhanced brightness temperature observed in C80. Fur-ther mid–frequency VLBA observations, currently underway, should provide the kinematical and flux evolution information necessary to obtain a better understandingof the nature of C80.This research has been supported in part by the Span-ish Ministerio de Ciencia e Innovaci´on grant AYA2007-67627-C03-03, the regional government of Andaluciagrant P09-FQM-4784, and by the U.S. National ScienceFoundation grant AST-0907893. I. A. acknowledges sup-port by an I3P contract by the Spanish Consejo Supe-rior de Investigaciones Cient´ıficas. We thank the anony-mous referee for helpful comments that improved signif-icantly our manuscript. The VLBA is an instrument ofthe National Radio Astronomy Observatory, a facility ofthe National Science Foundation operated under cooper-ative agreement by Associated Universities, Inc. This re-search has made use of data from the MOJAVE databasethat is maintained by the MOJAVE team (Lister et al.,2009, AJ, 137, 3718). This research has made use of datafrom the University of Michigan Radio Astronomy Ob-servatory which has been supported by the Universityof Michigan and by a series of grants from the NationalScience Foundation, most recently AST-0607523. Facilities: