Probing the innermost regions of AGN jets and their magnetic fields with RadioAstron IV. The quasar 3C 345 at 18 cm: Magnetic field structure and brightness temperature
F. M. Pötzl, A. P. Lobanov, E. Ros, J. L. Gómez, G. Bruni, U. Bach, A. Fuentes, L. I. Gurvits, D. L. Jauncey, Y. Y. Kovalev, E. V. Kravchenko, M. M. Lisakov, T. Savolainen, K. V. Sokolovsky, J. A. Zensus
AAstronomy & Astrophysics manuscript no. MAIN © ESO 2021February 9, 2021
Probing the innermost regions of AGN jets and their magneticfields with
RadioAstron
IV. The quasar 3C 345 at 18 cm: Magnetic field structure and brightnesstemperature (cid:63)
F. M. Pötzl , A. P. Lobanov , , E. Ros , J. L. Gómez , G. Bruni , U. Bach , A. Fuentes , L. I. Gurvits , , ,D. L. Jauncey , , Y. Y. Kovalev , , , E. V. Kravchenko , , , M. M. Lisakov , , T. Savolainen , , ,K. V. Sokolovsky , , and J. A. Zensus Max-Planck-Institut für Radioastronomie, Auf dem Hügel 69, 53121 Bonn, Germanye-mail: [email protected] Moscow Institute of Physics and Technology, Institutsky per. 9, Dolgoprudny, Moscow region, 141700, Russia Instituto de Astrofísica de Andalucía, CSIC, Apartado 3004, 18080, Granada, Spain INAF - Istituto di Astrofisica e Planetologia Spaziali, via del Fosso del Cavaliere 100, 00133, Rome, Italy JIVE - Joint Institute for VLBI ERIC, Oude Hoogeveensedijk 4, 7991 PD Dwingekoo, The Netherlands Dept. of Astrodynamics and Space Missions, Delft University of Technology, Kluyverweg 1, 2629 HS Delft, The Netherlands CSIRO Astronomy and Space Science, PO Box 76, Epping, NSW 1710, Australia Research School of Astronomy and Astrophysics, Australian National University, Canberra, ACT 2611, Australia Astro Space Center of Lebedev Physical Institute, Profsoyuznaya st. 84 /
32, Moscow, 117997, Russia INAF Istituto di Radioastronomia, Via P. Gobetti 101, Bologna 40129, Italy Aalto University Department of Electronics and Nanoengineering, PL 15500, FI-00076 Aalto, Finland Aalto University Metsähovi Radio Observatory, Metsähovintie 114, FI-02540 Kylmälä, Finland Center for Data Intensive and Time Domain Astronomy, Department of Physics and Astronomy, Michigan State University, 567Wilson Rd, East Lansing, MI 48824, USA Sternberg Astronomical Institute, Moscow State University, Universitetskii pr. 13, 119992 Moscow, RussiaReceived 22 September 2020 / Accepted 28 January 2021
ABSTRACT
Context.
Supermassive black holes in the centres of radio-loud active galactic nuclei (AGN) can produce collimated relativisticoutflows (jets). Magnetic fields are thought to play a key role in the formation and collimation of these jets, but the details are muchdebated.
Aims.
We study the innermost jet morphology and magnetic field strength in the AGN 3C 345 with an unprecedented resolution usingimages obtained within the framework of the key science programme on AGN polarisation of the Space VLBI mission
RadioAstron . Methods.
We observed the flat spectrum radio quasar 3C 345 at 1.6 GHz on 2016 March 30 with
RadioAstron and 18 ground-basedradio telescopes in full polarisation mode.
Results.
Our images, in both total intensity and linear polarisation, reveal a complex jet structure at 300 µ as angular resolution,corresponding to a projected linear scale of about 2 pc or a few thousand gravitational radii. We identify the synchrotron self-absorbedcore at the jet base and find the brightest feature in the jet 1.5 mas downstream of the core. Several polarised components appear inthe Space VLBI images that cannot be seen from ground array-only images. Except for the core, the electric vector position anglesfollow the local jet direction, suggesting a magnetic field perpendicular to the jet. This indicates the presence of plane perpendicularshocks in these regions. Additionally, we infer a minimum brightness temperature at the largest ( u , v )-distances of 1 . × K in thesource frame, which is above the inverse Compton limit and an order of magnitude larger than the equipartition value. This indicateslocally e ffi cient injection or re-acceleration of particles in the jet to counter the inverse Compton cooling or the geometry of the jetcreates significant changes in the Doppler factor, which has to be >
11 to explain the high brightness temperatures.
Key words. radio continuum: galaxies – galaxies: active – galaxies: jets – galaxies: magnetic fields – quasars: individual: 3C 345 (cid:63)
The reduced images (FITS format) are only available in elec-tronic form at the CDS via anonymous ftp to cdsarc.u-strasbg.fr(130.79.128.5) or via http://cdsweb.u-strasbg.fr/cgi-bin/qcat?J/A+A/
1. Introduction
A fraction of accreting supermassive black holes in the centresof active galactic nuclei (AGN) produce collimated relativisticoutflows (jets) manifesting themselves through broadband con-tinuum emission from the radio to gamma-ray regime. Blazarsare a subclass of AGN, where the jet is closely aligned with theline of sight to the observer. Strong Doppler boosting in these
Article number, page 1 of 10 a r X i v : . [ a s t r o - ph . H E ] F e b & A proofs: manuscript no. MAIN sources makes them brighter and thus easier to detect comparedto non-aligned AGN, and this makes them ideal laboratories forthe studies of jets.The physical processes of the formation of jets are still ac-tively debated. Two promising jet launching mechanisms assumethat either the jet is launched from the accretion disc (Blandford& Payne 1982) or from the rotating magnetosphere of the super-massive black hole itself (Blandford & Znajek 1977). The EventHorizon Telescope (EHT) observations of M87 (Event HorizonTelescope Collaboration et al. 2019) suggest that the jet is pow-ered by magnetic fields anchored in the black hole, as postulatedby the Blandford-Znajek mechanism. It might also be a com-bination of both launching mechanisms, as suggested from
Ra-dioAstron observations of 3C 84 (Giovannini et al. 2018), whichreveal a limb-brightened jet with a radius of 250 gravitationalradii ( r G ) already at a distance of 350 r G from the central en-gine. In either case, a dynamically important magnetic field isthought to play a crucial role in jet formation (Meier et al. 2001;Zamaninasab et al. 2014). The strength and morphology of themagnetic field within the innermost 10 to 10 r G can thereforegive crucial insight on how jets form and how they collimate andaccelerate on parsec scales.Often, the angular resolution of ground-based very longbaselines interferometry (VLBI) arrays is insu ffi cient to probethe innermost regions of distant AGN (Kovalev et al. 2020b).With the inclusion of a space-borne radio telescope orbiting theEarth into the array, the maximum baselines can be extended toseveral Earth diameters ( D ⊕ ), e ff ectively increasing the achievedangular resolution. The RadioAstron project (Kardashev et al.2013) began its in-orbit operations in 2011, with the
Spektr-R spacecraft launched on 2011 July 18, and remained operationaluntil January 2019. It was equipped with a 10-m dish and had ahigh-eccentricity elliptical orbit with a major axis of 350000 km,or ∼ D ⊕ , allowing for unprecedented µ as resolution at ob-serving frequencies of 0.32, 1.6, 4.8 and 22 GHz. The space ra-dio telescope (SRT) also provided, for the first time in SpaceVLBI, full polarisation capabilities, at 0.32, 1.6, and 22 GHz.The RadioAstron key science project (KSP) on AGN polar-isation (see Chapter 4.1 in Bruni et al. 2020, for the project de-scription) aims to develop, commission, and exploit the unprece-dented high angular resolution polarisation capabilities of
Ra-dioAstron to probe the innermost regions of AGN jets and theirmagnetic fields. With this it is possible to accurately determinepotential Faraday rotation gradients (e.g. Gómez et al. 2011; Za-maninasab et al. 2013) at the most compact angular scales, re-vealing changes of the magnetic field within the jet. Within theKSP, several of the brightest and also highly polarised AGN havebeen observed within the first four
RadioAstron observing peri-ods, AO-1, 2, 3 and 4, between 2013 and 2017. Results of thesepolarisation observations are reported for 0642 +
449 in Lobanovet al. (2015), for BL Lac in Gómez et al. (2016) and 0716 + + + ∼ ◦ (Pushkarev et al. 2009; Schinzel et al.2012) and exhibits apparent superluminal motions with speedsof ∼ c (e.g. Zensus et al. 1995; Schinzel et al. 2012; Listeret al. 2019). The source underwent several flaring episodes in theoptical, γ -rays and at radio wavelengths. Schinzel et al. (2012)were able to link the ejection of superluminal components to flar- − − − U / M λ − − − V / M λ l og ( T b , m i n / K ) Fig. 1. ( u , v )-coverage of the observations of 3C 345 described in thispaper. The colour coding shows the minimum brightness temperature T b , min estimated from the visibility amplitudes (see section 5.3 andLobanov (2015)). ing events in 2009. Our observations are close to a local maxi-mum in flux density of 3C 345 (see the OVRO archive ).We assume a flat Λ CDM cosmology with Ω m = . Ω Λ = .
7, and H =
70 km s − Mpc − (Planck Collaboration et al.2014), so that 1 milliarcsecond (mas) corresponds to 6.6 pc pro-jected distance for a redshift of z = .
593 ( D A = .
37 Gpc)(Marziani et al. 1996) for 3C 345.
2. Observations
Observations of 3C 345 at a central frequency of 1 . Spektr-R spacecraft.The ground array consisted of the Very Long Baseline Array(VLBA) with Brewster (BR), Fort Davis (FD), Hancock (HN),Kitt Peak (KP), Los Alamos (LA), North Liberty (NL), OwensValley (OV), Pie Town (PT) and Saint Croix (SC) (9 stations to-tal), the Green Bank Telescope (GB), and eight European VLBInetwork (EVN) stations (E ff elsberg (EF), Hartebeesthoek (HH),Jodrell bank (JB), Medicina (MC), Robledo (RO), Svetloe (SV),Torun (TR) and Zelenchukskaya (ZC)). Six more stations shouldhave been observing, but had technical problems (Badary (BD),Mauna Kea (MK), Onsala (ON25), Sheshan (SH), Urumqi (UR)and Westerbork (WB1)). The resulting ( u , v )-coverage from theremaining stations including the space baselines is shown inFig. 1.The data were recorded in dual-polarisation mode (Right-hand circular (RCP) and Left-hand circular (LCP) polarisations),with four intermediate frequency bands (IFs) of 16 MHz band-width each, yielding 64 MHz total bandwidth for the ground sta-tions, and two IFs for the space antenna, yielding 32 MHz band-width. The data were correlated with the Space VLBI dedicatedversion of the DiFX software correlator, developed and run atthe MPIfR in Bonn (Bruni et al. 2016).
3. Data reduction and calibration
The data were calibrated using standard
AIPS (Greisen 2003)procedures. The amplitudes were calibrated using the systemtemperatures ( T sys ) measured at the telescopes, where for SV andZC we used median values due to sparse T sys data. For RO, no https://sites.astro.caltech.edu/ovroblazars/ Astronomical Image Processing Software of the National Radio As-tronomy Observatory, USA;
Article number, page 2 of 10. M. Pötzl et al.: Probing the innermost regions of AGN jets and their magnetic fields with
RadioAstron T sys measurements were available, so default values were used.Due to severe amplitude miscalibration, that could not be re-solved later in the imaging process, stations JB and RO weredropped from further analysis, leaving us with 16 ground sta-tions. For the SRT, the accuracy of the a priori amplitude calibra-tion is considered at the level of 10-15 % (Kovalev et al. 2014).For the VLBA stations, an amplitude accuracy of ∼
5% can beexpected (Sokolovsky et al. 2011). Typical amplitude errors ofsome other stations are given in Lobanov et al. (2015), for exam-ple. For those stations with less accurate amplitude calibration,we used the well-calibrated antennas to gauge the overall cali-bration at the imaging stage. For the phase calibration (fringe-fitting), we performed a global antenna-based fringe-fitting withthe task
FRING , applying an signal-to-noise ratio (S / N) thresh-old of 6. We tested whether the number and the S / N of solu-tions could be improved by first phasing up the ground array, andthen using an exhaustive baseline search with baseline stackingwhile solving for the SRT. However, since it did not improvethe fringe detection rate, we calibrated the whole array at onceincluding the SRT. With this, ground-space fringes were foundup to 9 D ⊕ . We did not observe decorrelation due to a possibletime-dependent phase rate, that is to say the acceleration termwas small throughout the experiment. The receiver bandpass wascalibrated using standard AIPS routines, and for antennas withno good solutions for the bandpass, we flagged the outer 5 spec-tral channels on either side of IFs to minimise bandpass e ff ects. The phase delay between RCP and LCP signals was calibratedusing the task
RLDLY in AIPS . After its application, a residualphase o ff set between RCP and LCP visibilities still remained. Itwas compensated by rotating the electric vector position angles(EVPAs) in the final polarisation map so that their directions,if convolved with a large beam, aligned with the correspond-ing vectors obtained in single dish data. For that purpose weused polarisation observations of the target source made withthe E ff elsberg Telescope at 1 . > ff VLA observations (Sam-bruna et al. 2004). The E ff elsberg observations yield a total fluxdensity of S ν = . ± .
14 Jy, linearly polarised flux density S ν, P = . ± .
01 Jy ( S ν, P / S ν = . ± .
23 %) and χ = ± ◦ .Here and in the following, P = (cid:112) Q + U is the linearly po-larised intensity calculated from Stokes Q and U , m = P / I is thefractional polarisation (total intensity denoted by Stokes I ), and χ = . × arctan( U / Q ) is the EVPA, measured from north to east.The instrumental polarisation (the telescopes’ D -terms) wascalibrated using the AIPS task
LPCAL and the total intensity(Stokes I ) image of the source as input. The imaging proce-dure for the Stokes I map is presented in Sect. 4. The LPCAL task assumes constant fractional polarisation for defined sub-components of the total intensity structure, where several sub-components were automatically generated with the task
CCEDT to reflect the complex source structure. The resulting D -termswere generally within ∼
10 % for all antennas, except for GBand SV. For GB a likely explanation for the poor D -term deter-mination is the small parallactic angle coverage, as it only ob-served for a few scans. GB and SV were henceforth flagged outin the polarisation analysis because of the insu ffi cient instrumen-tal purity in our observations. Notably, we again confirm the po- − − − R e l a t i v e D E C [ m a s ] Fig. 2.
Total intensity image of 3C 345 at 1.6 GHz with the ground arraydata (orange scale) and all data including the space baselines (blue con-tours). The di ff erent beam sizes are displayed in the bottom right corner.We reach a resolution of 1 . × .
32 mas with
RadioAstron . Contour lev-els are in percents of peak emission of 0 .
39 Jy / beam: 2.83, 4, 5.65, 8,11.31, 16, 22.63, 32, 45.25, 64, 90.51. For the ground array image, theresolution is 3 . × . . .
68 Jy. larisation capabilities of the SRT which demonstrated the instru-mental polarisation of ∼
10 %, in agreement with the previouslyreported values (Pashchenko et al. 2015; Lobanov et al. 2015;Gómez et al. 2016). For the SRT’s LCP, we got D -term valuesof 6.6 % and 9.6 % for the two IFs, respectively. For RCP, weobtain D -term values as low as 3.0 % and 3.5 %, respectively.We used only the target source 3C 345 for the D -term estima-tion. We found this su ffi cient since the main target was strongenough. The calibrator sources (3C 286 and OJ 287) were notwell suited for D -term estimates since they were either observedonly with a subset of telescopes with a sparse parallactic anglecoverage and / or showed too much structure. Therefore we alsodo not provide error estimates of the D -terms. In addition to the
RadioAstron data at 1.6 GHz, we made use ofarchival MOJAVE observations at 15 GHz (Lister et al. 2018).These were conducted on 2016 March 5, less than a month apartfrom our RadioAstron observations, and will provide a reason-able comparison. This is justified by the observed variability inthe radio light curves, and by the median velocity of jet com-ponents of 0 . − . The 15 GHz data are available at theMOJAVE webpage .
4. Imaging
The data were imaged using the
Difmap software (Shepherd1997). Before imaging, we averaged the data into 90 s intervals.We first created an image of the ground array only using standard Monitoring Of Jets in Active galactic nuclei with VLBA Experiments Article number, page 3 of 10 & A proofs: manuscript no. MAIN
Fig. 3.
Visibility amplitudes (top) and phases (bottom) of the final cal-ibrated data. Blue data points show ground only data, while black datapoints highlight the space baselines. The source clean model is shownin red.
Fig. 4.
Visibility amplitudes (top) and phases (bottom) of data only fromground-based antennas. The source clean model is shown in red. clean and self-calibration procedures (see colour scale in Fig. 2).We then created a map using the baselines to the SRT based onthe ground-only map, where we applied phase self-calibrationdown to a time interval of 3 min and amplitude self-calibrationonly as overall gain factor to the SRT. The amplitude correctionswe applied for the SRT were 8 % and 3 % for IF3 and IF4, re-spectively.Fig. 2 shows the images of 3C 345 with the ground array onlymap in colour and with the full Space VLBI resolution in bluecontours, using a uniform weighting scheme in both cases. Thesynthesised beam size is 1 . × .
32 mas with
RadioAstron and3 . × . RadioAstron over the ground-only image is about a factor of 7along the jet direction. The image rms noise is ∼ . / beam.The visibility amplitudes and phases as a function of projected( u , v )-distance are displayed in Fig. 3 and Fig. 4 for the wholearray including the SRT and for the ground-array only data, re-spectively. An exemplary plot of the closure phases for the trian-gle R2-NL-ZC (R2 designating the SRT) is shown in Fig. 5 withthe source model displayed in red.The achieved resolution of ∼ µ as (minor axis FWHMbeam size) corresponds to a projected length of 2 pc or between ∼ ∼ r G for a black hole mass ranging between of M BH ∼ × M (cid:12) and M BH ∼ × M (cid:12) (Gu et al. 2001;Shen et al. 2011).
5. Results and discussion
Our images reveal several components in the inner 10 mas of thejet, which could not be resolved with data from the ground arrayonly. We find that the easternmost feature at the jet base is notthe brightest component, which is a characteristic already seenin 3C 345 with VSOP (VLBI Space Observatory Programme),the predecessor of
RadioAstron (Klare et al. 2000, 2005). Thisfeature likely corresponds to a partly synchrotron self-absorbedcore, which we designate as the ‘core’ in our subsequent anal-ysis. We also observe a visibly curved jet structure in the fewinnermost mas of the jet, where the jet direction changes rapidly.We consider the weak easternmost feature of the jet visible at theedge of Fig. 2 to be rather an imaging artefact than an indicationof a counter-jet, as it is only about 3 times the noise level.We fitted the visibilities with circular Gaussian componentsin the inner 8 mas of the jet using
Difmap . To find the minimumnecessary number of components required to describe the struc-ture, we used the criterion presented in Schinzel et al. (2012).The fitted flux densities, positions, and sizes are listed in Ta-ble 1 and displayed in Fig. 7. The errors on those quantities werealso calculated according to Schinzel et al. (2012). The errors inpolarised intensity, fractional polarisation and EVPA have beencalculated from the map rms errors in Stokes Q and U . In ad-dition, we calculated the brightness temperature of each com-ponent, which we describe in detail in Sec. 5.3. A minimumpossible size of a source structure which can still be resolvedby an interferometer is dependent on the S / N, which is ∼ . × .
32 mas, we estimate that features with an angular extentof θ lim = π (cid:115) π log(2) b maj b min log (cid:32) S / NS / N − (cid:33) ∼ µ as (1)can be probed by our observations, according to Lobanov(2005). Our component sizes are all larger than this limit. Liu et al. (2018) have investigated the variability properties ofa large samples of AGN observed with
RadioAstron in termsof their modulation index ¯ m at 5 GHz. 3C 345 did not showany signs of intra-day variability (IDV), although the sourceis known to exhibit long-term (months to years) variability, asshown in observations at the Green Bank interferometer at 2 and8 GHz (Rickett et al. 2006), with ¯ m = . m
15 GHz = .
129 for long-term variability. The overall lack ofIDV is not surprising considering the high Galactic latitude of3C 345, as IDV is likely caused by scintillation due to the Galac-tic interstellar medium (Rickett et al. 2006). From our small-est component size (L3), we can estimate the shortest variabilitytimescale according to Jorstad et al. (2017): τ ∼ . θ D L δ (1 + z ) . (2) Article number, page 4 of 10. M. Pötzl et al.: Probing the innermost regions of AGN jets and their magnetic fields with
RadioAstron
02 03 04 05 06 07 08 09 10UTC [h]15010050050100150 C l o s u r e P h a s e [] R2 NL ZC
Fig. 5.
Closure phases of the triangle R2-NL-ZC, where R2 designatesthe SRT. The source clean model is shown in red. The
RadioAstron perigee occurred during the last scans.
Here τ is the variability timescale in years, θ is the compo-nent FWHM in mas (as given in Table 1), D L is the luminos-ity distance and δ is the Doppler factor. Considering Dopplerfactors between 10 and 25 (see Sect. 5.3), the shortest variabil-ity timescale is estimated to be between 1 to 4 months, whereonly the lower estimate is broadly consistent with τ = . µ as is expected for sources with strong variability at higherfrequencies, while not showing signs of scintillation due to theISM (Koay et al. 2018). The ability of an interferometer to measure the brightness tem-perature is in principle independent from the observing wave-length, and depends only on the projected interferometer base-line (Kovalev et al. 2005). Accordingly,
RadioAstron uniquelyprobes the highest brightness temperatures (e.g. Gómez et al.2016; Kovalev et al. 2016; Kutkin et al. 2018; Pilipenko et al.2018; Kovalev et al. 2020a; Kravchenko et al. 2020).We calculate the brightness temperature in two ways. Wefirst estimate the minimum brightness temperature T b , min as wellas the maximum brightness temperature T b , max from the visibil-ity data and the visibility errors according to Lobanov (2015).Calculating the minimum brightness temperature requires the as-sumption of a circular or axially symmetric brightness distribu-tion. For the maximum brightness temperature, in addition, onehas to assume that the structure at the probed scale is marginallyresolved. Following that, we calculate the brightness temperaturefrom the fitted flux densities and FWHM of the modelfit compo-nents explained in Sect. 5.1. The values of the estimated bright-ness temperature from both methods are presented in Fig. 6, aswell as in the ( u , v )-coverage plot in Fig. 1 for the first method.We also did the same calculations for the archival MOJAVE ob-servations described in Sect. 3.2. We go into more detail aboutthe comparison of the data sets in Sect. 5.4.The estimates of the brightness temperature from the visi-bilities are most accurate for baselines >
200 M λ , where T b , min and T b , max provide a reliable constraint on T b (Lobanov 2015).As shown in Fig. 6, we binned the data into 10M λ bins, andfor the bin at the largest ( u , v )-distances we get an average T b , min = . × K and T b , max = . × K, which arethe values that we use in our following analysis.From the Gaussian modelfits, we calculate T b as: T b = S ν c k B ν Ω , (3)where S ν denotes the flux density, c the speed of light, k B theBoltzmann constant, ν the observing frequency and Ω the com-ponent solid angle. The highest T b , that we calculate for compo-nent L3 (see Table 1 and Fig. 7), lies between T b , min and T b , max at 8 . × K, close to the maximum value. So our values of T b , min and T b , max seem to provide a reasonable bracketing for thehighest component brightness temperature. Other studies alsofound a good agreement for both estimates (e.g. Nair et al. 2019).A decline in component brightness temperature downstream ofcomponent L3 along the jet is observed, that can be explained inthe framework of a jet with regions of relativistic plasma that ex-pand adiabatically and lose energy via radiation while travellingdownstream (Pushkarev & Kovalev 2012).It is generally thought that, for incoherent synchrotronsources such as AGN, if T b increases to values larger thanabout 10 K, the amount of energy released due to the InverseCompton (IC) process becomes too large to be sustainable. This‘IC catastrophe’ reduces T b again to values of about 10 K ontimescales of a day (Kellermann & Pauliny-Toth 1969). Thisthreshold is referred to as the IC limit. Readhead (1994) arguedthat the equipartition brightness temperature T b , eq might be a bet-ter constraint for the upper value of the brightness temperature,which is generally more of the order of 10 K. It assumes anequipartition of the energy of particles and magnetic fields. For3C 345, we estimate a value very close to 10 K ( T b , eq = . )as well, using equation (4b) in Readhead (1994) with an opti-cally thin spectral index α = − . S ν ∝ ν α ) (Liu et al. 2018)and our single dish flux density as a proxy for the peak flux den-sity. The equation gives an upper limit on T b , eq in case we are notusing the actual peak flux density of the spectrum. We considerthe source redshift z and the Doppler boosting according to T b , obs = δ T b , int (1 + z ) , (4)where T b , obs denotes the brightness temperature in the observer’sframe and T b , int in the source frame. The Doppler factor is de-noted by δ = (cid:112) − β (1 − β cos( θ )) − , where β is the jet bulkvelocity in units of the speed of light and θ is the jet viewingangle. We take δ = . ± . T b = . × K in the source frame.The visibility amplitudes imply the presence of emitting re-gions with observed brightness temperature in excess of thisIC limit. This suggests either locally e ffi cient injection or re-acceleration of particles in the jet to counter the inverse Comp-ton cooling, or that the geometry of the jet creates significantchanges in the Doppler factor, resulting in the su ffi ciently largeDoppler boosting. E ffi cient particle re-acceleration could, for ex-ample, be achieved with turbulent plasma flowing down the jetand crossing a standing shock (Marscher 2014). Alternatively,magnetic reconnection events can e ffi ciently accelerate particles(e.g. Sironi et al. 2015). Article number, page 5 of 10 & A proofs: manuscript no. MAIN
Fig. 6.
Minimum brightnesstemperature T b , min and maxi-mum brightness temperature T b , max as a function of ( u , v )-distance. The values wereestimated from the visibilitiesfollowing Lobanov (2015). Thesolid (dashed) lines show theaverage T b , min ( T b , max ) in bins of10 M λ for RadioAstron (black)and MOJAVE (red), while thepoints show the individual datavalues with the same colourscheme. The dashed horizontallines show T b calculated fromGaussian modelfits (see Ta-ble 1). T b , min and T b , max providea reasonable bracketing for thebrightness temperature at leastfor the RadioAstron data.
UV distance / M λ l og ( T b / K ) T b , min T b , max T b , min
15 GHz MOJAVE T b , max
15 GHz MOJAVE
Fig. 7.
RadioAstron image of thetotal intensity and polarised emis-sion of 3C 345 at 1.6 GHz. Map ofthe polarised intensity P in colour-scale, overlaid with contours dis-playing the total intensity emis-sion. The beam size is displayed onthe bottom right with a resolutionof 1 . × .
32 mas. The lines showthe EVPAs the length of which isproportional to P . Contours lev-els are (% of peak emission of0 .
39 Jy / beam): −
2, 2, 2.83, 4, 5.65,8, 11.31, 16, 22.63, 32, 45.25, 64,90.51. − − − −
202 Relative RA [mas] − − − R e l a t i v e D E C [ m a s ] Core L1 L2 L3 L4L5 L6 L7 . . . . . . . . Polarised Intensity [mJy / beam] Article number, page 6 of 10. M. Pötzl et al.: Probing the innermost regions of AGN jets and their magnetic fields with
RadioAstron
Table 1.
Circular Gaussian model fit parameters and inferred brightness temperature from the
RadioAstron and MOJAVE data. (1) (2) (3) (4) (5) (6) (7) (8)Comp. Flux Distance P.A. Size T b m χ [mJy] [mas] [ ◦ ] [mas] [K] [%] [ ◦ ] RadioAstron . ±
45 1.48 ± ± ± . ± . × ± ± ±
65 0.92 ± ± ± . ± . × ± ± ±
37 0.19 ± ± ± . ± . × ± ± ±
29 0.25 ± − . ± ± . ± . × ± ± ±
30 1.09 ± − . ± ± . ± . × ± ± ±
24 1.76 ± − . ± ± . ± . × ± ± ±
108 3.73 ± − . ± ± . ± . × ± ± ±
144 5.69 ± − . ± ± . ± . × ± ± ±
38 0.083 ± ± < . > . × ± ± ±
47 0.077 ± − . ± ± . ± . × ± ± ±
18 0.34 ± − . ± ± . ± . × ± ± ±
22 0.871 ± − . ± ± . ± . × ± ± ±
11 1.39 ± − . ± ± . ± . × ± − . ± ±
20 1.84 ± − . ± ± . ± . × ± ± ±
23 2.43 ± − . ± ± . ± . × ± ± ±
22 5.37 ± − . ± ± . ± . × ± ± ±
54 6.59 ± − . ± ± . ± . × ± ± N ote : Columns display the (1) component name (see Fig. 7 and 8), (2) flux density, (3) radial distance from the total intensity peak, (4)component position angle, (5) component FWHM size, (6) brightness temperature, (7) fractional polarization and (8) Electric Vector PositionAngle. Where the component size was smaller than the minimal resolvable source size in the map, we provide an upper limit. Errors arepurely statistical errors and may be underestimated. See text for details. Doppler boosting due to changes in the viewing angle alongthe jet has been investigated by Qian et al. (1996) for 3C 345,who find that the position and flux variability of a compo-nent could be explained with helical motion. A similar well-pronounced case of a helical jet pattern is known, for example,in the source 1156 +
295 (Hong et al. 2004; Zhao et al. 2011)or in 2136 +
141 (Savolainen et al. 2006). Variations in the jetorientation for the innermost 1 mas of the jet in 3C 345 withinabout 60 ◦ over 15 years also support such a scenario (Lister et al.2013), and the helical motion could possibly be explained byprecession of the accretion disc (Lobanov & Roland 2005). Qianet al. (1996) find variability in the Doppler factor between 7 and10.8, caused by the di ff erence in viewing angle. This would beinsu ffi cient to explain the high T b , min in our RadioAstron obser-vations, where we would need δ >
11. Schinzel et al. (2012)have investigated the Doppler factor for di ff erent components in3C 345 observed between 2008 and 2010 at 43 GHz. They findDoppler factors as high as 23 for one component. Jorstad et al.(2017) also find maximum Doppler factors of about 17. Suchhigh Doppler boosting could readily explain the high observedbrightness temperatures, however we can not identify individualcomponents in our RadioAstron map with the components pre-sented in these works.Still, our inferred brightness temperatures easily exceed theestimated T b , eq by an order of magnitude, suggesting that the jetin 3C 345 is not in equipartition. This indicates a flaring event,that is also supported by the radio light curve and the bright po-larization component we observe at 1 . RadioAstron , for ex-ample in 3C 273 (Kovalev et al. 2016) and BL Lac (Gómez et al. 2016). Kovalev et al. (2016) suggest refractive substructure asa possible source of high observed T b , which has been inves-tigated by Johnson et al. (2016). We also test the possible ef-fect of scattering on our estimated brightness temperature. Thee ff ect is more prominent at longer wavelengths and starts con-tributing to the observed signal at 18 cm at any baseline largerthan ∼ ,
000 km (5 . D ⊕ ), if the flux density at zero spacing is > T b is in the range of the values that we also obtainhere (see Fig. 1 and 3 in Johnson et al. 2016).We used Eq. 3 in Johnson et al. (2016) to calculate T b , min ,which accounts for both refractive substructure and angularbroadening. The former will lead to an overestimate, the latterto an underestimate of T b , min : T b , min = . × K (cid:18) B km (cid:19) / (cid:32) F B
20 mJy (cid:33) (5) × (cid:32) D (cid:33) / (cid:18) λ
18 cm (cid:19) (cid:32) θ scatt µ as (cid:33) − / , where B denotes the baseline length and F B the measured fluxdensity at this baseline. Adopting the NE2001 model (Cordes& Lazio 2002) for the Galactic distribution of free electrons,we estimate for the galactic coordinates of 3C 345 ( l = ◦ . b = ◦ .
95) an angular broadening of θ scatt = .
28 mas at 18 cmwavelength. At an assumed distance to the scattering screen of D = T b , min = . × K, which is evenhigher than our previous estimate without considering scatter-ing. So we conclude that refractive substructure likely does notplay a role for our observations. This is not surprising as 3C 345lies at high galactic latitude, so there is likely not enough scatter-ing material along the line of sight to cause significant refractivesubstructure.
Article number, page 7 of 10 & A proofs: manuscript no. MAIN
Fig. 8.
MOJAVE image of thetotal intensity and polarisedemission of 3C 345 at 15 GHz.Map of the polarised inten-sity P in colour-scale, overlaidwith contours displaying the to-tal intensity emission. The beamsize is displayed on the bot-tom right with a resolution of0 . × .
46 mas. The lines showthe EVPAs the length of whichis proportional to P . Contourslevels are (% of peak emis-sion of 3 .
44 Jy / beam): − . − − − − −
202 Relative RA [mas] − − R e l a t i v e D E C [ m a s ] Core U1 U2 U3 U4 U5 U6 U7 U8
20 40 60 80 100 120
Polarised Intensity [mJy / beam] We have compared the brightness temperatures obtained fromthe 1.6 GHz Space VLBI data with estimates obtained via thesame methods using MOJAVE 15 GHz observations. To make areasonable comparison of the two data sets, we applied a filteron the visibilities of the MOJAVE data, so that only data that oc-cupy the same location (within 10 %) in the ( u , v )-plane as our1.6 GHz data are used for the MOJAVE brightness temperatureestimate. We present T b , min as a function of ( u , v )-radius in Fig. 6.As for the RadioAstron data, we plot both the brightness tem-peratures obtained from the visibilities as well as those obtainedfrom modelfits.The range between T b , min and T b , max is larger for the MO-JAVE data compared to our RadioAstron data. This is likely dueto the underestimated visibility errors in the former, so T b , max might not be well defined, leading to a poor determination of T b , max . Nevertheless, the maximum brightness temperature frommodelfits to the MOJAVE data still lies between the limits pro-vided by T b , min and T b , max . Overall, the observed T b for the Gaus-sian components is higher at 1.6 GHz compared to 15 GHz. Thisis expected given Eq. 3 and the similar covered ( u , v )-distancesin both data sets. We concentrate on the comparison between thedi ff erent T b , min in the following.We see significantly higher values for T b , min in the 1.6 GHz RadioAstron data compared to the 15 GHz MOJAVE data. Thatis expected, as T b ∝ λ , and any di ff erences in the ratio T b , min , RA / T b , min , MOJ that di ff ers from (18 cm / as a functionof ( u , v )-distance seen in Fig. 6 can be interpreted as a spectralindex that is di ff erent from zero. Indeed we observe a trend ofdecreasing ratios for increasing ( u , v )-distances, which can beinterpreted as a progressive change of the jet opacity from opti-cally thin to thick with increasing baseline lengths. We observe multiple polarised components, the brightest beingroughly coincident with the total intensity peak close to compo-nent L2 (see Fig. 7). We see more polarised structure ∼ m = . ± .
67 %), where synchrotron self-absorption likelyalso leads to significant depolarisation. It can not be ruled outthat depolarisation due to blending of di ff erent unresolved fea-tures in the observing beam also contributes to the diminishedpolarisation degree. An optically thick core region has been ob-served already between 8.1 and 15.4 GHz by MOJAVE (Hovattaet al. 2014). At the total intensity peak the fractional polarisationreaches m = . ± .
51 %, where at the location of componentL6 it reaches m = . ± .
40 %. We observe a degree of polarisa-tion up to 60 % ∼ ∼
70 %. Kravchenko et al. (2020)also observed up to 50 % degree of linear polarisation in the jetof 0716 + / N region in Stokes I , which drives the uncertainty of this value to be ∼
20 %. In thatcase we can not rule out the possibility that uncertainties in the D -term estimation a ff ect the observed degree of polarisation inthat region substantially.Overall the EVPAs seem to be well aligned with the localjet direction, which was also observed at 43 GHz by MacDonaldet al. (2017). However, in the core, the EVPAs are oriented closerto the perpendicular orientation relative to the jet. This is consis-tent with a possible rotation of the EVPAs by π/ ff ects (Gomez et al. 1994; Gabuzda & Gómez 2001), as the coreis most likely optically thick (Pötzl et al., in prep., from now onpaper II). This would indicate a B-field closer to the perpendicu-lar relative to the jet direction also in the core. However, Wardle(2018) argues that higher optical depths of between 6 and 7 areneeded for a π/ Article number, page 8 of 10. M. Pötzl et al.: Probing the innermost regions of AGN jets and their magnetic fields with
RadioAstron − − − −
202 Relative RA [mas] − − − R e l a t i v e D E C [ m a s ] − . − . − . − . − . − . log( m ) Fig. 9.
Same as Fig. 7, but displaying the logarithm of fractional polar-isation log( m ) in colour-scale. along the jet can explain the observed bright polarised featureswith the EVPAs aligned with the jet direction, where the mag-netic field is quenched perpendicular to the jet direction (War-dle et al. 1994). Earlier multi-frequency, multi-epoch studies of3C 345 have favoured this scenario (Ros et al. 2000). Lobanov &Zensus (1999) argued that shocks likely do not play a significantrole in the dynamics and emission outside of the core region in3C 345. In this case, the EVPAs ∼ ffi cient to confirm the pres-ence of a helical magnetic field. This will be further tested withan analysis of the Rotation Measure (paper II), as well as withhigher resolution RadioAstron observations at 22 GHz made inMay 2016, close to our epoch at 1.6 GHz.Comparing our polarisation map at 1 . ∼ ∆ χ ∝ λ , thus the Fara-day rotation is stronger at lower frequencies. While we presenta deeper analysis of Faraday rotation in future work using a setof data at multiple frequencies, we briefly discuss the possiblemagnitude of Faraday rotation.Hovatta et al. (2012) have studied the RM in many AGN jets,with observations at four frequencies between 8 and 15 GHz.The results showed two distinct regions in 3C 345, one withRM = . ± . − in the core region and another onewith RM = − . ± . − at ∼ . ∼ ◦ , which would signifi-cantly change the EVPAs in the core. Motter & Gabuzda (2017) also studied the RM at four frequencies around 1.6 GHz in sixAGN, one of which was 3C 345. They found RM in the rangeof −
30 rad m − < RM <
30 rad m − , and report a statisticallysignificant RM gradient transverse to the jet direction. This sup-ports the presence of a toroidal magnetic field that may be partof a helical one. However, the di ff erence in beam size comparedto our RadioAstron observations is about a factor of 20 in theeast-west and a factor 10 in the north-south direction, and ourobservations only focus on the innermost ∼
10 mas of the jet.
6. Summary
The main conclusions of the paper are summarised in the follow-ing: – We present Space VLBI images obtained with the
RadioAs-tron mission in both total and linearly polarised intensity ofthe FSRQ 3C 345 at 1.6 GHz with an angular resolution of ∼ µ as. Several compact components that were not iden-tifiable with ground-only VLBI arrays at the same frequencyare resolved in our RadioAstron observations and the SpaceVLBI image reveals the complex, visibly curved inner jetstructure. – We identify several linearly polarised components, with analmost completely depolarised core, a high polarisation peakcoincident with the total intensity peak with ∼ – We compare several estimates of the brightness temperature T b for the RadioAstron data. We infer a minimum observedbrightness temperature of T b , min , obs = . × K and aminimum intrinsic brightness temperature T b , min , int = . × K. The latter is in slight excess of the IC limit, and anorder of magnitude larger than the equipartition brightnesstemperature limit, suggesting that 3C 345 is not in equipar-tition between particle and magnetic field energy during ourobservations. The most likely explanations of this excess areeither a variable Doppler factor ( δ >
11) due to changes inthe jet geometry along the flow or locally e ffi cient particlere-acceleration. We investigated the e ff ect of refractive sub-structure due to the galactic ISM and conclude that it doesnot dominate our estimate. We also confirm that the rangegiven by T b , min and T b , max accurately brackets the actual T b asmeasured from fitting the data with circular Gaussian com-ponents.These conclusions will be further tested with an analysis ofthe RadioAstron data presented here in conjunction with a mul-tiwavelength VLBI dataset in Pötzl et al. (in prep.).
Acknowledgements.
We thank N. R. MacDonald and J.-Y. Kim as well as theanonymous referee for valuable comments to the manuscript. The
RadioAstron project is led by the Astro Space Center of the Lebedev Physical Institute ofthe Russian Academy of Sciences and the Lavochkin Scientific and ProductionAssociation under a contract with the State Space Corporation ROSCOSMOS,in collaboration with partner organizations in Russia and other countries. Partlybased on observations performed with radio telescopes of IAA RAS (FederalState Budget Scientific Organization Institute of Applied Astronomy of Rus-sian Academy of Sciences). The European VLBI Network is a joint facility ofindependent European, African, Asian, and North American radio astronomyinstitutes. Scientific results from data presented in this publication are derived
Article number, page 9 of 10 & A proofs: manuscript no. MAIN from the following EVN project code(s): GG079A. Results of optical position-ing measurements of the Spektr-R spacecraft by the global MASTER RoboticNet (Lipunov et al. 2010), ISON collaboration, and Kourovka observatory wereused for spacecraft orbit determination in addition to mission facilities. The Na-tional Radio Astronomy Observatory and the Green Bank Observatory are fa-cilities of the National Science Foundation operated under cooperative agree-ment by Associated Universities, Inc. This research has made use of data fromthe MOJAVE database that is maintained by the MOJAVE team (Lister et al.2018). Partly based on observations with the 100-m telescope of the MPIfR(Max-Planck-Institut für Radioastronomie) at E ff elsberg. The data were corre-lated at the DiFX correlator (Deller et al. 2011; Bruni et al. 2016) of the MPIfRat Bonn. A.P.L., Y.Y.K. and E.V.K. were supported by the Russian Science Foun-dation (project 20-62-46021). L.I.G. acknowledges support by the CSIRO Dis-tinguished Visitor Programme. T.S. was supported by the Academy of Finlandprojects 274477 and 315721. J.L.G acknowledges the support of the SpanishMinisterio de Economía y Competitividad (grants AYA2016-80889-P, PID2019-108995GB-C21), the Consejería de Economía, Conocimiento, Empresas y Uni-versidad of the Junta de Andalucía (grant P18-FR-1769), the Consejo Superiorde Investigaciones Científicas (grant 2019AEP112), and the State Agency forResearch of the Spanish MCIU through the Center of Excellence Severo Ochoaaward for the Instituto de Astrofísica de Andalucía (SEV-2017-0709). This re-search has made use of NASA’s Astrophysics Data System. This research hasmade use of adstex ( https://github.com/yymao/adstex ). This research hasmade use of the NASA / IPAC Extragalactic Database (NED), which is operatedby the Jet Propulsion Laboratory, California Institute of Technology, under con-tract with the National Aeronautics and Space Administration.
References
Angelakis, E., Fuhrmann, L., Myserlis, I., et al. 2019, A&A, 626, A60Blandford, R. D. & Payne, D. G. 1982, MNRAS, 199, 883Blandford, R. D. & Znajek, R. L. 1977, MNRAS, 179, 433Bruni, G., Anderson, J., Alef, W., et al. 2016, Galaxies, 4, 55Bruni, G., Gómez, J. L., Casadio, C., et al. 2017, A&A, 604, A111Bruni, G., Savolainen, T., Gómez, J. L., et al. 2020, Advances in Space Research,65, 712Cordes, J. M. & Lazio, T. J. W. 2002, arXiv e-prints [ astro-ph/0207156 ]Deller, A. T., Brisken, W. F., Phillips, C. J., et al. 2011, PASP, 123, 275Event Horizon Telescope Collaboration, Akiyama, K., Alberdi, A., et al. 2019,ApJ, 875, L5Gabuzda, D. C. & Gómez, J. L. 2001, MNRAS, 320, L49Ghisellini, G., Padovani, P., Celotti, A., & Maraschi, L. 1993, ApJ, 407, 65Giovannini, G., Savolainen, T., Orienti, M., et al. 2018, Nature Astronomy, 2,472Gomez, J. L., Alberdi, A., Marcaide, J. M., Marscher, A. P., & Travis, J. P. 1994,A&A, 292, 33Gómez, J. L., Lobanov, A. P., Bruni, G., et al. 2016, ApJ, 817, 96Gómez, J. L., Roca-Sogorb, M., Agudo, I., Marscher, A. P., & Jorstad, S. G.2011, ApJ, 733, 11Greisen, E. W. 2003, Astrophysics and Space Science Library, Vol. 285, AIPS,the VLA, and the VLBA, ed. A. Heck, 109Gu, M., Cao, X., & Jiang, D. R. 2001, MNRAS, 327, 1111Hong, X. Y., Jiang, D. R., Gurvits, L. I., et al. 2004, A&A, 417, 887Hovatta, T., Aller, M. F., Aller, H. D., et al. 2014, AJ, 147, 143Hovatta, T., Lister, M. L., Aller, M. F., et al. 2012, AJ, 144, 105Johnson, M. D., Kovalev, Y. Y., Gwinn, C. R., et al. 2016, ApJ, 820, L10Jorstad, S. G., Marscher, A. P., Morozova, D. A., et al. 2017, ApJ, 846, 98Kardashev, N. S., Khartov, V. V., Abramov, V. V., et al. 2013, Astronomy Re-ports, 57, 153Kellermann, K. I. & Pauliny-Toth, I. I. K. 1969, ApJ, 155, L71Klare, J., Zensus, J. A., Lobanov, A. P., et al. 2005, in Astronomical Society ofthe Pacific Conference Series, Vol. 340, Future Directions in High ResolutionAstronomy, ed. J. Romney & M. Reid, 40Klare, J., Zensus, J. A., Ros, E., & Lobanov, A. P. 2000, in Astrophysical Phe-nomena Revealed by Space VLBI, ed. H. Hirabayashi, P. G. Edwards, &D. W. Murphy, 21–24Koay, J. Y., Jauncey, D. L., Hovatta, T., et al. 2019, MNRAS, 489, 5365Koay, J. Y., Macquart, J. P., Jauncey, D. L., et al. 2018, MNRAS, 474, 4396Kovalev, Y. A., Vasil’kov, V. I., Popov, M. V., et al. 2014, Cosmic Research, 52,393Kovalev, Y. Y., Kardashev, N. S., Kellermann, K. I., et al. 2016, ApJ, 820, L9Kovalev, Y. Y., Kardashev, N. S., Sokolovsky, K. V., et al. 2020a, Advances inSpace Research, 65, 705Kovalev, Y. Y., Kellermann, K. I., Lister, M. L., et al. 2005, AJ, 130, 2473Kovalev, Y. Y., Pushkarev, A. B., Nokhrina, E. E., et al. 2020b, MNRAS, 495,3576Kravchenko, E. V., Gómez, J. L., Kovalev, Y. Y., et al. 2020, ApJ, 893, 68 Kutkin, A. M., Pashchenko, I. N., Lisakov, M. M., et al. 2018, MNRAS, 475,4994Liodakis, I., Zezas, A., Angelakis, E., Hovatta, T., & Pavlidou, V. 2017, A&A,602, A104Lipunov, V., Kornilov, V., Gorbovskoy, E., et al. 2010, Advances in Astronomy,2010, 349171Lister, M. L., Aller, M. F., Aller, H. D., et al. 2018, ApJS, 234, 12Lister, M. L., Aller, M. F., Aller, H. D., et al. 2013, AJ, 146, 120Lister, M. L., Homan, D. C., Hovatta, T., et al. 2019, ApJ, 874, 43Liu, J., Bignall, H., Krichbaum, T., et al. 2018, Galaxies, 6, 49Lobanov, A. 2015, A&A, 574, A84Lobanov, A. P. 2005, arXiv e-prints [ astro-ph/0503225 ]Lobanov, A. P., Gómez, J. L., Bruni, G., et al. 2015, A&A, 583, A100Lobanov, A. P. & Roland, J. 2005, A&A, 431, 831Lobanov, A. P. & Zensus, J. A. 1999, ApJ, 521, 509Lovell, J. E. J., Rickett, B. J., Macquart, J. P., et al. 2008, ApJ, 689, 108MacDonald, N. R., Jorstad, S. G., & Marscher, A. P. 2017, ApJ, 850, 87Marscher, A. P. 2014, ApJ, 780, 87Marziani, P., Sulentic, J. W., Dultzin-Hacyan, D., Calvani, M., & Moles, M.1996, ApJS, 104, 37Meier, D. L., Koide, S., & Uchida, Y. 2001, Science, 291, 84Motter, J. C. & Gabuzda, D. C. 2017, MNRAS, 467, 2648Nair, D. G., Lobanov, A. P., Krichbaum, T. P., et al. 2019, A&A, 622, A92Pashchenko, I. N., Kovalev, Y. Y., & Voitsik, P. A. 2015, Cosmic Research, 53,199Pilipenko, S. V., Kovalev, Y. Y., Andrianov, A. S., et al. 2018, MNRAS, 474,3523Planck Collaboration, Ade, P. A. R., Aghanim, N., et al. 2014, A&A, 571, A16Pushkarev, A. B. & Kovalev, Y. Y. 2012, A&A, 544, A34Pushkarev, A. B., Kovalev, Y. Y., Lister, M. L., & Savolainen, T. 2009, A&A,507, L33Qian, S. J., Krichbaum, T. P., Zensus, J. A., Ste ff en, W., & Witzel, A. 1996,A&A, 308, 395Readhead, A. C. S. 1994, ApJ, 426, 51Richards, J. L., Hovatta, T., Max-Moerbeck, W., et al. 2014, MNRAS, 438, 3058Rickett, B. J., Lazio, T. J. W., & Ghigo, F. D. 2006, ApJS, 165, 439Ros, E., Zensus, J. A., & Lobanov, A. P. 2000, A&A, 354, 55Sambruna, R. M., Gambill, J. K., Maraschi, L., et al. 2004, ApJ, 608, 698Savolainen, T., Wiik, K., Valtaoja, E., et al. 2006, ApJ, 647, 172Schinzel, F. K., Lobanov, A. P., Taylor, G. B., et al. 2012, A&A, 537, A70Shen, Y., Richards, G. T., Strauss, M. A., et al. 2011, ApJS, 194, 45Shepherd, M. C. 1997, in Astronomical Society of the Pacific Conference Series,Vol. 125, Astronomical Data Analysis Software and Systems VI, ed. G. Hunt& H. Payne, 77Sironi, L., Petropoulou, M., & Giannios, D. 2015, MNRAS, 450, 183Sokolovsky, K. V., Kovalev, Y. Y., Pushkarev, A. B., & Lobanov, A. P. 2011,A&A, 532, A38Wardle, J. 2018, Galaxies, 6, 5Wardle, J. F. C., Cawthorne, T. V., Roberts, D. H., & Brown, L. F. 1994, ApJ,437, 122Zamaninasab, M., Clausen-Brown, E., Savolainen, T., & Tchekhovskoy, A.2014, Nature, 510, 126Zamaninasab, M., Savolainen, T., Clausen-Brown, E., et al. 2013, MNRAS, 436,3341Zensus, J. A., Cohen, M. H., & Unwin, S. C. 1995, ApJ, 443, 35Zhao, W., Hong, X. Y., An, T., et al. 2011, A&A, 529, A113en, W., & Witzel, A. 1996,A&A, 308, 395Readhead, A. C. S. 1994, ApJ, 426, 51Richards, J. L., Hovatta, T., Max-Moerbeck, W., et al. 2014, MNRAS, 438, 3058Rickett, B. J., Lazio, T. J. W., & Ghigo, F. D. 2006, ApJS, 165, 439Ros, E., Zensus, J. A., & Lobanov, A. P. 2000, A&A, 354, 55Sambruna, R. M., Gambill, J. K., Maraschi, L., et al. 2004, ApJ, 608, 698Savolainen, T., Wiik, K., Valtaoja, E., et al. 2006, ApJ, 647, 172Schinzel, F. K., Lobanov, A. P., Taylor, G. B., et al. 2012, A&A, 537, A70Shen, Y., Richards, G. T., Strauss, M. A., et al. 2011, ApJS, 194, 45Shepherd, M. C. 1997, in Astronomical Society of the Pacific Conference Series,Vol. 125, Astronomical Data Analysis Software and Systems VI, ed. G. Hunt& H. Payne, 77Sironi, L., Petropoulou, M., & Giannios, D. 2015, MNRAS, 450, 183Sokolovsky, K. V., Kovalev, Y. Y., Pushkarev, A. B., & Lobanov, A. P. 2011,A&A, 532, A38Wardle, J. 2018, Galaxies, 6, 5Wardle, J. F. C., Cawthorne, T. V., Roberts, D. H., & Brown, L. F. 1994, ApJ,437, 122Zamaninasab, M., Clausen-Brown, E., Savolainen, T., & Tchekhovskoy, A.2014, Nature, 510, 126Zamaninasab, M., Savolainen, T., Clausen-Brown, E., et al. 2013, MNRAS, 436,3341Zensus, J. A., Cohen, M. H., & Unwin, S. C. 1995, ApJ, 443, 35Zhao, W., Hong, X. Y., An, T., et al. 2011, A&A, 529, A113