FIRST J153350.8+272729: the radio afterglow of a decades-old tidal disruption event
Vikram Ravi, Hannah Dykaar, Jackson Codd, Ginevra Zaccagnini, Dillon Dong, Maria R. Drout, Bryan M. Gaensler, Gregg Hallinan, Casey Law
DDraft version February 12, 2021
Typeset using L A TEX twocolumn style in AASTeX62
FIRST J153350.8 + Vikram Ravi, Hannah Dykaar,
2, 3
Jackson Codd,
4, 5
Ginevra Zaccagnini, Dillon Dong, Maria R. Drout,
2, 6
B. M. Gaensler,
3, 2
Gregg Hallinan, and Casey Law Cahill Center for Astronomy and Astrophysics, California Institute of Technology, Pasadena, CA 91125, USA David A. Dunlap Department of Astronomy and Astrophysics, University of Toronto50 St. George Street, Toronto, Ontario, M5S 3H4 Canada Dunlap Institute for Astronomy and Astrophysics, University of Toronto, 50 St. George St., Toronto, ON M5S 3H4, Canada Cambridge Rindge and Latin School, Cambridge, MA, USA Department of Physics, Macalester College, Saint Paul, MN 55105, USA Observatories of the Carnegie Institution for Science, 813 Santa Barbara St., Pasadena, CA 91101, USA
ABSTRACTWe present the discovery of the fading radio transient FIRST J153350.8+272729. The source had amaximum observed 5-GHz radio luminosity of × erg s − in 1986, but by 2019 had faded by a factorof nearly 400. It is located 0.15 (cid:48)(cid:48) from the center of a galaxy (SDSS J153350.89 + γ -ray burst. This is only the second TDE candidate to be first discovered at radiowavelengths. Its luminosity fills a gap between the radio afterglows of sub-relativistic TDEs in the localuniverse, and relativistic TDEs at high redshifts. The unusual properties of FIRST J153350.8+272729(ongoing nuclear activity in the host galaxy, high radio luminosity) motivate more extensive TDEsearches in untargeted radio surveys. Keywords:
AGN host galaxies — black hole physics — radio transient sources — time domain astron-omy INTRODUCTIONThe ongoing Karl G. Jansky Very Large Array (VLA)Sky Survey (VLASS; Lacy et al. 2020) is the first ra-dio all-sky survey that is sensitive to several classes ofslowly evolving extragalactic radio transients, includingflares from active galactic nuclei (AGN), core-collapsesupernova afterglows, the orphan afterglows of off-axis γ -ray bursts, and tidal disruption events (TDEs) of starsby supermassive black holes (SMBHs). VLASS is aninterferometric survey in the 2–4 GHz band of the en-tire 33,885 deg of sky north of a declination of − ◦ ,with an angular resolution of 2.5 (cid:48)(cid:48) . When complete,the sky will be surveyed over three epochs spaced by 32months, to a continuum-image rms of µ Jy per beamper epoch. The exceptional survey grasp of VLASS pro-vides the opportunity to assemble samples of radio tran-sients without the need for external triggers, enablingradio-selected populations to be compared with thosefrom other wavelengths.Here we report on a remarkable radio transient discov-ered by jointly searching the first half of the first epochof VLASS (VLASS 1.1) and the VLA Faint Images of the Radio Sky at Twenty centimeters (FIRST) survey(Becker et al. 1995). The FIRST survey was conductedat a frequency of 1.4 GHz, and covered ∼ , deg of the northern sky mostly between 1994 and 1999with an angular resolution of 5 (cid:48)(cid:48) and a continuum-image rms of µ Jy per beam. We discovered asource (see §2), FIRST J153350.8 + + + µ Jy at 5 GHz.J1533 + A similar search yielded the discovery of the luminous ex-tragalactic radio transient FIRST J141918.9 + a r X i v : . [ a s t r o - ph . H E ] F e b (SDSS J153350.89 + + ∼ GHz is strong evidencefor its transient nature, and the preferred interpretationfor its origin is of a TDE (see §5).Although nearly a hundred TDE candidates are nowcataloged , only ∼ of TDEs exhibit radio emis-sion (Alexander et al. 2020; Anderson et al. 2019). Thethree most distant of the radio TDEs (Sw J1644 + +
05, and Sw J1112-82 at redshifts of 0.35, 1.18,and 0.89 respectively) were first discovered throughtransient γ -ray emission corresponding to the launch ofa nascent relativistic jet. The remaining radio TDEs arepowered by mildly relativistic outflows that drive shocksinto the circum-nuclear medium, and peak at radio lumi-nosities that are two to three orders of magnitude belowthe three relativistic radio TDEs. Only one TDE candi-date, CNSS J0019 +
00 (Anderson et al. 2019), has pre-viously been first discovered through its radio emission.In general, the radio emission generated by extragalacticexplosions (e.g., supernovae, γ -ray bursts, and TDEs)is enhanced in the presence of more energetic outflows,and denser circum-explosion material. TDEs discoveredthrough their radio emission, rather than through X-rayor optical emission associated with accreting material,may offer a novel selection of TDE phenomena and hostgalaxies.Throughout this work, we adopt the following cos-mological parameters: H = 67 . km s − Mpc − , Ω M =0 . , and Ω Λ = 0 . (Planck Collaboration et al.2016). DISCOVERY OF J1533+2727Two independent efforts discovered J1533 + , using either theAegean (Hancock et al. 2012, 2018) or PyBDSF (Mo-han & Rafferty 2015) source finding software. Detailsof how the source finding algorithms were applied willbe presented in future works that describe larger sam-ples of transients (e.g., Dong et al., in prep.). Oncethe VLASS 1.1 source catalogs were made, we per-formed cross-matches between unresolved sources inthe VLASS 1.1 and FIRST catalogs, with a specific fo-cus on finding FIRST sources not present in VLASS 1.1.In one of our efforts we only considered sources witha more than decrease in measured flux densitiesbetween FIRST and VLASS. In the other effort we only See http://tde.space. https://archive-new.nrao.edu/vlass/quicklook/ considered sources detected at > . mJy in FIRSTand undetected in VLASS (i.e., with 3 GHz flux den-sities < . mJy), which were additionally coincidentwith the nuclei of spectroscopically detected galaxies inSDSS DR14 (Abolfathi et al. 2018) at redshifts z < . .J1533+2727 was noteworthy as one of the brightestsources to pass all our selection thresholds. The posi-tion of this source in the FIRST catalog is (R.A. J2000,decl. J2000) = (15:33:50.884, + (cid:48)(cid:48) in each coordinate (Becker et al.1995). ARCHIVAL AND FOLLOW-UP RADIOOBSERVATIONSWe searched a selection of existing radio-survey cata-logs and data sets for detections of J1533+2727. Theresults are summarized in Table 1. J1533+2727 iscataloged in the FIRST survey with a flux density of . ± . mJy, on an observing epoch of 1995 Novem-ber 06. The source is also present in the NVSS catalogwith a flux density of . ± . mJy on 1995 April 16.Although the formal σ upper limit on the flux den-sity of J1533+2727 in the VLASS quick-look images is0.38 mJy (based on the per-pixel rms at the source lo-cation), we adopt an upper limit of 0.46 mJy to accountfor errors in the flux scale (Perley & Butler 2017). TheVLASS observation epoch was 2017 October 02. Wenext searched the VLA archive for data obtained at theposition of J1533+2727, and found that this source hadbeen observed (in a targeted observation, at the center ofthe primary beam) by the Cosmic Lens All-Sky Survey(CLASS; Myers et al. 2003) on 1998 May 22 at 8.46 GHz(VLA project AM0593), and by a wideband survey ofsources with rising spectra between 1.4 GHz and 4.8 GHzon 2001 September 14 (VLA project AG0617). We re-analyzed these data using standard tasks from the Com-mon Astronomy Software Applications (CASA, version5.1.1; McMullin et al. 2007), and finding no source atthe position of J1533+2727 derived the upper limits onits flux density reported in Table 1. As above, theseupper limits were derived using the per-pixel rms at thesource location.The selection of J1533+2727 as a CLASS source im-plied that J1533+2727 is present in the Green Bank300-foot telescope 6 cm survey catalog (GB6; Gregoryet al. 1996) with a flux density in excess of 30 mJy (My-ers et al. 2003) at 4.85 GHz. Indeed, the GB6 cataloglists a ± mJy source (GB6 J1533 + ± + ± https://archive.nrao.edu/archive/advquery.jsp Table 1.
Radio Observations of FIRST J153350.8 + Epoch Survey Frequency (GHz) Flux density (mJy)1968 Bologna 0.408 < < < ± ± < . ± . . ± . < . < < . < . < . < < < . . ± . . ± . . ± . Note —All upper limits are at the σ level. References: Bologna Sky Survey (Bologna;Colla et al. 1972), Texas Survey of Radio Sources at 365 MHz (Texas; Douglas et al.1996), Green Bank Northern Sky Survey (GBNSS; White & Becker 1992), Green Bank 6cm survey (GB6; Gregory et al. 1996), Gauribidanur Telescope (GEETEE; Dwarakanath& Udaya Shankar 1990), Faint Images of the Radio Sky at Twenty centimeters (FIRST;Becker et al. 1995), NRAO VLA Sky Survey (NVSS; Condon et al. 1998), Cosmic LensAll-Sky Survey (CLASS; Myers et al. 2003), VLA Low-frequency Sky Survey (VLSSr;Lane et al. 2014), TIFR GMRT Sky Survey Alternative Data Release (TGSS; Intemaet al. 2017), VLA Sky Survey (VLASS; Lacy et al. 2020). error is due to pointing errors of order 8 (cid:48)(cid:48) (Gregoryet al. 1996). Despite the 46.8 (cid:48)(cid:48) offset between the FIRSTand GB6 sources, the association is considered likelyas the next closest source in FIRST or VLASS to theGB6 position is offset by 379 (cid:48)(cid:48) , which is greater thanthe 3.5 arcmin full-width half-maximum of the GB6 sur-vey beam. The GB6 survey catalog is actually com-prised of observations obtained over two epochs, be-tween 1986 November 6 and December 13, and between1987 September 30 and December 1. Single epoch mapswere converted into source catalogs by Gregory et al.(2001) . The 1986 observations contain J1533+2727 The final CLASS sample was chosen by associating GB6sources with NVSS sources within a 70 (cid:48)(cid:48) separation cut, whichis explained by the much greater positional uncertainty of NVSSthan FIRST. Survey operations in late 1988 with the 300-foot telescopewere increasingly affected by pointing errors that rendered thedata unusable. Currently available at https://phas.ubc.ca/~gregory/RadioAstronomy.html. with a flux density of ± mJy, and the 1987 observa-tions showed a flux density of ± mJy.The remarkable fading of J1533+2727 over 31 yearsbetween 1986 and 2017 motivated follow-up VLA ob-servations (VLA program 19A-470). We obtained datain the B configuration (antenna separations between0.21 km and 11.1 km) on 2019 May 14 in the L (1–2 GHz) and C (4–8 GHz) bands, using standard contin-uum observing setups and CASA data-reduction proce-dures. The absolute flux-scale and bandpass calibra-tor was 3C286, and time-variable complex gain cali-bration was accomplished using J1513+2338. No self-calibration was conducted. We detected J1533+2727in both bands at a position consistent with the FIRSTposition within 0.2 (cid:48)(cid:48) . The measured flux densities were . ± . mJy at 1.52 GHz, . ± . mJy at 5 GHz,and . ± . mJy at 7 GHz. The measurementsare consistent with a single power law (flux density S ( ν ) ∝ ν α , where ν is the frequency) with spectral in-dex α = − . ± . . These measurements do notinclude a 3–5% uncertainty in the VLA flux-density scale(Perley & Butler 2017). HOST-GALAXY PROPERTIESThe centroid of the FIRST position of J1533+2727is located 0.150 (cid:48)(cid:48) from the optical center of the galaxySDSS J153350.89+272729. With a FIRST positional un-certainty of 0.400 (cid:48)(cid:48) , J1533+2727 is therefore consistentwith being coincident with the galaxy nucleus. Figure 1shows the SDSS DR16 (Ahumada et al. 2020) imageand spectrum of this galaxy, which we adopt as thehost of J1533+2727. The host galaxy lies at a red-shift of z = 0 . ± . (luminosity distance of . ± . Mpc), and the difference between the ra-dio position of J1533+2727 and the center of light ofthe host galaxy corresponds to a projected separationof just pc. An inspection of the SDSS optical spec-trum indicates weak Type II Seyfert activity, accordingto standard line-ratio diagnostics (Kewley et al. 2019),with log ([NII] / H α ) = − . and log ([OIII] / H β ) =0 . . Stellar population synthesis fits to the SDSS pho-tometry indicate a stellar mass of between . M (cid:12) and . M (cid:12) , and an ongoing star-formation rate of ∼ . M (cid:12) yr − (‘stellarMassFSPSGranEarlyDust’ table;Ahn et al. 2014). The quoted uncertainties are purelystatistical, and do not reflect the range of possible pa-rameterizations. The galaxy was classified morphologi-cally as a spiral by Kuminski & Shamir (2016).The absolute rest frame magnitudes of the bulge inthe g and r bands were calculated by Simard et al.(2011) using Galaxy IMage 2D (Simard 2010), as-suming extinction values obtained from the SDSSdatabase. We transformed these bulge absolute mag-nitudes, M g , bulge = − . and M r , bulge = − . tothe V band, M V , bulge = − . , according to formulasfrom Jester et al. (2005). We then applied the relationbetween SMBH mass and bulge luminosity from Mc-Connell & Ma (2013) to estimate the total black holemass to be log M BH /M (cid:12) = 7 . ± . . DISCUSSION5.1.
The nature of J1533+2727
We now interpret these observations in terms of threehypotheses for J1533+2727:• AGN variability.• An engine-driven transient associated with a su-pernova. • A jet or outflow powered by a TDE. We do not consider standard supernovae because the peakradio luminosity of J1533+2727 ( . × erg s − Hz − ) is a fac-tor of eight greater than even the most luminous radio supernova(PTF 11qcj; Palliyaguru et al. 2019). We first augment the observations presented abovewith archival ROSAT/PSPC pointings that includedJ1533+2727 in the field of view. We highlight two ob-servations in particular: a 415 s pointing on MJD 48102(1990 July 30; sequence id rs931238n00; around threeyears after the last Green Bank detection) obtainedas part of the ROSAT All-Sky Survey, and a 13932 spointing a year later on MJD 48449 (1991 July 12; se-quence id RP201103N00) obtained as part of a longexposure on α Cor Bor. We used the sosta tool inthe
XIMAGE package, and the exposure maps associatedwith the observations, to derive σ upper limits on thecount rates at the position of J1533+2727. These were0.125 cts s − and 0.0108 cts s − on the respective dates(in the standard PSPC 0.1–2.4 keV band). We then con-verted these to upper limits on the 2–10 keV luminosityassuming a photon index of 2, and a Galactic neutral-hydrogen column density of n H = 2 . × cm − de-rived from the HI4PI neutral hydrogen column map(HI4PI Collaboration et al. 2016). These upper limitswere L X < . × erg s − and L X < . × erg s − on the respective dates.These X-ray upper limits are low in comparison withthe radio luminosity of J1533+2727, if J1533+2727 rep-resents an active black hole. The black hole fundamen-tal plane relates the mass, 5 GHz radio luminosity, and2–10 keV X-ray luminosity of actively accreting objectsacross nine orders of magnitude in black hole mass, andhints at a fundamental link between accretion rate andjet power. We used the latest iteration of this relation(Gültekin et al. 2019) to derive predicted upper limitson the 5 GHz radio luminosity of J1533+2727 at theepochs of the X-ray observations. Given the derivedSMBH mass, by assuming Poisson statistics for the X-ray upper limits, and by using a Monte Carlo techniqueto account for the uncertainty and intrinsic scatter in theblack hole fundamental plane, we calculated 95% con-fidence upper limits on the expected radio luminosityof . × erg s − and . × erg s − correspond-ing to the two X-ray observations. These in turn implyradio flux-density upper limits of 3.7 mJy and 0.8 mJyrespectively, where we divided the luminosities by thefrequency to derive the spectral luminosities followingGültekin et al. (2019). If J1533+2727 was indeed thisfaint during the X-ray observations, a much more rapidevolution is implied between the Green Bank 4.85 GHzdetection and these epochs than at later times. Fur-thermore, unless the source re-brightened between theX-ray observations and the NVSS and FIRST detec-tions, an unrealistic spectral index steeper than − isimplied between 1.4 GHz and 5 GHz for the FIRST de-tection. We conclude that J1533+2727 was inconsistent F l u x ( − e r g s − c m − A n g s tr o m − ) H β [OIII][OII] OI [SII]H α [NII] Figure 1.
Left:
Three-color composite in the SDSS g , r and i bands of the host galaxy of J1533+2727, SDSS J153350.89+272729.The radio position of J1533+2727 is shown as a blue cross, and the spatial extent of the SDSS fiber input on the sky is shownas a red circle (Smee et al. 2013). Right:
SDSS spectrum (Ahumada et al. 2020) of SDSS J153350.89+272729 obtained on 2007March 21. Some relevant emission lines are labeled, which are indicative of weak Type II Seyfert activity. with the black hole fundamental plane when the ROSATX-ray observations were undertaken.This suggests that J1533+2727 was not actively ac-creting as an AGN at this time, which is in tension withthe hypothesis of ongoing AGN variability. This is how-ever not in tension with the TDE hypothesis, becausethe accretion could have ceased (e.g., Levan et al. 2016).When detected in the Green Bank survey, J1533+2727was also more luminous at a wavelength of 6 cm thanthe cores of any of the 52 nearby Seyfert galaxies ob-served by Ho & Ulvestad (2001), besides Perseus A(NGC 1275; Seyfert 1.5) and NGC 1167 (Seyfert II; be-yond the magnitude-completeness limit of the survey).Additionally, J1533+2727 is likely more variable thanany of the 12 Seyferts observed by Mundell et al. (2009),within whose sample the maximum variability in sevenyears at 8.4 GHz was a factor of three. We thereforeproceed to consider hypotheses (2) and (3) given above.Some insight can be gained by analyzing J1533+2727as a synchrotron transient, despite the lack of detailedspectral information. That J1533+2727 represents syn-chrotron emission is evident given the brightness tem-perature ( (cid:38) . × K) implied by the extreme vari-ability of the source. We can also derive rough con-straints on the source radius, R , the energy required topower the source, E , and the electron number density, n e , of the medium into which the source is expanding.The constraints are based on a fiducial mJy maximum A useful working definition of a variability timescale that canbe used to calculate a light-crossing time is a timescale over whichthe modulation index, defined as the variability range divided bythe mean source flux density, is greater than unity. Adoptinga timescale of six years (between the FIRST/NVSS observationsand L-band VLA observations in 2001), we find a brightness tem-perature in excess of . × K. flux density measured at . GHz, and the assumptionthat the optically thin spectral index observed in our2019 observations of the source of α = − . is repre-sentative of a non-evolving relativistic electron energydistribution N ( E ) ∝ E − p with p = − α + 1 = 2 . . Inthe following, we assume ( a ) equipartition between theenergy in relativistic electrons and in magnetic fieldswithin the source, ( b ) that the source is expanding sub-relativistically, ( c ) that the source is spherically sym-metric with a filling factor of unity, and ( d ) that therelativistic electrons are accelerated in a strong (for-ward) shock that deposits 10% of its energy in the elec-trons, and 10% of its energy in magnetic fields (i.e., (cid:15) e = (cid:15) B = 0 . in usual terms). The non-relativisticassumption further implies that the spectral peak wasassociated with synchrotron self-absorption rather thanthe minimum relativistic-electron energy (e.g., Chevalier1998).In this scenario, following Chevalier (1998), R ∝ S ( p +6) / (2 p +13) p ν − p , and E ∝ S (3 p +14) / (2 p +13) p ν − p , where ν p is the peak frequency and S p is the peak flux den-sity. If we are simply constraining the values of R and E when ν p = 4 . GHz, we are setting lower limits onboth quantities. This is also essentially the case if weare constraining the values of R and E during the 1986Green Bank observations. The electron number den-sity is derived by applying the Rankine-Hugoniot jumpconditions in the strong shock limit (Equation (15) of Ho In our calculation, we adopt c = 9 . × − and c =8 . × − from Pacholczyk (1970). If the spectrum was optically thick and ν p > ν = 4 . GHz, S p ∝ ( ν p /ν ) implies a nearly fixed estimate of R regardless ofthe true value of ν p , and a larger value of E . If the spectrum wasoptically thin, R and E would clearly both be larger. et al. 2019), with the further assumption of the shockvelocity being given by the source size divided by itslifetime, T . In this case, n e ∝ S − (2 p +16) / (2 p +13) p ν p . Insummary, using the 1986 measurement, we find R = 2 . × (cid:18) S p
65 mJy (cid:19) . (cid:16) ν p .
85 GHz (cid:17) − cm (1) E = 7 . × (cid:18) S p
65 mJy (cid:19) . (cid:16) ν p .
85 GHz (cid:17) − erg (2) n e = 1 . × − (cid:18) S p
65 mJy (cid:19) − . (cid:18) T (cid:19) (cid:16) ν p .
85 GHz (cid:17) cm − . (3)These equations, including the constants of proportion-ality, directly reproduce Equations (8), (12), and (16)of Ho et al. (2019), for our value of p . Although theradius estimate is only mildly sensitive to the assump-tions above, a departure from equipartition like that ob-served for Sw J1644 +
57 (Eftekhari et al. 2018), where (cid:15) B = 0 . was inferred assuming (cid:15) e = 0 . , would in-crease the energy estimate by two orders of magnitude.These results provide evidence that J1533+2727 rep-resents a relativistic jet/outflow from a TDE. First,the lower limit on the energy in the outflow is greaterthan that of any known stellar cataclysm (e.g., Mat-tila et al. 2018) besides classical, on-axis long gamma-ray bursts (LGRBs). Additionally, most LGRBs haveapparent expansion velocities of Γ β (cid:38) (here, Γ isthe bulk Lorentz factor of the emitting material, and β = v/c is the normalized expansion velocity). Howeverfor J1533+2727, following Sari et al. (1998), the char-acteristic frequency corresponding to radiation from thelowest-energy relativistic electrons, ν m , was likely lowerthan 4.85 GHz in 1986, because the source declines be-tween 1986 and 1987. This frequency is related to E and T in the case of adiabatic evolution, which impliesa lifetime: T (cid:38) (cid:18) E . × erg (cid:19) / days . (4)This in turn implies Γ β (cid:46) , and n e (cid:46) . − . Al-though this makes the LGRB scenario somewhat fine-tuned, the derived parameters are consistent with some Radio calorimetry of the ejecta of cosmic explosions tracesthe fastest ejecta, and therefore cannot be directly related to thetotal energies in outflows with a range of velocities, like supernovae(e.g., Berger et al. 2003).
GRBs (e.g., GRB 980703; Perley et al. 2017). Indeed,GRB 980703 exhibits late-time emission 16 years post-burst that is similar to J1533+2727 (Perley et al. 2017).However, the projected offset between J1533+2727 andthe center of light of its host galaxy of ≈ pc is in-consistent with more than 95% of the LGRB popula-tion (Lyman et al. 2017). Additionally, the high stellarmass and lack of evidence for an ongoing starburst in thehost galaxy of J1533+2727 are inconsistent with typicalLGRB hosts (Taggart & Perley 2019). We therefore fa-vor the TDE scenario.A comparison between the radio lightcurve of J1533+2727and the remainder of the TDE population is shown inFigure 2. The post-explosion time of the 1986 epoch (88days) was derived assuming a nominal expansion veloc-ity of c , which would imply n e ∼ cm − . Please notehowever that sub-relativistic expansion was assumed toderive the constraints above, and this is therefore forillustrative purposes only. The high radio luminosityand outflow energy relative to several radio-detectedTDEs is suggestive of a relativistic jet, rather than awide-angle outflow (Alexander et al. 2020). We notethat if the emission region were significantly aspherical,with a non-unity filling factor, some of the above con-clusions would be altered by factors of a few (BarniolDuran et al. 2013).5.2. Implications for the TDE population
We have established that J1533+2727 is a remarkableradio transient and a likely TDE afterglow. Using theFIRST survey, and assuming the detection of just onesuch source, we can calculate a lower limit on the occur-rence of sources like J1533+2727. The 1 mJy minimumflux density of the FIRST catalog and the peak lumi-nosity of J1533+2727 in FIRST implies a detectable dis-tance of 452 Mpc. The FIRST and VLASS 1.1 sky sur-veys have an overlapping sky coverage of ∼ deg ,and thus the detectable distance corresponds to an ob-servable volume of 0.056 Gpc . J1533+2727 emittedabove its detected FIRST luminosity for at least 8 yearsbetween the Green Bank and FIRST detections. We cantherefore infer a lower limit on the volumetric rate ofapproximately 2.2 Gpc − yr − , or ∼ % of the observedTDE rate (van Velzen 2018). We do not take this anal-ysis further because the search for TDEs detected inFIRST is ongoing.Our results add to the emerging picture of the di-versity of TDE-driven jets/outflows from supermassiveblack holes. Although it has long been known that ∼ of the mass of a disrupted star is likely to beunbound (e.g., Rees 1988), the geometry and kinemat-ics of such outflows are poorly constrained, as are any − − Approximate time since event (yr)10 R a d i o l u m i n o s i t y ( e r g s − ) J1533+2727 5 GHz J1533+2727 1.4 GHz
Sw J1644+57Sw J20508+05Sw 1112-82AT 2019dsgIGR J12580+0134CNSS J0019+00ASASSN-14liArp 299XMMSL1 J0740-85
Figure 2.
Lightcurves of all TDEs detected at radio wavelengths. All TDE data are at 5 GHz, except for the 1.4 GHz dataon J1533 + + jets/outflows powered by the accretion of the remain-ing 50% of the mass. The radio luminosity and derivedoutflow energy of J1533+2727 fills the gap between thethree relativistic TDEs identified through their prompthigh-energy emission, and the remaining TDE sample(Figure 2). The host galaxy of J1533+2727 and itscentral supermassive black hole appear otherwise unre-markable with respect to the TDE population (Frenchet al. 2020; Wevers et al. 2017). Although the blackhole mass is somewhat high relative to optically se-lected TDEs (Wevers et al. 2017), stars with a widerange of masses ( (cid:38) . M (cid:12) ) are expected to be dis-rupted by such black holes (Kochanek 2016). The opti-cal spectrum of the host of J1533+2727 shows emissionlines characteristic of the narrow-line regions of Type IISeyferts; this nuclear activity must have been ongoingprior to the transient event, given the large expectedsizes of narrow-line regions (e.g. Bennert et al. 2006).This is similar to the host of the radio-discovered TDECNSS J0019 +
00 (Anderson et al. 2019). The presenceof nuclear activity in the spectrum of the J1533+2727host makes it difficult to determine whether or not it is a post-starburst galaxy, although we note that TDEsare found to be over-represented in galaxies that areevidently post-starburst from their optical spectra (Fig-ure 3; French et al. 2016, 2020). We speculate that TDEsdiscovered in radio transient surveys will have substan-tially different selection effects, especially with regardsto AGN activity and extinction, than the selection ef-fects present in optical and soft X-ray surveys that dom-inate TDE discoveries today. CONCLUSIONSWe present the discovery of the candidate TDEFIRST J153350.8 + × erg s − ), and the implied energy in the out-flow generated by the TDE ( (cid:38) × erg), fill a gapbetween most radio-detected TDEs and the three high-redshift events that were first discovered through theirprompt γ -ray emission. Little more can be said about Figure 3.
Plot adapted from French et al. (2020) (their Fig-ure 5 – see French et al. (2020) and French et al. (2016) for de-tails) showing spectral indices of a sample of SDSS galaxies,and selected TDE hosts including J1533+2727. H α emissiontraces current star formation while H δ absorption traces star-formation activity in the past ∼ Gyr. Post-starburst / quies-cent Balmer-strong galaxies comprising 0.2% (solid box) and2% (dashed box) of the parent SDSS sample are at the lowerright of the plot. The hosts of optically and X-ray selectedTDEs are over-represented among post-starburst galaxies.Although only two radio-selected TDE candidates have beenidentified so far (J1533+2727 and CNSS J0019+00), neitheris hosted by a post-starburst galaxy. the nature of the outflow and the medium into which itpropagates, because we have only observed the opticallythin component of the radio spectral energy distribution.The host galaxy, at a distance of 147 Mpc, is largely un-remarkable (inferred supermassive black hole mass of × M (cid:12) Abolfathi, B., Aguado, D. S., Aguilar, G., et al. 2018,ApJS, 235, 42 Ahn, C. P., Alexandroff, R., Allende Prieto, C., et al. 2014,ApJS, 211, 17
Ahumada, R., Allende Prieto, C., Almeida, A., et al. 2020,ApJS, 249, 3Alexander, K. D., van Velzen, S., Horesh, A., & Zauderer,B. A. 2020, SSRv, 216, 81Anderson, M. M., Mooley, K. P., Hallinan, G., et al. 2019,arXiv e-prints, arXiv:1910.11912Barniol Duran, R., Nakar, E., & Piran, T. 2013, ApJ, 772,78Becker, R. H., White, R. L., & Helfand, D. J. 1995, ApJ,450, 559Bennert, N., Jungwiert, B., Komossa, S., Haas, M., &Chini, R. 2006, A&A, 456, 953Berger, E., Kulkarni, S. R., Frail, D. A., & Soderberg,A. M. 2003, ApJ, 599, 408Chevalier, R. A. 1998, ApJ, 499, 810Colla, G., Fanti, C., Fanti, R., et al. 1972, A&AS, 7, 1Condon, J. J., Broderick, J. J., Seielstad, G. A., Douglas,K., & Gregory, P. C. 1994, AJ, 107, 1829Condon, J. J., Cotton, W. D., Greisen, E. W., et al. 1998,AJ, 115, 1693Douglas, J. N., Bash, F. N., Bozyan, F. A., Torrence,G. W., & Wolfe, C. 1996, AJ, 111, 1945Dwarakanath, K. S., & Udaya Shankar, N. 1990, JAA, 11,323Eftekhari, T., Berger, E., Zauderer, B. A., Margutti, R., &Alexander, K. D. 2018, ApJ, 854, 86French, K. D., Arcavi, I., & Zabludoff, A. 2016, ApJL, 818,L21French, K. D., Wevers, T., Law-Smith, J., Graur, O., &Zabludoff, A. I. 2020, SSRv, 216, 32Graur, O. 2018, arXiv e-prints, arXiv:1809.08078Gregory, P. C., Capak, P., Gasson, D., & Scott, W. K.2001, in IAU Symposium, Vol. 205, Galaxies and theirConstituents at the Highest Angular Resolutions, ed.R. T. Schilizzi, 98Gregory, P. C., Scott, W. K., Douglas, K., & Condon, J. J.1996, ApJS, 103, 427Gültekin, K., King, A. L., Cackett, E. M., et al. 2019, ApJ,871, 80Hancock, P. J., Murphy, T., Gaensler, B. M., Hopkins, A.,& Curran, J. R. 2012, MNRAS, 422, 1812Hancock, P. J., Trott, C. M., & Hurley-Walker, N. 2018,PASA, 35, e011HI4PI Collaboration, Ben Bekhti, N., Flöer, L., et al. 2016,A&A, 594, A116Ho, A. Y. Q., Phinney, E. S., Ravi, V., et al. 2019, ApJ,871, 73Ho, L. C., & Ulvestad, J. S. 2001, ApJS, 133, 77Intema, H. T., Jagannathan, P., Mooley, K. P., & Frail,D. A. 2017, A&A, 598, A78 Jester, S., Schneider, D. P., Richards, G. T., et al. 2005,AJ, 130, 873Kewley, L. J., Nicholls, D. C., & Sutherland, R. S. 2019,ARA&A, 57, 511Kochanek, C. S. 2016, MNRAS, 461, 371Kulkarni, S. R., Frail, D. A., Wieringa, M. H., et al. 1998,Nature, 395, 663Kuminski, E., & Shamir, L. 2016, VizieR Online DataCatalog, 222Lacy, M., Baum, S. A., Chandler, C. J., et al. 2020, PASP,132, 035001Lane, W. M., Cotton, W. D., van Velzen, S., et al. 2014,MNRAS, 440, 327Law, C. J., Gaensler, B. M., Metzger, B. D., Ofek, E. O., &Sironi, L. 2018, ApJL, 866, L22Levan, A. J., Tanvir, N. R., Brown, G. C., et al. 2016, ApJ,819, 51Lyman, J. D., Levan, A. J., Tanvir, N. R., et al. 2017,MNRAS, 467, 1795Mattila, S., Pérez-Torres, M., Efstathiou, A., et al. 2018,Science, 361, 482McConnell, N. J., & Ma, C.-P. 2013, ApJ, 764, 184McMullin, J. P., Waters, B., Schiebel, D., Young, W., &Golap, K. 2007, in Astronomical Society of the PacificConference Series, Vol. 376, Astronomical Data AnalysisSoftware and Systems XVI, ed. R. A. Shaw, F. Hill, &D. J. Bell, 127Mohan, N., & Rafferty, D. 2015, PyBDSF: Python BlobDetection and Source Finder, , , ascl:1502.007Mundell, C. G., Ferruit, P., Nagar, N., & Wilson, A. S.2009, ApJ, 703, 802Murphy, T., Chatterjee, S., Kaplan, D. L., et al. 2013,PASA, 30, e006Myers, S. T., Jackson, N. J., Browne, I. W. A., et al. 2003,MNRAS, 341, 1Oosterloo, T., Verheijen, M. A. W., van Cappellen, W.,et al. 2009, in Wide Field Astronomy & Technology forthe Square Kilometre Array, 70Pacholczyk, A. G. 1970, Radio astrophysics. Nonthermalprocesses in galactic and extragalactic sourcesPalliyaguru, N. T., Corsi, A., Frail, D. A., et al. 2019, ApJ,872, 201Perley, D. A., Hjorth, J., Tanvir, N. R., & Perley, R. A.2017, MNRAS, 465, 970Perley, R. A., & Butler, B. J. 2017, ApJS, 230, 7Planck Collaboration, Ade, P. A. R., Aghanim, N., et al.2016, A&A, 594, A13Rees, M. J. 1988, Nature, 333, 523Salas, P., Bauer, F. E., Stockdale, C., & Prieto, J. L. 2013,MNRAS, 428, 12070