Constraints on off-axis jets from stellar tidal disruption flares
Sjoert van Velzen, Dale A. Frail, Elmar Koerding, Heino Falcke
AAstronomy & Astrophysics manuscript no. ruplim c (cid:13)
ESO 2018October 31, 2018
Constraints on off-axis jets from stellar tidal disruption flares
Sjoert van Velzen , Dale A. Frail , Elmar K¨ording , and Heino Falcke , , IMAPP, Radboud University, P.O. Box 9010, 6500 GL Nijmegen, The Netherlandse-mail: [email protected] National Radio Astronomy Observatory, Socorro, NM, USA ASTRON, Dwingeloo, The Netherlands Max-Planck-Institut f¨ur Radioastronomie Bonn, GermanyReceived 21 September 2012; Accepted 13 December 2012
Abstract
Context.
Many decades of observations of active galactic nuclei (AGN) and X-ray binaries have shown that relativisticjets are ubiquitous when compact objects accrete. One could therefore anticipate the launch of a jet after a star isdisrupted and accreted by a massive black hole. This birth of a relativistic jet may have been observed recently in twostellar tidal disruption flares (TDFs), which were discovered in gamma-rays by Swift. Yet no transient radio emissionhas been detected from the tens of TDF candidates that were discovered at optical to soft X-ray frequencies. Becausethe sample that was followed-up at radio frequencies is small, the non-detections can be explained by Doppler boosting,which reduces the jet flux for off-axis observers. And since the existing follow-up observation are mostly within ∼ Aims.
We wish to test the conjecture that all TDFs launch jets.
Methods.
We present 5 GHz follow-up observations with the Jansky VLA of seven known TDFs, a significant increaseof the number of radio observations of these events. To avoid missing delayed jet emission, our observations probe 1–8years since the estimated time of disruption.
Results.
None of the sources are detected, with very deep upper limits at the 10 micro Jansky level. These observationsrule out the hypothesis that these TDFs launched jets similar to radio-loud quasars. We also constrain the possibilitythat the flares hosted a jet identical to Sw 1644+57, the first and best-sampled relativistic TDF.
Conclusions.
We thus obtain evidence for a dichotomy in the stellar tidal disruption population, implying that the jetlaunching mechanism is sensitive to the parameters of the disruption.
1. Introduction
The disruption of a star by a massive black hole leads toarguably the most spectacular form of accretion onto thesecompact objects. The stellar debris that remains bound af-ter the disruption returns to the black hole at a rate thatinitially can exceed the Eddington limit ( ˙ M Edd ) by manyorders of magnitude. This fallback rate declines with apower law index of − / M Edd within a few to ten years. A tidal disruptionflare (TDF) may thus be used to sample different modes ofaccretion (e.g., Abramowicz & Fragile 2013) for a singlesupermassive black hole. Considerable effort is needed tosimulate the dynamics of the disruption (e.g., Nolthenius &Katz 1982; Evans & Kochanek 1989; Rosswog et al. 2009;Guillochon & Ramirez-Ruiz 2012) and to estimate the re-sulting optical to X-ray light curve of the flare (e.g., Loeb& Ulmer 1997; Bogdanovi´c et al. 2004; Strubbe & Quataert2009; Lodato & Rossi 2011). Efficient detection to obtain alarge sample of TDFs is much anticipated, as this will al-low, for example, a study of the demographics of dormantblack holes beyond the local universe (Frank & Rees 1976;Lidskii & Ozernoi 1979).Tens of (candidate) stellar tidal disruption events havebeen found by searching for flares in soft X-ray (Komossa& Bade 1999; Grupe et al. 1999; Komossa & Greiner 1999; Greiner et al. 2000; Esquej et al. 2008; Maksym et al. 2010;Lin et al. 2011; Saxton et al. 2012), UV (Gezari et al. 2006,2008, 2009, 2012), or optical surveys (van Velzen et al. 2011;Drake et al. 2011; Cenko et al. 2012a), or based on spec-tra with extreme coronal lines (Komossa et al. 2008; Wanget al. 2012). None of these thermal flares are associated witha radio transient, but only a handful have been followed-up at this frequency. The only tidal disruption candidateswith a detected transient radio counterpart are those dis-covered in γ -rays by Swift : Sw 1644+57 (Bloom et al. 2011;Burrows et al. 2011; Levan et al. 2011; Zauderer et al. 2011)and Sw 2058+05 (Cenko et al. 2012b). Since the radio andX-ray emission of these two events most likely originatesfrom a relativistic jetted outflow, they are often referred toas relativistic TDFs. In this paper we shall refer to the otherclass of TDFs as ‘thermal’, since they are all discovered atoptical to soft X-ray frequencies.One is left to wonder why the two TDFs discoveredwith
Swift are the only events with evidence for a newly-born jet. Interpreting this as a radio-loud/radio-quiet di-chotomy, similar to the devision of radio-loudness in quasars(Kellermann et al. 1989; Falcke et al. 1996; Sikora et al.2007), would require that the tidal disruption jet launchingmechanism is sensitive to the properties of the disruption(e.g., mass ratio, impact parameter, orientation of the orbitof the star with respect to the black hole spin, or circum- a r X i v : . [ a s t r o - ph . H E ] M a r joert van Velzen et al.: Off-axis tidal disruption jets nuclear environment). The explanation that quasars spendonly a fraction of their time as radio-loud objects, similarto the jets in the ‘hard intermediate state’ of X-ray binaries(e.g., K¨ording et al. 2006), does not apply to tidal disrup-tions because their accretion rate is not constant. On theother hand, based on the observed fundamental plane ofblack hole accretion (Merloni, Heinz, & di Matteo 2003;Falcke, K¨ording, & Markoff 2004), and the abundance ofjets in low luminosity AGN (Nagar et al. 2000) and X-ray binaries or microquasars (Mirabel & Rodr´ıguez 1999;Fender 2001), one may postulate that all stellar tidal dis-ruptions launch jets. Likewise, M¨ıller & G¨ultekin (2011)argue that the fundamental plane can be used to estimatethe black hole mass of a TDF (if X-ray and radio observa-tions of the flare are available).If all stellar disruptions are indeed accompanied by arelativistic outflow, the current upper limits on the radioflux of the thermal TDFs can be explained by the orienta-tion of this jet which can dramatically reduce the flux dueto relativistic Doppler boosting. The current non-detectionsmay also be explained by a delay of the radio emission ofthe jet with respect to the time of disruption. However, thenumber of TDFs that have been followed-up at radio fre-quencies is currently not sufficient to test this unificationbased on viewing angle.Recent advances in the hardware of the Very LargeArray (VLA) have made it possible to obtain very deepradio observations of stellar tidal disruptions in a relativelyshort amount of time. To use this opportunity, we selectedall thermal stellar tidal disruptions that occurred after 2004for follow-up observations. These observations significantlyincrease the number of TDFs with deep radio observations.And because our radio observations span a wide range oftimes since the disruption, we can, for the first time, testthe hypothesis that all stellar tidal disruptions launch jets.The remainder of paper is organized as follows. In sec.2 we present two different tidal disruption (TD) jet modelsand compute off-axis light curves. In sec. 3 we discuss theradio observations and sample selection. We use these ob-servations to constrain the jet models in sec. 4 and we closewith a discussion in sec. 5.
2. Tidal disruption jet models
To be able to interpret our radio observations, we need amodel that describes the radio emission of jets in accretingobjects. In this section we therefore review two models oftidal disruption jets and we present off-axis light curves forthese models. We have divided the models into two classes based on the origin of the emitting particles: external orinternal. In both models, some fraction of the accretionpower ends up in the jet and the emission mechanism issynchrotron radiation. The external model of radio emission from TD jets wasfirst presented by Giannios & Metzger (2011) and furtherdeveloped by Metzger, Giannios, & Mimica (2012). Shock Other models of TD jets (Lei & Zhang 2011; Krolik & Piran2012; De Colle et al. 2012), are not discussed here since thesemake no predictions for off-axis light curves at a given observerfrequency.
Time since Swift trigger (yr) − − − F l uxd e n s it y ( m J y ) ◦ ◦ ◦ ◦
22 GHz6 GHz
Figure 1.
The observed light curve of Sw 1644+57 (open sym-bols), with the predicted late-time light curve (dashed lines) fora total jet energy of E j = 10 erg (Berger et al. 2012). We showthe estimated 5 GHz light curve of different off-axis observers,Eq. 2, assuming that the Lorentz factor of the jet decreases withΓ j ∝ t − . , as inferred by Berger et al. (2012). We modifiedthe extrapolated light curve to match a Sedov-Taylor solution, L j ∝ t − . , when Γ j <
2. The 2- σ upper limits on the radio fluxof seven other TDF candidates (Table 2) are shown with blacktriangles (we scaled these limits to the redshift of Sw 1644+57,see sec. 4.1). interaction between the jet and the gaseous circumnuclearmedium powers the emission, similar to afterglow modelsof gamma-ray bursts (e.g., Sari & Piran 1995).The external model has been applied to the radio lightcurve of the relativistic TDF Sw 1644+57 (Metzger et al.2012; Berger et al. 2012). We show the fit and predictedlate-time light curve in Fig. 1. We note that this fit requiresa continuous increase of the isotropic jet power during thefirst year of observations.The scaling of the synchrotron peak and self absorptionfrequency in the Metzger et al. (2012) model of Sw 1644+57are based on spherical expansion of an ultra-relativisticshell and thus require θ j Γ j < θ j , Γ j the jet openingangle and Lorentz factor, respectively), plus an on-axis ob-server i < / Γ j ( t = 0); both requirements are supportedby the observed radio light curve (Metzger et al. 2012).To compute the light curve for an off-axis observer, wefirst boost the observed on-axis flux F ( ν ) into the jet rest-frame L j ( ν ) = d L δ − α F ( ν ) (1)(e.g., Lind & Blandford 1985; Jester 2008). Here we in-troduced the Doppler factor for the on-axis observer δ =[Γ j (1 − β j cos i )] − with β j = v j /c , α is the spectral indexdefined as F ( ν ) ∝ ν α , and d L is the luminosity distance.Next, we transform the jet luminosity to the off-axis ob-server using a different Doppler factor, δ . If the size of theemitting region is small compared to the distance to theblack hole, the time delay due to the geometrical separa-tion of the synchrotron peak with frequency can be ignored,and we can estimate the flux for an observer sitting at i : F ( t, ν ) = (cid:18) δ δ (cid:19) − α ( t ) F ( t, ν ) ≈ (cid:18) − β j ( t )1 − β j ( t ) cos i (cid:19) − α ( t ) F ( t, ν ) . (2) Time since disruption (yr) − − F l uxd e n s it y ( m J y ) ◦ ◦ ◦ ◦ Figure 2.
Predicted 5 GHz curves for the tidal disruption can-didate D3-13 (Gezari et al. 2009) for the internal jet model inthe optimistic scenario (Eq. 3 a ). The existing upper limits onthe 1.4 GHz flux density (Bower 2011) are shown with bluetriangles, pointing left and right to indicate the uncertainty onthe time of disruption. The downward pointing triangles labeled“2012” show our upper limits to the 5 GHz flux. For cos i < β j ,the peak of the light curve decreases in time and magnitude as i increases, because a fixed observed frequency corresponds toa smaller radius where the jet becomes optically thin to syn-chrotron self-absorption (i.e., z ssa ∝ δ/ν ). For cos i > β j thelight curve is compressed due to time retardation. Here t is measured in the observer-frame, α ( t ) is obtainedfrom the light curve, and we used cos i ≈ − Γ − j / ∼ j ( t ) and i . The latteris a free-parameter, which we shall constrain by our follow-up observations in sec. 4.To obtain Γ j ( t ) and the light curve beyond the last pub-lished radio observation of Sw 1644+57 (about one year af-ter the Swift trigger), we used the external model of TD jets(Metzger et al. 2012) applied to the Sw 1644+57 radio databy Berger et al. (2012). The off-axis light curve that is de-rived here may thus be viewed as a test for this model. Weconsider two scenarios. First, we set Γ( t > j (cid:46) j ∝ t − . (as inferred for Sw 1644+57), but modify the ex-trapolated light curve to match the non-relativistic Sedov-Taylor evolution when Γ j <
2. To estimate the light curvedecay in the Sedov-Taylor phase, we assume an electronpower-law index p = 2 . k = 3 / L j ∝ t − . (Granot et al. 1999; Leventiset al. 2012). We show the light curves in Fig. 1. The internal model of radio emission from TD jets wasfirst presented in van Velzen, Falcke, & Farrar (2010) andfurther developed by van Velzen, K¨ording, & Falcke (2011).The model is based on the simple idea of jet-disk coupling (Rawlings & Saunders 1991; Falcke & Biermann 1995): aconstant fraction of the accretion luminosity ( L d ) is fed intothe jet, Q j = q j L d . The conversion from jet power ( Q j )to radio luminosity ( L j ) follows by assuming equipartitionbetween the energy in relativistic particles and magneticfields, and has been calibrated using observations of AGN(Falcke et al. 1995; Willott et al. 1999; K¨ording et al. 2008).Stellar mass black holes show rapid switches from radio-loud ( q j = 0 .
2) to radio-quiet ( q j < . q j = . a )0 .
002 ˙ M ( t ) >
2% ˙ M Edd ( b )0 . t < t fallback ( c ) (3)where each scenario reverts to the preceding one if thecondition on t or ˙ M is not true (e.g., q j = 0 . M <
2% ˙ M Edd in all three scenarios). In the optimisticscenario ( a ), the TD jet behaves like a radio-loud quasarat all times. In the most conservative scenario ( b ), the jetbecomes radio-loud only when the accretion drops below <
2% ˙ M Edd (Maccarone 2003). In scenario c , the systemstarts with a radio-loud burst during the onset of the ac-cretion. We consider c the most realistic scenario, since itmost closely resembles the observed behavior of X-ray bi-naries.Besides q j (Eq. 3) the internal model requires a jetLorentz factor and the disk luminosity as a function of time.The latter is obtained from the fallback rate of the stellardebris for a pericenter passage at the disruption radius,capped at the Eddington limit. We vary Γ j between 5, thedefault value in van Velzen et al. (2011), and Γ j = 2. InFig. 2 we show an example light curve of the TDF candi-date D3-13 (Gezari et al. 2009) for Γ j = 5, using the blackhole mass as estimated from the luminosity of the host.In the internal TD jet model, the typical time scale ofthe light curve is set by the radius where the jet becomes op-tically thin to synchrotron self-absorption, z ssa ∝ δν − L / d .This is different from our off-axis version of the externalmodel, where we assumed that the emission is dominatedby the head of the jet, which explains the dissimilar scalingof the peak of the light curve with the Doppler factor forthe internal and external model.
3. Observations
In Table 1 we summarize the published radio follow-up ob-servations of TDF candidates that were discovered at op-tical to soft X-ray wavelengths. To increase this sample,we selected all TDF candidates with an estimated time ofdisruption after 2004 for follow-up observations. This limitis used since the internal model of TD jets is no longervalid when the jet slows down significantly. Our samplealso includes TDFs with existing radio upper limits, sincethe radio emission can be delayed with respect to the timeof disruption. We removed CSS100217 from the sample be-cause it is detected at 1 GHz before the time of disruption
Table 1.
Existing radio follow-up observations of TDF candi-dates that were discovered at optical to soft X-ray frequencies(our new observations are shown in Table 2).. name t D F ν ν ∆ t (yr) (mJy) (GHz) (yr)NGC 5905 < .
15 8.5 6.0D3 − < .
15 1.4 1.8TDE2 < .
10 8.4 1.1CSS100217 . ± .
03 7.9 0.3SDSS J1201+30 < .
22 4.8 1.4RX J1624+7554 < .
085 3.0 21.8IC 3599 . ± .
03 3.0 21.6RX J1420+5334 . ± .
02 3.0 21.6RX J1242 − < .
090 3.0 20.0SDSS J1323+48 < .
170 3.0 8.6SDSS J1311 − < .
095 3.0 8.4
Notes.
In the third column we show 5- σ upper limits on the ra-dio flux or the detected flux and 1- σ uncertainty. ∆ t denotes thetime of the radio observation with respect to the estimated timeof disruption ( t D ). The radio observation of the first five candi-dates were published before mid-2012, the last six are taken fromBower et al. (2013). We note that most of the TDF candidatesthat were discovered in the nineties also have post disruption ra-dio upper limits from large radio surveys (e.g., NVSS, Condonet al. 1998), these are listed in Komossa (2002). References. (1) Bade et al. (1996); Komossa (2002), (2) Gezariet al. (2008); Bower (2011), (3) van Velzen et al. (2011),(4) Drake et al. (2011), (5) Saxton et al. (2012), (6) Grupe et al.(1999), (7) Komossa & Bade (1999), (8) Komossa & Greiner(1999), (9) Komossa & Greiner (1999), (10) Esquej et al. (2007),(11) Maksym et al. (2010). with a flat spectral index, indicating an AGN origin for theradio emission (Drake et al. 2011). SDSS J1201+30 was notselected for follow-up observations because this TDF waspublished after our observations were scheduled. Details ofthe data reduction of the remaining six candidates are sum-marized below.The radio observations were carried out on the Karl G.Jansky Very Large Array on 29 January 2012 under pro-gram 12A-005. We observed at a central frequency of 5.0GHz with 16 subbands each with 64 2 MHz channels, span-ning 2 GHz of total bandwidth. The VLA was in the Cconfiguration yielding typical angular resolution of 4 arc-sec. The total observing time was 2.5 hrs, with integrationtimes for individual TDF sources varied from 18-30 min.Phase calibration was carried out by making short observa-tions of nearby point source calibrators every 10 minutes,while amplitude and bandpass calibration was achieved us-ing an observation of 3C 286 or 3C 48 at the beginning orend of each observing run. The data were reduced follow-ing standard practice in the Astronomical Image ProcessingSystem (AIPS) software package.In addition to these data, we identified one public dataset from the VLA archive (project AS 1020) for the TDFcandidate PS1-10jh (Gezari et al. 2012). These observationwere made with the VLA on 29 March 2011 in the B config-uration with two subbands (each with 64 2 MHz channels)centered at 4.83 and 4.96 GHz, for a total bandwidth of256 MHz. The calibration and imaging of these data wassimilar to the method described above.
Table 2.
Jansky VLA observations at 5 GHz of TDF candi-dates discovered at UV or optical frequencies.name redshift M BH t int σ ( F ν ) ∆ tM (cid:12) × (min) ( µ Jy) (yr)D1-9 Notes.
We list the redshift and estimated black hole mass ofthese candidates in the second and third column, respectively.No significant emission was detected at the phase center of theimages. We list integration time after removal of interference,the rms of the images, and the time of the observations withrespect to the estimated time of disruption.
References. (1) Gezari et al. (2008), (2) van Velzen et al.(2011), (3) Gezari et al. (2009), (4) Cenko et al. (2012a),(5) Gezari et al. (2012).
Our final sample that we shall use to constrain TD jetmodels thus consists of seven TDF candidates that wereobserved with the Jansky VLA. We summarize the resultsof these observations in Table 2.
4. Analysis
In this section we first compute the constraints that can beplaced on TD jet models using our Jansky VLA follow-upobservation and then consider the potential of radio tran-sient surveys.
If we assume that the angle between the observer and thejet is drawn from a uniform distribution (on a sphere), wecan calculate the probability of non-detections for a givenflux density limit. One simply has to find the largest anglefor which the predicted flux is above the flux limit and thencalculate the probability to observe a jet within this angle.The flux limit is set at twice the rms of the radio imageof each TDF. (This is lower than the limit for a blind-detection experiment since we use the threshold to find theprobability of a non-detection, not to claim a discovery.) InTable 3 we list the results of this exercise.The probability that all seven TDFs in our samplehosted jets, but were not detected due to Doppler boostingis P = Π i P i , with P i being the probability of the observa-tions of each TDF candidate, as listed in Table 3. We alsoconsider the possibility that, given our observations, at leastone of the seven TDFs hosted a jet, P ≥ . This is obtained bytaking the mean value of the product of all combinationsof the seven P i ’s (e.g., the probability that only one jetwas launched is P = (cid:80) i P i / j = 5, four of theseven TDF candidates (D23H-1, TDE2, PTF10iya, PS1-10jh) should have yielded a detection above the 2- σ level,hence P = 0, while P ≥ =2%. The probability that all ofthe other three TDF candidates hosted a jet is 1.7%. Forthe most conservative scenario (Eq. 3 b ), P = 48%, while Table 3.
Probability (%) that the jet orientation is such thatthe predicted flux is below the 2 σ -level of our 5 GHz observation.name Internal jet model Sw 1644+57, off-axis a c b Γ j = 2 Γ j ∝ t − . D1-9 39 78 83 49 17D3-13 62 89 91 52 26TDE1 7 92 100 0 0D23H-1 0 52 70 0 0TDE2 0 75 98 20 1PTF10iya 0 86 95 0 0PS1-10jh 0 95 97 0 0
Notes.
Zero probability implies that the predicted flux is abovethe threshold even for i = π/
2, while P i = 100% implies thedata cannot constrain the model. In the second to fourth columnwe list the results for the internal jet model, for the optimisticto the conservative scenario (Eq. 3), for Γ j = 5. In Fig. 3 weshow the results for lower Lorentz factors. In the fifth and sixthcolumn we give the probably of detecting a jet that is identical toSw 1644+57, but observed off-axis, using two different estimatesof the light curve past the last available observation (see sec.2.1). for the realistic scenario (Eq. 3 c ) this is lower at 21%. InFig. 3 we show P and P ≥ for lower Lorentz factors; atΓ j <
3, the hypothesis that all seven TDFs hosted a jet isruled out at 95% confidence for all three scenarios of theinternal jet model.Our upper limits also constrain the possibility that a jetsimilar to Sw 1644+57 was launched after the disruption.To place the Jansky VLA observations on the estimatedoff-axis light curve (Eq. 2), we equate the time of disrup-tion to the time of the
Swift trigger and we scale the fluxusing ( d L, Sw /d L ) , with d L, Sw the luminosity distance ofSw 1644+57. From Fig. 1 we see that our upper limits onthe radio flux of five TDFs (TDE1, D23H-1, PTF10iya,and PS1-10jh) are inconsistent with the estimated off-axislight curve of Sw 1644+57 for all viewing angles and bothversions of the late-time evolution we considered in sec. 2.1. A different method to test whether jets like Sw 1644+57 arecommon to stellar tidal disruptions is to compute the rateof these transients. The snapshot rate (or areal density) ata given flux density limit F ν, lim can be estimated directlyfrom the Sw 1644+57 light curve: R ( F ν, lim ) ∼ × − Γ − j (cid:18) F ν, Sw F ν, lim (cid:19) / ∆ T ˙ N TDJ − ρ BH × − Mpc − deg − . (4)Here ∆ T is the time in years that the flux of Sw 1644+57is above F ν, Sw , ρ BH is the black hole density, and ˙ N T DJ isthe rate of stellar tidal disruptions with jets. If Sw 1644+57was a typical stellar tidal disruption, this rate should be ofthe same order as the TDF rate inferred from soft X-ray(Donley et al. 2002) or optical (van Velzen & Farrar 2012)surveys, i.e., ˙ N TDJ ∼ − yr − .The 5 GHz light curve of Sw 1644+57 implies ∆ T ≈ F ν, Sw = 20 mJy. For Γ j = 2, we can thus obtain thesnapshot rate for a radio variability survey with a thresh-old at 10 mJy: R (10 mJy) = 5 × − deg − . This rate is . . . . . . . Γ j . . . . . . P r ob a b ilit y Always radio-loud (a)Conservative (b)Realistic (c)
Figure 3.
The probability of our data (i.e., no flux above thetwo times the image rms) for the three scenarios of the internalmodel (Eq. 3). The solid lines show P the probability that allseven TDFs we observed indeed hosted a jet; the dashed linesshow P ≥ the probability that at least one flare hosted a jet. ForΓ j < close to the existing upper limits on the snapshot rate at5 GHz (e.g., Scott 1996; Bower et al. 2007, 2011) – see Frailet al. (2012) for a review. Hence the observed light curveof Sw 1644+67 implies that near-future radio variabilitysurvey will either measure or constrain ˙ N TDJ .
5. Conclusion & Discussion
We obtained upper limits at the ∼ µ Jy level of the 5 GHzflux of seven stellar tidal disruptions events that were dis-covered with optical/UV imaging surveys. This is three or-ders of magnitude lower than the recently discovered TDFswith radio emission, suggesting that stellar tidal disrup-tions come in different flavors, ranging from radio-loud toradio-quiet (or radio-silent). To explore how this conclusionwould be biased by the large possibile parameter range in-herent to TDFs, we compared our upper limits to currentlyavailable jet models, taking into account Doppler boostingand temporal evolution of the radio emission.We used our observations to constrain the jet model ofvan Velzen et al. (2011). For a jet Lorentz factor of Γ j = 5,we can rule out the optimistic (“alway radio-loud”) scenariofor four of the seven flares. The probability that the otherthree TDF candidates did launch such jets, but are notdetected because Doppler boosting reduced the flux belowtwo times the image rms is only 4%. The hypothesis thatall events hosted jets that only becomes radio-loud whenthe fallback rate drops below 2% of the Eddington accre-tion rate (i.e., as observed in stellar mass black holes) isless constrained. Only for jets with Γ j < jet. Under the conservative assumption that jet Lorentzfactor is constant (Γ j = 2), the estimated off-axis lightcurves of this relativistic TDF are inconsistent with thenon-detection for four of the seven flares, for all possibleobserver angles. The hypothesis that all of the other TDFshosted jets identical to Sw 1644+57 is ruled out at the 95%confidence level.Our results are not sensitive to our assumption that thetime of disruption equals the time of the Swift trigger. Ifthe hard X-rays of the jet are emitted only after ten timesthe fallback time ( ∼ A V = 3 − Acknowledgements.
SvV would like to thank the anonymous ref-eree for the quick reply and useful comments. The VLA is oper-ated by the National Radio Astronomy Observatory, a facility of theNational Science Foundation operated under cooperative agreementby Associated Universities, Inc.
References
Abramowicz, M. A. & Fragile, P. C. 2013, Living Reviews inRelativity, 16, 1 Bade, N., Komossa, S., & Dahlem, M. 1996, A&A, 309, L35Berger, E., Zauderer, A., Pooley, G. G., et al. 2012, ApJ, 748, 36Best, P. N. & Heckman, T. M. 2012, MNRAS, 421, 1569Bloom, J. S., Giannios, D., Metzger, B. D., et al. 2011, Science, 333,203Bogdanovi´c, T., Eracleous, M., Mahadevan, S., Sigurdsson, S., &Laguna, P. 2004, ApJ, 610, 707Booth, R. S., de Blok, W. J. G., Jonas, J. L., & Fanaroff, B. 2009,ArXiv:0910.2935Bower, G. C. 2011, ApJ, 732, L12Bower, G. C., Metzger, B. D., Cenko, S. B., Silverman, J. M., &Bloom, J. S. 2013, ApJ, 763, 84Bower, G. C., Saul, D., Bloom, J. S., et al. 2007, ApJ, 666, 346Bower, G. C., Whysong, D., Blair, S., et al. 2011, ApJ, 739, 76Burrows, D. N., Kennea, J. A., Ghisellini, G., et al. 2011, Nature, 476,421Cenko, S. B., Bloom, J. S., Kulkarni, S. R., et al. 2012a, MNRAS,420, 2684Cenko, S. B., Krimm, H. A., Horesh, A., et al. 2012b, ApJ, 753, 77Condon, J. J., Cotton, W. D., Greisen, E. W., et al. 1998, AJ, 115,1693De Colle, F., Guillochon, J., Naiman, J., & Ramirez-Ruiz, E. 2012,ApJ, 760, 103Donley, J. L., Brandt, W. N., Eracleous, M., & Boller, T. 2002, AJ,124, 1308Drake, A. J., Djorgovski, S. G., Mahabal, A., et al. 2011, ApJ, 735,106Esquej, P., Saxton, R. D., Freyberg, M. J., et al. 2007, A&A, 462, L49Esquej, P., Saxton, R. D., Komossa, S., et al. 2008, A&A, 489, 543Evans, C. R. & Kochanek, C. S. 1989, ApJ, 346, L13Falcke, H. & Biermann, P. L. 1995, A&A, 293, 665Falcke, H., K¨ording, E., & Markoff, S. 2004, A&A, 414, 895Falcke, H., Malkan, M. A., & Biermann, P. L. 1995, A&A, 298, 375Falcke, H., Sherwood, W., & Patnaik, A. R. 1996, ApJ, 471, 106Fender, R. P. 2001, MNRAS, 322, 31Fender, R. P. 2012, in IAU Symposium, Vol. 285, IAU Symposium,ed. R. E. M. Griffin, R. J. Hanisch, & R. Seaman, 11–16Fender, R. P., Belloni, T. M., & Gallo, E. 2004, MNRAS, 355, 1105Frail, D. A., Kulkarni, S. R., Ofek, E. O., Bower, G. C., & Nakar, E.2012, ApJ, 747, 70Frank, J. & Rees, M. J. 1976, MNRAS, 176, 633Gezari, S., Basa, S., Martin, D. C., et al. 2008, ApJ, 676, 944Gezari, S., Chornock, R., Rest, A., et al. 2012, Nature, 485, 217Gezari, S., Heckman, T., Cenko, S. B., et al. 2009, ApJ, 698, 1367Gezari, S., Martin, D. C., Milliard, B., et al. 2006, ApJ, 653, L25Ghisellini, G. & Celotti, A. 2001, A&A, 379, L1Giannios, D. & Metzger, B. D. 2011, MNRAS, 416, 2102Granot, J., Piran, T., & Sari, R. 1999, ApJ, 527, 236Greiner, J., Schwarz, R., Zharikov, S., & Orio, M. 2000, A&A, 362,L25Grupe, D., Thomas, H.-C., & Leighly, K. M. 1999, A&A, 350, L31Guillochon, J. & Ramirez-Ruiz, E. 2012, arXiv:1206.2350Ho, L. C. 1999, ApJ, 516, 672Jester, S. 2008, MNRAS, 389, 1507Kellermann, K. I., Sramek, R., Schmidt, M., Shaffer, D. B., & Green,R. 1989, AJ, 98, 1195Komossa, S. 2002, in Lighthouses of the Universe, ed. M. Gilfanov,R. Sunyeav, & E. Churazov, 436–442Komossa, S. & Bade, N. 1999, A&A, 343, 775Komossa, S. & Greiner, J. 1999, A&A, 349, L45Komossa, S., Zhou, H., Wang, T., et al. 2008, ApJ, 678, L13K¨ording, E. G., Jester, S., & Fender, R. 2006, MNRAS, 372, 1366K¨ording, E. G., Jester, S., & Fender, R. 2008, MNRAS, 383, 277Krolik, J. H. & Piran, T. 2012, ApJ, 749, 92Lei, W.-H. & Zhang, B. 2011, ApJ, 740, L27Levan, A. J., Tanvir, N. R., Cenko, S. B., et al. 2011, Science, 333,199Leventis, K., van Eerten, H. J., Meliani, Z., & Wijers, R. A. M. J.2012, MNRAS, 427, 1329Lidskii, V. V. & Ozernoi, L. M. 1979, Soviet Astronomy Letters, 5,16Lin, D., Carrasco, E. R., Grupe, D., et al. 2011, ApJ, 738, 52Lind, K. R. & Blandford, R. D. 1985, ApJ, 295, 358Lodato, G. & Rossi, E. M. 2011, MNRAS, 410, 359Loeb, A. & Ulmer, A. 1997, ApJ, 489, 573Maccarone, T. J. 2003, A&A, 409, 697Maksym, W. P., Ulmer, M. P., & Eracleous, M. 2010, ApJ, 722, 1035Merloni, A., Heinz, S., & di Matteo, T. 2003, MNRAS, 345, 1057