Radio emission from tidal disruption events

Transcript

1. Radio emission from stellar tidal disruptions Sjoert van Velzen (Radboud

   University Nijmegen) Heino Falcke, Glennys Farrar, Joseph Gelfand, Suvi Gezari, David Hogg, Elmar Körding, Nidia Morrell, Linda Östman, Dennis Zaritsky TKSP Meeting, June 28, 2011, Amsterdam

2. jun-28-2011 Sjoert van Velzen Stellar tidal disruptions 2

3. jun-28-2011 Sjoert van Velzen Stellar tidal disruptions • Rather valuable

   for astrophysics • Rare events: once every 104-5 per year per black hole (eg, Wang & Merritt 2004) • Few found in X-ray and UV surveys (eg, Komossa+ 2002; Gezari+ 2008) • Recently found in at optical wavelengths (van Velzen+ 2010; Drake+ 2011, Cenko+ 2011) • Expect 10 (PTF, PanStarrs) to 1000 (LSST) per year 3

4. jun-28-2011 Sjoert van Velzen What about radio wavelengths? • No

   radio detections, yet • Accretion <--> Jets • Just like X-ray binaries (!) (eg, Fender+ 2004) • Let’s build a model: simple, conservative, empirical ‣ Synchrotron emission from jet core ‣ But coherent emission will be awesome 4

5. jun-28-2011 Sjoert van Velzen Extend jet-disk symbiosis 5 jet properties

   (Fender et al. 2004). In the quiescent mode (the hard-state) and during the onset of the burst, jets in X-ray binaries are radio-loud, while in the high-accretion mode (the soft-state) they are radio-quiet. The sudden enhancement of the accretion rate during a TDE, may move it through the different modes of accretion in two ways: directly into the radio-quiet soft-state, or into the soft-state via the radio-loud burst-state. Alternatively, the jet from a TDE may behave like a radio loud quasar at all times. We therefore consider three different scenarios for the fraction of accretion energy that is fed into the jet: qj =    0.2 all times (a) 2 × 10−3 ˙ M(t) > 2% ˙ M Edd (b) 0.2 t < t fallback (c) . (4) were each scenario reverts to the preceding one if the con- dition on t or ˙ M is not true (e.g., qj = 0.2 if ˙ M < 2% ˙ M Edd (Falcke & Biermann 1995; Körding+ 2008) 280 E. G. K¨ ording, S. Jester and R. Fender 21 22 23 24 25 26 27 28 29 I M − 24.5 25.0 25.5 26.0 26.5 27.0 27.5 28.0 28.5 29.0 z H M 4 7 ] r s / z H / W [ L g o L Figure 2. Normalizing the jet power measures from radio luminosities with sources of measured jet powers. On the ordinate (y-axis) we plot the 74- MHz radio luminosity and on the abscissa (x-axis) we give the absolute i-band magnitude. The line represents the linear relation fitted to the data (see text). Survey (VLSS1; Cohen et al. 2006, 2007a) using a matching ra- dius of 20 arcsec (the histogram of radial separations between the closest matched sources has a local minimum at this matching ra- dius). We use 74-MHz fluxes as this frequency is near the target of 151 MHz, especially for the large number of the quasars around Pjet Ljet Lν ∝ (qjLd )17/12 (Rees 1989) ˙ M(t) ∝ t−5/3

6. jun-28-2011 Sjoert van Velzen Light curves 6 Radio jets from

   stellar tidal disruptions L3 100 101 time since disruption (yr) 1027 1028 1029 1030 1031 1032 jet luminosity (erg s −1 Hz −1) M BH = 1×107M⊙ 10 GHz, always radio-loud (a) 1.4 GHz, always radio-loud (a) 200 MHz, always radio-loud (a) 1.4 GHz, loud for ˙ M < 2% (b) 1.4 GHz, burst (c)

7. jun-28-2011 Sjoert van Velzen Then we got lucky ... •

   Thanks to Rudy! 7

8. jun-28-2011 Sjoert van Velzen GRB 110328A / Swift 1644+57 •

   Over 46 Telegrams & Circulars • 5 GHz detection 2 days after Swift trigger (EVLA) • 1.7 mJy @ 8.4 GHz (VLBI) • 15 mJy @ 98 GHz (CARMA) 8 credit: HST NASA Birth of a Relativistic Outflow in Swift J164449.3+573451 109 1010 1011 1012 1013 1014 1015 1016 1017 1018 1040 1041 1042 1043 1044 1045 1046 1047 1048 2011 March 30 UT 2011 April 4 UT 2011 April 9 UT 2011 April 16 UT BL Lac Rest Frequency (Hz) !L ! (erg/s) Figure 2. Spectral energy distributions (SEDs) of Swift J164449.3+573451 from radio to X-ray (Zauderer+ 2011)

9. jun-28-2011 Sjoert van Velzen It fits “naturally” 9 Always radio-loud

   Burst M BH = 106 M⊙, i obs = 1◦, Γ j = 5 van Velzen et al. (2011) arXiv:1104:4105

10. jun-28-2011 Sjoert van Velzen Summary • Strong candidate stellar tidal

    disruption have been detected • We can expect many per year from optical surveys • Delay at radio frequencies • Should we care? 10

11. jun-28-2011 Sjoert van Velzen Radio variability surveys 11 10−1 100

    101 102 Flux density (mJy) 10−9 10−8 10−7 10−6 10−5 10−4 10−3 10−2 10−1 100 101 Snapshot Rate (deg−2) B07 B07, 2 month FIRST-NVSS ATATS MOST BS11 Scott96 deVries04 LOFAR VAST-Deep SKAMP 10 GHz, always radio-loud (a) 1.4 GHz, always radio-loud (a) 200 MHz, always radio-loud (a) 1.4 GHz, loud for ˙ M < 2% (b) 1.4 GHz, burst (c) 5 GHz, Giannios & Metzger (2011)

12. jun-28-2011 Sjoert van Velzen Finding long extra-galactic radio transients •

    Do MSSS twice ... • Use optical data: SDSS photometric galaxies ‣ over 14,000 sq degree (208,478,448 galaxies) ‣ median z~0.5 ‣ 0.1% match to FIRST (1.4 GHz, ~10,000 sq. degree) (Ivezic+ 2002) ‣ Interesting if detected at >30 mJy in LOFAR but not detected in FIRST 12 GMRT Bootes field (Intema PhD thesis)

13. jun-28-2011 Sjoert van Velzen Let’s go hunting! 13

14. jun-28-2011 Sjoert van Velzen Back-up: details of model 14 that

    we integrate using only the photons that will arrive simultaneously at the observer, tr (t, z) = t−z cos(i) c−1 with i the angle between the jet and observer, in the rest-frame of the jet. Note that for observed angles cos(i obs ) < βj , we have tr > t; the photons from the middle of the jet arrive simultaneous with photons emitted further ahead, i.e., the jet appears to be seen from behind in the observer frame; we refer to Jester (2008) for a detailed discussion of retardation in jets. While in the classic jet model the value of z ssa is ab- sorbed into the normalization (C eq ), for the time-variable model it sets the timescale of emission and thus needs to be determined. From τ ∝ zκ syn / sin(i) = 1, where κ syn ∝ B4 is the synchrotron emission coefficient, we get z ssa = 1 pc f GHz ν/δ ￿ qj (t) 0.2 Ld (t) 1045 erg s−1 ￿2 3 ￿ βj sin( i 30 ◦ ) 5 γj ￿1 3 (3) (FB95, Eq. 52), with γj the Lorentz factor of the jet and f ∼ 1, is a factor that dependents on the details of equipar- tition. We preform a check on the latter using observations zdec ∼ 10 pc ￿ qj 0.2 Ld 1045 erg s−1 ￿1/3 Lν (t) = C eq δ2 ￿ zdec 0 dz z2￿ syn (tr, z, ν/δ)Θ ssa (tr, z, ν/δ) .(2) Here Θ ssa (t, z, ν) is a step function that enforces a crude radiative transfer: it is zero for z < z ssa (t) and unity for z > z ssa (t). The retarded time, tr , is introduced to ensure that we integrate using only the photons that will arrive simultaneously at the observer, tr (t, z) = t−z cos(i) c−1 with i the angle between the jet and observer, in the rest-frame of the jet. Note that for observed angles cos(i obs ) < βj , we have tr > t; the photons from the middle of the jet arrive simultaneous with photons emitted further ahead, i.e., the jet appears to be seen from behind in the observer frame; we

15. jun-28-2011 Sjoert van Velzen Back-up: integration to obtain rate 15

    4 SNAPSHOT RATE Using the model presented in section 2, we can predict how many jets are visible above a certain flux limit (F lim ) at any moment in time, N(F lim , ν) = (4π)−1 ˙ N tde ￿ dΩ obs ￿ dz 4πd2 C (z) × ￿ dM BH φ BH τ eff (Lν , dL (z), F lim ) . (5) Here dC (z) and dL (z) are the co-moving and luminosity dis- tance1, respectively and φ BH is the black hole mass function. The integration over viewing angles, dΩ obs , accounts for the effects of Doppler boosting. Finally, our jet model enters via τ eff (Lν (M BH , i obs ), dL, F lim ) or the “effective time” given by the part of the light curve that obeys Lν (t)/(4πd2 L ) > F lim . ¨ ording and Falcke 100 101 ince disruption (yr) NGC 5905 TDE2 D3-13 z) flux for TDE2 MBH ∼ 5 × 107 M⊙ D3-13, MBH ∼ 2 × 107 M⊙ (Gezari 10−1 100 101 102 Flux density (mJy) 10−9 10−8 10−7 10−6 10−5 10−4 10−3 10−2 10−1 100 101 Snapshot Rate (deg−2) B07 B07, 2 month FIRST-NVSS ATATS MOST BS11 Scott96 deVries04 LOFAR VAST-Deep SKAMP 10 GHz, always radio-loud (a) 1.4 GHz, always radio-loud (a) 200 MHz, always radio-loud (a) 1.4 GHz, loud for ˙ M < 2% (b) 1.4 GHz, burst (c) 5 GHz, Giannios & Metzger (2011) Figure 3. The snapshot rate of TDE jets. We show 2-σ upper- limits from: Scott (1996) and Bower et al. (2007, B07) at 5 GHz,

16. jun-28-2011 Sjoert van Velzen Back-up: GRB 110328A 16 10 Zauderer

    et al. 100 101 4.9 GHz (m p = 0.38) 6.7 GHz (m p = 0.72) 15 GHz (m p = 0.07) 25 GHz (m p < 0.03) 0.3−10 keV (!30) a Flux density (mJy) 101 101 44 GHz 90 GHz 200 GHz 225 GHz 345 GHz b Flux density (mJy) Time since March 25 UT (d) Figure 3. Radio light curves of Swift J164449.3+573451 at 5 − 345 GHz re VLSS 2x2 deg