The rate of optical tidal disruption flares Featuring implications - PowerPoint PPT Presentation

The rate of optical tidal disruption flares Featuring implications for jet physics Sjoert van Velzen Hubble Postdoctoral Fellow The Johns Hopkins University Glennys Farrar, Suvi Gezari, James Guillochon, Elena Rossi, Heino Falcke, Elmar Körding, Dale Frail, Nadia Blagorodnova Aspen Jan-22-2015

More motivation ‣ Probing the evolution of stellar orbits: - Rate with galaxy mass, redshift, type - IMBHs (Hagai Perets talk) ‣ Connection with the Galaxy: - Hyper velocity stars and S- stars (eg, Bromley+ 2012) ‣ General relativity: a/M - Event horizon and spin (Kesden 2011) 2 S. van Velzen Aspen 2015

Non-trivial assignment • Systematic search • Well-sampled light curves • Decent model light curves 3 S. van Velzen Aspen 2015

Requirements to measure an event rate • Completed surveys: • ROSAT (3) • GALEX (3) • SDSS Stripe 82 (2) • Ongoing: • XMM ( ≈ 6) • PTF (3 or 4) • Pan-STARRS (2) • Future surveys: Gaia, eROSITA, BlackGEM, Atlas, ZTF , LSST 4 S. van Velzen Aspen 2015

SDSS Stripe 82 • 300 deg 2 , 10 yr, u,g,r,i,z • m < 22.5 • ~2 million galaxies Flux • 70 observations per galaxy • Systematic search for all nuclear flares in galaxies Time (day) 5 S. van Velzen Aspen 2015

Background removal: supernovae • Cut for nuclear flares: r < 0.2” • Quality cut: 3 detections in u,g , r • 42 nuclear flares • No additional variability: 2 flares r (van Velzen+ 2011) 6 S. van Velzen Aspen 2015

The SED of TDE is hot and slows little/no cooling Cooling PS1-11af Spectral energy distribution (van Velzen+ 2011) 7 S. van Velzen Aspen 2015

Detection rate in other surveys N obs ∝ f sky F − 3 / 2 ˙ lim Survey F lim (mag) f sky N obs (1/yr) GAIA 19 1 4 PTF 21.5 0.2 13 PS1 MD 24.5 0.0012 10 0.5 LSST 24.5 4000 (van Velzen+ 2011) 8 S. van Velzen Aspen 2015

Theoretical setup for finding the rate N gal N = N TDF ˙ ✏ i ˙ X N TDF = ⌧ N i N gal ⌧ ✏ i “E ff ective-galaxy-year” N X ✏ ≡ N � 1 ✏ i i E ffi ciency for given light curve: Monte Carlo 9 S. van Velzen Aspen 2015

Models & Scenarios • Correction for captures: ‣ Exponential (a ≈ 0.5) ‣ Step-function at 10 8 M ⊙ • M BH scaling: ‣ “Standard” (Harning & Rix 2008) ‣ “Broken” (Graham 2012) • Model light curves: M K ‣ Empirical: SDSS and PS1 ‣ Model light curves 10 S. van Velzen Aspen 2015

Lodato & Rossi (2011) Absolute mangitude (AB) Absolute mangitude (AB) TDE1 fit PS1-10jh − 20 − 20 PS1-11af TDE1 − 18 − 18 − 16 − 16 TDE2 fit PS1-10jh − 14 − 14 PS1-11af TDE2 0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350 Rest-frame days since disruption Rest-frame days since disruption 11 S. van Velzen Aspen 2015

Guillochon, Manukian, and Ramirez-Ruiz (2014 ) Absolute mangitude (AB) Absolute mangitude (AB) TDE1 fit TDE2 fit PS1-10jh PS1-10jh − 20 − 20 PS1-11af PS1-11af TDE1 TDE2 − 18 − 18 − 16 − 16 − 14 − 14 − 100 − 50 0 50 100 150 200 250 300 − 100 − 50 0 50 100 150 200 250 300 Rest-frame days since peak Rest-frame days since peak 12 S. van Velzen Aspen 2015

Effective-galaxy-year distribution van Velzen & Farrar (2014) 13 S. van Velzen Aspen 2015

Results • Uncertainty Rate Model ‣ Poisson: factor ~2 (yr -1 galaxy -1 ) ‣ Due to M BH scaling: ~2 Empirical 2.0 10 -5 ‣ Due to light curves models: 50% Lodato & Rossi 1.7 10 -5 ‣ Upper limit is model-independent 1.9 10 -5 Guillochon et al. Upper limit < 2 10 -4 14 S. van Velzen Aspen 2015

Comparison to theory • Theoretical rate ~10 times higher 500 10 6 � M h � M � � 10 7 Fallback Guillochon & Ramirez-Ruiz (tomorrow) ‣ Dust obscuration Accretion 400 Slowed ‣ TDE physics: circularization Rise affected 300 N Prompt 200 ‣ Occupation fraction (!) 100 • X-rays could help, however: 500 10 7 � M h � M � � 10 8 400 ‣ ROSAT: 9 x 10 -6 yr -1 (Donley+ 2002) 300 ‣ XMM: 2 x 10 -4 yr -1 (Esquej + 2009) N 200 100 0 � 2 � 1 0 1 2 3 Log 10 � t peak � � yr � 15 S. van Velzen Aspen 2015

Dust in TDE host galaxies: Mid-IR light curve, 6 months after optical detection Mendez & van Velzen (in prep) 16 S. van Velzen Aspen 2015

A two-minute radio detour… 17 S. van Velzen Aspen 2015

Implication for jetted TDEs Flux Density (mJy) Zauderer+ (2013) 1 10 0 10 15 GHz − 1 10 8.4 GHz (/2) 1 2 3 10 10 10 ✓ F ν , Sw Time (d) ◆ 3 / 2 R ( F ν , lim ) ∼ 8 × 10 − 3 Γ − 2 44+57 extending to 600 d. The data at 5 216 d were previously presented in Z j F ν , lim ∆ T ˙ N TDJ ρ BH 5 × 10 − 3 Mpc − 3 deg − 2 . 10 − 5 van Velzen+ 2013 ; Donnarumma+ 2015; Mimica+ 2015 18 S. van Velzen Aspen 2015

The most common transient on the radio sky? 1 0.843 GHz 10 Carilli+03 1.4 GHz B2 4.9 GHz Frail+03 0 10 B1 S w Areal Density (>f ν ) [deg − 2 ] i f t J 1 6 O − 1 4 10 4 Swift J1644+57 + 5 Stripe 82 (VLA) 7 type − II RSN deVries+04 − 2 10 Orphan long − GRB NS − NS mergers Croft+10 Bannister+10 Gregory & Taylor 86 − 3 10 Levinson+02/Gal − Yam+06 SN1998bw like − 4 Scott96 10 − 1 0 1 2 10 10 10 10 f ν [mJy] Frail et al. (2012), TDE jet rate from van Velzen et al. (2013) 19 S. van Velzen Aspen 2015

Tidal disruption jets: two models • Internal model • External model (van Velzen, Falcke & Farrar 2010 ; (Giannios & Metzger 2011 ; Metzger, Giannios, Mimica 2011) van Velzen, Körding & Falcke 2011) ‣ Inspired by AGN jets ‣ Inspired by GRB jets (eg, Granot & Sari 1999) ‣ Emission from matter ‣ Interaction of forward/reverse injected in the jet from shock with environment the disk ‣ On-axis or isotropic ‣ Include accretion state- transitions ‣ Function of inclination (Doppler boosting) 20 S. van Velzen Aspen 2015

Follow-up observations: JVLA, 5 GHz, 10 μ Jy rms • van Velzen et al. (2013): ‣ followed-up all optical/UV TDE ‣ No detections • Bower et al. (2012): ‣ Followed-up all X-ray TDE ‣ Two detected, both from ROSAT ( IC 3599 and RX J1420.4+5334 ) ‣ Very unlikely to be TDE jets • Soderberg et al. (in prep): ‣ No detections 21 S. van Velzen Aspen 2015

Off-axis light curves: conservative model 10 2 30 � 50 � Flux density (mJy) 10 1 70 � 90 � 10 0 22 GHz 6 GHz 10 � 1 10 � 2 10 � 3 0 2 4 6 8 Time since Swift trigger (yr) van Velzen+ (2013) 22 S. van Velzen Aspen 2015

Off-axis light curves: simulations Mimica+ (2015) 23 S. van Velzen Aspen 2015

Conclusions & Outlook • Jets from tidal disruptions: ‣ Not common (<10 % of TDE) ‣ Upcoming radio surveys could detect few per year • Rate based on systematic search: ‣ ~2 x 10 -5 yr -1 galaxy -1 • Discrepancy with theory ‣ Circumnuclear dust or something even more exciting? • Combine X-ray, UV, optical surveys 24 S. van Velzen Aspen 2015

. 25 S. van Velzen Aspen 2015

Efficiency: catalog selection + difference imaging 26 S. van Velzen Aspen 2015

Galaxy SEDs Mendez & van Velzen (in prep) 27 S. van Velzen Aspen 2015

Could there flares be supernovae? • Not normal SNe: more blue, little cooling • UV detection > 2 yr after the flare • Based on geometry: ‣ P (SN) < 2% • New kind of “nuclear” core collapse SN? SN UV upper limit ‣ Never observed before (?) ‣ Would require factor 1000 suppression outside nucleus 28 S. van Velzen Aspen 2015

TABLE 1 Light curve model efficiencies & resulting optical TDF rates. Name Mean e ffi ciency TDF Rate (yr − 1 galaxy − 1 ) (%) < 1 . 5 × 10 − 4 SDSS-only 0.13, 0.62 2 . 0 × 10 − 5 PS1 events (10jh, 11af) 1.0 1 . 5 × 10 − 5 Phenomenological 1.4 M BH scaling: Correction for captures: H¨ aring & Rix (2004) Step-function Exponential 1 . 2 × 10 − 5 1 . 7 × 10 − 5 Disk+Wind 0.83, 3.3 1 . 8 × 10 − 5 1 . 9 × 10 − 5 GMR14 1.2 M BH scaling: Correction for captures: Graham (2012) Step-function Exponential 2 . 1 × 10 − 5 3 . 2 × 10 − 5 Disk+Wind 0.22, 1.5 1 . 2 × 10 − 5 1 . 3 × 10 − 5 GMR14 1.6 29 S. van Velzen Aspen 2015

Could these flares originate from AGN? • Flares are more blue than QSO (in their high-state) • Host spectra show no sign of active black hole • Flux increases very large: P(AGN)~10 -7 ,10 -5 • No additional variability: P(AGN)~10 -6 ,10 -5 • Radio non-detection: F < 20 μ Jy, < 10 28 erg s − 1 Hz − 1 30 S. van Velzen Aspen 2015

Flare selection: catalog cuts χ 2 / DOF > 5 ∆ F/F mean > 0 . 1 ∆ F/ σ > 7 • Factor 100 reduction: 21,383 follow-up candidates 31 S. van Velzen Aspen 2015

Snapshot rate B07/F12 10 0 Snapshot rate (deg − 2 ) Mooley14 10 − 1 deVries04 10 − 2 Croft10 ThKAT FIRST-NVSS 10 − 3 VAST-Deep SKA Scott96 10 − 4 10 − 5 100% like Sw J1644 (5 GHz) 10 − 6 100% like Sw J1644 (1 GHz) van Velzen et al. (2011): Optimistic (1 GHz) 10 − 7 van Velzen et al. (2011): Conservative (1 GHz) 10 − 8 10 − 1 10 0 10 1 10 2 Flux density (mJy) Donnarumma+ 2015 32 S. van Velzen Aspen 2015

Observations: flaring state spectrum (TDE2) van Velzen+ 2011 33 S. van Velzen Aspen 2015