1. Tidal Disruptions of Stars by Supermassive Black Holes Suvi Gezari (Caltech) Chris Martin & GALEX Team Bruno Milliard (GALEX) Stephane Basa (SNLS)

2. Probing the mass of dormant black holes in galaxies Tidal disruption theory Candidates discovered by ROSAT Search for flares with GALEX GALEX tidal disruption flare detections Future detections Outline

3. Direct dynamical measurement of M BH is possible when R inf ≈ GM BH /  2 is resolved. Probing the Mass of Dormant Supermassive Black Holes Kormendy & Bender (1999) Ghez+ (2005) Milky Way M31

4. A dormant black hole will be revealed when a star approaches closer than R T ≈R star (M BH /M star ) 1/3, and is tidally disrupted. This is a rare event in a galaxy, occurring only once every 10 3 -10 5 yr depending on M BH and the nuclear density profile of the galaxy. Probing the Mass of Dormant Supermassive Black Holes Rees (1988)

5. L ≈ L Edd = 1.3x10 44 (M BH /10 6 M sun ) ergs s -1 Blackbody spectrum: T eff =(L Edd /  4  R T 2 ) 1/4. Start of flare: (t 0 -t D )  k -3/2 M BH 1/2 Power-law decay: dM/dt  (t-t D ) -5/3. The temperature, luminosity, and decay of the flare can be used as a direct probe of M BH. Probing the Mass of Dormant Supermassive Black Holes Evans & Kochanek (1989)  t -5/3

6. The ROSAT All-sky survey in 1990-1991 sampled hundreds of thousands of galaxies in the soft X-ray band (0.1-2.4 keV). Detected a large amplitude soft X-ray flare from 3 galaxies which were classified as non-active from ground based spectra. Follow-up narrow-slit HST/STIS spectroscopy confirmed the ground-based classifications of 2 of the galaxies (Gezari+ 2003). Previous Tidal Disruption Event Candidates Halpern, Gezari, & Komossa (2004) L flare /L 10yr = 240 L flare /L 10yr = 1000 L flare /L 10yr = 6000 HST Chandra

7. 50 cm telescope with a 1.2 deg 2 field of view. Simultaneous FUV/NUV imaging and grism spectra Data is time-tagged photon data (  t=5ms) accumulated in 1.5 ks eclipses. Some deep fields are revisited over a baseline of 2-4 years to complete deep observations. Take advantage of the UV sensitivity, temporal sampling, and large survey volume of GALEX to search for flares. Searching for Flares with GALEX 1350 Å 1750 Å 2800 Å | | |

8. Assume L=L Edd, and T eff =2.5x10 5 (M BH /10 6 M sun ) 1/12 K. The large K correction makes flares detectable out to high z. Estimated attenuation by HI absorption for z>0.6 from Madau (1995) Contrast with host early type spirals and elliptical galaxies not a problem for detection in the UV. Searching for Flares with GALEX 1350 Å 1750 Å 2800 Å | | | Gezari+ (in prep) 5x10 7 M sun 1x10 6 M sun

9. Estimate black hole mass function from Ferguson & Sandage (1991) luminosity function of E+S0 galaxies. Multiply by a factor of 2 for bulges in early-type spirals. Use M BH dependent event rate from Wang & Merritt (2004). Assume fraction of flares that radiate at L Edd from Ulmer (1999). Multiply by volume to which an L Edd flare can be detected in the FUV by a GALEX DIS exposure. Searching for Flares with GALEX 1350 Å 1750 Å 2800 Å | | | Gezari+ (in prep)

10. Match UV sources that vary between yearly epochs at the 5  level with the CFHT Legacy Survey optical catalog. Rule out sources with optical hosts with the colors and morphology of a star or quasar. Follow up galaxy hosts that do not have an hard X-ray detection with optical spectroscopy to look for signs of an AGN. Trigger Chandra TOO X-ray observations of our best candidates. Searching for Flares with GALEX 1350 Å 1750 Å 2800 Å | | | Gezari+ (in prep) stars QSOs galaxies x : X-ray source

11. AEGIS DEEP2 spectrum and ACS image of an early-type galaxy at z=0.3698. No evidence of Seyfert-like emission lines. No detection of hard X-rays. Archival Chandra observations during the flare detected a variable extremely soft X-ray source coincident with the galaxy. Tidal Disruption Flare Detections 1350 Å 1750 Å 2800 Å | | | Gezari+ (2006)

12. TOO VLT spectrum and CFHTLS image of an early- type galaxy at z=0.326. No evidence of Seyfert-like emission lines. No detection of hard X-rays. First optical detection of a tidal disruption flare. Triggered a Chandra TOO observation which detected an extremely soft X-ray source coincident with the galaxy. Tidal Disruption Flare Detections 1350 Å 1750 Å 2800 Å | | | Gezari+ (in prep)

13. Well described by a t -5/3 power-law decay. M BH =k 3 [(t 0 -t D )/0.11] 2 *10 6 M sun Tidal Disruption Flare Detections 1350 Å 1750 Å 2800 Å | | | Gezari+ (2006)Gezari+ (in prep) (t 0 -t D )/(1+z)=0.1-0.7 yr  k 3 (1-4)x10 7 M sun (t 0 -t D )/(1+z)=0.45±0.4 yr  k 3 (1.7±0.3)x10 7 M sun

14. T BB ≈ few x 10 5 K R BB ≈ 1 x 10 13 cm R T = 1.5 x 10 13 (M BH /10 7 M sun ) 1/3 cm R Sch = 3 x 10 12 (M BH /10 7 M sun ) cm Tidal Disruption Flare Detections Gezari+ (2006)Gezari+ (in prep) L bol = 6.5x10 44 ergs s -1 L bol > 1x10 44 ergs s -1

15. GALEX has proven to be successful in detecting tidal disruption flares. Goal is to measure the detailed properties and rate of the events to probe accretion physics, the mass of the black hole, and evolution of the tidal disruption rate. The next generation of optical synoptic surveys such as Pan-STARRs and LSST have the potential to detect hundreds of events. With a large sample we can probe the evolution of the black hole mass function, independent of studies of active galaxies. Future Detections

16. Stay Tuned for More Flares!