New Puzzles in Supermassive Black Hole Evolution

1. New Puzzles in Supermassive Black Hole Evolution Charles L. Steinhardt IPMU, University of Tokyo October 14, 2010 • Steinhardt & Elvis 2010, MNRAS, 402, 2637 (arxiv:0911.1355) • Steinhardt & Elvis 2010 MNRAS, in press (arxiv:0911.3155) • Steinhardt & Elvis 2010 MNRAS 406, L1 (arxiv:0912.0734) • Steinhardt, Elvis, & Amarie 2010, submitted

2. The supermassive black hole (SMBH) lifecycle • Seeding • Growth • Turnoff • Quiescence (well, almost)

3. The supermassive black hole (SMBH) lifecycle • Seeding • Growth: quasar phase (Soltan) • Turnoff • Quiescence (well, almost)

4. The supermassive black hole (SMBH) lifecycle • Seeding • Growth: quasar phase (Soltan) • Turnoff (M-s relation) • Quiescence (well, almost)

5. Quasar Luminosity Function Richards et al. (2006)

6. How to obtain black hole masses from one SDSS spectrum • Kepler’s Laws on broad emission line gas, so we need v,R. • Doppler broadening of spectral line  velocity • Supermassive black hole “mass ladder” • Continuum luminosity  radius • Comparison with reverbation masses implies ~0.4 dex uncertainty (more on this later!)

7. Quasar Mass Function Vestergaard et al. (2008)

9. Common beliefs about SMBHs • All quasars can radiate at the Eddington limit • Quasars are “light-bulbs”: either on (at Eddington) or off • Quasars “flicker” • Luminosity is a proxy for mass • Quasar dynamics come from host galaxy dynamics

10. Existing data Existing methods Existing catalogs But new methods

11. Existing data • Quasar catalog and spectra come from SDSS DR5 Existing methods Existing catalogs But new methods

12. Existing data • Quasar catalog and spectra come from SDSS DR5 • Virial Mass Estimation: Vestergaard/Peterson, McLure/Dunlop Existing methods Existing catalogs But new methods

13. Existing data • Quasar catalog and spectra come from SDSS DR5 • Virial Mass Estimation: Vestergaard/Peterson, McLure/Dunlop • Actual mass estimates: Shen et al. (2008) • Bolometric luminosities: Richards et al. (2006), Shen et al. (2008) Existing methods Existing catalogs But new methods

14. Existing data • Quasar catalog and spectra come from SDSS DR5 • Virial Mass Estimation: Vestergaard/Peterson, McLure/Dunlop • Actual mass estimates: Shen et al. (2008) • Bolometric luminosities: Richards et al. (2006), Shen et al. (2008) • Time to think two- (or three-) dimensionally! Existing methods Existing catalogs But new methods

16. 0.2 < z < 0.4, H‏

17. 0.2 < z < 0.4, H‏ SDSS Saturation Quasar Turnoff Detection Limit

18. 0.2 < z < 0.4, H‏ Quasar Turnoff Detection Limit

19. Virial mass estimation may be better than previously believed! Best-fit exponential decays: e-folding of 0.14-0.25 dex

20. 0.2 < z < 0.4, H‏ Quasar Turnoff Detection Limit

21. Quasars at 1.6 < z < 1.8

22. Quasars at 1.6 < z < 1.8

23. Quasars at 1.6 < z < 1.8

25. Best-fit sub-Eddington boundary slopes

26. Best-fit sub-Eddington boundary slopes Risaliti, Young, & Elvis (2009)

27. Common beliefs about SMBHs • All quasars can radiate at the Eddington limit • Quasars are “light-bulbs”: either on (at Eddington) or off • Quasars “flicker” • Luminosity is a proxy for mass • Quasar dynamics come from host galaxy dynamics FALSE!

28. Expected L/LE distribution at different M, 0.2<z<0.4 Normalized to peak

29. The L/LE distribution at different M, 0.2<z<0.4 Normalized to peak

30. The L/LE distribution at different M, 0.2<z<0.4 Normalized to peak

31. Common beliefs about SMBHs • Quasars radiate at the Eddington limit • Quasars are “light-bulbs”: either on (at Eddington) or off • Quasars “flicker” • Luminosity is a proxy for mass • Quasar dynamics come from host galaxy dynamics FALSE! TRUE! FALSE!

32. Common beliefs about SMBHs • Quasars radiate at the Eddington limit • Quasars are “light-bulbs”: either on (at Eddington) or off • Quasars “flicker” • Luminosity is a proxy for mass • Quasar dynamics come from host galaxy dynamics FALSE! TRUE! FALSE! MAYBE NOT?

33. SDSS quasar colors at high mass, low luminosity

34. Emission line ratios change at high mass Highest Mass Intermediate Mass Lowest Mass 1.2-1.4 0.8-1.0

35. Common beliefs about SMBHs • Quasars radiate at the Eddington limit • Quasars are “light-bulbs”: either on (at Eddington) or off • Quasars “flicker” • Luminosity is a proxy for mass • Quasar dynamics come from host galaxy dynamics FALSE! TRUE! FALSE! MAYBE NOT?

36. Luminosity at fixed mass, different z Redshift range 3.0-3.2 2.0-2.2 1.6-1.8 1.2-1.4 0.8-1.0

37. Common beliefs about SMBHs • Quasars radiate at the Eddington limit • Quasars are “light-bulbs”: either on (at Eddington) or off • Quasars “flicker” • Luminosity is a proxy for mass • Quasar dynamics come from host galaxy dynamics FALSE! TRUE! FALSE! MAYBE NOT? FALSE!

38. Comoving number density declines at different rates for different masses Log M (solar) 9.75-10.0 9.50-9.75 9.25-9.50 9.00-9.25

39. Timescales t(M), N(t) = N0e-t/t(M)

40. Common beliefs about SMBHs • Quasars radiate at the Eddington limit • Quasars are “light-bulbs”: either on (at Eddington) or off • Quasars “flicker” • Luminosity is a proxy for mass • Quasar dynamics come from host galaxy dynamics FALSE! TRUE! FALSE! MAYBE NOT? FALSE! SEEMINGLY FALSE!

41. Track sensitivity to 20% changes in parameters 20% changes in: M0 a k t0

42. Sample Track: 1.8 < z < 2.0

43. Sample Track: 1.6 < z < 1.8

44. Sample Track: 1.4 < z < 1.6

45. Sample Track: 1.2 < z < 1.4

46. Sample Track: 1.0 < z < 1.2

47. Allowed track parameters at M0=8.5, t0=3.5 Gyr Quasars are typically on for just 1-2 Gyr!

48. Allowed parameters for tracks originating at all times

49. What would we ideally use to study quasar accretion? • Mass and luminosity evolution of individual SMBH • All relevant host galaxy parameters Only one snapshot SDSS cannot see the galaxy

50. What would we ideally use to study quasar accretion? • Mass and luminosity evolution of individual SMBH • All relevant host galaxy parameters Quasars ARE like light bulbs! SDSS cannot see the galaxy