1. Coeval Evolution of Galaxies and Supermassive Black Holes : Cosmological Simulations J. A. de Freitas Pacheco Charline Filloux Fabrice Durier Matias Montesino Collaborators J. Silk – Oxford T.P. Idiart – USP Miguel Preto - Heidelberg

2. The Facts

3. Observational evidences for the existence of massive BH in the core of galaxies Sagittarius A* - galactic centre Ghez et al. 1998 DMO in M87, M84 and NGC4261

4. Black holes and galaxies Strong correlations are observed between the black hole mass and : Stellar velocity dispersion : Stellar bulge mass : Stellar bulge luminosity : Co-evolution of SMBH and galaxies Dark halo mass : Tremaine et al. 2002 ; Gebhardt et al.2000 Haring and Rix, 2004 Marconi and Hunt, 2003 Baes et al. 2003

5. Lower Mass Limits * Lower limits - negative searches for intermediate mass black holes - upper limits for M33 (3  10 3 M  ) and NGC 205 (3.8  10 4 M  ) -indirect evidence for IMBH in NLSeyf1 (8  10 4 - 8  10 6 M  )

6. Origin & Evolution of SMBH

7. Origin of Seeds (*) Intermediate mass black holes (10 3 -10 4 M  ) formed in : a) the collapse of primordial gas clouds (Haehnelt & Rees 1993) b) the core collapse of star clusters formed in starburts (Shapiro 2004) (*) Collapse of primordial massive stars (100-300 M  ) formed in high density peaks of the primordial fluctuation spectrum

8. Cosmological simulations Advantages follow up of seeds gas dynamics & merger tree follow up of the star formation history Difficulties Two extreme scales : Galaxies interactions : several kpc Black hole physics : sub-pc scales Number of ParticlesMass resolution (gas) V=(50 Mpc) 3 CPU/hours/run (128 processors) 2  160 3 5.35  10 8 M  4 000 2  192 3 3.09  10 8 M  12 000 2  256 3 1.30  10 8 M  60 000  v =0.7  m =0.3  b h 2 = 0.0224 h=0.70  8 =0.9

9. The Code GADGET-II Springel 2005 Gravitation (tree code) Hydrodynamics (SPH) DARK MATTER GAS SMBH Introduction of BH seeds at potential minima (z=15) BH Growth (« disk » and HLB mode) AGN activity (feedback) STARS Star formation (conversion of gas into tars) Ionisation, heating and radiative cooling Supernovae (type Ia and II) Galactic winds Metal enrichment SMBH coalescences during galaxy mergers

10. Code Parameters Energy injected by supernovae  Weight for the blast  energy per particle  ’’turbulent’’ diffusion efficiency  Accretion mode  spherical (Bondi – Hoyle)  ’’disk’’  AGN feedback gravitational energy  rotational energy  Jet angle  Jet length  100 - 400 kpc

11. Detection of Structures FoF SubFind Davis et al, 1985 ; Huchra and Geller, 1982 Springel et al, 2001 Structure determination

12. Properties of Galaxies

13. Dynamical Properties of Simulated Galaxies Faber-Jackson & Tully-Fisher relationsAngular momenta of blue & red galaxies ’’Red’’ galaxies  (U-V)>1.1 and (B-V)>0.8 ’’Blue’’ galaxies  otherwise

14. Properties of Simulated Galaxies Grey zone – SDSS data from Gallazzi et al. 2005

15. Properties of Simulated Black Holes

16. The mass function at z=0 All simulations give similar results, with BHMF slightly overestimated for M ● >10 7 M  BH seeds of 100 M  : evolution of massive pop III stars 192/160 : resolution disk/kerr : AGN feedback from accretion /rotation S : with higher SNIa efficiency

17. Evolution of the Black hole mass density Simulationρ ● (z=0) [M .Mpc -3 ] M BH,min [M  ] 160kerr5.0 x 10 5 2.4 x 10 4 160disk9.6 x 10 5 1.8 x 10 4 160diskS7.4 x 10 5 3.7 x 10 3 192disk8.2 x 10 5 1.2 x 10 3 Estimates : ρ ● = 2 - 9 x 10 5 M .Mpc -3 Chokshi and Turner 1992 ; Salucci et al., 1999 ; Aller and Richstone, 2002 ; Marconi et al, 2004, … 192/160 : resolution disk/kerr : AGN feedback from accretion /rotation S : with higher SNIa efficiency Assuming bolometric luminosity proportional to the accretion rate Black hole mass density at z=0

18. The M ● - σ relation Cygnus A, NGC 5252, NGC 3115 and NGC 4594 Good agreement, except for the four galaxies having black holes apparently too massive. 192disk simulation

19. The M ● - M halo relation Good agreement with Baes et al, 2003. M halo is directly extracted from simulations. 192disk simulation

20. Some problems: no SMBH at z ~ 6! No supermassive black holes at z=6  hierarchical growth No super-Eddington accretion rates are observed (resolution effect?)

21. Gravitational Waves from Coalescences

22. Coalescence of two massive BHs tThretThre Four regimes can be recognized: i)adiabatic – sequence of quasi-circular geodesic orbits ii)transition – near the innermost stable orbit iii)plunge – merger of the two horizons iv) ring-down – normal modes of the distorted black hole Maximum mean frequencies – adiabatic regime Total energy radiated under the form of GW in the adiabatic phase Frequency near the ’’last stable orbit’’

23. Expected contribution to the background Expected flux at the observer’s frame With  and total merger rate per comoving volume Equivalent density parameter  Fraction of mergers with a parameter  occurring at z

24. Ring-down contribution to the background Expected flux Equivalent density parameter  Duty cycle

26. Ring-down background (shot-noise – D <<1) Spin data: Daly 2009 Shot-noise

27. I. Conclusions Cosmological simulations are the best tool to study the coeval evolution of galaxies and their central black holes Properties of gas, galaxies and the growth of supermassive black holes depend strongly on feedback mechanisms, in particular the downsizing effect Simulated galaxies have adequate mass profiles, satisfying the Faber-Jackson (red galaxies), the Tully-Fisher (blue galaxies) and the mass-metallicity relation

28. II. Conclusions Seeds ( ~ 100 M  ) originated from the evolution of zero metallicity massive stars are able to explain SMBH, by growing through accretion and coalescences Simulated SMBH satisfy the mass distribution observed in the local universe as well as the evolution of the BH mass density, the M ● vs  and the M ● vs M halo relations Gravitational waves can put strong constraints on the coalescence history of seeds However, difficulties exist such: a) no SMBH are seen at z ~ 6 b) density of massive galaxies are overestimated c) simulated and Mg2 indices still do not reproduce adequately the observations d) blue galaxies do not have enough angular momentum