Black Holes in the Early Universe Accretion and Feedback

Transcription

1 1 1

2 Black Holes in the Early Universe Accretion and Feedback 1 1

3 Black Holes in the Early Universe Accretion and Feedback Geoff Bicknell & Alex Wagner Australian National University 1 1

4 High redshift radio galaxies: AGN Feedback z=2.42 radio galaxy MRC Nesvadba et al

5 Feedback from radio galaxies at z ~1-3 requires black holes ~ 10 9 solar masses What is the origin of these black holes? 3 3

6 Black holes in the early universe Fan 2006: The discovery of luminous quasars in SDSS at z > 6 indicates the existence billion-solar-mass black holes at the end of reionization epoch. Comoving spatial density of quasars at M1450 < Fan 2006, New Astr. Rev. 50,

7 Black hole masses Distribution of black hole masses for z>3 From Fan, 2006 Black hole masses up to ~ solar masses 5 5

8 Dust-free quasars further evidence of evolution Jiang, Fan, Brandt et al. Nature 2010 Disk emission Dust emission Low z quasars 6 6

9 Correlation between dust and black hole mass Correlation between measure of dust and black hole mass Formation of dust in quasar outflows? (Elvis et al. 2007) 7 7

10 Alternative to quasar outflow model Gall, Anderson and Hjorth, 2011 Rapid dust formation possible when SFR > 10 3 solar masses per year Models require top heavy IMF Input from SNe not AGB stars Does not rule out quasar outflow model 8 8

11 Growth by accretion (Shapiro ApJ 2005) Radiative efficiency: r = Normalised luminosity: ṁ = Accretion rate: dm dt L Ṁ 0 c 2 L L Edd = (1 r )Ṁ0 = ṁ(1 r) r M τ Accretion time scale: τ = cσ T 4πGm p = yrs 9 9

12 Radiative efficiency Comparison of quasar luminosity density with SMBH density at z<5 implies radiative efficiency εr > 0.1 (Soltan 1982; Aller & Richstone 2002; Elvis, Risaliti & Zamorani 2002; Yu & Tremaine 2002; Marconi 2004)

13 Time to grow a black hole by accretion t t 0 = yrs r 1 r 1 ṁ ln M M 0 r =0.1 ṁ =1 M = M M 0 = 100M yrs M = M M 0 = 100M yrs Time available z =6 t = yr z =3 t = yr 11 11

14 Effect of black hole spin: Amplification at z=6.43 Black hole spin increases radiative efficiency up to 0.42 Black hole amplification Amplification required from accretion alone 100<M0<600 Efficiency Amplification required assuming mergers account for growth ~10 4 in black hole mass Initial redshift 12 12

15 Enhancement of accretion by jets/winds (Jolley & Kuncic 2008) z R ms r Gravitational power directed into wind or jet decreases the radiative luminosity => Accretion rate larger for given luminosity 13 13

16 Modification of black hole growth Accretion efficiency = a = Radiative + Jet efficiency = r + j dm dt = ṁ(1 a) r M t t 0 = yrs τ a 1 a (1 j / a ) 1 ṁ ln M M 0 Effect of jet/wind is to reduce accretion time for a given luminosity 14 14

17 Black hole mass at z=6 M(z = 6)/M0 r > 0.1 j / a 15 15

18 Implications for AGN feedback Kinetic luminosity of jet/wind L wind = ergs s 1 w w /0.42 a r /0.1 ṁ M 10 9 M Powerful jets/winds such as this relevant for AGN feedback In this case winds may be more likely since, even at high z, most quasars are radio quiet

19 So far... Straightforward black hole growth by accretion difficult Driving black hole growth by Poynting-flux dominated jets/winds assists the formation of supermassive black holes by z~3, but still involves accretion at the Eddington limit Winds have other benefits Feedback in early epochs Early creation of dust 17 17

20 Next... Radio-Mode Feedback 18 18

21 The violent universe 19 19

22 The violent universe... We see gas being churned by explosions and huge black holes in the center of the cluster. We see how it's cooling down and how the cooling is being balanced by tremendous outbursts of jets and bubbles of hot gas

23 The violent universe... We see gas being churned by explosions and huge black holes in the center of the cluster. We see how it's cooling down and how the cooling is being balanced by tremendous outbursts of jets and bubbles of hot gas... Martin Rees quoted in review by McNamara & Nulsen Heating hot atmospheres by active galactic nuclei 19 19

24 Issues in Galaxy Formation (see Croton et al. 06) 20 20

25 Issues in Galaxy Formation (see Croton et al. 06) Hierarchical merging predicts more high mass galaxies than are observed (exponential cutoff in Schecter luminosity function) 20 20

26 Issues in Galaxy Formation (see Croton et al. 06) Hierarchical merging predicts more high mass galaxies than are observed (exponential cutoff in Schecter luminosity function) Requires feedback in addition to that provided by supernovae => Regulation of star formation 20 20

27 Issues in Galaxy Formation (see Croton et al. 06) Hierarchical merging predicts more high mass galaxies than are observed (exponential cutoff in Schecter luminosity function) Requires feedback in addition to that provided by supernovae => Regulation of star formation Downsizing: Star formation and AGN activity takes place more vigorously and in higher mass objects at z ~ 1-2. Thereafter there is a downsizing in the amount of activity that takes place

28 Galaxy downsizing semi-analytic models Croton et al. 2006: Effect of radiomode feedback on galaxy formation Feedback produces an exponential cutoff in luminosity distribution at bright end 21 21

29 Galaxy downsizing semi-analytic models Croton et al. 2006: Effect of radiomode feedback on galaxy formation No feedback Feedback produces an exponential cutoff in luminosity distribution at bright end 21 21

30 Galaxy downsizing semi-analytic models Croton et al. 2006: Effect of radiomode feedback on galaxy formation No feedback Feedback produces an exponential cutoff in luminosity distribution at bright end Feedback 21 21

31 Galaxy downsizing semi-analytic models Croton et al. 2006: Effect of radiomode feedback on galaxy formation No feedback Feedback produces an exponential cutoff in luminosity distribution at bright end Feedback Accretion rate orders of magnitude below Eddington => Low-powered radio galaxies providing the feedback 21 21

32 SPH simulations Booth & Schaye 2009 See also Schaye et al

33 SPH simulations Booth & Schaye 2009 Sub-grid prescriptions for effect of black hole on surrounding ISM See also Schaye et al

34 SPH simulations Booth & Schaye 2009 Sub-grid prescriptions for effect of black hole on surrounding ISM Accretion rate ~ 100 x Bondi rate See also Schaye et al

35 SPH simulations Booth & Schaye 2009 Sub-grid prescriptions for effect of black hole on surrounding ISM Accretion rate ~ 100 x Bondi rate Resolution ~ 2 kpc See also Schaye et al

36 SPH simulations Booth & Schaye 2009 Sub-grid prescriptions for effect of black hole on surrounding ISM Accretion rate ~ 100 x Bondi rate The secret life of an SPH particle Resolution ~ 2 kpc See also Schaye et al

37 Feedback in action: GPS and CSS sources 3C303.1 O Dea et al kpc Text z=0.267 CSS Radio Galaxy Evidence for shocks and star formation 23 23

38 High redshift radio galaxies: MRC z=2.42 radio galaxy MRC Nesvadba et al

39 Simulations of jet-ism interactions (Wagner & GB, ApJ Feb 2011) log(density) 25 25

40 Simulations of jet-ism interactions (Wagner & GB, ApJ Feb 2011) log(density) P jet = ergs s 1 Γ =

41 Simulations of jet-ism interactions (Wagner & GB, ApJ Feb 2011) log(density) P jet = ergs s 1 Γ = 10 p ISM k = 10 6, 10 7 T ISM = 10 7 n ISM = 0.1, 1.0 cm

42 Simulations of jet-ism interactions (Wagner & GB, ApJ Feb 2011) log(density) P jet = ergs s 1 Γ = 10 p ISM k = 10 6, 10 7 T ISM = 10 7 n ISM = 0.1, 1.0 cm 3 Fractal, truncated, log-normal distribution of clouds 25 25

43 Simulations of jet-ism interactions (Wagner & GB, ApJ Feb 2011) log(density) P jet = ergs s 1 Γ = 10 p ISM k = 10 6, 10 7 T ISM = 10 7 n ISM = 0.1, 1.0 cm 3 Fractal, truncated, log-normal distribution of clouds n clouds = cm

44 Standard jet 26 26

45 Standard jet 26 26

46 P jet = ergs s 1 p/k = 10 7 n clouds = 10 3 cm 3 f V =

47 P jet = ergs s 1 p/k = 10 7 n clouds = 10 3 cm 3 f V =

48 Comparison with MRC

49 P jet = ergs s 1 p/k = 10 6 n clouds = 10 2 cm 3 f V =

50 P jet = ergs s 1 p/k = 10 6 n clouds = 10 2 cm 3 f V =

51 log[vw (km/s)] 30 30

52 log[vw (km/s)] 30 30

53 Criterion for inhibition of star formation Parametrize jet in terms of Eddington luminosity: η = P jet L Edd Mean radial velocity of clouds exceeds velocity dispersion v cloud > σ 100 η 1/4 P 1/4 jet,45 km s

54 Effectiveness of jet feedback 32 32

55 Effectiveness of jet feedback σ 100 =1.0 η 1/4 P 1/4 jet,

56 Effectiveness of jet feedback σ 100 =1.0 η 1/4 P 1/4 jet,

57 Effectiveness of jet feedback p/k = σ 100 =1.0 η 1/4 P 1/4 jet,

58 Effectiveness of jet feedback p/k = p/k = σ 100 =1.0 η 1/4 P 1/4 jet,

59 Effectiveness of jet feedback p/k = 10 6 Low filling factor -5.0 p/k = σ 100 =1.0 η 1/4 P 1/4 jet,

60 Eddington factor velocity dispersion Parametrize jet power in terms of Eddington luminosity 33 33

61 Eddington factor velocity dispersion Parametrize jet power in terms of Eddington luminosity P jet = η M bh M ergs s

62 Eddington factor velocity dispersion Parametrize jet power in terms of Eddington luminosity Eddington factor P jet = η M bh M ergs s

63 Eddington factor velocity dispersion Parametrize jet power in terms of Eddington luminosity Eddington factor P jet = η M bh M ergs s 1 M-σ relation M bh M σ 100 km s

64 Eddington factor velocity dispersion Parametrize jet power in terms of Eddington luminosity Eddington factor P jet = η M bh M ergs s 1 M-σ relation M bh M σ 100 km s 1 4 σ 100 km s η 1/4 P 1/4 jet,

65 Recent low filling factor simulation D continues this sequence to low filling factor 34 34

66 Lower filling factor f V =0.027 f V =

67 Revised speed power diagram 36 36

68 Jet-Disk interaction Density 37 37

69 Radio and X-ray surface brightness 38 38

70 Radio and X-ray surface brightness 38 38

71 Radio and X-ray surface brightness 38 38

72 Simulations and observations GPS/CSO 4C31.04 Sutherland & GB 2007 Image courtesy of NRAO/AUI and Gabriele Giovannini, et al. 10 mas = 11.4 pc v 0.4c Reconciles difference between dynamic and spectral ages 39 39

73 Simulations and observations GPS/CSO 4C31.04 Sutherland & GB 2007 Image courtesy of NRAO/AUI and Gabriele Giovannini, et al. 10 mas = 11.4 pc v 0.4c Reconciles difference between dynamic and spectral ages 39 39

74 Simulations and observations GPS/CSO 4C31.04 Sutherland & GB 2007 Image courtesy of NRAO/AUI and Gabriele Giovannini, et al. 10 mas = 11.4 pc v 0.4c Reconciles difference between dynamic and spectral ages 39 39

75 Supporting evidence for Jet-ISM interactions from observations of ellipticals Kormendy et al., 2009, find that ellipticals with < MV < have extra light indicative of starbursts in wet mergers. For MV < no evidence of recent star formation AGN more effective in providing feedback in bright ellipticals Kormendy et al. interpreted in terms of high p/k of X-ray emitting ISM => more obstructive working surface for jet outflow Jet clumpy ISM interaction provides a more natural explanation 40 40

76 Main points Jets with Eddington factor > may disperse gas in the core of an evolving galaxy but porosity increases the critical Eddington factor Jets in a clumpy medium process all of the ISM Jets of all powers in excess of ergs s -1 could play a role Large fraction of the radio galaxy population involved 41 41

77 Increasing influence of radio galaxies at high redshift in view of the evolving radio luminosity function (Sadler et al. 2007) Important to consider the radio morphology of radio galaxies when assessing the role of AGN in influencing the evolution of the hosts 42 42