The Links Between AGN and Galaxy Formation

Omar AlmainiUniversity of Nottingham

The Links Between AGN and Galaxy Formation

The Links Between AGN and Galaxy Formation

Galaxy formation models …and their problems The AGN/galaxy connection Feedback and AGN winds Observing black hole/bulge growth Summary & future prospects

CDM cosmological model in excellent agreement with wide range of observations:

e.g. CMB, galaxy clustering, type 1a SN, element abundances, Cepheid distance scale, stellar ages, baryon fraction in clusters…

Merger Tree for the Growth of a Dark-matter Halo

Semi-analytic Galaxy Formation Models

+Messy physics

(gas cooling, star-formation, dust, SN feedback etc…)

N-body merger trees

=

Hierarchical Galaxy Formation

Benson (2003)

Producing the local K-band luminosity function -SN feedback is essential!

Benson (2003)

Problem can be solved with extreme super-winds >5x1049 erg per solar mass required

Kauffmann et al. 2003; Balogh et al. 2004

Local galaxies are strongly bimodal

e.g. Daddi et al. 2002; Roche, Almaini et al. (2002), Roche, Dunlop & Almaini (2003); Somerville et al. (2003)

Data Model

Can’t make enough red galaxies at z>1

The diversity of ERO galaxies

30-40% old, passive systems30-40% old, passive systems 30-40% dusty starburst30-40% dusty starburst Typical redshifts 1.0<z<1.5Typical redshifts 1.0<z<1.5 Strongly clustered Strongly clustered High space densityHigh space density

Many old, massive systems Many old, massive systems

already in place at z~1-1.5already in place at z~1-1.5

Recent discovery of many old, passive systems at z>1.5

Using J-K>2 colour selection find population of very red galaxies at z>2

Van Dokkum et al. 2003

‘EROs’ at higher redshift

Cimatti et al. 2004

Glazebrook et al. 2004

Summary of galaxy formation problems

I. Predicts too many dwarf galsII. Predicts “cuspy” coresIII. Why are there no galaxies with > 300 km/s?

Classic Problems

I. Fail to produce EROsII. Fails to predict SCUBA sourcesIII. Bimodal local galaxy pop.

Newer Problems

Need a mechanism for switching off star-formation in most massive galaxies

II. The AGN/galaxy connection

All massive galaxies contain a SMBH

The black-hole/bulge mass relationship.

Gebhardt et al. (2000), Ferrarese & Merritt (2000)c.f. Magorrian (1998)

MBH = 0.2 % Mbulge

MBH = 1.5x108 2004

M

Can the black hole influence its host galaxy?

MBH = 2x10-3 Mbulge

AGN radiative energy EAGN ~ 0.1 MBH c2 = 2 x 10-4 Mbulge c2

Binding energy of the bulge Ebulge ~ Mbulge 2 ~ 10-6 Mbulge 2

300 c2

Less than 1% of the AGN energy could blow the galaxy apart

The X-ray background: Most accretion activity is absorbed

Unobscured AGN

“AGN winds are ubiquitous” - Brad Peterson

“My wind model explains everything” - Martin Elvis

APM-08279 – lensed BAL quasar at z=3.91

Fe XXV K ? v1~0.2 cv2~0.4c

Chartas et al. (2002)

Massive X-ray outflow in PDS 456

Reeves et al. (2003)

XMM EPIC pn/MOS

Massive X-ray outflow in PDS 456

Reeves et al. (2003)

XMM RGS-2

V~50,000 kms-1 NH = 5x1023

~ 10 M yr-1

Momentum outflow:

Mwind v ~ L Edd/c.

King & Pounds 2003

Investigated supercritical accretion

M > M Edd

.

~

Assuming all excess matter ejected in a wind:

1. Compton-thick, quasi-spherical outflow

2. Momentum outflow: Mwind v ~ f L Edd/c.

.

King 2003(c.f. Silk & Rees 1998, Fabian 1999,

Blandford 1999, Di Matteo et al. 2004… )

Mwind v ~ L Edd/c.

Momentum outflow:

Outflow drives wind bubble through host galaxy

Shell velocity v2 Ledd / 2

M(R) 2 R

MBH 4

Mgas(R) 2 R

The black-hole/bulge mass relationship.

Gebhardt et al. (2000), Ferrarese & Merritt (2000)c.f. Magorrian (1998)

MBH = 0.2 % Mbulge

MBH = 1.5x108 2004

M

Observations of AGN-drivensuper- winds now required!

A cartoon model of massivegalaxy formation

SCUBA phase quasar elliptical?

II. AGN in galaxy clusters

The Entropy Problem(SN not enough?)

XMM RGS (Virgo)Sakelliou et al. 2002

The Cooling-Flow Problem

Chandra

VLA

II. Observing joint black-hole/galaxy formation

100100

Wavelength / m

f

101010001000

What is so special about submm sources?

The James Clerk Maxwell Telescope

(Hawaii)

• Highly luminous (ULIRG) systemsHighly luminous (ULIRG) systems• SFR ~ 1000 MSFR ~ 1000 M yr yr-1 -1

• Massive systemsMassive systems• Evidence for outflowing windsEvidence for outflowing winds

Progenitors of massive

elliptical galaxies?

Results of submm surveys

Mass determinations for submm sources

Genzel et al. (2003)

Massive (CO) Massive (H)

Swinbank et al. (2004)

Semi-analytic models have great difficultyproducing submm sources…

Baugh et al. (2005)

Most SCUBA sources DO NOT contain powerful quasars

Almaini et al. 2003

5 "

5 "

William Herschel Telescope 4m Depth R~27 Seeing 0.7”

HST ACS Imaging – (Almaini et al. 2005)

5 "

“True” color image0.5-2.0 keV 2.0-4.0 keV 4.0-8.0 keV

1.945 MsACIS-I

exposure

P.I.: W.N. Brandt

“True” color image0.5-2.0 keV 2.0-4.0 keV 4.0-8.0 keV

1.945 MsACIS-I

exposure

P.I.: W.N. Brandt

17/20 SCUBA galaxies

X-ray detected

Alexander et al. 2005

Joint spectral fitting

12Ms Chandra spectra!

Composite 2-20 keV spectra (12Ms)

Relatively low Lx/LFIR ratios

“True” color image0.5-2.0 keV 2.0-4.0 keV 4.0-8.0 keV

1.945 MsACIS-I

exposure

P.I.: W.N. Brandt

Alexander et al. 2005

Modest X-ray luminosities

Lx = 1043-1044 ergs-1

Suggest black holes <108MM

Evidence for growing black holes?

Archibald et al. (2002)

Fuel supply

Eddington limited growth

Where do we go from here?Where do we go from here?

Need detailed prescription for AGN feedback in semi-analytic models. Can we detect merging black holes with LISA (Madau 2004)? How prevalent are AGN-driven winds? How massive were first black holes? What influence did they have on the IGM at z>10?

Conclusions Overwhelming evidence for CDM hierarchical structure formation

Problems with semi-analytical galaxy formation models - mechanism required to terminate SF in massive gals - plus other problems…

AGN feedback is a likely solution - may be related to the origin of the M/ relation - could also explain high-mass cut-off & cluster heating problem