Hidden Black Hole Activity Hidden Black Hole Activity in the Early Universein the Early Universe

Ezequiel TreisterEzequiel Treister (Yale/U. de Chile)

Meg UrryMeg Urry (Yale)

The CYDER Team:

F. Castander, T. Maccarone, P. Coppi, E. Gawiser

The GOODS Team:

F. Bauer, D. Alexander, N. Brandt and others

Big

Bang

CM

B

Firs

t Sta

rsB

lack

Hol

es

Bla

ck H

oles

Toda

y

Qua

sars

Gal

axie

s &

time

z=0redshift z~2-3z~15

z~1000

Larson, Bromm, Coppi

Active Galactic Nucleus (AGN)Active Galactic Nucleus (AGN)

supermassive black hole

actively accreting matter

1-1000 x galaxy luminosity from few lt-hrs

Host Galaxy

AGN: signposts for SMBH in early Universe

AGN-Galaxy ConnectionAGN-Galaxy Connection

All (large) galaxies have supermassive black holes

Common AGN-Galaxy evolution 0<z<2

Tight correlation of MBH with AGN host galaxies are normal

AGN-Galaxy ConnectionAGN-Galaxy Connection

All (large) galaxies have supermassive All (large) galaxies have supermassive black holesblack holes

Common AGN-Galaxy evolution 0<z<2

Tight correlation of MBH with AGN host galaxies are normal

AGN-Galaxy ConnectionAGN-Galaxy Connection

All (large) galaxies have supermassive black holes

Common AGN-Galaxy evolution 0<Common AGN-Galaxy evolution 0<zz<2<2

Tight correlation of MBH with AGN host galaxies are normal

AGN-Galaxy ConnectionAGN-Galaxy Connection

All (large) galaxies have supermassive black holes

Common AGN-Galaxy evolution 0<z<2

Tight correlation of MTight correlation of MBHBH with with AGN host galaxies are normal

MMBHBH-- Relation Relation

Tremaine et al (2002)

Black Hole mass Black Hole mass is well correlated is well correlated with velocity with velocity dispersion, even dispersion, even though they are though they are dynamically dynamically independent.independent.

AGN-Galaxy ConnectionAGN-Galaxy Connection

All (large) galaxies have supermassive black holes

Common AGN-Galaxy evolution 0<z<2

Tight correlation of MBH with AGN host galaxies are normalAGN host galaxies are normal

O’Dowd et al. 2002

Kormendy Kormendy relation for relation for AGN same as AGN same as for normal for normal galaxiesgalaxies

(projection of fundamental plane)

AGN

Normal galaxies

Surface brightness versus sizeSurface brightness versus size

The AGN Unified ModelThe AGN Unified Model

blazars, Type 1 Sy/QSO

broad lines

Urry & Padovani, 1995

Type 1 AGN SEDType 1 AGN SED

Manners, 2002mm far-IR near-IR Optical-UV

X-rays

The AGN Unified ModelThe AGN Unified Model

radio galaxies, radio galaxies, Type 2 Sy/QSOType 2 Sy/QSO

narrow linesnarrow lines

Urry & Padovani, 1995

Type 2 AGN SEDType 2 AGN SED

Norman et al, 2002

Radio far-IR optical-UV X-rays

Narrow lines Narrow lines in total lightin total light

Broad lines in Broad lines in polarized lightpolarized light

Antonucci & Miller 1985, Barthel 1987

Evidence for UnificationEvidence for Unification

More Evidence: The XRBMore Evidence: The XRB

Gruber et al. 1999

unabsorbed AGN spectrum

Increasing NH

Population Synthesis ModelsPopulation Synthesis Models

Gilli et al. 1999Gilli et al. 1999Gilli et al. 2001Gilli et al. 2001

Giacconi et al. 1979Giacconi et al. 1979Setti & Woltjer 1989Setti & Woltjer 1989Madau et al. 1994Madau et al. 1994Comastri et al. 1995Comastri et al. 1995

Hasinger 2002

Gilli et al. population synthesis predictions

Hidden Population of Hidden Population of Obscured AGN at z>1Obscured AGN at z>1

Not Found in UV/Optical Surveys.

Multiwavelength Surveys needed:

Hard X-rays (Chandra) Far-IR (Spitzer) Optical Spectroscopy (Keck-VLT-Magellan)

Narrow Area/Deep SurveysNarrow Area/Deep Surveys Low luminosity sources Good to higher redshifts Less obscuration bias

Low numbers of high luminosity objects Large scale structure

Wide Area/Shallow SurveysWide Area/Shallow Surveys High luminosity sources Less “cosmic variance”

Low luminosity objects only at low redshift More obscuration bias

The CYDER SurveyThe CYDER Survey

CalánYaleDeepExtragalacticResearch

CYDER Survey CYDER Survey Optical and Near-IR Follow up of X-ray sources in Optical and Near-IR Follow up of X-ray sources in Deep Archived Chandra Fields.Deep Archived Chandra Fields.

FieldField RARA DecDec Chandra Chandra FieldField

Exp. Exp. TimeTime

C2C2 12:52:48 -09:24:30 HCG62 50 ks

C5C5 11:30:07 -14:49:27 Q1127 30 ks

D1D1 22:01:38 -31:46:30 HCG90 50 ks

D2D2 23:47:55 +00:57:21 Q2345 75 ks

D3D3 03:37:44 -05:02:39 SBS0335 60 ks

PopulationsPopulations

Object CYDER ChaMP CDF-N CDF-S

Unobs. AGN 47.2%47.2% 52.1% 12.4% 35.0%

Obs. AGN 35.8%35.8% 23.9% 39.4% 42.7%

Galaxies 10.4%10.4% 14.1% 43.3% 17.8%

Stars 6.6%6.6% 9.9% 4.9% 4.5%

Total 106106 125 284 168

V-magnitude DistributionV-magnitude Distribution

All X-ray Sources

Targeted for Spectroscopy

Identified Sources

Redshift DistributionRedshift Distribution

All X-ray Sources

AGN-Dominated Sources

Unobscured AGN

V-mag versus RedshiftV-mag versus Redshift

V-I versus RedshiftV-I versus Redshift

FFxx/F/Foptopt versus X-ray Luminosity versus X-ray Luminosity

FFXX/F/Foptopt Ratio Ratio

Fiore et al, 2003

Obscured to Total AGN RatioObscured to Total AGN RatioF

ract

ion

of O

bscu

red

AG

N Intrinsic

Hard X-ray Luminosity

Narrow Area/Deep SurveysNarrow Area/Deep Surveys Low luminosity sources Good to higher redshifts Less obscuration bias

Low numbers of high luminosity objects Large scale structure

Wide Area/Shallow SurveysWide Area/Shallow Surveys High luminosity sources Less “cosmic variance”

Low luminosity objects only at low redshift More obscuration bias

GOODSGOODSdesigned to find designed to find

obscured AGN obscured AGN at the quasar epoch, at the quasar epoch, z~2-3z~2-3

Chandra Deep Fields, Spitzer Legacy, HST Treasury (3.5+ Msec) (800 hrs) (600 hrs) (3.5+ Msec) (800 hrs) (600 hrs)

Very deep imagingVery deep imaging~70 times HDF area (0.1 deg~70 times HDF area (0.1 deg22))

Extensive follow-up spectroscopy (VLT, Gemini, …)Extensive follow-up spectroscopy (VLT, Gemini, …)

HST ACS fieldsHST ACS fields5 epochs/field, spaced by 45 days, simultaneous V,i,z bands + B band5 epochs/field, spaced by 45 days, simultaneous V,i,z bands + B band

CDF-S: Aug ‘02 - Feb ‘03 HDF-N: Nov ’02 - May ’03CDF-S: Aug ‘02 - Feb ‘03 HDF-N: Nov ’02 - May ’03

done 2003

ApJ Letters 2004, 600, …

B = 27.2

V = 27.5

i = 26.8

z = 26.7

B = 27.9

V = 28.2

I = 27.6

∆m ~ 0.7-0.8AB mag; S/N=10Diffuse source, 0.5” diameterAdd ~ 0.9 mag for stellar sources

ACS WFPC2

GOODS X-Ray SourcesGOODS X-Ray Sources

Chandra Deep Field North: Chandra Deep Field North:

Chandra Deep Field South: Chandra Deep Field South:

2 Ms2 Ms 503 sources503 sources 1.4x101.4x10-16 -16 ergs cmergs cm-2-2ss-1-1 (2-8 keV) (2-8 keV)

1 Ms1 Ms 326 sources326 sources 4.5x104.5x10-16 -16 ergs cmergs cm-2-2ss-1-1 (2-8 keV) (2-8 keV)

Modeling the AGN PopulationModeling the AGN Population• Grid of AGN spectra (LX,NH) with

– SDSS quasar spectrum (normalized to X-ray)– dust/gas absorption (optical/UV/soft X-ray) – infrared dust emission Nenkova et al. 2002, Elitzur et al. 2003– L* host galaxy

• Hard X-ray LF & evolution for Type 1 AGN Ueda et al. 2004

• Geometry with obscured AGN = 3 x unobscured, at all z, L

• Calculate expected redshift distribution – compare to measured redshifts of GOODS AGN

• Calculate expected optical magnitudes of X-ray sources in GOODS fields – compare to GOODS HST data

• Calculate expected N(S) for infrared sources – compare to GOODS Spitzer data

Modeling the AGN PopulationModeling the AGN Population• Grid of AGN spectra (LX,NH) with

– SDSS quasar spectrum (normalized to X-ray)– dust/gas absorption (optical/UV/soft X-ray) – infrared dust emission Nenkova et al. 2002, Elitzur et al. 2003– L* host galaxy

• Hard X-ray LF & evolution for Type 1 AGN Ueda et al. 2004

• Geometry with obscured AGN = 3 x unobscured, at all z, L

• Calculate expected redshift distribution – compare to measured redshifts of GOODS AGN

• Calculate expected optical magnitudes of X-ray sources in GOODS fields – compare to GOODS HST data

• Calculate expected N(S) for infrared sources – compare to GOODS Spitzer data

Treister et al. 2004

Treister et al. 2004

Modeling the AGN PopulationModeling the AGN Population• Grid of AGN spectra (LX,NH) with

– SDSS quasar spectrum (normalized to X-ray)– dust/gas absorption (optical/UV/soft X-ray) – infrared dust emission Nenkova et al. 2002, Elitzur et al. 2003– L* host galaxy

• Hard X-ray LF & evolution for Type 1 AGN Ueda et al. 2004

• Geometry with obscured AGN = 3 x unobscured, at all z, L

• Calculate expected redshift distribution – compare to measured redshifts of GOODS AGN

• Calculate expected optical magnitudes of X-ray sources in GOODS fields – compare to GOODS HST data

• Calculate expected N(S) for infrared sources – compare to GOODS Spitzer data

AGN Number Counts CalculationAGN Number Counts CalculationX-Ray Luminosity FunctionX-Ray Luminosity Function

Ueda et al, 2003

Modeling the AGN PopulationModeling the AGN Population• Grid of AGN spectra (LX,NH) with

– SDSS quasar spectrum (normalized to X-ray)– dust/gas absorption (optical/UV/soft X-ray) – infrared dust emission Nenkova et al. 2002, Elitzur et al. 2003– L* host galaxy

• Hard X-ray LF & evolution for Type 1 AGN Ueda et al. 2004

• Geometry with obscured AGN = 3 x unobscured, at all z, L

• Calculate expected redshift distribution – compare to measured redshifts of GOODS AGN

• Calculate expected optical magnitudes of X-ray sources in GOODS fields – compare to GOODS HST data

• Calculate expected N(S) for infrared sources – compare to GOODS Spitzer data

Dust emission models from Nenkova et al. 2002, Elitzur et al. 2003

Simplest dust distribution that satisfies

NH = 1020 – 1024 cm-2

3:1 ratio (divide at 1022 cm-2)Random angles NH distribution

Modeling the AGN PopulationModeling the AGN Population• Grid of AGN spectra (LX,NH) with

– SDSS quasar spectrum (normalized to X-ray)– dust/gas absorption (optical/UV/soft X-ray) – infrared dust emission Nenkova et al. 2002, Elitzur et al. 2003– L* host galaxy

• Hard X-ray LF & evolution for Type 1 AGN Ueda et al. 2004

• Geometry with obscured AGN = 3 x unobscured, at all z, L

• Calculate expected redshift distribution – compare to measured redshifts of GOODS AGN

• Calculate expected optical magnitudes of X-ray sources in GOODS fields – compare to GOODS HST data

• Calculate expected N(S) for infrared sources – compare to GOODS Spitzer data

redshifts of Chandra deep X-ray sources

GOODS-N

Barger et al. 2002,3, Hasinger et al. 2002, Szokoly et al. 2004

model

redshifts of Chandra deep X-ray sources

GOODS-N

Barger et al. 2002,3, Hasinger et al. 2002, Szokoly et al. 2004

R<24

GOODS N+S DistributionGOODS N+S Distribution

Treister et al, 2004

CYDER X-ray/optical survey

Treister, et al 2004

SEXSI X-ray/optical survey

Harrison, Helfand, et al.

Modeling the AGN PopulationModeling the AGN Population• Grid of AGN spectra (LX,NH) with

– SDSS quasar spectrum (normalized to X-ray)– dust/gas absorption (optical/UV/soft X-ray) – infrared dust emission Nenkova et al. 2002, Elitzur et al. 2003– L* host galaxy

• Hard X-ray LF & evolution for Type 1 AGN Ueda et al. 2004

• Geometry with obscured AGN = 3 x unobscured, at all z, L

• Calculate expected redshift distribution – compare to measured redshifts of GOODS AGN

• Calculate expected optical magnitudes of X-ray sources in GOODS fields – compare to GOODS HST data

• Calculate expected N(S) for infrared sources – compare to GOODS Spitzer data

Treister et al. 2004

brightfaint

GOODS N+S

Modeling the AGN PopulationModeling the AGN Population• Grid of AGN spectra (LX,NH) with

– SDSS quasar spectrum (normalized to X-ray)– dust/gas absorption (optical/UV/soft X-ray) – infrared dust emission Nenkova et al. 2002, Elitzur et al. 2003– L* host galaxy

• Hard X-ray LF & evolution for Type 1 AGN Ueda et al. 2004

• Geometry with obscured AGN = 3 x unobscured, at all z, L

• Calculate expected redshift distribution – compare to measured redshifts of GOODS AGN

• Calculate expected optical magnitudes of X-ray sources in GOODS fields – compare to GOODS HST data

• Calculate expected N(S) for infrared sources – compare to GOODS Spitzer data

Treister et al, 2004

Spitzer PredictionsSpitzer Predictions

Predicted Spitzer counts (3.6 Predicted Spitzer counts (3.6 m)m)

~1/2 AGN missing from deep Chandra samples

Will be detected with Spitzer IRAC

Total # infrared sources in GOODS: prediction=observed

Observed distribution of infrared fluxes matches prediction.

Spitzer supports model of obscured AGNSpitzer supports model of obscured AGN

Treister, Urry, van Duyne, et al. 2004

3.6 m flux (mJy)

N(>

S)

Observed IRAC fluxes of CDF X-ray sources

ConclusionsConclusions• Obscured AGN are host galaxy dominated in

optical light• Simple unification model explains:

– (faint) optical magnitude distribution– redshift distribution

• Previously reported dependence of obscured AGN ratio with luminosity can be a selection effect.

• This model is consistent with predictions by XRB population synthesis models:

- Broader redshift distribution with peak at z~1.3 - Type 2/Type 1~3

Treister et al.

X-ray background synthesis

X-ray background synthesis

Treister et al.

Treister et al.