The Cosmic X-ray Background Facts: a Nobel prize Fictions: The AGN synthesis models (?)

1
The Cosmic X-ray Background Facts a Nobel prize
Fictions The AGN synthesis models (?)

• Andrea Comastri
• (INAF- Oss. Astr. Bologna)

2
Extragalactic Background Light
3
The XRB spectrum data
(courtesy of R. Gilli)
4
The 2-10 keV XRB intensity
(courtesy of R. Gilli)
Vecchi et al. (1999)
Lumb et al. (2002)
Kushino et al. (2002)
Tozzi et al. (2001) Integrated flux from
resolved sources
5
The origin of the cosmic XRB
Diffuse gas can contribute significantly to the
cosmic XRB only below 2 keV (CMB argument, Wright
et al. 1994 lt10 Rosati et al. 2002) ? the
hard XRB is due to single sources
OBSCURED AGN
6
XRB model formulae
XRB spectrum
Source spectrum
X ray luminosity function (XLF)
Source counts
7
X-ray luminosity function
Luminosity Dependent Density Evolution (LDDE) Sof
t X-rays Miyaji et al. 2000 up to z1.5
evolution rate (1z)5 Hard X-rays La Franca et
al. (2002)
Miyaji et al.(2000)
8
Fit to the XRB spectrum
9
SOURCE COUNTS
10
Some Facts

• 60-70 of the XRB is made by sources around the
  knee of source counts (logfx-14-1)
• The 2-10 keV spectrum is a flat power law (0.4 ?
  HEAO1-ASCA-BeppoSAX-XMM,)
• Absorption distribution logNH21-24
• Luminosity distribution logLx 44-1.5

11
Several Fictions

• Compton thick (log NH gt 24) sources
• Optical continuum and emission line properties of
  hard X-ray sources
• Host galaxy properties

12
Main model assumptions

• The X-ray Luminosity Function and evolution of
  obscured AGN is unknown ? TypeIITypeI
  
• The absorption distribution of obscured AGN is
  unknown (but the nearby Universe)
• ? XRB fit
• Simple SED (power law absorption log NH 21-25)

13
Observational constraints N(NH), N(z)
Risaliti et al. (1999)
Observed in Seyfert 2s
Comastri et al. (1995)
Fitted to the XRB spectrum
Sy2/Sy1 ratio at z0 3-5 (Maiolino Rieke
1995)
14
Observational constraints need for QSO2
High Luminosity (logL(X)gt 44) highly
absorbed (log(NH)23/-1) sources are required to
fit bright counts Compton Thick QSO (logNHgt24)
are not energetically important
15
Hard counts (with/without Type 2 QSO)
16
Hard Counts (Compton thick dominated)
17
The deepest X-ray sky
CDFS
HDFN
18
The Deep X-ray Surveys
Survey X-ray observations Sensitivity/1e-
16 N. Sources Reference soft /
hard CDFS 1Msec Chandra ACIS-I
0.5 / 4 346 Giacconi et
al.(2002) 500 ksec XMM CDFN
2 Msec Chandra ACIS-I 0.2 / 1 503
Alexander et al. (2003) 230 ksec
XMM Lockman 100 ksec XMM 3.0 / 14
200 Hasinger et al.
(2001) 300 ksec Chandra HRC
1.5 Msec ROSAT 260 ksec ASCA Lynx
180 ksec Chandra 1.7 / 15
153 Stern et al. (2002)
300 ksec XMM SSA13 100 ksec Chandra
2.3 / 25 40
Mushotzky et al. (2000)

extended to 0.2 / 2 by Campana et al. (2001)
see also Moretti et al. (2002)
19
Problems

1. Redshift distribution
2. Absorption distribution
3. Quasar 2

20
Redshift distribution in the Deep Surveys
CDFS HDFN (Hasinger et al. 2003)
21
Gandhi Fabian 2003 Franceschini et al. 2002
link with IR?
Best fit to the peak, deficit at zgt1 most
likely due to spectroscopic incompleteness
BUT
22
Abs/unabs ratio as a function of z
CDFN
CDFSCDFN
CDFS
23
Luminosity function

• Cowie et al. 2003
• astro-ph/0301231

• Hasinger 2003
• astro-ph/0302574

24
Space density

• Cowie et al. 2003
• rho_BH 2x105

• Hasinger 2003

25
X/O -------- fX/fopt gt 10 Likely to be highly
obscured Undetected in the R-band at R24-25
(shallow), even Rgt27-28 (deep) Constant
fraction over a large range of fluxes Shallow
surveys pick-up brightest sources ? optical
identification is possible
26
317 sources (Hellas2XMM Lockman CDFN SSA13
-15 lt logFx lt -13.3 , 70 identified)
Fiore et al. 2003 astro-ph/0306556
27
Redshift distribution (Opt stat. ids)
28
The evolution of number and luminosity densities
(Fiore et al. 2003, astro-ph/030556)

• rho_BH 4-6x105

29
Average spectrum
Tozzi et al. 2001
30
2a) Absorption distributions2-10 keV
Chandra CDFN XMM Hellas2Xmm
HEAO-1 Piccinotti et al. ASCA Della Ceca et
al. XMM Piconcelli et al. XMM Mainieri et al.
31
2b) Absorption distribution5-10 keV
NHgt1022
NHgt1023
open BeppoSAX HELLAS filled ASCA Sheep
32
Absorption vs luminosity
CDFS
XMM Lockman Hole
From Mainieri et al. (2002)
33
3) Can we survive without QSO2 ?

• QSO2 (luminous obscured sources) do exist, but
  they are NOT luminous Seyfert 2.
• EROs, high fx/fopt sources , broad lines quasars,
  ULIRGs
• X-ray and optical properties do not go hand in
  hand
• Are they enough ? (see discussion)

34
CDFS 202 QSO-2 detected
LX 1045
erg/s
NH 1024cm-2 narrow high-excitation lines
Fe-line _at_ 6.4 keV
VLT-spectrum Chandra spectrum
1.4 keV 6.4/(1z)
Norman et al. 2002
35

• N(Z) predicted to peak around 1.5-2
• ?Now it seems to peak at 0.7-1
• Change the XLF parameters rather than assuming
  extremely fast evolution
• N(NH) ? HR vs. real spectra
• N(NH) ? lack of obscured sources at bright fluxes
  (30 obs vs. 50 expect)
• Dependence of NH vs. z and/or luminosity ?
• N(NH)sharp drop of ltalphagt ? things are rapidly
  changing around logfx-14 ?

36
Conclusions ?

• Forget the unification scheme ?
• Why the bull stop to moo at high luminosity ?
• Space density and evolution of obscured sources ?
• High energy spectrum ?