Build-up of black holes 

Paolo Padovani 
-----------------------------------------------

Census of black holes as a function of redshift. 

Scientific Justification 

The determination of the time-history of accretion is crucial to our
understanding of how supermassive black holes form and evolve. The first
step towards this goal is to identify them. However, it is now well
established that much of the accretion power in the universe is absorbed,
making it impossible to measure at optical wavelengths. The Chandra and
XMM-Newton observatories have revolutionized this field by making it
possible to map the history of the Active Galactic Nuclei (AGN) population
using hard (2 - 10 keV) X-ray surveys. This is because AGN are ideal
tracers of supermassive black holes, as their features make them easily
visible and recognizable and the hard X-rays can directly probe AGN
activity and are uncontaminated by star formation processes at the X-ray
luminosities of interest. Compton-thick sources, that is objects with
column density in excess of $1.5 \times 10^{24}$ cm$^{-2}$, will however
still be missed by hard X-ray surveys. The recently launched Spitzer
satellite, with its near- and far-infrared coverage, provides an even more
valuable probe of AGN demographics. The obscuring dust which hides AGN from
ultraviolet, optical, and soft X-ray surveys should be a largely isotropic
emitter at wavelengths $\ga 30\mu$. Furthermore, even Compton-thick sources
will have to re-emit the photons absorbed at shorter wavelengths in the
far-infrared.

The main aims of this project are the following:

1. expand the work already done in the X-ray band on small areas of the sky
(e.g., the GOODS fields) to larger areas to improve the statistics; include
sources with enough X-ray counts ($\ga 100$) to derive an X-ray spectrum,
which is vital to estimate the absorbing column (most previous works had to
rely on hardness ratios). This will also allow us to recognize
high-redshift absorbed sources, which can be mistakenly classified as soft
sources by using hardness ratios (e.g., at $z = 3$ the rest-frame $2-8$ keV
band shifts to $0.5-2$ keV). The XMM serendipitous catalogue will combine
a large area with the depth and counts we need;

2. utilize the infrared Spitzer data to complete the black hole census and
include also Compton-thick sources; quantify the fraction of such sources
so that Chandra and XMM surveys can be properly "corrected" for missing
them. 


Required Data

- XMM serendipitous catalogue
- Spitzer Legacy data: IRAC at 3.6, 4.5, 5.8 and 8.0 $\mu$, MIPS at 24, 70
  and possibly 160 $\mu$
- ancillary data in various bands: X-ray (XMM/Chandra), UV (Galex), optical
  (ground based), 

Strategy

1.
- estimate photometric redshifts for X-ray sources when no spectra are
  available
- correlate the X-ray catalogue with optical/near-IR catalogues and 
  catalogues in other bands; build a Spectral Energy Distribution (SED) to
  classify the source using all available data
- estimate the X-ray power by alternative means for the really faint
  sources, e.g., via the Fiore et al. (2003) relationship between
  X-ray-to-R-band flux ratio and X-ray power; calibrate first this
  relationship at faint magnitudes, as there might be a magnitude
  dependency
- derive absorbing columns and spectral indices from X-ray spectra to
  estimate the relative number of absorbed and non-absorbed sources. 

2. 
- select AGN using mid-IR (IRAC) colours (see Stern et al. 2004)
- improve on selection by using far-IR (MIPS) data
- use available spectra and multi-wavelength information to classify sources
- estimate photometric redshifts when no spectra are available
- examine AGN undetected in the X-ray band to look for Compton-thick sources
 

Why is the VO approach unique 

VO tools will allow to carry out the project in a much faster and reliable
way. Required tools include: 

- multi-wavelength tool to be able to cross-correlate and overlay images in
  different bands taking into account the different Point Spread Function
  (PSF)
- generation of photometric redshifts from catalogues/images by running 
  appropriate codes
- colour-colour tools to facilitate object selection
- data-mining tools to look for outliers
- classifier tools based on the properties of previously known sources 
- SED builder and comparison with synthetic spectral libraries
- availability of relevant models, e.g. X-ray spectrum for various 
  absorbing columns. 


*****YOUR PART******* 

- high-z QSO

  * sky area, Lyman-break ("SED")
  * XMM serendipitous survey and FIRST/SUMSS