I. Georgantopoulos NATIONAL OBSERVATORY OF ATHENS

A. Georgakakis, O. Giannakis, S. Kitsionas, A. Akylas, D. Gaga, M. Plionis, V. Kolokotronis, S. Basilakos

G.C. Stewart, M.G. Watson University of Leicester

E. Hatziminaoglou European Southern Observatory

T. Shanks, M. Vallbe University of Durham

B.J. Boyle AAO

Overview

XMM-Newton/2df survey NORTH 9 adjacent fields, 5-10 ksec exposure, ~2 deg2

200 sources in 0.5-8 keV in areas covered by both the 2-df and SDSS surveys

The survey main goals are :

a. Nature of various classes of AGN b. groups and poor clusters of galaxies (Basilakos et al. 2003 MNRAS submitted)

c. Galaxies (Georgakakis et al. 2003 MNRAS astro-ph/)

Motivation (why bother ?)

1. Explore the bright X-ray sky especially at hard energies where the ASCA and BeppoSAX surveys lack spatial resolution

Bridge the gap between the Chandra surveys and bright fluxes

2. Further explore the soft X-ray sky. With a 10 ksec exposure we reach ~3x10-15 (0.5-2 keV) similar to the PSPC

exposure in the LH (with better spatial positions and broad band spectra)

3.Find BRIGHT NEARBY counterparts of the enigmatic sources detected in the deep Chandra fields.

Introduction

Chandra surveys have shown that the X-ray background consists of various classesof sources:

1. QSOs 2. Obscured AGN 3. Optically ‘passive’ galaxies4. Optically faint sources 5. Star-forming galaxies at very faint fluxes

(Classes 3,4 may take into account the scarcity of obscured AGN)

The nature of these new populations is hard to explore mainly due to the fact that these are quite faint

ROSAT showed that the soft (0.5-2 keV) XRB is dominated by BL AGN

In harder energies (2-10 keV) the logN-logS shows an excess of sources Relative to the ROSAT logN-logS (obscured AGN ?). Couldn’t tell

Outstanding issues

What is the nature of ‘passive’ X-ray galaxies (XBONG) ?

Can the optically faint sources be obscured AGN at high z ?

Where are the type-1.9/2 QSOs ? (only 3 detected in Chandra surveys)

A number of Type-1 QSOs present large amounts of absorption ?

The last two points present great interest for AGN unification models (and XRB population synthesis models)

Survey Description

18 shallow 5-10 ksec pointings in the NGP/SGP covering ~4 deg2.

WHY THERE ?Extended optical follow-up :

Spectroscopic 2-df B<21 QSOs (NGP/SGP)Photometric B=22.5, spectroscopic by the SDSS (NGP)

ADJACENT fields They can be observed in a single 2-df exposure to go fainter

NGP details

0.5-8 keV 0.5-2 2-8keV---------------------------------------------

215 197 61 (>6σ) 135 134 49 with optical c’parts p>98.5% 36 31 4 with opt. C’parts at p<98.5% 44 32 8 No opt. C’parts down to B=22.5 63 26 Spectroscopic Redshifts 2-df/SDSS/Deeper 2-df

23 4 Photometric Redshifts

2 sources in the 2-8 keV catalogue only 13 sources in the soft 0.5-2 keV catalogueHence the 0.5-8 keV total catalogue gives a fair representation of the X-ray populations

What to remember:Spectroscopic Follow-up largely incomplete (possibly biased)Assigning probabilities helps to reject the chance coincidences

logN-logS

At these fluxes about 50% of the XRB is resolvedin both bands

Therefore our objects are typical of those makingthe XRB

The fx/fo diagram

Dominated by AGN (morphology suggests both nearby and distant AGN)

Some galaxies are observed with low fx/fo

Spectra:X BL AGN NL/� ABS Galaxy

Morphology:GalaxyStar • unclassified morph Low prob.

0.5-8 keV

Optical Ids

2-df/SDSS and deeper AAT/2-df spectroscopy Vast majority in the 0.5-8 keV sample associated with broad-line AGN

0

10

20

30

40

50

60

BL AGN NELG Star

Type 0.5-8 2-8---------------------------------QSO/Sy1 54 21NELG/Gal. 10 5Star 5 2No-ID 102 25

Photometric Redshifts

Using the Code of Hatziminaoglou et al. (1999)

It does not work for extended objects where the colours are contaminatedby the galaxy

Object selection:1. Stellar2. High fx/fo

70% of these have Dz<0.3 (Kitsionas et al. 2003)

The Redshift Distribution

Excess number of sources at low Redshift (even when we take into account only the type-1)

Blue= type-2Red= type-1Green=Photom

0.5-8 keV

2-8 keV

The 0.5-2/2-8 hardness ratios

Large columnns are observed, much higher than those encountered in type-1.

Still these are lower than those predicted by population synthesis models.

5

0.5-8 keV

2-8 keV

0.5-8 keV

The column density histogram

Two peaks with Galactic and moderate (logN_H~21.5) column densities

Surprisingly, the galaxies have A lower N_H than the point sources.Things get worse considering

That the point sources are furtherAway and thus that the rest-frameN_H will be even larger

0.5-8 keV

Spectra:X BL AGN NL/Galaxy�

Morphology:Galaxy Star

Most galaxies are red but these are not necessarily hardie there is not a simple picture where we find many Nearby Obscured AGN

Where are the type 1.9-2 QSOs ?

Only local type-2 AGN detected so farexception is RXJ at z=2.35 already discovered by ROSAT (Almaini et al. 1995, Georgantopoulos et al. 1999)

0.5-8 keV

2-8 keV

The 0.5-8 keV sampleis dominated by the blueAGN. In the hard band red galaxies and blueAGN play equal roles

The colour-colourdiagram easily picks out candidate type-2 AGN(Note that the HR of the red stellar objects is not hard)

Type-1 QSOs with large N_H

Correlation with redshift also found by Reeves & Turner 2000 (but logN_H~21 )

Corrected N_H for redshift

Population synthesis modelspredict 2/3 logN_H>2210-14 cgs 2-10 keV

An Example at z=0.8

N_H= 3 (+-2)x10^21 cm-2Photon index =1.9 fixed

Implications for AGN unification models

The apparent scarcity of type-2 AGN but mainly the observation of type-1 AGN with large amounts of X-ray absorption could possibly suggest that:

At high redshifts Dust cannot survive but neutral gas can. Dust sublimates in the strong radiation field ? sublimation radius = 0.2 pc for L~1045

But then the X-ray absorbing gas must avoid ionizationFor 0.2 pc, L~1045 the ionization constant is relatively neutral ξ =1 for densities n>109 cm-3

Dust coagulation is an alternative

Optically passive galaxies

Lx> 1043 for all 3 galaxies

N_H=0Photon index=1.9 fixedCf Severgnini et al. in press

Incompleteness of UVX surveys

Not significant as far as the colour selection is concerned The optical extension may be a far bigger problem

Summary

Large number of sources at low redshift (even more pronounced at hard energiessee Jahoda et al. 1991, Lahav et al. 1993)

There is an apparent scarcity of absorbed (narrow-line) AGN at high redshift (same in IR wavelengths)

Instead there is a number of type-1 QSOs with large column densities

These have interesting implications for unification models.Dust sublimates but the gas remains neutral and hence absorbs X-rays ?

No significant incompleteness of optical UVX surveys found due to colourselection.

Type-1 AGN are many more than type-2 ?

Galaxies

How we find them:

1. Objects with log(fx/fopt)<-12. Extended in the optical 3. Additional criterion hardness

ratio

2-4 galaxies in ALL 18 fieldsZ=0.05-0.09

Galaxies are quite rare:One would need tens Of XMM fields to form a decent sample

Another way of determining the propertiesof galaxies is by using stacking analysis (>200 2df galaxies in our fields z~0.1)

Extend at higher redshiftsusing radio galaxiesZ~0.4