The  optical identification  uses the  Maximum Likelihood  (LR) method
(Ciliegi et al.  2003, A&A, 398, 901; 2005, A&A, 441 879). This is the
ratio between the probability that  a given source at a given distance
from the X-ray centroid and with a given optical magnitude is the true
ID and the  probability that this source is  a spurious alignment. The
LR  ratio takes  into account  the X-ray  positional  uncertainty, the
background density of optical  galaxies and the apriori probability of
an optical counterpart with a given optical magnitude.

We adopt a search radius  of 4" and identify all possible counterparts
within that  distance. The table  lists only the counterpart  with the
higher LR. For  the cross-correlation with the DEEP2  catalogue we use
the R-magnitude to estimate the surface density of background optical
sources. For the CFHTLS we use the I-band selected catalogue.

For the positional accuracy of  the X-ray centroid we adopt a Gaussian
distribution with a standard deviation 0.5" for  off-axis angles (OAA)
<6arcmin and 1" for >6arcmin. This will be improved in the near future
to include a  more accurate OAA dependence and  also take into account
the  total  counts of  the  source,  i.e.   bright sources  have  more
accurate positions than fainter ones.

NOTE  THAT  ***NOT*** ALL  THE  LISTED SOURCES  IN  THE  TABLE HAVE  A
RELIABLE  COUNTERPART EVEN  IF  THERE  IS AN  ENTRY.  WHEN USING  THIS
CATALOGUE YOU NEED TO APPLY A CUTOFF IN LR AS LISTED BELOW. WE SUGGEST
THAT YOU USE ONLY THOSE IDS WITH:

 *** LH>0.5 FOR DEEP2 

 *** LH>0.2 FOR THE CFHTLS. 


The table columns are:

1. AEGIS X-ray source unique ID 

2. X-ray  centroid RA  in J2000

3. X-ray centroid DEC in J2000

4. Unique DEEP2 ID. This is  the counterpart with the higher LR within
a search radius of 4arcsec. -999  means no ID to the DEEP2 photometric
survey limit. 

5. offset between X-ray and DEEP2 position in arcsec 

6. DEEP2   ID  likelihood   ratio.   Should   be  >0.5   for  reliable
counterparts. For  this threshold the spurious  identification rate is
5%. 

7. Unique CFHTLS ID. This is the counterpart with the higher LR within
a search  radius of 4arcsec .  -999 means that the  sources is outside
the  CFHTLS or in  a masked  region (i.e.  nearby bright  stars). 1000
means no optical ID to the CFHTLS limit. 

8. offset between X-ray and CFHTLS position in arcsec 

9.   CFHTLS  ID   likelihood  ratio.  Should  be   >0.2  for  reliable
counterparts.    For   this   threshold   there   are   ~5%   spurious
identifications. 

10. Total  number of  DEEP2  IDs within  search  radius  of 4  arcsec

11. Total number of CFHTLS IDs within search radius of 4 arcsec


Questions   or   problems?   Please   contact   Antonis   Georgakakis,
age@ic.ac.uk.