Identification probabilities. We provide the probability of identification of each 2XMM DR3 source with several large catalogues (GSC2.2, USNO A-2, USNO B-1, 2MASS and SDSS DR7). Note that these identification probabilities are derived from a cross-correlation process different from that performed at the pipeline level with the 202 catalogues mentioned above.
In this process, we use the merged source coordinates and only consider catalogue entries as bright as the candidate to compute the likelihood of a spurious match.
Source classification method.
We present a tentative source classification in terms of two classes; i) stars and ii) extragalactic (AGN + "normal" galaxies). The learning sample defining class properties is mainly based on 2XMMi / SDSS DR7 spectroscopic identifications. We use the kernel density classification (KDC) of Richards et al. (2004) which is a non-parametric Baysian classifier computing lilelihoods using a kernel smoothing method.
Three sets of classification probabilities are provided, based on:
The four EPIC Hardness ratios with err(HR)< 0.2 (4-d space)
Paper presented at ADASS 2009: Compares the efficiencies of several classification methods on a star/extragalactic grouping and shows how the kernel density classification method behaves when handling more detailed classes.
A more detailed description of the way the leaning sample was built is found here.
The catalogue has been constructed by the XMM-Newton Survey
Science Centre (SSC) on behalf of ESA.
The XCATDB has been built with Saada 1.6 by the SSC
team of Strasbourg. This interface has been designed by L. Michel
The metadata describing XMM-Newton EPIC spectra according to
the IVOA Simple Spectral Access Protocol has been provided by Jai Won
Kim of the German Astrophysical Virtual Observatory (GAVO) and Frank
Haberl of the Max-Planck Institute for Extraterrestrial Physics. GAVO
is supported by a grant of the German Federal Ministry of Education
and Research (BMBF) under contract 05AC6VHA.