The XMM-Newton slew survey in the 2-10 keV band. (arXiv:1210.3992v1 [astro-ph.CO])

The XMM-Newton slew survey in the 2-10 keV band. (arXiv:1210.3992v1 [astro-ph.CO]):
The XMM-Newton Slew Survey (XSS) covers a significant fraction of the sky in
a broad X-ray bandpass. Although shallow by contemporary standards, in the
`classical' 2-10 keV band of X-ray astronomy, the XSS provides significantly
better sensitivity than any currently available all-sky survey. We investigate
the source content of the XSS, focussing on detections in the 2-10 keV band
down to a very low threshold (> 4 counts net of background). At the faint end,
the survey reaches a flux sensitivity of roughly 3e-12 erg/cm2/s (2-10 keV).
Our starting point was a sample of 487 sources detected in the XMMSL1d2 XSS at
high galactic latitude in the hard band. Through cross-correlation with
published source catalogues from surveys spanning the electromagnetic spectrum
from radio to gamma-rays, we find that 45% of the sources have likely
identifications with normal/active galaxies, 18% are associated with other
classes of X-ray object (nearby coronally active stars, accreting binaries,
clusters of galaxies), leaving 37% of the XSS sources with no current
identification. We go on to define an XSS extragalactic hard band sample
comprised of 219 galaxies and active galaxies. We investigate the properties of
this extragalactic sample including its X-ray logN-logS distribution. We find
that in the low-count limit, the XSS is strongly affected by Eddington bias.
There is also a very strong bias in the XSS against the detection of extended
sources, most notably clusters of galaxies. A significant fraction of the
detections at and around the low-count limit may be spurious. Nevertheless, it
is possible to use the XSS to extract a reasonably robust sample of
extragalactic sources, excluding galaxy clusters. The differential logN-logS
relation of these extragalactic sources matches very well to the HEAO-1 A2
all-sky survey measurements at bright fluxes and to the 2XMM source counts at
the faint end.