UCL Department of Space and Climate Physics


Extended Sources

The XMM-SUSS2 SRCLIST table provides a measure of source extent and profile on the sky. In these diagnostics, sources are represented crudely as two-dimensional Gaussian profiles with arbitrary orientation of the major axis relative to celestial north. Major and minor axes are characterised by Full-Width Half-Maxima (FWHM), estimated from Gaussian moments. To test the accuracy of the source extent diagnostics, we compared the XMM-SUSS source profile properties with those associated with optical counterparts within the Sloan Digitized Sky Survey (SDSS). We compare B band XMM-OM sources with g' band sources from SDSS DR6.

The nearest-neighbour cross-correlation of SUSS and SDSS DR6 tables, with a positional tolerance of 2 arcsec, yields 65,579 matches. The SDSS catalogue contains flags for source type, based upon empirical properties, and we segregate stars from galaxies and plot XMM-OM U band magnitude against source extent along the man or axis in Fig 1. 33,143 source are flagged as stars, 30,629 are flagged as galaxies.

Comparing XMM-SUSS source extension measurements with SDSS

Figure 1: left panel - XMM-OM U band magnitudes of SDSS DR6 counterparts plotted against XMM-OM U source extent along the major axis. SDSS source flagged as stars and galaxies are represented by different colours. Right panel - comparison of XMM-OM U band source extent with the counterpart SDSS scale diameter from exponential fits. Colour represents XMM-OM U band magnitude. These graphs were made using XMM-SUSS.

Neglecting source mismatches, the vast majority of flagged stars have XMM-OM detections consistent with a constant FWHM of 2.2 arcsec. This is larger than the quoted U band point spread function of FWHM 1.55 arcsec. Typical XMM-OM image have 2x2 pixel binning and Gaussian moments is not an exact method for calculating the extent of non-Gaussian profile, but it is unclear whether these properties can recover the difference completely. There is an anti-correlation between galaxy extent and magnitude which is presumably a selection effect where diffuse profile wings become increasingly difficult to detect from fainter magnitude objects. This plot provides some confidence that extended sources are being detected robustly by the SAS pipeline.

The right hand panel of Fig 1 compares XMM-OM U band galaxy extent along the major axis with the corresponding SDSS galaxy extent, represented by double the recorded scale radius in the u' band. The XMM-OM U band and SDSS u' bands are chosen for this demonstration because they are the filter bandpasses which most resemble each other. There are two families within this population. There is direct correlation between bright XMM-OM sources and their SDSS counterparts. However compared to the superior depth and angular resolution of the Sloan survey, the XMM-OM is less likely to detect the extended wings of faint sources, resulting in the second population of galaxies, recorded as faint, point sources within the XMM-OM SUSS2.