Accuracy of the recovered flux of extended sources obscured by bad pixels in the central EPIC FOV
A fraction of the XMM-Newton/EPIC FOV is obscured by the dysfunctional (i.e. bad) pixels. The fraction varies between different EPIC instruments in a given observation. These complications affect the analysis of extended X-ray sources observed with XMM-Newton/EPIC and the consequent scientific interpretation of the results. For example, the accuracy of the widely used cosmological probe of the gas mass of clusters of galaxies depends on the accuracy of the procedure of removing the obscuration effect from the measured flux. The Science Analysis Software (SAS) includes an option for recovering the lost fraction of the flux measured by a primary instrument by utilising a supplementary image of the same source. The correction may be accurate if the supplementary image is minimally obscured at the locations of the bad pixels of the primary instrument. This can be achieved e.g. by using the observation-based MOS2 image for correcting the pn flux, or by using a synthetic model image. By utilising a sample of 27 galaxy cluster observations we evaluated the accuracy of the recovery method based on observed images, as implemented in SAS 18.0.0. We found that the accuracy of the recovered total flux in the 0.5-7.0 keV band in the full geometric area within the central r = 6 arcmin is better than 0.1% on average while in some individual cases the recovered flux may be uncertain by ~1%.