EXSdetect: an end-to-end software for extended source detection in X-ray images: application to Swift-XRT data

Abstract: Aims. We present a stand-alone software (named EXSdetect) for the detection of extended sources in X-ray images. Our goal is to provide a flexible tool capable of detecting extended sources down to the lowest flux levels attainable within instrumental limitations, while maintaining robust photometry, high completeness, and low contamination, regardless of source morphology. EXSdetect was developed mainly to exploit the ever-increasing wealth of archival X-ray data, but is also ideally suited to explore the scientific capabilities of future X-ray facilities, with a strong focus on investigations of distant groups and clusters of galaxies. Methods. EXSdetect combines a fast Voronoi tessellation code with a friends-of-friends algorithm and an automated deblending procedure. The values of key parameters are matched to fundamental telescope properties such as angular resolution and instrumental background. In addition, the software is designed to permit extensive tests of its performance via simulations of a wide range of observational scenarios. Results. We applied EXSdetect to simulated data fields modeled to realistically represent the Swift X-ray Cluster Survey (SXCS), which is based on archival data obtained by the X-ray telescope onboard the Swift satellite. We achieve more than 90% completeness for extended sources comprising at least 80 photons in the 0.5-2 keV band, a limit that corresponds to 10^-14 erg cm^-2 s^-1 for the deepest SXCS fields. This detection limit is comparable to the one attained by the most sensitive cluster surveys conducted with much larger X-ray telescopes. While evaluating the performance of EXSdetect, we also explored the impact of improved angular resolution and discuss the ideal properties of the next generation of X-ray survey missions.