Applying the Background-Source separation algorithm to Chandra Deep Field South data

TL;DR

This paper introduces a probabilistic two-component mixture model that effectively separates diffuse background from celestial sources in X-ray data, enabling detailed source detection and characterization in deep field observations.

Contribution

The paper presents a novel one-step background-source separation algorithm using a mixture model with multi-resolution analysis, specifically applied to Chandra Deep Field South data.

Findings

Successfully separated background and sources in Chandra data

Automatically parametrized detected sources with positions, fluxes, and morphology

Demonstrated effectiveness on deep field X-ray observations

Abstract

A probabilistic two-component mixture model allows one to separate the diffuse background from the celestial sources within a one-step algorithm without data censoring. The background is modeled with a thin-plate spline combined with the satellite's exposure time. Source probability maps are created in a multi-resolution analysis for revealing faint and extended sources. All detected sources are automatically parametrized to produce a list of source positions, fluxes and morphological parameters. The present analysis is applied to the Chandra Deep Field South 2 Ms public released data. Within its 1.884 ks of exposure time and its angular resolution (0.984 arcsec), the Chandra Deep Field South data are particularly suited for testing the Background-Source separation algorithm.