Using the Gregory-Loredo Algorithm for the Detection of Variability in the Chandra Source Catalog

In searching for a reliable variability indicator for Chandra X-ray event data, we implemented the Gregory-Loredo (G-L) algorithm to assess variability and to generate light curves for the Chandra Source Catalog. A test was performed on 118 sources detected in a single observation, spanning the intensity range 5-24,000 photons over a total observing time of 102,000 sec. We conclude that the G-L algorithm is extremely robust, yielding both a reliable variability indicator and light curves with optimal resolution, while requiring a very modest amount of CPU time. The algorithm can also work on binned data, and is capable of handling data gaps as well as variations in effective area.