Enhancing Chandra-Gaia Crossmatching with Machine Learning

Understanding the nature of astrophysical X-ray sources may require a more nuanced, multi-wavelength approach than previously considered. While traditional cross-matching algorithms primarily rely on Bayesian statistics and spatial information, these methods may not capture the full scope of the intrinsic physical characteristics of astronomical objects across different wavelengths. This limitation raises the possibility of errors and misidentifications, particularly when ambiguity exists in the matches. In this work, we introduce a machine-learning-based approach to crossmatch the Chandra Source Catalog 2.1 with Gaia DR3. By exploring relationships between optical magnitudes, colors, X-ray hardness ratios, X-ray variability, among other properties, our model will generate likelihoods of two sources being a true match. Preliminary results suggest that a reliance solely on spatial separation for crossmatching could be misleading. When integrated with existing spatial crossmatch algorithms, our model shows potential to enhance the overall reliability of the probabilities. This method could be used to, for example, obtain more unambiguous cross-matches to stars, and might help fill gaps in our understanding of their rotational evolution. We acknowledge support from Chandra grant AR3-24002X and from the NASA Contract to the Chandra X-Ray Center NAS8-03060.