📝 SummarXiv

Abstract

We present a systematic search for spatial association between a high confidence sample of 3,166 morphologically selected galaxy mergers drawn from the initial convolutional neural network (CNN) catalog of 328,151 candidate mergers and refined through SDSS cross-matching and the Fermi Large Area Telescope Fourth Source Catalog Data Release 4 (4FGL-DR4) catalog of gamma-ray sources detected by the Fermi Large Area Telescope (LAT). Using a conservative 4$\sigma$ positional uncertainty threshold and a Poisson based statistical framework, we identify 21 statistically significant associations with match probabilities $p < 0.05$. Among these are known classes of gamma-ray emitters such as Flat-Spectrum Radio Quasars (FSRQs), BL Lacertae objects (BLL), and radio galaxies bolstering the hypothesis that merger driven processes can fuel high energy activity. Intriguingly, five of the associated sources remain unclassified in 4FGL-DR4, hinting at a possible link between galaxy mergers and a hitherto unrecognized population of gamma-ray sources. The dominance of AGN like matches supports the scenario in which mergers trigger accretion onto central supermassive black holes, initiating AGN activity observable at gamma-ray energies. Moreover, the recurrent presence of unassociated sources among secure matches underscores the potential of merger catalogs as physically motivated priors in gamma-ray source identification efforts. This work constitutes the first dedicated effort to explore associations between gamma-ray sources and a large, morphologically selected sample of galaxy mergers catalog, opening new avenues for understanding the role of galaxy interactions in high-energy astrophysics.