Pablo Huijse, Jordy Davelaar, Joris De Ridder, Nicholas Jannsen, Conny Aerts

Robust detections of supermassive black hole binaries (SMBHBs) are essential to unravel the role of galaxy mergers in galaxy evolution and for identifying potential sources of low-frequency gravitational waves. One of the most commonly used observational signatures of SMBHBs is periodic variability in the light curves of active galactic nuclei (AGN), which may arise from accretion rate modulation or relativistic Doppler boosting due to binary orbital motion. However, intrinsic stochastic AGN variability can mimic such periodic signals, complicating robust identification. We report the discovery of 181 new SMBHB candidates from a sample of approximately 770,000 AGN observed by the Gaia space observatory. Periodic signals were identified using a novel and computationally efficient Bayesian model selection framework, enabling unbiased source selection and quantifying the likelihood of periodicity over stochastic variability. These candidates nearly double the known SMBHB population and provide a prioritized target list for next-generation time-domain surveys.