Detection of Periodic Variability in Simulated QSO Light Curves

Periodic light curve behavior predicted for some binary black hole systems might be detected in large samples, such as the multi-million quasar sample expected from the Large Synoptic Survey Telescope (LSST). We investigate the false-alarm probability for the discovery of a periodic signal in light curves simulated using damped random walk (DRW) model. This model provides a good description of observed light curves, and does not include periodic behavior. We used the Lomb-Scargle periodogram to search for a periodic signal in a million simulated light curves that properly sample the DRW parameter space, and the LSST cadence space. We find that even a very conservative threshold for the false-alarm probability still yields thousands of "good" binary black hole candidates. We conclude that the future claims for binary black holes based on Lomb-Scargle analysis of LSST light curves will have to be interpreted with caution.