Description
The “null-signal template” method is a new, data-driven approach to testing the significance of potentially periodic signals in quasar light curves. NST employs Bayesian statistics to differentiate between supermassive black hole binary (SMBHB) signals and the complex noise of individual quasars. It accounts for correlated noise as a damped random walk and provides an alternative hypothesis for slow trends in the light curve using sinusoid-like variability with a randomized length for each period. NST is more computationally efficient than traditional significance testing methods, since it provides a testable alternative hypothesis to the periodic SMBHB signal without requiring simulations of quasar noise. We will share our tests on the validity of the NST method using simulated LSST-like light curves and discuss the prospects for its application in the upcoming survey.
