Extracting Information From AGN Variability: an LSST AGN Collaboration Proposal Vishal Kasliwal, University of Pennsylvania and Princeton University AGN exhibit rapid, high amplitude stochastic flux variability across the entire electromagnetic spectrum on timescales ranging from hours to years. The cause of this variability is poorly understood. We present a new method for using variability to (1) measure the time-scales on which flux perturbations evolve and (2) characterize the driving flux perturbations. We model the observed light curve of an AGN as a linear differential equation driven by stochastic impulses. Physically, the impulses could be local `hot- and cold-spots' in the accretion disk—the linear differential equation then governs how the spots evolve and dissipate. The impulse response function of the accretion disk material is given by the Green’s function of the linear differential equation. The timescales on which the spots radiate energy is characterized by the power spectral density (PSD) of the driving stochastic impulses. We analyze the light curve of the Kepler AGN Zw 229-15 and find that the observed variability behavior can be modeled as a damped harmonic oscillator perturbed by a colored noise process. The model PSD turns over on time-scale 385 days. On shorter time-scales, the log-PSD slope varies between 2 and 4, explaining the behavior noted by previous studies. We recover and identify both the 5.6 day and 67 day timescales reported by previous work. These timescales represent the time-scale on which flux perturbations grow, and the time-scale on which flux perturbations decay back to the steady-state flux level respectively. We present the software package used to study light curves using this method, KALI, which is freely available to the community. We conclude by discussing future prospects for the study of AGN variability.