Accretion Timescales from Kepler AGN
Abstract
We constrain AGN accretion disk variability mechanisms using the optical light curves of AGN observed by Kepler. AGN optical fluxes are known to exhibit stochastic variations on timescales of hours, days, months and years. The excellent sampling properties of the original Kepler mission - high S/N ratio (105), short sampling interval (30 minutes), and long sampling duration (~ 3.5 years) - allow for a detailed examination of the differences between the variability processes present in various sub-types of AGN such as Type I and II Seyferts, QSOs, and Blazars. We model the flux data using the Auto-Regressive Moving Average (ARMA) representation from the field of time series analysis. We use the Kalman filter to determine optimal mode parameters and use the Akaike Information Criteria (AIC) to select the optimal model. We find that optical light curves from Kepler AGN cannot be fit by low order statistical models such as the popular AR(1) process or damped random walk. Kepler light curves exhibit complicated power spectra and are better modeled by higher order ARMA processes. We find that Kepler AGN typically exhibit power spectra that change from a bending power law (PSD ~ 1/fa) to a flat power spectrum on timescales in the range of ~ 5 - 100 days consistent with the orbital and thermal timescales of a typical 107 solar mass black hole.
- Publication:
-
American Astronomical Society Meeting Abstracts #225
- Pub Date:
- January 2015
- Bibcode:
- 2015AAS...22522102K