Estimating Black Hole Masses in Hundreds of Quasars

Abstract

We explore the practical feasibility of AGN broad-band reverberation mapping and present first results. We lay out and apply a rigorous approach for stochastic reverberation mapping of unevenly sampled multi-broad-band flux measurements, assuming that the broad-line region (BLR) line flux is contributing up to 15 % in some bands, and is directly constrained by one spectroscopical epoch. The approach describes variations of the observed flux as the continuum, modeled as a stochastic Gaussian process, and emission line contribution, modeled as a scaled, smoothed and delayed version of the continuum. This approach is capable not only to interpolate in time between measurements, but also to determine confidence limits on continuum -- line emission delays. This approach is applied to SDSS observations in 'Stripe 82' (S82) providing flux measurements precise to 2 % at $\sim$ 60 epochs over $\sim$ 10 years. The strong annual variations in the epoch sampling prove a serious limitation in practice. Also, suitable redshift ranges must be identified, where strong broad emission line contribute to one filter, but not to another. Through generating and evaluating problem-specific mock data, we verify that S82-like data can constrain $\tau_{\mathrm{delay}}$ for a simple transfer function model. In application to real data, we estimate $\tau_{\mathrm{delay}}$ for 323 AGN with $0.225 < z < 0.846$, combining information for different objects through the ensemble-scaling relationships for BLR size and BH mass. Our analysis tentatively indicates a 1.7 times larger BLR size of H$\alpha$ and MgII compared to Kaspi2000 and Vestergaard2002, but the seasonal data sampling casts doubt on the robustness of the inference.