Recurrence Times of Large Earthquakes: Assimilating the Effect of Seismic Coupling into a Renewal Model

Abstract

ABSTRACT A renewal model for estimating recurrence times of large earthquakes is presented, which allows one to assimilate data from paleo, historic, and instrumental earthquake data as well as geodetic information in a simple and efficient way using three parameters: the Richter b-value, the mean recurrence time μT of paleo and historic earthquakes, and the seismic coupling coefficient χ. The core of the model is a state variable with linear drift in time reflecting Reid’s elastic rebound theory, decorated with random fluctuations modeling small and intermediate earthquakes. Furthermore, it is assumed that a fraction between 0% and 100% of the average seismic moment is released seismically, and the remaining part corresponds to aseismic processes such as creep. The model extends former renewal models of Matthews et al. (2002) and Zöller et al. (2008). Results for different regions indicate overall good performance and thus suggest that the model provides physically motivated constraints for recurrence time estimations in contrast to the standard fitting of a few data points to an assumed distribution.