The EEPAS forecasting model and the probability of moderate-to-large earthquakes in central Japan

Abstract

The EEPAS (“Every Earthquake a Precursor According to Scale”) model is a space–time point-process model based on the precursory scale increase (Ψ) phenomenon and associated predictive scaling relations. It has previously been fitted to the New Zealand earthquake catalogue, and applied successfully in quasi-prospective tests on the CNSS catalogue for California for forecasting earthquakes with magnitudes above 5.75 and on the JMA catalogue of Japan for magnitudes above 6.75. Here we test whether the Ψ scaling relations extend to lower magnitudes, by applying EEPAS to depth-restricted subsets of the NIED catalogue of the Kanto area, central Japan, for magnitudes above 4.75. As in previous studies, the EEPAS model is found to be more informative than a quasi-static baseline model based on proximity to past earthquakes, and much more informative than the stationary uniform Poisson model. The information that it provides is illustrated by maps of the earthquake occurrence rate density, covering magnitudes from 5.0 to 8.0, for the central Japan region as at the beginning of year 2004, using the NIED and JMA catalogues to mid-2003.