The Proportional Hazard Model applied to the CSEP testing area in Italy

Abstract

This paper presents the strategies adopted to modify the Proportional Hazard Model (Faenza et al., 2003; Cinti et al., 2004) to fit the requirement for the forecast test within CSEP experiment. The model has been originally proposed to study the spatiotemporal distribution of the M 5.5+ seismicity in Italy, using two spatial models, a regular grid and a seismotectonic zonation. A prospective 10-years forecast test is already ongoing since 2005 and the results are available at the webpage (http://www.bo.ingv.it/ earthquake/ ITALY/forecasting/M5.5+/). In that test, we report the probability maps of M 5.5+ earthquake for the next 10 years for two spatial distributions. Since the original model is time-dependent, it is updated every year and immediately after the occurrence of a target event, e.g., Mw 5.5. Despite that prospective test is continuing and that the model updates probabilities different from CSEP experiments, we argue that a full evaluation of the model could be achieved only through the CSEP testing experiment, where the performances of different models are compared using the same rules and tests. The major modification we have introduced into our model is the simulation of the expected numbers of events in the exposure time ∆τ. This is performed considering the probability that an event occurs in ∆τ and evaluating the change this will cause into the expected numbers of events. This procedure is implemented for the first and the second generation of aftershocks.