Bayesian nonparametric inference for earthquake recurrence time distributions in different tectonic regimes

Objective of the paper is to estimate the probability distribution of the time between consecutive earthquakes with the employ of two equally powerful tools: for the geology, a database of individual seismogenic sources, and for the statistics, assuming that the unknown distribution is a random measure, an estimation method based on the stochastic simulation of Markov chains. The resort to sophisticated instruments is motivated by the particular situation of Italy, where a complex tectonic model is combined with infrequent, medium-size earthquakes. The quality and the length of the parametric catalogue of Italian earthquakes provide for a generous data set, but bring with it problems of incompleteness and uncertainty regarding the parametrization of the events. The pointwise estimate of the inter-event time density functions makes it possible to calculate the occurrence probability depending on the date of the last event at different forecasting horizons.