Probabilistic analysis of macroseismic fields: Iceland case study

In this work our objectives are, first, to describe a probabilistic model aimed at forecasting macroseismic fields and, second, to present the results and findings obtained by applying the model to two of recent damaging earthquakes in the South Iceland Seismic Zone, namely, the earthquakes on 17 June, 2000 and 29 May, 2008, both of assessed epicentral intensity MMI X. The model considers the intensity decay as a random variable having a binomial probability distribution with parameter p, which is in its turn assumed to be a beta random variable. Estimation is carried out according to the Bayesian paradigm; on the basis of a learning set of macroseismic fields, we express our belief on the phenomenon by assigning prior distributions to the model parameters and we then update them with current data. The model has already been applied to Italian earthquakes, in particular, earthquakes occurring in the Mt. Etna volcanic environment.