The attenuation of macroseismic intensity in the volcanic island of Ischia (Gulf of Naples, Italy): comparison between deterministic and probabilistic models and application to seismic scenarios

In this paper, we tackle the problem of the intensity attenuation at Ischia, a critical parameter in a high seismic risk area such as this volcanic island. Starting from the new revised catalogue of local earthquakes, we select a dataset of 118 macroseismic observations related to the four main historical events and analyse the characteristics of the intensity attenuation according to both the deterministic and probabilistic approaches, under the assumption of a point seismic source and isotropic decay (circular spreading). In the deterministic analysis, we derive the attenuation law through an empirical model fitting the average values of ?I (the difference between epicentral intensity Io and intensities observed at a site Is) versus the epicentral distances by the least-square method. In the probabilistic approach, the distribution of Is conditioned on the epicentre-site distance is given through a binomial-beta model for each class of Io. In the Bayesian framework, the model parameter p is considered as a random variable to which we assign a Beta probability distribution on the basis of our prior belief derived from investigations on the attenuation in Italy. The mode of the binomial distribution is taken as the intensity expected at that site (Iexp). The entire calculation procedure has been implemented in a python plugin for QGIS® software that, given location and Io (or magnitude) of the earthquake to be simulated, generates a probabilistic seismic scenario according to the deterministic or probabilistic models of attenuation. This tool may be applied in seismic risk analyses at a local scale or in the seismic surveillance to produce real-time intensity shake-maps for this volcanic area.