SHORT-TERM SPACE-TIME VARIATIONS IN REAL-TIME OF THE ITALIAN SEISMIC ACTIVITY

The algorithm developed during the project INGV-DPC S4 provides the users with quantitative estimates of the probability of occurrence of new earthquakes on specific areas of the target terri-tory. The algorithm avails of the earthquake data detected by the Italian National seismic network. The software adopted for the estimation of the space-temporal seismic hazard is based on epidemic models of the ETAS type (Epidemic Type Aftershock Sequence) (Console et al., 2001, 2003, 2006, 2007). In epidemic models each event can be at the same time inducing another earthquake and indu-ced by a previous one. The expected seismicity rate in any particular point of the target area for a given threshold value can be determined through the contribution of all previous events using a ker-nel function that involves: magnitude, distance and time of occurrence of every previous events. The magnitude distribution assumed in the model follows the Gutenberg-Richter law. The parameters used by the software have undergone a first phase of training using the INGV data set, obtaining a maximum likelihood estimate of the parameters. The time span for training is from July 1987 to December 2005 for M>2.0 earthquakes. The statistical software ZMAP (Wiemer and Wyss, 1994) developed at ETH Zurich has been used to verify the data completeness