Analysis of short time-space range seismicity patterns in Italy

Abstract

In our paper we analyze the data base obtained from the observations of the Italian Seismological Network from 1975 to 1994 by using a simple algotithm to determine the rate of occurrence of seismic events condi- tioned by the occurrence of previous events after a period of quiescence. The number of observed pairs of earthquakes depends on several parameters: the magnitude threshold of the two events, the spatial and tempo- ral ranges of the quiescence period preceding the first (non aftershock) event, the time elapsed between the first and the second events and the spatial dimension of the alarm area. The Akaike information criterion was adopted to assess the optimal set of space-time parameters used in the definition of non-aftershock (events not related to a stronger previous one). In Central Italy, the rate of M ³3.8 earthquakes preceded by at least one M ~ 3.3 foreshock within 14.1 km and 2 days is 30%, while the rate of M ~ 3.3 earthquakes followed by a M ~ 3.8 mainshock in the same space time range is 7%. We observed that the probability that an earthquake of magnitude MI will be followed by an earthquake of magnitude M2 (success rate) fits the law log À = a+b (Mi -M2) with b approximately equal to l. By computing the success rate for given values of magnitude threshold of the first and the second events over a dense grid of spatial coordinates, we obtained maps of this feature over the investigated area. The results of this process document variations larger than a factor of five in the success rate over the Italian territory.

Keywords

• aftershocks
• foreshocks
• conditional probability
• probability gain