Soft-Computing Modelling of Seismicity in the Southern Hellenic Arc

Abstract

This letter investigates the possible coalition of time intervals and patterns in seismic activity during the preparation process of consecutive sizeable seismic events (i.e., <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">M</i> <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">S</sub> ges 5.9). During periods of low-level seismic activity, stress processes in the crust accumulate energy at the seismogenic area, while larger seismic events act as a decongesting mechanism that releases considerable amounts of that energy. Monthly mean seismicity rates have been introduced as a tool to monitor this energy management system and to divert this information into an adaptive neuro-fuzzy inference system. The purpose of the neuro-fuzzy model is to identify and to simulate the possible relationship between mean seismicity rates and time intervals among consecutive sizeable earthquakes. Successful training of the neuro-fuzzy model results in a real-time online processing mechanism that is capable of estimating the time interval between the latest and the next forthcoming sizeable seismic event.