Regularities in the time variations of seismic parameters and their implications for prediction of strong earthquakes in Greece

Time variations in the parameters of seismic activity in two regions in Greece, which are known to have different geodynamical conditions, are analyzed using the FastBEE algorithm suggested in (Papadopoulos and Baskoutas, 2009). The study is based on the data on weak earthquakes that occurred in two local regions. One region pertains to the zone dominated by intensive compression stress field, while another is located in the region of a relatively lower intensity extension stress field. It is shown that in the zone of compression the seismic parameters exhibit anomalous temporal behavior before strong earthquakes with Ms ≥ 5.7, whereas in the zones of extension, similar anomalies precede earthquakes with lower magnitudes of up to Ms ≥ 4.9. The most informative parameters for the purposes of predicting strong seismic events are the released seismic energy in the form logE 2/3 and the slope of the frequency-magnitude dependence, b-value. The seismic activity in the region, expressed in terms of the logarithmic number of earthquakes, per unit time in some cases does not exhibit any particular pattern of behavior before strong earthquakes. In the time series of the studied parameters, four stages in the seismic process are clearly distinguished before strong earthquakes. Typically, a strong earthquake has a low probability to occur within the first two stages. Instead, this probability arises at stage III and attains its maximum at the end of this stage coinciding with the occurrence of the strong earthquake. We suggest these features of the time series to be used for the assessment of seismic hazard and for the real-time prediction of strong earthquakes. The time variations in the b-value are found to be correlated with the time variations inlogE 2/3. This correlation is closely approximated by the power-law function. The parameters of this function depend on the geodynamical features of the region and characterize the intensity and the type of the regional tectonic stresses. The results of our study show that the FastBEE algorithm can be successfully applied for monitoring seismic hazard and predicting strong earthquakes.