Long-Term Clustering and Multifractality in Non-Stationary Earthquake Time Series. Insights from Corinth Rift, Greece

Abstract:

Earthquake time series are widely used to characterize the main features of seismicity and to provide useful insights into the dynamics of the seismogenic system. Properties such as fractality/multifractality, intermittency and non-stationary clustering are common in earthquake time series, highlighting the complex nature of the earthquake generation process. Here we use statistical physics to study the temporal properties of the earthquake activity in one of the most seismically active areas in Europe, the Corinth rift (central Greece). The earthquake activity in the Corinth rift is typically characterized by fluctuating behavior, where periods of low to moderate activity are interspersed by sudden seismic bursts, which are related to frequent earthquake swarms and the occurrence of stronger events, followed by aftershock sequences. A multifractal analysis reveals the degree of heterogeneous clustering in the earthquake activity and correlations acting at all time-scales that suggest non-Poissonian behavior. These properties are also displayed in the probability density function of the scaled inter-event times (i.e. the time intervals between the successive earthquakes), where for various time periods and threshold magnitudes the distribution presents scaling and two power-law regions at both short and long time intervals. In addition, we use generalized statistical physics and a stochastic dynamical mechanism with memory effects to model this behavior. During stationary periods where the mean inter-event time does not fluctuate, the solution of this mechanism is the gamma distribution, while for non-stationary periods the solution is a q-generalized gamma distribution, which exhibits power-law asymptotic behavior. These properties seem to be fundamental in non-stationary earthquake time series such that they should be considered in probabilistic hazard assessments, especially in localized seismicity where highly nonrandom activity may be expected.