Changes in the dynamics of seismic process observed in the fixed time windows; case study for southern California 1980–2020

Abstract In the present work we investigated changes in the extent of regularity (randomness) of seismic process in fixed time span windows. A comparison with a set of randomized catalogues was accomplished basing on spatial, temporal and energetic characteristics of a seismic process. Increments of cumulative times, increments of cumulative distances and increments of cumulative seismic energies have been calculated from the southern California earthquake catalogue, 1980 to 2020. The multivariate Mahalanobis' distance calculation, combined with the surrogate data testing procedure, was chosen as the analysis method. An analysis of variability in the extent of regularity of a seismic process has been accomplished for different completeness magnitude thresholds and sliding windows of different time spans. Analysing the features of temporal, spatial and energetic variability, in periods of supposedly aftershock activity, we found that the original seismic process is significantly different from a random process. Such periods, containing windows with nonrandom seismicity, are always ended by a series of windows in which a seismic process is indistinguishable from a random one. This was shown at different magnitude thresholds, comparing the original catalogue with the set of randomized catalogues. It was also found that at small magnitude thresholds (M2.6 and M3.0), the fixed time span windows with a seismic process significantly different from the random process might have occurred also prior to four large (M > 7.0) earthquakes in the considered catalogue. The amount of the released seismic energy in these windows is essentially smaller than in the ones after strong earthquakes of smaller magnitudes. Relying on our results we suggest that causes of regularity in the seismic process prior and after large earthquakes are probably different. At larger-magnitude thresholds, the total number of fixed time windows with a seismic process different from the random one gradually decreases. Moreover, at M3.8 and M4.2 thresholds there are practically no windows with regular seismic process prior to four large catalogue earthquakes. In the periods between strong earthquakes, mostly in the periods of relatively small earthquakes generation, the percentage of windows in which the seismic process is indistinguishable from the random one essentially increases with increasing magnitude threshold.