Statistical search for non-random features of the seismicity of strong earthquakes

Abstract

The space-time-magnitude relationship among worldwide earthquakes of magnitude M ⩾ 7 has been investigated with a view toward discerning the statistical reliability of possibilities of epicenter migration, gaps in seismic activity, and techniques of identifying aftershocks and foreshocks. The statistical procedure involves the calculation of the second-order moment of the sequences. The statistical validity of the results was obtained by modeling the interaction of discrete seismic events by a multidimensional branching stochastic point process. The results are that epicenters of large earthquakes migrate with velocities between 300 and 2,000 km/year with a significance level greater than 99.5%; the maximum likelihood estimate of the velocity of migration of large aftershocks of large earthquakes is about 1,400 km/year; foreshocks of large earthquakes have a maximum likelihood estimate of velocity of migration of about 2,800 km/year. A gap in seismic activity occurs before large earthquakes; with somewhat less confidence, a gap also occurs after large earthquakes, after the aftershock sequence has ended. The number of immediate foreshocks is 35–40% of the number of aftershocks. Aftershocks are comparatively weaker than foreshocks when compared with the main shock. There are hints, that are not fully statistically confirmed, of interaction among large earthquakes in different depth ranges.