Relative seismic quiescence before large aftershocks

In terms of the modified Omori's formula and Akaike's Information Criterion , thirteen aftershock sequences have been analysed quantitatively, to check if any anomalous change of seismic activity took place before the occurren ce of the largest aftershock. The analysis follows the method proposed by M ATSU'URA (1986). The epicentre distribution of the main shocks covers the A egean Sea (Greece) area and the surrounding regions. The earthquakes occurr ed from 1988 to 1998. Although difficulties existed for a quantitative treatment of seismic quies cence, the results of the temporal analysis of the aftershock activities se em to be promising. In eight of thirteen cases the observed temporal patter n is similar. The aftershock activity decreases significantly before the oc currence of the largest aftershock, with respect to the level expected acco rding to the modified Omori's formula. Thereafter, the activity recovers to its previous level just before or immediately after the occurrence of the main aftershock. Moreover, an estimation of the reliability of the method i s made, based on the results of this study as well as of previous similar i nvestigations. From a total of 41 aftershock sequences which have been anal ysed following Matsu'ura's method, 29 of them (71%) show a remarkable relat ive quiescence before the occurrence of the main shock. Although it is difficult to construct a model which explains the precursory pattern, the real time monitoring and check for seismic quiescence could a t least contribute to a 'qualitative' prediction of the largest aftershock which could at times be as large and disastrous as the main shock.