Spatial variability of seismicity parameters in aftershock zones

Abstract

The spatial variability of the b value of the frequency-magnitude relationship and the decay rate of aftershocks as described by the p value of the modified Omori law is investigated. By using dense spatial grids we map out the distribution of b and p values within the Landers, Northridge, Morgan Hill, and Kobe aftershock sequences. Considerable spatial variability is found, with b values of independent subvolumes ranging from 0.6 to 1.4, and p values ranging from 0.6 to 1.8. These systematic and statistically highly significant differences argue that it is an oversimplification to assign one single p and b value to an aftershock sequence that extends up to 100 km. The spatial distribution of these two parameters is compared with the slip distribution during the mainshock, suggesting that the areas of largest slip release correlate with high b value regions. We hypothesize that the frictional heat created during the event may influence the p value distribution within an aftershock zone, while applied shear stress, crack density and pore pressure govern the frequency-magnitude distribution. By investigating the frequency-magnitude distribution separately for preseismic and postseismic periods for the Morgan Hill mainshock, we find that only the volume in the vicinity of the highest slip release shows a significant increase in the b value, which decays to premainshock values within a year. Surrounding areas of the aftershock zone show an approximately constant b value with time. Because the aftershock hazard after a mainshock depends strongly on both the b and p value, we propose that aftershock hazard assessment can be improved by taking into account the spatial distribution of the parameters. Copyright 1999 by the American Geophysical Union.