Spatial analysis of the frequency-magnitude distribution and decay rate of aftershock activity of the 2000 Western Tottori earthquake

The b-value of the frequency-magnitude distribution and the parameters in the modified Omori law, describing the decay rate of aftershock activity, are investigated for more than 4000 aftershocks identified in the first four months after the Western Tottori earthquake (October 6, 2000). We used the JMA data catalog, containing aftershocks with magnitude larger than or equal to 2.0. The studied area is first divided into three areas: one region (A) corresponding to the main aftershock area and other two (B and C) corresponding to seismic activity probably triggered by the stress change caused by the main shock. For region A, the magnitude of completeness (Mc) decreases with time, from the largest value of 3.2 in the first two hours of the sequence, to 2.0, about four days after the main shock. Taking the threshold magnitude as 3.2, we estimated the b-value for the whole region A to be about 1.3 and p-value around 1. However, highly significant variations in both b and p values are found when analyzing their spatial distribution in region A. The seismic activity in the regions B and C started about 2.5 days after the main shock. The b-value for region B (Mc = 2) is 1.05. The decay rate of earthquake activity in Region B is well modeled by the modified Omori law and the p-value is found to be relatively low (0.83). The number of events in region C is too small for a meaningful study. The physical interpretation of the spatial variation of the parameters is not straightforward. However, the variation of b-value can be related to the stress distribution after the main shock, as well as the history of previous ruptures. Thus, the relatively low stress in the regions that have already experienced rupture is probably responsible for the larger value of b found in these areas. Regions with relatively low b-value, on the other hand, are probably regions under higher applied shear stress after the main shock. Alternatively, one can hypothesize that the areas that experienced slip are more fractured, favoring higher b-values. The larger p-values correlate well with the regions that experienced larger slip during the main shock, while small p-values are found generally in regions that have not ruptured recently. The variation of p-value can be related with the frictional heating produced during rupture. The crustal structure may explain some local features of b and p value spatial distribution. In order to verify our hypothesis we also analyzed the seismic activity that occurred before the Tottori earthquake, starting in 1978, using the data of DPRI, Kyoto University. It seems that the previous seismic activity associated with some moderate events in 1989, 1990 and 1997 had an influence on the following seismicity in the area-in particular on the spatial distribution of b and p values observed for the aftershocks of the Tottori earthquake. The aftershocks of the 1997 M5.5 earthquake have a larger p-value than previous aftershock sequences, while the b-value has a clear increase following the M5.5 event.