Abstract
Surface-rupturing earthquakes tend to cause larger damage than earthquakes without surface rupture (Takemura,1998). However, the degree of damage during an earthquake is largely influenced by population, surface geology and strength of houses in the epicentral area. I calculated the index valueIby eliminating the effects of population, surface geology and strength of houses from compiled damage data of 37 earthquakes occurred intraplate Japan during the past 115 years:I=D/(j⋅k⋅D7) whereDis number of collapsed houses during an earthquake, D7, is calculated number of collapsed houses when a point source of moment magnitude Mw: 7.0 (depth: Okm) is assumed at the epicenter, j is a coefficient of calibration for population chang e through time, and k is a coefficient of calibration for strength change of houses through time. A damage estimation program by National Land Agency of Japan is used for the damage calculation.
Ivalues for surface-rupturing events are expected to be at leas t a factor of ten larger than those for events without surface rupture in the range of Mw: 6.4-6.6 which is an over-lap zone of both types of events. The difference of depth to the upper edge of the seismogenic fault between both types of events possibly explains this observation. However, I interpret it to suggest that source property changes when rupture cuts free surface and this might cause larger damage.
All of the earthquakes ofI>0.5 are surface-rupturing events except ones which occurred beneath alluvial plains. Since all of surface-rupturing events occurred by movements of geomorphologically detectable active faults, I can empirically conclude that earthquake with large potential to cause great damage always occurs on geomorphologically detectable active fault except one beneath alluvial plain.
