活断層と地震 : その地質学的研究

This paper deseribes natures of surface faulting associated with earthquakes and its relation to Quaternary faults, emphasizing that the geological approach is indispensable to assess future earthquakes. Temporal and regional consistency between faultings in historic and Quaternary times gives us a basis for preestimation of location, magnitude and period of future earthquakes. For this, following items have been investigated: 1) Distribution of Quaternary faults. It is recognized that there are two types of fault system with respect to mode of strain release, the total release type (e.g. the Nobi fault system) and partial release type (e.g. the Aizu fault system). Still the dimension of fault system is one of the useful criteria for estimating probable maximum magnitude of the earthquake. 2) Long-term rate of faulting in Quaternary time. This represents a secular strain rate of the crust around a fault, and is practically the best indicator of degree of fault activity. It is shown that the long-term rate of faulting of the Japanese inland faults ranges from almost 10 m to 0.01 m per one thousand years. The Japanese Quaternary fault has been classified first according to this value into Classes A, B, C. Quaternary faults of more than 0.1 m/10^3 years (Classes A and B) in rate of faulting are easily detectable from the topographic method using aerial photos of scale 1:40,000. About 50% of historic great earthquakes that accompanied by surface faults took place at easily detectable fault of classes A and B, but the rest took place at hardly pre-detectable ones (class C fault). This suggests that the lower class faults are much greater in number than higher class faults. 3) Amount of fault displacement in a single event ("unit displacement"). This is used for estimating recurrence interval of earthquake from long-term rate of faulting. The values may be obtained from detail geologic and topographic observations or inferred from historic record. Class A faults have a recurrence interval of about 300 and 1300 years if we assume magnitude 7 and 8, respectively. 4) Age of the last activity or period elapsed without earthquake up to present in historic time. Comparison of the time clasped since the last event with the recurrence interval calculated as mentioned above will serve for a rough estimation of time of the next faulting or earthquake. Based on the above criteria, an attempt has been made to estimate the magnitude and recurrence period of some active faults in Japan. It is to be emphasized that there are many Quaternary faults in Japan, and that they are quite different in the long-term rate of faulting and in the recurrence periods, the latter ranging from an order of 10^2 to greater than 10^5 years. The present state of this line of approach can make an order estimation for recurrence period of earthquakes for a given fault.