This paper aims to estimate a long-term earthquake risk on the Japan islands from the data in “Active faults in Japan: sheet maps and inventories” (Research Group for Active Faults, 1980) and “Catalogue of historical damaging earthquakes in Japan” (Usami, 1975). The maximum magnitude of earthquakes, expectantly caused by active faults with Certainty I, is calculated from fault dimension using Matsuda's formula (1975). Seismic intensity (Japan Meteorological Agency Intensity Scale) is estimated for 161 grid points by Kawasumi's formula (1951). From recurrence interval of active faults, the probability (expectancy) in earthquake occurrence for arbitrary 1, 300 years is calculated. Accummulating the expectancy for each intensity scale, earthquake frequency per any 1, 300 years at each grid point is estimated. Considering the fact that earthquake motions with Intensity VI or higher have caused a great deal of damage in the historical period, it is adequate to discuss the earthquake risk based on the frequency of earthquakes with such intensities.
The earthquake frequency estimated from inland active faults is regarded as the inland earthquake potential, and it is compared with the frequency of historical earthquakes with the inland epicenter. Then the frequency of earthquakes in the future 1, 300 years is estimated, based on the idea that possibility of future earthquakes is high in the area where the earthquake potential has not been minimized sufficiently by historical earthquakes (Fig, 8)
In Japan, however, the damage caused by marine earthquakes has been very severe during the historical period. We must pay attention to such marine earthquakes particularly in coastal regions.