APPLICATION OF PALEOSEISMOLOGICAL DATA FOR SEISMIC HAZARD REDUCTION IN JAPAN

It was only after the devastating 1995 Kobe earthquake that the Japanese public and governments became aware of seismic risks of onshore active faults for hazard reduction. Since 1995, the Headquarters for Earthquake Research Promotion have worked to promote, review, and publicize research on seismic risks in Japan. Intensive studies have been conducted in more than 100 fault zones with grants from the Headquarters. The Headquarters have already evaluated 33 fault zones to officially announce seismic potential of each fault zone. At the same time, seismologists have improved the methods for probabilistic risk assessment and strong motion forecasting to cook the geological data. The massive efforts, however, seem still short for comprehensive results for the public preparedness while the 10th anniversary of the Kobe earthquake is approaching. This paper aims to analyze the difficulties for the evaluation and to seek for better way to utilize paleoseismological data. The Itoigawa-Shizuoka tectonic line fault system (ISTL) in Central Japan, for example, has been regarded as the most dangerous faults on land in Japan since the Headquarters' first official evaluation in 1996. In the evaluation, empirical relationship between slip and magnitude concluded an unusually large model earthquake of M8.0+/-0.5 while paleoseismological data suggested much smaller one. Likewise, in many cases when empirical estimation contradicts geological fact, and geological constraints were not good enough, empirical estimation was chosen. Such empirical models, however, lack reality as most hazardous events in 1990s in the world were beyond our ordinary models. Recent geological studies on the ISTL and other faults confirm the occurrence of out-of-average earthquakes in the past, but the data are not easy to use. Paleoseismological data by its nature have considerable range of uncertainty. We need to manage the uncertainty to model realistic earthquakes.