抄録
It is important to advance preparation for a tsunami disaster, one of the great concerns in Japan. Forecasting tsunami inundation is one such solution, which contributes to perceiving the danger of the tsunami, as the inundation is directly linked with the damage. Therefore, we developed a new real-time tsunami forecast system, aimed at rapidly and accurately forecasting tsunami inundation on land, based on offshore tsunami data observed by the seafloor observation network along the Japan Trench, S-net. The developed system takes a database approach. A database called a tsunami scenario bank was constructed by assuming all the possible tsunami sources affecting the target region and simulating offshore pressure data, coastal tsunami heights, and tsunami inundation. The forecast system searches for suitable tsunami scenarios whose offshore pressure data explain the observed data, based on the multi-index method. The multi-index method can evaluate the resemblance of offshore pressure data by using three indices, which are sensitive to different aspects of the pressure change distribution. When tsunami scenarios meet the criteria of the multi-index method, the system provides forecast information generated from coastal tsunami heights and tsunami inundation of the selected scenarios. A prototype system was constructed for the Pacific coastal region of Chiba prefecture as a target region and has been updated through a test operation. We also investigated the comprehensible visualization and effective disaster response using tsunami forecast information. Through workshops and tabletop exercises with local government officers using the forecast system, timelines and local disaster management plans for tsunamis were tested and revised. This led to the establishment of a standard operating procedure for tsunami disaster response through the use of tsunami observation and forecast information.
