[M-IS09] Interdisciplinary studies on pre-earthquake processes

[MIS09-P12] Creation of AI disaster mitigation sustainable studies

*服部 克巳^1,2,3、Josaphat Tetuko Sri Sumantyo^2,3 (1.千葉大学大学院理学研究院、2.千葉大学環境リモートセンシング研究センター、3.千葉大学災害治療学研究所)

For weather disasters such as torrential rains and typhoons, information on rainfall amounts, wind speeds, etc. is disclosed in advance, albeit with some uncertainty, to encourage society to respond to such disasters, and society has developed countermeasures in advance, taking into account the uncertainty of the forecasts. On the other hand, all information on earthquakes and tsunamis is currently available only after they occur, with little or no advance information available. Even in the case of a massive earthquake event of magnitude 7 or greater, it is expected that individuals and companies will be able to smoothly prepare in advance and take actions to mitigate damage if they have predictive (advance) information at normal times, one week, three days, one day, several hours, one hour, just before the arrival of seismic waves (Earthquake Early Warning), and 20 minutes before the arrival of tsunami It is also expected to facilitate individuals and businesses to prepare in advance and take actions to mitigate damage. In addition, as seen in the case of the Boso typhoon and Noto earthquake, there is a lack of information on the disaster during and immediately after the disaster (access to the disaster area by roads, ports, etc., and the situation of hospitals and other emergency response centers). Especially at night or when bad weather, power outages, the spread of infectious diseases, or nuclear accidents occur, information is even more scarce, making it difficult to conduct appropriate disaster response activities. In addition to conventional meteorological information and earthquake early warning systems, it would be expected to further reduce economic and human losses if accurate forecasting and disaster monitoring information could be utilized. This led to the conception of a world-first attempt to develop a system for extracting information from earth observation big data that contributes to disaster forecasting and monitoring and reconstruction assistance, and a crisis response navigator that uses this information, which will be refined and upgraded on a spatial AI platform and implemented in society. The creation of an AI disaster mitigation sustainable science that contributes to the realization of a disaster resilient society through collaboration and fusion of various fields of mathematical sciences such as remote sensing, disaster science, meteorology, and geophysics, social sciences that determine behavioral patterns, and various fields of disaster medicine and emergency response on the frontlines.