講演情報
[S25-04]Data assimilation for tsunami warning issuance and cancelation during the 2025 Mw 8.8 Kamchatka earthquake
*王 宇晨1、今井 健太郎1 (1. 海洋研究開発機構)
On 31 July 2025, at 08:24:52 JST, an Mw 8.8 megathrust earthquake struck off the eastern coast of the Kamchatka Peninsula. The earthquake generated a large tsunami, which arrived in Japan approximately 2 h after the earthquake and were observed in 22 prefectures.Along the Pacific coast of Japan, offshore observation networks have been established: S-net, DONET, and N-net. These networks contain bottom pressure sensors that can record water pressure in real time, enabling the detection of tsunamis before they reach the coast. The offshore records can be applied in the tsunami data assimilation approach to predict coastal tsunami waveforms. This method has already been tested in synthetic experiments (Maeda et al., 2015) and in moderate-scale tsunamis (e.g., 2016 Mw 6.9 Fukushima earthquake; Wang and Satake, 2021).The 2025 Kamchatka earthquake tsunami was the largest event since the completion of S-net. As a preliminary study, we selected Aomori, Iwate, and Miyagi Prefectures as the study area and used a part of S-net sensors for data assimilation. We predicted the arrival time and amplitude to provide reference for warning issuance, and we also predicted the attenuation process to provide reference for warning cancelation.Our results showed that data assimilation makes accurate predictions of arrival time and amplitude at least 30 min before the tsunami reaches the coast by assimilating S-net records. Particularly, it is able to accurately predict the 1.3 m tsunami recorded at the Kuji station. Meanwhile, data assimilation can continuously update predictions after the tsunami has reached the coast, forecasting the attenuation of following waves. This facilitates more timely and precise warning cancelation, providing support for disaster management.