The 2024 SSJ Fall Meeting
Oct 21 - Oct 23, 2024Toki Messe
The 2024 SSJ Fall Meeting
Oct 21 - Oct 23, 2024Toki Messe
[S20-05][Invited]Prediction of an Enigmatic Tsunami in October 2023 at Kii Peninsula, Japan
*Yuchen WANG1, Kentaro Imai1, Yutaka Hayashi2, Hiroki Horikawa1(1. Japan Agency for Marine-Earth Science and Technology, 2. Meteorological Research Institute)
On October 8, 2023, a tsunami was observed along the Pacific coast of Japan, but its cause was complicated. We applied tsunami data assimilation, a forecast method that does not require wave source information, to predict coastal tsunami waveforms in the Kii Peninsula. We assimilated offshore observations recorded by DONET and reconstructed the wavefield before the tsunami arrived at the coast. The waveforms at Kumano, Kushimoto, and Shirahama were accurately forecasted at least 20 min before the tsunami arrival, indicating that data assimilation is applicable for tsunami early warning. Quantitative analysis showed that the score and accuracy index generally increased after 21:00 (UTC).
Meanwhile, we investigated the tsunami decay process for tsunami warning cancellation by adopting the moving root mean squared (MRMS) amplitude. As data assimilation progressed, the prediction of tsunami later phase became increasingly accurate, and the errors between observed and predicted MRMS amplitudes decreased. Overall, the tsunami later phase was satisfactorily predicted at approximately 22:00 at coastal tide gauges. Hence, this data assimilation approach contributes to a comprehensive tsunami early warning process, from issuance to cancellation.
Meanwhile, we investigated the tsunami decay process for tsunami warning cancellation by adopting the moving root mean squared (MRMS) amplitude. As data assimilation progressed, the prediction of tsunami later phase became increasingly accurate, and the errors between observed and predicted MRMS amplitudes decreased. Overall, the tsunami later phase was satisfactorily predicted at approximately 22:00 at coastal tide gauges. Hence, this data assimilation approach contributes to a comprehensive tsunami early warning process, from issuance to cancellation.