[H-DS10] Tsunami and tsunami forecast

[HDS10-P04] Realtime tsunami damage prediction system using oceanfloor network system

*Narumi Takahashi^1,2, Kentaro Imai² (1.National Research Institute for Earth Science and Disaster Resilience, 2.Japan Agency for Marine-Earth Science and Technology)

Keywords:realtime tsunami prediction, oceanfloor network system, tsunami debris

We have developed realtime tsunami prediction system using oceanfloor network system since 2012 (Takahashi et al., 2017; 2018; 2021). The contents of the prediction are tsunami arrival times, the heights, the inundation area and the depth distribution. The system using the Dense Oceanfloor Network system for Earthquakes and Tsunamis (DONET) is installed on Wakayama and Mie prefectures, Chubu electric power and Kagawa University and implemented by themselves. On Chiba prefecture, we constructed the similar system using Seafloor observation Network for Earthquakes and Tsunami along the Japan trench (S-net) and users specified within disaster division people of the local government can receive the prediction using mobile phone. The concept is based on tsunami prediction for each area evaluated using oceanfloor networks. Even if the tsunami source is non-seismic and seafloor collapses, the system is worked when the tsunami is some observatories of the oceanfloor networks. We developed it and constructed newly tsunami damage prediction system for acceleration of the reconstruction after disaster. We considered that one of the important elements for the acceleration are tsunami debris. Tsunami debris prevents the support from others and access to the areas, and cause tsunami fire. Therefore, we focus on the tsunami debris and challenged prediction of the debris behavior. Tsunami debris is brought by fluid force of the inundated sea water, and the inundated volume leads to the evaluation of the debris behavior. We introduced two methods to evaluate the tsunami debris occurrence. The former calculates motion equation working on each house, and judges the collapse (Kozono et al., 2016; 2017). This method also evaluates move of cars and ships. And behavior of each produced debris after the collapse are calculated. Such iteration is repeated, and we get final location of each debris. The later evaluates the wooden debris distribution and the thickness based on the damage functions obtained by 2011 off Tohoku tsunami disaster (Imai et al., 2019; 2022). Concentrated wooden debris, cars and ships could be sources of tsunami fire. When evacuation place is surrounded by such debris, and it means that there is high risks to be damaged by tsunami fire. The debris distribution could be one of materials to evaluate health of infrastructure and emergency transportation. Drift of much debris to the sea means that it is difficult to access from the sea to support for there. Above realtime tsunami prediction system using DONET has a database of tsunami waveforms calculated using 1506 fault models changing the magnitude, depth, dip and location. We calculated the debris behavior using much fault models and obtained that 310 models with the inundated area of over 2 km² brings debris there. Therefore, we calculated the debris behavior using 310 fault models for six hours from arrival of tsunami, and added a function of the tsunami damage prediction on current realtime tsunami prediction system.