In this paper, we conducted a study of a tsunami inundation forecasting system with a case study in Owase City, Japan. In this study, a database that consists of precomputed tsunami inundation and waveforms from multiple scenarios is developed. The system is divided into two stages. In the first stage, preliminary earthquake information is used to find the appropriate tsunami inundation scenario in the database. In the second stage, a real-time tsunami waveform simulation is conducted to find the best-case scenario by minimizing the error between computed tsunami waveforms and those in database. Furthermore, this method is able produce good tsunami inundation forecasts in a reliable time.
The 2011 Tohoku earthquake left more than 15,000 people dead or missing (Kazama and Noda, 2012). It was the biggest earthquake followed by a tsunami ever recorded in Japanese history. Tsunami warnings and advisories were issued for areas along the coast of Hokkaido to Kyushu and the Ogasawara Islands (Ozaki, 2011). The Japan Meteorological Agency (JMA) estimated that the initial earthquake magnitude was Mjma 7.9, obtained within three minutes after the earthquake. Then, it was revised to be Mjma 8.4 in more than an hour after the earthquake (Ohta et al., 2012). Further study revealed that those magnitudes underestimated the actual earthquake magnitude of Mw 9.0 (e.g., Gusman et al., 2012; Satake et al., 2013).
A future Nankai Trough earthquake located in the southeast region of Japan is expected to occur in the near future and could be possibly more destructive than the 2011 Tohoku earthquake. On the basis of historical records, Ishibashi (2004) explained that a great earthquake has a recurrence interval of 100–200 years. Compared with the other earthquake zones in Japan, the Nankai Trough is located very close to the coast (less than 150 km) (Mulia et al., 2017); if an earthquake followed by a tsunami occurred, a resulting tsunami wave would need only a short time to reach the coast.
