Forecasting the Future Tsunami Potential of the Indo-Pacific Region Using Artificial Neural Network

Abstract

Tsunamis, one of the most devastating natural disasters on Earth, pose a substantial hazard and risk to coastal infrastructure and human life in the Indo-Pacific region. Tsunami hazard analysis is essential for safeguarding people and assets in coastal areas. However, a study on the comprehensive tsunami potential in the considered area needs further detailed investigation. For this purpose, the Indo-Pacific region (600N to -600S and 300E to1800E) is divided into eight tsunamigenic zones based on tsunamicity, physiography and seismotectonics. Tsunamigenic earthquake data from 1700–2022 is analysed using stochastic tools and Artificial Neural Network (ANN) algorithms to precisely forecast the time, magnitude, and location of the impending tsunamigenic earthquake. The study indicates that Zones 5,7 and 8 are highly vulnerable to tsunami occurrence between periods 2030–2033 (MW 8), 2030–2034 (MW 7) and 2031–2034 (MW 8.2), respectively. Whereas, Zones 1, 4 and 6 might turn vulnerable in the long run between periods 2051–2064 (MW 9.3), 2062–2081 (MW 7.3) and 2077–2120 (MW 7.7), respectively. It is also advocated that out of eight tsunamigenic zones considered, Zone 1 might experience the strongest tsunamigenic earthquake (MW 9.3) by the mid-century (2051–2064). Notably, the methodology is also validated through its successful application to already occurred events (the 2004 Indian Ocean Tsunami, 2011 Tohoku Tsunami and the recent Japan earthquake of January 01, 2024). The reliability of the methodology depends on the availability of a good amount of historical and instrumental tsunamigenic earthquake data.