ABSTRACT
Song, M.-J.; Kim, B.-H., and Cho, Y.-S., 2022. Estimation of maximum tsunami heights using probabilistic modeling: Bayesian inference and Bayesian neural networks. Journal of Coastal Research, 38(3), 548–556. Coconut Creek (Florida), ISSN 0749-0208.
Although tsunamis may occur relatively infrequently, they can cause substantial loss of life and property damage. Most studies on tsunamis have focused on developing numerical models to describe ocean propagation and the associated run-up process because laboratory experiments are expensive. Probabilistic techniques have been used in various scientific and engineering fields with reasonable results. In this study, numerical simulations were performed for historical and virtual tsunami events to explore maximum tsunami heights. Bayesian inference and Bayesian neural networks were then employed to predict the maximum tsunami height. These results can be used to develop hazard maps for unexpected tsunami events.
