An integrated Bayesian approach to the probabilistic tsunami risk model for the location and magnitude of earthquakes: application to the eastern coast of the Korean Peninsula

Abstract

We explored the distributional changes in tsunami height along the eastern coast of the Korean Peninsula resulting from virtual and historical tsunami earthquakes. The results confirm significant distributional changes in tsunami height depending on the location and magnitude of earthquakes. We further developed a statistical model to jointly analyse tsunami heights from multiple events, considering the functional relationships; we estimated parameters conveying earthquake characteristics in a Weibull distribution, all within a Bayesian regression framework. We found the proposed model effective and informative for the estimation of tsunami hazard analysis from an earthquake of a given magnitude at a particular location. Specifically, several applications presented in this study showed that the proposed Bayesian approach has the advantage of conveying the uncertainty of the parameter estimates and its substantial effect on estimating tsunami risk.