Prompt Tsunami Loss Estimation Using Satellite Imagery

Abstract

This study explores the development of prompt tsunami loss estimation methods that are based on inundation hazard parameters inferred from remotely-sensed images. Using a tsunami catastrophe model for the Tohoku region of Japan, approximate tsunami loss functions are derived for coastal communities in Miyagi Prefecture. Calibration of the prompt tsunami loss estimation models is carried out by utilizing predicted tsunami losses for 4000 stochastic earthquake scenarios of moment magnitudes between Mw7.5 and Mw9.1. Using numerous earthquake sources facilitates the robust calibration of the developed tsunami loss functions. Special considerations are given to investigate the effects of coastal topography and the potential bias due to errors in estimating inundation parameters.