Item

Abstract

This study discusses the spatio-temporal evolution of tsunami source process in real-time tsunami forecasting. Previously, we developed a tsunami forecasting method based on the real-time source inversion using offshore tsunami waveforms. In the method, the source process is assumed to finish instantaneously at the time of an earthquake, whereas the source process of actual huge earthquakes is spatio-temporally complex. The offshore tsunami data from the dense seafloor observation networks around Japan may enhance to constrain the spatio-temporal evolution. To design the improved forecasting method that is versatile for tsunamigenic earthquakes with various spatio-temporal source processes, this study conducted synthetic tests of tsunami forecasting, focusing on how well the offshore tsunami waveforms can constrain the time evolution of the source process. Synthetic data for ~170 offshore tsunami-meters along the Japan trench were produced by numerical tsunami simulation assuming the fault model of the 2011 Mw 9.0 Tohoku earthquake. Synthetic tests were conducted under various conditions by independently varying the source duration assumed in production of the synthetic data and the that used in the inversion. The results showed two suggestions for accurate forecasting of coastal tsunami heights. The first suggestion is that it is important to avoid situations that source duration used in the inversion is much shorter than that of the targeted earthquake. The second suggestion is that using a very long source duration for the inversion is effective for both the long- and short-source-duration earthquakes.This work was partially supported by JSPS KAKENHI Grant Numbers JP21K04621.