5:15 PM - 6:30 PM
[HDS09-P04] Numerical simulations of tsunamis caused by the 2015 Illapel earthquake and the 2017 Chiapas earthquake using the heterogeneous slip models
Keywords:Simulations of tsunamis
The 2015 Illapel earthquake and the 2017 Chiapas earthquake caused tsunamis in the Pacific Ocean. Previous studies estimated the high-precision slip distributions of the 2015 Illapel and the 2017 Chiapas earthquakes using inversion techniques. The high-precision slip distributions generally improve the tsunami prediction accuracy. However, we cannot acquire the high-precision slip distributions in real time. This study aims to answer what is the precision in the slip models needed in early tsunami predictions. Therefore, for the Illapel and the Chiapas earthquakes, we calculated the tsunamis using the high-precision slip distribution and uniform slip models. Also, we compared them to the tsunami waveforms observed by the DART systems.
Estimation of the crustal movement due to the faulting adopted USGS fault solutions. We incorporated the horizontal movement effect of the seafloor slope into the vertical displacement and a filter based on the linear potential theory to estimate the initial sea surface displacement. We used the nonlinear shallow water equations for tsunami propagations. The equations further included frequency dispersion, the earth's elasticity, and seawater density stratification. The root-mean-square error evaluated prediction accuracy between the observed and the calculated tsunami waveforms.
In the Illapel tsunami modeling, the observed tsunami waveforms agree better with the calculated tsunami waveforms using the high-precision slip distribution. However, in the Chiapas tsunami modeling, the observed tsunami waveforms agree better with the calculated waveforms using the uniform slip model. The amplitude of the calculated Chiapas tsunami waveforms using the high-precision slip model is almost double of the observed amplitude. Therefore, we calculated the Chiapas tsunamis again using the high-precision slip distribution in which the slip amount decreased half. That resulted in the root-mean-square error decreased much. The decreasing slip amounts to be half corresponds the earthquake magnitude smaller by 0.2. We conclude that accurate tsunami modeling requires the high-precision slip distribution and correct estimation of the earthquake magnitude.
Estimation of the crustal movement due to the faulting adopted USGS fault solutions. We incorporated the horizontal movement effect of the seafloor slope into the vertical displacement and a filter based on the linear potential theory to estimate the initial sea surface displacement. We used the nonlinear shallow water equations for tsunami propagations. The equations further included frequency dispersion, the earth's elasticity, and seawater density stratification. The root-mean-square error evaluated prediction accuracy between the observed and the calculated tsunami waveforms.
In the Illapel tsunami modeling, the observed tsunami waveforms agree better with the calculated tsunami waveforms using the high-precision slip distribution. However, in the Chiapas tsunami modeling, the observed tsunami waveforms agree better with the calculated waveforms using the uniform slip model. The amplitude of the calculated Chiapas tsunami waveforms using the high-precision slip model is almost double of the observed amplitude. Therefore, we calculated the Chiapas tsunamis again using the high-precision slip distribution in which the slip amount decreased half. That resulted in the root-mean-square error decreased much. The decreasing slip amounts to be half corresponds the earthquake magnitude smaller by 0.2. We conclude that accurate tsunami modeling requires the high-precision slip distribution and correct estimation of the earthquake magnitude.