HR: 0830h
AN: S51E-0091    [PDF]
TI: Should We Consider Sea in Simulating Strong Ground Motion? - Numerical Examination of Effects of Sea in the 2-D P-SV Wave Field -
AU: * Hatayama, K
EM: hatayama@fri.go.jp
AF: National Research Institute of Fire and Disaster, 3-14-1, Nakahara, Mitaka, 181-8633 Japan
AB: I investigated the effects of the sea on seismic ground motion through some numerical experiments and examined the necessity of considering the sea in simulating strong ground motion. To specify the effects of the sea, I compared the waveforms calculated for the models with and without the sea. The depth of the sea examined here ranges from several ten to several hundred meters. The deep sea several hundred meters deep exists over the source regions of potential big earthquakes anticipated along the Pacific Ocean trenches off Japan. The shallow sea several ten meters deep is located not only over the landward margins of the above source regions but also around cities with large population. Though such sea may influence ground motion, it is usual to ignore the sea in the numerical simulation of ground motion. This study is motivated by the question whether ignoring the sea is valid. Examined in this study are two-dimensionally simple subsurface structure models. The models are composed of a sea part and a land part. Both parts are horizontally stratified layers overlying elastic half-space and as regards the sea part the medium of the uppermost layer is water, which is regarded as the ideal fluid here. For such models, I solved point source impulse responses of the two-dimensional P-SV wave field in the frequency domain by using the direct boundary element method (BEM) and velocity waveforms are calculated by convolution with Ricker wavelets. The BEM here relies on the Green function for horizontally layered media and this enables to calculate Rayleigh waves alone as well as the total waves. Because of the shallow sea included in this examination, the impulse responses were calculated up to a high frequency of 16 Hz. This involved small boundary elements and I had to devise how to accelerate convergence of wave-number integrands of the Green function, by calculating zero-frequency components separately. Comparing the waveforms from the models with and without the sea, I found the following effects of the sea to date: (1) The Rayleigh waves are strongly influenced by the sea; (2) As the sea is deeper, it can influence the Rayleigh waves to longer periods; the intense effects of the sea with depths of 50, 200, 400 and 800 m appear up to periods around 0.2, 0.8, 1.5 and 3 s, respectively, although these periods are for the simple case of the sea overlying the bedrock with an S-wave velocity of 3 km/s and they may depend on velocity structures beneath the sea bottom; (3) Substituting sediments for the sea water, which is usual in ground motion simulation, can have bad influence particularly on vertical components and we would sooner replace the sea water by vacuum than substitute sediments for the sea water. The above findings strongly suggest the necessity of considering the sea in ground motion simulation and the effects of the sea on the three-dimensional wave field still remain to be evaluated.
DE: 7212 Earthquake ground motions and engineering
DE: 7223 Seismic hazard assessment and prediction
DE: 7260 Theory and modeling
SC: Seismology [S]
MN: 2003 Fall Meeting