| 英文摘要 |
In this study, we used three different seismic source models, the deterministic Haskell model, the stochastic ω-square model and the hybrid composite source model, to simulate the strong motion records of the 1994 NANAO earthquake. We compared the calculated ground motion with the observed data from the TSMIP. Under the condition of not considering the local topology effect and using a simplified layer velocity structure , we obtain the following results:
(1) For a vertical strike-slip fault, the stations located on the parallel side of the slip , the transvers component ground motions show faster attenuation than the radial component motions. This anisotropic decay role may be due to the rupture direction of the seismic source mechanism. (2) The calculated results using the deterministic Haskell model fit with the observation at low frequency range, but fail at high frequency range. This resault may due to we adopted a homogeneous layer velocity structure and simplified the rupture process to be linear bilaterally. (3) After comparing with the observed ground motion records, the predicted spectra using ω-square model show in good agreement with the observations for the frequency range greater than 1 Hz. The disadvantage of using the stochastic ω square model is that it can only produce a single component time series without seismic wave phase infomation. (4) The synthetic result of using the composite source model not only show a good agreement with the observation and Haskell model at low frequency range, but also fit with observation and ω square model at higher frequency range. This result shows that the calculation ability using the composite source model can cover whole frequency range. (5) Through this study, we observed that the local structure effect such as topography effect and the basin effect can be an important issue. Analytical studies on the causes of these effects will give helpful insight to the ground motion simulation. |
