Study on strong ground motion generated from dynamic rupture model

在近断层强地震动模拟中，需要构造物理上合理的震源模型。中村和宫武（2000）提出了基于动态破裂模型的滑动速率时间函数的近似表达式，并给出了如何利用滑动分布或运动学信息来估计上述公式的参数。Miyatake等人（2003）将其应用于具有多个微凸体的典型但简化的断层模型，并讨论了如何确定近似表达式的参数。在本文中，为了评估我们的近似函数，我们将其应用于过去的地震，2001年Geiyo地震，以评估我们的方法。近似时间函数与动力学模型的时间函数比较，结果令人满意。由近似模型得到的近场台站波形与动力学模型得到的波形吻合得很好。我们的功能的基础上的动力学模型将有助于准确模拟从一个可能的破裂情况。我们还研究了地震的动态破裂过程，并估计了动态参数。首先，研究了地震滑动的时空分布，通过近场强地面运动数据和地震波形数据的反演，研究了地震的运动学模型。其次，利用上述运动学模型，重建了一个有约束的动力学模型。Dc的分布也是根据反演的滑移率时间函数来估计的（Mikumo等人，2003年），发现小于70厘米。我们发现，动态参数是非常异质性。存在两个粗糙体。第一个微凸体上的峰值应力降约为20MPa，第二个微凸体上的峰值应力降约为10MPa。在两个微凸体之间存在很低的应力降区或负应力降区。强度过剩和断裂能分布也是不均匀的。

In near fault strong ground motion simulation, it is desired to construct physically reasonable source modeling. Nakamura and Miyatake (2000) have proposed an approximate expression of slip rate time function based on dynamic rupture model.We also showed how to estimate the parameters of the above formula by using the slip-distribution or the kinematic information. Miyatake et al. (2003) have applied to the typical but simplified fault models with several asperities and discussed how to fix the parameters of the approximate expression. In this manuscript, in order to evaluate our approximate function, we apply it to the past earthquake, the 2001 Geiyo earthquake in order to evaluate our method. Comparison between the approximate time functions and those of dynamic model are satisfactory. The waveforms at nearfield stations generated from our approximate modeling fit those from dynamic model very well. Our functions based on dynamic model will help accurate simulation from a probable rupture scenario. We also investigate a dynamic rupture process of the earthquake and estimate the dynamic parameters. Firstly, the space-time distribution of slip of the earthquake, i.e., the kinematic model, was investigated by inverting the nearfield strong ground motion data and teleseismic waveform data. Secondly, we reconstruct a dynamic model in constraint with the above kinematic model. The distribution of Dc is also estimated from the inverted slip rate time functions (Mikumo et al., 2003) and found to be less than 70cm. We found that dynamic parameters are very heterogeneous. There exists two asperities. The peak stress drop on the first asperity is about 20MPa, and 10MPa on the second one. Very low or negative stress drop zone is found between two asperities. The strength excess, and fracture energy distribution are also heterogeneous.