Study on estimation of dynamic faulting parameters for strong ground motion simulations

首先，我们改进了交错网格有限差分法中断层边界条件的实现，采用一个虚拟曲面来满足断层边界条件。在我们的实现中，速度（或位移）的网格设置在断层面上，应力网格从故障移动半个网格间距和破裂带的虚拟表面上的应力是这样的断层上的插值应力等于摩擦应力。与以前的版本相比，我们的实现有五个优点：（1）在破裂之前没有滑动的泄漏，以及（2）零厚度断层，（3）可靠地计算断层上的应力，（4）我们的实现适合于断层本构律的研究，因为滑动被定义为z =+0的平面上的位移与z =-0的平面上的位移之间的差，其次，研究了临界减滑距离（Dc）的估算方法， 关于我们 它是由Mikumo和Yagi [2003]提出的，用于最近地震的动态破裂模型。假设断层上的Dc均匀分布，模拟了断层的自发动态破裂过程，并生成了可在实际记录站观测到的合成波形。然后，对合成波形进行运动学反演，得到了各亚断层的滑动速率时间函数。我们估计Dc从这些功能，并讨论是否假设的Dc可以正确地恢复。我们还研究了破裂传播的影响，每个子故障与有限的尺寸和影响的波形反演和带通滤波过程中的估计直流。研究发现，传播效应可以使恢复的Dc0值与最终滑移量之间存在明显的相关性，这对阐明高频地震波的辐射机制，进行强震模拟和地震物理机制的研究具有重要意义。两种机制，即，提出了随机破裂过程和停止阶段。当前一种机制占主导地位时，高频波必须从粗糙体内部辐射。另一方面，在后一种机制的情况下，它必须从粗糙的边缘产生。在四次内陆地震的波形分析中，日本、三宅等。Al.（2003）发现，高频丰富的强运动产生区的大小和位置与粗糙体的大小和位置一致。然而，考虑到分析的分辨率，不可能区分哪种机制占主导地位。生成区域可以沿着粗糙体的边缘或者可以在粗糙体内部。通过计算机仿真，对该机制进行了讨论，特别是对哪种机制是有效的进行了讨论.由于地震波通常具有高频成分，有效机制比无效机制更合理。少

Firstly, we improve the implementation of the fault boundary conditions in the staggered grid finite-difference method by using a fictitious surface to satisfy the fault boundary conditions. In our implementation, velocity (or displacement) grids are set on the fault plane, stress grids are shifted half grid spacing from the fault and stress on the fictitious surface in the rupture zone is given such that the interpolated stress on the fault is equal to the frictional stress. Our implementation has five advantages over previous versions: (1) No leakage of the slip prior to rupture and (2) a zero thickness fault, (3) stress on the fault is reliably calculated, (4) our implementation is suitable for the study of fault constitutive laws, as slip is defined as the difference between displacement on the plane of z=+0 and that of z=-0, and (5) cessation of slip is achieved correctly.Secondly, we investigated the validity of a method for estimating the critical slip-weakening distance (Dc), w … More hich was proposed by Mikumo and Yagi [2003], for a dynamic rupture model of a recent earthquake. Assuming a uniform distribution of Dc on the fault, we simulated spontaneous dynamic rupture process and generated the synthetic waveforms that would be observed at actual recording stations. Then, we carried out kinematic inversion of the synthetic waveforms and obtained the slip-rate time functions on each subfault. We estimate Dc from these functions and discuss whether the assumed Dc could be recovered correctly. We also investigated the rupture propagation effect over each subfault with a finite dimension and the effects from the waveform inversion and band-pass filtering processes on the estimate of Dc. We found that the propagation effect could cause an apparent correlation between the recovered Dc0-values and the final slip.It is also important to clarify radiation mechanism of high frequency seismic waves for strong ground motion simulation and for study of physical mechanism of earthquake. Two mechanisms, i.e., stochastic rupture process and stopping phase, had been proposed. When the former mechanism is dominant, high frequency waves must be radiated from inside the asperity. On the other hand, in the case of the latter mechanism, it must be generated from edge of the asperity. In waveform analyses of four inland earthquakes, Japan, Miyake, et. al. (2003) found that the size and position of high frequency rich strong motion generation area coincide with those of asperities. Considering the resolution of the analyses, however, it is not possible to distinguish which mechanism is dominant. The generation area might be along the edge of the asperity or might be inside the asperity. We discussed the mechanism by using computer simulation, especially, which mechanism is efficient. Since seismic wave has usually high frequency components, effective mechanism is more plausible than ineffective one. Less