Strong Motion Estimation for Active Faults with High Earthquake Potential

我们的主要目标是通过评估活动断层的潜在危险来获得准确的长期地震预测，阐明断层动力学的物理原理，并通过计算和观测方法开发精确的地面运动模拟的数值和半经验方法。最终目标是建立预测假想活动断层地震的强烈地面运动的方法，并减轻未来大地震造成的地震破坏。本文开展了以下六个方面的研究工作：(1)对中构造活动断裂系统（MTL）的地震危险性进行了航拍解译和古地震沟测。基于地震反射剖面和钻孔资料，利用平衡剖面分析得到了震源断层的几何形态和平均滑动速率。此外，从近年来地震断层的破裂形态出发，研究了断层的分支和破裂传播的指向性。(2)利用BIEM研究了动态裂纹扩展的阻止机制，并与活动断层的表面几何形态进行了比较。采用带动态滑动率时间函数的近场表达式理论计算了地震断裂带附近的强脉冲。结合低频波形和高频包络反演结果，利用强地震动和宽带记录研究了震源非均质性。利用高质量的台网资料对内陆地壳地震和板块内大地震进行了震源表征。(3)利用折射和重力数据联合反演，实现了地下构造的精确三维建模。沉积盆地和浅层土层的场地效应可以通过微震阵列测量得到的s波剖面得到很好的估计，并通过地震记录的直接放大评估得到验证。(4)采用间接边界元法和平面半解析解的混合格式，建立了非规则界面传播介质的数值计算方法。通过对1997年鹿儿岛-北武地震震源区的观测和数值模拟，证实了地震各向异性。(5)对神户-阪神地区的三维地下构造进行了折射、反射和重力测量。为研究强地震动与结构损伤之间的关系，对重灾区及其周边地区进行了强地震动模拟。对俯冲和板块地震的观测和数值方法进行了考虑震源效应的地震动估计以及复杂构造引起的宏观和微观地震反应的研究。(6)通过与1995年兵库县南部地震和1994年北岭地震的观测资料和震害分布对比，建立了未来大地震的强震预报方法，并进行了验证。采用混合模拟方法对1948年福井地震的强地面运动进行了估计。利用波形反演对滑动分布进行统计分析，研究了内陆大地震的震源特征。将上述六个课题的结果结合起来，建立了假设活动断层地震的强地面运动估计方法。基于对地震断层的地质调查和震源模型的地震学研究，我们总结了目前用于描述地震破裂模型的程序。在此基础上，提出了一种预测由活动断层引起的情景地震的强地震动的方法。少

Our major objectives are to obtain accurate long-term earthquake prediction by assessing hazard potential of active faults, to clarify physics of the faulting dynamics and to develop the numerical as well as semi-empirical methods for precise ground-motion modeling by computational and observational approaches. Final target is to establish the methodology of predicting strong ground motions for hypothetical active-fault earthquakes and to mitigate earthquake damage from future large earthquakes.We have been developing the above six subjects as follows : (1) Aerial photograph interpretation and paleoseismic trenching were made to date late Holocene earthquakes along the active faults for the seismic hazard evaluation on the Median Tectonic active fault system (MTL). The geometry and average slip rate of seismic source faults were obtained using the analysis of balanced cross sections based on seismic reflection profilings and boring data in some representative areas, Shinjo and Tomakoma … More i. Fault branching and directivity of rupture propagation were studied from rupture patterns of recent earthquake faultings. (2) The arresting mechanism of dynamic crack growth was studied by BIEM and compared with surface geometry of active faults. Strong pulses near the earthquake faults were calculated by near-field expression theory with a dynamic slip-rate time function. Source heterogeneity was studied by combining the results of low-frequency waveform and high frequency envelope inversions using strong ground motion and broadband recordings. Source characterizations of inland crustal earthquakes and large intra-plate earthquake were done using high-quality network data. (3) Accurate 3-D modelings of underground structures are done by joint inversion of refraction and gravity data. Site effects for sedimentary basins and shallow soil layers are well estimated from S-wave profiles derived by microtrmor array measurements, validated by direct evaluations of amplifications from earthquake records. (4) Numerical methods for propagation media with irregular interfaces are developed as a hybrid scheme with the indirect boundary element method and semi-analytical solution for flat interfaces. Seismic anisotropy are confirmed by some observation and numerical simulation in the source region of the 1997 Kagosima-hokuseibu earthquakes. (5) The results of refraction, reflection, and gravity surveys were gathered for deriving the three-dimensional underground structure in the Kobe-Hanshin area. Simulations of strong ground motions in and around heavily-damaged areas were done for examining the relation between strong ground motions and structure damage. Ground motion estimations considering source effects as well as seismic responses due to complex structures in both macro and micro scales were studied in observation and numerical techniques for the subduction and slab earthquakes. (6) Methods of strong motion prediction for future large earthquakes were developed and validated by comparing synthesized results with the observed data and damage distribution during the 1995 Hyogo-ken Nanbu and the 1994 Northridge earthquakes. Estimation of strong ground motions during the 1948 Fukui earthquake was done by hybrid simulation method. The source characterization for large inland earthquakes has been studied from statistical analyses of slip distributions with the waveform inversion.The results of the above six topics was combined to establish the methodology of estimating strong ground motions for hypothetical active-fault earthquakes. We summarize the procedure that is currently used to characterize earthquake rupture models, based on geological investigations of capable earthquake faults and seismological studies of source models. Then we propose a recipe of predicting strong ground motion form scenario earthquakes caused by active faults. Less