基于虚拟地震的各类地震强地面运动模拟方法研究

Accurate prediction of strong ground motions of future catastrophic earthquakes is crucial for assessment and mitigation of seismic hazards. Current simulation methods rely strongly on the accuracy of the input 3D model, including both velocity and attenuation; both (especially the attenuation structure) are difficult to obtain in reality. Therefore, it is imperative to develop new method to simulate and predict strong ground motions for earthquake prone regions. An innovative technique for strong ground motion prediction, known as the virtual earthquake approach, recently developed by the Stanford group, is quite promising for future seismic hazard assessment. However, there is an urgent need to improve the reliability of Green’s functions from ambient seismic noise and compare the prediction with the realistic strong ground motions records of strong earthquakes (~ Mw 7). This project aims to significantly improve the virtual earthquake approach by quantifying the key factors that determining the accuracy of its prediction, and extend the novel approach, which was developed for major strike slip earthquakes, to the case of thrust and normal faulting events. Then, we will use strong motion records at the seismic stations adjacent to the Longmenshan fault of the Lushan earthquake to calibrate our simulation of strong ground motion based on the virtual earthquake approach. Through the calibration, we intend to quantify the effect of source mechanism, source rupture process, topography site-effect and basin amplification on strong ground motion with the new virtual earthquake approach.

准确的强地面运动模拟是评估和减轻地震灾害的有效途径。三维地球介质模型（波速和衰减）的不确定性是制约强地面运动模拟精度的主要因素，因此亟需发展新的强地面运动模拟方法。斯坦福大学提出的基于虚拟地震的强地面运动模拟新型方法，对于震灾评估等具有良好的应用前景，但是也存在着提取的背景噪声格林函数可靠性低、缺乏与大地震（~ Mw 7）实际强地面运动观测的定量对比等问题。本项目拟进一步发展和完善虚拟地震方法，系统分析影响虚拟地震方法精度的关键因素，发展适用于正断层和逆冲断层等震源类型的强地面运动计算方法，获得不同周期的地震动峰值速度空间分布，并和芦山地震等大地震的实际强地面运动观测数据进行定量对比。在此基础上，揭示复杂场地条件下震源破裂过程、山地效应、盆地效应等因素相互耦合的强地面运动分布规律。

本项目的研究目标是发展和完善长周期地震动计算方法，定量表征山地效应、盆地效应等因素耦合作用对地震动的影响。项目取得的主要成果和认识包括：（1）获得了盆地内部近垂直入射剪切波和面波的长周期地震动放大特征。由于激发方式和传播路径的差异，面波的地震动放大特征与剪切波存在显著差异，面波峰值放大系数约为剪切波的2倍。面波放大系数的空间分布和地质构造密切相关，中央松辽盆地和东边的三江盆地内部放大系数要显著高于周边的山区地带。山区地带的放大系数普遍在1.5以内，而盆地内部相对于参考台站的最大放大系数接近8，平均放大系数约为4。松辽盆地北部地区由于沉积层厚度较大，面波放大系数也要显著大于盆地南部；（2）联合采用频散曲线、面波振幅比和远震体波信号反演沉积盆地浅层速度结构，获得了盆地高精度横波速度模型，浅层速度模型与声波测井速度具有比较好的一致性；（3）利用反演获得的速度模型，获得了面波和剪切波长周期地震动放大系数，成功解释了观测得到的面波和剪切波的差异放大特征，揭示了沉积盆地内部复杂场地条件下长周期地震动分布规律。. 综合上述结果，本研究为发展了一套简便易行的盆地内部长周期地震动放大效应定量评估方法，为探讨不同地震波类型在盆地长周期地震动放大效应的作用和机制奠定了基础，对于研究复杂场地条件下山地效应和盆地效应等多因素耦合作用对地震动的影响，提高地震运动放大系数的估算精度，定量评估盆地内部潜在的地震灾害等方面具有重要意义，能够为盆地城市未来的防震减灾规划提供重要参考。