概率经验格林函数法和芦山Ms7.0地震强地面运动合成

发布时间：2018-03-26 12:33

本文选题：地震动模拟　切入点：经验格林函数　出处：《中国地震局地球物理研究所》2015年硕士论文

【摘要】：近场强地面运动研究既是探寻震源特性及传播介质信息的重要手段,也是确定某一地区地震动参数的衰减特征,并据此为抗震设计提供地震动参数的有效途径。目前有工程意义的近场强震记录数量仍然有限,但不断新建的核电站、大型堤坝、超高层建筑等特殊设防类工程,其设计地震动参数对加速度反应谱及加速度时间过程又有明确要求。因此,开展大地震近场强地面运动数值模拟有潜在的应用价值及重要的科学意义。经验格林函数方法是常用的地震动合成方法之一,其基本思想是用大震的前震或者余震作为经验格林函数合成大震记录。目前在应用经验格林函数法模拟近场强地面运动时,通常假定所使用的有关参数都是确定的。而实际上,震源参数如大小地震断层的长和宽、上升时间、断层面破裂的位错时间函数、断层的起始破裂点、剪切波速、破裂传播速度等,都有较大的不确定性,且这些参数的不确定性对模拟结果有显著影响。本文考虑剪切波速、小震上升时间、破裂传播速度及大小地震断层长(宽)之比等参数不确定性对基于经验格林函数法的强地面运动模拟结果的影响,将概率的方法应用于经验格林函数法的模拟计算中,确定这四种参数的可能取值及其对应的概率,以逻辑树的方式组合参数的可能取值及对应的概率,计算在不同情况下(参数组合)大震强地面运动的模拟结果及对应的概率,用统计的方法寻得最优解,即为概率经验格林函数法。论文开展了2013年4月20日芦山地震(Ms=7.0)近场强地面运动模拟,模拟结果表明：由概率经验格林函数法给出的峰值加速度、加速度时程、反应谱与实际大震记录符合的比较好,从而验证了该方法的可行性。另外,还分别研究了剪切波速、小震上升时间、破裂传播速度及大小地震断层长(宽)之比等每种参数的不确定性对2013年4月20日芦山地震(Ms=7.0)的近场强地面运动模拟结果的影响程度。采用控制变量法研究了每个参数的不确定性对模拟结果的影响程度,结果表明：小震上升时间的不确定对模拟结果的影响不大；剪切波速、破裂传播速度、大小地震长(宽)之比的不确定性对模拟结果影响明显,其中大小地震长(宽)之比的不确定性对模拟结果的影响最为明显。此外还研究了不同参数组合下模拟结果的实际误差与估计误差的分布情况,结果表明：使用考虑不确定性参数的概率经验格林函数法能减小模拟结果的误差。
[Abstract]:The study of near-field ground motion is not only an important means to search for the source characteristics and the information of transmitting medium, but also the attenuation characteristic of determining the ground motion parameters in a certain area. Based on this, the paper provides an effective way for seismic design to provide ground motion parameters. At present, the number of near-field strong earthquake records of engineering significance is still limited, but new nuclear power plants, large dams, super high-rise buildings and other special fortification projects are constantly being built. The design parameters of ground motion have definite requirements for acceleration response spectrum and acceleration time process. Numerical simulation of near-field ground motion of large earthquakes has potential application value and important scientific significance. Empirical Green's function method is one of the commonly used ground motion synthesis methods. The basic idea is to use the foreshocks or aftershocks of large earthquakes as empirical Green's function to synthesize the records of large earthquakes. At present, when the empirical Green's function is used to simulate the ground motion near the field strength, it is usually assumed that the relevant parameters used are determined. The source parameters, such as the length and width of large and small seismic faults, the rising time, the dislocation time function of fault plane rupture, the initial fracture point of the fault, the shear wave velocity, the fracture propagation velocity, etc., are all uncertain. The uncertainty of these parameters has a significant effect on the simulation results. In this paper, the shear wave velocity and the rise time of small earthquakes are considered. The influence of parameter uncertainty such as rupture propagation velocity and the ratio of length (width) of large and small earthquake faults on the simulation results of strong ground motion based on empirical Green's function method is studied. The probabilistic method is applied to the simulation calculation of empirical Green's function method. The possible values of these four parameters and their corresponding probabilities are determined, and the possible values and corresponding probabilities of the parameters are combined in a logical tree. The simulation results and the corresponding probabilities of the strong ground motions of large earthquakes under different conditions (parameter combinations) are calculated. The statistical method is used to find the optimal solution, that is, the probabilistic empirical Green's function method. In this paper, the ground motion simulation of the Lushan earthquake on April 20, 2013 (MsN 7.0) is carried out. The simulation results show that the peak acceleration is obtained by the probabilistic empirical Green's function method. The acceleration time history and the response spectrum are in good agreement with the actual earthquake records, which verifies the feasibility of the method. In addition, the shear wave velocity and the rise time of small earthquakes are studied respectively. The degree of influence of the uncertainty of each parameter, such as rupture propagation velocity and the ratio of length (width) of large and small earthquakes to the Lushan earthquake of April 20, 2013, on the ground motion simulation results of near-field strength is studied by using the control variable method. The degree to which uncertainty of the parameters affects the simulation results, The results show that the uncertainty of the rise time of small earthquakes has little effect on the simulation results, while the uncertainty of shear wave velocity, fracture propagation velocity and the ratio of length (width) of large and small earthquakes have obvious effects on the simulation results. The uncertainty of the ratio of magnitude to width has the most obvious effect on the simulation results. In addition, the distribution of the actual error and the estimation error of the simulation results under different parameter combinations are also studied. The results show that the method of probabilistic empirical Green's function with uncertain parameters can reduce the error of simulation results.
【学位授予单位】：中国地震局地球物理研究所
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TU352.11;P315.9

【参考文献】