论文部分内容阅读
利用Dieterich(1994)提出的物理模型模拟应力的变化而引起的地震活动时空演化过程。通过依赖于速率和状态的摩擦定律,将其应用于无限地震成核点群。模拟结果显示,地震活动速率变化取决于应力扰动的幅度、断层的基本物理特性(用参数Aσ表示)、应力加载速率和研究区的参考地震活动速率。为了使该模型具有一定的预报性,我们着重研究了物理模型的参数和参数间的相关关系对模拟结果的影响。首先,我们讨论了参考地震活动速率的不同定义,并以美国1992年的兰德斯地震序列为例,说明其对地震发生速率计算中的影响;其次,我们从物理上和统计上论证了模型各参数间均存在强相关。我们讨论了这种相关性并强调指出,参考地震活动速率、应力加载速率和Aσ值这些模型参数的估计与模拟再现观测到余震序列密切相关。我们推导出的解析关系证明了这些模型参数对于与大森定律相当的余震衰减律的意义,模型参数决定了大森定律中时间常数c值和常数因子K,并表示p值≤1.0。论文最后讨论了以准实时预测为目标的约束模型参数的最优策略。
Using the physical model proposed by Dieterich (1994) to simulate the temporal and spatial evolution of seismic activity caused by the change of stress. Through the friction law that depends on the velocity and state, it is applied to the infinite earthquake nucleation group. The simulation results show that the change of seismic activity rate depends on the magnitude of stress disturbance, the basic physical characteristics of the fault (indicated by the parameter Aσ), the stress loading rate and the reference seismic activity rate in the study area. In order to make the model have a certain degree of predictability, we focus on the physical model parameters and parameters of the correlation between the simulation results. First of all, we discuss different definitions of the reference seismic velocity. Taking the Ranlands earthquake sequence of 1992 in the United States as an example, we illustrate its influence on the seismic velocity calculation. Secondly, we demonstrate the physical and statistical models There is strong correlation between the parameters. We discuss this correlation and stress that estimates of model parameters that are referenced to seismicity rates, stress loading rates, and Aσ values are closely related to the simulation of aftershock sequences. The analytic relationship we deduced proves the significance of these model parameters for the aftershock decay law comparable to the Omori law. The model parameters determine the time constant c and the constant factor K in the Omori law and indicate that the p value is ≤1.0. Finally, the paper discusses the optimal strategy of the constraint model parameters aimed at quasi real-time prediction.