Stochastic Finite Fault Model Based On Empirical Formula To Simulate Ground Motion

A large number of earthquake damage investigations and statistical data show that the damage and collapse of building structures are the main causes of casualties and economic losses,and these damages are often caused by strong earthquakes.Despite the rapid accumulation of ground motion data,there are still few strong ground motion data.Seismic hazard analysis can provide a theoretical basis for seismic design of building structures in designated areas,thus reducing the losses caused by earthquakes.Therefore,when analyzing the seismic resistance of the structure,the commonly used way is to simulate and analyze the ground motion by building a perfect model.However,how to obtain the data and relevant parameters before the occurrence of a large earthquake remains the main difficulty at present.The main work and conclusions of this paper are as follows:(1)According to the theory of seismology,the relationship between rupture length,rupture width and moment magnitude of earthquake fault,the relationship between mean slip on fault surface and seismic moment,the relationship between surface wave magnitude and seismic moment is studied,and the theoretical relationship is derived.Based on a large number of seismic event observation data and seismic theory analysis,the semi-empirical relationship between key parameters in the process of earthquake rupture is obtained by using statistical analysis method.Based on the above relation,the empirical relation suitable for the study area is selected,and a set of practical relation of seismic source parameters is established.(2)The method of obtaining some parameters in stochastic finite fault model is improved,and a stochastic finite fault model based on empirical formula is established.Taking the M_W7.9earthquake that occurred in the Gulf of Alaska on January 23,2018 as an example,the acceleration time histories of ground motions at two different azimuths of 12 bedrock seismic stations at the epicenter were selected,and the ground motion time histories and response spectra of 12 bedrock seismic stations were simulated with the stochastic finite fault model set in this paper.The results show that the average value of the simulation error of each period of 12 seismic stations is small,and the simulation error is between 1.08 and 0.92.There is no significant relationship at the 95%confidence interval of the periodic variation,moreover,the error standard obtained from each cycle simulation meets the requirement of less than the standard value of 1.It is proved that the simulation has high average accuracy,reflects the average effect of ground motion,and proves the effectiveness of the empirical formula in this paper to obtain some source parameters.(3)Future large earthquakes will occur on active faults,but the exact rupture point and slip distribution on the fault plane before the earthquake is still unknown.Assuming the initial rupture point and quasi-random slip distribution of the fault plane,the average simulation error of 12seismic stations in a certain period of time is calculated again,and the conclusion similar to the initial simulation is obtained,which proves the effectiveness of the method in this paper.Therefore,on the premise of knowing the seismographic structure of future large earthquakes,and the initial rupture point and slip distribution on the fault plane are unknown,we propose to use the empirical equation to determine the length,width,seismic moment and slip distribution parameters of the fault plane of large earthquakes,and use the stochastic finite fault model to predict the future large earthquakes.The paper has 25 figures,3 tables,and 67 references.