Ground Motion Prediction Equation Based On Strong Ground Motion In Southwest China

The Ground Motion Prediction Equation(GMPE)is an important method for describing the effects of earthquakes and is an important factor in the results of seismic hazard analysis.Due to the lack of early seismic records in southwest China,existing studies of GMPEs have often been conducted through intensity conversion or supplemented of added records from other regions.Also,in southwest China,where earthquakes have been frequent,especially since 2008,when three earthquakes of magnitude 7 or higher occurred,it is necessary to consider the influence of these abundant data on the ground motion prediction equation.In the thesis,the ground motion data obtained from Strong Ground Motion Observation Network in China in recent years,and based on the magnitude Ms in the range of 5.0 to 8.0,6105 strong ground motion records from 2035 groups of stations were selected.With reference to the existing ground motion prediction equation fitting method,the filtered strong ground motion records are subjected to a random effect based multi-parameter nonlinear global optimization regression fits are obtained for the horizontal and vertical ground motion prediction equations in the region under different site conditions,respectively.The regression fit results,residuals,and standard deviations were also analyzed and compared with several existing horizontal and vertical ground motion prediction equation in southwest China to verify their reasonableness.The comparison with existing models shows that the predicted value in this paper is larger in the far-field segment,but the predicted value is closer to the ground motion records data,and At Mw=7 and 8,the predicted values in this thesis are closer to the actual ground motion records data.In addition,it was also found that the existing ground motion prediction equations predict better for Class I sites than for Class II sites,which are since the existing ground motion prediction equations fit a regression of ground motion records data from all sites.The Class I and II sites recorded data interact with each other,with the amplitude of the ground motion records data for Class I sites being significantly greater than that for Class II sites,which has a greater impact on the ground motion prediction equations.In general,the predicted values of the ground motion prediction equation in this thesis are more fit with the actual ground motion records data,which can better reflect the impact of the earthquake attenuation characteristics in the southwest China.