Characteristics Of Ground Motion Spectrum And Its Application In Earthquake Early Warning

As one of the three basic characteristicselements of ground motion,the seismic spectrum has great influence on the degree of earthquake seismic damage.The current earthquake early warning system（EEW）has not been able to quickly and effectively is unable to produce ground motin spectrum quickly,which makes it impossible to consider the influence of ground motion spectrum in the prediction of seismic damage.output the spectrum information of ground motion,so that the prediction of earthquake damage cannot consider the influence of the spectrum characteristics of ground motion.Therefore,based on the surface strong motion records of KiK-Net database in Japan,this paper investigates studies the influence influencing factors of typical ground motion spectrum characteristics,and proposes a method for fast rapid estimation of ground motion periods cycle parameters based on early warning parameters of first few seconds seismic P-waves.The main research contents arework is as follows:Firstly,according to the data requirements of earthquake early warning,a large number of high-quality strong earthquake records are selected from Japan KiK-Net database as the research data set.After comparing the site classification methods of China,America and Europe,664 stations in Japan KiK-Net database are classified by these methodsaccording to the site classification methods of different countries,the results show that the classification based on 30 m equivalent shear wave speed vs30 is the most universal,and V_s30 is used as the site classification in this study.and the most universal equivalent shear wave velocity V_s30 is selected as the site classification method.Subsequently,based on the above research data and site classification method,the influence of magnitude and site conditions on the spectral characteristics of ground motion is qualitatively analyzed.The results show that the larger the magnitude is,the slower the average response spectrum decreases.On the whole,the site conditions are also consistent with the softer the site is,the slower the average response spectrum decreases.This result confirms that magnitude and site conditions have significant indigenous effects on ground motion spectrum.Then,according to the results of the above qualitative analysis,the empirical relationships between the four typical periodic parameters(including the weighted average period of the response spectrum T_vag,the average period of the Fourier amplitude spectrum T_m,the smoothing predominant period of the response spectrum T_o,and the predominant period of the response spectrum T_p)and the magnitude and epicentral distance of the ground motion are further quantitatively analyzed.The results show that T_vag,T_m,T_o and T_p will increase with the increase of magnitude and epicentral distance under various site conditions.The T_vag of bedrock site,hard soil site and medium hard soil site has good correlation with magnitude.The T_vag,T_m,T_o and T_p of bedrock site,hard soil site and medium hard soil site showed weak correlation with epicentral distance.These results confirm the feasibility of using magnitude to predict ground motion period parameters.Finally,according to the feasibility of predicting the ground motion period parameters by the above magnitudes and the actual situation of predicting the magnitude by the first-break P-wave early warning parameters,the method of directly predicting the ground motion period parameters by the first-break P-wave early warning parameters is established.Three typical early warning parameters(including maximum predominant periodτ^p_max,average predominant periodτ_c and displacement amplitude P_d)were selected and compared with the empirical formulas of the four typical ground motion periodic parameters.The results show that the bedrock site and hard soil siteτpmax have the highest correlation coefficient and minimum error standard deviation with T_avg and T_m,and the medium hard soil siteτ^p_max has the highest correlation coefficient and minimum error standard deviation with T_avg.Based on these results,τ^p_max is proposed to predict ground motion period parameters.In this paper,the empirical formula of ground motion period parameters is established based on theτ^p_max of first break 3 s–6 s P wave,which can make up for the lack of ground motion spectrum information in the current earthquake early warning system and improve the accuracy of earthquake damage prediction.