Statistical Analysis Of Earthquake-induced Sources And Their Induced Forces In The North China Earthquake Region Based On The Logarithmic Gaussian Cox Process Mode

The North China seismic region is the main area of strong earthquake activity in eastern China,with magnitude and frequency second only to the "Qinghai-Tibet Plateau seismic zone".Due to the fact that the capital region is located in this seismic zone and there are densely populated large cities concentrated in this area,it is of great significance to explore the spatial distribution characteristics and the generation-driven mechanism of earthquake disasters in the North China seismic zone.At present,the research on earthquakes mainly focuses on the impact of earthquakes on economic society or the space-time distribution characteristics of earthquake disasters.Most of the research methods only use simple spatial correlation analysis to explore the spatial aggregation of earthquakes.Few scholars,starting from the perspective of earthquake inducement sources in the North China seismic zone,have deeply explored the impact of squeezing between the Eurasian plate and other adjacent plates on the occurrence of earthquakes in the North China seismic zone and quantified the impact through modeling.In this paper,the Log Gaussian Cox Process(LGCP)model is used to explore the spatial distribution and the inducement source of earthquakes in the seismic zone of North China.In order to quantitatively analyze the difference of the impact of the compression between different types of plates and the Eurasian plate on the occurrence of earthquakes in North China seismic zone.The earthquake disaster event data in the North China seismic zone is treated as point event,and the LGCP model is used to model the occurrence intensity of point events.In the intensity model,the spatial distance from the seismic point to the intersection points of the Indian Ocean plate,the Pacific Plate,the Philippine Sea plate and Eurasian plate are included in the mean structure of the model as environmental covariates.The error process is modeled by a spatial Gaussian process determined by a stochastic partial differential equation.After the prior distribution of each parameter in the model is determined,the direct calculation of the covariate coefficient and the posterior mean of the spatial Gaussian process super parameter requires the integration of the multidimensional posterior conditional density,which is computationally difficult.In this paper,the R-INLA package is used for rapid sampling to obtain the posterior distribution approximation of the model parameters.Finally,the DIC criterion is used to select the model,to eliminate the insignificant sources in the model.Based on the estimated results of the final model parameters,several suggestions on earthquake prevention and control are proposed for the reference of government divisions in the decision making of earthquake disaster reduction.This paper firstly divides earthquakes into weak earthquakes,sensitive earthquakes and moderately strong earthquakes according to their magnitudes.According to the seismic data of three magnitudes,the general situation of earthquake occurrence and spatial clustering characteristics in the zone are described and statistically analyzed.The results show that:(1)The five provincial regions with the most frequent earthquakes in the North China seismic zone are: Sichuan Province,Hebei Province,Shanxi Province Inner Mongolia and Ningxia Hui Autonomous Region,while Beijing and Anhui Province have fewer earthquakes.(2)The global Moran index of weak earthquakes,sensible earthquakes and moderately strong earthquakes are significant,indicating that the larger the magnitude of earthquakes,the higher the degree of spatial aggregation of seismic events.In the empirical analysis,the LGCP model is established to quantitatively analyze the induced forces of different earthquake sources in the North China seismic region,and the posterior distribution of each parameter in the model is obtained.The results show that:(1)According to the size of the global Moran index of earthquake events and the posterior mean value of super parameters,it is found that there are differences in the spatial aggregation of the three earthquakes,with different spatial distribution characteristics.(2)The occurrence of weak earthquakes,sensitive earthquakes and moderately strong earthquakes are all affected by the compression between the Asia-Europe plate and the Indian Ocean plate,followed by the compression between the Asia-Europe plate and the Pacific plate.While only the sensitive earthquakes are affected by the compression between the Asia-Europe plate and the Philippine Sea plate.In this paper,the LGCP model of seismic events in North China seismic region is constructed based on SPDE,and the impact of different types of inter-plate compression on earthquakes in North China seismic region is quantitatively analyzed.Furthermore,the Bayesian estimation results based on the model parameters provide a targeted test for local earthquake prevention.Control Embedding SPDE model in the LGCP model to carry out statistical analysis of seismic sources in North China seismic region is an attempt to quantitatively analyze seismic sources,which will provide an important reference for the study of seismic source exploration in other seismic zones.