Research On Division Of Seismic Hazard Zone From Mobile Gravity And Statistical Method

The earthquake is a highly harmful natural disaster and tens of thousands of people die every year from the earthquakes and their secondary disasters every year.Therefore,the early division of earthquake danger zones is of great significance for disaster prevention.However,the accurate predictions of earthquake are still a worldwide problem.The process of preparation,development and occurrence of earthquakes is extremely complicated,and there are many factors affecting earthquake occurrence,which are highly nonlinear with the earthquakes.The mobile gravity data has been used by researchers as effective earthquake precursor information for the division and judgment research of seismic danger zones.The relationship between gravity anomaly changes before the earthquake and seismogenic process is effectively verified in the analysis of some typical earthquakes.For example,Wenchuan Ms8.0,Yutian Ms7.3,Lushan Ms7.0,Yaan Ms6.0 and other earthquakes have significant precursor signs of gravity-gravity gradient zone or quadrant distribution of high and low variations.Based on the previous studies,this paper establishes a new seismic hazard zone division method using mobile gravity data,earthquake catalog data and national fault distribution data.Firstly,based on the earthquake catalog data,based on the spatiotemporal position of the earthquake,the kernel function algorithm is used to calculate and map the long-term earthquake background information,which provides a reference for seismic trend judgment.The concept of a fault buffer is applied around the occurrence of earthquakes and the spatial distribution of faults,and the fact that earthquakes are strip-shaped.Based on the fault buffer,this paper obtains the activity index of the main active fault,and proposes a concept about statistical activity and relative active fault,and marks the affected area of the fault zone with the ability to generate earthquakes.Finally,the fault buffer is applied to the calculation of the b-value,and the method of calculating the bvalue of the fault is utilized,and the possibility of applying the method to the seismic danger zone assessment is evaluated.In order to solve a problem,the gravity anomaly mode before the earthquake is mainly based on the experience of the researchers,and it is impossible to automatically identify and divide the gravity danger zone.This paper also uses the machine learning algorithm to automatically identify the precursor of the earthquake based on the space-time variation image of the gravity field.Firstly,based on the earthquake cases of the past years,the data sample library is built,and the data sample library is expanded by the algorithm of offline enhanced learning.Then use the machine learning algorithm to train the sample library data and establish a neural network model.Finally,the "Face Positioning" method is used to locate and identify the gravity anomaly in the gravity change image obtained by the mobile gravity monitoring,thereby automatically completing the image interpretation of the heavy mechanical earthquake precursor.Combined with the spatial range of the seismogenic fault,the calculation result of b value and the gravity anomaly area identified by the machine learning algorithm,the seismic danger zone is comprehensively combined,and the future earthquake can be predicted and evaluated.