Application Research Of Correlation Analysis Method On Earthquake Precursor Data Processing

Earthquakes are worldwide recognized as a large-scale natural disaster.The economic losses and casualties caused by the great earthquake are amazing.Strengthening earthquake monitoring and forecasting will help reduce the harm caused by the earthquake.Seismic activity characteristics,precursory anomalies,and co-seismic anomalies have been the focus of seismologists' research.A large number of studies have shown that the earthquake precursory data will change abnormally during the occurrence of earthquakes,indicating that these anomalies should have a certain degree of physics.Correlation analysis methods can tap the potential links between different items in a large number of datasets.Therefore,this paper attempts to use the classic Apriori algorithm to analyze the correlation between precursory observation data anomalies and earthquakes and the correlation between many items,and to try to analyze the precursor observation items with high correlation with earthquake,and provide a new analysis hand for the study of earthquake precursor anomaly.In this paper,we used the precursory data of some stations and theearthquake catalog data in Sichuan Province in the first half of 2008.The db4 wavelet was used to decompose seismic precursory data by wavelet decomposition.The method of sliding threshold was used to quantitatively extract the abnormal of detail components.The positive and abnormal values of precursory data and seismic catalog data were utilized binary model,then the Apriori algorithm was used to analyze the correlation between the earthquake and the items of different subjects,and some implicit correlation information was obtained.The implicit association rules obtained during the research process include:(1)The Apriori algorithm was used to generate correlation rules for some items in the disciplines of fluids,geoelectricity,geomagnetism,and deformation,and earthquakes above Grade 3.It was found that the correlation between the earthquake and the deep water temperature,the three component of the geoelectric field,the geomagnetic level and the vertical component and the observation of the borehole strain of the component was weak.Among them,there was a weak correlation rule between the deep water temperature of two fluid stations and the earthquake.The number of weakly related rules generated by the east-west components of the electric field observations and earthquake was the largest;the correlation between seismic and geomagnetic vertical components was stronger than the horizontal component;In the deformation data involved in the calculation,only the Gu Zan Deformation data that was closer to the earthquake occurrence area and the earthquake form a correlation rule,and the earthquake and the North-South component of borehole strain observation generated a weak association rule.(2)For the correlation analysis between many items of data,it was found that the correlation between different items at the same measurement point of the same subject in the same station was stronger,while the correlation between items from different disciplines was weak.The correlation betweenthe three components of the geoelectric field observation was strong,and a strong association rule was generated.There was a strong correlation between the north-south component and east-west or north-east components of borehole strain observations of deformation components.By calculating the lifting degree of strong association rules,it was found that the correlation strength between the preceding and the latter of these strong association rules was positively correlated.The items of different disciplines was mainly reflected in the weak correlation rules between the component of north-south component of the borehole strain observation and the three components of the electric field,the water level,the water temperature and the magnetic declination.