Seismic And Geomagnetic Data Analysis Based On Triple Exponential Smoothing Prediction Algorithm

There will be produce electromagnetic phenomena before the earthquake.In this paper,the geomagnetic research before Wenchuan earthquake found that: when the Wenchuan earthquake happened,some geomagnetic stations far away from the earthquake center observed geomagnetic anomalies at the same time.However,some geomagnetic stations close to the earthquake center did not observe geomagnetic anomalies.in-depth study of the abnormal phenomenon may provide some reference value for earthquake prediction,so in order to further study the abnormal phenomenon,need more geomagnetic data,and find the right algorithm to obtain the geomagnetic data for analysis.In this paper,the geomagnetic data of five earthquakes of magnitude 7.0 or above occurred in China since 2000 were obtained from the center of the national geomagnetic network.They were the pre-processed geomagnetic second records of 45 stations of the wenchuan earthquake,64 stations of the yushu earthquake,62 stations of the ya 'an earthquake,70 stations of the yutian earthquake and 65 stations of the jiuzhaigou earthquake seven days before the earthquake.The research work of this paper is as follows:The obtained geomagnetic data file is processed by the program.Eliminate empty and corrupt files,as well as incomplete data in available files.Then,the geomagnetic station name,IAGA CODE code,longitude and latitude,altitude and other geomagnetic station related information as well as geomagnetic element relative data such as magnetic declination angle D,horizontal component H,vertical component Z,total field strength F are extracted.The obtained geomagnetic data will have some continuous and identical invalid data with very large values.Based on the characteristics of these invalid data,this paper proposes an algorithm to eliminate these invalid data,which has good results in the obtained geomagnetic data.The magnetic declination D is greatly affected by the electromagnetic effect generated by the earthquake and less affected by the interference,so this paper uses the magnetic declination D to study geomagnetic anomalies.The existing geomagnetic anomaly processing algorithms are analyzed.There are simple difference method and complex difference method to detect geomagnetic anomalies in time domain.These two algorithms need the data of comparison stations to effectively detect geomagnetic anomalies,and the electromagnetic generated by earthquakes may also affect the geomagnetism of comparison stations,so they cannot effectively detect geomagnetic anomalies caused by earthquakes.From the perspective of frequency domain,the second sampling geomagnetic data in this paper can only detect the spectrum below 0.5 Hz.Previous studies have shown that geomagnetic field precursor anomalies before earthquakes are concentrated below 30 Hz,so this algorithm cannot effectively detect geomagnetic anomalies caused by earthquakes.The geomagnetic anomalies studied in this paper belong to local geomagnetic anomalies.Some algorithms that can effectively detect long-term geomagnetic variation anomalies are not suitable for processing local anomalies of geomagnetic data,such as low-point displacement method,so they cannot be used in the geomagnetic anomalies research in this paper.The geomagnetic anomaly detection algorithm suitable for this paper is improved and the obtained geomagnetic data are processed.The geomagnetic anomalies studied in this paper have the characteristics of high volatility,large amplitude and local occurrence in the geomagnetic data,and the geomagnetic data has a seasonal time series.Therefore,this paper improves the three-exponential smoothing prediction algorithm,which has good effect in predicting seasonal time series,to detect geomagnetic anomalies.Firstly,the three-exponential smoothing prediction algorithm is used to obtain the predicted value through the measured value.Secondly,in order to reduce the influence of geomagnetic interference,the difference value greater than half of the total number of time points in 11 consecutive time points is recorded to be greater than the predicted value and the measured value of the preset threshold.Finally,due to the relatively stable change of the predicted value,the large fluctuation and amplitude of the abnormal value,in order to improve the detection accuracy,the result of summing and dividing the recorded measured value and the predicted value after taking the absolute value is smaller than the measured value of the preset threshold,and the remaining measured value is considered as the abnormal value due to its strong fluctuation and large amplitude.The algorithm is used to detect the anomaly of magnetic declination D data of all geomagnetic stations of 5 earthquakes.Finally,the correlation analysis of the geomagnetic anomalies detected at each station in the five earthquakes was carried out using the correlation analysis algorithm of the abnormal time interval of multiple stations proposed in this paper,and the detected abnormal stations and abnormal electromagnetic peak-to-peak values and other information were plotted on the map,so as to analyze the acquired geomagnetic data.geomagnetic stations detected seismic geomagnetic anomalies on the day of Wenchuan earthquake.The geographical location of the nine stations and the detected abnormal peak-peak values further indicated that some stations far away from the epicenter of the earthquake detected stronger abnormal peak-peak values than those close to the epicenter.