TEC Prediction And Seismic Ionospheric Anomaly Analysis Based On EWT And Improved RVM

The ionosphere is an important part of the earth space.Ionospheric total electron content(TEC)is one of the important indicators to describe the temporal and spatial distribution and characteristic changes of the ionosphere.The prediction of ionospheric TEC can provide data basis for the study of ionospheric radio wave delay correction,earthquake precursory analysis and the occurrence mechanism of earthquake events.In recent years,Ionospheric TEC prediction and seismic anomaly detection have become the research hotspot of many scholars.However,due to the complexity of the ionosphere and the lack of interpretation theory of earthquake occurrence mechanism,the methods of TEC prediction and seismic anomaly detection need to be improved.Based on this,this article uses empirical wavelet transform and improved correlation vector machine to improve ionospheric TEC prediction and seismic anomaly detection methods,and combines specific earthquake examples for comprehensive analysis.The specific content is as follows:1.This article describes the structure of the ionosphere,introduces the influence of solar and geomagnetic activities on the Ionospheric TEC,and analyzes the period between the Ionospheric TEC and sunspot by using continuous and cross wavelet transform.It is found that there is a resonance period of about 27 days between the Ionospheric TEC and the sunspot in 2012,2014 and 2015.2.This article introduces chaos theory and the principle of adaptive kernel learning correlation vector machine,puts forward an a RVM prediction model based on chaos theory,and uses this method and time series method to predict and evaluate TEC in different solar activity intensity,longitude and latitude.It is concluded that the prediction effect of a RVM prediction model based on chaos theory is better than that of time series method.3.This article introduces the principle of empirical wavelet decomposition,puts forward an EWT-a RVM prediction model based on chaos theory,and uses it with a single a RVM prediction algorithm and time series method to model and analyze and evaluate the prediction accuracy of ionospheric TEC data.The experimental results show that the prediction effect of EWT-a RVM algorithm based on chaos theory is better than that of single a RVM and time series model in low and high years of solar activity,and can be well used in TEC prediction.It is considered as a TEC background value prediction model during earthquakes4.This article uses the TEC data of three earthquake cases in Wenchuan,Yushu and Jiuzhaigou to analyze,and uses the traditional quartile distance method,sliding time window method and EWT-a RVM prediction algorithm based on chaos theory to predict the background value and analyze the outliers.It is found that:(1)the abnormal phenomena on the 13 th,9th,6th and 3rd day before the Wenchuan earthquake may be the precursory information before the earthquake;(2)the abnormal phenomena on the 12 th day,the 2nd day and the day of the earthquake before the Yushu earthquake in Qinghai can be regarded as abnormal disturbance points caused by the earthquake;(3)TEC anomalies occur 12 to 11 days before the Jiuzhaigou earthquake in Sichuan,which can be used as precursory information before the earthquake.(4)the ionospheric TEC anomaly occurs about 12 days before the three earthquakes,which may be caused by the seismogenic earthquake and can be further analyzed.