Investigation of Ionospheric Disturbance and Seismic Events Based on Machine Learning

Abstract

 Earthquakes are natural disasters that endanger human life and cause the greatest loss of property. The study of anomalous disturbance of the ionosphere, one of the pre-earthquake anomalies, will help to further study the coupling effect on the ionosphere before the earthquake. This paper focuses on the analysis of the importance and influence of various parameters inside the seismic ionosphere under earthquake conditions, constructs a classification and prediction model of seismic ionospheric anomalies based on the gradient boosting decision tree GBDT algorithm, and analyzes the pre-earthquake ionospheric data. , the results show that NmE, nHe+, foF2 and TEC are important influencing factors, which lays a certain foundation for the further study of the internal structure and parameters of the pre-earthquake ionosphere