An Explainable Dynamic Prediction Method for Ionospheric foF2 Based on Machine Learning

To further improve the prediction accuracy of critical frequency ionospheric F2 layer (foF2), we use machine learning method (ML) to establish an explanatory dynamic model predict foF2. Firstly, according ML modeling process, three elements establishing a foF2 and four problems be solved are determined, idea concrete steps building determined. Then data collection is explained in detail, change mapping its parameters determined turn. Finally, established compared with International Reference Ionospheric (IRI-2016) Asian Regional Model (ARFM) verify validity reliability. The results show that IRI-URSI, IRI-CCIR, ARFM models, statistical average error decreased by 0.316 MHz, 0.132 0.007 respectively. Further, relative root-mean-square 9.62%, 4.05%, 0.15%,