Statistical and predictive study of the outer belt boundary and electron content using POES-15 and FY-3C observations

The dynamics of Earth’s radiation belts play a crucial role in space weather modeling. However, most existing studies have primarily focused on local electron fluxes, whereas systematic investigations of global parameters, such as the radiation belt content (RBC) and the outer belt boundary, remain limited. In this study, we conduct a superposed epoch analysis of different storm types to examine the distinct responses of electron fluxes, global parameters, solar wind conditions, and geomagnetic indices. Building on the statistical results, we further develop short-term predictive models for the RBC index and the outer boundary location by using support vector regression. By characterizing global variations through statistical analysis and incorporating physical understanding into machine learning models, this study not only advances the understanding of storm-driven radiation belt dynamics but also improves predictive capabilities for space weather applications.