An Hourly Prediction Model of Relativistic Electrons Based on Empirical Mode Decomposition

Abstract

Abstract During the recovery of a magnetic storm, the relativistic electrons at megaelectron volt energy from the outer radiation belt will be enhanced at the geosynchronous orbit. In particular, the >2‐MeV electrons could penetrate the surface of satellites and accumulate inside. After a long period, such an electron flux effect could cause satellites to be unable to function properly or to fail completely. Unlike previous prediction models of relativistic electrons focusing mainly on forecasting the daily value, we have developed an hourly prediction model to learn more detailed changes. In addition, due to external forcing by the solar wind, relativistic electron flux changes rapidly and is very nonstationary during magnetic storms. We use the method of empirical mode decomposition to reduce these effects. The result shows that the average prediction efficiency of hourly predictions of relativistic electrons from 2001 to 2006 is approximately 0.73. Moreover, we use this method to forecast daily fluence to validate empirical mode decomposition. The prediction efficiency of daily fluence from 2001 to 2006 is 0.80, and for the period in which the relativistic electrons change particularly rapidly during 2001–2004, the prediction efficiency is 0.79. Meanwhile, we compare with the prediction values without using empirical mode decomposition. The result shows a great improvement using this method. Furthermore, it shows that the nonstationary nature of the electron flux time series has a great impact on the prediction of relativistic electrons.