Global Validation of the Data‐Assimilative VERB‐3D Code for the Radiation Belts

Abstract Electrons within Earth's radiation belts exhibit large variability in space and time during geomagnetic storms, which could potentially damage satellites and harm astronauts in space. Physics‐based models describe the evolution of energetic electrons in the radiation belts, but they may suffer from uncertainties and errors, particularly in the initial and boundary conditions. Data assimilation can overcome these limitations by combining models with satellite observations, incorporating all available information to create a more reliable reconstruction. This study validates the data‐assimilative three‐dimensional Versatile Electron Radiation Belt code (VERB‐3D) model using data from three independent satellite missions: Arase and GOES for assimilation, and Van Allen Probes for validation. The data sets were cleaned and normalized to ensure their compatibility. The validation shows that the model effectively reproduces the radiation belt dynamics, where the reanalysis remains within a factor of 2 for 88.3% of the time, 98.4% within a factor of 5, and 99.4% within one order of magnitude for 1 MeV electrons. The results highlight the potential of data assimilation for space weather forecasting and as an input for specification models.