THE OVERARCHING GOAL OF THE PROPOSED WORK IS TO PROVIDE THE SCIENCE AND OPERATIONS COMMUNITIES THROUGH CMMC WITH A GRAY-BOX MODEL OF CIMI THAT WILL FACILITATE UNCERTAINTY QUANTIFICATION. THE GRAY-BOX MODEL WILL BE DEVELOPED THROUGH THE EFFECTIVE COMBINATION OF PHYSICS-BASED CIMI MODEL AND MACHINE LEARNING. THE OUTPUT OF THE MODEL WILL BE LOW AND HIGH ENERGY INTEGRATED ELECTRON FLUXES FOR ESTIMATING SURFACE AND INTERNAL SPACECRAFT CHARGING. SPECIFICALLY WE WILL INCORPORATE BAYESIAN NEUTRAL NETWORKS TO ESTABLISH AN ABILITY TO PROVIDE RELIABLE UNCERTAINTY ESTIMATION FOR NOWCAST AND FORECAST.