A stochastic ensemble model composed of three functional forecasting models has been developed to forecast >2 MeV electron flux at geosynchronous (GEO) orbit. The REFM model is based on a statistical link between electron flux and solar wind speed using empirically derived linear filter coefficients, the Li model solves a radial diffusion equation with a diffusion coefficient that is a function of the solar wind velocity and interplanetary magnetic field, and the Fluxpred model is a multi-layer feed-forward neural network with electron flux and summed Kp as input. Individual model results were combined using a multivariate regression to produce significantly better predictive results than any of the individual models alone. A stochastic model is then developed to forecast the probability that a fluence threshold will be exceeded. The regression technique, model optimization, and calculation of forecast probability will be discussed in reference to the ensemble model.
Ionospheric storms--Forecasting--Statistical methods, Atmospheric electricity--Mathematical models, Stochastic models.
Air Force Research Laboratory
Level of Degree
Electrical and Computer Engineering
First Committee Member (Chair)
Second Committee Member
Nelson, Steven. "A stochastic ensemble forecast model for geosynchronous relativistic electron fluxes." (2010). https://digitalrepository.unm.edu/ece_etds/191