Modeling Electron Density in the Topside of the Ionosphere using Machine Learning

Modeling the Earth's ionosphere is a critical component of forecasting space weather, which in turn impacts radio wave propagation, navigation and communication. This research focuses specifically on predicting the electron density in the topside of the ionosphere, using data collected from the Defense Meteorological Satellite Program (DMSP), a collection of 19 satellites that have been polar orbiting the Earth for various lengths of times, fully covering 1982 to the present. An artificial neural network with two hidden layers was developed and trained on two solar cycles worth of data, including features such as time series F10.7, time series average interplanetary magnetic field (IMF), time series Kp, solar azimuth, and location to generate an electron density prediction. We tested the model on six years of subsequent data, and found a correlation coefficient of 0.73 for a nowcast of electron density. Figure 1 depicts the predicted electron densities along with the true densities measured by a DMSP satellite over a 5-hour period. Upcoming work includes improved testing performance via modified model inputs, tweaking the model architecture, and further rounds of hyperparameter searching. A forecast will then be computed by providing the now-cast model with forecasted global inputs (solar wind, IMF, geomagnetic indices) as part of a larger space weather forecasting effort currently underway. Therefore, we eventually hope to better forecast the electron density of the Earth's ionosphere and in turn better predict space weather, mitigating its negative effects. In addition, these accuracy of these results will be assessed with out-of-training DMSP satellite data alongside the International Reference Ionosphere (IRI).