Auroral Oval Boundary Modeling Based on Deep Learning Method

Abstract

Research on the location of the auroral oval is important to understand the coupling processes of the Sun-Earth system. The equatorward boundary and poleward boundary of the auroral oval are significant parameters of the auroral oval location. Thus auroral oval boundary modeling is an efficient way to study the location of auroral oval. As the location of the auroral oval boundary is subject to a variety of geomagnetic factors, there are some limitations on traditional methods, which express the auroral oval boundary as a function of only one or several geomagnetic activity index. Deep learning method is used in this paper to learn the essential features of the inputs, which are a large number of geomagnetic parameters and the former locations of aurora boundary. Furthermore, a model is established to forecast the location of the auroral oval boundary. The experiment results show that our method can model and forecast the boundary of aurora oval efficiently on the data set obtained from Ultraviolet Imager UVI on Polar satellite and OMNI database on NASA.