Modeling of aurora borealis using the observed data

Abstract

In the field of Computer Graphics (CG), applications such as movies and games benefit from realistic visualizations of atmospheric phenomena because realistic visualizations increase applications' attractiveness. Among the atmospheric phenomena, auroras are some of the most beautiful phenomena, which makes their visualization highly desirable for such applications.Physically, the generation of auroras depends on the solar activity. Since auroras can be observed only in high latitudes and their observation depends on the weather condition, it is extremely rare to observe auroras directly in nature. Therefore, modeling and visualizing the aurora borealis using CG are important.In previous visual simulation methods for aurora borealis, the fluctuations and the curls of aurora are modeled by using a simple model that treats the aurora as fluids and therefore does not reflect the actual phenomena. One of the reasons for using such approximate simulations is that the motions of the aurora borealis are not fully understood yet among the researchers.In this paper, we propose a model for visualizing the aurora borealis using data gathered from satellite observations. The observed data represents the electric fields and the distributions of Field-Aligned Current (FAC) in high latitudes. By using the captured data and simulating the virtual current circuits that take place in the ionosphere, our method models the shapes of the aurora borealis. We demonstrate that our simulation method can model time-varying aurora borealis.