Illuminating the Physics of Aurora: A Predictive Model Based on Laboratory Simulations

Abstract

This study aims to simulate the generation of auroras in a laboratory environment and to explore changes in the intensity of artificial auroras under varying physical conditions through mathematical modeling. It introduces the mechanisms behind natural and artificial auroras, explaining the factors influencing auroral intensity. A sufficient data set is obtained by conducting a low-cost, classroom-based experiment while maintaining scientific validity. A predictive model is established using a radial basis function (RBF) neural network, with input parameters including magnetic field strength, current, voltage, vacuum level, electrode distance, and output corresponding to auroral intensity. The model enhances the controllability and accuracy of indoor auroral experiments and provides theoretical guidance for future artificial aurora studies, laying the groundwork for broader space environment simulations.