Empirical Modeling of Planetary Magnetospheres in Response to Solar Wind Dynamics Using EOF Analysis and Multivariate Linear Regression

Abstract

A popular approach for empirical modeling is through representing a phenomenon as a linear combination of a set of predefined basis functions, such as polynomial and harmonic functions. In comparison with predefined basis functions, the empirical orthogonal basis functions (EOFs) are constructed from the dataset according to the variance distribution, and typically summarize the data into representative features. This chapter reviews comparatively three applications of EOF analysis associated with multivariate linear regression in empirical modeling, namely in the models of the Earth’s ionospheric F ₂ -layer peak, the field-aligned currents at Earth, and the induced magnetic field near Venus (He et al. Geophys. Res. Lett., 38(14): L14101, 2011; He et al. Geophys. Res. Lett., 39, 2012; He et al. 2017; He et al. J. Geophys. Res., 121(4), 3362–3380, 2016). We illustrate the physical meaning represented by the most important EOFs, detail the model constructions and methodology, and highlight the revealed main scientific results.