Data-Driven Basis Functions for SuperDARN Ionospheric Plasma Flow Characterization and Prediction

Abstract

The archive of plasma velocity measurements from the Super Dual Auroral Radar Network (SuperDARN) provides a rich data set for the investigation of magnetosphere-ionosphere-thermosphere coupling. However, systematic gaps in this archive exist in space, time, and radar look-direction. These gaps are generally infilled using climatological averages, spatially smoothed models, or a priori relationships determined from solar wind drivers. We describe a new technique for infilling the data gaps in the SuperDARN archive which requires no external information and is based solely on the SuperDARN measurements. We also avoid the use of climatological averaging or spatial smoothing when computing the infill. In this regard, our approach captures the true variability in the SuperDARN measurements. Our technique is based on data-interpolating Empirical Orthogonal Function analysis. This method discovers from the SuperDARN data a series of dynamical modes of plasma velocity variation. We compute the modes of a sample month of northern hemisphere winter data, and investigate these in terms of solar wind driving. We find that the By component of the Interplanetary Magnetic Field (IMF) dominates the variability of the plasma velocity. The IMF Bz component is the dominant driver for the background mean field, and a series of non-leading modes, which describe the two-cell convection variability, and the substorm. We recommend our new technique for reanalysis investigations of polar-scale plasma drift phenomena, particularly those with rapid temporal fluctuations and an indirect relationship to the solar wind.