SuperDARN parameter estimation optimization and implementation of new techniques

Abstract

The Super Dual Auroral Radar Network (SuperDARN) is an international radar network to study the ionosphere and upper atmosphere. The primary target of SuperDARN is field-aligned plasma irregularities in the E- and F-region of the ionosphere. To quantify the characteristics of these irregularities, the radar measures power, Doppler velocity, and spectral width from auto-correlation functions of the received samples. Since the target of interest is overspread, the derived parameters suffer from errors related to cross-range interference. In this thesis, we propose two scenarios to address this problem. First, we implement new approaches to avoid the cross-range interference, and second, we develop new optimization techniques that are more robust and less sensitive in dealing with this interference. New methods include filtering techniques, spectral analysis, and use of inverse techniques. The filtering methods (mismatched and adaptive) offer improvement in both suppressing the side lobes associated with pulse compression techniques and optimal estimation of the main lobe signal-to-noise ratio. Spectral analysis, extracts multiple Doppler velocities in the range while the current time-domain analysis is only capable of measuring one. Instead of dealing with ambiguities, inverse theory applied to SuperDARN received samples can potentially remove the associated cross-range interference, which results in more detailed and accurate information in obtaining the structure and dynamics of the irregularities. More accurate and detailed empirical models resulting from new optimization methods give more information that can be mapped over the current in-progress theoretical models, which finally results in better understanding the physics of the ionosphere.