Improvement of SuperDARN Data Products by Signal-Derived Weights

Abstract:

Existing data products from the Super Dual Auroral Radar Network (SuperDARN) radars are obtained through least-squares fitting utilizing ad hoc weights (or ad hoc error bars). In low signal-to-noise and/or low signal-to-clutter regimes, these weights often lead to underestimated fitted parameter errors since they do not to consider the relative contributions of signal, noise, and clutter to the measurement uncertainty. We present signal-derived weights that include contributions of signal, noise, and clutter and their effects on extraction of ionospheric parameters (velocity, spectral width, etc.) from SuperDARN data. It is shown that using signal-derived weights, the fitted parameter errors are self-consistent and reliable quantitative measures of the uncertainty in extracted ionospheric parameters.