Using super-resolution to study small scale structures in the atmosphere

A frequent problem in radar meteorology is that the range resolution available is insufficient to investigate small scale atmospheric features such as thin layers. The problem is particularly acute for MST (mesosphere, stratosphere, troposhere) VHF radars which rarely have range resolutions better than 150 metres, however thin layers in the lower atmosphere are known to have vertical scale sizes less than this. Such layers may contain turbulence and/or have step-function changes in the radar refractive index. One approach to improving radar resolution is to switch rapidly between two closely-spaced frequencies in a technique known as frequency domain interferometry (FDI). This requires hardware changes to the system and can lead to some ambiguity in the results, but has nonetheless been successfully applied in a number of MST radar studies. The authors present a new approach that requires no hardware changes and can be applied to previously recorded data retrospectively, provided the system characteristics are known. The technique, known as best linear estimate super-resolution, improves range resolution through the use of `prior knowledge´ of the system impulse response, the scattering model and some of the target characteristics