X-band radar operation at superresolution

Especially at far range, high resolution is desired to acquire data with a better spatial resolution. In addition, more scanning capabilities become available when the initial resolution can be obtained in a shorter time (using superresolution). A superresolution retrieval technique using FMCW X-band radars in the Netherlands is proposed. In this work, the possibilities and benefits for superresolution at far range will be investigated for the optimization of offshore wind farm energy production, within the context of the PROWESS project. Superresolution is a step towards getting more out of the radar instrument. The superresolution technique is described as follows. The additional samples for the radar moments are acquired by using a shifting window over the voltage response signal from n transmitted chirps (for natural resolution blocks of n transmitted chirps are used). The extra samples are thus obtained by increasing the amount of processing of the voltage data samples (not by interpolation of radar moments). This superresolution is achieved in the scanning direction of the radar, and the range resolution stays the same. Specifically for PPI scans, the proposed superresolution implies that the azimuth resolution is increased, or the same azimuth resolution can be obtained in a shorter time. A Wiener deconvolution is applied to the spectral reflectivity to remove the smoothing effect of the radar power pattern, and the Fourier antialiasing windowing function (typically a Chebyshev window is used). First results on the following topics will be presented: 1. the implementation of the Wiener deconvolution algorithm and its effect on the quality of the radar moments; 2. spectral Doppler processing, which was upgraded to three dimensions (range, Doppler velocity, +time); and 3. the limitations due to computational requirements. The superresolution algorithm is currently being implemented in radar processing software from SkyEcho (SkyTorque) that is used by the IDRA, Rijnmond and MESEWI X-band radars in the Netherlands. The SkyTorque software is using CUDA processing capabilities to do real-time advanced spectral processing, with advanced filtering techniques such as Jensen–Shannon Distance-Based adaptive copolar correlation thresholding (C. Chen et al., 2022).