Hybrid-Scan Estimator: Using Split Cut Data to Improve the Quality of Polarimetric Variables

Volume Coverage Patterns (VCPs) used by the Weather Surveillance Radar – 1998 Doppler (WSR-88D) utilize “split cuts” whereby, to help mitigate range and velocity ambiguities, the same elevation angle is scanned twice using two different pulse repetition times (PRT). In other words, split cuts are used to maximize the spatial coverage of data through a surveillance scan with a long PRT and also to reduce the occurrence of velocity aliasing through a Doppler scan with a short PRT. With the dual-polarimetry upgrade of the WSR-88D complete, the polarimetric variables are computed using the Surveillance-scan data on split-cut elevations: no information from the Doppler scan is used. In this work, we present a gate-by-gate decision rule that chooses the polarimetric-variable estimate from either the surveillance or Doppler scans. The choice of estimate is controlled by signal characteristics and radar scanning parameters. The proposed hybrid-scan estimator chooses Doppler-scan estimates if both the theoretical bias and variance of the observed estimates are expected to be lower than those of the surveillance-scan estimates. Otherwise, estimates from the surveillance scan are kept. A comprehensive analysis is carried out for a wide range of simulated weather-signal parameters used on the WSR-88D, which confirms that Doppler-scan estimates can significantly improve the quality of the polarimetric variables under certain conditions. Furthermore, real radar data from WSR-88D testbed and operational systems are processed using both conventional (Surveillance only) and hybrid-scan estimators. Initial results confirm that a substantial improvement in data quality can be obtained with the proposed hybrid-scan estimators.