Quality Contral of Dual PRF Velocity Data for Operational Doppler Weather Radars

Doppler velocity information can be used for analysis and early warning of severe weather, removal of ground clutter, retrieval of wind profiles, construction of dual Doppler wind fields, and so on. Operational application of Doppler velocity data from weather radars is hampered, however, by the infamous limitation of the range–velocity ambiguity. The dual pulse repetition frequency (dual PRF) technique is available to extend the unambiguous velocity interval of Doppler weather radar, but the velocity data are typically contaminated by outliers in large areas of high quality data. In this paper, the radial velocity data from CINRAD-SA, SC, CC and CD type Doppler weather radar running in dual PRF mode have been analyzed quantitatively through calculating the deviations of each radial velocity from its local median velocity, results show the vast majority of the analyzed points obeys local continuity with hardly any deviation or small deviation from the local median velocity, the fraction of outliers are centered at different velocity deviations which roughly match the unambiguous intervals of the low PRF and high PRF measurements. On this basis, a three-step postprocessing algorithm is proposed that enhances the quality of the dual PRF velocity data significantly. First of all, the dual PRF velocity data are filtered based on the size to remove the isolated noise. Secondly, aliased velocity data that fall outside of the extended unambiguous velocity interval are primarily unfolded using local continuity constraints. This part of the algorithm first determines at most two sets of initial reference radials and gates by finding the weakest wind region, in which azimuth continuity constraint of adjacent tilt initial reference radials is applied to eliminate the inappropriate initial reference radials induced by false zero isodops. Then from these initial reference radials and gates, it checks continuities among adjacent gates and corrects for the velocity values with large differences that are larger than the given threshold. After twice round of radial-by-radial dealiasing, the algorithm will perform high gate-to-gate wind shear checking at radial and azimuth directions, if high wind shear still exists, it will second determine a set of initial reference radial and gate by finding the weak wind shear radial with the most valid velocity gates, then the preceding dealiasing procedures are repeated one time, which is needed for dealiasing velocities in isolated storm region away from the radar. Multiple passes of dealiasing are performed in this part of the algorithm that use very strict criteria and little shear threshold in earlier passes and relaxed criteria and large shear threshold in later passes to ensure a good reference for dealiasing in the later passes. And lastly, using the known error characteristics of the dual PRF data, velocity outliers are corrected in an efficient way. The algorithm has been tested using dual PRF velocity data from CINRAD-SA, SC, CD and CC type Doppler weather radars. It is concluded that the algorithm is very efficient and produces high quality velocity data with changeless spatial resolution.