On the use of differential phase for polarimetric rainfall measurements—a new approach to KDP estimation

Differential phase has many unique properties that make it a very attractive radar variable to measure and analyze. Among them is its immunity to radar miscalibration, attenuation in precipitation, and partial beam blockage. Specific differential phase K_DP is less sensitive to variations in drop size distribution at the higher end of the raindrop spectrum and is less contaminated by hail than radar reflectivity factor Z and differential reflectivity Z_DR as far as polarimetric rainfall estimation is concerned. On the negative side are extreme noisiness of K_DP for weak weather echoes, lack of adequate radial resolution, and its vulnerability to the gradients of Z and total differential phase Φ_DP within the radar resolution volume. The latter manifests itself as oscillatory behavior of Φ_DP causing spurious values of K_DP of both signs. As a result, accurate estimation of K_DP in localized convective cells is a challenge. Because of these problems, some of researchers avoid using K_DP for quantitative rain measurements and hydrometeor classification altogether and rely primarily on Z and Z_DR. In this paper, a new procedure for differential phase processing is suggested that ensures the radial resolution of K_DP similar to the one for Z (without degrading the statistical accuracy of the K_DP estimate) and substantially mitigates nonuniform beam filling effects. The proposed procedure involves automatic classification of radar echo types (performed without the use of differential phase) and spatial analysis of measured fields of Z and Φ_DP. Corrupted estimates of K_DP in rain are replaced by the values obtained from the self-consistency relations among K_DP, Z, and Z_DR. The effectiveness of the suggested technique is validated on the number of convective rain events observed with the polarimetric prototype of the WSR-88D radar in Oklahoma.