Technical Challenges in Transferring of the R(A) Methodology for Rainfall Estimation to Operational Systems

The technique for using specific attenuation A for rainfall estimation R(A) for S-, C-, and X-band polarimetric radars has been recently introduced and widely adopted internationally. The merits of the method include its immunity to attenuation, radar calibration errors, partial beam blockage, and wet radome. The algorithm has been developed and is planned to be deployed on a whole fleet of the WSR-88D operational radars. Many cases with different precipitation types over CONUS in different climatically regions have been selected to evaluate the robustness and performance of R(A) for WSR-88D radars. Many issues which affect the accuracy of R(A), such as hail contamination, automatic estimation of the net ratio α=A/K_dp which varies with drop size distributions and is a key parameter for rain rate estimation using the R(A) method, automatic determination of the range interval of the R(A) applicability, etc., are discussed and addressed.

To minimize the contamination from hail, segmentation of the ray is needed and the R(K_dp) relation is used in hail-affected area. Integration in the ZPHI formula for estimation of A has to be performed only over the rain range intervals and the contribution of hail to the total span of differential phase should be subtracted. Although the magnitude of α depends on differential reflectivity Z_dr, a slope of the Z_dr dependency on Z is utilized to estimate α in order to avoid Z_dr calibration errors. Linear and nonlinear relations between Z-Z_dr slope and α have been selected. Thus an optimal value of α is automatically estimated for each radar scan. It has been found that continuous refinements of R(A) are definitely needed before and after the transfer to an operational system. The preliminary results will be presented at the conference.