A Prototype Quantitative Precipitation Estimation Algorithm for Operational S-Band Polarimetric Radar Utilizing Specific Attenuation and Specific Differential Phase. Part II: Performance Verification and Case Study Analysis

The quantitative precipitation estimate (QPE) algorithm developed and described in Part I was validated using data collected from 33 Weather Surveillance Radar 1988-Doppler (WSR-88D) radars on 37 calendar days east of the Rocky Mountains. A key physical parameter to the algorithm is the parameter alpha alpha, defined as the ratio of specific attenuation A to specific differential phase K-DP. Examination of a significant sample of tropical and continental precipitation events indicated that alpha was sensitive to changes in drop size distribution and exhibited lower (higher) values when there were lower (higher) concentrations of larger (smaller) rain drops. As part of the performance assessment, the prototype algorithm generated QPEs utilizing a real-time estimated and a fixed alpha were created and evaluated. The results clearly indicated similar to 26% lower errors and a 26% better bias ratio with the QPE utilizing a real-time estimated alpha as opposed to using a fixed value as was done in previous studies. Comparisons between the QPE utilizing a real-time estimated alpha and the operational dual-polarization (dual-pol) QPE used on the WSR-88D radar network showed the former exhibited similar to 22% lower errors, 7% less bias, and 5% higher correlation coefficient when compared to quality controlled gauge totals. The new QPE also provided much better estimates for moderate to heavy precipitation events and performed better in regions of partial beam blockage than the operational dual-pol QPE.