By the end of July 2011, seven Weather Surveillance Radar 1988, Doppler (WSR-88D), radars were upgraded to Dual Polarization (DP) capabilities. As part of this effort, the Radar Operation Center's Applications Branch and the National Severe Storms Laboratory's (NSSL) Storm-Scale Hydrometeorology group are working together to evaluate the performance of the new DP Quantitative Precipitation Estimate (QPE) algorithm. An initial assessment was recently completed for two of the new DP radars: the KOUN DP prototype located in Norman, OK and KVNX located at Vance AFB, OK.

For KOUN, six warm season rain events were analyzed. DP QPE and the legacy Precipitation Processing System (PPS) estimates were compared with Oklahoma Climate Survey (OCS) rain gauges. Bias Error, defined as the radar rain estimate minus the gauge amount, and the Root Mean Square Error (RMSE) were calculated for each radar estimate paired to a co-located OCS rain gauge, henceforth called radar/gauge (R/G) pairs. Non-parametric statistical methods were used to calculate if the differences between DP QPE and PPS statistics were significant. Over 500 R/G pairs were examined. The results indicated DP QPE and PPS Bias errors were not significantly different. However, the RMSE was significantly lower for the DP QPE estimates when compared to the PPS estimates.

Similarly, ten warm season rain events and over 400 R/G pairs were examined using KVNX radar data. As with KOUN, the KVNX results were similar with RMSE values lower for DP QPE. For both KOUN and KVNX, the degree of improvement of the DP QPE RMSE over the PPS increased with heavier rain events/more convectively active weather. This result was expected as the DP QPE was expected to out-perform the legacy PPS system when hail contamination is present.