MRMS QPE Performance East of the Rockies during the 2014 Warm Season

Abstract: Multiple Radar-Multiple Sensor (MRMS) Quantitative Precipitation Estimate (QPE) Radar Only (Q3RAD), Q3RAD Local Gauge Corrected (Q3gc), Dual Pol, legacy Precipitation Processing System (PPS), and National Center for Environmental Prediction (NCEP) Stage IV (SIV) product performance was evaluated for data collected during the 2014 warm season. Precipitation events from 59 calendar days affecting 55 radars were collected with reference to single radars to highlight the advantages of a mosaic vs. single radar QPE. Product performance was evaluated at a range of 230 km and 100 km from the radar location, the latter distance functioning as a proxy for a best-case scenario mosaic for DPR and PPS estimates. Performance was assessed via the comparison of precipitation estimates to Community Collaborative Rain, Hail and Snow Network (CoCoRaHS) twenty-four hour accumulation gauge data while hourly automated gauge data (with quality controlled measures applied) were used for spatial and time series analysis. QPE data were evaluated over five regions: the Northern/Central Plains, Great Lakes/Midwest, the Northeast, the Southern Plains, and the Southeast/Mid-Atlantic.

Dual Pol QPE performed better than or similar to PPS estimates despite the continuing differential reflectivity calibration challenges. Although Dual Pol QPE challenges in the melting layer were present in some cases, particularly in the Northeast region, they were not as common as was observed during the 2013-2014 cool season. Q3RAD estimates performed better than Dual Pol and PPS over the Northeast, Southern Plains and the Southeast/Mid-Atlantic regions, with overall performance statistics second only to the forecaster quality controlled Stage IV estimates. Over the Great Lakes/Midwest and the Northern/Central Plains, a distinct Q3RAD overestimate bias was noted although Q3RAD remained competitive with the radar-only estimates. The overestimate bias was primarily due to the MRMS system too often classifying radar echoes as tropical. Hence, the 850–500 hPa lapse rate requirement, which affects tropical precipitation classification, was adjusted to favor weak CAPE environments. A test of the new precipitation classification requirement using 15 precipitation events revealed a significant reduction in the overestimate bias; the adjusted code should be operationally implemented before the 2016 warm season. Finally, Q3gc estimates generally improved upon the Q3RAD estimates and in some regions/cases were very close to the Stage IV performance statistics.