Inter-Comparisons of Radar-Based QPE Performance for Continental, Stratiform and Tropical Rain Regimes

Following a year-long evaluation of the operational Dual Pol, legacy precipitation processing system (PPS), multi-radar multi-sensor (MRMS) Q3RAD and the developmental MRMS Q3 Dual Pol (Q3DP) quantitative precipitation estimate (QPE) products across the CONUS US, an inter-comparison of QPE performance for continental, stratiform and tropical rainfall events was made to better understand the assessment results. QPE performance was examined for samples of continental (strong to severe convective events), stratiform (precipitation characterized by horizontally homogenous reflectivity and little convection) and tropical (tropical cyclone landfall and remnants thereof) rainfall. For the continental rainfall sample the operational Dual Pol QPE exhibited a better median bias ratio than PPS, Q3RAD and Q3DP QPEs. Q3DP exhibited the lowest errors although Dual Pol exhibited lower errors than PPS and similar errors to Q3RAD. The tropical and stratiform rainfall samples showed the operational Dual Pol QPE exhibited significant underestimates and either similar or higher errors than PPS QPE. However, the mosaicked MRMS Q3DP and Q3RAD QPEs exhibited better bias ratios & significantly lower errors than Dual Pol and PPS QPEs. Overall, the Dual Pol QPE results did not appreciably change even when differential reflectivity calibration or poor radar data quality was taken into consideration. Further examination indicated forecaster choice of what rain rate relations were used during some precipitation events sometimes significantly impact Dual Pol or PPS QPE performance.

For the Q3DP, the overestimate bias observed in some convective systems was chiefly related to 1) the use of a net value of alpha, key to estimating specific attenuation fields, for the radar field of view; 2) ice contamination from melting hail; 3) evaporation of smaller raindrops when appreciable amounts of dry air was near convection. Mitigation of factor 3) caused Q3DP to be competitive with the Dual Pol QPE bias ratio for continental events but with significantly lower errors. An updated Q3DP QPE product that takes into account the lessons learned from this evaluation will be implemented operationally.