Evaluation of Dual-Pol Snow QPE Methods as an Alternative to the NEXRAD Snow Accumulation Algorithm

The quantitative precipitation estimation (QPE) method currently used by the NWS during snow events applies a suite of S(Z) relations of the form Z=aS^2, where “a” can vary from 40 to 222. Initial case studies of snow events indicated that Z and the fixed S(Z) relations differ from observations because they do not sufficiently account for variations in snow particle size, density, shape, terminal velocity, and the degree of riming and aggregation. This study leverages polarimetric radar variables, especially KDP, in algorithms that attempt to explicitly treat the various snow particle characteristics. We present multiple possible implementations of dual-pol S(KDP, Z) algorithms and compare them against the SAA S(Z) and observed gauge data. One S(KDP, Z) relation uses fixed snow microphysical parameters, and other S(KDP, Z) relations estimate those parameters dynamically using nearby precipitation and extinction coefficient data from ASOS surface stations. Initial case studies indicated that the S(Z) relations perform better against observations in light snow regimes. The dynamic S(KDP, Z) performs best in heavy snow events where the 24-hour accumulated SWE is greater than 15 mm. These SWE relations will be further tested and refined on more cases encompassing different snow regimes and the outcome of the study should provide improvements to the NWS operational winter quantitative precipitation estimates (QPE) in in the operational WSR-88D product suite and in the Multi-Radar Multi-Sensor (MRMS) system.