A Methodology to Derive Radar Reflectivity–Liquid Equivalent Snow Rate Relations Using C-Band Radar and a 2D Video Disdrometer

The objective of this work is to derive equivalent radar reflectivity factor-liquid equivalent snow rate (Z sub(e)-SR) power-law relations for snowfall using the C-band King City operational weather radar and a 2D video disdrometer (2DVD). The 2DVD provides two orthogonal views of each snow particle that falls through its 10 cm 10 cm virtual sensor area. The 'size' parameter used here for describing the size distribution is based on the 'apparent' volume computed from the two images, and an equivolume spherical diameter D sub(app) is defined. The determination of fall speed is based on matching two images corresponding to the same particle as it falls through two light planes separated by a precalibrated separation distance. A new 'rematching' algorithm was developed to improve the quality of the fall speed versus D sub(app) as compared with the original matching algorithm provided by the manufacturer. The snow density is parameterized in the conventional power-law form , where alpha and beta are assumed to be variable. To account for strong horizontal winds that tend to decrease the measured concentrations from the 2DVD, a third parameter gamma is introduced. The methodology estimates the three parameters ( alpha , beta , and gamma ) by minimizing the difference between the radar-measured reflectivity and the equivalent reflectivity computed from the 2DVD in a least squares sense. The optimally determined values of alpha , beta , and gamma are used to estimate the SR and the coefficient and exponent of the Z sub(e) = a(SR) super(b) relation. For validation, the accumulation from the SR is compared with the manually recorded accumulations from the double-fence international reference (DFIR) gauge. The data were collected during the Canadian Cloudsat Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) Validation Project (C3VP) conducted in Ontario, Canada, during the 2006/07 winter season. A total of seven snow days were analyzed and the accumulation intercomparisons gave a fractional standard deviation of 26% and normalized bias 2.1%. The range of the a and b values for the seven days appear reasonable and similar to conventional Z sub(e)-R relations.