Microphysical characteristics of ice crystals and snowflakes as revealed by polarimetric radar measurements

Abstract

This research encompassed both observational and theoretical aspects of copolar and differential reflectivity in the less explored, yet important, winter season precipitation. The observational portion was conducted with the multiparameter, CSU-CHILL radar and supplemented by observers at the Fort Collins Weather Station on the Campus of Colorado State University (FCL) who recorded microphysical features of the snowfall such as snow type, composition, size, and degree of riming. Additionally, a 2-D video disdrometer, located at FCL, made particle size distribution measurements. In order to compare the appropriate radar data with the ground observations, the approximate trajectory of the snow was computed from the height it was interrogated by the radar to the surface. The trajectory, applied in reverse from FCL, identified the source region of the observed snow in the 0.5° and 1° elevation scans of the radar. The results of the observational analyses suggest that nearly homogeneous populations of aggregates can be distinguished from platelike crystals (i.e. dendrites, stellar crystals, and plates) using a combination of co-polar and differential reflectivity (Z and ZoR) radar observations. Furthermore, it appears possible to discern whether or not the platelike crystals are intensely rimed. Additionally, the results challenge the validity of the common assumption that aggregates always produce a ZoR value of O dB. Scattering model studies based on T-matrix theory and the Mueller matrix method were conducted to demonstrate the consistency of the observed radar variables with theoretical values and to test our speculations on which hydrometeor microphysical characteristics were responsible for the observed variations in those variables. Our modeling results suggest several conclusions. First, the aggregate shapes are more relevant and have more impact on ZoR than generally expected; therefore, the modeling assumption that all aggregates are nearly spherical can produce erroneously low ZoR values. Secondly, the size-dependent density formulas for aggregates which predict that bulk densities decrease with size may not always be applicable. Next, the canting of hydrometeors can overcome the influence of microphysical characteristics on Zoa; thus, in a model, it is important to include canting for turbulent situations. when it likely occurred and to exclude it for calmer situations where it was not likely to have occurred. Lastly, the model results validated the idea that intensely rimed platelike crystals could be distinguished from other platelike crystals and highlighted the problems that can be encountered if modelers depend upon size-dependent axis ratio formulas without considering the effects of riming.