Sensitivity of simulated snow cloud properties to mass-diameter parameterizations.

Abstract:

Mass to diameter (m-D) relationships are used in model parameterization schemes to represent ice cloud microphysics and in retrievals of bulk cloud properties from remote sensing instruments. One of the most common relationships, used in the current Global Precipitation Measurement retrieval algorithm for example, assigns the density of snow as a constant tenth of the density of ice (0.1g/m^3). This assumption stands in contrast to the results of derived m-D relationships of snow particles, which imply decreasing particle densities at larger sizes and result in particle masses orders of magnitude below the constant density relationship.

In this study, forward simulations of bulk cloud properties (e.g., total water content, radar reflectivity and precipitation rate) derived from measured size distributions using several historical m-D relationships are presented. This expands upon previous studies that mainly focused on smaller ice particles because of the examination of precipitation-sized particles here. In situ and remote sensing data from the GPM Cold season Experiment (GCPEx) and Canadian CloudSAT/Calypso Validation Program (C3VP), both synoptic snowstorm field experiments in southern Ontario, Canada, are used to evaluate the forward simulations against total water content measured by the Nevzorov and Cloud Spectrometer and Impactor (CSI) probe, radar reflectivity measured by a C band ground based radar and a nadir pointing Ku/Ka dual frequency airborne radar, and precipitation rate measured by a 2D video disdrometer.

There are differences between the bulk cloud properties derived using varying m-D relations, with constant density assumptions producing results differing substantially from the bulk measured quantities. The variability in bulk cloud properties derived using different m-D relations is compared against the natural variability in those parameters seen in the GCPEx and C3VP field experiments.