Validating physical assumptions of radar-based retrievals of snow microphysics during BAECC snowfall experiment

Radar-based retrievals of snow microphysics rely on a number of assumptions about microphysical properties of snow, i.e. shape and mass of snowflakes, and about their connection to scattering properties of ice particles. Validation of these assumptions is a challenging task and often is limited by the lack of coinciding radar and snow microphysics measurements and associated retrievals. During Biogenic Aerosol Effect of Clouds and Climate, where ARM mobile facility was deployed in Finland, a dedicated intensive observation period focusing exactly on these research questions was carried out. Three radars operating at X, Ka and W –bands were deployed at the University of Helsinki measurement station in Hyytiälä. At the same site a comprehensive suit of precipitation ground instruments was placed. To minimize wind effects on snow measurements a dual fence (DFIR) was build.

To link multi-frequency radar observations to snow microphysics retrievals of physical properties of snowflakes, namely mass-dimensional relations, terminal velocity- diameter and shapes and fall velocity were carried out. To insure that the retrievals are meaningful different methods utilizing different sets of instruments were applied for the retrievals. For example, mass-dimensional relations were retrieved using two different instruments, NASA Particle Image Package and 2D-video disdrometer, by utilizing Böhms method. At the same time these retrievals are validated by comparing measured X-band radar reflectivity and precipitation liquid equivalent accumulation to ones calculated from observed particle size distributions and retrieved m–D. A number of other cross checks was applied to insure consistence of different retrievals. After this step, different aspects of multi-frequency radar retrievals of snow microphysics were tested. For example, a connection between dual-frequency ratios and median volume diameter in a variety of snow cases was checked. It was found that performance of X/Ka and Ka/W retrievals differ and the biggest difference is observed for low-density aggregates. Other findings of this study will also be presented at the conference.