Characterizing falling snow using multi-frequency dual-polarization measurements

Multi-frequency observations of snow are becoming increasingly popular because of the potential to retrieve snow habit types. Past studies have shown that three-frequency observations, displayed on a dual-frequency ratio coordinate system can be used to distinguish between different snow types. The NASA D3R, which is a dual-frequency dual-polarization radar, has been deployed along side other operational and research radars at S, C, and X bands, and perhaps in the future along side W-band radars. This study investigates the various combinations of frequencies and polarizations, that can be utilized to perform snow retrievals. Towards that goal, we model dry snowflakes of various types at seven frequency bands: S, C, X, Ku, Ka, W, and 220 Ghz. The modeling results are analyzed with various cluster validation indices to check which combination of frequency, polarization and elevation angle produces better clustering and therefore better characterization of snow. In principle, this can reveal the optimal measurement strategy. Computational results are also compared to aircaft and ground-based measurements during the GCPEx campaign.