Reducing scattering signature ambiguities of frozen hydrometeors

Characterization of cloud composition and understanding of the microphysics and dynamics of atmospheric processes relies upon accurate radar measurements and the associated quality of the classification process to precisely identify the types of frozen hydrometers within a target volume. In principle, the signature of different types of hydrometeors can be interpreted from the polarimetric and multi-frequency measurements of backscattered power and attenuation. However, the applicability of this approach is complicated by the fact that there are many factors (such as size, shape, and material composition) that collectively contribute to the overall radiometric properties. In particular, we will focus here on spaceborne radar applications, where advanced algorithms base their decisions on relative strengths of the radar echo observed at multiple frequencies (i.e. dual-frequency and triple-frequency methods). Due to additional constraints imposed by frequency dependency, these methods are expected to provide more accurate estimates. This expectation hinges on the assumption that the difference in uncertainty inherently resulting from instrument characteristics and calibration processes at different operating frequencies can be resolved. The methods are proven to be effective according to previous simulation studies and field experiments. However, an ambiguity introduced by assumptions in the microphysics properties of the hydrometeors such as shapes and size distribution functions still remains and needs to be quantified.

This work examines the portion of uncertainty in the classification process that results solely from adopted microphysics assumptions and assesses the effectiveness of multiple frequency methods under various configurations including the presence of complex particle shapes and complex size distribution functions.

Electromagnetic scattering properties of simple/complex pristine ice crystals, as well as their aggregates are calculated using Discrete Dipole Approximation (DDA) and T-Matrix methods. A database of scattering signatures is built for a wide range of the particle shapes and sizes. Then effective electromagnetic signatures of the ensembles of hydrometeors are calculated assuming distinct size distribution functions including exponential and gamma probabilistic distribution functions. From this database of electromagnetic signatures, a degree of ambiguity in hydrometeor classification is estimated. Sensitivity of this ambiguity to the prior knowledge of cloud microphysics is also investigated.