Classification of particles in stratiform clouds using the 33 and 95 GHz polarimetric cloud profiling radar system (CPRS)

Abstract

This paper describes the identification of regions of ice, cloud droplets, rain, mixed-phase hydrometers, and insects in stratiform clouds using 33 and 95 GHz radar measurements of reflectivity, linear-depolarization ratio (LDR), dual-wavelength ratio, and velocity from a single-antenna radar system. First, the radar system, experiment, and data products are described. Then, regions are classified using a rule-based classifier derived primarily from LDR, velocity, and altitude. Next, a region-dependent attenuation-correction algorithm is developed to remove attenuation biases in the reflectivity estimate, and histograms of the corrected data are presented for each data product and class. The labeled regions and attenuation-corrected data are then used to train a neural net and maximum likelihood classifier. These agree with the rule-based classifier 96% and 94% of the time, respectively. Finally, the paper evaluates the importance of measuring dual-frequency parameters, velocity, and depolarization ratio.