Hydrometeor Identification from W-band Polarimetric Measurements

Automatic classifications of radar polarimetric measurements into hydrometeor types using fuzzy-logic or other similar algorithms are now routinely done using ground-based cm-wavelength polarimetric radars (e.g., Vivekanandan et al. 1999; Lim. et al. 2005). In most of these algorithms, the polarimetric thresholds for various hydrometeor types are based on computational studies. A comprehensive review of polarimetric thresholds used for hydrometeor identification from cm-wavelength radar is given by Straka et al. (2000). Most of the cm-wavelength based algorithms are focused on large-scale systems and typically lump ice crystals into 2-3 broad categories. However, as both observational and computational studies for mm-radar show (e.g., Wolde and Vali, 2001; Tang and Aydin 1995), ice crystals have diverse polarimetric signatures that depend on crystal shape, size, density and fall patterns. By using near coincident in-situ and airborne radar data, it is possible to refine the threshold obtained from computational studies and add more classes (mixed phase, mixed particle types, rimed vs. pristine crystals etc). This paper uses nearcoincident (within 100 m) W-band polarimetric radar and in-situ cloud microphysics measurements in developing fuzzy-logic based hydrometeor classification beyond what has been reported in the past.