Freezing drizzle detection with WSR-88D radars

Freezing drizzle represents a significant in-flight icing hazard and can even cause extensive engine damage to aircraft on the ground. Currently, measurements from WSR-88D radars are being examined in order to develop a radar-based algorithm to identify icing-threat areas from freezing drizzle. In this study, we will present characteristic features of radar reflectivity measurements in freezing drizzle and outline a possible detection scheme. Data analyzed were obtained from a number of freezing drizzle events that occurred at six selected operational radar sites. Radar returns are characterized by the areal-average and standard deviation of radar reflectivity factor as well as the texture of the reflectivity field computed from a Radar Echo Classifier algorithm. Cloud top temperatures from satellite and surface observations supplement the data.

Freezing drizzle typically forms via the collision-coalescence process rather than the classical melting process. Consequently, a reflectivity bright band is generally absent making detection difficult. The similarity of echo structures in freezing drizzle and light snow is also a problem for detection techniques based solely on radar reflectivity. However, detection should improve if additional information such as cloud top temperature and surface conditions are available as they are associated with different microphysical processes. For example, freezing drizzle is common for cloud top temperatures above -10C (Geresdi et al. 2005). The ensemble dataset showed that average reflectivity and its horizontal structure are, to some degree, related to cloud top temperature (CTT). For similar magnitudes, the reflectivity field is typically more horizontally uniform in freezing drizzle than in light snow during warm events (CTT>-10C). The standard deviation and average reflectivity are larger in cold events (CTT<-10C) for light snow, as cellularity in cloud increases; whereas the radar returns in freezing drizzle continue to have a relatively small standard deviation and weak reflectivity. Freezing drizzle is, however, difficult to identify in the presence of multiple cloud layers and in mixed-phase precipitation at the surface. In these cases, polarimetric-based discrimination of hydrometeors may be more useful. Snow particles and drizzle drops have characteristic polarimetric radar returns; thus they can at least be useful in delineating icing threat and non-icing threat areas. Prospects of enhanced freezing-drizzle detection with a polarimetric WSR-88D will be discussed.