Retrieval of Microphysical Parameters of Monsoonal rain Using X-band Dual-polarization Radar: Their Seasonal Dependence and Evaluation.

Multiyear measurements from Joss-Waldvogel disdrometer (5 years) and X-band dual-polarization radar (2 years) made at Gadanki (13.5 N, 79.18 E), a low latitude station, are used to i) retrieve appropriate raindrop size distribution (DSD) relations for monsoonal rain, ii) understand their dependency on temperature, raindrop size-shape model and season and iii) assess polarimetric radar DSD retrievals by various popular techniques (Exponential-Exp, Constrained Gamma - CG, Normalized Gamma - N-Gamma and ß methods). The coefficients obtained for different DSD relations for monsoonal rain are found to be different from that of existing relations elsewhere. The seasonal variation in DSD is quite large and significant and as a result the coefficients also vary considerably between the seasons. The slope of the drop size - shape relation, assumed to be constant in several studies, vary considerably between the seasons with warmer seasons showing smaller slope value than cold season. It is found that the constant (0.062) used in linear drop shape models is valid only for cold season. The derived coefficients for CG method for different seasons coupled with those available in the literature reveals that the warm seasons/regions typically have larger curvature and slope values than in cold seasons/regions. The coefficients of mass weighted mean diameter (Dm) - differential reflectivity (ZDR) exhibit strong dependency on drop shape model, while those for the derivation intercept parameter exhibit strong seasonal dependency. Using the retrieved relations and X-band polarimetric radar at Gadanki, four popular DSD methods are evaluated against disdrometer measurements collected over 12 events. All the methods estimated Dm reasonably well with small root mean square error, however failed to estimate intercept parameter accurately. Only N-gamma method estimated the normalized intercept parameter reasonably. Problems associated with specific differential phase (KDP)-based estimates close to the radar location, particularly during overhead convection, are also discussed. [ABSTRACT FROM AUTHOR]