Exploring Millimeter-Wavelength Radar Capabilities for Raindrop Size Distribution Retrieval

Abstract

Accurate precipitation characterization relies on the estimation of raindrop size distribution (RDSD) from observations. While various techniques using centimeter-wavelength radars have been proposed for RDSD retrieval, the potential of millimeter-wavelength polarimetric radars, offering enhanced spatial and temporal resolution while capturing light to moderate rain, remains unexplored. This study focuses on retrieving the mass-weighted mean diameter D_m using a dual-frequency cloud radar. Since the differential reflectivity Z_dr is ineffective for D_m retrieval at 94 GHz, and simulations demonstrate a strong dependence of the differential backscatter phase d_co on D_m, the estimation of d_co takes precedence in this paper. Notably, d_co remains unaffected by attenuation and polarimetric calibration. Addressing the initial require-ment of disentangling backscattering and propagation effects at millimeter wavelength, an automatic algorithm is proposed to detect Rayleigh plateaus in the spectral domain. Subsequently, a methodology for estimating d_co and its associated error is presented. Leveraging simulation results, confidence intervals for D_m that align with d_co confidence intervals are re-trieved. The assessment of D_m and its confidence interval at 35 and 94 GHz is conducted employing disdrometer-derived D_m. The results demonstrate a comprehensive concordance within a margin of 0.2 mm, underscoring the cloud radar’s efficacy in delineating nuanced variations in the raindrop mean diameter versus altitude. The validation process encounters difficulties for D_m below 1 mm, as the disdrometer-derived D_m may exhibit an overestimation, while the cloud-radar-derived D_m may exhibit an underestimation. The combination of 35 and 94 GHz serves to diminish the confidence interval associated with the retrieved D_m.