Performance Evaluations of Rain Microphysical Retrieval Using Gpm Dual-Wavelength Radar by Way of Comparison With the Self-Consistent Numerical Method

The primary concern of this paper is to evaluate the accuracy of drop size distribution (DSD) retrievals for the rain region using global precipitation measurement (GPM) dual-frequency precipitation radar (DPR). Direct comparison between ground and spaceborne radar might answer the question, but unfortunately, this is a difficult and challenging task that can result in largely unmeasurable uncertainty. In this paper, the self-consistent numerical method (SNM) was used to estimate microphysical parameters along rain profiles. The assessment of the SNM performance characterizes its accuracy in the microphysics retrieval. It follows that the profile data set used for the evaluation plays an important role in this process. The ideal solution would be to use real conditions. For this purpose, a new approach for providing vertical profiles closer to the natural vertical variability of rainfall microphysical characteristics and its global vertical structure can be derived from the observations of NASA dual-frequency, dual-polarization, Doppler radar (D3R)--a matched-beam, $Ku$- and $Ka$-band dual-frequency radar that is similar to the precipitation radar (DPR) on board the GPM core satellite. The profile data set was generated using reflectivity measurements collected in RHI mode, after correcting them for attenuation using the polarimetric measurements of the $Ku$-band. With this data set, the behavior of SNM is evaluated in the presence of the inaccuracy introduced by taking the constant value of μ and the measurement errors. The entire version 4 of the GPM DPR precipitation products obtained with data collected from September 1, 2014 to February 28, 2017 was used. In the presence of bright-band features, the rain region of matched scans, for which measurements were simultaneously collected by $Ku$- and $Ka$-band radars, was employed to retrieve DSD parameters with SNM and GPM algorithms. A direct comparison between the two procedures was performed in terms of merit factors. The results show a significant difference between the performance of the two methods by not taking into consideration the SNM procedures that, instead, are being performed by the GPM algorithms.