Hydrometeor Identification for GPM DPR

 

The GPM science implementation plan articulates the new frontiers of space-based precipitation observations, including new insights into vertical storm structure and microphysics [1]. The DPR algorithms are a key part of the process as discussed in the GPM science implementation the last updated DPR Algorithm Theoretical Basis Document (GPM/DPR ATBD 2021) document. Dual-frequency precipitation radar (DPR) onboard the GPM satellite has extended scan pattern to full swath mode for both Ku- and Ka-band since May 2018.  

 

The objective of the level-2 DPR algorithms are to derive radar only meteorological quantities meaning general characteristics of the precipitation, correction for attenuation and estimation of precipitation water content, rainfall rate and, when dual-wavelength data are available, information on the particle size distributions in rain and snow. The DPR Level-2 algorithms consist of several modules including the classification (CSF) module. Currently, our team plays a key role in the international DPR algorithm development group and is responsible for the dual-frequency algorithms development in the classification module. These algorithms include rain type classification, melting region detection, surface snowfall identification, graupel and hail detection etc. [2]-[4]. 

 

It is a unique advantage for space radar to provide a hydrometeor type over the globe while ground based observations are limited to the regions of deployment. Among the algorithms of current DPR classification module, most of the products are two dimensional with either a “flag” or “type” (or etc.) on a Latitude / Longitude surface. In this research, we will add a range-bin based hydrometeor type for DPR full swath data to enhance the classification module. This three-dimensional hydrometeor identification feature is the next frontier built upon our knowledge of hydrometeor classification development for DPR. 

 

This algorithm has been applied to various precipitation types and validated successfully with either ground based weather radar or airborne weather radar observations from field experiments such as OLYMPEX. Validation for cases with extreme hydrometeor type as hail are also performed with GMI-based approach and illustrate meaningful comparisons.