Identifying Precipitation Types Using Dual-Polarization-Based Radar and Numerical Weather Prediction Model Data

Abstract:

The recent upgrade of dual-polarization with NEXRAD radars has assisted in improving the characterization of microphysical processes in precipitation and thus has enabled precipitation estimation based on the identified precipitation types. While this polarimetric capability promises the potential for the enhanced accuracy in quantitative precipitation estimation (QPE), recent studies show that the polarimetric estimates are still affected by uncertainties arising from the radar beam geometry/sampling space associated with the vertical variability of precipitation. The authors, first of all, focus on evaluating the NEXRAD hydrometeor classification product using ground reference data (e.g., ASOS) that provide simple categories of the observed precipitation types (e.g., rain, snow, and freezing rain). They also investigate classification uncertainty features caused by the variability of precipitation between the ground and the altitudes where radar samples. Since this variability is closely related to the atmospheric conditions (e.g., temperature) at near surface, useful information (e.g., critical thickness and temperature profile) that is not available in radar observations is retrieved from the numerical weather prediction (NWP) model data such as Rapid Refresh (RAP)/High Resolution Rapid Refresh (HRRR). The NWP retrieved information and polarimetric radar data are used together to improve the accuracy of precipitation type identification at near surface. The authors highlight major improvements and discuss limitations in the real-time application.