Differentiating Freezing Drizzle and Freezing Rain in HRRR Model Forecasts

Abstract Supercooled large drop (SLD) icing poses a unique hazard for aircraft and has resulted in new regulations regarding aircraft certification to fly in regions of known or forecast SLD icing conditions. The new regulations define two SLD icing categories based upon the maximum supercooled liquid water drop diameter (Dmax): freezing drizzle (100–500 μm) and freezing rain (>500 μm). Recent upgrades to U.S. operational numerical weather prediction models lay a foundation to provide more relevant aircraft icing guidance including the potential to predict explicit drop size. The primary focus of this paper is to evaluate a proposed method for estimating the maximum drop size from model forecast data to differentiate freezing drizzle from freezing rain conditions. Using in situ cloud microphysical measurements collected in icing conditions during two field campaigns between January and March 2017, this study shows that the High-Resolution Rapid Refresh model is capable of distinguishing SLD icing categories of freezing drizzle and freezing rain using a Dmax extracted from the rain category of the microphysics output. It is shown that the extracted Dmax from the model correctly predicted the observed SLD icing category as much as 99% of the time when the HRRR accurately forecast SLD conditions; however, performance varied by the method to define Dmax and by the field campaign dataset used for verification.