A129-11
Analysis and Mitigation of Occasional Precipitation-Type Problems in NCEP Global Forecast System

Friday, 11 December 2020: 11:00
Virtual
Weizhong Zheng1, Jongil Han2, John S Kain3, Geoffrey Manikin3, Ruiyu Sun1, Fanglin Yang3, Vijay Tallapragada3, Jian-Wen Bao3 and Michael Barlage3, (1)IMSG at NOAA/NCEP/EMC, College Park, MD, United States, (2)SRG at NOAA/NCEP/EMC, College Park, MD, United States, (3)NOAA/NCEP/EMC, College Park, MD, United States
Abstract:
Recent diagnostic analyses by the National Centers for Environmental Prediction (NCEP) Model Evaluation Group (MEG) revealed that the NOAA Global Forecast System (GFS.v15) operational forecasts occasionally presented significant spurious warming in the lower troposphere during cold-season precipitation events. This behavior was detected in several events in the U.S. Midwest and Great Plains. In these cases, model-predicted warming led to inaccurate predictions of precipitation type, particularly a significant under-prediction of the areas where precipitation fell as snow. This presentation summarizes a diagnostic study of the mechanisms for the spurious warming through a series of experiments involving the planetary-boundary-layer (PBL) and microphysics parameterization (MP) schemes.

The GFS.v15 has 64 vertical levels and applies the GFDL MP scheme and the hybrid eddy-diffusivity mass-flux (K-EDMF) PBL scheme that includes dissipative heating and modified stable boundary layer mixing. In the series of experiments presented in this study, alternative parameterizations are used, including the Zhao-Carr and WSM6 MP schemes, along with the turbulent kinetic energy based EDMF (TKE-EDMF) PBL scheme, which is a 1.5-order closure scheme with a prognostic TKE for the eddy diffusivity and is slated for implementation in GFS.v16 in early 2021. In addition to the different PBL and MP schemes, sensitivities to both the GFS.v15 and the 127-level GFS.v16 configuration are tested.

It was found that anomalous lower-tropospheric warming was responsible for the erroneous precipitation-type diagnosis and the warming was caused by interactions between the K-EDMF PBL and GFDL MP schemes. It was determined that the overly vigorous mixing could be substantially eliminated by modifying one component of the K-EDMF scheme’s K-mixing profile. Nevertheless, further study needs to be conducted to understand what specific environmental conditions and model-physics interactions lead to this undesirable model behavior.

<< Previous Abstract | Next Abstract >>