A Locally Smoothed Terrain‐Following Vertical Coordinate to Improve the Simulation of Fog and Low Stratus in Numerical Weather Prediction Models. (19th August 2021)

• Record Type: Journal Article
• Authors: Westerhuis, Stephanie; Fuhrer, Oliver
• Is Part Of: Journal of advances in modeling earth systems. Volume 13:Number 8(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-08-19
• Subjects: fog -- low stratus -- numerical weather prediction -- vertical coordinate formulation; Geological modeling -- Periodicals; Climatology -- Periodicals; Geochemical modeling -- Periodicals; 551.5011
• DOI: 10.1029/2020MS002437
• Languages: English
• ISSNs: 1942-2466

Abstract

Abstract: The correct simulation of fog and low stratus (FLS) is a difficult task for numerical weather prediction (NWP) models. The Swiss Plateau experiences many days with FLS in winter. Most NWP models employ terrain‐following vertical coordinates. As a consequence, the typically flat cloud top is intersected by sloping coordinate surfaces above hilly terrain such as the Swiss Plateau. Horizontal advection across the sloping coordinate surfaces leads to spurious numerical diffusion which promotes erroneous FLS dissipation. To address this problem, we propose a new vertical coordinate formulation which features a local smoothing of the model levels. We demonstrate the positive impact of the new vertical coordinate formulation on a case study in detail and for a full month using the COSMO model. The improved vertical coordinate formulation is not yet sufficient to obtain perfect FLS forecasts, it is however a crucial aspect to consider on the way thereto. Plain Language Summary: In Switzerland, the Swiss Plateau is prone to occurrence of fog and low stratus clouds (FLS) in winter. High‐resolution weather prediction models are an important tool to predict FLS. However, they often struggle to provide accurate FLS forecasts. Among other model aspects which need to be improved, one issue is the determination of the computational mesh: Most weather models employ a mesh which follows the terrain at the Earth's surface, leading to grid cells that are tilted with respect to the horizontal at altitudes where FLS occur. The structure as well as the top of a FLS layer typically is horizontal. Thus, the cells of the computational mesh are at an angle with respect to the dominant physical processes that the model has to represent. We show that a sloping mesh is detrimental for the forecasting of FLS since it promotes premature dissipation of the clouds. For a widely used weather model, we propose a new way to determine the computational mesh leading to a smoother and flatter computational mesh over the Swiss Plateau. The model is still not able to yield perfect FLS forecasts, but in some cases the use of the new mesh leads to considerable forecast improvements. Key Points: Terrain‐following vertical coordinates feature sloping vertical coordinate surfaces in the atmospheric boundary layer Spurious numerical diffusion associated with advection across sloping coordinate surfaces promotes dissipation of fog and low stratus Local smoothing of the vertical coordinate surfaces improves forecasts of fog and low stratus