High-Resolution Modelling of the Boundary Layer over complex Terrain

The exchange of momentum, heat and mass in the atmospheric boundary layer is primarily governed by thermally and dynamically driven processes that occur on a wide range of scales. This can lead to challenges in accurately describing these processes in Numerical Weather Prediction (NWP) models, especially in mountainous regions. With the ability to use finer grid spacings, more processes can be partially resolved. As a result, some of the commonly used parametrisations are no longer appropriate and need to be adjusted or even deactivated.

The ICOsahedral Non-hydrostatic model is used in Limited Area Mode (ICON-LAM) with the aim of providing a skillful forecast of the Alpine area at 500 m in the framework of GLORI (GLObal to Regional ICON digital twin). The model performs simulations in two-way nesting mode with the operational ICON-D2, including a nest with a horizontal grid spacing of 1 km. This domain covers the entire Alpine region and its horizontal resolution lies within the so-called “grey zone”, where turbulence, drag and convection are neither fully resolved by large-scale processes nor can they be modelled as a small-scale, sub-grid process.

Deterministic experiments are performed for the month of May 2022, as strong convective events have been observed in the Alps during this period. The simulation results are compared with various meteorological stations located at different altitudes within the complex terrain to study the effect of increasing the horizontal resolution and the performance of the model physics on the boundary layer. Small modifications in the grey zone related parametrisations, such as e.g. the turbulent kinetic energy scheme, show that the bias of near-surface variables like temperature and wind speed can be reduced. However, a more sophisticated approach is required for the higher resolution model to represent the complex interactions in the boundary layer.