ACINN Graduate Seminar - WS 2024/25

2025-01-08 at 12:00 (on-line & on-site)

Utilizing high-resolution, full-physics simulations to unravel mountain boundary layer processes

Brigitta Goger

Center for Climate Systems Modeling, ETH Zurich, Zurich, Switzerland

Together with the rise of computational power, numerical weather prediction (NWP) models can be applied at hectometric and sub-hectometric grid spacings over mountainous terrain. Given the fine horizontal grid spacing, major topography features are resolved realistically in the model, which allows us to simulate mountain boundary layer processes such as orographically-induced gravity waves and thermally-induced flows explicitly.
In this presentation, we will present two model setups over complex terrain: First, we show large-eddy simulations (LES) at dx=48m with the WRF model over the Hintereisferner glacier, Austria. We validate the simulations with eddy-covariance stations from the HEFEX 1 experiment, and show that the local glacier boundary layer development and surface sensible heat fluxes strongly depend on the synoptic flow direction. In case of a North-Westerly flow direction, a gravity wave forms over the glacier valley, leading to the erosion of the local glacier boundary layer. As a second step, we modify the ice surfaces in the model to investigate whether glaciers nearby influence the formation and breaking pattern of the gravity wave over the glacier surface, and henceforth the response of the local boundary layer itself.
The second model setup we will show is a first model evaluation study of the ICON model (the operational NWP model of MeteoSwiss) over truly complex terrain, in the Inn Valley, Austria. We utilize the observations from the CROSSINN campaign to validate the model performance at horizontal grid spacings of 1km, 500m, 250m, and 125m, and identify resolution-dependent current challenges and shortcomings (e.g., the surface layer scheme and turbulence parameterizations). We evaluate the representation of thermally-induced flows such as the up-valley wind in the model, but also show the impact of the horizontal grid spacing on the formation of mid-level clouds.
For further analysis, we apply the multi-resolution Coherent Spatio-Temporal Scale Separation (mrCOSTS) method to vertically staring LIDAR observations of vertical velocity to identify coherent structures and the dominant spatial and temporal scales in MoBL flow. Applying the method to simulation data enables us to identify how coherent structures are represented at various mesh sizes in the model. Our findings suggest that scale interactions between large-scale coherent structures and smaller-scale flow features strongly impact the correct formation of the MoBL. This model validation aims to give a performance baseline for a large LES domain over Switzerland, which is currently experimental.

 

The content of this webpage is the author’s opinion and the author’s intellectual property. ACINN has the author’s consent to make this information available on the webpage for the announcement of the seminar presentation and in the seminar archive.