Incorporating three immersed boundary methods into the WRF-LES model to simulate atmosphereic processes over complex terrain

Large eddy simulations of atmospheric processes over complex terrain using the Weather Research and Forecasting (WRF-LES) face grand challenges due to incapability in dealing with complex topography. Here three immersed boundary methods (IBM) are implemented into WRF-LES, including the linear velocity reconstruction scheme, the log-law velocity reconstruction scheme, and the stress reconstruction scheme, to enhance the simulation capability of WRF-LES over complex topography. These schemes were examined extensively over a flat roughness surface, two-dimensional hills, and two field experiments (Askervein and Bolund hills). Our results show that the stress reconstruction scheme performed best in reproducing speed-up, followed by the linear velocity reconstruction scheme and the log-law velocity reconstruction scheme. The resolution sensitivity tests indicated that increasing the grid resolution can significantly improve LES simulations of atmospheric processes over complex terrain. On the lee side of the hill, all the simulations tended to underestimate the wind speed compared with observation. Compared with other studies, the overall simulations with IBM show promising results.