Multi-nested WRF simulations for studying planetary boundary layer processes on the turbulence-permitting scale in a realistic mesoscale environment

摘要

The Weather Research and Forecasting (WRF) model was applied in a nested configuration from a 2.7?km convection-permitting domain via grey-zone resolutions of 900?m and 300?m down to the 100?m turbulence-permitting scale. Based on sensitivity studies, this approach was optimized to investigate the evolution of small-scale processes in the PBL for a clear sky case during the HOPE experiment in western Germany on 24 April 2013. The results were compared with theoretical and experimental findings from literature and high-resolution lidar observations collected during the campaign. Simulations with parameterized turbulence were able to capture the temporal evolution of the PBL height, but almost no internal structure was simulated in the boundary layer. Only the turbulence-permitting simulations were capable of reproducing the morning transition from the stable nighttime to the daytime convective boundary layer and the following break-up into turbulent eddies. Comparisons with lidar data showed that the turbulence-permitting simulations reproduced the observed turbulence statistics. Nevertheless, the potential temperature in the boundary layer was 1?K cooler than observed, caused by a lower surface temperature mixed upward by the turbulent eddies. The simulated PBL height was underestimated by 200?m, reflected in a well-captured profile of specific humidity up to a height of 900?m and an overly strong decrease of moisture above. The general shape of the variance profiles of potential temperature and specific humidity were captured by the model. However, the simulated variability throughout the boundary layer was lower and the different heights of the variance peaks indicated that the model may not fully capture the turbulent processes at the top of the boundary layer. Identifying those systematic differences between nested simulations and observations demonstrated the value of this model approach for process studies and parameterization tests.