Large Eddy Simulation of Realistic Wind Fields in Daytime Atmospheric Boundary Layer

We carried out large eddy simulations (LES) of three cases of daytime convective boundary layer (CBL) evolution at two field data collection sites in Oklahoma. Boundary layer depth, mean temperature and wind profiles from LES were compared with sonde, profiler, and lidar observations and with Weather Research and Forecasting (WRF) model predictions. Three methods of accounting for the large scale forcing in the LES were tested: 1) imposing timeconstant large-scale pressure gradient force; 2) including temporal changes in the large-scale pressure gradient force and advection of temperature, humidity, and horizontal velocity components; and 3) applying a simpler force-restore technique that assumes the large-scale forcing is directly proportional to the deviation of horizontally averaged LES variables from their observed/modeled counterparts. Comparisons between LES and measurements revealed that the CBL evolution was influenced by large scale heterogeneity in fluxes at the underlying surface that were not included in LES. Additionally, neither temporal changes in the large-scale pressure gradient nor the effects of advection could be neglected. However, the straightforward incorporation of both these forces in the simulations did not greatly improve predictions of the mean flow fields. The force-restore method provided the only reasonably accurate depiction of mean wind, temperature, and humidity profiles.