Assessment of the impact of local surface heterogeneity on atmospheric boundary layer with large-eddy simulations (LES) over remotely-sensed surfaces

A computational study of the atmospheric boundary layer over remotely-sensed heterogeneous terrain is performed using large-eddy simulation (LES). A scale-dependent version of the Lagrangian Dynamic subgrid-scale model has been implemented in the LES thereby improving the predictions of the turbulence structure especially in the near-wall region. The heterogeneous terrain is implemented using remotely-sensed datasets of surface roughness and skin temperature from the USDA-ARS El Reno experiment. The second order structure function is used to determine the characteristic length scale of the various remotely-sensed datasets and LES runs using datasets with a variety of characteristic length scales are performed. Additional simulations using homogeneous boundary conditions are also performed where the boundary condition is prescribed as the mean of the remotely-sensed dataset. A blending height analysis is performed to distinguish between the variability induced by the local surface features and the random variability inherent to boundary-layer turbulence.