Assessing Atmospheric Boundary Layer Turbulence Over Heterogeneous Urban Canopies via LES-IBM

Accurate representation of surface fluxes and turbulence is critical for numerical weather prediction (NWP) models. However, parameterizing urban turbulence remains a significant challenge due to complex surface heterogeneities. To address this gap, we used high-resolution Large-Eddy Simulation with an Immersed Boundary Method (LES-IBM) to investigate Atmospheric Boundary Layer (ABL) flow over urban canopies. First, the model is validated against the in-house LiDAR measurements located at the University of Houston, reproducing mean wind profiles within the observational range. Next, idealized simulations over flat terrain and various canonical urban layouts (sparse and dense staggered arrays) are performed to isolate morphological effects. Dense canopies increase aerodynamic roughness length (z₀) by up to 200% relative to the typical urban value of 0.8 m in NWP models, while sparse arrays decrease z₀ by about 93%. Dense layouts also elevate turbulent kinetic energy (TKE) and Reynolds stresses within the canopy layer and generate stronger sheltering that increases the total drag by up to an order of magnitude. The canopy geometry influenced the characteristic eddy size and coherence, with dense canopies breaking down the conventional shear-induced coherent turbulent eddy streaks and elevating the ABL height. The overall goal is to characterize the surface heterogeneity impacts on TKE budget terms in the parameter space of roughness Reynolds, Richardson and Rossby numbers in the future. Finally, the eddy viscosity of various urban geometries are also calculated to provide significant inputs for the parameterization of urban canopies in NWP models.