Estimating Averaged Wind Characteristics for Generic Urban Neighbourhoods Using Drag-based Approaches

Spatially-averaged wind profiles over the neighbourhood scale are commonly characterized in the literature as exhibiting exponential or logarithmic shapes. Further evaluation of in-canopy averaged velocity and its components—the main aim of this work—can contribute to a more comprehensive characterization of canopy wind conditions. The conventional computational fluid dynamics approach, modeling buildings by fully resolved method (FRM), is taken as the basis to evaluate the drag-based approaches: the porous media model (PMM) and the velocity reconstruction method (VRM). The VRM is particularly proposed as an engineering solution, implemented in Python, that uses drag-velocity functions to inversely calculate the averaged velocities in target neighbourhoods. Both fully-developed flows and flow adjustments through generic neighbourhoods within simplified districts under parallel and oblique inflows are investigated. The neighbourhoods feature moderate size, uniform height, regular and staggered arrangements, and varying plan area density (λ_p, from 0.07 to 0.40). Results reveal that the normalized volumetric drag coefficient (C*_d) varies for the same neighbourhood under different flow conditions, e.g., with a median reduction of up to 20% for regular-medium cases. The comparison demonstrates the PMM’s ability to replicate wind profiles, despite with overestimation for windward neighbourhoods. The qualitative and statistical analyses (via R² and RMSE) confirm that VRM predicts reasonable averaged velocity components, though it tends to slightly overestimate due to neglecting turbulence processes. Drag correlations, fitted in trigonometric Fourier series against wind direction, provide a lightweight program for estimating neighbourhood-averaged velocities, showing promise for integration into urban building energy modeling tools.