Work

Impact of spatial representativeness of aerodynamic parameters on urban street-canyon wind simulations

Year: 2026
Type: article
Abstract: Accurate simulation of urban wind fields is critical for street-scale air quality modeling, yet remains challenged by the spatial heterogeneity of real cities. Street-network models enable citywide simulations at reasonable cost, and their aerodynamic parameterizations, which are often derived from idealized geometries, must be rigorously evaluated under realistic urban conditions. In this study, we evaluate wind parameterizations in the street-network model MUNICH using high-resolution CFD simulations (OpenFOAM, code_saturne, and PALM) over central Paris as references. Sensitivity experiments examine the spatial representativeness and physical consistency of key aerodynamic input parameters, including roughness length ( z 0 ), displacement height ( d ), and friction velocity ( u ∗ ), under both no-tree and with tree conditions. Results show that local-scale d and z 0 improve the representation of urban morphological heterogeneity. Moreover, both the roughness parameters and the CFD-derived u ∗ provide a more consistent scaling of rooftop wind speed and vertical wind attenuation compared to city-averaged geometry and regional-scale-based inputs. However, improvements from individual parameters alone remain limited and can even degrade performance when inconsistencies arise between their spatial scales and physical meanings. The best agreement with CFD is achieved when locally representative d and z 0 and consist u ∗ are combined, reducing global-average wind speed bias by about 1.0 m s −1 (64%) in no-tree conditions, and substantially improving the representation of wind attenuation from tree-lined streets, with a bias decrease of up to 53%. This study highlights that improving individual aerodynamic inputs is insufficient without accounting for their spatial representativeness and mutual consistency. Harmonizing the scale and physical basis of input parameters is essential for enhancing the reliability of urban wind simulations and street-scale dispersion modeling.
Language: English
Open Access status: hybrid
Estimated APC paid: $3,690
OpenAlex ID: W7233554537
FWCI:
Cites: 55
Cited by:
Related to: