Efficient city-scale wind mapping from building morphology: A CFD-based parameterization scheme

Abstract

The expansion and densification of urban areas weaken urban winds significantly. To provide a rapid and accurate estimation of city-scale wind, this study proposes a parameterization scheme. It employs two urban morphological parameters, namely plan area density ${\lambda }_p$ and frontal area density ${\lambda }_f$, to calculate wind speeds at both the urban canopy layer and pedestrian levels. The urban wind speeds under various morphologies are derived from CFD simulations. Then, the relationships between ${\lambda }_p$ and ${\lambda }_f$ and urban wind speed are established using nonlinear regression. The parameterization scheme is evaluated by comparing it with the multi-layer urban canopy model (MLUCM) and CFD simulations of real urban geometries. The results show that the MLUCM exhibits systematic errors in wind speed predictions, as it neglects the impact of ${\lambda }_p$. Moreover, the proposed parameterization scheme accurately reproduces wind speeds in complex geometries, although it is derived from idealized urban geometries. As an implementation, this parameterization scheme is used to map the wind environment in Guangzhou city. Due to its high building density, Guangzhou shows poor pedestrian-level wind environments in most areas. This study facilitates the identification of low wind speed hotspots and the development of mitigation strategies to enhance urban wind and thermal environments.