Analysis of the city-scale wind environment and detection of ventilation corridors in high-density metropolitan areas based on CFD method

Abstract

Urban wind corridors can improve the air exchange and ventilation within the city center and mitigate high UHIs. The current study simplified the complex underlying surface using a GIS clustering analysis model to establish the 3D digital model for CFD simulation, the Reynolds-averaged Navier–Stokes (RANS) model was used to simulate the city-scale wind environment with a horizontal resolution of 30 m. Machine learning methods were then employed to predict how ventilation potential could be improved within the identified wind corridors. The results are summarized as follows: (1) Building coverage ratio (BCRz) and floor area ratio (FAR) exerted significant influences on both horizontal and vertical wind fields, leading to variations in wind speed and direction. (2) BCRz and FAR have negative correlations with the average wind speed (Umean) at the heights ranging from 5 to 50 m above the ground, while NDVI and green plot ratio (GR) show positive correlations. (3) 14 potential ventilation corridors were detected based on the CFD simulations. Targeted recommendations to enhance urban ventilation were validated by machine learning methods, emphasizing adjustments to urban morphology and landscape types. These findings provide actionable insights for urban planning and design strategies aimed at improving ventilation in high-density metropolitan areas.