Abstract
Green infrastructure (GI) has emerged as an important nature-based solution for mitigating urban air pollution. However, a comprehensive understanding of how various GI types affect air quality remains limited, hindering the development of effective GI-based control measures. Computational fluid dynamics (CFD) modelling provides a powerful tool to explore the fundamental mechanisms by which GI influences air quality in complex urban environments. This paper critically examines the state-of-the-art CFD modelling of GI effects on urban air quality, with a focus on geometric, physical and chemical considerations. Despite its primary importance in capturing the airflow-pollutant interactions, current geometric modelling of GI is often simplified to uniform basic shapes (e.g., cubes or cylinders) while neglecting the morphological and spatial heterogeneities, which may affect the predicted air ventilation and pollutant dispersion patterns. Future studies need to use more realistic geometric representations to improve model fidelity. The performance of physical modelling of GI is largely limited by the empirical parameterisations for aerodynamic effects, pollutant deposition/resuspension/absorption, and biogenic emissions. Further investigations are needed to determine the key parameters of these source terms. Chemically, the reactions between NOₓ, O₃ and biogenic VOCs under varying radiation and shading conditions remain underexplored. By integrating diverse modelling strategies and empirical parameterisations, this paper provides a comprehensive framework to improve the modelling accuracy of the geometric, physical, and chemical interactions between GI and air pollutants in urban micro-environments, supporting the development of more effective GI-based solutions.