Modelling the Effects of Vegetation on Urban Air Quality Based on Multiple Environmental Data Sources

Air pollution is one of the biggest threats to public health worldwide. Vegetation has been shown to have some impact on urban air quality. Quantifying the effects of vegetation on air pollutant levels in urban areas is challenging due to the complex configurations of real cities. Such knowledge and methods are essential to urban vegetation planning and design that aims to improve urban air quality. Using high-resolution computational fluid dynamics (CFD) simulation to control a large set of variables is one of the existing methods to quantify vegetation effects. However, CFD simulation is often limited by the computing resources available for detailed modelling of the entire sphere of a real urban environment. This paper presents an alternative modelling framework for assessing vegetation effects on urban air quality based on the real urban environment data, collected from multiple data sources including satellite and ground-based sensing and imaging. First, based on the Multiangle Imaging Spectroradiometer (MISR) satellite aerosol optical depth (AOD) datasets and air pollutants datasets supplied by ground monitoring stations, a Random Forest based method is applied to generate distribution of PM2.5 concentrations at a high spatial resolution covering a large urban area. Then, an urban grid based clustering of non-vegetation related urban form (e.g., road length) is performed. Next, among the urban clusters, a deep learning image segmentation method is applied to a large set of street view images to capture the vegetation morphology characteristics for tracking the relationship with PM2.5 concentrations. The modelling framework was first tested with Harbin, a major city in northeast China suffering poor air quality. This research aims to provide a more effective and robust approach to quantifying the impact of urban vegetation on air quality in real urban environments.

Sustainable Cities: Pioneering Approaches to Green Urbanism and Climate Resilience