Associations between Google Street View-derived urban greenspace metrics and air pollution measured using a distributed sensor network. (December 2022)

• Record Type: Journal Article
• Authors: O'Regan, Anna C.; Byrne, Rosin; Hellebust, Stig; Nyhan, Marguerite M.
• Is Part Of: Sustainable cities and society. Volume 87(2022)
• Publication Date: 2022-12
• Subjects: Urban greenspace -- Google street view -- Normalized difference vegetation index -- Air pollution -- Air quality sensors -- Distributed network of sensors; Sustainable urban development -- Periodicals; Sustainable buildings -- Periodicals; Urban ecology (Sociology) -- Periodicals; 307.76
• DOI: 10.1016/j.scs.2022.104221
• Languages: English
• ISSNs: 2210-6707

Abstract

Highlights: · Urban greenspace was derived using Google Street View imagery and computer vision methods. · Air pollution was monitored using a distributed sensor network. · Urban greenspace was inversely associated with PM2.5 and PM10 . · The findings will have implications for urban planning and environmental policy. Abstract: With urbanisation occurring at an unprecedented rate worldwide, there is an urgent need to address air pollution-related health risks in urban areas. Although urban greenspace presents an opportunity to balance those risks, little research investigates the association between urban greenspace and air pollution. As such, in this first-of-a-kind study, we examined associations between street-level urban greenspace and air pollution metrics at multiple spatial scales using largescale Information Communications Technology (ICT) derived environmental datasets in an urban area. Urban greenspace was modelled using Google Street View images and computer vision methods in high spatial resolution and termed the Green View Index (GVI). Another greenspace metric, the Normalised Difference Vegetation Index, was quantified using Landsat satellite imagery and computational algorithms. Particulate matter (PM) air pollution of varying size fractions (PM2.5, PM10, PM1 ) was measured using a distributed network of air pollution sensors. Higher levels of urban greenspace were associated with decreases in PM air pollution. For example, an interquartile range (5.44) increase in GVI, within a buffer radius of 2000m around air pollution sensors, was significantly ( p <0.01) associated with a 1.94μg/m 3 [95% Confidence Interval: -3.03, -0.85] decrease in PM2.5 . The findings will inform the planning and design of smart, sustainable and healthy future cities. Graphical abstract: Image, graphical abstract