Investigating the influence of urban land use and landscape pattern on PM2.5 spatial variation using mobile monitoring and WUDAPT

• Published in: Landscape and urban planning Vol. 189; pp. 15 - 26
• Main Authors: Shi, Yuan, Ren, Chao, Lau, Kevin Ka-Lun, Ng, Edward
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.09.2019
• Subjects: Air pollution; Land use; Landscape pattern; Mobile monitoring; WUDAPT; Landscape pattern; Air pollution; Mobile monitoring; WUDAPT; Land use
• ISSN: 0169-2046; 1872-6062
• DOI: 10.1016/j.landurbplan.2019.04.004

•Investigating the spatial variation of PM2.5 in a compact urban scenario.•Mobile monitoring method was adopted to achieve a better spatial understanding.•An application of LCZ scheme and WUDAPT level 0 product in urban air quality study.•Land use/landscape pattern metrics were adopted as the predictors.•Land use/landscape patterns are influential to the spatial variation of PM2.5. Particulate matter that <2.5 µm in aerodynamic diameter (PM2.5) has been recognized as one of the principal pollutants that degrades air quality and increases health burdens. In this study, we employ the MLR and GWR modelling method to obtain estimation models for PM2.5 with a set of land use/landscape metrics as predictor variables. The study focused on investigating the influence of urban land use and landscape pattern on PM2.5 spatial variation, specifically, on identification of influential landscape classes/types that regulate PM2.5 concentration levels. The spatial PM2.5 concentration in the compact urban scenario of Hong Kong was sampled by conducting a series of mobile monitoring campaigns. The Local Climate Zone (LCZ) Scheme and World Urban Database and Portal Tools (WUDAPT) level 0 database were adopted as the basis of the calculation of land use/landscape metrics. These metrics were then adopted as the predictors to explain the spatial variations in PM2.5. 62% of the variance in PM2.5 can be explained by the resultant GWR model using only five land use/landscape classes, and without using any traffic-related variables or data from emission inventory. The findings can inform the urban planning strategies for mitigating air pollution and also indicate the usefulness of LCZ and WUDAPT in estimating the spatial variation of urban air quality.