Particulate matter with an aerodynamic equivalent dimeter less than 2.5 μm (PM2.5) and ozone (O3) are major air pollutants, with coupled and complex relationships. The control of both PM2.5 and O3 pollution requires the identification of their common influencing factors, which has rarely been attempted. In this study, land use regression (LUR) models based on the least absolute shrinkage and selection operator were developed to estimate PM2.5 and O3 concentrations in China's Pearl River Delta region during 2019. The common factors in the tradeoffs between the two air pollutants and their synergistic effects were analyzed. The model inputs included spatial coordinates, remote sensing observations, meteorological conditions, population density, road density, land cover, and landscape metrics. The LUR models performed well, capturing 54-89% and 42-83% of the variations in annual and seasonal PM2.5 and O3 concentrations, respectively, as shown by the 10-fold cross validation. The overlap of variables between the PM2.5 and O3 models indicated that longitude, aerosol optical depth, O3 column number density, tropospheric NO2 column number density, relative humidity, sunshine duration, population density, the percentage cover of forest, grass, impervious surfaces, and bare land, and perimeter-area fractal dimension had opposing effects on PM2.5 and O3. The tropospheric formaldehyde column number density, wind speed, road density, and area-weighted mean fractal dimension index had complementary effects on PM2.5 and O3 concentrations. This study has improved our understanding of the tradeoff and synergistic factors involved in PM2.5 and O3 pollution, and the results can be used to develop joint control policies for both pollutants.
Keywords: Land use regression; O(3); PM(2.5); Tradeoffs and synergies.
Copyright © 2021 Elsevier Ltd. All rights reserved.