[Spatial-temporal Relationships Between Urbanization Levels and Air Pollution Across Various Stages of Urbanization Nationwide].

This research investigates the spatial and temporal relationship between urbanization levels and air pollution in cities at different stages of urbanization in China, highlighting its significance for guiding cities towards green development with reduced pollution and carbon emissions. The study uses a range of datasets from 2005 to 2020, including per capita GDP raster data, land use type data, remotely sensed PM_2.5 and O₃ concentration data, and meteorological raster data. The urbanization stages for the years 2005, 2010, 2015, and 2020 were classified using the Chenery standard, facilitating a nuanced analysis of urban growth patterns. A one-way analysis of variance（ANOVA）was employed to examine the significance of differences in PM_2.5 and O₃ concentrations across urbanization stages, revealing distinct pollution profiles. Furthermore, multi-scale geographically weighted regression（MGWR）was applied to quantitatively analyze the spatial and temporal correlations between urbanization levels and the concentrations of PM_2.5 and O₃, offering insights into the complex dynamics at play. The findings indicate a progression through six urbanization stages from 2005 to 2020. In 2005, 110 cities were in the primary product stage （PPS）, and 118 were in the primary industrialization stage （PIS）. By 2010, the urbanization phase had shifted predominantly towards industrialization, with 139 cities in the medium-term industrialization stage （MIS） and 88 in the late industrialization stage （LIS）. The trend continued towards advanced stages, with the majority of cities in 2015 and 2020 being in the middle to late industrialization and developed stages. The number of cities in the primary developed stage （PDS）and the developed stage （DS）reached 80 and 91, respectively. The spatial distribution of PM_2.5 and O₃ concentration trends and their average values at different urbanization stages showed significant variance. From PPS to DS, the average PM_2.5 concentration initially rose and then declined, with concentrations during the industrialization stage higher than in the primary and developed stages. In contrast, the average O₃ concentration trended upward across all stages, reaching its peak in the developed stage. The MGWR results identified significant regional variations in the impact of urban built-up area proportions on PM_2.5 and O₃ concentrations. High-value areas for PM_2.5 regression coefficients in 2005 and 2010 were predominantly found in the Yunnan-Guizhou-Sichuan urban cluster, extending northeast by 2015 and 2020 to cover most of China. Conversely, high-value areas for O₃ regression coefficients from 2005 to 2020 were mainly in western and central China, with eastern regions, particularly in the south, showing significantly lower coefficients, indicating a negative correlation overall. Synergistic analysis of the data revealed that cities with concurrent increases in PM_2.5 and O₃ concentrations in 2005 and 2010 were concentrated in the Yangtze River Delta, Yunnan-Guizhou-Sichuan, and Shaanxi-Gansu-Ningxia regions. By 2015 and 2020, such cities were more broadly distributed across central and eastern China, highlighting the evolving nature of urban air pollution in relation to urbanization.