Detection of PM2.5 spatiotemporal patterns and driving factors in urban agglomerations in China

As important urban gathering places for new urbanization construction, urban agglomerations (UAs) have long suffered from high concentrations of PM2.5 pollution. By analyzing the changes of the temporal and spatial distribution patterns and the natural and human drivers of PM2.5 in UAs in China can provide an effective reference basis for atmospheric pollution prevention and control. In this study, based on PM2.5 concentration data from 2000 to 2020, PM2.5 pollution in 19 UAs in China was investigated. First, the annual average PM2.5 concentration was used to analyze the temporal variation, spatial distribution, and agglomeration characteristics of PM2.5 pollution in each UA. Then, factor detection was used to analyze the driving mechanisms of the effects of natural and human factors on PM2.5 pollution in UAs in China. Finally, the explanatory intensities of the interactions between different factors of PM2.5 pollution were analyzed based on interaction detection. The results show that from 2000 to 2020, the PM2.5 concentration exhibited an evolutionary trend of “M”-shape. An inflection point occurred in 2007, and 2020 had the lowest value for the study period. PM2.5 of high concentration took the Central Plains and the Beijing-Tianjin-Hebei UAs as the core of pollution. 2013 was an important turning point with a spatial pattern of expansion to the surrounding areas and then contraction to the center. UAs presented an elliptical spatial distribution, and outliers occurred in individual years. Humidity and temperature were the dominant natural factors for the variation of the PM2.5 concentration in 2013, 2018, and 2020, and their q-values were 0.276(humidity), 0.429(temperature), and 0.375(temperature), respectively. The total population and energy consumption were the dominant human factors for these three years, with q-values of 0.150(total population), 0.259(total population), and 0.196(energy consumption), respectively. The explanatory effect of the two-factor interaction was higher than that of a singular factor. Furthermore, the factor and interaction detection of natural factors found them to be more significant driving mechanisms than human factors. The results of this study not only allow for the understanding of the spatiotemporal variation of the PM2.5 concentration in UAs and the effects of different influencing factors, but also provide a scientific reference for the steady development of UAs and air pollution prevention and management in China