A comprehensive understanding of the relationships between PM2.5 concentration and socioeconomic factors provides new insight into environmental management decision-making for sustainable development. In order to identify the contributions of socioeconomic development to PM2.5, their spatial interaction and temporal variation of long time series are analyzed in this paper. Unary linear regression method, Spearman's rank and bivariate Moran's I methods were used to investigate spatio⁻temporal variations and relationships of socioeconomic factors and PM2.5 concentration in 31 provinces of China during the period of 1998⁻2016. Spatial spillover effect of PM2.5 concentration and the impact of socioeconomic factors on PM2.5 concentration were analyzed by spatial lag model. Results demonstrated that PM2.5 concentration in most provinces of China increased rapidly along with the increase of socioeconomic factors, while PM2.5 presented a slow growth trend in Southwest China and a descending trend in Northwest China along with the increase of socioeconomic factors. Long time series analysis revealed the relationships between PM2.5 concentration and four socioeconomic factors. PM2.5 concentration was significantly positive spatial correlated with GDP per capita, industrial added value and private car ownership, while urban population density appeared a negative spatial correlation since 2006. GDP per capita and industrial added values were the most important factors to increase PM2.5, followed by private car ownership and urban population density. The findings of the study revealed spatial spillover effects of PM2.5 between different provinces, and can provide a theoretical basis for sustainable development and environmental protection.
Keywords: Bivariate Moran’s I; PM2.5 concentration; socioeconomic factors; spatial lag model.