For a long time, air pollution caused by unreasonable urban spatial structure and excessive urban sprawl has been a prominent environmental problem in China. From the level of all cities, three economic zones and different city scales, panel data of 194 prefecture-level cities in China from 2006 to 2017 were used to construct a dynamic panel model and to analyze the impact of urban spatial structure on SO 2 , industrial smoke and dust emissions. The results showed that: (1) air pollution had a time cumulative effect year by year, the air pollution of the last year could add air pollution in the script year; (2) urban space expansion could effectively curb air pollution; (3) the urban spatial structure with high population compactness made the air pollution change in an inverted "U" shape; (4) in different economic zone levels and different urban scale levels, the direction of influence and intensity of urban spatial structure on air pollution was different. In the eastern region of China, the residential land, public facilities land and traffic land in the urban structure mainly affected the air pollution. In the central region, the residential land, industrial land, traffic land and municipal land in the urban structure had a significant impact on the air pollution, while the urban scale was the main cause of the air pollution in the western region. Based on this, we recommended the reasonable planning of land use structure, establishment of a population density regulation mechanism, and paying attention to regional differences and urban size differences. This study can help managers of different economic zones and cities of different sizes to improve urban spatial structure and control air pollution in the process of urban development.
KeywordsAir pollution • Urban spatial structure • Systematic generalized method of moments • Dynamic panel data model * Feng Wang