摘要：Untangling the reasons underlying changes in air pollution trajectories and regional emission differences is critical to guide efforts to attain air pollution reduction goals in China. Using Index decomposition analysis and the air pollution emission inventory, we identify the contributions of different regions and sectors to the socioeconomic factors driving China's sulfur dioxide and nitrogen oxides emissions change and difference from 2007 to 2012. We find that the reinforced end-of-pipe treatment, improved energy intensity and adjusted industrial structure dominated the emissions mitigation, while the provincial disparities in economy scale, energy pollution intensity and energy intensity led the emissions differences between regions. The emissions change was mainly driven by the utilities, capital-intensive manufacturing and service sectors of several key provinces in Central, North, Northwest and Southwest. Less developed provinces generally had a larger potential in decreasing emissions by improving energy intensity and adjusting industrial structure, especially for the utilities and capital-intensive manufacturing sectors. Our findings indicate that the successful emission reduction experiences for the power sector should be expanded to other major industrial sectors, further attention should be given to structure-focused policy actions, and different pathways of emissions reduction should be chosen for different types of provinces based on change and difference characteristics of each driver. Plain Language Abstract Identifying key factors, regions and sectors that driving China's sulfur dioxide (SO 2 ) and nitrogen oxides (NO x ) emissions change and regional differences is crucial for decision-makers to curb air pollution in a targeted way. By exploring this issue during 2007–2012, we found that end-of-pipe treatment, energy intensity and industrial structure represent the main factors offsetting SO 2 and NO x emissions. The emissions reduction was dominated by the utilities, capital-intensive manufacturing and service sectors of several key provinces in Central, North, Northwest and Southwest. We also found that disparities in economy scale, energy pollution intensity and energy intensity have largely contributed to provincial emission differences; less developed provinces generally showed poorer performances in energy intensity and industrial structure compared to developed provinces. Provinces are classified into four types based on each driver's temporal change and regional difference characteristics. We proposed that for different types of provinces, their emissions reduction potential can be further exploited by learning from other provinces instead of adjusting the mitigation measures depending only on their own historical air pollution emissions trajectories.
