T
he study contributed to the existing researches by depicting the complex network and efficiency-related drivers of PM
2.5
concentration, and by proposing two production-theoretical decomposition analysis (PDA) -based spatiotemporal decoupling models that incorporate efficiency-related factors in spatiotemporal decoupling analysis for the first time.
We comprehensively used social network analysis, extended PDA and the proposed spatiotemporal decoupling models, to investigate the network and driver PM
2.5
concentrations as well as the decoupling nexus with fiscal environmental expenditure (
FEE
) in a case of Chinese cities classified by population size and economic structure over 2007-2019. We found that first, the hierarchical network structure of PM
2.5
concentration has been broken to a certain extent. Second, GDP per capita and potential intensity of PM
2.5
to
FEE
(PPMI) were the largest positive and negative drivers to reduce PM
2.5
concentrations. Third, the temporal and spatial decoupling relationships between
FEE
and PM
2.5
were generally dominated by strong decoupling state and strong negative spatial decoupling states, respectively, where PPMI and
FEE
tendency were the positive drivers and significant heterogeneity of cities classified by economic structure and population size were found. The study highlights the importance of controlling air pollution by
FEE
according to local conditions.
