1) Research Objective: Cities are the main sources of the carbon emissions. To better achieve the dual carbon goals, we explore the impact path and mechanism of urbanization on carbon emissions from the perspective of city-level assessment. 2) Research methods: Based on the panel data of 30 provincial capital cities in China from 2006 -2020, this paper first classifies the cities based on their individual urbanization levels and geographical location. And then the Environmental Kuznets Curve (EKC) and Panel Vector Autoregression (PVAR) models are applied to quantitatively analyze the effects of four-dimension urbanization (population urbanization, economic urbanization, social urbanization and land urbanization) on per capita carbon emissions. 3)Research results include: There is no EKC pattern in West China regions, but varied forms exist in East China regions and middle of China; distinctive carbon emissions phase has been judged for the corresponding cities; there exists the bidirectional long-run Granger causality relationship between five urbanization factors and carbon emissions in East China regions, but not in the other two regions; the cities’ future focus-on carbon emissions reduction policies could be made according to the research results. 4) Conclusion: A fresh research framework for quantitively assessing the city-level impacts of urbanization on carbon emissions has been set up and applied to identify the cities’ carbon emissions phase (carbon peak time), reveal the causality relationship among the urbanization factors and carbon emissions in different regions, and at last help to clarify the focus-on carbon reduction measures.

urbanization; carbon emissions; EKC model; PVAR model