Accelerated urbanization is reshaping global carbon emissions worldwide, while city-level evidence still relies heavily on administrative boundaries and single proxies of urban development, limiting understanding of how emissions vary across urban contexts. We delineate census-consistent physical urban footprints for 292 Chinese cities and align them with 1-km ODIAC CO 2 data to construct a harmonized panel for 2005-2022. These physical urban areas occupy 2.7% of China’s land area but account for nearly one-third of national CO 2 emissions, with emissions growth consistently outpacing population growth. Using a STIRPAT-based empirical framework, we find that population growth and rising affluence are generally associated with higher emissions, whereas green innovation and greater compactness are associated with lower emissions. Regulatory impacts emerge with lags: contemporaneous effects weaken once lag terms are introduced, while the cumulative lagged effect is strongly negative. More importantly, these relationships are heterogeneous across the urban hierarchy and density regimes. Income-related turning points emerge earlier in denser cities, and the inverted-U pattern is most evident in large cities but becomes weaker toward the upper end of the urban hierarchy. These results recast urban emissions as trajectory-dependent and motivate city-type mitigation strategies paired with footprint-based accounting in gridded-emissions settings. • Physical urban footprints reveal strong CO 2 concentration in China’s top-tier cities. • A STIRPAT-based specification captures multidimensional urban carbon drivers. • Income turning points vary across city-size classes and density regimes. • Density lowers emissions overall, but its mitigating effect weakens at high levels. • Regulation effects emerge with lags, supporting differentiated city-type policies.
Yang et al. (Fri,) studied this question.