Understanding the spatial differentiation and spatiotemporal mechanisms of carbon emissions from urban transport

As the process of urbanisation accelerates, carbon emissions are reshaping the urban spatial structure. Current theoretical and empirical studies mostly discuss the spatiotemporal patterns of carbon emissions at different spatial scales, and distinguish the impacts of carbon emissions on urban growth from the perspective of spatial heterogeneity. Essentially, these studies focus on the spatial effects of carbon emissions caused by urbanisation. However, in the process of urban development, carbon emissions are also constantly reshaping the spatial structure and functional distribution of cities. Thus, this study aims to gain greater insight into the spatial differentiation and spatiotemporal mechanisms of urban carbon emissions. Using Beijing as a case study, we employ the community detection method, P-Z score analysis and a spatially interpretable machine learning approach to address the aforementioned questions, respectively. Our results show that Beijing demonstrates a total modular community structure, comprising seven identified communities. The patterns of carbon emissions observed for the top 10 metro stations in terms of importance within and between communities differ considerably. In addition, human activity intensity, socio-economic dynamics, and land-use complexity significantly influence spatial differentiation driven by urban carbon emissions, with distinct patterns emerging across different urban communities. The study contributes to existing knowledge in the following ways: first, it not only evaluates carbon emissions as a product of urbanisation, but also focuses on how carbon emissions can counteract urban spatial planning. Second, it departs from the traditional administrative functional zoning spatial units used in the analysis of carbon emissions and reconstructs urban spatial divisions based on carbon emissions. Third, it offers a sustainable indicator system with which to identify different stations in different spaces and analyse the spatial mechanisms of urban carbon emissions.