SGTWR Model with Spatial-Temporal Heterogeneity and Attribute Similarity for Urban Traffic Carbon Emission Driver Analysis

Against the backdrop of global climate change and carbon neutrality goals, the transportation sector has become a focal point in urban carbon emission research. This study develops a Spatiotemporal Geographically Weighted Regression (SGTWR) model that integrates spatial, temporal, and attribute similarity dimensions to identify the main driving factors of urban transportation carbon emissions (TCE) across 287 Chinese cities from 2000 to 2019. The model incorporates climatic and geographical variables to capture the spatiotemporal heterogeneity of emission patterns. The results indicate that population density, private vehicle ownership, and heating degree days have positive effects on TCE, while terrain elevation exhibits a mitigating effect. The SGTWR model demonstrates superior explanatory power and accuracy (adjusted R2 = 0.900) compared with traditional models, revealing significant spatial patterns and temporal trends in emission drivers. Based on coefficient clustering, six types of cities are identified, highlighting regional disparities in emission mechanisms. These findings provide methodological and theoretical support for formulating differentiated low-carbon transportation policies tailored to regional geographic and socio-economic contexts.