["Transactions in GIS, Volume 30, Issue 5, August 2026. ", "\nABSTRACT\nRoad traffic is a major source of global carbon dioxide (CO2) emissions. Existing studies often overlook well‐to‐wheel (WTW) emissions from fuel and electric vehicles and provide limited insight into how emissions interact with multiple factors. To address these gaps, this study uses license plate recognition (LPR) data to examine the spatiotemporal heterogeneity of WTW CO2 emissions. A unilateral geographically and temporally weighted regression (UGTWR) model is developed to analyze influencing factors. Results show distinct emission patterns between vehicle types. Fuel vehicle emissions averaged 1991.35 t on weekdays and 1797.80 t on weekends, while electric vehicle emissions remained stable at 215.13 t and 215.21 t. Fuel vehicles showed clear peak patterns, whereas electric vehicles did not. High emissions for both types were concentrated in urban centers and major corridors. The UGTWR model demonstrated superior performance compared with alternative models, with R2 more than 0.90, indicating its effectiveness in capturing spatiotemporal heterogeneity.\n"]