Spatiotemporal analysis of CO₂ emissions from heavy-duty trucks in belt and road port transportation using GPS trajectory data

• Reconstruct truck trajectories via HMM-based map matching. • Estimate CO₂ emissions considering load uncertainty. • Identify linear hotspot belts in port access roads. • Reveal low-speed driving as key emission factor. Amid global efforts toward carbon neutrality, the emissions from heavy-duty trucks (HDTs) within port transportation systems have become an increasing focus of research and policy discourse. This study constructs a comprehensive framework based on GPS trajectory data from overloaded container trucks, encompassing trajectory matching, carbon emission estimation, and spatial clustering analysis to systematically identify the spatiotemporal evolution characteristics of carbon dioxide (CO₂) emissions within the port transportation network. First, a Hidden Markov Model (HMM) is employed for high-precision trajectory matching, accurately reconstructing vehicle travel paths. Subsequently, a carbon emission estimation model that accounts for vehicle mass uncertainty is developed to quantify emission intensity across different time periods and road segments. Finally, Kernel Density Estimation (KDE) and the Getis-Ord G i * clustering method are utilized to identify emission hotspots and their dynamic changes. The results indicate that carbon emissions in the port collector-distributor roads exhibit a temporal pattern that precedes peak commute times on urban roads while spatially forming a “linear high-density belt” along major corridors, contrasting with the “point aggregation” characteristic of urban roads. Additionally, it is found that low-speed operating conditions, rather than high traffic volume, are key contributors to increased carbon emission intensity. This research provides theoretical support and quantitative evidence for carbon reduction in “Belt and Road” port transportation, offering decision-making references for constructing a low-carbon port logistics system.