https://doi.org/10.1021/acs.est.5c14619

Traffic emissions are unevenly distributed in dense cities, contributing to traffic-related environmental inequality. Addressing this issue requires identifying which vehicle fleets disproportionately drive these disparities and how their impacts vary throughout the day.

In this study, we integrate high-resolution traffic counts, traffic images, detector data, machine learning, and computer vision to estimate hourly road-segment-level traffic emissions for eight vehicle classes in Hong Kong. Our results show that light-duty goods vehicles, heavy-duty goods vehicles, and double-deck buses are the major contributors to both total emissions and emission-related inequality. Their contributions also vary by time of day: goods vehicles dominate inequality during daytime working hours, while double-deck buses become more important in the evening.