High-resolution mapping of regional traffic emissions using land-use machine learning models

Abstract

On-road vehicle emissions are a major contributor to significant atmospheric pollution in populous metropolitan areas. We developed an hourly link-level emissions inventory of vehicular pollutants using two land-use machine learning methods based on road traffic monitoring datasets in the Beijing-Tianjin-Hebei (BTH) region. The results indicate that a land-use random forest (LURF) model is more capable of predicting traffic profiles than other machine learning models on most occasions in this study. The inventories under three different traffic scenarios depict a significant temporal and spatial variability in vehicle emissions. NOx, fine particulate matter (PM2.5), and black carbon (BC) emissions from heavy-duty trucks (HDTs) generally have a higher emission intensity on the highways connecting to regional ports. The model found a general reduction in light-duty passenger vehicles when traffic restrictions were implemented but a much more spatially heterogeneous impact on HDTs, with some road links experiencing up to 40ĝ% increases in the HDT traffic volume. This study demonstrates the power of machine learning approaches to generate data-driven and high-resolution emission inventories, thereby providing a platform to realize the near-real-time process of establishing high-resolution vehicle emission inventories for policy makers to engage in sophisticated traffic management.