Characterizing spatiotemporal distributions of black carbon and PM2.5 at a toll station: Observations on manual and electronic toll collection lanes

Abstract

Toll stations are traffic-related pollution hotspots, but the contribution of various factors to pollutant distributions has not been clearly recognized. This study conducted a field investigation of the spatiotemporal variations of pollutants on different lanes at a toll station in Shanghai, China. Quantitative analysis showed that the concentrations of black carbon (BC) and fine particles (PM2.5) on the manual toll collection (MTC) lane were higher than those on the electronic toll collection (ETC) lane, and higher concentrations were identified on weekdays. Furthermore, the study introduced interaction terms between lane type and traffic conditions while constructing generalized additive models (GAMs) to reveal the internal relationship between concentrations and explanatory variables. The results of GAM analysis showed that the coupling effect of traffic volume and lane type was the most important force driving variations in BC, while background concentration was the dominant factor of PM2.5 variations. The Extreme Gradient Boosting (XGBoost) models were developed to separately evaluate the influence of lane type and traffic volume on BC and PM2.5, indicating that lane type contributed 29.4% and 36.5% of BC and PM2.5 variations respectively. Additionally, simulation demonstrated that in contrast to MTC, ETC can reduce BC pollution by 25.2% under the same traffic conditions. Finally, the respiratory deposition dose was used to compare the negative consequences of exposure between the ETC and MTC lane. This study revealed the general variation trends of pollutants and corresponding influencing factors at a toll station, and could provide implications for mitigating traffic-related pollution.