Modeling Impact of Traffic Conditions on Variability of Midblock Roadside Fine Particulate Matter

Abstract

The objective of this study was to examine the concentration variation of midblock roadside particulate matter less than 2.5 Μm (PM2.5) as a function of very high resolution meteorological and traffic data. Morning peak period measurements were taken at a midblock roadside location on an urban arterial commuter roadway. For the impact of dynamic traffic conditions to be captured, data were analyzed at 10-s intervals, a substantially higher resolution than that used in typical roadside air quality study designs. Particular attention was paid to changes in traffic conditions, including fleet mix, queuing, and vehicle platooning over the course of the study period, and the effect of these changes on PM2.5. Significant correlations were observed between vehicle platoons and increases in PM2.5 concentrations. Traffic state analysis was employed to determine median PM2.5 levels before and after the onset of congestion. A multivariate regression model was estimated to determine significant PM2.5 predictors while controlling for autocorrelation. Significance was found not only in the simultaneous traffic variables but also in lagged traffic variables; in addition, the effects of vehicle types and wind direction were quantified. Modeling results indicated that traffic state (e.g., congestion) and vehicle type had a significant impact on roadside PM2.5 concentrations. This study serves as a demonstration of the abilities of very-high-resolution data to identify the effects of relatively minute changes in traffic conditions on air pollutant concentrations.