Traffic Density and Traffic-Related Air Pollutants in NC
On this page:
Objective: Nitrogen dioxide (NO2), fine particulate matter (PM2.5), and carbon monoxide (CO) concentrations are measures of traffic-related air pollutants (TRAPs). TRAPs are associated with fossil fuels use in vehicles and vehicular exhaust, though it is unclear if they approximate traffic exposures. We evaluate how TRAPs are correlated with traffic density in North Carolina (NC).
Materials and Methods: Using average annual daily traffic counts and length of primary and secondary road segment data from the NC Department of Transportation, we calculated traffic density (vehicle miles traveled per square mile) for all NC census block groups for 2011 and 2015. Annual average estimates of NO2, PM2.5, and CO were obtained from land-use regression models for NC census block groups for 2011 and 2015. We calculated Spearman correlations between traffic density and TRAPs.
Results: Among TRAPs, NO2 was highly correlated with CO (ρ=0.77) and PM2.5 (0.80) in 2015. We observed moderate to high correlations between traffic density and the three TRAPs in 2015 (CO: 0.66, PM2.5: 0.58, and NO2: 0.76). We observed similar results for 2011. When stratified by roadway type, traffic density on primary roads was highly correlated with NO2 (0.76), and moderately correlated with PM2.5 (0.59) and CO (0.65), with comparable correlations observed for traffic density on secondary roads and TRAPs. When stratifying by low (<median) or high (≥median) traffic density, we observed similar patterns of correlations between traffic density and TRAPs in each stratum.
Conclusions: Estimates of census block group traffic density may be a moderate proxy of long-term exposure to NO2 and PM2.5 concentrations in NC, independent of roadway type or traffic density level. Additional analysis to explore spatial differences related to rural/urban and gentrification status will be useful. Understanding how measures of traffic are correlated may contribute to better adjustment of co-pollutant confounding.