Traffic-related air pollution near roadways: discerning local impacts from background

Abstract. Adverse health outcomes related to exposure to air pollution have gained much attention in recent years, with a particular emphasis on traffic-related pollutants near roadways, where concentrations tend to be most severe. As such, many projects around the world are being initiated to routinely monitor pollution near major roads. Understanding the extent to which local on-road traffic directly affects these measurements, however, is a challenging problem, and a more thorough comprehension of it is necessary to properly assess its impact on near-road air quality. In this study, a set of commonly measured air pollutants (black carbon; carbon dioxide; carbon monoxide; fine particulate matter, PM 2.5 ; nitrogen oxides; ozone; and ultrafine particle concentrations) were monitored continuously between June 01st, 2015 and March 31st, 2017 at six stations in Canada: two near-road and two urban background stations in Toronto, Ontario, and one near-road and one urban background station in Vancouver, British Columbia. Three methods of differentiating between local and background concentrations at near-road locations were tested: 1) differences in average pollutant concentrations between near-road and urban background station pairs, 2) differences in downwind and upwind pollutant averages, and 3) interpolation of rolling minima to infer background concentrations. The latter two methods use near-road data only, and were compared with method 1, where an explicit difference was measured, to assess accuracy and robustness. It was found that method 2 produced average local concentrations that were biased high by a factor of between 1.4 and 1.7 when compared with method 1 and was not universally feasible, whereas method 3 produced concentrations that were in good agreement for all pollutants except ozone and PM 2.5 , which are generally secondary and regional in nature. The results of this comparison are intended to aid researchers in the analysis of data procured in future near-road monitoring studies. Lastly, upon determining these local pollutant concentrations as a function of time, their variability with respect to wind speed (WS) and wind direction (WD) was assessed. With the exception of ozone and PM 2.5 , local pollutant concentrations were enhanced by a factor of 2 relative to their mean in the case of stagnant winds and were shown to be proportional to WS −0.6 ). Downwind conditions enhanced local concentrations by a factor of ~ 2 relative to their mean, while upwind conditions suppressed them by a factor of ~ 4.