Determinants of ultrafine particle exposures in transportation environments: findings of an 8-month survey conducted in Montréal, Canada

Abstract

An 8-month sampling campaign was conducted in Montréal, Canada to explore determinants of ultrafine particle (UFP) exposures in transportation environments and to develop models to predict such exposures. Between April and November 2006, UFP (0.02-1 mum) count exposure data were collected for one researcher during 80 morning and evening commutes including a 0.5-km walk, a 3-km bus ride, and a 26-km automobile ride in each direction. Ambient temperature, relative humidity, precipitation, and wind speed/direction data were collected for each transit period and the positions of bus and automobile windows were recorded. Mixing heights were also estimated. Morning UFP exposures were significantly greater than those in the evening, with the highest levels observed in the automobile and the lowest while walking. Wind speed and mixing height were highly correlated, and as a result only wind speed was considered in multivariable models owing to the accessibility of quantitative hourly monitoring data. In these models, each 10 degrees C increase in morning temperature was associated with decreases of 14,560/cm(3) (95% CI=11,111 to 18,020), 8160/cm(3) (95% CI=5060 to 11,260), and 11,310/cm(3) (95% CI=6820 to 15,810) for UFP exposures in walk, bus, and automobile environments, respectively. Likewise, each 10-km/h increase in morning wind speed corresponded to decreases of 8252/cm(3) (95% CI=5130 to 11,360), 6210/cm(3) (95% CI=3420 to 9000), and 6350/cm(3) (95% CI=2440 to 10,260) for UFP exposures in walk, bus, and automobile environments, respectively. Similar trends were observed in the evening hours. In an evaluation of model performance, moderate correlations were observed between measured and predicted UFP exposures on new bus (r=0.65) and automobile (r=0.77) routes. Further research is required to incorporate variables such as traffic density and vehicle ventilation settings into the models presented.