It’s widely known that traffic related air pollution is associated with long-term health effects like cancers, cardiovascular or respiratory illnesses. But there’s more that could be learned about the short-term exposure to traffic-related air pollution.  In the first SOCAAR Seminar for 2015-16, Marianne Hatzopoulou, Associate Professor in the Department of Civil Engineering at the University of Toronto discussed the application of transportation and emission models to investigate air pollution in the city of Montreal. The main research questions presented were: i) How can we improve measures of exposure? ii) What is the role of integrated transportation-dispersion models? Can transportation models replace atmospheric dispersion modelling?

In her talk, Hatzopoulou made comparisons amongst the distribution of spatial estimates of NO2 derived from a transportation/emission model, a land-use regression model and a dispersion model. Two case-control studies of breast and prostate cancer were presented to compare the estimates of health risk determined by the different exposure estimates.

First, land-use regression models created from the usage of Aeroqual micro-sensors to capture the spatial variability of NO2 in Montreal were presented.  Compared to the RSQA fixed stations, the micro-sensors under predict for low NO2 levels, but over predict for high NO2 levels.

Second, mobile monitoring was compared to fixed stations. Fixed stations are more time intensive compared to mobile monitoring. Mobile monitoring is less time intensive than fixed stations. However the larger coverage is at the sacrifice of obtaining temporal averages. Initial studies investigated ultra-fine particle concentrations, and found the traffic effect on air pollution is not linear.

Third, studies investigated air pollution exposures using the output of dispersion models. Exposure was evaluated by considering where individuals work, live, shop, and commute. 24-hour exposures from all the trips made by an individual on a typical day were evaluated to try to understand how one’s exposure varies. The hourly average NO2 from the dispersion model was correlated with the land-use regression model. The concentration of home exposure is higher than the outside exposure for most people since people spend more time at home. It was observed that individuals who live in an area with good air quality (i.e. better than other activity locations that you visit) have 90% probability of increasing their daily exposure by 20% compared to the air quality at home. On the other hand, individuals that live in area with bad air quality, have 87% of increasing exposure by 8% because of being on the road. Overall, looking at daily exposure more important than home exposure alone.

Finally, the question of how measures of exposure could be improved without running large panel studies and requiring everyone to wear a monitoring device was raised. Transport or dispersion models are not commonly used for exposure analysis. They are too coarse to reflect air pollution in urban microenvironments. Land-use regression models with spatially extensive data can capture “microenvironment” characteristics and update air pollution maps.