In a context of increasing concern for population exposure to ambient air pollution, it is crucial to develop tools to assist urban policies tackling air quality. In this thesis, techniques to investigate air quality at local and regional scales were developed and applied to analyze the impact of transportation policies on population exposure and health.
The first part of this thesis consists in a large data collection campaign involving short-term stationary and mobile measurements of air pollutant concentrations in Toronto. Land-use regression (LUR) models based on the two data collection protocols are developed and compared with data from a panel study. Recommendations are provided for the design of short-term monitoring campaigns.
In the second part of this thesis, we set-up a chemical transport model (CTM) and a plume-in-grid (PinG) model for the Greater Toronto and Hamilton Area (GTHA) to simulate the levels of various air contaminants at a scale of 1 km2. These models are based on a detailed traffic emission inventory and enable a refined analysis of population exposure to traffic-related air pollution. The CTM is combined with a health impact assessment tool to investigate the benefits of greening freight movements on urban air quality and health.
In the third module, we use our LUR and air quality models to analyze the impact of transportation policies on population exposure and health. The fine spatial resolution of the LUR exposure surfaces is optimal to assess a major urban planning strategy of the City of Toronto: the installation of new bike facilities. We use our CTM to investigate the health and climate benefits of renewing the fleets of private household vehicles, transit buses and commercial vehicles.
While we recommend the use of detailed air pollutant maps such as those obtained from LUR models to assist local policy making, spatially refined CTMs supplemented by comprehensive emission inventories are outstanding tools to assess and compare the benefits of transportation policies at a regional scale.