Emissions from mobile sources: improved understanding of the drivers of emissions and their spatial patterns

Emissions of greenhouse gases from the combustion of fossil fuels, in particular carbon dioxide (CO2), are a major contributor to global climate change. In the United States 28% of carbon dioxide emissions from fossil fuel combustion are produced by road vehicles. This dissertation reports the results of three studies that improve on our knowledge of the spatial and temporal distribution of vehicle CO2 emissions in the U.S. over the last 35 years. Using bottom-up data assimilation techniques we produce several new high-resolution inventories of vehicle emissions, and use these new data products to analyze the relationships between emissions, population, employment, traffic congestion, and climate change at multiple spatial and temporal scales across the U.S. We find that population density has a strong, non-linear effect on vehicle emissions, with increasing emissions in low density areas and decreasing emissions in high density areas. We identify large biases in estimates of vehicle CO2 emissions by the most commonly used national and global inventories, and highlight the susceptibility of spatially-downscaled inventories to local biases in urban areas. We also quantify emissions of several air pollutants regulated by the U.S. Environment Protection Agency, including carbon monoxide, nitrogen oxides and particulate matter, at hourly and roadway scales for the metropolitan area surrounding Boston, MA. Emissions of these pollutants show high emissions gradients across identifiable spatial hotspots, considerable diurnal and seasonal variations, and a high sensitivity to the presence or absence of heavy-duty truck traffic. We also find that the impact of traffic congestion on air pollution emissions across the region is minimal as a share of the total emissions. We show that policies that combine a reduction in the number of vehicles on the road with a focus on improving traffic speeds have greater success in reducing emissions of air pollutants and greenhouse gases than policies that focus solely on improving traffic speeds. Finally, we estimate that regional emissions of carbon monoxide will increase by 3% in 2050, but with numerous localized increases of 25-50%, due to an expected rise in mean regional temperatures due to global climate change