Air quality models (AQM) paired with epidemiological data are often used to estimate health related burdens from pollution exposure. The accuracy of these predictions depends, among other factors, on the horizontal resolution of the model. This thesis aims to quantify the impact of horizontal grid resolution by simulating health impacts due to O 3, NO 2 , and PM 2.5 exposure using various model and input resolutions. Adjoint sensitivity analysis was used to predict health impacts, producing

more »

... lts with respect to emissions sources. The results indicate that coarse modeled health impacts underestimate maxima in urban areas and overestimate in rural areas with proximity to urban cores. Additionally, coarse modelling does not sufficiently display spatial variance. Concentration exposure relationships applied to AQM predictions were also used to estimate health impacts. The resolution of individual inputs (meteorology, emissions, and population) were altered to examine which processes were most responsible for the differences due to resolution. The resolution of population had the largest impact on health impact results for all species. The resolution of meteorology and emissions impacted the species to different extents; O 3 was more impacted by meteorology, while NO 2 and PM 2.5 were more influenced by emissions. The influence of specific emission sources can be more adequately determined at fine resolution, benefiting air quality control. Aggregated estimates across the domain did ii not alter significantly between resolutions. It is suggested that justification based on the purpose of the analysis be considered prior to choosing a resolution. iii