Investigating the Potential of Land Use Modifications to Mitigate the Respiratory Health Impacts of NO2: A Case Study in the Portland-Vancouver Metropolitan Area

• Main Author: Rao, Meenakshi
• Format: Others
• Published: PDXScholar 2016
• Subjects: Atmospheric nitrogen dioxide > Oregon > Portland Metropolitan Area > Mathematical models; Atmospheric nitrogen dioxide > Washington (State) > Vancouver Metropolitan Area > Mathematical models; Urban forestry > Environmental aspects; Trees in cities > Environmental aspects; Atmospheric Sciences; Environmental Sciences
• Online Access: http://pdxscholar.library.pdx.edu/open_access_etds/2940; http://pdxscholar.library.pdx.edu/cgi/viewcontent.cgi?article=3949&context=open_access_etds

The health impacts of urban air pollution are a growing concern in our rapidly urbanizing world. Urban air pollutants show high intra-urban spatial variability linked to urban land use and land cover (LULC). This correlation of air pollutants with LULC is widely recognized; LULC data is an integral input into a wide range of models, especially land use regression models developed by epidemiologists to study the impact of air pollution on human health. Given the demonstrated links between LULC and urban air pollution, and between urban air pollution and health, an interesting question arises: what is the potential of LULC modifications to mitigate the health impacts of urban air pollution? In this dissertation we assess the potential of LULC modifications to mitigate the health impacts of NO2, a respiratory irritant and strong marker for combustion-related air pollution, in the Portland-Vancouver metropolitan area in northwestern USA. We begin by measuring summer and winter NO2 in the area using a spatially dense network of passive NO2 samplers. We next develop an annual average model for NO2 based on the observational data, using random forest -- for the first time in the realm of urban air pollution -- to disentangle the effects of highly correlated LULC variables on ambient NO2 concentrations. We apply this random forest (LURF) model to a 200m spatial grid covering the study area, and use this 200m LURF model to quantify the effect of different urban land use categories on ambient concentrations of NO2. Using the changes in ambient NO2 concentrations resulting from land use modifications as input to BenMAP (a health benefits assessment tool form the US EPA), we assess the NO2-related health impact associated with each land use category and its modifications. We demonstrate how the LURF model can be used to assess the respiratory health benefits of competing land use modifications, including city-wide and local-scale mitigation strategies based on modifying tree canopy and vehicle miles traveled (VMT). Planting trees is a common land cover modification strategy undertaken by cities to reduce air pollution. Statistical models such as LUR and LURF demonstrate a correlation between tree cover and reduced air pollution, but they cannot demonstrate causation. Hence, we run the atmospheric chemistry and transport model CMAQ to examine to what extent the dry deposition mechanism can explain the reduction of NO2 which statistical models associate with tree canopy. Results from our research indicate that even though the Portland-Vancouver area is in compliance with the US EPA NO2 standards, ambient concentrations of NO2 still create an annual health burden of at least $40 million USD. Our model suggests that NO2 associated with high intensity development and VMT may be creating an annual health burden of $7 million and $3.3 million USD respectively. Existing tree canopy, on the other hand, is associated with an annual health benefit of $1.4 million USD. LULC modifications can mitigate some fraction of this health burden. A 2% increase in tree canopy across the study area may reduce incidence rates of asthma exacerbation by as much as 7%. We also find that increasing tree canopy is a more effective strategy than reducing VMT in terms of mitigating the health burden of NO2. CMAQ indicates that the amount of NO2 removed by dry deposition is an order of magnitude smaller than that predicted by our statistical model. About one-third of the difference can be explained by the lower NO2 values predicted by CMAQ, and one-third may be attributable to parameterization of stomatal uptake.