Chapter 18 - Receptor Modeling of Epiphytic Lichens to Elucidate the Sources and Spatial Distribution of Inorganic Air Pollution in the Athabasca Oil Sands Region

Abstract

The contribution of inorganic air pollutant emissions to atmospheric deposition in the Athabasca Oil Sands Region (AOSR) of Alberta, Canada, was investigated in the surrounding boreal forests, using a common epiphytic lichen bioindicator species (Hypogymnia physodes) and applying multiple receptor models. Source materials from anthropogenic and natural emitters of air pollution in the AOSR were obtained and chemically characterized to aid in the assessment. The lichens selected for analysis were collected in 2008 using a stratified, nested grid approach radiating away from the central area of oil sands production at 121 sampling sites extending as far as 150 km. Source and lichen samples were extracted and analyzed for 43 elements using dynamic reaction cell quadrupole inductively coupled plasma mass spectroscopy (DRC-ICPMS). Source apportionment of the lichen tissue analytical results was conducted using principal component analysis (PCA), chemical mass balance (CMB), positive matrix factorization (PMF), and Unmix models. Initial Varimax-rotated PCA screening analysis indicated that there were five principal components that could explain 89% of the variance contained in the lichen data set, with the majority of the variance lumped into a fugitive dust factor. This fugitive dust source could be separated into tailing sand, haul road, and overburden components using CMB on lichen samples collected near the mining and oil processing facilities. However, the CMB model performance was limited by the similarity of sources and the lack of total nitrogen measurements in the emission source profiles. The PMF and Unmix models were found to perform best with this unique AOSR lichen data set, providing very similar results at near-source as well as remote lichen collection sites. The PMF results showed that sources significantly contributing to concentrations of elements in the lichen tissue include combustion processes (∼ 23%), tailing sand (∼ 19%), haul roads and limestone (∼ 15%), oil sand and processed materials (∼ 15%), and a general anthropogenic urban source (∼ 15%). The spatial patterns of CMB, PMF, and Unmix receptor models estimated that source impacts on the H. physodes tissue elemental concentrations from the oil sand processing and fugitive dust sources had a significant association with the distance from the primary oil sands surface mining operations and related production facilities. The spatial extent of the fugitive dust impact was limited to a radius of approximately 20 km around the major mining and oil production facilities, which is indicative of ground-level coarse particulate fugitive emissions from these sources. The impact of the general urban source was found to be enhanced in the southern portion of the sampling domain in the vicinity of the Fort McMurray urban area. The receptor model results showed lower Mn concentrations in lichen tissues near oil sands production operations suggesting a biogeochemical response. Overall, the largest impact on elemental concentrations of H. physodes tissue in the AOSR was related to fugitive dust, suggesting that implementation of a fugitive dust abatement strategy could minimize the near-field atmospheric deposition of any future mining-related production activities.