Spatial–Temporal Variations in Source-Specific PM2.5: Investigation of the Calgary Metropolitan Region, Alberta, Canada

Abstract The Calgary Spatial and Temporal Exposure Modeling (CSTEM) Study collected 2-week long filter samples in two different ways to study the spatial–temporal distribution of source contributions to PM 2.5 across Calgary, Alberta, Canada and its peri-urban region in 2015–2016. The spatial study involved 125 sites in each of two periods: August 2015 and January–February 2016. Alternatively, the temporal study collected samples each month at a subset of 4 sites within the Calgary city limits. The samples were analyzed for their chemical compositions and the data subjected to positive matrix factorization (PMF) for source identification and quantification. Five sources (soil/road dust, traffic, road salt, secondary inorganic aerosol/coal, and refineries) were identified from the spatial data set and four sources (soil/road dust, road salt, secondary inorganic aerosol, traffic/wood burning) from the temporal data set. Results demonstrated that PMF can resolve meaningful source types from a spatial or temporal dataset even when they are limited by longer integrated sampling times. Results from the spatial dataset modelling showed higher spatial heterogeneity in PM 2.5 contributions within the urban area, particularly for the local source types (vehicles, road dust, and refineries), demonstrating a key limitation in applying central site estimates of PM 2.5 source contributions across a large urban and non-urban area. Analysis of the temporal data from the 4 urban sites generally showed similarity in their source contributions, reflecting similar local sources, particularly highways and residential areas. Graphical Abstract