Key Atmospheric Processes in The Canadian Oil Sands Identified through Model Evaluation

We describe the current status of the ongoing model improvement and evaluation of the Oil Sands version of the Global Environmental Multiscale – Modelling Air-quality and CHemistry (GEM-MACH-OS) model.  GEM-MACH-OS was designed to provide 2.5km horizontal grid cell size model predictions for the chemical processing of gases and particulate matter emitted from industrial activities in the Canadian Oil Sands and other sources in the Canadian provinces of Alberta, Saskatchewan and neighboring regions.  Starting in 2022, a successive series of model updates and evaluations were carried out for the model simulation year October 1, 2017 through September 30, 2018.  We report here on several of these simulations how the comprehensive dataset from different monitoring networks was used to improve GEM-MACH-OS predictions, and identify key processes for Oil Sands chemistry.  The monitoring networks included the Wood Buffalo Environmental Association (WBEA, which provided hourly air concentration data for NO₂, SO₂, PM2.5, O₃, NO and CO, daily intermittent total and speciated PM2.5 and PM10, and passive monthly to bimonthly SO₂, NO₂, HNO₃, NH₃ and O₃), the National Trends Network (NTN, providing weekly precipitation totals and ions in precipitation for SO₄^2-, NO₃^-, NH₄⁺, Ca²⁺, Mg²⁺, K⁺, Na⁺ and Cl^-), the National Air Pollution Surveillance program (NAPS, providing continuous hourly samples of NO₂, SO₂, PM2.5, O₃, NO and CO, as well as daily intermittent samples of HNO₃, SO₂ speciated PM2.5, speciated total PM at CAPMoN stations), and the Canadian Air and Precipitation Monitoring Network (CAPMoN, providing daily intermittent samples of precipitation and ions in precipitation for the same species as NTN).

Examples of evaluation over 5 consecutive model versions will be shown, demonstrating both the improvement in model performance over time, and identifying chemical species for which further improvement is desired.  The evaluation also identified key processes governing chemical transformation in the region.  These included: (1) O₃:  relatively little photochemical production from local emissions takes place, but down-mixing from the upper atmosphere creates a substantial seasonal signal; (2) SO₂:  mostly emitted from large stacks, with the plume heights depending on a parameterization including latent heat release from combustion water (Fathi et al., 2024), and co-deposition potentially has a significant influence on SO₂ deposition; (3) NO₂:  a key reaction governing concentrations in the region is the reaction of NO₂ on particle surfaces to form HONO and HNO₃; (4) Forest fires in the region emit much lower levels of SO₂ and NO_x than standard inventory emission factors would suggest, and have a different particle speciation; (5) Particulate matter from Oil Sands fugitive dust sources is influenced both by vehicle-induced turbulence and meteorological modulation (with coarse mode emissions dropping off as temperatures drop below a fixed temperature when the ground is frozen, during rainfall and snowfall events, and as the surface soil water increases).  Planned next steps in model improvement will also be discussed.