The Government of Canada is actively addressing air pollution via multiple approaches including the implementation of regulatory policies. The Air Quality Policy-Issue Response Section (REQA) supports these efforts by providing technical and scientific analysis of Canadian emissions and air pollution. Our methodology involves conducting two simulations: a Base Case simulation, reflecting the state of the atmosphere without the proposed regulation, and a Scenario simulation, which models the effects of targeted emission reductions. These simulations are run using the Global Environmental Multiscale-Modelling Air Quality and Chemistry model (GEM-MACH). In recent years, as emission reductions targeted by regulations have become smaller, distinguishing the real signal from model inaccuracy has become increasingly challenging. Addressing this issue is crucial because small variations in AQ modelling can lead to significant discrepancies in cost–benefit analysis, particularly in populated areas. Therefore, it is essential to determine whether the differences between scenarios are statistically significant to accurately assess the health and environmental benefits of emissions reduction initiatives. Our current research focuses on testing several statistical methods to deliver air quality modelling products with relevant significant digits and/or confidence intervals based on sound scientific principles. Preliminary results suggest that the bootstrap and confidence ratio methods may serve as an effective tool in this regard, but some questions remain. These findings will be presented at the conference, highlighting the potential of statistical methods in improving the reliability of AQ scenario modelling.