“…Such uncertainties need to be carefully considered in air quality evaluations and policy decision‐making (Bei et al, ; Gilliam et al, ; Ludwig & Shelar, ; Pielke, ). While meteorological ensemble modeling has received much attention in both research and operational forecasting during the past three decades (e.g., Palmer et al, ; Stensrud et al, ; Toth & Kalnay, ), the use of ensembles of Eulerian dynamically generated meteorology‐air quality models to characterize uncertainties of air quality forecast has been lagging behind (e.g., Bei et al, ; Delle Monache et al, ; Delle Monache & Stull, ; Djalalova et al, ; Galmarini et al, ; Marecal et al, ; McKeen et al, ; Monteiro et al, ; Vautard et al, ; Zhang et al, ). To achieve skillful probabilistic air quality forecasts, different approaches were used to construct the ensemble, including using multiple air quality models (Delle Monache & Stull, ; Djalalova et al, ; Galmarini et al, ; Marecal et al, ; McKeen et al, ; Monteiro et al, ; Vautard et al, ), perturbing emissions (Delle Monache et al, ), and perturbing meteorological conditions (Bei et al, ; Zhang et al, ).…”