Size-resolved aqueous-phase atmospheric chemistry in a three-dimensional chemical transport model

[1] Three-dimensional chemical transport models typically include a bulk description of aqueous-phase atmospheric chemistry. Previously, this bulk description has been shown to be often inadequate in predicting sulfate production. The pH of the bulk mixture does not adequately describe the pH of the typically heterogeneous droplet population found in clouds and fogs. This often leads to an inability of bulk models to predict sulfate production when pH-dependent production pathways are important. A more accurate size-resolved aqueous-phase chemistry model, however, has long been considered infeasible for incorporation in a three-dimensional chemical transport model because of high computational costs. Here we investigate the feasibility of adding a computationally efficient size-resolved aqueous-phase chemistry module (Variable Size Resolution Model (VSRM)) to a three-dimensional model (the latest version of the Comprehensive Air Quality Model with extensions (PMCAMx)). The VSRM treats mass transfer between the gas phase and the different droplet populations and executes bulk or two-section size- resolved chemistry calculations in each step on the basis of the chemical environment of each computational cell. A fall air pollution episode in California's South Coast Air Basin is simulated, and model predictions are compared to observations. In an environment where clouds or fogs are present, the model without aqueous-phase chemistry severely underpredicts secondary sulfate formation. In cases where there is a high potential for sulfate production and widely varying composition across the droplet spectrum ( over the ocean and near the coast), there is a significant increase in sulfate production over bulk predictions with the activation of a size- resolved aqueous-phase chemistry module. Unfortunately, measurements were only available at inland sites, where the difference between bulk and size- resolved sulfate predictions was small. The effects of other uncertainties on sulfate production are also examined. For limited computational costs (5% overhead), the VSRM can be included in a three-dimensional chemical transport model.