Microphysical Modeling of Cloud Chemistry

A refined version of the Ayers and Larson (1990) cloud model, including explicit cloud microphysics and externally mixed size-dependent cloud drop chemistry, was employed to assess the effects of heterogeneous cloud chemistry on the processing of aerosols by clouds in both marine and continental air. The model was modified to include the labile acid, HNO_3, as an initial input parameter. Two different chemical modes, an acid accumulation mode and an alkaline coarse mode, were considered in the particle spectra. A warm stratiform cloud or a cumulus cloud was utilized in the simulation. Appropriate values for the various gas-phase species concentrations to use in conjunction with the particle spectra were derived from numerous literature sources. The following conclusions could be drawn from the modeling results. (1) Heterogeneity in the chemistry across the cloud droplet size distribution for both marine and continental air can have significant impact on the amount of sulfate produced in-cloud. (2) In-cloud sulfate production is most sensitive to variations in the updraft velocity, the initial SO_2 and NH _3 concentrations at cloud base, and the initial particle size distribution at cloud base. (3) The presence of HNO_3 neutralizes the NH_3 and the alkalinity of the coarse mode and increases the acidity of the droplets, thereby leaving less SO_2 to dissolve into the droplets and a decrease in the total sulfate mass produced. (4) The light-scattering efficiency of sulfate mass produced in clouds also depends on the initial particle size distribution in both marine and continental scenarios. (5) The presence of alkaline particles (especially in marine air) has a significant impact on the magnitude of sulfate produced in-cloud, its size distribution, and consequently on its light-scattering efficiency. (6) The relative differences in sulfate production predicted by the explicit model and that predicted by the simpler models increase with decreasing initial SO_2 concentration and this explicit-bulk discrepancy is time dependent. (7) The parameterization approach which is based on the results of linear regressions of explicit model predictions onto sets of bulk-model prognostic variables along with the bulk model predictions can significantly diminish the discrepancy between bulk and explicit cloud model predictions.