The tropospheric processing of acidic gases and hydrogen sulphide in volcanic gas plumes as inferred from field and model investigations
Abstract. Improving the constraints on the atmospheric fate and depletion rates of acidic compounds persistently emitted by non-erupting (quiescent) volcanoes is important for quantitatively predicting the environmental impact of volcanic gas plumes. Here, we present new experimental data coupled with modelling studies to investigate the chemical processing of acidic volcanogenic species during tropospheric dispersion. Diffusive tube samplers were deployed at Mount Etna, a very active open-conduit basaltic volcano in eastern Sicily, and Vulcano Island, a closed-conduit quiescent volcano in the Aeolian Islands (northern Sicily). Sulphur dioxide (SO2), hydrogen sulphide (H2S), hydrogen chloride (HCl) and hydrogen fluoride (HF) concentrations in the volcanic plumes (typically several minutes to a few hours old) were repeatedly determined at distances from the summit vents ranging from 0.1 to ~10 km, and under different environmental conditions. At both volcanoes, acidic gas concentrations were found to decrease exponentially with distance from the summit vents (e.g., SO2 decreases from ~10 000 μg/m3at 0.1 km from Etna's vents down to ~7 μg/m3 at ~10 km distance), reflecting the atmospheric dilution of the plume within the acid gas-free background troposphere. Conversely, SO2/HCl, SO2/HF, and SO2/H2S ratios in the plume showed no systematic changes with plume aging, and fit source compositions within analytical error. Assuming that SO2 losses by reaction are small during short-range atmospheric transport within quiescent (ash-free) volcanic plumes, our observations suggest that, for these short transport distances, atmospheric reactions for H2S and halogens are also negligible. The one-dimensional model MISTRA was used to simulate quantitatively the evolution of halogen and sulphur compounds in the plume of Mt. Etna. Model predictions support the hypothesis of minor HCl chemical processing during plume transport, at least in cloud-free conditions. Larger variations in the modelled SO2/HCl ratios were predicted under cloudy conditions, due to heterogeneous chlorine cycling in the aerosol phase. The modelled evolution of the SO2/H2S ratios is found to be substantially dependent on whether or not the interactions of H2S with halogens are included in the model. In the former case, H2S is assumed to be oxidized in the atmosphere mainly by OH, which results in minor chemical loss for H2S during plume aging and produces a fair match between modelled and measured SO2/H2S ratios. In the latter case, fast oxidation of H2S by Cl leads to H2S chemical lifetimes in the early plume of a few seconds, and thus SO2 to H2S ratios that increase sharply during plume transport. This disagreement between modelled and observed plume compositions suggests that more in-detail kinetic investigations are required for a proper evaluation of H2S chemical processing in volcanic plumes.