Publication Type:Journal Article
Source:Journal of Geophysical Research: Atmospheres, Volume 110, Issue D21, p.D21312 (2005)
Keywords:Atmosphere, DMS chemistry, sulfur cycle, Troposphere: composition and chemistry, Troposphere: constituent transport and chemistry
The gas-phase products of dimethylsulfide (DMS) oxidation are simulated in the global 3D Model of Atmospheric Transport and Chemistry (MATCH). The focus is on the sensitivities to the assumed mechanisms of DMS oxidation chemistry in large-scale models, and hence volcanic and anthropogenic sources of SO2 are ignored. Four representations of DMS chemistry are considered, including two comprehensive mechanisms (about 50 reactions) and two parameterized schemes (four and five reactions). The gas-phase yields of DMS, SO2, methanesulfonic acid (MSA) and H2SO4 are compared between these four cases as a measure of the sensitivity to uncertain DMS chemistry. Among the four cases, DMS is largely invariant, while SO2 using parameterized chemistry is three times higher than its levels using comprehensive chemistry in the tropical upper troposphere. For MSA and H2SO4, there are order-of-magnitude inter-mechanistic differences at high and low altitudes in the extratropics, respectively. The differences are attributed to fixed branching yields and the absence of important intermediate species and pathways in the parameterized mechanisms. Regional budgets are also analyzed within the remote Southern Ocean and central tropical Pacific. While the DMS budget varies primarily between the regions, the SO2, MSA and H2SO4 budgets vary strongly between the regions and mechanisms. The DMS oxidation products, therefore, are as sensitive to the assumed mechanisms as to the external conditions between the tropics and extratropics. To distinguish between the mechanism cases, the MATCH simulations are also compared to campaign measurements. The simulated values of DMS and SO2 are found to approximate the observations, but do not provide mechanism differentiation. The gas-phase H2SO4 and MSA simulations differ more extensively from the observations, yet the comprehensive chemistry cases give a better fit to the MSA measurements. Better fits to the MSA observations are also achieved for the two mechanisms that consider dimethylsulfoxide (DMSO) as an MSA precursor.