FY 1999

One-dimensional modeling of sulfur species during the First Aerosol Characterization Experiment (ACE 1) Lagrangian B

Mari, C., K. Suhre, R. Rosset, T.S. Bates, B.J. Huebert, A.R. Bandy, D.C. Thornton, and S. Businger

J. Geophys. Res., 104(D17), 21,733–21,749, doi: 10.1029/1999JD900022 (1999)

A one-dimensional Lagrangian model is used to simulate vertical profiles and temporal evolution of dimethylsulfide (DMS), sulfur dioxide (SO), aerosol methane sulfonate, and non-sea-salt sulfate (nss sulfate) that were measured during the three flights of the second First Aerosol Characterization Experiment (ACE 1) Lagrangian (Lagrangian B) experiment. Entrainment rate, mixing heights, and cloud occurrence are calculated prognostically in this type of model. The model is forced by geostrophic winds and large scale subsidence from European Centre for Medium-Range Weather Forecasts (ECMWF) analysis and sea surface temperature measured on board Research Vessel Discoverer. Gas phase oxidation and heterogeneous oxidation of SO₂ to nss sulfate in clouds and sea-salt particles are considered. The evolution of dynamical variables in the column is found to be well reproduced by the model. The model captures 82% of the variance of observed DMS assuming OH is the only oxidant and a DMS flux term calculated from Liss and Merlivat [1986] parameterization and seawater DMS concentrations measured aboard R/V Discoverer. However, uncertainties in DMS oxidation rates and regional seawater concentrations are too great to identify a best fit wind speed-transfer velocity relationship. SO₂ mixing ratios are correctly represented in the model (least squares correlation coefficient r² = 75%) using a DMS to SO₂ conversion efficiency of about 70%. Oxidation of SO₂ in sea-salt particle appears to be a dominant process and controls SO₂ lifetime during the Lagrangian B at least in the well mixed lower layer. Removing heterogeneous loss of SO₂ in sea salt significantly deteriorates the simulation (r² = 50%). Under cloudy conditions, heterogeneous loss in cloud droplets and in sea-salt particles are competitive (relative rates are 35% and 41%, respectively, during flight 26). Model-generated aerosol methane sulfonate mixing ratios agree with the observations (r² = 62.5%) when high branching ratio for an addition oxidation pathway is used. The model estimates nss sulfate mixing ratios with little bias (median simulated-to-observed concentration ratio 1.03 and slope of the regression line 0.7) but captures only one third of the observed variance of nss sulfate. Part of the discrepancy could be due to the assumption of a decrease of nss sulfate mixing ratios with altitude in the model, whereas observations revealed high concentrations at 4500 m during the last two flights suggesting that horizontal transport could be more important than vertical mixing in this region. Nss sulfate is found to be produced photochemically under non cloudy, low wind speed conditions encountered during the first flight. During the last two flights, nss sulfate is produced mainly by oxidation in cloud droplets (48% during flight 25 and 69% during flight 26) and sea-salt particles (50% during flight 25 and 22% during flight 26).