Sources, transport, and sinks of SO 2 over the equatorial Pacific during the Pacific Atmospheric Sulfur Experiment

The Pacific Atmospheric Sulfur Experiment (PASE) is the first sulfur-budget field experiment to feature simultaneous flux measurements of DMS marine emissions and SO2 deposition to the ocean surface. We make use of these data to constrain a 1-D chemical transport model to study the production and loss pathways for DMS and SO2 over the equatorial Pacific. Model results suggest that OH is the main sink for DMS in the boundary layer (BL), and the average DMS-to-SO2 conversion efficiency is ~73%. In an exploratory run involving the addition of 1 pptv of BrO as a second oxidant, a 14% increase in the DMS flux is needed beyond that based on OH oxidation alone. This BrO addition also reduces the DMS-to-SO2 conversion efficiency from 73% to 60%. The possibility of non-DMS sources of marine sulfur influencing the estimated conversion efficiency was explored and found to be unconvincing. For BL conditions, SO2 losses consist of 48% dry deposition, while transport loss to the BuL and aerosol scavenging each account for another 19%. The conversion of SO2 to H2SO4 consumes the final 14%. In the BuL, cloud scavenging removes 85% of the SO2, thus resulting in a decreasing vertical profile for SO2. The average SO2 dry deposition velocity from direct measurements (i.e., 0.36 cm sec−1) is approximately 50% of what is calculated from the 1-D model and the global GEOS-Chem model. This suggests that the current generation of global models may be significantly overestimating SO2 deposition rates over some tropical marine areas. Although the specific mechanism cannot be determined, speculation here is that the dry deposition anomalous results may point to the presence of a micro-surface chemical phenomenon involving partial saturation with either S(IV) and/or S(VI) DMS oxidation products. This could also appear as a pH drop in the ocean’s surface microfilm layer in this region. Finally, we propose that the enhanced SO2 level observed in the lower free troposphere versus that in the upper BuL during PASE is most likely the result of transported DMS/SO2-rich free-tropospheric air parcels from the east of the PASE sampling area, rather than an inadequate representation in the model of local convection.