Quantifying the importance of ocean-atmosphere exchange for atmospheric chemistry

Ocean atmosphere interactions are a vital part of the Earth system, especially for the composition of the atmosphere which impacts climate, and air quality. Our understanding of this system comes from numerical models, observations and laboratory studies. The representation of ocean-atmosphere interactions in these models is often simplistic. This work aims to extend our understanding of some of these interactions by improving their representation in the chemistry transport model GEOS-Chem. A more complete representation of oceanic ozone dry deposition results in a 50% reduction in ozone deposition velocity to the ocean, bringing measured and modelled ozone fluxes into better agreement. The resulting increase in tropospheric ozone is limited to the marine boundary layer (MBL), especially over the Southern Ocean where modelled ozone concentrations are brought closer to observations. Oceanic emissions of alkenes (ethene and propene) are parameterized based on observations and laboratory studies. An emission inventory is tuned to observations and extrapolated globally. The resulting increase of alkenes in the MBL improves model comparisons. Model predictions overestimate remote observations from aircraft over open ocean, implying further development is needed to better capture differences between coastal and open ocean emissions. Oceanic emissions of ethene and propene have a negligible impact on tropospheric oxidative capacity. Spatial resolution is an important consideration for marine environments. Representation of local emissions from islands and shipping plumes, transport and vertical mixing, and steep gradients in concentration can all contribute to differences between the predictions of models at different spatial resolution. Although there are some species which show large differences (likely when the chemical and dynamical timescales are comparable) the impact of spatial resolution on OH and O3 is small. When comparing model predictions and observations for remote islands the nearest grid box to the island might not be the most appropriate comparison for some species.