Two general circulation models with sophisticated aerosol treatments, ECHAM5-HAM and CAM-Oslo, have been used to investigate how the sea surface representation influences the simulated aerosol-climate response. Two experiments were performed with ECHAM5-HAM, one with fixed sea surface temperatures (SSTs) and one where a mixed-layer ocean (MLO) model was used. With CAM-Oslo, one experiment with an MLO was conducted. In all experiments equivalent emissions of anthropogenic aerosols and aerosol precursors were used, representing the increased emissions of these between the years 1850 and 2000.
The different ocean surface representations had minor effects on the simulated anthropogenic aerosol distribution. The model-specific aerosol treatment had a larger influence on the simulated anthropogenic aerosol optical depth than the change in the ocean boundary condition. The natural aerosols distribution was, on the other hand, sensitive to the ocean surface representation in ECHAM5-HAM.
The more substantial surface temperature change and altered atmospheric thermal structure supported in the MLO experiments influenced both the surface energy budget and the global circulation. Despite separating the analysis of the surface energy budget into land and ocean areas, the choice of ocean boundary condition influenced the surface energy-flux changes over land.
The larger cooling of the northern hemisphere compared to the southern hemisphere in the MLO experiments led to a southward shift of the tropical Hadley circulation as well as the eddy-driven circulation in the northern hemisphere. In the experiment with fixed SSTs, the winter Hadley cell intensity decreased in each hemisphere. The most prominent change in the zonal mean zonal wind in the fixed SST experiment was an Arctic Oscillation-like southward shift of the mid-latitude jet in the northern hemisphere during the northern hemisphere winter.
In the experiments presented in this study, the use of fixed SST and an MLO yielded distinctly different climate impacts despite the a small difference in the simulated anthropogenic aerosol burden.