General Circulation Models, also known as Global Climate Models (GCMs), apply the same physics as models used for Numerical Weather Prediction. GCMs are not used to predict tomorrow's weather, however. They are run for very long periods of time to study long-term variation in the Earth's climate system. The first generation of GCMs did not take the Earth's oceans into account. Yet there is an important relationship between the atmosphere and the ocean, with fluxes of heat, momentum, moisture and other properties occurring at the air-sea interface. Current thinking maintains that accurate prediction of the response of the global climate to stresses like increasing carbon dioxide concentrations requires that the world's oceans be included in the model. Funded by the EESD Programme, the NCAS Centre for Global Atmospheric Modelling at the University of Reading focused its attention on the Northern Atlantic. The methodology employed by the NCAS Centre for Global Atmospheric Modelling was to run four to five different GCMs coupling atmosphere and ocean. NCEP reanalysis data for years between 1958 and 1998 was fed to the models in order to identify natural decadal scale variation. Driven by fluctuations in sea surface temperature (SST), the North Atlantic Oscillation (NAO) phenomenon has been shown to strongly affect European climate. For instance, NAO can raise the temperature in some regions, while lowering it in others, bring abundant rain to some regions, but drought to others. Nearly all regions of the Atlantic are involved in the complex web of system responses, with a significant time lag developing. In particular, the Meridional Overturning Circulation, caused by the sinking and spreading of cold water, was strongly affected by the initial state of the ocean. It was found that model resolution on the order of forty kilometres was sufficient for capturing the essential physical phenomenon. As international concern for climate change grows, comprehensive efforts, such as this by the NCAS Centre for Global Atmospheric Modelling, to advance knowledge in the area of ocean-atmosphere interactions are crucial.