Simulated interannual to decadal variability irk the tropical and sub-tropical North Atlantic

The dominant pattern of tropical and subtropical North Atlantic sea surface temperature (SST) anomalies simulated in the GFDL coupled ocean-atmosphere model is identified and compared to observations. The spatial pattern and temporal variability of that pattern resemble observational results. On interannual time scales it is shown that anomalous surface heat fluxes, consistent with variations in the intensity of the subtropical high pressure system in the atmosphere and the associated Northeasterly Trade winds, appear to be the most important process for generating this SST pattern. This relationship is also true on decadal time scales, although the relative role of oceanic heat advection is somewhat larger than on the interannual time scales. SST variations in the tropical Atlantic with SST variations in a 200 years long climate simulation with the global coupled ocean-atmosphere model developed at the Geophysical Fluid Dynamics Laboratory (GFDL). It was shown in MD that the model-generated decadal SST variations in the tropical Atlantic resemble the observed decadal SST variations. Further, this preliminary analysis also showed that there is little coherence between model-generated decadal SST variations on the two sides of the Equator. The purpose of the present work is (a) to further clarify characteristics of the model-generated decadal SST variations in the tropical and sub-tropical North Atlantic in light of the more fiecent analyses of observations and (b) to analyze the physic!; of these model-generated decadal variations. We ask the following questions: (I) Do the decadal SST anomalies in the mo1jel show characteristics similar to the observed decadal SST anomalies? (2) What is the relationship among simulated anomalies in SSTs, sub-surface ocean temperatures, sea level pressure, near-surface winds, and ocean-atmosphere heat fluxes? (3) Which physical processes appear to be most important for generation of the decadal SST anomalies? 2. Des:cription of Model and Numerical Experiments