Decadal Air–Sea Interaction in the North Atlantic Based on Observations and Modeling Results

Abstract

The decadal, 12?14-yr cycle observed in the North Atlantic SST and tide gauge data was examined using the NCEP?NCAR reanalyses, the Comprehensive Ocean?Atmosphere Data Set, and an ocean model simulation. SST data are contrasted with tide gauge data, which shows that in SST the decadal mode is nonstationary with strong variability for the last 40?50 yr, but the sea level at the southeastern U.S. coast exhibits a robustly regular variability at this period for the last 75 yr. Thus, sea level variability gives credibility to the existence of a decadal mode. The main finding is that this 12?14-yr cycle can be constructed based on the leading mode of the surface heat flux, which implicates the participation of the thermohaline circulation. This cycle has potential to be a coupled mode because three necessary aspects of a coupled mode are found: a positive feedback between the atmosphere and ocean in the subpolar gyre, a negative feedback of the overturning variability on itself, and a delayed adjustment through midlatitude Rossby waves to these processes. During the cycle, starting from the positive index phase of the North Atlantic oscillation (NAO), positive SST and oceanic heat content anomalies exist in the subtropics. The warm anomalies advect to the subpolar gyre where they are amplified by local heat flux, that is, a positive feedback between the atmosphere and ocean. At the same time the advection of warm anomalies to the subpolar gyre constitutes of a negative feedback of the thermohaline circulation on itself. The effect of this internal feedback of the ocean is amplified by the positive feedback between the atmosphere and ocean. Consequently the oceanic thermohaline circulation slows down and the opposite cycle starts. The adjustment to the changes in overturning is mediated by midlatitude Rossby waves. They are responsible for the subtropical heat content anomalies that later advect to the high latitudes. This analysis suggests that the two principal modes of heat flux variability, corresponding to patterns similar to the NAO and the Western Atlantic, are part of the same decadal cycle.