Variability of the Overflow Water Transport in the Western Subpolar North Atlantic, 1950–97

One of the major topics in current field research is the question of whether or to what extent the North Atlantic Ocean is subject to changes in water mass transports, and how they are related to atmospheric phenomena like the North Atlantic Oscillation (NAO). Bottle and CTD data from the 1950s to 1990s are presented to reconstruct spatially and temporally the baroclinic contribution to the deep water transports in the western subpolar North Atlantic. The focus is on the two densest components of North Atlantic Deep Water: the Gibbs Fracture Zone Water (GFZW) and the Denmark Strait Overflow Water (DSOW). Direct velocity measurements in the considered time period are sparse. For this reason it was decided to calculate the geostrophic velocity relative to 1400 dbar. This level is located in the weakly stratified Labrador Sea Water. The combined baroclinic volume transport of GFZW and DSOW during the early 1990s was about 5 Sv (Sv ≡ 106 m3 s−1) in the Irminger Sea and 7–8 Sv in the Labrador Sea. Near the Flemish Cap, baroclinic transports reached 16–29 Sv. Because of the impact of the North Atlantic Current on the flow field resulting in steeply sloping isopycnals, the latter estimate is strongly dependent on the choice of the reference level, in contrast to other locations. Time series were obtained from data in the Irminger and Labrador Seas. In the Irminger Sea, the combined baroclinic transport of GFZW and DSOW increased from 4–5 Sv in the mid-1950s to 8–9 Sv in the 1980s, followed by a decrease to 5 Sv in the 1990s. In the Labrador Sea, the temporal variability was stronger (3–11 Sv), with interannual changes of 5–6 Sv. The importance of baroclinic transport variability is not easy to interpret. Results presented herein indicate that relations of the Irminger and Labrador Seas time series to the NAO remain ambiguous. Among other impacts the presence of eddies significantly affects the time series of baroclinic transport. Whether baroclinic variability represents the total variability of the flow (baroclinic and barotropic part) cannot be assessed without knowledge of the variability of the velocity field in the reference level.