Circulation and Deep-Water Export at the Western Exit of the Subpolar North Atlantic

Abstract

The current system east of the Grand Banks was intensely observed by World Ocean Circulation Experiment (WOCE) array ACM-6 during 1993?95 with eight moorings, reaching about 500 km out from the shelf edge and covering the water column from about 400-m depth to the bottom. More recently, a reduced array by the Institut für Meerskunde (IfM) at Kiel, Germany, of four moorings was deployed during 1999?2001, focusing on the deep-water flow near the western continental slope. Both sets of moored time series, each about 22 months long, are combined here for a mean current boundary section, and both arrays are analyzed for the variability of currents and transports. A mean hydrographic section is derived from seven ship surveys and is used for geostrophic upper-layer extrapolation and isopycnal subdivision of the mean transports into deep-water classes. The offshore part of the combined section is dominated by the deep-reaching North Atlantic Current (NAC) with currents still at 10 cm s?1 near the bottom and a total northward transport of about 140 Sv (Sv ≡ 106 m3 s?1), with the details depending on the method of surface extrapolation used. The mean flow along the western boundary was southward with the section-mean North Atlantic Deep Water outflow determined to be 12 Sv below the σ? = 27.74 kg m?3 isopycnal. However, east of the deep western boundary current (DWBC), the deep NAC carries a transport of 51 Sv northward below σ? = 27.74 kg m?3, resulting in a large net northward flow in the western part of the basin. From watermass signatures it is concluded that the deep NAC is not a direct recirculation of DWBC water masses. Transport time series for the DWBC variability are derived for both arrays. The variance is concentrated in the period range from ?2 weeks to 2 months, but there are also variations at interannual and longer periods, with much of the DWBC variability being related to fluctuations and meandering of the NAC. A significant annual cycle is not recognizable in the combined current and transport time series of both arrays. The moored array results are compared with other evidence on deep outflow and recirculation, including recent models of different types and complexity.