Variability of the North Atlantic Oscillation and the Tropical Atlantic dominate the climate of the North Atlantic sector, the underlying ocean and surrounding continents on interannual to decadal time scales. Here we review these phenomena, their climatic impacts and our present state of understanding of their underlying cause. Copyright

A new, speculative, and, we hope, provocative summary of the North Atlantic circulation is described, including both horizontal currents (wind‐driven) and the primarily (thermohaline) meridional flows that involve the transformation of warm to cold water at high latitudes. Our picture is based on a synthesis of a variety of independent investigations that are contained in the literature as opposed to a presentation of the results of one technique or the point of view of one author. We describe a thermohaline cell (the so‐called thermohaline conveyor belt) that is concentrated within the Atlantic and Southern oceans (rather than essentially global), with the most important upwelling sites being in the circumpolar and the equatorial current regimes. We concentrate on deep water formation and its replacement relative to intermediate‐water formation. It has been pointed out recently that the formation of 13 Sv (1 Sv = 106 m³ s−1) of southward flowing North Atlantic Deep Water is compensated for in the upper ocean by northward cross‐equatorial transport. We suggest that this thermocline layer flow passes through the Straits of Florida, transits the Gulf Stream system on its inshore side, and exits through the North Atlantic Current system after recirculation and modification. There is now a clear observational basis for the structure of recirculating gyres on the southern and northern sides of the Gulf Stream. We suggest a recirculation for the North Atlantic Current as well. We also describe a C‐shaped component to the southern Gulf Stream recirculation and identify a roughly 10‐Sv circulation in the eastern North Atlantic associated with the Azores Current. Recirculations play an important role in deep boundary current regimes and in water mass formation and modification. The transport of the deep western and northern boundary currents in the North Atlantic Ocean may be boosted (roughly doubled or tripled) by counterclockwise recirculating gyres and by additions of modified bottom or intermediate water. While the North Atlantic is the most completely observed ocean, there are still significant gaps in our knowledge of its circulation.

Observational evidence is presented for interannual to interdecadal variability in the intensity of the North Atlantic gyre circulation related to the atmospheric North Atlantic Oscillation (NAO) patterns. A two-point baroclinic pressure difference between the subtropical and subpolar gyre centers-an oceanic analogue to the much-used sea level pressure (SLP)-based atmospheric NAO indices-is constructed from time series of potential energy anomaly (PEA) derived from hydrographic measurements in the Labrador Basin and at Station S near Bermuda. Representing the upper 2000-db eastward baroclinic mass transport between the two centers, the transport index indicates a Gulf Stream and North Atlantic Current that gradually weakened during the low NAO period of the 1960s and then intensified in the subsequent 25 years of persistently high NAO to a record peak in the 1990s. The peak-to-peak amplitude difference was 15-20 megatons per second (MT s Ϫ1 ) with a 43yr mean of about 60 MT s Ϫ1 a change of 25%-33% occurring between 1970 and 1995. The timing of the ocean fluctuation is organized around the same temporal structure as the NAO index. The two are not directly covariant, but to first order, the ocean signal reflects a time integration, through mixed layer ''memory'' and Rossby wave propagation, of the atmospheric forcing.To some degree, the gyre PEA histories are fluctuating in antiphase reflecting latitudinal shifts of the surface westerlies across the North Atlantic. Differences in forcing mechanisms and baroclinic responses in each gyre, however, are reflected by divergences in the details of their PEA histories. The subpolar PEA changes are primarily thermally driven through diabatic mixing and surface buoyancy fluxes associated with water mass transformation. Salinity changes, stemming from the occasional passages of low-salinity surface lids (''Great Salinity Anomalies'') through the region, contribute relatively little to the Labrador Basin PEA variability. The interior subtropical gyre PEA history is dominated by quasi-adiabatic vertical displacements of the main pycnocline, and supplemented by changes in the locally formed subtropical mode water as well as by changes in middepth density structure related to advective-diffusive import of Labrador Sea Water.Multiyear composite fields of North Atlantic potential energy centered in time on the extreme high and low transport periods provide a broad geographic context for the transport index. Basin-scale shifts of oceanic baroclinic pressure gradients between the extreme phases reinforce the sense and amplitude of changes reflected in the Bermuda-Labrador Basin transport index. * Woods Hole Oceanographic Institution Contribution Number 10260.

No abstract

No abstract

Data collected in 1988–1989, as part of the South Atlantic Ventilation Experiment, have been combined with the historical database to study the circulation and water mass variability of the abyssal water in the Argentine Basin. A map of potential temperature at 4000 m used as an indication of geostrophic shear defines a south and western intensified crescent‐shaped abyssal recirculation. Within this recirculation, and its northward extension to the Brazil Basin, Antarctic Bottom Water (AABW) properties have undergone two modifications during the 1980s: (1) The water mass cooled (0.05°C) and freshened (0.008 in salinity ratio) on surfaces of constant density. (2) The densest layer of AABW was altered to less dense water through mixing or advection out of the study area. This water mass change does not appear to have affected the flow pattern. Data collected in 1983 and 1988 to the north in the Brazil Basin show penetration of the freshwater mass in the deep western boundary current to between 18°S and 10°S, indicating very rapid propagation of the anomaly from the Argentine Basin into the Brazil Basin as a deep western boundary current. It is suggested that open ocean convective events within the Weddell Sea contributed to the change in AABW documented here.

The spatial distributions of certain sea-surface properties, such as temperature, fluctuate on timescales from months to decades and in synchrony with the main regional atmospheric patterns comprising the global climate system 1 . Although it has long been assumed that the ocean is submissive to the dictates of the atmosphere, recent studies raise the possibility of an assertive, not merely passive, oceanic role in which water-mass circulation controls the timescales of climate fluctuations 2-6 . Previously held notions of the immutability of the physical and chemical characteristics of deep water masses are changing as longer time series of ocean measurements indicate that the signatures of varying sea-surface conditions are translated to deep waters 4,7 . Here we use such time-series measurements to track signals 'imprinted' at the sea surface in the North Atlantic Ocean's subpolar Labrador Basin into the deep water of the subtropical basins near Bermuda, and infer an approximately 6-year transit time. We establish a geographic and temporal context for a portion of the long-term warming trend reported for mid-depth subtropical waters over the past 40 or so years 8,9 , and we predict that waters at these depths will continue to cool well into the next decade.

No abstract