1. US DOVETAIL PROPOSAL OVERVIEW
This proposal describes the US portion of the international program for study of Deep Ocean Ventilation Through Antarctic Intermediate Layers (DOVETAIL), a program which has been conceived and organized by the international Antarctic Zone (iANZONE) group. The primary goal of iANZONE is to advance our quantitative knowledge and modeling capability of the seasonal cycle and interannual variability of the ocean and its sea ice cover, with emphasis on climate relevant fluxes which couple the Antarctic Zone to the atmosphere and to the Global Ocean. The first iANZONE field activity was Ice Station Weddell in 1992, directed at exploration of the environmental conditions along the western margin of the Weddell Sea [ISW Group, 1993] and at the formation and spreading of Antarctic Bottom Water. The second activity was the Antarctic Zone Flux (ANZFLUX) experiment in 1994, in which heat fluxes within the winter mixed layer, sea ice and atmospheric boundary layer were precisely measured [McPhee, 1995]. These two activities focused upon processes associated with ocean ventilation within the polar waters.
DOVETAIL builds upon the two preceding iANZONE programs and enters a second iANZONE phase whose purpose is to better define and understand the role of Antarctic waters and processes in the global ocean and climate system. DOVETAIL proposes to focus on escape of the recently ventilated deep water from the Weddell Sea into the Global Ocean - the final stage in its role of ventilating deep ocean waters. The DOVETAIL study region is shown in Figure 1.
Figure 1. Chart of the Weddell-Scotia Confluence region, showing proposed summer and winter cruise tracks (dashed and solid lines, respectively, with bold italicized and regular numbers indicating stations at transect ends) and mooring locations (large dots numbered 1-21). Heavy dashed lines north and south of the South Scotia Ridge show approximate respective locations of the Scotia and Weddell frontal systems. Moorings 7-18 are proposed for the US program, while 1-6 and 19-21 are planned as part of the coordinated foreign program.
DOVETAIL priorities parallel, and the results will contribute to, ongoing global change research. The processes responsible for vertical and horizontal fluxes within the ocean and associated interaction with the sea ice and atmosphere in polar regions must be properly represented. The DOVETAIL study region, off the tip of the Antarctic Peninsula, serves as the primary gateway between the southern polar waters and the global ocean. This region can therefore be considered as a "vital" location with respect to discharge of cold Antarctic Water into the global ocean. Results from the DOVETAIL experiment will aid in establishing a basis for long-range monitoring of this critical region, inasmuch as both the Global Ocean Observing System (GOOS) and the ocean component of the Global Climate Observation System (GCOS) have been established by a number of international bodies to provide such monitoring data.
The overall DOVETAIL goal can be stated as follows:
To understand physical processes in the Weddell-Scotia Confluence region sufficiently to quantify the ways in which it influences ventilation of the World Ocean by Weddell Sea water.
As discussed below, the Weddell-Scotia Confluence is thought to represent a gateway for the most direct and largest contribution of these Antarctic waters. It is imperative that we understand the associated physical processes in order that we be able to assess their sensitivity to changes in regional forcing, hence, the impact of such changes on Global Ocean ventilation.
1.2. Scientific Background
Approximately 57% of the deep ocean is colder than 2.0=F8C [Gordon, 1991]. The North Atlantic Deep Water, which is the northern hemisphere source for this deep water, has a characteristic temperature near or above 2.0C. Therefore most of this deep, cold water mass must be influenced to some extent through admixture of Antarctic Bottom Water, the only other available source having temperatures below 2.0C. Locarnini  estimates that 53% of the Antarctic water which ventilates the deep ocean originates in the Weddell Sea. It has been estimated [Gordon, 1975; Gordon and Taylor, 1975; Gordon and Huber, 1990] that a total of about 40 Sv Antarctic Bottom Water is needed to ventilate the World Ocean, so about 21 Sv must originate from the Weddell. Less than 6 Sv of this total can reasonably be supplied as topographically controlled bottom boundary currents which flow north through gaps in the Scotia Ridge [Fahrbach, 1994; Muench & Gordon, 1995].
It is hypothesized that the remaining 15 Sv of Weddell Sea water must be transferred north through the Weddell-Scotia Confluence region. Much of this transfer likely occurs in association with the Weddell and Scotia fronts via processes which include isopycnal and diapycnal mixing, parcel subduction and mean flow instabilities.
The Antarctic Circumpolar Current (ACC) in the Scotia Sea comprises three distinct fronts; the Subantarctic, Polar and southern ACC fronts [Orsi et al., 1995]. About 100 Sv of the estimated total transport though Drake Passage (134 Sv were estimated from direct measurements during ISOS [Nowlin and Klinck, 1986]) are carried continuously about Antarctica by these major fronts. Regional frontal zones like the Scotia and Weddell fronts have also been identified.
Water characteristics in the WSC region cannot be explained by lateral mixing of its adjacent waters from the ACC and Weddell Sea [Whitworth et al., 1994]. Among possible mechanisms which have been invoked to account for the WSC water are winter convection [Deacon, 1937], lateral and vertical boundary mixing [Patterson and Sievers, 1980], injection of meltwater from ice shelves farther south [Patterson and Sievers, 1980], and advection of shelf waters which have been conditioned along the eastern side of the Antarctic Peninsula [Gordon and Nowlin, 1978; Whitworth et al., 1994]. It seems probable that the last is the primary source of WSC water, but additional field observations and direct current measurements are needed to test and quantify this mechanism.
Transfer of water northward through the WSC region is, likewise, not well understood. We hypothesize above that 15 Sv of water must move north through the WSC in addition to the transport contained in topographically trapped bottom boundary currents. Available measurements of currents in the WSC region are inadequate to describe the regional circulation, however, currents derived using numerical model results suggest that the mean circulation is primarily zonal (Figure 2), offering little in the way of a meridional advective mechanism. Past field work in the region [e.g. Foster and Middleton, 1984; Muench et al., 1990] and instantaneous results from the Semtner and Chervin  model show energetic mesoscale activity, however, which might lead to significant meridional transports.
Figure 2. Four year mean currents at 117.5 m (topmost), 1542.5 m (middle) and 3025 m (bottom) derived from the Semtner and Chervin  model. Antarctic Peninsula (AP) is on the lower left, and the South Orkney Islands (SOI) are situated near the center of the figure. The Scotia (SF) and Weddell (WF) fronts show clearly as bands of stronger zonal north and south, respectively, of the South Orkneys. The deepest currents show northward flow through the gap west of the South Orkneys, but meridional flows are in general weak.
Neutral density surfaces [McDougall, 1987] are close approximations to isentropic surfaces, thus they can provide a useful approach to study water mass interactions in the WSC region because they include thermobaric and other nonlinear effects of the equation of state. For example, water lying at or near the shelf break of the northwestern Weddell Sea shelf can be traced on neutral surfaces to depths greater than 3500 m in the northern Scotia Sea (Figure 3). In this way, waters from the eastern shelf off the Antarctic Peninsula can be assumed to ventilate the Georgia and Scotia basins.
Figure 3. Meridional section showing neutral density surfaces from 62 to 57.5S across the South Scotia Ridge and demonstrating the contribution of mid-depth to deep Weddell Sea waters to Antarctic Bottom Water in the Global Ocean [Orsi, 1995 personal communication].
Tidal currents are a significant component of the total velocity field in many parts of the Weddell-Scotia Confluence region. Tides can affect the regional hydrography in several ways which include increasing the benthic and underice stress, generating mean lagrangian and eulerian circulations parallel to sloping topography, and generating baroclinic tidal and other higher frequency internal gravity waves that then provide energy to enhance diapycnal mixing rates in the pycnocline. Padman and Dillon  and Padman  consider some of these processes in the Arctic, while Foster et al.  consider them for the southern Weddell Sea.
Each of the tide-related processes has repercussions for the present proposed study. Increased benthic mixing raises the rate at which the boundary flows of dense WSBW are modified by entrainment of the ambient water through which it flows. Mean lagrangian flows paralleling bottom contours provide a mechanism which, in addition to vorticity conservation, steers the WSBW around topographic features. Turbulence associated with internal gravity waves modifies the diapycnal mixing rate which can then affect the entrainment rate for the WSBW bottom plumes, mixing in the pycnocline, and entrainment of the pycnocline by mixed layer turbulence whether it be driven by shear stress or convection.