I Introduction:

A group of US southern ocean oceanographers met in Seattle on 8 and 9 February to begin the process of developing a US contribution to the iAnZone Experiment #4 "Convection" as discussed in the iAnZone report of our Biosphere-2 meeting of December 1997.

A consensus was reached that the US group will develop, in collaboration with international partners (to the extent possible), a study of the role of the Southern Ocean shelf/slope front (Slope Front) in the exchange of mass, heat, freshwater between the shelf and open ocean regimes, including exchanges that lead to deep reaching convective plumes. Of particular interest is to identify the key processes that sustain such exchange of shelf and oceanic waters across the Slope Front, and to understand the dominant dynamics of the descending plumes.

The program is in its early stage of development, but at this time it seems clear from both logistics and scientific considerations that the Ross Sea Slope Front offers the most promising site. The Slope Front in the Ross Sea is generally ice free in the summer months, enabling use of mooring and Lagrangian technology, and a greater chance of sampling of the thermohaline and tracer fields at the proper scales. Dense shelf water abounds, including Ice Shelf Water. Its bathymetry is better surveyed than other segments of the continental margins. It is close to the logical bases of McMurdo and Christchurch. The US program can be linked with the Italian program on the Ross Sea shelf and with the Australian program along the Adelie coast to the west (which probably represents the downstream condition). The alternative site of the Weddell Sea, where the common wisdom dictates is the major source of Southern Ocean dense bottom water, while still a possibility, is considered too logistically difficult and remote to allow use of required technology.

A steering committee has been set up to drive the project development: Arnold L. Gordon (Chair); Laurence Padman; Alejandro H. Orsi; and Glen Gawarkiewicz. The steering committee's task is to: develop the scientific objectives, experiment design and implementation plan for the Slope Front program, in communication with interested scientists; and develop an integrated group proposal to the Office of Polar Programs for the 1 June, 2000 proposal deadline. The year long field effort will commence in Austral summer 2001/2002.

Clearly much work needs to be done before the project can be formulated in detail. Site selection is closely linked to the science objectives. If we accept the Ross Sea site, where along the Ross Sea margins would the experiment be positioned? east or west of Pennell Bank? In a canyon or along a relatively smooth segment? Careful study of existing data is needed to select the best site. The results of an international workshop planned for 27 to 29 September 1999 at the Lamont-Doherty Earth Observatory will help shape the Slope program. International Collaboration with the Italian Ross Sea program and the Wilkes Land, Adelie Coast research of Australia, and with other national programs, will be pursued under the SCOR iAnZone Program.

The following consideration were developed at and since the February 1999 Seattle meeting.

II Hypothesis:

The Antarctic Slope Front characteristics determine exchange of mass, heat, freshwater between the shelf and open ocean regimes, particularly those exchanges that lead to deep and bottom reaching continental margin plumes and convection. The characteristics of these dense water masses and of their descent over the continental slope are determined by complex feedback processes at the Front.

III Possible Shelf-Slope Front processes associated with deep reaching sinking:

1) Instabilities and eddy processes. Mesoscale processes occurring within the vicinity of the shelf-slope front are likely to contribute to local enhancement of dense water transport across the shelfbreak. Local topographic effects may also affect eddy characteristics. Changing density contrasts between shelf and slope water masses

may affect these processes by altering frontal stability characteristics;

2) Cross-slope frontal advection. Tides and other processes influence the movement of the front relative to the shelf break, which may induce sinking. Ekman veering may sinking along front;

3) Canyon-related plumes. Smooth slope topography may be expected to produce sheet like sinking, but canyons may channel the sinking plumes, allowing faster descent, thus possibly affecting mixing and entrainment processes;

4) Thermobaric effects and cabbeling (both diapycnal and isopycnal). Non-linear equation of state properties of sea water acts to encourage or enhance descent of near surface waters;

5) Lateral mixing and thermohaline intrusions: Fine structure intrusions, observed on summer crossing of the Slope Front, are an important contribution to lateral fluxes across the slope front.

IV Specific Objectives: Evaluation and Modeling of the candidate processes.

1. Determine the circulation and mixing within the shelf-slope front and its relationship to regional currents.

2. Determine the relationship between dense water sinking and shelf-slope front behavior, regional atmospheric and ice forcing.

3. Describe and determine the governing dynamics of Shelf-Slope Front structure and position variability at tidal, synoptic and seasonal scales.

4. Determine the mean frontal structure and dynamics of slope plumes and the relationship to local topography.

5. Use the observed behavior of the Antarctic shelf-slope front and associated deep and bottom reaching sinking to evaluate and further develop models.

V Methods: Meeting of specific objectives require in situ and satellite observational and development of appropriate models. The experiment may employ many of the following methods. What methods are employed depend on their cost effective relevance to addressing the projects hypotheses.

• moorings with current meters and CTD sensors;

• regional and high resolution (localized) CTD/LADCP and tracer field surveys at time of moorings deployment and during the winter period; a less detailed survey is anticipated for the moorings recovery in 2002/2003;

• Lagrangian methods: RAFOS floats; surface drifters, purposefully released tracers;

• SeaSoar, AUV to obtain high resolution measurements of the horizontal gradients.

• regional and plume models;

• Time series of sea ice and synoptic weather from satellites;

VI Ship support:

The Palmer is required for 40 days at the Ross Sea experiment site sometimes within the period December 2001 to April 2002, and similarly 40 days are needed within the period February 2003 to April 2003. Transient time is additional to the ‘at site’ 40 days. A winter cruise of approximately 30 day duration, sometimes within July to October 2002, is required to determine the winter period structure of the Slope Front.