a large number of measurements. On the basis of our estimate 

 of spatial variability, 23 current meters would be required to 

 estimate a spatially-averaged current within ± 1 cm /sec, equal 

 to a transport within ± 25 Sv. It is concluded that an array 

 of horizontally-integrating measurements, such as pressure or 

 sea-level measurements on each side of the passage, is the only 

 feasible method of long-term monitoring of the variations in 

 transport of the Antarctic Circumpolar Current through the 

 Drake Passage. 



The 3-week average speeds from the array of 15 cur- 

 rent meter moorings discussed above were used in conjunction 

 with geostrophic calculations to estimate the short-term trans- 

 port of the Antarctic Circumpolar Current. Five closely spaced 

 hydrographic sections were made across the Drake Passage 

 during this 3-week period in March 1975. The geostrophic 

 transport relative to 3,000 decibars averaged 95 Sv for the 

 five transects; this is consistent with estimates based on previ- 

 ous hydrographic measurements. Referencing the geostrophic 

 transport to the current meter measurements gave an adjusted 

 transport of 124 Sv to the east. 



During F DRAKE 76, the small-scale structure of the 

 polar front was studied in the Drake Passage. Two ships, the 

 RV Thompson (University of Washington) and the Chilean 

 naval vessel AGS Yelcho, combined efforts in the program 

 that took place during the onset of the austral autumn. De- 

 scribed here is the evolution of the polar front between Febru- 

 ary 29 and April 5. 1976. 



The polar front, in reality a transition zone between ant- 

 arctic and subantarctic water masses, is most clearly revealed 

 in vertical sections of temperature as shown in fig. 17. South of 

 the frontal zone is a temperature minimum layer between 100 

 and 300 m. This feature is insulated from seasonal heating at 

 the surface. Within the frontal zone, the minimum deepens 



abruptly and erodes, splitting into multiple extremes of inter- 

 leaving antarctic/ subantarctic waters. Although numerous defi- 

 nitions of the polar front have been suggested, the operational 

 definition used here is the presence of a well-defined tempera- 

 ture minimum layer as delineated by the 2°C isotherm. The 

 selection of the 2°C isotherm is somewhat arbitrary, but it 

 does provide a convenient, consistent, and objective method 

 of defining the position of the southern edge of the frontal 

 zone. 



Synoptic maps of the front were produced with the com- 

 bined data from the two vessels. Although some artistry is 

 involved in smoothing the observations, the horizontal density 

 of measurements is sufficient to resolve the structure. Figure 

 18, a composite of these maps, indicates that between March 

 10 and 24 a large meander developed in the polar front. This 

 was verified by thermistor chain measurements (position 

 shown in fig. 18) from a subsurface mooring in the vicinity 

 and by the distribution of surface properties (not shown). 

 The meander subsequently pinched off a ring of between 60 

 to 80 km diameter, which then drifted to the NE at a speed 

 of about 10 cm s-\ The ring formation process is reminiscent 

 of that observed in western boundary currents such as the 

 Gulf Stream. Continuous profiles of temperature and salinity 

 show that the antarctic water types inside the ring penetrate to 

 a depth of 2,500 m, suggesting that much of the water column 

 is involved. 



Subsurface neutrally buoyant floats equipped with fins to 



react to vertical water motion, were tracked acoustically from 

 the RV Thompson during the period of the ring formation. 

 The track of these floats indicates that three of the four vertical 

 current meters (VCM) were entrapped in the ring. Horizontal 

 velocities of the floats were 30 to 40 cm s-\ 



This process can potentially transport large amounts of 

 antarctic or subantarctic water across the front and may be 

 fundamental to the understanding of heat and salt transfers 

 toward the polar region. If the cold ring is reabsorbed into the 

 circumpolar current, the net transfer of properties will be less 

 than if the ring were to remain detached, slowly decaying with 

 time. Downstream of the Drake Passage, the circumpolar 

 current system is highly constrained by the Scotia Ridge, which 

 has only two large openings with a sill depth >2.5 km. It is 

 interesting to speculate upon the fate of this cold water ring 

 when it drifts into this rather formidable topography. 



During January and February of 1977, the RV Melville 

 recovered the array of tide gauges, deep-sea pressure gauges, 

 and current /temperature pressure recorders that were de- 

 ployed across Drake Passage during February 1976 as a part 

 of F DRAKE 76. While in this region, the Melville deployed 

 another year-long instrument array, and short-term experi- 

 ments and supporting hydrographic /STD work was done. This 

 field operation was called F DRAKE 77. 



As the final phase of F DRAKE, a 1-year clustered array 

 was deployed in January and February 1977 (figs 19 and 20) 

 to meet the following objectives: 



1. Determine the spatial variability of the currents with 

 meter spacings from 15 km to 60 km. 



2. Describe the movements of sharp bands of currents 

 through the cluster and to identify rings formed at the 

 Polar Front Zone. 



3. Aflow continuous geostrophic comparisons and to 

 compare local changes of temperature with horizontal 

 advection of temperature to understand the importance 

 of nonlinear effects. 



4. Extend time series of currents, temperatures, and 

 pressures for a third year. 



The CTD/STD program included measurements of tem- 

 peratures, salinities, and dissolved oxygen concentrations using 

 bottles on the CTD wire and a rosette sampler. The objectives 

 for making these measurements are to: 



1. Define the density field in the region of the moored 

 cluster and at the pressure recorders at the time of de- 

 ployment of the 1977 array. 



2. Make a section along the line of the 1976 moored 

 array to obtain another short-term estimate of transport 

 through the Drake Passage. 



3. Obtain clusters of hydrographic stations at a series 

 of spacings around a 1977 mooring heavily instrumented 

 in the vertical for use in comparing directly measured 

 with calculated vertical shears. 



Bottom pressure measurements at 500 m depth on either 

 side of the Drake Passage will be continued to monitor tem- 

 poral fluctuations in pressure difference across the Passage. It 

 is expected that this program of simultaneous pressure meas- 



33 



