pattern of alternative gyres was found to persist throughout the year, it was 

 found to be most prominent in July, August, and September. The two large 

 anticyclonic gyres, the Great Whirl or Prime Eddy and the Socotra Eddy, together 

 with the strong shear zone along the eastern edge of the Great Whirl observed 

 from USNS WILKES during late August and early September of 1979, are evidence 

 of this multi-cellular circulation. Circulation of this kind attests to the 

 considerable complexity and variability of the Somali Current system. 



One of the principal difficulties in obtaining a true picture of the 

 circulation in the area was the asynopticity of the data taken from USNS 

 WILKES. Unlike the meteorologist who relies on synoptic observations to 

 display and analyze weather observations on a map, the oceanographer must rely 

 on observations taken over a period of several days or weeks. Fortunately, 

 changes in the ocean are normally slow enough to permit meaningful analyses of 

 data taken from a single survey ship during a single cruise. However, recent 

 studies of satellite photographs and time-series XBT sections have indicated 

 much more rapid and pronounced oceanic variability than has been assumed in 

 classical oceanographic surveying and analysis. The satellite imagery from 18 

 to 27 August 1979 (figures 2 and 3, respectively) amply illustrates the magni- 

 tude of the changes that can occur during such a short time as nine days. The 

 ribbon of cold upwelled water between the Southern Eddy and the Great Whirl 

 that appears in the 18 August imagery is not evident at the- surface in the 

 27 August imagery. Such rapid changes lead to considerable difficulty in the 

 analyses of observations that are close to each other in space, yet taken 12 

 days apart at nearly the same geographical location (figure 4). Observations 

 154 and 344 are especially effective illustrations of the analytical diffi- 

 culties posed by a rapidly changing environment. Between 20 August 1979 and 

 1 September 1979 the 20°C isotherm rose 138 m. 



II. METHODS 



The survey was conducted from USNS WILKES (T-AGS 33) from 16 August until 

 5 September 1979. Four hundred fifteen successful shipboard expendable 

 bathythermograph (XBT) observations were taken, together with 27 salinity- 

 temperature-depth (STD) stations. The XBTs, Sippican Model T-4 and T-7 probes, 

 provided temperature profiles to approximately 450 m and 750 m respectively. 

 The STD stations, taken with the Plessey Model 9040 STD, were taken to a depth 

 of at least 1500 m. Also, sea surface temperature was monitored continuously 

 with a Model 2801A Hewlett Packard quartz crystal thermometer mounted on the 

 hull two feet below the water line. Hourly sea surface temperatures were 

 taken with a bucket thermometer, together with sea surface salinity samples 

 every two hours. At each STD station, salinity samples were also taken with a 

 Niskin rosette sampler to assist in field calibration of the salinometer. All 

 salinity samples were stored at room temperature for at least 24 hours in 

 citrate of magnesia bottles before being measured with an Autosal Model 8400 

 induction salinometer. The STD data were processed and field calibrations 

 were performed in accordance with instructions by Beller, et_ al^. (1975). 

 Offset corrections were derived and applied in the field, and the resultant 

 temperature and salinity values were found to be accurate within 0.02°C and 

 0.02 °/oo, respectively. Figure 5 presents track lines and locations of STD 

 and XBT drops during the survey. 



