Figure 25. Horizontal Temperature 



Distribution at 2000 Meters. 

 Contour Interva I - . 1 °C . 



5. Dynamic Topography. 



Figures 26 through 35 show the current systems in portions of 

 the survey region as determined from dynamic computations. 



The method of Defant (1961) was used to select a reference level. 

 The 700-decibar surface appeared to be the best choice (Figs. 26 through 

 31) although the data indicated that it was far from perfect. Other 

 levels, such as the 1000-decibar surface, also appear to be acceptable, 

 but the use of a deeper level would have reduced the number of stations 

 which could be included. Three charts using a 1000-decibar reference 

 level were constructed (Figs. 32 through 34), and they are presented 

 here for comparison. In those locations which overlapped, the charts 

 employing the 700-decibar reference surface and the charts using the 

 1000-decibar reference level were quite similar. 



Stations for which observations did not exist to the selected 

 reference level are not shown on the dynamic topography charts, and 

 the dynamic computations were not extended into shallow water. 



As shown above, the accuracy of the temperature and salinity 

 measurements made during the 1964 EDISTO survey were probably about 

 +0.02°C (or less) and approximately +0.02% , respectively. In cold 

 waters, temperature errors of about +0.02°C will have only a small 

 affect on dynamic computations. If there was a systematic error 

 in salinities of about +0.02 %o between stations, then with a reference 

 level of 700-decibars, differences in dynamic heights at the surface 

 could hardly be significant unless they exceeded 1 dynamic centimeter 

 (Stefansson 1962). 



28 



