i 9 o DISCOVERY REPORTS 



moves to the west, while below the water rises to the east. The trade wind, however, does not extend 

 right into the coast, and so inside its coastal boundary the surface-water is not drifted to the west, but 

 moves north purely by virtue of the distribution of density. This agrees with the conception of a one- 

 sided divergence which Defant derived from the surface-current pattern and showed in his fig. 7. 



Sverdrup (1938) was more fortunate than Defant in having at his disposal detailed repeated 

 observations off the coast of California. From these he was able to calculate directly a vertical 

 velocity-profile across the upwelling region, and to relate it to the prevailing wind vectors as recorded 

 at coastal stations during the surveys. From this, Sverdrup argues that as upwelling occurs along the 

 coast, the dense upwelled water is in time transported offshore as was the surface-water which it arose 

 to replace. Eventually conditions will be set up whereby a convection cell develops between the 

 denser water lying on the surface and the lighter offshore water, this cell becoming sharply enough 

 defined to be regarded as a boundary. 



That such a mechanism operates on the South-west African coast we cannot be certain, but the 

 available evidence strongly suggests that it does. The pattern of the isosteres on the Orange river line 

 (survey II, Fig. 37), which is probably the most complete section through an area of active upwelling, 

 is the best guide to an interpretation of such movements. Assuming the surface-waters to be under 

 the stress of a wind effective in their offshore transport, the distribution of mass on this section can be 

 interpreted on the basis of the water-movements suggested by the arrows. The convection cell between 

 stations WS 1053 and 1054 marks the boundary between the coastal and oceanic water types, to which 

 reference has already been made, and seems to act like a roller bearing between the two systems of 

 water-movement. Movement of this boundary to the west (left) would be accompanied by an 

 upwelling from about 300 m. depth. 



Depths affected by upwelling 

 The most straightforward method of determining the depth of upwelling is to utilize the T-S diagrams. 

 From these it is possible to determine from what depth in the offshore water the coastal waters 

 originate. On p. 179 it was shown that in the process of uplift from subsurface depths, the coastal 

 waters remained discrete and that a rise in their temperature was the only important change that took 

 place. If then we take the origin of the T-S curve for the coastal waters at the point where it arises 

 from the T-S curve of the offshore water, the depth on the offshore T-S curve will give us a measure 

 of the depth of origin of the coastal water. These depths are set out in Table 9. The first figure shows 

 the depth from which water is elevated to the surface. Where a range of depths are shown the second 

 figure represents the maximum depth affected, but it is not necessarily implied that water is brought 

 from this depth right to the surface. It is normally elevated from this depth to a somewhat lesser 

 depth on the continental shelf. 



Table 9. Depth of upwelling 



