The Tropospheric Circulation 649 



there is a partly closed circulation down to a depth of 80 m. In the upper half of this 

 circulation the water flows away from the coast, in the lower half towards the coast. 

 Near to the coast the water rises and in the region remote from the coast it sinks 

 along a boundary layer. This outer boundary layer itself moves away from the coast 

 and as a compensation a replacement has to be made from below (from depths of not 

 more than 200 m). In other cases dealt with by Sverdrup conditions are somewhat 

 more complicated but the essential characteristics are retained. 



In a study of the large amount of observational data, on the Californian upwelling 

 region, collected by the Scripps Institution of Oceanography in La Jolla, Defant 

 (1950, 1951) it has been shown that the piling up and upwelling processes are associated 

 with characteristic displacements of the sea surface and of the internal boundary layer 

 which gradually develop under wind influence and adjust with simultaneously formed 

 and normal to the coast occurring circulations. They finally tend towards a stationary 

 state. These condition can be illustrated by two opposite cases. During the first cruises 

 (28 February to 15 March 1949) it was found that the wind component towards the 

 coast predominated over the entire region with a maximum of 5 m/sec and caused 

 considerable piling up of water along the coast. During the second cruise (27 April to 

 15 May 1949), in contrast to the first case, the water was driven away from the coast 

 where as a consequence upwelling occurred. 



Cruise 1 thus is a typical example for a water accumulation along the coast, while 

 cruise 2 is typical for coastal upwelling. Figure 305 shows the dynamic topography of the 

 ocean surface represented by lines of equal positive and negative deviation from the 

 basic distribution produced by the Californian Current flowing south. This basic 

 distribution has been obtained by elimination of the disturbances caused by tide waves 

 and internal waves (Defant, 1950). The two cases show completely opposite trends. 

 First of all it may be noticed that the channels of positive and negative deviation 

 (shown by the contours) are more or less parallel to the coast following the wave-like 

 form of the disturbance, thereby forming a marked regular pattern. In cruise 1 the 

 coastal strip shows a pronounced positive deviation — with maximum values at the 

 coast. Outside this there is a strip of negative deviation, then farther out a strip of 

 postitive deviation, and finally a second negative strip forms the western border of the 

 region. Cruise 2 gave the same pattern with the signs reversed. 



In cruise 1 there is undoubtedly a piling up of water at the coast ; it was fully developed 

 at the beginning, but during the remainder of the cruise (about two weeks) it could be 

 maintained to this extent only if the tangential wind stress towards the coast exactly 

 balances the pressure gradient of the sloping physical sea surface. The water masses 

 piled up on the continental shelf are drawn from the oceanic strip just off the conti- 

 nental slope; there the sea level consequently lies somewhat deeper (trough-like 

 form). This disturbance then develops wave-like oscillations farther westwards and 

 generates the adjoining disorders. Exactly the same applies to cruise 2 but instead of 

 piling up of water a depression in water level occurs. Consequently, to these primary 

 disturbances the adjacent displacements in the sea level thus take place in the reversed 

 order. 



The dynamics of the processes of upwelling and removal of water as a surface drift 

 requires that the rise and fall of the physical sea surface should be accompanied by a 

 corresponding fall and rise in the density transition layer. In these processes (close to a 



