704 Main Features of General Oceanic Circulation and their Physical Exploration 



pointed out in previous work that for a time-variable wind effect the three-dimensional 

 structure of the sea tends to be baroclinic, which is in fact fairly readily understood. If 

 the physical sea level inclines due to a wind influence the correspondingly generated 

 horizontal pressure gradient at first extends down to the bottom also in the case of a 

 two-layered ocean. In addition to the flow in the uppermost layers, more or less in 

 the direction of the wind producing it a flow in the lower layers will thus also occur, 

 only in the opposite direction (see Fig. 339). This latter current will increase in intensity 

 until the slope of the internal boundary surface (opposite to that of the sea surface) 

 causes the horizontal pressure gradient to disappear in the lower layers, so that the 



Wind 



Fig. 339. Transition from a barotropic type in the first phase of wind influence to the final 



baroclinic type. 



lower part of the ocean will finally be at rest. In the upper layer there will then be a 

 drift current and underneath a geostrophic flow, while in the lower layers the ocean is 

 at rest. It is therefore to be expected that the mass field of the sea and the wind- 

 generated currents of the upper layers act and react on one another and this inter- 

 relation is such as to restrict the wind-driven currents to the upper layers of the ocean. 

 This striking compensation principle between the upper and lower layers is confirmed 

 by experience and is one of the most important experimental facts of oceanography. 

 If this were not the case it would not be possible to build up a picture of oceanographic 

 conditions in the deep layers and their mass displacements on the basis of wide-spaced 

 oceanographic observations; that is, it would be impossible from observations at 

 widely differing times to form a picture of the average conditions of stratification and 

 field of flow in the deep layers. This supports the theoretical results, since if these 

 were somewhat different there would undoubtedly be a contradiction with experience 

 and the model chosen would be unsuitable for such a purpose. This problem was 

 first discussed by Rossby (1938), but his results were unsatisfactory since they gave 

 more or less barotropic flow systems which is impossible. A barotropic state can only 

 persist for a short time and finally a baroclinic state must predominantly prevail. 

 After other attempts Veronis and Stommer'ihave re-examined the problem and 

 attempted its solution by means of a large mathematical apparatus. 



