406 BOUNDARIES OF THE SEA 



processes of importance may be found along concentrated diver- 

 gence or convergence bands in the oceans. In the model it is 

 assumed that the convergence in the surface layer is a smooth 

 function and is determined by the wind stress, but it is not im- 

 possible that real "cracks" also occur in the surface along which 

 water sinks or rises. The computation of the vertical motion below 

 the frictional layer from the geostrophic flow is also critical, and 

 even small nonlinear terms may become important. These points 

 may be of some general importance, since several later theories 

 dealing both with the wind stress currents and the thermohaline 

 circulation retain essential parts of the Sverdrup model. 



If we consider the ocean as a system "forced" by the atmos- 

 phere, then one comes easily over to the problem of the atmospheric 

 feedback. This is a most important one both for meteorologists 

 and oceanographers, but it is also one of the most difficult to 

 tackle. For a specific example of a feedback link, let us look at the 

 wind stress transport problem. Assume that for some reason there 

 is a small change in the wind stress acting on the sea. This will 

 change the motion of the ocean current somewhat, and, hence, 

 also the heat advected by it. Now, part of this heat is given to the 

 atmosphere. If the heat transport in the sea is changed, the 

 differential heating of the atmosphere is changed also. A change 

 in this differential heating finally changes the winds and thus also 

 the wind stress over the sea, whereby the circuit is closed. One 

 may now ask whether or not an original perturbation running 

 through the circuit is amplified or diminished. If it is amplified, 

 we may eventually expect an instability that can result in oscil- 

 lations, but if it is diminished, we may expect the system to return 

 to its equilibrium. In the present case we can make a rough numer- 

 ical estimate of the circuit. Consider, for example, the North Atlan- 

 tic. The w^ind stress w^ork over this basin is of the order of 10^ kw. 

 This figure gives the rate at which energy is given off to the 

 currents by skin friction and the pressure thrust on waves. Of the 

 heat advected by the currents an amount of the order 10^- kw 

 becomes available to change the differential heating of the atmos- 

 phere. Since the temperature differences are small, the efficiency 

 is low and, at most, a quantity of the order 1% of this effect can 



