Knowledge of permanent and semipermanent convergence and divergence 

 in the oceans is limited. Data on currents are scarce and uncertain; 

 therefore, computations of convergence and divergence rates and their 

 effects on the thermocline are not reliable. Monthly statistical compu- 

 tations of existing BT data for at least 5-degree squares would be helpful 

 for detection of permanent or semipermanent convergent or divergent areas 

 and would also be useful for predicting thermocline depth and other prop- 

 erties. Unfortunately, sufficient data for such computations are avail- 

 able only for very limited portions of the oceans. 



Mechanical Mixing 



With exception of areas of frequent advection as shown in Figure 7 > 

 the mixed layer in spring, summer, and early autumn is formed mainly by 

 mechanical mixing. This mechanical mixing is an indirect result of wind 

 stress on the water surface through intermediate action of wind waves and 

 pure wind current. 



Little mixing results from orbital motion of particles in regular 

 sinusoidal waves, because mass is not transferred from one level to another, 

 This conclusion is in good agreement with the general observation that 

 swell alone does not appreciably affect the mixed-layer thickness. Only 

 the interaction of several systems of orbital motions, resulting from 

 various origins, dimensions, and directions and with interfering patterns 

 of streamlines, can produce enough turbulence to overcome stability and 

 transfer momentum to adjacent levels. Mixing furnished by wave action 

 can have several causes: (l) interaction of particle motion because of 

 spectral distribution of periods and heights, (2) interaction of orbital 

 particle motion because of angular spreading of wave trains, (3) inter- 

 action of orbital motion between surface and internal waves (Figure 10), 

 and (k) breaking wind wave crests. Since the actual process of orbital 

 interaction is not known, all orbital motions may be combined and regarded 

 as orbital turbulence. 



Another important agent of mechanical mixing is drift or pure wind 

 current developing at the same time as wind waves. The velocity and 

 vertical extension of this current apparently depend on wind force, fetch, 

 and duration, as do wind waves. 



Flow in the ocean is generally considered to be nearly always turbu- 

 lent; consequently, stirring action is considerably more effective, so 

 that the rate of mixing depends on the rate of flow and the vertical 

 velocity gradient. However, it is doubtful whether a completely mixed 

 surface layer can be formed by an ocean current alone. If total mixing 

 could be caused by horizontal flow, the water would be mixed throughout 

 the vertical extent of permanent currents. 



The structure of upper layers in a permanent current is not essen- 

 tially different from the structure of these layers in areas with little 

 or no horizontal flow. There is a relatively shallow mixed layer and a 

 well-developed thermocline in a permanent current, such as the Gulf Stream, 



16 



