was used for hourly points over a 33-day period, with a standard devia- 

 tion of ± 13 feet (Fig 5). To check the validity of the prediction 

 scheme, equation (l) was applied to the hourly wind data points and the 

 results compared to the hourly measured thermocline depth averages 

 throughout the entire May-August period. The results were equally good 

 for the entire period; the ± 13 feet deviation calculated for the month 

 of August is probably typical of the accuracy that can be expected of 

 wind-based predictions. 



A similar check on the thermocline depths predicted as a function 

 of wind and tide did not support the validity of equation (2). Instead, 

 it was found that while the surface and internal tides are closely 

 coupled in freqi;ency and phase, they are only loosely coupled in 

 amplitude. Both schemes previously used to determine the ratio of 

 surface to internal tidal amplitudes have used long-term averages, 

 and thus have failed to account for variations in the ratio. Results 

 indicate, however, that the variations become quite large; these are as 

 yet unpredictable, but it is believed that the variations are at least 

 partially a function of the density gradient across the thermocline. 

 The determination of this effect is one purpose of a planned experimental 

 program; however, the depth prediction capabilities of equation (l) are 

 sufficient for most practical purposes. 



SUMMARY AND CONCLUSIONS 



This study has shown that the gross vertical migrations of the 

 thermocline along the coast are driven by the winds. Various modes of 

 layer motion were found to be: 



1. a long-term, upward trend throughout the summer, which occurs 

 in response to a similar trend in the wind, and may be related to the 

 annual heating cycle, 



2. an oscillation of 4 l/2 day period, not previously known to be 

 present in the thermocline motion, which is a direct result of a 4 1/2 

 day cycle in the wind, 



3. a diurnal oscillation that is also a wind effect, although 

 somewhat decreased by the diurnal tidal cycle, and 



4. a semidiurnal oscillation of tidal origin. 



131 



