The best agreement between the two variables is obtained when the wind 

 component from the WNW is used. 



The standard deviation is ± 13 feet when the axis parallel to the 

 coast is used for normalization. 



It should be mentioned that although equation (l) was used in the 

 direction search, the computed direction of best fit is independent of 

 the constants in the equation. That is, the parameters of the equation 

 have no effect on the choice of optimum direction since the equation 

 could simply be written as: 



Dw,t = -Wt-15 



without any change of the locations of the maxima and minima of the 

 curve as a function of direction. 



Tidal Relation to Isotherm Depth 



To further examine the data for short-term wind transport effects, 

 it was necessary to determine the extent of tidal influences on the 

 thermocline motion. To do this, equation (l) was applied to the 

 August diurnal wind cycle to generate a diurnal depth cycle predicted 

 as a function of wind only. The resulting curve was then compared with 

 the measured diurnal thermocline depth cycle averaged over the same 

 peri"od (Fig 6) . The pointwise difference between these two curves was 

 then computed and compared with that of the August diurnal tidal cycle. 

 Except for a small additive constant (which has been removed), the 

 pointwise differences between the measured and wind-predicted depth 

 cycles match the tidal cycle almost exactly (Fig 7). 



According to this fit, the internal tide is five times* as great in 

 amplitude as the surface tide and lags behind it by 3 hours. Using this 



*The factor of 4 appearing in equation (2) seems to disagree with this. 

 However, equation (2) predicts the depth from the surface, which is 

 itself oscillating with tidal amplitude. 



123 



