In addition, return currents in bottom waters may be caused by wind. To 

 illustrate this, we present a simulation of Baines and Knapp's (1965) 

 laboratory measurements on wind-driven currents in an open channel with two 

 solid boundaries along the wind direction. Using the same u*D/v (--51,000) of 

 their experiment, the model results in the middle of the channel are shown in 

 Figure 5.35. The computed horizontal velocities shown in Figure 5.35(a) agree 

 very well with the data and clearly exhibit the return currents caused by the 

 wind-induced set-up. Figure 5.35(b) shows the vertical turbulent velocity w' 

 and the vertical turbulent eddy viscosity A^. w'/u* agrees quite well with 

 Baines and Knapps' data. Although the vertical profile of eddy viscosity was 

 not measured, the simulated peak value of A^ is in agreement with their peak 

 value estimated from their measured turbulence profile and a length scale 

 equivalent to 20% of the water depth. 



Another situation in which return currents may exist corresponds to the 

 formation and deeping of a thermocline. Two example calculations are 

 presented in Figures 5.36 and 5.37. Both simulations start with a uniform 

 temperature of 6°C everywhere in a two-dimensional basin 20 m deep and 100 km 



2 



wide. In the first case, a uniform wind stress of 0.2 dyne/cm is applied 



2 



simultaneously with a surface heating rate of 0.01 cal/cm /sec. The 



temperature and velocity distribution in the middle of the basin at the end of 



96 hours are shown in Figure 5.36(a) and 5.36(b). The temperature 



distribution clearly indicates the existence of a thermocline. Velocities on 



the order of 50 cm/sec exist in the mixed layer while return currents on the 



order of 15 cm/sec exist below the thermocline. The second case was computed 



2 2 



with a wind stress of 0.5 dyne/cm and a heating rate of 0.01 cal/cm /sec. As 



shown in Figure 5.37(a), a thermocline exists very near the bottom. Figure 



5.37(a) shows mixed layer currents on the order of 15 cm/sec, but stronger 



return currents up to 25 cm/sec near the bottom. 



The mixed layer calculations presented above correspond to extended 

 periods of rather weak wind forcing and high heating rate. The presence of a 

 stronger wind and/or tidal mixing would cause stronger turbulent mixing which 

 may easily disrupt the sharp temperature and velocity profiles. During our 

 study periods of the Mississippi Sound, since winds are stronger and heating 

 rate are much lower, any pronounced vertical stratification is not expected to 



120 



