356 BosTwicK H. Ketchum and D. Jean Keen 



3-4 X 10* ftVday, in the total river flow used in calculating the flushing times, would 

 decrease them from about 1-6 years to 1-3 years — a change which is not significant 

 considering the other approximations which have been made. 



One tacit assumption has been made in the calculation of the horizontal mixing 

 coefficients which has not been discussed. It is assumed that the current produced 

 by the escaping river water is the only advective process of importance. Other 

 currents, such as density and wind-induced currents, are also present. From the 

 continuity of volume, the net transport of these currents must be zero, but it does not 

 follow that the net transport of salt and fresh water must be zero. If, as seems likely, 

 there are one or more large scale, counter clockwise eddies over the shelf, with an 

 onshore current in the northern, low salinity, part of the region, and an offshore 

 current in the southern higher sahnity area, there will be a net offshore transport of 

 salt. This would result in an underestimate of the coefficients of horizontal eddy 

 diff"usion. Unfortunately, our knowledge of the currents dver the continental shelf, 

 except for the narrow coastal strip, is inadequate to evaluate this effect. 



A brief discussion of some of the forces which could produce the turbulence over 

 the shelf may be of some interest. The tidal transport, the winds, the currents due to 

 river flow and density structure, and the upwelling along the coast as a result of 

 off'shore winds, are the most obvious turbulent forces. In estuaries the mixing process 

 has been related to the tidal volumes moving over a mixing length determined by 

 the excursion of the tidal currents (Arons and Stommel, 1951; Ketchum, 1951). 

 The approximate contribution of tidal flow over the continental shelf to the overall 

 mixing process can be evaluated. For turbulent flow, the mixing effect would be 

 proportional to the product of the velocity and the transport distance (vj^). The 

 coefficient of proportionahty will vary, depending upon the actual degree of turbu- 

 lence resulting from the ffow, but the potential eff'ect will always be less than this 

 product. Both velocity and transport distance depend on the shape of the bottom 

 contour and distance from shore, and the product will be greatest where the ratio of 

 distance from shore to depth is a maximum (Fleming, 1938). This product has been 

 calculated for various depth contours for a semidiurnal tide with a mean amplitude 

 of 50 cm, and has the following values, which may be compared with the coefficients 

 of eddy diffusion hsted in Table V. 



30 fathoms 0-65 x 10« cmVsec. 



40 fathoms 1-08 x 10« cmVsec. 



100 fathoms 0-32 x 10« cmVsec. 



This product increases to a maximum value at the 40 fathom contour and decreases 

 beyond this depth, whereas the mixing coefficients decreased with increasing depth 

 throughout this range. The tidal mixing product is always less than the coefficients of 

 horizontal diffusion and the coefficient of proportionality is probably much less than 

 unity. It may be concluded that the tides alone can contribute only a small fraction 

 of the total turbulence, and that the other forces must contribute substantially at all 

 times of year. 



REFERENCES 



Arons, A. and Stommel, H. (1951), A mixing-length theory of tidal flushing. Trans. Amer. Geophys. 



f/wo«, 32(3), 419-421. 

 Bigelow, H. B. (1913), Oceanographic cruises of the U.S. Fisheries Schooner Grampus 1912-1913. 



Sci. N.S. 38(982), 599-601. 



