The Tropospheric Circulation 631 



The theory of the Gulf Stream and similar boundary currents requires further 

 development. The double-layered model must be replaced by one with continuously 

 stratified water and the effects of friction in both vertical and horizontal directions 

 must be taken into account. Lateral friction against the coast should give a reduction 

 in the velocity of the current at the left-hand side as is shown by observations. 



The boundary current theory attributes the ocean boundary currents of the general 

 oceanic circulation, in so far as they have the character of a free jet, to the effects of 

 pressure and inertia and to the variation of the Coriolis parameter with latitude. It 

 has been pointed out above (p. 580) that the Sverdrup solution starting from an 

 eastern continental boundary and working westwards is unable to satisfy the boundary 

 conditions at the west coast of the ocean. Only by including the effects of a strong 

 lateral friction (mixing) Stommel and Munk have been able to satisfy the boundary 

 conditions at a western boundary and to give a general theory of a wind-driven 

 ocean circulation. However, along the eastern side of a continent (western side of 

 oceans) the currents apparently do not correspond to this theory. They are narrower 

 and more intense than would be expected from the general theory. The Charney 

 theory gives the explanation for this and yields in this way a western continuation to 

 the Sverdrup solution, without the addition of strong frictional effects but taking 

 into account the effects of inertial terms and the variation of the Coriolis parameter 

 with latitude. The density stratification of the water and the lateral inflow into a 

 meridionally directed jet current have been found to be of particular importance in 

 the formation of these boundary currents. These provide the connection with the 

 western transport of the zonal wind currents of lower latitudes. 



{d) Further Aspects of the Dynamics of the Gulf Stream 



Associated with the questions raised on p. 617 another one stands out concerning 

 the total current energy in a relatively thin top layer. This energy concentration in a 

 narrow current band occurring in the very upper layers persists for more than 2000 km, 

 from Cape Hatteras to the region east of the Newfoundland Banks while beneath this 

 top layer the velocities remain small. This remarkable phenomenon is probably 

 explicable by an association between momentum losses in the lower portion of the 

 current and the upper energy concentration. It should be stressed that the zonal width 

 and the high speed of the upper Gulf Stream layers rather definitely exclude an inter- 

 pretation of the current in this part of the Atlantic as the result of momentum added 

 locally by the prevailing winds. Rossby (1951) has attempted to find out what kind of 

 verticaly velocity profile would be formed in an immiscible stratified current subject 

 to momentum losses through contact with the underlying surface or at lateral boun- 

 daries. It would be of particular value to know the nature of the special velocity 

 profile corresponding to a minimum value of the momentum transfer in unit time 

 across a vertical plane normal to the current axis. It is reasonable to assume that 

 this profile represents a limiting state which would be gradually approached by any 

 stratified current subject to momentum losses but unable to escape to the sides. 



In a straight aparallel current of this type in which the water is considered to be 

 incompressible and the density varies with depth, the momentum transfer across a 

 vertical strip normal to the current axis is given by 



