Linear Theories — Viscous 



99 



and, as we shall see in the next chapter, there is therefore much reason to 

 formulate the boundary layer by using inertial terms rather than friction. 

 For the moment, however, we follow Munk's development. 



The total transport of the western current, ^^^ is independent of A, 

 and, using the value of X^ given in table 4, is simply 



^wc=-l-17ryff~^curl,T. 



(36) 



Table 5 contains a comparison of the computed transports of various 

 western currents: those computed by Munk (1950, table 2) from the wind 

 stress, and those computed from oceanographic observations. 



TABLE 5 



The Mass Transport of Some Western Cctrrents, Determined from 

 THE Wind Stress and from Oceanographic Observations 



Source: Munk (1950. table 2). 



The effect of the meridional wind component, which supports the zonal 

 winds, is introduced by a meridional wind factor 



curLr 



m" 



(37) 



The computed transports of the Gulf Stream and Kuroshio are too small 

 by about a factor of two. For example, Munk's computed value of the trans- 

 port of the Gulf Stream, using /?=l-9 x 10~^/cm./sec., r = 6500km., and 

 dTjdy = 10 X 10~^" g./sec./sec, is 36 x 10^^ g./sec, as compared to an actual 

 value of about 74 x 10^^ g./sec, if the reference level be taken as the depth 

 of 2000 m. 



The source of this discrepancy is not to be sought in the physics of the 

 western current itself. The results shown in table 5 apply equally weU to 

 the nonlinear theories set forth in Chapter VIII. Munk (1950, p. 92) 

 ascribes the discrepancy to an underestimate of wind stress for low wind 

 speeds: 



Finally we may examine the question of why the computed 

 transports of the Gulf Stream and the Kuroshio current amount to 



7-2 



