The effects of adjusting the dynamic topography to the geostrophic transport by 

 methods (i) and (iii) can be seen by comparing figures 42 to 47 (drag coefficient of 0.0026 

 and monthly mean pressures) with figures 49 to 54 (drag coefficient 0.0012 and twice-daily 

 mean pressures). These figures show the same general pattern and similar slopes with 

 discrepancies of only 1 to 2 dynamic centimeters in a few locations. 



(b) The calculations for wind-driven transport for each month are based on the stress- 

 distribution for that month's mean pressure distribution, without consideration of initial 

 conditions or the inertial response characteristics of the ocean. For the assumption to 

 hold that acceleration terms are negligible, the ocean system must be essentially in equi- 

 librium with wind-imposed forces during each month considered. Intuitively, a response 

 time shorter than one month would seem realistic for the barotropic mode, but intuition 

 as well as analysis of observations precludes such short response times for the baroclinic 

 mode. Veronis and Stommel (1956), in their treatment of variable wind stress on a two- 

 layered ocean without lateral boundaries, showed that for periods of the order of one month, 

 acceleration terms were unimportant in both the barotropic and baroclinic modes. An 

 explanation for this is that the barotropic mode has a response time much shorter than one 

 month and so it is at all times essentially in equilibrium with the driving force, while the 

 baroclinic mode primarily responds to forces with periods of the order of one year or more. 

 Presumably fluctuations of wind stress for periods shorter than one month will give in- 

 significant accelerations. 



In addition to wind stress, bottom and lateral friction must also be considered. Proud- 

 man (1953) calculated bottom stress using the square law 



Tft = CbPhU-l 



in which the drag coefficient, c b , is taken as 0.0025, p b , the density of water at the bottom 

 is roughly 1 gm/cm 3 , and u b , the velocity in cm/sec, is measured one meter above the bot- 

 tom. In the present investigation, deep velocities on the order of 1 cm/sec were found. 

 Under these conditions, bottom stresses would be of the order of 0.002 dynes/cm 2 and 

 negligible compared to surface stresses. 



Munk (1950) developed expressions for wind-driven transport in a central region of 

 the ocean in which horizontal friction terms were neglected. These expressions result 

 in a distribution of transport conforming to the distribution of properties, and thus support 

 the assumption that horizontal friction is negligible in central oceanic regions. 



(c) The assumption that divergence due to the effect of bottom topography is neg- 

 ligible in the area of study is examined here. Sverdrup (1947), Munk (1950). and Fofonoff 

 (1962) all specified that the transport was made up of only baroclinic and Ekman modes; 

 therefore, the treatment of the bottom convergence problem was not pertinent. If equa- 

 tions (22), (23), and (24) from appendix I are added and the term for bottom effects retained: 



n, 1/ , tr , ir * d T *V dT sx , I dZb . ()Zb\ 



^+^+v b )=-^-—- P 'f(u b -+v b -y 



The term for bottom effects can be expressed as 



p'/(r^|cos 0|Vz 6 |), 



where 9 is the angle between the bottom velocity and gradient vectors. Substituting and 

 rearranging, 



9t su djsA 

 dx by J p '/(|i^|cos 0| S7 z b \) 



p -v g -v E — -. 



