4. Summary and Conclusions 



A method has been derived for matching the geostrophic transport, calculated from 

 dynamic heights, with wind-driven transport for the same period. Geostrophic velocities 

 at 5,000 meters and above were determined covering a portion of the north central Pacific 

 Ocean for the months of September 1961, September 1962, and May 1963. These ve- 

 locities are reported as plan views of dynamic heights. The determined velocities were 

 generally zonal for the periods of observation. At the 5,000-meter level values on the 

 order of 1 cm/sec were found. Over most of the area at these times, the sense of the 

 velocity changed within the water column, so that the current direction at the 5,000-meter 

 level was opposite to that of the near-surface levels. 



The effects of varying the drag coefficient used in determining wind stress and of 

 alternate methods of determining the mean stress distribution for the one-month period 

 of consideration was investigated.' The use of a drag coefficient of 0.0012 and values for 

 stresses determined as the mean of stresses calculated from twice-daily reports of pressure 

 distribution seem to give more realistic transports than those obtained from other com- 

 binations of drag coefficients and pressures. 



The horizontal distribution of potential temperature and dissolved oxygen indicate 

 that the net flow at 4,000 meters at about 48°N must be easterly. Such a flow is in a direc- 

 tion contrary to the flow found at this location by the transport method. This has been 

 accounted for by examining the relative importance of the baroclinic and barotropic modes 

 at this level. The flow at 4,000 meters is mainly barotropic, and as the barotropic flow 

 must be periodic (Veronis and Stommel, 1956) it is concluded that the calculated deep 

 flow is a periodic barotropic flow. The calculated magnitude of deep velocities (about 

 1 cm/sec) is in good agreement with Veronis and Stommel's theoretical value. 



Vertical velocities are determined by graphically integrating — determined as a func- 

 tion of meridional velocity. Values as high as 10~ 4 cm/sec were calculated, but these 

 are at the level of significance of observational errors. Thus, the results shown may re- 

 flect the range of variation in vertical velocity, but details of the variation with depth cannot 

 be validated. 



For each period for which horizontal velocities were computed, plots were made of 

 the surfaces of no zonal or meridional motion. The determination of the depth of no mo- 

 tion can be very sensitive to observational errors and inaccuracies in the computed wind- 

 driven transport; consequently, the plots are subject to considerable error but do indicate 

 that the depth of no zonal or meridional motion varies widely with place and time. 



The assumption that terms due to the interaction of flow with the bottom can be ig- 

 nored was examined. The data suggest that bottom flow parallels isobaths and therefore, 

 the bottom and slope terms in the integrated momentum equations become very small. 

 However, determinations in greater detail over an area in which the bathymetry is well 

 defined would be needed to show conclusively that these terms are negligible. 



Because the velocities determined near the bottom appear to be periodic, comparison 

 with net velocities as inferred from the horizontal distribution of variables (Wooster and 

 Volkmann, 1960) or by carbon-14 dating or heat flow considerations (Knauss, 1962) is in- 

 appropriate. Perhaps the best corroboration of the deep flow determined in the present 

 study is in the tendency of the water to flow parallel to isobaths. 



It is concluded that the method employed here is a useful tool for the study of circula- 

 tion in central ocean areas. To best realize the possibilities of the method, oceanographic 

 data should be collected so that no additional limitations are placed on the results. A 



15 



