508 



Ocean Currents in a Non-homogeneous Ocean 



Table 137. Extreme positions of the isobaric surfaces at the ''Meteor'" St. 

 197 (8-7° S., 16-6° W.);for comparison ''Meteor" St. 198 

 (9-0° S., 19-8° W.) (reference-level at 3000 decibars) 



sea level showed displacements of about 10 dynamic cm and even at 1000 m the varia- 

 tion was still about 3-2 dynamic cm. Similar calculations have been made by Seiwell 

 (1932) for an "Atlantis" station to the north-west of Bermuda. These show a displace- 

 ment not only of the absolute position of the pressure surfaces but also the horizontal 

 pressure gradient is influenced. Comparison with the neighbouring "Meteor" St. 

 198 shows the magnitude of such variations in the pressure gradient due to the 

 passage of internal waves, unless these are only simulated. It is therefore not surprising 

 that the dynamic topographies remote from strong currents can be very confused and 

 only can be looked upon with utmost caution and criticism. 



8. The Determination of Water Transport in Density Currents 



The methods of topographical cartography of the isobaric surfaces allows an 

 insight, with the limitations discussed in the previous section, into the structure 

 of a density current, and when the pure drift and gradient currents are added, 

 a complete picture is obtained of the Ekman "elementar" current at any particular 

 place. In many cases it will scarcely be possible to give accurate details about each of 

 the constituents of the "elementar" current and the complete current structure will be 

 so complicated that it can only be shown graphically or by means of numerical 

 tables. Even a less detailed knowledge of the current conditions would be of con- 

 siderable value. Ekman (1929), by calculation of the current transport of a convection 

 current, first showed the possibility of obtaining information on the total mass of 

 water carried by a current in a relatively simple way without investigating the individual 

 layers. He was thereby able to show that this method of calculating the mass transport 

 is entirely independent of any arbitrary assumptions about friction, which seems to be 



