Ocean Currents in a Non-homogeneous Ocean 497 



is marked with a vertical double arrow. The reference-level for conversion of relative 

 into absolute topography must lie within this layer. Already these station pairs show 

 roughly the meridional distribution of the depth of the reference-level in the Atlantic : 

 lower depth in high latitudes (approx. 1500 m or deeper), rising up to 500 m at the 

 equator. The topography of the reference-level can thus be derived for the whole of 

 the Atlantic from a large number of such diagrams. Figure 228 presents the topography 

 determined by this method. The lines are drawn at 100 m intervals (or decibars); 

 for a reduction of the relative into absolute pressure values it is sufficient to know the 

 position of the reference-level to the nearest 50 decibars. It is clearly shown that the 

 assumption of a reference-level of constant depth can never do justice to the dynamic 

 structure of the water masses of the Atlantic Ocean; even over smaller oceanic areas 

 there are appreciable variations in its position. Along each meridian the depth of the 

 reference-level is least near the equator (up to 400 m), in the Southern Hemisphere it 

 sinks uniformly to great depths in high latitudes. But in the Northern Hemisphere 

 conditions are more complex. From the equator it sinks at first to a secondary mini- 

 mum between 5° and 10° N. (about 900 m), then rises again to another maximum 

 between 10-20° N. and from there begins the lowering towards the north-west to 

 greater depths. The irregularity in the northern subtropics has the same form as the 

 asymmetry in the position of the subtropical and tropical thermocline (see Pt. I, 

 p. 120). There is undoubtedly a causal coimection between the two phenomena. In the 

 Gulf Stream region there are considerable deviations from normal. Near to the 

 current core (intense flow) the reference level rises steeply upwards to a depth of 

 1000 m or less. This phenomenon, which belongs to the characteristic features of this 

 area, must be connected causatively with the inclination of the isosteres in a stratified 

 ocean with intense motion (see p. 331). 



From the chart shown in Fig. 228, Neumann (1954, 1955) has computed zonal 

 averages of the depth of no meridional motion D (zero level) for the North Atlantic 

 and has plotted them against the latitude (Fig. 228 a). Individual values along the 

 20°W-meridian were used for the South Atlantic, since the variation in D in the east- 

 west direction is small as compared with the variation of D in a meridional direction. 

 In Fig. 228 a the values of D are marked by circles and the full drawn curves represent 

 the function 



D= - K?,mcl>^- Kcos &. (XV.16) 



The constant ^ is different in the Northern and Southern Hemisphere but the increase 

 of D with latitude follows this function closely except in the equatorial regions, where 

 apparently another physical law applies (see Pt. I, p. 120). 



Excluding the equatorial regions, the relative variation of D with latitude is given by 



15 'I = - '^"*- "^^•'" 



Then, it follows from the Coriolis parameter, f=2w cos d that 



1 df 



-f-^^-- tan^. (XV. 18) 



2K 



