The Equatorial Region of the Atlantic Ocean
The Water Masses of the Equatorial Region of the Atlantic Ocean. Between depths of 100–200 m and 600–700 m, the water masses of the equatorial region are mainly of South Atlantic origin, but in the northern parts there is evidence of admixture of large quantities of North Atlantic Central Water which, as will be seen later, at temperatures between 8° and 15°C has a salinity which is nearly 0.5 ‰ higher than that of the South Atlantic Central Water. Figure 170 contains T-S values at selected equatorial stations, the locations of which are given on the inset map, and also shows the characteristic T-S relationships of the central water masses. At the two stations close to the Equator, stations M246 and M259, nearly typical South Atlantic water occurred, but at station M289 in 11°N and 50°W the water contained a considerable admixture of North Atlantic water and the same was true of station M267 in lat. 14°N off the African coast, where North Atlantic water appeared conspicuously at temperatures below 10.5°. At the most northern station, M281 in lat. 19°N, nearly pure North Atlantic water was encountered.
Temperature-salinity relation in the equatorial region of the Atlantic Ocean. Inset map shows location of Meteor stations used, and inset diagram shows vertical distribution of temperature and salinity at Meteor station 246 in 2°40′S, 16°37′W, and at Meteor station 289 in 11°02′N, 49°33′W.
In the figure are entered lines joining the T-S points at the different stations, which represent conditions at depths of 500, 1000, 2000, and 3000 m, and the line σt = 27.0 is shown. At 500 m there is a wide range in the temperature and salinity of the water but the density is practically constant, as is evident from the fact that the 500-m line in the diagram nearly coincides with a σt curve. At 1000 m there is still a considerable spread of temperatures and salinities, but a constant density, whereas at 2000 and 3000 m the variations in temperatures and salinities are nearly within the limits of the accuracy of the observation. The uniformity of the deep water is brought out by the close spacing of the 2000- and 3000-m marks.
In the equatorial region, the water masses which have been dealt with so far are separated from the surface waters by a layer within which the density increases so rapidly with depth that it has the character of a discontinuity layer. The existence of this discontinuity layer is illustrated by the inset diagram in fig. 170, showing the vertical distribution of temperature and salinity at Meteor stations 246 and 289. At the latter station a distinct salinity maximum was present in the discontinuity layer, such as is the case within large areas (Defant, 1936b).
The discontinuity layer is explained mainly as a result of heating of the surface water due to the absorption of radiation from the sun and forced mixing due to wind. The water comes from higher latitudes and has originally a relatively low temperature. As the heating and mixing penetrates to greater and greater depths, the differences in density between the warm surface waters and the colder subsurface strata are concentrated in a thinner and thinner layer within which the stratification becomes more and more stable. When the stratification has become so stable that the layer takes the character of a discontinuity surface, further mixing is inhibited because the eddy conductivity is greatly reduced by the exceedingly great stability (p. 476) and the amounts of heat which are conducted downwards become so small that they are carried away by weak currents. Additional absorption of solar energy is used mainly for evaporation, and the final temperature which is attained will depend upon the rate of evaporation and is therefore determined by the interaction between the atmosphere and the ocean.
The Currents of the Equatorial Region of the Atlantic Ocean. In a given locality the thickness of the upper warm layer depends not only upon the intensity of mixing but also upon the character of the circulation in the top layers, because in the presence of currents the discontinuity surface cannot remain horizontal but must slope (p. 444). On the basis of the Meteor data and all other observations available from
(A) Topography of the discontinuity surface in the Equatorial region of the Atlantic, shown by depth contours in meters, and corresponding currents. (B) Salinity in the layer of maximum salinity below the discontinuity layer and assumed currents. Regions without salinity maximum shaded. (Both representations after Defant.)
Evidently the countercurrent is related to the distribution of mass as demonstrated by the slope of the discontinuity surface. In a profile from south to north the discontinuity surface rises towards the Equator, reaches a maximum elevation at the Equator, and drops towards a minimum in 2° to 3°N latitude, at the northern boundary of the South Equatorial Current. Between 2° to 3°N and 6° to 8°N the discontinuity surface rises towards the north, reaching a second and more pronounced maximum in the latter latitude. This is the region of the countercurrent, and to the north of the second maximum, where the surface slopes downwards again, the North Equatorial Current is found. This distribution of mass was first recognized by means of the Carnegie observations in the Pacific (fig. 198, p. 710). On the basis of these, Sverdrup (1932) suggested that the countercurrent appeared because the trade winds maintaining the equatorial currents are not symmetrical to the Equator, the calm belt between them lying in the Northern Hemisphere, whereas the effect of the rotation of the earth is symmetrical in respect to the Equator.
The dynamics of the countercurrent have recently been discussed by Montgomery (1940) and by Montgomery and Palmén (1940). Montgomery writes:
The trade winds, by continually exerting a westward stress on the sea surface, produce a westward ascent of sea level along the Equator. This slope amounts to about 4 cm per thousand km, and the accompanying pressure gradient extends down to about 150 m in the Atlantic Ocean. The Equatorial Counter Currents are found in the doldrums and apparently result simply as a downslope flow in this zone where the winds maintaining the slope are absent.
The ascent of sea level from east to west is due to a greater accumulation of light surface water along the coasts of South America. Figure 171A shows that such accumulation exists because the thickness of the homogeneous surface layer above the discontinuity surface increases from less than 40 m in the east to about 140 m in the west. Above a depth of 150 m the isosteric surfaces, therefore, slope downwards from east to west, but below 150 m they are practically horizontal. Accordingly, Montgomery and Palmén find that at the Equator the geopotential height of the free surface referred to the 1000-decibar surface is 14 dyn cm greater in the west than in the east, but the 150-decibar surface is parallel to the 1000-decibar surface. The countercurrent is therefore a very shallow current which is confined to the surface layer above the discontinuity.
A swift current which is embedded between water masses moving in the opposite direction must be subject to a considerable retardation owing to friction. A certain amount of energy is therefore needed for maintaining such a current, and this energy is, according to Montgomery and Palmén, derived from the trade winds which maintain the slope of the sea surface. Montgomery and Palmén assume that the retarding friction may be due to lateral mixing with the westward-flowing equatorial currents, and estimate the coefficient of lateral eddy viscosity to be 7 X 107 g/cm/sec. This concept leads to the important conclusion that on both sides of the countercurrent the equatorial currents are subjected to great lateral frictional stresses which are directed opposite to the horizontal stresses exerted by the trade winds on the surface. Similar lateral stresses must be directed opposite to the general flow along the continental boundaries of the two large counterclockwise gyrals of the two
Schematic representation of the vertical circulation within the equatorial region of the Atlantic. The direction of the currents is indicated by the letters W and E. The water below the discontinuity surface, which is supposed to be at rest, is shaded.
Within the countercurrent and the adjacent equatorial currents, the frictional forces must lead to a transport of water across the isolines, that is, to a transverse circulation. This was examined theoretically by Defant (1936b), who considered friction due to vertical mixing, but the conclusions are probably applicable if the friction is due to lateral mixing. Defant's results are shown schematically in fig. 172, according to which four “cells” are present, representing gyrals with horizontal axes, neighboring gyrals rotating in opposite directions. Within the southern cell the water sinks in the region of the Tropical Convergence and rises at the Equator, and within the next cell, located between the Equator and the southern boundary of the countercurrent, the water rises at the Equator and sinks at the boundary of the countercurrent. Within the countercurrent the water rises at the northern and sinks at the southern boundary, and within the northern cell sinking motion takes place at the Tropical Convergence and rising motion at the northern boundary of the countercurrent. The data from the Atlantic Ocean indicate the existence of these cells and the Carnegie section across the Pacific countercurrent (fig. 198, p. 710) demonstrates their presence convincingly. As a consequence of these transverse circulations, the northern boundary of the countercurrent and the Equator represent lines of divergence, whereas
In summer, when the countercurrent is best developed, the effect of the divergence at the Equator appears on the charts of surface temperatures (Böhnecke, 1936) as a tongue of low temperature, but no effect of the divergence at the northern boundary of the countercurrent is visible. The divergences also appear to be of biological importance because the ascending motion maintains a replenishment of nutrients to the surface layers, thus favoring the development of phytoplankton. The tongues of water of high phosphate content and of abundant phytoplankton which extend towards the west from the coast of Africa (see figs. 216 and 217, p. 786) are probably related to the structure of the currents.
In order to study the currents in and below the tropical discontinuity layer, Defant examined the character of the salinity maximum in the discontinuity layer, which is present between lat. 10°S and 20°N except within two narrow bands in about 3°S and 11°N. In fig. 171B the isohalines in the layer of salinity maximum are entered and the areas with and without salinity maxima in the discontinuity layer are indicated (after Defant). Within the discontinuity layer the vertical turbulence must, as already stated, be so much reduced that in this layer the water spread without being mixed with overlying or underlying masses, and the water can therefore move over long distances without losing its characteristics provided no horizontal mixing takes place. In this case the direction of flow must be approximately parallel to the isohalines, and from the chart showing these isohalines it is possible on these assumptions to derive the direction of flow in the discontinuity layer. The arrows in fig. 171B demonstrate the direction of these currents. It should be observed particularly that the belts without salinity maxima appear here as regions of convergence, whereas at the surface they are regions of divergence. This relation is in agreement with the conclusion as to the character of the vertical circulation (fig. 172).
Montgomery (1939) has approached the problem of the subsurface currents near the Equator in an entirely different manner, assuming that the subsurface flow takes place along σt surfaces and that besides the vertical mixing, lateral mixing along σt surfaces determines the distribution of salinity and temperature. He points out that the layer of maximum salinity approximately coincides with the surface σt = 25.5, and has prepared a chart showing that in this surface the isohalines have in general the same form as those in fig. 171B. He furthermore assumes that the principal water motion follows the tongues of low or high salinity, whereas Defant assumes that lines of flow are roughly parallel to the isohalines. The pictures of the current as derived by Defant and
Montgomery points out that the observed distribution of salinity can remain stationary in the presence of the assumed currents if either vertical or lateral mixing dominates. He has computed the possible maximum values of the coefficients and obtained for the vertical one a value which was much greater than is consistent with the great stability of the stratification. If the vertical diffusion is small, the lateral diffusion must determine the character of the salinity distribution. The lateral diffusion was found to be about 4 × 107 cm2/sec, in agreement with the value of the lateral eddy viscosity which was derived by an entirely different method (p. 485). At the present time it is not possible to decide which of the two interpretations is the correct one, but the dynamic consequences of Montgomery's concept appear to fit better into a large-scale picture of the ocean currents.
The chart of the surface currents (chart VII) shows that on an average a transport takes place from the Southern to the Northern Hemisphere in agreement with the conclusions which were based on the Meteor sections between lat. 20°S and 33°S, according to which an average transport to the north of about 8 million m3/sec should take place above a depth of approximately 800 m. It appears that the greater part of this transport, about 6 million m3/sec, takes place within the warm surface layers, for which reason it is probable that this process contributes towards displacing the thermal Equator towards the north. The fact that the surface temperature of the Atlantic Ocean is, in the Southern Hemisphere, lower, but in the Northern Hemisphere higher than that of the Pacific Ocean (see table 31, p. 127), is perhaps related to this transport.