148 The Three-dimensional Temperature Distribution and its Variation in Time 



of the Gulf Stream extends down to about 3000 m and that of the Mediterranean 

 water down to 2000 m, so that there is always a considerable heat surplus even at 

 these great depths in the North Atlantic. Below 3000 m the cold Antarctic bottom 

 water first appears and at deeper levels spreads northward with slowly increasing 

 temperature. There are insufficient systematic data available for the Indian and 

 Pacific Oceans below 2000 m to allow any reasonably accurate description of the 

 horizontal temperature distribution in the deeper layers. 



The importance of horizontal charts of the distribution of temperature and other 

 oceanographic factors, as a geographic aid to the comprehension of the distribution 

 of these factors throughout the ocean, has in the past been somewhat overestimated. 

 Oceanic processes never, or very rarely, occur along horizontal planes or are quasi- 

 horizontally arranged. Because the three-dimensional field of oceanic elements is 

 arbitrarily intersected by horizontal planes, connected phenomena will therefore be 

 cut by such planes. They are thus, for example, quite insufficient for following water 

 movements in the depths of the oceans. The same is equally true for the study of 

 atmospheric phenomena. Before these were deduced in other ways it was difficult to 

 interpret the arrangement of the isotherms in horizontal sections. In all cases vertical 

 cross-sections must also be used to clarify the three-dimensional field of any oceano- 

 graphic element. 



Vertical temperature sections can be taken in any direction and thus can give a far 

 better idea of the thermal stratification of a water mass than a horizontal chart. It 

 is, of course, best and most convenient to take the vertical section either along the 

 axis of major spreading of the water mass in the ocean concerned or across it. 



At the present time there are several such longitudinal or transverse vertical sections 

 (relative to the direction of flow) for all three oceans, showing temperature, salinity 

 and in part also the oxygen content. Those for the Indian Ocean (Moller, 1929; 

 Clowes and Deacon, 1935) and for the Pacific Ocean (Wust, 1929; Sverdrup, 1942, 

 1945; ScHOTT, 1942) are less accurate because of the smaller number of stations than 

 those for the Atlantic Ocean (WiJST, Defant, 1936). It is neither possible nor appro- 

 priate to describe and interpret these vertical sections individually. An interpretation 

 can only suitably be given in conjunction with the phenomena of the oceanic circula- 

 tion in the deeper layers. Figure 62 shows, as an example, a longitudinal section along 

 the western side of the Atlantic giving temperatures and salinities (after WiJST, 1928). 

 This runs from 75° S. near the area of formation of the Antarctic bottom water, 

 through the Weddell Sea and the South Antilles Sea, along the western side of the 

 West Atlantic Trough to the Newfoundland Banks through the Labrador Basin to 

 the Davis Ridge. There is a vertical distortion of the section by a factor 1 : 1300. This 

 section is quite typical of all sections through the Atlantic Ocean and shows the im- 

 portant characteristics of the meridional vertical temperature distribution: the two 

 large warm-water accumulations in the subtropical troposphere of both hemispheres, 

 the approach of the cold-water mass in the equatorial subtroposphere towards the 

 surface, the concentration of the isotherms at the polar limits of the troposphere 

 between 40° and 50° S. and 45°-55° N., and the oceanic polar fronts. This western 

 section also shows at about 1000 m an intrusion of colder water from 55° S. towards 

 the north as a tongue-shaped bulge on the isotherms which is visible even across the 

 equator. In a central section this is only weakly developed, in an eastern section it is 



