TEMPERATURE OF THE SURFACE WATERS OF THE ATLANTIC OCEAN 



A further attempt to classify the whole ocean has 

 been made by G. Wiist.^^ In projecting upon the 

 surface of the sea the general forms of the sea bottom, 

 that is, the limits given by the undersea topography, 

 he uses purely geographical designations, such as 

 Angola Sea, Newfoundland Sea, etc. Without ques- 

 tioning the value of this nomenclature, it must be said 

 that it does not suffice for the treatment of many ques- 

 tions, especially that of the classification of the surface 

 of the ocean. The surface of the water is so intimately 

 related to the climatic phenomena that the latter must 

 not be neglected in the classification of the ocean. In a 

 manner similar to that in which the classification of the 

 solid surface of the earth according to geography and 

 geology takes into account the atmosphere above the 

 earth, a classification of the ocean should be based on a 

 principle that is adapted to the phenomena on the 

 boundary stratum between air and water. It is 

 obvious that these boundaries cannot be established in 

 the sense of Wiist's claim. But what climatic bound- 

 aries (for example, the snow or tree boundaries) are 

 established or form sharply defined lines? An attempt 

 will be made to classify the ocean into natural regions 

 on the basis of temperature, because there is more 

 information available on this phenomenom than on 

 other phenomena. Besides the salinity and other fac- 

 tors, the active processes of the water and the air 

 must also be considered. How much importance 

 Koppen attributed to them is clear from the fact that he 

 devised for the surface of the water a special system 

 of regions based upon the winds. 



An attempt was made first of all to determine the 

 zonal boundaries exactly. The method consisted in 

 establishing along several lines running mostly in a 

 north-south direction the gradients of the surface 

 temperatures. In order that the procedure might be pos- 

 sible and that the original values from the one-degree 



Table 2.- 



-Temperature gradients for lalitude differences of 1° 

 at 65° W. long. (Drake Passage) 



SOUTHERN LATITUDE 



Figures in boldface indicate maxima. 



Figures in parentheses indicate that the maximum is not 

 well established. 



field charts be used, the average values of the tempera- 

 ture for the 2 one-degree fields adjacent to these lines 

 were recorded in a system of coordinates with the 

 abscissa "Geographical Latitude" and the ordinate 

 "Temperature." To eliminate accidental errors con- 

 nectetl with these oljservations, the points of intersec- 

 tion of the isotherms taken from the monthly charts 

 were likewise recorded with these lines, and a curve 

 plotted through these points. On this curve, which 

 showed the average temperature along such a line, the 

 change of temperature was then read oft' from one 

 degree of latitude to another and entered in tables. 

 Table 2 gives as an example the temperature gradients 

 for such a profile through the Drake Passage. 



a. South Atlantic Ocean 



This procedure was followed for five surface profiles 

 of the South Atlantic Ocean: 1. through the Drake 

 Passage along 65° W; 2. through the longitudinal axis 

 of the Falkland Current, beginning in the north at 

 35° S. 53° W. through 47° S. 55° W.^and 54° S. 48° W. 

 to 60° S. 47° W; 3. at 30° W. from .30° S. to 55° S; 4. 

 at 5° W. from 30° S. to 55° S; and 5. through the Agul- 

 has Current at 21° E. from 35° S. to 55° S. Because 

 of lack of observations, the position of the fronts in the 

 higher south latitudes could not be established on 

 profiles 4 and 5 for every month. 



As table 2 shows, the gradients along 65° W., for 

 exantple, increase to the south, except for some which 

 are influenced by the land at 55° S. From 58° S. to 

 61° S., they increase from 0.5° C. to between 1.0° and 

 2.1° C. Farther south, they drop again to about 0.5° C. 

 The other profiles show similar differences in the gra- 

 dients. This increase of the gradients is brought about 

 by the meeting of water masses of varying temperature 

 and denotes the presence of the polar front. It can be 

 seen that within certain limits the zone of maximal 

 gradients fluctuates fiu-ther. This further fluctuation is 

 due principally, according to the general conditions, to 

 the irregidar distribution of the few observations. In 

 order to eluninate these irregularities, the gradients were 

 recorded in a system of coordinates, and an equalized 

 curve plotted through the points. Then the average 

 position of the front was read from these curves, as is 

 shown in tables 3 and 4. 



The temperature prevailing at the momentary posi- 

 tion of the front in the month in question was then 

 established from the charts and transferred to the tables. 

 Since these zones of maximal gradients are subject to 

 certain fluctuations, as shown by table 2, and since the 

 surface unit for the temperature value is the one-degree 

 field, the position of the fronts was indicated only in 

 whole degrees of latitude ^^ and the temperature de- 

 noting the kind of water in half degrees centigrade. 



23 G. Wiist. Vie Qliederurtg des Weltmeeres. Ver»uch einer systematiscfien geo- 

 graphischen Namengebung. Petermanns Geogr. Mitteilungen. Jabrg. 82. Gotha, 

 1S36. 



" It remains uncertain whether there is any justification whatever for giving such 

 boundaries with any greater c^tactitude in researches covering such huge spaces as. 

 fore.xamplc,Deacon(1932)does. Hedesignates the position of the front at 10 minutes 

 of latitude. 



