TEMPERATURE OF THE STJRFACE WATERS OF THE ATLANTIC OCEAN 



illustrates the temperature gradients between the annual 

 zone averages for two degrees of latitude which were 

 taken from Appendix II. It is obvious that in the Soulli 

 Atlantic Ocean the steepest gradients were reached at 

 50° S., so that the polar front which separates the cold 

 antarctic water from warmer water of low latitudes and 

 which is commonly foimd here, makes itself distinctly 

 felt even in the zonal averages. In the north, the situa- 

 tion is not so simple, although the boundary line for 

 the Cold Wall at 40° N. stands out clearly. As a result 



SO*N 



Figure 3. — Temperature gradients in degrees C between the 

 annual zonal averages of 2° lat. between 70° N. and 70° S. 



of the manifold interweaving of arctic, subarctic, and 

 Gulf Stream water which disturbs the more or less 

 zonal, smoothly running boundaries of the water masses 

 observed in the south, the gradient rises and falls 

 irregularly. Therefore, it is at first impossible to draw 

 a clear zonal boundary line between arctic and warmer 

 water. The drawing of such a line must be postponed 

 to a later treatment of individual areas. 



The further course of the curve shows to a certain 

 extent additional boundaries which are designated by 

 temperature. The charts of the temperature distribu- 

 tion indicate decreases in temperature from certain 

 latitudes. Such transitions are shown in figure 3 at 

 40°N. and 40°.S., where the curve takes a sharp bend. 

 It may be assumed that these latitudes are the bound- 

 aries between the temperate and subtropical water 

 masses, insofar as they are expressed in the tempera- 

 ture. It is scarcely possible to establish the scope of 

 the tropical zone any more accurately from these gra- 

 dients toward the south, since the transition to the low 

 gradients is not clearly discernible here. In the 

 equalization of the curve of figure 3, its southern 

 boundary might be set at about 10°S. The northern 

 boundary can be more easily designated. It is at 

 20°N., from which point the curve moves more or less 

 horizontally toward the equator. 



It becomes clear that after this first survey we can 

 divide the ocean into certain "climatic zones" which are 

 conditioned by the distribution of the surface tempera- 

 tures. In the south, the polar front at 50°S. and the 

 southern boundary of the subtropical water at 40°S. 

 stand out clearly, wliile in the north only the boundary 

 between the Gulf Stream and the Cold Wall, likewise 

 at 40° N. can be recognized. The polar front in the 



north, on the other hand, is not so clearly expressed, as 

 a result of the partly meridional course of the isotherms 

 in this consideration of annual zonal averages. The 

 extent of the tropical zone can be placed between 10°S. 

 and 20°N. 



3. The Demarcation of Zonal Areas on the Basis of 

 Surface Temperatures 



The survey of the temperature distribution made in 

 section 2 of this report made possible a general division 

 of the ocean into climatic zones. An attempt will 

 now be made to establish the position of these fronts, 

 that is, of the boundaries of bodies of water of oceanic 

 climatic zones of the surface stratum, by means of 

 the surface material at present available.'^ This first 

 attempt should be regarded merely as preliminary 

 treatment. Because of the lack of data on the broad 

 expanses of the ocean, it will be impossible to classify 

 the water masses as Koppcn," for example, classified 

 the atmosphere. 



Not only material for classification, but also tho 

 principles of classification are lacking. Koppcn was 

 able to construct his climatic system, which was made 

 principally for the atmosphere over the land surfaces, 

 on the basis of the two factors of temperature and 

 humidity or precipitation "which obtrude themselves 

 most upon human b(-ings and can be most easily ex- 

 pressed in comparable figures." However, such a 

 perfect system for the water of the ocean does not exist. 

 Schott ^° has made a classification according to "natural 

 regions" in his well-known geographies of the oceans, 

 but these "natural regions" are, in his own words, 

 "sti'ongly subjective" and in addition suffer from the 

 lack of a uniform system of nomenclature. In fact 

 Schott used terminology from the most widely differing 

 systems; for example, Brazilian Region, geographic; 

 Equatorial Region, climatological ; Sargasso Sea, bio- 

 logical; Gulf Stream, oceanographic. In a work 

 devoted exclusively to classification by "natural 

 regions," Schott '^ adds further to the structure of his 

 system by giving special attention to the surface of the 

 ocean and the climatological relationships above it. 

 Even this procedure is not free from subjectivity, for it is 

 not based on definite, comparable values for the factors 

 used in its construction. Too, it is not quite con- 

 sistent in nomenclature. To name one example of 

 inconsistency — 'the region corresponding to the North 

 Atlantic Trade Region (Passat-Region) is not a South 

 Atlantic Region, but an Ascension Region. 



i» Cf. footnotes 10-16. 



" W. Koppen, Orundriss der Klimakuvde. 2. Aufl. Berlin, 1931.— See also 

 W. Koppen, Das geographiscfie System der Klimate. Handbuch d. Klimatologie. 

 Bd. I, Teil 2. Berlin, 1936.— Since Koppen, especially in the latter work, gives a very 

 detailed review of the other climatic systems, previously developed by other authors 

 (Penck, Hettner, Passarge, and others), they are not further discussed at this point. 



'0 O. Schott, Geographie des Atlant. Ozeans. 2. Aufl. Hamburg, 1926. 



— Ceograpkie des Indischen u. Pazifischen Ozeans. Hamburg, 1935. 



21 Schott, Die Aufteitung der drei Ozeane in naturliche Regionen. Petermaons geogr. 

 Mitteilungen. Jahrg. 82, Heft 6. Ootha, 1936. 



