TEMPERATURE OF THE STJRFACE WATERS OF THE ATLANTIC OCEAN 



Table 3. — Average position of ihe southern polar front and 

 monthly averages of the water temperature at the front 



Parentheses indicate that the data have not been well established. 



Table 4. — Average position of the subtropical boundary and its 

 average monthly temperatures 



The tables show that in the treatment of the merid- 

 ional change of temperature, two zones of especially 

 steep temperature gradients of from 1° to 2° b^ecome 

 apparent. The most southerly of these zones lies where 

 Meinardus and Schott previously suspected a boundary 

 line within the west wind drift. It is most commonly 

 designated today as the Antarctic Convergence or Polar 

 Front. In the western part of the ocean, in the Drake 

 Passage, it is encountered at 60° S. At 50° W. it goes 

 north to 55° S.,^* and then at about 50° S. runs farther 

 eastward. Since in the east the data on temperature 

 are very scanty, they are to be regarded as merely 

 approximate values (table 3 and fig. 4). 



The temperatures met with in this area lie on the 

 average between 2° and 5°. Their origin is well 



>i On the current chart by H. H. F. Meyer, shown as Tafel XXXV in Q. Wiist, 

 Der Urspruna der Ailantmhen Tiefenwasser. Zeitschrift d. Oesellschaft fur 

 Erdkunde, Jubil.-Sonderhand Berlin, 1928, the western part of the polar front has, as 

 Deacon points out, obviously been confused with the subtropical convergence lying 

 farther to the north. 



explained by the previously cited works. According 

 to Sverdrup, the convergence is caused by an inter- 

 action of wind and thcrmohaline circulation. South 

 of the convergence, the water transport to the north, 

 which is caused by the wind, is obviously at work, 

 while the thermohaluie cu-culation conducts water to it 

 from the north. The thermohaline circulation in itself, 

 however, cannot be decisive, for the maximum of 

 density, which would be caused by it — at least in the 

 summer — lies considerably farther to the south than 

 the convergence. Plate XXIV dep'cts the distribution 

 of the surface density in the southern summer as cal- 

 culated from the average of the months December, 

 January, and February. It shows that although a 

 density ma.ximum is really present, it lies 4° to 5° of 

 latitude farther south than the polar front, as can be 

 deduced from the increases in temperature. 



Since the orientation of the details of the active 

 processes in the high southern latitudes are unknown, 

 direct comparisons with this zone of steeper temperature 

 gradients cannot be made. TJie only sources of infor- 

 mation are the vertical sections, such as were published 

 and discussed after the Discovery Expedition by Deacon 

 and Sverdrup, 1933.^' They show that the polar front 

 is to be found where the water masses sink down to the 

 subantarctic intermediate current and that this sinking 

 corresponds, even in individual cases, with its average 

 position, which was established ou the basis of tempera- 

 ture distribution. According to the definition intro- 

 duced by Defant, the polar front forms the common 

 northern boundary for polar and subpolar water. In 

 the interests of uniformity, this definition is being 

 used, although in reality subantarctic water can be 

 encountered, at least at the surface, even north of the 

 polar front. The concept of the oceanic polar front 

 with its effects upon the circulation below the surface 

 has, however, permeated the literature on oceanography 

 so thorouglily that it seems inadvisable for purely 

 practical reasons to suggest changes. Accordingly, 

 the polar front includes polar and subpolar water and 

 at the same time forms the pole-side boimdary of the 

 temperate zone. Even in the temperature sections of 

 the Meteor, this suilving at the front can be clearly 

 seen.^' Thus profile Va (fig. 5) through the Drake 

 Passage shows the position of the front between Sta- 

 tions 107 and 108, where the thermograph at 59° S. 

 records a temperature increase of 2° to 5.5°. Figure 6, 

 on which the hourly values from the registered tem- 

 perature curve are entered, clearly repeats this sudden 

 rise of the surface temperatm-es. The course of tlie 

 curve indicates by its fluctuations that individual warm 

 and colder bodies of water are separated in the vicinity 

 of the convergence as a result of vortex formations. 

 Even on profile Vd south of Africa (fig. 7) after a rise 



» Of. footnotes 11 and 13. 



* Q. Wust, Die Stratosphare des AUantischen Ozeant. 



Bd. VI. 



