ATLANT. DEEP-SEA EXPlil). 1910. VDI.. ll 



IIYSICAI, OCEANOGRAPHY AND METEOROLOGY 



47 



currents. In tlie other cases included in the above scheme 



the variation of temperature with time is unsettled as long 



S r 

 as the value of — is unknown, ll <■ is constant with 



5z 

 regard to depth we find that the temperature rises in the 

 cases III and V, while in the cases 1 and Vli it sinks. 

 If ) decreases with depth the temperature rises in case 

 II and sinks in case VI, while the opposite variation of 

 temperature takes place if r increases with depth. 



From what is said above it is evident that such 

 variations of temperature which are due to a virtual con- 

 duction of heat vertically, depend upon the energy of 

 turbulence and, therefore, on the velocity of the horizontal 

 currents and its variations with depth. As the currents 

 bring various amounts of heat with them, the variations 

 of temperature owing to vertical conduction are interwoven 

 with the variations caused by the horizontal flow of heat. 

 In some cases a lifting or sinking of large water-masses 

 complicate the conditions. 



32. Seasonal Variations. 



Synoptic charts like those now constructed every-day 

 for meteorological purposes can not be obtained from the 

 ocean. When we disregard short-period oscillations (chap- 

 ter V), a small area may, however, be examined by one 

 ship within so short a time that other temporal variations 

 may be neglected. An examination of larger regions as, 

 for instance, the eastern North Atlantic, within a short 

 space of time, would require a whole fleet of ships in 

 cooperation. If we wish to construct temperature charts 

 for the ocean we are at present obliged to utilize obser- 

 vations from many single expeditions which have worked 

 in different seasons and years. A discussion of seasonal 

 and annual variations is important in order to find means 

 for reducing all observations to a common epoch. 



The temporal variations of temperature may be re- 

 garded from two different points of view. We may either 

 investigate the, so to say individual, variations within 

 characteristic water masses, accounting for possible local 

 displacements. In this case we may speak of investigations 

 according to "oceanographic co-ordinates". Or, we may 

 study the changes according to ordinary geographic co- 

 ordinates. In the latter case we get a combination of 

 thermal variations in the individual water masses and 

 changes which are due to displacements of the currents. 



Seasonal variations in the distribution of currents 

 are known from different localities. The conditions in 

 the Norwegian Sea present a characteristic example. 

 The coastal water which moves along the Norwegian 

 coast, between this and the Atlantic current in the Nor- 



wegian Sea, exhibit marked seasonal changes in tempera- 

 ture and salinity so that the interior field of force varies 

 with the seasons. The result is that the coastal water in 

 summer spreads seawards and pushes the eastern limit of 

 the Atlantic current towards the west while the opposite 

 movement takes place in winter. These lateral variations 

 are combined with simultaneous changes in the depth of 

 the coastal water. Analogous variations seem to exist 

 with regard to the Gulf Stream off the coast of the U. S. A. 

 and the Labrador Current and its continuation southwards. 

 Apart from such local displacements which are caused by 

 variations in the internal field of force, considerable dis- 

 placements caused by seasonal variations in the wind con- 

 ditions may occur. In this connection it may be mentioned 

 that Professor Nansen and I [1917, 1920] when investi- 

 gating the annual variations of the surface temperature in 

 the North Atlantic found that differences of temperature 

 from one year to another were closely related to variations 

 in atmospheric pressure and, consequently, in wind. This 

 circumstance is chiefly due to the transport of the surface 

 water by wind in places where the horizontal temperature 

 gradient is fairly large. Similar seasonal changes and 

 variations without any definite periodicity undoubtedly 

 appear below the surface too. 



Professor G. Schott and others have made a statistical 

 research into the annual range of surface temperatures 

 of the ocean. In his "Geographie des Atlantischen Ozeans" 

 Schott [1926] has published a chart showing the geogra- 

 phical variations of the annual range of surface tempera- 

 tures in the Atlantic. The chart is here reproduced in 

 Fig. 9. It appears that there is an absolute minimum of 

 the annual range in the tropics. A secondary minimum, 

 with differences between summer and winter temperature 

 less than 5° C, occurs in the ocean south east of Green- 

 land. In the Azoric high pressure area there is a secondary 

 maximum, the annual range exceeding 8° C. Within a 

 broad belt, extending from Europe to America the tempera- 

 ture variation between summer and winter ranges from 

 5° to a little above 8° C. Off the east coast of the U. S .A. 

 and in the vicinity of the Newfoundland Banks there is 

 an absolute maximum of the annual range of temperature. 



The conditions represented by the chart are easily 

 explained. In the tropics the annual range is small, chiefly 

 because of the position of the sun, with great noon alti- 

 tude throughout the whole year and a relatively great 

 cloudiness, the radiation being subject to small seasonal 

 variations. In the Azoric high pressure area the seasonal 

 changes in the altitude of the sun have a considerably 

 more pronounced effect upon the alworption of heat, and 

 because of the prevailing clear sky the radiation from the 

 sea to the atmosphere in winter is fairly strong. Prevailing 

 cloudy conditions and a deep-reaching vertical convection 



