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



off is important, changes in salinity and density are too rapid to give correct forecast 

 values. 



The forecast of ice growth based on the ice potential can only be used during the 

 period when ice thickness is increasing. No theory has been given which accounts 

 for the decreasing ice thickness during the break-up period. A mathematical theory 

 for this period is still needed. 



The warm intermediate layer to be found at 250-750 m depth over the whole of 

 the North Polar Basin is of an advective nature. Its thickness depends on this, i.e. 

 on the strength of the oceanic circulation. It is thus not surprising that it shows strong 

 aperiodic variations especially in its upper boundary against the cold intermediate 

 layer. These variations may be as wide as 50-100 m and it is known that at the be- 

 ginning of this century this upper limit was at a depth of 1 50-200 m in the northern 

 Barents Sea and in the North Polar Sea. Since then it has risen to a depth of 75-100 m 

 due to the general climatic warming up of the Arctic, and in recent times the oceanic 

 circulation has undoubtedly increased in strength (WiJST, 1942; Weickmann, 1942). 

 There has been a strong increase both in the amount of ice transported into the Green- 

 land Sea from the central Arctic and in the transport of warmer, more saline Atlantic 

 water directed into the Arctic basin. Thus, since the "Fram" Expedition 1893-6, the 

 temperature of the warm intermediate layer of Atlantic water has risen noticeably, 

 as has been clearly shown by the observations of the "Sedow" Expedition 1937-40. 

 It is not impossible that a systematic study of these phenomena might show a close 

 correlation between the aperiodic variations of the boundary between the cold and 

 the warmer intermediate layers in the North Polar Sea and the variations in the 

 strength of the Atlantic oceanic circulation. 



5. Temperature Distribution in Horizontal and Vertical Sections 



The temperatures found at an oceanic station show the vertical temperature distri- 

 bution at that point, but only horizontal or vertical sections will give a two-dimen- 

 sional picture and thereby lead a step further towards a spatial conception of the 

 temperature distribution in the sea. A chart of the temperature distribution in the 

 Atlantic was first given by Maury in 1852 as a supplement to his charts of the winds 

 and currents in the Atlantic. The reliability of such horizontal temperature charts — 

 just as for vertical sections along fixed lines — depends on the amount of data available 

 and on its more or less uniform distribution over the entire section. The isotherms are 

 interpolated linearly between values given by observations, although it is known that 

 the linear interpolation does not always correspond to reality. However, the other- 

 wise sparse data leave too much freedom to the imagination of the analyst and the 

 resultant chart may soon be further from actual conditions than is tolerable. 



At the present time there are several recent temperature charts available covering 

 the entire ocean surface. The most comprehensive presentation of surface temperature 

 conditions in the Atlantic has been given by Bohnecke (1936) in the "'Meteor'" Report. 

 The same report also gives isothermal charts for different main levels in the Atlantic ; 

 WiJST (1936). A selection of surface charts and such for individual depths has been 

 given by Schott (1935, 1942) for the Atlantic, the Indian and the Pacific Oceans. 

 Recent surface charts have been published by the National Hydrographic Office 

 in Washington (1948), World Atlas of Sea Surface Temperatures. 



