Section 88. Temperature of Sea Ice 



Many observers have studied sea-ice temperatures and their vertical distribution, but 

 Malmgren and Sverdrup were the first to conduct observations that encompassed all the seasons. 

 These observations were made from October 1922 through June 1924, during the expedition on the 



Maud.* 



From these and other observations, it follows that the temperature of the lower surface of 

 ice is very close to the freezing point of sea water, i.e. , it is approximately constant, while the 

 temperature of the ice surface is close to the air temperature. Since the air temperature fluctu- 

 ates during the day, the temperature of the upper layers also fluctuates during the day. Further- 

 more, the daily variation of ice temperature is caused by the daily variation of radiation. Thus, 

 very often, under clear night skies, the temperature of the ice surface can be several degrees 

 lower than the air temperature, due to the radiation which is intense at this time. During the day, 

 when solar radiation penetrates the ice and is partially absorbed by it, the temperature of the ice 

 rises slightly, independently of the air temperature. As a result of such a daily temperature var- 

 iation, the ice strength, which also depends on temperature, has a daily variation. The ice is 

 strongest at about sunrise and weakest at about sunset. This fact should be considered when using 

 ice crossings. 



Table 78 gives some extracts from Malmgren' s observations during the winter of 1923-1924. 



TABLE 78. SEA-ICE TEMPERATURE IN °C. ACCORDING TO MALMGREN'S OBSERVATIONS 

 DURING 1923 AND 1924 



As was to be expected, the diurnal temperature variations, the mean annual temperature and 

 the absolute annual minimum, which characterize the temperature amplitude In the given case, 

 decrease with depth. 



Table 79 and figure 80 show the mean monthly temperatures at different depths from the ice 

 surface, computed by Malmgren from daily observations during the winter of 1923-1924. The 

 average temperatures of October and November, which were obtained from the averages of the 

 corresponding months in 1922 and 1923 , are exceptions . 



In figure 80, the average monthly temperatures are plotted along the Y-axis and the corre- 

 sponding months along the X-axis. If we exclude the January anomaly, which was due to the air 

 temperature anomaly, the curves of the average monthly temperatures are quite symmetrical. 



*The observations were made with resistance thermometers and thermocouples frozen into 

 the ice. 



232 



