By 12 October the surface water had cooled to -1.32° C; the warm- 

 est water was at the bottom. The temperatiire maximum of -0.62° C was 

 observed at a depth of 33 meters. The siirface salinity had increased 

 to 32. 34 °/oo, and convective mixing had produced an isohaline layer 

 in the upper 15 meters. The calculated freezing point of the surface 

 is -1.76° C. 



The density gradient below the 15-meter level, although weak, has 

 significant relevance to the bubbling system. Theoretical ice-poten- 

 tial calculations using the data of the deeper stations show that, 

 prior to initial ice formation , thermohaline convection takes place 

 to a depth of approximately 15 meters. Consequently, the water below 

 this level temporarily serves as a source of sensible heat. However, 

 as ice forms, the salinity of the upper layer increases, resiilting in 

 greater density and an increase in depth of convective mixing. As the 

 density gradient weakens and eventually disappears, cooling to the 

 freezing point will occur throughout the water column. At the known 

 average rate of heat loss from the sea surface in the latitude of Thiile, 

 the entire supply of warm water will be eliminated within two weeks 

 after initial formation of ice. 



The heat content of air issuing from the compressors is consider- 

 able. An appreciable amount of heat is possibly introduced into the 

 bottom water adjacent to the pier when compressed air cools in the 

 polyethylene pipes; however, the data do not indicate a temperature 

 differential attributable to this source within the bubbler field. A 

 layer of dirt covering four steel feed pipes provides insulation; how- 

 ever, heat loss through the rubber feed hoses is great. Between the 

 point where these hoses connect to the steel pipes and the point where 

 they enter the water, melting of snow within a radius of approximately 

 2 feet was observed. Ice formation due to moisture condensate in the 

 underwater sections of the feed pipes was removed by alcohol injection. 



The effectiveness of the bubbler system, when warmer bottom water 

 is available, is manifested by the temperatixre data in Table I, 







TABLE 



I 







13 



October 



■ 1959 





STATION 



1 







STATION ID 



Depth 

 (meters) 

 0.0 

 3.5 

 6.5 

 9.5 



Temp. 



(°c) 



-1.09 



-1.08 



-l.o6 



-l,o6 







Depth Temp, 

 (meters) (°C) 



-1.76 



3 -1.60 

 6 -1.25 

 9 -1.24 



