accretion is slow, brine drains out of the salt cells and a smaller amount of air bubbles and silt is 

 retained in the ice. Consequently, the layers of ice which formed during higher air temperatures 

 are fresher, less porous, and more monolithic than the layers which formed during low air tem- 

 peratures. The stratification of the upper layers of the sea has a still greater effect on the strati- 

 fication of ice. As has been noted, each new layer of the sea drawn into the vertical circulation 

 creates a new ice layer; the thinner the layer of sea water, the thinner will be the layer of ice; the 

 more sharply delineated the layer of the sea, the more sharply delineated will be the layer of ice. 



2. It is easy to calculate that if the upper layers (pre-mixed and pre-cooled to the freezing 

 point) were 20. 5 m thick instead of 10 m, the ice would not grow sporadically (and consequently, 

 the ice would not be stratified), and the ice thickness could increase to 170 m with 4000 freezing 

 degree-days (counting from the moment of cooling to the freezing point). Thus, in this example, 

 the ice thickness proved to be 44 cm less than it would have been under other conditions. 



3. It is also easy to calculate that if, at the same temperature of 1. 5°, the second layer 

 were 30. 4 m thick instead of 15 m, after 958 freezing degree-days had been expended on forming 

 the first 83 cm of ice, the remaining 3042 freezing degree-days would be expended solely on cool- 

 ing the upper layer mixed to a depth of 40. 4 m to the freezing point, and there could be no addi- 

 tional ice formation. The same result would be obtained with the same thickness of the second 

 layer (15 m), if its initial temperature were 3. 0° instead of 1.5°. 



4. It is also easy to calculate that after the ice becomes 83 cm thick, and convective mixing 

 of the upper layer with the second layer begins, the common temperature of the two upper mixed 

 layers, i. e. , to a depth of 25 m, will increase to 0. 26°, due to the temperature of the second 

 layer, as a result of which the ice might even begin to melt somewhat from below. Of course, if 

 we assume there is no motion other than convective motion in the water under the ice, a film of 

 melted fresh water will immediately form near the lower surface of the ice, which will limit the 

 heat exchange between the water the the ice. When water moves beneath fast ice, or when ice 

 moves, frictional mixing will constantly destroy this protective crust, and thus the effect of the 

 contact of the ice and warm water will intensify. 



5. A new increase in the water temperature beneath the ice will begin after 3735 freezing 

 degree-days, and by the end of winter the water temperature in the entire 50 m layer under the ice 

 will be -0. 19°, i. e. , slightly higher than the freezing point, which will undoubtedly intensify 

 spring thawing quite substantially. 



6 . The rate of heat release by the sea to the atmosphere is most intense at the initial 

 moment of ice formation; then it decreases parabolically. After 958 freezing degree-days, the 

 rate of heat release becomes constant, since ice accretion ceases; after 2498 freezing degree-days, 

 it again decreases, and finally, it remains constant after 3735 freezing degree-days until the end 



of the season. 



7. In the example studied, the freezing degree-days were computed from the moment the 

 upper layer was cooled to the freezing point. Let us now assume that a storm mixed the two upper 

 layers at exactly this moment. In such a case, the average temperature of these layers will be 



0. 26°, and consequently ice formation can begin only after the surface of the sea releases 4. 8 

 kg-cal/cm2 to the atmosphere. 



218 



