The Three-dimensional Temperature Distribution and its Variation in Time 1 39 



very precise and affords only an approximation to the time required for such con- 

 vection processes, but it does give a criterion of the readiness for ice formation in 

 polar waters. This advantage shows clearly when evaluating an oceanographic cross- 

 section from this aspect. Figure 59 shows such a section from the Murmansk coast 

 (69° N., 36° E.) in an E.N.E.-direction almost to Novaya Zembla in the southern 

 Barents Sea, based on measurements made by the "Murmansk" Expedition 1903. 



station 



883 



884 



885 886 887 888 889 



200 



Fig. 59. Section in the Barents Sea from the Murmansk coast (69° N. 36° E) north-east to 



nearly Novaya Zembla. 



These observations were made a little before the beginning of the convection period. 

 The full lines show the heat in kg cal/cm^ transmitted to the atmosphere from the 

 sea surface when the convection process extends to the corresponding depth. From 

 the Murmansk coast to about 42° E., where warm Atlantic water reaches to con- 

 siderable depth, conditions are uniform and there is no readiness for ice formation 

 even when the entire water mass down to the bottom is affected by the convection. 

 East of the centre of the Barents Sea towards Novaya Zembla the readiness for ice 

 formation increases considerably and while the amount of ice that can form is at 

 first not very large it reaches at the easternmost station 889 the respectable thickness 

 of 1 -5 m or more. 



ZuBOV (1938) has developed, as it seems, a similar method for the determination 

 of ice potential in the ocean, without putting it into practice. He and Simpson (1954) 

 have again dealt with the same problem of predicting ice formation and growth and 

 in addition have derived new formulae for computing ice growth in terms of known 

 or predicted oceanographic and meteorological data. The method was used to fore- 

 cast the general features of the ice distribution in the Baffin Bay-Davis Strait area for 

 the season 1952-3. The methods for ice potential calculation have proved in practice 

 to give a reasonable answer for open seas, and for inshore areas where local variations 

 in the physical properties of the water are not large. In harbours and areas where run 



