Table 105. (Continued) 



The calculated drift corrected in this way agrees with the observed drift, although the dis- 

 crepancies which appear indicate that the coefficients are somewhat too high. 



In the case of the drift of the Lenin , the nature of the discrepanices clearly indicate that 

 the decisive role here was played by the immediate proximity of the coastline. Actually, during 

 the entire winter the observed drift deviated to the left of the computed drift, namely in the direc- 

 tion in which it should have deviated with a coastline to the right. Possibly, in this case the shoals 

 also played a part. It is quite probable that the rule of ice drift along the isobars is not fully 

 applicable to ice over shoals. It is easy to show that if the shoals did have a substantial influence 

 on the direction of wind drift, it would deflect the drift to the left of the isobars. 



It has already been noted that figure 135 was constructed without consideration of the steady 

 current in the drift region of the Sedo u . After the elements of the steady currents in the drift 

 region of the Sedov had been computed, it appeared that there was some possibility of determining 

 the coefficient which connects the speed of pure wind drift with a distance between the isobars by 

 empirical methods. I did this for the ten-day pressure maps, and on an average of 35 cases 

 examined I found: 



1000 



where c is the pure wind drift of ice fields (steady current excluded), expressed in miles/ten days; 

 X is the distance in miles between the isobars on the ten day pressure charts, drawn every 1 mb. 



It should be emphasized that 37 cases are too few to establish a sufficiently reliable factor 

 of proportionality between the pressure gradient or the distance between isobars and the speed. 



In view of this, I feel personally that the law of movement along the isobars (in the deep part 

 of the Arctic Basin at any rate) may be considered sufficiently proven, but further data are required 

 to determine the factor of proportionality between the pressure gradient and the drift speed. 



Thus, I suggest that the maps of computed and observed drift given here as well as other 

 maps constructed on the basis of these formulas should be examined merely as an application of a 

 new method and an indication of the possibilities offered by this method. Further, although the 

 schemes obtained on the basis of the proposed method cannot be regarded as having an absolute 

 value, the relative data obtained by the new method certainly merits attention. 



It should be noted that attempts at a practical application of this method for short range ice 

 forecasting have already been made and have yielded positive results. For example, in the summer 

 of 1940 Ovchinnikov successfully traced the movements of ice edges in the Kara Sea, using the law 

 of movement of ice along isobars and calculating the ice drift speed from the atmospheric pressure 



376 



