As Drogaitsev currently points out, in the Arctic Seas where we encounter intimate expanses 

 of open water and ice masses in close proximity, the horizontal temperature gradients of the imder- 

 lying surfaces are large enough to change atmospheric pressure at sea level. , i.e. , that recorded 

 on synoptic charts. 



We must also add to Drogaitsev's discussion that it is not so much the actual air temperature, 

 as the processes of evaporization and condensation which absorb and release enormous quantities of 

 heat that play a role in the thermal interaction of the ocean and the atmosphere. These processes, 

 as we have seen, are determined by the water vapor pressure, while the later process (condensa- 

 tion) is not the same over ice and water, even when their temperatures are the same. Therefore, 

 there can never be equilibrium in the atmosphere above ice and above the sea. 



Changes in atmospheric pressure cause corresponding changes in wind direction, while this 

 latter case causes changes in air temperature, in the amount of advective heat, etc. ; these in turn 

 are all reflected in the ocean's regime. 



Up to this point we have examined anomalies which are created in individual regions due to 

 anomalies in the temperature of sea currents. We may note that the effect of even slight anomalies 

 of this type may be increased considerably by the pressure distribution anomalies imparted by 

 them, and even more so by the wind distribution anomalies. 



Actually, the corresponding wind distribution may, for example, break up the ice and create 

 between the pieces a considerable band of open water or, may drive the ice away from the corre- 

 sponding shore. The spaces of open water formed absorb solar radiation considerably better than 

 the ice, and therefore the horizontal air temperature gradient increases, the pressure gradient 

 intensifies, etc. Thus, the pressure and wind distribution anomalies created by the temperature 

 anomalies and sea currents may, under certain combinations of physical-geographic conditions, 

 play the role of resonators, strongly increasing the magnitude of these anomalies. 



Naturally, for the transfer of temperature anomalies by sea currents, the following are 

 characteristic: 1) All other anomalies of the physical-chemical characteristics of a water mass 

 are simultaneously transferred (e.g. , salinity) and 2) This process generally requires a long 

 period of time. For example, according to my calculations the temperature anomaly of Atlantic 

 waters, observed off northwestern Spitsbergen, should in some way or another be reflected in 

 Bering Strait only after 4-1/2 years. Dobrovol'skii, having determined the propagation rate of 

 Atlantic waters in the Arctic Basin for observations at the station "North Pole" and on the Sedov 

 arrived at the same results. 



Vize has told me that according to his 1943 calculations , the anomalies of temperature , pres- 

 sure, iciness, etc. , spread in the seas of the Soviet Arctic from west to east at a rate of 20° 

 longitude per year. 



We have seen that any sea temperature anomaly causes changes or shifts in one direction or 

 another of the pressure topography. However, a change or shift of the pressure topography in one 

 region of the sea causes definite shifts in that of the adjacent regions. 



Figure 27 (according to Ovchinnikov) shows the connection between the departures in the tem- 

 peratures of the January cross-section along the Kola Meridian (33°30' east), in the Barents Sea 

 and the deviations of the winter air temperatures (December-February) from the mean multi-annual 

 temperatures at the "Uellen" polar station and the iciness of the Chuckchee Sea in degrees of 

 concentration. 



87 



