State of the Greenland and Barents Sea Ice Cover in the Context of Current Climate Change

In the long-term variability of sea ice extent, a statistically significant negative linear trend has been identified for areas of the Greenland and Barents seas. Using the method of integral anomaly curves, periods of steady increase and decrease in sea ice extent are identified. The period of predominance of negative sea ice anomalies has been observed in the Greenland Sea since the winter of 2000/01 and in the Barents Sea since the winter of 2004/05, i.e., 4 years later. An analysis of the age structure of the ice cover shows that old ice predominated in the Greenland Sea throughout the whole winter period, occupying no less than 1/3 of the total ice area. Seasonal maxima of absolute values of the old ice area were observed in December and April. They correspond to two peaks in the seasonal course of ice exchange through the Fram Strait, which determines the amount of old ice in the sea area. The Barents Sea was characterized by the presence of old ice only in the waters of the northern regions, but the amount did not exceed 4% relative to the total area of the ice cover. A comparison of the estimates obtained in 1997–2022 with the results of earlier studies of the ice age from 1989–1992 for the Greenland Sea and from 1971–1976 for the Barents Sea is indicative of a change from a thick (old) ice stage of development to a thinner and younger ice (first-year) one and, as a consequence, a decrease in the average thickness of the ice cover. To reveal the dependence of changes in the sea ice area on various hydrometeorological factors, statistical analysis with use of multiregression models, namely the method of inclusion of variables, is applied. Various hydrometeorological parameters and climate indices are used as predictors. The regularities make it possible to construct statistical models of long-term variability of the sea ice extent for the winter and summer seasons, the reliability of which is 85–95%, with an efficiency of more than 10%. The reliability shows the percentage of justified forecasts to their total number (respectively, it is expressed in percent). The effectiveness of this forecast method (also expressed in percent) shows its preference compared to the climate prediction.