The Polyakov study also concludes that the warming trend alone cannot explain the retreat of Arctic sea ice observed in the 1980-1990s, which was probably caused by the shift in the atmospheric pressure pattern from anti-cyclonic to cyclonic.
The mechanism of sea ice changes is incredibly complex, and it is extremely difficult to identify the rather short-term anthropogenic influence from the background of natural phenomena, which are both long and short term. Depending on the period of time studied, the records containing only a few years to a few decades of data, yield different trends. For example, Winsor (40) reported that six submarine cruises between 1991-1997, transecting the Central Arctic Basin from 76°N to 90°N and around the North Pole (above 87°N), found a slight increasing trend in sea ice thickness. Vinje in 1999, 2001, and 2003 (41, 43) reviewed observations of the extent of ice in the Nordic Seas measured in April 1864-1998, and also back in time for a full 400 years. Sea-ice extent has decreased there by 33 percent over the past 135 years. However, nearly half of this decrease was observed over the period 1864-1900. The first half of this decline occurred over a period when the CO2 concentration in air rose by only 7 parts per million volume (ppmv), whereas for the second half of the decline, the CO2 content rose by over 70 ppmv. This suggests that the rise of CO2 content in the air has nothing to do with the sea-ice cover.
Vinje 42 stated that the
"annual melt-backs of the magnitude observed after about 1930 have not been observed in the Barents Sea since the 18th Century temperature optimum,"
which was followed by
"a fall in the Northern Hemisphere mean temperature of about 0.6°C over the last few decades of the 18th Century,"
which temperature has just now been finally erased by
"a rise of about 0.7°C over a period 1800-2000."
Consequently, the Northern Hemisphere would appear to be not much warmer now (and the extent of Barents Sea ice cover not much less now) than it was during the 1700s, when the CO2 air concentration was claimed to be 90 to 100 ppmv less than it is now. (The validity of this claim was criticized by Jaworowski in References 29 and 44.)
Even high-sensitivity short-term determinations of surface air temperature or sea-ice, covering one or two decades (for example, satellite observations between 1981 and 2001, appearing in the November 1, 2001, issue of the Journal of Climate, showing a 9% per decade decline of Arctic sea-ice), are not the best basis for the determination of man-made impact on the climate of polar regions. This is valid also for Antarctic studies, where over the past 18 years the net trend in the mean sea-ice edge has expanded northward by 0.011 degree of latitude per year, indicating that the global extent of sea-ice may be on the rise. (45)