A chart of Barents Ice Cycles looks a lot like the icicles above, except upside down since Barents Sea is usually all water by September. Notice the black lines in the graph below hitting bottom near zero.
Note also the anomalies in red are flat until 1998, then decline to 2007 and then flat again.
Why Barents Sea Ice Matters
Barents Sea is No. 1, being located at the gateway between the Arctic and North Atlantic. Previous posts (here and here) have discussed research suggesting that changes in Barents Sea Ice may signal changes in Arctic Sea Ice a few years later. As well, the studies point to changes in heat transport from the North Atlantic driving the Barents Sea Ice, along with changes in salinity of the upper layer. And, as suggested by Zakharov (here), there are associated changes in atmospheric circulations, such as the NAO (North Atlantic Oscillation).
Here we look at MASIE over the last decade and other datasets over longer terms in search for such patterns.
Observed Barents Sea Ice
Below is a more detailed look at recent years.
This graph shows that the last two years were outliers in opposite directions. 2014 was an exceptionally high annual average due to melting delayed until April, and then a much higher minimum and faster than average recovery. In contrast 2015 was high initially, became average by day 91, then dropped sharply to a meltout, followed by a slower recovery. 2012 shows the lowest Barents ice year contrasting with 2014, the highest annual extent in the last decade.
Annual average BSIE (Barents Sea Ice Extent) is 315k km2, varying between 250k and 400k over the last ten years. The volatility is impressive, considering the daily Maximums and Minimums in the record. Average Max is 781k, ranging from 608k to 936k. Max occurs on day 77 (average) with a range from day 36 to 103. Average Min is 11k on day 244, ranging from 0k to 77k, and from days 210 to 278.
In fact, over this decade, there are not many average years. Five times BSIE melted to zero, two were about average, and 3 years much higher: 2006-7 were 2 and 3 times average, and 2014 was 7 times higher at 77k.
As for Maxes, only 1 year matched the 781k average. Four low years peaked at about 740k (2006,07,08 and 14), and the lowest year at 608k (2012). The four higher years start with the highest one, 936k in 2010, and include 2011, 13, and 15.
This graph confirms that Barents winter extents (JFMA) correlate strongly (0.73) with annual Barents extents. And there is a slightly less strong inverse correlation with NAO index (-0.64). That means winter NAO in its negative phase is associated with larger ice extents, and vice-versa.
Comparing Barents Ice and Arctic Annual
Arctic Annual extents correlate with Barents Annuals at a moderately strong 0.46, but have only weaker associations with winter NAO or Barents winter averages. It appears that 2012 and 2015 interrupted a pattern of slowly rising extents.
NAO and Arctic Ice Longer Term
Fortunately there are sources providing an history of Arctic ice longer term and overlapping with the satellite era. For example:
Observed sea ice extent in the Russian Arctic, 1933–2006 Andrew R. Mahoney et al (2008)
Mahoney et al say this about Arctic Ice oscillations:
We can therefore broadly divide the ice chart record into three periods. Period A, extending from the beginning of the record until the mid-1950s, was a period of declining summer sea ice extent over the whole Russian Arctic, though not consistently in every individual sea. . . Period B extended from the mid-1950s to the mid- 1980s and was a period of generally increasing or stable summer sea ice extent. For the Russian Arctic as a whole, this constituted a partial recovery of the sea ice lost during period A, though this is not the case in all seas. . . Period C began in the mid-1980s and continued to the end of the record (2006). It is characterized by a decrease in total and MY sea ice extent in all seas and seasons.
Comparing Arctic Ice with winter NAO index
The graph shows roughly a 60 year cycle, with a negative phase 1950-1980 and positive 1980 to 2010. As described above, Arctic ice extent grew up to 1979, the year satellite ice sensing started, and declined until 2007. The surprising NAO uptick recently coincides with the anomalous 2012 and 2015 meltings.
As of January 2016 NAO has gone negative for the first time in months.
If the Barents ice cycle repeats itself over the next decades, we should expect Arctic ice extents to grow as part of a natural oscillation. The NAO atmospheric circulation pattern is part of an ocean-ice-atmosphere system which is driven primarily by winter changes in the North Atlantic upper water layer.