I'm cramming a load of papers about the Arctic Atmosphere, changes in Arctic atmospheric circulation will be my theme for the next few posts. In the process of re-reading papers and reading new ones it's occurred to me that I haven't been as prescise as I should have been in some of my previous statements.
In an earlier post I've claimed that thermodynamics best explain the loss of Arctic sea-ice, I still think that this is the case. however think some clarification may be appropriate as I don't discount atmospheric and oceanic factors and may have given the impression that I do.
In the 1990s the Arctic Oscillation (AO) went into a period of high index, several subsequent papers cited this as a major issue in sea-ice loss, e.g. Rigor 2002 "On the Response of Sea Ice to the Arctic Oscillation". A high index AO state implies a low presure or cyclonic anomaly, this doesn't actually mean the air is circling in a counter-clockwise manner around the Artic, according to Rigor 2002 the actual Arctic circulation only becomes cyclonic during extreme positive AO index events (two standard deviation events), generally the circulation in the Arctic is anti-cyclonic. What happened in the 1990s was that lower pressure than typical dominated the Arctic.
Since the 1970s during high AO index periods there is increased FY ice production off the coast of Siberia due to divergence away from the coast, the speed of ice movement in the Transpolar Drift also increases, producing a 10% increased flux through the Fram Strait. However after the 1990s the AO subsequently returned to a more neutral index, despite this the loss of sea-ice continued.
As Maslanik's recent paper has shown it's taken only a few years for multi-year (MY) sea-ice to recover. And Tietsche et al show, with a modelling study, that even after total removal of the sea-ice in a model recovery time is a rapid as 2 years. First year (FY) sea-ice can bounce back rapidly from a perturbation, MY ice's response time is limited by it's slower growth rate, however it will regrow if the sea-ice is below it's equilibrium point, as the crash of 2007 and Maslanik's recent findings show.
Therefore if the ice loss of the 1990s was solely due to the impacts of the AO why did the ice-pack not recover after the AO ceased it's positive excursion? Its true, as Maslanik has recently shown, that there was a slight recovery of MY ice following the period of loss from 1986 to 1995, however this recovery in MY ice was but a seventh of the previous loss. Furthermore the decline in extent doesn't show an impact of the early '90s high index AO. As Bitz and Roe argue, first year sea-ice is quickly able to grow back in response to a perturbation. So whilst Maslanik tells us that the MY ice extent declined due to the high AO, the extent of total sea-ice didn't change because FY ice was able to quickly make up the difference. It was able to do so because the ice was below it's equilibrium point at that time.
Modelling studies are useful at this point. Almost all models that include the increase of CO2 show declines in Arctic sea-ice. The control model runs, those that don't have increases of CO2, do not show reductions in sea-ice.
So to sum up whilst there have been changes in Arctic sea-ice circulation driven by changes in the atmosphere, these changes have happened against a background of CO2 driven warming. Without that constant background forcing the changes in circulation would have been rapdily compensated for by regrowth of sea-ice, both FY (and at a slower rate) MY.