Last spring I proposed Gompertz curves as simple models for thinking about past and future trends in Arctic sea ice extent, area and volume. Extent and area predictions made using these models in April came close to the observed September means:
NSIDC extent — predicted 4.4, observed 4.6
UB extent — predicted 4.6, observed 4.6
NSIDC area — predicted 3.1, observed 3.2
But the April volume prediction, based on what seemed at the time to be a pessimistic model, turns out to have been too high:
PIOMAS volume — predicted 5.2, observed 4.2
As estimated by researchers at the Polar Science Center of the University of Washington, using the PIOMAS model (website, data), the mean volume for September 2011 was only 4.2 thousand km3. Figure 8 shows this result together with a new Gompertz curve, which points toward a September 2012 mean of 4.0 thousand km3. Whereas the earlier model showed volume reaching very low levels (below 2 thousand km3) around 2016, this new version crosses that line one year sooner.
Arctic sea ice area and extent in 2011 reached one-day minimum values barely above 2007 (the previous record) according to some measures, or just barely below 2007 according to others. Sea ice volume, on the other hand, had already dropped well below the 2007 minimum in 2010. In 2011 it declined further, setting a clear new record: 4 thousand km3, down more than 75% from the 1979 minimum, and almost 40% below even the 2007 minimum (Figure 9).
If the PIOMAS series graphed in Figures 8 and 9 are anywhere close to correct, then sea ice volume is declining much faster than area or extent. That implies the remaining ice is thinner, but spread out over a wide area. Thinner, more intermittent ice was mentioned in reports from icebreakers on the Arctic Ocean this summer, and fits also with the fact that 2011 extent and area reached such low levels despite wind conditions that were more favorable than 2007. As noted often on this blog, thin ice could change state abruptly when conditions turn less favorable.