One of the main arguments on the web is that if one extends the volume trends shown by the PIOMAS they suggest that in summer the volume could reach zero by around 2016, e.g.
If you do the same with the winter trends you get a sea ice free state some time in the 2030s! Yet the sun still sets around September and rises in March, that will still happen in the 2030s, so during that cold winter period the ice will still freeze. Remember that this is a statistical extrapolation, so one can't support it by relying on changes one thinks will happen, e.g. Abbot & Tzipperman 2008. The statistics are based upon what's going in the period of data from which the trends are drawn. So for a start I think that if there's something fishy about what that technique says about winter, why trust it about the summer?
What I think(thought?) will happen is that as the ice thins and there is more open water at the end of the melt season, more heat will be vented to the atmosphere, and thence radiated to space. Furthermore
1) No rapid transition to a seasonally sea-ice free state: Because to a degree the enery lost due to atmosphere/space in winter will partially decouple the energy gains in preceding seasons. So whilst there may be drops ultimately there will be a long tail of lower rate of loss, i.e. gradual loss => rapid loss period => gradual loss. As seen in some climate model simulations e.g. Holland 2006.
2) A reduction of the trend in volume loss as the ice transitions to a mainly FY ice mass. i.e. the sea ice is in transition to a mainly FY sea ice state, this doesn't mean a transition to a sea-ice free state is imminent, even though it may seem so.
However I now see evidence which is causing me to doubt my position and wonder if we are indeed on a fast track to a sea-ice free Arctic sometime this decade.
I hate to rely once again on information from Ron Kwok, but his work of ICESat has been ground breaking and provides crucial information about what's going on in the Arctic. At present it seems to me that Cryosat is overstating sea-ice thickness, so I prefer not to use it.
This page from Kwok provides a neat summary of the key message from Kwok et al 2009 of the 2003. it's as well to click on the graphic to get a larger copy. Kwok et al 2009 "Thinning and volume loss of the Arctic Ocean sea ice cover: 2003–2008" is referenced in that page, and the publications page is excellent.
Basically what that page from Kwok shows is that the loss of volume has come from the loss of multi-year (MY) sea-ice, not first-year (FY) which has grown to replace the lost MY ice. From my previous post on sea ice thickness this is to be expected in light of the findings of Bitz & Roe: FY ice can grow more quickly than MY, so while MY ice has declined FY sea ice can grow back rapidly in the winter to take it's place. So most of the volume loss as shown by PIOMAS has been from the loss of MY ice volume.
Recently the PIOMAS model volume plot has shown an apparent increase in rate of volume loss.
Piomas Volume anomaly plot, Polar Science Center.
The most recent information from Kwok about MY sea ice is that the decline has levelled in 2008, 2009 & 2010, PDF. This levelling in the loss of MY sea-ice is what I had been expecting. But not at the same time as volume appears to be assuming a steeper rate of decline!
Then there is this year. I've downloaded all of the false-colour Bremmen AMSRE images for July, the critical month in the season as it sets up the ice pack for the rest of the melt. Running them as a slide show is staggering, even though I've followed previous years. The PIOMAS projection for 2011 uses 7 ensemble members that are hindcast with the weather of the past 7 years, the hindcast period runs up to the end of June, forecast from 1 July 2011. So that projection doesn't take into account the losses of July due to clear skies. Furthermore current forecasts imply that an Arctic Dipole type configuration may be about to replace the zonal high pressure, if so this will draw in warmer air from lower lattitudes into the Arctic Basin. A high rate of loss is to be expected now, and the rate will decline shortly: But the area rate will be acting upon a 'circle' of smaller and smaller radius. In terms of previous years we are currently closest to 2007.
Even allowing for the possibly imminent cloudier conditions, I now suspect that years of volume loss combined with the preconditioning of the first 2 weeks of July mean that the Arctic is now set up to meet or exceed the crash of 2007.
Abbott & Tzipperman, 2008, "Sea ice, high-latitude convection, and equable climates." PDF.
Holland, 2006, "Future abrupt reductions in the summer Arctic sea ice." PDF.