An important image.
As we come to the end of the Arctic "refreeze" season, in order to understand the coming melt, I think it useful to consider our recent history (weather). To do that, I think we should start by looking at the results of the most recent climatological winter - Dec 1 through Feb 28 - which as compared to the astronomical season Dec 21 - Mar 21, better captures the deepest cold of the year.
In the Arctic, this is the time of greatest heat transfer from the ocean to atmosphere, and by extention the period during which most of the ice is built. During the past, thinking particularly of the end of the 20th century that we have records for it - 1979 through 1999 - we would see increases of about 9.5 million square kilometers.
(Jim Pettit graph 1 - annual seasonal sea ice loss, area and ending volume)
https://sites.google.com/site/pettitclimategraphs/sea-ice-area#asiaamlirThis is actually less than we gain typically now in the 21st century, which over the last few years has averaged close to 10.5 million KM2, with a high of 11.5 million KM2 in 2012. However, previously, the refreeze was starting from a much higher value - and average of just under 8 million KM2 vs the more recent typical 4.5-5.0 million KM2.
(Jim Pettit graph 2 - Arctic seasonal sea ice extent, decadally averaged)
https://sites.google.com/site/pettitclimategraphs/sea-ice-extent#asiedsThe ice appearing and how much of it returns is a function of heat loss first convectively from the ocean surface, and conductively, through the ice first formed during the initial freeze. Without going too deeply into the mechanics, the rule of thumb I've seen both calculated and published is that for each degree celsius air temperature is below freezing (-1.8C for seawater), in the Arctic, you will develop and maintain approximately 10CM of ice. In the past, this meant during the winter, you'd end up with between 2.5 and 3 meters of ice forming to reestablish the pack. Development of additional thickness required mechanical ridging.
Recently, with warmer temperatures during the winter, and more complete melt out of ice, the average thickness of the pack has declined considerably.
https://nsidc.org/sites/nsidc.org/files/images//kwok_rothrock.gif(Kwok, R., and D.A. Rothrock. 2009. Decline in Arctic sea ice thickness from submarine and ICESat records: 1958-2008. Geophysical Research Letters 36: L15501.- complete article link
http://dx.doi.org/10.1029/2009GL039035 )
What this is generally illustrative of can be summarized thus:
The net seasonal energy budget in the Arctic has not dramatically changed over the last 50 years (less than 10%). Approximately the same amount of energy entered the system to melt the ice established during the refreeze.
The net enthalpy of the system over time, as reflected in the decreasing maximum sea ice volume and area, has INCREASED dramatically.
There are a lot of factors feeding into this over time, which I'm not going to explore particularly here, but here's my take aways from the temperature anomalie drawing.
Next summer will be very perilous for the Arctic. Considering the 90 day anomalies shown, most of the arctic likely has ice which is between 50 and 80CM *thinner* than was typical, and likely 10s of CM thinner even than most recent years. Some Navy HYCOM estimates as illustration:
http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticictn/nowcast/ictn2015030218_2015030300_040_arcticictn.001.gif(current)
http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticictn/nowcast/ictn2014030118_2014030200_038_arcticictn.001.gif(March 2014)
My base prediction right now is this: Given average conditions, 2015 melt season will almost certainly be among the three lowest in modern times, and I give it a greater than 50% chance of breaking the 2012 record.
So my initial SWAG predictions for 2015 Minimums
Extent - 3.5 million KM2 +/- .5 Million (2012 - 3.178 million KM2)
Area - 2.75 million KM2 +/- .75 million (2012 - 2.23 million KM2)
Volume - 3500KM3 +/- 500KM3 (2012 - 3261 KM3)