don't forget peripheral seas such as Kara, Okhotsk, Bering, Norwegian and earlier snow-free permafrost land.
As a public service, I digitized the key journal graphic of DK Perovich showing a full season of solar heat influx measured at the surface of the Beaufort. That's provided as the .txt attachment below which is really a .cvs spreadsheet file.
The initial graphic was a bit funky, not being provided at high resolution and using phong-rendered metallic balls (inset on first graphic) instead of points or cross-hatches to indicate the data points. However just like a Landsat can be grossly over-sampled yet still be informative, it proved feasible to position vertical bars precisely and use their pixel heights as watts/sq meter values (after adjusting for irrationality in the vertical scale rendering).
Powerful online curve-fitting tools (such as
http://www.xuru.org/rt/LR.asp) aren't helpful here because the data, being so affected by clouds, is just a single noisy sample from an unknown underlying distribution. However common sense suggests that function peaks at the summer solstice and tapers off monotonically on both sides without inflection points, constraining its first and second derivatives.
A rolling window convolution can then 'take out the clouds' to a certain extent (6th column of spreadsheet), yielding a reasonably tight upper bound on daily solar heat influx into open water at the latitudes of the Beaufort Sea. A top-down approach starting with the TOA solar black body spectrum and taking out atmospheric absorption isn't going to work as well, either via Ceres or modeling, because of the oblique optical path taken by light reaching the Arctic Ocean.
In summary, taking the 'dot product' of AMSR2 daily open water with column 6 solar heat influx over the 2016 season (coming shortly) will likely indicate that a quite respectable fraction of the total solar heat to be absorbed by the 'seasonally ice-free Arctic Ocean' is
already being taken up (since the remaining non-peripheral seas are at even more unfavorable solar latitudes).
Consequently, the
effects of polar radiator loss have already begun -- what mysterious physical process could be holding them off? There's little purpose served waiting upon tipping points or invoking changes of state, whistling in your tent won't keep the bear out. Yes, there is a distinction between polar bear, grizzly bear and black bear but it's still a bear.