Andreas - I think we were talking past one another, and are generally in agreement, but there are a couple of things I disagree with you on.
...What I wanted to point out on the brightness temperature image above is that the IR which is detected by the satellite is going out to space. What makes these ice surfaces so very cold over the winter is not cold air, it is their negative radiation balance.
Absolutely agree - in the image I posted we're looking at loss from black body radiation, not convection, though I expect there's quite a bit of that going on as well at the sea surface where we have open water.
... The ice cools the air, not the other way round. (the snow layer on top of the ice to be precise) ... Warm air needs to get to the surface to warm it, but there it gets cold and gets in the way of more warmer air which is less dense than the cold.
Considering the recent temperature profile of the mass balance buoy Jim Hunt posted, that's very hard assertion to swallow. I've seen buoy profiles which have a "reservoir" of cold - temperatures going from near freezing at the surface grading to -20 one hundred or so CM below the surface and then grading back up to freezing at the ice/water interface.
That however was at the end of the season, with ice that was at or over 3M in thickness - not something we have over much of currently; and even where we have it, I'm doubtful heat flow has permitted that ice to lose that much heat at this stage of the melting season.
When air cools at the ice surface, it's not a result of ice cooling the air by conductive transport at the ice surface, rather it's exactly what you point out - air losing heat via black body radiation - which is not replaced fast enough via convection to balance the lost energy.
Put simply - the balance of heat flow at the ice surface, especially when covered with snow, is still in the opposite direction - ocean -> to ice -> to atmosphere - not the other way around, until such time as SST's are well above -1.8C.
That these areas of freezing over water form without a noticeable increase in extent would suggest that there is compression happening elsewhere. What the consequence of ridges are for melting exactly I don't know but I guess they are mostly reducing melt as a fraction of volume.
I think it would be a very tall order to quantify how much compression is taking place, and I think it is easy to overstate it given current conditions. Expansion of leads in one area can also be taken up by closing of incompletely frozen leads elsewhere.
It can also be taken up by export; the reanimated Fram this refreeze season has easily exported over 500,000KM2 of ice. The lions share of that has been MYI. While some of that remains as the current extent in the Greenland sea, much of that export has met it's doom at the margins where it is encountering astonishing and persistent heat in the water.
Lastly, the margins of the ice, especially in the marginal seas like the Bering, Okhotsk and Barents where most of our growth is currently taking place, are highly volatile. They are far from the core cold (... or lack thereof) in the Arctic basin proper, building ice over warmer water, and can gain or lose thousands of KM2 (of extent) in a day on a turn of the wind.
In short: that it isn't possible to predict melt season from what we see now, has been demonstrated by the recent years. Looking at the physics of what cracks do to the ice also shows that they should help to increase ice volume.
I agree, it is impossible to precisely predict the melt season outcome from what we have now.
I'd give you qualified agreement on your second point - leads permit rapid growth of ice in exposed areas, up to the first meter or so - which will have a positive effect on volume. I'm doubtful given current and recent past weather that growth of volume is as great as we could hope for.
How much such an increase (of volume - jd) can do in a climate which generally reduces ice growth and increases ice melt I can't guess. The complexity of these competing processes is what makes watching ice melt so fascinating for me.
Fascinating (and terrifying...) for me as well.
Regarding volume increase from leads - I'm pessimistic about the net effect of their contribution this year. The heat loss that ice represents has been dwarfed by the massive imports of heat via moisture and direct sensible heat brought by the endless series of storms entering the basin. Leads may freeze, as -20C air will do that just as well as -30C air. What that warmer temperature won't do is thicken Ice. Most of the ice across the basin won't reach much more than 2M thickness. I think as a result we might get a zero sum outcome for volume increase over the entire season. If we are lucky.