ultimate driver is thermodynamic. How much heat enters the arctic. Direct heat in is a question of albedo, clouds hurt, reflective ice hurts, clear skies maximize, blue waters maximize.
Just so we are all clear on the research results (#2343) applicable to the Arctic Ocean and Greenland, clouds can and sometimes do make melt worse than straight sunlight. Indeed, heat transfer from moist imported clouds is the dominant driver of early melt of the Arctic Ocean, not sunlight through clear skies.
This seems counterintuitive because the top of the cloud reflects a portion of the sunlight back into space and so that energy never has the opportunity to warm the ice or ocean, whereas no clouds mean more solar energy reaches the surface.
However for thinner clouds like we have been seeing this spring*, quite a bit of the solar radiation does make it to the surface, either directly or after rounds of elastically scattering within the cloud. A portion of it is taken up and the rest reflected, some back up to the clouds to be partly back-scattered down again and so forth.
The counterintuitive part is where the heat (absorbed shortwave fraction) re-emits at longer wavelengths per the graybody spectrum appropriate to the ~0º C ice/water surface. The cloud is no longer transparent to this upwelling infrared so absorbs and re-radiates some of it downward. This amounts to an efficient near-surface greenhouse effect for low thin liquid-containing clouds.
After folding in all the transmissivity coefficients, the net result can be more heating of ice than would have happened had straight sunlight came down on high albedo ice or reflective water at the unfavorably oblique angles appropriate to spring/summer and Arctic latitudes.
Decades of in situ polar radiometric observations published by D Perovich and others have quantitated this for various conditions:
The Radiation Budget of Sea Ice during the Springtime Melt
http://tinyurl.com/gwsmcd8Such clouds are very common over the Arctic and Greenland:
G Cesani 2012
http://onlinelibrary.wiley.com/doi/10.1029/2012GL053385/abstractThe remarkable Greenland melt event of 12 July 12 even melted dry facies at Summit Station, 3216 m). The decisive effect was not clear skies from a stagnant ridge as initially thought:
Bennartz 2013
http://www.nature.com/nature/journal/v496/n7443/abs/nature12002.html "Here we show that low-level clouds consisting of liquid water droplets, via their radiative effects, played a key part in this melt event by increasing near-surface temperatures.
At the critical surface melt time, the clouds were optically thick enough and low enough to enhance the downwelling infrared flux at the surface. At the same time they were optically thin enough to allow sufficient solar radiation to penetrate through them and raise surface temperatures above the melting point.
Outside this narrow range in cloud optical thickness, the radiative contribution to the surface energy budget would have been diminished, and the spatial extent of this melting event would have been smaller.
We further show that these thin, low-level liquid clouds occur frequently, both over Greenland and across the Arctic, being present around 30–50% of the time... Global climate models fail in simulating the Arctic surface energy budget because they under-predict the formation of optically thin liquid clouds at supercooled temperatures"
Hanna 2013
http://onlinelibrary.wiley.com/doi/10.1002/joc.3743/full "In 2012, as in recent warm summers since 2007, a blocking high pressure feature, associated with negative NAO conditions, was present in the mid-troposphere over Greenland for much of the summer. This circulation pattern advected relatively warm southerly winds over the western flank of the ice sheet, forming a ‘heat dome’ over Greenland that led to the widespread surface melting."
van Tricht 2016 free full
http://www.nature.com/ncomms/2016/160112/ncomms10266/full/ncomms10266.html "The main drivers of Greenland ice sheet runoff remain poorly understood. Here we show that clouds enhance meltwater runoff by about one-third relative to clear skies, using a unique combination of active satellite observations, climate model data and snow model simulations. This impact results from a cloud radiative effect of 29.5 Wm
2.
Contrary to conventional wisdom, however, the Greenland ice sheet responds to this energy through a new pathway by which clouds reduce meltwater refreezing as opposed to increasing surface melt directly, thereby accelerating bare-ice exposure and enhancing meltwater runoff. The high sensitivity of the Greenland ice sheet to both ice-only and liquid-bearing clouds highlights the need for accurate cloud representations in climate models.
The dominating effect depends strongly on cloud properties such as vertically integrated ice and liquid water contents that determine cloud optical depth and emissivity, in addition to cloud temperature, sun position and surface albedo.
CloudSat and CALIPSO data suggest liquid-bearing clouds that contain both ice and (supercooled) liquid water are present 28% of the time, consistent with other work showing that such clouds are prevalent throughout the Arctic."
*You can see that on the 367 Modis over the Beaufort because floe details are still visible, ie sunlight has passed through the clouds, reflected off the ice, passed through the clouds again and reached the satellite sensors at sufficient levels for imaging despite absorption and scattering at each step.