http://www.cam.ac.uk/research/discussion/a-response-to-methane-mischief-misleading-commentary-published-in-nature
...
What is happening is that the summer sea ice now retreats so far, and for so long each summer, that there is a substantial ice-free season over the Siberian shelf, sufficient for solar irradiance to warm the surface water by a significant amount – up to 7C according to satellite data. That warming extends the 50 m or so to the seabed because we are dealing with only a polar surface water layer here (over the shelves the Arctic Ocean structure is one-layer rather than three layers) and the surface warming is mixed down by wave-induced mixing because the extensive open water permits large fetches. So long as some ice persisted on the shelf, the water mass was held to about 0C in summer because any further heat content in the water column was used for melting the ice underside. But once the ice disappears, as it has done, the temperature of the water can rise significantly, and the heat content reaching the seabed can melt the frozen sediments at a rate that was never before possible.
The 2008 US Climate Change Science Program report needs to be seen in this context. Equally, David Archer’s 2010 comment that “so far no one has seen or proposed a mechanism to make that (a catastrophic methane release) happen” was not informed by the Semiletov/Shakhova field experiments and the mechanism described above. Carolyn Rupple’s review of 2011 equally does not reflect awareness of this new mechanism.
...
That heat-transfer mechanism may be newly-discovered, but in reality it probably was in effect during past warming episodes (the paleoclimate counter-argument). I am just a lay person, but I think the counter-argument may lack a certain degree of validity due to the warming rate and initial conditions I mentioned earlier.
= new reply since forum marked as read