Current outlook as shown by the ECM, and to a lesser extent, the GFS, would have us easily beating the early August losses from 2012.
The general pattern starts off benign enough, but quickly turns by 3 days out, as the ridging over the Beaufort area strengthens the high pressure there, while errant jet stream fragments combine to cause a rapid intensification of the low over the ESS/Laptev sector.
t48: Nothing much going on, weak surface high to the west, slight low pressure development to the east, shown by the white oval.
Rapid intensification occurs over the next 24 hours, with the low deep enough and the pressure gradient steep enough to cause some damage. Main wind direction in wine.
By t96, the low pressure has grown to cover about a third of the Arctic, with a central low pressure below 980hPa and a very wide wind field, strengthened by the high pressure to the west of the Arctic.
We can see a strong, almost jet streak feature cross the N. Pole at t96. This feature is a strong contributor to the rapid low development, and continues fueling the storm up to t120.
t96
t120: The Arctic jet streak has the highest jet stream speed in the entire northern hemisphere!
Below shows the areas likely to be affected by the storm from t48 to t96 (white, white, black)
Looking at the details further than t96 is a bit dodgy at this stage, given how unusual this set up is. It does appear, with agreement from the ECM and GFS, that the low pressure will gradually weaken and hang around just north of Svalbard between t120 and t168 (5-7 days out), while high pressure builds over the Pacific sector of the Arctic, introducing milder air to the region.
From 8-10 days out, the 500hPa GPH chart suggests little change, with a stronger dipole like set up than we've seen all summer.
With all this in mind, some very high melt rates could be on the cards starting from the middle of next week. A mixture of storm damage, upwelling, export, compaction and general melting could well produce a second period with 7 day losses exceeding 1 million km2.