Travelling the Northwest Passage by kayak is about many things.  But sea ice is what dominates the journey's feasibility.  No open water, no kayaking.  Where things get complicated is how much open water one needs to be able to paddle between two resupply points before running out of food or summer.  I don't have any experience paddling in sea ice, nor do I know anyone who does, so this turns into a guessing game.  The ice is quite mobile and so the stakes aren't trivial, as Kabloona's 18 ice locked days on Hudson Bay show.  Mark Serreze observes, in his understated way in his talk at AGU last fall, that there's significant variability from year to year and the precipitous melt of 2007 makes it even harder than usual to estimate what will happen in the future.  The overall decline is stark and the risk of further steep declines in the near future is clear.  From a kayaking perspective this sends two messages.  If you want a more historically typical artic experience and are willing to take the risks that entails, go as soon as you can.  If you're willing to trade ice navigation risks for the somewhat uncertain climate changes of global warming, wait.  Going as soon as one can is involved, but doesn't seem likely to be all that soon.

As I look at the parts if the Northwest Passage I'd like to paddle in NSIDC's historical animations I find myself inclined to think in terms of three ice systems.  These are the Beaufort Sea and Amundsen Gulf, the ice between Canada's arctic islands and mainland between Dolphin and Union Strait and the Gulf of Boothia, and Baffin Bay.  Historically, what's happened is the Cape Bathurst polynya opens in the eastern Amundsen and the Mackenzie opens a polynya of its own in the Beaufort.  The ice then fractures and open water appears from Tuk most of the way to Kugluktuk fairly early in the melt season due to a combination of melt and ice transport toward the Chukchi Sea.  Unusually, this system operated this winter as well, creating open water from Dolphin and Union Strait to Barrow by July 2008.  The passage east of Amundsen Gulf---Dolphin and Union Strait, Coronation Gulf, Dease Strait, Queen Maud Gulf, and Simpson and Rae Straits---is not subject to this ice evacuation and hence melts out later.  The route from Taloyoak to Igloolik across the Gulf of Boothia, Pelly Bay, Committee Bay, and Fury and Hecia Strait is even more iced up, with fast and pack ice persisting past the end of paddling season.  Barring significant warming, the only paddling option in this area area is an out and back trip from Kugaaruk.  Fast and thick first year ice persists along Baffin's east coast though the end of paddling season as well.  So it's not particularly kayakable either aside from short, late season out and back trips from Qikiqtarjuag.

Looking at various ice coverage examples, it seems to me a 30% concentration is the upper end of what one can paddle without getting into lots of navigational complexity and probably ending up falling through weak ice while trying to land or portage.  Concentrations under 20% seem a good bit more attractive.  While Arctic Sea Kayaking Adventures ran trips in Pelly Bay this summer with concentrations above 35%, these presumably have the luxury of being out and back trips which can vary their route based on what the ice does and can probably call back to Kugaaruk for support.  The other data point I have is NSIDC data which suggests ice concentrations ranged from 0 to maybe 35% during Kabloona's 1994 paddle from Paulatuk to Gjoa Haven.  Ironically, despite the massive 2005 and 2007 melts, 1994 has the second lowest amount of ice the recent historical record for this part of the passage.  Only 1996 had more open water.

Conveniently, Canda publishes recent weekly and daily ice charts in the Annual Arctic Ice Atlas, which offer a great deal of insight into melt patterns and season to season variability.  The main challenge is melt times vary by about a month, which becomes quite significant as the paddling season's two months long at the outside. Paddling from Tuk to Kuglugtuk is the easiest part with respect to ice so long as one paddles from west to east, timing the transit of Dolphin and Union Strait and western Coronation Gulf to happen at the end of the paddling season.  Paddling the rest of the passage is harder as the ice melts out later.  The advantage of getting to Kuglugtuk is the southwest end of Coronation Gulf is the first to melt out, typically in late July, which means one has a decent shot at making a run to Cambridge Bay before the season closes by sticking to the south coast of Coronation Gulf and following the ice as it melts out through Dease Strait.  The same approach works for getting from Cambridge Bay to Gjoa Haven but is trickier as Queen Maud Gulf fractures a couple weeks after Coronation, with a tounge of ice persisting between Cambridge Bay and Gjoa Haven until the end of the season.  Rae and Simpson Straits fracture and melt out before Queen Maud Gulf, but the crossing of the gulf is likely to happen too late in the season to allow time to reach Taloyoak.

As the artic warms I suspect a year will eventualy turn up like 1996 where one can start from Kugluktuk early enough to arrive in Taloyoak.  However it depends on how the artic warms.  1994 and 1996 are near the end of a period of warmer than usual years in this area.  Recent years are cooler than usual, though the coolest parts are east of Taloyoak.  Lately pack ice has persisted through the end of kayak season in Committee Bay.  Will the cold spot persist?  Move west?  Go someplace else?  Who knows? But I figure it'll take a string of warmer years to make paddling between Kuglugtuk and Taloyoak viable.  Points eastward along the passage would require even more warming to be kayakable.  Be interesting to see how this winter's freeze and 2009's melt go.