One of the biggest climatic dangers out there is that warming in the Arctic will melt the permafrost. The tundra is heavily laden with methane – a potent greenhouse gas. In total, the ten million square kilometres contain about 1,000 gigatonnes of carbon (3,670 gigatonnes of carbon dioxide). The permafrost contains more carbon dioxide equivalent than the entire atmosphere at present.
If even a fraction of a percent of that gets released every year, it would blow our carbon budget. Even with enormous cuts in human emissions, the planet would keep on warming. Right now, humanity is emitting about 8 gigatonnes of carbon a year, on track to hit 11 gigatonnes by 2020. If we were to stabilize at that level, emitting 11 gigatonnes a year until 2100, the concentration of greenhouse gasses in the atmosphere will surpass 1,000 parts per million, creating the certainty of a vastly transformed world and a very strong possibility of the end of human civilization.
As such, it is vital to stop climate change before the planet warms sufficiently to start melting permafrost. This is especially challenging given that warming in the Arctic is more pronounced than warming elsewhere. There is also the additional challenge of the sea-ice feedback loop, wherein the replacement of reflective ice with absorptive water increases warming.
The actions necessary to prevent that are eminently possible. Unfortunately, people have not yet developed the will to implement them to anything like the degree necessary. Hopefully, the ongoing UNFCCC process for producing a Kyoto successor will help set us along that path before it becomes fantastically more difficult and expensive to act.
[Update: 4 February 2009] Here is a post on the danger of self-amplifying, runaway climate change: Is runaway climate change possible? Hansen’s take.
[Update: 19 February 2010] See also: The threat from methane in the North.