Entries RSS Comments RSS

Articles to 2011-12-01

First the link to this week’s complete list.

I find it very funny how the innocuous little story from the list of Oct 22nd draws out the feminists of both sexes in Gerardin. Structurally it’s very similar to the ones about Santa [nature 456 (2008), 1007–1008] and Neanderthals [nature 453 (2008), 562] but neither of those has such well organised lobby groups with vested interests behind them. With all the power and finance in lobbying I’m never sure whose interests those groups have at heart, their own or those of the group they claim to represent.

Michael reveals a typical fallacy usually committed by epidemiologists. Any rare event naturally looks improbable where and when it occurs if viewed in isolation. But events and event clusters have to happen somewhere and sometime.

Two independent articles deal with arctic sea ice. First there is the doomsday one by Kinnard et al., according to whom the present decline is unprecedented in the last 1450 years. By their own data this is true only for the Fram Strait, while the Chukchi Sea had significantly less cover from 1400 to 1600 AD than it has now. Their graph of total ice coverage does show a recent singular decline but also a continuous rise all through the mediaeval optimum followed by a decline through the little ice age, so clearly something besides global temperature is at work here.

Clegg et al. don’t show the last few hundred years but also demonstrate the arctic is more complicated than a simple linear response. Temperatures in the arctic were lower than today up to 7 ka BP with lots of ice cover, rose to a maximum around 5 ka BP, 3 ka later than in lower latitudes, then gradually fell again. Ice cover declined much later to an all-time low (over 14 ka) at 2 ka BC and rose to its all-time high at the height of the mediaeval optimum around 1 ka BP. Thus a current decline won’t be abnormal but rather a regression from an extreme toward a long-term average.

Leave a Reply

You must be logged in to post a comment.