I've just bought an read yet another climate book - David Archer's "The Long Thaw" ISBN: 978-0-691-14811-3.
As far as I was concerned, the most interesting comment in the book is that the geological records shows ice-sheets disintegrating an melting remarkably quickly. Archer thinks that the Greenland ice cap and some parts of the Antarctic ice sheet may slide off into the sea as icebergs relatively soon, making nonsense of the IPCC's predicted 0.5 metres of sea level rise by 2100. A five metre sea-level rise within a century or two looks rather more consistent with the geological record.
Archer ties this "sliding off" thawing to the occurence of Heinrich events during the last ice age. They happened every eight to ten thousand years, and seem to represent chunks of the Laurentide (Canadian) ice sheet sliding off into the Atlantic and melting - a process that seemed to take a a few hundred years. In today's warmer climate, the more northerly Greenland ice cap could slide off in the same way, and just as fast.
The Heinrich events were superimposed on the Dansgaard-Oeschger (D-O) cycles which are visible as saw-tooth oscillations in the Greenland ice-core temperatures, where 1470 years of slow cooling is followed by an abrupt warming. The cooling gets progressively more severe from cycle to cycle until you get a Heinrich event in the middle of a D-O cooling, where the cooling trend stops, and the next D-O cycle starts off warmer than anything seen since the previous Heinrich event.
This sort of regularity is enough to make me reconsider my stance on climatic chaos - when Phil Hobbs argued that climat was chaotic, I argued that the Younger Dryas event - which looks very like a Heinrich event - was probably better seen as a one-off rather than as one of many limit-cycles in a persistent set of coupled processes. This attitude now looks less defensible.
I still argue that even if climate is chaotic, this doesn't mean that it is unpredictable - as I pointed out to Phil Larkin in a simlar argument a few years ago, the solar system is also chaotic, but it is only unpredictable over periods of longer than 100 million years. Climate may be chaotic, but it certainly can be predicted much further in advance than weather - agriculture couldn't work otherwise
- and it is well worth trying to get a better handle on what is going on, not least because chaotic systems can be redirected by remarkably small effects if the stimulus is applied at the right place and the right time.
That a butterfly flapping its wing in Brazil might have caused hurricane Katrina does imply that a meteorologist waving a table- tennis bat in the right adjacent clearing, at the right time, could have prevented it.
-- Bill Sloman, Nijmegen