Wednesday, June 08, 2005

Predicting Climate Change (2)

Time to look at another prediction from the Foresight Exchange (see here for previous). This time I'll look at global temperature rise - the central question of climate change itself. The GW2030 claim (which I set up) pays $0.01 for each 0.02C by which the globally averaged surface air temperature in 2030 exceeds that in 2000 (using a centred 5-year average at each end). Data is taken from NASA GISS. I chose 30 years so as to be long enough to contain a meaningful anthropogenic warming signal on top of natural variability, but not too long for me to have a hope of finding out the answer.

The current market price of $0.40 implies a temperature rise of 0.8C in 30 years - a rate of 0.27C per decade. The last 30 years or so shows a substantially lower trend of 0.17C per decade. I've extrapolated out the trend line to 2030 (result: 0.5C warming) and added two alternative futures of accelerated warming, 0.2C/decade and 0.3C/decade starting at the 2000 temperature. Of course these give 0.6C and 0.9C warming respectively in 30 years.

I wouldn't rule out 0.8C as a possibility, but it seems to err on the high side to me. Modelling experiments don't usually quote 30 year forecasts explicitly, but my impression is that they also tend to generally display a lower rate of warming, except for some experiments with highly pessimistic emissions scenarios and/or extreme model sensitivities. The IPCC TAR has the following comment:
anthropogenic warming is likely to lie in the range of 0.1 to 0.2°C per decade over the next few decades
and I'm not aware of any strong evidence to change that much in the forthcoming AR4, Disclaimer: I'm selling on this claim!

No comments: