You need a model?

I can't get a response out of the Garnaut Review website—probably overloaded—to download the Review report. But here's an extract from Prof Garnaut's address to the Press Club on a 'curious turn' in AGW 'dissent':

"The dissent took a curious turn in Australia in 2008, with much prominence being given to assertions that a warming trend had ended over the last decade. This is a question that is amenable to statistical analysis, and we asked econometricians with expertise in analysis of time series to examine it. Their response, that the temperatures recorded in most of the last decade lie above the confidence level produced by any model that does not allow for a warming trend, is reported in Chapter 5 (Box 5.1)". extract from Herald-Sun report

Recent warming is a reality although I'm unconvinced that there has been any alarming degree of warming. I don't say "warming has ended." But the global temperature trend has been such (slightly declining) since 2001 that it presently disconfirms the projections in the IPCC's model-salad. If what the IPCC does is science then we must take the disconfirmation of its projections into account. That doesn't need another model. So far, statistical analysis appears to show that they're wrong.

Update: David Stockwell has a copy of the 'model' to which Prof Garnaut refers. More debate likely to follow.

Second update: In their paper, Drs Breusch and Vahid—the statisticians to whom Prof. Garnaut referred the question of whether warming continues—are unable to rule out a trend in the temperature data (HADCRUT, NOAA, GISS) since 1850. No more than that. They conclude that statistical evidence does not allow them to say anything about the trend "without corroborating data from other sources and close knowledge of the climate system." Taking the analysis a step further, the statisticians find that if they pick a starting date of 1958 i.e. 50 years ago, then the temperature anomaly data since then exceeds the 2-standard-deviations boundary for a projected 'no warming' trend after about 2001. This is their 'model'. But since the 1958 date coincides with a low point in the anomaly data (-0.3 degrees C less than zero in the HADCRUT series) their choice of starting point seems to be a case of 'cherry picking' a date.

Your thoughts?…

Earlier Entries

You'll find a complete list of entries here. You could also search by keywords.