Gerd Bürger: 20th century warming was statistically insignificant

In previous weekly doses of consensus-busting climatological literature in peer-reviewed magazines, we discussed articles in Nature, Geophysical Research Letters, and many other sources. Today we look in Science magazine, volume 316 (the volume 317 is already available on-line).

Gerd Bürger argues nothing less than that the global character of the 20th century warming was not statistically significant.

Bürger: Technical comment on "The Spatial Extent of 20th-Century Warmth in the Context of the Past 1200 Years"

The author explains that the authors of the original article, Osborn and Briffa, didn't take proxy screening into account when they calculated the significance levels. It really means that they neglected, in their quantitative analysis of the significance, that proxies may be correlated with the existing temperatures and included among temperature-sensitive proxies by chance.

Update: Later, I realized that Steve McIntyre has already discussed Bürger's paper in June 2007 when it became available online.

Once this mistake is corrected and the probability of this kind of error included in the calculation, the significance of the 20th-century warming anomaly disappears. The 99th percentile is almost never exceeded while the 95th percentile is surpassed in the early 20th century as well as the early 11th century while the 17th century (and the year 1700 in particular) were equally anomalously cold - nothing that would support the AGW theory. Just see that there is nothing unusual about the last 70 years on this figure:



Figure 1: Difference between the fraction of records that have smoothed and normalized proxy anomalies whose absolute value is greater than zero (top), greater than one (center), greater than two (bottom), using only proxies that are significantly temperature sensitive (bold), along with the (10th, 90th), (5th, 95th), and (1st, 99th) percentile bands (thin) from 1000 random-based series exceeding the critical correlation. The area of the signal outside the (5th, 95th) percentile band is filled black.

Osborn and Briffa reply with a "nothing matters" argument that I don't understand. They say that they used a small number of records which should remove Bürger's worries about the statistical significance. I would guess that this fact would, on the contrary, strengthen the worries.

Many people, including skeptics, have been saying for years that they had no doubts about the 20th-century warming, its global character, and the assertion that it can't be understood as noise. I, for one, have always had doubts about this statement. Whether there has been anything in the 20th century that would deserve to be called global warming is pretty much an open question.

WCR predicts that the MSM won't inform their readers and spectators that the Science magazine has printed a technical paper that shows that an unusual 20th century warmth didn't exist. We won't see NBC telling us that "Expert in the Science magazine proves that there was no global warming in the 20th century." Well, that's not too ambitious a prediction by WCR.

Hat tip: World Climate Report, JunkScience.com