Hot or cold?

7/11/2011 - Since 1998, climate scientists have attempted to reconstruct the last millennium’s global annual temperatures by using natural proxies such as tree rings and ice cores. However, a new study finds substantial uncertainty in these reconstructions.

Blakeley McShane, assistant professor of marketing at the Kellogg School, and Abraham Wyner of the Wharton School at the University of Pennsylvania find that natural proxies have been weak predictors of global annual temperatures over the last 1,000 years.

	Related Links Click here to read the study, “A Statistical Analysis of Multiple Temperature Proxies: Are Reconstructions of Surface Temperatures Over the Last 1,000 Years Reliable?” (pdf) To read the rejoinder, click here (pdf)

Proxy-based temperature reconstructions such as the renowned “hockey stick” graph have arguably become the most iconic illustration of global warming. The hockey stick is a temperature reconstruction which features a long flat “handle” indicating temperatures were relatively consistent for 900 years, followed by a sharp upward “blade” indicating that temperatures dramatically increased over the past 100 years.

“We conclude unequivocally that the evidence for a ‘long-handled’ hockey stick is lacking in the data,” the authors wrote in the study. The handle of the hockey stick “is best understood to be a feature of regression and less a reflection of our knowledge of the truth .... The fundamental problem is that there is a limited amount of proxy data which dates back to 1000 A.D.; what is available is only weakly predictive of global annual temperature.” While the professors do not deny the existence of climate change, they believe the climate-change case as it pertains to proxy-based temperature reconstructions has been overstated.

In their study, the statisticians assessed the reliability of temperature reconstructions and their statistical significance against various models. Additionally, they compared existing reconstructions of the northern hemisphere’s average annual land temperature over the last millennium to their own reconstruction. While their reconstruction produced temperature estimates over the last 500 to 1,000 years that largely matched those produced by climate scientists’ models, McShane and Wyner found substantially larger uncertainty intervals as compared to the other models.

“Natural climate variability is not well understood and is probably quite large,” the authors wrote. “It is not clear that the proxies currently used to predict temperature are even predictive of it at the annual or decadal scale over several centuries. Nonetheless, paleoclimatoligical reconstructions constitute only one source of evidence in the Anthropogenic Global Warming (AGW) debate.”

The study, “A Statistical Analysis of Multiple Temperature Proxies: Are Reconstructions of Surface Temperatures Over the Last 1000 Years Reliable?” was published in March 2011 issue of The Annals of Applied Statistics.