The Statistics Behind Climate Change

If you're like me, even the mention of statistics causes the head to swim and the feet to move toward the  nearest exit. But understanding statistics is a necessity for those who are interested in most scientific issues, including climate change.  The beliefs of those who claim that anthropogenic global warming  is accelerating beyond natural levels rests in large part on the  UN  IPCC's premise that the warming seen over the past half century is statistically significant. Most scientists agree that there has been warming over the last hundreds of years, but there is substantial debate as to whether the warming has increased as a direct result of rising levels of carbon dioxide in the atmosphere. Figuring out whether the record reflects an increase in warming requres evaluating the temperature record to determine whether the temperature increase is statistically significant - i.e., whether there is no reasonable  likelihood that the temperature has remained the same or fallen during that period.

This article from the Wall Street Journal does as good a job as any in explaining in layman's terms how climate data should be analyzed, and why the UN and US study groups looking into the matter may have gotten things wrong. It was written by Douglas Keenan, who was described in this fashion:

Mr. Keenan previously did mathematical research and financial trading on Wall Street and in the City of London; since 1995, he has been studying independently. He supports environmentalism and energy security. Technical details of this essay can be found at http://www.informath.org/media/a41/b8.pdf.