Monday, October 22, 2007

McIntyre v Peterson on Urban Heat Islands

Canadian mathematician and climate scientist Steve McIntyre has found another striking error in the academic work supporting the case for climate alarmism.


Recall, in August McIntyre found an error in the way NASA was collating the temperature records from 1200 North American monitoring stations operating since the 1870s. NASA admitted the error, acknowledging they had been overstating the warming in North America since 1930 by a factor of 1.75x. (We wrote about the incident here, here, here, and here.)



This week McIntyre got his hands on the source data cited in an influential 2003 paper claiming the Urban Heat Island effect not having an important impact on historical temperature measurements. The article suggestively titled, Assessment of Urban Versus Rural In Situ Surface Temperatures in the Contiguous United States: No Difference Found, claimed:


Contrary to generally accepted wisdom, no statistically significant impact of urbanization could be found in annual temperatures.


Using the same data (sent to him by the author Thomas Peterson) McIntyre this claim to be completely unsupported. Here’s McIntyre’s plot of data, separating urban and rural monitoring stations based on Peterson’s own definition (click to enlarge):


peters26.gif



Its painfully obvious to the most casual observer that there’s a strong warming trend among urban stations, and none for the rural ones.



McIntyre cuts the data another way. Rather than using Peterson’s definition of what constitutes an urban station, he looks at stations in cities with NFL teams (he calls them ‘major cities’) versus everything else. Here’s the graph (click to enlarge):


peters27.gif

3 comments:

Bishop Hill said...

It can't be "this week" can it? McIntyre's post is dated August.

Jonathan Lowe said...

yes true...

Philip_B said...

The news today is that CO2 levels are increasing much faster than the IPCC predicted since 2000, and the previous trend of decreasing energy input and hence CO2 emissions per unit of GDP and has become an increase in CO2 emissions per unit of GDP.

From Bloomberg.com; The growth rate of carbon dioxide, or CO2, emissions has averaged 3.3 percent a year since 2000, compared with 1.1 percent in the 1990s, according to a study in the Proceedings of the National Academy of Sciences. The rate rose 35 percent more than scientists had anticipated based on economic growth, said Corinne Le Quere, one of the authors of the paper.

Note the 35% rate increase is relative to economic growth. The actual rate increase from a rate of 1.1% to a rate 3.3% would be 200%.

And of course, not even the normally objective Bloomberg dares mention that under Kyoto CO2 emissions have accelerated 200% and Kyoto itself is substantial to blame for the CO2 emission versus economic growth increase, because Kyoto drives energy intensive industries like steel out of energy efficient countries like Germany, Japan and Australia and into energy inefficient countries like China and India.

I'd add that decreasing energy input per unit of GDP of about 0.6% per annum was one of those things that happened year after year. It was a fundamental empirical constant of economic growth.

Then the United Nations comes along and in the name of reducing CO2 emissions dreamed up Kyoto and caused the seemingly economic impossible feat of increasing CO2 emissions per unit of GDP.

This is a huge story and I bet it never gets mentioned in the mainstream media.