Global Warming Data Affected By Land-Use Change, Study Says

December 04, 2007 - News Release

Land-use modifications for urbanization and agriculture have affected climate change data more than previously thought, according to new research by a University of Guelph professor.

In a paper published online this week in the Journal of Geophysical Research-Atmosphere, economics professor Ross McKitrick says the resulting discrepancies may be leading to an overstatement of the role of greenhouse gases in the atmosphere. In fact, the study concludes that skewed data could account for as much as half the post-1980 warming trend over land.

"Much of the temperature data used by the Intergovernmental Panel on Climate Change (IPCC) to measure global warming comes from places where people have modified the land surface for economic activity, as well as from low-income countries where there are few resources for maintaining continuous climate records," McKitrick said.

"To identify climate changes due to greenhouse gases, scientists have to make adjustments to the data to remove biases created by these kinds of influences."

For the study, McKitrick and co-author Patrick Michaels, a meteorologist with the Cato Institute, a non-profit public research centre in Washington, D.C., examined how the pattern of warming and cooling trends around the world compares with the patterns of population growth, economic development, coal consumption and other socio-economic indicators.

According to standard assumptions, trends in adjusted global climate data should not be correlated with patterns of economic activity, McKitrick said.

“But we found large, statistically significant correlations exist, indicating that the climate ‘signal’ in a commonly used scientific database remains contaminated with sources of bias that were supposed to be removed at the adjustment stage.”

The researchers applied a series of tests to check their results. For example, they looked at data measured by weather satellites in the lower atmosphere.

"We found that the correlations with economic activity pretty much vanishes even though the effects of greenhouse gases should be similar at the two layers," McKitrick said. "That tells us there is a unique problem in the surface data.”

They also determined that the effects were especially strong in regions where economic growth is faster, which also points to local socio-economic development as the culprit, he said.

The scientists used their results to simulate what worldwide trends would look like if the contamination were removed.

“Our estimate is that the measured warming over land since 1980 would go down by nearly half, which implies that these data problems are larger than is currently supposed," McKitrick said.

This is the second study on this topic by the two researchers. In 2004, they examined part of the Earth’s land surface and found similar results. For the new study, they extended their coverage to all available land regions.

McKitrick said the findings may affect current interpretations of global climate data, including how much warming is happening and what is causing it.

"The IPCC and other users of climate data need to consider the possibility that the basic data being used to study global warming are contaminated.”

Prof. Ross McKitrick
519 824-4120, Ext. 52532

For media questions, contact Communications and Public Affairs: Lori Bona Hunt, 519-824-4120, Ext. 53338, or Deirdre Healey, Ext. 56982,

University of Guelph
50 Stone Road East
Guelph, Ontario, N1G 2W1