Part of the scientific consensus on global warming may be flawed, a new study asserts.
The researchers compared predictions of 22 widely used climate "models" — elaborate schematics that try to forecast how the global weather system will behave — with actual readings gathered by surface stations, weather balloons and orbiting satellites over the past three decades.
The study, published online this week in the International Journal of Climatology, found that while most of the models predicted that the middle and upper parts of the troposphere —1 to 6 miles above the Earth's surface — would have warmed drastically over the past 30 years, actual observations showed only a little warming, especially over tropical regions.
"Can the models accurately explain the climate from the recent past? It seems that the answer is no," said lead study author David H. Douglass, a physicist specializing in climate at the University of Rochester.
Douglass and his co-authors S. Fred Singer, a physicist at the University of Virginia, and John R. Christy, a climatologist at the University of Alabama at Huntsville, are noted global-warming skeptics.
However, Christy was a major contributor to the 2001 report by the Intergovernmental Panel on Climate Change, and is one of the world's premier authorities on collection and analysis of satellite-derived temperature data, having been commended by both NASA and the American Meteorological Society for his efforts.
"We do not see accelerated warming in the tropical troposphere," said Christy. "Instead, the lower and middle atmosphere are warming the same or less than the surface."
The difference between the climate models and the satellite data has been known for several years.
Studies in 2005 found that improper compensation for temperature differences between day and night was the cause of most of the satellite-data discrepancy, a correction that Christy has accepted.
No explanation has been put forth for the weather-balloon discrepancy.