Warmer still. Global warming is faster if satellite data are adjusted for stratospheric effects.

Getting Warmer, However You Measure It

Could thermometers be lying about global warming? The disagreement between satellite measurements of the atmosphere's temperature and thermometer measurements on the surface has been a persistent thorn in the side of climate researchers. Now, in this week's issue of Nature, a group of researchers describes a way to correct a long-standing problem in the satellite readings. Once adjusted, satellite temperatures show much the same warming in the lower atmosphere as do thermometers at the surface.

Thermometers show that Earth's surface has been warming by a substantial 0.17°C per decade. That's about what climate models predict under the growing greenhouse effect. But at least by some analyses, orbiting satellite sensors have detected a slower warming in the lower atmosphere, or even none at all. This gap between the surface thermometers and the globally comprehensive satellite readings has long been Exhibit A for contrarians arguing that the greenhouse effect is feeble.

The satellites have a tricky task. They take the temperature of the lower atmosphere by measuring microwave emissions with a series of microwave sounding units (MSUs) 850 kilometers up. The problem is that the satellites look down from above, and they get thrown off by the cool lower stratosphere that caps the lower atmosphere. Atmospheric scientists don't agree on how to tell temperature from satellite measurements of the atmosphere's microwave emissions. The general idea is that intensity of the emissions depends on temperature. There is agreement, however, that past analyses fail to take the lower stratosphere into account, where strong cooling has occurred due largely to the loss of ozone.

In the Nature paper, meteorologist Qiang Fu of the University of Washington, Seattle, and his colleagues offer a way to correct for the stratosphere's effect on the calculated lower atmosphere temperature by using stratospheric MSU measurements and balloon temperature measurements. When applied to recent warming trend estimates, the correction raises the trend from 0.10°C per decade to 0.18°C per decade, close in line with the surface warming trend.

Fu and his colleagues' approach “is a fairly simple but illuminating way of looking at the problem,” says meteorologist David Karoly of the University of Oklahoma in Norman. That would leave researchers to tussle with the uncertainties evident in the spread of the three published estimates. But for now satellite temperatures can no longer be used to portray a feeble greenhouse effect.