The world should be warming, if climate models predicting the effects of rising levels of greenhouse gases are to be believed. But satellite data seemed to throw cold water on these predictions, suggesting that the atmosphere has in fact become slightly cooler over the last 2 decades. Now two scientists, in a report in tomorrow's issue of Nature, challenge the satellite measurements, suggesting that they actually reveal a slight warming trend. The result, so far, is a heated debate if not a warming world.
Climatologists James Hurrell and Kevin Trenberth of the National Center for Atmospheric Research in Boulder, Colorado, plugged sea-surface temperatures into a model that predicts atmospheric temperatures, then compared the results with actual temperatures measured by microwave-sensing satellites. They found that satellite-recorded temperatures spiked downward twice with respect to the model's predictions--once in 1981 and again in 1991--about the same time that new satellites were put in orbit to collect data. The duo suggests that some of the apparent cooling may arise from a rough transition between satellites. "It's quite spectacular," says Trenberth. Remove those discrepancies, he claims, and "you can completely reconcile the records." The "corrected" satellite data show a weak warming trend of about a tenth of a degree per decade, in line with the climate models.
But the scientists who collect the satellite data are not so easily reconciled. Hurrell and Trenberth are wrong, asserts climatologist John Christy of the University of Alabama, Huntsville. "It's obvious they're not satellite meteorologists," he says. Christy says that in 1981, for instance, two satellites were in orbit collecting data and their data sets "merged perfectly." Meteorologist Alan Basist of the National Climatic Data Center in North Carolina suspects that, rather than indicating flaws in the satellite record, the discrepancy between the two data sets may arise from El Niño warming in the Pacific Ocean and from volcanic eruptions, including Mount Pinatubo in 1991, that occurred at roughly the same time as the spikes.
The debate is unlikely to die soon. Christy and Roy Spencer of the Marshall Space Flight Center in Huntsville have submitted a rebuttal to Nature. Trenberth happens to have been Christy's doctoral adviser. "There's a certain irony there," admits Trenberth. "I wish I had taught him statistics a little better," he adds with a laugh.