As globe warms, atmosphere keeps its cool.

A panel of top scientists last week doled out ammunition to both sides in the debate over global warming.

The National Research Council (NRC) committee concluded that Earth's surface has warmed dramatically over the past 2 decades, accelerating a trend observed throughout the 20th century. The panel also confirmed, however, that the lowermost atmosphere has warmed only slightly or not at all since 1979, backing up measurements made by satellites.

"As scientists, we really have our work cut out for us now, as we recognize that there really is a difference between the behavior of the temperature at the two levels that we don't fully understand," says panel chairperson John M. Wallace of the University of Washington in Seattle.

For years, skeptics of global warming have trumpeted the results of the satellite measurements while questioning the accuracy of surface measurements. Meanwhile, many climate scientists convinced of the surface warming have looked askance at the satellite data, especially after reports of problems with this record (SN: 9/4/99, p. 150).

The NRC committee concluded that the errors in the two records are not large enough to explain the discrepancy between them. The surface has warmed 0.25 [degrees] to 0.40 [degrees] C during the past 20 years, whereas the troposphere--the atmospheric layer up to 8 kilometers--has warmed 0 [degrees] to 0.2 [degrees] C, says the report.

Some of the difference may stem from recent natural events. The period has produced two of the largest El Nino warmings of the century, along with two large volcanic eruptions that cooled the climate. Such factors affect the surface and the troposphere differently.

Another complicating factor is Earth's ozone layer, just above the troposphere. Thinned by pollution over the past 20 years, the ozone shield now absorbs less solar radiation than it did before, thereby cooling the upper troposphere.

In the past, computer climate models predicted that the troposphere and Earth's surface should warm together as greenhouse-gas pollution builds up overhead. The atmosphere's reluctance to follow suit has led some researchers to question these models.

"Here is a set of observations that hasn't been well replicated [in models], and it gives us a bit of skepticism about other aspects of the climate models and their ability to predict for century timescales," says panel member John R. Christy of the University of Alabama at Huntsville, who developed the satellite record.

Recent climate simulations that incorporate added factors have done a better job at matching the temperature records, says Wallace. He also notes that data from before 1979--the start of the satellite record--reveal an atmospheric warming trend similar to that seen at the surface.

The panel ducked the question of what is causing the recent warming as well as the 20th-century trend. The year 1998 was the warmest since global records began in 1860. Last year ended shy of that mark; the tropical Pacific Ocean had cooled off in a phenomenon known as La Nina. Still, 1999 ended up fifth-warmest year overall, says panel member David E. Parker of the Meteorological Office in Bracknell, England.

This kind of heightened temperature, even in the face of La Nina, adds weight to the theory that greenhouse gases are now warming Earth, he says.

A reviewer of the report, Richard S. Lindzen of the Massachusetts Institute of Technology, debates that point: "There is the problem that if you don't have warming in the atmosphere and you do have it at the surface, the surface [warming] is not a greenhouse response." By definition, he says, the greenhouse effect comes from atmospheric warming of the surface.

All the scientists involved agree that 20 years is too short a span for drawing conclusions about climate change.