August 13, 1998
By WILLIAM K. STEVENS
The New York Times
As nearly as climatologists can tell by studying measurements dating back to the 1850s, the earth's average surface temperature has risen by about one degree over the last century. But measurements made by orbiting satellites have suggested that the lower atmosphere is cooling, leading some skeptics to argue that the world is not warming.
Now two scientists in California have reanalyzed the two-decade record of satellite data and concluded that it has been distorted by the inevitable decay, or lowering, of the satellites' orbits as they encounter atmospheric resistance.
Once this error is corrected, the scientists report Thursday in the journal Nature, the satellite record shows the atmosphere has become warmer, but not as much as the surface has.
The authors of the study, Frank Wentz and Dr. Matthias Schabel, atmospheric physicists with Remote Sensing Systems, a private research firm in Santa Rosa, Calif., report an atmospheric warming of about 0.13 of a degree Fahrenheit per decade.
The finding could signal "a sea change in the global warming debate," Dr. James Hansen and four colleagues write in a commentary that is to appear Friday in the journal Science. Hansen, the director of the NASA Goddard Institute for Space Studies in New York, has long been an advocate of the view that emissions of heat-trapping gases like carbon dioxide are warming the earth.
If the Wentz-Schabel finding stands up, Hansen said in an interview, it will make it "very difficult to deny the reality of global warming," though he and his colleagues note in the Science article that the finding is likely to be challenged and possibly revised.
Dr. John R. Christy of the University of Alabama, an originator and keeper of the satellite record, immediately offered a revision. While Christy called the report, "real" and "legitimate," he said other corrections involving, for example, east-west drift of the satellites and instrument temperatures had also recently been made. The net result of all the corrections, he said, shows a very slight warming in the lower atmosphere, less than the figures reported in Nature and the reading derived from surface measurements.
Christy acknowledged that this does not necessarily contradict the surface record. He and others have long maintained that the temperature profile of the surface and that of the lower atmosphere would not be expected to be identical, since they differ physically. Comparing the two records, many scientists say, is like comparing apples and oranges.
"The surface is doing something slightly different from the atmosphere above," Dr. Christy said, "and that's just the climate. They do not have to be the same."
Even so, he has long noted, the surface and satellite data agree well in many regions, including the Northern Hemisphere continents.
For the last 10 years, Christy and a colleague, Dr. Roy Spencer of NASA's Marshall Space Flight Center in Huntsville have been using data from weather satellites to put together a continuous global temperature record. The surface records have many gaps and inconsistencies resulting from changing methods of measurement.
And although analysts have tried to correct for these, Christy and Spencer saw their satellite data as having an advantage.
A disadvantage, however, has been that the satellite data record goes back only to 1979, and many experts say this is too short a period to detect a meaningful trend. And it has been challenged on its own methodological grounds, most recently in Thursday's Nature article.
Even before the new corrections of the satellite record, Christy had calculated that once the short-lived warming effects of El Nino and the cooling effects of volcanic eruptions were statistically removed, the satellite measurements showed an atmospheric warming trend of about 0.1 degree per decade. Now, with corrections in place, he calculates it at 0.05 degree to 0.18 degree per decade. This compares with about 0.25 of a degree for the surface record.
Experts say that over the long century and a half of the surface data record the effects of El Nino and volcanic eruptions tend to cancel each other out and that no adjustment for them is necessary.
Even with El Nino and the volcanoes included, Christy said, the short satellite record now showed a warming trend of 0.07 degree per decade.
This, he said, is because the 1997-98 edition of El Nino warmed the atmosphere so much that it counterbalanced the cooling effects of volcano eruptions in the 1980s and early 1990s, which emitted a haze that reflected sunlight.
From Nature Aug. 13,1998:
Effects of orbital decay on satellite-derived lower-tropospheric temperature trends
The 17-year lower-tropospheric temperature record derived from the satellite Microwave Sounding Unit (MSU) shows a global cooling trend, from 1979 to 1995, of 0.05K per decade at an altitude of about 3.5km. Air temperatures measured at the Earth's surface, in contrast, have risen by approximately +0.13K per decade over the same period. The two temperature records are derived from measurements of different physical parameters, and thus are not directly comparable. In fact, the lower stratosphere is cooling substantially (by about 0.5K per decade), so the warming trend seen at the surface is expected to diminish with altitude and change into a cooling trend at some point in the troposphere. Even so, it has been suggested that the cooling trend seen in the satellite data is excessive. The difficulty in reconciling the information from these different sources has sparked a debate in the climate community about possible instrumental problems and the existence of global warming. Here the authors identify an artificial cooling trend in the satellite-derived temperature series caused by previously neglected orbital-decay effects. They find a new, corrected estimate of +0.07K per decade for the MSU-based temperature trend, which is in closer agreement with surface temperatures. They also find that the reported cooling of the lower troposphere, relative to the middle troposphere, is another artifact caused by uncorrected orbital-decay
F J Wentz & M Schabel
Effects of orbital decay on satellite- derived lower-tropospheric temperature trends (Letter to Nature)
Nature 394, 661 (1998)