Warming Debate Highlights Poor Data

NEWS NATURE|Vol 436|18 August 2005

Weather balloons /SPL

used to collect VEY temperature data can

introduce errors into TIC SUR C

the climate record. AR BRITISH ANT

Warming debate highlights poor data

An unusual truce has been reached in the tur- But the three new papers add important new The second Science paper points out prob- bulent field of climate science. Scientists who results, says Carl Mears of Remote Sensing Sys- lems in the temperature record taken by have spent 15 years arguing over a discrepancy tems in Santa Rosa, California, and a co-author weather balloons, in part because different in certain data on global warming now say of one of the reports. “We are converging,” he manufacturers’ instruments heat up by differ- they all agree: the data are inadequate. says. “We are definitely getting closer.” ent amounts during the day3. The third The lowest layer of Earth’s atmosphere is Along with colleague Frank Wentz, Mears paper reviews predictions from 19 different indeed warming, but uncertainties in the data reanalysed raw satellite data, including those climate models and concludes that the differ- are as large as the trends the scientists are look- used in the original 1990 study. Their analysis2 ences between the models’ predictions and ing for, says Peter Thorne, a researcher with the suggests that the troposphere is warming at a observations are most likely to be the result of UK Met Office. “The fact that the uncertainty rate of 0.19 C per decade — far faster than the errors in the observations and in how they had has increased is actually a step forward,” he says. estimate of 0.09 C reported by a group at the been analysed4. Thorne is one of the lead authors on a new University of Alabama in Huntsville, who did Together, the reports conclude that the report commissioned by the US Climate the original study. observed tropospheric temperature trends are Change Science Program (CCSP), the govern- But recently the Alabama team issued its consistent with a warming world. Yet climate ment entity responsible for research on climate own revised data set, reporting a warming of experts acknowledge that the papers are just change. The report will be 0.12 C per decade. The group another data set to argue over, as happened released for public review in “What we all show is has adopted a new way to with an earlier analysis published in Nature5 the next few months, but that the troposphere is adjust the satellite data for the that failed to lay the argument to rest. three papers published online warming. The question time of day the observations “I don’t think this will be the last word,” says last week by Science preview were taken. The change comes climate researcher Phil Jones from the Univer- some of its findings. is by how much.” in response to a problem sity of East Anglia. ■ The papers advance a debate that began in pointed out during a meeting of authors for Jenny Hogan 1990, when an analysis of satellite observations the forthcoming CCSP report. suggested that temperature trends in the tro- “What we all show is that the troposphere 1. Spencer, R. W. & Christy, J. R. Science 247, 1558–1662 (1990). posphere, the lowest layer of the atmosphere, is warming. The question is by how much,” 2. Mears, C. A. & Wentz, F. J. Science doi:10.1126/ science.1114772 (2005). were inconsistent with the picture of global says climatologist John Christy of the Alabama 3. Sherwood, S., Lazante, J. & Meyer, C. Science doi:10.1126/ warming emerging from measurements made team. “When we are talking about precisions science.1115640 (2005). at the surface1. This finding became ammuni- of just a few hundredths of a degree per 4. Santer, B. D. et al. Science doi:10.1126/science.1114867 (2005). tion for sceptics arguing against evidence for decade, we’re not quite there with our observ- 5. Fu, Q., Johanson, C. M., Warren, S. G. & Seidel, D. J. Nature global warming. ing systems.” 429, 55–58 (2004).