Levels of the atmosphere's main cleansing agent—the hydroxyl radical (OH)—have risen and fallen substantially since 1978, according to an international team of researchers whose analysis appeared in Science (10.1126/science.1058673, Science Express Research Articles). However, some atmospheric scientists caution that the findings are not definitive because of the uncertainties that accompany the group's estimates.

Because most atmospheric chemists believe that OH levels do not fluctuate greatly, the data surprised the team of U.S., Australian, and U.K. scientists, led by atmospheric scientist Ronald Prinn, chairman of the department of Earth Atmospheric and Planetary Sciences at the Massachusetts Institute of Technology in Cambridge, MA. "If our current analysis is correct, it implies that present understanding of OH chemistry—and hence the capacity of the atmosphere to remove many anthropogenically and naturally produced trace gases—is incomplete," write the researchers. "This has important implications for mitigation of air pollution and climate change." OH is highly reactive and cleans the air by destroying pollutants, including gases involved in ozone depletion and the greenhouse effect.

Global average OH estimates cannot be obtained directly because OH concentrations are highly variable in time and space. As a result, scientists estimate global OH levels by measuring levels of a trace gas that is destroyed primarily by OH. In this case, the researchers inferred OH concentrations from measurements of methyl chloroform, an industrial solvent destroyed by OH. Past studies have also used this method to estimate OH concentrations.

The researchers estimated that the average amount of OH in the atmosphere rose 15 ± 22% from 1979 to 1989 and then decreased in 2000 to levels about 10 ± 24 % below 1979 values. This implies a 25% change in 11 years, but this estimate is quite uncertain. In fact, it is possible that there was no change in OH concentrations over that 11-year period.

"If these trends in OH are truly occurring, then there is cause for concern," says Stephen Montzka, an atmospheric chemist with the National Oceanic and Atmospheric Administration in Boulder, CO. "But it seems that there are a number of factors that could cause this to be incorrect," he says. "The data set is invaluable, and the result is reasonable, but the uncertainties they've considered include the possibility that there is in fact no trend in OH, and they haven't even considered all of the uncertainties," says University of California–Irvine, professor Michael Prather, who models atmospheric chemistry.

Other atmospheric scientists, also cautious about the findings, note that the new analysis depends on the accuracy of methyl chloroform emissions data. This is because scientists infer OH concentrations based on the imbalance between estimates of methyl chloroform emissions and global observations. In the early 1990s, the Montreal Protocol called for the phaseout of methyl chloroform. As a result, "Methyl chloroform emissions data are particularly uncertain for the period that emissions have diminished by complying with the Montreal Protocol," says Jos Lelieveld, an atmospheric scientist and director of the atmospheric chemistry at Max-Planck Institute for Chemistry, in Mainz, Germany, who studies processes that control tropospheric ozone, one of the precursors to OH.

The coincidence between the time when methyl chloroform emissions began to diminish and the beginning of the decreasing OH trend also raises suspicions, says Lelieveld. "This needs to be looked at in detail," he adds.

Methyl chloroform measurements of polluted air from the mid-Atlantic states also cast doubt on the Prinn team's analysis, says Steven Wofsy, professor of atmospheric chemistry at Harvard University. These measurements, which were not used in Prinn's analysis, suggest that methyl chloroform emissions in the northeastern United States have changed little since 1997. This could be from storage, leakage from dumps, or bootlegging.

But according to Prinn, the methyl chloroform emissions estimates used in the recent analysis are as good as it gets. "We've used several independent sources to carefully estimate emissions and uncertainties," he says. These estimates account for emissions from storage, dumps, and bootlegging, and they also account for changes in the way methyl chloroform was used following the Montreal Protocol phaseout, according to atmospheric chemist Archie McCulloch, the member of the Prinn team who did the methyl chloroform emissions estimates.

The uncertainties in this new analysis reflect the state of the science, says Prather, who predicts that it will be about five years before a clear answer emerges. To get there, researchers will need better three-dimensional models to handle shifts in atmospheric circulation, a more complete data set, and a better handle on methyl chloroform emissions.

"If these OH changes are truly happening, then we should see it in the trends of other gases like methane and selected CFC alternatives that are also destroyed by OH," says Montzka. —REBECCA RENNER