For seven puzzling years at the turn of the millennium, the most consequential greenhouse gas after carbon dioxide simply stopped rising. Between 1999 and 2006, global monitoring networks recorded almost no growth in atmospheric methane, even as economic expansion across Asia and the developing world was widely expected to push anthropogenic emissions upward. The pause, known to atmospheric scientists as the methane plateau, has been one of the most debated mysteries in the global carbon cycle. A new study published in Nature now offers what its authors describe as a coherent explanation: the plateau was sustained not because emissions stalled, but because the atmosphere’s own capacity to destroy methane strengthened at precisely the right time and in precisely the right places.

The research, led by Yu Zhu and Shushi Peng of Peking University together with Lu Shen and colleagues including Kelvin H. Bates of the University of Colorado Boulder, combines chemical transport model simulations with observations of methane and its carbon isotopes. The team concludes that global concentrations of the hydroxyl radical, OH, the highly reactive molecule that serves as the atmosphere’s principal methane-destroying agent, increased by 1.4 plus or minus 0.4 percent during 2000 to 2006 relative to 1999. Crucially, most of that enhancement occurred in the tropics, the band of warm, sunlit, humid air where the overwhelming majority of methane removal takes place. Because methane’s chemical loss scales directly with OH abundance, this tropical strengthening translated into a 2.2 plus or minus 0.7 percent amplification of global methane loss over the plateau period, and a 3.7 plus or minus 1.3 percent increase by 2006 relative to the 1999 baseline.

Those numbers matter because they are large enough to have absorbed the emissions growth that was happening at the same time. According to the study’s attribution analysis, the 2000 to 2006 enhancement of the methane sink was equivalent to 136 percent, with a plausible range of 90 to 184 percent, of the contemporaneous growth in anthropogenic methane emissions. In other words, the atmosphere was destroying methane faster by an amount that fully offset, and in the central estimate exceeded, the extra methane that industry, agriculture, and landfills were adding. The result was the flat concentration record that monitoring stations around the world actually observed, despite rising sources.

What drove the OH increase? The answer, the researchers find, lies in nitrogen oxides, collectively known as NOx, the pollutant family dominated by nitrogen dioxide and nitric oxide that is best known for producing urban smog. NOx chemistry has a complicated relationship with atmospheric oxidants: in polluted low-NOx regimes, additional NOx accelerates the recycling of hydroperoxyl radicals back into OH, boosting the oxidizing power of the air. The study shows that during the late 1990s and early 2000s, global NOx emissions rose and, just as importantly, were spatially redistributed. Economic growth in tropical developing countries lifted their emissions, while the expansion of global trade drove a marked increase in shipping emissions over the oceans. Meanwhile, developed regions in the Northern Hemisphere mid-latitudes were beginning to curb their own NOx outputs through air quality regulation.

The net effect was a southward and oceanward shift of the world’s reactive nitrogen emissions, moving NOx from regions where it is chemically less efficient at producing OH toward the tropical and marine environments where it is far more potent. The team quantified this using sensitivity simulations in the GEOS-Chem chemical transport model, diagnosing the sector-specific response of tropospheric OH to changes in land-based, shipping, and aircraft NOx emissions. All three sectors showed statistically significant positive relationships between NOx emission changes and OH changes, with the land and shipping contributions dominating the tropical enhancement. The spatial redistribution, not merely the total magnitude of emissions, emerges as a central part of the story.

The methodological approach is notable for the way it triangulates independent lines of evidence. Methane observations alone cannot cleanly separate changes in sources from changes in sinks, a long-standing ambiguity in the field. The researchers therefore brought in measurements of the carbon isotope ratio of atmospheric methane, delta-13C-CH4, which responds differently to emissions from fossil, microbial, and biomass-burning sources and to the kinetic isotope effect of OH destruction. By running Monte Carlo ensembles of emission combinations and retaining only those consistent with both the methane concentration record and the isotopic constraints, the team narrowed the space of plausible explanations and found that a strengthening tropical sink was required to reproduce the observed plateau.

The decomposition of the methane sink change adds further texture. Over 1999 to 2006, changes in OH accounted for roughly 65 percent of the cumulative increase in the global methane sink relative to 1999, while temperature-dependent changes in the reaction rate constant contributed about 30 percent, and rising methane concentrations themselves contributed the remaining 5 percent. This partitioning underscores that meteorology and chemistry both played roles, but that the OH response to the NOx redistribution was the dominant and most novel factor. It also helps explain why earlier studies, which often assumed a more static OH field, struggled to reconcile rising emission inventories with the flat concentration record.

The findings carry an uncomfortable irony that is likely to fuel public and policy debate. The very pollutant responsible for acid rain, photochemical smog, and hundreds of thousands of premature deaths each year appears to have been quietly doing the climate a favor, scrubbing methane from the air faster than it otherwise would have. As air quality regulations succeed and NOx emissions decline in the decades ahead, a portion of that hidden climate service will be withdrawn, potentially accelerating methane growth unless direct methane emission cuts compensate. The study’s authors and the broader community have long warned of this feedback, but the new results give it a concrete historical precedent: the atmosphere’s oxidizing capacity is not a fixed backdrop but a variable that responds to human activity on decadal timescales.

There are also implications for how scientists project future methane trajectories. Climate models that treat OH as constant, or that capture only its response to meteorology, may misjudge the methane lifetime and therefore the warming commitment associated with a given emissions pathway. The Peking University team’s results suggest that the geography of pollution matters as much as its quantity: a tonne of NOx emitted over a tropical ocean does different chemistry than the same tonne emitted over a mid-latitude industrial region. As trade patterns shift, shipping fuels change, and developing economies industrialize, the global OH field will continue to evolve in ways that either amplify or dampen methane’s climate forcing.

The methane plateau itself ended in 2006, after which atmospheric concentrations resumed a steep climb that continues today, driven by growing microbial emissions from wetlands, agriculture, and waste. Understanding why growth paused in the early 2000s is therefore more than an exercise in historical bookkeeping. It reveals the sensitivity of the methane budget to the atmosphere’s chemical state and identifies a lever, the NOx-OH relationship, that has already shaped climate outcomes once. Whether the world’s remaining carbon budget is spent faster or slower in the coming decades may depend in part on getting this chemistry right, and on ensuring that methane mitigation efforts are sized not against a static atmosphere but against a dynamic one whose cleaning capacity is itself changing.

Subject of Research: Drivers of the 1999–2006 atmospheric methane plateau through NOx-driven hydroxyl radical variability

Article Title: Atmospheric CH4 plateau sustained by NOx emission rise and southward shift

Article References: Zhu, Y., Shen, L., Liu, G., Bates, K. H., Cai, Y., & Peng, S. (2026). Atmospheric CH4 plateau sustained by NOx emission rise and southward shift. Nature, 657(8132), 674-679. https://doi.org/10.1038/s41586-026-10983-w

Image Credits: AI Generated

DOI: 10.1038/s41586-026-10983-w

Keywords: methane, hydroxyl radical, nitrogen oxides, atmospheric chemistry, greenhouse gases, NOx emissions, shipping emissions, tropics, GEOS-Chem, isotopes, methane budget, climate

Cite Scienmag News
APA MLA Chicago

Copy citation Download RIS

Tags: atmospheric chemistryatmospheric chemistry and climate changeatmospheric methane declinechemical transport model simulationsclimateclimate change mitigation and greenhouse gas dynamicsGEOS-Chemglobal methane cyclegreenhouse gas emission trendsgreenhouse gaseshydroxyl radicalhydroxyl radical increaseimpacts of atmospheric composition on greenhouse gasesisotopeslong-term methane monitoringmethanemethane budgetmethane destruction mechanismsMethane plateaunitrogen oxidesNOx emissionsrole of hydroxyl radicals in methane removalshipping emissionstropics