The stakes of getting OH right are enormous, particularly for methane. Methane is the second most important anthropogenic greenhouse gas, and its atmospheric lifetime is determined almost entirely by how quickly it reacts with OH. If models overestimate OH, they will underestimate how long methane lingers in the air and, in turn, miscalculate its climate impact. This is precisely the problem that has plagued the modelling community. Major model intercomparison projects, including HTAP, ACCENT and ACCMIP, produced multi-model mean methane lifetimes of roughly 9.7 to 10.2 years, whereas observationally constrained estimates, such as those by Prather and colleagues, point to a longer lifetime of about 11.2 years. The gap implies that most global models simulate too much OH, and identifying why has become one of atmospheric chemistry’s most persistent puzzles.

Hou and colleagues attacked the problem with the FRSGC/UCI chemistry transport model, an offline global model driven by meteorological data from the European Centre for Medium-Range Weather Forecasts. The team ran nine hindcast simulations spanning 2000 to 2017, each designed to switch a specific process on or off so that its individual contribution to OH could be isolated. Four recently revised or newly appreciated processes took centre stage: the absorption of ultraviolet radiation by water vapour, heterogeneous chemistry on clouds and aerosol particles, an updated rate coefficient for the reaction between OH and nitrogen dioxide, and emissions of acetaldehyde from the ocean surface. The model was evaluated against an impressive array of observations, including tropospheric ozone columns from the OMI/MLS satellite instruments, nitrogen dioxide columns from OMI QA4ECV, carbon monoxide columns from MOPITT, surface ozone from the TOAR database, and vertical profiles of OH, HO2 and acetaldehyde measured during NASA’s ATom aircraft campaigns across four seasons and nearly the full latitude range of the globe.

The evaluation revealed both strengths and a critical weakness. The model reproduced the spatial distribution and year-to-year variability of tropospheric ozone and nitrogen dioxide columns well, with correlation coefficients exceeding 0.6 in most regions, lending confidence to the simulated influence of changing emissions on OH. Along the ATom flight tracks, simulated OH fell within observational uncertainty with no clear systematic bias through the troposphere, a notable achievement given the widespread tendency of global models to overestimate OH. However, the model failed to capture the observed decline in the carbon monoxide column reported by MOPITT, instead simulating significant increases over China, India and Africa. Because reaction with CO is the single largest sink of OH, accounting for roughly 41 percent of total OH removal in the control run, this mismatch matters enormously. The authors conclude that the discrepancy, likely rooted in uncertain biomass burning emissions in the southern hemisphere and overestimated CO trends over Asia, may cause the model to underestimate any increase in OH, or overestimate any decrease, over the study period.

The sensitivity experiments delivered striking quantitative results. When the team removed the absorption of ultraviolet radiation by water vapour, a process long considered negligible until Prather and Zhu highlighted its importance in 2024, global mean OH rose by 3.6 percent and the methane lifetime shortened by 4.2 percent. In the humid tropical lower troposphere, where the effect is strongest, water vapour UV absorption suppresses OH by 6 to 10 percent at the surface by soaking up the radiation needed to photolyse ozone and produce the excited oxygen atoms that generate OH. Heterogeneous chemistry proved equally consequential. Neglecting reactions on cloud droplets and ice increased global mean OH by 2.4 percent, while neglecting reactions on the five aerosol types considered, sea salt, dust, sulphate, black carbon and organic carbon, increased OH by 3.2 percent. Together, omitting all heterogeneous chemistry inflated OH by 5.8 percent and shortened the methane lifetime by 5.2 percent. Intriguingly, clouds exert a disproportionately large influence on global OH despite occurring where nitrogen oxide levels are low, because they scavenge NOx in remote, high-altitude regions that aerosols, concentrated near industrial surfaces, cannot reach.

Perhaps the most surprising result concerned a single reaction. The termolecular reaction between OH and NO2 terminates radical chains and controls the lifetime of NOx, and its rate coefficient has recently been remeasured with unprecedented precision by Amedro and colleagues. When Hou’s team adopted the faster new rate coefficient, global mean OH fell by 7 percent and the methane lifetime lengthened by 8.4 percent, with the largest effects in the upper troposphere and at high latitudes. The reaction contributes only about 1 percent of total OH loss, so this outsized response reflects nonlinear adjustments in the coupled HOx-NOx chemical system and radical recycling. The finding underscores how sensitive the atmosphere’s oxidative capacity is to laboratory kinetics, and it argues forcefully for tighter experimental constraints on fundamental reaction rates.

Oceanic emissions of acetaldehyde, a reactive carbonyl compound that provides a direct sink for OH in the marine boundary layer, rounded out the analysis. The model without these emissions underestimated acetaldehyde by up to an order of magnitude over the remote Pacific, but adding ocean sources of 40.8 to 62.7 teragrams per year, estimates that differ substantially between the CESM2 and GEOS-Chem frameworks used, reduced the marine boundary layer bias from roughly 100 to 20 parts per trillion. The global impact was modest: OH decreased by up to 1.5 percent and the methane lifetime increased by up to 1.6 percent. The authors note that a missing chemical source of acetaldehyde likely remains in the free troposphere, so a stronger effect cannot be ruled out as emission estimates improve.

Beyond the process-level accounting, the study tackled the contentious question of OH trends. The tropospheric OH column showed pronounced spatial heterogeneity, with significant increases of more than 1.5 times ten to the eleventh molecules per square centimetre per year over India and China, exceeding 1.6 percent per year and driven principally by rising anthropogenic emissions of nitrogen oxides and ozone precursors. In contrast, OH declined over tropical Africa, the tropical Atlantic and the Indian Ocean by more than 1 percent per year, changes attributable to meteorological variability rather than emissions. When averaged globally, these regional patterns largely cancel, leaving a weak but statistically significant positive trend in global mean OH, strongest in summer. Crucially, the answer depended on the experimental design: with time-varying emissions there was little change in OH, while with emissions fixed at year-2000 levels, OH increased more strongly, revealing an important contribution from interannual changes in meteorology, temperature, water vapour, clouds and photolysis.

The interannual story proved equally revealing. Global mean OH dipped markedly in 2002 and 2015 and rose in 2007 and 2016, and these swings tracked fire activity closely. Elevated wildfire emissions during El Niño years pumped carbon monoxide into the atmosphere and consumed OH, while the drop in fire-related CO emissions associated with the 2006 to 2007 La Niña allowed OH to recover. Annual OH anomalies correlated negatively with fire-related CO emission anomalies at a striking coefficient of minus 0.75. Comparisons with independent estimates derived from methyl chloroform and hydrofluorocarbon inversions, and from other models, showed common temporal features despite differing amplitudes, including enhanced OH around 2012 to 2013 and lower values in 2014 to 2015, suggesting that global OH over this period is characterised primarily by interannual variability rather than a persistent long-term trend.

The broader message is one of cautious progress. Incorporating water vapour UV absorption, heterogeneous chemistry and the updated OH plus NO2 kinetics reduces simulated OH levels and lengthens methane lifetimes, thereby shrinking the systematic high-OH bias that has afflicted global models for two decades, although these processes alone cannot fully close the gap. The study also highlights where future effort must go: better constraints on the kinetics of the OH plus NO2 reaction, more reliable estimates of marine acetaldehyde emissions, consistent implementation of current process understanding across modelling groups, and resolution of the CO emission discrepancies that cloud any attribution of OH trends. For a molecule that lives for barely a second yet governs the fate of methane and the cleansing of the air everyone breathes, OH has never been better understood, and the path to even sharper answers has never been clearer.

Subject of Research: Tropospheric hydroxyl radical variability and its implications for methane lifetime

Article Title: Tropospheric OH and its trends: new insight into atmospheric processes and implications for methane lifetime

Article References: Tropospheric OH and its trends: new insight into atmospheric processes and implications for methane lifetime. (n.d.). https://doi.org/10.5194/acp-26-13767-2026

Image Credits: AI Generated

DOI: 10.5194/acp-26-13767-2026

Keywords: hydroxyl radical, tropospheric chemistry, methane lifetime, atmospheric oxidation capacity, chemistry transport model, heterogeneous chemistry, water vapour UV absorption, carbon monoxide, nitrogen oxides, acetaldehyde emissions, ATom campaign, satellite observations