The greenhouse effect and the role of methane from livestock
The greenhouse effect warms Earth's surface as atmospheric gases absorb and re-emit infrared radiation, and methane released through livestock digestion and manure contributes to that warming. Over the industrial era the added carbon dioxide accounts for 2.16 watts per square metre of effective radiative forcing and methane for 0.54, within a total human contribution of 2.72 watts per square metre (Forster et al., 2021).
This explainer sets out how the effect works, which gases carry it, why methane is treated separately from carbon dioxide, and how much of the methane in the air is traced to farmed animals. It also separates the parts that rest on direct measurement from the parts that depend on inventories, models and the choice of accounting metric.
What the greenhouse effect is and how it works
The surface of the Earth is warmed by sunlight and loses that energy again as infrared radiation. Several gases in the atmosphere absorb part of this outgoing radiation and re-emit it in all directions, including downwards, so the surface and the lower atmosphere settle at a higher temperature than they would under a transparent atmosphere (Forster et al., 2021).
The familiar name is a poor description of the physics. A garden greenhouse works mainly by stopping warm air from mixing away, while the atmospheric effect works through absorption and re-emission of radiation at particular wavelengths. The outcome is a change in the energy balance at the top of the atmosphere, not a lid over the planet.
The natural greenhouse effect can be described in terms of surface temperature and infrared energy flows. Effective radiative forcing, by contrast, measures the change in the energy balance after a perturbation such as increased greenhouse gas concentrations, in watts per square metre (Forster et al., 2021).
How a 1967 climate model quantified the effect
The modern estimate starts with a one-dimensional radiative-convective model. Manabe and Wetherald (1967) held relative humidity fixed and found that doubling the carbon dioxide content of the atmosphere raised the temperature of the air column by about 2 degrees Celsius.
That calculation also showed why water vapour matters so much. In a warmer column the air holds more vapour at the same relative humidity, and because vapour is itself a strong absorber, the initial change is amplified (Manabe and Wetherald, 1967).
The figure was a result for an idealised column rather than a forecast for the planet, and the authors presented it as such. Its importance is that the mechanism could be quantified at all, and that the order of magnitude has survived six decades of more detailed work.
What has been measured directly
Satellite spectra give the clearest observational check. Harries and colleagues (2001) compared the outgoing longwave spectrum recorded by the IRIS instrument on Nimbus 4 in 1970 with the IMG spectrum from the ADEOS satellite in 1997, using the same April to June window to avoid seasonal artefacts, and found differences consistent with higher concentrations of methane, carbon dioxide, ozone and two chlorofluorocarbons.
The same signal has been measured from the ground. Feldman and colleagues (2015) used spectra from two sites, the Southern Great Plains and the North Slope of Alaska, and found a surface radiative forcing trend attributable to carbon dioxide of 0.2 watts per square metre per decade across 2000 to 2010, alongside a rise of 22 parts per million in concentration.
Both results measure the mechanism rather than the warming. They show that the atmosphere is intercepting more infrared radiation at the wavelengths these gases absorb, which is a separate observation from the surface temperature record and is not a substitute for it (Harries et al., 2001; Feldman et al., 2015).
Which gases produce the greenhouse effect
The present-day effect is not mostly carbon dioxide. Using a radiation model to switch absorbers off one at a time, Schmidt and colleagues (2010) attributed about 50% of the effect to water vapour, 25% to clouds and 20% to carbon dioxide, with all other absorbers playing minor roles.
That allocation is often quoted as an argument that carbon dioxide hardly matters, which the paper itself warns against. Water vapour is controlled by temperature and leaves the atmosphere within days, so it responds to a change rather than causing one, while the long-lived gases are added directly and set the level the vapour responds to (Schmidt et al., 2010).
The long-lived gases have risen by measured amounts. Carbon dioxide went from 278.3 parts per million before industrialisation to 409.9 in 2019, methane from 729 parts per billion to 1,866, and nitrous oxide from 270.1 to 332.1 (IPCC, 2021).
Converted into radiative terms for 1750 to 2019, carbon dioxide accounts for 2.16 watts per square metre, reported range 1.90 to 2.41. Methane accounts for 0.54, reported range 0.43 to 0.65, and nitrous oxide for a further 0.21 (Forster et al., 2021).
A second figure for methane circulates alongside that one and causes confusion. Counted by emissions rather than by concentration, which also credits methane with the ozone and stratospheric water vapour it produces as it breaks down, methane accounts for about 31% of the additional forcing from human emissions, or 1.19 of 3.84 watts per square metre (Saunois et al., 2025).
The two numbers are answers to different questions rather than rival estimates. One asks what the methane now in the air is doing, the other what the emissions of methane have set in motion across the whole chemistry of the atmosphere (Forster et al., 2021; Saunois et al., 2025).
Why methane behaves differently from carbon dioxide
Methane is removed from the atmosphere by oxidation, with a perturbation lifetime of 11.8 years plus or minus 1.8, whereas a pulse of carbon dioxide is drawn down over centuries and millennia (Forster et al., 2021). The practical consequence is that stabilising methane emissions stabilises its concentration, and therefore its forcing, within a few decades (Saunois et al., 2025).
Per tonne, methane is also the stronger absorber, which is why the two gases are compared through global warming potential. Over 100 years one tonne of methane from a non-fossil source is rated at 27.0 tonnes of carbon dioxide equivalent, and fossil methane at 29.8, because the carbon in the latter adds to the long-lived pool as it oxidises (Forster et al., 2021).
Shorten the horizon and the same gas looks different. Over 20 years the figures are 79.7 and 82.5, and the only thing that has changed is the period over which the effect is added up (Forster et al., 2021).
Published figure for methane | Value |
Perturbation lifetime | 11.8 ± 1.8 years |
Global warming potential, 20 years, non-fossil | 79.7 ± 25.8 |
Global warming potential, 100 years, non-fossil | 27.0 ± 11 |
Global warming potential, 100 years, fossil | 29.8 ± 11 |
Forcing by concentration, 1750 to 2019 | 0.54 W/m2, against 2.16 for CO2 |
Forcing by emissions, 1750 to 2019 | 1.19 W/m2, about 31% of 3.84 |
Global mean concentration, 2025 | 1,936.03 ppb, against 729 in 1750 |
Note: these are metric values and measured forcings for the gas itself, not a measure of how much any sector is to blame. The lifetime figure describes how long a change in emissions takes to work through the atmosphere rather than how long a single molecule survives (Forster et al., 2021; Saunois et al., 2025; IPCC, 2021; NOAA, 2026).
Because one number has to stand for both the strength and the duration of the effect, the choice of horizon is an accounting decision. Lynch and colleagues (2020) set out an alternative called GWP*, which reports warming-equivalent emissions by treating a change in the rate of short-lived emissions as comparable to a one-off release of carbon dioxide.
On their comparison the conventional 100-year metric diverges most from modelled warming when methane emissions are steady or falling. A constant flow of a short-lived gas sustains a roughly constant level of warming rather than adding to it year on year, which the standard metric does not capture (Lynch et al., 2020).
The alternative has not replaced the standard. Reporting under the climate convention uses the 100-year potential endorsed at COP27, and the FAO assessment of the sector follows it, which is why most published livestock totals are in those units (FAO, 2023).
How much methane comes from livestock
The atmospheric accounts put the sector in context. For 2010 to 2019 the global methane budget gives total emissions of 575 million tonnes a year from atmospheric inversions, reported range 553 to 586, of which 369 million tonnes, about 65%, are attributed to direct human sources (Saunois et al., 2025).
Within that human total, enteric fermentation and manure management together account for 112 million tonnes of methane a year, reported range 107 to 118. The compilers of the budget describe that as about a third of anthropogenic methane emissions (Saunois et al., 2025).
Natural sources remain comparable in size, which is worth keeping in view. The bottom-up budget estimates 159 million tonnes a year from wetlands alone and 248 million tonnes from wetlands plus inland freshwaters after a correction for double counting. The top-down estimate of 165 million tonnes is labelled as wetlands, but the authors caution that it may already include freshwater emissions (Saunois et al., 2025).
The growth in human methane has not come from one sector. Jackson and colleagues (2020) found that the rise of about 40 million tonnes a year by 2017, compared with the early 2000s, was divided roughly equally between agriculture with waste and fossil fuels. Agriculture drove the increase in Africa, south Asia and Oceania, while fossil fuels drove it in China and North America (Jackson et al., 2020).
Expressed in carbon dioxide equivalent, the FAO assessment of livestock agrifood systems gives 6.2 billion tonnes for the reference year 2015. That is about 12% of an estimated human total of 50 to 52 billion tonnes, and cattle farming carries 62% of the sector figure (FAO, 2023).
The split between the animals and the supply chain is roughly three to two. Direct emissions from the animals, meaning enteric methane plus methane and nitrous oxide from manure, come to 3.7 billion tonnes or about 60%, while the remaining 2.6 billion tonnes are indirect emissions associated with feed production, fertiliser manufacturing, land-use change, energy use, processing and transport (FAO, 2023).
Methane is the largest single component. It carries 54% of the sector total in those units, against 31% for carbon dioxide and 15% for nitrous oxide, which is why discussion of livestock and climate keeps returning to the one gas (FAO, 2023).
An earlier assessment of the same sector reached a higher number. Gerber and colleagues (2013) estimated 7.1 billion tonnes of carbon dioxide equivalent a year for the 2005 reference period, attributing 45% to feed production, processing and transport, 39% to enteric fermentation and 10% to manure storage and processing, with cattle contributing 4.6 billion tonnes or 65% of the sector.
That report also produced the widely quoted 14.5% share. How the figure was derived matters: the 2005 livestock estimate was set against a global total of 49 billion tonnes for 2004, so the percentage mixes two reference years (Gerber et al., 2013). The conclusions that assessment drew for itself are summarised in the account of what the global livestock assessment found.
How such sector estimates are built, and what their published ranges actually cover, is set out at length in the review of livestock emission estimates. The ranges carry as much information as the central values, because each total depends on the reference year, the model version and the accounting rules behind it.
Why published estimates of livestock emissions differ
The move from 7.1 to 6.2 billion tonnes is not a measured decline in emissions. The FAO states directly that the newer figure is lower than previous estimates from the same model but that direct comparisons are not feasible, because the methodology, the input data and the global warming potential values all changed (FAO, 2023).
The metric change alone moves the total. Successive versions of the model used a 100-year potential for methane of 25, then 34, then 27, following the fourth, fifth and sixth assessment reports, so the same tonnage of methane can shift the sector total by roughly a quarter without any change on the ground (FAO, 2023).
The two families of methane estimate are built in opposite directions. Inventories work bottom up from animal numbers, feed and emission factors, while inversions work top down from measured concentrations, and for 2010 to 2019 the bottom-up total of 669 million tonnes a year sits well above the inversion figure of 575 (Saunois et al., 2025).
System boundaries matter as much as method. A life-cycle total includes feed, fertiliser manufacture, transport and processing, while a national inventory of enteric fermentation counts only the animals, so two correct figures for the same herd can differ by a factor approaching two (FAO, 2023; Gerber et al., 2013).
Timing is the other hidden assumption. When a flow of emissions is compared with a one-off release, the horizon chosen decides the answer, and the same question arises for biomass energy, as the debate over whether boreal forest bioenergy is carbon neutral shows.
None of this makes the figures arbitrary. It means a reader comparing two totals should check the reference year, the gases included, the system boundary and the metric before treating the difference as news (FAO, 2023).
What is proven about livestock methane and what remains uncertain
Several points are settled. The mechanism has been observed from space and from the ground (Harries et al., 2001; Feldman et al., 2015), methane has risen from 729 parts per billion before industrialisation to a global mean of 1,936.03 in 2025 (IPCC, 2021; NOAA, 2026), and ruminant digestion with manure accounts for roughly a third of human methane emissions (Saunois et al., 2025).
Neither headline describes the recent trend. Annual growth in the global mean fell from 17.69 parts per billion in 2021 to 4.77 in 2025, which is a slowdown in the rate of increase and not a decline in concentration (NOAA, 2026).
Methane from farmed animals accounts for a substantial estimated share of human methane emissions (Saunois et al., 2025). What remains contested is how that share should be weighed against carbon dioxide and how much of it can be reduced in practice.
The first open question is the metric. The 100-year potential remains the reporting standard (FAO, 2023), while the warming-equivalent approach gives a closer match to modelled temperature when methane emissions are flat or falling, and the two framings lead to different conclusions about what a stable herd contributes (Lynch et al., 2020).
The second is how much can be cut. A meta-analysis of 430 studies evaluating 98 mitigation strategies identified five approaches that reduced daily methane emissions by an average of 21% and three approaches that reduced methane per unit of meat or milk by an average of 12% (Arndt et al., 2022).
Those effects are measured under trial conditions, and the authors are explicit about what follows. Ruminant methane would have to fall by 11% to 30% by 2030 and by 24% to 47% by 2050 against 2010 levels to stay consistent with a 1.5 degree pathway, and full adoption of the most effective strategies meets the 2030 figure but not the 2050 one, because projected growth in demand for milk and meat offsets the gains (Arndt et al., 2022).
The third uncertainty sits in the policy arithmetic. The assessment by UNEP and the Climate and Clean Air Coalition attributes about 30% of warming since pre-industrial times to methane and finds that cutting human-caused methane by up to 45% this decade would avoid nearly 0.3 degrees of warming by 2045 (UNEP and CCAC, 2021).
That scenario covers every source of methane, from gas infrastructure to landfill, so it cannot be read as a result for livestock on its own. The honest summary is that the sector's share of methane is reasonably well established, while its achievable contribution to any temperature target depends on choices of metric, boundary and adoption rate that are still being argued (Lynch et al., 2020; Arndt et al., 2022; UNEP and CCAC, 2021).
Frequently asked questions
Is methane a stronger greenhouse gas than carbon dioxide?
Per tonne it absorbs far more energy, which is why one tonne is rated at 27.0 tonnes of carbon dioxide equivalent over a century and at 79.7 over twenty years. The comparison depends entirely on the period chosen, because methane is removed from the air while carbon dioxide largely is not (Forster et al., 2021).
How long does methane stay in the atmosphere?
A change in methane emissions works through the atmosphere over a perturbation lifetime of about 11.8 years, so concentrations respond to emission trends within decades. That is different from the fate of an individual molecule, and it is far shorter than the centuries over which added carbon dioxide is drawn down (Forster et al., 2021).
Which farmed animals account for most livestock emissions?
Cattle dominate the published totals. The most recent global assessment attributes 62% of livestock emissions to cattle farming, and the earlier assessment put the figure at 65%, splitting it between beef at 41% of the sector and milk at 20%, with pigs, poultry and small ruminants far behind (FAO, 2023; Gerber et al., 2013).
Do cows emit more methane than wetlands?
Wetlands and freshwater ecosystems exceed livestock methane emissions in the bottom-up budget. Enteric fermentation and manure management account for 112 million tonnes annually, compared with 159 million tonnes from wetlands alone and 248 million tonnes from wetlands plus inland waters. The top-down wetland category is 165 million tonnes but may also capture freshwater emissions (Saunois et al., 2025).
Is the amount of methane in the air still increasing?
Yes, although more slowly than at the start of the decade. The global mean rose by 17.69 parts per billion in 2021 and by 4.77 in 2025, so the concentration continues to climb each year while the rate of climb has fallen, which are two separate statements about the same record (NOAA, 2026).
Sources
- Forster P, Storelvmo T, Armour K, et al. The Earth's energy budget, climate feedbacks, and climate sensitivity. In: Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, Chapter 7. Cambridge University Press; 2021:923–1054. DOI 10.1017/9781009157896.009.
- Intergovernmental Panel on Climate Change (IPCC). Annex III: Tables of historical and projected well-mixed greenhouse gas mixing ratios and effective radiative forcing of all climate forcers. In: Climate Change 2021: The Physical Science Basis. Cambridge University Press; 2021:2139–2152. DOI 10.1017/9781009157896.017.
- Schmidt GA, Ruedy RA, Miller RL, Lacis AA. Attribution of the present-day total greenhouse effect. Journal of Geophysical Research: Atmospheres. 2010;115:D20106. DOI 10.1029/2010JD014287.
- Manabe S, Wetherald RT. Thermal equilibrium of the atmosphere with a given distribution of relative humidity. Journal of the Atmospheric Sciences. 1967;24(3):241–259. DOI 10.1175/1520-0469(1967)024<0241:TEOTAW>2.0.CO;2.
- Harries JE, Brindley HE, Sagoo PJ, Bantges RJ. Increases in greenhouse forcing inferred from the outgoing longwave radiation spectra of the Earth in 1970 and 1997. Nature. 2001;410(6826):355–357. DOI 10.1038/35066553.
- Feldman DR, Collins WD, Gero PJ, Torn MS, Mlawer EJ, Shippert TR. Observational determination of surface radiative forcing by CO2 from 2000 to 2010. Nature. 2015;519(7543):339–343. DOI 10.1038/nature14240.
- National Oceanic and Atmospheric Administration (NOAA), Global Monitoring Laboratory. Trends in Atmospheric Methane (CH4). 9 October 2026. Accessed 11 November 2026. https://gml.noaa.gov/aggi/aggi.html
- Saunois M, Martinez A, Poulter B, et al. Global Methane Budget 2000–2020. Earth System Science Data. 2025;17:1873–1958. DOI 10.5194/essd-17-1873-2025.
- Gerber PJ, Steinfeld H, Henderson B, et al. Tackling climate change through livestock: a global assessment of emissions and mitigation opportunities. Food and Agriculture Organization of the United Nations; 2013. archive record 20133417883
- Food and Agriculture Organization of the United Nations (FAO). Pathways towards lower emissions: a global assessment of the greenhouse gas emissions and mitigation options from livestock agrifood systems. Rome; 2023. DOI 10.4060/cc9029en.
- Lynch J, Cain M, Pierrehumbert R, Allen M. Demonstrating GWP*: a means of reporting warming-equivalent emissions that captures the contrasting impacts of short- and long-lived climate pollutants. Environmental Research Letters. 2020;15(4):044023. DOI 10.1088/1748-9326/ab6d7e.
- Arndt C, Hristov AN, Price WJ, et al. Full adoption of the most effective strategies to mitigate methane emissions by ruminants can help meet the 1.5 °C target by 2030 but not 2050. Proceedings of the National Academy of Sciences. 2022;119(20):e2111294119. DOI 10.1073/pnas.2111294119.
- United Nations Environment Programme (UNEP) and Climate and Clean Air Coalition (CCAC). Global Methane Assessment: benefits and costs of mitigating methane emissions. Nairobi; 2021.
- Jackson RB, Saunois M, Bousquet P, et al. Increasing anthropogenic methane emissions arise equally from agricultural and fossil fuel sources. Environmental Research Letters. 2020;15(7):071002. DOI 10.1088/1748-9326/ab9ed2.
Related
Livestock and Climate Change: Why Emission Estimates Differ
Tackling Climate Change Through Livestock: Emissions and Mitigation
Is Burning Boreal Wood Carbon Neutral? What a 2011 Model Showed
What is biodiversity and why it matters for food and farming
Invasive species and why some newcomers take over ecosystems
The nitrogen cycle and how microbes move nitrogen through soil