What Is the Global Methane Budget?

The global methane budget is an accounting of every source that puts methane into the atmosphere and every process that removes it. Current expert assessments place total global methane emissions in the range of roughly 580 to 900 teragrams per year (a teragram is a million metric tons), with more than three-quarters of that traceable to human activity or human-disturbed landscapes. The budget matters because methane is a powerful greenhouse gas, and the gap between how much enters the atmosphere and how much gets removed determines whether concentrations rise, hold steady, or fall.

How the Budget Adds Up

Think of the methane budget like a bathtub: sources are the faucet, sinks are the drain. When more methane flows in than drains out, the atmospheric level rises. A recent expert synthesis estimated total emissions at about 561 teragrams per year from sources that are moderately to fully influenced by humans, and roughly 174 teragrams per year from natural and minimally disturbed sources.¹ That split, roughly 76% anthropogenic and 24% natural, is one of the cleaner estimates available, though the boundaries between “natural” and “human-influenced” are genuinely blurry. A wetland that formed naturally but expanded because of upstream land-use changes is hard to categorize.

On the removal side, chemical reactions in the atmosphere destroy most of the methane each year. The best current estimates of that chemical sink fall in the range of roughly 470 to 610 teragrams per year, depending on the modeling approach and whether raw simulations or observation-corrected values are used.² Soils absorb a smaller additional amount. The persistent imbalance between total sources and total sinks is what drives the steady climb in methane concentrations that instruments have tracked since the early 2000s.

Wetlands, Lakes, and Rivers

Wetlands are by far the largest natural methane source on the planet. When organic material decomposes underwater without oxygen, microbes called methanogens produce methane as a metabolic byproduct. Multiple assessments agree that wetlands contribute roughly 30% of total global methane emissions from all sources combined, natural and human alike.³⁴ That makes them the single biggest source of uncertainty in the whole budget, because wetland area fluctuates with rainfall, temperature, and land management, and measuring emissions across millions of square kilometers of soggy ground is inherently difficult.⁵

Lakes and rivers add a significant amount on top of wetland emissions. A global observation-driven study estimated lake methane fluxes at about 42 teragrams per year, with tropical and subtropical lakes responsible for roughly half. Most of that methane escapes through bubbling from lake-bed sediments rather than quiet diffusion across the water surface.⁶ Rivers and streams contribute a comparable amount, estimated at roughly 28 teragrams per year globally, a figure that surprised researchers because running water had long been considered a minor player in the methane budget.⁷

Wildfires and Geological Seeps

Wildfires release methane when vegetation burns under oxygen-poor conditions. In a typical year, fire-related methane is a modest line item in the budget. But during extreme wildfire seasons, fire emissions can spike enough to cancel out years of emission reductions from anthropogenic sources in the affected regions.⁸ As wildfire seasons grow longer and more intense in many parts of the world, this source is getting harder to ignore.

Geological seeps, where methane escapes from underground reservoirs through faults, mud volcanoes, and the ocean floor, are another natural source, though their magnitude remains one of the more contentious numbers in the budget. Bottom-up estimates based on direct measurements at seep sites tend to be considerably higher than what radiocarbon analysis of ancient methane in ice cores would suggest.⁹ Submarine mud volcanoes around the world hold enormous quantities of methane locked in gas hydrates, though most of this stays trapped in sediment rather than reaching the atmosphere.¹⁰

Livestock and Rice Paddies

Agriculture is where the human fingerprint on the methane budget becomes most obvious. Livestock production is the largest single anthropogenic source of methane globally, dominated by enteric fermentation, which is a polite way of saying that cattle, sheep, and other ruminants belch methane as a natural byproduct of digesting their food.¹¹ The sheer number of ruminants on the planet, over a billion cattle alone, turns an ordinary digestive process into a climate-scale problem.

Rice paddies are the other major agricultural source. Traditional flooding practices create the same oxygen-free conditions found in natural wetlands, and methane-producing microbes thrive in submerged paddy soils. Rice cultivation accounts for an estimated 10 to 12% of global methane emissions. Newer water management techniques like alternate wetting and drying, which lets fields dry out periodically between floods, have cut emissions from participating farms by roughly 40%. Other approaches such as biochar application and mid-season drainage also show meaningful reductions.¹² The challenge is adoption: these practices require changes to deeply rooted farming traditions across Asia, Africa, and South America, and they sometimes come with trade-offs in water use and yield.

Oil, Gas, and Coal

The fossil fuel sector leaks methane at every stage, from extraction to processing to distribution. Natural gas is itself mostly methane, so any leak in a pipeline or at a wellhead is a direct methane emission. Oil production releases associated gas that is sometimes vented or flared. Coal beds contain trapped methane that escapes when seams are mined.

Measuring these emissions accurately has been a persistent headache. A landmark study of the U.S. oil and gas supply chain found that actual emissions were roughly 60% higher than what the Environmental Protection Agency’s official inventory estimated, largely because standard inventory methods miss releases during abnormal operating conditions like equipment malfunctions.¹³ A similar comprehensive synthesis of Canada’s oil and gas value chain also found broad agreement with updated national inventories only after those inventories incorporated atmospheric measurement data.¹⁴ The pattern is consistent: countries tend to undercount fossil fuel methane when they rely solely on engineering calculations rather than real-world atmospheric data.

Coal mining adds its own contribution, and the trajectory is not encouraging. Modeling of future coal mine emissions suggests that methane from active underground mines could quadruple by 2100 under some scenarios, while emissions from abandoned mines could increase eightfold, because abandoned mines keep leaking methane for decades after operations end.¹⁵ Even aggressive mitigation cannot fully stop abandoned mine emissions, which makes this a uniquely stubborn category.

Landfills and Wastewater

When organic waste decomposes in a landfill without access to air, the same anaerobic process that drives wetland and paddy emissions produces methane. Landfilling of solid waste accounts for roughly 60% of methane emissions from the waste sector, with wastewater treatment responsible for most of the remainder.¹⁶ Landfill gas capture systems can collect and burn this methane for energy, and many large landfills in wealthier countries already do so. But in lower-income countries, open dumping without gas capture remains common, and these uncontrolled sites are substantial emitters. Because waste generation is growing with urbanization and population growth, this sector’s contribution is likely to rise in the coming decades unless capture infrastructure expands alongside it.

How Methane Leaves the Atmosphere

The dominant drain in the methane budget is a chemical reaction with hydroxyl radicals in the lower atmosphere. Hydroxyl radicals, sometimes called the atmosphere’s “detergent,” are highly reactive molecules that break methane down into carbon dioxide and water over an average lifetime of about nine years. This reaction accounts for the vast majority of methane removal.¹⁷

Pinning down exactly how much methane the hydroxyl sink destroys is one of the harder problems in atmospheric chemistry. Different modeling approaches have produced estimates ranging from roughly 470 to over 600 teragrams per year, and the gap has real consequences for interpreting the budget. If the sink is large, then rising methane concentrations must be driven primarily by increasing emissions. If the sink is smaller or weakening, some of the rise could reflect the atmosphere becoming less efficient at cleaning itself. Research that corrected hydroxyl estimates using observed concentrations of precursor chemicals converged on a narrower range of about 470 to 510 teragrams per year for the 2000s, aligning with independent atmospheric inversion methods.² Soil bacteria that consume methane provide a smaller removal pathway, typically estimated at around 30 teragrams per year, though this number carries its own uncertainty.

Measuring Methane From the Ground and From Space

Scientists approach the methane budget from two directions. Bottom-up methods add together emission estimates from every known source category: counting cattle herds, measuring gas leaks at facilities, modeling wetland area from satellite imagery. Top-down methods use atmospheric measurements taken from ground stations, aircraft, or satellites and work backward to figure out what pattern of emissions would produce the observed concentrations. The two approaches should converge on the same answer. They often do not.

Geological methane is a clear example of the discord. Direct flux measurements at known seep sites, scaled up globally, imply substantially larger emissions than what radiocarbon analysis of methane in ice cores can accommodate.⁹ The disagreement matters for allocation: if natural geological emissions are smaller than bottom-up inventories suggest, then the anthropogenic share of the budget must be larger to explain what we observe in the atmosphere.

Satellites have transformed methane monitoring over the past decade. Instruments aboard several spacecraft can now map methane concentrations across the globe and identify individual “super-emitter” sites, facilities that release outsized amounts of methane, often from equipment failures or deliberate venting. One satellite system achieves 750-meter resolution and can revisit the same location multiple times per day, enabling near-real-time tracking of large emission events.¹⁸ Automated detection algorithms, including deep-learning approaches trained on satellite imagery, are making it possible to scan for methane plumes globally without relying on manual image inspection.¹⁹²⁰ The practical value is straightforward: you cannot fix leaks you do not know about, and satellite data has already revealed emissions far exceeding official self-reported numbers at numerous oil and gas facilities around the world.

Methane in Earth’s Deep Past

Ice cores drilled in Greenland and Antarctica have given researchers a methane record stretching back hundreds of thousands of years, and the picture is striking. Over the past 110,000 years, atmospheric methane has swung rapidly in lockstep with abrupt climate shifts called Dansgaard-Oeschger events, with concentrations jumping by 50 to 300 parts per billion within decades or less.²¹ These swings were driven by changes in Northern Hemisphere summer sunlight that expanded or contracted tropical wetlands, the same source category that dominates the natural budget today.

Dual isotope records from Antarctic ice cores have further clarified that tropical wetlands and seasonally flooded landscapes were the primary drivers of methane variability across glacial and interglacial cycles, while geological emissions appear to have played a steady but minor role.²² Isotope analysis of more recent atmospheric methane has also helped explain the plateau in methane concentrations that occurred in the early 2000s, suggesting it was linked to reduced emissions from the fossil fuel industry and possibly changes in the hydroxyl sink.²³ Since about 2007, concentrations have been climbing again, and the isotope signature points to biogenic sources, likely expanding tropical wetlands and growing livestock herds, as the main drivers of the renewed rise.

The Global Methane Pledge

In 2021, over 100 countries signed the Global Methane Pledge, committing to a 30% collective reduction in methane emissions from 2020 levels by 2030. A detailed assessment of the mitigation measures countries have outlined in their national climate plans found that, if every pledged measure were implemented to its maximum technical potential, the result would be roughly a 31% reduction in global methane emissions. That barely meets the pledge target and falls short of the approximately 40% cut that climate models indicate is needed to stay on a 1.5°C warming pathway.²⁴ The gap between “maximum technical potential” and what countries are actually doing is wide. Many national plans lack clear timelines, omit major emitting sectors entirely, or describe measures too vaguely to assess.

The economic case for action is stronger than the political momentum might suggest. A comprehensive assessment found that the benefits of global methane abatement outweigh the costs by a factor of at least six.²⁵ Methane’s relatively short atmospheric lifetime, about nine years compared to centuries for carbon dioxide, means that cutting emissions produces noticeable effects on warming within a decade or two. In the oil and gas sector specifically, mitigating methane leaks could avoid about 0.1°C of warming by 2050.¹⁴ That may sound small in isolation, but for a problem where every fraction of a degree matters, it represents one of the fastest levers available.