This page's content and links are no longer actively maintained. It is available for reference purposes only.
ICP Website Curator: Robert B. Schmunk — NASA Official: Gavin A. Schmidt


The Global Methane Cycle

(Note: This article was originally written in conjunction with the 1997 Global Methane Inventory.

1. Introduction to the Methane Cycle

Measurements of methane from Greenland and Antarctic ice cores indicate atmospheric concentrations of ~350 ppbv (parts per billion by volume) during the Last Glacial Maximum about 18,000 years ago, rising to 650 ppbv by about 200 years ago (Chappellaz et al., 1990). Researchers have estimated that natural methane sources totaled about ~180-380 Tg (1012 g) methane per year (Chappellaz et al., 1993). Wetlands were the dominant source with small contributions from wild fires, animals and oceans.

Since methane is chemically as well as radiatively active, atmospheric concentrations can increase because the terrestrial sources are increasing and/or because the sinks are declining. An important atmospheric sink for methane is the OH (hydroxyl) radical. The reaction of methane with OH radicals is the first step in a series of reactions which eventually leads to compounds that are readily removed from the atmosphere by precipitation or uptake at the surface. OH radicals also act as a chemical sink for other trace gases. For this reason, OH radicals are known as "the detergent of the atmosphere" (Crutzen, 1995).

During the last two hundred years, atmospheric methane concentrations have more than doubled to ~1800 ppbv and are still increasing. During the same period, the total annual emission of methane has increased to ~450-500 Tg (see Table above), about two times what it was during the pre-industrial period when natural sources dominated. Most of this increase in sources is due to the anthropogenic perturbation to the methane cycle, though climate variations may also contribute to changes in emission from wetlands and from wildfires.

The following sections provide a brief overview of the major methane sources -- background on what we know about them, the processes that produce methane, and what we still do not know (based on Matthews, 1993, 1994).

2. Natural Sources

Figure 2-1 below shows current estimates for individual sources. Although some uncertainties remain, the largest sources are natural wetlands, irrigated rice paddies, and domestic animals.

Global sources of methane
Figure 2-1: Global sources of methane

2.1 Wetlands

What we know:
  • Wetlands are most likely the largest natural source of methane to the atmosphere; their emissions are estimated to be about 100 Tg annually.
  • About 50% of the wetlands are peat-rich, temperature regulated northern wetlands and the remainder are low latitude systems dominated by precipitation and flood cycles.
  • Both flux measurements and large-scale modeling studies confirm the dominance of low-latitude wetlands and the smaller role of northern ecosystems in methane emissions under current climate conditions. Only about 30% of emissions are from peat-rich northern wetlands.
Processes that produce methane:
  • Methanogenic bacteria produce methane by anaerobic decomposition of organic materials.
  • Methane produced in the sediments is transported to the surface either through the water column (diffusion), through gas bubbles that rise from the sediments (ebullition) or via transport through the plants themselves.
  • Highly local controls such as temperature, topography, water table, and organic content as well as episodic events such as ebullition, degassing, hydrostatic pressure changes and wind have a large effect on methane fluxes.
  • Methane that is produced in water-logged soil sometimes moves upward through a drier surface soil and is oxidized, resulting in no methane emission.
  • There is general agreement concerning the global area and distribution of wetlands although uncertainties remain as to seasonal variations of wetland environments and dynamics of methane production periods.
  • The response particularly of high-latitude wetlands under a changing climate is highly uncertain; they may become larger or smaller methane sources or methane sinks.
  • Methane characteristics of most major wetland environments have been studied; measurements are still scarce for the wetlands of Russia which occupy about 25% of the global total, and for non-riverine tropical grasslands such as the Pantanal in Brazil.

2.2 Termites

What we know:
  • The source strength of termites has been estimated with a very large range, but recent estimates suggest emissions of 15-20 Tg/year.
  • The habitat distribution of the termite source is similar among several studies; termites are concentrated in tropical grasslands and forests.
Processes that produce methane:
  • Methane is produced by the activity of methane-oxidizing bacteria on the organic material consumed by the termites.
  • The source strength has been estimated with a very large range of 0-200 Tg methane per year.

2.3 Oceans

What we know:
  • The ocean source is very poorly known but considered minor. The emissions estimate is 10 Tg.
  • Coastal ocean regions exhibit higher and more variable concentrations.
Processes that produce methane:
  • Methane is produced at seepage areas in the seabed with organic-rich sediments (Judd, 2000).
  • Seabed flux rate estimates vary widely.
  • Losses to solution as bubbles rise to the sea surface are not known well.

2.4 Methane Hydrates

What we know:
  • Methane hydrates are rigid water cages surrounding methane molecules.
  • Hydrates are known or inferred to be found on the continental shelf at all latitudes.
  • Stability of hydrates requires high pressure and cold temperatures, meaning that most occur at depths and in regions insulated from climate change.
  • Methane source from hydrates is considered minor at present.
Processes that produce methane:
  • Hydrates are subject to destabilization from climate warming. Destabilization would lead to the release of the methane molecules.
  • The dispersed nature of this source makes it especially difficult to evaluate.
  • The large pool of methane in gas hydrates implies that a small perturbation under a changing climate could produce a considerable source of methane.

3. Anthropogenic Sources

3.1 Rice Cultivation

Photo of rice cultivation
What we know:
  • Rice production may account for 10% (~30-60 Tg) of the total annual methane emission.
  • Since 1980, rice production has risen by over 40% through the combined effects of increased harvest areas and higher yields.
  • Over 90% of the harvested area is confined to Asia.
Processes that produce methane:
  • Methanogenic bacteria produce methane by anaerobic decomposition of organic materials.
  • A large number of factors affect the production, transport, and efflux of methane in the flooded rice fields, among them temperature, water status, fertilizer application, soil properties, and plant phenology.
  • Methane that is produced in water-logged soil sometimes moves upward through a drier surface soil and is oxidized, resulting in no methane emission.
  • As much as 90% of methane produced in sediments may oxidize during transport to the water surface.
  • Information on local soil temperatures, type and application rate of fertilizers (especially non-commercial organics), seasonal and annual variations in water status, and soil chemistry is not readily available.
  • Even with sufficient data on the factors mentioned above, considerable uncertainties remain in quantitative relationships between methane flux and these factors, though they have been significantly reduced recently.

3.2 Domestic Animals

What we know:
  • Animals contribute about 80 Tg methane per year.
  • About half of the global emission come from India, China, the former USSR, USA, and Brazil.
  • Non-dairy cattle and dairy cows together contribute about 75% of the total methane source from animals; the remainder is from water buffalo, sheep, goats, pigs, camels, and horses.
  • Emissions from animals may be one of the better known sources in the methane budget because statistics on animal populations in developed countries are reasonably reliable.
Processes that produce methane:
  • Methane production from animals results from fermentation of carbohydrates in the rumen (stomach containing microbes capable of breaking down cellulose).
  • Production is affected by factors such as quantity and quality of feed, body weight, age, and activity level; therefore, it varies among animal species as well as among individuals of the same species.
  • Significant uncertainties exist with respect to population statistics for less developed countries.

3.3 Fossil Fuel

What we know:
  • Methane is the major component of coal gas and natural gas.
  • Methane emissions associated with fossil fuel sources range from ~16 to 24% of the total source strength equal to about 80-120 Tg.
Processes that produce methane:
  • Natural gas and coal gas both consist almost entirely of methane.
  • Fossil fuel sources of methane include coal mining and processing, as well as gas exploration, production, transmission, and distribution.
  • Methane released to the atmosphere during mining and processing of coal is associated directly with coal removed from mines as well as with releases from coal left in the mine in overlying and underlying seams.
  • Methane emissions associated with natural gas production and consumption include losses during extraction, venting and flaring at oil and gas wells, and losses during processing, transmission and distribution.
  • Most of the variation among estimates of methane released to the atmosphere during mining and processing of coal results from inclusion and/or exclusion of processes or particular coal products, and from differences in the emission factors used.
  • Natural gas transmission losses are the difference between gas purchased for delivery and gas sold, differences which may be due to theft and metering errors as well as to actual leaks.
  • The geographically- and sectorially-dispersed nature of the sources of methane emission from natural gas transmission makes direct estimates difficult.
  • Time series estimates using constant emission factors do not take into account changes in technology which would affect the amount of methane emitted.
  • Lack of data on venting and flaring of natural gas (collected by oil and gas companies but not available publicly) is a major source of uncertainty.

3.4 Biomass Burning

What we know:
  • Recent estimates indicate annual methane releases of ~10-50 Tg from biomass burning.
Processes that produce methane:
  • Methane is released when vegetation is burned. The amount is a function of burning technique and temperature, moisture and carbon content of the vegetation, amount and type of vegetation burned etc.
  • The contribution of biomass burning to the global emission of methane is highly uncertain due to the innate variability of the process itself as well as to severe data limitations.
  • A considerable portion of burning takes place in association with poorly documented agricultural activities in the tropics.
Photo of landfill

3.5 Landfills

What we know:
  • One early global estimate of methane emission from landfills gave a range of 30-70 Tg/year, although recent estimates suggest that the landfill methane source may be more in the range of ~25 Tg.
Processes that produce methane:
  • Decomposition of biodegradable organic material in landfills produces both carbon dioxide and methane.
  • Uncertainties remain with respect to the magnitude and composition of waste production and the fraction of waste placed in landfills.
  • There is very high variability in local factors such as climate, age of refuse, and landfill design, construction, and management which affect the amount of methane produced, consumed and emitted in these sites.

4. Summary

Methane sources under anthropogenic control currently account for approximately 70% of the total annual emission. Several of these (e.g., animals, rice cultivation, energy-related sources) may be prone to future increases due to demands of increasing human populations. The magnitude of the remaining natural sources, dominated by wetlands, is relatively well known. Currently, northern wetlands contribute about one-third of the world's total wetland emissions while the tropics account for most of the remainder. However, wetland response to climate change predicted for the next century is highly uncertain. Depending on local interactions among temperature, water status, nutrients etc., wetland ecosystems may become larger or smaller methane sources, or even methane sinks. A varied collection of additional sources such as volcanoes, oceans, seabed seepage, gas hydrates, and peat mining are highly uncertain but considered minor, probably totaling of ~20 Tg.

Currently, we understand a great deal about the processes that produce methane from various sources as well as the distribution of many of the sources such as wetlands or animals. In addition, we now have a 15 year record of measurements of atmospheric methane from a growing network of ground stations. However, the atmospheric record shows large and variable patterns in annual increases in atmospheric concentrations and uncertainties remain with respect to year-to-year variations in methane emissions from various sources. For example, short-term variations in temperature and precipitation can affect emissions from biological sources such as rice paddies and wetlands; political and economic changes can affect levels of industrial activity and consumption of fossil fuels; management practices can affect emissions from domestic animals and rice paddies; and development and recycling trends can affect methane emissions from landfills and from wastewater.

The challenge is to improve the collection of tools we have to understand the methane cycle and its variations over time. These include:

  1. atmospheric chemistry models that simulate chemical reactions and transport in the atmosphere,
  2. measurements of the atmospheric concentrations of methane and related gases,
  3. field measurements that quantify the influence of various factors on emission from sources (for example, how do methane emission from a northern wetland change when the summer is hotter than usual? wetter than usual?)
  4. inventories of the sources of methane and associated emissions at intervals of about 5 years beginning in the early 1980s and continuing in the future.

The methane inventory project described in the following section is a contribution to the last goal listed above. These global inventories are developed in part using field measurements (2 above) and used in the atmospheric chemistry models (1 above). Results from the models are evaluated by comparing them to the measurements of atmospheric methane (2 above).