Air Pollution as a Climate Forcing: A Workshop

Day 1 Presentations

Global Methane Emissions: Historical Trends, Controlling Factors, and Future Prospects

Elaine Matthews
NASA Goddard Institute for Space Studies, New York, NY, U.S.A.

Introduction. In situ measurements of atmospheric methane concentrations, begun in the early 1980s show decadal trends, as well as large interannual variations, in growth rate (Dlugokencky, this volume). Recent research indicates that the decadal trend may be the combined effect of increasing sinks and stabilizing sources. We present an analysis of new ~20-year histories of annual, global source strengths for all major methane sources except for biomass burning which remains highly uncertain. We highlight trends, variability, drivers, and uncertainties, together with some conclusions about reconciling the source history with observations.

Natural Wetlands. Natural wetlands are the world's largest methane source, and the only one whose emissions are controlled almost exclusively by climate. The mean annual emission from wetlands averages about 30% of the global source strength of about 525 Tg (1 Tg =1012 g) over the last 20 years and shows large interannual variations (Walter et al., 2001). Wetlands have been shown to explain large positive and negative anomalies in interannual growth rates of atmospheric methane (Dlugokencky et al., 2001), but they do not explain the trend of declining growth rates in atmospheric concentrations.

Rice Cultivation. Rice cultivation accounts for ~10% of global methane emissions (Khalil, 2000). The trend in emissions over the last two decades is almost flat; small positive and negative anomalies in emissions reflect small interannual variations in area. Emissions closely track area, although production and yield have increased substantially during the last 20 years. For example, a 38% increase in yield drove the 44% increase in rice production between 1980 and 1998. In other words, increases in rice production were achieved by increasing yields via inputs of fertilizers, pesticides and technology. Because expanding irrigated rice cultivation (either to new lands or by increasing the number of rice crops grown annually on land currently planted to rice) is not possible, the flat trend in emissions is likely to continue although the ramifications for food production are very uncertain.

Ruminant Animals. Ruminant animals, which produce close to 20% of total annual methane emission, consist of bovines (cattle and dairy cows), goats, sheep, pigs, camels, and water buffalo, although bovines account for about 80% of methane emissions from ruminants. Historical emissions from ruminant animals show a slow and steady increase over the last 2 decades. This trend is the result of two developments. Animal populations in developed countries, where emissions per head tend to be large because large animals are fed high-quality feed, are slowing while those in developing countries, where per capita emissions tend to be lower due to smaller animals consuming lower-quality feed, are increasing. The global result is large increases in bovine populations and small increases in emissions. Thus, emissions are becoming progressively decoupled from absolute population growth of animals. This trend is likely to continue for several reasons. First, animals provide not only food and dairy products, but power, fuel, and fiber; therefore, projected increases in populations in developing countries will likely be accompanied by increases in animals. Secondly, no major anomalies in emissions and populations are expected since animal husbandry comprises a system of animals, provision of grazing and/or growing of fodder etc., and related uses of animals which together provide some inertia. Finally, there are no substantial reserves of alternative fuels, power, fiber, or food that would signal a major shift from the use of animals in the immediate future.

Landfills. Emissions from landfills (~5-8% of all methane emissions) (Khalil, 2000) are governed by human populations, quantity and quality of generated waste, landfilling practices, and landfill management including methane recovery. The trend in landfill emissions, similar to that of animals, is modest but steady increases throughout the period 1980-1998 (Bogner and Matthews, 2002). The increase in global emissions is the result of declining emissions in developed countries and increasing emissions in developing nations. The latter is increasing because of expanding populations, higher per capita waste generation, and increasing urbanization. Urban dwellers produce more waste per capita than do rural populations. In addition, the process of urbanization itself tends to promote increased rates of landfilling in response to health concerns posed by large amounts of mixed trash accumulating in areas of high population densities (Christensen et al., 1999; World Health Organization, 1993). While trends in waste generation and methane production within landfills are expected to increase in developing countries, emissions are likely to decline as those countries increase recovery efforts for methane which can be used for fuel as well as to offset the cost of building landfills.

Fossil fuels. Fossil fuels, comprising oil, natural gas, and coal, account for about 15% of global methane emissions (Khalil, 2000). Methane is released during multiple activities related to the exploration, production, transmission and consumption of oil and gas, and to the mining and processing of coal. Total methane emissions from fossil fuels increased more than 50% between 1980 and 1998, a proportional and absolute increase larger than for any other source (Matthews and Sarma, in prep, 2002). The decadal trend of these emissions was driven primarily by increases in both hard coal and oil/gas production and consumption, and secondarily by a gradual conversion from hard coal to oil and gas in some regions. However, substantial uncertainties remain with respect to high emission factors for countries whose production of oil and gas rose the most during this time period making it difficult to predict whether such conversions will result in larger or smaller methane emissions per heat unit of fuel in the future. In addition, trends in fossil fuel production and consumption are controlled by a large number of economic, political, and social factors whose influences vary over time periods from years to multiple decades. Finally, the potential for fuel conversion in individual countries relies on a variety of factors such as presence of indigenous fuel resources, fuel prices and availability, economic and development status, and availability of funds for infra-structure and other support necessary for conversion.

Summary and Prospects. Anthropogenic emissions, which account for about 70% of global annual emissions, are the only potential targets for mitigation. Several of these sources (rice cultivation and ruminant animals) exhibit flat trends over the last two decades and therefore potentially little opportunity for mitigation; landfills show modest increases; and fossil fuels have experienced relatively steep growth in the 1980s, and slower and somewhat variable growth in the 1990s. Various approaches and emission levels have been identified for mitigation efforts. For example, the Kyoto Protocol generally recommends that countries reduce methane emissions to about 6% below those of 1990, while the alternative scenario of Hansen et al. (2000) calls for future anthropogenic methane emissions to decline to ~70% those of 1990 to limit climate change. However, it is important to note that sources with flat or modest growth rates over the last several decades may represent either promising or diminished opportunities for mitigation.

While co-benefits accruing to reducing methane emissions have been identified by both science and policy researchers, characteristics unique to methane emissions indicate potential obstacles to mitigation. For example, rice is the staple food for about 2/3 of the worlds' people and, with respect to land use, is the most efficient way to produce food on a large scale. Periodic drainage and reflooding of fields is the dominant strategy suggested for reducing emissions from this source, and has been shown to be effective. At the same time, this approach requires sufficient water for reflooding as well as guaranteed provision of that water which can depend upon local or regional resource managers. Recent work (Matthews and Cambronero, in prep, 2002) has confirmed that water management regimes in rice fields have remained essentially stable over the last 20 years, confirming that at least high-emission flooded fields are not on the increase. Mitigation of emissions from ruminant animals has also been identified with co-benefits: since methane emission represents incomplete conversion of food intake to dairy or meat products, reduction of emissions is associated with higher productivity or lower feed requirements. However, controlling emissions has been researched and accomplished in management systems such as those in the US where feed and activity are carefully controlled and where amount (dairy and beef animals) and speed (beef animals) of product are maximized via expensive cash inputs. In most of the world, where animals are typically of value for a variety of uses and animal management systems are extremely variable and locally adapted, mitigation of emissions may prove difficult on a large scale. As noted above, emissions from landfills are increasing in rapidly growing and urbanizing developing countries. However, recovery of methane from landfills is standard practice in these countries; since the methane must be removed to minimize the danger of explosion, new landfill sites have recovery systems included in their design. Fossil fuels may offer the most promise for mitigation of emissions in the future. The largest emission of methane associated with fossil fuels occurs from venting, flaring and leakage during production of oil and natural gas, i.e., at point sources. Historical statistics confirm the decline of venting and flaring of methane directly to the atmosphere, and the increase of capture and reinjection of gas into oil wells to increase oil production. These trends are expected to continue for engineering as well as financial reasons. Unlike the point sources noted above, losses such as leakage during transmission and distribution of natural gas are much more dispersed and therefore may be expensive as well as difficult to identify and control.


  • 1. Bogner, J.E., and E. Matthews, Global methane emissions from landfills: A new approach to modeling historical trends, Global Biogeochem. Cycles, in review, 2002.
  • 2. Christensen, T.H., R. Cossu, and R. Stegman, Leachate, Gas, Operation, and Health Effects in Landfills, in Sardinia 99, Seventh international Waste Management and Landfill Symposium, pp. 731, Environmental Sanitary Engineering Centre, Cagliari, Italy, 1999.
  • 3. Dlugokencky, E.J., B.P. Walter, K.A. Masarie, P.M. Lang, and E.S. Kasischke, Measurements of an anomalous global methane increase during 1998, Geophys. Res. Lett., 28, 499-502, 2001.
  • 4. Hansen, J., M. Sato, R. Ruedy, A. Lacis, and V. Oinas, Global warming in the twenty-first century: An alternative scenario, Proc. Nat. Acad. Sci., 97, 9875-9880, 2000.
  • 5. Khalil, M.A.K., Atmospheric Methane: Its Role in the Global Environment, Springer-Verlag, Berlin, 2000.
  • 6. Matthews, E., and C. Cambronero, Methane emissions from rice cultivation: Validating and improving historical estimates, in prep, 2002.
  • 7. Matthews, E., and D. Sarma, Fossil fuels and the global methane cycle, in prep, 2002.
  • 8. Walter, B.P., M. Heimann, and E. Matthews, Modeling modern methane emissions from natural wetlands, 2. Interannual variations 1982-1993, J. Geophys. Res., in press, 2001.
  • 9. World Health Organization, The Urban Health Crisis: Strategies for Health for All in the Face of Rapid Urbanization, World Health Organization, Geneva, 1993.

Workshop Homepage * Background
Summaries: Overview, Gases, Aerosols, Tech., Health, Agri./Eco.
Abstracts: Day 1, Day 2, Day 3, Day 4, Day 5 * Participants