In the early 1990’s, in defiance of IPCC projections, the methane concentration in the atmosphere abruptly stopped rising, and has remained nearly constant since then. Methane is a crouching tiger in the carbon cycle, with potentially enough available as hydrates and from peats to really clobber the Earth’s heat budget. The big question is, will atmospheric methane start rising again?
Climate impact of methane release
The climate impact of methane differs from that of CO2 in that methane is a transient gas, while CO2 accumulates. The climate impact of methane release depends on whether it’s released quickly or slowly, relative to the methane lifetime.
If it’s released quickly, over just a few years or less, there would be a decade-timescale warming spike, followed by a recovery toward the lesser warming from the CO2 that the methane would oxidize into. The amount of available methane is staggering. If just 10% of the ocean hydrate reservoir were to escape to the atmosphere within a few years, it would be the radiative equivalent of a ten times increase in atmospheric CO2, truly catastrophic.
On the long term (longer than a few decades) the transient methane concentration is determined by the chronic rate of methane emission to the atmosphere. A higher concentration in the atmosphere accelerates the overall rate of methane oxidation, to balance the greater input. More methane molecules are standing in line to compete for the limiting supply of the reaction catalyst molecule, OH radical. The oxidation product of the methane, CO2, builds up enough to impact the climate as well. In the long term, the radiative forcing from the accumulating CO2 may exceed that of the transient methane concentration [Archer and Buffett, 2005].
There have been two methane sources in the newspapers in the past few years. One is plants, measured recently [Keppler et al., 2006] to give off prodigious amounts of methane. If this is true, it might imply a small increase in chronic emissions in the future. The seasonal cycle of atmospheric CO2, the breathing of the biosphere, has been getting deeper over the years. If the biosphere is breathing deeper, maybe it’s also farting more.
The other is peats, on which there is a voluminous literature. Peat methane fluxes are notoriously patchy and difficult to generalize, but there are a few things you can depend on. One is that water makes a huge difference; a wet peat soil will emit methane while the same soil, dried, would actually consume methane. Two, there seems to be a reproducible methanogenesis poisoning effect by sulfuric acid deposition (acid rain), caused by the stimulation of sulfate-reducing bacteria displacing the methanogens. Three, melting starts things cooking. Fourth, peats release much more carbon as CO2 than as methane, and their strongest radiative impact will probably be from the CO2.
Methane hydrates are the giant reservoir on Earth. It takes a long time to warm the deep ocean and the clathrate zone, and no one has proposed a mechanism for getting much methane release from hydrates in the coming century. The only place where melting methane hydrates appear to be releasing methane to the atmosphere is on the Siberian margin, where hydrates associated with the permafrost relict from the last glaciation release methane to the shallow water column of the shelf waters.
Industrial emission of methane declined a bit with the collapse of the USSR in 1989 and thereafter. Industrial methane emission arises from leaks, more difficult to pin down than the amounts of deliberately released gases such as CO2. Agricultural emission, from rice farming and ruminant animals, is not so easy to quantify either, but we’ll leave a description of that to the reader’s imagination.
Methane degrades to CO2 by reaction with OH radical in the atmosphere. OH is produced by photochemical reaction with compounds such as ozone and NOx, the concentrations of both of which have been altered by industrial activity. OH reacts with methane and carbon monoxide (CO), the concentrations of both of which have also changed. The chemistry of the atmosphere is analogous to the chemistry in a candle flame, which can burn faster or slower depending on the conditions, for example how quickly it is primed by reactive NOx emission or solar UV (in the case of the atmosphere) or how long the wick is (in the case of the candle). The lifetime of methane in the atmosphere seems like it could quite easily be altered by human activity.
It is impossible to measure the global inventory of OH directly because its concentration is so low and so variable. The tracer for the global OH inventory with the longest pedigree is methyl chloroform, CH3 CCl3. Prinn et al  diagnosed 10% changes of OH concentration on decadal timescale, but Krol and Lelieveld  claim that this result is sensitive to small changes in the assumed source function of methyl chloroform. If the stuff were hoarded a bit before the onset of the Montreal Protocol banning its use, the OH change diagnosed by Prinn could go away. Emission of methyl chloroform stopped in 1992, so the signal now comes down to the decay time constant of the atmospheric concentration. This is complicated by other uptake fluxes such as invasion into the ocean [Wennberg et al., 2004].
The concentration of 14C in CO serves as another tracer for the OH inventory [Manning et al., 2005]. 14C is produced naturally, by cosmic ray neutrons impacting nitrogen gas, and it quickly oxidizes to 14CO, then after a few months to 14CO2. Changes in the solar magnetic shielding of the Earth can affect the production rate of 14CO, requiring a correction, and exchange with the stratosphere is important, but the competing source / sink problems do not appear to be as severe as they are for methyl chloroform. The lifetime of CO in the atmosphere is much shorter than that of methyl chloroform, making the 14CO concentration much more sensitive to, and diagnostic of, month-to-year timescale variability in OH. 14CO varies by factor of two or more over the seasonal cycle, whereas methyl chloroform only varies by a few percent. Manning et al  found no long-term trend, but short-term 10% variations from Pinatubo and Indonesian fires. Pinatubo brought a 12% decrease in solar UV flux [Dlugokencky et al., 1996], decreasing OH, while fires, in particular the Indonesian fire in 1991, bring an increase in CO and CH4 emissions, which can also deplete OH [Butler et al., 2005]. One gets a picture of a volatile but self-stabilizing OH cycle, a flickering flame.
Putting them together
One clue that might help unravel past changes in methane sources is that the rate of atmospheric increase of several gases all correlate. Langenfelds et al  found that CO2, 13C, H2, CH4, and CO growth rates all march in step with the Southern oscillation from 1992-99. Simmonds et al  find similar correlations in data from 1996-2003. Both authors point to fires as a potential common source, as opposed to wetland emissions (which don’t produce H2, for example).
Another seemingly useful clue is that, during a period of methane doldrums (no rise) from 1999-2002, the N/S gradient of methane relaxed a bit [Dlugokencky et al., 2003], suggesting that the doldrum was due to a decline in a methane source in the northern high latitudes. Since most fires burn in the tropics, rather than in the high latitudes, this clue would seem to be pointing us toward wetlands, i.e. in a different direction than clue #1.
A recent paper [Bousquet et al., 2006] attempts to bring all these pieces together into an inverse calculation of the methane sources. This is not a fundamentally new approach, but it does have the advantage of including the most recent several years of data, as methane stubbornly continues to refuse to rise. Changes in OH concentration are diagnosed from methyl chloroform. The spatial pattern of CH4 variations, plus 13CH4 data, provides the basis for partitioning the methane changes among the various sources and sinks.
Their conclusion is that rising human emission since 2000 has been masked by a probably temporary natural decline in wetland emission. Their diagnosed source fluxes are consistent with bottom-up models of wetlands and fires, and independent fossil fuel emissions estimates. But I have to wonder what they’d get if they considered some of the other trace gases mentioned above, such as CO, 14CO, and H2.
What are the implications of all this for our ability to predict the future of the methane cycle? Let’s summarize what you’ve just read. According to one set of papers, atmospheric methane could be suppressed in the future by controlling land fires. Or it could be that methane variations are mostly produced by wetland emission, driven by climate change as well as land use decisions, according to another set of papers. Or methane could resume its rise, toward a new steady state, because it is driven by increasing fluxes from melting permafrost peat and hydrates, according to observations on the ground.
The bottom line I take away from all this is that the available studies come to a strikingly divergent range of conclusions. We know a lot about the methane cycle, but as far as forecasting the near-term future, we have no clue. No one would build a nuclear reactor if our understanding of the underlying chemical dynamics were as fragile as this. Instead, we are taking the reins of a planetary biosphere. This is disturbing.
Archer, D.E. and Buffett, Time-dependent response of the global ocean clathrate reservoir to climatic and anthropogenic forcing, G-cubed, 6,3, doi: 10.1029/2004GC000854, 2005.
Bousquet, P., P. Ciais, J.B. Miller, E.J. Dlugokencky, D.A. Hauglustaine, C. Prigent, G.R. Van der Werf, P. Peylin, E.G. Brunke, C. Carouge, R.L. Langenfelds, J. Lathiere, F. Papa, M. Ramonet, M. Schmidt, L.P. Steele, S.C. Tyler, and J. White, Contribution of anthropogenic and natural sources to atmospheric methane variability, Nature, 443 (7110), 439-443, 2006.
Butler, T.M., P.J. Rayner, I. Simmonds, and M.G. Lawrence, Simultaneous mass balance inverse modeling of methane and carbon monoxide, Journal of Geophysical Research-Atmospheres, 110 (D21), 2005.
Dlugokencky, E.J., E.G. Dutton, P.C. Novelli, P.P. Tans, K.A. Masarie, K.O. Lantz, and S. Madronich, Changes in CH4 and CO growth rates after the eruption of Mt Pinatubo and their link with changes in tropical tropospheric UV flux, Geophysical Research Letters, 23 (20), 2761-2764, 1996.
Dlugokencky, E.J., S. Houweling, L. Bruhwiler, K.A. Masarie, P.M. Lang, J.B. Miller, and P.P. Tans, Atmospheric methane levels off: Temporary pause or a new steady-state?, Geophysical Research Letters, 30 (19), 2003.
Keppler, F., J.T.G. Hamilton, M. Brass, and T. Rockmann, Methane emissions from terrestrial plants under aerobic conditions, Nature, 439 (7073), 187-191, 2006.
Krol, M., and J. Lelieveld, Can the variability in tropospheric OH be deduced from measurements of 1,1,1-trichloroethane (methyl chloroform)?, Journal of Geophysical Research-Atmospheres, 108 (D3), 2003.
Langenfelds, R.L., R.J. Francey, B.C. Pak, L.P. Steele, J. Lloyd, C.M. Trudinger, and C.E. Allison, Interannual growth rate variations of atmospheric CO2 and its delta C-13, H-2, CH4, and CO between 1992 and 1999 linked to biomass burning, Global Biogeochemical Cycles, 16 (3), 2002.
Manning, M.R., D.C. Lowe, R.C. Moss, G.E. Bodeker, and W. Allan, Short-term variations in the oxidizing power of the atmosphere, Nature, 436 (7053), 1001-1004, 2005.
Prinn, R.G., J. Huang, R.F. Weiss, D.M. Cunnold, P.J. Fraser, P.G. Simmonds, A. McCulloch, C. Harth, S. Reimann, P. Salameh, S. O’Doherty, R.H.J. Wang, L.W. Porter, B.R. Miller, and P.B. Krummel, Evidence for variability of atmospheric hydroxyl radicals over the past quarter century, Geophysical Research Letters, 32 (7), 2005.
Simmonds, P.G., A.J. Manning, R.G. Derwent, P. Ciais, M. Ramonet, V. Kazan, and D. Ryall, A burning question. Can recent growth rate anomalies in the greenhouse gases be attributed to large-scale biomass burning events?, Atmospheric Environment, 39 (14), 2513-2517, 2005.
Wennberg, P.O., S. Peacock, J.T. Randerson, and R. Bleck, Recent changes in the air-sea gas exchange of methyl chloroform, Geophysical Research Letters, 31 (16), 2004.