By Matt Owens August 27, 2013
The potent greenhouse gas methane (CH4) is given short shrift in climate science, and that's not even considering the methane hydrate or permafrost thaw issues. In fact, sustained methane levels from microbe-generated methane (microbial methanogenesis) could dwarf an Arctic methane hydrate pulse from the East Siberian Shelf.
Current climate models ignore this issue. They use a set of prescribed greenhouse gas levels to run their simulations, and they generally follow the same standardized prescriptions, the RCP scenarios. These RCP scenarios assume that today's atmospheric methane level rises and falls only as a result of direct human activity. This activity is mostly things like methane-producing farming practices and methane releases from burning fossil-fuels.
But looking at the paleoclimate data for the past 800,000 years (when Earth fluctuated between a climate like today's and a colder one that featured large continental ice sheets) shows that methane is closely tied to carbon dioxide levels and temperature. I believe that it's reasonable this relationship will persist, and that therefore very faulty assumptions of methane levels underlie each of the four standard RCP (representative concentration pathway) scenarios. This is a serious problem because these pathways form the basis for almost all climate science studies and evaluations of potential impacts.
What's more, in the RCP scenarios, because methane is presumed to be entirely now under human direct control, its level remains at today's atmospheric concentration or even falls below it, despite significantly higher CO2 levels (even in the lowest RCP2.6, CO2 reaches about 450 ppm), except for just one scenario, the worst-case¹ scenario, RCP8.5. This worst-case scenario is used less often out of the four central scenarios for evaluating future impacts of global warming. In other words, the projections for agricultural output, sea level rise, temperature rise, rates of species extinction, droughts, and so on, are all based on what appears to be far-too optimistic assumptions about methane. In that worst-case scenario where methane rises, it reaches a peak of about 3.7 ppm in 2100,² as compared to today's value of ~ 1.85 ppm. Even this worst-case scenario could be an enormous underestimation of the appropriate methane level if methane is actually a function of temperature and CO2.
There are two connected lines of reasoning for why this is the case. First, the ice-core paleorecord shows that while CH4 rises in tandem with CO2 (and temperature), it rises at a faster pace than CO2. This relationship indicates that the CH4 equilibrium level is not a static fraction, but rather an exponential function, probably of global temperature and CO2 levels. Second, the main atmospheric source of CH4 today is microbial activity which fluctuates by orders of magnitude over time and location based on how favorable conditions are. And favorable conditions are expected to rise significantly with global warming as a result of rising temperatures, increased CO2 and the resulting increased organic carbon availability, and an accelerated hydrological cycle (more rain, more sudden rain, more erosion, and more flooded soil).
Above: CH4, N2O, and CO2 levels for the past 800,000 years, based on ice-cores. Each gas is indexed to its pre-industrial maximum (bottom set of lines) and minimum (top set of lines). So a value of 2 on the top set means that the particular gas has reached twice its pre-industrial minimum.
The CO2 and CH4 record for the past 800,000 years (chart above) shows that both gases rise in tandem (and I should add, along with N2O, another potent greenhouse gas). But during warm periods, CH4 rises much faster than CO2. If microbial methane production rates rise exponentially as a function of global temperature, soil moisture, and CO2 levels, then this relationship should continue.
A more detailed look reveals that CH4 reaches between 2.0 and 2.3 times its minimum value in sharp upward spikes, while also demonstrating a number of sustained plateaus of around 1.8 to 2.0 times its minimum. CO2 on the other has comparatively more muted spikes, only making it just barely and briefly to 1.75 times its minimum during the pre-industrial period. And when carbon dioxide does show a sustained plateau, it's around 1.5 times its minimum. Methane levels clearly fluctuate with a greater amplitude than CO2 levels.
So is this excitable methane function simply a coincidence of melting and retreating ice sheets that opened up more soil surface and standing water for methanogenesis? Or perhaps that and some other transient climate feature? The record shows that methane levels remained at elevated levels when CO2 was stable, and that when CO2 fell, methane fell too, but at much sharper rates. This again argues for CH4 being a function of CO2, if not temperature as well. The ice sheet retreat certainly didn't hurt, but there is a growing body of research showing that methane production will contribute significantly to global warming feedbacks.³,⁴,⁵,⁶,⁷,⁸
Today's level of methane has reached 5.5 times its pre-industrial minimum, while CO2 has increased to about 2.3 times its respective minimum. So, methane is continuing to outpace CO2, and at an accelerating rate!
To elaborate on this second line of reasoning:
A) The rate of microbial methane formation depends on substrate suitability and quantity. With more organic carbon available, such as in the form of dead plant leaves, there is a higher production of methane. And as CO2 levels rise, net primary production (the amount of plant biomass produced) increases,⁹ all things being equal. While all things are not equal, this increase in net primary production has already been shown to be well underway across diverse ecosystems and across the continents.
B) Temperature also mediates methane formation. Higher temperatures lead to higher methane production rates in soil, and potentially from living plants as well under drought stress.⁷
C) Methane production is highly variable in space and time, which makes it challenging to take meaningful measurements in the field. As I've personally seen, a few days of rain or one heavy burst of torrential rain can sweep large amounts of forest floor litter (leaves) into clumps and leave large areas of standing water on the forest floor (picture below shows flooded forest floor near Accotink Lake in 2013). In just a few days, the standing water is gone. But in these shallow pools, methane production can become exceptionally high, and with shallow water depth, a large ratio of the methane can escape into the air before other microbes have the chance to oxidize it; and perhaps before field researchers have the opportunity to take note of it.
D) With increasing rates of torrential downpour, there could easily be an increase in the distribution and frequency of temporarily flooded soils, leading to more methane production.
E) Coupling the above issues together with evidence for past increases in methane production, it seems clear that the RCP scenarios are ignoring a large amount of methane in all of the scenarios. Furthermore, it seems equally clear that while our actions have increased the amount of methane in the air today, bringing that level down could be impossible without first reducing CO2 levels and global temperature.
Ultimately, we are going to have to face the truth and admit that not only is the Earth's climate changing, but it's changing in a potentially very dangerous and hard to reverse way. We should do what we can to stop exacerbating the problem immediately and prepare to accommodate what could potentially become billions of people who will need to either relocate outside their country of origin or undergo surprising lifestyle adaptations to survive in their current location.
Based on the past relationship of methane and carbon dioxide to one another and to global temperature, it appears that atmospheric methane levels will potentially continue to rise to an unknown new equilibrium level. This level is probably impossible to determine based only on the ice-core record from the past 800,000 years because we have surpassed the CO2 levels seen during that time and are quickly surpassing global average temperature levels too.
Additionally, methane levels in the atmosphere continue to rise, despite expectations that they would stabilize by now. The poor integration of soil, aquatic, microbial, and plant systems into global climate projections creates a potentially massive blind spot. Climate modelers should take urgent measures to educate themselves in these areas and significantly increase the methane levels in their scenarios.
¹ The term worst-case scenario which I use to describe RCP8.5 (roughly equal to the precursor A1FI), is only meant relative to the other RCP scenarios. Of the four, RCP8.5 is the worst. However, if you read the literature on the pathways carefully, there is no claim that RCP8.5 is the worst-case possible. This aspect should be emphasized more strongly.
² Meinshausen et al. 2011 "The RCP greenhouse gas concentrations and their extensions from 1765 to 2300" (doi: 10.1007/s10584-011-0156-z).
³ Levy et al. 2012 "Methane emissions from soils: synthesis and analysis of a large UK data set" (doi: 10.1111/j.1365-2486.2011.02616.x).
⁴ Inglett et al. 2012 "Temperature sensitivity of greenhouse gas production in wetland soils of different vegetation" (doi: 10.1007/s10533-011-9573-3).
⁵ Pancost et al. 2007 "Increased terrestrial methane cycling at the Palaeocene–Eocene thermal maximum" (doi: 10.1038/nature06012).
⁶ Yvon-Durocher et al. 2011 "Warming increases the proportion of primary production emitted as methane from freshwater mesocosms" (doi: 10.1111/j.1365-2486.2010.02289.x).
⁷ Qaderi et al. 2009 "Methane emissions from six crop species exposed to three components of global climate change: temperature, ultraviolet-B radiation and water stress" (doi: 10.1111/j.1399-3054.2009.01268.x).
⁸ Christiansen et al. 2011 "Nitrous oxide and methane exchange in two small temperate forest catchments—effects of hydrological gradients and implications for global warming potentials of forest soils" (doi: 10.1007/s10533-010-9563-x).
⁹ See for example: Wang et al. 2006 "Effect of natural atmospheric CO2 fertilization suggested by open-grown white spruce in a dry environment" (doi: 10.1111/j.1365-2486.2006.01098.x). And also see that this change impacts soil carbon and soil chemistry, for example in: Liu et al. 2009 "Enhanced litter input rather than changes in litter chemistry drive soil carbon and nitrogen cycles under elevated CO2: a microcosm study" (doi: 10.1111/j.1365-2486.2008.01747.x).