Terraforming Earth is the effort to use large-scale engineering to affect geophysical processes in a way to avert radical changes to the environment -- that is, to make Earth "Earth-like" again. I touched on the idea first here, expanded on it here, and explored some of the more philosophical questions here. In all of these pieces, however, you'll note that this terraforming work is thought to be an option for some time down the road, after other solutions are exhausted. There's no argument in those three essays that we should start large scale engineering efforts now.

Today's email brought news that should make us think hard about how soon we might want to bring such efforts to bear.

Many of you sent me links to the article in today's Guardian UK newspaper (linking to a New Scientist article) outlining a "tipping point" in the Siberian arctic: the permafrost appears to be melting. This is happening due to a combination of natural arctic temperature cycles, global warming (Siberia is warming faster than any other place on Earth), and a feedback effect from melting snow -- the darker ground absorbs more heat, resulting in faster melting of adjacent permafrost. Siberian permafrost covers a million square kilometers of ground that's largely peat bog; the peat has been producing methane for centuries, but that methane has been trapped under the permafrost. With the permafrost melting, the methane would be released into the atmosphere, accelerating global warming by a substantial amount. How quickly the methane would be released remains an open question -- would it take years to release it all? Decades? A century or more? Clearly, this situation demands a great deal more study.

It's important to note that the source of this story is not a peer-reviewed, multiply-confirmed piece of research in Nature, Science or the PNAS. It's an article in New Scientist about a presentation from a group of researchers just back from Siberia. This doesn't mean that the findings are wrong, only that we should be skeptical until they've been confirmed. But that such permafrost melting would result in the release of abundant methane is not a new theory, and New Scientist notes that independent research points to methane "hot spots" already forming in the region.

For the moment, then, let's assume that the article is generally correct: the permafrost melt is getting faster, and the boggy ground beneath is releasing its pent-up methane. There are two important things to know about this situation: the amount of methane that would be released is projected to be in the multi-gigaton range -- one source says 70 billion tons, another says "several hundred" billion tons; and methane is 21 times more powerful a greenhouse gas than carbon dioxide. In essence, the release of (say) 100 billion tons of methane would be the functional heat-trapping equivalent of 2.1 trillion tons of CO2. To put that number into perspective, the total annual output of greenhouse gases from the US is about 7 billion tons of CO2 equivalent.

This is a big deal.


But there's actually a third important thing to know: although CO2 takes upwards of a century to cycle out of the atmosphere naturally, methane (CH4) takes only about ten years. Why the difference? Chemical processes in the atmosphere break down CH4 (in combination with oxygen) into CO2+H2O -- carbon dioxide and water. In addition, certain bacteria -- known as methanotrophs -- actually consume methane, with the same chemical results. These processes have their limits, however; an abundance of methane in the atmosphere can overwhelm the oxidation chemistry, making the methane stick around for longer than the typical 8-10 years, and the commonplace methanotrophic bacteria evolved in an environment where methane emerges gradually.

These are pretty much the only two natural methane "sinks." There are a few small-scale human processes that can make use of methane (for the production of methanol for fuel, for example) and function as artificial sinks, but such efforts would be hard-pressed to capture methane released across nearly a million square kilometers. This, then, is where we start to consider the option of planetary engineering.

Both of the natural processes are, in principle, amenable to human intervention. The oxidation of methane into CO2 and water is a well-understood phenomenon, and relies on the presence of OH (hydroxyl radical); upwards of 90% of lower atmosphere methane is oxidized through this process (PDF). But OH is something of a problem chemical, in that it's also a key oxidation agent for many atmospheric pollutants, such as carbon monoxide and NOx. Although we could produce OH to enhance the natural chemical oxidation process, the side-effects of pumping enough OH into the atmosphere to oxidize all of that methane would be unpredictable, but almost certainly quite bad.

So what about methanotrophic bacteria? Such bacteria have long been recognized in freshwater areas and soil, and have had limited use in bioremediation efforts. Methanotrophic Archaea -- similar to bacteria, but a wholly different kingdom of organism -- were recently identified in the oceans; research suggests that methanotrophic Archaea may be responsible for the oxidation of up to 80% of the methane in the oceans. Methanotrophic microbes can also be temperature extremophiles, as they were among the various species found after the Larsen B ice shelf collapsed.

We recently began to learn much more about how methanotrophic bacteria function, as a team from the Institute for Genomic Research sequenced the genome of the methanotroph Methylococcus capsulatus. The scientists discovered that Methylococcus has the genomic capacity to adapt to a far wider set of environments than it is currently found in. They also looked at the possibility of enhancing the microbe's ability to oxidize methane, although admittedly for purposes other than straight methane consumption.

You can see where I'm going with this.

Freshwater methanotrophs are increasingly well-understood, but present a limited means of methane remediation. Methanotrophic Archaea have demonstrated ability to act as a major methane sink, at least in the oceans, and to live in extreme temperature conditions. Neither is a good fit for Siberia. The Siberian arctic, while warming, remains damn cold, but the melted permafrost lakes will be freshwater settings.

It appears to me that what will be the most effective means of mitigating and remediating the gargantuan methane excursion from the Siberian permafrost melt would be using genetically-modified forms of methanotrophic bacteria, with greater oxidation capacity and the Archaea-derived resistance to extreme cold (these may well go hand-in-hand, as one way that deep sea methanotrophs survive the icy depths is through internal energy production from methane consumption). Given the size of the region, we'll need lots of them, but that's another advantage of biology over straight chemistry: the methanotrophs would be reproducing themselves.

It's unlikely that abundant reproduction of GMO methanotrophs would pose a larger risk -- at the very least, they'd be limited to the post-permafrost lakes, as they'd be based on freshwater-only species -- and a mass of methanotrophic organisms would undoubtedly be helpful for reducing overall atmospheric methane beyond the Siberian release. More importantly, the successful introduction of such organisms would give us practice for what would be a far, far greater problem: the undersea methane clathrates, which are believed to contain upwards of 500 billion tons of CH4. Undersea clathrate melts have been implicated in mass extinctions in the geologic past; the significant climate warming that would result from an unmitigated Siberian release would pale in comparison to the effects of a clathrate melt.

What are the outstanding questions we need to answer before we could consider creating GMO methanotrophs?

• Is it physically possible? Could a sufficient number of methane-eating bacteria even be produced to counter a fast release of methane from the Siberian bogs?

• Is it biologically possible? Could a species of methanotrophic bacteria be engineered to be able to thrive in Siberian conditions and consume large quantities of methane?

• What are the unrecognized risks? What are we missing in an initial risk analysis? Saying "we don't know the risks" doesn't, in and of itself, mean "we should not attempt this," it means "we need to do more research." Clearly, if the risks from enhancing the methane consumption and environmental adaptation capacities of a methanotroph could lead (through species-hopping genes or simple mutation) to even harder-to-manage problems than gigatons of atmospheric methane, this isn't an option. Boosting OH levels in the region would be the fallback position, as we have more experience with managing CO and NOx pollutants. 

If you think I'm suggesting this option in a casual or flippant manner, you need to read Terraforming Earth essays one, two and three. Planetary engineering -- including the widespread release of genetically modified organisms to combat atmospheric changes -- should only be considered when more readily reversed and managed solutions are no longer available or functional. In the case of the Siberian methane, the more cautious options are extremely limited. We're no longer in a position to stop the melting, even by ceasing all greenhouse gas production today; the temperature increases we're seeing now are the results of greenhouse gases put into the atmosphere decades ago.

In a way, among the different scenarios forcing us to consider "terraforming," this is probably close to the best choice. Failure would be drastic, but not utterly catastrophic (unless the resulting warming, in turn, melts the undersea clathrates, at which point all bets are off). The engineering options are enhancements of natural processes, as opposed to something beyond current experience (such as putting a "solar shade" between the Earth and the Sun to reduce overall insolation). At least with current understanding, a "runaway" condition for the terraforming effort would not mean widespread extinctions (such as would the "runaway" scenario for boosting phytoplankton blooms in the oceans). And, as noted, this would allow for better refinement of technique and understanding of choices in the face of a similar-but-greater in magnitude problem down the road (in this case, the aforementioned clathrates).

A further advantage is that this is a process that could begin after we start to see significant methane output and could still have a measurably positive result. Using microbes for bio-"scrubbing" of methane from the atmosphere would work on methane that was a decade old as readily as methane fresh from the bog. We'd still see some effect from the methane that makes it to the atmosphere, but eventual removal would help to reduce that effect. This means that, should we face a situation where questions still need to be answered before we could comfortably begin to use the GMO methanotroph option, but we're starting to see an impact from the Siberian release, we don't necessarily have to rush past our better judgment in response. With a process of this magnitude, it's worth taking the time to get it right.

I imagine that this, as with the previous Terraforming Earth essays, will trigger some heated questions and discussion -- or, at least, some deep reflection on human choices. I must emphasize again that I don't consider large-scale projects like this to be favored options; I have a strong preference for reversibility, flexibility and limited second- and third-order effects. Planetary engineering -- Terraforming -- embraces none of those three standards.

But if these reports are true (and it remains a very big if, for now), we would be facing a problem of a scale with few precedents in human history. No society on the planet would be unaffected; if left unmitigated, it would continue to affect the lives of our children, and our children's children, and generations beyond that. And -- again, if the reports prove accurate -- this is not a process that can be readily stopped or prevented from happening.

Our choices are few, and the risk of not acting is (potentially) immense. We may well be on the brink of a new era in planetary management. Let's hope we're up to the challenge.