Climate Change: High risk of permafrost thaw

Somewhat in the shadow of the big state-of-the-Arctic publication today is a related paper in Nature, an expert elicitation on the risks of methane releases from permafrost thaw by Schuur and Abbott is released today.

Climate change: High risk of permafrost thaw

Nature 480, 32–33 (01 December 2011)

Experts were fairly tightly clustered in their estimates of equivalent CO2 to be expected from this phenomenon. In a high warming scenario, 95% confidence bounds offered CO2 equivalents of:

“30 billion to 63 billion tonnes of carbon by 2040, reaching 232 billion to 380 billion tonnes by 2100 and 549 billion to 865 billion tonnes by 2300”

with roughly 1/3 of those values in a low warming scenario.

The low warming scenario presumably corresponds to the trillion ton scenario, under which we have 450 billion tons left to emit. So at a third of the approximately 300 billion tons by 2100, our safe operating range is substantially cut, from 450 B tons left to about 350 B tons left, with just this one geochemical feedback.

This is not the methane runaway some people lose sleep over, but it’s pretty nasty news just the same. As the paper says:

Knowing how much carbon will be released from the permafrost zone in this century and beyond is crucial for determining the appropriate response. But despite the massive amount of carbon in permafrost soils, emissions from these soils are unlikely to overshadow those from the burning of fossil fuels, which will continue to be the main source of climate forcing. Permafrost carbon release will still be an important amplifier of climate change, however, and is in some ways more problematic: it occurs in remote places, far from human influence, and is dispersed across the landscape. Trapping carbon emissions at the source — as one might do at power plants — is not an option. And once the soils thaw, emissions are likely to continue for decades, or even centuries.

Photo is from the article, captioned “Abrupt thaw, as seen here in Alaska’s Noatak National Preserve, causes the land to collapse, accelerating permafrost degradation and carbon release.”


  1. I like these expert elicitations. Not only do they take the guesswork out of figuring out what a given field thinks of recent research, if implemented widely they have the potential to become a more nimble alternative to WG I.

  2. Two points:

    1. Yes, the IPCC is an expert elicitation exercise, but one that must pass through a particular nasty political filter. We've seen the way that distorts the outcome ("embers diagram", etc.), so I would consider the IPCC its own sun-genre of eee.

    2. True, this is not the permafrost bomb so many fear, but compare this view of the situation with what we thought five or ten years ago, and we've taken a big step towards that awful scenario. This is what keeps many of us, including me, up at night: Because of the appropriate and necessary conservatism of scientists, we're in a state of making almost constant discoveries about things being worse than expected. Speaking as an economist, i.e. one trained to think it's possible to see around corners, it's not hard to look at published papers from the last few years and wonder just how much worse things are than we know right now. It would be a very serious mistake for scientists to speculate about such an extrapolation, and I'm glad they refrain from it. But for the rest of us it's at a bare minimum a question begging to be answered, one with potentially horrendous consequences if we underestimate the threat or take too long to comprehend it.

    This is why I think one of the most important ways governments can spend money over the next few decades is in funding greatly increased climate science efforts. Everything from lab work and boots on the tundra and ground-based sensors to satellites and data analysis, with the results all publicly available. We need to know as much as we can and as soon as possible, simply because what we know now and the trend line are not painting a cheery picture.

  3. Agree entirely, Lou.

    Something else we could do is start looking at the Arctic problem from the other end entirely, The facts seem indisputable, yet I for one haven't heard the problem put as follows.

    The large quantity of carbon-rich permafrost formations, shallow methane clathrates and methanogenetic sediments present in and around the East Siberian Shelf are almost certainly unique in the planet's history. It took a pretty unlikely combination of circumstances, the Pleistocene ice ages (only the third deep ice age of the last ~600,000,000 years) to keep the Arctic cool, a shallow Arctic Ocean with a huge shelf area that's dry land during glaciations (allowing permafrost formation over clathrates and sediments), a further coincidence of geography that keeps Siberia ice-free during the glaciations (prevailing winds lose their moisture over the European ice sheet) so that large quantities of carbon-rich permafrost (yedoma) can build up, and several of the world's biggest rivers dumping huge amounts of sediment onto the shallow shelf zones such that a little warming can induce rapid decomposition, well, suffice to say it's a whole lot of CO2 and methane lacking only a little assistance from us to get into the atmosphere. Rapid release of a substantial fraction of these Pleistocene carbon deposits would result in what could only be described as a global catastrophe.

    (Is there anything fundamentally wrong with that picture, Michael?)

    What could possibly go wrong? Oh yes, rapid warming of the Arctic generally and of Siberia and the shallow shelf currents in particular, already observed, that's what could go wrong.

    Flipping that around, what is our expactation that these deposits, which largely didn't exist during the mid-Pliocene warm period only ~3.3 mya, *won't* release rapidly if temperatures return to the same level or even warmer? (Insert Dirt Harry reference here.)

    So in my view it really is a bomb, and we really are setting it off. Exactly how much of the fuse we've already burned is of considerable academic interest, notwithstanding that people tend to work on pieces of the problem (just the permafrost e.g.) rather than the big picture, but essentially zero political interest, I think in part because the big picture hasn't been communicated (probably because it's so new and because so much of the problem has yet to be quantified, scientists being traditionally reticent under such circumstances). I have my doubts as to whether the AR5 will do the job.

    What a bummer.

  4. Come to think of it, I do have a disagreement with a couple of things Lou said.

    First, I think scientists *can* safely discuss the problem in the terms I described, as it's simply saying that these carbon deposits are geologically unique (in toto), of recent cool-climate origin, and guaranteed to release on our current warming trajectory, the only question being how quickly. By all means throw in all the uncertainties, but the central message remains.

    Second, while I think more research on this and other aspects of climate is essential, let's not forget that we already know enough to stomp on the brakes as hard as we can.

  5. The clathrate "bomb" has a very long fuse and will not go off all at once AIUI. It's not really all that bomblike, not that it isn't bad news and more stealing from our descendants.

    The Siberian thing specifically is new to me. To make a case you need numbers, specifically quantities and time constants. I am not going to get worked up about it until you point to someone who does that, in peer review, preferably in a major journal. Are you sure this isn't a genuine instance of alarmism?

  6. I'm surprised you haven't seen all this stuff, Michael, but I'll pull it together. That'll be a useful exercise for me anyway.

    Re Dickens' much-maligned "clathrate bomb" concept (sort-of abandoned, noting that at this point he seems to just be defending the idea that clathrates played a major, non-triggering role in the PETM, regarding which there's little disagreement AFAIK), that's not what I'm referring to. It's the shallow stuff on the ESS, which I doubt anyone would think would be able to survive MPWP-like temps.

    Of course the part that nobody can prove yet is how fast all of this can get into the atmosphere, and even with a partial solution to the land permafrost component (not the last word, I think we can be sure) that leaves no indication of how fast the much larger off-shore carbon deposits might outgas. My point is that with sufficient warming it is inevitable that they will to a great extent (else they would be older). That we're on that track and the apparent geologically unique extent of these deposits is what ought to give pause.

    Just to note that we should be hearing very soon from the big NSF-funded ESS obs mission (involving the usual Russian UAF suspects plus now Joye and Liefer, the leading experts on methane in water). This was their second year out there, so if there's a sharp methane emissions trend they may have been able to spot it. Even if not, any quantification of the extent of current emissions will be a help.

    If you haven't seen it, this NSF article is worth a read. It describes the results that caused them to fund the big-bucks effort with Joye and Liefer. I have to say they seem nearly as worried as I am.

    More to come.

  7. Boy, if that NSF "article" isn't an example of science becoming more like politics I don't know what is.

    But if you dig, you see the number 7 Tg/yr as the current release rate. Let's multiply by20 to get the CO2 equivalent, and convert to tons, so we can see how big that number is.
    7 Tg CH4 = 140 Tg CO2eq = 140 KTons CO2 equivalent. On a global scale, not much to worry about; in a whole year this matches the emissions of the Chicago metro area in a day. Correct me if I missed something, or wake me up when you get another three orders of magnitude out of this, please.

    And see that telltale "###" in the "article"? It means "press release, written by the press office". NSF is no better at press releases than the institutions it funds, apparently. The lack of context here is pretty much inexcusable.

  8. I'm not an idiot, Michael. I've beeen carefully following this for several years and *know* the number isn't big yet (or wasn't as of 2008, anyway). But as shelf waters have been warming, the concern about the rate increasing, potentially a lot, is real. Let us hope it's found not to be. Long-term, given the location/composition/size of those deposits and the rapidly warming waters, I'm entirely unsanguine about things. What concerns me is what present behavior is committing us to, not whether it happens now or in a century. Those deposits are going to go, the only question is how fast.

    Who knows if and how the release was vetted; the point is that the featured research caused the NSF to immediately put major resources into an expensive multi-year obs effort for which they recruited the field leaders. That seems notable to me. Unless you think the press office is making funding decisions?

    Maybe you'd prefer Joye's take (and follow the link for more).

  9. Correction: Dickens actually coined the phrase "clathrate gun," although others have used "clathrate bomb" somewhat interchangeably. Just to be clear, that discussion was all about deep ocean clathrates, not the ESS ones about which little was known at the time.

    IIRC Dave Archer and Raypierre were pretty involved in that debate, possibly while you were there, Michael?

  10. Naw, let's not be walking on eggshells here. Be gobsmacked if you're gobsmacked.

    By all means don't be overestimating ME. I'm so out of touch these days. U Texas, for all its advantages, is not like being in the Archer/Pierrehumbert orbit day in and day out. In fact, one selfish reason to try this whole crazy Planet3.0 stunt is to try to stay in the loop even though I'm not in Chicago anymore.

    Note though, this is really important. I AM WILLING TO BE WRONG. The whole point of P3 is thinking out loud without being terribly afraid of getting something wrong.

    I'm just expressing doubts and trying to keep the conversation going. Call it skepticism. It's a lost word, besmirched by the crap of the denial squad, but respectable enough.

    So, back to the question at hand, lacking an estimate of an inventory and the flux being three orders of magnitude too small to worry about, I am provisionally not very worried about it.

  11. Good question. I hadn't really thought it through and I'm at home and have no access to the primary literature at this moment, but I dug this up:

    “Most of the methane is produced during the warm summer months, but not all of it is emitted then,” says Christensen. At the Greenland site, for example, only the uppermost 30 to 50 centimeters of soil thaws each summer. In autumn, the ground freezes from above. The top layer of soil freezes and expands, pressurizing the soil beneath, Christensen contends. Because the underlying permafrost is impermeable, any methane that has accumulated in the thawed soil during the summer is squeezed out and forced to the surface

    Maybe that is what we see in the picture?

    via here via there.

  12. I think that's plain old slope failure, i.e. minor landsliding due to the active layer having grown too thick for the slope to hold together. Frozen earth is pretty strong, wet mushy earth not so much. The methane effect Michael describes would only make things worse.

  13. So, on the photo....If I understand Steve correctly, he says that the photo has nothing whatever to do with 'abrupt thaw', but is an example of something else -- slope failure.

    Are there, or have there been, any real life instances of 'abrupt thaw' anywhere? Have they been studied? Are there papers on the results? Are there photos?

    Mr. Tobis....Thank you for the links to the two articles on a related but entirely different subject which show that methane produced by biological activity in permafrost during the summer is released slowly over the year, even in the winter.

  14. Sorry to have confused things, Kip. Probably better to say that this effect is very commonly observed on slopes, for what should be obvious reasons. Of course it *is* abrupt, since the caption (in Nature) says so and it's Schuur's photo.

    Here's what the article says:

    "The ability to project how much carbon will be released is hampered both by the fact that models do not account for some important processes, and by a lack of data to inform the models. For example, most large-scale models project that permafrost warming depends on how much the air is warming above them. This warming then boosts microbial activity and carbon release. But this is a simplification. Abrupt thaw processes can cause ice wedges to melt and the ground surface to collapse, accelerating the thaw of frozen ground. Evidence of rapid thaw is widespread: you can see it in the ‘drunken’ trees that tip dangerously as a result of ground subsidence, and in collapsed hill slopes marked by scars from landslides. These are just some of the complex processes that models don’t include." (emphasis added)

    I think the upshot is that thaw was thought until recently to be largely a slow process where the ground stays intact while melting proceeds more or less evenly downward from the surface. Just in the last few years, it has become apparent from observations that thawing will involve physical rupture of the ground, not everywhere but extensively enough to vastly accelerate the rate of thaw-related outgassing of CO2 and methane.

    I didn't notice it in the photo as reproduced above, but in the article version the big chunk of ice (an ice wedge) over on the left became very apparent. There's a lot of that in permafrost, and when it goes there's not even any mud left, and so whatever's above it has to collapse to the bottom of where the ice used to be.

    I will nitpick something in the article, though:

    "Abrupt thaw processes can cause ice wedges to melt and the ground surface to collapse..."

    I think it should be "involve" rather than "cause" since the melting ice wedges and collapsing surface (including slope failure) pretty much *are* the larger part of the abrupt thaw processes.

    But see for yourself: The reference for that passage is to the first paper listed here, and the next two look to be updates from the same authors (none of them paywalled, fortunately). Click on "Cited by 115" under the first one to see everything.

    What's most interesting to me about this is that there's been a huge sea change in the scientific understanding of permafrost that in a very short period of time. The modelers are struggling to catch up, and I have a feeling that in a few more years things are going to look much worse than this early indication.

    That's not to blame the scientists involved in any way. They need data to base projections on, and they do the best they can with what they have. This is, unfortunately, a really good example of why it is that increasing data tends to worsen the climate picture.

    I think what it means is that once enough ice is lost, collapse is quick and all of the material involved completes the thawing process quickly. It's important because it's a mechanism for permafrost to go away quickly rather than having to wait for thawing to penetrate from top to bottom. Observations of the rapid spread of this process in the recent past have caused the

  15. Pingback: The Permafrost Trigger | CultureSwerve

  16. Mr. Bloom: Thank you for taking the time to consider and explain. It looks to me that this may actually be a photo of a collapsed bank along the river or estuary shown in the photo. I have queried the corresponding author directly to sort this out.

  17. Nosing around I happened on this video from the Vulnerability of Permafrost Carbon
    Research Coordination Network web site, the umbrella for the group that produced the Nature comment with the new estimate on permafrost carbon’s state and inventory. I don’t think it illustrates the mechanism for the entirety of the very substantial boost in expected methane (and ultimately CO2) addition to the atmosphere sourced to thawing permafrost, but I found it interesting.

    As best as I can guess, the location of the Horn Lake in the video is immediately on the north side of the Books Range and adjacent to the Trans Alaska Pipe Line.

    The site for the coordination network:

  18. Pete, the Arctic in the mid-Pliocene warm period (MPWP) was much too warm to allow for such a thing. Search on that phrase in Google Scholar and you'll turn up enough of the research for a start. Actually there's a rather large collaborative effort (PRISM, led by Harry Dowsett of the USGS; they have a website) working on characterizing this period since it's the best analog for where we're headed (although still imperfect since it was an equilibrium state and our most immediate concern is the conditions associated with a sharp transient, possibly taking us rather quickly to even warmer conditions than the MPWP).

    I say "largely" because obviously there still would have been carbon in river sediments.

  19. Hmm, refused embedded links and chopped off my comment. Eh.

    The bit at the end was just noting the very rapid change in collective view of of the permafrost field, and expecting that follow-on news will be worse. That sort of thing is why I'm way more worried about the East Siberian area shelf than is Michael.

  20. Very relevant to this discussion is this recent post by Steve Easterbrook on a recent talk by leading paleoclimate researcher Mark Pagani of Yale. Steve was most interested in the distinction between Charney climate sensitivity (the ~3C one we hear about all the time) and Earth System Sensitivity, which adds all known forcings and feedbacks (missing from Charney's since so much was unknown when it was defined in the '70s) such that it's a much better approximation of real sensitivity, and beyond that the apparent variation in ESS depending on initial climate state.

    But I was much more interested in this aspect of the talk:

    "Mark’s preferred explanation is a melting of the antarctic permafrost, caused by a shift in orbital cycles, and indeed he demonstrates that the orbital pattern leads to similar spikes (of decreasing amplitude) throughout the Eocene. Prior to the PETM, Antarctica would have been ice free for so long that a substantial permafrost would have built up, and even conservative estimates based on today’s permafrost in the sub-arctic regions would have enough carbon to explain the observed changes."

    So it was the permafrost wot triggered the PETM. As noted, this mechanism fits very well not just with the carbon isotope evidence but with the evidence that other hypotheses fall short on, i.e. the series of subsequent Eocene hyperthermals that steadily declined in strength. Steve's post wasn't clear on this point, but I expect that clathrates were the other major component of these events, just not the trigger. We'll see when Pagani publishes.

    Hmm, so lots of permafrost melting fast, and lots of clathrates available for melt. Well, fortunately we don't have any such thing to worry about. Oh wait...

  21. Oh foo. I am wrong by exactly the three orders of magnitude I asked for.

    Somebody wrote in to tell me this but I fat thumbed it on my phone and lost the comment instead of approving it.

    But I finally checked after blabbing on Twitter. Three orders of magnitude in the wrong direction. This does verge on significant after all. My mistake, and my apologies.

  22. Two images for Kip:
    NPR: Washing Away The Arctic Coastline.

    Breathtaking slope failures! Extra large images (not linked from above link):

    From the landscape of the last image I read there have been much larger slope failures some longer time ago. How long?

  23. Even though I was off in my calculation, I do agree with Revkin. The Russians have no evidence that the emissions they have discovered are increasing, and the time scale for propagation of surface warming to depth is long. This is not a real issue.

  24. The Revkin feather, scarce tipping the scales.

    Michael, Revkin talked to cited the wrong scientists, i.e. permafrost experts. What he needed to do is talk to one of the American expedition co-leaders, Samantha Joye and Ira Leifer, aka the world's leading authorities on methane in water. He didn't. Go figure.

    You should know better than to swallow a report like that so uncritically.

    We need to know what the observed trends are for hydrate loss into the water and any subsequent transmission into the air. The Independent article may have only given us anecdote, but Revkin's failed to refute the potential significance of that anecdote to any degree.

  25. I believe Revkin has made a mistake in using an older paper to reassure us about new findings. While the permafrost issue is impossible to judge at this point, Revkin's don't worry attitude is more or less based on current conditions. This doesn't speak to the risk involved. The short life of Methane in the atmosphere means that by the time there is a "bomb", if at all, we would already be heading for catastrophe anyway. Is that the way you read it?

  26. Not at all.

    The Russians have discovered an interesting phenomenon, but the fact that they just discovered it does not mean it had not been going on all along. There is no evidence of anything unusual in the methane record. The vulnerability is to surface methane, not to deep sea methane, because it takes a long time for the warming signal to propagate down to the ocean floor. So there is no reason to panic. And the investigators or their institutions implying that there is new bad news on the table seems to me unacceptable.

  27. Steve, I don't think it will take any prompting from me for David Archer to eventually get around to this on RC. But David is probably too much of a gentleman to address this idea that Joye and Leifer are "the world's leading authorities on methane in water". A search for "Joye Leifer clathrate" turns up a comment by Steve Bloom at Bart's as the first hit!

    I welcome comments from either of them, of course. But the fact remains that the Russian paper predicts detectable impacts in about a thousand years. That is nothing to lose sleep over.

    Which "American expedition" are you talking about?

    Joye and Leifer seem to be all about the Gulf of Mexico and the BP spill. More power to them for that, but there's no need to conflate these drastically different phenomena at the level of physics.

  28. Criminy, Michael, you are just not paying attention to this. Joye and Leifer did not get signed on to this because of their expertise on clathrates, but because of their expertise on the detection and behavior of methane in water. Try these results instead.

    I think the Russian/UAF scientists involved probably already consider themselves experts on clathrates and permafrost, although I wouldn't have any idea as to their standing in those fields.

  29. These may be all of the relevant NSF grants. Given that other countries are involved, it looks like a quite serious effort with quite serious money behind it. I suppose the expensive part almost goes without saying, given the nature of the required obs.

    Notice the timing of the grants, with Joye and Leifer being brought in recently. Shakhova and Semiletov actually started obs in 2003, so probably there were some prior, smaller grants to them, now expired. The search was only for the active ones.

    BTW, if you want to check for all current NSF grants to both Joye and Leifer, you'll find lists that I think even you'll agree are impressive.

  30. I have made no statement about Joye and Leifer other than that I would welcome their input.

    Update: Missed Steve's previous comment. Yes you got the same search results I did. So I qualified with "arctic" and Steve Bloom was the first hit.

    If they developed the instrumentation, that is very cool. This means it is the first observation of this phenomenon. To my understanding, it takes at least two data points separated in time to estimate a trend, and many to establish a trend with confidence.

    Until then we have strong theoretical reasons for believing that there is no substantial relationship between these events and global warming.

    My guess at this point would be "isostatic rebound" as the driver, which would mean we are looking at a fairly typical class of event in the interglacials. But admittedly I am guessing.

    This says nothing one way or another about tundral releases, which are a very different deal. I am not sure that either the inventory or the time scales are well-characterized. But that's a separate question.

    However some of them act in front of a camera, in formal publications they say stuff like the EOS summary:

    Forecasting the expected future permafrost thaw, the authors found that even under the most extreme climatic scenario tested this thawed soil growth will not exceed 10 meters by 2100 or 50 meters by the turn of the next millennium. The authors note that the bulk of the methane stores in the east Siberian shelf are trapped roughly 200 meters below the seafloor, indicating that the recent methane emissions observations were likely not connected to the modest modern permafrost thaw.

    Which is what I am saying, and what Revkin is saying. So what are you saying?

  31. Michael, coincidentally I have asked Revkin about this "bulk" of the sub-sea methane stores on David Appell's blog. I'll reproduce it here:


    Just as it isn't about the entire Greenland and Antarctic ice sheets, but about how fast the amount of melted ice sheet needed to elevate sea level by 1-2 meters could come about, it isn't about the "brunt of the sub-sea methane" either.

    If I've understood correctly, there could be about 1400 to 2200 billion tons of sub-sea methane (the estimates vary). In Shakhova et al. (2008) it is estimated that "release of up to 50 Gt of predicted amount of hydrate storage [is] highly possible for abrupt release at any time". That would increase the methane content of the planet's atmosphere by a factor of twelve. (source: Wikipedia)

    50 Gt of 1400 Gt (lowest estimate) is 3.6%. Methane currently has a forcing of 0.5 W/m2 (if I've understood correctly).

    I'm probably making some sort of mistake, but here goes:
    So if 3.6% of the sub-sea methane would be released - hardly the "brunt" I would say -, this forcing would increase twelvefold to let's say 6 W/m2 (CO2 currently is 1.6 W/m2).

    If this is correct, I'd be inclined to say 'Apocalypse perhaps not immediately but definitely not "Apocalypse not"'.


    We're also discussing this over at the Arctic sea ice blog. The Idiot Tracker has an interesting perspective on the Revkin piece.

    IMO these things are important: 1) the numbers that will be released by Semiletov in Spring, so we get a bit of a (further) grip on things, 2) global methane concentration in the atmosphere that has been going up since 2006, 3) Arctic sea ice, as it plays a crucial role in all of this.

    We'll know more in a couple of years. I hope research into this gets a big boost.

  32. To correct my own stupidity: the impact of a forcing doesn't increase linearly, so 12 x 0.5 W/m2=6 W/m2 is too simple. Nevertheless, my question still stands. What happens when 2.5%-3.5% of the sub-sea methane stores gets released (the non-brunt, if you will)? How long/much warming would that take? Is there a correlation with Arctic sea ice?

    Coincidentally global methane concentration started rising again in 2006, around the time when we saw the 'shift' in Arctic summer sea ice melt.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.