Why the “methane bomb” scenario is not “settled science” (yet)

Foreword: I have created this blog a few weeks ago with the intention of only using it for an identity when commenting to other good blogs that I like, for instance this and this. However this recent tweet from Gavin Schmidt made me reconsider things for a very quick and dirty post, I apologize in advance for the graphs.

A little context might help: recently there was some noise in the blogosphere about how (Russian) scientists Shakhova and Semiletov (authors of many papers including this and this) were not invited to speak to a recent Royal Society meeting where Gavin Schmidt was also talking in the context of the “methane bomb” (the tweet linked in the first paragraph above contains an audio of his presentation and a few other highly-recommended materials which I will actually reuse in this post).

Two other aspects must be also mentioned here for context, first is that the “methane bomb” was recently sensationalized to death in the non-scientific media (to some extent as a result of the papers from the two Russian authors mentioned above). The second is that as a result of that hype a few of the more mainstream scientists (most notably Gavin Schmidt and David Archer for instance here, also Chris Colose in some social media) have felt the need to defuse the sensationalism with a more conservative analysis. Which in itself is a great thing.

However I believe that Gavin Schmidt was far from convincing in his  recent Royal Society presentation, and what is more unsettling for me (given that I have a very solid respect for him and his scientific work) is that some aspects have been skipped-over in a very surprising way given the information in his own presentation.

I will start with this slide:


and I will only highlight his own number here – a sudden one-time release of  50 Gt of methane would only be visible with a 500 year resolution as a 40 ppbv peak. The resolution for the Eemian is currently a little better than that (300-400 years) so let’s say we should get a peak of around 50-60 ppbv above the longer-term average (which in itself might have certain calibration problems with very large values). This number also comes with the very important caveat that such a peak will only be seen assuming an ice proxy with a very high and relatively constant snow accumulation, those are obviously exactly what experts in the field select for such research but under fast-changing conditions (like today’s warming, and very likely also certain intervals during the Eemian) the high-snow accumulation might be quite variable.

Now let’s move to Gavin’s own methane graph two slides later:


EDML is Antarctica (EPICA) Dronning Maud Land site and EDC is Antarctica (EPICA) Dome C site, those are quite different sites and there might be certain differences in the rate of snow accumulation between them, hence the minor differences seen on the graph. I consider EDML a much better proxy for methane given how it also captures the sudden decrease in methane around 128-129kyr, which is very similar to what we see in the most recent deglaciation around 12kyr ago (with the obvious difference that the data in Eemian is a lot more smoothed as a result of the larger intervals compared to the much better accuracy in the 10 times more recent records).

Let’s add (please don’t laugh, just for context, quick and dirty by hand) on the same graph the solar forcing during polar summer (light blue, peaks around 128-130kyr ago):


Everything looks rather normal, right? Well, not so fast, now let’s also add (in green) for context the sea level from Kopp 2009 (Hearty 2007 would look even more damning but let’s work with the most conservative scenario) and let’s take a second look:


Now the sudden methane spike visible around 121 kyr no longer looks like an artifact, does it? And how large is that spike you ask? Well, I would say around 60-70 ppbv above the values around it. Which somehow happens to be Gavin’s own estimate on how we would detect a 50 Gt sudden methane release in a proxy with that kind of smoothing that we have here.

In that context the 2nd (most recent, 118 kyr ago) sea-level sudden peak does not have any other more obvious cause since CO2 itself does not have a major 2nd peak, this is from Möller 2013:

and would be even more unexpected to have a very sudden collapse point in the ice sheets themselves under reduced radiative forcing (but that is not completely excluded either).

Of course there is the question of why the extra methane release did not happen during the initial peak in radiative forcing around 128 kyr ago (or for that matter at the peak from the Holocene deglaciation around 10 kyr ago). The answer is of course the very slow warming in the deep ocean, it takes many thousands of years for the deep ocean to warm from the levels reached during the full glacial, and there is a very good chance that today we could already have the warmest deep ocean in at least 110 kyr and given the current astonishing rate of warming in the oceans it is not impossible to think that around 2100 we could eventually reach levels similar to (or even above) the peaks from 121 kyr ago (regarding the normal rate of ocean warming in the Holocene see AR5 WG1 for instance section 5.6.3).

So where does all this leave the “methane bomb” scenario? I would still say that it is a pretty unlikely scenario, actually very unlikely for the next 1-2 decades but under no circumstance we can say it is completely excluded a few decades after that. And the matter of ice sheet instability (probably much bigger than what was seen with MIS 5e) around the ocean heat levels that we will most likely reach by 2100 remains a huge problem, with or without an extra methane kick.

Extra references that might be helpful:


2) Kopp, Robert E., et al. “Probabilistic assessment of sea level during the last interglacial stage.” Nature 462.7275 (2009): 863-867.

3) Hearty, Paul J., et al. “Global sea-level fluctuations during the Last Interglaciation (MIS 5e).” Quaternary Science Reviews 26.17 (2007): 2090-2112.

4) Schilt, Adrian, et al. “Atmospheric nitrous oxide during the last 140,000 years.”Earth and Planetary Science Letters 300.1 (2010): 33-43.

5) Loulergue, Laetitia, et al. “Orbital and millennial-scale features of atmospheric CH4 over the past 800,000 years.” Nature 453.7193 (2008): 383-386.

6) Huber, Christof, et al. “Isotope calibrated Greenland temperature record over Marine Isotope Stage 3 and its relation to CH4.” Earth and Planetary Science Letters 243.3 (2006): 504-519.

EDIT: For some reason I was convinced that at 6) I had given a reference to Möller 2013 but that’s what you get when you write your posts during lunch break based on hand-drawn graphs made during the morning coffee break 🙂 Either way this one was now added here:

7) Möller, Lars, et al. “Independent variations of CH4 emissions and isotopic composition over the past 160,000 years.” Nature Geoscience 6.10 (2013): 885-890.