Arguments
Arctic methane outgassing on the E Siberian Shelf part 2 - an interview with Dr Natalia Shakhova
Posted on 19 January 2012 by John Mason
In December 2011, following a fresh flurry of sometimes conflicting media reports about methane outgassing on the East Siberia Arctic Shelf (ESAS), we decided to go and talk to the people doing the work on the ground. We are pleased to report that Dr Natalia Shakhova (NS below) of the University of Alaska in Fairbanks agreed to be interviewed by the author, on behalf of Skeptical Science, via email. Here are the responses, verbatim, to our questions.
Above: Bathymetric map (source - NOAA) of the Arctic with key features noted and the subject area highlighted in red.
SkS: In your JGR paper from 2010 you state that methane hydrate in Siberia can occur at depths as shallow as 20 m. Have any such remarkably shallow methane hydrate deposits on the ESAS been directly observed/sampled and if so, how could methane hydrate have formed at such depths?
NS: Yes, such shallow hydrates were sampled in Siberia. They form as a result of the so-called “self-preservation phenomenon” and they are termed “metastable”. This phenomenon has been intensively studied by Russian geologists starting in the late 1980s.
SkS: Your 2011 field season is reported to have located kilometre-diameter plumes of outgassing methane. Are these located in areas visited in previous seasons?
NS: These were new sites from that part of the ESAS that was investigated very sparsely before. In our previous investigations we mainly focused on the shallower part of the ESAS, which composes about 70% of the total area and provides a very short conduit for methane to escape to the atmosphere. Besides, because we worked mostly on small vessels, we were not allowed to navigate far enough from the coasts to reach the mid-outer shelf where water is relatively deep on the scale of the shallow ESAS (>50 m depth). That is why deeper waters were under-represented and were considered a minor contributor to annual emissions. Last summer’s findings made us re-consider our previous constraint on the annual emission budget; they highlight the need to further assess underestimated components of annual fluxes from the ESAS.
Searching for methane in such an extensive area is truly like searching for a needle in a haystack. The ESAS is more than 2 million square kilometers in extent. Even if we study ~10 000 km2 every year (100x100 km, which is a lot!), it will take >200 years to investigate the entire ESAS! Even then, the probability of finding a hot spot 1 km in diameter within the study area will still be only 0.01%.
SkS: Have you done any analyses/isotopic studies of the fugitive gas to see if anything can be learned about its provenance (i.e., biogenic, thermogenic, destabilized hydrate or a combination of these)?
NS: Yes, we conducted an isotopic analysis to obtain the isotopic signature of the methane dissolved in the water column. The isotopic signature indicates a mixture of methane of different origins. We are currently making an effort to investigate particular sources.
SkS: Do the observed methane outgassing sites tend to correlate with features seen on acoustic imaging of the sea bed (e.g. taliks, pock marks, fractures) or on deep seismic data (e.g. fault-zones, anticlines and other structures)?
NS: We believe that methane outgassing sites primarily correlate with features like those you list above. Our data, although they are still limited, clearly exhibit such a correlation. Unfortunately, there are some limitations in usage of both hydro-acoustic and deep seismic methods imposed by the shallowness of the water column and the ubiquity of shallow gas fronts in the sediments. In addition, our ability to obtain extensive records was constrained by our limited funds; to date we only have ~3000 nautical miles of such recordings.
SkS: A critical question at this point is whether the outgassing is a recent development as a consequence of the dramatic Arctic warming of the past thirty years, or an ongoing, long-term response to the Holocene inundation of the ESAS. What are your thoughts on this and, on a similar line of enquiry, would it be possible to determine the age of the organic matter the methane was originally derived from?
NS: An entire second paragraph of our paper published in Science (Shakhova et al., 2010) is devoted to addressing this question! We were the ones who hypothesized - and devoted our entire study to testing this hypothesis - that methane release from the Arctic shelf is determined by the change in thermal regime of permafrost inundated thousands of years ago. I do not understand why this question should arise over and over again or, moreover, be considered critical. As we deal with the long-lasting permafrost warming caused by the warming effect of the overlying seawater, is there any logic in negating the contribution of the recent warming, which caused additional warming of that overlaying sea water? I believe that there is absolutely no point in trying to determine who is responsible, Mother Nature or human beings. Whoever is responsible, the consequences will be the same.
As for determining the age of the organic matter the methane was derived from, it is very hard to distinguish between modern and ancient sources. The mean age of organic matter preserved even in the surface sediments in the ESAS is 6-8 thousand years, and when you go deeper, you find older organic matter. “Talik” is a term used to describe an unfrozen layer of ground within a still-frozen permafrost body. As taliks develop within the sub-sea permafrost, organic matter of different ages could provide the substrate for methanogenesis. This means that modern methane could be produced from organic matter of different ages, and this is also true of pre-formed methane.
SkS: The recent reports of substantial releases of methane on the ESAS prompt us to ask how these observed emissions could detectably change global atmospheric methane concentrations and in what timeframe?
NS: To date, we have only taken the very first steps down the long path of learning enough to answer this question. We officially reported only 8 Tg of methane was being released from the ESAS per year. This reported amount is <2% of the total annual global methane release and would not detectably change global atmospheric methane concentrations. However, we did not incorporate a few emission components – probably the most important ones - because of some uncertainties still remaining concerning their constraints. Newly obtained data, without question, indicate that annual methane emissions from the ESAS have been underestimated. To say how significant the underestimated components are, and to identify the mechanisms responsible for such substantial releases, we need to carefully analyze obtained data and, very likely, conduct further investigations on a broader scale. To be able to answer your question, which is a core question of our study as well, we need to establish at least a few observatory sites to trace dynamic atmospheric concentrations of methane; we need to develop a monitoring net to detect changes occurring in known plume areas; we also need to continue all-season observations in this region to study temporal and spatial variability in methane releases and the factors that determine this variability. We undoubtedly need to learn much more than we currently know. We call for the involvement of serious funding organizations to give this study the level of support that is consistent with the importance of this topic.
SkS: With respect to future events, in your EGU 2008 abstract it is stated that "we consider release of up to 50Gt of predicted amount of hydrate storage as highly possible for abrupt release at any time". This represents a colossal quantity of gas. How quickly could such a release occur and what would be the most likely mechanism?
NS: I believe that the non-gradual (massive, abrupt) emission mode exists for a variety of reasons. First, wherever in the World Ocean such methane outgassing releases from decaying hydrates occur, they appear to be torch-like with emission rates that change by orders of magnitude within just a few minutes. Note that there was no additional seal such as permafrost to restrict emissions for hundreds of thousands of years anywhere in the World Ocean. Imagine what quantity of methane has been stored beneath sub-sea permafrost if even now, when the permeability of permafrost is still limited, the amount of methane annually escaping from the ESAS is equal to that escaping from the entire World Ocean. Another important factor is that conversion of hydrates to free gas leads to a significant increase in the gas pressure. This highly-pressurized gas exerts a geological power that creates its own gas migration pathways (so-called “chimneys” within sediments). It is even more important to understand that the nature of the permafrost transition from frozen to unfrozen is such that this physical process is not always gradual: the phase transition itself appears to be a relatively short, abrupt transformation, like opening a valve. Remember that the gas “pipeline” is highly pressurized. There could be several different triggers for massive releases: a seismic or tectonic event, endogenous seismicity caused by sediments subsiding pursuant to hydrate decay, or sediment sliding on the shelf break; the shelf slope is very steep, and the sedimentation rates are among the highest in the ESAS. As for the amount that could possibly be released, this estimate represents only a small fraction of the total amount of methane believed to be stored in the ESAS (3.5% of 1400 Gt). Because these emissions occur from extremely shallow water, methane could reach the atmosphere with almost no alteration; the time scale of such releases would largely depend on the spatial distribution and capacity of the gas migration pathways.
SkS: A previous methane release of such a magnitude, occurring abruptly, would logically manifest as a spike in the global methane concentration record, yet the ice-core methane record has no such spikes during previous interglacials. Is there any evidence for massive methane release events having occurred further back - e.g. at any point during the Cenozoic?
NS: You would better address such a question to a specialist in paleo-climatology. To my knowledge, there are a few episodes in the Earth’s history attributed to abrupt methane releases. Interpretation of ice-core methane records may not be relevant, because these records are too short to reach back to the entire Cenozoic.
Skeptical Science would like to thank Dr Shakhova for her contributions.
Notes
The "self preservation phenomenon" mentioned by Dr Shakhova in her reply to the first question is well-known in Russian and other northern petrochemical industry circles, where much discussion may be found. It is temperature-dependent i.e. it requires fairly low temperatures to work. For more information, see Self-preservation of gas hydrates (PDF) for a briefing.
Summary
The research team have located new and large (~1km wide) plumes of outgassing methane, in areas not previously investigated, so this is not necessarily a recent development: at least, there are no previous data from these areas to compare the large plumes with.
That the area has seen warming over a prolonged time since the Holocene transgression and that there has been an additional, sharp recent warming event is well-documented. Whether there is increased outgassing caused by the additional recent warming is an important question that requires urgent investigation, a point indeed made in Shakhova et al's 2010 paper in Science - PDF - see last paragraph.
Further work will better constrain known current methane emissions to the atmosphere, currently estimated to be 8 Tg (1 Tg=1 million tonnes) per year. Clearly, because new sources have been identified, the figure is greater than 8Tg but how much so remains to be discovered.
A large (multi-gigaton) abrupt release event is considered possible, but when is not known. It is important to remember that hydrocarbons, including methane, migrate upwards through the Earth's crust from their source-rocks due to their low density. Basic oil geology tells us that recoverable oil and gas deposits occur where such upward migration has been blocked by an impermeable barrier (an oil- or gas-trap) such as a salt-dome or anticline including thick impermeable strata such as a clay-bed. In such places, the accumulation can build up to the point where the oil/gas is in an highly pressurised state - hence the "blowouts" that have been recorded over the years in some oilfields. What Shakhova is suggesting is that if buried gas hydrates destabilise, what could result is accumulations of pressurised methane capped off by permafrost, which because it is degrading might lose its effectiveness as a gas-trap. The same point would, I suggest, apply to non-hydrate derived methane i.e. gas that has remained as gas. What one would likely see in such a scenario would be a strong increase in outgassing, not in one great "burp" at one locality but via multiple pathways up through the defrosted sediment over a wide area.
David Archer, who has worked extensively with gas hydrates, looks at some release scenarios over at Realclimate, here and here.
Some afterthoughts
In February 2002, the then U.S. defence secretary Donald Rumsfeld famously said: "There are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns – there are things we do not know we don't know." He was lambasted and ridiculed for this at the time. However, at face value, all politics aside, the statement has more than a ring of truth to it.
My scientific background is in mineralogy, and I hope I may offer an analogy which illustrates what is meant by "known knowns, known unknowns and unknown unknowns". During the 1980s, I and a number of other mineralogists started to turn up samples of a mineral forming beautiful vivid pale-green spiky crystals. These were found at old metal-mines in Mid-Wales and the English Lake District: first at one locality then another and another. Despite having all the tools of mineral identification at our disposal, including x-ray diffraction and the electron microscope, we could not match it to any known species. Before we started to find specimens, it had lain there in the mine spoil, an unknown unknown, for centuries. Now we had specimens and could study it but not identify it, it was a known unknown. Finally, nearly twenty years after the first discovery, it became a known known - redgillite, named after one of its localities: a basic sulphate of copper with its crystallography and chemistry described in detail in the peer-reviewed journal Mineralogical Magazine.
above: from unknown unknown to known unknown to known known: redgillite, first described in 2004.
I offer this not as an off-topic distraction but as an example of the way in which science proceeds. In almost every case it is the same: only the timescale may vary. Something new is discovered yet is not fully understood. That uncertainty, however, gives pointers towards follow-up work: in time the uncertainties are whittled away one by one and the whole picture becomes clear. It is a journey which will be familiar to anyone working in any of the many branches of academic research. Shakhova and her colleagues have embarked on a journey of challenging proportions: less than a decade ago they discovered what had to that point been an unknown unknown. Now they have described some of the known knowns: we now know without doubt that methane is venting to Earth's atmosphere from parts of the ESAS seabed in copious quantities as a response - a positive feedback - to warming. Furthermore, they have identified many of the known unknowns for follow-up study. They have rightly called for the setting-up of an observational network in the area and they are planning their research for the coming years in a physically and logistically-challenging area.
Outside of science, people seem to work differently a lot of the time, wanting a black-or-white world where everything is a known known. That there are no absolute, cover-all conclusions yet from the ESAS might come as a disappointment to some. However, this is maybe a time to reflect that, 116 years ago, Arrhenius figured out that adding carbon dioxide to the Earth's atmosphere would raise Earth's surface temperature over time. That has been a known known ever since, reinforced by study after study after study, during which time we have raised carbon dioxide levels well beyond anything during the geologically recent glacial-interglacial cycles, so that we are now heading, decade by decade, towards the climate of the mid-Cenozoic, methane or no methane.
[DB] Fixed links.
[DB] Note: All graphics must include accompanying text which provides the context for why the graphic is relevant and germane to the htread on which it is posted. Future graphics-only or link-only will be deleted. FYI.
[DB] Fixed image ftp issues. Please resize images to no more than 500 pixels width. See here for posting tips.
[DB] Yes, saltspringson's comment was a nice example of a Gish Gallop. As such, should he/she feel up to the challenge of actually mounting a defense, the onus will be on he/she to mount said defense of each claim individually, on the most appropriate threads.
The age of the methane is of interest. If it is found to be greater than the 50,000 years that can be measured with carbon dating, this would point to a geological source such as outgassing from coal, shale or liquid hydrocarbon deposits. This would reinforce the idea that such seeps would be trapped under ice sheets as clathrates once the ice sheet had reached about 300m deep. Over the roughly 100,000 years that glacials last, a lot of methane can accumulated only to be released when a milankovitch nudge starts the melt.
It could be that no methane pulse would be seen in ice bubbles from Antactic or Greenland cores. The half life of methane is about 7 years and much of the oxidation is due to hydroxyl ions in the upper atmosphere. Whether this oxidation would continue inside the ice is a moot point. However, the top 70 or so meters of a forming ice sheet still exchanges gas from the atmosphere. The bubbles are not yet closed. This would tend to blur or even eliminate a methane signature in ice bubbles even if it did occur in the atmosphere.
Another unsettling effect of methane release, especially on the continental shelf is the loosening of the sediment both by the melting of the "clathrate permafrost" and the expansion of the gas, turning the sediment into a fluidized bed. Sudden slumps on continental shelves have been implicated in localized but very severe tsunamis. Will the coastal people of the Arctic ocean experience some of these in the not too distant future.