Thursday 2 February 2012

Arctic Methane: Imminent, Abrupt and Massive Release?

In April 2008 Natalia Shakhova and colleagues (Shakhova et al 2008) presented data to the European Geosciences Union General Assembly in Vienna about the threat posed by methane clathrates in the East Siberian Arctic Shelf (ESAS). Their abstract read:
...we consider release of up to 50 Gt of predicted amount of hydrate storage as highly possible for abrupt release at any time. That may cause ~12-times increase of modern atmospheric methane burden with consequent catastrophic greenhouse warming.

When I first read that, the impact was the same as it would be for any reasonable person reading such a statement, coming as it did from people who are acknowledged experts in their field. It was only years later that I finally downloaded and read a copy of David Archer's "Methane hydrate stability and anthropogenic climate change." (Archer 2007) Having just finished reading a stack of papers on Arctic methane I'm still of the opinion that Archer's paper is required reading for anyone interested in this matter, and certainly should be read before people call 'apocalpyse' over Arctic methane. Archer 2007 is a summary paper that concludes that the release of methane is more likely to be a chronic release rather than a catastrophic one.

For this issue a key point of Archer 2007 concerns the mechanisms by which methane could be released from marine methane clathrates. The first issue is that of temperature change rates of the ocean, however for the ESAS this is less of an objection, the Arctic is undergoing rapid climatic warming, its ocean is in a basin, and the ESAS is a shallow area of sea subject to influx of warm waters from the inflowing rivers. Indeed Semiletov et al (2012) note: "The ESAS acts as a Lena, Kolyma and Indigirka estuary." Archer then goes on to discuss pockmarks and landslides, however these issues are necessarily secondary to the thermodynamics of getting surface warming to impact buried clathrates in layers of sediment beneath the ocean. Here both Archer 2007 and McGuire et al seem to be in agreement, the melting of marine clathrates and resultant emission of methane is likely to be a relatively slow (chronic) problem as opposed to a rapid (catastrophic) threat. It is here that Shakhova et al 2008 appears to give cause for concern, as the authors suggest a mechanism by which large amounts of methane could be at risk of fast-track warming and release despite the slow nature of warming for the bulk of the sediments.

The abstract of Shakhova et al's 2008 presentation is talking about a rapid and imminent degassing, not rapid and imminent on a geological scale, but on a human scale. The key to this is that they discuss not only a large flux of methane but a large atmospheric concentration in response. As methane is a reactive gas, a large flux over a long period would not produce a large atmospheric flux. Shakova et al 2008 is not talking about massive widespread release of sedimentary methane, but release from limited areas where the sediment column is disturbed. Their abstract states:

The total value of ESS carbon pool is, thus, not less than 1,400 Gt of carbon. Since the area of geological disjunctives (fault zones, tectonically and seismically active areas) within the Siberian Arctic shelf composes not less than 1-2% of the total area and area of open taliks (area of melt through permafrost), acting as a pathway for methane escape within the Siberian Arctic shelf reaches up to 5-10% of the total area, we consider release of up to 50 Gt of predicted amount of hydrate storage as highly possible for abrupt release at any time.
(Gt = Gigaton, 1000,000,000 tons)

How do they come up with the stated 'not less than 1,400 Gt of carbon"?
Amount of methane hydrate deposited beneath and/or within submarine relic permafrost is estimated to be at least 540 Gt. Amount of free gas, accumulated beneath the hydrate deposits, is expected to be about 2/3 of the amount of hydrates or 360 Gt. Additionally as much as 500 Gt of carbon could be stored within as minimum as a 25 m-thick permafrost body of this type.
Shakhova & Semiletov 2006, pg228, reference Kvenvolden and Grantz as the source of their 'at least 540Gt'. They then state that "a total amount of 10^15 m^-3 of methane (540 Gt of carbon) exists in sediments of the offshore Arctic Basin (Knevolden 1988)." This is confusing as Knevolden & Grantz is a 1990 paper, and Knevolden 1988 uses Arctic inventory to estimate global inventory. However the Arctic's store of methane is not all in the ESAS, so the actual figure appears to be somewhat less than 540Gt, by how much less I do not know.

In their 2004 paper, Hornbach et al estimate an "upper bound" on the amount of free methane gas below hydrate regions, this paper is relevant with regards Shakhova et al 2008 stated "2/3 of the amount of hydrates or 360Gt". Hornbach et al's upper limit is stated to be "about one-sixth to two-thirds of the total methane trapped in hydrate", thus potentially substantially reducing the 360Gt of free methane below the clathrates. Furthermore Hornbach et al figure 3 and associated text shows that the layer of methane gas beneath the clathrate layer gets thicker the deeper the bottom-simulating reflection (BSR) is, however in areas around faults the gas layer is not dependent upon BSR depth (and hence thickness of the hydrate layer) it remains on average about 35m thick. I should explain that the BSR is a seismic feature that mimics the sea floor and is indicative of the base of the hydrate stability zone (see below). The point being that Shakhova et al 2008 states that they have used fault zones and open taliks to estimate the fraction of the gross amount of methane that is likely to undergo rapid emission. However if my understanding of Hornbach et al is correct then the concentration of free methane around such faults (and presumably taliks) is likely to be less than the stated 2/3 (maximum) of the amount of hydrates. This is because the pressure of free methane below the hydrate stability zone required to force tectonic faults to slip (or presumably force taliks to vent) is lower than the pressures that can build up where there are no faults. Hence around faults the amount of free methane is less than in undisturbed sedimentary basins. So by using the area around talkis and faults the depth of the gas layer would be even less than Hornbachs estimate of a maximum of 1/6 to 2/3. This would have a negligible effect on the initial 1400Gt estimate but would further reduce the 50Gt estimate of that methane at risk of abrupt release at any time.

Finally the inclusion of 500Gt carbon stored in a 25m thick permafrost is ambiguous as it does not specify whether this is carbon as methane or organic carbon, perhaps it was specified in the lecture the abstract refers to.

In summary if we accept for the moment that an abrupt release is imminent, I am not convinced that the amount at risk of such an abrupt release is actually 50Gt, my working assumption is now that the the amount at risk is much less than 50Gt. Leaving aside these quibbles over scale, there is likely to be a problem with limited regions of the ESAS, but what of the bulk of methane in that region?

Shakhova & Semiletov 2006 state that the sediments of the ESAS are mainly find grained mud and silt with around 1 - 2% organic carbon. As is noted in Archer 2007, fine grained muds and silts with methane hydrates in them have a low percentage of hydrate mixed with the silt, whereas course sands can allow the accumulation of larger masses of hydrate. This means that if some process exposes the hydrate holding sediment it is unlikely to float if in fine grained sediment because the mass of the sediment keeps the hydrate from being bouyant. So fine grained sediment is less likely to rapidly break up, favouring a chronic release as the upper edge of the hydrate stability zone moves down with warming of the sediment column from above. The hydrate stability zone being the region in which methane hydrates are stable between the warmer surface layers (warmed by the ocean) and warmer deep layers (warmed by geothermal heat). As mentioned previously, below the stability zone bubbles of methane can be found, this is a result of the continual geothermal heat flux 'thawing' hydrate that was previously in the stability zone. The free methane is then trapped beneath the impermeable hydrate above. So it seems reasonable to suppose that in undisturbed sediment the fate of hydrates under warming is to produce a chronic source of methane emmission as warming from the ocean conducts downwards into the sediment.

In the comments over at Realclimate it has been argued by some that we will most likely see an exponential increase of methane released from the Arctic. I strongly disagree with this. Using that viewpoint it is likely that when we do see large releases of methane from the ESS they will be the start of some kind of massive 'unzipping' of hydrate methane. However the evidence suggests that any large releases may occur episodically and be followed by periods of lower emissions. There are features in the ocean that suggest there have been large releases of methane, these are pockmarks and landslides.

Archer discusses pockmarks on the ocean bottom that seem to be associated with the release of methane, they range in diameter from meters to kilometres. One at Blake Ridge is estimated to have been able to release up to 1Gt carbon as methane, however pockmarks could be the result of either rapid explosive release or slow chronic release. I presume here that a localised penetration of warmth into the sediment could cause a small localised release that would then expose and disrupt sediments surrounding the initial event, this destabilisation then propagating outwards. If hydrate bearing sediments are then exposed to the ocean they can be brought out of the stability zone by either release of pressure from above sediment load or warming of the ocean. Following the reasoning of Hornbach et al any localised destabilisation could then lead to explosive release of underlying pressurised gasseous methane which could add to disruption of the surounding sediment column and lead to a signicant pulsed release of methane. However what the pockmarks suggest is that this is not an indefinitely self-sustaining process - each event does reach a natural end, and the biggest such event could have been no more than 1Gt methane.

Another candidate for rapid release of methane is where landslides occur in the sediment due to slope instability on undersea slopes. Here a long range of slope could destabilise exposing a large burden of hydrate taking it out of the stability zone, again by warming and/or pressure release. High sediment loads in river deltas can deliver sediment faster than the entrained fluid can be squeezed out, thus causing slides. As mentioned earlier, the ESAS is in effect the estuary for three rivers, so sediment deposition rates are high and it is reasonable to suppose that some of the slopes could destabilise due to entrainment of fluid. This is a longstanding situation, and as I'll discuss in a following post there is reason to suspect that the last few decades of rapid warming in the region have had less of an impact than the equally longstanding post-glacial warming. So whilst the methane contained in the upper layers would not be a problem, such slides can destabilise the hydrate containing layers beneath and the free gas below that, but the idea that this is likely to lead to an outgassing of such magnitude that it would have catastrophic climatic effects is not strongly supported. In terms of the public discussion, the most fequently referenced landlsides are the Storegga events, the most recent of these (what Archer refers to as 'Storegga proper') occurred about 8150 years ago and resulted in pockmarks, indicating gas expulsion. There was no catastrophic warming in response, indeed the 8200 Year Event happened at around that time and methane levels in the atmopshere decreased.

In Shakhov et al 2010, figure 3 is presented as evidence of methane pulses, these may seem to support the idea that a rapid catastrophic release is likely. Indeed recently Natalia Shakhova has been interviewed by Skeptical Science, and she noted that "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." However this does not automatically imply massive release of tens of gigatons of methane over short periods of years. Even a chronic long term release of methane is likely to be interspersed with such events, events which have happened since the Paleocene-Eocene Thermal Maximum, but have not lead to catastrophic climate changes leading in turn to mass extinction.

Continue to part 2.

Archer, 2007, "Methane hydrate stability and anthropogenic climate change."
Hornbach et al, 2004, "Critically pressured free-gas reservoirs below gas-hydrate provinces."

McGuire et al, 2009, "Sensitvity of the carbon cycle in the Arctic to climate change."

Shakhova & Semiletov, 2006, "Methane release and coastal environment in the East Siberian Arctic shelf."

Shakhova et al, 2008, "Anomalies of methane in the atmosphere over the East Siberian shelf: Is there any sign of methane leakage from shallow shelf hydrates?"

Shakhova et al, 2010, "Extensive Methane Venting to the Atmosphere from Sediments of the East Siberian Arctic Shelf."

No comments: