The Paleocene, before the PETM, was warmer than now, it's been hard to pin down by how much. Robert Rhode has produced this useful graphic, which suggests temperatures were of the order of 4 degC higher than preindustrial, however the PETM itself seems to have involved a temperature increase of about 6degC, the temperature rise in that graphic is much less. The PETM was a minor extinction event, there were no large losses of species on land, and the ocean extinction was restricted to bottom dwellers in the deeps of the oceans in certain regions, probably due to anoxia, which was likely due to changes in ocean overturning circulation. However with regards land species, during the PETM, mammals radiated profusely which indicates environmental stress, changes in the environment driving evolution to suit new niches and niches made available by the decline or movement of other species.
At the time of the PETM the Arctic was temperate, with temperatures not falling below zero, even during the winter. This presents the equable climates problem, in that models are not able to reproduce conditions; they produce tropics that are too hot, or poles that are too cold. Abbot & Tzipperman propose a cloud radiative feedback, with carbon dioxide levels and ocean/atmospheric heat flux keeping a blanket of cloud over the Arctic throughout the winter. The cloud backradiates infra-red radiation to the surface, keeping the region warm. In their 2007 paper they find that atmospheric heat transport of 140W/m^2 plus doubled CO2 could keep the Arctic ice free during the winter, for comparison, current atmospheric heat transport is of the order of 100W/m^2 at 80degN. Given what appears to be the unfolding of a rapid transition of the Arctic to a seasonally sea ice free state, finding of possible bifurcations (tipping points) in the transition to a perennially sea ice free state (Eisenman 2011), and the increasing evidence that methane hydrates in the Arctic are thawing, plus the near certainty that they will more actively do so in the decades to come: The prospect of substantial emissions of methane from the Arctic is a very real prospect. So the question posed is can we avoid a re-run of the PETM?
The following graphic is taken from this article. It's worth noting that it states 25 petagrams are being emitted each year, that's 25 gigatons (Gt the unit I'll use in this post). However from the CDIAC data used in the following paragraph, actual emissions from fossil fuels in 2008 were 8.75 Gt carbon, or 32.1 Gt CO2. This shows visually what Skeptical Science have noted, based on Cui et al, 2011, that current emissions dwarf the PETM's rate of emission, ref.
Using data from CDIAC: Total human emissions from fossil fuels are 346.8Gt carbon up to 2008, of this about 50% is from coal, 35% from oil. Gas totals 12%, but by 2008 was contributing 18%. Peak Oil is about now, but we may face a long plateau due to non-conventional fossil oil production because of the high price making previously uneconomic sources feasible. So we're about half the way through Oil. Gas is decades off peaking, we can conservatively double that. Coal again can be at least doubled. So the bare minimum, using the most pessimistic assumptions about fossil fuel availability, that we can emit from fossil fuels is about 693.6Gt, given the uncertainties I'll round that to 700Gt. If we take preindustrial to be 260ppm, then a doubling is 520ppm, coincidentally our bare minimum emissions should be able to double CO2 levels because the current ~390ppm is halfway to 520ppm from 260ppm. Assuming a 3degC climate sensitivity if we burn this minimum amount of fossil fuels we'll commit ourselves to well over 2 degrees warming from pre-industrial, the equilibrium warming takes centuries to manifest itself. However that still saves us from a re-run of the PETM, right? Because even if we double our minimum 700Gt fossil fuels to give a total of 1400Gt, that still only takes us under one third of the way to equivalent PETM emissions.
Well before we look at the issue of methane emissions it's as well to look more closely at the 5000Gt target for total emissions during the PETM. The source of the 5000Gt seems to be a 2006 paper from Higgins & Schrag. They examine the PETM using data available at the time, they discount methane hydrates as a major player and focus instead on thermogenic methane and CO2 emissions from vulcanism due to the seperation of Greenland from Europe. These sources have a lower d13C isotopic signature, the d13C signature is a measure of the ratio of carbon 12 and carbon 13 isotopes relative to a standard. The lower the d13C signature the more depleted in 13C, as a heavier isotope 13C is energetically less favoured in chemical processes, hence it is depleted in certain biological processes. Because the carbon released during the metamorphosis of rocks surrounding the volcanic activity would have been less 13C depleted than an equivalent methane source, more of it would be needed to explain the excursion in d13C observed during the PETM, hence the 5000Gt. Whereas Higgins & Schrag note that Dickens had previously found that only 1000 to 2000Gt of methane would be needed to reproduce the PETM d13C excursion.
Fig 1. The PETM d13C excursion, from figure 1 of Higgins & Schrag 2006.
So why do Higgins & Schrag find that oceanic clathrate methane was probably not a serious player in the PETM?
The first issue has been subsequently addressed in later work: Archer & Buffet previously estimated low clathrate inventories in the Eocene, due to the warmth of the oceans. However last year Gu et al published a study in which they found that because of the warmth of the oceans, biological activity, from carbon deposition on the ocean floor to methanogenic bacteria, was enhanced. So whereas now methanogenesis is found in cold deeps, during the Palaeocene it was theoretically capable of sustaining large deposits of clathrates, possibly larger than today, in warmer seas.
The second issue is large changes in the Calcite Compensation Depth (CCD). The CCD is the depth below which the disolution of calcite exceeds the deposition of calcite through the ocean column, this leads to a deficit of calcite at depth. During the PETM in the South Atlantic the CCD rose by 2000m in less than 10k years. Higgins & Schrag take this as indicative of a massive influx of CO2, due to an equally massive release of CO2. However a different interpretation may be offered by Sexton et al in their 2011 paper.
Sexton et al examine warming events, hyperthemals, during the Eocene. The PETM was a hyperthermal but was substantially different from the Eocene hyperthermals that they examined; the recovery from the PETM took over 100k years, whereas the other Eocene hyperthemals were up to around 40k years in duration. This suggests that whereas the PETM involved an external source of carbon, such as methane hydrates, these shorter events did not involve such a source. Sexton et al propose that the hyperthermals involved movement of carbon between the atmosphere/land/ocean reservoirs, hence whereas an external perturbation would need to be removed by geological timescale processes such as weathering, the carbon in the hyperthermals could move back from atmosphere to ocean/land, a much quicker process. Crucially this process resulted in substantial changes in the CCD, with dissolution intensity being found to be greatest in the South Atlantic. Sexton et al propose that changes in meridional overturning were probably responsible.
So it seems that the main objections of Higgins and Schraggs that lead to them placing methane clathrates aside to examine the greater implied emissions of volcanic activity have since been re-examined. Unfortunately that puts methane clathrates back 'on the table'. Furthermore it reduces the total carbon burden required to re-run the PETM, emissions may have been of the order of 2000Gt, rather than 5000Gt. So the low-end estimate of 700Gt becomes a more feasible initiator for a total 2000Gt emission. It's worth noting that the fossil fuel emissions don't take into account natural processes coming into play and emitting CO2 or methane into the atmosphere, or changing the amount sequestered in the deep ocean.
Archer 2006 finds that for ocean sediment methane at present:
The data seem strong enough to say that the Kvenvolden/MacDonald “consensus” value of 10,000 Gton is probably too high. The very high 78,000 Gton estimate from Klauda and Sandler is inexplicable. A potential range of hydrate inventories must span about 500-3000 Gton C, with the inclusion of bubble methane adding perhaps a similar amount.But this is just ocean methane hydrates. Lawrence et al find that during Rapid Ice Loss Events warming extends over the areas covered by land permafrost. Weather variability means that the Siberian pattern in NCEP/NCAR hasn't manifested itself as shown below, but the Canadian pattern is strong.
Fig 2. Figure 1c of Lawrence et al 2008, showing October/November/December warming in models during and outside of Rapid Ice Loss Events.
I think we are seeing the start of a process that, in terms of human timescales, will continue indefinitely. The millennia to come will see vast stores of frozen carbon in the Arctic melting and being release as carbon dioxide and methane.
But will what we face be as bad as the PETM? We're starting from lower baseline temperatures, and we probably haven't enough fossil fuels to get atmospheric CO2 up to near 2000ppm, so on the face of it, it seems it won't be as bad. The oceans are certainly colder, for example in the tropics at Demerara ridge temperatures were around 13degC at 3000m depth, equivalent temperatures are today around 5degC. However contrary to this, I think a re-run of the PETM in terms of species loss, and equivalent human impacts is the best scenario we can hope for. The reality will probably be much worse.
In opening this post I mentioned that the PETM was a minor extinction event, around 40% of benthic (deep ocean floor) foraminifera were lost. It's worth comparing that with the current situation, since the 1950s about 40% of phytoplankton have been lost from the non-coastal regions of almost all the oceans (Boyce et al, 2010), as all readers of this blog will know, phytoplankton are a major base of the ocean food chain. Then we have the effect of overfishing.
Then on land we're already seeing Hansen's Climate Dice come into play, I've blogged on it here, some 10% of the globe are now covered by 3 standard deviation warming events. We're seeing mid latitude effects from Arctic amplification that may be involved in floods, persistent weather patterns, and extreme winters. Tamino has done an excellent job of cataloguing weather disaster increases, wildfires, heatwaves, that's not to mention other bloggers and scientists doing sterling work cataloguing what's going on. These changes are happening after only around 0.8degC of AGW, with the last decade seeing surface warming abate due to the effects of the El Nino, sulphate pollution, and low solar output, e.g. Foster & Rahmstorf 2011 & Kaufman et al 2011. What will happen by the middle of this century? How much acceleration of warming, and further global weirding will the coming decades bring?
As the first graphic of this post shows, the pace of change we are forcing is far greater than in the PETM. With the impacts above intensifying, human agriculture covering 40% of land, and increasingly impinging on what wilderness that is still out there, we will probably see a greater mass extinction then during the PETM. We're giving natural systems neither the space nor time to adapt. Furthermore the super-interglacial warming we are prompting is driving the planet from a glacial/interglacial ecosystem, evolved over the last several million years, to a much warmer state, whereas the PETM was a warm event from an already warm baseline.
I still remain unconvinced by the claims of some that within decades we'll see a catastrophic methane blow out, it's un-necessary to reiterate my reasoning which is provided here, and here. The issue is not the amount of free gaseous methane that is available for a Hollywood disaster movie style blow out, but the amount of fossil clathrates stored during the Quarternary, this has the potential as it melts in the super-interglacial warmth of the anthropocene to massively amplify our emissions. And as Sexton et al find, if we warm the oceans enough they may add to the effective emissions as the overturning circulation reduces, causing abyssal anoxia and reducing sequestration of carbon into the abyssal deeps. But the PETM shows that this is a slow process when measured in terms of a human lifetime. Both the clathrate containing sediments and the oceans have large thermal mass and take a long time to respond to warming.
The current events in the Arctic region presage the awakening of a slow, cumbrous, but inexorable beast that will be with us for longer than human history, gradually adding to the damage we have done, warming the climate well beyond 3 degrees. The idea of stopping at 350ppm, or 2 degrees, of staying in safe levels of AGW will be totally out of our hands within a matter of decades. Indeed in view of what is practically achievable, given the time needed to shift from fossil fuels, time to develop and roll out alternatives, the problem that as the rich move off fossil fuels the price will drop, and the poor will take up the slack in consumption: It is already too late in terms of the "art of the possible". The notion of reducing to 350ppm, or limiting to 2 degrees is as dead as the broader aim of avoiding dangerous climate change.
How much further we drive the process depends on how much of the available fossil fuels we burn. But we are already at the stage where, barring a miracle, we are committed to dangerous climate change. As I fully expect us to burn all the fossil fuels economically and technically available, I expect that the best outcome is a re-run of the PETM.
The good news? We haven't anywhere near enough fossil fuels to re run the End Permian extinction event. That we cannot do so due to circumstance rather than due to intelligent choice is not something humanity can feel justly proud of.
Abbot & Tziperman, 2007, "Sea ice, high-latitude convection, and equable climates." PDF
Archer, 2006, "Destabilisation of Methane Hydrates: A Risk Assessment." PDF
Boyce et al, 2010, "Global phytoplankton decline over the past century." PDF
Cui et al, 2011, "Slow release of fossil carbon during the Palaeocene–Eocene Thermal Maximum" PDF
Eisenman, 2012, "Factors controlling the bifurcation structure of sea ice retreat" PDF.
Foster & Rahmstorf, 2011, "Global temperature evolution 1979–2010" PDF
Gu et al, 2011, "Abundant Early Palaeogene marine gas hydrates despite warm deep-ocean temperatures." PDF
Higgins & Schrag, 2006, "Beyond methane: Towards a theory for the Palaeocene–Eocene Thermal Maximum" PDF
Kaufman et al, 2011, "Reconciling anthropogenic climate change with observed temperature 1998–2008" PDF
Lawrence et al, 2008, "Accelerated Arctic land warming and permafrost degradation during rapid sea ice loss." PDF
Sexton et al, 2011, "Eocene global warming events driven by ventilation of oceanic dissolved organic carbon." PDF