Lately I have been reading a lot about the Paleocene-Eocene Thermal Maximum, or PETM, which is my favourite paleoclimatic event (is it weird to have a favourite?) This episode of rapid global warming 55 million years ago is particularly relevant to our situation today, because it was clearly caused by greenhouse gases. Unfortunately, the rest of the story is far less clear.

Paleocene mammals

The PETM happened about 10 million years after the extinction that killed the dinosaurs. The Age of Mammals was well underway, although humans wouldn’t appear in any form for another few million years. It was several degrees warmer, to start with, than today’s conditions. Sea levels would have been higher, and there were probably no polar ice caps.

Then, over several thousand years, the world warmed by between 5 and 8°C. It seems to have happened in a few bursts, against a background of slower temperature increase. Even the deep ocean, usually a very stable thermal environment, warmed by at least 5°C. It took around a hundred thousand years for the climate system to recover.

Such rapid global warming hasn’t been seen since, although it’s possible (probable?) that human-caused warming will surpass this rate, if it hasn’t already. It is particularly troubling to realize that our species has never before experienced an event like the one we’re causing today. The climate has changed before, but humans generally weren’t there to see it.

The PETM is marked in the geological record by a sudden jump in the amount of “light” carbon in the climate system. Carbon comes in different isotopes, two of which are most important for climate analysis: carbon with 7 neutrons (13C), and carbon with 6 neutrons (12C). Different carbon cycle processes sequester these forms of carbon in different amounts. Biological processes like photosynthesis preferentially take 12C out of the air in the form of CO2, while geological processes like subduction of the Earth’s crust take anything that’s part of the rock. When the carbon comes back up, the ratios of 12C to 13C are preserved: emissions from the burning of fossil fuels, for example, are relatively “light” because they originated from the tissues of living organisms; emissions from volcanoes are more or less “normal” because they came from molten crust that was once the ocean floor.

In order to explain the isotopic signature of the PETM, you need to add to the climate system either a massive amount of carbon that’s somewhat enriched in light carbon, or a smaller amount of carbon that’s extremely enriched in light carbon, or (most likely) something in the middle. The carbon came in the form of CO2, or possibly CH4 that soon oxidized to form CO2. That, in turn, almost certainly caused the warming.

There was a lot of warming, though, so there must have been a great deal of carbon. We don’t know exactly how much, because the warming power of CO2 depends on how much is already present in the atmosphere, and estimates for initial CO2 concentration during the PETM vary wildly. However, the carbon injection was probably something like 5 trillion tonnes. This is comparable to the amount of carbon we could emit today from burning all our fossil fuel reserves. That’s a heck of a lot of carbon, and what nobody can figure out is where did it all come from?

Arguably the most popular hypothesis is methane hydrates. On continental shelves, methane gas (CH4) is frozen into the ocean floor. Microscopic cages of water contain a single molecule of methane each, but when the water melts the methane is released and bubbles up to the surface. Today there are about 10 trillion tonnes of carbon stored in methane hydrates. In the PETM the levels were lower, but nobody is sure by how much.

The characteristics of methane hydrates seem appealing as an explanation for the PETM. They are very enriched in 12C, meaning less of them would be needed to cause the isotopic shift. They discharge rapidly and build back up slowly, mirroring the sudden onset and slow recovery of the PETM. The main problem with the methane hydrate hypothesis is that there might not have been enough of them to account for the warming observed in the fossil record.

However, remember that in order to release their carbon, methane hydrates must first warm up enough to melt. So some other agent could have started the warming, which then triggered the methane release and the sudden bursts of warming. There is no geological evidence for any particular source – everything is speculative, except for the fact that something spat out all this CO2.

Magnified foraminifera

Don’t forget that where there is greenhouse warming, there is ocean acidification. The ocean is great at soaking up greenhouse gases, but this comes at a cost to organisms that build shells out of calcium carbonate (CaCO3, the same chemical that makes up chalk). CO2 in the water forms carbonic acid, which starts to dissolve their shells. Likely for this reason, the PETM caused a mass extinction of benthic foraminifera (foraminifera = microscopic animals with CaCO3 shells; benthic = lives on the ocean floor).

Other groups of animals seemed to do okay, though. There was a lot of rearranging of habitats – species would disappear in one area but flourish somewhere else – but no mass extinction like the one that killed the dinosaurs. The fossil record can be deceptive in this manner, though, because it only preserves a small number of species. By sheer probability, the most abundant and widespread organisms are most likely to appear in the fossil record. There could be many organisms that were less common, or lived in restricted areas, that went extinct without leaving any signs that they ever existed.

Climate modellers really like the PETM, because it’s a historical example of exactly the kind of situation we’re trying to understand using computers. If you add a few trillion tonnes of carbon to the atmosphere in a relatively short period of time, how much does the world warm and what happens to its inhabitants? The PETM ran this experiment for us in the real world, and can give us some idea of what to expect in the centuries to come. If only it had left more data behind for us to discover.

Pagani et al., 2006
Dickens, 2011
McInerney and Wing, 2011


23 thoughts on “The PETM

  1. Good article, thank you.

    Is the theory that the uplift of the Himalayas from the collision of India with Asia resulted in/triggered the PETM now considered wrong? I know it was controversial but I’m not up to date on any resolution of that.

    Somewhat beside the main point of the article but:

    “…carbon with 13 neutrons (¹³C)…”

    ¹³C is carbon with 7 neutrons and 6 protons. ¹²C is six of one and half a dozen of the other.

    • The Mid-Atlantic (volcanic) Ridge was erupting about then, as was the Siletzia (now Yellowstone) supervolcano, as the west coast of North America slid slowly west across it. All that had to do was warm the planet enough to begin a methane excursion; methane burps are self-reinforcing feedback loops, and like all such, they grow exponentially. We’re venting methane hydrates all over the Arctic, already, especially from the shallow continental shelf of Eastern Siberia. We stop burning fossils right now, and cool the planet as fast as we can, or it’s already too late.

  2. (is it weird to have a favourite?)

    Why would you think it weird? /puzzled

    Then, over several thousand years, the world warmed by between 5 and 8°C.

    Scary to think that we’re looking down the throat of a similar temperature rise in a very, very much shorter timescale…

    It is particularly troubling to realize that our species has never before experienced an event like the one we’re causing today. The climate has changed before, but humans generally weren’t there to see it.

    Ain’t that the truth!

    There was a lot of rearranging of habitats – species would disappear in one area but flourish somewhere else – but no mass extinction like the one that killed the dinosaurs.

    I think that’s a strange assertion — particularly in the light of what you say next. My own gut feeling — and that is all it is — tells me that such upheavals would cause huge changes that would surely result in many extinctions (and if we haven’t found any sign of that, maybe it’s just a case of it is there but we’ve not yet found it?) … although, perhaps my reaction is influenced by the knowledge of what homo fatuus brutus is doing today, which is, in a nutshell, giving few beasts any wiggle room to escape when it finds itself in a bad place! But I should stop rambling, since I have absolutely no expertise in this field :)

    Glad to see you back!

  3. Very interesting read!

    I finally got my first set of data to analyze for my thesis research- things certainly got delayed longer than I would have liked!

    Taking a look at the Global Mean data from the tsi files, I found some interesting behaviour, at least for the 2000-5000GtC cumulative emission groups, though I suspect it would also be present in the 1000GtC group as well, if I ran the model for longer?

    What is happening is that in all my peak and decline, as well as my overshoot scenarios (my PULSE scenarios just finished and haven’t been analyzed yet), temperature reaches a preliminary peak, and then levels off for a while, before there is a sudden jump in temperature again! Interestingly, the timing of this feature seems to be inversely proportionally related to the rate of emissions release? That is, runs with a faster rate of emissions release all seem to show a later timing than runs with a slower, more prolonged emissions trajectory. I suspect this might be to do with ocean thermal inertia, but haven’t as of yet had time to investigate this. What throws me off, however, is that the timing of this heat release is quicker for runs with a slower, but more prolonged emissions release?

    • Tyler,

      I don’t know what model(s) you’re referring to, but in the UVic model I’ve seen a sudden jump in surface temperature after hundreds of years. This is apparently due to heat accumulating in the Southern Ocean and eventually destabilizing the water mass, causing an overturning that releases deep ocean heat back to the surface. I don’t know if any AOGCMs also exhibit this behavior. I’d check to see if the timing of this event is related to the cumulative ocean heat uptake in the deep Southern Ocean, or with some critical temperature there (over what depth range, I don’t know).

  4. An alternative, or perhaps supplement, to the methane hydrate hypothesis for fuelling the PETM is the Antarctic permafrost feedback model that I wrote about here:

    This model accounts for carbon isotope excursion and the fact that the PETM was followed by a series of similar but smaller events that seem to have been triggered by orbital wobbles. But this model is speculative, of course, and is by no means universally accepted.

    There’s also an article by my SkS colleague Rob Painting on the PETM that is worth reading:

    In the comment thread of that article, there are some very helpful comments by Gerald Dickens, who is one of the main proponents of the methane hydrate theory. He confirms that the only mass extinction event caused by the PETM involved the foraminifera that lived in the depths of the ocean.

    The collision of the Himalayas did not cause the PETM, rather, the uplift of the mountains in the caused greatly increased rock-weathering during the Eocene and Oligocene that drew down the then high CO2 levels in the atmosphere, eventually producing the “ice-house” climate (causing the Antarctic ice sheet to grow) that has prevailed for the past 30 million years.

  5. Does the behaviour of the planet during the PETM falsify Hansen’s “oceans might end up boiling” statements? If the planet “was several degrees warmer, to start with, than today’s conditions” and it then went on to warm 5-8 degrees C without causing a runaway, then his worst fears cannot be likely…

    • Remember that the rate of warming matters too. Positive feedbacks tend to respond more quickly than negative feedbacks (think melting permafrost vs carbon sequestration into the ocean floor) so if you change the temperature too quickly, the stabilizing effect of negative feedbacks is limited. That’s not to say that boiling of the oceans is probable or even plausible; just that it’s not necessarily impossible.

  6. Yes very nice article. But are foraminifers “snail-like”? I understood they are protozoans… surely not molluscs :-)

  7. “is it weird to have a favourite?”

    Mine was the Eocene-Oligocene transition – the beginning of Antarctic glaciation – out of southern-hemisphere patriotism.

  8. I don’t think it’s a given that the methane clathrate (aka methane hydrate) had to melt in order to release large amounts of methane. If I recall correctly the recent research by Shakhova and Semiletov suggests that there are also large amounts of free gas trapped under pressure under the seabed. If the containment was ruptured through either tectonic activity (or conceivably by puncturing the frozen permafrost cap through thawing) it seems possible large amounts of methane (gigatonnes) could be released very abruptly indeed.

    Just one of several rather ominous things that appears to be escalating in recent years.

  9. If you have time, I recommend the 2005 book by Michael Benton When Life Nearly Died. It is a bit of a scientific detective story about the discovery and then the search for the cause of the mass extinction. A gripping read.

  10. From the book details it seems that that Michael Benton book is about the end-Permian extinction. Indeed the biggie, but nearly five times longer ago than the PETM.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

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