PETM climate warming 56 million years ago strongly tied to igneous activity

Part 1 – The Rabbit Hole

This 3-part post is expanded from an article I originally wrote for Quanta Magazine and features quotes from interviews that appeared in that piece.

Back in 1991, James Kennett of UC Santa Barbara and Lowell Stott of UCLA reported a: “…rapid global warming and oceanographic changes that caused one of the largest deep-sea benthic extinctions of the past 90 million years.” This warming event 56 million years ago became known as the “Paleocene-Eocene Thermal Maximum” or “PETM.” Over the intervening years a considerable body of research has shown that the planet warmed by about 5ºC (9ºF), oceans acidified, sea levels rose, and land suffered an increase in downpours, while hot seas and dead zones stressed ocean life. Land animals underwent a high rate of extinction and replacement by dwarf species, and tiny shell-making creatures on the sea bed (benthic foraminifera) went extinct.

It’s the closest natural analog to modern climate change in the last 66 million years, but what caused it?

There have been four main ideas: methane emissions after the planet crossed a climate tipping point; volcanic activity; a comet impact; or burning peat. Evidence for the volcanic trigger has been growing over recent years, and now a new study ties the underground part of volcanic activity to the PETM in a quantified way, from mantle motion to climate warming, calculated independently from - yet matching - the sedimentary record of the carbon cycle and temperature changes. The new study, by a team from the University of Birmingham, UK, has helped to convince more scientists that volcanic activity probably triggered the warming, and this knowledge helps constrain its lessons for our own climate change today.

This post is in 3 parts – in part 1, I review competing explanations for the cause of the PETM, in part 2 I report on the results of the new study, and in part 3 I report the reactions of other scientists, and what the new results tell us about our warming in the years to come.

The Arctic looked like this 56 million years ago – modern Baldcypress Swamp in Louisiana. Photo by Jan Kronsell CC BY-SA 3.0

The Arctic looked like this 56 million years ago – modern Baldcypress Swamp in Louisiana. Photo by Jan Kronsell CC BY-SA 3.0

Causes and consequences

The trigger for the PETM has been debated since its discovery. The main ideas have been: methane emissions after crossing a climate tipping point; volcanic activity; a comet impact; or burning peat.

Methane tipping point:

The leading idea for what caused the PETM has long been a rapid release of methane from a reservoir of methane ice known as “methane clathrates” or “methane hydrates.” Scientists worried that, if that was the case, a similar thing could happen today, where seabed methane clathrates could be destabilized by human-driven warming, greatly magnifying our own climate change.

But even Gerald Dickens, the original proponent of the clathrate hypothesis in 1995, published a paper in 2011 titled: “Down the Rabbit Hole…” where he stated that it “lacks proof.” The main leg that hypothesis had to stand on was the carbon isotope signature in sediments around the world. This showed a strong increase in the isotope carbon-12 relative to carbon-13, indicating a big release of organic, rather than volcanic, carbon. Methane from clathrates is organic and very rich in carbon-12, so less is needed to explain the shift in the isotopes than if the carbon was volcanic, which has much less carbon-12. Yet that’s also a problem for clathrates – they are so rich in carbon-12 that if the PETM was triggered by a purely clathrate source, the volume you need to reproduce the isotope shift in sediments is insufficient to generate the warming. There must be additional carbon from a non-clathrate source mixed in.

Burning methane hydrates source USGS

Methane hydrates (clathrates) on fire - USGS

"there are no any two records which share the consistent pattern"

One problem with the clathrate hypothesis is that it, in turn, needs an external trigger to destabilize the clathrates. The idea was that Earth’s climate warmed just before the PETM, enough to destabilize seabed methane clathrates and unleash the really strong warming as a positive feedback. A crucial observation in favor of this is that the signal for warming precedes the shift in carbon isotopes in some mid- and high-latitude sedimentary records, suggesting that the organic carbon signal was in response to, not the initial driver of, the warming. But this has been hard to pin down definitively because of the short timescales involved compared to the slow accumulation of sediments, and the fact that “there are no any two records which share the consistent pattern of the [isotope shift],” as Chen et al put it in their 2014 paper. More recently it has been shown that, particularly for lake deposits like the ones studied by Chen et al, the temperature signal in some biochemical indicators of ancient temperature can be confused by how waterlogged they are.

Chen et al proposed that volcanic eruptions may have provided the initial carbon to destabilize clathrates, whereas the leading idea has long been that Earth’s orbit around the Sun was in an unusually warm configuration that caused Earth to cross a climate tipping point that liberated seabed methane from clathrates into the atmosphere.

The reason scientists thought that orbits must be involved is because the PETM is the most prominent of a number of warm periods (dubbed “hyperthermals”) spread across 3 million years of the Eocene, several of which are in sync with wobbles in Earth’s orbit (Milankovitch Cycles).

the PETM is quite different from those other hyperthermals

But chaos swamps calculations of orbits older than about 50 million years, and many sedimentary records have gaps, so it’s possible to calculate either an orbital configuration that supports an orbital push to begin the PETM, or one that is out of phase. Others have argued that the PETM is quite different from those other hyperthermals and so requires a different explanation: it was twice as big, and much more abrupt, and they argue that those subsequent hyperthermals are just the normal orbital climate drumbeat found in sediments throughout geological time, without the need for clathrate release to explain them.

Another problem for the clathrate hypothesis is that it needs a large reservoir of clathrates to be there in the first place. We know they exist in today’s seabed but the Paleocene ocean was much warmer than today’s, so the reservoir was probably as good as empty. A third problem is that even if there was a release of seabed methane, about half would never make it to the atmosphere (so double the reservoir is needed to explain the warming) and it would also have been too slow.

Last year it looked, briefly, like barium could be the smoking gun of seabed methane release. Joost Frieling of Utrecht University, with colleagues, observed that barium was deposited in sediments during the PETM at triple the normal rate. Since water in sediments around methane clathrates contains an abundance of dissolved barium, this uptick in barium could be a sign of seabed methane release. Unfortunately, as Luke Bridgestock of Oxford University and colleagues observed in a paper published around the same time, barium doesn’t last in the ocean long enough to explain the protracted barium burial, and there’s no evidence for an increase in barium sulfate saturation in ocean water over that time interval. The increased barium burial, Bridgestock et al argue, is instead due to increased biological activity after the initial phase of the PETM, and not seabed methane at its onset.

Permafrost is another reservoir of carbon-12 that could have been released by warming from orbital wobbles, much like clathrates. But here again there probably wasn’t much permafrost in the Paleocene because fossils show that the Arctic supported alligators, giant tortoises, palms and swamp-cypresses, while Antarctica was mostly ice-free and forested. Antarctic glaciation seems to have begun some 20 million years after the PETM and even mountains close to the South Pole were tree-covered well into the Miocene.

Comet impact:

an instant release of carbon such as from a comet impact doesn’t match the protracted sedimentary record

In 2013 Morgan Schaller of Rensselaer Polytechnic Institute in New York and James Wright of Rutgers University in New Jersey proposed that a comet impact may have driven the PETM but, as responses to their paper showed, an instant release of carbon such as from a comet impact doesn’t match the protracted sedimentary record of carbon release and warming. There are indeed traces of an asteroid or comet impact in some PETM sediments dated to about the right time, tentatively linked to the Marquez Crater in Texas, which reportedly overlies a petroleum reservoir. But the crater is only 12.7 km wide and would be too small to contain anything close to enough carbon. Geochemical tracers such as mercury are sustained or pulsed in a manner not consistent with a singular, instantaneous comet impact, but consistent with volcanic activity. Schaller recently proposed that an impact may have served as an initial trigger, with volcanic activity doing the main body of work. This echoes a recent theory linking the end-Cretaceous Chicxulub impact to eruptions in the Deccan Traps, but the comet idea doesn’t have much traction with most PETM scientists I’ve spoken to over recent years.

Comet Ison November 2013 - NASA

Peat fires:

In 2003 Andrew Kurtz of Boston University and colleagues suggested that burning peat might be the source of the organic carbon, rather than methane clathrates or permafrost. But the absence of charred peat from that time argues against a widespread peat conflagration as a trigger for the PETM, and Arctic wildfires seem to have increased later in the PETM, rather than at its onset.

Smoldering peat fire VA 2011

A smoldering peat fire on Great Dismal Swamp National Wildlife Refuge, Virginia, 2011 - Chris Lowie/USFWS

Volcanic Activity:

A volcanic trigger for the PETM was first proposed in 1993 and unlike the other theories there’s a substantial body of physical evidence to support it.

The vast assemblage of igneous rocks in Greenland, the British Isles, and under the North Atlantic seabed, are collectively known as the “North Atlantic Igneous Province” or “NAIP.” Its initial phase around 60 million years ago created the spectacular columned basalt landscapes of Giant’s Causeway in Northern Ireland, and Scotland’s “Fingal’s Cave” that inspired Mendelssohn’s composition.

Extent of NAIP based on Hansen 2006 Horni et al 2017

Present-day extent of the North Atlantic Igneous Province based on Hansen et al 2009 and Horni et al 2017. Red - igneous rocks including flood basalts, sills, dikes and central volcanic complexes.

In addition to the rocks, there are geochemical traces of volcanic mercury and osmium in sediments that formed at the time, and buried igneous rocks of that age are mapped in many seismic scans and encountered in many oil exploration boreholes in the North Atlantic region.

The NAIP is an example of a “Large Igneous Province” or “LIP.” LIPs are behind most of the large climate warming events since the dawn of animals, and behind most mass extinctions, so a link between the NAIP and the PETM would fit that oft-repeated pattern.

But coincidence isn’t causation, so the question remains: how exactly do you go from volcanic activity to climate warming? And how do you explain the apparently contradictory evidence pointing to a large release of organic – not volcanic – carbon?

To be continued in part 2