What Caused the Paleocene-Eocene Thermal Maximum (PETM)?

The Paleocene–Eocene Thermal Maximum (PETM) was a period of time beginning about  56 million years ago. Temperatures increased by 5–8°C[2][4] due to a large excursion of biogenic carbon. Temperatures increased extremely rapidly, and the perturbation of the carbon cycle led to ocean acidification and a mass extinction of benthic foraminifera. The warming event occurred suddenly, geologically speaking, perhaps in as little as 10,000 years[12], making it one the most rapid warming events detected in the Phanerozoic. The extreme warmth of the PETM lasted less than 220,000 years before returning to "normal" Eocene levels. The rapid warming warming associated with the PETM makes it a good analogue to current warming, so I think it would be helpful to cover this event as well as provide a bibliography for further reading on the subject. The PETM is also one among many examples in geologic history where it is clear that GHGs were driving global warming. CO2 led (and drove) the warming event; the evidence cannot be explained by some other forcing causing warming, and the oceans outgassed CO2 as a feedback. Toward that end, I'll cover a recent MDPI "paper" that tries (and fails) to explain the PETM as if that's the case.

Temperature Change and Impact

While proxy evidence suggests the PETM caused 5–8°C warming, a recent data assimilation (DA) analysis by Jessica Tierney suggests the amount of warming was more likely in the lower end of that range; her analysis "yields a rigorous estimate of global mean temperature change" of 5.6°C with a 95% range of  5.4°C to 5.9°C.[8] One of the benefits of her DA study is that the analysis provides a globally coherent spatial distribution of temperature changes, and her results show evidence of polar amplification; that is, warming Antarctica and the Arctic was significantly greater than in the tropics.

 

The biological impacts of the PETM were pretty enormous; the globe was essentially free of ice, and what we currently call tropical habitats could be found as far north as the what is now Wyoming. Ferns and redwood trees could be found inside the Arctic circle, as could crocodiles and alligators. Mammals diversified greatly during this time. The oceans, however, both warmed and acidified, and this lead to a mass extinction of benthic foraminifera; about 30-50% of species went extinct. Since there were no polar ice caps, so sea levels were much higher; what is now Florida would have been completely under water.

Sea Levels and the Positions of Continents During the PETM

Causes of Warming

The geological marker that identifies the beginning of the PETM is a very large negative carbon isotope excursion (nCIE), which can only be explained by a very large perturbation of the carbon cycle from a source of biogenic carbon added to the climate system. This can be seen in the graph below from Stokke et al 2020[5].

The nCIE at the base of the PETM is indicated by a sudden decrease in 13C/12C ratios, meaning that the carbon isotope ratio suddenly "lightened" - that is proportion of the heavier 13C isotope decreased in relation to the more typical 12C. This indicates a large influx 12C-rich biological carbon (either carbon dioxide or methane). Stokke et al 2020[5] (above) identified measured nCIE of ∼4.5‰ 𝛿13C in the Eastern North sea Basin, but other studies have discovered a range of CIE values between -3‰ and -7‰ 𝛿13C[10]. There's some difficulty constraining estimates for the size of the excursion because of the strong possibility diagenetic overprinting, where the original characteristics of rocks become modified by processes affecting those rocks through time after deposition, and this (among other things) makes it difficult to determine the exact size of the excursion. But the size of the nCIE is an indicator of the amount of biogenic carbon that must have been added to the climate system. It's essentially a mass balance calculation to determine the amount of 12C-rich carbon that would need to be released to create the nCIE. Wu et al 2025 concludes:

During the PETM event, significant amounts of 12C (∼3000 to 13,000 Pg) were released. The PETM is marked by pronounced negative excursions in δ13C, with some terrestrial basins exhibiting red-bed sediments. The primary sources of 12C release remain debated, with potential triggers including volcanic activity, seafloor methane release, and permafrost melting. Notably, negative carbon–oxygen isotope excursions in carbonate rocks serve as direct evidence for PETM events.[9]

To put these numbers into perspective, since humans have vented ~720 GtC (=PgC) to the atmosphere since 1750, and the PETM saw a minimum of 4x the increase caused by humans, though it was spread out over a much longer period of time. While there are several possible sources for this CO2 and/or CH4, the most likely explanations have to do with methane hydrates quite possibly associated with the North Atlantic Igneous Province (NAIP); these could have been released from warming oceans or permafrost. We won't be able to settle all the sources that were responsible for the PETM here, but the large nCIE indicates a vary large source of 12C-rich biogenic carbon was abruptly added to the carbon cycle, likely in several "pulses," and this large perturbation of the carbon cycle is what was responsible for the rapid warming. The oceans absorbed a large fraction of this extra carbon and acidified, resulting in the mass extinction of benthic foraminifera. And even if these pulses  were methane hydrates (I suspect this is most likely), methane quickly degrades to CO2, so either either way, we would see a large increase in CO2 concentrations.

Tierney's analysis[8] found that CO2 levels before the PETM were 1,120 ppm (with a 95% CI of 850 to 1,460 ppm), and at the peak of the PETM, CO2 nearly doubled to 2,020 ppm (with a 95% CI of 1,550 to 2,630 ppm). Her estimate of a 900 ppm increase in CO2 equates to an additional 2000 GtC in the atmosphere. If we take as a rule of thumb that about half of the PETM carbon emissions were absorbed by land and oceans, then her estimate would indicate a carbon source of about 4,000 GtC, on the lower end of the Wu et al 2025 range. Her analysis of the changes in GMST and the increase in CO2 lead to the conclusion that ECS may actually have been higher when the globe was warmer. She estimated ECS to be 6.5°C with a 95% range of  5.7°C to 7.4°C.  

The evidence is compelling that the PETM was a geologically sudden warming event driven by a large increase in GHGs to atmosphere. While carbon increases were geologically "sudden," the carbon emission rates on average were actually much slower than human emissions over the last century or so. This can be shown with some simple math. If the 900 ppm increase in CO2 occurred in just 10,000 years[12], this would indicate an average rate of increase on the order of ~0.1 ppm/year, compared to current rates exceeding 2 ppm/year (about 20x faster). And warming rates were also, on average, much slower than current warming rates. Using a high-end estimate of the amount of global warming, 8°C over 10,000 years, that's a warming rate on the order of  ~0.01°C/decade, compared to current warming rates of 0.2 C/decade (again, about 20x faster). Certainly within that 10,000 year span, there were pulses of increased carbon emissions and more rapid warming, and at least one analysis[20] suggests that warming pulses may have approached current warming rates on decadal time scales. But at bare minimum, the evidence suggests that the current human-caused perturbation of the carbon cycle and associated warming rates is on par with or exceeds the most rapid warming events in the Phanerozoic. 

Patrick Frank

A recent paper published by Patrick Frank[18] attempts to analyze the climate history of the Cenozoic (the last 66 million years) in terms of Henry's Law. Frank argues that the partial pressure of CO2 (pCO2) during the Cenozoic shows that concentrations of CO2 in the atmosphere is essentially only a product of ocean warming and cooling. Oceans warm and outgas CO2; they cool and absorb CO2. Consequently,

P(CO2) has been a molecular spectator of the Cenozoic climate. A generalizing conclusion is that the notion of atmospheric CO2 as the predominant driver of Cenozoic global surface temperature should be set aside.

In other words, the cumulative impact of 170 years of climate research should be discarded. The oceans may respond to warming events by outgassing CO2, but CO2 doesn't drive global warming events. Regarding the PETM, Frank suggests that the warming was triggered by the volcanic activity at NAIP. He proposed that "the elevated SST of the PETM was caused by thermal contact with NAIP magma." In other words, GHGs did nothing, but magma from the NAIP heated the entire planet, and the oceans outgassed CO2 as a result. There are several obvious fatal problems with this proposal:

1. The nCIE during the PETM points conclusively to a large perturbation of the carbon cycle; biological carbon must have been added to the atmosphere. Without a large 12C-rich source of CO2 or CH4, the nCIE could not occur. If the warming was driven by heat from the NAIP with outgassing of CO2 from the oceans occuring as a result, we would not see a negative CIE.[19]
2. If the PETM was caused by an event that resulted in the outgassing of CO2 from the oceans, the pH of the oceans would increase, not decrease. The ocean acidification that occurred with the warming during the PETM refutes the notion that the oceans were losing CO2.
3. If the PETM was caused by NAIP coming into thermal contact with ocean water, the globe would have warmed the most around the NAIP, near what is now Eastern Greenland. However, the above map shows the most warming likely occurred near Antarctica. The pattern of warming follows what would be expected from GHGs with polar amplification. 

Perhaps this explains why Frank's paper makes no mention of the nCIE that marks the onset of the PETM. If he had addressed the actual geologic evidence, it would prove his paper wrong. There is really nothing about Frank's paper that makes any sense of the available geologic evidence. The PETM could not have been caused magma warming the oceans, at least not in the manner described by Frank, with CO2 being a mere spectator of warming. Instead, the geologic evidence requires that biogenic carbon was the driver of the warming event, and the oceans absorbed (not outgassed) CO2 as a result.




References:

[1] Joost Frielinga et al, "Thermogenic methane release as a cause for the long duration of the PETM" PNAS 113 no 43 (October 25, 2016) 12059–12064
www.pnas.org/cgi/doi/10.1073/pnas.1603348113

[2] Alexander Gehler et al, "Temperature and atmospheric CO2 concentration estimates through the PETM using triple oxygen isotope analysis of mammalian bioapatite" PNAS 113 no 28 (July 12, 2016): 7739-7744.
https://doi.org/10.1073/pnas.1518116113

[3] Marcus Gutjahr et al, "Very large release of mostly volcanic carbon during the Paleocene-Eocene Thermal Maximum" Nature. 548/7669 (August 30, 2017): 573–577
https://escholarship.org/uc/item/1n988123

[4] McInerney, F. A. and S. Wing. “The Paleocene-Eocene Thermal Maximum: A Perturbation of Carbon Cycle, Climate, and Biosphere with Implications for the Future.” Annual Review of Earth and Planetary Sciences 39 (2011): 489-516.
https://www.researchgate.net/publication/234145841_The_Paleocene-Eocene_Thermal_Maximum_A_Perturbation_of_Carbon_Cycle_Climate_and_Biosphere_with_Implications_for_the_Future

[5] Stokke, E. W., Jones, M. T., Tierney, J. E., Svensen, H. H., & Whiteside, J. H. (2020). Temperature changes across the Paleocene-Eocene Thermal Maximum – a new high-resolution TEX86 temperature record from the Eastern North Sea Basin. Earth and Planetary Science Letters, 544. https://www.sciencedirect.com/science/article/pii/S0012821X20303320

[6] Zhu, J., Poulsen, C., & Tierney, J. (2019). Simulation of Eocene extreme warmth and high climate sensitivity through cloud feedbacks. Science Advances, 5(9). Retrieved from https://advances.sciencemag.org/content/5/9/eaax1874

[7] Keller, Gerta, et al. "Environmental changes during the Cretaceous-Paleogene mass extinction and Paleocene-Eocene thermal maximum: implications for the Anthropocene." Gondwana Research 56 (2018): 69-89.
https://www.sciencedirect.com/science/article/abs/pii/S1342937X17303702

[8] J.E. Tierney, J. Zhu, M. Li, A. Ridgwell, G.J. Hakim, C.J. Poulsen, R.D.M. Whiteford, J.W.B. Rae, L.R. Kump, (2022) Spatial patterns of climate change across the Paleocene–Eocene Thermal Maximum, Proc. Natl. Acad. Sci. U.S.A. 119 (42) e2205326119,
https://doi.org/10.1073/pnas.2205326119.

[9] Yuqi Wu, Tao Hu, Fujie Jiang, Jing Guo, Feilong Wang, Zhenguo Qi, Renda Huang, Zhou Fang, Xiaowei Zheng, Di Chen, Lacustrine records of Paleocene-Eocene Thermal Maximum (PETM) triggered by volcanic activity, Organic Geochemistry, Volume 200, 2025, 104899, ISSN 0146-6380,
https://doi.org/10.1016/j.orggeochem.2024.104899.
(https://www.sciencedirect.com/science/article/pii/S0146638024001645)

[10] Qinghai Zhang, Ines Wendler, Xiaoxia Xu, Helmut Willems, Lin Ding, Structure and magnitude of the carbon isotope excursion during the Paleocene-Eocene thermal maximum, Gondwana Research, Volume 46, 2017, Pages 114-123,
https://doi.org/10.1016/j.gr.2017.02.016.
(https://www.sciencedirect.com/science/article/pii/S1342937X17301417)

[11] Morgan F. Schaller et al. ,Impact ejecta at the Paleocene-Eocene boundary. Science 354, 225-229 (2016). DOI:10.1126/science.aaf5466

[12] Svensen, H., Planke, S., Malthe-Sørenssen, A. et al. Release of methane from a volcanic basin as a mechanism for initial Eocene global warming. Nature 429, 542–545 (2004). https://doi.org/10.1038/nature02566

[13] Philip A. E. Pogge von Strandmann et al., Lithium isotope evidence for enhanced weathering and erosion during the Paleocene-Eocene Thermal Maximum.Sci. Adv. 7 ,eabh4224 (2021). DOI:10.1126/sciadv.abh4224

[14] Secord R, Gingerich PD, Lohmann KC, Macleod KG. Continental warming preceding the Palaeocene-Eocene thermal maximum. Nature. 2010;467(7318):955-958. doi:10.1038/nature09441

[15] Londono et al, "Early Miocene CO2 estimates from a Neotropical fossil leaf assemblage exceed 400 ppm." Am J Bot. 2018 Nov;105(11):1929-1937.
https://www.ncbi.nlm.nih.gov/pubmed/30418663

[16] Panieri et all, "Methane seepages recorded in benthic foraminifera from Miocene seep carbonates, Northern Apennines (Italy)" Palaeogeography, Palaeoclimatology, Palaeoecology. Volume 284, Issues 3–4, 30 December 2009, Pages 271-282
https://www.sciencedirect.com/science/article/pii/S0031018209004246

[17] Zuoling Chen, Xu Wang, Jianfang Hu, Shiling Yang, Min Zhu, Xinxin Dong, Zihua Tang, Ping'an Peng, Zhongli Ding, Structure of the carbon isotope excursion in a high-resolution lacustrine Paleocene–Eocene Thermal Maximum record from central China, Earth and Planetary Science Letters,

Volume 408, 2014, Pages 331-340.

https://doi.org/10.1016/j.epsl.2014.10.027.

(https://www.sciencedirect.com/science/article/pii/S0012821X14006542)

[18] Frank, P. Cenozoic Carbon Dioxide: The 66 Ma Solution. Geosciences 2024, 14, 238. https://doi.org/10.3390/geosciences14090238

[19] Omta, A.W., Follett, C.L., Lauderdale, J.M. et al. Carbon isotope budget indicates biological disequilibrium dominated ocean carbon storage at the Last Glacial Maximum. Nat Commun 15, 8006 (2024). https://doi.org/10.1038/s41467-024-52360-z

[20] Kemp, D., Eichenseer, K. & Kiessling, W. Maximum rates of climate change are systematically underestimated in the geological record. Nat Commun 6, 8890 (2015). https://doi.org/10.1038/ncomms9890