The source of the greenhouse gas is still the subject of active debate. One thing we know for sure, the PETM CO2 could not have come from humans or those early primates! There are five potential candidate causes: (1) gas hydrates in marine sedimentary rocks, (2) coals in terrestrial sedimentary rocks, (3) extensive wildfires, (4) melting of permafrost, and (5) volcanism in the North Atlantic Ocean coincident with the opening of the sea between Norway and Greenland. Each of these sources could have liberated sufficient CO2 or methane (CH4) to cause the warming and the carbon isotope excursion. This is explained in the video below.
Geologists have several ways to decide between these different mechanisms. There are independent means of determining the sources of carbon. The best way is the carbon isotope ratio of the different sources, which is very different. For example, methane hydrate has a ratio of – 60; coal and permafrost are about -20, while volcanism has a ratio of about -6. This means to cause an isotope excursion of -4 to -5, there needs to be almost 10 times more volcanic carbon than methane hydrate carbon.
Another way is the volume of carbonate dissolved by acidification in the deep sea. This can be estimated by looking at changes in the amount of CaCO3 in different sites. The third is from the magnitude of the change in pH, determined from the boron isotope proxy. Since the results from one proxy are not unique, geologists must use two proxies to constrain the source of CO2. These estimates give quite different results, unfortunately. Estimates from carbon isotopes and carbonate point to a source such as permafrost or coal, or a mix of volcanism and methane while those from boron and carbon isotopes and also point to a mixture from volcanism and one of the other sources. Thus, it is likely that there was a mixture of sources of greenhouse gas that fueled the PETM.
One of the key aspects of this debate is the evidence for some of these mechanisms actually exist. We know that there was volcanic activity in the North Atlantic, and dates from these lavas are almost precisely the same as the PETM. We know that this volcanic magma was injected through coal. There are also signs of disturbance on the sea floor off the coast of Florida that could have been caused by the release of methane hydrates. The evidence is thus stronger for volcanism, coal, and methane hydrate than it is for permafrost and wildfire, but the problem is far from solved.
One other key piece of evidence is the rate of CO2 addition. Methane hydrate, coal, and permafrost tend to be released in a more catastrophic fashion, whereas volcanism tends to be a little slower over time. If CO2 is released quickly, then a large part of it is absorbed in the surface ocean, leading to surface ocean acidification. On the other hand, CO2 added slowly generally bypasses the surface ocean and acidifies the deep. The lack of evidence for significant surface ocean acidification is more consistent with slow emission and volcanism.
One of the most important parts of the PETM is that it allows us to learn how the Earth operates in a warm mode with higher CO2 levels than today. As we learned, the current CO2 concentration in the atmosphere is 400 parts per million (ppm) and the PETM levels were likely double or triple this concentration (800-1200 ppm). Warm conditions during the height of the PETM led to increased break down of minerals, a process called weathering, and this removed CO2 from the atmosphere. It could be that increased weathering in the warm PETM atmosphere was the beginning of the end of the event. In addition, the process whereby the ocean absorbed CO2 leading to dissolution of CaCO3 would also have led to the termination of the event.