If we want to know what to expect from our climate as it continues to warm over the next few centuries, looking at similar examples of climate change in Earth’s past would be helpful. But there certainly haven’t been any similar temperature excursions in the instrumental record. Using indirect measures, we can tell that there probably haven’t been any since the last ice age. Even the exit from that ice age isn’t especially relevant; while the planet warmed considerably, it was driven by a complicated mixture of orbital changes, greenhouse gases, and melting ice.
To find a sudden warming that’s driven entirely by greenhouse gases, you have to go back 56 million years to the Paleocene-Eocene Thermal Maximum (PETM). At the start of the PETM, a geologically sudden surge of carbon dioxide into the atmosphere caused warming and a large change in the ocean’s pH. It took well over 100,000 years for conditions to return to anything normal. During that time, the extinction rate rose, and many ecosystems were disrupted or shifted by thousands of miles.
But understanding the PETM has proven a challenge, as it’s not clear how much carbon entered the atmosphere or where it came from. A new paper in today’s issue of Nature takes existing information about carbon dioxide levels and isotope ratios and combines them with data on the amount of carbon that dissolved into the oceans. The results provide a new indication of how much carbon entered the atmosphere—10,000 gigatonnes—and suggests volcanoes put it there.
Somewhat disturbingly, however, the data reinforces past indications that the PETM was relatively slow compared to our current carbon binge.
The Earth as a whole was warming during the Paleocene. It was also undergoing a series of “hyperthermals” during which orbital changes drove short periods that were somewhat warmer than the baseline. Then, within a geological blink of an eye, there was a massive influx of carbon into the atmosphere, enough to throw off the ratio of carbon isotopes that had prevailed. Temperatures soared by at least 5 degrees Celsius, and the acidification of the oceans set off an extinction event. While the warming took less than 20,000 years to occur, the planet stayed unusually warm for about 200,000 years.
A variety of sources have been suggested for that carbon. These include everything from a comet that broke up in the atmosphere to the release of methane trapped on the ocean floor. The different sources all have somewhat different carbon isotope ratios, and so each would need to provide different amounts of carbon dioxide in order to drive the changes seen in the atmosphere. There’s uncertainty both in our measure of carbon isotopes during the PETM, in our estimates of the temperature change it created, and in the amount of carbon we think would be needed to drive that much temperature change. Combine those uncertainties with the fact that more than one source might have been responsible, and you get decades of scientists arguing over the details.
An international team of scientists decided to tackle the problem using a series of constraints. One constraint is temperature—you need enough carbon to raise the temperature by at least 5 degrees Celsius. Another is carbon isotopes. There needs to be enough carbon released to change those isotope ratios, although the precise value will depend on the source.
To those, the scientists added a third constraint: oceanic pH. Carbon dioxide dissolves in the oceans, lowering their pH in the process. The amount of pH change tells us something about how much carbon dioxide ended up in the atmosphere.
To get pH values, the authors turned to boron isotopes, which act as a proxy, tracking the pH of the oceans. They obtained these from the shells of organisms in a sediment core taken from a site in the North Atlantic. (The measurements seen at this single site are similar to those from elsewhere around the globe.) Their measurements show that pH dropped by about 0.27 during the PETM, which is a rather large change given the volume of the ocean.
To turn all these measurements into actual values, the authors turned to a coupled ocean-climate model. They ran simulations of the climate at weekly intervals across the entire 200,000-year period of the PETM. At each step, the team determined whether the amount of carbon in the atmosphere could cause the requisite changes in pH, temperature, and carbon isotope ratios. They would then adjust the carbon present according to these results and step the simulation forward.