Rocks from the Bighorn Basin in Wyoming were used to reconstruct Earth's climate at the time of the Paleocene-Eocene thermal maximum.

Rocks from the Bighorn Basin in Wyoming were used to reconstruct Earth's climate at the time of the Paleocene-Eocene thermal maximum.

Scott Wing, Smithsonian Institution

Greenhouse emissions similar to today’s may have triggered massive temperature rise in Earth’s past

About 55.5 million years ago, a burst of carbon dioxide raised Earth's temperature 5°C to 8°C, which had major impacts on numerous species of plants and wildlife. Scientists analyzing ancient soil samples now say a previous burst of the greenhouse gas preceded this event, known as the Paleocene-Eocene thermal maximum (PETM), and probably triggered it. Moreover, they believe humans are pumping similar levels of carbon dioxide into the atmosphere right now, raising concerns that our own emissions may also destabilize Earth's climate, triggering the planet to emit devastating bursts of carbon in the future.

The paper implies that even if we stopped emitting carbon dioxide right now, our descendants might still face huge temperature rises, says paleoclimatologist Gabriel Bowen of the University of Utah in Salt Lake City, the lead author of the new research. “It is a possibility,” he says, “and it's a scary one.”

Scientists accept that a massive injection of carbon into the atmosphere caused the PETM, but they don't agree about where the gas came from. Some researchers say it originated from the release of carbon locked up under the ocean by an undersea landslide; others blame a comet crashing into Earth, causing carbon from both the comet and Earth to be oxidized to carbon dioxide and potentially causing wildfires or burning of carbon-rich peat bogs on Earth. They also don't know how long the release lasted, with recent estimates ranging from 10 years to 20,000 years.

One of the best ways to measure the prehistoric release of carbon into the atmosphere is to look at the ratio of two types of carbon atoms called isotopes. Carbon has two stable isotopes: About 99% of natural carbon is carbon-12, whereas the remaining 1% is mainly the heavier carbon-13, with trace amounts of radioactive carbon-14 that decay within a few thousand years to nitrogen. Living organisms have a slight preference for the lighter isotope, so carbon derived from organic sources (such as fossil fuels) is slightly depleted in carbon-13. If that carbon gets returned to the atmosphere at a faster rate than normal, atmospheric carbon dioxide has less carbon-13 than normal. Plants taking up this carbon dioxide become even more carbon-13 depleted, and when they decompose, this depletion is recorded in the soil.

Sedimentary rock samples that have been compacted from soils formed at the time of the PETM contain less carbon-13 than normal. Sedimentary rocks of the Bighorn Basin in Wyoming contain one of the best records of soils from this period. Geologists have studied them for more than 100 years, but to obtain samples from soils of different periods, geologists had to analyze surface rocks from different parts of the basin and try to piece together a continuous geological history. Therefore, the Bighorn Basin Coring Project, run by the University of New Hampshire, Durham, drilled approximately 1 km of core from each of three different points in the basin to give geologists three clear, continuous records of how the soils had varied over time in a particular place.

Bowen and colleagues analyzed one of these cores, tracking the variations in carbon isotope ratios in greater detail than had been previously possible by examining surface rocks. They report online today in Nature Geoscience that in soils beneath those laid down during the main rise in temperature about 55.5 million years ago, there was a distinct drop in the proportion of carbon-13. In soils immediately on top of these, the ratio seemed to recover to its normal value. Finally, soils on top of these showed a large drop in the proportion of carbon-13 corresponding to the PETM itself.

So what was going on? The researchers concluded that there must have been two separate releases of carbon. The first, smaller release, about 2000 years before the main temperature rise, was followed by a recovery to normal climatic conditions. Later, a second, larger pulse caused the main event. “I'm fairly convinced that they're related,” Bowen says. “We see nothing remotely similar during the many hundreds of thousands of years before this event. To have within a few thousand years these two major carbon isotope shifts and have that be circumstantial would be quite remarkable.”

The researchers used climate models to investigate how the initial, smaller heating could have triggered the later surge in temperature. They estimate that the first thermal pulse is likely to have warmed Earth's atmosphere by 2°C to 3°C, but that the atmospheric temperature would have gradually returned to normal as the heat was absorbed into the deep ocean. However, when that heat finally reached the ocean floor, it might have melted methane ices called clathrates, releasing the methane into the ocean and allowing it to make its way into the atmosphere. As a greenhouse gas, methane is 21 times more potent than carbon dioxide, so a sudden spike in methane emissions could lead to huge climate change.

“The connection between these two pulses is something that it's going to be really important to get a handle on,” Bowen says. The researchers believe the rate at which carbon dioxide escaped into the atmosphere during both bursts is unlikely to have been greater than the rate at which humans are emitting it now, and it may have been considerably lower. “Carbon release back then looked a lot like human fossil fuel emissions today,” Bowen says. “So we might learn a lot about the future from changes in climate, plants, and animal communities 55.5 million years ago.”

“We think this is really good news for our contention that the release of carbon was very fast,” says marine geologist James Wright of Rutgers University, New Brunswick, an advocate of the comet impact hypothesis.

Wright is not convinced, however, about the importance of the first pulse in triggering the second. He suggests that the most logical interpretation of the apparent cooling after the first pulse is that its significance was less than Bowen's group believes, with limited effect on the overall ocean temperature, and that not just the atmosphere but rather the entire planet quickly returned to normal. “If that's the case, then the first has nothing to do with the second,” Wright says. That, in turn, would require an alternative explanation for the PETM such as a comet impact.