Recent climate projections may be underestimating the pace of global warming in an atmosphere damaged by greenhouse gas emissions, because the interaction of powerful climate feedback loops that can accelerate warming are not well-represented in key climate models, an international team of scientists concluded in a study published today in the journal One Earth. Their findings suggest that efforts to reduce emissions require even more urgency to avoid worst-case climate outcomes, the team reported.

“If amplifying feedbacks are strong enough, the result is likely tragic climate change moving beyond anything humans can control,” said co-author Bill Ripple, an ecologist at Oregon State University, and co-founder of the Alliance of World Scientists, which has 26,000 members in 180 countries urging decisive implementation of policies to curb global warming and meet the commitments governments made under the 2015 Paris Agreement. 

“We would like to see an IPCC special report that focuses on the many risky climate feedbacks and the possible acceleration towards planetary tipping points,” he said. “It’s important to understand the most optimistic estimates, but we also need to be informed of potential worst-case scenarios.”

Recent evaluations conclude that, if countries meet the emissions-reduction targets they’ve set for themselves, the average global temperature would warm 2.7 degrees Celsius from pre-industrial temperatures by 2100, which would have catastrophic impacts for people and ecosystems. But if some of the feedback loops shown in the new paper accelerate, warming could soar well above that level, toward 4 degrees Celsius, by the end of the century.

Climate feedbacks are physical, chemical and biological processes in the ocean, atmosphere and on land that, in most cases, reinforce each other to speed up warming. One example is melting sea ice that allows the ocean to absorb more sunlight, which warms the water, to melt more ice. The new paper also describes several climate feedback loops that could have a cooling effect, like when trees fertilized by increasing concentrations of carbon dioxide grow faster, thus taking up more CO2. 

The researchers examined 41 climate feedback loops and found 27 that significantly increase warming but may not be fully accounted for in climate models. Ripple said scientists generally understand the feedback loops individually, but that the models often overlook the cumulative effect all of them together might have over the next 50 to 80 years.

“We are particularly concerned about several biological feedback loops, including permafrost thawing, forest destruction, loss of soil carbon and smoldering peatlands,” Ripple said. “These feedbacks may contribute significantly to warming over the course of the century.”

Feedback Loops Increase Polar Vulnerability

The Arctic, warming now at four times the global average rate, shows how feedback loops can interact. Scientists know thawing permafrost releases greenhouse gases. A 2017 study showed the potential for carbon releases from the disintegration of an Alabama-sized area of permafrost. Researchers, however, don’t currently expect that process by itself to cause runaway warming in the next few decades.

But Arctic sea ice is dwindling too, exposing more dark ocean water to absorb more heat, which leads to yet more ice melting. And the changes to sea ice extent and ocean surface temperatures affect the atmosphere above the sea and the permafrost.

What researchers don’t yet fully understand or show in climate models, Ripple said, is how all those different processes can amplify each other, and whether their interactions will lead to sudden and irreversible changes in the next few decades. 

Other climate feedbacks in the ocean could also accelerate global warming, said co-author Johan Rockström, director of the Potsdam Institute for Climate Impact Research. The models used to calculate widely accepted global temperature increases, including those by the Intergovernmental Panel on Climate Change, “give us the pathways towards holding the 1.5 degrees Celsius limit,” Rockström said. The models assume “that the ocean will continue to operate more or less in the same mode as today, in terms of heat uptake and in terms of the solubility of carbon dioxide.” 

But the amplifying cycle of feedbacks could change that sooner than expected.

“We say, in this One Earth article, that we need to look very carefully into how the Atlantic overturning heat circulation, connected to the Gulf Stream, will behave,” he said. “How will the Labrador Sea Oscillation operate? How will the El Niño, La Niña oscillation develop? We know the ocean has taken up 95 percent of the heat caused by anthropogenic global warming. The question is, what dynamics could unexpectedly occur as a result of that rising stress?”

Deep Uncertainty Regarding Potentially Dangerous Tipping Points

The most recent science reports from the Intergovernmental Panel on Climate Change identify a handful of the most important feedbacks that could push the climate past tipping points, but since they still can’t be adequately assessed, the international science panel can’t determine their probability of happening, or accurately project how they will interact.

In its landmark climate science reports, IPCC estimates the possibility of specific climate impacts occurring—for instance, it assigns high likelihood to sea level rise—but the feedbacks and tipping points are still in the realm of “deep uncertainty,” said Frank Pattyn, an Antarctic glaciologist at Vrije Universiteit Brussels who was not involved with the paper.  

“In my domain, one known deep uncertainty is Antarctic ice sheet loss, which is related to feedbacks that we know from marine ice sheet instability and marine ice cliff instability,” he said. Research suggests that, in past times of global warming, big chunks of Antarctica’s ice have crumbled fast enough to melt ice sheets and raise sea levels faster than the IPCC projections.

The first papers identifying marine ice cliff disintegration as a potential contributor to rapid sea level rise appeared in about 2016, he said. “And now, after six, seven years, we have not profoundly advanced to where we can take that into account in a more physically understandable way, that would give more precise projections of sea level rise.”

Research on the identified tipping points is still “in its infancy,” he said, and the new paper points to the need for a “massive international mobilization … to quickly assess the impacts and interactions of feedbacks.”

“If you sum up the funding associated with sea level rise research, it’s peanuts compared to what is at stake with sea level rise,” he said. A 2022 study in EOS clearly showed this critical funding gap.

“There’s a deeper picture here,” said co-author Tim Lenton, director of the Global Systems Institute at the University of Exeter. Lenton was deeply influenced by scientist James Lovelock’s Gaia hypothesis, which proposes that life interacts with its inorganic surroundings on the planet to form a self-regulating system that perpetuates the conditions required for it to persist. “That’s part of what some climate models may be missing,” he said.

“Lovelock was a visionary in thinking about how humans and our actions are intertwined in that system, and of course, that’s also what the paper is about,” he said, “how to see ourselves as integral parts and actors in the system.”

There has been an effort to bring a more faithful representation of the nuanced influences of life on the Earth into climate modeling, he said, but they aren’t yet fully integrated. Even if global climate models can’t be rebuilt from the ground up, they should “bring those qualities of life into some model frameworks, at least just to explore how that would affect the feedback dynamics,” he said.

“You want to bring together evolution, ecology, biogeochemistry and the physical climate and you don’t want to go crazy,” Lenton said.  

Climate models began as weather forecast models focused on the atmosphere and short-term atmospheric processes, he said. 

“Gradually, they’ve had more and more stuff bolted on, and when we spot something that might be missing, that might be an important feedback, we kind of go to work bolting that into the model as well,” Lenton said. “You’re always playing catch up.” 

Lenton said the new paper shows how important it is to keep striving to limit global warming, even if the 1.5 degree Celsius limit is breached in coming years.

“I think many of us would say the evidence shows that the damage from global warming is going up in a nonlinear way, so every tenth of a degree more warming will cause more damage than the previous tenth,” he said.