Even if humanity manages to cut emissions, Earth may still slow our return to a cooler world. New research shows that as northern peatlands warm, they release methane that outweighs much of their CO₂ uptake. The effect is small in the short term but stretches the global heat peak and shrinks the remaining carbon budget – meaning that returning below 1.5°C will require far greater carbon removal than expected.
Imagine the world finally bending the emissions curve – only to find that nature itself resists the cooldown. Across Canada, Scandinavia and Siberia, vast northern peatlands hold carbon built up over millennia. But as these water-logged soils warm, they begin to exhale methane – a greenhouse gas far stronger than CO₂.
“It’s one of the planet’s quietest systems, yet it speaks loudly in the language of climate,” says Biqing Zhu from the International Institute for Applied Systems Analysis (IIASA) in Austria.
In a new international study co-led by Biqing Zhu and Chunjing Qiu from Laboratoire des Sciences du Climat et de l’Environnement, France, researchers show that for every extra degree of global warming, the remaining carbon budget – the CO₂ we can still emit before crossing 1.5°C – shrinks by about 37 gigatons. If temperatures overshoot, getting back below the line could require roughly 40 gigatons of extra carbon removal.
“Our goal was to understand how northern peatlands behave if the world temporarily exceeds the 1.5°C limit – what climate scientists call an overshoot,” Chunjing Qiu explains. “Even though peatlands continue to take up CO₂, the cooling effect is largely offset by rising methane emissions. The feedback may look small, but it stretches the peak and makes the return much slower.”
In practice, that means today’s climate plans may be too optimistic. “Basically, yes – current goals were derived from models that didn’t include these processes,” Zhu says. “For the amount of reduction and removal that’s currently planned, the effects of peatlands were not considered.”
“Overshoot needs serious attention,” Qiu warns, “because nature doesn’t respond linearly – and some changes are hard to reverse.”
A quiet ecosystem with loud consequences
Peatlands are among Earth’s quietest ecosystems – mossy, remote, and easily forgotten. Yet beneath their still, wet surfaces lie ancient stores of carbon that dwarf all the forests of the world combined. For decades, these northern wetlands were assumed to be stable, passive participants in the climate system. But as temperatures rise, that assumption no longer holds.
“It’s a system that’s often overlooked,” says Biqing Zhu. “For a long time, peatlands and permafrost were presumed fairly stable, and because most people rarely encounter these systems, we know least about the places that matter most for carbon, but newer findings show that stability can shift quickly with climate warming.”
Scientists now know that peatlands both absorb and emit greenhouse gases in complex ways. In cool conditions they act as a sponge, trapping CO₂ from the air; in warmer, wetter ones, they can become a potent source of methane.
“Even the systems that help us cool the planet can start to work against us,” Zhu notes.
The missing data that kept peatlands out of climate models
Until recently, the challenge was data – or rather the lack of it. Because most peatlands lie far from people, field measurements are scarce – and early climate models, built mainly for oceans and forests, left them out entirely. “The field is young,” Zhu says. “Our understanding has lagged simply because these cold, remote systems are so hard to measure.”
“Peatlands, to me, are part of that cold world that quietly shapes the Earth’s future – so for me it was both scientifically important and personally interesting to study them. When a system has been overlooked but you sense it’s important, it naturally sparks curiosity – you want to know a bit more about it.”
Now, as the climate warms faster in the Arctic than anywhere else, these quiet landscapes have become impossible to ignore. “Small on the map,” Zhu says, “but big in the carbon cycle.”
Simulating the slow heartbeat of the Earth
How do you measure the behaviour of something that covers millions of square kilometres and responds over centuries?
Zhu, Qiu and their colleagues turned to a computational shortcut that balances realism with flexibility. Instead of running full Earth-system simulations – which can take months – they built a simplified computer model – known as an ‘emulator’ – that imitates how the full Earth-system models behave.
“We used an emulator, a reduced-complexity model built on complex process-based models,” Zhu explains. “Complex models describe the Earth in small grid cells; we aggregate their behaviour so we can run hundreds of scenarios and quantify uncertainty across processes, climate, and pathways.”
The emulator – nicknamed OSCAR-PEAT – draws on five of the world’s leading land models, each describing peatland hydrology, vegetation, and soil chemistry a little differently.
“I think of it as an evidence chain,” Zhu says. “Field scientists measure, complex modelers encode those processes, and we synthesize their results to explore the full range of plausible futures.”
Rolling the dice on thousands of possible Earths
When data are scarce, different teams reach different conclusions. Rather than choose one model, Zhu’s team combined them – Monte Carlo style – to explore thousands of possible Earths.
“We used what scientists call a Monte Carlo approach – running thousands of simulations, like rolling dice, to see every possible outcome,” she says. “Each roll of the dice reveals how much uncertainty remains – and where the real policy risks hide.”
The real strength of the approach is its scale. Full Earth-system models can only test a handful of scenarios, whereas the emulator allowed the team to explore hundreds of emissions pathways – including those that temporarily overshoot global temperature targets – and estimate how peatland carbon and methane fluxes would behave in each.
“The point is not to predict one future,” Zhu says, “but to show the range of what’s possible – because the risks are hidden in that range.”
“We call it an emulator,” she adds, “but really it’s a bridge – connecting field data, complex models, and statistical synthesis into one map of uncertainty.” That map reveals not only what we know, but also where the blind spots remain.
Sink on paper, warming in practice
When the researchers let their emulator run, the pattern that emerged was both reassuring and unsettling. Peatlands are still doing their ancient job – pulling CO₂ from the air and locking it away underground – but at the same time they are emitting more methane than before. The result is a small but persistent source of extra warming, especially as global temperatures rise.
“Peatlands still accumulate carbon, just as they have for thousands of years,” says Zhu. “But warming – even without drying – boosts methane emissions enough to turn them into a small but growing warming source. It’s already happening, though still modest for now.”
The implications are clear: even if global society manages to cut emissions on schedule, methane from peatlands could offset part of the cooling benefit. “Because many models didn’t include peatlands,” Zhu notes, “meeting the same temperature targets will require more reduction and removal than previously planned.”
In their ensemble – a group of many model simulations – each additional degree of warming shortened the remaining carbon budget and stretched the time needed to bring temperatures back down. The trend held across hundreds of overshoot scenarios. “When we talk about positive or negative climate effects,” Zhu explains, “we usually refer to the warming impact. The net carbon balance can stay negative, but temperature feedback can still be positive.”
Or as Zhu puts it more simply: “Sink on paper, warming in practice – that’s the methane effect. It reminds us that nature can be both an ally and an obstacle in the same breath.”
Nature’s fine print in our climate plans
The message from Zhu’s study is not alarmist but precise: even the most optimistic climate plans must include nature’s feedbacks if the math is to hold. Peatlands, once thought too small to matter, could subtly stretch the timeline for cooling the planet back down.
“When we plan for overshoot, these natural side-effects must be counted too – otherwise we’re promising a future that physics won’t keep,” says Zhu. “If we ignore peatlands, we make an already difficult task slower and more expensive.”
That doesn’t mean the situation is hopeless. “Restoring peatlands isn’t just about saving ecosystems,” Zhu says.
“It’s about resetting the planet’s carbon rhythm. Early experiments show that rewetting drained bogs and re-establishing vegetation can sharply reduce CO₂ emissions, though the effects on methane are more complex. The data are still scarce, but they point to restoration as one of the few tools we have to regain balance.”
When patient science meets nature’s quiet power
Those experiments, scattered across northern Europe and Canada, are among the few real-world tests of how quickly a damaged peatland can recover its carbon-trapping power. “Most field experiments happen in places like Finland, Denmark, Russia, and Canada,” Zhu explains, “where there are cold peatlands, long winters, and patient scientists.”
The study itself, Zhu emphasizes, was a collective effort – an example of how synthesis can advance climate science. “This was truly a collaborative study,” she says. “Every piece, from field data to model design, depended on someone else’s expertise. It wouldn’t have been possible without the whole team, including my co-first author who is an expert in complex models.”
And amid the technical detail, Zhu’s takeaway remains clear. “Protecting peatlands buys time,” she says. “Add nature to the ledger – or the ledger lies. That’s the simplest way I can put it.”
“It’s not a topic that’s usually heard in the media,” she adds, “but these quiet systems may decide how fast we can cool the planet.”
For decades, climate policy treated peatlands as background scenery. Zhu and Qiu’s findings show they belong in the main equation. “Every fraction of a degree matters,” Zhu says. “And every hidden feedback counts.”
