CO2 is simultaneously cooling the stratosphere and warming the surface
For more than half a century, a quiet paradox has lived inside climate science: the same molecule warming Earth's surface has been steadily cooling the upper atmosphere. Now, a team at Columbia University's Lamont-Doherty Earth Observatory has done what decades of research could not — they have written down the precise mathematics of why. In resolving this puzzle, they have also revealed how the two effects are not opposites but partners, each amplifying the other in ways that matter deeply for how we understand our planet's future.
- The stratosphere has been cooling at a rate more than ten times steeper than natural models predict, a decades-long anomaly that climate science acknowledged but could never fully explain at the level of mechanism.
- The tension lies in CO2's split personality: in the lower atmosphere it acts as a heat-trapping blanket, but in the stratosphere it becomes a radiator — and no one had written down exactly why until now.
- Columbia researchers identified a 'Goldilocks zone' of infrared wavelengths that become increasingly efficient at driving stratospheric cooling as CO2 rises, confirming a prediction made in the 1960s that had never been mathematically grounded.
- The resolution carries a sting: a colder stratosphere emits less total energy to space, meaning more heat stays locked in the Earth system overall — stratospheric cooling and surface warming are not competing forces but reinforcing ones.
- The mathematical framework now stands as a foundation for sharper climate models and, unexpectedly, as a potential tool for decoding the atmospheres of other planets, from our solar system's neighbors to distant exoworlds.
There is a puzzle that has sat at the heart of climate science for more than fifty years. Earth's surface grows warmer, the lower atmosphere traps more heat — and yet the stratosphere, the band of air stretching from eleven to fifty kilometers above us, has been cooling sharply. Since the mid-1980s it has dropped roughly two degrees Celsius, a decline far steeper than any model without human-caused CO2 would predict. Scientists knew it was happening. They even knew, in broad strokes, why. But the precise mechanism — the kind you can write as equations — had never been worked out.
A team at Columbia University's Lamont-Doherty Earth Observatory has now done that work. Led by postdoctoral researcher Sean Cohen, alongside Robert Pincus and geophysicist Lorenzo Polvani, the study identifies the specific physics behind the paradox. The key is that CO2 does not behave the same way at all altitudes. In the lower atmosphere it acts as a blanket, trapping infrared energy. In the stratosphere it becomes a radiator, absorbing heat from below and sending it into space. The difference comes down to specific infrared wavelengths — what the researchers call a Goldilocks zone — that are particularly efficient at driving cooling. As CO2 concentrations rise, that zone expands. This was predicted by Nobel laureate Syukuro Manabe in the 1960s, but the underlying mathematics had never been fully derived.
Cohen's team built their framework through careful, iterative work — identifying key processes, assigning mathematical values, testing against climate simulations and real-world observations, then refining. They also checked whether ozone and water vapor might be significant drivers of the cooling and found their influence minor compared to CO2. What emerged fits cleanly with decades of data: stratospheric cooling grows more pronounced at higher altitudes, and each doubling of CO2 produces substantial cooling near the stratosphere's top.
There is a twist that makes the surface warming worse. A cooler stratosphere, despite being a more efficient radiator, ends up sending less total energy to space because it is colder. The net effect is that more heat remains trapped in the Earth system overall. CO2 is simultaneously cooling the stratosphere and intensifying surface warming — and the two effects feed each other.
Cohen is clear that this is not another proof of climate change; that case has long been settled. What the study offers is something more foundational: a precise mechanistic understanding of a process that has been part of climate science for half a century without ever being fully explained. Better models, more precise predictions, and a sharper picture of how the atmosphere actually works are the practical rewards. There is also an unexpected reach: the same physics applies, in principle, to the atmospheres of other planets — and potentially to exoplanets orbiting distant stars. The study appears in Nature Geoscience.
There is a puzzle at the heart of climate science that has bothered researchers for more than fifty years. While the surface of Earth grows warmer and the lower atmosphere traps more heat, something strange happens higher up: the stratosphere—the layer of air stretching from about eleven to fifty kilometers above our heads—has been cooling. A lot. Since the mid-1980s, it has dropped roughly two degrees Celsius, a decline more than ten times steeper than models without human-caused carbon dioxide would predict. Scientists knew this was happening. They even knew, in broad strokes, why. But they could not explain the mechanism. Not really. Not with the precision that lets you write down equations and say: here is exactly what drives this.
A team at Columbia University's Lamont-Doherty Earth Observatory has now done that work. Led by postdoctoral researcher Sean Cohen, with contributions from Robert Pincus and geophysicist Lorenzo Polvani, they have identified the specific physics that makes the stratosphere cool even as the planet below warms. The answer lies in understanding that the atmosphere is not one thing. Carbon dioxide behaves like a blanket in the lower atmosphere, trapping infrared energy and warming the surface. But in the stratosphere, it becomes something else entirely—a radiator, absorbing heat from below and sending it out into space.
The difference comes down to how carbon dioxide interacts with infrared light at different altitudes. Not all wavelengths of infrared radiation behave the same way when they encounter CO2 molecules. Some are far more efficient at driving heat away. The researchers identified what they call a Goldilocks zone of wavelengths that are particularly effective at cooling the stratosphere. As CO2 concentrations rise, that zone expands, making the radiating process more efficient. This was actually predicted back in the 1960s by climatologist Syukuro Manabe, whose models later earned him a Nobel Prize. But the underlying mathematics—the precise mechanism—had never been worked out.
Cohen and his colleagues built their understanding through careful, iterative work. They identified the key processes, assigned mathematical values to them, then tested their equations against both detailed climate simulations and real-world observations. They adjusted, refined, and repeated. They also checked whether other factors—ozone and water vapor, both involved in similar processes—might be driving the cooling. They found that compared to CO2, their influence was minor. What emerged was a clean mathematical framework that fit neatly with decades of observations: stratospheric cooling becomes more pronounced at higher altitudes, and each doubling of CO2 causes substantial cooling near the top of the stratosphere.
But there is a twist that makes the surface warming worse. A cooler stratosphere radiates less total energy out to space overall. Yes, CO2 makes the stratosphere better at radiating heat outward, which cools it. But because it becomes colder, it ends up sending less total energy into the void than it otherwise would. The net result is that more heat stays trapped in the Earth system as a whole, reinforcing the warming happening below. Carbon dioxide is simultaneously cooling the stratosphere and making the surface warmer, and the two effects are connected—one feeds the other.
Cohen is careful about what this study does and does not do. It is not another piece of evidence for climate change. That case has long been settled. What it offers instead is something more fundamental: a clearer mechanistic understanding of a process that has been part of climate science for half a century without ever being fully explained. Understanding which factors actually drive stratospheric cooling, and being able to express that mathematically, gives future researchers a more solid foundation. Better models. More precise predictions. A sharper picture of how the atmosphere actually works.
There is also an unexpected reach beyond Earth. The same physics that governs CO2 behavior in our stratosphere applies, in principle, to the atmospheres of other planets. A cleaner mathematical theory for stratospheric cooling could help scientists make sense of conditions on other worlds in the solar system and potentially on exoplanets orbiting distant stars. It is a long way from a quirk in Earth's temperature record to understanding alien atmospheres. But that is sometimes how basic science works. You set out to explain something that has puzzled people for decades, and you end up with a tool that reaches further than you expected. The study appears in Nature Geoscience.
Citas Notables
The existing theory was incredibly insightful, but we lacked a quantitative theory for CO2-induced stratospheric cooling— Sean Cohen, Columbia University
This is really telling us what is essential—understanding which factors actually drive stratospheric cooling gives future researchers a more solid foundation— Robert Pincus, Columbia University
La Conversación del Hearth Otra perspectiva de la historia
So CO2 is warming the surface but cooling the upper atmosphere. That seems backwards.
It's not backwards—it's two different jobs in two different places. In the lower atmosphere, CO2 traps heat like a blanket. Higher up, in the stratosphere, it acts like a radiator, pushing heat away into space.
But if it's pushing heat away, shouldn't that cool the whole planet?
That's the paradox. The stratosphere does cool. But because it's colder, it radiates less total energy outward. So more heat ends up staying in the Earth system overall. The cooling up there actually makes surface warming worse.
How did they finally figure out what was actually happening?
They identified that certain wavelengths of infrared light are much more efficient at driving the cooling. As CO2 increases, those wavelengths become even more effective. They built equations around that, tested them against real data, and it all fit.
This was predicted in the 1960s though, right?
Yes, but only qualitatively. Scientists knew it happened. They didn't know the precise mechanism—the actual physics that made it work. That's what took fifty years to nail down.
Does this change anything about climate predictions?
Not the big picture. But it gives researchers a solid mathematical foundation for better models and more precise predictions going forward. It's also useful for understanding other planets' atmospheres.