Climate sensitivity is the amount Earth’s average temperature will rise if the concentration of carbon dioxide in the atmosphere doubles. The best current estimate is 3°C (about 5.4°F), with a likely range between 2.5°C and 4°C. That single number carries enormous weight: it determines how quickly the planet warms, how much carbon humanity can still emit, and how aggressive emissions cuts need to be.
The Core Concept
At its simplest, climate sensitivity answers a specific question: if CO2 levels went from their pre-industrial concentration (around 280 parts per million) to double that (560 ppm), how much warmer would the planet get once everything settled into a new balance? Scientists call this “equilibrium climate sensitivity,” or ECS. It captures every amplifying and dampening process in the climate system, from shifting cloud patterns to melting ice, rolled into one number.
The reason ECS uses a doubling of CO2 as its benchmark isn’t because that’s a magic threshold. It’s a standardized thought experiment that lets scientists compare results across models, time periods, and methods. Whether you study ancient ice cores or run supercomputer simulations, you can express your findings in the same units.
Equilibrium vs. Transient Response
Equilibrium climate sensitivity describes the warming you’d see after the climate has fully adjusted to doubled CO2. That full adjustment takes a long time. The deep ocean absorbs heat slowly, so reaching true equilibrium can take several thousand years. In practical terms, ECS tells you where the planet is headed, not where it is right now.
For nearer-term planning, scientists use a different measure called the transient climate response, or TCR. This is the warming that occurs at the exact moment CO2 concentrations hit double their starting point, assuming CO2 has been rising at a steady rate of 1% per year. TCR is always lower than ECS because the ocean hasn’t finished absorbing heat yet. Think of ECS as the final temperature the oven reaches and TCR as the temperature when the timer goes off partway through preheating. Both numbers matter, but for different reasons: TCR is more relevant to what we’ll experience in the coming decades, while ECS tells us the full commitment we’re locking in.
What Drives the Number Up or Down
If CO2 were the only thing that changed, and nothing else in the climate responded, a doubling would warm the planet by roughly 1.1°C. The reason sensitivity is closer to 3°C is feedback loops: chain reactions where the initial warming triggers other changes that amplify or dampen it.
The most powerful amplifying feedback involves water vapor. A warmer atmosphere holds more moisture, and water vapor is itself a greenhouse gas, so it traps additional heat. This alone roughly doubles the warming you’d get from CO2 on its own. Ice and snow create another amplifying loop: as warming melts bright, reflective ice surfaces, the darker ocean or land underneath absorbs more sunlight, which drives further warming. This is especially pronounced in the Arctic.
Clouds are the wild card. Low-level clouds tend to reflect sunlight and cool the planet, while high-altitude clouds can trap heat. How cloud cover changes as the planet warms is the single largest source of uncertainty in climate sensitivity estimates. Climate models struggle to accurately represent low-level clouds, and small differences in how a model handles them can shift its sensitivity estimate by more than a degree. NOAA-funded research has shown that even advanced models capture some aspects of real cloud behavior but miss important details about cloud types near Earth’s surface, exactly the clouds that have the biggest impact on warming projections.
How Scientists Pin Down the Range
No single method can nail climate sensitivity precisely, so scientists triangulate from three independent lines of evidence.
The first is climate models: complex simulations of the atmosphere, ocean, ice, and land surface run on supercomputers. These models apply the laws of physics to project how the climate responds to added CO2. The latest generation of models (called CMIP6) produced ECS values ranging from 1.83°C to 5.67°C, a wider spread than the previous generation’s range of 2.1°C to 4.7°C. Several of the newest models ran hotter than expected, largely because of stronger cloud feedback and ice-reflectivity feedback in their code. However, those high-end values don’t match what paleoclimate records and real-world temperature measurements suggest, so the scientific community treats them cautiously.
The second line of evidence comes from the historical temperature record. Scientists compare the observed warming since the industrial revolution to the known increase in CO2 and other greenhouse gases, then work backward to infer sensitivity. This approach is grounded in actual measurements but has its own complications, particularly around how much warming aerosol pollution from factories and vehicles may have temporarily masked.
The third approach looks deep into Earth’s past. By reconstructing surface conditions during periods like the Last Glacial Maximum (roughly 20,000 years ago), the mid-Cretaceous period, and the early Eocene, researchers can estimate how much temperature changed relative to the CO2 forcing at the time. A landmark study using this paleocalibration method found that all three periods point to a sensitivity in the 2°C to 5°C range, consistent with model-based estimates. Ice core records have been especially valuable, providing direct measurements of past CO2 levels alongside temperature markers frozen in ancient layers of ice.
A Range That Barely Budged in 40 Years
One of the more striking facts about climate sensitivity is how stable the estimate has been. In 1979, a National Academy of Sciences panel led by meteorologist Jule Charney concluded that doubling CO2 would warm the planet by 1.5°C to 4.5°C, with a best guess near 3°C. That range held essentially unchanged through five cycles of major climate assessments.
It wasn’t until the IPCC’s Sixth Assessment Report in 2021 that scientists finally narrowed the window, tightening it to 2.5°C to 4°C with a best estimate still at 3°C. The lower bound moved up from 1.5°C to 2.5°C, meaning scientists grew more confident that very low sensitivity values are unlikely. This narrowing came not from any single breakthrough but from decades of accumulated evidence across models, observations, and paleoclimate records all converging.
Earth System Sensitivity: The Longer View
Standard climate sensitivity accounts for “fast” feedbacks: water vapor, clouds, snow and sea ice, and changes in how heat moves through the atmosphere. These operate on timescales of years to decades. But Earth also has slow feedbacks that play out over centuries and millennia, including the growth and collapse of massive ice sheets, shifts in vegetation, and changes in the carbon cycle itself.
Earth system sensitivity, or ESS, folds in those slow processes. It represents the full, long-term temperature response to a CO2 change after every feedback has had time to play out. ESS is generally higher than ECS because slow feedbacks tend to amplify warming. During glacial periods, for instance, the presence of enormous ice sheets over North America and Europe created a powerful reflectivity feedback: as those ice sheets melted, the exposed land absorbed more heat, pushing temperatures higher than fast feedbacks alone would predict.
ESS is harder to pin down than ECS because you can’t observe it in real time. It relies almost entirely on deep-time paleoclimate data, piecing together how temperature and CO2 co-evolved over millions of years. But it matters for understanding the long-term consequences of today’s emissions, the warming that will continue unfolding long after CO2 levels stabilize.
Why the Exact Number Matters
The difference between 2.5°C and 4°C of sensitivity might sound modest, but it translates into vastly different futures. A world with sensitivity at the low end of the range has more room to emit CO2 before crossing dangerous temperature thresholds. A world at the high end reaches those thresholds much sooner, with a smaller remaining carbon budget.
At 3°C sensitivity, we are on track to blow past 1.5°C of warming relatively soon and face a serious challenge staying below 2°C without rapid, deep emissions cuts. If sensitivity turns out to be 4°C, the math gets significantly worse: the same amount of emissions produces more warming, and the window for action shrinks. If it’s closer to 2.5°C, the task is still urgent but marginally more achievable. Every fraction of a degree in the sensitivity estimate reshapes the timeline for climate policy, the cost of adaptation, and the severity of impacts like sea level rise, heat extremes, and ecosystem disruption.
This is why narrowing the range remains one of the most consequential tasks in climate science. The difference between the low and high ends of that 2.5°C to 4°C window is, practically speaking, the difference between a difficult future and a dramatically harder one.