Menu

Earth science

How much will Earth warm if CO₂ doubles?

That more carbon dioxide warms the planet has been known for over a century. Exactly how much is a question that took decades to narrow — and is still not closed.

open since 1896 · partly answered · updated 18 October 2026

The question

Carbon dioxide lets sunlight through but absorbs part of the heat the Earth radiates back to space. More CO₂ therefore means a warmer planet. This basic physics has been understood since the 19th century and is not in dispute.

The open question is quantitative. Climate scientists summarise it in a single number, the equilibrium climate sensitivity: how much the average surface temperature would eventually rise if the amount of CO₂ in the air doubled compared with pre-industrial times and then stayed there. The number matters enormously. A low value would mean slower and milder change; a high one, faster and more severe change for the same emissions.

This question is marked as partly answered. The range has narrowed considerably, but it has not closed.

What we know

  • In 1896 the Swedish chemist Svante Arrhenius made the first calculation, by hand, and estimated that doubling CO₂ would warm the Earth by about 5–6 °C.
  • In 1967 Syukuro Manabe and Richard Wetherald built the first physically consistent model of how heat moves through the atmosphere and found about 2.3 °C. Manabe shared the Nobel Prize in Physics in 2021 for this line of work.
  • In 1979 a US National Academy of Sciences panel led by Jule Charney estimated the sensitivity at 1.5 to 4.5 °C. Remarkably, this range stayed almost unchanged through more than three decades of research and several assessment reports.
  • Doubling CO₂ alone, with nothing else changing, would warm the Earth by only about 1.2 °C. Most of the expected warming comes from feedbacks: warmer air holds more water vapour, itself a greenhouse gas; melting ice and snow reflect less sunlight; and clouds change.
  • In 2020 a large international assessment combined three independent lines of evidence: the physics of feedbacks, the warming observed since the industrial era, and past climates such as the last ice age. It concluded that the sensitivity is most likely between about 2.6 and 3.9 °C.
  • The Intergovernmental Panel on Climate Change, in its 2021 report, gave a best estimate of 3 °C, with a likely range of 2.5 to 4 °C — the first substantial narrowing since 1979. Newer studies pull in both directions — one combining ice-age evidence with temperature patterns narrowed the range to 2.4–3.5 °C, another argued for about 4.8 °C — but none has displaced the IPCC estimate.

Main hypotheses

Here the alternatives are not rival explanations but different values.

Around 3 °C

The central estimate supported by all lines of evidence together. Most current work aims to narrow the range around it.

Lower

Estimates based only on the warming observed so far tend to give lower values. But the pattern of recent warming — for example, parts of the Pacific have warmed less than expected — may temporarily mask the full sensitivity. If that pattern changes, warming could speed up.

Higher

Some of the newest climate models give values above 5 °C. Many of them struggle to reproduce past climates, and researchers now treat them with caution, but the possibility of a high value cannot be excluded.

Where understanding ends

The largest uncertainty is clouds. Low clouds over the oceans cool the planet by reflecting sunlight. Whether they thin out as the world warms, and by how much, depends on processes far smaller than a climate model's grid, such as turbulence and droplet formation. Small differences in how models represent them produce large differences in sensitivity.

A second difficulty is that sensitivity is not one fixed number. It may depend on the pattern of warming and on the climate's starting state, so values derived from the ice age, from today's warming and from models need not exactly agree.

And time matters. "Equilibrium" means centuries; what affects people in the coming decades depends more on how fast warming unfolds, which brings in the slow uptake of heat by the deep ocean.

What would settle it

  • Better understanding of low clouds, from satellite observations, field campaigns and high-resolution simulations that resolve the processes models now approximate.
  • Longer and more precise records of Earth's energy balance — how much more energy the planet absorbs than it emits.
  • Climate models that reproduce both the recent warming pattern and past climates, and agree on why.

Sources