A planet that receives from the Sun exactly what it returns to space has a stable temperature. The whole of global warming lies in the fact that this equality has been broken for about a century and a half, by a little over one watt per square metre — a gap of around half a percent between what comes in and what goes out.
The figure looks trivial. It is not, for a reason that is far from intuitive: it applies permanently, over the entire surface of the globe, and what it accumulates does not go away.
Where the energy goes
The first surprise, when you follow that energy, is that it barely goes where we feel it. Air is a poor heat reservoir: little is needed to warm it, and it holds no more than that. The ocean, for equal volume, has a heat capacity roughly a thousand times greater — and it is deep.
The consequence: ocean heat content is the least noisy measure of the imbalance. An El Niño year makes air temperature jump by a tenth or two of a degree, then fall back; the ocean rises without a jolt, because it integrates the imbalance rather than suffering it Johnson et al., 2020.
NOAA National Centers for Environmental Information — Global Ocean Heat and Salt Content · accessed 15 November 2026
This is why a headline such as "warming has paused" has been written several times about air temperature, and never about the ocean Cheng et al., 2024.
A mean temperature that exists nowhere
The quantity everybody quotes — "+1.2 °C since the pre-industrial era" — is not a direct measurement. No thermometer reads the mean temperature of the globe. What is built instead is an anomaly: station temperatures are not averaged, their departures from their own normal are.
That shift is not a technicality. Absolute temperatures depend on altitude, on distance from the sea, on exposure; two stations ten kilometres apart can differ by several degrees permanently. Their anomalies, on the other hand, resemble one another across hundreds of kilometres, because a cold front cools everyone at once. That is what makes interpolation between stations legitimate.
NASA Goddard Institute for Space Studies — GISS Surface Temperature Analysis (GISTEMP v4) · accessed 15 November 2026
Arranged by decade, the same series says something the annual curve drowns in its own noise: no decade has been colder than the one before it since the 1960s.
Global mean temperature anomaly, by decade
Bar heights are proportional to the gap from the period's minimum, -0.33 °C for the 1910 decade. The 2020 decade has fewer than ten years on record. NASA Goddard Institute for Space Studies — GISS Surface Temperature Analysis (GISTEMP v4) · accessed 15 November 2026
Land warms twice as fast
The global mean has one flaw: nobody lives in it. Two thirds of the Earth's surface is ocean, which warms markedly more slowly than the continents — it spreads the heat it receives through a thick layer, and devotes a good part of it to evaporation.
NOAA National Centers for Environmental Information — Climate at a Glance — Global Time Series · accessed 15 November 2026
The gap between the two curves is one of the reasons the global figure sits consistently below what people observe where they live. This is not a contradiction: it is the difference between a surface average and an experience lived on dry land.
What drives it: radiative forcing
Radiative forcing is the direct measure of the imbalance imposed on the radiative balance, in watts per square metre. NOAA publishes it every year for long-lived greenhouse gases, from measured concentrations — not from a model.
NOAA Global Monitoring Laboratory — The NOAA Annual Greenhouse Gas Index (AGGI) · accessed 15 November 2026
What is already committed
One question keeps coming back: if all emissions stopped tomorrow, how much further would the Earth warm? Intuition says "a lot", since the ocean has not finished absorbing. The answer from models is more subtle: stopping emissions makes CO₂ concentration fall faster than the ocean gives its heat back, and the two effects largely cancel. The additional committed warming is on the order of a few hundredths of a degree, not a few tenths MacDougall et al., 2020.
That result has a direct political bearing, and it is one of the few firmly established pieces of good news in the field: the level at which warming stops depends essentially on what is emitted from now on, not on an inescapable inheritance.
What projections say, and what they do not
End-of-century trajectories are not forecasts. They are answers to emission assumptions, and the spread between them measures a social choice first, not a scientific uncertainty. The constraint brought by the trend observed since 1970 narrows the high end appreciably, without changing the fact that most of the dispersion comes from the chosen scenario Tokarska et al., 2020.
IPCC WGI Interactive Atlas — Interactive Atlas — CMIP6 aggregated regionally · accessed 15 November 2026
The annual stocktake of climate change indicators, which updates these quantities every year with the same method, is the safest entry point for following them Forster et al., 2026.
Sources
- GISS Surface Temperature Analysis (GISTEMP v4) — NASA Goddard Institute for Space Studies
- Global Ocean Heat and Salt Content — NOAA National Centers for Environmental Information
- The NOAA Annual Greenhouse Gas Index — NOAA Global Monitoring Laboratory
Studies cited in this brief
Piers M. de F. Forster et al. · Earth system science data · 2026
- Climate observation
Steven C. Sherwood et al. · Reviews of Geophysics · 2020
- Modelling and projections
Lijing Cheng et al. · Advances in Atmospheric Sciences · 2024
- Ocean and sea level
- Climate observation
Gregory C. Johnson et al. · Nature Climate Change · 2020
- Ocean and sea level
- Climate observation
Katarzyna B. Tokarska et al. · Science Advances · 2020
- Modelling and projections
Is there warming in the pipeline? A multi-model analysis of the Zero Emissions Commitment from CO 2
Open accessAndrew Hugh MacDougall et al. · Biogeosciences · 2020
- Ocean and sea level
- Modelling and projections