The claim: global warming has paused, stalled or slowed down.

The verdict: not supported. The most recent analysis points the other way — a peer-reviewed study published this year in Geophysical Research Letters reports that warming has accelerated significantly over the past decade.

Where the claim comes from

The pause claim is not invented from nothing, which is why it keeps returning. It comes from a real feature of temperature records: they are noisy.

Global surface temperature moves year to year for reasons unrelated to the long-term trend. El Niño pushes it up; La Niña pulls it down. Large volcanic eruptions cool the surface for a year or two. Ocean heat exchange redistributes energy. The result is a jagged line with a clear upward slope, and a jagged line offers plenty of short segments that look flat if you choose the endpoints carefully.

Starting a trend line at an unusually hot year — a strong El Niño — and ending at a cooler one produces an apparent pause. This is a well-documented selection problem rather than a discovery, and it recurs in the same form every few years.

What the recent analysis did

Researchers at the Potsdam Institute for Climate Impact Research approached the noise problem directly: rather than picking start and end points, they statistically removed the short-term natural influences — the El Niño cycle and similar — to isolate the underlying human-driven trend.

Their conclusion is that the underlying rate of warming has increased in recent years. The study is Foster and Rahmstorf, published in Geophysical Research Letters in 2026, and it is a specific, checkable citation rather than a summary of opinion.

Two caveats belong here. A single paper is one contribution, not a consensus, and the causes of an acceleration — reduced aerosol pollution altering how much sunlight is reflected, versus changes in cloud cover, versus emissions — are actively debated among researchers. What is not seriously debated is the direction.

The test you can apply yourself

Any claim about a temperature trend can be checked with three questions, and they work regardless of which direction the claim points.

• What are the start and end years, and why those? A trend claim that depends on its endpoints is not a trend claim.
• Is the period long enough? Climatologists generally treat periods under roughly thirty years as too short for surface temperature trends to emerge from natural variability with confidence.
• Has natural variability been accounted for? A raw record and a record with El Niño removed can look quite different over the same decade.

The same discipline applies in reverse. A single record-hot summer is not by itself evidence of acceleration either, and treating it as such makes the opposite error.

Why this one keeps coming back

Because the underlying data really is noisy, the pause claim can be revived after almost any cool year without anyone needing to fabricate anything. That makes it unusually durable compared with claims that depend on outright false figures.

It also means the correction has to be about method rather than about numbers. The answer to “temperatures were flat from year X to year Y” is not a different set of temperatures. It is: those are the right numbers, and that is the wrong way to read them.

We applied the same reasoning this week to the phrase “hundred-year flood”, which misleads for a related reason: a real statistic, described in a way that invites the wrong inference.

Sources

• Potsdam Institute for Climate Impact Research via ScienceDaily — Foster and Rahmstorf, “Global Warming Has Accelerated Significantly”, Geophysical Research Letters, 2026