TODAY IN NUMBERS
WEATHER · SATIRE · COMPOSITION FALLACY

October 2026 Was the Warmest October on Record Globally, and Over Land It Came Joint Third

Copernicus puts the month 1.75C above pre-industrial levels, the third in a row above 1.5C. The land surface finished behind 2023 and 2024.

October 2026 was the warmest October on record globally, with an average surface air temperature of 15.35°C, according to the Copernicus Climate Change Service, which is implemented by the European Centre for Medium-Range Weather Forecasts.

The month was 0.90°C above the 1991-2020 average for October, surpassing the previous record for the month, set in 2023, which stood 0.85°C above average.

Measured against the estimated 1850-1900 pre-industrial average, October 2026 was 1.75°C warmer. It was the third consecutive month more than 1.5°C above that baseline. Before August 2026 the last month to exceed 1.5°C was November 2025, and the previous run of at least three consecutive such months ended in April 2025.

The 12 months from November 2025 to October 2026 averaged 0.63°C above the 1991-2020 reference and 1.51°C above pre-industrial levels.

For the year so far, the January-to-October average places 2026 second on record at 0.64°C above the 1991-2020 average: 0.07°C below the same period in 2024 and 0.05°C above 2023.

A statistical model trained on past ERA5 data puts the full-year 2026 anomaly between 0.68°C and 0.73°C above the 1991-2020 average, a 95 per cent prediction interval corresponding to 1.56 to 1.61°C above pre-industrial levels. On that model 2026 is most likely to finish as the second-warmest year on record, at about 80 per cent probability, with about a 20 per cent chance of tying or exceeding 2024, which at 0.72°C above average is the warmest year measured.

The same model makes 2026 virtually certain, at more than 99 per cent probability, to exceed 1.5°C, which would make it the second calendar year to do so in the ERA5 record after 2024. Because the 2023-2025 average is already above that level, a further such year would put the four-year average for 2023 to 2026 above 1.5°C as well.

In Europe the month averaged 11.08°C over land, 1.49°C above the 1991-2020 reference and the second-warmest October in the record, 0.43°C cooler than the warmest, in 2022. The largest anomalies were over Ukraine, up to 6°C above average. Southern Italy, the southern Balkans, Greece, the eastern Mediterranean and western Türkiye ran about 2°C below average.

At sea the extra-polar ocean, between 60°S and 60°N, averaged 20.95°C, the highest October figure on record and the fourth consecutive month to set a record for its month. In the Niño 3.4 region the October sea surface temperature reached 29.94°C, an all-time high, and its anomaly of 3.21°C above the 1991-2020 average beat the previous record of 2.99°C, set in September. Copernicus cites the World Meteorological Organization's assessment that the El Niño event is strengthening rapidly, is expected to peak in December and may become one of the strongest on record.

One figure in the bulletin sits awkwardly beside the headline, and Copernicus states it plainly. Over global land, October 2026 was the joint third warmest on record, at 10.82°C, level with October 2025 and behind only October 2023 and 2024.

So the warmest October ever measured was not the warmest October on any land surface. The record was set in the average, and the average includes the sea.

This newspaper notes the implication for anyone who spent the month outdoors. The land came third. The record belongs to a surface whose thermometers are mostly buoys.

The Department of Unnecessary Research has pursued the point into the annual figures, where it holds again. The year containing the warmest October on record is predicted to finish second. The record month sits inside a runner-up year, and the runner-up year sits inside a four-year average that clears the threshold no single one of its years was meant to reach.

Our analysts have arranged the findings in descending order of confidence. October was the warmest October. The land was not the warmest land. The year will probably not be the warmest year. The four years taken together will be warm enough to count.

It remains only to establish which of these is the weather. The Department's working position is that the 1.5°C limit has now been exceeded by an average of places, over a span of years, in a month that nowhere on land found exceptional.

JANUARY TO OCTOBER, ABOVE THE 1991-2020 AVERAGEQUESTIONABLE ANALYTICS
592023
712024
642026
January-to-October global temperature anomalies above the 1991-2020 average, in hundredths of a degree Celsius. Copernicus gives 2026 as 0.64C, which it states is 0.05C above 2023 and 0.07C below 2024; the 2023 and 2024 figures here follow from those two statements.
THE BORING FACT

October 2026 was the warmest October on record globally at 15.35C, 0.90C above the 1991-2020 average and 1.75C above the estimated 1850-1900 pre-industrial average. The average temperature over global land was the joint third highest on record for October, at 10.82C.

Source: Copernicus Climate Change Service
OUR PROFESSIONAL CONCLUSION

Because the record was set in a global average dominated by the ocean while the land finished joint third, and because the year holding the warmest October will most likely finish second inside a four-year average that passes 1.5C, the threshold has been exceeded by an average of places over a span of years in a month no land surface found exceptional.

REALITY CHECK · COMPOSITION FALLACY

Unfortunately, there's a small problem.

A global average combines land and ocean, and the ocean's far greater area makes it the larger term, so a record global month with land in joint third place is ordinary arithmetic rather than a contradiction. Land at joint third warmest in the whole record is itself an extreme result, not an unexceptional one.

THE BS DETECTOR

Where did reality end?

Tap the paragraph where this story stopped being true. Then see where everyone else bailed.

 

READER CREDULITY METER

Did you believe this story before the twist?

Be the first reader to answer.

YOU WRITE THE CONCLUSION

Think you can end this story better than our analysts?

One deadpan conclusion, 280 characters at most. Start from the facts above and take them one step too far. Our editor picks the best one each morning, and it runs in the next edition.

280 left
No real people's private lives, no tragedy, nothing that accuses anyone of anything. Our editor reads every entry and prints one; the rest are never published.