The October heat record puts 2026 on course to pass 1.5°C above pre-industrial levels, the Copernicus Climate Change Service says. It puts the global mean for October 2026 at 15.35°C. As a result, it is the warmest October in its ERA5 record.
Copernicus gives the full year a greater than 99% chance of passing that level. However, the World Meteorological Organization (WMO) says one year above 1.5°C does not mean the Paris Agreement’s long-term limit has been passed.
October heat record and the annual outlook
Copernicus compares October with both recent decades and the pre-industrial past. The table shows which figures refer to the month and which cover the whole year.
| Measure | Temperature above baseline | Comparison period |
|---|---|---|
| October 2026, global | 0.90°C | October average, 1991–2020 |
| October 2026, global | 1.75°C | October average, 1850–1900 |
| Full-year 2026 forecast (95% prediction interval) | 1.56–1.61°C | Annual average, 1850–1900 |
Source: Copernicus Climate Change Service / ERA5
Data as of: 2026-10-31
The monthly readings and the full-year forecast use different time spans and reference periods.
Copernicus gives 2026 about an 80% chance of ranking second-warmest. There is also about a 20% chance it will match or beat 2024, the record year. But these are forecasts from one model. Other data or methods may give different odds, the service says.
How to read the October heat record
Copernicus uses 1991–2020 to show how warm it is compared with the recent climate, in line with WMO guidance. For the Paris goals, though, it compares recent warmth with 1850–1900. That period stands for the pre-industrial past. The figures answer different questions, so each needs a stated baseline.
An “anomaly” means how far a value lies from that baseline, the Copernicus methods guide explains. However, it does not mean each place warmed by that amount. For example, the October bulletin shows Greenland and much of southern South America were colder than usual.
A hot year and a long-term limit
WMO explains in its State of the Global Climate 2024 report that warmth varies from year to year. El Niño and La Niña, volcanoes and shifts in ocean currents all play a role. In contrast, long-term warming chiefly stems from greenhouse gas emissions and spans a much longer time.
The report says the Intergovernmental Panel on Climate Change uses 20-year averages to define future warming levels. As a result, one hot month or year cannot settle whether that long-term level has been crossed. A year can pass 1.5°C without being the same thing as a breach of the Paris goal.
How scientists build a global record
Copernicus calls ERA5 a reanalysis. It blends past weather readings with a computer model to build a picture of the climate. The readings come from stations, ships, balloons and satellites. In this way, the model helps fill gaps where direct readings are scarce.
However, the method has limits, the service says: its quality depends on the data and model used. Copernicus also says changes from a baseline tend to be more reliable than estimates of the actual temperature.
To assess what follows the October heat record, Copernicus uses a model trained on ERA5 data from 1979–2025. It predicts November–December warmth from recent land-air and sea temperatures. Therefore, the full-year result is still a forecast.
Sources: Copernicus October bulletin, Copernicus outlook, Copernicus methods, Copernicus reanalysis guide, WMO climate report
