The Berkeley Earth surface temperature record is one of the most respected and widely followed projects tracking global temperatures. They have released their evaluation of 2025 global temperatures.
We conclude that 2025 was the third warmest year on Earth since 1850. It is exceeded only by 2024 and 2023. This period, since 1850, is the time when sufficient direct measurements from thermometers exist to create a purely instrumental estimate of changes in global mean temperature. Berkeley Earth’s analysis combines 23 million monthly-average thermometer measurements from 57,685 weather stations with ~500 million instantaneous ocean temperature observations collected by ships and buoys.
The last 11 years have included all 11 of the warmest years observed in the instrumental record, with the last 3 years including all of the top 3 warmest.
The warming spike observed in 2023 to 2025 has been extreme and suggests an acceleration in the rate of Earth’s warming. The spike has multiple causes, including both natural variability and man-made global warming from the accumulation of greenhouses gases; however, as discussed below, we believe additional factors are needed to explain the full magnitude of this event. Reductions in low cloud cover and man-made sulfur aerosol pollution are likely to have played a significant additional role in recent warming.
Over the previous 50 years, global warming has preceded in an almost linear fashion, consistent with an almost linear increase in the total greenhouse gas forcing. The warming spike in 2023 to 2025 suggests that the past warming rate is no longer a reliable predictor of the future, and additional factors have created conditions for faster warming, at least in the short-term.
The global annual average for 2025 in the Berkeley Earth analysis is estimated as 1.44 ± 0.09 °C (2.60 ± 0.17 °F) above the average during the period 1850 to 1900, which is traditionally used a reference for the pre-industrial period.
Unlike 2023 and 2024, which were El Niño years, 2025 began and ended with a modest La Niña event. La Niña events are typically associated with modest cooling in the global mean temperature. While the Earth cooled in 2025 compared to the previous years, the anticipated cooling effect was relatively modest, at best.
During 2025, 9.1% of the Earth’s surface had a locally record warm annual average, including 10.6% of land areas and 8.3% of ocean areas. These areas coincided with a number of major population centers. We estimate that 770 million people — 8.5% of Earth’s population — experienced a locally record warm annual average in 2025. The largest population centers affected by record warmth in 2025 were mostly in Asia, including ~450 million people in China.
None of the Earth’s surface had a record cold annual average in 2024.
Due to natural cooling patterns that have recently begun in the oceans, it is expected that 2026 will likely be similar to 2025. The most likely outcome is that 2026 ranks as roughly the 4th warmest year since 1850, though warmer or cooler outcomes are also possible. The likelihood of record warmth in 2026 is low, but can not be ruled out.
In Berkeley Earth’s analysis the global mean temperature in 2025 is estimated to have been 1.44 ± 0.09 °C (2.60 ± 0.17 °F) above the average temperature from 1850-1900, a period often used as a pre-industrial baseline for global temperature targets. This is ~0.08 °C (~0.14 °F) cooler than the previous record high observed in 2024, and also 0.03 °C (0.05 °F) cooler than 2023. As a result, 2025 is the 3rd warmest year to have been directly observed using thermometer measurements.
Though the availability of thermometers limits direct measurements to the period since 1850, indirect evidence suggests that the Earth is currently at its warmest global average temperature for at least several thousand of years, and possibly the warmest since the last interglacial ~120,000 years ago.
The last 11 years stand out as the 11 warmest years to have been directly observed.
Since 1970, global warming has proceeded at a roughly linear pace. This roughly linear pace has been consistent in both rate and magnitude with the expected effects of increasing greenhouse gases during this period.
However, the warming spike in 2023 to 2025 appears to have deviated significantly from the previous trend. If we were to assume that global warming was continuing at the same rate as during the 50-year period 1970-2019, then the 2023 to 2025 excursion would be by far the largest deviation from that trend, with less than a 1-in-100 chance of occurring solely due to natural variability.
While 1-in-100 chance events do sometimes occur, we consider it more likely that the recent rate of global warming has been larger than expected, exceeding both the previous trend and what would be expected when considering only the observed pattern of greenhouse gas emissions.
It appears that changes in low cloud cover and reductions in man-made aerosol pollution may be responsible for additional recent warming.
That’s it. Clearly global warming, which never existed and ended in 2015, ended in 2024. How long til that ice age?
Let’s ask Joe Bastardi….