January sets an unexpected temperature record

Despite long-faded El Niño conditions, 2025 saw the warmest January on record

Both 2023 and 2024 were exceptionally warm years, at just below and above 1.5C relative to preindustrial in the WMO composite of surface temperature records, respectively. While we are still working to assess the full set of drivers of this warmth, it is clear that a sizable portion of 2024’s elevated temperatures were driven by a moderately strong El Niño event that peaked in November 2023.¹

For this reason many of us expect that 2025 will be cooler than both 2023 and 2024, and is unlikely to be the warmest year in the instrumental record (though it will very likely be in the top three warmest years).

Temperature projections for 2025 from the UK Met Office, NASA’s Dr Gavin Schmidt, Berkeley Earth and Carbon Brief, relative to pre-industrial (1850-1900) temperatures and compared to the historical average of six different datasets produced by the WMO. From Carbon Brief'‘s 2024 State of the Climate.

However, at least at the start of the year nature seems not to be following our expectations. Global temperatures were in record territory for the first three weeks of the year in the Copernicus/ECMWF ERA5 dataset, only falling out of record territory over the past few days.

As a result, January 2025 beat the prior record set in January 2024 by a sizable margin. And unlike the prior record Januaries (2007, 2016, 2020, and 2024) there is currently no El Niño event boosting global temperatures; rather, the world is in modest La Niña conditions that should, all things being equal, result in lower global temperatures.

This means that January 2025 stands out as anomalous even by the standards of the last two years. We can see this by comparing the evolution of global temperatures after the peak of the 2023 El Nino event in the figure below (which is an update of the analysis I posted here back in November 2024).

Global temperature over the past few months have exceeded or been at the upper end of what we’ve seen after any other El Nino event in the historical record. And this analysis already removes an assumed acceleration from the surface temperature record (using a ~20 year LOWESS fit).

So what does this mean? Thankfully weather models expect global temperatures are set to drop next week as the Northern Hemisphere sharply cools, making it less likely that February will also set a new record. But an unexpected record to start things off may presage higher temperatures this year than many of us thought.

1. There has historically been a roughly three month lag between peak El Niño conditions in the tropical Pacific and the peak global surface temperature response to the event, though there are some indications that the 2023/2024 El Niño diverged from this pattern.