A record-hot ocean, year after year

If you want to know what the climate is doing, the air temperature is a noisy answer. The oceans absorb roughly 90 percent of the excess heat trapped by greenhouse gases, so the heat content of the sea is the cleaner instrument. On 10 January 2025, an international team of 54 scientists from seven countries published their annual account of that instrument in Advances in Atmospheric Sciences. ‘Record High Temperatures in the Ocean in 2024’, led by Cheng Lijing of the Institute of Atmospheric Physics, reports that 2024 set a new high for both global sea-surface temperature and the heat stored in the upper 2,000 metres, each one a record in the modern observational record.

The headline number is the growth in ocean heat content. Between 2023 and 2024, the 0-to-2,000-metre heat content rose by 16 plus or minus 8 zettajoules in the Chinese analysis, and by 18 plus or minus 7 zettajoules in the independent Italian reanalysis. A zettajoule is ten to the twenty-first joule. To make that legible, the authors note that the 16 zettajoules added in a single year is around 140 times the total electricity the world generated in 2023. Over the previous five years the upper ocean had gained roughly 15 to 20 zettajoules each year, a steady climb that the El Niño and La Niña cycles merely modulated rather than reversed.

Three independent datasets agree, which is the study’s real strength. The paper releases four international ocean products from China, Italy, the European Copernicus service and the United States NOAA, and all four show the same direction. The rise in global mean sea-surface temperature from 2023 to 2024 is reported as 0.07, 0.05 and 0.06 degrees Celsius across the three analysis centres, each a new modern high. Consistency across independent observing systems is what turns a single hot year into a trend rather than a measurement dispute.

The regional pattern is what a climate scientist would call instructive. The Indian Ocean, the tropical Atlantic, the North Atlantic, the Mediterranean, parts of the North Pacific and the Southern Ocean all set local heat-content records, a widening of warming across almost every basin rather than a single hot patch. The tropical Pacific, by contrast, swung with the El Niño cycle, which is exactly the variability the authors expect from natural modes riding on top of a long-term rise. The heat has even accumulated near both poles.

The authors tie the ocean’s state to the weather that followed on land. In the twelve months of the dataset, 104 countries recorded their hottest temperatures on record. The paper notes the coral bleaching on the Great Barrier Reef in March 2024, the successive hurricanes in the Caribbean and the US southeast, and the catastrophic Spanish flooding in late October. The causal link is not that a zettajoule caused any one event, but that a hotter ocean loads the atmosphere with water vapour, a powerful greenhouse gas in its own right, and sharpens the hydrological extremes that follow.

The honest caveats are real but they point the same way. Much of the deep record relies on the Argo array, a drifting network of nearly 4,000 profiling floats that only reached full coverage around 2020, so the uncertainty on the early decades is larger than on the recent years. The authors report those uncertainties rather than hide them. What survives the caveats is the central claim, stated plainly by the lead author: the breaking of ocean records has itself become a broken record. A year that sets another record is no longer the story. The story is that the ocean keeps absorbing the surplus heat, and there is no sign yet that it will stop.