Earth\u2019s oceans ended 2025 with more heat stored in their upper 2,000 meters than in any other year since modern measurements began, after an estimated increase of 23 zettajoules. The increase was recorded even though the average sea-surface temperature eased from the extraordinary high of 2024.<\/p>\n
The finding comes from an international annual synthesis led by Yuying Pan of the Chinese Academy of Sciences and published in Advances in Atmospheric Sciences<\/a>. Its central estimate was 23 plus or minus 8 zettajoules. In other words, the headline number is the middle of a substantial uncertainty range, not a perfectly exact reading.<\/p>\n This is one annual synthesis, not settled consensus on a quantity known to the last joule. But independent observing products used in the paper, including Copernicus Marine, agreed on the important point: ocean heat content reached another record in 2025.<\/p>\n How 23 zettajoules becomes 37 years<\/p>\n A zettajoule is 1021 joules. The prefix is so large that it hides rather than explains the scale. Converting the number into the unit used in global energy statistics helps: 23 zettajoules equals 23,000 exajoules.<\/p>\n The Energy Institute reported<\/a> that global primary-energy consumption reached about 620 exajoules in 2023. Dividing 23,000 by 620 gives 37.1. That is where the comparison in the headline comes from.<\/p>\n It is a scale comparison, not a physical explanation. The ocean did not absorb 37 years of electricity generated by people, and the calculation should not be confused with global electricity use, which is much smaller than total primary-energy consumption. It says that the estimated one-year increase in upper-ocean heat was about 37 times a recent year\u2019s consumption of energy from all human sources.<\/p>\n The benchmark is also not timeless. Annual energy use changes, and different statistical conventions can produce slightly different totals. \u201cRoughly 37 years\u201d is therefore the honest form of the comparison. ScienceBlog\u2019s first report on the 2025 result<\/a> used the same translation when the paper appeared in January. The useful next step is to keep its boundaries visible.<\/p>\n What ocean heat content actually measures<\/p>\n Ocean heat content is not the ocean\u2019s temperature at one point. It is an estimate of how much thermal energy is stored across a huge volume of water. Researchers build it from temperature profiles taken at many locations and depths, then account for the density and heat capacity of seawater.<\/p>\n Modern estimates rely heavily on robotic profiling floats, alongside ship observations, moorings and other instruments. A float measures a vertical column, not an entire basin. Scientists have to correct known instrument biases, combine observations made at different times and estimate conditions across the gaps. Coverage was especially sparse before the global observing network expanded in the 2000s.<\/p>\n That is why the uncertainty matters. Pan and colleagues report an increase of about 23 plus or minus 8 zettajoules for the surface-to-2,000-meter layer relative to 2024. It is a year-to-year change, not the total heat already in the ocean. It also does not directly include the deepest water below 2,000 meters.<\/p>\n A separate full-depth assessment in Nature Reviews Earth & Environment<\/a> estimated that the whole ocean gained 24 plus or minus 6 zettajoules from 2024 to 2025. The closeness of the central values is reassuring, but the estimates use different data combinations and cover different depth ranges. They should not be treated as two instruments returning the same exact reading.<\/p>\n The surface cooled while the ocean gained heat<\/p>\n In the datasets assessed by Pan\u2019s team, the global mean sea-surface temperature for 2025 was 0.49 degrees Celsius above the 1981 to 2010 average. It was 0.12 plus or minus 0.03 degrees below 2024, yet remained the third-highest annual value on record.<\/p>\n There is no contradiction here. Sea-surface temperature describes the thin, responsive boundary between ocean and atmosphere. It can move noticeably with winds, evaporation and natural patterns such as El Ni\u00f1o and La Ni\u00f1a. Ocean heat content adds temperature changes through a layer reaching two kilometers down. It is a measure of the reservoir, not merely its skin.<\/p>\n Conditions evolved toward La Ni\u00f1a during 2025. That change rearranged winds and currents across the tropical Pacific, affecting where heat sat and how readily it was exchanged with the atmosphere. The surface could step down from 2024\u2019s record while the much larger volume below continued to accumulate energy.<\/p>\n