Scientists Measured How Much Extra Heat the Ocean Holds. Then They Had to Find a Comparison Big Enough.

The number the researchers needed to describe last year's ocean heat content did not fit comfortably inside the usual vocabulary of climate statistics. Terawatts felt insufficient. Petajoules were in the right neighborhood but still too granular. They settled on zettajoules — a unit so large that it rarely appears outside discussions of planetary-scale energy budgets — and even then, the figure required a frame of reference before it became anything a human mind could grip. The 2025 study published in Advances in Atmospheric Sciences, led by researchers at the Institute of Atmospheric Physics of the Chinese Academy of Sciences and drawing on oceanographic data gathered across three continents, reported that the world's oceans absorbed approximately 23 zettajoules of heat above the previous year's record in 2024. To put that in terms that don't immediately dissolve, 23 zettajoules is roughly equivalent to 37 years of global primary energy consumption — every power plant, every engine, every furnace on Earth, running continuously for nearly four decades.

That is the single-year addition. The record broken beneath it was itself the previous all-time record, which was set in 2023, which broke the record set in 2022, which broke 2021's. Nine consecutive years of record ocean heat content, each year adding to the accumulated total, each data point sitting higher than the last on a chart that, to anyone who stares at it long enough, stops looking like variability and starts looking like a slope. The oceans have absorbed an estimated 90 percent of the excess heat^[2] trapped by rising greenhouse gas concentrations since pre-industrial times. They are not a passive register of atmospheric warming. They are its primary vault.

The measurement itself is worth understanding, because ocean heat content is not something you read off a single thermometer. The primary observational backbone comes from the Argo float network^[1] — roughly 4,000 autonomous profiling devices distributed across the world's ocean basins, each one cycling between the surface and about 2,000 meters depth every ten days, radioing temperature and salinity data back to shore as it ascends. That data is then integrated with measurements from ship-based CTD casts, moored buoy arrays, and satellite altimetry, which tracks ocean surface height as a proxy for thermal expansion. The 2025 study combined these streams using ensemble analysis methods developed across research groups in China, the United States, and Europe. The result is not a single temperature reading. It is a volumetric accounting of how much thermal energy the ocean column holds, expressed as a departure from a historical baseline.

The baseline matters. Ocean heat content measurements are referenced against a long-term mean, typically calculated from the 1981–2010 period, which itself reflects a climate system already measurably warmer than the pre-industrial state. That means the record-breaking surplus being reported is a surplus above a baseline that already embeds decades of prior warming. Every zettajoule added is piling onto a foundation that was not neutral to begin with.

How Heat Hides in Depth

One of the more counterintuitive aspects of ocean heat content as a climate metric is how poorly it correlates with what people experience at the beach. Surface temperatures are volatile — they respond quickly to wind mixing, seasonal cycles, El Niño and La Niña events, and localized atmospheric conditions. Ocean heat content, particularly when measured through the upper 2,000 meters of the water column, smooths over much of that noise. What it tracks is the slower, deeper accumulation: heat that has been mixed downward and stored in layers the atmosphere does not easily reach in the short term. The 2024 record was not the result of a single hot summer's surface warming. It reflected continued downward transfer of heat that has been building in the ocean system for years.

The Atlantic Ocean showed the most pronounced warming signal in the 2024 data, consistent with patterns observed across multiple studies over the past several years. The North Atlantic in particular has been running anomalously warm, a pattern that has raised questions among oceanographers about the behavior of the Atlantic Meridional Overturning Circulation — the large-scale current system that moves warm surface water northward and cold deep water southward, regulating heat distribution across the basin. Whether the AMOC is weakening, reorganizing, or simply experiencing natural decadal variability remains one of the more actively contested questions in physical oceanography. The 2025 study did not resolve that debate, but it added another data point to a baseline that researchers tracking AMOC behavior will need to account for.

“The oceans have absorbed roughly 90 percent of the excess heat trapped since the pre-industrial era. They are not a passive register of atmospheric warming. They are its primary vault.”

The Indian Ocean and the southern Pacific also showed continued warming, and the deep ocean — below 2,000 meters — is increasingly entering the picture. Argo floats only profile to around 2,000 meters, leaving the abyssal ocean below that threshold largely undersampled. Deep-Argo floats, capable of reaching 6,000 meters, are being deployed in growing numbers but do not yet provide the spatial or temporal coverage needed for robust global accounting. Researchers acknowledge this as a gap: warming signals that have already been detected at intermediate depths suggest heat is being mixed downward more aggressively than early climate models anticipated, which has implications for how much additional thermal energy is already locked in the deep ocean and committed to future surface expression.

What Stored Heat Actually Does

Heat stored in the ocean does not simply sit there. It drives sea level rise through thermal expansion — warmer water occupies more volume, independent of any contribution from melting ice. Current estimates attribute roughly one-third to one-half of observed global mean sea level rise to thermal expansion alone. It fuels tropical cyclone intensification by providing the surface energy budget that storms draw from as they develop; warmer sea surface temperatures and a deeper pool of warm water below the mixing layer mean hurricanes and typhoons have more energy available and take longer to exhaust it as they move. And it alters atmospheric circulation patterns by changing the temperature gradients that drive large-scale pressure systems, which is one mechanism linking the warming ocean to the increasingly erratic jet stream behavior observed over the Northern Hemisphere in recent years.

Marine ecosystems feel the heat differently. Coral reefs are particularly sensitive to sustained thermal anomalies — bleaching events triggered by even modest temperature increases above the threshold a reef has adapted to can kill coral tissue in weeks. The 2024 ocean heat record coincided with what NOAA's Coral Reef Watch^[3] classified as a global mass bleaching event, the fourth such event in recorded history and the second since 2016. Mass bleaching does not automatically mean permanent die-off; reefs can recover if water temperatures return to normal within a survivable window. But recovery requires time, cooler conditions, and absence of compounding stressors like runoff, overfishing, and coastal acidification — conditions that are becoming harder to guarantee as baseline temperatures climb.

“The record broken in 2024 was itself the previous all-time record, set in 2023, which broke 2022's — nine consecutive years of ocean heat records sitting higher than the last on a chart that stops looking like variability and starts looking like a slope.”

The Acceleration Problem

The finding in the 2025 study that deserves more attention than the headline record is the rate. Ocean heat content is not just reaching new highs — the pace at which it is climbing has itself been increasing. Analysis of the long-term trend shows that the rate of heat uptake over the most recent decade is measurably faster than the rate observed in the decade before it, which was faster than the decade before that. This is not simply more energy entering a warming system at a steady rate. The rate of accumulation is accelerating. In physical terms, that distinction matters enormously: a constant rate of warming implies a system approaching a new equilibrium; an accelerating rate implies a system still responding to a forcing that is itself still growing.

The forcing in question is the atmospheric concentration of carbon dioxide and other greenhouse gases, which continue to rise. The ocean's heat uptake will continue as long as atmospheric forcing outpaces the ocean's ability to radiate energy back. There is a concept in climate physics called committed warming — the additional temperature rise already guaranteed by the greenhouse gases already in the atmosphere, even if all emissions stopped today. The ocean's current heat content partially reflects this commitment: energy the ocean has already absorbed that will eventually resurface, influencing atmospheric temperatures for decades after the original source is removed. The 2025 figures are not just a snapshot of today's ocean. They are a lower bound on tomorrow's atmosphere.

The Problem With the Zettajoule

There is a legitimate communication problem embedded in all of this. Ocean heat content as a metric is almost uniquely resistant to intuitive framing. The number is large enough to require a unit most people have never encountered. The mechanism — energy absorbed into water column depth — is invisible and slow-moving. And the impacts, while real and measurable, arrive with enough delay and geographic displacement that connecting the stored heat to the visible consequences requires a chain of physical reasoning that most news cycles are not built to carry. The 37-years-of-global-energy-consumption comparison helps, but it also risks becoming the kind of statistic that registers as impressive for a moment and then evaporates. The challenge is not just measuring ocean heat. It is making the measurement do the work it should do in the public understanding of the climate system.

Researchers working on this problem have been advocating for ocean heat content to replace or at least sit alongside global mean surface temperature as the primary public metric of climate change progress. Surface temperature is what people feel, and it has obvious intuitive appeal — a half-degree of warming sounds small, one degree sounds moderate, two degrees sounds alarming. But surface temperature is noisy, sensitive to short-term variability like El Niño, and captures only a fraction of the total energy imbalance. Ocean heat content is more stable, more physically comprehensive, and directly proportional to the energy imbalance that drives long-term climate change. It is a slower, quieter number. It is also a more honest one.

What the Record Cannot Tell You

The 2025 study is careful, and notably so, about what the data does not yet resolve. The spatial distribution of warming within ocean basins is still uneven enough that attribution — identifying which specific regional changes are driven by greenhouse forcing versus natural variability — remains difficult in many cases. The deep-ocean observational gap is real and acknowledged. And the interaction between ocean heat content and ice sheet dynamics, particularly around Greenland and West Antarctica, is an area of active and genuinely uncertain research. Warmer ocean water intruding beneath ice shelves can accelerate basal melting in ways that surface temperature measurements miss entirely, which means the consequences of the stored heat may be feeding into ice loss faster than the observational record currently captures.

“The 2025 figures are not just a snapshot of today's ocean. They are a lower bound on tomorrow's atmosphere.”

There is also the question of what comes next on the measurement side. The international Argo program is in the process of expanding both its float count and its depth range, and new satellite missions designed to improve sea surface height accuracy and ocean color measurement will add additional layers to the heat content picture over the coming years. The observational infrastructure for tracking ocean heat has never been better, and it is still improving. The irony embedded in that progress is not easy to sit with: we are getting better and better at measuring something we are not, as yet, meaningfully slowing down. The ocean keeps a ledger. Nine consecutive record years means the ledger has had nine consecutive years of entries trending the same direction. The tenth year of measurements is already underway.

References

1. Argo float network (globalocean.noaa.gov)
   Describes the Argo float network of roughly 4,000 autonomous profiling devices that provide the primary observational data for ocean heat content measurements.
2. Climate Change: Ocean Heat Content (climate.gov)
   Establishes that oceans have absorbed roughly 90 percent of excess heat trapped by greenhouse gases since pre-industrial times.
3. NOAA confirms 4th global coral bleaching event (noaa.gov)
   Confirms the fourth global coral bleaching event on record occurred in 2024, coinciding with the record ocean heat content year.
4. Ocean Heat Content Sets Another Record in 2025 (link.springer.com)
   The 2025 study published in Advances in Atmospheric Sciences that reported oceans absorbed approximately 23 zettajoules of heat above the previous year's record in 2024.