Deep dive: Ocean circulation & heat uptake — what's working, what's not, and what's next

The ocean has absorbed roughly 90% of the excess heat trapped by anthropogenic greenhouse gas emissions since the 1970s, equivalent to approximately 14 zettajoules per year over the 2010-2025 period, according to data from NOAA's National Centers for Environmental Information. In 2024, ocean heat content reached its highest level in the instrumental record, with the upper 2,000 meters storing an additional 15 zettajoules compared to the 2005-2020 baseline. This relentless accumulation of thermal energy is altering global circulation patterns, accelerating ice sheet melt, and reshaping climate risk for coastal communities, marine ecosystems, and global food systems. Understanding the current state of ocean circulation and heat uptake research is essential for policymakers, compliance teams, and sustainability professionals working to translate climate science into actionable risk assessment.

Why It Matters

Ocean circulation and heat uptake are not abstract geophysical phenomena; they directly drive the climate impacts that organizations must now plan for under mandatory disclosure frameworks. The Atlantic Meridional Overturning Circulation (AMOC), which transports warm surface water northward and cold deep water southward, influences weather patterns across Europe, North America, and West Africa. Multiple studies published in Nature and Science between 2023 and 2025 have identified statistically significant weakening of the AMOC, with proxy reconstructions suggesting the system is at its weakest point in over a millennium. A further decline or potential collapse would trigger cascading impacts: disrupted monsoon systems in South and West Africa, accelerated sea-level rise along the US East Coast (0.5-1.0 meters above global average from dynamic effects), and altered hurricane tracks and intensities.

For organizations subject to SEC climate disclosure requirements (effective for large accelerated filers in 2026), the Task Force on Climate-related Financial Disclosures (TCFD) framework, or the EU's Corporate Sustainability Reporting Directive (CSRD), ocean circulation changes represent material physical risks that demand scenario analysis. The Intergovernmental Panel on Climate Change's Sixth Assessment Report assigned "medium confidence" to the projection that the AMOC will weaken over the 21st century under all emissions scenarios, and "low confidence" to the possibility of a complete collapse before 2100. That combination of high impact and deep uncertainty makes ocean circulation a defining variable in climate risk modeling.

The economic stakes are enormous. A 2024 study in Nature Communications estimated that AMOC weakening or collapse could reduce global GDP by 2-7% annually, with disproportionate impacts on Northern European agriculture, North Atlantic fisheries, and tropical food production. The marine heat wave that devastated North Atlantic fisheries in 2023-2024, with sea surface temperatures exceeding 1.4°C above the 1991-2020 average, provided a preview of economic disruption that ocean warming can deliver on much shorter timescales.

Key Concepts

Ocean Heat Content (OHC) quantifies the total thermal energy stored in the ocean, typically measured for the upper 700 meters (OHC700) and the upper 2,000 meters (OHC2000). OHC is calculated from temperature profiles collected by the Argo float network, ship-based measurements, and satellite altimetry. Since 2006, the Argo program has provided the primary observational backbone, with approximately 4,000 autonomous profiling floats distributed across the global ocean, each completing a 10-day cycle of diving to 2,000 meters and transmitting temperature and salinity data via satellite. OHC is expressed in joules, with the 2005-2025 trend showing accumulation of roughly 10-15 zettajoules per year in the upper 2,000 meters, equivalent to approximately 1 watt per square meter of ocean surface.

Atlantic Meridional Overturning Circulation (AMOC) is a system of ocean currents that transports roughly 17 sverdrups (17 million cubic meters per second) of warm surface water northward in the Atlantic, where it cools, densifies through evaporation and ice formation, and sinks to form North Atlantic Deep Water. This thermohaline circulation drives approximately 25% of global northward heat transport and has profound effects on regional climate. The RAPID-MOCHA monitoring array at 26.5°N, operational since 2004, provides continuous measurements of AMOC strength, revealing significant variability (13-20 Sv range) and a declining trend that has attracted intense scientific attention.

Deep Water Formation occurs in specific high-latitude regions, primarily the Labrador Sea, the Nordic Seas (Greenland, Iceland, and Norwegian Seas), and around Antarctica (Antarctic Bottom Water formation in the Weddell and Ross Seas). Freshwater input from melting ice sheets and increased precipitation reduces surface water density, potentially inhibiting convective sinking and weakening the overturning circulation. Greenland Ice Sheet melt has increased from approximately 50 Gt per year in the 1990s to over 250 Gt per year in the 2020s, contributing measurable freshwater anomalies to the North Atlantic subpolar gyre.

Marine Heat Waves (MHWs) are prolonged periods of anomalously high sea surface temperatures, defined as temperatures exceeding the 90th percentile of the local seasonal climatology for at least five consecutive days. MHW frequency has increased by 54% since the 1980s, with average duration increasing by 17%, according to a 2024 update to the Hobday et al. framework published in Progress in Oceanography. The 2023-2024 North Atlantic marine heat wave, partially linked to reduced Saharan dust (and corresponding reduction in aerosol cooling) combined with background warming trends, represented the most extreme event in the satellite record.

Ocean Deoxygenation is the progressive loss of dissolved oxygen driven by warming (reduced gas solubility and increased stratification limiting vertical mixing). The global ocean has lost approximately 2% of its dissolved oxygen since the 1960s, with oxygen minimum zones expanding by 4.5 million square kilometers. Deoxygenation interacts with circulation changes because the overturning circulation ventilates the deep ocean, and weakened circulation reduces oxygen supply to intermediate and abyssal depths.

What's Working

Argo and Deep Argo Observing Networks

The international Argo program represents one of the most successful sustained ocean observing systems ever deployed. With approximately 4,000 active floats, Argo provides near-real-time temperature and salinity profiles to 2,000 meters depth across the global ocean. The data are freely and publicly available within 24 hours of collection, enabling both operational applications (weather forecasting, seasonal climate prediction) and fundamental research. Argo has transformed ocean heat content estimates from uncertain regional analyses to globally coherent time series with quantified error bounds.

The Deep Argo extension, deploying floats capable of profiling to 6,000 meters, has expanded from pilot arrays to over 150 active floats by early 2026. Deep Argo is revealing that the abyssal ocean (below 2,000 meters) is warming at approximately 0.02°C per decade, a rate previously undetectable. While small in temperature terms, the vast volume of the abyssal ocean means this warming contributes significantly to global energy imbalance accounting and thermosteric sea-level rise. NOAA, the Japan Agency for Marine-Earth Science and Technology (JAMSTEC), Ifremer (France), and CSIRO (Australia) are the primary contributors to the Deep Argo fleet.

Satellite Altimetry and Gravimetry

Satellite missions including the Sentinel-6 Michael Freilich (launched 2020) and the GRACE-FO gravity mission provide complementary top-down constraints on ocean heat content and mass changes. Satellite altimetry measures sea surface height changes with millimeter precision, which can be decomposed into thermosteric (heat-driven expansion) and halosteric (salinity-driven) components. When combined with Argo in-situ profiles and GRACE-FO mass measurements, the satellite-Argo synthesis produces ocean heat content estimates with uncertainties reduced by 40-50% compared to either data source alone. The Surface Water and Ocean Topography (SWOT) satellite, launched in December 2022, is resolving ocean mesoscale and submesoscale features (10-100 km) for the first time from space, providing unprecedented views of eddy-driven heat transport processes.

AMOC Monitoring Arrays

The RAPID-MOCHA array at 26.5°N has provided over 20 years of continuous AMOC monitoring, establishing that the circulation exhibits substantial interannual variability (ranging from 14 to 20 Sv) superimposed on a multi-decadal weakening trend. The Overturning in the Subpolar North Atlantic Program (OSNAP), deployed since 2014 across a line from Labrador to Scotland, has revealed that the subpolar overturning (where deep water formation occurs) responds primarily to buoyancy forcing in the eastern subpolar gyre rather than the Labrador Sea, overturning a decades-old paradigm. These arrays are essential for validating climate model projections of future AMOC behavior. The RAPID dataset has become a benchmark against which every major climate model's AMOC representation is evaluated.

What's Not Working

Deep Ocean Observation Gaps

Despite the success of Argo, critical observation gaps remain. The deep ocean below 2,000 meters, representing over half the ocean's volume, remains grossly undersampled. The 150 Deep Argo floats deployed by early 2026 compare to a theoretical requirement of 1,200-1,500 floats for global coverage at the resolution needed to detect decadal trends in abyssal warming with confidence. Funding for Deep Argo expansion remains fragmented across national agencies, with no equivalent to the coordinated international funding that enabled standard Argo's rapid deployment in the early 2000s.

Additionally, under ice observation in polar regions presents fundamental technological challenges. The Arctic Ocean and waters beneath Antarctic ice shelves are among the most climate-sensitive yet least observed ocean regions. Autonomous underwater vehicles (AUVs) and under-ice floats have demonstrated feasibility in pilot deployments, but sustained, network-scale observations under ice remain years away.

Climate Model Disagreement on AMOC Trajectory

Climate models show wide disagreement on the rate, magnitude, and timing of future AMOC changes. In the CMIP6 multi-model ensemble, projected AMOC weakening by 2100 ranges from 15% to 55% under SSP3-7.0 (a high-emissions scenario), with several models showing recovery after initial weakening. This spread reflects fundamental uncertainties in how models represent deep convection, mesoscale eddy transport, ice sheet-ocean interactions, and atmospheric freshwater flux changes. A 2024 study in Science Advances by Ditlevsen and Ditlevsen used statistical early-warning signals to estimate a potential AMOC tipping point between 2025 and 2095, with a central estimate around 2050, but their methodology has been debated by other groups who argue that observational records are too short for reliable tipping-point detection.

The practical consequence for risk assessment is that policymakers and corporate climate teams face deep uncertainty about a potentially catastrophic risk. Existing scenario analysis frameworks (NGFS climate scenarios, IPCC SSP pathways) do not adequately capture the tail risk of abrupt AMOC changes, creating a disconnect between scientific possibility and financial risk modeling.

Oxygen Monitoring Deficiency

While ocean deoxygenation is recognized as a critical environmental stressor (affecting fisheries, nutrient cycling, and carbon sequestration), routine dissolved oxygen observations lag far behind temperature and salinity monitoring. Approximately 30% of Argo floats carry biogeochemical sensors including oxygen, but the spatial and temporal coverage is insufficient to detect trends in oxygen minimum zone expansion with the same confidence as temperature trends. The Global Ocean Oxygen Network (GO2NE), supported by UNESCO's Intergovernmental Oceanographic Commission, has called for tripling the biogeochemical Argo fleet, but funding commitments remain below the required level.

What's Next

OneArgo: The Full-Depth, Global, Multidisciplinary Vision

The OneArgo design, endorsed by the oceanographic community and outlined in a 2020 OceanObs decadal plan, calls for expanding to 4,700 standard floats, 1,200 Deep Argo floats, and 1,000 biogeochemical Argo floats. Full implementation would provide the first globally integrated, full-depth, multidisciplinary ocean observing system, enabling detection of deep ocean warming trends, oxygen changes, and carbon uptake with decadal-scale resolution. The estimated annual cost of OneArgo is approximately $80 million, roughly double the current Argo budget but modest relative to the societal value of improved climate projections. Several national agencies have increased commitments, but full funding remains a work in progress.

AI and Machine Learning for Ocean Analysis

Machine learning is accelerating ocean science by filling observational gaps and improving predictions. Neural network approaches are being used to reconstruct global ocean temperature fields from sparse Argo observations, achieving reconstruction skill comparable to adding 30-40% more physical floats. Deep learning models trained on satellite altimetry and in-situ data are improving seasonal AMOC predictions from 6 to 12 months lead time. NOAA's Geophysical Fluid Dynamics Laboratory and the UK Met Office are integrating ML-augmented ocean analysis into operational climate prediction systems, with initial results showing 10-20% improvements in seasonal forecast skill for North Atlantic sea surface temperatures.

AMOC Tipping Point Detection and Early Warning Systems

Developing reliable early warning indicators for AMOC collapse is among the highest priorities in climate science. Current approaches include statistical analysis of variance and autocorrelation changes in AMOC time series (potential indicators of approaching a tipping point), fingerprint methods based on spatial patterns of sea surface temperature and salinity, and targeted high-resolution modeling experiments. The EPOC (Explaining and Predicting the Ocean Conveyor) project, funded by the EU's Horizon Europe program, is coordinating international efforts to assess AMOC stability using a combination of observations, paleoclimate reconstructions, and model experiments. Results expected by 2027-2028 may significantly narrow uncertainty bounds on AMOC tipping risk.

Key Players

Established Leaders

NOAA (National Oceanic and Atmospheric Administration) operates the largest US ocean observing infrastructure, including contributions to Argo, RAPID-MOCHA, and satellite missions. NOAA's Pacific Marine Environmental Laboratory and Atlantic Oceanographic and Meteorological Laboratory lead US ocean heat content and circulation research.

UK National Oceanography Centre (NOC) manages the RAPID-MOCHA array and leads the OSNAP eastern subpolar array, providing two decades of continuous AMOC monitoring data that underpin scientific understanding of circulation variability.

Ifremer (French Research Institute for Exploitation of the Sea) is the single largest contributor to the global Argo float fleet and leads Deep Argo development in the North Atlantic, deploying over 200 Deep Argo floats since 2018.

Emerging Startups

Sofar Ocean develops low-cost ocean sensing platforms (Spotter buoys) and has deployed over 2,000 devices globally, providing real-time sea surface temperature and wave data that complement Argo subsurface profiles.

Saildrone operates autonomous surface vehicles capable of sustained ocean transects lasting 6-12 months, collecting atmospheric and oceanographic data in regions too remote or hazardous for crewed vessels, including the Arctic and Southern Oceans.

Terradepth is developing autonomous underwater vehicles for persistent deep-ocean observation, targeting applications in seafloor mapping and abyssal monitoring that could complement Deep Argo floats.

Key Investors and Funders

US National Science Foundation (NSF) provides primary funding for US oceanographic research, including the OSNAP program and Deep Argo technology development, with annual ocean science funding of approximately $450 million.

European Commission Horizon Europe funds major multinational ocean and climate research programs including EPOC and the Mission Starfish initiative for healthy oceans, allocating approximately EUR 350 million annually to ocean research.

Schmidt Ocean Institute (founded by Eric and Wendy Schmidt) operates the research vessel Falkor(too), providing ship time and technology platforms for deep-ocean research, autonomous vehicle testing, and ocean exploration at no cost to scientists.

Action Checklist

• Incorporate AMOC weakening scenarios into organizational climate risk assessments, testing both gradual decline (15-55% weakening by 2100) and abrupt collapse possibilities
• Review exposure to North Atlantic sea-level rise, which could exceed global average by 0.5-1.0 meters under AMOC weakening scenarios, affecting US East Coast and Northern European coastal assets
• Monitor RAPID-MOCHA and OSNAP real-time AMOC data (publicly available from RAPID project website) for early indicators of circulation changes
• Assess supply chain exposure to marine heat wave disruption, particularly for organizations dependent on North Atlantic fisheries, aquaculture, or marine-dependent tourism
• Update physical climate risk models to include ocean-driven coastal flooding scenarios using the latest IPCC AR6 sea-level projections with AMOC-related regional adjustments
• Engage with insurers and reinsurers on how ocean circulation changes affect coastal property risk pricing and availability
• Track regulatory developments around ocean-climate-related disclosure requirements, particularly under CSRD and ISSB standards
• Support sustained ocean observing through philanthropic or corporate ocean stewardship programs that fund Argo and Deep Argo operations

FAQ

Q: Is the AMOC really at risk of collapsing, and what would that mean? A: The scientific evidence indicates the AMOC has weakened and may be approaching a tipping point, though significant uncertainty remains about timing. A complete collapse (not experienced in over 12,000 years) would trigger abrupt cooling over Northern Europe (potentially 5-10°C), dramatic shifts in tropical rainfall patterns affecting billions of people, accelerated sea-level rise of 0.5-1.0 meters along the US East Coast from altered ocean dynamics, disruption of the Amazon and West African monsoons, and collapse of major fisheries. The IPCC rates a collapse before 2100 as "low likelihood, high impact," but recent observational and modeling studies suggest the risk may be higher than previously assessed.

Q: How does ocean heat uptake affect sea-level rise projections? A: Ocean thermal expansion (thermosteric sea-level rise) accounts for approximately 40% of observed sea-level rise since the 1990s. As ocean waters warm, they expand, raising sea levels independently of ice sheet melt and glacier contributions. Current ocean heat accumulation rates imply roughly 1.0-1.5 mm per year of thermosteric sea-level rise, which is essentially irreversible on human timescales because the deep ocean thermal equilibrium time is centuries to millennia. Even if atmospheric greenhouse gas concentrations were stabilized today, ocean thermal expansion would continue for centuries as the deep ocean slowly equilibrates.

Q: What data sources should we use for climate risk modeling related to ocean changes? A: For sea surface temperature and marine heat wave monitoring, use NOAA's OISST v2.1 dataset and the Marine Heatwave Tracker. For ocean heat content trends, reference NOAA/NCEI's global OHC time series and the Cheng et al. dataset maintained by the Institute of Atmospheric Physics, Chinese Academy of Sciences. For AMOC monitoring, access real-time data from the RAPID project at rapid.ac.uk. For sea-level projections incorporating ocean dynamics, consult the IPCC AR6 Sea Level Projection Tool and NASA's Sea Level Portal. All datasets are publicly accessible and regularly updated.

Q: How does ocean circulation research translate into actionable business decisions? A: Ocean circulation changes affect business through several channels. Coastal real estate and infrastructure face elevated flood risk from dynamic sea-level changes driven by circulation shifts. Agricultural operations in regions dependent on AMOC-influenced climate patterns (Northern Europe, West Africa) face productivity risk. Insurance and reinsurance companies must model changing marine peril frequencies. Fisheries and aquaculture operations face range shifts and productivity changes. Shipping routes may be affected by changing sea ice, currents, and weather patterns. The most immediate actionable step is incorporating ocean-driven scenarios into existing climate risk assessment processes required under TCFD, CSRD, and SEC frameworks.

Sources

• Cheng, L., et al. (2025). "Record-setting Ocean Warmth Continued in 2024." Advances in Atmospheric Sciences, 42(3), 1-10.
• IPCC. (2021). Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report. Cambridge: Cambridge University Press.
• Ditlevsen, P. and Ditlevsen, S. (2023). "Warning of a Forthcoming Collapse of the Atlantic Meridional Overturning Circulation." Nature Communications, 14, 4254.
• Smeed, D., et al. (2024). "Twenty Years of Continuous AMOC Observations at 26°N: Variability, Trends, and Early Warning Signals." Journal of Geophysical Research: Oceans, 129(5), e2023JC020500.
• Roemmich, D., et al. (2019). "On the Future of Argo: A Global, Full-Depth, Multi-Disciplinary Array." Frontiers in Marine Science, 6, 439.
• NOAA National Centers for Environmental Information. (2025). Global Ocean Heat Content Time Series. Silver Spring, MD: NOAA.
• Hobday, A.J., et al. (2024). "Marine Heatwaves: Updated Definition, Trends, and Future Projections." Progress in Oceanography, 218, 103120.