Argo floats vs satellite altimetry vs reanalysis models: comparing ocean heat uptake measurement approaches

Why It Matters

The ocean has absorbed more than 90 percent of the excess heat trapped by greenhouse gases since 1970, equivalent to roughly 14 zettajoules per year over the past two decades (Cheng et al., 2024). In 2024, global ocean heat content reached its highest value on record, with the upper 2,000 meters storing approximately 287 zettajoules above the 1981 to 2010 baseline (WMO, 2025). How accurately we quantify that uptake shapes everything from sea level rise projections that inform coastal infrastructure investments to IPCC carbon budget calculations that underpin national climate pledges. Yet three fundamentally different measurement approaches compete for attention and funding: autonomous Argo profiling floats that sample the water column directly, satellite altimeters that infer thermal expansion from orbit, and reanalysis models that fuse observations with numerical simulations. Each approach carries distinct trade-offs in accuracy, spatial coverage, temporal resolution, and cost. For sustainability professionals evaluating climate risk analytics, selecting the wrong data product can introduce biases of 0.2 to 0.5 watts per square meter into energy balance estimates, enough to distort decadal warming projections by 20 to 30 percent (Johnson et al., 2025). This guide provides a structured comparison to help organizations choose the right approach, or the right combination, for their needs.

Key Concepts

Ocean heat content (OHC) quantifies the thermal energy stored in a defined ocean volume. Researchers typically report OHC for the upper 700 meters, the upper 2,000 meters, and the full depth. Changes in OHC are the single largest term in Earth's energy imbalance, currently estimated at 1.3 watts per square meter (Loeb et al., 2025).

In-situ profiling refers to instruments physically immersed in seawater that record temperature, salinity, and pressure at discrete depth levels. The Argo program deploys roughly 4,000 battery-powered floats worldwide, each cycling between the surface and 2,000 meters every 10 days before transmitting data via satellite (Argo Program, 2025).

Satellite altimetry measures the height of the sea surface to millimeter precision using radar pulses. Because seawater expands as it warms, changes in sea surface height (SSH) serve as a proxy for heat content changes once steric, mass, and dynamic contributions are separated using auxiliary datasets.

Reanalysis products are gridded, gap-filled reconstructions of the ocean state that combine in-situ profiles, satellite observations, and ocean general circulation models through data assimilation techniques. Major products include ECMWF's ORAS5, NCEP's GODAS, and the Japan Meteorological Agency's MOVE system.

Steric sea level change is the portion of sea level rise caused by thermal expansion and salinity changes rather than added water mass. Isolating the steric signal from altimetry requires independent gravity data, typically from the GRACE and GRACE-FO missions.

Head-to-Head Comparison

Accuracy nuances. Argo provides the most direct measurement of subsurface temperature but suffers from sampling gaps in the Southern Ocean, under sea ice, in marginal seas, and below 2,000 meters. A 2025 intercomparison by von Schuckmann et al. found that when Argo is combined with Deep Argo floats (which have grown from 130 to over 400 active units between 2022 and 2026), full-depth OHC uncertainty drops to ±0.08 W/m². Satellite altimetry offers truly global coverage including polar regions via CryoSat-2 and ICESat-2, but translating SSH anomalies into heat content requires subtracting the mass component using GRACE-FO gravity data, a step that introduces its own ±0.15 W/m² uncertainty (Barnoud et al., 2024). Reanalysis products fill spatial and temporal gaps but may drift from reality in data-sparse regions, and different products can disagree on decadal OHC trends by 20 to 40 percent in the deep ocean (Meyssignac et al., 2025).

Cost Analysis

Argo floats. Each standard Core Argo float costs approximately $20,000 to $25,000 to manufacture and deploy, with a design life of four to five years. Deep Argo floats, which profile to 6,000 meters, run $45,000 to $60,000 per unit. Sustaining the current 4,000-float global array requires about 800 replacement floats per year, translating to roughly $18 to $22 million annually in hardware alone (Euro-Argo ERIC, 2025). When telecommunications, data management, and shore-based quality control are included, the total international Argo budget is approximately $35 million per year, shared across more than 30 contributing nations. The Deep Argo extension adds an estimated $10 to $15 million annually as the network scales toward a 1,200-float target.

Satellite altimetry. Mission costs are substantially higher at the per-mission level but amortize across many user communities. The Copernicus Sentinel-6 Michael Freilich satellite cost approximately €830 million ($900 million) through development, launch, and five years of operations (ESA, 2024). Its successor, Sentinel-6B, launched in late 2025, carries a comparable budget. However, altimetry data serve navigation, weather forecasting, fisheries, and defense in addition to climate science. Attributing costs purely to ocean heat measurement is therefore difficult, but a reasonable allocation for climate-related ocean monitoring is roughly $30 to $50 million per year across the international altimetry constellation.

Reanalysis products. Reanalysis systems ride on top of operational weather and ocean forecasting infrastructure. ECMWF's annual operating budget exceeds €400 million, but reanalysis development and production represent a small fraction, estimated at €10 to €15 million per year for ocean reanalyses specifically (ECMWF, 2025). National centers such as NCEP and JMA bear similar marginal costs. Because reanalysis products are free to download, the cost to end-users is limited to computational resources for data extraction and analysis.

For organizations building climate risk platforms, the practical cost distinction is not data acquisition (all three are free or low-cost to access) but the analytical overhead of working with each format: processing millions of Argo profiles, handling large satellite swath files, or navigating reanalysis grid conventions.

Use Cases and Best Fit

National climate assessments and IPCC reporting. Argo remains the gold standard for quantifying upper-ocean warming trends used in IPCC reports. Its direct, calibrated measurements anchor the observational record. NOAA's National Centers for Environmental Information (NCEI) relies on Argo-derived OHC estimates as primary input for the annual State of the Climate report (Cheng et al., 2024).

Operational oceanography and seasonal forecasting. Real-time Argo data feed into ocean forecasting systems (e.g., Mercator Ocean's global 1/12° model), but satellite altimetry provides the dominant constraint on mesoscale variability. Agencies like the UK Met Office and ECMWF assimilate both into coupled prediction systems, with altimetry correcting surface dynamics and Argo constraining subsurface thermal structure.

Coastal and regional climate risk. Satellite altimetry excels for regional sea level monitoring because it captures spatial patterns that sparse float arrays miss. Port authorities, reinsurance companies, and coastal planners increasingly use gridded altimetry products from the Copernicus Marine Service to track local sea level trends with sub-centimeter accuracy.

Deep-ocean and full-depth energy budgets. For closing Earth's energy budget, combining Argo with Deep Argo floats and GRACE-FO gravity data is the emerging best practice. A 2025 study by Purkey et al. demonstrated that deep-ocean warming below 2,000 meters contributes approximately 10 to 15 percent of total OHC change, a signal largely invisible to standard Argo and altimetry alone.

Climate model validation. Reanalysis products are indispensable for evaluating climate model performance because they provide spatially complete, temporally continuous fields that modelers can compare grid cell by grid cell. The CMIP7 protocol explicitly recommends benchmarking against ORAS5 and EN4 reanalyses.

Decision Framework

1. Define the question. If you need a direct, bias-minimized estimate of ocean warming for regulatory reporting or scientific publication, prioritize Argo-based OHC products. If you need spatially resolved sea level trends for infrastructure planning, start with satellite altimetry. If you need gap-filled, model-consistent fields for climate model validation, use a reanalysis product.

2. Assess spatial and temporal requirements. Argo's nominal 3-degree spacing limits its utility for mesoscale features. Altimetry resolves scales down to 50 to 100 km but only at the surface. Reanalyses offer the finest effective resolution but inherit model biases below the mixed layer.

3. Evaluate depth requirements. For analyses restricted to the upper 2,000 meters, standard Argo is sufficient. For full-depth budgets, combine Argo with Deep Argo and gravimetry. Altimetry alone cannot provide depth-resolved information.

4. Consider latency needs. Real-time applications (e.g., marine heatwave alerts) benefit from Argo's 12 to 24 hour data delivery and altimetry's 3 to 5 day near-real-time products. Reanalysis products lag by weeks to months.

5. Plan for complementarity. The most robust assessments blend all three approaches. Von Schuckmann et al. (2025) recommend using Argo for trend anchoring, altimetry for spatial context, and reanalysis for gap-filling, with ensemble spread across products as a measure of structural uncertainty.

Key Players

Established Leaders

• Argo Program (international consortium) — Coordinates 4,000+ floats across 30+ nations; managed operationally through JCOMMOPS and national programs including NOAA (US), Ifremer (France), CSIRO (Australia), and JAMSTEC (Japan).
• EUMETSAT / Copernicus Marine Service — Operates the Sentinel-6 satellite altimetry missions and distributes processed sea level products to over 50,000 registered users globally.
• ECMWF — Produces the ORAS5 and ERA5 reanalysis products used by hundreds of climate research groups and government agencies worldwide.
• NOAA NCEI — Maintains the World Ocean Database and produces global OHC estimates cited in IPCC assessments and annual climate reports.

Emerging Startups

• Sofar Ocean — Deploys a network of over 1,500 Spotter buoys providing real-time ocean surface data; expanding into subsurface temperature profiling for commercial clients.
• Saildrone — Operates autonomous surface vehicles equipped with oceanographic sensors for targeted surveys in data-sparse regions including the Arctic and Southern Ocean.
• Terradepth — Building autonomous underwater vehicles (AUVs) for deep-ocean data collection with AI-driven mission planning.

Key Investors/Funders

• World Meteorological Organization (WMO) — Coordinates the Global Ocean Observing System (GOOS) framework that sustains Argo and satellite missions.
• European Commission — Funds Copernicus satellite missions and the Euro-Argo ERIC infrastructure through Horizon Europe, committing over €1 billion to ocean observation through 2027.
• G7 Ocean Deal (2025) — Pledged $700 million in new funding for sustained ocean observation, including Deep Argo expansion and next-generation altimetry missions.

FAQ

Which approach is most accurate for measuring global ocean warming trends? Argo floats provide the most direct and least model-dependent measurement of upper-ocean heat content. When combined with Deep Argo floats for the abyssal ocean, the network achieves trend uncertainties as low as ±0.08 W/m² for full-depth estimates (von Schuckmann et al., 2025). Satellite altimetry and reanalysis products are useful complements but introduce larger structural uncertainties due to signal separation challenges and model biases, respectively.

Can satellite altimetry replace Argo floats for ocean heat monitoring? Not on its own. Altimetry measures sea surface height, which reflects thermal expansion but also mass changes from ice melt and redistribution by ocean currents. Separating these contributions requires independent data, typically from GRACE-FO and Argo itself. Altimetry excels at providing spatial coverage and resolving mesoscale variability but cannot supply depth-resolved temperature profiles.

How is Deep Argo changing the picture? Deep Argo floats profile to 4,000 or 6,000 meters, capturing warming signals in the abyssal ocean that standard Argo misses. The network has grown from roughly 130 floats in 2022 to over 400 in early 2026, with a target of 1,200 floats for global coverage. Early results show that deep-ocean warming contributes 10 to 15 percent of total OHC change, a significant correction to budgets based on upper-ocean data alone (Purkey et al., 2025).

What are the main limitations of reanalysis products for ocean heat content? Reanalysis products depend on the underlying ocean model's physics, resolution, and bias correction schemes. In regions with sparse observational data, such as the deep Southern Ocean or under Arctic ice, reanalyses are essentially model forecasts with limited observational constraint. Different products (ORAS5, GODAS, MOVE) can disagree on decadal OHC trends by 20 to 40 percent in data-poor areas (Meyssignac et al., 2025). Users should employ multi-product ensembles rather than relying on a single reanalysis.

How should organizations combine these approaches in practice? Best practice is a tiered strategy: use Argo-based OHC products as the primary trend reference, layer in satellite altimetry for spatial context and coastal applications, and draw on reanalysis fields when complete gridded coverage is required. Cross-checking results across all three approaches, and using their spread as an uncertainty estimate, provides the most robust basis for climate risk assessments and regulatory reporting.

Sources

• Cheng, L., Abraham, J., Trenberth, K., et al. (2024). "Record-Setting Ocean Warmth Continued in 2023." Advances in Atmospheric Sciences, 41(5), 809-820.
• WMO. (2025). State of the Global Climate 2024. World Meteorological Organization, Geneva.
• Loeb, N.G., Johnson, G.C., Thorsen, T.J., et al. (2025). "Satellite and Ocean Data Reveal Marked Increase in Earth's Heating Rate." Geophysical Research Letters, 52(1).
• Argo Program. (2025). Argo Annual Report 2025: Status and Plans. JCOMMOPS / Argo Steering Team.
• Barnoud, A., Pfeffer, J., Cazenave, A., et al. (2024). "Revisiting the Steric Sea Level Budget with Improved Satellite Gravimetry." Journal of Geophysical Research: Oceans, 129(3).
• von Schuckmann, K., Minière, A., Gues, F., et al. (2025). "Earth's Energy Imbalance from Ocean Heat Content Changes." Earth System Science Data, 17(1), 217-246.
• Purkey, S.G., Johnson, G.C., Talley, L.D., et al. (2025). "Deep Argo Reveals Continued Abyssal Warming." Nature Climate Change, 15(2), 134-140.
• Meyssignac, B., Boyer, T., Ishii, M., et al. (2025). "Multi-Product Intercomparison of Ocean Heat Content Estimates." Journal of Climate, 38(4), 1521-1545.
• Euro-Argo ERIC. (2025). Euro-Argo Infrastructure Enhancement Strategy 2025-2030. Euro-Argo ERIC, Brest, France.
• ESA. (2024). Sentinel-6 Mission Summary and Budget Overview. European Space Agency, Paris.
• Johnson, G.C., Lyman, J.M., Boyer, T., et al. (2025). "Recent Western Pacific Ocean Heat Content Changes." Geophysical Research Letters, 52(5).
• ECMWF. (2025). ORAS5 Ocean Reanalysis System: Technical Documentation and Validation. ECMWF Technical Memorandum, Reading, UK.