The team set out to evaluate two global ocean reanalysis products distributed through the Copernicus Marine Service, both built on the NEMO ocean model and constrained by data assimilation. The first, GLORYS2V4, runs at a horizontal resolution of one quarter of a degree, roughly 25 kilometers at mid-latitudes, which makes it eddy-permitting, meaning it can barely represent the larger eddies but smears out the smaller ones. The second, GLORYS12V1, developed by Mercator Ocean International, runs at one twelfth of a degree, about 8 kilometers, with 50 vertical levels, placing it firmly in the eddy-resolving regime. Both systems assimilate along-track satellite sea-level anomalies, sea-surface temperature, in-situ temperature and salinity profiles, and sea-ice concentration over the satellite altimetry era that began in 1993, producing a continuous, physically consistent four-dimensional picture of the ocean state.

Reanalyses are indispensable to ocean and climate science precisely because they blend sparse observations with model dynamics, filling gaps in space and time. But that blending raises an obvious question: how faithfully do they reproduce the real, turbulent ocean? To answer it, the researchers needed an independent benchmark, and they chose two satellite altimetry products, both mapped onto a one eighth of a degree grid. The first was the established AVISO DUACS product, derived from decades of multi-mission nadir altimeters that measure sea level along narrow ground tracks. The second was the experimental SWOT MIOST Science product, which incorporates observations from the Surface Water and Ocean Topography mission, launched in December 2022, whose wide-swath radar interferometer captures sea-surface height over broad strips of ocean at far finer spatial scales than conventional altimetry ever could.

The choice of reference data was itself a methodological statement. Previous work had shown that the effective resolution of mapped altimetry products is lower than their nominal grid spacing, because the optimal interpolation used to fill in the gaps between satellite tracks smooths away variability at wavelengths shorter than roughly 50 to 100 kilometers. Comparing an eddy-resolving model against a product that cannot see the smallest eddies would unfairly penalize the model. The SWOT-based product, while still a mapped Level-4 reconstruction rather than raw swath data, injects genuinely new fine-scale information that has not been directly assimilated into the reanalyses. The authors are careful to note that neither satellite product can be treated as absolute truth, since both share nadir-altimetry information with the reanalysis systems, so they used the two references as complementary yardsticks rather than oracles, comparing results over a common period from August 2023 to May 2025 across a North Atlantic domain spanning roughly 30 to 60 degrees north, a region dominated by the Gulf Stream and its famously energetic eddy field.

The verification machinery was built on the open-source py-eddy-tracker package, which identifies eddies as closed contours in sea-surface height fields that satisfy strict geometric criteria: a single extremum at the center, a roughly circular shape, and a minimum amplitude. Detection was performed on absolute dynamic topography rather than sea-level anomaly to avoid mistaking Gulf Stream meanders for eddies in this strongly sheared region. Each eddy was then tracked day by day using a cost function that weighs differences in amplitude, effective radius, and centroid position, with a minimum lifetime of four days to filter out transient noise. The crucial step came next: rather than merely counting eddies in each dataset independently, the team matched eddies between the reanalyses and the satellite references, classifying each detection as a hit, a miss, or a false alarm, and condensing those counts into two familiar verification scores borrowed from weather forecasting, the Probability of Detection and the False Alarm Ratio.

The headline result is unambiguous. GLORYS12V1 detected roughly 62 percent of the eddies observed by either satellite product, while GLORYS2V4 managed only about 46 percent, a relative improvement of more than 30 percent in the Probability of Detection. The price of that sensitivity was a higher False Alarm Ratio, rising from roughly 18 percent for the coarser model to about 24 percent for the eddy-resolving one, an increase of 22 to 26 percent depending on the reference dataset. That trade-off is not a flaw so much as a signature of resolution: the high-resolution model sees many more small eddies, and while many of them correspond to real features, a substantial fraction find no counterpart in the satellite maps, which struggle to resolve features below about 50 kilometers. A block-bootstrap statistical analysis with 95 percent confidence intervals confirmed that the detection advantage of GLORYS12V1 is robust and insensitive to the details of the resampling.

The study went beyond simply asking whether eddies were found, and examined how accurately the matched eddies were rendered. For eddies detected in both the model and the satellite data, GLORYS12V1 achieved a mean matching cost about 26 percent lower than GLORYS2V4 against both references. Its mean errors in amplitude, effective radius, and centroid distance were reduced by approximately 35, 32, and 13 percent respectively. In other words, when the eddy-resolving model gets an eddy right, it gets the eddy’s strength, size, and position substantially closer to what the satellites observe. Notably, the improvement was slightly more pronounced in the comparison against the SWOT-based product, hinting that wide-swath altimetry reveals fine-scale structure that only the high-resolution model can begin to reproduce.

The diagnostics also exposed where both systems still stumble. The overwhelming majority of misses and false alarms were concentrated among small eddies, those with radii below 50 kilometers and amplitudes below 4 to 5 centimeters. For large, strong eddies the agreement was striking: for features with radii above 100 kilometers, the Probability of Detection climbed to around 90 percent and the False Alarm Ratio fell to 10 to 15 percent in the DUACS comparison, while in the SWOT comparison the detection rate exceeded 80 to 95 percent for eddies with amplitudes greater than 10 centimeters. The authors caution that some of the smallest detections may be spurious artifacts of mapping, filtering, or noise rather than genuine dynamical features, and they argue that future verification studies should develop objective, scale-aware strategies for down-weighting these uncertain features rather than treating every closed contour as a physically meaningful eddy.

The broader implications extend well beyond model scorecards. Ocean reanalyses underpin climate monitoring, fisheries oceanography, marine safety, and seasonal prediction, and eddies are central to how the ocean stores and redistributes heat. Demonstrating quantitatively that eddy-resolving systems represent the mesoscale ocean markedly better gives the modeling community a concrete justification for the computational expense of high resolution, while the persistent false alarms among small features map out exactly where the next improvements are needed. Equally important, the study showcases SWOT wide-swath altimetry as a new verification tool of genuine power. Because its fine-scale observations are not assimilated into the reanalyses examined here, they provide a partially independent window on ocean variability that conventional nadir altimetry simply cannot offer. As the SWOT record lengthens and future work extends this framework to longer periods, other ocean basins, and more sophisticated eddy classifications that distinguish merged, split, and artefactual features, the combination of eddy-resolving models and wide-swath observations promises a far sharper view of the ocean’s restless, swirling interior.

Subject of Research: Evaluation of mesoscale eddy representation in North Atlantic ocean reanalyses using satellite altimetry

Article Title: Quantitative evaluation of mesoscale eddies in the North Atlantic using satellite altimetry and ocean reanalyses

Article References: Mauriello, P., Smith, G. C., Storto, A., & Yang, C. (2026). Quantitative evaluation of mesoscale eddies in the North Atlantic using satellite altimetry and ocean reanalyses. Ocean Science, 22(5), 2863-2902. https://doi.org/10.5194/os-22-2863-2026

Image Credits: AI Generated

DOI: 10.5194/os-22-2863-2026

Keywords: mesoscale eddies, ocean reanalysis, satellite altimetry, SWOT, North Atlantic, GLORYS12V1, GLORYS2V4, AVISO DUACS, py-eddy-tracker, data assimilation, ocean modeling, Gulf Stream

In a paper published in Ocean Science on 28 September 2026, Qifan Wu of Zhejiang University and colleagues compared three experimental Level-4 products that incorporate SWOT wide-swath data: MIOST, a multiscale statistical interpolation framework; 4DvarQG, a variational assimilation scheme constrained by a reduced-order quasi-geostrophic model; and 4DvarNET, a data-driven approach in which both the dynamical prior and the assimilation operator are learned by deep neural networks. All three draw on the same input observations processed through the DUACS system and distributed by AVISO, so differences among the products arise almost entirely from their mapping methodologies — a controlled comparison that isolates the effect of each reconstruction strategy.

The team focused on the North Atlantic between 25 and 50 degrees north and 80 and 10 degrees west, with particular attention to the Gulf Stream, one of the most energetic ocean current systems on Earth. Because all three products share this domain, the region served as a stringent testbed. The evaluation combined three complementary lenses: Eulerian mapping accuracy measured against surface drifters, Lagrangian trajectory prediction skill assessed by forecasting the paths of drifting buoys, and dynamical diagnostics built on the Rossby number and the finite-size Lyapunov exponent, which reveal how well each product captures vorticity and stirring structures across scales.

The drifter-based validation was substantial. A total of 176 drifting platforms deployed between 27 July 2023 and 20 April 2024 provided more than 690,000 velocity samples, which the authors partitioned into three dynamical regimes: a shelf–slope zone, an energetic quasi-geostrophic regime along the Gulf Stream, and a weakly nonlinear open-ocean regime that dominated the sampling. Because drifters measure the total upper-ocean current at 15 meters depth — including wind-driven, tidal, and inertial components — while the mapped products primarily represent geostrophic flow, the team relied on pairwise gain/loss ratios to remove this shared representativeness error and obtain a conservative estimate of relative performance.

The Eulerian results revealed a clear hierarchy in the Gulf Stream region. There, 4DvarQG consistently outperformed both competitors by approximately 4 to 5 percent in region-averaged velocity error, with improvements that were spatially coherent along the current’s core and its downstream extension. The authors attribute this to the quasi-geostrophic dynamical constraints, which are well matched to the mesoscale-dominated, strongly balanced flow of an energetic boundary current. In the weakly nonlinear open ocean, by contrast, 4DvarQG and MIOST performed comparably, while 4DvarNET showed a consistent degradation of about 1 to 2 percent relative to both, suggesting that its learned representations offer little advantage — and may even add noise — when the background flow is weak and the balanced component dominates the error budget.

The Lagrangian experiments told a complementary story. By initializing simulated trajectories from observed drifter positions and integrating them forward for lead times up to 20 days, the team quantified how velocity errors accumulate along predicted paths. For short lead times of roughly 0 to 4 days, 4DvarQG delivered the largest improvement over the DUACS benchmark, reflecting its superior reconstruction of short-timescale velocity variability. MIOST showed smaller gains, and 4DvarNET demonstrated no systematic improvement at any lead time. Beyond about 10 days, when mesoscale transport dominates, MIOST and DUACS outperformed the other products, with separation errors converging to within roughly one degree after 15 days — evidence of their robust representation of the large-scale circulation.

The dynamical diagnostics exposed the most intriguing differences. 4DvarNET exhibited systematically sharper sea surface height gradients, larger Rossby number magnitudes, and more filamentary strain structures in the finite-size Lyapunov exponent fields. In the weakly nonlinear open-ocean regime, its Rossby number enhancement exceeded 60 percent relative to the other products. Yet the authors urge caution: this amplification occurred precisely where the background flow is too weak to sustain intense submesoscale activity, raising the possibility that weak gradients or observational noise are being inflated into high-vorticity signals. The neural network’s training data, drawn from the NATL60 simulation, include no explicit tidal forcing and only partial representation of inertia-gravity waves, so unresolved wave-like variance may be misattributed to vortical structures. With available validation data, the physical realism of these sharpened features simply cannot be confirmed.

Three case studies of individual eddies crystallized the scale-dependent trade-off. For a roughly 100-kilometer mesoscale eddy near the Gulf Stream core, all three products reconstructed the vortex geometry reliably, differing mainly in amplitude fidelity — 4DvarQG smoothed the core, 4DvarNET preserved gradients but distorted the quasi-circular morphology, and MIOST produced the smoothest fields. For an intermediate 50-kilometer eddy embedded in the fast Gulf Stream core, 4DvarQG most accurately reproduced the observed elliptical geometry, though all products underestimated the frontal steepness across the eddy core. The starkest contrast emerged for two roughly 10-kilometer submesoscale vortices south of the Gulf Stream, where only 4DvarNET recovered closed sea surface height anomaly contours matching the SWOT Level-3 observations, with eddy-core position errors of about 0.05 degrees, while MIOST and 4DvarQG failed to resolve the coherent vortex structures at all.

For users, the practical guidance is straightforward. MIOST, the only global-coverage product of the three, offers stable and consistent performance across all regimes and is the recommended choice for researchers who need reliable, well-characterized fields without regime-specific biases. Physical oceanographers studying mesoscale circulation, eddy tracking, or trajectory prediction in energetic midlatitude regions will benefit most from 4DvarQG, while those hunting fine-scale fronts, filaments, or submesoscale eddies may find 4DvarNET more informative — provided its amplified small-scale signals are interpreted with caution. Looking ahead, the authors argue that future mapping systems may best exploit SWOT through hybrid approaches that marry the flexibility of data-driven learning with explicit physical constraints, a balance the satellite altimetry community is only beginning to strike.

Subject of Research: Intercomparison of SWOT-derived Level-4 sea surface height mapping products over the North Atlantic

Article Title: Intercomparison of three SWOT-derived Level-4 products: from mapping accuracy to multi-scale dynamical representation

Article References: Wu, Q., Zhou, C., Liu, B., Wu, W., Li, J., & Yang, J. (2026). Intercomparison of three SWOT-derived Level-4 products: from mapping accuracy to multi-scale dynamical representation. Ocean Science, 22(5), 2939-2956. https://doi.org/10.5194/os-22-2939-2026

Image Credits: AI Generated

DOI: 10.5194/os-22-2939-2026

Keywords: SWOT, satellite altimetry, sea surface height, ocean currents, submesoscale, mesoscale eddies, Gulf Stream, data assimilation, deep learning, quasi-geostrophic dynamics, Lagrangian drifters, North Atlantic

The team’s approach was to stop treating the North Atlantic as a single homogeneous basin and instead examine its internal geography. The mid-latitude North Atlantic is dominated by what oceanographers call the inter-gyre region, the frontal zone where the wind-driven subtropical gyre to the south meets the subpolar gyre to the north. Today, the line of zero wind stress curl, which marks the confluence of these two great rotating systems, tilts diagonally from southwest to northeast across the basin. This tilt coincides with the point where the Gulf Stream detaches from the North American coast and feeds into the North Atlantic Current, producing a band of roughly 10 degree Celsius water that stretches from Cape Hatteras toward Iceland. Cold, fresh subpolar water lies to the northwest of this boundary; warm, saline subtropical water lies to the southeast.

To reconstruct how this frontal system behaved through the Holocene, the researchers analyzed two deep-sea sediment cores recovered by the International Ocean Discovery Program, ODP Sites U1304 and U1308, positioned along the modern 10 degree isotherm. From these cores they extracted alkenones, organic molecules produced by marine algae whose chemical structure encodes the temperature of the water in which the algae grew. The team also counted planktonic foraminifera, single-celled organisms whose species composition tracks surface and subsurface water masses with remarkable fidelity. Age control came from radiocarbon dating of the foraminifer species Globigerina bulloides and from a volcanic ash layer associated with the Younger Dryas cold episode about 12,100 years ago.

Because alkenone-producing algae bloom seasonally, the researchers first quantified a potential seasonal bias. Using satellite-based estimates of monthly primary productivity combined with modern sea surface temperature fields, they showed that in the mid- and high-latitude North Atlantic, alkenone temperatures reflect warm-season conditions rather than annual means. This matters because skeptics have long argued that the apparent Holocene cooling is merely an artifact of declining summer insolation being recorded by summer-biased proxies. The new analysis acknowledges the warm bias but demonstrates that it cannot explain the spatial pattern the team uncovered.

That pattern is the study’s central revelation. Comparing sea surface temperatures between two time slices, 6,000 to 8,000 years ago and the last 2,000 years, the researchers found that late Holocene cooling was strongly zonally asymmetric. The western basin cooled dramatically, with a maximum drop of nearly 5 degrees Celsius at one sediment core, while cooling in the eastern basin, where subtropical waters dominate, was muted. In other words, the east-west temperature gradient across the mid-latitude North Atlantic intensified over the past 6,000 years. A simple, uniform response to declining local summer insolation would have cooled the basin evenly; it cannot produce this kind of asymmetry.

The explanation, the authors argue, lies in a basin-scale reorganization of the inter-gyre circulation. During the early and mid-Holocene, multiple proxy lines of evidence suggest the frontal system sat farther north and carried more subtropical heat toward high latitudes. Tropical and subtropical foraminifera species, including Globigerinoides ruber, appeared at the 53 degree north site U1304 between 8,000 and 10,000 years ago, and the Gulf Stream thermocline species Globorotalia inflata was abundant. Thermophilous mollusks reached Svalbard by 9,500 years ago, and diatom records show Atlantic water dominating the southern Nordic Seas. Meanwhile, neodymium isotopes measured in deep-sea corals, subsurface temperature and salinity estimates, and alkenone-based productivity proxies all point to a vigorously ventilating subpolar gyre that peaked between roughly 6,000 and 8,000 years ago.

Then, around 6,000 years ago, the system flipped. Sea surface temperatures at U1304 cooled in a stepwise fashion that strikingly mirrors the abrupt termination of the African Humid Period recorded off West Africa. Alkenone productivity declined, subsurface waters south of Iceland warmed and became saltier, and neodymium isotopes shifted toward subtropical values, all indicating a weakening of subpolar gyre ventilation and a reduced supply of cold, fresh subpolar mode water to the eastern basin. Crucially, the subtropical circulation did not compensate. Gulf Stream indicator species remained scarce, warm Atlantic water retreated from the Nordic Seas, and ice-rafted debris increased along Greenland, signaling expanded sea ice and colder conditions. The researchers link this coordinated shift to a weakening of the North Atlantic subtropical anticyclone as summer heating over northern continents declined with Earth’s orbital precession, allowing the inter-gyre front and the North Atlantic Current to shift equatorward.

When the team compared these reconstructions with the TraCE-21K transient climate model simulation, one of the most widely used tools for studying climate since the last ice age, the mismatch was stark. The model simulates basin-wide warming over the past 6,000 years, in flat contradiction to the proxy data, and shows no significant change in either the subpolar or subtropical temperature gradients. Even when the researchers examined summer-only temperatures to account for the seasonal bias of alkenones, the model still produced mid-latitude warming. A modified version of the simulation without Holocene freshwater forcing fared somewhat better in the subpolar gyre but still failed to reproduce the zonally asymmetric cooling. Digging deeper, the team found the model’s zero wind stress curl line sits nearly zonal at about 40 degrees north and barely moves through the entire Holocene, missing the characteristic southwest-northeast tilt of the real ocean and any temporal variability in the frontal system.

The authors propose two tentative explanations for this failure. First, the model cannot capture the substantial late Holocene cooling and freshening of the subpolar gyre driven by increased Arctic freshwater export through the East Greenland Current and intensified westerly winds, changes well documented in marine sediments after about 4,000 to 5,000 years ago. Second, and perhaps more fundamentally, the coarse horizontal resolution of long transient simulations is known to inadequately represent western boundary currents like the Gulf Stream and their associated eddies, which are essential to the geometry and variability of the inter-gyre circulation. Without that dynamic machinery, the model’s North Atlantic front is essentially frozen in place.

The implications reach well beyond academic bookkeeping. The Holocene conundrum bears directly on how scientists weigh the climate sensitivity of the Earth system to carbon dioxide versus orbital forcing, and on how much confidence to place in paleoclimate reconstructions themselves. By showing that the North Atlantic cooling was not a uniform seasonal artifact but a dynamically driven, spatially patterned reorganization of ocean circulation, the study reframes the debate: the proxies may be telling the truth, and the models may simply be missing the ocean’s role. If wind-driven gyre dynamics can reshape regional temperatures by several degrees over a few millennia without any change in greenhouse gases, understanding those dynamics becomes all the more urgent as rising CO2 continues to push the system in new directions.

Subject of Research: Late Holocene sea surface temperature cooling and inter-gyre circulation reorganization in the mid-latitude North Atlantic

Article Title: Middle to late Holocene cooling and increased zonal asymmetry in the mid-latitude North Atlantic

Article References: Si, W., Herbert, T., & Toggweiler, J. R. (2026). Middle to late Holocene cooling and increased zonal asymmetry in the mid-latitude North Atlantic. Climate of the Past, 22(9), 1729-1740. https://doi.org/10.5194/cp-22-1729-2026

Image Credits: AI Generated

DOI: 10.5194/cp-22-1729-2026

Keywords: Holocene, North Atlantic, sea surface temperature, alkenones, inter-gyre circulation, subpolar gyre, Gulf Stream, paleoclimate, climate models, Holocene conundrum, foraminifera, ocean circulation

Researchers led by Raquel Toste of the Federal University of Rio de Janeiro analyzed monthly output from 24 climate models participating in the Coupled Model Intercomparison Project Phase 6, or CMIP6, under the SSP2-4.5 scenario. This pathway represents an update of the RCP4.5 scenario used in earlier IPCC assessments, and corresponds to a world in which a two-degree warming relative to the start of the century is extremely likely. To give every model equal weight in the ensemble, the team retained a single realization from each model, a standard practice that avoids over-representing institutions that contributed large ensembles. The variables examined included surface winds, wind stress, sea-level pressure, sea surface height, and the horizontal velocity of seawater at every depth.

The physical heart of the analysis lies in volume transport, a measure of how much water a current carries through a fixed cross-section of the ocean, expressed in Sverdrups, where one Sverdrup equals one million cubic meters per second. Rather than relying on surface speed alone, the researchers computed transport by numerically integrating the velocity field cell by cell across carefully defined staircase sections on each model’s native grid, capturing the full vertical structure of every current. Sections were placed across the Kuroshio and Oyashio Currents in the North Pacific, the Alaska and California Currents in the eastern North Pacific, the East Australian and Humboldt Currents in the South Pacific, the Gulf Stream in the North Atlantic, the Brazil, North Brazil, and Benguela Currents in the South Atlantic, and the Agulhas Current in the Indian Ocean.

The headline finding is a stark hemispheric asymmetry. Northern Hemisphere western boundary currents exhibit an overall reduction in volume transport by 2090-2100. The Gulf Stream is projected to weaken by 15.89 percent, a loss of 7.36 Sverdrups, with all 24 models agreeing on the sign of the change. The Agulhas Current is projected to decline by 10.40 percent, or 7.03 Sverdrups, also with unanimous model agreement. In the equatorial Atlantic, the North Brazil Current loses 10.42 percent of its transport, about 3.08 Sverdrups. In sharp contrast, the Southern Hemisphere is projected to experience enhancements in the volume transport of the Brazil Current, which strengthens by 26.56 percent, and the East Australian Current, which gains 3.27 percent. In the subpolar North Pacific, the Oyashio intensifies by 13.49 percent, driven by the strengthening and shifting of the Aleutian Low, which appears to spin up the subpolar gyre.

The study goes beyond transport to examine where currents sit on the globe. The northern branches of the subtropical gyres generally shift northward, a movement consistent with the poleward expansion of the Hadley circulation that both CMIP5 and CMIP6 models project under rising greenhouse gases. The most dramatic displacement belongs to the North Atlantic Current Drift, which on average is expected to sit 2.54 degrees north of its present latitude by the century’s end. The Atlantic South Equatorial Current is the lone exception, drifting slightly southward. These migrations matter because the position of gyre branches partly determines how much water western boundary currents can collect and carry.

Crucially, the research tested whether these changes can be understood within a single dynamical framework, and the answer is a resounding no. Using multiple linear regression and variance decomposition across the model ensemble, the team found that future transport changes are associated with fundamentally different mechanisms in different basins. In the North Atlantic, Gulf Stream transport is jointly linked to changes in the North Atlantic Current Drift position and to wind stress curl anomalies, with the combined regression explaining 44.5 percent of the inter-model variance. Wind stress curl, the rotation of the wind field that drives the interior flow of subtropical gyres through Sverdrup dynamics, independently accounts for roughly a quarter of the explained spread, while gyre migration explains about 17 percent. This joint response is broadly consistent with the projected weakening of the Atlantic Meridional Overturning Circulation, which modulates atmospheric temperature gradients and generates surface wind anomalies that reshape the wind-driven circulation.

Elsewhere the picture inverts. In the South Pacific, changes in the East Australian Current are mostly related to the migration of the South Pacific Equatorial Current, which emerges as the dominant predictor in the regression and explains part of the spread among model projections. In the Indian Ocean, the Agulhas slowdown is primarily associated with wind stress curl anomalies rather than the latitude of the Indian South Equatorial Current, suggesting that large-scale wind forcing exerts stronger control over this system than gyre repositioning. In the South Atlantic and North Pacific, by contrast, the Brazil and Kuroshio Currents showed no significant relationships with either of the evaluated mechanisms, indicating that additional regional dynamical processes dominate their long-term evolution at the studied sections. The Kuroshio is particularly revealing: its regression explains a mere 2.2 percent of the inter-model variance, pointing to complex, topographically constrained behavior in which regional thermal gradients and localized atmospheric forcing compete with the large-scale poleward shift of the westerlies.

One of the most striking technical discoveries concerns the vertical structure of these changes. When the team analyzed transport restricted to the upper 200 meters alongside full-depth transport, they found decoupled behavior in several systems. The Kuroshio’s total transport declines slightly by 1.37 percent, yet its upper-200-meter transport intensifies significantly by 1.08 Sverdrups, with positive velocity anomalies confined to the upper layers and negative anomalies below. This vertical shear enhancement aligns with the mechanism proposed by earlier work showing that greenhouse-gas-induced surface warming amplifies upper-ocean density stratification, trapping momentum in the uppermost layers and altering vertical velocity profiles. The Brazil Current shows a similar surface concentration, with about 72 percent of its transport increase packed into the upper 200 meters, reflecting the shallow dynamic nature of this flow, which is vertically constrained by the opposing northward Intermediate Western Boundary Current directly beneath it. Deep-penetrating currents such as the Gulf Stream and Agulhas behave differently, with their deceleration extending below 1,400 meters, a signature that modifies how these currents interact with bottom topography and could trigger path instabilities.

The broader implications are considerable. Ocean currents redistribute heat, carbon, and nutrients, so hemispheric shifts in their strength and position will reverberate through regional climates, marine ecosystems, and coastal economies on both sides of the equator. The finding that no single framework, whether classical Sverdrup dynamics or simple gyre reorganization, can universally explain future changes in western boundary current transport underscores the sophistication of the ocean’s response to warming. The authors point to atmospheric reorganization as the deeper driver: subtropical high-pressure systems in the Southern Hemisphere expand, the North Pacific Subtropical High contracts, and the North Atlantic Subtropical High migrates poleward, with sea-level pressure anomalies and wind stress curl changes reshaping the wind fields that ultimately spin the great gyres. Future studies combining overturning diagnostics, stratification changes, wind-driven circulation, and momentum budget analyses, the team suggests, would help quantify the relative importance of these mechanisms and further constrain the physical drivers of projected ocean circulation change. What is already clear is that the ocean’s engine room is being rewired, and that each basin is rewiring in its own way.

Subject of Research: Projected changes in ocean current volume transport and positioning under the CMIP6 SSP2-4.5 warming scenario

Article Title: Projections for ocean currents’ transport and positioning from CMIP6 models

Article References: Toste, R., Soares, A. E., de Freitas Assad, L. P., & Landau, L. (2026). Projections for ocean currents’ transport and positioning from CMIP6 models. Climate Dynamics, 64(10), Article 428. https://doi.org/10.1007/s00382-026-08385-9

Image Credits: AI Generated

DOI: 10.1007/s00382-026-08385-9

Keywords: ocean currents, CMIP6, climate change, ocean circulation, western boundary currents, Gulf Stream, Agulhas Current, Brazil Current, subtropical gyres, wind stress curl, volume transport, Sverdrup dynamics

The IWMO workshops are managed by an international steering committee of scientists drawn from ten different countries, with local volunteers from the hosting institute handling organization and support. A defining tradition of the series is the publication of a special issue, or Topical Collection, of peer-reviewed papers in the journal Ocean Dynamics after each meeting. More than 200 papers have now appeared in these collections, making the workshop series one of the most consistently productive venues for ocean modeling research. To promote international collaboration and diverse participation, organizers deliberately rotate the meeting across continents: IWMO-2023 was held in Hamburg, Germany, and IWMO-2024 took place in Sapporo, Japan. Since the organization’s inauguration in Taiwan in 2009, six meetings have been held in Asia, four in Europe, four in North America, one in South America and one in Australia. Only two previous meetings were held in the United States, in Virginia in 2010 and in Ann Arbor, Michigan in 2022. Following the Stanford meeting, IWMO-2026 was held in Palma de Mallorca, Spain, and IWMO-2027 is planned for June 2027 in Hangzhou, China.

The scientific program at Stanford featured 32 oral presentations, including four keynote invited presentations by prominent scientists, along with 10 posters. Sessions spanned the breadth of modern ocean modeling: internal and surface waves, coupled physical-biological interactions, global and basin-scale circulation, mixing and sub-mesoscale processes, coastal and estuarine modeling, and numerical methods. Continuing the IWMO tradition of encouraging early-career participation, 13 graduate students and postdoctoral researchers presented their work as part of the Outstanding Young Scientist Award competition, with the top winners hailing from the United States, Japan and Germany. The accompanying Topical Collection in Ocean Dynamics includes eight peer-reviewed papers from workshop participants, each subjected to rigorous review by IWMO members and external experts under the stewardship of a team of guest editors. Together the papers showcase a variety of numerical ocean and wave models, diverse analysis methods, and study sites ranging from the German Bight and the coasts of the United States and Canada to seas along the Asian Pacific coasts.

Among the collection’s contributions, Zhang and colleagues developed an idealized coupled Computational Fluid Dynamics and Discrete Element Method model, known as CFD-DEM, to study how cross-flow velocities affect particle transport near a subsea pipeline. The simulations identified distinct critical velocity thresholds for different particle sizes, offering a practical engineering tool for subsea pipeline design and backfilling projects, where the loss of backfill material is a persistent operational concern. The work illustrates how idealized numerical experiments can distill complex sediment-structure interactions into design guidance that engineers can apply directly to offshore infrastructure.

Wave dynamics featured prominently in the collection as well. Imamura and Yoshikawa tackled the long-standing challenge of parameterizing surface-wave breaking in wave-resolving simulations using a nonhydrostatic numerical model. Rather than relying on the empirical criteria that many existing models employ, they introduced a fourth-order diffusivity into the surface elevation equation to explicitly represent wave breaking. The approach was validated successfully against laboratory experiments and theory, demonstrating a physically grounded alternative to heuristic breaking schemes. In a related vein of coastal prediction, Ezer used the Princeton Ocean Model to explore how high-frequency oscillations in the Gulf Stream, with periods ranging from one week to two months, can induce coastal sea level variability. Comparisons with observations revealed a statistically significant anticorrelation between Gulf Stream strength and coastal sea level, but the relationship depends nonlinearly on location along the coast and on the frequency of the current’s variability, complicating the prediction of coastal flooding.

Nature-based coastal protection emerged as another theme. Jacob, Pein and Staneva used the unstructured-grid SCHISM hydrodynamic model, extended with sediment dynamics and vegetation, to evaluate seagrass as a nature-based solution for coastal protection in the German Wadden Sea. Their simulations, which incorporated end-of-the-century sea level rise projections, showed that despite rising seas, seagrass meadows retain their damping capacity and can reduce wave heights by as much as 30 percent in shallow areas. This finding carries significant implications for coastal management, suggesting that restored ecosystems could complement or even substitute for hard engineering defenses in some settings. In the Pacific Northwest, Khangaonkar and colleagues tested a sigma-coordinates formulation with shaved cells in the SCHISM model to simulate the Salish Sea, a fjord-like deep estuary. The refined grid and higher resolution compared with a past FVCOM configuration of the region eliminated the need for bathymetric smoothing and improved predictions of currents and temperatures, resolving near-shore intertidal and deep estuarine circulation simultaneously.

Aquaculture’s footprint on coastal hydrodynamics was examined by Fu and colleagues, who applied the FVCOM numerical model with an enhanced dual-drag parameterization to simulate the impact of large-scale floating-raft aquaculture structures in the open waters near Zhangzi Island in the northern Yellow Sea of China. The simulations showed that floating rafts can significantly reduce tidal current velocities, especially near the surface, with consequences for regional circulation and the transport of material. On the neighboring Pacific coast of North America, Lin and Dunphy deployed the NEMO model with AGRIF two-way nesting to resolve a fjord tidal jet in Quatsino Sound, British Columbia, Canada. The nested model reproduced tidal mixing in good agreement with observations and generated internal waves during each flood phase, with barotropic-to-baroclinic energy conversion for the M2 tide identified along the tidal-jet slope and in shallow waters.

Artificial intelligence also made its mark on the collection. Hu and colleagues proposed a geospatially encoded dual-channel network with attention mechanisms and physics constraints for reconstructing complex seafloor topography. Bathymetric prediction based on gravity anomalies remains the dominant approach for mapping the seafloor, but the new study integrates traditional physical models with neural networks to enhance the model’s fitting capacity, pointing toward a hybrid future in which machine learning and process-based oceanography reinforce one another rather than compete.

Taken together, the IWMO-2025 proceedings capture a field in vigorous transition, one in which classical process models such as the Princeton Ocean Model and FVCOM share the stage with next-generation frameworks like SCHISM and NEMO, and in which data-driven methods are beginning to reshape how the seafloor and the flows above it are mapped. The breadth of applications, from pipeline engineering and aquaculture siting to flood prediction and ecosystem-based coastal defense, underscores why ocean modeling has become indispensable infrastructure for climate adaptation worldwide. The organizers extended thanks to the local organizing committee at Stanford and to all participants for sustaining the collaborative spirit of the IWMO tradition, and the community now looks ahead to Palma de Mallorca and Hangzhou, where the next chapters of this sixteen-year-old scientific conversation will unfold.

Subject of Research: The 15th International Workshop on Modeling the Ocean (IWMO-2025) and its associated peer-reviewed papers on numerical ocean modeling

Article Title: The 15th international workshop on modeling the ocean (IWMO-2025) in Stanford, California, USA, June 30th – July 1st, 2025

Article References: Fringer, O., de Camargo, R., Chang, Y.-L., Xue, H., & Ezer, T. (2026). The 15th international workshop on modeling the ocean (IWMO-2025) in Stanford, California, USA, June 30th – July 1st, 2025. Ocean Dynamics, 76(9), Article 95. https://doi.org/10.1007/s10236-026-01853-0

Image Credits: AI Generated

DOI: 10.1007/s10236-026-01853-0

Keywords: IWMO-2025, ocean modeling, Ocean Dynamics, Stanford University, coastal sea level, seagrass coastal protection, SCHISM, NEMO, FVCOM, Gulf Stream, wave breaking parameterization, seafloor topography reconstruction