The work was carried out by C.W.E. de Korte, M. Verlaan, A. W. Heemink and B. Backeberg, affiliated with Delft University of Technology and the research institute Deltares. Their starting point was a persistent problem: third-generation wave models such as SWAN, the Simulating WAves Nearshore model used in the study, are highly reliable in the open ocean but continue to struggle in coastal shelf seas. Wave-current interactions, uncertain wind forcing, shallow-water effects and imperfect parametrizations of the physical source terms all introduce errors that are difficult to eliminate by calibration alone. Data assimilation offers a different route, blending real observations with the model’s own physics to nudge the simulated ocean state closer to reality without rewriting the underlying equations.

What sets this study apart from most earlier wave data assimilation efforts is the choice of state variable. Conventional operational schemes, such as Optimal Interpolation and three-dimensional variational methods, typically apply corrections only to significant wave height and then scale those corrections back onto the full wave spectrum using simplifying assumptions about how wave energy is distributed across frequencies and directions. The Dutch team instead placed the complete directional wave energy spectrum in the model state. Each ensemble member carried the spectrum across 32 frequency bands and 36 directional bins at every point of a 567-cell grid, producing a state vector of more than 650,000 elements per member. Because the ensemble evolves under the full SWAN physics, the corrections the filter produces are automatically physically consistent with the model, and integral parameters such as mean wave period adjust themselves without any ad hoc scaling.

The DEnKF itself is a deterministic variant of the classic Ensemble Kalman Filter. Rather than perturbing observations with random noise to propagate uncertainty, it updates the ensemble mean and the ensemble anomalies separately, avoiding the sampling errors that stochastic perturbations introduce. This is particularly valuable for small ensembles, and the team settled on 64 members after previous synthetic twin experiments showed the error statistics fully converged at that size. Uncertainty was injected into the system through the wind forcing, treated as the control variable, using a first-order autoregressive noise model with a spatial Gaussian correlation structure. Parameters were derived from the difference between HARMONIE wind analyses and forecasts, giving a standard deviation of two metres per second, a decorrelation timescale of fifteen hours and a spatial decorrelation length of 500 kilometres.

The observations came from seven nadir satellite altimeters: CFOSAT, Haiyang-2B, Cryosat-2, Jason-3, the two Sentinel-3 satellites, and Saral/AltiKa, all retrieved from the Copernicus Marine Environment Monitoring Service. Over the three-month window the satellites contributed 713 tracks over the North Sea, an average of about eight passes per day, with a mean interval of roughly three hours between passes but gaps stretching to nearly fifteen hours. Tracks were sub-sampled every 120 kilometres to avoid overloading individual grid cells, and a coastal mask excluded measurements within 50 kilometres of shore, where altimeter retrievals are known to be unreliable. Observation errors were assumed to be uncorrelated with a standard deviation of 0.2 metres. Hamill localisation experiments comparing the standard EnKF with the DEnKF across localisation radii of 100 to 500 kilometres showed the DEnKF with a 200-kilometre radius performed best, and that configuration became the final set-up.

Validation against the North Sea’s dense network of independent wave buoys delivered strikingly consistent results. Significant wave height errors dropped by a mean of 20.5 percent, from 0.39 metres in the free-running coarse model to 0.31 metres, a performance essentially matching the much finer SWAN-DCSM benchmark model run at roughly 3.6-kilometre resolution. Mean wave period improved at 21 of 23 stations with a ten percent reduction in root mean square error, while the peak period improved at 13 of 16 stations by about five percent. Wind speed showed modest improvements at some stations, though the researchers caution that the station anemometers, corrected to ten-metre equivalent heights assuming a neutral wind profile, carry their own uncertainties over the frequently non-neutral marine boundary layer. Not every parameter benefited: swell wave height degraded slightly on average, and the low-frequency inverse moment period and mean wave direction, each measured at only a handful of stations, also worsened marginally.

Spectral analysis explained the pattern. In unimodal sea states dominated by wind-driven waves, the assimilation corrected the entire wave spectrum in a way that closely matched buoy observations, as demonstrated during a storm peak on 20 January 2022 at the offshore station A121, where the analysis tracks the measured spectra hour by hour. But in mixed sea states where wind-sea and swell are clearly separated, the ensemble spread remained concentrated in the mid and high frequencies, because wind perturbations barely touch an independently propagating swell field and the altimeters measure only total significant wave height. Detailed examination of the largest swell errors revealed two distinct mechanisms: during short-fetch, rapidly rotating local wind conditions, the wind-based error covariances failed to represent the spatial scales at which swell actually varied between neighbouring stations, while a second error type, premature swell arrival at coastal stations, proved to be a systematic bias of the coarse-resolution model rather than a failure of the assimilation itself.

One of the most practically important findings concerns timing. The researchers binned all validation samples by the number of hours elapsed since the last satellite pass and found that prediction errors rose steadily with the length of the gap, particularly for stations in the open central North Sea. Coastal stations, whose errors are dominated by shallow-water processes rather than wind-driven corrections, were less sensitive to the satellite schedule. Because the orbital geometry of the contributing satellites fixes the timing of the gaps, these gaps recur with the tidal cycle, a phase-locking effect the authors flag as deserving further study. The message for forecasters is clear: the temporal density of observations matters, and merging additional data sources could deliver substantial gains.

The authors are candid about the limitations. The coarse 0.5-degree grid, chosen so that a 64-member ensemble of full spectra could be run at all, degrades accuracy near the coast, where resolution, missing triad interactions and a simplified setup all take their toll. Running a high-resolution model within the ensemble framework would require major advances in computing power and memory. Still, the path forward is mapped out: refining the wind noise model, assimilating additional integral wave parameters or even full spectra, incorporating continuous buoy measurements, and adding satellite SAR observations from missions such as Sentinel-1, SWOT and Sentinel-6. Beyond operational forecasting, the team points to wave reanalyses for risk and climate studies, and to the growing demand for high-quality training data for machine-learning wave models. For a shelf sea as busy and economically vital as the North Sea, better wave information from the satellites already overhead is a prize worth the computation.

Subject of Research: Ensemble-based assimilation of satellite altimeter wave measurements in a regional North Sea wave model

Article Title: Exploring the added value of assimilating satellite altimeter measurements in a North Sea wave model using the Deterministic Ensemble Kalman Filter

Article References: de Korte, C., Verlaan, M., Heemink, A. W., & Backeberg, B. (2026). Exploring the added value of assimilating satellite altimeter measurements in a North Sea wave model using the Deterministic Ensemble Kalman Filter. Ocean Dynamics, 76(10), Article 102. https://doi.org/10.1007/s10236-026-01858-9

Image Credits: AI Generated

DOI: 10.1007/s10236-026-01858-9

Keywords: wave data assimilation, Deterministic Ensemble Kalman Filter, satellite altimeter, SWAN wave model, North Sea, significant wave height, wave spectrum, ensemble forecasting, swell, coastal shelf seas, Ocean Dynamics, CMEMS

Led by Md. Khairul Amin and G. M. Jahid Hasan of the Department of Civil Engineering at the Military Institute of Science and Technology in Dhaka, the study tackles a notorious modeling gap. Previous efforts along the Bangladesh coast typically validated either water level or wave height in isolation, leaving current velocity largely untested. The new model, built on the widely used Delft3D suite, closes that gap by coupling a hydrodynamic component with the SWAN spectral wave model and confronting both against real observations at multiple coastal stations. The payoff is impressive: Pearson correlation coefficients ranging from 0.79 to 0.99 and Nash-Sutcliffe efficiency values of 0.72 to 0.99 across all tested parameters.

The challenge the model confronts is extraordinary. The Bay of Bengal’s distinctive geometry, a wide continental shelf in the west that narrows sharply toward the southeast, combined with the enormous freshwater discharge of the Ganges-Brahmaputra-Meghna river system, creates a uniquely complicated hydrodynamic environment. Bangladesh’s coastal zone covers roughly 47,201 square kilometers and is home to approximately 46 million people, much of it sitting only a few meters above mean sea level. Tides here are predominantly semi-diurnal and amplify toward the head of the bay, while monsoon winds, river discharge, and tropical cyclones all modulate water levels, wave activity, and current patterns in ways that defy simple prediction.

Technically, the model domain stretches from 18 degrees north to 23 degrees north and from 83 degrees east near Vishakhapatnam, India, to 95 degrees east near Gwa beach in Myanmar. The hydrodynamic component employs a curvilinear-orthogonal grid of roughly 27,000 cells with variable resolution, refined to 100-200 meters along the intricately sculpted coastline near the Meghna Estuary and coarsened to 10-15 kilometers near the open ocean boundaries. Bathymetry was drawn from the GEBCO 2020 dataset at 15 arc-second resolution, while tidal forcing at the open boundaries came from the TPXO 8.0 global inverse tide model using 13 principal harmonic constituents. Upstream river boundaries were driven by hourly discharge measurements from the Padma and Upper Meghna rivers.

Atmospheric forcing derives from the European Centre for Medium-Range Weather Forecasts’ ERA5 reanalysis, with hourly 10-meter wind fields and mean sea level pressure at 0.25-degree resolution. The SWAN wave component runs on a separate rectangular grid and incorporates third-generation physics, including depth-induced breaking via the Battjes-Janssen formulation, JONSWAP bottom friction, nonlinear triad interactions, and a phase-decoupled refraction-diffraction approximation. Crucially, the researchers employed online two-way coupling between the wave and flow models, allowing radiation stress gradients, enhanced bed shear stress, wave boundary layer streaming, turbulence mixing, and Stokes drift to be exchanged dynamically between components at every coupling interval, capturing the feedback between waves and currents that is essential during energetic monsoon conditions.

Calibration against observed tide levels and ERA5 wave heights during June 2020, a period spanning a full spring-neap tidal cycle under active monsoon conditions, produced tide-level correlations of 0.88 to 0.99 and Nash-Sutcliffe values of 0.91 to 0.99. Independent validation for October 2020 under contrasting post-monsoon conditions confirmed the model’s robustness, with tide-level correlations holding between 0.86 and 0.99. Perhaps most notably, simulated current velocities at Payra Port achieved a correlation coefficient and Nash-Sutcliffe efficiency of 0.80 each, reasonably capturing the temporal rhythm and dominant directions of the semi-diurnal tidal currents, though the authors note a strongly negative percentage bias that reflects the metric’s known sensitivity to reversing tidal flows rather than a genuine failure of the model.

A year-long simulation for 2020 then revealed the striking spatial texture of the region’s coastal hydrodynamics. At Sandwip, an island adjacent to the tide-dominated Meghna Estuary, water levels ranged from about minus 1.8 meters to nearly plus 2.0 meters relative to mean sea level, and depth-averaged currents commonly reached 1.0 to 1.3 meters per second. At Kuakata on the more open south-central coast, the tidal range was somewhat smaller and currents generally stayed below 0.5 meters per second during neap tides, peaking near 1.0 to 1.1 meters per second at spring tide. Wave heights at Sandwip generally remained below 1.0 meter with a late-August peak of roughly 1.7 meters, while Kuakata’s waves stayed between 0.3 and 0.7 meters most of the time, with mean wave periods of 6 to 11 seconds that are longer and more variable than Sandwip’s estuarine 4.5 to 6.5 seconds.

The model also reproduced recognizable large-scale patterns, including the persistent anticlockwise residual circulation around Sandwip Island documented in earlier studies and the pronounced tidal asymmetry near major estuarine systems. Snapshot comparisons showed flood-tide currents reaching 0.69 meters per second and ebb currents 0.38 meters per second during the monsoon neap-tide snapshot, compared with up to 1.11 and 0.90 meters per second respectively during the post-monsoon spring-tide snapshot, with the strongest currents consistently concentrated in shallow nearshore zones and estuarine channels where bathymetry and channel geometry amplify flow. Mean wave directions clustered predominantly between 160 and 210 degrees, a southerly to southwesterly approach consistent with the monsoon wind field and swell arriving from the southern Indian Ocean. The authors are careful to note that the two snapshots differ simultaneously in tidal phase and season, so they illustrate spatial structure rather than isolating a pure seasonal effect.

The study is candid about its limitations. GEBCO bathymetry cannot fully resolve the fine-scale estuarine channels and shoals of the Meghna Estuary, and ERA5’s relatively coarse resolution may smooth localized nearshore wind gradients and coastline-induced variability, meaning wave validation is effectively a comparison against reanalysis fields rather than fully independent observations. Current-velocity validation rests on a single station, and the model is accordingly best suited for regional-scale assessment of tide-wave-current variability rather than absolute sediment-flux estimates or site-specific engineering predictions. The researchers recommend future evaluation against satellite altimetry wave products and in situ buoy records, along with higher-resolution hydrographic surveys and nested local models for areas of specific engineering interest.

Even with those caveats, the significance for Bangladesh is hard to overstate. A physically consistent, statistically validated coupled model of the northern Bay of Bengal provides exactly the platform needed for storm surge forecasting, disaster preparedness, early warning systems, coastal infrastructure design, and long-term climate change impact assessment in a region where sea-level rise and intensifying cyclones threaten millions. By demonstrating that a single coupled framework can simultaneously simulate tides, waves, and currents with good-to-excellent skill, the Dhaka team has established a regional baseline onto which future studies of compound flooding, sediment transport, morphodynamics, and coastal protection interventions can build. For the 46 million people living along this vulnerable deltaic edge, better models mean better warnings, and better warnings mean lives saved.

The statistical metrics used to judge the model deserve a brief explanation for readers outside the field. Pearson’s correlation coefficient measures how closely simulated and observed values move together in time, while the Nash-Sutcliffe efficiency compares the model’s errors against the simple baseline of using the observed mean; a value of one indicates a perfect match, and values above roughly 0.7 are generally considered good for coastal hydrodynamic simulations. The fact that the coupled framework achieved efficiencies as high as 0.99 for tide levels, across both monsoon and post-monsoon conditions, indicates that the model captures not only the timing of the semi-diurnal tidal signal but also its amplitude through spring-neap cycles.

The choice of a full annual simulation spanning 2020 is itself methodologically significant. Many earlier regional studies focused on single cyclone events or short calibration windows, which can overstate skill because a model tuned to one tidal or wind regime may fail when conditions shift. By testing against the June monsoon period, when southwesterly winds and heavy Ganges-Brahmaputra-Meghna discharge dominate, and again in October as conditions relaxed, the researchers exposed the model to the two most contrasting hydroclimatic states the coast experiences.

The two-way coupling approach also reflects a broader trend in coastal oceanography. Earlier generations of surge models treated waves and water levels as separate problems, but studies of cyclone events in the Bay of Bengal have repeatedly shown that wave setup and radiation stress gradients can materially alter coastal water levels during storms. By exchanging radiation stresses, bed shear stress, and turbulence terms between the wave and flow components continuously, the new framework provides a physically consistent foundation for the compound-flooding and storm-surge applications that regional disaster planners increasingly demand.

Subject of Research: Coupled wave-hydrodynamic modeling of tides, waves, and currents in the northern Bay of Bengal

Article Title: Two-dimensional coupled wave-hydrodynamic modeling of the northern Bay of Bengal

Article References: Amin, M. K., & Hasan, G. M. J. (2026). Two-dimensional coupled wave-hydrodynamic modeling of the northern Bay of Bengal. Discover Oceans, 3(1), Article 54. https://doi.org/10.1007/s44289-026-00167-9

Image Credits: AI Generated

DOI: 10.1007/s44289-026-00167-9

Keywords: Bay of Bengal, Bangladesh coast, Delft3D, SWAN wave model, tidal hydrodynamics, storm surge, coastal modeling, Meghna Estuary, monsoon, ERA5 reanalysis, current velocity, climate resilience