Evaluation of parametric tropical cyclone wind models for representing storm surge and wave heights in the Queensland region

D-WaveNet: A physics-informed transformer with cross-scale interaction for Significant Wave Height forecasting

Beach and shoreface nourishments are widely used to mitigate erosion of sandy coasts. Since 1990, large stretches of the Dutch coast have been repeatedly nourished, with efforts expected to increase further. However, the cumulative effects of repeated sand nourishments on coastal evolution remain poorly understood. This study uses satellite imagery to quantify how repeated nourishments have influenced shoreline dynamics and long-term trends between 1985 and 2025 at two sites with contrasting wave climates: (1) Egmond-Bergen, with a bimodal directional wave climate and limited net alongshore transport and (2) Ameland, dominated by a single prevailing wave direction and stronger alongshore transport. The two sites showed distinct responses to the repeated nourishments. At Egmond-Bergen, shoreface nourishments produced a cumulative but localised effect on the shoreline, leading to a positive trend of almost 2 m/yr. In contrast, shoreface nourishments at Ameland barely affected the shoreline, presumably because of the high alongshore transport. Beach nourishments at both sites, when not combined with a shoreface nourishment, only affected the shoreline temporarily ( < 1 − 2 years). Seasonal shoreline cycles (amplitude ≈ 10 m or less) were out-of-phase with seasonality in the offshore wave height and did not differ between nourished and non-nourished sections. Our results highlight how especially long-term shoreline dynamics can differ between sites, despite comparable nourishment strategies. • Contrasting long-term shoreline response to repeated shoreface nourishments. • Cumulative shoreline progradation observed when alongshore sand transport is limited. • Beach nourishments result in short-lived shoreline progradation only. • Directional wave climate may affect shoreline response to repeated nourishments.

Hybrid AI model for skillful significant wave height forecasting with validation in the South China Sea

Multi-step significant wave height forecasting using a hybrid VMD–GRU–MATNet model

Monthly extreme wave height prediction based on an LSTM-Stacking model

Seagrass meadows have been proposed as a nature-based coastal protection measure to reduce incoming wave energy. Although numerous studies have demonstrated the capability of seagrass meadows to attenuate waves, their real-world effectiveness in providing coastal protection remains uncertain. The aim of this study was to quantify the influence of a Zostera marina meadow located in a non-tidal fetch-limited environment on three coastal protection metrics: wave runup at the shore, the storm-induced erosion of dunes, and the longshore sediment transport. Field observations were combined with numerical wave simulations using the open-source model SWAN. The field study encompassed one year of wave observations along a transect from 1.5 to 8.0 m depth, using a wave buoy and six pressure sensors. Seagrass characteristics were mapped on four occasions to capture seasonal variability. The effect on wave attenuation of the seagrass meadow was isolated from other dissipation processes by comparing model scenarios with and without vegetation. Results showed that maximum wave attenuation occurred under high-energy conditions, with a maximum wave height attenuation of 12%. However, as depth-induced breaking became the dominant dissipation process, the contribution of the seagrass meadow diminished, leading only to modest reductions in wave runup (1.0%), storm erosion volume (4.0%), and longshore sediment transport (0.6%). These findings indicate that seagrass meadows situated in relatively deep, fetch-limited environments offer limited potential for wave energy dissipation and coastal protection.

A novel hybrid forecast framework for significant wave height based on VMD-WOA-BiLSTM

IVYA-FMGRU: A frequency-domain context interaction model with bio-inspired optimization for significant wave height prediction

ABSTRACT Post-tropical Hurricane Fiona (2022) crossed the Gulf of St. Lawrence in Atlantic Canada, generating large waves and strong wave-driven currents in this semi-enclosed coastal sea. In this study, surface waves and circulation conditions were simulated using a dynamically coupled wave-circulation model revealing significant wave heights Hs ≤ 9.4 m and current speeds >1.0 m s−1. Good agreement was obtained for bulk wave statistics (r2 > 0.86 for Hs ), and broad-scale validation was achieved over a large area using observations from seven satellite altimeters. Higher resolution modelling of wind-generated waves in Malpeque Bay, a back-barrier bay connected to the Gulf, used a nested grid to simulate the waves in sheltered areas with short fetch. Significant wave heights ≤ 2.2 m in back-barrier regions, and current speeds ≤ 1.5 m s−1 in coastal areas around the barrier islands were simulated. Limitations to wind-wave growth were examined through comparison of stationary and nonstationary simulations (time constraints) and analysis of the ratio of storm translational speed to wave celerity (spatial constraints). The results indicate that despite extreme winds, the wave field remained dominated by locally generated young wind sea. Because the storm footprint was comparable to the horizontal basin scale, translating-fetch resonance and sustained swell development were spatially constrained and stationary simulations overpredicted peak wave heights (by up to ∼25%) under rapidly evolving wind conditions. These findings demonstrate how basin geometry can fundamentally limit hurricane wave growth in marginal seas, even during extreme events.

Significant wave height retrieval from spaceborne GNSS-R via interpretable machine learning and multi-source fusion: a case study of the tianmu-1 constellation

Abstract. North Atlantic extratropical cyclones (ETCs) cause significant financial losses in Europe, particularly in winter. Previous work has shown seasonal relationships between ETC hazards and modes of North Atlantic atmospheric variability, including the North Atlantic Oscillation (NAO; PC1) and East Atlantic Pattern (EAP; PC2). This study examines the relationship between the most extreme ETC hazards experienced at a given location in a winter season with the NAO and EAP, focusing on the winter maximum 10 m wind gust, marine significant wave height, daily maximum precipitation and winter total precipitation. We examine compound effects where PC1 or PC2 have signals in multiple hazard types at the same location. Positive PC1 exhibits coincident increases in winter maximum wind gust and significant wave height hazards around most coastal regions in northern Europe. Positive PC2 exhibits coincident increases in winter maximum wind gust and daily precipitation hazards over land areas in southern UK, Portugal and Spain, with an additional compound effect from increased significant wave height near the southern UK, northern France and Spain coasts. We also consider compound effects where PC1 and PC2 show coincident signals in the same ETC hazard at a given location, potentially indicating an elevated hazard likelihood when circulation anomalies project onto both modes concurrently. PC1 and PC2 have coincident signals for wind gusts in southern Ireland, southern UK, Portugal and the Scandinavian coast. For significant wave height, PC1 and PC2 have coincident signals around the Scandinavian, southern UK and Ireland and Northern Portugal coasts. This study shows that large-scale modes of seasonal North Atlantic climate variability modulate the exposure to extreme ETC hazards in many parts of Europe. The results have the potential to be combined with skilful seasonal climate forecasts of PC1 and PC2 to inform the insurance sector.

ABSTRACT Wind power has become a vital alternative energy source for coastal countries. This study integrates construction data from European wind farms with wind speed and wave height data from the Taiwan Strait to identify available weather windows and required working windows for key installation tasks. A deterministic scheduling model was developed to optimize construction vessel navigation, foundation and transition piece installation, and wind turbine installation based on these weather windows. The model results indicate that parallel (overlapping) construction of foundations and turbines is necessary to meet project timelines. Project d uration can be minimized by starting construction in December, while start dates between April and July also result in relatively short durations. Furthermore, the optimal period for initiating wind turbine installation is between December and January, allowing effective use of favorable weather conditions from April to December to accelerate progress. The findings provide a valuable reference for planning and scheduling offshore wind farm construction in Taiwan and other regions.

ABSTRACT The novelty of this letter lies in preliminarily analysis of the wave mapping mechanism using Gaofen-3 (GF-3) acquired in extended wide (EW) mode at low incidence angle of 10–20°. In total, 35 Chinese GF-3 images in vertical-vertical (VV) were collocated with hindcasted wave spectra from the numeric model WAVEWATCH-III (WW3). Validation of WW3-simulated significant wave heights (SWHs) against the Haiyang-2 (HY-2) altimeter wave products yields a root mean squared error (RMSE) of 0.44 m with a correlation coefficient (r) of 0.94 and a scatter index (SI) of 0.19, confirming the reliability of the WW3 simulations for this study. Subsequently, the synthetic aperture radar (SAR) intensity spectrum is simulated by multiplying the hindcasted wave spectrum by three modulation transfer functions (MTFs), i.e. tilt, hydrodynamic modulation, and velocity bunching. Based on experimental results, the spectrum correlation coefficient (Cor) and the spectrum squared error (E) of total energy between SAR intensity spectra and simulations using three MTFs are 0.94 and 7.79, respectively. These values represent significant large compared to other MTF combinations, which yield higher errors (approximately 6 E) and lower correlations (around 0.91 Cor). Therefore, it is concluded that velocity bunching has less influence on SAR mapping mechanism at low incidence angle, which is further supported by the Shapley Additive exPlanations (SHAP) analysis performed using eXtreme Gradient Boosting (XGBoost). This study establishes the foundation of wave retrieval from SAR image at low incidence angle.

A study was conducted on the interaction of surface gravity waves with a relatively thin, free-floating ice floe compared to the height of the waves. Physical and numerical modeling, as well as analytical research, were used. An overview of scientific works on the research topic is presented. The physical model consisted of an experimental setup (wave flume) with a wooden plate exposed to gravitational harmonic waves of different lengths and periods. The numerical model is based on calculations performed in the LS-DYNA program, where the fluid was simulated using the Euler–Lagrange method, and solid bodies were considered rigid. Analytical studies use the theory of interaction of small-amplitude waves with floating breakwaters. It is shown that as the wave height increases for conditions of interaction between waves and ice floes of almost identical horizontal dimensions, one end of the floating body sinks into the water, which leads to a significant reduction in the drift speed of the ice floe. Formulas have been obtained that express the ratio of the ice floe’s speed to the wave velocity, as well as the ratio of the height of the incident waves to the height of the transmitted waves, depending on the ratio of the wavelength to the horizontal dimensions of the floating ice floe.

On influence of AI- and physics-based estimation of significant wave height on ocean wave prediction using non-coherent X-band radar measurement

Characterizing the swell-dominated wave climate off West Africa: an intercomparison of global wave reanalyses

Abstract The nadir and off-nadir ocean wind and wave products of the Surface Waves Investigation and Monitoring (SWIM) instrument onboard the China France Oceanography Satellite (CFOSAT) are validated against multiple wave buoy datasets. These include a moored buoy in the Southern Ocean, the Northern Hemisphere focused wave buoys of the National Data Buoy Centre (NDBC), and a globally distributed Sofar drifting spotter buoy archive. SWIM nadir comparisons of significant wave height and wind speed data against NDBC buoy measurements showed that SWIM can capture these parameters with high accuracy and temporal stability. The off-nadir comparisons of SWIM bulk statistics of significant wave height and peak wave direction indicate high correlation and low errors against all buoy datasets. The performance of wave period (both mean and peak) is relatively poorer and varies by buoy dataset considered, with best results obtained against the Sofar archive. The accuracy of SWIM mean wave period against all buoy datasets is significantly improved by using a recently published neural networks-based inversion approach. Finally, quantitative evaluation of SWIM omni-directional spectra indicated an overall best performance for SWIM 10° beam product amongst the four SWIM products and against all exploited wave buoy datasets. This calibration and validation work is comprehensive and valuable because it describes the performance of CFOSAT wind/wave products in global oceanic conditions and against both moored and drifting wave buoy networks, with satellite-buoy collocations varying from 1605 to 7233 respectively for nadir wave and wind comparisons and over 3000 for off-nadir comparisons across all platforms. It is the intention that this work will enable the uptake of SWIM data with better-informed quantification of errors and accuracies.

Steady hydroelastic cnoidal wave on a thin flexural plate floating on a shallow water surface

Accurate short-term significant wave height forecasting is critical for coastal navigation, offshore energy operations, and hazard preparedness. However, operational wave forecast skill is often limited by uncertainty in wind forecasts. This study develops attention-based Long Short-Term Memory (LSTM) networks that explicitly address wind forecast uncertainty through noise-augmented training strategy and probabilistic modeling. Using wind and wave observations from the Western Long Island Sound (WLIS) buoy and operational North American Mesoscale (NAM) wind forecasts at Sikorsky Station, we characterize wind forecast errors and incorporate their statistical properties in model design. Three model variants are evaluated: (1) a perfect-future model trained with ideal future winds, (2) a noise-augmented model trained with controlled wind perturbations to enhance robustness, and (3) a probabilistic model that outputs both mean forecasts and uncertainty estimates. Results show that noise-augmented training reduced RMSE and MAE by approximately 7% compared to the perfect-future model and enhanced robustness to a broader range of wind forecast uncertainties. Furthermore, it led to smoother and more consistent performance across lead times. The probabilistic model achieved comparable accuracy while producing well-calibrated uncertainty estimates. Both models outperform a baseline relying solely on past buoy observations at all forecast horizons. These findings demonstrate that explicitly incorporating wind forecast uncertainty significantly improves wave height forecasting accuracy and reliability. The proposed framework offers a practical and robust pathway for operational wave forecasting systems under imperfect atmospheric forcing, enabling better-informed risk management and coastal decision-making. • Attention-based LSTM models using future winds outperform past-only baselines. • Noise-augmented training improves robustness to large wind forecast errors. • Probabilistic modeling provides calibrated uncertainty for risk-aware decisions. • Under forecast winds, proposed models reduce RMSE and MAE by about 7%.