MorphoGP is a nonparametric Gaussian process framework designed to predict equilibrium beach profiles under tidal influence. It first classifies beach morphologies using a ContourCluster model based on contrastive learning, then trains a specialized Gaussian process expert for each category to learn relationships between environmental descriptors (waves, tides, sediments) and beach shape. A Gating Net probabilistically combines the experts’ outputs, achieving a 59.3% reduction in test RMSE compared to the best baseline, with a final RMSE of 0.297 m on over 180 Chinese coast beach profiles.
arXiv:2609.37565v1 Announce Type: new Abstract: Deep learning surrogates can produce high-resolution coastal flood maps orders of magnitude faster than physics-based hydrodynamic simulators, yet tran...
arXiv:2606. 13302v1 Announce Type: new Abstract: Wave parameters in the nearshore are crucial for coastal engineering, shoreline protection, marine hazard assessment, and coastal management for climate resilience.
arXiv:2609.25505v1 Announce Type: cross Abstract: Rapid intensification (RI) remains one of the most consequential and difficult aspects of tropical cyclone (TC) forecasting. Although full-physics nu...
arXiv:2606. 12141v1 Announce Type: new Abstract: Accurate forecasting of sea surface temperature (SST) in regional seas such as the East Sea is crucial for monitoring marine ecosystems, assessing climate risks, managing fisheries, and conducting naval operations.
arXiv:2602. 13010v2 Announce Type: replace Abstract: Accurate production forecasts are essential for the integration of renewable energy sources into the power grid.
The paper performs a systematic hyperparameter search across five deep learning architectures (DLinear, LSTM, PatchTST, ResAttLstm, and Mamba2) and nine context lengths for forecasting significant wave height (Hs) at a single buoy, then re‑evaluates the best configurations on a 47‑buoy, 37‑year dataset. Across all models, performance converges to a very small spread (SD = 0.0014 m², 0.8% of the grand mean), yet all models still outperform persistence, though none consistently outperforms the others. The study finds that persistence already captures the dominant linear‑inertial signal in univariate Hs, and that model‑class differences are dwarfed by cross‑buoy variance, suggesting diminishing returns for further architecture engineering under univariate input settings.
arXiv:2606. 04143v1 Announce Type: cross Abstract: Accurate flood forecasting is essential for mitigating disaster risks and protecting communities.
arXiv:2607. 07834v1 Announce Type: new Abstract: Pseudo-nitzschia diatoms pose recurrent risks to coastal ecosystems and shellfish harvesting along the Portuguese Atlantic coast.
arXiv:2606. 26389v1 Announce Type: cross Abstract: Sea state prediction is essential for operational maritime applications and coupled earth system modeling, yet current spectral wave models remain computationally prohibitive for many use cases, including online coupling to climate simulations and making probabilistic (ensemble-based) predictions.
arXiv:2608. 10022v1 Announce Type: cross Abstract: The large-scale oceanic and atmospheric forecasts provided by global climate models typically lack sufficient resolution to accurately capture the response of the coastal ocean to atmospheric forcing and coastal circulation that drive fine-scale SST variability.
The study evaluates how feature engineering (FE) affects machine learning models for ocean colour data, proposing a seven‑step optimisation framework that includes band selection, scaling, normalisation, index extraction, PCA, and feature scaling. Applied to Sentinel‑3 OLCI observations, the framework improves model accuracy for estimating Chlorophyll‑a and Secchi disk depth, achieving higher R values and lower mean absolute errors compared to standard algorithms. However, the optimal FE varies across targets and models, indicating that FE optimisation must be tailored to each application.
arXiv:2608. 19899v1 Announce Type: cross Abstract: Machine learning (ML) is widely used for the development of ocean colour algorithms, but most studies focus on model parameter training and hyperparameter tuning.