The ability to predict beach morphodynamic changes at short to medium timescales is crucial for sustainable coastal management in a changing climate. The use of data driven approaches such as Machine Learning (ML) techniques to predict coastal change have gained interest in recent years (Kim, 2022, Montano, 2020). These techniques are proving to be a more computationally efficient alternative to the traditional process-based models. In this study tests two high-performance ML algorithms to predict cross-shore beach change. A deep learning neural network, Long-Short Term Memory (LSTM) and an ensemble ML model, extreme gradient boosting (XGBoost) are explored, and performances analysed, compared and contrasted.

Recent video images of tsunami attacks clearly show that tsunamis transport sediment from the seabed and riverbeds to land. Some studies also pointed out that sediment-containing fluids have a greater density than fresh water, so the hydrodynamic force acting on structures is more significant than that of pure water. However, it is common practice to use the advection-diffusion equation to estimate sediment transport by the tsunami, and the effect of the sediment concentration in the water on tsunami propagation needs to be taken into account. In this study, we solved the motion equation, which can take into account the effects of sediment in the water on tsunami propagation and run-up. We used a sediment transport model that does not rely on the advection-diffusion equation to predict tsunami damage associated with the Nankai Trough earthquake. The present model shows that the propagation velocity of a tsunami with a high concentration is faster than that of a low concentration.

Mediterranean deltaic areas have suffered reduced sediment inputs during the last decades due to dams and coastal structures, resulting into significant losses of wetlands and beaches. To design and develop beach protection strategies for future adaptation of deltaic coasts to climate change impacts, a more comprehensive understanding of the inherent dynamics between the backshore and the nearshore is crucial, especially during extreme events, when these areas become highly vulnerable as they get submerged under the action of waves and surges (Cohn and Ruggiero, 2016). Therefore, this study aims to determine the effect of the topobathymetric variability, on theevolution of the backshore and the nearshore during extreme events. Special attention is given to beach erosion and the formation of backshore sand deposits.

Bubbles and foam generated by breaking waves are primary surface signatures that have profound effects on remote sensing applications. Recent studies showed that wave breaking-induced bubbles and foam significantly increase LiDAR backscatter and affect LiDAR data processing, reducing the accuracy of airborne LiDAR measurements(Wang et al., 2022). Over the past several decades, significant efforts have been made to improve the understanding of wave-breaking mechanisms and their connection to optical signatures, such as bubbles and foam, in both the deep ocean and the nearshore field(e.g., Reul and Chapron, 2003, Carini et al., 2015). In this study, we pro- posed to develop a surface signature component in the phase-resolving wave model, FUNWAVE- TVD. FUNWAVE-TVD is a depth-integrated two- dimensional model for modeling wave propagation, breaking, and wave runup in a nearshore domain.

Loss of sediment from the Lagoon of Venice has been a well-known problem for many years. The erosion has been pronounced in the central part of the lagoon, where the navigational channel leading from the Malamocco Marghera Mouth to the industrial harbour of Marghera on the mainland west of Venice is also located. Previous studies have shown high concentrations of suspended sediment in the water column close to and inside the channel during and after vessel passages in the channel. The ‘Channeling the Green Deal for Venice’ project has sought to find means to mitigate this problem while balancing the continued functionality of the harbour. This paper describes a numerical model used in the study to simulate vessel generated hydrodynamics and erosion from displacement waves. Another important topic of the project has been the navigational safety.

Shore protection strategies worldwide increasingly incorporate hybrid dunes, which combine sandy dunes with hard structures. These hybrid dunes often intentionally combine classic erosion-resistant hard elements (rock, concrete elements, sea walls, etc) with sandy dunes. Hereby they aim to combine the best two worlds, as hard structures need less space and have less uncertainty in performance, while sandy dunes enable recreation and nature and can be more resilient under sea level rise. At times, hard elements such as buildings or outfall pipelines are introduced for other purposes, but unintentionally affect dune erosion during storms.

Coasts are highly dynamic areas that are particularly sensitive to the effects of climate change (CC) due to their geographical location and the variety of physical processes that shape them. It is well known in the scientific community that CC modifies the characteristics of various marine variables; in particular, the latest IPCC report (AR6, 2022) indicates that, regardless of future greenhouse gas emissions, a global mean sea level rise of between 10 and 25 centimeters is expected by 2050. In recent years, the modeling capacity of coastal morphodynamics has developed significantly in low-lying sandy beaches; however, these advances have not been directly extended to transitional zones such as estuaries, tidal inlets, or river mouths, mainly due to the difficulty and high computational cost of modeling the multiple physical processes and their interactions (Toimil et al., 2020). Transitional zones promote the proliferation of species and ecosystems of high ecological value; moreover, these areas are under great social and economic pressure, as a significant proportion of the population (e.g., 40percent in Spain) lives near the coast.

Shoreline changes as a result of increased anthropogenic activities in nearshore zones have caused unexpected damage to the hinterland. However, the current status of shoreline change prediction remains at a level where the basic concept is unestablished, and the existing method using the longshore sediment transport formula has uncertainties in the related coefficients and is limited in its scope of application. This study aimed to characterize the ultimate shoreline change due to topographical alteration by applying the concept of Equi-Wave Phase Potential (EWPP) (Ria and Lee, 2023) and proposes a numerical methodology for converging to the new morphological equilibrium due to the change of wave phases after sand mining.

Beach nourishment from relict sand deposit is considered as the main soft engineering intervention for the coastal defense from extreme events and sea level rise. For the success of the intervention a detailed study of the physiographic units is needed as well as the study on the sand compatibility but even the management of the operation in the execution in phase play an important role. To optimize the dredging and nourishment activities, limit the overall impact of the intervention and obtain a more stable beach profile in the short term after the intervention, forecast waves for sediment transport prediction can be used during the operational phase.

The assessment of sea structures is a critical endeavor in coastal engineering, ensuring the safety and sustainability of coastal developments and infrastructures. Effective management and protection of coastal areas require reliable predictions of shoreline evolution when hard structures are present. In most cases, sea structure assessment involves dealing with both deterministic and random factors. Deterministic modeling is adopted to determine the shoreline response based on wave conditions, providing a baseline understanding on how the structure might behave under standard circumstances. However, the natural variability of the marine environment means that deterministic models alone are insufficient to capture the full range of possible scenarios. The present study addresses the inherent randomness in wave conditions by employing a stochastic technique to generate random sea states based on known parameters. This novel approach is applied to the coastal area of Milford-on-Sea, located in Hampshire (UK) where, in 2007, a field experiment was conducted involving the temporary installation of a 46-meter-long groyne.