Abstract:Coastal hazards such as flooding and erosion can cause large economic and human losses. Under this threat, early warning systems can be very cost-effective solutions for disaster preparation. The goal of this study was to develop, test, and implement an operational coastal erosion early warning system supported by a particular method of machine learning. Thus, the system combines Bayesian Networks, and state-of-the-art numerical models, such as XBeach and SWAN, to predict storm erosion impacts in urbanized areas. This system was developed in two phases. In the development phase, all information required to apply the machine learning method was generated including the definition of hundreds of oceanic synthetic storms, modeling of the erosion caused by these storms, and characterization of the impact levels according to a newly defined Erosion Impact index. This adimensional index relates the distance from the edge of the dune/beach scarp to buildings and the height of that scarp. Finally, a Bayesian Network that acted as a surrogate of the previously generated information was built. After the training of the network, the conditional probability tables were created. These tables constituted the ground knowledge to make the predictions in the second phase. This methodology was validated (1) by comparing 6-h predictions obtained with the Bayesian Network and with process-based models, the latest considered as the benchmark, and (2) by assessing the predictive skills of the Bayesian Network through the unbiased iterative k -fold cross-validation procedure. Regarding the first comparison, the analysis considered the entire duration of three large storms whose return periods were 10, 16, and 25 years, and it was observed that the Bayesian Network correctly predicted between 64% and 72% of the impacts during the course of the storms, depending on the area analyzed. Importantly, this method was also able to identify when the hazardous conditions disappeared after predicting potential consequences. The k -fold cross-validation procedure was applied to the peak of a set of varying storms and it demonstrated that the predictive skills were maximized (63%–72%) when including three nodes as input conditions of the Bayesian Network. In the operational phase, the system was integrated into the architecture of a forecast and early warning system that predicts emergencies in coastal and port zones in Portugal, and the alerts are issued to authorities every day. This study demonstrated that the two-phase approach developed here can provide fast and high-accuracy predictions of erosion impacts. Also, this methodology can be easily implemented on other sandy beaches constituting a powerful tool for disaster management.

Machine learning tool for wave overtopping prediction based on the safety-operability ratio

Oscillatory Responses of a Spatially Random Seabed in the Vicinity of Structures: Validation and Application of the Integrated CFD-SFEM

Numerical Investigations of Pedestrian Overboard on a Compound Slope Dike Due to Excessive Overtopping Waves

A copula-based model to describe the uncertainty of overtopping variables on mound breakwaters

Numerical modelling of the spatial distribution of wave overtopping water behind dikes under regular wave action using OpenFOAM

Development of a Bayesian network-based early warning system for storm-driven coastal erosion

Measuring Prediction Accuracy in a Maritime Accident Warning System

Computational Study of Overtopping Phenomenon over Cylindrical Structures Including Mitigation Structures

Physical and numerical modeling of random wave transformation and overtopping on reef topography

Resilient Coastal Protection Infrastructures: Probabilistic Sensitivity Analysis of Wave Overtopping Using Gaussian Process Surrogate Models

Phase-resolved wave prediction with varying buoy positions in the field using machine learning-based methods

Exceedance Probability Forecasting via Regression for Significant Wave Height Prediction

Predictive capabilities of data-driven machine learning techniques on wave-bridge interactions

Validation and Application of the Accu-Waves Operational Platform for Wave Forecasts at Ports

Evaluation of Pedestrian Safety for Wave Overtopping by Ship-Induced Waves in Waterfront Revetment

Cointegration strategy for damage assessment of offshore platforms subject to wind and wave forces

Experimental Study of Wave Overtopping Flow Behavior on Composite Breakwater

A Machine Learning Framework to Forecast Wave Conditions

Wave Overtopping Control Effect of AWOC Equipped With Self-Adaptive Floating Concrete Structure: A Numerical Study

A physics-informed machine learning model for time-dependent wave runup prediction

Intelligent prediction of wave loads based on multi-source data-driven methods