Data-driven modeling of wind waves in upper Delaware Bay with living shorelines

Machine Learning‐Based Wave Model With High Spatial Resolution in Chesapeake Bay

“…Unlike other ML wave models applied at a single station or selected limited stations (Adytia et al., 2022; Chen et al., 2021; Wang et al., 2023), we employed a dimensionality reduction approach to transform the high‐dimensional data into a lower‐dimensional space, and therefore, we only trained the temporal principal components. Although this approach will introduce model errors using limited PCs, the approach allows the model to predict wave heights in high spatial resolution.…”

“…The time series simulation shows that RMSE using 8 PCs is only slightly different from the RMSE using 16 PCs (Figure 5 and Figure S8 in Supporting Information ). This satisfactory performance can be attributed to the fetch‐limited environment, as both the simplified fetch‐limited model and the machine learning model accurately predict wave distribution inside the Bay (Cerco et al., 2010; Lin et al., 2002a; Wang et al., 2023).…”

“…In recent years, there has been a growing interest in utilizing machine learning (ML) models for wave simulation in estuaries and coastal waters, providing an alternative to numerical‐based dynamical and statistical downscaling approaches (O'Donncha et al., 2018). While ML techniques have been employed for wave downscaling, many applications focus on wave conditions at fixed or specific locations (Adytia et al., 2022; Chen et al., 2021; Wang et al., 2023). Michel et al.…”

Modeling of water waves and sediment transport using physics-based and machine learning methods

“…Many data-driven models have been developed to study nonlinear relationships between input features and labels for coastal engineering applications during the last two decades, such as artificial neural networks (ANN), Bayesian networks (BN), support vector machines (SVM), decision trees, and fuzzy inference systems (FIS) (e.g., Deo and Naidu, 1998;Deo et al, 2001;Jain et al, 2011;Peres et al, 2015;Cornejo-Bueno et al, 2016;Sadeghifar et al, 2017;Parker and Hill 2017;Oh and Suh, 2018;Stringari et al, 2019;Zheng et al, 2020;Wei, 2021;Miky et al, 2021;Jörges et al, 2021;Elbisy and Elbisy, 2021;Bento et al, 2021;Mares-Nasarre et al, 2021;Lee et al, 2021;Wang et al, 2022a;Wang et al, 2022c). For instance, Malekmohamadi et al (2011) compared the performance of ANN, BN, SVM, and Adaptive Neuro FIS methods for estimating wave height with wind data in Lake Superior, USA.…”