Extreme Storm Surge Prediction Using Hydrodynamic Modelling and Artificial Neural Networks

摘要

On coastlines with shallow shelf areas (e.g. North Sea), a combination of high tides, storm surges, wind waves and mutual interactions generally represent the major sources of coastal flood risks: The contribution of the mutual interactions between the various components still remains the most unknown, despite the now routine linking of tidal and surge components in the current operational hydrodynamic storm-tide models. In fact, a proper physically-based coupling of all constituents will probably take decades to be implemented in the current operational models due to the highly complex and stochastic nature of the entire storm-tide system. Meanwhile, rather a more pragmatic data-driven approach is required to assess the contributions of these non-linear interactions to the resulting extreme storm-tide. Such a pragmatic approach is proposed, which is based on two types of artificial neural networks (ANNs) models called NARX (Nonlinear AutoRegressive eXogenous inputs): (ⅰ) NARX neural network model to predict the extreme storm-tide (Type-A), (ⅱ) NARX neural network model to nonlinearly correct the numerical storm-tide results from TELEMAC2D and TOMAWAC (Type-B). Ensembles methods are then used to reduce variance and minimize error especially in extreme storm-tide events. The approach was applied for two pilot sites in the North Sea (Cuxhaven and Sylt). The results show that the ensemble models are able to extract the contribution of the nonlinear interaction between the different extreme storm-tide components at both sites by subtracting the results of the hydrodynamic models (linear superposition of storm-tide constituents) from the ensemble results. In most extreme storm-tide events considered in this study, the contribution of the nonlinear interaction resulted in the reduction of the extreme water levels when compared with the linear superposition of extreme storm-tide components. However, under certain conditions, the nonlinear interactions might result in higher storm-tides than the linear superposition (e.g. storm of January 2000 at Cuxhaven and Sylt).