Storm surge induced by severe typhoons has caused many catastrophic tragedies to coastal communities over past decades. Accurate and efficient prediction/assessment of storm surge is still an important task in order to achieve coastal disaster mitigation especially under the influence of climate change. This study revisits storm surge predictions using artificial neural networks (ANN) and effective typhoon parameters. Recent progress of storm surge modeling and some remaining unresolved issues are reviewed. In this paper, we chose the northeastern region of Taiwan as the study area, where the largest storm surge record (over 1.8 m) has been observed. To develop the ANN-based storm surge model for various lead-times (from 1 to 12 h), typhoon parameters are carefully examined and selected by analogy with the physical modeling approach. A knowledge extraction method (KEM) with backward tracking and forward exploration procedures is also proposed to analyze the roles of hidden neurons and typhoon parameters in storm surge prediction, as well as to reveal the abundant, useful information covered in the fully-trained artificial brain. Finally, the capability of ANN model for long-lead-time predictions and influences in controlling parameters are investigated. Overall, excellent agreement with observations (i.e., the coefficient of efficiency CE > 0.95 for training and CE > 0.90 for validation) is achieved in one-hour-ahead prediction. When the typhoon affects coastal waters, contributions of wind speed, central pressure deficit, and relative angle are clarified via influential hidden neurons. A general pattern of maximum storm surge under various scenarios is also obtained. Moreover, satisfactory accuracy is successfully extended to a much longer lead time (i.e., CE > 0.85 for training and CE > 0.75 for validation in 12-h-ahead prediction). Possible reasons for further accuracy improvement compared to earlier works are addressed.

Storm surges are one of the most devastating coastal disasters. Numerous efforts have continuously been made to achieve better prediction of storm surge variation. In this paper, we propose a parametric cyclone and neural network hybrid model for accurate, long lead-time storm surge prediction. The model was applied to the northeastern coastal region of Taiwan, i.e., Longdong station. A total of 14 historical typhoon events were used for model training and validation, and the results and questions associated with this hybrid approach carefully discussed. Overall, the proposed method reduced the complexity of network structure while retaining the important typhoon indicators. In particular, local pressure and winds estimated from the storm parameters through physically-based parametric cyclone models allow for inferring the possible future influence of a typhoon, unlike the simple collection and direct usage of observation data from local stations in earlier works. Meanwhile, the error-tolerance capability of the neural network alleviated some discrepancy in the model inputs and enabled good surge prediction. Further, the proposed method showed better and faster convergence thanks to the retention of storm information and the reduced dimensions of the search space. The hybrid model presented excellent performance or maintained reasonable capability for short lead-time and long lead-time storm surge prediction. Compared with the pure neural network model under the same network dimensions, the present model demonstrated great improvement in accuracy as the prediction lead time increased to 8 h, e.g., 33–40% (13–21%) and 32–37% (18–29%) RMSE and CE, respectively, in the training/validation phase.

A series of experiments exploring the propagation of a solitary wave over a submerged step were performed using a flow-field visualization measurement system, an image-connection technique as well as model simulations. The experimental data were used to validate a one-layer finite-element non-hydrostatic model and a multi-layer finite-difference non-hydrostatic σ model for various submerged step configurations and wave conditions—combinations of step height ratios d/h, width ratios B/h and solitary wave height ratios H/h, where d denotes the step height, B the step width, H the solitary wave height, and h the still water depth. The main differences between the numerical results and the experimental data are highlighted. The effect of the height and width of the submerged step as well as the wave height of the solitary wave are quantified in terms of reflection (R), transmission (T), and energy dissipation (D). Through a series of numerical experiments, an optimal combination of the height ratio d/h, width ratio B/h, and solitary wave height ratio H/h for breakwater design for coastal protection is suggested.

Marine energy development has been actively promoted in recent years. Analyzing wave data in the surrounding sea areas is crucial for improving wave power generator efficiency. This study collected unidirectional wave spectrum data from the northeast coastal area of Taiwan (e.g., Pengjia Islet, Fugui Cape, and NTOU). It aimed to analyze the main wave types (i.e., wind waves, swell, or mixed sea) and their proportions during the northeast and non-northeast monsoon periods. The results indicate that swell waves dominate at the NTOU buoy, while wind waves dominate at the other two stations. In addition, the analysis of wave spectra ratios revealed a low percentage of unimodal spectra (only about 0.5%), and bimodal and unformed spectra presented as predominant. Furthermore, this study also introduced a brand new JW-J wave spectrum, which combines the JW and JONSWAP spectra to describe low-frequency (swell) and high-frequency (wind wave) systems, showing superior performance compared to the Torsethaugen spectrum. Finally, this research analyzed the JW-J spectrum parameters from the Pengjia Islet and Fugui Cape stations and applied these results to the NTOU bouy for effectively describing wind/swell wave variations.

Green Island, located in the typhoon-active eastern Taiwan coastal water, is the potential Kuroshio power plant site. In this study, a high resolution (250–2250 m) shallow-water equations model is used to investigate the effect of typhoon on the hydro-dynamics of Kuroshio and Green Island wakes. Two typhoon–Kuroshio interactions—typhoon Soulik and Holland’s typhoon model—are studied. Simulation results of typhoon Soulik indicate salient characteristics of Kuroshio, and downstream island wakes seems less affected by the typhoon Soulik, because the shortest distance of typhoon Soulik is 250 km away from Green Island and wind speed near Green Island is small. Moreover, Kuroshio currents increase when flow is in the same direction as the counterclockwise rotation of typhoon, and vice versa. This finding is in favorable agreement with the TOROS (Taiwan Ocean Radar Observing System) observed data. Simulations of Kuroshio and Holland’s typhoon model successfully reproduces the downstream recirculation and vortex street. Numerical results reveal that the slow moving typhoon has a more significant impact on the Kuroshio and downstream Green Island wakes than the fast moving typhoon does. The rightward bias phenomenon is evident—Kuroshio currents increase (decrease) in the right (left) of the moving typhoon’s track, due to the counterclockwise rotation of typhoon.

A multi-layer hydrostatic shallow-water model was developed in the present study. The layer-integrated hydrostatic nonlinear shallow-water was solved with $\theta$ time integration and the least-squares finite element method. Since the least-squares formulation was employed, the resulting system of equations was symmetric and positive–definite; therefore, it could be solved efficiently by the preconditioned conjugate gradient method. The model was first applied to simulate the von Karman vortex shedding. A well-organized von Karman vortex street was reproduced. The model was then applied to simulate the Kuroshio current-induced Green Island vortex street. A swirling recirculation was formed and followed by several pairs of alternating counter-rotating vortices. The size of the recirculation, as well as the temporal and spatial scale of the vortex shedding, were found to be consistent with ADCP-CDT measurements, X-band radar measurements, and analysis of the satellite images. It was also revealed that Green Island vortices were affected by the upstream Orchid Island vortices.