Assessing wave energy as extreme events propagate near the coast

Abstract

Keywords Physical modeling; Wave flume; Extreme waves; Wavelet transform; Machine Learning; MLP model Physical modeling, spectral analysis, and artificial intelligence techniques were used to study extreme wave behavior and its evolution in shallow waters. A series of physical tests were conducted in a laboratory wave flume using different wave spectra, including JONSWAP (γ = 7), JONSWAP (γ = 3.3), and Pierson-Moskowitz, varying within a broad range of wave amplitudes. The dispersive focusing technique was used to generate these spectral waves. To account for the varying duration of extreme events, one, three, six, and nine wave trains were generated. A total of fifty-one wave gauges, located between 4 m and 14 m from the wave generator, provided comprehensive monitoring of the wave characteristics and their propagation along the wave flume [1]. The analysis incorporates wavelet transform to identify frequency components and their assigned energy using the Maximal Overlap Discrete Wavelet Transform (MODWT) method. The energy of the dominant frequency components, d5 and d4, which represent the peak frequency (fp = 0.75 Hz) and its first harmonic (2fp = 1.5 Hz), respectively, has significantly decreased. In contrast, the energy of the remaining components has increased. By investigating the energy of each frequency component along the wave flume, potential correlations between the dissipation of dominant frequency components and zones of higher energy dissipation are explored. Moreover, using the Multilayer Perceptron (MLP) machine learning algorithm [2], the study confirmed the repeatability of our findings regarding the energy of the frequency components with an accuracy of 98%. This study demonstrates the effectiveness of the MLP algorithm in improving wave prediction using field experimental data. [1] Zhang, J., Benoit, M., Kimmoun, O., Chabchoub, A., & Hsu, H. C. (2019). Statistics of extreme waves in coastal waters: large scale experiments and advanced numerical simulations. Fluids, 4(2), 99. [2] Abroug, I., Matar, R., & Abcha, N. (2022). Spatial Evolution of Skewness and Kurtosis of Unidirectional Extreme Waves Propagating over a Sloping Beach. Journal of Marine Science and Engineering, 10(10), 1475.