Multi-Step-Ahead Forecasting of Wave Conditions Based on a Physics-Based Machine Learning (PBML) Model for Marine Operations

Abstract

Short-term wave forecasts are essential for the execution of marine operations. In this paper, an efficient and reliable physics-based machine learning (PBML) model is proposed to realize the multi-step-ahead forecasting of wave conditions (e.g., significant wave height Hs and peak wave period Tp). In the model, the primary variables in physics-based wave models (i.e., the wind forcing and initial wave boundary condition) are considered as inputs. Meanwhile, a machine learning algorithm (artificial neural network, ANN) is adopted to build an implicit relation between inputs and forecasted outputs of wave conditions. The computational cost of this data-driven model is obviously much lower than that of the differential-equation based physical model. A ten-year (from 2001 to 2010) dataset of every three hours at the North Sea center was used to assess the model performance in a small domain. The result reveals high reliability for one-day-ahead Hs forecasts, while that of Tp is slightly lower due to the weaker implicit relationships between the data. Overall, the PBML model can be conceived as an efficient tool for the multi-step-ahead forecasting of wave conditions, and thus has great potential for furthering assist decision-making during the execution of marine operations.

Improving Significant Wave Height Forecasts Using a Joint Empirical Mode Decomposition–Long Short-Term Memory Network

“…As a result, forecast skill is significantly improved. With regards to high-frequency IMFs, it is hypothesized that the inclusion of wind information may increase the accuracy of wave forecasts, as shown in past research [40][41][42]. Although this is currently outside the scope of the present work, additional research into investigating this hypothesis is of significant value to overcome the present limitations of the EMD-LSTM model.…”

Forecasting Rogue Waves in Oceanic Waters

“…The interesting thing to note here is that while the performance of ET continues to decrease till 24 hours, the performance of LightGBM does not decrease similarly. From Figures (6)(7)(8)(9) and Table 3, we observe that the performance of the machine learning algorithms decreases as the time steps for forecasting increase. Specifically, for both algorithms, Coefficient Correlation, Variance, Adjusted R2 decrease rapidly for predictions till 15 hours but decrease slowly afterward.…”

“…Thus, every 3 hourly predictions add 6 steps to our proposed method. Figures (2)(3)(4)(5)(6)(7)(8)(9), Table 3, and Table 4 show the performance of LightGBM and ET for predictions up to 24 hours ahead in 3-hour intervals. In Figures (2)(3)(4)(5) and Table 2, we can see that the performance of Extra Trees decreases rapidly from 0 to 3 hours but then decreases slowly for MAE, RMSE, HH, and SI.…”

“…The prediction of ocean waves and their properties is important for offshore engineering projects like the exploitation of marine renewable energy, harbor construction, marine operations, marine, and navy marine forces operations, etc [2][3][4][5][6][7]. Since the physical process of wave generation is uncertain, nonlinear, and non-stationary, dominant wave period prediction remains difficult [8]. The dominant wave period is defined as the wave period associated with the highest energetic waves at a specific point or area in the total wave spectrum and is always the swell period or wind-wave period.…”

“…Although numerical models are considered accurate for the small-time domain, they are computationally costly and time-consuming, which limits their use for feedback control operations [14]. Furthermore, they have low generalization ability for different domains, and the models must be run periodically, resetting the boundary conditions for different regions [8]. To deal with these problems, various data-driven and statistical methods have been proposed.…”