Predicting Lake Erie wave heights and periods using XGBoost and LSTM. (August 2021)

• Record Type: Journal Article
• Authors: Hu, Haoguo; van der Westhuysen, André J.; Chu, Philip; Fujisaki-Manome, Ayumi
• Is Part Of: Ocean modelling. Volume 164(2021)
• Publication Date: 2021-08
• Subjects: Oceanography -- Periodicals; Océanographie -- Périodiques; Oceanography; Periodicals; 551.46
• DOI: 10.1016/j.ocemod.2021.101832
• Languages: English
• ISSNs: 1463-5003

Abstract

Abstract: Waves in large lakes put coastal communities and vessels under threat, and accurate wave predictions are needed for early warnings. While physics-based numerical wave models such as WAVEWATCH III (WW3) are useful to provide spatial information to supplement in situ observations, they require intensive computational resources. An attractive alternative is machine learning (ML) methods, which can potentially improve the performance of numerical wave models, while only requiring a small fraction of the computational cost. In this study, we applied novel ML methods based on XGBoost and a Long Short-Term Memory (LSTM) recurrent neural network for predicting wave height and period under the near-idealized wave growth conditions of Lake Erie. Data sets of significant wave height ( H ), peak wave period ( T p ) and surface wind from two offshore buoys from 1994 to 2017 were processed for model training and testing. We trained and validated the ML models with the data sets from 1994 to 2015, and then used the trained models to predict significant wave height and peak period for 2016 and 2017. The XGBoost model yielded the best overall performance, with Mean Absolute Percentage Error (MAPE) values of 15.6%–22.9% in H and 8.3%–13.4% in T p . The LSTM model yielded MAPE values of 23.4%–30.8% in H and 9.1%–13.6% in T p . An unstructured grid WW3 applied to Lake Erie yielded MAPE values of 15.3%–21.0% in H and 12.5%–19.3% in T p . However, WW3 underestimated H and T p during strong wind events, with relative biases of -11.76% to -14.15% in H and -15.59% to -19.68% in T p . XGBoost and LSTM improve on these predictions with relative biases of -2.56% to -10.61% in H and -8.08% to -10.13% in T p . An ensemble mean of these three models yielded lower scatter scores than the members, with MAPE values of 13.3%–17.3% in H and 8.0%–13.0% in T p, although it did not improve the bias. The ML models ran significantly faster than WW3: For this 2-year run on the same computing environment, WW3 needed 24 h with 60 CPUs, whereas the trained LSTM needed 0.24 s on 1 CPU, and the trained XGBoost needed only 0.03 s on 1 CPU. Highlights: Machine learning models XGBoost and LSTM applied in Lake Erie for wave prediction. XGBoost and LSTM accuracy comparable to WW3, but reduce storm peak underestimation. XGBoost and LSTM require only a small fraction of the computational expense of WW3.