A Machine Learning Framework to Forecast Wave Conditions

Scott C. James,Yushan Zhang,Fearghal O'Donncha

DOI: https://doi.org/10.48550/arXiv.1709.08725

2017-09-26

Abstract:A~machine learning framework is developed to estimate ocean-wave conditions. By supervised training of machine learning models on many thousands of iterations of a physics-based wave model, accurate representations of significant wave heights and period can be used to predict ocean conditions. A model of Monterey Bay was used as the example test site; it was forced by measured wave conditions, ocean-current nowcasts, and reported winds. These input data along with model outputs of spatially variable wave heights and characteristic period were aggregated into supervised learning training and test data sets, which were supplied to machine learning models. These machine learning models replicated wave heights with a root-mean-squared error of 9cm and correctly identify over 90% of the characteristic periods for the test-data sets. Impressively, transforming model inputs to outputs through matrix operations requires only a fraction (<1/1,000) of the computation time compared to forecasting with the physics-based model.

Atmospheric and Oceanic Physics

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to develop a machine - learning framework to predict ocean wave conditions. Specifically, through supervised training of machine - learning models, the authors utilized data from thousands of iterations of physics - based wave models to accurately represent significant wave height and period, thereby predicting ocean conditions. The goal of this method is to reduce computational costs while maintaining prediction accuracy comparable to that of physics - based models. In the paper, Monterey Bay was used as a test case. Through input data such as measured wave conditions, ocean current forecasts, and reported wind speeds, as well as the spatially varying wave height and characteristic period output by the model, training and testing data sets for supervised learning were constructed. These data sets were used to train the machine - learning model, which was able to predict wave conditions with extremely low error (for example, the root - mean - square error of significant wave height was 9 centimeters, and more than 90% of the characteristic periods were correctly identified), and its computation time was only a fraction (less than 1/1000) of that of the physics - based model.