A Data-Driven Approach to Offshore Wind Forecasting in the Celtic Sea

ORCID

Abstract

Accurate weather forecasting is crucial for various industries,including offshore wind, which is vital for global net zeroenergy goals. Machine learning models, trained on historicaldata, offer a new opportunity by replacing traditional physicsbasedequations. These models can learn patterns not alwaysrepresented by physical equations, potentially increasing the accuracyand efficiency of weather forecasting compared to traditionalNumerical Weather Prediction (NWP).This study applies a machine learning framework (MaLCOM)to offshore wind forecasting in the Celtic Sea. It uses anattention-based LSTM recurrent neural network to learn temporalpatterns and a random forest-based spatial nowcasting model,trained on ERA5 data, for spatiotemporal predictions. Windsderived from wave spectra measured by buoys are integrated,showing the framework’s value even with imperfect data.Validation with independent observations from floating lidarunits in 2023 confirms the framework’s suitability for regionalwind prediction. This work extends previous machinelearning-based predictions of ocean conditions to wind forecasting,demonstrating a new approach to metocean data. Theselightweight, data-driven predictions can be run on standard computers,improving real-time decision-making for offshore planning.

Publication Date

2025-01-01

Publication Title

44th International Conference on Ocean, Offshore & Arctic Engineering

ISSN

2153-4772

Embargo Period

2026-06-01

This document is currently not available here.

This item is under embargo until 01 June 2026

Share

COinS