 Prediction of hourly wind speed time series at unsampled locations using machine learningFreddy Houndekindo and Taha B.M.J. Ouarda Energy, 2024, vol. 299, issue C Abstract: Various models for wind speed mapping have been developed, with increasing attention on models focusing on mapping wind speed distribution. This study extends these models to predict hourly wind speed time series at unsampled locations. A model based on the quantile mapping (QM) procedure was compared to a traditional and machine-learning model to interpolate wind speed spatially. These proposed models were also used with inputs from the ERA5 reanalysis dataset, enabling them to consider local variation in orography and large-scale wind fields. A widely used procedure for mean bias correction of reanalysis based on the Global Wind Atlas (GWA) was implemented and compared to the proposed models. It was found that the QM and machine learning model, both using input from ERA5, significantly outperformed GWA bias correction in terms of time series correlation and probability distribution. Despite being more computationally intensive than GWA bias correction, both models are recommended due to their significantly (in a statistical sense) superior performance. Keywords: Bias-correction; ERA5; Light gradient-boosting machine; Quantile regression; Reanalysis; Wind resource assessment (search for similar items in EconPapers) Downloads: (external link) Related works: Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text Persistent link: https://EconPapers.repec.org/RePEc:eee:energy:v:299:y:2024:i:c:s036054422401291x DOI: 10.1016/j.energy.2024.131518 Access Statistics for this article Energy is currently edited by Henrik Lund and Mark J. Kaiser More articles in Energy from Elsevier |
Is your work missing from RePEc? Here is how to contribute.
Questions or problems? Check the EconPapers FAQ or send mail to .
EconPapers is hosted by the
School of Business at Örebro University.