Performance of Recent Wind Assessment Datasets in Coastal Areas

Coastal communities, particularly those on islands or in remote locations, have unique characteristics that impact the energy technologies available to support them. The U.S. Department of Energy’s Energy Transitions Initiative Partnership Project collaborates with island and remote communities to advance their energy portfolios and increase their energy resilience, and distributed wind energy is frequently considered for its potential to address each goal.

The need for accurate, user-friendly, and affordable wind resource assessments has driven the development of a variety of models and tools over the last decade. Global Wind Atlas 3 is a high-resolution wind database developed by the Technical University of Denmark and the World Bank Group to support policymakers, planners, and investors identify areas of high wind resource for wind energy generation at locations across the world. The 2023 National Offshore Wind dataset (NOW-23), developed by the National Renewable Energy Laboratory, the University of Colorado, and Veer Renewables, offers the latest wind resource information for offshore regions in the United States but provides wind information for large land-based areas as well. The WIND Toolkit Long-Term Ensemble Dataset includes a climate-scale simulation created by Argonne National Laboratory with a variety of wind data output heights. All three recent wind datasets are freely accessible in user-friendly web applications and provide varying degrees of spatial and temporal resolution.

The presented analysis will evaluate the performance of the three recent wind assessment datasets in previously unvalidated locations along United States coastlines. The validation heights (20 m – 60 m) in this work support coastal communities interested in adopting small or midsize wind energy, using turbines up to 1 MW in capacity. The performance of the recent datasets is compared with that of ERA5, a popular global reanalysis model widely utilised for wind energy resource assessments as a standalone product, as input boundary conditions for higher-resolution model runs, and as a reference dataset for measure-correlate-predict estimates with local observations. Validation is performed on multiple timescales, including annual average, interannual, seasonal, and diurnal in order to identify the strengths and limitations of each recent wind dataset.