Posters

Abstract

Wind energy is booming, but we need to accelerate its growth if we want to meet our climate goals. One way to do this is to use wind speed data from climate reanalysis to identify potential sites for wind energy projects. However, the low grid resolution of climate reanalysis data limits its accuracy at the microscale. To address this limitation, a variety of tools and models have been developed for downscaling wind speed from reanalysis data to a horizontal grid resolution lower than 300m. In this study, we evaluated the accuracy of two of the most widely used downscaling tools, the Global Atlas of Siting Parameters (GASP) and the Global Wind Atlas (GWA). The uncertainty benchmark was performed with the measured data from 50 met-masts located around the 5 continents over more than 20 countries and terrain complexities ranging from simple to complex. All measurement periods were longer than 1 year of data, operating between 2014 and 2022. Each of this measurement mast data was analyzed and extrapolated to the long term for the estimation of energy yield. The normalized wind speed bias for GASP ranges from -18.6% to 33.7% with an average of 8.2% and -1.3 to 2.5 m/s for the bias with an average of 0.1m/s. The standard deviation is 10.2% for the normalized bias and 0.8 m/s for the bias. The normalized bias for GWA ranges from -28.7% to 20.2% with an average of 8.5% and -2.1 to 1.5 m/s for the bias with an average of 0.1m/s. The standard deviation is 10.5% for the normalized bias and 0.8 m/s for the bias. Our results are similar to the validations performed by EMD and DTU for GASP (normalized bias of 6.9%) and GWA (normalized bias of 8.3%) respectively. This alignment of the studies provides confidence in the evaluation of the accuracy of the tools. This study provides additional support for the accuracy of GASP and GWA wind speed predictions, which could help wind developers make better decisions about where to site their projects. While wind maps or reanalysis data can provide a rough estimate of a site's potential, costly measurement masts are often required. Precise and accurate assessments of a site's energy potential based on reanalysis data alone would save time and money. However, current tools, including GWA and GASP, despite their increasing accuracy and correlation with real measurements, are still very limited for accurately estimating a site's potential. The available data for the validation was limited and there are several approaches to improve the quality of the validation. With the available data it was not possible to identify any correlation between the accuracy of the tools and the type of terrain. However, a large dataset could allow to find trends between these two as the ones identified by EMD in their validation of GASP.