Since its release in 2021, the global surface wind climatology data available in the ERA5-Land product have seen widespread application in studies on wind energy potential and wind farm deployment around the world, because the wind data are provided at a higher spatial resolution than its original source, the ERA5 reanalysis. The study presented here reveals that ERA5-Land surface wind fields do not correctly reflect the assimilated hind-cast model results of the ERA5 reanalysis. ERA5-Land significantly and consistently under-represents surface wind speeds by 20–30% compared with the original ERA5 reanalysis they are derived from, contrary to the claim that the wind fields in the former are simply a higher resolution interpolation of the latter. Results indicate that the wind speed discrepancies lead to a roughly 20–30% under-estimation of annual onshore wind power density at typical turbine hub height in the windy regions of NW-Europe, 10–20% in the Great Plains of the USA, and up to 10% in North and Western China. The study shows that the lower wind speeds in ERA5-Land cannot be explained by the interpolation procedures, suggesting a fundamental problem with the wind fields applied in the land surface model of ERA5-Land and potentially many other variables derived from it. The systematic discrepancy in surface wind speeds in ERA5-Land has important ramifications for numerous recent wind energy studies as well as for research in broader environmental and Earth sciences. • A hitherto unknown fundamental flaw has been found in a widely used wind climatology dataset. • ERA5-Land surface wind fields do not correctly reflect the original ERA5 reanalysis. • ERA5-Land under-estimates surface wind speeds by 20–30%. • Resulting Wind Power Density at hub-height is up to 20% lower than it should be. • The incorrect wind fields appear to originate from within the ERA5-Land surface model.
Andreas C.W. Baas (Tue,) studied this question.