Vattenfall has completed a validation of the SARWind Timeseries dataset developed by CLS and Meteodyn. The assessment covered a five-year benchmark period from 2018 to 2022, during which Meteodyn supplied hourly wind speed and wind direction time series at 100 metres height based on satellite synthetic aperture radar observations.
Vattenfall compared the dataset with mast measurements from an offshore wind site in the North Sea. The comparison showed close agreement with the reference data. The correlation coefficient reached 0.958, while the mean wind speed error was -1.0% and the median wind speed error -1.8%. The root mean square error was 1.40 metres per second and the mean absolute error 1.03 metres per second.
Estimated gross annual energy production, based on the International Energy Agency 15MW turbine power curve, showed an error of -0.70%.
According to the benchmark results, SARWind Timeseries reduced annual energy production estimation error by 1.42% compared with mesoscale models. The dataset also provided a detailed representation of wind direction distribution as well as annual and diurnal wind cycles, supporting offshore wind project planning and operational analysis.
