Using bias-correction to improve future projections of offshore wind energy resource: A case study on the Iberian Peninsula

The reduction of the error in climate model's meteorological variables representation is a key challenge to improve the reliability of future climate projections. It has special importance when analyzing wind power density (WPD) because this variable is proportional to the wind speed cubed. The first aim of this study is to determine whether bias correction improved WPD future projections from the Coordinated Regional Climate Downscalling project. With this purpose, two bias correction techniques have been applied over wind speed climatic projections. The first one was based on quantile mapping approach, while the second one was based on the correction in the frequency domain. It was found that the combination of bias correction techniques reduced biases both in terms of temporal variability and in the distribution of wind series. Regarding the sensitivity of WPD to bias correction techniques, it was detected that not corrected simulations tended to overestimate offshore wind energy in the area selected as case study, the Iberian Peninsula. Thus, a WPD reduction higher than 200 W m(-2) at an annual scale for the end of the 21st century was observed in most of the Western Iberia coastal areas when comparing the median WPD from not corrected and corrected simulations. A WPD reduction was observed for near, mid and far future by means of corrected projections, except for the northwestern corner of the Iberian Peninsula. At seasonal scale, an increase of about 20% was projected in summer, while a WPD decrease was observed in spring and, especially in autumn (20%).