Inter-calibration of SAR data series for offshore wind resource assessment

Wind observations in the marine environment are both costly and sparse. This makes wind retrievals from satellite Synthetic Aperture Radar (SAR) an attractive option in connection with planning of offshore wind farms. Because the wind power density is proportional to the wind speed cubed, it is important to achieve the highest possible absolute accuracy on SAR wind speed retrievals for wind energy applications. A method is presented for inter-calibration of SAR observations from Envisat and Sentinel-1A/B. Sensor-specific effects on the SAR-retrieved wind speeds are first quantified through comparisons against collocated ocean buoy observations. Based on global circulation model simulations of wind speed and direction, we retrieve the Normalized Radar Cross Section (NRCS) for different radar incidence angles. Residuals between the retrieved and the observed NRCS are used to inter-calibrate the observed NRCS before reprocessing to SAR wind fields. The inter-calibration leads to an improved agreement between SAR and buoy wind speeds with biases below 0.2 m s(-1) for all investigated SAR sensors. Estimates of the wind resource improve with respect to the buoy observations for ten of the twelve sites investigated. The average deviation between wind power densities is reduced from 20% to 8% as the SAR inter calibration leads to more conservative estimates of the wind resource.