Satellite based estimation of atmospheric wakes downstream offhore windparks using a new objective filter technique

The large number of already existing and planned offshore wind parks in the German Bight leads to challenging requirements with regard to reliable information on various processes in the atmosphere and the ocean. In particular wind shadowing effects play a major role for the optimal planning and operation of wind park installations. Synthetic Aperture Radar (SAR) satellites have proved their capability of giving a 2D view of the wakes generated behind wind farms at a  high spatial resolution. However, the estimation of wind speed deficits from SAR data is still a challenge, because undisturbed reference wind fields are usually not available at the exact location of the wake. A common approach is therefore to identify some reference areas on SAR scenes outside the wake region, which naturally leads to errors in the deficit computations. 
In this study a new filter approach for the deficit estimation is proposed, which allows to derive error bars for the deficits. The filter is based on a 2D convolution operation with a filter kernel, which has a shape depending on the wind park geometry and the wind direction. The errors depend on spectral properties of the background wind fields, which are estimated from SAR data as well. In this context the stability of the atmospheric boundary layer is shown to play a major role. Examples are shown using data acquired by the SENTINEL-1A/B satellites. The approach is seen as a contribution to make SAR based deficit computations more objective and automised, which is essential for the application of the method to larger data sets and to make wake analysis done in different regions more comparable.