Offshore wind retrievals at the hub height of wind turbine from the satellite observation in coastal waters

It is well known that microwave satellites can retrieve sea surface wind speeds at a height of 10 m above the sea level. In fact, such kind of wind information is used as initial and boundary values in simulations for daily weather forecasts. However, the wind vectors are only retrieved at 10 m height, and the application is limited. In this study, several methods are presented to use the retrieved wind speeds from satellite-borne Synthetic Aperture Radar (SAR) to convert wind speeds at 10 m height to wind speeds at tens to hundreds of meters height, and to clarify their applicability by comparing dual scanning Light Detection and Ranging（LiDAR）observations and numerical simulations, the Local forecast Model (LFM) of the Japan Meteorological Agency (JMA) and the Weather Research and Forecasting Model (WRF) around Japanese coastal waters. In particular, the hub heights required for offshore wind generation are the target of this study.

There are three target areas, Isumi, Ishikari and Saikai. The wind speeds at 10 m height retrieved from the Sentinel-1 SAR, which is operated by the European Space Agency (ESA), are converted to wind speeds to hub height using three methods based on the logarithmic wind law, the Monin–Obukhov similarity theory (MOST) and the differences (or ratios) between the wind speeds at 10 m height and those at hub heights from LFM and WRF, respectively. The results at Isumi showed that the biases of the SAR wind speed at 30 m height are 1.50 m/s and 1.27 m/s from the logarithmic law and MOST, respectively, compared to wind speeds from the LiDAR observation. The Root Mean Square Errors (RMSEs) of the SAR wind speeds are 2.97 m/s and 2.67 m/s, respectively. The results show that the correction for atmospheric stability is effective. On the other hand, the bias and RMSE when using wind differences between 10 m and 30 m heights are 0.70 m/s and 2.39 m/s, respectively. These bias and RMSE are lower than the previous two methods. It is suggested that the wind speed difference method may convert the SAR wind speed at 10 m height to hub height with higher accuracy.

This study will be further validated for other areas to determine the optimal method and its accuracy.