Combining Satellite Winds and NWP Modelling for Wind Resource Mapping Offshore
Badger, Merete; Pena, Alfredo; Hahmann, Andrea; Hasager, Charlotte
Technical University of Denmark, DENMARK

Ocean wind fields can be retrieved from satellite-borne active microwave sensors at different temporal and spatial resolutions depending on the sensor type. Processing of both scatterometer and synthetic aperture radar (SAR) observations to wind fields relies on a geophysical model function (GMF), which relates the sea surface backscatter to winds at the standard height 10 m above the sea surface. For wind energy applications it is not sufficient to know the 10-m wind conditions because offshore wind turbines operate at heights of 80 m and beyond where the wind power potential is higher. Extrapolation of the satellite wind speed to higher levels requires information about the thermal stratification (stability) in the atmospheric boundary layer. For the wind direction, turning between the sea surface and higher levels must be considered in order to achieve the highest possible accuracy of wind resource predictions. A combination of the satellite winds with other data sources is needed to account for both of these effects. In this paper, the problem of vertical wind speed extrapolation is examined by means of high-resolution satellite winds retrieved from ENVISAT ASAR and stability information from a Numerical Weather Prediction (NWP) model. The objective is to assess the accuracy of satellite wind speeds, which have been brought to higher levels (here 100 m), through comparisons with NWP wind speeds and wind observations from an offshore research platform. Further, we wish to demonstrate the advantage, in terms of accuracy, of correcting the mean wind speed with a mean stability correction as opposed to correction of each individual wind field sample. An area around the research platform Fino-2 in the Baltic is analysed. The satellite data archive at DTU Wind Energy holds a total of 1,600 ENVISAT ASAR scenes in WSM mode over this area. The scenes are processed to 10-m wind fields using the tool ANSWRS2.0 from the Johns Hopkins University, Applied Physics Laboratory with wind direction inputs from the US Navy Operational Global Atmospheric Prediction System (NOGAPS). The model data have a spatial resolution of 1° latitude/longitude and a temporal resolution of six hours. A second set of wind fields is retrieved with wind direction inputs from higher-resolution model simulations over the region using the Weather Research and Forecast (WRF) model for the period 2006-11. These model simulations have a spatial resolution of 5 km and a temporal resolution of one hour. Peña and Hahmann (2012) demonstrated how a mean stability correction can be calculated from heat fluxes and air temperatures of WRF in very good agreement with mast observations. Here we follow a similar approach for the wind fields retrieved from SAR. A map of the 10-m mean wind speed is first calculated from the SAR wind fields. For each of the WRF simulations, the Obukhov length scale, L is calculated based on the friction velocity, the heat fluxes, and the air temperature. A check is made to ensure that 1/L derived from WRF is distributed in a similar manner for the times concurrent with the satellite data acquisitions and the entire WRF data set. A very good agreement is found between the two and both follow the theoretical distributions well. The final step in the extrapolation procedure is to calculate the stability correction and apply it to each cell of the mean wind speed map from SAR. The work is still in progress and preliminary results look promising. We note especially, that the mean stability correction from WRF is more reliable than stability correction of each individual SAR wind field. It is therefore advisable to work with the mean wind conditions when extrapolating 10-m satellite wind fields to higher levels. Acknowledgements: Satellite data from the European Space Agency. Wind observations provided by the Bundesamt für Seeschifffahrt und Hydrographie (BSH), Germany. The work was funded through the EC FP7 project Norsewind (www.norsewind.eu). References: Peña, A. and Hahmann, A.H. (2012) Atmospheric stability and turbulence fluxes at Horns Rev – an intercomparison of sonic, bulk, and WRF model data. Wind Energy, 15, 717-731. DOI 10.1002/we.500.