Cartographie des vents dans le golfe du Saint-Laurent à l'aide de l'imagerie RADARSAT-1

Abstract

The development of wind power, both in Canada and around the world, often requires an assessment of the wind power potential in coastal regions, such as in the Gulf of St. Lawrence. Synthetic aperture radar (SAR) is becoming a more common tool for offshore wind mapping and is complementary to ground measurements and meteorological models. To test the applicability of RADARSAT-1 SAR scenes for this purpose, we use a hybrid model using the semi-empirical CMOD-IFR2 model and the polarization ratio to link wind speed to radar backscatter in dB, radar geometry, and wind direction. Wind direction is interpolated from ground wind direction measurements and co-located QuickSCAT data. This methodology gives accurate results when used with the Kirchhoff polarization ratio and ScanSAR Narrow (SNA) images. The complex geography and topography of the Gulf of St. Lawrence region do not seem to interfere with this process, as offshore wind speed accuracy is good, with a root mean square error (RMSE) of 1.30 m/s in comparison with QuickSCAT wind speeds. Closer to the coast as well, the precision of this technique in comparison with coastal ground measurements seems to be sufficiently good (RMSE of 2.12 m/s) to allow an estimate of the wind power potential in areas not covered by ground-based techniques.