Comparison of Sea Surface Wind Measurements Obtained from Buoy, Aircraft and Onshore Masts during the TOSCANE T Campaign

Abstract

The first European Remote Sensing satellite, known as ERS1, will be launched by 1989. Priority in the payload has been given to a comprehensive set of radar instruments designed to observe the surface wind and wave structure over the oceans: scatterometer, altimeter and synthetic aperture radar. The aim of the TOSCANE program is to contribute to the calibration and validation of the ERS1 C-band wind scatterometer. The accurate measurement of surface wind over the ocean is a problem in itself, but even if perfect surface measurements are done, the question arises whether time averaged wind measurements at a single point adequately represent the instantaneous spatially averaged satellite values. During the TOSCANE T experiment winds were simultaneously recorded from a network of coastal masts, a Marisonde buoy moored 6 km offshore, and a Piper Aztic aircraft flying at a nominal altitude of 100 m. Wind sensors were at 10 m on the masts and at 3 m on the buoy, Buoy and aircraft wind speeds were adjusted to the mast altitude (i.e., 10 m) using a constant-flux surface boundary model. Data analyses point out the large influence of local environment on coastal winds. Comparisons of the three kinds of wind measurements, however, reasonably concur. Between masts and buoy, mean differences (mast minus buoy) reach 0.4 m s?1 for speed and ?9° for direction. Truncation of the buoy data to the knot and to the 36-direction compass card account for almost 60% of these differences. Corresponding standard deviations are 0.8 m s?1 and 9°. Wind speed mean differences do not show any correlation with wave heights. Aircraft and mast mean differences (aircraft minus mast) are ?-0.4 m s?1 for speed and 7° for direction with standard deviations of 0.9 m s?1 and 10°. Comparison between aircraft and mast emphasizes that under inhomogeneous meteorological conditions, including convective clouds, great care must be taken in data analysis.