During the second survey of the Travaux d'Oc¿anographic Spatiale: Capteurs actifs dans l'Atlantique Nord-Est (TOSCANE T) experiment (February 14 to April 17, 1985), seven wind masts were operated on the shore of the ''Baie d'Audierne.'' Distances between them ranged from 1.5 to 13.7 km, and the data were sampled at 3 s. An important portion of the data was recorded under severe weather conditions. Results from 27 cases of wind blowing from the sea, which corresponded to synoptically stationary wind regimes, show that for both horizontal components the spectral energy in the transition region between mesoscale and Komogorov turbulence takes the shape of a well-marked dip when weather types are stable or slightly unstable. But, in more convective cases the dip disappears and the transition region becomes almost horizontal: spectral energy density follows an n-1 law (where n is equal to frequency) until the Kolmogorov region is reached. Coherences and cross correlations between masts show that in the 6 s to 1-hour period range, only mesoscale fluctuations are coherent. Turbulent fluctuations are not correlated for the separation distances of the masts. Under synoptically steady of slightly unstable conditions, such single-point measurements could reliably be time-averaged for use in satellite wind sensor calibration. In more convective conditions, especially for the ubiquitous open mesoscale cells found over mid-latitude oceans in cold air advections, interpretation problems might occur because mesoscale events, as time-averaged from coastal masts, buoys, or ships, could be different from those spatially integrated in the footprint of a satellite sensor. In these cases, some relationship must be used to relate single-point averaging times to the area illuminated by the satellite. To do so, Taylor's hypothesis is commonly extended to the mesoscale; but, the present data show that such an extension cannot be made under usual actual conditions because of the structure of the wind profile in the atmospheric layers concerned. ¿ American Geophysical Union 1989