A STUDY OF WIND STRESS DETERMINATION METHODS FROM A SHIP AND AN OFFSHORE TOWER

Comparisons are made between surface wind stress measurements obtained by the inertial-dissipation and direct covariance methods on a stable offshore tower and by the inertial-dissipation and bulk methods on a ship. The shipboard inertial-dissipation friction velocity measurement s agreed very well with both the tower inertial-dissipation and direct covariance values, to within +/-2% in the mean and with a 10% or lowe r rms scatter. The inertial-dissipation determinations also exhibited less scatter than the tower direct covariance measurements. A detailed error analysis indicates that shipboard inertial-dissipation wind str ess values can have an accuracy of better than 15% in near-neutral con ditions, as compared to an accuracy of roughly 30% for the bulk method . The accuracy of shipboard inertial-dissipation values was shown to b e equal to that of direct covariance measurements from a tower. Errors in inertial-dissipation wind stress values are most likely due primar ily to deviations from the assumed balance between turbulent kinetic e nergy production and dissipation and to errors in determining the wind speed variance spectra. Errors in direct covariance measurements are most likely due primarily to finite time averaging and to flow distort ion effects, unless great care is taken to minimize or correct for flo w distortion. The high accuracy of inertial-dissipation wind stress va lues found in this study, combined with the well-known difficulties in shipboard direct covariance measurements due to platform motion ana f low distortion, demonstrate that the inertial-dissipation method is th e best option at present for determining the wind stress from a ship.