AbstractSurface wind stress is a crucial driver of upper-ocean processes, impacting air–sea gas flux, wind-wave development, and material transport. Conventional eddy covariance (EC) processing requires imposing a fixed averaging window on the wind velocity time series in order to estimate the downward flux of momentum. While this method has become the standard means of directly measuring the wind stress, the use of a fixed averaging interval inherently constrains one’s ability to resolve transient signals that may have net effects on the air–sea interactions. Here we utilize the wavelet transform to develop a new technique for directly quantifying the wind stress magnitude from the wavelet coscalogram products. The time averages of these products evaluated at the scale of maximum amplitude are highly correlated with the EC estimates (R2 = 0.99; 5-min time windows), suggesting that stress is particularly sensitive to the dominant turbulent eddies. By taking advantage of the new method’s high temporal resolution, transient wind forcing and its dominant scales may be explicitly computed and analyzed. This technique will allow for more general investigations into air–sea dynamics under nonstationary or spatially inhomogeneous conditions, such as within the nearshore region.
