Abstract
It is pointed out that accounting for an ocean surface velocity dependence in the wind stress τ can lead to a significant reduction in the rate at which winds input mechanical energy to the geostrophic circulation. Specifically, the wind stress is taken to be a quadratic function of Ua − uo, where Ua and uo are the 10-m wind and ocean surface velocity, respectively. Because |Ua| is typically large relative to |uo|, accounting for a uo dependence leads only to relatively small changes in τ. The change to the basin-averaged wind power source, however, is considerably larger. Scaling arguments and quasigeostrophic simulations in a basin setting are presented. They suggest that the power source (or rate of energy input) is reduced by roughly 20%–35%.