Abstract
The observed and predicted response of the bulk – skin sea surface temperature difference (δT) to changes in the wind speed and net heat flux is analyzed. Observations of δT from the northern Atlantic and tropical Pacific Oceans demonstrate that the wind speed affects δT through the net heat flux and turbulent mixing. Increased winds typically increase the net heat flux, which increases the size of δT. At the same time, increased winds cause enhanced mixing, which decreases the size of δT. To predict the net change to δT, both effects must be properly modeled. The theoretical development of existing models for δT is traced and compared. All the models can be similarly derived from surface renewal theory with their differences resulting only from the corresponding definition of the dissipation rate. The differences are manifested in the predicted dependence of δT on the wind speed. The predicted δT values and wind speed dependencies are evaluated with the available δT observations to determine the most accurate approach. A new model for δT is developed to better reproduce the observed behavior of δT. The new model follows from surface renewal theory and includes timescales for both the shear-driven and free convection regimes. The model is shown to accurately reproduce both aspects of the observed effect of wind speed on δT and predict the value of δT to better than 0.1 K.
