Form of the Temperature-Salinity Relationship in the Central Water: Evidence for Double-Diffusive Mixing

Abstract

Ingham (1966) reported that the temperature-salinity relationships in the Central Waters were much better described by a curve of constant density ratio (R? = α?T/??S) than by a straight line. His result is quantitatively verified and a simple, but powerful, double-diffusive mechanism is proposed to explain the observed constancy of R? in the main thermocline. The mechanism is based on the evidence from theory, experiment and observation that the intensity of salt-finger convection is a strong function of R?. This dependence, plus the fact that more salt than heat is transferred by the fingers, causes any deviation from a constant R? to be the site of convergence or divergence of the vertical salt flux that acts to remove the perturbation in R?. A linear treatment of the mechanism shows that R? can be ?diffused? with an effective diffusivity that is much greater than the diffusivities of heat or mass. A few numerical examples illustrate the predicted effects of salt fingering on the T-S relation, showing that a constant R? is the basic state of a fluid in which some salt fingering is taking place. The model suggests that the large scale T-S relation may be controlled as much by the details of the microscale diffusive processes as by the large-scale atmospheric forcing.