Microstructure in and around a Double-Diffusive Interface

M. Fleury The Johns Hopkins University, Chesapeake Bay Institute, Baltimore, Maryland

Search for other papers by M. Fleury in
Current site
Google Scholar
PubMed
Close
and
R. G. Lueck The Johns Hopkins University, Chesapeake Bay Institute, Baltimore, Maryland

Search for other papers by R. G. Lueck in
Current site
Google Scholar
PubMed
Close
Restricted access

Abstract

The microstructure of temperature and velocity were measured for 35 km with a towed vehicle that moved in and around a single interface in a well-developed thermohaline staircase with a density ratio of 1.5. Although the microstructure was the result of double diffusion of the “salt-finger” type, there was no evidence of cells or sheets. The temperature gradient was quasi-isotropic and reminiscent of turbulence rather than strongly anisotropic. The ratio of horizontal to vertical temperature-gradient variance was 1.27 and 0.83 in the interface and mixed layers, respectively, for scales between 0.035 and 1 m. The temperature field was decorrelated over a vertical distance of 0.035 m, approximately the predicted width of cells and sheets. The spectral slope and peak wavenumber were similar in the interface and the mixed layers, and these features did not clearly distinguish turbulence from double diffusion. The spectral peak was at 15% of the Batchelor wavenumber in both regions. The spectra of ∂w/∂x and ∂T/∂x were not similar, the spectral level of ∂w/∂x was 10 times smaller than predicted, and the spectra of ∂T/∂x did not scale with the local mean vertical temperature gradient as predicted for vertical cells. The structure of the interface was far from the assumption of models, and the agreement between the predicted and observed average fluxes of heat, salt, and buoyancy is not sufficient to prove the adequacy of current models.

Other features of the interface were large-scale plumes, density-compensated inversions, and strong layering. The edges frequently showed billows and filaments entering the mixed layers.

Abstract

The microstructure of temperature and velocity were measured for 35 km with a towed vehicle that moved in and around a single interface in a well-developed thermohaline staircase with a density ratio of 1.5. Although the microstructure was the result of double diffusion of the “salt-finger” type, there was no evidence of cells or sheets. The temperature gradient was quasi-isotropic and reminiscent of turbulence rather than strongly anisotropic. The ratio of horizontal to vertical temperature-gradient variance was 1.27 and 0.83 in the interface and mixed layers, respectively, for scales between 0.035 and 1 m. The temperature field was decorrelated over a vertical distance of 0.035 m, approximately the predicted width of cells and sheets. The spectral slope and peak wavenumber were similar in the interface and the mixed layers, and these features did not clearly distinguish turbulence from double diffusion. The spectral peak was at 15% of the Batchelor wavenumber in both regions. The spectra of ∂w/∂x and ∂T/∂x were not similar, the spectral level of ∂w/∂x was 10 times smaller than predicted, and the spectra of ∂T/∂x did not scale with the local mean vertical temperature gradient as predicted for vertical cells. The structure of the interface was far from the assumption of models, and the agreement between the predicted and observed average fluxes of heat, salt, and buoyancy is not sufficient to prove the adequacy of current models.

Other features of the interface were large-scale plumes, density-compensated inversions, and strong layering. The edges frequently showed billows and filaments entering the mixed layers.

Save