Temperature-Salinity Oscillations, Sudden Transitions and Hysteresis in Laboratory Experiments and Layered Models

I review a number of simplified box models of the cooling of a salt-stratified ocean that we have constructed in the laboratory with a number of collaborators through the past five years. In the most recent one, a large isothermal basin of water has two layers with differing salinity. Beside this is a small basin connected to the large basin by horizontal tubes at the top, middle and bottom. Calculations indicate that there is a sudden transition and hysteresis between a shallow and a deep convection state if there is a relaxation temperature boundary condition and also if one tube has large flow resistance. Our laboratory studies to date do not clearly show hysteresis but have relatively sudden changes in properties for some parameters. The shallow state is frequently found as an oscillation, and the deep convection state is steady, although thermals produce small rapid fluctuations. Numerical models of the experiments produce qualitative agreement, but quantitative differences are large. In contrast, experiments with a cavity at the bottom of a fresh water reservoir, subjected to steady heating from below and steady salt-water inflow has two distinct states, and exhibits a hysteresis range. Oscillations and transitions like those seen in these experiments may exist in natural bodies with a layer of fresh water cooled from above such as fjords, polar bays, or larger polar regions. The oscillation periods are much greater than either the fresh water or the thermal time scale, making the oscillation mechanism a candidate for climate oscillations.