The average rates at which and depths to which convection occurs in the polar oceans sets the strength of the so-called 'thermohaline circulation' (THC); an equatorward transport of cold dense waters. In the North Atlantic, this transport is partially compensated by the familiar poleward return flow of warmer surface waters known as the Gulf Stream. This poleward transport of heat in the upper ocean plays an important role in the Earth's heat budget - being responsible, for example, for between a third and half of the total ocean-atmosphere poleward heat transport in the northern hemisphere.
The distribution of 'tracers', such as temperature, salinity, nutrient
and oxygen concentrations, CFCs, and various radioisotopes in the ocean
have been recently measured on a global scale during the WOCE (World Ocean
Circulation Experiment) program. The large-scale distributions of such tracers, along with knowledge of average surface-ocean concentrations, can be used to estimate the average rates at which water convectively sinks in the polar oceans.
From the measured concentrations of four tracers in the ocean whose sources and sinks are negligible or well-known, mass and tracer-budgets for sub-regions of ocean which are remarkably homogeneous in their tracer properties are enforced to estimate the combination of advective and diffusive fluxes which best reproduce the observed data. This nonlinear inverse problem is solved using the algorithm proposed by Tarantola and Valette (1982).