Modelling Convection Forced By Temperature and Salinity Fluxes - Implications for the Ocean
We are investigating the flow forced by differential heating and cooling (buoyancy fluxes) applied along a single horizontal boundary. This can be viewed as a highly simplified model of the ocean, which nonetheless captures many of the features of the large-scale meridional overturning circulation. The physical processes driving the circulation remain a controversial topic and a general consensus on which are the crucial processes has yet to be reached. This study aims to elucidate the thermally-driven component of the flow and reveals new results which may have implications for the way in which we model the ocean and the associated climate processes.
|
Principal Investigator Ross GriffithsGeophysical Fluid Dynamics, Research School of Earth Sciences, Australian National University |
Project x52 |
|
Co-Investigator Julia MullarneyGeophysical Fluid Dynamics, Research School of Earth Sciences, Australian National University |
RFCD Codes 260403 |
Significant Achievements, Anticipated Outcomes and Future Work
The project has been progressing well. The results for numerical runs with a range of forcing strengths match both the theoretical predictions and laboratory results well and have now been included in a paper submitted to J. Fluid Mechanics. Additional runs using a temperature-dependent thermal expansion coefficient have revealed no significant difference from our earlier results. This proves to be a useful result as it provides confidence in our use of an approximation of a constant expansion coefficient in the theory. Comparing animations of the converged solution to the time-averaged fields has highlighted the unsteady nature of the flow. The flow is turbulent and contains lage eddy structures even in these two-dimensional simulations. New runs have been commenced. Firstly we have begun an investigation into the sensitivity of the flow to changes in boundary conditions. Free-stress and fixed temperature (rather than fixed heat flux) conditions have been imposed. However, some difficulties have been encountered - the solution converges to an unphysical flow pattern. Preliminary investigation reveals a new dependence on grid spacing in these runs and future work will focus on resolving this issue. Runs with the 'pole-equator-pole' model have also been initiated. In this model the forcing is applied to four sections (rather than two) of the base, with the centre and end sections corresponding to the low and high latitudes, respectively, in each hemisphere. Early results look promising, revealing that a small perturbation away from the symmetric forcing leads to highly asymmetric flow patterns. These runs will continue with a wider parameter range.
Computational Techniques Used
Use of the CFD package FLUENT
Publications, Awards and External Funding
External Funding and Awards
None.
Publications
None.