Simulating interdecadal variation of the thermohaline circulation by assimilating time-dependent surface data into an ocean climate model

Li, Guoqing (1994) Simulating interdecadal variation of the thermohaline circulation by assimilating time-dependent surface data into an ocean climate model. Masters thesis, Memorial University of Newfoundland.

[English] PDF (Migrated (PDF/A Conversion) from original format: (application/pdf)) - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (3MB) [English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. (Original Version)

Abstract

We explore the feasibility of simulating interdecadal variations of the temperature, salinity and thermohaline circulation in the North Atlantic using an ocean general circulation model (OGCM) driven by time-dependent surface data. The natural way to drive the ocean is to use the surface heat and freshwater fluxes. In this thesis, we investigate the alternative of using surface temperature and salinity data since compared to heat and freshwater flux data, they are more accurate and more readily available. -- We do the experiments using idealized North Atlantic sized box geometry. In order to obtain a set of interdecadally-varying data, we first reproduced the results described by Zhang, Greatbatch and Lin (1993). Temperature, salinity, surface heat and freshwater flux are output from this control run and serve as "observations" in the further experiments. -- We can apply either a restoring boundary condition or a flux boundary condition at the ocean surface. To simulate the interdecadal variations, there are four choices for the surface boundary conditions: i) flux conditions on both temperature and salinity; ii) restoring conditions on both temperature and salinity; iii) a restoring condition on temperature and a flux condition on salinity (mixed boundary conditions); and iv) a flux condition on temperature and a restoring condition on salinity ("reversed mixed" boundary conditions). The restoring boundary conditions are to be understood in a sense of data assimilation. -- The experiments show that all the choices work well except mixed boundary conditions. It is found that a correct simulation of the thermohaline circulation is necessary to obtain a correct distribution of the sub-surface variables. Under mixed boundary conditions, a positive feedback between the development of a freshwater cap and heat-loss reduction results in either a collapsed or violent overturning thermohaline circulation quite unlike what happens in the control run. So mixed boundary conditions are not suitable for interdecadal simulation studies. The use of both-restoring boundary conditions does not allow a freshwater cap to develop because the surface salinity is constrained by the use of the restoring condition on salinity. This guarantees a realistic thermohaline circulation and, consequently, a correct distribution of the sub-surface temperature and salinity. In contrast with mixed boundary conditions, "reversed mixed" boundary conditions work well. This is because both surface heat flux and surface salinity are controlled, breaking the positive feedback that occurs under mixed boundary conditions. -- A common feature of the successful experiments is that all of them have a correct surface heat flux distribution (prescribed or implied), whereas the surface salt flux may defer (though not significantly). This shows that whatever the boundary condition is, it is necessary to get the surface heat flux seen by the ocean correct in order to have a realistic thermohaline circulation. It also shows that controlling surface salinity is more important than getting every detail of the freshwater flux correct (fortunately, freshwater flux data are the poorest). The results show that we should use both-restoring boundary conditions or reversed mixed boundary conditions to simulate interdecadal variations of the thermohaline circulation.

Actions (login required)

View Item