Reconstructing the General Circulation of the Ross Sea (Antarctica)...
123
uration of the model uses values of sea surface élévation and horizontal velocity
is set to zéro, while température and salinity values are computed according to the
T-S vertical profile [4]. We used a température linearly decreasing from the ice
melting pressure-depending one to -1.85°C at the bottom, while salinity linearly
ranges from 34.4 immediately under the ice cover to 34.8 in the bottom. The initial condition sets hâve been used in runs lasting 2 months.
In the numerical experiments a robust diagnostic intégration has been performed. This means that the prognostic values of T and S, adjusted to the velocity field, hâve been relaxed to their original values, with a relaxation time of
Y=l/r=0.1, corresponding to actual time of hours. Different values hâve been used
in order to test the sensitivity of the adjustment: large values of y almost prevented any change in the T-S distribution, while small values allow the System to
adjust better under the baroclinie forcings.
2.4 Open Boundary Velocities
In order to reproduce the forcing caused by the Ross Gyre current, the eastern
inflow and the western outflow hâve been simulated prescribing velocity values in
four grid points in the east and west zone (about 100 km). Both vertical averaged
velocities and those related to the single level hâve been imposed, using a decreasing profile.
These values intend to simulate the Ross Gyre current, having direction
west-east and entering the Ross Sea area in the eastern part following the continental shelf break, and outflowing the domain near Cape Adare. The computational points where the velocities are imposed hâve a depth of 1500 m, and are
completely ice-free.
A sensitivity analysis of the inflow points shows that the general circulation is
slightly sensitive to the forcing points spatial variation, reproducing results quite
close even if input points are moved 2-3 degrees. In these preliminary numerical
tests, the topography was expanded beyond the examiner! limits (202 °E, 70 °S),
thus creating a numerical buffer zone which allowed the local gyre to close and
bend to the output area.
Model forced only with these velocities at the open boundaries produced very
small transport values inside the Ross Sea, confirming wind to be the major forcing factor in driving the general circulation. After these numerical évidences, in
order to save computing time, the authors decided to reduce the extension of the
numerical buffer near the open boundary.
2.5 Wind Stress
A synthetic reconstruction of the prévalent summer winds has been forwarded,
in order to reproduce their direction and intensity over the whole basin. It is the
authors’ intention to analyse and use real wind data, but up until now a wind field
has been synthesized on the basis of Buzzi et al. [5]. In Fig. 3 the adopted wind
field pattern is shown; maximum velocities are about 10 m/s.
123
uration of the model uses values of sea surface élévation and horizontal velocity
is set to zéro, while température and salinity values are computed according to the
T-S vertical profile [4]. We used a température linearly decreasing from the ice
melting pressure-depending one to -1.85°C at the bottom, while salinity linearly
ranges from 34.4 immediately under the ice cover to 34.8 in the bottom. The initial condition sets hâve been used in runs lasting 2 months.
In the numerical experiments a robust diagnostic intégration has been performed. This means that the prognostic values of T and S, adjusted to the velocity field, hâve been relaxed to their original values, with a relaxation time of
Y=l/r=0.1, corresponding to actual time of hours. Different values hâve been used
in order to test the sensitivity of the adjustment: large values of y almost prevented any change in the T-S distribution, while small values allow the System to
adjust better under the baroclinie forcings.
2.4 Open Boundary Velocities
In order to reproduce the forcing caused by the Ross Gyre current, the eastern
inflow and the western outflow hâve been simulated prescribing velocity values in
four grid points in the east and west zone (about 100 km). Both vertical averaged
velocities and those related to the single level hâve been imposed, using a decreasing profile.
These values intend to simulate the Ross Gyre current, having direction
west-east and entering the Ross Sea area in the eastern part following the continental shelf break, and outflowing the domain near Cape Adare. The computational points where the velocities are imposed hâve a depth of 1500 m, and are
completely ice-free.
A sensitivity analysis of the inflow points shows that the general circulation is
slightly sensitive to the forcing points spatial variation, reproducing results quite
close even if input points are moved 2-3 degrees. In these preliminary numerical
tests, the topography was expanded beyond the examiner! limits (202 °E, 70 °S),
thus creating a numerical buffer zone which allowed the local gyre to close and
bend to the output area.
Model forced only with these velocities at the open boundaries produced very
small transport values inside the Ross Sea, confirming wind to be the major forcing factor in driving the general circulation. After these numerical évidences, in
order to save computing time, the authors decided to reduce the extension of the
numerical buffer near the open boundary.
2.5 Wind Stress
A synthetic reconstruction of the prévalent summer winds has been forwarded,
in order to reproduce their direction and intensity over the whole basin. It is the
authors’ intention to analyse and use real wind data, but up until now a wind field
has been synthesized on the basis of Buzzi et al. [5]. In Fig. 3 the adopted wind
field pattern is shown; maximum velocities are about 10 m/s.