Posted by Sabrina Speich on May 14, 2001 at 08:01:25:
In Reply to: comments on Speich et al. poster posted by Steve Rintoul on May 13, 2001 at 05:30:52:
Dear Steve,
thank you for your valuable comments. Here are
some answear/precisions on the poster.
>They find that
>the well-known "warm" and "cold" routes both contribute, but
>westward flow south of Tasmania is also significant. (They
>call this path the "tepid" path; since this word also means
>"half-hearted" or "lacking in enthusiasm," I hope we can
>find another word for it.)
The "tepid" term has been chosen only to contrast with
"cold" and "warm". With "tepid" we mean Mode Waters.
To be noted here that the we do noy identify "tepid" route
with only the TAS leakage ... as DRAKE contributes too.
Also "tepid" has the advantage to make people jump
and therefore they pay more attention to the subject.
Nevertheless, if it is really so misleading, we will change
for "temperate" or something else in the future.
Concerning the NADW exported from the Atlantic to the
Intermediate Waters passing to the three passage, we
are completing our analyses. What is clear in
models is a strong connection between NADW and upwelling
in the Southern Ocean (cf. Doos and Cowards, WOCE Newsletter, 1997).
Backward integrations of Intermediate Waters in the three
passages also end in the mixing layer in the Souther Ocean region (this
for all the three models). the problem is that
waters upwell and downwell more than once, so it
is difficult to link straightforward NADW to IW.
This part of the work is under progress.
>The trajectory technique seems to imply a
>long-term "coherence" to water parcels that seems hard to
>imagine in an eddy-rich (and at least in the model, diffusive)
>ocean. The intermediate water flowing west south of Tasmania
>is already "old" (relatively low in oxygen), and the path from
>Tasmania to the equatorial Atlantic is very long. Are the
>water properties of the equatorial Atlantic IW consistent with
>a substantial contribution of this old, oxygen-poor and
>nutrient-rich IW?
All the models simulating the actual global ocean gives
the same order of transport for the TAS water reaching
the North Atlantic. I am awear only of one simulation,
modelling tha last glacial maximum ocean circulation,
where the TAS leakage (as well as the ITFL contribution)
disappear.
Concerning water mass decomposition, the question is
under debate. The ORCA and OCCAM model (OCCAM is
eddy permitting) are more or less in agreement. GIM
model give TAS as the only TEPID contribution.
Also we are comparing more observations to try
to verify the TAS-EQATL link (collaboration with
M. Arhan and H. Mercier).
>With regard to the Lagrangian particle tracking, I have a
>few questions. Are diffusion or sub-grid-scale motions
>accounted for in some way? I gather the Gent and McWilliams
>eddy parameterisation was not used in this model: if it was,
>could the eddy-induced transport velocity be included in this
>method, and would it make much difference? Another important
>aspect of the large-scale interbasin exchange discussed here
>is the transformation of water masses by air-sea fluxes where
>layers outcrop. Is this process accounted for in the
>trajectory calculation? How about the interior restoring
>terms: the fictitious interior sources/sinks of heat and
>freshwater drive diapycnal fluxes - are these included in the
>trajectories?
The ARIANE technique doesn't account for the subgrid-scale motion
for the ORCA2 simulation. Comparisons with OCCAM, that
do have some eddy activity, are good for transport values.
We are now evaluating the impact of eddies on water mass
transformations.
Concerning the lagrangian tracking method, the issue
is addressed in Blanke et al. [1999], section 4 ("Water mass
decomposition"), subsection a ("Methodology"), equations (2) and (3).
We first assume that the monthly output of the GCM is able to sample most of
the variability simulated in the global ocean. As we deal with a coarse-
resolution model, forced with climatological winds and heat fluxes, this
assumption is fair.
We then compute three-dimensional trajectories in this monthly varying velocity
field. The way temperature (or salinity, or density) varies along these
trajectories is the signature of water mass conversions, i.e., the mean
effect of the physics parametrized in the model, whatever its origin:
- vertical mixing within the surface layer induced by air-sea interactions;
- lateral or vertical mixing in the ocean interior;
- restoring term on the Levitus' climatology (that acts to compensate for
imperfections in the model physics, or atmospheric forcing).
Water masses do transform in the model (as the atmospheric forcing does
include the effect of the wind stress and those of the atmospheric buoyancy
fluxes, and as the vertical and lateral turbulent mixing is accounted by
fairly relevant parametrizations).
Individual trajectories do exhibit tracer variations along their path, and
help us to analyze these conversions in a logical and natural way.
The restoring to Levitus does require a local input of heat or salinity (and
thus buoyancy).
Our philosophy is to consider this input as a necessary compensation for
imperfections in the model formulations (grid, physics, forcing, initial
state).
In other words, a perfect model and a perfect atmospheric forcing combined
with a perfect Levitus climatology, would lead to a simulation for which this
restoring term would be virtually equal to zero (the Levitus atlas would then
be the exact solution of the model equations !).
Consequently, we do not wish distinguish the mean effect of the true turbulent
mixing parametrized in the model from the effect of the restoring term to the
Levitus' climatology. The latter of course corresponds to a local addition or
subtraction of buoyancy, whereas the first one is a simple lateral or
vertical redistribution of heat or salt. As the simulation is in equilibrium
on a global and annual mean (the mean heat and salt fluxes at the surface
are equal to zero), the integrated effect of the Levitus restoring is also
zero over the simulation, and may be considered too as a transfer of heat or
salinity.
Thank you again!
Sabrina