Posted by Rick Lumpkin on December 03, 2001 at 13:18:45:
I enjoyed the side-by-side runs testing the impact
of various dynamics on the flow field, and they
lead to a few specific questions.
1. Flow through the Alenuihaha Channel: the contrast
is made between a linear (Munk) layer with a (too-narrow)
width set by an eddy diffusivity of approx. Ah=570 m^2/s,
and a (correct) inertial layer width set by a core speed
of 20 cm/s. Does this suggest that higher-viscosity
linear runs might be more "realistic" than the one
examined on this poster, at least with respect to the
transport of the NHRC?
This also highlights the significance of resolution
for simulating the NHRC - the model grid resolves
the Channel with something like 5 or 6 grid points
in latitude and 2 in longitude.
For nonlinear runs of varying resolution (which I
believe the authors performed), was the 2.0 Sv
through-channel flow robust, or did it vary with
resolution even at the high-res end of the runs?
Would further increased resolution - particularly
in the zonal direction of through-channel flow -
conceivably make an O(.5 Sv) difference in this
value?
2. Topographic steering influencing NHRC strength.
The authors show that, as the NEC approaches
Hawaii, a northward deflection is associated with
planetary vorticity conservation as water in the
deep layer passes over the Hawaiian Trough.
However, I can't read the anti-aliased labels in
Figure 6. The upstream latitude of bifurcation
appears to be about 20N in both runs, so that the
northward (then, as h shrinks against Hawaii, southward)
deflection appears to be a local feature which doesn't
strongly affect the volume split off into the NHRC.
What was the strength of the flow south of Hawaii
in the nonlinear, flat-bottom run of Fig.6b? Is it
about 5.5 Sv, as in the linear run of Fig.5 and Table 2?
I have a couple of additional, unrelated comments.
The abstract states "Interestingly, there is no
discernible impact of topography on the surface
ocean currents." I'm not sure what the authors mean
here - it seems directly contradicted by Figs. 4c, d
(for example). Second, is there a list of the
references? If so, please excuse me - my version of
Netscape has occasional problems!
Overall, I enjoyed this virtual poster, and agree
with the previous comments regarding variability -
a model like this one is an ideal "laboratory" for
such examinations.