12 Oceanic Planetary Waves and Eddies
203
12.4.1 Waveguides and Normal Modes
One important dynamical problem is why there appear to be distinctive waveguides
of enhanced westward propagation in the oceans, as noted for instance at 33–34 ◦ N
in the Atlantic (Cipollini et al., 1997; Cromwell, 2001). Simulations done with a ray
tracing approach (Killworth and Blundell, 2004, 2005) do yield zonal waveguides of
enhanced propagation energy due to the convergence of many rays, but sometimes
at different latitudes from those observed in the data.
Another significant problem for ocean dynamicists is to ascertain the occurrence and relative importance of the different normal modes of propagation. The
barotropic (i.e. depth-independent) mode of planetary waves is believed to propagate too fast to be properly resolved by altimetry, especially when the data are
gridded on 10- or 7-day time steps; however Fu (2004) has been able to demonstrate a significant presence of “fast” westward-propagating barotropic energy in
some oceanic basins by using TOPEX/Poseidon data gridded on a 3-day orbital
sub-cycle. This suggests that at least part of the barotropic energy could be mapped
by adopting spatially coarse grids with time resolution of the order of 1 day or less,
something that might become feasible with a constellation of small number (<10)
of altimeters.
Maharaj et al. (2007) have looked at the significance of the different baroclinic
(i.e. variable with depth) modes in the South Pacific by splitting the energy in
wavenumber-frequency spectra of SSH anomalies on the basis of “spectral boundaries”. These boundaries are dictated by the dispersion curves for the different
modes predicted by the various theories of planetary wave propagation. The adoption of Killworth and Blundell’s extended theory in place of the classical linear
theory results in explaining up to 60% more of the variance in the observed power
spectral energy as planetary waves.
As far as the relative importance of the different modes, Maharaj et al. (2007)
found that mode 1 is by far the most important, and that mode 2 is significant in
places, while modes 3 and 4 are negligible, as shown in Fig. 12.4. The dominance
of the first baroclinic mode is evident also in a modelling study carried out over
the north Atlantic by Lecointre et al. (2008), and based on the ATL6-ERS26 1/6 ◦
simulation (Penduff et al., 2004) performed during the French CLIPPER project.
In addition, Lecointre et al. (2008) found a puzzling result that calls for further
investigation: while at the surface the model wave speeds agree reasonably well with
their counterparts observed in altimetry, below the surface the westward propagating
disturbances in the model exhibit a systematic deceleration with increasing depth,
by a factor that appears to vary geographically.
This questions the usual normal mode assumption that the speed of propagation
of the disturbances is independent of depth. A crucial contribution to a better understanding and full 3-D characterization of the modal structure of planetary waves
(and eddies) is expected from the integration of altimetric data and vertical profiles
of temperature and salinity (hence density) from the ARGO floats (Gould et al.,
203
12.4.1 Waveguides and Normal Modes
One important dynamical problem is why there appear to be distinctive waveguides
of enhanced westward propagation in the oceans, as noted for instance at 33–34 ◦ N
in the Atlantic (Cipollini et al., 1997; Cromwell, 2001). Simulations done with a ray
tracing approach (Killworth and Blundell, 2004, 2005) do yield zonal waveguides of
enhanced propagation energy due to the convergence of many rays, but sometimes
at different latitudes from those observed in the data.
Another significant problem for ocean dynamicists is to ascertain the occurrence and relative importance of the different normal modes of propagation. The
barotropic (i.e. depth-independent) mode of planetary waves is believed to propagate too fast to be properly resolved by altimetry, especially when the data are
gridded on 10- or 7-day time steps; however Fu (2004) has been able to demonstrate a significant presence of “fast” westward-propagating barotropic energy in
some oceanic basins by using TOPEX/Poseidon data gridded on a 3-day orbital
sub-cycle. This suggests that at least part of the barotropic energy could be mapped
by adopting spatially coarse grids with time resolution of the order of 1 day or less,
something that might become feasible with a constellation of small number (<10)
of altimeters.
Maharaj et al. (2007) have looked at the significance of the different baroclinic
(i.e. variable with depth) modes in the South Pacific by splitting the energy in
wavenumber-frequency spectra of SSH anomalies on the basis of “spectral boundaries”. These boundaries are dictated by the dispersion curves for the different
modes predicted by the various theories of planetary wave propagation. The adoption of Killworth and Blundell’s extended theory in place of the classical linear
theory results in explaining up to 60% more of the variance in the observed power
spectral energy as planetary waves.
As far as the relative importance of the different modes, Maharaj et al. (2007)
found that mode 1 is by far the most important, and that mode 2 is significant in
places, while modes 3 and 4 are negligible, as shown in Fig. 12.4. The dominance
of the first baroclinic mode is evident also in a modelling study carried out over
the north Atlantic by Lecointre et al. (2008), and based on the ATL6-ERS26 1/6 ◦
simulation (Penduff et al., 2004) performed during the French CLIPPER project.
In addition, Lecointre et al. (2008) found a puzzling result that calls for further
investigation: while at the surface the model wave speeds agree reasonably well with
their counterparts observed in altimetry, below the surface the westward propagating
disturbances in the model exhibit a systematic deceleration with increasing depth,
by a factor that appears to vary geographically.
This questions the usual normal mode assumption that the speed of propagation
of the disturbances is independent of depth. A crucial contribution to a better understanding and full 3-D characterization of the modal structure of planetary waves
(and eddies) is expected from the integration of altimetric data and vertical profiles
of temperature and salinity (hence density) from the ARGO floats (Gould et al.,