THE NEAR-SURFACE LAYER OF THE OCEAN
For
1/ 4
Ri t
, but well below the inviscid criterion (5.65), the system
appears to be stable at the Langmuir number of 0.01 used in the calculations.
Leibovich and Lele (1982) found that sufficiently thick preexisting
thermoclines with sufficiently strong temperature gradients act as an
impenetrable “bottom” for the induced circulations. This is in accord with
Langmuir’s (1938) ideas.
b) Large Eddy Simulation (LES)
One problem of DNS is that the maximum Reynolds number is limited
in principle. This complicates the application of DNS to geophysical
situations. LES is able to handle typical geophysical situations with larger
Reynolds number than is possible with DNS. Skyllingstad and Denbo (1995)
performed LES studies of Langmuir circulations under a variety of
conditions. These conditions included wind- and convection-driven mixing
with and without Stokes drift to highlight their importance in the structure of
the upper layers. They indicated the limited effect of surface heating and
cooling on the near-surface structure of Langmuir circulations during their
simulation. Cases with the wind- and wave-forced simulations (with or
without cooling or heating) reveal the organized structure; conversely, runs
without surface wave forcing do not reveal this type of organization.
Remarkably, the characteristic elongated structures associated with
Langmuir cells were evident in the near-surface layer only but not at greater
depths in the upper ocean mixed layer. The enhancement of vertical velocity
variance and the increased entrainment heat flux in the case with Stokes drift
relative to the case with no Stokes drift led Skyllingstad and Denbo (1995)
to the conclusion that Langmuir circulations might be important to the
dynamics of the upper ocean mixed layer.
McWilliams et al. (1997) added Coriolis and pressure terms to the mean
momentum equation for Stokes drift that was ignored by Skyllingstad and
Denbo (1995). McWilliams et al. (1997) performed LES studies with the
improved model formulation and demonstrated an elevated TKE and
dissipation rate of TKE in the mixed layer, especially in the upper part, as
well as a significant increase in eddy viscosity. The velocity profiles in the
Ekman layer were also more homogeneous in the case with Langmuir
circulations.
LES results show significant increase of the dissipation rate in the upper
ocean mixed layer, which is in contradiction to the concept of selforganization. Organization should not increase, but rather reduce, chaos and
thus reduce dissipation in the system. LES might then not be completely
appropriate to the modeling of Langmuir circulations. LES models are based
on the idea of separation between large and small-scale motions (Deardorff,
382
For
1/ 4
Ri t
, but well below the inviscid criterion (5.65), the system
appears to be stable at the Langmuir number of 0.01 used in the calculations.
Leibovich and Lele (1982) found that sufficiently thick preexisting
thermoclines with sufficiently strong temperature gradients act as an
impenetrable “bottom” for the induced circulations. This is in accord with
Langmuir’s (1938) ideas.
b) Large Eddy Simulation (LES)
One problem of DNS is that the maximum Reynolds number is limited
in principle. This complicates the application of DNS to geophysical
situations. LES is able to handle typical geophysical situations with larger
Reynolds number than is possible with DNS. Skyllingstad and Denbo (1995)
performed LES studies of Langmuir circulations under a variety of
conditions. These conditions included wind- and convection-driven mixing
with and without Stokes drift to highlight their importance in the structure of
the upper layers. They indicated the limited effect of surface heating and
cooling on the near-surface structure of Langmuir circulations during their
simulation. Cases with the wind- and wave-forced simulations (with or
without cooling or heating) reveal the organized structure; conversely, runs
without surface wave forcing do not reveal this type of organization.
Remarkably, the characteristic elongated structures associated with
Langmuir cells were evident in the near-surface layer only but not at greater
depths in the upper ocean mixed layer. The enhancement of vertical velocity
variance and the increased entrainment heat flux in the case with Stokes drift
relative to the case with no Stokes drift led Skyllingstad and Denbo (1995)
to the conclusion that Langmuir circulations might be important to the
dynamics of the upper ocean mixed layer.
McWilliams et al. (1997) added Coriolis and pressure terms to the mean
momentum equation for Stokes drift that was ignored by Skyllingstad and
Denbo (1995). McWilliams et al. (1997) performed LES studies with the
improved model formulation and demonstrated an elevated TKE and
dissipation rate of TKE in the mixed layer, especially in the upper part, as
well as a significant increase in eddy viscosity. The velocity profiles in the
Ekman layer were also more homogeneous in the case with Langmuir
circulations.
LES results show significant increase of the dissipation rate in the upper
ocean mixed layer, which is in contradiction to the concept of selforganization. Organization should not increase, but rather reduce, chaos and
thus reduce dissipation in the system. LES might then not be completely
appropriate to the modeling of Langmuir circulations. LES models are based
on the idea of separation between large and small-scale motions (Deardorff,
382