183
cillations in a single buoyancy model are actually the prototype of those
in models with mixed Be's. Instead of a thermal instability of the heat
transport by the northeastward currents, which is associated with deep
water formation, he argues that oscillations arise from the instability of
this current hitting a no normal flow boundary with subsequent Kelvinlike propagation of anomalies. He points out that sufficient weak damping is essential for the anomalies to propagate along the boundary, and in
models with mixed boundary conditions a surface halo cline near the northern boundary actually shields subsurface anomaly from being damped by
surface fluxes. Since Kelvin-like propagation is the only way to release
instability in fixed flux experiment, it is not clear whether this propagation is essential for the existence of oscillations unless feedbacks from this
propagation are essential to the instability of the northeastward current
associated with deep water formation.
Another type of decadal variability, completely different from that described above, is induced through the nonlinear dynamics of boundary
currents. Spall (1995) has recently indicated in a model, how the Gulf
Stream, interacting with the southward deep western boundary current
fed by the sinking of North Atlantic Deep Water, can dynamically interact
to produce decadal oscillations.
5.4 Longer Term Variability
We may be sure that our understanding of the climate system will be on
considerably firmer foundation when the task of quantitatively modeling
the paleo-record has been completed. At the disposal of the climate theorist are variability mechanisms in a broad range of time scales, from the
ten thousand year time scale of insolation changes and ice sheet dynamics
down to the interannual time scale of low frequency atmospheric variability.
The ocean is quite flexible in this regard, with time scales for relevant processes ranging from seasonal (mixed layer) to thousands of years (vertical
diffusion) .
We have already seen the record in the Greenland ice cores (Fig. 3). A
plausible hypothesis for the temperature signal recorded by the ice cores
is that it reflects the variations of surface heat loss associated with variations of intensity of North Atlantic Deep Water production. There is
evidence that the low nutrient signal of the NADW diminished during the
cold Younger Dryas period (Boyle and Keigwin, 1987) and more complete
cillations in a single buoyancy model are actually the prototype of those
in models with mixed Be's. Instead of a thermal instability of the heat
transport by the northeastward currents, which is associated with deep
water formation, he argues that oscillations arise from the instability of
this current hitting a no normal flow boundary with subsequent Kelvinlike propagation of anomalies. He points out that sufficient weak damping is essential for the anomalies to propagate along the boundary, and in
models with mixed boundary conditions a surface halo cline near the northern boundary actually shields subsurface anomaly from being damped by
surface fluxes. Since Kelvin-like propagation is the only way to release
instability in fixed flux experiment, it is not clear whether this propagation is essential for the existence of oscillations unless feedbacks from this
propagation are essential to the instability of the northeastward current
associated with deep water formation.
Another type of decadal variability, completely different from that described above, is induced through the nonlinear dynamics of boundary
currents. Spall (1995) has recently indicated in a model, how the Gulf
Stream, interacting with the southward deep western boundary current
fed by the sinking of North Atlantic Deep Water, can dynamically interact
to produce decadal oscillations.
5.4 Longer Term Variability
We may be sure that our understanding of the climate system will be on
considerably firmer foundation when the task of quantitatively modeling
the paleo-record has been completed. At the disposal of the climate theorist are variability mechanisms in a broad range of time scales, from the
ten thousand year time scale of insolation changes and ice sheet dynamics
down to the interannual time scale of low frequency atmospheric variability.
The ocean is quite flexible in this regard, with time scales for relevant processes ranging from seasonal (mixed layer) to thousands of years (vertical
diffusion) .
We have already seen the record in the Greenland ice cores (Fig. 3). A
plausible hypothesis for the temperature signal recorded by the ice cores
is that it reflects the variations of surface heat loss associated with variations of intensity of North Atlantic Deep Water production. There is
evidence that the low nutrient signal of the NADW diminished during the
cold Younger Dryas period (Boyle and Keigwin, 1987) and more complete