180
then breaks down and a halo cline spreads over the sinking regions. The
detailed mechanism by which the circulation revives to give an oscillation
will be dealt with below.
5.3 Decadal Variability
Weaver and Sarachik (1991a, 1991b) were among the first to recognize the
importance of F. Bryan's halo cline catastrophe experiment. They carried
out extensive studies of the transient behavior of an ocean-only model upon
a switch from the equilibrium solution obtained under restoring boundary
conditions to mixed boundary conditions. They found that the model
went through an adjustment process during which there is large transient
and chaotic variability on interdecadal to centennial time scales. After a
detailed analysis of an short time scale (decadal) oscillation cycle, they
proposed an advective mechanism: the time scale it takes an anomaly to
be advected around a limited portion of the basin is decadal. In resolving
the dilemma that transient variability was observed in some models but not
in others, they found that the existence of transient variability depends on
the characteristics of the forcing, especially the magnitude of the surface
fresh water flux (Weaver et aI, 1993): variability increased as the braking
effect of the fresh water fluxes increased relative to the driving effect of the
thermal forcing.
Transient variability of an ocean model, upon switching from restoring
to mixed boundary conditions, is usually considered in the context of multiple equilibria: the equilibrium obtained under restoring Be's is unstable
upon a switch to mixed BC's, therefore variability arises spontaneously as
an instability, after which the model settles either into another equilibrium
or stays in a chaotic state. Yin and Sarachik (1995) showed that the ocean
model with mixed BC's can also oscillate interdecadally. They analyzed
the heat and salt budget in regions where largest variability occurs and
proposed an advective and convective mechanism. Horizontal advective
heat transports from subtropical regions warm up the subsurface water
in the subpolar region and thereby enhance convection. Convection in
turn induces surface cyclonic and equatorward flows, which, together with
horizontal diffusion and surface fresh water input, advect subpolar fresh
water into convective regions to weaken or suppress convection. During an
oscillation, convection vertically homogenizes the vertical water column, increases the surface salinity, creates a larger rate of meridionally decreasing
then breaks down and a halo cline spreads over the sinking regions. The
detailed mechanism by which the circulation revives to give an oscillation
will be dealt with below.
5.3 Decadal Variability
Weaver and Sarachik (1991a, 1991b) were among the first to recognize the
importance of F. Bryan's halo cline catastrophe experiment. They carried
out extensive studies of the transient behavior of an ocean-only model upon
a switch from the equilibrium solution obtained under restoring boundary
conditions to mixed boundary conditions. They found that the model
went through an adjustment process during which there is large transient
and chaotic variability on interdecadal to centennial time scales. After a
detailed analysis of an short time scale (decadal) oscillation cycle, they
proposed an advective mechanism: the time scale it takes an anomaly to
be advected around a limited portion of the basin is decadal. In resolving
the dilemma that transient variability was observed in some models but not
in others, they found that the existence of transient variability depends on
the characteristics of the forcing, especially the magnitude of the surface
fresh water flux (Weaver et aI, 1993): variability increased as the braking
effect of the fresh water fluxes increased relative to the driving effect of the
thermal forcing.
Transient variability of an ocean model, upon switching from restoring
to mixed boundary conditions, is usually considered in the context of multiple equilibria: the equilibrium obtained under restoring Be's is unstable
upon a switch to mixed BC's, therefore variability arises spontaneously as
an instability, after which the model settles either into another equilibrium
or stays in a chaotic state. Yin and Sarachik (1995) showed that the ocean
model with mixed BC's can also oscillate interdecadally. They analyzed
the heat and salt budget in regions where largest variability occurs and
proposed an advective and convective mechanism. Horizontal advective
heat transports from subtropical regions warm up the subsurface water
in the subpolar region and thereby enhance convection. Convection in
turn induces surface cyclonic and equatorward flows, which, together with
horizontal diffusion and surface fresh water input, advect subpolar fresh
water into convective regions to weaken or suppress convection. During an
oscillation, convection vertically homogenizes the vertical water column, increases the surface salinity, creates a larger rate of meridionally decreasing