283
assume that the atmosphere has no internal dynamics and responds passively to the oceanic changes. Since our ocean model carries no sea surface
temperature, we parameterize the SST variations in terms of the strength
of the western boundary current which is taken to be proportional to the
streamfunction difference across the two gyres near the eastern boundary.
The actual wind stress curl is given by:
\7 x T.. = a· g(q) . \7 x I, q = [(8'ljJj8y)wj(8tfjj8y)w]
(3)
Bars denote in (3) quantities that were derived from the steady state
solution of the uncoupled system. The subscript "W" indicates that the
pressure gradients are computed near the western boundary. The coupling strength is given by the parameter 0:. In order to describe the nonlinear equilibration of the growth, we assume a function g( q) proportinal to
[qj(l +'Y·lql)], with l' a parameter controling the degree ofthe non-linearity.
The equations are solved numerically on a 2° x 2° grid. Preliminary
results indicate that such a simple coupled model oscillates at decadal time
scales at sufficiently high values of the coupling strength 0:. An example
is shown in Fig. 8 which shows the normalized streamfunction anomaly
in the center of the subtropical gyre as function of longitude and time
for a rectangular basin that extends 50° in the meridional and 100° in
zonal direction. The other parameters are given in the caption of Fig. 8.
Although the oscillation is weakly damped for the set of parameters chosen,
a clear periodicity of the order of 30 years can be readily seen. Thus, the
results of our simple coupled model support our hypothesis that decadal
variability in mid-latitudes can originate from unstable air-sea interactions
in mid-latitudes themselves. A complete investigation of the sensitivity of
our simple coupled model is underway and will be described by Miinnich
et al. (in preparation).
5 The decadal variability in the North Atlantic
The type of decadal variability described above can exist in principal also in
the North Atlantic. Indeed, observations show some evidence of a decadal
cycle in the North Atlantic. This is shown in Fig. 9 which displays observations of subsurface temperature anomalies in the North Atlantic at 125 m
depth (Levitus et al. (1994)). The subsurface temperature measurements
show some remarkable oscillatory behaviour during the last few decades,
assume that the atmosphere has no internal dynamics and responds passively to the oceanic changes. Since our ocean model carries no sea surface
temperature, we parameterize the SST variations in terms of the strength
of the western boundary current which is taken to be proportional to the
streamfunction difference across the two gyres near the eastern boundary.
The actual wind stress curl is given by:
\7 x T.. = a· g(q) . \7 x I, q = [(8'ljJj8y)wj(8tfjj8y)w]
(3)
Bars denote in (3) quantities that were derived from the steady state
solution of the uncoupled system. The subscript "W" indicates that the
pressure gradients are computed near the western boundary. The coupling strength is given by the parameter 0:. In order to describe the nonlinear equilibration of the growth, we assume a function g( q) proportinal to
[qj(l +'Y·lql)], with l' a parameter controling the degree ofthe non-linearity.
The equations are solved numerically on a 2° x 2° grid. Preliminary
results indicate that such a simple coupled model oscillates at decadal time
scales at sufficiently high values of the coupling strength 0:. An example
is shown in Fig. 8 which shows the normalized streamfunction anomaly
in the center of the subtropical gyre as function of longitude and time
for a rectangular basin that extends 50° in the meridional and 100° in
zonal direction. The other parameters are given in the caption of Fig. 8.
Although the oscillation is weakly damped for the set of parameters chosen,
a clear periodicity of the order of 30 years can be readily seen. Thus, the
results of our simple coupled model support our hypothesis that decadal
variability in mid-latitudes can originate from unstable air-sea interactions
in mid-latitudes themselves. A complete investigation of the sensitivity of
our simple coupled model is underway and will be described by Miinnich
et al. (in preparation).
5 The decadal variability in the North Atlantic
The type of decadal variability described above can exist in principal also in
the North Atlantic. Indeed, observations show some evidence of a decadal
cycle in the North Atlantic. This is shown in Fig. 9 which displays observations of subsurface temperature anomalies in the North Atlantic at 125 m
depth (Levitus et al. (1994)). The subsurface temperature measurements
show some remarkable oscillatory behaviour during the last few decades,