Several recent models of midlatitude climate have speculated on the role of
the North Atlantic Ocean in modulating the North Atlantic oscillation (NAO
). Here this role is examined by means of numerical experimentation with a
quasigeostrophic ocean model underneath a highly idealized atmosphere. It i
s argued that the dominant midlatitude oceanic influence is due to the so-c
alled inertial recirculations, rather than linear baroclinic waves, as have
previously been studied.
In these experiments, the forced response of the inertial recirculations do
minates the leading-order ocean spatial mode, but that mode is energized by
oceanic intrinsic variability. The oceanic signals are amplified relative
to those predicted by wave models. The primary oceanic role of the coupling
is to damp sea surface temperature (SST) at longer timescales, and the int
erdecadal atmospheric variability is placed under the control of the ocean.
The SST damping reflects competition between intrinsically driven intergyr
e heat flux and an opposing feedback-driven advective heat flux. Spectral S
ST extrema can result near the transition point where the feedback heat flu
x approaches equilibrium, although these are secondary phenomena.
The picture of midlatitude climate variability painted here has qualitative
similarities to that obtained from the linear waves models, but differs fu
ndamentally from them both dynamically and philosophically. Most important,
ocean variability is a dominant, rather than passive, partner in all aspec
ts of the coupled system.