Decadal time scale climate variability in the North Pacific has implication
s for climate both locally and over North America. A crucial question is th
e degree to which this variability arises from coupled ocean/atmosphere int
eractions over the North Pacific that involve ocean dynamics, as opposed to
either purely thermodynamic effects of the oceanic mixed layer integrating
in situ the stochastic atmospheric forcing, or the teleconnected response
to tropical variability. The part of the variability that is coming from lo
cal coupled ocean/atmosphere interactions involving ocean dynamics is poten
tially predictable by an ocean/atmosphere general circulation model (O/A GC
M), and such predictions could (depending on the achievable lead time) have
distinct societal benefits. This question is examined using the results of
fully coupled O/A GCMs, as well as targeted numerical experiments with sta
nd-alone ocean and atmosphere models individually. It is found that coupled
ocean/atmosphere interactions that involve ocean dynamics are important to
determining the strength and frequency of a decadal-time scale peak in the
spectra of several oceanic variables in the Kuroshio extension region off
Japan. Local stochastic atmospheric heat nux forcing, integrated by the oce
anic mixed layer into a red spectrum, provides a noise background from whic
h the signal must be extracted. Although teleconnected ENSO responses influ
ence the North Pacific in the 2-7 years/cycle frequency band, it is shown t
hat some decadal-time scale processes in the North Pacific proceed without
ENSO. Likewise, although the effects of stochastic atmospheric forcing on o
cean dynamics are discernible, a feedback path from the ocean to the atmosp
here is suggested by the results.