275
be found in Latif and Barnett (1994)).
The leading CEOF mode, accounting for about one third of the variance
in the filtered heat content data, has a period of about 20 years. Anomalies
in upper ocean heat content reconstructed from this leading CEOF mode
(Fig. 3) are displayed at intervals of about two and a half years. When the
SST anomalies are fully developed and in a stage corresponding to that
shown in Fig. 2b (8=0) the main heat content anomaly is positive and
covers the majority of the western and central Pacific. A negative anomaly
extends to the southwest from North America and increases in area and
strength as it approaches the tropics. With time, through one-half of a
cycle, the large anomalies rotate around the Pacific in a clockwise fashion
reminiscent of the general gyral circulation. Thereafter, the whole sequence
of events is repeated, but with reversed signs, and that completes one full
cycle.
This evolution is characteristic of the transient response of a midlatitude
ocean to a variable wind stress, as described in many theoretical and modeling papers (e. g. Anderson and Gill (1975), Andere::m et al. (1979), Gill
(1983)). The response is mostly baroclinic at climate time scales longer
than several months and involves the propagation of long, relatively fast
planetary waves with westward group velocity and their reflection into
short, relatively slow planetary waves with eastward group velocity. However, the mean horizontal currents will affect the wave propagation. The
net effect of this wave propagation is to modify the strength of the subtropical gyre circulation. In particular, resultant fluctuations in poleward
transport of warm tropical waters by the western boundary current lead
to the generation of SST anomalies along the path of the Kuroshio and
its extension. The spin-up time of the subtropical gyre is several years to
a decade or even longer, which accounts for the decadal time scale of the
mode under discussion.
3.3 Atmospheric response experiments
The remaining task is to explain the oscillatory nature of this mode.
Our hypothesis is that it arises from an instability of the coupled oceanatmosphere system in the North Pacific. The characteristic SST anomaly
pattern exhibits a strong meridional gradient which either reduces or enhances the meridional SST gradient normally found in the central Pacific.
Suppose the coupled system is in its reduced meridional SST gradient state
be found in Latif and Barnett (1994)).
The leading CEOF mode, accounting for about one third of the variance
in the filtered heat content data, has a period of about 20 years. Anomalies
in upper ocean heat content reconstructed from this leading CEOF mode
(Fig. 3) are displayed at intervals of about two and a half years. When the
SST anomalies are fully developed and in a stage corresponding to that
shown in Fig. 2b (8=0) the main heat content anomaly is positive and
covers the majority of the western and central Pacific. A negative anomaly
extends to the southwest from North America and increases in area and
strength as it approaches the tropics. With time, through one-half of a
cycle, the large anomalies rotate around the Pacific in a clockwise fashion
reminiscent of the general gyral circulation. Thereafter, the whole sequence
of events is repeated, but with reversed signs, and that completes one full
cycle.
This evolution is characteristic of the transient response of a midlatitude
ocean to a variable wind stress, as described in many theoretical and modeling papers (e. g. Anderson and Gill (1975), Andere::m et al. (1979), Gill
(1983)). The response is mostly baroclinic at climate time scales longer
than several months and involves the propagation of long, relatively fast
planetary waves with westward group velocity and their reflection into
short, relatively slow planetary waves with eastward group velocity. However, the mean horizontal currents will affect the wave propagation. The
net effect of this wave propagation is to modify the strength of the subtropical gyre circulation. In particular, resultant fluctuations in poleward
transport of warm tropical waters by the western boundary current lead
to the generation of SST anomalies along the path of the Kuroshio and
its extension. The spin-up time of the subtropical gyre is several years to
a decade or even longer, which accounts for the decadal time scale of the
mode under discussion.
3.3 Atmospheric response experiments
The remaining task is to explain the oscillatory nature of this mode.
Our hypothesis is that it arises from an instability of the coupled oceanatmosphere system in the North Pacific. The characteristic SST anomaly
pattern exhibits a strong meridional gradient which either reduces or enhances the meridional SST gradient normally found in the central Pacific.
Suppose the coupled system is in its reduced meridional SST gradient state