Figure 7: Tracer concentration (solid) and density (dashed line) on zonal sections at (a) 30.5, (b) 19.5°N. Negative tracer values are shaded. Contour intervals are 0.125 at 19.5°N and 0.25 at 30.5°N, and are 0.50 for σθ.
The tracer balance at 5th year on the same trajectory (Fig. 5b) confirms that it is the isopycnal mixing that supplies the tracer around point F, the same replenishment shown in the steady state solution. At the 5th year, although the tracer subducted from the sea surface within the exchange window is advected to only around point C (Fig. 5a, b), the tracer advected along the recirculating route has moved to the north of the points E and F (Fig. 6) and is mixed isopycnally to the trajectory that leads eventually to the equatorial region. Thus, the isopycnal mixing replenishes the tracer around point F before the subducted tracer arrives, and provides a short cut to the equatorial region, reducing the journey from the North Pacific to the equator by 5 years compared with a purely advective route.
5. DISCUSSION AND CONCLUSION
In a realistic OGCM, the equatorward spreading of the tracer corresponding to the North Pacific decadal/interdecadal SST anomaly through the subduction-ventilation process has been examined. Here, we compare the simulated passive tracer fields with the observed interdecadal temperature anomalies. The meridional section of the tracer at 30.5°N (Fig. 7a) shows the eastern shallow positive tracers and the central deep negative tracers tilting east-downward reminiscent, of the observed temperature difference between two decades of 1976-85 and 1966-75 (Yasuda and Hanawa 1997). At 19.5°N where tracer is restored to zero at the surface. the negative tracers submerge into the subsurface and move southwestward with the core staying around σθ=25.5 (Fig. 7b), an isopycnal surface on which Deser et al. (1996) observed the migration of the negative temperature anomaly core during 1977-1991.
