Figure 4: (a) Tracer concentration on the trajectory (solid line with characters in Fig. 2b) for the standard experiment (solid), and for the weak isopycnal mixing experiment (dash-dotted line). (b) Tracer balance budget on the same trajectory as Fig. 4 at the 30th year. The horizontal axis is the along-the-trajectory distance. The solid (dashed) line shows the diapycnal (isopycnal) mixing.
4. ROLES OF DIFFUSION
4-1. STEADY STATE
The tracer transported from the North Pacific is strongly affected by the diffusion on the way to the equator. To illustrate this, we show the tracer concentration (Fig. 4a) and the tracer balance budget (Fig. 4b) along a trajectory linking the North Pacific to the equator. On this trajectory, the negative tracer subducted from the large SST anomaly region in the exchange window is advected southeastward and then southwestward. However, the subducted signals decay rapidly (from point A to D of Fig. 4a) because of isopycnal mixing (Fig. 4b). within the exchange windows, the SST anomalies change signs (Fig. 1) and contain strong horizontal gradient. The tracers of opposite signs mix up and compensate each other on the long way to the Tropics and the positive value of the tracer does not extend to the equator (Fig. 2a).
The second major isopycnal mixing event takes place in a narrow band in the southwestern part of the subtropical gyre. Part of the recirculating negative tracers in the subtropical tongue are diffused toward the south, and the water going to the equator is replenished with negative tracers, as clearly shown in Fig. 4a around point F as an increase in the negative concentration. In the real ocean, baroclinic eddies are the likely agents of mixing tracers across mean flow trajectories. High eddy activity is observed by satellite altimetry along 20｡? over the southwestern subtropical gyre (Fu and Smith 1996).
