The Kuroshio is known to take a distinct path depending on its volume transport; a straight path and a meandering path. Traditionally, it was believed that the Kuroshio would take a straight path at low volume transport, a meandering path at high volume transport, and a bimodal path at an intermediate volume transport. When the volume transport changes in time, the Kuroshio was thought to transition from one state to the other, but with a lag in the timing between the increasing and the decreasing transport cases; thus a hysteresis is formed.
In our study using high-resolution regional OGCM with realistic topographic features, we have seen that even at a fixed transport, the Kuroshio can switch between straight and meandering path. Thus, the bimodality is realized in a probabilistic sense. The distribution of the mean Kuroshio axis position shows double peaks at 35 Sv inflow volume transport whereas at lower transport of 25 Sv, only a single peak exists. This is quite different from the traditional picture, and it is important to point out that the probabilistic view suggests that even at a fixed inflow transport, the switching of the states occur. Which indicates also, that external perturbation can easily trigger a transition from one state to the other at a fixed inflow transport.
The robustness of this probabilistic view was tested by applying various winds forcing to the 35 Sv inflow case. The following wind forcing was used: i) Hellerman-Rosenstein annual mean wind; ii) QuikSCAT-derived daily wind; iii) QuikSCAT-derived monthly wind; iv) ECMWF operational model daily wind; v) no wind forcing. Comparing i), ii) and iii), we have proved that the switching between the two states can be triggered by high-frequency local wind forcing (external forcing). On the other hand, by comparing i), iv) and v), we have discovered that the presence of a well-established recirculation south of the Shikoku Island is necessary for realizing a double peaked probability density function of the mean Kuroshio distance.
In summary, we suggest that the number of relevant control parameters in determining the Kuroshio path state is more than one (not just the volume transport), and possibly includes the strength of the Shikoku recirculation gyre. By the time of the meeting, this hypothesis will be tested by conducting more experiments at different inflow Kuroshio transport rates. |
|
|