It is also revealed that a slow phase speed and a large amplitude of the meander are regarded as conditions for the stationary large meander, which is also applicable to the past large meanders. The sensitivity analysis reveals that the small meander and the intensified anticyclonic eddy significantly affect the phase speed and the amplitude of the meander, respectively. The eddy energetic analysis indicates that baroclinic instability is essentially responsible for this large meander formation.

Then a southward flow in the eastern part of the eddy carries the Kuroshio path offshore, bringing about the large meander. The deep-anticyclonic eddy grows up during its eastward propagation. At that time, a deep-anticyclonic eddy is generated below the Kuroshio by a cross-frontal flow. The small meander starts moving eastward around May 2004, when an anticyclonic eddy east of the Tokara Strait is intensified by coalescing of two anticyclonic eddies. This balance is achieved as a result of reducing the advection by a westward propagating negative-temperature-anomaly originating in the wind stress field.

The vorticity analysis indicates that this small meander is maintained by a balance between the advection and the bottom pressure torque on the continental slope. The small meander southeast of Kyushu that occurred in December 2003 remains stationary until the next spring. Formation process of the 2004 Kuroshio large meander has been examined using a data assimilation and prediction system.

Meander ending explained