Seasonal velocity variations over the entire Kuroshio path part I: data analysis and numerical experiments

Abstract

Herein, seasonal velocity variations from the sea surface to a depth of 1000 m over the entire Kuroshio path are investigated using satellite altimetry and reanalysis datasets. The data analysis results show that velocities in the upper layer (from 0 to approximately 500 m) reach a maximum in July and a minimum in autumn (October to November) or winter (December to February) with different tendencies in each region. However, those in the lower layer (> 500 m depth) show a reversed seasonal variation—reaching a maximum in winter—especially in the continental slope area from the east of Luzon Island to the east of the Ryukyu Islands chain, which is regarded as a route of the deeper Kuroshio flow. Using a realistic general circulation model, we performed numerical experiments to clarify the role of the local wind stress as the driving force in seasonal Kuroshio velocity variations in the upper layer. These experiments revealed that seasonal Kuroshio velocity variations in the upper layer are mainly caused by the local response to wind stress upon the current itself. These numerical results cannot be explained by conventional mechanisms, such as flow–topography interactions or coastal upwelling/downwelling. On the other hand, seasonal Kuroshio velocity variations in the lower layer can be explained by the Sverdrup theory, in which barotropic responses to the wind stress curl over the area west of the Izu–Ogasawara Ridge are responsible.

Appendix

1.1 Validation for the subsurface velocity field from the MOVE-WNP reanalysis data

Usui et al. (2013) validated the MOVE-WNP reanalysis data using the repeat hydrographic data on PN sections in the ECS and found good consistency in the interannual Kuroshio volume transport variation between the two datasets. However, the validation of the subsurface velocity field east of the Ryukyu Islands chain from their reanalysis data has not been conducted. Here, we compare the vertical velocity sections east of the Ryukyu Islands chain (Fig. 16) between the MOVE-WNP reanalysis (Fig. 17) and the observed data (Fig. 18). The observational velocity sections shown in Fig. 18 are a copy of Fig. 7 in Ichikawa et al. (2004), who produced those sections by combining data from moored current meters deployed at the positions shown by open circles in each panel and along-track satellite altimetry sea-surface height data (see their paper for the exact method). The MOVE-WNP reanalysis data well reproduced the current core at a depth of approximately around 500 m depth in the Ryukyu Current system. In addition, the MOVE-WNP reanalysis data indicate a year-to-year velocity variation with a maximum in 1999 and a minimum in 2001, which was found in the observational velocity sections (Fig. 18). On the other hand, the seasonal variations shown in Fig. 5 are consistent with the results of a previous study. Figure 19 shows a monthly time series of volume transport averaged over the area east of the Ryukyu Islands chain, reconstructed from Fig. 5b using the MOVE-WNP reanalysis data. The amplitude of seasonal volume transport variation was approximately 11 Sv and the maximum appeared in winter. These features are similar to the results shown by Thoppil et al. (2016, see Fig. 19 in their paper). Therefore, it is considered that the MOVE-WNP reanalysis data are reliable for the seasonal velocity variation analysis not only for the upper layer Kuroshio but also for the lower layer Kuroshio including the Ryukyu Current system.