The flow field east of the Ridge is not yet known well. Further observational studies are needed for estimating the transport there.
Seasonality of the transport of the Kuroshio system is somewhat complicated. The seasonal signal in the surface transport and total transport of the Kuroshio is small in the East China Sea and south of Japan, where the Kuroshio flows along the western boundary, compared with the seasonal fluctuation of the interior Sverdrup transport (10, 12). For the Kuroshio Extension, after leaving the western boundary, the seasonal signal of its surface transport is somewhat apparent, according to recent altimetry data analyses (13, 14), although some other studies show that the seasonal signal disappears when the transport associated with the recirculation on the southern side is removed (e.g., 15). The Izu-Ogasawara Ridge, located at about 140｡?, may play an important role in the seasonal fluctuation of the Kuroshio transport. Recent numerical models show that the joint effect of baroclinicity and bottom relief is important in the seasonal fluctuation of the western boundary current (16). Further studies need to be done on the seasonality of the Kuroshio transport.
The transport of the Kuroshio below the 1,000m depth is estimated to be negligibly small. The transport between depths of 1,000m and 2,000m under the upper layer Kuroshio is estimated approximately using velocities observed by moored current meters at a nominal depth of 1,500m. It is found that the fluctuation of transport of this layer is not significantly correlated with that of the upper layer transport and does not have any apparent seasonal signal. The mean transport of this layer over the first year observation period is estimated to be 3±5 Sv, indicating little contribution to the total transport of the Kuroshio.
One of the most important outcomes from estimating the Kuroshio transport will be an estimate of the net meridional heat transport of the mid-latitude North Pacific, by combining the transport with the transpacific hydrographic data and wind stress data along 30｡?. The effort is now under way. Someone might be interested in the relationship of this fluctuation of the Kuroshio transport with occurrence of the occasional stationary meander south of Japan. Throughout this five-year period, the Kuroshio took the path without the stationary meander and therefore no information on this has been obtained yet.
The present favorable relationship between the Kuroshio transport and the SSDT difference across the Kuroshio provides us with a practical way of long-term monitoring of the transport using satellite altimeter data.
ACKINOWLEDGEMENTS
The repeated hydrography data used in the present analysis were provided by members of the ASUKA Group, including M. Fukasawa, K. Hayashi, H. Ichikawa, S. Imawaki, H. Kinoshita, A. Misumi, I. Nakano, K. Okuda, T. Saito, Y. Sekine T. Shiga, H. Yoritaka and N. Yoshioka. Moored current meter data at 30｡? near the ASUKA line were kindly provided by D. Yanagimoto and used as reference velocities for transport estimates.
