The estimated transport of the Kuroshio is found to have a very tight relationship with the SSDT difference across the Kuroshio. A scatter plot of the volume transport of the Kuroshio for the upper 1,000m layer and the SSDT difference across the Kuroshio is shown in Fig. 1a for those 25 cases. Their correlation coefficient is high, and the rms (root-mean-square) difference from the regression line is small. This tight relationship can be understood by examining the vertical profile of transport of the Kuroshio per unit depth. Individual transport profiles for 25 cases vary much, but if the transport is normalized by the sea-surface transport, the profiles become almost the same (Fig. 1b). Therefore their vertical integrals, i. e., total transports normalized by the seasurface transport, are almost constant. Namely, the total transport is almost proportional to the surface transport, which is directly related with the SSDT difference across the Kuroshio under the assumption of geostrophy. In other words, the increase (or decrease) of the total transport results from proportional increase (or decrease) of all the layers of the Kuroshio. Note that the normalized transport at 1,000m depth is fairly small (0.09±0.05), which indicates that most of the transport is confined to the upper 1,000m layer.
TRANSPORT TIME SERIES
Having this relationship, SSDT difference across the Kuroshio derived from TOPEX/POSEIDON satellite altimetry data provides us with a time series of the volume transport of the Kuroshio with sufficiently fine resolution (10-day intervals) and as long a duration as the satellite operates. Figure 2a shows a five-year long record for 1992-1997 thus obtained, which is the first long time series of the Kuroshio transport with sufficient temporal resolution. Altimeter-derived transports agree very well with transports estimated from in situ data.
Here TOPEX/POSEIDON altimetry data are processed for the first 194 repeat cycles of 10day period to estimate SSDT profiles along the ASUKA line. The presently available geoid model is not accurate enough to be used with the altimetry data in order to estimate the absolute SSDT (deviation of the sea surface from the geoid). Therefore only anomalies from the temporal mean are usable in the altimetry data. In the present analysis, above-mentioned SSDT profiles estimated from in situ data along the ASUKA line are combined with SSDT anomalies from altimetry data to obtain the mean SSDT profile. The sum of this mean profile and the anomalies from the altimetry data gives us the absolute SSDT profile. Sea level data at Cape Ashizuri are used to improve the SSDT profiles near the coast, where the altimeter cannot work well.
The transport of the Kuroshio fluctuates much. The seasonal cycle is not apparent, and fluctuations of shorter periods are dominant. Part of the transport and its variability estimated above are associated with those of the stationary local anticyclonic warm eddy located on the offshore side of the Kuroshio (8). The present estimate of the Kuroshio transport includes the eastward transport of the northern part of this eddy, inevitably. To understand the western boundary current of the North Pacific subtropical gyre, those local transports should be excluded and the transport of the Kuroshio as a throughflow should be estimated.
