Figure 3 (a)Time series of zonal wind stress at 147.375°E. Low pass filter, cutoff period of which is at 150 days, is applied. (b) Time series of low pass filtered zonal current at 147°E at 40m (thick line). 60m (thin line), 80m (thick dash line) and 100m (thin dash line). (c) Monthly mean 29℃ isotherm depth at 147°E-0°N (circle) and at 147°E-2°N (asterisk). If available data are less than two weeks long, they are omitted except for July of 1996. 95% confidence limits computed from sample standard mean are shown as vertical solid lines.
5 Summary
JAMSTEC is conducting long term continuous current measurements by means of moored ADCP at seven sites, five sites at the equator and two sites off the New Guinean coast, in the western tropical Pacific. The vertical coverage of these observations are from 30-40m to 260-270m with some exceptions. Among these, observations at 142｡?-0｡? and at 147｡?-0｡? are about four years long and are the longest records. Zonal currents near the surface were strongly influenced by local wind. The mean zonal current at 40m at 147｡?-0｡? was westward in 1995 and in 1996 while it was eastward in 1997. The result of the analysis applied to the ECMWF wind data shows that low frequency variations of zonal wind speeds at positions closest to our moorings were highly correlated with each other. This may cause the correlation between the near-surface zonal current and the locale zonal wind speed in the western equatorial Pacific to be high even though it is low in the central equatorial Pacific (Kessler et al, 1995).
Below the depths where the wind had large influence, there was a westward flowing South Equatorial Current (SEC). Below the SEC, there was an eastward flowing Equatorial Undercurrent (EUC), and the mean zero-crossing depth between these two currents was about 140m. There was an annual variation of EUC; the upper limit of EUC shoaled in December to February and deepened gradually from March until November. The secondary core was separated from the main body of the EUC and was developed at about 100m or less in boreal summer. The development of this secondary core was synchronous with the seasonal shift of the zonal wind direction from westerly to easterly, and the meridional distribution of the pycnocline at the bottom of the mixed layer across the equator exhibited an upward curve centered at the equator in boreal summer. These results suggest that this upper core of the EUC in boreal summer was caused by a local equatorial upwelling due to the local easteriy wind (Philander, 1973; McCreary, 1981a and 1981b). The EUC temporarily disappeared from our observation depths from September 1994 throug November 1994.
