Ocean Model Studies of Upper-Ocean Variability at 0°N, 160°W during the 1982–1983 ENSO: Local and Remotely Forced Response

D.E. Harrison

NOAA, Pacific Marine Environmental Laboratory, 7600 Sand Point Way NE, Seattle, WA 98115

A.P. Craig

School of Oceanography, University of Washington, Seattle, WA 98195

Journal of Physical Oceanography, 23(3), 426-451 (1993)
Copyright ©1993 American Meteorological Society. Further electronic distribution is not allowed.

2. Comparison of observations at 0°N, 160°W and SADLER hindcast results

Figures 1a and 2a present the depth-time plots of zonal current and temperature from Firing et al. (1983). The interval between July 1982 and March 1983 will be focused upon because the currents went through the largest variations during this time. From mid-June 1982 to mid-September 1982, the core of the EUC decelerates from over 80 cm s maximum speed eastward to westward flow at about 120-m depth. From mid-October 1982 to mid-November 1982, eastward acceleration of the surface flow forms a strong surface-trapped jet with weak westward flow just beneath it. From mid-November 1982 to mid-January 1983, the eastward surface jet decelerates rapidly and the surface flow becomes westward while the deep flow accelerates eastward late in this period. By mid-February, the EUC has returned. During the spring, the EUC shallows slightly and surface flow returns to more normal conditions.

Figure 1. Zonal velocity during 1982-1983. (a) profiles from Firing et al. (1983) at (0°, 159°W), (b) results from the model using the SADLER wind stress field (Harrison et al. 1989) at 0°N, 160°W. The contour interval is 20 cm s in both plots.

Figure 2. As for Fig. 1, except for temperature. The contour interval is 2°C.

There is also interesting thermal evolution (Fig. 2a) during the same period. In July, there is rapid warming throughout the water column. From August 1982 to November 1982, there is weak warming above 160 m. Water warmer than 30° appears above 70 m in October. From late November 1982 through January 1983, there is shoaling and intensification of the thermocline. During this period the thermocline is very sharp and the 28° and 18°C isotherms are separated by only about 30 m. In January there is rapid warming of water colder than 18°C. From February 1983 through spring 1983, the thermocline gradient weakens dramatically as isotherms greater than 22°C continue to rise while isotherms less than 22°C deepen.

Figures 1b and 2b present the corresponding fields from the SADLER hindcast experiment. The hindcast reproduces all of the major features of the observational data, with scales of variability close to those observed. The largest discrepancies are in the delayed return of the EUC in 1983 in the hindcast, and the differences in the near-surface temperatures before July 1982 and after April 1983. Otherwise, these timing discrepancies are typically smaller than one month, which is the resolution of the imposed monthly mean wind stress field of the hindcast. This is at least as successful a longterm simulation of observations as any known to the authors; the hindcast has quantitative skill, in the sense discussed by Harrison et al. (1989), between July 1982 and January 1983.

Figure 3 presents the equatorial zonal and meridional wind stress fields for the SADLER hindcast experiment as a function of longitude and time. These wind stress fields were calculated by multiplying the SADLER pseudostress results by a drag coefficient of 1.2 × 10 and air density of 1.2 × 10 gm cm (Harrison et al. 1989). This C value is consistent with the values of Large and Pond (1981) for the windspeed range of this area over this period. Climatologically there are easterlies across the region between 160°W and 160°E throughout the seasonal cycle (see Harrison et al. 1989 for more comparison between the 1982-83 stresses and the climatological stresses), so that the westerlies and weak easterlies found during 1982 are strongly anomalous, as are the strong meridional winds in late 1982 through early 1983.

Figure 3. SADLER analysis of the monthly mean zonal and meridional wind stress along the equator, 1982-1983. The contour interval is 0.02 N m in both plots.

Figure 4 presents time series of zonal and meridional stress at 160°W as well as at several longitudes west of 160°W. With the short periods of enhanced westerly wind west of the date line and the abrupt termination of the period of westerlies in December 1982, linear theory predicts that Kelvin pulses will be generated and then propagate eastward away from the region of westerly winds. Eastward propagation of the zonal stress forcing occurs during some periods; this can affect the rate at which remote response occurs (Harrison and Schopf 1984) or lead to resonant Kelvin forcing (McCreary and Lukas 1986). A number of papers have discussed the elements of the physics of forced Kelvin pulses (see, e.g., Giese and Harrison 1990). At the observation site (0°, 160°W) there is also substantial local wind forcing, however, so the behavior there will likely also depend on the local forcing. In the next section, we examine how much of the response is produced by local forcing.

Figure 4. Time series of equatorial zonal and meridional wind stress at 160°W and at several longitudes west of 160°W.

Return to previous section or go to next section

PMEL Outstanding Papers

PMEL Publications Search

PMEL Homepage