FY 1989

Equatorial oceanic response to forcing on time scales from days to months

Giese, B.S.

NOAA Tech. Memo. ERL PMEL-87, NTIS: PB89-206775, 99 pp (1989)

Episodes of westerly wind in the western Pacific may be an important source of sea surface temperature variability in the eastern Pacific on monthly, seasonal and interannual time scales. In this report we use a combination of data, linear theory and an ocean general circulation model to examine remote response to western Pacific wind forcing. Characteristics of the wind anomaly are determined using daily averaged observations of wind from equatorial islands near the date line. In the model, wind anomalies generate a train of eastward propagating Kelvin pulses. When the wind anomaly is weak the Kelvin response agrees with predictions of linear theory. For more realistic strong forcing there are three important deviations from linear theory; the amplitude of low baroclinic modes increases, the amplitude of higher baroclinic modes decreases, and the phase speed increases. In the presence of realistic oceanic background conditions, response in the equatorial waveguide is complicated by the equatorial undercurrent, a sloping thermocline and instability waves. As Kelvin pulses propagate from western to eastern Pacific surface zonal velocity associated with the first mode decreases, whereas velocity associated with the second mode increases. These changes can be deduced by the principle of conservation of energy flux. In the central and eastern Pacific Kelvin pulses act to amplify and change the phase of existing instability waves. Thermal changes brought about by enlarged instability waves can be comparable in magnitude to changes brought about by zonal advection of the zonal temperature gradient by Kelvin pulses. At the coast of South America, model Kelvin pulses cause a warming of 2°C for 45 days. Current observations made at 140°W and 110°W subsequent to a strong westerly wind event in May 1986 indicate passage of Kelvin-like pulses which agree in magnitude and timing with those modeled. At the coast of South America observations of sea surface temperature show a warm anomaly that lasts for almost two months, comparable in duration and magnitude to changes found in the model.