Fig. 1. Time-longitude sections of anomalies in surface zonal winds (ms-1), SSTs (C) and 20℃ Cisotherm depth (m) for April 1996-March 1997. Analysis is from the TAO array 5-day means. Positive winds are westerly and values greater than 2ms-1 are shaded. SST anomalies greater than 1℃ are shaded and negative values are dashed. Buoy locations are given on bottom axis. From the TAO web page.
Another key feature of this event was the presence of intraseasonal oscillations which may have helped to initiate the event. A prevailing theory of ENSO is that it behaves as a delayed oscillator in which Rossby waves travel westwards off the equator to the western side of the Pacific where they are reflected and manifested as Kelvin waves along the equator and cause a reversal of the previous conditions. Evidence for such a sequence is not readily apparent in the onset of this event. Instead, the intraseasonal oscillations (40-50 day Madden-Julian Oscillation, MJO) were very prominent in the westerly wind anomalies in the far western Pacific) with westerly bursts in December 1996, February, May, August, October, and November 1997, and each one is traceable as a down-welling Kelvin wave propagating rapidly eastwards (across the Pacific in about 3 months) in the thermocline, as seen in the 20℃ isotherm anomalies, see Fig. 1. Thus within the overall envelope of the eastward developing subsurface temperature anomalies, there is important structure associated with the MJOs. The first warming at the surface in March 1997 coincided with the arrival of the downwelling Kelvin wave generated in Decembere 1996 (Fig. 1).
