Evolution and historical perspective of the 1997-1998 El Nino-Southern Oscillation event

The ocean thermal history of the 1997-98 El Nino episode is described in detail, with emphasis on developments along the equator and eastern Pacific coastlines. The temporal evolution of the warming and its causes are traced from the western Pacific, past the Galapagos Islands and on to the subpolar gyres off North and South America. Along the equator, the event was characterized by a subsurface warm anomaly that slowly made its way from west to east across the Pacific from mid-1996 until early 1997, whence it triggered the onset of surface anomalies at the eastern terminus of the equatorial waveguide. The thermocline depression off Ecuador intensified from mid-1997 through the end of the year, culminating in a mature phase with maximum sea surface temperature anomalies (SSTA) around November-December 1997. The event gradually abated thereafter until the beginning of the subsequent cool phase (La Nina) was detected in July 1998. Following their arrivals at the eastern boundary, equatorial Kelvin waves proceeded poleward into both hemispheres as coastal trapped waves, carrying the thermocline depression signal with them along with associated nutrient deficiencies and ecosystem impacts. The poleward propagation of SSTA was more uniform and faster south of the equator, reaching south-central Chile with amplitudes of 2 degreesC or greater. North of the equator the propagation was discontinuous, with decreased anomalies south of 20 degreesN and a revival of SSTA in excess of 2 degreesC, north of there, but with considerably larger time lags than observed off Chile. The possible reasons for these interhemispheric differences are discussed. The magnitude of the event is also discussed in an historical context, with emphasis on comparisons to the El Nino of 1982-83. Each of the two events, in its own way, set records. However, the two events are generally comparable in their magnitudes and the extent of their impacts, while both are top-ranked events for the period after 1950. In the centennial context, however, these events are not unprecedented, considering that they were probably enhanced by strong decadal warming during the 1980s and 1990s. An attempt is made to assess the accuracy of model forecasts of the 1997-98 event. Two recent studies are discussed which generally agree that statistical and dynamical models under-predicted the equatorial warming prior to its onset and failed to capture the strong, early onset at all. Predictions of the late-1997 climax, with shorter lead times, improved once the data showing large mid- 1997 anomalies were ingested into the models. However, the revised predictions were not in time to guide the successful atmospheric climate outlook for North America, which was issued in June 1997 on the basis of observed strong anomalies on the equator.