Abstract
There has been intense debate about the causes of the 1997--1998 El Niño. One side sees the obvious intense westerly wind events as the main cause for the exceptional heating in summer 1997, the other emphasizes slower oceanic processes. We present a quantitative analysis of all factors contributing to the onset of this El Niño. Specifically, we decompose the NINO3 index in the HOPE OGCM at 1 June 1997 into contributions from the fluxes and initial state at six months' lead time. The initial state thermal anomalies contribute about 40% compared with an average year, and the wind stress about 50%. Compared with the previous year, in which no El Ni\~nntilde;o developed, the main difference is in the zonal wind stress. This contribution is concentrated at the time and place of the strong westerly wind events in December 1996, March and April 1997. As westerly wind events are difficult to predict, this limited the predictability of the onset of this El Niño.
(top) Longitude-depth plot of the anomalies in the initial state temperature, the sensitivity of the June 1997 NINO3 index to these, and their influence.
(left) The zonal wind stress in the week around 11 March 1997, the sensitivity of the ocean to zonal wind stress, and their product: the influence of this week on the onset of the El Niño.