Extratropical Precursors of the El Niño–Southern Oscillation

Extratropical mechanisms thought to stochastically force ENSO are reviewed and investigated as potential ENSO predictors. Results indicate that these precursors have a stronger impact on the probability than amplitude of ENSO events and more strongly impact El Niño than La Niña. We identify the combination of precursors that best fit eastern and central Pacific ENSO indices using linear regression. The best models have poor fit to the EP index, representing only 4% of the variance of this index at 1 year lead time. However, the best model can represent 36% of the variance of the CP index at the same lead time. The extratropical precursors and their impact on the tropics can be seen a year in advance as opposed to events without an extratropical influence, which develop much more quickly beginning in the spring. As a result, extratropically forced ENSO events have a higher potential for long lead predictability. Although a subset of precursors is selected for the models, evidence of all of the precursors is present when fields are regressed with the modeled CP index. This demonstrates that the precursors are highly interrelated and should be evaluated in a holistic manner rather than as separate mechanisms.