Long-term non-linear predictability of ENSO events over the 20th century

Abstract

We show that the monthly recorded history of the Southern Oscillation Index (SOI), a descriptor of the El Niño Southern Oscillation (ENSO) phenomenon, can be well described as a dynamic system that supports an average nonlinear predictability well beyond the spring barrier. The predictability is strongly linked to a detailed knowledge of the topology of the attractor obtained by embedding the SOI index in a wavelets base state space. Using the state orbits on the attractor we show that the information contained in the Southern Oscillation Index (SOI) is sufficient to provide average nonlinear predictions for time periods of 2, 3 and 4 years in advance throughout the 20th century with an acceptable error. The simplicity of implementation and ease of use makes it suitable for studying non linear predictability in any area where observations are similar to those that describe the ENSO phenomenon.