A COMPOSITIVE PREDICTION OF EASTERN PACIFIC EQUATORIAL SST BASED ON WAVELET DECOMPOSITION AND ANFIS MODEL
Zhang Ren · Redai haiyang · 2002
Based on the method of wavelet decomposition combining with ANFIS, a compositive prediction model of the equatorial east Pacific sea surface temperature anomaly (SSTa) was established. By using this method, the complex SSTa system was separated into several relative simple bandpass signals, then the independent prediction models of the decomposed signals were set up, and the independent predicted results were integrated finally. The testing results showed that under the premise of keeping the object′s main characters, the prediction difficulty of SSTa system had been effectively decreased, the precision and durative of the compositive prediction model were evidently improved and promoted compared with those of traditional prediction technique.