The Prediction of Non-stationary Climate Series Based on Empirical Mode Decomposition

This paper proposes a new approach which we refer to as ``segregatedprediction to predict climate time series which are nonstationary. Thisapproach is based on the empirical mode decomposition method (EMD), whichcan decompose a time signal into a finite and usually small number of basicoscillatory components. To test the capabilities of this approach, someprediction experiments are carried out for several climate time series. Theexperimental results show that this approach can decompose thenonstationarity of the climate time series and segregate nonlinearinteractions between the different mode components, which thereby is able toimprove prediction accuracy of these original climate time series.