Nonlinear Canonical Correlation Analysis of the Tropical Pacific Climate Variability Using a Neural Network Approach

Abstract

Recent advances in neural network modeling have led to the nonlineargeneralization of classical multivariate analysis techniques such asprincipal component analysis and canonical correlation analysis (CCA).The nonlinear canonical correlation analysis (NLCCA) method is used tostudy the relationship between the tropical Pacific sea level pressure(SLP) and sea surface temperature (SST) fields. The first mode extractedis a nonlinear El Nino-Southern Oscillation (ENSO) mode, showing theasymmetry between the warm El Nino states and the cool La Nina states.The nonlinearity of the first NLCCA mode is found to increase graduallywith time. During 1950-75, the SLP showed no nonlinearity, while the SSTrevealed weak nonlinearity. During 1976-99, the SLP displayed weaknonlinearity, while the weak nonlinearity in the SST was furtherenhanced. The second NLCCA mode displays longer timescale fluctuations,again with weak, but noticeable, nonlinearity in the SST but not in theSLP.