Application of principal component analysis to the interpretation of rainwater compositional data

Abstract

Principal component analysis (PCA), based on non-linear iterative partial least squares (NIPALS) coupled with a cross-validation approach, is applied to data obtained from the chemical analysis of rainwater. The correlation between variables is obtained and their sources identified. The classification of samples into groups by PCA is also investigated. The problem of data scaling and the evaluation of methods for assessing the number of significant components in the data are also dicussed.