Multivariate data analysis in classification of must and wine from chemical measurements

A chemometric study was carried out; based on measurements of chemicals pre sent in musts and wines from the Tacoronte-Acentejo Designation of Origin a rea (Canary Islands, Spain) obtained in an artisanal winemaking procedure. Univariate and multivariate data analysis was used to distinguish between t wo different harvest years (1987 and 1988) in both musts and wines. The dat a was examined using the statistical techniques of discriminant analysis, p rincipal component analysis, biplot analysis and factor analysis, which all owed the study of the underlying structures of the data and the characteriz ation of the wines. This methodology was able to differentiate between must s and wines according to traditional variables used in wine analysis. Howev er, no clear differentiation was found with respect to the year of vintage.