Statistical methods for improving authentication of wines based on stable isotope ratios

Wine derives its economic value partly from its geographical origin and production technologies. Authenticity evaluation involves a check of several features like declaration of origin, year of vintage, and also the evaluation of any addition of unpermitted substances during the production process like water and sugar. In Italy, at least 400 reference samples (collecting a particular number of samples from every region) every year are used to build the isotope ratio databank ((D/H)1, (D/H)2, R, δ 18O, δ13C). In this work, we evaluate several methods based on this unique data repository for checking authenticity claims (wine origin, detection of unpermitted sugar and water addition). The results show that using multivariate methods false claims of origin as well as unpermitted sugar and water addition can be discovered much easier