Romanian researchers have applied FT-Raman spectroscopy and metabolomics to classify a sample set of 126 wines, originating from Romania and France, with respect to cultivar, geographical origin and vintage. Cultivar recognition was successfully achieved for four grape varieties (Sauvignon, Riesling, Chardonnay, Pinot Gris), and the subtle particularities existing between the Chardonnay wines, coming from the two countries, because of terroir influences were observed. In addition, a 100% separation in both initial and cross-validation procedures for geographical differentiation between the two origin countries, as well as, among the three Romanian areas (Transylvania, Muntenia and Moldova) were obtained. The limitations of this approach, and the importance of choosing meaningful data sets, in terms of representativity for each classification criterion, are addressed in this paper.
Read the full paper here