Can spectroscopy geographically classify Sauvignon Blanc wines from Australia and New Zealand?

Danial Cozzolino, Wieslawa Cynkar, N Shah, Paul Smith

    Research output: Contribution to journalArticlepeer-review

    54 Citations (Scopus)

    Abstract

    The combination of UV, visible (Vis), near-infrared (NIR) and mid-infrared (MIR) spectroscopy with multivariate data analysis was explored as a tool to classify commercial Sauvignon Blanc (Vitis vinifera L., var. Sauvignon Blanc) wines from Australia and New Zealand. Wines (n = 64) were analysed in transmission using UV, Vis, NIR and MIR regions of the electromagnetic spectrum. Principal component analysis (PCA), soft independent modelling of class analogy (SIMCA) and partial least squares discriminant analysis (PLS-DA) were used to classify Sauvignon Blanc wines according to their geographical origin using full cross validation (leave-one-out) as a validation method. Overall PLS-DA models correctly classified 86% of the wines from New Zealand and 73%, 86% and 93% of the Australian wines using NIR, MIR and the concatenation of NIR and MIR, respectively. Misclassified Australian wines were those sourced from the Adelaide Hills of South Australia. These results demonstrate the potential of combining spectroscopy with chemometrics data analysis techniques as a rapid method to classify Sauvignon Blanc wines according to their geographical origin.

    Original languageEnglish
    Pages (from-to)673-678
    Number of pages6
    JournalFood Chemistry
    Volume126
    Issue number2
    DOIs
    Publication statusPublished - 15 May 2011

    Keywords

    • Classification
    • Geographical origin
    • Mid-infrared
    • Near-infrared
    • Sauvignon Blanc
    • UV
    • Visible
    • Wine

    Fingerprint Dive into the research topics of 'Can spectroscopy geographically classify Sauvignon Blanc wines from Australia and New Zealand?'. Together they form a unique fingerprint.

    Cite this