Examination of the potential for using chemical analysis as a surrogate for sensory analysis

The application of a multi-block statistical analysis method, known as Common Components and Specific Weight Analysis. to the determination of connections between sensory descriptors and analytical data for Hunter Valley Semillon is described. Sixteen wines were used in the data analysis with 15 sensory descriptors and 10 analytical measurements available for each wine. The multi-block analysis simplifies the comparison between the data sets and allows relationships between the sensory and analytical parameters to be readily ascertained, more effectively than a linear regression approach. A sweetness zone established the connections between several sensory descriptors and analytical measurements based on fructose. Glucose was not part of the sweetness connections, although glycerol was connected to the sensory sweetness descriptors. Sensory assessment of acidity was positively related to the titratable acidity and pH was negatively related. The malic acid concentration was also negatively related to sensory acidity and the possible reasons for this are described. Several sensory descriptors including toast, honey and kerosene were found to be in opposition to the sweetness sensory parameters and not connected to any analytical parameters. The outcomes of this multi-block treatment indicate the potential for using analytical measurements as a surrogate for sensory analysis. (C) 2009 Elsevier B.V. All rights reserved.