Abstract: A descriptive analysis of the aroma of six samples of Keemum black tea (three grades of Maofeng tea and three grades of Congou tea) was carried out using sensory evaluation, quantitative descriptive analysis (QDA) and flash profile (FP). In QDA, ten trained sensory panelists evaluated 16 aroma attributes, and analysis by PanelCheck and analysis of variance (ANOVA) showed that 10 of the aroma attributes were significantly different among the samples. In FP, 18 untrained consumer panelists established 51 aroma attributes and ranked them in terms of their intensity. Using generalized procrustes analysis (GPA) and agglomerative hierarchical clustering (AHC), we obtained similar results to those using QDA. The FP method could easily and quickly distinguish the differences in aroma between tea samples, but could not provide specific aroma descriptors. The combined use of QDA and FP may provide a more scientific and practical approach for tea sensory evaluation.

Key words: Keemun black tea; aroma characteristics; traditional tea sensory evaluation; quantitative descriptive analysis; flash profile