关键词: 香茶, 品质因子, 产地判别, 随机森林回归

Abstract: The current studies on green tea quality factor analysis and origin-variety discrimination are mostly on famous green teas, while there are fewer studies on bulk green tea, which are more consumed in daily life. This study used Zhejiang Xiangcha tea, a bulk green tea representative, as its research subject. A total of 50 tea samples were gathered from three production areas and four varieties based on their conventional physicochemical indexes, catechin fractions, sensory factors, and wholesale prices in the origin trading market. Correlation analysis, linear discriminant analysis (LDA), and random forest regression (RF-R) were used to extract correlation factors and other information from the data. The results showed that the contents of free amino acids and caffeine were significantly correlated with many sensory factors, which were the key physical and chemical factors for the formation of flavor quality of fragrant tea. The content of free amino acids, catechins and the ratio of phenol to ammonia were the origin-related factors. There were significant differences in the total score, appearance and taste scores of different varieties of aromatic tea, and there were significant differences in aroma and leaf base scores. LDA demonstrated that Xiangcha tea had clear county-origin clustering and that county-origin discrimination was possible. The relative importance of price fitting was appearance > total review score > aroma > soup color > leaf base > taste > amino acids > water extract > phenol to ammonia ratio > caffeine > tea polyphenols, according to RF-well-fitted reference price model for wholesale trading of Xiangcha tea (R2 = 0.867, MAE = 7.907). This study serves as a research guide for the investigation of tea leaf quality factors, origin tracking, variety distinguish, and the development of trade reference prices.

Key words: Fragrant tea, Quality factor, Origin identification, Random forest regression