Mapping the Origins of Chinese Grape Wine: Unlocking Geographical Regions Through Chemometrics, Stable Isotopes, and Multielement Analysis

Developing a robust and reliable geographical origin discrimination model, unaffected by grape varieties, wine types, or the scale and distance of the wine‐growing area, is essential for the sustainable growth of China’s wine industry. A total of 967 wine samples were collected from 10 original regions in China, and 27 elemental profiles and two stable isotope ratios were determined by ICP‐MS and IRMS, respectively. PCA was used to evaluate the contributions of 29 regional markers in classifying the geographical regions of Chinese wine. Six chemometric methods were employed, including PCA, PLS‐DA, RF, ANN, SVM, and R‐Part, to verify the regions and subregions of Chinese wine. This study successfully proposed advanced simultaneous classification models for original regions and subregions of Chinese wine using multielements and stable ratios combined with random forest (100% accuracy) for the first time.