Systematic sampling and machine learning reveal the mechanisms of climatic cascades driving high-temperature Daqu ripening dynamics and liquor quality.

Abstract

High-temperature Daqu (HTD) is the key starter for Moutai-flavor Baijiu, yet the long ripening stage remains poorly understood. Here, we combined systematic sampling with industrial-scale validation and machine learning to elucidate how climatic cascades shape HTD ripening and liquor quality. Over 16 months, 576 HTD bricks from four seasonal batches were monitored for physicochemical properties, while 10-t fermentations using HTD ripened for 5, 6, 8 and 9 months were conducted to evaluate fermentation performance and liquor profiles. In contrast to acidity and starch content, HTD moisture showed more consistent associations with several key Baijiu flavor compounds, but only a weak association with ethanol yield. Moisture content showed a weak correlation with storage time but close associations with indoor humidity, which was largely influenced by outdoor climate. These findings suggest that climatic control of HTD moisture, rather than fixed storage duration, is the more informative target for managing consistent liquor quality.

Abstract as supplied by the indexing source. © the authors / publisher.