The paper explores to which extent narrative documentary records on wine production in Central Europe can be used as a proxy for summer temperatures. Here, we compiled 11 regional series spanning from the early 15th century to 1988. We detrended the data to adjust for long-term biases such as warfare-related population decline. The homogenised regional series were subsequently merged into three supra-regional series: (1) the Mosel series, starting in 1416 and consisting of data from the former city-republic of Metz (France) and the Grand Duchy of Luxembourg; (2) the series for Germany, starting in 1511 and mainly originating from the former city-state of Heilbronn; and (3) the third series, starting in 1529 and representing production on the Swiss Plateau. The residuals of the supra-regional yield series were averaged, divided into seven classes, and multiplied by five quality classes. Yield quality indices (YQI) varying between 35 (large and excellent) and 1 (small and undrinkable) significantly correlated with temperatures between May and July. Regression analysis of the composite series revealed that yield and quality primarily depend on the climate conditions from May to July as well as on those in June of the previous year. Crops with a YQI > 28 (rated “good” by traditional winegrowers) were related to above-average May–July temperatures, early grape harvest dates and high tree-ring maximum latewood density values resulting from frequent anticyclonic weather situations. Crops with YQI > 10 could not be uncritically assigned to cold summers since severe winter conditions, and spring frosts sometimes substantially reduced yields without affecting quality, particularly during the Little Ice Age in the marginal Mosel and German areas. Extreme yield fluctuations challenged winegrowing communities. Overproduction induced wastefulness, while crop failures sometimes triggered witch hunts.
Pfister et al. (2026) studied this question.