Climate factors driving wine production in the Portuguese Minho region

• Published in: Agricultural and forest meteorology 2014-02, Vol.185, p.26-36
• Main Authors: Fraga, H., Malheiro, A.C., Moutinho-Pereira, J., Santos, J.A.
• Format: Article
• Language: English
• Subjects: Agricultural and forest climatology and meteorology. Irrigation. Drainage; Agronomy. Soil science and plant productions; Atmospherics; Biological and medical sciences; Climate; Climatic factors; Climatology; Fermented food industries; Food industries; Fundamental and applied biological sciences. Psychology; General agronomy. Plant production; International trade; Logistic regression; Mathematical models; Minho Wine Region; Portugal; Regression analysis; Vitaceae; Weather; Weather regimes; Wine production modeling; Wines; Wines and vinegars
• ISSN: 0168-1923; 1873-2240
• DOI: 10.1016/j.agrformet.2013.11.003

•Climatic factors controlling wine production in the Minho Wine Region are studied.•A long wine production series is used from 1945 to 2010 (66 years).•Wine productions are classified as low, normal and high for logistic regression.•Modeling is carried out using ten regressors, including new weather regimes.•A 72% skill in modeling production classes is obtained after cross-validation. Establishing the role of climate on wine production is one major goal of the winemaking sector. Portuguese viticulture plays a key role in national exports of agro-food products. The Minho Wine Region, in particular, produces a unique wine type ‘Vinho Verde’ that has been taking its stand as a prominent international brand. The present study aims at improving the understanding of climate-yield relationships in this region. A long wine production series (1945–2010) is used and some transformations are undertaken for robust statistical relationships. A stepwise methodology is applied to select regressors for logistic modeling of production classes (low, normal and high). New weather regimes are developed to assess large-scale atmospheric forcing and cycles in production are isolated by a spectral analysis. Ten regressors are selected: dryness and hydrothermal indices, 3-yr lagged production, mean temperatures in March and June, precipitation in June and frequencies of occurrence of two regimes in May, and of one in February and September. Overall, moderate water stress during the growing season, high production 3-yrs before, cool weather in February–March, settled-warm weather in May, warm moist weather in June and relatively cool conditions preceding harvest are generally favorable to high wine production. Some of these relationships demonstrate the singularity of Minho Wine Region and justify the present study. The model shows high skill (72% after cross-validation), stressing not only the important role played by atmospheric conditions, but also its value for prediction and management.