Modelling past and future wine production in the Portuguese Douro Valley

Climate is a major factor driving wine productivity in the Douro Valley region of Portugal because vineyards are generally grown under marginal conditions for agricultural production. Using monthly means of climate variables (daily maximum, minimum and mean temperature, diurnal thermal amplitude and daily precipitation) and the normalised difference vegetation index, we analysed the vegetative cycle of vineyards and the vulnerability of Douro wine production to climate variability and change. A good wine production year reflects high photosynthetic activity during spring followed by reduced greenness during summer. Lower precipitation in March has a positive effect during the growing stage, and higher temperatures during late spring are beneficial for production. Two linear regression models were developed with the aim of forecasting Douro wine production at the early (March) and mid-season (July) stages of the vine vegetative cycle. Both models performed well in terms of robustness and reliability. The early season model has the advantage of providing a reliable production scenario immediately after the dormancy stage. All predictors of the mid-season model are meteorological parameters; combined with the very high value (90%) of explained variance, this suggests its usefulness to assess the crucial role of climate in wine production. Limitations result from the linear nature of the model, and effects due to the increase in CO2 were disregarded. Nevertheless, our results indicate an increase in production from 1956-1985 to 1986-2006 due to changes in the winter precipitation regime. Moreover, 2071-2100 climate scenarios (A2 and B2) are expected to favour an increase in wine production because of the combined effects of temperature and precipitation in late spring and early summer.