Résumé
Objectively evaluating the quality of a vineyard in the context of climate change is not alwayssimple. Bayesian networks are widely used for knowledge representation and reasoning underuncertainty in natural resource management. There is a rising interest for this methodology astools for ecological and agronomic modelling. We designed a probabilistic model that takes intoaccount the parameters defining the status of a vineyard with their associated interactions. Nosuch model has been developed before. It includes an inference engine and software. Data werecollected from vine-growing experts. The model includes a database with more than 660 grapevarieties. For climate, our model uses a classification method (Tonietto and Carbonneau, 2004)involving multivariate measurements of climate on the basis of three indices: heliothermal index(HI), cool night index (CI), and dryness index (DI). Our model should ease assessments of thelikely impact of the choices and decisions of vine growers on the quality of new vineyards to beplanted. Thanks to this mathematical model, any kind of simulation of climate change based onthe international indexes can be performed. Some examples will be presented. Same thingconcerns a primary evaluation of models of sustainable Viticulture. The general frame of theGiESCO charter of sustainable Vitiviniculture is reminded on that occasion.