© Académie du chenin. Créé avec Wix.com

 

Moderator : Jacques Carroger

1. Ilaria Brunetti - A Chenin decision help system to predict grape berries maturity and anticipate wine potentialities

The grape berries maturation is a complex process relying on physicochemical and biochemical reactions. These reactions depend on multiple factors of which the climate is the most influential, especially in the last weeks preceding the harvest. Since berries maturity plays a major role in determining wine potentialities, to anticipate the maturation process and determine the right harvesting date is a significant challenge for the wine industry. Different indicators to evaluate the maturation state can be considered, which might be chemical, and thus exactly measurable as the content of sugar, or sensory characteristics, as the seeds color, which requires an expert evaluation on a symbolic ordinated scale.
 

In this work we present a mathematical model able to predict the maturation process, considering both chemical and sensory indicators, by observing the weather conditions. The predicted indicators’ values are then used to evaluate the wine potentialities, according to the winemaker target.


We first built a Dynamic Bayesian Network (DBN), which allows us to obtain reliable dynamic predictions of sugar, total acidity, and malic acid, as well as of a selection of sensory indicators, by observing air temperature, rain fall, relative humidity and sunshine hours in the last three weeks preceding the harvest.


Since DBN models require discrete variables and the preliminary discretization process might be a complex and time-consuming task, we developed a semi-automatic discretization technique merging human knowledge and automatic optimization, which not only simplifies the discretization task, but also improves the DBN results.


Once we predict the maturation indicators, we evaluate the wine potentialities according to the winemaker expectation, by means of a Fuzzy Logic model, linking the maturity indicators to global wine quality.

2. Nicolas Bernard - Chenin Blanc ripening dynamics in Anjou: focus on the 5 last vintages through berry active sugar loading, berry volume and color evolution.

In 2014, a network of about forty Chenin Blanc blocks was set up as part of a collaboration between a group of Anjou winegrowers and the Vivelys company. One of the goals of this project was to describe the ripening dynamics of the Chenin and to understand better the relationships between grape potential, positioning of harvest dates and wine profile. This study summarizes the main results obtained between 2014 and 2018.


First, the ripening profiles of each vintage are described through the sugar loading kinetics of the berries. The analysis reveals a significant effect of the vintage on the duration and speed of active sugar loading, as well as on the date of cessation of sugar accumulation and the Potential Degree at the cessation of sugar accumulation. The cumulative summer rainfall seems to be the main factor explaining the observed variations. A comparison with the maturation dynamics observed in the Paarl region of South Africa confirms the influence of summer rainfall on sugar loading kinetics.


Secondly, an analysis of the quality ratings of the vintages reveals the links between summer rainfall, berry sugar loading and the overall quality of the vintage: the best rated vintages correspond to the driest summers and the most efficient sugar loadings.


Finally, a synthesis of the sensory analyses of the wines, carried out each vintage, enables us to propose a model for understanding the relationships between the sugar loading of the berries, the positioning of the harvest and the profile of the wines.

the Program

Session V - Viticulture and winemaking itineraries of Chenin blanc in a changing environment