GiESCO 2019 banner
IVES 9 IVES Conference Series 9 GiESCO 9 GiESCO 2019 9 Aroma and quality assessment for vertical vintages using machine learning modelling based on weather and management information

Aroma and quality assessment for vertical vintages using machine learning modelling based on weather and management information

Abstract

Context and purpose of the study ‐ Wine quality traits are usually given by parameters such as aroma profile, total acidity, alcohol content, colour and phenolic content, among others. These are highly dependent on the weather conditions during the growing season and management strategies. Therefore, it is important to develop predictive models using machine learning (ML) algorithms to assess and predict wine quality traits before the winemaking process.

Material and methods ‐ Samples in duplicates of Pinot Noir wines from vertical vintages (2008 to 2013) of the same winery located in Macedon Ranges, Victoria, Australia were used to assess different chemical analytics such as i) aromas using gas chromatography – mass spectrometry, ii) color density, iii) color hue, iv) degree of red pigmentation, v) total red pigments, vi) total phenolics, vii) pH, viii) total acidity (TA), and ix) alcohol content. Data from weather conditions from the specific vintages were obtained both from the bureau of meteorology (BoM) and the Australian Wine Availability Project (AWAP) climate databases. Such data consisted of: i) solar exposure from veraison to harvest (V‐H), ii) solar exposure from September to harvest (S‐H), iii) maximum January solar exposure, iv) degree days from S‐H, v) maximum January evaporation, vi) mean maximum temperature from veraison to harvest, vii) mean minimum temperature from V‐H, viii) water balance from S‐H, ix) solar exposure from V‐H, x) degree hour accumulation with base 25 – 30 °C, xi) degree hour accumulation with base 25 °C, xii) degree hour accumulation with base 30 °C, xiii) degree hour accumulation with base 35 °C, and xiv) total cumulative degree days accumulation with base 10 °C. All data were used to develop two machine learning (ML) regression models using Matlab® R2018b. The best models obtained were using artificial neural networks (ANN) with the Levenberg‐Marquardt algorithm with 5 neurons for Model 1 and 9 neurons for Model 2. Model 1 was developed using the 14 parameters from the weather data as inputs to predict 21 aromas found in the wines from the six different vinatges. Model 2 was developed using the same 14 parameters from weather data and the eight chemical parameters as targets and outputs.

Results ‐ Both models obtained presented very high accuracy to predict wine quality trait parameters. Model 1 had an overall correlation coefficient R = 0.99 with a high performance based on the mean squared error (MSE = 0.01), while Model 2 had an overall correlation coefficient R = 0.98 with a high performance (MSE = 0.03). These models would aid in the prediction of wine quality traits before its production, which would give anticipated information to winemakers about the product they would obtain to make early decisions on wine style variations.

DOI:

Publication date: June 22, 2020

Issue: GiESCO 2019

Type: Article

Authors

Sigfredo FUENTES, Claudia GONZALEZ VIEJO, Xiaoyi WANG, Damir D. TORRICO

School of Agriculture and Food, Faculty of Veterinary and Agricultural Sciences, University of Melbourne, VIC 3010, Australia

Contact the author

Keywords

wine quality, machine learning, weather, aromas

Tags

GiESCO 2019 | IVES Conference Series

Citation

Related articles…

Polysaccharide families of lyophilized extracts obtained from unfermented varietal grape pomaces

The recovery of bioactive compounds from grape and wine by-products is currently an important objective for revaluation and sustainability. Grape pomace is one of the main by-products and is a rich source of some bioactive compounds. The aim of this study was to evaluate the polysaccharide (PS) composition of extracts obtained from pomaces of different white and red grape varieties of Castilla y León. Grape pomaces were obtained after the pressing in the winemaking process.

Sensitivity of vis‐nir spectral indices to detect nitrogen deficiency and canopy function in cv. Barbera (Vitis vinifera L.) Grapevines

Precision nutrient management in viticulture can be addressed on the basis of a spatial characterization of within‐vineyard vine

IMPACT OF MUST NITROGEN DEFICIENCY ON WHITE WINE COMPOSITION DEPENDING ON GRAPE VARIETY

Nitrogen (N) nutrition of the vineyard strongly influences the must and the wine compositions. Several chemical markers present in wine (i.e., proline, succinic acid, higher alcohols and phenolic compounds) have been proposed for the cultivar Chasselas, as indicators of N deficiency in the grape must at harvest [1]. Grape genetics potentially influences the impact of N deficiency on grape composition, as well as on the concentration of potential indicators in the wine. The goal of this study was to evaluate if the che- mical markers found in Chasselas wine can be extended for other white wines to indicate N deficiency in the grape must.

The sea breeze: a significant climatic factor for viticultural zoning in coastal wine growing areas

La brise de mer est un facteur climatique important pour le zonage viticole des régions viticoles côtières car l’accélération du vent qui lui est associée l’après midi ainsi que l’augmentation de l’humidité relative et la réduction de la température concomitantes sont significatives pour le fonctionnement de la vigne et, par conséquent, la qualité du raisin et du vin

Ethyl esters interact with the major wine Thaumatin Like Protein VVTL1

The interactions among aromatic compounds and proteins is an important issue for the quality of foods and beverages. In wine, the loss of flavor after vinification is associated to bentonite treatment and this effect can be the result of the removal of aroma compounds which are bound wine proteins. This phenomenon was recently demonstrated for long chain fatty acids and their ethyl esters (1). Since these latter compounds are spectroscopically silent, their association with proteins is not easy to measure.