GiESCO 2019 banner
IVES 9 IVES Conference Series 9 GiESCO 9 Machines and fire: developing a rapid detection system for grapevine smoke contamination using NIR spectroscopy and machine learning modelling

Machines and fire: developing a rapid detection system for grapevine smoke contamination using NIR spectroscopy and machine learning modelling

Abstract

Context and purpose of the study – Bushfires are a common occurrence throughout Australia and their incidence is predicted to both rise and increase in severity due to climate change. Many of these bushfires occur in areas close to wine regions, which receive different levels of exposure to smoke. Wine produced from smoke-affected grapes are characterised by unpalatable smoky aromas such as “burning rubber”, “smoked meats” and “burnt wood”. These smoke tainted wines are unprofitable and result in significant financial losses for winegrowers. This study investigated the use of near-infrared (NIR) spectroscopy and machine learning (ML) modelling for the rapid and non-destructive detection of grapevine smoke exposure by analysing grapevine leaves and/or grape berries.

Materials and methods – The trial was conducted during the 2018/2019 season at the University of Adelaide’s Waite campus in Adelaide, South Australia (34° 58’ S, 138° 38’ E) and involved the application of five different smoke and water misting treatments to Cabernet Sauvignon grapevines at approximately seven days post-veraison. Treatment vines were exposed to straw-based smoke for one hour under experimental conditions described previously by Kennison et al. (2008) and Ristic et al. (2011). Near-infrared (NIR) measurements were then taken from berries and leaves a day after smoking using the microPHAZIR TM RX NIR Analyser (Thermo Fisher Scientific, Waltham, USA) which has a spectral range of 1600-2396 nm. The NIR spectra were then used as inputs to train different ML algorithms, which resulted in two artificial neural networks (ANNs) with the best classification performance for either berry or leaf readings according to the different smoke treatments.

Results – Both ANN models found were able to correctly classify the leaf and berry spectral readings with high accuracy. The leaf model had an overall accuracy of 95.2%, 97.7% accuracy during training with a mean square error (MSE) 0.0082, 90.9% during validation with a MSE of 0.0353 and 88.1% during the testing stage with a MSE of 0.0386, while the berry model had an overall accuracy of 91.7%, 95.2% accuracy during training with a MSE of 0.0173, 86.4% during validation with a MSE of 0.0560 and 80.2% during the testing stage with a MSE of 0.0560. These results showed the potential of developing a rapid, non-destructive, in-field detection system for assessing grapevine smoke contamination following a bushfire using NIR spectroscopy and artificial neural network modelling.

DOI:

Publication date: September 28, 2023

Issue: GiESCO 2019

Type: Poster

Authors

Vasiliki SUMMERSON, Claudia GONZALEZ VIEJO, Damir TORRICO, Sigfredo FUENTES*

The University of Melbourne, School of Agriculture and Food, Faculty of Veterinary and Agricultural Sciences, Parkville 3010, Victoria, Australia

Contact the author

Keywords

bushfires, machine learning, smoke taint, climate change, non-destructive

Tags

GiESCO | GiESCO 2019 | IVES Conference Series

Citation

Related articles…

Innovative sparkling wines, traditional grape varieties and autochthonous yeasts: emerging trends for regional products diversification

Italy, like all the major vine-growing and wine-producing countries, has experienced a decline in wine export volumes in recent years.

Development of a semi-controlled setup for manipulating drought and heat stress in open field trials

Drought and heat stress will pose challenges for the future of viticulture and wine quality, as grapevine biological processes are pushed beyond their optimum conditions. Efforts are increasing to study and predict the effects of drought spells and heatwaves on grapevine physiology and resulting harvest quality. This calls for the development of adequate systems to induce and manipulate the required stress, especially in open field trials where conditions are more difficult to control. We present a semi-controlled system for studying drought and heat stress in grapevine in the field.

Genotypic differences in early-stage root architectural traits and consequences for water uptake in three grapevine rootstocks differing in drought tolerance

Root architecture (RSA), the spatial-temporal arrangement of a root system in soil, is essential for edaphic resources acquisition by the plant, and thus contributes largely to its productivity and adaptation to environmental stresses, particularly soil water deficit. In grafted grapevine, while the degree of drought tolerance induced by the rootstock has been well documented in the vineyard, information about the underlying physiological processes, particularly at the root level, is scarce, due to the inherent difficulties in observing large root systems in situ. The aims of this study were (i) to determine the phenotypic differences in traits related to root distribution and morphology along the substrate profile in different Vitis rootstocks during early growth, (ii) to assess the plasticity of these traits to soil water deficit and (iii) to quantify their relationships with plant water uptake.

Recent advances in measuring, estimating, and forecasting grapevine yield and quality

Grapevine yield and fruit quality are two major drivers of input allocation and, ultimately, revenue for grape producers. Because yield and fruit quality vary substantially from year-to-year and within a single block, opportunities exist for optimization via precision management activities that could lead to more profitable and sustainable grape production. Here, we review recent advances in the techniques and technology used to measure, estimate, and forecast grapevine yield and fruit quality. First, we discuss direct “measurement” of yield and quality (i.e. ground-truth data generation), with an emphasis on potential for scalability and automation. Second, we discuss technology and techniques that do not directly measure yield and quality, but use correlated measurements for their estimation.

Terroir traceability in grapes, musts and wine: results of research on Gewürztraminer and Sauvignon Blanc grape varieties in northern Italy

In the study of terroir, a separate analysis of its many component factors can be of great help in accurately identifying a vineyard’s natural elements that impact wine quality and typicity. This research used a dedicated pluri-disciplinary approach to investigate the ecological characteristics, including geology and geographical features, of 14 vineyards that produce Gewürztraminer and Sauvignon Blanc cultivars in the alpine Alto Adige DOC wine region. Both the geopedological method using Vineyards Geological Identity (VGI) and the new Solar Radiaton Identity (SRI) topoclimatic classification method were used to provide analytical measurements and qualitative/quantitative characterisations. In addition, wide-ranging targeted and untargeted oenological and chemical analyses were carried out on grapes, musts and wines to correlate the soils’ geomineral and physical conditions with the biochemical properties of their fruits and wines. The research identified strong correlations between vineyard geo-identity and wine biofingerprint, confirming a mineral traceability of strontium rubidium ratio and some minerals distinctive to the local geology, such as K, Ca, Ag, Ba and Mn.  The study also discovered that particular geomineral and physical soil conditions of the studied vineyards are related to the different amount of amino acids, primary varietal aromas and polyphenols found in grapes, musts and wines. The research confirmed that winemaking technologies support oenological quality, although in some cases, human practices can overpower certain characteristic elements in wine, erasing the typical imprint left by the vineyards’ natural terroir, which becomes less traceable. Terroir abiotic ecological factors and vineyard identity can be classified in detail using the new VGI and SRI analysis methods to discover interrelationships between geo-pedological and topoclimatic conditions that impact wine quality. These methods are also helpful in identifying which ecological elements are exclusive to a particular vineyard or wine sub-region.