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…

How distinctive are single vineyard Gewürztraminer musts and wines from Alto Adige (Italy) based on untargeted analysis, sensory profiling, and chemometric elaboration?

Vitis vinifera L. ‘Gewürztraminer’ is a historical grape variety of Alto Adige (Südtirol), Italy, which is widely grown in the area of Tramin an der Weinstraße, but is also grown globally. It produces highly aromatic wines that are strongly influenced by the terroir of the vineyard sites where they are grown. This study looked at musts and young wines from ‘Gewürztraminer’ grapes harvested in seven distinct vineyards near Tramin and then processed at Cantina di Termeno, minimizing winemaking protocol variability. Samples were profiled using bidimensional gas chromatography–time-of-flight mass spectrometry, liquid chromatography coupled to electrochemical detection, and near-IR spectrometry. The data were subjected to Principle Component Analysis and Hierarchical Clustering Analysis. Sensory discriminant testing was undertaken using the sorting method with a semi-trained panel, and the data were processed using Multidimensional Scaling. Seven must/wine pairs could be distinguished based on their untargeted volatilome profiles and on sensory evaluation. As expected, there were greater differences in the volatile compounds between the wines than between the musts. The wines from vineyards 4 and 5 were nonetheless quite homogenous in terms of chemical and sensory analyses, as were the wines from vineyards 1 and 3. For the phenolic profile, differences were noted between the musts and wines of vineyards 2, 3, and 4, but the musts from vineyards 5 and 7 were similar. Sensory analysis showed the wines from vineyards 6 and 7 to be distinct from the rest. These results reinforce that the composition of ‘Gewürztraminer’ musts and wines is strongly determined by vineyard site, even in a small geographic area with high variability of the terroir (soil and microclimate), and that these differences are apparent in the flavours and aromas of the finished wines. Further confirmation would require a larger sample of wines, preferably from several vintages.

A predictive model of spatial Eca variability in the vineyard to support the monitoring of plant status

[lwp_divi_breadcrumbs home_text="IVES" use_before_icon="on" before_icon="||divi||400" module_id="publication-ariane" _builder_version="4.19.4" _module_preset="default" module_text_align="center" module_font_size="16px" text_orientation="center"...

Second pruning as a strategy to delay maturation in cv. ‘Touriga nacional’ in the Portuguese Douro region

The advance in maturation of wine grapes is an important climate change risk related effect that could affect warm regions like Portuguese Douro Wine Region. Indeed, the climate analysis over the past years registered a decrease in the precipitation, significant higher average temperatures, and a more frequent occurrence of extreme weather events, including heat waves. In these conditions the length from anthesis until maturation is shortened and the uncoupling of technical and phenolic maturity results in berries with higher sugar concentration (and lower acidity), but lower anthocyanins, tannins, and total phenolic concentration, which produce unbalanced wines.
In this work, an innovative strategy of crop forcing, based on forcing vine regrowth after a second pruning of green shoots, was tested, aimed at delaying ripening until the temperature becomes lower and, therefore, preventing acidity loss and increasing anthocyanin-to-sugar ratio. The experiments were conducted in 2019 and 2020 in a commercial vineyard of ‘Touriga Nacional’ located in the Douro Region. Crop forcing was conducted 15 (CF1) to 30 (CF2) days after fruit set. Vines pruned with conventional methods were used as control (CF0). Results confirmed that fruit ripening was shifted from the hot season (August/September), until a cooler period (October through early-November). At harvest, grapevine berries from CF1 and CF2 presented lower pH and higher acidity, than control, with no significant differences in colour intensity and phenolic levels composition. Sugar content was lower in CF2-treated vines in both seasons. However, in CF-treated vines the number and size of clusters were significantly lower (up to 88% reduction) than in control plants. A metabolomics analysis of mature berries from CF-treated vines and control is underway. Crop forcing was indeed effective in producing a more balance berry composition but severely reduced grapevine yield,

Analysis of Cabernet Sauvignon and Aglianico winegrape (V. vinifera L.) responses to different pedo-climatic environments in southern Italy

Water deficit is one of the most important effects of climate change able to affect agricultural sectors. In general, it determines a reduction in biomass production, and for some plants, as in the case of grapevine, it can endorse fruit quality. The monitoring and management of plant water stress in the vineyard

Grapevine yield-gap: identification of environmental limitations by soil and climate zoning in Languedoc-Roussillon region (south of France)

Grapevine yield has been historically overlooked, assuming a strong trade-off between grape yield and wine quality. At present, menaced by climate change, many vineyards in Southern France are far from the quality label threshold, becoming grapevine yield-gaps a major subject of concern. Although yield-gaps are well studied in arable crops, we know very little about grapevine yield-gaps. In the present study, we analysed the environmental component of grapevine yield-gaps linked to climate and soil resources in the Languedoc Roussillon. We used SAFRAN data and IGP Pays d’Oc wine yields from 2010 to 2018. We selected climate and soil indicators proving to have a significant effect on average wine yield-gaps at the municipality scale. The most significant factors of grapevine yield were the Soil Available Water Capacity; followed by the Huglin Index and the Climatic Dryness Index. The Days of Frost; the Soil pH; and the Very Hot Days were also significant. Then, we clustered geographical zones presenting similar indicators, facilitating the identification of resources yield-gaps. We discussed the number of zones with the experts of IGP Pays d’Oc label, obtaining 7 zones with similar limitations for grapevine yield. Finally, we analysed the main resources causing yield-gaps and the grapevine varieties planted on each zone. Mapping grapevine resource yield-gaps are the first stage for understanding grapevine yield-gaps at the regional scale.