terclim by ICS banner
IVES 9 IVES Conference Series 9 GiESCO 9 Deep learning based models for grapevine phenology

Deep learning based models for grapevine phenology


Context and purpose of the study – the phenological evolution is a crucial aspect of grapevine growth and development. Accurate detection of phenological stages can improve vineyard management, leading to better crop yield and quality traits. However, traditional methods of phenological tracking such as on-site observations are time-consuming and labour-intensive. This work proposes a scalable data-driven method to automatically detect key phenological stages of grapevines using satellite data. Our approach applies to vast areas because it solely relies on open and satellite data having global coverage without requiring any in-field data from weather stations or other sensors making the approach extensible to other areas.

Material and methods we leveraged historical phenological observations and developed a supervised deep-learning model that uses the land surface temperature estimated by the Copernicus Sentinel-3 satellite to estimate the current phenological stage at the parcel level. We compared the performances of our model with traditional approach based on Growing Degrees Days (GDD).

Results – we train our algorithm on manually collected phenological observations of four winegrape cultivars in three Europeanvineyards (Italy, Spain, and Portugal) from 2017 to 2022. Preliminary results indicated that our deep learning phenology model outperforms the traditional methods based on GDD, decreasing the Mean Absolute Error from 33.8 to 7.8 days (-76.5%).


Publication date: July 5, 2023

Issue: GiESCO 2023

Type: Poster


Federico OLDANI1*, Dario SALZA1, Giacomo BLANCO1, Claudio ROSSI1, Boris BASILE2*, Fabrizio CARTENI2, Núria PÉREZ-ZANÓN3, Antonio DENTE4, Fernando ALVES5, Joana VALENTE5, Montse TORRES6, Carlos EZQUERRA6, Rosa ARAUJO7

1LINKS Foundation, Turin, Italy
2Department of Agricultural Sciences, University of Naples Federico II, Portici (Napoli), Italy
3Barcelona Supercomputing Center, Barcelona, Spain
4Mastroberardino, Atripalda (Avellino), Italy
5Symington Family Estates, V. N. Gaia, Portugal
6Familia Torres, Vilafranca del Penedès, Spain
7Eurecat, Centre Tecnològic de Catalunya, Barcelona, Spain

Contact the author*


satellite imagery, earth observation, machine learning, Sentinel-3, Copernicus, climate change


GiESCO | GIESCO 2023 | IVES Conference Series


Related articles…

Impact of some agronomic practices on grape skins anthocyanin content

Wine colour is the first quality characteristic to be assessed, especially regarding red wines. Anthocyanins are very well known to be the main responsible compounds for red wine colour. Red cultivars can synthesize and accumulate anthocyanins in berry skin to express their colour. However, anthocyanin accumulation is often influenced by a series of factors, such as genetic regulation, phytohormones, environmental conditions and viticultural management.

Characterization of Glycosidically Bound Aroma Compounds of País cv. grapes of different Chilean zones

Úbeda-Aguilera, C., a Callejón, R. M., b Peña-Neira, A c. a Instituto de Ciencias Biomédicas, Facultad de Ciencias, Universidad Autónoma de Chile, Chile b Área de Nutrición y Bromatología. Facultad de Farmacia. Universidad de Sevilla. C/ P. García González nº 2, E- 41012. Sevilla. Spain c Department of Agro-Industry and Enology, Faculty of Agronomical Sciences, University of Chile, Post Office Box 1004, Santa Rosa 11315, La Pintana, Santiago, Chile País grape has been estimated to arrive to Chile almost 500 years ago, being the first strain grown in this country. Traditionally, this grape has been used to mix with other varieties, to produce poor quality wines, but today is beginning to be used in the production of high quality wines. However, very little is known about the chemical characteristics of this variety.

Ageing of Sauvignon Blanc white wines with Specific Inactivated Dry Yeasts: Effect on physical and chemical characteristics

Del Barrio-Galán, R.a, b, Gómez-Parrini, A.a, Peña-Neira, A.b a Lallemand Inc. Chile y Compañía Limitada, Rosario Norte 407, piso 6, Las condes, Santiago, Chile b Department of Agro-Industry and Enology, Faculty of Agronomical Sciences, University of Chile, Post Office Box 1004, Santa Rosa 11315, La Pintana, Santiago, Chile It is well known that polysaccharides, mainly mannoproteins, play an important role on physical, chemical and sensory quality of wines. The ageing of white wines on lees is used in order to release higher amounts of polysaccharides by the autolytic processes in order to obtain higher-quality wines. However, this technique is too slow, because the temperature and pH conditions are not the most suitable for this process. In addition, it can also involve certain disadvantages such as a greater demand on winery resources, a longer period of wine storage, the appearance of reduction notes and some microbiological alterations.

Effect of ageing with Specific Inactivated Dry Yeasts on the volatile composition of Sauvignon Blanc and Carménère wines

Úbeda-Aguilera, C a, b, Peña-Neira, A.b Del Barrio-Galán, R.b, c a Biomedical Sciences Institute, Science Faculty, Universidad Autónoma de Chile, Chile. b Department of Agro-Industry and Enology, Faculty of Agronomical Sciences, University of Chile, Post Office Box 1004, Santa Rosa 11315, La Pintana, Santiago, Chile c Lallemand Inc. Chile y Compañía Limitada, Rosario Norte 407, piso 6, Las Condes, Santiago, Chile The wine is a complex matrix made up of several compounds which can interact among themselves throughout the wine ageing process, thereby modifying their sensorial characteristics. It is well known that during ageing of wines on lees, polysaccharides (mainly mannoproteins) can be released and can interact with the aromatic fraction modifying its volatility.

New molecular evidence of wine yeast-bacteria interaction unraveled by untargeted metabolomic profiling

Bacterial malolactic fermentation (MLF) has a considerable impact on wine quality. The yeast strain used for primary fermentation can consistently stimulate (MLF+ phenotype) or inhibit (MLF- phenotype) malolactic bacteria and the MLF process as a function of numerous winemaking practices, but the molecular evidence behind still remains a mystery. In this study, such evidence was elucidated by the direct comparison of extracellular metabolic profiles of MLF+ and MLF- yeast phenotypes. Untargeted metabolomics combining ultrahigh-resolution FT-ICR-MS analysis, powerful machine learning methods and a comprehensive wine metabolite database, discovered around 800 putative biomarkers and 2500 unknown masses involved in phenotypic distinction.