TY - JOUR
T1 - Can Grapevine Leaf Water Potential Be Modelled from Physiological and Meteorological Variables? A Machine Learning Approach
AU - Damásio, Miguel
AU - Barbosa, Miguel
AU - Deus, João
AU - Fernandes, Eduardo
AU - Leitão, André
AU - Albino, Luís
AU - Fonseca, Filipe
AU - Silvestre, José
N1 - Funding Information:
This research work was developed in the context of the project AI4RealAg—Artificial Intelligence and Data Science solutions for the implementation and democratisation of digital agriculture—which was funded by both the Operacional Competitividade e Internacionalização program (POCI-01-0247-FEDER-069670) and the Operacional Regional de Lisboa 2020 program (LISBOA-01-0247-FEDER-069670). This project aims to help the agricultural sector transition into the new digital era through the adoption of artificial intelligence methods—in particular, techniques of data science. This project was formed by a consortium of three entities: Promotor: SISCOG, a world-leading technological company in the development of AI applications in the transportation sector, with more than 140 highly experienced technical professionals; Co-Promotor: INIAV, a Portuguese national institute which has domain expertise on agriculture, experimental farms, equipment, and necessary laboratories for data validation; Co-Promotor: BEYOND VISION, a technological company which specialises in the production of drones, that has vast experience in the area of image processing and fusion (in both the software and hardware ends), being capable of integrating data from multiple sensors.
Publisher Copyright:
© 2023 by the authors.
PY - 2023/12
Y1 - 2023/12
N2 - Climate change is affecting global viticulture, increasing heatwaves and drought. Precision irrigation, supported by robust water status indicators (WSIs), is inevitable in most of the Mediterranean basin. One of the most reliable WSIs is the leaf water potential ((Formula presented.)), which is determined via an intrusive and time-consuming method. The aim of this work is to discern the most effective variables that are correlated with plants’ water status and identify the variables that better predict (Formula presented.). Five grapevine varieties grown in the Alentejo region (Portugal) were selected and subjected to three irrigation treatments, starting in 2018: full irrigation (FI), deficit irrigation (DI), and no irrigation (NI). Plant monitoring was performed in 2023. Measurements included stomatal conductance ((Formula presented.)), predawn water potential (Formula presented.), stem water potential ((Formula presented.)), thermal imaging, and meteorological data. The WSIs, namely (Formula presented.) and (Formula presented.), responded differently according to the irrigation treatment. (Formula presented.) measured at mid-morning (MM) and mid-day (MD) proved unable to discern between treatments. MM measurements presented the best correlations between WSIs. (Formula presented.) showed the best correlations between the other WSIs, and consequently the best predictive capability to estimate (Formula presented.). Machine learning regression models were trained on meteorological, thermal, and (Formula presented.) data to predict (Formula presented.), with ensemble models showing a great performance (ExtraTrees: (Formula presented.), (Formula presented.) ; Gradient Boosting: (Formula presented.) ; (Formula presented.)).
AB - Climate change is affecting global viticulture, increasing heatwaves and drought. Precision irrigation, supported by robust water status indicators (WSIs), is inevitable in most of the Mediterranean basin. One of the most reliable WSIs is the leaf water potential ((Formula presented.)), which is determined via an intrusive and time-consuming method. The aim of this work is to discern the most effective variables that are correlated with plants’ water status and identify the variables that better predict (Formula presented.). Five grapevine varieties grown in the Alentejo region (Portugal) were selected and subjected to three irrigation treatments, starting in 2018: full irrigation (FI), deficit irrigation (DI), and no irrigation (NI). Plant monitoring was performed in 2023. Measurements included stomatal conductance ((Formula presented.)), predawn water potential (Formula presented.), stem water potential ((Formula presented.)), thermal imaging, and meteorological data. The WSIs, namely (Formula presented.) and (Formula presented.), responded differently according to the irrigation treatment. (Formula presented.) measured at mid-morning (MM) and mid-day (MD) proved unable to discern between treatments. MM measurements presented the best correlations between WSIs. (Formula presented.) showed the best correlations between the other WSIs, and consequently the best predictive capability to estimate (Formula presented.). Machine learning regression models were trained on meteorological, thermal, and (Formula presented.) data to predict (Formula presented.), with ensemble models showing a great performance (ExtraTrees: (Formula presented.), (Formula presented.) ; Gradient Boosting: (Formula presented.) ; (Formula presented.)).
KW - modelling
KW - precision irrigation
KW - predawn leaf water potential
KW - Vitis vinifera
KW - water status indicators
UR - http://www.scopus.com/inward/record.url?scp=85180671980&partnerID=8YFLogxK
U2 - 10.3390/plants12244142
DO - 10.3390/plants12244142
M3 - Article
AN - SCOPUS:85180671980
SN - 2223-7747
VL - 12
JO - Plants
JF - Plants
IS - 24
M1 - 4142
ER -
