A step towards new irrigation scheduling strategies using plant-based measurements and mathematical modelling
Source: Irrigation Science, Volume 26, Number 6, September 2008 , pp. 505-517(13)
Abstract:Because of the increasing worldwide shortage of freshwater and costs of irrigation, a new plant-based irrigation scheduling method is proposed. In this method, two real-time plant-based measurements (sap flow and stem diameter variations) are used in combination with a mathematical water flow and storage model in order to predict the stem water potential. The amount of required irrigation water is derived from a time integration of the sap flow profile, while the timing of the irrigation is controlled based on a reference value for the predicted stem water potential. This reference value is derived from the relationship between midday values of maximum photosynthesis rates and stem water potential. Since modelling is an important part of the proposed methodology, a thorough mathematical analysis (identifiability analysis) of the model was performed. This analysis showed that an initial (offline) model calibration was needed based on measurements of sap flow, stem diameter variation and stem water potential. Regarding irrigation scheduling, however, only sap flow and stem diameter variation measurements are needed for online simulation and daily model calibration. Model calibration is performed using a moving window of 4 days of past data of stem diameter variations. The research tool STACI (Software Tool for Automatic Control of Irrigation) was used to optimally combine the continuous measurements, the mathematical modelling and the real-time irrigation scheduling. The new methodology was successfully tested in a pilot-scale setup with young potted apple trees (Malus domestica Borkh) and its performance was critically evaluated.
Document Type: Research Article
Affiliations: 1: Faculty of Bioscience Engineering, Department of Applied Ecology and Environmental Biology, Laboratory of Plant Ecology, Ghent University, Coupure links 653, 9000, Ghent, Belgium, Email: kathy.steppe@UGent.be 2: Faculty of Bioscience Engineering, Department of Applied Mathematics, Biometrics and Process Control, KERMIT: Knowledge Based Systems, Ghent University, Coupure links 653, 9000, Ghent, Belgium, Email: dirk.depauw@UGent.be 3: Faculty of Bioscience Engineering, Department of Applied Ecology and Environmental Biology, Laboratory of Plant Ecology, Ghent University, Coupure links 653, 9000, Ghent, Belgium, Email: raoul.lemeur@UGent.be
Publication date: September 1, 2008