Author(s): Maria Belen Pedoja Balparda; Juan Antonio Rodriguez Diaz; Maria Pilar Montesinos Barrios; Rafael Gonzalez Perea

Linked Author(s):

Keywords: Crop modelling; Decision support system; Precision irrigation; Satellite

Abstract: Efficient water management is critical for environmental sustainability and food security. This paper presents the development of a Digital Twin (DT) for irrigated agriculture, implemented as a QGIS plugin. The DT integrates four functional components: hydraulic network simulation using EPANET, dynamic irrigation management, crop growth modelling using the FAO-AquaCrop model, and a Canopy Cover (CC) data assimilation module fed by remote sensing data. Three irrigation strategies were simulated for a broccoli plot in Tarazona de la Mancha (Spain) under contrasting climatic scenarios. To reduce uncertainty in crop growth simulations, a Multilayer Perceptron neural network was developed to estimate CC using high-resolution drone imagery and Planet satellite data (R² = 0.94 and RMSE = 0.085). The CC estimated was introduced in AquaCrop at multiple points during the growing season. Simulation results showed that Controlled Deficit Irrigation strategy achieved an 18.8% water saving compared to Maximum Irrigation Requirements strategy with only a 1.1% yield reduction Simulations under dry and wet historical years evidenced the system's sensitivity to climatic extremes, with yield reductions of up to 35.5% under excess rainfall conditions.. These findings demonstrate the DT's ability to identify optimal irrigation strategies and quantify risks associated with climatic variability, providing a robust decision-support framework for precision irrigation management.

DOI: https://doi.org/10.64697/iahr.proc.hic2026.304

Year: 2026