Recercat. Dipósit de la Recerca de Catalunya
Publicación
Contribución de congreso (ConferenceOutput)
A novel AI-based framework for forecasting root-zone soil moisture and optimizing irrigation strategies using the AI4WATER digital twin
Global warming intensifies water scarcity, posing significant challenges to sustainable agriculture worldwide. As rising temperatures and shifting precipitation patterns exacerbate drought conditions, the need for timely and effective irrigation becomes paramount to ensure crop productivity and resource conservation. Surface Soil Moisture (SSM) and Root-zone soil moisture (RZSM), defined as the water content within the soil layers where plant roots actively extract water—typically at depths of 0–15 cm for cereals and 40–100 cm below the surface for fruit trees—is a critical parameter for precision irrigation. Accurate forecasting of SSM and RZSM enables farmers to optimize water use, reduce waste, and enhance crop resilience in arid and semi-arid regions. This work proposes a novel AI-based framework for forecasting SSM and RZSM over a 14-day horizon and addressing irrigation optimization as a multi-constrained problem across the Iberian Peninsula, a region increasingly vulnerable to water stress. Building on the AI4WATER project—a Digital Twin (DT) initiative led by the Universitat Politècnica de Catalunya (UPC)—this framework integrates advanced computational techniques with real-world data to support sustainable agricultural practices., The AI4WATER project, funded by the Spanish Ministry of Science and Innovation (MCIN/AEI, project code TED2021-131877B-I00), Peer Reviewed, Postprint (published version)