SoPoSIS: An AI-Integrated Framework Linking Evapotranspiration and Photovoltaic Power Predictions for Sustainable Irrigation

Abstract

Solar smart irrigation is one of the significant techniques to enhance water and energy resources in sustainable agriculture. Despite this, current systems in common use typically treat water and energy separately with few operational indicators to predict irrigation adequacy, particularly in tropical growing environments. This research aims to design and validate the Solar-Powered Smart Irrigation System (SoPoSIS) as a holistic and predictive tool of irrigation management. Meteorological data and historical solar energy production were collected from field experiments in tropical locations for this research. The system is constructed based on an Extreme Gradient Boosting model for estimating reference evapotranspiration (ETo) and a Hybrid Autoregressive + Wavelet Gated Recurrent Unit residual model for predicting photovoltaic (PV) power output. The performance of the system was assessed by the use of Predictive Irrigation Adequacy Index (PIAI) to measure the suitability of PV power prediction for crop water demand. For the ETo model, a mean absolute error (MAE) of 0.006 mm day−1 and accuracy of 98% were obtained, whereas for the PV model, it was an MAE of 188.50 watts with an accuracy level of 89.9%. The equipment could irrigate up to 51.6 m3 of water per application, whereas crop water requirement was between 2.6 and 3.0 m3 ha−1. The PIAI values indicated that the irrigation was right in all weather conditions. These results show that SoPoSIS offers an operational means of integrated water and energy management and supports resource-efficient and low-carbon farming in tropical regions.



Received: 6 December 2025 | Revised: 7 April 2026 | Accepted: 19 June 2026



Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.



Data Availability Statement

Data are available from the corresponding author upon reasonable request.



Author Contribution Statement

Nurmalitasari Nurmalitasari: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Funding acquisition. Nurchim Nurchim: Conceptualization, Methodology, Software, Validation, Formal analysis, Resources, Writing – original draft, Writing – review & editing, Supervision. Retna Dewi Lestari: Validation, Writing – original draft, Writing – review & editing. Zalizah Awang Long: Validation, Writing – review & editing, Supervision.