Scalable quantum non-local neural network with electrical Eel foraging optimisation for groundwater quality evaluation and irrigation index prediction

In areas with irrigation, agriculture is crucial for global food security and local economies. This study proposes a novel deep learning framework for assessing groundwater (GW) quality and predicting its suitability for irrigation in the El Kharga Oasis, Egypt. The model predicts the key irrigation water quality indices (IWQIs), including IWQI, RSC, SSP, KI, SAR, and PS by using the Scalable Quantum Non-local Neural Network (QNL-Net) along with Electric Eel Foraging Optimisation (EEFO). Preprocessing used correlation coefficients and min-max normalisation to capture hydrogeochemical patterns, including TDS, EC, pH, and major ions. The results indicate that most GW samples are classified as ‘no restriction (93.6%)’, ‘excellent (100%)’, and ‘good (60%)’ for irrigation. EEFO optimised network parameters, while QNL-Net captured non-linear interactions, achieving better accuracy than conventional methods. This model offers valuable tools providing a decision-support tool for groundwater management in arid regions, ensuring irrigation feasibility and long-term agricultural sustainability.