Enhancing the prediction of groundwater quality index in semi-arid regions using a novel ANN-based hybrid arctic puffin-hippopotamus optimization model

Study region: The west of Minia Governorate, Egypt. Study focus: This study presents a novel hybrid arctic puffin–hippopotamus optimization (HPHO) algorithm combined with an artificial neural network (ANN) to improve irrigation water quality index (IWQI) predictions in semi-arid areas. A total of 88 groundwater samples from the study region were collected and analyzed using statistical and graphical tools. Hydrochemical facies showed a dominance of calcium bicarbonate (CaHCO₃) water, indicating the role of carbonate dissolution. Principal component analysis (PCA) and hierarchical cluster analysis (HCA) revealed that groundwater quality was influenced by both geogenic and anthropogenic factors. IWQI values exhibited good to excellent classifications, with lower-quality zones linked to higher contamination. New hydrological insights for the region: Most groundwater samples were found suitable for irrigation, while only a few showed poor-quality classifications due to contamination. The proposed HPHO–ANN model significantly outperformed other artificial intelligence models, achieving high predictive accuracy (correlation coefficient = 0.99). Sensitivity analysis revealed that sodium percentage (∼14.25 %), saturated sodium percentage (∼8.5 %), and sodium adsorption ratio (∼7.0 %) were the most influential variables affecting IWQI prediction. The HPHO–ANN framework supports sustainable development goals (SDGs) relevant to clean water (SDG 6), climate action (SDG 13), and life on land (SDG 15). This study can offer a practical and high-precision tool for groundwater quality forecasting, contributing to improved water management, agricultural planning, and environmental protection in semi-arid regions.