Groundwater quality assessment using artificial neural networks and clay treatment: A case study in Nineveh Governorate, Iraq

This study aimed to evaluate groundwater quality in ten selected wells located in rural villages in the northern part of Ninewa Governorate (Iraq). An intelligent model based on a Multilayer Artificial Neural Network (MLP) was developed and applied in this study. The simulation results and predictions for the groundwater quality index showed high accuracy. The correlation coefficient (R²) for the test dataset was ۰.۹۷۱, while the Mean Squared Error (MSE) was ۰.۰۰۰۴. Furthermore, the chemical data from groundwater samples were used to design an advanced water treatment protocol. A clay sample from the Hammam al-Alil area was used, after treatment with hydrochloric acid and chemical modification with aluminum chloride. The results showed that acid treatment completely removed calcite from the samples, thereby doubling their surface area to ۱۵۳.۹۷ m²/g from its initial value of ۶۸.۸۷ m²/g. In contrast, the surface area of the aluminum chloride-treated samples reached ۹۰.۳۱ m²/g, while maintaining their stable fibrous structure. The removal rates of sulfate and magnesium TDS from the groundwater were ۴۴%, ۷۹%, and ۷۰%, respectively. This study concludes that integrating an AI-powered prediction model with chemically modified natural clay represents an innovative, integrated, and sustainable strategy for groundwater resource management and rehabilitation.

کلیدواژه ها:

ANN ، Water Quality ، Drinking ، Groundwater ، Clay ، AI ، X-ray ، SEM