Machine learning-based assessment of irrigation water quality in a multilayer aquifer: a case study from the Oued Mya Basin, Algeria
摘要
Groundwater is a critical resource in arid regions, and the Northwestern Sahara Aquifer System (NWSAS), shared by Algeria, Libya, and Tunisia, represents one of the world’s largest transboundary aquifers. The Oued Mya Basin take a part of this aquifer with complex geological structures. The increasing groundwater exploitation in this zone have raised concerns about water quality and aquifer interactions. This study aims to characterize groundwater chemistry, identify potential interconnections between aquifers, and evaluate irrigation suitability to support sustainable water management. A total of 48 borehole samples, 27 from the Continental Intercalaire (CI) and 21 from the Complex Terminal (CT) were analyzed using the Self-Organizing Map (SOM) method to classify waters and determine the dominant geochemical processes. Three hydrochemical groups were identified: Complex Terminal waters (25%), mixed Complex Terminal–Continental Intercalaire waters (27%), and Continental Intercalaire waters (47.9%). Strong Na⁺-Cl⁻ correlations indicate halite dissolution as the principal geochemical process, while aquifer exchanges may occur through fractures or irrigation-induced infiltration. Continental Intercalaire waters generally exhibit better quality than Complex Terminal waters but belong to the same hydrogeochemical family. The findings highlight the need for strengthened groundwater monitoring, improved irrigation practices, and enhanced regional cooperation to ensure the sustainable and equitable management of transboundary water resources within the Northwestern Sahara Aquifer System.