Detectıon Of Faılure Potentıal In The Clean Water Network Usıng Multı-Crıterıa Decısıon-Makıng (Mcdm) Methods: The Case Of Yenişehir, Mersin

Türkiye was once regarded as a country rich in water resources; however, rapid population growth, climate change, and increasing water demand have accelerated the transition toward water scarcity. Water losses in distribution networks have reached critical levels, with annual averages often exceeding 50%. Mitigating these physical leakages through proactive infrastructure management is therefore essential. The primary aim of this study is to develop a robust spatial susceptibility model for water distribution network failures using an integrated Analytical Hierarchy Process (AHP) and Frequency Ratio (FR) approach. Focusing on the Yenişehir district of Mersin Province, nine environmental and anthropogenic conditioning factors were identified for failure prediction: soil type, lithology, slope, elevation, aspect, building density, population density, lineament density, and distance to roads. These factors were spatially analyzed and weighted by coupling data-driven FR statistics with expert-based AHP evaluations. The developed model was validated against an independent testing dataset comprising 30% of the actual historical failure inventory, achieving a reliable predictive accuracy rate of 92.40%. Ultimately, this research presents an effective geospatial decision support tool for municipalities to prioritize maintenance interventions and minimize urban water losses.