Flood risk assessment using modern models integrated with GIS for susceptibility mapping of the M'zab Valley watershed (Algeria)

A flood susceptibility mapping study of the M'Zab Valley, Algeria, revealed high predictive accuracy for two modeling approaches: frequency ratio (FR) and logistic regression (LR). Both models achieved over 92% success rates in validation tests, with FR and LR predicting 86% and 88.1%, respectively. The research enhanced bivariate probability analysis by integrating multiple variables—including distance from rivers, rainfall, soil type, land use/land cover, elevation, drainage density, flow accumulation, and slope—to assess their influence on flooding. By training models with 70% of flood inventory data and validating with the remaining 30%, the study demonstrated robust performance in susceptibility mapping, offering critical insights for flood risk management in this UNESCO World Heritage Site.

Brahim, Z., Boualem, R., & Azzeddine, M. (2024). Flood risk assessment using modern models integrated with GIS for susceptibility mapping of the M’zab Valley watershed (Algeria). Water Science & Technology, 90(8), 2352–2366. https://doi.org/10.2166/wst.2024.343