HYDROLOGICAL AND HYDRAULIC MODELING FOR FLOOD RISK ASSESSMENT IN WADI YESHBUM- SHABWA GOVERNORATE, IN YEMEN

Abstract

Flash floods in arid regions pose severe threats to life and infrastructure, yet quantitative hydrological data are frequently absent. This study presents the first hydrologic model for the ungauged Wadi Yeshbum basin (337.48 km²) in Shabwah Governorate, Yemen, using the HEC-HMS model within a GIS framework. Morphometric parameters were derived from a 30 m SRTM DEM, revealing a steep average slope (10.5%), a well-developed drainage network (drainage density 1.8 km/km²), and a short time of concentration (4.38 h). The spatially distributed curve number (CN) was computed from FAO soil data and ESA CCI land cover, yielding a weighted average of 83, indicative of high runoff potential. Historical daily rainfall (1994–2024) from CHIRPS was used to fit the Gumbel distribution; 24-h design storms for return periods of 20, 50, and 100 years were estimated as 34.3, 40.5, and 44.9 mm, respectively. The SCS unit hydrograph method was applied to transform excess rainfall into direct runoff. Simulated peak discharges were 733.8, 867.5, and 963.7 m³/s with corresponding runoff volumes of 8.88, 11.72, and 13.03 million m³. Because no observed streamflow data were available for calibration, a generalized likelihood uncertainty estimation (GLUE) based on 500 Monte Carlo runs gave a 95% confidence interval for the 100-year peak discharge of 712.7–1177.7 m³/s. Sensitivity analysis revealed that peak discharge is highly sensitive to CN (coefficient 3–4) and moderately sensitive to the time of concentration (coefficient ~0.7–1.0). The results demonstrate that the basin exhibits a rapid, flashy hydrological response and that a large fraction (77–86%) of extreme rainfall becomes direct runoff. The modeling framework is transferable to other data-scarce arid basins and provides quantitative baseline information for flood mitigation planning in Yemen, though all results should be considered uncalibrated screening-level estimates.