Reliable hydrological data from ephemeral desert streams are essential for infrastructure planning and flood risk assessment, yet such data are scarce in the northern Dead Sea basin due to geomorphic instability, limited gauging stations, and frequent failure of conventional instruments during extreme floods. This study presents the development and validation of a cost-effective and flexible method for flood monitoring and discharge reconstruction that overcomes key limitations of traditional hydrometric approaches in highly dynamic streams. The method integrates time-lapse video documentation of flood events, high-resolution drone-based photogrammetric surveys, and three-dimensional hydraulic modelling. Time-lapse cameras were deployed along several Dead Sea tributaries, while pre- and post-flood digital terrain models (DTMs) were generated using UAV photogrammetry. Flood flows were simulated with a 3D HEC-RAS model using synthetic hydrographs, producing an “atlas of discharges” that visually represents expected water surfaces for a range of flows from the camera viewpoint. Observed flood imagery was then systematically compared to the atlas to reconstruct flood hydrographs. The method was validated against measured hydrographs from the Hydrological Service of Israel stations for flood events during winter 2025. Results show excellent agreement in hydrograph shape and timing, with Pearson correlation coefficients of 0.95 and 0.98, indicating near-perfect synchronization of rising limbs, peak discharge, and recession. The proposed approach provides robust estimates of flood peaks and timing in channels where direct measurements are impractical, offering a valuable complementary tool for hydrological analysis and infrastructure design in fluvial environments.