The planning of infrastructures that cross flash-flooding streams presents a complex engineering challenge and require a deep understanding of flow characteristics during discrete flood events. This planning relies, among other factors, on the frequency and magnitude of extreme flow events, and necessitate a comprehensive hydrometric database, particularly of peak discharges. Existing measurement methods for reconstructing peak discharges in flash-flooding streams suffer from significant limitations, especially in arid regions. The scarcity of hydrometric stations, the challenges in maintaining them, and damage which they suffer during floods lead to a lack of continuous data. Additionally, the high spatial variability of precipitation results in many flood events going unrecorded by existing stations. Furthermore, in alluvial streams channel morphology changes, sediment erosion, and deposition processes dynamically influence the geometry of the streambed, making accurate flood event reconstruction challenging.
The research presented here proposes an innovative, efficient, and cost-effective method that integrates time-lapse cameras (TLC), drone-based photogrammetric surveys, and hydrological modeling. Preliminary results, compared with data measured at the hydrometric stations of the Hydrological Service, demonstrate the reliability of this method and its ability to bridge existing knowledge gaps. This method is not intended to replace classical approaches but provides a complementary and significant solution in data-scarce regions. Additionally, the obtained data are expected to improve the accuracy of peak discharge and recurrence interval calculations, serving as a key component in enhancing the hydrological and statistical models that guide planning and risk assessment for infrastructure.