Hybrid methodology for flood early warning indicators: Combining hydrodynamic modeling and community participation in Manizales, Colombia

This paper presents a hybrid methodology for the definition and validation of flash flood indicators and thresholds, combining hydrodynamic modeling with community participation. The proposed approach was developed within the framework of implementing Community-based Early Warning Systems in the Lusitania and Providencia neighborhoods -flood-prone areas- located in Manizales, Colombia. From the technical perspective, a hydraulic model was built in IBER, calibrated, and validated using historical records and field measurements. Multiple simulations were then performed under different boundary conditions to determine the water levels associated with flood flows. In addition, empirical hydrological methods reported in the literature were applied to estimate bankfull levels and discharges based on local morphological parameters. From the community perspective, participatory workshops were conducted to identify critical river sections and the corresponding overflow levels. The results show that community-defined levels aligned closely with the modeled values in Lusitania. In Providencia, however, the hydrological approach proved more accurate. These findings suggest that a hybrid approach -integrating technical analysis and local knowledge- can significantly enhance the accuracy and reliability of thresholds and indicators used in Flood Early Warning Systems, thereby improving their long-term sustainability.