Integrated Rainfall Modeling and Hydrometeorological Alert System for Extreme Flooding in Kosovo

Flash floods pose a significant threat to infrastructure and public safety in Kosovo, particularly in urban and river environments. This study analyzes a winter flash flood event that struck the small catchment of Skenderaj on January 19, 2023. The main objective was to develop a flash flood warning system by integrating high-resolution atmospheric modeling, hydrological simulations, and hydrometeorological hazard modeling. A 2-km nested configuration of the ARW model successfully captured the spatial distribution of intense rainfall and localized flooding. The integration of GloFAS forecasts and ERA5 reanalysis data enhanced flood risk assessment and comparative analysis with the EFAS platform, revealing strong agreement in identifying high-risk zones and thereby validating the reliability of the developed system. Additionally, the Flash Flood Potential Index (FFPI) enabled high-resolution mapping of flash flood susceptibility using topographic, soil, land cover, vegetation, and satellite-derived precipitation data. The results for the Klina catchment revealed the precise identification of high-risk zones, allowing for both retrospective and predictive discharge estimation. This geospatially driven approach provides a scalable and operationally viable tool for enhancing regional early warning systems. These findings underscore the effectiveness of coupled weather-hydrology models and geospatial risk assessment tools in enhancing early warning systems, with potential for broader application and further refinement through sensitivity studies.