Operational impact-based flood early warning in Lao PDR and Cambodia

Abstract

Operational flood forecasting and early warning systems are key tools for disaster risk prevention, with implementation rates steadily increasing worldwide. Yet, most operational systems focus on detecting only the hazard characteristics of upcoming floods, rather than the likely impacts on the society, which are the variables of highest interest for warning centers and emergency responders. This work describes the implementation of FloodPROOFS, a physically based modelling framework to predict the occurrence and the consequent impacts of riverine floods in five river basins in Cambodia and in Lao People's Democratic Republic. The system is updated twice daily with four numerical weather predictions for the next 5 days. It uses a modeling cascade to produce estimates of river discharges, water levels, exceedance of flood thresholds, inundation depth and extent, socio-economic impacts for seven exposure categories, including a novel approach for the automated estimation of impact-based warnings at the district level. The latter are included in a warning bulletin application co-designed with the two national meteo-hydrological services for daily monitoring and dissemination of alert messages ahead of impending disasters. The system is operational from summer 2024 and has already shown significant value in rising preparedness ahead of some moderate-magnitude floods, including the floods of the Prek Thnot river near Phnom Penh in July 2024.