Impact based forecasting to cope with riverine floods in Peruvian Andes—Amazon basin

Impact-based forecasting (IBF) represents a significant advancement in natural disaster risk management by focusing on the vulnerabilities of people, livelihoods, and assets. Here we introduce the methodology of PANDORA (Impact based forecasting to cope with riverine floods in the Huallaga river basin), a system designed to provide impact-based forecasts for a basin in the Andean-Amazon region of Peru. PANDORA integrates a large-scale hydrological model with precipitation forecasts resampled from historical meteorological data to produce 5-day probabilistic streamflow forecasts. These are compared against flood thresholds for 2, 5, and 10-year return periods, corresponding to moderate, heavy, and extreme severity levels. Subsequently, they are linked with key flood-exposed elements: (i) population, (ii) educational institutions, (iii) health centers, (iv) road networks, and (v) agricultural areas. Potential impacts can be assessed at various administrative levels, including districts, provinces, and departments. The system’s performance was evaluated during December 2023, when significant river floods occurred in the basin. Results show that flood events were primarily forecasted between December 27 and 30, while the IBFs indicated extreme severity (red level) for the exposed elements mainly on December 27, 30 and 31. These findings align with reports from the Information System for Response and Rehabilitation of Peru. Despite existing limitations, PANDORA is currently operational and demonstrates great potential to support local authorities in decision-making processes for flood risk management.