A Preliminary Evaluation of Season-ahead Flood Risks Globally

Abstract:

Globally, flood catastrophes lead all natural hazards in terms of impacts on society, causing billions of dollars of damages each year. While short-term flood warning systems are improving in number and sophistication, forecasting systems on the order of months to seasons are a rarity, yet may lead to further disaster preparedness. In this study, prospects of season-ahead flood forecasts are investigated by examining inter-annual climate variability on seasonal maximum floods, particularly how ENSO and other large-scale phenomena may modulate discharge and flood severity. Predictability of main seasonal floods is also assessed globally using PCR-GLOBWB simulations with large- and local- climate drivers, and validated with GRDC streamflow. Skillful prediction can lead to season-ahead flood probabilities, flood extent, estimated damages, and eventually integrate into early warning systems for informed advanced planning and management. This is especially attractive for regions with limited observations and/or little capacity to develop early warning flood systems.