Abstract/Summary

Anticipatory actions are increasingly being taken before an extreme flood event to reduce the impacts on lives and livelihoods. Local contextualised information is required to support real-time local decisions on where and when to act and what anticipatory actions to take. This study defines an impact-based early warning trigger system that integrates flood forecasts with livelihood information, such as crop calendars, to target anticipatory actions better. We demonstrate the application of this trigger system using a flood case study from the Katakwi District in Uganda. First, we integrate information on the local crop cycles with the flood forecasts to define the impact-based trigger system. Second, we verify the impact-based system using historical flood impact information and then compare it with the existing hazard-based system in the context of humanitarian decisions. Study findings show that the impact-based trigger system has an improved probability of flood detection compared to the hazard-based system. The number of missed events are fewer in the impact-based system while the trigger dates are similar in both systems. In a humanitarian context, the two systems trigger anticipatory actions at the same time. However, the impact-based trigger system can be further investigated in a different context (e.g., for livelihood protection) to assess the value of the local information. The impact-based system could also provide a valuable tool to validate the existing hazard-based system, which builds more confidence in its use in informing anticipatory actions. The study findings should therefore open avenues for further dialogue on what the impact-based trigger system could mean within the broader Forecast-based Action landscape towards building the resilience of at-risk communities.