A model for impact-based flood early warning and anticipatory actions in Uganda

Among the many disasters, floods are the most common disaster worldwide. The number of flood events worldwide has increased by 23% between 2000 and 2019. The trend is expected to increase due to climate variability and other environmental factors. Efforts to reduce the impacts of extreme events such as floods have been emphasised through various multilateral frameworks, resulting in the development of early warning systems. Technological advancement has also contributed to significant improvement in forecasting science leading to more accurate predictions and improved forecast skills. Despite these improvements, more focus has been on early warning of physical risks and less on incorporating the needs of the most at-risk populations and early action disaster responders. Thus, early warning systems (EWS) should be people-centred to ensure that at-risk populations can access tailored early warning information to inform their preparedness actions and protect their lives and livelihoods. The potential for early warning information (EWI) can be achieved if all the components of a people-centred early warning system are implemented through an integrated approach that involves the at-risk population. Therefore, there is a need to redefine the development of EWS by shifting from top-down to more bottom-up community-driven approaches. To address this gap, the research developed an impact-based flood early warning trigger system for anticipatory action through a community-led process. As shown through community and disaster management practitioners’ engagements, a more coordinated institutional response is needed to understand the gaps in the provision and use of EWI at the local level. Local context-specific information can also be used to verify forecast information to make them more acceptable in informing early and anticipatory actions in data-scarce regions. Further, such information could enhance the existing hazard-based systems by redefining the design of trigger thresholds and early actions. Overall, this thesis has shown that community-led approaches based on holistic engagements can effectively ensure EWSs are locally targeted to inform local anticipatory actions.