A high-resolution, human-in-the-loop flood early warning system for informal settlements

Residents of informal settlements globally are disproportionately exposed to severe floods and often lack access to early warning systems to prepare and respond. The project will develop a participatory flood early warning system (EWS) designed for informal settlements that both disseminates timely alerts and enables communities to report real-time experiences during and after flood events. The EWS leverages artificial intelligence to overcome long-standing barriers to flood risk mapping in data-sparse, ultra-dense urban environments, employs inclusive dissemination strategies to reach populations often excluded from traditional EWS, and improves post-flood aid targeting through real-time community feedback. The EWS will be piloted in Bangladesh while building a global coalition of EWS partners to continue co-development and scaling worldwide. By transforming EWS from one-way broadcasts into responsive platforms, this project seeks to systematically decrease flood-related morbidities by ensuring that recovery efforts reflect households’ real-time priorities.