EPFL explores hybrid early warning systems in East AfricaA new study
© 2026 EPFL
A new study led by researchers at EPFL examines how Indigenous Knowledge Systems (IKS) and scientific early warning approaches can be integrated to strengthen climate resilience and disaster preparedness in the Karamoja Cluster of Kenya and Uganda.
Published in Discover Sustainability, the research explores how pastoralist communities continue to rely on locally grounded environmental observations, including animal behavior, cloud formations, wind patterns, dew formation, and ecological indicators, to anticipate droughts, erratic rainfall, and climate-related shocks.
The study combines qualitative fieldwork, focus group discussions, stakeholder interviews, and policy analysis conducted across West Pokot (Kenya) and Karamoja (Uganda). Researchers examined how Indigenous and scientific Early Warning Systems (EWS) coexist, where they diverge, and how hybrid systems could improve climate-risk communication and anticipatory action.
The findings reveal important differences between the two countries. Kenya’s system is more technologically embedded, relying on SMS alerts, meteorological services, and institutional coordination mechanisms. Uganda’s approach remains more deeply rooted in oral traditions, community gatherings, and culturally transmitted environmental knowledge.
Despite these differences, both systems face common challenges, including language barriers, limited community feedback mechanisms, low literacy, and insufficient integration of Indigenous knowledge into formal disaster risk management frameworks.
The researchers propose a hybrid and community-centered model for Early Warning Systems that combines the spatial and temporal accuracy of scientific forecasting with the trust, contextual relevance, and cultural legitimacy of Indigenous knowledge systems. The study argues that integrating local and scientific knowledge can improve warning uptake, strengthen anticipatory action, and support more inclusive climate governance across vulnerable pastoral regions in Sub-Saharan Africa.
The study also contributes to broader international discussions on climate justice, participatory governance, and the decolonization of environmental knowledge systems. It emphasizes that effective Early Warning Systems are not only technical infrastructures, but also social systems shaped by trust, inclusion, and local participation.
The research was conducted by Jean-Claude Baraka Munyaka, Mohammed Hlal, Joseph Timu Lolemtum, Seyid Abdellahi Ebnou Abdem, Olivier Gallay, and Jérôme Chenal from EPFL, UM6P, Turkana University College, and the University of Lausanne.
Read the article: https://doi.org/10.1007/s43621-026-03271-0