Two country case reports from our ‘impact collaboration‘ on early warning systems (EWS) and dryland communities in the Horn of Africa have been published. The reports complement a global desk review and bring country perspectives to help answer the challenge question ‘how can we bridge the disconnects between food security, climate, and natural hazards early warning systems and the anticipatory actions that pastoral and agro-pastoral communities in the Horn of Africa can take to overcome recurring shocks and threats to their lives and livelihoods?’
Conclusions from the Ethiopia study include:
Pastoral and agro-pastoral communities possess extensive indigenous knowledge for weather forecasting and strong traditions of information sharing information. However, climatic hazards are becoming increasingly difficult to predict accurately using indigenous knowledge, as climate change disrupts historical weather patterns, and younger generations rely more heavily on modern technologies.
Communities reported significant early warning system limitations, noting that information often lacks relevance or fails to reach them in time. This compounds the scepticism in meteorological forecasts that is more prevalent among older community members. These limitations and previous negative experiences undermine trust in these systems and affects perceptions of the information conveyed.
Early warning dissemination is further constrained by structural barriers including inadequate mobile phone infrastructure and electrification, as well as organization-level obstacles. Similarly, findings revealed that women were more likely to report that they personally had not received early information from sources outside their communities, but rather they often learned of them through other community members or in communal gathering places.
Modern technologies can enhance the efficiency and reach of all types of information dissemination, aligning with communities’ preferences to receive early warning information through multiple channels including mobile phones, radio, and traditional in-person methods. Indeed, communities are already embracing modern technologies to spread early warning information, using mobile phones to share news and to facilitate practices such as Sahan.
Overall, integrating modern technologies with indigenous knowledge and involving communities in co-producing early warning systems provides opportunities to improve the reach, efficiency, and accuracy of information dissemination, strengthening trust and understanding of these systems.
Conclusions from the Kenya study include:
EWS are widely recognised by Kenyan pastoral and agro-pastoral communities as essential for preparing for drought, livestock disease, and other recurring shocks.
Across the two counties, communities highlighted that indigenous knowledge, rooted in environmental observation, animal behaviour, seasonal indicators, and the expertise of elders and traditional specialists, remains a trusted and accessible way to predict hazards. This knowledge is deeply embedded in everyday decision-making, especially for mobility, herd management, and assessing local conditions.
However, the reliability of indigenous forecasting has been increasingly disrupted by climate change. Seasonal patterns are less predictable, traditional indicators are no longer as reliable, and communities report that droughts, floods, and disease outbreaks now occur outside the cycles they previously relied upon. These changes have led to growing interest in complementing Indigenous knowledge with modern forecasting methods.
Despite this shift, perceived value of formal EWS remains uneven. Communities frequently cited late, unclear, or non-localized warnings as key reasons for scepticism, particularly among those who compare scientific forecasts unfavourably to their past experiences. Many participants emphasized that warnings often arrive after hazards have already begun to escalate or do not specify the precise areas at risk, making it difficult to take anticipatory action.
Dissemination barriers further restrict the effectiveness of EWS. Limited network coverage, low levels of mobile phone ownership among certain groups and challenges with charging and maintaining devices, all prevent timely access to information. County officials also face structural constraints, including resource limitations, inconsistent communication systems, and competing responsibilities during crisis periods. These challenges mean that many people still hear about hazards informally through neighbours, markets, chief barazas, or elders rather than directly from official channels.
At the same time, communities are already integrating modern tools into their own information-sharing practices. Mobile phones and local radio remain widely used, especially for cross-community updates, disease alerts, and conflict-related information. There is strong demand for blended systems that draw on both scientific and Indigenous knowledge, presented in clear, localized formats and communicated through multiple trusted channels.
Overall, strengthening Kenya’s early warning landscape requires improving the clarity, timing, and geographic precision of scientific forecasts, investing in reliable communication infrastructure, and involving communities more systematically in the design, interpretation, and feedback processes of EWS. Integrating indigenous knowledge with modern forecasting, and ensuring marginalised groups have equitable access to information, offers significant potential to build trust, enhance relevance, and support more effective anticipatory action.
More
Download the Kenya report: Muhandale, A., Kirwa, L., Walker, A., Allard, J., Brown, F., Adam, A., Mohammed, N., Puerta Fernandez, M. de la, Jensen, N.D., Lazarus, B., and Kristensen, K. 2026. Early warning systems and dryland communities in the Horn of Africa: Kenya country report. Edinburgh: Urban Foresight for the Jameel Observatory for Food Security Early Action.
Download the Ethiopia report: Adam, A., Mohammed, N., Walker, A., Allard, J., Brown, F., Kirwa, L., Muhandale, A., Puerta Fernandez, M. de la, Jensen, N.D., Lazarus, B. and Kristensen, K, 2026. Early warning systems and dryland communities in the Horn of Africa: Ethiopia country report. Edinburgh: Urban Foresight for the Jameel Observatory for Food Security Early Action.
Download the desk study: Puerta Fernandez, M.L. de la., Adam, A., Kirwa, L., Kiprop, C., Jensen, N., Lazarus, B., Kristensen, K., and Walker, A. 2025. Early warning systems and dryland communities in the Horn of Africa: A desk review of actors, challenges, and opportunities. Edinburgh. Urban Foresight for the Jameel Observatory for Food Security Early Action.