Enhancing forecasting and earth observation data and applications for drylands

by | Oct 10, 2025 | Community of Practice, Dialogue, Drylands, Eastern Africa, Forecasts, Horn of Africa, ICPALD, SPARC, University of Milan | , ,

Our recent joint event on ‘Building the Resilience and Prosperity of Pastoralists and Dryland Communities’ featured a session on earth observation (EO) developments and what they can mean for early warning and early action in drylands. Here we provide a short report from the session and links for further information.

Recent combinations of advanced earth observation sensors and investments in remote surveillance and forecast models have led to fast changes in data availability and forecasting products. This session highlighted some of those advances and their implications for early warning and early action in the drylands.

Moderator Maurine Ambani (World Food Programme) introduced the session, framing the roles of earth observation data and forecasting in the humanitarian space, especially with regards to anticipatory action.

She asked the audience to consider three questions: How does this research on EO support anticipatory actions in the drylands; what essential information do these types of EO-based solutions miss; and are there some key synergies emerging that link research on EO with other early action agendas?

The session comprised four short research presentations followed by audience interaction and closing summaries with a panel comprising Eugene Kayijamahe (IGAD Climate Prediction and Applications Centre), Vivianne Meta (LocateIT Ltd) and Maurine Ambani.

Key insights presented:

  • EO drought index design shows promise but lacks clear trigger mechanisms to link triggers to actions and react fast to shocks.
  • We need continuous co-learning to understand compound risks and cascading impacts associated with drought, moving from crisis response to opportunity creation.
  • Ground truthing is critical to address mismatches between satellite data and local conditions, particularly for identifying vegetation types valuable for livestock.
  • EO alone is insufficient – it must integrate with deep place-based knowledge of natural resources and environmental changes.
  • Timely surface water information and water mapping capabilities are advancing for drought preparedness and early action.
  • Longitudinal data covering entire nations enables monitoring of conflict impacts on agricultural activity and rural markets.

Main points of discussion or debate:

  • Consensus emerged that EO has limitations – it cannot capture mobile markets nor full market functionalities, and it may miss important local variations.
  • Debate on extent of EO capabilities and where to stop, with agreement on need to bring local voices into EO applications.
  • Discussion on economic analysis of different anticipatory actions – there is limited cost-benefit analysis available for decision-making.
  • Tension between availability/access (no longer major problems) and acceptance of products, requiring better communication of uncertainty.
  • Questions about animal movement as optimal strategy – may not always be best option depending on context.

Implications for policy, practice and research:

  • Policy needs include developing trigger mechanisms that effectively link EO data to rapid response actions and ensuring quality verification through institutions like IGAD.
  • Practice requires capacity building to bring EO science closer to ground practitioners, with simplified applications using AI and machine learning.
  • Research priorities include addressing cost barriers for in-situ sensors, developing economic analysis frameworks for anticipatory actions, and improving uncertainty communication.
  • Private sector involvement is essential for sustainability and commercialization, building solutions for different end users.

Some take-away messages:

  • EO presents significant opportunities, but challenges remain – we need greater awareness of limitations while maximizing potential.
  • Integration of satellite data with local knowledge and ground truthing is essential for effective applications.
  • Private sector roles are often forgotten but are crucial for sustainability – need commercialization strategies that serve diverse end users.
  • Capacity building and simplification through AI and machine learning are necessary to bridge science-practice gaps and ensure solutions match real needs.

More information – Speakers and presentations (with download links):

Other session reports