Building capacity for seasonal heat forecasting in West Africa: A case study with Burkina Faso’s Meteorological Agency (ANAM-BF) using the Objective Seasonal Outlook Package (OSOP)

Extreme heat is an increasing concern in West Africa, particularly in the Sahel which is already exposed to high temperatures.  There is growing interest from regional climate centres and national meteorological agencies in expanding capabilities to inform heat early warning systems, including enhanced seasonal forecasting.

During the Clima-Social project, part of the Weather and Climate Information Services (WISER) programme funded by the UK’s Foreign, Commonwealth and Development Office, Social Protection and Public Health stakeholders in Burkina Faso expressed interest in seasonal forecasts of extreme heat, which, unlike rainfall, are not yet produced by the National Meteorological Agency (ANAM-BF). To meet this need, the UK Met Office and the Red Cross Red Crescent Climate Centre are supporting ANAM-BF to strengthen its technical capability to deliver robust seasonal heat forecasts and advance its ambition to develop regional expertise in heat forecasting and early warning.

The Objective Seasonal Outlook Package (OSOP) is an open-source toolkit developed by the Met Office to support objective seasonal forecasting. The toolkit contains a set of Python and shell scripts that can be edited and tailored to the user's needs. OSOP generates hindcasts to evaluate model skill and probabilistic forecasts using data from Global Producing Centres, ensuring transparency and reproducibility. Using OSOP, we customised parameters to generate locally relevant, tailored seasonal forecasts from global datasets. Outputs were designed to meet ANAM-BF’s operational requirements, including specific thresholds relevant for Social Protection intervention and a sub-domain focused on Burkina Faso and surrounding regions.  

The process involved technical co-development, stakeholder engagement, and prototype creation, addressing challenges in communicating forecasts. Discussions highlighted the limitations of tercile-based approaches in the context of a warming climate and the need for alternative (e.g., deciles) categories to better reflect extreme heat seasons. OSOP was tested with ANAM-BF forecasters, providing insights into practical implementation and capacity-building needs, while identifying opportunities for future development. For instance, local weather station data are currently used to bias correct simulations as well as assessing hindcast skill. 

Our work demonstrates the value of co-producing new climate service capabilities that adopt an open-source access approach. This ensures stakeholders can continue to access and adapt global forecast data without relying on privately-operated systems, encouraging independent application and autonomy of services. The project offers a scalable model for seasonal heat forecasting across West Africa and beyond, with the potential for uptake by other meteorological services to supporting early action and resilience in the face of climate change.