Institutionalising co-production of weather and climate services: learning from the African SWIFT and ForPAc projects

There is growing recognition of the multiple benefits of co-production for forecast producers, researchers and
users in terms of increasing understanding of the skill, decision-relevance, uptake and use of forecasts. This
policy brief identifies lessons learnt from two operational research projects, African SWIFT and ForPAc, on
pathways for embedding co-production into operational weather and climate services as the new standard
operational procedure.
Experiences across these projects identifies the following potential pathways for institutionalising co-production
practises within operational weather and climate services:
• Changing mindsets and systems to enable co-production of enhanced forecasts and systematic approaches
for their use.
• Strengthening in-country institutional links between operational forecasting centres and academic
institutions to develop sustainable and improved forecasting capacities to meet users’ evolving weather and
climate information needs.
• Ensuring continued access to raw forecast data from global forecasting centres to continue and further
develop new and improved decision-relevant forecasts.
• Formalising user engagement in co-production, through agreeing standard and continuity of representation
and commitment to providing regular feedback.
• Mainstreaming stakeholder engagement and co-production in meteorological training, forecasting
operations and environmental research.
• Working through existing channels, such as agricultural and livestock extension services, and harnessing
social media and remote ways of working to develop sustainable forms of continuous user engagement.
• Establishing monitoring systems to demonstrate the benefits of investing in forecasting capacities.
• Incentivising collaboration between complementary initiatives.
• Addressing the risks of operationalising new and improved weather and climate services in resource-
constrained environments.