Use of AI in meteorological modelling has recently demonstrated the ability to produce state-of-the-art predictions with relatively small computational power. Thus, there is growing interest in leveraging AI technology to help countries without sophisticated super-computers to “leapfrog” to the latest most advanced prediction systems.

WMO’s Executive Council recently set up a new Joint Advisory Council on Artificial Intelligence to inform WMO activities and to balance opportunities and challenges. 

WMO is therefore excited about a new pilot project in Malawi, with funding from the Climate Risk and Early Warning Systems (CREWS) initiative. It will test the ground in leveraging a state-of-the-art AI-based Weather Prediction (AI-WP) system to improve the accuracy, timeliness, and accessibility of weather predictions in Malawi.

It aims to empower Malawi’s Department of Climate Change and Meteorological Services (DCCMS) to build operational capacity in AI-WP and early warning provision, and to evaluate how an AI-WP system can help in closing critical capacity gaps in Malawi and more generally in Least Developed Countries and Small Island Developing States.

The project builds on a high-resolution data-driven weather forecasting model named Bris developed by MET-Norway and the “forecast-in-a-box” concept developed by the European Centre for Medium-Range Weather Forecasts (ECMWF).

Through this pilot, meteorologists in Malawi will gain hands-on experience running AI-enabled forecasts locally, assessing their operational feasibility, forecast skill, and potential to support timely early warnings for high-impact weather events. 

Malawi – like many African Least Developed countries – is highly vulnerable to climate-related hazards. Its early warning systems face significant gaps due to limited observational infrastructure, constrained human resources, and outdated forecasting systems. 

“We firmly believe this initiative represents a strategic opportunity to strengthen Malawi’s early warning infrastructure, deliver actionable insights, and support long-term capacity development for our forecasting staff,” said Lucy Mtilatila, director for Climate Change and Meteorological Services and permanent representative of Malawi to WMO.

The project – which combines Norway’s AI expertise with Malawi’s local knowledge and data – was presented by Roar Skålin, permanent representative of Norway to WMO at the High-Level Open Consultative Platform on AI during the week of the WMO Executive Council. 

Forecast-in-a-Box

It is being launched in collaboration with the European Centre for Medium Range Weather Forecasing (ECMWF) which is developing early prototypes of Artificial Intelligence/Integrated Forecasting System (AIFS) packaged as a Forecast-in-a-Box as part of the AI-driven solutions of the Digital Twin Engine and the Destination Earth initiative of the European Commission. 

Data-driven models such as ECMWF’s AIFS and MET Norway’s Bris are fundamentally different from traditional numerical weather prediction systems. They are lighter, faster, and more portable, making them well suited to run outside large high-performance computing (HPC) infrastructures. 

This set-up enables to run forecasts closer to where the data is needed, offering key benefits: 

• Users can tailor the forecasting pipeline to their specific needs.
• Improved responsiveness and timeliness.
• Deployment across a range of environments.
• No deep expertise in system setup or infrastructure is needed.

WMO Integrated Processing and Prediction System 

Despite the huge possibilities, there question marks about the capability of AI to support forecasts and warnings of local high-impact weather and water hazards.  The WMO’s Executive Council therefore requested the development of technical guidelines on the use of AI-based technologies and how incorporate AI into the WMO Integrated Processing and Prediction System (WIPPS).  This is the worldwide network of operational centres of WMO’s Members and is the backbone of all forecasting.  

The planned Joint Advisory Group will be a coordination mechanism among WMO’s Infrastructure and Services Commissions, Research Board and other relevant WMO bodies. It will include experts from the public, private and academic sectors and will steer joint efforts to explore the opportunities and challenges of AI/ML technology.