AI and Forecasting: WMO Charts a New Course with Global Collaboration

The World Meteorological Organization (WMO) is taking decisive steps to integrate Artificial Intelligence (AI) into global forecasting systems, signaling a new era of innovation in weather, climate, and hydrological services. At the heart of this transformation is a growing collaboration with private tech firms and academic institutions aimed at enhancing the capabilities of National Meteorological and Hydrological Services (NMHSs)—especially in developing countries.

On 16 June, WMO hosted its sixth annual Open Consultative Platform on Artificial Intelligence, a strategic dialogue that laid the groundwork for the WMO Executive Council’s deliberations. The discussions underscored the urgency and potential of AI to redefine Earth system prediction and service delivery worldwide.

AI as a Game Changer in Forecasting

“The rapid and transformative development of AI is truly astonishing,” said Michel Jean, President of the WMO Commission for Observation, Infrastructure and Information Systems (INFCOM). “The convergence of next-generation AI models and global reanalysis datasets is poised to significantly enhance the accuracy of weather predictions.”

However, Jean emphasized that AI is not a replacement for traditional forecasting. The value of observational networks, physics-based modeling, and decades of meteorological infrastructure—like the WMO’s World Weather Watch—remains undiminished. Instead, AI is a tool to complement and expand the scope of prediction services.

Bridging Public, Private, and Academic Sectors

A major theme of the platform was fostering closer collaboration between public agencies, private tech companies, and academic research institutions. Representatives from Google Research, Microsoft Research, AccuWeather (on behalf of the HydroMeteorological and Environmental Industry Association), and the Shanghai Academy of AI for Science presented powerful use cases demonstrating AI’s growing role in weather prediction.

These organizations pledged not to compete with NMHSs but to support them as the official and authoritative voices in public weather warnings. Several ongoing collaborations illustrate this model:

• Google is using AI for global flood forecasting and partners with the U.S. National Hurricane Center, as well as NMHSs in the Czech Republic, Nigeria, Uruguay, and Viet Nam.

• Microsoft works with the UK Met Office on supercomputing, contributing to the UN’s “Early Warnings for All” agenda through its AI for Good Lab.

• The Shanghai Academy, alongside the China Meteorological Administration, offers a cloud-based solution to help countries with limited infrastructure access high-quality forecasting tools.

Supporting Small NMHSs: Democratizing Innovation

One of the most compelling promises of AI in weather services is its potential to empower smaller and under-resourced NMHSs. AI-based Earth System Prediction (AI-ESP) models require far fewer computational resources than traditional Numerical Weather Prediction (NWP) models once trained, making them ideal for countries without supercomputing capabilities.

Key benefits for developing countries include:

• Reduced computational burden, enabling forecast generation using modest infrastructure.

• Improved data coverage by using satellite, reanalysis, and crowdsourced data to fill observational gaps.

• Cost efficiency, as partnerships with private entities offer access to data and services at reduced costs.

• Scalable knowledge transfer, ensuring that new systems can be locally adapted and maintained.

Roar Skalin, Norway’s permanent representative to WMO, presented a pilot project involving the European Centre for Medium-Range Weather Forecasts (ECMWF), the Norwegian Meteorological Institute, and Malawi’s NMHS. The project’s architecture trains models in the global North, while operational implementation occurs in Malawi using simpler computing systems—showcasing an effective global-local collaboration.

Key Principles and Ethical Frameworks

The platform reinforced that the integration of AI into weather services must adhere to principles of openness, transparency, and traceability. Public investment in data systems has laid the groundwork for AI, and public-private initiatives must continue to share open and reliable datasets.

But as AI reshapes how—and in some cases by whom—forecasts are produced, important questions about roles, responsibilities, and accountability arise:

• Single Authoritative Voice: NMHSs remain the trusted providers of public warnings. Any AI-powered system must support and not conflict with this mandate.

• Ethics and Trust: Balancing innovation with responsible AI use is critical, especially when automated systems could influence decisions related to life and safety.

• Joint Coordination: A proposal was floated for a joint body to oversee the development of shared principles, ethical standards, and AI governance models across sectors.

Arlene Laing, WMO Representative for the British Caribbean Territories, presented early evidence that AI is already helping improve tropical cyclone forecasting. However, she emphasized the need to build capacity so that NMHSs can adapt to these tools effectively, ensuring sustainable and responsible implementation.

Toward a Shared Forecasting Future

The sixth Open Consultative Platform affirmed a collective vision: harnessing AI to strengthen, not sideline, the public institutions at the heart of global weather services. It also reiterated the importance of collaboration to make forecasting more inclusive, equitable, and accurate.

As Earth system challenges mount—from climate change to extreme weather events—the fusion of AI innovation and meteorological expertise may well define the future of global resilience. The WMO, through this evolving partnership model, is helping to ensure that no country—and no community—is left behind.

Tags: