AI's Expanding Role in Weather Forecasting: WMO Strengthens Global Partnerships

The World Meteorological Organization (WMO) is forging new pathways to harness the transformative power of Artificial Intelligence (AI) in weather forecasting, signaling a significant shift in the future of Earth system prediction. In a world grappling with climate change, extreme weather events, and escalating data needs, the WMO is opening the door to greater collaboration with the private sector and academic institutions to leverage AI-driven technologies—while reaffirming the foundational role of National Meteorological and Hydrological Services (NMHSs) as the authoritative sources of weather forecasts and warnings.

Setting the Stage: The Sixth Open Consultative Platform on AI

Held on 16 June, the WMO's sixth Open Consultative Platform on AI served as a prelude to high-level discussions at the WMO Executive Council. It gathered global leaders in meteorology, AI research, and technology, representing sectors with growing stakes in the future of forecasting. The central question: how can the world’s public, private, and academic sectors collaborate to responsibly and effectively integrate AI into Earth system prediction?

Michel Jean, President of the WMO’s Commission for Observation, Infrastructure and Information Systems (INFCOM), highlighted the rapid pace of AI development, noting that, “The convergence of next-generation AI systems with reanalysis data will likely drive significant gains in the accuracy of AI-based predictions.” However, he stressed that AI complements—rather than replaces—the foundational value of traditional observations and physics-based modeling that have long defined modern meteorology.

Strengthening Cross-Sector Collaboration

The WMO is actively working to unite the technological capabilities of private companies with the infrastructure and mission of public services. The workshop featured insights from industry leaders like Google Research, Microsoft Research, and AccuWeather (on behalf of the HydroMeteorological and Environmental Industry Association), as well as the Shanghai Academy of AI for Science.

Each organization emphasized a shared goal: to collaborate with NMHSs, not compete with them. Examples abound:

• Google is partnering with the U.S. National Hurricane Center and meteorological agencies in the Czech Republic, Nigeria, Uruguay, and Viet Nam to apply AI tools for real-time flood forecasting.

• Microsoft, through its “AI for Good” initiative, is working with the UK Met Office to develop supercomputing-powered forecasting models and support the UN’s “Early Warnings for All” initiative.

• The Shanghai Academy, in cooperation with the China Meteorological Administration, is offering a cloud-based platform that may help nations with limited infrastructure adopt high-quality forecasting systems.

These partnerships aim to provide scalable, cost-effective forecasting tools, especially for low-income and developing countries.

Supporting Developing Countries and Small NMHSs

A key focus of the dialogue was ensuring AI’s benefits extend to smaller and less-resourced NMHSs. By democratizing access to prediction tools, AI holds the potential to help these services "leapfrog" into more advanced capabilities without requiring costly infrastructure.

Several critical advantages of AI-powered Earth System Prediction (AI-ESP) models were highlighted:

• Reduced Computational Burden: AI-ESP models, once trained, require fewer resources than conventional Numerical Weather Prediction (NWP) models.

• Bridging Observational Gaps: AI can incorporate diverse data sources—including satellite imagery and reanalysis datasets—helping to fill data voids in under-observed regions.

• Affordability: Through collaborative networks, smaller NMHSs can gain access to data and systems at little or no cost, while avoiding the operational and implementation risks of building such tools from scratch.

One standout example came from Roar Skalin, Norway’s Permanent Representative to the WMO, who presented a joint pilot project with ECMWF and Malawi’s meteorological agency. It combines powerful computational training in the Global North with modest on-the-ground execution in Africa—showcasing a model of global-local synergy.

A Framework for Trust and Authority

As AI reshapes the operational landscape, WMO leaders emphasized the need for governance frameworks to ensure the integrity and public trust of forecasting systems. While AI can enhance decision-making and delivery of information, its deployment must support—not replace—the trusted role of NMHSs as the “Single Authoritative Voice” for public safety alerts and weather warnings.

Key principles and takeaways from the event include:

• Openness and Transparency: Public investments in reanalysis data and observation systems must remain open and accessible to drive innovation.

• Ethical Standards: The use of AI in forecasting must adhere to scientific and ethical norms to prevent misuse or misinformation.

• Training and Knowledge Transfer: Local adaptation is essential; global AI solutions must be supported with regionally appropriate training and technical support.

• Public-Private Coordination: Establishing a formal joint coordination body could enable continuous dialogue, ethical oversight, and shared goals across sectors.

Arlene Laing, Permanent Representative of the British Caribbean Territories, highlighted how AI is already helping improve tropical cyclone tracking. However, she cautioned that as AI systems evolve, care must be taken not to undermine the vital public service role of NMHSs.

Looking Ahead: Balancing Innovation and Responsibility

AI/ML presents an unprecedented opportunity to not only refine prediction models but also optimize decision-making workflows across disaster management systems. With that opportunity comes responsibility. WMO’s proactive engagement with private tech giants and academic institutions reflects a broader recognition that shared solutions are needed to protect lives and property in an increasingly unpredictable world.

The 2024 platform reinforced that future-ready forecasting will depend on openness, ethical collaboration, and capacity building. As WMO pushes forward with its "Early Warnings for All" agenda, AI is not just an innovation buzzword—it is becoming a cornerstone of global climate resilience.