How AI technologies transform national meteorological services delivery

Innovative AI technologies can drive a profound transformation in meteorological services as this new Amazon Web Services (AWS) short report describes.

AI, generative AI, and agentic AI offer opportunities to transform meteorological services and their business models. The new methods complement traditional physics-based Numerical Weather Prediction (NWP) models. They collect and analyze vast datasets rapidly. They learn from past weather patterns to deliver more accurate, detailed, and localized short-term and long-range forecasts. Agentic AI technologies can identify future weather events and patterns beyond human inference and develop appropriate responses.

Klemen Bergant, Executive Director of EUMETNET, a network of 33 European national meteorological and hydrological services, predicts that AI adoption will be widespread across the meteorological value chain: “AI will enter every segment of the meteorological value chain, from replacing some of the current forecasting instruments to developing new and more effective weather forecasting products and services.”

The paper explores the potential impact of AI on meteorological services and suggests:

1. Users of AI-enabled meteorological services seek weather impact, not only forecasts.
2. Public-private sector collaboration is important—even necessary—to realize AI’s potential.
3. Governance of meteorological services must change to keep pace with AI-driven innovations.

The future of weather forecasting lies in hybrid approaches combining the speed and pattern recognition capabilities of AI with the scientific rigor of traditional, physics-based NWP models, as technology investment company Battery Ventures research indicated. That study identified high-value opportunities for meteorological services and products, principally in the US and Europe. These include natural disaster mitigation and preparation, defense mission planning, airline route planning and ground operations, route optimization for logistics companies, energy demand forecasting for utilities, and conditions-based marketing for the travel and leisure industry.

Public-private partnership increases innovation and impact

Charles Ewen, Director of Technology at the UK Met Office, highlights the fundamental shift in user expectations. These have moved beyond “what will the weather be” to “what impact will the weather have and what actions should I consider.” To meet these expectations, he says, “We [meteorological agencies] need to learn how to work better with private-sector and tech companies to deliver composite value chains.” This is partly because public services face greater resource and recruitment challenges than private enterprises. Three public-private organization initiatives have found success recently, as described in the following paragraphs.

The World Meteorological Organization’s WIS 2.0 initiative is critical to achieving a UN commitment made in 2022 to protect every person on Earth with early warning systems for extreme weather within 5 years. Launched in January 2025, WIS 2.0 is a cloud-based, serverless, global cache providing real-time meteorological data to national forecasters. Built by AWS in collaboration with the UK Met Office, US National Weather Service, and Synoptic Data, the framework democratizes access to AI-based forecasting models and National Meteorological Service data without requiring substantial investment in computing infrastructure or technical expertise. Forecasters in resource-constrained nations can use WIS 2.0 to access the same quality data as their counterparts in developed economies.

The UK Met Office is pioneering generative AI to modernize weather communication, starting with the UK Shipping Forecast. This has strict format requirements as well as using multiple atmospheric and ocean datasets. The Met Office developed two complementary approaches in collaboration with AWS. The first uses large language models (LLMs) Amazon Nova Pro and Claude Sonnet by Anthropic in Amazon Bedrock, which employs a novel vision language model approach, converting numerical weather predictions into video format for visual processing through Amazon Nova customization on SageMaker AI. The prototype work has established a route to increasingly accurate results and offers the Met Office a roadmap to modernize more than 300 text-based products and services. AI handles initial drafting, freeing meteorologists to focus their expertise where human judgment adds greatest value—a blueprint for scaling specialized knowledge across government services. Watch Architecting AI solutions for mission-critical systems with UK Met Office, an AWS re:invent 2025 session, for a technical explanation of this initiative.

Sencrop, a French agritech company that uses AWS to collect and analyze data from local stations across Europe, has built a machine learning (ML)-based microclimate app to provide farmers with hyperlocal climate data to optimize crop yields while reducing environmental impact. AI-powered forecasting models use millions of in-field data points processed daily to enable data-driven decisions on irrigation, pesticide application, and harvest timing. Even small-scale farmers now have access to high-quality data. Sencrop also creates regional climate intelligence by aggregating anonymized data to help agricultural cooperatives and food companies improve supply chain resilience and sustainability planning. Farmers report reducing water consumption by up to 30 percent and decreasing pesticide use through precise application timing.

Governance and trust challenges

Trust remains critical for meteorological services. Concerns include potential cyberattacks, data quality disparities between regions, and potential lack of transparency or accountability for AI-generated forecasts. One way to provide assurance is for technology companies to continue to support open source codes and standards.

Conclusion

AI and innovative technologies are fundamentally transforming meteorological services, enabling faster, more accurate, and more personalized forecasts. However, success requires more than technology adoption alone. As industry leaders emphasize, effective transformation depends on public-private collaboration, addressing talent shortages, maintaining trust through transparency, and developing appropriate governance frameworks. To learn more, read this AWS global meteorological services report and find additional resources.