S1378-119

What This Bill Does

Requires NOAA to build curated weather forecasting training datasets, explore artificial-intelligence weather models, test public communication and wildfire-risk applications, keep advancing traditional numerical forecasting, support forecasters and emergency managers, create model-assessment frameworks, explore public-private research partnerships, release selected AI weather models and data when safe, and report on foreign-country risks to U.S. weather data.

Who Benefits and How

National Weather Service forecasters benefit from technical assistance, data access, best practices, testbed support, and model-assessment frameworks for AI and numerical weather models. Emergency managers benefit from support for operational decisions based on AI weather models, numerical weather models, or combined outputs. Firefighters and wildfire-prone communities benefit if AI improves readiness, risk mitigation, safety, and impact-based decision support. Weather researchers, private weather companies, academic institutions, and international partners benefit from novel R&D partnerships, co-investment strategies, public model releases, metadata, documentation, and access to releasable federal data.

Who Bears the Burden and How

NOAA weather model developers must curate comprehensive Earth-system training datasets within four years, assess existing reanalysis datasets, experiment with global, regional, and local AI models, quantify uncertainty, evaluate data-poor coverage, report every two years through 2035, and protect national security, intellectual property, trade secrets, contract restrictions, and the life-and-property mission. Department of Energy, NASA, NSF, NCAR, NIST, and the Interagency Council on Advancing Meteorological Services must consult or collaborate. Foreign countries of concern face scrutiny because NOAA must report in classified and unclassified form on economic and intellectual-security risks from access to U.S. weather data.

Key Provisions

• Defines AI, AI weather models, curated datasets, numerical weather models, observational data, synthetic data, and weather data.
• Requires NOAA to develop and curate comprehensive weather forecasting training datasets within four years.
• Authorizes development and testing of global, regional, and local AI weather models using NOAA data where possible.
• Requires continued support for observational data, numerical Earth-system modeling, post-processing, and data assimilation.
• Requires biennial reports through 2035 and support for forecasters, scientists, engineers, and emergency managers.
• Directs public-release planning for operational and experimental AI weather models, documentation, metadata, and releasable data.
• Requires a classified and unclassified report within one year on foreign-country risks to U.S. weather data.