Jane Edwards 
MITRE and The Weather Company, or TWCo, have partnered to advance global weather intelligence by integrating high-resolution weather data into artificial intelligence-driven forecasting tools.
The latest collaboration highlights the growing role of advanced data and AI in strengthening decision-making across the federal landscape. Through executive-level discussions and agency-driven perspectives, the 2026 FedCiv Summit this October will explore how federal civilian agencies are translating priorities into action, with topics including scaling AI across government, data, cloud and compute infrastructure, cybersecurity and compliance, and cross-agency initiatives. Register now!
What Does the Collaboration Intend to Achieve?
In a joint release published Thursday, the organizations said the effort will focus on improving forecast accuracy and enabling more informed decision-making for government and industry users.
In the initial phase, TWCo will integrate MITRE’s Weather 1K dataset into its Global High-Resolution Atmospheric Forecasting system, known as GRAF, to produce more detailed forecasts designed to address the gap between experimental weather technologies and real-world applications.
Peter Neilley, senior vice president of science and forecasting operations at TWCo, said the integration of high-resolution data into forecasting models is intended to support the development of a 1-kilometer AI-based forecasting system.
“This would enable us to provide hyper-local, decision-specific weather forecasting intelligence in high-impact weather circumstances supporting personal, business, aviation, defense, and many other use cases,” added Neilley.
Commenting on the partnership, MITRE President and CEO Mark Peters said, “We appreciate the opportunity to collaborate with The Weather Company and other industry partners, and we look forward to further engagements as we advance meaningful applications of weather intelligence.”
What Is Weather 1K?
Weather 1K is a high-resolution, continental-scale AI training dataset developed by MITRE using its Federal AI Sandbox, a computing capability designed to help federal agencies train and deploy AI foundation models to support critical government missions.
The dataset totals approximately 7 petabytes and is designed to support next-generation AI-driven weather models. MITRE developed Weather 1K in collaboration with NVIDIA’s Earth 2 team and used advanced modeling tools, including TempoQuest’s AceCast.
“Weather 1K is a tremendous technical achievement with national impact, possible only through deep collaboration with commercial industry and government,” said Peters, a 2026 Wash100 awardee.
In a recent Executive Spotlight interview, Peters discussed the role of federally funded research and development centers in helping government agencies accelerate the transition of emerging technologies into operational programs, emphasizing the importance of speed, scalability and alignment with mission needs.
