Earth Prediction Innovation Center – Artificial Intelligence in Weather Modeling Workshop
Event Date: June 27, 2025
Event Location: Virtual
The Earth Prediction Innovation Center (EPIC) hosted a highly successful virtual workshop on Artificial Intelligence (AI) in Weather Modeling on June 27, 2025. The event brought together experts and practitioners from government, academia, and industry to explore a range of AI use cases in Numerical Weather Prediction (NWP). Sessions covered model development, training workflows, and hands-on demonstrations using tools such as GraphCast, FourCastNetv2, PanguWeather, and Anemoi Core (https://github.com/ecmwf/anemoi-core).
Presenters represented a wide mix of organizations including federal agencies, national research centers, academic programs, and leading companies in the technology and defense sectors. The workshop reached full capacity with 577 registered participants and saw peak attendance of 210, with strong engagement throughout the eight hour session. The turnout and level of participation highlighted growing community interest in the application of AI in weather modeling.
A special thanks to all of our presenters, as this workshop would not have been as impactful without you: John Ten Hoeve (OAR/WPO), Kevin Garrett (NWS/OSTI), Daniel Abdi (OAR/GSL), Isidora Jankov (OAR/GSL), Vanderlei Vargas (OAR/GSL), Tim Smith (PSL), Andrew Justin (OU), Anil Kumar (OAR/EPIC), Paul Madden (OAR/GSL), Brandon McClung (OU/AFIT CI), Jun Wang (NWS/EMC), Jacob Carley (NWS/EMC), Microsoft and team led by Salar Adili, Corey Potvin (OAR/NSSL), Tomorrow.io and team led by Arun Chawla, and Fire Neural Network (FNN) led by Istvan Kereszy and Kyle Miller.
Agenda:
June 27, 2025 | Team | Content | Content link | Presenter and affiliations* | Session Recording Link |
| 10:00 AM – 10:10 AM | Introduction | Introduction from NOAA Leadership at AI4NWP | John Ten Hoeve, OAR/WPO Kevin Garrett, NWS/OSTI | ||
| 10:10 AM – 10:55 AM | AI model development and verification | HRRRCast: a data-driven emulator for regional weather forecasting | Daniel Abdi, Isidora Jankov | YouTube: https://youtu.be/Epq0aOrapLU | |
| 10:55 AM – 11:25 AM | Anemoi initial overview supporting NOAA | Anemoi Overview | Tim Smith, PSL | YouTube: https://youtu.be/q78Aa6VYbh0 | |
| 11:25 AM – 11:40 AM | Anemoi demo | Anemoi demonstration using Google Colab | Justin Andrew OU/EPIC via Raytheon | **Anemoi demo included in overview recording above** | |
| 11:40 AM – 11:45 AM | BREAK | BREAK | BREAK | BREAK | |
| 11:45 AM – 12:35 PM | AI modeling Verification | AI modeling Verification overview and walk through both AI modeling frameworks being looked at NOAA. | Anil Kumar, OAR/EPIC via Tomorrow.io Paul Madden, OAR/GSL via CIRES D. Alex Burrows (NOAA-EPIC & Redline Inc), Mariah Pope (NOAA-EPIC & Tomorrow.io) | YouTube: https://youtu.be/VJSv2EDnZ50 | |
| 12:35 PM – 1:05 PM | Front finder and Boltcast | AI application use cases for front finder and bolt cast. | Brandon McClung OU/AFIT CI, Justin Andrew OU/EPIC, Amy McGovern OU, John T. Allen CMU | YouTube: https://youtu.be/gAkFkXZZK8Q | |
| 1:05 PM – 1:10 PM | BREAK | BREAK | BREAK | BREAK | |
| 1:10 PM – 2:00 PM | Running Graphcast/FourCastNetv2/PanguWeather | This notebook will go over running AI models in a repeatable process using Google Collab notebooks. | Keven Blackman, OAR/EPIC via Raytheon | YouTube: https://youtu.be/3JeO92DMAPM | |
| 2:00 PM – 2:10 PM | BREAK | BREAK | BREAK | BREAK | |
| 2:10 PM – 3:00 PM | NOAA Environmental Modeling Center (EMC) advances in AI numerical modeling | Presenting on advances in AI numerical modeling on project EAGLE from EMC. | Jun Wang, NWS/EMC and Jacob Carley, NWS/EMC | YouTube: https://youtu.be/Rd78DhnennU | |
| 3:00 PM – 3:30 PM | Microsoft Azure | Microsoft presenting on advances from industry in the field of AI numerical modeling | Joe Madaus, Microsoft and Salar Adili, Microsoft | YouTube: https://youtu.be/rZ1Aob1-Xf4 | |
| 3:30 PM – 3:40 PM | BREAK | BREAK | BREAK | BREAK | |
| 3:40 PM – 4:15 PM | WoFSCast | Presenting on advances in WoFSCast: A GraphCast-based emulator for the Warn-on-Forecast System. | Corey Potvin, NSSL and Monte Flora, CIWRO | YouTube: https://youtu.be/j6Td9C-CNdM | |
| 4:15 PM – 5:05 PM | Tomorrow.io advances in industry in AI modeling | Tomorrow.io presenting advances from industry in the field of AI numerical modeling | Forest Cannon, Randy Chase, Brandon Taylor, Max Green, Kushal Keshavamurthy, Mariah Pope, Luke Conibear, Joe Munchak, Ethan Nelson, Arun Chawla with Tomorrow.io | YouTube: https://youtu.be/JerrXnkzSQc | |
| 5:05 PM – 5:15 PM | BREAK | BREAK | BREAK | BREAK | |
| 5:15 PM – 5:45 PM | Fire Neural Network (FNN) capabilities | AI for fire detection capabilities | Istvan Kereszy, FNN and Kyle Miller, FNN | YouTube: https://youtu.be/Q83m8Qw2CVY | |
| 5:45 PM – 6:00 PM | Closing | Closing remarks from NOAA Leadership at AI4NWP | John Ten Hoeve, WPO and Kevin Garrett, OSTI | YouTube: https://youtu.be/VrZPMG0o82Q |
Oceanic and Atmospheric Research (OAR), Weather Program Office (WPO)
OAR/Physical Sciences Laboratory (PSL)
OAR/Global Systems Laboratory (GSL)
OAR/National Severe Storms Laboratory (NSSL)
National Weather Service (NWS), Environmental Modeling Center (EMC)NWS/Office of Science and Technology Integration (OSTI)
Cooperative Institute for Severe and High-Impact Weather Research and Operations (CIWRO)
Oklahoma University (OU)
Air Force Institute of Technology Civilian Institutions (AFIT CI)
Microsoft Corporation
Tomorrow.io
Fire Neural Network (FNN)
European Centre for Medium-Range Weather Forecasts (ECMWF)
