Information

We are looking forward to welcoming you to Fort Collins from August 18 to 21. While in-person registration is currently full, we encourage you to reach out to the organizing committee if you're still interested in attending as spots may open up due to last-minute cancellations. Registration for online attendance is still open, and we'd be delighted to have you join us virtually.

Agenda

The workshop agenda is available here.

Additional Information

This document provides useful information about travel arrangements and workshop logistics.

A true color image of Hurrican Milton north of the Yucatan peninsula overlaid with contours of 89-GHz passive microwave observations.

About

AI foundation models (FMs), i.e., large machine learning models trained on extensive, generic corpora of atmospheric training data, are emerging as powerful tools for tackling complex challenges in atmospheric science. The AI Foundation Models for the Atmosphere workshop will bring together atmospheric scientists and model developers to assess the current state of the art of AI FMs and related validation frameworks, explore their applications, and identify remaining data, technology, and knowledge gaps hindering their adoption.

Program Highlights

  • Presentation and discussion sessions on the development of AI FMs and their application throughout the Atmospheric Sciences
  • Hands-on sessions demonstrating the use of FMs on diverse tasks ranging from satellite retrievals over weather forecasting to climate model downscaling.
  • Roadmap Development: Collaboratively outlining key opportunities, challenges, and strategic steps needed to unlock the full potential of FMs in atmospheric science.
The findings and discussions from this workshop will form the basis for a report detailing principal opportunities and challenges of FMs for atmospheric-science applications, and proposing a roadmap for advancing the use of AI FMs in atmospheric science.

Target Audience

We invite graduate students, postdocs, faculty, and research scientists interested in developing AI FM applications for atmospheric science. AI model developers eager to share their models and collaborate with the atmospheric science community are also highly encouraged to participate.

A masked two-meter-temperature field as used in the pre-training of the PrithviWxC foundation model.

Contact

For questions please contact fm4a2025@gmail.com.

Organizing Committee

  • Simon Pfreundschuh, Research Scientist, Colorado State University
  • Haonan Chen, Assistant Professor, Colorado State University
  • Mu-Ting Chien, Postdoctoral Fellow, Colorado State University
  • Itinderjot Singh, Postdoctoral Fellow, Colorado State University
  • Senne Van Loon, Postdoctoral Fellow, Colorado State University
  • Spencer Jones, Doctoral Student, Colorado State University
  • Tsengdar J. Lee, Program Manager, NASA
  • Michael M. Little, Principal Data Scientist, NASA
  • Ramachandran Rahul, Senior Research Scientist, NASA
  • Sujit Roy, Principal Research Scientist, University of Alabama in Huntsville
  • Johannes Schmude, Senior Research Scientist, IBM