NOAA AI for Earth Observation Research

NOAA AI for Earth Observation Research is sponsored by National Oceanic and Atmospheric Administration (NOAA). This grant funds research that applies AI and machine learning to improve the understanding and prediction of environmental phenomena and Earth systems, which can include wildlife-related observations.

## 8th NOAA AI Hybrid Workshop To expand community engagement and increase the accessibility of our activities, this year’s NOAA AI Workshop will host virtual events in March and June. The Workshop event series will conclude with a 3-day hybrid event in Boulder, Colorado with panels and sessions spanning all topics around AI for environmental sciences.

* **August 25-27, 2026 (hybrid; virtual* and in person)** * Registration details coming soon *Virtual events will be hosted via Zoom. We invite NOAA researchers and the broad AI community to submit AI research and applications offsite link that are relevant to NOAA’s mission. Due to limited slots, we cannot guarantee all submitted abstracts will be included.

Priority will be given to work that has not been presented at other meetings.

Time (ET)**Session Topics** 11:00**Workshop Welcome**_(Douglas Rao, Rob Redmon)_ 11:10**Session 1: AI for Environmental Observations**_(Chair: **Stephen Bieda**, National Weather Service)_ * Xiaoming Liu (NOAA/NESDIS STAR): Detection of ocean eddies from satellite ocean color and SST measurements using a deep learning approach * Shardae Douglas (NOAA NCEI): ADDIS: A Hybrid Ensemble Framework for Automated Anomaly Detection in the GHCN-D Dataset * Vijay Vardhan Tadimeti (Oak Ridge National Lab): AI-Ready by Design: How Modular, Data-Driven Development Enables On-the-Fly AI Functionality for Operations * Zachary Medley (Oak Ridge National Lab): Metadata Summary Creation using Retrieval-Augmented Generation with a Large Language Model 13:00**Session 2: AI for Environmental Modeling**_(Chair: **Isidora Jankov**, Global Systems Laboratory)_ * Kinya Toride (CIRES/NOAA PSL): Interpretable AI-Informed Model Analogs for Seasonal Precipitation: When and Why Optimization Adds Skill * Marc Alessi (Union of Concerned Scientists/Colorado State University): Southern Hemisphere Surface Warming Drives Southwestern U.S. Precipitation according to AI-Informed Climate Model Simulations * Jia Wang (NOAA OAR/GLERL): A Machine Learning Framework For Efficient Ice Cover Forecasting In The Great Lakes Leveraging Long-Term Extensive Simulation Data * Ellie Hojeily (University of Albany/NOAA NCAS-M): Evaluating the Aurora Machine Learning Model Utilizing NCEP Aerosol Initial Conditions for PM2.

5 Forecasting 15:00**Session 3: AI for Data Augmentation and Information Management**_(Chair:**David Turner,** Weather Program Office)_ * Manoj Nair (CIRES/NOAA NCEI): Autoresearch applied to the Prompt Penetration Electric Field Model * Jay Su (NOAA NESDIS/ACIO-S): Bridging Data Silos with the NESDIS Storefront: A Unified AI-Powered Access Point via the INDUS Embedding Model * Andrew Fritz (NWS/National Data Buoy Center): Project Lighthouse: cloud-native AI-assisted quality control system * Aaron Anderson (NWS/Norman Weather Forecast Office): Scaffolding NOAA AI Data Services: One Template, Many Domains 16:20–16:30**Closing Remarks** **March 19, 2026 Agenda** Time (ET)**Session Topics** 10:30**Workshop opening and welcome** 10:40**Session 1: AI for Environmental Observations** Chair: Hassan Moustahfid (NOAA/NOS) * Informing Great Lakes Observing System Design with Convolutional Gaussian Neural Processes (Dani Jones, CIGLR) * Earth Embeddings: Towards AI-centric Representations of our Planet (Esther Rolf, CU Boulder) * Rapid Annotation with VIAME Using AI-Assisted Tools for Environmental Imagery (Matt Dawkins, Kitware Inc) * Autonomous Tropical Cyclone Center Prediction via Computer Vision AI Models (Matthew McKinney, NESDIS/STAR) 12:15**Session 2: AI for Earth System Modeling** Chair: Dan Barrie (NOAA/OAR) * 4-D nowcasting of clouds and visibility with a hierarchy of model complexities (Ryan Lagerquist, ACME) * Accelerating Earth System Modeling by Leveraging a Standardized Coupling Framework for Hybrid Models (Ufuk Turuncoglu, NSF NCAR) * NeurOCAST: A Scalable Neural Operator Framework for Spatiotemporal Prediction (Atieh Alipour, NOS) * Improving HRRR 1-h APCP with a U-Net and rain-aware sampling (Oscar Chimborazo, Howard University/NCAS-M) 14:00**Session 3: AI for Data Management and Information Services** Chair: Rob Redmon (NOAA/NESDIS) * Agentic AI for Near-Real-Time Ocean Hazard Assessment (Kevin Lee, UCLA/NASA JPL) * SPEAR Earth System Data Assistant (Jake Zappala, OAR/GFDL) * Human perspectives on building and using an Intelligent Data Exploring Assistant (IDEA) for environmental data analysis (Matthew Widlansky, University of Hawaii) * From Assistants to Automation: Leveraging NOAA’s Gemini Enterprise to Modernize Portfolio Management and R2O Transition Planning (Abigail Arnold, OAR/WPO) 15:30**NOAA Leadership Fireside Chat** Moderator: Monica Youngman (NOAA OCIO & NOAA/NWS) * Steve Thur (Assistant Administrator for Oceanic and Atmospheric Research and Acting NOAA Chief Scientist) * Heather Kilcoyne (Director, NESDIS Office of Common Services) NCAI hosted a series of virtual and hybrid public workshops focused on GenAI development for environmental science applications.

Each event focused on different aspects of GenAI. The series concluded with a 2-day hybrid event with panels and sessions spanning topics around GenAI for environmental sciences. This workshop explored collaborations on leveraging AI to align with NOAA’s strategic priorities.

Groups formed around selected use cases that could be sustained after the workshop via online collaboration platforms. The interactive, virtual workshop focused on collaboration building and active development of AI-powered applications and community standards. The workshop collaborated around two themes:Benchmarking Framework and Research-to-Operation-to-Research (R2O2R).

September 19-21, 2023, virtual. An interactive workshop to build collaboration and initiate the active development of AI-powered applications and community standards, focusing on three themes: fire weather and impacts, AI in ocean conservation, and interoperable Digital Twin Earth. September 4-9, 2022, virtual.

The third event continued the successes of previous workshops and encouraged participation by scientists, program managers, and leaders from the public, academic and private sectors who work in AI and environmental sciences. The theme was “Transforming Weather, Climate Services, and Blue Economy with Artificial Intelligence. ” September 13-17, 2021, virtual.

The workshop consisted of weekly events including: extended presentations, mini sessions, panel discussions, tutorials, and networking opportunities. This WiSM (Workshop in Slow Motion) took advantage of the virtual format and reserved time for the development of presentations and Q&A sessions. July 30, 2020 - February 25, 2021, virtual.

The workshop gathered scientists, program managers, and leaders from the public, academic and private sectors. It enabled experts involved in the development and adaptation of AI tools and applications in a number of fields to meet and exchange experiences. April 23-25, 2019, College Park, Maryland.