IN027
Geospatial Foundations Models – Applications for Terrestrial Weather and Space Weather
Session ID#: 250398
Session Description:
Geospatial foundation models trained on satellite data have begun to outperform physics-based numerical simulations and other traditional techniques in certain use cases. These models have the potential to harness the power of artificial intelligence to efficiently address some of the planetâs most complex challenges across weather, land, water, and space. Recent advances demonstrate rapid forecast times for terrestrial weather and space weather forecasts measured in minutes rather than hours as compared to traditional approaches, and potentially for higher spatial and temporal resolution inferences. In this session, we invite experts and practitioners to discuss the recent results, ongoing research and developments, and future directions with the focus on but not limited to the applications in terrestrial weather forecasting for disaster preparedness, response and recovery for extreme weather events; space weather forecast of solar flares, filament eruptions, and solar winds; as well as carbon monitoring, agriculture, aquaculture and more.
Co-Sponsor(s):
Index Terms:
1922 Forecasting [INFORMATICS]
1942 Machine learning [INFORMATICS]
3399 General or miscellaneous [ATMOSPHERIC PROCESSES]
7924 Forecasting [SPACE WEATHER]
Primary Convener: David Bell, Universities Space Research Association (USRA), Research Institute for Advanced Computer Science (RIACS), Moffett Field, United States
Convener: Tsengdar J Lee, NASA Headquarters, Washington, United States
Climate Downscaling with Geospatial Foundation Models (1871182)
Weile Wang, NASA/AMES Research Center, Moffett Field, United States, Hirofumi Hashimoto, NASA Ames Research Center, Moffett Field, CA, United States, Taejin Park, NASA Ames Research Center/Bay Area Environmental Research Institute, Moffett Field, United States, Ian G Brosnan, NASA Ames Research Center, Earth Science Division, Moffett Field, CA, United States and Haonan Chen, Colorado State University and NOAA Physical Sciences Laboratory, Boulder, United States
Climate in a Bottle: A Multi-Modal Generative Foundation Model for On-Demand Kilometer-Scale Climate States (1925693)
David Hall, Noah Brenowitz, Tao Ge, Akshay Subramaniam, Peterq Manhausen, Morteza Mardani, AAyush Gupta, Arash Vahdat, Karthik Kashinath and Michael Pritchard, NVIDIA Corporation, Santa Clara, CA, United States
Learning Global Geostationary Representations with DINO-MAE: A Self-Supervised Foundation Model for Longwave IR Imagery (1968477)
Disha Shidham1, Olivia Morgan Alexander2, Tom Berg3, Jad Makki4, Ata Akbari Asanjan5, Cara Crawford6, Srija Chakraborty7, Stephen Peng4, Arun Ravindran4, Olivier Raiman8, David Potere4 and David Bell2, (1)Universities Space Research Association Moffett Field, Moffett Field, CA, United States, (2)Universities Space Research Association Moffett Field, Moffett Field, United States, (3)Boston Consulting Group (BCG) X, Los Angeles, United States, (4)Boston Consulting Group (BCG) X, Boston, United States, (5)Universities Space Research Association Moffett Field, Greenbelt, United States, (6)Boston Consulting Group (BCG) X, Washington DC, United States, (7)Universities Space Research Association Moffett Field, Washington DC, United States, (8)Universities Space Research Association Moffett Field, Paris, France
Predicting Persistent Contrail Formation with Satellite Sounding Fusion Data (1999005)
Sebastian Gutierrez-Nolasco1, Evan Cook1, David Bell1, Banavar Sridhar1, William L Smith Jr2 and David P Duda3, (1)Universities Space Research Association Moffett Field, Moffett Field, United States, (2)NASA Langley Research Ctr, Hampton, United States, (3)Analytical Mechanics Associates, Inc., Hampton, United States
