Interoperable Space Weather Capabilities
About this project
This project aims to advance the study of space weather by integrating new and innovative approaches into our atmospheric models to drive upper atmospheric phenomena at varying spatial and temporal resolutions. We plan to implement data-driven and machine-learning options to eliminate the need to simulate the full lower atmosphere, allowing the upper atmosphere to be simulated at ultra-high resolution more efficiently and making these simulations more practical due to lower computational cost.
Why this work is important
Currently, it is difficult to run ultra-high-resolution upper-atmospheric forecast simulations due to the complexity of lower-atmosphere models and the associated computational costs. Simplifying the approach with data-driven models and machine learning will enable a more practical way to run these high-resolution simulations, revolutionizing space weather science.
How does this fit within the CSF
- This project fits under Better Practices
- Increases the interoperability of scientific models and physics
- Integrates the use of AI/ML
- Allows us to run ultra-high resolution simulations more efficiently
NSF NCAR Labs involved in this project
HAO
CGD
MMM
CISL
External partners
More information