Thursday, 1 February 2024: 2:30 PM
316 (The Baltimore Convention Center)
Handout (4.1 MB)
We present our progress towards a digital twin with an integrated ability to monitor, analyze, and predict the Earth system. Our system is centered around visualizing Earth observations, of which more than 95% are satellite-derived; the output of Earth System Models; and the various conjunctions and combinations of these elements. This is a big data problem, and the ability to breakdown, compare, contrast, and prepare data sets for machine learning is one of the major use cases for such a digital twin system. We also demonstrate the ability to spin up cloud instances and run large Earth system models to make and visualize predictions from the same browser-based pane of glass, a capability that can allow scientists to research innovations more quickly and with minimal manual intervention. This work was done as part of NOAA’s Earth Observation Digital Twin (EO-DT) prototype effort. While still in development, digital twin systems stand to improve the speed of science and help decision makers utilize the best information available in real-time.

