A Brief Reflection on DOE’s AI4ESP Workshop
Artificial Intelligence for Earth System Predictability (AI4ESP) workshop
Last October the U.S. Department of Energy (DOE) welcomed participants to the Artificial Intelligence for Earth System Predictability (AI4ESP) workshop, as a part of BER-ASCR planning step to create a paradigm shift in DOE’s Earth System modeling and observational networks and capabilities. The workshop is particularly focused on the co-design of artificial intelligence methodologies with physics-based models, data assimilation, explainability/trustworthiness, and new observational technologies. The AI4ESP workshop extended over six weeks (October -December 2021) with a total of 10 workshop days and 17 workshop sessions, which were identified based on common themes from 156 White Papers (640 unique authors from 112 unique institutions). All sessions, 9 Earth system predictability science topics, and 8 cross-cutting AI/ML topics discussed the grand challenges, state-of-the-science, opportunities, research priorities, and near term and 10-year goals of the AI4ESP with respect to their specific discipline. The Opening Plenary on October 25, 2021, was attended by ~ 450 participants! By November 28, 2021, there were 724 unique participants from 181 institutions. All public plenary talks and reports out from the sessions are recorded and available for viewing on AI4ESP.org. The major outcomes of the workshop are currently being synthesized by the AI4ESP Team in a comprehensive workshop report that will be used by the Office of Science to inform future priorities.
Related Articles
This article is a part of the E3SM “Floating Points” Newsletter, to read the full Newsletter check:
