AGU Session of Interest ML/AI and Process-based Modeling

[Natasha Stavros]

Hey all,

I wanted to make you aware of an AGU session that may be of interest to you and some of your colleagues. Please forward as you see appropriate.

Thanks!

Natasha

Abstracts Due: 4 August 2021 23:59 EDT/3:59 +1 GMT

Session ID: 123315   
Session Title: IN031. Process-based modeling and AI/ML for Predicting Global Environmental Change  
Section: Earth and Space Science Informatics  
Virtual Only Session (selected by primary convener during session submission): No  
View Session Details: https://agu.confex.com/agu/fm21/prelim.cgi/Session/123315

Session Summary:  We are experiencing catastrophic global environmental change: global warming, population growth, and the sixth mass extinction. There is a pressing need to understand why our world is changing and predict changes to adapt. Despite progress made through emerging data analytics technologies (e.g., artificial intelligence - AI and machine learning - ML), there remains a need to understand causal effects of interconnected human-natural systems. AI/ML alone cannot capture the feedbacks and processes connecting land, water, atmosphere, and biosphere. For this, we rely on process-based models (e.g., radiative transfer, hydrology, fire behavior) but they often prohibit operational uses. We invite presentations on novel applications of AI/ML and process-based models to provide near-real-time predictive capabilities. We encourage submissions to describe environmental processes and discuss the use of AI/ML for data collection, integration, inference, or computational efficiency. Methods can include: synthetic data, emulation, scale-aware, reduced order, imputation, and data fusion.