C3S2 Seminar
Clarkson Center for Complex Systems Science
Data Assimilation: From Traditional Foundations to Broader Horizons
Professor Nan Chen, University of Wisconsin
Friday, April 25, 2025
12:00 PM
Science Center 166
ABSTRACT:
In this talk, I will present how data assimilation serves as a crucial bridge between models and data across various fields. I will begin by giving an overview of data assimilation, including state estimation, parameter estimation, and forecast initialization. Beyond these traditional uses, I will show how data assimilation can broadly facilitate other areas of study and innovation. First, I will demonstrate how data assimilation and machine learning can be synergistically combined by incorporating appropriate stochastic models. Second, I will discuss how data assimilation can connect multiple imperfect models, leveraging their individual strengths to create a more accurate and cohesive integrated system. Finally, I will highlight how data assimilation can contribute to a new way of studying causal inference, which provides a unique method to identify instantaneous cause-and-effect relationships and the causal inference range.