Fourth EoCoE webinar planned for the 11th of September 2018 - 10:00 (CEST).
Title: Introduction to ensemble simulations and data assimilation in the context of high performance scientific computing.
Harrie-Jan Hendricks-Franssen, Forschungszentrum Julich GmbH.
Whereas high-performance scientific computing in the 90´s of the 20th century focused mostly on forward single-physics problems, using parallel computing, the 21st century has seen a shift towards multi-physics problems, ensemble simulations and massively parallel computing. Recently the focus towards data-driven modelling has been re-enforced. In this context, this webinar will discuss different approaches for data driven modelling relevant for simulation models based on the numerical solution of partial different equations (PDE). The webinar will give an overview on different data-driven modelling techniques like (i) inverse modelling or variational data assimilation, (ii) sequential data assimilation (Ensemble Kalman Filter variants and Particle Filters), and (iii) Markov Chain Monte Carlo techniques. In addition, an outlook will be given on the potential role of machine learning for PDE-based model-data fusion techniques.
It will be operated by the GoToWebinar Platform.
Speaker: Harrie-Jan Hendricks-Franssen, Forschungszentrum Julich GmbH.
To register please follow this link.