Speaker
Francesco Fossella
(Telecom-Paris & University of Rome "Tor Vergata")
Description
"In chaotic systems, limited knowledge of initial conditions causes simulated dynamics to rapidly diverge from reality, severely undermining predictability and state estimation. However, integrating real-world observations, even if spatiotemporally sparse and noisy, effectively mitigates this divergence. The functional and optimal use of such imperfect data constitutes the core of Data Assimilation, which we leverage to improve state estimation in fully developed turbulence, a physical regime that is not only inherently chaotic, but characterized by complex interactions across multiple spatiotemporal scales."
Author
Francesco Fossella
(Telecom-Paris & University of Rome "Tor Vergata")
Co-authors
Prof.
Alberto Carrassi
(University of Bologna)
Prof.
Kiwon Um
(Telecom-Paris)
Prof.
Luca Biferale
(University of Rome "Tor Vergata")
Prof.
Massimo Cencini
(CNR-ISC)
Prof.
Mathieu Desbrun
(Inria & Ecole Polytehcnique)