Data-driven Spatio-temporal Prediction of High-dimensional Geophysical Turbulence using Koopman Operator Approximation
Fluid Dynamics
2019-03-05 v2 Dynamical Systems
Optimization and Control
Chaotic Dynamics
Abstract
We show the skills of a data-driven low-dimensional linear model in predicting the spatio-temporal evolution of turbulent Rayleigh-B\'enard convection. The model is based on dynamic mode decomposition with delay-embedding, which provides a data-driven finite-dimensional approximation to the system's Koopman operator. The model is built using vector-valued observables from direct numerical simulations, and can provide accurate predictions. Similar high prediction skills are found for the Kuramoto-Sivashinsky equation in the strongly-chaotic regimes.
Cite
@article{arxiv.1812.09438,
title = {Data-driven Spatio-temporal Prediction of High-dimensional Geophysical Turbulence using Koopman Operator Approximation},
author = {M. A. Khodkar and Athanasios C. Antoulas and Pedram Hassanzadeh},
journal= {arXiv preprint arXiv:1812.09438},
year = {2019}
}
Comments
Figures and texts need to be revised