Exact coherent structures in two-dimensional turbulence identified with convolutional autoencoders
Abstract
Convolutional autoencoders are used to deconstruct the changing dynamics of two-dimensional Kolmogorov flow as is increased from weakly chaotic flow at to a chaotic state dominated by a domain-filling vortex pair at . The highly accurate embeddings allow us to visualise the evolving structure of state space and are interpretable using `latent Fourier analysis' (Page {\em et. al.}, \emph{Phys. Rev. Fluids} \textbf{6}, 2021). Individual latent Fourier modes decode into vortical structures with a streamwise lengthscale controlled by the latent wavenumber, , with only a small number required to accurately represent the flow. Latent Fourier projections reveal a detached class of bursting events at which merge with the low-dissipation dynamics as is increased to . We use doubly- () or triply- () periodic latent Fourier modes to generate guesses for UPOs (unstable periodic orbits) associated with high-dissipation events. While the doubly-periodic UPOs are representative of the high-dissipation dynamics at , the same class of UPOs move away from the attractor at -- where the associated bursting events typically involve larger-scale () structure too. At an entirely different embedding structure is formed within the network in which no distinct representations of small-scale vortices are observed; instead the network embeds all snapshots based around a large-scale template for the condensate. We use latent Fourier projections to find an associated `large-scale' UPO which we believe to be a finite- continuation of a solution to the Euler equations.
Keywords
Cite
@article{arxiv.2309.12754,
title = {Exact coherent structures in two-dimensional turbulence identified with convolutional autoencoders},
author = {Jacob Page and Joe Holey and Michael P. Brenner and Rich R. Kerswell},
journal= {arXiv preprint arXiv:2309.12754},
year = {2024}
}