Exact coherent structures in two-dimensional turbulence identified with convolutional autoencoders

TL;DR

This paper employs convolutional autoencoders to analyze two-dimensional turbulence, revealing how flow structures evolve with Reynolds number and identifying unstable periodic orbits linked to high-dissipation events.

Contribution

It introduces interpretable autoencoder embeddings with latent Fourier analysis to uncover flow structures and UPOs across different turbulence regimes.

Findings

Latent Fourier modes decode vortical structures with small mode numbers.

High-Reynolds number flows are dominated by large-scale condensate structures.

Identified UPOs correspond to high-dissipation events and evolve with Reynolds number.

Abstract

Convolutional autoencoders are used to deconstruct the changing dynamics of two-dimensional Kolmogorov flow as $Re$ is increased from weakly chaotic flow at $Re=40$ to a chaotic state dominated by a domain-filling vortex pair at $Re=400$. 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, $l$, with only a small number $l \lesssim 8$ required to accurately represent the flow. Latent Fourier projections reveal a detached class of bursting events at $Re=40$ which merge with the low-dissipation dynamics as $Re$ is increased to $100$. We use doubly- ($l=2$) or triply- ($l=3$) 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 $Re=40$, the same class of UPOs move away from the attractor at $Re=100$ -- where the associated bursting events typically involve larger-scale ($l=1$) structure too. At $Re=400$ 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-$Re$ continuation of a solution to the Euler equations.