Revealing the state space of turbulence using machine learning

Abstract

Despite the apparent complexity of turbulent flow, identifying a simpler description of the underlying dynamical system remains a fundamental challenge. Capturing how the turbulent flow meanders amongst unstable states (simple invariant solutions) in phase space, as envisaged by Hopf in 1948, using some efficient representation offers the best hope of doing this, despite the inherent difficulty in identifying these states. Here, we make a significant step towards this goal by demonstrating that deep convolutional autoencoders can identify low-dimensional representations of two-dimensional turbulence which are closely associated with the simple invariant solutions characterizing the turbulent attractor. To establish this, we develop latent Fourier analysis that decomposes the flow embedding into a set of orthogonal latent Fourier modes which decode into physically meaningful patterns resembling simple invariant solutions. The utility of this approach is highlighted by analysing turbulent Kolmogorov flow (flow on a 2D torus forced at large scale) at $Re=40$ where, in between intermittent bursts, the flow resides in the neighbourhood of an unstable state and is very low dimensional. Projections onto individual latent Fourier wavenumbers reveal the simple invariant solutions organising both the quiescent and bursting dynamics in a systematic way inaccessible to previous approaches.

References in corpus (8)

• Learning data driven discretizations for partial differential equations
• The Significance of Simple Invariant Solutions in Turbulent Flows
• Recurrent motions within plane Couette turbulence
• Visualizing the geometry of state space in plane Couette flow
• Deep learning to discover and predict dynamics on an inertial manifold
• Spatiotemporal dynamics in 2D Kolmogorov flow over large domains
• An adjoint-based approach for finding invariant solutions of Navier-Stokes equations
• Koopman mode expansions between simple invariant solutions

Cited by in corpus (18)

• Evolutional Deep Neural Network
• Data-Driven Reduced-Order Modeling of Spatiotemporal Chaos with Neural Ordinary Differential Equations
• Explaining the physics of transfer learning a data-driven subgrid-scale closure to a different turbulent flow
• Data-driven low-dimensional dynamic model of Kolmogorov flow
• Reconstruction of three-dimensional turbulent flow structures using surface measurements for free-surface flows based on a convolutional neural network
• Compressing fluid flows with nonlinear machine learning: mode decomposition, latent modeling, and flow control
• Data-driven state-space and Koopman operator models of coherent state dynamics on invariant manifolds
• Exact coherent structures in two-dimensional turbulence identified with convolutional autoencoders
• Stabilisation of exact coherent structures in two-dimensional turbulence using time-delayed feedback
• Statistical properties of large data sets with linear latent features
• Dynamics of a Data-Driven Low-Dimensional Model of Turbulent Minimal Pipe Flow
• Learning a Reduced Basis of Dynamical Systems using an Autoencoder
• Real-time forecasting of chaotic dynamics from sparse data and autoencoders
• Data-assisted, physics-informed propagators for recurrent flows
• Data-driven guessing and gluing of unstable periodic orbits
• Dynamical relevance of periodic orbits under increasing Reynolds number and connections to inviscid dynamics
• On nonlinear transitions, minimal seeds and exact solutions for the geodynamo
• AI Emulation of Stochastic Sudden Stratospheric Warming with Interpretable Latent Structure