Implicit Augmentation from Distributional Symmetry in Turbulence Super-Resolution

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Balla, Julia, Bailey, Jeremiah, Backour, Ali, Hofgard, Elyssa, Jaakkola, Tommi, Smidt, Tess, McConkey, Ryley
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911175508230144
author Balla, Julia
Bailey, Jeremiah
Backour, Ali
Hofgard, Elyssa
Jaakkola, Tommi
Smidt, Tess
McConkey, Ryley
author_facet Balla, Julia
Bailey, Jeremiah
Backour, Ali
Hofgard, Elyssa
Jaakkola, Tommi
Smidt, Tess
McConkey, Ryley
contents The immense computational cost of simulating turbulence has motivated the use of machine learning approaches for super-resolving turbulent flows. A central challenge is ensuring that learned models respect physical symmetries, such as rotational equivariance. We show that standard convolutional neural networks (CNNs) can partially acquire this symmetry without explicit augmentation or specialized architectures, as turbulence itself provides implicit rotational augmentation in both time and space. Using 3D channel-flow subdomains with differing anisotropy, we find that models trained on more isotropic mid-plane data achieve lower equivariance error than those trained on boundary layer data, and that greater temporal or spatial sampling further reduces this error. We show a distinct scale-dependence of equivariance error that occurs regardless of dataset anisotropy that is consistent with Kolmogorov's local isotropy hypothesis. These results clarify when rotational symmetry must be explicitly incorporated into learning algorithms and when it can be obtained directly from turbulence, enabling more efficient and symmetry-aware super-resolution.
format Preprint
id arxiv_https___arxiv_org_abs_2509_20683
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Implicit Augmentation from Distributional Symmetry in Turbulence Super-Resolution
Balla, Julia
Bailey, Jeremiah
Backour, Ali
Hofgard, Elyssa
Jaakkola, Tommi
Smidt, Tess
McConkey, Ryley
Fluid Dynamics
Machine Learning
The immense computational cost of simulating turbulence has motivated the use of machine learning approaches for super-resolving turbulent flows. A central challenge is ensuring that learned models respect physical symmetries, such as rotational equivariance. We show that standard convolutional neural networks (CNNs) can partially acquire this symmetry without explicit augmentation or specialized architectures, as turbulence itself provides implicit rotational augmentation in both time and space. Using 3D channel-flow subdomains with differing anisotropy, we find that models trained on more isotropic mid-plane data achieve lower equivariance error than those trained on boundary layer data, and that greater temporal or spatial sampling further reduces this error. We show a distinct scale-dependence of equivariance error that occurs regardless of dataset anisotropy that is consistent with Kolmogorov's local isotropy hypothesis. These results clarify when rotational symmetry must be explicitly incorporated into learning algorithms and when it can be obtained directly from turbulence, enabling more efficient and symmetry-aware super-resolution.
title Implicit Augmentation from Distributional Symmetry in Turbulence Super-Resolution
topic Fluid Dynamics
Machine Learning
url https://arxiv.org/abs/2509.20683