Learning physics-constrained subgrid-scale closures in the small-data regime for stable and accurate LES

TL;DR

This paper demonstrates that incorporating physics constraints into CNNs enables stable, accurate large-eddy simulations in small-data regimes, outperforming traditional methods and physics-agnostic CNNs.

Contribution

It introduces physics-constrained CNN methods, including data augmentation, group convolutions, and enstrophy conservation, to improve LES accuracy with limited training data.

Findings

Physics-constrained CNNs outperform traditional closures.

Data augmentation and GCNNs improve small-data performance.

Combining structural and functional constraints yields best results.

Abstract

We demonstrate how incorporating physics constraints into convolutional neural networks (CNNs) enables learning subgrid-scale (SGS) closures for stable and accurate large-eddy simulations (LES) in the small-data regime (i.e., when the availability of high-quality training data is limited). Using several setups of forced 2D turbulence as the testbeds, we examine the {\it a priori} and {\it a posteriori} performance of three methods for incorporating physics: 1) data augmentation (DA), 2) CNN with group convolutions (GCNN), and 3) loss functions that enforce a global enstrophy-transfer conservation (EnsCon). While the data-driven closures from physics-agnostic CNNs trained in the big-data regime are accurate and stable, and outperform dynamic Smagorinsky (DSMAG) closures, their performance substantially deteriorate when these CNNs are trained with 40x fewer samples (the small-data regime). We show that CNN with DA and GCNN address this issue and each produce accurate and stable data-driven closures in the small-data regime. Despite its simplicity, DA, which adds appropriately rotated samples to the training set, performs as well or in some cases even better than GCNN, which uses a sophisticated equivariance-preserving architecture. EnsCon, which combines structural modeling with aspect of functional modeling, also produces accurate and stable closures in the small-data regime. Overall, GCNN+EnCon, which combines these two physics constraints, shows the best {\it a posteriori} performance in this regime. These results illustrate the power of physics-constrained learning in the small-data regime for accurate and stable LES.