Fourier Neural Operators for Rayleigh-Bénard Convection
Machine Learning
2026-07-02 v1 Fluid Dynamics
Abstract
We propose an improved Fourier Neural Operator (FNO) for modeling two-dimensional Rayleigh-B\'enard convection by predicting time increments instead of full solutions, achieving higher accuracy than a standard FNO baseline. The resulting model is compact (314k parameters, 1.26 MB) and fast (7 ms inference), while maintaining similar accuracy as demonstrated in previous benchmarks. We show that although FNOs generalize to finer meshes, accuracy remains limited by the resolution of the training data.
Keywords
Cite
@article{arxiv.2607.02088,
title = {Fourier Neural Operators for Rayleigh-Bénard Convection},
author = {Chelsea Maria John and Thibaut Lunet and Sebastian Götschel and Andreas Herten and Stefan Kesselheim and Daniel Ruprecht},
journal= {arXiv preprint arXiv:2607.02088},
year = {2026}
}
Comments
Accepted at Computational Science, ICCS 2026