English

Resolving Turbulent Magnetohydrodynamics: A Hybrid Operator-Diffusion Framework

Fluid Dynamics 2026-01-21 v2 Artificial Intelligence Machine Learning General Relativity and Quantum Cosmology Computational Physics

Abstract

We present a hybrid machine learning framework that combines Physics-Informed Neural Operators (PINOs) with score-based generative diffusion models to simulate the full spatio-temporal evolution of two-dimensional, incompressible, resistive magnetohydrodynamic (MHD) turbulence across a broad range of Reynolds numbers (Re\mathrm{Re}). The framework leverages the equation-constrained generalization capabilities of PINOs to predict coherent, low-frequency dynamics, while a conditional diffusion model stochastically corrects high-frequency residuals, enabling accurate modeling of fully developed turbulence. Trained on a comprehensive ensemble of high-fidelity simulations with Re{100,250,500,750,1000,3000,10000}\mathrm{Re} \in \{100, 250, 500, 750, 1000, 3000, 10000\}, the approach achieves state-of-the-art accuracy in regimes previously inaccessible to deterministic surrogates. At Re=1000\mathrm{Re}=1000 and 30003000, the model faithfully reconstructs the full spectral energy distributions of both velocity and magnetic fields late into the simulation, capturing non-Gaussian statistics, intermittent structures, and cross-field correlations with high fidelity. At extreme turbulence levels (Re=10000\mathrm{Re}=10000), it remains the first surrogate capable of recovering the high-wavenumber evolution of the magnetic field, preserving large-scale morphology and enabling statistically meaningful predictions.

Keywords

Cite

@article{arxiv.2507.02106,
  title  = {Resolving Turbulent Magnetohydrodynamics: A Hybrid Operator-Diffusion Framework},
  author = {Semih Kacmaz and E. A. Huerta and Roland Haas},
  journal= {arXiv preprint arXiv:2507.02106},
  year   = {2026}
}

Comments

16 pages, 6 figures, 1 table. Content synced with the published version

R2 v1 2026-07-01T03:43:56.598Z