Physics-Informed Neural Operator Surrogate for 2D Magnetohydrodynamic Reconnection

Abstract

Direct numerical simulation of magnetic reconnection is limited by the scale separation of resistive magnetohydrodynamics. At large Lundquist numbers S the current layer thins as S^−1/2, forcing fine grids and short time steps that make parameter scans prohibitively expensive. Deep learning neural operators offer an alternative by learning the map between function spaces rather than individual solutions, so that a single trained model returns the state for any parameter and time at inference cost. We have developed a Fourier Neural Operator (FNO) based Physics-Informed Neural Operator (PINO) surrogate for two-dimensional compressible, viscous, resistive reconnection in a wall bounded domain. We condition on the initial state, the Lundquist number, and a continuous query time. Predicting the magnetic flux function makes grad · bB=0 exact, direct time queries remove autoregressive error accumulation, and parity-aware spectral differentiation matches the wall treatment of the reference solver. Trained across 10³≤ S≤2×10⁵ with two values withheld, the surrogate recovers density, pressure, and guide field to better than 0.2%, velocity to $3.2$ to 3.9%, the flux function to 1.0%, and the spectral current density to 7.5%. It resolves the thin, narrow current sheet, Alfvénic jets, and spectral resolution beyond mode cutoff. Our surrogate reproduces the layer-averaged reconnection rate to within 3% through S=1.1×10⁴, recovering a reconnection time exponent of $0.497$ consistent with Sweet-Parker scaling. When queried zero-shot on the 2049² solver mesh, a full trajectory run is on the order of seconds on one GPU against several hours for the DNS, over two orders of magnitude faster, establishing learned operators as a practical route to surveying reconnection across parameter space.

Preconditioned Physics-Informed Neural Operator Training
Shizheng Wen, Siddhartha Mishra, Marius Zeinhofer
arXiv28 Sep 2026PreprintarXiv licenceFull textNumerical analysis & scientific computingarXiv
WINO: A Weak-Form Physics Informed Neural Operator for Hyperelasticity on Variable Domains
Bokai Zhu, Yizheng Wang, Qinghui Zhang +1
arXiv27 Aug 2026PreprintarXiv licenceFull textNumerical analysis & scientific computingarXiv
Wrong Operator or Blind Design? A Post-Training Diagnostic for Physics-Informed Coefficient Learning
Eric Fock
arXiv14 Oct 2026PreprintarXiv licenceFull textNumerical analysis & scientific computing StatisticsarXiv
Physics-informed random neural architectures on compact boundaryless multiply connected manifolds for random-operator models in machine learning and uncertainty quantification
Christian Soize
arXiv10 Oct 2026PreprintCC BY 4.0Full textStatisticsarXiv
Physics-informed random neural architectures on compact boundaryless multiply connected manifolds for random-operator models in machine learning and uncertainty quantification
Christian Soize
Computer Methods in Applied Mechanics and Engineering26 Sep 2026Peer-reviewedCC BY 4.0Full textStatistics General physicsHAL

Cite

BibTeX RIS

BibTeX

@misc{pothula2026physicsinformed,
  title = {{Physics-Informed Neural Operator Surrogate for 2D Magnetohydrodynamic Reconnection}},
  author = {Kushaal Kumar Pothula and Arunav Kumar},
  howpublished = {arXiv},
  year = {2026},
  month = sep,
  doi = {10.48550/arxiv.2609.29514},
  eprint = {2609.29514},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2609.29514},
  note = {CC BY 4.0},
}

RIS

TY  - UNPB
TI  - Physics-Informed Neural Operator Surrogate for 2D Magnetohydrodynamic Reconnection
AU  - Kushaal Kumar Pothula
AU  - Arunav Kumar
PY  - 2026
DA  - 2026/09/01
PB  - arXiv
DO  - 10.48550/arxiv.2609.29514
UR  - https://arxiv.org/abs/2609.29514
AB  - Direct numerical simulation of magnetic reconnection is limited by the scale separation of resistive magnetohydrodynamics. At large Lundquist numbers S the current layer thins as S^−1/2, forcing fine grids and short time steps that make parameter scans prohibitively expensive. Deep learning neural operators offer an alternative by learning the map between function spaces rather than individual solutions, so that a single trained model returns the state for any parameter and time at inference cost. We have developed a Fourier Neural Operator (FNO) based Physics-Informed Neural Operator (PINO) surrogate for two-dimensional compressible, viscous, resistive reconnection in a wall bounded domain. We condition on the initial state, the Lundquist number, and a continuous query time. Predicting the magnetic flux function makes grad · bB=0 exact, direct time queries remove autoregressive error accumulation, and parity-aware spectral differentiation matches the wall treatment of the reference solver. Trained across 10^3≤ S≤2×10^5 with two values withheld, the surrogate recovers density, pressure, and guide field to better than 0.2%, velocity to $3.2$ to 3.9%, the flux function to 1.0%, and the spectral current density to 7.5%. It resolves the thin, narrow current sheet, Alfvénic jets, and spectral resolution beyond mode cutoff. Our surrogate reproduces the layer-averaged reconnection rate to within 3% through S=1.1×10^4, recovering a reconnection time exponent of $0.497$ consistent with Sweet-Parker scaling. When queried zero-shot on the 2049^2 solver mesh, a full trajectory run is on the order of seconds on one GPU against several hours for the DNS, over two orders of magnitude faster, establishing learned operators as a practical route to surveying reconnection across parameter space.
ER  -

Abstract

Related

Cite