Linear-Stability-Guided Two-Stage Surrogate Model for Nonlinear ITG-TEM Turbulence

Abstract

We develop a two-stage surrogate model framework to predict the linear and nonlinear properties of ITG-TEM turbulence from GENE gyrokinetic simulations. The first stage uses a hierarchical ensemble of stability and mode classifiers, along with global and mode-specific regressors, to predict the linear growth rate and real frequency. The second stage uses these predictions as linear spectra in a Fourier neural operator (FNO)-based nonlinear surrogate model. The nonlinear model predicts saturated potential and heat flux spectra, together with zonal flow quantities. Two linear drive scalars, the maximum and integral of the linear growth rate over a low-wavenumber range, are also supplied directly to the networks predicting scalar quantities. For the architecture studied here, comparisons with models omitting either or both input paths show that linear information can improve spectral and scalar predictions under extrapolation toward stronger linear drive.

ELMO: An Uncertainty-Aware Simulation-to-Surrogate Workflow for Fast Pedestal Linear-Stability Prediction
Nami Li, X. Q. Xu, T. Osborne +5
arXiv28 Aug 2026PreprintarXiv licenceFull textPlasma physicsarXiv
Six-year data from three magnetic resonance-guided radiotherapy systems demonstrate long-term stability of linear accelerator quality assurance
Guido C. Hilgers, Simon J. Woodings, Thomas A. Foppen +4
Physics and Imaging in Radiation Oncology23 Sep 2026Peer-reviewedCC BY 4.0Full textApplied physics & instrumentationPMCEurope PMC
A Pre-trained Variational Autoencoder for Gyrokinetic Plasma Turbulence Surrogate Modeling
Minglei Yang, Marshall Nicholson, Diego Del-Castillo-Negrete +2
arXiv29 Sep 2026PreprintarXiv licenceFull textPlasma physicsarXiv
Microdroplet Generator Surrogate Enhancement and Inverse Design Through an Uncertainty-Guided Framework
David Gabriel Harispe, Matías Gerard, Pablo A. Kler
arXiv4 Nov 2026PreprintarXiv licenceFull textFluid dynamicsarXiv
Guided navigation in turbulence
Jingran Qiu, Kristian Gustavsson
arXiv31 Aug 2026PreprintarXiv licenceFull textFluid dynamicsarXiv

Cite

BibTeX RIS

BibTeX

@misc{cho2026linearstabilityguided,
  title = {{Linear-Stability-Guided Two-Stage Surrogate Model for Nonlinear ITG-TEM Turbulence}},
  author = {Y. W. Cho and C. G. Wan and Z. S. Qu and C. J. Ajay and K. P. Li and K. Lim and S. S. Kim and Y. -H. Lee and J. M. Kwon and X. Garbet},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.39381},
  eprint = {2610.39381},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.39381},
}

RIS

TY  - UNPB
TI  - Linear-Stability-Guided Two-Stage Surrogate Model for Nonlinear ITG-TEM Turbulence
AU  - Y. W. Cho
AU  - C. G. Wan
AU  - Z. S. Qu
AU  - C. J. Ajay
AU  - K. P. Li
AU  - K. Lim
AU  - S. S. Kim
AU  - Y. -H. Lee
AU  - J. M. Kwon
AU  - X. Garbet
PY  - 2026
DA  - 2026/10/29
PB  - arXiv
DO  - 10.48550/arxiv.2610.39381
UR  - https://arxiv.org/abs/2610.39381
AB  - We develop a two-stage surrogate model framework to predict the linear and nonlinear properties of ITG-TEM turbulence from GENE gyrokinetic simulations. The first stage uses a hierarchical ensemble of stability and mode classifiers, along with global and mode-specific regressors, to predict the linear growth rate and real frequency. The second stage uses these predictions as linear spectra in a Fourier neural operator (FNO)-based nonlinear surrogate model. The nonlinear model predicts saturated potential and heat flux spectra, together with zonal flow quantities. Two linear drive scalars, the maximum and integral of the linear growth rate over a low-wavenumber range, are also supplied directly to the networks predicting scalar quantities. For the architecture studied here, comparisons with models omitting either or both input paths show that linear information can improve spectral and scalar predictions under extrapolation toward stronger linear drive.
ER  -

Abstract

Related

Cite